Advances in Experimental Medicine and Biology 10

The Human Testis
ADVANCES IN EXPERIMENTAL MEDICINE AND BIOLOGY

Editorial Board:
Nathan Back Chairman, Department of Biochemical Pharmacology, School of Pharmacy,
State University of New York, Buffalo, New York
N. R. Di Luzio Chairman, Department of Physiology,
Tulane University School of Medicine, New Orleans, Louisiana
Alfred Gellhorn University of Pennsylvania Medical School, Philadelphia, Pennsylvania
Bernard Halpern College de France,
Director of the Institute of Immuno·Biology, Paris, France
Ephraim Katchalski Department of Biophysics, The Weizmann Institute of Science, Rehovoth, Israel
David Kritchevsky Wistar Institute, Philadelphia, Pennsylvania
Abel Lajtha New Y,ork State Research Institute for Neurochemistry and Drug Addiction,
Ward's Island, New York
Rodolfo Paoletti Institute of Pharmacology, University of Milan, Milan, Italy
Volume 1 Volume 8
THE RETICULOENDOTHELIAL SYSTEM BRADYKININ AND RELATED KININS:
AND ATHEROSCLEROSIS Cardiovascular, Biochemical, and Neural Actions
Edited by N. R. Di Luzio Edited by F. Sicuteri, M. Rocha e Silva,
and R. Paoletti· 1967 and N. Back· 1970
Volume 2
PHARMACOLOGY OF HORMONAL Volume 9
SHOCK: Biochemical, Pharmacological,
POLYPEPTIDES AND PROTEINS
and Clinical Aspects
Edited by N. Back, L. Martini,
Edited by A. Bertelli and N. Back· 1970
and R. Paoletti· 1968
Volume 3 Volume 10
GERM·FREE BIOLOGY-Experimental THE HUMAN TESTIS
and Clinical Aspects Edited by E. Rosemherg
Edited by E. A. Mirand and N. Back· 1969 and C. A. Paulsen· 1970
Volume 4
DRUGS AFFECTING LIPID METABOLISM Volume 11
Edited by W. L. Holmes, L. A. Carlson, MUSCLE METABOLISM DURING EXERCISE
and R. Paoletti· 1969 Edited by B. Pernow and B. Saltin - 1971
Volume 5 Volume 12
LYMPHATIC TISSUE AND GERMINAL MORPHOLOGICAL AND FUNCTIONAL
CENTERS IN IMMUNE RESPONSE ASPECTS OF IMMUNITY
Edited by L. Fiore·Donati Edited by K. Lindahl.Kiessling, G. AIm, and
and M. G. Hanna, Jr.· 1969 M. G. Hanna, Jr. - 1971
Volume 6
RED CELL METABOLISM AND FUNCTION Volume 13
CHEMISTRY AND BRAIN DEVELOPMENT
Edited by George J. Brewer· 1970 Edited by R. Paoletti and A. N. Davison -1971
Volume 7
SURFACE CHEMISTRY OF Volume 14
BIOLOGICAL SYSTEMS MEMBRANE·BOUND ENZYMES
Edited by Martin Blank· 1970 Edited by G. Porcellati and F. di Jeso • 1971
The Human Testis
Proceedings of the Workshop Conference held at
Positano, Italy, April 23.25, 1970
Edited by
Eugenia Rosemberg~ M. D.
Research Director, Medical Research Institute of Worcester, Inc.
Worcester City Hospital
Worcester, Massachusetts
and
C. Alvin Paulsen, M. D.
Professor of Medicine
University of Washington School of Medicine
Seattle, Washington
A SERONO FOUNDATION SYMPOSIUM
~ PLENUM PRESS . NEW YORK-LONDON . 1970

ISBN-13: 978-1-4615-9010-1 e-ISBN-13: 978-1-4615-9008-8
DOl: 10.1 007/978-1-4615-9008-8
First Printing - September 1970

Second Printing-March 1971
Library of Congress Catalog Card Number 70·129058
SBN 306·39010·8
@1970PlenumPress,New York
Softcover reprint ofthe hardcover 1st edition 1970
A Division of Plenum Publishing Corporation
227 West 17th Street, New York, N.Y. 10011
United Kingdom edition published by Plenum Press, London
A Division of Plenum Publishing Company, Ltd.
Donington House, 30 Norfolk Street, London W.C.2, England
All rights reserved
No part of this publication may be reproduced in any form
without written permission from the publisher
PREFACE
This volume describes the proceedings of the Workshop Conference

on The Human Testis which was held at Positano, Italy, April 23-25,
1970.
The format of the book has been arranged according to topics

discussed during the Conference. Each chapter includes individual
contributions followed by discussion. Thus, this volume provides
a current critical evaluation of the subjects discussed. The Editors
assume responsibility for any shortcomings and wish to thank all par~
ticipants for their unprecedented cooperation in making this publi-
cation possible.
The Editors are indebted to Miss Barbara Martin for her able
secretarial and administrative assistance and to Mrs. Griff T. Ross
who supervised the secretarial staff during the conference. Rec-
ognition is also due to Mrs. M. Flack for editorial assistance.
The Workshop Conference could not have taken place but for the
Serono Foundation, which supported this Meeting as well as the pub-
lication of the Proceedings of this Workshop. The physical as well
as the technical arrangements were under the careful direction of
Mr. Cesare Florimonte of the Serono Foundation. The support pro-
vided by the Serono Foundation is another proof of its generosity
in stimulating scientific exchanges and a demonstration of its con-
tinuous efforts to support basic and applied studies in the field
of human reproduction.
We express our thanks to our Publisher, Plenum Press, for their

fine cooperation and for ensuring the rapid publication of this vol-
ume.
May, 1970 Eugenia Rosemberg

C. Alvin Paulsen
v
ACKNOWLEDGMENTS
Individual participants wish to thank Editors and Publishers

of journals and books for permission to reproduce previously pub-
lished figures and tables.
Individual participants desire to express their gratitude to

the Endocrine Study Section, National Institutes of Health, USPHS,
Bethesda, Maryland, U.S.A., to the National Pituitary Agency,
Baltimore, Maryland, U.S.A.; to the Department of Biological Stan-
dards, National Institute for Medical Research, Mill Hill, London,
and to the Istituto Farmacologico Serono, Rome, Italy, for generous
supplies of hormones.
Grants supporting work presented in these proceedings are ac-

knowledged by individual participants and indicated in their re-
spective contributions.
FOREWORD
During the past twenty years many meetings concerning the re-
productive process have taken place. Both structure and function
of female and male gonads have been major topics at these gatherings.
However, in only a few instances has the male gonad, particularly
the human, been the main subject of discussion. This Conference was
designed to focus attention on the human testis. A group of inves-
tigators representing various sciehtific disciplines was invited to
this Workshop to discuss this subject from a morphologic,physiologic,
biochemical, genetic and pathologic point of view. In some areas,
especially that related to embryology, basic human data are lacking,
thus relevant animal studies were also presentedo Effort was made
in these comparative studies to illuminate rather than complicate
the understanding of fundamental phenomena in human reproductive phys-
iology.
This volume contains the Proceedings of this Workshop which served

as a forum for inter exchange of ideas which have enriched our knowl-
edge of the human testis. If the topics discussed excited our inter-
est for further research and have opened new and original lines of
study, then the organization of this gathering was amply justified.
Roberto E. Mancini, M. D.
May, 1970
~i
INTRODUCTORY REMARKS
Roberto E. Mancini, M. D.
Since its creation, mankind has been concerned with problems re-
lated to reproduction. Most of the ancient religions clearly ex-
horted the people to multiply, as this was considered the main if
not the only object of marriage. Although this preoccupation seemed
from the beginning to involve both the female and male aspects of
the pair, a preferential and unbroken thread of attention was paid
to the knowledge of the ovary and female genital tract. From the
earliest writers to modern researchers, this interest in the female
sexual function has also been reflected in the continuous recommen-
dation to seek a reliable method for the control of fertility appli-
cable only to women. In consequence a considerable amount of data
has been accumulated on the biology of the ovary as compared with
the smaller quantity on the testis. However, the importance and
significant symbolism of the untouchable male gonad appeared rein-
foreed in the ancient Assyrian times, when a woman or even a physi-
cian who consciously or unconsciously caused an injury to the testis
of a man, could be punished by having their hands cut off. Yet, the
shadows of this rather superstitious concept has persisted until to-
day, where in some countries, it is more difficult to get a biopsy
of the testis than to perform a similar one on the endometrium or
even on the ovary.
It was not before the decade following 1940 that our cognizance
of the human testis, thanks to the introduction of surgical biopsies,
became a reality and so the mysterious aura which had surrounded the
male gonad began to vanish. The biopsy was rapidly incorporated by
endocrinologists and urologists as a diagnostic tool. A few years
later, studies on the normal and abnormal histology of the different
structures of the testis corroborated and amplified previous necropsy
findings, and the pathological background of more than one endocrine
or non-endocrine disease of this gland was established. A pioneer
group of well-known North American, European and Latin American au-
thors have greatly contributed to the opening of this new field of
investigation. To all of them, dead or alive, we convey on this
x INTRODUCTORY REMARKS
special occasion our feelings of sympathy and admiration. Since

then, the development of laboratory research and biochemical tech-
niques has led to further progress in important areas of testicular
biology. Thus incontestable histophysiological evidence of some
basic mechanism operating in the male gonad has been elucidated.
This is with special reference to the cellular and subcellular mor-
phology of germinal epithelium and interstitial Leydig cells, and
to the final products of both structures, sperm cells and male hor-
mones and the feedback phenomena with the hypophysis.
Apart from the advanced studies on the relationship between Leydig

cells and steroidogenesis, we are just at the beginning of our under-
standing of: a) the dynamics of the human spermatogenic process which
is replacing the old wavecycle concept of classical cytologists; b)
the sustentacular and nutritional function of Sertoli cells related
to germinal cells and spermatozoa; c) the functional significance of
the physico-chemical composition of seminiferous tubular wall struc-
tures; d) the development and fate of sex chromosomes in the germinal
cells; e) some histochemical aspects of spermatogenesis; f) the ef-
fects of follicle stimulating and luteinizing hormones and some ste-
roids on the development of germinal and Sertoli cells and intersti-
tial tissue.
Unfortunately, an extension of these approaches is restricted by

the small quantity of tissue present in biopsies. However, it is
eagerly hoped that more refined instrumentation and methodology will
prove rewarding in a more profound analysis of human testicular struc-
ture and function, as has occurred with the animal testis. Tissue
organ culture techniques, which permit an observation of the direct
effect of different agents, molecular biological aspects of cell pop-
ulations, the metabolic pathway of germinal and Sertoli cell function,
the role of tubular wall structures and Sertoli cells in the intratu-
bular diffusion of different molecules, the relationship between
Leydig cell steroids and germinal epithelium, cellular and subcellular
receptors for the gonadotropic hormones, hypothalamus-hypophyseal-
testicular interrelation, radioimmunoassays for FSH and LH, and hemo-
dynamic aspects of testicular function are some of the topics which
urgently need the attention of investigators.
This is precisely what this Workshop Conference is intended to

do. It is not an exhaustive review of what is known today about the
physiology and pathology of the human testis, but is to add a new
dimension to our knowledge of this gland by outlining new evidence
and provoking additional thoughts and inspiration for further studies.
This is the reason why other pathological disorders of the testis,
such as inflammation, congenital abnormalities, infections and tumors
have been deliberately excluded. On the other hand some other recent
approaches which are beginning to contribute to the elucidation of
disturbances in the mechanism of spermatogenesis have been added,
INTRODUCTORY REMARKS xi
i.e., immunological factors, chromosomal anomalies of germinal cells,

and other topics concer~ing the significance of seminal plasma chem-
ical composition as an indicator of the adnexal gland function. It
was also fully realized that the introduction of a subject related
to the assessment of gonadotropin therapy in the still debatable
field of male sterility, should be at present of paramount importance.
This appears more than justified when it is considered that almost
half of infertility in couples is due to the male factor.
It is our intention that all these topics will form a solid guide
for present and future fundamental studies, which in turn will provide
a better understanding of why and how testicular structure and func-
tion may become abnormal. Needless to say the benefit of this view
will lead to a rational approach in the study of the heterogenous
group of sterile patients, or what today is considered equally im-
portant, as a counterpart, to new tools for male contraception.
Finally, the aim of this Workshop Conference was not only to

have the pleasure of meeting experts on a subject of common interest,
but also to demonstrate that the integrated efforts of embryologists,
biologists, geneticists, biochemists, pathologists and endocrinol-
ogists will be of great help in substantially increasing the infor-
mation at present available on the physiology and pathology of the
human testis.
LIST OF PARTICIPANTS
Cesar Bergada, M. D.
Director of Research
Department of Endocrinology
Hospital de Ninos
Gallo 1330
Buenos Aires, Argentina
Professor Mario Hector Burgos, M. D.

Instituto de Histologia y Embriologia
Facultad de Ciencias Medicas
Universidad Nacional de Cuyo
Casilla de Correo 56
Mendoza, Argentina
Ann C. Chandley, Ph. D.

Research Scientist
Medical Research Council
Clinical and Population Cytogenetics Research Unit
Western General Hospital
Crewe Road
Edinburgh 4, Scotland
A. Kent Christensen, Ph. D.

Associate Professor of Anatomy
Stanford University
Department of Anatomy
Stanford School of Medicine
Stanford, California 94305
Professor Yves Clermont, Ph. D.

McGill University
Strathcona Medical Building
3640 University Street
Montreal 112, Quebec, Canada
xiii
xiv LIST OF PARTICIPANTS
Michel Courot
Maitre de Recherches
Institut National de la Recherche Agronomique
Lab. Physiologie de la Reproduction
37 Nouzilly - France
Joseph R. Davis, M. D., Ph. D.

Professor of Pharmacology
Loyola University Stritch School of Medicine
2160 South First Avenue
Maywood, Illinois 60153
Professor Egon R. Diczfalusy, M. Do

Director, Reproductive Endocrinology Research Unit
Swedish Medical Research Council
Karolinska Sjukhuset
Stockholm 60, Sweden
Piero Donini, M. D.
Director of Research Laboratories
Istituto Farmacologico Serono
Via Casilina 125
Rome, Italy
C. E. Ford, D. Sc., F.R.S.

Head, Cytogenetics Group
Medical Research Council Radiobiology Unit
Harwell, Didcott
Berkshire, United Kingdom
Irving I. Geschwind, Ph. D.

Professor of Animal Science
Department of Animal Science
University of California
Davis, California 95616
Carl G. Heller, M. D., Ph. D.

Director,Division of Reproductive Physiology
Pacific Northwest Research Foundation
1102 Columbia Street
Seattle, Washington 98104
Bryan Hudson, Mo D., Ph. D.

Monash University
Prince Henry's Hospital
St. Kilda Road
Melbourne
Victoria 3004, Australia
LIST OF PARTICIPANTS xv
Jan E. Jirasek, M. D., D. Sc.

Assistant Professor of Obstetrics and Gynecology
University of Minnesota Medical School
Minneapolis, Minnesota 55455
Svend G. Johnsen, Mo Do
Hormone Department
Statens Seruminstitut
Copenhagen S., Denmark
Professor Alfred Jost

Director, Laboratory of Comparative Physiology
Faculty of Sciences
9. quai Saint-Bernard
Paris 5e, 75 - France
Marian Jutisz, Ph. D.

Research Director in C.N.R.S.
Laboratoire de Physiologie Cellulaire
College de France
11 Place Marcelin Berthelot
75 - Paris V, France
Juan Carlos Lavieri, M. D.

Chief Laboratory of Andrology
Centro de Investigaciones Sobre Reproduccion
Facultad de Medicina
Paraguay 2155,10 0 Piso
Mortimer B. Lipsett, M. D.
Chief, Endocrinology Branch
National Cancer Institute
National Institutes of Health
Bethesda, Maryland 20014
Professor Bruno Lunenfeld, M. D.

Director, Institute of Endocrinology
Tel-Hashomer Government Hospital
Tel-Hashomer, Israel
John MacLeod, Ph. D.

Professor of Anatomy
Cornell University Medical College
1300 York Avenue
New York, New York 10021
xvi LIST OF PARTICIPANTS
Professor Roberto E. Mancini, Mo D.

Director
Centro de Investigaciones Sobre Reproduccion
Paraguay 2155, 10 0 Piso
Professor Thaddeus Mann, M. D., Sc. D., Ph. D., F.R.So

Professor of Physiology of Reproduction
Director, Unit of Reproductive Physiology and Biochemistry
University of Cambridge
Downing Street
Cambridge, England
Luciano Martini, M. Do
Professor of Pharmacology
Institute of Pharmacology
University of Milan
Via Vanvitelli, 32
20129 Milano, Italy
Anthony R. Means, Ph. D.

Assistant Professor
Vanderbilt University School of Medicine
Department of Obstetrics and Gynecology
Nashville, Tennessee 37203
Susumu Ohno, D.V.M., Ph. D., D.Sc.

Chairman, Department of Biology
City of Hope Medical Center
1500 East Duarte Road
Duarte, California 91010
C. Alvin Paulsen, M. D.
Chief, Division of Endocrinology
USPHS Hospital
University of Washington School of Medicine
P.Oo Box 3145
Seattle, Washington 98114
Eugenia Rosemberg, M. Do
Research Director
Medical Research Institute of Worcester, Inc.
26 Queen Street
Worcester, Massachusetts 01610
LIST OF PARTICIPANTS xvii
Griff T. Ross, M. Do, Ph. D.

Head, Endocrinology Service and
Assistant Chief, Endocrinology Branch
National Cancer Institute
National Institutes of Health
Building 10, Room 10B09
Bethesda, Maryland 20014
Professor Kenneth Savard, D. Sc.

Director, Endocrine Laboratory
University of Miami School of Medicine
Endocrine Laboratory
P.O. Box 875, Biscayne Annex
Miami, Florida 33152
Professor Carl Schirren, Mo Do

Leiter der Andrologischen Abteilung
Universitatskrankenhaus Eppendorf
2 Hamburg 20, Martinistr. 52
Germany
Alberto J. Solari, M. D.
Centro de Investigaciones en Reproduccion
Paraguay 2155
Anna Steinberger, Ph. D.

Associate Member, Division of Endocrinology and Reproduction
Research Laboratories
Albert Einstein Medical Center
York and Tabor Roads
Philadelphia, Pennsylvania 19141
Emil Steinberger, M. D.
Chairman, Division of Endocrinology and Reproduction
Research Laboratories
Albert Einstein Medical Center
York and Tabor Roads
Philadelphia, Pennsylvania 19141
Philip Troen, M. D.
Physician-in-Chief
Montefiore Hospital
3459 Fifth Avenue
Pittsburgh, Pennsylvania 15213
xviii LIST OF PARTICIPANTS
Oscar Vi1ar, Mo Do
Associate Professor
Centro de Investigaciones en Reproduccion
Facu1tad de Medicina, 10 o piso
Paraguay 2155
J.-P. Weniger
Docteur de l'Universite de Strasbourg
Maitre de Recherche au Centre National de 1a Recherche Scientifique
Universite de Strasbourg
rue de l'Universite, No. 12
67 Strasbourg, France
Emil Witschi, M. D., Ph. D.

The Population Council
Bio-Medica1 Division
The Rockefeller University
York Avenue and 66th Street
New York, New York 10021
CONTENTS
CHAPTER I
EMBRYOLOGY OF THE MALE REPRODUCTIVE TRACT
Embryology of the Testis . . • • . . . . . . . . 3

Emil Witschi
Hormonal Factors in the Development of the Male

Genital System . • • 11
Alfred Jost
The Relationship Between Differentiation of the

Testicle, Genital Ducts and
External Genitalia in Fetal and
Postnatal Life . . . • • . • . • 19
Jan E. Jirasek
Heteroplastic Gonaduct-Testes Combinations:

A Biochemical Outlook 31
J. P. Weniger
General Discussion 39
CHAPTER II
HISTOLOGY OF THE TESTIS
Dynamics of Human Spermatogenesis 47

Yves Clermont
Electron Microscopy of the Human Seminiferous Tubules 63

Oscar Vilar, C. Alvin Paulsen, and Donald J. Moore
Fine Structure of Testicular Interstitial Cells in Humans 75

A. Kent Christensen
xix
xx CONTENTS
Histology of the Human Testis from Neonatal Period to

Adolescence . • • . . . . • . . 95
Oscar Vilar
CHAPTER III
CYTOGENETICS OF SPERMATOGENESIS
Morphological Aspects of Meiosis and Their Genetical

Significance . . . . . • 115
Susumu Ohno
Ultrastructure and Histochemistry of the Nucleus During

Male Meiotic Prophase . . . . . . . . 127
Alberto J. Solari and L. L. Tres
The Cytogenetics of the Male Germ Cells and the Testis

in Mammals 139
C. E. Ford
Chromosome Studies on Testicular Cells of Subfertile Men . , 151

Ann C. Chandley
CHAPTER IV
REGULATION OF TESTICULAR FUNCTION
SECTION 1: ROLE OF THE HYPOTHALAMUS
Mechanism of Action of FSH and LH ReleaSing Factors 171

Irving I. Geschwind
Hypothalamic Control of Gonadotropin Secretion in

the Male . . . . . . . . . . 187
L. Martini
Purification and Chemistry of Gonadotropin Releasing

Factors . . . . . . • • • • • . . 207
Marian Jutisz
Further Studies on Mechanism of Action of Luteinizing

Hormone Releasing Factor Using In Vivo
and In Vitro Techniques . . . . • . . 221
Marian Jutisz, Bernard Kerdelhue and Annette Berault
CONTENTS xxi
SECTION 2: TESTICULAR-PITUITARY INTERRELATIONSHIP
Investigations Into the Feed-Back Mechanism Between

Spermatogenesis and Gonadotropin Level
in Man . . . . . . . . . . . . . . • . 231
Svend G. Johnsen
The Role of FSH, ICSH, and Endogenous Testosterone

During Testicular Suppression by
Exogenous Testosterone in Normal Men 249
Carl G. Heller, Howard C. Morse, Mike Su, and
Mavis J. Rowley
SECTION 3: GONADOTROPINS, PURIFICATION

AND MEASUREMENT
Use of Standards in Gonadotropin Assays 263

Eugenia Rosemberg
Preparations and Biological Characteristics of HMG

Preparations Used Clinically 277
Piero Donini
Plasma FSH and LH Measured by Radioimmunoassay in

Normal and Pathologic Conditions in Men 289
Griff T. Ross
SECTION 4: METABOLIC EFFECTS OF GONADOTROPINS
Early Effects of FSH upon Testicular Metabolism 301

Anthony R. Means
Radioautographic Studies of Protein Synthesis in Cells

of the Seminiferous Epithelium 315
Joseph R. Davis and Casimir F. Firlit
Studies of Spermatogenesis and Steroid Metabolism in

Cultures of Human Testicular Tissue 333
A. Steinberger, M. Ficher and E. Steinberger
xxii CONTENTS
SECTION 5: INFLUENCE OF GONADOTROPINS ON

TESTICULAR FUNCTION
Effect of Gonadotropins on the Seminiferous Tubules of

the Immature Testis • • • • . • • • 355
M. Courot
Ultrastructural and Chemical Effects of LH upon the

Seminiferous Tubule • • • • • . • • 369
Mario H. Burgos, F. L. Sacerdote, R. Vitale-Calpe,
and D. Bari
Biologic Effect of Human Pituitary Luteinizing Hormone

and Human Chorionic Gonadotropin 381
E. Rosemberg, J. F. Crigler, Jr., W. F. Jan,
G. Bulat, R. Nakano, and S. G. Lee
The Effect of Gonadotropins on the Prepubertal Testis 393

Cesar Bergada
CHAPTER V
TESTICULAR STEROIDOGENESIS
Steroid Secretion by the Human Testis 407

Mortimer B. Lipsett
Testosterone Plasma Levels in Normal and Pathological

Conditions • • • • • • • . • • 423
Bryan Hudson, H. G. Burger, D. M. de Kretser,
J. P. Coghlan, and H. P. Taft
Relation of In Vitro Metabolism of Steroids in

Human Testicular Tissue to
Histologic and Clinical Findings 439
E. Steinberger, M. Ficher, and K. D. Smith
Subcellular Structure and Synthesis of Steroids in

the Testis . • • • • • • • • • • 459
Kenneth Savard
CONTENTS xxiii
CHAPTER VI
NON-HORMONAL FACTORS INFLUENCING

SPERM PRODUCTION AND TRANSPORT
The Biochemical Characteristics of Spermatozoa and

Seminal Plasma • • • • • • • 469
Thaddeus Mann
The Significance of Deviations in Human Spenn Morphology . • 481

John MacLeod
Pharmacological Studies on the Testicular Capsule in

Relation to Sperm Transport 495
Joseph R. Davis and George A. Langford
Effect of Gonadotropins and Androgens on Fructose and

Citric Acid of Seminal Fluid . . • • 515
Juan Carlos Lavieri and Juan C. Calamera
Immunological Aspects of Male Infertility 529

Roberto E. Mancini
CHAPTER VII
GONADOTROPIN THERAPY
SECTION 1: TREATMENT OF HYPOPITUITARISM
Effects of HCG, HMG, HLH, and HGH Administration on

Testicular Function 547
C. Alvin Paulsen, Duane H. Espeland, and
Edward L. Michals
Effect of Urinary FSH and LH on the Testicular Function

in Hypogonadal Patients • • • • • 563
Roberto E. Mancini
The Effects of Urinary Gonadotropins Following

Hypophysectomy and in Hypogonadotropic
Eunuchoidism • • • • • • • • • • • • • 577
John MacLeod
xxiv CONTENTS
SECTION 2: TREATMENT OF ADULT SEMINIFEROUS

TUBULAR FAILURE
Assessment of Gonadotropin Therapy in Infertile Males 591

P. Troen, T. Yanaihara, H. Nankin, T. Tominaga,
and H. Lever
Assessment of Gonadotropin Therapy in Male Infertility 605
Carl Schirren and Joseph O. Toyosi
Assessment of Gonadotropin Therapy in Male Infertility 613

B. Lunenfeld and R. Shalkovsky-Weissenberg
General Discussion • 631
Concluding Remarks • 637

Eugenia Rosemberg and C. Alvin Paulsen
Index • . . . . • • . . . . • . . • • • 639
CHAPTER.
EMBRYOLOGY OF THE MALE REPRODUCTIVE TRACT

EMBRYOLOGY OF THE TESTIS
Emil Witschi
The Population Council, Rockefeller University,

New York 10021
Since the gonochoristic higher vertebrates evolved from

hermaphrodite ancestors, it seems justified to turn to surviving
primitive species in order to gain an understanding of genetic
and developmental problems of sex determination. About a hundred
years ago it was first realized that rudimentary hermaphrodism
still occurs in some local races of the European brown frog (Rana
temporaria). The males start out differentiating ovaries which,
usually before maturing, transform into testes. It was left to
me to find out that in a low percentage the change is still in-
complete at the adult stage. Artificial self fertilization and
outcrossing of three hermaphrodites to males and females of truly
gonochoristic race could be made. As far as I know these are the
only reported and proven cases of fully fertile tetrapod verte-
brates (1).
Gross differences between X and Y chromosomes, as seen in the

human, have not been found in any amphibian species. But in frogs
with gonochorism a first step in the evolution of sex chromosomes
is taken, showing incomplete conjugation of the fourth largest
chromosomes in meiosis and precocious separation in the first
meiotic anaphase (2,3). The comparative study of gonad develop-
ment in hermaphrodite and purely gonochoristic frogs reveals a
distinct composite of cortex and medulla. The observation that
germ cells of identical constitution become eggs in the cortex
and sperms in the medulla led to the assumption of specialized
local environments and thus, to the inductor theory of sex differ-
entiation (4). Germ cells, being capable of becoming male or fe.
male, evidently carry both female (F) and male (M) genes as do the
somatic cells of cortex and medulla. Cortical environment acti-
vates F genes; medullary environment activates M genes. Carry-
3
4 E. WITSCHI
ing this trend of thought a little further, the cytoplasms of somat-

ic inductor cells appear to produce substances which pass through
cell and nuclear membranes and activate the corresponding F or M
genes. This will start chains of further reactions resulting next
in female differentiation of germ cells and somatic elements of the
cortex and correspondingly in male differentiation in the medulla.
In hermaphrodites simultaneous differentiation may occur in both
systems. In gonochorists prevalence of one or the other is usually
determined by genetic factors. The names corticin and medullarin
have been given to the cortical and medullary inductor substances.
The inductor theory has survived over half a century having adapted
to special conditions in mammals, and was only rarely criticized
adversely (5).
As to the chemical nature of the inductive substance we know

nothing. Evidence, however, demonstrates quite clearly that even
before the establishment of indifferent gonads the preprimordia of
cortex and medulla are astonishingly active producers of attrac -
tion substances, telopherons (possibly telopherases), which guide
the migration of primordial germ cells from dorsal endoderm and
splanchnopleure to the gonadal site (6). Clever experiments by
Dubois (7) show that in chick embryos attraction is at a maximum
during the third day (sex differentiation occurs at 5~ days). At
this time the presumptive gonad tissues are even able to free and
capture germ cells that were already embedded in graft ovaries of
7~ days and in graft testes of an age up to 13 days. In some go-
nochorists cortex and medulla exhibit separate attraction, as is
clearly seen in birds with asymmetric distribution of the germ
cells (8). In Triturus the telopherons show sexual differentia -
tion, the male tending to impede germ cell migration in female
parabiontic partners (9). Should telopherons be identical with
corticin and medullarin which are produced by the same complexes
of cells? There is no sure answer to this question, though the
experiments of Dubois suggest an end to telopheron production be-
fore the stage of sex differentiation.
Since amphibians lend themselves to sex reversal and the Afri-

can frog Xenopus can readily be bred under laboratory conditions,
we have made available male as well as female breeders of either
ZZ, SW or WW genetic constitution. This makes it possible to pro-
duce all-male and all-female offspring in practically unlimited
numbers. Male and female cultures have been challenged with two
questions. Firstly, what can you do with offered hormones? Sec-
ondly, what can you do with a supply of steroid-precursors? In
these investigations I was joined by Giovanni Chieffi, Edward Dale,
Heinz Breuer and Kazuya Mikamo (10,11,12). The answers are pre-
sented in Tables 1 and 2. Estrogens and androgens are degraded in
the livers to less active substances and eventually are completely
denatured. This can be done in the early developing livers even
before sex glands have become organized (Table 1). On the other
hand, labelled precursors remain largely unchanged and no C19 and
EMBRYOLOGY OF THE MALE REPRODUCTIVE TRACT 5
Table 1. Catabolism of Sex Steroids (12)

INCUBATION WITH C19 AND C18 (4~4C) STEROIDS
Substrates Xenopus Metabolites
Dehydroepi- ~5Androstenediol
Larvae 25c
androsterone no ~4-3Ketosteroids
Androstenedione
Androsterone
Larvae 25b
Androstanediol
Testosterone to Met. 31
Androstanedione
Adult Liver
Hxdroxy Compounds
Polar Fractions
Estrone
Larvae Estriol
17 [3-Estradiol
25a to 25d ~ -Hydroxyestrone
15a -Hydroxyestrone
Met: animals in metamorphosis, Added numbers in second column
indicate standard stages of development. Morphologically recog-
nizable sex differentiation occurs at Stage 27.
Table 2. Steroidogenesis from (4. l4C) Labelled Precursors (12)

Material Substrate Metabolites Enzymes
L.25c Acetate no C19 Steroids no Desmolases
L.25c Cholesterin " "
L.25c
L.25b
Pregnenolone
17a-Hydroxy-
"
Corticoids
"
Hydroxylases
L.25d progesterone Cortisol "
Testes %S
17a-Hydroxy- Androst-
L.29/30 progesterone enedione .030 C17-C20 Desmolase:
Met.31
Met.33
" " .072
.064 17 -Hydroxyproges-
II
" terone C17 -C20 -lyase
Fr.3mo.
Frog ad.
" " .065
"
" " and
Testosterone
Ovaries 17a-Hydroxy- Androstene C17-C20 Desmolase
L.29/30 progesterone dione .029 "
Met.31 " " .026 "
Met.33 " " .035 "
Frog ad. " " and
17[3 -Estradiol
"
L: larvae, Met: animals in metamorphosis, %S: percent sub-

strate radioactivity per animal, mo: months of age, ad: adult,
added numbers indicate standard stages of development.Morphological-
ly recognizable sex differentiation occurs at Stage 27.
6 E. WITSCHI
C18 steroids are formed before or during sexual differentiation.
In short, after the supply of 17a-hydroxyprogesterone, which

is a likely natural precursor, androstenedione appears near the
end of larval life, simultaneously in female and in male cultures.
Estradiol and testosterone can be identified only after metamor-
phosis. Both increase considerably at five to seven months with
the approach of sexual maturity. Evidently these findings lend no
support to claims that estradiol and testosterone are the chemical
inductors of sexual differentiation of the gonads.
The assumption that ovary and testis do not only produce

steroid hormones but also start differentiating themselves as a
consequence of induction each by its own hormone, still has some
strong defenders (13). A review of pertinent literature shows
that the search for fetal and embryonic sex hormones in other ver-
tebrates is always based on already differentiated gonads, usually
of quite advanced stages. The most valuable series is one on male
rats presented by Noumura et ale (14). It starts with testes of
13~ day old rats. They can not be more than half a day beyond in·,
itial differentiation. After incubation with tritiated progres-
terone almost infinitesimal traces of testosterone were found on
this (stage 28) and the next day (stage 30). A total of 336 testes
yielded 0.1 m~M testosterone and 0.03 tllIJ.M androstenedione. After
another day (stage 32) conversion rates rose twentyfold and from
here until birth there was a steady but only modest increase.
Comparison with Xenopus shows that sex differentiation in relation
to anatomic body conformation (the basis of standard stages) occurs
slightly later (stage 27). Apparently, the first steroid formed
is androstenedione but conversion to testosterone follows immedi-
ately in the rat, whereas it is delayed considerably in Xenopus.
Other reports on mammals (15,16,17,18) give qualitatively similar
results but on relatively late fetal stages. Of interest is the
failure to find estrogens in ovaries even at midpregnancy (17,18).
Since the human fetus is flooded with high concentrations of pla-
cental estrogens, this may be a matter of secondary hormonal sup-
pression.
Extensive work on chick embryos by a Paris school has given

the remarkable result that female embryos of seven or more days,
if explanted in adjusted nutritional media for one to three or
more days, release estradiol and estrone in appreciable quantities
but no androgens (19). This is surprising for not only in Xenopus
but also in other amphibians (20) and in mammals (18) estrogen
synthesis very commonly follows that of androgens. The particular
pathway for chick embryos has essentially been clarified now by
Haffen and Cedard (21). Incubation with radio-labelled dehydro-
epiandrosterone (DHA) yields estrone and estradiol with female
and testosterone with male embryos. The latter also produced min-
or amounts of estradiol. The females seem to synthesize estrone
by short cut aromatisation of DHA.
Morphologic sex differentiation in the chick is recognizable

at 5\ days (stage 30). Latent determination must precede this by
another half-day (stage 28). Seven-day old embryos (stage 33)
correspond developmentally to the metamorphosing Xenopus. However,
even at this stage the sex type of gonads in freshly dissected
embryos is difficult to assess. This, together with the rapid
course of development during the first 10 days, has probably led
to certain statements on specific hormone production preceding
sexual differentiation (13). Critical consideration of all facts
leads to the realization that the chemical nature of corticin and
medullarin as also that of the telopherons and of all morphogenetic
substances produced by gonad primordia remains unsolved.
As early as 1931 (22) the existence of an antagonism between

cortex and medulla had come fully to light. First analyzed in
cases of amphibian parabiosis, evidence was soon found throughout
the vertebrate classes. Almost forgotten now are the classical
experiments of Benoit, Zawadowsky, Padoa and Donnn (ably reviewed
by the last named) on partial castration of poultry (23). The
operations simply remove the cortical repressor which in intact
hens prevents hermaphrodite development. Sini.stral ovariectomy
permits complete testicular organization and growth of the right-
side medullary rudiment. But incomplete operation, leaving in
situ the left medulla, likewise is followed by development of a
testis. Left as well as right "female testes" can fully maintain
cock plumage and male-type combs. Spermatogenesis is complete
even though the female sex chromosome arrangement is maintained.
In some experiments Mintz and Wolff (24) had peeled off the cor-
tices of ovaries of chick embryos aged 9 and 10 days and implanted
the medullae into body cavities of indifferent second-day embryos.
After 11 to 13 days well developed grafts had assumed a testicular
structure.
A spectacular example of medullary-cortical antagonism is

presented by freemartinism of cattle. In this case, however, the
medulla of the male calf is the dominant system. suppressing the
cortical development in the female twin. Contradicting the widely
accepted interpretation of the Chicago school, I have early ex-
pressed the opinion that the blood-borne principle is not the an-
drogenic steroid but a quite different type of medullary antagon-
ist (25). It gives me great satisfaction that this view is now
fully validated by recent experiments of Jost (26). They prove
the inability of testosterone to inhibit ovarian development.
Recognition of a second principle, a cortex inhibitor released by
the testes of the bull twin, is strictly necessary.
The same anti-cortex factor finds application in female to

male sex conversion in Xenopus. Testosterone administrations to
8 E. WITSCHI
larvae in no way impede ovarian development. But testis implants

almost completely inhibit ovarian differentiation of the cortex.
The medullary nodes, however, remain unaffected. Later removal
of the implanted testis is followed by compensatory medullary de-
velopment that often leads to the production of fertile males with
testes of female ZW or WW constitution (27).
This now calls for a short consideration of the mechanism by

which estrogens feminize genetic males (28). The experiment is
successful only during the third week (larvae stage 25, 2.0 - 2.5
cm total length). Later administrations are without effect. The
larvae develop like normal females. The medulla nodes do not en-
large but assume the female character. Does the estradiol during
the short application stimulate the cortex or inhibit the medulla?
We have come to assume the latter. This conclusion is supported
by the effect observed aft-er administration of the anti-androgen
Cyproterone acetate. L If applied during the third week it some-
what irregularly damages the medullar nodules of male larvae which
then become irregular hermaphrodites. It seems that here an initi-
al tissue damage is set while with estradiol the genetically favor-
ed inductor - the M activating principle - is suppressed.
In 1953 zoologists and medical embryologists were startled

by an announcement of Jost (29) that testicular steroids only stim-
ulate male secondary sex characters but do not stop growth of the
oviducts nor cause their regression. The possibility was consider-
ed that the fetal testis may produce a second, inhibiting, factor.
We shall probably hear today more about this from Dr. Jost himself.
In the meanwhile, I cannot forsake the opportunity of presenting
the amphibian version of the same situation. In Xenopus oviducts
develop in males and females immediately following metamorphosis.
Left and right ones join in a short unpaired uterine segment before
opening into the cloaca. Development continues for 7 months. Then,
while the testes start full spermatogenesis, the oviducts regress
and soon disappear. If, however, the males are castrated before the
seventh month, the oviducts persist indefinitely. Injections of
testosterone do not reduce but rather stimulate them. Their reduc-
tion can only be brought about by implanted testes. Step by step
we have come to realize that testes are not simply manufacturing
sperms and testosterone. Exercising their important functions in
development and reproduction, they maintain correlations by messen-
gers and receptors, the chemical nature of which is still largely
unknown.
I have dwelt largely on amphibian work. In a time when ge-

neticists earn the highest honors occupying themselves with bac-
teria and molds, an apology should not be expected. Embryology,
1 supplied by Schering AG Berlin, through the courtesy of President
K. H. Kimbel, New York.
sndocrinology and anatomy of the sex organs remain essentially the

same in all tetrapods in spite of complications added in amniotes
and particularly in mammals. Why then can we not change sex in
mammals as in amphibians? There is no full answer. Certainly a
major impediment is the higher evolution of the cytogenetic sex
determining mechanisms. However, the defenses are not impregnable
as shown by the possibility of full spermatogenesis in poulards
with female sets of sex chromosomes; or the permanent masculin-
ization of the mode of secretion of hypophyseal gonadotropins by
implantation of a testis in the newborn female rat. A broadened
knowledge of existing mechanisms will provide new approaches lead-
ing to the control of reproductive processes in the mammalian field.
REFERENCES
1. Witschi, E.) J.Exp. Zool., 54: 157, 1929.

2. Witschi, E., J.ind. Abst. V-;;'erb., 29: 31, 1922.
3. Witschi, E. and Opitz, J.M. In Intersexuality, ed. Overzier,
Academic Press, 1963.
4. Witschi, E., Arch. mikr. Anat., 86: 1, 1914.
5. Wolff, E., Ann.Fac. Sc. Clermont-Ferrand, 26: 17, 1965.
6. Witschi, E., Carnegie Contr. Embr., 16,: 67, 1948.
7. Dubois, R., C. R. Acad. Sci. (Paris), 260: 5108, 1965.
8. Witschi, E., Ames.J. Anat., 12: 119, 1935.
9. Witschi, E. and McCurdy, H.M., Proc. Soc. Exp. BioI. Med.,
26: 655, 1929.
10. Rao, G.S., Breuer, H. and Witschl, E., Experientia, 24: 1258,
1968.
11. Rao, G.S., Breuer, H. and Witschi, E., Gen. Compo Endocr.
ll.: 119, 1969.
12. Witschi, E. Appleton-Century-Crofts (1970, in press).
13. Wolff, E., Haffen, K. and Scheib, D., Ann. Histochim, 11:
353, 1966.
14. Noumura, T., Weisz, J. and Lloyd, C.W., Endocrinology, ~:
245, 1966.
15. Acevedo, J.F., Axelrod, L.R., Ishikawa, E. and Takaki, F.,
J. Clin. Endocr., 23: 885, 1963.
16. Lipsett, M.B. and Tullner, W.W., Endocrinology, 77: 273, 1965.
17. Bloch, E., Endocrinology, 74: 833, 1964.
18. Jungmann,.R.A. and Schweppe, J.S., J. Clin. Endocr., 28:
1599, 1968.
19. Weniger, J.P., C. R. Acad. Sci. (Paris)~: 809, 1965.
20. Ozon, R., Gen. Compo Endocr., J!: 214, 1967.
21. Haffen, K. and Cedard, L., Gen. Compo Endocr., 11: 220, 1968.
22. Witschi, E., J. Exp. Zool., 58: 113, 1931. ---
23. Domm, L.V., In Sex and Internal Secretions, ed. Allen, E.,
Williams andlWilkins, Baltimore, 227,1939.
24. Mintz, B. and Wolff, E., J. Exp. Zool., 126: 511, 1954.
25. Witschi, E. In Sex and Internal Secretions, ed. Allen, E.
Williams and1Wilkins, Baltimore, p 145, 1939.
10 E. WITSCHI
26. Jost, A., Excerpta Medica International Congress Series,

132: 74, 1966.
27. Mikamo, K. and Witschi, E., Genetics, 48: 1411, 1963.
28. Witschi, E., Arch. Anat. Micr. Morph. Exp., 39: 215, 1950.
29. Jost, A.,Recent Progr. Hormone Res., .§.: 379-,-1953.
HORMONAL FACTORS IN THE DEVELOPMENT OF THE MALE GENITAL SYSTEM
Alfred Jost
Laboratory of Comparative Physiology, Faculty of Sciences,

Paris, France
During normal development, sex differences appear, in accor-

dance with the initial genetic sex,at the level of many structures,
namely the gonads, the genital tract and external genitalia, the
mammary glands, the nervous structures in the brain controlling
hypophyseal gonadal stimulating function or mediating sexual behav-
ior and many other body secondary sex characteristics. Testicular
organogenesis precedes in time any other sex mark and the testes soon
take over the control of the other sex characteristics. In female
fetuses, ovarian organogenesis lags behind and many fundamental fem-
LnLne physiological features do not depend upon ovarian development.
There is a basic lack of symmetry in the developmental processes in
males and in females, and until the age of prepuberal changes, the
testis is the active sex differentiator (1,2).
In the frame of the hormonal theory of sex differentiation, it

has long been assumed not only that somatic sexual characteristics
develop under the control of gonadal hormones, but that the initial
sexual differentiation of the gonads themselves is imposed by hormones
or diffusible inductors produced inside the gonads. The latter part
of the concept originally arose from evidence of partial gonadal sex
reversal in freemartins in cattle and in experimental heterosexual
combinations in Urodele amphibians. It is noteworthy that in both
cases ovarian changes toward testicular organogenesis were imposed
secondarily in gonads which were already marked as presumptive
ovaries (3).
The present paper will be limited to mammals since basic differ-

ences between mammals and for instance, birds, are likely (4).
11
12 A. JOST
HORMONAL FACTORS IN GONADAL DEVELOPMENT
The undifferentiated gonadal primordium appears on the internal

side of the mesonephros, in the vicinity of the glomeruli. It is
made of cells of local origin (their derivation cannot be discussed
here) and of primordial germ cells which originate in endodermic
organs and migrate toward the gonadal anlage. The primordial germ
cells most likely obey a chemotaxis, and are attracted by a substance
released by some gonadal cells. The identity of these cells (the
presumptive sustentacular or follicular cells which throughout life
remain associated with the germ cells are obvious candidates) and
the biochemical nature of their effect remain unknown. Neither an-
drogens given to the mother animal during early pregnan,cy (for
instance high dosages of methyltestosterone given to pregnant rabbits
from day 7 on (5» nor the anti androgen cyproterone acetate given to
pregnant rats from day 10 or 11 on (up to 100 mg/day) (6,7) prevented
the germ cells from reaching the gonads in seemingly normal numbers.
Unless the androgen or the anti androgen is not transferred from
mother to fetus at early stages of pregnancy, it appears that an-
drogens are not likely to play a role in the migration of the germ
cells.
The series of organogenetic changes which produce either a tes-

tis or an ovary from the early anlage are yet poorly understood, even
on purely descriptive bases. A fortiori the mechanisms, especially
the humoral or hormonal factors involved, remain conjectural or un-
known.
There is a common consensus that testicular organogenesis pre-

cedes ovarian organogenesis, and that at first presumptive ovaries
are recognized as such mainly because they are not testes. True
ovarian organogenesis characterized by follicular formation and de-
lineation of the ovarian stroma is a late event. Some triggering
mechanism seems to divert the male gonad from this slow ovarian evo-
lution and to impose a more rapid and early testicular differentiation.
Morphological data. According to a vast literature and to pre-

liminary personal observations in calf, rabbit and rat fetuses some
of the first steps in testicular organogenesis can be summarized as
follows:
Testicular differentiation becomes first recognizable in the

gonadal anlage when a process of cellular differentiation and seg-
regation separates: a)cordlike strands of germ cells and presumptive
sustentacular cells which aggregate at a certain distance from the
superficial coelomic layer and b) the presumptive intertubular ma-
terial. In the former, the future seminiferous tubules, the cells
soon increase in size and acquire larger cytoplasmic areas whereas
the intertubular cells first remain packed and undifferentiated.
The superficial layers of the testis (future tunica albuginea) are
EMBRYOlOGY OF THE MALE REPRODUCTIVE TRACT 13
rapidly deprived of germ cells, perhaps because these cells moved

toward the seminiferous tubules. In the inner part of the gonad
near the mesonephric glomeruli, the presumptive rete testis is still
morphologically undifferentiated but it becomes increasingly distinct.
Some time elapses before true interstitial or Leydig cells appear.

In the human fetus separation of seminiferous tubules and intertubular
cells begins at the stage of 15-17 mm, whereas Leydig cells showing
their distinct characteristics under the electron microscope are just
appearing in a 29 mm fetus (8), and they display 3 ~-ol-steroid de-
hydrogenase activity in 30 mm fetuses. Similar time relations be-
tween the early appearance of the seminiferous cords and the later
cytological and histochemical differentiation of the Leydig cells hold
true in several species.
Hormonal factors in gonadal sex differentiation. The existence,

the source or the role of hormonal agents governing testicular organo-
genesis in mammals still remain unknown.
The Leydig cells of the adult type, showing the tubular organi-
zation of the agranular endoplasmic reticulum, can hardly be concerned
with the early testicular organogenesis since they differentiate at a
later stage (8). Moreover, androgens given to pregnant mammals were
never seen to impose testicular organogenesis on the primordium of
female fetuses. To quote one experiment, androgens given in large
amounts to pregnant cows did neither masculinize the ovaries of the
female fetuses, nor reduce their ovaries to the freemartin condition,
nor prevent premeiosis in the ovocytes, although they masculinized
the external genitalia at an early stage (4,9). On the other hand,
cyproterone acetate given from day 11 on to pregnant rabbits (80 mgt
day) (9) or rats (up to 100 mg/day) (6,7) did not prevent testicular
organogenesis. Similarly, in the human syndrome of "testicular fem-
inization", testicular organogenesis occurs despite the admitted in-
sensitivity of the tissues to androgens. Such observations do not
favor the hypothesis that conventional androgens play any important
role in the initial testicular organogenesis.
A converse question concerning the Leydig cells is: under which

influence do they appear in the fetal testis? This has been studied
very little so far: interstitial cells were observed to appear and
to acqui~e 3 ~-ol-dehydrogenase activity in vitro in rat testes taken
from 14.5 day-old rat fetuses, a stage at which they still cannot
be detected (10). Leydig cell differentiation thus could occur in
the absence of a hypophyseal or extragonadal influence. Nonetheless,
the further activity and thriving of the Leydig cells will depend
upon the pituitary gland.
Pituitary hormones can hardly be expected to direct the sexual

orientation of the gonadal primordium and they do not seem to be
necessary for this process. Testes and ovaries are found in infants
14 A. JOST
congenitally deprived of their anterior pituitary in Guernsey calves

suffering from hypophyseal aplasia or in mesonephric + gonadal primor-
dium of mice embryos cultivated in vitro (11). In the latter case
eventual placental influences are also ruled out.
Another important point concerns intragonadal influences on germ

cells. In the testis these cells do not enter into premeiosis until
prepuberty, whereas ovocytes display the first premeiotic changes
before the definitive ovarian structure is established, and before
they are enclosed in primordial follicles and stopped at the diplotene
phase (12). Examples of complete sex reversal of germ cells in
amphibians or birds suggest an environmental control of their out-
come. It is still unknown whether a stimulus for meiosis is acting
in the ovary or whether a meiosis-preventing influence prevails in
the testis, or both as suggested by Witschi (13). In some inverte-
brates the ovogenetic evolution is autodifferentiating whereas sper-
matogenesis is imposed by an extragonadal hormone (14). In the mam-
malian testes the germ cells are precociously confined in the semi-
niferous tubules and surrounded by sustentacular cells; it is not un-
likely that this leads to their control.
The mechanism of gonadal differentiation remains an open problem.

The leading cells in the initial changes are not yet identified.
These cells could belong either to the intertubular cells (immature
Leydig cells?), to the rete anlage or they could be the sustentacular
cells. The role of the germ cells themselves has also to be eluci-
dated. Reciprocal and successive influences can be suspected to
exist in the differentiation of the two major types of somatic tes-
ticular cells, the Sertoli cells which will impose spermatogenetic
metamorphosis upon the germ cells and the Leydig cells which will
release androgens.
HORMONAL CONTROL OF THE DEVELOPMENT OF THE GENITAL TRACT
Testicular control of the genital tract. Castration of 19-day

old rabbit fetuses in utero showed that in the absence of testes
(i.e., in normal or in castrated females and in castrated males) all
the body sexual characters become feminine; complementary experiments
reinforced the conclusion that the basic developmental trend of the
gonadless body is essentially fe~inine, and that the testes impose
masculinity and repress femininity (1, 15, 16,2).
The fetal testes cause the retrogression of the Mullerian ducts

(presumptive uteri and tubes which persist in the absence of the
testicular influence), stimulate the development of the epididymes,
vasa deferentia and seminal vesicles from the wolff ian ducts (these
ducts retrogress in the absence of the testes) and impose masculinity
on those parts of the genital tract which derive from a common un-
differentiated primordium (derivatives of the urogenital sinus, and
external genitalia which become feminine if not masculinized by the
fetal testes).
These results were confirmed in vivo on mice (17) and in vitro;

for instance the Mullerian ducts persist in parts of the genital
tract of male fetuses cultivated in vitro unless they are exposed to
the action of a fetal testis in mouse or rat experiments (18).
Androgens such as testosterone can replace fetal testes in

their masculinizing effects but fail to inhibit the Mullerian ducts
(1,16,4). It was surmised that the fetal testis produces two kinds
of morphogenetic substances: a Mullerian duct inhibitor, which still
has to be uncovered and a masculinizing factor probably similar to
androgens. Early testicular synthesis of androgens has been amply
demonstrated in rabbit (19) and rat (20) fetuses as well as in
other species. The concept of a dual testicular control fits most
of the c1ini~a1 cases of genital abnormalities (2). It also permits
one to interpret the prominent effects of the antiandrogen, cyproterone
acetate, in rabbit fetuses (21,9). Male fetuses of treated does
develop no masculine character probably because cyproterone acetate
opposed the fetal androgens. In the meantime their Mullerian ducts
retrogress, presumably because the antiandrogen does not oppose the
testicular Mullerian inhibitor.
In male rats the effect of cyproterone acetate (G.A.) is not as

clear since it does not impair the development of the wo1ffian ducts
and seminal vesicles (6). This lack of effect might be due either
to lack of placental transfer of G.A. to the fetus before day 18 or
to the inability of G.A. to antagonize the effects of the hormone
produced at that stage by the fetal testis (6). In order to further
analyze this problem, experiments were carried out in another way
(7). Pregnant rats of the Sherman strain were given testosterone
propionate (T.P.) alone (2.5 mg/day) or T.P. (2.5 or 10 mg/day) plus
G.A. (25 mg/day)and the effects on the female fetuses were studied.
In seven female fetuses who received only testosterone, Wo1ffian
ducts and seminal vesicles which were more or less perfect were
present. In twelve females who received testosterone (2.5 or 10 mg/
day) plus G.A., both the wo1ffian ducts (excepting inconstant ante-
rior remnants) and the seminal vesicles were absent. Thus at the
level of the wo1ffian derivatives the antiandrogen antagonized
high doses of testosterone in females, but it fails to ant~~onize
the secretion of the fetal testis (even if the mother is given 100
mg G.A./day). This difference still has to be explained; it could
suggest that the testis of the fetal rat produces on day 18 some
hormone which is not antagonized by cyproterone acetate.
Pituitary control of testicular function. Hypophysioprivia

affects testicular activity on the genital tract to varying degrees
in different species (22).
In rabbits, hypophysectomy by decapitation carried out before the

16 A. JOST
onset of body sex differentiation (on day 19), does not completely
stop testicular activity since the anterior parts of the sex ducts
are largely normal, but it introduces a definite testicular insuf-
ficiency which is conspicuous at the level of the prostate, external
genitalia and interstitial cells. Restoration of development is
achieved in those fetuses given gonadotropins (23).
In mouse and rat fetuses, absence of the pituitary gland does

not prevent masculine organogenesis, nor the multiplication of the
germ cells (24), but the synthesis of testosterone has been reported
to be reduced at term in the testes of those fetuses who were de-
capitated on day 16.5 (20). It would appear that during the period
covering the major sexual organogenetic processes (days 16 to 19 in
rats), and even in the absence of the pituitary, the fetal testes
produce enough of their hormone to ensure masculine development.
In human infants congenitally deprived of their pituitary gland,

masculinization of the genital tract takes place, but in some in-
stances a small penis or small testicles have been recorded. It is
uncertain whether during the period of organogenesis of the genital
tract, the human fetal testis needs gonadotropic stimulation or if
it is stimulated by the chorionic gonadotropins of the mother which
are especially high at that stage of pregnancy. During the second
half of pregnancy the fetal testis undoubtedly needs pituitary
stimulation for normal functioning (details in 22).
Role of the hypothalamus in hypophyseal gonadal stimulating

activity. The fetal hypothalamo-hypophyseal relations have received
little attention so far. Experiments performed on fetuses deprived
of their hypothalamus by a surgical encephalectomy which leaves the
pituitary gland in situ (22,25,26) could be compared with observations
made on human anencephalic infants whose anterior pituitary suffers
from disturbed or absent hypothalamic connections. In the rat fetus
the pituitary corticostimulating activity is more strictly dependent
upon the hypothalamus than is the thyrotropic activity. The same
situation seems to prevail in the human anencephalic fetuses (26).
In male anencephalic fetuses masculinization is complete. How-

ever, hypoplastic testes, reduced number of interstitial cells (27)
and inconstant hypoplasia of the penis and scrotum (28) have been
observed. The condition of the male genital tract in these infants
thus resembles that found in infants without pituitary glands; in
both cases sexual differentiation of the genital tract which occurs
before the 9th or 10th week is normal. This may be due to testicular
stimulation by chorionic gonadotropin. Thus in hypopituitarism and
anencephalia, testicular deficiency occurs at a later developmental
stage. In the case of the anencephalic fetuses the anterior pitu-
itary might suffer from a lack of hypothalamic influence during the
second part of pregnancy (29).
Experiments pertaining to this problem were made on the rabbit

since in that species hypophysioprivia produces testicular deficiency
(23). Preliminary results (25) will be summarized. Rabbit fetuses
were encephalectomized on days 17 to 19 and studied on day 28. Despite
the absence of the hypothalamus the pituitary gland differentiated
and the genital tract developed almost normally in nine of these fe-
tuses, whereas it was abnormal in three fetuses of the same strain
which were decapitated on day 19. The gonadal stimulating activity
of the pituitary is not depressed even if the hypothalamus is re-
moved as early as day 17 or 18.
In contradistinction to these experiments the genital tract was

of the decapitate type in two spontaneously abnormal male fetuses
(anencephaly and otocephaly) collected in another strain of rabbits
in which the pituitary gland was present but deprived of hypothalamic
connections. Furthermore, the pituitary gland was strikingly con-
gested, and rather small in one of them. It still is difficult to
assess whether the difference between the two groups of fetuses can
be accounted for by species differences in the onset of functional
hypothalamo-hypophyseal relationships or by other factors. These
points need further research.
CONCLUSION
We still do not know the hormonal or inductive factors which

impose early testicular organogenesis upon the gonadal prjmordium
and appropriate function of the cells within the seminiferous tu-
bules and intertubular spaces. It is obvious that the differentiated
testis soon becomes the major source of androgens. But the differ-
entiating testis also releases substances the effect of which can-
not be duplicated by conventional androgens. The fetal testis pro-
vokes the retrogression of the Mullerian ducts; it inhibits ovarian
growth and differentiation when fetal testes and ovaries are grown
together either in vitro (30) or in vivo (31). In heterosexual twin
calf fetuses, the testis is most likely responsible for the free-
martin effect. In the latter case ovarian and Mullerian inhibition
begins almost simultaneously. From a chronological point of view
it is not inconceivable that both effects might be produced by the
same testicular factor (9). Such problems still await solution and
it certainly would not be wise to restrict research in fetal testic-
cular endocrinology to testosterone or akin androgens of the adult
type.
REFERENCES
1. Jost, A., Arch. Anat. Micr. Morph. Exp. 36: 271, 1947.
2. Jost, A., In Hermaphroditism, Genital An~alies and Related
Endocrine Disorders, ed. Jones, H. W., and Scott, W. W., Williams
and Wilkins, Baltimore, 2nd ed. in press, 1969.
18 AJO~
3. Jost, A., Phil. Trans. Roy. Soc.,in press, 1970.

4. Jost, A., In Organogenesis, eds. De Haan, R. L., and Ursprung,
H., Holt, Rinehart and Winston, New York p 611, 1965.
5. Jost, A., Arch. Anat. Micr. Morph. Exp. 36: 242, 1947.
6. Jost, A., Research on Steroids, }: 207, 1967.
7. Jost, A., Unpublished data.
8. Pe11iniemi, L. J., and Niemi., M., Z. Ze11forsch., 22: 507, 1969.
9. Jost, A., Excerpta Medica Foundation International Congress
Series No. 132, 74, 1966.
10. Picon, R., Arch. Anat. Micr. Morph. Exp., 1i: 281, 1967.
11. Wolff, Et., C. R. Acad. Sci. (Paris), 234: 1712, 1952. _
12. Ohno, S., and Smith, J. B., Cytogenetics, 1: ~, 1964.
l3. Witschi, E., In The Biochemistry of Animal Development, ed.
Weber, R., Vol. 2, Academic Press, New York, p 193, 1967.
14. Streiff, W., Ann. Endocr. (Paris), ~: 641, 1967.
15. Jost, A., Gynec. Obstet. (Paris), 49: 44, 1950.
16. Jost, A., Recent Progr. Hormone Res.,~: 379, 1953.
17. Raynaud, A., and Fri11ey, M., Ann. Endocr. (Paris), 8: 400, 1947.
18. Picon, R., Arch. Anat. Micr. Morph. Exp., ~: 1, 1969.
19. Lipsett, M. B., and Tu11ner, W. W., Endocrinology, 12: 273,
1965.
20. Noumura, T., Weisz, J., and Lloyd, C. W., Endocrinology, 78: 245,
1966.
21. E1ger, W., Arch. Anat. Micr. Morph. Exp. 55: 657, 1966.
22. Jost, A., In The Pituitary Gland, eds. Ha;ris, G. W•• and
Donovan, B. T., vol. 2, Butterworths, London, p 299, 1966.
23. Jost, A., Arch. Anat. Micr. Morph. Exp. ~: 247, 1951.
24. Creasy, R. K., and Jost, A., Arch. Anat. Micr. Morph. Exp. ~:
561, 1966.
25. Jost, A., Co110que Neuroendocrino10gie. Paris: CNRS, in press,
1969.
26. Jost, A., and Ge10so, A., C. R. Acad. Sci. (Paris), ~: 625,
1967.
27. Zondek, L. H., and Zondek, T., Bio1. Neonat., 8: 329, 1965.
28. Bearn, J. G., Lancet, 2: 464, 1959. -
29. Bearn, J. G., Excerpta-Medica Foundation International Congress
Series No. 157, 21, 1968.
30. Holyoke, E. A., and Beber, B. A., Science, 128: 1082, 1958.
31. MacIntyre, M. N., Anat. Rec., 124: 27, 1956:--
THE RELATIONSHIP BETWEEN DIFFERENTIATION OF THE TESTICLE, GENITAL
DUCTS AND EXTERNAL GENITALIA IN FETAL AND POSTNATAL LIFE
Jan E. Jirasek
University of Minnesota Medical School, Minneapolis,
Minnesota 55455
The purpose of this study is to elucidate the relationship be-

tween testicular differentiation, regression of the paramesonephric
(Mullerian) duct system and masculinization of the external genita-
lia in normal human embryos and fetuses and in some male pseudoher-
maphrodites.
The results described are based on a study of 300 human embryos

and fetuses from therapeutic abortions and from 35 male pseudoher-
maphrodites in which laparotomy was performed and whose testicles
were subjected to histological and histochemical study.
1. RELATIONSHIP BETWEEN THE PRIMORDIAL GERM-CELLS (GONOCYTES) AND

THE MESONEPHRIC RIDGES
Primordial germ-cells are histochemically recognized by high

alkaline phosphatase activity and a considerable quantity of glyco-
gen (1-5). The early entoderm, however, has the same histochemical
properties and does not allow identification of germ-cells until the
embryos show development of the chordomesodermal process (embryos of
about 0.5 rom, 15 to 17 days SH":VII). In embryos ·showing differentia-
tion of early somites (1.5-5 rom., 17-30 days, SHVIII-XII), the alka-
line phosphatase and glycogen disappear from the entodermal cells of
the gut, but remain high in primordial germ-cells. These cells be-
gin to migrate from the gut-entoderm into the adjacent mesenchyme in
embryos of about 1-2 rom (19 to 20 days, SH IX). Some migrating pri-
mordial germ-cells with long cytoplasmic processes are also present
in the mesenchyme of the mesentery. Cytoplasmic processes suggest
*SH refers to Streeter's Horizons of differentiation of the human

embryo.
19
20 J. E. JIRASEK
the possible ameboid movement of these cells. These findings are

in full agreement with the classical concept of Witschi (6).
Primordial germ-cells are regarded as a basic embryonal cell-

line (7) and, to date, there is not a single human embryo known
showing absence of these cells. Germ-cells were found in embryos
of XO or XXY-chromosomal constitution (8) and in embryos showing
severe malformations, such as posterior duplication (9). In embryos
5 to 8 rom (32 to 35 days, SH XIV), the primordial germ-cells begin
to accumulate in the area of the future genital ridge in the medio-
ventral part of the mesonephric ridge.
Mesonephric ridges in embryos 5 to 8 rom (32 to 35 days, SH XIV)

are longitudinal folds on each side of the mesentery related to the
development of the mesonephros. The differentiation of the mesone-
phric (Wolffian) duct has been described (10). Wolffian duct pri-
mordia originate from ectodermal material which invaginates from sur-
face ectoderm lateral to somites 8 to 13 as glycogen-rich ectodermal
buds which join the pronephric blastema. The epithelium of prone-
phric and mesonephric ducts is glycogen-rich and the pronephric and
mesonephric blastema is glycogen-free (11). The mesenchyme of the
mesonephric ridge originates from somites. The lateral surface of
the mesonephric ridge is covered by a single-layered epithelium with
a distinct basement membrane (mesoderm). The medial surface is cov-
ered by a layer of cells which are not separated from the underlying
mesenchyme (mesoblast). There is an affinity between the mesoblastic
cells of the genital ridge and the primordial germ-cells. A positive
chemotropism between the germ-cells and mesoblastic cells has been
proposed (12). However, if this chemotropism exists, its chemical
nature is unknown.
2. THE INDIFFERENT GONAD (GENITAL RIDGE)
The proliferation of mesoblasts on the medioventral side of the

mesonephric ridge gives rise to the genital ridge (embryos 7 to 12
rom, 37 to 42 days, SH XVI). The genital ridge contains mesoblastic
cells of coelomic origin, mesenchymal cells penetrating into the
genital ridge from the mesonephric mesenchyme, and primordial germ-
cells (Fig. 1).
3. DEVELOPMENT OF PARAMESONEPHRIC (MULLERIAN) DUCTS AND THEIR

RELATIONSHIP TO MESONEPHRIC (WOLFFIAN) DUCTS
In embryos 14 to 19 rom (44 to 48 days, SHoXVIII), cylindrical

mesodermal epithelium on the lateral side of the mesonephric ridge
invaginates laterally to the cranial part of the genital ridge and
approaches the epithelium of the Wolffian duct giving rise to the
primordium of the Mullerian duct. The cells of the Mullerian duct
grow laterally along the Wolffian duct as a solid cellular cord which
gradually becomes luminized and separated from the Wolffian duct by
EMBRYOlOGY OF THE MALE REPRODUCTIVE TRACT 21
Fig. 1. Indifferent gonad in a 9 mm embryo. The basement membrane

underneath the mesodermal epithelium ends lateral to the
genital blastema (arrow). Primordial germ cells are rich
in glycogen. Colloidal iron - PAS stain. 150x.
mesenchymal connective tissue. Caudal to the mesonephric ridges,

the Mullerian ducts cross the Wolffian ducts and approach each other
near the midline. Parallel-lying Wolffian and Mullerian ducts con-
tact the dorsal wall of the entodermal urogenital sinus. The Wolf-
fian ducts have previously penetrated into the urogenital sinus. The
Mullerian ducts contact only the entodermal epithelium between the
Wolffian duct openings and this area becomes the Mullerian tubercle.
Sex-chromosomal constitution is determined at fertilization and

sex-chromatin is observed in embryonal tissues beginning the 16th
day of development (13,14). In the embryos studied, no sexual dif-
ferences were observed in the migration of primordial germ-cells,
in the formation of the Wolffian and Mullerian ducts nor in the for-
mation of indifferent gonads.
4. CHANGES OF THE GERM-CELLS AND EARLY DEVELOPMENT OF THE TESTICLE
The transformation of the indifferent genital ridge into the

testicle is observed in embryos 15 to 19 mm (44 to 48 days, SH
XVIII). At this time, the primordial germ-cells temporarily lose
their alkaline phosphatase, glycogen and round shape and become in-
corporated into the testicular cords which differentiate from the
blastema of the indifferent genital ridge (Fig. 2). Every testicu-
lar cord has a distinct basement membrane. These changes in pri-
mordial germ-cells were observed only in normal chromatin-negative
embryos. Testicular cords differentiate in situ. No ingrowth of
22 J. E. JIRASEK
Fig. 2. Differentiation of the testicular cords in a 15 mm embryo.

Glycogen-rich germ cells are elongated (arrows) and are
attached to the basement membranes of the testicular cords.
Colloidal iron-PAS stain, 300x.
cords from the surface epithelium was observed. Four cellular types
may be distinguished in the testicular primordium at this time ••
Germ-cells and primitive epithelial cells form the testicular cords.
Mesenchymal cells of somitic origin fill the interstitial spaces
between the testicular cords adjacent to the mesonephros (primordium
of the rete testis) and are mixed with mesenchymal cells of meso-
blastic (coelomic) origin in the interstitial spaces between the
testicular cords. It is believed that the transformation of the
genital blasteme into testicular cords takes place only in the pre-
sence of germ-cells containing an active XY-chromosomal complex (XY,
XYY, XXY, XXXY, XXXXY cell lines).
5. DIFFERENTIATION OF THE TESTICULAR CONNECTIVE TISSUE AND

PRIMITIVE TUNICA ALBUGINEA AND REGRESSION OF THE MULLERIAN DUCTS
In embryos 18 to 32 mm (48 to 60 days, SH XIX-XXIII) the reti-

cular connective tissue differentiates between the sex-cords and
beneath the surface epithelium. Connective tissue forms a lamina
propria of each cord of rete testis and each testicular cord and,
finally, a primitive tunica albuginea (embryos 28 to 32 mm, SH
XXIII). The development of primitive tunica albuginea ends the
period of transformation of the gonadal blastema into the testicu-
lar cord. No epithelial cells of the peripheral portion of each
cord are slightly PAS-positive and may be called primitive Sertoli
cells. The epithelial cells of rete cords are PAS-negative and
represent primitive cells lining the rete testis. Strongly PAS-

positive, glycogen and alkaline phosphatase-rich germ-cells (sper-
matogonia) are present among the epithelial cells in both portions
of the testicular cords. At the same time as the connective tissue
of the tunica albuginea differentiates, the mesenchyme around the
Mullerian ducts gives rise to reticular connective tissue. Regres-
sion of the Mullerian ducts is first observed at the crossing of
the Mullerian ducts and caudal testicular ligament in the area
where the Mullerian ducts are nearest to the caudal pole of the
testicles (embryos 30 to 35 mm, approximately 60 days). Narrowing
of the lumen of the Mullerian duct follows the differentiation of
the testicular reticular connective tissue. Release of lysosomal
acid hydrolase-acid phosphatase and non-specific esterase is ob-
served in the degenerating Mullerian epithelium. In male fetuses,
45 to 55 mm, as Wolffian ducts grow, the Mullerian ducts fragment
and gradually disappear, except at both ends (appendix testis, pro~
static utricle). The histogenesis of testicular tunica albuginea
and the regression of the Mullerian ducts are closely related. The
factors responsible for the regression of the Mullerian ducts as
well as for differentiation of the testicular connective tissue, is
unknown. The regression of the Mullerian duct is related to the
normal development of the ipsilateral testicle. One normal testi-
cle is unable to produce regression of the contralateral Mullerian
duct.
6. DIFFERENTIATION OF THE LEYDIG CELLS AND MASCULINIZATION OF

THE EXTERNAL GENITALIA
Fetal Leydig cells are epithelioid with acid phosphatase and

PAS-positive glycoprotein granules in the paranuclear zone. They
contain lipids, high activity of glucose-6-phosphate dehydrogenase
and 3 ~-hydroxysteroid dehydrogenase (Fig.3). Cells with these
characteristics may be found initially in fetuses 32 to 33 mm in
the connective tissue between the seminiferous cords under the tu-
nica albuginea (15,16). They originate from the fibroblast-like
cells of the testicular interstitium. The differentiation and num-
ber of the fetal Leydig cells seem to be under the influence of HCG
(15). Incubated in vitro, only testicles containing Leydig cells
are able to produce androgens from dehydroepiandrosterone or proges-
terone. Testicles from fetuses 8 to 9 weeks old are able to con-
vert dehydroepiandrosterone to androstenedione, but not testoster-
one. Testosterone can be synthesized in vitro by testicular tissue
from fetuses 15 to 24 weeks old (17,18,19). Fetal testicles able
to produce androgens do not show any maturation of Sertoli cells.
The masculinization of external genitalia begins in fetuses

35 to 40 mm (65 to 70 days old) with the lengthening of the anogen-
ital distance. Shortly after follows the fusion of the labioscro-
tal swellings developing the scrotum and closing the scrotal rhaphe.
Later, in fetuses 70 to 75 days old, this fusion involves the rims
24 J. E. JIRASEK
Fig. 3. 3 ~-hydroxysteroid dehydrogenase (substrate dehydroepi-

androsterone) in the Leydig cells of a 45 mm fetus, 300x.
of the urethral groove. As the penile rhaphe closes, the phallic

part of the urogenital sinus gives rise to the cavernous urethra
(20). If this fusion does not take place, the external genitalia
may feminize or may remain ambiguous, similar to the indifferent
stage. If the formation of the scrotal and/or penile rhaphe is
incomplete, the external genitalia are hypospadic. Masculiniza-
tion of external genitalia in normal males depends upon the produc-
tion of androgens by the testicular Leydig cells. One normal tes-
ticle is able to produce complete masculinization of external geni-
talia.
Female and malformed external genitalia in patients showing

androgen-insensitivity suggest that, in addition to androgens, sen-
sitivity of the target tissues forming labioscrotal swellings and
a genital tubercle is necessary for normal masculinization of the
external genitalia.
7. TESTICULAR DYSGENESIS AND PERSISTENCE OF THE MULLERIAN DUCT IN

MALE PSEUDOHERMAPHRODITES
In some male pseudohermaphrodites, a special type of a congen-

ital testicular maldevelopment, testicular dysgenesis, may be recog-
nized (Fig. 4). The following criteria are required for classify-
ing a testicle as dysgenic: 1. Thin insufficiently developed tu-
Fig. 4. Testicular dysgenesis. Tunica albuginea is thin and con-

tains cellular connective tissue. Seminiferous tubules
underneath are very narrow and degenerating. The other
seminiferous tubules are luminized and lined with mature
Sertoli cells, spermatogonia and some spermatocytes.
Leydig cells are present. Nineteen year old androgen
sensitive male pseudohermaphrodite with uterus, bilateral
testicular dysgenesis and a normal XY-karyotype. Hema-
toxylin-eosin, 150x.
nica albuginea; 2. Degenerating, narrow seminiferous tubules un-

derneath the tunica albuginea; 3. Areas of cellular connective
tissue resembling ovarian stroma (not always present).
Testicular dysgenesis must be distinguished from gonadal dys-

genesis (streak gonads without any germinal elements), from tubular
testicular dysgenesis (21) (eosinophilic, enlarged, immature Serto-
Ii cells in some seminiferous tubules of cryptorchic testes), and
from tubular, testicular, sclerotizing degeneration found in chro-
matin-positive males.
Testicular dysgenesis is associated with persistence of the

ipsilateral Mullerian duct and is congenital. Before puberty, the
dysgenic testicles do not contain Leydig cells. In post-puberal
patients, Leydig cells are present and usually hyperplastic and
26 J. E. JIRASEK
hypertrophic. They contain 3 ~-hydroxysteroid dehydrogenase, a

high activity of glucose-6-phosphate dehydrogenase and cytoplasmic
lipids. The seminiferous tubules are luminized and, in postpuberal
patients, lined with mature Sertoli cells. Some are lined with
Sertoli cells only. In others, numerous spermatogonia are present.
The spermatogonia show a very high activity of alkaline phosphatase.
Hyalinization of the basement membranes of the seminiferous tubules
is a regular finding. Bilateral and unilateral forms of testicular
dysgenesis may be recognized.
In cases showing unilateral testicular dysgenesis, a streak

gonad or well-developed cryptorchid, or normal, testicle may be pre-
sent contralaterally. Unilateral testicular dysgenesis with a streak
gonad contralaterally is known under the following names: mixed go-
nadal dysgenesis; unilateral testicular differentiation or asymme-
trical differentiation of gonads (22,23,24).
Patients with unilateral testicular dysgenesis and streak gonad

contralaterally and patients with bilateral testicular dysgenesis are
phenotypically similar. They show varying degrees of clitoral hyper-
trophy; they have a vagina and uterus, and they are androgen sensi-
tive. After puberty, their breasts are not developed, although some-
times a slight gynecomastia is present. Cytogenetically, in testi-
cular dysgenesis, XO/XY mosaicism is usual. However, in some pa-
tients, only a normal male XY-karyotype has been found. In other
cases chromosomal mosaicisms showing, in addition to XO and XY cells,
other cell lines such as XXY and XYY-cells have been observed. Pa-
tients with unilateral testicular dysgenesis and with a normal or
cryptorchid testicle contralaterally are phenotypically males and
described as "males with a uterus" (25,26,27). Their external geni-
talia are male or hypospacid. Fertility may be preserved in some.
The uterus is present on the same side as the dysgenic testicle and
usually lies in the groin inguinal hernia of the uterus or within
the scrotum in a hernial sac.
Dysgerminomas, embryonal dysgerminomas and gonadoblastomas aris-

ing from dysgenic testicles are very common. They were present in 7
of our 11 cases of testicular dysgenesis. Embryonal dysgerminomas
and gonadoblastomas were observed in three siblings, all showing ty-
pical testicular dysgenesis (28).
8. TESTICULAR DIFFERENTIATION IN PATIENTS WITH ANDROGEN

INSENSITIVITY
In male pseudohermaphrodites showing androgen insensitivity,

some specific histological changes may be recognized. Three clinical
syndromes are related to androgen-insensitivity: Complete testicular
feminization, incomplete testicular feminization and androgen insen-
sitive form of Reifenstein's syndrome. They differ phenotypically,
primarily in the formation of the external genitalia. In complete
testicular feminization, the external genitalia are malformed and

ambiguous, usually showing a hypertrophied clitoris. In Reifen
stein's syndrome the external genitalia are hypospadic. No uterus
is found. Female breasts are present in all postpuberal patients
insensitive to androgens. Absence of 5 a-androgen reductase, which
coverts testosterone to dihydrotestosterone, was observed in skin
and sexual hair follicles and is regarded as being related to andro-
gen-insensitivity (29,30). Testicular histology in androgen-insen-
sitive prepuberal patients is normal (16). After puberty, charac-
teristics are testicular tubular adenomas with hypertrophic and hy-
perplastic Leydig cells. The seminiferous tubules of the adenomas
are non-luminized and lined with immature Sertoli cells. The inter-
stitial spaces are either completely filled with hyperplastic Ley-
dig cells or contain only dense collagenous tissue. The discrepancy
between the differentiation of Sertoli cells, which are immature,
and Leydig cells, which show hypertrophy and hyperplasia, is patho-
gnomonic for patients with testicular tissue insensitive to andro-
gens (Fig. 5). In complete testicular feminization, Sertoli cells
do nat show maturation. Histochemically, the hypertrophied and hy-
perplastic Leydig cells in androgen-insensitive patients are similar
to the normal ones and contain 3 ~-hydroxysteroid dehydrogenase (16,
31). Sertoli cells do not contain any activity of this enzyme. Al-
kaline phosphatase, suggesting growth activity of spermatogonia, was
demonstrated in only one of 12 cases studied histochemically. Non-
specific degenerative changes, such as hyalinization of the basal
membranes of some tubules and degeneration of the immature Sertoli
cells, are common and develop at puberty or later. Abnormalities
of the tunica albuginea were not observed. Testicular tubular ade-
nomas are benign. Malignant tumor"s originating from non-luminized
tubules lined with immature Sertoli cells have not been reported.
Malignant tumors of testicular origin in androgen insensitive pa-
tients are rare (32).
The regression of the Mullerian ducts in androgen insensitive

patients agrees with the normal development of the testicular con-
nective tissue and tunica albuginea observed in the prepuberal pa-
tients. The abnormalities of the Wolffian system and the differen-
tiation of the external genitals reflect the complete or incomplete
androgen insensitivity of these structures at the time of masculine
development.
SUMMARY
Testicular development was studied in material from 300 human

embryos and fetuses and 35 male pseudohermaphrodites. The following
relationships were observed:
1. In normal chromatin-negative embryos of 15 to 19 mm (44 to

48 days) the primordial germ cells lose their round shape, glycogen
and alkaline phosphatase. At the same time, the gonadal blastema
28 J. E. JIRASEK
Fig. 5. Biopsy of an intra-abdomi.nal testis showing hyperplastic

and hypertrophied Leydig cells and immature Sertoli cells
lining non-luminized seminiferous tubules. Discrepancy
in the differentiation of the Leydig cells (hyperplasia
and hypertrophy) and Sertoli cells (immature) reflects the
testicular androgen insensitivity of the patient. Complete
testicular feminization in an 18 year old patient, hema-
toxylin-eosin, 150x.
differentiates into the testicular cords.

2. In embryos of 28 to 35 mm (55 to 60 days) the reticular
tissue of the tunica albuginea differentiates ending the transfor-
mation of gonadal blastema into testicular cords. At the same time,
the regression of the Mullerian ducts begins.
3. In fetuses of 32 to 35 mm (approximately 60 to 65 days) the
first Leydig cells showing 3 ~, -hydroxysteroid dehydrogenase activity
are present. No testicular 3~ -hydroxysteroid dehydrogenase activi-
ty was detected before differentiation of the Leydig cells.
4. Only testicles containing Leydig cells are able to synthe-
size androstenedione. In vitro, dehydroepiandrosterone is converted
to androstenedione by the testicular tissue of fetuses 8 to 9 weeks
old. Testosterone was synthesized by testicular tissues of fetuses
15 to 24 weeks old.
5. Masculinization of the external genitalia begins in fetuses
of 35 to 40 mm which are 65 to 70 days old. The scrotum and cavern-
ous urethra are formed between days 70 to 75. The presence of Ley-
dig cells is necessary for normal masculinization.
6. In some androgen-sensitive male pseudohermaphrodites, a

special type of testicular dysgenesis is present. A dysgenic testi-
cle is characterized by a thin insufficiently developed tunica al-
buginea, narrow degenerating seminiferous tubules and areas of cel-
lular connective tissue. Dysgenic testicles are unable to produce
regression of the Mullerian duct. Tumors arising from seminiferous
epithelium of dysgenic testicles were found in 7 of 11 cases studied.
7. In testicular biopsies, the non-luminized testicular tubules
with immature Sertoli cells, which are in contrast to the hyper-
plastic and hypertrophic Leydig cells, suggest the testicular andro-
gen insensitivity of the postpuberal patient.
ACKNOWLEDGMENT
This study presented is a part of Czechoslovakian State re-

search in human reproduction.
REFERENCES
1. McKay, D.G., Hertig, A.T., Adams, E.C. and Danziger, S., Anat.
Rec., 117: 201,1953.
2. McKay, D.G., Adams, E.C., Hertig, A.T. and Danziger, S., Anat.
Rec., 122: 125, 1955.
3. McKay, D.G., Adams, E.C., Hertig, A.T. and Danziger, S., Anat.
Rec., 126: 433,1956.
4. Rossi, F., Pescetto, G. and Reale, E., J. Histochem. Cytochem.,
5: 221, 1957.
5. Jirasek, J .E., Acta Histochem. (Jena), Q:" 220, 1962.
6. Witschi, E., Carnegie Inst., Washington, Contrib. Embryol.,
32: 67, 1948.
7. Streeter, G.L., Carnegie Inst., Washington, Developmental
Horizons in Human Embryo, Embryology Reprint Vol. II, 1951.
8. Singh, R.P. and Carr, D.H., Anat. Rec., 155: 369,1966.
9. Jirasek, J.E., Unpublished data.
10. Wenick, M. and McGory, W.W., Birth Defects, 4: 1, 1968.
11. Jirasek, J.E., Unpublished data.
12. Witschi, E., Recent Progr. Hormone Res., 6: 1, 1951.
13. Glenister, T.W., Nature, London, 177: 1135, 1956.
14. Park, W.W., J. Anat., 21.: 369, 1957.
15. Niemi, M., Ikonen, M. and Hervonen, A., Ciba Found. Colloq.
Endocr., 16: Endocrinology of the Testis, 31, 1967.
16. Jirasek, J.E., Ciba Found. Colloq. Endocr. 16: Endocrinology
of the Testis, 3, 1967.
17. Jirasek, J.E., Sulcova, J., Capkova, A., Rohling, S. and
Starka, L., Endocrinologie (Dresden), 54: 173-183, 1969.
18. Acevedo, H.F., Axelrod, L.R., Ishikawa, E. and Takaki, F.,
J. Clin. Endocr., l!.: 1611, 1961.
19. Bloch, E., Tissenbaum, B. and Benirschke, K., Biochim. Biophys.
Acta., 60, 182, 1962.
20. Jirasek, J.E., Raboch, J. and Uher, J., Amer. J. Obstet. Gynec.,
101: 830, 1968.
30 J. E. JIRASEK
21. Sohval, A.R., J. Urol, J.1.: 693, 1954.

22. Sohval, A.R., Amer. J. Hum. Genet., 12: 155, 1963.
23. Bergada, C., Cleveland, W.W., Jones, H.W., Jr. and Wilkins, L.
Acta Endocr. (Kobenhavn), 40: 521, 1962.
24. Zourlas, A.P. and Jones, H.W., Jr., Obstet. Gynec., 26: 48,
1965.
25. Young, D., J. Obstet. Gynaec. Brit. Comm., 58: 830, 1951-
26. Prichard, R.W., Arch. Path. (Chicago), 56: 505, 1953.
27. Motyloff, L., Z. Geburtsh. Gynaek., 99: 330, 1931.
28. Jirasek, J.E. and Prem, K.A., Unpublished data.
29. Northcutt, R.C., Island, D.P. and Liddle, G.W., J. Clin. Endocr.,
29: 422, 1969.
30. Mauvais-Jarvis, P., Bercovici, J.P. and Gauthier, F., J. Clin.
Endocr., 29: 417, 1969.
31. Jirasek, J.E. and Raboch, J., Fertil. Steril., 14: 237, 1963.
32. Hauser, G.A. In Intersexuality, ed. Overzier, C., Academic
Press, London, New York, 1963.
HETEROPLASTIC GONADUCT-TESTES COMBINATIUNS: A BIOCHEMICAL OUTLOOK
J. P. Weniger
Laboratoire de Zoologie et d'Embryologie experimentale,

Universite de Strasbourg, France
Ten years ago, I performed heteroplastic combinations of embry-

onic chick and mouse gonads and observed the feminization of the
chick testis by the mouse testis (1-3). This result prompted the
following experiments with gonaducts.
The question arises as to whether the embryonic chick and mouse

testes exert different effects on one and the same gonaduct. This
would imply that the hormone secretion of the two kinds of testes is
different. Our first consideration is the embryonic mouse wolffian
duct.
As shown by Jost (4-8) in his castration and grafting experiments,

the testes are essential for the development of the wolff ian ducts
in the male mammalian embryo. This statement is also valid for in
vitro conditions, as shown by Price and Pannabecker (9,10) in their
organ culture studies of embryonic rat reproductive tracts. In my
own cultures (11), embryonic mouse testes maintained embryonic mouse
wolffian ducts and stimulated their growth (Fig. 1). On the contrary,
when cultured in combination with embryonic chick testes, the mouse
wolffian ducts retrogressed (Fig. 2), just as they do when cultured
alone.
Secondly, let us consider the embryonic chick Mullerian duct.

This duct is a unique test organ for its responsiveness both to
masculinizing and feminizing actions. When subjected to masculinizing
influences, it retrogresses (12); feminizing actions produce stimu-
lation of the duct (13).
Experimentally (14), the chick testis brought about the regres-

sion of the chick Mullerian duct (Fig. 4), whereas the mouse testis
31
32 J. P. YVENIGER
1 2
Fig. 1. The left wo1ffian duct and the testes from a 17-day-01d
mouse embryo have been grown together in vitro for 3 days: curling
of the wo1ffian duct occurred.
Fig. 2. The right wo1ffian duct has been cultured for 3 days in
combination with the testes from an 8-day-01d chick embryo: retro-
gression occurred.
Fig. 3. The right Mullerian duct from an 8-day-01d female chick

embryo has been cultured for 5 days in combination with a testis
from an 18-day-01d mouse embryo: stimulation was observed.
Fig. 4. The left Mullerian duct has been cultured for 5 days in
combination with the testes from an 8-day-01d male chick embryo:
retrogression occurred.
produced stimulation (Fig. 3). Thus, the chick and mouse testes
exert opposite effects on the chick Mullerian duct. It may be con-
cluded from this as well as from the foregoing results that there is
a difference in the hormone secretion of the embryonic chick and
mouse testes. How far do the results of biochemical studies fit this?
Experiments by Jost (4-8) and Price and Pannabecker (9,10) have

shown that testosterone prevents the wolffian ducts from degener-
ating in the castrated male mammalian embryo or in vitro conditions.
Indeed, testosterone and androstenedione secretion has been demon-
strated in a great variety of embryonic mammalian testes (human:
Acevedo et al., (15, 16); Bloch et al., (17); Rice et al., (18);
Ikonen and Niemi (19); armadillo: Bloch and Benirschke, (20);
rabbit: Lipsett and Tullner, (21); rat: Noumura et al., (22); guinea
pig and dog: Bloch, (23); horse: MacArthur et al., (24); mouse:Weniger
et al., (25); calf: Weniger et al., (26); sheep: Attal, (27). The
question might then be posed as to whether the failure of the
embryonic chick testis to promote the development of the embryonic
mouse Wolffian duct may be attributed to the lack of testosterone
secretion by the chick testis.
Embryonic chick testes were cultured according to the method of

Wolff and Haffen (28) in the presence of 1_14C sodium acetate, 4_l4C
progesterone or 4_l4C androstenolone (Table 1). Analysis for
radioactive testosterone was carried out in the media, using multiple
thin-layer chromatography, derivative formation and recrystallization
to constant specific activity, but in no case could it be detected
(Figs. 5 and 6, Tables 2 and 3). Thus the lack of testosterone
secretion explains the failure of the chick testis to stimulate the
mouse wolffian duct.
Table 1. Experimental Conditions for Embryonic Chick Testes Cultures
Culture
Total
Experiment Precursor Number and Age of Testes Period
Activity (days)
(Eairs) (days)
1 Acetate 4.5 mC 889 10 1
2 Acetate 2 mC 373 14 1
3 Pro 5 fl.C 29 19 1
4 Pro 10 fl.C 60 12 3
5 Alone 5 fl.C 103 15 3
Pro: progesterone. Alone: androstenolone.

34 J. P. WENIGER
~G)
.~ Attenuation setting' 3000
u
a
o CI-lCI 3 -An 9:1
U
a
'-
®
Att. : 300 Be-EtAc 3:2
® Alt. :30 Be-EtAc 3:2
Alt. :3 Cy-EtAc 1:1
5
Fig. 5. Radiochromatograms pertaining to the experiment No. 4
(Table 1).
1: Total androgen fraction. There is no definite peak in

the testosterone region.
2: Androstenedione obtained by testosterone oxidation.
3: Testosterone obtained by androstenedione reduction. One-

fourth was acetylated (radiochromatogram No.4), the rest
recrystallized (Table 2).
4: Background level in the testosterone acetate region. Thus,

all the activity was contaminating.
Attenuation &etting : 100

u
o &olvent
o
'-g CHCI3 -An9:1
'-
® Atl.: 10
Cy-EtAc 1:1
® Att.:30
Cy-EtAc 1:1
5 10
Fig. 6. Radiochromatograms pertaining to the experiment No. 5
(Table 1).
1: Total androgen fraction. A peak seems to be associated

with carrier testosterone
2: Testosterone acetate.
3: Rechromatography of testosterone acetate: there is a

quite good correspondence with a peak, but recrystalliza-
tions will show that all is contaminating activity (Table 3)
36 J. P. WENIGER
Table 2. Recrystallization of Testosterone (Experiment No. 4)

SEecific Activity (c~Lmg)
Crystallization
Crystals Mother Liguors
1 258 760
2 83 580
3 74 132
4 58 61
Testosterone was recrystallized (from step 3, Fig. 5) in chloroform-
petroleum ether with 20 mg of carrier. Constant specific activity
could not be obtained. The background was 30 cpm.
Table 3. Recrystallization of Testosterone acetate (Experiment No. 5)

SEecific Activity (c~/mg)
Crystallization
Crystals Mother Liguors
1 560 1170
2 347 1740
3 170 810
4 124 605
5 83 242
6 230
7 36 69
Testosterone acetate (from step 3, Fig. 6) was recrystallized in
methanol-petroleum ether with 20 mg of carrier. Constancy of
specific activity could not be achieved. The background was 30 cpm.
Table 4. Experimental Conditions for Embryonic Mouse Testes Cultures
Total Culture
Experiment Precursor Number and Age of Testes Period
Activity
(Eairs) (days) (days)
1 Acetate 1 mC 179 14-19 1
2 Pro 10 ~C 28 15-17 3
3 Alone 5 ~C 69 14-18 3
Pro: progesterone. Alone: androstenolone.
It has been shown by Weniger, Ehrhardt and Fritig (29) that the
embryonic chick ovary secretes estrone and estradiol. Since the
embryonic mouse testis exerts on both the embryonic chick testis
and Mullerian duct the same feminizing effects as the chick ovary,
does this mean that it secretes estrone and estradiol too? Embryonic
mouse testes were cultured in the presence of 1_14C sodium acetate,
4_14C progesterone or 4_14C androstenolone (Table 4) but in no
case could estrogen formation be demonstrated.
On the other hand, it is unlikely that the secretion of testos-

terone or androstenedione, or both, would explain the feminizing
effects of the mouse testis. Indeed, if it is true that andros-
tenedione feminizes the chick testis, as shown by Wolff, Strudel
and Wolff (30) in their injection experiments, both testosterone
and androstenedione bring about the retrogression of the chick
Mullerian duct. In the same way, the chick Mullerian duct retro-
gresses when cultured on testosterone propionate-containing media
(12).
Thus, the biochemical outlook 1 ventured upon is strongly neg-

ative: The hormone or mechanism responsible for the feminization
of the embryonic chick testis and Mullerian duct by the embryonic
mouse testis remains to be discovered.
REFERENCES
1. Weniger, J. P., C.R. Acad. Sci. (Paris), 253: 2410, 1961.

2. Weniger, J. P., Arch. Anat. Micr. Morph. Exp., 52: 261, 1963.
3. Weniger, J. P., Ann. Endocr. (Paris), 26: 467, 1965.
4. Jost, A., Arch. Anato Micr. Morph. Exp:: 36: 271, 1947.
5. Jost, A., Arch. Anat. Micr. Morph. EXPe, 39: 577, 1950.
6. Jost, A., Recent Progr. Hormone Res. , ~:~79, 1953.
7. Jost, A., Mem. Soc. Endocr., 4: 237, 1955.
8. Jost, A., Mem. Soc. Endocr., 1:49, 1960.
9. Price, D., and Pannabecker, R., Cib.a Foundation Colloquia on
Ageing, 2: 3, 1956.
10. Price, D., and Pannabecker, R., Arch. Anat. Micr. Morph. Exp.,
48: 223, 1959.
12. Wolff, E., Lutz-ostertag, Y., and Haffen, K., C. R. Soc. BioI.
(Paris), ill,: 1793, 1952.
15. Acevedo, H. F., Axelrod, L. R., Ishikawa, E., and Takaki, F.,
J. Clin. Endocr.,2l: 1611, 1961.
16. Acevedo, H. F., Axelrod, L. R., Ishikawa, E., and Takaki, Fe,
J. Clin. Endocr.,23: 883, 1963.
17. Bloch, E., Tissenbaum, B., and Benirschke, K., Biochim. Biophys.
Acta.,60: 182, 1962.
18. Rice, B. F., Johanson, C. A., and Sternberg, W. H., Steroids, 7:
79, 1966.
19. Ikonen, M., and Niemi, M., Nature (London), 212: 716, 1966.
20. Bloch, E., and Benirschke, K., Endocrinology,76: 43, 1965.
21. Lipsett, M. B., and Tu1lner, W. W., Endocrinology,77: 273, 1965.
22. Noumura, T., Weisz, J., and Lloyd, C. W., Endocrinology,78: 245,
1966. --
23. Bloch, E., Steroids, 2: 415, 1967.
24. MacArthur, E., Short, R. V., and O'Donnell, V. J., J. Endocr.,
38: 331, 1967.
38 J. P. WENIGER
25. Weniger, J. P., Ehrhardt, J. D., and Fritig, B., C. R. Acad.

Sci. (Paris), 264: series D, 1069, 1967.
26. Weniger, J. P., Ehrhardt, J. D., and Fritig, B., C. R. Acad.
Sci. (Paris), 264: series D,1911, 1967.
27. Atta1, J., Endocrinology, 85: 280, 1969.
28. Wolff, E., and Haffen, K., J. Exp. Zool., 119: 381, 1952.
29. Weniger, J. P., Ehrhardt, J. D., and Fritig:-B., C. R. Acad.
Sci. (Paris), 264: series D, 838, 1967.
30. Wolff, E., Strudel, G., and Wolff, E., Arch. Anat. (Strasb),
31: 237, 1948.
GENERAL DISCUSSION
MANN: Professor Jost referred to the masculinizing effect of the

male fetus on the female co-twin, that is, on the freemartin. It is
somewhat surprising that in contrast to the enormous amount of work
that has been done on the influence of the male on the female, the
reciprocal influence of the female on the male co-twin has been
largely ignoredo An opportunity to investigate this aspect in a
quantitative manner arose in conjunction with a recent study of free-
martinism by Short et ale (Cytogenetics, ~:,369, 1969). We found
that although spermatogenesis was progressing in the testes of the
mature bull co-twin, the testes were small and immature with respect
to the testosterone/androstenedione ratio. The male accessory organs
were also reduced in size. The secretory function of the seminal
vesicles, as reflected by the formation of fructose and citric acid,
was one tenth of what it should have been in a normal bullo Thus it
would seem that the harm inflicted by the male on the female met with
vengeance, and that the male twin was unable to develop its own re-
productive tract normally.
FORD: Weiss and Hoffman from Giessen in Germany published a paper

last year in "Cytogenetics", in which they claimed the presence of
XX cells in the bull twin of the freemartin in tissue cultures pre-
pared from testicular material. I do not know which cells grew out
from the explants, but it is a very remarkable observation if it is
true and I would like to see it confirmed. If the stromal cells
grew out, how do the XX cells get there? How do they get into the
fixed tissue? And if the germ cells grew out, this is even more
remarkable.
JOST: When one studies freemartins after birth, one is studying the
sum of all the events which have occurred during the long pregnancy
period. In rats, it is obvious that early post-natal treatment with
estrogen can have these effects on the testis. It is also known
that in the calf fetus the ovary has the capacity to produce estro-
gens briefly before birth. This would suggest that the estrogenic
influence could occur late during pregnancy, when the morphological
sex differentiation of the young has long been completed. In our
39
40 GENERAL DISCUSSION
studies and in any others I know of, devoted to sex differentiation

in freemartins, i.e. during younger stages, no visible effect of the
female twin on the male twin has ever been observed. The physiolog-
ical significance as to sex and function of the exchange of cells be-
tween the male and female twin in cattle, mentioned by Dr. Ford, cer-
tainly needs more research.
E. STEINBERGER: We have studied a case of testicular feminization

whose testicular biopsy showed masses of Leydig cells and semini-
ferous tubules with Sertoli cells which appeared immature. The tis-
sue was capable of metabolizing progesterone to testosterone. Both
FSH and LH were detected in this patient's urine. Why should these
Sertoli cells remain in the immature state in spite of the presence
of FSH? Well, if we assume something different from the generally
accepted idea that Sertoli cells are under FSH control, and consider
the possibility that these cells are under testosterone control, i.e.
that their maturation depends on testosterone, then we may have here
an example of insensitivity to testosterone not only of peripheral
tissues, but also of the Sertoli cells.
JOHNSEN: We have made exactly the same observation as Dr. Steinberger,

in the syndrome of testicular feminization.
DICZFALUSY: Could Prof. Jost comment on the nature of the so-called

"Mullerian inhibitor"? Could this be the steroid with 21 carbon
atoms, or a compound which is non-steroidal?
JOST: A "Mullerian" inhibiting substance has not been isolated so

far, and of all the steroids assayed none has shown this effect.
LIPSETT: Was a systematic study made of the human testis in anen-

cephaly? Has the state of the pituitary gland been correlated with
the histology of the testis?
JOST: Studies on the testis of anencephalic infants have usually

been made at birth. There is no real developmental study, as far
as I know, for obvious reasons. It was found by Zondeck and Zondeck
that the interstitial cells in this condition were reduced. In some
cases the penis was somewhat reduced in size. But it was a true penis
showing that masculinization was complete. The condition of the gen-
ital tract would suggest that during the early weeks of pregnancy the
testis functioned normally. Thereafter, it lacked stimulation. This
fits the curve of excretion of chorionic gonadotropins during preg-
nancy.
LIPSETT: We tend to assume that, in the human, chorionic gonadotro-

pin plays a role in the development of the Leydig cell. We have had
the experience of having to abort a woman at four and a half months
of pregnancy. This patient presented metastatic choriocarcinoma and
rather high titers of HCG. The fetal testis showed Leydig cell hyper-
plasia.
MARTINI: I would like to know if an attempt has been made to dupli-

cate with crude testicular extracts the effects which are obtained
by testicular transplantations.
LUNENFELD: Was an attempt ever made to decapitate or encephalectomize

rat embryos prior to the 13th day and assess the fate of oogenesis ?
JOST: First Dr. Martini asked if anyone has tried testicular extracts.
This has been attempted in the past but probably not in the best way,
and should be repeated. Dr. Lunenfeld asked about meiosis in decapi-
tated rat fetuses. This has been studied recently by Mauleon in
France in unpublished experiments. In rat fetuses decapitated on
day 16.5, premeiosis occurred in the ovaries. This experiment must
be repeated at earlier stages since premeiosis starts between days
17 and 18. Premeiosis was not observed in the sections on day 14.
Other authors also state that in the rat premeiosis starts between
day 17 and 18. In any case we did not study fetuses decapitated on
day 13. In freemartins, if we assume that the effect of the male
twin on the female twin is produced by some testicular hormone (this
actually has never been demonstrated) then the following chronolog-
ical events are of great interest: the abnormalities in the ovaries
of the female twin begin to appear approximately at, or shortly be-
fore the time when the Mullerian ducts are inhibited by the testis
both in the male and in the female twin. Chronologically there is
a possibility that the same hormone inhibits the Mullerian ducts in
both the male and the female partners and the ovarian cortex in the
freemartin.
BERGADA: I want to make a brief comment on a human syndrome which

resembles very much the unilateral castrated rabbit experiment dem-
onstrated by Dr. Jost. In 1961, in collaboration with Dr. Wilkins, we
described this syndrome which we call asymmetrical gonadal differ-
entiation. It is generally induced by an XO/XY chromosomal mosaicism.
Surprisingly, the differentiation of the gonads is exactly what their
chromosomal constitution would theoretically induce, i.e., a testis
on one side and an atrophic or rudimentary gonad on the other side
which is similar to the gonadal streak observed in the pure XO cases
or Turner's syndrome. This suggests that in the human sexual differ-
entiation is similar to what Dr. Jost and others have showed in ex-
perimental animals.
JIRASEK: The Sertoli cells in testical feminization are undiffer-

entiated at least in some portions of the testis. The androgen in-
sensitivity of the individual is reflected in the testis by the lack
of maturation of Sertoli cells. With respect to testicular dysgenesis
this syndrome was present in all individuals with an XO/XY karyotype.
It is possible to distinguish three types of testicular dysgenesis:
1. Bilateral, with dysgenetic testes bilaterally; 2. Unilateral
testicular dysgenesis with a dysgenetic testis and a contralateral
gonadal streak; and 3. Unilateral testicular dysgenesis with a
42 GENERAL DISCUSSION
normal testis on the opposite side. Phenotypically, the cases of

the last group are described as males with a uterus. In some such
cases the phenotype, including external genitalia, is completely
normal male. The uterus is present ipsilaterally to the dysgenic
testis. With respect to hypophysial aplasia, it seems that up to
the seventh month of fetal development, in the absence of the pi -
tuitary, the testes are completely normal.
JOST: I was very interested to see Dr. Jirasek's cases of male

pseudohermaphrodites in which there is no inhibition of Mullerian
ducts, probably as a result of defective production of the Mullerian
inhibitor during fetal life. In these cases there was also an in-
complete differentiation of the albuginea. I wonder whether it might
be possible to make a parallel between the deficient albuginea differ-
entiation and lack of Mullerian inhibition in the pseudohermaphrodites,
on the one hand, and the inhibition of the ovarian cortex and of the
Mullerian ducts in freemartins, on the other.
CHRISTENSEN: There has been some discussion of Sertoli cell differ-

entiation this morning. Usually we think of cellular differentiation
in terms of the cell's ability to carry out its specific function.
Sertoli cells in a normal seminiferous tubule have a very complex
relationship with the germinal cells, and probably nurture and sup-
port them. I wonder how much we can say about Sertoli cell differ-
entiation where spermatogenesis is not being carried out. I do nat
know whether the size and appearance of the nucleus is an adequate
indication that the cell is capable of carrying out its full rol~
in supporting spermatogenesis.
JIRASEK: Infertile patients with Sertoli cells only in the testi-

cular biopsy,show no relationship between the germ cells and Sertoli
cells because there are no germ cells. In all these cases, after
puberty, the Sertoli cells are fully differentiated or mature. My
other remark is just a terminological one. I would be opposed to
the use of the term hormone for a substance producing regression of
the Mullerian duct because its action in male pseudohermaphrodites
is strictly ipsilateral. The dysgenetic testis or the streak gonad
is unable to induce regression of the ipsilateral Mullerian duct.
However, the regression of the contralateral Mullerian duct may be
normal.
MARTINI: Is there any evidence that testosterone can be converted

into dihydrotestosterone by the Wolffian and Mullerian ducts?
MANN: There have been several attempts to demonstrate the occurrence

of dihydrotestosterone in the mature bull testis but without any sU'c-
cess so far.
LIPSETT: I cannot remember whether Jean Wilson found the conversion

but she demonstrated the localization of dihydrotestosterone in the
Wolff ian duct, similar to the localization found in the prostate and
seminal vesicle.
BERGADA: We have studied several patients with different testicular

pathology who showed the relationship of interstitial tissue steroid
production and tubular development. One was a 19-year-old boy with
unilateral cryptorchidism. Another was a l7-year-old boy with hypo-
gonadotropic eunuchoidism and cryptorchidism. In the second case
there were no changes in the seminiferous tubules which coincided
with an absence of functional interstitial tissue, whereas the first
patient demonstrated tubular development. A third type of case is
exemplified by sexual precocity and a pseudo-tumor of the testis.
We studied a 5-year-old boy with marked hyperplasia of the Leydig
cells in one part of the testis, and complete development of the sem-
iniferous tubules with full spermatogenesis within this hyperplasia.
In the same testis, away from the area of Leydig cell hyperplasia,the
seminiferous tubules were infantile. Finally the testicular biopsy
of a 21-year-old patient with a feminizing testis revealed no changes
in the seminiferous tubules, which looked infantile. Since all the
androgen target organs of these patients do not respond to this hor-
mone, this phenomenon, and the previous situation, support the hy-
pothesis that the seminiferous tubule, independently of gonadotropin
action, is a selective target organ for the androgens secreted by the
surrounded Leydig cells.
JOST: It may seem strange that the testicular secretion which is

responsible for the disappearance of the Mullerian ducts has not
been better studied. As a matter of fact, this study is difficult
in mammals for technical reasons. In vivo experiments are not easy
to make on a large scale. In vitro experiments can be expected to
afford better possibilities. But Dr. Weniger in mice and Mrs. Picon
in my laboratory in rats observed that inhibition of the Mullerian
ducts by fetal testicular tissue grown together in vitro can be ob-
tained only during a very early and limited period of time. In ex-
periments with gUinea pig fetuses Dr. Dorothy Price has not succeeded
so far in maintaining the Mullerian ducts of male fetuses in vitro.
This is probably due to the fact that they disappear at a very early
stage. A convenient in vitro system must be devised in order to per-
mit this type of study.
WITSCHI: In my introductory remarks, I attempted to stimulate in-

terest in amphibians. They have not been entering the discussion at
all. Possibly they are not well enough known generally, and one
does not sufficiently realize that all the essential hormones found
in the higher forms,that is in mammals, and in birds, are already
present in amphibians. The anatomical situation is often very much
clearer, due to more primitive organization of the gonads. I wish
again to direct attention to the favorable features for work on pri-
mary sexual differentiation and initiation of steroidogenesis in
amphibia.
CHAPTER"
HISTOLOGY OF THE TESTIS

DYNAMICS OF HUMAN SPERMATOGENESIS
Yves Clermont
Department of Anatomy, McGill University, Montreal,
Canada
INTRODUCTION
The production of spermatozoa by the seminiferous epithelium

of mammals has been and still is the object of numerous cytological
and histological investigations (1,2,3). Spermatogenesis in man
has not been completely ignored by microscopists, even if it is dif-
ficult to obtain properly fixed sections of testis from normal adult
individuals for investigative purposes. From the morphological
point of view the phenomena occurring within the seminiferous tu-
bules may be classified as being either cytological or histological.
Amongst the cytological phenomena there are: the mode of prolifer-
ation, renewal and differentiation of spermatogonia, the meiotic
divisions of spermatocytes, the metamorphosis of spermatids into
spermatozoa, and the release of spermatozoa from the seminiferous
epithelium. At the histological level, considering the seminifer-
ous epithelium as a whole, there is a grouping of germ cells into
distinct cellular associations. The chronological evolution of
these cellular associations is known as the "cycle of the seminifer-
ous epithelium", while the orderly distribution of the cell associa-
tions along the seminiferous tubule is referred to as the "~Tave of
the seminiferous epithelium". In the following brief discussion
three topics concerning the dynamics of human spermatogenesis will
be examined: firstly the steps of spermatogenesis, secondly the
cycle of the seminiferous epithelium and its duration, and thirdly,
the mode of renewal of spermatogonia.
MATERIAL AND METHOD

Part of the material analyzed consisted of testicular biopsies
from healthy prisoners collected by Dr. C.G. Heller of the Pacific
47
48 Y. CLERMONT
Northwest Research Foundation, Seattle, U.S.A. This material was

fixed with Zenker-formol or with Bouin's fluid as modified by Cle-
land. Following fixation and embedding in paraffin, 5-~ sections
were stained with hematoxylin-eosin and with the periodic acid-
Schiff-hematoxylin techniques. The best biopsies collected were
serially sectioned and stained with hematoxylin-eosin. The topo-
graphical arrangement of spermatogonia along the tubular limiting
membrane and the distribution of the cell associations formed by
spermatogonia, spermatocytes and spermatids along the tubular wall
was investigated in reconstructions from these serial sections.
The topographical arrangement of spermatogonia was also investigated
in dissected tubules, fixed in Bouin's fluid, stained with hematoxy-
lin and mounted "in toto" between glass slide and coverslip.
OBSERVATIONS AND DISCUSSION
Spermatogenesis
Spermatogenesis is the elaborate process whereby spermatogoni-

al stem cells transform into spermatozoa, and involves the produc-
tion of several types of germ cells that will now be briefly de-
scribed.
Spermatogonia. With the light microscope the large majority of

spermatogonia can be classified into one of the following three ca-
tegories, respectively referred to as dark type A (Ad), pale type A
(Ap) and type B (B) spermatogonia (4,5). The dark type A spermato-
gonia have an ovoid nucleus containing a deeply stained dust-like
chromatin. In the central part of the nucleus a cavity containing
a pale stained material is usually present. One or two nucleoli
closely applied to the nuclear membrane are also visible (Fig. 1).
The pale type A spermatogonia have a discoid or ovoid nucleus con-
taining a pale stained granulated chromatin and show one or two nu-
cleoli attached to the nuclear envelope (Fig. 1). The type B sper-
matogonia show a spherical nucleus characterized by clumps or gran-
ules of heavily stained chromatin distributed along the nuclear mem-
brane, attached to a centrally located nucleolus or floating free in
the nuclear sap (Fig. 1). There is some variation in nuclear size
and morphology of spermatogonia and some cannot be identified as any
one of the three types just described (6,7). These spermatogonia
may be either distinct classes of spermatogonia or spermatogonia of
the three types described above but at various phases of the cdl
cycle (i.e. Gl, S or G2).
When examined under the electron microscope, the various clas-

ses of spermatogonia appear difficult to differentiate due to the
poor characterization of the chromatin (8,9). However, recently,
the three main types of spermatogonia mentioned above have been
identified in electron microscope photographs and the existence of
a fourth spermatogonial type has been postulated (10).
HISTOLOGY OF THE TESTIS 49
Fig. 1. Series of drawings illustrating the steps of spermatogene-

sis in man. Spermatogenesis begins with the spermatogonia
at the top left and terminates with a spermatozoon at the
bottom right. Labels: Ad, dark type A spermatogonia; Ap,
pale type A spermatogonia; B, type B spermatogonia; Pi,
pre leptotene primary spermatocyte; L, leptotene spermato-
cyte; Z, zygotene spermatocyte; EP, early pachytene sper-
matocyte; MP, mid-pachytene spermatocyte; LP; late pachy-
tene spermatocyte; II, secondary spermatocytes. Spermatids
are shown at various steps of spermiogenesis. A sper-
matozoon is illustrated as seen from ts lateral (left)
and frontal (right aspects. RB, residual body.
Spermatocytes. The pre leptotene primary spermatocytes have a

spherical nucleus containing a deeply stained granulated chromatin
which accumulates on the nuclear membrane and nucleolus. As these
primary spermatocytes evolve through the various steps of prophase
of the first maturation division, the nucleus undergoes a progres-
sive swelling, and the chromatin assumes the characteristic confi-
gurations known as leptotene, zygotene, and pachytene (Fig. 1).
Following diakinesis, the primary spermatocytes complete the first
maturation division and give rise to secondary spermatocytes. These
cells, which because of their short life span are not seen frequent-
ly within the seminiferous epithelium, have a spherical nucleus con-
taining a finely granulated chromatin and some larger deeply stained
globular masses (Fig. 1). They undergo the second maturation divi-
sions which yield spermatids.
50 Y. CLERMONT
Spermatids and spermiogenesis. Human spermatids embark upon

an extraordinarily complex series of cytological changes which have
attracted the attention of light and electron microscopists (3,11,
12,13,14). While we will limit ourselves to a brief description
of the main steps in spermatid development as seen with the light
microscope, it should be remembered that underlying the gross chan-
ges to be mentioned, extremely intricate submicroscopical phenomena
are taking place. Newly-formed spermatids are characterized by a
small spherical nucleus and the usual array of cytoplasmic organel-
les, among which the Golgi zone, mitochondria, centrioles and chro-
matoid body are particularly prominent. A well demarcated Golgi
zone, which is visible next to the nucleus, soon starts to elaborate
small granules which stain vividly with the PA-Schiff technique.
These small corpuscles, referred to as proacrosomic granules, coa-
lesce to form a single larger granule, the acrosomic granule, which
becomes closely attached to the surface of the nucleus (Fig. 1).
The head cap then appears, expanding around the acrosomic granule
and growing over the surface of the nucleus. The acrosomic granule
and head cap, both well stained by the PA-Schiff technique, toge-
ther constitute the acrosomic system, which has been shown, with
the electron microscope, to be a modified secretory granule and a
membrane bound organelle. The Golgi zone, which contributed mater-
ial to the acrosomic system throughout its formation and develop-
ment, eventually separates from it to float freely in the cytoplasm
(Fig. 1).
At the same time this is going on at one pole of the nucleus,

centrioles at the opposite pole have begun to sprout the flagellum.
A short time later, a delicate fibrillar structure, the caudal tube
or manchette, appears in the cytoplasm, and extends from a close
association with the nucleus to surround the flagellum.
As spermiogenesis continues, the acrosomic system rotates to-

ward the limiting membrane of the seminiferous tubule; this reori-
entation is accompanied by a displacement of the nucleus toward the
periphery of the cytoplasm. Following this displacement, the nu-
clear chromatin starts to condense and becomes more chromophilic.
Its shape changes from a sphere to an elongated fusiform body with
a conical apex covered with the acrosomic structure (a distinction
between acrosome and head cap cannot be made from here on). Around
the flagellum, and close to the centriolar apparatus, a delicate
"ring" or annulus appears. This ring eventually slides down the
flagellum for a short distance and comes to rest at the caudal ex-
tremity of the middle piece. The middle piece is that portion of
the flagellum between the centrioles and the ring around which the
mitochondria will accumulate in a regular covering. The caudal tube
disappears soon after the formation of the middle piece.
As spermiogenesis reaches completion, the nucleus condenses

further and flattens slightly (front and lateral views are shown in
Fig 1). The anterior two thirds of the nucleus is covered by the
delicate acrosomic system, now rather pale staining with PA-Schiff.
Much of the cytoplasm, which was seen along the flagellum during
the second half of spermiogenesis, flows toward the nucleus and sep-
arates from the cell to become the residual body. (The Sertoli
cells are actively involved in the process of residual body forma-
tion, 15).
Some Characteristics of Spermatogenesis
It is of value at this point to present certain characteristics

of spermatogenesis that will serve to introduce the histological
phenomenon known as the cycle of the seminiferous epithelium. It is
soon apparent to anyone investigating the development of mammalian
germ cells, of rodents in particular, that the series of changes
taking place during spermatogenesis must be a rigidly programmed
and timed process. It has in fact been observed from radioautogra-
phic studies on rodents in which 3H-thymidine was used as a label
of germ cells that the whole process of spermatogenesis has a con-
stant duration (9,16,17,18). The same can be said for each indivi-
dual step of spermatogenesis. Thus, the duration of spermatogene-
sis or of any of its component steps for a given species and strain
of animal has a given and fixed duration.
In various mammals, including man, it appears that several

spermatogonial stem cells form a group and enter spermatogenesis
simultaneously. Moreover, topographical analyses on spermatogonial
stem cells of the rat (20), bull (21), monkey (22) and man (4) have
clearly indicated that when these cells divide, they do so synchro-
nously in groups. Thus, clusters of germ cells evolve through the
entire spermatogenic process synchronously, and as a consequence,
large numbers of cells all at the same step of development are seen
in the seminiferous epithelium. One such group of cells will be
referred to as a "generation" in the following discussion.
New groups of spermatogonial stem cells are periodically enter-

ing into spermatogenesis before the previous generations have com-
pleted their transformation into spermatozoa. Because of this, the
seminiferous epithelium is composed of several generations of germ
cells, the youngest generations being close to the limiting membrane
of the tubule and the more mature germ cells being closer to the lu-
men of the tubule (Fig. 3).
THE CYCLE OF THE SEMINIFEROUS EPITHELIUM
It has long been recognized that the various generations of

germ cells are not associated at random, but form cellular associa-
tions of fixed composition (1,23). Thus for example, spermatids at
a given step of development are always associated with the same
types of spermatocytes and spermatogonia. Consequently, only a lim-
52 Y. CLERMONT
ited number of cell associations can be seen in various cross sec-

tions of the seminiferous tubules. This characteristic arrangement
of the germ cells implies that in anyone area of the seminiferous
epithelium, the aforementioned cellular associations follow each
other in time. The resulting orderly sequence gives rise to the
so-called "cycle of the seminiferous epithelium" (23).
Each typical cell association which may then be considered as

a "stage of the cycle" can be classified and identified using va-
vious arbitrary cytological or histological criteria. Our previous
studies of the seminiferous epithelium in several mammals identi~
fied the stages by making use of the steps of development of the
spermatids stained with the periodic aCid-Schiff-hematoxylin tech-
nique; the periodic acid-Schiff component demonstrated particularly
well the changes taking place in the acrosomic structure of the
spermatids (23).
The remarkable arrangement of the germ cells described for

most mammals was more difficult to demonstrate in the human testis.
Some authors, impressed by the poor organization of spermatogonia,
spermatocytes and spermatids in the seminiferous tubules of man,
concluded that in contrast to other mammals there was little, i.f
any, correlation between the developmental steps of the germ cells
(24,25,26). The author (27), following an investigation of serial-
ly sectioned human testicular biopsies fixed in Zenker-formol, found
that man was not fundamentally different from other mammals in that
germ cells formed typical cell associations. A classification of
these typical cell associations into six stages of a cycle was then
proposed. The cellular composition of these stages is illustrated
in Fig. 2. For a detailed description of the cycle the reader is
referred to the original publication (27). In the course of this
study it was emphasized, however, that the typical cell associations
were frequently ,obscured by irregularities resulting from the fol-
lowing features. Firstly, in man (unlike most other species where
cell associations occupy large segments of tubule), each typical
cell association occupies small areas of the tubular lining, such
that in tubular sections usually more than one cell association can
be identified (Fig. 3). Thus, cells at the interface of two adja-
cent cell associations usually intermix to produce groupings which
are difficult to classify. Secondly, cellular associations are
often incomplete due to the absence of one or two generations of
germ cells. Thirdly, even a slight pressure on the delicate semini-
ferous tubule.s at the time of collection of the biopsy displaces
germ cells, disorganizes the seminiferous epithelium and makes the
identification of the stages of the cycle impossible. All these
factors and possibly others contribute to a disorderly appearance
in human tubular histology.
The frequency with which the various stages of the cycle in

man appear in tubular sections was taken to be an index of their
(I:)
..
~'
s
- .,
..
,-
J
'-'
, S
~ tgl
:11 :
p
i~
.c@ ~
l
(~)
:z~
. ~;
AP~ J();
Ad~j )0)
III IV V VI
STAGES 0# THE CYCLE
Fig. 2. Diagram illustrating the cellular composition of the six

stages of the cycle of the seminiferous epithelium in man.
The stages of the cycle are labelled with Roman numerals
(I-VI) and correspond to cell associations which succeed
one another in time in any given area of the seminiferous
tubule according to the sequence from left to right. Fol-
lowing stage VI, stage I reappears and the sequence starts
over again. The space allotted to the stages of the cycle
are proportional to their relative duration. Lettering:
Ad, Ap, B, respectively dark type A, pale type A, type B
spermatogonia ; Pi, L, 2, P, respectively pr~leptotene,
leptotene, zygotene, pachytene primary spermatocytes; II,
secondary spermatocyte; S, spermatids; S2, spermatozoon,
illustrated from its ventral aspect. The arrows corres-
pond to the most advanced labelled cells at various time
intervals (indicated on the arrows) after an intratesticu-
lar injection of 3H-thymidine (modified from Heller and
Clermont, 28). See discussion in the text.
relative duration (27). The stages were of unequal duration, so the

space allotted to each stage of the cycle in Fig. 2 was made to cor-
respond approximately to its relative duration.
The duration of the cycle of the seminiferous epithelium in man

was estimated by using 3H- t hymidine and radioautography (28,29).
The radioactive label was injected locally in the testis and biop-
sies were collected from the sites of injection at various intervals
thereafter. Some observations that were made on the radioautographs
54 Y. CLERMONT
Fig. 3. Low power microphotograph of a tubular cross section show-

ing side by side three typical cell associations (numbered
by Roman numerals) as well as their approximate demarca-
tion lines. Fixation: Zenker-formol; staining: hematoxy-
lin -eosin. Magn. x 40.
are shown in Fig. 2. One hour after 3H-thymidine injection some sper-
matogonia and many pre leptotene spermatocytes were labelled as they
were duplicating their DNA in preparation for mitosis or meiosis. No
other germinal elements showed a radioautographic reaction. There-
fore, the pre leptotene spermatocytes, the oldest germ cells to be
labelled one hour after injection, were considered as the "most ad-
vanced labelled cells". This group of cells was followed through
the cycle of the seminiferous epithelium at various time intervals
after 3H- t hymidine injection. Arrows in Fig. 2 give the stages of
the cycle in which the most advanced labelled cells, obviously de-
rived from the labelled preleptotene spermatocytes at one hour, were
seen at 8, 16, 24 and 32 days after 3H-thymidine administration. At
one hour, 16 and 32 days after injection, the most advanced labelled
cells, i.e. the preleptotene, pachytene spermatocytes and spermatids
respectively, were located in stage III of the cycle. Therefore, it
took 16 days for the pre leptotene spermatocytes to go through a com-
plete series of six stages (i.e. IV, V, VI, I, II, III) and reappear
in stage III as pachytene spermatocytes. Similarly, it took another
16 days for these pachytene spermatocytes to go through another com-
plete series of six stages to become spermatids in stage III. The
HISTOlOGY OF THE TESTIS 55
duration of a complete series of six stages or of one cycle of the

seminiferous epithelium would thus be close to 16 days. This es-
timate was confirmed by observations made at 8 and 24 days after
injection. At these two time intervals, the most advanced labelled
cells were pachytene spermatocytes and early spermatids, both found
at the onset of stage I of the cycle (Fig 2). In this case again
it took 16 days for the stage I - pachytene spermatocytes to go
through the complete series of six stages to reappear in stage I as
newly-formed spermatids.
MODE OF RENEWAL OF SPERMATOGONIA
It is obvious that the continuous production of spermatozoa by

the seminiferous tubules, production which in quantitative terms is
astronomical over the life span of a normal individual, requires
not only an active proliferation of spermatogonia but a maintenance
or renewal of the stock of spermatogonial stem cells. While the
mode of renewal of spermatogonia has been extensively analyzed in
several mammalian species (1,3) the kinetics of the human spermato-
gonial population has been the object of relatively few investiga-
tions. Roosen-Runge and Barlow (26) were the first to tackle the
problem systematically by means of a quantitative analysis of the
spermatogonial population. From their study they proposed the exis-
tence of seven classes of spermatogonia; they regarded the apparent
existence of seven differently sized groups of dividing spermato-
gonia in metaphase as confirmation of this conclusion. They also
suggested that new spermatogonial stem cells were produced by any
one of the first three or four spermatogonial divisions.
We have re-examined the problem (4,5) following a quantitative

analysis of spermatogonia and pre leptotene spermatocytes at the va-
rious stages of the cycle and a study of the topographical distri-
bution of the dark and pale type A spermatogonia along the tubular
limiting membrane. From these data it was found that for each dark
type A spermatogonium, there is one pale type A cell at all times
during the cycle of the seminiferous epithelium, an observation that
was confirmed by the quantitative data of Steinberger and Tjioe (30).
It was also calculated that the proportion of either of these two
classes of type A spermatogonia to type B cells was 1:2 and finally
that the ratio of dark type A or pale type A spermatogonia to pre-
leptotene spermatocytes was 1:4. In summary, the proportions of the
various cells counted was as follows: Ad:Ap:B:Pl = 1:1:2:4:. It was
also observed that the dark and pale type A spermatogonia, seen in
maps which show their topographical arrangement along the tubular
wall, were arranged in homogeneous groups and that within these
groups the spermatogonia formed pairs of identical cells. This ar-
rangement of the dark and pale type A spermatogonia throughout the
duration of the cycle was taken to indicate that the mitoses of
spermatogonial stem cells are "equivalent" in nature (Le. mitoses
56 Y. CLERMONT
giving rise to two morphologically identical cells); therefore, the

possibility of having "differential" mitoses (i.e. divisions giving
rise to two morphologically distinct spermatogonia) to explain the
renewal of spermatogonial stem cells should be abandoned. A similar
finding was made in several other mammals (19,20,21). These various
data (i.e. quantitative and topographic) were used to build a ten-
tative model to illustrate the mode of renewal of spermatogonia in
man (Fig. 4). In this model the dark type A spermatogonia were
considered as stem cells and according to this scheme one member of
a pair of type Ad spermatogonia would give rise to a pair of new
stem cells while the other member would produce a pair of the more
differentiated type Ap spermatogonia. These in turn would produce
type B spermatogonia which would give rise to spermatocytes (Fig.
4).
Ap PI
B
PI
Ap PI
B
PI
Ad
Ap
Ad
Ad
IV V VI II III IV V VI I II III IV V VI
STAGES OF THE CYCLE
Fig. 4. Model illustrating the development and renewal of sperma-
togonia in man. Lettering: Ad, Ap, B, respectively dark
type A, pale type A, type B spermatogonia; Pl, prelepto-
tene spermatocytes. Roman numerals indicate the stages of
the cycle. This model (taken from Clermont, 4) was mainly
based on the following two observations: a) spermatogoni-
al mitoses are equational in nature; b) the ratios of cell
counts were the following: Ad:Ap:B:Pl = 1:1:2:4. This
model is considered only as a tentative illustration of
spermatogonial renewal (see discussion in the text).
Although supported by the counts, this model is based on a

number of assumptions that remain to be confirmed by additional ex-
perimental data. These assumptions are: 1) There is only one gen-
eration in each class of spermatogonia; 2) Spermatogonia of each
type divide once but only once during each cycle of the seminife-
rous epithelium; 3) The dark type A spermatogonia constitute the
most primitive spermatogonial elements and a class of renewing stem
cells; the pale type A spermatogonia would be differentiated ele-

ments which arise from the dark type A and in turn give rise to
type B cells; 4) Once spermatogonia embark on their course of dif-
ferentiation into spermatocytes (Ap onward) their evolution is un-
idirectional and irreversible; they do not revert back to the stem
cell condition; 5) The incidence of spermatogonial degeneration is
not important enough to alter the cell ratios. Since so many as-
sumptions remain to be confirmed, it is clear that the model pro-
posed should not be considered as being definitive but, on the con-
trary, as a first approximation suggested by the data collected so
far.
SUMMARY
In man, as in other mammals, spermatogenesis is a long and

elaborate process which involves several distinct classes of ger-
minal elements. There are the spermatogonia of which there are at
least three main types in human tubules, the dark type A, the pale
type A and the type B spermatogonia. Spermatogonia proliferate to
give rise to primary spermatocytes. These go through a long pro-
phase which shows characteristic morphological configurations of
the chromosomes (leptotene, zygotene, pachytene). This prophase is
followed by the sUbsequent steps of the first maturation division
that yields secondary spermatocytes. These cells, after a short
interphase, go through the second maturation division which results
in the formation of spermatids. In turn, spermatids metamorphose
through an extremely complex series of morphological changes into
spermatozoa.
The timing of the various steps of spermatogenesis appears to

be rigidly fixed and since several spermatogonial stem cells enter
spermatogenesis simultaneously, large groups of germinal cells
evolve through spermatogenesis synchronously. One such group is
referred to as a "generation" of germ cells. The seminiferous epi-
thelium is composed of four or five distinct generations of germ
cells. In man, as in other mammals, such generations are not ran-
domly arranged, but form cellular associations of fixed composition.
These cell associations, six in man, follow each other chronologi-
cally in a regular sequence in any given area of the seminiferous
tubule, to form the cycle of the seminiferous epithelium. Such
cell associations are often overshadowded by irregularities mainly
resulting from the mixing of germ cells. The cycle of the semini-
ferous epithelium lasts approximately 16 days.
The mode of renewal of spermatogonia was investigated by means

of a quantitative and topographical analysis on the various types
of spermatogonia; from this, a tentative model to illustrate the
development of spermatogonia was proposed. According to this model
the dark type A spermatogonia would divide to produce equal numbers
of new dark type A spermatogonia and of pale type A spermatogonia.
58 Y. CLERMONT
Each one of the latter would give rise to two type B spermatogonia
which in turn would produce four spermatocytes. This model, based
on several assumptions not yet verified by observation, should be
considered only as a preliminary approximation; the problem of
spermatogonial renewal in man remains an open question.
ACKNOWLEDGEMENTS
This work was supported by the Population Council, Inc. The

help of Dr. Claire Huckins in the preparation of the manuscript is
acknowledged. The drawings were prepared by Mrs. M. Oeltzschner.
REFERENCES
1. Roosen-Runge, E.C., Biol. Rev., 12: 343, 1962.

2. Leblond, C.P., Steinberger, E. and Roosen-Runge, E.C. In Mecha-
nisms Concerned with Conception, Pergamon Press, New York, 1963.
3. Clermont, Y., Arch. Anat. Micr. Morph. Exp. (Suppl.),56: 7,
1967.
4. Clermont, Y., Amer. J. Anat., 118: 509, 1966.
5. Clermont, Y., Fertil. Steril.,12: 705,1966.
6. Barr, A., M.Sc. Thesis, McGill University, Montreal, Canada,
1967.
7. Mancini, R.E., Narbaitz, R. and Lavieri, J.C., Anat. Rec., 136:
477, 1960.
8. Tres, L.L. and Solari, A.J., Z. Zellforsch., 91: 75, 1968.
9. Vilar, O. and Paulsen, C.A., Anat. Rec., 157:336, 1967 (Abstr.)
10. Rowley, M.J., Berlin, J.D. and Heller, C.G., Proceedings of the
Society for the Study of Reproduction, 1969.
11. Gatenby, J.B., Beams, H.W. and Woodger, J.H., Quart. J. Micr.
Sci., 65: 137, 1921.
12. Clermont, Y. and Leblond, C.P., Amer. J. Anat., 96: 229,1955.
13. Fawcett, D.W. and Burgos, M.H., Ciba Foundation Colloquium on
Ageing, 2: 86, 1956.
1.4. Fawcett,-D.W., Int. Rev. Cytol., 2: 195, 1958.
15. Fawcett, D.W. and Phillips, D.M., J. Reprod. Fertil. (Suppl.),
.£: 405, 1969.
16. Clermont, Y., Leblond, C.P. and Messier, B., Arch. Anat. Micr.
Morph.Exp " 48: 37, 1959.
17. Clermont, Y. and Harvey, S.C., Endocrinology, 76: 80, 1965.
18. Desclin, J. and Ortavant, R., Ann. Biol. Anim.~ioch. Biophys.,
1: 329, 1963.
19. Clermont, Y. and Leblond, C.P., Amer. J. Anat., 104: 237, 1959.
20. Clermont, Y. and Bustos-Obregon, E., Amer. J. Anat., 122: 237,
1968.
21. Hochereau, M.T., Proceedings VI Congo Intern. Reprod. Anim.
Insem. Artif., Paris 1: 149, 1968.
22. Clermont, Y., Amer. J. Anat., 126: 57,1969.
23. Leblond, C.P. and Clermont, Y., Ann. N.Y. Acad. Sci., 55: 548,
1952.
24. Branca, A., Arch. Zool. Exp. Gen., 62: 53, 1924.
25. Roosen-Runge, E.C., Fertil. Steril.~2: 251, 1956.
26. Roosen-Runge, E.C. and Barlow, F.D., Amer. J. Anat., 93: 143,
1953.
27. Clermont, Y.,Amer. J. Anat. 112: 35, 1963.
28. Heller, C.G. and Clermont, Y., Science, 140: 184, 1963.
29. Heller, C.G. and Clermont, Y., Recent Progr. Hormone Res.,
Academic Press, 20: 545, 1964.
30. Steinberger, E. and Tjioe, D.Y., Fertil. Steril., 19: 960,
1968. --
DISCUSSION
SOLARI: Several years ago we stated that different bodies can be

identified in the nuclei of human spermatogonia (Tres, L. and Solari,
A.J., Z. Zellforsch mikr. Anat., 91: 75, 1968). One kind of sper-
matogonium has a type of nuclear body which we called type 1; iden-
tified by histochemical methods and ultrastructural techniques as a
typical nucleolus. Other spermatogonia have another kind of nuclear
body, which is not a characteristic nucleolus, but is attached to
the nuclear membrane. It is not chromatin, hence it cannot be a
sex chromosome. We have called it type II body. The type B sper-
matogonium has chromatin bodies which are never closely associated
with the nuclear envelope, so again here we have no evidence of a
heteropyknotic sex chromosome. We have some histochemical reactions
on another kind of nuclear body, which we called number 4; the fifth
kind of nuclear body is the only one which is chromatin in nature,
but is too small to be a sex chromosome. In summary, we have five
types of nuclear bodies in human spermatogonia, and these facts re-
late to what Prof. Clermont has said, that human spermatogonia must
have several steps in their evolution, and these steps in evolution
are probably expressed by the characteristic nuclear structures.
CLERMONT: Since I have not studied these cells with the electron
microscope I have no comments to make on this point.
VILAR: Among the prepuberal spermatogonia one cannot see the typical
adult A pale or A dark spermatogonia. They appear during puberty.
PAULSEN: Dr. Clermont, what is the status of the so-called A0 sper-

matogonia in human beings? I believe that you have described the Ao
spermatogonia in the rat. Also, would the Ao spermatogonia corre-
spond to the Ai spermatogonia of the monkey?
CLERMONT: The type Ai spermatogonia, which we have described in the

monkey, are probably the equivalent to what we have called the type
Ao spermatogonia in the rat, judging from their proliferative activ-
ity. As far as the human is concerned, we still do not know of
course, but it is quite possible that the A dark spermatogonia would
be the equivalent of the type Ao spermatogonia of the rat and the
60 Y. ClERMONT
type Ai of the monkey.
E. STEINBERGER: Dr. Clermont, I was particularly interested in your

expression of ambiguity concerning the question of which is the stem
cell - the dark or the pale type A spermatogonia. You have proposed
in your publications that it is the dark type A although no definite
evidence is available for this. I would like to suggest that the
type A pale spermatogonium may be the stem cell, at least in the
mature type of spermatogenesis. This suggestion is made on the basis
of two lines of evidence: 1) that in culture, gradually, the more
mature cell types degenerate, and the ratio between the dark and
pale changes and shifts towards the pale. In other words, with time,
we have numerous pale type spermatogonia which keep dividing, while
the dark disappear. 2) the observation on two biopsies from oligo-
spermic individuals, in which no spermatocytes or spermatids were
present and the seminiferous tubules were lined with type A pale
spermatogonia. I feel that these observations provide some evidence
that it is the pale spermatogonia that may be the stem cells.
CLERMONT: What was the age of the testicular tissue that you used
for your culture?
E. STEINBERGER: The age varied anywhere between 20 and 30 to 60

days.
CLERMONT: As I said in my presentation, most of the spermatogonia

that we see in the human seminiferous tubules could be classified
amongst the three classes of cells that were mentioned, i.e. the
dark and the pale type A spermatogonia and the type B spermatogonia,
but there is still the possibility that other classes of spermato-
gonia exist. Dr. Heller's group, from their electron microscopic
studies, postulated the presence of a fourth type of spermatogonium,
so it is not impossible that other spermatogonia with type A pale
characteristics would be present along the limiting membrane and
could serve as stem cells. We still know too little about the human
spermatogonial stem cells.
JIRASEK: I would like to know what is the metabolic significance of

the dark or clear spermatogonia. I always thought that the clear sper-
matogonia were related to the accumulation of glycogen in these cells.
The spermatogonia containing glycogen in the cytoplasm appear in the
sections as clear. I do not know what cell type begins to accumulate
glycogen.
CLERMONT: I should say first, that when I talk of dark or pale type
A spermatogonia, I always refer to the staining of the chromatin.
What we observed in PA-Schiff stained preparations is that glycogen
is present within the cytoplasm of dark type A spermatogonia, while
the pale type A have a rather pale cytoplasm. The significance of
this observation remains to be clarified.
JOHNSEN: I would like to know if one germinal cell association is

associated with one particular Sertoli cell, so that one Sertoli cell
and its germ cell association form a unit.
CLERMONT: I do not think so; the area occupied by the given cell
association is extremely variable. In some tubules, a given asso-
ciation may occupy a large area, and in this case many Sertoli cells
will be seen in it. In other tubules, the same cell association
many occupy a small and narrow area of the seminiferous epithelium,
and in this case one or a few Sertoli cells will be seen in it. I
would say, however, that generally, there are many Sertoli cells per
typical cell association.
BURGOS: In favor of a metabolically low active type of spermatogonium

in the human testis, the electron microscope shows some cells which
contain cytoplasmic crystalloids of Lubarsch and clusters of closely
packed mitochondria. These cells usually have a dark nucleus and
sometimes contain a portion of clear nucleoplasm which appears under
the light microscope as a vacuole.
JOST: Dr. Clermont, you emphasized the fact that the chronology of
all the events during spermatogenesis is very rigid. I would like
to know if you have any evidence that this is controlled by some oth-
er cells, as for instance the Sertoli cells, or if the germ cells do
it by themselves?
CLERMONT: We have no evidence favoring one or the other possibility.

I do not really know what could control the timing of spermatogenp
esis.
ELECTRON MICROSCOPY OF THE HUMAN SEMINIFEROUS TUBULES
Oscar Vilar, C. Alvin Paulsen and Donald J. Moore
Centro de Investigaciones sobre Reproduccion, Facultad

de Medicina, Buenos Aires, Argentina and Department of
Medicine, Washington State University, Seattle, Wash.
The germinal epithelium of the testis is composed of different

cell types which constantly undergo changes through a complicated
combination of proliferation, differentiation and renewal. Clermont
has postulated (1-3) that human germ cells are systematically ar-
ranged into six basic stages, and has proposed models for the devel-
opment and renewal of spermatogonia in man. On the other hand other
authors have stated that the cellular division and combinations in
the human testis lack the synchronism and steadiness which is seen
in other mammals (4-6).
With the development of the electron microscope, certain ultra-

structural characteristics of the human testis have been investigated.
Data have been accumulated with respect to the testicular connective
tissue (7,8), the Sertoli cells (9-13) and the spermatids (14-16).
The present paper is directed towards obtaining information

which would provide a basis for more detailed investigation into the
histophysiology and histopathology of the germinal epithelium. In
recent years research in cellular pathology has emphasized the im-
portance of cytoplasmic(17) rather than nuclear changes, as these ap-
pear to be the earliest cellular modifications following diverse
types of injury. Therefore, identification and description of the
normal germinal cell ultrastructural characteristics are essential
to the understanding of morphological events following injury or di-
sease processes.
MATERIAL AND METHODS
For these studies biopsy specimens were obtained from 10 healthy
63
64 O. VILAR, C. A. PAULSEN, AND D. J. MOORE
men who had no record of former testicular disease. Bippsy was per-
formed by a modification of Charny's technique (18). The material
was prepared for electron microscopy following the method already
reported (19).
RESULTS
Tubular Wall
Electron microscopical studies of the wall of the human semin-

iferous tubules clearly showed two main components: a) The internal
one, attached to the germinal epithelium is the basal membrane pre-
senting a lamellar structure, the lamellae are separated by spaces
of low electron-density(18). This structure presented infoldings
which appeared simple or rather labyrinthic and included some col-
lagen fibers or cytoplasmic processes of the wall cells. The depth
of indentation reached one or two microns. b) The external layer
showed an alternate arrangement of collagen fibers oriented at dif-
ferent angles together with cells. The cells are extremely elon-
gated, their endoplasmic reticulin was poorly developed and was main-
ly in the form of lamellae. Fine filaments aggregated in bundles
are seen in the cytoplasm of the cells which were found to be some-
what similar to smooth muscle fibers (20).
Germinal Epithelium
The germinal epithelium is composed of spermatogenic or germ-

inal cells and Sertoli cells. It is now a widely accepted belief
that there are two independent cell lines in the mammalian testis.
The Sertoli cells are most conspicuous in this epithelium. The

germinal cells, with the exception of the earlier spermatogonia are
dispersed amongst the Sertoli cells and are separated from the base-
ment membrane by the cytoplasmic processesoof the Sertoli cells,
which in many cases are not wider than 200A.
Sertoli Cells. These showed a well developed endoplasmic reti-

culum which appeared as flat and parallel aggregated cisternae, with
associated ribosomes arranged on the membranes in single strands.
This type predominated in the cytoplasmatic processes surrounding
the germinal cells. In the basal portion of the cell, which contains
its characteristically lobulated nucleus with a prominent nucleolus,
ribosomes were arranged in membrane-free clusters, and the endoplas-
mic reticulum appeared in Caufield's solution-fixed preparations as
numerous empty smooth-surfaced vesicles of about 600 to 1800X in di-
ameter. Fixation with glutaraldehyde changes this appearance into a
rather tubula~ form: short, tortuous, interconnected tubules aver-
aging 240-300A in diameter with localized dilatations were seen in
the central portions of these cells, while in the periphery they
formed arrays of long, flat membranous profiles. The ground sub-
HISTOLOGY OF THE TESTIS 6S
stance of the cytoplasm appeared as a dense material made of gran-

ules smaller than ribosomes. As the density of the ground substance
is higher than that of the germinal cells, the body of the Sertoli
cells and their processes clearly stood out over the surrounding
elements. Regarding the density of the ground substance, two types
of Sertoli cells can be distinguished; dark and light. At present
we are not able to relate these two types to any functional stage
or ascertain if they correspond to the two different types of cells
demonstrated by Johnsen at light microscopical level (21). Ground
substance also included bundles of microtubules of about 300~. Mi-
crofibrils with diameter averaging 3si were particularly numerous
around the nuclear membrane. They were prominent near the head of
the spermatids where they appear in bundles parallel to the axis of
the spermatid (12). Sertoli cells show wide attachment to the la-
mellar basement membrane, whose characteristic "pegs" protrude into
these cells (7,8).
Spermatogonia. Different types of spermatogonia can be dem-

onstrated with the electron microscope. In the following descrip-
tion we are going to avoid identification by using the alphabet.
There are two reasons for this: first, there. is more than one paper
dealing with spermatogonia where similar letters refer to different
cells (1,22). Secondly, when following alphabetical order it could
be assumed that this order is related to the sequence of the cells
in the spermatogenic line.
The flat type spermatogonium appeared as a rather small, oval-

shaped cell. Its larger diameter was parallel to the basement mem-
brane to which it was attached. The nucleus contained one or two
conspicuous nucleoli composed of coarse, dense strands which branch-
ed and anastomosed in an irregular network. The ground substance of
the cytoplasm contained disseminated m~crotubules and very abundant,
uniformly scattered fine granules, 120A in diameter, which made
these cells appear darker than other spermatogonia and spermatocytes,
although lighter than Sertoli cells. The endoplasmic reticulum was
poorly developed and consisted of a few flat membranous profiles,
while occasional scattered clusters of four or five larger and den-
ser ribosomal granules were seen. The round or ovoid mitochondria
were relatively large and contained a dense matrix with clearly
marked parallel cristae; they tended to aggregate close to the poles
of the nucleus. The unremarkable Golgi apparatus also appeared in
the vicinity of the nucleus.
Following glutaraldehyde fixation some of the flat spermatogo-

nia showed a circular rarefaction in the center of the nucleoplasm,
which appeared as a "vacuole" of less density, but without any li-
miting membrane or special structure within it. This characteris-
tic was not exclusive to flat spermatogonia as similar "vacuoliza-
tions" could also occasionally be seen in round spermatogonia, pre-
leptotene spermatocytes and in spermatids.
The round type spermatogonium, a larger, circular cell, was

also attached to the basement membrane. Its round, homogeneous nu-
cleus was lighter and contained a small nucleolus which frequently
did not show up in the ultra-thin sections. The ground substance
of the cytoplasm was less dense and contained no microtubules, al-
though the Golgi apparatus was more conspicuous. While the mito-
chondria were smaller, they showed well-developed cristae in a
dense matrix. In contrast to the flat type spermatogonia, they
were scattered throughout the cytoplasm. Small crystal like struc-
tures were seen. They presented an irregular shape and were formed
of bundles of fibrils or tubules aligned in the long axis. The nu-
cleus of human spermatogonia has been studied recently by Tres and
Solari (23). Five different types of nuclear bodies were observed
in type A (flat) spermatogonia. Type I bodies are typical nucleoli.
Types II, III and V are considered to be atypical nucleoli. Type
IV bodies are small chromatin condensations. These authors stated
that they could not establish a valid criterion for distinguishing
the different kinds of A gonia at the ultrastructural level although
they suggested the existence at least 2 types of A gonia. In these
authors' criteria the different nuclear bodies probably reflect dif-
ferent metabolic activities.
A third type of cell was noted which seemed to correspond to

the usual descriptions for pre leptotene primary spermatocyte. The
cytoplasmic subcellular characteristics of this cell were very si-
milar to those of round spermatogonia and showed little resemblance
to those of primary spermatocytes. They differed from the round
type spermatogonia in that they were separated from the basement
membrane, as well as from neighboring cells, by the previously de-
scribed ramifications of the Sertoli cell cytoplasm. It was fre-
quently noted that intercellular bridges existed between these
pre leptotene cells and round spermatogonia.
Although in the course of meiosis primary spermatocytes under-

went remarkable changes in size as well as in the characteristics
of the nucleus, the subcellular features of the cytoplasm remained
stable. These changes in the nucleus will be presented later on by
Dr. Solari. The numbers of polyribosomes dispersed throughout the
ground substance were greater than in any other germ cell type. But
since the ground substance contained little else, the primary sper-
matocytes were the lightest cells in the germinal epithelium. Short
and very flat membranous profiles were scattered throughout the cyto-
plasm, and flat saccules were located throughout the nucleus. The
prominent Golgi apparatus, round and rather compact in form, was sit-
uated at one of the poles of the nucleus. The cristae mitochondri-
ales did not show the straight parallel pattern which was seen in
the spermatogonia; instead, they followed an irregular course and
frequently coalesced with the outer membrane, giving the mitochon-
dria a "vacuolated" appearance (24). Microtubules were scattered
throughout the cytoplasm, and although their appearance was undis-
tinctive, they seemed to be unrelated to the microtubules of the

spindle during the first meiotic division.
The earliest modification of primary spermatocytes, which en-

abled those cells to be distinguished from spermatogonia, was the
appearance of very fine single leptotene threads. These were always
present even in the youngest spermatocytes. We were, therefore, un-
able to find a "resting" stage that could be interposed between the
spermatogonial mitosis and the meiotic prophase.
The secondary spermatocytes were relatively small cells, slight-

ly larger than young spermatids, and obviously smaller than the
neighboring spermatocytes. They were seen infrequently and appeared
close to or partially in the lumen of the seminiferous tubules. The
nucleus was round with a rather homogeneous aspect; the nucleolus
was very inconspicuous and was absent much of the time. The endo-
plasmic reticulum was made of rosaries of small flat saccular pro-
files which were arranged concentrically in relation to the nucleus.
The limiting membranes of these saccules were agranular and tortuous.
Scattered among the endoplasmic reticulum,round mitochondria were
found,which present the same vesiculization seen in the primary sper-
matocytes and spermatids. Resting and dividing secondary spermato-
cytes were generally seen co-existing in the same portion of the tu-
bule.
During the course of spermiogenesis, the spermatids showed

changes in the configuration of the acrosome which did not differ
from those classically known (14,15). Five different stages of pro-
gressive change in the morphology of the spermatid were pasily char-
acterized.
In Stage A, corresponding tG the beginning of spermiogenesis,

the spermatids were round, with a round homogeneous nucleus. To-
ward the end of this phase the acrosome granule, as well as the
acrosome vesicle, could be seen.
The cap finally became established during Stage B, and the nu-
cleus could be seen starting its elongation, although the appearance
of the nucleoplasm remained unchanged. During this stage the micro-
tubule of the manchette first appeared, while within the cytoplasmic
branches of the Sertoli cell similar microtubules became arranged
parallel to the axis of the spermatid.
Elongation proceeded during Stage C, but the most remarkable

change occurred in the nucleoplasm where chromatin-like material be-
came condensed in flocs or large irregular granules, which gave the
nucleus a "tigroid" appea.rance.
During Stage D, progressive reduction in nuclear size could be

observed, the chromatin flocs formed wider clusters, and the nucleus
became more and more compact until all the interstices among the
flocs were obliterated and the nucleus again presented a homogeneous
appearance.
In the final phase, Stage D2, most of the cytoplasm of the sper-
matid became eliminated, and residual bodies could be seen; either
partially free in the lumen of the tubule, or in the process of being
"phagocytozed" by the Sertoli cells.
DISCUSSION
Each cell in the human germinal epithelium shows, in both the

cytoplasm and the nucleus, particular subcellular characteristics
which allow its identification in electron micrographs. For the
cytologic study of the testis, electron microscopy offers undoubted
advantages over light microscopy. The distinctive features of the
germinal cell types that can be established with the light micro-
scope, either in normal or pathological conditions, represent pri-
marily morphological configuration of the chromatin and the nucleo~
lus, since the cytoplasmic details are poor and the cellular limits
are difficult, or impossible to determine. Furthermore, different
fixatives give diverse chromatin patterns, and it is sometimes dif-
ficult to classify spermatogonia or correlate the different classi-
fications which appear in the literature (3,22).
Clermont (1) recognized two types of A spermatogonia (A dark

and A pale), the first one having a large vacuole-like cavity in the
nucleus. He gave a ratio of Ad:Ap of 1:1. We were not able to es-
tablish with the electron microscope the same discrimination among
spermatogonia. Following fixation with glutaraldehyde, we observed
in some nuclei a rarefaction of the nucleoplasm which might appear
as a vacuole-like structure, but since procedures in the preparation
of the specimen are different, it is difficult to affirm that this
"vacuole" corresponds to that described by light microscopists. On
the other hand, our ratio between the "vacuolated" and "non-vacuo-
lated" flat spermatogonia is not 1:1, but closer to that of Roosen-
Runge and Barlow (5) who found vacuoles in 96 of 1000 nuclei. We
observed in the fragments fixed in glutaraldehyde, that similar
"vacuoles" could be found in other types of germinal cells (gono-
vytes, spermatogonia, round spermatids, etc.). Thus it appears that
this "vacuole" is not a distinctive characteristic of any type of
cell, or of any specific stage, but a kind of modification of the
nucleus of the germinal cells due to the aldehyde fixation. Its
significance we are not yet able to evaluate.
We did not find an identifiable resting stage of primary sper-

matocyte that would be interposed between the last spermatogonial
mitosis and the meiotic prophase. The youngest spermatocyte seemed
to he already in meiosis. Sotelo and Trujillo-Cenoz (25) reported
the same finding in other species, and stated that "the appearance
of very fine chromosome threads (leptotene threads) seems to be the

better mark to recognize the transformation of spermatogonia into
spermatocytes. The pre leptotene cell, here described, presented
subcellular characteristics, which corresponded to those of B sper-
matogonia rather than spermatocytes, except for the fact that they
were not attached to the basement membrane but were surrounded by
Sertoli cells and so separated from the other cells and the base-
ment membrane. The existence of cellular bridges between these
cells and B spermatogonia unquestionably indicates their origin.
The preleptotene stage probably represents a more differentiated
type of B spermatogonia moving into the core of the germinal epi-
thelium.
It could be speculated that the establishment of the above men-

tioned relationship between the Sertoli and the pre leptotene cell is
an unavoidable determining factor that induces, after the last sper-
matogonial division, the transformation of the spermatogonium
into a primary spermatocyte, which in time will be able to accom-
plish meiotic division. The function of the Sertoli cells is not
yet known, but it was suggested long ago that the products of the
secretion of these cells could supply nutrition to the germinal
cells (26). Also, it has been shown that they can act as an active
transporting system as labelled substances injected into adult rats
leave the circulation in the intertubular spaces and rapidly move
into the seminiferous tubules, diffusing through the cytoplasm of
the Sertoli cells (10). Christensen and Fawcett( 27) have advanced
the idea that the site of hormone production in steroid producing
cells is in the agranular endoplasmic reticulum. The agranular
endoplasmic reticulum of the Sertoli cell could be considered a mor-
phological base to support the claim that these cells produce ste-
roid hormones, but no consistent proof has been found at the present
time. As was also speculated by Christensen (8) the parallel micro-
tubules in the Sertoli cytoplasm could be related to the mechanism
of the release and holding of the spermatids.
The supporting cells reach maturity at pubertal ages, simultan-

eously with the beginning of spermatogenesis, and an obvious regu-
lation by the hypophysis occurs. We can only conjecture as to whe-
ther gonadotropins act upon both germinal and Sertoli cells or wh~
ther the primary target is the supporting cells (Sertoli) which then
could influence the germinal ones. Courot (28) injected gonadotro-
pic preparations into lambs whose testes were not yet showing sper-
matogenic activity and observed that while no important evolution
of the gonocytes was seen, a fairly constant stimulation of precur-
sors of Sertoli cells was observed.
Mancini et ale (22) have described additional types of sperma-

togonia in the adult human testis; binucleated, hypertrophic and
pyknotic. While binucleated spermatogonia, as well as other binu-
cleated types, are seen frequently and merely indicate a delay of
the cytokinesis as Fawcett and Burgos (14) have reported, hypertro-

phic and pyknotic cells are seen rather seldom, and in our opinion
have little significance. Large spermatogonia have been interpreted
as polyploid cells (5), and pyknosis probably indicates failure of
isolated cells to pursue the complete evolution toward spermatozoa.
Fig. 1. Human Sertoli cell. Note the well developed nucleolus.

x 17,000
Fig. 2. Flat spermatogonium in contact with the basal lamina.

x 30,000.
Fig. 3. Flat spermatogonium. Note the "vacuolization of the nu-

cleus" x 27,300.
Fig. 4. Round spermatogonium. Note that the basal lamina forms

"pegs" which project into the Sertoli cells. x 18,000.
Fig. 5. Spermatocyte II. x 17,000.

Fig. 6. Spermatid B. Note the tubules of the "manchette" as well

as those in the cytoplasm of Sertoli cell. x 40,000.
Fig. 7. Spermatid D. Note condensed nuclear material and the con-

necting piece. x 27,300.
ACKNOWLEDGEMENTS
Supported by the National Research Council (Argentina) and

by AEC contract No. At (45-1) - 2225.
REFERENCES
1. Clermont, Y.,Amer.J. Anat. 112: 35,1963.

2. Clermont, Y. ,Amer.J. Anat. 118: 509, 1966.
3. Clermont, Y., Fertil. Steril. 12: 705, 1966.
4. Branca, A., Arch. Zool. Exp. et Gen. ~: 53, 1924.
5. Roosen-Runge, E.C. and Barlow F.B.,Amer.J. Anat. 93: 143, 1953.
6. Markewitz, M., Veenema, R. and Fingerhut, B., Int. J. of Fertil.,
~: 22, 1969.
7. Burgos, M.H., Rev. Soc. Argent. Biol. 35: 309, 1959.
8. Mancini, R.E., Vilar, 0., P. del Cerro, M. and Lavieri, J.C.,
Acta Physiol. Lat. Amer., ~: 382, 1964.
9. Fawcett, D.W. and Burgos, M.H., Anat. Rec. 124: 401, 1956.
10. Vilar, 0., Perez del Cerro, M.P. and Mancini, R.E., Exp. Cell
Res., !:2: 158, 1962.
11. Bawa, S.R., J. Ultrastruct.Res., 9: 459, 1963.
12. Nagano, T., Z. Zellforsch., 89: 39, 1968.
13. Nagano, T., Z. Zellforsch.,.21: 89, 1966.
14. Fawcett, D.W., Burgos, M.H., Ciba Colloquia on Aging of Tran-
sient Tissues, 2: 86, 1955.
15. Horstmann, E., Z. Zellforsch., 54: 68,1961.
16. De Kretser, D.M., Z. Zellforsch., 98: 477, 1969.
17. Trump, B.F., Goldblatt, P.J. and Stowell, R.E., Lab. Invest.
~: 1969, 1965.
18. Paulsen, C.A., In Textbook of Endocrinology, ed. Williams, R.H.
R.E. Saunders, Philadelphia, p. 405, 1962.
19. Vilar, 0., Steinberger, A. and Steinberger, E., Z. Zellforsch.
74: 529, 1966.
20. Clermont, Y., Exp. Cell Res. 15: 438, 1958.
21. Johnsen, S.G., Acta Endocr. (Kobenhavn) 61: 111, 1969.
22. Mancini, R.E., Narbaitz, R. and Lavieri, J.C., Anat. Rec. 136:
477, 1960.
23. Tres. L. and Solari, A.J., Z. Zellforsch., ~: 75, 1968.
24. Andre, J., J. Ultrastruct. Res. Suppl. 1: 1, 1962.
25. Sotelo, J.R. and Trujillo-Cenoz, 0., Z. Zellforsch., 51: 243,
1960.
26. Regaud, C., Arch. Anat. Micr. Morph., ~: 231, 1901.
27. Christensen, A.K. and Fawcett, D.W., J. Biophys. Biochem. Cytol.
9: 653, 1961.
28. Courot, M., Ann. Biol. Anim. Bioch. Biophys. 5: 145, 1965.
FINE STRUCTURE OF TESTICULAR INTERSTITIAL CELLS IN HUMANS
A. Kent Christensen
Department of Anatomy, Stanford School of Medicine
Stanford, California 94305
It is now well established that the interstitial or Leydig

cells are the principal site of androgen production in the mammalian
testis. In rats, the interstitial tissue can be separated from the
seminiferous tubules and shows considerably more activity than the
tubules in synthesizing androgen from cholesterol in vitro (1,2).
Proof that the interstitial cells are indeed the source of this ac-
tivity within the interstitial tissue comes from histochemical
studies in which the reaction product of 3~ -hydroxysteroid dehydro-
genase appears mainly in these cells. This approach has been ap-
plied to a variety of vertebrate species with consistent results,
although it appears somewhat more difficult to demonstrate in the
human testis (3).
For a number of years we have been interested in the fine

structure and correlated function of testicular interstitial cells,
especially in guinea pigs (4-6), rats (7), mice (8) and opossums
(9). In the present paper I will summarize the fine structure of
human testicular interstitial cells, based on the literature and on
recent electron microscopy done in this laboratory on human testes
fixed by perfusion.
There is already a considerable literature on the light and

electron microscopy of human interstitial cells. Observations with
the light microscope have been reviewed by Stieve (10), Ito and
Oinuma (11), and Sniffen (12). The most useful general descriptions
of human interstitial cells as seen by electron microscopy are those
of Fawcett and Burgos (13), Yamada (14) and de Kretser (15,16); pa-
pers by Nishimura and Kondo (17), Schmidt (18), Hatakeyama (19) and
Nagano (20) are also available. Pelliniemi and Niemi (21) have fol-
lowed the development of interstitial cells with the electron micro-
75
76 A. K. CHRISTENSEN
scope in the human fetus. The fine structure of interstitial cells

in the testicular feminization syndrome (22) and in an androgen-
producing tumor (23) have been described. Restricted topics have
been dealt with by Desolle (24) Sisson and Fahrenbach (25) and Ya-
suzumi et ale (26).
With all these papers available on the fine structure of hu-

man interstitial cells, why publish another one here? Aside from
whatever use a brief review may serve, the main justification lies
in the improved preservation for electron microscopy obtained by
perfusion-fixation in the present material. Steroid-
are notoriously difficult to fix for electron microscopy, and this
has led to considerable dispute about some important aspects of
their fine structure. The most notable uncertainty concerns the
normal structural organization of the smooth endoplasmic reticulum
(ER), which is characteristically abundant in steroid-secreting
cells. Most of the enzymes of androgen biosynthesis from small pre-
cursors are probably either attached to the smooth ER or lie in the
cytoplasm near it (see (7) for an extensive review of the fine
structure and cell fractionation literature on steroid-secreting
cells). It is of considerable functional interest to know whether
this reticulum is a continuous system of interconnected membrane
tubules, which would allow transport within a stable closed system,
or whether it is a discontinuous system of membrane vesicles, posing
very different functional problems. The evidence one obtains from
electron micrographs on questions such as this depends on how the
tissue was preserved for electron microscopy.
The method of preservation used in all the previous studies on

human interstitial cells is to cut the testis tissue into small
pieces in a drop of fixative. Aside from the mechanical damage as
the tissue is diced, the slow penetration of the fixative gives am-
ple time for degenerative change and a differential response by
cells to the gradual increase in fixative concentration around them.
As a result, the quality of fixation may vary markedly from one
place to another in a single small piece of tissue. In contrast,
fixation by perfusion supplies fixative rapidly and uniformly to all
parts of the testis without disturbing structural relationships, and
the preservation obtained is of consistent high quality throughout
the whole organ. Perfusion-fixation has been used previously in
this laboratory to study testicular interstitial cells in guinea
pigs (4) and rats (7).
In the present paper I will describe the fine structure of tes-

ticular interstitial cells from human testes that have been fixed
by perfusion with a formaldehyde-glutaraldehyde-picric acid fixative
(27), used 2/3 strength). One testis came from a 64-year old man
undergoing orchiectomy because of prostatic cancer. The man had
otherwise been healthy and active, and the testes appeared rela-
tively normal by light microscopy. The other testis used in this
study was from a 22-year old man, and was removed because it had
become lodged under the skin above the external inguinal ring. This
testis appeared cryptorchid by light microscopy. In both cases the
testes were received from surgeons within a minute or so after the
blood supply had been clamped off, and the stubs of two main branches
of the testicular artery in the spermatic cord were then cannulated
with blunted 20 gauge needles. Details of the perfusion apparatus
and method have been published previously (4). Perfusion continued
for about an hour, after which the testes were allowed to remain
another hour before being sliced and were then left in fixative for
another two or three hours. The material was stored in the refrig-
erator overnight or longer in 0.1M sodium cacodylate buffer. Small
pieces of tissue were postfixed in 1.3% osmium tetroxide buffered
with s-collidine (28), and were then dehydrated and embedded in Epon
(29). The material was stained in block with 1% uranyl acetate in
ethanol during the dehydratiQn. Pale gold sections were stained
with lead citrate and were viewed in an RCA EMU-3F or a Siemens
Elmiskop lA electron microscope. Sections 2~ in thickness for light
microscopy were cut from the Epon blocks and were stained by the
method of Richardson et ale (30).
In the rest of this chapter I will first catalogue and briefly

describe the various structures seen in human interstitial cells,
basing the description on the electron micrographs of perfused mat-
erial that illustrate the paper,with supplemental information from
the literature. I will then attempt to correlate functional infor-
mation with these fine structure observations to the modest extent
that is possible at present, and will also compare the organelles
as seen in humans with those in steroid-secreting cells of other
species. A more extensive description of the interstitial tissue
in the human testis fixed by perfusion is being prepared for publi-
cation elsewhere.
GENERAL DESCRIPTION OF THE INTERSTITIAL CELLS
In a light micrograph of perfusion-fixed testis from a 64-year

old man (Fig. 1), the dark interstitial cell clusters are seen be-
tween lighter seminiferous tubules. Small blood vessels, appearing
white and somewhat swollen here because of the perfusion, often have
a few interstitial cells associated with them or may pass through
interstitial cell clusters. The white rectangle in Fig. 1 encloses
an area similar to that shown in a very low power electron micro-
graph in Fig. 2.
The large cluster of interstitial cells in Fig. 2 lies in the

interstitial tissue outside the tunica propria, a sheath of flat-
tened (and perhaps contractile) cells surrounding the seminiferous
tubule. A capillary (cap) receives hormone produced by ·the cluster.
Below the capillary is an adventitial macrophage (mac). Three inter-
stitial cell nuclei are visible at the top of the cluster. The white
circles in the interstitial cells are lipid droplets whose content
Fig. 1. Light micrograph of human testis fixed by perfusion. Dark

clusters of interstitial cells lie between the seminiferous
tubules. The area enclosed by the rectangle is similar to
that shown in a low power electron micrograph in Fig. 2.
From 64-year old man; 2~ Epon section of material fixed
and embedded for electron microscopy. x 450.
Fig. 2. Edge of a seminiferous tubule and adjacent interstitial

cells, showing general level of preservation in perfused
material. This very low power electron micrograph shows
an area resembling that enclosed in the rectangle in figure
1. A capillary (cap) and adventitial macrophage (mac) are
included in the field. The white circles in the intersti-
tial cells are extracted lipid droplets. From a 64-year
old man. x 2400.
has been extracted by the solvents used for electron microscopy.

The dark gray structures scattered throughout the cytoplasm are mi-
tochondria, while the black granules, especially evident at lower
right in the cluster, are lipofuscin pigment granules. Note the
uniform texture of the cytoplasm in the interstitial cells of this
perfused material. There are no "dark" and "light" cells, as have
been described in human testis fixed by routine methods (for exam-
ple, 15,16). There is also no indication that the smooth endoplas-
mic reticulum is vesiculated in any of the cells. The preservation
shown here is typical of that seen in all the sections that have
been examined from both perfused testes.
Fig. 3. Human interstitial cells, showing general cellular organi-

zation. See text for description and abbreviations. Ecto-
pic testis from 22-year old man; perfusion-fixed. x 7,100.
Fig. 3 is an electron micrograph showing most of one intersti-

tial cell and parts of two others, still at fairly low magnifica-
tion. The nucleus (n) is large and round, and contains a prominent
nucleolus (nu) made up of a network of nucleolonemata. There is
little heterochromatin in the nucleus, except for a scanty layer,
here almost lacking, just inside the nuclear envelope (it is more
prominent in the nuclei of Fig. 2). The cytoplasm contains a few
large, partially extracted lipid droplets (lp) and many mitochondria
(m), generally rod-shaped and about 0.3~ in diameter, that are cut
in various planes of sections. The cytoplasm between the mitochon-
dria is filled with smooth endoplasmic reticulum (ser), which will
be better seen in subsequent electron micrographs. The cell con~
tains an unusually large patch of rough endoplasmic reticulum (rer),
consisting of five flattened cisternae; isolated small cisternae are
scattered here and there in the cytoplasm. Elements of the Golgi

complex (g) are seen in various parts of the cell. No basal lamella
is visible on the outer surface of the plasma membrane. Small white
spots in the extracellular space are negative images of collagen
fibers (co) cut in cross section.
A view of the cytoplasm at higher magnification is seen in Fig.

4. The most striking feature of the interstitial cell cytoplasm,
as in most steroid-secreting cells, is the abundant smooth endoplas-
mic reticulum (ser) composed of interconnecting tubules extending
throughout the available cytoplasmic space. The tubules are gener-
ally 600-700 ~ in diameter and branch and anastomose freely in three
dimensions, although in this two-dimensional view many are seen only
in cross or oblique section. At upper left are small profiles of
rough endoplasmic reticulum (rer) which connect (arrow) with the
smooth reticulum. Free ribosomes (r) lie between the tubules of the
smooth ER, but are not attached to them. Some of these ribosomes
occur in clusters that probably constitute polysomes (ps). The mi-
tochondria (m) are generally rod-shaped but here are mostly cut in
cross or oblique section. Their internal structure will be seen to
better advantage in Fig. 5. A large partially extracted lipid drop-
let (lp) is present in the lower part of the field. An element of
the Golgi complex (g) consists of two (usually more) flattened sacs
lying parallel to one another. At right is part of a crystal of
Reinke (cr), showing one of the several crystal patterns seen as
this structure is cut in various ways relative to its three crystal
axes. Within the nucleus at the bottom of the micrograph is scanty
heterochromatin (h) and a fibrous lamina (fl) lining the inner sur-
face of the nuclear envelope, which is penetrated by occasional
pores (p). The surface of the cell at upper left is thrown into
irregular folds which lie against processes from adjacent cells;
in one area the adjacent plasma membranes have fused as a tight or
gap junction (tj).
Fig. 5 is a thinner section at higher magnification to give

further detail of some structures seen previously. The tubules of
the smooth ER are cut in longitudinal (ser) and various oblique or
cross sections. No content can be made out within these tubules.
The mitochondria are generally rod-shaped, of variable diameter,
and may branch in complex ways (m). I t is difficult to interpret
the three-dimensional form of the cristae. In some cases they ap-
pear to be lamellar, extending across the mitochondrion, while
others seem to be tubular. These profiles can be explained if one
assumes that most of the cristae are fenestrated lamellae. The
cristae presumably arise from the inner membrane of the mitochon-
drion, but only occasionally can a direct connection be found (ar-
row). The usual mitochondrial granule (mg) is often seen, and it
is not uncommon to find what appear to be intramitochondrial lipid
droplets (not shown here). Within the mitochondria shown in Fig. 5
Fig. 4. Cytoplasm of human interstitial cell, showing abundant

smooth endoplasmic reticulum (ser) and crystal of Reinke
(cr). See text for other abbreviations. Ectopic testis
from 22-year old man; perfusion-fixed x 24,000.
Fig. 5. Detail of cytoplasm in human interstitial cell. The mito-

chondrial cristae are probably in the form of fenestrated
cisternae. Occasional cristae clearly arise from the in-
ner mitochondrial membrane (arrow). The mitochondria also
contain conventional matrix granules (mg) and presumptive
mitochondrial ribosomes (mr). Ectopic testis from a 22-
year old man; perfusion-fixed. x 33,000.
are also numerous small particles of a size and appearance one would
expect for mitochondrial ribosomes (mr). A nucleus (n), a crystal
of Reinke (cr) and presumptive lysosomes (ls) are also present in
the field. Lysosomes may allow the development of residual bodies,
in which the polymerization products of autoxidized unsaturated fat-
ty acids may accumulate to produce a pigment, giving rise to lipo-
fuscin pigment granules (shown only at low magnification in Fig. 2).
In this description I have emphasized the structures whose pre-

servation is improved by perfusion-fixation, and have not pretended
to cover the extensive literature on Reinke's crystals and their
precursors, on cytoplasmic filaments, on lipofuscin pigment or on
some other topics of interest in the fine structure of human inter-
stitial cells.
COMPARATIVE AND FUNCTIONAL CONSIDERATIONS
In conclusion I will summ3rize briefly how the organelles of

human testicular interstitial cells compare with those seen in the
steroid-secreting cells of other mammals. In addition, I will out-
line some of the correlations that might be made between structure
and function in human interstitial cells, if one is allowed to sup-
plement the meager evidence from human material with data from ani-
mal studies on steroid-secreting tissues. Such extrapolations are
hazardous because of possible tissue and species differences, but
they may at least give a heuristic functional outlook to the fine
structure that has been presented. To attempt either of these sum-
maries with full documentation would be cumbersome and would sur-
pass the bounds allowed for this brief chapter. Those who desire
the complete references are referred to our review (7) describing
the fine structure and cell fractionation studies on steroid-secret-
ing cells in mammals.
Human interstitial cells are moderate steroid-secreting cells

from several points of view. The celts average about 15\-L in diame-
ter and are thus larger than interstitial cells of rats (about 10\-L),
but considerably smaller than those of the boar (about 35\-L). The
smooth ER is more abundant than that seen in rat interstitial cells
(certainly more than rat adrenal cortical cells), but is not as vol-
uminous as that of guinea pig or opossum interstitial cells, or
guinea pig granulosa lutein cells. Human interstitial cells have
less rough ER than many other steroid-secreting cell types. The mi-
tochondria resemble morphologically those of the guinea pig inter-
stitial cell, and are very different from the larger mitochondria
with tubular cristae seen in mouse interstitial cells. Even these
are unimpressive compared to the very large mitochondria with vesi-
cular cristae that dominate the cytoplasm of adrenal fasciculata
cells in the rat adrenal. The development of adrenal mitochondria
reflects their additional role in carrying out the 11 (3 -hydroxylase
and (in the glomerulosa) il~ -hydroxylase reactions. Interstitial

cells have very large mitochondria in the human fetus. The Golgi
complex has a sameness about it throughout steroid-secreting cells,
being somewhat concentrated at one pole of the nucleus, around the
centrioles, but also sending arms out to other parts of the cell.
Reinke's crystals are supposedly unique to man, although Stieve (10)
says they occur in dogs. The human lipofuscin granule resembles
that seen in mouse interstitial cells, but that found in guinea pigs
is organized differently- rat and opossum interstitial cells lack
lipofuscin granules. Human interstitial cells are moderate in their
content of lipid droplets. They contain considerably fewer than
adrenal fasciculata cells of rats, but do contain some, in contrast
to the interstitial cells of mature rats, opossums and boars, where
lipid droplets are uncommon.
Fig. 6 summarizes the probable movements of substrates within

a cell over the course of steroid biosynthesis, according to the
cell fractionation literature. Cholesterol biosynthesis seems to
take place on or near the smooth endoplasmic reticulum. The wide
variation in the amount of smooth ER in steroid-secreting cells led
to the suggestion (4) that the amount of smooth ER reflects the ex-
tent to which the cell makes its own cholesterol rather than taking
it in from the plasma, and some correlations with biochemical liter-
ature were cited. One interesting example is the insect prothoracic
gland, source of the steroid ecdysone, where cells almost entirely
lack a smooth ER. It turns out that their cholesterol is derived
entirely from the diet (31).
Whether or not human interstitial cells make their own choles-

terol, the question arises of how the cholesterol is supplied to the
mitochondria for side chain cleavage. A glance at an electron mi-
crograph like that shown in Fig. 4 immediately makes it obvious that
this involves a logistic problem How does cholesterol get from its
site of synthesis (or uptake from the plasma) to the mitochondria?
The smooth ER usually occupies a much larger percentage of the cyto-
plasmic volume than the mitochondria, and yet it appears that to en-
ter steroid synthesis the cholesterol must be funneled through the
mitochondria. Cholesterol is a normal constituent of biological
membranes and tends to be taken up by membranes, so it seems unlike-
ly that cholesterol could pass by simple diffusion across a field
of smooth ER. It has been suggested that the newly synthesized
cholesterol may be sequestered in the smooth ER, and that this or-
ganelle may constitute a vast reservoir of cholesterol for steroid
synthesis (32). One could imagine a membrane "circulation" that
would bring ER membranes into contact with mitochondria, although
we see little indication of this in static electron micrographs.
The smooth ER in most steroid-secreting cells is abundant enough so
the mitochondria must be constantly jostled by it, but the problem
still remains how to channel cholesterol from a large volume of
"tltllttGtlt
Jld IO(tBI .
f$frOlfnS
Fig. 6. Summary diagram of intracellular substrate movements dur-

ing steroid biosynthesis, according to the cell fraction-
ation literature. Cholesterol is either synthesized in
the smooth ER and adjacent cytoplasm, or taken up from
the plasma. Cholesterol then passes to the mitochondria
for cleavage of its side chain. Pregnenolone emerges from
the mitochondrion and is then transformed to androgens by
enzyme systems in the smooth ER. Used by permission from
A.K. Christensen and S/W. Gillim, in The Gonads, ed. Mc-
Kerns, K.W., Appleton-Century-Crofts, p. 415, 1969.
smooth endoplasmic reticulum to mitochondria that may constitute a

much smaller volume. The egress of pregnenolone from the mitochon-
dria presents similar problems, although it is possible that the
subsequent steps of androgen biosynthesis are not equally distri-
buted throughout the smooth ER.
If the cholesterol side chain cleavage system is present on the

inner membranes of human interstitial cell mitochondria, as seems to
be the case in mitochondria of hog adrenal (33), this means that
cholesterol must pass the outer membrane on its way in, and pregnen-
olone must do the same on its way out. The large globules some-
times seen in mitochondria of human interstitial cells, and whose
fine structure suggests they are lipid droplets, may well be pools
of either cholesterol or pregnenolone within the mitochondrion.
The function of the Golgi complex in steroid-secreting cells

is unknown. In cells that make protein for export, such as the pan-
creatic acinar cell (34), the newly-synthesized proteins are trans-
ported to the Golgi complex within the endoplasmic reticulum, and in

the Golgi complex are dehydrated to form secretory granules to be
secreted from the cell. It is unknown whether any similar function
takes place in steroid-secreting cells, although the Golgi complex
is often well developed in these cells and there is some evidence
that it hypertrophies during times of heightened steroid production
(35,36). Other suggestions as to possible functions for the Golgi
complex in steroid-secreting cells have been made, but are rather
tenuous at the present stage of our understanding.
Our meager knowledge of the pathways and time-course by which

substrates and products move through steroid-secreting cells under-
scores the need for a means of localizing steroids within cells at
the electron microscope level. This is extremely difficult because
the conventional techniques of electron microscopy that might permit
autoradiography involve solvents that would either remove or displace
the steroids from their original positions. It has been shown by
Stumpf and Roth (37) that stringent requirements must be met to lo-
calize steroids by autoradiography even at the light microscope le-
vel: 1) The tissue must be frozen without fixation, 2) The tissue
must be kept at a temperature of below _60 0 throughout cutting and
freeze-drying of frozen sections, and 3) no solvents of any kind can
contact the tissue throughout the whole process. Over the last few
years we have been developing a means of cutting frozen thin sections
that will satisfy these criteria at the electron microscope level
(38,39), and we are just beginning experiments that we hope will al-
low us to follow radioactive steroid substrates through the cells
after a pulse-label in vitro. If successful, this should allow us
to chart the path of substrates through the cell and also to show
whether or not these substrates accumulate in the mitochondria at
any time period or in the Golgi complex, and how long it takes the
materials to clear the cell. Through the use of specific inhibitors
it should be possible to show where individual steps occur.
ACKNOWLEDGEMENTS
I am indebted to Dr. Norris Finlayson, of the EI Camino Hospi-

tal, Mountain View, California, and to Drs. Gilbert LeBlanc, Lynn
Frary and Richard Weinstein of the Oakland Naval Hospital for their
cooperation in obtaining the material used in this study.
This work was supported by research grant HD-01512 from the

U.S. Public Health Service.
REFERENCES
1. Christensen, A.K., and Mason, N.R., Endocrinology, ~: 646,

1965.
2. Hall, P.F., Irby, D.C. and de Kretser, D.M., Endocrinology,
84: 488, 1969.
3. Baillie, A.H., Ferguson, M.M. and Hart, D.McK., Developments

in Steroid Histochemistry, Academic Press, London, 1966.
4. Christensen, A.K., J. Cell Biol., 26: 911, 1965.
5. Frank, A.L. and Christensen, A.K., J. Cell Biol., 36: 1, 1968.
6. Black, V.H. and Christensen", A.K., Amer. J. Anat.,124: 211,
1969. -
7. Christensen, A.K. and Gillim, S.W., In The Gonads, ed. McKerns,
K.W., Appleton-Century-Crofts, New York, p 415, 1969.
8. Christensen, A.K. and Fawcett, D.W., Amer. J. Anat., 118: 551,
1966. -
9. Christensen, A.K. and Fawcett, D.W., J. Biophys. Biochem. Cytol.,
~: 653, 1961.
10. Stieve, H., Mannliche Genitalorgane, Handbuch der mikroskopi-
schen Anatomie des Menschen,ed. v. Mollendorf, W., vol 7 (Harn-
und Geschlechtsapparat), part 2, Julius Springer, Berlin, 1930.
11. Ito, T. and Oinuma, S., Okajimas Folia anat. Jap., 18: 497,
1939. -
12. Sniffen, R.C., Arch. Pathol., 50: 259,1950.
13. Fawcett, D.W. and Burgos, M.H., Amer. J. Anat., 107: 245, 1960.
14. Yamada, E., Gunma Symposia on Endocrinology, 2: 1, 1965.
15. de Kretser, D.M., Z. Zellforsch, 80: 594, 1967.
16. de Kretser, D.M., Z. Zellforsch, 83: 344, 1967.
17. Nishimura, R. and Kondo, I., Urologia Internationalis, 18: 1,
1964.
18. Schmidt, F .C., Z. z"ellforsch, 63: 707, 1964.
19. Hatakeyama, S., Acta Path. Jap., 12: 155, 1965.
20. Nagano, T., Gunma Symposia on Endocrinology, 2: 19, 1965.
21. Pelliniemi, L.J. and Niemi, M., Z. Zellforsch~ 99: 507, 1969.
22. Gordon, G.B., Miller, L.R. and Bensch, K.G., Lab. Invest., 13:
152, 1964. -
23. Cervos-Navarro, J., Tonutti, E. and Bayer, J.M., Endokrinologie,
47: 23, 1964.
24. Dessolle, N.~ Comptes Rendus Acad. Sci., 264: 2018, 1967.
25. Sisson, J.K. and Fahrenbach, W.H., Amer. J. Anat., 121: 337,
1967.
26. Yasuzumi, G., Nakai, Y, Tsubo, I, Yasudo, M. and Sugioka, T.,
Exp. Cell Res., 45: 261, 1967.
27. Ito, S. and Karnovsky, M.J., J. Cell Biol., 39: 186A, 1968.
28. Bennett, H.S. and Luft, J.H., J. Biophys. Biochem. Cytol., 6:
113, 1959.
29. Luft, J.H., J. Biophys. Biochem. Cytol., ~: 409, 1961.
30. Richardson, K.C., Jarett, L. and Finke, E.H., Stain Technol.,
35: 313, 1960.
31. Osinchak, J., Z. Zellforsch, 72: 236, 1966.
32. Fawcett, D.W., In Intracellular Membraneous Structure, eds.
Seno, S. and Cowdry, W.V., Japan Soc. for Cellular Chern.,
Okayama, Japan (supplement for Symposia of the Society for
Cellular Ch£mistry, vol 14), p. 15, 1963.
33. Yago, N. and Ichii, S., J. Biochem., 65: 215, 1969.
34. Jamieson, J.D. and Palade, G.E., J. Cell Biol., 34: 577, 1967.
35. McDonald, D.M., Seiki, K., Prizant, M. and Goldfien, A.,

Endocrinology, 85: 236, 1969.
36. Rouiller, G. an~Schirnmer, B., Anat. Rec., 163: 342, 1969.
37. Stumpf, W.E. and Roth, L.J., J. Histochem. Cytochem., 14: 274,
1966.
38. Christensen, A.K., In Autoradiography of Diffusible Substances,
eds. Roth, L.J. and Stumpf, W.E., Academic Press, New York,
p. 349, 1969.
39. Christensen, A.K., Proceedings 7th International Congress of
Electron Microscopy (Grenoble, France), in press, 1970.
DISCUSSION
LUNENFELD: Dr. Vilar, did you ever find cellular bridges between
Sertoli and germinal cells, which could explain how information or
substances could be directly transferred from the Sertoli to the ger-
minal cells, or do you think that information has to be transferred
to germinal cells by diffusion?
VILAR: If you mean real intercellular bridges, not at all. But some
authors have described pinocytocis vesicles either in the cytoplasm
of the Sertoli or germinal cells close to the plasma membranes.
Whether or not this means any kind of transmission or transference
of substances, I do not know. The possibility has been mentioned.
HUDSON: I would like to ask a question that could be directed ei-

ther to Dr. Vilar or to Dr. Christensen Do they believe that the
secretions of the interstitial cells pass across the rather thick
lamina propria that surround the tubule and so nourish the germinal
cells?
VILAR: I cannot tell specifically about the transference of the se-

cretion of the interstitial cells, but we have studied the transfer-
ence into the seminiferous tubule of different substances. For in-
stance, if you put isolated seminiferous tubules in a medium con-
taining 20% of a mixture of sodium and potassium ferrocyanide, you
can see, after ten minutes, with the prussian blue reaction that the
ferrocyanide is able to pass through the structures of the tubular
wall to the germinal epithelium via the Sertoli cells. This expe-
rience could be repeated in vivo either with ferrocyanide or with
labelled albumin. For example, when the testis of a rat was injected
with rat albumin labelled with a fluorescent dye, the fluorescence
diffused well in the intertubular spaces, and by way of the germinal
epithelium passed into the lumen of the tubules. This means that
the albumin is able to pass through the complicated tubular wall.
When we used labelled globulins, they apparently did not go through
the tubular wall. When albumin labelled with iodine 131 was injec-
ted into the testis, the albumin was accumulated around the tubular
wall and in the lumen of the seminiferous tubule. The same phenom-
enon can be seen with the electron microscope, labelling the proteins
with ferritine. After 30 minutes ferritine labelled albumin appears
in pinocytosis vesicles and also in lysosomal bodies of the Sertoli
cells which apparently act as an extracellular compartment for the
germinal cells, (Vilar and Mancini, Acta Europea Fertilitatis, 1970,
in press).
CHRISTENSEN: The work of Waites and Setchell has shown that the semi-
niferous epithelium and its tunica propria may be rather selective
about what they allow to passo But I see no a priori reason why
steroids could not penetrate the cellular layers of the tunica pro-
pria, each of which is no thicker than a capillary endothelium.
MANCINI: I would like to mention that the data reported by Dr.

Vilar as regards the origi~ of seminiferous tubule fluid are in con-
trast with those of Shetler et ale (J. Physiol. 189: 63, 1967) who
have worked with rams. They collect a fluid at the level of the rete
testes and they assume that this fluid is not a consequence of dis-
placement or diffusion from the intertubular vessels to the seminif-
erous tubules, but a product of an active process inside the tubules.
However, they do not provide direct evidence of such a process.
A. STEINBERGER: I would like to ask Dr. Vilar, and perhaps Dr.

Christensen as well, whether they had the opportunity to examine tes-
ticular tissue under the conditions of gonadotropic stimulation, as
well as gonadotropic deprivation? And, if so, what were the ultra-
structural changes, particularly in the smooth endoplasmic retic-
ulum and mitochondria of the Sertoli cells and Leydig cells?
CHRISTENSEN: I have not studied the human testis after gonadotropin

stimulation, but the literature on mammals (including humans) indi-
cates that the changes are mostly quantitative rather than quali-
tative. In other words, the organelles become more abundant, but
their appearance remains much the same.
PAULSEN: I should like to expand Dr. A. Steinberger's point relative

to changes in the interstitial cells following gonadotropin stimu-
lat10n. Dr. de Kretser has shown that there is an increase in the
smooth endoplasmic reticulum, mitochondria and lysosomes. These
changes were quantitative rather than qualitative. With respect to
the crystalloids of Reinke, Dr. de Kretser has demonstrated that
these may be found in the nucleus as well as in the cytoplasm of the
interstitial cells. This is in contrast to our previous beliefs
using light microscopy studies that the crystalloids were present
only in the cytoplasm. If the crystalloids of Reinke are both in
the cytoplasm and nucleus of two cells of the body that secrete tes-
tosterone, i.e. the interstitial cells of the testis and the hilar
cells of the ovary, they may have some function.
CHRISTENSEN: I do not think anyone has much of an idea of what the

crystals are doing. Dr. Yamada (Gunma Symp. on Endocr. 2 : 1-17)
described intranuclear crystalloids in 1965, so it has been known
for some time that they and what appear to be their precursors occur
in the nucleus, as well as in the cytoplasm. However, it is well
known from early work (reviewed by Stieve 1930 and others) that the
crystals vary greatly in abundance from person to person, and it has
not been possible to correlate this variation with any functional
condition. Crystals seem to be all but lacking in some men of nor-
mal fertility. Also, they do not seem to occur in any other mammal
besides man, although the testes of all mammals presumably secrete
androgens. I therefore find it difficult to believe that they play
an essential role in androgen synthesis or secretion in human inter~
stitial cells.
SOLARI: I want to add to what has just been said about the Reinke
crystals. Dr. Vilar found that the optical diffraction pattern of
the crystals is almost hexagonal; this pattern was repeated in crys~
tals from several individuals.
JOHNSEN: We would think that the crystals of Reinke may have some
bearing on the functional state of the Leydig cells. We have found
that they are absent in Leydig cell hyperplasia, whatever the cause.
They are absent in Klinefelter's syndrome, in testicular feminiza-
tion, and it would appear that they are, in fact, absent in any
kind of poor testis.
PAULSEN'~ I should like to remind Dr. Johnsen that we presented data

at the Laurentian Hormone Conference in 1955 which showed that crys-
talloids of Reinke were evident in patients with Klinefelter's syn-
drome. Also Dr. de Kretser demonstrated these crystalloids in his
hypogonadotropic patients before as well as after gonadotropic ther-
apy.
DAVIS: Dr. Vilar, are smooth muscle cells present in the tunica pro-
pria surrounding the seminiferous tubule?
VILAR: I personally think they are not, but I know that Dr. Cler-
mont considered these cells myoepithelial cells related to contrac-
tion, and responsible for the movements of the seminiferous tubules.
CLERMONT: We have seen cells of the limiting membrane pulsating in

the rat, so there must be contractile elements in this location.
Ox. M. Ross working with the electron microscope also demonstrated
that similar cells in man have characteristics of smooth muscle fi-
bers, and were considered as contractile cells. Dr. Fawcett now
calls them myoid cells, I think. So he must be convinced that they
have characteristics of smooth muscle cells.
MANN: It is now fairly clear that testosterone is secreted by the

testes along three routes, namely, spermatic vein blood, lymph and
rete testis fluid. The concentrations seem to be of the same order
of magnitude. The concentration of testosterone is only a little low-
er in the lymph than in the spermatic vein blood; and only a little
lower in the rete testis than in the lymph. One would think that there
must be some common mechanism which regulates the movement of testos-
terone from the interstitial cells by all three avenues. Much has
been learned already, particularly from Dr. Mancini's work, on the
permeability of testicular blood capillaries, and from the recent
work by Fawcett on the testicular lymphatic sinusoidal vessels, that
makes it possible to visualize how testosterone enters the blood and
lymph. But what is the mechanism of testosterone entry into the semi-
niferous tubules? Dro Vilar showed us in his electron micrographs
how parts of the Sertoli cell protoplasm fill the spaces between the
interst~tial cells and the membranes around the seminiferous tubules.
Is it possible that the Sertoli cells are actually able to transport
testosterone?
CHRISTENSEN: It is always possible that the Sertoli cells may trans-

port steroids from the interstitial tissue into the seminiferous ep-
ithelium, but this would be difficult to prove. Certainly the de-
scriptive fine structure would not allow us to settle this question.
But if we are eventually successful in following steroids by auto-
radiography on frozen thin sections of fresh tissue, then it should
be possible to demonstrate whether or not labelled steroids are
taken up by the Sertoli cells.
MANCINI: We have been successful in injecting intratesticularly a

very small amount of labelled homologous albumino Provided there
was no disruption or rupture of seminiferous tubules by the needle,
this labelled albumin goes in large part into the intertubular lym-
phatic vessels and only a small part diffuses into the seminiferous
tubules through the Sertoli cell cytoplasm.
OHNO: Mitochondria have their own ribosome which is a bacterial

type. So translation by the mitochondrial ribosome is inhibited
by cycloheximide. Therefore, I would like to ask two questions:
first, are the ribosomes inside the mitochondria of the interstitial
cell? And, second, regarding messenger RNA for steroid enzymes, are
they translated by mitochondrial ribosome, and therefore sensitive to
chloramphenicol? Or does the translation actually occur outside the
mitochondria, and the enzymes get in afterwards?
CHRISTENSEN: It is a very interesting question about which I have

no information. We do see what could be interpreted as mitochon-
drial ribosomes inside the mitochondria. However, I know of no ev-
idence as to whether mitochondrial steroidogenic enzymes (20-, 22-,
11- and l8-hydroxylases) are coded by mitochondrial DNA.
LUNENFELD: Dr. Christensen, have you formed any working hypothesis

on the sequential transportation of such traits and products from
the different particulate fractions in interstitial tissue and on
the regulation of this transport?
CHRISTENSEN: As you know, cholesterol forms an important component

of biological membranes, and may be taken up by membranes·; conse-
quently, it seems unlikely that cholesterol could diffuse across an
extensive field of smooth endoplasmic reticulum. The membranes of
smooth reticulum may constitute an extensive reservoir of cholester-
ol, as has been suggested by Fawcett, and movements of the reticulum
may serve to bring fresh cholesterol to the mitochondria for side-
chain cleavage. One would look in electron micrographs for a very
close association between the smooth ER and mitochondria. This is
sometimes seen, but not with a frequency that would be expected if
this means of transport were a basic mechanism.
HISTOLOGY OF THE HUMAN TESTIS FROM NEONATAL PERIOD TO ADOLESCENCE
Oscar Vilar
Centro de Investigaciones sobre Reproduccion. Facultad

de Medicina. Buenos Aires, Argentina
The history of the testicular structures from birth to the end

of the adolescence period is to be discussed, from the histological
point of view, with the consideration that this gonad contains three
fundamental structures: a) the stroma, including albuginea, the in-
tertubular connective tissue and the seminiferous tubule wall; b)
the Leydig cells and c) the germinal epithelium. All these three
structures show definite morphological, histochemical and ultra-
structural changes which can be interrelated as well as correlated
with the age of the individual.
Few data are available regarding the human prepuberal testes.

Information given in the literature is often fragmentary since it is
very difficult to obtain material of different ages with a satisfac-
tory sequence. Our material was obtained either from biopsies ac-
cording to the technique of Charny and Meranze (1) or from fresh
necropsy material. We are deeply indebted to Dr. Sebastian Rosasco
and Dr. E. Carpaneto who kindly provided us several biopsies. Prep-
aration of the tissue for light microscopy, histochemistry and elec-
tron microscopy is reported elsewhere (2-6).
A. THE STROMAL CONNECTIVE TISSUE
Some authors have shown that the tunica albuginea has the typ-
ical appearance of juvenile connective tissue in the fetal gonad (7).
During the post-natal ages maturation of this tissue takes place (8).
This structure shows progressive changes from the new-born to the 8-
year-old child. During the first two months of life, the thickness
of the albuginea is about 300~. Its histological features match
those corresponding to a juvenile connective tissue arranged in
three layers: 1) the outer one, beneath the vaginal tunic, contains
95
96 o. VILAR
young fibroblasts, very fine collagen and reticulin fibers and few
blood vessels; 2) the middle layer shows a similar structure but
less collagen fibers; positive reactions for mucopolysaccharides
are observable in the interfibrillar spaces; 3) the inner layer is
richer in cells, mucopolysaccharides and blood vessels, but only
reticular fibers are seen. During the first year of age some chang-
es occur in the cells and fibers. Fibroblasts diminish in number
and undergo a slight maturation; collagen fibers become thicker,
more abundant and oriented; reticular fibers are scarcer, reactions
for mucopolysaccharides gradually disappear and vascularity dimin-
ishes. These changes take place mainly in the outer and middle
layers since the deeper one preserves most of the characteristics
of the juvenile connective tissue. From then, thinness of the albu-
ginea accentuates with age, with an increase in collagen fibers,
reaching its peak between six and eight years when the width of the
albuginea is close to 230 \-l.'
At early puberty the main feature in the albuginea is an in-

crease in the thickness (about 380 \-l.). The middle layer appears
edematous and numerous blood vessels and hypertrophied fibroblasts
are present; the collagen fibers and bundles are swollen with ra-
ther diffuse limits. The inner layer shows less striking changes
since a very loose network of reticulin fibers predominates and the
presence of edema is very difficult to assess.
In the following stages corresponding to the middle and late

phases of puberty, the thickness of the albuginea gradually increas-
es. Numerous collagen fibers are arranged in parallel, edema dis-
appears and fibroblastic cells now show the features of the so-
called fibrocytes. Reticular fibers as well as interfibr.illar muco-
polysaccharides progressively diminish. The most prominent feature
in the outer layer is the coalescence of collagen fibers in a s~m~
lar way as that seen in dense connective tissue. At the end of
puberty the albuginea is 450 l-.l. thick.
The previously described histological modifications also de-

velop in the connective tissue of the septae.
Intertubular spaces. From birth to the fourth year of age,

juvenile fibroblasts are predominant in these areas, but later de-
crease in number and become of a more mature type. The intertubu-
lar spaces are filled by a homogeneous mass of an amorphous plasma-
like substance which is eosinophilic and PAS positive. Reticulin
fibers are abundant in the testes of the new born, but diminish
progressively thereafter while the intertubular spaces widen and the
cellular population becomes more mature and scarcer.
At the beginning of puberty, proliferation of juvenile fibro-

blasts again takes place; congestion of blood vessels and an increase.
in the amorphous plasma-like intercellular substance are striking
changes. A very loose network of reticulin fibers persists.
At the end of puberal maturation there is a reduction in the

degree of hyperemia; mucopolysaccharides are no longer detected
histochemically and the whole cell population decreases; the juve-
nile fibroblasts almost disappear whereas the mature ones persist
although in reduced number. These cells have a slender fusiform
shape with the cytoplasm forming a narrow rim around the nucleus.
The nucleus is relatively large, showing scarce invaginations and
rather uniform karyoplasm, except for the periphery, where dense
granules attached to the nuclear membrane sharply define the bound-
aries of the nucleus. The Golgi apparatus is not particularly
outstanding and the endoplasmic reticulum is represented by a few
vesicles with ribosomes attached to them. Inclusion bodies, myelin-
like figures and lipids are also present. Light and electron micro-
scopical studies show changes in the wall of the seminiferous tu-
bules which proceed from birth to maturity.
The basement membrane is very thin in the newborn. It is PAS

positive, pale blue with Azan's stain; silver impregnation reveals
a delicate net of reticulin fibers. The electron microscope shows
that it is composed of two parts: the internal and the external.
The former, a continuous band about 1100 ~ thick (lamina basalis),
shows a rather homogeneous structure consisting of electron-dense
material with discernible and regularly distributed fine granules.
Surrounding the basement membrane is an external layer com-

posed of cells with multiple expansions. The endoplasmic reticulum
shows dilated cisternae studded with ribosomes which are also seen
free in the cytoplasmic matrix; fibrils are abundant in the cyto-
plasm; the intercellular spaces are occupied by dense bundles of
typical protofibrils and collagen fibers with characteristic peri-
odicity. No elastic fibers were seen.
This structure presents no striking changes until the begin-

ning of puberty. At this time the tubular wall increases in thick-
ness as well as in complexity. At early puberty the continuous in-
ternal lamina basalis preserves the same characteristics mentioned
above but pronounced changes soon develop. The following striking
features are seen: a) a lamellar structure separated by low elec-
tron-dense spaces replaces the original homogeneous band; b) in-
foldings of this lamina basalis develop and are preferentially lo-
cated in between the spermatogonia and Sertoli cells. Th~se pro-
cesses appear simple or rather labyrynthic, including some collagen
fibers and fine cytoplasmic branches of the underlying connective
cells (9). The depth of indentation of these infoldings reaches
one or two microns in some cases and can be seen with the light
microscope.
The external arrangement of collagen fibers and cells is more

98 O. VILAR
complex. Collagen fibers appear abundant and alternately strati-

fied with cells considered either fibrocytes (5,9,10) or smooth
muscle fibers (11,12). When techniques for elastic fibers are ap-
plied, they show that during the development of puberty, short and
fine elastic fibers appear. Nevertheless no evidence of them could
be shown with the electron microscope (5,13).
B. THE LEYDIG CELLS
These cells originate from fibroblasts during fetal life (14,

15), puberty or adult stage (16-18). Different types of cells,
with special distribution according to age were seen in the inter-
tubular spaces and were designated as type A, B, C and D fibroblast-
like cells, fetal and adult Leydig cells and degenerating cells.
These different types of cells represent progressive steps in the
transformation from the juvenile fibroblast of the mesenchyme (fi-
broblast A) to the fetal or adult Leydig cell. This differentia-
tion means changes in size and shape, as we:l as modifications in
histochemical reactions. Loss of basophils, appearance of lipids,
especially those grouped as 3-beta steroids, increased activity for
oxidative enzymes, lipases, esterases (4) and hydroxy-steroid dehy-
drogenases (19) are characteristics which appear progressively in
all these transitional stages. Leydig cells are fully differenti-
ated about the 12th week of fetal life (20). They are present in
the testes of the newborn (they reach a peak just before birth) and
several weeks after birth Leydig cells have disappeared completely.
No evidence of transformation of Leydig cells into other cell types
is found; on the contrary cell degeneration seems to be the rule.
The ultrastructural characteristics of the fully mature fetal Ley-
dig cells are much like their adult counterparts except that no
crystals of Reinke are visible. These cells show a number of histo-
chemically demonstrable enzyme activities and such a gradual devel-
opment of some of them has been demonstrated in the human(19,20).These
enzyme activities strongly suggest that steroids are actively pro-
duced in the fetal Leydig cells. Parallel with the depletion of
fetal Leydig cells at birth, a proliferation of less differentiated
fibroblasts begins and increases during infancy. Although the ju-
venile type of fibroblast predominates, there are also more mature
cells having a tendency to be located close to the tubular wall.
When puberty begins there is a decrease in the number of ju-

venile fibroblasts and an increase in more mature fibroblasts and
immature Leydig cells as well as transitional and degenerating el-
ements.
When puberty is complete, only mature fibroblasts and Leydig

cells remain. Leydig cells of the more mature type increase with
the development of puberty whilst degenerating forms decrease.
The differentiation of fibroblasts into mature Leydig cells is

characterized not only by modification in the light and electron

microscopical features (18) but also because a number of histochemi-
cal characteristics develop. Cytochemical methods give positive re-
actions for lipids, cholesterol, ascorbic acid, lipases, esterases,
leucylaminopeptidase, oxidative enzymes and for a wide variety of
hydroxysteroid dehydrogenases including 3-a, 6-~, 11-13, 16-~, 17-13,
20-a and 20-~ hydroxysteroid dehydrogenases. Most of these enzymes
can be fitted into a broad biochemical pattern (19,21) but several
have not been satisfactorily integrated. Most of these enzyme ac-
tivities are present not only in the mature Leydig cells but also
in the precursor fibroblast-like cells.
c. THE GERMINAL EPITHELIUM
There is at present general agreement sustaining the hypothe-

sis that the cell population of the germinal epithelium consists of
two different lines and that the early segregated cells (primordial
germ cells) in the yolk sac are the only source of the definite sex
cells while Sertoli cells are originated in the so-called supporting
cells derived from the coelomic epithelium. But this theory is
based on personal interpretation of morphological findings and it is
quite obvious that more work is needed to establish precisely what
happens (22-24). The available literature on the human developing
testis is incomplete (16,17,25). The migration of gonocytes in the
human embryo seems to have been demonstrated (26) but not their fu-
ture destiny.
Newborn to 2-month-old baby. The seminiferous epithelium of

the newborn human testis is composed of solid cords about 50 ~in di-
ameter where two main types of cells are easily differentiable. The
most numerous are the precursors of Sertoli cells (supporting or in-
different cells) which are cuboidal or columnar, forming a multilay-
er tissue which is attached to the basement membrane. These cells
have a round or elongated nucleus sometimes with deep indentations
surrounded by a thin rim of cytoplasm. One, two or more dense bod-
ies are recognized as possessing nucleolar structure without the
complexity seen in the adult Sertoli cell. Mitochondria are elon-
gated, small in diameter and possess a dense matrix and short crist-
ae oriented transversally. Polyribosomes, fibrils and few lysosomal
bodies are contained in the ground substance. The "intercellular
spaces" between neighboring supporting or immature Sertoli cells
narrow from time to time while there is accumulation of dense mate-
rial beside the plasma membrane. These are the junctional struc-
tures already described in the mouse (27). Either after stain with
polychrome blue or under the electron microscope (EM), the support-
ing cells were of two types: dark and light. The last ones are
more frequently seen. The significance of these two types of Sertoli
cells is uncertain. Our data differ from that reported by Johnsen
(28) in that we find differences not only in the nucleus but in the
cytoplasm as well.
100 o. VILAR
The second type of cells found at this age are those belonging
to the germinal line. Among them we can first describe the gonocyte
or primordial germ cell which is large and posesses basophilic cyto-
plasm. They are the most striking cells of the seminiferous cord.
They are sometimes found in the center of the cord, sometimes later-
ally, but always isolated from the basement membrane by the support-
ing cells. They show one or two large and quite characteristic nu-
cleoli, the rest of the nucleoplasm being clear and homogeneous (28).
In the cytoplasm sparse endoplasmic reticulum and scattered clusters
of free ribosomes are seen. In the light ground substance the mito-
chomdria are characteristically large, with well-defined cristae
without a definite pattern. They show a tendency to polarization.
The so-called A-bodies (29,30) are not present in our human testes,
but multi lamellar figures and residual bodies are not an unusual
finding.
A second type of germinal cell is already present in the two

month old testis. They are smaller, more irregular in shape and
show attachment to the basal membrane. These cells, which are still
large and have many similar nuclear and cytoplasmic characteristics,
could be identified with the "transitional" (29) or primitive type
A spermatogonia (31) of the rat. They have a higher number of more
spread mitochondria.
A halo or clear zone can appear in the nucleus of some of these

"transitional cells" and this change is also frequently associated
with vacuolization. While the proportion of cells which have estab-
lished contact with the lamina basalis increases, those unattached
cells diminish progressively.
Infancy and childhood. The following changes take place at

this stage. Supporting cells divide actively, especially during in-
fancy and then remain stationary until the beginning of puberty.
Consequently they appear more numerous than in the younger animals,
forming a palisade-like layer. The diameter of the tubules remains
unchanged. Big gonocyte cells occupy the center of the tubule and
become less numerous with age. Some of them appear "swollen" (20 IJ.)
but the size of the nucleus is not modified; branches of them are
interposed among the supporting cells. These cells are called hy-
pertrophic spermatogonia (2) and show in their structure evidence
of cellular degeneration. Some of them also appear as bi-or multi-
nucleated cells.
The so called intermediate cells divide, especially during child-

hood, and originate smaller round cells, attached to the basement
membrane, frequently shoWing intercellular bridges. Their cytoplasm
is light, with few flat membranous profiles, big mitochondria with
defined cristae and a moderately dense matrix. The nucleoli are
well-developed. These cells have been considered both in animal
and human as spermatogonia (29,32). By the observation of multiple
sections it is evident that numerous intermediate forms exist be-

tween the transitional large cell and the smaller spermatogonium.
It is also a fact that the number of these spermatogonial cells di-
minishes until the beginning of puberty. There are two possible
reasons for this fact: 1) the rate of mitosis of these cells is
very low; 2) the cells degenerate.
Puberty. At the onset of puberty the germinal epithelium is

lined with numerous palisade supporting cells, scarce spermatogonial
cells and occasional gonocyte-like cells (hypertrophic spermatogo-
nia). When puberal development starts changes in the germinal epi-
thelium take place very rapidly. With the fragmentary material
which it is possible to obtain in humans, the precise interpretation
of the facts is very difficult.
Spermatogonial proliferation is the first hi stological evidence

of puberty, and it proceeds generally after the clinical features of
the age are present. These are growth of pubic and axillary hair,
change in voice, etc. which are indicators of the presence of cir-
culating androgens. When spermatogonial proliferation is present,
all adult types can be seen simultaneously. We were not able to
find one single biopsy in which spermatogonial cells correspond in
toto to a unique generation of cells.
Almost at the same time the supporting cells start their trans-
formation into adult Sertoli cells. They stop dividing and increase
in size but remain attached to the basal membrane. The cytoplasm
contains an increased amount of tubular and vesicular agranular en-
doplasmic reticulum. Lipids become frequent inclusions. The nuclei
become more irregular and the nucleolus develops the characteristics
of the adult type with three components: amorphous, granular and
fibrillar. When the lumen becomes evident, the cytoplasm sends nu-
merous lateral expansions between the neighboring germinal cells
which also interdigitate with those of the Sertoli cells. Numerous
intermediate forms (immature Sertoli cells) are present at the be-
ginning of puberty. At this time, spermatocytes in the leptotene
and pachytene stages are present and are followed by spermatids
which complete their development at the end of puberty.
COMMENTS
From the results presented here it is evident that all the

structures of the testes show changes from birth to puberty, which
is the moment when the final and full organization of the gonad
takes place. The existence in the Leydig cells of a definite cycle
of growth and differentiation during the fetal life followed by a
period of involution after birth is evident. The connective tissue
structures show two developmental phases, the first one during in-
fancy which is characterized by a progressive fibrogenesis and par-
allel reduc'tion of the mucopolysaccharides.
102 o. VILAR
The definite organized maturation of the human testes starts at
puberty which cannot be matched with any chronological age. The pro-
liferation of the connective cells, precursors of the future Leydig
cells, is the first step and it is followed by the proliferation of
spermatogonial cells and maturation of Sertoli cells. Spermatogene-
sis will be complete thereafter.
Although it is generally believed that the genesis of Leydig

cells occurs later than spermatogenesis (16,32), it is also accepted
that the proliferating interstitial cells provide the source of an-
drogens throughout the puberal period (34) even without achieving
the characteristic morphology of the adult Leydig cells.
Also, apparently under the influence and control of hormones,

the connective tissue structures go through their second developmen-
tal phase which begins at the onset of puberty and extends through-
out adulthood.
The history of the germ cells after birth is more difficult to

explain. From the observation of many authors it is evident that
two different cellular lines exist. The Sertoli cell line passes
during infancy and childhood through a period of cellular prolifera-
tion. Then at puberty proliferation stops and they differentiate
into mature types. The germinal line is derived from the primary
gonocyte. During the course of infancy and childhood a series of
changes takes place which leads to the development of several inter-
mediate types of spermatogonial cells.
At this moment several questions arise: 1) Which one of these

cells is a "definite spermatogonium" and what are the features that
make it different from the preceding cells? (This question has been
posed before (35). 2) Could this definite spermatogonium be con-
sidered as the stem cell of the seminiferous tubules? 3) Which cell
of the seminiferous tubules of the adult testes could be identified
with thi s ce 11 ?
It is very difficult to answer this question properly. What

we have in hand is an incomplete set of static histological observa-
tions. Therefore the explanation of the dynamics of the renewal of
this cell population in the human testes is based on subjective in-
terpretations.
ACKNOWLEDGEMENTS
This work was supported by the Consejo Nacional de Investiga-

ciones Cientificas y Tecnicas, Argentina.
The technical assistance of Miss Lourdes Farinati is most ap-
preciated.
Fig. 1 Testis of a 2-month-old boy. Maraglas

embedded section stained with blue polychrome.
Numerous gonocytes and spermatogonial cells
(some vacuolated) are seen in the seminiferous
cords. x 760.
104 o. VIlAR
Fig. 2 Testis of a 2-month-old boy. Electron

microscopical picture of a "transitorial" cell
showing broad attachment to the basement mem-
brane. It is surrounded by light and dark sup-
porting cells. x 9200
Fig. 3 Testis of an ll-year-old boy. Only supporting cells can be

seen. x 2600.
106 O. VILAR
Fig. 4 Testis of an l1-year-old boy showing supporting cells (dark

and light). x 8500.
REFERENCES
1. Charny, C.W. and Meranze, D.M., Surge Gyn. and Obst., 74: 836,
1942.
2. Mancini, R.E., Narbaitz, R. and Lavieri, J.C., Anat. Rec.,
136: 477, 1960.
3. Vilar, 0., Perez del Cerro, M. and Mancini, R.E., Exp. Cell Res.,
]2: 158, 1962.
4. Mancini, R.E., Vilar, 0., Lavieri, J.C., Andrada, J.A. and
Heinrich, J.J., Amer. J. Anat., 112: 203, 1963.
5. Mancini, R.E., Vilar, 0., Perez del Cerro, M. and Lavieri, J.C.,
Acta Physiol. Lat. Amer., ~: 382, 1964.
6. Vilar, 0., Steinberger, A. and Steinberger, E., Z. Zellforsch.,
74: 529, 1966.
7. Gillman, J., Contributions to Embryology, Carnegie lnst., 32:
83, 1949.
8. Savoie, J.C. In Evolution morphologique et histochimique du
testicule che~le nourrisson et l'enfant, Masson et Cie, ed.
Paris, p.19, 1957.
9. Burgos, M., Rev. Soc. Argent Biol., 35: 309, 1959.
10. Vilar, 0., Steinberger, A. and Steinberger E., Z. Zellforsch,
~: 221, 1967.
11. Clermont, Y. Exp. Cell Res., 12: 438, 1958.
12. Ross, M.H. and Long, I.R., Science, 153: 1271, 1966.
13. de la Balze, F.A., Bur, G., Scarpa Smith C. and Irazu, J.,
J. Clin. Endocr., 14: 626, 1954.
14. Bouin, P. and Ancel, P., C.R. Soc. Biol. (Paris), 55: 1682,
1903.
15. Esaki, S., Z. Mikr Anat. Forsch, 12: 368, 1928.
16. Charny, C.W., Conston, A.S. and Meranze, D.M., Fertil. Steril.,
1: 461, 1952.
17. Sniffen, C., Ann. N.Y. Acad. Sci., 21: 609,1952.
18. Fawcett, D.W. and Burgos, M.H., Amer. J. Anat., 107: 245, 1960.
19. Baillie, A.H., Fergusson, M.M. and McK Hart, D., Developments
in Steroid Histochemistry, Academic Press, N.Y., 1966.
20. Niemi, M., Ikonen, M. and Hervonen, A., In Ciba Found. Colloq.
Endocr., Vol. 16. Endocrinology of the Testis., ease Wolsten-
holme, G.E. and O'Connor, M., J.A. Churchill Ltd. London, p.31,
1967.
21. Seilicovich, A. y A. Perez Lloret, Actas de la XIV Reunion Anual,
Soc. Arg. Inv. Clinica. Bariloche, Argentina 1969,p 97.
22. Gier, H.T. and Marion, G.B., Bioi. Reprod. Suppl., 1: p. 1,
1969.
23. Clermont, Y. and Perey, B., Amer. J. Anat., 100: 241, 1957.
24. Sapsford, C.S., Aust. J. Zool., 1..Q: 178, 1968.
25. Stieve, H., Der Hoden in Hand. der mikros. Anat. des menschem.,
W. von Mollendorf, 7(2). 1930.
26. Witschi, E., Contributions to Embryology, 32: 69, 1948.
27. Flickinger, Ch. and Fawcett, D.W., Anat. Rec:, 158: 207, 1967.
28. Pierce, G.B. and Beals, D.F., Cancer Res., 24: 1553, 1964.
108 O. VILAR
29. Franchi, L.L. and Mandl, A.M., J. Embryol. Exp. Morph., 12:
289, 1964.
30. Roosen-Runge, E. and Leik, L., Amer. J. Anat., 122: 275, 1968.
31. Steinberger, E., Steinberger, A. and Perloff, W.H., Anat.
Rec., 148: 581, 1964.
32. de Kretzer, D.M., Virchow Arch. (Zellpath) 1: 283, 1968.
33. Wershub, L.P., The Human Testis: A Clinical-Treatise. Charles
C. Thomas, Springfield, Illinois, 1962.
34. Segal, S.J. and Nelson, W.O., In Recent Progress in Endocrin-
ology of Reproduction, ed. Lloyd, C.W., Academic Press,
New York, 1959,pl07.
35. Roosen-Runge, E., In Charny, C.W., Conston, A.S. and Meranze,
D.R., Ann. N.Y. Acad. Sci., 55: 543, 1952.
DISCUSSION
JOHNSEN: I would like to comment upon Dr. Vilar's statement that

the first puberal step is proliferation of Leydig cell precursors
which are producing androgens. Dr. Vilar stated that these events
are followed by development of spermatogenesis. I am willing to
accept that we get Leydig cell precursors early and also that such
cells do contain steroid enzymes, but I am not willing to accept
that in man, Leydig cell function occurs before initiation of tu-
bular development. This in fact contradicts the clinical sequence
of events observed at puberty. Tanner has shown in his book "Growth
at Adolescence" (2nded., Oxford, 1962) that puberty starts with
growth of the testes which reflects tubular development. This con-
tinues for two years and when pubic hair appears, the testes are at
least half adult size and the development of the tubular system is
far advanced with near complete spermatogenesis. Development of
spermatogenesis before Leydig cell function is shown beautifully in
fertile eunuchs. In 1967, we (Johnsen, S.G.: Acta Endocr. (Kobenhavn)
1967, Supple 124, 17) showed some such cases. One was a completely
infantile pat~t of 20 years with no pubic hair and an androgen
level 15 times lower than in normal men but nevertheless with ad-
vanced tubular development.
Pubic hair is a very sensitive index and responds quickly as an in-

dicator of androgen level. Pubic hair can be seen in any pre-puberal
boy two or three weeks after administration of HCG. Presence of pu-
bic hair is no proof of Leydig cell function. But its absence proves
that there can be no more than an extremely small androgen produc-
tion. Hence, in man, advanced development of the tubules and the
germinal epithelium occurs first followed by Leydig cell function,
and not vice versa.
VILAR: Sorry, Dr. Johnsen, but I do not agree with you. The micro-
photograph I showed you which corresponds to the first stage of
puberty belonged to a child having pubic hair but no growth of the
testis. The histology showed, as you say, no histological puberal

development. I would like Dr. Lipsett to comment on the secretion
of androgens by the prepuberal and puberal testis.
LIPSETT: I think one problem is that Dr. Vilar is referring only to

morphology whereas Dr. Johnsen is referring not only to morphology
but also to a clinical effect of a hormone which we might be able
to relate to secretion. I think that one should try to utilize both
parameters: secretion and morphology, in trying to clarify such pro-
blems. Recently,working with Dr. Vanha Pertula from Finland, we
studied development of the testis in rats from 20 to 70 days of age.
In the immature rat there were a few Leydig cells present and, as
Dr. Mancini showed some time ago in the human, most of these cells
had perinuclear basophils and the rate of testosterone secretion
was very low. However, as the perinuclear basophils decreased, the
rate of testosterone secretion increased to adult levels. Here, at
least, p~rinuclear basophils seemed to be a morphologic feature which
correlated with Leydig cell function. In other similar situations,
I feel that some of the discrepancies may be resolved when we are
able to follow several parameters.
JOST: Dr. Vilar showed us very well develop.ed interstitial cells in

the two month old baby. These well developed Leydig cells later
disappear. Could Dr. Vilar tell us how the system of intracyto -
plasmic tubules,which characterize the adult type of Leydig cell,
disappear, or do all the differentiated Leydig cells degenerate
during childhood before being replaced by another set of Leydig
cells at puberty? My second question relates to the correlation
between the degenerating germ cells of the final stages of spermato-
genesis and the immaturity of the Leydig cells; could you comment
upon that?
VILAR: The disappearance of Leydig cells after birth is due to cell

degeneration, i.e. they do not involute again into fibroblast type
cells. The Leydig cells degenerate very soon after two months of
age, so in a very young child, say three years old, there are no
interstitial cells at all except fibroblasts.
MANN: I do not think we should exclude the possibility that in the

course of the sexual development of the male, there is an early
phase when testosterone is produced, followed by a second stage
when the testosterone production diminishes and a third stage when
testosterone production increases again. In support of that con-
cept two examples may be given. In the bull calf, at birth and
during the first few months of life, androstenedione but very little
testosterone is present in the testes. This high androstenedione/
testosterone ratio is maintained until five months of age. Only
later does testosterone become the dominant androgen. However, in
the bovine fetal testis, there is more testosterone than androstene-
dione. Thus, the ratio of testosterone to androstenedione is high
110 o. VILAR
before birth, low during the first few months of life, and then it
goes up again. The second example is derived from the work by Attal
(Endocrinology, 85: 280, 1969) who has shown that in the fetal testis
of the ram there-rs a high testosterone/androstenedione ratio early
in pregnancy, but during the later stage of fetal development that
ratio decreases. Towards the end of pregnancy it increases once
more.
BERGADA: With regard to pubic hair development after HCG treatment,

we have a wide experience in the treatment of cryptorchidism in
children, and we have never seen pubic hair development after HCG
treatment with doses of 1000 IU weekly given for 2 to 8 months. How-
ever, congestion and erections of the penis have been noted.
MANCINI: I fully agree with Dr. Lipsett that until we are able to
measure, in the same individual, different parameters which must be
integrated, we will not be able to reach a definite conclusion re-
garding puberal development. Second, it is difficult to accept what
Dr. Johnsen has said. Our studies showed that proliferation and
appearance of different types of spermatogonial cells and partial
maturation of Sertoli cells may develop in the absence of morpholog-
ically identifiable Leydig cells, but in the presence of fibroblast
like cells. Moreover, the beginning of meiosis may be present in
the absence of the Leydig cells, but completion of meiosis, develop-
ment of spermatids and spermatozoa and maturation of Sertoli cells
were observed when fully mature Leydig cells were present.
JOHNSEN: Dr. Mancini, I would say that I think we will have to ac-
cept the existence of the fertile eunuch. We have described several
such patients, and Faiman and Ryan from the Mayo Clinic have described
another patient with measurable FSH, but undetectable LH, no Leydig
cells, and undetectable testosterone. Dr. MacLeod showed in a hypo-
physectomized male that when FSH is administered, development of
spermatogenesis without Leydig cells, and without a rise in plasma
testosterone occurred. I think we will have to accept that advanced
development of the germinal epithelium in the absence of Leydig cells
can happen in normal puberty.
E. STEINBERGER: A few weeks before this meeting, we saw a patient

who may shed some light on the discussion here. He was a 6-year-
old boy with a Leydig cell tumor of one testicle. He showed some
penile enlargement and very early pubic hair. The tumor showed
typical microscopic characteristics of a Leydig cell tumor. The
tumor tissue avidly converted progesterone to testosterone in vitro.
The testicular tissue on the tumor side showed spermatogenesis up
to early spermatids while the contralateral testicle was small (nor-
mal for the age) and microscopic study revealed small tubules lined
with Sertoli cells and gonocytes, as expected in the testicle of a
6-year-old boy.
JIRASEK: According to my experience, using the 3-~hydroxysteroid

dehydrogenase reaction, I was able to demonstrate that the trans-
formation of the fibroblast-like cells into the Leydig cells con-
taining the 3-~hydroxysteroid dehydrogenase, precedes the onset of
spermatogenesis. I would like to inquire if the luminization of the
seminiferous tubules precedes or follows the differentiation of
Leydig cells.
VILAR: In our biopsies, the formation of the lumen apparently co-

incides with the appearance of few mature Leydig cells.
WENIGER: Dr. Vilar, in the human, do the Leydig cells originate in

the interstitial tissue? It has been shown by Dr. Sheib in the chick
embryo that the Leydig cells grow out from the germ cords.
VILAR: In the human testis, either in the fetal or in the puberal

period, progressive transformation is seen from very juvenile fibro-
blasts to fully mature Leydig cells.
CHAPTER "'
CYTOGENETICS OF SPERMATOGENESIS
MORPHOLOGICAL ASPECTS OF MEIOSIS AND THEIR GENETICAL SIGNIFICANCE
Susumu Ohno
Department of Biology, City of Hope Medical Center
Duarte, California
The ultimate aim of meiosis is to segregate four haploid cells

from a single diploid cell as a result of two successive cell divi-
sions. The prerequistte for this segregation of haploid sets is
the exact longitudinal pairing between two homologous chromosomes
which takes place during early phases of first meiosis. Further-
more, peculiar behavior of the centromere at first meiotic anaphase
is required for this segregation. In ordinary mitosis, the centro-
mere of each chromosome splits, and, during anaphase, two chroma-
tids move away from each other toward opposite division poles. As
a result, two diploid daughters are produced from one diploid cell.
In the case of first meiotic anaphase, on the other hand, two chro-
matids of each chromosome are held together by the unsplit centro-
mere, and the repulsive movement is between one whole chromosome
against the whole of its homologous counterpart. As a result, two
haploid daughters result from a diploid mother cell. Inasmuch as
each chromosome thus produced and included in the haploid nucleus
is already made of two chromatids, new DNA replication does not
take place between first and second meiosis. The second meiosis is
purely mechanical in that it merely serves to separate one chroma-
tid of each chromosome from the other.
Either coincident with or immediately following the longitudi-

nal pa~r~ng of homologous chromosomes (synapsis), an event of ge-
netic importance takes place. This is recombination which is due
to exchanges of pieces of chromatids between paired homologues.
The consequence of such exchange is seen in the form of visible
chiasmata between homologues at later phases of first meiosis, i.e.,
diplotene and diakinesis. Such intrachromosomal recombination be-
tween the maternally derived and paternally derived homologues, as
well as the random nature of segregation of haploid sets, insures
115
116 S.OHNO
that the finally produced gametes are of many different genetic

constitutions and represent mixtures of varying degrees of parental
genetic material.
Deviations from the above described norm of meiosis do occur

in certain exceptional species. For example, no crossing-over (re-
combination) takes place during meiosis of the male fruit fly (Dro-
sophila), and the male mealybug (hemipteran insect) transmits only
the maternally derived haploid set to its gametes. Human male mei-
osis, however, faithfully follows the norm of meiosis. There is
nothing exceptional about the process.
Various stages of male meiosis are described below with empha-

sis on underlying functional significance.
COMMITMENT TO MEIOSIS AND THE SYNTHETIC PHASE
Primary spermatocytes are direct derivatives of spermatogonia.

The mitotic cycle of spermatogonia must be interrupted at some point
and a commitment made by their daughter cells to go into meiosis
instead of returning to the mitotic cycle. According to Monesi (1)
in the laboratory mouse, such commitment is automatically made when
type A (dusty) spermatogonia differentiate into type B (crust-like).
The type B spermatogonia divide just once before becoming primary
spermatocytes.
It is of interest to note that certain abnormal sex chromosome

constitutions appear to preclude differentiation from spermatogonia
to spermatocytes. The fertile human XYY-male is expected to pro-
duce four types of gametes; normal X- and Y-bearing, as well as ab-
normal XY- and XY-bearing spermatozoa. As a result, he should sire
three times more sons than daughters and two-thirds of his sons
should again be abnormal; XYY and XXY. Yet the available records
show that only normal XY sons and XX daughters are found among his
progeny. Melnyk and his colleagues (2) have recently shown that
spermatogonia of the XYY man exhibit considerable mitotic instabi-
lity. Consequently, the normal XY-constitution is restored to some
of the spermatogonia, and only those with the normal constitution
differentiate into spermatocytes and complete meiosis. The situa-
tion found in the XYY man is reminiscent of that normally found in
a particular rodent species (Microtus oregoni). In this species,
the female is normally XO, and this remarkable feat is accomplished
by selective elimination of the X from spermatogonia of the XY male
(3).
After commitment, primary spermatocytes do not proceed immedi-

ately into first meiotic prophase. Necessary preparations for mei-
osis must first be made. The intervening period between the last
spermatogonial telophase and the leptotene stage of first meiotic
prophase is commonly referred to as the premeiotic resting stage
CYTOGENETICS OF SPERMATOGENESIS 117
which is an unjustifiable misnomer for this most metabolically ac-

tive stage. During this stage, primary spermatocytes replicate
their DNA. There appears to be a subtle but very important differ-
ence between premeiotic S-phase and premitotic S-phase. On plant
materials, Stern and Hotta (4) have shown that a small but signi-
ficant DNA component of the chromosomes, about 0.3% of the total
DNA, fails to replicate by the termination of the premeiotic S-
phase. Its replication is delayed to the zygotene stage of first
meiotic prophaseo This delayed DNA synthesis may be an essential
prerequisite for synapsis of homologous chromosomes which begins
at the zygotene stage.
Aside from DNA replication, premeiotic primary spermatocytes

engage in intense transcriptional and translational activities.
Analysis of the enzyme phenotype of mature sperm produced by genet-
ic heterozygotes suggests that most of the enzymes contained in ma-
ture spermatozoa were specified by the diploid nucleus of primary
spermatocytes and have been stored in the cytoplasm either in the
form of messenger RNA or already made enzyme molecules. The hap-
loid nucleus of secondary spermatocytes does not appear to direct
spermiogenesis. It has been shown that the elimination of a number
of chromosomes from secondary spermatocytes does not affect subse-
quent spermiogenesis. Functional and fertile nullosomic spermato-
zoa are produced. (5).
SYNAPTIC PHASES
When chromosomes are seen for the first time as individual en-
tities inside the nuclear membrane, primary spermatocytes are said
to have entered the "leptotene" stage of first meiotic prophase.
By definition, at this earliest prophase stage, all the chromosomes
(46 in the case of man) should be distributed at random with no hint
of homologous pairing. Thus, there is no foolproof criterion which
enables us to distinguish the leptotene nucleus of primary sperma-
tocytes from the early prophase nucleus of spermatogonia. It is
probable that the "leptotene" stage, as such, does not exist in a
strict sense in that necessary preparations for homologous pairing
have already been made in the interphase-like nucleus of the pre-
meiotic stage.
Longitudinal pa1r1ng between two homologous chromosomes, which

is accomplished in primary spermatocytes, is very exact as witness-
ed by the formation of a quadrivalent by an individual heterozygous
for a reciprocal translocation. When one chromatid of a metaphase
chromosome is completely stretched out, it should become a single
strand of DNA double helix hundreds of microns long. The unique-
ness of an individual chromosome no doubt lies in its DNA base se-
quence. Proteins associated with DNA are too short in variety to
be homologue specific. It then follows that the exact pairing be-
tween two homologues must be based on the near identity of their
118 S.OHNO
DNA base sequences. However, unless many parts of two homologous

chromosomes are first placed in close vicinity of each other either
by chance or by design, such attraction cannot become sufficiently
powerful to accomplish exact longitudinal pairing of homologues.
Viewed in this light, "bouquet" orientation of individual

chromosomes seen at the "zygotene" stage of the first meiotic pro-
phase gains added significance. Furthermore, if the "leptotene"
stage does not exist, as I am inclined to believe, "zygotene" marks
the beginning of the first meiotic prophase.
At zygotene, individual chromosomes are still in an extremely

fine, thread-like state, but both ends (telomeres) of all the chro-
mosomes are oriented toward one point in the nuclear membrane; thus
giving the appearance of a "bouquet". Because of this orientation,
the chromosomes of like size tend to be placed in close proximity
to each other along their length. Indeed, homologous pairing has
already begun in parts wherever two homologous segments happen to
lie near each other. What is the force which induces partial homo-
logous pairing at zygotene? DNA replication, transcription of DNA
base sequence to messenger RNA base sequence, and, in fact, life
itself, owes its existence to the inherent complementality that
exists between two purine and pyrimidine base pairs; adenine and
thymine as well as guanine and cytosine. Therefore, the attraction
to form a double helix exists only between a single uncoupled DNA
strand and its complementary single strand; there is no attraction
between two identical strands of DNA double helices. If the DNA
replication has already been 10oto completed during the premeiotic
interphase-like stage, each zygotene chromosome, being made of two
complete DNA double helices, is already equivalent to two sister
chromatids. No attraction is left between two satisfied homologous
chromosomes. However, it should be recalled that Stern and Hotta
(4) have shown that about 0.3% of the DNA fails to replicate. If
parts of a zygotene chromosome contain uncoupled single DNA strands
not yet replicated, such single strands can form stable double he-
lices with their complementary counterparts contained in its homo-
logue. In fact, this appears to be the most likely mechanism by
which the exact longitudinal pairing between two homologous chro-
mosomes is initiated during zygotene. Subsequent completion of
longitudinal pairing appears to be aided by a sort of protein ce-
ment formed between homologous chromosomes. The presence of such
protein cement was first noted by Moses (6) with the use of an elec-
tron microscope and termed "synaptonemal complex".
When the process of longitudinal pairing is completed, pri-

mary spermatocytes are said to h~ve entered the pachytene stage.
Thus, the human pachytene nucleus contains 22 autosomal bivalents
and the XY-bivalent (Fig. 1). An autosomal bivalent of pachytene
has the appearance of a rather thick thread showing a more or less
distinct transverse banding pattern. The position of a centromere
Figures 1 and 2 are pachytene and diakinesis or first meiotic

m~taphase figures obtained from testicular biopsy of a normal human
male subject. The air-dry preparation is shown courtesy of Dr.
John Melnyk of this institute.
-.-
.'
Fig. 1. Pachytene. An intensely stained sex vesicle enclosing
the X and the Y is seen at 12:30 o'clock.
Fig. 2. Diakinesis or 1st meiotic metaphase. 22 autosomal biva-

lents and the X and the Y in typical end-to-end association are
seen. The XY-bivalent is at 6 o'clock inside.
120 S.OHNO
is, as a rule, marked by an adjacent knob of heavily stained heter-

ochromatin (centromeric heterochromatin). In sharp contrast to the
thread-like autosomal bivalents, the XY-bivalent exists as a con-
spicuous mass of heterochromatin embedded in the sex vesicle (7)
which is rich in RNA (8).
Unlike autosomes, which always exist as a pair in diploid or-

ganisms, the X and the Y of mammalian males are the determiners of
the opposite sexes, and there is very little homology between the
X and the Y. The Y carries none of the many structural genes un-
related to the act of sex determination carried by the X. Although
an extremely short arm of the acrocentric human Y may be homologous
to the distal end of the short arm of the metacentric X (9,10) and
genuine longitudinal pairing appears to occur within the sex ves-
icle between the extremely short homologous segments of the X and
the Y (11), the presence of such a short homologous segment appears
to offer insufficient attraction to bring the X and the Y together
during zygotene and pachytene. Yet, the segregation of the X a-
gainst the Y at the end of first meiosis can be assured only if the
largely nonhomologous X and Y form a respectable bivalent. The
formation of a peculiar sex vesicle in male pachytene nuclei appar-
ently reflects the special device mammalian species invented to
bring the X and the Y together. Heteropyknosis (condensation) may
be the means employed to deemphasize the great difference that ex-
ists between the DNA base sequences of the X and the Y. Indeed,
two X-chromosomes during female meiosis b have in exactly the same
manner as any other autosomal pairs (12). From the end of last
spermatogonial telophase to the end of the pachytene stage is a
period of long duration. From what is known in the mouse(l) it may
be inferred that in man nearly 10 days would be taken up by this
period.
FROM DIPLOTENE TO THE END OF FIRST MEIOSIS
Genetic recombination (crossing-over) which is due to exchange

of parts of chromatids between paired homologous chromosomes is
most likely to occur either during the process of homologous pair-
ing (zygotene) or immediately after the completion of longitudinal
pairing (pachytene). First, breaks must occur at corresponding
positions of the DNA strands of paired homologues;subsequent re-
joining of broken ends by a repair enzyme may unite the broken end
of one with that of its homologue thus accomplishing an exchange.
Although we have no way of catching the process of recombination
"in the act" under the microscope, the consequence of crossing-
over becomes evident at the stage following pachytene which is re-
ferred to as the "diplotene" stage.
At the beginning of "diplotene", homologous autosomes which in

pachytene were paired side-by-side along their entire length fall
apart from each other except at certain places where chiasmata have
been formed. As expected, more chiasmata are seen on larger biva-

lents. The largest autosomal bivalent seen in the human diplotene
nucleus may show as many as five chiasmata; three interstitial and
two terminal. Although diplotene chiasmata most likely represent
the consequence of crossing-over events which took place in the
preceding stages of first meiotic prophase, the position of a chi-
asma does not appear to coincide with the precise point of an ac-
tual exchangeo Our genetic knowledge tells us that crossing-over
occurs essentially at random along the entire length of "an autosome.
In sharp contrast, diplotene chiasmata tend to be positioned with
remarkable regularity in a given autosomal bivalent. Furthermore,
the mean number of chiasmata exhibited by a large autosomal bivalent
appears to be too small to equal the number of genetic crossing-
overs expected of so large a genetic unit (linkage group). If fall-
ing apart of pachytene homologues is due to the repulsion between
pairs of centers located along the length of a bivalent, all actual
points of crossing-over contained in the region between two adja-
cent repulsive centers would move away toward the middle as a result
of the repulsive movement and be seen as a single chiasma. This
would account for the relatively small number of chiasmata actually
seen and also for the tendency of chiasmata to occur at fixed sites
of a given bivalent.
Soon diplotene bivalents begin rapid condensation, and primary

spermatocytes reach the "diakinesis" stage, which marks the end of
the first meiotic prophase. This progressive condensation is ac-
companied by further terminalization of chiasmata. Two or three
interstitial chiasmata of diplotene can now be seen as a single
terminal chiasma. Thus, many of the autosomal bivalents of dia-
kinesis assume an O-shape (two terminal chiasmata) and some small
ones an X-shape (a single terminal chiasma) as shown in Fig.2. It
is my opinion that counting the number of chiasmata on diakinesis
bivalents is a rather futile endeavor, for it will tell us about the
degree of chiasma terminalization, rather than the frequency of
genetic crOSSing-over. At diakinesis, the X and the Yare clearly
seen in a characteristic end-to-end association. Chen and Falek
(10) have recently presented rather convincing evidence that this
association in man is between the extreme short arm of the acro-
centric Y and the terminal end of the short arm of the X.
Diakinesis is the best stage in which to count the number of

bivalents and detect the presence of a quadrivalent expected in an
individual heterozygous for a translocation. Some of the trans-
locations which cannot b-a detected by examination of somatic dip-
10id metaphase figures can be detected at diakinesis.
At the end of diakinesis, disappearance of the nuclear mem-

brane is followed by the alignment of already condensed bivalents on
the equatorial plate of the first meiotic spindle. The difference
between diakinesis and first meiotic metaphase is in the alignment
122 S.OHNO
and not in the appearance of individual bivalents.
At anaphase of ordinary mitosis, the centromere of each chro-

mosome splits, and two sister chromatids move away from each other
toward the opposite division poles. Thus, in man, each daughter
cell receives 46 chromatids as a result of mitosis. In the case
of first meiotic anaphase, on the other hand, two sister chromatids
of each chromosome are held together by the unsplit centromere, and
the repulsive movement is between one whole chromosome against the
whole of its homologous counterpart. Thus, as a result of first
meiosis, each daughter cell receives 22 autosomes and the X or the
Y. The segregation of two haploid sets from a diploid set is there-
fore accomplished.
SECOND MEIOSIS
The stage between the first meiotic telophase and second mei-
otic prophase is called "interkinesis". On a pair of daughter
secondary spermatocyte nuclei, the condensed X can clearly be seen
on one and the condensed Y on the other. Interkinesis is a rather
inert stage. There is no need for S-phase (DNA replication), since
each chromosome in the haploid nucleus is already made of two chro-
matids. Furthermore, most of the enzymes and proteins needed for
the ensuing stage of spermiogenesis have already been synthesized
and stored in their cytoplasm by the mother cell during the pre-
meiotic stage. Thus, after the interkinesis stage which is of
short duration, secondary spermatocytes merely go through the mo-
tion of haploid mitosis, which is the true nature of second meiosiso
At the end, each daughter cell receives 23 chromatids including an
X-chromHtid or a Y-chromatid. It can be said that the spermatozoa
are in the G-1 stage of haploid interphase.
A major cause of known types of chromosome anomalies should

be sought in meiotic errors occurring either during spermatogenesis
or oogenesis. Analysis of second meiotic metaphase figures is the
only way to accurately estimate the frequency of errors committed by
first meiosis. Unfortunately, the morphology of second meiotic
metaphase chromosomes is such that they are not readily amenable
to analysis. Sister chromatids of each chromosome which are loosely
held together at the centromere do not attain the fully condensed
state; instead they remain stringy in appearance.
ACKNOWLEDGEMENTS
This work was supported in part by a grant (CA 05138) from

the National Cancer Institute, U.S. Public Health Service.
REFERENCES
1. Monesi, V., Amer.Zool., 1: 374, 1961.

2. Melnyk, J., Thompson, H., Rucci, A.J., Vanasek, F., and Hayes,
S., Lancet. ii: 797, 1969.
3. Ohno, S., Jainchill, J., and Stenius, C., Cytogenetics, 2: 232,
1963.
4. Stern, H., and Hotta, Y., Genetics, ~: 27, 1969.
5. Lindsley, D.L., and Grell, E.H., Genetics, ~: 69, 1969.
6. Moses, P.B., Ann. Rev. Genet., ~: 364, 1968.
7. Sachs, L., Ann. Eugen.,..!§: 255, 1954.
8. Ohno, S., Kaplan, W.D., and Kinosita, R., Exptl. Cell. Res.,
11.: 520, 1956.
9. Ferguson-Smith, M.A., Lancet, ii: 475, 1966.
10. Chen, A.T.L., and Falek, A., Science, 166: 1008, 1969.
11. Solari, A.J., Genetics, 61: 113, 1969-.-
12. Ohno, S., Klinger, H.P., and Atkin, N.B., Cytogenetics, 1: 42,
1962.
DISCUSSION
CLERMONT: Dr. Ohno, you said that it was possible to prevent meiosis
experimentally. Could you explain how this could be done?
OHNO: This of cOUrse was done on a plant, by Hatta and Stern, where
the cells in the same step of meiosis can be recovered in mass. By
increasing the temperature, nucleic acid synthesis is completed and
the cells go into mitosis instead of meiosis.
E. STEINBERGER: I would like to emphasize the fact that most of

these studies have been performed in a plant (lily). If I recall
correctly these plant cells are capable of going through meiosis and
completing the reduction division in vitro with no difficulty. In cul-
tures of mammalian testis meiotic division fails to take place suggest-
ing that conditions necessary for completion of meiotic process in a
plant spermatocyte must differ greatly from those which occur in the
mammalian testis.
OHNO: I have to disagree with you on this point. What we learned

from comparing the sequences of direct gene products is the amazingly
conservative nature of natural selection. Really important things
have never really changed in evolution, the examples being transfer
RNA and histone IV. Exchange of genetical information is based on
the complementality of base sequences. Such a process must have begun
at the very beginning of life. For this reason, I believe that the
principle of meiotic process must be the same in all living creatures.
E. STEINBERGER: I want to apologize to Dr. Ohno. I did not mean to

imply that the basic mechanisms involved in the lily are not neces-
sarily similar to those of higher species. However, there are also
basic differences. For example, bacteria may reproduce sexually, but
still we cannot compare this with mammalian sexual reproduction, par-
ticularly with regard to the various regulatory mechanisms which in
124 S.OHNO
higher species are obviously highly developed.
FORD: Dr. Ohno, I was a little surprised at your opposition to the

chiasma-type hypothesis, that is the assumption of a direct 1 to 1
relationship between chiasmata observed in the spermatocyte and ge-
netic cross-over. I would like to ask how you explain the classic
experiments of Beadle, for example, which showed a very close cor-
relation? We have unpublished information in the mouse making use
of overlapping translocations, and correlating our chiasma counts
with the position of the break points on the genetic map. We find
extremely good agreement.
OHNO: Well, I believe that one has to distinguish the intergenic

crossing-over from intragenic crossing-over and what has tradition-
ally been measured is the crossing over between the genes. There
is a reason to explain the fact that intergenic-crossing over should
not happen as often as intragenic crossing-over. This is because
if you think of a chromosome as a single strand of DNA double helix,
there has to be a number of signals which say "here is the beginning
of one structural gene, and there is a beginning of the other."
Such a signal might be a promoter base sequence which is inserted
between structural gene base sequences. Thus, evolution has to re-
duce the number of intergenic crossings-over in order to avoid too
many unequal crossings-over which would be lethal. For this reason
I think that there are far more frequent intragenic crossings-over
than intergenic crossings-over.
FORD: This may be, but the fact remains that there is extremely
good agreement between cytologically observed chiasmata and genet-
ically identified cross-overs. One can estimate the total genomic
length of the mouse chromosomes and estimate it fairly accurately.
I have been watching my predictions over a number of years, and at
one stage the genetic map of one of the mouse chromosomes was much
longer than I had anticipated any mouse chromosome could be. True
enough, two or three years ago it was found that there were two
separate linkage groups involved and now the map again is consistent
with chiasma counts.
OHNO: Within the H2 locus of the mouse the intragenic crossing-

over occurs at the rate of 3 x 10-3 • And the rate of crossing-over
between the H2 locus and the next TL-locus is, I believe, consider-
ably less.
FORD: I do not deny, of course, what you are saying. We must find
some means of reconciling that with the counts of chiasmata in di-
plotene and diakinesis.
SOLARI: I think that there is a lot of evidence about the relation-

ship between chiasmata and genetic crossing-over. For instance,
Taylor's experiments strongly support these relationships, but this
refers to intergenic crossing-over. The intragenic crossing-over

is a rather open problem and I find Dr. Ohno's suggestion a very
interesting one. I think that there may be two mechanisms for intra-
genic and intergenic crossing-over during meiosis and this would be
perfectly in agreement with some models of the molecular structure
of the synaptonemal complex.
ULTRASTRUCTURE AND HISTOCHEMISTRY OF THE NUCLEUS DURING MALE MEIOTIC
PROPHASE
Alberto J. Solari and L. L. Tres

de Medicina - Paraguay 2155, Buenos Aires, Argentina
The life of the human first spermatocyte lasts about 23 days (1).
During this period predetermination of meiosis, synapsis. chiasmata
formation and segregation of homologues occurs; it is also possible
that some abnormal phenomena such as meiotic non-disjunction occur
during this time.
These processes are accompanied by ultrastructural and histo-

chemical changes that will be reviewed here, with attention focused
on human cells.
THE PROPHASE STAGES
Preleptotene. If incomplete replication of DNA is a prerequi-

site for meiosis, as it seems to be in the lily (2), then meiosis
determination begins in the preleptotene stage, during the synthesis
phase of DNA (S-phase). In the lily, meiosis determination is com-
pleted later, in the post-synthetic (G-2) phase (2)~
However, from the ultrastructural viewpoint, we do not have any

morphological basis for identification of the human preleptotene
stage. The suspected cells can be considered either as leptotene,
because of the presence of chromatin threads, or they cannot be
distinguished from type B spermatogonia. Recent work by Lima-de-
Faria et al. (3) attempts to explain this. These authors found
that DNA synthesis continues in leptotene and leptotene-zygotene sper-
matocytes in culture. It could be that true preleptotene is a short-
lived stage which rapidly evolves to nuclear condensation which is
characteristic of the leptotene stage, while continuing with the
functions usually associated with the preleptotene, the main one
being the premeiotic synthesis of DNA. The peculiar transformation
127
128 A. J. SOlARI AND L. L. TRES
of mitochondria from the regular morphology observed in the sper-

matogonia to the swollen aspect seen in spermatocytes (produced by
the dilation of the intracrista1 space) is observed only in cells
which have lost all connection with the base membrane by the inter-
posing cytoplasm of Serto1i cells. As similar mitochondrial changes
have been related with functional stages (4), a relatively abrupt
metabolic change may be associated with pre1eptotene and meiosis
determination.
Leptotene. Leptotene nuclei are characterized by the condensa-

tion of chromatin, which forms many thread-like patches (Fig. 1).
Axial condensations or single cores 300 R wide may be observed in
some cells which are more advanced in development. A small granular,
horseshoe shaped nucleolus is present in leptotene and in zygotene.
Zygotene. The beginning of zygotene is marked by the appearance

of synaptonemal complexes (SC) inside the nucleus. The number of
SC increases up to the end of zygotene. The sex pair ("sex vesicle")
appears during zygotene as a homogeneous zone of chromatin fibrils
containing single cores.
Pachytene. This stage occurs most frequently in the sections

(Fig. 2). Early pachytene nuclei contain long stretches of SC asso-
ciated with densely packed chromatin fibrils. During mid and late
pachytene the chromatin becomes gradually less packed with the ex-
ception of the "basal knobs" (5) and the sex pair (6). Nucleoli
grow during mid-pachytene and attain their maximum volume at late
pachytene (Fig. 2), when the whole nucleus is swollen and the sex
pair appears as a prominent dense mass attached to the nuclear en-
velope (Fig. 2).
Diplotene. The diplotene stage is marked by the virtual dis-

appearance of SC inside the nucleus. Single cores are observed in
the loose masses of chromatin fibrils. A special tripartite and
convoluted structure has been observed in one human bivalent during
prometaphase and metaphase (7).
THE SYNAPTONEMAL COMPLEX
This is the most distinctive structure found in meiotic nuclei,

(8,9) and its occurrence during synaptic stages in meiosis seems to
be universal (10). The SC is a plane, tripartite structure composed
of two lateral elements flanking a third, central element (Fig. 3)
The SC lies in the axis of each bivalent during zygotene and pachytene,
that is, it is located at the place where the two homo10gues face
each other. The lateral elements of the SC are rods denser and
wider (about 400 i wide) than the central element which is about
150 R wide. In frontal sections the centr~l element appears to be
formed of three longitudinal filaments 40 A thick. Transverse
Fig. 1. Human leptotene spermatocyte. Chromatin threads with

axial condensations (arrows). N, nucleolus. X 24,500
(without reduction).
130 A. J. SOlARI AND L. L. TRES
Fig. 2. Human pachytene spermatocyte. XY, XY pair; N, nucleolus.

Synaptonemal complex (arrow). X 20,000 (without reduction).
Fig. 3. High magnification micrograph of a human pachytene nucleus.

LE, lateral element of a synaptonemal complex. CE, central
element. M, dilated intracristal space of mitochondria.
X 152,000.
fibri.ls cross from the lateral to the central element. Each later-
al element is surrounded by chromatin fibrils of one homologue,
except at the medial side. The SC is present all along the syn-
apsed homologues (11) and its two ends are attached to the nuclear
membrane (12). Histochemical tests have shown that DNA is not pres-
ent except in the narrow inner layer of the lateral elements (13).
Thus, the main components of the SC must be different from those
of chromatin fibers. Sheridan and Barrnett (14) have shown that
in the lily the lateral elements contain a basic component, prob-
ably a lysine-rich protein, which is stained with alcoholic phos-
132 A. J. SOLARI AND L. L. TRES
photungstic acid, and which is probably different from common his-

tones. Thus, the main mass of the SC is formed by the deposition
of a special substance on specific parts of zygotene chromosomes
(9). However, small amounts of DNA can exist in the transverse
fibrils (8) and perhaps in the central element.
The formation of the SC has been studied with serial sections

in a few cases only (11,15). These show that leptotene chromosomes
bearing single cores pair to form the SC. Thus, each lateral ele-
ment pertains to each homologue, while the central element seems to
result from the condensation of the transvers~ fibrils. The fate
of the SC at diplotene varies with different species. In the human
the SC disappears while the homologues come apart from each other
forming chromatin patches which bear single cores, that is, lateral
elements remain, at least for a time, joined to single chromosomes.
Recent evidence indicates that at least two biochemical events

may be related to the SC formation. In the lily, 0.3% of the total
DNA does not replicate during the premeiotic S-phase, but it is re-
plicated specifically during zygotene (2). If this delayed DNA
synthesis is inhibited, the SC formation is also inhibited (16)-.
Certain non-histone proteins are synthesized at zygotene; if this
synthesis is inhibited with cycloheximide the arrest of nuclei at
zygotene results.
It has been assumed that the presence of the SC expresses the

synapsis of homologous segments of chromosomes (10) and that it is
a prerequisite for chiasma formation (8). A number of observations
contradicting these assumptions have been recorded (17,9) but they
generally refer to stages different from zygotene and early pachytene.
Thus, zygotene synapsis is accompanied by formation of the se.
THE SEX PAIR IN HUMAN SPERMATOCYTES
The XY pair is formed at zygotene as a dense body often called

the "sex vesicle." This heteropyknotic sex pair is formed by chro-
matin fibrils with a special degree of packing (6,18,19). This
special condensation of the human sex chromosomes is present in the
leptotene stage, but it is absent in the human spermatogonia (20).
The XY pair forms an oblong mass associated with the nuclear

envelope (Fig. 2). Inside the chromatin mass, 400 i wide sections
of single cores are observed. The pattern of the cores in the human
XY pair has been studied with serial sections and model building of
whole, reconstituted, XY pairs (21). Two cores are present at all
times in the XY pair, and they have constant features (Figs. 4 and
5). They are the long and the short cores, corresponding to the
axes of the X and the Y chromosomes respectively. Both cores begin
and end on the nuclear envelope, but they have a common end (Fig. 5)
Fig. 4. Photograph of a partially reconstructed XY pair from human

spermatocyte. L, long core. SC, synaptonemal complex formed
at the common end of both cores.
formed by the union of the two c,: )res. In that region a synaptonemal
complex is formed which measures 0.4 to 0.8 microns in length. This
SC is formed in zygotene and early pachytene and it shortens and be-
comes unrecognizable during mid and late pachytene. A similar pattern
of the cores has been observed in the mouse and in the raot (22).
The remaining length of the cores, especially the long core, is

convoluted at some places; this coiling is exaggerated during mid
and late pachytene, forming the ring bodies (6). As the centromeres
have not been identified with the electron microscope, no evidence
is provided that is associated with one of the two extremes of each
sex chromosome. However, from the optical picture of the human XY
pair, it has been assumed that the short arm of the X chromosome is
involved in the end-to-end joining with the Y (23). From the data
cited above, which agree with those concerning other mammals, it is
concluded that in the human there is partial synapsis between the
X and the Y chromosomes. Pairing is restricted to a small part of
both chromosomes, involving the tip of the X chromosome. Chiasmata
and cross.ing-over might be possible in that region, and the latter
would be reflected in the partial X-linkage of genes located in that
segment.
134 A. J. SOLARI AND l. l. TRES
pX
Fig. 5. Schematic drawing of a human spermatocyte during pachytene.

cs, synaptonemal complex ending on a basal knob associated
with the nucleolus (N). pXY, XY pair; ec, common end; 1,
long core; 2, short core. C, nuclear envelope.
NUCLEOLI OF HUMAN SPERMATOCYTES
It has been proved (24,6) that in human spermatocytes, nucleoli

are not regularly associated with the sex pair, but are regularly
associated with "basal knobs". These basal knobs are centered by
one or more SC that end immediately on the nuclear envelope (Fig. 5).
Thus, nucleoli are associated with the heteropyknotic tips of some
autosomes (6). In the human these autosomes must be the satellited
acrocentrics, that is, the members of groups 13-15 and 21-22.
The ultrastructure of the main and secondary nucleoli shows two

clear-cut regions: a granular, outer region and a central, dense
and fibrillar region(there is also a connecting region associated
with the basal knob).
Thus, a sort of natural segregation of zones occurs in the pachy-

tene nucleoli of humans as well as other mammals (19).
Histochemical tests show that the human XY pair does not contain
RNA (6). In the main nucleolus, RNA is found in the granular zone.
The inner dense zone gives negative reaction for RNA as well as for
DNA tests (7).
CONCLUSION AND SUMMARY
During meiotic prophase special structures are formed which are

intimately related to synapsis and chiasmata formation. The syn-
aptonemal complex is formed at zygotene by the union of single homo-
logues on which a deposition of SC material,including basic pro-
tein,forms definitive lateral elements. The SC material is depos-
ited on a small part of each homologue, as the main mass of the
chromosome is formed by chromatin fibrils which form loops at the
sides of each lateral element. It is assumed that this linear part
of each homologue associated to SC material is the one involved
with synapsis and crossing-over. The cores of the sex pair behave
like the lateral elements of the SC, and are collinear to the axes
of the X and Y chromosomes. The spatial reconstruction of the hu-
man XY pair shows the presence of a small SC formed by both chromo-
somes. It is concluded that effective partial synapsis exists in
the XY pair. The tip of the short arm or the X chromosome is prob-
ably involved in that pairing. Crossing-over and partial sex-linkage
of some genes may exist.
Nucleoli of spermatocytes show a segregation of their regions

and a zone without detectable RNA.
ACKNOWLEDGMENTS
We thank Prof. R. E. Mancini for his support and Dr. J. C.

Lavieri for providing the human biopsies. The authors are Members
of the Scientific Career, Consejo Nacional de Investigaciones
Cientificas y Tecnicas. This work was partially supported by a
Grant of the Population Council.
REFERENCES
1. Heller, C. G. and Clermont, Y., Recent Progr. Hormone Res., 20:

545, 1964.
2. Stern, H. and Hotta, Y. Genetics, Suppl., &1: 27, 1969.
3. Lima-de-Faria, A., German, J., Ghatnekar, M., McGovern, J. and
Anderson, L., Hereditas, 60: 249, 1968.
4. Hackenbrock, C.R., Proc. N;t. Acad. Sci. U.S.A., 61: 598, 1968.
5. Woollam, D.H., and Ford, E.H., J. Anat., 98: l63,-r964.
6. Solari, A.J. and Tres, L., Chromosoma ,Q: 16, 1967.
7. Tres, L., Thesis. Fa.cultad de Medicina, Univ. de Buenos Aires,
"no.
136 A. J. SOLARI AND L. L. TRES
8. Moses, M.J., Genetics, Supple 61: 41, 1969.

9. Moses, M.J., Ann. Rev. Genet., l: 363, 1968.
10. Moses, M.J. ~ Cytology and Cell Physiology, ed. Bourne, G.,
Academic Press, New York, p.423, 1964.
11. Wettstein, R. and Sotelo, J.R., J. de Microsc., 6: 557, 1967.
12. Woollam, D.H., Ford, E.H. and Millen, J., Exp. C;ll Res., 42:
657, 1966. -
13. Coleman, J.R. and Moses, M.J., J. Cell Biol., 23: 63, 1964.
14. Sheridan, W.F. and Barrnett, R.J., J. Ultrastr~t. Res., 27:
216, 1969. --
15. Moens, P.B., Chromosoma, 23: 418, 1968.
16. Roth, T.F. and Ito, M., J:-Cell Biol., 35: 247, 1967.
17. Sotelo, J.R. and Wettstein, R., Chromos~a, 15: 389, 1964.
18. Solari, A.J., Exp. Cell Res., 38: 160, 1964.-
19. Solari, A.J., J. Ultrastruct. Res., 27: 289, 1969.
20. Tres, L. and Solari, A. J., Z Zellfo~ch., 91: 75, 1968.
21. Solari, A.J. and Tres, L., J. Cell Biol. (i~press).
22. Solari,' A.J., Genetics Supple 66: 113, 1969.
23. Hulten, M., Lindsten, J., Ming:-P.M. and Fraccaro, M., Ann.
Hum. Genet., 30: 119, 1966.
24. Ferguson-Smith; M.A., Cytogenetics, 3: 124, 1964.
DISCUSSION
CLERMONT: Dr. Solari, you have indicated that in the nuclei of

pachytene spermatocytes there are many nucleoli. Could you tell us
if all these nucleoli are identical, or different morphologically
when examined with the electron microscope?
SOLARI: The nucleoli in human spermatocytes are generally three or

four. One is a main nucleolus. The others are secondary nucleoli.
This fact has been observed independently by Ferguson-Smith (Cyto-
genetics l: 124, 1964) and in our laboratory. Secondary and main
nucleoli are identical in their ultrastructure, and all of them are
related to acrocentric chromosomes.
FORD: I would like to ask Dr. Solari how he distinguishes between

a terminal chiasma and a terminal association which is not a termi-
nal chiasma.
SOLARI: I would like to interpret your question as referring to the

XY pair. In the human and in the mouse, the XY pair as seen by the
light microscope has no non-terminalized chiasmatic relationship.
What we have seen is that there is an ultrastructural basis for syn-
apsis, that is, a synaptonemal complex, and thus it points to the
possibility of the existence of a chiasma. We have not said that
there are chiasmata in the XY pair. We have demonstrated synapsis.
We think that we are on the road to demonstrating chiasmata.
FORD: All the evidence that I have seen of association between X

and.Y, in man, either of my own or published by other workers, does
not exclude the possibility that there was an earlier chiasma which
terminalized by the time that observation was made. Equally, I have
not hitherto encountered evidence on the other side for parasynap-
tic pairing, chiasma formation and terminalization. I contend that
the issue is still an open one.
SOLARI: This is a point of disagreement. I think that our evidence

of synapsis points to the possible existence of chiama. Your evi-
dence and the evidence of other authors of the end-to-end relation-
ship meaning chiasma terminalization has been opposed by other au-
thors who thought that the relationship could be explained by a
joining of the X and Y by the remaining material of the sex vesicle.
This was one interpretation different from a chiasmatic mechanism,
so I would say that the question merits more proof, but that our re-
sults strongly suggest chiasma formation between the X and the Y.
E. STEINBERGER: Dr. Solari, have you observed any extrusion of nu-

cleoli in spermatocytes?
SOLARI: I have observed extrusion of nuclear material which I have

not identified histochemically in secondary spermatocytes. I could
not positively say that nucleoli are extruded.
JOST: Dr. Solari, you emphasized several times the fact that, in
humans, there is no relation between the nucleolus and the sex chro-
mosomes, whereas there is such a connection in the mouse. What is
the significance of this difference?
SOLARI: Dr. Ohno (Ohno S., Kaplan W. and Kinosita, R., Exp. Cell.
Res. 13: 358, 1957), proved that nucleolus organizers are on the
X-chromosome of the mouse, and that in the human, nucleolus organ-
izers are not in the sex chromosomes.
JIRASEK: There is an association between the nucleolus and sex

chromatin in mammalian ganglionic cells. It was described in the
original paper by Barr in cats, and by others, including myself in
the human. Could you comment on this association?
SOLARI: I cannot comment on human ganglionic cells, of which I have

no experience. But I would say that there is a close approachment
between a secondary nucleolus and the XY pair in the human spermato-
cyte. This occurs with a low frequency and we have explained this
fact by the use of serial sections. (Solari, A. J. and Tres, L.,
J. Cell BioI., April, 1970). The formation of the nucleolus is re-
lated to a heterochromatic knob, and the sex chromosomes are also
formed by heterochromatin. Also, as Dr. Ohno said, there is affin-
ity between the different heterochromatic parts of the chromosomes.
Thus, the hypothetical explanation of the infrequent proximity but
THE CYTOGENETICS OF THE MALE GERM CELLS AND THE TESTIS IN MAMMALS
C. E. FORD
Medical Research Council, Radiobiology Unit,

Harwell, Didcot, Berkshire, England
INTRODUCTION
The testis was the classical source of material for chromosome

study in mammals. The first objectives were the determination of
chromosome number, the identification of the sex chromosomes and
the characterization of the normal course of meiosis in spermato-
cytes. Later, chiasma counts, the analysis of meiotic behavior
in structural heterozygotes and mapping the distribution of chro-
momeres in pachytene with a view to the recognition of individual
bivalents were all attempted. These latter tasks are barely more
than begun even today. However, cytogenetic methods can contribute
in a different way to an understanding of some aspects of testicu-
lar biology, the main operational theme being the identification of
specific cell types in a mixed population by specific features of
the karyotype. Four of these aspects are considered below.
Determination of the Gonads in Embryogenesis
The question of whether the nature of the mammalian gonad is

determined during embryogenesis by germ cells or by cells of the
mesenchymal stroma is still not answered satisfactorily. A cyto-
genetic study of some mosaics and chimeras may contribute to the
solution in a unique way. The point can be made by reference to
results from a study of two human 45,X/ 46,XY (i.e. XO/XY) mosaic
cases. Both were children with ambiguous external genitalia raised
as boys, and at laparotomy both were found to have a testis one side
and a streak gonad the other. Cultures were set up from biopsy spe-
cimens taken from all four gonadal structures and eventually prepa-
rations were made and samples of mitotic cells identified as XO or
139
140 c. E. FORD
XY (1). In the one case 30 percent of the cells in the sample
from the testis culture were XY compared with 10 percent from the
streak gonad, the difference being formally significant at the 1
percent level. The proportions of XY cells were in the second case
20 per cent from the testis and again 10 percent from the streak
gonad. This time the difference bordered on significance at the 5
percent level. If it may be assumed successively that these data
were representative of the explants, that the explants were repre-
sentative of the stromata of the whole structures from which they
were taken, and that the stromata were derived from the original
mesenchymal cells of the gonadal ridges, then they suggest three
conclusions. First, that the nature of the primitive gonad is in-
fluenced by the somatic cells of the gonadal ridge. Second, that
relatively few XY cells in a mixture with XO cells are sufficient
to initiate testicular differentiation of the gonad primordium.
And third, that there is a threshold proportion of XY cells below
which testicular development fails and ovarian development takes
its place. Singh and Carr (2) have demonstrated the presence of
normal or nearly normal ovaries in 45,X fetuses and there is there-
fore good reason to suppose that streak gonads are derived from
pre-existent ovaries, at least in subjects with a 45,X cell line.
A corollary would be that cells of the gonadal ridge that contain
a Y chromosome can influence the morphogenetic behavior of neigh-
boring cells that do not contain a Y chromosome.
Before firm conclusions can be drawn it will be necessary to

assemble more information of the same kind from other 45,X/46,XY
human mosaics and to test the validity of the assumptions carefully.
Meanwhile, some support for the basic idea that a minimal propor-
tion of mesenchymal XY cells in the gonad primordium is necessary
for testicular differentiation is provided by information obtained
from cultures of gonadal tissue from human 46,XX/46,XY chimeras.
The noteworthy feature is that XY cells were not recognized in
either of two cultures from ovarian biopsies but were present in
all three cultures established from testicular explants. Original
references have been given elsewhere (1).
Most informative of all could be the intensive study of exper-

imental 40,XX/40,XY chimeras in the mouse as begun by Mystkowska
and Tarkowski (3). In such animals it should eventually be possi-
ble to relate the proportion of XX to XY in both germ cells and in
cultures grown from gonadal explants to the histological nature of
each gonad.
Competition between Germ Cells
Mystkowska and Tarkowski (3) studied in detail a series of

nine mice obtained subsequent toemb~yo fusion. The inbred stocks
used, CBA-£ and CBA-T6T6, differed in both coat pigment and karyo-
type. Seven were overt chimeras as determined by inspection of the
coat and retinaeo Examination of the chromosomes in direct prepa-
rations from bone marrow of six of them showed that three were 40,
XX/40,KY chimeras; two phenotypic mal.es and a hermaphrodite. Breed-
ing tests with the two males showed that only one of the two compo-
nents had contributed to the functional spermatozoa, and, as the
progeny included males, that component must have had an XY sex
chromosome constitution. Subsequent direct examination showed that
the spermatocytes in the testes of these two animals, and in the
one testis of the hermaphrodite, were all KY.
The important point is that all the germ cells detected in the
testes of these three animals were 40,XY cells, whether identified
directly or inferred from the breeding data. To account for this
result it might be postulated that the primordial germ cells arise
from a relatively small number of precursors and, in a chimera, are
therefore commonly all derived from the one component or all from
the other. But if this is so it may be supposed that in an XX/XY
chimera the probability that they will all be XY is approximately
1/2 and the probability that they are all XY in three successive
XX/XY animals is (~)3 or 1 in 8. In a 39,X/41,XYY mosaic mouse
studied in this laboratory it was found that although both cell
types were present in bone marrow, all the germ cells were XYY (4).
If now the cell types compared are taken as those with one or more
Y chromosomes and those that lack a Y chromosome, the probability
that the germ cells of the testis will all be of the first type in
four out of four animals is 1 in 16. Although this is insufficient
for outright rejection of the hypothesis of "purity" of the primor-
dial germ cell population, it renders it somewhat unlikely.
The alternative is to suppose that the initial population of

primordial germ cells may be mixed and that in the developing testi-
cular environment germ cells that contain a Y chromosome are fa-
vored over those that do not. The idea of a mixed initial germ
cell population is favored by data assembled by Russell (5) on
mosaic mutant mice. She showed that mice with mosaic coat pheno-
types attributable to presumptive gene mutation very early in de-
velopment are almost invariable germ cell mosaics as well. But the
most telling evidence is again provided by Mystkowska and Tarkowski
(3). A further two animals in their series of seven were 40,XY/40,
KY chimeras and their breeding records showed quite deciSively that
both components were present in the germ cell population. The evi-
~e as a whole therefore strongly favors the view that XX germ
cells may not be able to multiply in the testicular environment of
the mouse, and if they are able to multiply, that they are elimin-
ated by the time the animals are of breeding age. This could be by
differential proliferation or differential survival or both and
142 C. E. FORD
could be regarded as a form of competition.
It is appropriate now to mention observations made on germ

cells in testicular preparations from human 47,XYY males. Results
have been reported from 11 subjects (6-10). Despite the absence
of any hint of mosaicism in blood cultures, the vast majority of
spermatocytes are of the constitution of 46,XY; they have lost one
of the two Y chromosomes. Only in one of the 11 subjects have a
few presumptive 47, XYY cells been identified. These were cells
with a univalent X and an apparent YY bivalent (9).
A special elimination mechanism for one of the two Y chromo-

somes has been suggested (7). But there is no precedent in mam-
malian cytogenetics for such a special mechanism and the alterna-
tive view is therefore preferred that once a Y chromosome had been
lost by chance lagging or non-disjunction, the resultant cell with
a normal 46,xy complement would enjoy selective advantage in pro-
liferation or survival or both.
Clonal Proliferation of Germ Cells in the

Irradiated Mouse Testis
The preparations are made from the testes of mice after re-
covery from exposure to ionizing radiation some of the spermato-
cytes are found to contain a quadrivalent configuration indicative
of the presence of a reciprocal translocation. Other spermatocytes
may contain two or more quadrivalents or multivalent configurations
of a higher order. In some preparations from single testes several
cells were found to exhibit the same unique multivalent configura-
tion or set of configurations and were interpreted to be clones
produced by the multiplication of single spermatogonia in which
trans locations had been induced by the irradiation (11). Two sper-
matocytes from one clone are shown in Figs. 1 and 2.
Figs. 1 and 2.
Two spermatocytes from one testis of a mouse killed nine

months after exposure to 600 R 250 kVp X-rays. Both con-
tain a chain of eight chromosomes and are members of the
same clone. The uniqueness resides in the relative lengths
of the successive chromosomes in the chain. Variation in
the position and number of the chiasmata are irrelevant but
can modify its gross appearance. Accordingly other sperma-
tocytes of the same clone might be expected with, for
example, a ring of eight or two chains of four.
Fig. 1.
Fig. 2.
144 C.E. FORD
In further unpublished work on mice exposed to an acute X-ray

dose of 600 R the occurrence of such clones has been repeatedly con-
firmed, some of them making up three or four percent of the whole
sample. It is assumed that the samples of cells analyzed are drawn
randomly from the spermatocyte population since a separate cell sus-
pension is made from each whole testis as the initial step in the
preparative schedule. Oakberg (12) has reported that only some
three per cent of type A spermatogonia in the mouse survive acute
exposure to 600R X-rays. If the conservative assumption is made
that there are 5 x 105 type A spermatogonia in the normal mouse tes-
tis, there should still be 10 4 survivors. If all contributed equal-
ly to the regeneration of the seminiferous tissue, the probability
of identifying members of the same clone in the samples of 200 or
400 spermatocytes normally examined would be negligibly low. It fol-
lows either that the numbers of surviving spermatogonia that contri-
bute to regeneration must be very much less than Oakberg's estimate
of the proportion of survivors would suggest or that there is very
strong differential proliferation as between different clones. Or
both could be responsible in part.
Clermont and Huckins (13) have shown that there are some 30
tubules in the normal rat testis that have connections to the rete
testis but not to one another. If the mouse testis has a similar
structure and if type A spermatogonia cannot move from one tubule
to another, the maximum clone size expected if the whole of a tubule
were repopulated by the descendants of one surviving spermatogonium
would be four or five per cent, allowing for some variation in size
between tubules. The observation of clones constituting three to
four percent of the sample suggests that repopulation of a tubule
by descendants of a single spermatogonium may indeed occur.
It is hoped that careful squash preparations of separate short

lengths of tubule will confirm the validity of the clone interpreta-
tion and also give direct information about the length of tubule
that may be occupied by a single clone.
Genetic Information and Spermatogenesis
Mice of either sex that are heterozygous for a reciprocal trans-

location between autosomes (as opposed to a Robertsonian transloca-
tion) are normally semi-sterile. That is to say, in matings to nor-
mal, litter size is reduced to about 50 per cent of that recorded for
parallel matings of their structurally homozygous sibs. However,
Lyon and Meredith (14) reported five stocks in which the heterozygous
males were completely sterile. These stocks were established from
semi-sterile daughters of irradiated males and were maintained
through the female line. In another experiment, Cattanach et al (15)
treated males with ethane methane sulphonate and found seven com-
pletely sterile sons out of 17 that were heterozygous for autosomal

reciprocal translocations.
Semi-sterility of reciprocal translocation heterozygotes is

satisfactorily explained by segregation of unbalanced haploid game-
tic genomes. Full sterility, however, cannot be accounted for in
these terms. The supposition is that it is a direct consequence
of the presence of the translocation itself and so should be attri-
buted to one of the classical three alternatives of position effect,
or gene mutation associated with one of the break points, or minor
duplication or deletion associated with one of the break points.
But to demonstrate that the translocation itself is responsible
(through whichever of these mechanisms) would require proof that
sterility remained associated with the translocation on further
breeding, i.e. could not be separated from it by crossing over.
Given this proof, the inference follows that genetic information lo-
cated at or near to one or both of the break points is in some way
concerned with spermatogenesis. And the fact that the break points
of the mouse trans locations that have been mapped are widely scat~
tered over many of the chromosomes suggests that many different
genetic loci may be involved. Specific supporting evidence is pro-
vided by the persistent association of male sterility with the trans-
location on further breeding of four of the five translocation stocks
reported by Lyon and Meredith. (The fifth was lost). Furthermore,
three of the eight break points have now been located on different
chromosomes (Lyon, personal communication).
The alternative formal explanation that has been advanced by

some;that sterility is a consequence of the disturbance of the meio-
tic process by the rearrangement ~ a rearrangement, can be dismissed.
Apart from the great difficulty of envisaging a causal sequence ini-
tiated by a mechanical factor and terminating in physiological fail-
ure there are two specific reasons for rejection: the property of
male heterozygote sterility is shown by some trans locations but not
by others that are mechanically equivalent, and in two of the cases
reported by Cattanach and his colleagues meiotic cells were reported
to be absent,the presence of a presumptive reciprocal translocation
having been established by observations of chromosomes in mitosis.
Meiosis not begun cannot be disturbed by abnormalities of pairing.
The proportion of all reciprocal trans locations between auto-

somes in the mouse that bring about sterility in male heterozygotes
has not yet been estimated and may not be great. Lyon and Meredith's
five stocks were established from 46 Fldaughters with significantly
reduced fertility. Translocations were inferred to be present in 36.
Heterozygotes of both sexes were semi-sterile in 26 and in the re-
maining five some of the male heterozygotes were semi-sterile and
some completely sterile.
146 C. E. FORD
In contrast to the purely autosomal trans locations it is re-

markable that heterozygous males are sterile in eight of the nine
X-autosome trans locations recorded in the mouse (16,17). Most are
proved or assumed reciprocal trans locations with the distal and
proximal portions of the X chromosome physically separated from one
another. It may therefore be significant that the one exception
with some semi-sterile male heterozygotes is a 3-point rearrange-
ment with a segment of the chromosome bearing linkage group I in-
serted into the X chromosome, which otherwise preserves its struc-
tural integrity (18). Merely to suppose that the X chromosome is
rich in loci that are in some way concerned with spermatogenesis
would not readily account for the exception. A somewhat more plau-
sible explanation, however, could be provided if the physical se-
paration into two parts of the X chromosome in the reciprocal trans-
locations should interfere with the normal development of the sex
vesicle at early prophase in the primary spermatocyte and thereby
disturb later events in spermatogenesis.
SUMMARY
Cytogenetic study of certain natural human mosaics and chime-

ras and experimentally produced XX/XY chimeras of the mouse may
eventually provide a decisive answer to the question whether the
testicular or ovarian nature of the embryonic gonad is determined
by the germ ce11s themselves or by somatic ce11s of the mesenchymal
stroma.
There is evidence from experimental mouse chimeras and from a

natural mosaic mouse that in the testicular environment, germ cells
containing at least one Y chromosome are favored over germ cells
that lack a Y chromosome. In human 47,XYY male subjects, one of
the two Y chromosomes is eventually lost from the overwhelming ma-
jority of the cells of the germ line. Strong selective prolifera-
tion of karyotypically normal 46,XY cells after accidental loss of
one Y chromosome could account for this observation.
Clones of spermatocytes characterized by specific chromosome

rearrangements can be identified in the testes of mice that have
recovered after radiation - induced depletion of the seminiferous
epithelium. Numerical considerations suggest either that fewer
surviving A-type spermatogonia contribute to the regeneration of
the seminiferous tissue than had hitherto been supposed, or that
there is a strong proliferative differential between different
clones, or both.
Considerations ar1s1ng from the association of male sterility

with heterozygosity for certain specific reciprocal translocations
between autosomes in the mouse suggest that many autosomal loci
may be concerned in some way with spermatogenesis. All males he-

terozygous for a reciprocal translocation between an autosome and
the X chromosome yet reported are sterile. This suggests either
that the X chromosome is particularly rich in loci concerned in
spermatogenesis or that a special unitary explanation, possibly
disturbance of the formation of normal sex vesicle in early meiotic
prophase, may apply.
REFERENCES
1. Ford, C.E. J. Biosocial Science Supplement~: 7, 1970.

2. Singh, R.P. and Carr, D.H., Anat. Rec. 155: 369, 1966.
3. Mystkowska, E.T. and Tarkowski, A.K., J:-Embryol. Exp. Morph.
20: 33, 1968.
4. Evans, E.P., Ford, C.E. and Searle, A.G., Cytogenetics 8: 87,
1969.
5. Russell. L.B •• In The Role of Chromosomes in Development.
ed. Locke, M. A~demic Press, N.Y. p 153, 1964.
6. Thompson, H., Melnyk, J. and Hecht, F., Lancet, (ii): 831,1967.
7. Melnyk, J., Thompson, H., Rucci, A.J., Vanasek, r;-and Hayes,
S., Lancet (ii): 797, 1969.
8. Hulten, M., Lancet (!): 717, 1970.
9. Evans, E.P., Ford, C.E., Chaganti, R.S.K., Blank, C.E. and
Hunter, H., Lancet (!): 719, 1970.
10. T(~ttenborn, U., Schwinger, E. and Gropp, A. Deutsche Med.
Wschr. 95: 158, 1970.
11. Searle, A.G., Evans. E.P., Ford, C.E. and West, B.J., Mutat.
Res. 6: 427, 1968.
12. Oakberg, E.F., Jap. J. Genet., 40 (Suppl.): 1191 1965.
13. Clermont, Y. and Huchins, C., Amer. J. Anat., 108: 79, 1961.
14. Lyon, M.F. and Meredith, R., Cytogenetics, 2: 335, 1966.
15 • Cattanach, B.M., Pollard, C.E. and Isaacson, J.H., Mutat. Res.,
.£: 297, 1968.
160 Russell, L.B. and Montgomery, C.S., Genetics 63: 103, 1969.
17. Lyon, MoF., Searle, A.G., Ford, C.E. and Ohno, S., Cytogenetics,
1: 306, 1964.
18. Ohno, S. and Cattanach, B.M., Cytogenetics, 1: 129, 1962.
DISCUSSION
ROSS: Dr. Ford, do you accept the evidence that apparently morpho-
logically normal testicular differentiation can occur in the absence
of a demonstrable Y chromosome, and if you accept this evidence, how
do you rationalize it in relation to the possible role of the Y chro-
mosome in testicular differentiation?
FORD: I think the best answer in man is the hypothesis of Ferguson-

Smith, that in instances of testicular differentiation in the human
148 C. E. FORD
species, in the absence of the Y chromosome, more specifically in

the presence of two X chromosomes, as in the majority of true her-
maphrodites and the so-called "XX male", this can be accounted for
by exchange of material between the X chromosome and the Y during
spermatogenesis in the father of the exceptional individual. The
evidence favoring this interpretation is mounting, and is derived
from the study of the Xg blood group system in the parents and in
the exceptional children by Race and Sanger. There are three in-
stances on record out of the small number of families that have been
analyzed in which the mother was Xg negative and the father Xg pos-
itive, yet the child, XX, true hermaphrodite or XX male was Xg negative
like the mother, implying that there had been no contribution of the
Xg locus from the father to his exceptional child. Now, when two
exceptional events occur together like that, it is reasonable to
suppose that ~here may be some association between them, and that
the association is an actual exchange of material between the X-chro-
mosome and the Y, which at the same time could be associated with
the transference of male-determining genetic information. Now, we
need the further background of the Lyon hypothesis. Given the pres-
ence of active male-determining material on some of the X-chromosomes,
not on others, then depending on the proportion of the two types of
cell in the early gonads, one could, in theory, get either ovarian
or testicular development.
PAULSEN: With respect to those phenotypic males having testes but

where only a single stem-cell line, namely XX, has been found, is
it not possible, Dr. Ford, that in vivo or in vitro cellular selec-
tion may play a role in this finding? Dr. Gartler has published
data which emphasize the importance of these phenomena.
FORD: I agree, it is a formal possibility, and it is one that I

think de la Chapelle favors. On the whole, though, although I am
prepared to accept the existence of cryptic mosaics, I do not think
that it can be sufficient to account for the considerable number of
these XX males and XX true hermaphrodites, in whom none but XX cells
have been identified.
PAULSEN: I agree that there may be several explanations. One study

of a phenotypic male with scrotal testes revealed only an XX line
examining the chromosomes from tissue culture of testes, skin and
blood (Hecht, F., et ale Pediatrics, ~: 982, 1966). These studies
were repeated by obtaining additional material from each tissue.
Two stem cell lines including one with XXY sex chromosomes were
demonstrated.
A. STEINBERGER: Does Dr. Ford, or any of the other speakers, have

information regarding the duration of Gl, Sand G2 periods in the
spermatogonial cells, and how they compare with the somatic cells
in the human?
SOLARI: There is information but not in humans. In some mammals,

the duration of the 8 phase before the last division of type B sper-
matogonia is much longer than somatic S phase, and this seems to be
an indication of the transition between the classic S phase of the
mitotic cycle to the very long S phase of the preleptotene.
WENIGER: In relation to the first point discussed by Dr. Ford, it

has been shown by L. Bounours that the chick embryo, the germinal
crescent of which has been destroyed by x-irradiation, develops
sterile testes or ovaries, in accordance with its genetic sex.
ORNO: Dr. Ford, may I ask you two questions. One, Beatrice Mintz
has her own idea which can be expressed as an "all or none" theory
to explain why so few of her allophenic mice are hermaphrodites.
She believes that most of the XX/XY mosaics develop either as nor-
mal males or normal females. What do you think of her explanation?
And another question is that in your spermatogonial clone, the clone
is presumably heterozygous for a translocation or trans locations,
so presumably you get two types of viable gametes produced. Do you
recover these two gametic types in expected one-to-one frequency
among the progeny?
FORD: In reply to the first qu~stion, I had not seen Beatrice Mintz's
own interpretation, but I am aware that her data appeared to differ
very strongly from the data obtained by Mystkowska and Tarkowski on
the one hand, and McLaren and Bowman on the other. Mintz obtains
almost equal numbers of apparent females and apparent males with
very few true hermaphrodites, whereas Mystkowska and Tarkowski and
McLaren and Bowman obtain very great excesses of males and a greater
proportion of hermaphrodites. I think that the only satisfactory
resolution to this problem will be when all these chimeras are iden-
tified chromosomally. With respect to the second question, let me
say that after irradiation of the testis and following recovery,
there are very many clones of spermatocytes, some carrying one, some
two, some three reciprocal translocations and some none at all. 80
the sperm population produced by such an individual will be a very
complex one indeed, and the problem of working out what its consti-
tution would be is qUite difficult. Nevertheless, we have attempted
it, and our conclusion is that only about half the expected number
of semi-sterile offspring are produced by such males as would be
expected from the chromosome observations. Now to the specific
question as to the number of types of viable gametes or of gametes
that would give viable individuals. Assuming two-by-two segregation
of the association of four, there are ten possible gametic types of
which two will give fully viable individuals, one homozygous normal,
the other carrying the translocation. That is, of course, when a
single translocation is segregating.
CHROMOSOME STUDIES ON TESTICULAR CELLS OF SUBFERTILE MEN
Ann C. Chandley
Medical Research Council, Clinical and Population

Cytogenetics Unit, Western General Hospital,
Edinburgh, 4
GENETIC CAUSES OF INFERTILITY
Infertility in man is recognized as a reduced capacity for

producing live offspring.
The causes of this reduction may be many and varied, but in a

proportion of cases, the infertility can be brought about by genetic
factors which may not directly harm the individual carrying them,
but which may produce defects in gametogenesis. Such defects are
known in other species. For example, spermatogenic depression or
arrest has been reported in mutant strains of Mastomys (1), and the
guinea pig (2), and mutants which affect synapsis and disjunction
at meiosis are well known in several plant species and in Drosophila
(3,4,5). In the bull, a change in sperm morphology resulting in
loss of reproductive function is known to be brought about through
a recessive gene acting in spermiogenesis (6) and it is possible
that gene action is also responsible in man for producing abnormal-
ities of this kind. For example several reports have been made of
infertile patients whose infer ility is attributable to a defect in
sperm morphology (7,8,9) and recently an example of what may be a
desynaptic mutant in man has been reported by Pearson (10). Among
the patients attending a subfertility clinic in Oxford, England,
Pearson found an azoospermic male in whom the mitotic karyotype was
normal, but in which many of the chromosomes at diakinesis were pre-
sent as unpaired univalents. (Fig. 1) The average number of chias-
mata per cell was four compared with the normal mean frequencies of
55.9 and 52.7 per cell which have been reported by Ford and Hamerton
(11) and Kjessler (12). The fact that some chiasmata at least were
present in the cells at diakinesis suggests that the chromosomes
had been paired in early prophase but had later undergone desynapsis,
and the presence of some pairs of univalents of approximately equal
151
152 A. C. CHANDLEY
Fig. 1. Diakinesis in a desynaptic male showing many chromosomes

present as univalents and a very low number of chiasmata.
(Reproduced by kind permission of Dr. p. Pearson).
size lying in juxtaposition gave further support to this idea.
CHROMOSOME ABERRATIONS IN SUBFERTILE PATIENTS
More information is available regarding the association between

infertility and chromosome aberrations in men, and much of it has
been obtained through systematic surveys of patients attending sub-
fertility clinics. One of the earliest surveys, carried out by
Ferguson-Smith et al. (13) in Glasgow, revealed that 11% of all pa-
tients having a sperm count of less than 1 million sperm per mi. of
seminal fluid were sex chromatin positive on a buccal smear. In a
later survey, carried out in Sweden, Kjessler (12) found the inci-
dence of such males to be 16%, a figure in agreement with that re-
ported earlier by Ferguson-Smith et al. (13). Mitotic chromosome
studies in the four sex chromatin positive males found by K;essler
(12) out of a total of 25 patients having sperm counts of <1 mill.
sperm/mi. revealed that two had the usual XXY sex chromosome com-
plement of an uncomplicated Klinefelter while the other two were
mosaics of the XY/XXY type. Meiotic studies were not possible since
all four males lacked histological signs of spermatogenesis in the

seminiferous tubules.
Among 131 sex chromatin negative subfertile individuals, Kjes-

sler found six patients with abnormal mitotic karyotypes. (A fur-
ther two patients with Y chromosomes of unusual size were also found
but variations in the size of the Y chromosome are known to occur in
about 3% of the normal male population (14) and these latter two pa-
tients will therefore not be considered further.) All but one of
the six patients with abnormal karyotypes belonged to classes with
less than 40 million sperm per mi., and no individuals with an ab-
normal karyotype were found among patients with more than 80 million
sperm per mi. (Table 1).
In a similar survey on subfertile males carried out in Edin-

burgh by McIlree et ale (15), abnormal mitotic karyotypes were found
in three sex chromatin negative individuals out of a total of 50 pa-
tients having sperm counts of less than 20 million sperm per mi. and
in addition, two patients were found in which the mitotic karyotype
was normal but in which a multivalent was thought to be present at
diakinesis in the primary spermatocytes. In a current Edinburgh
survey, one patient with an abnormal mitotic karyotype has so far
been found out of a total of 45 sex chromatin negative individuals
with sperm counts of less than 20 million per mi.
TYPES OF CHROMOSOME ABNORMALITIES FOUND IN SUBFERTILE PATIENTS
Sex Chromosome Mosaicism
Among the six patients with abnormal karyotypes reported by

Kjessler, four were sex chromosome mosaics. Three patients were 45,
X/46,XY mosaic. The testicular histology on these patients was ap-
parently normal and primary spermatocytes from two of the XO/XY mo-
saics showed some cells at diakinesis which Kjessler claimed had an
XO sex chromosome constitution. In addition other cells were chro-
mosomally normal and had a normal XY bivalent.
Autosomal Abnormalities
Among the subfertile patients reported by Kjessler was one azo-

ospermic male, whose testicular histology showed the typical appear-
ance of germinal cell aplasia and whose mitotic karyotype showed 47
chromosomes. The extra chromosome was approximately the size of a
G group chromosome but was apparently metacentric. In our current
Edinburgh survey, we have found an oligospermic male with a similar
abnormality, and it is of interest that the patient's brother has
now also presented at the infertility clinic and he too has been
found to show the same abnormality in blood. A testicular biopsy
has not yet been performed, however. The patient was a 30 year old
TABLE 1 01
..,.
CYTOGENETIC DATA ON SEX CHROMATIN NEGATIVE SUBFERTILE PATIENTS FROM THREE INDEPENDENT SURVEYS
Total Abnormal
Survey Sperm Count Mitotic Chromosomes Meiotic Chromosomes
Patients Karyotypes
21 1 Azoospermic 47,XY,mar+ Extra univalent present at

diakinesis.
24 2 1-20 mill./ml. 45,XY,D-,D-,t(DqDq)+ Trivalent present in most
spermatocytes.
" 46 , XY / 47 , XXY Normal
Kjessler 21 2 21-40 mill./ml. 45 ,XI 46 ,XY XO spermatocytes present
1966 " 45 ,xl 46 ,XY(?) "
14 o 41-60 mill./ml.
20 1 61-80 mill./ml. 45,X/46,XY Not studied
31 o >81 mill./ml.
131 6
34 2 Azoospermic 46 ,XY ,Gr Two univalents present at

diakinesis.
Mcllree " 45,X/46X,Ydic X and Y present as
1966 univalents in most cells.
16 3 <20 mill./ml. 45,X/46,Xinv(Yp+q-) Normal
" 46,XY Multivalent present at
diakineSis." '1>
" " ('I
50 5 ()
:J:
»
Z
Chandley 45 1 <20 mi 11. /ml. 47,Xy,mar+ Extra univalent at o
r-
1969 diakenesis ~
male who had counts of 7 and 3 million sperm respectively on two

separate occasions and on each occasion, less than 30% of the sperm
were morphologically normal and no more than 30% were motile. The
mitotic karyotype showed a small extra chromosome, slightly smaller
than a G group chromosome which was apparently metacentric (Fig. 2)
'to
:t
B »
2 3 4-5
HI
6-12+x
A6 ii
13-15 16 17-18
X ~1: "6 A_ I
19-20 21-22 Y Ab.Ch.
Fig. 2. Mitotic karyotype from an oligospermic patient with a small

extra centric chromosome.
Meiotic preparations made by the method of Evans et al. (16)

showed many cells in diakinesis I but very few second meiotic divi-
sions. At diakinesis, the small extra chromosome could be seen in
most cells as a univalent lying free from the other bivalents, and
no abnormal configurations were seen (Fig. 3). It was also observed
in a few cells in Metaphase II as a small dot-like chromosome.
Many diakineses showed degenerative changes in the bivalents

and it appeared that meiotic arrest was occurring in many cells be-
tween the first and second meiotic divisions.
156 A. C. CHANDLEY
A 6
4
c
o E
•
F G XV Ab.ch .
47 , XYmar +
Fig. 3. Diakinesis and meiotic karyotype from the patient with a

small extra centric chromosome. The extra chromosome is
seen as a univalent in most primary spermatocytes.
The presence of supernumary chromosomes in oligospermic males

has also been reported by Smith et ale (17) and by Hulten et ale
(18), but their origin is not known, and direct evidence that the
infertility is in any way related to the presence of the extra chro-
mosome is lacking. Supernumary chromosomes have been reported, for
example, in many apparently normal individuals as well as in patients
showing a variety of pathological conditions (19,20).
The sixth patient with a mitotic chromosome abnormality, re-

ported by Kjessler, was an oligospermic male who was the carrier of
a heterologous DID translocation and whose testicular histology
showed spermatogenic arrest of many gametes at the spermatid level.
We have also had the opportunity to study meiosis in an oligo-

spermic male with a translocation of this kind.
The patient did not originally present at the infertility clin-

ic but was found through his sister, also a carrier of a DID trans-
location, who reported with secondary infertility and a history of
several abortions. He was found to have a sperm count of 17 million
per mi., only 30% of which were morphologically normal, and his tes-
ticular histology showed reduced spermatogenic activity, but all
stages of maturation up to mature sperm were present in the tubules.
The karyotype in mitotic cells from the blood showed 45 chro-

mosomes including 4 D group chromosomes and a large metacentric
chromosome similar in size to a No.3 chromosome (Fig. 4). This
was interpreted as consisting of the major parts of the two missing
D group chromosomes joined together by translocation of a centric
fusion type.
j#
,
,
y
eo 'I
.p I'
..
, .. +
K+
.
1-.
....
, ...
~
3 4-5
II II x
6-12+X
• 13-15
t-
17-18
15 • • ~ lOA .110
19-20 21-22 y
Fig. 4. Mitotic karyotype in an oligospermic patient with a het-

erologous DID translocation.
158 A. C. CHANDLEY
At meiosis, a large trivalent representing the pairing config-

uration of the two translocated D chromosomes with their normal
homologues was seen in most cells at diakinesis (Fig. 5). The num-
ber of cells in second meiotic metaphase was approximately equal
in number to those in diakinesis and there was no evidence of meio-
tic arrest at this stage.
9 ~
A 6
0 4.~ (t a- I
c
4- ~ C ,.
~
a , t
E
CI C •f
O-CD D)-D
F G X+Y
t ri valent
4S.XY.D-.D-. (OqOq) +
Fig. 5. Diakinesis and meiotic karyotype from the carrier of the

DID translocation. Twenty-two structures are present in
the primary spermatocyte including a trivalent formed by
the pairing of the two translocated D chromosomes with
their normal homologues.
The various types of trivalent configuration found are shown

in Fig. 6. Various numbers of chiasmata were found ranging from 2
to 5.
Segregation from a trivalent of this kind is expected on theo-

retical grounds to produce gametes of two main types: one type
carrying a genetically balanced set of chromosomes and another type
carrying an unbalanced set. However, the observations from known
TIl : 2
Iii: : 2
lIT : 2
ill 3
ill : 4
ill 5
10u
Fig. 6. Cut-out trivalents from primary spermatocytes of the DID

translocation heterozygote and suggested interpretations
of the configurations observed.
carriers of DID trans locations suggest that only sperm which are
balanced with regard to their D chromosomes are efficient at fertil-
ization. Aneuploid offspring with the D trisomy syndrome are rare
among DID carriers and the frequency of spontaneous abortions in
unions involving a DID translocation carrier is no higher than the
levels found for the population as a whole (21,22).
Hamerton (22) has suggested that preferential segregation at

Anaphase I and selective fertilization of the egg by euploid sperm
may both be factors which favour the production of balanced off-
spring. Another possibility is that eggs fertilized by aneuploid
sperm are lost as preimplantation zygotes and therefore go undetec-
ted. Kjessler (12) has suggested that the reduction in numbers of
sperm observed in his DID translocation carrier reflects the elim-
ination, during spermiogenesis, of unbalanced sperm and this may
160 A. C. CHANDLEY
also account for the oligospermy observed in our DID translocation

carrier.
Multivalent associations at diakinesis of the type expected in

carriers of a balanced translocation were also reported in two oli-
gospermic males by McIlree et al. (15). The chromosomes involved
in these cases could not however be identified in mitotic cells from
peripheral blood. Both patients showed meiotic arrest in some cells
between diakinesis and metaphase II. Meiotic arrest at this stage
was also observed by McIlree et al. in an oligospermic male show-
ing a ring 21/22 chromosome in mitotic cells from blood and skin
(Fig. 7). Cells at diakinesis showed 22 bivalents, including the
XY bivalent, plus two G group chromosomes which were present as un-
ivalents. In some cells the X and Y chromosomes were also present
as univalents (Fig. 8), and many of the bivalents showed degenera-
tive changes and a reduced number of chiasmata.
3 4-5
6-12+x
13-15 16 17-18
,....
/ 1\
19-20 21-22 T
Fig. 7. Mitotic karyotype from an oligospermic male showing a ring

chromosome replacing a Group 21-22 autosome. (From: McIlree
et ale Lancet, Mar. 26, 679-682, 1966).
A B
D E
\
F G X+Y
Fig. 8. Diakinesis and meiotic karyotype from the oligospermic

male with a ring 21/22 chromosome in the blood, showing
degenerative changes in the bivalents, very low numbers
of chiasmata and two small chromosomes present as univa-
lents. (From: McIlree et ale Lancet, Mar. 26, 679-682,1966)
McIlree has suggested that the failure to pair of the two homo-
logues in this case may be due to one of these being a ring chromo-
some, and the pairing failure has led in some way to the develop-
ment of degenerative changes in the chromosomes and the breakdown
of spermatogenesis.
Structural Alterations of the Y Chromosome
Degenerative changes in the bivalents at diakinesis and meiotic

arrest between diakinesis and metaphase II were also observed by
McIlree et al. (15) in an azoospermic male showing a Y chromosome
with a dicentric appearance in mitotic cells from peripheral blood
(Fig. 9). The Y chromosome of the patient's father was normal. The
blood of the patient also showed a 45, X cell line (14/100 cells)
which may have some relevance to the finding of infertility. At
162 A. C. CHANDLEY
'2 3 4 - 5
6 - 1'2 + X
xx
13 - 15
X" .,.
16 17 - 18
iliA
,--
19 - '20 '21 - 2'2 Oic.Y
Fig. 9. Mitotic karyotype in an oligospermic male with a dicentric

Y chromosome. (From: McIlree, M.E., et ale Lancet, 2: 69, 1966.)
diakinesis, the X and Y chromosomes were present as univalents in

many cells and it is possible that here also, the failure to pair
has led in some way to the degenerative changes and meiotic arrest
observed.
Another structural alteration of the Y chromosome, possibly due

to a peri centric inversion (23) was reported by McIlree et ale (15)
in an oligospermic male who also showed a 45,X cell line in peri-
pheral blood. The Y chromosome was a metacentric similar in size
to a 19-20 group chromosome and was also found in the patient's
father and all his male relatives. As none of the patient's male
relatives were subfertile, it seems unlikely that the abnormality
is related to the patient's subfertility. However, the patient was
the only member of the family in whom a 45,X cell line was demon-
strated and it may well be that this bears some relationship to his
subferti li ty.
At diakinesis, primary spermatocytes from this patient showed

a normal end-to-end association.
The study of meiosis in subfertile men is of great value both

to the clinician in diagnosing the cause of the infertility and also
in providing basic information on meiosis in the human male. How-
ever, it is still not possible to say with certainty whether any
particular chromosome abnormality is a cause of infertility, al-
though there is sometimes strong evidence that this may be so.
Further studies should, however, lead to a greater understand-

ing of the processes involved and may eventually lead to a recogni-
tion that a specific group of abnormalities are associated with a
particular chromosome aberration.
ACKNOWLEDGEMENT
The present investigations were carried out in collaboration

with Dr. Melville Kerr and Mr. Peter Edmond of the Edinburgh Royal
Infirmary who kindly supplied the biopsies and Dr. N. MacLean of
the Department of Pathology, Western General Hospital, Edinburgh,
who provided the pathological reports. The technical assistance of
Mrs. Sheila Christie and Miss Judy Dennison is gratefully acknow-
ledged.
REFERENCES
1. Menzies, J.L, Nature, 179: 1142, 1957.

2. Sinks, J., Anat. Rec., 121: 367, 1955.
3. Blakeslee, A.F., Yearbook Carnegie Institute (Washington),
!:J...: 42, 1928.
4. Beadle, J.W., CQ~nell Univ. Agriculture Experiment Station
Mem., 129: 1, 1930.
5. Sandler:-L., Lindsley, D., Nicoletti, B. and Trippa, G.,
Genetics, 60: 525, 1968.
6. Donald, H.P. and Hancock, J.L., J. Agric. Sci., 43: 178, 1953.
7. Williams, W.W., J. Urology, 64: 614, 1950.
8. Kagan, C.A. and Kirov, S.M.,-Urologica, 28: 34, 1963.
9. Ross, A. and Kerr, M., Personal communication, 1969.
10. Pearson, P.L., Personal communication, 1969.
11. Ford, C.E. and Hamerton, J.L., Nature, (London), 178: 1020,
1956. -
12. Kjessler, B., Monographs in Human Genetics, 1: Ed. Karger, S.,
Basel, 1966.
13. Ferguson-Smith, M.A., Lennox, B., Mack, W.S. and Stewart, J.S.
S., Lancet, ii: 167, 1957.
14. Court Brown, W.M., Jacobs, P.A. and Brunton, M., Lancet ii:
561, 1965.
15. McIlree, M.E., Price, W.H., Court Brown, W.M., Selby Tulloch,
W., Newsam, J.E. and MacLean, N., Lancet, 1i 69, 1966.
164 A. C. CHANDLEY
16. Evans, E.P., Breckon, E. and Ford, C.E., Cytogenetics, 2: 289,

1964.
17. Smith, K.D., Steinberger, E., Steinberger, A. and Perloff, W.
H., Cytogenetics, ~: 219, 1965.
18. Hulten, M., Lindsten, J., Fraccaro, M., Mannini, A., and
Tiepolo, L., Lancet, ii: 22, 1966.
19. Armendares, S., Buentello, L., Cuevos-Sosa, A. and Cantu-Garza,
J., Cytogenetics, 8: 177, 1969.
20. Mukerjee, D. and B~rdette, W.J., Amer. J. Hum. Genet., 18: 62,
1966.
21. Walker, H. and Harris, R., Brit. Med. J., 2: 25, 1962.
22. Hamerton, J.L., Cytogenetics, 7: 260, 1968.
23. Jacobs, P.A. and Ross, A., Nat~re, (London), 210: 352, 1966.
DISCUSSION
LUNENFELD: In the last five years, quite a number of laboratories

have become interested in studying mitosis and meiosis in testicular
biopsies. Now I wonder whether you could speculate, based on avail-
able data today, what is the percentage of genetic factors in in -
fertility?
CHANDLEY: By genetic factors, you mean gene mutation, the kind of

case I was showing in the desynaptic male. Well, to my knowledge,
this is the only one which has been found in human meiosis. Perhaps
Dr. Ford or Dr. Ohno know of others?
OHNO: Speaking of this achiasmate mutant, I assume that he was a

homozygote. It follows then that both parents are obligatory hetero-
zygotes. I wonder if you looked at his father to see what the het-
erozygous condition was like.
CHANDLEY: I am not sure whether the parents were looked at but I

do not think they were.
E. STEINBERGER: Dr. Chandley, you have shown a karyotype 46 plus

extra metacentric chromosome in a patient with oligospermia. About
5 years ago we published a case almost identical to yours, and in-
terestingly enough in our case the patient's mother had an identical
karyotype. My question is, have you examined the parents of this
patient? Since in our case the patient's mother has this extra
chromosome and obviously was fertile, the presence of this extra
chromosome cannot interfere with fertility, at least in the female.
Could the oligospermia in this patient be simply coincidental to
the cytogenetic finding? An as a follow-up to this question how
many individuals with perfectly normal sperm count and fertility
did you examine and look for abnormalities in the meiotic metaphaseZ
CHANDLEY: In reply to the first question, as far as I can recall,

neither of the parents of this patient have been examined, but as
I said, the brother of the patient showed it. Now whether this has
any relationship to the infertility is certainly not clear at all,
because there have been reports of supernumary chromosomes in a
variety of patients who have been perfectly normal and others who
showed a variety of pathological conditions.
E. STEINBERGER: Was the brother oligospermic?
CHANDLEY: Yes, the brother was also oligospermic, but we have not
yet taken a testicular biopsy; in fact, he would not allow us to do
so.
PAULSEN: Sergovich from Toronto has published a survey which in-

dicates that DID translocation is one of the most common chromosomal
abnormalities, appearing in one-tenth percent of the general popu-
lation. Thus, it is important for us who are not geneticists to
understand the relationship or relevance of DID translocation to the
pathogenesis of oligospermia. We recently studied a patient with
oligospermia. His sperm count ranged from three to fifteen millionl
ml. His serum and urinary FSH levels were normal. In his testis
we found DID 13-15 translocation in mitotic and meiotic chromosomes.
Histologically the testis revealed a decrease in numbers of late
pachytene spermatocytes and spermatids. I would assume from what
has been said that this represents an unbalanced chromosomal abnor-
mality which has led to a loss of germ cells during meiosis. Would
this be a correct interpretation?
CHANDLEY: Kjessler suggested in his DID translocation that the oligo-

spermia which he detected was in fact due to the loss of the unbal-
anced sperm in spermiogenesis, but whether this is so, we do not
know. Dr. Ford would argue that unbalanced sperm are perfectly cap-
able of fertilizing eggs but they may then cause a breakdown in the
zygote. Perhaps Dr. Ford would like to comment at this point.
FORD: For 15 years or so I have been testing all information that

came my way for the possibility that unbalanced sperm were not cap-
able of functioning in fertilization and I have not encountered any
contrary information. Now the fact that I have never encountered
any does not exclude the possibility that in certain circumstances
if may occur. The next point I would like to make is that Robert-
sonian translocations, like the human DID trans locations, would ap-
pear to have played a major part in karyotype evolution of the mam-
mals. Indeed there are several species in which there is a balanced
polymorphic system involving translocations of that form. A common
shrew in Europe, for example, may be heterozygous for as many as four
of these changes at one time and such individuals must obviously have
a nearly normal fertility or the selection pressure operating against
them would be enormous. Now this is not to deny the information from
our own species. I would appear to be a matter of what specific ge-
netic material is involved and what the nature of the genetic lesion
is.
166 A.C.CHANDL~
ROSS: Is it not true that the male may be the carrier of these DID
trans locations, transmitting them to his progeny? If this is true
then certainly not all spermatozoa containing the abnormal chromo~
somes can be eliminated.
CHANDLEY: Well, transmitting the translocation to his progeny does

not mean that the progeny are unbalanced. If the progeny are car~
rying the translocations, then they themselves are balanced. Is
that what you meant?
ROSS: No, the sperm necessarily must have carried the translocated
chromosomes, if he transmitted it to his offspring.
CHANDLEY: But in a balanced fO~1ffi. If the offspring is a carrier

himself, this is a balanced condition. The offspring of DID trans~
locations occur in the ratio of 1:1 normal to balanced, in those
few cases in which offsprings have been shown. In fact, very few
carriers of DID translocations have been shown to have live off~
spring.
JIRASEK: I would like to refer to the pairing between X and Y chro~

mosomes. In the karyotypes it seemed to me that X and Y chromosomes
were paired end to end.
CHANDLEY: From a meiotic preparation, it really is not possible to

say exactly what the nature of the X~Y pairing is. If you want to
know whether you can see a chiasma configuration or anything like
that, you certainly can not.
SOLARI: I think that from light microscopic evidence, we cannot be

certain about the existence of a chiasma between the X and the Y.
But from our electron microscopic studies, we can say that there is
strong evidence for normal, partial synapsis between the X and Y
chromosomes, in the rat, in the mouse and in the human. If there
is a chiasma, this is now an open quesLion to be really proved with
strong evidence which I think will be provided very soon.
BERGADA: I just want to emphasize what Dr. Chandley said about the
incidence of chromatin positivity in man. In an article published
in Acta Helvetica Paediatrica last year, we indicated that one in
every 80 boys with cryptorchidism is chromatin positive; and this
is really a very high incidence. Furthermore, in a recent survey
performed in Buenos Aires during medical examination of candidates
for military service, approximately one in every 500 normal men
was chromatin~positive. So this is something that we have to bear
in mind when we study the infertile patient.
CHANDLEY: I should like to present two more cases worthy of note.

The first was a 37~year~old male who presented at the subfertility
clinic and who was found, on a semen analysis, to be completely azo-
ospermic. His mitotic karyotype showed no obvious abnormality. His

testicular histology showed numerous dividing cells but a deficiency
of spermatids and a complete absence of mature sperm.
Meiotic investigations, however, revealed that he was the carrier

of a balanced translocation between a C group chromosome and either
an E or an F group chromosome (since E and F bivalents often appear
alike in size in meiotic preparations it was difficult to reach a
decision on this point).
An examination of 35 spermatocytes at diakinesis showed that 21 nor-

mal bivalents, including the XY bivalent, plus a quadrivalent, were
present in the majority of cells. Few 2nd metaphases were observed,
and it seemed that meiotic arrest was occurring in most cells after
diakinesis.
The second case was a 39-year-old azoospermic male who was found to
be the carrier of a reciprocal translocation between a C and a G
group chromosome. His mitotic karyotype, and that of his mother
showed 46 chromosomes including three normal G chromosomes and an
abnormal chromosome about the size of a G group chromosome. His
testicular histology showed that the tubules contained a high pro-
portion of dividing cells, but in most tubules, the end stages of
apermiogenesis were deficient, and few mature sperm could be seen •
Meiotic studies revealed that the majority of spermatocytes at di-

akinesis had 21 normal bivalents, including the XY bivalent, and
either a quadrivalent, which was in the form of a long chain, or a
trivalent plus a univalent. Chiasma counts were made on 20 diakineses
and showed a mean chiasma frequency of 41.5 compared with a frequency
of 50-55 for normal males. The breakdown of meiosis and azoospermia
in these two patients is undoubtedly related to the presence of the
translocations.
CHAPTER IV

SECTION 1: ROLE OF THE HYPOTHALAMUS
MECHANISM OF ACTION OF FSH AND LH RELEASING FACTORS
Irving I. Geschwind
Department of Animal Science
University of California, Davis
The mechanism of action of the hypothalamic releasing factors

has been a subject of intense study during the past three years.
For such studies, either crude hypothalamic or stalk-median emi-
nence extracts or more purified releasing factors have been employ-
ed. Pure releasing factors for the pituitary gonadotropins, FSH
and LH, in which we are interested, have not as yet been isolated,
and indeed there is some question whether separate releasing fac-
tors exist (see, for example, White [lJ).
Initially, the approach to the study of mechanism of action

reflected contemporary involvement with inhibitors of peptide bond
synthesis or of DNA-dependent RNA synthesis. The particular para-
meter investigated was gonadotropin release. Subsequently, limit-
ed explorations were made of the metabolic or energetic require-
ments for hormone release, but by far the greatest effort in this
area has gone into investigations of the roles which various cations
play in the process. The other major parameter which has been the
subject of experimentation is gonadotropin synthesis. Whereas the
isolated rat pituitary incubated for a few hours in appropriate
medium has been the substrate of choice for demonstrating modifi-
cations of gonadotropic hormone release, longer term incubations
and organ culture of the pituitary have been widely employed for
studies of the control of pituitary hormone synthesis. In order
to demonstrate hormone release, a specific assay for the gonado-
tropin in question must be employed. Generally, bioassays have been
used, but more recently immunoassays have come to the fore and are
increasingly replacing bioassays due to the considerable advantages
of the former. Immunological techniques are also employed in the
study of hormone synthesis, with specific antibodies being used to
171
172 I. I. GESCHWIND
selectively precipitate the desired hormone after a gland has been

incubated in the presence of a radioactive amino acid or monosac-
charide (the gonadotropins are glycoproteins). However, most of
the evidence purportedly demonstrating de novo synthesis of hormon-
al protein has been acquired by bioassay of the total amounts of
hormone in experimental and control systems (gland plus incubation
medium). When the amount of hormone in the experimental system
significantly exceeds that present originally in the gland or in
the control system, de novo synthesis has been assumed. The limi-
tations of these assumptions will be discussed below.
The purpose of this paper is to present a review of our know-

ledge of the mechanism of action of LH- and FSH-releasing factor(s).
Much of the information has already been analyzed and compared with
what is known of the mechanism of action of the other hypothalamic
hypophysiotropic factors and of secretagogues for other endocrine
and exocrine glands (2,3). The comparisons effected have stressed
the details of the release process. In this paper greater atten·
tion will be given to the synthetic process.
For reasons of economy no references will be given for studies

of non-gonadotropic hormones or proteins which have been previous-
ly reviewed. The reader is referred to two reviews (2,3) for the
appropriate references to the original literature.
HORMOOE RELEASE
It is now accepted that a primary action of the hypophysio-

tropic factors is modification of release of anterior pituitary and
intermediate lobe hormones. There is also general agreement that
the modification imposed on the release of gonadotropins is a
stimulatory one, as is evident from the increases in circulating
gonadotropins found in man and other animals after administration
of the factors, and from the increases in the. amounts of hormone
released into the medium upon incubation of pituitary glands in
the presence of the factors. The mechanics of the release process
have not been elucidated specifically for the gonadotropins, but
there is no reason to believe that they differ from what has been
described for other proteins in involving a process of reverse
pinocytosis, emiocytosis. The most recent studies of this process
are those of Amsterdam et ale (4) on the secretory cycle of the
rat parotid gland. Only a limited number of ultrastructural stud-
ies of releasing factor action have been reported, and all have
been restricted to the action of the growth hormone releasing factor.
The Role of De Novo Protein and RNA Synthesis in the

Release Process.
A number of studies have been carried out in which the effects

ROLE OF THE HYPOTHALAMUS 173
on the release process of the administration of puromycin, an

inhibitor of protein synthesis, or actinomycin D, an inhibitor of
DNA-dependent RNA synthesis, have been determined (Table 1). When
Table 1. Dependence of In Vitro Releasing Factor-Stimulated

Pituitary Gonadotropin Release on Continuing Protein or
RNA Synthesis in the Rat Anterior Pituitary
Hormone Effect on
Released Inhibitor Release Reference
1. LH Puromycin (100)* 46% Normal (5)

" (200) Normal ( 6)
" (80) " (7)
Actinomycin D (10) " (5)
" " (80) " (7)
2. FSH
" " (1-2)**
Puromycin (100)
" (8)
61% Normal (9)
" (10)
Actinomycin D (10)
Prevented (10)
79% Normal (9)
" " ( 2) "No block" (11)
" " (10) Prevented (10)
" " (1)** Unaffected (11)
*Concentration expressed as ~g per ml medium. ** In vivo, rather
than in vitro. Dose expressed as ~g per g body weight.
no effects of inhibitor administration are to be found, one can

state with reasonable certainty that continuing protein or RNA syn-
thesis is not reqUired for the release process. When modification
of release does occur in the presence of inhibitor one must consid-
er the possibility that the inhibitor is not restricted in its
action to the process being studied. Nevertheless, because most of
the experiments reported are for the in vitro situation, and be-
cause the experiments with LH and FSH, although frequently perform-
ed under essentially similar conditions, lead to different results,
the results of the experiments can be accepted as consequences of
the effects which the inhibitors have on the two synthetic processes.
LH release from pituitaries incubated with hypothalamic extract or
purified LRF plus an inhibitor is obviously not affected by the in-
hibitors, and similar results have been found for TSH release. On
the other hand, FSH release is extensively, if not completely in-
hibited. In a kinetic study Jutisz and de la Llosa have demon -
strated that in the presence of cycloheximide, an inhibitor of pro-
tein synthesis, FSH release is essentially normal for the first
half hour, and thereafter little release takes place (12). A sim-
ilar partial inhibition has been seen in studies of antibody and of
insulin release. It would seem most reasonable to assume that the
continuing protein synthesis reqUired for release of FSH, as for
these other two non-pituitary proteins, involves synthesis of the
174 I. I. GESCHWIND
secreted proteins. If indeed FSH synthesis is required, the fact

that the considerable amount of preformed or storage hormone pre-
sent in the gland could not be released would indicate that this
latter hormone is not part of a readily-releasable pool, but in-
stead turns over much more slowly than hormone that can be released.
It is not apparent, however, why the control of FSH release should
differ from that of LH release in this important aspect.
The Role of Metabolic Energy in the Release Process
Earlier experiments (2,3) carried out on the release of several

proteins, especially enzymes from exocrine glands, had suggested
that inhibitors of oxidative phosphorylation inhibit release. How-
ever, more recently it has been reported that in catecholamine re-
lease from perfused bovine (13) and cat (14) adrenals either gly-
colysis or oxidative phosphorylation could supply the energy re-
quired for release. Both processes had to be inhibited to prevent
release of the hormone in response to stimulation, and carbamyl -
choline-induced chymotrypsinogen release from the incubated rat
pancreas was unaffected by the presence of 8 ~g/ml oligomycin, an
inhibitor of oxidative phosphorylation (15). Only a single study
has been reported on the effect of metabolic inhibitors on release
of LH, as stimulated by crude hypothalamic extract (16). Neither
2 x 10-4M dinitrophenol nor 1 x 10-3M cyanide affected release,
whereas 0.1 ~g/ml of oligomycin produced a partial inhibition. An
inhibitory effect of oligomycin has also been observed by others
studying TSH release, but the reported effects of dinitrophenol
on TSH release have been variable. Insensitivity of the release
process to the presence of cyanide has also been found for TSH,
as well as for catecholamines (13,14). In all other systems in-
vestigated, however, both dinitrophenol and cyanide have been
found to inhibit release, and we have no idea why the process in
the pituitary should be so different in its sensitivity to these
inhibitors.
A partial list of those sites in the release process which may

be energy-dependent is found in Table 2. The background for each
of these suggestions has been given elsewhere (3), and I would like
to add some information about two points only now. One is that
although others had found that ATP itself promoted release of cate-
cholamines, vasopressin, and oxytocin from their respective isolated
secretory granules, it has failed to promote insulin release from
isolated ~ -cell granules (17). The other concerns adenyl cyclase
and its product 3', 5'-cyclic adenylic acid (cAMP). The latter, and
its derivative 0', N-dibutyryl cyclic AMP (DBC), have been demon-
strated to promote the release of parotid and pancreatic amylases,
insulin, growth hormone, TSH, and ACTH. Moreover, TRF had been shown
to elevate pituitary cAMP levels. Although none of these early (i.e.,
1 to 3 years old) reports dealt with gonadotropins, there was no
reason to believe that what had been found for the other pituitary
hormones would not apply to the gonadotropins as well.
Table 2. Steps in the Release Process Which May be Energy-

Dependent.*
1. De novo protein synthesis, when required (e.g., for FSH

release).
2. Protein transport between rough endoplasmic reticulum
and Golgi complex.
3. Fission-fusion of cell and granule membranes.
4. Modification of cell membrane structure, brought about by
a phospholipase upon activation of the latter.
5. Conformational change in cell membrane promoted by the
interaction of ATP and membrane-bound ATPase.
6. Postulated microtubular, microfilamentous system, which
links secretion granules to the cell membrane and which upon con-
traction promotes fusion of the granule and cell membranes.
7. Membrane-associated adenyl cyclase.
*The reader is referred to ref. 3 for a more extended discussion
of each of these suggestions.
Within the past year we have seen reports that both cAMP and
DBC stimulate FSH release from the isolated rat pituitary (18),
that theophylline" an inhibitor of the phosphodiesterase which
hydrolyzes cAMP, while ineffective by itself at a 2.5 roM concen-
tration, did potentiate the effect of FSH-RF on release (18), and
that a hypothalamic extract which promoted a 3-fold increase in
LH release during a 6-hour incubation of rat anterior pituitary
glands, concomitantly stimulated a two-fold increase in cAMP and
adenyl cyclase levels in the gland (19). Such extracts were spe-
cific for the anterior pituitary, and the increases in cAMP levels
could be observed within three minutes of incubation of the
glands and increased with the duration of incubation. Moreover,
the effect of extract on adenyl cyclase levels was already marked
one minute after its addition to pituitary homogenates (19).
These latter results of course do not prove a cause-effect rela-
tionship between elevated cAMP levels and LH release, since a
crude hypothalamic extract and entire anterior pituitary glands
were employed. Nevertheless, the results are consistent with the
developing hypothesis that the releasing factors stimulate adenyl
cyclase activity, and the cAMP formed thereby acts as a "second
messenger" for release. Similar increases in incubated pituitary
adenyl cyclase and cAMP levels have been reported after addition
of a stalk-median eminence extract which stimulated TSH and growth
hormone secretion (20). In these experiments the extract was
shown to have no effect on phosphodiesterase activity.
Finally, it should be emphasized that if this is an energy-

requiring step in the release process, it is not the only one, for
dinitrophenol can block the exogenous cAMP-induced release of
parotid amylase.
176 I. I. GESCHWIND
The Role of Cations in the Release Process
We are all indebted to Douglas for the initiation of studies

on the part cations play in the release process. The many exper-
iments which he and his co-workers performed on the release of
catecholamines and of vasopressin established the following:
Elevated [~J itself induces release, with an optimum [~] reached
at between 10 and 15 times the normal medium concentration, and
for that release and for specific secretagogue-stimulated release
Ca++ is re~ired: in the presence of normal concentrations of the
latter, Mg in elevated concentrations acts as an inhibitor of
release, but the antagonism can be reversed by elevating the
[Ca++]: the response to a secretagogue increases with [Ca++] up
to a certain optimum concentration of the latter; and depolariza-
tion of cell membranes by secretagogues is not a sufficient stim-
ulus for the extrusion process. A theory of "stimulus-secretion
coupling" was advanced. The theory, which has found many advocates,
postulates that a secretagogue modifies the cell membrane so as to
make it more permeable to Ca++ and Na+ , thereby leading to the
depolarization, but it is the entry of Ca++ which triggers the
release process.
Limited studies of the anterior pituitary have established

that elevated [~ ] stimulates release of LH, FSH, TSH, ACTH, and
growth hormone; that for LH a five-fold increase in [K' ] does
not stimulate release (22), but that for both LH and FSH an opti-
mum effect is found with 60 mM~; that the high [~] effect is
additive with the effects of releasing factors on gonadotropin
release (12,16,22); that in the absence of Ca++, both high [~]
and releasing factor-induced release of LH and FSH are inhibited;
and that elevated [Mg++] also inhibits release of the gonadotrop-
ins (Table 3). Table 3 also records the effects of the absence
of Na+, ~, or Mg++ from the medium. Except for the lack of an
effect of Mg++-deficiency, which is consistent with what has been
found for other secretory processes, it is difficult to comment
on the effects produced by deficiencies of the other two cations,
since 0 mM K+ has been found to stimulate insulin and catecho_
lamine release, and to inhibit parotid amylase release, while a
Na+-free medium has been found to inhibit insulin release, and
by one group of investigators, catecholamine release (26), while
another group has found it to potentiate catecholamine release.
The theory of "stimulus-secretion coupling" does not account

for the varied effects of Na+ and K+ deficiency or for the site
of action of Ca++, nor does it concern itself with the activation
of adenyl cyclase by secretagogues, or with the release-inducing
effects of cM-lP and theophylline. The effect of elevated [~]
can best be attributed to depolarization of the cell membranes,
although it is possible that it could act indirectly since high
[K+]-induced increases in cAMP concentrations have been reported
Table 3. Effects of Cations on Pituitary Gonadotropin Release

From the Incubated Rat Anterior Pituitary.
Hormone
Released Stimulation Inhibition* Normal
1. LH 60 roM I(f- o roM Ca+t

(16,21,22) (16,22,23)
20 roM Mg+t (22) o roM Mg+t
(22)**
2. FSH 50-60 roM ~ (12,22) o roM K+ (24.)
o roM Na+ (22) o roM Ca+t (22,25)
20 roM Mg+i- (22) I
*Inhibition of releasing factor and/or high [~J-induced release.

**Effect on high [~J and releasing factor-induced release.
in non-pituitary tissues. Zor and his co-workers (27) have found,

however, that 10-fold [~J did not stimulate an increase in an-
terior pituitary cAMP levels in their experiments.
The relationship between adenyl cyclase - cAMP levels and

[Ca+t] may be quite complex, and recent reports would indicate a
variable effect of Ca+t or its deprivation on adenyl cyclase
activitiesin different tissues. It is known, however, that high
[~]- or epinephrine-provoked increases in cAMP levels in the
parotid gland, and hypothalamic- (20) or stalk-medium eminence-
induced (27) increases in pituitary adenyl cyclase or cAMP levels
are unaffected by the absence of Ca+t from the medium. The ab-
sence of Ca+t, however, is sufficient to prevent endogenous cAMP
from promoting protein release, as well as to inhibit completely
the effect of exogenous cAMP on FSH release (25). Thus, Ca+t is
apparently not needed for cyclase function, and increased intra-
cellular cAMP concentrations are not, of themselves, sufficient
stimulus for release.
The exact function of Ca+t is unknown, and any suggestion of-

fered should take into account the reported stoichiometry between
release and [Ca+t] at suboptimal concentrations of the latter.
Table 4 lists a number of possibilities for the action of Ca++.
The bases for each of these suggestions, other than the last, are
to be found elsewhere (3). The last is considered in the para-
graph which follows.
The mode of action of cAMP is also clothed in mystery. I have

suggested in Table 4 an effect of Ca++ on a protein kinase, and
one may visualize a kinase which may be activated either by Ca+t
or by Ca+t and cAMP. Kuo and Greengard have suggested that a
wide variety of cAMP effects may be mediated by stimulation of
protein kinases (28). Certainly if some stimuli for secretion
178 I. I. GESCHWIND
Table 4. Possible Modes of Action of Ca++*
1. Promotion of the dissociation of a hormone-carrier protein

complex.
2. Production of a conformational change in the macromole-
cules of the granule and/or plasma membranes, perhaps by bridging
between neighboring negatively-charged groups in the membranes,
resulting in a change in membrane reactivity.
3. Activation of a phospholipase, which by modifying mem-
brane lipids alters membrane reactivity.
4. Promotion of the adhesion of the granule to the cell mem-
brane, either by divalent bridging (see 2 above) or by enhancing
the solvation of a component of the granule membrane in the lipid
phase of the plasma membrane.
5. Alteration of the colloidal state of the cytosol, so as
to promote granule movement.
6. Relief of the inhibition of an ATPase, thus promoting
contraction of the postulated actomyosin-like microfilaments link-
ing the granules to the cell membrane.
7. Activation of a protein kinase, which by phosphorylating
a membrane protein either activates an enzyme or produces a can _
formational change in a structural protein. The net result of
either is an altered membrane reactivity or an induced capability
of membranes for fusion.
*The reader is referred to ref. 3 for a more extended discussion of

most of these suggestions.
do not provoke an increase in adenyl cyclase activity and if cAMP

does not stimulate release in the absence of Ca++, some independent
effect of Ca++must be postulated which must be distal to the pro-
duction of cAMP, and yet also involve available cAMP.
HORMONE SYNTHESIS
The methods employed to determine whether hypophysiotropic

factors affect the synthesis of the pituitary hormones whose re-
lease they promote have been mentioned above. The method of demon-
strating synthesis by an increase in the total amount of hormone in
a system has been the one most extensively employed, but it has
been criticized because too often the increments have been so small
as to tax the precision of the bioassays used. Occasionally, as is
reported below, large differences have been found, but then gener-
ally no proof has been offered which would eliminate activation of
a prohormone or inhibition of an intracellular hydrolytic system
as the cause of the increase. The more direct method of quanti-
tating protein or glycoprotein synthesis, by measuring the incor-
poration of a radioactive amino acid or monosaccharide into the
macromolecule, also has its limitations: conditions for protein
synthesis in vitro must be optimal, the intracellular pools of the

amino acid or monosaccharide must be unchanging, and adequate meth-
ods must be available for resolving the macromolecule from any rap-
idly labelled contaminants.
The evidence for stimulation of synthesis by the gonadotropin

releasing factors is given below.
1. In 13 of 17 female rats bearing pituitary autografts be-
neath the kidney capsule for 5 to 96 days prior to continuous per-
fusion of the graft with hypothalamic extract for an additional 6
to 38 days, the presence of PAS-positive cells was noted in the
grafts. Eleven of the 13 rats were also found to have large fol-
licles in their ovaries, and in 9 of these 11 vaginal estrus oc-
curred durring the infusion period (29). Since such grafts are
inactive and probably contained little hormone at the onset of the
infusion, synthesis of sufficient quantities of hormones to pro-
duce the observed effects must be assumed.
2. When normal female rat anterior pituitary is maintained
in organ culture, the addition of hypothalamic extract produces an
increase of the ovarian ascorbic acid depleting activity of the
culture and an increase in the uptake of tritiated leucine by the
PAS-positive cells of the pituitary (30).
3. Incubation of pituitaries derived from castrated, estra-
diol-progesterone treated rats for two hours with an LRF prepara-
tion, led to a sometimes significant increase in the total LH of
the system (23,31).
4. In preliminary studies, an increase in total immunoassay-
able LH has been found only after incubation of rat pituitary
halves with hypothalamic extract (32).
5. Organ culture of normal female rat pituitaries for a
period of three days in the presence of highly purified prepara-
tions of LRF led to a mean increase of 20% in the total LH content
of the system as determined by radioimmunoassay. The greatest in-
creases, 34% and 67%,were produced by the highest concentrations
of LRF (33).
All of the above evidence supports a stimulatory role of re-

leasing factors on LH synthesis.
6. However, the incorporation ot radioactive leucine or
glucosamine into LH was not increased above control values during
4-hour incubations of pituitaries from ovariectomized rats, some
of which were pretreated with estradiol-progesterone, or during
6-or 24-hour incubations of pituitaries from normal rats, or during
6-hour incubations of pituitaries from testosterone-treated rats,
all in the presence of hypothalamic or sta1k-median eminence ex-
tracts or crude LRF (6,34).
It should be noted that investigators using both total hor-

mone and radioactive amino acid incorporation methods are in agree-
180 I. I. GESCHWIND
ment that high [~J does not stimulate the rate of LH synthesis
(16,23,35), although it has been claimed that by the criterion of
the former method FSH synthesis is increased (12).
The data on FSH are far more varied and interesting.

7. In 2 or 4 hour incubations, with a crude FSH-RF, of pitu-
itaries from normal male rats or ovariectomized rats pretreated
with estradiol-progesterone, the total amount of FSH in the system
increased by 16 to 40% (9,1i+25). The increase in the total FSH
activity was found to be Ca -dependent (25), whereas others had
previously claimed that the incorporation of radioactive leucine
into LH was Ca++-independent (16). Rather interestingly the au-
thors (25) state "that synthesis of FSH or activation of a pre-
cursor of FSH may take place in the presence of FRF", although in
an earlier publication (12) they had stated that since FRF had no
effect on homogenates it could not be activating a precursor.
8. Under conditions identical with those given in 5 above,
the presence of an LRF stimulated a mean increase in total immuno-
assayable FSH of 106% (33).
9. Rat stalk-median eminence extract injected intrajugularly
into intact mature male rats 45 minutes after an initial dose of
the extract, at a time when depletion of pituitary FSH was maximum,
stimulated the reaccumulation of pituitary FSH, so that within two
hours the pituitary FSH content was more than 30% greater than the
initial, pre-injection levels (36). Moreover, in mature male rats
with median eminence lesions placed seven days earlier, in which
the pituitary FSH concentration is about 65~~ of normal, injections
of such extracts produce a 75% increase ~in pituitary FSH concen-
tration 45 minutes later (37), and a log dose-response relation-
ship for such "FSH-synthesizing factor" activity has been obtained
(38). Corbin has stated that "the hypothalamus controls both the
synthesis and release of FSH and whether or not synthesis or re-
lease of FSH occurs appears to depend on the initial level of FSH
in the pituitary" (38).
Corbin's findings are especially interesting, particularly if

the reported increases in pituitary FSH levels do indeed represent
synthesis of the hormone, for the question of the control of pro-
tein synthesis by the hypothalamic factors has been particularly
vexing. Although the bioassay evidence clearly indicates increases
in total system hormone for many of the pituitary hormones when
glands are incubated, or preferably cultured, in the presence of
crude or purified releasing factors, incorporation studies have
failed to support such claims, with a possible exception in the
case of ACTH. Such a discrepancy could result from any of the pre-
viously mentioned shortcomings of the two methods of analysis.
If it is assumed that synthesis is indeed stimulated, then one

may ask whether the effect is a primary one, or whether it is sec-
ondary to release. Although either release or synthesis can be
uniquely inhibited (e.g., for LH by Ca++.deprivation or by puro-

mycin, respectively [16J), there is no clear-cut evidence available
from studies on pituitary hormones to indicate whether an increase
in the rate of synthesis may be tied to release. Corbin's findings
suggest that it may not. From studies on insulin secretion in in
vitro or perfused systems, there is evidence that substances such
as ribose, xylitol, and tolbutamide may promote release without en-
hancing synthetic rates (39), so that there need not be any link
between the two processes. Recently, Grand and Gross (40) have
attempted to resolve the question by studying the effect of epi-
nephrine, a secretagogue for amylase release from the parotid gland,
on amylase synthesis. In slices of the gland, epinephrine produces
more than a two-fold increase in the incorporation of radioactive
amino acids into amylase during the period of active enzyme extru-
sion into tQe medium. Subsequently, the rate of protein synthesis
declines as enzyme release returns to near its original rate, and
if epinephrine is again added during this period, the rate of in-
corporation again doubles, but at this point there is no change in
the rate of amylase secretion.
Samli and Geschwind(6) originally raised the possibility that the

release process itself, as stimulated by a releasing factor, could
lead to feedback mechanisms allowing the synthesis of additional
pituitary hormone, a view also espoused by others (e.g., 23), and
more recently I have suggested (2,3) that cAMP may serve to link
release and synthesis. This could be accomplished if cAMP has a
direct, independent effect on both processes; if activation of
cell membrane-associated adenyl cyclase causes a Ca++-dependent
change in membrane activity, while the cAMP formed stimulates
protein synthesis; or if cAMP in some fashion affects the cyto-
structure, the alteration of which then leads independently to both
secretion and synthesis. Much of the evidence supporting a role of
cAMP in selected protein biosynthesis, especially in microorganisms,
has been reviewed (3). To that evidence we may add the following
additional support derived from experiments in higher organisms:
cAMP stimulates the protein synthetic capacity of thyroid polysomes
and monosomes in vitro (41); cAMP and DBC induce tyrosine transam-
inase in explants of fetal rat liver maintained in organ culture,
the induction process being inhibited by cycloheximide (42); and
DBC stimulates the incorporation of amino acids into parotid amy-
lase at the same time that it provokes release of the enzyme (40).
Directly pertinent to our considerations is the report that when
pituitaries are incubated for 4 hours in the presence of cAMP, the
total FSH activity of the system increases by about 35% (23).
These considerations suggest as a working hypothesis that a

secretagogue stimulates both processes by primarily activating cell
membrane-associated adenyl cyclase; The hypothesis, in the form
presented in the paragraph above, could account for both Corbin's
and Grand and Gross' results if the following stipulations are
182 I. I. GESCHWIND
accepted: when release is occurring from a pituitary in which the

initial level of FSH is high, synthesis is taking place simultan-
eously, as the inhibitor studies would suggest; and in a cell (par-
otid or pituitary) from which most of the readily-releasable pro-
ducts have been discharged (e.g., amylase or FSH), neither cell
membrane nor cytostructure alt'eration nor cAMP itself can promote
much further release of product, but protein synthesis is still
capable of being stimulated. The hypothesis, however, must allow
for secretion to occur by more than a single process, in order to
accomodate, for example, the finding that compounds such as ribose
and tolbutamide can only stimulate insulin release and not its
synthesis. Furthermore, it appears that at times the cAMP effects
can be mimicked by high [~], (e.g., the reported increases in FSH
synthesis and release), although both the failure of high [K+] to
promote increases in adenyl cyclase and cAMP levels in the pitu-
itary, and the additive effect on release of high [K+] and LRF or
FSH-RF, would suggest that elevated [~] and the releasing factors
are not operating through the same mechanism, i.e., activation of
adenyl cyclase.
ACKNOWLEDGMENTS
The research from my laboratory which is reported in this re-

view was supported by USPHS Research Grant HD-00394.
I am indebted to Drs. Alan Corbin, Marian Jutisz, S.M. McCann,

Andrew V. Schally, and Wilfrid F. White for their kindness in mak-
ing findings from their laboratories available to me prior to pub-
lication.
REFERENCES
1. White, W.F. In Workshop Conference on the Bioassay and Chem-

istry of the Hypophysiotropic Hormones of the Hypothalamus,
ed. Meites, J., Williams and Wilkins, in press, 1970.
2. Geschwind, 1.1. In Frontiers in Neuroendocrinology, 1969
eds. Ganong, W.F. and Martini,L., Oxford University Press
p. 389, 1969.
3. Geschwind, 1.1. In Workshop Conference on the Bioassay and
Chemistry of the Hypophysiotropic Hormones of the Hypothalamus,
ed. Meites, J., Williams and Wilkins, in press, 1970.
4. Amsterdam, A., Ohad, I., and Schramm, M., J.Cell Bioi., 41:
753, 1969.
5. Jutisz, M., Berault, A., Novella, M.A., and Chapeville, F.
C. R. Acad. Sci. (Paris), 263: 664, 1966.
6. Samli, M., and Geschwind, 1.1. Endocrinology,~: 835, 1967.
7. Crighton, D.B., Watanabe, S., Dhariwal, A.P.S., and McCann,
S.M. Proc. Soc. Exp. Bioi. Med., 128: 537, 1968.
8. Schally, A.V., Bowers, C.Y., Carter, W.H., Arimura, A.,
Redding, T.W., and Saito, M. Endocrinology, 85: 290, 1969.
9. Jutisz, M., and de la Llosa, M.P. Endocrinology jU: 1193,

1967.
10. Watanabe, S., Dhariwal, A.P., and McCann, S.M. Endocrinology,
82: 674, 1968.
11. Schally, A.W., Saito, T., Arimura, A., Sawano, S., Bowers,
C.Y., White, W.F., and Cohen, A.I. Endocrinology, 81: 882
1967.
12. Jutisz, M., and de la Llosa, M.P. Bull. Soc. Chim. BioI.,
50: 2521, 1968.
13. Kirshner, N., and Smith, W.J. Life Sci.,~: 799, 1969.
14. Rubin, R.P. J. Physiol., 202: 197, 1969.
15. Bauduin, H., Colin, M., and Dumont, J.E. Biochim. Biophys.
Acta, 174: 722, 1969.
16. Samli,~, and Geschwind, 1.1. Endocrinology, 82: 225, 1968.
17. Howell, S.L., Young, D.A., and Lacy, P.E. J. CeIl BioI.,
41: 167, 1969.
18. SUtisz, M., and de la Llosa, M.P. C. R. Acad. Sci. (Paris),
268: 1636, 1969.
19. ~,U., Kaneko, T., Schneider, H.P.G., McCann, S.M., Lowe,
I.P., Bloom, G., Borland, B., and field, J.B. Proc. Nat.
Acad. Sci. USA, 63: 918, 1969.
20. Steiner, A.L., Peake, G.T., Utigar, R., and Kipnis, D.
J. Clin. Invest., 48: 80a, 1969.
21. Vale, W., and Guillemin, R. Experientia,23: 855, 1967.
22. Wakabayashi, K., Kamberi, I.A., and McCann, S.M. Endocrin-
ology, 85: 1046, 1969.
23. Jutisz, M:, de la Llosa, M.P., Berault, A., and Kerdelhue,
B. Proceedings of the "Colloque de Neuroendocrinologie"
of the C.N.R.S. (Paris), in press.
24. Jutisz, M., and de la Llosa, M.P. Excerpta Medica Interna-
tional Congress Series, l2l: 137, 1968.
25. Jutisz.M., and de la Llosa, M.P. Endocrinology, in press,
1970.
26. Banks, P., Biggins, R., Bishop, R., Christian, B., and
Currie, N. J. Physiol., 200: 797, 1969.
27. Zor, U., Kaneko, T., Schneider, H.P.G., McCann, S.M., and
Field, J.B. J. Clin. Invest., 48: 93a, 1969.
28. Kuo, J.F., and Greengard, P. J. BioI. Chern. 244: 3417, 1969.
29. Evans, J.S., and Nikitovich-Winer, M.B. Neuroendocrinology,
4: 83, 1969.
30. Kobayashi, T., Kobayashi, T., Kigawa, T., Mizuno, M.,
Amenomori, Y., and Watanabe, T. Endo. Japon.,ll: 430, 1966.
31 Jutisz M., Berault, A., Novella-M.A., and Ribot, G. Acta
Endocr. (Kobenhavn), ~: 481, 1967.
32. Midgley, A.R., Jr., Gay, V.L., Caligaris, L.C.S., Rebar, R.W.,
Monroe, S.E., and Niswender, G.D. In Gonadotropins 1968,
ed. Rosemberg, E., Geron.X, Los Altos,Calif, p 307, 1968.
33. Mittler, J.C., Arimu~a, A., and Schally, A.V., personal
communication.
34. Wakabayashi, K., and Seldin, D.W. Fed. Proc., 28: 831, 1969.
184 I. I. GESCHWIND
35. Wakabayashi, K. Personal cormnunication from S.M. McCann.

36. Corbin, A., and Daniels, E.L. Experientia, 24: 1260, 1968.
37. Corbin, A. In Workshop Conference on the Bioassay and
Chemistry of the Hypophysiotropic Hormones of the Hypothal-
amus, ed., Meites, J. Williams and Wilkins, in press, 1970.
38. Corbin, A., personal cormnunication.
39. Taylor, K.W., presented at the Third International Congress
of Endocrinology, Mexico City, 1968.
40. Grand, R.J., and Gross, P.R. J. Bioi. Chern., 244: 5608, 1969.
41. Lissitzky, S., Mante, S., Attali, J.C., and Cartouzou, G.
Biochem. Biophys. Res. Cormnun., 35: 437, 1969.
42. Wicks, W.D. Science, 160: 997, 1968.
DISCUSSION
MEANS: Dr. Geschwind, as you probably know, if cyclic AMP is going

to do all the things that it does in all of these various target or-
gans, somewhere along the line some specificity must be invoked. If
increases in protein and nucleic acid synthesis occur there has to
be some manner of control. Dr. Jensen has recently reported that
calcium is very necessary for the stability of the estrogen receptor
molecule. If we couple this with L. Garren's findings that in the
adrenal gland there may be a protein that in fact specifically binds
cyclic AMP, this could in fact impart a means of specificity. If
this occurs, then this might also provide a reason why one cell type
with two distinctly different hormones acting upon it could possibly
exhibit differential effects if the receptor sites for cyclic AMP
were either different, or if the receptor proteins for the cyclic
AMP were different.
GESCHWIND: One is dealing with a release process and I have pointed

out that a high potassium ion concentration also requires calcium
ion in order to stimulate release. Thus, the consideration of re-
ceptors is important and one has to look for something much more
basic. I have also pointed out that if cyclic AMP is already pre-
sent, then calcium ion is also required in order to effect release.
Everyone who has studied release processes from both endocrine or
exocrine glands has always found the same thing; calcium is abso-
lutely necessary for the release process.
HELLER: I would ask you to differentiate between pharmacological

effects of hormones which can invoke feedback mechanisms and physi-
ological effects which can be as simple as utilization of FSH and
ICSH by the testis. I am sure I will be challenged by some of you
on this, but please let us remember that two mechanisms may be in-
volved.
SAVARD: I would like to speak on this matter of adenyl-cyclase. I

believe that the concept of cyclic AMP with its apparently ubiquitous
presence in so many cellular systems, is entirely acceptable, but

only if one invokes morphologic considerations. This, I admit,
might be considered heretical for a biochemist to express, but never-
theless it makes sense. Sutherland himself has suggested that the
system of adenyl cyclase - 3'S'-AMP is a very primitive control sys-
ten, which may have developed very early in biochemical evolution.
One of these may be the activation of phosphorylase; there may be
others which are very basic to cellular rnetabolism. With this as
a primitive control system, at the simple unicellular level, as more
sophisticated systems evolved, their mechanism utilized this basic
cellular system of adenyl cyclase. Consequently, we must introduce
into our thinking the idea that perhaps each cell has its own adenyl
cyclase which is activated by secondary, but more sophisticated con-
trol mechanisms. These latter have the so-called specificity pro-
perties, but in each case they may activate a particular cell's
adenyl cyclase. The process then becomes strictly an intracellular
one with the generated cyclic AMP exerting its stimulatory message
to whatever particular process the cell is programmed for. The
point I wish to make is that neither adenyl cyclase nor cyclic AMP
should be considered as circulating or peripheral factors otherwise
all endocrine specificities are lost. These agents must be viewed
as strictly intracellular agents, waiting to be turned on by a spe-
cific extracellular factor (a hormone for instance) and serving to
stimulate a process unique to the particular cell involved. In this
way it makes sense, at least to me.
LIPSETT: Dr. Pastan at the National Cancer Institute recently has

shown that in a soluble system derived from E. coli, cyclic AMP pro-
motes the synthesis of inducible enzymes. In particular, with res-
pect to ~-galactosidase, he can localize this effect to the promoter
gene of the lac operon. Thus here is a direct effect of cyclic AMP
on transcription localized at a particular site. Even though it is
in a bacterial system, it may be relevant to our considerations of
cyclic AMP and the synthesis of enzymes necessary for steroid syn-
thesis.
HYPOTHALAMIC CONTROL OF GONADOTROPIN SECRETION IN THE MALE
L. Martini
Department of Pharmacology, University of Milan, Italy
I. LOCALIZATION OF THE NUCLEI WHICH SYNTHESIZE THE GONADOTROPIN-

RELEASING FACTORS.
Thcee groups of experiments have been performed in order to

clarify which regions of the hypothalamus synthesize the releasing
factors (RF) controlling the secretion of pituitary gonadotropins
in the male rat.
A. Hypothalamic Lesions
The first technique used was that of placing separate electro-

lytic lesions in each of the hypothalamic areas which are believed
to play some role in the control of gonadotropin secretion (1), and
studying whether such lesions modify the concentrations of FSH-re-
leasing factor (FSH~RF) and LH-releasing factor (LH-RF)l at the
level of the median eminence (ME), i.e. of the region of the hypo-
thalamus which is specifically responsible for the storage of hypo-
physiotropic principles, and for their delivery to the anterior
pituitary through the portal circulation. It was assumed that a
lesion placed exactly in an area in which a RF is produced would
prevent its synthesis, and consequently reduce its accumulation in
the ME.
1. When dealing with the male, the name of Interstitial Cell Stim-
ulating Hormone-Releasing Factor (ICSH-RF) would probably be more
appropriate; however, the abbreviation LH-RF) will be used through-
out this paper, because it is the only one usually employed in the
neuroendocrine literature.
187
188 l. MARTINI
Three independent areas of the brain were bilaterally lesioned;

they will be referred to as: a) paraventricular area; b) supra-
chiasmatic area; and c) arcuate-ventromedial area. Five days fol-
lowing placement of the lesions the animals were killed and their
ME were collected in order to evaluate their content in FSH-RF and
in LH-RF. For the assays of these principles the "pituitary de-
pletion methods" described by Fraschini et al. (2) and by Motta et
al. (3) were used. Using this approach it has been possible to
localize, within the hypothalamus of the male rat, a circumscribed
region in which FSH-RF is synthesized; it has actually been shown
that lesions in or around the paraventricular nuclei are the only
ones which reduce the concentration of this RF at ME level (Fig. 1).
The synthesis of LH-RF takes place apparently in two different re-
gions, since the content of this RF in the ME is reduced when le-
sions are placed either in the suprachiasmatic area or in the ar-
cuate-ventromedial nuclei (Fig. 1) (4,5).
B. Hypothalamic deafferentation
The second approach devised to study the localization of the

nuclei synthesizing the RF, which control the secretion of gonado-
tropins in the male,was that of evaluating the concentrations of
FSH-RF and of LH-RF in the hypothalamic islands of rats submitted
to a complete hypothalamic deafferentation (6). This operation
permits a total separation of the paraventricular region from the
rest of the hypothalamus (Fig.2); consequently, if it is true that
FSH-RF originates in the paraventricular area, the complete dis-
appearance of FSH-RF from the island a few days after the operation
would be expected. On the other hand, the suprachiasmatic and the
arcuate-ventromedial regions are still included within the hypotha-
lamic island; consequently, if LH-RF is really produced by these
two zones, its concentrations in the island should not be reduced
following the operation.
The content of FSH-RF and of LH-RF in the deafferented island

of adult male rats was measured eight and fifteen days following
the operation, using the techniques described by Fraschini et ale
(2) and by Motta et ale (3). The results shown in Fig. 3 indicate
that FSH-RF stores are significantly reduced in the hypothalamic
island eight days after a complete deafferentation; FSH-RF disap-
pears completely from the isolated hypothalamus fift~en days after
the operation. Eight and fifteen days after deafferentation LH-RF
is still present in the isolated island (Fig.4); surprisingly, the
concentration of this RF in deafferented animals is even higher
than usual (7,8).
These data confirm with a new technique that FSH-RF is synthe-

sized in the paraventricular region and that LH-RF is probably manu-
factured in the suprachiasmatic and in the arcuate-ventromedial re-
gions. The fact that the stores of LH-RF are increased following
ROlE OF THE HYPOTHALAMUS 189
0/0Deplefion of
pituitary tropins
70
&0
•
50
• •
40 •
30
20
10
C PV Sch ArVM C PV Sch ArVM

rSH-Rr LH-RF
Fig. 1. Effect of lesions localized in the paraventricular (PV),
suprachiasmatic (Sch) and arcuate-ventromedial (ArVM)
areas on FSH-RF and LH-RF ac ti vi ty of the median emLlence
of male rats. Columns represent the depletion of pitu~
itary tropins induced in normal male rats by the intra-
carotid injection of hypothalamic extracts prepared from
non-lesioned controls (C) or from animals with the dif-
ferent hypothalamic lesions.
deafferentation suggests that, after the operation, all the LH-RF

which is synthesized is accumulated in the island, because the
extrahypothalamic neural stimuli necessary for its release are no
longer transmitted to the hypothalamus (see below) (7,8).
C. Hypothalamic Implants of Inhibitors of Protein Synthesis
In the third group of experiments, the hypothesis that FSH-RF

might be synthesized in the paraventricular nuclei was tested by
implanting cycloheximide (Actidione), an inhibitor of protein syn-
thesis, into the paraventricular region of adult, castrated, male
rats, and by evaluating the effects of such implants on FSH-RF
stores in the ME. Cycloheximide was implanted either unilaterally
or bilaterally. When unilateral implants were performed, at the
time of autopsy, ME tissue was collecteo in a way which permitted
the separation of the half ME corresponding to the implanted side
190 l. MARTINI
Fig. 2. Frontal section of the rat brain following total hypotha-

lamic deafferentation and removal of the hypothalamic
island o OT = Optic Tract; PV = Paraventricular Nuclei;
V3 = Third Ventricle; HI = removed Hypothalamic Island.
from the half corresponding to the non-implanted one; the two halves
of the ME were then tested separately for their content of FSH-RF
and of LH-RF (2,3).
The data shown in Fig.5 indicate that, five days after uni-
lateral implantation of cycloheximide, FSH-RF disappears only from
the ipsilateral half of the ME; complete disappearance of FSH-RF
from the ME is induced by bilateral implants. It may be concluded
from these results that inhibition of protein synthesis in the cells
of the paraventricular nuclei interfere with some biochemical pro-
cess which is essential for the synthesis of FSH-RF in this region;
these data, however, are not taken as indicating that FSH-RF itself
is a protein or a polypeptide (9). It is also clear from the re-
sults that fibers originating in one paraventricular nucleus do not
cross and carry FSH-RF only to the ipsilateral half of the ME.
Fig. 6 shows once more that the paraventricular region is not

o NORMAL
FSH-RF'
(J.l9 of F'SH
depletion/pit)
150 ~ DEAF'F'ERENTED
•
100
•
50
•
•
8 days 15days
Fig. 3. Effect of hypothalamic deafferentation on the FSH-RF con-

tent of the hypothalamic island of adult male rats (eight
and fifteen days after the operation). Columns represent
the depletion of pituitary FSH induced in normal male rats
by the intracarotid injection of hypothalamic extracts
prepared from normal or from deafferented animals.
strictly involved in the synthesis of LH-RF. Animals implanted with

cycloheximide either unilaterally or bilaterally in this region of
the brain have normal amounts of LH-RF in the ME. These results,
when considered in conjunction with the evidence previously described
indicating that LH-RF is synthesized in the suprachiasmatic region,
suggest that the effect of cycloheximide is very localized, and that
no significant diffusion of the drug toward the basal hypothalamus
takes place.
II. DATA SUGGESTING THE EXISTENCE OF A FOLLICLE STIMULATING HOR-

MONE-SYNTHESIZING FACTOR
The three sets of results presented in the preceding sections

of this paper agree in indicating that, in male rats, FSH-RF is
synthesized in the area of the paraventricular nuclei. It is inter-
esting that, both following paraventricular lesions and after hypo-
thalamic deafferentation, the concentration of FSH in the anterior
pituitary of the operated animals was perfectly normal, although
192 l. MARTINI
o
LH-RF"
()Jg of LH NORMAL
depletion/pit)
30 ~ DEAF"F"ERENTED
•
20
•
•
•
10
8 days 15 days
Fig. 4. Effect of hypothalamic deafferentation on the LH-RF con-
tent of the hypothalamic island of adult rats (eight and
fifteen days after the operation). Columns represent the
depletion of pituitary LH induced in normal male rats by
the intracarotid injection of hypothalamic extracts pre-
pared from normal or from deafferented animals.
the synthesis of FSH-RF had been considerably reduced. This result

is surprising, since it is generally believed that RF control not
only the release of anterior pituitary hormones, but also their
synthesis (10). In accordance with this view, one whould expect
the synthesis of a pituitary hormone to be interrupted when the
corresponding RF is no longer available.
The presence of normal amounts of FSH in the anterior pituitary

of animals in which the synthesis of FSH-RF had been experimentally
suppressed, might be tentatively explained by suggesting that the
hypothalamus controls FSH secretion through the production of two
independent humoral factors: one devoted to the control of FSH re-
lease (FSH-RF), the other to the control of FSH synthesis ( FSH -
synthesizing factor or FSH-SF). It was believed that such a
FSH-RF
(~g of FSH
depletiorVpit)
300 CASTRATED rf RATS
250
"c
200
..
1!!.
Q.
E
'0
c
150 ·i
•
"c
..
100 1!!.
Q.
50 ..
.§
"
OIl
CONTROL SHAM MONOLATERAL BILATERAL

IMPLANTS IMPLANTS
Fig. 5. Effect of implants of cycloheximide (Actidione) in the
paraventricular region on the FSH-RF content of the medi-
an eminence of adult castrated male rats. See text for
more details. Columns represent the depletion of pitui-
tary FSH induced in normal male rats by the intracarotid
injection of hypothalamic extracts prepared from controls
or from implanted animals.
Control: unimplanted animals
Sham; animals implanted with empty cannulae
Monolateral implants: animals implanted with cycloheximide
in one paraventricular nucleus
Bilateral implants: animals implanted with cycloheximide
in both paraventricular nuclei.
hypothesis might be tested by evaluating whether hypothalamic ex-

tracts prepared from animals submitted to a complete hypothalamic
deafferentation (and consequently devoid of any FSH-RF activity)
might induce the synthesis of FSH in animals in which this had been
interrupted by appropriate experimental manipulations. Corbin and
Daniels (personal communication) have recently shown that ME lesions,
performed in adult male rats, result, in one to two weeks, in a
significant decrease in the amounts of FSH stored in the anterior
pituitary. The decrease of FSH in the pituitary of lesioned ani-
mals has been interpreted as being due to the fact that hypophysio-
tropic principles endowed with synthesizing capacities cannot be
194 L. MARTINI
LH-RF
(,ugofLH
depletion/pit)
30 CASTRATED if RATS
~
1H.
r:
2S !!
~
.1H.
Q.
r:
E .!!
Q.
'0
20 ., .,
r: o§
o~ ~
iii
'" °
1S
• • •
10
CONTROL SHAM MONOLATERAL BILATERAL

IMPLANTS IMPLANTS
Fig. 6. Effect of implants of cycloheximide (Actidione) in the
paraventricular region on the LH-RF content of the median
eminence of adult castrated male rats. See text for more
details. Columns represent the depletion of pituitary
LH induced in normal male rats by the intracarotid injec-
tion of hypothalamic extracts prepared from controls or
from implanted animals.
Control: uni~planted animals
Sham: animals implanted with empty cannulae
Monolateral implants: animals implanted with cycloheximide
in one paraventricular nucleus.
Bilateral implants: animals implanted with cycloheximide
in both paraventricular nuclei.
transported to the anterior pituitary. Daniels and Corbin (per-

sonal communication) have also shown that the intravenous injection
of crude hypothalamic extracts prepared from the brain of normal
rats induces a rapid and significant increase of FSH stores in ME-
lesioned recipient animals. They have concluded that, under the
conditions of their experiments, crude hypothalamic extracts are
able to facilitate the resynthesis of FSH.
A similar approach has been recently used by Daniels, Motta

and Martini (unpublished observations). These authors have con-
firmed that electrolytic lesions placed in the ME region of adult
male rats are followed in 10 days QY a significant decrease in pi-
tuitary FSH concentration (Table 1). As in the experiments of
Daniels and Corbin, the intravenous injection of a hypothalamic
extract prepared from the brain of normal animals induced, in 45
minutes, a significant increase in the amounts of FSH stored in the
pituitary of ME-lesioned rats. In addition, Daniels, Motta and
Martini have shown that the concentration of FSH in the pituitary
of lesioned animals is also brought back to normal by the intra-
venous injection of a hypothalamic extract prepared from the brain
of deafferented animals (Table 1) in spite of the fact that this
hypothalamic extract does not contain any FSH-RF.
The results indicate that both normal hypothalamic extracts

and hypothalamic extracts prepared from animals submitted to com-
plete hypothalamic deafferentation contain a factor which is able
to increase the concentration of FSH in the pituitary of ME-lesioned
animals. There a~e two mechanisms through which this factor may
operate: a) to facilitate the synthesis of new FSH; or b) to in-
hibit the release of stored FSH. The first possibility is the one
favored by the authors, since it is not believed that, in ME-le-
sioned animals, release of FSH occurs so rapidly as to permit a
significant reaccumulation of the hormone 45 minutes after block-
ade of the release process. In addition, no factors able to sup-
press FSH release have been described so far; inhibiting factors
have been recognized to exist for the control of the secretion of
prolactin (11), melanocyte stimulating hormone (12) and growth
hormone (13).
The results obtained using the hypothalamic extract prepared

from the deafferent animals suggest that the hypothetical FSH-SF
is different and independent from FSH-RF; moreover, its site of
origin is within the deafferented island and not in the paraven-
tricular region.
III. MEDIATORS INFLUENCING THE NUCLEI WHICH SYNTHESIZE THE LUTEIN-

IZING HORMONE-RELEASING FACTOR
It is a generally accepted concept that extrahypothalamic ner-

vous structures (e.g., amygdala, cortex, hippocampus, midbrain) may
influence the neurons which secrete the gonadotropin-releasing fac-
tors (1,6,14). The validity of this concept is proved once more
by the observation, reported in a previous section of this paper,
that LH-RF is accumulated in the deafferented hypothalamic island
when all inputs from extrahypothalamic structures are eliminated.
...
~
Table 1. Effect of Median Eminence (ME)Lesions and of Intravenous Injections of Median Eminence
Extracts (MEE) on Pituitary Concentrations of FSH in Adult Male Rats
Intravenous Pituitary FSH b

Groups a P
Treatment (~g/ mg)
Normal (20) Saline 42.'54 + 3-.82 c
ME-lesioned (25) Saline 32.27 ± 3.38 < 0.05

Vs Normal
ME-lesioned (18) 2 MEE from 43.19 ± 3.10 < 0.05

Normal Rat Vs ME-lesioned,
saline treated
ME-lesioned (18) 2 MEE from 45.68 + 5.21 < 0.05

Deafferented Rat Vs. ME-lesioned,
saline treated
a. Number of rats in parentheses; b. Microgram equivalents of NIH-FSH-S-5 per milligram wet

weight of pituitary tissue; c. Values are means + SEe
r
~
~
~
The chemical nature of the humoral mediator(s) involved in trans-

ferring the information from extrahypothalamic centers to the hypo-
thalamic nuclei has not been fully clarified so far. A lot of data
seem to suggest that traditional central nervous system mediators
(e.g., epinephrine, norepinephrine, serotonin, dopamine, acetyl-
choline, etc.) may be involved in such a process (15,16,17,18).
The possibility that acetylcholine(Ach) might be the mediator in-

volved in transmitting extrahypothalamic influences to the nuclei
synthesizing LH-RF, has been tested recently using an in vitro pro-
cedure by Fiorindo and Martini (unpublished observations). They
have shown that the anterior pituitary of normal adult male rats,
when incubated alone, secretes very low amounts of LH. The release
of LH from the incubated pituitaries is increased, but not signi-
ficantly, by the addition of fragments of male hypothalamus; such
fragments have been prepared in a way to include the ME and the
arcuate-ventromedial region, i.e. one of the areas in which LH-RF
is synthesized. Ach, added to incubation flasks containing only
anterior pituitary tissue, does not increase LH output. The cho-
linergic mediator does increase LH secretion in flasks containing
hypothalamic fragments together with anterior pituitary tissue;
under these circumstances, the increase in LH release is proportion-
al to the dose of Ach added to the medium. The data have been in-
terpreted as indicating that Ach stimulates the release of LH-RF
from the hypothalamic fragments; the LH-RF released under the in-
fluence of the drug enhances the secretion of LH from the incuba-
ted pituitary.
This interpretation is supported by the results of additional

experiments performed by Fiorindo and Martini (unpublished observa-
tions) using drugs which either counteract or prolong Ach action.
Atropine, a drug inhibiting the peripheral effects of Ach, has been
shown to counteract in vitro the LH-releasing effect of Ach added
to the medium containing anterior pituitary tissue and hypothalamic
fragments. It is interesting that atropine also inhibits the small
release of LH which is observed when hypothalamic tissue is added to
anterior pituitaries in the absence of Ach; this result suggests
that atropine may be able to block the effects on LH-RF release of
endogenous Ach present in the incubated hypothalamic tissue. Pro-
stigmine,a drug which inhibits cholinesterases, enhances the LH-
releasing effect of Ach, when this mediator is added to media in
which hypothalamic fragments are incubated together with anterior
pituitary tissue. Prostigmine also increases the release of LH from
anterior pituitaries incubated with hypothalamic fragments, but
without exogenous Ach. This result may be explained by the fact
that prostigmine enhances the activity of the endogenous Ach present
in the hypothalamic tissue incubated.
In a similar set of experiments, Schneider and McCann (19)

have found that the adrenergic mediator dopamine is able to release
198 L. MARTINI
LH in vitro when added to incubation media containing hypothalamic

fragments and anterior pituitary tissue. They have suggested that
dopamine might be the mediator liberating LH-RF. The possibility
then exists that both cholinergic and adrenergic mechanisms parti-
cipate in the control of the release of LH-RF from the neurons in
which this factor is synthesized. The relative role of the two
types of mediators remains to be ascertained.
IV. FEEDBACK CONTROL OF GONADOTROPIN SECRETION IN THE MALE
There are three different levels of feedback control of the

activity of the hypothalamo-pituitary-gonadal axis of the male; a)
a "long" feedback system, in which the inhibiting messages are pro-
vided by the hormonal steroids (mainly, but not exclusively, andro-
gens) produced by the testes and possibly by the adrenals; b) a
"short" feedback system, in which the inhibitory signals are repre-
sented by the gonadotropic hormones synthesized in the anterior pi-
tuitary gland; and c) an "ultrashort" feedback system, in which the
releasing factors directly control their rate of production (20,
21).
Ao "Long" Feedback System
The classic findings about the "long" feedback control of go-

nadotropin secretion in the male may be found in two reviews by
Davidson (22,23) and in two articles by Martini and co-workers (5,
24). Consequently, only one aspect of the problem will be dealt
with in detail in this paper: the type of androgenic hormone ac-
tive on feedback receptors.
Recent evidence suggests that the potent androgenic steroid

17-betahydroxy-5-alpha-androstan-3-one (androstanolone, dihydro-
testosterone, DHT) may be the active form of testosterone on andro-
gen-sensitive target structures (see 25 for references). Conver-
sion of testosterone into DHT has been reported to occur in the
prostate , the seminal vesicles, the preputial glands and in other
accessory organs of male reproduction (see 25 for references). If
the formation of DHT is a general feature of testosterone action,
one might expect this derivative to be formed also in the sites in
which androgens exert their feedback effect on gonadotropin secre-
tion. This hypothesis has been recently tested by Kniewald and
Martini (unpublished observations). They have incubated in vitro
labelled testosterone with slices of hypothalamus, pituitary gland,
amygdala, cerebral cortex and prostates and identified the meta-
bolites formed. All tissues were taken from adult male rats. The
results indicate that testosterone is converted into DHT by all
tissues examined; the prostates were the structures which effected
such a conversion to the greatest extent; they were followed by the
pituitary gland and by the hypothalamus; the cerebral cortex and
the amygdala were also able to transform testosterone into DHT, but
the rate of conversion was not as great as that found in the other
tissues- Androstenedione (a metabolite of testosterone which is
formed also in other androgen-sensitive target tissues) (26- 28 )
was found together with DHT in all incubation media; traces of an-
tlrostanediol and of androstenedione were also formed by some of
the tissues.
These data, as well as the similar ones recently reported by

Jaffe (29), suggest that the transformation of testosterone into
DHT is probably necessary also for initiating androgen-induced
feedback responses. They also indicate that the hypothalamus is
not the only target organ of the feedback effect of testosterone,
as it is generally assumed: (22,23); apparently, the anterior pi-
tuitary and other nervous structures (e.g., the amygdala, the ce-
rebral cortex) must also be incorporated into androgen feedback
circuits. That cerebral receptors for androgens are not exclusive-
ly localized in the hypothalamus is supported also by the study of
the distribution of labelled androgens in vivo (27,28,30).
B. "Short" Feedback System
In an attempt to assess whether FSH might modify its own rate

of secretion through a "short" feedback effect, Fraschini and co-
workers (31,32) have studied the effects of the administration of
exogenous FSH on the pituitary stores of FSH and on the hypotha-
lamic content of FSH-RF in castrated adult male rats. Orchidecto-
mized animals were used in order to avoid the possibility that the
injected FSH might activate the gonads and operate through a "long"
feedback effect. FSH was measured according to the assay of Steel-
man and Pohley (33) and FSH-RF using the "pituitary depletion meth-
od" described by Fraschini et ale (2). The data sunnnarized in Fig.
7 indicate that treatment with exogenous FSH results in a signifi-
cant drop of pituitary stores of FSH and of the hypothalamic con-
tent of FSH-RF. The data shown in the figure also suggest that the
inhibitory signal provided by FSH is specific; it has been impos-
sible to duplicate its effects by injecting ACTH and LH into the
castrated animals.
The demonstration that an artificial increase in FSH in the

general circulation reduced hypothalamic FSH-RF content suggested
that the opposite phenomenon could be studied also, i.e., what
happens to the secretion of FSH-RF when endogenous FSH is elimi-
nated. Different groups of adult male rats were subjected to cas-
tration, hypophysectomy, or castration followed by hypophysectomy.
Their hypothalamic stores of FSH-RF were then evaluated utilizing
the in vivo procedure of Fraschini et ale (2). Three weeks follow-
ing castration a significant increase in the amounts of FSH-RF
stored in the hypothalamus was observed (Fig.B). One week follow-
ing hypophysectomy a slight, but insignificant, increase of FSH-RF
was detected. It is uncertain whether this small increase was due
200 l. MARTINI
o
F'SH
( Jjg/pit)
•
F'SHRF'
(J.lg of F'SH
depletion/pit)
800 CASTRATED cf RATS 200
•
600 150
400 100
200 50
0--010-- o
CONTROL F'SH LH+ACTH
Fig. 7. Effect of systemic administration of FSH on pituitary FSH

content and on hypothalamic FSH-RF stores of castrated
male rats.
to the elimination of the "short" signal or to the reduction of

testosterone secretion brought about by hypophysectomy, but the lat-
ter appears a likely explanation since all testosterone-dependent
structures (prostates and seminal vesicles) were atrophied in this
group of rats. In the animals subjected to the removal of the pi-
tuitary and the testes, hypothalamic FSH-RF stores reached a level
well above that observed following the elimination of either gland
alone (20,21). These data indicate that the effects of castration
and of hypophysectomy are additive. It is also clear that in cas-
trated animals FSH-RF stores do not reach the high levels found
following castration plus hypophysectomy because of the "short" neg-
ative feedback signals provided by FSH. The fact that the elimina-
tion of the "short" feedback signal increases release as well as
synthesis of FSH-RF is documented by the observation that, after
hypophysectomy, FSH-RF appears in the peripheral plasma (34,35).
Additional information and data on the "short" feedback mech-

anism controlling LH secretion may be found in the chapter by Motta
et al. (21).
rSHRr
()l9 of rSH
depletion/pit)
200
150
100
50
0 - ' - -....
NORMAL ex HYPOX eX+HYPOX
204.0*10.3 8.5 *0.3 70.0*3.4 10.8*0.4 Slrnin~1 ftsitllS ( mg/l00g b.w.)

70.0*2.7 2.8*02 25.8*0.9 *
3.4 0.3 Proslates (mg 1100 9 b.w.)
Fig. 8. Effect of castration (Cx), hypophysectomy (Hypox) and

castration plus hypophysectomy (Cx + Hypox) on hypothalam-
ic FSH-RF stores and on seminal vesicle and prostate
weight of male rats.
C. "Ultra-short" Feedback System
It is clear from the data presented in previous sections of

this paper that the concentration of FSH-RF in the hypothalamus can
be modified by changes in the level of the hormones produced either
in the peripheral target glands or in the anterior pituitary. Motta
et ale (20,21) have recently explored the possibility that the syn-
thesis, storage, and release of FSH-RF might also be influenced by
changes in titer of this same principle in the general circulation.
Two groups of male rats that had been castrated for three weeks and
hypophysectomized for one week were used. One of these groups was
injected subcutaneously with an extract of rat hypothalamus rich in
FSH-RF (1 ME/rat per day, for five days); the other was treated in
a similar way with saline solution. Animals were killed three hours
after the last injection. Normal controls were also studied. The
content of FSH-RF in the hypothalami of the different groups of an-
imals was measured by the "pituitary depletion method" described by
Fraschini et ale (2).
202 L. MARTINI
rSH- Rr
(,ug of rSH
depletlon/ pit)
,CASTRATED-HYPOPHYSECTOMIZED - ,
cf RATS
200
•
150
100
• •
50
o
NORMAL SALINE MEE
Fig. 9. Effect of treatment with a median eminence extract (MEE)
rich in FSH-RF on hypothalamic FSH-RF stores in castrated-
hypophysectomized male rats.
Fig. 9 shows that the treatment with the hypothalamic extract

brings back to normal the elevated hypothalamic stores of FSH-RF
typical of castrated hypophysectomized rats. This result is not
due to contamination of the hypothalamic extract with FSH; the ex-
tract was assayed by the Steelman and Pohley procedure (33), and
this hormone was not detected. It is difficult, of course, to es-
tablish whether this reduction of stores is due to inhibition of
the synthesis of FSH-RF or to stimulation of its release. However,
it appears that brain elements may exist which are sensitive to
changing levels of releasing factors.
ACKNOWLEDGEMENT
The experimental work performed in the author's laboratory

and here described has been supported by funds of the Department of
Pharmacology of the University of Milan and by the following grants:
67-530 of the Ford Foundation, New York; 5-R01 AM 11783-01-02 of the
National Institutes of Health, Bethesda, Marylando Gifts of FSH

and LH were made by the National Institutes of Health, Bethesda,
Maryland. All such support is gratefully acknowledged.
REFERENCES
1. Szentagothai, J., Flerko, B., Mess,B., and Halasz, B. Hypotha-

lamic Control of the Anterior Pituitary, Akademiai Kiado,
Budapest, 1968.
2. Fraschini, F., Motta, M., and Martini, L., In Methods in Drug
Evaluation eds. Mantegazza, P. and Piccinini, F., North Holland
Publishing Company, Amsterdam, p 424, 1966.
3. Motta, M., Piva, F., Fraschini, F., and Martini, L., In Chem-
istry of Hypophysiotropic Hormones of the HypothalamuS:ed.
Meites, J., In press.
4. Mess, B., Fraschini, F., Motta, M., and Martini, L.,In Hormonal
Steroids,eds. Martini, L., Fraschini, F. and Motta, M., Excerpta
Medica, Amsterdam, p. 1004, 1967.
5. Martini, L., Fraschini, F., and Motta, M., Recent Progr. Hormone
Res.,24: 439, 1968.
6. Halasz, B., In Frontiers in Neuroendocrinology, 1969,eds.
Ganong, W.F., and Martini, L., Oxford University Press, New
York, p. 307, 1969.
7. Tima, L., Motta, M., and Martini, L., Program of the 51st Meet-
ing of the Endocrine Society, p. 194, 1969.
8. Motta, M., Piva, F., Tima, L., Zanisi, M., and Martini, L,
Mem. Soc. Endocrinology,1§: 407, 1970.
9. McCann, S.M., and Dhariwal, A.P.S., In Neuroendocrinology, eds.
Martini, L., and Ganong, W.F., Vol. I, Academic Press, New York,
p. 261, 1966.
10. Geschwind, 1.1. In Frontiers in Neuroendocrinology, 1969, eds.
York,p 389, 1969.
11. Meites, J., In Neuroendocrinology,eds. Martini, L., and Ganong,
W.F., Vol. I,~cademic Press, New York, p. 669, 1969.
12. Kastin, A.J., and Schally, A.V., Gen. Compo Endocr., ~: 344,
1967.
13. Krulich, L., Dhariwal, A.P.S., and McCann, S.M. ,Endocrinology,
83: 783, 1968.
14. Mess, Bo and Martini, L., In Recent Advances in Endocrinology,
ed. James, V.H.T., J. and A. Churchill, London, p. 1, 1968.
15. Kobayashi, H., and Matsui, T., In Frontiers in Neuroendocrino-
ology, 1969,eds. Ganong, W.F., and Martini, L., Oxford Univer-
sity Press, New York, p. 3, 1969.
16. ruxe, K., and Hokfelt, T., In Frontiers in Neuroendocrinology,
1969,eds. Ganong, W.F., andlMartini, L., Oxford University
Press, New York, p. 47, 1969.
17. Ganong, W.F., and Lorenzen, L. In Neuroendocrinology, eds.
Martini, L., and Ganong, W.F., Vol. II, Academic Press, New
York, p. 583, 1967.
204 L. MARTINI
18. Piva, F., Sterescu, N., Zanisi, M., and Martini, L., Bull
WId. Hlth. Org., 41: 275, 1969.
19. Schneider, H.P.G.-,-and McCann, S.M., Endocrinology, 85: 121,
1969.
20 0 Motta, M., In Progress in Endocrinology,ed., Gual, C.,
Excerpta Medica, Amsterdam, p. 523, 1969.
21. Motta, M., Fraschini, F., and Martini, L., In Frontiers in
Neuroendocrinology, 1969, eds. Ganong, W.F., and Martini, L.,
Oxford University Press, New York, p. 211, 1969.
22. Davidson, J.M o , In Neuroendocrinology, eds. Martini, L., and
Ganong, W.F., Vol: I, Academic Press, New York, p. 565, 1966.
23 0 Davidson, J.M., ~ Frontiers in Neuroendocrinology, 1969, eds.
York, p. 343, 1969.
24. Martini, L., Fraschini, F., and Motta, M., In Endocrinology
and Human Behaviour, ed. Michael, R.P., Oxford University
Press, London~ p. 175, 1968.
25. Gloyna, R.E., and Wilson, J.D., J. Clin. Endocr., 29: 970,
1969. --
26. Harding, B.W., and Samuels, L.T., Endocrinology, 70: 109, 1962.
27. Resko, J.A., Goy, R.W., and Phoenix, C.H., Endocrinology, 80:
490, 1967.
28. Whalen, R.E., Luttge, W.G., and Green, R. Endocrinology, 84:
217, 1969.
29. Jaffe, R.B., Steroids, 14: 483, 1969.
30. Pfaff, D.W., Experientia, 24: 958, 1968.
31. Fraschini, F., Motta, M., and Martini, L., Experientia, 24:
270, 1968.
320 Fraschini, F., In Progress in Endocrinology, ed. Gual, C.,
Excerpta Medica:-Amsterdam, p. 637, 1969.
33. Steelman, S.L., and Pohley, F.M., Endocrinology, 53: 604, 1953.
34. Saito, T., Sawano, S., Arimura, A., and Schally, A.V., Endo-
crinology, ~: 1226, 1967.
35. Negro-Vilar, A., Dickerman, E., and Meites, J., Endocrinology,
82: 939, 1968.
DISCUSSION
LUNENFELD: Dr. Martini, you mentioned that progest~rone exerts a

positive feedback mechanism; could you elaborate on this assumption?
Also, I seem to be at a loss about the present state of releasing
factors. Do we have a releasing factor for FSH and a releasing fac-
tor for LH or not?
MARTINI: The "positive" feedback effect of progesterone is on LH

and not on FSH. As for your second question, I still believe that
there are two separate hypothalamic factors for the control of gon-
adotropic secretion: one controlling FSH, the other LH. I think
that our data, which indicate that following a total hypothalamic
deafferentation FSH-RF disappears while LH-RF is still present,
present good proof for the existence of two separate releasing fac-
tors.
GESCHWIND: It is a little difficult to reach a decision unless one

personally works with both of these releasing factors. I mentioned
one experiment from Dr. Schally's group in which they observed ap-
proximately a 100% increase in the total amount of FSH in the sys-
tem upon culturing the pituitary gland. It so happened that the re-
leasing factor used for this experiment was LRF and pot FRF. They
used one of Dr. Schally's purest preparations. W. White at Abbot,
who has collaborated with Dr. Schally in this work, is also of the
opinion that they are both one and the same substance. There is one
very important experiment which was done about a year ago by Paul
Nakane, who obtained from Dr. Midgley antibodies to FSH which did
not cross react with LH, and antibodies to LH which did not cross
react with FSH, and hooked peroxidase on to both of these. Dr.
Nakane then attempted to localize both of these hormones in the an-
terior pituitary of the rat, and found that both labelled antibodies
localized in the same cells. It is a little difficult to understand
how you can get separate control of FSH granules and LH granules
within the same cell with two different releasing factors.
JUTISZ: We did not purify LRF and FRF to the sat;.'. t::xtent as Dr.
Schally did, but as we purified FRF the LRF activity steadily di-
minished. We never eliminated this activity completely, but we ob-
tained an FRF preparation with a low LRF activity and an LRF prep-
aration with very low FRF activity.
LIPSETT: Dr. Martini, I have never had any difficulty accepting

your data on the short feedback and I am quite willing to accept
your data on the ultra short feedback, but as a physiologist I have
real difficulty in comprehending how such a system would possibly
work.
MARTINI: It is our impression that the hypothalamus is able to re-

alize and evaluate how much of each releasing factor is stored in
the median eminence. We have indicated that unilateral implants
with a blocker of protein synthesis, such as cycloheximide, per-
formed in the paraventricular region of the hypothalamus, inhibit
FSH-RF synthesis only on one side and make FSH-RF disappear from
the ipsilateral side of the median eminence. I do not know whether
you noticed that the non-implanted side contained twice the amount
of FSH-RF which is found in normal animals. We think that this may
be due to the fact that the paraventricular nuclei of the non-im-
planted side have got the indication that half of the median emin-
ence was devoid of any FSH-RF.
E. STEINBERGER: Dr. Martini, in the experiments where the hypothal-

~us was separated, you showed a decrease in FSH-RF with normal lev-
els of FSH in the pituitary and interpreted this finding by bringing
206 L. MARTINI
in the possibility of a synthesizing factor. If FSH-RF was dimin-

ished, and the FSH release was impaired,could these findings be due
to simple storage of FSH in the pituitary? I think that simulta-
neous determinations of pituitary and plasma levels of FSH and LH
could have shed some light on this question. In experiments with
castrated animals, you made an assumption that when an animal is
castrated you have removed the feedback related to testosterone.
Other experiments suggest different explanations for these findings.
MARTINI: I agree with you on the second question; actually our

data indicate that also estrogen and progesterone may participate
in the feedback control of gonadotropin secretion in the male rat.
With regard to your first question, it is still difficult to pro-
vide an answer mainly because of the lack of a good FSH-radioimmuno-
assay to be used in the rat. It is however interesting to note that
the weight of the testes of the animals whose hypothalamus had been
completely deafferented for two weeks was perfectly normal. Appar-
ently, even in the deafferented animal, there is enough leakage of
FSH from the pituitary to maintain the testicular weight, if tes-
ticular weight is an indication of circulating FSH.
PURIFICATION AND CHEMISTRY OF GONADOTROPIN RELEASING FACTORS
Marian Jutisz
Laboratoire de Physiologie Cellulaire, College de France,
11, Place Marcelin Berthelot, Paris Ve, France.
INTRODUCTION
Historical Data
Although the humoral mechanism for the control of secretion of

pituitary gonadotropins has been proposed for many years (1,2), the
first attempts to demonstrate the presence of an LH (luteinizing
hormone)-releasing factor (LRF) in hypothalamic extracts are rel-
atively recent. It was actually Harris and Campbell et ale (3,4)
and McCann et ale (5,6) who in 1960 independently demonstrated the
presence of LRF in the acid extracts of cattle and rat hypothalamus.
The existence in hypothalamic extracts of an LRF capable of stimu-
lating release of LH has now been established in the following nine
species: rat (5-8), cattle (4,9-11), sheep (8,11-14), pig (14),
rabbit and dog (10,15), monkey (10), guinea pig (16,17) and man (18).
Although LRF from some of these species has been prepared in a high-
ly purified state, nobody, to our knowledge, has obtained pure LRF.
The presence of an FSH (follicle stimulating hormone) releas-

ing factor (FRF) in rat hypothalamic extract was first demonstrated
by Igarashi and McCann (19) in vivo and by Mittler and Meites (20)
in vitro. Independently, Kuroshima et ale (21) demonstrated the
presence of FRF in hypothalamic extracts of cattle and sheep. The
existence of an FRF in hypothalamic extracts of pig (22) and man
(18) have also been demonstrated. Highly purified preparations of
FRF have been obtained only from some of these species.
A well documented review on the purification and chemistry of

hypothalamic neurohormones regulating anterior pituitary function
has recently been published by Schally et ale (23).
207
208 M. JUTISZ
Nomenclature
The most widely used term for designating the hypothalamic

principles is the term "releasing factors" (RF) proposed by Saffran
et al. (24). Thus the principle acting on the secretion of lutein-
izing hormone is called LH-RF or LRF, that acting on the secretion
of .follicle-stimulating hormone, FSH-RF or FRF. Schally et a1. (23)
recently suggested designating the releasing factors by the term
"releasing hormones", i.e., LRF by the abbreviation LH-RH or LRH and
FRF by FSH-RH or FRH. In order to avoid confusion, we thought it
preferable to keep the old nomenclature which is understood and
accepted by a large number of scientific workers.
BIOLOGICAL TESTS
It is very important for purification work to have available

a sensitive, specific and relatively simple biological test or assay
for rele~sing factors.
LRF
1. In vivo. To assay LRF, Campbell et al. (4,10) and Nikito-

vitch-Winer (9) have used the test of ovulation obtained in the fe-
male rabbit or rat by direct intrapituitary infusion of an acidic
extract of median eminence of cattle or rat. This same extract,
when injected intravenously, was unable to produce ovulation. The
cycle of the female rat, an animal with spontaneous ovulation, is
blocked at the time of proestrus by the administration of pentobar-
bital (9). To assay LRF, the ovulation test has also been used in
androgenized female rats (13,25) and in rats rendered anovulatory by
hypothalamic lesion (26).
The most frequently employed test for routine assay of LRF was
the ovarian ascorbic acid depletion (OAAD) test of Parlow (27). In
this test (5,28), a stalk-median eminence extract (SME) is infused
intravenously directly into immature female rats pretreated with
pregnant mare serum-gonadotropin (PMS) and human chorionic gonado-
tropin (HCG) and the ovarian ascorbic acid depletion is taken as
the response. To allow for the possible pr~sence of LH-like acti~
vity in LRF preparations, the animals were hypophysectomized, trea-
ted with prolactin and used for assay one hr after hypophysectomy
(29). It should be pointed out, however, that this assay has a low
sensitivity for LRF (30), and that sufficient doses of vasopressin
can deplete ovarian ascorbic acid (31).
A much more sensitive and specific test is the determination

of the elevation of plasma LH levels in chronically ovariectomized,
estrogen and progesterone-treated rats (32). The extracts are in-
fused i.v., and the rats bled 10 minutes later. The plasma LH le-
vels were determined by the OAAD-assay (27).
2. In vitro. Pituitaries from chronically ovariectomized, es-

trogen and progesterone-treated rats (32) were also used for demon-
strating the action of LRF in vitro (11). The pituitaries were bi-
sected and incubated for one or two hrs in a Krebs-Ringer bicarbon-
ate-glucose medium. At the end of the incubation, the LH was as-
sayed in the incubation medium using the OAAD test. It has been
shown that a linear log-dose response curve was obtained (33).
FRF
1. In vivo. FRF or a crude stalk median eminence (SME) raise

plasma FSH activity in chronically ovariectomized rats pretreated
with estrogen and progesterone (19,29). FRF or a crude SME also
deplete the pituitary FSH content in normal male rats (34,35),
castrated male rats pretreated with testosterone (36), and in ovar-
iectomized female rats pretreated with estrogen and progesterone
(37). The Steelman-Pohley test (38) was used for the assay of FSH
in the pituitary tissue.
2. In vitro. Mittler and Meites showed that crude rat SME

stimulates FSH release from normal male rat pituitaries cultured
for 9 days (20) or incubated in tissue culture medium for 6 hours
(39). A linear log-dose response curve was obtained. Using incu-
bation of pituitaries from chronically ovariectomized, estrogen and
progesterone-treated rats in a Krebs-Ringer bicarbonate-glucose me-
dium, a significant release of FSH under the influence of FRF or
crude SME was also observed after a one hr incubation (21) or after
a 2 hr incubation (40). Pituitaries from normal male rats may also
be employed in the short time incubation method (41) and give a
linear log-dose response curve.
PURIFICATION STUDIES
Purification of LRF
1. Ovine LRF. The first attempt to partially purify LRF,

starting from ovine hypothalamus, was made in 1963 by Guillemin et
ale (12). The method proposed consisted of the following steps:
preparation of acetone-ether dried powder from frozen sheep hypo-
thalamic tissue; extraction with 2 N acetic acid at 4 oC; gel fil-
tration on a Sephadex G-25 column in 0.1 M pyridine acetate buffer
pH 5.0; and chromatography on a carboxymethylcellulose (CMC) column
equilibrated with 0.01 M ammonium acetate buffer, pH 4.5. The LRF
activity was eluted with a gradient to pH 7, O.lM buffer. This
method has since been used by several authors as a general proce-
dure for the purification of all hypothalamic hormones. It was
further slightly modified in our laboratory (42) and adapted for
large batches of tissue. In order to completely eliminate inor-
ganic salts, one step was added to the original method; the phenol
extraction (43). An example of the new version of this method is
210 M. JUTISZ
given below: frozen fragments of sheep hypothalamic tissue (Sorga,

Paris) are lyophilized and the region of median eminence containing
the pituitary stalk (SME) is cut off from the dry fragments by means
of a stainless steel cutter. Tissues are ground in acetone with an
Ultra-Turrax 1 homogenizer, defatted with acetone (50 ml, then 25
ml/g) and peroxide-free ether and extracted twice with 2N acetic
acid at 40 C (5 ml/g for each extraction).
A lyophilized extract from 20,000 hypothalamic fragments (84

g) is extracted twice with 600 and 300 ml of cold 0.1 M pyridine
acetate buffer, pH 5.0, and applied to a Sephadex G-25 column (19
x 105 em) maintained in a cold room at 40 C. Fractions of 240 ml
are collected every 20 min. Fig. 1 shows the chromatographic pat-
tern and biological activities observed during gel filtration •
•- . U V 280mp
~ _ -0 GOA 330 mp
i\
....... ARG VP
, _., MSH z
o
Ca ..'------------_~ I-
1.000 K+--------------~~ 100 U
I. \ Na+'-------------~~ «
IZ:
FRF LRF u..
~
..I
I X " o
z
I \ I' I-
~
>
u ',..'" • j~\ "f\
R Z
',,' ".':!.. ('\
~QSOO 50 ::J 500 II)
, \ "f ' ""'. ...... 1 ~\t".1 \ •
" '\ ,\ 1\ IZ: I-
:,J.\.' , . . ", \ o
ID • y. . .' Z
IZ:
o \:'~ II)
II)
::J
II)
I' \'\ /!'~./'V' .; '.'~ " " • W
ID
« ': ,./,. o' .<. • i.·.· ·0. 00. ~ ......... ~,._ IZ:
A.
I \ . . _ ..Jo " ~"~:-:'Y. Y-y' Y,,'
I \a:'" - Y
·Ie . •
•. '1(. . . . .
-.y _.y
__ ~ -0-..0
50 100 150
FRACTION NUMBER (240 ml)
Fig. 1. Gel filtration of sheep hypothalamic extract (53 g of ma-

terial corresponding to 20,000 SME) on a column of Sephadex G-25
(19 x 105 cm) in 0.1 M pyridine acetate buffer, pH 5.0. Fraction
size, 240 ml; 300 fl.l aliquots taken for biuret colorimetry (Goa,
44). From Berault (42).
The concentration of different products in the fractions is

determined by the measurement of absorbancy at 280 mfl. and biuret
colorimetry at 330 mfl. (44). The first large peak (Fig. 1) corre-
sponds to substances excluded from Sephadex G-25 and contains ac-
tivities of the type of LH, TSH (thyroid stimulating hormone) and
ACTH. LRF emerges in fractions 73 to 84. The FRF, melanocyte
stimulating hormone (MSH) and vasopressin activities are also deter-
1
Janke & Kunkel Kg., Staufen i. Br. (Germany).
mined. There is some overlappin~of FRF and LRF activities. The

concentrations of Na+, K+ and Ca in the fractions are measured
using flame photometry. Starting from the fractions 55 to 60,
these ions are present in all fractions. The LRF activity is de-
termined using either the OAAD test or the incubation method as de-
scribed above. The lyophilized LRF fraction (3.9 g) is desalted by
phenol extraction. The fraction is dissolved in 80 ml water satu-
rated with H2S and this solution is extracted twice wi'th 40 and 20
ml phenol saturated with water and H2S, The phenol extracts con-
taining LRF are washed 4 or 5 times with water and 80 ml of water
together with 500 ml of ether saturated with H2S are added. After
equilibration the ether-phenol phase is washed five times with wa-
ter, the aqueous phases are combined and extracted several times
with ether to remove traces of phenol. The aqueous phases are
heated under vacuum to remove ether and lyophilized. All the LRF
activity is concentrated in this material (0.38 g) which is dis-
solved in 30 ml of 0.01 M ammonium acetate buffer, pH 4.5 and chro-
matographed on a CMC column (2.4 x 40 cm) equilibrated with the same
buffer. After 22 fractions of 10 ml had been collected, a gradient
to pH 7, 0.1 M buffer through a 1000 ml mixing flask is set up. Fig.
2 shows the results of chromatography on CMC •
• - . UV 280mp
0 __M Folin & Lowry 750 mp
, ....., Conductivity
E
1D
~
Do
:I
5000 "-
rn
o
%:
:E
~M "":
C
u ~ ... u
~
~ . c
E
".t." c
E ...
>
>
o c ~: .'
cC
>
t
:l:~
zQ.5 2500 ...
c FRF LRF o
ID ~r-i----: :;)
II:
o Ir. t. \ ."1\ Q
rn 'I' ~, • ·*i" lL.. Z
o
ID
c ~V \'
' .... ,.., 0'
__
"J'IJ '\7",
I.
,......
...........:.:.-•• - --...
.,1( ....
If--::..
o
x-x·" .-.-.~~~ . ............
50 100 150
FRACTION NUMBER (lOml)
Fig. 2. Chromatography of FRF and LRF concentrates from Sephadex

column and phenol extraction on CMC column (2.4 x 40 cm) equili-
brated with 0.01 M ammonium acetate buffer, pH 4.5. Gradient to
pH 7.0, 0.1 M buffer through 1000 ml mixing flask started at tube
22. Gradient to pH 7.0, 1.0 M buffer started at tube 125. Frac-
tion size, 10 ml; 200 ~l aliquots taken for Folin-Lowry colorime-
try. From Berault (42).
212 M. JUTISZ
The concentration of different products in the fractions is

determined by the measurement of absorbancy at 280 m~ and Folin-
Lowry colorimetry at 750 m~. The conductivity of the fractions is
also determined. LRF emerges in fractions 102 to 137 (Fig. 2), well
separated from vasopressin (fractions 138 to 163). Some overlapping
is observed between fractions containing LRF and FRF. After lyophi-
lization of the fraction containing LRF, the yield is 15.5 mg of a
dry product which is active in vivo at a dose of 3 to 6 ~g on the
elevation of plasma LH levels in ovariectomized estrogen and pro-
gesterone-treated rats. In vitro, a dose of 0.08 to 0.15 ~g of the
same product per mg or pituitary tissue is able to release a signi-
ficant amount of LH from pituitaries of the same type of rats (see
assay for LRF above).
A very similar method (but without phenol extraction) was used

by Dhariwal et ale (45) for the purification of ovine LRF. Nikito-
vitch-Winer et ale (46) dialyzed a boiled acid extract of SME be-
fore fractionating the dialysate on a CMC column under conditions
similar to those described above. Fawcett et ale (47) accomplished
a partial purification of LRF by a procedure in five steps: Two cy-
cles of gel filtration of a crude extract on Sephadex G-25 column;
chromatography on DEAE-Sephadex A-25; gel filtration on Sephadex
G-25 in ethanol - 0.2 N-acetic acid (1:5, v/v) medium; thin-layer
chromatography on microcrystalline cellulose in a butanol~acetic
acid-water solvent. Gel filtration in a partially organic medium
allowed the recovery of two zones of activity separated by an in-
active fraction. To prepare partially purified ovine LRF, Johnson
(25) extracted the tissue with an acetate buffer at pH 4.6 and pre-
cipitated the activity with 4 x vol of acetone. Precipitation of
the active material with acetone was repeated twice more. The final
product was assayed using the OAAD-test and also the weaver finch
assay of Witschi (48).
2. Bovine LRF. Schally et ale (49,50) purified bovine LRF

using a similar method to that described above for ovine LRF: an
extract of SME in 2 N acetic acid was heated to boiling point,
chilled, and lyophilized. After reextraction with glaCial acetic
acid, the lyophilized extract was submitted to Sephadex gel filtra-
tion, phenol desalting, and chromatography on CMC. This material
was then rechromatographed once more on an analytical column of CMC.
A partial purification of bovine LRF was also obtained by Ramirez
et ale (51) by extensive heat treatment of a crude acidic extract
followed by gel filtration on a Sephadex G-25 column.
3. Porcine LRF. LRF with the highest specific activity and

free from other contaminant hormones was prepared from pig SME ex-
tracts by Schally et ale (14,23); the hypothalamic fragments were
extracted in batches of 25,000 to 35,000 as described above for bo-
v~ne LRF. The concentrates made from these extracts were passed
through a preparative column of Sephadex G-25 equilibrated with 1M
acetic acid. The results of this chromatography are shown in Fig. 3.
o.yloCin
- Peptide cone
..
i
68
••· .. MSH ,,i 60
.
•• - 0 PrnSOf
(;
......•
.B-MSH a-MSH
~
f·' :
.-. - VoIocIeprnwr
:::::::::;: LRH
~"~ TRH
- FSH·RH
,i 52 ~ 26
22
~ ~
~
0;::. + ~;«~ GRH
.:.":.:.. ,', CRH 18 ~
.... "=
a0
~
~
~ ,
.§ ~ 28 , 14 ~
~ ~
~ 0
~
::>:. 20 1
i
10 :§
0 ! ~
.i 12
I 6 ct
0
4 i 2
100 120 140 160 180 200 220 240 260 280
Fraction Number
Fig. 3. Gel filtration of pig SME concentrates (10 g of material

corresponding to about 3000 pig SME) on a column of Sephadex G-25
(7 x 170 cm) in 1.0 M acetic acid. Fraction sizes 25 ml; 50 1J.1
aliquots taken for Folin-Lowry analysis. (From Schally et ale
Recent Progress in Hormone Research, 24: 505, 1968.)
LRF activity emerged in fractions that also contained cortico-

tropin releasing factor (CRF), thyrotropin releasing factor (TRF)
and FRF activities. Zones containing these activities were combined
and extracted together with phenol. Phenol desalted material was
then submitted to CMC chromatography. This chromatography separated
TRF and most of the vasopressin from CRF, LRF and FRF which emerged
in the same zone together with a~SH. Separation of CRF from a zone
containing LRF activity and FRF activity was accomplished by a sec-
ond chromatography on an analytical column of CMC. A complete se-
paration of LRF from FRF was effected by preparative column electro-
phoresis illustrated in Fig. 4.
As shown in Fig. 4, FRF activity moved strongly towards the

cathode and was well separated from the L~F zone. LRF thus prepared
was active in vivo at a dose of 5 ng on the elevation of plasma LH
level in ovariectomized, estrogen and progesterone-treated rats.
In vitro, a dose of 20 ng/ pituitary was able to release a signifi-
cant amount of LH (see LRF - 2 above). It was free of other known
hypothalamic RF and of vasopressin.
4. LRF from other sources. Barry et ale (17) reported the

partial purification of an LRF activity present in the lateral hypo-
thalamus of guinea-pig. These authors used gel filtration on Sepha-
214 M. JUTlSZ
Q
N
"'
0. 4
',<>/.::::;;;»z LRH
::::::;:;:::::::;:: F 5 H· RH
0. 3
o. 2
0. 1
10 20 30 50 60 70 90 100 110 120
TUBE NO. (0. 25 MLfTUBEI

Fig. 4. Column electrophoresis of FRF (FSH-RH), LRF (LRH) concen-
trate from rechromatography on CMC. Elution pattern from vertical
LKB apparatus No. 3340-c fitted with column 0.9 x 38 cm; supporting
medium Sephadex G-25; fraction size, 0.25 ml/tube . A210 readings
taken on fractions after removal of buffer by lyophilization.
Starting material: 3.2 mg of solids. Conditions: 1000 V for 2.5
hr; cathode at bottom. Buffer was pH 6.5 pyridine acetate. (From
Schally et at. Recent Progress in Hormone Research, 1i: 543, 1968).
dex G-50 and a micro scale preparative electrophoresis (52) at pH

3.9. In this latter procedure, they isolated a small amount of a
material that was active on the depletion of ovarian cholesterol in
immature rats pretreated with PMS and HCG (53).
Starting from rat hypothalamic tissue Kobayashi et al. (54)

partially purified LRF and separated it from FRF using gel filtra-
tion on a Sephadex G-25 column. LRF activity was assayed in an in
vitro system in which rat anterior pituitaries were cultured by the
organ culture method (55). The LH activity was estimated in the
culture medium using the OAAD-test (27) and the FSH activity using
the Steelman-Pohley method (38).
Purification of FRF
1. Ovine FRF. Partial purification of FRF has been accom-

plished in our laboratory (40,42) using the same method as that de-
scribed above for ovine LRF. The following steps were followed ;
Extraction of SME with 2 N acetic acid; reextraction of lyophilized
extract with glacial acetic acid; gel filtration on a Sephadex G-25
column; phenol extraction; chromatography on a CMC column. As shown
in Fig. 1, the FRF activity emerges from Sephadex just before the
LRF activity in fractions 59 to 71, but the two activities overlap
to some extent. FRF is assayed using the in vitro method of incu-
bation of pituitaries of either ovariectomized rats pretreated with
estrogen and progesterone, or normal male rats (see FRF - 2 above).

During the chromatography on CMC, FRF emerges in fractions 82 to
102 and overlaps slightly with LRF (Fig. 2). Starting with 20,000
SME, the yield is 9.2 mg of FRF. This product is active in vitro
at a dose of 5 ng per mg of pituitary tissue (41).
Dharival et al. (56) were the first to obtain a partial sepa-

ration of ovine FRF from LRF on a Sephadex G-25 column. In another
paper, Dhariwal et al. (57) reported on the chromatographic beha-
vior of ovine FRF on CMC.
2. Bovine FRF. The same method was used for the purification
of bovine FRF as for bovine LRF (50). FRF activity was estimated
by measuring depletion of FSH from pituitaries of normal or testos-
terone-treated male rats, as well as elevation of plasma FSH acti-
vity in ovariectomized rats, pretreated with estrogen and proges-
terone (see FRF - 1 above). FRF activity was found to emerge from
Sephadex before LRF and vasopressin and was found to be essentially
free of LRF. The FRF zone from Sephadex was further purified by
phenol extraction and chromatography on CMC. This step separated
FRF from TRF. The FRF thus purified was able to deplete pituitary
FSH content in vivo at doses of 10 to 47 ~g.
3. Porcine FRF. Schally et al. (22,23) prepared highly puri-

fied porcine FRF essentially by the same steps as for porcine LRF.
As shown in Fig. 3, FRF activity emerged from Sephadex in the same
fractions as LRF, TRF and CRF, in contrast to bovine material.
After extraction of the combined fractions with phenol, chromato-
graphy and rechromatography on a CMC column left FRF and LRF in the
same fraction, completely separated from TRF and CRF. As shown in
Fig. 4, a complete separation of FRF from LRF was achieved by pre-
parative column electrophoresis. As in the case of bovine material
FRF activity was estimated by the depletion of pituitary FSH con-
tent in rats. The highly purified fraction was found to be active
at doses of the order of 10 ng. This product was free of LRF and
other known hypothalamic RF.
4. FRF from other sources. Kobayashi et al. (54) prepared a

partially purified rat FRF by gel filtration, as described above
for rat LRF.
CHEMICAL PROPERTIES AND STABILITY
While this review was being written (December 1969), the exact
chemical nature of LRF and FRF was still unknown. However, our
knowledge in this field is progressing very rapidly and it is pro-
bable that in some months' time we should have more information on
this subject. The generally accepted view that LRF and FRF are
simple polypeptides has recently been questioned (58,59). In our
opinion the question however still remains open. The story of the
nature of TRF is a good example which shows how dangerous these
216 M. JUTISZ
kind of presumptions may be.
LRF
McCann (28) was the first to show that LRF activity was de-
stroyed by incubation with trypsin, slightly reduced by pepsin, and
that it was thermostable. In an early note from our laboratory
(60), we reported that pepsin destroys and that trypsin greatly di-
minishes the LRF activity of ovine origin. The destruction of LRF
activity (ovine) by proteolytic treatment with pepsin was also re-
ported by Fawcett et al. (47). The stability of LRF to heat was
generally accepted with the exception of Johnson (25) who found
that heating to 100 0 C for 1 min inactivated LRF. Schally and
Bowers (49) were the first to report an amino acid composition for
their bovine LRF and stated that the purified material showed only
one major ninhydrin-positive component on paper chromatography or
electrophoresis. Similar results concerning ovine LRF were reported
by Guillemin (61). The absence of cystine in bovine LRF has been
demonstrated by Ramirez and McCann (62). All these results and also
the position of LRF in the chromatographic pattern led several au-
thors to the conclusion that LRF activity was associated with a
basic polypeptide having a molecular weight of 1200 to 2000 (45,47,
49,60,61,63,64). Nikitovitch-Winer et al. (46) reported that their
purified ovine LRF "is probably a small polypeptide which is dialyz-
able, heat stable, has little absorbancy at either 260 or 280 m ~,
and does not give a positive reaction with ninhydrin after paper
chromatography."
The polypeptidic nature of LRF has recently been questioned

(23,58,59). Schally et al. (23)reported that pepsin does not in-
activate highly purified LRF, and trypsin does so only at high en-
zyme/substrate ratios. They further reported that chymotrypsin
inactivates LRF even at an enzyme/substrate ratio of 1/30. Over-
night incubation at 38 0 C at pH 8 did not affect LRF activity, but
a similar incubation at pH 2 slightly decreased it. They concluded
that an apparent effect of chymotrypsin and trypsin can perhaps be
accounted for by contamination with other enzymes or the ability of
these enzymes to hydrolyze other bonds in addition to peptide link-
age (ester, hydrazide, hydroxyamide, amide). This fact and also
the observation that acid hydrolysis of highly purified porcine and
bovine LRF revealed only trace amounts of amino acids (14,23) led
Schally and Kastin (59) to the conclusion that LRF consisted of a
non peptide, hydrosoluble molecule with a molecular weight of less
than 1500 and with a basic isoelectric point.
FRF
Very little information is available on the nature of FRF.

Several authors (21,56,57) reported that FRF is a small polypeptide
with a molecular weight of 1200 to 2000. Further, Dhariwal et al.
(57) indicated that their purified FRF was a rela"tively small basic
polypeptide (but less basic than LRF), heat stable, and that it was
inactivated by tryptic digestion.
Recently White et ale (65) isolated five polyamines from por-

cine hypothalamus: histamine, putrescine, spermidine, spermine
and lysine. They stated that taken together these substances ac-
count for most of the FSH-depleting activity in vivo of the crude
extract. These and other results led Schally et ale (22,23) and
White et ale (65) to consider that FRF is probably a polyamine
derivative with a molecular weight of not more than 300. With de
la Llosa (41) we checked the action of these polyamines in our in
vitro system and we found that they were inactive on the release
of FSH at doses of 0.05 to 1.25\,Lg per mg of pituitary tissue. This
was confirmed recently by Mittler and Schally (23) who indicated
that putrescine, cadaverine, spermine, spermidine and agmatine at
doses of 2 to 10 I,Lg per mg of pituitary tissue had no effect on re-
lease of FSH in vitro. They concluded that FRF is not one of these
polyamines. Kamberi and McCann(66) reported that some biogenic
amines enhanced the release of FSH in vitro, but only at high, non-
physiological doses. Thus the in vivo response stated by White et
ale (65) may be indirect in nature and it seems evident from these
results that it is not yet possible to assign FRF activity to a
specific feature of the chemical structure.
SUMMARY
The most frequently employed tests for assay of LH-releasing

factor (LRF) are, in vivo: The ovarian ascorbic acid depletion test
of Parlow and the determination of the elevation of plasma LH lev-
els in chronically ovariectomized, estrogen and progesterone-trea-
ted rats (EBP rats); in vitro: The incubation of pituitaries of
EBP rats. FSH-releasing factor (FRF) is assayed in vivo by either
determination of the elevation of plasma FSH levels in EBP rats or
the determination of the depletion of the pituitary FSH content in
different types of rats; in vitro by the incubation of pituitaries
of normal male rats or EBP rats.
Highly purified preparations of LRF and FRF from three species

have been obtained: sheep, cattle and pig, but they still seem
heterogeneous. The most widely used method for purifing hypotha-
lamic hormones consists of the following steps: Extraction of hy-
pothalamic tissue with 2 N acetic acid; lyophilization and reextra-
tion with glacial acetic acid; lyophilization and gel filtration
on Sephadex G-25; phenol extraction (desalting) and chromatography
on CMC. The most highly purified preparations of LRF and FRF were
obtained from porcine material.
The exact chemical nature of LRF and FRF is still unknown.

The generally accepted view that these hormones are simple polypep-
tides has recently been questioned.
218 M. JUTISZ
ACKNOWLEDGEMENT
This study was supported in part by research funds (R.C.P. no.

59) from the French National Research Council (C.N.R.S.).
REFERENCES
1. Harris, G.W., Neural control of the pituitary gland, Arnold,

London, 1955.
2. Benoit, J. and Assenmacher, I., J. Physiol. (Paris), 47: 427,
1955.
3. Harris, G.W., in Villee, C.A., Control of Ovulation, Pergamon
Press, New York, p. 56, 1961.
4. Campbell, H.J., Feuer, C., Garcia, G. and Harris, G.W., J.
Physiol. (London), 157, 30p, 1961.
5. McCann, S.M., Taleisnik, S. and Friedman, H.M., Proc. Soc.
Exp. Bioi. Med., 104: 432, 1960.
6. McCann, S.M., and Taleisnik, S., Endocrinology, 68: 1071, 1961.
7. Piacsek, B.E. and Meites, J., Endocrinology, 79:~32, 1966.
8. Courrier, R., Guillemin, R., Jutisz, M., Saki~ E. and Aschheim,
P., C.R. Acad. Sci. (Paris), 253: 922, 1961.
9. Nikitovitch-Winer, M.B., Endocrinology, 70: 350, 1962.
10. Campbell, H.J., Feuer, G. and Harris, G.W., J. Physiol.
(London), 170: 474, 1964.
11. Schally, A.V. and Bowers, C.Y., Endocrinology, 75: 312, 1964.
12. Guillemin, R., Jutisz, M. and Sakiz, E., C.R. Acad. Sci. (Paris),
256: 504, 1963.
13. Johnson, D.C., Endocrinology, 21: 832, 1963.
14. Schally, A.V., Bowers, C.Y., White, W.F. and Cohen, A.I.,
Endocrinology,~: 77, 1967.
15. Endroczi, E. and Hilliard, J., Endocrinology, 12: 667, 1965.
16. Barry, J., Dautrevaux, M., Lefranc, G. and Fourlinnie, J.C.,
C.R. Acad. Sci. (Paris), 257: 4217, 1963.
17. Barry, J., Moschetto, Y. and Leonardelli, J., C.R. Acad. Sci.
(Paris), 265: serie D, 1141, 1967.
18. Schally, A.V., Mueller, E.E., Arimura, A., Bowers, C.Y., Saito,
T., Redding, T.W., Sawano, S. and Pizzolato, P., J. Clin.
Endocr. ,!:J..: 755, 1967.
19. Igarashi, M. and McCann, S.M., Endocrinology, 74: 446, 1964.
20. Mittler, J.C. and Meites, J., Proc. Soc. Exp. Bioi. Med., 117:
309, 1964.
21. Kuroshima, A. Ishida, Y., Bowers, C.Y. and Schally, A.V.,
Endocrinology, ~: 614, 1965.
22. Schally, A.V., Saito, T., Arimura, A., Sawano, S., Bowers, C.Y.,
White, W.F. and Cohen, A.I., Endocrinology, 81: 882, 1967.
23. Schally, A.V., Arimura, A., Bowers, C.Y., Kastin, A.J., Sawano,
S. and Redding, T.W., Recent Progr. Hormone Res., 24: 497, 1968.
24. Saffran, M., Schally, A.V. and Benfey, B.G., Endocrinology,
57: 439, 1955.
25. Johnson, D.C., Experientia, 20: 311, 1964.
26. Schiavi, R., Jutisz, M., Sakiz, E. and Guillemin, R., Proc.
Soc. Exp. Biol. Med., 114: 426, 1963.
27. Parlow, A.F., In Albert, A., Human Pituitary Gonadotropins,
Thomas, C.C., Springfield, p. 300, 1961.
28. McCann, S.M., Amer. J. Physiol., 202: 395, 1962.
29. Guillemin, R. and Sakiz, E., Endocrinology, ~: 813, 1963.
30. Novella, M.-A., Alloiteau, J.-J. and Aschheim, P., C.R. Acad.
Sci. (Paris), 259: 1553, 1964.
31. McCann, S.M. and Taleisnik, S., Amer. J. Physiol., 199: 847,
1960.
32. Ramirez, V.D. and McCann, S.M., Endocrinology, 73: 193, 1963.
33. Jutisz, M., Berault, A., Novella, M.-A and Ribot, G., Acta
Endocr. (Kobenhavn), 55: 481, 1967.
34. David, M.A., Fraschini, F. and Martini, L., Experientia, 21:
483, 1965.
35. Corbin, A. and Story, J.C., Experientia, ~: 694, 1966.
36. Saito, T., Arimura, A., Mueller, E.E., Bowers, C.Y. and
Schally, A.V., Endocrinology, 80: 313, 1967.
37. Kuroshima, A., Arimura, A., Saito, T., Ishida, Y., Bowers, C.
Y. and Schally, A.V., Endocrinology, 78: 1105, 1966.
38. Steelman, S.L. and Pohley, F.M., Endocrinology, 53: 604, 1953.
39. Mittler, J.C. and Meites, J., Endocrinology, 78: 500, 1966.
40. Jutisz, M. and de la Llosa, M.P., Endocrinology, ~: 1193,
1967.
41. Jutisz, M. and de la Llosa, M.P., Bull. Soc. Chim. Bioi., 50:
2521, 1968.
42. Berault, A., Thesis, University of Paris, 1969.
43. Schally, A.V., Bowers, C.Y. and Redding, T.W., Endocrinology,
78: 726, 1966.
44. Goa, J., Scand. J. Clin. Lab. Invest., 1: 218, 1953.
45. Dhariwal, A.P.S., Antunes-Rodriques, J. and McCann, S.M.,
Proc. Soc. Exp. Bioi. Med., 118: 999, 1965.
46. Nikitovitch-Winer, M.B., Pribble, A.H. and Winer, A.D., Amer.
J. Physiol., 206: 1286, 1965.
47. Fawcett, C.P., Reed, M., Charlton, H.M. and Harris, G.W., Bio-
chem. J., 106: 229, 1968.
48. Witschi, E~Mem. Soc. Endocr., 4: 149, 1955.
49. Schally, A.V. and Bowers, C.Y., Endocrinology, 21: 608, 1964.
50. Schally, A.V., Saito, T., Arimura, A., Muller, E.E., Bowers,
C.Y. and White, W.F., Endocrinology, 79: 1087, 1966.
51. Ramirez, V.D., Nallar, R. and McCann, S.M., Proc. Soc. Exp.
Bioi. Med., 115: 1072, 1964.
52. Hannig, K., Hoppe Seyle Z. Physiol. Chem., 338: 211, 1964.
53. Bell, E.T., Mukerji, A. dnd Loraine, J.A., J. Endocr., 28: 321,
1964.
54. Kobayashi, T., Kobayashi, T., Kigawa, T., Mizuno, M., Amenomori,
Y., Watanabe, T. and Ichikawa, H., Endocrinol. Jap_, 14: 101,
1967.
55. Kobayashi, T., Kobayashi, T., Kigawa, T., Mizuna, M., Amenomori,
Y. and Watanabe, T., Endocrinol. Jap., 13: 430, 1966.
220 M. JUTISZ
56. Dhariwal, A.P.S., Nallar, R., Batt, M. and McCann, S.M.,

57. Dhariwal, A.P.S., Watanabe, S., Antunes-Rodrigues, J. and
McCann, S.M., Neuroendocrinology, ~: 294, 1967.
58. Guillemin, R., Ann. Rev. Physiol., 29: 313, 1967.
59. Schally, A.V. and Kastin, A.J., Tri~gle (Journal Sandoz des
Sciences Medicales), ~: 19, 1969.
60. Jutisz, M., de la Llosa, P •• Sakiz, E., Yamazaki, E. and
Guillemin, R., C.R. Soc. Bioi. (Paris), 157: 235, 1963.
61. Guillemin, R., Proc. 23rd Intern. Congr. Physiol. Sci. Excerpta
Medica I.C.S., 87: 284,1965.
62. Ramirez, V.D. and McCann, S.M., Amer. J. Physiol., 207:441,
1964.
63. McCann, S.M., Antunes-Rodrigues, J. and Dhariwal, A.P.S., Proc.
23rd Intern. Congr. Physiol. Sci., Excerpta Medica I.C.S., 87:
292, 1965. -
64. Jutisz, M., Europ. Rev. Endocrin. Supple 2, part 1, Pergamon
Press, p. 47, 1966.
65. White, W.F., Cohen, A.I., Rippel, R.H., Story, J.C. and Schally,
A.V., Endocrinology, 82: 742, 1968.
66. Kamberi, I.A. and McCann, S.M., Endocrinology, 85: 815, 1969.
FURTHER STUDIES ON MECHANISM OF ACTION OF LUTEINIZING HORMONE RE-
LEASING FACTOR USING IN VIVO AND IN VITRO TECHNIQUES
Marian Jutisz, Bernard Kerdelhue and Annette Berault
Equipe de recherche du CNRS, Laboratoire de Physiologie

Cellulaire, College de France, 11 Place Marcelin,
Berthelot, Paris Ve, France.
Some attempts have recently been made to discover the mechan-

isms whereby hypothalamic luteinizing hormone releasing factor (LRF)
stimulates the release of luteinizing hormone (LH) from the anterior
pituitary and its subsequent biosynthesis (1-3).
We reinvestigated some of the reported results using a highly

purified preparation of LRF and a new procedure of incubation of
pituitary glands. The recent development in our laboratory of a
radioimmunoassay for rat LH (4,5) has provided a sensitive and pre-
cise technique for these studies. Some of our preliminary results,
obtained by in vivo and in vitro techniques, are reported in the
present paper.
The highly purified preparation of LRF was prepared by gel

filtration of a crude extract of sheep hypothalamus on Sephadex G-
25 followed by chromatography on carboxymethyl cellulose (6). Cyclic
AMP was supplied by Boehringer A.G., Mannheim, Germany.
Rats used for this work were Wistar females (CF strain, CNRS,
Gif-sur-Yvette, Yvelines, France) ovariectomized when 40 days old
(140 g). Thirty to 40 days later, and three days before sacrifice,
they were injected with 25 ~g of estradiol benzoate and 12 mg of
progesterone (EBP rats). This treatment with steroids was selected
after an investigation of the influence of different doses of ste-
riods on the plasma LH concentration in rats injected or not with
LRF.
For in vivo studies, groups of three to five animals per treat-
221
222 M. JUTISZ, B. KERDElHUE, AND A. BERAUlT
ment were anesthetized with ethylurethane (150 mg/lOO g body wt;

solution 300 mg/ml) and LRF solutions or normal saline injected in-
to the external jugular vein. 500 ~l of blood were withdrawn from
the same vein 5 to 8 min, 14 to 15 min, and 22 to 28 min after in-
jection. LH was assayed in serum using a radioimmunoassay previous-
ly described (4,5).
For in vitro experiments, hemipituitaries (4 to 5 per flask)

were rinsed twice with 2 ml each time of Krebs-Ringer bicarbonate-
glucose buffer (KRB) and "pre incubated" for a period of 30 min in
2 ml of the same medium saturated with 93% 02 /7% C02 to remove LH
liberated from non-viable cells. After changing the medium, incu-
bation proceeded for 4 hr in 2 ml of either normal or modified KRB.
A high [~] medium was prepared by increasing the concentration of
~ from 5.9 to 59 mM and removing an equivalent amount of Na+. In
one case, cyclic AMP and LRF were added at the beginning of the in-
cubation period in flasks containing hemipituitaries of each pair;
the other hemipituitaries served as control. After a 2 hr period
the media were replaced by fresh media of the same composition, and
incubation proceeded for a further 2 hr period. In other experi-
ments, incubation was started in normal or high [~] KRB, in the
presence or absence of cyclic AMP or theophylline. After 2 hr of
incubation, and without changing the media, LRF or cyclic AMP were
added to some flasks, media were gassed and incubation proceeded
for a further 2 hr period. Ten ~l of media were withdrawn at 30 to
60 min periods during the incubation, using Eppendorf pipettes, and
were stored frozen until assayed. Results are expressed in terms
of a highly purified rat LH laboratory reference preparation (prep-
aration S15B; mean potency, 0.95 x NIH-LH-Sl; see ref. 4). Following
incubation groups of hemipituitaries were weighed to 0.05 mg.
RESULTS
Alteration of the Plasma LH Concentration After Injection of LRF.
Two doses of the LRF preparation were injected, 6 and 12 ~g per

rat. Fig. 1 shows that the concentrations of plasma LH increase
rapidly during the first 5 to 8 min following injection to a value
2-3 times higher than that of the controls. Maximal levels of LH
are reached in 14 to 15 min; the LH concentrations then decrease.
Kinetics of the Response of In Vitro Pituitaries to LRF, Cyclic AMP

and High [K+]
As shown in Fig 2, the response of pituitaries in vitro to LRF

(4 ~g/ml), cyclic AMP (15 mM), and high [~] is very rapid but the
amount of LH released after 2 hr of incubation is hisher after stim-
ulation with an optimal dose of LRF and with high [KT] than in the
presence of an optimal dose of cyclic AMP. After a 2 hr incubation
100
75 5
~
J
II:
W
50
'" --LRF
CONTROLS
E - _0
\m _--
.,.,.. ,.,0_ - ___
....... 2
-03
II) 25
~ ..,.,""""
'"
')..",..
_ _ _ _ __ 0"
:t
...I
DI - - - - __ u _______ -,,1
c:
10 20 30
TIME (min)
Fig. 1. Alteration of the plasma LH concentration

in EBP rats after injection of LRF. At time 0,
animals are injected into the external jugular vein
with normal saline (1,2,3), 6 ~g LRF (4,5), or 12 ~g
LRF (6,7). Assays of LH were performed using a
radioimmunoassay method and results were expressed
in terms of a laboratory reference preparation of
rat LH (From: Kerdelhue et al. C. R. Acad. Sci.
Ser. D., 270: 1010, 1970).
and changing the media, the release of LH was readily reinitiated

under the same conditions as existed at the beginning of incubation,
but the levels of LH were lower than during the first 2 hr. period.
224 M. JUTISZ, B. KERDELHUE, AND A. BERAULT
0.8 - LRF
_cAMP
- K+59mM
w
:)
U)
U) 0.6
t-
al ~---.
E
\o o
-
0.4
w a
."....-------
U)
c ;'
W ./ o
..J /
W /0
~ 0.2 I
% I
---
_____ 1
..J
I
1 3 4
TI ME (hr. )
Fig. 2. Kinetics of the response to LRF, cyclic AMP

(cAMP), and high [K+] of pituitaries of EBP rats in
vitro. Four groups of pairs of 4-5 hemipituitaries
were incubated in 2 ml of either norma'! KRB (controls)
or in KRB containing LRF (4~g/ml), cyclic AMP (15 mM),
or 59 mM~. One half of the pituitary served as a con-
trol for the other half. After a 2 hr incubation,
media were replaced by fresh media of the same compo-
sition and incubation proceeded for a further 2 hr
period. Assays were performed using a radioimmuno-
assay method and results were expressed in terms of
a laboratory reference preparation of rat LH.
Potentiation of the Action of LRF by Theophylline
Two groups of pairs of hemipituitaries were incubated, one pair

in each group in normal KRB, and another in KRB containing 1 mM of
theophylline. As shown in Fig. 3, the addition of the same dose
(4~g/ml) of LRF to both groups provokes the release of LH, which
is significantly greater in the medium containing theophylline.
Under the same conditions, the addition of cyclic AMP (15 mM) in-
stead of LRF, failed to alter significantly the release of LH,
either in the presence, or in the absence of theophylline.
0.8 0 ___0 KAB + LAF
• • THEOPH. + LAF
q ••••••••• q KAB + cAMP
,
w .-. -. THEOPH. + cAMP
:;)
(I) 0.6
(I)
KAB or THEOPH. LAF or cAMP
....
Ol
E
I
'\.
o 0.4
w
(I)
c(
,..
./
W <00 ......- . . .- . - . -
..J ./.~ .
W
_ _.
II:
0.2
.""/
/ ~ ...... .
....... "U •••••• •••• .. ·0
...-. / ............
~
..",.... • ••"..,.-&••••••••
_.~~
.-. ~ ......··:.-o
•••••• a _ _
~.:..:. ..;:...g:.:.:.__ --0
10 30 60 90 120 150 180 210 240
TIME (min)
Fig. 3. Kinetics of the response to LRF and cyclic AMP
(cAMP) of pituitaries of EBP rats incubated in the pres-
ence of theophylline. Two groups of pairs of 4-5 hemi-
pituitaries were incubated, one pair in each group in
a normal KRB, and another in KRB containing 1 mM of
theophylline. After a 2 hr incubation, LRF (4 ~g/ml)
was added to one group of pairs and cyclic AMP (l5mM)
to the other group, and incubation proceeded for a
further 2 hr period. Assays were performed using a
radioimmunoassay method and results were expressed in
terms of a laboratory reference preparation of rat LH.
Enhancement of the Effects of "Optimal" High [K+] by LRF on the

Release of LH
One group of pairs of hemipituitaries was incubated separately

for 2 hr in a high [K+] medium, then LRF (4 ~g/ml) was added to one
flask only and incubation continued for a further 2 hr period.
Similarly another group of pairs of hemipituitaries was incubated
in KRB containing 15 mM of cyclic AMP; after 2 hr incubation, LRF
(4 ~g/ml) was added to one flask, and the same amount of cyclic AMP
as in the starting solution (15 mM) to the second flask. As shown
226 M. JUTISZ, B. KERDElHUE, AND A. BERAUlT
in Fig. 4, when high [K+] and LRF were both present, the effects were
greater than either alone produced. On the other hand, the results
obtained with the second group confirmed those shown in Fig. 2, i.e.,
the amount of LH released with an optimal dose of LRF was higher than
that released with an optimal dose of cyclic AMP. Fig. 4 shows also
that the release of LH starts later in the presence of cyclic AMP
than in the presence of LRF.
Q8 cAMP+LRF
. - - . cAMP+cAMP
59mM K+
.- .-.
.-'
/'
W ,/
:l
--.-.~.-.-.-
~ 0.6
LRF or / . . - / /
I-
cAMP. / /
,I
cAMP or 59mM K+
.I /
01
E I I
/
"
/
o 0.4
w .Ii I.······
/
/ .. ' ....
en
c( / •.. G.···'!···
w / . - •••••••G.. I
....I
w ..J-I.. ···· I
a: 0.2 ...• I
••• ; / 0
•••o~· I
:I:
.. ,;... I
....I
.,;.. -1
' ;;;".. ;.t.
.'H"c- j ..
___ --0--
:
_0
:
- -
10 30 60 90 120 150 180 210 240
TIME (min)
Fig. 4. Kinetics of the response to LRF and to a "double

dose" of cyclic AMP (cAMP) of pituitaries of EBP rats in-
cubated previously in the presence of cyclic AMP or in a
high [K+] medium. Two groups of pairs of 4-5 hemipitu-
itaries were incubated, one group in a normal KRB con-
taining cyclic AMP (15 mM), and another in KRB containing
59 mM K+. After 2 hr of incubation, in the first group,
LRF (4 ~g/ml) was added to one flask and cyclic AMP to the
second flask; in the second group, LRF (4 ~g/ml) was added
only to one flask. Incubation proceeded for a further 2
hr period. Assays were performed using a radioimmunoassay
method and results were expressed in terms of a laboratory
reference preparation of rat LH.
DISCUSSION
Studies of the kinetics of the response of rats to LRF injected
intravenously indicate that the response is rapid with maximal levels
of LH being reached in 14 to 15 min. Similar results with a highly

purified porcine LRF were obtained by Schally (maximal effect in 10
to 12 min, personal communication), and with a crude hypothalamic
extract by Gay et ale (7).
Studies of the kinetics of the response of rat pituitary tis-

sue to LRF, high [K+] medium and cyclic AMP indicate that the levels
of LH released in 59 mM K+medium and in a normal KRB medium con-
taining an optimal dose of LRF are very similar. The amount of LH
released in the presence of cyclic AMP is much lower. Similar ob-
servations have been made in the case of FSH release as stimulated
by FRF or by cyclic AMP (8). It is possible that penetration of
cyclic AMP into the receptor cells is slower than is the action of
LRF which may activate a cell membrane enzyme and therefore does
not need to penetrate inside the cell (9). Another possibility is
that ·cyclic AMP may be rapidly destroyed by phosphodiesterase (10).
That the action of cyclic AMP on the release of LH is slower than
the action of LRF is also shown in Fig. 4. Potentiation of the
action of LRF by theophylline (Fig. 3) is good presumptive evidence
of the participation of cyclic AMP. In the experiment shown in Fig.
3, the action of cyclic AMP was not potentiated by theophylline but
it should be noted that this potentiation is often difficult to
demonstrate as is the action of cyclic AMP itself. The reason for
this difficulty is unknown.
The effects of "optimal" high [K+] are enhanced by LRF (Fig. 4)

as has been already shown by Samli and Geschwind (2), and this may
mean that high [K+] acts on the release of LH in a different way
than does LRF. The release of neural hormones by a high [K+] was
attributed by Douglas (11) to the depolarization of the specific
cell membrane. Therefore, LRF may not act by the same mechanism.
REFERENCES
1. Samli, M.H. and Geschwind, 1.1., Endocrinology, ~: 835, 1967.

2. Samli, M.H. and Geschwind, 1.1., Endocrinology, 82: 225, 1968.
3. Jutisz, M., De La Llosa, M.P., Berault, A. and Kerdelhue, B.,
Proceedings of the Workshop Conference on the Integration of
Endocrine and Non endocrine Mechanisms in the Hypothalamus,
Eds. Martini, L. and Motta, M., Academic Press, New York, 1970
(in press).
4. Kerdelhue, B., Berault, A., Courte, C. and Jutisz, M., C.R.
Acad. Sci., Sere D, 269: 2413, 1969.
5. Kerdelhue, B., Berault, A., Ribot, G. and Jutisz, M., C.R.
Acad. Sci., Sere D, 270: 1010, 1970.
6. Jutisz, M, Berault, A., Novella, M.-A., and Ribot, G., Acta
7. Gay, V.L., Rebar, R.W. and Midgley, A.R. Jr., Proc. Soc. Exp.
Biol. Med., 130: 1344, 1969.
228 M. JUTISZ, B. KERDElHUE, AND A. BERAULT
8. Jutisz, M., and De La Llosa, M.P., C.R. Acad. Sci., Sere D,

268: 1636, 1969.
9. Sutherland, E.W., Oye, I. and Butcher, R.W., Recent Progr.
Hormone Res., 11: 623, 1965.
10. Robison, G.A., Butcher, R.W. and Sutherland, E.W., Ann. Rev.
Biochem., 37: 149, 1968.
11.. Douglas, W.W., Nature, (London) 197: 81, 1963.
CHAPTER IV

SECTION 2: TESTICULAR-PITUITARY INTERRELATIONSHIP
INVESTIGATIONS INTO THE FEED-BACK MECHANISM BETWEEN SPERMATOGENESIS
AND GONADOTROPIN LEVEL IN MAN
Svend G. Johnsen
Hormone Department, Statens Seruminstitut, Copenhagen,

and the Male Hypogonadism Study Section, Medical Out-
patients' Department, University Hospital of Copenhagen
It is a well established fact that there is a close relation-

ship between spermatogenesis and gonadotropins. Selective damage
of the germinal epithelium can easily be induced experimentally or
found clinically and it has been shown by many authors that sper-
matogenetic failure is associated with castration changes in the
hypophysis and increase in the gonadotropin level. The cause of
spermatogenetic damage is unimportant. Hypophyseal changes have
been shown to occur after anti-spermatogenetic compounds (1,2),
cadmium(3), radiation (4), testicular ischemia (5), cryptorchidism
(6, 7,8) and in various testicular disorders in man (9-14). When
there is a failure of spermatogenesis, the gonadotropin rise pri-
marily affects FSH (7,14,15), although LH also seems to be involved
(16) •
In the literature there have been two different theories on the

cause of this gonadotropin increase. Nelson (17) and Heller et al.
(11) believed that gonadotropins were used by the germinal epithe-
lium, and if the germinal epithelium ceased to function the unused
hormone was excreted. This explained the increased urinary gonado-
tropin excretion in cases of spermatogenetic failure. This "utili-
zation hypothesis" therefore implies that FSH production is running
at full speed whether or not gonads are present. Although still
maintained by Heller (cf.Moore et ale (2», the theory is incompat-
ible with the fact that castration causes an increase of gonadotro-
pins not only in the urine, but also in the hypophysis (18,19). It
is also incompatible with the occurrence of castration changes in
the hypophysis after spermatogenetic failure. Furthermore, no one
has ever shown "utilization" of a hormone in a target organ.
Discarding the "utilization" theory, the only other possibility
231
232 S. G. JOHNSEN
is that during spermatogenesis, a substance is produced which is

capable of inhibiting FSH release from the hypophysis. According to
this "inhibin" theory there is a true feed-back mechanism between
spermatogenesis and the hypophysis: FSH stimulates spermatogenesis
which in turn releases a substance depressing FSH.
Unfortunately, however, neither the site of production of the

FSH inhibitor nor its chemical nature have been disclosed (cf. dis-
cussion) and this means that fundamental knowledge of the regulation
of testicular function is still lacking.
This study is concerned with the identification of the stage of

spermatogenesis which is involved in the testicular-hypophyseal feed-
back mechanism in man. In 284 patients with a variety of primary
testicular disorders, the degree of spermatogenesis was determined
from testicular biopsies by a new method (testicular biopsy score
count) and correlated with gonadotropin analyses. The study involved
a number of analyses which are described in detail elsewhere (20).
This survey presents an outline of the general findings.
Urinary hypophyseal gonadotropins(HG) were determined by the

method of Johnsen (21,22). Results are expressed in mouse uterus
units (MUU) per 24 hours. The MUU were compared with an HMG standard
(1 ampoule of 2nd IRP-HMG = 133 MUU). At FSH/LH ratios found in
clinical samples, the MUU expresses FSH + LH (23). In an analysis
of a large series of normal men the log HG showed no deviation from
a normal distribution and log HG is used throughout the study.
The evaluation of degree of spermatogenesis was performed using

the testicular biopsy score count method (24). The full comprehen-
sion of the present paper requires familiarity with the principle of
this method and the reader is referred to the cited paper as only a
brief description follows. In this score count method, each tubular
section in one testicular biopsy section is given a score from 10 to
1 according to the following (abbreviated) criteria: Score 10 = full
spermatogenesis, normal epithelium, 9 = many spermatozoa, but dis-
organized epithelium, 8 = only few spermatozoa, 7 = no spermatozoa,
but many spermatids, 6 = no spermatozoa and only few spermatids, 5 =
no spermatozoa, no spermatids, but many spermatocytes, 4 = no sper-
matozoa, no spermatids and only few spermatocytes, 3 = spermatogonia
are the only germ cells present, 2 = no germ cells, but Sertoli cells,
1 = no cells in tubular section. "Spermatozoa"are here defined as
cells having achieved the small head form of the spermatozoon and
correspond to Clermont's spermatid b 2, c, d l and d 2 (25).
In the testicular degeneration the tubules lose their cell con-

tent in a fixed sequence so that the more mature forms usually dis-
appear before the next stage is seriously affected (cf. Jopqsen (24».
TESTICULAR-PITUITARY INTERRELATIONSHIP 233
PLOT ORDINATE LG.HG IIOU1
·..., . .
LG.HG 220221 A8SCISSA "5 M MEANS OVER· 9 VALUE ~
2,.1
2,4,
Z.I'
2.1' 2 2
2.21
Z.ZI 2'
. ..
· .. ...
2.11
2.11
Z.OI • 2
,.
Z.OJ
1.9.
··, .. .
1.9, 2 •
1."
1.1' • 2
1.7'
1.7)
1.'8
..
1 •• ,
.
1.51
• • .*
· ...
1."
1.... • 2 ••• , •• "
1.43
1.,.
1."
1.2'
*2 Z... ••• .Z
. ..... .
1.2' 2 •
1.11 2. • • • Z•••
1.U •• •• Z Z •
. ..., .
1.0.
1.0J
0.9.
0.9,
0,"
0 •• )
'2'
0.71 2 •
0,1,
0."
0.6,
o.n
0,1'
z.
.
Z •
0 •• ' 2.44 6,44 1.44 9.44
Fig. 1. Computer plot of the mean score of the testicular

biopsy score count (abscissa) versus the log value of the
excretion of hypophyseal gonadotropins in MUU per 24 hours
(log HG) (ordinate). * = 1 value. M = over 9 values. 284
cases. (From: Johnsen, S.G., Acta Endocr.(Kobenhavn) in
press).
The parameter in the present study is the cumulative percent (cum. %)

of tubules scored at each step formed successively from step 10 to
2 ( 10 + 9, 10 + 9 + 8 etc). Thus, step 4 cum. % o"f 25 means that
25% of all tubules contained few spermatocytes, and that 75% did not
show spermatocytes (these 75% might contain spermatogonia only,
Sertoli cells only, or no cells at all). Cum. % 0 means absence of
the cell recorded at the particular step and at all spermatogenetic
stages above it (step 8 cum. % 0 = total absence of spermatozoa,
step 6 cum. % = total absence of spermatids and higher stage, step
3 cum. % 0 = total absence of germ cells, etc.). A cum. % of 100
at a step means that all tubules contained at least the cell corre-
sponding to this step.
In order to calculate a mean score (MS) the number of tubules

recorded for each score is multiplied by that score and the sum of
all 10 multiplications is divided by the total number of tubules.
The mean number of tubules scored per biopsy was 113.
284 patients, all having primary testicular disorders, were used

in the study. 98 had various degrees of oligospermia, 9 Sertoli-
cell-only syndrome, 33 sequelae of testicular retention, 42 Kline-
234 S. G. JOHNSEN
felter's syndrome, 2 simple gynecomastia, 7 occlusion of the sperm

tract, 73 varicocele ( these were collected by Dr. P. Agger for a
special study to be published elsewhere), 8 miscellaneous disorders
and 12 were normal men. No evidence of impaired hypophyseal func-
tion was found in these groups.
RESULTS
Fig. 1 shows a plot of the mean score (MS) and log HG in the
284 patients. A very high correlation between the two was found.
The correlation coefficient is 0.73 (p <0.001) and the regression
is given by: log HG = 2.125 - 0.113 x MS. The close relationship
between spermatogenesis and gonadotropin output was thus confirmed.
It was thought that systematic plots and correlation between

cumulative precentages (cum. %) and log HG at all score count steps
in all 284 patients would be an effective tool in determining the
stage of spermatogenesis involved in the feed-back mechanism. The-
oretically, any stage could be involved and hypothetical curves for
the correlations were drawn placing the determinative step at steps
8, 6, 4 and 3 corresponding to spermatozoa, spermatids, spermato-
cytes and spermatogonia. It was found that the determinative step
would be the first step showing absence of low HG values at low
cum. % (for example: if the presence of spermatocytes would determine
the release of a gonadotropin inhibitor, then cases showing absence
of spermatocytes (that is, low step 4 cum. %, cf.above) would show
absence of low HG values. Cases with low HG values could, however,
be present at low cum. % in higher steps because spermatocytes may
be present in the absence of spermatozoa or spermatids. Therefore,
taking the steps from above, step 4 would in this case be the first
step showing the change). The determinative step would be the
first showing a correlation between cum. % and log HG. The step
above the determinative step would be the last step having low HG
values at low cum. %, and the step below the determinative would be
the first having some high HG values at high cum. % (high cum. % =
cell scored fully present). If more than one spermatogenetic stage
was involved, the HG at low cum. % should be expected to rise to a
certain level at one step and further at another.
On this background, the systematic analysis at once revealed

that the determinative step was step 8 (few or many spermatozoa
present). Figs. 2, 3, 4 and 5 show the plots between log HG and
cum. % from step 10 to step 7. Step 10 (Fig. 2) shows both high
and low HG values at low cum. % and only low HG values at medium
and higher cum. %. At step 9 (Fig. 3) there are less low HG values
at low cum. %. When step 9 cum. % approaches zero the testis can
at best contain tubules with few spermatozoa. Step 8 (Fig. 4) shows
a great change. Low HG values at low cum. % have disappeared.
When step 8 cum. % approaches zero spermatozoa are absent and it is
seen that low HG values are absent also. At high cum. % at step 8
the HG values are, with a single exception, low. Step 7 (many sper-
matids present) is similar to step 8 at low cum. %, but a number
of rather high HG values appear at high cum. %. Thus, whereas step
8 only had 1 log HG value> 1.64 at cum. 'Yo >80, step 7 has 8 values
and step 6, 10 values.
Table 1 shows the log HG values at low cum. % for all steps. It
is seen that the gonadotropin level rises significantly from step
10 through step 9 to step 8. However, after step 8 no further rise
is observed. When the cum. % is low at step 6 ~permatids disappear,
at step 4 spermatocytes disappear and at step 3 spermatogonia dis-
appear, but the disappearance of these cells from the testis is
associated with no further rise in gonadotropin level as seen when
spermatozoa disappear.
LDGHGUDUl PLOT ORDINATE L,OGHGUDZU ,1,85(15$'" TR 10 ItUM PtII M MEANS OYER , YAL.U£S
1
12
S 2
1
1
•
0
I·· ..
•
6 '2
0
z* ••
• .1 Z •• z
• z· • •• ••
... .
Ize • Z • •
.... , .
*2*2
z·· ... *2 •• Z.
o
•
•
... •.
la •••
• *2 ••
• Z·
-20.00 -1.00 ,.99 15.99 27.99 19.99 11,99 ".99 7'." '7.99 100.
Fig. 2. Computer plot of the percent of tubules

scored at step 10 of the testicular biopsy score
count (abscissa) versus log HG. (From: Johnsen,
S. G., Acta Endocr., (Kobenhavn) in press).
Correlation analyses between log HG and cum. % in the region

10-90 % showed no correlation at step 10 and step 9 (p >0.10) but
a high correlation at step 8 ( p < 0.001). This correlation was
236 S. G. JOHNSEN
LOGHGU0Z21 PLOT ORDINATE LOGHG220221 A85CI55A TR 9 I(.UM PR M MUNS uvER 9 VAL.UE,
·..
·..,. ..
• 2'
2"
•
10 •
· ..
2
6
•
•• 2 • .... '2
... .
2* • 2
2'
..
• •• 2 •
..
-20.00 -1.00 '.99 U.9' 21.,9 51.99 '7." 100.
Fig. 3. Computer plot of the percent of tubules scored

at step 10+9 (step 9 cumulative percent = 9 cum.%)
(abscissa) versus log HG. (From: Johnsen, S. G., Acta
Endocr., (Kobenhavn) in press).
LoGttGUOUI PLOT OIlOINAT! LOCIHGZlOZlI AI,e .,'" TR • IUM ,.. M MIA"S olllR 9 VA~UI'
I,.'
2.0\.
z.1t
,.,.
J.lt
,.11
2.24
.••• . .
1,14
1.0'
1.04
I."
I."
1."
1,1. ,,.
I •
....
h19
1.7.
I."
I
,. •
....
hi.
.
I
.....
1.1' '1
1.14
I.'"
. ... ,
1 .....
·...
lell
I."
1.29
I
1.2. I Z • J. • ••••
.,. . ..
1,19
1.14 I
.. .
1.09 I
0."
.. .. ..
hOlt
a•••
·'1
·
0.1' • I
0.1.
0.79
0,7.
• •
0."
0 •••
0."
0."
-10.00 ".00 n." , 61."
T... 10'." 120.
21 •• 9 ....
Fig. 4. Computer plot of step 8 cum. % versus log HG.
(From: Johnsen, S.G., Acta Endocr., (Kobenhavn) in press.)
LOGHGUOZU PLOT ORO INATE LOGHGUOZZl ",seuSA TR 7 !tUM PR M MEA",S OVER 9 yALU(S
2.49
Z.~.
'It
2.'9
,.
Z.
2.29
Z.ZIt
l.19
.., ..
Z.I.
2.09
2.04
1.99
1.9.
1.19
1.... , 2
..·· .......
1. '79 1 •
1.74
,
2 •
1.69 6 •
1•••
1.19
..
1.14 2 ••
1.49
1•••
1. " • "2
1. '4
1.29 2
1.a4 ••• 2 • •••• '2*2'
1.19 • Z 2 *2 • 2-
1.14 2"
1.09 2 •• a*,
1.e4 2' 2
... ... . ,
0.99
• • 2
0.9. "2
0 •• 9
0.'.
0.19
0.".
0 •• 9
0.'.
0."
o.s.
-20.00 -6.00 21.99 ".99 .9.99 61.99 91.99 101." uo.
Fig. 5. Computer plot of step 7 cum. % versus log HG.

(From: Johnsen, S.G., Acta Endocr., (Kobenhavn) in press).
P against
Mean: 95%
Step Number Min. Max. Mean s. d. preceding
fide limits
steE
10 180 0.70 2.54 b2l 0.44 1.57 - 1.70
9 123 0.78 2.54 1.82 0.35 1.76 - 1.88 < 0.01
8 92 1.36 2.54 b.2l 0.28 1.87 - 1.99 < 0.01
7 88 1.38 2.54 1.93 0.28 1.88 .. 1.99 n. s.
6 85 1.38 2.54 1.94 0.28 1.88 - 2.00 n. s.
5 81 1.38 2.54 1.95 0.28 1.89 - 2.01 n. s.
4 78 1.45 2.54 1.96 0.27 1.90 - 2.02 n. s.
3 70 1.45 2.54 1.97 0.28 1.91 - 2.04 n. s.
2 11 1.62 2.46 2.02 0.25 1.87 - 2.l7 n. s.
Table 1- Log values of excretion of hypophyseal gonadotropins when,

at the steps of the testicular biopsy score count, the cumulative
percent is 0 - 10 (= few or no cells, indicated by step number,
present). n. s.: not significant. (From: Johnsen, S. G. Acta
Endocrinologica, (Kobenhavn) in press).
238 S. G. JOHNSEN
p against
Type of Biopsy Elimination No. MEAN s. d.
first group
No germ cells
step 3 < 5
but Sertoli cells 22 0.24
step 2 >90
present
No spermatozoa < 5 <cO.10
step 8
but spermatids >10 3 2.06 0.27
step 6 >0.05
present
No spermatids
step 6 < 5 <0.20
but spermatocytes 3 1.66 0.18
step 4 >90 >0.15
present
No spermatocytes
step 4 < 5 <0.25
but spermatogonia 4 1. 70 0.13
step 3 >40 >0.20
present
No germ cells step 3 < 5
11 0 0 25 <0.01
No Sert~li cells step 2 <10
Table 2. Influence on log HG of spermatids in the absence of sper-

matozoa, spermatocytes in the absence of spermatids and higher stages,
spermatogonia in the absence of spermatocytes and higher stages and
Sertoli cells in the absence of germ cells. (From: Johnsen, S.G.,
Acta Endocrinologica, (Kobenhavn) in press).
maintained through all the following steps. All these showed a

continuous shift of low HG values to the right. The correlation
between log HG and cum. % took place at a still increasing HG level
until at step 2 very high HG values were only found at cum. % lower
than 100. However, depending on the number of tubules recorded at
each step, these findings are more or less complicated by a mutual
relationship between the cum. % at several steps and cannot be taken
as evidence of a separate influence upon HG levels of steps below 8.
Screening procedures revealed that three patients possessed

spermatids in the complete absence of spermatozoa, three had many
spermatocytes in the complete absence of spermatids (complete meiotic
arrest) and four had spermatogonia in the complete absence of sper -
matocytes. In Table 2 these 10 patients are compared with patients
with only Sertoli cells and, as seen, no difference is found. This
is a further indication that spermatids, spermatocytes and spermato-
gonia have no influence upon the HG level.
There was, however, a small but significant difference between

patients possessing only Sertoli cells and patients having very few
or no tubular cells (Table 2) and this indicated that Sertoli cells
may have some influence upon the gonadotropin level although their
depressing action is very small as compared with the effect of sper-
matozoa. It was here considered that the first group consists of
cases of Sertoli-cell-only syndrome and the second of cases of
Klinefelter's syndrome, and that the poor Leydig cell function des-
cribed in the latter might be the explanation through an extra in-
crease of LH. This possibility was, however, eliminated in a sep-
arate analysis where the group of 42 Klinefelter patients was com-
pared with a group of 44 patients with severe hypospermatogenesis
of non-chromosomal origin. In order to eliminate the influence of
the germinal epithelium, the log HG was in all cases transformed to
correspond to a mean score of 1 using the regression stated earlier
(log HG transformed = log HG + 0.113 (MS - 1)). The XXY group thus
had a mean value for log HG of 2.04 (s. d. 0.37) and the XY group
2.09 (s. d. 0.26). Thus, no separate extra-tubular influence upon
HG was found in Klinefelter's syndrome and the demonstrated effect
of Sertoli eells upon the HG level seems to be real.
The study has thus shown that the high correlation between sper-
matogenetic status and gonadotropin excretion is due to an effect
of the spermatozoa (latest spermatids) in the testis. The last stage
of spermatogenesis involving the maturation of the spermatozoa is
the only spermatogenetic stage involved in the testicular-hypophyseal
feed-back mechanism in man. Sertoli cells, either alone or accom-
panied by immature germ cells, seem to have a separate influence on
the gonadotropin level. This is very small and only demonstrable
when testes containing spermatozoa are excluded.
DISCUSSION
Possible Mechanism of the Inhibitory Action of the Spermatozoa
From a biological standpoint, the placement of the feed-back

mechanism at the end of spermatogenesis is rational. The question
remains how this end stage exerts the FSH inhibition.
Previously the author (12) put forward the hypothesis that the
cytoplasm split off from the spermatozoa, as the residual body,
which is rich in newly formed fat globules (26), contains a gonado-
tropin inhibitory substance. This theory places the feed-back at
the spermatogenetic stage which we have shown here to be the effec-
tive one, but neither we, nor others, have published evidence to
prove or disprove the theory.
Another possibility is that, by contact with the finished sper-

matozoon, a change is induced in the Sertoli cell so that this cell
produces an FSH inhibitor. Brokelmann(27) found in the rat that,
concurrently with spermatogenesis, structural changes indicating
steroid secretion (cf. below) are induced in the Sertoli cell.
Brokelmann showed that these changes occurred when the spermatozoa
240 S. G. JOHNSEN
dived into the epithelium immediately prior to their escape. It is

interesting that induction of Sertoli cell changes could be the
first biological explanation of the meaning of the mysterious round
trip journey of the nearly ripe spermatozoon down into the epithelium
to the Sertoli cell nucleus.
The two theories are combined by Lacy (28-30) who showed that,
in the rat, the residual bodies are phagocytosed by the Sertoli cells.
He claimed that the residual bodies convey both information and ma-
terial to the Sertoli cells which induce steroid production in them.
Lofts (31) believes that the lipids found basally in Sertoli cells
originate in the residual bodies. However, in our group, P. F.
Larsen (unpublished observations) has, using electron microscopy
studies in the human, been unable to find residual body remnants in
the Sertoli cells.
Although the local effect of the spermatozoon is still open for

speculation, the occurrence of changes in the Sertoli cell concur-
rently with spermatogenesis is well documented and suggests the pos-
sibility that the gonadotropin inhibitor is produced in this cell.
The Chemical Nature of the FSH Inhibitor
Androgens given in physiological doses suppress LH but not FSH

(32-36). Gestagens have no physiological gonadotropin suppressing
effect. It is, however, well documented that estrogens even at small
doses suppress FSH (34-37). In fact estrogens are the only type of
compound now known to be capable of suppressing FSH at a physiologi-
cal dose level. This means that either spermatogenesis liberates
estrogen, or an entirely new testicular hormone must exist. The
second possibility appears very unlikely, and the key to the under-
standing of the regulation of the testis is therefore to look for
estrogen production in the Sertoli cells.
Indications of Steroid Production in the Sertoli Cells
Brokelmann (27) found that the fine structure of the Sertoli

cell "fits well into the cytoplasmic morphology of steroid producing
cells." He found, concurrently with spermatogenesis, agranular
endoplasmic reticulum, mitochondria with tubular christae and lipid
droplets.
It has been found that Sertoli cells in higher animals do not

possess a steroid 3 t3 -ol-dehydrogenase (39-44) and thus presumably
cannot perform a de novo steroid biosynthesis. However, they are
known to possess certain steroid enzymes such as 3-, 16- and 17 -
hydroxylases (45-47).
Christensen and Mason (45) found that in the tubules the complete
biosynthesis from progesterone to testosterone can be carried out,
and the presence of testosterone in the Sertoli cells was demonstrated

immunologically by Woods and Domm (42).
Lacy et ale (48) claimed to have isolated estrone and estriol

from the lipids which accumulate in rat Sertoli cells in large amounts
when spermatogenesis is destroyed by irradiation, and according to
Lofts (31) estrogens have been found in Sertoli cells of Lower ani-
mals. Although further evidence is required, certain indications of
estrogen production in the Sertoli cells exist and Sertoli cells
definitely possess part of an androgen synthesis pathway.
It is a well-known fact that there is a considerable testicular

estrogen production, and is believed that this estrogen is derived
from the Leydig cells (49-52) because the estrogens increase when
the testis is stimulated with RCG. If, however, the Sertoli cells
can utilize a precursor from the Leydig cells for estrogen produc-
tion, the argument favoring the Leydig cell origin of the estrogens
is not valid. Another possibility is that Leydig cells can produce
one estrogen and Sertoli cells another. In fact, estrogens comprise
a whole family of compounds and the relative gonadotropin suppressive
effect of them is practically unknown, particularly in man.
It would appear that important information can be derived from

granulosa cells which are presumably homologous to the Sertoli cell
(53). Ryan and Short (54) and Ryan and Petro (55) showed that iso-
lated granulosa cells can convert progesterone or testosterone to
estrogen, but a steroid 3 ~-ol-dehydrogenase is lacking in the fol-
licle (39). Falck (56) demonstrated that granulosa cells cannot per-
form de novo estrogen biosynthesis unless theca interna cells are pre-
sent (presumably producing the necessary precursor for aromatization).
If we believe that the two gonads have very much in common, this sug-
gests that for estrogen production the granulosa cell homologue, the
Sertoli cell, may be equally dependent upon precursors from its theca
interna homologue, the Leydig cell.
In conclusion, it appears highly probably that the Sertoli cell

is capable of producing the FSR inhibitor.
Suggestion for the Complete Arrangement
An interposition of the Sertoli cell between the spermatozoa and

gonadotropin inhibition fits well with the findings in this study. A
possible interpretation of the results is that the Sertoli cell alone
(or the Sertoli cell with only immature spermatogenetic cells at -
tached) has a small, basal level of gonadotropin inhibitor production
which is enormously stimulated when maturation of spermatozoa takes
place at the cell.
On the basis of the present state of knowledge, the complete

arrangement for regulation of spermatogenesis could tentatively be
242 S. G. JOHNSEN
described as follows:
The stimulator of spermatogenesis is FSH which penetrates the

tubules (57). The FSH inhibitor is an estrogen produced by the
Sertoli cells from a precursor from the Leydig cells. A certain
Leydig cell stimulation with LH is therefore indirectly of impor-
tance. Left alone, the Sertoli cell has only a minimal production
of FSH inhibitor. However, either by direct contact with the sper-
matozoon during its diving into the epithelium down to the Sertoli
nucleus, or perhaps in some animals via information from the phago-
cytosed residual body, the secretory apparatus in the Sertoli cell
is changed, and significant amounts of FSH inhibiting estrogen are
released. The factor controlled by this arrangement will then be
the number of spermatozoa produced per unit time.
It is interesting that a system involving the Leydig cell, the

Sertoli cell, the spermatozoon (late spermatid) and perhaps the
residual body, could offer a logical explanation for the development
of the entire testicular structure.
ACKNOWLEDGEMENT
Supported by the Ford Foundation.
REFERENCES
1. Nissim, J.A., Lancet, 1: 304, 1957.

2. Moore, D.J., Roscoe, R.T., Matson, L.J. and Heller, C.G., Clin.
Res. 10: 88, 1962.
3. Allan~n, M. and Deanesly, R., J. Endocr., 24: 453, 1962.
4. Heller, G.G., Wootton, P., Rowley, M.J., Lalli, M.F. and Brusca,
D.R., Proc. Vlth Pan-Amero Congr. Endocrinol. p 408, 1965.
5. Hinman, F., Jr. and Smith. G.I., J. Urol. (Baltimore) 83: 706,
1960.
6. Tornblom, N., Upsala Lak.-Foren. Forh. 48: 1, 1943.
7. Taira, A.M. and Tarkham, A.A., Acta End~r. (Kobenhavn) 40: 175,
1962.
8. Morehead, J.R and Morgan, C.F., Fertil. Steril.,~: 530,
1967.
9. Klinefelter, H.F., Jr., Reifenstein, E.C., Jr. and Albright, F.,
J. Clin. Endocr., 2: 615, 1942.
10. McCullagh, E.P., Sirridge, W.T and McIntosh, H.W., J. Clin.
Endocr., 1Q: 1533, 1950.
11. Heller, C.G., Paulsen. C.A., Mortimore, G.E., Junck, E. C. and
Nelson, W.O., Ann. N. Y. Acad. Sci., .21: 685, 1952.
12. Johnsen, S.G.,Acta Endocr. (Kobenhavn) Supple 90: 99, 1964.
13. Johnsen, S.G.,Acta Endocr. (Kobenhavn) Supple lli: 17, 1967.
14. Paulsen, C.A., ~ Clinical Endocrinology, eds. Astwood, E. B. and
Cassidy, G.E., Vol 2, Grune and Stratton, Inc. New York p 569, 1968.
15. Paulsen, C.A., In Gonadotropins 1968, Rosemberg, E. ed. Geron-X,
Inc. Los Altos,-Calif. p 163, 1968 •
16. Wide, L. and Kjess1er, B., Acta Endocr. (Kobenhavn), 62: 299,
1969.
17. Nelson, W.O., Ciba Found. Colloq. Endocr • , ~: 271, 1952.
18. Bahn, R.C., Lorenz, N., Bennett, W.A. and Albert, A., Endocri-
nology, 53:455, 1953.
19. Davidson;-J.M., Coutopou10s, A.N. and Ganong, W.F., Acta Endocr.,
(Kobenham1), 34: 169, 1960.
20. Johnsen, S.G., (1970) to be published.
21. Johnsen, S.G., Acta Endocr. (Kobenhavn), 28: 69, 1958.
22. Johnsen, S.G., Acta Endocr. (Kobenhavn),~ 209, 1959.
23. Rosemberg, E. and Joshi, S.R., ~ Gonadotropins 1968, Rosemberg,
E. ed. Geron-X, Inc., Los Altos, Calif. p 91, 1968.
24. Johnsen, S. G., Hormones, 1, 1970, January issue.
25. Clermont, Y., Amer. J. Anat., ~: 35, 1963.
26. Kingsley Smith, B.V. and Lacy, D., Nature (London), 184: 249,
1959.
27. Broke1mann, J., Z. Ze11forsch., 59: 820, 1963.
28. Lacy, D., J. Roy. Micr. Soc. 79:~09, 1960.
29. Lacy, D., Brit. Med. Bull., 18: 205, 1962.
30. Lacy, D., Endeavour, 12: 101, 1967.
31. Lofts, B., In Perspectives in Endocrinology, Barrington, E.J.W.
and Barker Jorgensen, C., eds., Academic Press, London and New
York, p 239, 1968.
32. He11baum, A.A. and Greep, R. 0., Endocrinology 32: 33, 1943.
33. Davis, T.E., Lipsett, M.B. and Korenman, S.G., J7 C1in. Endocr.,
11: 476, 1965.
34. Kirschner, M.A., Lipsett, M.B., and Collins, D.R., J. C1in.
Invest., 44:657, 1965.
35. Peterson,N.T., Midgley, A.R. and Jaffe, R.B., J. Clin. Endocr.,
1.§.: 1473, 1968.
36. Swerd10ff, R.S. and Odell, W.D., ~ Gonadotropins 1968, Rosemberg,
E., ed. Geron-X, Inc., Los Altos, Calif. p 155, 1968.
37. Biddu1ph, C., Meyer, R.K. and Gumbreck, L.G., Endocrinology, 26:
280, 1940. --
38. Greep, R.O., 1£ Sex and Internal Secretions. ed. Young, w.e.,
3rd ed. Williams & Wilkins, Baltimore, p 240, 1961.
39. Wattenberg, L.W., J. Histochem. Cytochem. ~: 225, 1958.
40. Levy, H., Deane, H.W. and Rubin, B.L., Endocrino10gy,65: 932,
1959. --
41. Jirasek, J.E. and Raboch, J., Ferti1. Steri1., li: 237, 1963.
1963.
42. Woods, J.E. and Domm, L.V., Gen. Compo Endocr., 7: 559, 1966.
43. Koudstaa1, J., Frensdorf, E.L., Kremer, J., Mudd;, J.M. and
Hardonk, M.J., Acta Endocr. (Kobenhavn), 22: 415, 1967.
44. Sengoku, K., Bull. Tokyo Med. Dent. Univ., 14: 51, 1967.
45. Christensen, A.K. and Mason, N.R., Endocrin~ogy, 76: 646, 1965.
46. Baillie, A.H. and Mack, W.S., J. Endocr., 35: 239,-r966.
47. Baillie, A.H., CaIman, K.C., Ferguson, M.M:-and Hart D. McK.,
J. Endocr., 34: 1, 1966.
48. Lacy, D., Lofts, B., Kinson, G., Hopkins, D. and Dott, H., Gen.
Compo Endocr., 1: 693, 1965.
244 S. G. JOHNSEN
49. Maddock, W.O. and Nelson, W.O., J. Clin. Endocr., 12: 985, 1952.
50. Plate, W.P., Ned. T. Verlosk. 63: 83, 1963.
51. Lipsett, M.B., Wilson, H., Kirschner, M.A., Korenman, S.G.,
Fishman, L.M., Sarfaty, G.A. and Bardin, C.W., Recent Progr.
Hormone Res., ~: 245, 1966.
52. Fishman, L.M., Sarfaty, G.A., Wilson, H. and Lipsett, M.B., In
Endocrinology of the Testis, Wolstenholme, G.E.V., ed., J. a~
A. Churchill Ltd., London, p 156, 1967.
53. Witschi, E. and Chang, C.Y., ~ Comparative Endocrinology,
Gorbman, A. ed. John Wiley, New York and London, p 149, 1959.
54. Ryan, K.J. and Short, R.V., Endocrinology, ~: 108, 1965.
55. Ryan, K.J. and Petro, Z., J. Clin. Endocr., 26: 46, 1966.
56. Falck, B., Acta Physiol. Scand., Supple ~: 163, 1959.
57. Mancini, R. E., Castro, A. and Seiguer, A.C., J. Histochem.
Cytochem., 11:516, 1967.
DISCUSSION
LUNENFELD: Dr. Johnsen, it is difficult for me to understand that a

change in pituitary gonadotropin secretion will occur after the change
in spermatogenesis.
JOHNSEN: I think that this question can be answered quite easily.

If failure of spermatogenesis is induced experimentally, gonadotro-
pin excretion rises.
PAULSEN: Dr. Lunenfeld, if you expose the testes of normal males to

acute doses of irradiation, the germinal epithelium is damaged. Serum
and urinary FSH levels increase from normal to pathologically ele-
vated levels. When spermatogenesis recovers FSH levels return to
normal.
CHRISTENSEN: This question of possible estrogen secretion by the

Sertoli cells has been debated, of course, for many years. Some of
the evidence comes from tumors, which are always a little suspect,
but as I understand it, interstitial cell tumors are commonly mas-
culinizing, Sertoli cell tumors are sometimes feminizing (in dogs)
and seminomas of the germ cells usually produce no hormones. A
number of years ago I developed a way of separating the interstitial
tissue from the seminiferous tubules in the rat testis, and N. Mason
and I tested the transformation of progesterone to androgens by
these components in vitro (Christensen and Mason 1965). In parallel
runs we incubated with androgens to see if they would be transformed
into estrogens, and we obtained no activity in either the seminiferous
tubules or the interstitial tissue. I understand that Lofts has taken
this up and has claimed some activity, although it has only been pub-
lished in abstract form. With regard to electron microscopy, it is
true that Sertoli cells have smooth endoplasmic ,reticulum to an ex-
tent, but not nearly as well developed as in the interstitial cells.
It varies a good deal with the species. Jost Brokelmann studied the
rat, and I am not aware that he saw any changes at the time when the
sperm are descending in the trunks of Sertoli cell cytoplasm toward
the basement membrane, which one would expect if your interpretation
is correct. That would be good to examine with the electron micro-
scope. The mitochondria in Sertoli cells are not unusual, and in
fact are quite slender and not particularly characteristic of those
seen in steroid secreting cells. But one should not be surprised
about this i f the cells lack the 3 f3 -hydroxysteroid dehydrogenase.
If this is so, then they probably also lack the 20, 22-lyase, and
the mitochondria would not therefore be playing any direct role in
steroid biosynthesis.
JOHNSEN: I think I quoted Brokelmann correctly in saying that he

found these functional changes when the spermatozoa had dipped down
in the epithelium. Dr. P. F. Larson in our group is continuing this
kind of research in the human. It is unpublished yet, but Dr. Larsen
thinks that he is able to demonstrate marked changes in the smooth
endoplasmatic reticulum in the Sertoli cell at the particular stage
of spermatogenesis when the spermatozoon performs its diving and not
in the others. That means that when a Sertoli cell has the dipping
down of its own spermatozoa, these changes are observed.
CHRISTENSEN: The clusters of spermatids dip twice toward the Sertoli

cell nucleus in rats.
LIPSETT: I cannot eliminate the possibility of a second hormone but

I question whether it is an estrogen. After all, in men, the quan-
tities of estradiol and of estrone are the same as in post-menopausal
women, and yet we know very well that the levels of FSH and LH are
greatly different in these two groups. To account for the differ-
ences you either have to postulate an entirely different type of hor-
mone or a different estrogen.
BURGOS: Dr. Johnsen, you mentioned that only those stages which have
spermatozoa or late spermatids are significant in producing an in-
hibitory substance. If this is true, the Sertoli cell related to
late spermatids must be different from the others. Have you any
comments on this possibility?
JOHNSEN: Dr. Larsen in our group has found that the Sertoli cell
having late spermatids attached is different, but I am not able to
go further into this at this moment.
COUROT: Could you comment further on the relationship between Ser-

toli and Leydig cells? What is the evidence for that?
JOHNSEN: It was purely hypothetical. If we think that the Sertoli

cell produces steroids, we must realize that it cannot proceed to
bio novo synthesis of steroids. Therefore, one must find some pre-
246 S. G. JOHNSEN
cursors for it. The Leydig cell may provide precursors.
DICZFALUSY: I wonder whether it would be possible to use specific

anti-testosterone and anti-estradiol sera for localizing intracell-
ular steroids?
PAULSEN: Certain aspects of Dr. Johnsen's data have deficiencies.

Dr. Johnsen is still using the mouse uterine weight assay which mea-
sures the net effect of FSH and LH. According to Dr. Diczfalusy
and others this assay is more sensitive to changes in LH than FSH
levels. More importantly, there are patients with oligospermia who
do not show spermatids in their testicular biopsy and who have nor-
mal FSH levels. I would like to show some data that Drs. J. Leonard,
R. Leach and I have accumulated which bear on this issue.
Although the precise feedback mechanism controlling FSH secretion

is unknown, it has been postulated that FSH levels are reciprocally
related to spermatogenic activity. This concept is based largely
on observations (Nelson, W.O., Fertil. Steril., 1: 477, 1950, Howard,
R.P., Sniffen, R.C., Simmons, F.A. and Albright, F., J. Clin. Endocr.,
10: 121, 1950, Heller, C.G., Paulsen, C.A., Mortimore, G.E., Jungck,
E:C. and Nelson, W.O., Ann. N.Y. Acad. Sci., ~: 685, 1952, and
Johnsen, S.G., Acta Endocr. (Kobenhavn), Supple 90: 99, 1964) made
prior to the availability of specific methods for measuring FSH.
To re-examine this thesis, the following studies were performed.
Serum FSH was measured by radioimmunoassay in 64 androgenically
normal but oligospermic males; all had normal serum LH as deter-
mined by radioimmunoassay. To our surprise, the majority had FSH
levels within the range observed in our control population, that is
54 patients exhibited serum FSH levels between 230-580 m 11 g LER 907/
m!. In the 10 patients that demonstrated elevated FSH levels, the
values ranged from 600-1400 l1g LER 907/ml. Our control population
consisted of 42 androgenically normal males with sperm counts >60
million/mI. Furthermore, no direct correlation between FSH levels
and sperm counts existed in this oligospermic group. Johnsen
(Johnsen, S.G., Acta Endocr. (Kobenhavn), Supple 2Q: 99, 1964) ini-
tially suggested that in man, the residual body which is split off
from the spermatid is the regulator of gonadotropin production. To-
day he has proposed that the Sertoli cell is the source of an un-
defined FSH inhibitor, possibly an estrogen, and that the release
of this inhibitor by the Sertoli cell is directly dependent on the
late stages of spermatogenesis. Our studies do not support the
thesis of a direct reciprocal relationship between spermatogenic
activity and FSH secretion. To study this question, the quantita-
tive histologic characteristics of testicular biopsy specimens were
examined in twelve additional oligospermic males in whom FSH mea-
surements and sperm counts were also available. These men were
divided into two groups on the basis of their urinary FSH titers
(Steelman-Pohley method): a) normal FSH (n= 7) 3.0 to 18.0 I.u.1
24 hours; b) elevated FSH (n = 5) 26.0 to 54.0 I.U./24 hours. The
severity of the oligospermia in the two groups was comparable, i.e.

mean of 11.1 million/ml vs. 13.3 million/ml. Germinal cell quanti-
tation was performed in the following manner. The histologic sec-
tion of an entire biopsy was photographed and those tubules which
had been cut perpendicular to their long axis were identified and
numbered. The individual germinal cells were counted in 10 such
tubules per biopsy specimen. The cell counts were expressed as the
mean cell count per tubule and the data for the two groups compared.
There was no significant difference in the germinal cell counts be-
tween these two groups, and therefore no direct correlation between
germinal cell populations and FSH levels. One of the patients with
normal titers of FSH and less than 1 million/ml sperm count had vir-
tually no (fewer than 6) spermatids in an entire testicular biopsy
section. Therefore, we have to reject the concept that the germ
cells per se directly control FSH secretion. We have to consider
that there is an "independent station" which controls FSH secretion.
This station may be the Sertoli cell or the Sertoli cell plus the
limiting membranes. You may have a normal "station" with abnormal
spermatogenesis or an abnormal "station" with abnormal spermatogenesis.
The nature of the feedback signal from this independent station re-
mains unknown.
JOHNSEN: I am of course very well aware that a total gonadotropin

assay is not the right answer; but the cases presented were collected
over a five year period. We are now pursuing the matter by specific
FSH and LH determinations in the urine by bioassay. Dr. Christensen,
can confirm that, in the poor testis lacking spermatozoa, he finds
a rise in FSH excretion. So, I do not have an explanation of the
difference in our findings.
Dr. Paulsen, in your patient completely lacking spermatids you did

not observe an FSH rise. Was hypopituitarism the cause of the sper-
matogenic defect?
PAULSEN: It does not seem reasonable to me to postulate that all

patients with oligospermia and normal FSH levels have hypopituitarism.
Our patients had no evidence of a pituitary lesion.
MACLEOD: Dr. Paulsen, I have a slight quarrel with you about your
definition of oligospermia. For example, if we do take a population
of men at large, we find that the mean sperm count will be in the
region of 95 million/ml. But it does .not follow that levels below,
even far below that are not able to produce conception with relative
ease. So that if you start at the 40 million level as your defini-
tion of oligospermia, I think you may be a little bit far afield in
terms of defining it. I fully agree, though, with what you have said:
that there is probably no relationship between even severe oligosper-
mia (that is, under the 10 million level) and gonadotropin levels.
PAULSEN: I would agree, Dr. MacLeod. For our studies we arbitrar-

248 S. G. JOHNSEN
ily selected 40 million/ml as the cut-off point. However, if you

select cases with sperm counts below 10 million/ml, no correlation
with FSH levels exists.
THE ROLE OF FSH, ICSH, AND ENDOGENOUS TESTOSTERONE DURING TESTICULAR
SUPPRESSION BY EXOGENOUS TESTOSTERONE IN NORMAL MEN
Carl G. Heller, Howard C. Morse, Mike Su, and Mavis J.

Rowley
Division of Reproductive Physiology, Pacific Northwest

Research Foundation, 1102 Columbia Street, Seattle,
Washington 98104
Testosterone administered to normal men is know to depress sper-

matogenesis to azoospermia, as reflected in the ejaculate, and to
depress Leydig cell morphology to the point that only mesenchymal
cells and fibroblasts are identifiable. It is also know that uri-
nary total gonadotropins are concomitantly depressed to the pOint
that they are undetectable in 24 hour urine aliquots administered
to single assay rats (1). It has been widely assumed, but not
proved, that the depression of human testicular function by admin-
istering testosterone is secondary to pituitary suppression.
Three possible mechanisms of action of testosterone upon the

testis were originally postulated in the following investigation:
1) that the depression is solely due to total pituitary gonadotropin
suppression, 2) that the depression is solely due to a direct effect
of testosterone upon the germinal epithelium and/or Leydig cells or
3) that the depression is due to a combination of direct testicular
and indirect gonadotropin suppression.
To test which of these three postulates apply, the following

protocol was planned. Testosterone enanthate (200 mg intramuscular-
ly (I.M.) per week) was to be administered to normal men until the
following three events occurred: azoospermia, Leydig cell disap-
pearance and disappearance of gonadotropins in 24 hour urine ali-
quots. Then, while continuing the testosterone administration at
the same level, human chorionic gonadotropin (HCG) was to be added
for its interstitial cell-stimulating hormone (ICSH) effect (4000
International Units (I.U.) HCG I.M. per two days). If Leydig cell
morphology could thus be restored to normal, as revealed by testicu-
lar biopsy examination, then a direct effect of testosterone upon
249
250 C. G. HELLER, H. C. MORSE, M. SU, AND M. J. ROWLEY
Leydig cell depression could be eliminated from further considera-

tion.. The suppression of Leydig cells by testosterone would then
be attributed to the suppression of pituitary ICSH (if no restora-
tion occurred the direct effect of testosterone on Leydig cells
must still be considered). Assuming restoration of Leydig cell mor-
phology, the plan was to add follicle-stimulating hormone (FSH) in
the form of Pergonal in order to restore spermatogenesis, while con-
tinuing to administer the testosterone enanthate and continuing the
HCG at the same levels for each. If spermatogenesis were now re-
stored we could conclude that the inhibitory effect of testosterone
upon the germinal elements was mediated by suppression of FSH.
To this end three healthy men, ages 37, 33 and 29 were chosen
to serve as subjects. They were determined to be reproductively
normal by hormonal studies, sperm concentration and examination of
control testicular biopsy specimens. Testicular biopsies were ob-
tained and prepared using the methods of Rowley and Heller (2).
Testosterone enanthate (Delatestryl, Squibb) was administered

at the rate of 200 mg I.M. once weekly for eight to 12 months. Hu-
man chorionic gonadotropin (APL, Ayerst, 4000 I.U. I.M. every second
day) was added beginning four months after the initiation of testos-
terone enanthate administration. The duration of HCG administration
was four and one-half months. Testosterone enanthate was continued
alone on two subjects after HCG was discontinued.
Seminal fluid specimens were collected once weekly and examined

for sperm concentration, motility and volume. Concentration was
measured using the hemocytometer method. Each subject was used as
his own control. Sperm morphology was examined by Dr. John MacLeod
of Cornell University Medical College.
Twelve-hour (overnight) urine samples were collected and pooled

in 48 hour collections. Total gonadotropins and ICSH determinations
were performed on aliquots of the urine pools. Urines were kept
below SoC until the extraction procedures were carried out. Urinary
gonadotropins were extracted using the kaolin-acetone method of
Albert (3). Total gonadotropin levels were determined by measuring
the ovarian weights of immature rats and expressed as mgU/24 hours
of a standard human urinary postmenopausal preparation (Pergonal P-
22C). Urinary ICSH was determined biologically using the method of
Greep et al. (4), measuring the increase in ventral prostate weight
in hypophysectomized immature male rats. The values were expressed
as mgU/24 hours of the standard preparation (Pergonal P-26E). Tes-
ticular biopsies were examined quantitatively using the germ celli
Sertoli cell ratio method as developed by Heller et al. (5).
The administration of testosterone enanthate 200 mg I.M. weekly

for four months resulted in barely detectable or absent total uri-
nary gonadotropins, azoosp~rmia as judged from ejaculated seminal
TESTICULAR-PITUITARY INTERRElATIONSHIP 251
fluid samples, disappearance of Leydig cells as judged by testicular

biopsy examination and marked reduction in numbers of Sd, Sc, Sb and
Sa spermatids, marked reduction in pachytene, zygotene and leptotene
spermatocytes, slight reduction in pre leptotene spermatocytes and type
B spermatogonia, questionable reduction of type A pale spermatocytes
and no apparent reduction in type A dark spermatocytes (Figs. 1-2)
200 '"''III qw ~-_ . ,
30. ,.
-HCG
250 .000 IU q.2d
,
v
v
:I
200
~
e
!
100
•• 0.'
20 30 40 .0
WilKS
Fig. 1. Sperm count, total gonadotropin (gth) and ICSH values

shown in a subject during control, testosterone and testosterone-
HCG administration periods. Dashed lines across control period in-
dicate range of normal gth and ICSH values. Bars with arrows point-
ing to bottom line indicate absent or barely detectable gth or ICSH.
Jc_. ~ I I I I
_,,.\1 HCG tOOO IU.24 .hli. 1<11l1li119_ ~~
:; I ~ _ • l ' "..,.,.., I+CO .ho,. '."9'''''-' 4_1o._cf

..j
~
t,
I
.
~ .t
i::ll.J .' I' '.

II ~
•
I
Sa
11 II
'"s,.",fIIIOt.li. .
n
~
Id
Fig. 2. Comparison of Sertoli cell ratios in biopsies taken during

testosterone administration and twice during the additional admin-
istration of HCG. Control values for each cell type were determined
from the average of the quantitation of 50 normal testicular biop-
sies. Arrows pointed to bottom line indiaate zero values.
Adding 4000 I.U. of HCG administered I.M. every second day

brought about complete restoration of Leydig cell morphology in all
five biopsies taken during HCG administration. Note should be made
that two of these biopsies were taken as early as three weeks after
the initiation of HCG administration.
Therefore, since testosterone caused Leydig cell disappearance

and adding HCG caused restoration of Leydig cells to normal, we con-
clude that testosterone-caused Leydig cell depression is mediated
by suppressing anterior pituitary ICSH secretion or release.
In the meanwhile, much to our surprise, the weekly seminal

fluid examinations revealed an increase in sperm concentration by
one month in one subject and by the second or third months in the
other two. Between the third and the fourth month of HCG adminis-
tration each subject had reached pre-testosterone treatment control
levels (Fig. 1). Two reached the average of the control levels and
one rose above the minimal control levels but did not rise as high
as his pre-treatment average.
Quantitation of the elements of the germinal series in the two

testicular biopsies obtained three weeks after the initiation of
HCG revealed a partial restoration of leptotene and pachytene sper-
matocytes and significant but small increases in each of the four
spermatid types. The second biopsy, obtained at four and one-half
months following initiation of HCG, revealed complete restoration
of each of the depressed germinal cells to pre-treatment levels in
the two subjects whose sperm concentration had returned to average
pre-treatment levels (Fig. 2). The third subject, whose sperm con-
centration had not reached the average of his pre-treatment levels,
revealed definite but incomplete restoration of each of the germinal
cells.
These results suggest that 1) testosterone does not exert a

direct depressive effect upon either the germinal cells or the Ley-
dig cells since both elements can be restored to normal by adminis-
tering HCG, 2) administering RCG during testosterone administration
restores spermatogenesis either by providing sufficient FSR to ob-
tain this effect or because testosterone does not depress FSR secre-
tion or release in normal male human subjects, and further 3) RCG
stimulates enough Leydig cell androgen production to aid in the re-
storation of spermatogenesis in the presence of FSR.
As a consequence of the unexpected germinal cell response to

HCG administration without the addition of FSH there was no reason
to follow the original protocol. Thus no Pergonal was administered
and the RCG was discontinued. In order to confirm the effect of
testosterone upon the two testicular systems the testosterone was
continued for an additional four months in the two subjects who had
complete restoration of spermatogenesis.
The two subjects ~esponded as before in that the weekly semi-

nal fluid values decreased to azoospermia, and the gonadotropins
dropped to levels barely detectable or absent. The Leydig cells
disappeared and the germinal elements (except type A dark spermato-
gonia) again became depressed in the subject whose biopsy was avail-
able for examination. Parenthetically, observations on sperm mor-
phology revealed a drop in normal forms during the initial testos-
terone enanthate administration and a return to normal during the
administration of HCG.
As a general proposition, spermatogenesis appears to reach

completion in most mammals under the influence of two different hor-
mones, testosterone and FSH. First testosterone must be present in
sufficient amounts to bring about maturation of the seminiferous
tubules. The tubules may be considered a part of the rest of the
genital system which is dependent upon testosterone for its main-
tenance and development, e.e., epididymis, was deferens, prostate,
seminal vesicle, etc. In man, when such maturation of the semini-
ferous tubules is induced by testosterone, whether exogenous or en-
dogenous, there is an accompanying differentiation of the previously
undifferentiated Sertoli cells, spermatogonia and the occasional
spermatocyte, but no noticeable progression of germinal cell matura-
tion. This can be judged from administering HCG to true hypogonado-
tropic eunuchoids (6). After testosterone has provided basic tes-
ticular support, the second hormone, FSH, is able to initiate pro-
gressive and complete spermatogenesis. Follicle-stimulating hormone
however is unable to influence human spermatogenesis when adminis-
tered alone as HMG (Pergonal) (7).
With the above in mind we wished to explore the proposition

that testosterone might differentially depress ICSH completely with-
out concomitantly completely depressing FSH. The preferred way of
testing this proposition would be, of course, to determine plasma
FSH and ICSH levels. These determinations were not performed on
the three subjects reported and must await a repetition of the ex-
periment. As a substitute that could be performed with relative
rapidity normal subjects were chosen for administration of a high
potency and more rapidly acting androgen. Thus, testosterone pro-
pionate (Oreton, Schering) was administered I.M. at 100 mg levels
daily for five days. Five control observations were recorded for
plasma FSH and ICSH using modifications of Odell's radioimmunoassay
methods (8,9). Daily observations were made during androgen admin-
istration and every two to three days for two weeks following ces-
sation of androgen. ICSH disappeared entirely within 24 hours of
the first injection and remained undetectable for the five day in-
jection period (any value below 2.0 mU/ml is considered as zero or
undetectable in our laooratory). In distinct contrast the plasma
FSH values, determined simultaneously, remained unaltered during and
following testosterone propionate administration. Thus under the
specific circumstances of this experiment it must be concluded that
testosterone can produce a dichotomous effect upon plasma FSH and

rCSH levels (Table 1).
Table 1. Plasma Gonadotropin Before, During and After Testosterone

Administration.
SUBJECT PM 315 SUBJECT PM 214

reSH FSH rCSH FSH
Date
(mU/ml)'~ (mU/ml)'~ (mU/ml)>'< (mU/ml)>'<
Control
3-9 5.66 8.07 4.43 7.03
4-11 3.13 7.47 3.98 7.30
4-12 6.00 7.03 4.23 7.30
4-14 5.25 7.20 3.58 6.75
10-11 4.40 7.75 3.45 6.50
Testosterone propionate 100 mg r.H. for five days

10-12 1.90 7.65 1.43 5.40
10-13 1.70 7.17 2.08 6.75
10-14 1.63 7.25 2.05 6.50
10-15 1.73 6.63 2.03 6.18
10-16 1.75 6.50 1.68 6.60
Post-
Testosterone
10.20 8.25 7.20 5.23 6.45
10-23 7.38 8.25 6.90 6.68
10-27 5.03 8.03 3.70 6.73
10-30 5.15 7.53 3.93 6.85
* Standard:Human Pituitary Extract Reference Preparation LER 907
Swerdloff and Odell (10) administering 20 or 50 mg of fluoxy-

mesterone daily for five days to five normal men and Franchimont
(11) using a single injection of 50 mg testosterone propionate re-
vealed a similar dichotomy in response, i.e., no change in FSH and
undetectable ICSH (LH).
Neither Swerdloff and Odell's nor our experiments to date have

been of sufficient duration to extrapolate the data to long-term
effects. Bogdanove (12) has demonstrated a delayed effect of andro-
gens on FSH release in rats.
The effect of androgen upon ICSH in the normal human male can
be accep~ed as causi~g marked suppression. This is inferred from
our observations of depression of total urinary gonadotropins, and
suppression of urinary ICSH, in the one case, following testoster-
one enanthate administration. It is also inferred from the complete
early plasma ICSH suppression following administration of fluoxymes-

terone and testosterone propionate. Thus there appears to be a
strong likelihood that a dichotomy exists between ICSH and FSH
suppression.
Assuming that FSH is not suppressed or only incompletely sup-

pressed by administering 200 mg testosterone enanthate I.M. per week,
we speculate that the restoration of spermatogenesis following the
addition of HCG is due wholly or in part to the presence of plasma
FSH of anterior pituitary origin.
Continuing with this assumption, each result of the testoster-

one enanthate plus HCG experiment is readily accounted for except
one: the suppression of spermatogenesis by administering testoster-
one enanthate (or other androgens). The enigma that requires ex-
planation is based upon the assumption that the presence of only two
hormones, testosterone and FSH, is required for completion of sper-
matogenesis in man. In this experiment each hormone is assumed to
be present; testosterone being administered exogenously and FSH not
having been suppressed. The only immediately apparent interference
with regulation of testicular function is the suppression of ICSH.
This in turn stops the endogenous production of androgens, which in
turn provides the explanation for the cessation of complete sperma-
togenesis. It is likely that the concentration of androgen provided
by the Leydig cells, which are in such close approximation to the
seminiferous tubules, is far higher than that provided to the germi-
nal cells by an injection of 200 mg of testosterone enanthate given
I.M. once weekly. Thus the effect of testosterone enanthate given
I.M. is to lower the androgen levels available to the germinal epi-
thelium, due to the inhibition of ICSH.
In applying this explanation, reference must be made to obser-

vations on the maintenance of spermatogenesis in the hypophysecto-
mized rat by exogenously administered androgens (13-16). Woods and
Simpson (17) found that spermatogenesis in hypophysectomized rats
could be poorly maintained by ICSH alone but well maintained by ICSH
plus FSH or testosterone propionate plus FSH. In these and many
confirmatory observations the dosage administered and the concentra-
tion of plasma testosterone attained are far in excess of those pos-
sible for so much larger a mammal as man.
That the location of the Leydig cell so close to the germinal

epithelium is an important factor in providing concentrated amounts
of androgen is seen in man in the case of the "fertile eunuch" (18).
Maintenance of spermatogenesis has been produced in hypophysecto-
mized rats by Dvoskin (19) and in hypophysectomized rhesus monkeys
by Smith (20) by intratesticular administration of testosterone in
pellet form. Smith found tubules nearest the deposited androgen
undergoing complete spermatogenesis and those more remote remaining
in the degenerate state. Dvoskin reported that if the same amount
of androgen was administered subcutaneously or intratesticularly the

latter was far more effective in maintaining spermatogenesis.
We conclude that in normal men testosterone causes testicular

Leydig cell and germinal cell depression without exerting a direct
effect upon testicular activity, but does so indirectly by inhibit-
ing production of ICSH. We postulate that FSH production is un-
affected or only partially suppressed. Under these circumstances
we assume that the inhibition of spermatogenesis is due to a de-
crease in the normally high androgen concentration at the germinal
cell site due to the absence of ICSH effect upon Leydig cells.
ACKNOWLEDGEMENTS
We wish to thank Edward C. Reifenstein, Jr., M.D. of The

Squibb Institute for supplying the Delatestryl, Jack Howard, M.D.
of the Schering Corporation for the supply of Ore ton and John
Jewell, M.D. of Ayerst-McKenna for a supply of APL.
This investigation was supported by grants-in-aid from The

Ford Foundation (680-0906), the Public Health Service (HD 00804)
and the U.S. Atomic Energy Commission (AT (45-1)-1780).
REFERENCES
1. Heller, C.G. and Nelson, W.O., Recent Progr. Hormone Res.,

1: 229, 1948.
2. Rowley, M.J. and Heller, C.G., Fertil. Steril., .!.Z.: 177, 1966.
3. Albert, A., Proc. Staff Meeting Mayo Clinic, 30: 552, 1955.
4. Greep, R.O., VanDyke, H.B. and Chow, B.F., Proc. Soc. Exp. BioI.
Med., 46: 644, 1941.
5. Heller, C.G., Heller, G.V. and Rowley, M.J., In: Progress in
Endocrinology, ed. Gual, C., Excerpta Medica Found., p. 1012,
1969.
6. Heller, C.G., Lalli, M.F. and Rowley, M.J., In: Pharmacology of
Reproduction, Proc. 3rd Internat. Pharmacol. Meeting July 1966,
Pergamon Press, Oxford and New York, ~: 61, 1968.
7. Paulsen, C.A., In: Gonadotropins, 196~ ed. Rosemberg, E., Geron-
X Inc., Los Altos, Calif., p. 491, 1968.
8. Odell, W.D., Parlow, A.F., Cargille, C.M. and Ross, G.T.,
J. Clin. Invest., 47: 2551,1968.
9. Odell, W.O., Ross, G.T., and Rayford, P.L., J. Clin. Invest.,
46: 248, 1967.
10. Swerdloff, R.S. and Odell, W.D., In: Gonadotropins, 1968, ed.
Rosemberg, E., Geron-X Inc., Los Altos,Calif., pISS, 1968.
11. Franchimont, P., Proc. Internat. Symp. on Protein and Polypep~
tide Hormones, ed. Margoulies, M., Excerpta Medica Found.,
Amsterdam, p. 99, 1968.
12. Bogdanove, E.M., Anat. Rec., 157: 117, 1967.
13. Walsh, E.L., Cuyler, W.K. and McCullagh, D.R., Amer. J. Physiol.,
107: 508, 1934.
TESTICULAR· PITUITARY INTERRELATIONSHIP 257
14. Nelson, W.O. and Gallagher, T.F., Science, 84: 230, 1936.
15. Cutuly, E., McCullagh, D.R., and Cutuly, E.C., Amer. J. Physiol.
119: 121, 1937.
16. Clermont, Y. and Harvey, S.C., In: Ciba Found Colloq. Endocr.
~: 173, 1967.
17. Woods, M.C. and Simpson, M.E., Endocrinology, 69: 91, 1961.
18. McCullagh, E.P., Beck, J.C. and Schaffenburg, C.A., J. Clin.
Endocr., 13: 489, 1953.
19. Dvoskin, 8:, Anat. Rec., 99 : 329, 1947.
20. Smith, P.E., Yale J. Biol.Med., 12.: 281, 1944.
DlSCUSSIOO
ROSEMBERG: Dr. Heller, you indicated that testicular function was

suppressed under testosterone administration; that 1S, the sperm
count dropped. However, after subsequent addition of HCG, the sperm
count increased. Hence, you postulated that HCG stimulated the Ley-
dig cells to produce endogenous testosterone by virtue of the fact
that testosterone may be near the germinal epithelium and then ex-
erted a local effect. Do you have experimental proof for your state-
ment? It seems to me that when exogenous testosterone is administered
one is flooding the system.
HELLER: We do not have proof for our statement. The testosterone

levels that we can conveniently produce in man by intramuscular in-
jection are relatively low. In other words, the amount of testoster-
one theoretically reaching the germinal epithelium is much different
in amount than what one can do, let us say in a hypophysectomized rat
where one can maintain s¥ermatogenesis by injecting testosterone.
All the other evidence is rather indirect, as has been shown at this
meeting. Dr. Steinberger showed a case with a testicular tumor in
one testis in which the tubules that were near the testosterone-pro-
ducing tumor continued to have normal spermatogenesis whereas on the
other side no spermatogenesis was observed.
ROSEMBERG: Dr. Steinberger, what was the level of testosterone in

this patient?
E. STEINBERGER: We are presently measuring the circulating hormone

level, so 1 do not have the answer but I will be surprised if it is
not extremely low.
ROSEMBERG: If this is so, it is difficult for me to visualize that

the amount produced by the Leydig cells under HCG stimulation could
overcome whatever is already there.
HELLER: The Leydig cells were morphologically altered by exogenous

testosterone administration, suggesting lowered testosterone output.
But you ask me for proof; and I have no proof of this; it remains
a logical assumption, but just an assumption.
SAVARD: There is a most important intra-testicular, as well as in-

tra-ovarian, circulation of hormones produced within the gonad. I
call your attention to the work of Lindner, some years ago in Dr.
Mann's laboratories in Cambridge, who demonstrated the high testos-
terone content of the lymph of the testis. He found it many times
greater than the peripheral testosterone concentrations, comparable
and sometimes exceeding the concentrations of the spermatic venous
blood. Lindner also suggested that this could perhaps explain the
phenomenon of the syndrome of the fertile eunuch, who has sufficient
local testosterone concentration to support spermatogenic function,
yet not enough peripherally to masculinize him.
HUDSON: With the administration of fluoxymesterone, plasma levels

of testosterone are suppressed. Continuing treatment with fluoxy-
mesterone and administering HCG, peripheral plasma levels of testos-
terone are restored to normal. I really believe that Dr. Savard's
comments are highly relevant: namely, what is the local concentra-
tion of testosterone around the tubules, and what is its signifi-
cance on nurturing the germinal epithelium?
JOHNSEN: In Dr. Steinberger's case, there was only growth of the

tubules and presence of a limited number of spermatocytes.
E. STEINBERGER: Dr. Johnsen, you are absolutely correct; as I

clearly pointed out when discussing the photomicrographs, the sper-
matogenesis did not proceed beyond young spermatids. Dr. Heller's
paper, together with data reported by Dr. Bergada on a case of Ley-
dig cell hyperplasia showing localized areas of spermatogenesis and
with our demonstration of spermatogenesis in a tumor-bearing testi-
cle of a 6-year old boy, presents a good case for the critical role
of testosterone in the process of initiation of spermatogenesis in
the human testicle. Furthermore, we provided evidence that in the
rat, FSH is necessary for the late stages of spermatid development.
I think that demonstration of FSH by Dr. Heller in his experimental
subject suggests that, in the human, FSH may also play an important
role in maturation of the spermatids.
HELLER: Many years ago Carl Moore implanted crystals of testosterone.

Wherever the crystals were, spermatogenesis was maintained; while
peripheral to that, they were not.
PAULSEN: Dr. Heller, it occurs to me that something curious is go-

ing on which is not understandable. For example, when Maddock and
Nelson administered HCG to normal adult men, oligospermia or azoo-
spermia developed. Histologically, the seminiferous tubules be-
came hyalinized and the Leydig cells became small and darkly stained.
They postulated that these changes were due to excessive steroid

hormone production. Now suppose they had added testosterone to
their HCG regimen. This would have further increased the intrates-
ticular testosterone concentration. Thus the data of Maddock and
Nelson appear to be in conflict with your experiment. Can you ex-
plain this?
HELLER: We have repeatedly confirmed the Maddock-Nelson work. And,

for those of you who are not familiar with it, it consists simply
of administering human chorionic gonadotropin long enough, in large
enough amounts, resulting in azoospermia in normal men. During
this time, not only are testosterone and the l7-KS steroids in-
creased, but the hormone, in the urine at least, that increases the
most is estrogen. I would postulate that human chorionic gonado-
tropin, under the circumstances of Maddock and Nelson's experiments
and under th~ circumstances that we have given it, is causing fail-
ure of spermatogenesis because of the high concentration of estro-
gen.
PAULSEN: But you would still be getting the same equivalent amount
of estrogen in the urine, Dro Heller, with testosterone and HCG
administration. Therefore, your explanation does not appear to be
plausible. It still remains doubtful how you were able to restore
spermatogenesis in your study. I would like to believe that it is
the retention of normal endogenous FSH levels which was fundamental
to the restoration of spermatogenesis in your study. But is it real-
ly fair, Dr. Heller, to compare your twenty-week administration of
testosterone enanthate to the five-day administration of testoster-
one propionate?
HELLER: No it is not. Dr. Paulsen brings up the fact that we did

not determine plasma ICSH and plasma FSH in the experiment with
testosterone enanthate, which was a long-term experiment. This ex-
periment is now under way.
DICZFALUSY: May I suggest another experiment. In the human female,

those who tried to work on the mechanism of action of contraceptive
steroids usually concluded that estrogen will inhibit FSH release
and secretion, provided it is given for a long period of time. Dr.
Heller, if you would give estrogen to your patients, you would see
the following: 1) whether or not it was indeed the estrogen which,
in combination with HCG, caused hyalinization, 2) whether or not
you really need FSH from the pituitary and 3) whether you could re-
place FSH with HMG.
CHAPTER IV

SECTION 3: GONADOTROPINS, PURIFICATION AND MEASUREMENT
USE OF STANDARDS IN GONADOTROPIN ASSAYS
Eugenia Rosemberg
Medical Research Institute of Worcester, Inc. ,
26 Queen Street, Worcester, Massachusetts
By definition a standard is "that which is set up and estab-

lished by authority as a rule for a measure of quantity weight, ex-
tent, value or quality." Principles governing the appropriate use
of standards in assays, whether the assays are biological, chemical,
immunological or microbial, have been established for a long time.
The most important principle governing the use of standards in

gonadotropin assays, is that a standard should be as alike the ma-
terial tested as possible. This implies that the use of one hormone
as a standard for another hormone represents a violation of the
principle of similarity.
One of the major sources of confusion in the literature con-

cerning both biologic and immunologic potency estimates of gonado-
tropic hormones has been the use of several standards. As an exam-
ple, in the bioassay field, a disparity of some 12-fold in the assay
of urinary LH was found when the assays were conducted by the ven-
tral prostate weight method (VPW) and by the ovarian ascorbic acid
depletion method (OAAD) (1). The discrepancy was understandable on
the basis that it was attributable to the physical and chemical dif-
ferences and possibly also to metabolic differences in the standard
used (NIH-LH-SI, an ovine pituitary LH) and the material (urinary
LH) tested (1-2). The discrepancy disappeared when a urinary stan-
dard was used for the assay of urinary LH. A similar observation
was made testing an LH preparation obtained from human pituitary
glands (R-469-2, kindly supplied by Drs. L.E. Reichert, Jr., and A.
Wilhelmi), in both the VPW and OAAD methods using ovine NIH-LH-SI
as standard (3).
As shown in Table 1, another example of this discrepancy is
263
264 E. ROSEMBERG
illustrated in the assay of ovine NIH-FSH-Sl and NIH-LH-Sl and the

Second International Reference preparation of Human Menopausal Go-
nadotropin (2nd IRP) (WHO/BS/723) a gonadotropin preparation ob-
tained from human urine.
TABLE 1
SPECIFIC ACTIVITY OF NIH-FSH-Sl AND NIH-LH-Sl, IU 2nd IRP/mg
Assay Weighted Mean Potency Ratio

No. No.
Assay Characteristics (95% CL)
of of
System Assays NIH-FSH-Sl NIH-LH-Sl
Animals bc A (ru/mg) (IU/mg)
26.5
AR 10 245 239 0.14
(24.3-28.8)
51.3
VPW 6 167 18.6 0.21
(43.9-59.9)
588
OAAD 21* 417 -26 0.30
(500-714)
*Lack of parallelism: 4 assays; Index of Discrimination ~~ = 11.5

The details of the assay methods used as performed in our labora-
tories have been described (1-4). (From: Rosemberg, E., In Gona-
dotropins 1968, ed. Rosemberg, E., Geron-X, Inc., Los Alto~ Calif.,
p. 383, 1968.)
In the bioassay field, at least six reference preparations or

standards were used to report values of gonadotropic activity in
pituitary tissue and body fluids. These were HMG-20A (5) obtained
from the urine of menopausal and post-menopausal women; AMW (6)
from the urine of normal males: the 1st I.oternational Reference
Preparation for Human Menopausal Gonadotropin (1st IRP, WHO Tech.,
Rep. Series, P. 9, 1959); the ovine pituitary preparations NIH-FSH-
Sl and NIH-LH-Sl (as well as subsequent batches), and the Second
International Reference Preparation of Human Menopausal Gonadotropin
(2nd IRP). Hence, one of the major sources of confusion concerning
biologic potency estimates has been the use of various standards
and therefore confusion arose concerning the conversion factors
used to relate one standard to another.
GONADOTROPINS, PURIFICATION AND MEASUREMENT 265
Rosemberg, E.
1 mg. 1st IRP ~ 0.004 mg. NIH-FSH-Sl AR* et ale (J. Clin.
1 mg. 1st IRP ~ 0.009 mg. NIH-LH-Sl vpw** Endocr. 24: 673,
1964)
1 mg. NIH-FSH-Sl 9.2 rat units AR Albert, A., Mayo

1 mg. NIH-LH-Sl 26.5 rat units vpw Clinic Proc. 40:
1 mg. NIH-LH-Sl 166.0 rat units OAADt 216, 1965.
1 mg. 1st IRP 0.14 IU 2nd IRP AR DRB/DF/82

1 mg. 1st IRP 0.5 IU 2nd IRP vpw DRB/DF/82
1 mg. NIH-FSH-Sl 26.5 IU 2nd IRP AR Rosemberg, E.

1 mg. NIH-LH-Sl 51.3 IU 2nd IRP vpw In Gonadotropins
1 mg. NIH-LH-Sl 588.0 IU 2nd IRP OAAD 1968, ed. Rosem-
berg, E., Geron-
X, Inc., p • 383 ,
1968.
*AR: augmentation reaction assay (7); ** VPW: ventral prostate
weight assay (1); tOAAD: ovarian ascorbic acid depletion assay (4).
(From Rosemberg, E. In Gonadotropins 1968, ed. Rosemberg, E., Geron-
X Inc., Los Altos, Calif., p. 387, 1968.)
It is well known that a protein molecule may lose functional or

biologic activity without concomitant loss of immunologic activity.
Hence, with the advent of immunological techniques for the measure-
ment of gonadotropic activity in pituitary tissue and body fluids,
the establishment of proper standards to be used in these systems
is of paramount importance.
Ideally the test material and the standard should be identical

at the molecular level. Hence, individual standards should be made
available for each gonadotropic hormone of human or animal origin.
However, the problem of setting up suitable standards for all forms
of gonadotropins is a real dilemma. Moreover, antisera to indivi-
dual hormones must be regarded as unique reagents. For example, al-
though cross reactivity has been found between HCG and LH antisera,
some antisera have been found to cross-react identically and others
only partially. Moreover, some antisera to ovine LH have been shown
to cross-react with human pituitary LH (HLH) (personal observations
and personal communication from L.E. Reichert). Also, some antisera
prepared against HFSH cannot recognize the difference between human
FSH, LH, HCG or TSH.
In order to set up proper reagents to be used in immunoassays,

it is obvious that it will be necessary to know the exact relation-
ship between immunologically active and biologically active mole-
cules; it would be also necessary to develop antisera specific to
266 E. ROSEMBERG
the hormone being measured. However, these conditions have not

been met as yet. In the interim it was found necessary to make
available antigens, antisera and a reference preparation for FSH
and LH radioimmunoassays (RIA). Consequently, the Institute of
Arthritis and Metabolic Diseases, NIH,through the National Pitui-
tary Agency (NPA) made available the reagents for RIA of human pi-
tuitary follicle stimulating (HFSH) and luteinizing (HLH) hormones
and a reference preparation designated as LER 907 derived from a
human pituitary tissue extract which contains both FSH and LH ac-
tivity.
The collaborative study conducted by the NPA (8) determined,

by both bioassay and radioimmunoassay techniques, the FSH and LH
content of four human pituitary tissue extracts differing widely in
FSH/LH ratio and of three hi2hlv purified (immunochemical grade)
preparations (one of HFSH and two preparations of HLH). The stan-
dard used was the 2nd IRP. However, the study was designed so that
all preparations could be assayed in terms of any, such as the ref-
erence preparation LER 907, or all of the remaining preparations.
The NPA study confirmed that a tissue pituitary extract (LER

907) is satisfactory as a radioimmunoassay reference preparation
for FSH and LH in pituitary tissue extracts. The study also showed
that the use of the 2nd IRP (urinary) as a radioimmunoassay standard
is associated with over estimation of the FSH and LH content of hu-
man pituitary tissue extracts. It has also been shown ( 9) that the
use of a human pituitary tissue reference preparation is associated
with radioimmunoassay under estimation of FSH and LH present in hu-
man urinary extracts. Moreover, it has been shown that the urinary
2nd IRP is a satisfactory standard for the radioimmunoassay of FSH
and LH in urinary gonadotropic extracts, the index of discrimination
(radioimmunoassay/bioassay) being close to unity (10). (A compre-
hensive review on the comparison of biologic and immunologic potency
estimates of human luteinizing hormone and human follicle stimulat-
ing hormone has been published by R.J. Ryan (11».
The availability of the NPA radioimmunoassay reagents (8) made

possible the study of the LH biologic and immunologic activity of
four human pituitary LH preparations and that of HCG. The standards
used were the 2nd IRP and LER 907. The radioimmunoassay technique
used was essentially that described by Odell et ale (12). For io-
dination, a purified human pituitary LH fraction DEAE-2-II, with an
LH potenc~ of 1,350 IU 2nd IRP/mg was used. The specific activity
of the 11 1-LH varied from 149 to 429 ~c/~g.
BIOASSAY RESULTS
All bioassays were calculated by appropriate computer programs.

For bioassay of LH activity, the OAAD method was employed; for the
bioassay of FSH, the augmentation reaction assay (AR) was used. The
concomitant standard and reference preparations tested with all bio-

assays were the 2nd IRP and LER 907, respectively.
Table 2 gives the bioassay results in terms of specific activi-

ty - i.e., International Units (IU) of FSH and LH activity per mil-
ligram of five preparations (including the reference material LER
907) and the specific activity by weight of the four pituitary pre-
parations in terms of weight of the reference preparation LER 907.
The FSH/LH ratios are also shown. The LH content of LER 907 was 54.2
lU/mg. One preparation, LER 856-1 exhibited an activity of 510
IU/mg. The other three preparations contained similar LH activity
(IU/mg) which was about 2.5 times higher than that of LER 856-1.
The four preparations differed in FSH/LH ratio by 74 to 185-fold
from the reference preparation. However, the FSH/LH ratio of LER
856-1 differed by about 4-fold from the other three materials.
LH RADIOIMMUNOASSAY RESULTS
All preparations were tested at several dose levels. The con-

comitant standard and reference preparation used were the 2nd IRP
and LER 907, respectively. All RIA assays were calculated by appro-
priate computer programs for parallel line assays. Table 3 presents
the LH RIA assays and the comparison of bioassay and immunoassay re-
sults.
The RIA results are given in terms of specific activity by

weight of the preparations in terms of weight of LER 907, and by
specific activity in IU's of the 2nd IRP per milligram. One prepa-
ration, LER 856-1 differed in LH activity by about 5-fold from LER
907, and between 3.5 to 5.7-fold from the other three preparations.
The LH activity of the last three preparations differed by about 20
to 30-fold from that of LER 907. The LH radioimmunoassay/bioassay
(imm/Bio) ratio was close to unity when a pituitary preparation was
used as reference material. However, the imm/Bio ratio varied from
3.2 to 6.9 when a urinary preparation (2nd IRP) was used as a con-
comitant standard.
These results confirm the observations recorded in the litera-

ture (8,13,14) in that bioassay and radioimmunoassay potency esti-
mates obtained for pituitary preparations using a pituitary refer-
ence preparation are in agreement, and that bioassay potency esti-
mates of pituitary preparations are lower than radioimmunoassay po-
tency estimates when a urinary reference preparation is used.
Several factors might be responsible for the radioimmunoassay

over estimation of LH content of the pituitary preparations when the
2nd IRP is used as the standard reference material. One is the pos-
sibility that a urinary preparation, such as the 2nd IRP, could have
lost immunoreactive sites in the LH molecule as a result of metabo-
lism attendant to its passage from pituitary to blood to urine, or
0-
'"
00
TABLE 2
BIOLOGIC CHARACTERISTICS OF HUMAN PITUITARY GONADOTROPIN PREPARATIONS
. OAAD Assay (LH)-Relative Potency (95% Confidence Limits) ARtassa (FSH) FSH/LH
Preparat10ns y Ratio
IU/2nd IRP./mg IU LER 907/mg mg LER 907/mg IU/2nd IRP/mg (2nd IRP std)
54.2
LER 907 20 0.37
(31.7 - 98.7)
510 430 7.9

LER 856-1 2.7 0.005
(320 - 830) (280 - 660) (6.7 - 8.7)
1,320 1,110 20.5

LER 1371 1.7 0.0013
(760 2,290) ( 640 - 1 , 910) (19.3 - 20.6)
1,465 1,420 26.2

LER 960 1.1 0.0008
(630 3,460) (740 - 2,770) (23.0 - 28.0)
1,350 1,320 24.4

DEAE-2-IItt (560 3,250) 2.5 0.0019
(680 - 2,550) (21.5 - 24.8)
!'"
t: Augmentation Reaction Assay; ttPrepared by Dr. Parlow for the NPA. "o
C/)
m
~
00
m
G)
"
8
~
o
o-t
TABLE 3 o
."
Z
LH RADIOIMMUNOASSAY BY HUMAN PITUITARY GONADOTROPIN PREPARATIONS .VI
."
C
Relative Potency (95% Confidence Limits) ~
"TI
Preparations
n
No. of No. of IMM/BIO (OAAD) IMM/BIO (OAAD) ~
mg LER 907/mg IU 2nd IRP/mg 5
Assays Assays LER 907 std* 2nd IRP std** z
310 >
z
LER 907 (320 340) 13 5.7 o
~
~
VI
5.5 1,633 C
LER 856-1 9 8 0.70 3.2 ::0
( 4.8 6.3) (1,500 1,780) m
~
m
Z
18.9 6,235 -t
LER 1371 6 (6,000 6 0.92 4.7
(17.9 19.8) 6,560)
28.1 8,526
LER 960 3 (8,230 8,800) 3 1.08 5.8
(27.2 29.1)
29.7 9,349
DEAE-2-II 7 (8,960 9,720) 7 1.22 6.9
(28.1 31.4)
*mg LER 907; **IU 2nd IRP.
i'.)
$
270 E. ROSEMBERG
from the extraction procedure used in the preparation of this stan-

dard. Another explanation has been introduced by Ryan (11) based
on the hypothesis of the formation of hybrid molecules which could
account for the immunologic deficiency of urinary LH. This hypo-
thesis implies that a dissociated biologically inactive but immuno-
logically recognizable form of LH, as well as the biologically ac-
tive assiciated species of LH, is excreted in urine.
With the reservations pointed out by Albert (15) regarding the

use of standards in the bioassay and radioimmunoassay of gonadotro-
pins, it is hoped that an official body such as the World Health
Organization (WHO) will undertake the preparation and distribution
of suitable standards and reagents for radioimmunoassay. At present
it seems very likely that the WHO will consider the distribution of
the NPA reference preparation LER 907. If so, this material will
become the WHO standard for FSH and LH radioimmunoassays and the
values will probably be referred to in terms of the unitage assigned
by the WHO to this preparation.
Finally, the use of one hormone as a standard for another hor-

mone represents an unnecessary, confusing and potentially dangerous
violation of the principle of similarity. Let us refer to the use
of HCG International Standard (IS) for the bioassay and immunoassay
of LH. While some LH bioassay systems, such as the OAAD, may ful-
fill the validity criteria for the bioassay, other biologic systems
may not. Similarly, the use of HCG IS or commercial preparations
of HCG of varying grades of biologically active purity as standard
with or without the use of human pituitary LH preparations or the
2nd IRP as second standards in the LH radioimmunoassay system pre-
sumes complete cross-reactivity between LH and HCG.
Table 4 shows the biologic (OAAD assay) and radioimmunologic

potency of a human pituitary LH preparation LER 856-1 (referred to
in Tables 2 and 3) distributed by the NPA for clinical investiga-
tion, and that of HCG 2nd IS. The activity of LER 907 (shown in
Tables 2 and 3) is included in this table.
The biopotency estimates of LER 856-1 and HCG 2nd IS were in

good agreement both in terms of LER 907 and the 2nd IRP. The im-
munopotency estimates expressed in terms of LER 907 (for both pre-
parations) agreed very closely with those obtained by bioassay us-
ing the same standard; the Immuno/Bio ratio being 0.70 and 1.3, re-
spectively. However, the immunopotency estimates expressed in
terms of the 2nd IRP were much greater than those obtained by bio-
assay; the Immuno/Bio ratio being 3.2 to 4.8, respectively. It is
worth noting that the Immuno/Bio ratio of HCG is very similar to
that observed with other HLH preparations (Table 3).
All assays fulfilled the necessary statistical criteria for

validity. Specifically, all preparations (tested in the radioim-
G)
oz
»
o
o-t
o'"
."
Z
TABLE 4 ~
."
C
BIOLOGIC AND IMMUNOLOGIC POTENCY OF HLH-LER 856-1 t AND HCG 2nd IS ::!!
.."
ri
~
BIO Potency>~ IMMUNO Potency Ratio 6
in terms of in terms of z
Preparation IMM/BIO IMM/BIO »
z
LER 907 std 2nd IRP std o
LER 907 2nd IRP LER 907 2nd IRP ~
m
»
en
C
LER 907 54.2 (lU/mg) 310 (IU/mg) 5.7
'"m~
m
Z
-t
(HLH)
7.9 (mg/mg) 510 (IU/mg) 5.5 (mg/mg) 1633 (IU/mg) 0.70 3.2
LER 856-1
HCG 2nd IS 10.8 (mg/mg) 592 (IU/mg) 14.0 (mg/mg) 2846 (IU/mg) 1.3 4.8
t Distributed by the NPA for clinical investigation; * OAAD assay.
0..:>
......
272 E. ROSEMBERG
munoassay at several dose levels) i.e., 2nd IRP, LER 907, LER 856-1,
and HCG 2nd IS, elicited parallel dose-response curves. The pre-
sence of parallelism in the radioimmunoassay, although a valid cri-
terion of assay, is not alone a sufficient criterion for judging
identity of materials being tested in this case, human pituitary
and urinary LH and HCG.
When adequate quantities of HFSH and HLH of sufficient chemi-

cal purity for use as immunologic reagents, appropriate specific
antisera and suitable standards of acceptable and known stability
will be made available by an official body, the present confusion
in radioimmunoassay values for the levels of FSH and LH in man will
disappear.
ACKNOWLEDGEMENTS
This work was supported by Grant AM-07564, USPHS, National

Institutes of Health, Bethesda, Maryland.
We are indebted to the National Institute of Arthritis and

Metabolic Diseases, NIH, and to the National Pituitary Agency fo.r
the generous supply of purified pituitary hormones and for the sup-
ply of immunologic reagents. We are also indebted to the Endocrin-
ology Study Section, NIH, Bethesda, Maryland, for the supply of NIH-
FSH-Sl and NIH-LH-Sl and to Dr. D. R. Bangham, Department of Bio-
logical Standards, Medical Research Council, Mill Hill, London for
the gift of the 2nd IRP used in this study. The valuable technical
assistance of George Bulat, Lawrence E. Fournier and Margaret M.
Shea is gratefully acknowledged.
REFERENCES
1. Rosemberg, E., Solod, E.A. and Albert, A., J. Clin. Endocr.,

24: 714, 1964.
2. Albert, A., Derner, I., Rosemberg, E. and Lewis, W.B., Endocrin-
ology, ~: 139, 1965.
3. Albert, A., Hanten, C., Rosemberg, E. and Bulat, G., Endocrin-
ology, 77: 588, 1965.
4. Rosembe~, E. and Engel, I., J. Clin. Endocr., 11: 1063, 1961.
5. Loraine, J. A. and Brown, J.V., J. Clin. Endocr., ~: 1180, 1956.
6. Albert, A., Proc. Staff Meet. Mayo Clinic., 11: 341, 1956.
7. Steelman, S.L. and Pohley, F.M., Endocrinology, ~: 604, 1953.
8. Albert, A., Rosemberg, E., Ross, G. T., Paulsen, C. A. and Ryan,
R. J., J. Clin. Endocr., 28: 1214, 1968.
9. Faiman, C. and Ryan, R. J7; Proc. Soc. Exp. Biol. Med., 111:
ll30, 1967.
10. Faiman, C., Ryan, R.J. and Albert, A., J. Clin. Endocr., ~:
1076, 1968.
11. Ryan, R.J., In Immunoassay of Gonadotropins, ed. Diczfalusy, E.,
Bogtrykkerie~Forum, Publish., Copenhagen, Denmark, p 300, 1969.
12. Odell, W.D., Ross, G.T. and Rayford, P.L., J. Clin. Invest.,
46: 248, 1967.
13. Albert, A., In Gonadotropins 1968, ed. Rosemberg, E., Geron-X
Inc., Los Altos, Calif., p. 393, 1968.
14. Odell, W.D., Reichert, L.E. and Swerdloff, R.S., In Gonadotro-
pins 1968, ed. Rosemberg,E., Geron-X Inc. Los Altos, Calif.,
p. 401, 1968.
15. Albert, A., J. Clin. Endocr., 28: 1683, 1968.
DISCUSSION
DONINI: I agree with Dr. Rosemberg thot the use of appropriate stand-
ards in radioimmunoassays is extremely important. When we used a
urinary standard, i.e., the second IRP-HMG, to determine the poten-
cies of the different urinary preparations with different grades of
purity and different FSH:LH ratios, we obtained a very good agree-
ment between biopotency and the immunopotency. The ratio between
the FSH and LH potencies as determined by bioassay and radioimmuno-
assay was close to unity. Our radioimmunoassay system for FSH and
LH is a total urinary system. We use highly purified FSH and anti-
urinary FSH sera; HCG for labelling, anti-HCG sera and the 2nd IRP-
HMG as standards. We thus obtain very good agreement between bio-
potency and immunopotency.
ROSEMBERG: The reference preparation called LER 907 which is dis-

tributed by the NPA in the United States and abroad, is a mixture of
human pituitary FSH and LH, and will probably be distributed by the
WHO if the collaborative study will show that this material is suit-
able as a reference preparation. Hence, in the future, we may have
this material labelled with a specific unitage assigned to it by the
WHO. In the United States, we have used this material for some time
and it is recommended to express results of radioimmunoassays in terms
of weight of this reference preparation. However, in the future, if
the WHO will decide on its distribution, things will be changed. I
do hope and I am very hopeful that some continuity regarding unitage
will prevail.
We have studied the biological and immunological LH activity of uri-

nary gonadotropin preparations obtained from the urine of postmeno-
pausal women, normal men and eunuchs.
These urinary extracts obtained from these three sources were kindly
supplied by Dr. Ao Albert, Mayo Clinic, Rochester, Minnesota. The
urinary extracts were subjected to the purification steps described
by Dr. Albert, in that fraction A, which represented the absorption
of HPG on kaolin and subsequent dilution with ammonium, was further
274 E. ROSEMBERG
purified to fraction B, utilizing 10% ammonium acetate and 70% eth-

anol; from fraction B a more purified fraction, fraction C was ob-
tained by removal of impurities by adsorption of fraction B with
diethylaminoethylcellulose (DEAEC).
In our studies we only analyzed the LH activity of these prepara-

tions utilizing bioassay (BIO) and radioimmunoassay (RIA) systems.
The bioassays used were the Ventral Prostate Weight (VPW) and Ovar-
ian Ascorbic Acid Depletion (OAAD) assays. The purified human pitu-
itary LH was used for labelling with 1-131 and rabbit anti HCG serum
was used as the first antibody. Goat anti-rabbit gamma globulin was
used as the second antibody to separate free from bound labelled hor-
mone.
Purified pituitary LH and anti HCG serum were kindly provided by the
National Institute of Arthritis and Metabolic Diseases through the
National Pituitary Agency and the goat anti-rabbit gamma globulin
was kindly supplied by Dr. Alice Reiss, Ortho Research Foundation.
The Second International Reference Preparation for Human Menopausal
Gonadotropins (2nd IRP) was the reference standard. The relative
potencies (RP) of fractions A, B, and C derived from the three uri-
nary sources showed good agreement. Fraction A obtained from post-
menopausal and eunuch urine showed good agreement between BIO and
RIA values; this was not observed with fraction A from male urine.
Fractions Band C obtained from these three sources showed higher
activity in the RIA system as compared to that obtained with the use
of BIO systems.
The mean purification corresponding to fractions A, B, and 0, taking

fraction A as 1 in the BIO system was similar to that originally
published by Dr. Albert. The mean purification taking fraction A
as 1 in the immunoassay system showed a similar trend, although. of
a different magnitude than that obtained with bioassay. The oV~
ratio was quite similar for the three fractions from PM, Mana
E urine; the IMM/BIO ratio increased with increased purification of
these fractions.
It has been reported that the IMM/BIO activity of urinary fractions,

specifically FSH fractions of urinary extracts, does not differ from
unitage in a RIA system which utilizes purified urinary FSH for la-
belling and a rabbit anti-urinary FSH antiserum.
In this study, which refers to the IMM/BIO LH activity of urinary

fractions, the RIA system involved the use of a human pituitary LH
for labelling and anti HCG serum. It is possible that the use of
this system may have resulted in the differences between IMM/BIO
potency of the purified fractions. However, this study suggests
that this immunoassay system could be utilized to check the degree
of purification obtained when urinary fractions are subjected to var-
ious purification procedures.
DICZFALUSY: Chorionic gonadotropin preparations do possess FSH-like

activity; this FSH-like activity can be removed, for instance, by
Sephadex chromatography. However, there is another activity remain-
ing which we could not remove thus far. If one immunizes rabbits
with HCG preparations, one obtains very potent anti-FSH sera. This
is due to a material that we call ANF, anti-gonadotropin neutralizing
factor, and Dr. Petrusz and I have devised a bioassay for measuring
it. The problem is that there are large quantities of biologically
inactive, immunologically active materials in highly purified cho-
rionic gonadotropin preparations, which will induce antibodies.
These antibodies will neutralize FSH activity of pituitary as well
as urinary preparations. Still, in most laboratories, including
ours, such HCG preparations are used to obtain antisera which are
later on considered to be specific to LH and HCG. By fractionation
of chorionic gonadotropin on a Sephadex column one finds low molec-
ular weight material which possesses little biological activity but
which still contains a high amount of immunological activity. One
obtains an entirely different ratio here than in a high molecular
weight material. Therefore, I think that for the time being we
should consider the distribution of different standards for bioassays
and immunoassays.
ROSEMBERG: I would favor the preparation and distribution of sev-

eral standards.
ROSS: In human chorionic gonadotropin there is good evidence that

antigenic determinants exist that are similar to antigenic determi-
nants in follicle stimulating hormone extracted from either human
urine or human pituitary tissue. During purification of human cho-
rionic gonadotropin, one removes substances with the apparent bio-
logic activity of follicle stimulating hormone without completely
removing substances with immunologic activity. This is not partic-
ularly surprising if one considers that both FSH and HCG are prob-
ably composed of subunits, all of which are required for biologic
activity of either hormone, but which in biologically active mate-
rial might still be capable of stimulating antibody formation. The
important feature with respect to the nature of antibodies generated
against these substances in my opinion relates to the differences
in what the individual animal sees in the way of antigenic determi-
nants, in the immunogen. I would like to ask Dr. Diczfalusy if all
of the rabbits that are challenged with this particular preparation
of chorionic gonadotropin make equally specific and equally potent
antisera with respect to the capacity to neutralize the biologic ac-
tivity of follicle stimulating hormone.
DICZFALUSY: One never obtains equally potent antisera when immu-

nizing 30 rabbits. However, I would submit that in every rabbit we
immunized thus far it was possible to detect measurable quantities
of anti-FSH. I would say that with one exception we have not seen
any rabbit serum thus far which contained less than 50 to 100 anti-
276 E. ROSEMBERG
FSH units, and I wonder whether Dr. Lunenfeld would like to add fur-
ther information to this point.
LUNENFELD: I wish to confirm Dr. Diczfalusy's statement. We have

shown that all HeG preparations which we have utilized for immuni-
zation, also produced antibodies to FSH.
PREPARATIONS AND BIOLOGICAL CHARACTERISTICS OF HMG PREPARATIONS
USED CLINICALLY
Piero Donini
Research Laboratories, Istituto Farmacologico
Serono, Rome, Italy
After the divergent and inconclusive effects obtained in

treatment of female sterility with gonadotropin extracts obtained
from animal pituitaries or from pregnant mare serum (PMS), our
group, which was involved in the preparation of chorionic gonado-
tropin (HCG), started in 1949 to study the possibility of the pre-
paration of human menopausal gonadotropin (HMG) suitable for treat-
ment in humans. In the same year we published our first paper (1)
on the preparation of HMG suitable for ctinical investigation and
Pergonal became available to physicians in Italy in 1950.
Following the same technique we also prepared Pergonal-23

which, in 1964, became the 2nd IRP ~MG. Each ampoule of this
reference preparation contains 5 mg of HMG, equivalent to 40 IU of
FSH and 40 IU of LH.
Another possible source of human gonadotropins are human pitu-

itaries obtained at autopsy. Although in some countries, e.g.
U.S.A., England, and Sweden, collection banks for human pituitaries
have been organized, this source does not permit the preparation of
human gonadotropins on a large enough scale for general clinical
use.
With respect to purification studies, Donini and Marchetti (2)

reported the first attempt of chemical characterization of HMG in
1952.
The clinical effect of this HMG preparation in women with pri-

mary amenorrhea was reported by Borth et al. (3) and the results
confirmed by Rosemberg et al. (4). Due to the relative low specific
activity of the original preparations used in these clinical trials,
277
278 P. DONINI
further purification was attempted. Donini et ale (5) increased

the relative potency of the HMG preparation so that the 75 IU of
FSH and 75 IU of LH contained in one ampoule of Pergonal-500 (trade
mark of purer HMG preparation) were equivalent to about 1-2 mg of
HMG.
In diagram 1, reported by Lunenfeld (6), the extraction and puri-

fication procedure of our HMG preparation is illustrated. The
extraction of gonadotropins from menopausal urine was done accord-
ing to the method of Bradbury et ale (7) and the crude material
obtained was then purified following the method of Albert et ale
(8). This is based on the extraction of crude HMG with 10% ammo-
nium acetate in 70% ethanol followed by absorption of the impurities
on DEAE-C at pH 7. At this point of purification, we used Permu-
tit column chromatography which, I believe, is the most important
step for purification. With this step one can obtain an HMG pre-
paration with a biological potency about 130-300 times higher than
crude gonadotropic extract obtained by the first step or the kaolin-
acetone precipitate.
It should be emphasized that by this single step of purifica-

tion it was possible not only to increase the biological potency
but to completely eliminate the phosphate salts contained in the
buffer solution used for the step on DEAE-C.
By following the purification procedure described above, im-

proved human menopausal gonadotropins could be prepared for therapeu-
tic purposes; the LD50 in mice was very low, the specific FSH and
LH potencies varied from batch to batch but the range was 40-90 IU
per mg and the FSH:LH ratio was about 1:1. In some clinical trials
the contents of 52 ampoules of Pergonal-500 were dissolved in 5 ml
of saline and injected in one woman without relevant side-effects.
Initial clinical trials with Pergonal-500 were carried out and

reported by Lunenfeld et ale (9,10). In one patient with primary
amenorrhea, pregnancy with subsequent delivery of a normal infant
was achieved.
It is well known that the gonadotropic extracts prepared from

human pituitaries or menopausal urine contain two different hor-
mones with different biological activities, i.e. follicle stimulat-
ing hormone (FSH), and luteinizing hormone (LH or ICSH). For a
better understanding of the mechanism of these two gonadotropins,
it was necessary to purify further and to separate these two bio-
logical activities. Due to the short time available for this pre-
sentation, I cannot present the details of studies carried out by
several workers over the last ten years. But it would be of inter-
est to examine the table previously published (11) which lists the
biological potencies of urinary preparations obtained by different
authors up to 1966 and expressed in terms of IU of the 2nd IRP-HMG.
Diagram 1. EXTRACTION AND PURIFICATION OF HUMAN MENOPAUSAL GONADO-

TROPINS.
Step 1: KAOLIN EXTRACTION
The kaolin is activated by washing with lM-HC1, allowed to settle, the
supernatant decanted and the kaolin subjected to repeated washings
with water. Human Menopausal Urine adjusted to pH 4.5 with glacial
AcH, then shaken with celite + 20% activated kaolin suspension.
Allowed to settle overnight at room temp.

I
I
sediment
supernatant discarded
I centrifugation
I I
supernatant discarded precipitate washed twice with tap
water adjusted to pH 4.5 with gla-
cial AcH
elution twice with lM-NH40H eluate (pH 11-11.3) adjusted to pH 8.5

with glacial acetic acid
Centrifugation
I
I I
supernatant acidified to pH 5-5.5 precipitate discarded
with glacial acetic acid
I
2 volumes acetone added with stirring and left overnight
I
Precipitate washed with acetone, then
supernatant discarded
with 95% EtOH, then with ether (HMG-K)
Step 2: CONDITIONING FOR DEAE-CELLULOSE TREATMENT

(HMG-K)
I
extracted twice by stirring with 70% EtOH + 10% ammonium acetate and
filtered
I
ppt. discarded clear extract
2 vols. of abs. EtOH + 10% ammonium acetate added allowed to settle
overnight at 4_5 0 C, then ppt. centrifuged.
I
j I
ppt. washed with 95% EtOH, then with ether and supernatant
finally dried (HMG-PD) discarded
280 P. DONINI
Step 3: DEAE-CELLULOSE TREATMENT
The DEAE-cellulose is previously washed with 0.5 N-HC1, then with

0.5 N-NaOH and finally with 0.05 M-phosphate buffer (pH 7)
HMGiPD
dissolved in 0.05 M-phosphate buffer(pH 7) shaken with DEAE-cellulose,
filtered through Buchner funnel, then ppt. washed twice with 0.05 M-
phosphate buffer (pH 7).
filtrate acidified to pH 5.4 with glacial ppt. discarded

acetic acid then cooled to 4_5 0 C (HMG-D)
Step 4: CHROMATOGRAPHY ON PERMUTIT COLUMN
The permutit column is previously washed with IM-NH 40H, then with
1M-acetic acid, and finally equilibrated with 0.05 M-sod. acetate
buffer (pH 5.4) I
HMG-D
clear liquid poured into column in
a cold room at 4_5 0 C
Column washed with 0.05 M acetate buffer outflow discarded

(pH 5.4) until optical density of outflow
is near 0 at 280 mil-
elute adsorbed proteins with 40% EtOH + discard outflow

10% ammonium acetate
I
add to eluate 95% EtOH (chilled at 2_4 0 C) with stirring, allow to
settle overnight at 2_40 C, centrifuge ppt.
ppt. washed with abs. EtOH, then

supernatant
with ether and finally dried.
discarded
(HMG-P)
(From: Excerpta Medica Proceedings of Second International Congress

of Endocrinology, London, 1964. Excerpta Medica International Congo
Series, No. 83, p 814).
Later the urinary FSH and LH hormones were purified further.

With respect to FSH, homogeneous preparations were obtained by Roos
(12) and by our group (13,14). Both of us had the opportunity to
make the physical chemical characterization and, in general, the
findings obtained in the different laboratories agreed except for
the sialic acid content which was much higher in our FSH prepara-
tion compared to that of Roos.
Diagram 2 depicts the different steps of purification which

allowed us to prepare our purest and most active urinary follicle
stimulating and luteinizing hormone preparations. In fact the FSH
preparation contained 1255 IU-FSH as determined biologically ac-
cording to the method of Steelman and Pohley, and only 3.2 IU-LH
per mg as determined by radioimmunoassay according to the method
of Donini et al. (16). Although the purest LH preparation was not
homogeneous, as demonstrated by the different criteria commonly
used for the physical-chemical characterization of proteins, it
showed one of the highest biological and immunological activities
reported, i.e. 982 IU and 1166 IU-LH, respectively. No FSH activ-
ity could be detected in this preparation at a dose level of 1260
I-Lg •
These highly purified preparations of urinary FSH and LH were
important not only to clarify the physical-chemical properties but
also for other reasons. Other workers as well as our group have
used these purified FSH preparations for labelling in radioimmuno-
assay systems and specificity studies of urinary anti FSH sera
have also been carried out (17).
Furthermore, these same highly purified FSH preparations were

used clinically by Mancini et al. who studied the effect of urinary
FSH on the testes of hypophysectomized patients. The same group
also studied the effects of highly purified LH and FSH and LH com-
binationsat different ratios on the testis of the same patients.
Their results will be reported elsewhere.
Bompiani and his group utilized the same highly purified FSH
and LH preparations to study the effect of these hormones in the
ovaries of one hypophysectomized patient. These results will also
be published elsewhere.
With respect to the HMG preparations, which are used clinically

in the therapy of female and male sterility, another point to be
considered is their FSH:LH ratio. As reported herein, we can now
prepare the HMG preparations with different FSH:LH ratios, but un-
fortunately we do not have good evidence as to which ratio can be
considered the best in order to have the optimal effectiveness for
induction of ovulation and pregnancy and to avoid hyperstimulation
and multiple ovulation.
282 P. DONINI
Diagram 2. PURIFICATION OF FOLLICLE-STIMULATING AND LUTEINIZING

HORMONES FROM HUMAN MENOPAUSAL GONADOTROPIN. THE BIO-
LOGIC ACTIVITIES ARE EXPRESSED IN IU (2nd IRP-HMG)
HMG (Pergonal, 82.7 IU of FSH and 58.3 IU of LH per mg) 5.46 g

Batchwise separation of FSH and LH on DEAE-C
I
I I
FSH fraction (2.88 g; 133 IU-FSH/ LH fraction (1.74 g; 86 IU-LH/
mg) DEAE-C column chromatography mg) BatcQwise chromatography
on CMC-70
Eluted by borate phosphate Eluted by 0.005 M phosphate

buffer pH 8.0 +0.1 M NaCl buffer pH 6 + 0.5 NaCl (1.03 g;
(828 mg; 351 IU-FSH/mg) Gel 102 IU-LH/mg) Gel filtration on
filtration on Sephadex G-lOO Sephadex G-lOO
Impurities
Other fractions with with lower
lower activity activity
S etric part of the 2nd First part of ascending limb of

protein peak (323.7 mg ; 692 elution curve (240 mgj 263 IU/
IU-FSH/mg). Preparative disc mg) DEAE-Sephadex chromatography
electrophoresis
Other fractions with Impurities

with lower
lower activity
activity
Part of 2nd protein peak. Eluted by 0.002 M phosphate

Highly purified FSH 104.3 mg; buffer pH 6.6 + 0.01 M NaCl
1255 IU-FSH/mg, 3.2 IU-LH/mg (36.8 mg; 807 IU/mg) Gel
filtration on Sephadex G-200
Other fractions with
Impurities
lower activity
with lower
activity
Symmetric and sharp part of

the first peak. Highly
purified LH 20.4 mg; 1166 IU/mg.
Fractions with
lower activity
(From: Donini et aI, In: Gonadotropins 1968, Rosemberg, E. ed., Los
Altos, Calif. p 40, 1~8).
Rosemberg and Nwe (18), Jones et ale (19), Taymor (20), and
Bettendorf et ale (21) have studied this problem. In general we
agree with Rosemberg's conclusion that, "Because the number of
courses of medication is relatively small, investigation on the im-
portance of FSH:LH ratios contained in HMG preparations used clin-
ically should be continued." Nevertheless, it seems that if the
FSH:LH ratio is too high, the effectiveness of the preparation is
lower than when the FSH:LH ratio is about 1:1 with regard to the
follicle ripening and estrogen excretion prior to ovulation. In
other words, the total amount of FSH activity needed for follicular
ripening is higher when the HMG preparation used has a high FSH:LH
ratio (20). On the contrary, as reported by Rosemberg (18), "with
preparations containing a hfgher proportion of LH activity relative
to FSH, small increases in dosage were shown to cause an excessive
pregnanediol excretion," and according to Bettendorf (21), "it is
of interest that a higher incidence of ovarian enlargement was
found in the group of patients who received the low FSH:LH ratio
preparation than those patients who received a preparation having
a high FSH:LH ratio."
In the last few years, our knowledge of the ovulatory mechan-

ism in normal women is much better than in the past. We know that
through the positive or negative feedback due to the action of ovar-
ian steroids on the hypothalamus, the FSH and LH releasing factors
regulate the release of follicle-stimulating and luteinizing hor-
mones from the pituitary gland in such a way that these gonadotro-
pins act in due time and in due amount on the ovaries. In the nor-
mal menstrual cycle, we also know the pattern of FSH and LH levels
as determined in the urine or serum by bioassay or radioimmunoassay.
Up to now, as reported in several papers, induction of ovula-

tion and pregnancy was tried in anovulatory women using HMG prepa-
rations having different FSH:LH ratios, i.e.: very high, 1:1, and
very low. In these studies the same preparation and the same FSH:
LH ratio were used during the entire course of a given treatment.
Another problem which should be kept in mind is that after

stimulation of the follicle by HMG, to induce ovulation and to pro-
duce a functional corpus lutem, HeG has been injected instead of LH
because LH was not available for extensive clinical trial. We know
(22) that the disappearance rate from the blood is greater for LH
than HeG. This finding could explain the ovarian hyperstimulation
and the multiple pregnancy when HeG was injected. It is worth no-
ticing that, according to the data obtained by radioimmunoassay,
the ovulatory peak of LH is only about 24 hours long.
I believe that a new therapeutic approach with human gonado-

tropins could be tried in anovulatory women who are sterile due to
a relative or absolute lack of pituitary gonadotropins. In the
last two years several authors have reported that in the first part
284 P. DONINI
of the follicular phase of the menstrual cycle, most of the normal

women studied showed in their serum and urine an FSH:LH ratio of
about 2:1. Then in the second part of the follicular phase and be-
fore the ovulatory peak, the FSH:LH ratio was about 0.3:1. Later
on in the cycle, the FSH:LH ratio is again about 1:1.
It would be worthwhile to evaluate the response to HMG prepa-

rations using different FSH:LH ratios during a course of treatment.
It would also be worthwhile to use urinary or pituitary LH instead
of HCG to induce the ovulation. One should keep in mind that the
responsiveness to human gonadotropins varies from one patient to
another and thus follicular stimulation should be carefully con-
trolled by the rapid determination of estrogens by the method of
Brown (23) or by other parameters like the fern test, karyopyknotic
index, etc.
The same problems of FSH:LH ratio, dosage, and length of ther-

apy, occur when male infertility has to be treated with human gona-
dotropins. Unfortunately, we do not have enough information on the
physiology and the pathology of the human testis and I hope that
this meeting will increase our knowledge on these problems.
In conclusion, using the methods of extraction and purifica-

tion which have just been reported on, it is possible to have avail-
able a large amount of HMG with different grades of purity and with
different FSH:LH ratios.
REFERENCES
1. Donini, P. and Montezemolo, R., Rass. Clin. Ter., 48: 143, 1949.
2. Donini, P. and Marchetti, E., 11 Farmaco., 2: 4l8,-r952.
3. Borth, R., Lunenfeld, B., and Menzi, A., In Human Pituitary
Gonadotropins. A Workshop Conference, ed:-Albert, A., C. Thomas,
Springfield, Ill., p 255, 1961.
4. Rosemberg, E., Coleman, J., Demany, M. and Garcia, C.R., J.
C1in. Endocr., ~: 181, 1963.
5. Donini, P., Puzzuo1i, D., and Montezemo10, R., Acta Endocr.,
(Kobenhavn), 45: 321, 1964.
6. Lunenfeld, B.:-Proc. 2nd Intern. Congress of Endocr. Publ.
Excerpta Medica International Congress Series 83, p 814, 1964.
7. Bradbury, J.T., Brown, E.S., and Brown, W.E., Proc. Soc. Exp.
BioI. Med., 11: 228, 1949.
8. Albert, A., Kobi, J., Lei ferman, J., and Derner, J., J. Clin.
Endocr., 21: 1, 1961.
9. Lunenfeld:-B., Sulimovici, S., Rabau, E., and Eshko1, A.,
C.R. Soc. Franc. Gynecol. 11: 346, 1962.
10. Lunenfeld, B., Sulimovici, S., and Rabau, E., Proc. of Tel-
Hashomer Hospital. 1: 25, 1962.
11. Donini, P., Puzzuoli, D., D'Alessio, I., Lunenfeld, B., Eshkol,
A., and Parlow, A.F., Acta Endocr. (Kobenhavn) 52: 169, 1966.
12. Roos, P., Acta Endocr. (Kobenhavn) 59: Supple 131, 5, 1968.
13. Donini, P., Puzzuoli, D., D'Alessio, I., Bergesi, G., and
Donini, S. III Gonadotropins 1968, ed. Rosemberg, E.,
Geron-X, Los Altos, Calif., p. 37, 1968.
14. Donini, P., Puzzuoli, D., D'Alessio, I., Bergesi, G., and
Donini, S., In Proceedings of meeting on 'Chemistry of
Gonadotropins', Birmingham, 1969, in press.
15. Steelman, S.L. and Pohley, F. M., Endocrinology, 53: 640,
1953. -
16. Donini, S., D'Alessio, I., and Donini, P., In Gonadotropins
1968, ed. Rosemberg, E., Geron-X, Los Altos, Calif., p. 263,
1968.
17. Donini, S. and Donini, P., In 1st Karolinska Symposium.
Immunoassay of Gonadotropins, ed. Diczfalusy, E.
Bogtrykkeriet Forum, Copenhagen, p. 57, 1968.
18. Rosemberg, E. and Nwe, ToT., Fertil. Sterile 19: 197, 1968.
19. Jones, G.S., De Moraes, R., Johanson, A.J., Ruiti, S. and
Blizzard, R.M., Fertil. Steril., 20: 14, 1969.
20. Taymor, M.L. In Progress in Infertility, ed. Behrman, S.J.
and Kistner, R.W., p. 393, 1968.
21. Bettendorf, G., Breckwoldt, M., and Neale, C., In
Gonadotropins 1968, ed. Rosemberg, E., Geron-X, Los Altos,
Calif., p. 453, 1968.
22. Yen, S.S.C., Llerena, 0., Little, B., and Pearson, O.H.,
J. Clin. Endocr., 28: 1763, 1968.
23. Brown, J.B., MacLeod, S.C., Macnaughtan, C., Smith, M.A., and
Smith, B., J. Endocr., 42:, 5, 1968.
DISCUSSION
JUTISZ: Dr. Donini, what physiochemical criteria did you use to

show that your urinary FSH is a homogeneous protein? Can you also
tell us about the contamination of this preparation with LH and
other hormones? Finally, do you have any information about the
amino acid and carbohydrate composition of this preparation, and is
this the same as in human pituitary FSH? In other words, there is
always a possibility of an alteration of the original molecule. I
think that Mori showed some time ago that human pituitary FSH cross-
reacts only partially with urinary FSH.
DONINI: The criteria used to demonstrate the homogeneity of our FSH

preparation were the commonly used criteria, that is, 1) the pattern
of sedimentation in the ultracentrifuge, where we observed only one
peak; 2) disc electrophoresis on polyacrylamide gel. The disc elec-
trophoresis pattern of the different fractions was as follows: in
the less purified fraction (fraction A) many bands were seen; frac-
tion B showed a smaller number of bands; with highly purified FSH
286 P. DONINI
only one band was observed. Also, with regard to the carbohydrate
moiety we have demonstrated that urinary FSH is very similar to pi-
tuitary FSH. For example, we found that in our preparation one of
the most important carbohydrate components, the sialic acid, was
almost the same as that found in pituitary FSH; 8.4% of sialic acid
was found in urinary preparations. About the same value was found
by Dr. Reichert and by other investigators in pituitary FSH prepar-
ations. Some differences between urinary and pituitary FSH exists.
The immunological behavior of some pituitary and urinary FSH pre -
parations are different. We have demonstrated that our radioimmuno-
assay system with reagents of urinary origin is not suitable for
radioimmunoassay of FSH in plasma. This means that at least some
immunological differences exist between urinary and pituitary FSH.
The activity of the highly purified FSH was determined biologically
by the method of Steelman and Pohley. The LH contamination was
determined by radioimmunoassay, because it was almost impossible,
or difficult to waste so many milligrams of this preparation, to
test its LH activity by biological methods. The FSH potency of this
preparation was 1250 IU per mg and the LH contamination was 3.2 IU
per mg. The differences between the amino acid composition of uri-
nary and pituitary FSH are very small.
ROSEMBERG: Dr. Donini pointed out that in the preparation of HMG

used clinically, the consideration of FSH:LH ratios is of importance.
Dr. Diczfalusy, as well as ourselves, some years ago indicated that
the FSH:LH ratio contained in HMG preparations was of importance in
the treatment of female infertility. With the availability of puri-
fied material of urinary or pituitary origin it will be much easier
to study this problem experimentally, mixing both hormones at a cer-
tain desired FSH:LH ratio. We have used purified HFSH and HLH at
ratios similar to those described by Dr. Ross in his study of the
blood levels of FSH and LH during the normal menstrual cycle. We
have treated eight patients during 13 or 14 courses of medication.
Most patients ovulated as judged by indirect indices of ovulation.
Three of the patients became pregnant. There were no untoward ef-
fects. These were patients with intact pituitaries. We should test
these combinations further in hypophysectomized patients. I think
we are approaching the time when with the availability of purified ma-
terials,many of the previously unanswered questions will be clarified.
LUNENFELD: I would like to add that we now currently use urinary

testosterone to monitor the amount of HCG to be given in addition
to Pergonal in the treatment of males. We try to keep the testos-
terone at a desired level. I think this may be one of the ways to
monitor treatment.
DONINI: Dr. Lunenfeld, how many international units of HCG or LH

do you think are useful for the treatment of male infertility?
LUNENFELD: I do not think that we can make any general statement

here. Before we start treating a patient we always evaluate Leydig
cell function by dynamic tests. I was very impressed with Dr.
Steinberger's presentation because it demonstrated that the testic-
ular biopsy by itself might sometimes be misleading. We rely to a
certain extent on urinary testosterone and estrogen as indicative
of Leydig cell response to HCG. If, in a certain patient, there
is very little or no response to HCG we treat this patient for 14
days with HCG alone and then re-estimate testosterone. If testos-
terone is increased, Pergonal therapy is initiated with the same
amount of HCG which would keep the testosterone value at the desired
level. This amount will vary from patient to patient.
PAULSEN: Dr. Lunenfeld, I assume you are referring to the treat-

ment of patients with hypogonadotropic eunuchoidism.
PLASMA FSH AND LH MEASURED BY RADIOIMMUNOASSAY IN NORMAL AND
PATHOLOGIC CONDITIONS IN MEN
Griff T. Ross
Endocrinology Branch, National Cancer Institute,

National Institutes, of Health, Bethesda, Maryland,
20014
Rational therapy of hypogonadal states in men depends in part

upon determining whether the problem results from failure of the
pituitary to secrete gonadotropins or from failure of the testes to
respond. The use of double antibody radioimmunoassays for plasma
follicle stimulating hormone (FSH) and luteinizing hormone (LH) in
attempts to distinguish these pathophysiologic alternatives among
patients seen on the clinical service of the Endocrinology Branch,
National Cancer Institute will be reviewed here.
MATERIALS AND METHODS
1. Assays
Radioimmunoassays for FSH and LH activity in plasma were per-

formed by methods which have been described in detail (1,2,3). All
samples from an individual patient were assayed in the same assay.
Not all samples from all subjects were included in a single assay,
since this was not feasible for this study.
In both FSH and LH assays, dose response relationships of the

Second International Reference Preparation of Human Menopausal Gona-
dotropin (IRP 2 HMG) have proved to be satisfactory for dose inter-
polation of concentrations of these hormones in samples of both
plasma and urine. The immunoreactivity of 1 milli International
Unit (mIU) of biological activity of this reference preparation is
equivalent to that of 33 and 4 nanograms of the LER 907 reference
preparation (Distributed by the National Pituitary Agency, USA) in
these assays for FSH and LH respectively. In this discussion all
concentrations will be expressed in milli International Units of
IRP 2 HMG per milliliter (mIU/ml) of plasma.
289
290 G. T. ROSS
The methods of Rodbard and associates (4) and of Weaver (5)

were used for dose interpolation in most instances. Statistical
quality control was maintained by the method of Rodbard et ale (6).
In the assay of LH when 300 mi. plasma samples were used, a minimal
detectable quantity was usually around 6 mIU per mi. Similarly,
for FSH when 200 mi. plasma samples were used, the quantity was
around 4 mIU per mi. In both assays potency estimates were less
precise when concentrations were near either the upper or lower
limits than when near the middle of the proportionate range of the
dose response curve (6,7).
2. Clomiphene Tests
Clomiphene citrate was given in doses of 200 mg per day orally

for six days. This regimen usually elicited significant increases
in concentrations of immunoreactive FSH and LH when samples taken
prior to and during administration of drug to normal men were com-
pared (8-11). When coupled with simultaneous measurements of plas-
ma testosterone concentrations, some evidence of Leydig cell re-
sponse to LH was obtained (8,10).
3. Patients
Six eugonadal men, ages 19 to 57 years, served as controls.

Eighteen men with hypogonadism were studied. Seven of these hypo-
gonadal men had a syndrome characterized by retarded sexual and so-
matic development associated with skeletal anomalies and hyposmia,
but with 46, XY karyotypes in cells cultured from peripheral blood
(10). Seven of them were bypogonadal due to either neoplastic or
idiopathic pituitary disease. The four anorchic patients included
two who had been orchiectomized for treatment of carcinoma of the
prostate and two with the "vanishing testis" syndrome (12).
RESULTS
Twenty-eight plasma samples taken on separate days, usually

consecutively, from the six eugonadal men were assayed for FSH and
LH. The mean, standard deviation of the mean and range of values
obtained in each subject are summarized in Table 1. Plasma FSH con-
centrations varied from two to four-fold, within subjects between
days. Values ranged from 4.7 to 27.3 mIU per ml when all determina-
tions for all subjects were considered. Plasma LH values were some-
what less variable than plasma FSH values in this small series of
samples although, on occasion, concentrations of LH in plasma sam-
ples randomly collected from other eugonadal men fell below the
limits of detection in a given assay.
Means of FSH and LH concentrations in all specimens from the

group of eugonadal men were compared with means of concentrations
in randomly collected samples from the three groups of hypogonadal
PLASMA LH PLASMA FSH

50~--~==~~~~----.----r----.---~~
40
6 2 4 6
CLOMIPHENE TREATMENT (Days)
Fig. 1 Plasma FSH and LH concentrations prior to and during admin-
istration of Clomiphene to men with hypogonadism and hyposmia. Sha-
ded areas depict range of responses in eugonadal men on the same re-
gimen. (From: Bardin, C.W. et al., J. Clin. Invest., 48 2049,1969.)
PLASMA LH PLASMA FSH

50 r----r----.---~----_r----._--_.----.__,
E
.......
~
246 024 6
CLOMIPHENE TREATMENT (Days)
Fig. 2 Plasma FSH and LH concentrations prior to and during admin-

istration of Clomiphene to hypogonadal men with hypopituitarism.
The shaded areas represent the range of values in eugonadal men on
the same regimen. (From: Bardin, C.W. et al., J. Clin. Invest., ~:
2049, 1969.)
292 G. T. ROSS
FSH* LH*
Number
Subject
Samples Mean ± SO Range Mean ±SO Range
A 4 6.9 ± 2.8 4.7 - 10.9 14.0 ± 4.0 11.1-19.9

B 5 16.7 ± 6.2 11.9 - 27.3 11.1 ± 1.0 10.4 - 12.8
C 5 9.9 ± 2.2 7.4 - 12.9 13.1 ±3.0 9.9 - 17.6
0 4 10.0 ± 5.0 6.5 - 17.4 15.2 ± 2.1 12.5 - 17.5
E 5 15.3 ± 6.6 5.5 - 22.8 13 ± 1.3 11.7 - 14.5
F 5 12.3 ± 3.3 8.1 -16.6 10.7 ± 1.3 9.5 - 12.9
* milli International Units per milliliter of plasma (radioimmunoassay. I RP 2 HMG)
Table 1 Mean, standard deviation and range of FSH and LH concen-

trations in plasma samples from six eugonadal men.
men. Results are shown in Table 2. Comparisons showed that mean

concentrations for FSH and LH in each group of hypogonadal men were
significantly different from mean concentrations for the group of
eugonadal men. Thus, comparison of means of concentrations of FSH
and LH in single random plasma samples distinguished the groups of
hypogonadal men from the group of eugonadal men.
Therapeutic decisions must be made for individuals rather than

for groups, however. Ranges of values in the six eugonadal men
were compared with ranges of values in the three groups of hypogo-
nadal men (Table 3). Concentrations in single samples discriminated
between the six eugonadal men and the four anorchic hypogonadal men
in every instance since the ranges of values obtained showed no
overlap. In contrast, measurement of either FSH or LH or of both
in single samples did not discriminate between individual eugonadal
men and individual hypogonadal men, either hypopituitary or hypos-
mic. Thus, determinations of concentrations of FSH and LH in a
single random plasma sample appeared to be of limited diagnostic
value for any except anorchic subjects.
The possibility that assays of multiple specimens taken from

a single subject might be diagnostic was considered. To test this
possibility, values for LH were determined in four samples taken
daily from two subjects with hypogonadism and hyposmia and the means
calculated (Table 4). Tests for significance of differences were
made using means for similar numbers of samples from the six indi-
vidual eugonadal men(Table 1). Results were significant (p <.05)
in only two of six contrasts for subject K, but in all six contrasts
for subject L. Thus, means of determinations of plasma concentra-
tions of LH in four daily specimens also failed to be diagnostic
when hypogonadism resulted from failure of adequate pituitary gona-
dotropin secretion.
Number
Group FSH* LH*
Subjects
Eugonadal men 6 12.1 ± 1.0 12.7 ± 0.5
Hypogonadal men
A. Hypopitu itary 7 5.7 ± 0.6 t 9.4 ± 1.1t
B. Hyposmic 7 5.3 ± 0.5 t 9.7 ± 1.2t
C. Anorchic 4 93.1 ± 6.6 t 54.1 ± 3.6 t
* milli International Units per milliliter plasma (radioimmuno-
assay, I RP 2 HMG).
t p < 0.05 when compared with eugonadal controls.
Table 2 Mean and standard error of mean plasma FSH and LH concen-
trations in eugonadal and hypogonadal men.
Number
Group FSH* LH*
Subjects
Eugonadal men 6 4.7- 27 9.5 - 19.9
Hypogonadal men
A. Hypopituitary 7 4- 8 <6-14
B. Hyposmic 7 4- 7 <6 -14
C. Anorchic 4 66 - 120 37 -86
* milli International Units per milliliter plasma (radioimmuno-

assay, I RP 2 HMG)
Table 3 Range of FSH and LH concentrations in plasma samples from
eugonadal and hypogonadal men.
Subject Range Mean S.E.M.
K 8.8 - 11.2 10.0 0.5
L 6.4- 8.0 6.8 0.4

Table 4 Range, mean, and standard error of the mean LH concentra-
tions in four plasma samples from each of two hypogonadal
men with hyposmia. Units are milli International Units
per ml of plasma (radioimmunoassay, IRP 2 HMG).
294 G. T. ROSS
When other tests had failed to be diagnostic, results of clom-

iphene tests were examined since plasma concentrations of FSH and
LH always increased when eugonadal men were given clomiphene in
doses of 200 mg per day for six days (8,9.10). Results of similar
tests performed in the seven hypogonadal men with hypopituitarism
and the seven hypogonadal men with hyposmia are shown in Figs. 1 and
2 respectively. Uniformly, these hypogonadal men failed to increase
plasma FSH and LH concentrations in response to clomiphene, clearly
discriminating them from eugonadal men.
DISCUSSION
The hypogonadal syndromes examined in this series are repre-

sentative of the pathophysiologic varieties encountered by physi-
cians in clinical practice. These include failure of gonadal func-
tion presumably due to either an intrinsic gonadal abnormality or
to inadequate pituitary gonadotropin secretion, with a variety of
nosologic entities in each group.
In this study, determination of concentrations of FSH and LH

in a single randomly collected sample of plasma clearly discrimi-
nated men with normal gonadal function from men with hypogonadism
secondary to gonadal failure per see It is conceivable that in a
larger group of anorchic men or in a larger group of normal men
some might be found in whom concentrations of FSH and LH in random
plasma samples would be identical. It seems likely, however, that
in the majority of instances of hypogonadism due to gonadal failure,
diagnostic information could be obtained from a single determination.
In contrast, the diagnosis of hypogonadism due to inadequate

pituitary gonadotropin secretion was more complex. Hypogonadism
associated with low levels of gonadotropin secretion in syndromes
characterized by chronologically delayed spontaneous pubescence
presented particularly difficult problems. It is noteworthy that
in normal prepubertal males, as in hypogonadal males in this study,
determination of FSH and LH activity in a single specimen of plasma
occasionally failed to discriminate an individual prepubertal boy
from a sexually mature normal male (5,13-17). Similarly, low but
variable urinary total gonadotropin activity has been observed in
concentrates of 24-hour urine collections in prepubertal boys (13).
Thus, results of a single determination of either plasma or urinary
gonadotropin activity may not permit definitive categorization of
pituitary gonadotropin secretion among these patients.
Failure of measurements of plasma FSH and LH to provide diag-

nostically decisive results might be supposed to be due uniquely to
reagents and methods of measurement used in this study. Since sim-
ilar results have been obtained by others using different reagents,
this explanation does not appear to be tenable (14,16,17).
The failure of measurement of FSH and LH in randomly collected

plasma samples to be diagnostic is methodologic in part since it has
been shown that both precision and specificity are less dependable
when concentrations of gonadotropins are near the limits of detec-
tion by radioimmunoassay (6,7). When concentrations are low, it is
frequently not possible to detect activity at sample volumes less
than the maximum which can be tolerated without influencing speci-
ficity of measurements in some assay systems (18). Under these
conditions, then, inability to demonstrate similarity in dose re-
sponse relationships of unknown and reference preparation makes it
impossible to employ this test for specificity.
When specificity of the assay is uncertain, discrepancies be-

tween results of these measurements and the degree of hypogonadism
might be expected and, indeed, occur (13). Whether the discrepan-
cies relate to substances which are antigenically active but bio-
logically inactive gonadotropins cannot always be determined. It
is tempting to speculate that differences between immunologic and
biologiC potency estimates shown for buffer solutions of a variety
of pituitary extracts (19) may also exist for gonadotropins in plas-
ma. Developments of methods for extraction and concentration of
gonadotropins from plasma may permit systematic investigation of the
antigenic and biologiC properties of plasma gonadotropins. In addi-
tion, such techniques may serve to enhance both the precision and
specificity of the determination, increasing the diagnostic value
for the individual patient of either a single measurement or alter-
nately of measurements of activity in a series of specimens.
In summary, then, when low plasma gonadotropin concentrations

prevail, that is, wren pituitary gonadotropin secretion is reduced,
measurements of concentrations in four or five randomly collected
plasma samples may not be completely reliable for diagnostic cate-
gorization. Under these circumstances, failure of concentrations
of immunoreactive FSH and LH in plasma to rise in response to clomi-
phene is consistent with a diagnosis of hypogonadism secondary to
inadequate pituitary gonadotropin secretion. Used in this manner,
radioimmunoassays for plasma gonadotropins provide information diag-
nostically useful for evaluation of hypogonadism in men.
REFERENCES
1. Odell, W.D., Rayford, PaL. and Ross, G.T., J. Lab. Clin. Med.,
70: 293, 1967.
2. Odell, W.D., Ross, G.T. and Rayford, P.L., J. Clin. Invest.,
46: 248, 1967.
3. Cargille, C.M., Rodbard, D. and Ross, G.T., J. Clin. Endocr.,
28: 1276, 1968.
4. Rodbard, D., Bridson, W.E. and Rayford, P.L., J. Lab. Clin.
Med., 74: 770, 1969.
5. Weaver, C.K. and Cargille, C.M., unpublished observations.
296 G. T. ROSS
6. Rodbard, D., Rayford, P.L., Cooper, J.L. and Ross, G.T., J.

Clin. Endocr., 28: 1412, 1968.
7. Rodbard, D., Cooper, J.A., Proc. International Symposium on
Radio Isotopes in Medicine: In vitro Studies. International
Atomic Energy Agency, 1969.
8. Bardin, C.W., Ross, G.T., and Lipsett, M.D., J. Clin. Endocr.,
:?J..: 1558, 1967.
9. Cargille, C.M., Ross, G.T. and Bardin, C.W., Lancet, 1: 1298,
1968.
10. Bardin, C.W., Ross, G.T., Rifkind, A.B., Cargille; C.M. and
Lipsett, M.B., J. Clin. Invest., 48: 2046, 1969.
11. Peterson, N.T., Midgley, A.R., and Jaffe, R.B., J. Clin. Endocr.,
28: 1473, 1968.
12. Abeyaratne, M.R., Aherne, W.A. and Scott, J.E.S., Lancet, 1:
822, 1969.
13. Rifkind, A.B., Kulin, H.E. and Ross, G.T., J. Clin. Invest.,
46: 1925, 1967.
14. Schalch, D.S., Parlow, A.F., Boon, R.C. and Reichlin, S., J.
Clin. Invest., 47: 665, 1968.
15. Cargille, C.M.,~ayford, P.L., and Howland, L.A.,Excerpta
Medica Foundation: International Congress Series No. 157:
187, 1968.
16. Faiman, C. and Ryan, R., J. Clin. Endocr., :?J..: 444, 1967.
17. Saxena, B.B., Demura, H., Gandy, H.M. and Peterson, R.E., J.
Clin. Endocr., 28: 519, 1968.
18. Burr, I.M., Brant, D.B., Sizonerho, P.C., Kaplan, S.L. and
Grumbach, M.M., J. Clin. Endocr., 29: 948, 1969.
19. Albert, A., Rosemberg, E., Ross, G.T., and Paulsen, C.A. and
Ryan, R.J., J. Clin. Endocr., 28: 1214, 1968
DISCUSSION
TROEN: I should like to report briefly on studies carried out by Dr.

Howard Nankin and myself measuring serum gonadotropin levels in 11
men following bilateral orchiectomy for prostatic carcinoma. Six pa-
tients with normal preoperative FSH levels and five patients with
elevated levels showed a steady increase in FSH during the days fol-
lowing surgery. The average increase each day was the same in both
groups. In seven patients of this group with normal preoperative LH
levels there was an early and marked increase in LH levels peaking on
the second day and then falling to a lower but still elevated level.
The different response of FSH and LH following orchiectomy suggests
that release from gonadal inhibition (or metabolism of gonadotropin)
yields a different pattern of synthesis and/or secretion of the two
gonadotropins. This may be an inherent characteristic of the hypo-
thalamic-pituitary axis or may reflect different inhibitory factors
from the testis for each gonadotropin.
LUNENFELD: Dr. Ross, I was very interested in your clomiphene test.

I think you have confirmed, once and for all, that this rules out
clomiphene therapy for hypogonadotropic males.
ROSS: I have no experience in the use of clomiphene in the treat-

ment of hypogonadotropic hypogonadism. However, I would suppose
that clomiphene would be ineffective in instances of hypo gonadotropic
hypogonadism such as those we have studied in this series.
PAULSEN: Dr. Santen in our laboratory has studied the response to

30 days of clomiphene administration in patients with hypogonadotro-
pic eunuchoidism. There was no LH rise.
ROSS: In our cases, there was clinical evidence of gonadal failure

manifest in terms of low plasma concentrations of testosterone asso-
ciated with loss of libido and potentia in these hypopituitary sub-
jects. Some of them had failure of development of secondary sexual
characteristics. The only patients in whom we had testicular biop-
sies were the seven subjects with hyposmia and hypogonadism. In a
word, "immaturity" describes the nature of the testicular lesion.
Functionally these individuals have what we call a "double defect",
in that in addition to hypogonadotropism they fail to elevate plasma
testosterone concentrations in response to administration of 4,000
IU of HeG per day for 5 days.
E. STEINBERGER: Dr. Ross, have you looked for a dose-response to

clomiphene in individuals who did respond? Have you followed the
gonadotropin levels in treated individuals after discontinuation of
therapy? If so, for how long?
ROSS: In response to the first question, there is a dose response

relationship, in that the mean values observed after 100 mg of clo-
miphene per day for six days are about half the mean values observed
for 200 mg of clomiphene per day for six days. I cannot answer the
second question since we have n0t systematically followed the plasma
concentrations of FSH and LH serially after the discontinuation of
clomiphene.
HUDSON: Dr. Lunenfeld asked about the treatment of hypogonadotropic

males with clomiphene. We have observed two patients who have levels
of between 110-200 ng/lOO ml of testosterone when treated with clo-
miphene in doses of between 50 and 100 mg. per day. They had a fur-
ther fall in the levels of their plasma testosterone. I have no ex-
planation of this phenomenon.
ROSS: I am unable to explain this observation either. I would draw

your attention to a paper by Kulin and his associates in Grumbach's
laboratory in which effects of the administration of clomiphene on
urinary FSH excretion by prepubertal males were reported. A decline
in urinary excretion of follicle stimulating hormone associated with
a decline in concentrations of testosterone in plasma was observed.
The doses of clomiphene to which these prepubertal subjects were sen-
298 G.T.ROSS
sitive were of the order of magnitude of l mg or less per square

meter of body surface, or less, per day. These patients differ
from the classical patients with Kallman's syndrome in that these
occurred sporadically, that is, not in family constellations in
which the syndrome was recurrent. In addition, some of the signs
and symptoms occurring in the classical syndrome were not seen in
our patients. Our patients failed to detect odors at concentrations
which differ by orders of magnitude from the concentrations at which
any individual with normal sense of smell would be able to detect
them and were thus classified as hyposmic rather than anosmic as in
classical Kallman's syndrome. None of the nine subjects that we
have had the opportunity to study up to now has had midline defects
such as hare lip or cleft palate and only two were color blind. Our
patients may have a syndrome which might be regarded as a subset of
Kallman's syndrome.
CHAPTER IV
SECTION 4: METABOLIC EFFECTS OF GONADOTROPINS

EARLY EFFECTS OF FSH UPON TESTICULAR METABOLISM
Anthony R. Means
Departments of Ob-Gyn and Physiology, Vanderbilt
University School of Medicine, Nashville, Tennessee
Growth and maturation of the testis culminating in the onset of

complete spermatogenesis are clearly some of the most profound phy-
siological changes which occur during the postnatal development of
the male. These events are regulated to a large degree by two pep-
tide hormones elaborated in the anterior hypophysis, i.e. LH and
FSH. However, the exact role of the pituitary gonadotropins in the
initiation and maintenance of spermatogenesis is not clearly under-
stood. The current concept, based primarily on morphological evi-
dence, is that FSH maintains the germinal epithelium and LH facil-
itates the completion of spermatogenesis by stimulating the Leydig
cells to produce testosterone (1).
It has been demonstrated that FSH, when given to immature or

mature hypophysectomized rats, markedly increases the size of the
testis but will not accelerate the appearance of mature sperm nor
increase the secretory activity of the Leydig cells (2,3). More-
over, administration of FSH alone does not result in formation of
germinal cells more advanced than spermatocytes (4,5). In addition,
for final stages of spermatogenesis as well as for hypertrophy of
the male accessory glands, an androgenic influence is needed (2,4,
5). This androgenic stimulation can be provided either indirectly
by administration of LH or directly by testosterone (6). These
studies, coupled with the demonstration by many investigators that
LH stimulates testicular steroidogenesis (7,8), offers support to
the suggestion that the primary function of LH is to direct the syn-
thesis of testosterone which in turn is necessary for completion of
the spermatogenic process. On the other hand the function of FSH
remains uncertain.
There is a gFeat lack of information concerning the biochemi-

301
302 A. R. MEANS
cal effects of FSH in the mammalian testis. Furthermore, those

studies which exist are primarily concerned with the long-term ef-
fects of FSH on testicular lipid metabolism. Gambal and Ackerman
(9) have reported that following hypophysectomy the concentration
of individual phospholipids in the testis changes,and administra-
tion of FSH helped to maintain the concentrations observed in un-
treated controls. FSH has also been shown to increase the incor-
poration of radioactive linoleic acid into Lipid in testes from
hypophysectomized rats (10).
Since LH was also shown to affect lipid metabolism (9,10) we

undertook some studies to determine whether FSH stimulates phos-
pholipid synthesis in the testis of hypophysectomized rats and if
so whether the effect could be attributed to contamination with LH.
It was demonstrated that FSH stimulated incorporation of various
precursors into testicular phospholipids. Stimulation was demon-
strated by incorporation of 32p into a total phospholipid fraction
as well as by incorporation of labelled choline and ethanolamine
into total and individual phospholipids (1). Finally the stimula-
tion was shown to be specific for FSH since treatment of this hor-
mone with urea or with specific antibody to LH failed to destroy
the stimulating activity.
Clearly the early effects of FSH upon testicular metabolism

have been neglected. Since FSH increases testicular growth, ef-
fects upon protein and nucleic acid biosynthesis seem logical start-
ing points in any attempt to study the mechanism of action of this
gonadotropin. Therefore a series of studies were initiated in this
laboratory designed to study the regulation of testicular protein
biosynthesis following the administration of a single injection of
FSH.
FgH was shown to enhance the testicular incorporation of var-

ious 1 C-labelled amino acids into acid-precipitable material with-
in 30 minutes following a single injection to rats aged 20 days (12).
This respoose to FSH was time-dependent (rnaximal stimulation was
reached one hour after injection), was dose dependent within the
range of 0.2 to 200 ~g per rat, and was abolished by addition of
puromycin in vitro or administration in vivo. Table 1 demonstrates
this effect and also illustrates that stimulation of testicular pro-
tein biosynthesis by FSH was specific for that hormone and for other
gonadotropins with FSH-like activity. It can be seen that PMS caus-
ed a significant stimulation at a dose of 100 I.U. whereas boiled
FSH, LH, prolactin, TSH, HCG, or bovine serum albumin failed to
exert any demonstrable increase in protein synthesis (12). Finally,
treatment of the FSH preparation with neuraminidase, which inacti-
vates this hormone, abolishes the stimulation, whereas the stimulat-
ing activity of FSH was unaltered under conditions where the LH con-
tamination was destroyed by pretreatment with 6 M urea or 4 M guan-
idine-HCl.
METABOLIC EFFECTS OF GONADOTROPINS 303
Table 1. Specificity of the FSH-Mediated Stimulation of Testicular

Protein Biosynthesis
Specific Activity
Treatment (dpm/mg protein) P
Control 1620
FSH 2829 .001
Boiled FSH 1602 NS
FSH + Neuraminidase 1631 NS
FSH + Urea 2861 .001
PMS 2209 .01
HCG 1802 NS
ICSH 1611 NS
Prolactin 1654 NS
TSH 1620 NS
Serum Albumin 1642 NS
Saline 1615 NS
All substances were injected intraperitoneally 1 hr prior to

killing of animals. Amounts of substances injected were as follows:
20 ~g each of FSH, boiled FSH, ICSH, prolactin, TSH, or bovine se-
rum albumin; 150 I.U. HCG; 100 I.U. PMS. All compounds were dis-
solved in 0.2 ml of 0.9% aqueous saline. All rats were 20 days old
and each served as its own control as follows: one testis was re-
moved and 1 hr following injection of test compound the other testis
was removed. Testis tissue was incubated at 37 0 for 30 minutes in
Krebs-Ringer bicarbonate buffer containing 0.5 ~C of lysine- 14C
(0.1 roM). Values represent the mean from 6 animals (12). The hor-
mone preparations used were ovine FSH (NIH-FSH-S3), ovine ICSH (NIH-
LH-S9), ovine TSH (NIH-TSH-S2), bovine prolactin (NIH-P-Bl), PMS
(Ayerst Equinex) and HCG (Ayerst A.P.L.).
The stimulation of testicular protein synthesis by FSH was

shown to be dependent upon the age of the animal, that is stimula-
tion was observed in testes of rats aged 15-24 aays, but not in
testes of rats older than 25 days (13). On the other hand the tes-
tes of rats of any age will respond to FSH within 18 hours of hypo-
physectomy and will continue to respond for at least 25 days after
operation. Glucose was also shown to stimulate testicular protein
biosynthesis in an age-dependent fashion (14). However, sensitivi-
ty to glucose and to FSH were shown to occur at different ages (Ta-
ble 2). Glucose did not increase protein synthesis in the testis
before the age of 28 days and the stimulation observed in the adult
gradually declined following hypophysectomy until no response was
observed 25 days after operation.
Spermatids first appear between 26 and 28 days of age in the

immature rat and disappear from the adult animal within 25 days of
304 A. R. MEANS
Table 2. Effects of FSH or Glucose Upon Protein Biosynthesis in

Testes of Intact or Hypophysectomized rats.
Intact Specific Activity (dprn/ mg protein)

Age (Days) Control FSH Glucose
15 900 1260 892
20 1605 2875 1640
24 1847 3368 1900
28 1954 2035 2262
30 1150 1203 1665
35 1222 1203 2005
40 1106 1112 2200
60 1210 1193 2492
Hypox
Duration Control FSH Glucose FSH + Glucose
4 hr 900 884 1863
18 hr 920 1082 1850
3 days 842 1312 1706
5 days 530 922 980
15 days 277 452 398 536
25 days 235 408 230 420
Each animal served as its own control as previously described

(13). FSH (50 ~g/100 g body weight) was administered intraperito-
neally 1 hr. before killing of animals. Glucose when present was
added to the incubation at a final concentration of 9mM. Testis
tissue (60-100 mg/flask) was incubated with 0 5 ~ C (0.1 mM) lysine-
0
14C for 30 min. at 37° as previously described (12). Numbers re-

present the mean value from 6 animals. The FSH preparation used
was ovine NIH-FSH-S4.
hypophysectomy. Davis and Firlit (14) have reported autoradio-

graphic studies on the incorporation of lysine- 14C into cells of the
mature rat testis which indicate that the incorporation of the amino
acid resulting from the addition of glucose to the medium occurs for
the most part in spermatids. Our results would support this conclu-
sion. On the other hand the effect of FSH is not related to the
presence or absence of spermatids since the hormone stimulates pro-
tein biosynthesis in the young rat before these cells have appeared,
fails to stimulate in the normal adult testis which contains sperma-
tids, but increases protein synthesis within a few hours following
hypophysectomy when spermatids are still present in large numbers.
It is likely therefore that glucose and FSH exert their effects on
different cell types within the testis.
The manner in which glucose stimulated testicular protein bio-

synthesis in the adult rat was then investigated. It was demon-
strated that the effect of glucose was indirect, that is, through
its ability to act as a source of energy and thus maintain testic-

ular levels of ATP (15). Furthermore,the ATP produced from the
oxidation of glucose was shown to serve at least a dual purpose;
firstly in the transport of amino acids into the testicular cells,
and secondly at some step(s) beyond the intracellular entry of
amino acids (16). Finally, glucose had no effect upon cellular con-
centrations of ATP in the immature rat testis, which is in keeping
with the inability of this substance to stimulate protein biosyn-
this.
Stimulation of testicular protein biosynthesis by FSH appears

to be of a more direct nature. Two lines of evidence demonstrate
that the stimulation cannot entirely be attributed to enhanced ami-
no acid transport into testicular cells (12). In the first place
FSH, under conditions where protein biosynthesis was stimulated,
did not increase t'1e cellular transport of the model compound, a -
amino-isobutyric acid, which is transported in a manner similar to
naturally occurring amino acids but is not metabolized or incorpo-
rated to any demonstrable degree. Secondly, the stimulation by FSH
was not diminished by increasing the concentration of lysine in the
incubation medium, i.e. further addition of substrate did not result
in increased incorporation into protein. Likewise, FSH did not in-
crease the activation of lysine- 14C into lysyl-transfer RNA.
Since FSH administration resulted in accelerated protein bio-

synthesis, but this gonadotropin did not increase the transport or
activation of amino acids, we decided to investigate the effect of
the hormone upon the activity of testicular polyribosomes. There-
fore, polyribosomes were prepared from rat testis. Characteriza-
tion of this preparation by sucrose gradient centrifugation, elec-
tron microscopy and analytical ultracentrifugation revealed a large
predominance of ribonucleoprotein particles which contained more
than two ribosomes (17). Furthermore, conditions were described for
optimal incorporation of valine- 14C into peptide by a cell-free sys-
tem. This ribonucleoprotein preparation was then used to investi-
gate the mechanism by which FSH stimulates testicular protein bio-
synthesis.
When testes from rats which had received a single injection of

FSH one hour before killing were incubated with valine- 14C before
polyribosomes were prepared, the specific activity of protein asso-
ciated with all polysomal fractions was higher than in similar pre-
parations from control animals (Fig. 1) (18). On the other hand,
FSH neither demonstrably influenced the proportion of testicular
ribosomes appearing as polyribosomes nor altered the relative pro-
portion of various polyribosomal species. Incorporation of valine-
14C into protein by polyribosomes in vitro was increased by prior
treatment of the animal with FSH in vivo. No effect of the hormone
upon protein synthesis was demonstrable until 1 hour after admini-
stration (Table 3). The response reached a maximum at 2 hours and
306 A. R. MEANS
TOP BOTTOM
250
200
150
E
c..
"0 A254
100 .300
.200
50
.100
O~--~----~----~-----L-----L--~
o 5 10 15 20 25 30
FRACTION
Fig. 1. Effect of FSH administered in vivo upon the radioactive
labelling of testicular polyribosomes in vitro. FSH
(200 ~g/100 g body weight) was administered to hypophys-
ectomized rats as a single intravenous injection 1 hr be-
fore killing. Testis tissue (2.5 g) was incubated with
10 ~C of valine- 14C for 30 min at 37° and analyzed as pre-
viously described (18). The FSH preparation was ovine
NIH-FSH-S6. (From: Means, et ale Biochemistry, ~: 4293,
1969).
Table 3. Influence of FSH upon Protein Biosynthesis by Testicular

Polyribosomes in vitro as a Function of Time Following
Intravenous Injection
Specific Activity
Duration of FSH (pmoles val- 14C/mg
(minutes) ribosomal protein)
o 14.14
15 14.08
30 14.58
60 19.51
120 23.19
180 19.52
240 15.13
480 14.39
All animals were hypophysectomized at 150-160 g and used be-

tween 3 and 4 weeks postoperatively. FSH (200~ /100 g body weight)
was injected intravenously at various times before sacrifice of an-
imals. Polyribosomes (400~ /tube) were incubated in a cell-free
system containing 1.0~ C valine- 14C for 30 min. at 37°. The values
represent the mean of determinations performed in triplicate and
have been corrected for zero controls (17). The FSH preparation
was ovine NIH-FSH-S6.
then began to decrease until by 8 hours no detectable difference

was observed between incorporation activity by polyribosomes from
control or FSH-treated animals. Moreover, it was demonstrated that
the stimulation of peptide synthesis by FSH is independent of the
source of enzymes necessary for amino acid activation (i.e. no dif-
ference was observed when pH 5 enzyme was from control or FSH-treat-
ed testis, or liver).
Injection of actinomycin D one hour before injection of FSH

was shown to prevent the stimulation of protein biosynthesis by
testicular polyribosomes in vitro without demonstrably decreasing
the rate of synthesis by particles from control animals (18). Fur-
thermore, although polyuridylic acid increased protein synthesis in
vitro by polyribosomes from treated and from control animals, in-
corporation was equal with polyribosomes from these two sources in
presence of excess poly U. It appears, therefore, that the capaci-
ty of polyribosomes to support peptide synthesis is the same wheth-
er these particles are isolated from hormone-treated or control an-
imals. These observations suggested that FSH stimulates testicular
protein synthesis by increasing the synthesis of RNA (possibly mes-
senger RNA), although additional effects upon translation of mRNA
cannot be excludedo
We then attempted to ascertain whether FSH increased testicular

308 A. R. MEANS
Table 4. Effect of FSH upon Synthesis of Rapidly-Labelled Nuclear

RNA and Uptake of Uridine-5- 3H in Testes of Rats Aged 20
Days
Tissue Radioactivity Specific Activity
Time After Acid Soluble Acid Insoluble Nuclear RNA

FSH (min) (dpm/mg DNAxl0- 3 ) ('70) (dpm/mg DNA) (%) (dpm/mg)
0 100.6 98.9 1020 1.0 2500

15 109.0 96.9 1800 3.1 3680
30 107.0 96.5 2760 3.5 5100
45 103.0 97.0 2230 3.0 4060
60 106.0 97.4 2110 2.6 2600
120 105.0 98.0 1800 2.0 2220
FSH (50 ~g/rat) was injected intravenously at various times be-

fore killing and all animals received 100 ~C uridine-5- 3H intra-
venously 10 min prior to killing. Testes were homogenized and ali-
quots were taken for determinations of acid-soluble and acid-insol-
uble radioactivity as well as for DNA. Nuclei were isolated from
the remainder of the homogenate and rapidly-labeled RNA was extract-
ed. All of these methods have been described previously (19). Re-
sults represent the mean values of 6 experiments. Testes of 10 an-
imals were pooled for each experiment. The FSH preparation was
ovine NIH-FSH-S7.
RNA synthesis prior to the observed effect on protein biosynthesis.

It can be seen from Table 4 that the specific activity of rapidly-
labeled nuclear RNA was increased significantly within 15 minutes
following a single injection of FSH. This response reached a maxi-
mum at 30 minutes and by one hour was declining rapidly toward con-
trol values. Moreover, simultaneous measurements of the acid soluble
pool of uridine-5- 3H during these experiments showed that uptake of
the nucleotide was unaltered by FSH. These changes in nuclear RNA
synthesis clearly precede the enhancement of testicular polyriboso-
mal incorporation activity first observed within one hour following
FSH administration and argue for an early effect of the hormone upon
genetic transcription.
The precise mechanism of action of FSH is still uncertain. Re-

cent reports indicate that some effects of this gonadotropin may be
mediated through a stimulation of adenyl cyclase (20,21). ReguJa~
tion of protein synthesis may occur at two principal sites, at the
level of transcription in the nucleus or in the cytoplasm at the
level of translation. The early changes in rapidly-labeled nuclear
RNA, the inhibitory action of actinomycin D and the time-course of
events following FSH would suggest a very early action of this hor-
mone at the nuclear level of protein synthesis. This then might re-
sult in new gene transcriptions and the subsequent de novo synthesis
of protein molecules necessary for further growth and development of

the seminiferous epithelium.
ACKNOWLEDGEMENT
The work presented herein was supported in part by Grant 66-

16023 from the Australian Research Grants Committee (to P.F. Hall),
Research Grant RA-5 from the Morrison Trust of San Antonio, Texas,
and by U.S. Public Health Service Health Sciences Advancement Award
5-S04-FR06067 to Vanderbilt University. Thanks are extended to the
Endocrine Study Section at NIH for gifts of peptide hormones used
in these studies.
REFERENCES
10 Gemzell, C. and Roos, P., In The Pituitary Gland, eds. Harris,

GoW., and Donovan, B.T., London, Butterworth Press Vol. 1,
492, 1966.
2. Greep, R.O., van Dyke, H.B. and Chow, B.F., Endocrinology,
30: 635, 1942.
3. Simpson, M.E., Li, C.H. and Evans, H.M., Endocrinology, 48:
370, 1951.
4. Lostroh, A.J., Acta Endocr. (Kobenhavn), 43: 592, 1963.
5. Lostroh, A.J., Johnson, R. and Jordan, C.W., Acta Endocr.
(Kobenhavn), 44: 536, 1963.
7. Hall, P.F. and Eik-Nes, K.B., Biochim. Biophys:-Acta, 63: 411,
1962.
8. Hall, P.F. and Young, D.G., Endocrinology, 82: 559, 1968.
9. Gambal, D. and Ackerman, R.J., Endocrinology, 80: 231, 1967.
10. Go swami , A. and Williams, W.L., Biochem. J., 105: 537, 1967.
11. Yokoe, Y., Means, A.R. and Hall, P.F., Biochim. Biophys. Acta,
187: 278, 1969.
12. Means, A.R. and Hall, P.F., Endocrinology, 81: 1151, 1967.
13. Means, A.Ro and Hall, P.F., Endocrinology, ~: 597, 1968.
14. Davis, J.R. and Firlit, C.F., AIDer. J. Physiol., 209: 425, 1965.
15. Means, A.R. and Hall, P.F., Endocrinology, 83: 86~968.
17. Means, A.R., Hall, P.F., Nicol, L.W., Sawyer, W.H. and Baker,
C.A., Biochemistry, ~: 1488, 1969.
18. Means, A.R. and Hall, P.F., Biochemistry, ~: 4293, 1969.
19. Means, A.R. and Hamilton, T.H., Proc. Nat. Acad. Sci., U.S.A.
50. 686, 1966 •
20. Murad, Fo, Strauch, B.S. and Vaughn, M., Biochim. Biophys. Acta,
177: 591, 1969 0
21. Kuehl, F.A., Patanelli, D.J., Tarnoff, J. and Humes, J.L.,
BioI. Reprod. (in press).
310 A.R.MEANS
DISCUSSION
SOLARI: Dr. Means' results agree with our unpublished results on

the effect of FSH on rapidly labelled RNA in the rat testis. I would
say from our data on the action of actinomycin on the adult rat tes-
tis that this action seems to be restricted to the spermatogonia and
the Sertoli cells. Thus, it could be that the transient effect of
FSH is also restricted to the Sertoli cells and the spermatogonia.
MEANS: One has to be very careful, I think, in interpreting data

concerning actinomycin D. Firstly, actinomycin D is a very potent
drug that has all sorts of systemic effects as well as affecting
RNA systhesis. Secondly, it preferentially affects ribosomal RNA
synthesis, that is to say even if one obtains an inhibition of RNA
synthesis up to as much as 90 percent, messenger RNA synthesis
(which may comprise only 2 to 3 percent of all the RNA synthesis in
the cell) could continue merrily along its way. With regard to what
cell types are targets for FSH, I purposely avoided that subject.
We used rats aged 20 days and Dr. Clermont was kind enough to tell
me that at least 50 percent of the cells present in the testis at
that age are primary spermatocytes. I would find it difficult to
imagine that these large increases in rapidly labelled nuclear RNA
synthesis would be localized to one cell type unless this cell com-
prised a large proportion of the mass. I would rather suggest that
we may unfortunately be looking at both specific and non-specific
effects of this hormone. We are in the process now of isolating
testicular DNA and separating on hydroxyapatite columns the unique
sequences of the DNA from the repeating or the redundant sequences.
When we hydridize our rapidly labelled nuclear RNA back against
these two types of DNA sequences, I believe that this will obtain
much more information about what is going on. If we find only in-
creased competition in the redundant species this would indicate
that we have a non-selective effect of FSH. Finally, in our hands
this is a very transient effect and always when one has a transient
effect one wonders about the reasons. This is particularly true in
this case since we know that if one gives multiple doses of FSH to
immature hypophysectomized rats one obtains an increase in testicu-
lar weight. Therefore, we probably have two things going on, but I
would not be willing to. make any comments at the present time about
what cell types FSH is specifically affecting, except to rule out
spermatids and spermatozoa.
LUNENFELD: We have found that Dr. Means' experiments on the effect

of FSH or role of protein on nucleic acid synthesis in the testis
give results extremely similar to ours on the effect of FSH on the
mouse ovary. Furthermore, I was extremely interested in the effect
of HCG. We found a similar effect of HCG on the ovary. We postu-
lated that this might be due to an endogenous release of FSH and
this was found to be true since when HCG was injected together with
an antiserum to rat gonadotropin the effect of HCG was completely
abolished.
CHRISTENSEN: I wonder if ordinary autoradiography at the light

microscope level would not suffice to identify the cell type in-
volved, since one administers soluble amino acids and can test for
insoluble proteins. In the polyuridine experiments there was the
implication that free ribosomes were being taken up into polysomes,
and yet in your polysome profile, Dr. Means, you had an extremely
small free ribosomal peak. Where did the ribosomes for that large
an increase in protein synthesis come from?
MEANS: With regard to the autoradiography, I agree that this is a

reasonable type of approach. We have some preliminary results that
indicate very sim~lar data to what Prof. Mancini has reported pre-
viously. We find radioactivity localized over the primary spermato-
cytes and there may be also an increase over the Sertoli cells. I
feel obligated to stress that these studies are very preliminary in
nature. With regard to the second question, you should realize
that this is an absorbancy tracing and in these particular studies
we are selectively isolating polyribosomes. Therefore, the monomer
peak should be small. This is not to say, however, that all of the
polyribosomes so isolated are active in protein biosynthesis, and
in fact this is probably not the case. Secondly, we were applying
only 10 A 254 NM of polyribosomes to each gradient since we were
interested primarily in the optimal conditions for separation of the
various species of polyribosomes which we had present in our prepa-
ration. Finally, it is well known that in any cell-free system as
protein synthesis continues in vitro polyribosomes dissaggregate
monomers and that the addition of an exogenous messenger will re-
establish these ribosomes into functional polysomes. I tried not
to make any particular conclusions from the data with Poly U except
one simple point: if one believes the validity of such a study the
data would indicate that the capacity of the ribosomes to synthesize
protein is similar whether isolated from testes of control or FSH-
treated rats.
OHNO: I wonder if cyclic AMP is mediating the generalized stimula-

tion which you observed after FSH?
MEANS: Two recent reports by Murand and Kuehl bear directly on this
subject. FSH was demonstrated by these two groups of workers to en-
hance production of cyclic AMP when added to a whole tissue prepara-
tion from the testis. This response was also very age-dependent as
we have demonstrated for protein biosynthesis, with maximal activity
occurring at 20 days of age. FSH added in vitro will actually cause
a stimulation of adenyl cyclase activity. We have injected dibutyryl
cyclic AMP directly under the tunica and indeed find an increase in
protein synthesis with similar kinetics to the effect of FSH. We
find that when we analyze the pep tides synthesized by polyribosomes
in vitro as we showed for the FSH they are again similar. However,
at this point, we have one discrepancy that I have no explanation
for at present. That is, that under no conditions does dibutyryl
312 A. R. MEANS
cyclic AMP plus or minus methyl xanthine derivatives exert any ef-
fect on rapidly labelled nuclear RNA synthesis. The mechanism of
action of most trophic hormones is probably mediated by some sort
of intercellular second messenger. Cyclic AMP is most definitely
a likely candidate.
ROSEMBERG: Dr. Means, would you please indicate the dose of FSH
used in the experiments presented tonight? Is this one of the NIH
ovine materials?
MEANS: The most common dose given was about 200 micrograms per rat.
All of the preparations used were NIH ovine preparations. In each
case, the LH contaminating activity was removed.
A. STEINBERGER: Dr. Means, are you certain that nothing happens

in the Leydig cells? Was there any particular reason for using
ovine NIH-FSH which has some LH activity?
MEANS: We did get rid of the LH contaminating activity. I still

cannot really say whether there is any effect on the Leydig cells
or not. I think some information from several workers, one of whom
is Dr. Hall, I believe, is important in this regard. They have iso-
lated interstitial cells and seminiferous tubules, and looked at the
effect of LH and FSH on steroidogenesis. They find that FSH has no
effect on production of testosterone by isolated Leydig cells.
CHRISTENSEN: Even if the interstitial cells were carrying on pro-

tein synthesis, I doubt that it would show in your results, because
the interstitial cells in rats only constitute about 1.5% of the tes-
tis volume.
PAULSEN: I think it is important to consider what effect testoster-

one might have in stimulating protein synthesis in the seminiferous
tubules. Also in that regard, did you indicate that ICSH had an ef-
fect on protein synthesis?
MEANS: We have tried testosterone, in fact we have tried a number

of naturally occurring steroids, none of which exert any demonstra-
ble effect on the parameter measured. LH (ICSH) showed no effect.
PMS and HCG did have some effect; but both of these substances have
follicle stimulating activity.
MANCINI: Dr. Means, in order to localize the site of action of FSH,

did you consider it useful to try to separate the seminiferous tu-
bules from the interstitial tissue? Would you consider the effect
of FSH a direct one or could it be mediated by ~ome changes in the
blood flow, which in turn by extravascular diffusion may offer to
the seminiferous tubules an increased amount of nutrients?
MEANS: Blood flow has of course been shown by many people to be

very important. Larry Ewing's very elegant studies with the per-
fused testis have indicated this point. I think that this is cer-
tainly a possibility, and would not rule it out. One of the prob-
lems readily apparent from our studies is that all of these data
were obtained in vivo; that is to say the hormone was administered
exogenously.
RADIOAUTOGRAPHIC STUDIES OF PROTEIN SYNTHESIS IN CELLS OF THE
SEMINIFEROUS EPITHELIUM
Joseph R. Davis and Casimir F. Firlit 1
Dep~rtment of Pharmacology and Therapeutics,

Loyola University Stritch School of Medicine,
Maywood, Illinois, U.S.A.
Because of the complex histological pattern of the spermatogen-

ic cycle, it has been exceedingly difficult to ascertain the speci-
fic metabolic functions of the various cell types involved in mam-
malian spermatogenesis. In addition, very little is known about the
metabolism of cells undergoing meiotic division and of the biochem-
ical events associated with the transition from somatic mitosis to
meiosis. However, the continuous nature of the spermatogenic cycle
which begins with a stem cell spermatogonium and ends with the ma-
ture spermatid seems to be uniquely suited for a comparative study
of the three main phases of growth, namely mitotic division, meiotic
division and morphogenic differentiation. It would appear that this
aspect of testicular biochemistry is of extreme importance since not
only is meiosis the mechanism by which the diploid number of chromo-
somes is reduced to the haploid number of chromosomes found in the
gametes, but the recombination of chromosomal segments during the
crossing-over phase of meiosis provides for constant changing of the
cell's genotype.
For these reasons, our laboratory has initiated a series of ex-

periments which have been designed to investigate protein biosynthe-
sis as it occurs in the various cells of the germinal epithelium of
the testis. Since it is not possible at the present time to physi-
cally separate and isolate the individual cells of the testicular
germinal epithelium, we have utilized two experimental approaches
designed to determine which of the several cell types of the semini-
ferous germinal epithelium of the testis were particularly suscepti-
ble to our previously reported effects of temperature (1), cryptor-
chidism (2) and glucose (3) on testicular protein biosynthesis. The
1predoctoral Trainee, U.S. Public Health Service.
315
316 J. R. DAVIS AND C. F. FIRLIT
first approach involves the measurement of the incorporation of 14C_

lysine into acid-precipitable protein of various preparations of
testes consisting primarily of only a very few cell types. The sec-
ond and more specific approach involves the radioautographic deter-
mination of the incorporation of tritiated lysine into each of the
successive cells of the cycle of the seminiferous germinal epithe-
lium.
PROTEIN SYNTHESIS IN THE CRYPTORCHID RAT TESTIS
Cryptorchidism can easily be induced in the rat by gently

forcing the testis into the abdominal cavity and anchoring it to the
lateral abdominal wall. The contralateral testis of the same animal
can then be allowed to remain in the scrotal sac to serve as a con-
trol. The predominant histological feature of the cryptorchid tes-
tis appears to be the total absence of spermatids (4). On the other
hand, the immature testis of the rat appears to consist largely of
primary spermatocytes (5).
Fig. 1 presents the in vitro incorporation of L_Iysine-U- 14C

into acid-precipitable protein of slices of immature, mature and
cryptorchid rat testes. The normal descent with age of the immature
abdominal testis into the scrotal sac was accompanied by a remark-
able decrease in testicular protein labelling. In addition, protein
labelling of the cryptorchid testis was observed to increase marked-
ly following abdominal transplantation from the scrotal sac. It
seemed quite surprising to us that a tissue such as the cryptorchid
testis which was rapidly regressing in size should display an in-
creased synthesis of protein. The question ~rose as to whether this
increased protein labelling of the cryptorchid testis, on the one
hand, was a reflection of an unmasking of the spermatogonia, primary
spermatocytes and Sertoli cells resulting from the disappearance of
the spermatids or, on the other hand, represented a true tempera-
ture-induced change in the protein-synthesizing capacity of these
remaining testicular cell types.
Fig. 2 shows photomicrographs of the three types of seminifer-

ous tubules which occurred in approximately equal numbers in the
cryptorchid rat testis 30 days after abdominal transplantation.
Only four cell types were found in the germinal epithelium of these
tubules, namely "crust" spermatogonia, pachytene primary spermato-
cytes, normal-appearing Sertoli cells and atrophic-appearing Sertoli
cells, accounting for 31, 3, 28 and 38 percent, respectively, of the
total cells remaining in the seminiferous germinal epithelium of the
cryptorchid rat testis.
Fig. 3 presents the radioautographic incorporation of tritiated

lysine into each of the remaining cells of the cryptorchid rat tes-
tis 30 days after abdominal transplantation of the testis. Slices
of testes were incubated at 37.5 0 C in a Warburg apparatus with 100
INCORPORATION OF L-LYSINE-U-C I4 INTO PROTEIN

4,000 OF RAT TESTIS SLICES
z 3,000
w
t-
o
Q:
a..
2,000
A
0'
E
"-
E
Q. SCROTAL TESTIS
u 1,000
o 15 30 45 60 75 90 105 120
AGE OF ANIMAL IN DAYS
ii- '\
~ 60010:''----.......__
B •
!~ §r 200 '",', CRYPTORCHID TESTIS
"'o---------V--------------<l
~
0~~5~~10~~,5~-2~0-~2~5--~~~~3~5--~~
DAYS AFTER ABDOMINAL TRANSPLANTATION
iii~ 16,000 0
~ !!! ,----------0---------------0
~ E 12,000 / CRYPTORCHID TESTIS
i~ /'/
~~8,000
i. _0"
/ C
.. t;
~. 4,000
_-----------
~
SCROTAL TESTIS
~r ~~--~-------+------
~
o 5 10 15 20 25 ~ 35 ~
DAYS AFTER ABDOMINAL TRANSP'LANTATION
Fig. 1. Incorporation of L-lysine-U 14C into protein of rat testis

slices. A. Effect of age of animal on protein labelling
of normal testis. B. Effect of abdominal transplantation
on testicular weights of experimentally-induced unilateral
cryptorchid rats which were 60 days old at the time of sur-
gery. C. Corresponding testicular protein labelling of
experimentally-induced unilateral cryptorchid rats which
were 60 days old at the time of surgery.
W
<Xl
PERCENTAGE DISTRIBUTION OF CELL POPULATION IN SEMINlFEROUS

TUBULES OP THE CRYPTORCHID RAT TESTIS 30 DAYS APTER ABDOMINAL
TRANSPLANTATION
Typd Tj'/J' B TY/J4 C

QUs IJJbuU lJibuu IJJbuk
'Crust' spermatogonia 55·0 ± O·S 4-3·0 ±2 ·7 3-6 ± O·3

Pachytene lO 'permatocYlcs 8·6± 1· 1
Normal Serloli cells 36·4-± 1·4- 57-0±3·0
° °
Atrophic Serloli cells 0 96·4- ± O·g
° '-
. ~~
- ------- °
The number of cells were determined in approximately twenty-five seminiferous ""o
tubules of each type for each of the fou r a nimals used.
AI! figures are percentagcs of total tubule cells± . tandard error. >
<
en
Fig. 2. Percentage distribution of the rema~n~ng cells of the semi- >
z
niferous tubules of the cryptorchid rat testis 30 days after o
()
abdominal transplantation. The animals were 60 days of age
;-n
at the time of surgery. (From: Davis, J.R. et ale J. Reprod.
:::!!
Fertil.,11: 125-131, 1966.) r-
""
=l
~c of tritiated lysine in Krebs-Ringer bicarbonate buffer for 1

hour under aerobic conditions (6). At the end of the incubation
period, the flask contents were poured through a 90-mesh stainless
steel sieve in order to collect the testicular slices. The tissue
held back by the sieve was then washed three times with cold Krebs-
Ringer bicarbonate buffer containing 0.1% non-radioactive lysine
and gently placed in cold Carnoy's solution by means of forceps.
Following fixation overnight, the tissue was dehydrated, cleared,
embedded in paraffin and sectioned at 5~. The sections were freed
from paraffin, dipped in molten Kodak nuclear track emulsion type
NTB3 and stored in the dark for 2 days. The exposed slides were de-
veloped, stained in hematoxylin solution and examined. Alternate
sections which were not dipped with nuclear emulsion were stained
with either hematoxylin alone or the periodic acid-Schiff reagent
as outlined by Leblond and Clermont (7). Grain counting was per-
formed under oil immersion using a Whipple micrometer eyepiece grid.
Grains were counted within four squares in the Whipple micrometer
and recorded as grains per 200 ~ 2. A very dense radioautographic
labelling pattern was observed over both the spermatogonia and the
Sertoli cell cytoplasm in Type A tubules of the cryptorchid rat
testis. The labelling of the Sertoli cell cytoplasm in Type B tu-
bules of the cryptorchid rat testis was found to be quite similar
to the labelling over the Sertoli cell cytoplasm present in Type A
tubules. However, in contrast, the atrophic-appearing Sertoli cells
found in Type C tubules of the cryptorchid rat testis were charac-
terized by a marked decrease in the uptake of tritiated lysine into
cellular protein. In comparison to the scrotal testis, the experi-
mental induction of cryptorchidism was found to produce an approxi-
mate four-fold increase in protein labelling of both "crust" sper-
matogonia and pachytene primary spermatocytes. The induction of
cryptorchidism was also found to result in a forty-fold increase in
protein labelling from tritiated lysine in the normal-appearing Ser-
toli cells of both Type A and Type B cryptorchid tubules. It would
therefore appear that the overall increase in labelling of acid-
precipitable protein of slices of the cryptorchid rat testis is due
not only to an unmasking of the remaining cell types with an in-
herent higher capacity for protein synthesis but is also due to a
temperature-induced stimulation of protein synthesis occurring in
these same testicular cell types, namely the "crust" spermatogonia,
pachytene primary spermatocytes and the Sertoli cells.
EFFECT OF GLUCOSE ON RAT TESTICULAR PROTEIN SYNTHESIS
Although glucose has long been shown to maintain not only the
oxygen uptake (8-12) but also the morphological integrity of adult
testicular tissue (13), few studies have been carried out on the
effect of this important substrate on testicular protein biosynthe-
sis. Our laboratory has reported that the addition of 0.009 M glu-
cose is capable of increasing the aerobic incorporation of radio-
Co)
~
o
RADIO-AUTOGRAPHIC INCORPORATION OF L-LYSINE- JR INTO CELLS OF THE

CERML'IAL EPITHEUUM OF THE CRYPTORCHID RAT TESTIS 30 DAYS AFTER
ABDOMINAL TRANSPLANTATION
CrypwTChid testis
C.11s Scrotal
/<Stis Typd TypeD Tyr., G
tubuk tubule Iu "I,
·Crusl' spermatogonia IH±O·7 67·I± 1·0 66·2±1·8 6·7tO·3
Pachytene 1° .permatocytes 7·0tO·7 27·StO·7 - - '-
Normal Sertoli cell. 1·6 ± 1·6 67·0tO·6 61 ·S± 1· 1 -
Atrophic Sertoli cells - - - - 19·I±2·4 ?"
-- o
Grains/2oo/1'wercdetcrmined overeaeh of the designaled cell types in approximately 100
»
<
c;;
microscopic field. for each of the four animals used.
The data arc presented", grains/2oo II' abo,·e background t standard error. »
z
o
Fig. 3. Radioautographic incorporation of tritiated lysine into n
cells of the germinal epithelium of slices of the cryptor- -n
chid rat testis 30 days after abdominal transplantation. -n
:;0
The animals were 60 days of age at the time of surgery. .-
::::j
(Davis, J. R. et al. J. Reprod. Fertil. 11: 125-131, 1966).
active lysine into acid-precipitable protein of slices of adult rat

testes by 600 percent (3). In marked contrast, the addition of
0.009 M glucose was found to increase protein labelling of the head
of the epididymis by 150 percent while causing no change to only a 50
percent enhancement of protein labelling in slices of jejunal mucosa,
spleen, thymus, submaxillary gland, pancreas, tail of epididymis,
kidney, brain, diaphragm, heart, lung, seminal vesicle, normal liver,
regenerating liver, Walker 256 carcinosarcoma, Jensen sarcoma and
Morris hepatoma 5123. It therefore appeared that with respect to a
total of 15 different normal tissues of the adult male rat, as well
as 3 varieties of transplantable rat tumors and a rapidly-growing
non-malignant tissue such as the regenerating rat liver, glucose
exerted a unique and characteristic stimulation of testicular pro-
tein biosynthesis.
Fig. 4A indicates the effect of exogenous glucose on the in

vitro incorporation of L-lysine-U- 14C into acid-precipitable protein
of slices of four different testicular preparations, each having its
own individual histological structure. Whereas the absolute levels
of protein labelling of the immature testis, the cryptorchid testis
and the interstitial-cell testicular tumor were all higher than the
adult scrotal testis, these tissues, which in no instance contained
any spermatids"displayed essentially no effect of glucose on pro-
tein labelling as compared to the scrotal testis. In view of the
fact that only the adult scrotal testis contained spermatids, these
data pointed to the possibility that the spermatids were the most
sensitive cell type of the seminiferous germinal epithelium with re-
gard to glucose regulation of testicular protein synthesis. As sug-
gested by Means and Hall, this remarkable stimulation of spermatid
protein biosynthesis may be due at least partially to the mainten-
ance of testicular ATP levels by glucose (14). Additional evidence
that the spermatids are highly sensitive to glucose with regard to
regulation of protein synthesis is presented in Fig. 4B. A marked
increase in the stimulatory effect of glucose on protein labelling
of slices of the normal rat testis was observed with an increase in
the age of the animal from 20 to 50 days, corresponding to the ap-
pearance of the spermatids in the developing seminiferous tubular
epithelium.
In order to investigate this possibility further, as well as

to compare protein labelling in mitotic versus meiotic cell divi-
sion, in vitro radioautographic studies were initiated on the ef-
fect of exogenous glucose on protein labelling from tritiated ly-
sine in each of the successive cells of the cycle of the seminifer-
our epithelium of the rat. The fourteen stages of the cycle of the
seminiferous epithelium in the rat have been classified according
to the description of Leblond and Clermont (7). Slices of normal
scrotal testes of the adult rat were incubated in Krebs-Ringer bi-
carbonate buffer under aerobic conditions wi th 100 u c of tritiated
322 J. R. DAVIS AND C. F. FIRlIT
lysine in the presence and absence of 0.009 M glucose. It was

found that the seminiferous epithelium of the rat testis retained
its morphological integrity in the Warburg apparatus while shaking
at a rate of 140 oscillations/minute. At the end of the 1 hour in-
cubation period at 37.5 0 C, the tissues were sectioned and dipped in
molten Kodak nuclear track emulsion. Exposure was carried out for
four days, after which the slides were developed and stained in
hematoxylin solution. Grains were counted within fo~r squares in
the Whipple micrometer and recorded as grains/200~ • A total of
four cells occupied an area of 200~ 2 for each of the successive
cell types of the seminiferous germinal epithelium with the excep-
tion of the larger pachytene primary spermatocytes of whic~ a total
of only two cells were observed to occupy an area of 200 ~ •
Fig. 5 demonstrates representative radioautographs of sections

taken from slices of rat testes incubated with tritiated lysine in
the presence and absence of exogenous glucose. A marked stimulation
of protein labelling upon the addition of exogenous glucose was ob-
served in the grain distribution over the pachytene primary sperma-
tocytes present in stage VII. Glucose was also found to cause a
great increase in the number of grains appearing over the area of
the spermatids.
Fig. 6 presents a plot of the average grain count per 200 ~ 2

above background over each of the successive cells of the cycle of
the seminiferous germinal epithelium of the rat testis one hour
after incubation with tritiated lysine both in the presence and ab-
sence of exogenous glucose. In the control system, the cells which
were found to incorporate the greatest amounts of lysine into pro-
tein were the early primary spermatocytes. In fact, immediately
after the division of type B spermatogonia into the resting primary
spermatocytes, there was a mar~ed increase in the degree of label~
ling over these cells. In the glucose-supplemented system, a marked
increase in the overall degree of protein labelling from tritiated
lysine was observed in all the cells of the spermatogenic cycle,
with the greatest grain count occurring over the pachytene primary
spermatocytes. During the two meiotic cell divisions, shown over
the secondary spermatocytes, a sharp decrease in the grain count
was observed. Whereas few grains were seen over the spermatids in
the control system, the addition of exogenous glucose was found to
produce a marked increase in the number of grains appearing over
the spermatids. These data indicate that of the various cell types
found in the seminiferous germinal epithelium of the rat, the pri-
mary spermatocytes demonstrate the highest degree of protein label-
ling from radioactive lysine. On the other hand, protein labelling
of the spermatids appears to have a great sensitivity to the addi-
tion of glucose. The data also suggest that glucose is capable of
exerting a pronounced stimulation of protein labelling in the pri-
mary spermatocytes whose nuclei are at pachynema and whose chromo-
somes are undergoing crossing-over. The possibility therefore ex-
EFFECT OF GWCOSE ON THE INCORPORATION OF L-LYSINE -u.C'~
A INTO PROTEIN OFTESTIS SLICES
...2
~
<II
:>-
£1 --.
.MUTUIU TUTI,
t'!'~
~~-- : ~
'" . .~ ·a·
'f
...j
~ ' 1Q,
0
""z
~ If'• # . '
8 lNTfltSnTw.-.cUL
...0
ADULT~
TlSfiCUlM'ft.III)ft TlIm
z
;;;
I
S
""0.
co
....IE
2
0.
'"
0
TESTIS CELL TESTIS
TESnCULAR
TUIOOII
8 ~
~ 7.0
o
~ 60
"
z
...
~~
cr ~ 5.0
~'"
qlc!
~e: 4.0
g~
d ...
~ ~ 3,0
zw
wI>
bf 2 .0 I NFLUENCE Of AGE ON THE
."
11' ..
CI)
o
r,O
EFFECT OF GLUCOSE ON THE
INCORPO RATION OF L"LYSINE·U·CJ4
INTO PROTEIN or RAT TESTIS SuCES
;::
: 0~-7.~~--~~~~~~~~
20 40 60 80 100 120
AGE OF ANIMAL IN OAYS
Fig. 4. Effect of glucose on the incorporation of L-lysine-U- 14C

into protein of testis slices. A. Slices of four differ-
ent testicular preparations, each with their own character-
istic histological structure, were incubated in the pre-
sence and absence of 0.009 M glucose for 1 hour at 37.So C.
Height of each bar represents the control specific activity
(cpm/mg dry protein) while the depth of each bar represents
percent stimulation of protein labelling by glucose. B.
Influence of age on the effect of 0.009 M glucose on pro-
tein labelling of slices of the normal rat testis.
w
'".....
CONTROl SYSTEM
GLUCOSE -SUPPLEMENTED
SYSTEM
~
v VII IX XIII ~
Fig. 5. Radioautographs of sections taken from slices of rat tes- o
»
tes incubated for 1 hour with tritiated lysine in the pre- <
(i;
sence and absence of 0.009 M glucose. 1 and 2 represent »
stage I; 3 and 4, stage V; 5 and 6, stage VII; 7 and 8, z
o
stage IX; 9 and 10, stage XIII. Sections were cut at 5 ~ ()
and stained with hematoxylin. Magnification ~pproximately :-n

x 1,000. (From: Davis, J. R. et al. Amer. J. Physiol., lQ2: :;0
...,....
425-432, 1965). =i
I ZO
~
::
:~
·: ~,
· ••
, ,
, I
I ,
0
z 90
, ,,
:>
••
•• •••
0
cr
..
""~
w
110
4 ~:
•
• ••~~\ •
:\ .. ~,'" i) 'tJ
~c
,: \ ,.c/ \. ': ,'~\ ~"\.
-8,
,
N
:
.~"
b'f!'! \
If,
~
1
\ :
GLUCOSE -SUPPLEMENTED SYSTEM
~ ~~
: ~:\ ~,'''d'\Q
'"z<i : "'' l,'Ip' '"
\t,
•
~:
I ,
....
c \ "n
cr
,/~<\ " ~ ~-! \
"z
0
\b'""'~" " \ ,:Q,.t1'
t :
\!
:
... ,.. ::
"" ~
~
>
" i \'0
zo
10 CONTROL SYS TEM
0
I: li .. S I , • , 10 • III I) loiII "11 II 1' 17 II . . "
m -iI 111-/% ;&,D D .Dr-V lZI -D II IIIr I·. 111'·11 l1li._ ID

I I
SPtlmo tOQorll O PtH'''O'w SDf'I'motocyt,. Seoandary Sp."nOllCII
Spt1'mO'oqolt
SUCCESSIVE CELLS Of' THE CYCLE Of' THE SEMINIFEROUS EPI THELIUM
Fig. 6. Effect of glucose on the incorporation of tritiated lysine

into protein of the successive cells of the seminiferous
epithelium of the rat testis. Each point on the curves
represents the average from 4 individual experiments, com-
prising 20 microscopic fields from each animal or a total
of 80 determinations for each of the designated cells of
the cycle of the seminiferous epithelium. Representative
standard errors for cells in the control system averaged
±3 grains/200~ 2 above background; that for cells in the
glucose-supplemented system averaged ± 7 grains/ 200 j..1. 2 I-
XIV refer to stages of the cycle of the seminiferous epi-
thelium of the rat according to the description of Leblond
and Clermont (7). Numbers 1-19 refer to spermatids at va-
rious steps of spermiogenesis. (From: Davis, J.R. et al.,
Amer. J. Physiol., ~: 425-432, 1965).
ists that glucose may be involved in the mechanisms of recombina-

tion of chromosomal segments during meiosis, thereby playing an
important role in the constantly changing patterns of the genome.
In addition, the data also suggest that the transition from mito-
sis to the first meiotic prophase is associated with a marked in-
crease in protein labelling from lysine.
EFFECT OF GLUCOSE ON HUMAN TESTICULAR PROTEIN SYNTHESIS
The question then arose as to whether the presently-described

pattern of protein labelling of the various successive cells of the
seminiferous germinal epithelium of the rat would be similar in sli-
ces of the human testis incubated with tritiated lysine. The pro-
blems of comparing experimental animal data with human data are well
known and the difficulties involved in performing dynamic biochemi-
cal studies on viable human testicular slices provide no exception.
We were fortunate to obtain large segments of normal testes of two
males aged 65 and 70 years which were removed as a therapeutic mea-
sure for prostatic carcinoma. Incubation of human testicular slices
with tritiated lysine was begun within 30 minutes of removal of the
testis at surgery. At no time prior to the beginning of the incu-
bation period was the temperature of the removed testis allowed to
rise above 4oC. Slices of the human testis, weighing approximately
250 mg, were therefore incubated for one hour under aerobic condi-
tions with 100 IJ-c of tritiated lysine both in the presence and ab-
sence of 0.009 M glucose in a manner identical to that for the rat.
At the end of the incubation period, slices of the human testis were
sectioned, dipped in molten Kodak nuclear track emulsion and expo-
sure carried out for five days. The slides were then developed,
stained in hematoxylin solution and the six stages of the cycle of
the seminiferous epithelium of the human testis identified as de-
scribed by Clermont (15).
Fig. 7 demonstrates representative radioautographs of sections

taken from slices of human testes incubated with tritiated lysine
in the presence and absence of exogenous glucose. A definite en-
hancement of protein labelling was observed over the spermatogonia,
primary spermatocytes and spermatids.
Fig. 8 presents a plot of the average grain count per 200~ 2

above background over each of the successive cells of the cycle of
the seminiferous germinal epithelium of the human testis one hour
after incubation with tritiated lysine both in the presence and ab-
sence of exogenous glucose. The pattern of protein labelling over
each of the cell associations of the human seminiferous germinal
epithelium is remarkably similar to that just described for the rat.
Not only do the primary spermatocytes most actively incorporate ly-
sine into protein, but there is also a rapid increase in protein
~
!!l
»
or-'"
R
m
"TI
"TI
m
()
-I
(f)
o
"TI
CONTROL SYSTEM C>

oz
»
o
o
-I
AI
o
""
Z
CI>
GLUCOSE-SUPPLEMENTED
SYSTEM
II III IV VI
Fig. 7. Radioautographs of sections taken from slices of human tes-
tes incubated for 1 hour with tritiated lysine in the pre-
sence and absence of 0.009 M glucose. 1 and 2 represent
stage II; 3 and 4, stage III; 5 and 6, stage IV; 7 and
8, stage VI. Sections were cut at 5 f.L and stained wi th w
t-.)
hematoxylin. Magnification approximately x 1000. ......
labelling following division of the spermatogonia into primary sper-

matocytes. A sharp decrease in protein labelling again occurred
over the secondary spermatocytes in metaphase which was even more
pronounced than that previously found in the rat. It is of interest
to note that the meiotic divisions occurring during the development
of Trillium anther have been reported to be associated with a fall
in protein sulfhydryl groups (16). The only difference in the pat-
tern of protein labelling in the successive cells of the seminifer-
o~s epithelium of the human testis as compared to the rat testis
appears to involve the degree of stimulation that glucose exerts
over the pachytene primary spermatocytes. The greater enhancement
of protein labelling by glucose in the pachytene primary spermato-
cytes occurring in the rat seems to be due to the presence of lower
control values of these cells in the rat. One possible explanation
may be that in the absence of adequate levels of glucose, the ear-
lier resting and leptotene primary spermatocytes have a higher ca-
pacity for protein synthesis than the later pachytene primary sper-
matocytes in younger animals as opposed to older animals. Experi-
ments involving the influence of age on the effect of glucose on
the radioautographic incorporation of tritiated lysine into the suc-
cessive cells of the cycle of the seminiferous epithelium of both
the rat and the human testis are now in progress in our laboratory.
The present studies have demonstrated characteristic quantita-

tive radioautographic patterns of protein labelling in the various
successive cells of the seminiferous germinal epithelium of both
the rat and human. Although similar quantitative patterns for nu-
cleic acid labelling are presently lacking for these two species,
Monesi (17,18) has investigated the radioautographic incorporation
of nucleic acid precursors in the mouse testis. DNA was found to
be synthesized during the resting stage of the primary spermatocytes
with RNA synthesis reaching a peak in middle pachytene, declining
during late pachytene and then ceasing completely during metaphase
and anaphase I and II. These observations on nucleic acid labelling
in the mouse testis seem to be consistent with our findings for pro-
tein labelling in the various successive cells of the seminiferous
germinal epithelium of both the rat and human testis. Any differen-
ces in protein biosynthesis of those cells of the spermatogenic cy-
cle responsible for cell proliferation and renewal as compared to
the more mature spermatids may have importance not only in investi-
gations dealing with regulation of male fertility but also in in-
vestigations dealing with potential side-effects of drugs that may
cause genetic damage with resulting congenital malformations. It
is our opinion that the presently-described radioautographic pat-
terns of protein labelling over each of the successive cells of the
seminiferous germinal epithelium may provide an extremely practical
initial screening procedure for the detection of any such possible
side-effects of drugs on spermatogenesis. Drugs which are commonly
administered to the male prior to the breeding performance may alter
the radioautographic pattern of protein labelling over the succes-
sive cells of the spermatogenic cycle in a more sensitive manner
EFFECT OF GLUCOSE ON THE INCORPORATION OF L-LYSINE-H 3 INTO

PROTEIN OF CELLS OF THE SEMINIFEROUS EPITHELIUM OF
HUMAN TESTIS SLICES
160
. .°-0I
150 0'
I I
140 I I
I
GLUCOSE- SUPPLEMENTED
0
z 130
SYSTEM I
:::> I I
0
,."
0:
120 I
I
I
I
•
C,)
110 I o
ell I I
/1
w 100 I I /1
> \ I I
0 / I
I
•
0_0
90
\
ell
.
N
I /
"- 80 I I
--
C> I I
N
70 I 1
VI
z I 1
:;(
0: 60 I
I
,
I
"z0 50 I
I
I
I ,
w
40 o o
"w•
0:
30
I
\ /
\
\
•
>
20
\ I
1/ .............---+--41
\
0- 0
10
0
AAAAAABBRLLzpppppn~~~ • • ~.~~
"---..---''--------v-----~ '---------.r--- ~
I-VI I-VI I-VI I-VI HI
I~ Spermatogonia )I( Primary 2o )I( )l(

Spermatids )1
Spermatocytes Spermatocytes
SUCCESSIVE CELLS OF THE CYCLE OF THE SEMINIFEROUS EPITHELIUM
Fig. 8. Effect of glucose on the incorporation of tritiated lysine

into protein of the successive cells of the cycle of the
seminiferous epithelium of the human testis. Each point
on the curves represents the average from 2 individual
experiments, comprising 10 microscopic fields from each
human testis or a total of 40 determinations for each of
the designated cells of the cycle of the seminiferous epi-
thelium. I-VI refer to stages of the cycle of the semini-
ferous epithelium of the human according to the descrip-
tion of Clermont (15). Sa-Sd refer to spermatids aOt va-
rious steps of spermiogenesis.
than a usual histological examination of the testis. Any drugs that

are found to influence initially the radioautographic pattern of
protein labelling of the successive cells of the seminiferous epi-
thelium could then be studied further employing the more laborious
technique of mating experiments with subsequent examination of off-
spring.
ACKNOWLEDGEMENT
This investigation was supported by U.S. Public Health Service
Research Grant HD-01573 from the National Institute of Child Health
and Human Development.
REFERENCES
1. Davis, J.R., Firlit, C.F. and Hollinger, M.A., Amer. J. Physiol.,
204: 696, 1963.
2. Davis, J.R., Morris, R.N. and Hollinger, M.A., Amer. J. Physiol.,
207: 50, 1964.
3. Davis, J.R. and Morris, R.N., Amer. J. Physiol., 205: 833, 1963.
4. Steinberger, E. and Nelson, W.O., Endocrinology, 56: 429, 1955.
5. Davis, J.R. and Firlit, C.F., Fertil. Steril., 17: 187, 1966.
6. Firlit, C.F. and Davis, J.R., J. Reprod. Fertil~ 11: 125, 1966.
7. Leblond, C.P. and Clermont, Y., Amer. J. Anat., 90~167, 1952.
8. Schuler, W., Helv. Chim. Acta, 12: 1796, 1944. --
9. Tepperman, J., Tepperman, H.M. and Dick, H.J., Endocrinology,
45: 491, 1949.
10. Paul, H.E., Paul, M.F., Kopko, F., Bender, R.C., and Everett,
G., Endocrinology, 53: 585, 1953.
11. Steinberger, E. and Wagner, C., Endocrinology, 69: 305, 1961.
12. Ewing, L.L., Baird, E.R. and VanDemark, N.L., Compo Biochem.
Physiol., 12: 455, 1966.
13. Mancine, R.E., Penhos, J.C., Izquierdo, I.A. and Heinrich, J.J.,
Proc. Soc. Exp. Biol. Med., 104: 699, 1960.
15. Clermont, Y., Amer. J. Anat., 112: 35, 1963-.-
16. Stern, H., J. Biophys. Biochem. Cytol., 4: 157, 1958.
17. Monesi, V., J. Cell Biol. 14: 1, 1962. -
18. Monesi, V., Exp. Cell Res. 39: 197, 1965.
DISCUSSION
BURGOS: In guinea pig testes, we have found by electron microscopy
that from the spermatogonia B to the primary spermatocyte in the
pachytene stage, there is a remarkable synthesis of cytoplasmic mem-
branes at the same stage at which the peak of maximal protein label-
ling occurs. The membranes start to appear at the leptotene stage
and reach their greatest development at the end of the pachytene stage.
CHRISTENSEN: Dr. Davis, I might suggest that you use perfusion to

administer the labelled amino acid to testes that have been removed
from human beings. I think you might obtain a more uniform adminis-
tration. On that first autoradiogram that you showed, it looked to

me as if the grains were evenly distributed over the nuclei and over
the cytoplasm. I wonder whether you are claiming nuclear protein
synthesis?
DAVIS: With regard to your first comment, we have previously shown

that the in vitro testicular slice technique for radioactive tracer
experiments allows for a uniform penetration of the isotope through-
out the entire seminiferous germinal epithelium. With regard to
your second comment, we have expressed our results as grains per
200~2 over the entire cell inasmuch as it was not possible to rule
out any cytoplasmic grains that might appear over areas of the nu-
cleus in 5~ sections.
CHRISTENSEN: In the autoradiograph you showed, many of the grains

were over the nuclei.
DAVIS: We are certainly keeping the possibility of a selective glu-

cose-stimulated nuclear protein synthesis in the seminiferous ger-
minal epithelium in mind and are presently designing experiments to
investigate this point.
MEANS: I have one comment to add to what Dr. Christensen said about
perfusion. We must realize that pool sizes are neglected in tissue
slices because we have many cut surfaces of cells. A perfusion might
be a very interesting approach to use if one is really trying to
categorize and look at the clinical type of phenomenology that Dr.
Davis was talking about.
JOST: I should like to ask Dr. Davis if the influence of glucose

on the synthesis of protein is insulin-dependent, and what happens
in diabetic patients?
DAVIS: We are certainly planning to study the possible influence

of insulin on the radioautographic incorporation of tritiated lysine
into the successive cells of the seminiferous germinal epithelium,
employing our in vitro testicular slice technique, both in the pre-
sence and absence of exogenous glucose. Dr. Mancini has indicated
that hypoglycemia will produce testicular atrophy and a number of
investigators have reported that male diabetics can have reduced
sperm counts. In addition, alloxan-diabetic rats have been shown
to have testicular atrophy. Until we determine whether the testis
is dependent upon in$ulin for glucose transport, all I would like
to state at the present time is that it is apparent that the intra-
cellular content of glucose is a most important regulator of protein
synthesis occurring in the seminiferous germinal epithelium. We are
currently quite interested in whether one possible explanation for
any damaging effects of diabetes on the testicular germinal epithe-
lium may involve the extremely sensitive regulation of testicular
protein synthesis by glucose.
332 J. R. DAVIS AND C. F. FIRlIT
MACLEOD: I would like to amend the statement made by Dr. Davis

that diabetics are infertile. We have found the contrary to be
true. We have not found any higher incidence of infertility amongst
diabetics than we have in any other population. It is true that
diabetics may become impotent, but that is quite another matter.
DAVIS: All I can say is that I have read several papers expressing
both viewpoints. I would like to suggest that the possibility of
male infertility in diabetics be more extensively studied in non-
treated cases but of course this would be extremely difficult to do.
In answer to Dr. MacLeod's comment, there have been experiments per-
formed in alloxan diabetic rats; and these have demonstrated destruc-
tion of the seminiferous germinal epithelium.
STUDIES OF SPERMATOGENESIS AND STEROID METABOLISM IN CULTURES OF
HUMAN TESTICULAR TISSUE
A. Steinberger, M. Ficher and E. Steinberger
Division of Endocrinology and Reproduction Research

Laboratories, Albert Einstein Medical Center,
Philadelphia, Pa. 19141
INTRODUCTION
Study of the physiology of any organ in vivo is complicated

by numerous interactions taking place within a living organism, and
attempts to identify factors controlling testicular functions have
been faced with a similar problem. Thus, while it has been estab-
lished that pituitary gonadotropins play an integral part in the
control of testicular physiology, the specific role of follicle
stimulating hormone (FSH) or luteinizing hormone (LH), as well as
the role of other factors in the spermatogenic and steroidogenic
functions of the testis have not been clearly defined.
We have sought to alleviate some shortcomings of the in vivo

systems by utilizing an in vitro culture method, which permits a
more rigid control of the environmental conditions and offers an
experimental model for the study of direct effects of various fac-
tors on testicular physiology. A culture method would seem parti-
cularly useful for the investigation of human testicular functions,
since besides simplifying the experimental model from a biological
viewpoint, it also alleviates many moral and legal problems connec-
ted with human experimentation in vivo. In view of the fact that
relatively small amounts of tissue are needed to set up a culture,
mUltiple cultures can be initiated from an average size testicular
biopsy.
This communication summarizes our observations on the mainten-

ance and differentiation of human testes grown as organ culture un-
der several in vitro conditions, as well as the capacity of the cul-
tured tissues to metabolize progesterone. Formation of steroid hor-
mones from acetate by fetal human testes grown as organ culture was
333
334 A. STEINBERGER, M. FICHER, AND E. STEINBERGER
reported recently (1).
a. Source of Tissue. Testicular tissues were obtained from adult

individuals undergoing diagnostic biopsy or therapeutic orchiectomy.
In one instance, testes from a 2-day old male were obtained at au-
topsy two hours following accidental death. Only testes with com-
plete spermatogenesis and ability to metabolize progesterone to
testosterone were utilized for these studies.
b. Culture Method. The method for organ culture of human testes

was similar to that used in our laboratory previously for culture
of testes from other mammalian species (2). Briefly, small tissue
fragments (1 mm 3 ) were grown on top of a millipore filter (0.45 ~
pore size) supported by a stainless steel wire grid at the liquid-
gas interphase. The culture medium was maintained at pH 6.9-7.1
and was replenished every 3 to 4 days. Penicillin (100 IU), strep-
tomycin (100 ~g) and fungizone (5 ~g) per ml of culture medium were
used to minimize the incidence of microbial contamination. The
various culture medium supplements, incubation temperatures and gas
phases used in an attempt to define optimal conditions for cultiva-
tion of human testes will be discussed in the Results section.
c. Histologic Evaluation. Fresh testicular tissue and tissues cul-

tured for various time intervals were fixed in Cleland's or Bouin's
fixative and processed by standard histologic methods. Four micron
sections were stained with periodic acid-Schiff followed by hema-
toxylin (PAS-H). The histologic appearance of the cultured tissues
was evaluated microscopically, using the following criteria: 1)
Maintenance of the overall tubular structure and the viability of
supporting cells. 2) Duration of the survival of specific cell
types in the germinal epithelium. 3) Degree of differentiation of
the germinal cells in vitro.
d. Radioautography. Tritiated thymidine (specific activity 6.7

Ci/mmole) was added aseptically to the culture medium in a final
concentration of 1 ~i/ml. After 40 minutes the 3H-thymidine con-
taining medium was replaced by "cold" medium. At various time in-
tervals following the pulse, several tissue fragments were fixed in
Bouin's fixative and processed by routine histologic methods •• Four
micron thick sections were stained with PAS and coated with liquid
Kodak NTB-2 emulsion. After 2 or 3 weeks of exposure in the dark
at 4 oC, the emulsion was developed and the sections stained with
hematoxylin. This procedure yielded better preparations than sec-
tions stained with PAS-H, either before or after development of the
emulsion. The location of grains over cell nuclei was determined
using a high magnification objective. A nucleus was considered
labelled when the density of grains over the nucleus was more than
two standard deviations above that of the background.
e. Incubation with 3H-progesterone. The incubation procedure was

similar to that described previously (3). Twenty to 80 mg. wet
weight of fresh or cultured testicular tissue was minced in a small
flask containing 1.5 ml of incubation medium and the substrate 3H_
progesterone (20 x 10 6 dpm, 3 mpmole). The incubation was carried
out in a Dubnoff shaker for 3 hours at 37 0 , in an atmosphere of 95%
air and 5% C02' After the incubation, 14C-Iabelled tracers (40,000
dpm of each) dissolved in 10% ethanol in benzene were added to the
incubate in order to correct for losses in subsequent manipulations.
At this point, 10 ml of ethyl acetate was also added and the mater-
ial was stored at -20 0 until further processing. The incubation
medium (1.5 ml/flask) had the following composition:
1 ml Hank's balanced salt solution
125 ~1 1.3% NaHC03
175 ~l NADP,Na solution (3.257 ~mole/ml)
175 ~l G-6-P,Na2 solution (36.17 ~mole/ml)
25 .~l e thano I
f. Extraction Procedure. Following repeated extractions with ethyl

acetate and methanol-hexane partition, the material was extracted
with chloroform. Appropriate cold carriers (50 ~g of each) were
added to the dry residue obtained from the chloroform extract.
g. Purification and Identification of Steroids. For the initial

separation of steroids,aBush A chromatograph system (heptane, me-
thanol, water - iOO:80:20) was utilized, and the radioactive peaks
were related to known steroid standards; the A4-3-ketosteroids were
located on the paper strips as dark zones under short-wave ultra-
violet illumination. The materials from the radioactive peaks of
the initial chromatogram were eluted and rechromatographed in a
Bush 3 system (heptane, benzene, methanol, water - 66:34:80:20).
Further identification of the steroids was accomplished by deriva-
tive formation and by the technique of isotopic dilution and re-
crystallization to constant specific activity or constant ratio
3H/14C•
h. Measurement of Radioactivity. Paper chromatograms were scanned

in a Packard model 7200 radiochromatogram scanner. Eluates from
paper chromatograms were dried, dissolved in 10 ml of scintillation
solution (0.4 percent w/v of 2,5-bis-[2'(5'-tert-butylbenzoxazolyl)]
-thiophene in toluene) and counted in a Packard liquid scintillation
spectrometer, model 3003, set for simultaneous counts of 3H and 14C•
Further details on counting efficiencies and purification of chemi-
cals used in these procedures were described previously (3). Cal-
culation of percent conversion from progesterone was based on the
following formula:
of 14C steroid added X 100

progesterone in incubation medium
RESULTS
A. Maintenance of Human Testes in Organ Culture
1) The effect of basic culture conditions
Prior to initiating studies on the effect of hormones or other

specific factors on mammalian testes cultured in vitro, basic en-
vironmental conditions had to be defined which would best support
the over-all tissue viability for prolonged periods of time. Ini-
tial studies on the effect of several culture conditions were car-
ried out, using organ cultures of rat testes (4,5,6). Conditions
found to be optimal for maintaining rat testes in vitro were later
utilized for culture of testes from several other mammalian species
(2). In an attempt to define optimal conditions for cultiyation of
human testes in vitro, the effects of incubation temperature, oxy-
gen tension and medium composition were re-investigated, using hu-
man testes.
a. Effect of incubation temperature: Multiple cultures of

adult human testes were grown at 310 , 33° or 35° for varying peri-
ods of time. Microscopic examination of stained histologic sec-
tions revealed that with all other conditions kept uniform, best
overall survival of germinal elements and differentiation of sper-
matogonia to late pachytene spermatocytes occurred in cultures
grown at 31°. At temperatures of 33° or 35°, the results were much
more variable and extensive degeneration of germ cells was often
observed in cultures grown at 35°. The incubation temperature of
31 0 was, therefore, considered to be optimal for cultivation of
human testes.
b. Effect of oxygen tension: To test the effect of higher

oxygen tension, cultures were incubated in a gas atmosphere com-
posed of 5% C02 and 95% air, or 5% C02 and 95% oxygen. The pre-
sence of 5% C02 in the gas phase was essential for the maintenance
of pH (6.9-7.1). The increased oxygen tension was found to be dam-
aging to the overall survival of testicular tissue and the survival
of germ cells in particular.
c. Effect of culture medium composition: Several commercially

available media were tested for their efficiency in supporting the
viability of human testes in organ culture. Minimum Essential Me-
dium (7) supplemented with 1 mM of Na pyruvate and 0.1 mM of each
of the amino acids alanine, asparagine, aspartic acid, glutamic acid,
glycine, proline and serine, appeared consistently somewhat superior
to other media tested and was, therefore, adapted for routine use.
In most cases, the medium was also supplemented with 10% calf serum,
since it appeared beneficial for maintenance of the tissues for per-
iods longer than four weeks. Until the fourth week of cultivation,
no difference was observed in the morphological appearance of tis-
sues grown in chemically defined media or media supplemented with

serum. The calf serum was consistently superior to homologous se-
rum. Additional supplements tested for enhancing effect on either
maintenance or differentiation of the germinal cells in culture will
be discussed below.
2) Sexually Mature Testes
A microscopic cross-section of a seminiferous tubule from a

sexually mature human testis is shown in Fig. 1. Several cell as-
sociations (stages of spermatogenesis) are present within a single
tubular cross-section, which is characteristic for the human testis
(8). In organ cultures grown under optimal conditions, spermatids
were present for approximately seven days of cultivation (Fig. 2),
and spermatocytes for appr9ximately three weeks of cultivation (Fig.
3)0 The basic tubular architecture was maintained and the Sertoli
cells and spermatogonia remained viable for at least seven weeks
(Fig. 4). Mitotic activity diminished gradually with time of cul-
tivation, but was still evident after seven weeks (Fig. 4). The
pale type A spermatogonia (Ap) survived in culture for longer per-
iods than the dark type A (Ad) spermatogonia. This results in an
increased ratio of Ap:Ad, which is close to 1.0 in normal tissues,
and may become as high as 4.6 after several weeks of cultivation(9).
3) Differentiation of premiotic germ cells in vitro
The presence of spermatocytes in cultures for periods up to

four weeks raised the question whether these cells were formed by
in vitro differentiation of less mature germ cells or represented
cells which were present in the tissues prior to cultivation. This
question was resolved by using radioautography techniques following
pulse-labelling of cell nuclei with 3H-thymidine.
It has been shown (10) that one hour following injection of

3H- t hymidine into human testes in vivo, the most advanced germ cells
which incorporate the precursor into nuclear DNA are the early lep-
totene or pre-leptotene spermatocytes. It was, therefore, possible
to follow the development of these cells into more mature forms by
obtaining testicular biopsies at time intervals and examining them
following radioautography.
We utilized a similar approach to determine the fate of pre-

leptotene spermatocytes in culture. The initial labelling was ac-
complished in vitro b~ growing the cultures for the first 40 minutes
in medium containing H-thymidine. At various time intervals fol-
lowing the pulse, several tissue fragments were fixed and processed
for radioautography. After 40 minutes exposure to 3H-thymidine,
the most advanced germ cells that were labelled were pre-leptotene
spermatocytes. After seven days of cultivation, the label was as-
sociated with leptotene or zygotene spermatocytes, and after two to
Fig. 1. Cross-section of a seminiferous tubule from a sexually

mature human testis. The germ cells are present in several specific
associations (stages of spermatogenesis) Ap-pale type A spermatogo-
nia. Ad-dark type A spermatogonia. Sc-spermatocytes. St-sperma-
tids. S-Sertoli cells. Magn. X 480.
Fig. 2. Microscopic appearance of a sexually mature human testis

after 7 days of cultivation. Magn. X 480.
Fig. 3. Microscopic appearance of a sexually mature testis after

21 days of cultivation. Note presence of mitotic figures (M) and
absence of spermatids. Magn. X 480.
Fig. 4. Microscopic appearance of a sexually mature testis after

48 days of cultivation. The only germ cells remaining are sperma-
togonia. Magn. X 480.
three weeks of cultivation, with spermatocytes in the late pachy-

tene stage of miosis. Differentiation of germ cells was observed
only when either vitamins A (12 u ), C (50 I-1g) and E (60 I-Lg) per ml,
or 4 mM of glutamine were present in the culture medium. Addition
of various concentrations of testosterone or gonadotropic hormones
(HCG, NIH-LH-B5, HIH-FSH-S-1, PMS)to the culture medium did not re-
sult in improved maintenance or stimulation of differentiation of
the germ cells. In fact, high concentrations of the gonadotropins
appeared toxic and produced pronounced vacuolation of Sertoli cell
cytoplasm and scattered non-specific necrosis.
These results show unequivocally that in human testes differen-

tiation of pre-leptotene spermatocytes through the leptotene, zygo-
tene and pachytene stages of the miotic prophase takes place under
appropriate culture conditions in the absence of gonadotropic hor-
mones. Under no culture conditions tested could we maintain the
entire spermatogenic process or show progression of differentiation
beyond the late pachytene stage of spermatocytes. These findings
are in agreement with our earlier observations on organ cultures of
several other mammalian species (2).
4) Sexually immature testes (2-day old male)
The histologic appearance of normal testes from a 2-day old

male, and of similar tissue following 28 days of growth in organ
culture, are shown in Figs. 5 and 6, respectively. The cultures
were grown at 31 0 C in an atmosphere of 5% C02and 95% air, in a me-
dium containing a 4 mM concentration of glutamine. It is apparent
that the tubular structure, the supporting cells and the germinal
cells were well maintained under these culture conditions, with
some evidence of morphologic differentiation. The lumen of the tu-
bules became somewhat larger and the number of germ cells increased.
Since the human testes do not become sexually mature for about 13
years, we obviously did not expect to observe drastic differentia-
tion in vitro in a matter of weeks.
B. Metabolism of Progesterone in Fresh and Cultured Human Testes
Although the qualitative results obtained in three separate

experiments, using tissue from three different adult individuals,
were similar, there were considerable quantitative differences in
the amounts of converted progesterone and the amounts of the result-
ing metabolites. For this reason, the results obtained in each ex-
periment (Exp. A,B,C) are shown in Tables 1-4. Tracings of the
radioscans from the first chromatograms (Bush A system) obtained
from fresh tissue and tissue cultured for seven days (Exp. A) are
shown in Figs. 7 and 8, respectively. Both radioscans showed a
major radioactive peak of highly polar material at the origin and
four additional peaks of radioactivity with chromatographic mobili-
ties corresponding to the following steroid standards: Testosterone
Fig. 5. Section of testis from a 2-day old human male. The semin-
iferous tubules contain only Sertoli cells (S) and undifferentiated
germ cells (G). Magn. X 480.
Fig. 6. Microscopic appearance of testis from 2-day old human male

after 28 days of cultivation. The seminiferous tubules appear more
differentiated. Magn. X 480.
w
.....
.....,
TABLE 1. ISOTOPIC DILUTION AND RECRYSTALLIZATION OF TESTOSTERONE ACETATE TO CONSTANT 3H/14C RATIO
Days in Culture
EXE. A EXE. B Exe. C
FRACTION 0 7 14 0 14 0 3 31
Crystallization #1 xis a 24c 3.7 c 3.OC 9.3c 0.503c 41c 3.9C .580
methanol/water MLb 27 7.4c 6.1 4.0 0.413 15 1.6 2.426
Crystallization #2 xis 23 c 3.7 c 2.9 c 9.0 c 0.445 c 41c 3.9c 0.5le c '!>
en
methanol/water ML 24 c 4.8 3.2c 8.5 0.603 26 3.4 1.660 -I
m
Z
Crystallization #3 xis 24c 3.8 c 2.9 c 9.2c 0.50S c 43 c 3.7 c 0.540c m
""
.0
methanol/water ML 24c 3.7 c 3.0 c 9.7 c 0.539 c 39 c 5.2 0.790 (;)
m
,.0
~
23.80 3.73 3.00 9.30 0.516 41 3.8 .525 -n
MEAN + S.E. ()
+0.20 +0.03 +0.06 +0.15 +0.012 +0.8 +0.07 +0.015 :r:
m
?"
»
a - crystals Z
0
b - mothzr liquor rn
c - 3H/1 C ratios used for calculation of the mean + S.E. en
-I
m
Z
""m
.0
(;)
m
.0
~
~
OJ
or-
n
m
TABLE 2. ISOTOPIC DILUTION AND RECRYSTALLIZATION OF 17u-HYDROXYPROGESTERONE TO CONSTANT 3H/ 14C RATIO ."
."
m
()
--I
en
Days in cui ture o
."
G)
Exp. A Exp. B Exp. C oz
»
o
FRACTION 0 7 14 0 14 0 3 31 o
--I
::<:1
33 c 23 c 165 c 20 c 48 c 7.4c 56 c
o
."
Crystallization #1 xls a 110 c
methanol/water MLb 79 44 27 158 18 c 44 c 13.3 45 Z
en
Crystallization #2 xis 110 c 33 c 21 c 172 c 20 c 47 c 7.4 c 52 c

methanol/water ML 103 c 33 c 23 c 172 c 20 c 45 c 8.2 c 56 c
Crystallization #3 xis 106 c 32c 24 c 174 c 20 c 47 c 8.4 c SSG

methanol/water ML 106 c 34c 22 c 167 c 19 c 47 c 7.4 c 57 c
107.00 33.00 22.60 170.00 19.50 46.30 7.8 55.60

MEAN + S.E. +0.34 +0.22 +0.88
±1.30 ±0.32 ±0.51 ±1.70 +0.60
a - crystals
b - mother liquor
c - 3H/ 14C ratios used for calculation of the mean + S.E.
w
",..
w
t
TABLE 3. ISOTOPIC DILUTION AND RECRYSTALLIZATION OF 2Ott-DIHYDROPROGESTERONE
TO CONSTANT SPECIFIC ACTIVITY (dpm/ I.L Mole)
Days in Culture
EXE. A EXE. B EXE- C
FRACTION 0 7 14 0 14 0 3 31
Crystallization #1 xlsa 443 c 1,139 c 1,076 c 380 c 95 320 c -d 1,430 c

methanol/water MLb 601 2,184 1,392 475 76 c 1,520 1,980
Crystallization #2 xIs 443 c 1,139 c 1,044 c 380 c 70 c 310 c 1,540 '?»

(II
methanol/water ML 475 c 1,266 1,044 c 443 79 c 320 c 1,460 c m
.....
Z
Crystallization #3 xIs 443 c 1,155 c 886 348 c 70 c 310 c 1,310 c ""m::1:1
methanol/water ML 443 c 1.202 c 1,170 380 c 57 320 c 1,490 c G'm>
,::1:1
449 1,159 1,055 372 73.8 316 1,422

MEAN ± S.E. +6.4 +15.0 +10.7 +8.0 +40
...i"i~
+2.3 +2.5 :::t:
m
,::1:1
a - crystals »
b - mother liquor Z
1:1
c - specific activities used for calculation of the mean ± S.E. !1'
(II
d - not done ;ri
z
m""
Q
m
::1:1
TABLE 4. PERCENT CONVERSION OF PROGESTERONE TO KNOWN METABOLITES
Daxs in Culture
EXE· A EXE· B EXE· C
METABOLITES 0 7 14 0 14 0 3 31
Progesterone a 13 68 61 29 84 7 64 72
(unconverted substrate)
Testosterone b 4.5 0.7 0.5 2.5 0.14 10 0.9 0.14
17~hydroxyprogesteroneb 17 5 4 38 4.5 9 1.6 12
20~dihydroprogesteronea 7 6 10 5 3 3 6 6
Androstenedione a 3 2 2 1 0.4 2 1 0.5
a - calculated on basis of recovered radioactivity in the first

chromatogram (Bush A)
b - calculated on basis of recrystallization to constant 3H/ 14C
ratio
and 17a-hydroxyprogesterone, 20U-dihydroprogesterone, androstene-

dione and progesterone.
It is evident that following seven days of cultivation, the

amount of unconverted progesterone increased considerably, while
the amount of radioactivity in the area corresponding to testoster-
one and 17a-hydroxyprogesterone showed a three-fold decrease. The
peaks in the zones of 2Da-dihydroprogesterone and androstenedione
showed no significant change. The radioscans obtained from the
cultured tissues also showed an additional peak of radioactivity
in the area of less polar material than progesterone. This material
has not been identified. The change observed in the first chromato-
gram after seven days of cultivation were slightly emphasized with
additional time of cultivation.
The material from the radioactive peak corresponding in mobil-

ity to testosterone and 17a-hydroxyprogesterone was eluted and re-
chromatographed in a Bush 3 system following acetylation Two peaks
of radioactivity were formed; one with an Rf value similar to that
of 17a-hydroxyprogesterone, and another peak with the Rf value of
testosterone acetate. Final identification of the materials in
these two peaks was obtained by isotopic dilution and recrystalliza-
tion to const~nt 3H/ 14C ratio (Tables 1 and 2).
The materials migrating in the first chromatogram, like andro-

stenedione and 20a-dihydroprogesterone standards, respectively,
yielded single peaks upon rechromatography in the Bush 3 system.
The chemical purity of 20a.dihydroprogesterone was further asserted
by recrystallization to constant specific activity (Table 3).
w
f>.
?>
RADIOSCAN VI
-l
m
C BUSH A ) Z
00
m
;<0
G)
m
,;<0
%RECOVERED
33 2 7 3 13 ~
RAD'OACTIVITY 42 ~
"
n
STANDARD :J:
m
(fOr~ <:) ,;<0
STRIP
I G;~~ OJ -I >
Z
o
Fig. 7. Radioscan of the first chromatogram (Bush A obtained from incubate of fresh testicular rn
tissue (adult human) with 3H-progesterone in Exp. A. VI
-l
m
Z
00
m
;<0
G)
m
;<0
~
~
o
"",...
R
m
-n
-n
~
-I
(1'1
o
-n
(;)
oz
»
o
o-I
;;0
o
"1:J
Z
(1'1
RADloaCAN
( BUSH A )
'" RECOVER.ED
8 0 II 6 2 68 4
RADIOACTIVITY .. m
STANDARD
IG·~) g- 0 - ~
STRIP
GJ-
Fig. 8. Radioscan of the first chromatogram (Bush A) obtained from incubate of similar tissue
with 3H-progesterone after 7 days of cultivation in Exp. A.
w
~
The percent conversion of progesterone to several known meta-

bolites obtained in all three experiments and the percentages of
radioactivity associated with unconverted substrate are summarized
in Table 4. The percent conversion of progesterone to testosterone
and 17a-hydroxyprogesterone, respectively, are based on the results
of recrystallizations to constant 3H/14C ratio using 14C labelled
carriers. The percent conversion of progesterone to 20~dihydropro
gesterone and to androstenedione was calculated on the basis of re-
covered radioactivity in the first chromatograms, since repeated
chromatographies and recrystallizations to constant specific activi-
ties did not reveal any impurities. It is evident that the amount
of progesterone converted to testosterone drops significantly with-
in the first three days of cultivation, with a further small drop
occurring upon additional culture period. The capacity of testi-
cular tissue maintained in organ culture to convert progesterone to
17a-hydroxyprogesterone also appears to diminish, although not as
drastically as the conversion of progesterone to testosterone. In
one experiment (exp. C), the percent conversion of progesterone to
17a-hydroxyprogesterone diminished after three days of cultivation,
but after 31 days of cultivation it was somewhat higher than in
fresh tissue (Table 4).
DISCUSSION
While in vitro conditions, which would successfully maintain

the entire spermatogenic process in the mammalian testes, have not
been defined, some progress has been made in that direction. In
organ culture of human testes, the overall tubular structure and vi-
ability of Sertoli cells and spermatogonia were maintained for at
least seven weeks. The spermatids and spermatocytes survived for
approximately seven days and 28 days, respectively. It was shown
with the aid of radioautography that resting primary spermatocytes
differentiate through the leptotene, zygotene and pachytene stages
of the miotic prophase, in the absence of gonadotropins or testos-
terone, providing the culture medium contains vitamins A, C and E
or a 4 mM concentration of glutamine. Addition of the hormones had
no effect on either maintenance or differentiation of human testes
in vitro. These results are generally similar to the results ob-
tained with cultures of testes from several other mammalian species
(2) •
However, a unique characteristic of human testes is the reten-

tion of morphologic integrity of type A spermatogonia throughout the
entire culture period, while in testes from other mammalian species
the type A spermatogonia, after several days of cultivation, are re-
placed by primitive type A spermatogonia (11). The Ap spermatogo-
nia survive in vitro for longer periods of time than the Ad sperma-
togonia, causing the Ap:Ad ratio to increase 3 to 5 fold within sev-
eral weeks of cultivation. The significance of these observations
is not clear.
The steroid studies indicate that the capacity of human tes-

ticular tissue to convert progesterone to l~-hydroxyprogesterone
and testosterone diminishes drastically within the first three days
of cultivation, with only a slight further decrease occurring upon
additional cultivation time. 20l-dihydroprogesterone and andro-
stenedione were formed from progesterone by the cultured tissues,
as well as by the fresh tissues. Since the major change in the pat-
tern of progesterone metabolism occurs in culture rather rapidly,
it could be due to the sudden removal of gonadotropic stimulation
rather than a gradual loss of tissue viability. If the latter were
the case, one would expect the changes to be more directly related
to the time of cultivation. The decrease in the activity of 17a-hy-
droxylase, and particularly the decrease of 17-20 lyase activity
could very well be a reflection of lack of gonadotropins, since dim-
inution of the activity of these two enzymes was found in testicular
tissue of human males with suppressed pituitary function (12).
Major metabolites, including testosterone, formed from pro-

gesterone in incubates of fresh tissue, were also formed by tissues
cultured for up to 31 days. This would imply that all the neces-
sary enzymes in the progesterone to testosterone metabolic pathway
must have been functional under the culture conditions employed.
Whether the quantitative relationships of progesterone metabolism
can be affected by adding gonadotropic hormones to the cultures is
presently being investigated.
Formation of a number of steroids, including testosterone from

14C-acetate, was recently demonstrated in organ culture of human
fetal testis (1). It is interesting that the amounts of testoster-
one and androstenedione formed by these cultures were also very
small and were not significantly affected by addition of HCG to the
culture medium.
It appears that organ culture may be a useful experimental

model for the study of steroid biosynthesis and metabolism in testes
exposed to various environmental conditions.
ACKNOWLEDGEMENTS
This study was supported, in part, by u.S. Public Health Ser-

vice Grant No. HD 00399 and by a grant from the Ford Foundation. The
excellent technical assistance of A. Gdowik, o. Luckyj, R. Ham and
I. Haltrecht is gratefully acknowledged.
The folloWing trivial names and abbreviations are used in this

paper:
Androstenedione (A) 6 4-Androstene-3, 17-dione
17a-Hydroxyprogesterone (17-0H-P) 6 4 -Pregnene-1Ja-ol-3,20-dione
20a-Dihydroprogesterone (200-DP) 6 4 -Pregnene-20a-ol-3-one
Progesterone (p) 6 4 -Pregnene-3,20-dione

Testosterone (T) 64-Androstene-17 ~-ol-3-one
NADP,Na Nicotinamide adenine dinucleo-
tide phosphate, sodium salt
G-6-P, Na2 Glucose-6-phosphate, sodium salt
REFERENCES
1. Rice, B.F., Johanson, C.A. and Sternberg, W.R., Steroids, 2: 1,

1966.
2. Steinberger, A. and Steinberger, E., J. Reprod. Fertil. Suppl.,
2, 117: 1967.
3. Steinberger, E. and Ficher, M., Steroids, 11: 3, 1968.
4. Steinberger, A., Steinberger, E. and Perloff, W.R., Exp. Cell
Res., 12.: 19, 1964.
5. Steinberger, E., Steinberger, A. and Perloff, W.R., Endocrin-
ology, 74: 5, 1964.
6. Steinberger, A. and Steinberger, E., Exp. Cell Res., 44: 429,
1966.
7 Eagle, R., Science, 130: 432, 1959.
8. Clermont, Y., FertU-:-8terU., 17: 6, 1966.
9. Chowdhury, A.K., Steinberger, A:-and Steinberger, E., Unpub-
lished data.
10. Reller, C.G., Matson, L.J., Moore, D.J. and Clermont, Y.,
Procedures International Symposium on Effects of Ionizing Radi-
ation in the Reproductive System, Pergamon Press, New York
1963. '
11. Steinberger, E., Steinberger, A. and Perloff, W.R., Anat. Rec.,
148: 581, 1964.
12. Tamaoki, B. and Shikita, M., Steroid Dynamics, eds. Pincus, G.,
Nakao, T. and Tait, J.F., Academic Press, p. 493, 1966.
DISCUSSION
DICZFALUSY: Dr. Steinberger, do you have any information on the

chemical nature of the metabolite which was less polar than proges-
terone? Could this possibly be a protein-bound form of progesterone?
Have you done or do you intend to repeat your beautiful incubations
with acetate instead of progesterone? The reason for asking this
question is that in fetal perfusions using progesterone we found ex-
tremely limited incorporation of progesterone into testicular tissue.
Using acetate we found that we could isolate as much as 10 to 15% of
the total testicular radioactive material as cholesterol. Further-
more, when cholesterol and acetate were used together it was found
that some 100 times more acetate was incorporated into the esterified
cholesterol fraction than into unesterified cholesterol. A similar
relationship was found in the fetal adrena1s where much more acetate
was incorporated into different conjugated steroids than into uncon-
jugated forms. What I would like to find out is whether this is a
particular situation in the fetal glands or does it also apply to

the adult glands?
A. STEINBERGER: We have not identified the material, which is less

polar than progesterone. The answer to the second question is that
we are indeed planning to use acetate as a precursor in our incuba-
tions to test de novo synthesis of testosterone. The progesterone
was used as a preliminary experiment. In this connection I would
also like to quote a paper by Rice et ale (Steroids, l: 79, 1966)
who showed that acetate incubated with organ cultures of human fetal
testis was incorporated into several steroids and I believe among
others into testosterone and l7-hydroxyprogesterone.
MACLEOD: Dr. Steinberger mentioned that above the oxygen tension of

air they did not obtain good culture development. I wonder if Dr.
Steinberger would tell us under what experimental conditions this
was found. Has the catalase activity of the various generations of
germinal epithelium ever been measured? This question pertains to
the observation we made about 25 years ago, namely that mature human
spermatozoa are extremely sensitive to high oxygen tensions, and in-
deed extremely sensitive to any oxidations produced in their pre-
sence.
A. STEINBERGER: We found that in cultures grown by the usual method

but under high oxygen tension (5% carbon dioxide and 95% oxygen)
the germinal cells survive for a much shorter period of time, and
do nat show the progressive differentiation in the presence of glu-
tamine or vitamins A, E and C. These observations are similar to
our previous findings with cultures of marranalian testis of other
species. Whether certain germ cells are specifically more affected
by the oxygen I am not able to say, but the overall survival of the
germ cells was impaired in higher oxygen tensions, and this is why
we prefer using a gas phase composed of 95% air and 5% carbon dioxide.
The carbon dioxide, incidentally, is necessary for the proper buf-
fering capacity of the bicarbonate system used in the culture medium.
FORD: I understood Dr. Steinberger to say that in her organ cultures

the spermatocytes developed up as far as the late pachytene stage.
What total span of spermatocyte development would be covered? Would
it be from zygotene to late pachytene or leptotene through to late
pachytene?
A. STEINBERGER: The radioautography experiment showed definitely

development of preleptotene spermatocytes into late pachytene sper-
matocytes. Possibly some spermatogonia also proceed into the late
pachytene stage, but in this experimental design, this would be more
difficult to assess since we were primarily following the most ad-
vanced labelled cells.
TROEN: I should like to report on some findings obtained by Dr.

Takumi Yanaihara and myself in an organ culture study of steroid bio-
synthesis in testicular tissue of infertile males. Tissue obtained

at diagnostic testicular biopsy from nine patients with oligospermia
was studied in 24 to 48 hour organ cultures, using 3.8 to 6 mg and
in one instance, 20 mg of tissue. Tritium and C14 labelled precur-
sors were used in varying combinations and the metabolites identi-
fied and isotopic ratios determined. ~ 5-androstenediol and testos-
terone, and to a smaller extent,~ 4-androstenedione were identif~ed
from precursor dihydroepiandrosterone. With mixed precursors,~ -
androstenediol and ~4-androstenedione there was greater conversion
to testosterone from ~4-androstenedione. From mixed precursors, C14 _
l7-hydroxyprogesterone with tritiated l7-hydroxypregnenolone or C14_
pregnenolone with tritiated DHA or C14_pregnenolone with tritiated
l7-hydroxypregnenolone, there was a consistently greater tritium!C 14
ratio in testosterone than in~4-androstenedione. The radioactive
intermediates in both ~4 and A5 pathways were identified. It ap-
pears that DRA and 17-hydroxypregnenolone favor the ~5 pathway to
testosterone more readily than does pregnenolone. Dr. Tominaga and
I conducted experiments wi~h adult human testis tissue in organ cul-
tures for four days under hyperbaric oxygen, and have found that
mitosis persists in the germinal epithelium and steroid biosynthesis
continues or is even increased.
MANN: Among the interesting findings on which you reported, Dr.

Steinberger, is the observation that organ cultures can maintain
steroidogenesis during prolonged incubation periods, and as you
qUite rightly pointed out, that must mean that all the enzymes re-
quired for steroidogenesis have been preserved intact. However, you
appear to be somewhat worried about the decreasing rates of steroid-
ogenesis, particularly the rate of testosterone formation. I wonder
if you are giving the best possible chance to all the enzymes to act
under optimal conditions. Perhaps one might look into the possibil-
ity that, although the enzymes are peeserved excellently, the co-
enzymes which they normally require have disappeared. I mean, for
example, the diphosphopyridine nucleotide and the triphosphopyridine
nucleotide. Would it not be worthwhile to look into the possibility
that these co-enzymes are in fact, decreasing during incubation and,
if so, one might perhaps add them to the organ cultures.
CHANDLEY: Dr. Steinberger, in your labelling studies, have you

found any incorporation of tritiated thymidine during the leptotene!
zygotene stage of meiosis? This has been claimed by Lima de Faria
in human spermatocytes. This would mean looking at cells within
one hour or half an hour of labelling, and seeing if there was any
incorporation at this stage, in addition to the incorporation that
is occurring at the premeiotic interphase or preleptotene stage.
A. STEINBERGER, We have not observed labelled leptotene or zygo-

tene spermatocytes in sections of human testes following a short 40
minute pulse with 3.H thymidine. The label was limited to the pre-
leptotene stages. Similar results were obtained with other mammalian
species as well.
CHAPTER IV

SECTION 5: INFLUENCE OF GONADOTROPINS ON
TESTICULAR FUNCTION
EFFECT OF GONADOTROPINS ON THE SEMINIFEROUS TUBULES OF THE IMMATURE
TESTIS
M. Courot
I.N.R.A. - Station de Physiologie de la Reproduction,

37 - Nouzilly, France
INTRODUCTION
It is important first to define the "immature testis", at least

with respect to the spermatogenetic function of the organ. We know
that an "adult" or "mature" testis is that in which complete sper-
matogenesis occurs, with a satisfactory yield (although this may be
related to the time of year or to the environment in which individ-
uals are placed). This brings to light the characteristics of the
immature testis; namely, a non-existent or incomplete spermatogenesis
or a spermatogenesis not yet having attained normal cell yield.
Thus there are several types of immature testes, depending on the
degree of their spermatogenetic activity. A terminology has just
been proposed (1) in which the following terms are used to describe
the immature testis:
Impuberal testis type: the sex cords, future seminiferous

tubules do not yet have a lumen and have no spermatogenetic activity.
The germ cells present are gonocytes or stem spermatogonia; they
divide by mitosis, but do not differentiate into primary spermato-
cytes. The somatic cells are Sertoli cell precursors, the supporting
cells o
Prepuberal testis type: The germ cells evolve and different

stages leading to the formation of the first spermatozoa may be
observed as the Sertoli cells differentiate from the supporting
cells. Puberty terminates this phase, i.e., the first spermatozoa
are produced, but the testis is not yet mature because the cellular
yield of the spermatogenetic process is very low.
Postpuberal testis type: during this stage, the cellular yield

355
356 M.COUROT
of the spermatogenetic processes increases, and the quantity of

spermatozoa produced increases to become progressively comparable
to that of the adult.
All of these stages constitute the immature phase of the testis,

and we shall study the effect of gonadotropins on the immature semi-
niferous tubules thus defined. Since the literature available for
the human generally concerns pathological cases in which organ
receptivity may be changed, we shall also use the experimental re-
sults obtained on mammals for our analysis and then compare them to
each other.
HYPOPHYSECTOMY
It is known that all stages of testis development, even the

earliest ones, are regulated by the pituitary gland. After hypo-
physectomy, the weight of the testis in operated animals decreases
in the impuberal (monkey (2); lamb (3»,prepuberal and postpuberal
phases (rat (4». Likewise, early pituitary imbalance of accidental
or pathological origin prevents, or retards, testis development and
onset of spermatogenesis in man (5-11). These effects concern both
the germinal and endocrine testis functions, but we shall only dis-
cuss the former here. Hypophysectomy always involves a decrease in
seminiferous tubule diameter, but its effect on the tubule cell
population depends on the period in which it is performed.
In the impuberal period (Table 1) in the lamb, hypophysectomy

mainly affects the somatic cells (supporting cells) which are the
precursors of the Sertoli cells. Their number rapidly decreases
(from day 5 onwards) and then remains relatively constant, the cell
appearance remaining that of undifferentiated cells. On the other
hand, germ cells seem less sensitive to pituitary endocrine regula-
tion. Their number is unmodified by hypophysectomy, and it even
continues to increase after the pituitary has been removed (Table 1),
although less rapidly than in normal animals. However, the evolu-
tion of these cells is affected because they do not differentiate
into primary spermatocytes (3). It is not possible to perform the
equivalent experiment in man, but several pathological cases have
been described. When the surgical operation is done at a young age,
it prevents germ cell differentiation and spermatogenesis will not
be established (11).
In the prepuberal period (Table 4); the number of Sertoli cells

decreases in the rat testis following hypophysectomy. The somatic
cells in the tubules are thus also under pituitary control, as they
could be in the adult (8,12-14). The development of germ cells is
greatly affected, the elongated spermatids disappear, the young
spermatids quantitatively considerably decrease and in most instances
are said to disappear. The number of primary spermatocytes also
diminishes (only 12% of leptotene spermatocytes are formed in the
INFLUENCE OF GONADOTROPINS ON TESTICULAR FUNCTION 357
Table 1. Effect of Hypophysectomy on the Cellular Content of the

Sex Cords in the Impuberal Lamb Testis.
Initial Delay After Surgery (days)

Control----3~~~~6-=~~1~5~~~~3~0~~--~2~5~0~
No. of Animals 19 7 15 22 9 5
Testis Weight 3.4+ 2.9 2.3 1.9 1.9 2.1
(g) (0.13) (0.40) (0.15) (0.13) (0.14) (0.30)
Supporting Cells/ 2540 2550 1930 1680 1540 1600
testis x 10 6 (134) (333) ( 170) ( 147) ( 130) ( 252)
Spermatogonial
37 49 48 53 55 76
Cells/
testis x 106 (35) (72) (3.3) (6.0) (5.0) ( 18.0)
+ m (± sm) - 6 days and more after surgery, experimental data and

initial control are significantly different (P';; 0.05)
rat 48 days after hypophysectomy performed at 27 days). Nothing

precise has been published on stem spermatogonia in the impuberal
testis, but it may be that their number decreases as observed in
the adult (13,15). This point should be carefully studied in the
light of recent studies on stem spermatogonia (16) to determine if
the renewing stem cells, AI, and the reserve stem cells, AO' are
equally affected. Could there not be reserve stem cells not destroyed
by hypophysectomy as is the case for the ge~ cells of the impuberal
testis, the only affected cells being the renewing stem cells?
GONADOTROPIC SUPPLEMENTATION IN THE IMPUBERAL MALE
To our knowledge, only the monkey (2) and the lamb (3) have been
the objects of analytical experiments on endocrine control of the
impuberal testis, but since in the work on the monkey only slightly
purified products were used, we will not present these results. In
man, cases of impuberal testis have been studied in boys showing gen-
ital infantilism (5-10), treated after cranial surgery (11,17) or
treated for unilateral cryptorchidism (11). In the last case, we
shall limit the discussion to the normal testis.
In the lamb, spermatogenesis commences several months after birth

thus leaving time for experimentation on a testis not yet presenting
any spermatogenetic activity; a situation which does not exist in
laboratory animals such as the rat or the mouse. Using hypophysec-
tomized lambs, supplemented immediately after the surgical operation
twice a day for 5 to 60 days,we have shown that ovine acetonic ante-
rior pituitary powder helps in the maintenance of testis growth asso-
ciated with the growth of the sex cords,with an effect on their diam-
eter as well as on their somatic and germinal cell populations. Germ
cells even differentiate into primary spermatocytes. However, it
358 M.COUROT
should be noted that there is a wide individual variability of re-

sponse. In the most favorable cases, there is a strong stimulation
with a high testicular weight increase and the establishment of
spermatogenesis in the whole organ. Spermatids have even been ob-
tained in the maturation phase, that is, in the ultimate phase of
their formation in the seminiferous epithelium (3). In this case,
cellular associations observed in the seminiferous epithelium re-
semble those described for the adult. This indicate~ that spermato -
genesis takes place at the same rhythm as that in the normal individ-
ual. The constant speed of the spermatogenetic process has already
been demonstrated in different conditions of hormonal balance (18,
19). In only one instance has this concept been questioned because
of slight variations (20). The results obtained with impuberal
testes are logical. There is no acceleration of s~ermatogenesis,
but rather an early beginning of spermatogenetic activity due to
pituitary hormones. Precocity and acceleration are two very differ-
ent things, and only the first one may be obtained for spermato-
genesis before it starts.
The search for active factors in the anterior pituitary prepara-

tion led to separate supplementation with ovine FSH or LH (ICSH)
(Table 2). It was thus observed that LH has the most clearly measur-
able action on the testis of the hypophysectomized lamb, i.e. organ
weight increase and enlargement of sex cord diameter. However, one
of the two gonadotropins maintains the supporting cell reserve but
the two hormones seemed separately to have no clear quantitative
effect on the germ cells (same levels as the hypophysectomized con~
troIs, Table 2). Nevertheless, they also allowed the differentiation
of some spermatogonial cells into primary spermatocytes as did the
anterior pituitary powder but to a lesser degree. This important
action of the two gonadotropins on the onset of spermatogenesis is
less conspicuous than that related to the supporting cells. It is
rather difficult to observe because it affects only a small number
of cells in a way similar to that observed at the onset of normal
spermatogenesis. When both hormones are administered simultaneously
to the hypophysectomized lamb, there is a synergistic effect leading
to a stronger testts response which resembles that obtained with an-
terior pituitary powder. This response concerns all criteria stud-
ied: organ weight, tubule diameter, and even germ cell reserve;
which is significantly increased as spermatogenesis commences
(Table 2).
Some cases of gonadotropic treatment in boys deprived of a

functional pituitary gland have been reported in the literature (pa-
tient operated on for craniopharyngioma). A 3-month treatment of
pure FSH, 3700 U 2nd IRP, does not seem to definitely stimulate the
testis (17). On the other hand, testis stimulation is provoked by
more complex supplementation with HMG, which has both FSH and LH
activities, or with HCG (11). The Sertoli cells mature and sper-
matogenesis unobtrusively commences (shown by the presence of some
Table 2. Effect of Hormone Supplementation on the Hypophysectomized

Lamb Testis.
Treatment No. of Testis Sex Cord Supporting Spermato-
and Dose Animals Weight Diameter Cells/Testis gonial Cells/
~mg) ~g) (U) x 10 6 Testis x 10 6
2
FSH (4.5)1,4 4 2.9 ± 0.2 39 ± 2.5 2570 + 75+ 65 + 9 NS
LH (2.4) 3 3.8 ± 0.5 53 ± 0.7 2370 + 350+ 49 + 12 NS
FSH + LH
(4.5 + 2.4)
4 8.9 ± 0.6 64 + 3.1 5465 + 104+ 91 + 8 NS
LH + cyproterone
acetate 4 3 3.2 ± 0.2 52 + 1.7 2010 + 192+ 44 + 9 NS
Cyproterone
acetate (20)
3 1.7 ± 0.2 31 + 2.3 1200 + 160 40 + 5
Testosterone
(3.75) 3 7 2.2 ± 0.3 38 + 1.4 1810 + 220 55 + 10
Initial Control 19 3.4 ± 0.1 49 + 1.0 2540 + 134 37 + 4

Hypox. Control 22 1.9 ± 0.1 36 + 0.6 1680 + 147 53 + 6
1
Dose given twice a day for 15 days. FSH-CNRS-P 15 (1.06 FSH-NIH-S3)
LH-CNRS-P8 (0.67 LH-NIH-S3).
2 m(± sm)' The level of significance is determined in relation to the

hypophysectomized control (NS, non-significant; + significant P~0.05).
3
Stronger doses have no greater effect.
4 FSH and LH were kindly supplied by the Endocrinology Coordination

Committee of the National Center for Scientific Research (CNRS),
which is gratefully acknowledged. The author also wishes to thank
Schering AG for supplying the Cyproterone acetate.
primary spermatocytes) in conditions somewhat similar to those re-

ported in the hypophysectomized lamb with FSH or LH. However, these
supplementations, even when given over a long period, do not estab-
lish complete spermatogenesis in these boys.
In the normal lamb, gonadotropins have the same effects on the

impuberal testis as in the hypophysectomized animal, but stimulation
is stronger, and FSH then becomes the most efficacious hormone, caus-
ing a higher testis response than that obtained with LH (21). The
cause of this may be a synergy effect of the gonadotropins injected
with those secreted by the animal's own pituitary gland under the
effect of the treatment.
360 M. COUROT
The impuberal testis of the normal boy responds to gonadotropic

stimulation in the same manner. While the administration of HMG to
unilaterally cryptorchidic boys may have no effect on the non-func-
tioning testis, it does stimulate the scrotal testis. The latter
presents "histological changes similar to those observed at approx-
imately the second phase of normal puberty" (11). The diameter of
the tubules increases, as does the number of their spermatogonial
cells. Sertoli cell differentiation begins as well as that of the
spermatogonia and primary spermatocytes at the pachytene stage. In
boys with genital infantilism, the situation seems to be more com-
plicated because one can obtain stimulation of the seminiferous tu-
bule with HMG or HCG, the best results appearing with simultaneous
administration of both hormones, but there are also some failures
with these various treatments (5-10).
When the mechanism of action of FSH and LH is analyzed, it ap-

pears that the hormones probably have a direct effect on the sex
cords of the impuberal testis (7) (Table 2). In the hypophysectom-
ized lamb, FSH acts without stimulating the interstitial gland, for
the accessory glands are umodified. LH action on the supporting
cel1s is the result of direct hormonal action on these cel1s. It
does not come from the increased testosterone secretion of the inter-
stitial gland under the effect of the gonadotropin; even injecting
a very strong dose of testosterone does not stimulate the supporting
cells. Moreover, administering an antiandrogen (cyproterone acetate)
simultaneously with LH does not significantly reduce gonadotropic
effect on the supporting cells (21).
GONADOTROPIC SUPPLEMENTATION IN THE PREPUBERAL MALE
The prepuberal rat (often referred to as an immature, a juvenile

or an impuberal animal) has been used in most studies analyzing the
effects of gonadotropins on the testis during the establishment of
spermatogenesis. Therefore, we shall take this animal as an example,
comparing paral1el results obtained in pathological human cases where
spermatogenesis cannot progress to term. Ovine FSH stimulates devel-
opment, or partially corrects post-operative testis regression with-
out any accessory gland reaction in the prepuberal hypophysectomized
rat (not older than 40 days), treated subcutaneously immediately, or
after a given delay (Table 3) (22). On the other hand, even at a
dose (8 ~g/lOO g body weight) twice that effective in the adult hypo-
physectomized rat, ovine LH does not either maintain nor restore the
testis in the prepuberal hypophysectomized rat. The seminiferous tu-
bules do not present an aspect different from that of the hypophysec-
tomized controls for LH treatment, while with FSH there is stimulation
Table 3. Effect of Ovine FSH and LH on the Maintenance of the Genital

Organ Weight in the Prepuberal Hypophysectomized Male Rat.
Treatment No. of Testis Seminal Vesicles

and Dose Animals Weight Weight (mg)
( \J.g/D) (mg)
Initial Control 5 269 + 10 16.1 + 0.8

+ 11 +3.5
±0.3
Terminal Control 5 589 30.8
Hypox. + Saline 4 67 +4 4.9
Hypox. + FSH (15) 4 330 +36 6.3 ± 0.9
Hypox. + LH (5) 4 126 +6 10.8 ± 1.2
Hypox. + FSH + LH
(15 + 5)
3 466 + 42 13.2 ± 4.4
Hypophysectomy at 27 days - 12 day treatment with FSH-CNRS-M (5.86

FSH-NIH-S3) and LH-CNRS-Ml (1.83 LH-NIH-S3).
From Ortavant, R., Courot, M. and de Reviers, M.M., In La Specificite
Zoologique des Hormones Hypophysaires et leurs Activites, CNRS ed.,
Paris, p 369, 1968.
or partial restoration of spermatogenesis, which is less advanced

than in the normal animal (Table 4). In cases of experimental or
pathological imbalance, an appropriate hormonal supplementation
should re-establish conditions in the environment which prevent par-
tial cellular degeneration. This permits spermatogenetic' evolution
to proceed to term. Thus, when spermatogenesis is established,
hormones do not accelerate the process, but prevent cellular degen-
eration.
Quantitative histological analysis of spermatogenesis shows

that FSH acts particularly on the first part of the spermatogenetic
cycle (Table 4). FSH permits a large restoration (about 2/3 of the
normal control) of spermatogonial division and of meiotic prophase
efficiency, at least until the pachytene stage. However, the follow-
ing critical stages are only partially overcome by the administration
of ovine FSH; the transition of primary spermatocytes from the pachy-
tene to the diplotene stage (42 per 100 normal control for diplotene
primary spermatocytes), and the maturation divisions (16 per 100 of
normal control for young spermatids). After that, there are no longer
enough cells to measure cellular yield accurately. It may be that
the combined administration of FSH and LH helps to overcome the
critical stages (analysis in progress) as it generally improves the
spermatogenetic yield. We already know the synergic effect of the
mixture of these two hormones on testis weight (Table 3).
Thus, the separate, subcutaneous administration of FSH or LH

shows that FSH is the gonadotropic hormone having the most effect on
362 M. COUROT
Table 4. Efficiency of Spermatogenesis in the Prepuberal Hypophy-

sectomized Rat, Supplemented with FSH or LH after 24-day
Post-operative Regression (1).
PRIMARY
No. SPERMATOCYTES SPERMATIDS
Treatment of In
(2) Rats L* p** Dt Round Elon- Elon- Sertoli
gat ion gated Cells
Terminal 4 9.0 8.5 805 35.8 46.8 31.9 115
control (3) (0.6) (0.4) (005) (0.4) (1.2) (3.8)
Hypox. + 4 12% 3% 1% < 17. 0 0 55%
Saline (4) (0.3) (0.1) (0.1) (0.1)
Hypoxo + 4. 61% 62% 427. 16% 57. <1% 68%
FSH (4)
(0.5) (0.4) (0.7) (1.3) (1.5) (0.2)
Hypox. + 3 14% 4% 2% 1% 0 0 57%
LH (4) (0.3) (0.1) (0.1) (0.2)
(1) Adapted from Ortavant et al (14), 24-day treatment, (2): hormones

and doses, see Table 3, L*: Leptoto , P**: Pacyht., D t: Diplot.,
(3): terminal control: results are expressed in daily production for
germ cells (x 10 6 ), and in testis content for Sertoli cells (x 10 6 ).
(4): treated animals: results are expressed in per 100 of the ter-
minal control; ( ): sm of the original value.
The leptotene primary spermatocytes permit the measurement of the

overall yield of spermatogonial divisions. Note the lack of LH ef-
ficiency and the decrease of FSH efficiency with progression of sper-
matogenesis.
spermatogenesis in the prepuberal animal. This agrees with overall

results reported in the literature (see review 20), in spite of some
studies which seem to have contradictory conclusions, but in which
the histological pictures appear to support the results reported here.
It should be emphasized that only a correct, quantitative histological
analysis gives objective and solid conclusions.
The results reported in men having incomplete spermatogenesis

(who may be parallel led to the individuals with prepuberal testis)
seem more complicated than those obtained in the rat. After treat-
ment with HMG, complete spermatogenesis and viable sperm in post-
treatment ejaculates have been obtained in patients with hypogonadism
(9). Positive results of spermatogenesis stimulation have also been
reported after treatments with HCG alone (6), especially if there is
already a preexisting although small spermatogenetic activity. How-
ever, the best results are obtained after a combination of FSH and
LH (5-8, 10,23).
A tentative explanation for the variability in the human results

might be the existence of variable final FSH to LH ratios. An opti-
mum might exist for the ratio but it is still disputed (24). In
supplementation with exogenous hormones, this ratio has to be con-
sidered, taking into account the patient's own gonadotropic balance.
One might also point out that in the human it could be a relation-
ship between the physiological state of the testis and its respon-
siveness to gonadotropic treatment as is the case in the rat where
the importance of age in relation to spermatogenetic response to
gonadotropins has been demonstrated (22). After puberty and in the
adult, the main hormone is LH whereas it is FSH in the prepuberal
rat, this transition occurring during the post puberal period (22).
Most often, this observation explains the various conclusions in the
literature concerning endocrine control of spermatogenesis.
CONCLUSION
Thus, in addition to the information about the endocrine balance

of the patient, a good knowledge of the state of the abnormal testis
is necessary to propose an adapted treatment. However, it should be
remembered that combined hormonal action always gives the best re-
sults, provided it is administered for an appropriate period of time.
It is at present the most often proposed treatment for spermatogenetic
accidents in man.
SUMMARY
This short analysis shows that testis seminiferous tubules are

under early pituitary control, when there is still no spermatogenetic
activity (impuberal testis). They respond to hypophysectomy or to
gonadotropic supplementation. Pituitary LH or/and FSH is necessary
to the onset of spermatogenesis (differentiation of spermatogonial
cells). FSH appears to playa leading role in the prepuberal testis,
while LH is the principal hormone after puberty, but there is always
synergy between the two hormones. This article also emphasizes the
importance of somatic cells in the seminiferous tubules and their
dependence in relation to the pituitary hormones, either before
(supporting cells) or after their differentiation (Sertoli cells).
ACKNOWLEDGMENT
This work was supported by the French National Institute for

Agronomic Research (I.N.R.A.).
REFERENCES
1. Courot, M., Hochereau-de Reviers, M. T., Ortavant, R., ~ The

Testis, Vol. I, Chap. 6, Eds. Johnson, A.D., Gomes, W.R., Van
Demark, N. L., Academic Press (in press).
364 M. COUROT
2. Smith, P.E., ~ Les Hormones Sexuelles, ed. Brouha, L., Hermann,

Paris, p 201, 1938.
3. Courot, M., J. Reprod. Fertil., Supple£: 89, 1967.
4. Ortavant, R., and Courot, M., Arch. Anat. Micr. Morph. Exp.
(Paris),~: (Suppl. 3-4) 111, 1967.
5. Heller, C. G. and Nelson, W.O., J. Clin. Endocr., 8: 345, 1948.
6. Paulsen, C.A. In Proc. 6th Pap Amer. Congress Endocrinol.,
Mexico City, 1965, Excerpta Medica Foundation ICS 112: 398, 1966.
7. Davies, A.G. and Crooke, A.C., Proc. Roy. Soc. Med:-58: 580,
1965. -
9. Lunenfeld, B. In Recent Research on Gonadotrophic Hormones,
Proceedings 5th Gonadotrophin Club Meeting, Edinburgh 1966,
eds. Bell, E. T. and Loraine, J.A., E & S Livingstone Ltd.,
p 284, 1967.
10. Crooke, A.C., Davies, A.G. and Morris, R., J. Endocr.,i£: 441,
1968.
11. Rosemberg, E., Mancini, R.E., Crigler, J.F. and Bergada, C. ~
Gonadotropins 1968, ed. Rosemberg, E., Geron-X, Inc., Los Altos,
California,p 527, 1968.
12. Mancini, R.E., Seiguer, A.C. and Perez-Lloret, A., J. Clin.
Endocr.,29: 467, 1969.
13. Courot, M7 and Courte, M.T., unpublished data.
14. Ortavant, R., Courot, M. and de Reviers, M.M., to be published.
15. Clermont, Y. and Morgentaler, H., Endocrinology,ll: 369, 1955.
16. Clermont, Y. and Bustos-obregon, E., Amer. J. Anat.,~: 237,
1968.
17. Mancini, R.E. In Gonadotropins 1968, ed. Rosemberg, E., Geron-X,
Inc., Los Altos, California, p 541, 1968.
18. Desclin, J. and Ortavant, R., Ann. BioI. Anim. Biochim. Biophys.,
3:329,1963.
19. Clermont, Y. and Harvey, C., Endocrinology,22: 80, 1965.
20. Boccabella, A.V., J. Reprod. Fertil., Supple 2: 139, 1967.
21. Courot, M., to be published.
22. Ortavant, R., Courot, M. and de Reviers, M.M. In La Specificite
Zoologique des Hormones Hypophysaires et leurs~ctivites, C.N.R.S.
ed., Paris, p 369, 1968.
23. Heller, C.G., ~ Estrogen Assays in Clinical Medicine, ed.
Paulsen, C.A., University of Washington Press, p 275, 1965.
24. Johnsen, S.G., Rosemberg, E., Lunenfeld, B., Crooke, A.C. In
Recent Research on Gonadotrophic Hormones, Proceedings 5th-
Gonadotrophin Club Meeting, Edinburgh,1966, eds. Bell, E.T. and
Loraine, J.A., E & S Livingstone Ltd. p 301, 1967.
DISCDSSION
SOLARI: Dr. Courot, have you seen signs of nuclear differentia-

tion in the pre-Sertoli cells of the immature testis after the in-
jection of LH and FSH? In the human, the mature Sertoli cell has
a nucleolus-chromatin complex which is very characteristic. In the
mature Sertoli cells from mice, the nucleolus is flanked by two het-
eropyknotic, DNA-containing bodies. Are there similar structures
in your material?
CODROT: I performed some observations with the electron microscope

on the testes of impuberal lambs. It seems to me that in normal
lambs, as well as in hypophysectomized animals, nucleoli with asso-
ciated chromatin are present in the supporting cell nuclei; but I
would say that in normal lambs it also appears that the endoplasmic
reticulum with associated ribosomes is well developed. After hypo-
physectomy, what the most important fact seems to be is the disap-
pearance of ribosomes along the endoplasmic reticulum, but as far
as the nucleolus is concerned, the postulated changes are very dif-
ficult to observe. After supplementation with LH or FSH it seems
that the first effects that could be observed concern the endoplas-
mic reticulum. But these are only preliminary results, this work
being now in progress.
BERGADA: Dr. Courot, when you injected FSH and testosterone you
did not obtain increases in testicular weight. Have you tried to
implant testosterone in the testes simultaneously with FSH adminis-
tration?
CODROT: No, I have not. Some results with injected exogenous tes-
tosterone were reported. I have tried implantation of testosterone
crystals but not simultaneous implantation of crystals and injec-
tion of FSH.
LIPSETT: Dr. Courot, do you have information as to what sort of

plasma testosterone level you achieved with the dose used or about
the dose of testosterone with respect to the amount you think an
adult ram would normally produce? In other words, were you giving
a very large dose or a replacement dose?
CODROT: May I remind you that the animals I am talking about are
lambs of 15 kg. body weight. I tested different doses of testos-
terone and the results presented here have been obtained with 7.5
mg. per day. We have chosen this dosage of testosterone according
to the response of the sexual glands i.e. seminal vesicles, and it
seems that it is a little too high because with this dose of testos-
terone we obtained seminal vesicle weights which are a little high-
er than those of normal animals. With respect to the tubular cell
content, we tried a much higher dosage of testosterone, 120 mg. per
day, and the effect on the tubules was not different from that ob-
366 M.COUROT
tained with the smaller dosage of 7.5 mg/day. This work is still
in progress and I have as yet no results to present with respect
to blood plasma concentration of testosterone. I recall that FSH
plus testosterone does not parallel the results obtained with FSH
plus LH on testis weight, tubular diameter and on the supporting
cells. This can be interpreted as a direct effect of LH on this
cell category.
A. STEINBERGER: Dr. Courot, could you tell us whether the Sertoli

cells normally divide in lambs of the age that were used in these
experiments or do they divide only following hypophysectomy and
gonadotropin stimulation?
COUROT: Normally, in lambs, the supporting cells divide for a long

time after the period at which I operated on the animals. In hypo-
physectomized animals, the total population of supporting cells de-
creases rapidly after pituitary removal and then remains at the
same level for as long as 8 ~ months; some mitotic figures can be
observed in the supporting cell population a long time after hypo-
physectomy. So it seems that the total population remains quite
the same but with some slow replacement of cells.
E. STEINBERGER: What happens to the germinal epithelium in the ram

after hypophysectomy? How much cyproterone did you administer and
did you have any definite evidence that the cyproterone blocked
testosterone at the germinal epithelium level?
COUROT: I have hypophysectomized adult rams and some animals have

been maintained without gonadotropic supplementation for 120 days
after surgery. In such animals I can observe some young round nu-
clear spermatids at a stage of evolution corresponding to the begin-
ning of spermiogenesis, but in very very small number. Hence, in
the adult hypophysectomized animal spermatogenesis seems to develop
at a very low quantitative level until the formation of the young
spermatids. This is in opposition to the young animals, where hypo-
physectomy performed before any spermatogenetic activity is observed
prevents development of spermatogenesis. As regards your other ques-
tion, in one experiment, I used 20 mg. per day and in a second exper-
iment 40 mg. per day of cyproterone acetate in animals of 15 kg.
body weight. Cyproterone acetate was injected intramuscularly twice
a day. We obtained a complete inhibition of testosterone effect on
the vesicles. We do not have to report relative to the effect of
this compound at the tubular level.
VILAR: Dr. Courot, since when you talk about stimulation you show
no increase in the number of Sertoli cells, could you explain what
you mean when you refer to stimulation? What parameter did you use?
Is it a kind of maturation of differentiation of the Sertoli cells
as we see in the human?
COUROT: I have been wrong with the expression stimulation, because

effectively when we compared results in animals supplemented imme-
diately after surgery with those obtained in normal animals, we main-
tained the number of supporting cells at about the same level. The
results were expressed as the total population of cells. We counted
the number of cells per cross section of tubules and calculated the
total population considering testis weight, diameter of tubules,
and relative percentage of seminiferous tubules. I was then able
to calculate the total population of supporting cells in the testis.
JUTISZ: Dr. Courot, you did not mention the dosage of FSH and LH
used in your work or the ratio of the two hormones in the case of
combined injections. I recall an old assay method for FSH in which
the weight of the testis of hypophysectomized rats was used as a
criterion. Your results show that FSH alone increases the weight
of the testis to a very small extent and that there is a synergy be-
tween FSH and LH.
COUROT: I used preparations obtained from your laboratory. I adapt-

ed the dosage in order to give the equivalent of 10 mg. per animal
per day of NIH FSH Sl and 2.5 mg. of NIH LH Sl, either in separate
or in simultaneous administration of the gonadotropins. The gonado-
tropins were injected subcutaneously twice a day, beginning immedi-
ately after hypophysectomy. With respect to the second question,
I would say that the bioassay you referred to was developed on young
rats. As you know young rats present a spermatogenetic activity
while lambs have no spermatogenetic activity. These are two differ-
ent physiological situations.
ULTRASTRUCTURAL AND CHEMICAL EFFECTS OF LH UPON THE SEMINIFEROUS
TUBULE 1
Mario H. Burgos, F. L. Sacerdote 2 , R. Vitale-Calpe 3 and

D. Bari 4
The interstitial cells have been pinpointed as the only target

of LH action on the testis. Greep and Fevold and Lacy et al. (1,2)
suggest that any tubular action of this hormone is mediated through
the androgen produced by the Leydig cells in response to LH. On the
other hand, recent studies on the effect of associated gonadotropins
(FSH and LH) have demonstrated that LH is necessary to maintain sper-
matogenesis (3-6) independently of the presence of androgens. The
latter have been shown to inhibit this effect (7,4).
The evidence for a direct action of LH on the seminiferous tu-

bules has received additional support from the comparative study
done in our laboratory on the mechanism of spermiation in amphibia
(8) and in mammals (9,10). In these studies we have shown that
following endogenous or exogenous administration of LH the semini-
ferous tubule is able to change the ultrastructural characteristics
of those Sertoli cells related to the last stages of spermatid
differentiation and to induce the release of late spermatids into
the tubular lumen. This mechanism of spermiation is characterized
in amphibia and mammals by cytoplasmic swelling and vacuolization
of the Sertoli cells followed by outburst of apical bubbles into the
1
This work was supported by grant M-63-l2l from the Population
Council, Inc.
2
Member Scientific Career Consejo Nacional de Investigaciones
Cientificas y Tecnicas, Argentina.
3
Research Assistant Population Council, Inc. USA.
4 Member of Technician Career, Consejo Nacional de Investigaciones
Cientificas y Technicas, Argentina.
369
370 M. H. BURGOS, F. L. SACERDOTE, R. VITALE-CALPE, AND D. BAR I
tubular lumen, increase in testicular sodium and water content in

amphibia (8) and by increase in testicular fluid as measured in the
rete testis in mammals (11).
Since these observations indicated a new target of LH in com-

petent Sertoli cells, and also on the basis of the marked swelling
and vacuolization found in electron micrographs, we have assumed
that an important portion of the testicular fluid increase evoked
by LH should be located in the territory occupied by such Sertoli
cells. Inasmuch as more than 80% of the testicular volume is occu-
pied by seminiferous tubules and that approximately 1/4 of the total
tubular mass belongs to the Sertoli cells associated with the late
spermatids, we thought it of interest to study in testicular homo-
genates and cell fractions whether this movement of water is accom-
panied by changes in some enzyme activities. We therefore measured
the response of the ATPases to LH. These are known to play a role
both in fluid transport and in the energy cycle. We also investi-
gated the effect of LH on mitochondrial respiration and determined
whether this hormone acts directly on mitochondrial fluid content.
Enzymatic study of testicular homogenates of toads, hamsters and
mice showed that the AtPase activity decreased approximately 50% in
response to a single dose of LH or to the secretion of endogenous LH
during copulation. Similar results were obtained in the study of the
oxygen consumption (10). Following this line of work we have studied
the ultrastructural characteristics of the Sertoli cell-late spermatic
association, the response to LH in the guinea pig and the enzymatic
activity of cell fractions of rat testis after addition of LH.
METHODOLOGICAL CONSIDERATIONS
Electron microscopy: Adult male guinea pigs weighing approxi-

mately 400 g were distributed in two groups: (1) Controls: kept in
cages, 2 or 3 per cage, separated from females; (2) Animals mated
with females in estrus and killed one hour after the beginning of
copulation.
The testes of Nembutal anesthetized animals were fixed by 5%

glutaraldehyde in s-collidin buffer, according to the perfusion
method of Aoki and Vitale-Calpe (12). The tissue was cut in thin
slices, washed in buffer and refixed in cold 1.33% OsO 4 in s-collidin
buffer with the addition of sucrose and calcium (13). Blocks de-
hydrated with acetone were embedded in Epon-Araldite (14). Thin
sections were examined in a Siemens Elmiskop I after uranyl acetate
and lead citrate stain (15). One-micron sections stained with
toluidine blue were studied with the light microscope for orientation
and photomicrography.
Chemical determinations: Adult male rats weighing approximately

250 g were decapitated and the testis homogenized in 0.44 M sucrose-
EDTA (Fig. 1); the mitochondrial and microsomal pellets were used for
HOMOGENATE
2.000 x g/15 min.
I
PELLET SUPERNATANT
Washed
2.000 x g/15 min.
PELLET
I
SUPERNATANT
10.000 x g/30 min.
I NUCLEI I
I
PELLET SUPERNATANT
Washed
10.000 x g/30 min.
I
PELLET SUPERNATANT
MITOCHONDRIA I 105.000 x g/30 min.
I
PELLET SUPERNATANT
I MI CROSOME I
Fig. 1. Fractionation method of rat testis homogenized in 0.44 M
sucrose-EDTA.
the enzyme measurements. Total ATPases were assayed in both pellets

in a medium containing Na+, ~, Mg++ and Ca++ by measuring the
liberated phosphorus (16) both in controls and in the presence of
50 or 100 ~g/ml LH (Ovine LH, lot NIH-LH-S-12 was kindly supplied
by the Endocrinology Study Section of the National Institutes of
Health.)
Respiration of mitochondria in succinate was measured polaro-

graphically in a Gilson model KM oxygraph according to the method
of Stoppani and Vallejos (17) with and without LH, 50 ~g and 100 ~g
per ml of substrate.
The optical densities of mitochondrial suspensions in 0.25 M

sucrose were read over 30 min. in a Beckman DU spectrophotometer at
520 m~ after the addition of LH (100 ~g/ml) and after the addition
of a well-known mitochondrial swelling agent such as 3 x 10_3M
phlorizin (18).
372 M. H. BURGOS, F. L. SACERDOTE, R. VITALE-CALPE, AND D. BARI
In some preliminary experiments 0.1 ml serum obtained from con-

trol or copulating rats was added to the substrate for ATPases and to
the substrate for respiratory enzymes and its effect assayed by the
above mentioned methods in the presence of testicular mitochondria
or microsomes.
RESULTS
Ultrastructure of Sertoli Cell-Late Spermatid Association in

the Guinea Pig
1) Control group. As described in the hamster only the last

stage of the Sertoli cell-late spermatid association presents the
characteristics of spontaneous spermiation. This is stage number
7 in the guinea pig and corresponds to number 8 in the hamster. The
~ain characteristics of this stage are the focal swelling of the
apical endoplasmic reticulum, the concomitant release of late sper-
matids and the retention of the residual bodies (Fig. 2).
2) Experimental group (one hour after starting copulation).

It has been shown in the rat (19) and in the hamster (20) that cop-
ulation is associated with a marked secretion of pituitary LH. The
Sertoli cells related to the stages V, VI and VII (Fig. 3) show a
marked swelling of the apical cytoplasmic matrix, associated with
vacuolization and distension of the apical endoplasmic reticulum.
The apical recesses become erased and the spermatids are released
together with the residual bodies. Large bubbles expel into the
lumen a fluid which helps the spermatids move toward the rete testis
(Fig. 4). We have measured in the hamster rete testis the increase
in fluid coming from the testis in response to intravenous injection
of LH (ll).
Biochemical Observations
Optical density of mitochondria. The addition of LH (100 ~g/ml)

to a mitochondrial suspension produces a significant decrease in op-
tical density in less than 10 minutes (Fig. 5) which continues for
20 minutes. This effect is comparable to the mitochondrial swelling
produced by 3 x 10-3M phlorizin.
ATPase activity. The addition of 50 ~g/ml or 100 ~g/ml LH to

the substrate in the presence of isolated testicular mitochondria
increases ATPase activity by approximately 25% or 35% respectively;
the addition of LH inactivated by boiling for 10 minutes in a water
bath causes no increase (Fig. 6). Experimental serum as compared to
control serum also increases mitochondrial ATPase activity by 25%.
On the other hand, microsomal ATPase activity shows a moderate

depression after 100 ~g/ml of LH (Fig.6) •
fp~ t'
I.S
8{1o
1-$ Q Q G
r
~ ~ ~ ~
11
® ® ~ ~
A @P) cQJ @§) (Q) FICI 2
ry
11(, .«
Figs. 2 and 3. Schematic representation of stages 5 to 8 of guinea

pig spermatogenesis. Fig. 2 shows the spontaneous spermiation of
stage 7 with the retention of the residual body and release of
the spermatid. Fig. 3 shows the post-coital spermiation, with
release of spermatids in stages 5,6 and 7 without retention of re-
sidual bodies. In both figures the apical cytoplasm of the Sertoli
cells (in grey) shows swelling and vacuolization during spermiation
of late spermatids (LS). The other cell types are: A spermatogonium
A; B spermatogonium B; P - primary spermatocyte, pachytene; ES-
early spermatid.
374 M, H, BURGOS, F, L, SACERDOTE, R, VITALE-CALPE, AND D, BARI
l'()!:T-('ornl SI'I:1I"1 \TI(),\ I 11 \ "'TI:n /s,- \(;1 : \'1\
CIIA ,\ ca::-; I SEllIOl.1 ('1'1 I \1'1( ' \1 ('\ 1IlI'I ,\';\1
A II
P )ST-COITAL SPE IUII ,\ 1'10.' I,\, I:\EA PI. (STA ;~: \ ')
IIANGJ;;S I~ SEHTOI.I CEI.I. APICAl , CYTOI'I.Mi\1
A B C
Figo 4. Schematic representation of post-coital spermiation in the

hamster (stage VI) and in the guinea pig (stage V). A shows the
beginning, B the middle and C the end of the mechanism of spermiation.
The apical recesses become erased by swelling of the cytoplasmic
matrix and endoplasmic reticulum, and late spermatids together with
the residual bodies are released.
RAT TESTIS MITOCHONDRIA
5~ 0
54 0
53 0
CO;-.lTROr.
520
510
0
N
oJ'>
500 -3
0 P HLORI ZIN 3 x 10 M
490
480
0 . 44 M Sucrose
470
480
0 5 10 15 20 25 min .
Fig. 50 Optical density of a mitochondrial suspension. It shows

the mitochondrial swelling produced by addition of LH or phlorizino
R AT TESTIS ATPase
+ + ++ ++
Na K Mg Ca
10 o CONTROL
9 m LH 50 II
8 • LH 100 "/1
ffi] CONTROL SE RU M
~ EXPER . SERUM·
I - - MITO<':II0N DRIA - - . MICROSOMES
Fig. 6. A typical experiment in isolated mitochondria shows the

increase in ATPase activity after addition of LH or post-coital
serumo Isolated microsomes show a moderate depression after LH.
376 M. H. BURGOS, F. L. SACERDOTE, R. VfTALE-CALPE, AND D. BARf
ISOLATED MITOCHONDRIA
RAT TESTIS
100
90
0 CONTROL
0 LH 50 Y
80
!1il UI 100 Y
•
z
..
0 70
.: LH 200 Y
>: 60
::>
'"0z SO
0 HEATED LH
'-'
..,z 40
~ )0
1'i
...
,.,
0 20
10
o
Fig. 70 A typical experiment on oxygen consumption. Mitochondrial
respiration is markedly depressed after 200 ~g LH. A similar dose of
heated LH 'produces no effecto
5 min.
The values indi ca te IT\}J atoms 02 / min.
Fig. 8. Polarographic record of the respiratory response of isolated

mitochondria in succinate substrate, after the addition of LH.
Oxygen consumption. The addition of LH to the succinate substrate

decreases mitochondrial respiration by about 25% (Figs. 7 and 8).
The mitochondrial and microsomal fractions were always checked

in the electron microscope and were consistently free of contamin-
ation by microsomes and mitochondria, respectively.
DISCUSSION AND CONCLUSIONS
Post-coital changes in the seminiferous tubules of the guinea pig

are comparable to those already described in the hamster (9). In the
guinea pig from stages V to VII, the apical conglomerate of smooth
endoplasmic reticulum of the Sertoli cells swells markedly and forms
large vacuoles and dilated cisternae. The rest of the tubule shows
no significant changes. The present observations therefore offer
additional evidence in favor of our hypothesis that the Sertoli cell-
late spermatid association is a target of LH action on the tubule.
More specifically, the target is the apical cytoplasm of those com-
petent Sertoli cells.
The enzymatic mechanisms involved in such action remain obscure

but some evidence has been presented that the sodium pump (Na+ de-
pendent ATPase) is related to the LH action and to the swelling of
the Sertoli cells (10).
Experiments with total homogenates (10) showed that the activity

of Na+ dependent ATPase is blocked by LH approximately 50%. The
results presented in this report prove that the microsome fraction
is also depressed, but to a much lesser extent than the total homo-
genate. This may indicate that most of the LH sensitive ATPase is
located in the cell membrane fraction (first pellet). On the other
hand, the mitochondria, which only contribute 10% of the total ATPase
activity of testis homogenates, contain an ATPase which is signifi-
cantly stimulated by LH.
Isolated testis mitochondria swell in the presence of LH. This

phenomenon is associated with an increase in fluid uptake and with
a fall in mitochondrial ATP concentration. The fact that in the
work reported here LH inhibits respiration and at the same time
stimulates ATPase activity agrees perfectly with the mitochondrial
swelling after LH. Other hormones have been also found to produce
mitochondrial swelling (oxytocin, vasopressin,insulin, thyroxine,2l).
In conclusion, the ATPase sensitive to LH is mainly related to

the cell membrane fraction.LH is able to induce mitochondrial swelling
directly.
Both effects can help the understanding of the testicular fluid

increase evoked by LH in mammals.
378 M. H. BURGOS, F. L. SACERDOTE, R. VITALE-CALPE, AND D. BARI
REFERENCES
1. Greep, R.O.and Fevold, H.L. Endocrinology, 21: 611, 1937.

2. Lacy, D., Vinson, G.P., Collins, P., Bell, J., Fyson, P.,
Pudney, J. and Pettitt, A.J., Proc. 3rd Int. Congress Endoc.,
Mexico,Excerpta Med. Found., 1019, 1969.
3. Lostroh, A.J. Endocrinology, 85: 438, 1969.
4. Heller, C. G., Lalli, M.F. and Rowley, M.J., Ciba Found. Colloq.
on Endocr., 1£: 1967.
5. Courot, M., J. Reprod. Fertil., Suppl.-1: 89, 1967.
6. Mancini, R. E., Seiguer, A. C. and Perez Lloret, A., In Gonado-
tropins 1968, ed. Rosernberg, E., Geron-X, Inc., Los Altos,
California, p 503, 1968.
7. Rosernberg, E., Mancini, R.E., Crigler, J. F. and Bergada, C. In
Gonadotropins 1968, ed. Rosernberg, E., Geron-X, Inc. Los Altos,
8. Burgos, M.H. and Vitale-Calpe, R., Amer. Jo Anat., 120: 227,
19670
9. Vitale-Calpe, R. and Burgos, M.H., J. Ultrastruct. Res., 1969
(in press).
10. Burgos, M.H., Vitale-Calpe, R. and Russo, J., In Gonadotropins
1968, ed. Rosernberg, E., Geron-X, Inc., Los Altos, California,
p 213, 1968.
11. Burgos, M.H., Vitale-Calpe, R., Proc. 3rd Int. Congress Endoc.,
Mexico, Excerpta Med. Found., 1030, 1969.
12. Aoki, A. and Vitale-Calpe, R., personal communication, 1969.
13. Burgos, M.H., Vitale-Calpe, R. and Tellez De Inon, M.T., J.
Microsc., !: 457, 1967.
14. Mollenhauer, H., Stain Techn., 39: 111, 1964.
15. Reynolds, EIS., J. Cell Biol., 17: 208, 1963.
16. Fiske, Ch.H. and Subbarow, H., J: Biol. Chern., 66: 375, 1925.
17. Stoppani, A.O.M. and Vallejos, R.H., Arch. Biochern. 117: 573,
1966.
18. Burgos, M.H., Aoki, A. and Sacerdote, F.L., J. Cell Biol., 23:
207, 1964.
19. Taleisnik, S., Caligaris, L. and Astrada, J., J. Endocr. 79:
125, 1966. -
20. Donoso, A.O. and Santolaya, R.C., Acta Physiol. Lat. Amer., 12:
70, 1969.
21. Lehninger, A.L., The Mitochondrion, Benjamin, New York, 1964.
DISCUSSION
JOHNSEN: Dr. Burgos, Dr. Mancini showed that whereas F"SH enters
the tubule, LH does not.
BURGOS: When we inject LH, there are changes in the later stages
of spermatids associated with Sertoli cells.
CLERMONT: Dr. Burgos, you are probably aware that, in recent studies
by Dr. Fawcett, he does not observe any vacuolization or any loss of
cytoplasm at the time of spermiation. Secondly, in our own laboratory
we have never succeeded in provoking premature release of spermato-
zoa from the seminiferous epithelium of hypophysectomized or normal
rats injected with various hormones. In your electron microscopic
study, have you observed any sign of premature maturation of sperma-
tids in your LH-injected animals?
BURGOS: I am aware of the work of Dr. Fawcett on spontaneous re-

lease of spermatids at stage 8 in the guinea pig. I think that, in
his preparations, there is some swelling of the cytoplasmic projec-
tions of the Sertoli cells. However, we work at much earlier stages.
We have studied testicular preparations between 30 minutes and one
hour after copulation or after injection of the hormone, and we
never have found release of any other type of cell apart from late
spermatids.
BIOLOGIC EFFECT OF HUMAN PITUITARY LUTEINIZING HORMONE AND HUMAN
CHORIONIC GONADOTROPIN
E. Rosemberg, J. F. Crigler, Jr., W. F. Jan, G. Bulat,

R. Nakano and S. G. Lee
Medical Research Institute of Worcester, Inc., Worcester

City Hospital, Worcester, Massachusetts and The Harvard
Medical School, Boston, Massachusetts
Bioassay [ovarian ascorbic acid depletion (OAAD) and ventral

prostate weight (VPW) assays] and radioimmunoassay of Human Pitu-
itary Luteinizing Hormone (HLH) and of Human Chorionic Gonadotropin
(HCG) indicate differences in biologic activity and immunologic re-
activity of these preparations. Information concerning the compar-
ative effect of these gonadotropin materials in the human testes is
meager. Hence, it was decided to conduct acute studies with HLH
and HCG in four normal adult male volunteers, and in two adolescent
male subjects with immature testes.
The preparations used in this study were a commercial prepara-

tion of HCG (Follutein, Squibb) with an activity of 10,000 IU HCG
per vial (biologic potency) and a purified HLH preparation supplied
by the National Pituitary Agency (Lot AI, LER 856-1) with an LH ac-
tivity of 200 IU 2nd IRP per vial; the FSH contamination was 2.5 IU
2nd IRP per vial (biologic potency).
ASSAYS
Plasma testosterone levels were determined by the competitive

protein binding assay (1). The LH activity in plasma was determined
using the double antibody radioimmunoassay described by Odell et ale
(2). All samples from an individual subject were assayed in the
same assay at two dose levels. The reference materials used in all
assays were: LER 907 (distributed by the NPA, USA); the Second In-
ternational Reference Preparation for Human Menopausal Gonadotropin
(2nd IRP); LER 856-1, and HCG Second International Standard (2nd IS).
381
382 E. ROSEMBERG ET AL.
All reference preparations were tested at four dose levels. Radio-

immunoassay results were calculated by appropriate computer programs
for parallel line assays. In our assay system, the immunoreactivity
of 1 milli International Unit (mIU) of biological activity of the
2nd IRP is equivalent to that of 3.2 ng of LER 907, 0.63 ng of LER
856-1, and 0.28 mIU of biological activity of HCG-2nd IS; 10.6 mIU
of biological activity of HCG-2nd IS is equivalent to that of 1 ~g
of LER 907. In this discussion all concentrations will be expressed
in terms of ~g of LER 907 per milliliter of plasma.
SUBJECTS
Four normal male volunteers of ages ranging from 26 to 35 years,

and two sexually immature male patients were selected for these
studies. The first patient (P.R.), aged 16 years, presented bilat-
eral undescended testes at birtho At adolescence, he became con-
cerned about the lack of sexual development and was referred at age
13 7/12 years to the Childrens Hospital Medical Center, Boston for
evaluation. His height and bone age corresponded to a chronological
age of 13 years, and his weight was 108.2 lbs. Lack of sexual de-
velopment was observed i.e., the penis was small with a stretched
length of 5.3 cm and a diameter of 1.4 cm and the testes were not
felt in the scrotal region or in the inguinal canal. Axillary and
pubic hair were absent. Bilateral orchidopexy was performed at age
14. The testicular biopsy obtained at that time revealed the pres-
ence of small tubules with arrest of spermatogenesis at the sper-
matocyte stage, presence of immature Sertoli cells, and absence of
Leydig cells. Prior to the present studies the patient was never
treated with gonadotropin preparations.
The second patient (J.K.) age 18 6/12 years, was referred to

the Childrens Hospital Medical Center at 8 years of age because of
short stature. On examination, his height age was 4 6/12 years, his
bone age 3 2/12 years, his weight 40.2 lbs. and his genitalia were
normal for his age. The usual laboratory procedures revealed defi-
ciences in GH, TSH and ACTH and a diagnosis of idiopathic hypopi-
tuitarism was made. The patient was treated with I-thyroxine (T4 ),
0.1 to 0.15 mg per day from age 10 3/12 years to age 11 9/12 years.
At 13 years of age, the patient was treated with human growth hormone
(2 mg q.o.d.); T4 was re-started while on growth hormone at age 15
3/12 years. Both medications were given continuously until the time
of the present studies. At 18 years of age, the patient did not
show secondary sexual development. On examination, the testes mea-
sured 3.8 x 2 cm, the phallus measured 5.3 x 1.8 cm, and axillary
and pubic hair were absent. A testicular biopsy performed at 18
years of age revealed the presence of small tubules without a lumen,
arrest of spermatogenesis at the spermatocyte stage, immature Sertoli
cells, some thickening of the tubular wall and a diminished number of
Leydig cells. Prior to the present studies, the patient was never
treated with gonadotropin preparations.
EFFECT OF HLH IN NORMAL MALE SUBJECTS
This study was designed to determine the effect of a single dose

administration of HLH. Each of the four subjects received a dose of
2,000 IU HLH given intramuscularly at 9:00 a.m. Blood samples were
drawn prior to HLH administration and at intervals of 15'; 45'; 75';
2 hr 15'; 3 hr 15'; 4 hr 15'; 5 hr 15'; 6 hr 15'; 24; 28 and 31 hours
following HLH administration.
Fig. 1 depicts the LH concentration in plasma and the plasma

testosterone levels after a single dose of 2,000 IU HLH (1M) on
each of the four subjects.
Time After Administration

15' 45' 75' 2"15· 3"1$ 4'15' 5"1s 6'15· 24' 28' 31"
i i i i i , I t ' i f
1.5
1.3
1.1
0.9
§e
~8 0.7
"a:
~~ 0.5
j!
~.~
0.3 ,..-._____.. 2
.. ~ 0.1 4
it O.OrL"~__________________
. 1200
4
~ 1000
•
~ BOO 2
!s_ BOO
!8 ...0
~K
."
400 C;
it g' 200
Fig. 1. Plasma LH concentration and plasma testosterone

levels after a single dose administration of 2,000 IU HLH
to four normal male subjects.
Plasma LH concentration increased to 3 to 6 times the control levels

in all subjects 15 minutes after HLH administration. Except for one
(sujject 2), who did not show significant changes in LH concentration
throughout the period of observation, all other subjects showed max-
imum plasma LH concentration (22 to 40 times control values) 3 hours
and 15 minutes after HLH administration. The concentration of LH in
plasma remained above control levels ( 3 to 9-fold ) in all subjects
24 hours after HLH administration and 3 to 5 times control values in
two subjects (subjects 2 and 4), 31 hours after HLH administration.

Testosterone levels, although remaining above control levels through-
out the period of observation, varied from subject to subject. How-
ever, all subjects showed increased testosterone levels 24 hours
after HLH administration. Testosterone levels in two subjects
(subjects 2 and 4) were above control levels 28 and 31 hours after
HLH administration, respectively.
Fig. 2 depicts the mean increase in plasma LH cOhcentration,

and the mean percent increase in testosterone levels in the four
subjects studied.
Time After Administration
15' 45' 75' 2"15' 3"15' 4"15' 5"15' 6"15' 21( 2ft 31·
I i i i I I I I I i I
.§
et: 30 Mean Values ( !SE)
•. S
Four Sub,ecls Two Sublects
Q)
g 20
8 39 39
l:
e
...J 10
i
13 13
o~__~____~~__=-~~__~________________
Increase Above Control Levels
80
~
1.
!60
~
.!!40
i
F 20
~
til
cu
~ OL-__~__________~~~__~~____________
Percent Increase Over Control Levels
Fig. 2. Mean values (four subjects). Plasma LH con-
centration and plasma testosterone levels after a
single dose administration of 2,000 IU HLH.
Maximum plasma LH concentration was observed 3 hours and 15

minutes after HLH administration. The LH concentration declined
steadily throughout the period of observation. However, LH levels
in all subjects were 6 times the control values 24 hours after HLH
administration. In two subjects LH levels showed a 4-fold increase
compared with control levels 28 and 31 hours after HLH administra-
tion. Testosterone levels were increased 15 minutes after HLH ad-
ministration, and remained elevated (65% above control), 24 hours
after HLH administration. In two subjects (2 and 4) testosterone
levels were 90.6 and 5.4% above control, and 66.4 and 5.5% above
control 24 and 21 hours after HLH administration, respectively.
EFFECT OF HLH AND HCG IN THE IMMATURE TESTIS
Patients P.R. and JoK. received four identical courses of med-

ication. During the first course, 5,000 IU of HCG was given as a
single intramuscular injection each day for three consecutive days;
during the second course, 5,000 IU of HLH was given intramuscularly
in three divided doses each day for three consecutive days;during
the third course HLH was administered using the dosage and mode
of administration used in the second course. During the fourth
course HCG was administered using the dosage and mode of administra-
tion used in the first course. Each course of medication was sep-
arated by an interval of two to four weeks. Blood samples were
drawn three times daily at 8:30 a.m., 2:30 p.m. and 8:30 p.m., the
day prior to initiation of medication, during medication days and
for two days after cessation of therapy. A final blood sample was
obtained at 8:30 a.m. on the third day after withdrawal of medica-
tion.
Fig. 3 shows the plasma gonadotropin concentration and testos-

terone levels observed during each of the four courses of medica-
tion given to patient P.R. There was no significant variation be-
tween samples in a 24 hour period. Hence mean values for a 24 hour
period for both plasma gonadotropin concentration and testosterone
levels will be reported in the present discussion (Patients P.R. and
J .K.)
P.R.
Day6
!3i:~1
i!~...J 0.7
<3 .; 5. 0.5
~~
~..,~ -~ ~CG(4)SOOOIU
~_________ ~L....
H (2) 5000 IU
'lI~Z 0.3 ~
; '" 0.1 HCG (1) HLH(3) 5000 IU
~ ~ 0.011L----'='-"'--"--''"'--_ _ _ _ _ _ _ _ _ _ _ _ _ _SOOO
_ _I_U_-----"1
~:~!
200
180
G(4)
160 ~LH(3)
140t LH(2)
j ;~~r
!l 80 HCG(1)
~ ~ 60
Ii ~ 40
:ii-i~ 20
~3~ 0L----------------------~
Fig. 3. Patient P.R. Plasma gonadotropic activity
(LH) and plasma testosterone levels during four
courses of medication with HCG, HLH, HLH and HCG,
respectively.
During each of the four courses of therapy, plasma gonadotropin

concentration increased to 11 to 37 times the control levels on the
first day of therapy; 16 to 43 times on the second day of therapy;
18 to 47 times and 10 to 28 times on the first and second day after
withdrawal of medication, respectively. Plasma gonadotropin con-
centrations remained elevated (6 to 20-fold) on the morning of the
third day after cessation of therapy. Plasma testosterone levels
were not significantly elevated until the first day after cessation
of therapy (courses 2, 3 and 4). Increased testosterone levels
were observed on the second day after withdrawal of medication
(courses 2, 3 and 4) and on the morning of the third day after
cessation of medication (all courses).
It should be remembered that this patient exhibited undescended

testes until age 14, at which time the testicular biopsy revealed
absence of Leydig cellso Hence, it is important to note that two
years after orchidopexy, successive courses of medication steadily
increased the response of the Leydig cells to gonadotropin stimu-
lation.
Fig. 4 depicts the plasma gonadotropin concentration and tes-

tosterone levels observed during each of the four courses of med-
ication given to patient J.K.
1.9
1.7
1.5
E 1.3
.~ ~
=Xo::
1.1
0.9
~A,HCG(1)5000 IU
~~(3)5000IU
-8..J w
~~~ 0.7
~~! ~:~
In m 0.1
HLH (2) 5000 IU HCG (4) 5000 IU
~ ~ O.Q1'1'--"'-"L..lI!-=-_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __
2000
1800
1600
1400
I _1:~
i!! 1200 HLH (2)
~ ~ 600
E" & 400
lIl1" 2°OL.J;,::Q::£~§::=~~
a:..Jg' 0 t:r _ _ _ _ _ _ _ _ _ _ _ _ _ _ __
Fig. 4. Patient J.K. Plasma gonadotropic activity (LH)
and plasma testosterone levels during four courses of
medication with HCG, HLH, HLH and HCG, respectively.
Plasma gonadotropin concentration increased in a m~nner similar

to that observed in patient P.R. The response to gonadotropin ad-
ministration as judged by the plasma testosterone levels was qUite
striking. During each of the four courses of medication, testos-
terone levels were significantly increased (10 to 18%) over control
levels on the second day of therapy. Plasma testosterone concentra-
tion increased steadily thereafter: 25 to 38% on the second day of
therapy, 28 to 52% and 29 to 63% on the first and second day after
cessation of medication. Testosterone levels were still elevated
(29 to 99%) on the morning of the third day after withdrawal of
medication.
COMMENTS
The data suggest that both HLH and HCG at the dosage given in-
duced stimulation of the Ley,dig cells in two subjects with immature
testes. Increased testosterone levels were observed in both sub-
jects and were maintained 60 hours after the last HLH injection and
72 hours after the last HCG injection. However, the magnitude of
the increase in plasma testosterone levels differed in each subject.
The difference in response could be attributed to differences in
the functional state of the end organ in each individual patient,
and also to the fact that patient J.K. had received growth hormone
medication for over three years prior to and throughout the course
of the present studies. The possibility exists, therefore, that
human growth hormone may exert a stimulatory action on the testis
complementing that of HLH. In both patients, successive courses of
medication steadily increased the response of the Leydig cells to
gonadotropin stimulation.
The pattern of gonadotropin concentration in plasma during the

four courses of medication given was similar in both subjects. The
slow disappearance of plasma immunoreactive gonadotropin (HLH and
HCG) was coupled with a steady increase in testosterone levels after
withdrawal of both hormones. It could be postulated, therefore,
that the immunoreactive circulating gonadotropins exerted a pro-
longed biologic effect on the Leydig cells.
Our studies showed that in normal subjects, plasma levels of

immunoreactive HLH remained above control levels 31 hours after the
administration of 2,000 IU of HLH given as a single intramuscular
dose. Testosterone levels were also increased over control levels
31 hours after HLH administration.
The disappearance rate of endogenous or of exogenously adminis-

tered gonadotropins has been studied in the human by several in-
vestigators. These data are summarized in Table 1.
w
00
00
Table
DISAPPEARANCE RATE OF ENDOGENOUS AND EXOGENOUS GONADOTROPINS
EXOGENOUS
Material ENDOGENOUS Mode of t 1/2 MET~OD Investigator
Administrat ion
iv 1 hr Bioassay Parlow (1965, Ref 3)
iv 1.16 hr RIA Schalch e t a1. (1968, Ref 4)
HLII 3O-6O' (initial component) RIA Kohler et ale (196S, Ref 5)
iv
21' (initial component) Yen et ale (1968, Ref 6)
After hypox. - 3.9 hr (second component) RIA

BFSR 3.9 hr (initial component) RIA
After hypox. Yen et a1. (1970, Ref 7)
- 70.4 hr (second component)
5.1- 5.6 hr (initial component)
iv 23.6-23.9 hr (second component) RIA Rizsllah et a1. (1969, Ref 8)
iv 8 hr RIA Wide et a1. (1968, Ref 9)
11 hr (initial component)
RCG After delivery - 23 hr (second component) RIA Yen et ale (1968, Ref 6)
8.9 hr (initisl component) RIA Midgley-Jaffe (1968, Ref 10)
After delivery - 37.3 hr (second component)
1M 32 hr
(10,000 LU) RIA Midgley-Jafle (196H, Ref 10) !'1
;;0
1M 30.1 hr (maximum at 6 hr) o
RIA Rizallah et a1. (1969, Ref H) (n
,(l(l,OOO LU) -~--
m
~
cr:>
m
;;0
Q
!:!l
>
r
The studies utilizing the i.v. infusion of a tracer dose of

hormone indicate that the half life (t~) of HLH is approximately
1 hour, that of HFSH (human follicle stimulating hormone) is approx-
imately 3 hours, and that the t~ of HCG shows a fast and slow com-
ponent of approximately 5 and 23 hours, respectively. The study of
the disappearance rate of endogenous gonadotropins indicates that
the t~ of HLH has a fast and slow component of 21 minutes and 3.9
hours, respectively, and the t~ of HFSH has a fast and slow com-
ponent of 3.9 and 70.4 hours, respectively. The t~ of HCG has a
fast and a slow component of 8 to 11 hours and 23 to 37.3 hours,
respectively. The calculations of metabolic clearance rates (MCR)
(5 and 8) of HLH and HCG indicate that the MCR of HCG is approxi-
mately 1/10 that of HLH.
Because of the apparent difference in the metabolism of these

hormones it was of interest to note that the only two reported
studies on the disappearance rate of intramuscularly administered
HCG showed a disappearance rate of approximately 30 hours, which
is similar to that observed in the present studies after a single
intramuscular dose of HLH. The slow release of HLH and HCG into
the circulation could provide an explanation for these findings.
The only support for this explanation could be found in experiments
(ll) in which the kinetics of HCG- l.3l.I adsorption after intra-
muscular injection were studied, showing that 50% of the radioac-
tivity disappeared from the injection site in 110 minutes, while
10% remained in the site of injection six hours after the HCG_l. 3 l.I
administration. However, it remains to be shown that the kinetics
of HLH_l.31. I adsorption after intramuscular injection are similar
to that of HCG.
ACKNOWLEDGEMENTS
This work was supported by Grants AM-07564 and RR 00128, USPHS,

National Institutes of Health, Bethesda, Maryland, and by a Grant
from the Theodore Schulze Foundation, New York, New York.
We are indebted to the National Institutes of Arthritis and

Metabolic Diseases, NIH, and to the National Pituitary Agency,
Baltimore, Maryland for the supply of immunologic reagents and the
supply of HLH and HGH used in this study. We wish to thank Dr. E.
C. Reifentstein, Jr., the Squibb Institute for Medical Research, New
Brunswick, New Jersey for the gift of Follutein (human chorionic
gonadotropin), and Dr. D. R. Bangham, Department of Biological
Standards, Medical Research Council, Mill Hill, London for the
gift of HCG 2nd IS and the 2nd IRP.
REFERENCES
1. Mayes, D. and Nugent, C. A., J. Clin. Endocr., ~: 1169, 1969.

2. Odell, W. D., Ross, G. T. and Rayford, P. L., J. Clin. Invest.,
~: 248, 1967.
3. Parlow, A.F., Recent Progr. Hormone Res., 11: 2Ql, 1965

(Discussion).
4. Schalch, D.S., Parlow, A.F., Boon, R.C. and Reichlin, S., J.
Clin. Invest., i[: 665, 1968.
5. Kohler, P.O., Ross, G.T. and Odell, W.O., J. Clin. Invest., i[:
38, 1968.
6. Yen, S.S.C., Llerena, 0., Little, B. and Pearson, O.H., J. Clin.
Endocr., 28: 1763, 1968.
7. Yen, s.s.C7, Llerena, L.A., Pearson, O. H., and Littell, A.S.,
J. Clin. Endocr., 30: 325, 1970.
8. Rizkallah, T., Gurpide, E. and Vande Wiele, R.L., J. Clin.
Endocr., 12: 92, 1969.
9. Wide, L., Johannisson, E., Tillinger, K.-G., and Diczfalusy,
E., Acta Endocr. (Kobenhavn), 22: 579, 1968.
10. Midgley, A.R. and Jaffe, R. B., J. Clin. Endocr., ~: 1712,
1968.
11. Parker, M.L., Utiger, R.D. and Daughaday, W.H., J. Clin. Invest.,
lli 262, 1962.
DISCUSSION
TROEN: Dr. Rosemberg, do you have comparable data on the pattern of

the blood levels of FSH after intramuscular injection of HMG or HFSH?
ROSEMBERG: No, we do not.
PAULSEN: We have studied a normal male who received an intramuscular

injection of 450 IU of FSH and 450 IU of LH. Serum LH was increased
significantly two hours after injection. Within twenty-four hours,
the serum LH levels had returned to control levels. However, the
FSH levels remained elevated at 24 hours. These data agree in gen-
eral with information obtained by Dr. Ross's laboratory. They dem-
onstrated that the metabolic clearance rate for LH is essentially
twice that observed for FSH.
ROSEMBERG: Dr. Paulsen, you realize that the amount of LH given con-
tained in the HMG administered was only 450 IU. We administered 2000
IU. If pituitary LH (HLH) and urinary LH follow the same metabolic
pathway, then, after a single injection of a much higher dose of HLH,
circulating levels of HLH will remain elevated 24 hours after the ad-
ministration of HLH. We have to bear in mind the dose and type of
hormone being given.
PAULSEN: Yes, I agree. I will show you later this morning data
which indicate that the pituitary HLH at a maximum dose of 400 IU is
cleared fairly rapidly.
ROSS: I have no data on the disappearance from plasma when Pergonal

is given intravenously to the human subject. Rowever in validating
the use of iodinated human pituitary extract LR preparations as a
tracer in our studies of metabolic clearance rates, we examined the
disappearance of this radioactively labelled tracer material and of
Pergonal given intravenously simultaneously to rhesus monkeys. We
showed essentially identical kinetics of disappearance of the radio-
actively labelled human pituitary extract and the LR component of
Pergonal.
SAVARD: I believe there are two lines of thought running through

our discussion. One is the fate of the gonadotropin as it circulates
peripherally. The other is the fate of the gonadotropin as it sits
in the binding sites of the target cells. Are these one and the same
thing? I do not believe so. I believe that it is possible to have
all kinds of peripheral changes in concentration, clearance, and so
forth; while all target cells are very happily saturated with mole-
cules of the tropic hormone. It seems to me that once above a crit-
ical plasma concentration, most variations are of little consequence
to the target cell. I think that, in attempting to relate gonadotro-
pin to cellular function, we must look to the half-life of the gonado-
tropin at the cell site. This is no easy problem at the present time.
LUNENFELD: In answer to Dr. Savard, Mrs. Eshkol and myself have

actually tested this problem; we tested the pattern of ReG exchange
between the ovaries and the circulation in mice. Four phases of ReG
uptake in the ovary were identified: a) during the first hour only
inflow of ReG was detected, b) between 60 and 90 minutes inflow from
the circulation was greater than the outflow from the ovary to the
circulation, c) between 90 and 180 minutes inflow was equal to out-
flow and in the fourth phase between 3 and 6 hours the outflow was
greater than the inflow. I think it is these kind of experiments
which could answer the questions which you have asked. (Eshkol, A •
and Lunenfeld, B., ~ Gonadotropins, 1968, p 188).
BERGADA: Dr. Rosemberg, I was very enthusiastic about the results

of plasma testosterone on your two patients, in whom testosterone
levels increase after treatment with ReG. In our experience, before
puberty, we have not seen this result. In one case, after five days
of 5,000 IU of ReG daily, plasma testosterone rose and came down on
the third day after ReG was discontinued. In some patients treated
for five days with ReG followed by 1,000 units weekly, after one or
two months the testosterone level was less than after the acute ReG
test. I wonder whether the testis before puberty responds, as far
as steroid synthesis is concerned, equally as well as after puberty.
I am referring to the testis of a 7- or 8-year-old boy as opposed to
that of a 16- or l7-year-old boy with or without pituitary function.
ROSEMBERG: In our cases, we know that the patients did see endogenous
gonadotropins; they had measurable levels prior to administration.
This means that, although histological evidence indicated that Ley-

dig cells were absent or immature, the testis, however, had been ex-
posed to some gonadotropin stimulation. Both responded to medication.
One responded better than the other. This is the patient rece~v~ng
growth hormone throughout these studies. We cannot state that a pre-
pubertal testis has never seen gonadotropin. Dr. Ross is here to
testify that there is measurable gonadotropin in prepubertal children.
So, I really do not have an explanation for your findings.
BERGADA: With regard to the human growth hormone administration to-

gether with HCG, we have experience with three patients: one with a
congenital idiopathic hypopituitarism and two with organic lesions.
We administered HCG during a period of three or four months, and did
not obtain any androgen response. In idiopathic congenital hypopi-
tuitarism, when we added human growth hormone to HCG we obtained de-
velopment of secondary sexual characteristics.
ROSEMBERG: It is possible then to conceive that growth hormone acts

synergistically with LH on the Leydig cells.
DICZFALUSY: I have a question and a comment, Dr. Rosemberg. Is there

any information available on the amount of LH which is excreted in
the urine following a single injection? We have estimated both by
bioassay and by radioimmunoassay the amount excreted in the urine
and it was around 20% with both methods. Actually, the half-life
time was between 8 and 11 hours. The results of bioassays and radio-
immunoassays showed a close agreement.
ROSEMBERG: We are in the process of completing all the urinary studies

on the LH excretion after administration of LH.
OHNO: If the patient himself was making the mutant FSH or LH, the
normal FSH or LH would serve as an antigen and he would start making
the antibodies against it. Have you had any cases where patients
started making the antibodies against the injected LH or FSH?
ROSEMBERG: We did not study this problem.

THE EFFECT OF GONADOTROPINS ON THE PREPUBERTAL TESTIS
Cesar Bergada
Servicio de Endocrinologia, Hospital de Ninos, Buenos

Aires, Argentina
The response of the infantile testis to the exogenous adminis-

tration of gonadotropins has been occasionally studied in detail in
the human (1-6) as it is reported as a functional test in isolated
cases. This is understandable as there is no justification for the
administration of gonadotropins to a child unless it is necessary
to study testicular function in cases with a probable diagnosis of
primary hypogonadism, abnormality of sexual differentiation (6-8),
or in the presence of bilateral or unilateral cryptorchidism (1,6,
7,9). This latter condition has been considered as the ideal sit-
uation in which to study the effect of gonadotropins on the prepu-
bertal testis and its capacity to secrete testosterone and differ-
entiate the germinal epithelium. These studies not only increase
understanding of the role of different gonadotropins in testicular
maturation but also permit investigation into the pathogenesis and
behavior of some diseases before puberty; such as cryptorchidism
associated with primary or secondary hypogonadism, Klinefelter's
syndrome, congenital or acquired hypopituitarism, delayed or pre-
cocious puberty, male pseudohermaphroditism, etc.
Before discussing the effect of gonadotropins in children, it is

necessary to review briefly normal differentiation of the human tes-
tis and its relation to sexual development.
During early embryonic life, at approximately the sixth week,

the gonad in the male begins its differentiation into the testis
(10,11). The interstitial cells of the primitive gonad reveal marked
3 ~-ol-dehydrogenase enzymatic activity suggesting steroid synthesis
(12-l5), and simultaneously the embryonic cords, composed of primi-
tive coelomic cells and primary gonocytes, differentiate into semi-
niferous tubules. At the same time, testicular hormones induce mas-
393
394 C. BERGADA
culinization of the wolff ian ducts and external genitalia. All these
changes occur at the time of maximal development of Leydig cells
thereby demonstrating their active participation in the development
and differentiation of the testis and genitalia which terminates at
approximately the fourth to fifth month of pregnancy. From this time
on, the number of Leydig cells decreases but some remain until a few
weeks after birth when the stimulus of the chorionic gonadotropin
disappears (16). During infancy and childhood the interstitial tis-
sue is composed only of fibroblasts and for this reason the endocrine
function of the testis remains apparently inactive during this period.
The urinary excretion of testosterone glucuronide remains very low
before puberty, approximately from 5 to 11 ~ g/24 hrs. with no differ-
ence observed between boys and girls (17-20). Plasma testosterone
concentration has been recently determined by several authors with
somewhat similar results. Frasier and Horton (21) postulate that
testosterone production in children of either sex is low and plasma
concentration prior to the onset of puberty is not dependent either
on age or sex; their values being of 42± 9 m~g/lOO mI. in prepubertal
children. Saez and Bertrand (22) obtained values of 31.8 ± 10 m~g/
100 ml in normal boys aged from two to thirteen years. August et ale
(23) obtained 18 m~g/lOO mI. with a range between 10 and 75 m~g for
boys between one and twelve years. More recently, Rivarola et ale
(24) determined plasma testosterone levels by a competitive protein
binding method in children between four and eight years and obtained
values of 21.8 ± 28 ml-l-g/lOO ml and in boys aged between nine and
12 years the levels were 73.4 ± 44 ml-Lg/lOO ml.
With regard to the seminiferous tubules, after the complete

differentiation of the testis they increase their size and at birth
reach a diameter of approximately 501-1-. After birth, no important
changes are observed until puberty when hypophyseal gonadotropins
induce their development and stimulate the germinal epithelium (25).
During this prepubertal period, the tubules are composed of susten-
tacular cells, precursors of Sertoli cells, and germ cells. The
nuclei and cytoplasm of these spermatogonia acquire different as-
pects during fetal and postnatal life receiving different names by
some investigators (1,25-29) who attributed this to different stages
of cell activity. Two different types of Sertoli cells have also
been demonstrated by Johnsen (30). The significance of this is at
present difficult to explain. At approximately five to six years
of age, a tubular lumen appears without significant changes in the
germinal or Sertoli cells. Although this spermatogonial activity is
not really marked and no variation in the tubular diameter is ob-
served before puberty, whether or not the small amount of circulating
gonadotropins detected during infancy and childhood could be respon-
sible for this cellular process remains to be clarified. Both LH
and FSH have been detected in the urine of prepubertal children
either by biological or radioimmunological assays (31-34). Simultan-
eously several authors measured these gonadotropins in the plasma
of children by radioimmunoassay (35-41). Most of these investigators
gave similar results for LH plasma levels ranging between 1.5 and 15
mIU/ml in boys and girls under eight years of age. Yen et al. (40)
reports higher amounts for girls than boys after age eight. FSH was
found in much higher amounts than was LH (37,41).
As a consequence of the lack of change in the size of the semi-

niferous tubules and the inactivity of the fibroblasts of the inter-
stitial tissue, the size of the testes do not vary much from infancy
to puberty. At this time, and before any other sexual or somatic
signs appear, the testes increase their volume which coincides with
the rise of plasma and urinary gonadotropins and testosterone, and
usually precedes the growth spurt. Tanner clinically classified
the sexual development of boys in five stages (42): Stage 1: pre-
adolescent: testes, scrotum and penis are about same size and shape
as in early childhood. Stage 2: testicular enlargement begins with
some scrotal changes. Stage 3: penis is slightly enlarged, testes
and scrotum are further enlarged and pubic hair appears. Stage 4:
penis, testes and scrotum are further enlarged and sexual hair is
now adult in type but its area is still considerably smaller than in
the adult. Stage 5: the genitalia and sexual hair are adult in size
and shape. Johnsen (43) mentioned that he found the first appear-
ance of gonadotropins in the urine of boys two and a half years be-
fore their change of voice. He also pointed out that growth of the
testes comes first, followed by signs of androgen activity. Hudson
et ale (44) demonstrated a significant rise in plasma testosterone
levels about six months to one year prior to the beginning of sexual
maturation. On the other hand, Albert (45,46), in testes obtained
at autopsy, observed that the seminiferous tubules develop before
Leydig cell maturation. Mancini et ale (16,47) pointed out that
the first change observed in the testis is an active proliferation
of the interstitial fibroblasts, precursors of Leydig cells, and
then enlargement of the seminiferous tubules occurs followed by
maturation of the germinal epithelium and appearance of mature Ley-
dig cells. In a study of the urinary excretion of gonadotropins,
Rifkind et ale (32) observed that during childhood the testis is
exposed to some FSH stimulation. At puberty both FSH and LH increase
significantly, stimulating the seminiferous tubules and the inter-
stitial cells, respectively. More recently Yen et ale (41), studying
plasma gonadotropins by radioimmunoassay, detected a progressive rise
in LH and FSH concentration as sexual maturity was reached. The rise
in LH was more marked than in FSH. Therefore in boys he found that
the FSH:LH ratio decreases around puberty from 9 to 2 accompanied
by a progressive rise in plasma testosterone. Raiti et al. (34)
have recently correlated the plasma and urinary FSH determined by
radioimmunoassay with the clinical stages of sexual development
mentioned by Tanner (42). In the urine of preadolescent boys at
Stage 1, they found 2.2 IU of 2nd IRP-HMG ±
lS.D. 1.1 IU/24 hr. In
Stage 2, when testicular enlargement begins, the level was 4.4 ± 2.5
IU/24 hr and in Stage 3, with sexual hair and penis enlargement it
was 7.1 ± 3.0 IU/24 hr.
396 c. BERGADA
On the basis of all these findings we may conclude that growth
of the seminiferous tubules, stimulated by FSH which causes the in-
crease in the size of the testis, takes place before co~plete matur-
ation of the interstitial cells, induced by LH, and before full
androgen development occurs.
Administration of gonadotropins to prepubertal boys would permit

evaluation of the response capacity of an immature testis in order
to appreciate more specifically the role of FSH and LH on the semi-
niferous tubules and interstitial cells not previously exposed to
gonadotropic stimulation.
THE EFFECT OF HUMAN CHORIONIC GONADOTROPIN (HCG) ON THE PREPUBERTAL

TESTIS
Treatment of cryptorchidism with HCG has been widely used and in

the experience of more than 1500 patients treated in the Depart-
ment of Endocrinology of the Hospital de Ninos of Buenos Aires, 30
to 40% of the cases with unilateral or bilateral cryptorchidism were
cured with gonadotropin therapy. This is in accordance with data
reported in the literature. Clinical, endocrinological and histo-
logical studies of the testes which did not descend with this treat-
ment and underwent surgery, were considered an id~al situation in
which to evaluate the effect of different gonadotropins on the im-
mature testis.
Intramuscular administration of HCG at a dose of 1,000 to 1,500

IU weekly provokes, at any age, clinical signs of androgen activity,
such as congestion of the scrotum and penis and erections, after the
first or second week of treatment. This reaction is considered a
good sign to be taken into account when HCG is given to a boy with
no palpable testes since the lack of any response would suggest an
anorchidism (6-8). This effect d.emonstrates that the interstitial
cells are able to respond before puberty despite the age of the
child (1-7).
In 1967 Lipsett (48) observed that very young children are able
to increase their testosterone following HCG administration. Saez
and Bertrand (22) gave 1,500 IU of HCG every two days for 15 days
to 16 children aged from two to thirteen years, and the plasma tes-
tosterone level increased from 31.8 + 10 to 554 + 121 ml-L g/100 mI.
More recently Rivarola et al. (24) ~asured plas;a testosterone in
18 boys with cryptorchidism aged from four to twelve years before
and after daily injections of 800 to 5,000 IU of HCG for only five
days. From control values of 26.3 ± 27 m I-L g/lOO mI., the testosterone
rose to 378.9 + 188 ml-Lg/lOO mI. (range: 84-678). Some children
under six year; of age reached values above 500 ml-Lg/IOO mI. with
only 800 IU daily for five days. The same study performed in boys
with clinical signs of androgen activity such as slight testicular
enlargement or scanty pubic hair, revealed higher testosterone levels
in the basal state and a more marked response to HCG was observed
in some patients.
Intramuscular administration of HCG for evaluation of plasma

testosterone response should be given daily or every other day due
to the short life of this gonadotropin. Rizkallah et ale (49) gave
10,000 IU of HCG i.m. in a single injection and obtained a maximal
plasma concentration of this gonadotropin, 2,000 IU/ml, at 6 hr.
The half-life of the tail part of the curve was approximately 30 hr.
in their experiments. When 5,000 IU was injected daily for four
days, a maximal plasma concentration of HCG was noted each day, six
hours after injection. The values achieved ranged from 1200 to
1800 IU/ml.
Microscopic examination of the interstitial tissue of the testes

treated with short term HCG therapy, despite the androgen signs, did
not demonstrate full maturation of the Leydig cells, suggesting that
the precursor cells may be able to synthesize steroids (50). Histo-
logical studies performed in the scrotal testes of patients with
unilateral cryptorchidism treated with a prolonged series of gonado-
tropins were selected to evaluate the effect of these hormones in
the normal immature testis (51). Administration of HCG (Elea Lab-
oratories, Buenos Aires) to children between six and nine years of
age at a dose of 1,000 to 1,500 IU weekly for periods ranging from
three to six months, produced marked stimulation of the interstitial
tissue with development of immature and mature Leydig cells. Tubular
diameter was markedly increased to 70 or 90~. The normal median
for these ages is 53~. Thickness of the tubular wall was not ob-
served even in a case treated for six months. Germinal epithelium
was also markedly stimulated up to the appearance of primary sper-
matocytes, being more marked with the more prolonged treatment. The
same occurred with the Sertoli cells. Cytochemical and cytoenzymatic
studies revealed an increase in the enzymatic activity of the inter-
stitial cells, more marked in the scrotal than in the cryptorchidal
testes (52). The clinical signs of androgen activity observed were
congestion of the skin of the scrotum and penis, and erections. No
sexual hair appeared during or after treatment.
THE EFFECT OF PREGNANT MARE SERUM GONADOTROPIN (PMS) IN THE PREPUBERTAL

TESTIS
This preparation has seldom been used in children with cryptor-

chidism because it primarily exerts an FSH activity which does not
induce testicular descent. For this reason PMS is not indicated in
the treatment of cryptorchidism. We had the opportunity to study
three boys aged between six and nine years with unilateral cryptor-
chidism associated with visible hernia in whom surgical procedure
was indicated (51). PMS (Elea Laboratories, Buenos Aires) was given
in daily or weekly injections in total doses ranging between 7,500
and 28,000 IU over 15 to 60 days. Absolutely no clinical signs of
398 c. BERGADA
androgen activity were observed. Microscopic examination of the
testes revealed a slight increase in tubular diameter, up to 77~
in one case. Germinal epithelium was also slightly stimulated as
well as the Sertoli cells. In the nine year old patient treated
for 60 days some spermatocytes were observed. A very small effect
was seen in the interstitial cells. Appearance of anti-PMS in the
serum of these patients was not investigated.
EFFECT OF HCG PLUS PMS IN PREPUBERTAL TESTIS
Combined administration of both HCG and PMS provoked greater

effects than HCG alone (51). Five patients from six to nine years
of age were treated for one to six months with total doses of FSH
between 1,254 and 2,508 IU of 2nd IRP-HMG and of LH between 6,520
and 19,308 IU. The FSH and LH activity of the HCG used was checked
using the augmentation test and ascorbic acid depletion technique.
One IU of HCG contains 0.001 IU of FSH and 0.924 IU of LH (53). The
FSH:LH ratio was completely different with higher amounts of LH.
The FSH:LH ratio of 1:8 employed in three patients produced more
changes than ratios with less amounts of LH. Marked stimulation of
the Leydig cells, increased tubular diameter and marked stimulus of
the Sertoli cells and germinal epithelium up to spermatocytes were
seen. These changes were more marked when both gonadotropins were
administered simultaneously instead of subsequently. Again no
thickness or hyalinization of the tubular wall was observed. Clinical
signs were identical to the cases treated with HCG alone.
EFFECT OF HUMAN MENOPAUSAL GONADOTROPIN (HMG) IN PREPUBERTAL TESTIS
With the results previously mentioned, it was decided to employ

a preparation with both FSH and LH activity. In 1968, Rosemberg et
al.(54) reported the effect of Pergonal with different FSH:LHratios
in four boys with unilateral cryptorchidism aged six to seven years
who were treated for periods ranging from 16 to 75 days. FSH:LH
ratios varied from 1.2: 1 to 2 : 1. Stimulation of the interstitial
cells was observed, but more marked changes were seen at the semi-
niferous tubules. Tubular diameter increase up to 90 ~ and meiotic
spermatocytes reaching the pachytene stage were observed even in the
case treated for only 16 days. Since treatment was prolonged in the
longest case to two and a half months, it was questioned whether the
dose or duration of treatment was not sufficient to bring about full
spermatogenesis.
EFFECT OF HMG PLUS HCG IN THE PREPUBERTAL TESTIS
Several authors have reported better results in restoring sper-

matogenesis with the combination of HMG and RCG (55,53). In order
to induce a greater stimulation of the testes in children with
cryptorchidism, administration of both gonadotropin preparations with
varied FSH:LH ratios was performed in seven children with unilateral
undescended testis aged from six to nine years (51). Treatment

lasted from three and a half to eight months. Patients received
doses of FSH from 3~100 to 8,400 IU of 2nd IRP-HMG and of LH from
14,200 to 43,400 IU with FSH:LH ratios between 1 : 5 and 1 : 25.
The interstitial tissue showed marked stimulation with development
of mature Leydig cells comparable to adult testes. Tubular diameter
was markedly increased up to 110 ~ and active development of the
tubular wall was observed without signs of hyalinization. Germinal
epithelium showed more marked stimulation than with previous treat-
ments, as far as spermatogonia and spermatocytes are concerned,
since full spermatogenesis was not obtained in any case. Some sper-
matids were occasionally seen in the patient who received the high-
est dose of LH (43,380 IU) with 1,800 IU of FSH (FSH:LH ratio of
1 : 25) and who was treated for a period of six months. Another
patient treated for eight months with 37,968 IU of LH and 8,432 IU
of FSH, with a ratio of approximately 1 : 5, showed very marked
stimulation of the interstitial cells and germinal epithelium al-
though spermatogenesis was only developed up to the spermatocytes.
Plasma testosterone was markedly increased after several months on
treatment in accordance with the stimulation of the Leydig cells,
although the level reached was not so high as the values obtained
during the acute test of gonadotropin stimulation. In the two
patients mentioned above, plasma testosterone reached values of
450 and 620 m~g/IOO ml. respectively.
CONCLUSIONS
Administration of gonadotropins to prepubertal boys induced

changes in their testes which confirm previous concepts that FSH
stimulates the germinal epithelium and LH stimulates the intersti-
tial cells. However, the response capacity of these two structures
in the prepubertal testis varied greatly with that of the puber -
tal or adult testis.
These investigations have determined that the inactive fibro-

blasts of the interstitial tissue of the prepubertal testis are
capable of synthesizing testosterone within 48 to 72 hours after
the administration of daily injections of HCG, with a 15 to 20-fold
increase in plasma testosterone levels (22,24) even before they
differentiate completely to mature Leydig cells. At the onset of
puberty, when endogenous LH begins to increase and stimulates these
cells, testosterone production also increases and the exogenous
administration of HCG produces more marked response in plasma tes-
tosterone. In both instances mature Leydig cells and adult levels
of plasma testosterone could be induced.
The seminiferous tubules do not respond in the same manner.

These studies have determined that combined administration of small
amounts of FSH and LH are able to produce changes in the size of
the tubules and initiate spermatogenesis. However, in all our
400 C. BERGADA
studies performed in normal prepubertal boys using different ratios

of FSH and LH for prolonged periods of time up to eight months, full
spermatogenesis was not induced. It appears that the germinal epi-
thelium may be stimulated up to first meiotic division very easily,
but despite the dose of gonadotropin employed or the duration of
treatment, it was impossible to stimulate the spermatogenic process
any further. In only one case (51) a few spermatids were observed
with 43,380 IU of LH and 1,800 IU of FSH over six months. Perhaps
the ideal FSH:LH ratio has not been found, but it seems that the
prepubertal testis should acquire some sort of maturation associated
with adequate steroid synthesis and be stimulated by a proper FSH:LH
ratio in order to develop full spermatogenesis. Data reported here
suggest that in the immature human testis small amounts of FSH would
first stimulate the seminiferous tubules with minimal steroid func-
tion of the interstitial tissue, and then, gradual increase of both
FSH and LH with predominance of LH would produce active stimulation
of the Leydig cells whose steroids will contribute to complete sper-
matogenesis (already stimulated by FSH) from meiotic spermatocyte
up to spermatozoa.
Finally, it should be pointed out that in studies performed in

a group of adult men who were treated for cryptorchidism during
their childhood (56), all the patients with unilateral cryptorchidism
who received treatment with HCG and surgery with testicular biopsy
are fertile.
ACKNOWLEDGEMENT
This work was performed in part with grants from the Fundacion
de Endocrinologia Infantil and from Consejo Nacional de Investigaciones
Cientificas y Tecnicas, Buenos Aires, Argentina.
REFERENCES
1. Charny, C.W. and Wolgin, W. !E Cryptorchidism, Hoeber-Harper,

New York, p 80, 1957.
2. Loras, B., Ollagon, C. and Bertrand, J., Pediatrics 21: 455, 1966.
3. Paulsen, C.A., In Endocrinology, ed. Williams R.H., W7 B. Saunders
Co., p 441, 19687
4. Folk, R.L., Taulor, J.N., Sotos, J.F., Vorys, N. and Wieland, R.
G., Amer. J. Med. Sci., 225: 221, 1968.
5. Mancini, R.E., Cullen, M:;-Andrada, J.A., Vilar, 0., Lavieri, J.
C. and Saraceni, D.J., V Panamer. Congress of Endocrinology,
Lima, 1961, p 150 (Abstract).
6. Bergada, C., In Genital Anomalies, eds Nabil Rashad, M and
Morton, W.R.M:; Charles C. Thomas, Springfield, Ill. p 591, 1969.
7. Cullen, M., Bergada, C., Mora, H. andSoubie, C., Rev. Argent.
Endocrinol., Supple 1: 70, 1957.
8. Bergada, C., Cleveland, W.W., Jones, H.W., Jr. and Wilkins L.,
Acta Endocr. (Kobenhavn) 40: 493, 1962.
9. Bongiovanni, A.M., Pediatrics, 1i: 786, 1965.

10. Witschi, E., Carnegie Inst. Wash. Publ. 209, Contrib. to
Embryol., 32: 67, 1948.
11. Gillman, J7; Carnegie Inst. Wash.Publ. 209. Contrib. to
Embryol., 32: 81, 1948.
12. Bloch, E.,-rissenbaum, B., and Benirschke, K., Biophys.Acta 60
182, 1962. --
13. Baillee, A.H., Niemi, M. and Ikonen, M.,Acta Endocr. (Kobenhavn)
48: 429, 1965.
14. ~ldman, A.S., Yakovac, W.C. and Bongiovanni, A.M., J. Clin.
Endocr., 26: 14, 1966.
15. Schlegel.lR.J., Farias, E.,Russo, N.C., Moore, J.R. and Gardner,
L.I., Endocrinology, 81: 565, 1967.
16. Mancini, R.E., Vilar,-o., Lavieri, J. C., Andrada, J. A. and
Heinrich, J.J., Amer. J. Anat., l!1: 203, 1963.
17. Rosner, J.M., Conte, N.F., Briggs, J.H., Chao, D.Y., Sudman, E.
M. and Forsham, P.H., J. Clin. Endocr., ~: 95, 1965.
18. Vermeulen, A., !E Androgens, Vermeulen, A. and Exlev, D., eds.
Excerpta Med. Found. Amsterdam, 1966.
19. Knorr, D., Acta Endocr. (Kobenhavn) 54: 457, 1967.
20. Gupta, 0 and Butler, H., Steroids, 14: 343, 1969.
21. Frasier, S.D. and Horton, R., Steroids, 8: 777, 1966.
22. Saez, J.M. and Bertrand, J.,Steroids, 127749, 1968.
23. August, G.P., Tkachuk, M. and Grumback, M.M., J. Clin. Endocr.,
29: 891, 1969.
24. Rtvarola, M.A., Bergada, C., and Cullen, M., IX Meeting of the
Latinoamerican Society for Pediatric Research, Valdivia, Chile,
1969, J. Pediatrics (in press) (Abstract) 1970.
25. Albert, A., In Sex and Internal Secretions 3rd Ed. Williams and
Wilkins, Baltimore, p 305, 1961.
26. Evertt, N.B., Biol. Rev., 20: 45, 1945.
27. Roojen-Runge, E.C., Barlow, F.D., Amer. J. Anat., 21: 143, 1953.
28. Clermont, Y., Amer. J. Anat. 112: 35, 1963.
29. Mancini, R.E., Narbaitz, R. a~Lavieri, J.C., Anat. Rec.,136:
477, 1960. ---
30. Johnsen, S.G., Acta Endocrinol. (Kobenhavn), 61: 111, 1969.
31. Fitschen, W. and Clayton, B.E., Arch. Dis. child., 40: 16, 1965.
32. Rifkind, A.B., Kulin, H.E. and Ross, G.T., J. Clin.-rnvest., 46:
1925, 1967. --
33. Faiman, C., Ryan, R.J. and Albert, A., J. Clin. Endocr., 1&: 1078,
1968.
34. Raiti, S.M.B., Light, C. and Blizzard, R.M., J. C1in. Endocr.,
29: 884, 1969.
35. Odell, W.O., Rossi, G.T. and Rayford, P.L., J. Clin. Invest.,
46: 248, 1967.
36. ~halch, D.S., Parlow, A.F., Boon, R.C. and Reichlin, J., J.
Clin. Invest., 47: 665, 1968.
37. Taymor, M.L., A~o, T. and Pheteplace, C., !E Gonadotropins,
1968. ed. Rosemberg, E., Geron-X, Inc., Los Altos, Calif., p 349,
1968.
402 C. BERGADA
38. Raiti, S., Johanson, A., Light, C., Migeon, C.J. and Blizzard,
R.M., Metabolism, 1&: 234, 1969.
39. Saxena, B.B., Demura, H., Gandy, H.M. and Peterson, R.E., J.
Clin. Endocr. 28: 519, 1963.
40. Yen, S.S.C., V~ic, W.J. and Kearchner, D.V., J. Clin. Endocr.,
29: 382, 1969.
41. ¥;n. S., Vicic, W., Wieland, R. and Pohlman, C., Fifty-first
Meeting Endocrine Society, New York, 1969 (Abstract).
42. Tanner, J.M., In Growth at Adolescence, 2nd ed., Blackwell Sci.
Publ., Oxford,~ngland p 32, 1962.
43. Johnsen, S.G., ~ Gonadotropins 1968, ed. Rosemberg, E., Geron-X,
Inc., Los Altos, Calif. p 542, 1968.
44. Hudson, B., Coghlan, J.P. and Dulmanis, A., Ciba Found. Colloq.
Endocr. Vol. 16, p 140, 1967.
45. Albert, A., Mayo Clinic Proceedings, 28: 409, 557 and 698, 1953.
46. Albert, ~., Mayo Clinic Proceedings, 12: 131, 317 and 368, 1955.
47. Mancini, R.E., Cullen, M., Rosemberg, E., Lavieri, J.C., Vilar,
O. and Bergada, C., J. Clin. Endocr., ~: 927, 1965.
48. Lipsett, M.B., IV International Congress in Endocrinology,
Mexico, 1967.
49. Rizkallah, T., Gurpide, E., Vande Wiele, R.L., J. Clin. Endocr.,
29: 92, 1969.
50. Mlancini, R. E., ~ Gonadotropins 1968, ed. Rosemberg, E., Geron-
X, Inc., Los Altos, Calif. p 543, 1968.
51. Bergada, C and Mancini, R.E., in press.
52. Seilicovich, A and Perez Lloret, A., XIV Meeting of the Soc. Arg.
Invest. Clin., Bariloche, December 1969, Medicina, 1970 in press
(Abstract).
53. Mancini, R.E., Seiguer, A.C. and Perez Lloret, A., J. Clin.
Endocr., 12: 467, 1969.
54. Rosemberg, E., Mancini, R.E., Crigler, J.F. and Bergada, C. In
Gonadotropins 1968, ed. Rosemberg, E., Geron-X, Inc. Los Alt~,
Calif. p 527, 1968.
55. Paulsen, C.A., In Gonadotropins 1968, ed. Rosemberg, E., Geron-X,
Inc., Los Altos;-Calif. p 491, 1968.
56. Zamudio, R., Bergada, C. and Cullen, M.,Meeting of the Soc. Arg.
Invest. Clin., Bariloche, Medicina 1970, Chile, December 1969,
in press (Abstract).
DISCUSSION
SCHIRREN: Dr. Bergada, we know that in most cases of unilateral

cryptorchidism, the scrotal testis shows a histologic picture similar
to that of the undescended testicle. This has been our experience
in 300 cases, where we demonstrated in 80 percent of our cases hypo-
plasia of the scrotal testis. In evaluating results, one should con-
sider the age of the patients. For example, would a high dose of HCG
cause damage to the testes of a four year old child?
BERGADA: The lesion of the scrotal testis in unilateral cryptorchid-

ism in our experience is minimal in comparison with the total popu-
lation of cryptorchid children. We are studying the fertility pat-
tern in patients treated for cryptorchidism in our department for the
last ten years. Of 21 of our patients with unilateral cryptorchid-
ism treated before puberty, all but one are fertile at the present
time. With respect to the dosages of ReG administered; we did not
observe any damage as far as testicular histology is concerned.
CHAPTER V
TESTICULAR STEROIDOGENESIS
STEROID SECRETION BY THE HUMAN TESTIS
Mortimer B. Lipsett
National Cancer Institute, Endocrinology Branch,
National Institutes of Health, Bethesda, Maryland 20014
To prove. the secretion of a steroid, one must show that the

venous effluent from the gland has a higher concentration of the
steroid than has peripheral blood. This rigorous criterion is nec-
essary since there is conversion among steroids in peripheral tis-
sues as well as secretion by the adrenal cortex and gonads. There
are, of course, important inferences that can be obtained abou~ se-
cretion without sampling the venous effluent by devising appropri-
ate experimental conditions both in vitro and in vivo. For example,
the failure to demonstrate 11 ~-hydroxylation in the normal testis
and ovary under any conditions almost surely places the origin of
the 11~ -hydroxysteroids in the adrenal cortex. Suppression of
adrenal cortical function in men does not alter plasma testosterone
levels; thus testosterone must either be secreted by the testis or
be a product of peripheral metabolism of another steroid secreted
by the testis. This sort of inference has been made extensively and
has been confirmed on many occasions. It should be clear however
that the demonstration that a particular steroid is synthesized in
vitro or in vivo from a radioactive precursor gives no information
about secretion. One further caution is necessary; there are dif-
ferences in the testicular secretory patterns among species so that
extrapolation to man may not be warranted.
It is appropriate here to state two hypotheses that have not

as yet been contradicted. First, in normal men, testosterone is
the single important plasma androgen and second, the single impor-
tant function of the Leydig cell is the synthesis and secretion of
testosterone. Testosterone was isolated from the testes during the
infancy of steroid endocrinology (1) and 22 years later it was iden-
tified in human spermatic venous blood (2). Subsequently, there
have been many measurements of testosterone in spermatic venous
407
408 M. B. LIPSETT
blood (3-7). The data of Jeffcoate et ale (6) and Dupre et ale (5)
yield an average spermatic venous testosterone level of 50 ~g/100ml
with a very wide range. In older men, the level was 17 ~ g/100 ml
(7) again with a wide range. If testosterone secretion rates of 8
and 5 mg daily are assumed for young and old men respectively, then
an estimate can be obtained for testicular blood flow (Table 1).
The total testicular blood flow was estimated at 21 to 33 ml/min or
0.5 to 0.8 ml/g/min. These are reasonable estimates since Hudson
et ale (8) reported blood flows of 7 to 20 ml/min per testis in man,
and in the rat an adrenal cortical blood flow of 2 ml/g/min has been
reported (9). In the subsequent discussion a value of 25 ml/min for
testicular blood flow will be assumed.
Table 1. Testicular Blood Flow
Young Men Old Men
Testicular Venous Plasma T 50 ~g/100 ml 17 ~/100 ml

Hematocrit 47% 40%
Testicular Venous Blood T 27 ~g/100 ml 10 ~/100 ml
T Blood Production Rate 5.5 ~g 3.5 ~g/min
Total Testicular Blood 21 ml/min 33 ml/min
T = testosterone
The testis probably secretes testosterone in fetal life al-

though the evidence is entirely inferential. Fetal Leydig cells are
prominent from the ninth week of gestation in man (10); the human
fetal testis has the capacity to synthesize testosterone (11) and
testosterone produces development of external genitalia in the or-
chiectomized animal fetus (12). Within a few months of birth the
Leydig cells dedifferentiate not to appear again until shortly be-
fore the beginning of pubescence.
There are no data in man about the secretion of testosterone

by the immature testis. The studies of Lindner (13,14) have shown
that testes of the lamb, calf, and piglet secrete testosterone at
rates one-tenth to one-fiftieth those of the adult. Similarly Resko
(15) and we (16) found that the testes of the 20-day-old rat secrete
testosterone. In our study, a few atypical Leydig cells were pre-
sent at this timeo It may be surmised that in man, the testis se-
cretes small amounts of testosterone before the start of pubescence
since plasma testosterone levels (17) are higher in boys than in
girls. The identification of luteinizing hormone in the blood (18,
19) and urine (20) of prepubertal boys supports the contention that
there is a constant low level of a trophic hormone necessary for
Leydig cell function.
In the above discussion we have assumed that the blood produc-

tion rate of testosterone in men is a consequence solely of secreted
TESTICULAR STEROIDOGENESIS 409
testosterone.1 Since the testicular blood flow in man is not know from
independent measurements, the absolute secretion rate of testoster-
one cannot be calculated. However, quantitative studies of the per-
ipheral transformation of such possible precursors of testoster~ne
as androstenedione (21), 17-hydroxyprogesterone (22) dehydroepian-
drosterone (23) and its sulfate (24) make it unlikely that signifi-
cant amounts of plasma testosterone in normal men are derived from
any or all of these steroids (Table 2). Another possible precursor,
androstenediol~ has not been studied but the conversion of other
compounds such as dehydroepiandrosterone (23) and 17-hydroxypreg-
nenolone (2S) containing the tJ. 5_ 3 !3-hydroxy function to a tJ. 4_3_
ketosteroid is low so that it is improbable that androstenediol
could be an important peripheral precursor of plasma testosterone
in man.
Table 2. Testicular Secretion Rates and Transfer Constants
Testosterone mg/24 hr Fraction of

T blood production
rate (%)
Testosterone 7
Androstenedione 3 O.OS 2
Androstenediol ( 1-3 )>'<* (0.02-0.1)>'<>'< Small
17-Hydroxyprogesterone 2 <0.01 low
17-hydroxypregnenolone 1 < 0.01 low
Dehydroepiandrosterone 7 0.006 0.4
*Transfer constant of blood steroid precursor to blood testosterone

>'<*Best estimate
The steroid-synthesizing glands probably secrete each steroid

that lies in the biosynthetic sequence from pregnenolone to the
chief secretory product. It is our concept that this is an epiphe-
nomenon secondary to the process of steroid synthesis and has been
dubbed a consequence of the "leaky" gland by Short (26). The
lproduction rate signifies the total rate of entry of the steroid

from all sources into the blood compartment. This would include
secretion from the glands and entry from such. organs as liver and
skin. Secretion rate refers solely to the rate of entry of the
steroid from a gland.
2The following trivial names have been used for steroids:

Androstenedione = androst-4-ene-3,17-dione
Androstenediol = androst-S-ene-3 13,17!3 -diol
17-hydroxyprogesterone = 17a-hydroxy-SI3 -pregn-4-ene-3,20-dione
17-hydroxypregnenolone= 3 ~ ,17a-dihydroxypregn-S-ene-20-one
Dehydroepiandrosterone = 3 i3 -hydroxyandrost-5-en-17 -one
410 M. B. LIPSETT
amounts of steroids secreted will depend on such variables as the

activity of the several biosynthetic pathways and the velocity of
reaction of each enzymatic step of synthesis. Although steroid
synthesis will be discussed in a subsequent paper an outline of the
several steps in the biosynthetic sequences leading to testosterone
is shown in Fig. 1. We propose to discuss the secretion rates of
each of these steroids where data are available.
Autol.
o
t
Chol •• t.rol
cif
~H3
.rxY'
:
o
OH Oehydrottpitlndros'ttrontt
".9"8"olo"tt
CH,
I
(=0
~ Andro,tenadiel
I
o
17.hydro.llyproge.t.rone
ct:t
OH
o
Testo,t.rone
Androstenedione
Fig. 1 Pathways of testosterone biosynthesis in the Leydig cell.
Androstenedione. a proximate precursor of testosterone, is se-

creted by the Leydig cell. Since its spermatic venous concentra-
tion is about one-tenth that of testosterone in man (7) and in other
species (13,14), the secretion rate by the human testes would be
about 700 ~g/day. Using the transfer constant of 5.0% (10), only
35 ~g of the testosterone production rate would originate with tes-
ticular androstenedione. It is significant that in the calf (13),
piglet (14), and prepubescent rat (15) the spermatic venous andro~
stenedione level equaled the testosterone level. Thus, in the non-
stimulated state, androstenedione is as important a secretory pr~
duct as is testosterone. The change with pubescence then lies not
only with increased secretion of these steroids but in a major shift
in the relative amounts secreted. How the equilibrium of the andro-
stenedione-testosterone 17-oxido reductase system is affected by
pubescence is not clear.
Androstenediol, the other possible proximate precursor of tes-

tosterone, has not been measured either in peripheral or spermatic
venous blood. Androstenediol was identified in canine spermatic
vein blood after HCG stimulation at levels 1/2 to 1/20 those of
testosterone (27). If a similar concentration were present in man,
the secretion rate could be several milligrams daily. Since the
transfer constants for the conversion of dehydroepiandrosterone to
androstenedione (6%) (23) and 17-hydroxypregnenolone to 17-hydroxy-
progesterone (2-11%) (25) are low, it is unlikely that androstene-
diol could be an important precursor of plasma testosterone in man.
There is considerably more information about the testicular

secretion of another 6 5-3~-hydroxysteroid, dehydroepiandrosterone.
It was identified in canine testicular venous effluent by Ibayashi
et ale (28). Its secretion in man was established by Gandy and
Peterson (7) who reported a spermatic venous gradient ranging from
0.13 to 11 ~g/100 ml, the mean being approximately 1.3 ~/100 mi.
Using the earlier estimates of testicular blood flow, the estimated
secretion rate would be about 400 ~g/24 hr or 6% of the estimated
dehydroepiandrosterone blood production rate. Thus, their finding
that orchiectomy did not alter plasma dehydroepiandrosterone levels
is in accord with these estimates since a small decrease would not
have been detected. Rivarola et ale (29) reported considerably
higher gradients of dehydroepiandrosterone in patients with testi-
cular feminization. If their data were accepted, then the testes
in these subjects would be the most important source of dehydroepi-
androsterone.
17-hydroxyprogesterone is a precursor of androstenedione in

the testis and there is now considerable evidence bearing on its
secretion rate. Its blood production rate was estimated at 2 mg/24
hr (30). Since suppression of LH by an exogenous androgen reduced
plasma levels and production rates by 90%, and adrenal cortical
suppression had little effect, it was concluded that the testis was
the source of most of the plasma 17-hydroxyprogesterone (30). The
possibility that 17-hydroxypregnenolone was secreted in large enough
amounts and converted peripherally to 17-hydroxyprogesterone to ac-
count for its production rate was ruled out by studies showing that
no more than 200 ~ could originate in this way (25). Thus, there
must be a high testicular secretion of 17-hydroxyprogesterone. The
older findings of an increase in urinary pregnanetriol in response
to HCG (31) suggested that the origin of this metabolite was 17-
hydroxyprogesterone secreted by the testis.
17-Hydroxypregnenolone may be the precursor of 17-hydroxypro-

gesterone or dehydroepiandrosterone within the Leydig cell or after
secretion in peripheral tissues. We have estimated that only 30-
40% of the blood production rate of 2 to 3 mg daily is secreted by
the testis (25). The larger moiety is clearly due to adrenal corti-
cal activity.
412 M. B. L1Psm
Pregnenolone and progesterone have not been measured in sper-

matic venous blood and there is no evidence regarding their secre-
tion. Progesterone plasma levels are so low in men (30) that sup-
pression studies are difficult to interpret. Our preliminary esti-
mates of plasma pregnenolone concentration show it to be present at
about 60 ng/l00 ml so that the amount secreted by the testis must
be small. The secretion rates of the steroids by the testis and
their transfer constants to testosterone are summarized in Table 2.
CONJUGATES
The roles of steroid sulfates as intermediates in the biosyn-

thetic process and as secretory products of the adrenal cortex (32,
33) led to similar studies of the testis. Although a sulfokinase
has been identified in the human testis (34), the secretion of dehy-
droepiandrosterone sulfate has not been demonstrated conclusively.
In the syndrome of testicular feminization, Morris and Mahesh (35)
and Pion et ale (36) found high spermatic venous levels of the sul-
fate whe~eas Rivarola et ala (29) and Lastikainen et ale (37) re~
ported similar levels in peripheral and spermatic venous blood. In
one in vivo perfusion of a normal human testis, the secretion of
dehydroepiandrosterone sulfate was not shown (38). The best evi-
dence for secretion of this conjugate is the finding that HCG treat-
ment for 14 days increased peripheral dehydroepiandrosterone sulfate
levels threefold in young children (39) but this stimulation did not
occur in older children (30) or in adults (40). These data and the
findings in testicular feminization suggest that the testis has the
capacity to secrete dehydroepiandrosterone sulfate but that in nor-
mal men, the secretion rate is very low.
The human testis has the capacity to synthesize testosterone

sulfate (34). Laatikainen et ale (37) found the spermatic venous
plasma level to average 1.5 ~g/100 mI. This would be equivalent to
a secretiort rate of 270 ~/24 hr. The peripherial testosterone sul-
fate concentration was about 0.1 ~g/100 ml (37,41). Since Baulieu
et ale (33) reported that testosterone sulfate is not metabolized
further, it should not be a precursor of plasma testosterone. The
identification of testosterone sulfate in the urine in very small
amounts (42,43) suggests that perhaps the route of excretion is oth-
er than the urine if no metabolism is possible.
Androstenediol monosulfate has been identified in peripheral

blood (44) at a level of several micrograms per 100 mI. The origin
of this steroid, a putative precursor of testosterone, has not been
examined.
ESTROGENS
The complex problems associated with the secretion of estrogens

by the testis are derived from observations that HCG increased urin-
ary estrogens in men (45-49) and that castration caused a marked

fall in their excretion. These data proved that the testis secreted
estrogens or their precursors. The synthesis of estrogens from ac-
etate (50,51~,cholesterol (52), testosterone (52) and androstene-
dione (53) has been demonstrated in the testes of several species
including the human. In all cases, the conversions even from the
immediate precursors, such as androstenedione, were low. The iden-
tification of estrone and estradiol in spermatic venous blood from
the normal human testis at concentrations higher than in peripheral
blood has not been made. The early observations of Steinach and
Kun (54) and the later studies of West et al. (55) showed that tes-
tosterone could be converted to estrone and estradiol in man. There
were many subsequent reports of the conversion of radioisotope la-
belled androgens to urinary estrogens (56-59) and to blood estrogens
(60,61). The sites of conversion were not defined by these studies
but human liver (62,63) has been shown to synthesize estrogens from
androgens in vivo. There is suggestive evidence that human mammary
cancer (64) and rat liver and kidney (65) may also be able to aro-
matize androstenedione to estrone. Since aromatization of androe
stenedione took placecin an ovariectomized adrenalectomized woman
(59), it is not necessary to postulate that the process of aromati-
zation of androgens normally occurs in the adrenal cortex or gonads.
MacDonald et ale (59) reported the first complete quantitative

studies of the source of estrogens in men. After infusing tritiated
androstenedione and labelled estrone and measuring the conversion of
androstenedione to urinary estrone, they concluded that 1.3% of
blood androstenedione was converted to blood estrone, thereby ac-
counting for 18 ~g or an important fraction of the daily estrone
production rate. Longcope et al. (61-) and we (66) found essential-
ly the same transfer constants (1.5%) between blood androstenedione
and estrone. From our estimates of blood androstenedione produc-
tion rates, we (61,66) concluded that 30 to 40 ~g of estrone en-
tered the blood compartment daily as ~ result of conversion from
blood androstenedione. Since the metabolic clearance rate of es-
trone is about 2000 1/24 hr (67,68) and the 9:00 a.m. plasma estrone
concentration was 6 ng/l00 ml (69), the calculated blood estrone
production rate was 120 ~g/24 hr. However, there is a diurnal va-
riation of blood estrone in men (69) mQst probably related to the
diurnal variation of androstenedione secreted by the adrenal cortex
(70). Thus a mean 24 hour plasma estrone level would be closer to
4 ng/l00 ml and the 24 hour production rate about 80 ~g. If this
is the case then androstenedione blood production would account for
one-half of estrone production.
The possibility then remains that some estrone is secreted by

the testis. Baird (71) demonstrated that there is a small adrenal
cortical secretion of estrone but this could account for only a
small fraction of the estrone blood production rate. Other possi-
ble precursors such as dehydroepiandrosterone are unimportant
414 M. B. LIPSETT
sources of estrogens in men (72). We would thus conclude that the

testis secreted about 40 ~ of estrone daily. From the estimate of
testicular blood flow, a plasma estrone concentration of about 0.3
~g/100 ml should be present in spermatic venous blood. Methods of
requisite sensitivity have not been applied as yet to this problem.
Only in testicular feminization have estradiol and estrone been iden-
tified as testicular secretory products (36,73) and the testis in
this syndrome can scarcely be equated with the normal testis.
The conclusion that the testis secretes estrone depends criti-

cally on the estimate of the estrone blood production rate. Data ob-
tained by several workers using urinary isotope dilution methods place
the estrone production rate considerably lower than that estimated
from the blood level and clearance rates. Based on these lower esti-
mates, the testicular estrone secretion would be negligible. This
question will be settled by analysis of spermatic venous estrogen
levelso
The contribution of blood testosterone to blood estradiol has

been examined similarly, and the transfer constant was considerably
lower, values of 0.4% (59) and 0 0 6% (66) being reported. Using a tes-
tosterone blood production rate of 7 mg/24 hr., we estimate that 35 ~g
pf estradiol is derived from blood testosterone. In addition, the es-
trone-estradiol blood transfer constant is 15% (67) so that an addi-
tional 12 to 15 ~g of estradiol would originate from plasma estrone.
This accounts for all of the estimated estradiol production in men.
Thus, there is no need to assume appreciable secretion of estradiol
by the normal testis. These data are summarized in Fig. 2 and 3.
8% III
T ~
7 mg " 3% 3 mg
~ 0.5 0/0 ~ 1.5 0/0
E2
15 ,"0
... El
"C
32~g 5% 80~g
Fig. 2 Transfer constants and blood production rates of estrogens

and their precursorso The numbers within the blocks give the esti-
mated daily blood production rates and the transfer constants between
compartments are shown with the appropriate arrowso
Blood
Production
Rale
Jl.g/24 hr
20
Figo 3 Contribution of androgens to estrogen production rates in men.

The amount of estrone derived from blood androstenedione and estradiol
is shown within the total estrone production. All of the blood estra-
diol is derived from blood testosterone.
In summary then, there is no evidence that the testis secretes

any steroid other than testosterone in amounts sufficicient for sig-
nificant biologic effects. None of the steroids secreted can be con-
verted with high enough efficiency in target tissues to exert impor-
tant androgenic or estrogenic activit yo It is probable that estradiol
is not secreted by the testis whereas estrone may be. All of the
intermediates in steroid biosynthesis may escape the biosynthetic pro-
cess and thereby be secreted by the testis.
REFERENCES
1. David, K., Dingmanse, E., Freud, J., and Laqueur, E., Physiol.
Chern., 233: 281, 1235.
2. Lucas, W.M., Whitmore, W.F., Jr., and West, CoD., J. Clin. Endocr.,
..!2: 465,1957.
3. Hollander, N., and Hollander, V.P., J. Clino Endocro, ~: 966,
19580
4. Hudson, B., Coghlan, J., DuJmanis, A., Wintour, M., and Ekkel,
I., Aust. J. Exp. BioI. Med. Sci o, 41: 235, 1963.
5. Dupre, J., Brooks, R.V., Hyde, R.D., London, D.R., Prunty, F.T.G.,
and Selp, JoB., J. Endocr., 29: vii, 19640
6. Jeffcoate, SoL., Brooks, R.V., Lim, N.Y., London, D.R., Prunty,
F.T.G., and Spathis, G.S., Jo Endocr., 12: 401, 1967.
416 M. B. LIPSETT
7. Gandy, H.M. and Peterson, R.E., J. Clin. Endocro, 28: 949, 19680
8. Hudson, B., Coghlan, J.P., Dulmanis, A., and Wintour, M., Proc.
Second Intern. Congr. Endocr., 2: 1127, 1964.
9. Kramer, R.J. and Saperstein, L.A., Endocrinology, ~: 403, 1967.
10. Van Wagenen, G. and Simpson, M.E., Embryology of the Ovary and
Testis, Yale University Press, New Haven, 1965.
11. Bloch, E., Endocrinology, 74: 833, 1964.
12. Jost, A., Recent Progr. Ho~one Res., ~: 379, 1953.
13. Lindner, H.R., J. Endocr., 23: 139, 1961.
15. Resko, J.A., Feder, H.H. and Goy, R.W., J. Endocr., 40: 485,
1968.
16. Knorr, D., Vanha-Perttula, T., and Lipsett, M.B., Endocrinology,
in press.
170 Fr~sier, S.O. and Horton, R., Steroids,~: 777, 1960.
18. Johanson, A.J., Guyd'a, H. Light, C., MigeIDn, C.J., and Blizzard,
R.M., J. Pediat., 74: 416, 1969.
19. Sciarra, N., Leone:-V., and Pastorino, G.F., J. Endocr., 42:
167, 1968. --
20. Rifkind, A.B., Kulin, H.E., and Ross, G.T., J. Clin. Invest.,
46: 1925, 1967.
21. Horton, R. and Tait, J.F., J. Clin. Invest., 45: 301, 1966.
22. Camacho, A.M. and Migeon, C.J., J. Clin. Invest., 43: 1083,
1964. --
23. Horton, R. and Tait, J.F., J o Clin. Endocr., lZ: 79, 1967.
24. Slaunwhite, W.R., Jr., Burgett, M.J., and Sandberg, A.A.,
J. Clin. Endocr., 27: 663, 1967.
25. Bermudez, J.A., Strott, C.A., and Lipsett, M.B., Clin. Res.,
17: 587, 1969.
26. Short, R.V., Biochemical Soc. Symposia, 1&: 59, 1960.
27. Yamaji, T., Motohashi, K., Tanioka, T., and Ibayashi, H.,
28. Ibayashi, H., Nakamura, M., Uchikawa, T., Murakawa, S. Yoshida,
S., Nakao, K., and Okinaka, S., Endocrinology, 2£: 347, 1965.
29. Rivarola, M.A., Saez, J.M., Meyer, W.J., Kenny, F.M., and Migeon,
C.J., J. Clin. Endocr., 27: 371, 1967.
30. Strott, C.A., Yoshimi, T., and Lipsett, M.B., J. Clin. Invest.,
48: 930, 1969.
31. Landau, R.L. and Lanes, M.L., J. Clin. Endocr., 11: 1399, 1959.
32. Vande Wiele, R.L., MacDonald, P.C., Gurpide, E., and Lieberman,
S., Recent Progr. Hormone Res., 19: 275, 1963.
33. Baulieu, E.E., Corpechot, C., Dray, F., Emiliozzi, R., Lebeau,
R., Mauvais-Jarvis, P., and Robel, P., Recent Progr. Hormone
Res., l!.: 411, 1965.
34. Dixon, R., Vincent, V., and Kase, N., Steroids, ~: 757, 1965.
35. Morris, J. MeL. and Mahesh, V., Amer. J. Obsfet. Gynec., 87:
731, 1963.
36. Pion, R.J., Dignam, W.J., Lamb, E.J., Moore, J.G., Frankland,
M.V., and Simmer, H.H., Amer. J. Obstet. Gynec., 93: 1067, 1965.
37. Laatikainen, T., Laitinen, A., and Vihko, R., J. Clin. Endocr.,
29: 219, 1969.
38. Oertel, G.W., Wendlberger, F., Menzel, P., Treiber, L., and
Knapstein, P., Europ. J. Steroids, ~: 299, 1967.
39. Saez, J.M. and Bertrand, J., Steroids, 1l: 749, 1968.
40. Yamaji, T., and Ibayashi, H., J. Clin. Endocr., 29: 273, 1969.
41. Dray, F., Mowszowicz, I., and Ledru, M.J., Steroids, 1Q: 501,
1967.
42. Dessypris, A., Drosdowski, M.A., McNiven, N.L. and Dorfman,
R.I., Proc. Soc. Exp. Bioi. Med., 121: 1128, 1966.
43. Tamm, J., Volkwein, V., and Voigt, K.D., Experientia, 23: 299,
1967.
44. Sjovall, J. and Vihko, R., Steroids, 2: 447, 1966.
45. Leach, R.B., Maddock, W.O., Tokuyama, I., Paulsen, C.A., and
Nelson, W.O., Recent. Progr. Hormone Res., 1l: 377, 1956.
46. Jayle, M.F., Baulieu, E.E., Scholler, R., Lisboa, B.P., and
Savoie, J.C., Rev. Franc. Etud. Clin. Bioi., 3: 455, 1958.
47. Givner, M.L., Bauld, W.S., Hale, T.R., Vagi, K., and Nilsen,
M., J. Clin. Endocr., 20: 665, 1960.
48. Morse, W.I., Clark, A.F., MacLeod, S.C., Ernst, W.A., and
Gosse, C.L., J. Clin. Endocr., ~: 678, 1962.
49. Jayle, MLF., Scholler, R., Sfikakis, A., and Heron, M., Clin.
Chim. Acta, 2: 212, 1962.
50. Rabinowitz, J.L. and Oleksyshyn, 0., Arch Biochem. Biophys.,
64: 285, 1956.
51. Rabinowitz, J.L. and Dowben, R.M., Biochim. Biophys. Acta,
1i: 96, 1955.
52. Rabinowitz, J.L., Atompraxis, ~: 85, 1958.
53. Bedrak, E. and Samuels, L.T., Abstracts Third Intern. Congr.
Endocr., 1: 35, 1968.
54. Steinach,-E., and Kun, H., Lancet, ~: 845, 1937.
55. West, C.D., Damast, B.L., Sarro, S.D., and Pearson, O.H.,
J. Bioi. Chern., 218: 409, 1969.
56. Braun-Cantilo, J.A., LaRoche, G., Novitsky, M., and Lawrence,
J.H., Acta Isotop., 1: 351, 1961.
57. Breuer, H., Acta Endocr., 40: 111, 1962.
58. Epstein, B.J., Raheja, M.C., Prow, E., and Morse, W.I., Canad.
J. Biochem., 44: 971, 1966.
59. MacDonald, P.C., Rombaut, R.P., and Siiteri, P.K., J. Clin.
Endocr., 27: 1103, 1967.
60. Fishman, L:M., Sarfaty, G.A., Wilson, H., and Lipsett, M.B.,
Ciba Found. Colloq. Endocr., 1i: 156, 1967.
61. Longcope, C., Kato, T., and Horton, R., J. Clin. Invest., 48:
2191, 1969.
62. Bolt, W., Ritzl, F., and Bolt, H.M., Munchen Med. Wschr.,
108: 875, 1966.
63. Bolt, W., Ritzl, F., and Bolt, H.M., Verh. Deutsch Ges. Inn.
Med. , 72: 461, 1966.
64. Adams, ~B., and Wong, M.S.F., J. Endocr., 41:41, 1968.
65. Frieden, E.H., Patkin, J.K., and Mills, M., Proc. Soc. Exp.
Bioi. Med., 129: 606,1968.
41 B M. B. LIPSETT
66. Hembree, W.C., Bardin, C.W., and Lipsett, M.B., Abstracts

Third Intern. Congr. Endocr., 1: 181, 1968.
67. Longcope, C., Layne, D.S., and Tait, J.F., J. Clin. Invest.,
47: 93, 1968.
68. Hembree, W.C., Bardin, C.W., and Lipsett, M.B., J. Clin. Invest.,
48: 1809, 1969.
69. Baird, D.T. and Guevara, A., J. Clin. Endocr., 29: 149, 1969.
70. Crafts, R., Llerena, L.A., Guevara, A., Lobotsky, J., and Lloyd,
C.W., Steroids, 12: 151, 1968.
71. Baird, D.T., Uno-,-A., and Melby, J.C., J. Endocr., 45: 135,
1969.
72. Baird, D.T., Horton, R., Longcope, C., and Tait, J.F., Recent
Progr. Hormone Res., 25: 611, 1969.
73. French, F.S., Baggett-,-B., Van Wyk, J.J., Talbert, L.M.,
Hubbard, W.M., Johnston, F.R., Weaver, R.P., Forchielli, E.,
Rao, G.S., and Sarda, I.R., J. Clin. Endocr., 25: 661, 1965.
DISCUSSION
JOHNSEN: Dr. Lipsett, I would like to comment on your statement

that most or all of the estrone/estradiol in men is derived from pe-
ripheral conversion of androgens. I would say, that if there were
only one significant testicular hormone we would be unable to under-
stand testicular function since we have two different gonadotropins
both being regulated more or less independently by feedback mechan-
isms from the testis. Next, if most or all of the estrogens were
derived from the androgens, one would would expect to find a certain
correlation between the levels of these two hormones in the male. We
have been unable to find such a correlation. In 29 normal men aged
20 to 40 years we performed, under continuous suppression of the
adrenals with dexamethasone, a HCG stimulation test and studied the
rise of urinary estrone/estradiol and the rise of urinary androsteronel
etiocholanolone which we believe to be a good parameter for testicular
androgen during adrenal suppression. We have made a plot of the log
value of the rise of androsterone/etiocholanolone (i.e. the stimula-
tion value minus the value before stimulation) versus the log value
of the rise of urinary estrone/estradiol. There was no correlation.
The correlation coefficient was 0.28 (P value >0.1). This is sur-
prising because both parameters show a wide variation so that a cor-
relation could easily show up. At one end we have a man with a 160%
rise in androgen but only a 36% rise of estrogen. At the other end
there is a man with only a 38% rise of androgen but a 157% rise of
estrogen. For statistical reasons we have excluded three extreme
values. If these are included the correlation coefficient drops to
0.14 and there is one man with 104% androgen rise but only 14% estro-
gen rise and another with only 14% androgen rise but no less than
2l~!. estrogen rise. One' may speculate as to where the estrogens
came from. We think that the lack of correlation between androgen
and estrogen levels in the male speaks against the concept that most
or all of the estrogen is derived from peripheral conversion of an-

drogen. We even think that the lack of correlation, particularly
between the rise of androgen and estrogen after HCG,is an indication
that these two may not have the same testicular cellular origin. Dr.
Lipsett, did you ever try to correlate the levels of androgen and
estrogen in the male?
LIPSETT: No, we have not tried to correlate the levels of androgen

and estrogen. Very rough correlations have been obtained by many
workers showing that when the level of androgen is low the level of
estrogen is low and so on. In terms of looking at correlations be-
tween plasma testosterone and plasma estradiol: this has not been
done. I cannot comment on the question of two hormones of the tes-
tis. However, with estrone and estradiol there are at least three
groups, all of whom have essentially the same data, which show all
of plasma estradiol can be accounted for by conversion from testos-
terone. Now, as I indicated, the picture with estrone is not nearly
as clear, because it depends not only on some problems of measure-
ment of secretion and production rates but also on what figure one
uses for the androstenedione level. There is a diurnal variation of
androstenedione much of it coming from the adrenal cortex. In six
cases of feminizing testis syndrome, secretion of either estradiol
or estrone has in fact been shown.
TROEN: In covering the steroid secretion very nicely, I wonder

whether Dr. Lipsett would also care to discuss briefly the role of
conjugated steroids in biosynthetic processes in the testis. Dr.
Uanaihara and I have been studying the synthesis of testosterone in
minces of testicular tissue obtained at orchiectomy for prostatic
cancer, during herniorrhaphy or because of severe testalgia. One
gram of tissue was incubated for one hour in a standard incubation
mixture to which labelled steroid precursors were added. After in-
cubation with DHA, identification of the metabolites revealed the
major free compounds to be ~5-androstenediol and testosterone, with
only a small amount of ~4-androstenedione. In the approximately 5%
conju~ated fraction, with labelled dehydroepiandrosterone sulfate
and 6 -androstenediol sulfate, additional incubations with mixed
precursors were performed. When labelled t5-androstenediol was in-
cubated with either labelled DHA or DHA sulfate, b 5-androstenediol
sulfate was found from all three. The tritium/Cl4 ratios indicate
that ~ 5-androstenediol sulfate is formed more readily from ll5-andros_
tenediol than from either DHA or DHA sulfate. From mixtures of DHA
and DHA sulfate the ll5-androstenediol sulfate preponderantly comes
from DHA. In addition there is far greater conversion of DHA to ,,;5
androstenediol than DHA sulfate to ~S-androstenediol sulfate. These
findings indicate that the pathway is DHA to ~ 5-androstenediol to
A5-androstenediol sulfate. To determine the role of t 5 -androstenediol
sulfate in the biosynthesis of testosterone, particularly since no
testosterone sulfate was identified, additional studies were performed.
The effect of added carrier ll5-androstenediol or ~ 5-androstenediol
420 M. B. LIPSETT
sulfate on the conversion of labelled DRA to testosterone was deter-

mined. The addition of the conjugated material had no inhibitory
effect on the conversion of DRA to testosterone whereas a progressive
and marked inhibition of conversion resulted from the addition of the
free compound. These data suggest that A5 -androstenediol sulfate may
function as an available reservoir but not as a significant precursor
in the biosynthetic pathway to testosterone.
LIPSETT: Well, it is clear that there is a sulfokinase in the testis

and sulfation of testosterone has been clearly shown by Mason and
Payne. DRA sulfate has been synthesized by the testis. DRA sulfate
has been shown to be a poor precursor by Eik-Nes and his co-workers
on perfusion of the testis, however, the fact is that no matter what
one says about the biosynthetic sequence, DRA sulfate does not seem
to be a secretory product. There is a report by Saez and Bertrand
concerning children who, when treated with HeG for 14 days, show a
rise in DRA sulfate, whereas it does not increase in the adult. You
mentioned that you had some difficulty showing synthesis of testos-
terone sulfate, but I think the data of Vihko and Laatikinin using
both gas chromatography and mass· spectrometry very clearly show its
secretion. Thus we may always find some discrepancies between a
particular set of in vitro incubation results and measurement of a
steroid in peripheral blood or in the venous effluent from a gland.
ROSEMBERG: I am very glad that both Dr. Lipsett and Dr.Hudson brought
up the question of diurnal variation of testosterone correlated to
levels of plasma LH as measured by radioinnnunoassay. Dr. Hudson
mentioned the report from Dr. Saxena's group which points to some
diurnal variation in testosterone level. I would like to ask both
Dr. Lipsett and Dr. Hudson if there is enough data from various
laboratories which will indicate one way or another that there is
indeed some diurnal variation of LH levels, as measured by radio-
innnunoassay and if there is some variation in testosterone levels
at least in a 24-hour period. Dr. Lipsett, when you presented the
effects of HeG, how much HeG was given and how soon did you record
the increase in testosterone levels?
LIPSETT: It is now the consensus that there is a small decrease in

plasma testosterone in the evening as compared to early morning. I
would agree that this occurs in some patients and not in others, and
when one averages them out one finds some small decrease. No one
has found the large decreases that were reported by Dr. Saxena.
HUDSON: We first reported on the study of the diurnal variation of

testosterone in 1964 in an experiment done with Dr. Bartter when he
was visiting our laboratories. We used nine normal male medical
students, and when we averaged the data we were unable to find any
circadian rhythm in plasma testosterone. Because a number of groups,
in particular Eik-Nes and Resko, Drey and others found evidence to
the contrary, we repeated this experiment with another group of med-
ical students. On this occasion, LH was also estimated simultane-

ously. Analysis of this data showed a small diurnal variation with
high morning and low evening values for both LH and testosterone in
some individuals only. The changes in LH preceded those for testos-
terone and were of a magnitude of from plus to minus 17% around the
mean. I believe we must look at this data realistically. We are
asking quite a lot from our methods, I believe,that methods can mea-
sure 10 to 15% changes in these compounds. However, despite all the
care that is taken in the laboratory, within assay variation can be
at least plus or minus 10%, and this is what I mean by methodological
"noise" and biological "signal."
TESTOSTERONE PLASMA LEVELS IN NORMAL AND PATHOLOGICAL CONDITIONS
Bryan Hudson, H.G. Burger, D.M. de Kretser,
J. P. Coghlan and H.P. Taft

Departments of Medicine and Anatomy, Monash University,
Medical Research Centre Prince Henry's Hospital and
Howard Florey Laboratories of Experimental Physiology,
University of Melbourne, Melbourne, Australia
The purpose of this paper is to review what is now a fairly

substantial body of knowledge about the concentration of testos-
terone in human plasma both in normal and abnormal conditions.
For the most part, this review will be directed to testosterone in
the male; however, other closely related steroids such as andros-
tenedione and dehydroepiandrosterone (DHEA) will be discussed in
so far as these are relevant; likewise, plasma levels of all these
steroids in the female will be discussed particularly when it may
be appropriate to compare and contrast them between the two sexes.
This presentation will thus aim to synthesize this knowledge, and
will draw upon many contributions from different laboratories,
including our own.
The reason why there is now such a substantial body of know-

ledge is undoubtedly related to the existence of methods for the
measurement of testosterone in plasma and other biological fluids.
This paper cannot possibly review the many methods that now exist
(1 - 14), but clearly methods that depend upon the interaction
between testosterone and macromolecules have now displaced earlier
procedures because of their relative simplicity, adequate speci-
ficity and reasonable precision.
TESTOSTERONE IN NORMAL ADULT MALES
In the normal, sexually mature, adult male, there is good

agreement that the concentration of testosterone in peripheral
plasma is between 3.5 and 10.5 ~g/l. These data, which are not
423
424 B. HUDSON ET AL.
generally disputed, have been derived from many different labora-

tories using most of the methods that have been mentioned. In
round numbers, these values are about 10 times those found in the
normal adult female.
About 10 years ago Slaunwhite and Sandberg (15) and later

Chen et al. (16) suggested that testosterone was bound to plasma
proteins. Since then, many workers have independently confirmed
these observations and have done much to characterize the nature
of this binding protein (17-23). Although testosterone in plas-
ma is bound to albumin, this binding is relatively weak and non-
specific in comparison with the binding to what is probably a ~
globulin which has a high affinity not only for this steroid but
also for estradiol, with which it probably shares a common bind-
ing site (or sites). Rivarola et al. (23) have shown by in vitro
studies that in the male 92.~ 1.6% is bound, a figure that is
significantly different from the female in whom about 95.5+ 1.2%
is bound. Treatment with estrogens, and pregnancy, further in-
crease this binding, so that a value of 9~ 0.2% is observed
under these conditions. The biological implication of these in
vitro observations is that in the male, the unbound and presumably
active form of testosterone circulates in a concentration of about
80 ng/l. The comparable value for the adult female is between 4
and 5 ng/l. Extensive studies on the nature of this testosterone
binding affinity have also been made by August et al. (24) who
have shown that a fairly wide spectrum of affinities exists under
physiological conditions; pregnancy plasma showing the highest
affinity and umbilical venous plasma the lowest.
Blood levels of testosterone reflect the momentary balance

between entry into and removal from the blood. In the male, about
1000 liters of plasma are cleared of testosterone each day. This
figure is significantly correlated with liver blood flow; the fact
that it is only about 60-70% of liver plasma flow is presumed to
reflect the degree to which the steroid is normally bound to pro-
tein, a presumption which is supported by the finding that in the
female, in.whom protein binding is greater, the clearance rate is
less, and in both sexes may be further reduced by treatment with
estrogen (25-30).
Although the liver is undoubtedly the principal organ respon-

sible for the metabolic removal of testosterone from plasma, extra-
hepatic metabolism of testosterone certainly occurs (31,32), but
the precise contribution by skin, muscle and other tissues to
testosterone clearance has not been precisely determined. It may
be, for instance, that reduced clearance in the female not only
reflects increased protein binding but also a lower extrahepatic
metabolism by such sensitive tissues as skin, muscle and sexual
structures. With plasma levels ranging between 3.5 and 10.5 ~g/l,
and clearance rates of between 900 and 1200 l/day, the rate of
entry of testosterone into blood may be calculated, assuming a

steady state, to be between 3 and 13 mg/day. This statement begs
two further questions: first, is the state steady, and, second,
is the rate of entry synonymous with secretion?
It seems likely that the state is not entirely steady. In the

past there has been some controversy as to whether the level of
this steroid changes within the day or between days (11,33-36).
The weight of evidence would now suggest that within the day there
is a variation, with higher morning and lower evening values; the
amplitude of these swings is probably small when compared, for
instance, to that of cortisol. A possible reason for past uncer-
tainty is that the methodological noise has tended to obscure bio-
logical signals. Several authors (37,38) have been unable to show
that a significant circadian rhythm exists for luteinizing hormone
(LH) or follicle stimulating hormone (FSH), although in a recent
report Saxena and Gandy (39) have shown a circadian rhythm of
large amplitude for LH and FSH to which is entrained a rhythm of
much smaller amplitude for testosterone. In limited longitudinal
studies on normal males, we have been unable to demonstrate sys-
tematic variations in the plasma concentration of testosterone
over a period of five weeks; in fact values were found to range
widely in an apparently random fashion (40).
The original studies of Tait and Horton (25,41), which have

now been clearly substantiated, have shown that in the adult male
more than 95% of testosterone in blood is secreted as such, the
remaining 5% being derived as a peripheral conversion product
mainly from androstenedione(27,42). This whole subject has been
extensively reviewed by Baird et ale (43). It is also clear that
the testosterone secreted originates from the testis and not from
the adrenal cortex. Thus, following adrenocortical suppression
or after adrenalectomy, or in patients with Addison's disease,
testosterone levels are not significantly different from other
normal males (40); further, the administration of ACTH to some
normal male patients is associated with a significant decrease
in plasma levels of testosterone (27). Why this should be so is
not entirely clear. Is it that ACTH exerts a direct effect on the
testis, or is it that increased blood levels of adrenal androgens,
androstenedione and DHEA, or adrenal estrogens act to inhibit the
secretion of tropic hormones?
Methods for the measurement of testosterone in plasma have now

made it possible to examine more formally those proposals previous-
ly shown by classic techniques in endocrinology that the testis
and the anterior lobe of the pituitary constitute an endocrine
axis, and that the secretion of testosterone is directly regulated
by the hypothalamus acting through the anterior pituitary. The
pituitary hormone responsible is "the interstitial cell stimulating
hormone (ICSH) for which there is putative evidence of a common
426 B. HUDSON ET Al.
identity with pituitary LH. A similar action is also shared by

human chorionic gonadotropin (HCG) which, when administered par-
enterally, is associated with prompt elevations in the level of
plasma testosterone in normal, hypophysectomized and prepubertal
males and as such provides a useful method for the evaluation of
Leydig cell function. Administration over a three or four day
period is almost certainly a more discriminatory procedure than
a single injection and the measurement of changes within a few
hours of this stimulus (44-46). Although there is good evidence
that FSH acts on the seminiferous epithelium, it seems unlikely
that this gonadotropin plays a role in the regulation of Leydig
cell function and the secretion of testosterone (47,48). It is
of interest, however, that Steinberger and Duckett (49) have de-
scribed depletion of pituitary FSH and a rise in plasma FSH fol-
lowing castration in the male rat, changes that can be reversed
by the administration of testosterone.
The neuro-endocrine mechanisms involved in the regulation of

testosterone secretion appear to be of the usual type; high levels
of plasma testosterone inhibit the secretion of LH (11,36), while
low levels are associated with an increased secretion of this
hormone (40). There is good evidence that the receptor sites of
this neuro-endocrine loop are hypothalamic, and are susceptible to
exogenous influences that may stimulate or inhibit the formation
or secretion of releasing factors. The anti-estrogen clomiphene,
when given in doses of 200 mg/day over a five-day period to nor-
mal subjects has been shown to cause rises in both plasma LH and
testosterone to between 80 and 160% of basal values, an action
that can be blocked by fairly substantial doses (80 mg/day) of the
synthetic androgen fluoxymesterone. Bardin and others (50) have
proposed this as a useful test of hypothalamic and pituitary func-
tion and have shown that these changes do not result from the di-
minished disposal of the steroid or from a direct effect of the
compound on testicular steroidogenesis. In addition to androgens,
estrogens (51,52) and progesterone analogues (53) cause a fall in
the levels of plasma LH and testosterone. This inhibition of LH
secretion, which has useful therapeutic implications, is achieved
only relatively slowly (as compared, for instance, with ACTH)
thus, following the administration of substantial inhibitory doses
of estrogens or androgens, LH levels fall over the course of three
or four days; likewise, plasma levels of testosterone also take
several days to reach a nadir and from our own studies the secre-
tion of neither hormone is completely suppressed even after weeks
of such treatment (51).
INFANCY, CHILDHOOD, ADOLESCENCE AND OLD AGE
At the extremes of life there are distinct differences from

the normal adult male pattern of testosterone secretion. In the
immediate neonatal period, plasma levels are elevated in both
male and female infants, but there is no significant sex differ-

ence (23,54). Both are significantly lower than the levels found
in maternal plasma, but testosterone binding affinity is low in
umbilical venous plasma and high in maternal plasma. From the
data of Grumbach and his colleagues (24), it is not readily pos-
sible to calculate the proportion bound to protein, but presum-
ably the level of free testosterone in fetal plasma is higher
than is reflected by the total level. The fact that there should
be no sex difference in plasma levels of testosterone at birth is
of interest, taking into consideration the intense stimulation of
the Leydig cells of the fetal testis that occurs during the second
trimester. In one series there was a significant difference be-
tween male and female neonates in the levels of androstenedione,
the values being higher in the former (54). If conventional male
sex hormones are responsible for influencing the development of
the male or female brain, then it can only be presumed that these
influences have operated at an earlier stage of fetal development.
Immediately after birth, plasma levels fall to values that are at,
or below, the normal female range, and remain at this level until
puberty. Frasier and Horton (55) have shown that during this pe-
riod plasma levels of androstenedione are of the same order as is
found in the adult male, that the androstenedione/testosterone
ratio is the order of two or thereabouts, and that about 40/
0
of testosterone is derived by conversion from androstenedione.
As puberty approaches in the male there is a gradual increase

in the level of testosterone in plasma, a rise which follows the
secretion of LH at that time (44,56,57). These measurable chem-
ical changes precede by many months the physical manifestations of
this metamorphosis, and can usefully be employed in the management
of those adolescents about whom there is some concern for their
sexual maturation. The finding of plasma levels of testosterone
above the prepubertal but below the adult male range usually
indicates that puberty is likely to proceed normally, and that
watchful expectancy is the appropriate medical management.
At the other extreme of life, there is no abrupt fall in the

level of plasma testosterone that might be taken to indicate the
existence of a male climacteric, although low levels are sporad-
ically encountered in septa-, octo- and nonogenarians. An inter-
esting study in this context is that of Kent and Acone (58) who
showed that the clearance of testosterone from blood fell progres-
sively with advanced years; mean plasma levels of testosterone
were also lower in older men, so that it would seem that Leydig
cell activity does become impaired with age, and thece is a fall
in the secretion of testosterone. Is it possible that this re-
duced clearance might also be taken to imply a lower extrahepatic
metabolism by which normally sensitive tissues are utilizing less
testosterone with advancing years?
ABNORMALITIES OF TESTOSTERONE SECRETION
In the male, an apparent clinical deficiency of androgen

secretion is more commonly encountered than androgen excess. Low
levels of plasma testosterone can result from disorders that pri-
marily and directly affect the testis and Leydig cell function,
or from conditions in which these tissues are secondarily affect-
ed. Sometimes this is a consequence of a more generalized endo-
crine disorder, but at others the primary disease is not identi-
fiably endocrine in origin.
Patients with Testicular Disorders
Following castration, the concentration of testosterone in

plasma falls to very low levels within two or three days, and
although testosterone can be measured in the plasma of these
patients, there is reason to believe that it is derived by con-
version from androstenedione. Patients with anorchia show sim-
ilar findings (59).
Low levels of plasma testosterone are observed in patients

with bilateral cryptorchidism, but these values are not usually
as low as those found in the castrate or anorchic male. Although
Leydig cell function in the cryptorchid male is normally reduced,
or apparently absent, significant changes in the level of plasma
testosterone can be observed following treatment with HCG, in-
dicating the presence of Leydig cells or their precursors that are
normally responsive to LH. The practical significance of this
observation is that it may be used to indicate that testicular
tissue is present outside the scrotum should surgical removal be
contemplated.
Numerous studies have been made on the levels of plasma tes-

tosterone in Klinefelter's syndrome, the most common cause of
male hypogonadism (36,45,46,60). These have shown that a fairly
wide range of values may exist in this disease, from those found
in the normal female to those observed in the adult male. At-
tempts by a number of workers to correlate testosterone levels
with the chromatin pattern have not yielded useful correlates;
that is, patients who were chromatin negative showed no consistent
differences from those who were chromatin positive; nor was there
any systematic correlation between plasma levels of testosterone
and the presence of mosaicism or with an XXXY karyotype. We have
attempted to correlate plasma levels with interstitial cell mor-
phology, but have not found this to be helpful. Paulsen et al.
have also pointed out the difficulty of correlating plasma levels
of testosterone with the degree of secondary sexual development
(46). It is of interest that in the few studies that have been
made, plasma androstenedione appears to be normal in this disorder.
A number of studies have demonstrated the damage inflicted on

the germinal epithelium by the presence of an additional X chro-
mosome; what has yet to be satisfactorily explained is the mor-
phology of the Leydig cells which frequently appear nodular,
clumped or hyperplastic, the high levels of plasma LH and the
lower than normal levels of plasma testosterone. One might well
speculate that the additional X chromosome is also injurious to
these cells, but that their biosynthetic capacity is not irrepar-
ably damaged. Is it that these cells need greater stimulation than
normal in order to secrete testosterone? This could explain the
high levels of LH and the hyperplastic appearance of these cells,
the hormonal output of which mayor may not be adequate, but still
does not account for those patients with normal levels of testos-
terone and high levels of LH. Whatever the explanation, it seems
clear that in these patients the biosynthetic capacities of the
Leydig cells are reduced; the administration of additional ex-
ogenous HCG over a three or four day period may sometimes produce
an increment in the level of plasma testosterone seen in the nor-
mal subject, but in some patients there is no such increment (46,
60).
Although not primarily a testicular disorder, it would seem

relevant to refer to a number of studies that have been undertaken
on males with an XYY chromosomal complement. Although these sub-
jects may show distinctive physical and psychological features,
abnormalities of testicular function do not appear to be present.
We have investigated several such subjects and have found plasma
levels of testosterone and LH to be normal (61,62). Other inves-
tigators have measured the rate of urinary excretion of testoster-
one glucuronide and have found it to be elevated (63,64). Method-
ological differences may explain these apparent discrepancies,
particularly as plasma levels were not measured in the latter two
studies.
Primary disorders of the testicle other than those that have

been mentioned do not commonly show abnormalities of testosterone
secretion. We have now investigated in excess of 40 patients with
infertility associated with oligo- or azoa;permia without apparent
clinical abnormalities of androgen secretion. These patients have
been shown to have germinal cell aplasia (the Sertoli cell only
syndrome), germinal cell arrest, and other disorders of spermato-
genesis, the nature of which was often quite obscure. Commonly,
levels of plasma testosterone in these patients are within the
normal range and respond appropriately to normal stimuli. In
about25% of these patients, however, plasma levels of testosterone
are lower than normal (less than 300 ng/100 ml) and in this group
the common morphological change is that of germinal cell aplasia.
Such patients have high levels of FSH but normal LH. (65).
The testes may fail to mature because of primary gonadotropic

failure or following removal of the gonadotropic stimulus before

puberty; regression may also occur following removal of this
stimulus after puberty. In such patients plasma levels of tes-
tosterone are usually lower than normal, the ultimate level de-
pending upon the extent and duration of the failure of gonadotro-
pic secretion. In patients with the rather unusual disorder of
primary hypogonadotropic hypogonadism, plasma levels are in the
female range and fail to respond to clomiphene; patients with
hypopituitarism likewise show levels which vary from very low
normal to normal, levels that generally correlate with the out-
put of LR which is not invariably lost in all patients.
Disorders not Primarily Affecting the Testis.
Plasma levels of testosterone have been found to be abnormal

in a variety of disorders that are not normally regarded as being
primarily testicular in origin, although the testis may be in-
volved incidentally in the disorder. Perhaps the best recognized
of these is liver cirrhosis in which the triad of spiderangiomata,
testicular atrophy and gynecomastia is well recognized, particu-
larly when the disorder results from the ravages of alcohol. It
has long been speculated that an excess of circulating estrogen
is present and that plasma androgens may be normal or reduced in
amount. Shaver et ale (66) have reported studies which suggested
that blood levels of estrogen were elevated and Zumoff et ale (67)
have shown that the metabolism of estradiol was abnormal in cir-
rhosis. Only recently have these suggestions been confirmed by
the direct measurements made by Korenman et ale (68) who found
that levels of plasma estradiol in men with cirrhosis were almost
twice those of normal men. It is relevant that Tarvenetti et al.(69)
have shown that the plasma binding of this steroid was increased
in male and female patients with cirrhosis. Plasma levels of
testosterone have been found to be low in patients with alcoholic
cirrhosis (59). It has also been observed that levels of gonad-
otropins are lower than normal (71) and that the response to ReG
appears to be blunted in this condition. These observations sug-
gest that in cirrhosis the elevation of circulating estrogen le-
vels inhibits testosterone secretion either directly or by reduc-
ing gonadotropin secretion, and it is the combination of high cir-
culating estrogens and low testosterone which combine to produce
the classical endocrine abnormalities of this disorder. The path-
ogenesis of this condition has yet to be proved formally.
Diminished sexual potency, features of hypogonadism and tes-

ticular atrophy are well recognized features of hemochromatosis.
Stocks and Martin (72) have shown that in this disorder plasma
levels of testosterone were below normal in about 50% of the sub-
jects they studied; these values correlated well with levels of
plasma LR which were also at or below the lower limit of normal,
and with the appearance of the test.is or biopsy. Tubular atrophy,
reduction in the Leydig cell population and a virtual absence of

stainable iron in the testis, were other important features. Col-
lectively, these findings indicate that the lesion in this dis-
order is primarily a failure of pituitary function, a proposition
that is supported by other evidence: low plasma thyroxine levels
and diminished cortisol and growth hormone responses to hypogly-
cemic stress. This is thus in contrast to the situation that pre-
vails in liver cirrhosis. Likewise, we have investigated three
patients with transfusion hemisiderosis due to iron storage.
These were adolescents with severe hypoplastic anemia who had re-
ceived repeated blood transfusions over many years. All patients
showed hypogonadism and testicular atrophy and were found to have
abnormally low levels of plasma testosterone, but following the
administration of RCG there was an increase in these levels. At
necropsy in one patient, extensive deposition of iron was observed
in the sinusoids of the anterior pituitary with minimal iron de-
position in the testis.
The testicular lesions of leprosy have been known for some

years (73), and have been likened to those observed in Klinefel-
ter's syndrome with tubular hyalinization and clumping of Leydig
cells. In a series of 10 patients with this disorder, some with
gynecomastia, we have found high levels of urinary gonadotropins
and low levels of testosterone in plasma, these changes in general
correlating with the severity of hypogonadism and gynecomastia(74).
At present, the reason for the predilection for tubular involve-
ment in leprosy is obscure, but the hormonal changes would appear
to be a direct consequence of testicular involvement in this dis-
ease.
ELEVATED PLASMA LEVELS OF TESTOSTERONE
Levels of plasma testosterone that are higher than normal are

less frequently encountered. Undoubtedly the most dramatic ex-
amples are those unusual patients with sexual precocity, which
may result from the inappropriate secretion of LR; idiopathic
sexual precocity, from adrenal hyperplasia or tumor, or much less
commonly, from interstitial cell tumors of the testis. In patients
with inappropriate LR secretion, plasma levels of testosterone are
usually elevated but these may vary from below the normal adult
male range to values that are within this range. Administered
RCG produces a further rise in these levels, while the administra-
tion of estrogens or medroxyprogesterone acetate (Provera) will
lower the values (36,53). It is of interest that Provera has been
found to be associated with a decrease in testosterone binding,
and it is possible that while this drug will induce a lowering of
total testosterone content in plasma (bound and unbound), the
value of the unbound and active moiety may increase, or at least
may remain unchanged, and thus may prevent the complete therapeu-
tic efficacy of this compound.
An increased concentration of testosterone in plasma is also

found in patients with both salt-losing and non salt-losing forms
of congenital adrenal hyperplasia (CAR). Migeon (75) has shown
that in all these patients there is a striking increase of the
normal testosterone/androstenedione ratio with androstenedione
levels of between 0.5 and 5 ~g/100 mI. Further studies have shown
that the increased concentration of testosterone in patients with
CAR not only reflects an increased secretion of testosterone, but
also the conversion to testosterone of androstenedione which may
contribute up to 60% of the total blood production of testosterone.
The abnormal secretion of androstenedione and the production of
testosterone can be suppressed by the administration of cortico-
steroids.
Some solution to the apparent paradox of a normal feminine

appearance and normal adult male levels of plasma testosterone in
patients with the syndrome of testicular feminization has recently
been suggested (76,78). It has been shown that skin taken from
patients with this disorder did not convert testosterone to di-
hydrotestosterone as did normal controls (32,79). These observa-
tions provide additional evidence that at a tissue level the active
form of testosterone is dihydrotestosterone, a proposition that
has been strongly suggested by the work of Wilson et al. (80).
Levels of plasma testosterone have now been measured in a va-

riety of clinical and experimental conditions, many but not all of
which have been summarized in this paper. What has emerged from
these fairly intensive studies over the past four to five years is
that no level or group of levels may be entirely meaningful unless
a proper account can be taken of the form in which the steroid is
presented to the tissues; that is, unless there is some know-
ledge of the degree to which the steroid is bound. In this con-
text it is of interest that plasma levels of, testosterone have
been found to be high in patients with untreated thyrotoxicosis
(81,82). It has been shown further that the metabolic clearance
of testosterone in these patients is reduced, an observation that
has been related to an increased testosterone binding affinity of
the plasma in these patients. The importance of steroid binding
is further emphasized by the studies of Southren et al. (83) who
showed that in hirsute female patients with the syndrome of poly-
cystic ovaries and elevated levels of plasma testosterone, meta-
bolic clearance was increased and binding reduced, so that by
extrapolating in vitro studies as much as a six-fold rise in the
unbound levels of testosterone might be predicted. In the re-
verse situation in the hypogonadal male, it is possible that sim-
ple measurement of the level of testosterone in blood may not yield
a full measure of the androgenic stimulus unless some account is
taken of the level of free or unbound steroid which, of course,
may be significantly reduced if plasma levels of estrogen (known
to increase binding) are also increased. From these studies it
would seem important to point out that while measurements of plas-

ma testosterone have been of material value in understanding some
of the problems of the normal and disordered physiology of andro-
gen secretion, future endeavors must be directed to determining
the levels at which testosterone is present in the free or unbound
form in the circulation and of the tissue or plasma levels of con-
version products such as dihydrotestosterone.
ACKNOWLEDGEMENTS
This work has been supported by grants from the Australian

National Health and Medical Research Council, and from the Anti-
Cancer Council of Victoria.
REFERENCES
1. Murphy, B.E.P., Canad. J. Biochem., 46: 299, 1968.

2. Kato, T. and Horton, R., Steroids, ll: 631, 1968.
3. Mayes, D. and Nugent, C.A., J. Clin. Endocr., 28: 1169, 1968.
4. Hallberg, M.C., Zorn, E.M. and Wieland, R.G.St;roids, 21:
241, 1968. -
5. Rosenfield, R.L., Eberlein, W.R. and Bongiovanni, A.M.,
J. Clin. Endocr.,~: 854, 1969.
6. Niswender, G.D. and Midgely, A.R., Proc. 51st Meeting Endocrine
Society, 1969,(Abstract No. 22).
7. Finkelstein, M., Forchielli, E. and Dorfman, R.I.,
J. Clin. Endocr., 11: 98, 1961.
8. Hudson, B.,. Coghlan, J.P., Dulmanis, A., Wintour, M. and Ekkel,
I., Aust. J. Exp. Bioi. Med. Sci., &:.235, 1963.
9. Riondel, A., Tait, J.F., Gut, M., Tait, S.A.S., Joachim, E.and
Little, B., J. Clin. Endocr.,~: 620, 1963.
10. Burger, H.G., Kent, J.R. and Kellie, A.E.,
J. Clin. Endocr.,~: 432, 1964.
11. Kirschner, M.A., Lipsett, M.B. and Collins, D.R.,
J. Clin. Invest., ~: 657, 1965.
12. Rivarola, M.A. and Migeon, C.J., Steroids, 7: 103, 1966.
13. Gandy, H.M. and Peterson, R.E., J. Clin. Endocr., ~: 949, 1968.
14. Brownie, A.C., Van der Molen, H.J., Nishizawa, E.E. and
Eik-Nes, K.B., J. Clin. Endocr.,~: 432, 1964.
15. Slaunwhite, W.R. and Sandberg, A.A., J. Clin. Invest., 38:
384, 1959.
16. Chen, D.S., Mills, I.H. and Bartter, F.C., J. Endocr., 23:
129, 1961.
17. Mercier, C. and Beaulieu, E.E., Ann. Endocr. (Paris), 29
(Suppl): 159, 1968.
18. Pearlman, W.H. and Crepy, 0., J. Bioi. Chern., 242: 182, 1967.
19. Rosner, W., Christy, N.P. and Kelly, W.G., Biochemistry (Wash),
8: 3100, 1969.
20. Vermeulen, A. and Verdonck, L., Steroids, 11: 609, 1968.
21. Kato, T. and Horton, R., J. Clin. Endocr., 28: 1160, 1968.
22. Barlow, J.J., MacLaren, J.A. and Pothier, L.,

J. Clin. Endocr., ~: 767, 1969.
23. Rivarola, M.A., Forrest, M.G. and Migeon, C.J.,
J. Clin. Endocr., 28: 34, 1968.
24. August, G.P., Tkac~k, M. and Grumbach, M.M.,
J. Clin. Endocr., 29: 891, 1969.
25. Horton, R. and Tait; J.F., J. Clin. Invest., 45: 301, 1966.
26. Hudson, B., Dulmanis, A., Coghlan, J.P. and Wi~our, M.,
Aust. Ann. Med., 15: 236, 1966.
27. Rivarola, M.A., S;;z, J.M., Meyer, W.J., Jenkins, M.E. and
Migeon, C.J., J. Clin. Endocr., 26: 1208, 1966.
28. Bardin, C.W. and Lipsett, M.B., J: Clin. Invest., 46:891,
1967.
29. Southren, A.L., Gordon, G.G., Tochimoto, S., Pinzon, G.,
Lane, D.R. and Stypuckowski, W., J. Clin. Endocr., 27: 686,
1967. -
30. Bird, C.E., Green, R.N. and Clark, A.F.,
J. Clin. Endocr., 29: 123, 1969.
31. Gomez, E.C. and Hsi;, S.L., Biochemistry (Wash), 7: 24, 1968.
32. Wilson, J.D. and Walker, J.D., J. Clin. Invest., ~: 371,
1969.
33. Dray, F., Reinberg, A. and Sebaun, J.C.R.,
C.R. Seance Acad. Sci., 261: 573, 1965.
34. Southren, A.L., Gordon, G:G. and Tochimoto, 5.,
J. Clin. Endocr., 25: 1441, 1965.
35. Resko, J.A. and Eik:Nes, K.B., J. Clin. Endocr., 26: 573,
1966.
36. Hudson, B., Coghlan, J.P. and Dulmanis, A., Ciba.Found.
Colloq. Endocr., Endocrinology of the Testis, 16: 140, 1967.
37. Faiman, C. and Ryan, R.J., Nature, (London) 215: 85, 1967.
38. Peterson, N.T., Midgley, A.R. and Jaffe, R.~
J. Clin. Endocr., 28: 1473, 1968.
39. Saxena, B.B. and G~dy, H.M., Symposium on in vitro Procedures
with Radioisotopes in Clinical Medicine and Research.
Vienna 1969.
40. Burger, H.G., Brown, J.B., Catt, K.J., Hudson, B. and
Stockigt, J.S., In Protein and Polypeptide Hormones,
ed. Margoulies, M:, Excerpta Medica Foundation International
Congress Series, Amsterdam, 161: 412, 1968.
41. Tait, J.F. and Horton, R., St;roids, 4: 365, 1964.
42. Korenman, S.G. and Lipsett, M.B. J. Clin. Invest., 43: 2125,
1964.
43. Baird, D.T., Horton, R., Longcope, C.T., Tait, J.F.,
Perspect. Biol. Med., 11: 384, 1968.
44. Hudson, B., Coghlan, J:P., Dulmanis, A. and Wintour, M.
Excerpta Medica Foundation International Congress Series,
Amsterdam., 83: 1127, 1967.
45. Lipsett, M.B:; Davis, T.E., Wilson, H. and Canfield, C.J.,
J. Clin. Endocr',lli 1027, 1965.
46. Paulsen, C.A., Gordon, D.L., Carpenter, R.W., Gandy, H.M. and
Drucker, W.D., Recent Progr. Hormone Res., 24: 321, 1968.
48. de Kretser, D.M., Taft, H.P., Brown, J.B., Evans, J.H., and
Hudson, B., J. Endocr., 40: 197, 1968.
49. Steinberger, E. and Duckett, G., Acta. Endocr., (Kobenhavn)
57: 289, 1968.
50. Bardin, C.W., Ross, G.T. and Lipsett, M.B.,
J. Clin. Endocr., 27: 155, 1967.
51. Alder, A.B., Burger;-H.G., Davis, J., Dulmanis, A., Hudson, B.
Sarfaty, G.A. and Straffon, W., Brit. Med. J., 1: 28, 1968.
52. Young, H.H. and Kent, J.R., J. Urol., 99: 788: 1968.
53. Rivarola, M.A., Camacho, A.M. and Migeo~ C.J.
J. Clin. Endocr., 28: 679, 1968.
54. Mizuno, M., Lobotsky, J., Lloyd, C.W., Kobayashi, T.
and Murasawa, Y., J. Clin. Endocr.,~: 1133, 1968.
55. Frasier, S.O. and Horton, R., Steroids,~: 777, 1966.
56. Boon, D.A., Keenan, R.E., Slaunwhite, W.R. and Aceto, T.
Excerpta Medica Foundation International Congress Series,
Amsterdam, 157: 52, 1968 (Abstract).
57. Johanson, A:J7, Guda, H., Light, C., Migeon, C.J. and
Blizzard, R.J., J. Pediat. 74: 416, 1969.
58. Kent, J.R. and Acone, A.B., irl Androgens in Normal and Path-
ological Conditions, eds.Ver;;ulen, A. Exley, D., Excerpta
Medica Foundation International Congress Series, Amsterdam,
.!.Q!: 31, 1966.
59. Coppage, W.S. and Cooner, A.E., New Eng. J. Med. 11l: 902,
1965.
60. Lipsett, M.B., Wilson, H., Kirschner, M.A., Korenman, S.G.,
Fishman, L.M., Sarfaty, G.A. and Bardin, C.W., Recent Progr.
Hormone Res., 11: 245, 1966.
61. Wiener, S., Sutherland, G., Bartholomew, A.A. and Hudson, B.,
Lancet, 1: 150, 1968.
62. Hudson, B., Burger, H.G., Wiener, S., Sutherland, G. and
Bartholomew, A.A., Lancet, 2: 699, 1969.
63. Ismail, A.A., Harkness, R.A:, Kirkham, K.E., Loraine, J.A.,
Whatmore, P.B. and Brittain, R.P., Lancet,!: 220, 1968.
64. Rudd, B.T., Galal, a.M. and Casey, M.D., J. Med. Genet.,
5: 286, 1968.
65. de Kretser, D. M., Studies on the Structure and Function of
the Human Testis. M.D. Thesis, Monash University, 1968.
66. Shaver, J.C., Roginsky, M.S. and Christy, M.P.,Lancet,
1,: 335, 1963.
67. Zumoff, B.~ Fishman, J., Gallagher, T.F. and Hellman, L.,
J. Clin. Invest., 47: 20, 1968.
68. Korenman, S.G., Per~n, L.E. and McCallum, T., Proc. 61st
Meeting American Society Clin. Invest., 1969, (Abstract No.144).
69. Tarvenetti, R.R., Rosenbaum, W., Kelly, W.G., Christy, N.P.
and Roginski, M.S., J. Clin. Endocr., 1L: 920, 1967.
70. Taft, H.P., Seah, C.S., Burger, H.G. and Hudson, B.,
Unpublished Observations.
71. Rosselin, G. and Dolais, J,., Press Med., 75: 2027, 1967.
72. Stocks, A.E. and Martin, F.I.R., Amer. J.~ed. Sci., 45:
839, 1968. --
73. Grastbald, H. and Swan, L.L., J.A.M.A., 149: 1287, 1952.
74. Martin, F.I.R., Maddocks, I., Brown, J.B:-;nd Hudson, B.,
Lancet,~: 1320, 1968.
75. Migeon, C.J., Ann. Intern. Med. ,68: 1329, 1968.
76. Southren, A.L., Tochimoto, S., Carmody, N.C. and Isurugi, K.,
J. Clin. Endocr., 25: 1441, 1965.
77. Deshpande, N., Wang; D.Y., Bulbrook, R.D. and McMillan, M.,
Steroids, 4: 437, 1965.
78. French, F.S., Van Wyk, J.J., Baggett, B., Easterling, W.E.,
Talbert, L.M., Johnston~ F.R., Forchielli, E. and ~y, A.C.,
J. Clin. Endocr., 26: 493, 1966.
79. Northcutt, R.C., r;iand, D.F. and Liddle, G.W.,
J. Clin. Endocr., 29: 422, 1969.
80. Wilson, D.J., Brucb;vsky, N. and Chatfield, J.H., Excerpta
Medica Foundation International Congress Series, Amsterdam,
184: 17, 1968.
81. Dreyfus, G., Sebaoun, J., Delzant, G'., Dray, F., Ledru, M.J.
and Crepy, 0., Excerpta Medica Foundation International
Congress Series, Amsterdam, 122: 55, 1968.
82. Gordon, G.G., Southren, A.L., Tochimoto, S., Rand, J.J. and
Olivo, J., J. Clin. Endocr., 29: 164, 1969.
83. Southren, A.L., Gordon, G.G., fuchimoto, S., Olivo, J.,
Sherman, D.H. and Pinzon, G., J. Clin. Endocr., 29: 1356,
1969. --
DISCUSSION
PAULSEN: Dr. Hudson, is the basis for the discrepancy between plasma
testosterone and serum HLH levels in patients with Klinefelter's syn-
drome known? Do you have any data which indicate that there may be
a difference in testosterone binding in the plasma of these patients
as opposed to the normal male?
HUDSON: We do not, but in a paper by Vermeulen and Verdonk in the

Journal of Clinical Endocrinology (November, 1969) they report that
binding in hypogonadal males was increased. I am not certain that
patients with Klinefelter's syndrome were included among these.
SAVARD: Dr. Hudson, in your values for testosterone secretion in the

prepubertal and the neonatal children, did you make concomitant mea-
surements of androstenedione? In other words, is testicular function
in the youngsters reflected in a mo~e abundant production of andro-
stenedione? Your data showed at least in the prepubertal children
that a substantial amount of testosterone was derived from andro-
stenedione. The testis itself in its prepubertal state may conceiv-
ably be elaborating androstenedione as its major product. I think

that this has certainly been observed in animals by Lindner.
HUDSON: I do not have that data from my own laboratory, but in a

paper from Lloyd's group in which levels of androstenedione in cord
blood were estimated, they showed high levels for this steroid in
males.
DICZFALUSY: Dr. Hudson, would you consider that in cord blood you
will find some testosterone because the placenta secretes andro-
stenedione and this is converted by the fetal liver to testosterone?
HUDSON: Yes, I would consider this.
LIPSETT: In the prepubertal child Fraser and Johnson showed that

testosterone levels are higher in boys than in girls although andro-
stenedione levels were the same. We have recently performed studies
in the rat. At 20 days of age when there is absolutely no evidence
of increase in secondary sexual characteristics in prostate or sem-
inal vesicle weight, there are only few Leydig cells, but one can
clearly demonstrate testosterone secretion, although peripheral
testosterone levels still remain unmeasurable. Only subsequently
does Leydig cell activity increase and peripheral values of testos-
terone begin to rise.
BERGADA: We have treated prepuberal boys with HCG. Three groups

of patients were treated. We gave 800 IU of HeG daily in one group,
2,000 and 4,000 in the others, and the mean response in three groups
was approximately the same with a large range. It seems that there
is no dose response relationship with an HCG stimulation test given
for five days.
HUDSON: I am very interested in those data, because there has been

some discussion about what is the appropriate dose of HCG. We have
used 1500 IU of ReG daily for four days. It would seem that this
is an adequate dose as a test of testicular function, and this is
the reason we have used it.
E. STEINBERGER: I would like to make a comment on Dr. Hudson's pa-

per. He stressed the "methodological" noise interference in detec-
tion of fine changes in testosterone production. I would just like
to add one more noise, namely the "patient noise". Doctors Persky
and Smith of our group obtained data on a group of college students,
aged 17 to 22 years, who were given a series of psychologic tests
to determine the levels of aggression and hostility. The psychologi-
cal test scores were plotted against testosterone production rates.
An excellent positive correlation between these two parameters was
observed (R= + 0.66; p <001, for 17 dt). The predicted testoster-
one production rates of these patients based on their psychologic
scores again showed an excellent correlation. These results suggest
an additional parameter related to testosterone production rate.
RELATION OF IN VITRO METABOLISM OF STEROIDS IN HUMAN TESTICULAR
TISSUE TO HISTOLOGIC AND CLINICAL FINDINGS
E. Steinberger, M. Ficher and K.D. Smith
Division of Endocrinology and Reproduction, Research

Laboratories, Albert Einstein Medical Center,
Philadelphia, Pa. 19141
INTRODUCTION
The first definitive evidence for androgen secretion by the

human testis was provided by experiments demonstrating the formation
of radioactive testosterone and androstenedione in spermatic venous
effluent following perfusion of the testis with acetate 14C (1).
Lucas et al. (2) isolated testosterone from spermatic vein blood,
and subsequently a number of investigators demonstrated conversion
of suitable steroidal precursors to androgens by human testicular
tissue in vitro (3-7). Deficiency of the side-chain cleaving enzyme
in testicular tissue of an older man was shown by Axelrod (5). Car-
stensen (3) demonstrated in vitro an increased conversion of 17a-hy-
droxypregnenolone to testosterone in the presence of ICSH. He sug-
gested that the side-chain cleaving enzyme may be activated by go-
nadotropins. A suppression of the activity of the side-chain cleav-
ing enzyme and the 17U-hydroxylase, but no change, or possibly an
increase, in the activity of 17 ~-hydroxysteroid dehydrogenase was
demonstrated in the microsomal fraction of testicular tissue from a
patient with prostatic carcinoma treated with estrogens (8). If the
assumption is made that estrogen suppressed patients' endogenous
gonadotropins, the decrease in 17-20 lyase may be interpreted as
confirmation of Carstensen's findings.
On the other hand, Schoen (7), also utilizing testicular tis-

sue from prostatic cancer patients treated with estrogens, demon-
strated suppression of testosterone formation from androstenedione
and concluded that gonadotropins may be essential for the activity
of 17 ~-hydroxysteroid dehydrogenase. Similarly, Slaunwhite et al.
(4), utilizing testicular tissue from patients with prostatic car-
cinoma, showed that long-term estrogen therapy prevented the reduc-
439
440 E. STEINBERGER, M. FICHER, AND K. D. SMITH
tion of androstenedione to testosterone, but did not affect the con-

version of 17a-hydroxyprogesterone to androstenedione.
The above briefly summarized studies clearly indicate that hu-

man testicular tissue can incorporate simple precursors like acetate
into the steroid molecule and convert suitable steroidal precursors
to androgens. Also, these studies have elucidated, at least in part,
the biogenetic pathways leading to formation of androgens in human
testicular tissue. However, information concerned with the hormonal
control of this biosynthetic process is still controversial and
fragmentary. Furthermore, most studies reported to date were per-
formed on testicular tissue from elderly patients with carcinoma of
the prostate. Consequently, almost no information is available on
the steroid biosynthetic activity of testes from patients of repro-
ductive age, with various spontaneously occurring disorders of tes-
ticular function. Since the only practical way to obtain testicu-
lar tissue in such patients is by means of surgical biopsy, tech-
niques for study of steroid biosynthesis in small amounts of testi-
cular tissue are required.
Danezis (9) and Schoen (7) have shown that it is possible to

demonstrate in vitro conversion of steroidal precursors to androgens
in as little as 20 mg. of testicular tissue. In the study reported
here, we re-evaluated the feasibility of utilizing small amounts of
human testicular tissue and made an attempt to correlate the pattern
of progesterone metabolism in testicular tissue, obtained either by
orchiectomy or surgical biopsy, with the histologic appearance of
the testes, the clinical picture of the patient, and in some instan-
ces, with in vivo biochemical parameters of androgen secretion.
A. Source of testicular tissue. The age and diagnosis for each

patient are summarized in Tables 1 and 2.
B. Hormonal studies: Urinary levels of FSH were determined by the
Steelman-Pohley technique (10), and LH by the rat ovaria~ ascorbic
acid depletion technique (11). Plasma testosterone (P.T.) levels
were determined by the technique of Horton et al. (12), and the
testosterone production rates (T.P.R.) by the technique of Rivarola
et al. (13).
C. Chromosomal studies: A micro-technique employing whole blood
was utilized (14).
D. Steroid biosynthetic studies: Minced teRticular tissue was in-
cubated with 3H-progesterone (20 x 10 6 dpm,3m~ mole, dissolved in
1.5 ml of incubation medium) for 3 hr. in a Dubnoff shaker at 37 0 C
in an atmosphere of 95% air and 5% C02. The incubation was termin-
atqd by the addition of 10 ml ethyl acetate. At this point, appro-
priate 14C labelled tracers (40,000 dpm of each) were also added to
the incubate in order to correct for losses in subsequent manipula-
tions. The material was then frozen until further processing. Com-
Table 1. Testicular Tissue Obtained at Orchiectomy
Case Age Diagnosis
A 71 years Carcinoma of the prostate

B 65 years Carcinoma of the prostate
C 79 years Carcinoma of the prostate
D 22 years Testicular feminization
Table 2. Testicular Tissue Obtained at Biopsy
Case Age Diagnosis
E 24 years Klinefe 1 ter' s syndrome

F 26 years Carcinoma of contralateral testicle
G 36 years Azoospermia and impotence
H 32 years Azoospermia and impotence
I 18 years Oligospermia and gynecomastia
J 35 years Oligospermia and varicocele
K 29 years Oligospermia and liver disease
L 26 years Oligospermia
M 25 years Oligospermia
position of incubation medium (for total volume of 1.5 ml);

1 ml Hank's balanced salt solution
125 ~l 1.3% NaHC03
175 ~l NADP,Na solution (3.257 f.JlIlole/ml)
175 \-Ll G-6-P,Na2 solution (36.17 \-Lmole/ml)
25 ~l ethanol
The extraction procedure and the techniques of separation and iden-
tification of the steroid metabolites have been described elsewhere
(15). The final identification and quantitation of some of the ste-
roids was accomplished by the technique of derivative tormation and
isotopic dilution and crystallization to constant 3H/l C ratio or
constant specific activity.
RESULTS
1) Metabolism of Progesterone by Various Amounts of Testicular

Tissue
Tissue from Case A was utilized in amounts varying from 10 to

100 mg. per incubation flask, providing for approximate ratios of
substrate to tissue from 1:15,000 to 1:100,000. The results ob-

tained after first chromatography in a Bush A system are summarized
in Table 3. No significant difference was observed at tissue ra-
tios varying from 1:30,000 to 1:100,000; a diminution in capacity
to metabolize progesterone was observed at the ratio of 1:15,000.
About 50% of the precursor was converted to metabolites with

the chromatographic mobility of testosterone and 17 a hydroxyproges-
terone. This material was further worked up and the percent con-
version to testosterone and 17U-hydroxyprogesterone determined (Ta-
ble 3). No significant differences in conversion of progesterone
to testosterone were observed at ratios 1:30,000 to 1:100,000.
Small amounts of progesterone were converted to material with

the chromatographic mobility of 20U-dihydroprogesterone and andro-
stenedione, and about 20/ to highly "polar" metaboli tes. The iden-
0
tity of 20U-dihydroprogesterone and androstenedione was confirmed

by repeated chromatography in several systems and isotopic dilution
and crystallization to constant specific activity.
2) Progesterone Metabolism in Testicular Tissue of Patients with

Carcinoma of the Prostate
In all cases, progesterone was actively metabolized. However,

differences in the amounts of progesterone converted to specific
metabolites were observed (Table 4). Utilization of 14C labelled
tracers and crystallization to constant 3H/14C ratios permitted
relatively precise determination of the percent conversions of pro-
gesterone to testosterone and 17U-hydroxyprogesterone. The tissue
from Case A was very actively converting progesterone to 17U-hydrox-
yprogesterone (50% was converted), but only 1.0 percent of proges-
terone was converted to testosterone. Low level metabolism of pro-
gesterone was observed in tissue from Case B, and fairly active
metabolism was observed in tissue from Case C. In all cases, in
addition to 17a-hydroxyprogesterone and testosterone, 200-dihydro-
progesterone and androstenedione were formed. These metabolites
were identified by isotopic dilution and crystallization to con-
stant specific activity. The "polar" materials in tissues from
Cases A and C were rechromatographed, utilizing thin layer tech~
nique. Material with the chromatographic behavior of 17 ~-estradi-
01 was detected. Its identity was confirmed by isotopic dilution
and crystallization to constant specific activity (Table 5). Mi-
croscopic examination of the biopsies revealed spermatogenic acti-
vity, including mature spermatozoa in all specimens (Fig.l). The
Leydig cells in the biopsies varied markedly, and did not correlate
with the biochemical data. Tissue with the greatest capacity for
conversion ot progesterone to 17U-hydroxyprogesterone and testos-
terone (Case A) showed only sparse but well formed Leydig cells
(Fig. la), while tissue showing the least conversion (Case B) re-
vealed the largest masses of Leydig cells (Fig. lb).
Table 3. Metabolism of Progesterone by Various Amounts of Testicular

Tissue*
METABOLITES SUBSTRATE TO TISSUE RATIOS

1:15,000 1:30,000 1:50,000 1:100,000
p. pc p_ _ pc p pc
"Polar Compounds 13 27 15 17
0.7 0.9 1.0
Testosterone 20 39 52 53
17a-hydroxyprogesterone 38.0 50.0 50.0
20a-dihydroprogesterone 6 5 9 6
Androstenedione 1 1 2 2
Progesterone (unconverted
substrate) 60 28 21 21
p - percent of radioactivity recovered in respective zones after t:r.e

fi=st chromatography in Bush A system
pc -percent conversion of progesterone to testosterone and 17a-hy-
droxyprogesterone
* - tissue from Case A (see Table 1) was used
Table 4. Conversion of Progesterone by Testicular Tissue From Pa-

tients with Carcinoma of the Prostate
METABOLITES SOURCE OF TISSUE
A B C
p pc p pc p pc
"Polar" compounds 17 13 23
1.0 0.3 2.0
Testosterone 53 8 29
17a-hydroxyprogesterone 50.0 8.0 26.0
20a-dihydroprogesterone 6 13 10
Androstenedione 2 3 2
substrate) 21 37 27
p - percent of radioactivity recovered in respective zones after the

first chromatography in Bush A system
pc - percent conversion of progesterone to testosterone and
17a-hydroxyprogesterone
Table 5. 17 !3 -Estradiol Crystallization to Constant 3H/14C Ratio

Case A Case C Case E
Crystallization #1 xls a 0.86b 0. 86 b 1.04b
Crystallization #2 xis 0. 82b 1.09b
Crystallization #3 xis 1.06 0.9 7b
MEAN + S.E. 0.90+0.03 0.84+0.02 1.03+0.035
a - crystals
b - 3H/14C ratio used for calculation of the mean + S.D.
3) Metabolism of Progesterone in Testicular Tissue Obtained by

Means of Surgical Biopsy
Unfortunately, we do not have data on progesterone metabolism

in testicular tissue from perfectly normal individuals. Thus, no
absolute comparisons can be made. However, the information obtained,
when related to the data on plasma testosterone levels and testos-
terone production rates of the same patients and compared with the
P.T. levels and T.P.R. of normal individuals, provides interesting
information (Table 6)
a) Azoospermia. Two patients (G and H) with spontaneously

occurring a~oospermia and a gradual onset of impotence of three to
four years' duration, and one patient (E) with azoospermia associ-
ated with Klinefelter's syndrome were studied. The laboratory data
are summarized in Table ~ Unfortunately, no P.T. levels or T.P.R.
are available for the pa~ient with Klinefelter's syndrome. Both
patients with impotence (G and H) showed markedly depressed P.T.
levels, while no urinary gonadotropins were detected in Case H.
Histologic study of the testis showed almost complete degeneration
of seminiferous tubules in Case G, with masses of Leydig cells form-
ing "pseudoadenomas" in some areas (Fig. ld). Case H showed imma-
ture seminiferous tubules lined with "supporting cells" and no evi-
dence of Leydig cells (Fig. le). Testes from the patient with
Klinefelter's syndrome (Case E) revealed occasional seminiferous
tubules containing germinal epithelium cells and large masses of
Leydig cells (Fig. if).
Results of steroid studies are summarized in Table 8., In in-

cubates of tissue from Case G, most of the progesterone (71%) re-
mained unmetabolized, in spite of the fact that the tissue contained
masses of Leydig cells and the levels of gonadotropins were elevated.
This low steroid metabolic activity of the tissue was also reflected
by the extremely low P.T. levels and T.P.R. Similarly, low levels
of steroid metabolism were observed in testicular incubates from
Case H, also reflected by low P.T. levels and T.P.R., but associated
with atrophy of Leydig cells. In incubates of testicular tissue
Table 6. -Plasma Testosteoone Levels and Production Rates, Urinary

LH and FSH Levels in a Group of Normal Males 18 to 26
Years of Age with Sperm Counts Over 80 Million
P.T. T.P.R. LH* FSH*
II\J.g/l00 ml mg/24 hrs. I.U./24 hrs. I.U./24 hrs.
MEAN 685 10.1 7.4 1.8

S.D. 195 3.6 4.8 1.4
N 18 18 25 24
*LH and FSH expressed in terms of 2nd I.R.P.
Table 7. Summary of Laboratory Data for 3 Patients with Azoospermia

P.T. T.P.R. LH FSH
Case mb!g/100 ml mg/dax I.U./24 hrs. I.U/24 hrs. Karyotype
G 303 1.82 52.3 14.9 46/XY
H 196 1.52 *not detect. >~not detect. 46/XY
E 1.2 12.5 47/XXY
*Tested in 24 hr. aliquots of urine

Gonadotropins expressed in terms of 2nd I.R.P.
Table 8. Metabolism of Progesterone by Biopsies from Patients with

Azoospermia (G and H), Klinefelt'er's Syndrome (E) and
Testicular Feminization (D)
Case G Case H Case E Case D

METABOLITES P pc P pc P pc ~
"Polar" compounds 0 15 29 40
Testosterone 0.7 0.7 9.0 12

18 18 43 27
l1tt-hydroxyprogesterone 8.3 1.2 30.0 22
20a-dihydroprogesterone 5 4 3 3
Androstenedione 1 1 2 2
substrate) 71 59 10 15
p- percent of radioactivity recovered in respective zones after

first chromatography in Bush A system
pc- percent conversion of progesterone to testosterone and

17a-hydroxyprogesterone
from the patient with Klinefelter's syndrome (Case E), progesterone

was avidly metabolized and readily converted to testosterone (Table
8). The identity of the small amounts of materials with chromato-
graphic mobilities of androstenedione and 20~dihydroprogesterone
was confirmed by isotopic dilution and crystallization to constant
specific activity.
b) Oligospermia. Six patients with oligospermia were studied.

Except for Case F, with carcinoma of the contralateral testicle, the
oligospermia was of unexplained origin. The laboratory findings in
these patients, except for Case F, are summarized in Table 9. The
appearance of biopsies is illustrated in Fig. 2. The distribution
of radioactivity after first chromatography in Bush A system is
illustrated in Table 10. Again, a marked difference in the capacity
of the tissu~s from the different patients to metabolize progester-
one can be seen. In two patients (land J), we were able to obtain
P.T. levels and T.P.R. It is of interest to note that patient J
had both depressed P.T. levels and T.P.R. and no elevation of gona-
dotropin levels. His testicular tissue metabolized progesterone
avidly, but the conversion to testosterone was very low, only 1.1%.
In the Bush A system, the bulk of the material showing the Rf value
of testosterone and 17a-hydroxyprogesterone did not turn out to be
either of these two metabolites. The identity of this metabolite
and the metabolite migrating like androsterone are presently being
investigated. Patient I's testosterone production rates and plasma
levels were within normal limits for this age male (Table 9), and
while his testicular tissue did not metabolize progesterone as
avidly as J, it formed testosterone more readily and converted pro-
gesterone to 17~hydroxyprogesterone very actively.
In spite of the biochemical differences, the histological

appearance of the testicular biopsies of patients I and J was simi-
lar (Fig. 2b,c). Both show spermatogenic activity, and the Leydig
cells in Case I seem to be actually less prominent than in Case J,
the latter showing large clumps of Leydig cells in the interstitial
area. Comparison of biopsies from Cases K, Land M and the bio-
chemical data again reveals lack of correlation. The most active
metabolism of progesterone and highest conversion to testosterone
was observed in tissue from Case M. Histological examination of the
biopsies, however, showed well developed Leydig cells in Cases K and
L, and only sparsely distributed Leydig cells in Case M (Fig. 2e,d,
f). The least active progesterone metabolism observed in Case F was
associated with numerous Leydig cells forming large clumps in the
biopsy tissue (Fig. 2a).
4) Metabolism of Progesterone by Testicular Tissue from a Patient

with the Syndrome of "Testicular Feminization"
The histologic appearance of the biopsy is illustrated in Fig.

3. The tubules are lined with Sertoli cells and surrounded by mas-
Table 9. Laboratory Data on Patients with Oligospermia
Sperm Count
Case m~/100 ml mg/24 hrs. I.U./24 hrs. I.U./24 hrs. million/ml
I 981 9.82 1.7 2.1 <1

J 343 5.74 1.0 3.9 18-50
K 74.0 3.2 8-37
L 38.5 6.2 10-40
M 4-10
sesof Leydig cells. Distribution of radioactivity in a Bush A chro-

matogram and percent conversion to testosterone and 17a-hydroxypro-
-gesterone are summarized in Table 8. The progesterone was actively
metabolized (15% remained unconverted) and considerable amounts of
17~hydroxyprogesterone (22%) and testosterone (12%) were formed.
A large proportion of progesterone (40%) was converted to highly
polar metabolites. Thin layer chromatography of the polar metabo-
lite revealed material with the mobility of 17 ~-estradiol. Its
identity was confirmed by crystallization to constant specific ac-
tivity (Table 5).
DISCUSSION
In most instances the etiology of spermatogenic defects in the

human male still falls into the category of "idiopathic". A number
of reasons are responsible for this. Our understanding of the mech-
anisms involved in hormonal control of mammalian, and particularly
human, spermatogenesis is meagre (16). There are a lack of precise
techniques for quantitative evaluation of sperm production by the
testes, and the techniques for assessment of testicular androgen pro-
duction are difficult. While it has been clearly established that
gonadotropins are essential for testicular function, the role of an-
drogens, particularly in human spermatogenesis, remains obscure.
The possibility that steroids other than testosterone may play an
important role in spermatogenesis has been suggested.
Techniques for relatively precise measurements of testosterone

plasma levels and production rates are now available. However, the
problems of extratesticular conversion of testosterone precursors,
as well as testosterone and its metabolites, create difficulties in
studies of lesions in the steroid biogenetic pathways.
A technique permitting an investigation of pathways leading to

the biogenesis of androgens in testicular tissue has been applied in
the present study in an attempt to look for possible enzymic lesions
in testes with various forms of spermatogenic abnormalities. This
t:
0)
TABLE 10. METABOLISM OF PROGESTERONE BY TESTICULAR TISSUE FROM PATIENTS WITH OLIGOSPERMIA
PATIENTS
I J K L M F
P pc P pc P pc P pc P pc P pc
METABOLITES
"Polar" compounds 18 47 15 21 12
Testosterone 3.0 1.1 1.2 1,2 2.0 0.6
49 50 23 30 44 12
17~hydroxyprogesterone 32.0 2.0 16.0 18.0 11.0
20~dihydroprogesterone 2 4 10 6 6 2 !'"
(f)
Androstenedione 1 2 2 1 0.6 1 iT!
Z
c:>
Androsterone 33 m
::a
Progesterone (unconverted G'l
m
,::a
substrate) 29 6 41 36 28 59
~
p - percent of radioactivity recovered in respective zones after first chromatography in Bush A
()
"
system :J:
pc - percent conversion of progesterone to testosterone and 17a-hydroxyprogesterone m
,::a
>
Z
o
A
~
(f)
~
::::j
:J:
technique can only provide information concerning the presence or

activity of the various enzymes involved in the steroidogenic path-
way, and its quantitative aspects are difficult to evaluate when in-
formation concerning endogenous pools of the steroid precursors and
metabolites is not available. Concentrations of the various ste-
roids in testicular tissue, their production rates or the secretory
activity of the tissue cannot be assessed by this technique. Fur-
thermore, utilization of progesterone in experiments described here
limits the evaluation of the pathways and provides no information
concerning the capacity of tissue to synthesize the steroid nucleus
de novo from acetate, nor its capacity to convert cholesterol to
C-21 and C-19 compounds. In spite of these limitations, this tech-
nique provides a sensitive tool for in vitro studies of the enzymes
involved in steroid biosynthesis.
A positive correlation between plasma levels and production

rate of testosterone and the in vitro capacity of testicular tissue
to metabolize progesterone to testosterone has been demonstrated.
In patient I, the normal testosterone production rate and plasma
level were associated with the active conversion of progesterone to
testosterone. In patients G and H, the markedly depressed testos-
terone production rates and plasma levels were associated with mar-
ginal conversion of progesterone to testosterone in vitro.
The isolation and identification of 17tt-hydroxyprogesterone,

androstenedione, 20a-dihydroprogesterone and testosterone indicate
the presence of 17~-hydroxylase, 17-20-lyase, 17 i3 -hydroxysteroid
dehydrogenase and 201-hydroxysteroid dehydrogenase in testicular
tissues of older males, males of reproductive age with various forms
of oligospermia, and patients with Klinefelter's syndrome and testi-
cular feminization. Also, the capacity to aromatize ring A and form
17!3 -estradiol by tissues from elderly individuals and a patient
with testicular feminization was demonstrated.
The above mentioned enzymic activities have been amply demon-

strated to occur in testicular tissue from a variety of mammalian
species. However, except for a number of studies with testicular
tissues from patients with testicular feminization, only a limited
number of experiments with non-tumorous human testicular tissue has
been reported (3,5,6,8); and those reported, except for study of one
specimen of tissue from a 16-year old male, were performed on older
individuals suffering with carcinoma of the prostate.
In studies with immature rats treated with HCG, Steinberger

and Ficher (17) called attention to the lack of correlation between
the histological appearance of Leydig cells and the ability of the
tissue to convert progesterone to testosterone. The studies re-
ported here demonstrated a similar lack of correlation in human tes-
ticular tissue. Of partial interest is the biopsy from patient G
showing masses of Leydig cells in the form of "pseudoadenomas", but
only marginal capacity for in vitro metabolism of progesterone,

with low plasma levels and production rates of testosterone, asso-
ciated with elevated urinary levels of gonadotropins.
Similar lack of correlation was observed in biopsies of pa-

tients with oligospermia. This can be illustrated by comparing
Cases I and J. The tissue from Case J shows large clumps of Leydig
cells, while Case I shows a small number of dispersed Leydig cells.
Both tissues metabolize progesterone avidly, but in an entirely
different manner, tissue from patient I converting progesterone
more readily to 17a-hydroxyprogesterone and testosterone, while tis-
sue from patient J was almost unable to perform 17~hydroxylation.
These findings clearly demonstrate a dichotomy between morpho-

logical characteristics and steroid biogenetic capacity of Leydig
cells, a phenomenon not appreciated in the past, but one which pro-
vides further evidence for the concepts of steroid biosynthetic le-
sions in Leydig cells. Furthermore, these observations suggest that
microscopic interpretation of testicular biopsies needs to be tem-
pered with considerable caution when judgments concerning the sec-
retory capacity of the Leydig cells are made.
Studies with tissue from the patient with Klinefelter's syn-

drome (Case E) demonstrated its capacity to convert progesterone to
17~hydroxyprogesterone, androstenedione and testosterone. In fact,
it showed the greatest capacity for testosterone formation of all
tissues studied, except for that from testicular feminization.
These studies suggest that the pathway of progesterone to testos-
terone is intact in the testis of a patient with Klinefelter's syn-
drome. Since, to our knowledge, this is the first report of such a
study in Klinefelter's syndrome, evaluation of further cases will be
necessary before a generalization can be made.
Demonstration of an active steroid biogenetic pathway in the

testis from a case of testicular feminization in this study con-
firms a number of similar findings in the literature. The detec-
tion of 17 ~-estradiol confirms the findings of some investigators
(18-21), but not of others (22-24).
In summary, we have demonstrated that it is feasible to utilize

minute amounts of human testicular tissue obtainab"le by means of sur-
gical biopsy for the study of steroid biogenetic pathways. Formation
of testosterone from progesterone was demonstrated in testicular
tissue of elderly individuals with carcinoma of the prostate and in
the testicular tissue of man of reproductive age suffering with azo-
ospermia, oligospermia, Klinefelter's syndrome and testicular femin-
ization. In some cases, the lack of correlation between the ability
of the testicular tissue to form testosterone and the microscopic
characteristics of the Leydig cells was observed. In cases where it
was studied, a correlation between the ability of testicular tissue
Fig. 1. A-F
Fig. 2. A-F
Fig. 1. A) Case A - note presence of spermatogenesis, including

mature spermatids and a small group of Leydig cells (arrow). B)
Case B - note presence of spermatogenesis, including mature sper-
matids and a number of large clumps of Leydig cells (arrows).
C) Case C - note presence of spermatogenesis, including mature
spermatids and occasional small groups of Leydig cells (arrows).
D) Case G - note two completely atrophic seminiferous tubules
(arrows) and a Leydig cell "pseudoadenoma" (see A). E) Case H -
note immature seminiferous tubules lined by Sertoli cells and pri-
mitive spermatogonial cells and lack of Leydig cells in the inter-
stitial areas. F) Case E - note seminiferous tubules containing
germinal cells, including advanced spermatids and masses of large
Leydig cells in the interstitial areas.
Fig. 2. A) Case F - note seminiferous tubules containing Sertoli

cells and spermatogonia, and masses of Leydig cells in the inter-
stitial area (arrows). B) Case I - note spermatogenic activity in
seminiferous tubules and occasional small groups of Leydig cells
(arrow). C) Case J - note spermatogenic activity and large clumps
of Leydig cells (arrow). D) - Case K - note spermatogenic activity
and clumps of Leydig cells (arrow). E) - Case L - note spermato-
genic activity and clumps of Leydig cells (arrows). F) Case M -
note spermatogenic activity and sparsity of Leydig cells.
Fig. 1. Note immature seminiferous tubules lined with Sertoli cells

embedded in a mass of Leydig cells.
to convert progesterone to testosterone and plasma testosterone

levels and production rates was demonstrated. A variation in the
steroid biosynthetic pathways in testicular tissue from different
patients with oligospermia was demonstrated. The observations
support the concept of steroid biosynthetic lesions in some cases
of testicular disorders.
ACKNOWLEDGEMENTS
This study was supported by a grant from the Ford Foundation.

The excellent technical assistance of A. Gdowik and o. Luckyj is
gratefully acknowledged. Acknowledgement to Drs. A.L. Schneeberg,
J.M. Schneeberg and Dr. R. Klaus for providing the tissues of pa-
tients undergoing orchiectomy and for performing surgical biopsies.
REFERENCES
1. Savard, K., Dorfman, R.I. and Poutasse, E., J. Clin. Endocr.,

Q: 935, 1952.
2. Lucas, M., Whitmore, W.F. Jr. and West, C.D., J. Clin. Endocr.,
17: 465, 1957.
3. Carstensen, H.C.H., Acta Soc. Med. Upsal., 66: 129, 1961.
4. Slaunwhite, W.R. Jr., Sandberg, A.A., Jackson, J.E. and Staubitz,
W.J., J. Clin. Endocr., 22: 992,1962.
5. Axelrod, L.R., Biochim. Biophys. Acta, 97: 551, 1964.
6. Sharma, D.C., Racz, R.I. and Schoen, E.J., Acta Endocr. (Koben-
havn) , 56: 726, 1967.
7. Schoen,~.J., Acta Endocr. (Kobenhavn), 56: 56, 1967.
8. Tamaoki, B.I. and Shikita, M., In Steroid Dynamics, eds. Pincus,
G., Nakao, T. and Tait, J.F., Academic Press, New York, p. 493,
1966.
9. Danezis, J., Folia Bioch. et Biol. Graec~l: 41, 1966.
10. Steelman, S. and Pohley, F.M., Endocrinology, 53: 604, 1953.
11. Parlow, A.F., In Human Pituitary Gonadotropins, ed. Albert, A.,
Charles C. Thomas, Springfield, Ill., p. 300, 1961.
12. Horton, R., Kato, T. and Sherrins, R., Steroids, 10: 245, 1967.
13. Rivarola, M.A., Saez, J.M., Meyer, W.J., Jenkins, M.E. and
Migeon, C.J., J. Clin. Endocr., 26: 1208, 1966.
14. Steinberger, A., Smith, K.D., Steinberger, E. and Perloff, W.H.,
J. Albert Einstein Med. Center,Q: 5, 1964.
15. Steinberger, E. and Ficher, M., Steroids, 11: 351, 1968.
16. Steinberger, E., Physiol. Rev., 1970 (in press)
17. Steinberger, E. and Ficher, M., Biol. Reprod. 1: 119, 1969.
18. Sharma, D.C., Dorfman, R.I. and Southren, A.L., Endocrinology,
76: 966, 1965.
19. Neher, R., Kahnt, F.W., Roversi, G.D. and Bompiani, A., Acta
20. Kase, N. and Morris, J. MeL., Amer. J. Obstet. Gynec., 21: 102,
1965.
21. Rice, B.F., Cleveland, W.W., Sandberg, D.H., Ahmad, N.,
Politano, V.T. and Savard, K., J. Clin. Endocr., 12: 29, 1967.
22. Griffiths. K., Grant, J.K. and Whyte, W.G., J. Clin. Endocr.,
23: 1044, 1963.
23. David, R.R., Wiener, M., Ross, L. and Landau, R.L., J. Clin.
Endocr., 25: 1393, 1965.
24. Wade, A.P., Wilkinson, G.S., Davis, J.C. and Jeffcoate, T.N.A.,
J. Endocr., 42: 391, 1968.
DISCUSSION
HUDSON: Dr. Steinberger, I was intrigued that the patient with

Klinefelter's syndrome had a normal value for plasma LH. Secondly,
in one of the patients, there was a testosterone production rate of
5.7 mg. per day and a plasma level of 343 nanograms per 100 m1. I
think these are correctly quoted. A rough calculation would give
a metabolic clearance rate of about 1700 liters per day. This is
certainly well outside the limits of normal males. Thirdly, in one
of the incubation studies, 33% pregnenolone and 4% 20-a -hydroxy-
progesterone was found. This is of relevance in view of the paper
by Lacey and Petit in the British Medical Bulletin (Jan. 1970) in
which they gave a clear indication that Sertoli cells have two met-
abolic pathways, one leading to testosterone and the other leading
to reduced metabolites of progesterone. I wonder whether in your
study this represented Serto1i cell activity.
E. STEINBERGER: In answer to your first question; yes, the urinary

LH levels in the patient with Klinefelter's syndrome were within
normal levels and FSH was elevated. In answer to the second ques-
tion, I do not recall offhand what was the specific patient's clear-
ance rate but the value in our laboratory for a group of 18 normal
men was about 1500 + 100 (S.E.) liters per day. Recently Lipsett
et a1. reported a clearance rate of over 1200 liters and Southern
et a1. a rate of about 1300 liters in normal individuals. These
values are very close to your "rough" calculation. In answer to
your third question, I have not read Dr. Lacey's paper and cannot
comment on the possibility that the reductase activity resides in
the Sertoli cells. However, as a sheer speculation, I would doubt
this since in rat Leydig cell cultures we can detect 5-a-reductase
while in organ cultures composed primarily of Sertoli cells we do
not.
CHANDLEY: Dr. Steinberger, you were quite emphatic in stating that

your four or five Klinefelter patients showing spermatogenesis in
the testis were in fact not mosaics. I wonder how many tissues you
actually looked at to establish this fact. It is much easier to say
that a patient is a mosaic than to say that he is not, until you have
looked at almost every cell in the body. I know that there have been
one or two previous reports claiming spermatogenesis in Klinefelter's
syndrome; but there has always been the doubt expressed that these
may be XY/XXY mosaics.
E. STEINBERGER: I agree with you fully. As you recall, in our

first paper on Klinefelter's syndrome with spermatogenesis published
about five years ago, we analysed only blood and we found no mosaic
in that case. Since that time we studied a case where blood, skin
and testicular tissue were analysed and again no mosaic was found.
However, you can always come back and argue the point that we did
not look at all cells and have missed cells with XY. Unfortunately
one is limited, particularly when studying tissue like the testis,
in the number of good metaphases when dealing with a biopsy spec-
imen. So I did not mean to be emphatic that these patients are not
mosaics. I meant to state emphatically that we were'unable to prove
mosaicism in these patients.
PAULSEN: In one of our patients with Klinefelter's syndrome we had

to examine 209 metaphase plates from testis tissue culture before
finding a single example of XXY configuration. It is essential to
have sufficient metaphase plates for study before one can disprove
the existence of mosaicism.
FORD: My point is also concerned with the possibility of mosa1C1sm

in the Klinefelter patients. I would like to raise the possibility
of a rather special kind of mosaicism involving the germ cells, but
not the soma, arising out of. the observations on the XXY human males.
I believe I am right in saying that 15 cases have been examined so
far and, in 14 of the 15, the spermatocytes were all XY, one of the
Y chromosomes had been lost. In the other, the majority of the
spermatocytes were XY. So, may I suggest the possibility of a par-
allel process occurring sometimes in the Klinefelter with loss of
one of the two X chromosomes to give normal XY cells?
E. STEINBERGER: I think that we should bear in mind that study of

karyotypes in cultured testicular tissue should not be taken as syn-
onymous with study of karyotypes of germinal cells because most like-
ly we are not looking at karyotypes of germinal cells but probably
of fibroblasts. So your point is very well taken, the germ cell may
very well exhibit a different karyotype.
MANCINI: Dr. Steinberger, as far as I know, this is the first time

there has been an in vitro demonstration of a disturbance in the
conversion of progesterone to testosterone in oligospermic patients.
In terms of the relationship between spermatogenesis and in vitro
steroidogenesis I would like to know if you have had the chance to
study the "fertile" eunuch, because in this condition it has been
claimed that Leydig cells cannot be identified morphologically, but
that spermatogenesis is almost complete.
E. STEINBERGER: It is a fascinating question. I did not have the

opportunity to study a case of "fertile" eunuch. Incidentally, the
patient you mentioned was treated with Pergonal. We elected this
form of therapy because we felt that by simply increasing the rate
of steroid production we might elevate testosterone production.

On therapy the patient's sperm count increased to 150 million and
his wife became pregnant. The other patient with oligospermia and
low gonadotropin levels was also treated with Pergonal with good
results. Thus in a group of six oligospermic patients we were able
to make a specific diagnosis in two and both of them responded to
treatment.
LIPSETT: I think Dr. Steinberger's findings may be very important.

I therefore want to comment a little critically about them. First,
I would like to ask you why you decided to use a low substrate to
enzyme concentration? If one wants to test the enzymes, on"e would
like to use a concentration of substrate to enzyme that is 100:1
or 200:1, because when one uses a low ratio, one may be affecting
many factors, such as pool size, rate of penetration and so on,
which can be overcome by using very high levels.
E. STEINBERGER: In response to your comment concerning substrate

to tissue ratios, I would like to state that the ratio we employed
was comparable to that used by most workers who studied steroid
metabolism in testicular tissue slices or minces. Furthermore,
your concern of possible lack of saturation of the enzyme system
by substrate in our incubates is well taken. However, since we had
considerable quantities of unconverted substrate remaining in the
flask by the end of the incubation period this concern may not be
warranted. We should also keep in mind the problem of substrate in-
hibition.
LIPSETT: You are not limited by that, because you can always in-
crease the amount of substrate and have a higher substrate to tissue
ratio.
ROSEMBERG: Dr. Steinberger, I wondered what was your rationale for

using Pergonal in oligospermic patients? You mentioned that they
had a normal production rate of testosterone. Now if they did, did
you want to see the effect of FSH plus that LH in the patients?
How much Pergonal did you use? What was the total FSH and LH activ-
ity given and how soon did you obtain satisfactory results?
E. STEINBERGER: The reason why we used Pergonal was as follows:

Since we did not know how much testosterone the testes would pro-
duce in response to the gonadotropins, we were worried that the tes-
tosterone levels might become sufficiently high to block the pa-
tient's own FSH and thus possibly interfere with spermatogenesis.
By giving Pergonal we were assured of the presence of at least ex-
ogenous FSH. As far as the doses are concerned, we used two ampoules
three times weekly for about six months. He responded wi~hin a
month with the first increase in sperm count.
CHRISTENSEN: One of the generalities that seemed to emerge from

Dr. Steinberger's presentation was that the apparent number of in-

terstitial cells does not seem to correlate very well with the lev-
el of androgen production from progesterone in vitro. In fact, it
seemed inversely correlated. I wonder whether this is really sur-
pr1s1ng. If interstitial cells are not performing as they should,
then the pituitary will respond by producing LH, which may cause
some interstitial cell hyperplasia which would be in line with your
findings.
E. STEINBERGER: I agree with you, but I would like to correct one

point. I did not mean to correlate testosterone production rate
and Leydig cells, but only the pattern of steroid metabolism, which
really can not be very well correlated with the production rate.
SUBCELLULAR STRUCTURE AND SYNTHESIS OF STEROIDS IN THE TESTIS
Kenneth Savard
Endocrine Laboratory, Department of Biochemistry and
Medicine,University of Miami School of Medicine, Miami,
Florids, U.S.A.
It is becoming increasingly evident that subcellular localiza-

tion of enzymes, substrates, cofactor systems, transport systems
for oxygen, electrons, etc. are factors to be considered in the pro-
cess of elaboration of steroid hormones. The production of the
characteristic array of steroids by specific endocrine tissues, and
more important, its control by gonadotropin, cannot at the present
time be explained exclusively in terms of enzymes alone. Studies of
various gonadal tissues in recent years reveal that the same enzymes
or families of enzymes (3~-hydroxysteroid dehydrogenase-isomerase,
17-hydroxylase, C17,20-lyase, the 19-hydroxylase-aromatase system,
17~-hydroxysteroid dehydrogenase) are present in testicular tissues,
in follicular tissue, in the human corpus luteum and in ovarian in-
terstitial tissue. Yet each of these gonadal elements elaborates
its own u~ique steroidal product.
It seems appropriate at this time to review the field in its

present status. It is to be hoped that calling attention to the im-
portance of the subcellular distribution of the enzymes of the ste-
roidogenic pathway will cause an awareness of what is well document-
ed in the testis, what is not clearly established and what work re-
mains to be done.
The overall picture of the process of steroidogenesis in the

cell is depicted in Fig. 1. This is a highly idealized generaliza-
tion which is only partially true for the testis, and some data are
drawn from studies of ovarian and adrenal tissues where the process
and distribution seem to be largely the same.
459
g
LH
------.------------------- ..... --------
(ICSH)
Q ATP
inactive ADENYL
cell adenyl phospho-
membrane
cyclase ? CYCLASE [)
C I-S-'5-'--A-M-P-'1
diesterase
'-- .." 5-AMP
?
•
---------------------0.------------
plasma, }
tissue stores, CHOLESTEROL .. [20(1-HO-cholesterol ]
mito de novo syntb, ~
chondrial I 5-pregnenolone I. [20(1, 22-diHO-cholesterol ]
I
,
_._._._._._._._._._.L._._._._._._._._._._._._
t •
5-pregnenolone _ progesterone
microsomal J
17-HO-pregnen. _
J
17-HO-progest.
and + t
DHA ---.. and.dione - E-l
soluble ~ t +
and.diol ---.. testosterone --+- E-2
?<:
(f)
>
Fig. 1. Steroidogenesis in the testis. ~
:0::1
o
ENZYMES OF THE STEROIDOGENIC PATHWAY
The enzymes involved in the elaboration from pregnenolene of

progesterone, androgen, and estrogen appear to be localized largely
in the microsomal fraction of testicular homogenates. Much of this
important work has been carried out by Tamaoki and his group in
Japan, (1-7) and by Samuels and his associates (8,9). The Japanese
workers have carefully characterized, by electron microscopy, the
purity of their centrifugal fractions and have concluded that except
for the 20-hydroxysteroid dehydrogenase, which is in the soluble
fraction, the 17-hydroxylase-lyase system, the 19-hydroxylase-aro-
matase system (7), the 17~-hydroxysteroid-isomerase and the 3~ -hy-
droxysteroid dehydrogenase-isomerase system are microsomal. A re-
view of this literature has appeared (1,10). The Japanese workers
have studied this intracellular distribution in the testes of rats
(3), mice (4), rabbits (5), guinea pigs (6) and men (7).
It has been reported that seminiferous tubules have the capa-

city to convert progesterone to androgens in vitro (11); more re-
cent work has shown that only the Leydig cells and not the tubules
can convert cholesterol into androgen (12). Both studies however
point to the presence of certain enzymes of the biosynthetic path-
way in the tubular elements. The physiologic significance of these
observations is hard to interpret since the cnncensus is that the
Leydig cells are the major source of androgens in the testis.
THE FORMATION OF PREGNENOLONE
This principal and presumably obligatory precursor of steroid

hormones in the testis, as well as in other tissues, is derived
from cholesterol (Figo 1) by the enzyme complex called the "choles-
terol side-chain cleavage" enzyme system (12,14). This enzyme com-
plex is totally located in the mitochondria of all steroidogenic
tissues examined, including the testis (15), and consists of three
discrete enzymes: 20a-hydroxylase, 22-hydroxylase, and C-20,22
lyase (or 20a, 22-dihydroxycholesterol side-chain cleavage enzyme).
This mitochondrial enzyme complex has been studied more thor-

oughly in adrenal tissue than in the testis, and the sequence of re-
actions shown in Fig. 1 may not necessarily be the correct one, ac-
cording to recent studies (16) in homogenates of adrenal tissue.
The "cholesterol side-chain cleavage" enzyme complex of tes-

ticular tissue appears to be the limiting step in the rate of over-
all steroidogenesis in the testis, (17) as it appears to be in adre-
nal and ovarian tissues. The system is dependent on gonadotropin
which restores, or prevents the decline, in the rate of this system
following hypophysectomy (17).
Stimulation of steroid hormone production by appropriate tropic

462 K. SAVARD
hormones, in intact cell preparations of adrenal, corpus luteum,

ovarian interstitium (of the rabbit) and testis tissues, is consid-
ered to be mediated by the cyclic nucleotide, 3 1 , 5 ' -AMP. In the
testis, the complete picture has not yet been drawn. However, con-
version of labelled cholesterol to testosterone by testicular pre-
parations has been reported by several laboratories and is stim-
ulated by gonadotropin (18) and by 3 1 , 5 ' -AMP (19). In perfused pre-
parations, 3 1 , 5 1 -AMP, like ICSH-LH, causes increased biosynthesis of
the total steroidal product of the testis which includes testoster-
one, DRA, androstenedione, l7-hydroxyprogesterone, l7-hydroxypreg -
nenolone and progesterone (van der Molen and Eik-Nes, quoted in ref.
20). This has also been observed in intact slice preparations (21,
22).
The mechanism by which 3 1 , 5 ' -AMP stimulates the rate of the

transformation of cholesterol to pregnenolone is entirely unknown.
It is under investigation in the adrenal (23) and ovarian tissue
(24,25) but is relatively unexplored in testicular tissues (26).
The stimulating action of 3 1 , 5 ' -AMP can be inhibited by inhibitors
of protein synthesis such as puromycin and chloramphenicol (24,27).
The nature of this phenomenon in adrenal (23), ovarian and testic-
ular tissues is unknown.
Although the transformation of cholesterol to steroid is a mi-

tochondrial process, the source of cholesterol is not known. It
can be presumed that much of testicular cholesterol originates in
the plasma. However, l4C from small molecular precursors like ace-
tate and mevalonate is incorporated into steroids (27,28) and ste-
trols (28) by isolated testicular preparations. De novo synthesis
of cholesterol therefore occurs in testicular tissue and in Leydig
cells (12,21) although the subcellular component or components in-
volved are not known.
ADENYL CYCLASE AND 3 ' , 5 ' -AMP
The mediating role of 3 1 , 5 ' -AMP in the stimulation of steroid-

ogenesis in adrenal. ovarian and testicular tissues is now general-
ly accepted, though its mechanism is not understood. It is also
generally considered that the step in the biosynthetic process stim-
ulated is the mitochondrial "cholesterol side-chain cleavage" en-
zyme system,since this seems to be the rate-limiting step. An
early effect of gonadotropin and ACTH (in their respective tissues)
would consequently be to cause increased concentrations of endogen-,
ous 3', 5 ' -AMP in the tissue. This has been observed in ovarian (29)
and in testicular tissues (30). It is assumed that the tropic hor-
mone activates adenyl cyclase (31,32) in the membrane of the cell,
where transformation of ATP to 3 1 , 5 1 -AMP presumably occurs (Fig. 1).
The transport or migration of the newly formed 3 1 , 5 ' -AMP from the
membrane to mitochondria where its steroidogenic effect takes plac~
is unexplained.
A little light is being shed in the area of the activation of

gonadal adenyl cyclase by ICSH-LH, the nature of which is virtually
unknown in any tissue. It would now appear that a prostaglandin,
PGE? mav mediate the effect of the gonadotropin on adenyl cyclase.
Recent s~udies by J.M. Marsh of the Endocrine Laboratory in Miami have
shown that PGE2, added to incubating slices of bovine corpora lutea,
causes a rapid and significant activation of adenyl cyclase with a
presumed concomitant increase in endogenous 3',5'-AMP concentrations
(35). This same prostaglandin has been recorded as mimicking both
ICSH-LH and exogenous 3',5'-AMP on steroidogenesis in this same
tissue (36). This postulated mediating role of PGE2 on adenyl cy-
clase will require the demonstration that the gonadotropin will in-
crease concentrations of PGE2 in the tissue; this has not been
done.
EFFECTS OF GONADOTROPIN IN THE TESTIS
It is well known that both testicular weight and protein syn-

thesis are influenced by purified preparations of FSH and ICSH-LH
(33,34). It seems however that ICSH-LH is more effective (on a
weight basis) than FSH in stimulating steroidogenesis in testicular
slices (27); whether the steroidogenic effect of FSH is due to its
ICSH-LH contaminant is not clear (27), although it is interpreted
to be so (20). Murad and associates (29) have recently reported
the stimulation of adenyl cyclase in testicular homogenates by both
purified preparations of ICSH-LH and FSH at reasonably low concen-
trations. The activity of the FSH preparation was greater than
could be attributed to its LH content. Murad et ale however point
out that their testicular preparation contained elements of connec-
tive tissue, tubules, nerves, blood vessels, besides Leydig cells,
and that there could be more than one cellular type of cyclase re-
sponding to the gonadotropins.
To summarize and perhaps to predict, it would seem that ste~

roidogenesis in the testis is part of a larger natural scheme where
the process in the male gland is not unlike that in ovarian and
adrenal tissues. Biochemical phenomena in the membrane and the mi-
tochondria are comparable and almost identical, save for the bind-
ing sites for ACTH or gonadotropin. Distinctiveness and physio-
logical uniqueness, that of the production of androgen as major
product, appears to reside in parameters associated with the endo-
plasmic reticulum. It is here, in the microsomal fraction, that
the enzymes and other factors are localized and determine the
course of steroidogenesis leading to androgen formation and secre-
tion by the testicular cell.
The ultra structure of the.steroid-forming cell thus is be-

coming more prominent as knowledge of the subcellular distribution
of the enzyme system of the steroidogenic process evolves. The
well-established changes in ultrastructure of the cell following
464 K. SAVARD
hypophysectomy, and their prevention or reversal by treatment with

appropriate tropic hormone (reviewed by Christensen and Gillim (10)
point to the role of tropic hormone in the mainta~ance of these
fragile structures. It is obvious from Fig. 1 that a high degree
of organization and transport in the steroidogenic cell is implied.
Even more complicated is the manner by which tropic hormone accel-
erates steroidogenesis. If the schematic representation of Fig. 1
is true, it is plain why disruption of the cell's integrity by
freezing or by homogenization abolishes the stimulatory effects of
tropic hormone or of exogenous 3',5'-AMP. It follows then that
far more attention and awareness in maintaining and preserving
cellular integrity must be given than in the past. Only in this
way will biochemical studies in vitro have physiological relevance,
and provide further insight into this complex process.
ACKNOWLEDGEMENT
Supported in part by grant HD-03248 National Institute of

Child Health and Human Development, U.S. Dept. of Health, Education
and Welfare.
REFERENCES
1. Tamaoki, B., Inano, H. and Nakano, H., In The Gonads, ed.

McKerns, K.W., Appleton-Century-Crofts, Inc. New York p. 547,
1969.
2. Inano, H., Harl, Y. and Tamaoki, B., ~ Endocrinology of the
Testis, eds. Wolstenholme, G.E.W. and O'Connor, M. J. and A.
Churchill Ltd. London p. 105, 1967.
3. Shikita, M. and Tamaoki, B., Endocrinology, 76: 563, 1965.
4. Murato, S., Shikita, M. and Tamaoki, B., Steroids, 1: 509,1965.
5. Tamaoki, B. and Shikita, M. In Steroid Dynamics,eds. Pincus, G.,
Nakao, T. and Tait, J.F., Academic Press, New York, p. 493,
1966.
6. Inano, H., Egusa, M. and Tamaoki, B., Biochim. Biophys. Acta,
144: 165, 1967.
7. Murato, S., Shikita, M. and Tamaoki, B., Biochim. Biophys. Acta,
117: 241, 1966.
8. Mahajan, D.K. and Samuels, L.T., Fed. Proc.,ll: 209, 1962.
9. Matsumoto, K. and Samuels, L.T., Endocrinology, 85: 402, 1968.
10. Christensen, A.K. and Gillim, S.W. In The Gonads, ed. McKerns,
K.W., Appleton-Century Crofts, Inc., New York, p. 415, 1969.
11. Christensen, A.K. and Mason, N.R., Endocrinology, 76: 646, 1965.
12. Hall, P.F., Irby, D.C. and de Kretser, D.M., Endocrinology,
84: 488, 1969.
13. Shimizu, K., Hayano, M., Gut, M. and Dorfman, R.lo, J. Biol.
Chem., 236: 695, 1961.
14. Shimizu~., Gut, M. and Dorfman, R. I., J. Biol. Chemo, 237:
966, 1962.
15. Toren, D., Menon, K.M.J., Forchielli, E. and Dorfman, R.I.,
Steroids,l: 381, 1964.
16. Burstein, S. and Gut, M., Steroids, 14: 207, 1969.

17. Forchielli, E., Menon, K.M.J., and Dorfman, R.I., In The Gonads,
ed. McKerns, K.W., Appleton-Century-Crofts, Inc. New York,
p. 519, 1969.
18. Hall, P.F., Endocrinology, 78: 690, 1966.
19. Sandler, R. and Hall, P.F.,~ndocrinology, 79: 647,1966.
20. Connell, G.M. and Eik-Nes, K.B. In The Gonads, ed. McKerns,
KoWo, Appleton-Century-Crofts, New York p. 491, 1969.
21. Rice, B.F., Cleveland, W.W., Sandberg, D.H., Ahmad, A.,
Politano, V.To and Savard, K., J. Clin. Endocr., 27: 29, 1967.
220 Leymarie, Po and Savard, K., unpublished --
23 0 Garren, L.D., Vitamins and Hormones, 26: 119, 1968.
240 Marsh, J.M. and Savard, K., J. Reprod. Fertil. Suppl., 1: 133,
1966. -
25. Dorrington, J.H. and Kilpatrick, R., Biochem. J., 104: 725,
1967.
26. Sandler, R. and Hall, P.F., Endocrinology, 79: 647, 1966.
27. Hall, P.F. and Eik-Nes, K.B., Biochim. Biophys. Acts, 63: 411,
1962.
28. Tsai, S.E., Ying, B.P. and Gaylor, J.L., Arch. Biochem. Biophys.
105: 329, 1964.
29. Marsh, J.M. Butcher, R.W., Savard, K. and Sutherland, E.W.,
J. Biol. Chern., 241: 5436.
30. Murod. F., Chi, Y.M., Rall, T.W. and Sutherland, E.W., J. Biol.
Chern., 237: 1233, 1962.
31. Murod, F., Strauch, B.S. and Vaughn, M., Biochim. Biophys. Acta,
177: 591, 1969.
32. Marsh, J.M., J. Biol. Chern., in press (1970).
33. Tanoka, K., Wilson, W.O. and McFarland, E.Z., Amer. J. Vet. Res.,
27: 1067, 1966.
35. Marsh, J.M., FEBS Letters, in press (1970)
36. Speroff, L. and Ramwell, P.W., J. Clin. Endocr. in press (1970).
DISCUSSION
JOHNSEN: Dr. Savard, you made reference to a patient with testicular

fibrosis without production of estrogens. Did this fact surprise you
or did it have anything to do with the fact that there was testicular
fibrosis?
SAVARD: The patient was a case of incomplete testicular feminization,

and therefore virtually consisted of a pure Leydig cell preparation,
but we were very pleased that we had data that seemed to be in agree-
ment with Dr. Lipsett; that the Leydig cells are a very poor source
of estrogen, because the quantities of both estradiol and estrone com-
bined are extremely small.
CHAPTER VI
NON-HORMONAL FACTORS INflUENCING

SPERM PRODUCTION AND TRANSPORT
THE BIOCHEMICAL CHARACTERISTICS OF SPERMATOZOA AND SEMINAL PLASMA
Thaddeus Mann
Director, Unit of Reproductive Physiology and
Biochemistry, University of Cambridge,
Cambridge, England
Semen is made up of two components, viz. the spermatozoa and

seminal plasma, which differ from each other physiologically as well
as chemically. Moreover, the physiological and chemical properties
of these two components of semen do not remain constant but undergo
characteristic changes in the course of the period of sperm passage
from the testes to the site of final destination in the female re-
productive tract.
TESTICULAR SEMEN
Testicular semen is a dilute suspension of testicular sperma-

tozoa which can be obtained by means of a catheter from the rete
testis (1). It has few properties in common, physiologically or
chemically, with ejaculated semen. Testicular spermatozoa are im-
motile and incapable of effecting fertilization. Their immaturity
is also reflected in structural pecularities, such as the persis-
tent occurrence of kinoplasmic droplets as well as in certain chem-
ical and metabolic properties, best illustrated by the behavior of
sperm phospholipids.
In ejaculated ram spermatozoa the principal phospholipid is the

choline-based plasmalogen, that is, phosphatidalcholine, which in
the course of endogenous sperm respiration yields oxidizable fatty
acids (2,3). In both testicular and ejaculated ram spermatozoa the
amount of phosphatidalcholine appears to be approximately the same,
but the concentration of other phospholipids, such as phosphatidyl-
choline (lecithin), phosphatidylethamolamine and phosphatidaletha-
nolamine (ethanolamine-based plasmalogen) is distinctly higher in
469
470
testicular spermatozoa, and the fatty acids extracted from these

phospholipids include m~~e palmitic acid and less myristic acid (4).
When incubated with (U- C) glucose, testicular ram spermatozoa ex-
hibit a much higher rate of radioactivity incorporation into the
glycerol component of phospholipids and neutral lipids than the
ejaculated spermatozoa. However, none of that radioactivity appears
in phosphatidalcholine. Like the biosynthetic ability, so too the
rate of metabolic breakdown of lipids differs in testicular and
ejaculated ram spermatozoa. Thus for example, ram testicular sper-
matozoa are much less active than ejaculated spermatozoa in hydro-
lysing either added ethanolamine phosphoglycerides or added leci-
thin (5).
Equally pronounced are the differences between testicular and

ejaculated seminal plasma. A review listing recent findings per-
taining to testicular plasma, that is the fluid collected directly
from the rete testis (mainly from rams) has been published by
Setchell, Scott, Voglmayr and Waites (6). It is evident that al-
though testicular plasma has the same osmolality as other body flu-
ids, the rete testis fluid is watery so that its content of protein
(measured as biuret-reactive material) is only 0.28 g/100 mI. Its
total mineral (ash) content, on the other hand, is similar to that
of blood plasma; but the ionic composition is not the same, there
being less sodium but about three times as much potassium in the
testicular fluid. Most of the free amino acids are present in low-
er concentration in the rete testis fluid than in either testicular
lymph or blood plasma. An exception in this respect is the behavior
of glutamic acid and aspartic acid; these two amino acids are pres-
ent in testicular plasma in concentrations which exceed about ten-
fold those of testicular lymph or blood plasma. There is no glucose
or fructose in the rete testis fluid. An interesting chemical fea-
ture of that fluid is the occurrence of free inositol.
Inositol has been known for some time to be a characteristic

constituent of mammalian seminal plasma. Its concentration is par-
ticularly high in the boar, where it is produced in the seminal ves-
icles. The fluid secreted by the boar seminal vesicles may contain
as much as 3 g inositol/l00 ml, and it is from that fluid that in-
ositol was isolated in a pure crystalline state (7). Though in
much smaller concentrations, inositol occurs also in the seminal
plasma of other mammals, including ram and man (8,9). More recent-
ly it has also been shown to occur in the rete testis fluid, in
concentrations ranging from 0.97 to 0.15 g/100 ml (10).
Of much interest is yet another recent observation pertaining

to the chemical composition of ram rete testis fluid, namely the
occurrence in that fluid of testosterone (11,6), in quantities com-
parable to those in testicular blood and lymph (12). If future
studies should prove conclusively that the plasma of testicular
semen collected from the rete testis does, in fact, originate in
SPERM PRODUCTION AND TRANSPORT 471
the seminiferous tubules, and furthermore, if one accepts as valid

the view that testosterone is generated in the intertubular tissue
(that is not inside but outside the seminiferous tubules), then the
inescapable conclusion is that testosterone can pass through the
wall of the seminiferous tubule and enter its lumen. The movement
of testosterone, or for that matter of any other organic material,
across the wall of the seminiferous tubules is, as yet, poorly un-
derstood. High molecular weight substances at any rate, such as
serum proteins, seem to be unable to pass into the seminiferous
tubules, in spite of the fact that testicular interstitial blood
capillaries are unusually permeable to serum proteins and that such
proteins tend to accumulate readily and at a high rate in extravas-
cular spaces (13,14). Nor is there much information with regard to
the possible involvement of Sertoli cells in the movement of sub-
stances across the wall of seminiferous tubules, but of some signi-
ficance in that respect may be the observation that Sertoli cells
are able to take and transport peroxidase from blood (15). We shall
know more about the formation of the testicular plasma when the ul-
trastructural relations between the elements which compose the walls
of the seminiferous tubules and those which compose the intertubular
tissue are cleared up. In this connection there may be significance
in the recent finding (16) that in the guinea-pig and chinchilla the
testicular interstitium contains in addition to blood capillaries an
extensive system of sinusoidal lymphatics, located in close proximi~
ty on one side to the Leydig cells, and on the other side to the tu-
nica propria of the seminiferous tubules.
EPIDIDYMAL SEMEN
During the period of about two weeks which the spermatozoa re-
quire to pass along the epididymis, they undergo a number of changes,
described as "maturation" or "ripening". In addition, the epididy-
mus also plays a part in the disposal of aged, that is, superfluous
spermatozoa. Ripening is associated with morphological and physico-
chemical changes. The former include the phenomenon of kinoplasmic
droplet migration as well as ultrastructural changes in both the
acrosome and mitochondrid sheath (17,20). The latter involve a
process of progressive dehydration and a rise in both specific grav-
ity and permeability as well as changes in sperm composition, par-
ticularly in the lipid constituents. As spermatozoa pass along the
epididymis of the ram, the concentration of total phospholipid and
non-esterified cholesterol gradually decreases (21). This may be
due either to the metabolic activity of the spermatozoa themselves
or to the influence exerted by the epididymal tissue. It is by no
means certain in what measure the changes associated with sperm rip-
ening depend on reactions within the sperm cells themselves, and to
what extent they are triggered off by extraneous factors originating
in the epididymal epithelium. Apart from phospholipids, epididymal
and ejaculated spermatozoa differ in other properties, and one of
the most interesting differences is the much higher level of pyri-
472 T. MANN
dine nucleotides in epididymal than in ejaculated spermatozoa, a

phenomenon discovered by the Italian research-workers, Bistocchi,
d'Allessio and Leone (22).
The fluid medium for the spermatozoa within the epididymis is

the "epididymal plasma": together, the epididymal sperm and plasma
constitute epididymal semen. Chemically the epididymal plasma is
distinct from both the testicular fluid and the ejaculated seminal
plasma. This is particularly striking in respect of the ionic com-
position. When the sodium and potassium ions are measured at dif-
ferent levels of the bovine epididymis, the K/Na ratio is distinct-
ly higher in the corpus epididymis than at the point of sperm en-
try from the efferent ducts into the caput. In the cauda epididy-
miclis, however, the combined contents of K and Na are lower than in
the corpus, some of the osmotically active ions having been replaced
by organic substances (23,24). Among the organic constituents of
epididymal plasma two deserve special mention on account of their
unusually high concentration, namely glycerylphosphorylcholine (25)
and carnitine (26).
Glycerylphosphorylcholine occurs in such a high concentration

in the epididymis that in some species at least, notably ram, bull,
boar and stallion, a determination of glycerylphosphorylcholine in
semen can be used as a criterion for the evaluation of the epididy-
mal contribution to the whole ejaculate. Glycerylphosphorylcholine
also occurs in human semen but its concentration does not exceed
400 ~ moles or 100 mg/100 ml, i.e. ten times less than in the ram.
On the other hand, human seminal plasma is much richer in another
choline derivative, phosphorylcholine, than the semen of either ram,
bull, boar or stallion. However, phosphorylcholine is probably de-
rived from accessory organs other than the epididymis. Like most
secretory products of the male reproductive tract glycerylphospho-
rylcholine formation is under the influence of the male sex hormone;
its formation is arrested by castration, but can be stimulated by
exogenous testosterone (27).
Carnitine (~-oxy-y-butyrobetaine, C7H15 03N), together with ac-

etylcarnitine, has so far been studied mainly in the rat reproduc-
tive tract. The cauda epididymidis of adult rats contains nearly
100 t.1 moles carnitine, as well as about 5 ~moles acetylcarnitine,
per g dry weight tissue. Most of that carnitine seems to be derived
from the epididymal secretion (26). Whether the high content of
carnitine arises from synthesis de novo, or from accumulation by the
epididymal epithelium, is not known for certain. Much more is known
about the intermediary reactions in carnitine formation. In that
respect investigations on the relationship between the occurrence
in edpdidymal plasma of carnitine and glutamic acid could be of much
interest, particularly in relation to the possibility of carnitine
arising by decarboxylation of glutamic acid, viaY-aminobutyric acid
and butyrobetaine. The high level of carnitine in the epididymal
plasma, and the high level of activity of the enzyme carnitine

acetyl transferase in the spermatozoa, both strongly suggest that
carnitine and acetylcarnitine may play an essential role in semen
metabolism, perhaps in the oxidation of fatty acids.
EJACULATED SEMEN
Practically all that is known at present about the testicular

and epididymal semen derives from studies on species other than man.
But our knowledge of the biochemistry of ejaculated semen includes
a great deal of information on human semen. The account below is
intended as a brief introduction to the biochemistry of the human
ejaculate, and provides an outline of certain chemical methods which
are applicable to human semen.
The intracellular constituents of human spermatozoa appear to

be similar to those encountered in sperm cells of other mammals.
The human sperm nucleus is haploid and contains only half the amount
of deoxyribonucleic acid which one would expect to find in the dip-
loid cells of somatic cells; its average content appears to vary
slightly within the narrow range of 2.2 to 2.6 x 10- 9 mg DNA/sperm
cell (28). This constancy, however, characterizes only normal hu-
man spermatozoa; in subfertile and sterile men (or animals) the DNA
content is said to be much more variable (29-31)0 The intracellular
enzymes present in human spermatozoa also resemble those found in
animals. Human spermatozoal lactate dehydrogenase (LDH) shares with
the animal LDH the occurrence of the unusual LDHIV or X isozyme
This isozyme appears to be testis-specific for man and animals alike.
In the species that have been investigated its formation is linked
up with the process of spermatogenesis. The isozyme first appears
during the late spermatogonial or spermatocyte stage, but not be-
fore. Its substrate specificity and certain other properties are
quite distinct from the other lactate dehydrogenase isozymes (32-
37). Another enzyme characteristic of spermatozoa and of the mature
testis, but not present in the pre-pubertal testis, is ketose re-
ductose (sorbitol dehydrogenase). Discovered in ram spermatozoa
and purified from that material (38), this enzyme was later shown
also to occur in testicular homogenates of a variety of other ani-
mals, and in man (39).
At one time the human spermatozoa as present in ejaculated

semen were erroneously believed to differ from animal spermatozoa
by being incapable of respiring. However, in the light of numerous
studies carried out during the last decade (40-47), this view can
no longer be sustained. It is now certain that human spermatozoa
possess marked respiratory activity. They also contain, in common
with other species, the complete cytochrome system (48,49). The
variations, however, which one encounters in the metabolic rates
of human spermatozoa, appear to be much greater than in animal semen.
This variability characterizes not only respiration but also the
474 T. MANN
anaerobic metabolism, that is fructolysis.
Human seminal plasma represents the combined secretions of the

epididymis, vas deferens, ampulla, seminal vesicle, prostate gland,
and the glands of Cowper and Littre. In chemical composition it
differs quite significantly from the seminal plasma of other mam-
mals. Among its unusual and characteristic components are nitro-
genous bases, such as choline, spermine and spermidine; a whole
range of specific proteolytic enzymes, which are associated with
the process of semen coagulation and liquefaction and responsible
for the breakdown of large molecular weight proteins to smaller
peptides and free amino acids; various other oxidative and hydro-
lytic enzymes, including the powerful acid phosphatase and 5'-nu-
cleotidase; and many other organic substances, including at least
thirteen prostaglanains, a large group of free as well as bound
carbohydrates, and organic acids such as citric, lactic and ascor-
bic acid.
The prostaglandins are a group of unsaturated C20-fatty acids,

characterized by the presence of a cyclopentanone ring which broad-
ly speaking can occur in four forms, known as E, F, A and B. The
\) \0 00
o o 0
OH
extensive literature on prostaglandins was summarized by Pike and

Weeks (50) in their "Prostaglandin Bibliography", and the whole
field has been reviewed by several authors (51-53). The currently
held views on prostaglandins in relation to the reproductive func-
tions in both the male and female were discussed in recent reviews
by Ingelman-Sundberg (54) and Horton (55). Of the thirteen pros-
taglandins so far detected in human 'seminal plasma', the two most
prominent ones are prostaglandin Ei (iia 15-dihydroxy-9-keto-prosta-
i3-enoic acid and prostaglandin E2 (11a15-dihydroxy-9-keto-prosta-
5,13-dienoic acid), each present at a concentration of approximate-
ly 2 mg/l00m1. So far as is known at present, prostaglandins ac-
count for practically the whole of the smooth-muscle stimulating
activity of human seminal plasma. Serotonin (5-hydroxytryptamine)
which at one time was believed to account for at least one part of
the 'oxytocic' activity of semen, has been shown to be absent from
human seminal plasma (56).
The carbohydrates of human seminal plasma are partly free and

partly bound to proteins. The principal free sugar is fructose (57)
but in addition, human seminal plasma contains small amounts of
various other free carbohydrates, including sorbitol, inositol, glu-
cose, ribose and fucose. The bound carbohydrate is composed chiefly
of amino sugar, sialic acid and orcinol-reactive material which on
acid hydrolysis yields galactose, mannose and fucose, but only a
trace of glucose. Glycogen occurs in human semen in traces; the
bulk of bound carbohydrate is present in the form of glycoproteins
(58) •
In addition to the glycoproteins human seminal plasma contains

various other proteins, some of which break down during storage of
semen into smaller fragments (59-61). According to one recent es-
timate (62), based on optical density measurements of polyacrylamide
columns prepared by disc electrophoresis, out of a total of 5.8 g
protein/l00 ml seminal plasma, 13% was present as prealbumin (a
fraction migrating ahead of albumin), 14% as albumin, 36% as ul-
globulin, 157. as u 2-g10bulin, 87. as ~1-g10bulin, 67. as ~2-g10bulin,
and 7% in the form of immunoglobulins; plasma separated from semen
samples with sperm counts below 20,000/~1 contained more prealbumin
and ~1-g10bulin but less albumin. Another significant deviation
from the pattern of protein distribution in human seminal plasma has
been reported by Herrmann (63): in two cases of hypogonadism he
could find no evidence for the occurrence of the so-called 3,72S-
protein (mol. wt. 50,000) which according to this author, is normal-
ly secreted by the seminal vesicles.
Among the organic acids secreted in seminal plasma, citric acid

deserved special attention because of its high concentration in hu-
man semen (up to 0.5%, or even more), and its origin in the prostate,
it provides a good indicator of prostatic function. Just as fructose
can serve as a quantifiable criterion of ampullary and vesicular
activity, so citric acid, and also acid phosphatase, represent a
similarly sensitive and reliable indicator of prostatic activity.
A combined citric acid - fructose determination can be particularly
useful, e.g. as a diagnostic aid for the detection of defects in the
male reproductive tract such as bilateral aplasia of the seminal
ducts and vesicles or occlusion of the ejaculatory ducts. In cases
of this kind, there are no spermatozoa and no fructose in the ejac-
ulates which, however, show at the same time an abnormally elevated
concentration of citric acid (64). There are other instances in
which the chemical analysis of semen can be applied as a diagnostic
aid of reproductive failure in man (65). One to which we are direct-
ing our special attention at present concerns chemical methods of
semen analysis in the detection of disturbances in the ejaculatory
process (66). Interference with the normal nervous mechanism which
controls the pattern of ejaculation, caused e.g. by the administra-
tion of certain drugs, can seriously disturb the order in which the
476 T. MANN
various seminal fractions are ejaculated. In such conditions the

chemical characteristics of the so-called split ejaculates, that is,
of the pre-sperm, sperm-rich, and post-sperm fractions, are markedly
altered, and the characteristic sequence with which the various
fractions are normally voided, is upset or even altogether reversed.
REFERENCES
1. Voglmayr, J.K., Scott, T.W., Setchell, B.P. and Waites, G.M.H.,

J. Reprod. Fertil, 14: 87, 1967.
2. Hartree, E.F. and Mann, T., Biochem. J., 21: 423, 1959.
3. Hartree, E.F. and Mann, T., Biochem. J., 80: 464, 1961.
4. Scott, T.W., Voglmayr, J.K. and Setchell,iB.P., Biochem. J.,
102: 456, 1967.
5. Scott, T.W. and Dawson, R.M.C., Biochem. J., 108: 457, 1968.
6. Setchell, B.P., Scott, T.W., Voglmayr, J.K. and Waites, G.M.H.,
BioI. Reprod. Suppl., 1: 40, 1969.
7. Mann, T., Proc. Roy. Soc. (BioI.) 42: 21, 1954.
8. Hartree, E.F., Biochem. J., 66: 1317 1957.
9. Nixon, D.A., J. Reprod. Fertil.,~: 419, 1964.
10. Setchell, B.P., Dawson, R.M.C. and White, R.W., J. Reprod.
Fertil.,1L: 219, 1968.
11. Voglmayr, J.K., Waites, G.M.H. and Setchell, B.P., Na1!\:re,
210: 861, 1966.
13. Everett, N.B. and Simmons, B , Circ. Res., 6/3, 307, 1958.
14. Hancini, R.E o, Vilar, 0., Alvarez, B. and Seigner, A.C.,
J. Histochem. Cytochem., 13: 376, 1965.
15. Reddy, J.K. and Svoboda, D.J., J. Cell BioI., 35: 379, 1967.
16. Fawcett, D.W., Heidger, P.M. and Leak, L.V., J.JReprod. Fertil,
.!2.: 109, 1969.
17. Fawcett, D.W. and Hollenberg, R.D., Z. Zellforsch., 60: 276,
1963.
18. Bedford, J.H., J. Anat., 99: 891, 196.5.
19. Nicander, L. and Hellstrom, B., Exp. Cell Res., 48: 622, 1967.
20. Fawcett, D.W. and Phillips, D.H., J. Reprod. Fertil.,Suppl.
~: 405, 1969.
21. Quinn, P.J. and White, I.G., Aust. J. BioI. Sci., 20: 1205,
1967.
22. Bistocchi, M., d'Alessio, G. and Leone, Eo, J. Reprod. Fertil.,
16: 223, 1968.
23. Crabo, B. and Gustafsson, B., J. Reprod. Fertil.,l: 337, 1964.
24. Wales, R.G., Wallace, J.C. and White, I.G., J. Reprod. Fertil .• ,
12: 139, 1966.
25. Dawson, R.H.C., Hann, T. and White, I.G., Biochem. J., 65:
627, 1957.
26. Marquis, NoR. and Fritz, I.B., J. BioI. Chern., 240: 2197, 1965.
27. Dawson, R.H.C. and Rowlands, I.W., Quart. J. ExP:-Physiol., 44:
26, 1959.
28 0 Heyhofer, W., Arch. Klin. Exp. Derm., 216: 556, 1963.
29. Leuchtenberger, C., J. Dairy Sci., 43: (Suppl.), 31, 1960.

30. Me,yhofer, W., Arch. Klin. Exp. Derm., 219: 925, 1964.
31. Joel, C.A., Fe~til. Steril., ll: 374, 1966.
32. Goldberg, E., Science, 139: 602, 1963.
33. Goldberg, E., Arch. Biochem. Biophys., 109: 134, 1965.
34. Goldberg, E. and Hawtrey, C., J. Exp. Zool., 165: 309, 1967.
35. Zinkham, W.H., Blanco, A. and Kupchyk, L., Pediatrics, 37:
120, 1966.
36. Stanbaugh, R. and Buckley, J., J. BioI. Chern., 242: 4053, 1967.
37. Schatz, L. and Segal, M.L., J. BioI. Chern., 244: 4393, 1969.
38. King, T.E. and Mann, T., Proc. Roy. Soc. Bio~ 151: 226, 1959.
39. Bishop, D.W., Testicular enzymes as fingerprints in the study
of spermatogenesis. In Reproduction and Sexual Behavior
ed. Diamond, M., Indiana University Press, p. 261, 1968.
40. Rothschild, Lord, Proc. Roy. Soc. BioI. 152: 298, 1960.
41. Terner, C., Amer. J. Physiol., 198: 48, 1960.
42. Terner, C., In Spermatozoan Motility. Amer. Ass. Advan. Sci.
Publ. No. 7~89, 1962.
43. Nevo, A.C.-,-J. Reprod. Fertil.,11.: '19, 1966.
44. Murdoch, R.N. and White, I.G., J. Reprod. Fertil.,~: 351,
1968.
45. Eliasson, R., Murdoch, R.N. and White, I.G., Acta Physiol.
Scand., 73: 379, 1968.
46. Peterson, R.N. and Freund, M., J. Reprod. Fertil.,ll: 357,
1968.
47. Peterson, R.N. and Freund, M., BioI. Reprod., 1: 238, 1969.
48. Mann, T., Biochem. J., 48: 386, 1951. -
49. Mann, T., Biochemistry of Semen and of the Male Reproductive
Tract, Methuen., London, 1964.
50. Pike, J.E. and Weeks, J.R., Prostaglandin Bibliography. The
Upjohn Company (Kalamazoo, Michigan, 1969.
51. Bergstrom, S. and Samuelsson, B. (ed), Prostaglandins. Nobel
Symposium 2, held in Stockholm. Interscience Publishers,
New York, 1966.
52. Pickles, V.R., BioI. Rev., 42: 614, 1967.
53. Euler, U.S. and Eliesson, R., Prostaglandins. Academic Press,
New York and London, 1967.
54. Ingelman-Sundberg, A., In Progress in Infertility eds.
Behrman, S.J. and Kistner, R.W. Little, Brown & Co., Boston,
1968.
55. Horton, E.W., Physiol. Rev., 49: 122, 1969.
56. Mann, T., Seamark, R.F. and Sharman, D., Brit. J. Pharmacol.,
11.: 208, 1961.
57. Mann, T., Biochem. J., 40: 481, 1946.
58. Mann, T. and Rottenberg, D.A., J. Endocr 0' 34: 257, 1966.
59. Shulman, S. and Bronson~ P., Fertil. Steril., 12: 549, 1968.
60. Quinlivan, W.L.G., Arch. Biochem., 127: 680, 1968.
61. Herrmann, W.P. and Hermann, G., Arch. Klin. Exp. Derm., 234:
100, 1969. -
62. Quinlivan, W.L.G., The Physiologist, 11, No.3, Aug., 1969.
478 T. MANN
63. Herrmann, W.P., Der Hautarzt, 12: 365, 1968.

64. Marberger, M., Marberger, E., Mann, T. and Lutwak-Mann, C.,
Brit. Med. J., i, 835, 1962.
65. Mann, T., Chemical analysis of semen as a diagnostic aid in
reproductive failure. Proc. World Congr. Fertil. Steril.,
Tel-Aviv, 1969.
66. Mann, T., J.Reprod. Fert. Suppl., ~: 101, 1968.
DISCUSSION
DICZFALUSY: Dr. Mann, is there new information regarding the factors

which initiate sperm motility in the epididymis? This question came
up recently in a recent scientific group of WHO. It was felt that,
if it were possible to interfere with sperm motility specifically,
we would probably have an ideal male contraceptive.
MANN: Unfortunately, we do not know for certain what initiates mo-

tility in the epididymis, apart from oxygen. At one time it was
believed that the migration of the kinoplasmic droplet forms a very
important prerequisite. But now we know that that is not necessarily
so. Even a sperm-flagellum severed from the sperm-head can still be
motile. For example, we have at the Animal Research Station in
Cambridge, a bull which is completely fertile, and ejaculates de-
capitated spermatozoa, that is sperm heads and flagellae separated
from each other. These flagellae consist of midpieces and tails
which although completely decapitated, are perfectly motile. Epi-
didymal plasma per se is also unable to induce motility, but ejacu-
lated seminal plasma can induce motility of immotile but viable sper-
matozoa not only aerobically but also anaerobically, since it con-
tains fructose.
E. STEINBERGER: I seem to recall that Dr. Clermont reported some

years ago motile flagellae in immature spermatids in the testis.
MANN: Yes, indeed one can occasionally observe some motility even
in testicular spermatozoa, but these are very weak oscillatory move-
ments. There is no progressive motility.
COUROT: Dr. Mann, is there something known about the factors involved
in the initiation of the fertilizing ability of spermatozoa?
MANN: We know that the spermatozoa in the caput epididymidis at least

in the rabbit, bull and ram, are still non-fertile; in the cauda epi-
didymidis, however, they become fertile (though not fully capacitated).
Fertilizing ability must be acquired, therefore, during the epididymal
passage, however, not necessarily simultaneously with motility. We
know that under certain conditions, motility and fertilizing ability
can be dissociated from each other. For example, by exposing a sample
of ram semen to a heavy dose of x-rays, so much so that the glass. of
the test-tube acquires a brownish color, one can render the spermato-
zoa completely infertile without disturbing their motility. So, again,
we do not know what renders the spermatozoa fertile in the epididymis.
We do know, however, that during the epididymal passage the spermatozoa
change as regards their ultrastructure, metabolism and composition.
As regards the chemical changes,we know, for instance, that the chem-
ical character of the protein which complexes with nucleic acid changes
during the epididymal passage.
MACLEOD: Dr. Mann, would you agree that the first substrate that
would be utilizable for motility is probably not met until the sper-
matozoa reach the ampulla of the ductus deferens, where clearly there
is production of fructose by the epithelium of that organ?
MANN: Yes, the first chance that the spermatozoa have to acquire mo-
tility anaerobically occurs in the ampullae when they come in contact
with fructose. There is no fructose, however, in the epididymis.
Epididymal spermatozoa depend, therefore, not on fructose, but on
their aerobic endogenous respiration for induction of motility. The
endogenous sperm respiration requires oxygen and it depends mainly
on the oxidation of fatty acids liberated from phospholipids. The
RQ (the oxidation quotient) is 0.7, as one would expect, in the ab-
sence of fructose. In the presence of fructose, it becomes 1.
PAULSEN: Discounting a blockage in the ductile system, where one

observes an absence of fructose, do you recognize any specific or u-
nique biochemical change in the seminal fluid plasma of androgenical-
ly normal but oligospermic males?
MANN: Not really.
MANCINI: Are there any data related to the appearance of sorbitol

dehydrogenase in the cells of the germinal epithelium, and at which
cell stage?
MANN: David Bishop at Baltimore investigated the behavior of sorbi-

tol dehydrogenase not only in ejaculated spermatozoa, as we did, but
in the immature and mature testis. He established convincingly that
sorbitol dehydrogenase appears at the time when the testis becomes
spermatogenically active. The appearance of sorbitol dehydrogenase
is, therefore, dependent on spermatogenesis. But the same applies
to certain other sperm enzymes, for example, lactate dehydrogenase.
Among the isozymes of that enzyme is the so-called LDH IV (thus
named because it migrates in electrophoresis as no. 4). This partic-
ular enzyme also appears at the time of spermatogenesis. In general,
the formation of intracellular sperm enzymes must be regarded as a
process which depends on spermatogenesis.
OHNO: In man, a number of genetic defects,such as Gauchey's disease,

Nieman-Pick's and Marfan's syndrome are now known to be due to the
480 T.MANN
deficient lysosomal enzymes, such as beta-glucosidase and beta gluco-

saminidase. I was wondering if the study of the sperm of these pa-
tients might give us pertinent information as to which acrosomal
enzyme is essential for fertility.
MANN: It is our hope to relate some of the forms of male infertility

in animals to enzyme abnormalities. The question which you raised,
Dr. Ohno, is indeed very important.
THE SIGNIFICANCE OF DEVIATIONS IN HUMAN SPERM MORPHOLOGY
John MacLeod 1
Department of Anatomy, Cornell University Medical College,
New York, N.Y.
If the premise is accepted that the spermatozoa released from

the germinal epithelium do not, in the light microscope at least,
undergo marked changes in head-structure in their subsequent pas-
sage through the duct system to the ejaculate; and if it is further
assumed that, under physiological conditions, the germ cells are
not absorbed in numbers by the ductal epithelium, then the cellular
components of the ejaculate can be accepted, qualitatively and quan-
titatively,as truly representing the output of the germinal epithe-
lium. It is certain that the spermatozoa in all species studied do
undergo further maturation in the epididymis and possibly in the
ductus deferens. The size, shape and internal structure of the
acrosome are modified (1,2) as is the organization of the mitochon-
dria of the middle piece (3). In confirming earlier work that the
capacity for motility increases as the spermatozoa pass through the
epididymis, Blandau (4) and Gaddum (5) have shown in the rabbit
that this potentiation of motility is produced partly by an altera-
tion in the head-tail relationship of the cell. Undoubtedly chemi-
cal changes in the cells occur in the epididymal transit which, in
turn, may induce changes in their metabolic properties and an in-
creasing capacity for the ultimate utilization of substrates.
However in man, and probably in other speCies, it seems rea-

sonably certain that the gross shape and size of the sperm head is
not modified to an appreciable degree between release from the germ-
inal epithelium and the appearance of the cells in the ejaculate.
1Career Scientist, Health Research Council of the City of New York.
481
482 J. MACLEOD
The authors were fortunate in having access to a variety of

populations (medical students, patients, prisoners) in which the
cellular content of serial ejaculates obtained over extended per-
iods could be examined in relation to all aspects of semen quality.
Only sperm morphology will be discussed here.
The validity of the sperm morphology examination is dependent

upon at least two essentials. The first is that the same although
arbitrary criteria of cell classification be· followed throughout.
The second is that a single observer perform all examinations. These
conditions have been followed in all the studies presented below and
have been described in detail elsewhere (6). Figs. 1 and 2 are dia-
grannnatic representations of the dominant types of cells in the hu-
man ejaculate under normal and pathologic conditions. Table 1 shows
a laboratory work sheet and the cell classifications in four hypo-
thetical patients. The system of cell classification used may not
suit all experienced workers in the field but it has proved valuable
in this study. It is believed that the major changes in seminal
cytology found as a result of trauma, illnesses, the use of antisper-
matogenic compounds, etc. occur within these cell classes.
In this study, the "mature" cell is defined arbitrarily as

possessing a head, middle-piece and tail, regardless of aberrations
in any of these structures. When structural defects in head-shape
or size are found and cannot be classed as oval, large, small or
tapering or bicephalic (Table 1) they are put in the amorphous ca-
tegory.
The spermatids and other immature cells form a single class

and the ratio is expressed as a percentage of the total number of
germinal cells in the ejaculate. This percentage is not additiv~
in the mature-cell ratios.
EXPERIMENTAL RESULTS
Under stable environmental conditions the morphology of the

germinal cells in the human ejaculate, in terms of the cell classi-
fication in Fig. 1 and Table 1, is remarkably constant (Table 2).
In men of known fertility and otherwise normal semen quality the
dominant cell is the "oval" form. The mean head-length of this
cell is approximately five microns but it may vary from 2.5 to 10
microns, in which case the cell is classified in the "small" or
"megalo" class. We have considered the "oval" form to be the "nor-
mal" cell in terms of potential fertility although it is quite cer-
tain that pregnancies can occur when other cell types predominate
in the ejaculate. For example, in certain individuals of known
fertility, the "small" form ratio may be as high as 6070. But since
"oval" forms of normal dimensions ~ present in these ejaculates
and since no one yet has been privileged to observe directly the
types of human spermatozoa capable of fertilization, one cannot
5 7
4
Fig. 1. Major types of spermatozoa in human ejaculate.

1) Normal; 2) Small; 3) Megalo; 4) Pyriform; 5) Tapering;
6) Acutely-tapering ("stress" cells); 7) Bicephalic.
Fig. 2. Abnormal spermatozoa and immature cells in human ejaculate

1) Immature form with abnormal chromatin;
2,3,4,5,6,17 Types of amorphous forms; 7, Three spermatid
nuclei in common cytoplasm; 8,11 "spermatid" "ghosts"
-nuclei extruded; 9,10,15 Bizarre tapering and pyriform
cells; 12,13,14,16 Mononuclear spermatids.
484 J, MACLEOD
Table 1
PATIENT
...
OVAL
•
MORPHOLOGY WORK SHEET
LARGE
..
SMALL
" ,
TAPERING AMORI'HOOS
,.
DOUBLE
•
'f
IMMATURE
,
_..
(NCfiMAU HEAD (spermatlds
" obtuonl)
====
----
----
-
A
----
50 9 5 3 2 I
-
----
_<lMf _ _
C
----
20 __ _ Ill'
- ' 60
----
-" ..
20
_ _ ""OJ
lUI
29 29
..
5
---- ..
19 14 4
----
_ _ III
_1111
_ _ II
D
21 1 2 52 9 2 13
A typical work sheet of differential germ cell enumeration on 4

hypothetical cases.
Table 2. Sperm Morphology Readings on Same Individual Over 4 Year

Period.
' Immature
Date Oval Large Small Taper. Amorph. Blceph·(S permatl'd s )
12-27-62 43 3 8 12 17 17 4
1-18-63 38 3 2 18 19 20 16
12-3-63 38 7 2 10 23 20 6
11-4-64 34 4 6 17 19 20 6
*5-19-66 38 3.2 5.85 16.8 7 29 2.5
*1000 cells counted
speculate unduly on fertility in relation to head-shape and size.
The important facts are that each individual has a characteris-

tic pattern of sperm morphology, whether it be perfection or one in
which certain deviations in head-shape are found in constant ratios.
Table 2 demonstrates this point in the semen of a medical student
over a four-year period. In his freshman year, this young man had
emotional problems but he also was found to have a large varicocele.
His sperm morphology featured an unusually high percentage of bi-
cephalic forms although other deviations were present
(tapering, amorphous and immature forms). It will be seen that over
a four-year period (he refused the varicocele ligation) the ratios
of these cell types fell constantly within a narrow range. The ratio
of immature forms varied and at the end of 3~ years the bicephalic
ratio was higher than before although the basic pattern of spermato-
genesis throughout the study was essentially the same.
This stability in sperm morphology can be demonstrated in large

populations. In recent years our laboratory has been fortunate in
its access to large clinical populations (husbands in sterile mar-
riages) and to prison populations participating in studies of anti-
spermatogenic compounds. Medical students also have been studied
over extended periods. Table 3 presents data from over 900 prison-
ers in four penitentaries, from 1500 patients seen in sequence dur-
ing a randomly-selected period and in 146 patients in whom varico-
cele had been found. It will be seen that the sperm morphology of
the prisoners not only is good in terms of the "oval" forms but that
the ratios of the various cell types are very similar in the group
from each penitentiary. It can be said at this time that the pris-
oner morphologies show a remarkable lack of pathology.
In contrast, the patient groups, while quite similar numerical-

ly within themselves in so far as cell types are concerned, show
much fewer "oval" forms. The ratio shift is towards the "large",
tapering and amorphous forms and is accompanied by a nearly 10-fold
increase in the presence of spermatids. The conjunction of the lat-
ter cells with the tapering and amorphous forms we have corne to re-
cognize as the response of the testes to a variety of stresses (7,8,
9). It is found after certain illnesses, (Table 4) allergic reac-
tions (Table 5),and can be produced experimentally by antispermato-
genic compounds (Table 6). In all of these instances too, sperma-
togenesis is severely depressed in the quantitative sense but both
the sperm count and the morphology eventually return to the control
levels after periods ranging from several weeks to months. The de-
viations in the sperm morphology may appear before the count depres-
sion and usually are gone before the count fully recovers.
The same "stress pattern" (tapering and immature forms) is

found to a striking degree in the presence of varicocele. Indeed,
it is diagnostic of varicocele if the individual does not give a
clinical history which could otherwise account for the seminal pic-
ture. When this pattern is found in varicocele it persists in con-
stant ratios until the varicocele is removed and may continue at a
diminished level of pathology thereafter. It has also been found
that bicephalia is common in the presence of varicocele. Table 7
illustrates a case of bilateral varicocele in which the semen qual-
ity was followed before and after the bilateral ligation. This
husband was close to the sterile level. His seminal cytology showed
the classic "stress" picture (tapering, amorphous and immature) plus
a high percentage of bicephalic forms. Within six weeks after op-
eration, a marked improvement in sperm motility was found although
the sperm count remained low. In addition, the seminal cytology
showed a marked drop in the ratios of bicephalic and immature forms.
Five weeks later or about 11 weeks post-operatively, the sperm count
had risen sharply to 62 million/ml, the motility remained at a good
.....
00
0-
Table 3
SEMINAL CYTOLOGIES OF PRISONERS AND PATIENTS
Prisoners
Location & Oval Large Small Taper Amorph Bicephalic Spermatids %

Number % % % % /0 % of Total Germinal
Oklahoma 625 73.0 3.0 9.0 6.0 8.0 1.0 0.30

San Quentin 125 73.0 3.0 9.0 6.0 8.0 1.0 0.10
Atmore 103 73.0 2.0 8.5 5.5 10.0 1.0 0.40
Oregon 91 73.0 2.6 8.0 7.0 8.4 1.0 0.80
Patients in
Sequence as seen
1-11-61 to 5-16-62
1st 500 56.0 7.0 11.0 12.0 12.5 1.5 4.00

2nd 500 58.6 3.1 9.S 13.6 13.7 1.5 4.60
3rd 500 57.0 5.7 10.3 12.8 12.4 1.8 3.50
Varicocele
!-
146 34.0 5.0 10.0 23.0 25.0 3.0 16.00 ~
»-
()
Patterns of sperm morphology in prison, infertile marriage and varicocele populations. ....
m
0
0
Table 4
EFFECT OF VIRAL INFECTION ON CYTOLOGY OF THE EJACULATE
Date Oval Large Small Taper Amorph Double Inunature

1961 % % % % % Head % %
11-12 17 6 25 46 3 47
11-21 42 10 38 8 2 3
11-28 22 9 1 53 15 0 6
12-6 32 3 60 5 0 2
12-11 51 14 1 29 5 0 1.5
12-13 40 18 2 34 6 0 2.0
12-23 77 4 3 9 4 1 0.3
12-30 S2 6 2 31 8 1 2.5
1-8-62 64 4 3 27 1 1 0
1-13 S6 5 0 37 2 0 0
2-8 77 12 0 8 2 1 0
Effect of viral infection on sperm morphology following acute stage

and during recovery.
level and the bicephalic forms remained relatively low. The "stress"
picture, particularly the exfoliation of spermatids, virtually had
disappeared. A conception occurred about this date and a normal
pregnancy ensued. This case offers an excellent example not only of
the initiation of fertility but of a virtually complete reversal of
testicular pathology.
DISCUSSION
We believe these data show that the human testes are extremely
sensitive to changes in their environment and that this sensitivity
can be measured in the cellular components of the ejaculate with a
considerable degree of precision. That the response of the testes
to a variety of trauma is non-specific probably means that a parti-
cular level of spermatogenesis is vulnerable. Since the "tap"ering"
forms and the spermatids in the ejaculate are prominent cells in the
"stress" reaction, a strong argument can be made that the early and
middle stages of spermatid development are peculiarly susceptible.
The premature exfoliation of spermatids seems to occur predominantly
between the Sbl and Sb2 levels. The "tapering" form appears to be
a cell which has undergone a marked over-elongation of the nucleus
at and beyond the Sb2 level where the normal elongation process oc-
curs. Harvey (11), however, believes that the "tapering" (spindle)
~
0:1
0:1
Table 5
SPERMIOGRAM - BEFORE, DURING AND AFTER SEVERE ALLERGIC REACTIONS
Vol. Count Total Motility Oval Large Small Tapering Amorph. Biceph. Irrnnature
Days
(ml) ml Count Index 10 10 10 10 10 10 10
(millions)
Control
(12 spec- 4.20 360 1512 303 87 2 2 4 4 1 0.7
imens)
After
start of
3.0 115 345 0 47 7 3 7 35 1 17
reaction
35 days
49 3.2 6 19 0 50 12 10 11 15 2 31
77 4.0 6 24 3 31 20 3 32 12 2 30
90 2.6 7 18 150 53 3 6 12 23 3 20
106 2.0 175 350 388 68 3 9 7 9 4 3.0
126 3.2 309 989 270 82 1 2 7 6 2 0 :-

~
>
()
r-
m
0
0
<II
;>II
"m
Table 6 ~
;>II
"
6-MEDROXY-PROGESTERONE ACETATE 0
0
c
()
AVERAGE SEMEN QUALITY OF FOUR MEN BEFORE AND AFTER ::j
0
z
INJECTION OF 1000 MGMS IN A SINGLE SITE >
z
0
~
Total ;>II
Days Sperm z>

<II
After Vol. Count Mot. 70 70 70 % 70 70 70
Injection (ml) (Millions) Index Oval Large Small Taper Amorph Dup Sp.
"0
;>II
~
2.20 1200 212 82 3 5 4 6 1

3.35 1029 104 71 4 5 7 13 1 2
Controls 2.55 669 320 84 3 2 2 8 1 0
2.35 390 309 81 1 11 4 3 0 0
2.45 619 230 76 4 9 3 7 1 0
7 2.80 391 220 81 2 9 2 6 0 0.5

28 2.10 229 248 54 3 6 9 17 1 2
49 2.10 51 49 14 2 20 23 40 1 8
70 2.10 5 2 12 4 16 20 47 1 5
80 1.72 4 0 20 6 13 33 26 2 12
98 2.30 10 66 29 4 16 22 28 1 14
The effect on sperm morphology of antispermatogenic agent (6-medroxy progesterone acetate).
.l>-
e»
'0
8
Table 7
SEMEN QUALITY BEFORE AND AFTER VARICOCELE LIGATION
Date Vol/ml Count Total Mot. Oval Large Small Taper Amorph Biceph Sperm-
10 6 / ml Count Index atids
(millions)
11-12-65 3.0 1.0 3 5 10 4 32 26 28 57
12-9-65 2.0 3 6 20 24 1 25 25 25 32
12-21-65 L I GAT ION PER FOR M E D
2-4-66 2.8 4 11 225 45 9 23 17 6 10
>~3-10-66 3.0 62 186 225 38 2 10 12 30 8 1
ok
conception occurred about this date
!-
~....
oo
form is a result of abnormal behaviour of the chromosomes at both

the first and the second maturation divisions. In this regard, one
cannot preclude the possibility that the predilection towards the
"tapering" form is laid down at a pre-spermatid level. But the pre-
mature exfoliation of spermatids from the Sertoli cell which so often
appears in conjunction with the "tapering" form is more likely to be
a pathologic reaction in spermiogenesis and may perhaps be the re-
sult of a traumatic assault directly upon the Sertoli cell.
The spermatids which appear in the ejaculate are usually mono-

nuclear but often are found in clusters sharing a cOmmon cytoplasm
(Fig. 2). These concretions of spermatid nuclei do not appear to be
related to the presence of intercellular bridges at the spermatid
level described by Fawcett et ale (12) in which groups of spermatids
are joined one to the other in protoplasmic continuity. It seems
more probable that the normal syncytial process described by these
authors either has not occurred or has broken down. The difficulty
still lies in explaining how the spermatid nuclei are lost from the
parent cells and find themselves in a common cytoplasm apparently as
discrete entities.
A further point of interest is that in the mononuclear and in

the multi-nucleate cells, the nuclei are usually located peripherally
near or contiguous to the cell border (Fig.2), a position they would
occupy in normal spermiogenesis just prior to the first stage of
elongation of the nucleus. Indeed, they often appear to be in the
process of extrusion. In examining these cells in the seminal plas-
ma, this peripheral positioning of the nuclei is obvious, and in
certain cases, we actually have seen obvious bulging of the cell sur-
face and extrusion as if the nuclei were being forced outwards by
internal pressure. It is surprising not only that these cells sur-
vive the long passage through the duct system (at least 15 days from
testes to the ejaculate) without undue degeneration but that these
nuclear positions are retained in the ejaculated cells as they were
when they first were exfoliated into the lumens of the seminiferous
tubules.
In conclusion, it should be emphasized that while severe path-

ology in the germinal epithelium and the cellular contents of the
ejaculate can be produced experimentally, by illnesses, etc. hese
changes are fully reversible. And in this regeneration and restora-
tion, the pattern of spermatogenesis, as seen in the ejaculate, prior
to the trauma, is retained. In other words, each individual has an
inherent pattern of spermatogenesis which possibly may be determined
at the level of the spermatogonium.
ACKNOWLEDGEMENTS
The research was supported by Grant HD-00481 from the National

Institutes of Health.
492 J. MACLEOD
REFERENCES
1. Fawcett, D.W. and Hollenberg, R.D., Z. Zellforsch, 60: 276,

1963.
2. Bedford, J .M., J. Reprod. Fertil. ,2: 169, 1963.
3. Anberg, A., Acta Obstet. Gynec. Scand.~: Supple 2 1, 1957.
4. Blandau, R.J. and Rumery, R.E., Fertil. Steril.,15: 571, 1964.
5. Gaddum, P., Anat. Rec., 161: 471, 1968. --
6. MacLeod, J., Fertil. Steril., 13: 29, 1962.
7. MacLeod, J., Fertil. Sterile, 20: 545, 1969.
8. MacLeod, J., Fertil. Steril., Q: 531, 1962.
9. MacLeod, J., J. and A. Churchill, Ltd. London p. 93, 1965.
10. Heller, C.G. and Clermont, Y., Recent Progr. Hormone Res., 20:
545, 1964.
11. Harvey, C. Studies on Fertility ed. R.G. Harrison, Blackwell,
Oxford, p. 8, 1955.
12. Fawcett, D.W. Susumma Ito Slatterback, D., J. Biophys. Biochem.
Cytol., 2: 453, 1959.
DISCUSSION
LUNENFELD: We have recently re-investigated Dr. Amelare's findings

on the split ejaculate in oligospermic males. In most of the 80 pa-
tients which we investigated in the last seven months, the first part
of the split ejaculate (0.5 ml. of a 2 ml. ejaculate) was of a sig-
nificantly better quality than the second part. Could you elaborate
on the reasons for these findings?
MACLEOD: The first part of the split ejaculate is richer than the
second by nature of the ejaculatory process. It would seem that the
ampulla of the ductus deferens is the first to contract and empties
into the duct system the spermatozoa which are at that point at the
time of ejaculationo I have not observed any real differences in
morphology in these two systems; but I have observed that the motil-
ity in the first portion is usually much better than in the second.
On the other hand, the second part of the ejaculate is usually bad
in cellular content.
MANN: Dr. Lunenfeld may be interested in some observations which we

made on thoroughbred stallions. Ejaculation in the stallion is a
prolonged process, up to half-a-minute, so that it is relatjvely easy
to collect by means of an artificial vagina, several separate frac-
tions of the ejaculate. Normally, when one examines the split ejac-
ulate one finds that the pre-sperm, sperm-rich and post-sperm frac-
tion follow each other in an orderly fashion. However, under certain
conditions, e.g. under the influence of drugs, the sequence of ejac-
ulatory events is severely disturbed, and consequently the seminal
fractions do not appear during ejaculation in the normal order.
HELLER: I would like to comment on the multinucleated cells that

Dr. MacLeod demonstrated in the ejaculate as well as in the seminif-

erous epithelium. The explanation for finding groups of cells to-
gether was offered some years ago by Dr. Don Fawcett. He found in-
tercellular bridges connecting groups of spermatids. In our labora-
tory Mavis Rowley has recently demonstrated intercellular bridges
between human pachytene spermatocytes. This phenomenon occurs early
in the lifetime of spermatogenesis. Type A dark spermatogonia which
can be identified because of the presence of vacuoles lying next to
the basal membrane are clearly connected by a large intercellular
bridge. I think that each of us is struck with the great similarity
of cells during the process of spermatogenesis, in other words, we
see whole families of cells that look exactly alike 'in a given cell
association. When we look at another cell association and see the
same cell, however, it appears slightly different from those in the
first cell association but these cells again resemble each other ex-
actly like members of a family.' I wish to raise the question if this
represents not a rare phenomenon but a usual phenomenon. In other
words as these cells multiply they stay together as a family connec-
ted together by intercellular bridges. Now, assuming any and all of
this, I want to ask Dr. MacLeod if during spermiation he thinks that
maybe something goes wrong.
MACLEOD: The observation of intercellular bridges does not complete-

ly explain the phenomenon of multinucleated cells in the ejaculate.
I agree with Dr. Heller that when one sees these multinucleated cells
in the ejaculate, the nuclei are already separated and are located
at distinctly separate points in the cytoplasm. The presence of more
immature cells with multinucleation may be related to the phenomenon
of the intercellular bridges. Similarly, the bicephalic mature sper-
matozoa are probably a manifestation of disruption late in spermio-
genesis. This disruption may involve the bridges. At the present
time, I would consider that the late stages in spermatogenesis are
most vulnerable to "trauma".
E. STEINBERGER: Dr. MacLeod, do you see patients with oligospermia

who do not show histologic abnormalities in the spermatozoa?
MACLEOD: Yes, I am very disappointed indeed when I do not find it.

These are the cases which we have now relegated to the category of
idiopathic infertility. Since our understanding has improved with
respect to the influence of varicocele on spermatogenesis, only a
relatively small group of oligospermic males are in the idiopathic
category. These cases may exhibit abnormal sperm cytology but not
the distinct shift to the right. We associate this with a "stress-
ful" situation.
TROEN: Dr. MacLeod, do the tapering forms which you refer to as

stress fprms have normal motility?
MACLEOD: Yes, the tapering form may show very good motility and may
494 J. MACLEOD
migrate into the cervical mucus. Indeed, in performing the post-

coital examination of the cervical mucus no other abnormal forms are
seen aside from the tapering forms.
PHARMACOLOGICAL STUDIES ON THE TESTICULAR CAPSULE IN RELATION TO
SPERM TRANSPORT
Joseph R. Davis and George A. Langford 1
Department of Pharmacology and Therapeutics,

Loyola University Stritch School of Medicine,
Maywood, Illinois
The testicular capsule, which forms the thin outer covering

of the testis, has long been considered an inert tissue with no
other function than to contain the underlying seminiferous tubular
mass. In the course of carrying out in vitro metabolic studies on
rat testicular tissue in our laboratory, we have continually peeled
off and discarded the capsule of the testis prior to slicing the
testis with a Stadie-Riggs microtome. It recently occurred to us,
however, that it might be quite easy to prepare and save the cap-
sule of the rat testis as an intact tissue provided one first re-
moved the seminiferous tubules from the interior of the organ (1).
Procedure for the Isolation of the Testicular Capsule.
The testicular capsule has been commonly called the "tunica

albuginea", even though the tunica albuginea is but one of three
layers comprising the complete capsular membrane. The three layers
of the testicular capsule include the tunica vaginalis visceral,
which is an outer thin serous layer, the tunica albuginea which
forms the substance of virtually the entire capsule and the tunica
vasculosa which is a thin delicate layer of loose areolar tissue
directly beneath the tunica albuginea. Fig. 1 presents a schematic
representation of the procedure that we devised for the isolation
of the adult rat intact testicular capsule and its subsequent use
as an isolated tissue preparation suitable for pharmacological stu-
dies (2). Following sacrifice of the animal, a single testis is
removed and placed in Tyrode1s solution. A small piece of the in-
ferior end of the testis is cut away. The testicular parenchymal
1predoctoral Trainee, U.S. Public Health Service.
495
496 J. R. DAVIS AND G. A . LANGFORD
A B c D
E F G H
Fig. 1. Schematic representation of the procedure for the isola-
tion of the testicular capsule of the adult rat and its use as an
isolated tissue proparation for pharmacological studies. A. Cut
a small piece off the inferior end of the testis. B. Grasp both
the testicular parenchymal mass protruding through the resulting
opening as well as the lower rim of the testicular capsule with
forceps and gently pull apart. C. After removing most of the par-
enchymal tissue mass, turn the testicular capsule inside-out. D.
Diagram of the testicular capsule in the inside-out position with
only a few seminiferous tubules still attached to its inner surface.
E. Cut away the few remaining seminiferous tubules from the testi-
cular capsule, along with the large testicular artery. F. Photo-
graph of the isolated testicular capsule in the inside-out position,
illustrating the complete removal of all the interior tissue of the
testis. The removed testicular artery is shown on the left. G.
Tie the superior end of the testicular capsule with a long piece of
silk thread and place a short tie at the inferior end of the cap-
sule, leaving a small portion of the original hole in the capsule
open. H. Mount the testicular capsule in a 10 ml isolated tissue
bath with the short inferior thread attached to a s~pport rod and
the long superior thread leading to a transducer for detection of
tissue contractions.
mass protruding through this small hole in the testis is then

grasped with a forceps and removed from the interior of the testis.
The lower rim of the testicular capsule is then grasped with a sec-
ond forceps and the capsule is turned inside-out. The testicular
capsule is then separated from the testicular parenchymal tissue as
well as from the large testicll1ar artery and the few remaining se-
miniferous tubules still attached to the inside surface of the cap-
sule are carefully cut away. The intact isolated testicular cap-
sule is then tied at its superior end with a long piece of silk
thread. A second smaller tie is placed at the inferior end of the
testicular capsule with care taken to leave approximately one-fourth
of the original small hole in the capsule open. The intact isolated
testicular capsule is then mounted in a 10 ml standard isolated tis-
sue bath assembly with the upper long thread leading to a linear
motion transducer connected to a suitable recorder.
The resulting isolated testicular capsule of the adult rat was

found to resemble a thin hollow sac measuring approximately 2 cm in
length and 1 cm in width with an average wet weight of 75 mg. The
corresponding dimensions for the isolated testicular capsule of the
adult rabbit were quite similar to those of the rat. However, the
weight of the rabbit isolated testicular capsule was found to be
almost twice that of the rat, averaging 125 mg.
Fig. 2 presents a photograph of the testicular capsule mounted

within the complete isolated tissue assembly employed in our labor-
atory. The physiological salt solution in the isolated tissue bath
was Tyrode's solution (3) warmed to a constant temperature and gas-
sed with air. The paper speed of the recorder was 0.25 mm per sec-
ond. Each drug was dissolved in Tyrode's solution and added to the
isolated tissue bath in a volume of 0.1 ml by means of a suitable
pipet. Any contraction or relaxation of the isolated testicular
capsule then appears as a deflection of the pen on the recorder
chart paper.
Pharmacological Studies on the Isolated Testicular Capsule

of the Rat.
Now that the testicular capsule of the rat had been obtained
as an intact isolated tissue for the first time, it seemed of in-
terest to investigate the possible effects of pharmacological agents
on this preparation. It also seemed of interest to compare the pos-
sible effects of pharmacological agents on the testicular parenchy-
mal mass which had been obtained in an isolated form during the pre-
viously mentioned procedure for the isolation of the testicular cap-
sule. It was therefore a simple matter to attach a superior and an
inferior thread to the ends of the isolated testicular parenchymal
mass and to mount the entire isolated testicular parenchyma in an
organ bath in a manner similar to that just described for the iso-
lated testicular capsule.
498 J. R. DAVIS AND G. A. LANGFORD
Fig. 2. Photograph of the isolated tissue assembly employed in the

authors' laboratory for measurement of contractions of the
isolated testicular capsule. The testicular capsule moun-
ted within the isolated tissue bath is indicated by the
arrow. A linear motion transducer above the tissue bath
is connected to a recorder shown at the left. A constant-
temperature water bath, shown at the right, serves to main-
tain the desired temperature of the isolated tissue bath.
Fi~ 3 presents the response of the isolated testicular capsule

of the adult rat to various autonomic drugs. Both acetylcholine
and norepinephrine were observed to produce a marked contraction of
the isolated testicular capsule. In each case, the contractions
produced by the neurohumoral agent were found to be dose dependent,
with maximal responses occurring at a final organ bath concentra-
tion of 1 ~/ml. Maximal contraction of the isolated testicular
capsule to both acetylcholine and norepinephrine was reached approx-
imately three min~tes following addition of the drug. The testicu-
lar capsule was found to relax slowly after it had been caused to
contract so that a slight extra weight was added to the initial
resting tension load of 100 mg in order to assist its recovery.
This procedure was found to conveniently allow doses to be added
about once every 15 minutes.
Two additional parasympathomimetic agents, carbachol and pilo-

carpine, were compared with acetylcholine as to their effect on the
isolated testicular capsule of the rat. Carbachol, an example of a
ISOLATED
CAPSULE
(,100)
ISOLATED
PARENCHYMA
(,1 00)
ISOLATED
CAPSULE
(, 100)
ISOLATED
PARENCHYMA
(.100)
ISOLATED
CAPSUlE
(.100)
ISOlATED
PARENCHYMA
(.100)
Fig. 3. Comparison of the response of the isolated capsule versus

the isolated parenchymal tissue of the adult rat testis to
various autonomic drugs. Two large squares shown on the
chart paper measure 10 rnrn. The response magnification was
x 100 for both the isolated testicular capsule and isolated
testicular parenchyma. The vertical bar represents the
actual rnrn of tissue contraction for each preparation; i.e.,
for both the isolated testicular capsule and isolated tes-
ticular parenchyma, two large squares on the , chart paper
are equal to an actual tissue contraction of 0.1 rnrn. In
each case, the load on the lever arm was 100 mg. Drug con··
centrations are expressed per ml of organ bath volume. The
temperature of the isolated tissue bath was maintained at
32 o C.
synthetic cholinomimetic agent, resembled acetylcholine in producing

a marked contraction of the isolated testicular capsule. Pilocar-
pine, an example of a naturally-occurring cholinomimetic agent, was
also found to cause a contraction of the isolated testicular cap-
sule but the response to pilocarpine appeared much less sensitive
as compared to acetylcholine. The effects of two additional sympa-
thomimetic agents, namely epinephrine and isoproterenol, were com-
pared with norepinephrine as to their effect on the isolated testi-
cular capsule of the rat. Epinephrine, on the one hand, was ob-
served to cause a contraction of the testicular capsule while iso-
proterenol, on the other hand, was found to cause a pronounced re-
laxation of the testicular capsule.
Tetramethylammonium (TMA), which is a ganglionic stimulating

agent, was found to produce a contraction of the isolated testicu-
lar capsule of the rat. This observation would seem to suggest,
for the first time, the possible presence of parasympathetic gan-
glia located in the testicular capsule. Histamine was also found
to cause a marked contraction of the isolated testicular capsule of
the rat. However, barium chloride was observed to produce just a
slight contraction of the isolated testicular capsule of the rat
and only at the extremely high final concentration of 1 mg/ml of
organ bath volume.
Fig. 3 also presents a comparison of the response of the iso-

lated testicular parenchymal tissue of the adult rat to the same
nine autonomic drugs just mentioned for the isolated testicular cap-
sule. The isolated parench)~al mass of the adult rat testis had a
wet weight which averaged 1,450 mg as compared to only 75 mg for the
isolated testicular capsule. The load on the lever arm was 100 mg
for both the testicular capsule and the testicular parenchyma. Of
the nine autonomic drugs studied, only norepinephrine and epine-
phrine were observed to have any effect on the isolated testicular
parenchyma and these two agents caused only a negligible degree of
contraction. For example, whereas the addition of 1 ~/ml final
concentration of norepinephrine to the isolated testicular capsule
produced a tissue contraction of 0.286 mm/lOO mg, the same addition
of norepinephrine to the isolated testicular parenchyma produced a
tissue contraction of only 0.001 mm/lOO mg. Whereas 1 ~g/ml epine-
phrine produced a tissue contraction of the isolated testicular cap-
sule of 0.267 mm/lOO mg, the same addition of epinephrine to the
isolated testicular parenchyma produced a tissue contraction of only
0.002 mm/lOO mg.
Pharmacological Studies on the Isolated Testicular Capsule

of the Rabbit.
The preparation of the isolated testicular capsule of the adult

rabbit is similar to that shown for the rat in Fig. 1 with the ex-
ception that it is somewhat more laborious to remove the internal
parenchymal tissue of the rabbit testis because of the numerous thin
septa extending from the capsule into the testicular parenchyma.
In contrast to the adult rat testicular capsule, the isolated testi-
cular capsule of the adult rabbit was observed to undergo marked
spontaneous contractions within one hour of mounting in the organ
bath and before the addition of any drug. However, it should be
kept in mind that while the demonstration of spontaneous contrac-
tions employing the rabbit testicular capsule has important physio-
logical implications, the absence of demonstrable spontaneous con-
tractions employing the rat testicular capsule has distinct pharma-
cological advantages involving assay of drug effects. Indeed, bio-
assay procedures in pharmacology are often designed to eliminate
spontaneous contractions of an isolated tissue in order to determine
more accurately the extent of drug-induced contractions. It would

seem, therefore, that both the rat and rabbit isolated testicular
capsule preparations have their own unique advantages for p~armaco
logical investigations. The amplitude of the spontaneous contrac-
tions of the rabbit isolated testicular capsule was found to aver-
age a 5 percent shortening of the actual entire length of the
mounted capsule and indeed, it was even possible to observe these
rhythmic movements of the capsule with the unaided eye. The fre-
quency of the spontaneous contractions of the isolated testicular
capsule of the adult rabbit ranged from 3 to 5 per minute.
Fig. 4 presents typical spontaneous contractions recorded from

the rabbit isolated testicular capsule as well as from the whole
rabbit testis without removal of the capsule. In marked contrast
to the periodic spontaneous contractions recorded from both the
isolated testicular capsule and the whole testis, no spontaneous
contractions of the isolated testicular parenchymal tissue were ob-
served, even at a magnification response of x 100. It would there-
fore appear that the testicular capsule alone is responsible for the
endogenous rhythmic spontaneous contractions observed with the whole
rabbit testis. In addition, the fact that the whole testis with its
contained seminiferous tubules and interstitial tissue offering a
mass of tissue resistance was observed to undergo spontaneous con-
tractions would seem to lend great physiological importance to the
contractions described for the isolated capsule alone.
Fig. 5 presents a comparison of the response of the isolated

capsule versus the isolated parenchymal tissue of the adult rabbit
testis to the same autonomic drugs employed for the adult rat tes-
tis. Like the isolated testicular capsule of the rat, the rabbit
isolated testicular capsule was found to undergo a contraction fol-
lowing the addition of acetylcholine and norepinephrine. A final
concentration of 1 !!g/ml of acetylcholine in the organ bath re-
sulted in a contraction of the rabbit isolated testicular capsule
which amounted to approximately a 10 percent shortening of the
actual entire length of the mounted capsule. A final concentration
of 1 !!g/ml of norepinephrine in the organ bath produced a 20 percent
shortening of the actual entire length of the mounted capsule.
The rabbit isolated testicular capsule was also found to re-

semble the isolated testicular tapsule of the rat in that contrac-
tions were produced by the addition of carbachol, epinephrine, tet-
ramethylammonium (TMA) and histamine. The rabbit isolated testi-
cular capsule was found to be unusually sensitive to histamine in
that a final concentration of only 5 ~g/ml of histamine resulted in
a marked contraction. In addition, isoproterenol also produced a
prolonged relaxation of the isolated testicular capsule of the rab-
bit, as well as completely abolishing capsular spontaneous contrac-
tions. However, two differences were noted between the isolated
testicular capsule of the rat and rabbit in that pilocarpine, in a
502 J. R. DAVIS AND G. A . LANGFORD
WHOLE
TESTIS
(xIO)
ISOLATED
CAPSULE
(x25)
ISOLATED
PARENCHYMA
(X 100)
Fig. 4. Spontaneous contractions of the adult rabbit testis in the

absence of any added drugs. Two large squares shown on
the chart paper measure 10 rom. The top recording was ob-
tained from the whole testis weighing 1,700 mg at a re-
sponse magnification of x 10. The middle recording was
obtained from the isolated testicular capsule weighing 135
mg at a response magnification of x 25. The bottom record-
ing was obtained from the isolated testicular parenchyma
weighing 1,550 mg at a response magnification of x 100.
In every case, the load on the lever arm was 100 mg. The
37 o C.
comparable dose, did not elicit any response while barium chloride,
in a comparable dose, produced a very large contraction of the rab-
bit testicular capsule.
In a similar fashion to that observed in the rat, the isolated

testicular parenchymal tissue of the adult rabbit displayed only
extremely small and quite negligible contractions with several of
the autonomic agents studied, even at a response magnification of
x 100. For example, the addition of 1 ~/ml final concentration of
both acetylcholine and norepinephrine to the isolated testicular
capsule of the rabbit resulted in tissue contractions of 0.975 and
1.185 rom/l00 mg respectively. However, similar additions of both
acetylcholine and norepinephrine to the isolated testicular paren-
chyma of the rabbit resulted in tissue contractions which averaged
ISOLATED
CAPSULE
(1.10)
ISOLATED
PARENCHYMA
(a100)
ISOL.ATED
CAPSULE
hl IO)
ISOLATED
.•
PARENCHYMA
bIOO)
," '''''
.......-
ISOLATED
CAPSULE
hlO)
ISOLATED
PARENCHYMA
hIOO)
Fig. 5. Comparison of the response of the isolated capsule versus

the isolated parenchymal tissue of the adult rabbit testis
to various autonomic drugs. Two large squares shown on
the chart paper measure 10 mm. The response magnification
was x 10 for the isolated testicular capsule and x 100 for
the isolated testicular parenchyma. The vertical bar re-
presents the actual mm of tissue contraction for each pre-
paration; i.e., for the isolated testicular capsule two
large squares on the chart paper is equal to an actual
tissue contraction of 1.0 mm while for the isolated testi-
cular parenchyma two large s·quares on the chart paper are
equal to an actual tissue contraction of only 0.1 mm. In
each case, the load on the lever arm was 100 mg. Drug con-
centrations are expressed per ml of organ bath volume. The
37 o C.
not more than 0.001 mm/100 mg.
Pharmacological Studies of the Isolated Testicular Capsule

of the Human
The question then arose in our mind as to whether the present-

ly-described pharmacological responses of the isolated testicular
capsule of the rat and rabbit would also occur in the human testi-
cular capsule. Segments of normal testes from males aged 60-70
years which had been removed as a therapeutic measure for prostatic

carcinoma were brought to our laboratory within 15 minutes of re-
moval of the testis at surgery. Inasmuch as the entire isolated
testicular capsule of the human was much too large to be mounted in
our organ baths, strips of the testicular capsule were cut and uti-
lized for pharmacological investigations. Each strip of the iso-
lated testicular capsule of the human was cut to an approximate
measurement of 20 mm in length and 6 mm in width, with an average
wet weight of 400 mg. The strip was then tied with silk thread at
both ends and mounted in a 10 ml isolated tissue bath with the up-
per long thread leading to a transducer connected to a recorder.
Fig. 6 presents the results of our initial pharmacological

studies on the isolated testicular capsule of the human. The human
testicular capsule was observed to undergo very marked spontaneous
contractions which reached an actual tissue contraction of 1.381
mm/l00 mg. These extremely strong spontaneous contractions repre-
sented a remarkable 39 percent shortening of the actual entire
length of the mounted strip of the testicular capsule of the human.
The average frequency of these spontaneous contractions of the hu-
man testicular capsule was such that one extremely powerful con-
traction occurred every 13.6 minutes. It is interesting to note
that in most instances a large spontaneous contraction was followed
by a slightly smaller spontaneous contraction and that the larger
contraction was usually biphasic in nature. One might well specu-
late that spontaneous contractions of the human testicular capsule
may occur even more frequently in males who are younger than the
ones studied in the present experiments, especially in view of the
finding of Mancini et ale (4,5) that the tunica albuginea undergoes
hyalinization with increasing age.
Fig. 7 presents the effects of various autonomic drugs on the

isolated testicular capsule of the human. Each drug was added to
the organ bath shortly after the recording of a spontaneous con-
traction had returned to baseline so as to be sure that the response
was due to the added drug alone. Acetylcholine, carbachol, nore-
pinephrine, tetramethylammonium (TMA) and barium chloride were all
observed to induce a contraction of the isolated testicular capsule
of the human. The human testicular capsule seemed extremely sensi-
tive to norepinephrine and indeed, norepinephrine was found to in-
duce spontaneous contractions of the testicular capsule at the peak
of its effect, which was within two minutes of addition to the or-
gan bath. Whereas norepinephrine was found to produce a shortening
of the actual entire length of the mounted capsule of the rat and
rabbit testis which amounted to approximately 2 and 10 percent, re-
spectively, norepinephrine produced a 20 percent shortening of the
actual entire length of the mounted testicular capsule of the human.
;- 1 min
_-, mt
15 min 14 min
A B
12 min 12 min
C o
+ -f
j .... ,eeJ:'.
, + cont'ktlon
2.0 """
12 min 13m!n
E F
12m!n 14 min
G H
13 min
t
(X 5)
Fig. 6. Spontaneous contractions of the isolated testicular cap-

sule of the human in the absence of any added drugs. Two
large squares shown on the chart paper measure 10 rom. The
response magnification was x 5. The vertical bar repre-
sents the actual rom of tissue contraction; i.e., two large
squares on the chart paper equals an actual tissue contrac-
tion of 2.0 m. A through J represent consecutive spontan-
eous contractions observed throughout a time period of 134
minutes. The time beneath each spontaneous contraction
represents minutes after the preceding contraction. The
32 oC. The preparation employed was a longitudinal strip
of the isolated testicular capsule 18 rom in length and 7
rom in width which was obtained from the superior half of
the posterior border of the human testis. The load on the
lever arm was 250 mg.
'''''.
'.*~
~-
~
~
I Jlo'mi
ACETYLCHOLINE
I JlI'ml
CARBACHOL
•
100 JlI'ml
TNA
I JlI"ol
NOREPINEPHRINE
.:]
tr:ll ":ft::
j :~
rt,
t ·rl-;-:+·I'
Fig. 7. Effect of autonomic drugs on the isolated testicular cap-

sule of the human. Two large squares shown on the chart
paper measure 10 mm. The response magnification was x 10.
The vertical bar represents the actual mm of tissue con-
traction; i.e., two large squares on the chart paper equals
an actual tissue contractions of 1.0 mm. Drug concentra-
tions are· expressed per ml of organ bath volume. The tem-
perature of the isolated tissue bath was maintained at
32 oC. The preparation employed was a circular strip of
the isolated testicular capsule measuring 20 mm in length
and 6 mm in width which was obtained from the middle third
of the medial surface of the human testis. The load on the
lever arm was 250 mg.
Histological Studies of the Testicular Capsule of the Rat,

Rabbit and Human
Because of the presently-observed pharmacological responses of

the isolated testicular capsule of the rat, rabbit and human, it be-
came of particular interest to attempt to ascertain the presence of
smooth muscle in the testicular capsule of these three species.
Fig. 8 presents typical sections of the testicular capsule obtained
from an adult rat, an adult rabbit and a 65-year-old human. By em-
ploying Masson's trichrome stain, it was possible to identify occas-
ional smooth muscle fibers as well as smooth muscle nuclei which
were located within the collagenous tissue of the rat testicular
capsule. In contrast to the sparse amount of smooth muscle present
in the testicular capsule of the rat, two distinct layers of smooth
RAT RABBIT HUMAN

cap cap
low
magnification
high
magnification
Fig. 8. Representative sections of the testicular capsules of the

rat, rabbit and human stained with Masson's trichrome stain
illustrating the comparative thickness of each capsule as
well as the presence of smooth muscle. A-C. Low magnifi-
cation of the testicular capsule (cap) surrounding the
parenchymal tissue of the rat, rabbit and human testis, re-
spectively (x 22). D. High magnification of the rat testi-
cular capsule demonstrating occasional smooth muscle nuclei
(smn) within the collagenous tissue (c) of the tunica al-
buginea. fib, fibroblast; int, interstitial tissue cell
(x 1125). E. High magnification of the rabbit testicular
capsule demonstrating the superficial layer of longitudinal
smooth muscle (lsm) as well as the deeper layer of circular
smooth muscle (csm). st, seminiferous tubule (x 490). F.
High magnification of the human testicular capsule demon-
strating the abundance of smooth muscle fibers (smf) loca-
ted within the dense collagenous tissue of the tunica al-
buginea (x 490).
muscle fibers were found in the tunica albuginea of both the rabbit
and the human testicular capsule. A superficial layer of longitudi-
nal smooth muscle appears to run parallel to the long axis of the
testis while a second, deeper layer of circular smooth muscle ap-
pears to be oriented along the circumference of the testis at right
angles to the superficial layer. In addition, numerous fibroblasts
were observed within the dense collagenous tissue of the tunica al-
buginea. A definite relationship appears to exist between the ex-
tent of the observed pharmacological responses and the amount of
smooth muscle present in the testicular capsules of the three species
studied. The fact that the testicular capsule was found to contain
smooth muscle seems to offer a reasonable anatomical explanation for
both the spontaneous and the drug-induced contractions which have

been observed with the isolated testicular capsule.
Pharmacological Studies of the Intact Rabbit Testis In Vivo
While in vitro isolated tissue experiments are known to accu-

rately reflect an actual in vivo situation, it nevertheless seemed
of importance to attempt to answer the difficult question of whether
spontaneous contractions as well as drug-induced contractions and
relaxations of the testicular capsule do indeed occu~ in a living
animal. Fig. 9 presents an in vivo isolated tissue apparatus which
we designed to measure spontaneous contractions of the intact rab-
bit testis in an anesthetized animal. An adult male rabbit weighing
approximately 3 kg was anesthetized with urethane and a small inci-
sion made through the various layers of the scrotum in order to ex-
pose the testis. A long piece of silk thread was carefully placed
through the testicular capsule at the superior end of the testis
with a second smaller tie placed through the capsule at the inferior
end of the testis. The intact testis was then gently drawn through
the glass tube in the rubber stopper and anchored to its top lip by
means of the inferior tie. The scrotal sac was then sealed around
the bottom lip of the glass tube by means of a strong circular tie.
A small water-jacketed bath was then placed over the rubber stopper,
with the superior thread from the testis leading through a top hole
in the bath to a transducer connected to a recorder. The bath was
then filled through an inlet tube with 20 ml of Tyrode's solution
kept at a constant temperature of 37 o C, corresponding to the normal
scrotal temperature of the rabbit, in order to prevent the testis
from drying out. The intact rabbit testis was completely surrounded
by this buffer solution while still being attached via the spermatic
cord to the anesthetized animal. It was found that the head of the
animal had to be raised slightly to prevent drainage of the buffer
solution into the peritoneal cavity. It therefore seemed possible
with this apparatus to be able not only to measure spontaneous con-
tractions of the intact testis in vivo, but also to attempt to pro-
duce drug-induced responses of the intact testis by injection of a
drug into the rabbit ear vein.
Fig. 10 presents the results of our initial in vivo pharmaco-
logical studies of the rabbit testis. The photograph illustrates
the anesthetized rabbit with its head slightly raised as well as the
intact testis inside the temperature-controlled bath simulating an
extension of the scrotal sac. As shown in the top recording, spon-
taneous contractions of the intact rabbit testis were found to occur
in vivo with an average frequency of 4 per minute, corresponding al-
most exactly to our previous in vitro results obtained with the iso-
lated testicular capsule. As shown in the bottom recording, the in-
travenous injection of 25 ~g/kg body weight of carbachol was observed
to produce a contraction of the intact rabbit testis in vivo which
in vivo ISOLATED TISSUE APPARATUS
J:.'C""'\M'i~~~ ...:........- TRANSDUCER
TO
RECORDER
SUPERIOR THREAD FROM --~+-----:.I~ ~ WATER JACKET EXIT

TESTIS TO TRA~DUCER
20 ml. BUF~R
EPIDIDYMIS - -.........- .. i~
INTACT RABBIT TESTIS --~t--~. ..,
~ WATER JACKET INTAKE

INFERIOR THREAD FROM ---t.I-::::~~~
T&:STIS ANCHORED TO
GLASS TUBE REHOVABI1.: RUBBER STOPPER
SURROUNDING GLASS TUBE
SPERK.\TIC CORD STILL SCROTAL SAC SEALED

ATTACHED TO AROUND GLASS TUBE
ANESTHETIZED ANIMAL
Fig. 9. Schematic representation of the in vivo isolated tissue

apparatus designed for measurement of spontaneous and drug-
induced contractions of the intact rabbit testis in an
anesthetized animal. The intact rabbit testis still at-
tached to the anesthetized animal by the spermatic cord is
kept at normal scrotal temperature by a constant-tempera-
ture circulating water bath through the outer water jacket.
Drugs can be administered intravenously into the systemic
circulation of the rabbit by employing either the ear vein
or the femoral vein.
was characterized by a marked increase in tone. The intravenous in-

jection of 75 ~g/kg body weight of isoproterenol, on the other hand,
was found to produce a marked relaxation of the intact rabbit testis
in vivo, again confirming our previous in vitro results. These data
indicate that both spontaneous contractions as well as drug-induced
contractions and relaxations of the rabbit testis related to the
testicular capsule can be demonstrated in a living animal using the
in vivo isolated tissue apparatus presently described.
ADULT RABBIT TESTIS 1n v1vo
25 I1g/Kg 75 l1g/Kg
CARBACHOL ISOPROTERENOL
Iv Iv
Fig. 10. Spontaneous contractions and drug responses of the intact
testis of an anesthetized rabbit employing the in vivo
isolated tissue apparatus described in Fig. 9. Two large
squares shown on the chart paper measure 10 mm. The re-
sponse magnification was x 25. The temperature of the
circulating water bath was maintained at 37 oC. Both car-
bachol, which produced a sustained contraction of the in-
tact testis, and isoproterenol, which produced a relaxa-
tion of the intact testis, were administered intravenously.
Relation of the Testicular Capsule to Sperm Transport
It has long been established that the sperm formed in the semi-
niferous tubules of the testis are not capable of motility (6) and
that the sperm first attain their motility in the epididymis (7).
The factors responsible for the initial transportation of non-motile
sperm out of the testis and into the epididymis have remained un-
clear, although a number of possibilities have been offered, includ-
ing fluid flow resulting from ciliary movements in the rete testis
and vasa efferentia (8) and in vitro undulating motions of indivi-
dual seminiferous tubules involving both Sertoli cells (9) and va-
rious forms of peri tubular cells (10-13). However, Leeson (14) in
investigating the fine structure of the rete testis by electron mi-
croscopy concluded that the number of cilia present in the rete epi-
thelium were insufficient to have any effect on sperm transport.
In addition, Cross (15) noted that when large coiled masses of semi-
niferous tubules were viewed directly with a microscope through an
incision of the abdominal wall of an anesthetized rabbit, the semi-
niferous tubules did not exhibit any contractions. Furthermore,
there have been suggestions that the testis as a whole organ might
be capable of undergoing spontaneous movements, as indicated by
rhythmic interstitial pressure changes obtained with a cannula (16)
as well as the observation of irregular movements of the entire tes-
tis (17). It therefore seems reasonable, in our opinion, to expect
something other than ciliary movement or possible minute in vitro
undulating motions of the seminiferous tubules as observed with a
microscope to be responsible for movements of the entire testis
which can be seen with the unaided eye. It would appear, based on
the data of the present report, that the responsible factor for tes-
ticular movements is the contraction and relaxation of the testicu-
lar capsule.
The present finding that the isolated testicular capsule of

both the rabbit and the human exhibits periodic and powerful spon-
taneous contractions in the absence of any added drug would seem to
indicate that under normal conditions, the testicular capsule is in
a constant state of dynamic movement, capable of exerting great
force against the contained seminiferous tubular mass. It seems
likely that these rhythmic contractions of the testicular capsule
then serve to massage the seminiferous tubules and in so doing pro-
vide a pumping action which transports the non-motile sperm from
the seminiferous tubules out of the testis and into the epididymis
where the sperm can then attain their motility.
The present experiments have also demonstrated that both of

the naturally-occurring neurohumoral agents of the autonomic nervous
system, namely acetylcholine and norepinephrine, are capable of caus-
ing contractions of the isolated testicular capsule of the rat, rab-
bit and human. The possibility therefore exists that one partial
explanation for the loss of fertility in males with sympathetic
nerve damage may involve a lack of norepinephrine, thereby resulting
in an impairment of testicular capsular contractions necessary to
propel sperm into the epididymis. It is interesting to speculate
that such cases of male infertility might be corrected if the im-
paired contractions of the testicular capsule could be artificially
stimulated by either a sympathomimetic agent or a parasympathomimetic
agent such as carbachol.
On the other hand, one possible approach to male contraception

may involve the administration of a drug which can cause a prolonged
relaxation of the testicular capsule, thereby preventing the sperm-
propelling action of the testicular capsule. It would also seem of
interest to explore the possibility that some drugs which are com-
monly administered to males may have as yet undetected side-effects
involving either the stimulation or inhibition of testicular capsu-
lar contractions, thereby affecting sperm transport from the testis.
Such experiments are now in progress in our laboratory.
SUMMARY
The testicular capsule of the rat, rabbit and human has been
prepared for the first time as an isolated tissue suitable for phar-
macological investigation. Both cholinergic and adrenergic agents
were found to cause a contraction of the isolated testicular capsule.
In addition, periodic spontaneous contractions of the isolated tes-
ticular capsule have been observed in the absence of any added phar-
macological agents. The present experiments have also demonstrated
that the testicular capsule contains smooth muscle, thereby offering
a reasonable anatomical explanation for both spontaneous and drug-
induced contractions of the testicular capsule. It therefore appears
likely that the rhythmic contractions and relaxations of the testi-
cular capsule provide a pumping action capable of explaining the
transport of non-motile sperm out of the testis and into the epidi-
dymis where they can then attain their motility.
ACKNOWLEDGEMENT
This investigation was supported by U.S. Public Health Service

Research Grant HD-01573 from the National Institute of Child Health
and Human Development.
REFERENCES
1. Davis, J.R. and Langford, G.A., Nature (London), 222: 386, 1969.
2. Davis, J.R. and Langford, G.A., J. Reprod. Fertil., 12: 595,
1969.
3. Tyrode, M.V., Arch. Int. Pharmacodyn., 20: 205, 1910.
4. Mancini, R.E., Arrillaga, F., Vilar, o. and De La Baize, F.A.,
Rev. Soc. Argent. Bioi., 11.: 161, 1955.
5. Mancini, R.E., Vilar, 0., Perez del Cerro, M. and Lavieri, J.C.
Acta Physiol. lat. amer., ~: 382, 1964.
6. Rendenz, E., Wurzb. Abhandl. ges. Med., 24: 107, 1926.
7. Yochem, D.E., Physiol. Zool., 3: 309, 1930.
8. Reid, B.L. and Cleland, K.W., AUst. J. Zoo1., 2: 223, 1957.
9. Roosen-Runge, E.C., Anat. Rec., 109: 413, 1951.
10. Clermont, Y., Exp. Cell Res.,.!2.: 438, 1958.
11. Lacy, D. and Rotblat, J., Exp. Cell Res., 11: 49, 1960.
12. Ross, M.H. and Long, I.R., Science, 153: 1271, 1966.
13. Niemi, M. and Kormano, M., Ann. Med. Exp. Fenn., 43: 40, 1965.
14. Leeson, T.S., Anat. Rec., 144: 57, 1962. --
15. Cross, B.A., In Recent Progress in the Endocrinology of Repro-
duction, ed. Lloyd, C.W., Academic Press, New York, p. 172.
16. Holstein, A.F. and Weiss, Ch., Z ges. expo Med., 142: 334,
1967.
17. Wojcik, K., Acta Physiol. Pol., 12: 78, 1966.
DISCUSSION
JOST: Dr. Davis, did you study the effect of oxytocin on the albugi-
nea?
DAVIS: We have just recently begun to investigate the effect of

oxytocin and vasopressin on the isolated testicular capsule of the
adult rat. Our preliminary experiments have indicated that both are
capable of causing contractions of the testicular capsule.
MANN: I am particularly impressed by the regularity of the spontane-

ous contractions which you observed, e.g. 13 minutes in the human
testis capsule. This seems to suggest that some chemical subst~nce
must be discharged at the time of spontaneous contraction, and that
during the period of relaxation that substance is gradually replen-
ished. Could this phenomenon be explained further by studying the
action of, for example, the parasympathetic and sympathic blocking
agents, such as atropine, on the spontaneous contractions?
DAVIS: We are presently investigating the effect of a number of

autonomic blocking agents on the drugs which we have shown to cause
both contraction and relaxation of the isolated testicular capsule.
We have found that atropine is capable of blocking acetylcholine-
induced testicular capsular contractions in the rat. However, we
have not yet investigated the effect of atropine on spontaneous con-
tractions of the rabbit testicular capsule.
TROEN: Dr. Davis, in your in vivo experiments do the contractions

cease for a while after ejaculation?
DAVIS: We have not yet attempted to induce ejaculation in the anes-

thetized rabbits, which were used for our in vivo isolated tissue ex-
periments with the testicular capsule.
BURGOS: I always ask myself how lymphatic vessels drain from the
testis as far as they occupy the largest part of the interstitial
tissue, so I find reasonable to accept as one of the main functions
of the permanent contractions of the capsule to press out the lymph
content of the testis.
DAVIS: I think that is a very interesting suggestion and we have

also been considering the possibility that testicular capsular con-

tractions may play an important role in the flow of testicular lym-
phatic fluid. We are also presently contemplating a speculation
that testicular capsular contractions and relaxations may result in
changes in the interstitial tissue pressure surrounding the Leydig
cells, which might possibly influence the secretion of testosterone
from the Leydig cells. We are now planning to place a micro pressure
transducer in the middle of the testis and to measure interstitial
tissue pressure simultaneously with OUr recordings of testicular
capsular contractions. We are also planning to cannulate the rete
testis, and following the addition of pharmacological agents which
cause contractions and relaxation of the testicular capsule, measure
both the flow rate and the sperm content of the rete testis fluid.
MANCINI: Dr. Davis, would you be able to separate the capsule of the
testis from that of the epididymisl Anatomically, there is a con-
tinuity between the capsule which surrounds the seminiferous tubules
and the capsule of the epididymis. If this could be done, you will
be able to see if differences in contraction exist. Have you al-
ready studied th~ capsule of the immature testis? In the human tes-
tis, there are some differences during development as regards the
morphology of the layers and the thickness of the albuginea.
DAVIS: In answer to your first question, the entire epididymis is

removed in the process of preparing the isolated testicular capsule.
In our in vivo experiments, we are always very careful to anchor the
inferior pole of the capsule which eliminates any possible influence
of the epididymis on the testicular capsules. In answer to your sec-
one question, we are planning to include the investigation of the ef-
fect of age as well as various hormones in our new research program
on the pharmacology and physiology of the testicular capsule.
EFFECT OF GONADOTROPINS AND ANDROGENS ON FRUCTOSE AND CITRIC ACID
OF SEMINAL FLUID
Juan Carlos Lavieri and Juan C. Calamera

de Medicina, Paraguay 2155, Piso 10, Buenos Aires,
Argentina
INTRODUCTION
The existence of a reducing sugar in human seminal fluid was

reported many years ago. It was then characterized as glucose (1-4).
Afterwards the substance was shown to be fructose (5); later on,
Mann (6) identified it as D(-) fructose. These findings were con-
firmed by others (7). Many papers have been devoted to the occur-
rence of this substance in both human and animal seminal plasma (8-
12); as well as to the site of its origin in different species in-
cluding man (13-17). In the human species the seminal vesicles are
the main source of its production. Moreover, the relationship be-
tween androgens and seminal fructose has been established (18), so
that the levels of this sugar have been used as indicators of an-
drogenic function (13,19).
Citric acid was found in semen by Schersten who placed its or-
igin in the adnexal genital glands (20). It is secreted by the sem-
inal vesicles in the boar and bull (21) and by the prostate in man
(22). As the production of citric acid depends on androgens (15,16,
23,24), both fructose and citric acid determinations were proposed
as tests of androgenic activity (25-27).
Considering the importance of semen characteristics in the eval-'

uation of male sterility, the authors carried out both biochemical
examination and morphologic studies of animal fluid. Furthermore,
the functional capacity of the adnexal glands as well as androgenic
effects upon them can be assessed. Low seminal concentrations of
fructose have been found in animals and man in hypoandrogenic condi-
tions (26,28,29), and are also seen in inflammatory and infectious
diseases of the adnexal glands (28,30,31). Information regarding
515
516 J. C. LAVIERI AND J. C. CALAMERA
oligospermic patients of unknown etiology without demonstrable en-

docrine, metabolic or urologic alteration is fragmentary (28,32,33).
Therefore, we considered it important to investigate the levels of
these seminial substances in the oligospermic patients and compare
these levels with those encountered in normospermic subjects. The
action of different hormones upon the level of these substances in
seminal plasma was also studied.
NORMOSPERMIC SUBJECTS
Seminal fluid showing the following characteristics was con-

sidered normal: over 40 million spermatozoa per ml, 2 ml ejaculate,
60% motility, 80% normal forms and no pus. Fructose (34) and citric
acid (35) concentrations were determined in 70 subjects with normal
seminal fluid.
Fructose. We found a mean of 266.4 mg % (range of 121-490 mg%);

these data were in keeping with those of other workers (26,36-42),
(see Table 1) •
Citric Acid. A mean of 421.4 mg %, (range 174-658) mg% was

found; these levels were also in keeping with those of other inves-
tigators (479 mg % (43), 376 mg % (26) and 347 mg % (42), see Table
1).
Table 1. Seminal Concentration of Fructose and Citric Acid in Nor-

mospermic and Oligospermic Subjects.
Number Concentration of Concentration of

SUBJECTS of FRUCTOSE CITRIC ACID
cases mg % mg 1'0
NORMOSPERMICS 70 266.4 + 9.2 421.4 + 14.8
OLIGOSPERMICS 64 286.4 + 11.6 511.8 + 22

(P < 0.20) (p< 0:001
OLIGOSPERMIC PATIENTS
The same investigations were applied to 64 patients with oli-

gospermia not related to either endocrine, metabolic or urologic
disturbance. The number of spermatozoa was below 40 million per ml
and generally, oligospermia was accompanied by reduced motility.
Each specimen was free of signs that indicated infection.
Fructose. A mean 286.4 mg % (range 121-462 mg %) was obtained.

Concentrations were similar to those levels found in normal subjects.
The means reported by other authors were: 211 mg % (41), 264 mg %
(33), 336 mg % (44) and 357 mg % (45), (see Table 1).
Citric acid. There was scarce information about its content

in oligospermic semen (32,33). Hence we were especially interested
in assessing the level of this substance in this category of patients.
We found a mean of 511.8 mg% (range 188-893 mg%). Comparison with
normal reveals a significant increase in the mean in oligospermic
patients (see Table 1).
HORMONAL EFFECTS
The oligospermic patients were divided into five groups, accord-

ing to the different hormonal treatment they received. The effect
of the hormones was evaluated by the difference in fructose and cit-
ric acid concentrations before and after treatment. Hormones admin-
istered were: human chorionic gonadotropin (HCG); pregnant mare
serum (PMS); human postmenopausal gonadotropin (HMG); testosterone
propionate; 17- trimethylsiloxy-4-androsten-3-one ether (SC-16l48)
(Silandrone. Searle and Co., Chicago, Illinois) (C22 H36 02 Si).
Statistical analysis consisted of: 1) Spot graphs using the

horizontal axis for basal values and the vertical axis for differ-
ences between basal levels and those levels encountered after hormon-
al treatment: 2) Calculation of the correlation coefficient. 3)
Fitting the regression line of differences from initial values. 4)
Construction of 95% confidence intervals for mean differences and
regression coefficients. These intervals were equivalent to tests
of significance at the 0.05 level. The following results were ob-
tained with the administration of the hormones above mentioned.
HCG. (1000 I.U. given intramuscularly, twice a week for six

weeks in each patient. This group comprised 17 cases.) The mean
fructose concentrations before and after treatment were not signifi-
cantly different. However, the negative slope was significantly
different from zero. This finding established a relationship between
the values obtained with HCG administration and the basal values. It
was also observed that the fitted lfne crossed the horizontal axis
near the midpoint of the range of initial values. This suggests that
with HCG treatment, low values tended to increase while high values
tended to decrease (see Fig. 1).
Results obtained on citric acid levels were similar to those

found for fructose (see Fig. 2).
PMS The dose used was 2000 I.U. given intramuscularly, twice a
week, for three weeks in each patient. This group comprised 20
cases. With regard to fructose concentration there was no signifi-
cant difference between the mean values noted prior to and after PMS
administration. Since the slope of the line did not differ signifi-
cantly from the horizontal line, the effect of treatment cO:lld not
be considered to be altered from the initial values (see Fig. 3).
HCG
Fructose u, .•
.-
, • - 0 •• 7 z •
1I1t.I.n.- • .,.) • - 0.50
6,.
~
- 5}.4) ~ 26.95
lOO - 0.90 6 fi ~ - 0.04
.
100
100
.. -
141\1.1 .""••
- 100
..
'laO
Fig. 1
1111.
200
.-
1.".- bot.) HCG
Citric Acid , • _ 0 •• 2
r __ O.72
& • 204 ••
- '7.9 " ,. ~ ... ,

- 0." ~fi oS - 0.20
100
.. -
.1.00 400 80 iaJUa! CODC •
-100
'200
•
Fig. 2
A significant difference in citric acid concentrations between

the mean values before and after HCG treatment were noted. However,
the slope was not significantly different from zero, indicating that
the difference between the increases and basal values was not treat-
ment related (see Fig. 4).
HMG The dose employed was: 1 ampoule (80 ~ FSH and 80 ~ LH

of the 2nd IRP-HMG) given intramuscularly every second day, for 90
days in each patient. This group comprised eight cas~s. Measure-
ment of fructose concentration revealed no significant differences
between the means. According to the confidence interval for the
true slope, again, it could not be accepted as an effect of treat-
ment (see Fig. 5).
Results on citric acid concentration were similar to those

found for fructose (see Fig. 6).
Testosterone Propionate. The dose used was 25 mg given intra-

muscularly, once a week, for four or six weeks in each patient.
This group comprised 22 cases. With respect to fructose concentra-
tion no significant difference was seen between the mean values be-
fore and after treatment. Nevertheless the significant negative
slope indicated that a relationship existed between final and basal
concentrations. Furthermore, the fact that the fitted line crossed
the horizontal axis in the vicinity of the mean of initial values,
suggested that low initial values increase and high ones decrease
under hormonal action (see Fig. 7). Similar results were found with
citric acid concentration (see Fig. H).
Silandrone. The dose employed was 30 mg/day, per os, for 30

days in each patient. This group comprised 15 cases. There was a
significant difference between means of fructose concentrations be-
fore and after hormonal administration. However, the slope of the
straight line did not differ from the horizontal line. In conse-
quence, increases and decreases could not be accepted as being re-
lated to treatment (see Fig. 9). Results similar to those on fruc-
tose concentration were found for citric acid concentration (see
Figo 10).
COMMENTS
Although we have used a different method for the determination

of citric acid our results are comparable with those obtained in
normal subjects by other workers (26,42,43). With regard to fruc-
tose our technique was the same as previously recommended (26).
The possibility of individual variations must be taken into

account (27). However, the subjects under study were requested to
adopt a determined sexual rhythm. In this way their seminal fluid
did not show major variations in basal conditions in coincidence
Dit.(an.- ht.) PMS

-" Fruct ose , • - 0.26 • • 93.'
r • - 0.27
_ 26.11' ~,. • 54.U
- 0.72 Sft " 0.20
zoo
-100
-100
Fig. 3
IIlI.(att.- lIel.)
PMS
.. - Citric Acid , • - 0.12 •• 192.11
-
400 r.-0.18
110.2 , P $ 1M ••
-0." .ft'O.20
IGO
•
aoo
-.. •
•
too
•
• -
...
• lalUal ...0
-"
•
•
100
• •
-too
Fig. 4
HMG
VIr.(att.- t>.t.J
Fructose
... ~
7 • - 0.)4 • • 118.6~
r • - 0.57
- 4).4) 4 /II ~ 78.68
200
- 0.82 6 fi ~ 0.1&
100
1118_
JII1U.1 _e.
100
-100
Fig. 5
100
.-
DlI.(an.- MI.)
HMG
Citric Acid
, • - 0.42 & • 2)9.'
r • - 0.28
- 37." 6 J' ~ 109.8
- 1.8) ~fi 6 o.~
,00
.,,-.
1aJU.1 c ....c •
100 800
-'00
Fig. 6
522 J. C. LA VIER I AND J. C. CALAMERA
TESTOSTERONE
Fructose
D1,.(an.- lie,.) , • _ 0.40 " • 115.6
.. II r • - 0.52
- 47.78 'J1 ' 19••
+100 _ 0.70'~ , - 0.10
.. o
o la1Ual _....
100 500
..-
-100
·0
-zoo
Fig. 7
TESTOSTERONE
Citric Acid
+200 1 • - 0.40 " • 15•• 7
r • - 0.81
- 108.~ 6)' • Z9.3
• - 0.56 oft ~ - 0.24

+100
-100
-200
D1t.(an.- lIet.)
.. II
Fig. 8
SC-16148
Fructose 1 • - 0.05 • • 100.05
DH. (art.- .,.r.) r _ - 0.1)
""" 51.52 fo,. ~ 118.48

- 0.27 ~ fi ~ 0.17
+200
+100r-------------~____~~___________________________
200 4/00 .00

.. "
la1 \1.1 00110 •
-100
Figo 9
.. "
Illt.(.t\.- .,.t.)
SC-16146
Citric Acid
+ 1100 1 - - 0.11
r __ O.lO
I • 228.2
43.7 6 )'.268.1
- 0.94 -)3 • 0.12
t 300
+ 200
+ 100
hlltlal CODe.
l1li"
4100 600 AOO
Fig. 10
with other investiSltors (46). When there were such variations the
samples were discarded. The initial value was the mean of two or
three ejaculates. We would like to emphasize that urinary steroid
and gonadotropin assays were not performed in our group of patients.
In oligospermic patients the mean concentration of fructose

does not differ from normals. However, the mean of citric acid con-
centration is significantly increased compared with normals. Within
each group there are subjects with low values and others with high
values indicating the existence of a wide range. Schirren (28) has
already reported several cases of oligospermia with low concentra-
tions of fructose. Kuhnau and Nowakowski (32) have reported other
cases with low fructose and high citric acid levels. They postulate
that these elevated concentrations might be due to the fact that the
seminal vesicles contribute a lesser volume. We feel that while
this interpretation might fit the cases with reduced volume of se-
men it may not be applied to those with normal volumes.
Although the dependence of these seminal substances on testos-

terone levels has been demonstrated in castrated animals (26), in
certain pathological conditions the facts are not quite as clear.
In oligospermic patients, the doses of HCG or testosterone used do
not produce variations in the mean of fructose and the citric acid,
but they do induce an increase in initial low values and a decrease
in high values.
In patients with "postpuberal Leydig cell insufficiency and

normospermia", methyltestosterone, HCG and methylandrostanolone
(mesterolone) increase fructose (29,32) and lower citric acid (32).
As the treatment is discontinued, the concentrations return to their
initial values. When the treatment is resumed the results are in-
fluenced as previously reported (32).
Although in our patients no hormonal steroid levels were de-

termined it could be speculated that the initial low values might
be due to an androgenic deficit or that the effectors do not respond
to the hormonal stimuluso We are not able to explain initial high
values on the same grounds unless we accept that androgens act by
regulating the concentration of these substances within given limits.
We deem as unsatisfactory the theory of functional exhaustion of
glands by overstimulation, since Silandrone further elevates initial
high values in similar conditions. Therefore, the mechanism of ac-
tion still remains undisclosed.
Silandrone induces an increase of both fructose and citric acid.

There is still not enough data on its effects in human, but in cas-
trated rats a peak of androgenic stimulation is seen at 20 to 30
days after administration; its effect upon the weight of seminal
vesicles and prostate is greater than that of testosterone (Searle,
Division of Biological Research. Personal communication). There
is also the likelihood that the dose employed by us is more andro-

genic than that of testosterone. All this would add support to the
hypothesis of a direct action upon adnexal glands as is seen in the
rat.
PMS was administered for only three weeks in order to prevent

the formation of anti-gonadotropins (47,48). No variation in fruc-
tose levels was observed with its use. Owing to its low LH activi-
ty and predominant FSH action, the dose employed would be insuffi-
cient to stimulate steroid production. The unexpected increase of
citric acid induced by PMS is by now a matter of mere speculation.
When HMG was used no variations were observed; this fact agrees
with its proved inability to produce steroids (49).
Further studies are needed to establish the relationship be-

tween the concentration of fructose and citric acid and the concen-
tration of spermatozoa in the seminal fluid under hormonal action.
CONCLUSIONS
1) Fructose and citric acid concentrations found by us in normal

subjects agree with those reported by other workers.
2) In oligospermic patients, seminal fructose concentration is sim-
ilar to that in normals; but citric acid is significantly increased,
as reported by other investigators.
3) In oligospermic patients HCG and testosterone tend to regulate
the levels of these substances. The marked elevation by Silandrone
may be connected to the relationship existing between the concentra-
tions of fructose and citric acid on the one hand and androgenic
function on the other hand, as demonstrated by Mann in animals.
4) It is difficult to interpret the mechanism of action of these
hormones in the absence of information regarding urinary or plas-
matic steroidal levels.
ACKNOWLEDGMENT
The authors wish to express their gratitude to the Centro de

Bioestadistica y Demografia, Facultad de Medicina de la Universidad
de Buenos Aires for the statistical analysis of our results.
REFERENCES
1. McCarthy, J.F., Stepita, C.T., Johnson, M.B. and Killian, J.A.,

Proc. Soc. Exp., 25: 54, 1928.
2. Ivanov, E. E., Z. Zucht., B, 20: 404, 1931.
3. Huggins, C.B. and Johnson, C.-,-Amer. J. Physiol. 103: 574, 1933.
4. MacLeod, J. and Hotchkiss, R.S., J. Urol., 48: 225, 1942.
5. Yamada, K., Japan J. Med. Sci. II, Biochem. 2, 93: 245, 1933.
6. Mann, T., Biochem. J., 39: 458, 1945.
526 J. C. LAYIERI AND J. C. CALAMERA
7. Pryde, J. ,Nature, (London), 157: 660, 1946.

8. Mann, T., Biochem. J., 40: 481, 1946.
9. Mann, T., Advances in Endocrinology, ~: 329, 1949.
10. Mann, T., Recent Progr. Hormone Res., 12: 353, 1956.
11. Mann, T. and Lutwak-Mann, C., Physiol.~ev., 11: 27, 1951.
12. Mann, T. and Wilson, E., J. Endocr., 25: 407, 1962.
13. Davies, D.V. and Mann, T., Nature, (London), 160: 295, 1947.
14. Mann, T., J. Agric. Sci., 38: 323, 1948. --
15. Humphrey, G.F. and Mann, T~ Nature, (London), 161: 352, 1948.
16. Humphrey, G.F. and Mann, T., Biochem. J.,44: 97, 1949.
17. Bonadonna, T., Zootecnica e Veterinaria, Milano.
18. Mann, T. and Parsons, U., Nature, (London), 160: 294, 1947.
19. Landau, R.L. and Loughead, R., J. Clin. Endocr., 11: 1411, 1951.
20. Schersten, B., Skand. Arch.,58: 90, 1929.
21. Mann, T., Davies, D V. and Humphrey, G.G., J. Endocr., 6: 75,
1949.
22. Huggins, C. and Neal, W., J. Exp. Med., ~: 527, 1942.
23. Lutwak-Mann, C., Mann, T., and Price, D., Proc. Roy. Soc. (Biol),
136: 461, 1949.
24. Mann, T. and Parsons, U., Biochem, J., 46: 440, 1950.
25. Nowakowsky, H. and Schirren, C., Klin, Wschr., 34: 19, 1956.
26. Mann, T., The Biochemistry of the Semen and of the Male Genital
Tract. Methuen, London, 1964.
27. Mann. T., Ciba Found. Colloq. Endocr., 16: 233, 1967.
28. Schirren, C., J. Reprod. Fertil., 5: 34~ 1963.
29. Schirren, C" 5th World Congress Fertil. Steril., Stockholm,
1966. Excerpta Medica Found. International Congress Series,
133 p. 808, 1967.
30. Eliasson, R., Fredricsson, B., Johannisson, E. and Leander, G.,
5th World Congress Fertil. Steril., Stockholm, 1966. Excerpta
Medica Found. International Congress Series 133, p. 625, 1967.
31. Eliasson, R., Fertil. Steril., 11: 344, 1968.
32. Kuhnau, J. Jr. and Nowakowski, H., Endokrinologie, 40: 1, 1960.
33. Calamera, J.C. and Lavieri, J.C., Bol. Clin. Endocr. Metab.,
7.. : 149, 1968.
34. Roe, J.H., J. Biol. Chern., 105: 15, 1934.
35. Chambon, P., Ann. Pharm. Franc., l!: 78, 1963.
36. Tyler, E.T., Fertil. Steril.,~: 247, 1955.
37. Harvey, C., Proc. Soc. Study of Fertility, §: 3, 1957.
38. Vaishwanar, P.S., Amer. J. Obstet. Gynec., 22: 139, 1958.
39. Schirren, C., Fertilitatsstrorungen des Mannes. Stuttgart,
Enke., 1961.
40. Shet, A.R. and Rao, S.S., Indian J. Med. Sci., 1£: 709, 1962.
41. Moon, K.H. and Bunge, R.G., Fertile Steril., 19: 186, 1968.
42. Ackerman, D.R. and Roussel, J.D., J. Reprod. Fertil., 12: 563,
1968.
43. Harvey, C., Proc. Soc. Study of Fertility, 1: 56, 1951.
44. Abelli, G. and Falagario, M., Attualita Ostet. Ginec., 12: 723,
1966.
45. Pasetto, N., Monitore ost. Gin. di Endocrin. e del Metab., 27:
53, 1965.
46. Eliasson, R., J. Reprod. Fertil" .2: 331, 1965.
47. Maddock, W., Epstein, M. and Nelson, W.O., Ann. N.Y. Acad. Sci.,
55: 657, 1952.
48. Oestergaard, E.- XXI Assisses Franc. de Gynecologie. Clermont
Ferrand, 1962.
49. Mroueh, A., Lytton, B. and Kase, N., J. Clin. Endocr., 12: 53,
1967.
IMMUNOLOGICAL ASPECTS OF MALE INFERTILITY
Roberto E. Mancini

de Medicina, Paraguay 2155, Piso 10, Buenos Aires,
Argentina.
INTRODUCTION
It is very well known that formation of antibodies in a given

organism toward a foreign molecule, virus, bacteria or tissue, con-
stitutes the normal mechanism of defense. Moreover, the concept of
an autoimmune response against self tissue antigens in mammals has
been widely accepted in the last ten years. Encephalomyelitis,
thyroiditis and adrenalitis experimentally induced by sensitizing
animals with homologous or autologous antigens extracted from these
glands added with adjuvants have met their equivalent in some clin-
ical endocrine diseases. However, the role played by the induced
Circulating and delayed types of antibodies in conditioning the
glandular tissue destruction is still a matter of debate.
Allergic orchitis similarly induced and subsequently followed

by azoospermia has been an excellent model to substantiate some
cases of male sterility with unexplainable etiology. Consequently,
it was admitted that man may be immunized against his own sperm ma-
terial and a woman against the same antigen from a man. Two reasons
may explain why autosensitization is possible in males; firstly the
sperm cells are anatomically isolated from the circulation in the
genital tract and secondly they develop at puberty long after the
neonatal age where immunological tolerance can take place.
In the following sections, the animal experiments and clinical

findings supporting this new pathogenic contention in cases of male
sterility and the participation of the testis, adnexal glands and
semen will be briefly described.
529
530 R. E. MANCINI
I EXPERIMENTAL BACKGROUND
Testis
It has long been demonstrated that a large variety of animal

testicular tissues are antigenic in heterologous sensitization pro-
cedures. The "spermotoxic" antisera were capable of irrnnobilizing
normal spermatozoa in vitro (1). Nevertheless, an impact in the
field of irrnnunologic impairment of reproductive male function was
made when selective damage of the germinal epithelium of adult
guinea pigs could be obtained after auto- or homo-sensitization
with a single dose of testicular homogenate plus complete Freund's
adjuvant (2,3).
The allergic response developed between one and eight weeks

and was characterized by: a) edema in the intertubular spaces of
the gonad and epididymis, foci of perivascular cellular infiltra-
tion including mostly plasmocytes and lymphocytes; b) vacuolization
and sloughing of the germinal cells, beginning with spermatids, and
finally involving spermatocytes and spermatogonial cells. Only Ser-
toli cells showing striking vacuolization but unchanged nuclei re-
mained; c) no damage was demonstrable in Leydig cells or in other
connective tissue cells, such as fibroblasts and scattered mast
cells; d) the testicular lesion was accompanied by the appearance
of circulating antispermatic and cell-bound antibodies; e) as a
consequence of the germinal cell destruction, azoospermia rendered
the animal sterile for several months; f) signs of complete recovery
of fertility were apparent about 165 days after sensitization and
detection of normal motile spermatozoa in the ejaculate occurred
after about five months of azoospermia; g) the reaction was not
evoked by injections of adjuvant alone, irrnnature testis or even im-
munologically-damaged testis (4,5). As well as guinea pig, other less
susceptible species such as inbred strains of mice (6) or adult rats
(7,8) may respond to similar auto or homosensitization. Significant
testicular damage in an inbred strain of rats was also accomplished
with one injection of testicular antigen incorporated in complete
Freund's adjuvant (9). Recently (10) autoallergic orchitis was pro-
duced in primates. Rhesus monkeys were auto or isosensitized with
testicular homogenates in Freund's adjuvant. After six to eight
weeks they developed a patchy lesion similar to that observed in
guinea pigs. Circulating antibodies were detected in few animals
at very low titers; however, following castration the titers in-
creased. No cell bound antibodies could be detected. The produc-
tion and pathway of steroid hormones were also studied by in vitro
incubation with pairs of the differentially-labelled precursors (11).
The rate of production of testosterone by the testicular tissue was
somewhat decreased but the results were not statistically signifi-
cant and were in accordance with the lack of damage of the Leydig
cells in the irrnnune process (12).
The need for the simultaneous presence of germinal cell homo-

genetes and adjuvant in the production of unilateral allergic or-
chitis has been substantiated by the induction of a local inflamma-
tion in the contralateral testis together with an intradermal in-
jection of the adjuvant (13) or by the injection in the contralater-
al testis of the adjuvant alone (14). In contrast to the admitted
role played by complete Freund's adjuvant in conditioning the tes-
ticular and antibody response, it was claimed that adult testicular
extract,if repeatedly administered in guinea-pigs for three months,
may elicit gonadal damage (15), as was shown in other autoimmune
diseases such as thyroiditis.
Histology. The vascular and inflammatory changes in allergic

orchitis are minimal as compared to the dramatic breakdown of the
germinal epithelium. There is general agreement that slight vascu-
lar changes and histiocyte-mononuclear cell infiltration appear
around the second week and may precede the germinal cell lesion.
However, the known sensitivity of this epithelium to different kinds
of agents has made controversial the specificity of sloughing and
cytolysis of germinal cells (16). Nevertheless, the immunological
character of this type of orchitis appears reinforced by its cor-
relation with the presence of both circulating and cell-bound anti-
bodies and by the negative findings in animals injected with other
homologous tissue extracts (17).
The cellular basis of the seminiferous tubule lesion has been

reinvestigated in our laboratory and the findings of previous au-
thors have been extended to the subcellular level. The dissocia-
tion of the unit integrated by Sertoli and germinal cells mediated
by the rupture of peripheral areas of Sertoli cells constitutes the
main phenomenon of the induced aspermatogenesis. Development of
numerous large reticuloendoplasmic vesicles and of dense bodies
showing acid phosphatase reaction in the Sertoli cells (lysosomes)
coincides with the cytolysis of spermatids and sloughing of sperma-
tocytes and spermatogonial cells. Subsequently, the perinuclear
acrosomes appear free in the lumen of the tubules and then in the
epididymus, where they are reabsorbed (18). Changes in some other
hydrolytic enzymes have been reported (19). The Sertoli cells,
acting as a sustentacular and transporting cell system between the
intertubular circulation and closely associated germinal cells (20),
would permit the antibodies to reach the antigen present in the
spermatogenic cells.
The appearance of a bradykinin-like substance in the testicu-

lar tissue during the first weeks of the allergic response is also
interesting because of its role as mediator in the antigen-antibody
interaction in allergic inflammation (21).
Antigens. The heterogeneous nature of testicular homogenates

as revealed immunoserologically by the presence of other tissue
532 R. E. MANCINI
antigens common to kidney, liver and serum proteins points to the

problem of the specific antigens responsible for the antigenicity
of the testis (22). Other protein fractions obtained from the tes-
tis have been reported as being capable of inducing allergic orchi-
tis, but a testicular polysaccharide-polypeptide complex appears to
be a highly potent antigen (23). Recently a glycoprotein sub-
stance isolated from guinea pig germinal cells, devoid of hyaluron-
idase activity or tissue and serum protein contaminants, was able
to produce allergic orchitis (24).
The site of localization of the antigens was assessed by the

immunofluorescent technique and it was demonstrated that the homo-
logous autologous-anti testis antibodies react specifically with the
acrosome of spermatozoa (25) and spermatids (26). This finding and
the fact that this structure reacts positively to PAS, together
with the polysaccharidic nature of testicular antigens, supports the
earlier contention that the later cell stages of spermatogenesis and
particularly the spermatids and spermatozoa share the antigenic po-
tency of the testis.
Antibodies. Two types of different immunological significance

(circulating and cell bound) have been described in the guinea pig
as a result of a complete sensitization. The former are of low ti-
ter and may be detected by systemic anaphylaxis, local anaphylaxis
(immediate inflammatory skin reaction), passive cutaneous anaphy-
laxis (PCA), Schultz-Dale, complement fixation, hemagglutination
and precipitation. The immune serum also shows in vitro cytotoxic
properties such as immobilization and agglutination of motile sper-
matozoa (27) and a lytic effect on the acrosome of spermatids and
spermatozoa (28). The cell bound antibodies are involved in the
delayed type of hypersensitivity. These antibodies may be detected
by the appearance of a typical granuloma after intradermal injection
of the antigen alone.
In spite of the lack of detailed studies it is assumed that

the circulating antibodies are inconstant and of low titer and that
no strict correlation exists between them and the gonadal lssion.
On the contrary, a higher frequency of delayed skin reaction in sen-
sitized animals is generally admitted. Moreover, a sequential fol-
low up of the animals showed that delayed hypersensitivity appears
first around the sixth day and circulating antibodies can be detec-
ted approximately three days later at the time when tubular damage
begins to develop (29,30).
That complete adjuvant is necessary for the allergic orchitis

and delayed skin reaction is supported by the fact that testicular
tissue alone or with the addition of incomplete adjuvant may only
elicit systemic or local anaphylaxis, PCA and immunofluorescent
antibodies (31).
Pathogenesis. It has definitely been accepted that induced

allergic orchitis may be included in the category of a delayed type
of sensitization due to cell bound antibodies. This inference has
been drawn from the following facts: a) the use of complete adju-
vant is essential; b) testicular lesions are correlated more with
delayed skin reaction than with circulating antibodies; c) histolo-
gy of the testis shows not only typical aspermatogenesis but also
mononuclear cell infiltration and weak inflammatory signs; d) mono-
nuclear cells from lymph nodes of sensitized animals are capable
of passively inducing skin delayed hypersensitivity when they are
transferred to normal receptors (32). Moreover, fetal lesions of
"aspermatogenesis" have been evoked by intratesticular injections
of lymph node cells obtained from sensitized animals (33). On the
contrary, circulating antibodies cannot induce aspermatogenesis
passively. Also the immune globulin labelled with fluorescein can-
not be detected inside the seminiferous tubules after intravenous
injection (33). However intratesticular injection of immune serum
may elicit local specific changes (34).
Although the role played by the different types of antibodies

remains to be elucidated, the dual necessity for cell bound and
circulating antibodies in the pathogenesis of experimental asperma-
togenesis has recently been assumed (35). It is possible that
among the various antibodies, some may be concerned with damage to
the testis (cytotoxic and cell bound) while others are of an ana-
phylactic type and exert a "protective effect" (36). Nevertheless,
a sufficient body of experimental evidence on the significance of
all these antibodies as well as on the antispermatic properties of
different globulin fractions is still needed.
The prevention of induced allergic aspermatogenesis by thy-

mectomy early in the post-natal period (37) or by the administra-
tion of suppressive drugs (38) prior to sensitization has been
claimed. On the other hand the induction of immunotolerance early
in the postnatal period failed, because germinal cell destruction
also develops when the animals reach puberal maturation (39).
Semen
Earlier studies have shown that free spermatozoa or total se-

men may induce by heterologous sensitization a high titer of sperm
immobilizing antibodies.
More recent investigations have emphasized the complexity of

semen and the importance of considering sperm and seminal plasma
antigens separately. That spermatozoa are mainly responsible for
the antigeaicity of testis extract was deduced from the production
of homologous allergic orchitis using guinea pig seminal or epidi-
dymal sperm (40). In addition, cross immunoserological reactions
have been observed between testis and epididymal or seminal sperma-
534 R. E. MANCINI
tozoa (41). As in the testis, more than one spermatozoal antigen,

which seem to be polysaccharide (42) or glycoprotein substances
(43), is able to elicit, with the addition of complete adjuvant,
testicular damage and circulating and cell bound antibody response.
However, based on serological, immobilizing and agglutinating tests,
some discrepancies still stand with regard to sperm species speci-
ficity (44). Also three antigens have been recognized after ultra-
sonic vibration in spermatozoa from various species including man;
one in the head, a second in the tail and a third common to both por-
tions (45). Based on the immunofluorescent reaction of spermato-
zoal acrosomes, it has been shown that isolated acrosomes have an
alcohol precipitable fraction rich in aminosugars and sialic acid
with a higher antigenic potency than other soluble fractions and
residual structures of this cell (46).
The antigenic properties of seminal plasma in heterologous or

homologous sensitization procedures are well known. Applying sero-
logical techniques and immunoelectrophoresis, several fractions
which are immunologically common to serum proteins, epididymal or
prostate. fluids and to seminal spermatozoa but not to testicular
tissues have been studied (47). In contrast to discrepant reports
(48), the finding of specific antigens in the seminal spermatozoa
which are not of seminal plasma origin was corroborated by direct
and cross immunoserological reactions (49). Ejaculated rabbit
semen shows about nine antigens, seven of which are common to sem-
inal plasma, two to blood serum and two appeared to be sperm speci-
fic and originated in the testis (50). It is then likely that dur-
ing ductal transit spermatozoa are firmly coated with antigenic
material from seminal plasma (51). Indeed washed sperm are consi-
derably less antigenic (52).
Nevertheless, a detailed study of different antibodies induced

by both types of spermatozoal antigens, and their relation with
those elicited by seminal plasma and testicular antigens, is not
complete until now.
Adnexal Glands
The possibility that seminal plasma antigens may be present

in the accessory glands before they reach the semen opened a new
approach to possible autoimmunological damage of these glands and
of sperm cells related to anti-seminal plasma antibodies.
As referred to above, antigens common to seminal plasma, sperm

cells, prostate and seminal vesicles secretions have been demon-
strated. Also rabbit auto- and isoantibodies have been detected in
prostate, seminal vesicles, coagulating and bulbo urethral glands.
These antigens are also contained in seminal plasma but absent in
testicular tissue (53). Although the isoantibodies do not induce
lesions in the glands, those against guinea pig seminal vesicle or
prostate agglutinate spermatozoa at lower titer as compared with

those against the epididymis (54).
II HUMAN SPERMATIC TISSUE
Testis. Earlier findings showed that extracts of human testi-

cular tissue seem to be antigenic in sensitized animals. A prepa-
ration isolated from the human testis containing proteins and car-
bohydrates has shown antigenic potency in guinea pigs. Antibodies
demonstrated by precipitating, Schultz-Dale and immunoelectrophore-
sis techniques cross react with guinea pig testis antigen (55).
The possible role played by the human testis in the immunolo-

gic response to auto- or homosensitization was substantiated trying
to induce an allergic orchitis in man, as has already been obtained
in animals. Volunteer patients with prostate carcinoma were divided
into three groups according to age, previous treatment, general con-
dition and testicular biopsy. Patients of the first group were se-
parately auto- or homosensitized with testicular homogenate, a pro-
tein fraction, a mucoprotein preparation provided by Dr. S. Katsh,
(Denver, U.S.A.) or a glycoprotein substance obtained from human
germinal cells. The second group of patients was sensitized with
the same antigens alone or incorporated into complete adjuvants,
or with the adjuvants alone. The third group not sensitized at all
served as a control. Circulating as well as cell bound antibody
determinations and testicular biopsies were performed during and at
the end of the experiment, which was followed by therapeutic cas-
tration. As compared with the controls, some patients of the ex-
perimental group exhibited low titer circulating antibodies (com-
plement fixing, sperm immobilizing, precipitating and passive cu-
taneous anaphylaxis) and positive skin test as well. Biopsies of
the testis showed tubular lesions with sloughing and cytolysis of
germinal cells. Immofluorescent techniques using the patient's
own sera revealed positive results in the head of spermatozoa and
perinuclear area of spermatids (56,57).
More recently another group of similar patients was submitted

before castration to a mild unilateral orchitis in order to induce
antibodies against spermatic autoantigens emerging from germinal
epithelium destruction. However, no immunological response could
be detected and no lesions developed in the contralateral gland (58).
Semen and Adnexal Glands. Species specific antibodies were

demonstrated by complement fixing, immobilizing and sperm agglutin-
ating tests against these antigens (59). Some experimental evidence
indicates that human spermatozoa extracted from spermatocele lack
the antigens which characterize the seminal spermatozoa. The pos-
sibility that a great part of the antigenic material is taken up
during transit, as occurs in animals,was substantiated by direct and
cross reactions with heterologous antihuman semen and spermatozoal
536 R. E. MANCINI
or seminal plasma antibodies (60). This has been recently substan-

tiated using histoimmunochemical techniques which show several anti-
genic sites in the head, neck and tail of human seminal spermatozoa
whereas only one is demonstrable in testicular sperm cells (61).
Human semen contains at least 16 antigens and the spermatozoa

seven, four of the latter being common to seminal plasma (62). Im-
munochemical analysis of seminal plasma revealed the presence of
mucoproteins which seem to have antigenic properties. Prostate
specific antigens, as well as cross reaction between this gland or
acid phosphatase and seminal spermatozoa have also been shown (63).
Also antibodies against seminal plasma and prostate may cause ag-
glutination of seminal spermatozoa and higher hemagglutinating ti-
ters than antibodies against testis, epididymis, seminal vesicle or
spermatozoa. Cross agar diffusion reactions showed that seminal
plasma has a high common antigenicity with prostate and seminal ves-
icle and a lower one with seminal spermatozoa, testis or epididymis.
By immunoelectrophoresis, one of the prostate antigens was identi-
fied as acid phosphatase and -that of the seminal vesicle as lac to-
ferrin, the latter being claimed as the more potent sperm coating
antigen (64). From a comparative study, it has recently been con-
cluded that the immobilizing, agglutinating and cytotoxic properties
of anti seminal plasma antibodies are higher than those of antisper-
matozoal and antitestis antibodies (65).
It is then possible to assume that the testis contributes a few

antigens to semen which appear to be sperm specific, as they are
present in washed epididymal and seminal spermatozoa but not in sem-
inal plasma, whereas prostate and seminal vesicles provide the major
part of secondary sperm antigens.
III CLINICAL FINDINGS
The possibility that some cases of male sterility of unknown

etiology have an immunological pathogenesis has attracted the atten-
tion of investigators during the past ten years. The interest began
when sperm agglutinins were found in human blood and semen in some
sterile patients (66). Microscopic and macroscopic agglutination
techniques were recommended, and patients' spermatozoa or sperma-
tozoa from normal donors were currently used. With macroscopical
agglutination techniques 3.3% of the sera of infertile males were
found to be positive while serum samples of fertile males did not
show appreciable agglutinins (67). In a random group of 172 azo-
ospermic patients without spermagglutinins in their serum~ oligo
or azoospermia was found in 94 cases, obstruction in 45, and unqua-
lified in 33 patients (68). Applying the same agglutination tech-
nique, 3% was found among cases of oligo and azoospermia (69) where-
as 14% of a group of cases of inflammatory diseases of the genital
tract was reported (70). Using the supernatant of freeze-thawed
spermatozoa as antigen, hemagglutination gives 18% of positive re-
actions in the sera of infertile men (71). The same technique

using washed sperm and seminal plasma as antigens in the hemagglu-
tination technique showed no parallelism with fertility potency in
the male (72). It is interesting to note that some cases showing
spermagglutinins in their sera do not give a complement fixation
response with spermatozoa or seminal plasma but they produce a zone
of precipitation with seminal plasma in the gel diffusion technique
(73). No strict correlation was observed in a large group of ster-
ile patients, between sperm agglutination and hemagglutination
techniques using seminal plasma as antigen. Moreover, blood group
isoantigens known to be present in sperm cells or in seminal fluid
may induce false positive results when the hemagglutination is used
(74).
Applying several immunoserological tests, cases showing a his-

tory of retention and of inflammation of the testicle and epididymis
gave a 22% positive response with more than one of the tests used
(75). The presence of injury, ligation or infection of the vas
deferens or epididymis in a high percentage of oligo or azoospermic
patients led to the hypothesis that sperm reabsorption into the
intersti.tial tissue could elicit the immunological response (76).
However, some patients who had surgical relief of the obstruction
continued demonstrating positive spermagglutinin titers and pro-
duced pregnancies (77).
Immunofluorescent studies, using serum having spermagglutinins

and seminal or testicular spermatozoa, showed positive results.
Specific fluorescence appeared to be located in the head and occa-
sionally in the mid-piece or the tail (78). To check the partici-
pation of germinal cell destruction during inflammatory conditions
of the gonads, antispermatic antibodies were investigated in pa-
tients suffering mumps orchitis. Applying complement fixation, gel
diffusion, antiglobulin consumption, and skin tests, the presence
of antibodies related to delayed hypersensitivity rather than the
circulating type was found. Negative results appeared in fertile
men, children and in primitive or secondary diseases of the testis
having non-developed or involutive germinal epithelium (79).
Although the widely used sperm agglutination techniques do not

provide the certainty of an immunological reaction, it has been
proved that the same sera may immobilize sperm in the presence of
complement and that both antibodies were mainly of the 7S type.
Moreover the levels of IgG and IgM were significantly higher in sera
from men with sperm antibodies than in control sera (80). Then it
is evident that antisperm antibodies may occur in sterile patients,
but the independent or combined participation of testis, adnexal
glands and seminal plasma in the immunologic response are not yet
clarified. Although active germinal cell breakdown in the testis
or spermiostasis in the efferent duct may exist, simultaneous pre-
sence of an inflammatory process seems necessary in order to ini-
tiate the antibody response.
538 R. E. MANCINI
CONCLUSIONS
An evaluation of the immunological factors in the male repro-

ductive process may be made as follows: 1) Alteration of spermato-
genesis or impairment of seminal plasma formation may be accompa-
nied by the presence of antispermatic antibodies. 2) Oligo or azo-
ospermia might reflect the immunological destruction of germinal
cells whereas immobilization and agglutination of seminal spermato-
zoa would depend on allergic reaction originating in the adnexal
glands. 3) Maturation of spermatozoa appears to be related to dif-
ferences in their antigenic potency, for antibodies against testis
do not affect adnexal glands but may immobilize epididymal sperma-
tozoa, whereas antiseminal plasma antibodies react only with adnex-
al gland and seminal spermatozoa but not with testicular cells. 4)
Acute or subacute inflammatory lesions of gonad and genital tract
appear more related to the presence of antibodies than chronic le-
sions or endocrine diseases of the testis. Thus, the presence of
adjuvants added to the antigen in the experimental induction of
these antibodies, might be replaced in the naturally occurring dis-
eases of the genital tract by hap tens resulting from different
kinds of injury to the tissues. 5) While the sperm antigens seem
to be of a glycoprotein nature accumulated in the acrosome, neck
and intermediate segment, those belonging to the adnexal glands and
seminal plasma should be better known. 6) The site of reaction of
sperm or adnexal gland antigens with the antibody in the male gen-
ital tract, the nature of the phenomenon and its consequence on the
reproductive process need to be elaborated. 7) The significance
and pathogenic role of detectable antispermatic antibodies in ster-
ile male patients, both circulating and cell bound types, are not
yet sufficiently clarified. Their presence, apparently correlated
with unexplained cases of infertility, does not necessarily mean
that they are specifically responsible for this abnormal state and
only would be a reflection of the immunologic process involved.
ACKNOWLEDGEMENT
This work was supported by a grant from The Population Council,

Inc. New York.
REFERENCES
1. Metchnikoff, E., Ann. Instit. Pasteur (Paris), 14: 1, 1900.

2. Voisin, G., Delaunay, A. and Barber, M., Ann. Instit. Pasteur
(Paris), ~:48, 1951.
3. Freund, J., Lipton, M.M. and Thompson, G.E., J. Exp. Med., ~:
711, 1953.
4. Freund, J., Lipton, M.M. and Thompson, G.E."Proc. Soc. Exp.
Biol. Med., 87: 408, 1954.
5. Katsh, S. an~Bishop, D., J. Embryol. Exp. Morph., 6: 94, 1958.
6. Pokorna, Z., Vojtiskova,M. and Chutna, J. Folia Biologica
(Praha), 1: 203, 1963.
7. Freund, J., Lipton, M.M. and Thompson, G.E., J. Exp. Med.,

101: 591, 1955.
8. Mancini, R.E., Davidson, O.V., Vilar, 0., Fertil. Steril.,
12: 695, 1964.
9. Andrada, J.A., Andrada, E.C. and Witebsky, E., J. Immunology,
1969 (in press).
10. Andrada, J.A., Andrada, E.C. and Witebsky, E., Proc. Soc. Exp.
Bioi. Med., 130: 1010, 1969.
11. Hoschoian, J .C. and Brownie, A., Steroids, 10: 49, 1967.
12. Andrada, J.C., Proc. First Argentinian Congo Sterl1ity, 1: 65,
1968.
13. Boughton, B. and Spector, W.G., J. Path. Bact., 86: 69, 1963.
14. Eyquem, A. and Krieg, H., Ann. N.Y. Acad. Sci., 124: 270, 1965.
15. Bishop, D.V., Proc. Soc. Exp. Bioi. Med., 107: 116, 1961.
16. Waksman, B.H., J. Exp. Med., 109: 311,1959.
17. Baum, J., Boughton, B., Mongat, L. and Schild, O.H., J. Immu-
nology, ~: 95, 1961.
18. Mancini, R.E., Proc. 1st. Symp. Intern. Committee Immunol.
Reprod. Geneva, p. 55, 1968.
19. Katsh, S. and Aguirre, A., Int. Arch. Allergy, 11: 141, 1968.
20. Mancini, R.E., Vilar, 0., Alvarez, B. and Seiguer, A.C., J.
Histochem. Cytochem, 1}: 376, 1965.
21. Mancini, R.E., Huidobro, H., Fernandez Collazo, G. and Mona-
stirsky, R., Proc. Soc. Exp. Bioi. Med., 123: 227, 1966.
22. Alonso, A., Bueno, M.P., Scacciati, I.M., Gonzalez, N. and
Mancini, R.E., Acta Europea Fertil., 1: 459, 1969.
23. Bishop, D. and Carlson, G.L., Ann. N.Y. Acad. Sci., 124: 247,
1965.
24. Scacciati, I.M., Gonzalez, N., Bueno, M.P., Alonso, A. and
Mancini, R.E., J. Reprod. Fertil., 1970 (in press).
25. Swanson Bech, J., Edwards, R.G. and Young, M.R., J. Reprod.
Fertil., ~: 103, 1962.
26. Mancini, R.E., Davidson, O.W., Nemirovsky, M. and Bueno, M.P.,
Proc. Soc. Exp. Bioi. Med., ~: 435, 1962.
27. Voisin, G.A., Toullet, F. and Maurer, P., Ann. N.Y. Acad. Sci.,
73: 726, 1958.
28. Mancini, R.E., Monastirsky, R., Fernandez Collazo, E., Seiguer,
A.C. and Alonso, A., Fertil. Steril., 70: 779, 1969.
29. Boughton, B., Immunology, 2: 522, 1962.
30. Brown, P.C., Glynn, L.E. and Holborow, E.J., J. Path Bact.,
86: 505, 1963.
31. Chutna, J. and Rychlikova, M., Folia Biologica (Praha), 10:
177, 1964.
32. Laurence, K.A., Carpuk, B.A. and Perlabachs, M., Int. Fertil.,
1Q: 13, 1965.
33. Mancini, R.E., Barquet, J. and Alonso, A., Proc. Soc. Exp. Bioi.
Med., 1970 (in press).
34. Mancini, R.E., Alonso, A. and Barquet, J., Acta Europea Fertil.,
1970 (in press).
540 R. E. MANCINI
35. Brown, P.C., Glynn, L.E. and Holborow, E.J., Immunology, 11:
307, 1967.
36. Chutna, J. and Rychlikova, M., Folia Biologica (Praha), 10:
188, 1964.
37. Vojtiskova, M., and Pokorna, Z., Lancet, 11: 644, 1964.
38. Vojtiskova, M., Vikliky, V., Jirsakova, A., Nouza, K. and
Pokorna, Z., Folia Biologica (Praha), 11: 364, 1965.
39. Bishop, D., Narbaitz, R. and Lessof, M., Developmental Biology,
1: 444,1961.
40. Katsh, S., Int. Arch. Allergy, ~: 241, 1960.
41. Maruta, H. and Moyer, D.L., Fertil. Steril., ~: 649, 1967.
42. Katsh, S. and Katsh, G.F., Fertil. Steril., 12: 522, 1961.
43. Voisin, G.A. and Toullet, F., Ann. Instit. Pasteur (Paris), 114:
727, 1968.
44. Pehle, W., J. Immunology, 34: 325, 1928.
45. Henle, W. Henle, G. and Chambers, L.A., J. Exp. Med., 68: 335,
1938.
46. Mancini, R.E., Bueno, M.P., Alonso, A., Fernandez Collazo, E.,
Scacciati, I.M. and Gonzalez, N., Fertil. Steril., 1970 (in
press)
47. Rao, S.S. and Sadri, K.K., J. Compar. Pathol., 70: 1: 1960.
48. Stevens, K.M. and Fost, C.A., Proc. Soc. Exp. Biol. Med.,
117: 125, 1964.
49. Hunter, A.G. and Hafs, H.D., J. Reprod. Fertil., 2: 357, 1964.
50. Menge, A.C. and Protzman, W.P., J. Reprod. Fertil., 11: 31,
1967.
51. Weil, A.J., Ann. N.Y. Acad. Sci., 124: 267, 1965.
52. Weil, A.J., J. Reprod. Fertil., ~: 25, 1967.
53. Shulman, S., Yantorno, C., Soanes, W.A., Gonder, M.J. and
Witebsky, E., Immunology, 1Q: 99, 1966.
54. Moyer, D.L. and Maruta, H., Fertil. Steril., 18: 497, 1967.
55. Katsh, S., J. Urol., 87: 896, 1960. --
56. Mancini, R.E., Andrada, J.A., Saraceni, A., Bachmann, A.E.,
Lavieri, J.C. and Nemirovsky, M., J. Clin. Endocr., 12: 859,
1965.
57. Mancini, R.E., Proc. 3rd Int. Pharmacology Meeting, ~: 75, 1966.
58. Mancini, R.E., Andrada, J.A., Guirlanda, A. and Pascale, R.,
Acta Europea Fertil., 1970 (in press).
59. Popiranov, R. and Vulchanov, V.H., Acad. Bulg. Sci., 12: 865,
1964.
60. Weil, A.J. and Rodenburg, J.M., Proc. Soc. Exp. Bioi. Med.,
105: 43, 1960.
61. Mancini, R.E., Gutierrez, O. and Fernandez Collazo, E., J.
Reprod. Fertil., 1970 (in press).
62. Rao, S.S. and Sadri, K.K., Proc. Sixth Int. Confer. Planned
Parenthood, New Delhi, p. 243, 1959.
63. Flocks, R., Bandhauer, K., Patel, C. and Begley, R.J., J. Urol.,
87: 457, 1962.
64. Hekman, A. and Rumke, Ph., Fertil. Steril., 20: 312, 1969.
65. Fernandez Collazo, E., Gutierrez, o. and Mancini, R.E., Fertil.
Steril., 1970 (in press).
66. Wilson, L., Proc. Soc. Exp. BioI. Med., 85: 652, 1954.
67. Rumke, Ph. and Hellinga, G., Amer. Clin. Path., 32: 357, 1959.
68. Rumke, Ph., Ann. N.Y. Acad. Sci., 124: 696, 1965.
69. Nakabayashi, N.T., Tyler, E.T. and Tyler, A., Fertil. Steril.,
12: 544, 1961.
70. Bandhauer, D., Urologica Internationalis, 21: 247, 1966.
71. Rao, S.S., Sadri, K.K. and Sheth, A.K., Proc. Symp. Proteins
Central Food Techno!. Res. Instit. Mysore, India, 1961.
72. Southam, A.L., J. Reprod. Fertil., 2: 458, 1963.
73. Weil, A.J., Lotsevalov, o. and Wilson, H., Proc. Soc. Exp.
BioI. Med., 92: 606, 1956.
74. Schwimmer, W.B., Ustay, K.A. and Behrman, S.J., Amer. Obst.
Gynec., 30: 192, 1967.
75. D'Almeida, M., Belaisch, J., Eyquem, A., Guillon, G. and Palmer,
R., Citedby Cohen, In Acta Europea Fertil., 1: 193, 1969.
76. Rumke, Ph., Proc. Roy. Soc. BioI., ~: 275,-1968.
77. Ppadke, A.M. and Padukone, K., J. Reprod. Fertil., 2: 163, 1964.
78. Feltkamp, T.E.W. and Kruyff, K., Ann. N.Y. Acad. Sci., 124:
702, 1965.
79. Andrada, J.A., Loyzaga, P. and Mancini, R.E., Fertil. Steril.,
1970 (in press).
80. Fjallbrant, B., Acta Obstet. Gynec. Scand., 48: 131, 1969.
DISCUSSION
A. STEINBERGER: Dr. Mancini, you have indicated that the testicular

antigen is primarily located in the spermatids or spermatozoa. As
you have so beautifully illustrated in vitro, fluorescent tagged an-
tibody localizes in the spermatozoa, particularly in the acrosome.
How do you explain then the finding that when testicular homogenates
are injected in vivo together with complete Freund's adjuvant you
not only get degeneration of spermatids and spermatozoa but also de-
generation of spermatocytes and spermatogonia? Do you think that
there is some immunologic cross reactivity between these cells?
MANCINI: The first cells which degenerate are the spermatids and the
spermatozoa in the seminiferous tubules, then the spermatocytes which
may appear in the lumen or are phagocytosed by the Sertoli cells.
What is most surprising is that also spermatogonial cells may exfo-
liate and degenerate, in most but not in all 6f the tubules. With
immunofluorescence techniques, the antigen appears located in the
acrosome but we do not know if the resolution of the immunohistochem-
ical technique is high enough to detect the precursors of the anti-
gens which may be contained in spermatocytes. This may be a partial
explanation. I do not know of any other unless one speculates on
possible damage in the sustentacular function of Sertoli cells.
JOHNSEN: Could you indicate how often you find good evidence that
immunologic factors are the cause of infertility in males?
542 R. E. MANCINI
MANCINI: I cannot answer your question because I am not yet en-

gaged in exploration of sterile patients with immunological tech-
niques. Various investigators have found spermagglutinins and sperm
immobilizing antibodies in the serum of infertile individuals which
will increase if an injury such as an inflammatory condition is pre-
sent in the testis or in the genital male tract. I would like to
stress the fact that nobody can explain yet whether or not there is
a causal relationship between antigenic antibodies and sterility in
men.
JIRASEK: Did you use oxidase labelled or just the fluorescent la-
belled antibodies for analysis of your sections?
MANCINI: We have used in the last years as label, horseradish per-

oxides as recommended by Pearse and Nakane. We saw that the horse-
radish peroxidase, at least in our systems, is very useful and we
are able to detect the presence of antigenic areas, which point to
the site of interaction with the antibody in human spermatozoa bet-
ter than the immunofluorescent techniques.
JIRASEK: I refer to the sections. Were the results with the perox-
idase identical to those with the fluorescein labelled antibodies?
MANCINI: We did not apply the horseradish peroxides as a label of

antibodies in the sections of human testis. If one applies this
technique to sections of the human testis, some lipids, located in
the Leydig and spermatogonial cells, may react. It is well known
that some lipids may have a peroxide activity, and in this sense the
reaction will give a false positive reaction.
PAULSEN: I think that the evidence with respect to the immunologic

factors in testicular disease presented by Dr. Mancini is exceedingly
important. Several years ago we described that after unilateral tes-
ticular biopsy (without any clinical evidence of an inflammatory re-
action) about 40% of the patients showed a significant reduction in
sperm count. Working with Dr. William Clarke's group, we have been
able to demonstrate that the sperm from some of these patients fol-
lowing a unilateral biopsy become "tagged" with fluorescent labelled
antibodies. This was not present prior to the biopsy. This demon-
strates that in some patients you can elicit an antigen-antibody re-
action by incising the seminiferous tubules.
MANCINI: In terms of potential antigenicity for antibody formation,

it is not necessary that a great amount of antigen should be liber-
ated at the site of the injury. Also, the reactivity of the immune
system of the individual is of importance.
ROSEMBERG: The number of patienta undergoing testicular biopsy is

considerable. What would be then the proportion of individuals who
become infertile after a testicular biopsy? Are we justified in ask-
ing for it?
PAULSEN: I must emphasize that this is a very transient phenomenon

in the majority of patients. It usually clears within 90 days. We
do not find a permanent reduction in spermatogenesis as a consequence
of testicular biopsy.
MANN: May I be allowed to make a general comment on the antigens and

antibodies in semen? We have to draw a very clear distinction be-
tween two classes of immunologically active factors, namely those
that are coming from the accessory gland secretions, and those that
a~e found in spermatozoa. As regards the first class, these can be
identified easily in the accessory secretions, and they can be ex-
plored by experiments performed directly on, for example, the pro-
static and seminal vesicle secretion. However, I am somewhat doubt-
ful about the true origin of antigens and antibodies which are sup-
posed to occur in spermatozoa. One of the biochemical peculiarities
of spermatozoa is the extraordinary ease with which they can be coat-
ed with various proteins and mucoproteins and indeed, as Weil has
shown, some of the so-called sperm antigens and antibodies have been
trapped by the spermatozoa from the accessory secretions. There is,
therefore, the possibility that already in the epidididymis, those
antigens and antibodies which we find in spermatozoa may have origi-
nated in the epididymal plasma. I do feel that progress on the true
origin of immunologically active factors in spermatozoa as such, can
only be made when testicular spermatozoa become available for that
kind of investigation.
MANCINI: Dr. Mann, you raise the question which appeared in Dr.
Weil's papers more than ten years ago, related to the existence of
intrinsic antigens and coating antigens. Our present studies are
confirming what you assumed. It is difficult to separate from the
surface of the human spermatozoon, in spite of repeated washing and
some other treatment, the coating antigen which depends on the semi-
nal plasma to see what are really the intrinsic antigens. However,
using immunohistochemical techniques, testicular spermatocele showed
a weak head cap reaction in contact with antihuman testicular anti-
bodies.
DICZFALUSY: Why do you only mention epididymal antigens? I believe

that in prostatic secretions, at least three highly specific proteins
have been found, which are not present in any other organ.
MANCINI: Yes, the prostate has a strong antigenicity as compared

with that of the epididymis. Dr. Rumke has found that one of them
is a lactoferrin, which is also present in milk. This antigen is
apparently responsible for the antigenicity of the prostate. This
of course does not exclude the existence of some other antigenic pro-
teins in this gland.
CHAPTER VII
SECTION 1: TREATMENT OF HYPOPITUITARISM
EFFECTS OF HCG, HMG, HLH AND HGH ADMINISTRATION ON TESTICULAR
FUNCTION
C. Alvin Paulsen, Duane H. Espeland and Edward L. Michals
Department of Medicine, University of Washington School

of Medicine, Division of Endocrinology, U.S. Public
Health Service Hospital, Seattle, Washington
Proper patient selection together with the use of purified

gonadotropic hormones having known biologic activity are of para-
mount importance when experiments are conducted to define the ef-
fects of follicle stimulating hormone (FSH) or luteinizing hormone
(LH) on human testicular function. This is particularly true when
assigning physiologic roles to these gonadotropins.
With regard to patient selection, males who have normal or el-

evated endogenous FSH and LH levels and present with either oligo-
spermia or azoospermia are unsuitable for such studies irrespective
of their testicular histology. This is due to the fact that the
involved etiologic and pathologic factors are largely unknown.
Although still not ideal, patients with markedly diminished

pituitary gonadotropin secretion are best suited for studies of
this nature. Included in this group of patients are males with the
genetic disorder, hypogonadotropic eunuchoidism (1) or males who
demonstrate in addition to gonadotropin deficiency diminished se-
cretion of one or more than one of the other pituitary tropic hor-
mones.
The human gonadotropic hormones which have been available for

study in experimental systems up to the present time have not been
optimal agents due to either their lack of purity or to the fact
that they contain both gonadotropic hormones, i.e. FSH and LH.
For example, human menopausal gonadotropic (HMG) preparations ex-
hibit FSH and LH potencies of varying ratios (2). Even the purest
preparations contain both FSH and LH. Also human chorionic gona-
dotropin (HCG), which behaves like luteinizing hormone (LH) in
stimulating the Leydig cells, possesses an unknown FSH or follicle-
547
548 C. A. PAULSEN, D. H. ESPELAND, AND E. L. MICHALS
stimulating factor (FSF) activity in certain animal systems (3,4).

Thus the response to HCG can not be directly equated to that which
might be obtained with human (HLH) administration.
Despite these limitations, certain working concepts have emer-

ged from clinical studies in patients with hypogonadotropic hypogo-
nadism which in the main correspond to the information derived from
animal studies where more pure pituitary gonadotropins of animal
origin have been used.
This report will summarize our experience with respect to the

testicular response following HMG and HCG administration. Prelim-
inary studies with HLH or human growth hormone (HGH) treatment will
also be presented.
A. EFFECT OF HMG ON THE IMMATURE GERMINAL EPITHELIUM
Although FSH and LH act in concert during normal puberty to

produce testicular maturation, we desired to determine whether or
not FSH could stimulate immature germinal epithelium. Therefore
HMG was administered to three hypogonadotropic patients as their
initial form of gonadotropic therapy (D.L., C.H., D.D., see Tables
1, 2, 3, 4). The details of these patients' clinical, hormonal and
testicular status before and after treatment have been documented
previously (5-7)0 The dose of HMG was adjusted in an attempt to
administer that dose of FSH which had been shown to stimulate ovar-
ian follicular growth and later to restore spermatogenesis after
hypophysectomy (8). At the same time we endeavored to keep the LH
dose low enough to avoid Leydig cell stimulation. This latter point
had no precedence since the minimal effective dose for LH action on
Leydig cell function is unknown. The daily FSH dose ranged from
27.3 (raised to 81.9 I.U. in patient D.L. after two weeks) to 148.6
I.U. and for LH was 42.5 to 261.2 IoU. Treatment was carried out
for five and one-half to six weeks. Essentially two types of re-
sponse were noted. In patient D.L. the seminiferous tubules re-
mained the same size and no proliferation of the immature germinal
epithelium occurred. (See photomicrographs published in reference
5) •
On the other hand, in patient C.Ho the mean seminiferous tu-

bular diameter increased from 58.6\.l±1.5 to 71.0 ~ ± 1.4 (5). How-
ever no proliferation of germinal cells was noted, e.g. control,
42.7 ± 2.1 cells per tubular cross' section and 49.1 ± 2.3 cells per
tubular cross section following HMG therapy. C.H.'s LH dosage was
the highest of the three patients, ranging from 130.6 to 261.2 I.U.
per dayo Furthermore his plasma testosterone level increased from
a control of 0.05tJ.g% to O.l1~g%, which suggests that his inter-
stitial cells were stimulated by the administered LH in his HMG
preparation. In patient D.D. two different sized tubules existed
before treatment (Table 1). This observation might be related to
TREATMENT OF HYPOPITUITARISM 549
Table 1. Seminiferous Tubule Diameter in Normal Adult Males and

in Males with Prepuberal Hypogonadotropic Hypogonadism.
Age Mean Diameler (~)

Patient Diagnosis
(yrs) of 10 Circular Tubules
RV 11 27 Normal male 189.6

RV 16 36 " 185.4
RV 20 50 ". 162.6
Group Mean
RV 28 30 " 166.6
171.8 ± 12.0
RV 32 46 " 174.0 (S.D.)
RV 38 24 " 156.6
RV 53 41 " 168.0
Hypogonadotropic
M.B. 23 55.4
eunuchoidism
B.S. 25 " 57.3
J.P. 20 " 62.7
Group Mean
F.H. 19 " 59.3 62.1 ± 8.8
(S.D.)
C.H. 17 " 58.6
D.O. 22 Craniopharyngioma a 79.5 b
(96.1)
Idiopathic selective
E.J. 28 66.6
pituitary deficiencyc
a. Patient had TSH and gonadotropin deficiency. ACTH secretion

was presumed to be normal by testing; HGH secretion presumed
normal because of his normal height. He was on thyroid replace-
ment therapy.
b. Two populations of tubules in biopsy specimen. Smaller tubules

used for group mean calculations.
c. Sella turcica as well as suprasellar area was normal. Decreased

HGH, FSH, LH secretion by radioimmunoassay or bioassay testing.
Presumed normal TSH and ACTH secretion by appropriate testing,
e.g., RAI uptake, PBI, response to metapirone, 17-ketogenic
steroid excretion.
550 C. A. PAULSEN, D. H. ESPELAND, AND E. l. MICHALS
Table 2. Hypogonadotropic Hypogonadism Pretreatment Studies
Urinary Gonadotropins Urinary Plasma

Patient Age GGA FSH Estrogens Testosterone
D.L. 20 <0.7 <0.2

J.P. 20 <0.7 <0.2
M. S. a 20 <3.8, 1.6 <0.2 0.05
H.G. 24 < 0.7, 0.9 0.02, 0.02,
0.15
G.G. a 19 <0.7
D.D. 21 <0.7 <0.2
P.N. 16 <0.7 ,,-0.2
Hormonal data before treatment. GGA refers to a general go-

nadotropin assay method which measures the combined effect of FSH
and LH. (Normal adult male range 0.7 - 3.8 mg. -eq. per 24 hours).
Urinary estrogens determined biologically. (Normal adult male
range 0.3 - 3.0~g. -eq. estradiol benzoate per 24 hours). Plasma
testosterone by double isotope derivative method. (Normal adult
range 0.28 - 1.44 ~. %.) a Patients with anosmia.
previous therapy with methyl testosterone, since exogenous androgen

can stimulate the immature seminiferous tubule to some extent. On-
ly general gonadotropin assays were conducted prior to treatment
so that the precise status of his endogenous FSH and LH levels is
unknown. Following HMG therapy the mean seminiferous tubular di-
ameter of D.D. increased to 109.3 ± 4.5~. At this time an in-
crease in germinal cell numbers was also noted, but maturation had
not progressed beyond the cell type noted before treatment, i.e.,
the pachytene spermatocyte.
These data could be interpreted in the following manner: FSH

by itself does not affect the immature seminiferous tubule unless
Leydig cells are stimulated by concomitant administration of LH or
unless the seminiferous tubules have been previously exposed to
endogenous or exogenous testosterone. de Kretser (9) noted early
seminiferous tubular maturation and an increase in plasma testos-
terone levels in a patient with hypogonadotropic eunuchoidism when
the FSH/LH dosage was 200/500 I.U. three times weekly for five
months. Rosemberg et alo (10) in their studies on prepuberal males
with cryptorchidism noted germinal cell stimulation and early Ley--
dig cell maturation following HMG therapy of equivalent or higher
FSH and LH doses.
-i
;;0
~
m
Z
-i
o-n
::I:
-<
~
Table 3. Summary of Histologic Findings in Hypogonadotropic Hypogonadism before and after HCG
o
~
Therapy ::::j
c
~
~
Duration (II
Most Advanced Germ Cell Type ~
Patient Previous Therapy HCG Therapy
(weeks) Before HCG After HCG
D.L. HMG (7 wks) 23 Immature germinal Early differentiation

epi thelium of spermatogonia
J.P. None 59 " Sd Spermatid
D.O. 1) intermittent methyl 25 Pachytene Sd Spermatid
testosterone (? yrs) spermatocyte
2) HMG (6 wks)
P.N. None 56 No biopsy Pachytene
spermatocyte
E.J. 1) "Pi tui tary" shots 10 Pachytene Sc spermatids
(1 mo duration, age 14) spermatocyte
2) HGH therapy (15 wks)
01
~
Table 4. Studies to Induce Spermatogenesis in Hypogonadotropic

Hypogonadism
HCG HCG-HMG
Duration Sperm Duration Sperm Highest
Patient Age of treatment Count of treatment Count Sperm
(weeks) mill/ml (weeks)C mill/ml Count
J.P. 20 59 Azoo 7 1-2/hpf 39 b

a
M.S. 20 16 59 71 b
H.G. 22 85 (42) (42)
a
G.G. 19 78 Azoo 5 l/hpf 1-3/hpf
D.D. 22 25 Azoo 15 1-2/hpf 14.7
E.J. 28 30 d Azoo 13 0.3 48
P.N. 16 56 Azoo 19 occasional dead
sperm
In patient H.G., the bracket surrounding the sperm count in-

dicates that this was the only specimen obtained. a. Patients with
anosmia. b. Patients who were married and achieved a pregnancy~ c.
Duration in weeks refers to that interval between onset of treatment
and the first appearance of sperm in the ejaculate. d. During the
last twenty weeks HGH therapy was added.
B. EFFECT OF HCG ON THE IMMATURE TESTIS
Seven patients received HCG therapy at a dosage of 2000-5000

I.U. three times weekly for 16 to 85 weeks. Four patients (D.L.,
J.P., P.N., G.G.) exhibited the classic form of hypogonadotropic
eunuchoidism, ioe., no discernible evidence of effective endogenous
gonadotropin secretion. Two of the seven patients (M.S., H.G.) were
considered as variants of the syndrome resembling the subcategory
"Fertile Eunuch." Both men were devoid of androgen effects and dem-
onstrated very low plasma testosterone levels (see Table 2). On the
other hand, urinary FSH excretion was at the lower limit of the nor-
mal adult male range on one occasion in patient M.S. Patient H.G.
exhibited a normal urinary GGA (general gonadotropin assay) titer on
one of two occasions while his control testicular biopsy specimen
revealed seminiferous tubular stimulation. The mean tubular
diameter was 116.4 ~ ± 18.6 and spermatogenesis had progressed to
the Sd spermatid level. The remaining patient, D.O., has been de-
scribed before (Table 1 and section A).
In the patients who demonstrated positive evidence of endogen-

ous FSH secretion prior to treatment, therapy with HCG alone pro-
duced full spermatogenesis including spermatozoa in the ejaculate.

This response occurred within 16 and 85 weeks, respectively. Pa-
tient H.G. was unable to submit a seminal fluid specimen until the
85th week of treatment. Therefore we are uncertain as to the exact
time that full spermatogenesis developed.
The remaining five patients were still azoospermic at the com-

pletion of HCG therapy alone. Treatment lasted 25 to 78 weeks (Ta-
ble 4). Despite the azoospermia, examination of the testicular bi-
opsy specimens in D.L., J.P., D.O., and E.J. revealed that sperm-
atogenesis had been stimulated by HCG therapy in each patient where
a pre-HCG biopsy was available for comparison (Table 3, Figs. 2,3).
The Question as to why HCG therapy by itself should initiate sperm-
atogenesis in these hypogonadotropic "prepuberal" patients remains
unclear. The fact that sufficient 'endogenous FSH ~ecretion existed
in patients M.S. and H.G. affords a plausible explanation as to why
Leydig cell function plus full spermatogenesis was achieved by only
HCG therapy. However no such evidence was present in patients D.L.
and J.P. It may be that the "FSF" activity present in HCG permits
germinal cell stimulation. On the other hand, it may be that the
endogenous testosterone secretion initiated by HCG therapy has a
definite but limited action with respect to stimulating spermato-
genesis. Lastly it is possible that LH acts directly on the ger-
minal epithelium. Definitive conclusions will have to wait until
purified HLH is used in similar patients.
C. EFFECT OF COMBINED HCG-HMG TREATMENT
Five patients received combined HCG-HMG therapy of varying

dose schedules when solo HCG treatment of 25 to 78 weeks was still
associated with azoospermia (Table 4). The treatment schedule is
listed below:
HCG FSH (HMG)

(I.U., t.i.w.) (I.U.)
J.P. 2000 64.5 - 74.3 daily
D.O. 2000-4000 63.0 - 74.3 daily
G.G. 4000 150 t.i.w.
E.J. 4000 75 t.i.w.
P.N. 4000 75 t.i.w.
The addition of HMG (FSH) to the therapeutic regimen was suc-

cessful in four out of the five patients as measured by the appear-
ance of spermatozoa in their seminal fluid (Table 4). It required
from 5 to 19 weeks of combined HCG-HMG therapy for this event to
occur. Seminal fluid specimens were submitted for analysis once
weekly. The varying HMG (FSH) dosage did not appear to affect the
response to treatment in this limited study. HCG-HMG administra-
tion was continued to a maximum of four years (J.P., see reference
6). The highest seminal fluid sperm concentration was achieved in
E.J. (48 mill/ml) after a total of 60 weeks 1 of combined therapy.

In J.P., the highest sperm count noted was 39 mill./ml. HMG ther-
apy was discontinued in P.N., D.D. and G.G. after 34, 45 and 32
weeks, respectively.
It is difficult to explain why combined HCG-HMG therapy did

not produce numerically greater sperm counts. Fundamental infor-
mation pertinent to this point is lacking. That is, at what point
in time do males attain normal sperm production during their puber-
al process? Until these data are obtained, we have to consider the
following possibilities: 1) treatment did not last long enough,
2) the proper FSH/LH ratio was not employed, 3) HCG exerts an action
which is qualitatively different from that exerted by HLH, 4) the
testis as well as the pituitary-hypothalamic system is defective in
patients with this syndrome. (11)
It is also possible that the mUltiple biopsies employed to

document the response to treatment might have influenced sperm pro-
ductiono However, no testicular biopsy was performed in patient
G.G.
Only two patients were married while rece1v1ng treatment (J.P.,

M.S.). J.P.'s wife became pregnant twice and delivered two normal
male infants. M.S.'s wife became pregnant once and delivered a
normal female infant.
D. EFFECT OF HLH ON LEYDIG CELL FUNCTION
The dose of administered HLH required to produce normal Leydig

cell function has not been established. Therefore we initiated
studies in our patients with hypogonadotropic eunuchoidism in an
effort to obtain this information. Data from our HMG therapy ex-
perience (see discussion patient C.H.) suggested that the effective
HLH dosage might approximate 200 I.U. per day.
M.S. (Tables 2,4) who had received HCG for over three years
was selected for study. His current HCG dose was 2000 I.U. three
times weekly and his sexual maturation was essentially complete.
Just prior to starting HLH and twenty-four hours after his last
dose of HCG blood was sampled for plasma testosterone assay (see
Table 5). The value was 1.2\-l.g 'Y. (normal adult male range 0.28 -
1.44 ~g % (12» which confirmed our clinical judgment that he was
receiving optimal HCG therapy.
During the four weeks of HLH treatment (400 IoU. three times
weekly) M.S. stated that he experienced vasomotor symptons, e.g.
hot flashes, on those days he did not receive his medication. Li-
1. There were two periods of therapy, 28 weeks and 32 weeks, witha
one-year interval between.
xl
~
~
m
~
o
"T1
:t:
Table 5. Comparison Between HCG and HLH Administration in terms of Leydig Cell Function -<
(Patient M.S.)
o
"
~
Duration Time Interval Serum Plasma ~
~
of Therapy Between Treatment HLH Testosterone ( T) u;
Hormone Dosage Schedule at Indicated and Blood Sampling Levels c Levels d ~
Dosage (hrs) (mIU/mt) (~g 10)
HCG a 2000 I.U. t.i.w. 43 wks 24 1.2

b
HLH 400 IoU. t.i.w. 4 wks 12 9.4 0.23
HLH 400 IoU./d. 4 d 4 33.0 0.34
HLH 200 IoU./d. 4 d 8~ 6.8 0.11
HCG 2000 I.U. t. i .w. 3 wks 24 0.63
a. HCG, Ayerst Laboratories; b. HLH-LER 856, lot A-l, 200 I.U./vial (VPW) , kindly supplied by
the National Pituitary Agency; c. Serum HLH assayed by radioimmunoassay method of Midgley
(16); d. Plasma testosterone levels determined by a modification of the displacement
technique of Nugent (17).
01
01
01
556 C. A. PAULSEN, D. H. ESPElAND, AND E. L. MICHALS
bido was reported to be normal but his sexual potentia had decreased
slightly. Twelve hours after an injection of 400 I.U. HLH, his se-
rum HLH level was 9.4 m.I.U./ml. but his plasma testosterone was
0.23 ~g %, which is below the normal male range. His plasma testos-
terone increased to 0.34 ~g % after four days of 400 I.U. HLH per
day and dropped to 0.11 ~g % after four days of 200 I.U. per day.
Serum LH levels were 33 m.I.U./ml four hours after 400 I.U. of HLH
and 6.8 m.I.U./ml eight and one-half hours after 200 I.U. He then
returned to HCG therapy at his previous dose schedule. Androgen
withdrawal symptoms disappeared one week after the resumption of
HCG therapy, and his plasma testosterone titer was 0.63~g % two
weeks later.
These preliminary data indicate that for optimal maintenance

of Leydig cell function in patients without endogenous LH secretion,
HLH therapy using saline as the diluent will have to be administered
daily at doses slightly above 400 I.U. These observations as to
daily dose requirements are in keeping with the recent data on half-
life studies for HCG and HLH (a fast and slow component for each:
t 1/2 HCG = 11 hr, 23 hr; t 1/2 HLH = 21 min, 235 min). (13,14,15)
E. EFFECT OF HGH ON TESTICULAR FUNCTION
It has been suggested that when growth hormone and gonadotropin

deficiency exist together the testes of prepuberal males are less
developed than when only gonadotropin deficiency is present. (18)
If this is true, HGH may exert a stimulatory action on the testis.
To evaluate this hypothesis we studied the effect of HGH ther~

apy on patient E.J.'s testicular function (Table 1). This patient
demonstrated deficient HGH and gonadotropin secretion with intact
thyroid and adrenal cortical function. Testicular biopsy specimens
were obtained prior to and following HGH2 therapy (5mg thrice week-
ly). Before treatment, examination of his biopsy specimen revealed
an immature testis except for an occasional seminiferous tubule con-
taining maturing germinal cells. The most advanced cell type ob-
served was the pachytene spermatocyte (Fig.l). The mean tubular
diameter in E.J.'s control biopsy specimen was comparable to that
noted for patients with presumably normal HGH but deficient gonado-
tropin secretion (Table 1), 66.6 ~ vs 62 ± 8.8~. After fifteen
weeks of HGH therapy his biopsy specimen was unchanged (Fig. 2),
including the mean tubular diameter (65.9 ~). At this time his
plasma testosterone level was 0.04 ~g %.3 Since a control testos-
terone titer was not obtained, HGH was discontinued for three weeks
2. Kindly supplied by the National Pituitary Agencv. Bioassay of
this HGH preparation revealed the presence of 2.0 I.U. HLH and 4.0
USP units Prolactin per mg. No FSH activity was detected.
3. Performed by Dr. Hortense Gandy using the double isotope
derivative method.
Fig. 1. (x 63.) Testicular biopsy specimen before HGH. Semini-

ferous tubules contain mainly an immature germinal epi-
thelium with an occasional germ cell undergoing differen-
tiation.
Fig. 1A (x 160.) E.J. before HGH, higher magnification. Note

pachytene spermatocytes.
Fig. 2. (x 63.) E.J. testis specimen after HGH. Essentially no

change in the number of cells undergoing maturation.
Tubules are of the same size as before HGH.
Fig. 3. (x63.) Testicular biopsy specimen following HCG therapy

Definite increase in tubular diameter. Spermatogenesis
has been stimulated .
at which time his plasma testosterone level was 0.02 ~g %. His

body height increased 5/B" during HGH therapy. Thereafter his
treatment consisted of HCG for ten weeks, then combined HGH and HCG
for twenty weeks and finally HCG plus HMG for seven months. His
third testicular biopsy specimen obtained after his combined HCG-
HGH therapy (Tables 3 and 4) revealed active spermatogenesis (Fig.
3) which had progressed to the Sc spermatid stage. The mean tubu-
lar diameter had now increased to 98.8~. Maturation of his inter-
stitial cells into Leydig cells was evident and his plasma testos-
terone ti ter was 1. 17 ~ g %.
There was no evidence in this patient that HGH had any effect
on his testicular status. However these results do not reject the
possibility that HGH might supplement the action of FSH and LH dur-
ing testicular maturation in normal puberty.
ACKNOWLEDGEMENTS
These studies supported in part by NIH grants AM05161 and

AM05436 and AEC contract AT(45-1)-2225-T6.
REFERENCES
1. Paulsen, C.A., The Testis, In Textbook of Endocrinology, ed.

Williams, R.H., 4th edition, W.B. Saunders Co., Philadelphia,
196B.
2. Rosemberg. E., In Gonadotropins 1968, ed. Rosemberg, E.,
Geron-X, Inc., Los Altos,Calif. p 426, 1968.
3. Crooke, A.C. and Butt, W.R. J. Obstet. Gynaec. Brit. Corum.,
~: 297, 1959.
4. Albert, A., J. Clin. Endocr. ~: 1504, 1969.
5. Paulsen, C.A., ed., Estrogen Assays in Clinical Medicine,
University of Washington Press, Seattle, 1965.
6. Paulsen, C.A., In Proceedings of the Sixth Pan American Congress
of Endocrinology, Mexico City 1965, Amsterdam, Excerpta Medica
Foundation, p. 212, 1966.
7. Paulsen, C.A., In Gonadotropins 1968, ed. Rosemberg, E.
Geron-X, Inc., Los Altos, Calif. p 491, 1968.
B. MacLeod, J., Pagianos, A. and Ray, B., Fertil. Steril., 12;
7, 1964.
9. deKretser, D.M., Thesis, M.D., Monash University, Melbourne,
Australia, 1969.
10. Rosemberg, E., Mancini, R.E., Crigler, J.F. and Bergada, C.,
In Gonadotropins 1968. ed. Rosemberg, E., Geron-X, Inc.,
Los Altos, Calif., p 527, 1968.
11. Bardin, C.W., Ross, G.T., Rifkind, A.B., Cargille, C.M. and
Lipsett, M.B., J. Clin. Invest., 48: 2046,1969.
12. Paulsen, C.A., Gondon, D.L. Carpenter, H.M., Gandy, H.M. and
Drucker, W.D., Recent Progr.Horrnone Res., 24: 321, 1968.
13. Kohler, P.O., Ross, G.T. and Odell, W.D., J. Clin. Invest.,
!iJ...: 38, 1968.
562 C. A. PAULSEN, D. H. ESPELAND, AND E. l. MICHALS
14. Yen, S.S.C., Llerena, 0., Little, B. and Pearson, O.H.

J. Clin. Endocr., 28: 1763, 1968.
15. Rizkallah, T., Gurpide, E. and Vande Wiele,R.L., J. Clin.
Endocr., 29: 92, 1969.
16. Midgley, A.R. Jr., Endocrinology, 22: 10, 1968.
17. Mayes, D. and Nugent, O.A., J. Clin. Endocr., 28: 1169, 1968.
18. Albert, A., Underdahl, L.O., Green, L.F. and Lorenz, N.,
Proc. Staff Meeting Mayo Clinic 28: 698, 1953.
DISCUSSION
MACLEOD: Dr. Paulsen, in your studies with long-continued adminis-

tration of HCG in eunuchoid patients in whom you observed some degree
of maturation of spermatogenesis, have you taken into account the
FSH-like activity contained in HCG?
PAULSEN: Yes, we have to consider that HCG has several effects on

the germinal epithelium. One is by stimulating endogenous testos-
terone secretion. The other may relate to a direct action on the
germinal epithelium by the FSH property of HCG. What we need is
comparable data as to the effect of HLH on the immature testis.
ROSEMBERG: The amount of FSH activity intrinsic to the HCG molecule

could be estimated and this will be negligible in the studies reported
here.
DICZFALUSY: Making an approximation, I think it is fair to say that

in most HCG preparations the FSH-like activity corresponds to 1 IU
for every 1,000 IU of HCG. Hence, this amount of FSH will not really
have any effect.
ROSS: Dr. Paulsen, when you report detecting no follicle-stimulating

hormone activity in the urine of thes.e eunuchoidal patients, what
quantity of urine in relationship to the 24-hour output have you in-
jected per animal? What is the nature of the extract which you used:
fraction A, B, C? I ask these questions to emphasize the difference
between "failure to detect" and "total absence" of activity.
PAULSEN: I tried to be careful by stating that the urinary gonado-

tropins were "undetectable" as opposed to "absent". In general we
inject a l2-hour aliquot per assay animal. We have not defined the
small amounts of urinary gonadotropins which may be present in these
patients. All we can state is that there is less gonadotropin ex-
creted in these patients than is present in normal adult males. We
use fraction A for our assays.
EFFECT OF URINARY FSH AND LH ON THE TESTICULAR FUNCTION IN HYPO-
GONADAL PATIENTS
Roberto E. Mancini
Centro de Investigaciones sobre Reproduccion.

Facultad de Medicina, Paraguay 2155, Piso 10,
INTRODUCTION
The effects of human chorionic (HCG), pregnant mare serum (PMS)

or other animal pituitary gonadotropin preparations on the develop-
ment of spermatogenesis in prepuberal and postpuberal cases of hy-
pogonadotropic hypogonadism have generally been disappointing (1-5).
Although it has long been admitted that large doses and long term
administration of HCG clearly stimulate the development and function
of Leydig cells together with the appearance of secondary sexual
signs, some discrepancies still remain with regard to its effect on
the germinal epithelium and connective tissue structures of the tu-
bular wall. It seems that only those cases with previous histologi-
cal evidence of advanced meiotic process may respond successfully
with production of spermatozoa (6,7). On the other hand, gonado-
tropins of animal origin seem to be able to stimulate weakly both
Leydig and germinal cells; this restricted action may be imputable
to the long known induction of antigonadotropins (1,3). However,
results obtained in similar hypogonadal conditions seem to be more
encouraging, with the use of human urinary menopausal (HMG) and pi-
tuitary gonadotropins (HPG) administered singly or combined with
HCG. Their human origin and both FSH and LH activity are the main
advantages of these preparations. It is now generally agreed that
HMG or HPG alone may effectively stimulate the spermatogenic process
in hypogonadotropic subjects, but that completion of spermatogenesis
is achieved when full development of Leydig cells is produced by the
addition of HCG (8-15). Consequently, the contention at present is
in favor of the simultaneous need for FSH-LH and an excess of LH for
the development of interstitial and seminiferous tubule structures.
In spite of this considerable progress in the understanding of
563
564 R. E. MANCINI
the hormonal regulation of the spermatogenic process several points

still remain to be clarified: a) The separate action of different
doses and lengt~of treatment of highly purified preparations of
FSH and of LH. b) The effect of the successive or simultaneous ad-
ministration of both hormones in different ratios. c) Cell stages
reached by germinal and Sertoli cell populations in the qualitative
progression of the spermatogenesis induced by FSH or LH. d) The
extent to which Leydig cell differentiation and function is neces-
sary for germinal, Sertoli cell and tubular wall structure develop-
ment. e) Whether regression of peri tubular hyalinization is under
the direct control of FSH or LH or is correlated with Leydig or
germinal epithelium stimulation.
This rather ambitious program of study was made possible with

the collaboration of Dr. P. Donini (Serono Foundation, Rome) who
providedche purified preparations of human urinary FSH and LH,
Ors. o. Vilar and A. Perez Lloret, who performed histological stu-
dies and Drs. C. Bergada, J. Reforzo Membrives, A. Arguelles, M.
Molosnick and J.C. Lavieri (Buenos Aires School of Medicine) who
carried out the endocrine studies of the patients. In the present
report, emphasis will be placed on the microscopical changes of the
gonad. More detailed results concerning clinical data, hormonal
assays and electron microscopic observations of testis will be pub-
lished elsewhere.
The specifications of the FSH and LH urinary preparation were

as follows: FSH: 1255 I.U./mg of the 2nd. IRP-HMGj LH contaminant:
3.3 I.U./mg of the 2nd. IRP. LH: 1025 I.U. per ampoule/2nd. IRP.
FSH contaminant: 39.9 I.U./ampoule. The strength of both prepara-
tions was verified by the Steelman and Pohley and the O.A.A.D. me-
thods and the radioimmunoassay according to Donini et ale (16). As
it was estimated that 74 + 4-5 days are required for the develop-
ment of spermatogonia to the level of spermatozoa (17), FSH and/or
LH were given to our groups of patients for three or four months.
Biopsies of the testes and urinary steroid determinations were made
before and at the end of each course of treatment. The evaluation
of the histological changes induced by both gonadotropins was based
on cell count according to previous work (18,19), where a detailed
description of the various types of spermatogonial, Sertoli and
Leydig cells was made. However, as the aim of the present study
was to establish the effect of these hormones on the cell population
of the three main phases of spermatogenesis, the spermatogonial
cells (mitotic included), the spermatocytes and the spermiogenic
cells (spermatids and spermatozoa) were considered separately. The
estimation of the variations in the relative number of germinal and
Sertoli cells was made by cell count in 20 approximately circular
transverse sections of seminiferous tubules for each specimen. Cal-
culations of the average number of each cell type per one tubular
cross section were made and statistically verified by the "t" test.
The diameter of the seminiferous tubules, excluding the basement
membrane, was also measured in 30 transverse sections, expressed as

an average and statistically verified by the "t" test. Changes in
the wall of seminiferous tubules were estimated as disappearance of
peritubularhyalinmaterial and its replacement by normal fibrillar
structures. With reference to the intertubular spaces, the fibro-
blast-like cells as precursor Leydig cells and the immature, mature
and degenerating types of Leydig cells were estimated applying a
differential count over 200 cells of this cell population and sta-
tistically verified by the binomial distribution (20).
In the following text the results obtained in the induction or

recovery of spermatogenesis in prepuberal and postpuberal hypogonad-
al patients treated with urinary FSH and/or LH will be summarized.
These findings will also be compared with those published by others
who have used preparations of HMG or HPG containing different ratios
of FSH and LH activities.
I - FSH AND LH IN THE INDUCTION OF SPERMATOGENESIS IN PREPUBERAL

HYPOGONADOTROPIC HYPOGONADISM
Several reports have dealt with the action of gonadotropin

preparations containing variable FSH-LH ratios in cases with clas-
sical signs of eunuchoidism, undetectable or low levels of urinary
gonadotropins and a variety of steroids. Testicular biopsies showed
in these cases the very well known histology of an immature or pre-
puberal gland i.e., tubules of small diameter containing only sper-
matogonia and various degrees of immature Sertoli cells; occasional-
ly some leptotene, zygotene or pachytene spermatocytes were seen.
Correspondingly the tubular wall is thin and simple and does not
show the complex organization of the well developed tubules of nor-
mal adult testes as is seen under the light and electron microscope
(21,22). Different spermatogonial cell types may appear as well as
some unidentified intermediate forms, but few cytological studies
have been made at the electron microscope level (23,24). It is dif-
ficult to explain the finding, in some of these eunuchs, of a mod-
erate peri tubular hyalinization which mimics that present in post-
puberal hypogonadism (23,24). Finally the intertubular spaces are
filled with so-called fibroblast-like cells, known as the precursors
of immature and mature Leydig cells (19); however, the erratic pre-
sence of the latter cells cannot be excluded in all cases.
Based on the existence of an infantile testis and on the theo-

retical absence of gonadotropins, the eunuchoidal condition has been
considered a suitable one to verify the effect of exogenous gonado-
tropins in the initiation or induction of spermatogenesis. There-
fore, in previous publications, an HMG preparation with an FSH-LH
ratio equal to unity or favoring FSH, given at a daily dose of 24 to
60 I.U. of FSH for several months has induced, in some cases of eu-
nuchoidism, an increase in tubular diameter, differentiation of Ser-
toU cells, increase in the number of spermatogonial cells and de-
566 R. E. MANCINI
velopment of a few spermatocytes; no changes were observed in the

preexistent peri tubular hyalinization, no development of Leydig
cells and no ostensible modification in the urinary androgen assays
(11). With 12.5 I.U. of FSH contained in another HMG preparation
of unknown LH activity, administered three times weekly for three
months, no modifications could be observed in the testis biopsy,
in the steroid hormonal assays or in sexual signs (9). In some
other cases, daily doses for six and one half weeks of HMG (27.3 to
81.9 I.U. of FSH and 42.5 to 127.5 I.U. of LH) had no effect on
testicular histology or steroid levels (8). Doses of 225 I.U. of
FSH per week (HMG containing equal amounts of FSH and LH) for 17
weeks produced no results in any of the histological, clinical or
hormonal parameters (12) and the same is true of another case trea-
ted for 22 weeks with a weekly dose of 750 I.U. of FSH contained in
an HMG preparation in which LH activity was not known (13). In an-
other two cases 180 I.U. of FSH (HPG containing an FSH-LH ratio of
1:2) given three times a week for four or five months, resulted in
an increase in tubular diameter and of spermatogonial and Sertoli
cells, but no changes were seen in tubular wall thickness or in
Leydig cell development (7).
In our studies the separate, successive and simultaneous ef-

fects of FSH and LH on the testes of similar cases of hypogonado-
tropic hypogonadisms were considered. Three groups of eunuchs each
of three patients, aged from 18 to 25, were treated with the puri-
fied urinary FSH and LH already described. In the first and second
groups the separate action of both hormones was investigated.
In the first group a weekly dosage (divided into three doses)

of about 300, 500 and 800 I.U. of FSH (LH contaminant in the total
dose given varied from 9 to 22 I.U.) was given to the first, second
and third patients respectively for 12, 14 and 16 weeks. By com-
paring the biopsies, irrespective of the dose and length of time of
administration of the hormones, an increased number in the popula-
tion of Sertoli cells was observed with a marked predominance of
the mature forms over the precursor cell types. No significant
changes were detected in the diameter of the seminiferous tubules
or in spermatogonial cells and primary spermatocytes. The preexis-
tent normal structures of the tubular wall or the peri tubular hya-
linization remained the same as did the fibroblast-like cells of
the intertubular spaces. In the second group of patients, treat-
ment was scheduled with approximately the same amount of LH (FSH
contaminant in the total dose varied from 123 to 383 I.U.) and dura-
tion of treatment in each one of the three patients. As with FSH,
an increased number of Sertoli cells with a higher proportion of
the mature types was seen in two cases, being less clear in the re-
ma1n1ng one. Only in the patient who received the highest dose of
LH was there in addition a slight increase in the number of sperma-
togonial cells, which was coincident with an increased population
of fibroblast like cells and the appearance of few immature and

mature types of Leydig cells. In no case could clear modifications
of seminiferous tubule diameter be observed nor development of tu-
bular wall structures or regression of peri tubular hyalinization.
In a following step of our studies the successive effects of

both hormones was investigated. Two patients of the first group
who had been treated with FSH at low and high doses were subse-
quently treated with the same amount of LH during the same period
of time, 12 and 16 weeks respectively. Similarly, another two pa-
tients of the second group treated with low and high doses of LH
were subjected without interruption to another course of the same
doses of FSH during the same length of time. By comparing in the
same patient the biopsy after LH with the previous one after FSH,
the following was seen: a) Slight increase in the number of sper-
matogonial cells. b) No change in the number of spermatocytes but
more maturation of Sertoli cells. c) Slight increase in the tubu-
lar diameter, decrease in the peri tubular hyalinization and devel-
opment of immature and mature types of Leydig cells. d) No change
in the level of urinary steroids. These effects were specially
seen in the case given the higher dose of LH. In the patients of
the second group treated with FSH preceded by LH the comparison
of both biopsies showed: a) No changes in spermatogonial or sper-
matocyte cells. b) Increased maturation of Sertoli cells. c) No
major changes in the diameter of seminiferous tubules, in the peri-
tubular hyaline material or in the character of tubular wall struc-
tures. d) Disappearance of the already developed Leydig cells and
reappearance of fibroblast like cells. e) Levels of urinary ster-
oids remained the same.
In the third group of patients, the simultaneous action of

both gonadotropins was tried. The first patient was given a weekly
dose of 800 I.U. of FSH and 200 I.U. of lH for 16 weeks. The sec-
ond subject during the same period received 400 I.U. per week of
both FSH and LH. A weekly dose of 800 I.U. of LH plus 200 I.U. of
FSH was administered to the third patient. When comparing the bi-
opsies, the first case showed similar results to those obtained in
the patients of the first group treated with FSH alone. In the
second patient there was a moderate increase in spermatogonial
cells, development of earlier stages of primary spermatocytes, mat-
uration of Sertoli cells, increase in tubular diameter together
with slight regression of peritubular hyalinization, appearance of
a small number of immature and mature Leydig cells but no striking
modification in the urinary steroids. The third patient showed a
higher increase in spermatogonial cells, a marked maturation of Ser-
toli cells, development of primary spermatocytes up to the pachy-
tene stage, increase in tubular diameter, a slight development of
tubular wall connective tissue structures and higher levels of ur-
inary steroids o
568 R. E. MANCINI
II FSH AND LH IN THE RECOVERY OF SPERMATOGENESIS IN POSTPUBERAL

HYPOGONADOTROPIC HYPOGONADISM
Previous reports have pointed out the simultaneous need of both

HMG and HCG preparations for the recovery and maintenance of sperma-
togenesis in hypogonadotropic hypophysectomized patients (15,25).
The influence of the amount of pituitary gland removed, the persis-
tence of some pituitary functions and the time which elapsed after
operation upon the extent of the regression of germinal epithelium
and Leydig cell population, have been discussed (15,25). It was
also reported that patients showing in their control biopsies not
only Sertoli and spermatogonial cells but also primary spermatocytes
may respond better than those having only the first two types of
cells (25). Moreover HMG, with a FSH-LH ratio equal to unity in
weekly doses of 200 to 300 I.U. may recover spermatogenesis up to
the spermiogenic phase, while the addition of HCG at a weekly dose
of 2000 I.U., may induce production of spermatozoa after three or
four months treatment (15,25).
In the present report the separate or successive effect of pu-

rified preparations of urinary FSH and LH were studied in two adult
hypophysectomized patients aged 28 and 33. The operations were per-
formed seven and ten months respectively before the testicular bi-
opsies and hormonal assays were made. Histological study showed
small hyalinized seminiferous tubules, with only spermatogonial
cells, mature Sertoli cells with signs of nuclear atrophy and ab-
sence of Leydig cells. Urinary gonadotropins and steroids were un-
detectable or very low. The first patient was initially treated
with a weekly dose of 800 to 1000 I.U. of FSH for 16 weeks; after
withdrawal of the hormone the same amount of LH was given during
the same period. The second patient was similarly treated but first
with LH and afterwards with FSH.
By comparing biopsies taken before and after treatment, no

change was visible in any of the testicular structures of the first
patient treated with FSH; the exception being that the morphological
nuclear abnormalities previously observed in Sertoli cells tended
to disappear. No change was induced in the level of urinary ster~
oids. In contrast, the subsequent course of LH in the same patient
provoked a slight increase in the spermatogonial cells, appearance
of primary spermatocytes, morphologic recovery of nuclei of Sertoli
cells, partial regression of hyalinization, slight enlargement of
tubules and development of a few immature and mature Leydig cells.
No substantial change was detected in the level of urinary steroids.
In the second patient the first course of treatment with LH induced
similar changes as those described in the first patient but a sub-
sequent course of FSH showed the disappearance of all the stimula-
tory effects achieved with that hormone; the exception being those
induced in the Sertoli cells which persisted.
III COMMENTS
It is evident that the discrepancies in the results obtained

in patients with hypogonadotropic eunuchoidism may be attributable
to several factors, the most important being the variable potencies
of both FSH and LH activity in the gonadotropin preparations used.
Different doses and dUration of treatment, the variable degree of
susceptibility of testes to the hormones and the incompleteness of
the histological examination of the biopsies also playa part.
Nevertheless, taking into consideration the more detailed micro-
scopical studies of the testis, it may be accepted that HMG with a
FSH-LH ratio equal to unity or slightly favoring LH, administered
in weekly doses of about 300 I.U. for several months, may provoke
enlargement of tubules, increase the number of spermatogonial and
Sertoli cells, but does not apparently modify the interstitial cell
population or the clinical signs and steroid hormonal assays.
Assuming that the testicular response might be dose dependent,

we have used in our studies on cases of prepuberal hypogonadism,
different doses and lengths of treatment with FSH and LH. The fact
that all cases had approximately a similar degree of undeveloped
germinal epithelium and Leydig cells, enabled us to evaluate on a
uniform histological background the action of both hormones on the
stimulation of different phases of spermatogenesis.
Bearing in mind that differences in the individual response to

the gonadotropins may exist in eunuchoidal patients, our results
indicate that purified urinary FSH alone given in doses from 300 to
800 I.U. for several months is able to increase the number and in-
duce a maturation process in the Sertoli cells; this interesting
change includes the development of nuclear structures, mainly the
nucleoli which are related to the synthesis of RNA. The absence of
apparent changes in spermatogonial cells suggest that this hormone
is incapable of initiating or inducing spermatogenesis. This pre-
dominant effect on Sertoli cells appears to depend only on the FSH
activity as the amount of LH contaminant is negligible. This find-
ing is also in accordance with those reported by others who have
used HMG preparations rich in FSH activity (11). Moreover, an in-
creased number of Sertoli cells have been induced in the prepuberal
period of the lamb by injecting pituitary ovine FSH (26); hypertro-
phy of these cells was also obtained after intratesticular injec-
tion of the same hormone in the rat (27). Using immunohistochemical
techniques, a selective accumulation in Sertoli cells of the rat was
reported with light and electron microscopic studies after intra-
venous injection of native or labelled FSH; participation of vesi-
cles and lysosome-like bodies in the uptake of the hormone was evi-
dent (28,29). Lack of effect on the remaining structures of the
testis and on the urinary steroids favor the idea of a direct type
of action of FSH upon Sertoli cells. The earlier contention that
this cell plays a role in the spermatogenic process appears backed
570 R. E. MANCINI
not only by its participation in the intratubular diffusion of se-

rum proteins (30) but also by the fact that during normal puberty,
advanced meiosis and spermiogenesis are accompanied by the matura-
tion of Sertoli cells (31). This latter effect appears also re-
flected histologically in immature testicles of prepuberal boys
subjected to the effect of HMG (32). It has been said that these
cells reach a high glycogen content and show enzymes related to
carbohydrate metabolism, especially at puberal ages (32,33). All
these data strongly suggest that the development and function of
the Sertoli cells may be under gonadotropic control and also that
they may act as an intermediate site of FSH storage for the ulti-
mate action on the spermatogenic process. The nature of this phe-
nomenon will undoubtedly require biochemical studies at molecular
biological level.
With regard to the effect of LH, it is interesting to note

that the same doses and period of administration (contaminant of
FSH was low) as that of FSH sh~wed a similar effect on Sertoli cells
but none on the remaining structures of the testis. The fact that
only the highest dose of LH induced in addition an increment in
spermatogonial cells coinciding with a slight development of inter-
stitial Leydig cells, suggests but does not prove that LH might have
a direct effect on Sertoli cells and an indirect one on spermatogo-
nial cells mediated by Leydig cells. This dual action of LH which
would indicate that this hormone may initiate spermatogenesis, is
also supported by long experience in the use of large doses of HCG
alone which induces spermatogenesis up to the pachytene stage, dif-
ferentiation of Leydig cells, secretion of steroids, enlargement of
tubular diameter and regression of hyaline material in hypogonado-
tropic eunuchoidal patients (6,8,32). The same results were ob-
tained in prepubertal boys similarly treated (34). However, the
view that the direct effect of LH might also be mediated by some
activated function of the intertubular fibroblast-like cells cannot
be discarded.
The successive use in our studies first of FSH and then of LH

or vice-versa points to the lack of an additive effect of both hor-
mones, except in the maturation of Sertoli cells. However, it
should be noted that the higher doses of LH can preserve the bene-
ficial action achieved by previous treatment with FSH, to which a
stimulating effect on spermatogonial cells and appearance of Leydig
cells is added. On the other hand, the inefficiency of FSH, with
the exception of its effect on maturation of Sertoli cells, to pre-
serve the stimulating action of previously administered LH was evi-
dent. These findings, which may merely reflect the separate speci-
fic action of each hormone on the testis, not only verify the more
striking effect of LH but also demonstrate the simultaneous action
of both hormones. This problem as faced in our studies suggests
that similar or even lower doses of both gonadotropins administered
for an identical period of time reinforced the separate effect of
each, which does not apparently occur if an excess of FSH is given.

Moreover, a higher ratio favoring LH (1:3), during similar periods
of time, resulted in an efficient initiation and progression of
spermatogenesis at least up to the pachytene stage, together with
remarkable maturation of Sertoli cells, enlargement of tubules, re-
gression of hyalinization, development of tubular wall structures
and differentiation of Leydig cells. As has been proved, a much
longer period of administration with even higher doses of LH given
in the form of HCG together with lower dose of HMG may induce a
complete spermatogenesis in hypogonadotropic eunuchoids (7,8,10).
The necessity of FSH for further advance of meiosis is backed by
those cases successfully treated first with large doses of HCG with
resultant progression up to primary spermatocytes to which HMG was
added (8,32,35). The fact that only higher doses of LH may influ-
ence the tubule wall structure development is backed by the speci-
fic localization of intravenously injected labelled LH, not only in
both types of Leydig cells (28,36) but also in peri tubular fibro-
blasts and less clear in Sertoli cells (28,29). This would rather
suggest the existence of a direct effect of LH on Sertoli cells and
tubule wall structures and an indirect and stronger one on germinal
cells mediated by secretory Leydig cells. Consequently, the possi-
bility that FSH together with LH may act directly and at the same
time on seminiferous tubules and that a certain amount is essential
for this effect could not be ruled out.
Whether or not some degree of Leydig cell function is necessary

for full tubule development in the human being has not yet been def-
initely proved. However, and opposed to some reported data (37), it
has been demonstrated that in the human testis during puberal matu-
ration, spermatogenesis may start in the absence of Leydig cells but
in the presence of their precursors such as fibroblast-like cells,
but the full meiotic and subsequent spermiogenic processes are ac-
companied or may be preceded not only by development of mature Ser-
toli cells, but also by secretory Leydig cells (18,19,31). That
Sertoli cells may be subjected to gonadotropin control paralleled
with Leydig cells is supported by our findings on the simultaneous
stimulation exhibited by both cells when the high dose of LH alone
was used. Although Leydig cell activity is needed for germinal cell
development, whether the same may be true for Sertoli cell function
is at present a matter of mere speculation.
Results obtained in postpuberal hypogonadal patients confirmed

those obtained in hypogonadotropic eunuchoids with respect to the
impossibility of using low or high doses of FSH of LH alone or in
successive administration to recover the spermatogenesis in adult
hypophysectomized subjects; it also corroborates the predominant
effect of FSH and LH in repairing the Sertoli cells and the more
extended action of higher doses of LH. As happens with hypogonado-
tropic eunuchoid patients, this latter hormone seems to be capable
of producing recovery in the germinal epithelium by stimulating the
572 R. E. MANCINI
spermatogonial cells which then pass on to the meiotic phase. How-

ever, this cell progression as well as the regression of hyaliniza~
tion and enlargement of tubules,is concomitant with the appearance of
immature and mature Leydig cells. It would then appear likely that
smaller doses of LH would have been ineffective in bringing about
such changes, but would selectively affect the Sertoli cells. This
is also supported by the administration in the same patients of
much larger doses of HCG alone for the same periods, which resulted
in more intense stimulation of germinal cells up to the pachytene
stage, together with marked development of mature Leydig cells (25).
Although these studies also showed the need for simultaneous use of
FSH and LH with an excess of LH, it has been claimed that in eu-
nuchs, HMG containing FSH and LH in equal amounts, administered for
about three months, stimulates germinal cell progression up to the
advanced meiotic or even the spermiogenic phases. However, Leydig
cell differentiation and regression of hyalinosis were poor in
these cases, and addition of HCG induced the completion of both cell
populations, bringing about a normal condition of the tubular wall
structures (15,25). The need for a much longer treatment in eunuchs
might be due to the fact that induction of spermatogenesis would re-
quire the participation of more complicated cellular and hormonal
mechanisms, like those which are present during normal puberal dev-
elopment. In hypophysectomized patients the gonad undergoes atro-
phy to a variable degree, but gonadotropins and spermatogenesis are
present before operation.
Therefore the results obtained in both conditions would indi-

cate that administration of both FSH and LH in a ratio equal to uni-
ty is not effective, a higher LH-FSH ratio and duration of treat-
ment being required. Coincidentally it has been shown in hypophy-
sectomized rats that FSH stimulates Sertoli cell development but
only both hormones establish complete spermatogenesis (38). A FSH-
LH ratio favorable to LH is required for the recovery of spermato-
genesis, and the "gametogenic" action of LH includes some stimula-
tion of Leydig cells, which may not appear reflected by changes in
the adnexal glands (39).
Current studies on adult hypophysectomized patients given tes-

tosterone propionate alone or simultaneously treated first with HMG
and HCG and then with HMG plus testosterone propionate at weekly
doses of 50 to 200 mg, seem to indicate that the beneficial effect
of LH given in HCG cannot be replaced by the androgen. By compar-
ing the biopsies taken after a second course of treatment with those
after HMG plus HCG where recovery of all testicular structures could
be seen, a regression of spermatogenesis up to primary spermatocytes
or even spermatogonia, atrophy of Leydig cells and reappearance of
peri tubular hyalinization was detected (40). These findings are in
marked contrast with previous ones as regards the maintenance of
spermatogenesis with testosterone in hypophysectomized rats (38,41)
and resemble those obtained in adult normal subjects treated with
high doses of androgens with resultant inhibition of pituitary go-

nadotropins, increase in estrogens and atrophy of the testes (42).
With regard to the effect of FSH and LH on the tubular wall

structures it was postulated that peri tubular hyalinization is a
nonspecific process consequent to spermatogenic failure (43), or
that it damages secondarily the seminiferous epithelium (44) or
that there is an apparent synchronization between hyalinization and
regression of germinal epithelium and Leydig cells (25). Histo-
chemical and ultrastructural studies of this process have been made
in adult hypophysectomized patients (45) and results obtained seemed
to be similar to those observed in our group of eunuchs. In adult
normal subjects the seminiferous tubule wall is mainly composed of
an acellular basal lamina adjacent to the germinal epithelium and
an outer cellular layer in which fibrils arranged in compact bundles
and flattened cells are present (22). Immunochemical analysis of
this structure revealed the existence of two classes of glycopro-
tein complexes, one corresponding to collagen and the other to re-
ticulin substance (46). When peritubular hyalinization develops
the basal lamina appears dissagregated in some areas. The under-
lying cellular layer increases in thickness, the cells exhibit re-
gressive changes, and a deposition of an amorphous substance appears
in between the fibrils, which show absence of characteristic elec-
tron density and cross striation. After HCG or HHG plus HCG admin-
istration, the thickness of the tubular wall tends to diminish, the
cells recover their normal aspect, the amorphous substance disap-
pears and normal cross striated fibrils reappear. That this effect
is due to LH and not to FSH is confirmed in our group of eunuchs as
shown above. As only the higher dose of LH was able to induce such
changes, and assuming that peritubular cells are responsible for
the production and maintenance of tubular wall structures, it is
postulated that in this condition developing Leydig cells may in-
fluence the function of peri tubular cells. However the theory that
some other stimulated adjacent cells,like Sertoli cells, may parti-
cipate in this phenomenon could not be discarded.
CONCLUSIONS
Based on our present and previous studies on the separate, suc-

cessive and simultaneous effect of human urinary gonadotropins in
prepuberal and postpuberal hypogonadism, the following is proposed:
1) Neither the separate nor the successive administration of FSH
and LH may stimulate the germinal epithelium adequately. 2) Both
FSH and LH are needed simultaneously for a long time and with a
higher LH-FSH ratio for the inductio~ or the recovery of spermato-
genesis. 3) Low or high doses of FSH or a low dose of LH only in-
duce hyperplasia and maturation of Sertoli cells in the testis of
hypogonadotropic eunuchoids and repair of these cells in hypophysec-
tomized patients. 4) As doses are increased LH stimulates the in-
tertubular fibroblastic cells and might initiate spermatogenesis
574 R. E. MANCINI
through the proliferation of spermatogonial cells which then pro-

ceed to the next meiotic phase until pachytene. At this stage FSH
might participate in the further progression of meiosis. 5) An ex-
cess of LH with subsequent Leydig cell development and steroid se-
cretion would be required for full maturation of Sertoli cells to
reach the spermiogenetic phase and the completion of spermatogene-
sis. 6) Leydig cell function, by influencing the development of
tubular wall connective tissue structures or the regression of hy-
alinization, might contribute to the enlargement of seminiferous
tubules. 7) Nevertheless, it is hard to say if this hypothetical
interpretation of the relationship between gonadotropins and sper-
matogenesis in the hypogonadotropic hypo gonadal subjects, has its
eqUivalent in the events that take place during normal development
of the testis.
ACKNOWLEDGEMENTS
This work was supported by a grant from The Population Council,

Inc., New York.
REFERENCES
1. Jailer, J.W. and Leathem, J., Proc. Soc. Exp. BioI. Med., 45:
506, 1940.
2. Heller, C.G. and Nelson, O.W., J. Clin. Endocr., 8: 345, 1948.
3. Maddock, O.W., J. Clin. Endocr., 2: 213, 1949. -
4. Hurxtal, L.M., Bruns, H.J. and Mussulin, M., J. Clin. Endocr.,
2: 1245, 1949.
5. Bartter, F.C., Sniffen, R.C., Simmons, F.A. and Albright, F.,
J. Clin. Endocr., g: 1532, 1952.
6. Nelson, W., J. Urology, 69: 325, 1953.
7. Crooke, A.C., Davies, A.G. and Morris, R., J. Endocr., 42: 441,
1968.
8. Paulsen, A., ~ Gonadotropins 1968, ed. Rosemberg, E., Geron-X,
Inc. Los Altos, California, p. 491, 1968.
9. Lytton, B. and Kase, N., New England J. Med., 274: 1061, 1966.
11. Johnsen, S.G. In Gonadotropins, 1968, ed. Rosemberg, E., Geron-
X Inc. Los Altos, California, p. 515, 1968.
12. Lunenfeld, B., Mor, A. and Mani, M., Fertil. Steril., ~: 581,
1967.
13. Martin, F.I.R., J. Endocr., 38: 431, 1967.
14. Gemzel, C. and Kjessler, B., Lancet, 1: 644, 1964.
15. McLeod, J., Pazianos, A. and Bronson, R., Fertil. Steril., 17:
7, 1968.
16. Donini, P., Puzzuoli, D., D'Alessio, I., Bergesi, G. and Donini,
S., In Gonadotropins, 1968, ed. Rosemberg, E. Geron-X Inc. Los
Alt~ California, p 37, 1968
17. Heller, C.G. and Clermont, Y., Recent Progr. Hormone Res., 20:
545, 1964.
18. Mancini, R.E., Narbaitz, R. and Lavieri, J.C., Anat. Rec., 136:
477, 1960.
19. Mancini, R.E., Vilar, 0., Lavieri, J.C. and Andrada, J.A.,
Amer. J. Anat., 112: 2, 1963.
20. Dixon, W. and Massey, F.J., In Introduction to Statistical
Analysis, McGraw-Hill Book Co. Inc. New York, p. 258, 1957.
21. Mancini, R.E., Vilar, o. and Perez del Cerro, M., Acta Physiol.
Latino Americana, 14: 382, 1964.
22. Ross, M.H., Amer. ~ Anat., 121: 523, 1967.
23. Conti, C. and Fabrini, A., In Atti Congr. Soc. Med. Inter.
Edizioni L. Pozzi, Roma, p. 104, 1968.
24. DeKretser, D.M., Virchow Arch. (Zellpath), 1: 283, 1968.
25. Mancini, R.E., Seiguer, A.C. and Perez Lloret, A., J. Clin.
Endocr. 29: 467, 1969.
26. Courot, M., J. Reprod. Fertil., 89: 2, 1967.
27. Murphy, H.D., Proc. Soc. Exp. Biol. Med., 118: 1202, 1965.
28. Mancini, R.E., Castro, A.E. and Seiguer, A.C., J. Histochem.
Cytochem., ~: 516, 1967.
29. Castro, A.E., Seiguer, A.C. and Mancini, R.E., Proc. Soc. Exp.
Bioi. Med., 1970 (in press).
30. Mancini, R.E., Vilar, 0., Alvarez, B. and Seiguer, A.C., J.
Histochem. Cytochem., 11: 376, 1965.
31. Mancini, R.E., Rosemberg, E., Cullen, M., Lavieri, J.C., Vilar,
C., Bergada, C. and Andrada, J.A., J. Clin. Endocr., ~: 927,
1965.
32. Rosemberg, E., Mancini, R.E., Crigler, J.F. and Bergada, C.,
In Gonadotropins 1968, ed. Rosemberg, E., Geron-X Inc., Los
Altos, California, p 527, 1968.
33. Fabrini, A., Re, M. and Conti, C., J. Endocr., 43: 499, 1969.
34. Bergada, C. and Mancini, R.E., (In press).
35. Heller, C.J., In Estrogen Workshop Conference, ed. Paulsen, A.,
University of Washington Press, Seattle, p. 276, 1965.
36. De Kretser, D.M., Catt, K.J., Burger, H.G. and Smith, G.C.,
J. Endocr., 43: 105, 1969.
37. Johnsen, S.G-:-; Acta Endocr. (Kobenhavn), 66: 1, 1962.
38. Lostroh, A.J., Endocrinology, 85: 438, 1969.
40. Mancini, R.E. and Perez Lloret, A., Acta Europea Fertil., 1970
(in press).
41. Nelson, W.O. and Merckel, C., Proc. Soc. Exp. Biol. Med., 36:
825, 1937.
42. Heller, C.G., Nelson, W.O., Hill,I.B., Henderson, B., Maddock,
E., Junck, E.C., Paulsen, A.C. and Mortimore, G.E., Fertil.
Steril., 1: 415, 1950.
43. Nelson, W-:O. and Heller, C.G., J. Clin. Endocr., 5: 13, 1945.
45. Mancini, R.E., Seiguer, A.C. and Perez Lloret, A., In Gonado-
tropins 1968, ed. Rosemberg, E., Geron-X Inc., Los Altos,
California, p. 503, 1968.
46. Denduchis, B., Gonzalez, N., Lustig, L. and Mancini, R.E., Acta
Europea Fertil., 1: 810, 1969.
THE EFFECTS OF URINARY GONADOTROPINS FOLLOWING HYPOPHYSECTOMY AND
IN HYPOGONADOTROPIC EUNUCHOIDISM
John MacLeod 1
Department of Anatomy, Cornell University Medical
College, New York, N.Y.
I should like to present summaries of two new cases treated

with gonadotropins. The first is that of a 28-year-old married man
with two children who had undergone hypophysectomy for an infarcted
adenoma of the pituitary. I saw him for the first time one and a
half years after this operation. He had returned to the neuro-sur-
geon (Dr. Bronson Ray) because his wife desired another child. In
the interim, he had been supported only by cortisone and thyroid.
He had not received androgens. He complained of low libido, loss of
ejaculate and of pubic, axillary and body hair. He could attain
erections and orgasm after considerable stimulation. He had not ex-
perienced a spontaneous erection since the operation. It was ascer-
tained that he had been under-active sexually during his married
life and before the illness that precipitated the operation. The
post-operative loss of libido therefore, had not bothered him unduly.
He could not produce an ejaculate for examination in spite of erec-
tion and orgasm.
Since the desired end was to initiate a pregnancy for this cou-
ple, it was decided to put him immediately on the combined HMG-HCG
regimen which had been successful in the first case reported by the
author (1). The respective doses were HMG (Pergonal-batch 20714 75
IU FSH/2nd IRP - 75 IU LH/2nd IRP) and HCG (APL) 4000 IU (HCG Std)
on alternate days. Prior to treatment a bilateral testicular biopsy
was requested and obtained.
1Career Scientist, Health Research Council of the City of New York.
577
578 J. MACLEOD
PRE-TREATMENT BIOPSIES
These biopsies (Figs. 1-7) showed all stages of spermatogenesis

up to mature spermatozoa in selected tubules but a marked quantita-
tive depletion at all levels. The nuclei of many of the spermatogo-
nia (Figs. 1 and 2) appeared to be unduly large but were otherwise
normal in appearance. The pachytene stage of the primary spermato-
cyte (Fig. 3), while present in many of the tubules, was not con-
spicuous. The early and late spermatids (Figs. 4 and 5) were nu-
merically high in certain tubules. Mature spermatozoa apparently
ready for exfoliation (Figs. 6 and 7) also were obvious although
bizarre head-shapes were easily identifiable, particularly the so-
called "tapering" forms (Fig. 7). I t is of interest to note that
when the first spermatozoa appeared in this man's ejaculate follow-
ing treatment, a rather high percentage were of the tapering variety.
Peri tubular fibrosis was absent in both testes except in a few

of the peripheral tubuleso The basement membranes were normal as
were the tubular di~eters. Complete maturation was more evident in
the left testis o
The interstitial tissue in both testes was edematous ( a char-

acteristic noted in the testes of all the hypox subjects seen) and
relatively acellular. True Leydig cells were not identified.
Although the total epithelium was depleted, the presence of all

stages of spermatogenesis in selected tubules, 1~ years after opera-
tion, suggested that some degree of hormonal stimulation remained in
this individual. The urinary gonadotropin assay prior to treatment,
while very low (3 m.u./24 hours), did indicate some residual activi-
ty of hypophyseal tissue.
RESPONSE TO GONADOTROPINS
Twenty-eight days after the initiation of treatment, an acel-

lular ejaculate of 1.4 ml was obtained. At 77 days, the ejaculate
volume was 2.0 ml and an occasional sluggishly-active spermatozoon
was seen. At 83 days, the volume again was 2.0 ml and there was a
detectable sperm count of 2 million/mt with sluggish progression.
A rather high percentage of these cells were of the "tapering" va-
riety (Fig. 1), bicephalic forms were obvious as were spermatids.
From about the 15th post-treatment day, the subject was again ex-
periencing spontaneous erections.
At 111 days, a sperm count of 44 million/ml with excellent pro-

gression was recorded. The tapering and immature forms had reached
a normal level and the sperm morphology in general was attaining a
normal equilibrium which it was to maintain throughout the study
as long as any spermatozoa were present in the ejaculate. The peak
sperm count of 114 million/ml was reached at 150 days and the semen
Fig. 1. Testicular Biopsy 1~ years after hypophysectomy. Large

spermatogonia, early spermatids and Sertoli cells are
conspicuous.
Fig. 2. Same biopsy. Megalo-type spermatogonium (note bulging of

basement membrane). Late spermatids (probably spermatozoa)
present.
580 J. MACLEOD
•
Fig. 3. Same biopsy. A few pachytene cells and spermatozoa are
obvious.
Fig. 40 Same biopsy. Good population of early spermatids.

Fig. 5. Same biopsy. Middle and late stage spermatids •
•• ••
• • •
,,
•
• •,• •
•
Fig. 6. Same biopsy. Late spermatids (probably spermatozoa) with

bizarre head-shapes.
582 J. MACLEOD
Fig. 7. Same biopsy. Tapering forms are conspicuous.
quality in general was excellent.
Unfortunately, at this stage and when a pregnancy could be ex-

pected, the patient informed me that his wife no longer wished for
a child and had been on the contraceptive pill since she knew that
spermatozoa had been found in the ejaculate.
The decision was made, therefore, to withdraw the Pergonal but

to continue him on HCG for androgenic support. This approach added
to the study in allowing us (1) to follow any involution of sperma-
togenesis if it occurred and (2) to determine if HCG alone would
maintain spermatogenesis.
The second objective became most obvious within a few weeks at

which time it was apparent that spermatogenesis was being maintained .
The semen quality remained good at a level between 21 million and
54 million/ml for one year thereafter. In the interim, a decision
to try for a pregnancy again was made by the couple. Conception
occurred in the fourth cycle and a normal child resulted.
The patient refused a post-treatment biopsy but his semen qual-

ity was sufficient proof of good spermatogenesis.
With the pregnancy established and with the continued coopera-

tion of the patient, androgen support in the form of HCG was with-
drawn. We should note that from the early stages of HCG therapy i n
this subject, he had reported normal sexual vigor (for him), con··
firmed by his wife, and restoration of normal hair distribution.
Within 45 days after withdrawal of HCG, the ejaculate volume

was reduced to 0.2 ml with only an occasional sperm present. At 59
days, only a drop of fluid was obtained. It appeared certain that
involution of the reproductive tract again had occurred. At this
point, testosterone enanthate (Delatestryl-20Omg 1M/biweekly) was
given as a substitute for HCG in androgen support. After 84 days
of this regimen, the ejaculate was not restored and the patient com-
plained, as he had done after the withdrawal of HCG, of a feeling of
lethargy.
He then was returned to HCG (2000 IU on alternate days). An

ejaculate appeared within 30 days (1.4 ml) and at 51 days the vol-
ume was 2.2 mi. Spermatozoa were absent in both specimens. At
intervals up to 6 months thereafter with continuation of HCG, the
semen was examined. No cells of the germinal line were found at
any timeo In all other respects, the reproductive tract appeared
normal. His sexual activity and general well-being were good.
DISCUSSION AND CONCLUSIONS
The salient features of this study are that 1) some degree of

complete spermatogenesis remained 1~ years after hypophysectomy; 2)
after complete restoration of spermatogenesis with HMG-HCG, this
state could be continued for at least one year thereafter with HCG
alone; 3) HCG alone could not re-initiate spermatogenesis after
withdrawal of all gonadotropins and 4) testosterone could not re-
place HCG in restoring the ejaculate.
With regard to the first point, the urinary gonadotropin assay

was positive at a low level which was enough, perhaps, to maintain
a degree of spermatogenesis but not enough to support the Leydig
cells and the more distal portions of the reproductive tract.
This conclusion gains support when considering point (2). The

maintenance of normal spermatogenesis with HCG alone would mean that
if normal function can be established with HMG-HCG, the FSH-like
activity now known to be present in pregnancy urine extracts (2),
coupled with small amounts of endogenous FSH, is enough to maintain
spermatogenesis. In this regard, we agree with Johnsen (3) in that
we can restore spermatogenesis in these subjects with smaller doses
of gonadotropins than previously used and can maintain such function
with minimal levels of FSH.
However, the maintenance level of FSH or FSH-like activity is

not enough to re-initiate spermatogenesis ~ven if large amounts of
chorionic gonadotropins (LH) are given simultaneously. It would
584 J. MACLEOD
seem, therefore, that FSH does play an essential role in initiating

spermatogenesis in the hypox subject.
Lastly, the failure in this case of testosterone to substitute

for HCG in maintaining the ejaculate is not surprising. In another
subject hypophysectomized for acromegaly, who refused testicular
biopsies and attempts to restore his fertility, testosterone en-
anthate failed to maintain sexual vigor and ejaculate over extended
periods. The replacement of testosterone with HCG fully restored
the ejaculate and his masculinity within 30 days but did not return
sperm to an ejaculate which, prior to operation, had shown excellent
cellular qualityo We found the same to be true in the following
case of hypogonadism.
CASE 2 - HYPOGONADOTROPIC HYPOGONADISM
The patient, a medical resident 25 years old and married for

18 months, presented himself with a semen specimen because of in-
fertility. The facies and general physical appearance suggested
hypogonadism. The voice was high-pitched and while scalp hair was
luxuriant, temporal recession was not present. Shaving was infre-
quent.. While he was reluctant to admit that he had considered him-
self as hypogonadal, he remembered having been given methyl testos-
terone at about age 14 for retarded pubertal development. He did
not continue this treatment nor did he take androgens at any time
thereafter. His sexual activity prior to marriage was low. He did
not recall spontaneous erections and ejaculations. In his marriage
and with adequate stimulation he could perform intercourse once or
twice weekly. The semen specimen he brought was obtained by mastur-
bayion after one week of continence. The ejaculate volume was ex-
tremely low (0.30ml), a few spermatozoa of normal morphology but
sluggish motility were present. Eleven days later, a second speci-
men of even lower volume (0.10ml) was seen which contained an occa-
sional spermatozoon. He was then referred to Dr. Melvin Horwith
and the endocrinology department.
Unfortunately, the patient did not consult him for some time
thereafter. The case history does not record a physical examina-
tion until the patient had been on HCG (5000 JU on alternate days)
for 22 days. A pre-treatment biopsy was not obtained. The first
complete physical confirmed the hypogonadism, the testes being on
the small side (3 cm x 1.5 cm) but firm. The penis was considered
to be normal in size. Body hair distribution was normal but axil-
lary, pubic and extremity hair were only moderate. Breast tissue
was prominent but obvious gynecomastia was not recorded. Facial
acne, never present in the past, was already apparent after 22 days
of HCG. Gonadotropin levels prior to treatment, determined by ra-
dioimmunoassay, were detectable but low. Plasma testosterone was
0.02 m!J.&%.
After 22 days of HCG treatment, the ejaculate volume was ap-

preciably greater (1.0ml) but virtually devoid of spermatozoa. At
50 days (7-31-68) the ejaculate volume had increased to 3.0ml with
an occasional sperm and a fructose concentration of 90 mg%. The
voice pitch was markedly deepened, facial acne was prominent and
shaving had become more frequent. The patient's sexual drive and
aggressiveness had increased. Thereafter, neither he nor a semen
specimen was seen for six months but a bilateral testicular biopsy
was obtained four months after the initiation of HCG. The germinal
tissue in both testes contained an abundant population of normal
spermatogonia and all levels of spermatogenesis and spermiogenesis
were obvious. Ample evidence of mature spermatozoa was seen in ma-
ny tubules but the epithelial appearance in general was highly dis-
organized. A considerable amount of premature exfoliation of middle
and late spermatids into the tubular lumens had taken place. Peri-
tubular fibrosis was absent in tubules of normal diameter and the
basement membranes were not thickened. Leydig cells were present
in normal interstitial tissue. Both testes were essentially simi-
lar. The findings could be interpreted as quantitatively good but
disorganized germinal function. On the basis of the biopsy alone
more spermatozoa could have been predicted in the ejaculate than
had been found. The question remained, too, that the long course of
HCG which preceded the biopsy may have stimulated maturation in the
germinal epithelium.
Following the biopsy his physician withdrew HCG and substituted

clomiphene citrate. He remained on this regimen for six weeks dur-
ing which his sexual drive and feeling of well-being waned consider-
ably. At the end of the clomiphene therapy ~onadotropin was detect-
able by radio immunoassay but low. HCG therapy was resumed on 1-6-
69. A semen specimen 52 days later (2-24-69) showed a volume of
2.4 ml, no spermatozoa and 290 mg% fructose. Pergonal therapy was
begun on 3-18-69 (75 IU FSH - 75 IU LH) on alternate days with con-
tinuation of HCG (4000 IU three times weekly). Thirty-six days la-
ter, in an ejaculate of 3.5 ml, a sperm count of 4 million/ml was
recorded, 85 percent showing excellent progression. The sperm mor-
phology was disturbed. On the 52nd day of therapy the count reached
10 million/ml and on the 65th day, in a volume of 4.6ml, the count
remained at the same level. The sperm motility continued to be ex-
cellent and the sperm morphology had stabilized in a reasonably good
pattern. The count level on this date was the highest attained.
Thereafter and until the Pergonal was discontinued the sperm count
ranged from 8 to 10 million/mI. A pregnancy occurred about 123 days
after Pergonal was begun or 87 days after the first obvious sperm
count and good motility was found in the post-Pergonal ejaculate.
Pergonal was withdrawn 14 days after the pregnancy was con-

firmed, the HCG being continued at a reduced level of 2000 IU three
times weekly.
586 J. MACLEOD
At the time of writing, 127 days after withdrawal of Pergonal,

the total sperm count level of approximately 30 million has been
maintained, accompanied by excellent sperm motility, for this period.
The highest total sperm count achieved by the combined HMG-HCG ther-
apy was only 46 million, the mean total being approximately 33 mil-
lion. HCG alone, therefore, has kept up the sperm count level for
at least 127 days.
In contrast to the findings of others (2) in similar hypogona-

dotropic subjects,in this patient, we could not elavate the extreme-
ly low sperm counts in the pre-treatment ejaculate with the long
continued use of HCG alone. However, the expected pubertal changes
and higher ejaculate volumes were readily induced by the same thera-
peutic approach.
The addition of HMG t.o the HCG regimen promptly (36 days) pro-
duced the first sperm count elevation in the ejaculate and the ini-
tiation of excellent sperm motilityo But continuation of this com-
bined therapy for another 100 days did not push the sperm count
beyond 10 million/mi. Paulsen (2) and Johnsen (3) have recorded
similar failures in hypogonadotropic enuchoids.
The pregnancy probably occurred at a sperm count level between

5 and 10 million/ml but we should emphasize that the sperm motility
was excellent. We have furnished evidence elsewhere (4) that preg-
nancies can be expected at these low count levels provided sperm
motility is good.
ACKNOWLEDGEMENTS
The research was supported by Grant HD-00481 from National

Institutes of Health. I am grateful to C.R. Thompson of the Cutter
Labs. for the Pergonal and to Dr • .Tohn Jewell for the APL used in
all subjects in these experiments.
REFERENCES
1. MacLeod, J. J Pazianos, A. and Ray, B.R., Fertil. Steril, 17:

7, 1964.
2. Paulsen, C.A. , Clin. Endocr., l: 509, 1968.
3. Johnsen, S., In Gonadotropins 1968. ed. Rosemberg, E. , Geron-X,
Inc. Los Altos, Cal., p. 515, 1968.
4. MacLeod, J. , Fertil. Steril., 20: 545, 1969.
DISCUSSION
JOHNSEN: At the gonadotropin meeting in 1968 (Gonadotropins, 1968,

E. Rosemberg, ed. Geron-X, Inc., Los Altos, California) in Vista
Hermosa, Mexico, I reported a case of a male having been deprived of
pituitary function for nine years. We administered HCG to this pa-
tient but we did not observe the presence of spermatozoa. Then the
patient received HMG and HCG. We observed spermatozoa up to 20 mil-
lion per total count. His wife became pregnant and delivered a nor-
mal child. Therefore, we stopped HMG treatment and continued with
HCG alone to maintain his sexual function. Exactly as Dr. MacLeod
has pointed out, spermatogenesis continued and has continued ever
since, for nearly two years, at the level of 40 to 70 million of
total sperm count. On HCG alone his wife became pregnant for the
second time and delivered his second son.
MACLEOD: I have nothing to add to what Dr. Johnsen has said. In

our first case, the biopsy prior to treatment showed all stages of
spermatogenesis; testosterone levels were extremely low, and there
was no evidence of Leydig cell activity.
JOHNSEN: In my patient we observed only spermatogonia in nearly

all tubules and in a few tubules a very limited number of spermato-
cytes.
E. STEINBERGER: If we make the assumption that HCG does stimulate

testosterone secretion, then our discussion seems to indicate that
in the human, the situation may be similar to that of the rat, name-
ly that the initiation of spermatogenesis may require different hor-
monal milieu than the maintenance of spermatogenesis. Apparently
spermatogenesis can be maintained with HCG or testosterone.
MACLEOD: I think that the situation is quite different in the rat.

Dr. Clermont pointed out that in the testis of hypophysectomized
rats there is apparently no diminution in the spermatogonial popula-
tion whereas in the human male there is no question at all that fol-
lowing hypophysectomy there is a very severe depletion in the sper-
matogonial cell population.
LUNENFELD: I think you should have added, Dr. MacLeod, that it is

not possible to maintain spermatogenesis with testosterone alone.
PAULSEN: If there is general agreement that it is possible to stim-

ulate spermatogenesis with HCG and HMG in the hypophysectomized male
and then maintain spermatogenesis with HCG treatment alone, this
points to factors present in patients with hypogonadotropic eunuch-
oidism other than just gonadotropin deficiency. We have been unable
to maintain spermatogenesis with HCG alone in those patients except
when there was evidence of effective endogenous FSH secretion. As
Dr. Ross' group suggested, a double defect probably exists in some
588 J. MACLEOD
of the patients with hypogonadotropic eunuchoidism.
JlRASEK: I would like to raise a question which remained entirely

out of. consideration, that is the sex of offspring of the hypophys-
ectomized males. What was the sex of offspring of the hypophysec-
tomized or hypogonadotropic eunuchoid males who received gonadotropin
therapy?
JOHNSEN: Two boys.
MACLEOD: One boy so far. One miscarriage, and one still in utero.
PAULSEN: Three pregnancies. Patient J. P.'s wife had two sons.

Patient M. S.'s wife had a girl.
CHAPTER VII
SECTION 2: TREATMENT OF ADULT SEMINIFEROUS TUBULAR FAILURE

ASSESSMENT OF GONADOTROPIN THERAPY IN INFERTILE MALES
P. Troen, T. Yanaihara, H. Nankin, T. Tominaga,
and H. Lever
Department of Medicine, Montefiore Hospital and

University of Pittsburgh School of Medicine
Although human follicle-stimulating hormone (FSH) has been

demonstrated to be effective in the initiation or restoration of
human spermatogenesis (1), its use in the treatment of male in-
fertility has had variable results (2). Various reports indicate
that FSH has no effect on steroid biosynthesis in the testis (3,4)
while others suggest a stimulating effect on testosterone and es-
trogen production (5). The present investigation was carried out
to test the effect of a human FSH preparation on males with in-
fertility of varying pathogenesis. In addition to following the
response of the sperm count, the study included assay of pituitary
and gonadal hormones in blood and urine and investigation of bio-
synthetic pathways in small amounts of testicular tissue obtained
at biopsy. Criteria were sought for grouping infertile patients
in relation to response to therapy.
Pertinent data on the 11 patients studied are summarized in

Table 1. Human memopausal gonadotropin (HMG - supplied by Cutter
Labs. as Pergonal) was administered intramuscularly from ampules
containing 63 to 85 I.U. (2nd IRP) FSH and 62 to 123 I.U. (2nd IRP)
luteinizing hormone (LH). Human chorionic gonadotropin (HCG - sup-
plied by Ayerst Labs. as A.P.L.) was administered intramuscularly
three times weekly in doses of 5,000 I.U.
Seminal fluid was collected by masturbation after 2 days of

abstinence. Volume, duplicate sperm counts and motility were de-
termined within one hour of collection of the sample.
Serum FSH and LH levels were determined by double antibody

591
til
IS
TABLE I
CLINICAL MATERIAL
Pt.
(Age) Clinical Data Testicular Biopsy Clinical Response to Rx
L.R. Fertile 5 yrs. ago; infertile Disorderly spermatogenesis; sloughing; Moderate weight gain on
(33) 4 yrs; testes: 4.0 cm no arrest; ? decreased Leydig cells __ high dose HMG
R. D. Infertile Z yrs; exposure to high Focal disorderly spermatogenesis; slight No change
(Z9) temp 8 yrs ago; testes 3. 5xZ. 0 cm sloughing. no arrest; Leydig cells normal
M.R. Infertile 5 yrs; left varicocele; Arrest at 1°-~ spermatocyte; No change
(27) testis: R 4. 5xZ. 0 cm marked tubular variation; ? decreased
L 3. Ox2. 0 cm LeYclig cells
B.G. Infertile Z yrs; Arrest at spermatocyte; disorderly with Gynecomastia during
(32) testes: 4.5 cm slough; Leydig cells normal HMG and HCG
W.J. Infertile I liZ yrs; Arrest at spermatogonium or 1° sperma- Gynecomastia after
illil ___ testes:...4. OxZ. 5~_ __toc.Y!e; slo~hing; l Leydig cells increased HMG and HCG
L.B. Fertile (?) 13 yrs. ago; Arrest at ZU spermatocyte;? decreased No change
(33) infertile II yrs; Leydig cells
testes: 4.0 cm
R.A. Infertile 3 yrs; X-ray to penis 7 Peritubular hyalinization; tubular epithelium No change
(38) fl\os. ago; ttlSt.es..: 3. ~xZ. O..em clliefly Sertoli cells; ') ~eydig cells increased
I. L. Infertile 6 nlOS; Peritubular hyalinization; tubular epithelium No change
(Z5) testes: Z. 8xl. 5 cm, soft chiefly Sertoli cells; Leydig cells normal
E. G. Infe rtile Z 1/ Z yrs; Arrest at spe rmatogonium or I U sperma- No change
(3Q) _ testes: 4.0 cm tocyte; slough; decreased Leydig cells
P. W. Prepubertal; Immature tubules; absent Leydig cells Pubertal changes with HCG;
(Z4) testes: I. OxO. 5 cm regression on HMG alone
S. J. Prepubertal; Immature tubules; almost absent Pubertal changes with HCG; :-a
(Z8) testes: I. 5xl. 0 cm Leydig cells regression on HMG alone .....
;0
om
Z
m
.....
»
:
ADULT SEMINIFEROUS TUBULAR FAILURE 593
radioimmunoassay (6) utilizing reagents provided by the Endocrine

Study Section and National Pituitary Agency (7). Samples were as-
sayed in duplicate in a single FSH and LH assay respectively.
Serum testosterone analyses were performed by Bio-Science

Laboratories (8) using the technique of competitive protein-bind-
ing radioassay following extraction and chromatography. Urinary
17-ketosteroids and 17-ketogenic steroids were measured by the me-
thods of Drekter et al. (9) and Metcalf (10) respectively.
Urinary estrogens were measured using the Brown (11) and

Ittrich techniques (12) modified to extend the sensitivity of the
method to 0.10 ~g. Urinary estrogen production rates were deter-
mined following intravenous infusion of approximately 2.0 ~c of
14 C-estradiol or 3 H-estrone (13,14).
Organ cultures of testicular biopsy tissue were performed for

48 hours using 3-6 mg of tissue with radioactive precursors 3H_
dehydroepiandrosterone (0.25 mc/0.005 mg) and 14C-pregnenolone
(0.05 mc/0.302 mg). Radiochemical homogeneity of metabolites was
demonstrated on countercurrent distribution and by recrystallia-
ation to constant specific activity.
RESULTS
1. Testicular Histology
Repeat biopsies were performed on- the patients with hypo-

gonadotropic eunuchoidism (PoW. and S.J.) after therapy with HMG
and HCG. P.Wo showed the appearance of mature Leydig cells plus
increase in tubule size with spermatogenesis to the primary and
secondary spermatocyte level and, in several foci, spermatids.
There was no associated increase in testis size. Repeat biopsy
on SoJ. showed a mature tubule pattern with mature Leydig cells.
Spermatogenesis was present progressing to mature spermatozoa.
Clinical evidence of tubular maturation was an increase in testis
size to 2.3 x 1.5 cm during HMG therapy.
2. Seminal Fluid Examination
The findings on examination of seminal fluid during the

course of gonadotropin therapy are indicated in Table 2. Despite
the improvement after therapy in the histologic picture of P.W.
and S.J., no spermatozoa appeared in the ejaculate. Treatment
with HCG was required for measurable seminal fluid to be present.
Of the three patients with sperm counts below 1 million perml,
patients IoLo and LoB. showed no change with therapy. The sperm
count of E.G. increased on HMG therapy but the suggested improve-
ment did not continue with the addition of HCG to HMG therapy.
594 P. TROEN ET Al.
TABLE 2
RESPONSE OF SPERM COUNT TO GONAOOTROPIN THERAPY
Treatment Seminal Fluid AnaIX.i.
HMG Duration No. of Ave. Vol. Ave. Count Range Ava.
Patient {aml!/weekl {weeksl Sa!!!l!le s {eel {106/ eel {10 6 /eel Motili~ !!
L.R. 0 14 6 3.6 3.7 2.5-4.6 35
3 19 9 3.8 8.0 1. 3-17.4 20
6 16 8 4.3 17.7 5.1-45.4 30
14 43 8 3.7 11. 0 1.4-44.2 30
14 + HCG 31 7 4.4 9.8 4.5-30. I IS
0 30 4.5 2.5 410
H.D. 0 4 2.6 6.5 5.5-8.2 40
J 8 2. I 8.4 1.6-15.7 35
5 13 10 2.5 8.3 1.4-27.9 30
10 4 2.3 7.6 I. 9-18. 0 25
0 4 2 1.9 6.8 5.5-8. I 35
M.R. 0 26 4 4. I 4.7 3.8-5.6 40
3
I} 4 2. 3 18.8 6.5-40.4 50
6
14 4 3. 2 10.7 60
0 21 I 3. 5 9.9 90
B.G. 0 9 9 4. 8 12.4 4.7-44.5 30
3 16 5 4.8 4.8 2.0-9.8 25
6 13 9 4.7 11. 1 4.6-17.5 30
14 22 5. 5 9.0 3.7-14.3 15
14 + HCG 20 6.1 7.1 3.3-9.6 35
W.J. 0 17 6.8 2.5 0.6-11.0 15
3 13 8 7.5 1.7 O. 1-3. I 10
6 9 5 7.7 1.6 0.1-3.7 25
6 + HCG 7 4 8.1 2.2 0.1-4.2 10
L.B. 0 4 2 2.4 0
3 30 1.3 0
R.A. 0 17 1.7 2.4 0.5-7.9 30
16 4 2.0 2.3 0.5-7.0 30
I.L. 0 2 2 2.8 0.2 0.1-0.3 5
3 10 2 2.7 0.4 0.1-0.7 5
E.G. 0 6 2.7 0.5 0.37-0.73 60
14 21 4.2 3.0 2.3-4.3 60
14 + HCG 6 1 4.0 0.63 40
0 6 2 3.8 1.2 0.9-1.5 50
P.W. 0 0
HCG 28 2 0.5 0
0 40 0
14 17 0
14 + HCG 8 2 0.5 0
6 + HCG 8 0.5
S.J. 0 0
HCG 26 2 0.75 0
0 22 0
14 17 0
14 + HCG 13 2 1.5 0
5 + HCG 5 1.0 0
Four patients, B.G., R.D., R.A. and W.J. with initial average
sperm counts from 2.4 to 12.4 million per ml showed no improvement
in sperm count on varying treatment regimens of up to 14 ampules of
HMG per week. The addition of HCG to HMG in B.G. and W.J. also
yielded no improvement.
From initial average sperm counts of 3.7 and 4.7 million per
ml, L.R. and M.R. had increased counts on HMG therapy. There was
no further increase at the highest HMG dosage or with added HCG
(patient L.R.). No patient's wife became pregnant during the treat-
ment program.
3. Serum Gonadotropin Levels
Before therapy, P.W. and S.J. had low total urinary gonadotro-
pin excretion and E.G. had normal gonadotropin excretion by bio-
assay. In Table 3 the values for serum FSH and LH by radioimmuno-
assay are given. Levels of gonadotropin were determined after a
minimum of 17 days and an average of 84 days on a given treatment
schedule. The time from last injection of HMG to collection of
blood for assay was 34 hrs. at the dosage of 3 ampules per week,
24 hrs. at 6 ampules per week and 14 hrs. at 10 and 14 ampules per
week.
I.L. and R.A. twice had elevated baseline FSH values. Both
patients had testicular biopsy pictures characterized by loss of
germinal epithelium with preservation of Sertoli cells and peri-
tubular hyalinization. Baseline FSH values of the other six pa-
tients were in the normal range. Administration of three ampules of
HMG per week resulted in increased FSH levels in B.G., R.D. and L.
R., the three patients with the lowest pre-treatment levels. At
10 and 14 ampules per week, all patients studied showed further
elevation of serum FSH reaching similar levels at the highest dosage
of HMG. After cessation of therapy, serum FSH values fell, return-
ing very closely to the pre-treatment values (R.D. and L.R.).
I.L. and R.A., in addition to elevated FSH levels, twice had

values of LH which seemed abnormally ~igh. The exact degree of LH
elevation was not determined. The baseline serum LH values of the
other patients were in the normal range. Treatment with HMG result-
ed in no consistent alteration of serum LH for all patients, al-
though B.G. showed a progressive decrease in serum LH. The varying
content of LH in different lots of the HMG preparation appeared to
have no correlation with changes in serum LH.
4. Plasma Testosterone
Baseline plasma testosterone values showed a wide range from

242 ng% (L.B.) to 540 ng% (R.D.) with the majority of patients
596 P. TROEN ET AL.
TABLE 3
VALUES OF BLOOD FSH, LH AND TESTOSTERONE AND URINARY
17 KETOSTEROIDS AND 17 KETOGENIC STEROIDS IN RELATION TO
HMG FSH LH Testosterone 17-KS 17-KGS

Patient (amp/wkl (mIU / mil (mIU/mll (ng/IOO ml) (mg/l4 hrl (mgl l4 hr)
L.R. 0 l. I 5.0 486 13.9 Il.4
3 3.9 3.3 336 14.6 14.3
6 14. I 14. l
~
14 8.l 6.9
14.9 15. l
14 6.6 7.6 413
14 + HCG 6.4 >14.0 1330 17.4 15.7
0 3. I 3.9 387 15.5 17. I
R.D. 0 l. I 5.9 540 16.8 16. l
3 3.8 7.5 408 16.6 16.6
10 5. 3 5.2 339 17.6 18.6
0 2.9 5. 7 444 15.4 14.8
Iffi-
M.R. 0 5.2 5. I
0 4.6 2.9 15.7 13.4
0 5. I 2.9 23
3 5.4 10. I 267 14.6 12. l
6 16.7 15.2
14 7. I 5.4 15.8 II. I
0 248
B.G. 0 3.0 12.2 255 13.0 9.6
3 5.0 9.4 222 14.2 14.7
6 II. 3 Il.7
14 7.7 8.3 162 13.4 Il.4
W.J. 0 370 20.3 16.7
3 5.7 5.2 388 23.4 19.0
6 6. I 11. 9 478 20.6 18.9
0 5.7 9.4 412
0 6.6 5. I
~~
L.B.
9.8 19.6
0 7.0 6.6
R.A. 0 21. I :>14.0 439 17.3 17.4
0 Zl.6 ;;'14.0 410 13.7 15. 8

1. L. 0
0
17.5
Zl.2
:>14.0
>14.0 !I} 12.4 15.9
11. 0 10.9
Normal Values: Serum FSH: mean 3. 6 mIU / ml; range 1. 6 - 8.l5

Serum LH: mean 4.9 mIU/ml; range 2.3 - 15
Plasma testosterone: range 400 - 1200 ng I 100 ml
Urina ry 17-ketoste roids: range 5 - 25 mgmll4 hrs
Urinary 17-ketogenic steroids: range 5 - 21 mgm/24 hrs
Ave rage values during treatment pe riod a re given for

17-KS and 17-KGS
having values at or below 400 ng% (Table 3). None of these patients
had clinical androgen deficiency. In two patients with low plasma
testosterone (L.B. and M.R.) there was a suggestion of decreased
Leydig cells on biopsy. There was no correlation between plasma
testosterone values and the sperm count.
During the several stages of HMG therapy, the plasma testos-

terone level had a variable course. In two patients (B.G. and R.
D.), there was progressive decrease in plasma testosterone, asso-
ciated in B.G. with a decreasing serum LH value. In contrast, the
plasma testosterone of W.J. increased with therapy and was accom-
panied by a higher serum LH value. L.R. had an initial decrease of
plasma testosterone and serum LH but then an increase in both values
particularly when HCG was added to the therapy (15). In contrast,
M.R. showed no correlation between changing values of LH and plasma
testosterone.
5. 17-Ketosteroid and 17-Ketogenic Steroid Excretion
Values for 17-ketosteroid and 17-ketogenic steroid excretion

were within the normal range without significant change during
gonadotropin therapy. The lowest baseline 17-KS excretions observ-
ed were for L.B., who had the lowest plasma testosterone, and S.J.
(5 mg/24 hrs.), a hypogonadotropic patient.
6. Urinary Estrogens
The excretion values for estrone (E1), estradiol (E2) and

estriol (E3) were similar to those reported for normal males (13,
14) except for a high pre-treatment value for estriol in R.D. (Table
4). During HMG therapy, estrogen escretion remained normal. Uri-
nary production rates of estradiol, although higher in L.R. and R.
D. toward the end of HMG therapy, were within the normal range be-
fore and during HMG therapy. One measurement of urinary estrone
production rate during HMG therapy was also normal. The addition
of HCG to the treatment of L.R. resulted in a marked increase both
in excretion of each estrogen and in estradiol production rate (15).
7. Organ Culture of Testicular Biopsy Tissue
Following organ culture with 3 H-DHA and 14 C-pregnenolone, the

incubating medium yielded 49% to 78% of the 3 H and 31% to 58% of
the 14 C-radioactivity. Conversions to specific metabolites (as
percent of recovered radioactivity) were calculated, without cor-
rection for losses, following recrystallization to constant specific
activity (Table 5). The polar compounds in the neutral fraction as
well as the water soluble (conjugated) fraction are under study.
The organ cultures showed a significant metabolism of added

precursors to testosterone. Ready conversion of 65 to ff com-
(.n
00
'"
TABLE 4
URINARY ESTROGEN EXCRETION AND ESTROGEN PRODUCTION RATE DURING GONADOTROPIN THERAPY
Excretion (fLg/24 hr.) Prod. fLg/ % Recove ry Radioactivity

HMG Rate 24 hr.
Patient (aInE/week) Estrone Estradiol Estriol E2 El Estrone Estradiol Estriol Total
L.R. 0 5. 10 0.74 2.03 13.5 9.0 5.6 5. 7 20.3
3 4.41 0.56 2.21 13.5 7.2 4. I 4.2 15.5
14 4.64 0.43 1. 78 9.6 9.4 4.4 4.9 18.7
14 4.00 0.67 1. 85 33.9 10.2 2.3 2.0 14.5
14 + HCG 22.n 3.06 8.47 160.6 9.5 2.9 3.5 15.9
R.D. 0 3. 90 0.84 12.90 19.0 6.4 4.5 5.4 16.3
3 4.10 0.99 4.00 21.7 6.1 4.6 4.5 15.2
10 3. 10 0.54 0.84 31. 5 9. 5 5.6 1.4 16.5
0 4. 50 0.97 3.82 21. 3 7.4 4.6 3.7 15.7
M.R. 0 2.84 0.58 2.01 17.3 6.3 3.2 6.1 15.6
0 2.83 0.28 3.21 7.4 6.0 3.7 8. 1 17.8
8 6. 1 4.3 7.3 17.7
14 3. 14 O. 58 3. 38 -L. 49.8 6.4 2.6 6.9 15.9
Normal values: mean 4.42 O. 94 3.n 29 46.3 9.3 3.9 6.9 20. 1
(Ref. 13, 14) S. D. 1. 30 O. 34 I. 73 14.3 12.7 3. 1 1. 1 2.2 5.3 ;t>
--I
0""
m
Z
m
--I
»
:-
TABLE 5 >
0
cr-
RADIOACTIVE STEROIDS IDENTIFIED IN NEUTRAL FREE FRACTION FROM MEDIUM OF TESTIS ORGAN -i
CULTURE USING PRECURSOR 3H-DEHYDROEPIANDROSTERONE AND l4C_PREGNENOLONE U'I
m
~
Z
""T1
Patient M.R. W.J. R.A. I. L. m
'"
0
14C 3 14C c
3H 14C 3H/ 14C 3H 14C 3H/ 14 C 3H 3H /14 C H 3H /4 C U'I
-i
Precursor 0.4 O. I 4. 0 2. I 0.7 3.0 2. I 0.7 3.0 2. I 0.7 3.0 c
0:>
(f!.c) C
r-
>
Recovery 59.0 50.4 4.7 49.6 31. 5 4. 7 78.0 47.1 5. 0 65.1 57.7 3.4 ""T1
'"
% ~
r-
C
Te stoste rone 40. I 3.3 57.9 6.7 3. 9 8.2 53.0 3. I 85.2 59.2 16.9 11. 9 m
'"
Pregnenolone 7.3 31. 7 1.2 19.4
Progesterone 2. 5 2.2 0.4 1.6

.~
17 OH-Preg. 1.0 0.3
:tl
u
0 170H-Prog. 3. 7 6.4 3.2 8. 1
:a'"
~ DHA 1.4 8.8 0.7 59. 5 1.2 19.2 0.3 227.6
'tl
..,.,:> 6 4 _A 0.5 0.9 2.6 1.7 2. 9 2.8 0.5 0.7 0.6 3.9
0
.,p:;u 6 5 _A 2.2 II. 8 2.0 0.7
~I Polar 4. 1 16. I 1.2 42.2 53. 3 3.8 4.7 30.7 0.8 6.0 31. 3 0.6
(- ) not identified
17-0H-Preg 3~. 17a-dihydroxypregn-5-en-20-one DHA = 311-hydroxyandrost- 5-en-17 -one

17-0H-Prog 17a-hydroxypregn-4-ene-3, 20-dione 6 4 _A O!
= androst-4-ene-3. 17-dione '0
6 5 _A = androst-5-ene-311, l711-diol '0
600 P. TROEN ET Al.
pounds took place. The higher 3 H/ 14C ratio for testosterone than
for androstenedione suggested that a significant portion of the
testosterone arose from DHA by a pathway other than through andro-
stenedione (16). We are continuing studies of these pathways (17).
No correlation was noted between patterns of conversion of

steroids in these organ cultures and sperm counts, levels of plasma
testosterone or serum gonadotropins, or response to HMG therapy.
DISCUSSION
The 11 infertile males in this study include a wide range of

abnormalities of testicular histology and function. Four groups
of patients can be separated. 1. The two hypogonadotropic patients,
following combined HMG and HCG therapy, showed a histologic response
of the testis, but there was no correction of the azoospermia. In
these patients with a presumably responsive condition (1,4), the
incomplete results may be due to a combination of factors includ-
ing duration of therapy, dosage and ratio of FSH/LH therapy. 2.
Three normogonadotropic patients with disorderly spermatogenesis or
spermatogenic arrest had increased sperm counts during HMG therapy,
but remained at oligospermic levels. Whether this reflects the
limit of responsivenesss or inadequate knowledge concerning effec-
tive treatment regimens in not known. 3. Four other normogonado-
tropic patients showed no increase in sperm count on HMG therapy.
They were not distinguishable from the previous three patients by
such criteria as histologic appearance of the testis .or the pre-
liminary data on response to therapy of serum gonadotropin, plasma
testosterone or urinary estrogen excretion and production rates.
4. Two hypergonadotropic patients showed no response to less
intensive treatment programs. These patients would appear to be
unlikely to respond to exogenous HMG therapy because the endogenous
FSH is at a level higher than that produced by treatment.
Under the time relations of injection and venipuncture used

here, levels of serum FSH remained within the range for normal
adult males although they increased during HMG therapy. Attempts
to improve spermatogenesis in normogonadotropic males by use of
additional FSH stimulation may require levels of serum FSH above the
normal range. Monitoring of serum FSH levels during therapy is sug-
gested for future treatment programs. It is not known whether con-
tinuous high levels or intermittent peaks of FSH would be effective
in stimulating spermatogenesis. Further information concerning the
half-life of injected HMG (18) may help to establish optimal time-
dose relationships.
The finding of low plasma testosterone values in association

with oligospermia but normal secondary sex cha~acteristics warrants
further study. The possibility of a dissociation between levels of
circulating testosterone required for spermatogenesis and levels
reqUired for secondary sex characteristics would have therapeutic

significance. The question of possible interaction between germinal
epithelium and Leydig cells in maintaining normal steroid biosyn-
thetic activity is again raised.
The significance of the changes noted in plasma testosterone

during HMG therapy is not clear. A direct relation to changing
serum LH levels is suggested in some of the patients but other pa-
tients showed no correlation. The finding of high LH values with
low normal plasma testosterone values .in two patients suggests al-
tered testis responsivenesss to circulating LH.
The administration of HMG despite its content of LH did not

result in any consistent change in serum LH values. This may be due
to the shorter half-life of LH compared to FSH (19,20). There was
no clinical indication of physiologic LH activity of the administer-
ed HMG in the hypogonadotropic patients and no evidence for testis
stimulation as measured by estrogen excretion or production. The
plasma testosterone levels varied during treatment, but, as indicat-
ed above, there was no consistent pattern of increased LH nor did
the plasma testosterone rise above mid-normal levels until HCG was
added.
The possibility of abnormalities of steroid biosynthesis being

associated with disorders of spermatogenesis suggested the study in
vitro of testicular tissue from infertile males (21). Using an
organ culture technique, small amounts of tissue(3-6 mg), as would
be obtained with conventional biopsy procedures, have now been used
for investigation of steroid metabolism in these patients. Conver-
sion of pregnenolone to testosterone with identification of various
intermediates has been demonstrated. Studies are in progress con-
cerning pathways, response to stimulation and comparison with tis-
sue with normal spermatogenesis.
ACKNOWLEDGEMENTS
This research was supported in part by U.S.P.H.S. Grant AM-

08375 and the B.F. Anathan Fund. The technical assistance of Miss
Harriet Savasten, MISS Linda Kuntz, Mrs. Lisa Paraska and Miss
Masumi Kubomoto is gratefully acknowledged. The assistance of Dr.
Harvey Mendeiow in obtaining aud examining the testicular biopsies
is appreciated.
REFERENCES
1. Mancini, R.E., Seiguer, A.C. and L1oret, A.P. J. C1in. Endocr.,

29: 467, 1969.
2. Lunenfe1d, B., Goldman, B. and Sha1kovsky. R. In Progress in
Endocrinology, edt Gua1, C., Excerpta Medica Foundation, Amsterdam
p 1044, 1969.
3. Mroueh, A., Lytton, B. and Kase, N. J. Clin. Endocr.,
l:2:53, 1967.
4. Paulsen, C.A., In Estrogen Assays in Clinical Medicine,
ed. Paulsen, C.A. Univ. of Washington Press, Seattle, ~274,
1965.
5. de Kretser, D.M., Taft, H.P., Brown, J.B., Evans, J.H., and
Hudson, B. J. Endocr., 40: 107, 1968.
6. Odell, W.O., Ross, G.T., and Rayford, P.L., J. Clin. Invest.,
46: 248, 1967.
7. Albert, A., Rosemberg, E., Ross, G.T., Paulsen, C.A. and
Ryan, R.J., J. Clin. Endocr., ~: 1214, 1968.
8. Demetriou, J.A. et all Clin. Chern., to be published.
9. Drekter, I.J., Heisler, A., Scism, G.R., Stern, S., Pearson, S.
and McGavack, T.H., J. Clin. Endocr., ~: 55, 1952.
10. Metcalf, M.G., J. Endocr., 26: 415, 1963.
11. Brown, J.B., Advances Clin. Chern., ed. Sobotka, H., Academic
Press, New York, Vol. 3, p. 157, 1960.
12. Salokangas, R.A.A. and Bulbrook, R.D., J. Endocr., ~: 47,
1961.
13. Crowell, G.e., Turner, M.E., Jr., Schmidt, F.H., Howard, C.M.
and Preedy, J.R.K., J. Clin. Endocr., 1]: 807, 1967.
14. Eren, S., Reynolds, G.H., Turner, M.E., Jr, Schmidt, F.H.,
Mackay, J.H., Howard, C.M. and Preedy, J.R.K. J. Clin.
Endocr., 27: 819, 1967.
15. Lipsett, M:D., Wilson, H., Kirschner, M.A., Korenman, S.G.,
Fishman, L.M., Sarfaty, G.A., and Bardin, C.W., In Recent
Progr. Hormone Res., ed. Pincus, G. Academic Press, New York,
Vol. 22, p. 245, 1966.
16. Perez-Palacios, G., Lamont, K.G., Perez, A.E. and Jaffe, R.B.,
J. Clin. Endocr., ~: 786, 1969.
17. Yanaihara, T. and Troen, P., Proc. Central Soc. Clin. Res.,
~: 122, 1969.
18. Rosemberg, E., Joshi, S.R., and Nwe, T.T., In Gonadotropins
1968, ed. Rosemberg, E. Geron-X, Los Altos, California,
p. 139, 1968.
19. Yen, S.S.C., Llerena, 0., Little, B., and Pearson, O.H.,
J. Clin. Endocr., 28: 1763, 1968.
20. Stevens, V.C., J. Clin. Endocr., 29: 904, 1969.
21. Danezis, J.M., Fertil. Steril., 177 488, 1966.
DISCUSSION
HUDSON: Dr. Troen commented that his patients exhibited low levels
of plasma testosterone. I think that this is a relevant question in
view of Dr. Heller's presentation. What does Dr. Troen regard as the
lower extreme of the normal range in his laboratory? Among our own
group of infertile men without hypogonadotropic syndromes, the only
patients who consistently show abnormalities in levels of plasma tes-
tosterone are those patients with the Sertoli cell only syndrome. In
25% of these patients, the plasma levels of testosterone are less than
300 ng/100 ml of plasma, which we regard as significantly lower than

normal. Oligospermic patients with germinal cell arrest rarely show
abnormalities in testosterone secretion, clinically or chemically.
TROEN: We consider the low limit of the normal range to be 400 ng/
100 ml. The patients referred to had values of plasma testosterone
as low as 240 ng/100 ml.
Eo STEINBERGER: Dr. Troen, in the incubates with l4C pregnenolone

and tritiated DHA you found the ratio of tritium to carbon to be
greater in testosterone than in androstenedione. You interpreted
these findings as indicating that the pathway does not go via andro-
stenedione?
TROEN: Yes, in the organ culture studies, we incubated 3H dehydro-

epiandrosterone and l4 C_ pregnenolone and found that the tritium/
l4C ratio of the recovered testosterone was greater than that for
androstenedione suggesting therefore that a significant portion of
the testosterone arose from DHA by a pathway other than through an-
drostenedione. As we indicated in discussion earlier during this
meeting additional data indicate this is via /), 5-androstenediol.
ASSESSMENT OF GONADOTROPIN THERAPY IN MALE INFERTILITY
Carl Schirren and Joseph O. Toyosi
Andrology Department, University Clinic for Skin Diseases

Hamburg-Eppendorf, Bundesrepublik Deutschland
The treatment of male subfertility has involved several differ-

ent gonadotropin preparations. We shall describe our experience
with the following: a) Pregnant mare serum (PMS), b) Human chori-
onic gonadotropin (HCG), c) Human menopausal gonadotropin (HMG), d)
Human hypophyseal gonadotropin (HHG) and e) various combinations of
these agents.
Emphasis will be laid on the influence of these gonadotropins on

the seminal fluid (sperm count, sperm morphology and sperm motility).
Ou:..- studies were carried out between the years 1953 and 1969 and in-
volved 449 cases. (See Table 1).
Table 1. Survey of Patients Treated with Gonadotropins.
PMS
Oligospermia 284
Hypozoospermia 79
Hypokinesis 58
421 cases
HMG and HCG
Oligospermia 26
Azoospermia 1
27 cases
HHG and HCG
Azoospermia 1 1 case
Total 449 cases
605
606 C. SCHIRREN AND J. O. roVOSI
Until recently, the most widely used gonadotropin preparation

for male subfertility was PMS. PMS was first used by Tonutti (1).
Although PMS is placental in origin, it primarily exerts a follicle
stimulating hormone (FSH) action rather than an interstitial cell
stimulating hormone (ICSH) action. Tonutti concluded that ICSH has
no direct or indirect influence on spermiogenesis when used to stim-
ulate testicular function following hypophysectomy. He also con-
cluded that the presence of FSH was necessary for maturation from
primary spermatocytes to secondary spermatocytes.
The first consideration was the influence of PMS on sperm mor-

phologyo PMS was initially administered at the dose rate of 1,000
I.U. twice weekly for six weeks. Recently the dosage rate has been
increased to 2,000 IoU. twice weekly for three weeks because of the
antigonadotrupic effects of PMS. The results obtained are summarized
in Table 2. There was a mean increase of 11% in the number of nor-
mal structured sperm. A corresponding decrease of 4.9% was observed
in the "clLanged middlepart" sperm. Both the mean sperm count and
the motility increased. Data derived from an additional 151 cases
are summarized in Table 3. An improvement in the three important
factors, i.e., motility, morphology and sperm count was noted.
Table 2. The Effect of PMS on Human Spermatozoa

(1,000 I.U. PMS twice weekly for 6 we~ks) (72 cases)
Morphology Sperm Motility

a b c d e f g h Count I II III
before PMS:
44.2 10.8 0.66 17.5 14.5 3.1 1.7 2.1 41.1 39.4 15.7 44.9
after PMS:
55.7 7.7 O.ll 12.6 14.4 3.1 2.0 4.4 82.1 52.0 14.0 34.0
a: normal sperm, b: small head, c: giant head, d: changed

middlepart, e: double head, f: malformed head, g: phantom,
h: abaxial implantation, Sperm Count: (in millions/mI.),
I: high motility, II: moderate motility, III: immotile.
Table 3. The Effect of PMS on Sperm Count, Motility and Morphology

in Patients with Subfertility (151 cases).
Group
(millions/ml) Before/After Moti li ty Morphology Sperm Count
(cases) Treatment (%) normal (%) mi 11ions/m1.
1 -10 before 48 48.3 4.3
(16) after 52 61.2 20.2
11-20 before 35 53.6 13.6
(29 ) after 55 54.1 25.5
21-30 before 56 67.3 26.9
(28 ) after 65 72.3 62.8
31-40 before 54 68.2 34.5
(39) after 59 70.0 42.8
41-60 before 53 71.7 41.4
(18) after 66 74.9 63.4
hypokinesis before 39 36.1 131.2
(21 ) after 47 60.0 120.0
From our experience with PMS, the following pOints are worthy
of note: 1) treatment with PMS in the human is risky because of
antihormone production. When used we recommend 2,000 I.U. twice
weekly for a limited period of three weeks. 2) With careful pa-
tient selection, PMS alone will usually improve sperm count, sperm
motility and sperm morphology. 3) We have modified our PMS treat-
ment by adding androgen (2). The addition of testosterone augments
the PMS and to a certain extent maintains the achieved result for
longer periods of time. We recommend Mesterolone at a daily dosage
of 20 mg for five months.
We believe that by using the guidelines of PMS therapy our

treatment can be improved by using homologous gonadotropins. To
date, we have treated 27 patients with HMG, 21 with Humegon and 6
with Pergonal.
HMG treatment was given to 20 males using a dosage of 500 I.U.

(IRP= 75 I.U. FSH) thrice weekly for 12 weeks. In another group
(6 patients) the weekly dose was doubled. The results of these
studies are seen in Fig. 1.
608 C. SCHIRREN AND J. O. roYOSI
cases
10 r--
sperm count 9 morphology motility
cases 8 cases
7 r-- 7 f-- 7 r--
6 6 I--
6
5 r-- 5 5 .---
4 - 4 4
3 3 3
2 I-- 2
~
2 ~r--
~
o~ I
o .. + ~ - -- ++ +
~
t6 - -- I o ++ + t6 - --
The effect of HMG on human spermatozoa (20 cases)

o 3 inj. weekly a
500IU (12 weeks)
a
B8I 3 inj weekly 1000 W (12 weeks)
Fig. 1. An improvement in sperm morphology only is seen; the sperm

count increases slightly, no change in sperm motility. Between the
dosage of 500 I.U. and 1,000 I.U. there is no difference in the re-
sults obtained. ++: important improvement ( >50%), +: moderate
improvement (11-49%),~: no change (+ 10%), -: moderate decrease
(11-4970), --: important decrease (>50%).
Table 4. Clinical Data of Six Patients Treated with HMG + RCG

Year of Testes Diseases in Urinary
No. Birth Volume Anamnesis Gonadotropins
(FSH-I.U.)
1 1925 normal f/J 6.6
(soft)
2 1930 normal hepatitis 6.6
diphtheria
3 1937 normal tonsi 11ec tomy 13.2
4 1927 normal f/J 6.6
5 1932 normal f/J 26.4
6 1931 normal ~ 26.4
Table 4 shows the clinical data of patients treated with HMG and
RCG alternately (Pergonal and Choragon). In three cases the gonado-
tropin content in urine was low, in the other cases the levels were
normal. The sperm counts differed between below 3 million/ml. and

12.3 million/ml. before the hormone treatment. The control spermato-
gram after three months of treatment shows an improvement of sperm
count in one case only, in the other five cases the sperm count
diminished or remained unchanged. The morphology and motility re-
mained unchanged in all cases. In one case we noted gynecomastia
during the treatment, and the hormone treatment was interrupted
because of swollen and painful breasts.
The last case in this study was a 33-year old man with a clinical
diagnosis of azoospermia. The histological picture shows germinal
cell arrest (Case 2 Tables 4 and 5). HMG (Humegon) can produce
sperm in the ejaculate but no motility. Later HHG was administered
for 10 days at a dosage of 500 I.U. daily. In connection with this
treatment we also injected 5,000 I.U. HCG daily for 10 days. The
spermatogram of this patient remained unchanged.
Table 5. Effect of HMG/HCG on Human Testis
No. Testicular Biopsy Sperm Morphology Motility Side

(Diagnosis) Tubule Diameter Count (% normal) ('70) Effects
( t.L ) (mi11/ml)
a'i'( bt a* bt a* bt
1 l44-1BO 6 <3 30 40 30 40 f/J
(0)
2 10B-1BO 0 <3 0 0 (J (J (J
(A)
3 10B-216 11.4 19.5 70 65 70 65 gyneco-
( 0) 12.3 mastia
4 120 6.B 5.0 45 40 40 40 f/J
(0) 7.2
5 95 5.4 < 3.0 45 49 40 40 (J
( 0) 6.2
6 145 11.4 10.2 35 40 40 45 (J
( 0) 9.7
a*: before treatment, bt : after treatment for 3 months,

(A) : azoospermia, (0) : Oligospermia.
The hormone status in his urine showed an increase in 17-

ketosteroids, estrogens, pregnanediol and gonadotropins.
610 C. SCHIRREN AND J. O. TOYOSI
Table 6. Effect of lIMG, HHG and HCG in Azoospermia
Date Hormone Status in Urine

Treatment Spermatogram ct dtt
( 1963) a"l: b~':"l(
22.1 lIMG (Humegon) azoospermia

75 I.U. 3 x
weekly for 12
weeks
13.6 some sperms 11.06 20.9 1.70 6.6
20.8 HHG+ (500 I.U. some sperms

daily for 10
days)
1.10 some sperms 14.0 35.2 1.92 26.4
10.10 HCG (5,000 I.U. some sperms 14.2 34.8 1.90 26.4
daily for 10
days).
+ Sponsored by G. Bettendorf.
a*: 17-ketosteroids in mg., b"l:*: estrogens in I-L g, c t preg-
nanediol in mg, d tt: gonadotropin in mouse U.U.
The application of lIMG is limited. This oplnlon is supported

in the literature (3-7). This treatment is mainly for hypogonado-
tropic hypogonadism. Even in this disease it is not possible in
all cases to produce an improvement in the sperm factor. According
to Lunenfeld (4) the following criteria are important if a positive
reaction of the seminiferous tubules is to be obtained: a) tubule
diameter from 50 to 70 I-L , b) germ cells with Sertoli cells, spermato-
gonia- and spermatocytes , c) hypogonadotropin content in urine.
CONCLUSION
The treatment of the subfertile male having oligospermia and

hypokinesis with gonadotropin preparations should not be compared
with the use of these agents in men following hypophysectomy. In
the latter gonadotropin therapy is highly successful.
Since the causes of oligospermia are quite extensive, therapy

is difficult at the present time. Now that human gonadotropin
material is available it is imperative that our diagnostic capa-
bilities are improved. Prior to treatment, it is important to
measure the endogenous gonadotropin levels and examine a testicular
biopsy specimen. We do not support the notion that repeated testi-
cular biopsies are necessary during treatment. We hold the view
that the testes are vulnerable to repeated operations and that

permanent damage may result. The spermatogram affords an insight
into the effectiveness of gonadotropin preparations after a longer
period of observation.
Since our experience with HMG, alone or in combination with

HCG, is somewhat limited, we cannot give a positive prognosis. It
is possible that with a thorough selection and examination of the
patients, better results could be achieved.
REFERENCES
1. Tonutti, E., Zo Urolo 11, 1955.

2. Schirren, C., Internist (Berlin) ~:2, 1967.
3. Johnsen, S. G., and Christiansen, P., In Gonadotropins 1968,
ed. Rosemberg, E., Geron-X, Inc., Los Altos, California, p 516,
1968.
4. Lunenfeld, B., Goldman, B., and Ismajovich, B., ~Gonadotropins
1968, ed. Rosemberg, E., Geron-X, Inc., Los Altos, California,
p 5l3, 1968.
5. Mancini, R. E., Seiguer, A. C., and Levret, P., In Gonadotropins
1968, ed. Rosemberg, E., Geron-X, Inc., Los AltoS;- California,
p 503, 1968.
6. Paulsen, Co A., In Gonadotropins 1968, ed. Rosemberg, E.,
Geron-X, Inc., Los Altos, California, p 491, 1968.
7. Rosemberg, E., Mancini, R. E., Crigler, J. F. and Bergada, C.,
In Gonadotropins 1968, ed. Rosemberg, E., Geron-X, Inc., Los
Altos, California, p 527, 1968 0
ASSESSMENT OF GONADOTROPIN THERAPY IN MALE INFERTILITY
B. Lunenfeld and R. Shalkovsky-Weissenberg
Institute of Endocrinology, Tel-Hashomer Government Hos-

pital, Israel and Department of Life Sciences, Bar-Ilan
University, Ramat-Gan, Israel
Male infertility may be produced by genetic, mechanical, infec-

tious, inflammatory or hormonal causes. Correct diagnosis is essen-
tial before any therapy is started. In this presentation we will
analyze and discuss therapeutic results only in such cases where in-
fertility was presumably due to hormonal factors.
In order to evaluate correctly the results of hormonal therapy

in male infertility, it was found necessary to summarize existing
knowledge on the hormonal control of testicular function.
Hypophysectomy causes morphological changes in Leydig cells and

Sertoli cells as well as atrophy of androgen-dependent parts of the
male reproductive system (1-4).
It is universally accepted that Leydig cell proliferation and

function is under the control of ICSH (interstitial cell stimulating
hormone). This is also evidenced by the fact that administration of
ICSH or HCG (human chorionic gonadotropin - biological activity sim-
ilar to ICSH) promotes excessive Leydig cell function (5), and that
fluorescein-labelled ICSH is selectively localized in Leydig cells
(6). It seems clear that the main secretory product of the testis
is testosterone and that this is produced mainly by the Leydig cells.
Several pathways of testosterone biosynthesis have since been

proposed (7). Furthermore, it has been shown that the secretion of
the Leydig cells is stimulated by ICSH and HCG and is suppressed by
the administration of androgen and estrogen (8).
Gonadotropins also have a particular action on supporting cells
613
614 B. LUNENFELD AND R. SHALKOVSKY-WEISSENBERG
(immature Sertoli cells). Courot (9) showed that in the impuberal

lamb the number of supporting cells was reduced by almost half of
the initial control within a few days following hypophysectomy. The
supporting cells which survived hypophysectomy were not transformed
into Sertoli cells. Vilar (10) showed that following hypophysectomy
in the rat the Sertoli cells became smaller, atrophic and the amount
of cytoplasm of these cells decreased. Sertoli cells seem to present
the first target of gonadotropic action in the seminiferous tubules;
therefore they may have an important role in spermatogenesis. Mancini
(6) demonstrated that fluorescein-labelled FSH (follicle stimulating
hormone) appears mainly in Sertoli cells, while the labelled ICSH is
selectively localized in Leydig cells. This finding shows that Ser-
toli cells are influenced by gonadotropins.
It has been postulated by de Robertis et ale (11,12) that in

amphibians ICSH provokes release of spermatozoa from Sertoli cells;
this has been further clarified in electron microscope studies by
Burgos et ale (4). Furthermore, Burgos et ale showed in preliminary
experiments with male hamsters that a similar mechanism may exist in
mammals, i.e. that ICSH may be responsible for sperm release from
the Sertoli cells. On the basis of these studies the authors pos-
tulated that the Sertoli cell cytoplasm is a target for ICSH activity;
the cytoplasmic swelling and the unfolding of the apical recesses
which is accompanied by the release of spermatozoa in the hamster
was considered by these authors as the counterpart of ovulation.
Hypophysectomy also causes an arrest in spermatogenesis in prac-

tically all mammalian species (3). The exact role of gonadotropins
on the germinal epithelium is still debated. Some authors (2,13,14)
claim that gonadotropins act on the first stages of spermatogenesis.
Steinberger and Duckett (15) concluded that the progression of type
A spermatogonia to late stages of pachytene spermatocytes may not
require hormonal factors. Smith (16), Crooke and Gilmour (17) and
Tonutti (18) believe that gonadotropins act on spermatogenesis at
the late stages of meiotic division of the primary spermatocytes.
Clermont and Morgentaler (19) stated that, in the hypophysectomized
rat, the process of spermatogenesis may proceed as far as step 7 of
spermiogenesis. Lostroh et ale (20) observed spermatogonia as well
as primary spermatocytes in the tubules of hypophysectomized male
rats even six months after operation. Since these authors came to
the conclusion that after "complete" hypophysectomy residual gonado-
tropic activity might be present, they attempted to neutralize this
activity with antisera to ovine ICSH; this resulted in further atro-
phy of the testes, but degenerating spermatocytes and mitosis of
spermatogonia, although infrequent, were still observed.
The same author (21) demonstrated that administration of ICSH to

hypophysectomized animals elicited an increase in the population of
young primary spermatocytes. Spermatogonia in mitosis were much in evi-
dence, but no spermatids were observed. Since hyperplasia of Leydig
cells was noted, and since the functional activity of these cells
was demonstrated by increased weight of seminal vesicles and ventral
prostate, one cannot conclude from the experiment whether the sper-
matogenic activity seen was due to the ICSH or to the androgens pro-
duced by the interstitial cells through ICSH stimulation. Injection
of FSH to hypophysectomized rats initiated an extensive and orderly
repair of a large number of seminiferous tubules; most tubules con-
tained several rows of primary spermatocytes in addition to young
spermatids. Although from these results one could claim that FSH
stimulated the formation of spermatocytes, it is unfortunate that
the FSH preparation used was probably contaminated with ICSH. This
becomes evident from the fact that, although the interstitial tissue
was still deficient, it showed signs of stimulation over and above
that of the controls. Lostroh's experiments, therefore, are not
conclusive as to the selective role of FSH, ICSH and testosterone
on the first stages of spermatogenesis in the mature hypophysectom-
ized rat.
Cutly and Cutly (22) showed that spermatogenesis in the testes

of hypophysectomized rats can be maintained with testosterone pro-
pionate. The testes of hypophysectomized testosterone-injected ani-
mals did not differ histologically from those of the normal controls.
The above mentioned experiment demonstrated that hypophysectomy

of the mature animal will severely affect spermatogenesis and sper-
miogenesis and indicates clearly the necessity of circulating gonado-
tropins to maintain normal testicular function. The authors could
not conclusively pinpoint any particular stage in the process of
hormonal restoration of spermatogenesis.
The study of mechanisms controlling initiation of spermatogenesis

was hampered by the fact that hypophysectomy at birth or in early
infancy was practically impossible, since such young hypophysectom-
ized animals either die or are killed and abandoned by their mothers.
Moreover, by its very nature hypophysectomy may lead to inconclusive
results due to deprivation of other vitally important pituitary hor-
mones. Steinberger and Duckett (15) thus attempted to suppress go-
nadotropic activity by injecting estradiol from birth. In rats
receiving estradiol daily from birth spermatogenesis commences, and
at the age of 15 days the testes show progression of spermatogenesis
similar to that of normal uninjected controls (23). However, the
pachytene spermatocytes are unable to complete the meiotic division
and degenerate at diakinesis. In testes of 60 and 90 day old rats
the tubules contain Sertoli cells, spermatogonia and only occasional
spermatocytes.
These results suggest that in the absence of gonadotropins and

additionally in the secondary absence of testosterone production, the
initial wave of spermatogenesis commences at puberty. However, the
spermatocytes are unable to complete the reduction division. The
question arises whether estrogen administration blocks gonadotropic

release and therefore the few spermatocytes found after estrogen
administration in the adult animal may be due to small amounts of
gonadotropins released. Furthermore, one would have to exclude a
direct effect of estrogens on the testes.
The purpose of our investigation was to elucidate the factors

governing testicular development and initiation of spermatogenesis.
This includes the hormonal requirements in spermatogenesis and sper-
miogenesis, as well as the conditions which lead to normal sperm
formation. An experimental approach to the above mentioned problem
was to compare animals deprived of gonadotropins from birth to nor-
mal littermates, at different stages of testicular development. A
continuous deprivation of circulating endogenous gonadotropins in
mice, from birth onwards, is made possible by the continuous admin-
istration of antisera to rat gonadotropin (24).
From Table 1 it can be seen that there was no significant differ-

ence in testicular development during the first week of life. At
the age of 14 days the difference between the control and the gonado-
tropin-deprived animals is apparent inasfar as the number of sper-
matogonia and spermatocytes is significantly reduced in the experi-
mental animals.
During the third week of life the first spermatids could be seen
in some of the tubules. In the experimental animals no tubules con-
taining spermatids were found (for each testis at least 50 cross -
sections were examined). The number of spermatogonia and spermato-
cytes in the gonadotropin-deprived animals remained significantly
reduced. The mean diameter of the tubules of the experimental ani-
mal was significantly smaller than the control.
In the fourth and fifth weeks of life spermatids in the normal

animals became abundant, and a mean of 102 per tubule could be counted
on day 30. In the gonadotropin-deprived animals spermatid formation
was totally inhibited, and the number of spermatogonia and spermato-
cytes in the gonadotropin-deprived animals remained similar to the
numbers found on day 14, and these were significantly less than in
the control animals. Whereas the mean total number of cells of the
germinal layers increased continuously up to day 30 in the control
animals, no change was observed in the gonadotropin-deprived ani-
mals. It can therefore be concluded that in the absence of gonado-
tropins during puberty, spermatid formation ceases and there is no
increase in the total number of cells of the germinal layers from
day 14 onwards. The increase of spermatogonia and their transfor-
mation to spermatocytes from day 7 to day 14 in the antiserum-treated
animals cannot yet be interpreted (Table 1).
When gonadotropin-deprived animals received substitution of FSH

+ ICSH, a spermatogenic pattern ai~ilar to that of the normal was
Table 1 »
0
C
r-
--I
EFFECT OF GONADOTROPIN DEPRIVATION AND SUBSTITUTION ON TESTICULAR DEVELOPMENT IN MICE en
m
~
Mean No. of Cells Per Tubule Cross Section and (S.D.) Tubule Wt. of
~
"T1
Diam. Testes m
Age Treatment Spermatogonia Spermatocytes Spermatids Mean as % of 0
Total Serto1i '"c
(days) Mean P Mean P Mean mean ~ body wt. VI
Mean --I
C
OJ
7 NRS* 5.0 ( 1.0) 0 0 5.0 22 (3.0) 45.0 cr-
< 0.05 »
"T1
'"
7 ARG** 3.2 ( 2.1) 0 0 3.2 23 ( 2.8) 40.5 ~
r-
C
14 NRS 25.0 ( 9.8) 20.4 ( 6.9) 0 45.4 20.4(5.1) 72 .0 0.35 '"m
< 0.0025 < 0.01
14 ARG 15.0 ( 5.2) 14.1 ( 6.1) 0 29.1 20.5(7.8) 66.6 0.14
19 NRS 28.6 (10.3) 38.0 ( 3.9) 7.6 74.2 14.8(7.3) 117.0 0.44
< 0.01 < 0.005
19 ARG 12.6 ( 4.3) 12.4 ( 5.4) 0 25.0 19.8(3.0) 33.6 0.17
25 NRS 52.0 (10.0) 40.2 (15.8) 33 125.2 12.0(3.0) 132.0 0.30
< 0.005 < 0.005
25 ARG 18.3 ( 8.0) 14.0 ( 5.0) 0 32.3 17.0(5.0) 54.0 0.15
30 NRS 47.3 (26.8) 54.0 ( 9.0) 102 203.3 15.4(4.4) 140.0 0.55
< 0.005 < 0.005
30 ARG 11.0 ( 5.5) 14.2 ( 5.6) 0 25.2 17.2(4.5) 66.0 0.20
25 ARG+HMG 65.0 (31.9) 72.1 (23.0) 48.0(46.0) 185.1 20. 2( 6. 7) 90.0 1.30
25 ARG+HCG 26.6 ( 9.1) 25.8 ( 5.9) 31.5(22.0) 83.9 12.3(3.4) 68.0 0.35
25 ARG+FSH 32.7 (13.8) 33.0 (l2.5) 8.0( 6.6) 73.2 14.2(4.8) 66.0 0.50
0-
* Normal rabbit serum. ** Anti-rat gonadotropin.
'I
618 B. lUNENFElD AND R. SHAlKOVSKY-WEISSENBERG
obtained. Antiserum to rat gonadotropins, together with non-cross-

reacting human FSH qualitatively restores spermatogenesis up to the
spermatid stage, although quantitatively less spermatids are found
than in the normal littermates. It is of interest that HeG in the
presence of anti-rat gonadotropins permitted the meiotic process to
progress to completion, as shown by the occurrence of spermatids
(Table 1).
It is interesting to note that when Steinberger and Steinberger

(25) cultured testes from five day old rats, only differentiation
up to the pachytene stage of meiosis was shown to take place in
vitro in the total absence of hormones. This seems to be in accor-
dance with our in vivo results in mice up to the age of 14 days,
where spermatogonia proliferated into pachytene spermatocytes in
the absence of gonadotropins.
The effect of hypophysectomy in men was investigated by several

authors. Mancini et ale (5) examined testicular biopsies in men one
month to six years after hypophysectomy. When examining the biopsies,
the authors found that the extent of regression of the seminiferous
epithelium, tubular wall and interstitial cells were of different
degrees. Some biopsies were characterized by Sertoli cells, almost
complete spermatogenic line, slight hyalinization of the tubular wall
and an appreciable number of Leydig cells. In other biopsies the tu-
bules contained Sertoli cells and primary spermatocytes up to the
pachytene stage, moderate hyalinization of the tubular wall and a
low number of Leydig cells. There was a correlation between the
time elapsing after operation and the degree of involution of the
testes, indicating that the degenerative process took from a few
weeks to several years. The authors found that PMSG (pregnant mare
serum gonadotropin) and HeG treatment acted upon increase of sper-
matogonial cells, thus giving rise to primary spermatocytes. HMG
(human menopausal gonadotropin) caused a further augmentation of
testicular function, and combined therapy of HMG and HeG caused a
significant stimulation of all types of cells, especially from the
spermatocyte onward-, and even up to spermatozoa. It also caused a
marked increase in the number of mature Leydig cells.
Restoration of spermatogenesis in a hypophysectomized patient

with HMG was reported by MacLeod et ale (26,27). Treatment of this
patient was started when the process of involution of the seminif-
erous tubules was virtually complete, except for the presence of a
few primary spermatocytes at the pachytene stage.
After 67 days of daily HMG injections, testicular biopsy revealed

that maturation had occurred close to the stage of exfoliation of
spermatozoa into the lumen. However, t~is stage of maturation was
achieved only in some of the tubules. A testicular biopsy, performed
100 days after initiation of therapy, demonstrated a fully quanti-
tative restoration of spermatogenesis. There was no evidence of
Leydig cell proliferation or successful ejaculation by the patient,

and blood testosterone was exceedingly low. This indicated that the
ICSH content of HMG was insufficient for the restoration of the in-
terstitial tissue. This seems to imply that spermatogenesis and
spermiogenesis can proceed without any morphological or biochemical
evidence of Leydig cell proliferation and function. This should
not be surprising,since Pasqualini and Bur (28) and Johnsen (29)
have discussed the so-called "fertile" eunuchs, a condition of tu-
bular development and spermatogenesis with complete absence of Ley-
dig cells.
Fifty-one days after initiation of combined HMG and HCG therapy,

a measurable volume of seminal plasma was obtained in which the sperm
count was 60,000,000 per ml. According to this study, it would ap-
pear that daily HMG (200 IU FSH + 200 IU ICSH) is sufficient to re-
store spermatogenesis and spermiogenesis within 100 days, but con-
tains insufficient ICSH to detect any stimulation of Leydig cells.
Mancini et ale (5), studying hypophysectomized subjects aged 18

to 63, demonstrated that HCG was capable of stimulating Sertoli cells,
increasing tubular size and causing maturation of germinal epithelium
up to pachytene stage, regression of "hyalinization" and proliferation
of Leydig cells. HMG (75 IU FSH+ 75 IU ICSH), similar to HCG, not
only provoked growth of seminal tubules and maturation of Sertoli
cells, but also advanced proliferation of germinal epithelium with
the production of spermatids and occasional spermatozoa.
HMG + HCG significantly reinforced the effect of HMG alone on

the production of spermatids and spermatozoa. However, within the
74 + 5 days of this treatment regime, the quantitative level of sper-
mat~genesis and spermiogenesis seen in normal adult patients was not
reached. These authors also demonstrated that testosterone, given
together with gonadotropins, had an inhibitory effect on the response
to HMG therapy.
This study seems to confirm the earlier work of MacLeod et ale

(26,27), in which HMG + HCG restored spermatogenesis in a hypophy-
sectomized patient. The qualitative resemblance but quantitative
difference between these two groups of authors might be due to the
different amounts of HMG administered (MacLeod et al., 200 IU FSH +
200 IU ICSH; Mancini et al., 75 IU FSH + 75 IU ICSH).
The partially hypophysectomized patient of Gemzell and Kjessler

(30) showed a pre-treatment biopsy with spermatogenic activity up to
early spermatid stage. After three weeks of human pituitary gonado-
tropin (HPG) therapy, spermatozoa could be seen in the ejaculate.
After 13 weeks of treatment, the number of spermatozoa in the ejac-
ulate rose to 72,000,000 per ml. In this study all stages of the
first meiotic division and the metaphase of the second meiotic
division, together with immature spermatids, could be demonstrated

prior to HPG treatment. The authors therefore claim, in contra-
diction to MacLeod, that spermatogenesis was independent of gonado-
tropic stimulation and that the effect of HPG was specific for sper-
miogenesis.
However, it could be argued that if this patient was only partially

hypophysectomized, enough gonadotropic activity could have been pre-
sent to prevent complete involution of the germinal epithelium,but
not sufficient for the final stages of maturation. With HPG therapy,
spermiogenesis was rapidly restored and sperm appeared in the ejac-
ulate within only 21 days of treatment. The effect of HPG on the
earlier stages of sperm production (i.e. spermatogenesis) may be
indicated by the steady increase in sperm counts up to the 13th week
of treatment. The full restoration of sperm production by HPG alone
indicates either a higher ICSH content of this preparation or the
presence of endogenous ICSH.
The results of these earlier investigations can therefore be

interpreted as showing the need for both FSH and ICSH in order to
completely restore spermatogenesis after partial or complete gonado-
tropic deprivation. There is no conclusive evidence that any stage
of spermatogenesis in man may be independent of gonadotropic stimu-
lation.
It must be realized that data on restoration of spermatogenic

activity with human gonadotropic preparations after partial or com-
plete gonadotropic deprivation does not necessarily imply that a
response can be obtained in immature testes of children or eunuchoids
with complete or partial lack of gonadotropins. It might be that
gonadotropin therapy promotes spermatogenesis only in the previously
stimulated germinal epithelium.
After administration of HMG (16-75 days) to four children aged

six to seven years with unilateral cryptorchidism, Rosemberg et ale
(31) demonstrated maturation of Sertoli cells and a significant in-
crease in spermatogonial cells, as well as maturation of germinal
epithelium up to primary spermatocytes.
Bergada and Mancini (32) compared the differential effects of

pregnant mare serum (PMS), HCG, PMS followed by HCG, and HMG + HCG
in 17 children aged six to nine years with unilateral cryptorchidism.
In the HMG-HCG therapy group, a greater increase in tubular size,as
well as a greater stimulation of the germinal epithelium, was pro-
voked compared to the other therapeutic regimes. No regime provoked
maturation beyond the spermatocyte stage. Sertoli cell maturation
and Leydig cell proliferation were found both with HMG + HCG as well
as with PMS + HCG. The authors conclude that both FSH and ICSH-like
activity might be necessary for initiation of spermatogenesis.
Furthermore, data has now accumulated on the effect of human gon-

adotropins on adult eunuchoidal men. Heller (33) treated a eunuch-
oidal male first with HCG and then with HCG and HMG simultaneously.
Eight weeks after the latter treatment, testicular biopsy showed com-
plete spermatogenesis. Davies (34) treated a hypogonadotropic eunuch-
oid with HPG, 600 mg eq. of the International Reference Preparation
(IRP HMG) three times weekly. After four months, increase of testic-
ular size and enlargement of tubules was seen. The dose was then in-
creased to 2000 mg eq. IRP HMG three times a week for two more months
and a still greater increase of testicular size was noted; there was
no maturation of the germinal epithelium nor were any Leydig cells
seen. When 12,000 IU of HCG was added to this regime for another
four months, Leydig cells were prominent and spermatozoa were found
in a testicular biopsy. One month later spermatozoa appeared in the
ejaculate.
Martin (35) demonstrated in a patient with an apparently isolated

deficiency of pituitary gonadotropins and with a testicular biopsy
showing only spermatogonia and isolated spermatocytes, that prolonged
therapy with HCG (12 months) did not modify testicular morphology
significantly. There were still only two or three layers of cells
within the tubules, although mitosis was more frequent and more ob-
vious. No spermatids or spermatozoa were found. In contrast, in-
terstitial cells were much more numerous and obvious, and there was
a slight increase in the fibrous elements of the interstitial tissue.
After HPG (500 IU FSH twice weekly) + HCG (1,000 IU once a week) for
61 days, motile, morphologically normal spermatozoa were seen in the
ejaculate. The sperm count rose steadily and reached a level of
25,000,000 per ml eight months after treatment.
Johnsen's extensive study (36) of two adult men with severe hypo-
gonadotropic hypogonadism with infantile testes treated with HCG for
23 and 6 months respectively, revealed the development of secondary
sex characteristics and probably stimulation of t.ubular diameter.
HMG treatment alone, following the HeG regime, did not significantly
modify testicular morphology, except that in both cases the number
of Sertoli cells per tubule increased (28.5± 6.4 to 39.4± 9.7;
38.9 + 10.9 to 51.8 + 20.8). The combined HMG-HCG regime for 12 and
9 months respectively induced a complete gonadal maturation in both
cases.
Paulsen (37) studied five hypogonadotropic hypogonadic eunuch-

oids. From his study it seems that HMG alone for up to six weeks
stimulated increase of tubular diameter in two out of three cases,
an increase of mean cell count in one out of three cases, but with-
out significant maturation of the germinal cell epithelium. For two
cases who received HCG alone for 42 to 56 weeks, Sertoli and Leydig
cell maturation were stimulated; in one patient the proliferation
advanced to Sd spermatids, and in the other only up to the stage of
pachytene spermatocytes, and varying degrees of germinal cell matura-
tion were noted.
Two of the patients who initially received HMG therapy alone

were subsequently treated with HCG for 20 to 25 weeks. One of these
patients showed germinal cell maturation to the Sd spermatid and the
other patient remained unresponsive. In the one patient who received
HMG, followed by HCG, and finally HMG + HCG therapy, full spermato-
genesis and spermiogenesis were obtained, and an occasional sperm
was encountered in the ejaculate. Of the two patients who initially
received HCG and subsequently received HMG + HCG, one responded with
full spermatogenic and spermiogenic stimulation with 13.6 million
sperm in the ejaculate, and the second showed only pachytene sper-
matocytes.
Lytton and Kase (38) studied a 19 year old primary panhypopitu-

itary hypogonadic male after partial evacuation of a craniopharyn-
gioma and radiation therapy, whose testicular biopsy revealed fetal
type testes with germinal arrest and absence of Leydig cells. After
the administration of HMG (12.5 IU FSH three times a week for three
months), no changes were observed in either gonad or secondary sex-
ual characteristics. He was then given HCG (2000 IU three times a
week for some six and a half months). Secondary sex changes devel-
oped and a testicular biopsy revealed an increase of interstitial
cells artd germinal epithelium stimulation to the stage of spermatids.
He was then given a combined HMG-HCG regime (12.5 IU HMG + 2000 IU
HCG) three times a week for another two months. A testicular biopsy
at that time revealed marked focal areas of mature tubules, showing
the presence of spermatozoa.
Therefore it can be noted that in most cases(eight out of nine

patients)the combined HMG-HCG therapy induced complete gonadal matu-
ration (Table 2).
Table 2
Complete Gonadal
Patient Maturation
Heller (33) 1965 1 1
Davies (34) 1965 1 1
Paulsen (37) 1965 3 2
Lytton and Kase (38) 1966 1 1
Johnsen (36) 1966 2 2
Martin (35) 1967 1 1
The demonstration that combined HMG-HCG treatment can induce com-

plete gonadal maturation in some patients has been confirmed by Joel
(39) who induced complete gonadal maturation in three cases of azoo-
spermia whose pre-treatment testicular biopsies revealed the presence
of spermatogonia. (No effect was found in eight other cases whose
pre-treatment biopsies showed tubules with only Sertoli cells).

Furthermore, Lunenfeld et al. (40) demonstrated that in all six cases
of hypogonadotropic azoospermic males whose pre-treatment biopsies
revealed infantile testes with mean tubular diameter of 50 to 70~
containing Sertoli cells and cell elements up to various forms of
spermatogonia, complete gonadal maturation could be obtained within
80 to 190 days of combined theraFY. Five out of 15 patients who had
tubules of normal or near-normal size (170 to 250~) with normal,
moderate, or thickened walls, containing spermatogonia, Sertoli cells
and, in some tubules, occasional spermatocytes, had normal sperm
counts after HMG-HCG treatment. Based on the knowledge of the length
of the spermatogenic cycle, treatment was given for 80 to 144 days in
this group of patients.
Sixteen out of the 17 patients who had tubules of normal or near

normal size and which contained only Sertoli cells were unresponsive
to therapy. This may be explained by the fact that no spermatogenic
response can be expected in the absence of cellular elements per-
taining, to the germinal line. Since biopsy material might not rep-
resent the status of all the tubules, patients diagnosed as having
a "Sertoli-only cells" syndrome might actually have mixed forms. In
such cases gonadotropic therapy may be beneficial therapeutically
or it may at least serve as a diagnostic tool.
In a different series of patients showing normal tubules which

contained spermatogonia, treatment was continued without any fixed
time limitation (75 IU FSH + 75 IU ICSH + 2500 IU HCG; three times
per week). One patient responded after 17 months with occasional
spermatozoa in the ejaculate; a month later a normal sperm count was
obtained and his wife became pregnant. Three other patients responded
after 12 to 15 months of treatment, and the remaining patients are
still under therapy.
It can be speculated that administration of human gonadotropins

may stimulate spermatogenesis in hypogonadotropic oligospermic men
and also in oligospermic patients with a hyposensitive response to
normal gonadotropic stimulation, thus improving the sperm count and/
or quality.
Treatment with various gonadotropic preparations with FSH and

ICSH activity have been used in males with primary spermatogenic
disorders, but the results have been variable and, on the whole,
unsatisfactory (see Tables 3 and 4).
These unsatisfactory results can be explained by the relative

lack of knowledge of other pathological conditions, besides gonado-
tropin deficiency, which lead to oligospermia ( e.g. varicocele, in-
fection, inflammation, genetic conditions), and the lack of data on
the relationship between dose and scheme of gonadotropic treatment
and the spermatogenic response.
~
./>.
Table 3
Results of Gonadotropic Treatment in Group I (Severe Oligospermia 0.1 - 10 x 10 6 /m1.)
Improvement Improvement
No. of Improvement No. of
from 1 to from I to
Author Cases from I to II Pregnancies
20-30 x 10 6 /m1 > 30 x 106/m1
Schoysman (41) 4 1 1 0 2
Lytton and Mroueh (42) 9 1 0 1 0
Abe11i and Fa1agario (43) 6 2 0 0 2 ?"
r-
c
Cittadini and Quartaravo (44) 11 5 1 1 1 z
m
Z
"'T1
Po1ishuk et a1. (45) 9 1 2 1 1 m
r-
0
Mroueh et a1. (46) 4 0 0 0 0 »
z
Danesis and Batrinos (47) 9 2 1 0 1 0
~
Debiasi and Misura1e (48) 8 1 0 2 1 (f)
::I:
De Kretser et a1. (49) 2 1 0 0 0
»
r-
;><;
0
Mak1er et a1. (50) 9 1 2 2 1 <
(f)
;><;
Lunenfe1d et al. (40) 27 6 0 8 1 -<
~
!!!
(f)
(f)
m
Z
co
m
;:0
G'>
Table 4. Results of Gonadotropic Treatment in Group II (Moderate

Oligospermia 10-20 x 106/ml.)
Improvement
No. of No. of
from II to
Author Cases Pregnancies
>30 x 106/ l
m
Schoysman (41) 4 2 o
Lytton and Mroueh (42) 9 6 2
Abelli and Falagario (43) 26 9 7
Cittadini and Quartaravo (44) 2 2 o
Polishuk et ale (45) 5 1 o
Mroueh et ale (46) 2 1 o
Danesis and Batrinos (47) 1 1 1
Debiasi and Misurale (48) 7 4 2
De Kretser et ale (49) o o o
Makler et ale (50) 2 2 1
Lunenfeld et ale (40) 24 12 4
It should be noted that the mechanisms by which FSH and ICSH or

HCG effect spermatogenesis, spermiogenesis or release of sperm and
contribute to the normal behavior of the sperm and its motility are
poorly understood. The role that the FSH/ICSH ratio may play in any
one of these processes is still unknown. It can be speculated that
these processes are regulated by a feedback mechanism between the
pituitary and the testes, as well as by an interplay between the two
major testicular components, the Leydig cells and the germinal epi-
thelium.
It can be expected that when oligospermic conditions are defined

and separated into specific categories and when the pathological mech-
anism leading to each condition is better understood, certain groups
will be found which would benefit from individually adjusted gonado-
tropic treatment.
When treating female patients, ovarian response to gonadotropins

will be reflected by changes in cervical mucus, in vaginal smear and
the endometrial epithelium as well as by levels of ovarian steroids
in body fluids.
In the male the steroid excretion pattern of Leydig cells is not

directly correlated to spermatogenic activity. Furthermore, the
ovarian cycle is relatively short (follicular ripening is approxi-
mately 15 days), whereas in the male the spermatogenic cycle is
approximately 70 days. Sperm analysis and quantitative histology
of testicular biopsies are the only practical tests now available
to indicate, although late, the testicular response to gonadotropic

stimulation. Due to the delay in appearance of response, these tests
cannot be used for simultaneous monitoring of treatment in males.
In female patients the daily dose requirements to induce ovulation

vary from 75 IU to 375 IU FSH and ICSH daily, and the mean daily dose
was 188 IU FSH and ICSH. Most authors reporting on the use of gonado-
tropins in males administered dosages significantly lower than have
been used for the induction of ovulation. Thus it is not surprising
that in the absence of basic information about the physiological re-
quirement of gonadotropic hormones at different stages of the sper-
matogenic cycle and in the absence of clinical indices for monitoring
treatment, the results obtained so far are unsatisfactory.
Before a final evaluation of the use of gonadotropins in cligo-

spermic men can be attempted, it would only be fair to administer to
a sufficiently large group of well defined patients the daily mean
dose of gonadotropins which was sufficient to induce ovulation in
females. The analysis of such a study would probably produce a guide-
line for empirical treatment.
Treatment with gonadotropins must remain empirical until appro-

priate parameters reflecting responses suitable for monitoring treat-
ment become available.
ACKNOWLEDGMENT
This study was supported in part by Ford Foundation Grant No.

67-470.
REFERENCES
1. Greep, R. 0., Fevold, H.L. and Hisaw, F.L., Anat. Rec., ~: 261,
1936.
2. Greeo. R.O. and Fevold, H.L., Endocrinology,ll: 611, 1937.
3. Greep, R.O. In Sex and Internal Secretion, 3rd ed., ed. Young,
W.C., Willia;; and Wilkins, Co., Baltimore, Maryland, Vol. 1,
p 240, 1961.
4. Burgos, M.H., Vitale-Calpe, R. and Russo, J. ~ Gonadotropins
1968, ed. Rosemberg, E., Geron-X, Inc., Los Altos, California,
p 213, 1968.
5. Mancini, R.E., Siegueur, A.G. and Perez Lloret, A. In Gonado-
tropins 1968, ed. Rosemberg, E., Geron-X, Inc., Los~ltos,
6. Mancini, R.E., J. Histochem. Cytochem., 11: 516, 1967.
7. Samuels, L.T., Uchikawa, T. and Huseby, R.A. ~ Endocrinology
of the Testis, Ciba Foundation Colloq. Endocr., eds. Wolstenholme,
G.E.W. and O'Connor, M., J. & A. Churchill, London, Vol. 16, P
211, 1966.
8. Hudson, B., Coghlan, J.P. and Dulmanis, A. !E Endocrinology of

the Testis, Ciba Foundation Colloquia on Endocrinology eds.
Wolstenholme, G.E.W. and O'Connor, M., J. & A. Churchill, London,
Vol. 16, p 140, 1966.
9. Courot, M., J. Reprod. Fertil. Supple 2: 89, 1967.
10. Vilar, O. !E Gonadotropins 1968, ed. R-;;-semberg, E., Geron-X, Inc.
Los Altos, California, p 205, 1968.
11. De Robertis, E.D.P., Burgos, M.H. and Breyter, E., Rev. Soc.
Argent. Biol., ~: 21, 1945.
12. De Robertis, E.D.P., Burgos, M.H. and Breyter, E., Proc. So~.
Exp. Biol. Med., .§l: 20, 1946.
13. Greep, R.O., Van Dyke, H.B. and Chow, B.F., Anat. Rec., ~:88,
1940.
14. Albert, A. !E Sex and Internal Secretion, 3rd ed., ed., Young,
W.C., Williams and Wilkins, Baltimore, Maryland. 1961.
15. Steinberger, E. and Duckett, G.E., J. Reprod. Fertil., Supple
1,: 75, 1967.
16. Smith, P.E., Amer. J. Anat., 45: 205, 1930.
17. Crooke, A.C. and Gilmour, J.B., J. Path. Bact., ~: 525, 1938.
18. Tonutti, E., Z. Zellforsch, 11.: 495, 1943.
19. Clermont, Y. and Morgentaler, H., Endocrinology, 57: 369, 1955.
20. Lostroh, A.J., Johnson, R. and Jordan, C.W., Acta:Endocr.
(Kobenhavn) 44: 536, 1963.
21. Lostroh, A.J:7 Acta Endocr. (Kobenhavn) 43: 592, 1963.
22. Cutly, E. and Cutly, E.C., Endocrinology, 26: 503, 1940.
23. Steinberger, E. and Duckett, G.E., Endocrinology,~: 1184, 1965.
24. Lunenfeld, B. and Eshkol, A. !E Gonadotropins 1968, ed., Rosemberg,
E., Geron-X, Inc., Los Altos, California, p 197, 1968.
25. Steinberger, A. and Steinberger, E., J. Reprod. Fertil., Supple
2, 117, 1967.
26. MacLeod, J., Pazianos, A. and Ray, B., Lancet, 1197, May 30,
1964.
27. MacLeod, J., Pazianos, A. and Ray, B., Fertil. Steril., 12: 1,
1966.
28. Pasqualini, R.A. and Bur, G., Fertil. Steril., £: 144, 1955.
29. Johnsen, S.G., Acta Endocr. (Kobenhavn) Suppl., 66: 1, 1962.
30. Gemzell, C. and Kjessler, B., Lancet, 644, March~l, 1964.
31. Rosemberg, E., Mancini, R.E., Crigler, J.F., Jr., and Bergada.
C. In Gonadotropins 1968, ed., Rosemberg, E., Geron-X, Inc., Los
Altos, California, p 527, 1968.
32. Bergada, C. and Mancini, R.E. , J. Reprod. Fertil., in press.
33. Heller, C.J. In Estrogen Workshop Conference, ed. Paulsen, C.A.,
University of-Washington Press, Seattle, Washington, p 276, 1965.
34. Davies, A.G., Proc. Roy. Soc. Med., ~: 580, 1965.
35. Martin, F.I.R., J. Endocr., 38: 431, 1967.
36. Johnsen, S.G., Acta Endocr. (Kobenhavn) 22: 315, 1966.
37. Paulsen, C. A. In Proc. of Med. Vlth Pan-American Congress of
Endocrinology, Mexico City, 388, 1965.
38. Lytton, B. and Kase, N., New Eng. J. Med., ~: 1061, 1966.
39. Joel, C.A., Harefuah, 11: 281, 1966.

40. Lunenfeld, B., Mor, A. and Mani, M., Fertil. Steril., ~: 581,
1967.
41. Schoysman, F., Bulletin de la Societe Royale BeIge de Gynecologie
et d'Obstetrique, 34: 399, 1964.
42. Lytton, B. and Mro;;h, A., Fertil. Steril., 12: 696, 1966.
43. Abelli, A. and Falagario, M., Attualita Ostet. Ginec., 11: 723,
1966.
44. Cittadini, E. and Quartaravo, P., Agg. Ost. Gin. Palermo, 1,
1966.
45. Polishuk, W.Z., Palti, Z. and Laufer, A., Fertil. Steril., 18:
127, 1967.
46. Mroueh, A., Lytton, B. and Kase, N., J. Clin. Endocr., 12: 53,
1967.
47. Danesis, J.M. and Batrinos, M.L., Fertil. Steril., ~: 788,
1967.
48. Debaisi , E. and Misurale, F., Monit. Ost. Ginecol. Endocr. e
del Metab., ~: 5, 1967.
49. DeKretser, D.M., Taft, H.P., Brown, J.P., Evans, J.H. and Hudson,
B., J. Endocr., 40: 107, 1968.
50. Makler, A., Paldi, A., Peretz, A. and Barzalai, D., Harefuah,
11: 102, 1968.
DISCUSSICN
A. STEINBERGER: Dr. Lunenfeld, you were somewhat concerned about not

getting an increase in numbers of Sertoli cells in the 14-day or older
mouse when treated with antigonadotropic sera. I would like to point
out that in the rat, the Sertoli cells completely cease to divide
after about the 13th day of life. We did a study with a 3» thymidine
pulse at various ages and found that the percent of labelled Sertoli
cells decreases from about 30% at birth to zero at about 13 days of
age. Similar results were also obtained in culture. Considering that
in the mouse, spermatogenesis takes only about 36 days as compared
with about 48 days in the rat, I would assume that the Sertoli cells
stopped dividing in the mouse even earlier. Once the Sertoli cells
stop dividing nothing seems to be able to stimulate their division
and increase their number.
ROSS: Dr. Lunenfeld, is it true that testicular weight expressed as

a function of body weight remained constant over the period of 14 to
30 days in animals treated with antigonadotropic serum? Were these
animals gaining weight during this period?
LUNENFELD: The weight increase in these animals was normal. We con-

sider this information as indicative that the antiserum is not con-
taminated with anti growth hormone.
ROSS: How do you account for the gain in gonadal weight if the anti-
gonadotropic serum was entirely effective in neutralizing endogenous
gonadotropin?
LUNENFELD: We said that the testis weight as a percentage of the

animal weight did not increase. Absolute testis weight increased
but this can be expected for growing animals. A general weight in-
crease does not have to be gonadotropin dependent.
ROSS: If you had eliminated gonadotropic activity by hypophysecto-

m~z~ng these animals would you have expected the gonads to increase
in weight?
LUNENFELD: To my knowledge hypophysectomy at birth has never been

attempted. Furthermore, hypophysectomy would deprive the animal
also of growth hormone and other functions. In this experiment we
intended to deprive the animal only of gonadotropic hormones.
CLERMONT: Concerning the Sertoli cells in growing rats, we found

them proliferating until about 20 to 21 days of age.
E. STEINBERGER: Dr. Lunenfeld, I think that your experiments are

exceedingly interesting because this approach permits us to examine
the effect of gonadotropin deprivation on spermatogenesis in the
presence of the pituitary. Experiments using ablation of the pitu-
itary have many drawbacks. However, we should not forget that in
the type of experiments you reported we are looking only at the
mechanisms concerned with initiation of spermatogenesis rather than
maintenance, and I think these are different. Furthermore, I wish
you had done a quantitative analysis of the germinal epithelium us-
ing slightly different techniques which would have given us more
precise information.
LUNENFELD: This was a progress report of an experiment which is

still going on. All the material will be analyzed. I am sure that
more information will be derived from this experiment at its final
stage.
HUDSON: Dr. Lunenfeld, was the anti-rat gonadotropin an antiserum

to crude pituitary extract or specifically to LH or FSH?
LUNENFELD: This was an antiserum prepared to a gonadotropin extract

obtained from rat pituitary glands. It neutralizes both FSH and LH
activity. The characteristics of this antiserum were published in
Acta Endocrinology, 1967, 59: 311.
ROSEMBERG: Could we turn our attention to a discussion of gonado-
tropin therapy in males with idiopathic oligospermia and assumed
normal pituitary function?
GENERAL DISCUSSION: Treatment of Male "Infertility"
JOHNSEN: We have tried to select a group of men in whom we had eve-

ry good reason to believe that the cause of oligospermia was an FSH
deficiency. We think that we can select these patients from the
characteristics of their testicular biopsy. We also perform FSH as-
says. With respect to dosages used in the treatment of male infer-
tility, we know fairly well how much FSH is excreted in the urine.
Hence it is possible to calculate how much FSH will have to be ad-
ministered to replace the amounts of FSH secreted by the pituitary
in the normal male. We proceeded along this line. Despite the fact
that we treated men with an assumed hormonal deficiency, which was
substituted by the administration of HMG, we did not observe the
desired results. However, it is conceivable that the FSH dose ad-
ministered (160 IU 2 to 3 times per week) was not sufficient.
LUNENFELD: I think that we cannot use the comparative amount of go-

nadotropin excreted in order to set up dosages. We have to take into
account also the half-life and the transport of the hormone from the
site of injection to the target organ. Dr. Ross has informed me that
their data related to pituitary production of FSH in normal women
during the follicular phase of the cycle range from 200 to 800 IU per
day. They do not have comparable data on normal men.
TROEN: I think the point is to distinguish, as we know in clinical

medicine, between physiologic and pharmacologic doses. I think it
should be qUite apparent that there are at least two possibilities
in the approach to therapy: one would be to replace an amount of
gonadotropin that the body lacks, but with an otherwise responsive
end organ. It is clear that this approach has not been successful
in a large number of normogonadotropic males. Another possibility
open to us is to give pharmacological doses of HMG which we know to
be an effective agent. The dosage of HMG that we have given was not
in excess of that which we would expect to be normally produced either
by our knowledge of production rates or by studying blood levels.
Therefore, my suggestion is to give larger doses in order to attempt
to achieve a pharmacological action upon the germinal epithelium.
631
632 ADULT SEMINIFEROUS TUBULAR FAILURE
E. STEINBERGER: I fully agree with Dr. Troen. The fact is that one
of the most important problems is the ability to make a proper diag-
nosis and to understand the pathophysiology. If we had the means to
establish the pathophysiology of each testicular abnormality seen in
our offices, we could offer therapy in a rational fashion either by
replacement therapy or using gonadotropins in pharmacologic dosages.
Therefore, I think that it is of extreme importance to learn how to
diagnose the various types of testicular disorders.
PAULSEN: Dr. Steinberger, you indicated previously that good results

were achieved with HMG therapy in two patients. One was an example
of hypopituitarism while the second patient had primary testicular
disease with normal pituitary function. Would you please repeat the
details with respect to the second patient? We could then try to
characterize those patients who might respond favorably to HMG ther-
apy.
E. STEINBERGER: I think this particular patient was an example of

abnormal steroid biosynthetic activity in the testis. While the bio-
synthesis of steroids was proceeding, a large amount of the material
was shunted to a biologically inactive steroid (pregnenolone and
androstenediol). By increasing the total biosynthetic activity of
the Leydig cells, using pharmacologic doses of gonadotropins, we in-
creased the total steroid synthesis and thus also increased the tes-
tosterone output. I would assume that gonadotropins were responsible
for the effects observed.
ROSEMBERG: From the experimental evidence presented during this

meeting, it seems that initiation of spermatogenesis may be accom-
plished by the use of mixtures of FSH and LH with a ratio favorable
to LH. Would you agree, Dr. Lunenfeld, that, if we were to have a
preparation of gonadotropins with a mixture of FSH and LH, with a
ratio favorable to LH, this type of preparation will give better clin-
ical results. Of course, we should have to consider the type of pa-
tient to be treated with such mixtures.
LUNENFELD: Dr. Rosemberg, why would you think it worthwhile to add

to the Pergonal preparation a high amount of LH? I think it would
be just as good to use the existing Pergonal preparation and inject
HCG simultaneously, in the specific amounts necessary for each pa-
tient, monitored by testosterone levels.
ROSEMBERG: I think this is of practical importance. Dr. Donini can

tell us how easy it is for him to prepare an LH:FSH combination in
an HMG preparation with an LH:FSH ratio of 3 to 1, for example.
Would it be just the same, Dr. Donini, to add HCG instead of increas-
ing the amount of urinary LH in it?
GENERAL DISCUSSION 633
DONINI: I think that for some small clinical trials, it would be

possible to prepare ampoules with a ratio of LH to FSH of 3 to 1 for
example, but it would be difficult to have available such an amount
of LH for large clinical trials.
PAULSEN: I fail to see where there is any data to suggest that we

need more LH in the patient with adult seminiferous tubule failure.
It appears that we have to explore the effect of FSH in those patients
who do not exhibit a disturbed biosynthetic pathway and who are andro-
genically normal. Furthermore, we need to study the pharmacologic
effect of FSH on the abnormal germinal epithelium.
ROSEMBERG: If we accept the theoretical implication that the local

effect of testosterone is very important, then, if a preparation
with a higher amount of HCG is administered, the amount of testos-
terone secreted by the Leydig cells will be increased. This may be
necessary.
TROEN: I stressed the finding, which to us was surpr1s1ng, of low

values of plasma testosterone in patients who are otherwise clin-
ically normal, to point out that there is great variation in pa-
tients. To try to make the right combination of FSH and LH in ad-
vance might therefore be difficult. More to the point would be to
obtain a complete profile of each and every patient with regard to
what his actual function may be, to find out how this may vary from
whatever norms we may have and then apply the proper therapy.
E. STEINBERGER: I agree with Dr. Troen. However, it is possible

that some patients may exhibit local "intratesticular" androgen de-
ficiency. Thus circulating levels and production rates of testos-
terone may not necessarily tell us whether in these patients the
levels of testosterone at the target cell are adequate.
HELLER: May I remind you that we already have a circumstance in

the human where we can increase LH, FSH and testosterone and concom-
itantly increase spermatogenesis. This is accomplished by the ad-
ministration of clomiphene. We have all three hormones elevated,
so I do not think you should abandon the idea that LH may not be
important in this mixture.
PAULSEN: One final comment about HCG administration. We have gone

through some 20 years of clinical studies which have dealt with
treating oligospermic males with HCG. I have not seen any evidence
to indicate that this has been beneficial.
ROSEMBERG: These were high doses of HCG.
PAULSEN: Not necessarily so. Many investigators have administered

small doses of HCG. In some instances HCG alone and in other in-
634 ADULT SEMINIFEROUS TUBULAR FAILURE
stances HCG plus exotic substances such as liver extract were ad-
ministered.
MACLEOD: The emphasis during this discussion has been placed on

evaluating the response to pharmacologic doses of HMG. I would pre-
fer the contrary, let us say administering 15 instead of 75 IU.
Some of the evidence presented today indicated that high dosages of
FSH were not necessary.
LUNENFELD: Dr. Heller, I think that it was shown earlier during

this meeting by Dr. Ross that in hypogonadotropic patients clomiphene
had no effect on FSH and LH levels. Dr. MacLeod, have you any evi-
dence that with 15 IU of FSH you can obtain any response or is this
speculation in your part?
MACLEOD: No, I have no direct evidence. All I know is that a 75 IU

dose produced full spermatogenesis in the hypogonadotropic male.
Indeed I think that we are giving too much in view of the fact that
one can maintain spermatogenesis with what we presume to be extremely
low levels of FSH. Therefore, I would prefer to start at much lower
levels of FSH in hypophysectomized male subjects.
E. STEINBERGER: I have to be the devil's advocate and re-emphasize

Dr. MacLeod's statement. There is some evidence in the lower species
which supports his comments considering the possible damaging ef-
fect of high levels of gonadotropins. There is a paper by Dr. Dorner
showing the damaging effect of high doses of FSH in the rat. We
have found high doses of PMS to be damaging to spermatogenesis.
Thus, we should keep our minds open to the possibility that over-
dosing with gonadotropins may produce a result opposite to the one
we desire.
ROSEMBERG: We are forgetting that comparison with animal experiments

is not absolutely fair in this situation. Animals do not have ova-
ries or testes which have been exposed to any damage and they differ
from patients having some sort of defect in the ovaries or testes.
We are administering pharmacological doses to female patients who
do not ovulate on their own. They may secrete endogenous FSH and
LH but the end organ does not function. In the infertile male there
is a defect in testicular function. Hence, the consideration of
dosage should be related to the type of end organ which has to be
treated.
JOHNSEN: Dr. Lunenfeld requested me to indicate the characteristics

of the testicular tissue in hypopituitarism (FSH deficiency) in
adult males. There is a greater or minor inhibition of spermato-
genesis at the spermatid level so that, in the score count method,
the majority of the figures will be placed at steps 8, 7 and 6 with-
out finding any signs of degenerative processes, such as hyaliniza-
tion: Finally, there will be no Leydig cell hyperplasia.
GENERAL DISCUSSION 635
LUNENFELD: I think that we could summarize the problem in the fol-

lowing way. There seems to be no controversy as to our first task
which is to correctly diagnose the patient for gonadotropic therapy.
Unfortunately to date we have no clear cut way to arrive at the pro-
per diagnosis. Some of the participants presented improved ways for
selection of patients for HMG therapy; 2) I think all of us agree
that more experimental work is needed to elucidate the specific
pathological mechanism leading to oligo or azoospermia; 3) I believe
that we do not have the right to deprive a patient of possible bene-
fits which could be derived from gonadotropin therapy. Consequently,
the "empirical" choice of patients as well as improvement of methods
for monitoring gonadotropic therapy has to be continuously revised
as new data become available. I do believe that we are administering
low doses of gonadotropin. With respect to monitoring gonadotropin
therapy, I again suggest that the testosterone level has to be kept
slightly above normal. We also have to consider that hypogonado-
tropinism may be due to either too low endogenous gonadotropin
production or, that the target cell may be relatively insensitive
to endogenous gonadotropin stimulation. Such patients may need
larger dosages of HMG because endogenous gonadotropins which ap-
pear to be normal may be insufficient. Furthermore, sperm counts,
especially in oligospermic patients, vary. A sufficient number of
pre-treatment sperm examinations are necessary to assess a possible
response to gonadotropic therapy. Lastly it must be emphasized that
not all testicular pathology is of an endocrinological nature and
that gonadotropic therapy cannot be expected to be of value in all
conditions of testicular dysfunction.
CONCLUDING REMARKS
Although many topics were discussed during this three-day meet-

ing, it is evident that a Workshop Conference cannot be all inclu-
sive in its scope. In this regard, it should be noted that less
established genetic factors influencing male fertility were con-
sidered rather than entities caused by sex chromosomal aberrations.
It was also beyond the scope of this Conference to include a dis-
cussion on the surgical treatment of disorders leading to testicular
dysfunction such as cryptorchidism or varicocele.
The main thrust was directed toward placing present concepts

and hypothesis in a more appropriate perspective and in presenting
newer material which in some instances was embryonic in nature.
Finally, it was intended to blend the above with established data
so that guidelines for future investigations could be set forth.
Eugenia Rosemberg
C. Alvin Paulsen
637
INDEX
ADENOHYPOPHYSIS, see under pi- CHROMOSOMES, see sex chromo-

tuitary. somes,
in, infertility, 151 to 163.
ANDROGENS, see under individual meiotic prophase, 128 to
hormones. 135.
effect on seminal fluid,
fructose and citric acid, 519, CLOMIPHENE, effect on LH-FSH
522-524. levels in hypo gonadal men,
metabolism in testis, 439-454. 292-294.
ANDROSTENEDIONE, secretion by CORTICOTROPIN RELEASING FACTOR,

testis, 410. separation from other re-
synthesis in embryo, 6. leasing factors, 213.
ANTI-GONADOTROPIN SERA, effect

on testicular function, 616
to 618. E
ANTI-SEMINAL PLASMA ANTIBODIES, ELECTRON MICROSCOPY,

533, 534. of, Leydig cells, 75 to 87.
, seminiferous tubules, 63
AUTOIMMUNITY, 535, 536-538. to 73, 370, 373.
, Sertoli cells, effect of,
B LH, 372, 374.
BRAIN, hypothalamus, see under EMBRYOLOGY, cytogenetics in

FSH-RF and LH-RF. man~als, 139 to 147.
control of gonadotropin secre- of the testis in amphibia,
tion, 187 to 191. 3 to 9.
ESTROGENS, secretion by testis,

C 412 to 415.
syntheSis in embryo, 6.
CHORIONIC GONADOTROPIN, see urinary levels after gonado-
Human Chorionic Gonadotropin. tropin therapy, 597.
639
640 INDEX
F G
FEEDBACK CONTROL, of gonado- GENITAL TRACT, development,

tropin secretion, 198 to hormonal control of, 14, 15.
202.
spermatogenesis and gonado- GENETICS, germ-cell cytogenetics,
tropin, 231-242. l39 to 147.
relation to infertility, 151
FETAL TESTIS, see under testis. to 163.
FOLLICLE STIMULATING HORMONE, GERM-CELLS (PRIMORDIAL), cyto-

effect, on, immature testis, genetics in mammals, 139
355 to 363. to 147.
,on, spermatogenesis, developmental histology, 99,
578 to 582. 100.
,on, steroidogenesis in, amphibian testis, 3.
in testicular , human embryo, 19 to 23.
tissue, 463, 464. , mammalian testis, 12.
,on, testicular func-
tion, 563 to 574. GERMINAL EPITHELIUM, developmen-
,on, testicular metab- tal histology, 99 to 106.
olism, 301 to
309. GLUCOSE, effect on testis pro-
feedback, control of secre- tein biosynthesis, 319 to
tion, 198 to 202. 330.
levels, after clomiphene test,
290 to 294. GONADAL HORMONES, see under
, during testosterone individual hormones.
administration, 253,
254. GONADOTROPINS, see also under
, in normal males, 290 individual hormones. FSH,
to 294. HCG, LH and PMS.
standards, in bioassay and
radioimmunoassay,263 to 272. GONADS, hormonal factors in,
urinary, purification and development, 12, 13.
characterization, 278 to 282. sex differentiation, 13, 14.
FREEMARTINISM, in cattle GROWTH HORMONE, effect on tes-

medullary-cortical antag- ticular function, 387, 556
onism, 7. to 561.
FSH-RELEASING FACTOR, see FSH-RF.

chemical properties, 216, 217. H
FSH-RF, assay of, 209.
mechanism of, action, 171 to HUMAN CHORIONIC GONADOTROPIN,
182. biologic effect of, 381 to
, release, 187 to 389.
202. effect on, immature testis,
purification, 214, 215. 396 to 399.
INDEX 641
HUMAN CHORIONIC GONADOTROPIN effect of gonadotropin therapy

cont. on, 577 to 586, 591 to 601,
effect on, seminal fluid, 605 to 611, 613 to 626, 631
fructose and to 635.
citric acid, 517, immunology of, 529 to 538.
518. INTERSTITIAL CELL STIMULATING
, spermatogenesis, HORMONE, ICSH, see LH.
577 to 586
testicular func-
tion, 552-554, L
591 to 601, 608
to 610, 618 to LEYDIG CELLS, developmental
625. histology, 98,99.
effect of, LH and HCG in im-
HUMAN MENOPAUSAL GONADOTROPIN, mature testis,
effect on, immature testis, 385 to 387.
398, 399. , testosterone on,
, seminal fluid, 249 to 256
fructose and fetal, development of, 13, 14.
citric acid, 519, , differentiation in, 23,
521. 24.
, spermatogenesis, function, effect of gonadotro-
577 to 586. pins, 547 to 561.
testicular func- steroid secretion by, 407 to
tion, 547 to 554, 412.
591 to 601, 607 ultrastructure of, 75 to 87.
to 610, 618 to
625. LH-RELEASING FACTOR, see LH-RF.
extraction and purification,
279 to 284. LH-RF, assay of, 208, 209.
chemical properties of, 215,
HYPOGONADISM, effect urinary 216.
FSH and LH on,testicular mechanism of, action of, 171
function, 563 to 574. to 182, 221
to 227.
HYPOPHYSECTOMY, effect on, , release, 187 to
testicular function, 356, 202.
577 to 582, 618. purification, 209 to 214.
HYPOTHALAMUS, see under FSH-RF LUTEINIZING HORMONE, LH,

and LH-RF. effect on, immature testis, 355
role in genital tract develop- to 363.
ment, 16, 17. , seminiferous tubules,
369 to 377.
, steroidogenesis in
testicular tissue,
I 463, 464.
, spermatogenesis,
INFERTILITY, see also sterility. 578 to 583.
chromosome structure in, 151 , testicular function,
to 163. 554 to 556, 563 to 574.
642 INDEX
LUTEINIZING HORMONE, LH, cont. P

levels, after testosterone
administration, 249 PITUITARY, control of testicular
to 256 function, 15, 16.
, after clomiphene ad- feedback control of gonado-
ministration, 292 tropin secretion, 198, 202.
to 294. feedback mechanism with sper-
, in hypo gonadal males, matogenesis, 231 to 242.
290 to 294.
, in normal males, 290 PMS GONADOTROPIN, effect on,
to 294. immature testis, 397 to 398.
pituitary, biologic effect effect on, seminal fluid,
of, 381 to 389. fructose and
secretion, influence of LH-RF, citric acid,
221 to 227. 517, 520.
standards, in bioassay and , testicular func-
radioimmunoassay, 263 to tion, 606, 607.
272.
urinary, purification and PREGNENOLONE, biosynthesis in
characterization, 278 to 281. testis, 461, 462.
M
PREPUBERAL TESTES, see testis.
MALE GENITAL SYSTEM, hormonal effect of gonadotropins.
factors in fetal develop-
ment, 11 to 17. PROGESTERONE, metabolism
in, azoospermia, 444 to 447.
MALE INFERTILITY, see infertil- ,testicular tissue, 333 to
ity. 349.
MEIOSIS, functional significance

in male, 115 to 122. R
nuclear ultrastructure in
prophase, 127 to 135. RADIOIMMUNOASSAY, of FSH and
LH in diagnosis hypogonadism,
MENOPAUSAL GONADOTROPIN, see 289 to 295.
Human Menopausal Gonadotro- standards for, LH and FSH,
pin. 266 to 272.
MULLERIAN DUCT, development,

20, 21. S
inhibition by testis, 14, 31.
persistence in pseudoherma- SEMEN, effect of gonadotropin
phrodites, 24 to 26. therapy on, 593 to 595,
regression, 22, 23. 605 to 611.
ejaculated , biochemistry of,
o 473, 474.
epididymal, biochemistry of,
OLIGOSPERMIA, effect of gonado- 471 to 473.
tropin therapy on, 605 to fructose and citric acid of,
611, 623 to 625. 515 to 525.
INDEX 643
SEMEN, cont. SPERMATOCYTE, development in

immunology of, 533, 534. human testis, 49, 51 to 57.
sperm morphology of, 481 to nucleus, ultrastructure in
491. meiotic prophase, 127 to
testicular, biochemistry of, 135.
469 to 471. stages in meiosis, 116 to 122.
ultrastructure in human testis,
SEMINAL PLASMA ANTIGENS, 535, 66, 67.
536.
SPERMATOGENESIS, dynamics in
SEMINIFEROUS TUBULES, develop- human testis, 47 to 58.
mental histology of, 99 to effect of,gonadotropins in,
101. 547 to 561, 613 to
effect of, gonadotropins in 626.
immature testis, ,HCG, on, 577 to
355 to 363, 393 586.
to 400. ,HMG, on, 577 to 586.
, LH in immature ,immature testis,
testis, 369 to 355 to 363, 393 to
378. 400.
function, effect of gonado- ,urinary FSH, on,
tropins, 547 to 561. 563 to 574.
protein synthesis in, 315 to ,urinary LH, on, 563
330 to 574.
spermatogenesis in, 47 to 58. feedback mechanism with pi-
ultrastructural changes in, tuitary, 231 to 242.
63 to 73. in tissue culture, 336, 337.
suppression by testosterone,
SERTOLI CELLS, role in gonado- 249 to 256.
tropin inhibition, 239, 240.
steroid production of, 240, SPERMATOGONIA, see also germ
241. cells.
ultrastructure, effect of LH, Classification and develop-
372 to 377. ment in human testis, 48,
, of, human 51 to 57.
testis, 64,
65. SPERMATOZOA, in human semen,
biochemistry of, 469.
SEX CHROMOSOMES, characteriza- morphology, deviations, 481-491.
tion in mammals, 139 to 147. transport, relation to tes-
in, infertility, 151 to 163. ticular capsule, 506 to 512.
, meiosis, 116 to 122.
, meiotic prophase, 139 to SPERM COUNT, evaluation of
to 147. spermatogenesis, 231 to 242.
SPERMATID, development in human SPERMIOGENESIS, see also sper-
testis, 50 to 55. matid.
ultrastructure in human
testis, 67 to 68. STERILITY, see also infertility.
644 INDEX
STEROIDS, see under individual protein biosynthesis, in, 315

hormones. to 330.
steroid, metabolism, in, 439
to 454, 459 to 464.
T , secretion by, 407 to
415.
TESTICULAR CAPSULE, histology suppression of spermatogenesis
of, 506 to 508. by testosterone, 249 to 256.
pharmacology of, 497 to 506. testicular capsule relation
relation to sperm transport, to transport, 510 to 512.
510 to 512. tissue culture, spermato-
genesis and steroid metab-
TESTIS, biopsy, evaluation of olism, 333 to 350.
spermatogenesis, 232 to 238. ultrastructure, of Leydig
developmental histology, 95 cells, 75
to 106. to 87.
effect of, FSH on metabolism ,seminiferous
of, 301 to 309. tubules, 63
, LH on, 383 to 384. to 73.
fetal, differentiation, 19 to
29. TESTOSTERONE, effect of seminal
influence of steroids fluid, fructose and citric
on regression, 31 acid, 519, 522.
to 37. plasma levels, after gonado-
function, influence of FSH and tropin ther-
LH, 563 to 574. apy, 595-597.
influence of gonado- ,in, infancy,
tropins and HGH, childhood,
547 to 561. adolescence
influence of HMG and and old age,
HCG, 547 to 561, 426 to 428.
577 to 586, 591 to ,in,non-testic-
601. ular dis-
germ cell cytogenetics, 139 orders, 430
to 147. to 433.
histological grouping in sper- ,in,normal males,
matogenesis, 47 to 58. 423 to 426.
immature, effect of, gonado- ,in, testicular
tropins on, 355 disorders,
to 363, 396 to 428 to 430.
399. role in, spermatogenesis
effect of HCG on, suppression, 249
385 to 387. to 256.
, effect of LH on, , FSH-LH production,
385 to 387. 249 to 256.
immunology of, 529 to 533. secretion by testis, 407 to
pharmacology, in vivo, 508 410.
to 510. synthesis, from progesterone
, of, testicular in testicular
capsule, 497 tissue, 439 to 454.
to 506. in embryo, 6.
INDEX 645
THYROTROPIN RELEASING FACTOR, TISSUE CULTURE, cont.

separation from other re- steroid metabolism,in, human
leasing factors, 213. testicular tissue, 333-350.
TISSUE CULTURE, of human tes-

ticular tissue, 333 to 350. W
steroid metabolism, after go-
nadotropin therapy, 597 to WOLFFIAN DUCT, development, 20
600. to 22, 31 to 37.

Advances in Experimental Medicine and Biology 10

Uploaded by

Copyright:

Available Formats

Advances in Experimental Medicine and Biology 10

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Advances in Experimental Medicine and Biology 10

Uploaded by

Copyright:

Available Formats

The Human Testis

ADVANCES IN EXPERIMENTAL MEDICINE AND BIOLOGY

Rodolfo Paoletti Institute of Pharmacology, University of Milan, Milan, Italy

A SERONO FOUNDATION SYMPOSIUM

~ PLENUM PRESS . NEW YORK-LONDON . 1970

First Printing - September 1970

Library of Congress Catalog Card Number 70·129058

This volume describes the proceedings of the Workshop Conference

The format of the book has been arranged according to topics

We express our thanks to our Publisher, Plenum Press, for their

May, 1970 Eugenia Rosemberg

Individual participants wish to thank Editors and Publishers

Individual participants desire to express their gratitude to

Grants supporting work presented in these proceedings are ac-

This volume contains the Proceedings of this Workshop which served

special occasion our feelings of sympathy and admiration. Since

Apart from the advanced studies on the relationship between Leydig

Unfortunately, an extension of these approaches is restricted by

This is precisely what this Workshop Conference is intended to

i.e., immunological factors, chromosomal anomalies of germinal cells,

Finally, the aim of this Workshop Conference was not only to

Professor Mario Hector Burgos, M. D.

Ann C. Chandley, Ph. D.

A. Kent Christensen, Ph. D.

Professor Yves Clermont, Ph. D.

Joseph R. Davis, M. D., Ph. D.

Professor Egon R. Diczfalusy, M. Do

C. E. Ford, D. Sc., F.R.S.

Irving I. Geschwind, Ph. D.

Carl G. Heller, M. D., Ph. D.

Bryan Hudson, Mo D., Ph. D.

Jan E. Jirasek, M. D., D. Sc.

Professor Alfred Jost

Marian Jutisz, Ph. D.

Juan Carlos Lavieri, M. D.

Professor Bruno Lunenfeld, M. D.

John MacLeod, Ph. D.

Professor Roberto E. Mancini, Mo D.

Professor Thaddeus Mann, M. D., Sc. D., Ph. D., F.R.So

Anthony R. Means, Ph. D.

Susumu Ohno, D.V.M., Ph. D., D.Sc.

Griff T. Ross, M. Do, Ph. D.

Professor Kenneth Savard, D. Sc.

Professor Carl Schirren, Mo Do

Anna Steinberger, Ph. D.

Emil Witschi, M. D., Ph. D.

EMBRYOLOGY OF THE MALE REPRODUCTIVE TRACT

Embryology of the Testis . . • • . . . . . . . . 3

Hormonal Factors in the Development of the Male

The Relationship Between Differentiation of the

Heteroplastic Gonaduct-Testes Combinations:

HISTOLOGY OF THE TESTIS

Dynamics of Human Spermatogenesis 47

Electron Microscopy of the Human Seminiferous Tubules 63

Fine Structure of Testicular Interstitial Cells in Humans 75

Histology of the Human Testis from Neonatal Period to

Morphological Aspects of Meiosis and Their Genetical

Ultrastructure and Histochemistry of the Nucleus During

The Cytogenetics of the Male Germ Cells and the Testis