(Biotechnology in Agriculture and Forestry 30) V. Nuti Ronchi, L. Giorgetti (auth.), Professor Dr. Y. P. S. Bajaj (eds.)-Somatic Embryogenesis and Synthetic Seed I-Springer-Verlag Berlin Heidelberg (1.pdf

Volumes already published
Volume 1: Trees I (1986)

Volume 2: Crops I (1986)
Volume 3: Potato (1987)
Volume 4: Medicinal and Aromatic Plants I (1988)
Volume 5: Trees II (1989)
Volume 6: Crops II (1988)
Volume 7: Medicinal and Aromatic Plants II (1989)
Volume 8: Plant Protoplasts and Genetic Engineering 1(1989)
Volume 9: Plant Protoplasts and Genetic Engineering II (1989)
Volume 10: Legumes and Oilseed Crops I (1990)
Volume 11: Somaclonal Variation in Crop Improvement I (1990)
Volume 12: Haploids in Crop Improvement I (1990)
Volume 13: Wheat (1990)
Volume 14: Rice (1991)
Volume 15: Medicinal and Aromatic Plants III (1991)
Volume 16: Trees III (1991)
Volume 17: High-Tech and Micropropagation I (1991)
Volume 18: High-Tech and Micropropagation II (1992)
Volume 19: High-Tech and Micropropagation III (1992)
Volume 20: High-Tech and Micropropagation IV (1992)
Volume 21: Medicinal and Aromatic Plants IV (1993)
Volume 22: Plant Protoplasts and Genetic Engineering III (1993)
Volume 23: Plant Protoplasts and Genetic Engineering IV (1993)
Volume 24: Medicinal and Aromatic Plants V (1993)
Volume 25: Maize (1994)
Volume 26: Medicinal and Aromatic Plants VI (1994)
Volume 27: Somatic Hybridization in Crop Improvement I (1994)
Volume 28: Medicinal and Aromatic Plants VII (1994)
Volume 29: Plant Protoplasts and Genetic Engineering V (1994)
Volume 30: Somatic Embryogenesis and Synthetic Seed I (1995)
Volume 31: Somatic Enbryogenesis and Synthetic Seed II (1995)
Volumes in preparation
Volume 32: Cryopreservation of Plant Germplasm I (1995)
Volume 33: Medicinal and Aromatic Plants VIII
Volume 34: Plant Protoplasts and Genetic Engineering VI
Volume 35: Trees IV
Volume 36: Somaclonal Variation in Crop Improvement II
Volume 37: Medicinal and Aromatic Plants VIII
Volume 38: Plant Protoplasts and Genetic Engineering VII
Biotechnology in
Agriculture and Forestry 30
Somatic Embryogenesis
and Synthetic Seed I
Edited by YP.S. Bajaj
With 164 Figures and 54 Tables
Springer-Verlag Berlin Heidelberg GmbH

Professor Dr. Y.P.S. BAlAl
A-137
New Friends Colony
New Delhi 110065, India
ISBN 978-3-642-08183-5 ISBN 978-3-662-03091-2 (eBook)

DOI 10.1007/978-3-662-03091-2
CIP data applied for

This work is subject to copyright. AII rights are reserved, whether the whole or par! of the material is
concerned, specifically the rights of translation, reprinting, reuse of iIlustrations, recitation, broadcasting,
reproduction on microfilms or in any otherway, and storage in data banks. Duplication ofthis publication
or parts thereofis permitted only under the provisions ofthe German Copyright Law of September 9, 1965,
in its current version, and permission for use must always be obtained from
Sprioger-VerJag Berlin Heidelberg GmbH
Violations are liable for prosecution under German Copyright Law.
© Springer-Verlag Berlin Heidelberg 1995
The use of general descriptive names, registered narnes, trademarks, etc. in this publication does not imply,
even in the absence of a specific statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.
Dedicated to
Professor Dr. Jacob Reinert
with whom I had the privilege of working
at the Freie Universitiit, Berlin, Germany
from 1974-1976
Preface
While working in the laboratory of Professor Dr. Jacob Reinert at the Freie
Universitat Berlin (1974-1976), I had the opportunity to become deeply
involved in studying the intricacies of the fascinating phenomenon of
somatic embryogenesis in plant cells and protoplasts. In numerous stimu-
lating discussions with Professor Reinert on this subject, I was fully
convinced that somatic embryogenesis would become one of the most
important areas of study, not only regarding basic and fundamental
aspects, but also for its application in crop improvement. During the last
decade, we have witnessed tremendous interest and achievements in the
use of somatic embryos for the production of synthetic seeds, for micro-
propagation, genetic transformation, cryopreservation, and conservation
of germplasm. The en masse production of somatic embryos in the
bioreactors has facilitated some of these studies. Somatic embryos have
now been induced in more than 300 plant species belonging to a wide range
offamilies. It was therefore felt that a compilation ofliterature/state of the
art on this subject was necessary. Thus, two volumes on Somatic Embryo-
genesis and Synthetic Seed have been compiled, which contain 65 chapters
contributed by International experts.
Somatic Embryogenesis and Synthetic Seed I comprises 31 chapters,

arranged in 3 sections:
Section ICommitment of the cell to somatic embryogenesis; early

events; anatomy; molecular basis; gene expression; role of
polyamines; machine vision analysis of somatic embryos.
Section II Applications of somatic embryos; technology of synthetic
seed; fluid drilling; micropropagation; genetic transfor-
mation through somatic embryos; cryopreservation.
Section III Somatic embryogenesis in various tree species of Aesculus,
Betula, Carica, Citrus, Cocos, Corylus, Elaeis, Hevea, Jug-
lans, Larix, Liriodendron, Magnolia, Olea, Picea, Populus,
and Theobroma.
Somatic Embryogenesis and Synthetic Seed II contains 34 chapters,
arranged in 4 sections:
Section I Cereal and Grasses - wheat, rice, maize, rye, oat, ryegrass,
fescue, orchardgrass, bluestem grasses, sugarcane.
VIII Preface
Section II Vegetables and Fruits - asparagus, chicory, cucurbits,

cucumber, okra, carrot, banana.
Section III Legumes and Oilseed Crops - peanut, soybean, cotton, white
mustard, meadowfoam.
Section IV. Ornamental, Medicinal and Miscellaneous Plants - daylily,
freesia, celery, coriander, Coptis, Panax, Rauwolfia, Arabi-
dopsis, Bellevalia, Brimeura, Dendrophthoe, Rumex,
Fagopyrum, and Ranunculus.
These books will be of interest to students, teachers, and research

workers in the field of botany, horticulture, forestry, tissue culture, general
plant biotechnology, and to those involved in micropropagation.
New Delhi, January 1995 Professor y.P.S. BAJAJ

Series Editor
Contents
Section I Basic and Fundamental Aspects

of Somatic Embryogenesis
1.1 The Cell's Commitment to Somatic Embryogenesis

V. NUTI RONCHI and L. GIORGETII (With 2 Figures)
1 Introduction ......................................... 3
2 Comparison Between Carrot Zygotic
and Somatic Embryogenesis ............................ 4
3 Somatic Chromosome Segregational Events . . . . . . . . . . . . . . . . 6
4 Evidence of the Occurrence ofSegregational Events. . . . . . . . . . 9
5 Cellular Competence for DNA Reprogramming ............ 13
6 Concluding Remarks .................................. 15
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.2 Early Events in Embryogenesis

F. Lo SCHIAVO (With 2 Figures)
1 Introduction ......................................... 20
2 Characteristics of Cell Lines
and Morphogenetic Competence . . . . . . . . . . . . . . . . . . . . . . . . . 21
3 Importance of the Primary Explant. . . . . . . . . . . . . . . . . . . . . . . 21
4 Biological Significance of Modulation . . . . . . . . . . . . . . . . . . . . . 23
5 Biochemical Identification of the Various Classes of ABPs .... 25
6 Cellular Mechanisms That Generate Totipotency . . . . . . . . . . . . 26
7 Summary and Conclusions ............................. 28
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
I.3 Molecular Basis of Somatic Embryogenesis

R. KAWAHARA and A. KOMAMINE (With 7 Figures)
1 Introduction ......................................... 30
2 Expression of Polarities in Early Stages
of Somatic Embryogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3 Molecular Aspects of Somatic Embryogenesis .............. 34
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
x Contents
1.4 Gene Expression in Somatic Embryos

H.D. WILDE, W.S. SEFFENS, and T.L. THOMAS (With 3 Figures)
1 Introduction ......................................... 41
2 Embryogenesis from Somatic Carrot Cells ................. 41
3 Patterns of Gene Expression
During Carrot Somatic Embryogenesis . . . . . . . . . . . . . . . . . . . . 43
4 Conclusions ......................................... 48
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
1.5 Role ofPolyamines in Somatic Embryogenesis

S.c. MINOCHA and R. MINOCHA (With 1 Figure)
.1 Introduction ......................................... 53
2· Polyamine Metabolism ................................ 54
3 Polyamines and Somatic Embryogenesis . . . . . . . . . . . . . . . . . . . 56
4 Polyamines in Carrot .................................. 58
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
1.6 Anatomy of Somatic Embryogenesis

G. SCHUMANN, U. RYSCHKA, J. SCHULZE, and E. KLOCKE
(With 15 Figures)
1 Introduction ......................................... 71
2 Embryoid Induction and First Cell Division . . . . . . . . . . . . . . . . 72
3 Maturation and Growth ............................... 82
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
1. 7 Machine Vision Analysis of Plant Cells

and Somatic Embryos
M.A.L. SMITH (With 5 Figures)
1 Overview............................................ 87
2 Challenge of Cell Culture Analysis ....................... 88
3 Advantages of Machine Vision .......................... 89
4 Approaches to Vision Analysis
of Cell and Somatic Embryo Culture ..................... 91
5 Potential for Machine Vision
in Culture System Automation .......................... 97
6 Remaining Challenges for Image Analysis
of Somatic Embryos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
References ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 100
Contents XI
Section II Applications of Somatic Embryos; Technology

of Synthetic Seed; Fluid Drilling; Micropropagation
and Genetic Transformation Through Somatic Embryos;
Cryopreservation
11.1 Somatic Embryogenesis and Its Applications

for Crop Improvement
Y.P.S. BAJAJ (With 12 Figures)
1 General Account ..................................... 105
2 Applications of Somatic Embryogenesis ................... 114
3 Summary ........................................... 118
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 119
11.2 Somatic Embryogenesis and the Technology

of Synthetic Seed
D.J. GRAY, M.E. COMPTON, R.C. HARRELL,
and D.J. CANTLIFFE
1 Introduction ......................................... 126
2 Somatic and Zygotic Embryo Development ................ 127
3 Genetic Variation from Cell Culture ...................... 131
4 Structural Aspects of Synthetic Seed ...................... 132
5 Automation of Synthetic Seed Production ................. 138
6 Estimated Cost of Synthetic Seed ........................ 141
7 Crop Applications for Synthetic Seed ..................... 142
8 Conclusion .......................................... 145
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 145
11.3 Role of Maturation and Desiccation of Somatic Embryos

in the Production of Dry Artificial Seed
B.D. McKERSIE, S. VAN ACKER, and F.-M. LAI (With 5 Figures)
1 Introduction ......................................... 152
2 Concepts of Artificial Seeds ............................. 153
3 Somatic Embryogenesis in Alfalfa - a Model
for the Development of Dry Somatic Embryos . . . . . . . . . . . . .. 154
4 The Water Replacement Hypothesis. . . . . . . . . . . . . . . . . . . . .. 164
5 Conclusions ......................................... 166
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 167
11.4 Artificial Seed Production Through Encapsulation

of Hairy Root and Shoot Tips
N. UOZUMI and T. KOBAYASHI (With 4 Figures)
1 General Introduction .................................. 170
2 Materials and Methods ................................ 171
3 Results ............................................. 171
XII Contents
4 Discussion .......................................... 178

References ........................................... " 180
II.5 Fluid Drilling as a Delivery System

for Somatic Embryo-Derived P1antlets
S.L. KITTO, W.G. PILL, and D.M. MOLLOY (With 5 Figures)
1 Introduction ............... . . . . . . . . . . . . . . . . . . . . . . . . .. 181
2 Somatic Embryo Development .......................... 182
3 Fluid Drilling ........................................ 182
4 Post-Fluid Drilling Environment. . . . . . . . . . . . . . . . . . . . . . . .. 183
5 Fluid Drilling as a Delivery System
for Somatic Embryo-Derived Plantlets of Carrot ............ 183
7 Protocol ............................................ 189
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 191
II.6 Micropropagation Through Somatic Embryos

P.D. DENCHEV and A.I. ATANASSOV (With 5 Figures)
1 Introduction ......................................... 193
2 A System for Direct Somatic Embryogenesis in Alfalfa ....... 195
3 Future Trends ...................................... " 199
4 Conclusions and Prospects . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 203
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 204
II. 7 Genetic Transformation of Somatic Embryos

D. ELLIS
1 General Account ..................................... 207
2 Transformation of Somatic Embryos ..................... 211
3 Transformation of Picea glauca Using Somatic Embryos ..... 215
4 Conclusions ......................................... 217
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 218
II.8 Cryopreservation of Somatic Embryos

y.P.S. BAJAJ (With 1 Figure)
1 Introduction ......................................... 221
2 Potential of Freeze Preservation of Somatic Embryos ........ 221
3 Methods for the Cryopreservation of Embryos ............. 222
4 Studies on the Cryopreservation of Somatic Embryos ........ 225
5 Conclusions ......................................... 227
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 227
Contents XIII
Section III Somatic Embryogenesis in Trees
III. 1 Somatic Embryogenesis in Horse Chestnut

(Aesculus hippocastanum L.)
P. PROFUMO and P. GASTALDO (With 6 Figures)
1 Introduction ......................................... 233
2 Somatic Embryogenesis ................................ 235
3 Summary ........................................... 244
4 Protocol ............................................ 244
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 244
III.2 Somatic Embryogenesis in Birches (Betula spp.)

A.M. NUUTILA, U. KURTEN, R. PUUPPONEN-PIMIA, J. HAMALAINEN,
L. MANNONEN, and V. KAUPPINEN (With 6 Figures)
1 Introduction ......................................... 246
2 Somatic Embryogenesis in Betula ........................ 246
3 Summary ........................................... 256
4 Protocol ............................................ 257
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 258
IIL3 Somatic Embryogenesis in Papaya (Carica papaya L.)

M.M.M. FITCH (With 5 Figures)
1 General Account ..................................... 260
2 Somatic Embryogenesis in Papaya ....................... 267
3 Conclusions ......................................... 274
4 Protocols............................................ 275
References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 276
III.4 Somatic Embryogenesis in Citrus Species

H. KUNITAKE and M. MIl (With 9 Figures)
1 Introduction ......................................... 280
2 Induction of Embryogenic Calli . . . . . . . . . . . . . . . . . . . . . . . . .. 283
3 Induction of Somatic Embryos (Embryoids) ............... 287
4 Plant Regeneration from Embryoids . . . . . . . . . . . . . . . . . . . . .. 290
5 Histological Observation of Somatic Embryogenesis ......... 291
6 Stability of Embryoid-Derived Plants ..................... 293
8 Protocol ............................................ 295
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 296
III. 5 Somatic Embryogenesis in Coconut (Cocos nucifera L.)

J.L. VERDEIL and J. BUFFARD-MoREL (With 10 Figures)
1 General Account ..................................... 299
2 Achieving Embryogenesis in Coconut. The Different Stages ... 302
XIV Contents
3 Discussion and Conclusion ............................. 313

4 Protocol ............................................ 314
References ................................... . . . . . . . . .. 315
I1L6 Somatic Embryogenesis in Hazelnut (Corylus Species)

B. BERROS, M. REY, C. DiAZ-SALA, M. ALBUERNE,
and R. RODRiGUEZ (With 8 Figures)
1 Introduction ......................................... 318
3 Histological Studies ................................... 328
References .... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 333
III.7 Somatic Embryogenesis in Oil Palm (Elaeis guineensis Jacq.)

Y. DUVAL, F. ENGELMANN, and T. DURAND-GASSELIN
(With 9 Figures)
1 Introduction ......................................... 335
2 Somatic Embryogenesis and Regeneration in Oil Palm ....... 338
3 Conclusions and Prospects . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 349
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 350
I1L8 Somatic Embryogenesis in Rubber Tree

(Hevea brasiliensis Miill. Arg.)
M.P. CARRON, H. ETIENNE, N. MICHAUX-FERRIERE,
and P. MONTORO (With 6 Figures)
1 Introduction ......................................... 353
4 New Protocol for Somatic Embryogenesis in Hevea . . . . . . . . .. 367
References .... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 368
I1L9 Somatic Embryogenesis in Walnut (Juglans Species)

W. TULECKE, G.H. MCGRANAHAN, AND C.A. LESLIE
(With 6 Figures)
1 Introduction ......................................... 370
3 Summary and Conclusion .............................. 375
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 376
III. 10 Somatic Embryogenesis in Western Larch

(Larix occidentalis)
P. VON ADERKAS, R.G. THOMPSON, M. ZAKI, and L. BENKRIMA
(With 7 Figures)
1 Introduction ......................................... 378
Contents XV

4 Protocol for the Induction of Somatic Embryos .. . . . . . . . . . .. 386
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 387
111.11 Somatic Embryogenesis in Magnoliaceae

(Liriodendron and Magnolia)
S.A. MERKLE (With 4 Figures)
1 Introduction ......................................... 388
4 Protocol for the Induction of Somatic Embryos. . . . . . . . . . . .. 401
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 402
111.12 Somatic Embryogenesis in Olive (Olea europaea L.)

E. RUGINI, A. PEZZA, M. MUGANU, and G. CARICATO
(With 3 Figures)
1 Introduction ......................................... 404
3 Summary and Conclusion .............................. 411
4 Protocol for the Induction of Somatic Embryogenesis ........ 411
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 413
111.13 Somatic Embryogenesis in Norway Spruce (Picea abies)

S. VON ARNOLD, D. CLAPHAM, U. EGERTSDOTTER,
I. EKBERG, H. Mo, and H. YIBRAH (With 5 Figures)
1 Introduction ......................................... 415
3 Conclusions ......................................... 428
References ............................................ , 429
III.14 Somatic Embryogenesis in Black Spruce

[Picea mariana (Mill.) B.S.P.] and Red Spruce (P. rubens Sarg.)
L. TREMBLAY and F.M. TREMBLAY (With 8 Figures)
1 Introduction ......................................... 431
4 Protocol ............................................ 443
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 443
111.15 Somatic Embryogenesis in Poplars

(Populus nigra L. x P. maximowiczii Henry)
Y.G. PARK and S.H. SON (With 3 Figures)
1 Introduction ......................................... 446
XVI Contents

References .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 454
111.16 Somatic Embryogenesis in Cacao (Theobroma cacao)

v.c. PENCE (With 2 Figures)
I Introduction ......................................... 455
4 Protocols............................................ 464
References ............ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 466
Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 469

List of Contributors
ALBUERNE, M., Laboratorio de Fisiologia Vegetal, Departamento de

Biologia Organismos y Sistemas, Facultad de Biologia,
Universidad de Oviedo, 33006 Oviedo, Spain
ATANAssov, A.I., Institute of Genetic Engineering,

2232 Kostinbrod-2, Bulgaria
BAJAJ, y'P.S., Former Professor of Tissue Culture,

Punjab Agriculture University, Ludhiana, India
(Present Address: A-137 New Friends Colony, New Delhi 110065, India)
BALDAN, B., Department of Biology, University of Padova,

Via Trieste 75, 35121 Padova, Italy
BENKRIMA, L., Celex Laboratories Inc., 409 Granville Street,

Vancouver, British Columbia, V6C IT2, Canada
BERROS, B., Laboratorio de Fisiologia Vegetal,

Departamento de Biologia Organismos y Sistemas,
Facultad de Biologia, Universidad de Oviedo,
33006 Oviedo, Spain
BUFFARD-MoREL, l., ORSTOM, Laboratoire des Ressources Genetiques

et Amelioration des Plantes Tropicales, 911 avenue d' Agropolis,
BP 5045, 34032 Montpellier Cedex, France
CANTLIFFE, D.l., Department of Horticultural Sciences,

University of Florida, 1251 Fifield Building, Gainesville,
FL 32611-0514, USA
CARICATO, G., Dipartimento di Produzione Vegetale sez.

Ortofloroarboricoltura, Universita della Tuscia, Facolta di Agraria,
via S. Camillo de Lellis, 01100 Viterbo, Italy
CARRON, M.P., CIRAD-Cultures Perennes, Programme Hevea,

Laboratories BIOTROP, BP 5035, 34032 Montpellier Cedex,
France
XVIII List of Contributors
CLAPHAM, D., Swedish University of Agricultural Sciences,

Uppsala Genetic Center, Department of Forest Genetics,
Box 7027, S-752 27 Uppsala, Sweden
COMPTON, M.E., Central Florida Research and Education Center,

University of Florida, 5336 University Avenue, Leesburg,
FL 34748-8203, USA
DENCHEV, P.D., Institute of Genetic Engineering, 2232 Kostinbrod-2,

Bulgaria
(Present Address: The University of Tennessee, Department of Plant
and Soil Science, Knoxville, TN 37901-1071, USA)
DtAZ-SALA, c., Laboratorio de Fisiologia Vegetal, Departamento de

Biologia Organismos y Sistemas, Facultad de Biologia,
DURAND-GASSELIN, T., CIRAD-CP, Station Principale

IDEFORIDPO de La Me, 13 BP 989, Abidjan 13, Ivory Coast
DUVAL, Y., ORSTOM, Laboratoire des Ressources Genetiques et

Amelioration des Plantes Tropicales, 911 avenue d' Agropolis,
BP 5045, 34032 Montpellier Cedex, France
EGERTSDOTIER, U., Swedish University of Agricultural Sciences,

EKBERG, I., Swedish University of Agricultural Sciences,

ELLIS, D., Department of Horticulture, University of Wisconsin,

1575 Linden Dr., Madison, WI 53706, USA
ENGELMANN, F., IPGRI, Via delle Sette Chiese 142,00145 Rome, Italy
ETIENNE, H., CIRAD-Cultures Perennes, Programme Hevea,

Laboratoires BIOTROP, BP 5035, 34032 Montpellier Cedex, France
FITCH, M.M.M., Department of Horticulture, University of Hawaii,

Honolulu 96822, USA
(Present Address: US Department of Agriculture, Agricultural Research
Service, Experiment Station HSPA,
PO Box 1057, Aiea, HI 96701, USA)
GASTALDO, P., Institute of Botany, University of Genoa,

Corso Dogali lIC, 16136 Genoa, Italy
List of Contributors XIX
GIORGETTI, L., Institute of Mutagenesis and Differentiation,

CNR (Italian National Research Council), Via Sevezia 10,
56124 Pisa, Italy
GRAY, D.J., Central Florida Research and Education Center,

University of Florida, 5336 University Avenue, Leesburg,
FL 34748-8203, USA
HAMALAINEN, J., VTT Automation, PO Box 1301,

02044 VTT (Espoo), Finland
HARRELL, R.C., Department of Agricultural Engineering,

University of Florida, 9 Frazier Rogers Hall, Gainesville,
FL 32611-0361, USA
KAUPPINEN, V., VTT Biotechnology and Food Research,

PO Box 1505,02044 VTT (Espoo), Finland
KAWAHARA, R., Gene Bank, Tsukuba Life Science Center,

The Institute of Physical and Chemical Research (RIKEN)
3-1-1 Koyadai, Tsukuba, Ibaraki 305, Japan
KITTO, S.L., Department of Plant and Soil Sciences,

University of Delaware, Newark, DE 19717-1303, USA
KLOCKE, E., Institute of Vegetable, Medicinal and Spice Plant Breeding,

Federal Centre for Breeding Research on Cultivated Plants (BAZ),
Neuer Weg 22/23,06484 Quedlinburg, Germany
KOBAYASHI, T., Department of Biotechnology, Faculty of Engineering,

Nagoya University, Chikusa-ku, Nagoya 464-01, Japan
KOMAMINE, A., Department of Chemical and Biological Sciences, Japan

Women's University, Mejiro, Tokyo 113, Japan
KUNITAKE, H., Laboratory of Plant Cell Technology, Faculty of

Horticulture, Chiba University, 648 Matsudo, Chiba 271, Japan
(Present Address: Laboratory of Plant Biotechnology, Saga
Prefectural Agricultural Research Center, 1088 N anri,
Kawasoe-cho, Saga-gun, Saga 840-23, Japan)
KURTEN, U., VTT Biotechnology and Food Research,

PO Box 1505, 02044 VTT (Espoo), Finland
LAI, F.-M. Department of Crop Science, University of Guelph,

Guelph, Ontario, NIG 2Wl, Canada
xx List of Contributors
LESLIE, C.A., Department ofPomology, University of California,

Davis, CA 95616, USA
Lo SCHIAVO, F., Department of Biology, University of Padova,

Via Trieste 75,35121 Padova, Italy
MANNONEN, L., VTT Biotechnology and Food Research, PO Box 1505,

FIN-02044 VTT (Espoo), Finland
MCGRANAHAN, G.H., Department ofPomology,

University of California, Davis, CA 95616, USA
McKERSIE, B.D., Department of Crop Science, University of Guelph,

Guelph, Ontario, NIG 2Wl, Canada
MERKLE, S.A., Daniel B. Warnell School of Forest Resources,

University of Georgia, Athens, GA 30602-2152, USA
MICHAUX-FERRIERRE, N., CIRAD-Cultures Perennes, Programme Hevea,

Laboratoires BIOTROP, BP 5035, 34032 Montpellier Cedex, France
MIl, M., Laboratory of Plant Cell Technology, Faculty of Horticulture,

Chiba University, 648 Matsudo, Chiba 271, Japan
MINOCHA, R., USDA Forest Service, Northeastern Forestry Experiment

Station, Concord/Mast Road, PO Box 640, Durham, NH 03824, USA
MINOCHA, S.c., Department of Plant Biology, University

of New Hampshire, Durham, NH 03824, USA
Mo, H., Swedish University of Agricultural Sciences,

MOLLOY, D.M., Alfred A.1. Dupont Institute, 100 Rockland Road,

Wilmington, DE 19899, USA
MONTORO, P., CIRAD-Cultures Perennes, Programme Hevea, Laboratoires

BIOTROP, BP 5035, 34032 Montpellier Cedex, France
MUGANU, M., Dipartimento di Produzione Vegetale sez. Ortofloro-

arboricoltura, Universita della Tuscia, Facolta di Agraria,
via S. Camillo De Lellis, 01100 Viterbo, Italy
NUTI RONcHI, V., Institute of Mutagenesis and Differentiation, CNR

(Italian National Research Council), Via SvezialO, 56124 Pisa, Italy
List of Contributors XXI
NUUTILA, A.M., VTT Biotechnology and Food Research,

PARK, YG., Department of Forestry, College of Agriculture,

Kyungpook National University, Daegu 702-701, Republic of Korea
PENCE, v.c., Center for Reproduction of Endangered Wildlife,

Cincinnati Zoo and Botanical Garden, 3400 Vine Steet,
Cincinnati, OH 45220, USA
PEZZA, A., Dipartimento di Produzione Vegetale sez. Ortofloro-

PILL, W.G., Department of Plant and Soil Sciences,

University of Delaware. Newark, DE 19717-1303, USA
PROFUMO, P., Institute of Botany, University of Genoa,

Corso Dogali lIC, 16136 Genoa, Italy
PUUPPONEN-PIMIA, R., VTT Biotechnology and Food Research,

REY, M., Laboratorio de Fisiologia Vegetal, Departamento de Biologia

Organism os y Sistemas, Facultad de Biologia, Universidad de Oviedo,
33006 Oviedo, Spain
RODGRiGUEZ, R., Laboratorio de Fisiologia Vegetal, Departamento de

Biologia Organismo y Sistemas, Facultad de Biologia,
RUGINI, E., Dipartimento di Produzione Vegetale Sez. Ortofloro-

RYSCHKA, u., Institute of Vegetable, Medicinal and Spice Plant Breeding,

Federal Centre for Breeding Research on Cultivated Plants (BAZ),
Neuer Weg 22/23,06484 Quedlinburg, Germany
SCHULZE, J., Institute of Botany, Technical University Braunschweig,

Humboldstr. 1,38106 Braunschweig, Germany
SCHUMANN, G., Institute of Vegetable, Medicinal and Spice Plant

Breeding, Federal Centre for Breeding Research on Cultivated Plants
(BAZ), Neuer Weg 22123,06484 Quedlinburg, Germany
SEFFENS, W.S., AFESClRDVW, Tyndall Air Force Base, FL 32403, USA

XXII List of Contributors
SMITH M.A.L., Department of Horticulture, University of Illinois,

Urbana, IL 61801, USA
SON, S.H., Laboratory of Biotechnology, Forest Genetics Research

Institute, Forestry Administration, PO Box 24, Suwon,
Kyonggido 440-350, Republic of Korea
THOMAS, T.L., Department of Biology, Texas A & M University,

College Station, TX 77843, USA
THOMPSON, R.G., Crop Development Centre, University of

Saskatchewan, Saskatoon, Saskatchewan, S7N OWO, Canada
TREMBLAY, F.M., Centre de Recherche en Biologie-Forestiere,

Pavillon Charles-Eugene-Marchand, Universite Laval,
Sainte-Foy, Quebec, G 1K 7P4, Canada
TREMBLAY, L., Centre de Recherche en Biologie Forestiere,

Universite Laval, Faculte de Foresterie et de Geomatique,
Pavillon Charles-Eugene-Marchand, Sainte-Foy, Quebec,
G1K 7P4, Canada
TULECKE, W., Antioch College, Yellow Springs, OH 45387, USA
UOZUMI, N., Department of Biotechnology, Faculty of Engineering,

Nagoya University, Chikusa-ku, Nagoya 464-01, Japan
VAN ACKER, S., Department of Crop Science, University of Guelph,

Guelph, Ontario, N1G 2Wl, Canada .
VERDEIL, J.-L., CIRAD/CP - ORSTOM, Laboratoire des Ressources

Genetiques et Amelioration des Plantes Tropicales, 911 avenue
d' Agropolis, BP 5045, 34032 Montpellier Cedex, France
VON ADERKAS, P., Centre for Forest Biology, Department of Biology,

University of Victoria, Victoria, British Columbia V8W 2Y2 Canada
VON ARNOLD, S., Swedish University of Agricultural Sciences,

WILDE, H.D., Department of Crop and Soil Sciences,

University of Georgia, Athens, GA 30602, USA
YIBRAH, H., Swedish University of Agricultural Sciences,

ZAKI, M., Department of Plant Sciences, Institute for Efficient

Productivity, Zagazig University, Zagazig, Egypt
Section I
Basic and Fundamental Aspects
1.1 The Cell's Commitment to Somatic Embryogenesis
V. NUTI RONCHI and L. GIORGEITI i
1 Introduction
Embryogenesis may be defined as the developmental program which, starting

from two independent fertilization events, proceeds through coordinated stages
to form a dormant embryo which is preserved and sheltered in the maternal
ovary tissues by means of specific, and often elaborated structures, Le., the seed
and the fruit. In the life cycle of the flowering plant, this process plays a
determinate role, gathering all the expectations for the future harvest and the
performances of the forthcoming sporophytic organs. The functions of the
maternal involucre and of agents or factors which may eventually influence the
embryogenetic program, are largely unknown. Very little experimental work has
been accomplished in this field, due to the small size of the zygote and its location
deep within the maternal tissue. Notwithstanding these difficulties, the plentiful
and accurate reports tracing the various stages of the embryological events,
described with painstaking precision for most flowering plants, have to be
acknowledged (Maheshwari 1950; Johri 1984).
This imposing quantity of work has been little considered in relation to the
more recent experimental studies of somatic embryogenesis, on the prejudicial
idea that the phenomenon has no relationship to embryology in vivo, but was to
be ascribed mainly to that never well-specified attribute of the plant cell,
totipotency.
The difficulties imposed by the female organs to embryo manipulation are
reflected also in the advances that molecular biology has recently achieved in the
identification of genes specific to the floral organ systems; however, very few
concern pistils and related functions.
The purpose of this review, with regard to the commitment to somatic
embryogenesis, will be to recollect how embryogenesis is accomplished in the
maternal tissues, particularly in carrot; and with the published descriptions of
the developmental phases of the reproductive plant organ in mind, to afford a
comparison between these embryological events and what underlies the decision
of a somatic cell to undergo embryogenesis. From this perspective, the com-
parison has to be extended also to other embryogenic systems, not generally
considered to be similar, e.g., androgenesis and parthenocarpy.
I Istituto di Mutagenesi e Differenziamento. CNR. Via Svezia 10, 56124, Pisa, Italy
Biotechnology in Agriculture and Forestry, Vol. 30

Somatic Embryogenesis and Synthetic Seed I (ed. by Y.P.S. Bajaj)
©Springer-Vedag Berlin Heidelberg 1995
4 v. Nuti Ronchi and L. Giorgetti
In this chapter the influence, on cell embryogenesis commitment, of factors

such as nutrients, growth regulators, explant origin, and genotype are not
discussed except when necessary, as several reviews have recently covered this
topic (Williams and Maheswaran 1986; Carman 1990; Dudits et al. 1991;
Komamine et al. 1992). However, attention has been focused on the new
approaches to understanding somatic embryogenesis, starting from cellular,
cytological, histological, and molecular aspects studied in the author's
laboratory or derived from experimental data published by other workers. Some
proposals may, in some cases, still await confirmation, or may be, up to now, just
provocative hypotheses. Unresolved questions in somatic embryogenesis must
be urgently dealt with, thus this chapter has been written with the hope of helping
to attain this aim.
2 Comparison Between Carrot Zygotic

and Somatic Embryogenesis
2.1 Zygotic and Somatic Embryo Properties
Examining one of the earliest reports concerning the ability of carrot cell to
undergo embryogenesis in vitro (Halperin and Wetherell 1964, 1965), it appears
clearly that the competence of carrot for somatic embryogenesis was as mys-
terious, at that time, as it is now. Particularly interesting, in these two quoted
papers, is the fact that two different patterns to form the proembryos were
described, both starting from a small, starch-filled initial embryo cell: one
spherical and the other filamentous (this latter typical of proembryos in the
Daucus family). Regardless of the initial segmentation pattern, a normal pre-
globular proembryo ultimately differentiates. Halperin and Wetherell (1964,
1965) mention a filamentous pattern, similar to the one described by Borthwick
(1931a) that occurs during zygotic embryogenesis, as the first stage of somatic
embryo formation in carrot. This remains the only single report, since all other,
more recent studies refer to a spherical proembryo mass from which somatic
embryos formed (Steward et al. 1969; McWilliam et al. 1974; Nomura and
Komamine 1985; Smith and Krikorian 1989; Komamine et al. 1992). The same
holds true in other embryogenic systems developing from suspension cultures of
dicots (Dudits et al. 1991).
When first discovered, the discussions concerning the phenomenon of som-
atic embryogenesis focused mainly on the similarity to zygotic events. Attention
was given particularly to properties known to belong to the egg cell, e.g., isolation
from neighboring cells and the formation of a special callose wall (Mackenzie et
al. 1967; Williams et al. 1973). All these assumptions and questions have
subsided, particularly since the more sophisticated and refined methods of
obtaining embryos directly from single cells, even protoplasts (Dijak et al. 1986;
Dijak and Simmonds 1988), no longer draw attention to parallel events occurring
in surrounding cells or tissues. Nor do these new exciting possibilities stimulate
investigations to compare similar developmental phases going on in reproductive
organs.
The Cell's Commitment to Somatic Embryogenesis 5
2.2 Embryogenic Cell Size
Several authors maintain that large vacuolate cells of the suspension cultures of
either carrot or alfalfa are not embryogenic (Konar et al. 1972; Nomura and
Komamine 1985; Smith and Krikorian 1990; Dudits et al. 1991). Most authors
agree that, although their origin is unknown, small cells, densely cytoplasmic and
rapidly dividing, are the primary unit of development of a somatic embryo,
which is always preceded, however, by the formation of a small clump of cells
(Backs-Husemann and Reinert 1970).
Komamine's group has successfully established a system for the identi-
fication and isolation of single cells producing embryos at high frequency in
carrot (Nomura and Komamine 1985): single cells that the authors considered
predetermined for embryogenesis. In both Daucus and Medicago it clearly
appears that in order to confer embryogenic competence to such a cell, the first
division has to be asymmetric, producing two cells of different sizes (Dudits et al.
1991).
Unequal asymmetrical divisions are well known, in the plant kingdom, to
occur in relation to specific developmental phases; the first division in pollen
grain, generating the vegetative and regenerative cells, is a well-known case of
asymmetrical division, and is similar to the first division occurring in the Daucus
zygote (Borthwick 1931a). A close relationship with in vivo embryogenesis may
be recognized, but a question immediately arises: is it this type of division that
confers embryogenic competence to this cell, or does the asymmetrical division
occur just because the cell is predetermined? The question is also pertinent
because, in most embryogenic species, the unequal division does not form an
embryo directly, but forms a proembryogenic mass (PEM), whereby only one or
a few cells of this clump can subsequently develop into an embryo.
Turning again to what normally occurs in plants in vivo, and relevant to our
comparison to in vitro processes, it is interesting that, in plants, the reproductive
cells which undergo unequal divisions are in fact predetermined for a final
differential stage, i.e., a functional pollen grain or an embryo. On the other hand,
it is worth noting that, under suitable conditions, a pollen grain at this stage may
also develop into an embryo (by means of an androgenetic process). Moreover,
as reported by Meyer (1966), in some plants, pollen grains can be transformed
into an embryo sac, theoretically able, therefore, to bear embryos; and ovules
may also contain pollen grains (Salter 1863; Goebel 1908; Stow 1930, 1933; de
Mol 1933; Naithani 1937; Geitler 1941).
2.3 Origin of the Small Embryogenic Cells
The origin of the small cells, from which most authors agree that PEMs develop,
is a much discussed subject and many different and conflicting reports have not
reached a consensus (Williams and Maheswaran 1986; Staceyet al. 1990; Dudits
et al. 1991; van Engelen et al. 1991; Komamine et al. 1992). The enigma of the
provenance of small cells comes from the fact that the cells produced from
the primary explants of Daucus and Medicago are large, since the cells of the
heterogeneous cell population of the cultures growing in the presence of 2,4-D

are also predominantly large. In contrast to these observations, as already
mentioned, most authors have demonstrated the involvement of very small cells
(l0-20 Jllll) in the formation of PEMs, whereas the cells constituting the initial
cellular composition of the suspension cultures were considerably larger (50-500
~m). Here, it may be pertinent to recall that Stacey et al. (1990) characterized, in
carrot, by indirect immunofluorescence, the pattern of expression of the cell-
surface arabinogalactan-protein epitope defined by monoclonal antibody JIM 4.
Interesting is the fact that the JIM 4 epitope, which appears to have a
developmental role during embryogenesis in culture, upon removal of 2, 4-D
from the medium, is mostly expressed, in the cultured hypocotyls, along the
vascular strands and, in the derived suspension cultures, in the large vacuolated
cells, considered to be nonembryogenic. In contrast, only infrequent, isolated,
small cells at the surface ofPEMs are found to express JIM 4. Whether these few
cells are the only ones in the entire PEM committed to develop into embryos is
not known.
Supposed nonembryogenic, large vacuolate cells have again the role of
protagonist insofar as the localization of an extracellular cell wall glycoprotein
(related also to embryogenic development) from carrot embryogenic suspension
culture is concerned. This epiprotein, EP 1, was only found to be associated with
pectin-containing cell wall material of the large vacuolate cells of carrot cultures
(van Engelen et al. 1991).
3 Somatic Chromosome Segregational Events
3.1 In Vitro Differentiation of Gamete-Like Cells
We have attempted to solve the mystery of the origin of small cells with an
experiment aiming at careful and continuous observations ofthe living, cultured,
carrot cell population. Observation started from the first day that cells were
released from the hypocotyls into the medium up to the time (15th day) that the
cortical layers easily dissociated from the internal structures developed around
the vascular strands, discharging large or very large vacuolated cells into the
medium. According to our hypothesis, these particular vacuolated cells are the
predecessor (and generative) cells, from which the predetermined small
embryogenic cells would originate.
This last experiment enabled us to determine that the cells, released into the
medium from the hypocotyls, go through a distension phase, most cells appear-
ing as female or male gametophyte cells, i.e., embryo sac and immature pollen
grains. Both are often indistinguishable from in vivo gametic cells, the embryo
sac-like cell reaching the 8-nucleate (or 7-nucleate due to fusion of two nuclei)
stage and the pollen-like cell undergoing the unequal division. From the vacu-
olate cell, similar to the female gametophyte, small single uninucleate cells are
discharged through a process similar to budding, or by the separation of the
nuclei by means of a cytokinesis process followed by the rounding off and release
Fig. 1. Different aspects in the formation of proembryogenic masses (PEMs) from gametic-like cells in
carrot cultures. A Asymmetrical division in a pollen-like cell. Arrows point to the double cell wall.
D Four-celled stage similar to the zygotic proembryo; the rounding off of the walls will allow the release
ofthe small cells into the medium. C Large embryo sac-like cell which forms at both distal ends, small
cells released into the medium after rounding off of the cell wall (arrows) ; CC central cell. D Initial
early stages of PEMs, reminiscent of carrot zygotic stages; arrows further growth with different
division planes. E PEM formed from initial early stages similar to those observed in zygotic carrot
embryos and still recognizable in the form
8 V. Nuti Ronchi and L. Giorgetti
of the resulting small cells (Fig. 1C), or from the breakdown of the parental cell
wall.
PEMs are formed after the first unequal division, from these small cells in
two ways: (1) following, in part, the early divisions described by Borthwick
(193la) for Daucus zygote giving first a filamentous 8- or l6-celled proembryo
(Fig. lD), followed by less regular divisions, so forming a PEM (Fig. IE); (2) or
as described by Street and Withers (1974), the single elongate cell divides with
two planes of division, forming a clump of few cells. Other modes of formation
are possible, depending on the cultivar, temperature shift and differences in cell
lines. Pollen-like cells may produce a PEM directly after the first unequal division
(Fig. lA), but sometimes the smallest cell rounds off separating from the larger
(Fig. IB). The different and often complicated ways in which large cells adapt to
discharge small cells are described in detail elsewhere (Giorgetti and Nuti
Ronchi, submitted).
At the moment, it is not possible to determine whether there are differences
in the ability to form PEMs and/or embryos among, e.g., the egg-like cells or any
of the other embryo sac-derived cells (synergid- or antipodal-like). Since these
cells are known in vivo to be involved sometimes in the formation of apomictic
embryos, it is presumed that they can regenerate embryos in culture.
The cells of the PEM clusters may also round off, allowing their subsequent
release into the medium, when the cultures are maintained in the presence of2,4-
D. Since the cultures are composed of a very heterogeneous cell population, some
of the cells released in this way from PEMs probably undergo the same pheno-
mena already described for the initial stages of the culture, i.e., cell wall distension
and vacuolation. Since the supposed nonembryogenic, large vacuolate cells are
easily isolated from the culture, it has been possible, once separated from the
others and cultured for few days under the usual conditions with 2,4-D, to
demonstrate that embryogenic capacity is recovered and that the heterogeneous
cell population of the culture is restored (L. Pitto, unpubl.).
3.2 The Somatic Meiosis Process
The surprising plasticity of plant cells, and their ability to transform and to
interchange functions between organs and tissues, have suggested the possibility
that all the generative functions, meiosis included, could also be expressed, when
triggered by stress, in cells which are not normally capable of such functions. Our
investigations on the hidden capacity of plant cells to express all the functions of
the reproductive cycle, once triggered by specific agents or factors, were based on
the above-mentioned finding that embryogenic carrot cell suspension cultures
could behave as reproductive cells (pollen and embryo sac-like). Drawing a
parallel between in vitro and in vivo behavior, cultured cells should also be able
to divide in a manner closely resembling meiosis (Nuti Ronchi 1990; Nuti Ronchi
et al. 1990, 1992a,b; Giorgetti et al. 1991).
A cytological and histological analysis of the first 20 days of culture of carrot
hypocotyl explants, cultured in B5 medium in the presence of2,4-D (2,2 IlM), has
revealed that the chromosome-segregating events (somatic meiosis and prophase
reduction) occurred with a modality described in detail elsewhere (Nuti Ronchi

1990, 1991; Nuti Ronchi et a1. 1992a,b,c,d; Giorgetti et a1., in press).
Both processes allowed homologous chromosome segregation, but after this
segregating phase was accomplished, a precocious centromere division of the
sister chromatids could restore the diploid condition, of course at 2C DNA
content.
The reducing events involve about 3% of the division, although this estimate
may be low, since prophase reduction is not always easy to recognize in the
squashes.
3.3 Somatic Meiosis As a Reprogramming Event
It is important to note that the passage from mitosis to meiosis (the two processes
differing only in a few basic steps) occurs possibly under stress, and is more wide-
spread than normally presumed (Huskins 1948). Therefore, it may be associated
with the primary expression of totipotency. However, the constant presence of
these events in the cultures and in the early phases of embryo development could
also respond to a general inductive hormonal condition of the culture. Thus, a
constant induction of embryogenically determined cells may be inferred.
The old carrot nonembryogenic lines WI and W2, permanently expressing
an aberrant meiotic phenotype (Nuti Ronchi et a1. 1992b), and other lines with a
similar phenotype, easily obtained after temperature shocks (M.G. Tonelli
unpub1.), show that this feature may be completely independent of embryogenic
capacity. Out data suggest that only when the segregating event is followed by the
development of a functional gametophytic-like cell, can embryogenesis be
accomplished. Thus, according to this reasoning, cell commitment to embryo-
genic capacity appears related, in carrot, to a reprogramming phenomenon
which involves not only a meiotic-like process, but also a further development of
the affected cell, i.e., mimicking a micro- or macrogametophyte. Moreover, a
dramatic decrease in DNA sequences occurs along with these phenomena (Geri
et a1. 1992), suggesting the need of the cells, before starting a new developmental
process, to erase all previous sporophytic information.
4 Evidence of the Occurrence of Segregational Events
4.1 Microdensitometric and Molecular Demonstration of DNA Diminution
We first ascertained, by means of chromosome counts and microdensitometric

DNA content measurements, that a haploid condition and a total DNA
diminution was progressively acquired by the hypocotyls and suspension culture
cells, and maintained by the regenerated embryos up to the plantlet stage.
Interesting enough was the demonstration, by means of micro densitometric
DNA measurements (Nuti Ronchi et a1. 1992a,b) and molecular slot-blot
hybridization with specific problems (Geri et al. 1992), that, besides the shift to
the haploid DNA content, during the same culture times, a considerable loss of
DNA sequences accompanied the acquisition of embryogenic capacity. The

DNA loss was evident also in the regenerated embryos up to the plantlet stage,
when the normal DNA content, comparable to the root tip condition, was
recovered. Similar DNA variations during plant development have been found
by several authors (see Bassi 1990).
The phenomenon involves medium repetitive and unique chalcone
synthetase DNA sequences. The DNA loss was maximal at the cell culture phase
and early stages of embryo development. A return to normal DNA values of
leaves was evident only at the embryo plantlet stage (Geri et al. 1992).
It may be recalled that similar phenomena have been shown to occur in Scilla
siberica cultures (Deumling and Clermont 1989) in response to environmental
conditions and have been suggested to be a prerequisite to render the cells
"omnipotent".
4.2 RFLPs as Markers of Segregation
To provide the genetic proof that a true chromosomal segregation occurs in

carrot, experiments have been performed using restriction fragment length
polymorphisms (RFLPs) as markers of segregation, comparing the patterns
shown by carrot pure lines to those of their sexual hybrid and regenerants
obtained from the sexual hybrid via somatic embryogenesis. Segregation was
demonstrated in 8 out of 10 regenerants. These data support, therefore, the
hypothesis that, in carrot cell commitment to embryogenesis requires a process of
genetic reprogramming similar to the one producing the gametes.
4.3 Somatic Meiosis and Epigenetic Homeotic Events
The differentiation of reproductive structures (Fig. 2) (sometimes resembling

primitive ancestral phenotypes), situated along the vascular bundles, occurred
in the hypocotyls concomitantly with chromosome-reducing events. These
epigenetic rudimentary organs may be defined as homeotic, accepting the
definition of home osis as the "assumption by one part of an organism oflikeness
to another part" (Sattler 1988).
Anther- and pistil-like (Fig. 2A) organs were formed, often separately, in the
smallest cuttings (2- 4 mm) or symmetrically on both sides of the longer explants
(Giorgetti et al. 1991; Nuti Ronchi 1991). Often a floral primordium (Fig. 2D), at
one or at both ends of the explant, was also evident using a suitable stain.
The two ploidy reducing mechanisms, somatic meiosis and prophase
reduction, were detected along the vascular strands, often very close to the xylem.
The most frequent, newly formed structures were the organs comparable to the
carrot pistils. The carpels were formed by the transformation of the hypocotyl
cortical cylinder (Fig. 2C), whose cells undergo a distension process except for
two symmetrical longitudinal zones. These zones appear to turn inwards,
simulating carrot carpel margins, dividing the structure into two loculi where the
ovules and styles developed at the two sides of the hypocotyl vascular strands
(Fig. 2A). Similarly, the vascular bundles assumed the function ofthe filament
The Cell's Commitment to Somatic Embryogenesis II
~
o
Fig. 2. Reproductive structures developed on carrot and tomato hypocotyls cultured in vitro for 20
days. A Symmetrical, specular, pistil-like organs developed on carrot hypocotyls. 0 Ovaries; S styles;
C carpels. B Ovary-like structure formed on tomato hypocotyl. 0 V Ovules. C Carpel-like structure
with carpel margins and few ovules (or) of different sizes. D Floral primordia formed at one end of the
carrot hypocotyl. ST stamen; VS vascular strands
when an anther-like structure was formed. After 20 days of culture the

resemblance of the newly developed structures to similar organs in vivo was
striking, even if several variants were possible, e.g., inflorescence primordia
developing into carpel-like structures or ovules being freely attached to the
hypocotyl ends. The epidermal layers of these structures seemed, in most cases,
poorly defined, cells being, therefore, easily released into the medium.
When hypocotyls were maintained for more than 20 days in the same
medium, extensive callus formation covered the structures. No further
development occurred when the growth factor was omitted from the mf"dium,
but at the proper cell density PEMs and, subsequently, embryos were regenerated
from the derived cell culture.
4.4 Epigenetic Events and Cell Commitment
The most surprising conclusion coming from these data is that only a few days of
culture in the presence of a growth factor are sufficient to form, on the cut ends
of the hypocotyl explants, a meristem, previously not present. It rapidly
undergoes a transition from vegetative to reproductive development or is directly
reproductively competent. Moreover, the competence for reproductive organ
formation is also developed by a ring (or cylinder) or fascicular cambium which,
instead of differentiating into a pro vascular tissue, forms structures, often
resembling a style, continuously discharging cells into the medium. From those
cells which originate from a segregating event (as demonstrated by the molecular
data) PEMs are formed, sometimes directly when still in the explant (particularly
around and in the ovary-like structures) or after few passages in fresh medium,
depending on the cultivar.
4.5 Carrot Somatic Embryos Are of Secondary Origin
In the carrot system thus described, somatic embryos may be defined as being of
indirect and secondary origin, since they develop from an unsuccessful attempt to
form an embryo. Recently, some authors have described refined systems in carrot
and in other species, obtaining embryos directly from zygotic embryos or
mesophyll protoplast. (Magnusson and Bornman 1985; Dijak et a1. 1986; Smith
and Krikorian 1988, 1989, 1990; Jones and Rost 1989; Tetu et a1. 1990; Stolarz
et a1. 1991). Only few of these papers give histological details of embryo
formation, and no cytological data are presented. There is therefore, no means of
ascertaining whether in these cases also the totipotency acquired by cells and
leading to somatic embryogenesis can be ascribed to cytological events similar to
those occurring during micro- or macro gametogenesis in the plant. Moreover, it
is worth noting that the genetic proof of a segregation process is only available
when somatic embryos are actually formed from a single somatic cell, a case that
does not appear to be frequent.
A recent paper (Smith and Krikorian 1992) reporting the finding that low
external pH prevents cell elongation but not multiplication of embryogenic
carrot cells does not appear to be in contrast to our data. In our experiments large
vacuolate cells are potentially embryogenic, continuously discharging new small
cells, when submitted to suitable conditions. Growth at low pH probably allows
the continuous multiplication, on hormone-free medium, of already committed
embryogenic cells only, in a sort of cleavage polyembryony, where all cells are
predetermined for embryonal development (Williams and Maheswaran 1986).
As it is clearly shown by Smith and Krikorian (1990, 1992), pH changes in
growth media can affect morphological patterns, may alter the internal pH, and
may be implicated as a second messenger in intracellular signaling. The
possibilities to change cellular morphology and fate only by means of a pH
change may be determinate to confirm, with appropriate experiments, the role of
large cells in carrot suspension cultures.
4.6 Segregational Events Occur Also in Nonembryogenic Plants
All plant species tested have been shown to carry over chromosome-reducing
divisions (somatic meiosis and/or prophase reduction) when triggered by
wounding (in vitro culture or other stress conditions). In tomato, besides the
reducing events occurring in the cultured hypocotyls, structures simulating
reproductive organs are formed along the vascular cylinder, and inflorescence
primordia are visible at times comparable to carrot. Particularly clear, immature
mononuekate pollen and embryo sac-like cells are extruded into the medium, but
abnormal cdl wall differentiation and the inability to go beyond the eight-celled
stage, respectively, impeded further growth of these cells (Pitto et al. 1992,1993).
It is worth noting that, in tomato, we have obtained preliminary con-
firmation of the true reproductive nature of these structures by means of in situ
hybridization with tomato pistil-specific molecular probes (Nuti Ronchi et al.
1992d; Pitto et al. 1992,1993).
With slight differences, other species such as Helianthus (P. Belloni unpubl.)
and Malus (Blando et al. 1990) have shown a similar behavior. Although a
complete study has not been carried out in monocotyledonous plants,
preliminary observations have confirmed the presence, at least, of prophase
reduction divisions.
Our hypothesis is that the response of plants to culture (or to other stress
conditions) is always an attempt to develop reproductive structures which,
in different ways, could propagate the species, overcoming the unsuitable
environment by means of the extraordinary method of increasing genetic
variability such as the one offered by meiotic segregation.
5 Cellular Competence for DNA Reprogramming
5.1 Specific Cells May Be More Responsive to Stress
The reprogramming process does not seem to occur in all cells of the cultured
hypocotyls, but instead in specific cells properly located in tissues deeply affected
by wounding and culture medium, and triggered to a fast transformation from
vegetative to reproductive development.
The term "specific" to define the cells undergoing the reducing events is not
casual, but may imply a peculiar chromatin asset or a special location (both states
possibly due to a still embryonic condition), which provides these particular cells
with a remarkable plasticity. Data are scarce on this point, but other authors
have noted that, in some plant species, the most responsive cells to cultural
triggers were those situated close to the vascular bundles, especially xylem
elements (Lu and Vasil 1981; Lipucci di Paola et al. 1987; Barcelo et al. 1991,
1992). These cells have the ability to vary rapidly, within such a short period as
few hours, start division when the tissue is excised and cultured in vitro.
5.2 Modulation of Nuclear DNA Content and Development
Again, data on the DNA content and on the cytological behavior of newly
explanted plant tissues are scanty, so there is little possibility to compare our
findings with other published results. The papers concerning the regeneration
property of young leaves or immature inflorescences of Gramineae plants seem,
in our opinion, to confirm our hypothesis of the existence, in embryonic organs,
of a particular cell population prone to react to stress as described for carrot. Due
to the well-known difficulty to induce differentiation and embryogenesis in
monocotyledonous plants, immature tissues or organs, close to an embryonic
state, have been successfully used with this aim in mind. The picture coming from
an analysis of the published data reveals a surprising fact regarding the sectors of
the youngest leaves more prone to regenerate, either via organogenesis or
embryogenesis. During the first days of culture, when cell division was more
active, as often shown also by the mitotic index, the cells of these leaves prone to
regenerate were most often in the 2C phase. Only a small proportion of the cells
was in the 4C phase, as would be the case in a fast dividing cell popUlation
(Wernicke and Milkovits 1984; Joarder et al. 1986; Karlsson and Vasil 1986;
Taylor and Vasil 1987). Hesemann and Schroeder (1982) have shown a remark-
able DNA diminution, even under the normal diploid level, in leaves of rye.
The disproportionate prophase index reported for Lotium multiflorum
(Joarder et al. 1986) is probably due to the prophase reduction phenomenon,
which is most frequent in cell culture, and may be confused with normal
prophases.
The prophase reduction phenomenon (Nuti Ronchi et al. 1992a,b) is
signaled by a disproportionate number of prophase stages which always occur in
pairs in the same cell, appearing often still partly intermingled and partly divided,
the chromosomes being directly unthreaded into two reduced prophase
configurations. According to a careful cytological observation, the phenomenon
leads to the formation of two cells by subsequent cytokinesis. As in carrot, a cell
population with a proportion of cells dividing accordingly, will cycle mainly
between C and 2C DNA content, only a small proportion being in 4C.
Our interpretation of these data is that in the Gramineae also, as in carrot,
the tissues which have retained also embryonic capacity are in a particular con-
dition (physical?, physiological?, genetic?) which, under stress (e.g., wounding
and culture medium), may enter directly a reprogramming phase which includes
chromosome segregation and DNA reorganization. Some authors have sug-
gested the presence, during the early stage of differentiation of the leaves, of a
nuclear condition (most likely G 1) as a phase, of a cell popUlation, competent for
further, even unscheduled, reprogramming (Gould 1983; Taylor and Vasil 1987).
Large nuclear changes, possibly related to the morphogenetic response

in vitro, have also been reported in Nicotiana tabacum tissues (Altamura et al.
1987). The DNA loss concerned particularly highly repeated sequences and
seemed to progressively impair morphogenetic capability. These results,
apparently in contrast to our data, should however, be considered in relation to
the system, consisting in regeneration of roots, buds, or flowers directly from
epidermal and subepidermal cell layers excised from different portions of
flowering tobacco stem, completely devoid of vascular tissue, which in our
hypothesis has a determinate role in the future destiny ofhypocotyl explants. It
is worth noting that also in Nicotiana glauca pith tissue (devoid of vascular tissue)
the dedifferentiation phase was characterized by only DNA amplification
(Parenti et al. 1973; Martini and Nuti Ronchi 1974).
5.3 Segregational Events and Plant Transformation
A special aspect, which is worth considering in relation to these phenomena

occurring in embryonic tissues when triggered by stress, is the possibility offered
by these peculiar genetic and cytological states to accomplish a process of
transformation. It is well known that the production of transformed plants
depends on the ability of the cells to enter a transformation-competent state.
In a review on gene transfer in plants (Potrykus 1991), a series of important
questions to be studied in order to achieve further progress in this field is
proposed. Among others, the need to understand what makes a cell competent
for dedifferentiation, and what makes the cells in this phase more competent for
transformation. We believe that our data may, in part, help to answer these
questions. The response to wounding and to culture of hypocotyl cells starting
processes such as cell wall autolysis (which detaches the cortical layers from the
vascular strands), along with chromosome segregation, and with a stage of
haploidy and DNA turnover, make this system extremely promising for
transformation experiments. The system we propose is based on the formation of
gamete-like cells which, after suitable manipulation, could directly produce
bipolar embryos or shoots, offering "a method which will allow routine and
efficient gene transfer into all desired genotypes of any plant species" (Potrykus
1991).
Wounding and dedifferentiation have proved to be an important step in gene
transfer protocols, and already it has been suggested that embryonic organs such
as hypocotyls might give the best response in transformation (Dekeyser et al.
1990; Dong and McHughen 1991; Schlappi and Hohn 1992).
6 Concluding Remarks
Here, it is proposed that the phenomena we have discovered have a more general
meaning than the one related to somatic embryogenesis, being in fact that
expression of specific differentiated states as an answer to challenging conditions.
Preliminary observations support the hypothesis that the findings in carrot,

tomato, and other plants, are not restricted to hypocotyls, but occur when tissues
are excised from the plant and cultured in vitro. Similar reproductive structures
may be recognized, although with more difficulty, because they are extremely
abnormal, in cultured leaves and stems of tomato and carrot explants, at least as
an early response to culture conditions.
Clearly, more experimental work is needed to confirm our data and our
interpretation and to acquire knowledge, at the molecular and physiological
level, of the phenomenon, ascertaining also whether it occurs in other systems.
Our hypothesis is that these epigenetic homeotic structures, which we have
found to occur in all plant species studied up to now (tomato, Helianthus, Malus),
are always developing on plant tissue explants when exposed to culture media
and growth factors. They are the ultimate products of a series of events that
switch on a determinate program leading to floral induction, including a
gametophytic phase. The complete course of events takes only a few days, the
developmental reproductive program often showing ancestral primordial
features and a high level of deregulation, with phenotypes already described in
abnormal flowers (Meyer 1966; Sattler 1988). Totipotency, in this context, might
only be an effort to survive stress by means of an unscheduled reproductive
program.
Acknowledgments. The author's research group work quoted in this chapter was supported by the
National Research Council ofItaly, Special Project RAISA, Subproject No.2, Paper No.1 125 and by
the Minister of Agriculture and Forestry "Sviluppo Tecnologie Alternative" program.
References
Altamura MM, Bassi P, Cavallini A, Cionini G, Cremonini R, Monacelli B, Pasqua G, Sassoli 0, Tran
Thanh Van K, Cionini PG (1987) Nuclear DNA changes during plant development and the
morphogenetic response in vitro of Nicotiana tabacum tissues. Plant Sci 53: 73-79
Backs-Husemann D, Reinert J (1970) Embryo formation by isolated single cells from tissue cultures of
Daucus carota. Protoplasma 70: 49-60
Barcelo P, Lazzeri PA, Martin A, Lorz H, (1991) Competence of cereal leaf cells. I. Patterns of
proliferation and regeneration capability in vitro of the inflorescence sheath leaves of barley, wheat
and fertile amphiploids Hordeum chilense X Triticum tritordeum. Plant Sci 77: 243-251
Barcelo P, Lazzeri PA, Martin A, Lorz H (1992) Competence of cereal leaf cells. II. Influence of auxin,
ammonium and explant age on regeneration. II. Plant Physiol 139: 448-454
Bassi P (1990) Quantitative variations of nuclear DNA during plant development: a critical analysis.
Bioi Rev 65: 185-225
Blando F, Giorgetti L, Tonelli MG, Nuti Ronchi V (1990) Cytological characterization of cell
suspension cultures offruit trees. Acta Hortic 300: 377-380
Borthwick HA (193Ia) Development of the macrogametophyte and embryo of Daucus carota. Bot Gaz
92: 23-44
Borthwick HE (193Ib) Flower development in Daucus carota. Am J Bot 18: 784-796
Carman JG (1990) Embryogenic cells in plant tissue cultures: occurrence and behaviour. In Vitro Cell
Dev Bioi 26: 746-753
Dekeyser RA, Claes B, De Rycke RMU, Habets ME, Van Montagu M, Caplan AB (1990) Transient
gene expression in intact and organized rice tissues. Plant Cell 2: 591-602
de Mol WE (1933) Die Entstehungsweise anormaler Pollenkorner bei Hyazinthen, Tulpen und
Narzissen. Cytologia 5: 31-63
Deumling B, Clermont L (1989) Changes in DNA content and chromosomal size during cell culture
and plant regeneration of Scilla siberica. Chromo soma (Berl) 97: 439-448
Dijak M, Simmonds DH (1988) Microtubule organization during early direct embryogenesis from
mesophyll protoplasts of Medicago sativa L. Plant Sci 58: 183-191
Dijak M, Smith DL, Wilson, n, Brown DCW (1986) Stimulation of direct embryogenesis from
mesophyll protoplast of Medicago sativa. Plant Cell Rep 15: 468-480
Dong JZ, McHughen A (1991) Patterns of transformation intensity on flax hypocotyls inoculated with
Agrobacterium tumefaciens. Plant Cell Rep 10: 555-560
Dudits D, Bogre L, Gyorgyey J (1991) Molecular and cellular approaches to the analysis of plant
embryo development from somatic cells in vitro. J Cell Sci 99: 475-484
Geitler, L. (1941) Embryosacke aus Pollenkornren bei Orinthogalum. Ber Dtsch Bot Ges. 59:419-423
Geri C, Giorgetti L, Turrini A, Nuti Ronchi V (1992) Modulation of DNA sequences in relation to the
commitment of carrot somatic embryogenesis. Atti AGI 38: 139-140
Giorgetti L, Nuti Ronchi V. Carrot somatic embryogenesis is related to an epigenetic phenomenon of
homeosis (submitted)
Giorgetti L, Pitto L, Tonelli MG, Vergara MR, Martini G, Collina Grenci F, Luccarini G, Evangelista
M, Nuti Ronchi V (1991) Auxin-induced formation of home otic structures (anther- and pistil-like)
in cultured hypocotyl explants of different plant species. Atti AGI 37: 201-202
Giorgetti L, Vergara MR, Evangelista M, Lo Schiavo F, Terzi M, Nuti Ronchi V. On the occurrence
of somatic metosis in embryogenic carrot cell cultures. Mol Gen Genet (in Press)
Goebel K (1908) Einleitung in die Experimentelle Morphologie der Pflanzen. Druck und Verlag Bg
Teubner, Leipzig
Gould AR (1983) Control of the cell cycle in cultured plant cells. CRC Crit Rev Plant Sci I: 315-344
HalperinW, Wetherell, DF. (1964) Adventive embryony in tissue cultures of the wild carrot, Daucus
carota. Am J Bot. 51: 274-283
Halperin W, Wetherell DF (1965) Ontogeny ofadventive embryos in wild carrot. Science 147: 756--758
Hesemann CU, Schroeder G (1982) Loss of nuclear DNA in leaves of rye. Theor Appl Genet 62:
325-328
Huskins CL (1948) Segregation and reduction in somatic tissue. J Hered 39: 311-318
Joarder 01, Joarder NH, Dale PJ (1986) In vitro response ofleaftissues from Lolium multiflorum - a
comparison with leaf segment position, leaf age and in vivo mitotic activity. Theor Appl Genet 73:
286-291
Johri BM (ed) (1984) Embryology of angiosperms. Springer, Berlin Heidelberg New York
Jones TJ, Rost TL (1989) The developmental anatomy and ultrastructure of somatic embryos from rice
(Oryza sativa L.) scutellum epithelial cells. Bot Gaz 151: 41-49
Karlsson SB, Vasil IK (1986) Growth, cytology and flow cytometry of embryogenic cell suspension
cultures of Panicum maximum Jacq. and Pennisetum purpureum Schum. J Plant Physiol 123:
211-227
Komamine A, Kawahara R, Matsumoto M, Sunabori S, Toya T, Fujiwara A, Tsukahara M, Smith J,
Ito M, Fukuda H, Nomura K, Fujimura T (1992) Mechanisms of somatic embryogenesis in cell
cultures: physiology, biochemistry and molecular biology. In vitro Cell Dev BioI 28P: 11-14
Konar RN, Thomas E, Street HE (1972) Origin and structure of embryoids arising from epidermal
cells of the stem of Ranunculus sceleratus L. J Cell Sci II: 77-93
Lipucci di Paola M, Fossi D, Tognoni F, Nuti Ronchi V (1987) Adventitious bud induction in vitro
from juvenile leaves of Cupressus arizonica Green. Plant Cell Tissue Organ Cult 10: 3-10
Lu C, Vasil IK (1981) Somatic embryogenesis and plant regeneration from leaf tissue of Panicum
maximum Jacq. Theor Appl Genet 59: 275-280
Lu C, Vasil V, Vasil IK (1983) Improved efficiency of somatic embryogenesis and plant regeneration
from tissue cultures of maize (Zea mays L.) Theor Appl Genet 66: 285-289
Mackenzie A, Heslop-Harrison J, Dickinson HG (1967) Elimination of ribosomes during meiotic
prophase. Nature 215: 997-999
Maheshwari P (1950) An introduction to the embryology of angiosperms. McGraw-Hill New York
Martini G, Nuti Ronchi V (1974) Microdensitometric and autoradiographic analysis of cell
proliferation in primary culture of Nicotiana glauca pith tissue. Cell Differ 3: 239-247
Mascarenhas JP (1989) The male gametophyte of flowering plants. Plant Cell I: 657-664
McWilliam AA, Smith SM, Street HE (1974) The origin and development of embryoids in suspension
cultures of carrot (Daucus carota). Ann Bot 38: 245-250
Meyer VG (1966) Flower abnormalities. Bot Rev 32: 165-218
Naithani SP (1937) Chromosome studies in Hyacinthus orientalis L. Reversal of sexual state in the
anthers of Hyacinthus orientalis L., var. Yellow Hammer. Ann Bot (Lond) NS 1: 369-377
Nomura K, Komamine A (1985) Identification and isolation of single cells that produce somatic
embryos at a high frequency in a carrot suspension culture. Plant Physiol 79: 988-991
Nuti Ronchi V (1990) Cytogenetics of plant cell cultures. In: Bhojwani SS (ed) Plant tissue culture:
application and limitation. Elsevier, Amsterdam, pp 276-300
Nuti Ronchi V (1991) Biological and genetic features of cell cultures in relation to plant morpho-
genesis. IAPTC News Lett 65: 2-12
Nuti Ronchi V, Giorgetti L, Tonelli MG (1990) The commitment to embryogenesis, a cytological
approach. In: Nijcamp HJJ, Van Der Plas LHW, Van Aartijk J (eds) Progress in plant cellular and
molecular biology, Kluwer, Dordrecht, pp 437-442
Nuti Ronchi V, Giorgetti L, Tonelli MG, Martini G (I 992a) Ploidy reduction and genome segregation
in cultured carrot cell lines. I. Prophase chromosome reduction. Plant Cell Tissue Organ Cult 30:
107-114
N uti Ronchi V, Giorgetti L, Tonelli M G, Martini G (1992b) Ploidy reduction and genome segregation
in cultured carrot cell lines. II: Somatic meiosis. Plant Cell Tissue Organ Cult 30: 115-119
Nuti Ronchi V, Giorgetti L, Tonelli MG, Belloni P, Martini G (1992c) Analogies of chromosome
reducing events in somatic cultured cells and in microsporogenesis. In: Mulcahy DL, Bergamini
Mulcahy G, Ottaviano E (eds) Angiosperm pollen and ovules. Springer, New York, pp 245-247
Nuti Ronchi V, Giorgetti L, Geri C, Pitto L, Vergara R, Martini G (1992d) Cytological, anatomical
and morphological aspects of somatic embryogenesis. In: Griga M, Tejklova E (eds) Regulation of
plant somatic embryogenesis. Research Institute Technical Crops and Legumes Sumperk,
pp 67-71
Parenti R, Guille' E, Grisward J, Durante M, Giorgi L, Buiatti M (1973) Transient DNA satellite in
dedifferentiating pith tissue. Nature New BioI 246: 237-239
Pitto L, Miarelli C, Giorgetti L, Colella C, Luccarini G, Nuti Ronchi V (1992) Occurrence of floral
homeotic structures in carrot, tomato and Helianthus hypocotyls in liquid cultures. Atti AGI 38:
107-108
Pitto L, Giorgetti L, Miarelli C, Luccarini G, Evangelista M, Colella C, Nuti Ronchi V (1993)
Hormone induced floral reprogramming in cultured hypocotyls: induction of gametic-like cells and
expression of floral homeotic MAD genes. Atti AGI 39: 111-112
Potrykus I (1991) Gene transfer to plants: assessment of published approaches and results. Annu Rev
Plant Physiol Plant Mol BioI 42: 205-225
Salter SJA (1863) On a sexual monstrosity, consisting in the development of polleniferous ovules, in
two species of Passiflora. Trans Linn Soc 24: 143-150
Sattler R (1988) Homeosis in plant. Am J Bot 75: 1606-1617
Sch1appi M, Hohn B (1992) Competence of immature maize embryos for Agrobacterium-mediated
gene transfer. Plant Cell 4: 7-16
Smith DL, Krikorian AD (1988) Production of somatic embryos from carrot tissues in hormone-free
medium. Plant Sci 58: 103-110
Smith DL, Krikorian AD (1989) Release of somatic embryogenic potential from excised zygotic
embryos of carrot and maintenance of proembryonic cultures in hormone-free medium. Am J Bot
76: 1832-1843
Smith DL, Krikorian AD (1990) Somatic embryogenesis of carrot in hormone-free medium: external
pH control over morphogenesis. Am J Bot 77: 1634-1647
Smith DL, Krikorian AD (1992) Low external pH prevents cell elongation but not multiplication of
embryogenic carrot cells. Physiol Plant 84: 495-501
Stacey NJ, Roberts K, Knox P (1990) Patterns of expression of the arabinogalactan-protein epitope in
cell cultures and during somatic embryogenesis in Daucus carota L. Plant 180: 285-292
Steward FC, Ammirato PV, Mapes MO (1969) Growth and development of totipotent cells. Some
problems, procedures and perspectives Ann Bot 34: 761-787
Stolarz A, Macewicz J, Li:irz H (1991) Direct somatic embryogenesis and plant regeneration from leaf
explants of Nicotiana tabacum L. J Plant Physiol137: 347-357
Stow I (1930) Experimental studies on the formation of the embryosac-like giant pollen grains in the
anthers of Hyacinthus orientalis. Cytologia I: 417-439
Stow I (1993) On the female tendencies of the embryo sac-like giant pollen grain of Hyacinthus
orientalis. Cytologia 5: 88-108
Street HE, Withers LA (1974) The anatomy of embryogenesis in culture. In: Street HE (ed) Tissue
culture and plant science. Proc 3rd Int Congr Plant tissue and cell culture. University of Leicester.
Academic Press London, pp 71-100
Taylor MG, Vasil IK (1987) Analysis of DNA size, content and cell cycle in leaves of Napier grass
(Pennisetum purpureum Schum.). Theor Appl Genet 74: 681-686
Tetu T, Sangwan RS, Sangwan-N orree1 BS (1990) Direct somatic embryogenesis and organogenesis in
cultured immature zygotic embryos of Pisum sativum. J Plant Physiol137: 102-109
Van Engelen FA, Sterk P, Booij H, Cordewener JHG, Rook W, Van Kammen A, De Vries S (1991)
Heterogeneity and cell type-specific localization of a cell wall glycoprotein from carrot suspension
cells. Plant Physiol 96: 705-712
Wernicke W, Milkovits L (1984) Developmental gradients in wheat leaves - response of leaf segments
in different genotypes cultured in vitro. J Plant Physioll15: 49-58
Williams EG, Maheswaran G (1986) Somatic embryogenesis: factors influencing coordinated
behaviour as an embryogenic group. Ann Bot 57: 443-462
Williams EG, Heslop-Harrison J, Dickinson HG (1973) The activity of the nucleolus organising region
and the origin of cytoplasmic nucleoids in meiocytes of Lilium. Protoplasma 77: 79-93
1.2 Early Events in Embryogenesis
F. Lo SCHIAVO l
1 Introduction
Plants, as compared to animals, are characterized by a great flexibility in their

ability to differentiate. This flexibility is manifested, in vivo, in the capacity to
generate and regenerate parts of the organism and, in vitro, with the acquisition
of totipotency. This characterization of general nature is not applicable as such to
all plant species. Cumulative work of many years in plant tissue culture shows
that primary explants of various types (hypocotyls, cotyledons, leaves, roots,
etc.) respond to synthetic media supplemented with auxins, typically 2,4-
dichloro-phenoxyacetic acid (2,4-0) and cytokinins (dispensable in some
systems) by dedifferentiating the tissues so treated. However, the acquisition of
morphogenetic capacity can occur with greater or lesser ease in the different
species, or even, within a species, according to the genetic background of the
varieties.
Analyzing the various events that lead, from the primary explant to the
formation of an embryogenic line, should help to answer some of the questions
posed in this chapter, namely:
1. Why, using the same type of explant and the same culture conditions, do some
species generate lines with, and other species lines without, morphogenetic
capacities?
2. Why, among the species that generate lines with morphogenetic capacities, are
some species (e.g., carrot) restricted to embryogenesis, and others (e.g.,
tobacco) to organogenesis?
In an attempt to answer these questions, the following points are dealt with:
1. Characteristics of morphogenetic lines and general rules used to choose the
proper primary explant.
2. The type of cells capable of generating embryogenic cell lines.
3. The response to hormones of the cells that will give rise, eventually, to
morphogenetically competent lines.
I Department of Biology, University of Padova, Via Trieste 75,35121 Pad ova, Italy

Early Events in Embryogenesis 21
2 Characteristics of Cell Lines and Morphogenetic Competence
Established cell lines originating from different explants and plants are very
similar in their behavior. Their differences in physiological parameters are
usually small while their range of morphogenetic capabilities is usually wide. In
fact, a cell line may: (1) contain or even consist of early embryonal stages that,
upon reduction of the phytohormones in the medium, complete their
developmental program and regenerate a plantlet; (2) regenerate organs in
sequence and, following a program of variation in the ratio of auxin to cytokinin
in the medium, give rise to entire plantlets; and (3) be incapable of expressing
developmental capacity and die in the absence of hormones.
The morphogenetic capacities of a cell line are acquired early in this history.
These capacities are maintained over a certain time until differentiation events
become, slowly, rarer and rarer. The morphogenetic potential of any cell line,
once acquired, cannot be changed at our will: we are unable to manipulate the
cell culture in such a way as to obtain, in a controlled way, embryogenesis from
a nonembryogenic line or even to cause loss of morphogenetic potential from a
morphogenetic line.
3 Importance of the Primary Explant
The choice of the starting material can be, in some species, of paramount
importance for obtaining a cell line with morphogenetic capabilities. Plant
species exist which are capable of expressing their morphogenetic potential
regardless of the explanted tissue (carrot, alfalfa, tobacco, etc.); in others, in
particular monocotyledons (Vasil and Vasil 1986) , only tissues with embryonal
character give rise to cell lines with morphogenetic capacities. In general, in a
given species, a primary explant will produce, under constant culture conditions,
lines that always show the same morphogenetic capabilities. For example, carrot
hypocotyls explanted in B5 medium + 2,4-D will give rise to embryogenic cell
lines (de Vries et al. 1988). In tobacco, piths explanted in MS+ NAA and kinetin
will give rise to organogenic cell lines (Skoog and Miller 1957; Nuti Ronchi 1981).
3.1 Identification of Progenitor Cells
Cell cultures have been used for quite some time in biochemical, physiological,
and genetic studies but very few attempts have been made so far to identify the
cells starting the culture nor their sensitivity to the hormones that cause cell
proliferation in culture.
Our experimental system is based on carrot embryogenic cell lines, which is
the model system to study somatic embryogenesis (Terzi and Lo Schiavo 1990).
With this system a histological analysis was carried out to identify the progenitor
cells. The starting material consisted ofhypocotyl fragments (2-3 mm long) from
22 F . Lo Schiavo
freshly germinated carrot seeds, submitted to various hormone treatments.

These hypocotyls gave rise to embryogenic cultures, always derived from the
procambium. After a short period of induction (3-6 days) the procambium cells
of the hypocotyls responded to the hormone treatment by dividing rapidly along
both axes: longitudinally and transversely. At the same time, the cells of the cor-
tical parenchyma and the epidermis elongate - but do not divide - and separate
from one another, liberating a procambial cylinder, actively proliferating, that
releases proliferating cells from its surface (Fig. 1); the same cells that will give rise
to the embryogenic culture (Guzzo et al. 1994).
The main results of the analysis, reported above, enable: (1) the identification
of the type(s) of cells capable of responding to hormonal stimuli; (2) the
determination of the temporal sequence of cell divisions leading to the formation
of proembryogenic masses (PEM) (Halperin 1966; Halperin and Jensen 1967),
which represent the first morphological evidence of embryo-competent cells; and
(3) the determination of the cell composition of the newly formed cell line.
The detailed knowledge, thus acquired, in a model embryogenic system such
as carrot, should help to identify and select the proper type of cells and to define
the culture conditions for cells of species recalcitrant to morphogenesis in vitro.
Fig. 1. Longitudinal section observed under the light microscope of the procambial proliferating mass
of a carrot hypocotyl soaked in medium with auxin for 10 days. At the outer surface number of cells
(arrowheads) separating from the proliferating mass are detectable. These are the cells that will give rise
to the embryogenetic culture; x 637
3.2 Differential Response of Cell Culture to Auxins
Another experimental approach is the study of cell "sensitivity" to auxin. A

primary explant may respond to the auxin-containing synthetic medium by
generating a class of cells capable of dividing and proliferating, or it may generate
cells that divide and acquire embryogenic or organogenic capacity. The
alternative explanation, i.e., that the culture conditions provide a selective
advantage to the types of cells referred to above, cannot be excluded. It would
then be possible, by changing the tissue of origin and the culture conditions, t6
achieve the desired cell composition in the newly formed line.
There are already classical examples of cells that respond differently to
hormone treatments: cells from tobacco, regardless of the tissue of origin, have
organogenetic properties, whereas cells from carrot, independent of their origin,
are embryogenic. Can we speak of selection, also in this case, or can we instead
consider differential responses to hormone treatments?
We investigated the behavior and the sensitivity of carrot and tobacco cell
lines to 2,4-D. 2,4-D was chosen because it is the auxin that most efficiently
induces embryogenic capacity in carrot (and in other species: e.g., alfalfa; Dudits
et al. 1991), which is our system of choice. Sensitivity to auxin was measured as
the auxin-binding capacity (ABC) of a crude membrane preparation of cells
exposed to different concentrations of2,4-D. An increase in the concentration of
the hormone in the medium causes an increase in ABC (Lo Schiavo et al. 1991).
This increase is shown by both carrot embryogenic cells and tobacco organogenic
cells although, in the former, the increase is much more pronounced. Other
differences concern the range of usable hormone concentrations (more limited in
tobacco, where a concentration ofl mglI2,4-D is already toxic) and the timing of
ABC induction (2 h in carrot and 10 days in tobacco). We repeated this type of
experiment, i.e., auxin-dependent induction of ABC, on primary explants. In
carrot, the analysis was carried out on cotyledons, hypocotyls, and roots
incubated in the presence of hormones. Hypocotyl was the strongest modulating
tissue and cotyledons the weakest: they have a constitutive low level of ABC that
remains constant in the presence of auxin. Roots showed an intermediate level
between the two other types of organs analyzed.
These experiments are limited in that one is dealing in each case with several
types of cells. Histological analysis will allow one to focus experiments on the
cells known to be the progenitors of the cell line.
Experimental induction of ABC has also been carried out on leaves and
plantlets of Nicotiana tabacum, grown in vitro; the results were in line with those
obtained on cell cultures.
4 Biological Significance of Modulation
ABC modulation is shown by both Nicotiana and carrot, our model systems for
embryogenesis and organogenesis, respectively. Alterations of the modulation
process were sought in order to gain some hints on the biological significance of
24 F. Lo Schiavo
the process. To this end we used mutant carrot cell lines, altered in the
differentiation processes, isolated in our laboratory, previously or during this
investigation.
Modulation was measured in:
- wild-type lines having different embryogenic capacities (indicated in Tables 1
and 2 as A + and derivatives);
- auxin-resistant lines (incapable of embryogenesis) that were obtained by
plating cells in the presence of 451lM 2,4-D (indicated in Tables 1 and 2 as the
w series);
isolated emb-lines incapable of embryogenesis (indicated as the MF series in
Table 1);
- a habituated line (E9) characterized by alterations in IAA metabolism.
The results, presented in Table 1, indicate that every time modulation was
reduced or absent - due to a high constitutive level or low inducibility-
embryogenesis was also altered or totally abolished (Filippini et al. 1992).
Knowing that the amount of ABC correlates, over a certain range, with the
amount of 2,4-D in the medium, we treated the wand MF series with higher
concentrations of2,4-D and were able to increase the amplitude of modulation
and, concomitantly, to regain embryogenicity (Table 2).
In tobacco, where we had no mutant cell lines, a cell line derived from a plant
transformed with rolB (Cardarelli et al. 1987) was used. It is known that rolB
transformation confers to the cell an increased sensitivity to auxin in
electrophysiological assays (Barbier-Brygoo et al. 1989). Modulation in trans-
formed cell lines is also altered in a characteristic way: greater amplitude and the
Table 1. IAA binding assay in vitro pmoles of IAA bound to 50 J.1g of membrane protein extracted
from cells kept for 4 days in the presence or absence of 2.3 J.1M 2,4-D
Cell line 2,4-D Embryogenic potential'

+
A)
NTZ 6.0 39.6 ++
A+TI 7.1 36.2 +
A+ 14.6 24.5
B)
MFI 3.0 6.6
MF2 2.8 7.2
MF4 3.2 6.4
MF7 2.7 3.1
E9 17.7 26.6 ±
wI 3.9 10.1
w2 4.0 17.0
w3 3.7 21.7
w4 4.3 14.4
a++, Embryogenic efficiency greater than 0.5 embryos formed per cell unit; +, efficiency between 0.05
and 0.4 embryos/cell unit; ±, efficiency on the order of 0.001 to 0.04; -, an efficiency always lower than
0.0001 (i.e., embryos never observed under the conditions used).
Table 1. Correlation between ABC and embryogenesis measured at

different 2,4-D concentrations
Lines 2,4-D concentration UtM)

2.3 45 90
ABC' EEb ABC EE ABC EE
A+T2 33.9 ++
MF2 7.2 - 16.4 - 26.6 +
wi 10.1 - 22.8 +
w2 17.0 - 30.5 +
w3 21.7 - 29.5 +
w4 14.4 - 26.1 +
a ABC reported as pn101es bound by 50 Ilg membrane preparation.
bEE denotes embryogenic efficiency, measured as in Table I.
possibility of using a greater range of auxin concentrations. Roots, practically

absent in cultures of nontransformed cells, are generated with a characteristic
high frequency by rolB- transformed cells.
The isolation, from carrot embryogenic cell lines, of cell lines that proliferate
normally but without modulating capacity and without differentiation capacity,
indicates that the type of hormone response that leads to cell division is
independent of the type of response that leads to acquisition of totipotency and
generation ofPEMs in the presence of auxin. Consequently, it is possible that two
classes of auxin-binding proteins (ABP) exist one responsible for cell division,
and the other, capable of inducing, besides cell division, regeneration. The class
of ABPs that are induced in response to auxin determines the fate of cell
morphogenesis. Plant species recalcitrant to differentiation events may perhaps
generate their lines by selecting cells only capable of expressing ABPs involved in
cell division. By assaying modulation on primary explants it can be determined
whether the starting material is good, or whether the wrong selection has been
made for generating cell lines. If the starting material is not modulated in
response to various auxin concentrations, we are using either inappropriate
culture conditions or inappropriate tissue.
5 Biochemical Identification of the Various Classes of ABPs
The claim that different classes of ABPs exist with different physiological roles
has been verified in tobacco and carrot, where the ABP class induced in the
phenomenon of modulation could be removed from the membrane with salt
treatments that leave the constitutive ABP class unaffected. Further definition of
the ABP classes can be achieved with structural auxin analogs unable to induce
some of the classical auxin-stimulated phenomena (cell elongation, ethylene
production), but capable of interacting with auxin-binding sites involved in
differentiation responses. Competition experiments were performed with such
26 F. Lo Schiavo
analogs and redioactive auxin and the results demonstrated that there is
competition for different sites by different analogs. In particular, 1,2-
benzisoxazole-3-one (BOO) (Branca et al. 1991) binds the labile site involved in
differentiation and 1,2-benzisoxazole-3-acetic acid (BOA) (Branca et al. 1990)
binds the stable site involved in cell division. From a physioloagical point of view,
BOO induces malformations in the differentiation pathway of carrot embryo,
whereas BOA causes in the embryo an early block (reversible) in the radical pole.
Other drugs that alter differentiation pathways, possibly by interfering with
auxin-binding sites, are oligo saccharides (Branca et al. 1988); these same drugs
may be useful in the further classification of distinct auxin-binding sites. The
identification of distinct auxin-binding sites, specifically recognized by auxin
analogs, could help us to isolate ABP mutants, with the subsequent possibility of
attributing the various physiological roles to those of ABPs.
6 Cellular Mechanisms That Generate Totipotency
The first step, in which some cells of the primary explant respond to auxin and
give rise, in carrot, to cell lines capable of modulation, is followed by a second,
which is morphologically defined by the appearance ofPEMs, which characterize
the embryogenic lines. PEMs have been purified from cultures of proliferating
cells and incubated with different auxin concentrations. This subpopulation of
cells is "insensitive" to auxin during all embryonal stages up to the mature
globular embryo. Thus, in a proliferating cell culture, two cell populations are
present: modulating cells and PEMs, the latter generated from modulating cells
but insensitive to auxin. At later embryonal stages (from heart-shaped embryos
on) some tissues lose the characteristics of the primary meristem and start
responding to auxin. The question arises as to how the progenitor embryonal
cells - whose response to auxin does not consist in division and elongation - are
generated. By examining the behavior of the first identifiable embryonal stage
(PEM) in auxin, we noticed that the cell, endowed with a conspicuous vacuole,
differs markedly from an embryonal meristem.
After PEMs are transferred to a medium without auxin, cells fill with
cytoplasm in a matter of hours, so that they acquire the appearance of an
embryonal meristem; frequent divisions follow and in few hours (48-72 h) the
globular stage is achieved (Fig. 2).
In order for a totipotent cell to arise perhaps it is necessary - and also
sufficient - to attain a certain configuration of the auxin receptor, which, after
complete loading, remains unable to bind further amounts of the hormone. This
hypothesis might explain why some plant species are embryogenic, whereas
others, unable to reach the postulated configuration of the auxin receptor, never
gain totipotency. If this hypothesis is correct, it should become possible, in a
more or less distant future, to purify, characterize and clone single components of
the receptor complex from, e.g., carrot, and upon gene transfer, to express the
proper component, capable of providing embryogenic capacity to recalcitrant
species.
Fig. 2. Section, seen under the transmission electron microscope, of a PEM after 3 days in a medium
without auxin. Some modifications occurred in the PEM cells so that they now appear rich in cytoplasm
with numerous small vacuoles (V). N Nucleus; P Plastid with starch; x 2475
Another suggested mechanism capable of giving totipotency is somatic

meiosis. Nuti Ronchi et al. (see Chapter I.l, this Vol.) have described in cell
cultures a series of events exactly matching the steps of meiosis as it occurs during
gametogenesis and they suggest that this mechanism generates in vitro a cell (the
progenitor cell of the somatic embryo) which assumes a role similar to the
fertilized ovule at the onset of embryogenesis. A consequence of such a
mechanism would be haploidization and segregation with generation of a great
amount of genetic variability, such as that seen, e.g., in the phenomenon of
somaclonal variation. To this haploidization, an endoreduplication giving rise to
a completely homozygous diploid, will soon follow.
28 F. Lo Schiavo
7 Summary and Conclusions
It is a general rule that plant somatic tissues, when exposed to auxin in vitro,
dedifferentiate and generate proliferating cell lines. Not all cell lines, however,
acquire morphogenetic capacities.
Embryogenic cell lines of carrot have been shown to induce microsomal
auxin-binding proteins in relation to the amount of auxin present in the culture
medium. This modulating capacity is reduced or lost whenever the embryogenic
capacity is altered or absent. These data suggest that the cells that, in response to
auxin, proliferate but do or do not acquire morphogenetic capacity, should
possess different sets of auxin receptors and that the auxin receptors responsible
for cell proliferation are not the same receptors responsible for morphogenetic
responses. It is also likely that different morphogenetic responses are under the
responsibility of different auxin receptors. All these receptors are not necessarily
present all the time on the cell membrane, but their presence can be stimulated by
the hormone itself. The great variability encountered on the same tissue from
different species indicates that, within a certain range of nontoxic auxin
concentrations, some species show all types of auxin receptors (or, perhaps, all
components of the auxin receptor complex), whereas other species do not
respond to variations in auxin concentration and do not use modulation to vary
their receptor constitution.
The identification and purification of the various auxin receptors (or of the
various components of the receptor complex) will no doubt contribute to our
understanding of the cellular and molecular mechanisms underlying morpho-
genetic capacity and behavior of the plant cells.
Acknowledgment. This research was supported by the National Research Council of Italy, Special
Project RAISA, Subproject No.2, Paper No. 1566.
References
Barbier-Brygoo H, Ephritikine G, Kliimbt D, Ghislain M, Guern J (1989) Functional evidence for an

auxin receptor at the plasmalemma of tobacco mesophyll protoplasts. Proc Nat! Acad Sci USA 86:
891-895
Branca C, De Lorenzo G, Cervone F (1988) Competitive inhibition of the auxin-induced elongation by
-D- oligogalacturonides in pea stem segments. Plant Physiol91: 889-897
Branca C, Torelli A, Bassi M (1990) Effects of benzisoxazole and benzisothiazole on tomato plant
regeneration in vitro. Plant Cell Tissue Organ Cult 21: 17-19
Branca C, Ricci A, Fermi P, Bassi M (1991) Activity of 1,2-benzisoxazole-3 -one and indole -2,3-dione
on plant regeneration in vitro and on cell elongation. Plant Cell Rep 10: 498-500
Cardarelli M, Mariotti D, Pomoni M, Spano L, Capone I, Costantino P (1987) Agrobacterium
rhizogenes T-DNA genes capable of inducing hairy root phenotype. Mol Gen Genet 20: 475-480
de Vries SC, Booij H, Meyerink P, Huisman G, Wilde HD, Thomas TL, van Kammen A (1988)
Acquisition of embryogenic potential in carrot cell-suspension cultures. Planta 176: 196-204
Filippini F, Terzi M, Cozzani F, Vallone D, Lo Schiavo F (1992) Modulation of auxin binding proteins
in cell suspensions. II. Isolation and initial characterization of carrot lines impaired in somatic
embryogenesis. Theor Appl Genet 84: 430-434
Guzzo F, Baldan B, Mariani P, Lo Schiavo F, Terzi M (1994) Studies on the origin of totipotent cells
in explants of Daucus carota L. J Exp Bot 45: 1427-1432
Halperin W (1966) Alternative morphogenetic events in cell suspensions. Am J Bot 53: 443-453
Halperin W, Jensen WA (1967) Ultrastructural changes during growth and embryogenesis in carrot
cell cultures. J Ultrastruct Res 18: 428-443
Lo Schiavo F, Filippini F, Cozzani F, Vallone D, Terzi M (1991) Modulation of auxin binding proteins
in cell suspensions. I. Differential responses of carrot embryo cultures. Plant Physiol 97: 60-64
Nuti Ronchi V (1981) Histological studies of organogenesis in vitro from callus cultures of two
Nicotiana species. Can J Bot 59: 1969-1977
Skoog F, Miller CO (1957) Chemical regulation of growth and organ formation in plant tissue cultured
in vitro. Symp Soc Exp BioI II: 118-140
Terzi M, Lo Schiavo F (1990) Somatic embryogenesis. In: Bhojwani SS (ed) Plant tissue culture:
applications and limitations. Elsevier, Amsterdam, pp 54--66
Vasil IK, Vasil V (1986) Regeneration in cereal and other grass species. In:Vasil IK (ed) Cell culture
and somatic cell genetics of plants vol3. Plant regeneration and genetic variability. Academic Press,
Orlando, pp 121-150
1.3 Molecular Basis of Somatic Embryogenesis
R. KAWAHARA! and A. KOMAMINE2
1 Introduction
Somatic embryogenesis is an ideal system for the investigation of the

differentiation process in plants. In contrast to zygotic embryogenesis, somatic
embryogenesis can easily be observed, the external conditions of the embryo can
be controlled, and large quantities of embryos can be easily obtained, etc.
Since the first reports of somatic embryogenesis (Reinert 1958; Steward et al.
1958), the carrot (Daucus carota L.) has been used as a model plant. Generally,
carrot cells are maintained in a medium that contains auxin and proliferates in an
unorganized manner. Embryogenesis is easily induced by removal of auxin from
the medium whereupon subsequent differentiation to globular, heart-shaped,
and then torpedo-shaped embryos occurs.
Removal of auxin from the medium induces somatic embryogenesis in
carrot suspension cultures, although at low frequencies and asynchronously. In
such systems, biochemical and molecular events specific for embryogenesis are
diluted by the activities of cells not engaged in embryogenesis. Furthermore, only
average values for biochemical parameters related to various stages of embryo-
genesis can be determined when asynchronous systems are used. Thus, high
frequency, synchronous embryogenesis systems are required for investigation of
somatic embryogenesis mechanisms, especially at the molecular level.
We established suitable systems for this purpose using carrot suspension
cultures (Fujimura and Komamine 1979a). Embryogenic cell clusters, which are
designated State 1 cell clusters, were selected by sieving with nylon screens and
subsequent density gradient centrifugation in Ficoll solution (Fig. 1). Isolated
State 1 cell clusters were transferred to media lacking auxin, but containing
zeatin, and synchronous embryogenesis occurred at about 90% frequency. By
using this system, it has become possible to investigate the molecular mechanisms
of somatic embryogenesis, especially during early phases of differentiation.
I Biological Institute, Faculty of Science, Tohoku University, Sendai 980 Japan.

Present address: Gene Bank, Tsukuba Life Science Center, The Institute of Physical and Chemical
Research (RIKEN), 3-1-1 Koyadai, Tsububa, Ibaraki 305 Japan.
2 Departmentof Chemical and Biological Sciences, Japan Women's University, Mejiro, Tokyo, 113
Japan.

~Springer.Verlag Berlin Heidelberg 1995
Molecular Basis of Somatic Embryogenesis 31
sieving
.
18% Ficoll
- - !!!i$
.:.
3.000rpm 5min
G
oQ Q
State 1
Torpedo Globular
Fig. 1. Induction of synchronous and high frequency somatic embryogenesis by sieving and density
gradient centrifugation
2 Expression of Polarities in Early Stages

2.1 Relationship Between Expression of Polarities and Auxin
Auxin is the most important factor for regulation of induction and development
of embryogenesis. Usually carrot suspension cultures are maintained in a
medium containing 2,4-D and cells proliferate in an unorganized manner.
Somatic embryogenesis is induced by transferring embryogenic cell clusters to
auxin-free medium. Auxin is inhibitory to embryogenic development (Fig. 2),
and if added to the medium after prior withdrawal, development of embryos will
stop. After the heart-shaped stage, addition of 2,4-D deforms embryos, and in
both cases, cells begin proliferating in an unorganized manner.
If auxin inhibits development, anti auxins would be expected to promote
embryogenic development. However, the antiauxins, 2,4,6-trichlorophenoxy-
acetic acid, and p-chlorophenoxyisobutyric acid (PCIB) inhibited development
(Fujimura and Komamine 1979b), even though embryos contained endogenous
auxin during culture.
These results (Fig. 3) may be explained as follows: morphological polarity is
expressed through the first unequal division in the developing embryo. This is
initially manifested through endogenous auxin in the embryo which, through its
32 R. Kawahara and A. Komamine
:....
2r
...... .
. .
Fig.2. Process of somatic embryogenesis from State I cell cluster. Embryo development from State I
cell cluster is inhibited by 2,4-D. Addition of 2,4-D after the heart-shaped stage deforms the embryo
polar distribution, brings about morphological polarity. If auxin is added to the

medium, the acquired endogenous gradient is disrupted through the diffusion of
exogenous auxin in the embryo and results in the inhibition of further
development. In the case of antiauxins, they inhibit the action of endogenous
auxin, resulting in inhibition of embryogenic development.
. ~X I N
; v::;, AN T~Q
B * E MBRYO
DE VEL OPME NT
~c
Fig.3A-C. Hypothesis for the effect of auxin and antiauxin on embryo development. A Endogenous
auxin is distributed in the embryo with polarity. B Exogenous auxin diffuses in the embryo and disrupts
the acquired endogenous gradient. C Antiauxins inhibit the action of endogenous auxin
2.2 Polarized DNA Synthesis and Cell Division at Early Stages

As mentioned above, State I cell clusters differentiate synchronously to form

globular embryos by transferring them to auxin-free medium. By serial
observation, rapid cell division was found to occur during this process in certain
Statel State2 Globu l ar
Fig. 4. Polarized cell division in the transition from State 1 embryogenic cell cluster to globular
embryo. Hatching shows actively dividing cells
a b
( kD :
- 116
92
- 66
- 45
- 31
Fig. 5. SDS-PAGE of protein extracted from 21

ADS and NDS cells. ADS-specific bands are
14
indicated by arrows. a ADS cells; bNDS cells
parts of cell clusters. We designated such cell clusters State 2 cell clusters in which
cell division was very rapid during the formation of globular embryos. Doubling
times were estimated as 58 h during the formation of State 2 cell clusters from
State 1 cell clusters, 6.3 h in State 2 cell clusters, which form globular-stage
embryos, and 20 h after the globular stage (Fujimura and Komamine 1980).
To confirm the polarity of cell division, three-dimensional reconstructions of
serial sections from 3H-thymidine pulse-chase labeled cells were performed. In
the first 3 days of culture in the absence of auxin, DNA synthesis occurred
randomly. However, polarized DNA synthesis was evident during the transition
ofthe State 2 cell cluster to the globular embryo. In globular embryos, high DNA
synthesis activity was observed in procambial and proepidermal cells, while no
DNA synthetic activity was detected in the suspensor-like structure (Fig: 4).
State 2 cell clusters were used to produce protoplasts and were subsequently
fractionated by density gradient centrifugation in Percoll to investigate the
difference, at the molecular level, between actively DNA synthesizing cells (ADS
cells) and nonactively DNA synthesizing ones (NDS cells). Protein patterns were
compared in both cell types using 35S-methionine and SDS-PAGE (Fig. 5). These
proteins were detected in ADS cells, but not in NDS cells, and may be potential
markers of polarity during DNA synthesis specific for embryogenesis.
3 Molecular Aspects of Somatic Embryogenesis
3.1 Molecular Markers of Somatic Embryogenesis
Much work has gone into finding specific molecular markers for somatic
embryos. Two-dimensional gel electrophoresis of proteins has identified several
polypeptides that are specifically expressed in association with somatic
embryogenesis. Some polypeptides were expressed in particular stage during
embryo development (Sung and Okimoto 1981; Racusen and Schiavone 1988),
some were in embryonic endomembranes and plasma membranes (Slay et al.
1989), and some were in extracellular medium of embryogenic cultures (de Vries
et al. 1988).
We also found several molecular markers in the carrot system. Using in vitro
translation and two-dimensional gel electrophoresis, more then 99% of the
polypeptides produced were found to show the same patterns between
embryogenic and nonembryogenic cultures. Four different translatable mRNAs
encoding polypeptides could be detected, with two appearing in embryogenic
cultures, while the other two appeared in nonembryogenic cultures.
These results indicate that only a few proteins may play important roles
during embryogenesis, and that changes in protein patterns are regulated at the
transcriptional level.
Smith et al. (1988) reported a nuclear protein associated with cell division
which reacted with the monoclonal antibody designated as 21D7. Using the
21D7 antibody in our system, it was found that the expression of the 21D7
protein may be essential for expression of totipotency. The cDNA clone corre-
sponding to the 21D7 protein was isolated from a cDNA library of carrot root
by an immunoscreening method and the nucleotide sequence was determined.
The predicted amino acid sequence was found to have sequence similarity with
the amino acid sequence of the mouse gene, tum-transplantation antigen P91A
(Lirquin et al. 1989), the function of which is unknown.
3.2 Gene Expression During Somatic Embryogenesis
The most attractive approach to elucidate the mechanisms of somatic embryo-

genesis is to isolate genes which are expressed specifically during embryogenesis
and to reveal their function. Choi et al. (1987) isolated several cDNA clones that
were preferentially expressed during somatic embryogenesis in carrot, by a
combined immunoadsorption and epitope selection method. Regulation of the
expression of two clones, DC8 and DC59, was analyzed in detail, and it was
found that their mRNAs were accumulated not only in somatic embryos but
also in mature zygotic embryos (Borkird et al. 1988). In addition, their expres-
sion was found to be controlled by abscisic acid (Hatzopoulos et al. 1990a,b).
Sequence analyses showed that the proteins encoded by DC8 and DC59 were
similar to Lea (late embryogenesis abundant) proteins (Dure et al. 1989) and
lipid body membrane proteins (Hatzopoulos et al. 1990b), respectively. By
differential screening, Wilde et al. (1988) and Ulrich et al. (1990) also isolated
cDNA clones (Dc3 and EMB-l, respectively) that encoded proteins similar to
Lea proteins in carrot. The Lea proteins contain phylogenetically conserved
elements and are thought to function in protecting cellular structures during seed
desiccation (Dure et al. 1989). The accumulation of EMB-l mRNA was specific
for embryos and increased during somatic embryogenesis (Wurtele et al. 1993),
while the Dc3 mRNA was accumulated at almost the same rate in both
unorganized cells and somatic embryos (Wilde et al. 1988). In situ hybridization
showed a similar distribution of EMB-l mRNA in both zygotic and somatic
embryos (Wurtele et al. 1993). These results suggest that somatic embryos share
the same developmental program as zygotic embryos, although the function of
the products of those genes is unclear. The expression pattern also indicates that
such Lea-like proteins playa role in embryogenesis at a later stage during embryo
development. In order to understand the regulation mechanisms of the
expression of these genes, the genomic sequences of these clones have been
isolated, and extensive analysis has been carried out to find cis elements of
promoter regions and trans-acting factors that are responsible for
developmental, hormonal, and environmental regulation (Franz et al. 1989;
Hatzopoulos et al. 1990b; Seffens et al. 1990; Goupil et al. 1992; Vivekananda
et al. 1992).
Sterk et al. (1991) reported tissue-specific expression of a lipid transfer
protein gene, EP2, during embryogenesis in carrot. A cDNA clone for the gene
was obtained by the expression screening with an antiserum raised against all
proteins secreted into the medium of somatic embryo cultures. Although the
temporal accumulation of its transcripts was not restricted in embryo
development, in situ hybridization showed specific accumulation of the tran-

scripts in protoderm cells of both somatic and zygotic embryos. Based on the
extracellular location ofEP2 protein and protoderm-specific expression of EP2,
it is proposed that EP2 protein plays a role in the transport of cutin monomers
through the extracellular matrix to sites of cutin synthesis.
So far, only a few genes have been isolated that are transiently expressed
during early stages of somatic embryogenesis. Aleith and Richter (1990) isolated
several cDNA clones whose expression was roughly associated with the first
morphogenetic or globular stage by differential screening. From the sequence
analysis, it was found that two of them encode glycine-rich proteins and one of
the remaining clones encodes a polypeptide with a proline-rich domain.
However, their function is still unknown.
We constructed Agt11 cDNA libraries from poly (Ar RNA of hypocotyls
and roots of carrot seedlings, and screened the cDNA libraries differentially to
isolate hypocotyl or root-specific cDNAs. Two cDNAs were isolated, CAR3 and
CAR4, which were specifically expressed in hypocotyls, and two additional
cDNAs, CARS and CAR6, were found to be specifically expressed in roots.
Expression of these four cDNAs was investigated during embryogenesis by
Northern hybridization.
Relative expression of CAR4 (Fig. 6) and CARS increased after the form-
ation of globular embryos and that of CAR6 increased after the formation of
heart-shaped embryos, while CAR3 was expressed earlier (before globular
c 1
0
'iii
I/)
CP
~
0.
-
)(
CP
0
Qi
>
~ 0.5
-
CP
.:: Fig. 6. Expression of mRNA of
CQ the hypocotyl-specific gene, CAR4,
Qi during somatic embryogenesis.
CI:
Relative expression of CAR4
increase after the globular stage
(.). In cells cultured in the medium
containing 2,4-D. CAR4 mRNA
O~==~==~L-__•____~~______~__ is expressed at a very low level (.).
o 7 14 21 The expression is suppressed by
adding 2,4-D to heart-shaped
Time of culture ( day) embryos (0)
embryos). Expression of these cDNAs was at a very low level in cells which were
cultured in medium containing 2,4-D, and was strongly suppressed when 2,4-D
was added to heart-shaped embryos (Fig. 6). In situ hybridization analysis
revealed that CAR4 was expressed in the epidermis and in the regions
surrounding tracheary elements in torpedo-shaped embryos.
The predicted amino acid sequence of the protein encoded by CAR4 was
rich in proline (N-terminal region) and in leucine (C-terminal region). A charac-
teristic repeated motif was found in the proline-rich region, which resembled
repeated sequences found in proline-rich cell wall proteins, such as p33 (carrot)
(Chen and Varner 1985) or PRP (soybean) (Hong et al. 1987). The predicted
amino acid sequence of CARS was found to have sequence similarity with the
amino acid sequence of membrane channel proteins (Yamamoto et al. 1990).
Isolation of genes specific for earlier stages of embryogenesis was also
attempted. We succeeded in cloning five cDNAs by differential screening
between State 1 cell clusters, which were cultured in the absence of auxin for 5
days (preglobular embryos), and those which were cultured in the presence of
auxin for 5 days. These clones were designated CEM 1,2,3,4, and 5. One of them,
CEM 1, was shown to be expressed preferentially prior to and after the globular
stage of embryogenesis (Kawahara et al. 1992). The nucleotide sequence and the
predicted amino acid sequence of the protein encoded by CEMI was found to
show high similarity to the elongation factor (EF-la) of eukaryotic cells. The
similarities were 76.4% for human, 76.8% for Xenopus, 73.1 % for yeast, 81 % for
Euglena, and 94.2% for Arabidopsis. EF-la is an essential protein in the elon-
gation of the peptide chain in protein synthesis. The distribution of CEMI
Fig. 7A,B. In situ hybridization analysis of globular embryos. Carrot globular embryos were
hybridized to 35S-labeled antisense (A) or sense CEMI RNA (B)
mRNA during somatic embryogenesis in carrot cells was determined by in situ

hybridization analysis. Accumulation of specific mRNA was observed in the
spherical region of globular embryos (Fig. 7) and in the meristematic region of
heart- and torpedo-stage embryos. Distribution of CEMI mRNA was also
closely associated with cell division activity during embryogenesis.
Several other cDNA clones that are expressed specifically in earlier stages of
somatic embryogenesis have been cloned by the substraction method. One of the
clones isolated by this method, CEM6, was shown to be expressed in State 2 cell
clusters and globular embryos. This clone's expression was not detectable in
suspension cultures or State 1 cell clusters cultured in the presence of auxin, and
the predicted amino acid sequence contains many repeats of "Gly-Gly-*".
By using a high frequency and synchronous embryogenesis system, the following

findings regarding mechanisms of embryogenesis at the molecular level,
especially in the early stage, have been revealed.
Expression of Polarities. Rapid and polarized cell division occurs in the tran-
sition from State I embryogenic cell cluster to globular embryo. Auxin is
important for this process and for maintenance of polarities during embryo-
genesis.
Gene Expression During Embryogenesis. The EF-la gene was expressed during
embryogenesis, correlating with cell division activity in the embryo. The organ-
specific CAR genes were expressed at an early stage of the embryo (globular to
heart-shaped embryos), and their expression increased with the progress of
embryogenesis and was suppressed by auxin. The embryo-specific gene, CEM6,
which was expressed in State 2 and globular embryos, was cloned by the
substraction method.
However, the final goal, i.e., the complete elucidation of inductive and
control mechanisms during embryogenesis, is still far from being realized. The
list of genes affecting embryogenesis continues to grow, however, the elucidation
of their biochemical functions lags seriously behind, thus, creating a bottleneck in
our current knowledge. Therefore, through the use of molecular techniques, it is
anticipated that many functions can be matched to known genes. For example,
through the use of antisense RNA and microinjection or particle gun and
subsequently following the developmental fate of treated cells and tissues, we
should be able to more easily identify gene function, thereby separating
constitutive from embryo-specific functions. Developmental mutants are also
invaluable tools and will help in the genetic analysis of embryogenesis (Lo
Schiavo et al. 1990; de long et al. 1992). It must also be considered that
embryogenesis-related gene function and control may be complicated by
coordinate and temporal regulation of inductive and developmental "gene
families" and regulation by differential expression and cascade mechanisms.
References
Aleith F, Richter G (1990) Gene expression during induction of somatic embryogenesis in carrot
cell suspensions. Planta 183: 17-24
Borkird C, Choi JH, Jin Z, Franz G, Hatzopoulos P, Chorneau R, Bonas R, Pelegri F, Sung ZR
(1988) Developmental regulation of embryonic genes in plants. Proc Natl Acad Sci USA 85:
6399-6403
Chen J, Varner JE (1985) Isolation and characterization of cDNA clones for carrot extension and a
proline-rich 33-kDa protein. Proc Natl Acad Sci USA 82: 4399-4403
Choi JH, Liu LS, Borkird C, Sung ZR (1987) Isolation of cDNA clones for rate embryo-specific
antigens in carrot cell cultures. Proc Nat! Acad Sci USA 84: 1906-1910
de Jong AJ, Cordewener J, Lo Schiavo F, Terzi M, Vandekerckhove J, van Kammen A, de Vries SC
(1992) A carrot somatic embryo mutant is rescued by chitinease. Plant Cell 4: 425-433
de Vries SC, Booij H, Jassens R, Vogels R, Saris L, Lo Schiavo F, Terzi M, van Kammen A (1988)
Carrot somatic embryogenesis depends on the phytohormone-controlled presence of correctly
glycosylated extracellular proteins. Genes Dev 2: 462-476
Dure III L, Crouch M, Harada J, Ho TD, Mundy J, Ralph Q, Thomas T, Sung ZR (1989) Common
amino acid sequence domains among the LEA proteins of higher plants. Plant Mol Bioi 12:
475-486
Franz G, Hatzopoulos P, Jones TJ, Krauss M, Sung ZR (1989) Molecular and genetic analysis of an
embryonic gene, DC8, from Daucus carota L. Mol Gen Genet 218: 143-151
Fujimura T, Komamine A (1979a) Synchronization of somatic embryogenesis in a carrot cell
suspension culture. Plant Physiol64: 162-164
Fujimura T, Komamine A (1979b) Involvement of endogenous auxin in somatic embryogenesis in a
carrot cell suspension culture. Z Pflanzenphysiol95: 13-19
Fujimura T, Komamine A (1980) The serial observation of embryogenesis in a carrot cell suspension
culture. New Phytol 86: 213-218
Goupil P, Hatzopoulos P, Franz G, Hempel FD, You R, Sung ZR (1992) Transcriptional regulation
ofa seed-specific carrot gene, DC8. Plant Mol Bioi 18: 1049-1063
Hatzopoulos P, Fong F, Sung ZR (1990a) Abscisic acid regulation ofDC8, a carrot embryonic gene.
Plant Physiol 94: 690-695
Hatzopoulos P, Franz G, Choy L, Sung ZR (1990b) Interaction of nuclear factors with upstream
sequences of a lipid body membrane protein gene from carrot. Plant Cell 2: 457-467
Hong JC, Nagao RT, Key JL (1987) Characterization and sequence analysis of a developmentally
regulated putative cell wall protein gene isolated from soybean. J Bioi Chern 262: 8367-8376
Kawahara R, Sunabori S, Fukuda H, Komamine A (1992) A gene expressed preferentially in the
globular stage of somatic embryogenesis encodes elongation factor la in carrot. Eur J Biochem
209: 157-162
Lirquin C, van Pel A, Mariame B, de Plaen E, Szicora J -P, Jansens C, Reddenhase MJ, Lejeune J, Boon
T (1989) Structure of the gene of tum-transplantation antigen P9IA: The mutated exon encodes a
peptide recognized with Ld by cytolytic T cells. Cell 58: 293-303
Lo Schiavo F, Giuliano G, de Vries SC, Genga A, Bollini R, Pitto L, Cozzani F, Nuti-Ronchi V, Terzi
M (1990) A carrot cell variant temperature sensitive for somatic embryogenesis reveals a defect in
the glycosylation of extracellular proteins. Mol Gen Genet 223: 385-393
Racusen RH, Schiavone FM (1988) Detection of spatially- and stage-specific proteins in extracts from
single embryos of the domesticated carrot. Development 103: 665-674
Reinert J (1958) Untersuchungen iiber die Morphogenese an Gewebekulturen. Ber Dtsch Bot Ges
71: 15
Seffens WS, Almoguera C, Wilde HD, Von der Haar RA, Thomas TL (1990) Molecular analysis of
phylogenetically conserved carrot gene: developmental and environmental regulation. Dev Genet
11: 65-76
Slay RM, Grimes HD, Hodges TK (1989) Plasma membrane proteins associated with undifferentiated
and embryonic Daucus carota tissue. Protoplasma 150: 139-149
Smith JA, Krauss MR, Borkird C, Sung ZR (1988) A nuclear protein associated with cell divisions in
plants. Planta 174: 462-472 .
40 R. Kawahara and A. Komamine: Molecular Basis of Somatic Embryogenesis
Sterk P, Booij H, Schellekens GA, van Kammen A, de Vries SC (1991) Cell-specific expression of the
carrot EP21ipid transfer protein gene. Plant Cell 3: 907-921
Steward FC, Mapes MO, Mears K (1958) Growth and organized development of cultured cells. II.
Organization in cultures grown from freely suspended cells. Am J Bot 45: 705-708
Sung ZR, Okimoto R (1981) Embryonic proteins in somatic embryos of carrot. Proc Nat! Acad Sci
USA 78: 3683-3687
Ulrich TH, Wurtele ES, Nikolau BJ (1990) Sequence of an mRNA accumulating specifically in
embryos of carrot. Nucleic Acids Res 18: 2826
Vivekananda J, Drew MC, Thomas TL (1992) Hormonal and environmental regulation of the carrot
lea-class gene Dc3. Plant Physiol100: 576-581
Wilde HD, Nelson WS, Booij H, de Vries SC, Thomas TL (1988) Gene-expression programs in
embryogenic and nonembryogenic carrot cultures. Planta 176: 205-211
Wurtele ES, Wang H, Durgerian S, Nikolau BJ, Ulrich TH (1993) Characterization of gene that is
expressed early in somatic embryogenesis of Daucus carota. Plant Physiol102: 303-312
Yamamoto YT, Cheng CL, Conkling MA (1990) Root specific genes from tobacco and Arabidopsis
homologous to an evolutionarily conserved gene family of membrane channel proteins. Nucleic
Acids Res 18: 7449
1.4 Gene Expression in Somatic Embryos
H.D. WILDE l , W.S. SEFFENS2, and T.L. THOMAS 3
1 Introduction
The developmental biology of somatic embryos is similar in many respects to

that of plant zygotic embryos. For that reason, the analysis of gene expression
during somatic embryogenesis has been viewed as an alternative means to
investigate the genetic control of zygotic embryo development, which can be
technically difficult. In the study of gene expression during somatic
embryogenesis, the molecular events associated with embryo development must
be distinguished from those associated with embryonic determination. Research
into the cytology and physiology of embryonic determination in somatic carrot
cells has made this a useful system for molecular analysis (Komamine et al.
1992). A comprehensive review of molecular and cellular aspects of somatic
embryogenesis, with emphasis on the alfalfa system, has been published by
Dudits et al. (1991). In this chapter the work of our laboratory and others on the
molecular biology of carrot somatic embryogenesis is discussed. (also see
Chapter 1.3, this Vol.)
2 Embryogenesis from Somatic Carrot Cells
A population of single cells (Type 1) can be isolated from carrot suspension

cultures that gives rise to embryos at a high frequency when removed from the
presence of the auxin analog, 2,4-dichlorophenoxyacetic acid (Nomura and
Komamine 1985a; Komamine et al. 1992). Paradoxically, Type 1 cells are not
embryogenic in auxin-free medium; instead they form elongated cells that do not
proliferate. Type 1 cells differentiate into embryos only after undergoing cell
division in the presence of2,4-D and zeatin, a cytokinin. It was observed that the
first division of this somatic cell was unequal and that continued mitosis in one of
the daughter cells produced an embryogenic cluster of cells (Backs-Hiisemann
and Reinert 1970; Nomura and Komamine 1986). As they divided, cells of this
embryogenic cluster decreased in size and became highly cytoplasmic. Further
I Department of Crop and Soil Sciences, University of Georgia, Athens, GA 30602, USA
2 AFESC/RDVW, Tyndall Air Force Base, FL 32403, USA
3 Department of Biology, Texas A&M University, College Station, TX 77843, USA

42 H.D. Wilde et al.
embryonic development was inhibited by 2,4-D at the 10-12 cell stage (Nomura
and Komamine 1985a, 1986). In the presence of 2,4-D, cells of these small
embryogenic clusters continue to divide to form larger, spherical structures with
multiple meristematic centers on the periphery (Halperin and Jensen 1967;
Ammirato 1987). The cell walls of the older cells in the center of this structure
break down with further culture, causing fragmentation into single cells and
smaller, morphologically polar clusters (Halperin and Jensen 1967; McWilliam
et al. 1974). The origin of Type 1 cells has not been elucidated, although a
continuous population of these cells may be maintained as a result of cluster
fragmentation. This is supported by the observation that fragmentation of carrot
zygotic embryos stimulates somatic embryo differentiation (Smith and
Krikorian 1989).
The formation ofthe embryogenic cluster, termed a proembryogenic mass
or PEM (Halperin and Jensen 1967), in the presence of auxin can complicate the
molecular analysis of somatic embryogenesis. Embryonic determination may
occur early during PEM development (Nomura and Komamine 1985a,b;
Komamine et al. 1992). The first unequal cell division apparently establishes
polarity in RNA synthesis (Nomura and Komamine 1985b), DNA synthesis
(Nomura and Komamine 1986), and Ca2+ distribution (Nomura 1987) in small
«12-cell) clusters. The establishment of cell division-related polarity is a
fundamental step in developmental events, such as organogenesis (Schnepf
1986). Polar organization within the small clusters is lost with continued
exposure to 2,4-D. The dual role of 2,4-D in carrot culture, inducing embryonic
determination and inhibiting embryo development, has been well established
(Sung et al. 1985; Ammirato 1987). This synthetic auxin disrupts electrical
patterns (Goldsworthy and Mina 1991), and possibly other tissue-organizing
networks (Racusen and Schiavone 1990), in cell cultures. Polar organization may
be reestablished with the removal of 2,4-D. For example, an increase in mem-
brane-bound calcium and activated calmodulin could be detected in defined
regions of the outer layer of cells after the transfer and dilution of PEMs into
auxin-free medium (Timmers et al. 1989).
Exposure of PEMs to auxin-free medium permits somatic embryo
development to take place. Somatic embryos are initiated from one or more cells
on the surface of the PEM (McWilliam et al. 1974; Haccius 1978; Jones 1979;
Fujimura and Komamine 1980). Somatic embryogenesis from carrot cell sus-
pensions can be synchronized by fractionation procedures which enrich for small
clusters of densely cytoplasmic cells (Lo Schiavo 1984). Small PEMs (48-120 j.1m
diameter) give rise to single embryos at a high frequency (Giuliano et al. 1983).
The remaining cell suspension consists of single cells and larger clusters of cells
that are of varying embryogenic potential (Halperin and JensenI967). Synchro-
nization of the development of cultured cells by size and density fractionation
provides an experimental system in which somatic embryo development can be
analyzed. The pattern of induction and development of embryos from somatic
carrot cells is not unique to suspension cultures. A similar sequence of events can
be observed during the induction of somatic embryogenesis directly from cells of
hypocotyls (De Vries et al. 1988; Stacey et al. 1990), cotyledons (Smith and
Krikorian 1989), and shoot tips' (Kiyosue et al. 1989).
Gene Expression in Somatic Embryos 43
3 Patterns of Gene Expression

During Carrot Somatic Embryogenesis
3.1 Expression of Dc3 in Suspension Culture
Polypeptides and mRNAs of a number of genes expressed by embryogenic carrot

cultures have been identified by 2-D PAGE analysis, nucleic acid hybridization,
and immunological techniques. The pattern of expression of these genes reveals
A 1.8
1.6
1 1.4
! 1.2
-«
~ 1.0
CI:
-« 0.8
~
-«
~
0.6
=- 0.4
~ H ~UO ~ P G
,
T L R
7500t -
,..
H C P G T L
..,.
,.. H C P G T L
- Actio
Dc3 SAc3
Fig. IA,B. Expression of Dc3 during carrot somatic embryogenesis. A Total RNA (3 ~g), prepared
from carrot culture fraction, somatic embryo stages, and vegetative tissues, was applied to
nitrocellulose using a slot-blot apparatus. Blots were hybridized with nick-translatedDc3, washed, and
autoradiographed. Lower panel represents 50-h exposure. Upper panel represents results of a
densitometric analysis of autoradiograph in lower panel. B Total RNA (7.5 fig) was size-fractionated,
transferred to nitrocellulose, and hybridized with probes for Dc3 (left panel) and actin, SAc3 (right
panel). Hybridization with the actin probe, SAc3, illustrates the developmental regulation of Dc3
expression. Rehybridization of all blots with an rDNA probe confinned equal RNA loading (not
shown). Co cotyledon; Ca callus; Hhypocotyl; J20> 120-fim fraction; 48 < 48-fim fraction; P PEM; G
globular stage; Ttorpedo stage; L leaf; R root. (Adapted from Wilde et al. 1988)
developmental regulation (transcriptional and translational) during embryo

induction and embryo development. For example, several lines of evidence
suggest that expression of the gene Dc3 correlates with PEM development in the
presence of 2,4-D. Steady state levels of Dc3 mRNA were found to be low in a
fraction of carrot cell suspension «48 Jlm diameter) containing single cells and
very small cell clusters (Wilde et al.1988). Dc3 transcripts were detected at levels
10- to 20-fold higher in a fraction containing small PEMs (48-120 J.Lm) and a
fraction> 120 J.Lm containing older cell clusters with multiple meristematic areas
(Fig 1A). When somatic embryogenesis was induced directly from hypocotyl
explants in a liquid culture, Dc3 expression was detected only after PEMs had
begun to form (De Vries et a1.1988). Dc3 was not expressed in seedling organs
and only at low levels in hypocotyl-derived callus (Fig. 1). No expression of Dc3
was detected in carrot culture variants asH (Vergara et al. 1982), C15 (Lo Schiavo
et al. 1983), and FGlO that cannot form PEMs (Wilde et al. 1988). The steady
state levels of other genes (Del 1, Del 6, Del 8) were also found to vary among the
fractions of crudely sized cell clusters «48, 48-120, > 120 Jlm) from suspension
cultures (Wilde 1988). These patterns of gene expression could result from cycles
ofPEM development taking place in cell suspensions maintained in the presence
of auxin.
3.2 Expression of Dc3 in Somatic Embryos and Transgenic Plants
Populations of staged somatic embryos were obtained by culturing a cell

suspension fraction enriched for small PEMs (48-120 Jlm) in auxin-free medium.
The 750 base pair transcript of Dc3 was present at high levels in both globular
and torpedo-stage embryos (Fig. IB). The steady state levels of the mRNA in
somatic embryos were within the range found in small PEMs (less than a 1.6-fo1d
increase). Dc3 was not normally expressed at detectable levels in vegetative
tissues from somatic embryo- or seed-derived carrot plants (Wilde 1988).
However, transcripts of Dc3 could be detected in reproductive tissues (carrot
seeds and flowers), and expression in vegetative tissue may be induced in response
to stress, for example, the low level expression of Dc3 in roots (Fig.lA) may be
the result of a stress response resulting in ABA production.
Osmotic stress was found to induce low levels of expression of a Dc3
promoter-reporter gene fusion in transgenic tobacco seedlings (Seffens et al.
1990). Transcriptional control sequences from a Dc3 genomic clone (DcG3) were
fused to the coding sequence of the bacterial enzyme p-glucuronidase (GUS), and
the chimeric gene was introduced into tobacco by Agrobacterium-mediated
transformation. A 1.5-kb 5' upstream region of Dc3 conferred rigorous embryo-
specific GUS expression in transgenic tobacco when oriented in the same
direction as the GUS transcription unit (Fig. 2). The Dc3 upstream regulatory
region contains DNA sequences similar, but not identical, to cis-acting, ABA-
responsive elements of wheat Em and rice RAB21 genes (Marcotte et al. 1989),
both late embryo-abundant (lea) genes (Galau et al. 1986).
In contrast to unstressed controls, 12 day-old transgenic tobacco seedlings
expressed Dc3-driven GUS when desiccated or exposed to 150 mM NaCl or
Fig. 2. Developmental accumulation of 10000,-------------,
GUS driven by DcG3 upstream sequence

elements. Developing seeds from transgenic
tobacco containing a 1.5-kb upstream region 1000
of Dc3 fused to the p-glucuronidase (GUS)

gene in pBllOI (Jefferson et al. 1987) were
assayed fluorometrically for GUS activity. 100
(Adapted from Seffens et al. 1990)
10
.1 +--_----,--~-...,...--'
10 20 30
Days after flowering (DAF)
10 11M ABA (Seffens et al. 1990). These treatments resulted in expression at levels
1, 0.4, and 0.6%, respectively, of those in developing transgenic tobacco seeds.
More recently, Vivekananda et al. (1992) demonstrated ABA and desiccation
induction of Dc3-driven GUS expression in transgenic tobacco seedlings and
mature leaves. In addition, ABA/desiccation response elements have been
uncoupled from seed-specification cis-regulatory elements in the promoter
proximal region of the Dc3 gene (Thomas et al. 1991). In seedling tissues, Dc3-
driven GUS expression may have been in response to osmotic stress-induced
ABA production, as is typical of some lea genes (Mundy and Chua 1988).
Alternatively, regulated GUS expression could have resulted from stress-induced
embryonic determination. For example, Kiyosue et al. (1989) observed somatic
embryo development on the surface of shoot tips of carrot seedlings exposed
to osmotic stress (l00 mM NaCI or 700 mM sucrose) or heavy metal ions
(0.25~ 1.0 mM CdCI 2 , CoCl 2, NiCI 2, or ZnCI 2). Furthermore, it has been demon-
strated that epidermal cells of tobacco leaves have the capacity to undergo
embryonic determination (Stolarz et al. 1991).
3.3 Lea Gene Expression During Somatic Embryogenesis
Sequence and DNA hybridization (Seffens et al. 1990) indicated that DcG3
belonged to a gene family whose members encode polypeptides that contain an
II-amino acid repetitive motif that is a distinguishing characteristic of LEA
proteins (Dure et al. 1989). A different lea gene (Dc8) has been isolated from
carrot somatic embryos that encodes mRNA whose steady state level increases
50- to 100-fold in somatic embryos exposed to 10 11M ABA (Hatzopoulos et al.
1990). Unlike Em, RAB21, and Dc3/GUS, the expression of Dc8 was not
inducible in nonembryonic tissues by ABA or desiccation. Franz et al. (1989)
reported that De8 was a single copy gene with alleles that differed slightly in
sequence and length. Like Dc3, De8 was found to be expressed in carrot seeds but
not roots or leaves (Borkird et al. 1988). The II-amino acid motif of Dc3 and DcB
was found in tandem arrays in LEA proteins (group 3) of cotton (Baker et al.
1988), barley (Hong et al. 1988, and rape (Harada et al. 1989). A proposed
function of LEA proteins, based on the structure of LEA peptide units and the
gene expression patterns of lea-class genes, is the protection of cellular structures
in mature embryos during seed desiccation (Dure et al. 1989).
Ulrich et al. (1990) isolated a eDNA (EMB-J) of an embryo-regulated
mRNA encoding a LEA protein of a different classification (group 1). A partial
amino acid sequence of an embryogenic cell protein ECP31 (Kiyosue et al. 1992),
indicated that it was 70-90% homologous to the predicted amino acid sequence
of the unclassified cotton lea eDNA, D34 (Baker et al. 1988), that is unusual in its
hydropathy. ECP31 was localized primarily on peripheral cells ofPEMs from
carrot suspension culture, and its expression correlated with embryogenic callus
growth from hypocotyl explants (Kiyosue et al. 1991). This protein was detected
immunologically in carrot seeds but not in nonembryogenic cultured cells or in
vegetative (unstressed) tissue. Like the Dc3-GUS gene fusion, expression of
ECP31 could be induced by osmotic stress in vegetative tissue (seedling shoot
tips) with embryogenic capacity (Kiyosue et al. 1990)
In contrast to the temporally restricted expression of lea genes during late
zygotic embryogenesis, Dc3 mRNA was expressed in early and late stages of
somatic embryo development (Wilde et al. 1988). DcB transcripts could also be
detected in all somatic embryo stages, although their expression appeared to be
associated primarily with the development of heart-stage embryos (Borkird et al.
1988). During somatic embryogenesis, ECP31 protein was detected
immunologically in the central region of globular embryos and was undetectable
in torpedo-stage embryos (Kiyosue et al. 1991), unless exposed to 3.71lM ABA
(Kiyosue et al. 1992). The protein product of Dc8 was observed in all somatic
embryo stages but not in nonembryonic tissues (Choi et al. 1987).
In zygotic carrot embryos, DcB protein was detected primarily within
vacuoles and protein bodies of torpedo-stage (30 days after flowering, DAF)
embryos (Franz et al. 1989). Expression ofECP31 began to appear in carrot fruit
at 28 DAF (Kiyosue et al. 1990). Dc3-driven GUS expression was maximal
during the maturation stage (22-25 DAF) of transgenic tobacco zygotic embryos
(Fig. 2). Dc3, DcB, and ECP31 expression could also be detected in cell clusters in
the presence of2,4-D (Borkird et al. 1988; Wildeet al. 1988; Kiyosue et al. 1991).
Together, these data indicate that the control of lea gene expression is either less
rigorous or more complex during somatic embryogenesis than during zygotic
embryogenesis.
3.4 Expression of Specific Embryo-Regulated Genes by PEMs
The detection in cell suspensions (+2,4-D) of mRNA and proteins of genes

expressed developmentally in somatic embryos is not uncommon. Stacey et al.
(1990) immunochemically identified a cell-surface arabinogalactan protein (J4e
epitope) during both PEM and somatic embryo development. Like Dc3,
expression of J4e correlated with PEM formation directly from cells ofhypocotyl
explants in auxin-containing medium. In large cell clusters (> 120 11M), J4e was
expressed in patches of peripheral cells. Within 24 h of transfer of PEMs into
auxin-free medium, an increase in the level of the epitope was detected. J4e was
localized in the outer surface layer cells of globular embryos and its expression
was gradually restricted during embryo development. In mature torpedo-stage
embryos, J4e was expressed primarily by cells forming two regions of the future
stele and by cells associated with the cotyledonary provascular tissue.
Transcripts of the gene that encode a lipid transfer protein (EP2) were
detected in PEMs by in situ hybridization and Northern blot analysis (Sterk et al.
1991). Expression of EP2 mRNA was not observed in nonembryogenic cell
clusters. In both somatic and zygotic carrot embryos, EP2 was expressed
specifically in cells of the protoderm, where it is presumed to be involved in
cutinization of the embryo surface. In addition to EP2, De Vries and coworkers
have identified other proteins that are present in the culture medium of
embryogenic carrot cell suspensions (reviewed in Van Engelen and De Vries
1992). Of these secreted proteins, a cationic peroxidase (Cordewener et al. 1991)
and acidic endochitinase (De Jong et al. 1992) can, respectively, restore globular
embryo development in carrot lines impaired biochemically (by tunicamycin) or
genetically (temperature-sensitive mutation) in somatic embryogenesis.
The mRNAs of several other embryo-regulated genes could be detected in
embryogenic cell cultures. Like Dc3, the steady state levels of De5, Dell, Del3,
and Del8 transcripts were approximately equal in small PEMs (48-120 11m) and
somatic embryos (Wilde 1988; Wilde et al. 1988). Aleith and Richter (1990)
identified a different set of genes (De2.l5, De4.2, and DeJa.l) that expressed
mRNAs in suspension culture ( +2,4-D) that temporarily increased in abundance
(6- to lO-fold) soon after auxin removal.
3.5 General Expression Patterns of Proteins and mRNAs in Carrot Cultures
Analyses of protein populations have demonstrated that there are relatively few
differences in gene expression between carrot cultures under conditions
permissive (-2,4-D) or repressive (+2,4-D) for embryo development (Sung and
Okimoto 1981; Choi and Sung 1984). It has been suggested that there are only a
small number of new genes that are expressed at detectable levels during somatic
embryo development (Choi and Sung 1984; Komamine et al. 1992).
Alternatively, developmentally regulated gene expression may be obscured by
comparisons of unstaged somatic embryos and unfractionated suspension
culture. For example, Racusen and Schiavone (1988) observed polypeptides
regulated temporally and spatially during somatic embryogenesis using staged
embryos and microsurgical techniques. Several of the stage- or tissue-specific
polypeptides (9 to 15) were expressed in auxin-maintained suspension culture
and, of these, three were not detected again until the torpedo-stage of embryo
development. Analysis of staged somatic embryos of temperature-sensitive
carrot variants at permissive and nonpermissive temperatures revealed
polypeptides with stage-specific expression patterns (Schnall et al. 1991).
The similarity of protein populations of carrot cultures maintained in the

presence and absence of 2,4-D is reflected at the mRNA level for moderately to
abundantly expressed genes (Fujimura and Komamine 1982). The mRNA
population expressed by small PEMs (48-120 ~M) can account for the
correspondence in expression patterns of suspension cultures and somatic
embryos detectable by 2-D PAGE analysis of in vitro translation products
(Wilde et al. 1988). The removal of 2,4-D from the medium in which isolated
PEMs ("state I") are cultured resulted in few changes « 1%) in this mRNA
population (Komamine et al. 1992). The similarity at the mRNA level suggests
that to some extent at least gene expression during somatic embryogenesis may
be regulated posttranscriptionally. Apuya and Zimmerman (1991) reported that
the control of expression of the genes encoding carrot homologues of an ATP
synthase subunit (ATP-2) and translational elongation factor (EF-l a) during
somatic embryogenesis is at the level of translation. ATP-2 and EF-la
transcripts were found at equal levels in total RNA of callus and somatic
embryos, but there was preferential loading of these mRNAs (two- to threefold)
onto polysomes in globular stage embryos. Komamine et al. (1992) described a
different accumulation pattern of an mRNA encoding a carrot EF-la.
homologue. However, the observation of embryo-regulated proteins (Borkird
et al. 1988; Racusen and Schiavone 1988; Stacey et al. 1990; Kiyosue et al. 1991)
in cell cultures (+ 2,4-D) indicates that posttranscriptional control is not a general
mechanism of gene regulation during somatic embryogenesis. This is supported
by experiments with transgenic plants demonstrating, for example, that the Dc3
promoter is sufficient to confer embryo-regulated expression to a bacterial
reporter gene encodingfi-glucuronidase (Seffens et al. 1990).
4 Conclusions
Although several developmentally regulated genes have been identified, the

molecular mechanisms underlying embryo development from somatic cells
remain enigmatic. In 1987, we published a model of somatic embryogenesis
(Thomas and Wilde 1987) that incorporated experimental results from our
laboratory and those of others (Choi and Sung 1984; Nomura and Komamine
1985b). This model differed from an earlier version proposed by Sung (1985) by
depicting Type 1 cells (12-~m cells) as "ground state" cells within suspension
cultures which require the presence of2,4-D to induce embryonic determination.
Recent results (Wildeet al. 1988; Timmerset al. 1989; Staceyet al. 1990; Kiyosue
et al. 1991; Sterk et al. 1991) allow this model to be refined (Fig 3).
One hypothesis for the observed gene expression pattern is that genes
regulated during embryo development are expressed in suspension culture during
PEM development from Type 1 cells. For example, Dc3, J4e, EP2, and ECP31
are genes that are developmentally regulated during somatic embryogenesis. Dc3
and EP2 are also detected in carrot cell cultures (+2,4-D), but only in cell clusters
larger than 48 ~m in diameter, suggesting that a few cell divisions are required
before it is expressed (Fig3A). Expression of Dc3, J4e, and ECP31 correlate with
+ 2,4-D - 2,4-D
A
SUSPENSION CULTURE SOMATIC EMBRYOS (A)
Diameter
0 Type 1
12"m
•
~
PEM
60 ~m
I
(Dc3,EP2) (J4e,ECP31,Dc3,EP2)
~
Enlarged
Clusters
(J4e,ECP31,Dc3)
t
>1000 pm FRAGMENTATION (J4e,ECP31,Dc3,EP2)
• PEM(liquid medium)
B ~(Dc3, J4e)
Hypocotyl 1+ 2,4-0
'------------' ~ e:::===- Embryogenic callus

(solid medium)
(ECP31)
Fig. 3A,8. Model of embryogenesis from somatic cells of carrot suspension cultures (A) and
hypocotyls (8). A The suspension culture is a dynamic and unsynchronized system. A crude size
fraction (48-120 flm; stippled bar), however, can give rise to single somatic embryos that develop
synchronously in the absence of2,4-D. The small PEMs that compose this fraction are derived from
Type I cells and from fragmentation of large clusters. Shaded areas indicate the expression of the
embryo-regulated genes shown in parentheses. 8 Embryonic determination in hypocotyl explants in
liquid or solid medium is associated with the expression of embryo-regulated genes. Two of these genes
(Dd and ECP3I) are lea genes that, in zygotic embryos, are restricted in expression to a late stage
(maturation) of embryogenesis
the acquisition of embryogenic potential in cells ofhypocotyl explants (Fig. 3B).

In auxin-treated cultures, J4e and ECP31 expression was localized in peripheral
cells ofPEMs, indicating that only the meristematic cells on the surface continue
to express these genes. Timmers et al. (1989) have suggested that the presence of
activated calmodulin in subset of the PEM surface cells may be correlated with
embryo development from these cells. The biological significance of these
observations is not clear, but they may explain the similarity in gene expression
detected in suspension cultures and somatic embryos.
In conclusion, the study of the regulation of genes normally expressed in

differentiated embryos (e.g., lea genes) during embryonic determination may
provide a means of analyzing the genetic mechanisms controlling somatic
embryogenesis. It should also be noted that rapid progress is now being made
in the application of molecular genetic approaches to study embryogenesis in
Arabidopsis thaliana (Meinke 1991). It is apparent that a combination of these
approaches should provide the range of tools required ultimately to define the
molecular and cellular rules governing plant embryogenesis.
Acknowledgments. The authors gratefully acknowledge the contributions of the somatic

embryogenesis community, especially the laboratories of S. de Vries, A. Komamine, and Z.R.
Sung. Work from the laboratory of T.L. Thomas was supported by grants from the USDA
Competitive Grants Program (84CRCRI1391 and 86CRCR12143) and the Texas Advanced
Research Program (10366038).
References
Aleith F, Richter G (1990) Gene expression during induction of somatic embryogenesis in carrot cell
suspensions. Planta 183: 17-24
Ammirato PV (1987) Organizational events during somatic organogenesis. In: Green CE, Somers DA,
Hackett WP, Biesboer DD (eds) Plant tissue and cell culture. Alan R Liss, New York, pp 57-82
Apuya N, Zimmerman JL (1991) Differential gene expression in carrot somatic embryos is
accomplished through translational control. In: Hallick RB (ed) Abstr 3rd Int Congr Plant Mol
BioI, Tucson, No 610
Backs-Hiisemann D, Reinert J (1970) Embryobildung durch isolierte Einzellen aus Gewebekulturen
von Daucus carota. Protoplasma 70: 49-60
Baker JC, Steele C, Dure L III (1988) Sequence and characterization of 6 Lea proteins and their genes
from cotton. Plant Mol BioI 11: 277-291
Borkird C, ChoiJ,JinZ, FranzG, Hatzopoulos P, Chomeau R, Bonas U, Pelegri F, SungZR (1988)
Developmental regulation of embryonic genes in plants. Proc Nat! Acad Sci USA 85: 6399-6403
Choi JH, Sung ZR (1984) Two-dimensional gel analysis of carrot somatic embryonic proteins. Plant
Mol BioI Rep 2: 19-25
Choi JH, Liu L-S, Borkird C, Sung ZR (1987) Cloning of genes developmentally regulated during plant
embryogenesis. Proc Nat! Acad Sci USA 84: 1906--1910
Cordewener J, Booij H, van der Zandt H, van Engelen F, van Kammen A, de Vries S (1991)
Tunicamycin-inhibited carrot somatic embryogenesis can be restored by secreted cationic
peroxidase isoenzymes. Planta 184: 478-486
De J ong A, Cordewener J, Lo Schiavo F, Terzi M, VandekerhoveJ, van Kammen A, de Vries SC (1992)
A carrot somatic embryo mutant is rescued by chitinasae. Plant Cell 4: 425-433
De Vries SC, Booij H, Meyerink P, Huisman G, Wilde HD, Thomas TL, van Kammen A (1988)
Acquisition of embryogenic potential in carrot suspension cultures. Planta 176: 196--204
Dure L III Crouch M, Harada J, Ho T-HD, Mundy J, Quatrano R, Thomas T, Sung ZR (1989)
Common amino acid domains among LEA proteins of higher plants. Plant Mol BioI 12: 475-486
embryonic gene, DeB, from Daucus carota. Mol Gen Genet 218: 143-151
Fujimura T, Komamine A (1980) The serial observation of embryogenesis in a carrot cell suspension
culture. New Phytol 86: 213-218
Fujimura T, Komamine A (1982) Molecular mechanisms of somatic embryogenesis in a synchronous
system in a carrot suspension culture. In: Fujiwara A (ed) Plant tissue culture 1982, Maruzen,
Tokyo, pp 105-106
Galau GA, Hughes DW, Dure L III (1986) Abscisic acid induction of cloned cotton late embryogenesis
abundant (Lea) mRNA. Plant Mol Bioi 7: 155~ 170
Giuliano G, Rosellini D, Terzi M (1983) A new method for the purification of different stages of carrot
embryoids. Plant Cell Rep 2: 216-218
Goldsworthy A, Mina MG (1991) Electrical patterns of tobacco cells in media containing indole-3-
acetic acid or 2,4-dichlorophenoxyacetic acid. Planta 183: 368~373
Haccius B (1978) Question of unicellular origin of non-zygotic embryos in callus cultures.
Phytomorphology 78: 74-81
cultures. J Ultrastruct Res 18: 428-433
Harada J, Delisle A, Baden C, Crouch M (1989) Unusual sequence of an abscisic acid-inducible mRNA
which accumulates late in Brassica napus development. Plant Mol Bioi 12: 395-401
Hatzopoulos P, Fong F, Sung ZR (1990) Abscisic acid regulation of De8, a carrot embryonic gene.
rIong B, Uknes SJ, Ho T-HD (1988) Cloning and characterization of a cDNA encoding a mRNA
rapidly induced by ABA in barley aleurone layers. Plant Mol Bioi II : 495~506
Jefferson RA, Kavanagh TA, Bevan MW (1987) GUS fusion: p-glucuronidase as a sensitive and
versatile gene marker in higher plants. EMBO J 6: 3901: 3907
Jones LH (1979) Factors influencing embryogenesis in carrot cultures. Ann Bot 38: 1077~1O82
Kiyosue T, Kamada H, Harada H (1989) Induction of somatic embryogenesis by salt stress in carrot.
Plant Tissue Cult Lett 6: 162~164
Kiyosue T, Dong JG, Satoh S, Kamada H, Harada H (1990) Detection of an embryogenic cell antigen
in carrot. Plant Cell Physiol 31: 947~950
Kiyosue T, Satoh S, Kamada H, Harada H (1991) Purification and immunohistochemical detection of
an embryogenic cell protein in carrot. Plant Physiol95: 1077~1083
Kiyosue T, Nakayama J, Satoh S, Isogai A Suzidi, Kamada H, Harada H (1992) Partial amino-acid
sequence of ECP31, a carrot embryogenic-cell protein, and enhancement of its accumulation by
abscisic acid in somatic embryos. Planta 186: 337~342
Komamine A, Kawahara R, Matsumoto M, Sunabori T, Toya T, Fujimura A, Tsukahara M, Smith
J, Ito M, Fukuda H, Nomura K, Fjimura T (1992) Mechanisms of somatic embryogenesis in cell
cultures: Physiology, biochemistry, and molecular biology. In Vitro Cell Dev Bioi 28P: 11~14
Lo Schiavo F (1984) A critical review of the procedures for embryo purification. Plant Mol Bioi Rep 2:
15~18
Lo Schiavo F, Giovinazzo G, Terzi M (1983) 8-Azaquanine resistant carrot cell mutants and their use
as universal hybridizers. Mol Gen Genet 192: 326-329
Marcotte WR, Russell SH, Quatrano RS (1989) Abscisic acid-responsive sequences from the Em gene
of wheat Plant Cell I: 969~976
McWilliam AA, Smith SM, Street HO (1974) The origin and development of embryoids in suspension
cultures of carrot. Ann Bot 38: 62~72
Meinke DW (1991) Perspectives on genetic analysis of plant embryogenesis. Plant Cell 3: 857~866
Mundy J, Chua NH (1988) Abscisic acid and water-stress induce the expression of a novel rice gene.
EMBO J 7: 2279~2286
Nomura K (1987) Mechanisms of somatic embryogenesis in carrot suspension cultures. PhD
Dissertation, Tohoku University
Nomura K, Komamine A (1985a) Identification and isolation of single cells that produce somatic
embryos at a high rate in carrot suspension cultures. Plant Physiol79: 988~991
Nomura K, Komamine A (1985b) Physiological and biochemical aspects of somatic embryogenesis
from single cells. In: Terzi M, Pitto L, SungZR (eds) Somatic Embryogenesis. IPRA, Rome, pp 1~5
Nomura K, Komamine A (1986) Polarized DNA synthesis and cell division in cell clusters during
somatic embryogenesis from single carrot cells. New Phytol 104: 23~32
Racusen RH, Schiavone FH (1988) Detection of spatially- and stage-specific proteins in extracts from
single embryos of the domesticated carrot. Development 103: 665-674
Racusen RH, Schiavone FM (1990) Positional cues and differential gene expression in somatic
embryos of higher plants. Cell Differ Dev 30: 159~ 169
Schnall JA, Hwang CH, Cooke TJ, Zimmerman JL (1991) An evaluation of gene expression during
somatic embryogenesis of two temperature-sensitive carrot variants unable to complete embryo
development. Physiol Plant 82: 498~504
52 H.D. Wilde et al.: Gene Expression in Somatic Embryos
Schnepf E (1986) Cellular polarity. Annu Rev Plant Physiol 37: 23-47
Seffens WS, Almoguera C, Von der Haar RA, Wilde HD, Thomas TL (1990) Molecular analysis of
phylogenetically conserved carrot gene: environmental and developmental regulated expression.
Dev Gen II: 65-76
Smith DL, Krikorian AD (1989) Release of somatic embryogenic potential from excised zygotic
embryos of carrot and maintenance of pro embryonic cultures in hormone-free medium. Am J Bot
76: 1832-1843
Stacey NJ, Roberts K, Knox JP (1990) Patterns of expression of the JIM4 arabinogalactan-protein
epitope in cell cultures and during somatic embryogenesis in Daucus carota L. Planta 180: 285- 292
Sterk P, Booij H, Schellekens GA, van Kammen A, de Vries SC (1991) Cell-specific expression of the
carrot EP2 lipid transfer protein gene. Plant Cell 3: 907-921
Stolarz A, Macewicz J, Liirz H (1991 Direct somatic embryogenesis and plant regenerati·on from leaf
explants of Nicotiana tabacum L. J Plant Physiol137: 347-357
Sung ZR (1985) Development states of embryogenic culture. In: Terzi M, Pitto L, Sung ZR (eds)
Somatic embryogenesis. IPRA, Rome, pp 117-121
USA 78: 3683-3687
Sung ZR, Feinberg A, Chorneu R, Borkird C, Furner I, Smith J, Terzi M, Lo Schiavo F, Guiliano G,
Pitto L, Nuti Ronchi V (1984) Development biology of embryogenesis. Plant Mol Bioi Rep 3:
3-14
Thomas TL, Wilde D (1987) Analysis of gene expression in carrot somatic embryo gene expression
programs. In: Green CE, Somers DA, Hackett WP, Biesboer DD (eds) Plant tissue and cell culture.
Alan R Liss, New York, pp 83-93
Thomas TL, Vivekananda J, Bogue MA (1991) ABA regulation of gene expression in embryos and
mature plants. In: Davies WJ (ed) Topics in environmental physiology. BIOS, London, pp 125-135
Timmers ACJ, de Vries SC, Schel JHN (1989) Distribution of membrane-bound calcium and activated
calmodulin during somatic embryogenesis of carrot (Daucus carota L.). Protoplasma 153: 24-29
Ulrich TV, Wurtele ES, Nikolau BJ (1990) Sequence of EMB-I, an mRNA accumulating specifically
in embryos of carrot. Nucl Acids Res 18: 2826
Van Engelen FA, De Vries SC (1992) Extracellular proteins in plant embryogenesis. Trends Genet 8:
66-70
Vergara MR, Biasini G, Lo Schiavo F, Terzi M (1982) Isolation and characterization of carrot cell
mutants resistant to a-amanitin. Z Pflanzenphysioll07: 313-319
Vivekananda J, Drew MC, Thomas TL (1992) Hormonal and environmental regulation of the carrot
lea class gene Dc3. Plant PhysiollOO: 576-581
Wilde HD (1988) Patterns of gene expression during carrot somatic embryogenesis. PhD Dissertation,
Texas A&M University, College Station, Texas
Wilde HD, Nelson WS, Booij H, de Vries SC, Thomas TL (1988) Gene-expression programs in
embryogenic and nonembryogenic carrot cultures. Planta 176: 205-211
1.5 Role of Polyamines in Somatic Emhryogenesis*
S.c. MINOCHA 1 and R. MINOCHA2
1 Introduction
The aliphatic amines putrescine, spermidine, and spermine are present in all
living organisms. Since the demonstration of "an essential nutritional function"
for putrescine in the bacterium Hemophilus parainjluenzae (Herbst and Snell
1948), polyamines have attracted a great deal of attention from workers in
diverse fields of the life sciences. The first reports of the existence of putrescine in
plants date back to 1911 (see Smith 1991 for a historical summary).
Several physiological functions of polyamines in plants have been reviewed
(Slocum et al. 1984; Smith 1985, 1990; Evans and Malmberg 1989; Slocum and
Flores 1991). While no hypothesis on the specific mode(s) of polyamine action
has been advanced, their positive charges at cellular pH and ability to bind
(interact) with several cellular macromolecules are obviously important for their
functions. It has been suggested that polyamine metabolism plays an important
role in growth, development, and stress responses in plants (Evans and
Malmberg 1989; Slocum and Flores 1991). Most arguments regarding their
putative functions are based upon two types of experimental observations: (1)
correlative changes in cellular polyamine levels during and prior to the growth
and developmental process in question; and (2) chemical inhibition of the
biosynthesis of polyamines and its effects on the growth and developmental
process. Among the developmental processes in which a major role for
polyamines has been proposed in cell and tissue cultures are cell division and
morphogenesis. The following discussion is focused primarily on the current
status of our knowledge on the metabolism of polyamines in relation to somatic
embryogenesis in plants.
* Scientific Contribution Number 1853 from the New Hampshire Agricultural Experiment Station
I Department of Plant Biology, University of New Hampshire, Durham, NH, 03824, USA
2 USDA Forest Service, NEFES, Concord/Mast Road, PO Box 640, Durham, NH, 03824, USA

Somatic Embryogenesis and Synthetic Seed I (ed. by y.P.S. Bajaj)
r9Springer-Veriag Berlin Heidelberg 1995
S4 S.c. Minocha and R. Minocha
2 Polyamine Metabolism
Biosynthesis of polyamines has been studied in a number of organisms (Pegg

1986; Smith 1990; Slocum and Flores 1991). In animals, the primary pathway for
putrescine biosynthesis involves decarboxylation of ornithine by ODC 3 (Fig. 1).
Plants and microorganisms utilize an additional pathway for putrescine
biosynthesis via the enzyme ADC which converts arginine into agmatine, which
is subsequently metabolized into putrescine in two steps. The distribution of the
two pathways seems to be tissue-, organ-, and species-specific.
Spermidine and spermine are synthesized by the sequential addition of an
aminopropyl moiety to putrescine, using the enzymes spermidine synthase and
spermine synthase, respectively. Decarboxylated SAM, which is the product of
SAMDC reaction, donates the aminopropyl groups. ADC, ODC, and SAMDC
are considered to be key regulatory enzymes for the biosynthesis of polyamines.
Each of these enzymes has a relatively short half-life (20-120 min) and its activity
generally shows a positive correlation with cellular levels of polyamines. The
catabolism of polyamines is regulated by diamine and polyamine oxidases
(Federico and Angelini 1991). Since polyamines can also serve as precursors for
a number of secondary metabolites, including nicotine, anabasine, and other
Ethylene
t
Aminocyclopropane
ARGININE ORNITHINE Carboxylic Acid (ACC)
ADCI i
S-Adenosylmethionine
Agmatine
!
(SAM)
N-Carbamoyl- -
putrescine
pur,·,n. j SAMDC
- - - Decarboxylated
spermidi~ S-Adenosylmethionine
1
Spermine
Fig. 1. Pathway for the biosynthesis of polyamines and ethylene in plants. The enzymes listed are:
ADC, ODC, and SAMDC
3Abbreviations: ACC, l-aminocyc1opropane-l-carboxylic acid; ADC, arginine decarboxylase;

AOA, aminooxyacetic acid; AVG, aminoethoxyvinylglycine; ODC, ornithine decarboxylase;
CHAP, cyc1ohexylammonium phosphate (same as DCHA); DFMA, a-difluoromethylarginine;
DFMO, a-difluoromethylornithine; MGBG, methylglyoxal bis (guanylhydrazone); SAM,
S-adenosylmethionine; SAMDC, S-adenosylmethionine decarboxylase.
Role ofPolyamines in Somatic Embryogenesis 55
alkaloids (Flores and Martin-Tanguy 1991), their cellular pools are subject to
modulation via biosynthesis, interconversion, degradation, and conversion
to secondary compounds. Total cellular polyamine levels are further subject to
changes in the soluble, bound, and conjugated forms of each polyamine (Smith
1985, 1990). In addition to the three common polyamines mentioned above,
some plants contain other polyamines, e.g., cadaverine, nor-spermidine, nor-
spermine, canavanine, etc.
Cellular levels of polyamines respond to a variety of external stimuli, e.g.,
physical stress, chemical stress, biological stress, plant growth regulators, etc. In
recent years, a number of specific inhibitors have been used to successfully
modulate the biosynthesis ofpolyamines (Pegg 1986; McCann et al. 1987). The
activities of ADC and ODC can be inhibited by the substrate analogs DFMA
and DFMO, respectively. Both are considered to be irreversible, or so-called
suicide inhibitors. While DFMA is almost universally effective in the inhibition
of ADC, DFMO does not inhibit ODC in all plants (Galston 1983; Flores and
Galston 1984; Slocum and Galston 1985; Robie and Minocha 1989). Other
inhibitors of ODC and ADC include methylornithine, monofluoromethylorni-
thine, D-arginine, and canavanine. Methylglyoxal bis (guanylhydrazone)
(MGBG) is a commonly used inhibitor of SAM DC, which causes a decrease in
the biosynthesis of both spermidine and spermine. MGBG also causes
stabilization of SAMDC in certain tissues, thus long-term use of MGBG may
actually result in elevated levels of these two polyamines (Malmberg and Hiatt
1989). A number of somewhat specific inhibitors of spermidine and spermine
synthases have also become available but only cyclohexylamine has been tested
for inhibition of spermidine in plants. Recent studies from our laboratory show
that MCHA (trans-4-methylcyclohexylamine) is an effective inhibitor of
spermidine synthase in carrot tissue, causing a reduction in cellular levels of
spermidine and an increase in putrescine. Spermine levels are either unaffected
or show a slight increase during the 96 h oftreatment (Andersen and Minocha,
unpubl.). APCHA [N-(3-aminopropyl) cyclohexylamine], which is an inhibitor
of spermine synthase in rat tissue (Shirahata et al. 1993), also shows similar
effects on carrot cells, i.e., it causes a reduction in cellular spermine and a
concomitant increase in spermidine. Putrescine is not affected by APCHA in
carrot cells.
Most enzymes involved in polyamine biosynthesis (Fig. 1) have been purified
and characterized in several microbial and animal systems (Tabor and Tabor
1984; Pegg 1986; Boyle et al. 1989). The genes for ADC, ODC, SAMDC, plus
spermidine and spermine synthases have been isolated, cloned, and characterized
from E. coli, yeast, and several mammalian tissues (Boyle et al. 1984, 1989;
Kontula et al. 1984; McConlogue et al. 1984; Glass et al. 1987; van Kranen et al.
1987; Pajunen et al. 1988; Kahana 1989; Tabor and Tabor 1989; Xie et al. 1989).
With the exception of preliminary reports on the cloning of ADC and SAMDC
genes from plants (Bell and Malmberg 1990; Taylor et al. 1992; Rastogi et al.
1993), not much is known about genes coding for other enzymes of the
polyamine biosynthetic pathway.
It is also obvious from Fig. 1 that the pathway for spermidine and spermine
biosynthesis shares a common precursor with the biosynthesis of ethylene in
56 S.C. Minocha and R. Minocha
plants. Whereas for polyamine biosynthesis SAM is irreversibly decarboxylated

by SAMDC, ACC synthase utilizes SAM for the production of ethylene. The fact
that many of the physiological functions of polyamines and ethylene are
antagonistic to each other has led to speculation that regulation of polyamine
biosynthesis may have a physiological function via its effects on ethylene
biosynthesis (Apelbaum et al. 1981, 1985; Even-Chen et al. 1982; Minocha 1988;
Flores et al. 1990). As described later in this chapter, several laboratories have
demonstrated that ethylene may be inhibitory to somatic embryogenesis in
carrot; therefore, it is conceivable that increased polyamine biosynthesis could
promote somatic embryogenesis via reduction in ethylene production (Minocha
1988; see also Nissen and Minocha 1993).
3 Polyamines and Somatic Embryogenesis
The roles of polyamines in the growth of shoot apex, root formation, and tuber
formation in plants have been summarized by Galston and Flores (1991). The
first report of a critical role of polyamines in somatic embryogenesis was given by
Montague et al. (1978). Since then, several studies have been aimed at the
changes in cellular levels of polyamines during somatic embryogenesis in plant
tissues.
Analysis of polyamines in several tissues of mango (Mangifera indica)
revealed no positive correlation with either the development of nucellar embryos
or the effect of 2,4-D (Litz and Schaffer 1987). Polyamine levels were higher in
nonembryogenic than embryogenic calli, and the addition of polyamines to the
medium had either no effect or suppressed the formation of nucellar callus in
most cultivars. Likewise, exogenous polyamines did not affect the development
of somatic embryos from nucellus or nucellar callus. It was concluded by. the
authors that polyamine levels in these tissues were correlated more with increased
growth rate than with morphogenesis.
Meijer and Simmonds (1988) demonstrated a sharp increase in putrescine
during the callus induction phase in petiole explants of two cultivars of clover
(Medicago sativa). This was followed by a decrease in putrescine upon transfer to
embryogenic media. DFMA and DFMO inhibited putrescine accumulation in
both cultivars, while embryo development was only inhibited in one cultivar. The
effect of inhibitors was not reversed by exogenous putrescine. In the same year,
Altman et al. (1988) reported a promotion of somatic embryo development by
spermidine and its inhibition by MGBG. Unfortunately, no data on cellular
polyamines were presented.
Fobert and Webb (1988) studied the effects of exogenous polyamines
and enzyme-activated inhibitors of putrescine biosynthesis on somatic
embryogenesis from cotyledons of eggplant (Solanum melongena). The addition
of arginine, ornithine, agmatine, putrescine, spermidine, or spermine generally
had no significant effect on somatic embryogenesis, except that high con-
centrations (10 mM or more) were inhibitory. DFMO was more potent than
DFMA in inhibiting somatic embryogenesis. Exogenous putrescine restored the

inhibition by DFMO. Explants growing on auxin-containing medium that
produced somatic embryos had higher levels of putrescine than those not
producing somatic embryos (minus auxin medium). DFMO inhibited cellular
putrescine as well as spermidine.
EI Hadrami and D'Auzac (1992) observed an increase in embryogenic
potential of callus in the rubber tree (Hevea brasiliensis) through the addition of
arginine, spermidine, or a combination of putrescine, spermidine, and spermine.
The maximum effect was seen with the combination of polyamines. DFMA,
DFMO, and MGBG, while having no effect on callus formation, inhibited
somatic embryogenesis. The inhibition was partially reversed by spermidine.
During the early period of growth (0-20 days), neither DFMO nor MGBG had
much effect on cellular polyamines. However, between 40 and 60 days of growth,
MGBG caused a decrease in all three polyamines, while DFMO caused a
decrease in spermidine and spermine but a small increase in putrescine. DFMA
caused a decrease in putrescine but an increase in spermidine and spermine.
Although no direct enzyme assays were performed, it was concluded that ADC
was probably the major pathway for putrescine biosynthesis in this tissue.
Proembryogenic tissues of Norway spruce (Picea abies) were used by
Santanen and Simola (1992) to study changes in polyamines during the
maturation of somatic embryos. The results showed a substantial decrease in
polyamines in tissue grown on the maturation medium. The results were further
confirmed by Minocha et al. (1993) in both Norway spruce and red spruce (Picea
rubens). In contrast to the tissue on maturation medium, that on the proliferation
medium always had two- to threefold higher levels of putrescine and spermidine.
Neither DFMO nor DFMA showed an appreciable effect on somatic embryo-
genesis. However, it should be pointed out that the development of somatic
embryos in conifers follows a very different pattern compared to that in
herbaceous angiosperms. In conifers, the proliferation of proembryogenic
masses involves relatively fast growth on a medium containing an auxin and a
cytokinin. The maturation of somatic embryos that is accompanied by slow
growth and senescence of a large part of the callus mass occurs in the presence of
abscisic acid (Tautorus et al. 1991). Thus, the decrease in polyamines in the tissue
grown on the maturation medium may reflect the physiological status of the
whole tissue rather than the developing/maturing somatic embryos. In contrast
to conifers, the development of somatic embryos in most angiosperms entails a
continued fast growth.
It is obvious from the above discussion that for most species only
preliminary data are available on the metabolisms of polyamines in relation to
somatic embryogenesis. No common pattern seems to emerge from these
reports. Sufficiently detailed studies have been conducted only with carrot cell
cultures; therefore, the remainder of this chapter is devoted primarily to this
tissue.
58 s.c. Minocha and R. Minocha
4 Polyamines in Carrot
In carrot cell cultures, either putrescine or spermidine is the predominant

polyamine depending upon the time of analysis; spermine being present only in
limited amounts based upon fresh weight. Trace amounts of cadaverine have also
been reported (Baker and Yon 1983). Furthermore, it is known that polyamines
in carrot cell cultures are largely present in the soluble form; conjugated
polyamines being generally less than 10% of the total. Arginine decarboxylase is
the major regulatory enzyme for putrescine biosynthesis in most carrot cell
cultures (Montague et al. 1979; Feirer et al. 1984; Robie and Minocha 1989; also
see Mengoli et al. 1989 for an exception). However, in green somatic embryos, as
well as in leaf and floral tissues of mature plants, high levels of ODC are present.
Since ODC is absent from both cell cultures and the root tissue of mature plants,
it is conceivable that ODC activity in carrot may be localized in the chloroplast.
Carrot ADC is sensitive to DFMA and canavanine (Feirer et al. 1984; Robie and
Minocha 1989; Mengoli et al. 1989). On the other hand, both the mature plant
ODC as well as the ODC in cell cultures (where present) are not inhibited by
DFMO (Mengoli et al. 1989; Robie and Minocha 1989).
The activity of SAMDC is correlated with the cellular levels of spermidine
and spermine (Roustan et al. 1989b; Minocha et al. 1991b). This enzyme is
inhibited by MGBG, which also causes a reduction in cellular spermidine and
spermine levels (Minocha et al. 1991a). Cyclohexylammonium phosphate
(CHAP), a potent inhibitor of spermidine synthase, causes a significant decrease
in spermidine and a concomitant increase in both putrescine and spermine (Khan
and Minocha 1991). A combination of CHAP + MGBG causes a synergistic
increase in putrescine. On the other hand, the increase in spermine with CHAP
treatment is completely counteracted by MGBG (Minocha and Khan 1991).
4.1 Polyamines and Somatic Embryogenesis in Carrot
In earlier studies on the metabolism of polyamines in carrot cell cultures,

Montague et al. (1978) reported that putrescine levels increased sharply when
cells were transferred to fresh media, regardless of the presence or absence of2,4-
D. However, embryogenic cultures maintained a higher level of putrescine than
nondifferentiating cells. The rate of biosynthesis of putrescine in the absence of
2,4-D was almost double that in its presence. While spermidine showed no
significant difference in response to treatment with auxin, spermine levels were
actually lower in cultures undergoing somatic embryogenesis. The amounts of
radioactive putrescine and spermidine synthesized from 14C-arginine remained
higher in the absence of auxin throughout the 72 h of study as compared to those
in the presence of auxin. During the same period, incorporation of a label into
spermine was lower in the differentiating cultures. However, in pulse-labeling
studies, the label disappeared from putrescine at a faster rate than from
spermidine, whereas spermine was quite stable during the 24-72 h of incubation.
Based on these results, it was concluded that the metabolism of polyamines was
different in the presence and absence of2,4-D.
In a later study by Montague et al. (1979), it was demonstrated that increases

in putrescine synthesis were correlated with increased cellular activity of ADC.
Within 6 h of incubation, ADC levels were higher in the absence of2,4-D vs +2,4-
D cultures, and this difference was sustained for at least 3 days. It was further
shown that increase in ADC activity was dependent upon the biosynthesis of
both the mRNA as well as the ADC protein. Following inhibition of protein
synthesis by cycloheximide, the half-life of ADC in carrot cell cultures was
calculated to be about 30 min. While SAMDC activity also increased sharply on
transfer of cells to fresh medium, there was no major difference in SAMDC
activity in the presence or absence of auxin.
The role of ADC in the biosynthesis of putrescine in cultured cells of carrot
was further documented by the studies of Feirer et al. (1984). Using specific
inhibitors of ADC and ODC, it was shown that ADC was the predominant
pathway for putrescine biosynthesis in these cells. Both putrescine biosynthesis
and somatic embryogenesis were inhibited by D FMA. Furthermore, the effect of
DFMA was reversible by the addition of exogenous polyamines. DFMO had
little effect on either cellular putrescine or somatic embryogenesis.
A parallel study by Fienberg et al. (1984) confirmed the importance of
polyamine biosynthesis during the differentiation of somatic embryos in carrot
cell cultures. They compared changes in the cellular content of soluble
polyamines in an embryogenic and a nonembryogenic cell line of carrot grown in
the presence or absence of2,4-D. It was seen that both the embryogenic cultures
grown in the presence of auxin and the nonembryogenic cultures grown in (+) or
(-) auxin media had similar profiles of changes in putrescine, spermidine and
spermine. This involved the following: (1) a sharp decrease in putrescine within
24 - 48 h followed by low levels of this diamine until 8-10 days, and then an
increase by 12-14 days; (2) little changes in spermidine levels during the first 8
days, followed by a slow increase during the next 4 - 6 days; and (3) only small
fluctuations in spermine during most of the culture period. In contrast, the
embryogenic cultures grown in the absence of auxin showed significantly higher
levels of putrescine during the period of 4 - 8 days, a sharp increase in spermidine
during the first 2 days, maintenance of higher levels of this polyamine through
most of the 14 day culture period, and a peak of spermine during the first 3 days
of culture. No significant differences in growth rates of the two lines were
observed under the two growth conditions. These results appeared to contradict
the observations of Montague et al. (1978), but the differences could be
attributed to differences in cell density and the method of somatic embryo
induction (see Fienberg et al. 1984 for further discussion). Whereas a sharp
decrease in putrescine was seen in embryogenic cultures grown either in plus or
minus auxin medium, a substantial increase in ADC activity was evident during
the same period; the latter treatment (minus auxin) sustained higher ADC
activity during the first 4 days as compared to the former treatment. The
apparent contradiction between putrescine levels and ADC activity, which is not
unique to carrot (similar observations have been made with Picea tissue; R.
Minocha and s.c. Minocha, unpubl.), may be attributed to a rapid metabolism
of putrescine in the absence of 2,4-D. The activity of ADC was always lower in
the nondifferentiating cell line regardless of the presence of2,4-D.
The activity of SAMDC was substantially higher in differentiating than in
nondifferentiating cultures. Both DFMA and DCHA inhibited cellular poly-
amines in a predictable manner and also inhibited somatic embryogenesis. The
effects of inhibitors were reversed by the addition of spermidine to the medium.
Exogenous spermine was generally inhibitory to somatic embryogenesis. DFMO
had no effect.
Bradley et al. (1984) published a brief report on the effects of exogenously
supplied arginine and putrescine on somatic embryogenesis in carrot. When cells
were grown for 40 days in the presence of 0.03 IlM putrescine in a medium
containing 2 mg/12,4-D and a PRL-4-D M amino acid mixture which included 40
mg/l L-arginine (Gamborg 1966), and then transferred to a medium lacking 2,4-
D and putrescine, but containing arginine, only globular stage embryos were seen
and no mature embryos were formed. If the later growth was allowed to occur in
the absence of arginine, about half of the globular embryos matured within 4
days. Control cultures that were not treated with putrescine produced normal
embryos within 3 weeks of transfer to 2,4-D-free medium. While this work was
done with the same cell line (WOOIC) used by Fienberg et al. (1984), the results
are, unfortunately, difficult to compare with most other published data for
several reasons: (1) the culture period was relatively long (i.e., 40-60 days); (2) no
rationales for the selection of medium supplements and concentration of
putrescine and arginine nor for the duration of treatment were provided; (3) no
information on scoring of somatic embryogenesis was given; and (4) cellular
polyamine levels were not measured.
In a study of the long-term effects of DFMO, Mengoli et al. (1987)
demonstrated that repeated culture of carrot cells in the presence of 5 mM
DFMO had no effect on growth rate, but caused an increase in cellular protein
content and polyamines and a decrease in ADC and ODC activities. In this
study, the control and DFMO-treated cells did not differ in their ability to form
somatic embryos (cf. Robie and Minocha 1989).
In a later study, Mengoli et al. (1989) reported that DFMO supplied in the
presence of 2,4-D prolonged the commitment of cells to embryogenesis on
transfer to an auxin-free medium. Canavanine, an inhibitor of ADC, inhibited
both the growth of cells and (consequently) their ability to form somatic
embryos. In contrast to the results of Montague et al. (1979), Feirer et al. (1984),
Robie and Minocha (1989), and Roustan et al. (1992), they observed a higher
ODC activity than ADC during the preembryogenic phase in their cells. They
also found DFMO to have no effect on ODC activity in vivo, while it inhibited
ODC activity in vitro (cf. Robie and Minocha 1989).
During the past several years, research in our laboratory has focused on the
effects of several inhibitors of polyamine biosynthetic enzymes on the
metabolism of polyamines and ethylene and the activities of ODC, ADC,
SAM DC, and SAM synthetase in carrot cell cultures. A series of papers have
been published (Robie 1987; Minocha 1988; Papa 1988; Robie and Minocha
1989; Minocha et al. 1990a,b, 1991a,b; Samuelsen 1990; Khan and Minocha
1991; Minocha and Khan 1991; Nissen and Minocha 1993) in which we have
shown that:
1. The only polyamines seen in carrot cells are putrescine, spermidine, and
spermme.
2. Whereas ADC is the primary enzyme for putrescine biosynthesis in un-
differentiated cell suspensions, ODC activity is relatively high in fully
developed somatic embryos and in leaves and flowers of mature plants.
3. Inhibition of ADC by DFMA significantly inhibits polyamine biosynthesis
and also inhibits somatic embryogenesis.
4. DFMO, which does not inhibit ODC (from mature plants), promotes ADC-
derived polyamine biosynthesis and also inhibits the production of ethylene.
5. DFMO promotes somatic embryogenesis in the absence of auxin and allows
some somatic embryogenesis even in the presence of auxin.
6. MGBG is a potent inhibitor of SAMDC and thus inhibits spermidine and
spermine biosynthesis while promoting the accumulation of putrescine as
well asACC.
7. MGBG inhibits somatic embryogenesis.
8. Cyclohexylammonium phosphate (CHAP), an inhibitor of spermidine bio-
synthesis, while promoting the accumulation of putrescine, has no effect on
cellular ACC or the production of ethylene from cells.
9. CHAP does not inhibit somatic embryogenesis but only retards the develop-
ment of somatic embryos.
10. The activity of SAM synthetase increases rapidly (up to 25-fold) within 5
days and is not significantly affected by any ofthe inhibitors except MGBG.
11. AVG strongly inhibits ACC and ethylene biosynthesis and promotes
spermidine and spermine biosynthesis.
12. At low concentrations (5-20 11M), AVG and AgN0 3 promote somatic
embryogenesis.
l3. Exogenously supplied spermidine and spermine (1-10 mM but not lower
concentrations) inhibit somatic embryogenesis and promote ACC and
ethylene production, probably through a feedback inhibition of their own
biosynthesis in the cells. These results are in general agreement with earlier
work, except for the DFMO effects on somatic embryogenesis.
In addition to the use of inhibitors, the modulation of cellular polyamines in

tobacco and carrot was achieved by the expression of mammalian genes for ODC
and SAMDC (De Scenzo and Minocha 1993; Noh and Minocha 1993; Bastola
1994). Transgenic carrot cell lines that overexpress a mouse ODC cDNA show
several-fold higher levels of putrescine, with little or no change in spermidine and
spermine. These cell lines are highly embryogenic in the absence of auxin and
show substantial somatic embryogenesis even in the presence of auxin in the
medium at concentrations that normally inhibit somatic embryogenesis.
Furthermore, these cell lines are highly tolerant to DFMA both for growth as
well as for somatic embryogenesis. Transgenic cell lines expressing a human
SAMDC cDNA show significantly higher levels of spermidine and also produce
somatic embryos at a high frequency (Bastola 1994).
4.2 Polyamines and Ethylene in Carrot Somatic Embryogenesis
Another aspect of somatic embryogenesis that is influenced by polyamine

metabolism is the role of ethylene. Like auxin, ethylene has been implicated as a
requirement for somatic embryogenesis and as having a strong inhibitory effect.
Tisserat and Murashige (1977) observed only a slight effect of exogenously
applied ethylene on somatic embryogenesis in carrot. Roustan et al. (1989a,b)
reported that inhibitors of ethylene biosynthesis (salicylic acid and acetylsalicylic
acid) and ethylene action (C0 2+ and Ni2+) had a strong stimulatory effect on
somatic embryogenesis in carrot. They emphasized that the effects of these
inhibitors were mediated through ethylene synthesis and action. Exogenously
supplied ethylene (by the addition of ethephon in the medium) caused a
substantial decrease in somatic embryogenesis in this tissue (Roustan et al. 1990).
In an unrelated study, Kiyosue et al. (1990) demonstrated that a number of heavy
metal ions including Coz+, NF+ and Cdz+were capable of inducing direct somatic
embryogenesis in apical tips of carrot seedlings without the formation of callus.
Keeping in mind that the polyamine and ethylene biosynthetic pathways
share a common precursor (Fig. 1), Robie and Minocha (1989) studied the
production of ethylene by carrot cell cultures under embryogenic and
nonembryogenic conditions. It was observed that significantly more ethylene was
generated by cells grown in the presence of2,4-D. Furthermore, the promotion of
polyamine biosynthesis by DFMO was accompanied by a reduction in ethylene
accumulation. Aminoethoxyvinylglycine (AVG) was found to stimulate somatic
embryogenesis. It was suggested that auxin-induced ethylene biosynthesis may
play an important role in the inhibition of somatic embryogenesis by 2,4-D
(Minocha 1988; Robie and Minocha 1989; Minocha et al. 1991a).
In line with this suggestion are the results of Roustan et al. (1992) who
studied the incorporation of 3,4-p4C]-methionine into polyamines and ethylene
during somatic embryogenesis in carrot. They showed that in control cultures
(-2,4-D), the incorporation of radioactive methionine into SAM and spermidine
was about eight to nine times higher than that recovered in ACC or ethylene.
Furthermore, incorporation into spermine was almost twice as high as in
spermidine. Treatments that promoted somatic embryogenesis, including cobalt
chloride and salicylic acid (Roustan et al. 1989a, b), stimulated the incorporation
of methionine into spermidine and spermine while reducing its incorporation
into ethylene. The amount of radioactivity present in ACC was also higher in the
presence of CoCl2 and salicylic acid (probably due to its reduced utilization in
ethylene synthesis). It should be pointed out that the release of 14COZ was not
affected under these conditions, nor was there any effect on the uptake of
methionine. Exogenously supplied ethylene (as ethephon) caused an inhibition of
somatic embryogenesis and a reduction in the incorporation of methionine into
spermidine and spermine. Activities of ADC as well as SAMDC were higher in
the presence of either CoCl2 or salicylic acid than in the control cultures.
Ethephon caused a reduction in the measurable activity of both these enzymes. It
was concluded that the reduction in polyamine biosynthesis could be due to the
effects of ethylene on ADC and SAMDC activities. Their results are in agreement
with the work of Even-Chen et al. (1982), who had observed an increased flux of
radioactive methionine into spermidine in orange peel tissue in response to

treatment with aminoethoxyvinylglycine (AVG). Biondi et al. (1990) have also
reported a similar effect of A VG on the incorporation of radioactive methionine
into spermidine in Prunus avium shoot cultures.
The interaction of auxin, ethylene, and polyamines has been discussed by
Nissen and Minocha (1993) in the light of some recent results from our
laboratory. However, ethylene has also been found to be suboptimal for somatic
embryogenesis in a different cell line of carrot where addition of low
concentrations of ACC or ethephon actually stimulated somatic embryogenesis
(Nissen 1993). For this cell line the inhibitors of ethylene biosynthesis actually
caused a slight inhibition rather than a promotion of somatic embryogenesis.
The observation that the carrot cell cultures can be suboptimal with respect to
ethylene under certain conditions has obvious implications for the mechanism of
action of 2,4-D in the inhibition of somatic embryogenesis.
The morphogenetic role of2,4-D, DFMO, and ethylene in somatic embryo-
genesis is complicated by the observations that: (1) inhibitors of ACC
biosynthesis (AOA and A VG), while promoting embryogenesis at lower concen-
trations, actually inhibit this process when used at higher concentrations (Nissen
1993; Minocha, unpubl.); (2) low levels of exogenously supplied ethylene are
actually stimulatory to embryogenesis under certain conditions (Nissen 1993);
(3) while some of the inhibitors of ethylene biosynthesis or action promote
somatic embryogenesis in the absence of auxin, none are able to reverse the
inhibitory effect of2,4-D; and (4) DFMO can restore embryogenesis even in the
presence of relatively high concentrations of exogenously applied ethylene in
the form of ethephon.
The situation becomes even more complex when one compares the studies of
Robie and Minocha (1989) and Roustan et al. (1992) with the data obtained by
Samuelsen (1990) who found that the inhibition of spermidine biosynthesis
by MGBG was accompanied by a decrease in ethylene biosynthesis and a
concomitant inhibition of somatic embryogenesis. Detailed analysis showed,
however, that ACC levels were significantly higher in MGBG-treated cells than
in controls. The inhibition of ethylene production in this case could be due to
some direct effects ofMGBG on ethylene-forming enzyme (ACe oxidase). It is,
therefore, quite possible that not only ethylene but also ACC itself could playa
morphogenetic role in carrot cell cultures. This is in agreement with the findings
that moderate to high levels of Aee (50--100 /J.M) inhibit somatic embryogenesis
in carrot (Verma and Tarka 1984; Minocha, unpubl.). Since there is currently no
information available on the effects of salicylic acid, C0 2+, Ag+, and exogenously
applied ethylene (ethephon) on the cellular levels of ACe, the role of ethylene vs
ACC cannot be easily resolved. Moreover, the data obtained from ethylene
measurement in the culture vessel do not necessarily represent the actual cellular
levels of ethylene. Being a gas, the cellular effects of ethylene must be exerted
largely at the time and site of production. Undoubtedly, the physiological
functions of ethylene in somatic embryogenesis and its competitive interaction
with polyamines deserve further analysis. Studies are currently underway to
accurately measure the rates of biosynthesis of ACC and ethylene in response to
treatments with the various inhibitors and promoters of somatic embryogenesis.
Any explanation for the role of polyamines in somatic embryogenesis is

further complicated by the following apparently contradictory observations:
While DFMO-promoted somatic embryogenesis is accompanied by increased
cellular polyamine levels and inhibition of their synthesis inhibits somatic
embryogenesis (Robie and Minocha 1989; Nissen and Minocha 1993), an
exogenous supply of either spermidine or spermine (1-: 10 mM) actually inhibits
somatic embryogenesis (Minocha, unpubl.). Since polyamines are rapidly taken
up by carrot cells (Pistocchi et al. 1988; Kanchanapoom et al. 1991), it can be
argued that the inhibitory effects of spermidine and spermine and the promo tory
effects of DFMO are mediated through cellular metabolism and are not due to
the mere presence of these compounds in the medium. Moreover, the fact that
DFMO effects cannot be mimicked by other analogs (e.g., methylornithine, D-
ornithine, L-ornithine, and putrescine) points to a specific physiological effect of
DFMO. Whether this effect is mediated through the DFMO effects on cellular
polyamine metabolism or through an independent mechanism is still not clear
and needs further investigation (see Nissen and Minocha 1993, 1994). The
arguments in favor of the former mechanism are: (1) DFMO causes a significant
increase in putrescine biosynthesis in the cells both in the presence and absence of
auxin; (2) this increase in putrescine biosynthesis is correlated with an increase in
ADC activity; and (3) the increased rate of putrescine biosynthesis by the
overexpression of a mouse ODC cDNA induces a behavior that mimics the
DFMO effects on the untransformed cells. On the other hand, an equally strong
argument can be made against the idea that DFMO effects are mediated largely
through polyamines. The observations that the inhibitory effects ofDFMA and
MGBG can be counteracted by an exogenous supply of putrescine and
spermidine, but the same polyamines cannot replace DFMO for promotion of
somatic embryogenesis, indicate that the need for polyamines during somatic
embryogenesis is independent of the stimulation caused by DFMO.
A third possibility that is presently being explored in our laboratory is that it
is not merely the presence of high(er) concentrations of putrescine and
spermidine that is sufficient for somatic embryogenesis, but their rapid turnover
may even be more critical for the development of somatic embryos. It is
envisioned that increased turnover of polyamines may somehow regulate
nitrogen balance in the cell and also affect ethylene biosynthesis through the
increased consumption of SAM. This hypothesis reconciles the contradiction
between the effects of increase in cellular polyamines via increased synthesis (by
DFMO treatment or transgene expression) and those due to an exogenous
supply in the medium. The latter could cause a decrease in polyamine turnover by
feedback inhibition and thus inhibit somatic embryogenesis.
A few generally accepted conclusions with respect to somatic embryogenesis in

carrot are:
1. Tissue explants must be treated with aUXlll to produce organized

proembryogenic cell masses.
2. On transfer to an auxin-free medium, these proembryogenic cell masses
undergo a programmed sequence of cell divisions and cell enlargement to
produce heart- and torpedo-stage embryos that continue to grow into small
plantlets.
3. The development of somatic embryos can be halted, and reversion to
the proembryogenic stage is seen if auxin is added back to the medium at any
time.
4. While a number of physical and chemical treatments can suppress the
development of somatic embryos in the auxin-free medium, until recently
nothing was known to overcome the inhibitory effect of auxin.
5. Biochemical and molecular analyses of cell clumps grown in the presence and
the absence of auxin reveal only minor differences, indicating that gross
changes in gene expression are probably not involved in early stages of
somatic embryogenesis.
While a few authors have speculated on the acquisition of embryogenic potential

by carrot cells and the role of auxin in this process (de Vries et al. 1988; Chasan
1993; Cooke et al. 1993; De Jonget al. 1993; and references therein), many more
hypotheses have been advanced to explain the role of auxin in the suppression of
somatic embryo development beyond the globular stage (Nissen and Minocha
1993 and references therein). Many of these suggestions are based upon
correlative changes in the physiology of cells and involve either the production of
some inhibitory compounds (e.g., ethanol, acetaldehyde, ethylene, etc.) or the
production of certain promotory factors. Some of these suggestions have
recently been experimentally tested in the light of the ability of DFMO to
counteract the auxin effects (Nissen and Minocha 1993).
A strong argument has been made for the need of an auxin gradient to
establish a bilateral symmetry (polarity) in the developing proembryo (Liu et al.
1993; see also comments in Cooke et al. 1993 and references therein). The
question as to whether this polarity is already present in the proembryogenic cell
masses growing in the presence of auxin, or if it is generated subsequent to the
transfer of these cell masses to auxin-free medium, still remains to be answered.
It has been suggested that the bilateral symmetry may involve (or depend upon)
the polar transport of auxin itself. However, we must be cautious in our
assumptions with respect to the nature of the auxin responsible for this polarity,
i.e., endogenously produced auxin vs the exogenously supplied auxin. There is
ample evidence to show that carrot cells growing in the presence of exogenously
supplied synthetic or natural auxin (i.e., 2,4-D or lAA) do produce endogenous
auxin (Micha1czuk et al. I 992a,b). Furthermore, it is quite apparent that the
polar transport of auxin is essential for the development of somatic embryos. It
is possible that the globular stage proembryogenic masses present in the 2,4-D-
containing medium already possess polarity with respect to the distribution and
polar transport of auxin and the exogenous auxin simply overwhelms the
endogenous gradient. Thus, the removal of auxin from the medium permits the
auxin gradient to influence further development. Alternatively, it is also
66 S.c. Minocha and R. Minocha
conceivable that the globular masses do not actually possess an auxin gradient
and it is only after removal of the exogenous auxin that a gradient is formed.
Having established that: (1) the presence ofDFMO in the medium promotes
polyamine biosynthesis and can counteract the auxin effect, and (2) the
overexpression of mouse ODC cDNA in carrot also has the same two effects; the
question that can be asked is: Does the increased production (see discussion
above for rate of synthesis vs cellular content) of polyamines have something to
do with the establishment or maintenance of the endogenous auxin gradient, or
are the two observations a consequence of different mechanisms for counter-
acting the auxin effects? In the first case, a possibility exists that DFMO has some
direct effects on auxin metabolism or transport that are unrelated to cellular
polyamine metabolism. The latter situation is, however, more difficult to
reconcile with the above explanation because the transgenic cells are grown under
the same set of conditions as the controls. The reversal of the effects of TIBA,
CPIB and other auxin transport inhibitors by DFMO (Nissen, pers. commun.)
suggests that DFMO is either able to protect the endogenous gradient of lAA
directly or the increased production of polyamines somehow helps to maintain or
(re)establish the auxin gradient.
A provocative suggestion that can be made to centralize the role of poly-
amines in in vitro morphogenesis is as follows:
Polyamines being positively charged at cellular pH could contribute to: (1) either
the establishment of polarity through their own unequal distribution in the
cell, thus causing the unequal distribution of some other negatively charged
macromolecule(s); or (2) the maintenance of an asymmetrical distribution of a
negatively charged (macro)molecule by neutralizing its charge. In either case
an asymmetrical distribution of polyamines would be required. At present,
however, nothing is known about the cellular compartmentation/distribution
of polyamines in plant cells.
Acknowledgments. The authors are thankful to Drs. Arthur Mathieson and Curtis Givan for useful
suggestions on the manuscript and to Nancy Jackson for word processing. Research discussed in this
chapter was partially supported by National Science Foundation Grant DCB-8615915.
References
Altman A, Levin N, Cohen P, Schneider M, Nadel B (1988) Polyamines in growth and differentiation
of plant cell cultures: the effect of nitrogen nutrition, salt stress and embryogenic media. In: Zappia
V, Pegg AE (eds) Progress in polyamine research. Plenum, New York, pp 559-571
Apelbaum A, Burgoon AC, Anderson JD, Lieberman M, Ben-Arie R, Mattoo A K (1981) Polyamines
inhibit biosynthesis of ethylene in higher plant tissue and fruit protoplasts. Plant Physiol 68:
453-456
Apelbaum A, Goldlust A, Icekson I (1985) Control by ethylene of arginine decarboxylase activity in
pea seedlings and its implication for hormonal regulation of plant growth. Plant Physiol 79:
635-640
Baker SR, Yon RJ (1983) Characterization of ornithine carbamoyltransferase from cultured carrot
cells oflow embryogenic potential. Phytochemistry 22: 2171-2174
Bastola DR (1994) Genetic engineering of the polyamine biosynthetic pathway and somatic
embryogenesis in carrot (Daucus carola L.). PhD Thesis, University of New Hampshire, Durham,
NH, 173 pp
Bell E, Malmberg RL (1990) Analysis of a cDNA encoding arginine decarboxylase from oat reveals
similarity to the Escherichia coli arginine decarboxylase and evidence of protein processing. Mol
Gen Genet 224: 431-436
Biondi S, Diaz T, Iglesias I, Gamberini G, Bagni N (1990) Polyamine and ethylene in relation to
adventitious root formation in Prunus avium shoot cultures. Physiol Plant 78: 474-483
Boyle SM, Markham GO, Hafner EW, Wright JM, Tabor H, Tabor CW (1984) Expression of the
cloned genes encoding the putrescine biosynthetic enzymes and ethionine adenosyltransferase
of E. coli (Spe A, Spe B, Spe C, met K). Gene 30: 129-136
Boyle SM, Wright JM, Satishchandran C, Buch 1 (1989) Regulation of the putrescine biosynthetic
genes (SpeA, SpeB, SpeC) in Escherchia coli. In: Bachrach U, Heimer YM (eds). The physiology of
polyamines II. CRC Press, Boca Raton, pp 73-84
Bradley PM, El-Fike F, Giles KL (1984) Polyamines and arginine affect somatic embryogenesis of
Daucus caroia. Plant Sci Lett 34: 397-401
Chasan R (1993) Embryogenesis: new molecular insights. Plant Cell 5: 597-599
Cooke TJ, Racusen RH, Cohen 10 (1993) The role of auxin in plant embryogenesis. Plant Cell 5:
1494-1495
De long Al, Schmidt EDL, De Vries SC (1993) Early events in higher plant embryogenesis. Plant Mol
Bioi 22: 367-377
De Scenzo RA, Minocha SC (1993) Modulation of cellular polyamines in tobacco by transfer and
expression of mouse ornithine decarboxylase cDNA. Plant Mol Bioi 22: 113-127
de Vries SC, Hilbert B, Peter M, Huisman G, Wilde HD, Thomas TL, van Kammen A (1988)
Acquisition of embryogenic potential in carrot cell-suspension cultures. Planta 176: 196-204
El Hadrami I, 0' Auzac J (1992) Effects of polyamine biosynthetic inhibitors on somatic embryogenesis
and cellular polyamines in Hevea brasiliensis. J Plant Physiol140: 33-36
Evans PT, Malmberg RL (1989) Do polyamines have roles in plant development? Annu Rev Plant
Physiol Plant Mol Bioi 40: 235-269
Even-Chen Z, Mattoo AK, Goren R (1982) Inhibition of ethylene biosynthesis by AVG and by
polyamines shunts label from 3,4-[14C] methionine into spermidine in aged orange peel slices. Plant
Physiol 69: 385-388
Federico R, Angelini R (1991) Polyamine catabolism in plants. In: Slocum RD, Flores H E (eds)
Biochemistry and physiology ofpolyamines in plants. CRC Press, Boca Raton, pp 41-56
Feirer RP, Mignon G, Litvay 10 (1984) Arginine decarboxylase and polyamines required for
embryogenesis in the wild carrot. Science 223: 1433-1435
Feirer RP, Wann SR, Einspahr OW (1985) The effects of spermidine synthesis inhibitors on in vitro
plant development. Plant Growth Regul3: 319-327
Fienberg AA, Choi IH, Lubich WP, Sung ZR (1984) Developmental regulation of polyamine
metabolism in growth and differentiation of carrot culture. Planta 162: 532-539
Flores H, Galston AW (1984) Osmotic stress-induced polyamine accumulation in cereal leaves. Plant
Physiol 75: 102-109
Flores HE, Arteca R N, Shannon lC (eds) (1990) Polyamines and ethylene: biochemistry, physiology,
and interactions. Am Soc Plant Physiol, Rockville, MD, pp 1-425
Flores HE, Martin-Tanguy 1 (1991) Polyamines and plant secondary metabolism. In: Slocum RD,
Flores HE (eds) Biochemistry and physiology of polyamines in plants. CRC Press, Boca Raton, pp
57-76
Fobert PR, Webb DT (1988) Effect ofpolyamines, polyamine precursors, and polyamine biosynthetic
inhibitors on somatic embryogenesis from eggplant (Solanum melongena) cotyledons. Can 1 Bot
66: 1734--1742
Galston AW (1983) Po1yamines as modulators of plant development. Bioscience 33: 382-388
Galston AW, Flores HE (1991) Polyamines and plant morphogenesis. In: Slocum RD, Flores
HE (eds). Biochemistry and physiology of polyamines in plants. CRC Press, Boca Raton, pp
175-186
68 S.c. Minocha and R. Minocha
Gamborg OL (1966) Aromatic metabolism in plants. II. Enzymes of the shikimate pathway in
suspension cultures of plant cells. Can J Biochem 44: 791-799
Glass JR, MacKrell M, Duffy JJ, Gerner E W (1987) Ornithine decarboxylase production in vitro by
using mouse eDNA. Biochem J 245: 127-132
Herbst EJ, Snell EE (1948) Putrescine as a growth factor for Hemophilus parainjiuenzae. J BioI Chern
176: 989-990
Kahana C (1989) Molecular genetics of mammalian ornithine decarboxylase. In: Bachrach U, Heimer
YM (eds) The physiology ofpolyamines l. CRC Press, Boca Raton, pp 281-295
Kanchanapoom M, Antognoni F, Pistocchi R, Bagni N (1991) Effect of auxins on spermidine uptake
into carrot protoplasts. Physiol Plant 82: 19-23
Khan AJ, Minocha SC (1991) Polyamines and somatic embryogenesis in carrot. II. The effects of
cyclohexylammonium phosphate. J Plant Physiol 137: 446-452
Kiyosue T, Takano K, Kamada H, Harada H (1990) Induction of somatic embryogenesis in carrot by
heavy metal ions. Can J Bot 68: 2301-2303
Kontula KK, Torkkeli TK, Bardin CW, Jiinne OA (1984) Androgen induction of ornithine
decarboxylase mRNA in mouse kidney as studied by complementary DNA. Proc Nat! Acad Sci
USA 81: 731-735
Litz RE, Schaffer B (1987) Polyamines in adventitious and somatic embryogenesis in mango
(Mangifera indica L.). J Plant Physiol128: 251-258
Liu C-M, Xu Z-H, Chua N-H (1993) Auxin polar transport is essential for the establishment of
bilateral symmetry during early plant embryogenesis. Plant Cell 5: 621-630
Malmberg RL, Hiatt AC (1989) Polyamines in plant mutants. In: Bachrach U, Heimer UM (eds) The
physiology ofpolyamines II. CRC Press, Boca Raton, pp 148-159
McCann PP, Pegg AE, Sjoerdsma A (1987) Inhibition of polyamine metabolism. Academic Press, New
York, 371 pp
McConlogue L, Gupta M, Wu L, Coffino P (1984) Molecular cloning and expression of the mouse
ornithine decarboxylase gene. Proc Nat! Acad Sci USA 81: 540-544
Meijer EGM, Simmonds J (1988) Polyamine levels in relation to growth and somatic embryogenesis in
tissue cultures of Medicago sativa L. J Exp Bot 39: 787-794
Mengoli M, Bagni N, Biondi S (1987) Effect of a-difiuoromethylornithine on carrot cell cultures.
J Plant Physiol 129: 479-485
Mengoli M, Bagni N, Luccarini G, Ronchi VN, Serafini-Fracassini D (1989). Daucus carota cell
cultures: polyamines and effect of polyamine biosynthesis inhibitors in the preembryogenic phase
of different embryo stages. J Plant Physiol 134: 389-394
Michalczuk L, Cooke TJ, Cohen JD (l992a) Auxin levels at different stages of carrot embryogenesis.
Phytochemistry 31: 1097-1103
Michalczuk L, Ribnicky DM, Cooke TJ, Cohen JD (1992b) Regulation of indole-3-acetic acid
biosynthetic pathways in carrot cell cultures. Plant Physiol 100: 1346-1353
Minocha R, Kvaalen H, Minocha SC, Long S (1993) Polyamines in embryogenic cultures of Norway
spruce (Picea abies) and red spruce (Picea rubens). Tree Physiol 13: 365-377
Minocha SC (1988) Relationship between polyamine and ethylene biosynthesis in plants - its
significance in morphogenesis in cell cultures. In: Zappia V, Pegg AE (eds) Progress in polyamine
research. Plenum, New York, pp 601-616
Minocha SC, Khan A J (1991) Effects of combinations of polyamine biosynthetic inhibitors on cellular
polyamines in carrot cell cultures. J Plant Physiol137: 507-510
Minocha SC, Minocha R, Robie CA (I 990a) A high-performance liquid chromatographic method for
the determination of dansyl-polyamines. J Chromatogr 511: 177-183
Minocha SC, Robie CA, Khan AJ, Papa N, Samuelsen Al (l990b) Polyamine and ethylene
biosynthesis in relation to somatic embryogenesis in carrot (Daucus carota) cell cultures. In: Flores
HE, Arteca RN, Shannon JC (eds) Polyamines and ethylene: biochemistry, physiology and
interactions. Am Soc Plant Physiol, Rockville, MD, pp 339-342
Minocha SC, Papa NS, Khan AJ, Samuelsen AI (199Ia) Polyamines and somatic embryogenesis in
carrot. III. Effects of methylglyoxal bis(guanylhydrazone). Plant Cell Physiol 32: 395-402
Minocha SC, Minocha R, Komamine A (1991 b) The effects of polyamine biosynthesis inhibitors on
S-adenosylmethionine synthetase and S-adenosylmethionine decarboxylase activities in carrot cell
cultures. Plant Physiol Biochem 29: 231-237
Montague MJ, Koppenbrink JW, Jaworski EG (1978) Polyamine metabolism in embryogenic cells of
Daucus carata. I. Changes in intracellular content and rates of synthesis. Plant Physiol 62:
430-433
Montague MJ, Armstrong TA, Jaworski EG (1979) Polyamine metabolism in embryogenic cells of
Daucus carata L. II. Changes in arginine decarboxylase activity. Plant Physiol63: 341-345
Nissen P (1993) Stimulation of somatic embryogenesis in carrot by ethylene. In: Pech JC, Latch<: A,
Balague C (eds) Cellular and molecular aspects of the plant hormone ethylene. Kluwer, Dordrecht,
pp 359-364
Nissen P, Minocha SC (1993) Inhibition by 2,4-D of somatic embryogenesis in carrot as explored by
its reversal by difluoromethylomithine. Physiol Plant 89: 673--{j80
Noh E W, Minocha SC (1994) Expression of a human S-adenosylmethonine decarboxylase cDNA in
transgenic tobacco and its effects on polyamine biosynthesis. Transgen Res 3: 26-35
Pajunen A, Crozat A, Janne OA, Ihalainen R, Laitinen PH, Stanley B, Madhubala R, Pegg AE (1988)
Structure and regulation of mammalian S-adenosylmethionine decarboxylase. J Bioi Chern 263:
17040-17049
Papa NS (1988) The effect of methylglyoxal bis(guanylhydrazone) on somatic embryogenesis and
polyamine levels in Daucus carata L. cell cultures. MS Thesis, University of New Hampshire,
Durham, NH, 86 pp
Pegg AE (1986) Recent advances in the biochemistry ofpolyamines in eukaryotes. Biochem J 234:
249-262
Pistocchi R, Bagni N, Creus JA (1987) Polyamine uptake in carrot cell cultures. Plant Physiol 84:
374-380
Pistocchi R, Keller F, Bagni N, Matile P (1988) Transport and subcellular localization ofpolyamines
in carrot protoplasts and vacuoles. Plant Physiol87: 514-518
Rastogi R, Dulson J, Rothstein SJ (1993) Cloning of tomato (Lycapersican esculentum Mill) arginine
decarboxylase gene and its expression during fruit ripening. Plant Physiol 103: 829-834
Robie CA (1987) The role of polyamines in the regulation of somatic embryogenesis in wild carrot
(Daucus carata L.). MS Thesis, University of New Hampshire, Durham, NH, 182 pp
Robie CA, Minocha SC (1989) Polyamines and somatic embryogenesis in carrot. I. The effects of
difluoromethylomithine and difluoromethylarginine. Plant Sci 65: 45-54
Roustan J-P, Latch<: A, Fallot J (1989a) Stimulation of Daucus carata somatic embryogenesis by
inhibitors of ethylene synthesis: cobalt and nickel. Plant Cell Rep 8: 182-185
Roustan J-P, Latche A, Fallot J (I 989b) Effet de l'acide salicylique et de I'acide acetylsalicylique sur la
production d'ethylene et I'embryogenese somatique de suspensions cellulaires de carotte (Daucus
carata L.). CR Acad Sci Paris 308: 395-399
Roustan J-P, Latch<: A, Fallot J (1990) Control of carrot somatic embryogenesis by AgN03, an
inhibitor of ethylene action: effect on arginine decarboxylase activity. Plant Sci 67: 89-95
Roustan J-P, Latche A, Fallot J (1992) Influence of ethylene on the incorporation of 3,4-[14C]
methionine into polyamines in Daucus carata cells during somatic embryogenesis. Plant Physiol
Biochem 30: 201-205
Sala F, Galli MG, Nielson E, Magnien E, Devreux M, Noy GP, Spadari S (1983) Synchronization of
nuclear DNA synthesis in cultured Daucus carata L. cells by aphidicolin. FEBS Lett 153: 204-208
Samuelsen AI (1990) Effects of inhibitors of polyamine and ethylene biosynthesis on ACC and ethylene
production in carrot (Daucus carota L.) cell suspension cultures. MS Thesis, University of New
Hampshire, Durham, NH 101 pp
Santanen A, Simola LK (1992) Changes in polyamine metabolism during somatic embryogenesis in
Picea abies. J Plant Physiol140: 475-480
Shirahata A, Takahashi N, Beppu T, Hosoda H, Samejima K (1993) Effect of inhibitors of spermidine
synthase and spermine synthase on polyamine synthesis in rat tissues. Biochem Pharmacol 45:
1897-1903
Slocum RD, Flores HE (eds) (1991) Biochemistry and physiology of polyamines in plants. CRC Press,
Boca Raton, 264 pp
Slocum RD, Galston A W (1985) In vivo inhibition of polyamine biosynthesis and growth in tobacco
ovary tissue. Plant Cell Physiol26: 1519-1526
Slocum RD, Kaur-Sawhney R, Galston AW (1984) The physiology and biochemistry ofpolyamines in
plants. Arch Biochem Biophys 235: 283-303
Smith TA (1985) Polyamines. Annu Rev Plant Physiol36: 117-143
Smith T A (1990) Plant polyamines - metabolism and function. In: Flores HE, Arteca RN, Shannon JC
70 S.c. Minocha and R. Minocha: Role ofPolyamines in Somatic Embryogenesis
(eds) Polyamines and ethylene: biochemistry, physiology and interactions. Am Soc Plant Physiol,
Rockville, MD, pp 1-17
Smith TA (1991) A historical perspective on research in plant polyamine biology. In: Slocum RD,
Flores HE (eds) Biochemistry and physiology of polyamines in plants, CRC Press, Boca Raton, pp
1-22
Sung ZR, Fienberg A, Chorneau R, Borkird C, Furner I, Smith J (1984) Developmental biology of
embryogenesis from carrot culture. Plant Mol Bioi Rep 2: 3-14
Tabor CW, Tabor H (1984) Polyamines. Annu Rev Biochem 53: 749-790
Tabor CW, Tabor H (1989) Microbial mutants deficient in polyamine synthesis. In: Bachrach U,
Heimer YM (eds) The physiology ofpolyamines II. CRC Press, Boca Raton, pp 63-72
Tautorus TE, Fowke LC, Dunstan DI (1991) Somatic embryogenesis in conifers. Can J Bot 69:
1873-1899
Taylor MA, Mad ArifSA, Kumar A, Davies HV, Scobie LA (1992) Expression and sequence analysis
of cDNAs induced during the early stages of tuberisation in different organs of the potato plant
(Solanum tuberosum L). Plant Mol Bioi 20: 641-651
Tiburcio AF, Campos JL, Figueras X, Besford RT (1993) Recent advances in the understanding of
polyamine functions during plant development. Plant Growth Regu112: 331-340
Tisserat B, Murashige T (1977) Effects of ethephon, ethylene, and 2,4-dichlorophenoxyacetic acid on
asexual embryogenesis in vitro. Plant Physiol 60: 437-439
van Kranen HJ, van de Zande L, van Kreije CF, Bisschop A, Wieringa B (1987) Cloning and
nucleotide sequence of rat ornithine decarboxylase cDNA. Gene 60: 145-155
Verma DC, Tarka T (1984) Influence of I-aminocyclopropanecarboxylic acid (ACC) and consequent
ethylene biosynthesis on growth and somatic embryogenesis in wild carrot (Daucus carota L.) cell
suspensions. In: Terzi M, Pitto L, Sung Z R (eds) Somatic embryogenesis. IPRA, Rome, pp
151-158
Xie Q-W, Tabor WC, Tabor H (1989) Spermidine biosynthesis in Escherichia coli: promoter and
termination regions of the SpeD operon. J Bacteriol171: 4457-4465
1.6 Anatomy of Somatic Embryogenesis
G. SCHUMANN l , U. RYSCHKA\ J. SCHULZE2, and E. Klocke l
1 Introduction
The regeneration of plants in tissue culture is generally possible through two

morphological and anatomic different pathways: (I) somatic embryogenesis and
(2) shoot/root organogenesis involving adventitious shoot bud development (see
Reinert et al. 1977).
Somatic embryogenesis is the process by which somatic cells develop into
differentiated plants through characteristic embryological stages without fusion
of gametes. Somatic embryos, which are morphologically and physiologically
similar to zygotic embryos, attain bipolarity early, and have a closed vascular
system formed in this process (Ammirato 1983; Williams and Maheswaran 1986;
Vasil 1987). The somatic embryos are also referred to as "embryoids".
Organogenesis leads to root formation or to the production of unipolar
adventitious buds, which develop into shoots that have to be rooted for plantlet
formation. The young shoot is connected via a procambial strand with a
preexisting vascular strand in the mother tissue. In contrast to a bud, the
embryoid does not develop a vascular connection to the mother plant and has a
closed radicular pole (Haccius 1978). The embryoid is like the zygotic embryo, a
bipolar structure with a root/shoot axis. Therefore, the most distinctive
characteristic of any embryoid is its anatomical discrete and closed vascular
system.
Since the first studies carried out with carrot (Steward 1958; Steward et al.
1958; Backs-Hiisemann and Reinert 1970), somatic embryoids have usually been
assumed to be of unicellular origin. Unlike plants regenerated via organogenesis,
it is believed that embryoids, like their zygotic counterparts, arise from single cells
either directly or after the formation of a mass of proembryogenic cells. This is
based on detailed histological studies (Botti and Vasil 1984; Lu and Vasil 1985;
Magnusson and Bornman 1985; Wang et al. 1990). Haccius (1978) described the
embryogenic cell complex as being formed from segmenting single cells that
produce embryoids through cleavage polyembryony, a sort of embryo budding.
Detailed studies of regenerates from embryogenic callus of Panicum (Hanna
et al. 1984) and Pennisetum (Swedlund and Vasil 1985; Rajasekaran et al. 1986)
1 BAZ, Institute of Vegetable, Medicinal and Spice Plant Breeding, Neuer Weg 22/23, D-06484
Quedlinburg, FRG
2 Institute of Botany, Technical University, Humboldstr, D-38106 Braunschweig, FRG

©Springer-Verlag Berlin Heidelberg 1995
72 G. Schumann et al.
showed the absence of chimeras in the F 1 generation. These results suggest a

single-cell origin and, thus, if mutation has occurred or exogenic genes have
been introduced, whole regenerates are genetically modified.
Nevertheless, other authors (Halperin and Wetherell 1965; Wernicke and
Brettelll982) have proposed that the embryoids are broadly based and hence of
multiple cell origin. Recent histological studies on different species have shown
the multicellular origin of somatic embryoids obtained by adventitious budding
(El Maataoui et al. 1990; Michaux-Ferriere et al. 1992). Nonetheless, the multi-
ple cell proembryogenic tissue may still have originated from a single cell
(Maddock 1985).
In this chapter early anatomical events of somatic embryogenesis in
microspores, immature embryos, and callus tissue are described and discussed in
the light of contemporary literature.
2 Embryoid Induction and First Cell Division
2.1 Microspores
The earliest indication of embryoid initiation is the appearance of very small,

round, elliptical bodies. For example, in cereals, direct pollen embryogenesis
from isolated microspores has only recently been reported, although there are
several studies reporting on globular structures as embryoids which never
develop into plants. Direct embryoid formation from microspores in cultured
anthers has been known for a longer time and was observed by Pan and Kao
(1978) as well as Henry et al. (1984). Initial divisions in the microspores have been
critically examined in the family Poaceae (Sunderland and Evans 1980; Kruger
1987; Schumann 1987a, 1990) and general routes to embryogenesis have been
distinguished. Pathway A is comparable with the normal development of pollen
grains. In this route the first division of microspores is unequal, giving rise to a
small generative cell and a large vegetative cell. The embryoids are formed by
repeated divisions of the vegetative cell. The generative cells remain quiescent
(one or two divisions are possible) and degenerate eventually. In pathway B the
microspore divides into two cells of equal size. The embryoid is formed by
divisions in one or both cells. In pathway C the first cell division is unequal, but
the generative cell also participates in embryogenesis. Irrespective of the nature
of the first pollen mitosis, further divisions in the absence of cell enlargement lead
to the formation of pollen grains, in which cells become progressively smaller
(Schumann 1990). Cell structures formed in this manner develop into pro-
embryoids when liberated from the exine.
2.2 Immature Zygotic Embryos
The scutellum of immature embryos, which has been extensively investigated

during the last decade in regard to somatic embryogenesis (Ozias-Akins and
Vasil 1982; Kott et al. 1985; Magnusson and Bornman 1985; Ryschka et al.
Anatomy of Somatic Embryogenesis 73
1991), is responsive only at a particular developmental stage (Thomas and Scott

1985; McCain and Hodges 1986; Kott et al. 1987). Immature embryos harvested
9-14 days post-anthesis are well differentiated and organized. The shoot-root
axis is developed and the procambium is visible as a strand of elongated cells
located centrally in the embryo axis ranging from the scutellar node to the upper
half of the scutellum (Fig. 1). Placing the embryo with the scutellum uppermost
on a callus induction medium resulted in the following development:
- precocious germination,
- formation of soft, watery, and translucent callus consisting of elongated cells,
- formation of compact, weak, yellowish callus consisting of small isodiametric
cells,
- formation of globular and embryo-like structures at the scutellar surface.
Since precocious germination of the original embryo and soft callus
formation have a negative effect on somatic embryogenesis, a number offactors
were investigated to optimize and enhance the frequency of somatic embryo-
genesis (Lu et al. 1983; Muller et al. 1987; Qureshi et al. 1989)
Histological studies of the early stages of callus formation in combination
with scanning electron microscopy and stereo microscopy on immature embryos
of wheat (Magnusson and Bornmann 1985) and barley (Ryschka et al. 1991)
gave evidence of different origins of somatic embryoids. Using optimized culture
conditions, after two days in culture, the periphery of the scutellum becomes
slightly enlarged (Fig. 2A,B), which is the macroscopic result of elongation of
cells from the abaxial scutellar epidermis. Single cells of the scutellar epidermis
start to divide periclinally and give rise to pseudothallus-Iike structures. The
parenchymatous cells of the scutellum and cells of the coleorhiza enlarge and
vacuoles coalesce, leading to separation of cell groups. Furthermore, in the area
of the hypocotyl and radicular periclinal cell divisions arising from the
procambium from a radial cell layer and 2 to 4 days later, first callus cells can be
observed. At the same time, division activity in the abaxial cells is intensified and
expands to the subepidermal cell layers resulting in the formation of distinct
Fig. 1. Longitudinal section of an immature embryo 9- 14 days after anthesis with a well-differentiated
procambium (arrow heads)
A B
Fig. 2A,B. Peripheral surface of the scutellum of an immature embryo two days after culture
initiation. ALongitudinal section; B Scanning electron micrograph of pseudothallus-like structures
Fig. 3. Longitudinal section of the scutellum of an immature embryo with a meristematic center
in the subepidermal cell layers (4 days after culture initiation)
meristematic areas without polarity (Fig.3), also described as meristematic units

(Fransz and Sche11991a) or pro embryonal cell complex (Trigiano et al. 1989),
and a swelling of the scutellar surface.
Five days after culture initiation callus proliferation originating from the
procambium of the hypocotyl increased and a meristematic ligament developed.
It is characterized by radial cell files initiated by periclinal divisions of pro-
cambium initial cells. The cells of this area contain more than one nucleolus in
one nucleus (Fig. 4) and many mitotic phases are visible. At the end of these
radial cell files the cells divide asymmetrically and form soft callus. Moreover, a
number of single cells of the procambium, branched into the scutellum, show
high division activity. Internal segmenting divisions in these single cells result in
the formation of discrete two- and four- celled proembryoids or proembryogenic
aggregates (Fig. 5). The latter consists of many small, isodiametric, cytoplasmic,
and starch-containing cells. These proembryoids and proembryogenic com-
plexes are separated from each other by loose, highly vacuolated, parenchy-
matous scutellar cells. Continuous periclinal and also anticlinal cell divisions
of the pseudothallus-like epidermal cells lead to two-, four- and eight-celled
proembryoids with a distinguishable bipolar structure (Fig. 6). Furthermore,
compact multicellular complexes are visible which presumably may be formed
from single, densely crowded proembryoids which coalesced (Fig. 7). These
proembryogenic complexes are recognizable externally as globular structures
on the surface of the scutellum. The formation of proembryoids and pro-
embryogenic complexes can also be initiated from single cells ofthe subepidermal
celllayers.
Fig. 4. Actively dividing cells in the area of the

procambium with 2 or 3 nucleoli (arrow heads) in
one nucleus
Fig. 5. A few-celled proembryoid arising

from procambium cells of the scutellum
To summarize three distinct origins of proembryoid formation can be

determined:
1. Single cells of the abaxial scutellar epidermis elongate, divide periclinally, and
give rise to pseudo thallus-like structures leading to bipolar proembryoids of
proembryogenic complexes.
Fig. 6. A few-celled proembryoid in the subepidermal layers 5 days after culture initiation
Fig. 7. Multicellular proembryogenic complex
2. Cells of the subepidermal cell layer with high division activity give rise to two-
and more celled proembryoids.
3. Cells of the scutellum's branched pro cambium initiate two-, four-, or more
celled proembryoids without any contact with the pro cambium from which
they originate.
The often found proembryogenic areas may be considered a result of
coalition of densely crowded proembryoids loosening their bipolarity, resulting
in the formation of compact embryogenic callus (Ryschka et al. 1991).
2.3 Callus Tissue
Induction of embryogenic callus has been reported in many cereal grass species
(Vasil and Vasil 1986). Embryogenic callus is predominantly produced by
immature and young tissues of inflorescences, embryos, leaves, and microspores
(Lu and Vasil 1981; Thomas and Scott 1985; Hoffmann et al. 1990).
Indirect plant regeneration after the production of callus from cells (e.g.,
protoplast, microspore culture) or tissues is the most common pathway. Detailed
morphological and histological studies (Schumann 1987b, 1990; Hoffmann et al.
1990) enable us to characterize three types of callus tissue with varying color,
structure, and morphogenetic response:
1. Compact globular, nodular, or convoluted callus, light yellow and very
dense, consisting of small isodiametric cells;
2. Irregular, loose callus, consisting ofloosely arranged cells, more semitrans-
lucent in appearance with dense regions, straw-colored
3. Irregular, extremely loose callus tissue of long, tubular, giant cells trans-
lucent and nonmorphogenic.
Fig. 8A-C. Longitudinal sections of a compact callus. A Presumptive procambium-Iike cell layer.
B Development of proembryoids by periclinal cell divisions (arrow heads). C Embryogenic mass
(arrow heads direction of cell divisions)
These three callus types are similar to those described by Heyser et al. (1985) and
Ozias-Akins and Vasil (1982).
The development of a presumptive procambium-like cell layer (Fig. 8A)
showed only the first callus type in a period coinciding with the phase of
proliferation. Internal segmenting divisions in these single cell layers gave rise to
a series of mainly periclinal divisions (Fig. 8B). Subsequently, anticlinal
divisions resulted in the formation of a pro embryonic mass (Fig. 8C). It appears
that the periclinal division found in this cell layer of the embryogenic callus may
Fig. 9. Longitudinal section of an irreg-

ular loose callus. Separated, darkly
stained, non vacuolated cells (arrow heads)
surrounded by vacuolated callus cells
mark the beginning of somatic embryogenesis. This developmental stage showed

in its morphology a striking analogy to that of the pathways demonstrated by
Magnusson and Bornmann (1985) in tissue of immature zygotic embryos of
Triticum aestivum. Maheswaran and Williams (1985) showed that the first sign of
direct somatic embryogenesis was the shift from regular anticlinal division to
irregular periclinal and oblique divisions. These histological studies suggested
that proembryoids arise de novo from single cells of the procambium-like cell
layer.
The second callus type in the phase of active callusing showed internally
less organized regions with single, darkly stained, nonvacuolated cells separated
from one another by groups of vacuolated cells (Fig. 9). Furthermore, the
development of organized regions, which contained cytoplasmically dense,
mitotically active cell groups with small vacuoles, was observed. The single cells
of the less organized regions divide to form groups of two cells and later, three to
four cells, but did not play any role in the formation of embryoids or shoot buds.
In contrast, cell division activity of more organized, compact, cytoplasmically
dense cell groups resulted in the formation of meristemoids, the origin of
adventitious shoot bud differentiation beneath the callus surface (Fig. 10). The
meristematic characteristics and the high division activity of these cells make
them analogous to embryonic cells observed during the development of the
zygote in the globular embryo (Jones and Rost 1989). It is evident that the
meristemoids arise de novo from proliferating callus tissue, but they are probably
multicellular in origin (D' Amato 1985) and may give rise to chimeras (Irvine
1984) which are unsuitable for clonal propagation, genetic analyses, breeding,
etc.
The de novo origin of shoot buds or embryo ids has also been proved by other
authors (Ho and Vasil 1983; Botti and Vasil 1984; Lu and Vasil 1985). According
to their morphogenic potency, these calli may be divided into ones producing
Fig.tO. Longitudinal section of an irregular, loose callus with shoot bud formation beneath the callus
surface
somatic embryos and those undergoing organogenesis to produce adventitious

shoots. The anatomy of the calli is quite distinct, but is difficult to distinguish
morphologically during the early stages of embryoid or shoot bud development.
These calli may occur alone or together in mixed cultures.
In an ultrastructural study on the early development in Zea mays Fransz and
Schel ( 1991 b) examined the development and structure of irregular, loose callus
(Type II). It was found that aggregates of cytoplasmic-rich, so-called embryo-
genic units, form the embryogenic potential of this callus type. They proposed the
presence of a transition structure between the embryogenic unit and the somatic
embryoid that marks the transition from unorganized to organized growth and
reflects the initiation of somatic embryogenesis (Fransz and Schel 1991a,b).
However, the results do not provide evidence for a unicellular origin of the
somatic embryoids.
The third callus type was composed of elongated tubular cells that were
occasionally as large as 500 /lm in length. They were loosely associated with each
other and non morphogenic (Schumann 1987b). This callus has also been referred
to as nonembryogenic (Heyser et al. 1985) and nonregenerative. Visually, the
nonmorphogenic callus sometimes appears to give rise to more compact callus
tissue. In the histological section the same kinds of compact areas were com-
monly found, but differentiation into leaf or shoot primordia never occurred.
While Nabors et al. (1983) suggested that nonmorphogenic callus can
become embryogenic, Heyser et al. (1985) stated that nonembryogenic cells are
Fig. 11. Scanning electron micrograph

of elongated embryoid structure with
lateral notch (arrow head)
Fig. 12. Scanning electron micrograph of two

embryoids showing the scutellum and the formation
of the coleoptile
not convertible to embryogenic cells. The formation of scattered sectors of

embryogenic callus in non morphogenic callus is more likely to be the result of
divisions into a few embryogenic cells which were maintained in a suppressed
stage (Morrish et al. 1987).
82 G . Schumann et al.
Fig. 13. Scanning electron micrograph of

the coleoptile with tenninal pore (arrow
head)
3 Maturation and Growth
Continued cell division and further organization of rounded, globular pro-

embryo ids formed elongated structures. A distinct macroscopic sign of further
Fig. 14. Longitudinal section of a young proembryoid

embryoid differentiation was the development of a lateral notch in the terminal

part of lobed, elongated structures (Fig. 11). Further lateral growth above the
formed notch gave rise to a hooded structure, resembling the classic stage in
the development of the scutellum (Fig. 12) in zygotic embryos. Precocious
proliferation in the area of the notch resulted in the development of a prominent
tubular structure, the coleoptile (Fig. 13) with its terminal pore (Vasil and Vasil
1982; Schumann 1987c).
Besides this embryoid development, the formation of abnormal structures is
a common phenomenon. Different embryonic structures develop without
passing through recognizable stages of zygotic embryogenesis. These structures
are characteristically compact, nodular, and creamy white to pale yellow. Often
the scutellum showed either an aberrant morphology or the embryoids were
without a scutellum. Such abnormal complex embryoid formation has also been
described in anther culture (Schumann 1987b, 1990; Schumann et al. 1991) and
was classified as an embryoid cell complex.
Michaux-Ferriere et al. (1992) consider that the formation of abnormal
embryoids, particularly joined twin or triplet embryos in Hevea brasiliensis, could
result from a polarization defect in embryogenic units of multicellular origin.
Histological examination of embryoids and also of embryoid cell complexes
in this period of differentiation showed that the ground tissue consists of small,
Fig. 15. Longitudinal section of an embryoid cell complex showing small, densely packed cytoplasmic
cells (se), medium-sized, highly cytoplasmic cells (me), and medium-sized vacuolated cells (mv)
thin-walled cells. In separated embryoids the polarity was established quite early
in their development by organization of the root meristem at one end and the
shoot meristem at the other end (Fig. 14). In contrast to separated embryoids,
embryoid cell complexes produce many meristematic regions in all parts of the
dissected tissue. Small cells with dense cytoplasm, a prominent nucleus, and no
vacuole; medium-sized highly cytoplasmic cells; and medium-sized vacuolated
cells, which may indicate disturbed development (Fig. 15), could be distin-
guished.
Although the bipolarity in embryoid cell complexes is not clearly evident,
anatomic integrity as the most distinctive characteristic of any embryo (Haccius
and Hausner 1975) is guaranteed. The loss of normal polarity might be
interpreted as an expression of aberrant embryoid development. One important
cause of this phenomenon might be the lack of synchronism of the developing
embryoids, especially when some embryoids form secondary embryoids.
Significant progress has been made in the past with histological and
morphological studies on regeneration of different species in many plant families
via somatic embryogenesis. More recently, special efforts were made in the
ultrastructural study on the induction up to the later globular stage.
In many reports the question of unicellular versus multicellular origin is
discussed. Both concepts are supported by histological evidence. Up to now it
appears that culture conditions influence the uni- or multicellular mode of
embryogenesis. This is supported by observations on callus of Elais
(Schwendiman et al. 1988) and Hevea (Michaux-Ferriere et al. 1992) in which the
two embryogenic modes were induced. Regardless of the origin of somatic
embryoids, the induction of embryoid development seems to be a major step in
somatic embryogenesis. However, available data are still insufficient.
References
Ammirato PV (1983) Embryogenesis. In: Evans DA, Sharp WR, Ammirato PV, Yamada Y (eds)
Handbook of plant cell culture, vol 1. Techniques for propagation and breeding. Macmillan, New
York, pp 82-123
Backs-Hiisemann D, Reinert J (1970) Embryoidbildung durch isolierte Einzelzellen aus
Gewebekulturen von Daucus carota. Protoplasma 70: 49-60
Botti C, Vasil IK (1984) Ontogeny of somatic embryos of Pennisetum americanum. II. In cultured
immature inflorescences. Can J Bot 62: 1629-1635
D'Amato F (1985) Cytogenetics of plant cell and tissue cultures and their regenerates. CRC Crit Rev
Plant Sci 3: 73-112
EI Maataoui M, Espagnac H, Michaux-Ferriere N (1990) Histology of callogenesis and somatic
embryogenesis induced in stem fragments of cork oak (Quercus suber) cultured in vitro. Ann Bot
66: 183-190
Fransz PF, Schel JHN (1991a) Cytodifferentiation during the development of friable embryogenic
callus of maize (Zea mays). Can J Bot 69: 26-33
Fransz PF, ScheIJHN (1991 b) An ultrastructural study on the early development of Zea mays somatic
embryos. Can J Bot 69: 858-865
Haccius B (1978) Question of unicellular origin of nonzygotic embryos of callus cultures.
Phytomorphology 24: 74--81
Haccius B, Hausner G (1975) Adventivknospen und nichtzygotische Embryonen - Grundlagen und
Anwendungen. Planta Med Suppl, pp 35-41
Halperin W, Wetherell DF (1965), Ontogeny of adventive embryos of wild carrot. Science 147:
756-758
Hanna WW, Lu C, Vasil IK (1984) Uniformity of plants regenerated from somatic embryos of
Panicum maximum Jacq. (Guinea grass). Theor Appl Genet 67: 155-159
Henry Y, de Buyser J, Guenegou T, Ory C (1984) Wheat microspore embryogenesis during in vitro
anther culture. Theor Appl Genet 67: 439-442
Heyser JW, Nabors MW, Mackinnon C, Dykes TA, de Mott KJ, Kautzman DC, Mujeeb-Kazi A
(1985) Long-term, high-frequency plant regeneration and the induction of somatic embryogenesis
in callus cultures of wheat (Triticum aestivum L.). Z Pflanzenziicht 94: 218- 233
Ho WJ, Vasil IK (1983) Somatic embryogenesis in sugarcane (Saccharum officinarum L.). I. The
morphology and physiology of callus formation and the ontogeny of somatic embryos.
Protoplasma 118: 169-180
Hoffmann B, Schumann G, Kriiger H-U (1990) Histological observations of morphogenesis
from androgenetic tissues of Triticum aestivum L. I. Callus tissue. Arch Ziichtungsforsch 20:
179-186
Irvine JE (1984) The frequency of marker changes in sugarcane plants regenerated from callus tissue.
Plant Cell Tissue Organ Cult 3: 201-209
Jones TJ, Rost TL (1989) Histochemistry and ultrastructure of rice (Oryza sativa) zygotic embryo-
genesis. Am J Bot 76: 504--520
Kott LS, Howarth MJ, Peterson RL, Kasha KJ (1985) Light and electron microscopy of callus
initiation from haploid barley embryos. Can J Bot 63: 1801-1805
Kott LS, Kott E, Howarth MJ, Kasha KJ (1987) A comparative study of initiation and development
of embryogenic callus from haploid embryos of several barley cultivars. I. Development stage of
embryo explants and callus potential. Can J Bot 65: 53-56
Kriiger H-U (1987) Zytologische Charakterisierung androgenetischer Entwicklungsphasen bei
Weizen (Triticum aestivum L.) Arch Ziichtungsforsch 17: 297-309
Lu C, Vasil IK (1981) Somatic embryogenesis and plant regeneration from freely-suspended cells and
cell groups of Panicum maximum Jacq. Ann Bot 48: 543-548
Lu CY, Vasil IK (1985) Histology of somatic embryogenesis in Panicum maximum (Guinea grass). Am
J Bot 72: 1908-1913
Lu C, Vasil V, Vasil IK (1983) Improved efficiency of somatic embryogenesis and plant regeneration
in tissue cultures of maize (Zea mays L.) Theor Appl Genet 66: 285-289
Maddock SE (1985) Cell culture, somatic embryogenesis and plant regeneration in wheat, barley, oats,
rye and triticale. In: Bright SWJ, Jones MGK (eds) Cereal tissue and cell culture. NijhofflW. Junk,
Dordrecht, pp 915-926
Magnusson I, Bornman CH (1985) Anatomical observations on somatic embryogenesis from scutellar
tissues of immature zygotic embryos of Triticum aestivum. Physiol Plant 63: 137-145
Maheswaran G, Williams EG (1985) Origin and development of somatic embryos formed directly on
immature embryos of Trifolium repens in vitro. Ann Bot 56: 619-630
McCain, JW, Hodges TK (1986) Anatomy of somatic embryos from maize embryo cultures. Bot Gaz
147: 453-460
Michaux-Ferriere N, Grout H, Carron MP (1992) Origin and ontogenesis of somatic embryos in
Hevea brasiliensis (Euphorbiaceae). Am J Bot 79: 174--180
Morrish F, Vasil V, Vasil IK (1987) Developmental morphogenesis and genetic manipulation in tissue
and cell cultures of the Gramineae. In: Scandalios JG, Caspari EW (eds) Advances in genetics, vol
25. Academic Press, New York, pp 431-499
Miiller B, Schulze J, Ryschka U, Ryschka S (1987) The effect of sorbitol on the induction of callus from
immature embryos of some cereals. In: Symp 100. Wiederkehr der Berufung von W. Pfeffer auf den
Leipziger Lehrstuhl fiir Botanik, Leipzig, 10.-12. Juni, Abstr, p 68
86 G. Schumann et al.: Anatomy of Somatic Embryogenesis
Nabors MW, Heyser JW, Dyles TA, de Mott KJ (1983) Long-duration, high-frequency plant
regeneration from cereal tissue cultures. Planta 157: 385-391
Ozias-Akins P, Vasil IK (1982) Plant regeneration from cultured immature embryos and inflorescences
of Triticum aestivum L. (wheat): evidence for somatic embryogenesis. Protoplasma 110: 95-105
Pan CL, Kao KH (1978) The production of wheat pollen embryo and the influence of some factors on
its frequency of induction. Acta Bot Sin 20: 122-128
Qureshi JA, Karatha KK, Abrams SR, Steinhauer L (1989) Modulation of somatic embryogenesis in
early and late-stage embryos of wheat (Triticum aestivum L.) under the influence of abscisic acid
and its analogs. Plant Cell Tissue Organ Cult 18: 55-69
Rajasekaran K, Schank SC, Vasil IK (1986) Characterisation of biomass production, cytology and
phenotypes of plants regenerated from embryogenic callus cultures of Pennisetum americanum x P.
purpureum (hybrid triploid napiergrass). Theor Appl Genet 73: 4-10
Reinert J, Bajaj YPS, Zbell B (1977) Aspects of organization - organogenesis, embryogenesis,
cytodifferentiation. In: Street HE (ed) Plant tissue and cell culture. Blackwell, Oxford, pp 389-427
Ryschka S, Ryschka U, Schulze J (1991) Anatomical studies on the development of somatic embryoids
in wheat and barley explants. Biochem Physiol Pflanz 187: 31-41
Schumann G (1987a) Zytologisch-histologische Untersuchungen in Triticale-Antherenkulturen. Arch
Ziichtungsforsch 17: 17-25
Schumann G (l987b) Zur Morphologie und Entwicklung androgenetischer Makrostrukturen in
Antherenkulturen von Triticale. Arch Ziichtungsforsch 17: 245-257
Schumann G (1987c) Morphologisch-anatomische und physiologische Untersuchungen zur
Antherenkultur von 6x-Triticale. Berlin, Akad Landwirtschafts-Wiss DDR, Ber Pflanzen-
produktionsforsch, Diss A 1987,123 pp
Schumann G (1990) In vitro production of haploids in Triticale. In: Bajaj YPS (ed) Biotechnology in
agriculture and forestry, vol 13. Wheat. Springer, Berlin Heidelberg New York, pp 382-402
Schumann G, Hoffmann B, Kruger H-U (1991) Histological observations on morphogenesis from
androgenetic tissues of Triticum aestivum L. II. Embryoids and embryo cell complexes. Arch
Ziichtungsforsch 21: 161-168
Schwendiman J, Pannetier C, Michaux-Ferriere N (1988) Histology of somatic embryogenesis from
leafexplants of the oil palm Elaeis guineensis. Ann Bot 62: 43-52
Steward FC (1958) Growth and development of cultivated cells. III. Interpretation of the growth from
free cell to carrot plant. Am J Bot 45: 709-713
Steward FC, Mapes MO, Mears K (1958) Growth and organized development of cultured cells. II.
Organization in cultures grown from freely suspended cells. Am J Bot 45: 705-708
Sunderland N, Evans LJ (1980) Multicellular pollen formation in cultured barley anthers. J Exp Bot
31: 501-514
Swedlund B, Vasil IK (1985) Cytogenetic characterization of embryogenic callus and regenerated
plants of Pennisetum americanum (L.) K. Schum. Theor Appl Genet 69: 575-581
Thomas MR, Scott KJ (1985) Plant regeneration by somatic embryogenesis from callus initiated from
immature embryos and immature inflorescences of Hordeum vulgare. J Plant Physiol121: 159-170
Trigiano RN, Gray DJ, Conger BV, McDaniel JK (1989) Origin of direct somatic embryos from
cultured leaf segments of Dactylis glomerata. Bot Gaz 150: 72-77
Vasil IK (1987) Developing cell and tissue culture systems for the improvement of cereal and grass
crops. J Plant Physiol128: 193-218
Vasil V, Vasil IK (1982) The ontogeny of somatic embryos of Pennisetum americanum (L.) K. Schum
I. In cultured immature embryos. Bot Gaz 143: 454-465
Vasil V, Vasil IK (1986) Plant regeneration from friable embryogenic callus and cell suspension
cultures of Zea mays. J Plant Physiol124: 399-408
Wang L, Huang B, He M, Hao S (1990) Somatic embryogenesis and its hormonal regulation in tissue
cultures of Freesia refracta. Ann Bot 65: 217-276
Wernicke W, Brettell R (1982) Morphogenesis from Sorghum bic%r leaves. Nature 287:138-139
behaviour of cells as an embryogenic group. Ann Bot 57: 443-463
I.7 Machine Vision Analysis of Plant Cells
and Somatic Embryos
M.A.L. SMITH l
1 Overview
Machine vision (video image analysis) is a strategy for collecting visual

information (with a camera) from a specimen, and allowing the visible
(appearance) characteristics of the specimen to be rigorously, numerically inter-
preted with the aid of computer. The image, stored in the memory of a computer,
provides an extremely detailed representation of a specimen. Consequently, the
image, rather than the live specimen, can be subjected to rigorous examination,
measurement, and manipulation. For these reasons, image analysis advances a
very useful tactic for analyzing characteristics of a biological/in vitro specimen in
detail, without invading the environment of the actual specimen itself, except for
the brief moment of image capture (acquisition of an image with a camera).
All in vitro experimental measurements usually require discrimination of
subtle, difficult-to-quantify changes in culture appearance, growth, or other
aspects of quality. Evaluations of the somatic embryogenesis (SE) process also
require identification of stages in embryo development and maturation, which
demands characterization of multiple qualitative parameters in solid culture or
liquid suspensions. Image analysis, with its capacity for shape recognition, and
ability to discern small growth, color, or textural changes, is presented as a
reliable means to collect exacting, dynamic information applicable to control
many in vitro experimental and production systems. At the whole plant or organ
level in vitro, machine vision has been adapted for micropropagated shoot
culture in order to assess growth characteristics and tissue quality (Smith et al.
1989; McClelland and Smith 1990), and comparatively evaluate in vitro versus ex
vitro root systems in terms of both morphological character and anatomical
development (McClelland et al. 1990; Smith et al. 1990; Rogers and Smith 1992).
Because the technology has an extraordinary capacity for capturing and
quantifying multiple, small details in a culture, it has been a valuable tool for
determining the developmental patterns of in vitro regenerated plantlets (Iriondo
and Perez 1991), or assessing the stress responses at the whole plant level (Spomer
and Smith 1988; Smith et al. 1991, 1993; Shibli et al. 1992). Image analysis has
been particularly advantageous in applications to cell culture and somatic
I Department of Horticulture, University of Illinois, Urbana, IL 61801, USA

88 M.A.L. Smith
embryo systems, and has resulted in superior operator control over data
collection without intruding on the tissue culture growing environment. In this
chapter, basic requirements for video image analysis are described, applications
of image analysis to measurement of cell and somatic embryo culture are
surveyed, and current strategies for using machine vision to automate and
control the somatic embryo bioprocess are presented.
2 Challenge of Cell Culture Analysis
Cell, and somatic embryo cultures are evolving as excellent tools in both scientific
research and plant production. Each of these systems, however, are characterized
by common, inherent obstacles which inhibit accurate measurement and
evaluation. In vitro culture growth and development occur on a small scale. For
this reason, significant changes in overall culture biomass (growth) attributed to
factors such as novel growth regulators or temperature regimes are not easily
quantified. Treatment effects take place within a sterile microenvironment.
Measurements which intrude on the microenvironment (for example, fresh
weight analysis or packed cell volume) will contaminate a culture, or at least
disturb the delicate balance of chemical and physical factors in the medium and
culture headspace. These typical contraints to direct measurement are familiar to
most in vitro culture systems. When a culture's development must be terminated
in order to extract a direct measurement, the opportunity to evaluate
experimental factors over time is lost. Quality changes (for example, degree of
callus culture friability or incidence of culture vitrification) may not even be
marked by measurable differences in fresh weight or other measurement
parameters (Smith and Spomer 1987).
Somatic embryos, which undergo a distinct set of stages during develop-
mental ontogeny, have additional, intrinsic measurement challenges that need to
be addressed. Once conditions for inducing SE are met, the bipolar structure
progresses through globular, torpedo, and cotyledonary stages (dicots) or
globular, scutellar, and colyoptilar stages (monocots). A goal in SE research is to
advance efficiently through each stage towards maturation, to achieve a high
degree of culture synchronization (assurance that most developing embryos are
at the same stage of ontogeny), and ultimately accomplish conversion of the
somatic embryos.
However, heterogeneity in terms of SE quality and degree of maturity is
typical in most embryogenic cultures. Production of high quality somatic-
embryos is always a multistage process, with different chemical and physical
requirements, depending on the stage of development. SE quality, development
stage, and conversion rate can clearly be manipulated by altering the
composition of nutritive media or changing other culture factors (Kamada and
Harada 1979; Senaratna 1992). Conversion frequency, for example, can be
dramatically improved by adjusting nutrient and growth regulator supply and
composition during the production cycle. Physical micro environmental
Machine Vision Analysis of Plant Cells and Somatic Embryos 89
manipulation, such as a high relative humidity treatment following embryo

maturation, can similarly improve conversion frequency dramatically (Roberts
et al. 1990). In order to assess these important influences on SE, culture samples
must usually be examined (microscopically) to visually determine the stage of
development, and the degree of culture synchronization. These observations are
by nature subjective, time-intensive, and tedious, yet absolutely essential in light
of the inherent variability in cultures. The variability in the culture forces viable
embryos to separate from immature or atypical units prior to ex vitro handling.
Because scaled-up production of somatic embryos most efficiently takes place
in liquid suspension cultures or bioreactors, measurement difficulties are in-
tensified, since the culture is flowing a three-dimensional suspension.
When the SE procedure is carried one step further through synthetic seed
production, the stages in the process include callus bulk -up, multiple maturation
stages in embryo development, desiccation, germination, and container planting.
Evaluation and measurement challenges, especially those featuring irregular,
heterogeneous objects on the microscopic scale, complicate each of these
production phases. Unless treatment effects can be rapidly, accurately detected,
progress towards system improvement tends to be slow and inefficient. In fact,
SE is well recognized as a phenomenon which will require a great deal of further
intensive research before optimum control of the process can be achieved.
3 Advantages of Machine Vision
Video image analysis, or machine vision, is a measurement technology that

extracts quantitative data from a visual image of a cell or SE culture, and thus
permits objective interpretation at any stage in a oell culture production cycle.
The technique, which is conceptually straightforward and simple, evaluates the
photometric (spectral) and morphometric (spatial) visual qualities of a specimen
in an image. It is of particular importance for in vitro evaluations that image
analysis is capable of detecting small changes in growth or development with
greater precision than human observation, and can be used to automatically
classify objects. In many cases, the measurements can be accomplished through
the walls of a culture vessel without intruding on the microenvironment.
Machine vision measurements applied to in vitro cultures have proven to be
extremely sensitive (Smith and Spomer 1987), can override many of the problems
inherent to microculture evaluations, and provide a level of information that
previously had never been available. Image analysis can replace the repetitive
microscope evaluations which are otherwise necessary to assess culture status,
and eliminates the possibility that different technicians may insert a bias into their
classifications. The same methods can be adapted for macroscopic culture
Images.
Basic components for image analysis of cell and embryo cultures include:
(1) staging devices for consistent presentation of culture samples; (2) an image
capture device (camera); (3) a digitizing board housed in a microcomputer for
90 M.A.L. Smith
storage of image data and measurements; and (4) a display/output device for
measurement data and images (Fig. 1). To collect information, a sample is placed
in the staging device and viewed in situ with a camera. The image area is
calibrated to preserve exacting spatial relationships. A two-dimensional ruled
template can be placed in the field of view for a macroscopic image, or a
hemocytometer with known dimensions can be used for microscopic images. The
image analysis technique captures a visual image of a specimen and maps out the
visual information into a numeric format (digitized). During spatial calibration,
the precise area of each image point (pixel) that comprises the image is
determined. For example, for many standard systems, the image will be
comprised 480 x 512 pixels, each with a unique number assignment for location
on the screen image. Calibration of spectral properties can be achieved by
including a gray-level chart or color chart in an initial image to standardize
brightness levels. Each pixel in the image also has a unique number corre-
sponding to the gray levellbrightness (for a black and white camera) or color (in
a color system) assigned to that point in the image. In effect, a digitized,
numerically defined image in machine vision will "look" identical to the counter-
part video image. As an added advantage, all of the visible (vision-detectable)
features of this image can be recognized and numerically identified by a machine
vision system, even when they are not easily discerned by the human eye.
For cell cultures, the staging device may be an inverted microscope
standardized in terms of focus and light for display of a sample. Later stages of
cell or callus growth on a culture plate may be suited for macroscopic staging in
a uniform, consistent light environment. A microscopic image of a cell culture
will magnify details of objects so that they can be easily classified by size, shape,
and spectral characteristics. However, sample size is quite small and inexact.
Macroscopic staging of cultures allows the investigator to view larger sample
sizes and directly calculate the volume of the sample to project the calculation
back to the entire culture broth. Macroscopic staging is particularly suited for
viewing solid plate cultures and growing callus masses as they develop over time.
Charge-coupled device (CCD) video cameras are widely used as imaging
devices for machine vision applications, including culture analysis (Fig. 1). An
I IMAGE CAPTURE DEVICEI----I.~ IDIGITIZER AND MICROCOMPUTER I
H
ISTAGINGI
DISPLAY OUTPUT
macro or microscopic
uniform lighting
for sample display or presentation
Fig. 1. Basic components and options for an image analysis system for cell and somatic embryo
measurements
image digitizer can analyze a photographic image on film, but direct, live capture
of images using a video camera allows for rapid, real-time measurement of
multiple cell samples. Microcomputers, widely available in laboratory settings,
are convenient devices for housing digitzer boards, implementing image analysis
algorithms, and storage of data in files for easy analysis.
As a result of the numeric classification of spectral and spatial data in a
digitized image, exacting calculations of length, area, shape, and size
(morphometric data) or color, brightness, tissue quality, and density
(photometric data) are available from the image of a cell culture specimen. The
image analysis method for cell and SE culture measurement provides more detail
than available to routine observation, and allows more samples to be rapidly
counted or measured. Cell samples can often be viewed noninvasively and in
natural form in situ. Often, for conventional measurement such as cell counts,
which require pretreatment of samples with chromic acid, the cultures are
removed from the in vitro environment for measurement. Both pretreatment and
extraction to the ex vitro environment may create measurement artifacts. Image
analysis is essentially a hands-off, quantifiable means of assessing culture
performance.
4 Approaches to Vision Analysis

of Cell and Somatic Embryo Culture
4.1 Cell Culture Biomass Measurement
Straightforward evaluations of cell culture area (growth in two dimensions) or

cell number are standard criteria for many types of in vitro systems. Cell counts
are calculable via image analysis of microscopic samples. Usually, the lowest
microscope magnification is preferred to maximize cell culture volume viewed
in a single image, and also allowing good discrimination of image features. In
simple evaluations, cell cultures are imaged, and cell features are separated
(threshholded) from the remainder (background) of the image. The number of
image points comprising the total area of the cells automatically calculates the
area of all the cells in sample. When the size of individual cells is fairly uniform,
cell number can be rapidly estimated by dividing total area of cells by the size of
an individual cell. As a demonstration of this technique, I-ml samples from
suspension cultures of tobacco were viewed at fourfold magnification with an
inverted microscope, and cells within the field of view were counted manually.
Next, an image of the same microscopic view was captured, pixels corresponding
to cells were separated from background image points, and total pixel counts
were plotted against manually obtained cell counts (Fig. 2), revealing excellent
correlations between the two sets of data (Smith and Spomer 1987). Simple
evaluations using these techniques have proven to be quite accurate, and a much
larger number of samples can be assessed in the same amount of time it would
take to manually count or observe a single microscope sample.
92 M.A.L. Smith
Fig. 2. Image analysis measurements of

Nicotiana tabacum L. cv. Xanthi cultures on a
8 I + cell number basis. Samples (I ml) from
T suspension cultures diluted over a range of
cell densities were viewed at fourfold
magnification with an inverted microscope.
Cells within the field of view were manually
0 6 counted prior to imaging. (Smith and Spomer
o 1987)
-
o
o 5 10 15 20
Number of Cells in Field
As cells in a callus colony grow in three dimensions, and cell colonies increase
in depth, growth is mirrored by a corresponding increase in the area and visual
density of the culture image. The ability of machine vision to very accurately
record incremental increases in three-dimensional cell growth, using extracted
information about pixel gray level in addition to spatial data, was conclusively
demonstrated again using a cell line of tobacco (Smith and Spomer 1987).
Suspension cultures were diluted to four distinct concentrations, dispensed in
10 ml volumes into culture plates, then mixed with an equal volume of medium
with 1.2% molten agar. The cell plates, which initially contained single cells
evenly dispersed within the layer of medium, grew into aggregates and small
colonies over a culture period of 14 days. In order to record machine vision data
corresponding to incremental increases in culture growth, the cell plates were
macroscopically viewed in a custom-constructed viewing unit (Fig. 3) which
excluded outside light, allowed cells to be consistently imaged in the same areas
of the plate repeatedly, and standardized the camera focal distance and
backlighting under the plate being viewed.
As cell colonies grew over the culture period, both the area of cell cultures in
the image the density increased. The sample-digitized images, collected
periodically over the course of a growth cycle, faithfully recorded cell/callus
growth increments which generated the corresponding visual results. Pixel gray-
level values were darkest near the centers of callus masses (where cell depth/
density was highest), and gradually diminished near the outer margins of a
colony, where single cells formed the periphery between the colony and the
culture medium. Computer-derived data recording of these features can be
Fig. 3. Staging unit constructed for

macroscopic viewing of cell culture plates. A
A CCD video camera; B macro-lens; C
lens mount; D coarse focusing tubes; E
sliding stage for registering cell culture
plates, allowing repeated, time-course
observation of exactly the same views of
growing cell colonies; F plate inset; G
aperture for radiation transmission; H
lamp. The unit blocked all extraneous light
from the cell image for consistent, uniform
image analysis. (Smith and Spomer 1987)
applied in a weighted density formula (Smith and Spomer 1987; Smith and
McClelland .1991) to derive growth curves, which correspond well to the
measured fresh weight of the cultured tissues. Accurate evaluations of the effect
of plating density on culture growth were derived nonintrusively using the
machine vision method of assessment. The resolution of recently introduced
CCD cameras and digitizers is currently so refined that image analysis can be
applied as an accurate tool to nonintrusively assess subtle changes in pigment
development in a cell culture (as well as morphological/growth changes), in
experimental and industrial systems for production of secondary metabolites
(Fig. 4).
Image analysis has been used in a similar way to directly evaluate callus
growth from carrot suspension cell inoculum on solidified plate cultures
(Olofsdotter 1993). In this case as well, machine vision proved to be more
accurate, and better able to discriminate intermediate growth stages than
manual, visual counting of cell aggregates. Other investigators have proposed
94 M.A.L. Smith
....._.-............
.... j
.ISail :
....... ....
nil
Fig. 4A,B. Digitized images of Ajuga replans cell cultures. Image analysis is used to monitor and
quantify the growth and production of anthocyanin pigments in the cells. A A digitized view of a callus
colony, showing how the callus image can be segmented to allow morphometric calculation of volume
and density. The morphometric data are very highly correlated to callus fresh weight. B Digitized views
of a suspension culture sample containing both pigmented and colorless cells. Pigment concentration
per volume of suspension is calculated from the image, cells characterized by different pigment contents
can be sorted, and estimates of total culture productivity can be obtained
more complex algorithms and equations, using digital image analysis data, to
predict callus fresh weight, however, simple, straightforward measurements of
image density and area data based on integrated gray level and data already
provide an excellent alternative to destructive measurements.
4.2 Embryo Recognition and Evaluation
Image analysis for evaluation of embryogenic callus and somatic embryos

requires not only estimation of total culture biomass, but also some capacity to
recognize shape (embryo or embryogenic callus morphology). Embryos which
have reached a certain predeterminded status of maturity (typically the torpedo
stage) must be separated from callus, embryogenic callus, and immature or
abnormal SE, prior to transfer to the next appropriate set of culture conditions
(Fig. 5.)
The human eye detects, interprets, and judges the quality of somatic embryos
and embryogenic callus using a complex set of microscopically observed
characteristics including size, shape, color, degree of curvature, etc. Image
analysis substitutes for human judgement in somatic embryo selection by
analyzing culture images using computer vision algorithms, which are capable of
judging these same morphological and chromatic characteristics. Various
approaches to making SE quality assessments with image analysis have been
reported; each concentrate on different visual features for making classifications.
Many use a combination of spectral density and morphological features; some
require image thinning algorithms to simplify identification of key features. In
general, the photometic qualities (density and transparency) recorded via image
analysis have been most useful for classifying and sorting embryogenic and
nonembryogenic callus, whereas spatial characteristics (size, shape, curvature,
etc.) have been exploited to sort and identify embryo classes. For example, Grand
d'Esnon et al. (1989) used a macro-lens to view sweet potato somatic embryos
through a Plexiglas window connected to a bioreactor. The research team
developed a program to identify embryogenic callus clusters based on visual
density (embryogenic cell aggregates are darker when backlighted). In addition,
for later stages in development ontogeny, the group implemented a set of
standard templates to mirror observed, torpedo-shaped sweet potato somatic
embryo morphologies, then developed a shape-recognition algorithm based on
matching the contours of the embryo image.
Using a different approach, Cazzulino et al. (1991) relied on a combination
of image area, image diameter, and image perimeter data to classify carrot
somatic embryos. HamaUiinen and colleagues (1992) successfully classified SE
using algorithms that initially discard objects that are clearly not SE, then
proceed in a second grading to classify selected objects as to embryo stage, and
provide an index of the embryo stage. Harrell and Cantliffe (1991) used a
combination of 17 image geometric features processed by a neural network
(including symmetry, area, length, and elongation) to rank SE suitability for
harvesting. Kurata et al. (1991) developed classifications applied to SE
recognition based on a thinning procedure for discernment of a normal embryo
96 M.A.L. Smith
CALLUS BULK UP
---_.
!
image analysis
recognition
EMBRYOGENIC CALLUS
!
SOMATIC EMBRYOS
_ _ _ _ image analysis
classification
machine vision
evaluation! sorting
robotic ~andling
torpedo shaped
embryo t
_ _ _~._ENCAPSULATION
SYNTHETIC SEED
/ \
+
CRYOPRESERVATION ---l"~ GERMINATION
DESICCATION
CONTAINERIZATION
Fig. S. Schematic presentation of stages in a complete somatic embryogenesis production system,

indicating opportunities for using image analysis to automate or improve the efficiency of the process
in an image. In this case, successful torpedo stage embryos were classified from
the skeletonized, thinned embryo images based on two image parameters:
number of branches and curvature of the longest branch of the skeleton. Later,
an algorithm based on the Fourier descriptor (for image recognition) was tested
by the same group for the task of embryo recognition (Kurata and Shono 1992),
however, the thinning method was judged to be more sensitive and useful.
Similarly, Cheng and Ling (1992) have recently used an embryo image thinning
process and pattern recognition algorithm for evaluating normality of torpedo-
shaped coffee somatic embryos, and predicting their potentials to further
germinate into plants. For this study, SE were spaced on a culture plate for
viewing with three video cameras with zoom lensing. Processing was first applied
to improve image clarity, then a thinning alogrithm was used to simplify the
images prior to extraction of significant measurement features. Vision-based

computer judgements of embryo normality were highly correlated with human
judgements of embryo quality.
Most of these image classification schemes have required destructive
sampling, however, Harrell et al. (1992) developed a modified apparatus that
cycles samples from a bioreactor into an imaging chamber (vision vessel) and
recycles the sample aseptically.
In order to use image analysis to accurately recognize embryos, the system
must be capable of recognizing SE over the normal range of shapes and sizes.
Once a broad array of shapes is established in this way, the ability of machine
vision to recognize the range of features can be determined. Sucrose con-
centrations and population densities were varied experimentally in order to
induce a wide range of different somatic embryo morphologies in carrot cell
culture. Somatic embryos were categorized by shape according to the relative
development of root and shoot axes. These experiments were aimed towards use
of somatic embryo morphology as criteria to select embryos with highest
probability for survival at transplant, and to provide a thorough calibration set
for machine vision interpretation (Kurata and Shono 1992).
Sophisticated computer algorithms can effectively assess multiple
programmed characters simultaneously to arrive at a classification scheme. A
key point on the use of machine vision for embryo recognition is that specific
recognition algorithms need to be created to maximize efficiency for each unique
plant species in production (Hamalainen et al. 1992). Although machine vision
can reliably classify many SE for many species without false selections, current
research shows that computer judgement is not as accurate as the human eye to
distinguish fine differences in embryo shape (Kurata and Shono 1992).
5 Potential for Machine Vision in Culture System Automation
Because machine vision analysis can be useful in many stages of in vitro culture,
and for a range of different culture systems, the technology has great promise as
a tool for automating culture protocols. In particular, the technology is
intensively applied to automation of the bioreactor-based mass harvest of
somatic embryos, towards commercial scale-up.
Mass production (commercialization) ofSE demands that mechanization be
implemented during some or all of the stages in embryo development and plant
regeneration (Fig. 5). Bulk handling of somatic embryos involves growing the
units in high volumes in controlled bioreactor systems. In the bioreactor stages of
production, machine vision can be assigned to the task of sorting and classifying
embryos as described earlier. Robotics and computer vision may be used hand in
hand to standardize the somatic embryo product.
Novel, automated procedures have very recently been reported for
determining the developmental status of a culture (proportion of units at each
developmental stage), selection of embryogenic callus, recognition of SE at
predefined stages of maturation, sorting of suitable embryo units, encapsulation,
98 M.A.L. Smith
and other tasks in the multistep process, by several laboratories operating on a

global scale. In many cases, image analysis is a key component of the automated
system.
Because of their small, uniform size, highly synchronized somatic embryos
are suited to mechanization, encapsulation, automation, and bulk handling.
Bulk-up of embryogenic callus and SE maturation can be adapted to liquid
phase bioreactor systems for mass production. Somatic embryos also provide
a convenient propagule for transport and production, as they can be cryo-
preserved, desiccated, or cold-stored (see Chap. 11.8 by Bajaj). These features
enhance the flexibility of SE as compared to many other tissue culture systems,
and increase the competitive economics of the process for clonal crops, or high-
value transgenic plants.
Embryogenic callus is frequently selected by size in somatic embryo
production systems, by sieving through a series of meshes. Automated, rather
than manual sieving, can be substituted at this production stage. The size-
recognition capability of image analysis can be effectively applied to simplify the
process of sorting by size, and increase the accuracy of the sorting. Color/density
data are especially useful in recognizing embryogenic versus nonembryogenic
callus (Grand d'Esnon et al. 1989).
As described previously, significant hindrances to mechanization ofSE are
culture asynchrony, and the mixture of normal and malformed units in the same
culture broth. Commercialization and scale up ofSE has been limited in the past
by insufficient biological control over the quality and degree of synchronization
of somatic embryos. Economically viable scale up will absolutely require
efficient, reliable methods to assess embryo quality and rapidly evaluate
conditions that promote top quality production. Machine vision, in particular
the sophisticated mathematical programming to categorize somatic embryos
based on perceived external morphology, is an innovation which is developing
currently to facilitate automated handling, and to move the process towards
economic feasibility. Machine-vision guided robotic harvesters for somatic
embryos of birch and Norway spruce have demonstrated that reliable,
automated embryo recognition can be achieved both on solid plate cultures and
in liquid bioreactor production (Hiimiiliiinen et al. 1992).
Harrell et al. (1992) developed a fluidic, in vitro embryo monitor and
harvester for sweet potato liquid-based embryos grown in bioreactors. In this
system, sterile aliquots of culture broth were introduced to a glass conduit,
imaged with video camera, and embryos were ranked according to harvest
suitability, without intruding on the culture environment. Embryo units ranked
as harvestable were automatically ejected from the culture broth by a timed pulse
stream of medium created by a peristaltic pump; nonharvestable units were
recycled to the culture bioreactor for further growth and development.
During later harvest and processing of mature somatic embryos, robotics
and machine vision have been combined in quality control to select ideal embryos
for encapsulation, which will enhance conversion frequency. Simple size
screenings are adequate to sort some embryos, which is an easy task for
computerized image analysis. Morphological evaluation of embryos may be
required to assess quality, which requires more sophisticated vision analysis.
The economic viability of the SE method of propagation is more likely to be

achieved ifthe problem-prone, labor-intensive transplanting and acclimatization
stages can be reduced or eliminated. Synthetic seed technology for somatic
embryos, which permits embryos to be seeded directly into a planting site, is
continually improving. For example, Sakamoto et al. (1992) developed a bead
sorter to handle gel-encapsulated carrot SE, which relied on machine vision for
automated sorting. Gel beads were transferred to the surface of a rotating drum
and imaged using a CCO camera, at a rate of 18 000 beadslh. The color images
were analyzed to determine the proportion of green area in the encapsulated unit;
synthetic seeds which met or exceeded the minimum standard were collected for
direct seed planting. The use of the machine-vision aided sorter resulted in an
elevated conversion frequency for carrot SE. The use of image analysis in sorting
synthetic seeds can be combined with automated gel bead seeders to further
increase the efficiency of the process.
6 Remaining Challenges for Image Analysis

of Somatic Embryos
The shift from experimental systems to full commercial use of cell and somatic
embryo evaluation and sorting hinges on future improvements in efficiency and
accuracy. New image analysis system refinements for individual cell and somatic
embryo cultures are aimed at increasing the accuracy of classifications (fewer
false identifications of embryos, or discards of viable embryos during sorting).
More sophisticated programming is under development in many laboratories, to
account for a broader range of embryo variability and to refine selection criteria.
The automated classification and harvesting systems described for SE have
demonstrated that this process can be significantly enhanced by machine vision
analysis, however, improvements in programming (to increase the accuracy of
embryo sorting) and aseptic adaptations are still required (Harrell et al. 1992).
The relative merit of different algorithms applied to the problem of SE
recognition and classification remains to be determined. In most cases described
above, the correctness of machine vision algorithms in judging embryo quality
has been based on comparison with visual observations, rather than on
evaluation of subsequent embryo germination and survival.
Currently, species produced by somatic embryos must have a high value
to justify production costs, and to date no viable full-scale somatic embryo
production system has been commercialized.
Individual researchers or commercial producers desiring to use some form of
machine vision to aid in vitro measurements currently have a broad range of
options to select from. Complete commercially marketed image analysis systems
have been developed with specific attention to in vitro researchers. Emerging
technical refinements in individual system components, in particular the rapidly
evolving capabilities of new CCO color video cameras, suggest that building
individual systems from off-the-shelf components may be a more cost-effective
option. In the latter case, an image analysis system can adapt to changing
100 M.A.L. Smith
research and production needs as required, and new, superior instruments can be
periodically purchased to replace less sensitive or versatile components.
Acknowledgment. The author expresses appreciation to Dr. John Reid, a collaborator in machine
vision analysis of pigmented plant cell cultures in Fig. 4.
References
Cazzulino D, Pedersen H, Chin CK (1991) Bioreactors and image analysis for scale-up and plant
propagation. In: Vasil IK, (ed) Cell culture and somatic cell genetics of plants, vol 8. Scale-up and
automation in plant propagation. Academic Press, San Diego, pp 147-177
Cheng Z, Ling PP (1992) Computer vision for plant embryo quality evaluation. Amer Soc Agric Eng,
1992 Int Winter Meet, Chicago, Illinois, Pap 923575
Grand d'Esnon A, Chee R, Harrell RC, Cantliffe DJ (1989) Qualitative and quantitative evaluation of
liquid tissue cultures by artificial vision. Biofutur 76: S3
Hiimiiliiinen JJ, Kurten U, Kauppinen V, Heilala J (1992) Automated classification of somatic plant
embryos. Acta Hortic 319: 601-605
Harrell RC, Cantliffe DJ (1991) Automated evaluation of somatic embryogenesis in sweet potato by
machine vision. In: Vasil IK (00) Cell culture and somatic cell genetics of plants, vol8. Scale-up and
automation in plant propagation. Academic Press, San Diego, pp 179-195
Harrell RC, Bieniek M, Cantliffe DJ (1992) Noninvasive evaluation of somatic embryogenesis.
Biotechnol Bioeng 39: 378-383
Iriondo JM, Perez C (1991) Use of image analysis techniques for development studies in "in vitro"
regenerated plants of Lavatera oblongifolia. Acta Hortic 289: 335-336
Kamada H, Harada H (1979) Studies on the organogenesis in carrot tissue cultures. II. Effects of
amino acids and inorganic nitrogenous compounds on somatic embryogenesis. Z Pflanzenphysiol
91: 453-464
Kurata K, Shono H (1992) An application of Fourier descriptor for selecting somatic embryos. Acta
Hortic 319: 591-594
Kurata K, Terada M, Komine M, Liyanage KH (1991) Computer vision for selecting somatic
embryos. ASAE Pap 91-3054
McClelland MT, Smith MAL (1990) Influence of vessel type, closure, and explant orientation on
in vitro performance of five woody species. HortScience 25: 797-800
McClelland MT, Smith MAL, Carothers Z (1990) The effects of in vitro and ex vitro root initiation on
subsequent microcutting root quality. Plant Cell Tissue Organ Cult 23: 115-123
Olofsdotter M (1993) Image processing: a non-destructive method for measuring growth in cell and
tissue culture. Plant Cell Rep 12: 216-219
Roberts DR, Sutton BCS, Flinn BS (1990) Synchronous and high frequency germination of
interior spruce somatic embryos following partial drying at high relative humidity. Can J Bot
68: 1086b
Rogers RB, Smith MAL (1992) Consequences of in vitro and ex vitro root initiation for miniature rose
production. J Hortic Sci 67: 535-540
Sakamoto Y, Mashiko T, Suzuki A, Kawata H, Iwasaki A (1992) Development of encapsulation
technology for synthetic seeds. Acta Hortic 319: 71-76
Senaratna T (1992) Artificial seeds. Biotechnol Adv 10: 379-392
Shibli R, Smith MAL, Spomer LA (1992) Osmotic adjustment and growth responses of three
Chrysanthemum morifolium Ramat. cultivars to osmotic stress induced in vitro. J Plant Nutr 15:
1373-1381
Smith MAL, McClelland MT (1991) Gauging the quality and performance of woody plants produced
in vitro. In Vitro Cell Dev BioI 27P: 52-56
Smith MAL, Spomer LA (1987) Direct quantification of in vitro cell growth through image analysis.
In Vitro Cell Dev BioI 23: 67-74
Smith MAL, Spomer LA, Meyer MJ, McClelland MT (1989) Non-invasive evaluation of growth
during plant micropropagation. Plant Cell Tissue Organ Cult 19: 91-102
Smith MAL Spomer LA, McClelland MT (1990) Direct analysis of root zone data in a microculture
system. Plant Cell Tissue Organ Cult 23: 21-26
Smith MAL, Wagner R, Anderson JS, Spomer LA (1991) A simple in vitro assay for Phytophthora
megasperma f. sp. glycinea symptom expression on soybean taproots. Crop Sci 31: 1364-1366
Smith MAL, Meyer J, Knight SL, Chen GS (1993) Gauging turfgrass salinity responses in whole plant
microculture and solution culture. Crop Sci 33: 566-572
Spomer LA, Smith MAL (1988) Image analysis for biological research: camera influence on
measurement accuracy. Intell Instrum Comput July/August 1988, 11: 201-216
Section II
Applications of Somatic Embryos; Technology
of Synthetic Seed; Fluid Drilling; Micropropagation,
and Genetic Transformation Through Somatic Embryos;
Cryopreservation
11.1 Somatic Embryogenesis and Its Applications
for Crop Improvement
Y.P.S. BAJAi
1 General Account
One of the most spectacular achievements in plant tissue culture has been the
discovery of the induction of somatic embryogenesis in cell cultures (Reinert
1958; Steward et al. 1958). It demonstrated the persistence of totipotency in cells
of higher plants, i.e., the regeneration of whole plants (see Reinert et al. 1977).
These embryos are bipolar, in contrast to monopolar shoots or roots. Various
developmental stages of somatic embryos, i.e., proembryo, globular, heart- and
torpedo-shaped, are all strikingly similar to those occurring after the fertilization
of the egg cell (Reinert 1959; see Figs. 1-12). Various terms, such as embryo-like
structures, embryoides, adventitious embryos, accessory embryos, secondary
embryos, asexual embryos, somatic embryos, etc., have been applied to these
structures by various authors. In this chapter, the term somatic embryo has been
used. Perhaps the best definition of such structures (Haccius 1971) states that the
decisive feature for categorizing a plant structure as an embryo, besides other
morphological properties, is its bipolarity and the fact that at the earliest
developmental stage it has at opposite ends a shoot and a radicular pole.
Furthermore, this system must not be connected with the vascular tissue of the
mother plant or the explant during its initiation and development. With mono-
polar buds and roots, on the other hand, it is often possible to show their
connection with the vascular elements of the mother plant or in the callus.
The phenomenon of somatic embryogenesis has now been reported in over
300 species, in a wide range of plants, i.e., trees (Section III, Chapters 1-16, this
Vol.), cereal and grasses, vegetables and fruits, legumes and oilseed crops,
ornamental, medicinal and miscellaneous plants (see Somatic Embryogenesis and
Synthetic Seed II).
During the last decade, considerable work has been done on factors affecting
somatic embryogenesis, ultrastructural studies, molecular aspects, genetic
transformation, and gene expression in somatic embryos (SE). Large-scale
production of SE in bioreactors, mass propagation, synthetic seed and fluid
I Former Professor of Tissue Culture, Punjab Agriculture University, Ludhiana (Punjab), India
(Present address: A-137 New Friends Colony, New Delhi 110 065, India)

106 y.P.S. Bajaj
Figs. 1-9. Somatic embryogenesis in carrot. Figs. 1-3. One-, two-, and four-celled stages formed in
the parenchyma of carrot; these stages closely resemble those occurring in vivo. Figs. 4--6. Globular
and heart-shaped stages. Fig. 7. Torpedo-stage embryo isolated by immersion of a carrot culture in
liquid medium, note the suspensor-like cells at the radicular end. Figs. 8,9. Plantlets with a well-formed
root, shoot and leaves. (Reinert et al. 1977)
Somatic Embryogenesis and Its Applications for Crop Improvement 107
drilling, and the cryopreservation ofSE are some of the applications which have
attracted much attention. These recent developments are summarized in this
introductory chapter, which is based primarily on information extracted from
various chapters of this book.
1.1 Early Events and Ultrastructural Studies on Somatic Embryogenesis
Extensive studies have been conducted on the cell's commitment to, and early
events in, somatic embryogenesis (see Chapters 1.1, I.2, this Vol.). Histological
analysis was done to identify the progenitor cells; another approach was to
determine the sensitivity to auxin. The sensitivity to auxin was measured as the
auxin binding capacity (ABC) of a crude membrane preparation of cells exposed
to different concentrations of 2,4-D. Biochemical identification of various
classes of auxin binding proteins (ABP) in carrot cell cultures has been made
(Lo Schiavo et al. 1991). When carrot proembryos (PEMs) are transferred to a
medium without auxin, cells are filled with cytoplasm in a matter of hours,
frequent divisions follow, and in 2 to 3 days a globular embryo is formed.
Recently, Nuti Ronchi et al. (1992) have observed a novel mechanism of
"somatic meiosis", capable of producing totipotency. They described a series of
events in cell cultures which match the steps occurring during meiosis
(gametogenesis). They suggested that this mechanism generates a cell in vitro
(progenitor cell of the somatic embryo) which assumes a role similar to that of the
fertilized ovule at the onset of embryogenesis.
Earlier ultrastructural studies were conducted on Atropa belladonna
(Thomas et al. 1972), Citrus sinensis (Button et al. 1974), and carrot (McWilliam
et al. 1974). The precise patterns of early events in embryogeny in somatic cells in
vitro may not necessarily be the same as in vivo (Borthwick 1931), but there is at
least a very close similarity. The ultrastructure of carrot embryogenic clumps
Fig. to A-C. Somatic embryogenesis in protoplast-derived callus cultures of Atropa belladonna. A An

early stage of embryo formation in a callus transferred to liquid differentiation medium (NAA-free MS
supplemented with 8% coconut water and 0.1 mg/I kinetin). B,C Maturing embryos from 2- and 6-
week-old cultures in liquid differentiation medium; note the regeneration of a root in B. (Gosch et al.
1975; Bajaj et al. 1978)
\08 Y'P.S. Bajaj
su
A B
o
Fig. II A-D. Induction of polyembryony in in vitro cultured globular proembryos of Dendrophthoe
falcata. (Johri and Bajaj 1965)
(Street and Withers 1974) shows that the superficial cells destined to become
embryos are highly cytoplasmic and possess a large, diffusely staining nucleus
with a single, darkly staining nucleolus. These cells also show intense staining
with the protein dyes nigrosine and coomassie blue. With the RNA dye
gallocyanin, a high content of RNA is observed in the nucleoli and the
cytoplasm. In addition, these cells showed high dehydrogenase activity. On the
other hand, the central cells in a clump show weak or negative responses.
Somewhat similar observations were made earlier by Halperin and Jensen
(1967).
Recently, extensive studies have been conducted on monocots, especially
cereals and grasses. Observations on the earliest stages of somatic embryogenesis
in these plants show evidence for a unicellular as well as multicellular origin ofSE
(Vasil et al. 1985; McCain and Hodges 1986; Fransz and ScheI1994). Evidence
Somatic Embryogenesis and Its Applications for Crop Improvement \09
B c
Fig. 12 A-c' Anatomy of somatic embryos of Sinapis alba, from 25-week-old differentiating callus,
at various stages of development. (Bajaj and Bopp 1972; Bajaj et al. 1973)
for a unicellular origin in maize scutellar tissue was proposed by Fransz and Schel
(l991a,b), who observed discrete clusters of small cytoplasm-rich cells in
immature embryos after 10 days of culture. A unicellular origin was further
demonstrated in the scutellum epithelium of mature rice embryos (Jones and
Rost 1989) and in leaf segments of orchard grass (Trigiano et al. 1989). Evidence
for a multicellular origin after a callus phase comes from Sorghum bicolor
(Wernicke et al. 1982), and for a unicellular origin from Pennisetum americanum
(Botti and Vasil 1984).
1.2 Role of Polyamines in Somatic Embryogenesis
Polyamines are involved in cell division, growth, and morphogenesis in animals

and plants (Bagni 1989; Heby 1989). The most common polyamines are
putrescine, spermidine, and spermine. Their biosynthesis was studied in a
number of organisms (Pegg 1986; Smith 1990; Slocum and Flores 1991). A
number of studies have examined the involvement of polyamines in somatic
embryogenesis (see Chap. I.5, this Vol.). The following is a summary of the work
on various plant species.
110 y'P.S. Bajaj
Extensive studies have been conducted on the role of polyamines in the

somatic embryogenesis of carrot. In carrot cell cultures, putrescine or spermidine
is the predominant polyamine, spermine being present in lesser amounts; traces
of cavaderine have also been reported (Baker and Yon 1983). In earlier studies,
Montague et al. (1978) reported that putrescine levels increased sharply when
cells were transferred to fresh medium, regardless of the presence or absence of
2,4-D. However, embryogenic cultures maintained a higher level of putrescine
than the nondifferentiating cells. In the absence of2,4-D , the rate of biosynthesis
of putrescine was almost twofold greater than when it was present. Fienberg et al.
(1984) confirmed the importance of polyamine biosynthesis during the
differentiation of SEs in carrot. They observed that both the embryogenic
cultures grown in the presence of 2,4-D and the nonembryogenic cultures in (+)
or (-) auxin media have similar changes in profiles with regard to putrescine,
spermidine, and spermine.
A study was conducted on the role of a suspensor in the synthesis and
translocation of materials which can be essential for developing embryos of
Phaseolus (NagI1990). There was rapid translocation of putrescine through the
funiculus and the suspensor, a nutritive embryonic haustorium. The results
indicated that embryo development in vivo required polyamines, which may
be one ofthe limiting factors for embryogenesis in vitro in rubber (El Hadrami
et al. 1989).
In attempts to increase the efficiency of embryogenesis in celery (Altman
et al. 1990) and grape (Faure et al. 1991), it was observed that it is not the
presence of a high polyamine titer that is important, but adequate polyamine
levels and putrescine/spermidine ratio. In alfalfa, putrescine levels increased up to
32-fold during exposure to medium containing 2,4-D and kinetin, but fell sharply
after transfer to hormone-free differentiation medium. This rapid decline
coincided with early embryogenesis (Meijer and Simmonds 1988).
Roustan et al. (1989) reported that inhibitors of ethylene biosynthesis
(salicylic acid and acetylsalicylic acid) and ethylene action had a strong
stimulatory effect on embryogenesis in carrot. Exogenously supplied ethylene
(ethephon) caused a substantial decrease in SEs. Keeping in mind that the
polyamine and ethylene biosynthetic pathways share a common precursor,
Robie and Minocha (1989) studied the production of ethylene by carrot cell
cultures under embryogenic and nonembryogenic conditions, and observed
that significantly more ethylene was generated by cells growing in the presence
of2,4-D.
Studies by Roustan et al. (1992) indicated that the diminished polyamine
biosynthesis caused by ethylene could result in the inhibition of somatic
embryogenesis in carrot. On the other hand, the stimulation of SEs, occurring
when ethylene synthesis is inhibited, could be attributed to an enhancement of
polyamine formation (Mengoli and Bagni 1992).
Minocha and coworkers (Robie and Minocha 1989; Khan and Minocha
1991; Minocha and Khan 1991; Minocha et al. 1991a,b; Nissen and Minocha
1993) focused on the effect of several inhibitors of polyamine biosynthetic
enzymes on the metabolism of polyamines and ethylene and the activities of
ODC (ornithine decarboxylase), ADC (arginine decarboxylase), SAMDC
Somatic Embryogenesis and Its Applications for Crop Improvement III
(S-adenosylmethionine decarboxylase), and SAM (S-adenosylmethionine)

synthetase in carrot cell cultures. They showed that the only polyamines seen in
carrot cells are putrescine, spermidine, and spermine. For more details, see
Mengoli and Bagni (1992) and Chapter I.5 (this Vol.).
1.3 Molecular Basis of Somatic Embryogenesis

and Gene Expression
Somatic embryogenesis in carrot cells has become an attractive model system for
studying the molecular mechanisms underlying this fundamental process of
plant development (Aleith and Richter 1990). Carrot cell cultures proliferate in a
synthetic medium containing 2,4-D, and upon its removal from the medium
the cells form embryos (Backs-Hiisemann and Reinert 1970). The ease of
transforming most of the proliferating cells to SEs allows the comparison of
unorganized cells and developing embryos with regard to the identification
of genes involved in embryonic organization.
In carrot cell suspensions, the removal of the auxin induces SEs, although
generally at low frequencies and asynchronously. However, for investigations of
SE mechanisms, especially at the molecular level, high-frequency, synchronous
embryogenesis systems are required. Thus, for this purpose, Fujimura and
Komamine (1979a) established such a system. Research on the cytology and
physiology of embryonic determination in somatic carrot cells made this a useful
system for molecular analysis (Komamine et al. 1992).
Studies show that many of the molecular processes of embryogenesis have
already been established in PEMs in the presence of auxin, and indicate the
usefulness of Dc3 as a molecular marker to enhance the embryogenic potential.
Synchronous embryogenesis is frequently initiated by enriching the carrot
cell culture fraction containing clusters of small, densely cytoplasmic cells
(Fujimura and Komamine 1979b; Giuliano et al. 1983). Nomura and
Komamine (1985) reported that a particular class of single carrot cells (type 1)
can diffeJentiate into embryos, but only if these cells are cultured in the presence
of low revels of auxin where they divide to form small aggregates operationally
identical to PEMs.
Much work has gone into finding specific molecular markers for SEs.
Komamine and his coworkers (1992) studied several such markers in carrot.
Using in vitro translation and two-dimensional gel electrophoresis, 99% of the
polypeptides produced showed the same patterns between embryogenic and
nonembryogenic cultures. Four different translatable mRNAs encoding
polypeptides were detected, with two appearing in embryogenic and two in non-
embryogenic cultures. This showed that only a few proteins play an important
role in somatic embryogenesis. Timmers et al. (1989) suggested that the presence
of activated calmodulin in the surface cells of PEMs may be correlated with
embryo development from these cells. Analysis of staged SEs of temperature-
sensitive carrot variants at permissive and nonpermissive temperatures revealed
polypeptides with stage-specific expression patterns (Schnall et al. 1991).
112 Y.P.S. Bajaj
The analysis of gene expression during somatic embryogenesis has been

viewed as an alternative means of investigating the genetic control of zygotic
embryo development. The most attractive approach to elucidate the mechanism
of embryogenesis is to isolate genes which are expressed specifically during
embryogenesis. However, as the list of genes affecting embryogenesis continues
to grow, the elucidation of their biochemical functions lags behind. Choi et al.
(1987) isolated several cDNA clones which were preferentially expressed during
embryogenesis. Developmental regulation of two clones was studied in detail,
and the expression of their genes was associated with heart-stage embryos
(Borkird et al. 1988).
Extensive studies of gene expression programs in carrot SEs identified a
gene, Dc3, that serves as a reliable molecular marker enchancing embryogenic
potential (Seffens et al. 1990). A1eith and Richter (1990) isolated cDNA clones,
and observed that the gene expression of several clones was associated with
early stages of embryogenesis. Isolation of genes specific for earlier stages of
embryogenesis was also attempted (Kawahara et al. 1992).
Analysis of gene expression programs in SEs maintained in the presence of
2,4-D revealed relatively few differences in prevalent and moderately prevalent
gene products (Choi and Sung 1984; Wilde et al. 1988). Wilde et al. (1988)
described a cDNA sequence, Dc3, that was derived from mRNA expressed only
in carrot cell cultures that contained SEs or proembryogenic masses (PEMs).
Analysis of Dc3 expression during the initial phases of induction of embryogenic
carrot cultures and in nonembryogenic mutant cell lines indicated that Dc3
mRNA expression is a useful molecular marker for the embryogenic potential in
plant cell cultures (de Vries et al. 1988). Polypeptides and mRNAs of a number
of genes expressed by embryonic carrot cultures have been identified by 2-D
PAGE analysis, nucleic acid hybridization, and immunological techniques.
The pattern of expression of these genes reveals developmental regulation
(transcriptional and translational) during embryo induction and development.
For example, several lines of evidence suggest that the expression of the gene Dc3
correlates with PEM development in the presence of 2,4-D. The 750 base-pair
transcript of Dc3 was present at high levels in both globular and torpedo-stage
embryos. A lea gene (Dc8) was isolated from carrot SE that encodes mRNA,
whose steady state level increase 50 to 100 fold in SE exposed to 10 /!M ABA
(Hatzopoulos et al. 1990)
In contrast to the temporally restricted expression of lea genes during late
zygotic embryogenesis, Dc3 mRNA was expressed in early and late stages of SE
development (Wild et al. 1988). Dc3 transcripts could also be detected in all SE
stages, although their expression appeared to be associated primarily with the
development of heart-stage embryos (Borkird et al. 1988). During embryo-
genesis ECP31 protein was detected in the central region of globular embryos and
was undetectable in torpedo-stage embryos (Kiyosue et al. 1991).
The detection in cell suspensions of mRNA and proteins of genes expressed
developmentally in SEs is not uncommon. Stacey et al. (1990) immuno-
chemically identified a cell-surface arabinogalactan protein (J4e epitope) during
both PEM and SE development. Within 24 h of transferring PEMs into auxin-
free medium, an increase in the level of the epitopewas detected. J4e was localized
in the outer surface layer of cells of globular embryos and its expression
was gradually restricted during embryo development. In mature torpedo-stage
embryos, J4e was expressed primarily by cells forming two regions of the future
stele and by cells associated with the cotyledonary provascular tissue.
Analysis of protein populations demonstrated that there are few differences
in gene expression between carrot cultures grown under conditions permissive
(-2,4-D) or repressive (+2,4-D) for embryo development (Sung and Okimoto
1981). It has been suggested that there are only a small number of new genes that
are expressed at detectable levels during SE development (Choi and Sung 1984;
Komamine et al. 1992). The study of the regulation of genes normally expressed
in differentiated embryos (i.e., lea genes) during embryonic determination may
provide a means of analyzing the genetic mechanisms controlling somatic
embryogenesis.
1.4 Image Analysis of Somatic Embryos
Image analysis is a strategy of collecting visual information with a camera from

a specimen, allowing the visible characteristics to be interpreted with the aid of a
computer. The image, stored in the memory of a computer, provides details of
the specimen. Thus, the image, rather than the live specimen, can be subjected to
examination, measurement, and manipUlation (Spomer and Smith 1988). Image
analysis is nondestructive (Olofsdotter 1993), and with its capacity for shape and
color recognition is a reliable means of collecting information about the specimen
(Smith et al. 1989). As the technology has an extraordinary capacity for
capturing small details in culture, it is a valuable tool for determining the
developmental patterns of in vitro regenerated plantlets (Iriondo and Perez
1991). Image analysis has been particularly advantageous in applications to cell
culture and the somatic embryo system. In order to assess the effect of various
factors on the development of SEs, culture samples are usually examined
microscopically, however, these observations are usually time-consuming and
tedious. Moreover, as the large-scale production of SEs usually takes place in
suspension cultures and in bioreactors, measurement difficulties are intensified.
Video-image analysis extracts quantitative data from a visual image of a SE, and
thus permits objective interpretation at any stage of culture. Grand d'Esnon et al.
(1989) used a macro-lens to view sweet potato SE through a Plexiglass window
connected to a bioreactor. The machine-vision measurements applied to in vitro
cultures have proved to be extremely sensitive (Smith and Spomer 1987), can
overcome many of the problems inherent to microculture evaluations, and
provide a level of information that has never been available previously.
Hiimiiliiinen et al. (1992) successfully classified SEs using algorithms
that initially discard objects that are clearly not SEs, then, using a second
classification selected objects at the embryo stage, and provided an index of
the embryo stage. Kurata et al. (1991) developed classifications regarding SE
recognition based on a thinking procedure for discernment of a normal embryo
in an image. Recently, Cheng and Ling (1992) used an embryo image thinning
process and pattern recognition algorithm to evaluate the normality of torpedo-
114 y'P.S. Bajaj
shaped coffee SEs, and to predict their potential to germinate further into plants.
Harrell et al. (1992) developed a modified apparatus that cycles samples from a
bioreactor into an imaging chamber and recycles the sample aseptically.
Machine-vision technology has great promise as a tool for automating
culture protocols. In particular, it can be extensively applied to the automation of
the bioreactor-based mass harvest of SEs for commercial scaleup. Robotics and
computer vision may be used hand in hand to standardize the SE product.
Sakamoto et al. (1992) developed a bead sorter to handle gel-encapsulated carrot
SEs, which relied on machine vision for automated sorting. The use of this
machine vision-aided sorter resulted in increased conversion frequency for SEs.
Now complete commercially marketed image analysis systems have been
developed with special emphasis on in vitro culture work (see Chapter I.7, this
VoL).
2 Applications of Somatic Embryogenesis
In addition to the use of somatic embryogenesis as a system in basic research in

plants, there are some aspects which have attracted attention due to their
potential in agricultural research/crop improvement programs.
2.1 Genetic Transformation of Somatic Embryos and the Production

of Transgenic Plants
Somatic embryos have proved to be an excellent material for genetic

transformation studies, especially in tree species. SEs are competent to express
introduced DNA, and regeneration can be initiated from single cells on the
surface of the embryos, enabling the regeneration of nonchimeric plants.
The regeneration system is of prime importance for genetic engineering
studies. Zygotic embryos have been used for DNA uptake, and they develop into
complete plants (Sato et al. 1993). However, different developmental stages of
zygotic embryos differ in their competence to respond to embryogenic induction
calli. On the other hand a large number of developmentally uniform SEs are
available for use in place of zygotic embryos. Such a strategy was crucial,
especially in the development of a transformation system for Picea glauca (Ellis
et al. 1993).
Embryos responded favorably to Agrobacterium tumefaciens (Mathews
et a1. 1992), microinjection (Neuhaus et a1. 1987), and particle bombardment
(Wilde et al. 1992). There are now a number of successful reports on the
regeneration of transgenic plants from well-developed SEs or embryogenic cell
cultures. Some notable examples are Carica papaya (Fitch et a1. 1990), Carya
illinoensis (McGranahan et al. 1993), luglans regia (McGranahan et al. 1990),
Liriodendron tulipfera (Wilde and Merkle 1994), Mangifera indica (Mathews
et al. 1992), Picea glauca (Bommineni et al. 1993; Ellis 1993), Vilis rupestris
(Mullins et al. 1990), etc., for more examples see Chapter II.7, this Vol.
2.2 Micrografting of Somatic Embryos
Micrografting, a relatively new technique of grafting shoot tip/apex from a

mother plant to a decapitated young plant grown from a seedling under aseptic
conditions, has resulted in the rejuvenation and elimination of viruses in a
number of species, especially citrus (see Jonard 1986; Navarro 1992). Aguilar
et al. (1992) applied the technique of micrografting to somatic embryos of coco
which are normally difficult to differentiate into complete plants (Novak et al.
1986). They induced SEs on segments of cotyledons cultured in vitro, and the
differentiated embryos were grafted on 3-week-old rootstocks obtained from
germinated seeds. The plantlets used as rootstocks were cut at the hypocotyl
level, a longitudinal incision of 0.5 cm was made, and then a SE was inserted in
the incision. Grafts that developed leaves and roots were transferred to soil. After
18 months two micro grafted plants underwent flowering. This technique may
now be extended to other plant species in which SEs do not normally differentiate
into complete plants.
2.3 Bioreactor Production of Somatic Embryos

and en masse Micropropagation
Micropropagation of ornamental and medicinal plants, fruit, and forest trees

offers not only a means for the mass multiplication of existing stocks of
germplasm for biomass energy production, but also for the conservation of
important, elite, and rare plants that are threatened with extinction. For the
clonal propagation of such plants, afforestation, and orchard plantations, a large
number of plants are needed. Conventional micropropagation is time-
consuming and labour-intensive. During the last decade, there has been increased
interest in problems affecting scaleup operations (see Bajaj 1991a). To increase
the efficiency and scaleup, the production of plants by means of automation has
been employed. For this purpose bioreactors, robots, and various other
mechanized systems are being developed.
Different types ofbioreactors have been used for the large-scale production
of cell cultures and somatic embryos by various workers, i.e., rotating drum
bioreactor (Tanaka et al. 1983), gaseous phase bioreactor (Ushiyama et al.
1984), airlift column bioreactor (Smart and Fowler 1984), spin filter bioreactor
(Sty1er 1985), light-introducing bioreactor, etc. (Margaritis and Wallace 1984).
For cell cultures, bioreactors with a capacity for 450 (Ulrich et al. 1985) to
75000-1 fermenters (Ritterhaus et al. 1990) have been used.
Alfalfa SEs have been grown in various types of bioreactors (Chen et al.
1987; Stuart et al. 1987; McDonald and Jackman 1989). Preil et al. (1988)
studied the development of SEs of Euphorbia in Biostat M - a bioreactor with
silicone tubes for bubble-free aeration. More than 60 000 SEs can be obtained
from 1 L of carrot cell suspensions in a bioreactor (Ammirato and Styer 1985).
Embryogenic cultures of conifers such as Pinus radiata (Smith et al. 1991) and
various species of Picea (Tautorus et al. 1992) have been successfully grown in
bioreactors. Merkle et al. (1991) developed a protocol for SEs of Liriodendron,
and over 5500 plants were field-tested.
116 Y.P.S. Bajaj
Now that SEs are being increasingly employed for genetic transformation
(Ellis 1993), the production of synthetic seed and fluid drilling (Kitto et al. 1991;
Redenbaugh et al. 1991), and for micropropagation, their large-scale production
in bioreactors (Denchev et al. 1992; Ducos et al. 1993) will greatly facilitate these
studies.
2.4 Somatic Embryo-Derived Synthetic Seed
One of the most important applications of SEs is to use them for synthetic seed
- a concept first put forward by Murashige in 1978. Since then extensive studies
have been conducted to realize this goal, and today SE-derived synthetic seed is
reality in a number of plant species (Redenbaugh et al. 1991; Gray and Purohit
1991; Fuji et al. 1992; Redenbaugh 1992; Senaratna 1992; Tay et al. 1993; Attree
and F owke 1993). The production of synthetic seed assumes great significance in
(1) crops where seeds are not produced and vegetative propagation is
cumbersome, (2) where seeds are produced in small quantities, and (3) crops in
which the zygotic seed is short-lived (recalcitrant) and the germplasm cannot be
conserved. Thus, mass production of synthetic seed through somatic embryos
has far-reaching implications in the seed industry.
Studies, especially on carrot (Kitto and Janick 1982, 1985a,b; Wake
et a1.1992) and alfalfa (Redenbaugh et al. 1984,1987; McKersie et al. 1989;
Senaratna et al. 1990) have yielded encouraging results. Various types of gel
coatings, such as calcium alginate, gelrite, gelatin, carrageenan, polyethylene
oxide, etc., have been used for encapsulation, However, calcium alginate,
although it poses some problems, has proved to be the best so far.
The synthetic seed may consist of either a quiescent or nonquiescent SE with
or without protective encapsulation. Naked, nonquiescent SEs germinated in soil
plugs could be used to propagate certain ornamentals, while dehydrated,
quiescent SEs without encapsulation would be useful for germplasm storage (see
Chapter 11.2, this VoL). Encapsulation may provide physical protection to the SE
as well as nutrients, growth regulators, antibiotics, fungicides, etc., (Kitto and
Janick 1985 a,b; Redenbaugh et al. 1991).
Recently, considerable attention has been paid to the desiccation of SEs.
Desiccation causes a developmental switch from storage reserve synthesis to
storage reserve catabolism in developing zygotic embryos (Kermode and Bewley
1985). It is therefore anticipated that seedlings from dry SEs will be more
vigorous and exhibit more rapid and uniform development than seedlings from
precociously germinated SEs (McKersie et al. 1986). Desiccation studies have
been conducted with a view to arrest embryo growth, to provide storage ability,
and to provide normal dry seeds with the ability to be sown (Kitto and Janick
1985a,b; Gray et al. 1987). Desiccation is not only favorable for the storage and
sowing of synthetic seed, but also for the development and maturity of embryos.
Wake et al. (1992) reported the promotion of germination of synthetic seed of
carrot by the extract prepared from a marine cyanobacterium (Synecohococcus
sp.). For storage of the encapsulated seed, liquid paraffin and liquid nitrogen
have been suggested.
Attempts are being made for automated encapsulation and production of

synthetic seed. Sakamoto et al. (1992) developed a system that automatically
encapulated carrot SEs in gel beads, which were sorted visually. They then
germinated them in conventional growing trays. The self-breaking gel beads were
composed of modified hydrogel and contained a sucrose-substained release
microcapsule. The system was capable of producing and sowing 70 000 synthetic
seeds per day; the plant recovery rate was 52%.
2.5 Fluid Drilling of Somatic Embryo-Derived Plants
Fluid drilling is a crop establishment technique that includes seed germination,

incorporation of the germinated seeds into a gel, and planting of the gel-seed
mixture in the seed bed (Pill 1991). Fluid drilling can give (1) earlier, greater, and
more uniform seedling emergence, (2) earlier and greater yields, and (3) in some
crops, more uniform maturity than conventional methods of sowing dry seeds
(Gray 1984). The fluid drilling technique has the potential for bulk handling of
many small plantlets without the need for individual handling. In addition, the
protective carrier gel can contain additives such as fertilizer salts, plant growth
regulators, pesticides, etc., thereby creating a "packaged" environment for the
seed and seedling (Salter 1978).
Early research examining fluid drilling as a delivery system for somatic
embryos demonstrated both its feasibility and drawbacks (Baker 1985). While
SEs of carrot survived up to 7 days postfluid drilling, they failed to elongate and
grow normally. The effect of an amended hydroxyethyl cellulose gel on SEs of
sweet potato to allow maximal SE maturation and subsequent development into
normal plants has been examined (Schultheis and Cantliffe 1992). These workers
observed that both nutrients and the carbohydrate source of the gel influenced
the normal development of SEs. Field planting of alfalfa SEs demonstrated the
importance of maintaining proper moisture levels around SEs. Field-sown SEs
provided with a protective covering (Styrofoam) survived and grew better than
SEs sown directly in the seedbed (Fujii et aI1992).
Kitto et al. (1991) carried out extensive studies on fluid drilling of carrot SEs.
The composition and volume of the fluid-drilling gel, the role oflight and chilling,
and the percent embryo conversion were determined. The SE-derived plantlets
converted into plants when incubated for 8 days in gel containing 2% sucrose and
250 mg/l Truban fungicide. Plants appeared normal 12 weeks after transplanting
into growth medium under glasshouse conditions. Light or chilling during
suspension culture, followed by a lighted incubation period in fluid-drilling gel
having a reduced water potential and containing sucrose and fungicide or
chitosan, favored the conversion of SE into plantlets. Sucrose, chitosan, and
Truban appeared to be beneficial. Thus, it is clear from these experiments that
SEs can be transferred successfully to the seedbed using fluid drilling. For further
details, see Chapter II.5 (this Vol.).
118 Y.P.S. Bajaj
2.6 Cryopreservation of Somatic Embryos for Conservation

of Germplasm
Somatic embryos are an excellent material for cryopreservation studies; their

potential is manifold. Since SEs can be produced in large numbers in bioreactors,
and are employed for mass propagation, genetic transformation, and for the
production of synthetic seed, methods are being developed for their storage. In
this regard, cryopreservation in liquid nitrogen has attracted considerable
attention, especially for the conservation of germplasm (Bajaj 1986, 1991b).
Considerable improvements have been made in the conventional method of
freeze storage, and new refined methods (vitrification, desiccation, gel coating,
encapsulation of somatic embryos) have enabled high viability and plant
regeneration in a number of plant species such as coffee (Tessereau et a1. 1991),
carrot (Bajaj 1976), citrus (Marin and Duran-Vila 1988), cassava (Sudar-
monowati and Henshaw 1990), oil palm (Engelmann and Duval 1986),
sugarcane (Eksomtramage et a1. 1992), asparagus (Uragami et al. 1989), etc.
(also see p. 222, this Vo1.).
Cryopreservation of encapsulated somatic embryos/synthetic seed is yet
another viable proposition for the long-term conservation and exchange of
germplasm. The SEs are looked upon as "seeds" for plants which do not set seed,
set seed only in small quantities, or that are short-lived. Since in many plant
species they can now be produced in large numbers in bioreactors, it would be
rewarding to freeze them not only for the preservation of germplasm, but also for
large-scale propagation. This is an area which can be commercially exploited by
the seed industry.
For the international exchange of germplasm similar to the routine used of
meristem cultures of potato and cassava (Bajaj 1995), cultures ofSEs may also be
used. However, since SEs can be mass produced only in liquid media, in shake
cultures, or in bioreactors which require continuous aeration, they cannot be
transported without getting damaged. Thus, cryopreserved materiaVsynthetic
seed would be ideal for this purpose.
3 Summary
1. Since the discovery of somatic embryogenesis in carrot in 1958, there are now
over 300 plant species in which the phenomenon has been observed.
2. Somatic embryos are bipolar, in contrast to monopolar roots and shoots.
3. Ultrastructural studies have shown both a unicellular and multicellular origin
of somatic embryos.
4. Polyamines, i.e., putrescine, spermidine, and spermine, play an important
role in cell division and the induction of somatic embryogenesis.
5. Many of the molecular processes of embryogenesis have been established in
proembryos in the presence of auxin.
6. Isolation of genes specific for earlier stage of embryogenesis has been
attempted (Kawahara et al. 1992), and there are only small number of new
genes that are expressed at detectable levels during somatic embryo

development (Komamine et al. 1992).
7. Machine-vision technology has great promise for the automation of
bioreactor-based mass harvest of somatic embryos for commercial scaleup.
8. Somatic embryos have proved to be an excellent material for genetic
transformation and for the production of transgenic plants, specially in tree
species.
9. Micrografting of somatic embryos on young rootstocks may be employed in
cases where somatic embryos normally do not germinate into complete
plants.
10. Bioreactor production of somatic embryos in large numbers has enabled en
masse micropropagation of some tree species.
11. Synthetic seed encapsulated somatic embryos assume great importance in
crops where seeds are not produced and vegetative propagation is
cumbersome. An automated system capable of producing and sowing
70000 synthetic seeds per day has been developed (Sakamoto et al. 1992).
12. Fluid drilling of somatic embryo has been successful; the plants appeared
normal after transplanting under glasshouse conditions (Kitto et al. 1991).
13. Cryopreservation of somatic embryos/synthetic seed is a viable proposition
for the long-term conservation and exchange of germplasm.
References
Aguilar ME, Villalobos VM, Vasquez N (1992) Production of cocoa plants (Theobroma cacao L.) via
micrografting of somatic embryos. In Vitro 28: 15-19
Altman A, Nadel BL, Falash Z, Levin N (1990) Somatic embryogenesis in celery: induction, control
and change in polyamines and proteins. In: Nijkamp HJJ et al. (eds) Progress in plant cellular and
molecular biology. Kluwer, Dordrecht, p 454
Ammirato PV, Styer DJ (1985) Strategies for large-scale manipulation of somatic embryos in
suspension culture. In: Zaitlin M, Day P, Hollaender A (eds) Biotechnology in plant science.
Academic Press, New York, pp 161-178
Attree SM, Fowke LC (1993). Embryogeny of gymnosperms: advances in synthetic seed technology of
conifers. Plant Cell Tissue Organ Cult 35: 1-35
Backs-Hiisemann D, Reinert J (1970) Embryobildung durch isolierte Einzellen aus Gewebekulturen
von Daucus carota. Protoplasma 70: 49-60
Bagni N (1989) Polyamines in plant growth and development. In: Bachrach U, Heimer YM (eds) The
physiology ofpolyamines, vol.2. CRC Press, Boca Raton, p 107
Bajaj YPS (1976) Regeneration of plants from cell suspensions frozen at -20,-70, and -196 °C. Physiol
Plant 37: 263-268
Bajaj YPS (1986) In vitro preservation of genetic resources - techniques and problems. Int Symp
Nuclear techniques and in vitro culture for plant improvement. IAEAlFAO, Vienna 1985, pp
43-57
Bajaj YPS (1991 a) Automated micropropagation for en masse production of plants. In: Bajaj YPS (ed)
Biotechnology in agriculture and forestry, vol 17. High-tech and micropropagation I. Springer,
Berlin Heidelberg New York, pp 3-16
Bajaj YPS (1991 b) Storage and cryopreservation of in vitro cultures. In: Bajaj YPS (ed) Biotechnology
120 y.P.S. Bajaj
in agriculture and forestry, vol. 17. High-tech and micro propagation I. Springer, Berlin Heidelberg
New York, pp 361-381
Bajaj YPS (1995) Cryopreservation of germplasm of potato (Solanum tuberosum L.) and cassava
(Manihot esculenta Crantz). In: Bajaj YPS (ed) Biotechnology in agriculture and forestry, vol 32.
Cryopreservation of plant germplasm 1. Springer, Berlin Heidelberg New York (in press)
Bajaj YPS, Bopp M (1972) Growth and organ formation in Sinapis alba tissue cultures. Z
Pflanzenphysiol 66: 378-381
Bajaj YPS, Bopp M, Bajaj S (1973) Patterns of peroxidases and differentiation in Sinapis alba.
Bajaj YPS, Gosch G, Ottma M, Weber A, Griibler A (1978) Production of polyploid and aneuploid
plants from anthers and mesophyll protoplasts of Atropa belladonna and Nicotiana tabacum.
Indian J Exp Bioi 16: 947-953
Baker CM (1985) Synchronization and fluid sowing of carrot, Daucus caroia somatic embryos. MS
Thesis, Univ Florida, Gainesville
Baker SR, Yon RJ (1983) Characterization of ornithine carbamoyltransferase from cultured carrot
cells oflow embryogenic potential. Phytochemistry 22: 2171-2174
Bommineni VR, Chibbar RN, Dalta RSS, Tsang EWT (1993) Transformation of white spruce (Picea
glauca) somatic embryos by microprojectile bombardment. Plant Cell Rep 13: 17-23
BorkirdC, ChoiJH, Jin Z, FranzG, HatzopoulosP, Chorneau R, Bonas R, Pelegri F, SungZR (1988)
Developmental regulation of embryonic genes in plants. Proc Nat! Acad Sci USA 85: 6399-6403
Borthwick HA (1931) Development of the macrogametophyte and embryo of Daucus carota. Bot Gaz
92: 24-44
Botti C, Vasil IK (1984) Ontogeny of somatic embryos of Pennisetum americanum. II. In cultured
immature inflorescences. Can J Bot 62: 1629-1635
Brettell RIS, Wernicke W, Thomas E (1980) Embryogenesis from cultured immature inflorescences of
Sorghum bicolor. Protoplasma 104:141-148
Button J, Kochba J, Bornman CH (1974) Fine structures of and embryoid development from
embryogenic ovular callus of 'Shamouti' orange (Citrus sinensis Osb.). J Exp Bot 25: 446-457
Chen THH, Thompson BG, Gerson DF (1987) In vitro production of alfalfa somatic embryos in
fermentation systems. J Ferment Technol 65: 353-357
Cheng Z, Ling PP (1992) Computer vision for plant embryo quality evaluation. Am Soc Agri Eng
1992, Int Winter Meet, Pap 923575
Choi JH, Sung ZR (1984) Two-dimensional gel analysis of carrot somatic embryonic proteins. Plant
Mol Bioi Rep 2: 19-25
Choi JH, Liu LS, Borkird C, Sung ZR (1987) Isolation of cDNA clones for rare embryo-specific
antigens in carrot cell cultures. Proc Nat! Acad Sci USA 84: 1906--1910
De Vries SC, Booij H, Meyerink P, Huisman G, Wilde HD, Thomas TL, van Kammen A (1988)
Acquisition of embryogenic potential in carrot suspension cultures. Planta 176: 196--204
Denchev PD, Kuklin AI, Scragg AH (1992) Somatic embryo production in bioreactors. J Biotechnol
26: 99-109
Ducos JP, Bollon H, Pettiard V (1993) Production of carrot somatic embryos in a bioreactor. Appl
Microbiol Biotechnol 39: 465-470
Eksomtramage T, Paulet F, Guiderdoni E, Glaszmann JC, Engelmann F (1992) Development of a
cryopreservation process for embryogenic calluses of a commercial hybrid of sugarcane
(Saccharum sp.) and application to different varieties. Cryo-Lett 13: 239-252
EI Hadrami I, Michaux-Ferriere N, Carron MP, D'Auzac H (1989) Les polyamines, facteur limitant
possible de l' embryogenese somatique chez I'Hevea brasiliensis. CR Acad Sci Paris Ser III, 308:
205-211
Ellis DD (1993) Transformation in spruce (Picea species). In: Bajaj YPS (ed) Biotechnology in
agriculture and forestry, vol 23. Plant protoplasts and genetic engineering IV. Springer, Berlin
Heidelberg New York, pp 315-330
Ellis DD, McCabe DE, McInnis S, Ramachandran R, Russell DR, Wallace KM, Martinell BJ,
Roberts DR, Raffa KF, McCown BH (1993) Stable transformation of Picea glauca by particle
acceleration. Biorrechnology II: 84· 89
Engelmann F, Duval Y (1986) Cryopreservation d'embryons somatiques de palmier it huile (Elaeis
guineensis Jacq.): resultats et perspectives d'application. Oleagineaux 41: 169-173
Faure 0, Mengoli N, Nougarhde A, Bagni N (1991) Polyamine pattern and biosynthesis in zygotic and
somatic embryo stages of Vilis vinifera. J Plant Physiol138: 545-549

Fienberg AA, Choi JH, Lubich WP, Sung ZR (1984) Developmental regulation of polyamine
metabolism in growth and differentiation of carrot culture. Planta 162: 532-539
Fitch MM, Manshardit RM, Gonsalves 0, Slightom JL, Sanford JC (1990) Stable transformation of
papaya via microprojectile bombardment. Plant Cell Rep 9: 189-194
Fransz PF, Schel JHN (199Ia) Cytodifferentiation during the development of friable embryogenic
callus of maize (Zea mays). Can J Bot 69: 26-33
Fransz PF, Schel JHN (l99Ib) An ultrastructural study on the early development of Zea mays L.
somatic embryos. Can J Bot 69: 858-865
Fransz PF, Schel JHM (1994) Ultrastructural studies on callus development and somatic
embryogenesis in Zea mays L. In: Bajaj YPS (ed) Biotechnology in agriculture and forestry, vol 25.
Maize. Springer, Berlin Heidelberg New York, pp 50-65
Fuji JA, Slade 0, Aguirre-Rascon J, Redenbaugh K (1992) Field planting of alfalfa artificial seeds. In
Vitro Cell Dev Bioi 28: 73-80
Fujimara T, Komamine A (1979a) Synchronization of somatic embryogenesis in a carrot cell
suspension culture. Plant Physiol64: 162-164
Fujimara T, Komamine A (1979b) Involvement of endogenous auxin in somatic embryogenesis in a
carrot cell suspension culture. Z Pflanzenphysiol 95: 13-19
Giuliano G, Rosellini 0, Terzi M (1983) A new method for purification of the different stages of carrot
embryoids. Plant Cell Rep 2: 216-218
Gosch G, Bajaj YPS, Reinert J (1975) Isolation, culture and induction of embryogenesis in protoplasts
from cell suspensions of A tropa belladonna. Protoplasma 86: 405-410
Grand d'Esnon A, Chee R, Harrell RC, Cantliffe OJ (1989) Qualitative and quantitative evaluation of
liquid tissue cultures by artificial vision. Biofutur 76: S3
Gray 0 (1984) The role of fluid drilling in plant establishment. Aspects Appl Bioi 7: 153-172
Gray OJ, Conger BV, Songstad DO (1987) Desiccated quiescent somatic embryos of orchardgrass for
use as synthetic seeds. In Vitro Cell Dev Bioi 23: 29-33
Gray OJ, Purohit A (1991) Quiescence and dormancy in somatic embryos. In: Bajaj YPS (ed)
Biotechnology in agriculture and forestry, vol 17. High-tech and micropropagation 1. Springer,
Haccius B (1971) Zur derzeitigen Situation der Angiospermen-Embryologie. Bot Jahrb 91: 309
cell cultures. J Ultrastruc Res 18: 428-443
Hiimiiliiinen 11, Kurten U, Kauppinen V, Heilala J (1992) Automated classification of somatic
embryos. Acta Hortic 319: 601-605
Harrell RC, Bieniek M, Cantliffe OJ (1992) Noninvasive evaluation of somatic embryogenesis.
Biotech Bioeng 39: 378-383
Hatzopoulos P, Fong F, Sung ZR (1990) Abscisic acid regulation of De8, a carrot embryonic gene.
Heby 0 (1989) Polyamines and cell differentiation. In: Bachrach U, Heimer YM (eds) The physiology
ofpolyamines, vol 2. CRC Press, Boca Raton, p 83
Iriondo JM, Perez C (1991) Use of image analysis techniques for developmental studies in 'in vitro'
regenerated plants of Lavalera oblongifolia. Acta Hortic 289: 335-336
Johri BM, Bajaj YPS (1965) Growth responses of globular proembryos of Dendropthoefaleata (LJ.)
Ettings in culture. Phytomorphology 15: 292-300
Jonard R (1986) Micrografting and its applications to tree improvement. In: Bajaj YPS (ed)
Biotechnology in agriculture and forestry, vol I. Trees I. Springer, Berlin Heidelberg New York, pp
31-48
Jones TJ, Rost TL (1989) The developmental anatomy of somatic embryos from rice (Oryza sativa L.)
scutellum cells. Bot Gaz 150: 41-49
Kawahara R, Sunabori S, Fukuda H, Komamine A (1992) A gene expressed preferentially in the
globular stage of somatic embryogenesis encodes elongation factor a in carrot. Eur J Biochem 209:
157-162
Kermode AR, Bewely JD (1985) The role of maturation drying in the transition from seed development
to germination. J Exp Bot 36: 1906-1915
Khan AJ, Minocha SC (1991) Polyamines and somatic embryogenesis in carrot II. The effects of
cyclohexylammonium phosphate. J Plant Physiol 137: 446-452
122 Y.P.S. Bajaj
Kitto S, Janick J (1982) Polyox as an artificial seed coat for asexual embryos. HortSci 17: 448
Kitto S, Janick J (1985a) Production of synthetic seeds by encapsulating asexual embryos of carrot.
J Am Soc Hortic Sci 110: 227-282
Kitto S, Janick J (1985b) Hardening treatments increase survival of synthetically-coated asexual
embryos of carrot. J Am Soc Hort Sci 110: 283-286
Kitto SL, Pill WG, Molloy DM (1991) Fluid drilling as a delivery system for somatic embryo-derived
plantlets of carrot (Daucus carota L.). Hort Sci 47: 209-220
Kiyosue T, Satoh S, Kamada H, Harada H (1991) Purification and immunohistochemical detection of
an embryogenic cell protein in carrot. Plant Physiol95: 1077-1083
Komamine A, Kawahara R, Matsumoto M, Sunabori T, Toya T, Fujimura A, Tsukahara M, Smith
J, Iro M, Fukuda H, Nomura K, Fujimura T (1992) Mechanisms of somatic embryogenesis in cell
cultures: physiology, biochemistry, and molecular biology. In Vitro Cell Dev Bioi 28P: 11-14
Kurata K, Terada M, Komine M, Liyanaga KH (1991) Computer vision for selecting somatic
embryos. ASAE Pap. 91-3054
Lo Schiavo F, Fillippini F, Cozzani F, Vallone D, Terzi M (1991) Modulation of auxin binding
proteins in cell suspensions. I. Differential responses of carrot embryo cultures. Plant Physiol 97:
60-64
Margaritis A, Wallace JB (1984) Novel bioreactor systems and their applications. Biorrechnology 2:
447-453
Marin ML, Duran-Vila N (1988) Survival of somatic embryos and recovery of plants of sweet orange
(Citrus sinensis (L.) Osb.) after immersion in liquid nitrogen. Plant Cell Tissue Organ Cult 14:
51-57
Mathews H, Litz RE, Wilde HD, Merkle SA, Wetzstein HY (1992) Stable integration and expression
of b-glucuronidase and NPT II genes in mango somatic embryos. In Vitro Cell Dev Bioi 28:
172-178
McCain JW, Hodges TK (1986) Anatomy of somatic embryos from maize embryo cultures. Bot Gaz
147: 453-460
McDonald K, Jackman A (1989) Bioreactor studies of growth and nutrient utilization in alfalfa
suspension cultures. Plant Cell Rep 8: 455-458
McGranahan GH, Leslie CA, Uratsu SL, Dandekar AM (1990) Improved efficiency of walnut somatic
embryo gene transfer system. Plant Cell Rep 8: 512-516
McGranahan GH, Leslie CA, Dandekar AM, Uratsu SL, Yates IE (1993) Transformation of pecan
and regeneration of transgenic plants. Plant Cell Rep 12: 634--638
McKersie BD, Senaratna T, Bowley SR, Brown DCW, Krochko E, Bewley JD (1989) Application of
artificial seed technology in the production of hybrid alfalfa (Medicago sativa L.). In Vitro Cell
Dev Bioi 25:1183-1188
McWilliam AA, Smith SM, Street HE (1974) The origin and development of embryoids in suspension
cultures of carrot (Daucus carota). Ann Bot 38: 234--250
Meijer EGM, Simmonds J (1988) Polyamine levels in relation to growth and somatic embryogenesis in
tissue culture of Medicago sativa. J Exp Bot 39: 787-794
Mengoli M, Bagni N (1992) Polyamines and somatic embryogenesis in higher plants. IAPTC Newslett
68: 2-8
Merkle SA, Schlarbaum SE, Cox RA, Schwarz OJ (1991) Mass propagation of somatic embryo-
derived plantlets of yellow-poplar for field testing. In: Proc 21st Southern Conf Forest tree
improvement, Knoxville, pp 56-68
Minocha SC, Khan AJ (1991) Effects of combinations of polyamine biosynthetic inhibitors on cellular
polyamines in carrot cell cultures. J Plant Physiol137: 507-510
Minocha SC, Papa NS, Khan AJ, Samuelsen AI (1991a) Polyamines and somatic embryogenesis in
carrot. III. Effects of methylglyoxal bis (guanylhydrazone). Plant Cell Physiol32: 395-402
Minocha SC, Minocha R, Komamine A (1991 b) The effects of polyamine biosynthesis inhibitors on S-
adenosylmethionine synthetase and S-adenosylmethionine decaboxylase activities in carrot cell
cultures. Plant Physiol Biochem 29: 231-237
Montague MJ, Koppenbrink JW, Jaworski EG (1978) Polyamine metabolism in embryogenic cells of
Daucus carota.I. Changes in intracellular content and rates of synthesis. Plant Physiol62: 430-433
Mullins MG, Archie Tang FC, Facciotti D (1990) Agrobacterium-mediated genetic transformation of
grapevines: transgenic plants of Vitis rupestris Scheele and buds of Vitis vinifera. Bio/Technology
8: 1041-1045
Murashige T (1978) The impact of plant tissue culture on agriculture. In: Thorpe T (ed) Frontiers of
plant tissue culture 1978. IAPTC, Univ Calgary, Alberta, pp 15-26
Nagl W (1990) Translocation of putrescine in the ovule, suspensor and embryo of Phaseolus coccineus.
J Plant Physiol136: 587-591
Navarro L (1992) Citrus shoot tip grafting in vitro. In: Bajaj YPS (ed) Biotechnology in agriculture
and forestry, vol 18. High-tech and micropropagation II. Springer, Berlin Heidelberg New York,
pp 327-338
Neuhaus G, Spangenberg G, Mittelsten Scheid 0, Schweiger HG (1987) Transgenic rapeseed plants
obtained by the microinjection of DNA into microspore-derived embryoids. Theor Appl Genet 75:
30-36
Nissen P, Minocha SC (1993) Inhibition by 2,4-D of somatic embryogenesis in carrot as explored by
its reversal by difluoromethylomithine. Physiol Plant 89: 673-680
Nomura K, Komamine A (1985) Identification and isolation of single cells that produce somatic
embryos at high rate in carrot suspension cultures. Plant Physiol79: 988-991
Novak FJ, Donini B, Owusu GK (1986) Somatic embryogenesis and in vitro plant development of
cocoa. In: Proc Int Symp Nuclear techniques and in vitro culture for plant improvement. IAEAI
FAO, Vienna, pp 443-449
Nuti Ronchi V, Giorgetti L, Tonelli MG, Martini G (1992) Ploidy reduction and genome segregation
in cultured carrot cell lines. II. Somatic meiosis. Plant Cell Tissue Organ Cult 30: 115-119
Olofsdotter M (1993) Image processing: a non-destructive method for measuring growth in cell and
tissue culture. Plant Cell Rep 12: 216-219
Pegg AE (1986) Recent advances in the biochemistry of polyamines in eukaryotes. Biochem J 234:
249-262
Pill WG (1991) Advances in fluid drilling. HortTechnology I: 59-65
Preil W, Florek P, Wix U, Beck A (1988) Towards mass propagation by use of bioreactors. Acta
Hortic 226: 99-106
Redenbaugh K (ed) (1992) Synseeds: applications of synthetic seeds to crop improvement. CRC Press,
Boca Raton
Redenbaugh K, Nichol J, Kossler M, Paasch B (1984) Encapsulation of somatic embryos for artificial
seed production. In Vitro 3: 256-257
Redenbaugh K, Paasch B, Nichol J, Kossler M, Viss P, Walker K (1986) Somatic seeds: encapsulation
of asexual plant embryos. Biorrechnology 4: 797-801
Redenbaugh K, Slade D, Viss P, Fujii J (1987) Encapsulation of somatic embryos in synthetic seed
coats. Hort Sci 22: 803-809
Redenbaugh K, Fujii J, Slade D, Viss P, Kossler M (1991) Artificial seeds: encapsulated somatic
embryos. In: Bajaj YPS (ed) Biotechnology in agriculture and forestry, vol 17. High-tech and
micropropagation I. Springer, Berlin Heidelberg New York, pp 395-416
Reinert J (1958) Untersuchungen iiber die Morphogenese an Gewbekulturen. Ber Dtsch Bot Ges 71: I 5
Reinert J (1959) Uberdie Kontrolle der Morphogenese and die Induktion von Adventivembryonen an
Gewebekulturen aus Karotten. Planta 53: 318-333
Reinert J, Bajaj YPS, Zbell B (1977) Aspects of organization - organogenesis, embryogenesis,
cytodifferentiation. In: Street HE (ed) Plant tissue and cell culture. Blackwell, London, pp 389-427
Ritterhaus E, Ulrich J, Westphal K (1990) Large-scale production of plant cell cultures. IAPTC
Newslett 61: 2-10
Robie CA, Minocha SC (1989) Polyamines and somatic embryogenesis in carrot I. The effects of
difluoromethylomithine and difluoromethylarginine. Plant Sci 65: 45-54
Roustan JP, Latche A, Fallot J (1989) Stimulation of Daucus carota somatic embryogenesis by
inhibitors of ethylene synthesis: cobalt and nickel. Plant Cell Rep 8: 182-185
Roustan JP, Latche A, Fallot J (1992) Influence of ethylene on the incorporation of 3,4 C4C)
methionine into polyamines in Daucus caTota cells during somatic embryogenesis. Plant Physiol
Biochem 30: 201-205
Sakamoto Y, Mashiko T, Suzuki A, Kawata H, Iwasaki A (1992) Development of encapsulation
technology for synthetic seeds. Acta Hortic 319: 71-76
Salter PJ (1978) Fluid drilling of pregerminated seeds: progress and possibilities. Acta Hortic 33:
245-249
Sato S, Newell C, Kolacz K, Tredo L, Finer J, Hinchee M (1993) Stable transformation via particle
bombardment in two different soybean systems. Plant Cell Rep 12: 408-413
124 y'P.S. Bajaj
Schnall JA, Hwang CH, Cooke TJ, Zimmerman JL (1991) An evaluation of gene expression during
somatic embryogenesis of two temperature sensitive carrot-variants unable to complete embryo
development. Physiol Plant 82: 498-504
Schultheis JR, Cantliffe DJ (1992) Growth of somatic embryos of sweet potato (Ipomoea batatas (L.)
Lam) in hydroxyethyl cellulose gel amended with salts and carbohydrates. Sci Hortic 50: 21-33
Seffens WS, Almoguera C, Wilde HD, Von der Haar RA, Thomas TL (1990) Molecular analysis ofa
phylogenetically conserved carrot gene: developmental and environmental regulation. Dev Genet
II: 65-76
Senaratna T (1992) Artificial seeds. Biotech Adv 10: 379-392
Senaratna T, McKersie BD, Bowley SR (1990) Artificial seeds of alfalfa (Medicago sativa L.).
Induction of desiccation tolerance in somatic embryos. In Vitro Cell Dev BioI 16: 85-90
Slocum RD, Flores HE (1991) Biochemistry and physiology ofpolyamines in plants. CRC Press, Boca
Raton
Smart NJ, Flower MW (1984) An airlift column bioreactor suitable for large-scale cultivation of plant
cell suspensions. J Exp Bot 35: 531-535
Smith DR, Donaldson LA, Warr A (1991) Bioreactor technology for mass propagation of Pinus
radiata by somatic embryogenesis. In Vitro 27:43A
Smith MAL, Spomer LA (1987) Direct quantification on in vitro cell growth through image analysis.
In Vitro Cell Dev BioI 23: 67-74
Smith MAL, Spomer LA, Meyer MJ, McClelland MT (1989) Non-invasive evaluation of growth
during plant micropropagation. Plant Cell Tissue Organ Cult 19: 91-102
Smith T A (1990) Plant polyamines - metabolism and function. In: HE Flores et al. (eds) Polyamines
and ethylene: biochemistry, physiology and interactions. Am Soc Plant Physiol, pp 1-17
Spomer LA, Smith MAL (1988) Image analysis for biological research: camera influence on
measurement accuracy. Intelligent Instruments and Computers, July/Aug 1988, 11: 201-216
Stacey NJ, Roberts K, Knox JP (1990) Patterns of expression of the HIM4 arabinogalactan-protein
epitope in cell cultures and during somatic embryogenesis in Daucus carata L. Planta 180: 285-292
Steward FC, Mapes MO, Smith J (1958) Growth and organized development of cultured cells. 1.
Growth and division of freely suspended cells. Am J Bot 45: 693-703
Street HE, Withers L (1974) The anatomy of embryogenesis in culture. In: Street HE (ed) Tissue culture
and plant science. Academic Press London, pp 71-100
Stuart DA, Strickland SG, Walker KA (1987) Bioreactorproduction of alfalfa somatic embryos. Hort
Sci 22: 800-803
Styler DJ (1985) Bioreactor technology for plant propagation. Basic Life Sci 32: 117-130
Sudarmonowati E, Henshaw GG (1990) Cryopreservation of cassava somatic embryos and
embryogenic tissue. Abstr Viith Int Congr Plant tissue and cell culture, Amsterdam, pp 378
USA 78: 3683-3687
Tanaka H, Nishijima F, Suwa M, Iwamoto T (1983) Rotating drum fermenter for plant cell suspension
cultures. Biotech Bioeng 25: 2359-2370
Tautorus TE, Lulsdorf MM, Kikcio SI, Dunstan DI (1992) Bioreactor culture of black spruce and
interior spruce somatic embryos. Abstr Proc 3rd Can Worksh Plant tissue culture and genetic
engineering, Univ Guelph, p 36
Tay LF, Khoh LK, Loh CS (1993) Alginate-chitosan conservation in production of artificial seeds.
Tessereau H, Lecouteux C, Florin B, Schlienger C, Petiard V (1991) Use of simplified freezing process
and dehydration for the storage of embryogenic cell lines and somatic embryos. Rev Cytol BioI Veg
Bot 14: 297-310
Thomas E, Konar RN, Street HE (1972) The fine structure of the embryogenic callus of Ranunculus
sceleratus L. J Cell Sci II: 95-109
Timmers ACJ, De Vries SC, Schel JHN (1989) Distribution of membrane-bound calicum and activated
calmodulin during somatic embryogenesis of carrot (Daucus carata L.). Protoplasma 153: 24-29
Trigiano RN, Gray DJ, Conger BV, McDaniels JK (1989) Origin of direct somatic embryos from
cultured leaf segments of Dactylis glomerata. Bot Gaz 150: 72-77
Ulrich B, Wiesner W, Arens H (1985) Large-scale production of rosmarinic acid from plant cell
cultures of Coleus blumei Benth. In: Neumann KH, Barz W, Reinhard E (eds) Primary and
secondary metabolism of plant cell cultures. Springer, Berlin Heidelberg New York, pp 293-303
Uragami A, Sakai A, Nagai M, Takahashi T (1989) Survival of cultured cells and somatic embryos of
Asparagus offieinalis cryoprcserved by vitrification. Plant Cell Rep 8: 418-421
Ushiyama K, Miyamoto Y, Oda H, Ishida Y (1984) Gaseous phase culture equipment for continuous
culture of plant tissues. Jpn Patent Kokai Tokayo Koho 59-45873
Vasil V, Lu C, Vasil IK (1985) Histology of somatic-embryogenesis in cultured immature embryos of
maize (Zea mays L.). Protoplasma 127: 1-8
Wake H, Akasaka A, Umetsu H, Ozeki Y, Shimomura K, Matsunaga T (1992) Enhanced germination
of artificial seeds by marine cyanobacterial extract. Appl Microbiol Biotechnol 36: 684-688
Wernicke W, Potrykus I, Thomas E (1982) Morphogenesis from cultured leaf tissue of Sorghum
hieolor. The morphogenetic pathways. Protoplasma III: 53-62
Wilde HD, Nelson WS, Booiji H (1988) Gene-expression programs in embryogenic and non-
embryogenic carrot cultures. Planta 176: 205-211
Wilde HD, Meagher RB, Merkle SA (1992) Expression of foreign genes in transgenic yellow-poplar
plants. Physiol 98: 114-120
Wilde HD, Merkle SA (1994) Genetic transformation of Liriodendron tulipifera L. (yellow poplar). In:
Bajaj YPS (ed) Biotechnology in agriculture and forestry, vol 29. Plant protoplasts and genetic
engineering V. Springer, Berlin Heidelberg New York (in press)
11.2 Somatic Embryogenesis and the Technology
of Synthetic Seed*
D.J. GRAY!, M.E. COMPTON!, R.C. HARRELL2 , and D.J. CANTLIFFE3
1 Introduction
Synthetic seeds are functionally defined as somatic embryos engineered to be of

use in commercial plant production (Gray 1990b). The actual form of synthetic
seed (i.e., presence or absence of a synthetic seed coat, whether they are hydrated
or dehydrated, quiescent or not, etc.) may vary depending on the specific crop
application.
The first reference to the potential use of somatic embryos for propagation is
generally credited to Murashige (1978) and efforts to engineer them into synthetic
seed have been ongoing ever since (e.g., Gray et al. 1984, 1992; Kitto and Janick
1985a,b,c; Redenbaugh et al. 1986, 1987a,b, 1988, 1991, 1994; Fujii et al.
1987a,b, 1989; Gray 1987a, 1990b; Gray and Mortensen 1987; Stuart et al. 1987;
Carman 1989; Janick et al. 1989; Kim and Janick 1989a,b; McKersie et al. 1989;
Redenbaugh 1990; Senaratna et al. 1990; Gray and Purohit 199Ia,b; Parrott
et al. 1991; Gray and Compton 1993; Senaratna 1992; Attree and Fowke 1993).
However, basic developmental mechanisms that contribute to the desirability of
seed, such as onset of quiescence and ability to withstand dehydration, are either
missing in all but a few embryogenic systems or have been simply overlooked
(Gray 1990b).
* This is an updated version of a previous review (Gray and Purohit 1991a). Partial support for this
project was provided by the International Board for Plant Genetic Resources (IBPGR) and the
Program in Science and Technology Cooperation, Office of the Science Advisor, US Agency for
International Development under Grant No. DHR-5600-G-00-0057-00. Florida Agricultural
Experiment Station Journal Series No. R-03398.
I Central Florida Research and Education Center, University of Florida, 5336 University Avenue,
Leesburg, FL 34748-8203, USA
2 Department of Agricultural Engineering, University of Florida, 9 Frazier Rogers Hall, Gainesville,
FL 32611-0361, USA
J Department of Horticultural Sciences, University of Florida, 1251 Fifield Bldg., Gainesville, FL
32611-0514, USA

Somatic Embryogenesis and the Technology of Synthetic Seed 127
2 Somatic and Zygotic Embryo Development
2.1 Developmental Morphology
Somatic and zygotic embryos share similar gross ontogenies, with both typically
passing through globular, torpedo, and cotyledonary stages for dicots (e.g.,
Ammirato 1987; Gray and Mortensen 1987) and conifers (Becwar et al. 1989), or
globular, scutellar, and coleoptilar stages for monocots (e.g., Conger et al. 1983;
Gray and Conger 1985b,c). However, significant differences exist that currently
limit the use of somatic embryos for propagation. Monoembryonic species that
produce zygotic embryos with a discrete filiform suspensor (such as most seed-
propagated crops) often produce clusters of somatic embryos from a mass of
embryonic tissue termed a pro embryonal cell complex (Haccius 1978). This basic
change in developmental pattern is likely due to differences between the in vivo
(seed) and in vitro environments since immature zygotic embryos dissected from
seeds and cultured often develop abnormally (Norstog 1965; Norstog and Klein
1972). Soma tic embryos growing from pro embryonal complexes tend to develop
asynchronously so that several stages are present in cultures at any given time.
Such somatic embryos initiated at different times are subjected to changing
nutrient regimes as medium becomes depleted than replenished between and
during subcultures. This leads to relatively extreme differences in development,
even among embryos from a single culture. With these variable and nonregulated
environmental conditions, somatic embryos often bypass maturation altogether,
becoming disorganized, forming new embryogenic cells and contributing to
asynchrony (Conger et al. 1989). Among embryos that grow to a relatively
mature stage, embryonic organs may develop at different rates. This leads to
precocious germination, in which only a root or a shoot is typically formed but
normal rapid germination and growth to a plant do not occur. Somatic embryos
also often exhibit structural anomalies such as extra cotyledons and poorly
developed apical meristems (Ammirato 1987). These problems appear to be due
to culture conditions and not factors intrinsic to somatic embryos since immature
zygotic embryos exhibit similar irregularities when removed from seed and
allowed to develop in vitro (Norstog 1965). Lack of synchronous cultures that
produce uniformly mature somatic embryos is a serious obstacle to propagation,
since embryo production systems that approach the uniformity of commercial
seed production will be required for many applications (Haccius 1978; Gray
1987c). Methods to remedy some of these problems by regulating embryo
development were discussed previously (Gray and Purohit 199Ia).
2.2 Functions of Nutritive and Seed Coat Tissues in Seeds
Tissues that enclose embryos in seed can be broadly distinguished as having

either a nutritive or protective function. A major difference between somatic and
zygotic embryos is that the former develop naked without these tissues. Beyond
protection and nutrition, such tissues create barriers to regulate gas exchange
128 DJ. Gray et al.
(Nikolaeva and Katkevich 1961). For example, oxygen availability, which is

limited by these tissues, influences embryo respiration (Nikolaeva and Katkevich
1961; Nutbeam and Duffus 1978), regulates embryo development, and prevents
precocious germination (Norstog and Klein 1972; Carman 1988). In contrast, in
embryogenic cell culture, availability of oxygen and other gases depends on the
subculture interval, whether or not the vessel is tightly sealed as well as gas usage
and evolution by cultured tissues. Studies to regulate oxygen availability in wheat
(Triticum aestivum L.) cell culture demonstrated that a low oxygen concentration
promoted embryogenic callus formation and suppressed nonembryogenic callus
(Carman 1989). Since immature zygotic embryos that are removed from seed
often do not develop normally, it can be inferred that the presence of seed tissues
is a contributing factor to normal embryo development. The effects of nutritive
and seed coat tissues were previously considered in greater depth (Gray and
Purohit 1991a).
2.3 Quiescence and Dormancy
Perhaps the most significant developmental difference between zygotic and

somatic embryos is that somatic embryos lack a quiescent resting phase (Gray
1987b). Generally, the resting phase is categorized as either "quiescent" or
"dormant" (Bewley and Black 1985). Quiescence is a resting phase that can be
reversed solely by the addition of water. Dormancy is a form of quiescence that
requires factors in addition to water, such as cold or heat treatments, for
resumption of growth to occur.
2.3.1 Desiccation Tolerance
Dehydration is of particular importance in onset and regulation of the resting

phase. The embryo (and seed) is able to withstand dehydration only during
a specific phase of development, termed the desiccation-tolerant stage (Gray
1987b; Senaratna et al. 1987,1990; Koster and Leopold 1988; Neumann et al.
1989). Prior to dehydration, metabolism is oriented toward accumulation of
storage compounds, whereas after rehydration, storage reserves are consumed
for germination. Lack of such developmental phases in conventional embryo-
genic culture systems may explain typical poor germination of somatic embryos.
Physiological studies show that dehydration can dramatically alter biosynthetic
pathways in somatic embryos. For example, Saranga et al. (1992a,b) showed
that treatments with various osmotica resulted in increased endogenous proline
levels and increased proline was associated with improved desiccation tolerance.
Similarly, treatment of white spruce [Picea glauca (Moench) Voss] somatic
embryos with ABA and polyethylene glycol, the latter of which functioned as a
non permeating osmoticum, increased that amount of triacylglycerol ninefold
(Attree et al. 1992). The same treatments that increased triacylglycerol also
induced desiccation tolerance.
2.3.2 Quiescence in Somatic Embryos
Several studies with carrot (Daucus carota L.) suggested that a resting phase
could be induced in somatic embryos by drying (Jones 1974; Nitzsche 1978; Kitto
and Janick 1985c). However, actual plant recovery from dehydrated somatic
embryos was first documented in orchardgrass (Gray and Conger 1985a; Gray
et al. 1987). In that study, somatic embryos were exposed to 70% RH air and
stored for up to 21 days at 23°C. Under these conditions embryos dehydrated
rapidly, became discolored, decreased in size, and their outer cell walls collapsed.
Embryo water content dropped from 83 to 13% within 24 h and was maintained
over the entire storage period. This water content is similar to that of seeds
maintained at 70% RH and is adequate for maintenance of viability during
prolonged seed storage. The embryos were rehydrated by placement on solidified
medium, during which they swelled rapidly, regained their normal white
coloration, and germinated. However, only well-developed (possessing a visible
scutellum and coleoptile), white, opaque somatic embryos were responsive. Such
somatic embryos were structurally mature and contained starch and lipid storage
compounds (Gray and Conger 1985b). When these embryos were stored in a
dehydrated state for 21 days, 4% germinated and produced plants after
imbibition (Table 1). The occurrence of quiescence in somatic embryos was
demonstrated by this study.
Grape somatic embryos were subsequently dehydrated in the method
described for orchardgrass and quiescence was induced (Gray 1987b,d, 1989).
During dehydration, grape somatic embryos underwent morphological changes
similar to those of orchardgrass. Their water content equilibrated to 13% when
stored at 70% RH and they resumed a normal appearance after rehydration.
Genotypic differences in response were noted; those that produced relatively
well-developed somatic embryos were most responsive. After 21 days of
dehydrated storage, 34% of embryos from one grape genotype produced plants
following imbibition (Table 2). This study demonstrated that higher germination
percentages were possible. Grape is discussed further in the next section in
relation to dormancy.
Table 1. Gennination of dehydrated orchardgrass somatic embryos after storage at 23°C for 0.7. and
21 days'. (Gray et al. 1987)
Response b Days of dehydrated storage
o 7 21
No gennination 126/28' 333/74 396/88

Gennination - no further growth 180/40 81118 36118
Gennination - viable plants 144/32 36/8 18/4
Total 4501100 4501100 4501100
aEmbryos were imbibed on solidified medium after test storage periods.

bEmbryos that produced root hairs, roots, coleoptiles, and/or shoots but failed to develop further were
scored as genninated - no further growth. Those that produced green leaves and continued to grow
were considered to be viable.
'Number/percentage of total.
130 OJ. Gray et al.
Table 2. Comparison of dehydration and benzyladenine (BA) for inducing germination in grape
somatic embryos." (Gray 1989)
Treatment Percent germination response b
Hypocotyl Root Cotyledon Shoot
Dehydration 77 68 65 34
0.5 liM BA 100 92 98 12
Control 76 88 36 0
"Well-developed embryos were either dehydrated for 21 days at 70% RH and 27°C, placed directly on
medium with BA, or placed on basal medium (control).
b Germination response was based upon either enlargement and greening of hypocotyls and
cotyledons or emergence of roots or shoots.
Quiescence also has been observed in somatic embryos of com (Zea mays L.)
by Compton et al. (1992), who showed that plants could be obtained from
previously dried embryos, as well as in soybean (Glycine max Merr.) (Parrott
et al. 1988). For soybean, plant formation from somatic embryos typically was
spread over a period of 9 months. However, after 2 weeks, when results from
seven genotypes were pooled, plant regeneration averaged 60% from embryos
that were dehydrated compared with 1% for embryos that were not dehydrated.
This study differed from the preceding dehydration studies in that it included a
maturation period of 28 days.
The longevity record for survival of dehydrated somatic embryos was
reported for alfalfa (McKersie et al. 1989). When high quality embryos were
selected and treated with ABA before drying, 60% produced plants after 1 year of
dried storage (Senaratna et al. 1990). Ability of the alfalfa embryos to withstand
dehydration was due, in part, to ABA pretreatments, which induced a stage of
desiccation tolerance as described in the previous section (Senaratna et al. 1987,
1989a,b). Provision of desiccation tolerance to somatic embryos clearly demon-
strates that they are capable of entering complex developmental pathways
normally associated with those of seed embryos when proper environmental
conditions are provided.
Recently, methodology to induce quiescence in somatic embryos
of white spruce was reported by Attree et al. (1991,1992). A nonpermeating
osmoticum, such as polyethylene glycol, was required to induce desiccation
tolerance, in addition to ABA. Desiccation tolerance did not occur in the absence
of ABA or with permeating osmotica, such as sucrose. Gradual drying to about
29% water, with a total storage time of 14 days, allowed up to 81 % of the embryos
to germinate into plants after imbibition. This represents a significant
advancement in light of the useful applications of synthetic seed technology in
conifer improvement (see later discussion).
2.3.3 Dormancy in Somatic Embryos
Embryogenic cultures of grape represent a good example of an in vitro dormancy

system since somatic embryos become well developed but still germinate poorly.
Dormancy of grape seed can be alleviated by cold stratification, after which

germination occurs (Flemion 1937). Somatic embryos of grape are similar to seed
in that they germinate better following cold treatments (Gray and Mortensen
1987). This suggests that the somatic embryos also are dormant. Studies of ABA
concentration in embryogenic cultures showed a rapid increase during embryo
development, which reached a peak at maturation (Rajasekaran et al. 1982).
Cold stratification of somatic embryos resulted in a rapid decrease in ABA.
Exogenously supplied ABA inhibited somatic embryo germination (Gray 1989).
Because ABA is implicated as a controlling factor of dormancy in many types of
seeds (e.g., Walton 1980; Bewley and Black 1985; Thevenot et al. 1987), it is
plausible to consider that ABA also functions in grape somatic embryos. In
contrast to ABA, exogenously supplied gibberellin (GA3) caused grape somatic
embryos to germinate and the concentration of endogenous GA-like compounds
increased during cold stratification (Pearce et al. 1987). This suggests a simple
endogenous control of embryogenesis and germination whereby ABA inhibits
precocious germination and thus promotes normal development while GA
causes germination to occur. Other examples of dormancy in somatic embryos
were previously discussed (Gray and Purohit 1991a,b).
3 Genetic Variation from Cell Culture
Somaclonal variation has been documented for a number of culture systems

(Redenbaugh et al. 1988; Bajaj 1990). In instances where somaclonal variation is
present, inability to control it would invalidate the basic rationale behind
synthetic seed technology. Rather extreme examples of this phenomenon include
sterility arising in plants regenerated from long-term embryogenic cultures of
carrot (Sussex and Frei 1968) as well as albinism in somatic embryo-derived
plants ofbromegrass (Bromus inermis L.) (Gamborg et al. 1970). Certain species
differ in tendency to yield culture-induced genetic changes than others. For
example, embryogenic response can be intentionally intensified in orchardgrass
by recurrent cycles of culture initiation and plant regeneration, and the increased
embryogenic response appears to be heritable (Hanning and Conger 1986).
Conversely, repeated attempts to select increased embryogenic response in grape
have been unsuccessful (Gray unpubl.). Further, two other studies have found
very little variation in regenerated plants of barley (Hordeum vulgare L.) and rice
(Oryza sativa L.) (Ogura et al. 1987; Luckett et al. 1989).
Recently, the genetic stability of somatic embryo-derived poinsettia plants
(Euphorbia pulcherrima Willd. ex Klotzsch) was compared with that of the
embryogenic cells from callus, Erlenmeyer flask, and bioreactor cultures from
which they were regenerated (Geier et al. 1992). Whereas the progenitor
embryogenic cell cultures became increasingly unstable over time, with nuclear
DNA contents ranging from 2C to 32C, the regenerated plants exhibited high
uniformity in both ploidy level and morphology. These results suggest that the
potential to form somatic embryos becomes impaired in cells with genetic
abnormalities. These studies strongly indicate that the process of embryogenesis,
132 D.J. Gray et al.
at least for some species, is remarkably stable and resistant to random genetic
change.
Possibilities for suppressing variation, when present, include frequent
initiation of new cultures (Krikorian 1982) as well as the use of short culture
growth intervals (Evans and Gamborg 1982), both of which reportedly minimize
instability. Producing somatic embryos directly from other embryos without an
intervening callus stage was used to reduce variation in oil palm (Elaeis guineensis
Jacq.) (Brackpool et al. 1986). Despite the obvious importance of managing
genetic variation in synthetic seed, it may be premature to anticipate problems
before suitable prototype systems are in place.
4 Structural Aspects of Synthetic Seed
Synthetic seed will consist of either a quiescent or nonquiescent somatic embryo

with or without a protective encapsulation. The exact form of synthetic seed that
is required will depend upon its specific applications. Naked, nonquiescent
somatic embryos, germinated in soil plugs, could be used to propagate certain
ornamental crops that are now laboriously micropropagated by tissue culture
(Table 3). Manpower reduction achieved by producing plants by somatic
embryogenesis, when compared to existing micropropagation, would confer a
cost advantage. Dehydrated, quiescent somatic embryos without encapsulation
would be useful for germplasm storage since they can be hand manipulated
and carefully stored in protective containers. Cost of manipulating somatic
embryos for germplasm storage would be similar to that of seed. Nonquiescent
encapsulated embryos could be useful for crops that are first grown in green-
houses before transplanting to the field, such as carrot and celery (Fujii et al.
1987a). However, for mass propagation of field crops, a protective encapsulation
will be necessary.
Somatic embryo encapsulations are analogous to the seed coat and endo-
sperm of normal seed. Encapsulations may provide physical protection to the
somatic embryo and carry nutrients, growth regulators, antibiotics, fungicides,
etc. to assist in germination and plant survival (Kitto and Janick 1985b,c;
Redenbaugh et al. 1986,1988; Janick et al. 1989). Encapsulated somatic embryos
could conceivably be handled as seed using conventional planting equipment.
Both hydrated and dry encapsulations have been envisioned.
Table 3. Types of synthetic seed and crop applications
Synthetic seed type Relative develop- Relative cost Example of crop

ment cost per seed application
Naked embryo
Nonquiescent Low High Ornamentals
Quiescent (dried) Low High Germplasm conservation
Encapsulated embryo
Nonquiescent Medium High Vegetable transplants
Quiescent (dried) High Low Agronomic and conifers
Nonquiescent somatic embryos placed in a hydrated encapsulation may be

cost effective for certain field crops that pass through a greenhouse transplant
stage such as celery (Apium graveolens L.) (Redenbaugh et al. 1986). A hydrated
encapsulation, termed as hydrogel, that could provide protection and allow
convenient handling has been developed (Redenbaugh et al. 1986, 1987a,b).
4.1 Delivery Methods for Somatic Embryos
To date, most embryo-to-plant systems require several intermediate steps prior

to successful plant establishment in the field. For example, Haydu and Vasil
(1981) germinated napier grass (Pennisetum purpureum Schum.) somatic
embryos in tissue cultures until they grew into plants 1-5 cm long. These small
plants were then transferred to culture tubes to establish a more vigorous root
system. The next step was potting the plantlets into soil-vermiculite in a growth
chamber, then acclimating the plants to lowerrelative humidities. Finally, plants
were transplanted in the greenhouse and/or the field.
For some species, plantlet formation is a very slow process. Celery somatic
embryos required 5 weeks before forming roots and leaves (Dunstan et al. 1982).
After adequate growth was obtained, the plants were transferred to soil in
propagation boxes in the greenhouse, then to individual pots. Spiegel-Roy and
Vardi (1984) developed a procedure for plant regeneration of citrus that required
several transfers from solid and liquid media. After plants reached a certain size,
they were cultured in tubes on paper bridges for further development prior to
transfer to soil. Sixteen to 18 weeks were required before plants obtained from
somatic embryos were growing in the greenhouse.
Extensive transfer steps were also required to establish plants from papaya
(Carica papaya L.) somatic embryos (Litz and Conover 1982). Embryos were
germinated and formed plants on White's medium supplemented with 0.1-2 mg/l
NAA and 0.05-0.2 mg/l BAP. Plants were then moved to a soilless potting mix
and hardened off under intermittent mist for 2-2.5 weeks. These few examples
illustrate the laborious and time-consuming task before plants from somatic
embryos could be transplanted in the field.
Several delivery methods for propagating somatic embryos directly from in
vitro conditions to the field or greenhouse have been proposed. They include: (1)
simultaneous dehydration of somatic embryos using a water-soluble resin and
planting in a wafer or seed tape (Kitto and Janick 1985a,b); (2) dehydration, then
planting dried somatic embryos with a conventional drill (Gray et al. 1987); (3)
encapsulation of singulated somatic embryos in an alginate gel capsule, then
planting using a conventional drill (Redenbaugh et al. 1984, 1986; Jeon et al.
1986); and (4) gel seeding somatic embryos with fluid drilling equipment (Drew
1979; Baker 1985; Schultheis et al. 1986a,b; 1990; Cantliffe et al. 1987).
4.1.1 Dehydration and Encapsulation in Polyethyleneoxide
Dehydration of somatic embryos has been investigated with the intent of

providing storage capability, and the ability to plant as normal dry seeds (Gray
134 D.l. Gray et al.
and Conger 1985a; Kitto and Janick 1985a,b,c; Gray 1987a,b; Gray et al. 1987).
Kitto and Janick (l985a,b) were the first to successfully dehydrate encapsulated
somatic embryos of carrot and have them survive upon rehydration. Unspecified
numbers of carrot embryos from suspension were incorporated with
polyethyleneoxide (Polyox), a water-soluble plastic resin, and dried to form
wafers under sterile conditions in a laminar flow hood at ambient temperatures
and humidities (Kitto and Janick 1985b). A constant weight was achieved after
6.5 h of drying. Three percent of the embryos survived encapsulation once
rehydrated, although it was unclear in the report as to whether plants formed.
Embryo survival decreased rapidly over drying time. No survival was obtained
when embryos were dried to a constant weight and not coated; however, some
embryos (fewer than 1%) survived after 32 h when they were coated with Polyox.
Hardening treatments, including 12% sucrose, chilling at 4°C, High inoculum
density, and/or 1 J..lM ABA prior to desiccation reportedly increased embryo
survival compared with the nontreated control (Kitto and Janick 1985c). The use
of a Polyox seed coat seemed to damage the embryos during the coating and
drying process as supported by low percentage survival data. Dehydration of
somatic embryos in Polyox does not appear to be a practical seeding method,
especially since coating somatic embryos singly was difficult and that no plants
formed after the coating process.
Recently, ABA-treated somatic embryos of carrot were successfully dried in
an alginate encapsulation. Up to 68% of the embryos germinated from capsules
that had been dehydrated by 92% after 10 days of storage (Liu et al. 1992).
4.1.2 Dehydration Prior to Encapsulation
Dehydration of somatic embryos as a separate step prior to encapsulation with

a coating material was previously proposed (Gray 1987b,c; Gray et al. 1987).
Somatic embryos of orchardgrass became quiescent when dried from 75 to 13%
water (Gray et al. 1987). Developmental stage before desiccation was critical for
embryo germination and plant formation. Only those embryos with well-
differentiated scutellar and coleoptilar regions prior to desiccation were able to
germinate and/or form plants after imbibition (see Sect. 2.3.1 for more details on
dehydration).
4.1.3 Encapsulation in Soft Gel Capsules
A third delivery method proposed was encapsulation of somatic embryos in a

soft gel capsule (Redenbaugh et al. 1984, 1986; Jeon et al. 1986). Several com-
pounds were tested for suitability for encapsulation, but calcium alginate, a food
thickener derived from brown algae, was selected due to its low embryo toxicity
and ease in forming capsules around the somatic embryo (Redenbaugh et al.
1986).
Gel capsules can be amended with nutrients, plant growth regulators, and
carbohydrates to facilitate the rapid growth and survival of embryos. Most
encapsulation research has dealt with alfalfa somatic embryos, which have
consistently averaged 60% plant formation under in vitro conditions and 20%
when grown in the greenhouse (Fujii et al. 1987a,b). The best plant recovery
frequency was 65% with celery somatic embryos when grown under in vitro
conditions utilizing carefully selected embryos.
Examples of other crops where encapsulated somatic embryos were shown
to form plants in vitro were Brassica species, carrot, cotton, lettuce, and corn
(Redenbaugh et al. 1987a). Although encapsulated somatic embryos have been
successfully converted to plants in several species, there are problems with the
alginate capsules. Water-soluble nutrients have been reported to rapidly leach
out of the capsule (Redenbaugh et al. 1987b). Respiration has been shown to be
reduced in encapsulated seeds due to poor gas exchange (Redenbaugh et al.
1993). Root and shoot emergence have also been inhibited. In addition,
Redenbaugh et al. (1987b) states that there is "lack of automation technology to
allow medium to large-scale production of single-embryo capsules." Finally,
storage may be a problem since embryo viability may decline over time due to
inhibition of embryo respiration in the capsule (Fujii et al. 1987a,b). This seeding
technology is relatively new and will require much more intensive study before it
can be used.
4.1.4 Fluid Drilling
Fluid drilling is the sowing of seeds in a protective flowable gel. The system was
developed in England during the 1970s (Gray 1981). The concept of fluid drilling
zygotic seed was first conceived by Currah et al. (1974), while Drew (1979) was
one of the first to suggest the idea of fluid drilling somatic embryos.
Actively growing or pregerminated embryos can be sown in the gel without
damage. The gel can also be amended with nutrients, carbohydrates, beneficial
microorganisms (mycorhizae, bacteria), and pesticides to stimulate embryo
growth and improve plant survival. The microenvironment created around the
seeds when fluid drilled can be modified and optimized such that the above
advantages can be realized (see Chap. 11.5, this VoL).
4.1.5 Gel Type
Due to their various chemical compositions, certain gels may be better carriers
for fluid drilling. For example, some gels may facilitate better oxygen movement,
some may contain nutrients which enhance plant growth, and others may hold
more moisture. Several gels have been used for drilling zygotic seed. Some of the
more commonly used gels are: Laponite (magnesium silicate clay), Liqua-gel
(potassium starch acrylamide), Planta-gel (copolymer of potassium acrylate
and acrylamide), Terrasorb (starch or synthetic copolymer), Hydrozorb 30
(potassium acrylate), and N-gel (various cellulose-based materials, formerly
called Natrosol).
A laboratory study by Brocklehurst (1979) determined that the polyacrylate
mineral colloid gel was superior to the natural gum gels since the latter caused
lettuce seedlings to senesce 4 days after planting. So sa-Coronel and Motes (1979)
evaluated seven commercial gels for planting carrot, onion, and pepper seed.
Pregerminated seeds were suspended in gels for 4 and 24 h prior to seeding in the
greenhouse. Superior seedling growth was obtained for all species sown with
Viterra 2, Natrosol250, and Bacto-Agar gels whereas Gum Blend and CLD were
toxic.
Oxygen movement is dependent on gel type and temperature. Frazier et al.
(1982) determined that oxygen movement in gels was slower as temperature
decreased. Slower oxygen movement in the gel could be critical since low
temperatures at planting might restrict the metabolism of the pregerminated
seed. Seed viability was related to oxygen diffusion in the gel. Of all the gels tested,
N atrosol had the highest oxygen diffusion and led to the most radicle emergence
and growth of snapdragon seedlings when suspended in gel. Pill and Fieldhouse
(1982) ran a similar comparison and also found that Natrosol was the only gel
that tomato seed could be stored in for more than 4 days without reducing the
percentage and rate of emergence. Bryan et al. (1982) observed slower plant
emergence and reduced pepper stands when germinated seeds were sown in
Laponite gel than without gel under wet growing conditions. Laponite has been
reported to reduce O 2 uptake by the seed (Brocklehurst 1979).
4.1.6 Protection/rom Damage and Desiccation
Gels will protect actively growing seeds with or without roots. In addition, gels
absorb water to many times their weight. This keeps moisture around the seeds
and reduces the possibility of desiccation. Gels may retain moisture around the
seeds several days after planting (Bryan et al. 1978).
4.1.7 Gel Additives
Ghate (1982) studied the rheological properties of gels when amended with
chemical additives. Little change occurred in the gel without additives after 1 day
when temperatures varied between 10 and 37°C. The addition offertilizer caused
gels to become thin, while relative consistency was maintained with the addition
of more gel.
Nutrient salts have been consistently added to gels at planting to enhance
germination and early seedling growth from zygotic embryos. Nitrogen,
phosphorus, and potassium were added to guar gum gel with little effect on
lettuce or cabbage growth (Costigan and Locascio 1982). Finch-Savage and Cox
(1982) reported improved early seedling growth of carrots with the addition of
monosodium phosphate at concentrations up to 30 gil, but nitrate (5-20gll)
reduced seedling growth. In a similar study, Finch-Savage and Cox (1983)
obtained earlier maturity in lettuce and onion when pregerminated seeds were
fluid drilled in gel supplemented with either sodium phosphate or ammonium
nitrate compared with nontreated, dry sown seed. No benefit was obtained when
less than 10 gil ammonium nitrate was incorporated in the gel.
Faster, more complete germination of pepper seeds was reported when

cytokinins, humic acids, or alpha-keto acids were added to the gel and planted
under wet environmental conditions (Bryan et al. 1982). The addition of GA3
and GA4I7 to gel at 2 and 2.5 mg/l increased the growth of tomato seedlings
compared with gel without GA (Ohep and Cantliffe 1980). Tomato plants were
taller when the gel was supplemented with either Cytex or Cytozyme, while
increased plant height, dry weight, and leaf area under field conditions were
obtained when humic acids were incorporated. GA and diphenamide hastened
the emergence rates of pepper by 1 to 3 days (Ghate and Phatak 1983).
Various pesticides have improved plant establishment. Ohep et al. (1984)
reduced damping off of tomato seedlings by incorporating ethazol plus
thiophanatemethyl, fenaminosulf, and ethazol alone or combined with benomyl,
chloroneb, or captan into the gel. White (1979) controlled white rot of onions
more efficiently by placing 25 mg active ingredientll metalaxyl in the gel, whereas
much higher concentrations of metalaxyl were required to overcome the problem
when incorporated in the soil. Similarly, iprodione, when placed in gel, effectively
controlled white rot of salad onion at one-quarter the rate used for dry seeds
(Entwistle 1979).
Hadar et al. (1984) added the biocontrol agents Trichoderma harsianum and
T. kininoii to gel to prevent seed rot in peas and navy beans when sown in
Pythium-infested soil. They determined that 95% of the plants emerged when
seeded in gel amended with T. harsianum and T. kininoii compared with 15%
when planted with dry seeds. Inclusion of Rhizobium in the gel resulted in more
than three times more root nodules on navy bean (Phaseolus vulgaris L.)
compared to inoculation of dry seed (Hardwick and Hardaker 1977).
The growth of Trichoderma hamatum, a biocontrol and antagonist used to
control Rhizoctonia solani, was compared in Planta-gel, Natrosol, and Laponite
gel carriers (Mihuta-Grimm and Rowe 1986). Superior growth of Trichoderma
isolates was obtained in Natrosol versus Planta-gel and Laponite gels. Due to
better growth of the antagonist in Natrosol, this gel was supplemented with
Trichoderma to control damping off when fluid sowing radish seeds. Four
delivery systems for application of Trichoderma isolates were compared: (1) fluid
drilling in Natrosol; (2) coated seed; (3) application of wheat-bran grown isolated
applied in the furrow before planting; and (4) soil drenching with 5-day broth-
grown cultures. The best disease control was consistently obtained with fluid
drilling.
4.1.8 Gel Studies with Somatic Embryos
Baker (1985) determined that a mixture of peat and vermiculite was superior to
vermiculite alone as a soil planting medium for carrot somatic embryos since the
former alleviated water stress. However, vermiculite was the best soil cover since
it did not crust and inhibit emergence. Baker also compared Natrosol and
Planta-gel for suitability as carriers for fluid drilling somatic embryos. No
difference was measured between gels with respect to embryo emergence, which
was approximately 5 to 30% under in vitro conditions. The polyacrylamide

material was used in subsequent greenhouse studies.
Schultheis et al. (1986a, b, 1990) reported that somatic embryos of sweet
potato grew quite rapidly in N-gel when compared to three other gel carriers. All
embryos suspended in the gel were viable after 6 days and up to 62% of the
embryos formed plants. Both NAA and BAP improved plantlet formation in
the gel and a sucrose concentration of 1.6% resulted in the highest plantlet
formation. Up to 25% more plants were obtained when elongated, torpedo-stage
(expanded mature) embryos were sown in the gel carrier compared with
embryos sown at the torpedo (immature) or cotyledonary (mature) stages.
Although somatic embryos are similar to zygotic embryos, the former lack
many of the stored reserves zygotic embryos have, available for their early
seedling growth. Thus, it is important to determine which "growth factors" are
necessary to achieve rapid and complete plant formation from somatic embryos.
It is also necessary to use a synthetic seeding method which could be easily
adapted for field use. The incorporation of additives around the embryos to
sustain and improve their growth and development into plants would be required
for the successful implementation of synthetic seeding.
5 Automation of Synthetic Seed Production
The commercial application of somatic embryogenesis for many crops requires

high volume, low cost production, which invariably translates to a requirement
for automated production and harvest. Relative need for automation will vary
depending on crop application. Synthetic seed production can be separated into
several discrete steps, each of which are amenable to automation: (1) production
of cell cultures and somatic embryos; (2) sorting and harvesting of somatic
embryos; and (3) encapsulation and dehydration.
5.1 Automation of Cell Culture and Somatic Embryo Production
Production unit operations vary considerably depending on the culture system

used. While mass production of somatic embryos is probably best accomplished
in liquid medium, which can be scaled up in bioreactors (Redenbaugh et al.
1988), a number of possible alternatives have been explored. For Petri plate-
based protocols, production involves subculturing callus from plate to plate
under aseptic conditions. Subculturing involves medium preparation, plate lid
removal, harvesting calli, sieving and dispensing calli on new plates. Kurata and
Futaya (1992) developed an automatic sieving system applicable to many of
these operations. The goal of this system was to select appropriately sized calli for
the production of carrot somatic embryos. However, the system was not as
effective as manual sieving, but did demonstrate the validity of the concept.
For a liquid-based production protocol (primarily bioreactors), unit

operations are typical of chemical process control tasks. Durzan and Durzan
(1991) list the process control objectives as: (1) suppression of the influence of
external disturbances on the somatic embryogenesis process; (2) assurance of the
stability of embryo development; and (3) optimization of the overall, long-term
performance of somatic cells and embryos. Process control requires identifying
the outputs to be controlled (e.g., embryo production rate and maturity), how
these outputs are to be quantified, and the inputs to be used to manipulate the
outputs. In some instances production objectives may be met by controlling
elemental physical and chemical parameters such as temperature and p02
without the need to measure the output variables. A successful example of this
type of process control is the model system for the bioreactor production of
poinsettia globular somatic embryos (Preil 1991). The manipulated variables
(inputs) for this system included temperature, agitation, pH, p02' redox
potential, and pC02. The controlled variable (output) was embryo production
rate. Relationships between the controlled and manipulated variables were
established by off-line experimentation which avoided the need for on-line
monitoring of embryo production.
On-line monitoring of the callus production phase of the somatic
embryogenesis process was demonstrated by Harrell et al. (1991). A machine
vision system was used to nondestructively monitor the growth of sweet potato
callus during a lO-day culture period in an airlift bioreactor. Growth data
obtained with the system included overall reactor popUlation and population
estimates for the 200-1200 Ilm fractions at 200-Jlffi intervals. A model of callus
growth was developed to explain the mechanics of callus enlargement. The model
was based on the assumptions that (1) the calli could not shrink or subdivide, (2)
there was a fixed percentage of the initial popUlation within each fraction that
was nonviable, and (3) growth rates did not vary with time during the culture
period. It was determined with this system that growth rates and nonviable ratios
decreased as fraction size increased.
Obstacles to automated production lie primarily in the biological systems.
While well-developed somatic embryos capable of surviving direct field planting
have been produced in liquid flask culture (e.g., Gray et al. 1992), there have been
difficulties involved with scale-up of cultures in larger bioreactors. The main
obstacle is that somatic embryos generally do not develop as well in liquid culture
medium, when compared to solid medium. For example, plant recovery from
suspension culture-derived somatic embryos of alfalfa was less than from callus
and decreased still more when scale-up to a larger bioreactor was attempted
(Stuart et al. 1987). Improvements in bioreactor design are needed as well as a
better understanding of somatic embryo development under liquid culture
conditions to achieve progress in large-scale production.
5.2 Automation of Somatic Embryo Selection and Harvest
Separation of viable embryos from other tissue is necessary due to lack of

synchronization of embryo maturity and the proliferation of aberrant structures
in many systems. Machine vision has been the primary technology used to sense
viable embryos (see Chapter 1. 7, this VoL). Robotic and fluidic approaches have
been developed to do the physical sorting. Postharvest encapsulation has been
developed to help improve the embryo's chance of survival and germination.
Sakamoto et al. (1992) developed a system that automatically encapsulated
somatic embryos in gel beads, which were optically sorted, and then sowed in
cells of a conventional growing tray. The self-breaking gel beads were composed
of modified hydrogel as described above and contained a sucrose sustained-
release microcapsule. The embryos were approximately 5 mm in length at the
time of encapsulation and were subjected to a pre-encapsulation treatment
process, which included mild dehydration (Onishi et al. 1992). Approximately
one-third of the beads produced by the system contained no embryo, one-third
contained a single embryo and one-third contained multiple embryos. Optical
sorting was implemented to collect only those beads whose green content was
higher than an established threshold. This eliminated empty beads and those
which contained pale green embryos. The system was capable of producing and
sowing 70 000 synthetic seeds per day and test results showed a plant recovery
rate of 52%.
Manual sorting of somatic embryos prior to germination is often performed
in an attempt to enhance the plant recovery rate. Visual cues predominate in this
process and several researchers have attempted to replicate and improve upon
human sorting using machine vision. The early work by Grand d'Esnon et al.
(1988) demonstrated the technical practicality of recognizing torpedo-shaped
sweet potato embryos with machine vision. Cazzulino et al. (1991) extended the
concept and quantified embryo production experiments with machine vision by
counting the occurrence of globular, heart-, and torpedo-shaped embryos in
carrot suspension cultures. Kurata et al. (1991) developed a thinning algorithm
to recognize torpedo-shaped carrot embryos. In this approach, the outline of an
embryo was degenerated to a skeleton figure, which facilitated detection and
analysis of cotyledons. The skeleton approach was also employed by Cheng and
Ling (1992) to analyze images of coffee somatic embryos. The algorithm was 74%
successful at distinguishing between 69 torpedo and non torpedo stage embryos
judged by a human expert. Among misclassifications, 13 human-judged torpedo
embryos were classified as non torpedo stage embryos and five nontorpedo stage
embryos were incorrectly classified as torpedo embryos.
Kurata and Shono (1992) developed a recognition algorithm based on
Fourier coefficients extracted from embryo outlines. Under limited testing, this
algorithm coincided with human judgment at the 74% level. Birch and Norway
spruce somatic embryos were classified with a machine vision system by
Hamalainen et al. (1992b, 1993). The classification algorithms were based on
generic size and shape features in addition to morphological specific features,
which quantified certain cotyledon characteristics. When tested on an indepen-
dent set of birch somatic embryos (a set that excludes any images used to develop
the classification algorithm), only 0.1 % of nonembryo objects were classified as
embryos and only 17% of the good embryos were discarded by the classifier.
A novel approach to image classification of somatic embryos was described
by Molto and Harrell (1992). This classification algorithm utilized a neural
network to emulate the judgment of a human experienced in selecting viable

embryos. The network processed 17 geometric features extracted from an
embryo's outline. The feature set included size measures such as embryo length as
well as shape measures such as circularity and roughness. Alternatively, the
neural network could process 32 radii lengths from the perimeter centroid to 32
perimeter points at a constant angular increment of 21t/32 radians. Angular
intervals were taken counterclockwise starting from the principal axis of the
embryo. On an independent test set, the neural network classifier correctly
selected 75% of the embryos deemed desirable by the human expert and correctly
rejected 74% of the malformed, immature, or nonembryo structures.
Hamalainen et al. (1992a) discussed a prototype system that could me-
chanically harvest somatic embryos targeted by the machine vision classifier
using a robotic manipulator. A fluidic, in vitro approach to embryo monitoring
and sorting was developed for liquid-based, bioreactor culture systems by Hood
(1992). A vision system was used to rate the viability of embryos as they were
pumped through a 3-mm, square glass conduit. Embryos were ranked with the
neural network described in Molto and Harrell (1992). Embryos targeted by the
network were ejected from a gap in the flow conduit by a precisely timed injection
of culture medium from a control nozzle. The positions and velocities of embryos
between the imaged section of the conduit and the harvest gap were monitored by
an object tracker employing 30 LED/photo-diode pairs mounted along the
conduit. Objects not harvested were routed back into the reactor. The harvester
was designed to sort one embryo per second. In nonaseptic tests 28% of the
desirable embryos from a test population were harvested improving the homo-
geneity of the harvested population from 56 to 88% (Harrell et al. 1992).
5.3 Automation of EncapsUlation and Dehydration
It may be possible to combine the steps of embryo dehydration and en-

capsulation in an automated process. Dehydration and encapsulation may be
accomplished simultaneously by using an osmotically active synthetic seed coat
(Gray 1987b). Hydrated embryos placed in such a material would begin to lose
water during the capsule hardening process. Final water content could then be
controlled by composition of the encapsulation.
6 Estimated Cost of Synthetic Seed
The overall expense of introducing synthetic seed technology for a given crop will
result from a combination of development and seed production costs and will
vary depending on the level of sophistication required for a given application
(Table 3). For example, use of naked, nonquiescent embryos will involve less
preliminary development expense but the cost of manipUlating each embryo to a
plant will be relatively high. In contrast, development of automated culture and
encapsulation systems will be relatively expensive but the mass production of

synthetic seed with convenient storage and handling characteristics will result in
relatively low plant production costs. It is impossible to determine actual costs
for all types of synthetic seeds at this early stage of development since several key
components of technology (i.e., automated culture systems and hardened encap-
sulations as described above) have not yet been refined. However, use of hydrated
encapsulated alfalfa somatic embryos, which involves much hand manipulation
to produce, would result in a cost of US $0.033 for each greenhouse transplant
(Redenbaugh et al. 1987a). This level of synthetic seed technology would be
effective if adapted to certain crops with high per-plant value as discussed below.
Further, cost per plant would undoubtedly be greatly reduced with an automated
system.
7 Crop Applications for Synthetic Seed
Potential applications of synthetic seed will vary from crop to crop depending on
the relative sophistication of existing production systems and the opportunities
for improvement. Whether or not a cost advantage results from synthetic seed
will ultimately determine its commercial use. As discussed above, for seed-
propagated agronomic crops, relatively sophisticated quiescent, encapsulated
somatic embryos produced en masse will be necessary in order to achieve
adequate planting efficiencies. However, for vegetatively propagated crops,
particularly those with a high per-plant value, naked, hand-manipulated,
nonquiescent embryos may be cost effective. A discussion of potential
applications of synthetic seed technology for several specific crop types follows.
7.1 Ornamental Crops
Ornamental crops have a yearly value of over $ 2.4 billion in the USA
(Anonymous 1990). Many are laboriously micropropagated via tissue and organ
culture where per-plant production costs can exceed $ 0.50 (Florkowski et al.
1988). Most of this cost is due to manpower needed for multiple culture and
rooting steps. Substitution of embryogenic culture systems for such crops would
greatly reduce labor costs since somatic embryos could be mass produced
from callus, then hand selected and placed directly into planting fiats, resulting
in rooted plants. This would eliminate several labor-intensive steps. The
implementation of synthetic seed for ornamental crops is compelling since a
relatively modest level of technological development would result in lowering
existing costs. Somatic embryogenesis has been reported for a number of
ornamental species (Gray and Purohit 1991a). Additional ornamental systems
surely will be developed and refined as pressure to decrease labor costs intensifies.
7.2 Vegetable Crops
Certain vegetable crops are also early candidates for implementation of synthetic
seed technology due to high seed and/or propagation costs in conjunction with
high per-plant value. Generally, hybrid varieties bring higher prices. Asparagus
(Asparagus ojjicinalis altiUs L.) and certain cucurbits have seed costs that
approximate those estimated for encapsulated, nonquiescent somatic embryos
(i.e., > $ 0.033). For seedless watermelon (Citrullus lanatus [Thunb.] Matsum. &
Nakai), seed cost can be so high (up to $ 0.35 per seed) as to approach that of
tissue culture micropropagation. Considering that often only 30% of seedless
watermelon seeds germinate, the cost of producing a single seedling can be as
high as $ 1.05. For this crop, synthetic seed could reduce per-plant costs by
circumventing barriers to seed production. However, seed costs are volatile and
the first report of somatic embryogenesis has just occurred for watermelon
(Compton and Gray 1993). Therefore, although the application is compelling,
much research and development would be necessary to commercialize synthetic
seed technology for this crop. For sweet potato, 1 acre of nursery space is
consumed to produce enough plants for 10 acres of production field (Cantliffe
et al. 1987). In this instance, implementation of synthetic seed would dramati-
cally lower production costs by eliminating nursery requirements.
7.3 Conifers
Forest conifers can be propagated economically only by seed. Improvement via

conventional breeding is extremely time-consuming due to the long conifer life
cycle. Furthermore, conifers are highly heterozygous so that seed from out-
standing individuals does not necessarily result in improved progeny. Superior
breeding stock has been developed, but there is a lag time to produce improved
seed in an orchard setting. Synthetic seed offers the possibility of cloning
outstanding elite trees at reasonable costs, thus circumventing years of develop-
ment (Farnum et al. 1983). To date, well-developed somatic embryos have been
ot>tained for a number of conifer tree species (e.g., Gupta and Durzan 1987;
Becwar et al. 1989; Attree et al. 1991).
7.4 Forage Crops
Synthetic cultivars of seed-propagated, self-incompatible crops such as alfalfa

and orchardgrass are developed laboriously by selecting phenotypically uniform
but genetically distinct lines (Gray et al. 1987, 1992; McKersie et al. 1989; Gray
1990a). These lines are then allowed to cross-pollinate for seed production. Such
seed is nonuniform and each resulting plant is a potentially distinct genotype.
The nature ofthis breeding system makes it difficult to incorporate specific new
genes into existing lines. Use of synthetic seed would allow single, outstanding
hybrids to be utilized as cultivars since self-fertilization would not be needed for
seed increase. Such cultivars would be genetically uniform. Although excellent
144 DJ. Gray et al.
embryogenic culture systems exist for both alfalfa (e.g., Walker and Sato 1981)
and orchardgrass (e.g., Gray et al. 1984), a limitation to this application of
synthetic seed technology is the low per-plant value and the low cost of existing
seed. An intermediate use of synthetic seed for these crops may be to increase
parental lines prior to establishment in open crossing blocks (McKersie et al.
1989; Gray 1990a; Gray et al. 1992).
7.5 Fruit and Nut Crops
Planting efficiency of crops that are currently vegetatively propagated due to self-
incompatibilities and long breeding cycles, such as fruits and nuts etc., could
theoretically be increased by using synthetic seed instead of cuttings. But, since
existing methods tend to be cost effective, developmental costs of synthetic seed
would not likely be justified. Direct planting via synthetic seed is further
complicated for varieties that require grafting to a rootstock. Use of synthetic
seed for germplasm conservation of these crops could be highly advantageous,
however. For example, germplasm of unique clones maintained as living plants
in field gene banks (Towill 1988). This method of conservation is expensive and
subject to loss from environmental disasters. The use of in vitro conservation
methods for these crops, in addition to recalcitrant-seeded species, has been
emphasized (Bajaj 1986; Withers 1989). Use of synthetic seed would allow clonal
germplasm of grape to be conserved in seed repositories. More genotypes could
be conserved since space problems would be eliminated. Moreover, development
costs would be reduced since automated production equipment for mass
production would not be needed for the relatively small number of synthetic seed
required. This method of germplasm conservation would be particularly useful
for tropical species where existing conservation is inadequate or nonexistent.
Grape is a good experimental prospect since well-developed somatic embryos
have been obtained (Gray 1987b,1989; Gray and Mortensen 1987). The pros-
pects of using synthetic seed technology for grape germplasm conservation was
previously discussed in depth (Gray and Compton 1992).
7.6 Cotton and Soybean
Commercial quantities of hybrid seed are difficult to produce for certain seed-
propagated crops such as cotton (Gossypium hirsutum L.) and soybean due to
cleistogamous flowers and/or problems with flower abscision. Thus, seed of most
existing cultivars is derived from self-pollination. However, relatively small
numbers of hybrids can be produced laboriously by extensive hand pollination
and subsequently mass produced by use of synthetic seed. Hybrid vigor could
then be exploited at the production level. Somatic embryogenesis has been
obtained for both cotton (Finer 1988) and soybean (Ranch et al. 1985), although
the commercial potential of synthetic seed for such crops is unclear at this time,
considering the moderate cost of existing seed and the possibility of developing
chemical emasculants to produce male-sterile plants in the field.
7.7 Hybrid Cereals
Synthetic seed technology provides the possibility of circumventing the need for
inbred and male-sterile parental lines in hybrid seed production. For example,
the hybrid corn industry relies on inbred parentals to produce uniform hybrid
seed. Mass hybridization is possible by using male-sterile lines as females.
Increased production costs over open-pollinated seed that are incurred by use of
inbred and male-sterile lines are more than offset by the resulting yield and
quality conferred by hybrid vigor. However, development and maintenance of
these parental lines consume much of the time and resources of a breeding
program and integration of new germplasm is slow. An intriguing possibility is
the use of synthetic seed to propagate new hybrids and eliminate the need for
parental inbreds and male steriles altogether. This would facilitate com-
mercialization of new hybrids and would probably stimulate competition since
newcomers could produce cultivars without an existing stock of parental inbreds.
Although somatic embryogenesis of corn is well described (Kamo et al. 1985), it
is unclear whether the successful hybrid corn industry would accept this concept.
8 Conclusion
The structural similarities between somatic embryos and zygotic embryos are the
basis for interest in synthetic seed technology. However, functionally, somatic
embryos have yet to match the convenience of seed. A number of obstacles, for
example, problems with mass production, encapsulation, and uniformity, must
be overcome before synthetic seed can be useful for most crops. Many of these
problems are probably related to shortcomings of contemporary in vitro culture
systems. In vitro culture presents a very different environment compared to that
in which zygotic embryos develop (i.e., seeds). The environment of the
developing seed is dynamic and complex, probably causing many functional
attributes of zygotic embryos that we would desire to have expressed in somatic
embryos.
References
Ammirato PV (1987) Organizational events during somatic embryogenesis. In: Somers DA,
Gengenback BG, Biersboer BD, Hackett, WP, Green CE (eds) Plant tissue and cell culture. Alan
Liss, New York, pp 57-81
Anandarajah K, McKersie BD (1 990a) Enhanced vigor of dry somatic embryos of Medicago sativa L.
with increased sucrose. Plant Sci 71: 261
Anandarajah K, McKersie BD (1990b) Manipulating the desiccation tolerance and vigor of dry
somatic embryos of Medicago sativa L. with sucrose, heat shock and abscisic acid. Plant Cell Rep
9: 451--455
Anonymous (1990) Floriculture crops, 1989 summary. National Agric Stn USDA, Washington DC
Attree SM, Fowke LC (1993) Embryogeny of gymnosperms: advances in synthetic seed technology of
conifers. Plant Cell Tissue Organ Cult 35: 1-35
Attree SM, Moore D, Sawhney VK, Fowke LC (1991) Enhanced maturation and desiccation tolerance
of white spruce [Picea glauca (Moench) Voss] somatic embryos: effects of non-pi asmolysing water
stress and abscisic acid. Ann Bot 86: 519-525
Attree SM, Pomeroy MK, Fowke LC (1992) Manipulation of conditions for the culture of somatic
embryos of white spruce for improved triacylglycerol biosynthesis and desiccation tolerance.
Planta 187: 395-404
Nuclear techniques and in vitro culture for plant improvement. IAEAlFAO, Vienna 1985, pp
43-57
Bajaj YPS (ed) (1990) Biotechnology in agriculture and forestry, vol 11. Somaclonal variation in crop
improvement I. Springer, Berlin Heidelberg New York
Baker CM (1985) Synchronization and fluid sowing of carrot, Daucus carota somatic embryos. MS
Thesis, University of Florida, Gainesville
Becwar MR, Noland TL, Wycoff JL (1989) Maturation, germination and conversion of Norway
spruce (Picea abies L.) somatic embryos to plants. In Vitro Cell Dev Bioi 25: 575-580
Bewley JD, Black M (1978) Physiology and biochemistry of seeds in relation to germination, vol I.
Springer, Berlin Heidelberg New York
Bewley JD, Black M (1985) Seeds: physiology of development and germination. Plenum Press, New
York, 367 pp
Brackpool A, Branton R, Blake J (1986) Regeneration in palms. In: Vasil IK (ed) Cell culture and
somatic cell genetics of plants, vol 3. Academic Press, Orlando, p 345
Brocklehurst PA (1979) Fluid drilling. Oxygen supply to the developing seedling. Nat! Veg Res Stn
Annu Rep 1978: 100-102
Bryan HH, Stall WM, Gray D, Richmond NS (1978) Fluid drilling of pregerminated seeds for
vegetable gardening. Proc Fla State Horic Soc 91: 88-90
Bryan HH, Mass FA, Sherry M (1982) Fluid drilling of pregerminated pepper seed. Proc Fla State
Hortic Soc 95: 353-356
Cantliffe DJ, Liu JR, Schultheis JR (1987) Development of artificial seeds of sweet potato for clonal
propagation through somatic embryogenesis. In: Smith WH, Frank JR (eds) Methane from
biomass - a systems approach. Elsevier, New York, pp 183-195
Carman JG (1988) Improved somatic embryogenesis in wheat by partial simulation of the in-ovulo
oxygen, growth-regulator and desiccation environments. Planta 175: 417-424
Carman JG (1989) The in ovulo environment and its relevance to cloning wheat via somatic
embryogenesis. In Vitro Cell Dev Bioi 25: 1155-1162
Cazzulino DL, Pedersen H, Chin CK (1991) Bioreactors and image analysis for scale-up and plant
propagation. In: Vasil IK (ed) Cell culture and somatic cell genetics of plants, vol8. Scale-up and
automation in plant tissue culture. Academic Press, San Diego, pp 147-177
Cheng Z, Ling PP (1992) Computer vision for plant embryo quality evaluation, ASAE Pap 92-3575.
Am Soc Agric Eng, St Joseph, MI, 49085-9659
Compton ME, Gray DJ (1993) Somatic embryogenesis and plant regeneration from immature
cotyledons of watermelon. Plant Cell Rep 12: 61-65
Compton ME, Gray DJ, Songstad DD (1992) Induction of quiescence in maize somatic embryos by
dehydration. In Vitro Cell Dev Bioi Plant 28: 197-201
Conger BV, Hanning GE, Gray DJ, McDaniel JK (1983) Direct embryogenesis from mesophyll cells
of orchardgrass. Science 221: 850-851
Conger BV, Hovanesian JC, Trigiano RN, Gray DJ (1989) Somatic embryo ontogeny in suspension
cultures of orchardgrass. Crop Sci 29: 448-452
Costigan PA, Locascio SJ (1982) Fertilizer additives within or around the gel for fluid drilled cabbage
and lettuce. HortScience 17: 746-748
Currah IE, Gray D, Thomas TH (1974) The sowing of germinated vegetable seeds using a fluid drill.
Annu Appl Bioi 76: 311-318
Drew RLK (1979) The development of carrot (Daucus carota L.) embryoids (derived from cell
suspension culture) into plantlets on a sugar-free basal medium. Hortic Res 19: 79-84
Dunstan DI, Short KC, Merrick MA, Collin HA (1982) Origin and early growth of celery embryoids.
New Phytol 91: 121-128
Durzan DJ, Durzan PE (1991) Future technologies: model-reference control systems for the scale-up of
embryogenesis and polyembryogenesis in cell suspension cultures. In: Debergh PC, Zimmerman
RH (eds) Micropropagation. Kluwer, Amsterdam, pp 389-423
Entwistle A (1979) Fluid drilled salad onions. Natl Veg Res Stn Annu Rep 1978: 72-73
Evans DA, Gamborg OL (1982) Chromosome stability of cell suspension cultures of Nicotiana species.
Plant Cell Rep I: 104-107
Farnum P, Timmis R, Kulp JL (1983) Biotechnology offorest yield. Science 219: 694-702
Finch-Savage WE, Cox CJ (1982) Effects of adding plant nutrients to the gel carrier used for fluid
drilling early carrots. J Agric Sci 99: 296-303
Finch-Savage WE, Cox CJ (1983) Effects of adding plant nutrients to the gel carrier used for fluid
drilling germinating lettuce and onion seeds. Ann Appl Bioi 102: 213-217
Finer JJ (1988) Plant regeneration from somatic embryogenic cultures of cotton (Gossypium hirsutum
L.). Plant Cell Rep 7: 399-402
Flemion F (1937) After-ripening at 5 °C favours germination of grape seeds. Contr Boyce Thompson
Inst 9: 7-15
Florkowski WJ, Lindstrom 0, Robacker C, Simonton W (1988) Biological, technical and economic
aspects of micro propagation. Georgia Agric Exp Stn Res Rep 556, Athens, Georgia, pp 1-17
Frazier DC, Wiest SC, Wooton RD (1982) Storage of pre germinated seed of snapdragon (Antirrhinum
majus L.) in hydrophilic gels. J Am Soc Hortic Sci 107: 660-664
Fujii FA, Slade DT, Redenbaugh K, Walker KA (1987a) Artificial seeds for plant propagation. Trends
Biotechnol5: 335-340
Fujii JA, Slade DT, Redenbaugh K, Walker KA (1987b) How APV puts the technology into
biotechnology. BioTechnology 5: 335-339
Fujii FA, Slade D, Redenbaugh K (1989) Maturation and greenhouse planting of alfalfa artificial
seeds. In Vitro Cell Dev Bioi 25: 1179-1182
Gamborg OL, Miller RA, Ojima K (1968) Nutrient requirements of suspension cultures of soybean
root cells. Exp Cell Res 50: 148-151
Gamborg OL, Constabel F, Miller RA (1970) Embryogenesis and production of albino plants from
cell cultures of Bromus inermis. Planta 95: 355-358
Geier T, Beck A, Preil W (1992) High uniformity of plants regenerated from cytogenetically variable
embryogenic suspension cultures of poinsettia (Euphorbia pulcherrima Willd. ex Klotzch). Plant
Cell Rep 11: 150-154
Ghate SR (1982) Rheological properties of gels used as carriers of germinated seeds. Trans ASAE 26:
1757-1762
Ghate SR, Phatak SC (1983) Growth of pepper seeds planted with a gel chemical mixture. Trans ASAE
26: 687--691
Grand d' Esnon A, Chee R, Harrell RC, Cantliffe DJ (1988) Qualitative and quantitative evaluation of
liquid tissue culture by artificial vision. Pap 88.395. Presented at AGENG 88, Agricultural
Engineering Int Conf, Paris, March 2-5, 1988, 12 pp
Gray DJ (1981) Fluid drilling of vegetable seeds. Hortic Rev 3: 1-27
Gray DJ (1987a) Introduction to the symposium. In: Proc Symp Synthetic seed technology for the mass
cloning of crop plants: problems and perspectives. HortScience 22: 796-797
Gray DJ (1987b) Quiescence in monocotyledonous and dicotyledonous somatic embryos induced by
dehydration. In: Proc Symp Synthetic seed technology for the mass cloning of crop plants:
problems and perspectives. HortScience 22: 810-814
Gray DJ (I 987c) Concluding remarks. In: Proc Symp Synthetic seed technology for the mass cloning
of crop plants: problems and perspectives. HortScience 22: 814
Gray DJ (1987d) Effects of dehydration and other environmental factors on dormancy in grape
somatic embryos. HortScience 22: 1118
Gray DJ (1989) Effects of dehydration and exogenous growth regulators on dormancy, quiescence and
germination of grape somatic embryos. In Vitro Cell Dev Bioi 25: 1173-1178
Gray DJ (1990a) Somatic embryogenesis and cell culture in the Poaceae. In: Kasperbauer MJ (ed)
Biotechnology in tall fescue improvement. CRC Press, Boca Raton, pp 25-57
Gray DJ (I990b) Synthetic seed for clonal production of crop plants. In: Taylorson RD (ed) Recent
advances in the development and germination of seeds. Plenum Press, New York, pp 29-45
Gray DJ, Compton ME (1993) Grape somatic embryo dormancy and quiescence: potential of
dehydrated synthetic seeds for germplasm conservation. In: Redenbaugh K (ed) Synseeds:
applications of synthetic seeds to crop improvement. CRC Press, Boca Raton, pp 367-379
Gray DJ, Conger BV (1985a) Quiescence in somatic embryos of orchardgrass (Dactylis glomerata)
induced by desiccation. Am J Bot 72: 816
Gray DJ, Conger BV (1985b) Somatic embryo ontogeny in tissue cultures of orchardgrass. In: Henke
RR, Hughes KW, Constantin MJ, Hollaender A (eds) Tissue culture in forestry and agriculture.
Plenum Press, New York, pp 49-57
Gray DJ, Conger BV (1985c) Time-lapse light photomicrography and scanning electron microscopy of
somatic embryo ontogeny from cultured leaves of Dactylis glomerata (Gramineae). Trans Am
Microsc Soc 104: 395-399
Gray DJ, Mortensen JA (1987) Initiation and maintenance oflong-term somatic embryogenesis from
anthers and ovaries of Vitis longii 'Microsperma'. Plant Cell Tissue Organ Cult 9: 73-80
Gray DJ, Purohit A (199Ia) Somatic embryogenesis and development of synthetic seed technology.
Crit Rev Plant Sci 10: 33-61
Gray DJ, Purohit A (199Ib) Quiescence and dormancy in plant somatic embryos. In: Bajaj YPS (ed)
Biotechnology in forestry and agriculture, vol. 17. High-tech and micropropagation I, Springer,
Gray DJ, Conger BV, Hanning GE (1984) Somatic embryogenesis in suspension and suspension-
derived callus cultures of Dactylis glomerata. Protoplasma 122: 196-202
Gray DJ, Conger BV, Songstad DD (1987) Desiccated quiescent somatic embryos of orchardgrass for
use as synthetic seeds. In Vitro Cell Dev Bioi 23: 29-33
Gray DJ, Trigiano RN, Conger BV (1992) Liquid suspension culture of orchardgrass somatic embryos
and their potential for the breeding of improved varieties. In: Redenbaugh K (ed) Synseeds:
Gupta PK, Durzan DJ (1987) Biotechnology of somatic polyembryogenesis and plantlet regeneration
in loblolly pine. BioTechnology 5: 147-152
Haccius B (1978) Question of unicellular origin of non-zygotic embryos in callus cultures.
Hadar Y, Harmon GE, Taylor AG (1984) Evaluation of Trichoderma koningii and Trichoderma
harzianum from New York soils for biological control of seed rot caused by Pythium spp.
Phytopathology, 70: 106-110
Hiimiiliiinen JJ, Kurten U, Kappinen V, Heilala J (1992a) Automated classification of somatic plant
embryos. Acta Hortic 319, Transplant Production Systems, pp 601-605
HiimiiJiiinen JJ, Yang H, Jokinen KJ (1992b) Selection of Norway spruce somatic embryos by
computer vision. Optics in agriculture and forestry, SPIE's DE/Technology '92,15-20, Nov, 1992,
Boston
Hiimiiliiinen JJ, Kurten U, Kappinen V (1993) Classification of somatic plant embryos by computer
vision. Biotech Bioeng 41: 35-42
Hanning GE, Conger BV (1986) Factors influencing somatic embryogenesis from cultured leaf
segments of Dactylis glomerata . J Plant Physiol123: 23-29
Hardwick R, Hardaker J (1977) Fluid carrier for Rhizobium bacteria. Natl Veg Res Stn Annu Rep
1976: 68-69
Harrell RC, Bieniek M, Cantliffe DJ (1991) Noninvasive evaluation of somatic embryogenesis.
Harrell RC, Hood CF, Molto E, Munilla R, Bieniek M, Cantliffe DJ (1992) Automatic identification
and separation of somatic embryos in vitro. Acta Hortic 319, Transplant Production Systems, pp
595-600
Haydu Z, Vasil IK (1981) Somatic embryogenesis and plant regeneration from leaf tissues and anthers
of Pennisetum purpureum Schum. Theor Appl Genet 59: 269-273
Hood CF (1992) Automated harvest of somatic embryos. PhD Thesis, Agricultural Engineering
Department, University of Florida, Gainesville, 195 pp
Janick J, Kitto SL, Kim Y-H (1989) Production of synthetic seed by desiccation and encapsulation. In
Vitro Cell Dev Bioi 25: 1167-1172
Jeon J, Liu J, Yang S, Lee H, Joung H, Han M (1986) Development of a model system for artificial seed
production. I. Encapsulation of somatic embryos by alginic acid. Korean J Plant Tissue Cult 13:
119-128
Jones LH (1974) Long-term survival of embryoids of carrot (Daucus carota L.). Plant Sci Lett 2:
221-224
Kamo KK, Becwar MR, Hodges TK (1985) Regeneration of Zea mays L. from embryogenic callus.
Bot Gaz 146: 327-334
Kim Y-H, Janick J (1989a) Origin of somatic embryos in celery tissue culture. HortScience 24: 671-673
Kim Y-H, Janick J (1989b) ABA and polyox-encapsulation of high humidity increases survival of
desiccated somatic embryos of celery. HortScience 24: 674-676
King MW, Roberts EH (1979) The storage of recalcitrant seeds. Report for the International Board for
Plant Genetic Resources Secretariat, Rome, IBPGR, p 96
Kitto SL, Janick J (1985a) A citrus embryo assay to screen water-soluble resins as synthetic seed coats.
HortScience 20: 98-100
Kitto SL, Janick J (1985b) Production of synthetic seeds by encapsulating asexual embryos of carrot.
J Am Soc Hortic Sci 110: 277-282
Kitto SL, Janick J (1985c) Hardening treatments increase survival of synthetically coated asexual
embryos of carrot. J Am Soc Hortic Sci 110: 283-286
Koster KL, Leopold CA (1988) Sugars and desiccation tolerance in seed. Plant Physiol 88: 829-832
Koul AK, Karihaloo JL (1977) In vitro embryoids from anthers of Narcissus biflorus Cur. Euphytica
26: 97-102
Krikorian AD (1982) Cloning higher plants from aseptically cultured tissues and cells. Bioi Rev 57:
151-218
Kurata K, Futaya Y (1992) Automation of selecting embryogenic callus by sieving. Acta Hortic 319,
Transplant Production Systems, pp 563-566
Kurata K, Shono H (1992) An application of Fourier Descriptor for selecting somatic embryos. Acta
Hortic 319, Transplant Production Systems, pp 591-594
Kurata KM, Terada M, Komine M, Liyange KH (1991) Computer vision for selection of somatic
embryos. ASAE Pap 91-3054. Am Soc Agric Eng, St Joseph MI, 49085-9659
Litz RE, Conover RA (1982) In vitro somatic embryogenesis and plant regeneration from Carica
papaya L. ovular callus. Plant Sci Lett 26:153-158
Liu JR, Jeon JH, Yang SG, Lee HS, Song NH, Jeong WJ (1992) Dry type carrot (Daucus carota L.)
artificial seeds. Sci Hortic 51: I-II
Luckett DJ, Rose D, Knights E (1989) Paucity of somaclonal variation from immature embryo culture
of barley. Austr J Agric Res 40: 1155-1159
McKersie BD, Senaratna R, Bowley SR, Brown DCW, Krochko JE, Bewley JD (1989) Application of
artificial seed technology in the production of hybrid alfalfa (Medicago sativa L.). In Vitro Cell Dev
Bioi 25: 1183-1188
McKersie BP, Bowley SR (1993) Synthetic seeds in alfalfa. In: Redenbaugh K (ed) Synseeds
Mihuta-Grimm L, Rowe RC (1986) Trichoderma spp. as biocontrol agents of Rhizoctonia damping-
off ofradish in organic soil and comparison offour delivery systems. Phytopathology 76: 306-312
Molto E, Harrell RC (1992) Neural network classification of sweet potato embryos. Optics in
agriculture and forestry, SPIE's OElTechnology '92, 15-20 Nov, 1992, Boston
Murashige T (1978) The impact of plant tissue culture on agriculture. In: Thorpe T (ed) Frontiers of
plant tissue culture 1978. International Association for Plant Tissue Culture, University of
Calgary, Calgary, Alberta, pp 15-26
Neumann D, Norer L, Parthier B, Rieger R, ScharfKD, Wollgiehn R, Nieden UZ (1989) Heat shock
and other stress response systems of plants. Bioi Zentralbl108: 1-146
Nikolaeva MG, Katkevich YY (1961) Study of respiration in dormant seeds in relation to temperature.
Fiziol Rast 8: 42-50
Nitzsche W (1978) Erhaltung der Lebensfahigkeit in getrocknetem Kallus. Z Pflanzenphysiol 87:
469-472
Norstog K (1965) Development of cultured barley embryos. I. Growth ofO.l-O.4-mm embryos. Am J
Bot 52: 538-546
Norstog K, Klein RM (1972) Development of cultured barley embryos. II. Precocious germination
and dormancy. Can J Bot 50: 1887-1894
Nutbeam AR, Duffus CM (1978) Oxygen exchange in the pericarp green layer of immature cereal
grains. Plant Physiol62: 360-362
Ogura H, Kyozuka J, Hyashi Y, Koba T, Shimamoto K (1987) Field performance and cytology of
protoplast-derived rice (Oryza sativa L.): high yield and low degree of variation in four Japonica
cultivars. Theor Appl Genet 74: 670-676
150 DJ. Gray et al.
Ohep J, CantIiffe DJ (1980) Growth regulators as gel additives for fluid drilling of pregerminated
tomato seeds. HortScience 15: 407
Ohep J, McMillan R, Bryan H, CantIiffe D (1984) Control of damping-off of tomatoes by
incorporation offungicides in direct seeding gel. Plant Dis 68: 66-67
Onishi N, Mashiko T, Okamoto A (1992) Cultural systems producing encapsulatable units of synthetic
seeds in celery. Acta Hortic 319, Transplant production systems, pp 113-118
Parrott WA, Dryden G, Vogt S, Hildebrand DF, Collins GB, Williams EG (1988) Optimization
of somatic embryogenesis and embryo germination in soybean. In Vitro Cell Dev Bioi 24:
817-820
Parrott WA, Merkle SA, Williams EG (1991) Somatic embryogenesis: potential for use in
propagation and gene transfer systems. Adv Methods Plant Biotechnol4: 158-200
Pearce D, Pharis RP, Rajasekaran K, Mullins MG (1987) Effects of chilling and ABA on [3H]
gibberellin A. metabolism in somatic embryos of grape (Vitis vinifera L.x V. rupestris Scheele).
Plant Physiol80: 381-385
Pill WG, Fieldhouse DJ (1982) Emergence of pregerminated tomato seed stored in gels up to twenty
days at low temperatures. J Am Soc Hortic Sci 107: 722-725
Preil W (1991) Application of bioreactors in plant propagation. In: Debergh PC, Zimmerman RH
(eds) Micropropagation. Kluwer, Amsterdam, pp 425-445
Rajasekaran K, Vine J, Mullins MG (1982) Dormancy in somatic embryos and seeds of Vitis: changes
in endogenous abscisic acid during embryogeny and germination. Planta 154: 139-144
Ranch JP, Oglesby L, Zielinsky AC (1985) Plant regeneration from embryo-derived tissue cultures of
soybean. In Vitro Cell Dev Bioi 21: 653-658
Redenbaugh K (1990) Application of artificial seed to tropical crops. HortScience 25: 251-255
Redenbaugh K, Nichol J, Kossler M, Paasch B (1984) Encapsulation of somatic embryos for artificial
seed production. In Vitro Cell Dev Bioi 20: 256
Redenbaugh K, Paasch BD, Nichol JW, Kossler ME, Viss PR, Walker KA (1986) Somatic seeds:
encapsulation of asexual plant embryos. BioTechnology 4: 797-801
Redenbaugh K, Slade D, Viss P, Fujii JA (1987a) Encapsulation of somatic embryos in synthetic seed
coats. HortScience 22: 803-809
Redenbaugh K, Viss P, Slade D, Fujii J (1987b) Scaleup: artificial seeds. In: Green C, Somers D,
Hackett W, Biersboer D (eds) Plant tissue and culture. Liss, New York, pp 473-493
Redenbaugh K, Fujii JA, Slade D (1988) Encapsulated plant embryos. In: Mizrahi A (ed)
Biotechnology in agriculture. Liss, New York, pp 225-248
Redenbaugh K, Fujii J, Slade D, Viss P, Kossler M (1991) Artificial seed - encapsulated somatic
Redenbaugh K, Fujii JA, Slade D (1993) Hydrated coatings for synthetic seeds. In: Redenbaugh K
(ed) Synseeds application of synthetic seeds to crop improvement. CRC Press. Boca Raton, pp
35-45
Sakamoto Y, Mashiko T, Suzuki A, Kawata H (1992) Development of encapsulation technology for
synthetic seeds. Acta Hortic 319, Transplant Production Systems, pp 71-76
Saranga Y, Kim YH, Janick J (1992a) Changes in tolerance to partial desiccation and in metabolite
content of celery somatic embryos induced by reduced osmotic potential. J Am Soc Hortic Sci 117:
342-345
Saranga Y, Rhodes D, Janick J (I 992b) Changes in amino acid composition associated with tolerance
to partial desiccation of celery somatic embryos. J Am Soc Hortic Sci 117: 337-341
Schultheis JR, Chee R, Cantliffe DJ (1986a) Effect of growth regulators and gel carriers on growth and
development of sweet potato (Ipomoea batatas) somatic embryos. HortScience 21: 762
Schultheis JR, Chee R, CantIiffe DJ (1986b) Growth of sweet potato [Ipomoea batatas (L.) Lam.] in
hydroxyethyl cellulose gel amended with salts and carbohydrates. Sci Hortic 50: 21-33
Schultheis JR, CantIiffe DJ, Chee RP (1990) Optimizing sweet potato [Ipomoea batatas (L.) Lam.] root
and plantIet formation by selection of proper embryo developmental stage and size, and gel type for
fluidized sowing. Plant Cell Rep 9: 356-359
Senaratna T (1992) Artificial seeds. Biotechnol Adv 10: 379-392
Senaratna T, McKersie BD, Borochov A (1987) Desiccation and free radical mediated changes in plant
membranes. J Exp Bot 38: 2005-2014
Senaratna T, McKersie BD, Bowley S (I989a) Desiccation tolerance of alfalfa (Medicago sativa L.)
somatic embryos - influence of abscisic acid, stress pretreatments and drying rates. Plant Sci 65:
253-259
Senaratna T, McKersie BD, Bowley S, Bewley JD, Brown DCW (1989b) Process to induce desiccation
tolerance in somatic embryos. European Patent Application 88306589.8 Publ No 0 300 730
Senaratna T, McKersie BD, Bowley S (1990) Artificial seeds of alfalfa (Medicago sativa L.) induction
of desiccation tolerance in somatic embryos. In Vitro Cell Dev BioI 26: 85-90
Sosa-Coronel J, Motes JE (1979) Evaluation of gels for fluid drilling carrot, pepper, and onion
pregerminated seed. HortScience 14: 36 (Abstr)
Spiegel-Roy P, Vardi A (1984) Citrus. In: Ammirato PV, Evans DA, Sharp WR, Yamada Y (eds)
Handbook of plant cell culture, vol 3. Macmillan, New York, pp 355-372
Stuart DA, Strickland SG, Walker KA (1987) Bioreactor production of alfalfa somatic embryos. In:
Proc Symp Synthetic seed technology for the mass cloning of crop plants: problems and
perspectives. HortScience 22: 800-803
Sussex 1M, Frei KA (1968) Embryoid development in long-term tissue cultures of carrot.
Thevenot C, Ralambosoa J, Simond-Cote E (1987) Influence of abscisic acid and oxygen supply on
germination of more or less dormant apple embryos. Israel J Bot 36: 101-112
Towill LE (1988) Genetic consideration of germ plasm preservation of clonal materials. In: Proc Symp
Genetic considerations in the collection and maintenance of germplasm. HortScience 23: 91-95
Walker KA, Sato SJ (1981) Morphogenesis in callus tissue of Medicago sativa: the role of ammonium
ion in somatic embryogenesis. Plant Cell Tissue Organ Cult I: 109-121
Walton DC (1980) Biochemistry and physiology of abscisic acid. Annu Rev Plant Physiol31: 453-489
White J (1979) Brassica seedling establishment. Natl Veg Res Stn Annu Rep 1978: 76-77
Withers LA (1989) In vitro conservation and germplasm utilisation. In: Brown AHD, Frankel OH,
Marshall DR, Williams JT (eds) The use of plant genetic resources. Univ Cambridge Press, New
York, pp 309-334
11.3 Role of Maturation and Desiccation
of Somatic Embryos in the Production
of Dry Artificial Seed
B.D. McKERSIE, S. VAN ACKER, and F.M. LAI!
1 Introduction
Since the beginning of agriculture plants have been propagated whenever

possible by seeds as a convenient means of multiplication, storage, and distri-
bution. At the end of every harvest, the highest quality portion of the seed crop
is set aside for the next year's planting. In many respects, seeds are an ideal
propagation system because they (l) contain substantial quantities of food
reserves to support the growth and establishment of the young seedlings; (2) are
dormant or quiescent to minimize respiration and maximize longevity; and (3)
are covered with a protective seed coat to allow handling and reduce pathogen
attack. Yet, this system which has served so well has limitations in certain
commodities, and alternative means of propagating plants by technologies such
as micropropagation and artificial seeds have merits.
Seeds derive from the sexual recombination of male and female gametes and
thus are not genetically identical to their parents. This genetic variability is
minimal in self-pollinated crops, such as barley and wheat, but can be quite
substantial in open-pollinated species, such as alfalfa or hybrid varieties of corn
or tomato. Seeds also harbor a number of pathogens which can be inadvertently
spread from contaminated seed production fields. Many important crops,
primarily tropical crops or ornamental plants, are sterile and do not set viable
seeds. This necessitates propagation by cuttings or other vegetative means, and
prevents convenient storage.
Artificial seeds are seen by many as a means of overcoming ~hese limitations
and allowing the clonal propagation oflarge number of disease-free propagules.
The technology has applications in a diverse range of plant species, as an aid in
the breeding of hybrid varieties, as a system of storage of valuable genetic
resources, and as a means of large-scale multiplication. Several recent reviews
describe these applications and the development of technology in a number of
plant species (Gray and Purohit 199Ia,b; Redenbaugh et al. 199Ia,b; McKersie
and Bowley 1993). This review is therefore restricted to the more recent
! Department of Crop Science, University of Guelph, Guelph, Ontario, Canada Nl G 2Wl

©Springer·Verlag Berlin Heidelberg 1995
Role of Maturation and Desiccation of Somatic Embryos 153
development of dry artificial seeds, an advance that combines the potential

benefit of the previous artificial seed technology with the storage and handling
advantages associated with dry seeds.
2 Concepts of Artificial Seeds
Artificial or synthetic seeds are functionally defined as somatic embryos

engineered to be of use in commercial plant production (Gray and Purohit
1991a). There have been several different concepts which can be classified on the
basis of whether the somatic embryos are naked or encapsulated and whether
hydrated or dry. The exact type of artificial seed required depends on the specific
application and commodity. Naked, hydrated somatic embryos can be used to
propagate high value plants such as ornamentals, where the efficiency of somatic
embryogenesis reduces production costs compared to the current labor-intensive
methods of micropropagation. Encapsulated, hydrated somatic embryos can be
handled mechanically and directly planted into peat plugs for greenhouse
production. These propagules could then be transplanted to the field and would
be suitable for vegetables such as carrot or celery. The capsule surrounding the
somatic embryo could be alginate or a similar product which would maintain the
hydration of the embryo, allowing temporary storage, and could contain
nutrients, fungicides, and other growth supplements to enhance the germination
and establishment of the seedling.
Drying the somatic embryo, whether naked or subsequently encapsulated,
allows longer term storage of the propagule, and the artificial seed then becomes
a true analog of conventional seed. These dry somatic embryos could be used for
germplasm conservation in seed storage banks, or simply for bulking up prop a-
gules prior to greenhouse production. Many commodities have very seasonal
value, such as flowers on Valentine's Day, or vegetables for spring planting.
Therefore, greenhouse production of these commodities must be timed precisely
to meet the high value markets. If these commodities are micro propagated, the
tissue culture production must also be very carefully timed. However, if somatic
embryos can be dried and stored, a much smaller, and less seasonal tissue culture
facility could be continually used to produce dry somatic embryos year-round.
These propagules could be stored like conventional seeds and planted in the
greenhouse at the appropriate time to synchronize production. In addition, the
dry somatic embryos could be shipped, or even mailed, from a central tissue
culture facility to a number of diverse greenhouse production facilities.
Desiccation of somatic embryos may also have the added advantages of
acting as a developmental cue in the maturation process of embryos. Desiccation
and rehydration cause a developmental switch from storage reserve synthesis to
storage reserve catabolism in developing zygotic embryos (Kermode et al. 1989)
It would therefore be anticipated that the seedlings from dry somatic embryos
would be more vigorous and exhibit more rapid and uniform development than
seedlings from precociously germinated somatic embryos (McKersie et al. 1989).
154 B.D. McKersie et al.
3 Somatic Embryogenesis in Alfalfa - a Model

for the Development of Dry Somatic Embryos
Saunders and Bingham (1972) were the first to report somatic embryogenesis in
alfalfa and this was followed by many reports of regeneration from callus,
suspension, and protoplast cultures (for a detailed review, see Bingham et al.
1988). Two major factors influence regeneration, the genotype of the plant and
the quantity and type of growth regulators in the induction medium. Many other
factors from the physiological state of the donor plant, to the nitrogen content of
the induction and regeneration media, to light quantity and quality modulate the
embryogenic response (Brown and Atanassov 1985; Bingham et al. 1988; Kris
and Bingham 1988). The following cell culture system is designed to produce
large number of dry, quiescent somatic embryos. The modifications that we have
SOMATIC EMBRYO
PRODUCTION
PHASE I MEDIUM
CALLUS ··INDUCTioN ... Modified SH agar

I> ·······1
SUSPENSION 1< SYNCHRONIZATION •.1 Modified B5 liquid
DEVELOPMENT I ••• • . >DEVELOPMENT "1 BOi2Yagar
MATURATION I I RESERVE DEPOSITION I BOl2Yagar

with high glutamine
MATURATION II
DRYING
1<
1
,
DESICCATION TOLERANce
QUIESCENCE
BOi2Yagar
with abscisic acid
Varying Relative
Humidity
Fig. 1. A schematic summary of the alfalfa (Medicago sativa) tissue culture system used to produce
dry, quiescent somatic embryos for use as artificial seeds
recently made to the system increase the efficiency of the induction process,
synchronize embryo development, facilitate the deposition of storage reserves,
and induce the embryo to express tolerance of desiccation. The protocol as
summarized in Fig.l has many steps involving (1) induction; (2) synchronization;
(3) embryo development; (4) reserve deposition (maturation phase I); (5)
induction of desiccation tolerance (maturation phase II); and finally (6) drying
and storage.
3.1 Embryo Induction
The first step in the culture procedure involves the induction of competent cells to
form proembryogenic structures that directly or indirectly form somatic
embryos. To accomplish this, petiole explants offully expanded, young leaves are
cultured on SH basal medium (Schenk and Hildebrandt 1972) containing 1 mg/l
2,4-D as an auxin source and 0.2 mg/l kinetin. Both B5 and SH basal media are
effective, but SH seems moderately better. Responsive genotypes produce
somatic embryos from subepidermal cells, whereas the callus is initiated from
rapid cell divisions in the vascular cambium (Wenzel and Brown 1991). The
auxins vary in their efficacy to induce this response. IAA is ineffective; NAA will
promote callus formation; and 2,4-D will promote both callus and somatic
embryo initiation (Bingham et al. 1988; Finstad 1992; Stuart and McCall 1992).
There is an apparent time-dose response to auxin; high doses for short periods, or
low doses for longer periods are effective (Dudits et al. 1991).
One well-documented function of auxin is to stimulate the proton pump on
the plasmalemma, leading to hyperpolarization of the membrane (Briskin and
Hanson 1992). This alters solute transport and signaling across the membrane.
Although our information is still rudimentary about this cell signaling process, it
is apparent that potassium plays a key role, which has been almost completely
overlooked in plant tissue culture (with a few notable exceptions: Reinert et al.
1967; Brown et al. 1976; Galiba and Yamada 1988). Our initial interest in this
was sparked by a report by Nichol et al. (1991), indicating that the potassium
salts of citric acid and C4-carboxylic acids stimulated somatic embryogenesis in
alfalfa. Potassium is broadly required for many very important physiological
functions, such as photosynthesis, respiration, enzyme synthesis, and stability as
well as being a major contributor to osmotic potential of cells (Salisbury and
Ross 1985). The possible coupling of auxin with embryogenesis may be
associated with potassium channels and proton pumps. Potassium is closely
associated with signal pathways during animal development (Dubios and
Dubios 1991). In plants, potassium is involved with H+-ATPase linked proton
pumping and therefore is indirectly associated with solute transport and
metabolite allocation (Briskin and Hanson 1992).
Increasing the potassium content of the basal SH induction medium from 25
to 75 mM by the addition of potassium chloride stimulated formation of alfalfa
somatic embryos (Table 1). This stimulation was confirmed with other K +-salts
such as potassium sulfate and potassium phosphate (Shetty and McKersie 1992).
Other cations such as Na+ or anions like Cl-, SO;2, or PO;3 were not involved in
Table 1. Effect of proline, thioproline, and potassium chloride additions to

basal SH induction medium on the number of total somatic embryos
per explant of Medicago sativa L. (after Shetty and McKersie 1992)
Proline addition Potassium addition"
o 50mM
None 4.5 18.3

Proline (100 mM) 7.6 24.6
Thioproline (0.4 mM) 12.5 20.6
Proline and thioproline 17.8 34.9
"Basal SH medium contains 25 mM KN0 3•
the stimulation of alfalfa somatic embryogenesis. Therefore, K+ may be a critical

link in directing hormone-induced somatic embryogenesis, implying that high
activity ofK+-ATPase, K +- transport, and K +-accumulation are needed to direct
hormone-induced somatic embryogenesis in alfalfa more efficiently. Cellular
activity related to K+ could couple auxin-mediated interactions at the cell
membrane with other signal pathways, such as inositol (Boss 1989) and Ca2+_
linked phosphorylations (Blowers and Trewavas 1989).
Previous reports have noted an interaction between potassium and nitrogen
in Daucus (Reinert et al. 1967; Brown et al. 1976). We have also noted an
interaction between potassium and proline in the induction phase of somatic
embryogenesis in alfalfa (Table 1). Proline was particularly interesting because of
its possible link to purine metabolism, and previous reports of its simulation of
embryogenesis in other systems (Shetty and Asano 1991a,b). The earliest in-
dication of somatic embryo formation in petiole explants of alfalfa is a periclinal
cell division about 72 to 120 h after auxin treatment, followed by a series of cell
divisions leading to the formation of globular-shaped embryo (Wenzel and
Brown 1991). Prerequistite to cell division is DNA replication, and therefore one
assumes that during the 72 h prior to the first visual indications of somatic
embryogenesis, DNA synthesis occurs, which requires previous synthesis of
purines. Therefore, increased or altered purine metabolism must be an early
biochemical event in the process. Proline has been shown to stimulate proline-
linked purine metabolism in certain tumor cells that contain proline oxidase
(Phang et al. 1982; Phang 1985; Hagedorn and Phang 1986). Further,
thioproline, an analog of proline, can reduce certain types of animal tumors
(Brugarolos and Gosalvez 1980). Although this pathway has not yet been
identified in plants, proline and its analog thioproline stimulated somatic embryo
production in alfalfa (Table 1). The observed interaction between proline and
potassium suggests that potassium may regulate metabolic pathways related to
purine metabolism and nucleotide turnover.
3.2 Cell Suspension and Embryo Synchronization
After 14 days on the induction medium, the proembryonic cells are embedded in
a mass of callus, and it is necessary to disperse the callus in a suspension to free
these structures. The callus is transferred to liquid Bsg medium (Anandarajah

and McKersie 1992) also containing 1 mg/12,4-D, but no kinetin, at a ratio of I
to 1.5 g callus/25 ml medium. This medium serves to break the mass of callus into
small fragments and allows further cell multiplication and embryo initiation.
Three predominant cell types form in this suspension culture: Large clumps of
nondifferentiated callus and abnormal somatic embryos; small cell clusters
composed of nondifferentiated meristematic cells, which under appropriate
conditions develop into globular embryos; and large "banana" -shaped single
cells, which are terminally differentiated (McKersie et al. 1989). These classes of
cells are separated by screening the suspension first through 500-j.lID nylon screen
and then a 200-~m screen. The latter screen contains small clusters of pro-
embryonic cells which are more or less synchronized in their development.
3.3 Embryo Development
To allow the development of somatic embryos, that fraction collected on the

200-llm screen is spread in a thin layer on hormone-free BOi2Y medium modified
to contain 5% sucrose. After approximately 4 days, green globular and heart-
shaped embryos are apparent protruding from the bed of callus. The
development of these embryo continues up to the late torpedo stage at day 7.
Rapid growth occurs between 7 and 27 days after sieving (Fig. 2). Based on this
pattern of growth, somatic embryo development was divided for experimental
purposes into three phases, namely, development, maturation phase I, and
maturation phase II. In the development phase, the somatic embryo progresses
through the globular and heart-shaped stages to the torpedo stage of
development. Maturation phase I is considered to be the rapid growing in phase
during which most storage reserve deposition occurs, whereas the induction of
desiccation tolerance occurs in maturation phase II (Fig. 2).
A number of factors are critical during the development phase, including
plating density and light intensity. The greatest number of somatic embryos
occurred at plating densities of 10 to 30 mg cells/cm2 (Table 2) but at the low
Table 2. Effect of plating density on development of proembryonic cell

clusters to somatic embryos in Medicago sativa L. (After Anandarajah
and McKersie 1992)
Plating density Cotyledonary Abnormal

(mg/cm2) embryos(%) embryos(%)
5 16 57
10 27 43
30 58 5
60 54 4
Cotyledonary and abnormal classifications were made visually.
Cotyledonary embryos were bottle-shaped with distinctive cotyledons.
Abnormal embryos were green globular structures which were not in an
identifiable heart, torpedo, or cotyledonary stage of development.
Table 3. Effect of light intensity during Medicago sativa somatic embryo development
and maturation on embryo yield, survival after desiccation, and days from sieving to
drying (maturity). (After Anandarajah and McKersie 1992)
Light intensity Embryo yield Survival Maturity

(Ilmolfm2/s) (per mg petiole) (%) (days)
15 9 26 41
35 18 55 34
75 28 92 30
150 0 0 0
75/150' 27 81 25
aIntensity was 75llmolfm2/s during development and 150 !1molfm2/s during maturation
phases.
densities (5 and 10 mglcm2) substantially more somatic embryos appeared

visually abnormal as the result of secondary embryo formation (Anadarajah and
McKersie 1992).
Light intensity has a profound effect when varied during the embryo
development phase, with an optimum at 75 ~mol/m2/s PPFD (Table 3). The
sensitivity to high light intensities was limited to the globular stage of
development. A biochemical explanation for this sensitivity is lacking but
globular somatic embryos are missing one Cu/Zn superoxide dismutase isozyme
which may reduce the embryo's ability to scavenge activated forms of oxygen
produced from chlorophyll (V Vleeshouwers and BD McKersie, unpubl.). When
embryos were grown at 75 ~oVm2/s PPFD on the development medium and
then transferred to maturation medium at 150 ~moVm2/s PPFD, the high light
intensity hastened somatic embryo development by as much as 16 days between
15 and 75/150 (elongation/maturation) ~01lm2/s PPFD (Table 3). On a
practical level, the critical importance of optimal light conditions is often
overlooked in cell cultures, but the present data emphasize that practices such as
stacking plates in incubators will introduce substantial variation in culture
response.
3.4 Embryo Maturation and Deposition of Storage Reserves
The most rapid increase in embryo fresh weight occurs between 7 and 17 days
after sieving, after the somatic embryo reaches the torpedo stage (Fig. 2) and it
is during this period that the majority of storage reserves in the form of
carbohydrates and protein are deposited. The deposition of storage reserves is
facilitated by transfer of the late torpedo stage embryo to BOi2Y medium
containing 50 to 100 mM glutamine and 50 gil sucrose (Anandarajah and
McKersie 1990a,b; Lai et al. 1992).
Sucrose and glutamine have independent additive effects on somatic embryo
dry weight, and maximal weight was recorded at 50 mM glutamine and 50 gil
sucrose (Table 4). Sucrose did not affect the accumulation of storage proteins in
alfalfa somatic embryos (Fig. 2; Lai et al. 1992) but increased the levels of starch
Table 4. Effects of sucrose and glutamine applied during maturation

phase I on dry weight of Medicago sativa L. somatic embryos
Sucrose Glutamine Drywt.

gil (mM) (mg)
30 0 1.47
30 50 1.68
50 0 2.22
50 50 2.55
in the embryos to approximately 10% dry wt. (F. Lai and B.D. McKersie,
unpubl.). A twofold increase in the absoulte amount ofS-2 protein in the mature
somatic embryos was observed when 50 mM glutamine was included in the
phase I maturation medium (Lai et al. 1992).
In order to characterize the patterns of protein synthesis during maturation,
proteins were extracted into two fractions. The S-1 (low salt-soluble) protein
fraction contains water-soluble enzymatic proteins, and the high molecular
weight (HMV), low molecular weight (LMW), and alfin (7S) storage proteins.
The S-2 (high salt-soluble) protein fraction contains predominantly the
medicagin (11S) storage proteins (Krochko and Bewley 1990). Compared to
seeds, the somatic embryos contained relatively low levels of the S-2 protein
fraction but the polypeptide composition of the fraction was similar. Five major
maturation
development I phase I phase II I
~.~... ............. ~ ..... ............. ~
globular torpedo cotyledon late cotyledon
-..
.10
II
o
1:1. .....
III
.z:"iii
() u
.. UJ
,f! II
UJ >
~;;
.c III
J!! 'i
III
....II:
..
~
>-
c
o 7 17 27
Days after sieving
Fig. 2. Accumulation of dry weight, starch, and storage protein during the development and
maturation of alfalfa somatic embryos
acidic polypeptide bands (AI, A2, A3, A5, and A6) and three basic peptide
components BI/B2, and B3) were resolved from the S-2 protein fraction in the
somatic embryos. The addition of 50 mM glutamine increased the level of both
the acidic and the basic polypepide components of the 11 S storage protein (Lai
et al. 1992).
Gultamine is converted to 5-oxoproline (pyroglutamic acid) and ammonium
by autoclaving the culture medium, and it is the 5-oxoproline that affects somatic
(a) II III IV V VI VII VIII
kD
66 -
HMW2
-"']
45 -
36 -
a2 7S
29 - a3 (allin)
24 -
20 -
a4
14 -
LMW
(S-1 )
(b) V VI VII VI II
II III IV
kD
66 -
45 -
36 -
29 - 11S
24 - (medicagin)
20 -
14 -
(S-2)
Fig. 3. SDS-PAGE analysis of the effect of glutamine, 5-oxoproline, and sulfate additions during
maturation of alfalfa somatic embryos on storage protein accumulation. Lane I control; lane II 50 mM
autoclaved glutamine (5-oxoproline and ammonium); Lane III 50 mM filter-sterilized glutamine; Lane
IV 50 mM autoclaved glutamine + 25 mM ammonium sulfate; Lane V 50 mM autoclaved glutamine
+ 25 mM potassium sulfate; Lane VI 50 mM 5-oxoproline + 25 mM ammonium nitrate; Lane VII 50
mM 5-oxoproline +25 mM ammonium nitrate + 25 mM ammonium sulfate; Lane VllI 50 mM 5-
oxoproline +25 mM ammonium nitrate + 25 mM potassium sulfate
embryo maturation. Although the addition of filter-sterilized glutamine to the

maturation medium increased S-l and S-2 proteins in mature embryos compared
to the control, these embryos elongate (precociously germinate) on maturation
medium even in the presence of 20 /-lM ABA, and therefore did not survive
desiccation. Sulfate salts, in combination with 5-oxoproline, enhanced the
deposition of all storage proteins, including the 7S alfin, the 2S LMW storage
proteins in the S-l fraction, and the lIS medicagin in the S-2 fraction (Fig. 3). In
contrast, ammonium nitrate increased the deposition of only the alfin storage
protein, but not the deposition of the LMW protein in the S-l fraction nor the
BlIB2 (basic polypeptides) subunits of medicagin in the S-2 fraction. The
elevated levels of LMW and medicagin storage proteins positively correlated
with enhanced embryo quality, desiccation tolerance, germination, and
converSIOn.
These nutrient responses are very much dependent on the density of embryos
on the maturation medium. The embryogenic potential of alfalfa is very high and
the tissue culture system previously described can yield up to 1000 globular
somatic embryos on a Petri plate (15 cm diameter). If all 1000 somatic embryos
are to develop to 2 mg dry wt. with a 20% protein content, then the medium must
provide a total of 400 mg protein or 25 mg N. One Petri plate of the basal BOi2y
maturation medium contains 25 ml of 26 mM N03, 12 mM NH;, 27/-lM
glycine, and 2 gil yeast extract providing an approximate amino acid content of
3mM, which in total provides only 14 mg N. A 50 mM glutamine supplement
provides an additional 35 mg N, which clearly overcomes the nutritional lim-
itation and partially overcomes the block in storage protein synthesis (Lai et al.
1992). The quantity as well as the quality of the nutritional components of the
medium influences embryo maturation, and is a factor which must be recognized
if the goal of the tissue culture procedure is to produce a dry artificial seed
containing a comparable quantity of storage reserves to true seeds.
3.5 Desiccation Tolerance
Finally, the embryos are transferred to basal BOi2Y medium containing 20 /lM
ABA (maturation phase II) at which time the embryos acquire desiccation
tolerance (Senaratna et al. 1990). ABA serves several functions. The first is to
prevent precocious germination. The second is to induce a genetic program in the
embryo that initiates biochemical and physical changes in the way the proto-
plasm binds water and its tolerance of water loss.
The somatic embryo acquires sensitivity to ABA at approximately the late
torpedo or early cotyledonary stages of development, just prior to precocious
germination. Earlier or later applications do not induce the expression of
desiccation tolerance. A number of other factors can replace ABA, such as heat
shock, partial drying, or osmotic stress with polyethyleneglycol. These treat-
ments act by inducing the somatic embryo to synthesize ABA (Senaratna
et al. 1989; McKersie et al. 1990). The role of ABA seems to be common in many
species including alfalfa, canola (Senaratna et al. 1991), geranium (Marsolais
et al. 1991), and spruce (Attree et al. 1991).
Not only are there positive factors which promote the acquisition of
desiccation tolerance, but there are negative factors which prevent or reduce
survival and vigor after desiccation. For example, the presence of ammonium or
maltose in the maturation medium with ABA reduced embryo survival and vigor
after drying (Anandarajah and McKersie 1990b).
Once the somatic embryos have been induced to express desiccation
tolerance, the cotyledonary embryos are dried slowly by transfer through
atmospheres of progressively reduced relative humidity. The embryos are
routinely stored between 10 and 15% moisture content at room temperature for
several months.
Physically drying somatic embryos can be accomplished by simply air drying
the embryos in a laminar flow bench overnight, once the embryos have acquired
full tolerance. However, much more consistent responses, and improvements in
embryo quality, occur if the embryos are slowly dried with a progressive, linear
loss of water (Senaratna et al. 1989). For slow drying, a series of relative
humidities are generated in desiccators over saturated salt solutions; the embryos
in Petri plate with no nutrient medium are equilibrated at each humidity for 1 day
and then air dried to a final moisture of 10-15%.
3.6 Desiccation and Water Binding
Seed desiccation is a natural event in the life cycle of many plant species. Seeds
that are tolerant to desiccation are known as orthodox, while those that cannot
withstand desiccation are recalcitrant. The role of desiccation in embryo
development is not fully defined. Desiccation may be just one of several
temporally discrete developmental program occurring within the embryo, each
program initiated by a separate endogenous factor inducing the expression of a
particular set of genes (Galau et al. 1991). Alternatively, seed desiccation may be
a bypass of the plant's ordinary developmental program due to physical and
environmental restraints on the embryo (Walbot 1978). The embryo may be
genetically programmed to germinate while still immature but the genes required
for germination are transcribed only when the embryo is removed from the plant
(Dure 1975). Thus, the basic control of zygotic embryo desiccation appears to be
genetic with regulation by hormones and the embryo environment.
Cellular changes that occur during desiccation involve structural alterations
(Webb and Arnott 1982), fluctuations in the type of water binding (Vertucci and
Leopold 1984), and shifts in molecular organization and gene expression
(Bianchi et al. 1991; Shin et al. 1991).
Water sorption can be defined using the D'Arcy/Watt equation:
W -- KK'(P/po) + c(P / Po ) + kk'(p/po) ,

I+K(p/po) 1 -k(P/po)
where W is the amount of water sorbed per gram of tissue (D' Arcy and Watt
1970). The other terms are as follows:
p/po - relative vapor pressure;

K - relates to the attraction of strong binding sites for water;
K' - relates to the number of strong binding sites;
c - relates to the number and strength of weak binding sites;
k - relates to the water activity of multimolecular water; and
k' - relates to the number of multi molecular sorption sites.
A typical isotherm contains three regions corresponding to different types of
water binding: strong binding at RVP (relative vapor pressure) of 0 to 0.2, weak
binding atRVP of 0.20 to 0.60, and multimolecular water above an RVP of 0.60.
For desiccation tolerant seeds, the isotherm is reverse sigmoidal in shape,
whereas intolerant seeds exhibit isotherms which are often hyperbolic or linear in
shape (Vertucci and Leopold 1987 a,b).
To compare the water-binding characteristics, zygotic embryos (removed
from the endosperm) and somatic embryos were desiccated over saturated salt
solutions at 0.76 RVP (NaCl), 0.63 RVP NH4N0 3, 0.51 RVP [Ca(N03)2] 0.43
RVP (K2C0 3oHP), 0.25 RVP (CH 3COOK), 0.08 RVP (LiCl), and 0 RVP (Gel
Rite silica stones). Each sample was weighed after 24-28 h in each desiccation
chamber and then moved to the next chamber.
Desiccation tolerance was associated with changes and differences in the
strength and number of strong water-binding sites in the embryos. The region of
Water Binding Isotherms
5.0 Zygotic embryos Somatic embryos
3.0
4.0
§'
c 3.0
--
01 2.0
S
c
CD 2.0
c
0
u
"- 1.0
CD intolerant
1ii 1.0
:=
tolerant .' tolerant
o 0.2 0.6 o 0.2 0.6
Relative Vapour Pressure
Fig. 4. Water binding in zygotic and somatic embryos of alfalfa recorded as the amount of water
bound at varying relative vapor pressures
Strong Binding Sites
INTOLERANT
1.2 1.2
ZYGOTIC SOMATIC
EMBRYOS 1.0 EMBRYOS
1.0
0.8 0.8
§'
0 0.6
J!I 0.6
S
~ 0.4 0.4
INTOLERANT
0.2 0.2
0.0 0.0
Fig. 5. The relative number of strong water-binding sites in zygotic and somatic embryos which are
intolerant and tolerant of desiccation
strong binding was not present in intolerant embryos, as evidenced in the linear
loss of water between 0.20 R VP, but gradually appeared as the embryos matured
(Fig. 4). The strength and number of strong binding sites was reduced in a
qualitatively similar manner as embryos acquired desiccation tolerance (Fig. 5).
Differences in water binding were apparent between the mature embryos.
Mature, desiccation-tolerant somatic embryos did not follow the typical
isotherm in the strong binding region; the shape was concave instead of convex.
Water loss from somatic embryos was complete at 0.08 RVP, whereas the zygotic
embryos still retained substantial water at this humidity. It is possible that the
somatic embryos did not contain strong water-binding sites, or that these sites
were lost as desiccation proceeded beyond a critical point. Regardless, somatic
embryos seem to be qualitatively different from zygotic embryos in the way they
bind water at low RVP. This may have significance because differences in water
binding may be associated with longevity in the dry state and/or vigor after
imbibition. These differences in water binding reflect biochemical and composi-
tional differences between somatic and zygotic embryos. This may be the result of
inherently different mechanisms of desiccation tolerance between the two types
of embryos, or simply differences in the quality and quantity of storage reserves.
4 The Water Replacement Hypothesis
Water is the universal solvent in biological systems and stabilizes the molecular
bilayer structure of cell membranes through hydrophobic and hydrophilic inter-
actions with the amphipathic lipid molecules. Therefore, removal of water from
the membrane surface would logically be expected to change membrane structure
and promote injury symptoms such as cytoplasmic solute leakage (Simon 1974).
In sensitive plant cells, the membranes may form gel phase domains at low water
contents, which are readily reversed on rehydration (Crowe et al. 1989). This
reversible lipid phase change may be lethal if there is extensive solute leakage,
reorganization of membrane protein complex, or other cellular damage during
dehydration or rehydration. The basis of the water replacement hypothesis lies in
the replacement of structural water on membranes or macromolecules during
desiccation by polyhydroxyls such as sugars and sugar alcohols (Crowe and
Crowe 1988). In dehydration-tolerant cells, the hydroxyl groups of the sugar
form hydrogen bonds with the polar head groups of the lipid, providing the
hydrophilic interactions necessary to maintain the liquid-crystalline phase. In
animal and microbial species, the disaccharide trehalose has been found to
replace water (Crowe and Crowe 1988). A temporal correlation between the loss
of desiccation tolerance and decreased sucrose content has been reported in
germinating soybean, pea, and corn axes (Koster and Leopold 1988).
Furthermore, the sucrose content in Papaver dubium L. pollen increased with the
acquisition of desiccation tolerance (Hoekstra and van Roekel 1988). Sugar
measurements in developing alfalfa seeds and somatic embryos have not con-
firmed that accumulation ofthese carbohydrates is essential for the acquisition of
desiccation tolerance (Table 5), but measurements at the seed or embryo level
Table 5. Sugar composition (mg/g dry wt.) of alfalfa seeds at desiccation-intolerant (24 days
after pollination) and tolerant (36 days after pollination), and somatic embryos at
desiccation-intolerant (without abscisic acid) and tolerant (with abscisic acid) stages. (Data
from K. Anandarajah, T. Senaratna, and BD. McKersie, unpub!.)
Sugar Seed Somatic embryo
24 days 36 days -ABA +ABA
Stachyose 2 2 nd nd
Sucrose 13 11 23 24
Glucose 25 7 37 3
Fructose 15 9 18 9
Galactose 14 3 0.5 0.6
Table 6. (l- Tocopherol content of alfalfa seeds at desiccation-intolerant (24 days after
pollination) and -tolerant (36 days after pollination) and somatic embryos at
desiccation-intolerant (without abscisic acid) and tolerant (with abscisic acid) stages.
(Data from K. Anandarajah, T. Senaratna, and BD. McKersie, unpub!.)
Seeds Somatic embryos

24 days 36 days -ABA +ABA
Dry weight (nmol/g) 108 337 443 785

Phospholipid (mmol/mol) 5.9 14.7 22.7 35.8
may not adequately reflect the subcellular changes in these carbohydrates and
their association with membranes. However, the monosaccharides glucose
and fructose decreased in these embryos and in all tissues so far reported as
they acquired tolerance (K Anandarajah, T Senaratna and BD McKersie,
unpubl.). The significance of this observation is that these are reducing sugars
which react with metal ions such as iron to generate oxidizing agents in reactions
such as the Maillard reaction. The involvement of oxygen-based free radicals in
desiccation tolerance and seed longevity has been shown in membrane studies
(Senaratna et al. 1988; McKersie 1991). Dehydration of sensitive cells promoted
lipid peroxidation and other oxygen-free radical reactions causing the
accumulation of free fatty acids and other products in the membrane bilayer.
These lipid degradation products increase the lipid phase transition temperature,
and thereby cause the irreversible formation of gel phase domains, which are
lethal when the cell is rehydrated. It is therefore interesting to note that in alfalfa,
a-tocopherol accumulated approximately 2.5 fold in seeds and somatic embryos
coincident with the acquisition of desiccation tolerance (Table 6). Tolerance of
desiccation in germinating soybean seeds has been associated with the presence of
antioxidants which scavenge free radicals before the radicals attack biomolecules
(Senaratna et al. 1985; McKersie 1991).
5 Conclusions
The tissue culture procedures which are necessary to produce a dry quiescent
somatic embryo differ significantly from standard somatic embryogenesis
procedures in the complexity required to achieve synthesis of storage reserves and
acquisition of desiccation tolerance. If the dry somatic embryos are to be used as
artificial seeds, they require comparable quantities of storage reserves as conven-
tional seed to support germination and early seedling growth in competitive and
stressful environments. Somatic embryos do not readily synthesize storage
reserves and comparatively little research has been done to elucidate the control
of this process in somatic embryos. Part of the limitation in highly embryogenic
systems such as alfalfa is simply inadequate quantities or species of nutrients in
the medium. In addition, the signals required to sustain synthesis and deposition
of reserves are not well defined, although 5-oxoproline and sulfate may playa
role in this regard. Therefore, the maturation process for dry somatic embryos is
by necessity complex because the somatic embryo requires significant changes in
the levels and types of nutrients in the medium during its development.
Somatic and zygotic embryos develop through analogous stages which are
visually similar such as the globular, heart-shaped, and torpedo stages. Somatic
embryos differ from their zygotic counterparts in a number of ways. For
example, cotyledon development is rudimentary. Storage protein accumulation
is low. Although both types of embryos acquire desiccation tolerance, it seems
that they bind water in distinctly different ways at low relative vapor pressures.
This may indicate significantly different mechanisms of desiccation tolerance,
different chemical composition, or different longevity in the dry state. These

differences may reflect abnormal embryonic development, on the part of the
somatic embryo, as a result of poorly defined tissue culture media, or they may
reflect inherent differences in the maturation of the embryos. A more thorough
understanding of the control of the developmental processes in zygotic and
somatic embryos is required before our objective of plant propagation by dry
artificial seeds can be achieved.
Acknowledgments. The authors gratefully acknowledge the financial assistance of the Natural Sciences
and Engineering Council of Canada (Strategic Grant ), the University Research Incentive Fund of the
province of Ontario, and Somatica Plant Technologies in the conduct of this research. Susan Van
Acker is a recipient of the McConkey Scholarship for postgraduate studies.
References
Anandarajah K, McKersie BD (I 990a) Enhanced vigor of dry somatic embryo of Medicago sativa L.
with increased sucrose. Plant Sci 71: 261-266
Anandarajah K, McKersie BD (1990b) Manipulating the desiccation tolerance and vigor of dry
somatic embryos of Medicago sativa L. with sucrose, heat shock and abscisic acid. Plant Cell Rep
9: 451-455
Anandarajah K, McKersie BD (1992) The influence of plating density, sucrose and light during
development on the germination and vigour of Medicago sativa somatic embryos after desiccation.
Seed Sci Res 2: 133-140
Attree SM, Moore D, Sawhney VR, Fowke LC (1991) Enhanced maturation and desiccation tolerance
of white spruce [Picea glauca (Moench) Voss] somatic embryos-effects of non-piasmoly sing water
stress and abscisic acid. Ann Bot 68: 519-525
Bianchi G, Murelli C, Bochicchio A, Vazzana C (1991) Changes oflow-molecular weight substances
in Boea hygroscopia in response to desiccation and rehydration. Phytochemistry 30, 2: 461-466
Bingham ET, McCoy TJ, Walker KA (1988) Alfalfa tissue culture. In: Hanson AA, Barnes DK, Hill
RR (eds) Alfalfa and alfalfa improvement. American Society of Agronomy Madison, WI, pp
903-929
Blowers DP, Trewavas AJ (1989) Second messengers: their existence and relationship to protein
kinases. In: Boss WF, Morre DJ (eds) Second messengers in plant growth and development. Alan
R Liss, New York, pp 1-128
Boss WF (1989) Phosphoinositide metabolism: its relation to signal transduction in plants. In: Boss
WF, Morre DJ (eds) Second messengers in plant growth and development, Alan R Liss, New York,
pp 29-56
Briskin D, Hanson JB (1992) How does the plant plasma membrane H+ -ATPase pump protons? J Exp
Bot 43: 269-289
Brown DC, Atanassov A (1985) Role of genetic background in somatic embryogenesis in Medicago.
Brown S, Whetherell DF, Dougall DK (1976) The potassium requirement for growth and
embryogenesis in wild carrot suspension cultures. Physiol Plant 37: 73-79
Brugarolas A, Gosalvez M (1980) Treatment of cancer by an inducer of reverse transformation. Lancet
I: 65-70
Crowe LM, Crowe JH (1988) Effect of water and carbohydrates on membrane fluidity. In: Aloia RC,
Curtain CC, Gordon LM (eds) Physiological regulation of membrane fluidity. Alan R Liss, New
York, pp 75-99
Crowe JH, Hoekstra FA, Crowe LM (1989) Membrane phase transitions are responsible for
imbibitional damage in dry pollen. Proc Natl Acad Sci USA 86: 520-535
D'Arcy RL, Watt IC (1970) Analysis of sorption isotherms of non-homogeneous sorbents. Trans
Faraday Soc 66: 1236-1245
Dubios BR, Dubios 1M (1991) A quantitative analysis of the role of K + channels in mitogenesis of
neuroblastoma cells. Cell Signal 3: 333 339
Dudits D, Bogre L, Gyorgyey 1 (1991) Molecular and cellular approaches to the analysis of plant
embryo development from somatic cells in vitro. 1 Cell Sci 99: 475-484
Dure LS (1975) Seed formation. Annu Rev Plant Physiol26: 259-278
Finstad K (1992) Biochemical and developmental markers of induction of somatic embryogenesis in
alfalfa tissue culture. PhD Thesis, Carleton University, Ottawa
Galau GA, lakobsen KS, Hughes DW (1991) The controls of late dicot embryogenesis and early
germination. Physiol Plant 8 J: 280--288
Galiba G, Yamada Y (1988) A novel method for increasing the frequency of somatic embryogenesis in
wheat tissue culture by NaCl and KCI supplementation. Plant Cell Rp 7: 55-58
Gray Dl, Purohit A (l99Ia) Somatic embryogenesis and development of synthetic seed technology.
Crit Rev Plant Sc 10: 33-61
Gray Dl, Purohit A (1991b) Quiescence and dormancy in somatic embryos In: Bajaj YPS (ed)
Biotechnology in agriculture and forestry, vol 17. High-tech and micropropagation 1. Springer,
Hagedorn CH, Phang 1M (1986) Catalytic transfer of hydride ions from NADPH to oxygen by the
interconversion of proline and A'-pyrroline-5-carboxylate. Arch Biochem Biophys 248: 166-174
Hoekstra FA, van Roekel T (1988) Desiccation tolerance of Papaver dubium 1. pollen during its
development in the anther. Possible role of phospholipid composition and sucrose content. Plant
Physiol 88: 626-632
Kermode AR, Bewley ID (1985) The role of maturation drying in the transition from seed development
to germination. 1 Exp Bot 36:1906-1915
Kermode AR, Oishi MY, Bewley ID (1989) Regulatory roles for desiccation and abscisic acid in seed
development: a comparison of the evidence from whole seeds and isolated embryos. In: Stanwood
PC, McDonald MB (eds) Seed moisture. CSSA Spec Publ 14 Crop Science of America Inc,
Madison, pp 23-50
Koster, KL, Leopold AC (1988) Sugars and desiccation tolerance in seeds. Plant Physiol 88: 829·-832
Kris MHS, Bingham ET (1988) Interactions of highly regenerative genotypes of alfalfa (Medicago
sativa) and tissue culture protocols. In Vitro Cell Dev BioI 24: 1047-1051
Krochko JE, Bewley ID (1990) Identification and characterization of the seed storage proteins from
alfalfa (Medicago sativa). 1 Exp Bot 41: 505-514
Lai F, Senaratna T, McKerise BD (1992) Glutamine enchances storage protein synthesis in Medicago
sativa 1. somatic embryos. Plant Sci 87: 69-77
Marsolais AA, Wilson DPM, Tsuijita Ml, Senaratna T (1991) Somatic embryogenesis and artificial
seed production in zonal (Pelargonium x hortorum) and regal (Pelargonium x domesticum)
geranium. Can 1 Bot 69: 1188-1193
McKersie BD (1991) The role of oxygen free radicals in mediating freezing and desiccation stress in
plants In: Pell E, Steffer K (eds) Active oxygen/oxidative stress and plant metabolism. American
Society of Plant Physiologists, Rockville MD, pp 107-118
McKersie BD, Bowley SR (1993) Synthetic seeds of alfalfa. In: Redenbaugh K (ed) Synseeds:
McKersie BD, Senaratna T, Bowley SR, Brown DCW, Krochko E, Bewley ID (1989) Application of
artificial seed technology in the production of hybrid alfalfa (Medicago sativa). In Vitro Cell Dev
BioI 25: 1183-1188
McKersie BD, Senaratna T, Bowley SR (1990) Drying somatic embryos for use as artificial seeds. In:
Proc Plant Growth Regul Soc Am 17: 199-207
Nichol JW, Slade D, Viss P, Stuart DA (1991) Effect of organic acid pretreatment on the regeneration
and development (conversion) of whole plant from callus cultures of alfalfa. Medicago sativa 1.
Plant Sci 79: 181-192
Phang JM (1985) The regulatory functions of proline and pyrroline-5-carboxylic acid. Curr Top Cell
Regul25: 19-132
Phang 1M, Downing SJ, Yeh GC, Smith RJ, Williams JA, Hagedorn CH (1982) Stimulation of hexose
monophosphate-pentose pathway by pyrroline-5-carboxylate in cultured cells. J Cell Physiolll 0:
255-261
Redenbaugh K, Fujii lA, Slade D (199Ia) Synthetic seed technology. In: Vasil IK (ed) Cell culture and
somatic cell genetics of plants, vol 8. Academic Press, New York, pp 35-74
Redenbaugh K, Fujii J, Slade D, Viss P, Kossler M (l991b) Artificial seeds - encapsulated somatic
micropropagation I. Springer, Berlin Heidelberg New York, pp 395--416
Reinert J, Tazawa M, Semenoff S (1967) Nitrogen compounds as factors of embryogenesis in vitro.
Nature 216:1215-1216
Salisbury FB, Ross CW (1985) Mineral nutrition. In: Plant physiology 3rd edn. Wadsworth, Belmont,
CA, pp 96--113
Saunders JW, Bingham ET (1972) Production of alfalfa plants from callus tissue. Crop Sci 12: 804-808
Schenk RU, Hildebrandt AC (1972) Medium and techniques for induction and growth of
monocotyledonous and dicotyledonous plant cell culture. Can J Bot 50: 199-204
Senaratna T, McKersie BD, Stinson RH (1985) Antioxidant levels in germinating soybean seed axes in
relation to free radical and dehydration tolerance. Plant Physiol 78: 168-171
Senaratna T, Gusse J, McKersie BD (1988) Age-induced changes in cellular membranes of imbibed
soybean seed axes. Physiol Plant 73: 85-91
Senaratna T, McKersie BD, Bowley SR (1989) Desiccation tolerance of alfalfa (Medicago sativa L.)
somatic embryos. Influence of abscisic acid, stress pretreatments and drying rates. Plant Sci 65:
253-259
Senaratna T, Mckersie BD, Bowley SR (1990) Artificial seeds of alfalfa (Medicago sativa L.).
Induction of desiccation tolerance in somatic embryos. In Vitro Cell Dev Bioi 16: 85-90
Senaratna T, Kott L, Beversdorf WD, McKersie BD (1991) Desiccation of microspore derived
embryos of oilseed rape (Brassica napus L.). Plant Cell Rep 10: 342-344
Shetty K, Asano Y (199Ia) Specific selection of embryogenic cell lines in Agrostis alba L. using the
proline analog thioproline. Plant Sci 79: 259-263
Shetty K, Asano Y (1991 b) The influence of organic nitrogen sources on the induction of embryogenic
callus in Agrostis alba L. J Plant Physiol 139: 82-85
Shetty K, McKersie (1992) Proline, thioproline and potassium mediated simulation of somatic
embryogenesis in alfalfa (Medicago sativa L.) Plant Sci 88: 185-193
Shin DH, Virigool S, Shinozaki KY, Dono K (1991) Survival mechanism of dried calli and
regeneration of plants in rice. Jpn J Genet 66: 13-25
Simon EW (1974) Phospholipids and plant membrane permeability. New Phytol 73: 337--420
Stuart DA, McCall CM (1992) Induction of somatic embryogenesis using side chain and ring modified
forms of phenoxy acid growth regulators. Plant Physiol 99: 111-118
Vertucci CW, Leopold AC (1984) Bound water in soybean seed and its relation to respiration and
imbibitional damage. Plant Physiol 75: 114-117
Vertucci CW, Leopold AC (1987a) Water binding in legume seeds. Plant Physiol85: 224-231
Vertucci CW, Leopold AC (I 987b) The relationship between water binding and desiccation tolerance
in tissues. Plant Physiol 85: 232-238
Walbot V (1978) Control mechanisms for plant embryogeny. In: Clutter ME (ed) Dormancy and
developmental arrest. Academic Press, New York, pp 113-164
Webb MA, Arnott HJ (1982) Cell wall conformation in dry seeds in relation to the preservation of
structural integrity during desiccation. Am J Bot 69: 1657-1668
Wenzel CL, Brown DCW (1991) Histological events leading to somatic embryo formation in cultured
petioles of alfalfa. In Vitro Cell Dev Bioi 27: 190--196
11.4 Artificial Seed Production Through Encapsulation
of Hairy Root and Shoot Tips
N. UOZUMI and T. KOBAYASHII
1 General Introduction
The delivery system, using artificial seed, is of great value in germplasm storage,
propagation, and in the production of chemicals. Artificial seed for plant
propagation has been investigated in somatic embryogenesis (Kitto and Janick
1985; Redenbaugh et al. 1991). We have extended this concept to the hairy root
because of the successful regeneration of the whole plant from Ri-transformed
cells. The micro propagation system using these hairy roots promises practical
application in the fields of cellular biology, agriculture, and bioengineering.
Agrobacterium rhizogenes is responsible for hairy root induction. The
phenomenon is due to the transfer, integration and expression in the plant cell
genome of DNA (T-DNA) originating from a large plasmid called Ri (root-
inducing plasmids (Tepfer 1990). Genetic modification using A. rhizogenes
plasmids as vectors is believed to be feasible for the improvement of plant
properties and for the production of transgenic plants. The induced hairy root
has other superior properties, such as higher inherent genetic stability and
growth rate increment than tissue induced by growth regulators.
Since tissue-specific behavior and production are organized, the specific
tissue or organ (e.g., root) is suitable for production of secondary metabolites
and useful chemicals in hairy root. Some plants regenerated from hairy root have
a higher content of the target metabolites in leaves. In this case, regeneration
from the roots is necessary for the production ofthe chemicals. Considering these
properties of hairy root, transformed "hairy root" provides a promising
alternative to the biotechnological exploitation of plant cells.
In this chapter, attention is focused on the development of a micro-
propagation procedure using the hairy root in an artificial seed system from the
standpoint of bioengineering (Uozumi et al. 1992b). The use of the following
explants is discussed:
1. hairy root or root fragment;
2. adventitious shoot primordia formed in the dark;
3. plantlet produced from hairy root in the light.
I Department of Biotechnology, Faculty of Engineering, Nagoya University, Chikusa-ku, Nagoya
464-01, Japan

Artificial Seed Production Through Encapsulation of Hairy Root and Shoot Tips 171
The regeneration frequency from horseradish hairy roots as a model species

is evaluated by excision and encapsulation combined with supplementation of
growth regulators to develop the manipulation procedure for artificial seeds.
2 Materials and Methods
Root Culture and Medium. Horseradish (Armoracia rusticana) hairy root-in-

duced by the leaf-disk method (Tanaka et al. 1985; Noda et a11987; Taya et al
1989b) was used in all experiments. The hairy root was maintained by regular
subculture in the dark, over 3 weeks at 25 cC, on hormone-free MS medium
(Murashige and Skoog 1962) supplemented with 2% (w/v) sucrose. It was then
transferred aseptically to an Erlenmeyer flask containing MS medium with 2%
sucrose (the final inoculum volume of hairy roots was about 2 gil fresh weight) on
a rotary shaker (100 rpm). The solid medium contained 1.5% agar. Under light
conditions the photoperiod (ca. 45 ~mollmls), with a fluorescent white light tube,
was 14 h a day.
Encapsulation. A suspension of excised explants (root fragments, adventitious
shoot primordia, plantlets) in a twofold concentrated MS medium containing 4%
sucrose and the required growth regulators was mixed at room temperature with
the same volume of 4% (w/v) sterilized sodium alginate. The mixture was added
dropwise to a sterile solution of 100 mM CaCl2 using pipettes, thus forming
calcium alginate beads of a size large enough to cover an entire root fragment or
an adventitious shoot primordium. The resultant beads were rinsed with a small
amount of water and then placed directly on agar plates in the light.
Scanning Electron Microscopy. Explants were fixed in 2% glutaraldehyde in
10 mM phosphate buffer (pH 7) for 2 h at room temperature. Afterwards, they
were dehydrated with an increasing acetone series and then soaked in isoamyl
acetate for 4 h. After critical-point drying and sputtering, micrographs were
made using a Hitachi S-570 scanning electron microscope.
3 Results
3.1 Regeneration from Encapsulated Root Fragments
3.1.1 Dependency of Shoot Formation Frequency on Portions

of Hairy Roots
From the preliminary experiments, shoot formation occurred over the entire
root except the root apical meristem. A larger number of shoots occurred at
that portion further away from the root apical meristem.
172 N. Uozumi and T. Kobayashi
After the root was cut into various fragments (1 to 10 mm) and encapsulated
with alginate, the shoot formation frequency was examined. In this case, the
excised root fragment, including the root apical meristem, exhibited a high shoot
formation frequency. In the root fragment containing a lateral root, shoot
formation was observed at the center of the beads, and the shoot formation
frequency was comparable with that of the apical meristem fragment. The shoot
formation frequency of root fragments without a root apical meristem and
lateral root (intermediate portion) was significantly low.
We determined the minimum root length which will enable shoot formation,
choosing the root fragment with the apical meristem as the encapsulated root
fragment. The shoot formation frequency increased with increasing root length
up to 5 mm (data not shown). There was no increased formation frequency when
fragments were longer than 5 mm. The results indicated that root fragments of
5 mm length were suitable for encapsulations, thus they were used in subsequent
experiments.
3.1.2 Effects of Auxin on Root Morphology and Shoot Formation Frequency
For somatic embryogenesis, the callus is maintained in medium supplemented

with auxin, and is then transferred to auxin-free medium. In preliminary
experiments, it was found that auxin stimulated the emergence of the root apical
meristem and lateral roots. In order to harvest a large number of root fragments
which have root apical meristems or lateral root portions from the whole root
culture, we examined the effects of auxin on the morphological changes in the
root (Table I). After the root was treated with various concentrations of auxin,
the randomly picked root fragments were encapsulated and transferred to MS
medium in the light. The root treated with 0.1 and I mg/l naphthalene acetic acid
Table 1. Induction of branching (lateral root formation) and increase in shoot formation frequency by
auxin supplementationa
Nontreated NAA NAA NAA IAA IBA 2,4-D
(mg/I) 0.1 10
Shoot formation
frequency (%) 21 72 60 19 3 60 0
Branching (lateral
root) (%)b 17 70 58 16 3 60 0
Others (%)' 4 2 2 3 0 0 0
Percentage oflateral
root portion (%)d 60 85 88 53 58 88 50
a All cultures were raised for 24 days.

b Shoot formation frequency from root fragments having branching (later root).
, Shoot formation frequency from root fragments without branching (lateral root).
d The number of beads included root fragments with branching (lateral root)/total number of beads
(Uozumi et al. I 992b).

(NAA) or 1 mg/l indole-3-butyric acid (IBA) showed a markedly high shoot

formation frequency, compared with the nontreated root. In particular, the
highest shoot formation frequency was obtained when the root was placed on
medium containing 0.1 mg/1 NAA. On the other hand, 2,4-dichloro-
phenoxyacetic acid (2,4-D) and indole-3-acetic acid (IAA) resulted in a decrease
in shoot formation frequency.
To explain the effect of auxin treatment on shoot formation, the frequency of
NAA-treated roots with or without lateral roots was also examined. With the
hairy root containing a lateral root at day 24, shoot formation was significantly
higher than the root without branching. There was an apparent correlation
between lateral root emergence and shoot formation frequency. The higher
frequency of shoot formation was due to the large number of lateral roots
induced by NAA (O.l-lmg/I) or IBA (1 mg/I) treatment.
As described previously, the shoot formed from branch portion of the root
fragment in the bead, and then two leaves grew in the liquid preculture without
NAA. Healthy plantlets were also grown in the preculture with 0.1 mg/l and
1 mg/l NAA, which appeared larger than the non treated plantlet. However,
abnormal morphologies of the plantlets were observed in the liquid preculture
with 5 mg/l NAA. From the plantlet development frequency and morphological
observations, the optimum NAA concentration in the preculture was thus
determined to be 0.1 mg/I.
After encapsulation, NAA inhibited plantlet formation extensively (data not
shown). Hence, for lateral root emergence, plantlet development required the
removal ofNAA from the medium in the light after the preculture with NAA in
the dark.
F or the production of artificial seeds, carbohydrate should be contained only
in the beads. Various concentrations of sucrose in the beads which were placed on
a plastic sheet, were tested without sucrose supplementation in agar. A higher
shoot formation frequency of the root was observed when the sucrose con-
centration in the beads was above 3%. Once a leaf emerges from the root, the
energy of differentiation and proliferation could be supplied by photosynthesis.
Thus, supplementation with carbohydrate in the beads is necessary to grow
plantlets from a root fragment.
3.2 Encapsulation of Adventitious Shoot Primordia
3.2.1 Morphology of Adventitious Shoot Primordia
When horseradish hairy roots were cultured in liquid culture in the dark,
neoplastic tissue appeared on the roots after a few weeks. One day after being
transferred to light, the neoplastic tissue turned green, and shoots emerged a few
days later. From this observation, the neoplastic tissue was considered to be
adventitious shoot primordia. At the beginning of culture, adventitious shoot
primordia developed only in the center portions of whole roots in the flask,
which, being further away from the root apical meristem, are composed of older
cells. Over the course of the culture, however, primordia also emerged close to the
Fig. 1 A,B. Scanning electron micrographs of adventitious shoot primordia from horseradish hairy
root cultured in the dark. A and B were derived from the same culture
root apical meristem. Although some of the primordia developed into etiolated
plantlets in the dark, most of them remained 0.5- 3 mm in size during the
subculturing period (3 weeks) in the dark. Taking the hairy root morphology into
account, excision of the root with adventitious shoot primordia facilitated the
handling of the primordia for encapsulation. After adventitious shoot primordia
formed in the dark were excised, encapsulated, and placed on agar medium in the
light, plantlets grew out of the beads after a few weeks.
To clarify the morphology of the adventitious shoot primordia, scanning
electron microscope observations were carried out. Figure I shows small
primordium clusters on the basal portions of the lateral roots. An adventitious
shoot primordium consisted of primordial leaves.
3.2.2 Relationship Between Adventitious Shoot Primordium Formation

and Culture Time Before Encapsulation
We tested whether the culture time in the dark was related to primordium
formation. The number of primordia increased and individual primordia grew
larger as root culture proceeded. The number increased with the culture time up
to approximately day 20, and then reached a plateau.
Table 2 shows the plantlet formation frequency of the excised adventitious
shoot primordia. Multiple shoot formation resulted when more than one
adventitious shoot primordium formed on the excised root fragments. The
Table 2. Effect of time in culture in the dark on the frequency of plantlet

formation from adventitious shoot primordia
Culture time Plantlet formation MUltiple shoot formation

(days) frequency (%) frequency (%)
8 24 0
18 34 36
26 53 47
31 43 57
40 40 60
The hairy root fragments were cultured in the dark at 25°C for the indicated
periods, after which adventitious shoot primordia were encapsulated and
cultured in the light for 40 days. The experimental data were obtained from
25-36 adventitious shoot primordia.
maximum plantlet formation frequency with a low multiple shoot formation

frequency was obtained for the primordia in the 26-day culture. At 40 days, 60%
of the adventitious shoot primordia formed multiple shoots, which were found
difficult to grow into healthy plants. These results suggest that in the shorter
cultures primordia were generally too premature to enable plantlets to develop,
whereas the 40-day culture was too long and the plantlet formation frequency
decreased. The primordia from the 26-day culture were most suitable to produce
artificial seeds, both in terms of the number and the plant formation frequency
and, hence, were used in subsequent experiments.
3.2.3 Effect of Auxin or Cytokinin Supplementation

on Adventitious Shoot Primordia in the Dark
To test the effect of auxin and cytokinin on primordium formation from

horseradish hairy root, roots were cultured in MS medium containing various
concentrations of naphthalene acetic acid (NAA) (0-5 mg/l) or benzyl adenine
(BA) (0-5 mg/l) in the dark. NAA supplementation stimulated the formation of
callus which had a low plantlet development capability. Neither auxin nor
cytokinin in the dark improved the number of adventitious shoot primordia or
their regeneration frequency, although with cytokinin adventitious shoot
primordia were enlarged.
3.2.4 Root Inductionfrom Adventitious Shoot Primordia After Encapsulation
Root emergence and elongation from encapsulated artificial seeds are important
for nutrient uptake. To stimulate root elongation after encapsulation, the auxin
supplementation conditions were examined. At first, both beads containing
adventitious shoot primordia and agar were supplemented with various
concentrations ofNAA (0.1-5 mg/l); the development potential of adventitious
shoot primordia was completely inhibited.
To gradually lower the NAA concentration in the beads after encapsulation,

beads containing various concentrations ofNAA were placed on agar medium
without NAA. The beads in which roots emerged out of the beads were counted.
The optimal NAA concentration was 1 mg/l for root emergence. NAA
supplementation in the beads did not adversely effect the plantlet formation
frequency or root elongation. This result was apparently due to the diffusion of
NAA from the beads to the agar, suggesting that treatment with NAA effectively
triggered root formation.
When embryogenesis occurs in somatic adventitious shoots, root growth is
sometimes carried out in vitro under a high auxin concentration (Pierik 1987). In
contrast, hairy root adventitious shoot primordia overcome the difficulties of
root emergence and elongation in a natural manner. Thus, this is considered an
important advantage of root regeneration from hairy roots compared with the
use of adventitious shoots derived from somatic embryos (Uozumi et al. 1994).
3.3 The Encapsulation of Plantlets Regenerated from Hairy Root
To obtain a higher plant development frequency, the plantlet regenerated from

hairy root is more suitable for encapSUlation than root fragments and
adventitious shoot primordia. However, it is difficult for light, which is necessary
for plantlet formation, to reach the inner space of root materials and the
bioreactor when the density of hairy roots in liquid cultured becomes too high.
To overcome this problem, after hairy roots were cultured in MS medium
containing 2% sucrose, they were excised by razor and the resultant root
fragments were cultured in MS medium containing 2% sucrose in light. At day 7,
many adventitious shoots emerged from the excised root. At day 15, the plant
cells were harvested. Healthy plantlets were produced efficiently when the excised
hairy root was less than 10 mm in length.
To form shoots on the root fragment, auxin should be removed from the
medium since it often inhibits regeneration. On the other hand, auxin could
promote the cell growth rate of horseradish hairy root by extensive lateral root
emergence. In these experiments, hairy root was cultured in MS medium supple-
Table 3. Effect of kinetin supplementation on the number and size of plantlets
Kinetin Number of plantlets(l-I) Biomass yield Percentage

concentration Size (mm) [gr.wtll] of plantlets'
(mg/l) 0.4-1 1-2 2--4 4-10 10< Root Plantlet (%)
0 2580 1450 100 0 0 17.0 8.2 32.5

om 2320 2400 180 0 0 14.7 9.7 39.8
0.1 1350 1650 1650 930 0 8.5 39.5 82.3
I 280 970 580 670 850 1.0 125.5 92.2
Weight ofplantlets
------- x 100.
Total biomass weight
mented with 2% sucrose and Img/l NAA for 15 days. Hairy root culture was
excised by razor, transferred to fresh MS medium containing 2% sucrose,
and then cultured in the light for 15 days. Although shoots appeared at the
end of this culture, many roots producing the shoots grew to mUltiple plantlets
which were not useful for artificial seed because of their inability to grow to
mature plants. A root fragment with a single shoot is required for an artificial
seed.
To produce single plantlets through mass production, mechanical excision
was carried out by a commercial blender. In addition, the effect of kinetin of
single shoot formation was evaluated. Table 3 shows that for single shoot
formation the optimum kinetin concentration was 0.1 mg/!. Approximately 30%
of the obtained single shoots were of a suitable size (2--4 mm) for encapsulation
(Fig.2A).
After harvesting, the plantlets were dehydrated slightly and encapsulated
with gel (Fig. 2B). Most of the beads produced healthy plantlets.
The mass production of plantlets from hairy root was performed in several
types of bioreactors as shown in Fig. 3. The excised hairy root fragments were
cultured in the bioreactors in the light. A sufficient number of plantlets were
obtained in all types ofbioreactors.
Fig. 2 A,B. Plantlet formation and encapsulation. A Plant let from horseradish hairy root excised by
blender; BPlantlet encapsulated with calcium alginate gel
Fig. 3 A-D. Plantlet formation from hairy root in various types ofbioreactors. A, B, C Teardrop-type
airlift bioreactor; D spherical airlift bioreactor
4 Discussion
There are two main strategies for obtaining biochemical products in hairy root.
1. High density culture of hairy root by using bioreactors in combination with an
effective control technique.
2. The micropropagation of hairy root.
The use of cultured hairy roots focuses mainly on the large-scale production
of useful products or secondary metabolites such as pigments and alkaloids (See
Bajaj 1993). The optimization of fermenter-scale plant cell culture requires
knowledge of substrate requirements and utilization rates (Kondo et al. 1989).
To obtain a high density culture of plant cells, the culture conditions should be
maintained at an optimum level. To optimize culture conditions, the cell mass
must be monitored correctly. A decrease in conductivity in the medium has a
linear relation with an increase in cell mass (Taya et al. 1989a); on-line estimation
of cell mass is possible throughout culture. If a constant growth yield and
maintenance coefficient are assumed, consumption rates for nutrient
components such as carbon source are evaluated, nutrient components are
maintained at their optimum levels, and a high density culture of plant cells is
f
Dark ~" N\ Light

-~. . &,
Excision
,~ Root fragment. . ~
P:-- ......... jJID-- Exc,soon@/,
Excision
:~ ~ tpi~:~Uosdia / 1
Plantlet Plant
Fig. 4. Three routes for artificial seed production from hairy root
possible (Uozumi et al. 1991, 1993). The chemicals produced by plants are often
localized in specific tissues, and some of them are stored in the roots. The
excretion and recovery of the biochemicals are required in order to attain an
efficient culture (Kilby and Hunter 1990; Kato et al. 1991 ; Shim omura et al.
1991; Taya et al. 1992; Uozumi et al. 1992a).
Artificial seeds are expected to be a reliable delivery system for the clonal
propagation of elite plants. The delivery system has the potential for genetic
uniformity, high yields and low production costs. Generally, genetic improve-
ment of plants through conventional breeding and selection methods takes a long
time. Within an acceptable time period, new gene transfer technologies offer the
opportunity to easily produce plants having desirable traits such as disease or
herbicide resistance. Successful reports on elite transgenic plant cells and their
advantageous properties stimulated interest in developing a regeneration and
delivery system for hairy roots. A proper system for plant regeneration is
necessary to produce transgenic plants from hairy roots efficiently. In particular,
the production process should be constructed and improved so that these plants
can produce artificial seeds at an industrial scale.
Hairy roots must be cut to produce the artificial seeds. In this study, three
tissues from hairy roots were distinguished: root fragments, adventitious shoots,
and plantlets. Figure 4 presents schematically regeneration from three kinds of
explants. Root fragments are the most suitable for encapsulated hairy roots.
Adventitious shoot primordia, however, have a high potential to regenerate and
to green more uniformly under the light conditions than the two other tissue cells.
It is necessary to increase the number of the adventitious shoot primordia formed
in an artificial seed system. The mass production of adventitious shoots and the
separation from hairy roots are difficult. Although the production of plantlets
from hairy roots is required for long-term cultures, the plantlets have the highest
regeneration frequency. For the production of artificial seed from hairy roots, an
efficient culture in combination with bioreactors should be developed.
180 N. Uozumi and T. Kobayashi: Artificial Seed Production Through Encapsulation
We investigated the use of hairy roots in the production of artificial seed and
evaluated the efficiency of regeneration from horseradish hairy root in
combination with excision and encapsulation. The fundamental method and
time schedule could be developed on three kinds of explants. Hairy root has
superior properties because of plant transformation due to infection of a part of
the Ri plasmid. An efficient regeneration system, e.g., for artificial seed, provides
information on the micropropagation of elite plant organs.
References
Bajaj YPS (ed) (1993) Biotechnology in agriculture and forestry, vol. 21. Medicinal and aromatic
plants IV. Springer, Berlin Heidelberg New York
Kato Y, Uozumi N, Kimura T, Honda H, Kobayashi T (1991) Enhancement of peroxidase production
and excretion from horseradish hairy roots by light, NaCI, and peroxidase-adsorption in situ. Plant
Tissue Cult Lett 8: 158-165
Kilby NJ, Hunter CS (1990) Repeated harvest of vacuole-located secondary product from in vitro
grown plant cells using 1.02 MHz ultrasound. Appl Microbiol Biotechnol33: 448-451
Kitto SL, Janick J (1985) Production of synthetic seeds by encapsulating asexual embryos of carrot. J
Am Soc Hortic Sci 110: 227-282
Kondo 0, Honda H, Taya M, Kobayashi T (1989) Comparison of growth properties of carrot hairy
root in various bioreactors. Appl Microbiol Biotechnol33: 291-294
Murashige T, Skoog F (1962) A revised medium for rapid growth and bioassays with tobacco tissue
cultures. Physiol Plant 15: 473-497
Noda T, Tanaka N, Mano Y, Nabeshima S, Ohkawa H, Matsui C(1987) Regeneration of horseradish
hairy roots incited by Agrobacterium rhizogenes infection. Plant Cell Rep 6: 283-286
Redenbaugh K, Fujii J, Slade D, Viss P, Kossler M (1991) Artificial seeds - Encapsulated somatic
Shimomura K, Sudo H, Saga H, Kamada H (1991) Shikonin production and secretion by hairy root
cultures of Lithospermum erythrorhizon. Plant Cell Rep 10: 282-285
Tanaka N, Hayakawa M, Mano Y, Ohkawa H, Matsui C (1985) Infection of turnip and radish storage
roots with Agrobacterium rhizogenes. Plant Cell Rep 4: 74-77
Taya M, Yoyama A, Kondo 0, Kobayashi T (1989a) Growth characteristics of plant hairy roots and
their cultures in bioreactors. J Chern Eng Jpn 22: 84-89
Taya M, Yoyama A, Nomura R, Kondo 0, Matsui C, Kobayashi T (1989b) Production of peroxidase
with horseradish hairy root cells in a two step culture system. J Ferment Bioeng 67: 31-34
Taya M, Mine K, Kino-oka M, Tone S, Ichi T (1992) Production and release of pigments by culture of
transformed hairy root of red beet. J Ferment Bioeng 73: 31-36
Tepfer D (1990) Genetic transformation using Agrobacterium rhizogenes. Physiol Plant 79: 140-146
Uozurni N, Kohketsu K, Kobayashi T (1993) Plant hairy root growth and metabolism in fed-batch
culture on monosaccharide medium. J Chern Techno! BiotechnoI57:155-161
Uozumi N, Asano Y, Kobayashi T (1994) Micropropagation of horseradish hairy root by means of
adventitious shoot primordia. Plant Cell Tissue Organ Cult 36: 183-190
Uozumi N, Kohketsu K, Kondo 0, Honda H, Kobayashi T (1991) Fed-batch culture of hairy root
using fructose as a carbon source. J Ferment Bioeng 72: 457-460
Uozurni N, Yoshihiro K, Nakashimada Y, Kobayashi T (1992a) Excretion of peroxidase from
horseradish hairy root in combination with ion supplementation. Appl Microbiol Biotechnol 37:
560-565
Uozumi N, Nakashimada Y, Kato Y, Kobayashi T (1992b) Production of artificial seed from
horseradish hairy root. J Ferment Bioeng 74: 21-26
11.5 Fluid Drilling as a Delivery System
for Somatic Embryo-Derived Plantlets*
S.L. KITIO\ W.G. PILL2, and D.M. MOLLOy2
1 Introduction
Four systems have been proposed to facilitate the bulk handling and delivery of
in vitro-derived somatic embryos: fluid drilling (Drew 1979), encapsulation of
somatic embryos in alginate gel (Redenbaugh et al. 1986), desiccation of somatic
embryos (Gray et al. 1987), or desiccation of somatic embryos encapsulated in
water-soluble resin (Kitto and Janick 1985).
Although fluid drilling of somatic embryos, the extrusion of a gel-embryo
mixture into a seedbed, was conceptualized as early as 1979 (Drew), there has
been little work in this area to date. Early research examining fluid drilling as a
delivery system for somatic embryos demonstrated both its feasibility and
drawbacks (Baker 1985). While somatic embryos of carrot (Daucus carota L.)
survived up to 7 days post-fluid drilling, they failed to elongate and grow
normally (Baker 1985). Fluid drilling of somatic embryos is a complex system
since gel composition and somatic embryo development and their associated
characteristics and interactions must be considered. Also, the post-fluid drilling
environment is critical to somatic embryo survival, growth, and development.
Gels must provide a favorable environment for somatic embryo growth while
being capable of tolerating modifications. Somatic embryos must be fully
developed or developmentally mature and capable of continued normal post-
fluid drilling growth (Gray and Purohit 1991). Environmental and growth
conditions outside the laboratory (i.e., greenhouse, field) must not impede
continued growth of the somatic embryos.
* Research supported in part by a Delaware Research Partnership Grant which was funded jointly by
the State of Delaware and the Aqualon Group, Wilmington, DE 19899, USA
'Delaware Agricultural Experiment Station, Department of Plant and Soil Sciences, College of
Agricultural Sciences, University of Delaware, Newark, DE 19717-1303, USA
2Former Research Associate. Currently Research Assistant, Alfred A.I. DuPont Institute, 100
Rockland Rd., Wilmington, DE 19899, USA

182 S.1. Kitto et al.
2 Somatic Embryo Development
A mature true seed, essentially a young plant in an arrested state of development,

is composed of an embryo (embryonic axis plus nutritive tissues) and covering
structures (seed coats). Seed coats playa role in both the regulation of nutrition
and protection during embryo development. Somatic embryos produced in vitro
not only lack covering structures, but they are produced under relatively low-
stress laboratory conditions. Therefore, survivability of the unprotected, non-
hardened somatic embryos outside the tissue culture environment is low. Systems
for handling somatic embryos with minimal damage must be developed if the
embryos are to survive delivery to seedbeds in the greenhouse or field.
Suitable treatments imposed on somatic embryos during development that
mimic natural zygotic embryo-developmental conditions might allow for the
development of mature, normal, autotrophic somatic embryos. Somatic
embryos not fully developed have reduced vigor and develop more slowly
(McKersie et al. 1989).
Carbohydrates as osmotic agents have been linked to acquisition of
desiccation tolerance (maturity) by somatic embryos of carrot (Kitto and Janick
1985) and of celery (Apium graveolens L.) (J. Janick, pers. comm.). Exposure to
water stress induced by polyethylene glycol (PEG) also has been associated with
increased stress tolerance in carrot cells (Fallon and Phillips 1989) and in somatic
embryos of celery (J. Janick, pers. comm.).
3 Fluid Drilling
Fluid drilling is a crop establishment technique that includes: seed germination

(primary root just protruding through the seed coat) under ideal conditions,
incorporation of the germinated seeds into a gel, and planting of the gel-seed
mixture in the seedbed (Pill 1991). Fluid drilling can give (1) earlier, greater, and
more uniform seedling emergence; (2) earlier and greater yields; and (3) in some
crops, more uniform maturity than conventional methods of sowing dry seeds
(Gray 1984).
The fluid-drilling technique provides the potential for bulk handling of many
small plantlets without the need for individual handling. In addition, the
protective carrier gel can contain additives such as fertilizer salts, plant growth
regulators, pesticides, and microorganisms, thereby creating a "packaged"
environment for the seed and seedling (Salter 1978).
The effect of an amended hydroxyethyl cellulose gel on somatic embryos of
sweet potato [Ipomoea halalas (L.) Lam] to allow for maximal somatic embryo
maturation and subsequent development into normal plants has been examined,
and it was noted that both nutrient formulation and carbohydrate source of gel
amendments influenced normal development of somatic embryos (Schultheis
and Cantliffe 1992).
Fluid Drilling as a Delivery System for Somatic Embryo-Derived PlantIets 183
4 Post-Fluid Drilling Environment
Proper seedbed conditions are critical to the successful establishment of fluid-

drilled true seeds (Pill. 1991) and, therefore, should be no less important, and
may be more important, for fluid-drilled somatic embryos. Watering regime has
been associated with successful somatic embryo conversion and growth under
greenhouse conditions (Fujii et al. 1989). Field planting of somatic embryos of
alfalfa (Medicago sativa L.) has demonstrated further the importance of
maintaining proper moisture levels around somatic embryos (Fujii et al. 1992).
Field-sown somatic embryos provided with a protective covering (i.e., styrofoam
beverage cup) survived and grew better than somatic embryos sown directly in
the seedbed (Fujii et al. 1992).
5 Fluid Drilling as Delivery System

for Somatic Embryo-Derived Plantlets of Carrot
The results of a series of studies that examined techniques for fluid sowing of
somatic embryo-derived plantlets (SEPs; torpedo embryos that have pre-
cociously "germinated", Fig. 1) of carrot (Daucus carota L. 'Orlando Gold') are
reported here.
Fig. 1. SEPs derived from somatic embryos incubated under light for 6 days with 1 x MS salts and 2%
sucrose (bar = 1 cm). (Kitto et al. 1991)
184 S.L. Kitto et al.
Table 1. 'Orlando Gold' carrot SEP vigor 5 days after incorporation in N-gel containing 1x MS
medium, sucrose, and two fungicides; and SEP conversion in the greenhouse 20 days after
transplanting. (Kitto et al.. 1991)
Sucrose Fungicide SEPvigor SEP conversion

(%, w/v) (mga.i.ll)" (mean± SD? (%y
0 None 0 4.0 ± 1.02 0

0 Truban 250 4.0±1.17 0
0 Truban 500 4.0 ± 1.05 0
0 Banrot 250 2.9 ± 0.64 0
0 Banrot 500 4.1 ± 0.78 0
None 0 2.7±1.03 0
Truban 250 3.2 ± 1.14 0
Truban 500 3.4 ± 0.91 0
Banrot 250 2.7 ± 0.75 0
Banrot 500 3.6 ± 1.22 0
2 None 0 2.4 ± 0.64 7.0
2 Truban 250 2.8 ± 0.55 12.5
2 Truban 500 3.1 ± 0.52 4.8
2 Banrot 250 2.6 ± 0.64 7.1
2 Banrot 500 3.0 ± 0.50 4.8
, a.i. = active ingredient.
b SEP = somatic embryo-derived plantlet; n = 25; scale from 1 = dead to 5 = green and healthy.
C SEP conversion = plants with primary leaves.
5.1 Composition and Volume ofthe Fluid-DriUing Gel
Incubation of SEPs in gel variously amended with sucrose, chitosan, PEG, or

fungicides was examined as a prefluid drilling SEP-maturation step. Sucrose,
chitosan glutamate, and Truban fungicide appeared to be beneficial during SEP
incubation in the fluid-drilling gel. SEP conversion was greatest after incubation
in gel containing 2% sucrose (Table 1). Sucrose functions as a nutritional carbon
source and as an osmotic agent in vitro. Lack of SEP conversion with 0 or 1%
sucrose (Table 1) may be due to an insufficient supply of carbon and energy
during the period when SEPs become autotrophic. Gel supplemented with
chitosan glutamate (0.01 gil) during the incubation period before fluid drilling led
to a high SEP conversion into plants 4 weeks after fluid drilling, a conversion rate
that was sustained until 7 weeks (Table 2).
The total water potential of the SEPs, estimated by vapor pressure
osmometry of crushed SEPs, was -0.39 MPa (20°C). Water potentials of 1.67%
(w/v) N -gel (hydroxyethyl cellulose, Aqualon, Wilmington, Delaware) prepared
with 1 x MS salts (Murashige and Skoog 1962) and 0, 1, and 2% (w/v) sucrose
were -0.25, -0.31, and -0.39 MPa, respectively. Thus, the SEPs should have
received a net influx of water until water potential equilibrium was reached.
Decreasing the water potential of the gel from -0.39 MPa (0% PEG) to -1.99
MPa (25% PEG) during a I-week incubation period before fluid drilling did not
increase SEP conversion (Table 2).
Fluid Drilling as a Delivery System for Somatic Embryo-Derived Plantlets 185
Table 2. Percentage conversion of somatic embryos conditioned for I week under light with PEG or
chitosan in different gel volumes then fluid drilled septically onto RediEarth and placed under light in
the laboratory. (Kitto et al. 1991)
Gel Gel SEP length after SEP conversion"

amendement volume b incubation in 4 weeks 7 weeks
gel' (mm)
0 10 21.4 0 I
0 15 23 . 1 3 10
0 20 23.6 I 5
0 25 23.3 I 3
0 30 20.7 I I
15% PEG 30 0 0
20% PEG 30 14.8 0 0
25% PEG 30 14.1 0 0
0.01 g Chitosan d 30 18.7 5 6
0.05 g Chitosan 30 13.3 0 0
0.09 g Chitosan 30 14.0 I 0
LSD 0.05 2.69
a SEP conversion = number of plants with primary leaves at 4 and 7 weeks after fluid drilling.
b Volume of gel per 10 ml of SEP.
, Initial SEP length 9.8 ± 2.3 mm.
d Water-soluble chitosan glutamate (Seacure).
Fig. 2. Four-week-old (A) and full-grown

(B) 'Orlando Gold' carrot plants
Table 3. SEP length at the time of fluid drilling and SEP cotyledon emergence I week after fluid drilling
as influenced by duration of culture exposure to continuous light at 25°C, and subsequent duration of
SEP incubation in 1.67% N-gel prepared with 1 x MS salts and 2% sucrose. (Kitto et al.199 I)
Duration Incubation SEP length after SEP emergence 1

oflight in gel incubation in gel week after fluid
(days) (weeks) (mm) drilling" (%)
4 0 6.4 4.1
1 10.9 31.0
2 12.2 52.5
3 26.2 18.3
4 14.4 20.0
6 0 8.2 9.7
1 11.8 41.8
2 13.2 45.0
3 15.1 16.8
4 11.9 7.5
8 0 9.6 8.2
I 15.2 22.5
2 22.2 15.2
3 12.1 12.5
4 10.8 15.0
LSD 0.05 2.14 24.43
a SEP emergence =emergence of cotyledons above the growth medium surface.
SEPs converted into plants when incubated for 8 days in gel containing 2%
sucrose and 250 mg active ingredient (a.i.) Truban fungicide per liter (Tablel).
Plants appeared normal 12 weeks after transplanting into growth medium under
greenhouse conditions (Fig. 2). SEPs incubated in gel containing Benlate
fungicide did not survive after transplanting (data not presented). Increasing the
gel volume above 15 ml per 10 ml ofSEP suspension reduced the SEP conversion
rate (Table 2).
5.2 Roles of Light and Chilling During Embryo Development
In an attempt to develop a system for maturing somatic embryos into auto-

trophic SEPs, somatic embryos were treated either with light (PAR 60 J.lmoV =
m2/s!, cool white fluorescent lamps) and/or chilling (4°C) during development in
suspension culture or during incubation in gel.
SEPs provided 4 to 6 days oflight and incubated for up to 2 weeks in gel had
the greatest vigor (Table 3). Regardless of chilling duration (3 or 6 days),
increasing the subsequent incubation period at 25°C in the light from 0 to 6 days
increased SEP growth before fluid drilling (Fig. 3A). Light or dark during chilling
for up to 6 days initially had no effect on SEP length. However, SEP length 2
weeks after fluid drilling increased dramatically when suspension cultures were
chilled for 3 days in light before SEP incubation in gel for 3 days in light (Fig. 3B).
60
A 8
50 -
I ./
/
//
E ~./
,5 40 --- ---
.c
+-'
~--
OJ
c 30
..!!1
0.. /
.- .-
ill
CI) // _---0
(ij
20 - 6- - - - - -::--::-~f'----------'
0
t-
10 o-------_~-=- --e
• -----0
Olo-~ _ _ _.....Ii-_ _ ____

o 3 6 o 3 6
Days of chilling in light or dark
Fig.3. SEP length at the time offluid drilling (A) or 2 weeks after fluid drilling (8). SEPs were subjected
to 0,3, or 6 days of chilling (4°C)in light (open symbols) or dark (closed symbols) followed by 0 (O,e),
3 (Y',.A.) or 6 (0, .) days oflight at 25°C before fluid drilling. Vertical bars == LSD 005' n == 25. (Kitto
et al. 1991)
Initially, light appeared to have minimal influence on subsequent SEP

growth. However, greater SEP length 2 weeks after fluid drilling resulted from
exposure to light, rather than dark, during chilling. This growth differential due
to irradiation during chilling may reflect a delayed, albeit promotive, effect of
light on SEP maturation. Although SEP length was greatest when 6 days of
chilling was followed by 6 days of incubation in light at 25°C, these SEPs were
too large to fluid drill (Fig. 3). The greatest increase in SEP length 2 weeks after
fluid drilling occurred when SEPs were treated for 3 days at 4 °C in light followed
by 3 days at 25 °Cin light. Although the 18 mm-Iong SEPs of this treatment could
be fluid drilled, from a practical standpoint, SEPs should be no longer than 1 cm
to be fluid drilled. A decrease in SEP length might be accomplished by com-
mencing the entire chilling! post-chilling regime earlier in embryo development.
SEP length increased with up to 3 weeks of incubation in gel if treatment at
25°C in the light prior to gel incubation was::;; 6 days (Table 3). SEPs were longest
when treated for 4 days at 25 0 C in the light during liquid suspension culture,
followed by 3 weeks of incubation in gel. SEPs from liquid suspension cultures
treated at 25°C in the light for> 6 days required shorter incubation in gel for
maximal growth. SEPs incubated in gel, compared to those not incubated in gel,
were longer and had a greater emergence percentage. Cotyledonary emergence 1
week after fluid drilling was maximal when SEPs were incubated in gel for I to 2
Table 4. Percentage cotyledon emergence of SEPs 5 weeks after fluid drilling as influenced by growth
medium and irrigation. Before fluid drilling, the SEPs were given either 6 days oflight at 25°C (non-
chilled SEPs) or 3 days at 4 °C then 3 days light at 25°C (chilled SEPs). (Kitto et al. 1991)
Growth medium" Pr PrY PrP Re ReV ReP
Percentage SEP cotyledon emergence

Mist
Nonchilled SEPs 10.1 1.5 5.7 57.3 0 8.5
Chilled SEPs 5.0 0 16.7 4.8 0 7.8
No Mist b
Nonchilled SEPs 18.8 8.0 21.4 25 .3 16.9 10.9
Chilled SEPs 4.5 14.6 24.2 25.1 5.0 0
LSD 0.05 = 21.34 (3-way interaction)
a Growth medium: Pr = Pro-Mix BX; PrY = Pr plus vermiculite (V, v/v); PrP = Pr plus perlite (P, v/v);
Re = RediEarth; ReV = Re plus V (v/v); ReP = Re plus P (v/v).
b Surface-irrigated once daily.
weeks, and was greatly reduced when suspension cultures were exposed to 8 days
oflight.
5.3 Post-Fluid-Drilling Environment in the Greenhouse
Although SEPs were capable of conversion under greenhouse conditions

(natural irradiance) in early February (Table 1), no conversion occurred from
March to August (Tables 3 and 4; Fig. 3). However, SEPs were still capable of
conversion into plants having primary leaves after fluid drilling as occurred
under septic laboratory conditions (Table 2; Fig. 4). We concluded that the
factor(s) in the greenhouse environment preventing SEP conversion, while not
preventing SEP emergence, may be associated with higher average temperature
and reduced water availability between irrigations.
Fig. 4. Expanding primary leaf from plants

derived from somatic embryos in vitro
Chilling, growth medium, and irrigation all influenced the percentage of

cotyledon emergence of SEPs 5 weeks after fluid drilling under greenhouse
conditions (Table 4). A greater percentage of chilled SEPs emerged compared to
nonchilled SEPs (15.4% versus 9.1%). Growth medium interacted with mist
versus no mist. Compared to other growth media Pro-Mix BX + perlite (v/v) or
RediEarth generally resulted in a greater percentage emergence from fluid-drilled
SEPs (Table 4). When provided daily surface irrigation, chilled or nonchilled
SEPs gave 21 to 25% emergence in these media. Nonchilled SEPs that were fluid
drilled into the peat-lite RediEarth and provided intermittent mist gave the
greatest (57.3) percentage emergence.
Results of these experiments showed that SEPs were capable of conversion
into plants under either septic laboratory or greenhouse conditions following
fluid drilling.
The objective of this study was to develop fluid-drilling protocols for tissue
culture-derived somatic embryos of 'Orlando Gold' carrot. Light or chilling
during liquid suspension culture followed by a lighted incubation period in fluid-
drilling gel of low water potential that contained sucrose and fungicide or
chitosan favored SEP conversion. It is clear from this study that somatic embryos
of carrot can be delivered successfully to the seedbed using fluid drilling; however,
more work will be required to define further the variables involved with (1)
maximizing the development of normal somatic embryos; (2) maximizing
interactions between somatic embryos and gel (additives); and (3) developing
seedbed conditions conducive to continued somatic embryo growth.
7 Protocol
The overall protocol from explant extraction to fluid drilling of SEPs is shown in Fig. 5. The major
methodologies that have been reported in more detail in Kitto et al. (1991) are summarized below.
7.1 Production and Conditioning ofSEPs
Callus and cell suspensions of Daucus carota L. 'Orlando Gold' were initiated and maintained as
described by Kitto and Janick (1985). To initiate embryos from cell suspensions, the suspensions were
transferred to fresh medium every 3 days. After 10 to 14 days, torpedo embryos precociously
"germinated" to produce SEPs. SEPs (still in liquid suspension culture) were conditioned either under
cool-white fluorescent lamps (60 ~mol/m2/s of PAR at suspension culture level; 16-h photoperiod) at
25120 ·C (light/dark) and/or by chilling at 4 ·C in the light or dark for 3 to 8 days.
Gel was hydrated by adding powder (1.67%, w/v) to MS solution [l x MS salts; 0, I, or 2% (w/v)
sucrose; 0, 250, or 500 mg/I active ingredient of wettable powder formulations of the fungicides
SEED
\0
CARROT S o
T~A~~~~n HVPO'rOTViEDLING
" _. ~A~;A~~TE'NN'T'ATION
3-D AV TRANSFER ,,',';'; ,_ .l'-' -

(For 2 Week, ) ~ fi~~ ~ .\~ ~,,-.-i ~...,.:
.. ..,·;:.... ..···r ~
~
.>.:?~\,. /
CAnROTS Gn
EMBRVO COLLECTION ~ CelL flUID DRlllE~WN FRO M
EMBRYOS
¥ SUSPENSION
.. j;
-' ~
r-~
" . (0." "-"

~ ~~,
It., .. II ., EMsnVO-GEl
t M"'l''' .. ''' INCUBATION
• T t.t"" PRE T RE ATM ENT
r."~ 1. .,"
~
EXTnUSION OF THE
~ GEL· EMBRYO
MIXTURE ItlTO A
:~ GREENHOUS E flAT
l:~I:~\
,---- , EMBRVO
.: ~ " TRAN SFER TO
FlU IO·DR ILLI NG Gel
Fig. 5. Schematic representation of protocol begins with the initiation of callus from a carrot seedling hypocotyl. During
subsequent suspension subcultures, somatic embryos were induced and subjected to various conditioning treatments. The
VJ
conditioned embryos or SEPs were then incorporated into fluid-drilling gel. The SEP-gel mixture was incubated before
being fluid drilled into the seedbed. (Kitto et al. 1991)
r
~
0-
~
~
Fluid Drilling as a Delivery System for Somatic Embryo-Derived PlantIets 191
Truban, Benlate, or Banrot; 0, 150, 200, or 250 gil of PEG 8000; or 0, 10, 50, or 90 mg/I of chitosan
glutamate (Seacure, Protan Co., Redmond, Wash.)] that was continuously agitated with a magnetic
stirrer. The gels, after autocIaving at 125 kPa and 121 DC for 15 min, were of sufficient viscosity to
suspend the SEPs for at least I h.
Water potentials for various solutions and N-gel preparations were determined by vapor pressure
osmometry (Wescor Model 5500XR, Logan, Utah). A 10-111 sample was added to the chamber for
aqueous solutions. For the viscous N-gel preparations, a filter paper disk was dipped into the fluid and
the excess material removed before placement of the disk in the sample holder. Osmometer readings
(milliosmolality, mOsm) were converted to -MPa using the formula: Ijf, = i.m.R.T, where Ijf, =
osmotic potential in -MPa, i.m. = osmolality (mOsmlkg) of the sample, R= gas constant (8.31 x
10.6 m 3 MPalmollK) and T = temperature (K).
SEPs were incubated in gel contained within parafilm-sealed eight-well polystyrene tissue culture
multiplates or 125 x 80 x 20mmpolystyrene boxes. Amultiplatewell(26 x 33 x 10mrn) contained five
SEPs and I ml gel of the appropriate composition. Multiplates were maintained for 5 days at 25/20 DC
in the light after which SEP vigor was rated subjectively from 1 = dead to 5 = green and healthy. To
determine the effect of the ratio ofSEP suspension to incubation gel, 10, 15,20,25, or 30 ml ofN-gel
added to 10 ml of SEP suspension was poured into polystyrene boxes to provide a gel layer 1 to 2 mm
thick. Incubation of the SEP-geI mixture under 16-h Iight/8-h dark regime, at 25/21 DC, was for 0, I, 2,
3, or 4 weeks, after which representative SEP lengths were measured.
After incubation, SEPs plus gel (adjusted to a total volume of 15, 20, 30, or 50 ml) were mixed and
placed in 125-ml plastic bags (6 oz Whirlpak, NASCO, Ft. Atkinson, Wisconsin) ready for fluid
drilling by squeezing the gel-SEP mixture through a cut corner.
7.2 Post-Fluid-Drilling Environment in the Greenhouse
SEPs either were transplanted manually or fluid drilled into IO-cm plastic Petri dishes or 17 x 12 x 6 cm
plastic flats. Petri dishes were placed under 16-h light/8-h dark regime, at 25/ 20 DC, for 2 weeks when
SEP length was determined. Flats contained five 12-em-long (I-em-deep) furrows pressed into a
growth medium [pro-Mix BX (Pr), Premier Brands Inc., Stanford, Connecticut; RediEarth (Re), W.R.
Grace, Fogelsville, Pennsylvania; vermiculite (V); perlite (P) or 50% (v/v) combinations of Re or Pr
with either V or P]. Sown SEPs either were not covered or covered with ca. 5 mm of the appropriate
growth medium. Flats in the greenhouse either were placed under intermittent mist (6 severy 6 min) or
were surface-irrigated once daily. Flats removed from the mist were surface-irrigated once daily. Flats
in the laboratory were placed in clear plastic boxes under coolwhite fluorescent lamps (60 I1mol PAR
m2/s, 16-h day/8-h night, 25/20 DC).,
References
Baker CM (1985) Synchronization and fluid sowing of carrot, Daucus carota somatic embryos. MS
Thesis, University of Florida, Gainesville
Drew RLK (1979) The development of carrot (Daucus carota L.) embryoids (derived from cell
suspension culture) into plantIets on a sugar-free basal medium. Hortic Res 19: 79-84
Fallon KM, Phillips R (1989) Responses to water stress in adapted and unadapted carrot cell
suspension cultures. J Exp Bot 40: 681-687
Fujii JA, Slade D, Redenbaugh K (1989) Maturation and greenhouse planting of alfalfa artificial
seeds. In Vitro Cell Dev BioI 25: 1179-1182
Fujii JA, Slade D, Aguirre-Rascon J, Redenbaugh K (1992) Field planting of alfalfa artificial seeds. In
Vitro Cell Dev BioI 28P: 73-90
Gray D (1984) The role of fluid drilling in plant establishment. Aspects Appl BioI 7: 153-172
Gray DJ, Purohit A (1991) Somatic embryogenesis and development of synthetic seed technology. Crit
Rev Plant Sci 10: 33-61
192 S.L. Kitto et. a1.: Fluid Drilling as a Delivery System for Somatic Embryo-Derived Plantlets
Gray DJ, Conger BV, Songstad DD (1987) Desiccated quiescent somatic embryos of orchard grass for
use as synthetic seeds. In Vitro Cell Dev BioI 23: 29-33
Kitto SL, Janick J (1985) Hardening treatments increase survival of synthetically coated asexual
embryos of carrot. J Am Soc Hortic Sci 110: 283-286
Kitto SL, Pill WG, Molloy DM (1991) Fluid drilling as a delivery system for somatic embryo-derived
plantlets of carrot (Daucus carota L.). Sci Hortie 47: 209-220
McKersie BD, Senaratna T, Bowley SR, Brown DCW, Krochko JE, Bewley JD (1989) Application of
artificial seed technology in the production of hybrid alfalfa (Medicago sativa L.). In Vitro Cell Dev
BioI 25: 1183-1188
Pill WG (1991) Advances in fluid drilling. Hort Technology 1: 59-65
Redenbaugh K, Paasch PD, Nichol JW, Kossler ME, Viss PR, Walker KA (1986) Somatic seeds:
encapsulation of asexual plant embryos. Biorrechnology 4: 797-801
Salter PJ (1978) Fluid drilling of pregerminated seeds: progress and possibilities. Acta Hartic 33:
245-249
Schultheis JR, Cantliffe DJ (1992) Growth of somatic embryos of sweet potato (Ipomoea hatatas (L.)
Lam) in hydroxyethyl cellulose gel amended with salts and carbohydrates. Scient Hartic 50 : 21-33
11.6 Micropropagation Through Somatic Embryos
P.D. DENCHEV and A.I. ATANASSOV i
1 Introduction
1.1 Micropropagation - Commercial Outlook
Commercial micropropagation is a challenging plant biotechnology industry

that offers new methods of plant production. A micropropagation system
exploits the morphogenic potential of existing growing points or meristems
within the plant (Giles and Morgan 1987).
Microclonal propagation is, however, labor-intensive, involving several in
vitro steps in which plants must be gradually acclimatized from culture to the
greenhouse and to the field (Murashige 1989). Because of extensive costs,
commericialization has been limited to high value crops (Redenbaugh et al. 1987;
Tisserat 1991).
The most expensive part ofthe tissue culture production process, however,
remains the multiplication phase. It requires repeated cutting, separation and
transfer of cultures, preparation and dispensing of media in the culture vessels,
and enough space and maintenance of the growth room (Kozai and Iwanami
1988; Levin et al. 1988; Kozai 1990; Bajaj 1991). Cost reduction at this stage
could be achieved by increasing the rate and efficiency of multiplication (Levin
et al. 1988; Ziv 1991), and reducing manual labor associated with the repeated
cutting, separation, and transfer needed for multiplication (Levin et al. 1988;
Redenbaugh 1990; Gray and Purohit 1991).
Over the last few years attention has been directed toward the optimization
of the multiplication rate by manipulating the growth regulator supplements to
the media (Flegmann and Wainwright 1981), by strictly controlling environ-
mental factors (Kozai 1990), and by reducing manual labor by using robotics
systems. These systems can detect, pick up, and transfer plantlets, etc. (Miles and
Kutz 1991). Several companies, mainly in Japan, such as Kirin, Toshiba, etc.,
have focused their efforts on the development of automated tissue culture
systems. Although automation will increase profits for certain crops, it cannot
replace current procedures of micropropagation for the majority of plants.
1 Institute of Genetic Engineering, 2232 Kostinbrod 2, Bulgaria

©Springer.Veriag Berlin Heidelberg 1995
194 P.D. Denchev and A.I. Atanassov
1.2 Somatic Embryogenesis and Micropropagation
Another culture system, which is amenable to a higher degree of automation and

holds much promise for the mass propagation of plants at low cost, is based on
the production of somatic embryos. Large-scale somatic embryogenesis in vitro,
on agar or in cell suspension cultures, is the most attractive method for the mass
cloning of plants, since very large numbers of somatic embryos can be produced
in a short period of time in a limited volume of medium. Furthermore, methods
for mechanical encapsulation of the embryos and production of artificial seeds
are available.
Artificial seeds, consisting of somatic embryos enclosed in a protective
coating, have been proposed as a low-cost, high-volume propagation system
(Redenbaugh et al. 1986). The inherent advantages of artificial seeds are the
production of many somatic embryos and the use of conventional seed handling
techniques for embryo delivery. Over the past 10 years much progress has been
made with regard to various aspects of artificial seed technology. Plants have
been produced from synthetic seeds for more than 20 species (Redenbaugh et al.
1991).
As a model system alfalfa has been used extensively to study the factors
controlling the process. Somatic embryos of alfalfa have been used to develop
hydrated synthetic seeds by applying various hydro gels (Redenbaugh et al.
1991). At the University of Guelph, Canada another approach has been devel-
oped. Somatic embryos of alfalfa were dried to a moisture content of 8 to 15%
and after ABA maturation the conversion frequency for randomly picked
embryos reached 69%. However, all somatic embryos were considerably less
vigorous than true seeds and appeared to be less biochemically mature
(McKersie et al. 1989; Senaratna et al. 1990). Therefore, the economical way to
produce large amounts of high-quality, dry-alfalfa somatic embryos will be to
mass produce embryos and select only the high-quality somatic embryos based
on shape, size, and density (McKersie and Bowley 1993).
Plants propagated in suspension or liquid culture via somatic embryogenesis
provide a valuable alternative to the traditional propagation systems (Lutz et al.
1985; Redenbaugh et al. 1988; Denchev et al. 1990). Such a liquid-base system
could use bioreactors, which even on a small scale could produce thousands of
embryos, and many of the steps could be mechanized (Styer 1985; Akita and
Takayama 1988). This would help towards making such a system economically
feasible for many plant species.
A critical step in the development of such a system is high-frequency
somatic embryo production in liquid culture. Two types of in vitro somatic
embryogenesis are recognized:
1. Direct embryogenesis in which embryos are formed without callus formation.
2. Indirect embryogenesis in which callus formation proceeds before the em-
bryogenesis process.
The potential applications of somatic embryogenesis in plant improvement
depend to a large extent on whether embryos develop through callus or directly
from explant cells. The process of cell dedifferentiation causes a higher frequency
Micropropagation Through Somatic Embryos 195
of somaclonal variation while direct somatic embryo formation appears to give

rise to relatively uniform clone material. This embryo multiplication process
offers a number of potential applications:
1. Direct cloning of commercial F I hybrids for species where material can be sold
to growers in "seedling" transplant form. The ideal system might involve
continuous, direct embryogenesis from F I sexual embryos with periodic
harvesting of clonal, somatic embryos for further growing.
2. Rapid cloning of valuable breeding stocks at the earliest possible stage of the
life cycle after crossing.
3. In vitro selection and screening of genotypes for whole-plant characters at the
earliest possible stage of the life cycle.
4. Generation of "seedling" clones in outbreeding species where each seed
normally represents a different genotype.
2 A System for Direct Somatic Embryogenesis in Alfalfa
High-frequency somatic embryo production was obtained for several alfalfa

genotypes either on solidified media or in suspension culture (Kao and
Michayluk 1981; Walker and Sato 1981; Atanassov and Brown 1984; Lupotto
1986; Stuart and Redenbaugh 1987). However, for all embryogenic systems
developed so far an indirect origin of somatic embryo formation was reported,
excluding protoplast systems (Kao and Michayluk 1981; Dijak and Brown
1987). Our results showed that somatic embryos could be induced indirectly in
callus or cell suspension culture and directly in explant cells. The process is
affected by several factors, e.g.: initial genotype, type and physiological state of
the explant, cultural conditions etc., which are the focus of our studies.
Research and development of a complete embryogenic system require a large
amount of single, mature embryos capable of successfully undergoing the
conversion step. At the Institute of Genetic Engineering, Kostinbrod, Bulgaria,
we have developed a general procedure for plant regeneration via direct somatic
embryogenesis in several alfalfa species and varieties (Denchev et. al. 1990,
1991a,b,c). Two different protocols for direct somatic embryogenesis have been
elaborated -liquid and drop culture. They are based on the ability of the
embryogenic cells in the explant to form embryos after wounding and exogenous
application of auxins, mainly 2,4-dichlorophenoxyacetic acid (2,4-D):
Liquid culture. Young trifoliate leaves from M. Jalcata, M sativa (Denchev et al.
1991a,b), and M. trautwettery (Denchev et al. 1991c) were directly introduced
into liquid induction medium.
Drop culture. Leaf petioles were separated from the plants and cultured in a drop
of induction medium in a Petri dish, without shaking (Denchev et al.I990).
2.1 Induction of Somatic Embryos
It has been claimed that development of a system for alfalfa somatic embryo-
genesis is limited by a number of factors such as initial genotype and explant
(Kao and Michayluk 1981; Brown and Atanassov 1985), type and duration of
treatment with a morphogenesis-inducing factor (Stuart and Redenbaugh 1987),
and composition of the culture medium (Dos Santos et al. 1983; Seitz and
Bingham 1988). One of the most important steps in developing such a system
turned out to be the induction of somatic embryos. It was found that the
induction time period strongly affected both the number of embryos obtained
and their quality and ability for further development from a single somatic
embryo up to a vigorous plant (Stuart and Redenbaugh 1987).
Direct embryo formation in selected clones was observed within 2 weeks.
Consequently, in order to better assess the effect of the induction period on
somatic embryo formation, induction periods of 15, 20, 25, and 30 days were
examined. The results showed a significant difference between different induction
periods with respect to the total number of embryos detected for M. falcata 47/
1-5 (Denchev et al. 1991 b). The largest number of embryos was observed after 25
days of induction. At this time the highest number of torpedo stage embryos was
induced. Increase in the length of the induction period after the 15th day
appeared to cause the induction of embryos at the cotyledon stage, decreasing the
number of globular embryos.
The experiments performed with M . sativa No. 2/9R showed a lower total
embryogenic yield in comparison with M. falcata 47/1-5, but a higher synchrony
in embryo development (there were no significant differences between the
embryos produced after lO, 15, and 20 days)(Denchev et al. 1991b).
The highest number of embryos was detected after 10 days of induction.
After 30 days of 2,4-D treatment embryo development was strongly inhibited.
The small number of globular and polyembryos detected after 25 and 30 days of
induction could be explained by the initiation of secondary cell differentiation.
Fig. 1. Prominent suspensor

quite similar to that found in
zygotic embryo
It is worth noting that most of the structures observed possessed a prominent

suspensor (Fig. 1) quite similar to that found in zygotic embryos. From these
results it can be assumed that somatic embryos induced in leaf explants go
through analogous developmental stages and originate from a single initial cell.
These are supported by the type of successive cell divisions preceding embryo
formation after leaf mesophyll protoplast isolation. These data would suggest
the presence of embryogenic cells in the explants showing a direct embryogenic
response.
Previous studies with Medicago have shown that the somatic embryogenic
capacity to form embryos varies widely among individual cultivars and lines
(Atanassov and Brown 1984; Brown and Atanassov 1985; Redenbaugh et al.
1986; Seitz and Bingham 1988). Moreover, in many systems, the developmental
stage of the mother plant was shown to be crucial to embryogenic response
(Williams and Maheswaran 1986). One possible explanation to all these vari-
ations is that they are due to differences in the hormonal status in the starting
explant (Denchev et al. 1990; Okudo et al. 1991). In order to confirm this
explanation a set of experiments was carried out to determine the levels of
endogenous cytokinins, and abscisic and indole-3-acetic acid in tube plants of
Medicago falcata lines possessing different embryogenic capacities. The highest
level of all cytokinins studied (Z, ZR, 2ip, and 2iPA) was found in the leaves of
the most embryogenic line. The line with a moderate embryogenic potency had
on average a 50% lower cytokinin level, and the level in the nonembryogenic lines
was very low (Fig. 2). The content of the zeatin-type cytokinins was much higher
than that of the 2iP-type, which is in good agreement with the results of
Hashizume et al. (1985).
The abscisic acid level was highest in the nonembryogenic line, 50% lower
than that in the line with moderate embryogenic potency and still lower
in the most embryogenic line. No significant differences were found in the free
IAA level (Fig. 2); the IAA level was slightly higher in the nonembryogenic
line. Many authors stated that the developmental and physiological stage
of the explant is crucial to the capacity for somatic embryogenesis (Tisserat
et al. 1979; Williams and Maheswaran 1986). However, to our knowledge,
differences in embryogenic capacity have not been proved so far to be related to
the hormonal status of the original explant with the exception of orchardgrass, in
which basal parts of leaves of embryogenic clones contained a much lower
content (three to five times) of cytokinins than that of nonembryogenic clones.
The lAA level was slightly higher in nonembryogenic clones (Wenck et al. 1988).
It is difficult to explain the opposite trend in cytokinin level in relation to
embryogenic potency found by Wenck et al. (1988) and in our work, while
changes in lAA levels were roughly the same. One difference which might be of
importance is that we determined, besides zeatin and zeatin riboside,
isopentenyladenine and isopentenyladenosine; the most pronounced differences
being found in the level ofisopentenyladenosine. Wenck et al. (1988) determined,
besides zeatin and zeatin riboside, dihydrozeatin and its riboside. Thus, the
results are not fully comparable, since the levels of Z and ZR show opposite
trends, i.e., a higher level in embryogenic lines in alfalfa and a lower one in
orchardgrass.
198 P.O. Denchev and A.I. Atanassov
100,---------------. 1000r---------------.
8 8
6 6
4 4
2 2
4 60
50
40
30
20
10
0
21PA
2500 ~----~------~
200
150
100
50
•• - M.falca ta 47/1-165-iowembryog ni
- M.falcata 47/1 - 5 - moderate e mbryoge nic line

lin e
D -M .fa l ata 47/1-150 - high mbryog e ni line
Fig. 2. Endogenous phytohonnonallevels in alfalfa M. fa/cata (data presented in ng/g)
It is also difficult to say whether there are some differences between mono-
cotyledonous and dicotyledonous plants. Until analyses in further plant species
can b~ performed, no reasonable explanation for the described discrepancy can
be given. Also, the decreasing ABA level with rising embryogenic potency might
be of importance. To our knowledge, this is the first report on ABA level in
original explants used for direct somatic embryogenesis.
Leaving aside the role of ABA in the process of somatic embryogenesis, the
embryogenic capacity may be a function of a high cytokinin/ABA ratio; as
suggested by Michler and Lineberger (1987).
2.2 Somatic Embryo Development and Plant Regeneration
Several factors have been identified that can influence embryo development: cell
density (Halperin 1970), the role of ammonium (Walker and Sato 1981),
carbohydrates, and osmolarity (Stuart and Redenbaugh 1987). Our experiments
proved that globular embryos underwent further development only after culti-
vation in a medium containing PEG 6000 (Serva) (Denchev et al. 1991 b). These
results suggested that a concentration of 2.5% PEG was optimum for somatic
embryo development. However, when PEG was added to the medium at higher
concentrations (5 and 10%) embryo development was strongly suppressed.
It was shown that the embryos of alfalfa germinated at a frequency of 85%
after maturation for 2 weeks in the presence of 10 11M ABA (Redenbaugh et al.
1991). Torpedo-shaped embryos originating from selected clones were cultured
on B5 solid medium, or B5liquid medium containing different concentrations of
ABA. The optimum concentration for both media was found to be 30 11M. All
embryos developed into vigorous plants on regenerative medium.
Embryo conversion to plants was carried out on MS medium containing
27 11M GA. This step was critical for further embryo development. All plants
produced showed a normal morphology.
3 Future Trends
Attempts at large-scale embryo production by bioreactors in different species

have been reviewed in several papers (Styer 1985; Chen et al. 1987; Stuart et al.
1987; Moo-Young and Chisti 1988; Preil et al. 1988; Preil 1990; Bajaj 1991;
Takayama 1991; Gupta et al. 1993). Bioreactor application in somatic embryo
production does not necessarily imply the use oflarge-volume bioreactors used in
microbial fermentations since the embryos are counted individually.
Several factors contribute to the application of laboratory discoveries in
industrial processes. An industrial bioprocess based on somatic embryogenesis
procedures would consist of two major operations: a bioreactor process and
downstream processing.
Surprisingly low activities have been observed in bioreactor application for
plant propagation on a commercial scale. This is perhaps due to the widespread
skepticism based on expected large somaclonal variation in suspension culture-
derived progenies. Those difficulties could be surmounted either by finding stable
genotypes (PreiI1990), or applying direct somatic embryogenesis in large-scale
embryo production.
The attention paid to the bioreactor design, scaleup, and operating
parameters reflects its central role in successful commercial ventures. The choice
of an optimal bioreactor configuration for a given process depends on a number
of factors which might appear mutually contradictory (Moo-Young and Chisti
1988; Takayama 1991; Denchev et al. 1992). For example, airliftfermenters exert
low shear force and are the most suitable for growing alfalfa cells. However, they
200 P.O. Denchev and A.I. Atanassov
have proved unsuitable for producing alfalfa somatic embryos via indirect
somatic embryogenesis (Chen et al. 1987).
3.1 Bioreactor Production of Alfalfa Somatic Embryos
In Canada, l.l - 1.7 x lOb alfalfa seeds per acre are planted. The retail cost of
seeds in North America is approximately $300 million. Therefore, greater
quantities of "convertible" embryos are needed in comparison with high-value
crops. This goal could be attained not only by volume increase but by
optimization of the process. Different types of fermentation systems for alfalfa
embryo production via indirect somatic embryogenesis and factors that influence
the process have been examined (Chen et al. 1987; Stuart et al. 1987). Bioreactor
studies of growth and nutrient utilization in alfalfa suspension cultures have also
been carried out (McDonald and Jackman 1989).
The availability of embryogenic systems based on liquid media makes it
possible to produce somatic embryos in large vessels and to scaleup the whole
process to an economically feasible level. However, a basic understanding of the
kinetics associated with embryo development up to the torpedo stage and
substrate consumption especially sucrose utilization in flasks is needed. On the
Plant ma te rial Embryo induction
@$
@ =:>
ExpJa nt
Embryo development
1 - .- 1 =:>
Cuting
1
21 851V
Meas ure me nts

Filtration
•
cell d en sity
viability
number of
embryos
fresh weJght
pH
HPLC analysis
85M
Fig. 3. Experimental protocol based on general procedure for direct somatic embryogenesis in
Medicago
other hand, explant preparation will require some form of mechanization if the
method is to be used on a large scale.
Taking these factors into consideration, we conducted a set of experiments
with Medicago falcala line 47/1- 5 (see Fig. 3; Denchev et al. 1993).
3.2 Explant Preparation and Somatic Embryo Development
Microscope examination showed that material prepared by scalpel contained

larger fragments than that prepared with a mechanical homogenizer. The release
of cells was delayed by approximately 5 days in the material prepared by the
homogenizer and the number of cells released after 10 and 30 days at the end of
induction was 1.3 and 7.l times less than that released after preparation by
scalpel. However, the viability of the released cells during the induction period
was much higher after homogenizer preparation. At day 25 of the induction
period viability of the released cells after 30-s homogenization was 1.7 times
higher in comparison with cutting by scalpel (Fig.4). Differences were also seen
in the embryo yields at the end of incubation. The total number of embryos was
highest in the (30-s) homogenizer preparation and the proportion of torpedo
embryos was also much higher. The conversion rate was about 80% and no
significant difference between the two cutting methods was observed.
Data concerning mechanical cutting showed that both differentiation and/or
dedifferentiation are strongly influenced by the cutting method. The process of
300 , - - - - - - - - - - - - - - - - - - , 100
o
250
o 80
o
: 200
E
"-
<IJ
"§ 150
40 _ homogenizer 10s(Hl0)
100 ~ homogenizer 30s(H30)
o sca l pel cutting( S K)
20 Viabi lit y Hl0
50
-- 0- -- V i ab ilit y H30
Viab ility SC
o
o 5 10 15 20 25
days of induction
Fig. 4. Cell density and viability after homogenizing ex plants

Table 1. The increase in total fresh weight, cell density, embryo number and pH during somatic
embryo development
Days' Fresh weight Cell density Number of embryos pH

(g/501 m!) (xJ0 6/m1) (per 50 m!)
4 0.86±0.09a 0.29±0.03a 9.2±J.9Ia 5.05±0.70abc

6 0.98±0.04a 0.45±0.02ab 25.2±3.20a 5.18±0.!2abc
8 1.26±0.0!ab 0.88±0.14b 57.6±6.0Ib 5.26±0.17abc
11 J.84±0.07bc 0.80±0.12b 70.8±4.25bc 4.52±0.54a
13 2.07±0.03cd J.49±O.l4c 85.4±6.00bc 5.45±0.06bc
IS 2.6!±0.2Id 2.!5±O.l5d 72.6±7.80bc 5.06±O.l4abc
17 3.30±0.03e 2.89±0.0ge 62.8±8.l4 5.79±0.0Id
a Values followed by the same latter are not significantly different at the 5% level.
dedifferentiation during homogenizer preparation was strongly depressed, thus

creating possibilities for embryogenic cells to give rise to embryos. These data can
perhaps be interpreted as a result of wound response (Potrykus 1990) to the kind
of wounding (slicing or tearing off). Released cells were not able to form embryos
under these conditions, but they could be used for embryo production in
suspension culture after further induction. The time-course experiments showed
a strong correlation between the time of culturing and the increase in fresh weight
(r = 0.92; Y= 4.50114 + 8.12593 x). Over the 17-day incubation period the fresh
weight increased nearly sevenfold (from 0.5 g/50 ml to 3.3 g/50 ml), and the cell
density reached 2.89/g x 106/ml (Table 1). It should be pointed out that the
increase in the FW over 8-11 days was due to the increasing number of
developing embryos. There were no significant differences between the average
amount of released cells for those two terms, whereas the embryo yield showed an
increase (p =0.05) (Table 1). The total number of embryos was maximum at day
13 (85.4 ± 6.0) and the proportion of globular embryos decreased with time,
whereas torpedo embryos increased. The proportion of heart-shaped embryos
remained fairly constant and cotyledons and polyembryos only appeared at day
16. However, there were no significant differences between the embryo yields
observed at days 11,13, and 15 (Table 1). Regression analysis showed a good
correlation (r = 0.6) between total number of embryos and cultivation term.
Particularly, the coefficient values for heart and torpedo stage embryos were
= =
estimated to be rh 0.87 and rt 0.86, respectively.
Similar results were obtained by Cazzulino et al. (1990) in carrot somatic
embryo culture where the number of embryos dropped rapidly after day 16. The
authors proposed that the death rate was dependent on carbohydrate source.
3.3 Sucrose Utilization During Embryo Development
HPLC analysis showed that sucrose was hydrolyzed completely on day 13, while
the monosaccharides, glucose and fructose, reached their peak (Fig. 5). However,
in comparison with the results obtained from alfalfa (McDonald and Jackman
1989) and carrot (Cazzulino et al. 1990) suspension cultures, in our experiments,
sucrose hydrolysis took longer.
After day 13 a large sugar uptake occurred which correlated with the fresh
weight and cell density increase (Table 1). At the same time, the number of
embryos tended to decrease (Fig. 5).
These results encourage us to propose the hypothesis that the increased
quantity of both monosaccharides up to day 13 is due to their role as an osmotic
factor and not as a carbohydrate source. The observed pH variations typically
characterized growing alfalfa suspension cultures (Stuart et al. 1987; McDonald
and Jackman 1989). One day after embryo inoculation the initial pH of the
medium dropped by about 1 pH unit. Between days 1 and 3 the pH began to rise
and at the end of the culture period (day 17) it reached 5.7. The analysis of
variance showed that there were no significant differences between the means
=
estimated for pH values at days 4, 6,8, and 11 (p 0.01) (Table 1).
4 Conclusions and Prospects
Since somatic embryos can now be produced in large quantities, they are ideally
suited for automated mass micropropagation (see Bajaj 1991). Somatic embryo-
derived plants have been successfully transferred to soil, and trees are being field
sugars gil SENo

30.--------------------------------.100
25
80
20
60
15
40
10 -A- SUCROSE
~ GLUCOSE
20
5 -e- FRUCTOSE
D SE
o 1 246 8 U D ~ ~
days
Fig. S. Sucrose hydrolysis during somatic embryo development

tested by a number of workers (Gupta et al. 1993). Extensive studies are being
conducted especially on conifers (Attree and F owke 1991), i.e., various species of
Picea (Krogstrup 1990; Webster et al. 1990) and Pinus radiata (Smith 1991).
Merkle et al. (1991) developed a protocol for somatic embryos of yellow poplar
(Liriodendron), and over 5500 plants were field tested.
The large-scale production of somatic embryos from many high-value
crops is now possible. More effort should be made to elucidate the theoretical
and practical problems, so that the technology can become commercially
competitive. This can lead to a rapid introduction of new valuable plant varieties
to the market obtained through selection, breeding programs, or genetic
manipulation.
References
Akita M, Takayama M (1988) Mass production of potato tubers using jar fermenter techniques. Acta
Hortic 230: 55-61
Atanassov AI, Brown DC (1984) Plant regeneration from suspension culture and mesophyll
protoplasts of Medicago sativa. Plant Cell Tissue Organ Cult 3: 149-162
Attree SM, Fowke LC (1991) Micropropagation through somatic embryogenesis in conifers. In: Bajaj
YPS (ed) Biotechnology in agriculture and forestry, vol 17. High-tech and micropropagation I.
Springer, Berlin Heidelberg New York, pp 53-70
Bajaj YPS (1991) Automated micropropagation for en masse production of plants. In: Bajaj YPS (ed)
Brown D, Atanassov A (1985) Role of genetic background in somatic embryogenesis in Medicago.
Cazzulino LD, Pedersen H, Chin Ch, Styer D (1990) Kinetics of carrot somatic embryo development
in suspension culture. Biotechnol Bioeng 230: 781-786
Chen THH, Thompson BG, Gerson DF (1987) In vitro production of alfalfa somatic embryos in
fermentation systems. J Ferment Technol 65: 353-357
Denchev PD, Velcheva MR, Dragijska R, Kuklin AI, Atanassov AI (1990) Somatic embryogenesis in
Medicago. Biotechnol and Biotechnol Equip 4: 66--70
Denchev PD, Kuklin AI, Ve1cheva MR, Atanassov AI, Kaimaktchiev AC (199Ia) A new approach to
large scale micropropagation in alfalfa. Biotech Forum Europe. Int J Biotechnol3: 126--129
Denchev PD, Velcheva M, Atanassov A (199Ib) A new approach for direct embryogenesis in
Medicago. Plant Cell Rep 10: 338-341
Denchev PD, Svanbaev E, Stefanov I, Atanassov A, Dudits D (199Ic) A tissue culture system for high
frequency somatic embryogenesis in Medicago trautwettery. Biotechnol Biotechnol Equip 3: 48-51
Denchev PD, Kuklin AI, Scragg AH (1992) Somatic embryo production in bioreactors. J Biotechnol
26: 99-109
Denchev PD, Kuklin AI, Atanassov AI, Scragg AH (1993) Kinetic studies of embryo development and
nutrient utilization in an alfalfa direct somatic embryogenic system. Plant Cell Tissue Organ Cult
33: 67-73
Dijak M, Brown DC (1987) Patterns of direct and indirect embryogenesis from mesophyll protoplasts
of Medicago sativa. Plant Cell Tissue Organ Cult 9: 121-130
Dos Santos V, Cutter E, Davey M (1983) Origin and development of somatic embryos in Medicago
sativa (alfalfa). Protoplasma 117: 100-115
Flegmann AW, Wainwright H (1981) Shoot doubling time: a quantitative parameter for
characterizing shoot cultures in vitro. Plant cell Tissue Organ Cult I: 85-92
Giles KL, Morgan WM (1987) Industrial-scale plant micropropagation. TIBTECH 5: 35-39
Gray DJ, Purohit A (1991) Somatic embryogenesis and development of synthetic seed technology. Crit
Rev Plant Sci 10: 33-61
Gupta PK, Pullman G, Timmis R, Kreitinger M, Carlson WC, Grob J, Welty E (1993) Forestry in the
21st century. The biotechnology of somatic embryogenesis. BioI Technology II (4): 454-459
Halperin W (1970) Embryos from somatic plant cells. Symp Int Soc Cult Bioi 9: 161-191
Hashizume T, Kawakita K, Senda S, Sugiyama T (1985) Isolation and identification of endogenous
cytokinins from alfalfa (Medicago sativa L.). Agri Bioi Chern 49: 3481-3484
Kao KN, Michayluk MR (1981) Embryoid formation in alfalfa cell suspension cultures from different
plants. In Vitro 17: 645-648
Kozai T (1990) Micropropagation under photoautotrophic conditions. In: Debergh PC, Zimmerman
RH (eds) Micropropagation. Kluwer, Dordrecht, pp 447-469
Kozai T, Iwanami Y (1988) Effect of CO 2 enrichment and sucrose concentration under high proton flux
on plantlet growth of carnation (Dianthus caryopyllus L.) during preparation stage. J Jpn Soc
Hortic Sci 57: 279-288
Krogstrup P (1990) Effect of culture densities on cell proliferation and regeneration from embryogenic
cell suspensions of Picea sitchensis. Plant Sci 72: 115-123
Levin R, Gaba V, Hirsch S, DeNola S, Vasil I (1988) Automated plant tissue culture for mass
propagation. Bio/Technology 6: 135-140
Lupotto E (1986) The use of single somatic embryo culture in propagating and regenerating lucerne
(Medicago sativa L.) Ann Bot 57: 19-24
Lutz J, Wong J, Rowel, Tricoli D, Lawrence R (1985) Somatic embryogenesis for mass cloning of crop
plants. In: Henke R, Hughes K, Constantline M, Hollaender A (eds) Tissue culture in forestry and
agriculture. Plenum Press, New York, pp 105-116
McDonald K, Jackman A (1989) Bioreactor studies of growth and nutrient utilization in alfalfa
McKersie BD, Bowley SR (1993). Synthetic seeds of alfalfa. In: Redenbaugh K (ed). Synseeds.
Applications of synthetic seeds to crop improvement, CRC Press, Inc. Boca Raton, pp 232-255
McKersie B, Senaratna T, Bowley S, Brown D, Krochko J, Bewley J (1989) Application of artificial
seed technology in the production of hybrid alfalfa (Medicago sativa L). In Vitro Cell Dev Bioi 25:
1183-1188
derived plantlets of yellow-poplar for field testing. In: Proc 21st Southern forest tree improvement
Conf, Knoxville, TN, pp 56--68
Michler CH, Lineberger RD (1987) Effect oflight on somatic embryo development and abscisic acid
levels in carrot suspension cultures. Plant Cell Tissue Organ Cult II: 189-207
Miles GE, Kutz LJ (1991) Robots in transplanting. In: Bajaj YPS (ed) Biotechnology in agriculture
and forestry, vo117. High-tech and micropropagation 1. Springer, Berlin Heidelberg New York, pp
452-468
Moo-Young M, Chisti Y (1988) Considerations for designing bioreactors for shear-sensitive culture.
Bio/Technology 6 (II): 1291-1296
Murashige T (1989) Plant propagation by tissue culture: practice with unrealized potential. In:
Ammirato PV, Evans DA, Sharp WR, Bajaj YPS (eds) Handbook of plant cell culture, vol 5.
Ornamental species. McGraw-Hill, New York, pp 3-9
Okudo H, Wada K, Uemoto S (1991) In vitro morphogenetic response and distribution of endogenous
plant hormones in hypocotyl segments of snapdragon (Antrirrhinum majus L.) Plant Cell Rep 10:
501-504
Potrykus I (1990) Gene transfer to cereals: an assessment. Bio/Technology 8: 535-542
Preil W (1990) Application ofbioreactors in plant propagation. In: Debergh P, Zimmerman RH (eds)
Micropropagation. Kluwer, Dordrecht, pp 427-448
Preil W, Florek P, Wix U, Beck A (1988) Towards mass propagation by use ofbioreactors. Acta Hortic
226: 99-106
Redenbaugh K (1990) Application of artificial seed to tropical crops. HortScience 25: 251-255
Redenbaugh K, Paasch B, NicholJ, Kossler M, Viss P, Walker K (1986) Somatic seeds -encapsulation
of asexual embryos. Bio/Technology 4: 797-801
Redenbaugh K, Viss P, Slade D, Fujii JA (1987) Scaleup: artificial seeds. In: Green CE, Somers DA,
Hackett WP, Biesboer DD (eds) Plant tissue culture, Alan R Liss, New York, pp 473-493
206 P.D. Denchev and A.I. Atanassov: Micropropagation Through Somatic Embryos
Redenbaugh MR, Slade D, Fujii JA (1988) United States Patent Number 4,777,762. Oct 18,1988
Redenbaugh K, Fujii J, Slade D, Viss P, Kossler M (1991) Artificial seeds - encapsulated somatic
micropropagation I. Springer, Berlin Heidelberg New York, pp 395--416
Seitz CM, Bingham ET (1988) Interactions of highly regenerative genotypes of alfalfa (Medicago
sativa) and tissue culture protocols. In Vitro Cell Dev Bioi 24: 1047-1052
Senaratna T, McKersie N, Bowley S (1990) Artificial seeds of alfalfa (Medicago sativa L.) Induction of
desiccation tolerance in somatic embryos. In Vitro Cell Dev Bioi 26: 85-90
Smith DR (1991) An automated bioreactor system for mass propagation of Pinus radiata. Agricell Rep
17:2
Stuart D, Redenbaugh K (1987) Use of somatic embryogenesis for the regeneration of plants. In:
LeBaron JH, Mumma RO, Honeycutt RC, Duesing MH (eds) Biotechnology in agricultural
chemistry, Am Chern Soc Washington, DC, pp 87-96
Stuart DA, Strickland SG, Walker KA (1987) Bioreactor production of alfalfa somatic embryos.
Styer DJ (1985) Bioreactor technology for plant propagation. In: Henke RR, Hughes K, Constantin
M, Hollender A (eds) Tissue culture in forestry and agriculture. Plenum Press, New York, pp
117-130
Takayama S (1991) Mass propagation of plants through shake- and bioreactor-culture techniques.
In: Bajaj YPS (ed) Biotechnology in agriculture and forestry, vol 17. High-tech and micro-
propagation I. Springer, Berlin Heidelberg New York, pp 490-515
Tisserat B (1991) Automated systems. In: Bajaj YPS (ed) Biotechnology in agriculture and forestry,
vol 17. High-tech and micropropagation I. Springer, Berlin Heidelberg New York pp 419--431
Tisserat B, Esan E, Murashige T (1979) Somatic embryogenesis in angiosperms. Agric Rev I: 1-78
Walker KA, Sato SJ (1981) Morphogenesis in callus tissue of Medicago sativa: the role of ammonium
ion in somatic embryogenesis. Plant Cell Tissue Organ Cult I: 109-121
Webster FB, Roberts DR, McInnis SM, Sutton BCS (1990) Propagation of interior spruce by somatic
embryogenesis. Can J For Res 20: 1759-1765
Wenck AR, Conger BV, Trigiano RN, Sams CE (1988) Inhibition of somatic embryogenesis in
orchardgrass by endogenous cytokinins. Plant Physiol 88: 990-992
behavior of cells as an embryogenic group. Ann Bot 57: 443--462
Ziv M (1991) Morphogenic patterns of plants micropropagated in liquid medium in shaken flasks or
large-scale bioreactor cultures. Isr J Bot 40: 145-153
11.7 Genetic Transformation of Somatic Embryos
1 General Account
Although DNA can be introduced into virtually any plant cell, several biological
constraints still exist for the regeneration of transformed plants. These biological
constraints are linked to the ability of an individual cell to respond to both the
introduced DNA and to a regeneration stimulus. For transformation success, the
plant cell must have certain attributes which include (1) the competence to
express the introduced DNA; (2) the competence for stable integration of
introduced DNA; and (3) the competence to regenerate into a whole plant. The
competence to express introduced DNA at a high level as well as ability to
differentiate plants from single cells has made somatic embryogenesis a useful
tool in plant transformation. In fact, somatic embryogenesis has been
instrumental in the regeneration of transformed plants from recalcitrant species
in such divergent plant groups as monocots and gymnosperms. Examples of
somatic embryogenesis used in the regeneration of transformed plants are listed
in Table 1.
While many reports use somatic embryogenesis for the regeneration of
transformed plants, the DNA is actually introduced into immature zygotic
embryos or seedling tissues. In this chapter, the focus is on targeting of somatic
embryos for gene transfer. Because the differentiation of embryos follows a
dynamic continuum from a single-celled zygote to a multi tis sued embryo, the
distinction of when a multicelled embryogenic cluster becomes an embryo is
blurred. This is best illustrated with conifers, where somatic embryogenic
cultures have been referred to as embryonal suspensor masses (Gupta and
Durzan 1986). These cultures consist of well-defined "proembryos" made up of
a multicellular embryonal head region subtended by suspensor cells. It is
therefore unclear when this multicelled cluster becomes a pro embryo or embryo
and when it is still a cluster of cells with embryogenic potential. For the purpose
of this chapter, the term proembryo will refer to the multicelled clusters in
embryogenic cultures which have the potential to form an embryo. Once a
somatic proembryo has a distinguished protoderm, it will be referred to as a
somatic embryo.
J Department of HorticuIture, University of Wisconsin, 1575 Linden Drive, Madison, WI 53706, USA

Somatic Embryogenesis and Synthetic Seed I (ed. by y'P.S. Bajaj)
(iJSpringer-Verlag Berlin Heidelberg 1995
Table 1. Examples of the use of somatic embryos/embryogenic cell suspensions for the transformation of plants ~
00
Plant species Tissue culture Transformation Reference

system used system used
A vena sativa Embryogenic cultures Microprojectiles Somers et al. (1992)

Brassica napus Microspore embryoid Microinjection Neuhaus et al. (1987)
Brassica napus Microspore proembryo Agrobacterium Pechan (1989)
Carica papaya Embryogenic cultures Microprojectiles Fitch et al. (1990)
Carica papr.ya Somatic embryos Microprojectiles Fitch et al. (1990)
Carica papaya Embryogenic cultures Agrobacterium Fitch et al. (1993)
Carya illinoensis Somatic embryos Agrobacterium McGranahan et al. (1993)
Daucus carota Embryogenic cultures Agrobacterium Scott and Draper (1987)
Glycine max Embryogenic cultures Microprojectiles Finer and McMullen (1991); Sato et al. (1993)
Gossypium hirsutum Embryogenic cultures Microprojectiles Finer and McMullen (1990)
Juglans regia Somatic embryos Agrobacterium McGranhan et al. (1988, 1990)
Liriodendron tulipifera Embryogenic cultures Microprojectiles Wilde et al. (1992)
Mangifera indica Somatic proembryos Agrobacterium Mathews et al. (1992)
Picea glauca Somatic embryos Microprojectiles Ellis et al. (1993); Bommineni et al. (\993)
Saccharum spp. Em bryogenic cultures Microprojectiles Bower and Birch (1992)
Triticum aestivum Embryogenic callus Microprojectiles Vasil et al. (1992)
Vilis rupestris Somatic embryos Agrobacterium Mullins et al. (1990)
Zeamays Embryogenic cultures Microprojectiles Gordon-Kamm et al. (1990); Walters et al. (1992)
~
gJ
0;'
Genetic Transformation of Somatic Embryos 209
Emphasis in this chapter is placed on the use of somatic embryos with

morphologically distinct tissues for DNA insertion. However, a brief overview of
the transformation of embryogenic multicellular clusters is also included as these
clusters have been referred to as proembryos, embryos, or embryoids in the
published literature. This overview is followed by specific examples oftransfor-
mation using somatic embryos. Finally, the transformation of Picea glauca is
discussed in detail as an example of the advantages of using somatic embryos in
the development of a transformation system.
1.1 Somatic Embryogenesis and Genetic Engineering
The general pattern of competence for in vitro regeneration in plants is that the
most responsive tissues are embryonic tissues and competence for regeneration
declines with time as the plant develops. For example, generally immature
embryos are the most competent tissues to respond to an embryogenic induction
stimulus, however, this competence is rapidly lost as the embryo matures (Merkle
et al. 1990; Becwar 1993).
Of particular importance for the genetic engineering of plants is that those
cells in the immature embryo that differentiate into embryogenic cultures are
surface cells and therefore are accessible to gene transfer by numerous methods.
This has allowed the use of immature zygotic embryos for the transformation of
several plants (Fitch et al. 1990; Christou et al. 1992; Sato et al. 1993). Unfortu-
nately, different development stages of zygotic embryos differ in their
competence to respond to embryogenic induction stimuli. For example, in
conifers there is a small development window when maximum embryogenic
potential can be realized (Webb et al. 1989; Becwar 1993). Therefore, the
identification and large-scale excision of the proper zygotic embryo development
stage for the induction of embryogenic cultures is difficult. Since somatic
embryos follow the same development program as zygotic embryos, large
numbers of developmentally uniform somatic embryos can be induced and used
in place of zygotic embryos. Such a strategy was crucial in the development of a
transformation system for Picea glauca (Ellis et al. 1993)
1.2 Gene Transfer Using Somatic Proembryos
Since embryogenic suspension cultures consist of multicellular clusters,

transformation invariably involves the transfer of DNA into cells within these
clusters. Although these clusters have been referred to as proembryos or
embryoids in several systems, the actual verification that these cell clusters are
proembryos has been done in only a few systems. This verification involves the
isolation of the clusters and the subsequent formation of embryos directly from
the isolated clusters. To further complicate the terminology, embryogenic multi-
celled clusters are referred to as embryoids, proembryos, and calli when no
morphological differentiation exists between these clusters. Clearly, in many
cases these multicelled clusters in embryogenic suspensions are cytologically
210 D. Ellis
similar to embryogenic callus cultures growing on solid medium. Therefore, the

distinction may be somewhat arbitrary. However, because of the discrepancy in
terminology, included here is a brief discussion on the transformation of plants
using embryogenic multicelled clusters.
In rapeseed (Brassica napus), transformed plants were regenerated following
the microinjection of microspore-derived embryoids (Neuhaus et al. 1987).
Twelve-celled embryoids proved optimal, yet microinjection of later stages of
somatic embryos, up to heart-shaped embryos, was also successful. Mter
microinjection, these multicelled embryoids were individually cultured in drop
cultures and allowed to mature for 6 to 10 days. Under the proper conditions,
transformation frequencies between 27 and 51% were reported. The primary
transformants were chimeric, yet secondary embryogenesis of the microinjected
embryoids proved successful at regenerating nonchimeric secondary embryos.
Although neomycin phospho transferase (NPT) II activity was assayed, no
selection was reported in the regeneration of the transformants.
Rapeseed microspore-derived proembryos have also been transformed using
Agrobacterium tumefaciens (Pechan 1989). With this system, both hygromycin
and kanamycin selections were used, with hygromycin selection being preferable.
Only 1% of the proembryos survived the cocultivation regime compared with
2.5% survival of the nonselected controls. Of the proembryos surviving the
cocultivation, 29% developed into plantlets on hygromycin selection.
In several systems, enrichment of embryogenic suspension cultures for small
multicelled clusters of cytoplasmically dense cells has been important. In maize
(Zea mays) embryogenic suspension cultures were allowed to settle, and the
denser cell fraction was subcultured. Using particle bombardment, these cultures
were transformed and fertile transformed plants were regenerated (Gordon-
Kamm et al. 1990). The use of the selectable marker gene bar allowed the
proliferation of transformed callus lines on the herbicide bialaphos.
In yellow poplar (Liriodendron tulipifera) embryogenic clusters were sub-
jected to particle bombardment and kanamycin-resistant calli were proliferated
under kanamycin selection (Wilde et al. 1992). As with the maize system, the
enrichment of the culture for small multicelled clusters was important. With
yellow poplar this was achieved by screening the cultures prior to bombardment.
Kanamycin-resistant colonies were identified within 5-6 weeks and plants
expressing both selectable and marker genes were regenerated. Cocultivation of
similar cultures with a disarmed Agrobacterium tumefaciens strain was not
successful due to a lack of survival of the cultures after selection (Wilde and
Merkle 1994).
Somatic proembryos of mango (Mangifera indica) have been successfully
transformed with Agrobacterium tumefaciens (Mathews et al. 1992). Proembryo
masses < 1000 f.lm in diameter were lightly macerated and cocultivated for 3 days.
Following cocultivation, the proembryo masses turned black within 7 days on
both solid and liquid medium. On solid medium containing kanamycin selection,
localized areas of cell division proliferated within 150 days from these blackened
Agrobacterium-treated tissues. Although all the lines which initially proliferated
were chimeric, the nontransformed cells could be selected by increasing the
kanamycin level to 400 /lg/ml. Mature somatic embryos expressing
Genetic Transfonnation of Somatic Embryos 211
p-glucuronidase (GUS) in all cells have been differentiated. Vitrification of plants

germinated from these somatic embryos has been a problem with both
transformed and nontransformed somatic embryos.
Embryogenic suspension cultures of soybean (Glycine max) proliferate by
the division of surface cells on globular somatic embryos. This division results in
a repetitive somatic embryo proliferation with each somatic embryo developing
to the globular stage. Although the authors (Finer and McMullen 1991) refer to
this system as embryogenic suspension cultures, it approaches the use of somatic
embryos for DNA insertion as the culture contains globular embryos that show
clear signs of epidermal differentiation. DNA was introduced into soybean
embryogenic suspension cultures by particle acceleration and stable trans-
formants were regenerated using hygromycin selection. Selection was delayed for
1 to 2 weeks after particle acceleration and the hygromycin level used did not
immediately kill the cells. As with Picea glauca, the survival of neighboring
nontransformed cells was crucial for the proliferation of those individual,
transformed cells.
2 Transformation of Somatic Embryos
Examples of the use of somatic embryos for transformation include walnut

(McGranahan et al. 1988, 1990), papaya (Fitch et al. 1991, 1993), grape (Mullins
et al. 1990), spruce (Bommineni et al. 1993; Ellis et al. 1993), and pecan
(McGranahan et al. 1993). It is interesting that all the reports on the successful
use of somatic embryos for the transformation of plants are with woody plant
species. While this may be coincidence, of the four examples listed above, only
papaya and grape have been transformed using explants other than somatic
embryos. In walnut, repetitive embryogenesis from the somatic embryos
provides a proliferative system where plants arise from individual surface cells.
Similarly, in spruce, embryogenic callus is induced from the division of individual
surface cells. In all these systems, somatic embryos must be competent to respond
to an embryogenic induction system, often in the presence of a selection agent.
Details of each system are summarized below.
2.1 Juglans regia (Walnut)
The first report of the successful regeneration of a plant following transfer of

DNA into somatic embryos was in walnut (McGranahan et al. 1988). Repetitive
embryogenesis from walnut somatic embryos relies on the formation of secon-
dary embryos from single cells on the surface of the cotyledons and hypocotyl
(Polito el al. 1989). This single-celled origin and repetitive embryogenesis occurs
without a callus stage and at a very high frequency which is crucial to the
successful regeneration of transformed plants.
The infection of walnut somatic embryos by Agrobacterium tumefaciens is
dependent on the age of the embryos. The infection frequency is highest in young
212 D. Ellis
embryos, with older somatic embryos infected at a significantly lower frequency.

Galls formed within 3 weeks and although infection frequency was highest in cut
embryos, sonication of the embryos during cocultivation was also successful,
inferring that a physical wound, such as cutting or slicing ofthe embryo, may not
be needed for infection (McGranahan et al. 1988).
For the regeneration of transformed walnut plants, somatic embryos were
infected with a disarmed A. tumefaciens strain and were initially allowed to go
through two rounds of repetitive embryogenesis prior to selection. These subse-
quent rounds of embryogenesis avoided chimeras which may have arisen due to
the infection of a cell in an already differentiating embryo on the surface of the
original explant. A third round of repetitive embryogenesis under kanamycin
selection (75 Ilg/ml) was used to screen out escapes. This kanamycin level was not
lethal to the somatic embryos yet it suppressed nontransformed somatic embryo
proliferation. Transformation was confirmed by both Southern blot analysis and
NPT II assays (McGranahan et al. 1988).
The walnut transformation system has since been improved by the use of a
more virulent A. tumefaciens strain (EHAlOl) (Hood et al. 1986), the use of
marker genes (gus), and the early selection of transformants with kanamycin.
Although initial detection of transformants after 1-2 weeks using gus was not
possible due to Agrobacterium contamination, the detection of GUS enzyme
activity did allow the detection of transformants in the secondary embryos. After
exposure to x-gluc, transformed secondary embryos turned blue initially at the
cut or wounded surface and within hours the entire embryo would be blue. With
the early application of selection, 1-2 days after cocultivation, transformed
somatic embryos were induced from 21% of the cocultivated embryos, in
contrast to only 3% without early kanamycin selection. These improvements in
the transformation of walnut allow the positive identification of transformants
after 5-6 weeks and have eliminated the need for a third round of repetitive
embryogenesis (McGranahan et al. 1990).
Plants regenerated from the transformed somatic embryos express gus in
both the stems and leaves (McGranahan et al. 1990). Transformed walnut plants
containing a Bacillus thuringiensis (B.t.) endotoxin gene have been placed in a
field trial to assess gene expression in these long-lived perennials. Although the
trees have not reached sexual maturity, it is anticipated that these field-planted
transgenics will be allowed to flower.
2.2 Carica papaya (Papaya)
Three papaya tissues competent to respond to an embryogenic stimulus

(immature zygotic embryos, hypocotyl sections, and somatic embryos) were
compared as starting explants for transformation by Agrobacterium (Fitch et al.
1993) and particle bombardment (Fitch et al. 1990, 1992). The somatic embryos
were not isolated individually, yet were well-differentiated somatic embryos on
callus. The number of somatic embryos was not counted, making comparison of
the three explants based on transformation frequency per explant difficult.
However, the use of these three explants does provide important comparisons of
differential responses to gene transfer by different explants and provides further

advantages to the use of somatic embryos.
With particle bombardment, immature zygotic embryos were pretreated on
an embryogenic induction medium for up to 23 days prior to bombardment.
Under this induction regime, somatic embryos were induced within 5 weeks from
the apex of the zygotic embryo. During the 23-day pretreatment period prior to
particle bombardment, early stage somatic embryos were already differentiating
from the apex (Fitch and Manshardt 1990). Since this differentiating apical
region of the zygotic embryo is the target for gene transfer, the DNA may
actually be inserted into a differentiating, somatic, globular embryo which
undergoes further secondary embryogenesis to form transformed somatic
embryos.
Following bombardment, all tissues are placed on embryogenic induction
medium for 3-5 weeks prior to selection with kanamycin (75 Ilg/ml). This
kanamycin level suppresses the induction of embryos yet it does not suppress the
development of somatic embryos already induced. Tissues were subsequently
transferred to 15 J.lg/ml kanamycin after 4-6 weeks. Histochemical detection of
transient gus expression 3 weeks after particle acceleration indicated that the
zygotic embryos have a higher level of gene expression compared to the other
explants, although expression was low even in the zygotic embryos. Of the 240
zygotic embryos tested, 47 GUS-positive spots were detected on 23 embryos
compared with only 8 spots on an equivalent surface area of somatic embryos.
Hypocotyl explants have the lowest level of transient GUS expression and were
not extensively tested (Fitch et al. 1990). It would be interesting to determine if
pretreatment of the hypocotyl sections could increase the level of transient
expreSSIOn.
Despite the higher transient expression in zygotic embryos, the co-integra-
tion of the npt II and the gus genes, based on expression, was higher in somatic
embryos. Thirty percent of the transformants derived from zygotic embryos
expressed both genes, while 75% of those derived from somatic embryos
expressed both gus and npt II (Fitch et al. 1990). The lack of co-integration was
suggested to be due to the size of the plasmid inserted (18 kb), which the authors
suggest would be more susceptible to fragmentation during bombardment than
a smaller plasmid. This does not account, however, for the higher frequency of
co-integration in the somatic embryos. Further, the differences may not be
significant due to the small number of regenerates from the somatic embryos.
In a subsequent set of experiments, the transformation frequency in zygotic
embryos was 1.4%, with 74% of the regenerated lines developing normally. This
is contrasted with somatic embryogenic cultures where only 25% of the trans-
formed lines regenerated and of those, only two regenerated normal plants
(Fitch et al. 1992).
The transformation of papaya by A. tumefaciens has also been reported. As
with particle bombardment, hypocotyl sections were not a suitable explant.
Although somatic embryos showed hypersensitive reactions, evidenced by the
formation of brown spots following cocultivation, transformants were
regenerated after 6-9 months. This is in contrast to particle acceleration where
transformants were identified in approximately 4 months. Interestingly, in
214 D. Ellis
contrast to particle acceleration, all transformants from the Agrobacterium-

infected embryos tested positive for all the inserted genes (Fitch el al. 1993). Since
both studies used the same plasmid, this supports the suggestion that the lack
of co-integration of genes in the earlier studies could have been due to
fragmentation of the plasmid during particle bombardment (Fitch et al. 1990).
A gene encoding the papaya ringspot virus coat protein has been inserted
into papaya and, as with the expression of gus, the expression of the coat protein
is variable among the different transformed lines. Ofthe four transformed lines
tested for resistance to the ringspot virus, one line showed 100% resistance in
greenhouse studies. This line is currently being field-tested to determine whether
it retains resistance under a continual challenge of virus inoculation by aphids
(Fitch et al. 1992).
2.3 Vitis rupestris (Grape)
Although the transformation of grape rootstock did not use somatic embryos per
se, it did use hypocotyl disks derived from somatic embryos. These hypocotyl
disks were cocultivated with the disarmed A. tumefaciens strain LBA4404 and
transformed shoots were regenerated via organogenesis under kanamycin
selection (Mullins et al. 1990). The selection regime was crucial for the
regeneration of transformed shoots. Like spruce (Eillis et al. 1989), grape is very
sensitive to kanamycin, with levels as low as 5 J.lglml being inhibitory to plant
formation. At higher kanamycin levels, 15 and 20 J.lglml, callus and even a few
GUS expressing buds formed, yet no transformed plants could be recovered
under these higher kanamycin levels. At 10 J.lglml kanamycin transformed plants
were regenerated which expressed GUS in all tissues tested. In the absence of
selection, no transformed shoots were recovered. It appeared that kanamycin
affected the bud to plant conversion in grape yet it allowed some organogenic and
caulogenic response (Mullins et al. 1990).
2.4 Carya illinoensis (Pecan)
Using the system developed for walnut, three pecan genotypes have been trans-
formed with the disarmed A. tumefaciens strain EHA101 (Hood et al. 1986).
Following cocultivation of somatic embryos with the Agrobacterium, transfor-
mants were regenerated by secondary embryogenesis under kanamycin selection.
Since pecan was more sensitive to kanamycin than walnut, a kanamycin level of
50 Ilglml was tested, although no difference in the number of transformants was
detected between this and a higher (l00 Ilglml) kanamycin level. Differences in
transformation frequency were seen between the genotypes tested, with one
genotype, Schely 7, yielding no transformants. The highest frequency of transfor-
mation observed, 6.1%, was with Elliot 6 (McGranahan et al. 1993).
Table 2. The expression of GUS in pretreated zygotic and stage 5 somatic embryos of white spruce at
various times following particle bombardment. GUS was regulated by an enhanced CaMV 35s
promoter. GUS expression is expressed as the number of blue foci (± SE) per embryo assayed
histochemically by x-glue. (Ellis et al. 1993)
Days following No. of GUS spots! No. of GUS spots!

bombardment zygotic embryos somatic embryos
2 36.1 (4.6) 18.4 (0.8)

7 30.3 (3.3) 17.0 (l.l)
14 5.1 (1.0) 15.3 (1.7)
21 7.4 (2.2) 15.6 (1.3)
28 1.4 (0.7) nd
35 nd" 11.7 (1.2)
56 2.6 (1.3) 3.6 (0.6)
and, No data for time point.
3 Transformation of Picea glauca Using Somatic Embryos
White spruce (Picea glauca) provides an example where the use of somatic
embryos has definite advantages over other explants for transformation.
Detailed studies on the time course of gene expression in zygotic embryos
revealed that expression of introduced genes declined prior to the onset of cell
division. Since the expression of antibiotic resistance genes is required for
selection, the expression of these genes is necessary during differentiation. In
contrast to zygotic embryos, the high initial expression level of introduced genes
in somatic embryos is maintained. In order to understand the importance of
somatic embryos for the transformation of white spruce, a detailed comparison
of the use of particle acceleration for the introduction of DNA into both zygotic
and somatic embryos follows.
The expression of gus following particle acceleration in white spruce zygotic
embryos has been characterized for numerous promoters (Ellis et al. 1991; Ellis
1993). Further, pretreatment ofthe zygotic embryos on a bud induction medium
increases the competence of the embryos to transiently express gus. Although the
initial level of transient expression 2 days following particle acceleration can be
relatively high, this level of expression rapidly declines after 1 week to a low
baseline level of expression which slowly declines over the next 8 weeks (Table 2).
This rapid decline in gene expression is very consistent, and efforts to alter this
expression pattern by either environmental or hormonal manipulations have not
been successful. Because cell division and the differentiation of adventitious buds
take weeks to induce, a continual high level of gene expression is important for
the expression of genes encoding resistance to the selective agent.
Since all previous work was done with mature zygotic embryos, it was
reasoned that different developmental stages of embryos could also differ in their
competence to express the introduced DNA due to the dependence on a specific
developmental stage for a high level of transient gene expression. However,
because of seasonality, the labor involved in embryo excision, and the variation
216 D. Ellis
within a single cone in embryo development, it is not practical to obtain large

numbers of different development stages of zygotic embryos. The use of somatic
embryos alleviated this problem because large numbers of different embryo
development stages can be readily obtained.
Our early work showed that proembryos from a proliferating embryogenic
culture were highly variable in the level of transient gus expression. However, this
variability may have more to do with physical constraints of the gene transfer
system than the competence to express the DNA (Ellis et al. 1991). Somatic
embryos differentiated from these cultures were competent to express gus and, as
expected, the level of transient expression was correlated to the developmental
stage of the explant. Early globular somatic embryos had a relatively low level
of transient expression and this level progressively increased as the embryos
developed (Ellis et al. 1993). Stage 5 cotyledonary embryos had the highest level
of transient GUS expression. Since the earlier development stages of somatic
embryos are smaller, it is reasonable that the differences in transient gene
expression could be a result ofless surface area on the earlier embryo stages. This
explanation is unlikely because whether expressed on a surface area basis or on a
per embryo basis, the trends are the same for the transient expression, inferring
that as the embryo develops, the competence to express introduced DNA
increases.
As mentioned, one limitation to the use of zygotic embryos was the rapid
decline in gus expression. In contrast, maximal gus expression was maintained for
up to 3 weeks in the somatic embryos (Table 2). Interestingly, all somatic embryo
development stages had a similar pattern of gus expression over time despite
differences in the initial transient level of expression. Further, early and late
cotyledonary embryos, although different in the level of transient expression, had
similar levels of GUS expression after 8 weeks. Like the zygotic embryos, this
pattern of GUS expression over time was independent of the promoter used to
express GUS as both an enhanced CaMV 35s and the wheat Em promoters
expressed similarly over time (Ellis et al. 1993).
This sustained initial high level of transient gene expression is crucial
because the proliferation of embryogenic callus can be delayed up to 2 weeks
following particle acceleration compared to nonbombarded embryos. This infers
that cell division following particle acceleration may be delayed until the cell
recovers from the bombardment. Damage to the cell following particle bom-
bardment has been reported (Russell et al. 1992). Therefore, due to the sustained
high level of transient gene expression, when cell division finally occurs in the
somatic embryo, cells are still expressing the npt II gene and are therefore
resistant to kanamycin.
As with transient gene expression, the induction of embryogenic cultures
from white spruce is very dependent on the developmental stage ofthe embryo.
With zygotic embryos, the developmental window for optimum embryogenesis
can be as short as 2 weeks. Due to variation in embryo maturation, the time
when this optimum developmental stage occurs can vary considerably from
year to year. Somatic embryos, like zygotic embryos, can be induced to
form embryogenic cultures (Eastman et al. 1991). The frequency at which
embryogenic cultures are reinduced from somatic embryos is also dependent on
the developmental stage of the somatic embryo, yet the optimal developmental
stage may vary with differences in treatment of the somatic embryos (Ellis et al.
1993).
The development of a selection regime for the recovery of transformed
cultures was difficult because white spruce is very sensitive to antibiotics, such as
kanamycin, used for selection (Ellis et al. 1989). The use of a lethal level of
selection to kill the nontransformed cells has not worked with white spruce. This
is due not only to the relatively slow rate of cell division, but also to the fact that
dying cells of many woody plants, including white spruce, produce compounds
which are detrimental to the surrounding cells. Therefore, the only selection
strategy that works is to apply a kanamycin level which suppresses yet does not
kill non transformed cells.
Although kanamycin is lethal to early stage 2 and 3 somatic embryos at a
concentration of 51lg/ml, this kanamycin level is inhibitory but not only lethal to
stage 4 and 5 cotyledonary somatic embryos. Even at this low kanamycin level
the earlier stages of somatic embryos would turn brown and die within 4 weeks.
Since the first visible signs of cell proliferation occur between 4 and 6 weeks, death
of the embryo eliminates the recovery of transformed cultures. The higher
antibiotic sensitivity of the early developmental stages of somatic embryos
prevented the differentiation of embryogenic material even with a lower (lllg/ml)
kanamycin level. At a higher kanamycin level (10 Ilg/ml), although transformed
callus could be induced, this callus was no longer embryogenic (Robertson et al.
1992). This sensitivity to higher kanamycin levels and the loss of regenerative
capacity is similar to the response noted earlier in grape (Mullins et al. 1990).
At 5 Ilg/ml kanamycin, embryogenic callus differentiated from 25% of the
stage 5 somatic embryos. Of these, approximately 95% were nontransformed
escapes. However, histochemical staining for gus expression proved 100%
effective at screening for the transformed lines. Further, based on PCR screening
of over 100 non-gus expressing callus lines, no non-gus expressing lines were
transformed. This is in contrast to both poplar (McCown et al. 1991) and
cranberry (Serres et al. 1992) where numerous non-gus expressing transformed
lines were identified. After the callus grew to a size of 25 mg, 10 Ilg/ml kanamycin
was also effective in screening for transformants. Nontransformed lines stop
growing at this kanamycin level after 4 weeks, while transformed lines continue
to grow.
Differentiation of somatic embryos and plants has been achieved from over
20 transformed lines. All transformed somatic embryos express gus and stain a
uniform blue in the presence ofx-gluc. Preliminary data suggest that the uniform
blue color of the embryo is due to GUS expression in the epidermal layer of the
somatic embryo. Seedlings germinated from these embryos also express gus and
have been transferred into the greenhouse and the field. Current emphasis is on
the germination of seedlings from all the transformed lines and the testing of the
expression of heterologous promoters in white spruce.
218 D. Ellis
4 Conclusions
Somatic embryos offer several advantages as an explant for transformation in

plant species where either repetitive embryogenesis or the reinitiation of embryo-
genic cultures is possible from somatic embryos. Somatic embryos are highly
competent to respond to hormonal stimuli for in vitro manipulation, are com-
petent to express introduced DNA, and regeneration can be initiated from single
cells on the surface of the embryos enabling the regeneration of nonchimeric
plants. However, the success in the transformation of somatic embryos has been
restricted to tree species. The reason for this could be that these species tend to be
relatively recalcitrant in culture and it is simply that other tissues from these
plants are not competent for genetic manipulation. This inability to use other
tissues presents the one current limitation of the use of somatic embryos for
transformation. Most somatic embryogenic cultures are initiated from immature
zygotic embryos and cannot be initiated from parental tissues of known and
proven superior genotypes. Despite this limitation, somatic embryos provide a
means of supplementing tree improvement programs by inserting unique genes
such as viral and insect resistance.
Acknowledgments. The author thanks Drs. Abhaya Dandekar, Maureen Fitch, Gale McGranahan,
and Scott Merkle for sharing unpublished results, and Dr. Rod Serres for critically reviewing this
chapter.
References
Becwar MR (1993) Conifer somatic embryogenesis and clonal forestry. In: Ahuja MR, Libby WJ (eds)
Clonal forestry I. Genetics and biotechnology. Springer, Berlin Heidelberg New York, pp 200-223
Bommineni VR, Chibbar RN, Dalta RSS, Tsang EWT (1993) Transformation of white spruce (Picea
glauca) somatic embryos by microprojectile bombardment. Plant Cell Rep 13: 17-23
Bower R, Birch RG (1992) Transgenic sugarcane plants via microprojectile bombardment. Plant J
2(3): 409-416
Christou P (1992) Genetic transformation of crop plants using microprojectile bombardment. Plant
J(3): 275-281
Eastman PA, Webster FB, Pitel JA, Roberts DR (1991) Evaluation of somac1onal variation during
somatic embryogenesis of interior spruce (Picea glauca engelmannii complex) using culture
morphology and isozyme analysis. Plant Cell Rep 10: 425- 430
Ellis DD (1993) Transformation in spruce (Picea species). In: Bajaj YPS (ed) Biotechnology in
Ellis DD, Lazaroff WR, Roberts DR, Flinn BS, Webb DT (1989) The effect of antibiotics on
elongation and callus and bud formation from embryonic tissue of Picea glauca. Can J For Res 19:
1343-1346
Ellis DD, McCabe D, Russell D, Martinelli B, McCown BH (1991) Expression of inducible
angiosperm promoters in a gymnosperm, Picea glauca (white spruce). Plant Mol Bioi 17: 19-27
Ellis DD, McCabe DE, McInnis S, Ramachandran R, Russell DR, Wallace KM, Martinell BJ,
Roberts DR, Raffa KF, McCown BH (1993) Stable transformation of Picea glauca by particle
acceleration. Bioffechnology 11: 84-89
Finer JJ, McMullen JJ (1990) Transformation of cotton (Gossypium hirsutum L.) via particle
bombardment. Plant Cell Rep 8: 586-589
Finer JJ, McMullen MD (1991) Transformation of soybean via particle bombardment of embryogenic
suspension culture tissue. In Vitro Cell Dev Bioi 27: 175-182
Fitch MM, Manshardt RM (1990) Somatic embryogenesis and plant regeneration from immature
zygotic embryos of papaya (Carica papaya L.) Plant Cell Rep 9: 320-324
Fitch MM, Manshardt RM, Gonsalves D, Slightom JL, Sanford JC (1990) Stable transformation of
papaya via rnicroprojectile bombardment. Plant Cell Rep 9: 189-194
Fitch MM, Manshardt RM, Gonsalves D, Slightom JL, Sanford JC (1992) Virus resistant papaya
plants derived from tissues bombarded with the coat protein gene of papaya ringspot virus. BioI
Technology 10: 1466-1472
Fitch MM, Manshardt RM, Gonsalves D, Slightom JL (1993) Transgenic papaya plants from
Agrobacterium-mediated transformation of somatic embryos. Plant Cell Rep 12: 245-249
Gordon-Kamm WJ, Spencer TM, Magano ML, Adams TR, Daines RJ, Start WG, O'Brien JV,
Chambers SA, Adams Jr WR, Willetts NG, Rice TB, Mackey CJ, Krueger RW, Kausch AP,
Lemaux PG (1990) Transformation of maize cells and regeneration of fertile transgenic plants.
Plant Cell 2: 603-618
Gupta PK, Durzan DJ (1986) Somatic polyembryogenesis from callus of mature sugar pine embryos.
BiolTechnology 4: 643-645
Hood E E, Chilton W S, Chilton M-D, Fraley R T (1986) T-DNA and opine synthetic loci in tumors
incited by Agrobacterium tumefaciens A281 on soybean and alfalfa plants. J Bacteriol 168:
1283-1290
Mathews H, Litz RE, Wilde HD, Merkle SA, Wetzstein HY (1992) Stable integration and expression
of p-glucuronidase and NPT II genes in mango somatic embryos. In Vitro Cell Dev Bioi 28:
172-178
McCown BH, McCabe DE, Russell DR, Robison DJ, Barton KA, Raffa KF (1991) Stable
transformation of Populus and incorporation of pest resistance by electric discharge particle
acceleration. Plant Cell Rep 9: 590-594
McGranahan GH, Leslie CA, Uratsu SL, Martin LA, Dandekar AM (1988) Agrobacterium-mediated
transformation of walnut somatic embryos and regeneration of transgenic plants. BiolTechnology
6: 800-804
McGranahan GH, Leslie CA, Uratus SL, Dandekar AM (1990) Improved efficiency of the walnut
somatic embryo gene transfer system. Plant Cell Rep 8: 512-516
McGranahan GH, Leslie CA, Dandekar AM, Uratsu SL, Yates IE (1993) Transformation of pecan
and regeneration of transgenic plants. Plant Cell Rep 12: 634-638
Merkle SA, Parrott WA, Williams EG (1990) Applications of somatic embryogenesis and embryo
cloning. In: Bhojwani S S (ed) Plant tissue culture: applications and limitations. Elsevier, New
York, pp 67-101
Mullins MG, Archie Tang FC, Facciotti D (1990) Agrobactrium-mediated genetic transformation of
grapevines: transgenic plants of Vitis rupestris scheele and buds of Vilis vinifera. BiolTechnology
8: 1041-1045
Neuhaus G, Spangenber G, Mittelsten Scheid 0, Schweiger H-G (1987) Transgenic rapeseed plants
obtained by the microinjection of DNA into microspore-derived embryoids. Theor Appl Genet
75: 30-36
Pechan PM (1989) Successful cocultivation of Brassica napus microspores and proembryos with
Agrobacterium. Plant Cell Rep 8: 387-390
Polito VS, McGranahan G, Pinney K, Leslie C (1989) Origin of somatic embryos from repetitively
embryogenic cultures of walnut (Juglans regia L.): implications for Agrobacterium-mediated
transformation. Plant Cell Rep 8: 219-221
Robertson D, Weissinger AK, Ackley R, Glover S, Sederoff RR (1992) Genetic transformation of
Norway spruce (Picea abies (L.) Karst) using somatic embryo explants by rnicroprojectile
bombardment. Plant Mol Bioi 19: 925-935
Russell JA, Roy MK, Sanford JC (1992) Physical trauma and tungsten toxicity reduce the efficiency of
biolistic transformation. Plant Physiol98: 1050-1056
Sato S, Newell C, Kolacz K, Tredo L, Finer J, Hinchee M (1993) Stable transformation via particle
bombardment in two different soybean regeneration systems. Plants Cell Rep 12: 408-413
220 D. Ellis: Genetic Transfonnation of Somatic Embryos
Scott RJ, Draper S (1987) Transfonnation of carrot tissues derived from proembryogenic suspension
cells: a useful model system for gene expression studies in plants. Plant Mol Bioi 8: 265-274
Serres R, Stang E, McCabe D, Russell D, Mahr E, McCown B (1992) Gene transfer using electric
discharge particle bombardment and recovery of transfonned cranberry plants. J Am Soc Hortic
Sci 117(1): 174-180
Sommers DA, Rines HW, Gu W, Kaeppler HF, Bushnell WR (1992) Fertile, transgenic oat plants.
BiolTechnology 10: 1589-1594
Vasil V, Castillo AM, Fromm ME, Vasil IK (1992) Herbicide resistant fertile transgenic wheat plants
obtained by microprojectile bombardment of regenerable embryogenic callus. Bio/Technology 10:
667-674
Walters DA, Vetsch CS, Potts DE, Lundquist RC (1992) Transfonnation and inheritance of a
hygromycin phosphotransferase gene in maize plants. Plant Mol Bioi 18: 189-200
Webb D T, Webster F, Flinn BS, Roberts DR, Ellis DD (1989) Factors influencing the induction of
embryogenic and caulogenic callus from embryos of Picea glauca and P. engelmanii. Can J For Res
19: 1303-1308
Wilde HD, Merkle SA (1994) Genetic transfonnation in Liriodendron tulipifera L. (yellow poplar). In:
Bajaj YPS (ed) Biotechnology in agriculture and forestry, vol 29. Plant protoplasts and genetic
engineering V. Springer, Berlin Heidelberg New York, pp 337-348
plants. Plant Physiol 98: 114-120
11.8 Cryopreservation of Somatic Embryos
Y.P.S. BAJAi
1 Introduction
Since somatic embryos (SEs) can be produced in large numbers in bioreactors

(Denchev et al. 1992), and are being employed for mass micropropagation
(Merkle et al. 1991; Bajaj 1991 a), genetic transformation (Ellis 1993), and for the
production of synthetic seed (Redenbaugh et al. (1991) methods need to be
developed for their storage so that they can be used when desired. Experiments
have been conducted on methods such as desiccation (Gray and Purohit 1991;
Tessereau et al. 1991), nonfrozen low-temperature storage (Hiraoka 1988),
coating with gels (Dumet et al. 1993), and cryopreservation in liquid nitrogen
(LN). This last aspect has attracted considerable attention, especially for the
conservation of germplasm (see Bajaj 1986, 1991b). Somatic embryos/embryo-
genic cultures of a number of plant species have now been successfully cryo-
preserved, and plants have been obtained from retrieved cultures (Table 1). The
prospects of cryopreservation of various types of embryos, i.e., zygotic, nucellar,
somatic, and pollen embryos have been discussed earlier (Bajaj 1985). In this
chapter, studies on the freeze-storage of SEs in liquid nitrogen are discussed.
2 Potential of Freeze Preservation of Somatic Embryos
From a practical point of view the potential of cryopreservation of SEs is four-

fold:
1. Plants with Recalcitrant Seeds. The storage of seed is the customary method
of conservation of germplasm; however, in a number of tree species and
plantation crops the seeds are short-lived recalcitrants (Pence 1990). Such seeds
are sensitive to temperature and humidity, and thus cannot be preserved under
ordinary conditions for long periods due to degeneration of the embryo. Their
maximum longevity varies from a few weeks to a few months. Some common
examples are cacao (Theobroma cacao L.), rubber (Hevea brasiliensis), mango
1 Former Professor of Tissue Culture, Punjab Agricultural University, Ludhiana (Punjab), India
Present address: A-137, New Friends Colony, New Delhi 110 065, India

222 y.P.S. Bajaj
Table 1. Summary of the literature on the cryopreservation of somatic embryos
Plant species References
Aesculus hippocastanum (horse chestnut) Jekkel et al. (1992)

Asparagus officinalis (asparagus) Uragami et al. (1989)
Citrus sinensis (sweet orange) Marin and Duran-Vila (1988)
Coffea arabica (coffee) Bertrand-Desbrunais et al. (1989);
C. canephora Tessereau et al. (1991); Mycock
and Berjak (1992)
Cucumis melD (melon) Shimonishi et al. (1991)
Daucus carota (carrot) Bajaj (1976); Withers et al. (1979),
Lecouteux et al. (1991)
Elaeis guineensis (oilpalm) Engelmann and Duval (1986);
Dumet et al. (1993)
Manihot esculenta (cassava) Sudarmonowati and Henshaw (1990)
Saccharum species (sugarcane) Eksomtramage et al. (1992)
Xanthosoma brasiliense (tanier spinach) Jayos et al. (1993)
(Mangifera indica), chestnut (Castanea crenata), horse chestnut (Aesculus

hippocastanum), and walnut (Juglans spp.) Since in all these plant species somatic
embryos can be induced, cryopreservation would enable the conservation of their
germplasm.
2. Substitute for Seeds and Mass Propagation. Somatic embryos are regarded as
"seeds" in plants which do not set seeds. Since somatic embryos can be produced
in large numbers from cell suspensions grown in bioreactors (Stuart et al. 1987;
Ducos et al. 1993), or from the in vitro culture of segments and tissues, they can
be cryostored and used for mass multiplication of elite and desirable plants/
hybrids in which seeds are either produced in lesser quantities or not set at all.
This aspect can be exploited by the seed industry.
3. Synthetic seed. Though somatic embryo-derived synthetic seeds have now
been reported in a number of plant species (Redenbaugh et al. 1991; also see
Chapters 11.2, 11.3, this Vol.), their viability period is rather short. Cryopreser-
vation would enhance the storage period of synthetic seed.
4. International Exchange of Germplasm. Like the routine use of meristem
cultures/in vitro plantlets of potato and cassava for the exchange of germplasm
(Bajaj 1995), cultures of SEs may also be used. However, since SEs can only be
mass produced in liquid media in shake cultures or in bioreactors which require
continuous aeration, they cannot be transported without being damaged. Thus,
cryopreserved material/synthetic seed would be ideal for this purpose.
3 Methods for the Cryopreservation of Embryos
Cryopreservation involves bringing the cultures to a state of non-dividing and

inactive metabolism. This can be achieved by freezing them in liquid nitrogen
(-196°C) by various methods.
Cryopreservation of Somatic Embryos 223
Fig. lA-H. Regeneration of plants from embryogenic cell suspensions of carrot frozen at various
temperatures. A12-day-old cell suspension containing vacuolated free cells and a proembryoid (at the
time offreezing). B A proembryoid from cultures frozen at - 20°C for I week; C same, in ultraviolet
light; note complete survival. D,E A proembryo along with free cells subjected to -20°C (\ h) and
- 70°C (30 min), and then immersed in liquid nitrogen (cryoprotectant DMSO 5%, thawing at 37°C);
note complete survival of the proembryo, and the death of single cells. F Maturing embryos obtained
from proembryos frozen in liquid nitrogen. G,H Various stages in the regeneration of plantlets in a 2,4-
D-free agar-solidified medium. (8ajaj 1976)
224 Y.P.S. Bajaj
3.1 Conventional Methods
The cultures/somatic embryos are treated with an appropriate cryoprotectant

solution and subjected to freezing. The following three methods have been
generally used for freezing plant tissue cultures (see Bajaj 1979):
1. Fast freezing by directly immersing the cultures in LN.
2. Prefreezing the cultures in various steps to -20, -50, -70°C, and then
-196°C. This cooling is periodic and at definite intervals.
3. Gradual freezing at a determined and slow rate of cooling, generally
I-5°C/min.
3.2 Recently Developed and Modified Methods
Although conventional methods form the basis of any cryopreservation study,

the following modifications and refinements have helped to increase viability.
Moreover, in some cases the expensive sophisticated cryostats for controlled
freezing rates are not needed as the processed material is directly immersed ip.
LN.
1. Vitrification. Vitrification involves treating the cultures with higher con-
centrations of cryoprotectants. The main advantage of vitrification is that the
rate-controlled cryostats are not needed, and the time required for cooling is
reduced.
Vitrification has been successfully used for asparagus SE by Uragami et al.
(1989). They used a mixture of glycerol (22%), ethylene glycol (15%), propylene
glycol (15%), and DMSO (7%) in an MS medium enriched with 0.5 M sorbitol.
Surviving viability of the vitrified SE was 50% and the plantlets were produced
directly from the revived SE. In earlier studies on somatic embryos of carrot
(Withers 1979) and orange (Marin and Duran-Vila 1988) frozen using a
conventional method, the recovery was not due to the survival of a whole
embryo, but due to the recovery of proliferating structures from surviving cells by
secondary embryogenesis.
Almost 100% survi.val of SE of carrot and coffee was observed when using a
high concentration of sucrose (0.85-1 M) as the sole source of cryoprotectant
(Tesserrau et al. 1991). A combination of high sugar and dehydration also
proved successful for immature zygotic embryos of coconut (Assy-Bah and
Engelmann 1992).
2. Alginate Coating. Carrot SE first encapsulated in calcium alginate, followed
by incubation in 10.2% sucrose (as sole source of cryoprotectant), desiccated for
4 h in a laminar flow, and then subjected to rapid freezing in LN gave a viability
of about 92%. These embryos developed directly into plants (Dereuddre et al.
1991).
3. Silica Gel. Dumet et al. (1993) considerably improved the cryoability of oil
palm SE through additional dehydration by keeping them (for 6--18 h in the
dark) in an air-tight box containing 40 g silica gel before immersing them in LN.
This improved process was successfully applied to seven different clones.

Desiccation with silica gel is likely to ensure more reproducible dehydration
conditions than those obtained with a laminar flow cabinet,which can vary,
depending on the air humidity and the speed of the air flow.
4 Studies on the Cryopreservation of Somatic Embryos
A list of plant species in which somatic embryos have been successfully cryo-
preserved is given in Table I. The cryopreservation of embryos is a multiple event
involving dehydration, cryoprotectants, freeze storage, thawing and culture. A
fault at any stage may cause irreparable damage. The survival of the frozen
embryos depends on the extent of damage caused during the pre- and post-
freezing phases, and is influenced by a number offactors such as: (1) genotype
and physiological state of the material; (2) size and stage of development;
(3) water contents/desiccation; (4) nature and concentration of cryoprotectants;
(5) method of freezing; (6) storage temperature; (7) thawing temperature; and (8)
estimation of viability. These factors are discussed below.
1. Genotype and Physiological State of the Material. The genotype plays an
important role in determining the capacity of the plant to withstand low
temperatures. It is now an established fact that different species, varieties,
cultivars, lines, clones, hybrids, etc. show variability and different responses
under the same in vitro conditions. The same is true for reproducibility of results
in cryopreservation work. Not only the genotype as such, but also the
physiological state of the cultures/SE and the conditions under which they are
grown before freezing effect the extent of survival. Tropical plants in general are
presumed to be more sensitive than plants grown under subzero temperatures.
2. Size and Stage of Development. Young and relatively small embryos (Fig.l)
generally show higher survival values than do the maturing embryos, which need
desiccation before freezing. Withers (1979) mentions that as the carrot embryos
proceeded through the torpedo stage toward plantlet formation, recovery
declined rapidly. Lecouteux et al. (1991) observed slightly higher (83%) viability
of heart-shaped embryos than of the torpedo-shaped embryo (79%). Likewise,
pollen embryos of Atropa at the globular and heart-shaped stages withstood
freezing to -196°C and gave higher viability as compared to latter stage embryos
(Bajaj 1977). Moreover, fully differentiated embryos were cryosensitive and only
some of them survived. This was also the case with the zygotic embryos of wheat;
the immature embryos survived the usual protocol, but the maturing embryos
needed prefreezing dehydration (Bajaj 1984a).
3. Water Contents/Desiccation. The large embryos are particularly vulnerable to
cryodamage, and methods of reviving them after freezing still need not to be
refined. The main task is to prevent recrystallization of the ice, and also
intracellular freezing, which is invariably lethal. Thus, the water content of these
embryos needs to be controlled (by desiccation) to bring it to a level at which
226 Y.P.S. Bajaj
intracellular freezing is minimal. For large tissues/embryos, dehydration before

freezing is an important factor. Various methods of desiccation, e.g., exposure
to (1) calcium hydroxide in a desiccator (Bajaj et al. 1987); (2) air current of a
laminar flow chamber (Chin et al. 1988), and (3) silica gel (Dumet et al. 1993),
have been employed.
4. Nature and Concentration of Cryoprotectants. Cryoprotectants have in-
variably been used for the SE, however, in Citrus sinensis no difference was
observed in viability ofSE frozen with (DMSO 10%) or without a cryoprotectant
(Marin and Duran-Vila 1988). In most studies on embryos, cryoprotectants such
as DMSO, sucrose, glycerol, glycol, and sorbitol have been used singly or in
combinations with varied degrees of success. In the author's work on various
types of embryos, i.e., SE of carrot ( Bajaj 1976), pollen embryos of Arachis,
Atropa, Brassica, and Nicotiana (Bajaj 1978, 1983), and zygotic embryos of
wheat and rice (Bajaj 1984a), a mixture of DMSO, sucrose, and glycerol was
successfully used. In studies on SE of asparagus (Uragami et al. 1989) and oil
palm (Dumet et al. 1993) materials treated with high concentrations of cryo-
protectants followed by sudden immersion in LN yielded excellent results.
5. Method of Freezing. Experiments on somatic embryos using various freezing
methods, i.e, prefreezing, slow cooling rate, and sudden immersion in LN, have
been conducted with various degrees of success (Table 1). However, in view of the
recent introduction of the vitrification method, methods of controlled cooling
seem redundant, as the material can be directly immersed and stored in LN.
Thus, in future studies, sudden freezing in LN is likely to be a preferred method.
6. Storage Temperature. Storage of SE would be yet another method for the
conservation of germplasm of plants with recalcitrant seeds. For the storage of
such species, segments of the embryos (Bajaj 1984b) or the embryonal axes
(Pence 1990) have been used in experiments. For short-term storage, methods
such as slow growth, desiccation, and quiescence (Gray and Conger 1985;
Senaratna et al. 1989), and encapsulation of embryos, (Janick et aI. 1989) have
been used. During the storage of embryos there can be two main problems:
precocious germination and the loss of viability with the passage of time. For
instance, Janick et al. (1989) reported a decrease in viability of embryos en-
capsulated in the resin Polyox WSR-N.
When specimens are not stored at a sufficiently low temperature additional
injury to the cells may be caused. For the storage of cells, -20°C is unsuitable, as
metabolic processes continue (see White 1963) and cause denaturation of
proteins (Brandts 1967), which results in cell death. The SE of carrot could not be
stored at -20 or -70°C, whereas those stored at -196°C for 3 months regener-
ated complete plants (Bajaj 1976). The survival rates of carrot SE decreased
significantly at increasing duration of storage at -20°C (Tessereau et al. 1991).
Likewise, in pollen embryos stored at -20°C, the viability was nil at the end of
3 months, whereas cultures stored at -196°C revived (Bajaj 1978). The somatic
embryos of asparagus (Uragami et al. 1989) and oil palm (Engelmann and
Duval 1986) have been successfully stored at -196°C for 6 and 15 months,
respectively.
7. Thawing Temperature. Cryodamage can be avoided by adopting a method of

thawing which discourages growth of ice crystals. Thus, the method of thawing
is the most important postfreezing factor, as most recrystallization of ice takes
place during this phase. A survey of the literature shows that in most studies fast
thawing at 35--40 °C has been used (Table 1).
8. Estimation of Viability. The method of estimation of the retrieved cultures is
of utmost importance. Staining reactions alone, in addition to being insufficient,
may even be misleading in determining viability. Sometimes, immediately after
freezing and thawing, the cells are in a state of cryoshock and show no sign oflife.
However, after an extended period of culture (even up to 6 months), they may
resume growth (Bajaj 1976). Thus, from a practical standpoint, the survival data
should be based on the resumption of growth, which may be judged by the
embryo's ability to (1) increase in size, (2) turn green, (3) proliferate to form
callus, and (4) regenerate shoots, roots, secondary somatic embryos, etc.
5 Conclusions
Somatic embryos are excellent material for the storage and conservation of
germplasm, especially that of plant species with recalcitrant seed. Considerable
improvement has been made in the conventional method of freeze storage, and
new refined methods (such as vitrification, desiccation, gel coating, and
encapsulation), employed singly or in combination have enabled high viability of
somatic embryos after freeze-storage in liquid nitrogen in a number of plant
species. Retrieved cultures have produced normal plants. Cryopreservation of
encapsulated somatic embryos/synthetic seed is yet another important method
for the long-term conservation and exchange of germplasm.
Somatic embryos are looked upon as "seeds" in plants which do not set seed
or only set seed in small quantities, or are short-lived. Since they can be produced
in bioreactors in large numbers from many plant species, including those of
recalcitrant ones, it would be rewarding to freeze them not only for the
preservation of germplasm, but also for large-scale propagation. This is an area
which can be commercially exploited by the seed industry.
References
Assy-Bah B, Engelmann F (1992) Cryopreservation of immature embryos of coconut ( Cocos nucifera

L.). Cryo-Lett 13: 67-74
Bajaj YPS (1976) Regeneration of plants from cell suspensions frozen at -20, -70, and -196 DC. Physiol
Plant 37: 263-268
Bajaj YPS (1977) Survival of Atropa and Nicotiana pollen embryos frozen at -196 DC. CUff Sci
46: 305-307
228 Y.P.S. Bajaj
Bajaj YPS (1978) Effect of super-low temperature on excised anthers and pollen-embryos of Atropa,
Nicotiana, and Petunia. Phytomorphology 28: 171-178
Bajaj YPS (1979) Establishment of germplasm banks through freeze storage of plant tissue cultures
and their implications in agriculture. In: Sharp WR, Larsen PO, Paddock EF, Raghavan V (eds)
Plant cell and tissue culture - Principles and application. Ohio State Univ Press, Columbus, pp
745-774
Bajaj YPS (1983) Regeneration of plants from pollen embryos of Arachis, Brassica and Triticum spp.
cryopreserved for one year. Curr Sci 52: 484--485
Bajaj YPS (1984a) The regeneration of plants from frozen pollen embryos and zygotic embryos of
wheat and rice. Theor Appl Genet 67: 525-528
Bajaj YPS (1984b) Induction of growth in frozen embryos of coconut and ovules of citrus. Curr Sci 53:
1215-1216
Bajaj YPS (1985) Cryopreservation of embryos. In: Kartha KK (ed) Cryopreservation of plant cells
and organs. CRC Press, Boca Raton, pp 227-242
Nuclear techniques and in vitro culture for plant improvement. IAEA/FAO, Vienna 1985, pp
43-57
Bajaj YPS (1991a) Automated micropropagation for en masse production of plants. In: Bajaj YPS (ed)
Bajaj YPS (1991 b) Storage and cryopreservation of in vitro cultures. In: Bajaj YPS (ed) Biotechnology
in agriculture and forestry. vol 17 . High-tech and micropropagation I. Springer, Berlin Heidelberg
Bajaj YPS (1995) Cryopreservation of germplasm of potato (Solanum tuberosum L.) and cassava
(Manihot esculenta Crantz). In: Bajaj YPS (ed) Biotechnology in agriculture and forestry, vol 32.
Cryopreservation of plant germplasm I. Springer, Berlin Heidelberg New York (in prep)
Bajaj YPS. Korneva SB, Gutierrez R, Maribona RH (1987) Freeze preservation of plantlets, excised
meristems and cell cultures of sugarcane (Saccharum officinarum L.) at -196 ·C In: Bose A,
Sengupta P (eds) Proc Int ConfCryogenics. McGraw-Hill, New Delhi, pp 222-226
Bertrand-Desbrunais A, Fabre J, Engelmann F, Dereuddre J, Charrier A (1989) Adventive
embryogenesis recovery from coffee (Coffea arabica L.) somatic embryos after freezing in liquid
nitrogen. CR Acad Sci Paris Ser III 307: 795-801
Brandts JF (1967) Heat effects on proteins and enzymes. In: Rose AH (ed) Thermobiology. Academic
Press, New York
Chin HF, Krishnapillay B, Alang ZC (1988) Cryopreservation of Veitchia and Howea palm embryos:
nondevelopment of the haustorium. Cryo-Letters 9: 372-379
Denchev PD, Kuklin AI, Scragg AH (1992) Somatic embryo production in bioreactors. J Biotech 26:
90-109
Dereuddre J, Blandin S, Hassen N (1991) Resistance of alginate-coated somatic embryos of carrot
(Daucus carota L.) to desiccation and freezing in liquid nitrogen. I. Effect of preculture . Cryo-Lett
12: 135-148
Ducos JP, Bollon H, Pettiard V (1993) Production of carrot somatic embryos in a bioreactor. Appl
Microbiol Biotech 39: 465-470
Dumet D, Engelmann F, Chabrillange N, Duval Y (1993) Cryopreservation of oilpalm (Elaeis
guineensis Jacq.) somatic embryos involving a desiccation step. Plant Cell Rep 12: 352- 355
Eksomtramage T, Paulet F, Guiderdoni E, Glaszmann JC, Engelmann F (1992) Development of a
cryopreservation process for embryogenic calluses of a commercial hybrid of sugarcane
(Saccharum sp.) and application to different varieties. Cryo-Letters 13: 239-252
Ellis D (1993) Transformation in spruce (Picea species). In: Bajaj YPS (ed) Biotechnology in
Engelmann F, Duval Y (1986) Cryoconservation d'embryons somatiques de palmier a huile (Elaeis
guineensis Jacg.): resultats et perspectives d'application. Oleagineux 41: 169- 173
Gray DJ, Conger BV (1985) Quiescence in somatic embryos of orchardgrass (Dactylis glomerata)
induced by desiccation. Am J Bot 72: 816
Gray DJ, Purohit A (1991) Quiescence and dormancy in somatic embryos. In: Bajaj YPS (ed)
Hiraoka N (1988) Nonfrozen storage of cell cultures and its effect on metabolites. In: Bajaj YPS (ed)
Biotechnology in agriculture and forestry, vol4. Medicinal and aromatic plants I. Springer, Berlin
Janick J, Kitto SL, Kim YH(l989) Production of synthetic seed by desiccation and encapsulation. In
Vitro Cell Dev Bioi 25: 1167- 1172
Jayos EE, Hoekstra FA, Staritsky G (1993) Cryopreservation of somatic embryos of tanier spinach
(Xanthosoma brasiliense Desf. Engler). Plant Cell Tissue Organ Cult (in press)
Jekkel Zs, Kiss J, Kiss E, Gyulai G, Heszky LE (1992) Freeze preservation of somatic embryos of
recalcitrant seed species: horse chestnut (Aesculus hippocastanum L.) Proc XIIIth EUCARPIA
Congr, Angers, France, July 6-11,1992
Lecouteux C, Florin B, Tessereau H, Bollon H, Petiard V (1991) Cryopreservation of carrot embryos
using a simplified freezing process. Cryo-Letters 12: 319-328
Marin ML, Duran-Vila N (1988) Survival of somatic embryos and recovery of plants of sweet orange
(Citrus sinensis (L.) Osb.) after immersion in liquid nitrogen. Plant Cell Tissue Organ Cult 14:
51-57
derived plantiets of yellow-poplar for field testing. In: Proc 21st Conf. Southern forest tree
improvement, Knoxville, TN, pp 56-68
Mycock DJ, BeIjak P (1992) Cryostorage of somatic embryos of coffee. Proc IVth lnt Worksh on seeds
(in press)
Pence VC (1990) Cryostorage of embryo axes of several large-seeded temperate tree species.
Cryobiology 27: 212-218
Redenbaugh K, Fujii J, Slade D, Viss P, Kossler M (1991) Artificial seeds - encapsulated somatic
Senaratna T, Bryan D, McKersie BD, Bowley SR (1989) Desiccation tolerance of alfalfa (Medicago
sativa L.) somatic embryos. Influence of abscisic acid, stress pretreatments and drying rates. Plant
Sci 65: 253-259
Shimonishi K, Ishikawa M, Suzuki S, Oosawa K (1991) Cryopreservation of melon somatic embryos
by desiccation method. Jpn J Bread 41: 347
Stuart DA, Strickland SG, Walker KA (1987) Bioreactor production of alfalfa somatic embryos.
HortScience 22: 800--803
Sudarmonowati E, Henshaw GG (1990) Cryopreservation of cassava somatic embryos and
embryogenic tissue. Abstr VIIth Int Congr Plant tissue and cell culture, Amsterdam, p 378
Tessereau H, Lecouteux C, Florin B, Schilenger C, Petiard V (1990) Use of a simplified freezing process
and dehydration for the storage of embryogenic cell lines and somatic embryos. Rev Cytol Bioi
Vegetal Bot 14: 297-310
Uragami A, Sakai A, Nagai M, Takahashi T (1989) Survival of cultured cells and somatic embryos of
Asparagus officinalis cryopreserved by vitrification. Plant Cell Rep 8: 418-421
White PR (1963) The cultivation of animal and plant cells. Ronald Press, New York
Withers LA (1979) Freeze preservation of somatic embryos and clonal plantiets of carrot (Daucus
carota.). Plant Physiol 63: 460-467
Section III
Somatic Embryogenesis in Trees
111.1 Somatic Embryogenesis in Horse Chestnut
(A esculus hippocastanum L.)
P. PROFUMO and P. GASTALDO'
1 Introduction
1.1 Botany
Aesculus hippocastanum L. (Horse chestnut, Hippocastanaceae) is an extremely

variable species with a number of varieties differing in growth habit, leaf shape,
and flower features. It is a tree up to 25-30 m high with opposing, long-stalked
leaves, composed of five to nine palmate leaflets. The varicolored flowers are
borne in showy, upright, terminal, branched clusters. The fruit is a large, round,
spiny, fleshy capsule that can be opened by two to three valves. It contains one to
four smooth, shiny brown, globular seeds with a prominent, grayish hilum that
have no endosperm. The embryo has two large, differently sized cotyledons; from
these, a sheath arises surrounding the seedling which is placed in a hollow between
the two cotyledons.
Great interest in the studies on Aesculus hippocastanum is due to the presence
of active substances extracted from the cotyledons, chiefly aescin, which is widely
used in the prevention and treatment of various peripheral, vascular disorders.
This species is also of remarkable phytogeographic significance as a Tertiary relict
and it is extensively planteq as an ornamental tree for its shady crown and highly
decorative flowers.
Aesculus hippocastanum was widely distributed throughout Europe during
the Tertiary, however its natural range was greatly reduced due to past
glaciations. It survives as a native species only in the mountain woods of the
central Balkan peninSUla (former Yugoslavia, Greece, Albania) and in parts of
eastern Bulgaria (Tutin et al. 1968). It was introduced into central Europe in the
sixteenth century, probably from Turkey, and became widespread as an
ornamental tree in Europe and nontropical countries of the world. It is found
locally naturalized in central and western Europe.
I Institute of Botany, University of Genoa, Corso DogaJi lIC, 16136 Genoa, Italy

234 P. Profumo and P. Gastaldo
1.2 Significance of Embryogenesis
If somatic embryogenesis in A. hippocastanum can be induced from explants of

the adult plant, then it can be useful for true to type micropropagation. In cases,
however, even if the starting material is tissue originating directly from the seed,
such as cotyledons or the first leaves of the plantlet, the variations induced by
embryogenic processes could affect the metabolism of the plant. In this case,
embryogenic tissue and/or embryoids might produce abundant active substances
induced through propagation of callus and/or embryoids.
1.3 Review of Earlier Work
The first report on the in vitro culture of A. hippocastanum dates back to 1963,
when Trippi cultured immature and mature branches of this species in spring and
autumn. He obtained more vigorous calli using mature explants without seasonal
differences. In contrast, the ability of young branches to proliferate varied,
probably depending on the seasonal variation of the cytokinin level in the plant
tissues. Heller medium with agar was supplemented with growth regulators at
various concentrations: NAA and K (kinetin) gave the best results for callus
formation.
Radojevic (1978) cultured anthers, isolated from flower buds at various
stages of development, in MS medium with agar supplemented with growth
substances. Androgenic haploid plantlets (n =20) were obtained in the presence
of auxin and kinetin. The author reported that the addition of 2,4-D was
necessary for the induction of androgenic embryoids; furthermore, embryoids
developed into plantlets when transferred to medium without growth regulators.
Later, RadojeviC et al. (1980) conducted an ultrastructural study of androgenic
embryos and showed that at the globular stage they consisted of cells richer in
cytoplasmic organelles than cells of the cotyledonary stage. The abundance of
ribosomes in the youngest embryonal cells was related to highly active synthesis
at this stage. Embryogenesis was intensified by both 2,4-D and GA 3. The possible
growth arrest during development into haploid plantlets (n = 20) could be over-
come by treatment with GA3 and ablation of cotyledons. In the same year,
RadojeviC studied somatic embryogenesis and androgenesis in certain wood
species, and reported that androgenic embryos arose from microspores, and that
anther explants gave best results when they were excised in the uninuclear
microspore stage. In 1988, RadojeviC published a paper concerning somatic
embryogenesis and subsequent plant regeneration from in vitro culture of imma-
ture zygotic embryos. Embryogenic and nonembryogenic calli were obtained in
MS medium with agar supplemented with sucrose, 2,4-D, K, CH (casein hydro-
lysate), and proline. Reduction of the 2,4-D concentration (from 3 to 1 mg/l)
produced an increase in both callus mass and number of embryoids. Germination
of somatic embryos and their development into plantlets were achieved in liquid
MS medium supplemented with glutamine, IAA (or IBA), and GA 3; cytokinins
inhibited plantlet formation. The chromosome number of these plantlets was
2n =40.
Somatic Embryogenesis in Horse Chestnut (Aesculus hippocastanum L.) 235
In addition to the above-mentioned work, Saito (1980) and Chalupa (1990)

induced embryogenesis from callus. Embryogenic callus production was also
observed by Jorgensen (1989) on staminal filaments of A. hippocastanum on
WPM (wood plant medium) supplemented with BAP and 2,4-0; development of
embryoids was induced upon transfer of the callus to the same medium without
2,4-0. Recently, Kiss et al. (1992) have reported highly efficient adventitious
embryogenesis in somatic and zygotic proembryos of horse chestnut (see also
RadojeviC 1991).
2 Somatic Embryogenesis
2.1 Induction of Somatic Embryogenesis
First, high frequency somatic embryos were obtained in explants excised from
primary leaves of horse chestnut seedlings produced in vitro from zygotic em-
bryos (Dameri et al. 1986). MS medium with agar supplemented with sucrose, K,
NAA, and 2,4-D (KND) was used. A whitish, compact callus, E[ (Fig. la), often
appeared before a yellow, friable embryogenic callus, E2 (Fig. 1b), which gave rise
to embryoids. A soft, whitish (Fig. Ie), nonembryogenic (NE) callus differen-
tiated from portions of E[ after mechanical shaking in liquid medium and
subsequent culture on solid medium. The procedure and the results are sum-
marized in Fig. 2. Somatic embryos were cup-shaped or slightly ovoid (Fig. Id,e)
and developed into plantlets when transferred to hormone-free basal agar me-
dium. While the shoot apex developed rapidly into a stem with leaves composed
of three to five leaflets (Fig. If), the lower part of the embryos grew more slowly.
This part was characterized by a sheath similar to that which, in nature, covers
first the root tip, then the radicle. Embryos formed without well-developed
cotyledons produced thin roots not covered by the sheath. This might suggest
that swollen cotyledons on somatic embryos inhibit the precocious development
of roots. Cytological examination of the small embryoids showed that the
genome was composed of 40 chromosomes, which is typical for the horse
chestnut.
In another study, Profumo et al. (1990) cultivated in vitro cotyledonary
explants, not only to induce somatic embryogenesis, but also to investigate
whether a liquid culture system could be used to determine the hormones
eventually necessary for embryogenesis, and to compare the results of these
experiments with data previously obtained with leaf explants. IAA, NAA, K, and
GA3 used separately did not induce embryogenesis as well as the hormone-free,
liquid or solid medium. If growth regulators were added to both liquid and
MS-agar medium, a low concentration of K, NAA, and 2,4-D used in combina-
tion (KND) and of 2,4-D alone produced embryogenic callus. In the agar medium
the callus gave rise to numerous, well-developed embryoids; in the liquid medium
the embryogenic callus formed few embryoids, but it produced a great number of
somatic embryos when transferred to the agar medium.
Fig. 1. a Compact type of callus (E I ) "precursor" ofE,. b Friable type of embryogenic callus (E,).
c Whitish, nonembryogenic callus (NE). d,e Embryogenic callus (E,) with embryoids at various
stages of development. f Plantlet grown in gelled medium without hormcnes. (Dameri et al. 1986)
As regards the comparison between zygotic and somatic embryos obtained

by leaf explants, Dameri et al. (1986) observed that in both somatic embryos with
well-developed cotyledons and zygotic embryos, the lower part of the embryos
was characterized by a sheath with the same origin and morphology (Profumo
et al. 1986). There is, however, a physiological difference: in seed plantlets, the
root actually grows before the cauline apparatus; in vitro, in contrast, the
development of the stem and leaves is rapid, while the development of the radicle
is slow. In 1989, Profumo et al. elucidated the role of the cotyledons in the
morphogenesis of the plant: when the embryoid has developed cotyledons, the
sheath formed by vascular and parenchymatous cotyledonary tissues of zygotic
and somatic embryos first surrounds the axis, then contributes to form the
hypogeal apparatus.
The embryo ids obtained from cotyledonary explants of mature seeds hardly
ever developed into plantlets. This phenomenon could be related to seed
dormancy (Profumo et al. 1987a). When the fruit falls off the tree in autumn, the
LEAF :XP'.ANTS
I
25 days
on coer medium
(MS. KON)
25 cays on agar meciium

on shaKen liauid
(MS·KDN)
medium (MS;KDN)
I
! !
I
on agar medium MOR~ ~I
(MS' KDN)
~
NE CALLUS . E, TLUS
!
MORE NE
EMBRYOIOS
1
MORE NE
Fig.2. Procedure and results in the culture of primary leaves. (Profumo et al. 1987b)
horse chestnut seed is dormant; if sown immediately in the soil, it will germinate
the following spring, after exposure to the winter cold. In 1991, Profumo et al.
regenerated A. hippocastanum plants produced by cotyledonary explants,
keeping somatic embryos for 6 months at 6 °C in the dark and then transferring
them to 25°C (12-h photoperiods). Within about 18 days the plantlets produced
roots and shoots and could be transplanted to soil.
2.2 Histological and Ultrastructural Studies
Histological study (Profumo et al. 1986) showed, first of all, that both calli and
embryoids grew free of a vascular connection with the maternal tissue inside the
leaf explant, and pushed the epidermal tissue outwards until it ruptured. The El
callus showed transverse divisions; the E2 callus showed many highly vacuolated
and elongated nonembryogenic cells (Fig. 3a) with clusters of proembryoids. In
some cases, the embryoids seemed to originate directly from the leaf, but the
histological transformations previously undergone by the explant indicated initi-
ation via a callus intermediary. The embryonic stages were typical, namely,
globular (Fig. 3b), heart-shaped (Fig. 3c), torpedo-shaped (Fig. 3d), and finally
cotyledonary. The differentiated embryoid showed a vascular continuum
Fig. 3. a E, callus: vacuolated and elongated nonembryogenic cells. b Globular embryoid (arrow)
originating from a leaf explant. c Heart-shaped embryoid (small arrow) in a leaf fragment (large
arrow). d Torpedo-shaped embryoid. (Profumo et al. 1986)
between the provascular cylinder, cotyledons, apical shoot, and radicular pole.
The apical shoot differed, depending on age and cutting plane; the root region
completed the embryoid bipolarity. It was covered with a thick sheath as in nature and
often bore new embryoids. Observed under the microscope, the sheath appeared at
the junction of the cotyledonary bases which bore, in a single structure, the vascular
bundles. Inside, another structure belonging to the embryonal axis differentiated. At
first, the axis had hypocoty I characteristics and showed a change from the eustelic to
the actinostelic form. The root appeared after a much longer period of time than was
necessary for the formation of the epigeal parts of the plantlet.
The cell ultrastructure in three types of callus obtained from leaf explants
was studied (Profumo et al. 1987b) and remarkable differences were observed
between the cells of the El callus and those of the callus arising from it. The El
cells obviously lack intercellular spaces (Fig. 4a), the small vacuoles scattered
around the voluminous central vacuole show autophagic activity, the numerous
mitochondria possess very few cristae (Fig. 4b), and the plastids, almost lacking
Fig. 4. E\ callus: a small cells, without obvious intercellular spaces but with a particularly well-
developed vacuole (v) (x 3400). b Mitochondria (m) with few cristae (x48 000). c A plastid (p) almost
devoid of an inner membrane system; no starch is visible (x40 000). E2 callus: d polysomes
(x 100000). e Plastids (p) with large starch grains (s) and a system with a low count of inner
membranes (x33 000). (Profumo eta!. 1987b)
in the inner membrane system, are always devoid of starch (Fig. 4c). Moreover,
in E" the dictyosomes and the endoplasmic reticulum profiles are rare. In E2, the
cytoplasm is very rich in organelles, the small vacuoles have no autophagic
activity, the ribosomes are grouped mainly in poly somes (Fig. 4d), the mito-
chondria sometimes exhibit very long cristae, and the plastids, almost devoid of
an inner membrane system, often have one or more starch grains (Fig. 4e). Active
dictyosomes can be seen in the cells and profiles of endoplasmic reticulum
(Fig. 5a) are frequent . A feature of the embryogenic callus is the diffuse cell wall
degradation, which accounts for the E2 friability. The digestion processes gene-
rally start from the cell corner with the appearance of holes (Fig. 5a), and proceed
along the parietal areas, essentially affecting the middle lamella (Fig. 5b). This
Fig. 5. E, callus: a profiles of endoplasmic reticulum (er) can be seen, with a plastid ( p) with starch
(s) and the beginning of parietal degradation at the cell corner (arrow) (x27 000). b Vacuolated cell
wall area (arrow) shows parietal alteration (v) (x8 000). c Degradative processes (arrow) of the cell
wall (x 10 000). d A massive outgrowth of fibrillar material is visible on the outer side of the cell wall
(cw) (xJ9 000). e An elongated and vacuolated nonembryogenic cell with damaged cytoplasm is
shown (x J 600). (Profumo eta!. J987b)
undergoes most peculiar alterations, with the formation, in the affected extra-
cellular region, of multivesiculated masses (Fig. 5c) and abundant outgrowths of
fibrillar material (Fig. 5d). This phenomenon leads to the detachment of small
embryogenic cellular aggregates and single cells; the latter, elongated as observed
under the light microscope (Fig. 3a), and vacuolated, have damaged cytoplasm
(Fig. Se). In the embryogenic aggregates, whose cells are separated from one
another by thin, undamaged walls, small, unvacuolated cells with large nuclei
(Fig. 6a), rather dense cytoplasm and plastids which are particularly rich in
Fig.6. a Part of an embryogenic cell aggregate in E, callus. Note the thin-walled vacuolated cells and
the small unvacuolated one with a large nucleus, dense cytoplasm and starch-rich plastids (x 5 500).
NE callus: b Very large and vacuolated cells and the reduced cytoplasm areas (x2 300). c The
cytoplasm appears rich in organelles (x 16 000) d The microbody exhibits an inner crystal (x80 000).
V Vacuole; er endoplasmic reticulum; mb microbody; m mitochondrium; cr crystal; p plastid;
d dictyosome; n nucleolus; cw cell wall (Profumo et al. 1987b)
starch, can be distinguished among the large, vacuolated cells. The NE callus
consists of very large cells, delimited by thin walls, with developed intercellular
spaces. The cells have large, central vacuoles (Fig. 6b) and reduced cytoplasm.
The cytoplasmic regions (Fig. 6c) appear well organized and rich in organelles,
with numerous dictyosomes, endoplasmic reticulum profiles, ribosomes (fre-
quently grouped in polysomes), and several mitochondria with a rather electron-
dense matrix. Microbodies are delimited by a single membrane as in E2, but
always contain a large crystal (Fig. 6d) similar to that found in unspecialized
peroxisomes of different achlorophyllous tissues. The plastids are almost devoid
of an inner membrane system and lack starch grains.
Table 1. Aescin content in calli and embryoids from leaf explants of Aesculus hippocastanum L.
cultured in vitro; ripe cotyledons in vivo are used as control. (Profumo et al. 1991 a)
Sample Dry wt. (% fresh wt.) Aescin content/dry wt. (%)
Ripe cotyledons 56.7 10.7 ± 0.5

E, callus 5.9 12.2 ± 1.9
E, callus after 20 days 4.3 62.2 ± 7.3
E, callus after 70 days 3.4 50.0± 2.9
E, callus after 2 years 3.4 31.7 ± 1.8
Embryoids 5.2 69.5 ±4.6
Data are mean ± SEM values; the number of observations is 6 for each sample.
2.3 Aescin Content
We primarily chose in vitro culture of Aesculus hippocastanum due to the

importance of the active substances extracted from its seeds (Profumo et al.
1976), aescin in particular. Our first experiments (Profumo et al. 1976, 1980) on
cotyledon explants grown in vitro did not give results of any interest regarding
aescin content and embryogenesis. By varying the medium composition, extrac-
tion procedure, and aescin determination method, quite interesting results were
obtained in leaf explant cultures. The leaves have a definite structure, however, in
botanical-pharmaceutical literature they have not been considered to be selective
accumulators of aescin. Profumo et al. (1987b) found that E2 vacuoles sometimes
contained a dark, peripheral precipitate similar to that observed in cells of
androgenic embryos of A. hippocastanum and regarded by Radojevic et al.
(1980) as aescin. HPLC determination of the aescin content in calli and em-
bryoids from the same leaf explants gave new results (Profumo et al. 199Ia). The
difference between aescin concentration in E) callus and in cotyledons was not
significant (Table 1). On the contrary, in E2 callus, the aescin content during
the more active phase of proliferation and differentiation is considerably higher
than in controls. Moreover, aescin is about seven times more concentrated in
embryoids than in ripe seed cotyledons in vivo.
2.4 Discussion
In somatic embryogenesis from leaf explants of Aesculus hippocastanum it is

interesting to note that three different calli, E), E2, and NE, were obtained. An
ultrastructural study was carried out in order to elucidate the subcellular features
corresponding to the different morphological and physiological characteristics.
We deemed this survey of some interest because, to our knowledge, such a
comparison has not yet been reported in literature. The differentiation of E2 cells
from E) callus is connected with the acquisition of characteristics indicative of
higher metabolic activity. The polysomes are indicative of active protein
synthesis, the rise in the number of mitochondrial cristae is related usually to a
higher respiratory level, the proliferation of dictyosomes and endoplasmic
reticulum profiles is probably related to the synthesis and secretion of cell wall
lytic enzymes, responsible for the friability shared by E2 and several other
embryogenic callus types. The appearance of starch only in E2 cells may be
related to the presence of carbohydrate exceeding the metabolic demand in the
cells and this leads to the supposition that the appearance of starch in E2 callus
is due to an enhanced ability to absorb sugar from the substrate. The numerous
autophagic vacuoles in El cells, which are not found is E2 callus, recall phe-
nomena of autophagy generally linked to subsequently drastic changes in cell
genetic programs, and interpreted by many authors as a mechanism for the
elimination of long-term, information-carrying molecules. Therefore, the mech-
anical treatment and change in the physical state of medium leading to the
differentiation of nonembryogenic NE callus from El are not easy to ascertain.
Nevertheless, it is logical to maintain that the modified contact and relationship
between El cells and the continuously shaken liquid media may have produced a
different hormonal and/or nutritional situation within the cells, thus being
responsible for the differentiation of cells with a high degree of growth and
enlargement, but lacking the potential to attain embryogenic ability. The NE
cells have an ultrastructural feature that suggests a good metabolic level, but they
are devoid of starch, and this supports the fact that the appearance of starch is a
phenomenon linked to the acquisition of embryogenic competence. The increase
in microbody number and the appearance of a crystal core, which is known to
contain catalase and perhaps different oxidases, have sometimes been connected
with phenomena of cell aging or degradation. In NE callus, continuous tissue
proliferation suggests that the frequency of microbodies is not related to cellular
aging, but may rather depend on a kind of metabolism which requires high
oxidase and catalase activities.
As regards the phytoregulators necessary for embryogenesis, research results
indicate different behavior between leaf and cotyledon explants and lead us to
believe that the chemical composition of the cotyledons, with respect to their
storage function, plays a decisive role in the requirement of exogenous phyto-
regulators in order to achieve embryogenesis. Moreover, the macroscopic,
histological, ultrastructural, and physiological characteristics show an analogy
between zygotic and somatic embryos.
The study by Profumo et al. (l991b) on plant regeneration from
cotyledonary explants after chilling suggests that the interruption in embryoid
development is related to seed dormancy, that the embryoids obtained from
cotyledons of dormant seeds are dormant somatic embryos, and that the cause of
dormancy is to be found in the cotyledons.
Finally, with regard to the aescin content, the studies show that morphological
differences in histological and ultrastructural characteristics are also, correlated with
biochemical differences.
3 Summary
In vitro culture of primary leaf explants from Aesculus hippocastanum resulted in

different types of calli and high frequency embryogenesis. E2 embryogenic callus
or NE (nonembryogenic) callus originated from the E1 callus, according to the
culture conditions. Histological features of embryo ids as well as histological and
ultrastructural features of three types of calli were studied. Cotyledonary
explants also produced embryoids, which acquired the capacity to evolve into
plants only after chilling. Somatic embryos were compared with zygotic embryos.
Finally, the determination of the aescin content indicated that embryogenic calli
and embryoids produce a greater amount of active substances than that of horse
chestnut seeds.
4 Protocol
Explants were excised from primary leaves of Aesculus hippocastanum L. seedlings grown from
sterile zygotic embryos in vitro on a medium consisting of 0.9% agar and 3% sucrose. For callus and
embryoid formation the explants were placed on MS medium supplemented with K, NAA, 2,4-D
(2 mg I of each) and with 3% sucrose, solidified by means of 0.9% agar, at pH 5.5. The explants as
well as calli isolated from them were transferred and subcultured once more on the same medium.
When transferred to hormone-free basal medium, the embryoids gave rise to plantlets. The
material was kept in a culture room at 25 ± I °C and exposed to 3 000 Ix by daylight fluorescent
tubes under a 12/12 h light-dark regime. The same conditions were adopted for the embryogenesis
of cotyledonary explants.
Acknowledgments. The authors are grateful to Prof. Nicoletta Rascio for her competence regarding
ultrastructural interpretation, to Dr. R.M. Dameri and Dr. A.M. Caviglia for their cooperation, and
to Mrs. S. Carli for her technical assistant.
References
Chalupa V (1990) Somatic embryogenesis and plant regeneration in Quercus petraea (Matt.) Lieb.
Tilia platyphyllos Scop., and Aesculus hippocastanum L. Lestnictvi. Forestry 36: 599-604
Dameri RM, Caffaro L, Gastaldo P, Profumo P (1986) Callus formation and embryogenesis with
leaf explants of Aesculus hippocastanum L. J Plant Physiol126: 93-96
Jorgensen J (1989) Somatic embryogenesis in Aesculus hippocastanum L. by culture of filament
callus. J Plant Physiol135: 240-241
Kiss J, Heszky LE, Kiss E, Gyulai G (1992) High efficiency adventive embryogenesis on somatic
embryos of anther, filament and immature proembryo origin in horse-chestnut (Aesculus
hippocastanum L.) tissue culture. Plant Cell Tissue Organ Cult 30: 59-64
List A, Steward FC (1965) The nucellus, embryo sac, endosperm, and embryo of Aesculus
hippocastanum and their interdependence during growth. Ann Bot 29 New Ser (113): 1-14
Murashige T, Skoog F (1962) A revised medium for rapid growth and bioassay with tobacco tissue
culture. Physiol Plant 15: 473-497
Profumo P, Dameri RM, Cremona Orsino I (1976) Frammenti cotiledonari di Aesculus
hippocastanum L. coltivati in vitro. Primi dati suI comportamento dell' amido e dell'escina. G Bot
Ital11O: 155-171
Profumo P, Dameri Orsino RM, Modenesi P (1980) Aescin content in callus from explants of
Aesculus hippocastanum L. cotyledons grown in vitro. G Bot lta1114: 25-28
Profumo P, Gastaldo P, Dameri RM, Caffaro L (1986) Histological study of calli and embryoids
from leaf explants of Aesculus hippocastanum L. J Plant Physiol126 : 97-103
Profumo P, Gastaldo P, Martinucci R (1987a) Variations in aescin content in Aesculus
hippocastanum seeds during the year. Fitoterapia 58 (3): 184-186
Profumo P, Gastaldo P, Rascio N (1987b) Ultrastructural study of different types of callus from leaf
explants of Aesculus hippocastanum. Protoplasma 138: 89-97
Profumo P, Gastaldo P, Dameri RM (1989) Studio preliminare sui rapporti cotiledoni-asse
embrionale in Aesculus hippocastanum L. Boll Soc Ital BioI Sper 65: 603-608
Profumo P, Gastaldo P, Caviglia AM, Dameri RM (1990) Somatic embryogenesis from
cotyledonary explants of Aesculus hippocastanum L. Acta Embryol Morphol Exp New Ser 11 (2)
:101-106
Profumo P, Caviglia AM, Gastaldo P, Dameri RM (199Ia) Aescin content in embryogenic callus
and in embryoids from leaf explants of Aesculus hippocastanum. Planta Med 57: 50-52
Profumo P, Gastaldo P, Bevilacqua L, Carli S (199Ib) Plant regeneration from cotyledonary
explants of Aesculus hippocastanum L. Plant Sci 76: 139-142
Radojevic L (1978) In vitro induction of androgenic plantlets in Aesculus hippocastanum.
Protoplasma 96: 369-374
RadojeviCL (1980) Embryogenese somatique et androgenese chez certaines especes ligneuses. Bull
Soc Bot Fr 127. Actual Bot (3/4): 99-107
RadojeviC L (1988) Plant regeneration of Aesculus hippocastanum L. (horse chestnut) through
somatic embryogenesis. J Plant Physiol132: 322-326
Radojevic L (1991) Horse chestnut (Aesculus spp.) In: Bajaj YPS (ed) Biotechnology in agriculture
and forestry, vol 16. Trees III. Springer, Berlin Heidelberg New York, pp 111-141
RadojevicL, Zylberberg L, Kovoor J (1980) Etude ultrastructurale des embryons androgenetiques
d' Aesculus hippocastanum L. ZPfianzenphysiol98: 255-261
Saito A (1980) In vitro differentiation of embryoid from somatic callus tissues in Aesculus. J Jpn For
Soc 62: 308-310
Trippi VS (1963) Studies on ontogeny and senility in plants. II. Seasonal variation in proliferative
capacity in vitro of tissues from branches from juvenile and adult zones of Aesculus
hippocastanum and Castanea vulgaris. Phyton 20 (2): 146-152
Tutin TG, Heywood VH, Burges NA, Moore DM, Valentine DH, Walters SM, Webb DA (1968)
Flora Europaea, vol2. Cambridge University Press, Cambridge
111.2 Somatic Embryogenesis in Birches (Betula spp.)
A.M. NUUTILAi, U. KURTEN l , R. PUUPPONEN-PIMIA l , J. HAMALAINEN2 ,
L. MANNoNEN l , and V. KAUPPINEN l
1 Introduction
Members of the genus Betula are deciduous trees, bushes, or shrubs with brown
branches and a trunk, the upper part of which is covered with smooth, silver-
white to white bark and the base with blackish, fissured bark. Birches are
monoecious: the purple-brown male flowers are housed is drooping catkins and
the green female catkins are small and erect but hang when mature. The fruit is
a winged achene (Jury 1978).
Birch trees (Betula pendula Roth, B. pubescens Ehrh., B. papyrifera Marsh.,
B. allegheniensis Britt., etc.) are common and widespread throughout Europe,
North America, and Asia where they grow in woods, scrubs, and heaths,
preferring sandy or gravelly soils and open habitats. A woody bush, B. nana L.,
is common in the northern hemisphere, especially in the field district, and prefers
marshes, swamps, and heaths (Jury 1978).
In the Holarctic zone birch, especially B. pendula Roth, is economically and
ecologically very important and is also used in landscaping as well as in
afforestation. In Scandinavia the silver birch is an important raw material for the
wood refining industry as a source of short fiber pulp.
Techniques for the cloning of high quality birch families have been widely
studied in recent years. In Finland for vegetative mass production of birch
(Betula pendula Roth) plantlets have now been used rnicropropagation tech-
niques (Tormiilii 1990). However, somatic embryogenesis offers an interesting
alternative for mass production of birch plantlets in the future.
2 Somatic Embryogenesis in Betula
The first report of the early stages of somatic embryogenesis in birch (Betula
pendula Roth) was published in 1981 by Srivastava and Steinhauer, but fully
expressed somatic embryogenesis in its cell cultures was not reported until 1990
(Kurten et al. 1990; Figs. I and 2).
I VTI Biotechnology and Food Research, P.O. Box 1505, FIN-02044 VTT (Espoo), Finland
2 VTI Automation, P.O. Box 1301, FIN-02044 VTT (Espoo), Finland

~Springer-Verlag Berlin Heidelberg 1995
Fig. la-d. Development of somatic embryos of birch (Betula pendula Roth). a Proembryos in
embryogenic callus on N7 growth medium; b different stages of embryo development: globular stage,
heart stage, torpedo stage, and late torpedo stage; c somatic embryo germinated on N70 medium;
d potted birch plants derived from somatic embryos.(U. Kurten, unpub!.}
Fig. 2. Cross section of an embryo at

the cotyledonary stage, showing shoot
and root apex. (Kurten et a!. 1990)
248 A.M. Nuutila et al.
2.1 Initiation of Embryogenic Cell Cultures
Juvenile tissues are considered to be the most suitable for induction of somatic
embryogenesis. The tissues used to induce embryogenic cell cultures in di-
cotyledons include the cotyledons, ovule, leaf, megagametophyte, embryo,
hypocotyl, anthers, microspores, and endosperm (Tulecke 1987).
The effect of explant type on the induction of embryogenic calli from
B. pendula has been studied using material of different development stages: seeds,
l-week-old seedlings, and leaves from 1-, 2-, 3-year-old plants. Explants from
different birch families have also been tested to investigate the effect of genotype
(Kurten et al. 1990). The explant type does not have a major effect on the
induction of embryogenic calli of birch. When the genotype is suitable, embryo-
genically competent cultures can be derived from seeds and seedlings as well as
from leaf tissue (Kurten et al. 1990).
Table 1. Media' (mg/l) used for embryogenic cell cultures of birch (Betula pendula Roth)
Compound N7 growth N73 N70 WPMO 115 WPMO

medium medium medium medium medium
NH4N0 3 400 80
KN0 3 2830 2830 2830
CaCI2·2Hp 166 166 166 96 19.2
MgS04·7Hp 185 185 185 370 74
KH 2P04 400 400 400 170 34
(NH4)2S04 463 463 463
Ca(NO)2.4HP 556 111.2
K 2S04 990 198
H 3B0 3 6.2 6.2 6.2 6.2 1.24
MnS04.4Hp 22.3 22.3 22.3 22.3 4.46
ZnS04·7Hp 8.6 8.6 8.6 8.6 1.72
KI 0.83 0.83 0.83
NaMo04·2Hp 0.25 0.25 0.25 0.25 0.05
CuS04·5H 2O 0.025 0.025 0.025
CoCI 2CHP 0.025 0.025 0.025
FeS0 4·7Hp 27.8 27.8 27.8 27.3 5.46
Na2EDTA.2HP 37.3 37.3 37.3 37.3 7.46
m-Inositol 100 100 100 100 20
Nicotinic acid 0.5 0.5 0.5 0.5 0.1
Pyridoxine. HCI 0.5 0.5 0.5 0.5 0.1
Thiamine.HCI 0.1 0.1 0.1 1.0 0.2
Glycine 2.0 2.0 2.0 2.0 0.4
Casein hydrolysate 1000 50
Sucrose 20000 20000 20000 10000 20000
2,4-D 2 0.1
Kinetin 0.5 0.02
Agar 8000 8000
Gelrite 3000 3000 6000
pH 5.8 5.8 5.8 5.8 5.8
'N7 basal medium: Simola (1985); N7 growth medium: Kurten et al (1990); N73 medium: Nuutila and
Kauppinen (1992); WPM basal medium: Smith and McCown (1982/1983)
Somatic Embryogenesis in Birches (Betula spp.) 249
Different birch families vary in their ability to produce embryogenic calli.

Only three families out of the eight tested by Kurten et al. (1990) produced
embryogenic calli. The best embryogenic callus induction was achieved with N7
growth medium (Table 1), which was also used for the maintenance of the cell
lines (Kurten et al. 1990). In the best family 29% of the induced cell lines were
embryogenic. The genotypic dependence of embryogenic competence in cell
cultures has been established by following the induction of embryogenic calli
from selected individuals over three successive years.
Using micropropagation, leaves of superior embryogenic genotypes can be
obtained for callus induction independently of the growth season.
2.2 Effect of Growth Conditions on Embryo Production
Embryogenic development in cell cultures of birch is induced by reducing the

hormone concentration, and by replacing agar with Gelrite in the solidified
medium. The removal of larger aggregates by sieving is also beneficial prior to
inoculation to embryo production medium. The cells are spread evenly on the
medium surface or diluted 1:30 [packed cell volume (PCV): ml medium] in shake
flasks (see Sect. 4, Protocol). The embryogenic potential of calli tends to decrease
during continuous subculture. The embryogenic potential can be elevated by
inducing new cultures from somatic embryos.
2.2.1 Optimization of Embryo Production Media
Systematic optimization of growth conditions, which is widely used with

microbes, has also been found useful with plant cells (Tuominen et al. 1989;
Toivonen et al. 1991; Nuutila et al. 1991). Using statistical experimental design
and regression analysis, it is possible to save time and reduce the number of
experiments needed for the optimization. Factorial experimental design has been
used in the optimization of some of the compounds of an embryo production
medium for birch (Nuutila et al. 1991).
Both nitrogen and sucrose are known to have profound effects on the
somatic embryogenesis of different plant species (Ammirato 1983; Gleddie et al.
1983). The ratio of ammonium to nitrate was optimized in birch cell cultures in
a stepwise manner in callus culture using 11 different nitrate:ammonium molar
ratios (Fig. 3). The best embryo production was clearly achieved with a 4 : I
nitrate: ammonium ratio. In suspension culture no clear optimum was found
(Fig. 3; unpubl. results).
The optimum total inorganic nitrogen and sucrose concentrations for
embryogenesis in callus and suspension cultures were estimated using statistical
experimental design and response surface methodology (Fig. 4; Box and Wilson
1951; Cochran and Cox 1957). The optimum sucrose concentration (20.8 gil) was
found to be the same in suspension (unpubl.) and in callus culture (Nuutila et al.
1991). The optimum nitrogen concentration obtained (30 mM) was rather low
when compared with the nitrogen concentration of 60 mM in MS medium
260
240
220
4Il 200
.3c. 180
Ci 160
...>C
~ 140
I:;::l
--120
~100
"8
III
80
60
40
20
o
0:1 1:16 1:8 1:4 1:2 1:1 2:1 4:1 8:1 16:1 1:0
Callus culture ~ Suspension culture
Fig. 3. Effect of the nitrate:ammonium ratio on somatic embryo production in birch callus and
suspension cultures. (A.M. Nuutila, unpubl.)
(Murashige and Skoog 1962), which is widely used for somatic embryogenesis of
various species, including some woody plants, with good results (Ammirato
1983; Meijer and Brown 1987).
All media used for birch somatic embryogenesis were sterilized by
autoclaving. We also tested filter-sterilized media, but embryo production was
less than 50% of the production in autoclaved media.
2.2.2 Nutrient Uptake of an Embryogenic Cell Line
The nutrient uptake of an embryogenic cell line of birch differs considerably from
that of a nonembryogenic birch cell line, the behavior of which is very similar to
that of dicotyledonous plant cell suspensions in general (McDonald and
Jackman 1989; Sagishimaetal. 1989; Treatetal.I989). The embryogenic cell line
rapidly hydrolyzes the sucrose to glucose and fructose, but consumption of the
monosaccharides as well as ammonium and nitrate commences only after the
embryos start to germinate after 3 weeks of culture. This implies that
the embryogenic cell line of birch is self-sufficient with respect to sugars and
nitrate during embryo formation (Nuutila and Kauppinen 1992). By contrast,
the sugar uptake of an embryogenic suspension culture of carrot did not differ
from that of an undifferentiated suspension culture (Cazzullino et al. 1990).
When the cells were grown in a medium without sucrose or without both
sucrose and inorganic nitrogen, the total viability of the embryogenic suspension
decreased but the viable cells left maintained their embryogenic potential. The
34.14.------~----------_<:::""-----__,
. ' ' . '.
///~ ,,
27.07 ", \
I / . . - - - _ .................
,,
f
/
/
/ '-
"- ,
I / I/ ________ \ \ \
\
\
! \
I
I / ""
\ \
\
'" \
.g : I \
l\ )
0 \
c: I
8 20.00 : r 1 I
; \ I I
/
CD
en
e0 \ \ / I
I
::J
\ \ "-- ___ .---/ I I
I
en \ \ / I
\ " /
f
'- ',, ' -,

, /
/ /
/
/
f
"" //
12.93 '" ~~
5. 86 +------,-----===--;,--===-------,--------j
20.86 27.93 35.00 42.07 49.14
Nitrogen cone. (mM)

Number of embryosl100mg inoculum 7 -------- 60 ----- 90
- - - 110 --125
Fig. 4. Embryo production (number of embryos/I 00 mg inoculum) of birch after 28 days as a function
of total inorganic nitrogen concentration and sucrose concentration (Nuutila et al. 1991)
starved suspension produced four times as many embryos per viable cell than the
control. The behavior of the embryogenic cells in suspension provides a
possibility to enrich these cells in a suspension by starvation (Nuutila and
Kauppinen 1992).
2.2.3 Effect of Hormones, Cell Density, and Light
The optimum concentrations of 2,4-D and kinetin for embryo production in

callus cultures of birch were found to be only 5% of the concentrations used for
callus growth. In suspension culture the optima shifted towards higher 2,4-D and
lower kinetin concentrations (unpubl.).
Embryo production in birch cultures, calculated on a fresh weight basis, can
be increased by using an inoculum consisting of single cells and cell aggregates
less than 300 /!M in diameter. The sieving removes larger aggregates that would
otherwise overgrow the smaller, slow-growing embryogenic cells and aggregates.
The effect of inoculum size in callus culture was studied by varying the
amount of sieved cells plated per N73 agar plate between 0.1 to 1 g. The highest
embryo production was achieved with the lowest inoculum (0.1 glplate).
In suspension, several dilutions were tested. The dilutions between 1:20
and 1:40 (sieved cells PCV: ml medium) were found to be best for embryo
production. When smaller or higher dilutions were used, the embryo production
gradually decreased (unpubl.).
Light had no significant effect on birch embryo production in suspension
culture. Embryo production on solidified medium in light was three times that
obtained in the dark (unpubl.).
2.2.4 Production of Somatic Embryos in Bioreactors
An embryogenic cell line of birch was grown on N73 medium in four different
types of bioreactors: a sparged, blade-stirred bioreactor, a diffusion-aerated
bioreactor with a marine impeller (ChemCell, Chemap), an airlift, and a bubble
column bioreactor. In all four bioreactors an aeration of 0.5 . 11-1 min- I was used.
During the cultivation (28 days), the cultures were sampled every 3 days and the
embryo development was monitored by computer vision (see Sect. 2.3). Globular
stage embryos were formed after 14 days in all the bioreactors but in
most cases, especially in the mechanically agitated bioreactors, further develop-
ment ceased.
The highest embryo production after 28 days (2362 embryos/I) was achieved
with the bubble column bioreactor. This production was only 20% of the
production in control shake flasks (12200 embryos/I) (Table 2; unpubl. results).
Table 2. Production of embryos in four types of bioreactors.

(Nuutila, unpubl.)
Bioreactor type Embryo production

(embryosll)
Blade-stirred 570
ChemCell 950
Kluyver flask 1500
Bubble column 2360
Shake flask 12200
2.3 Embryo Classification by Computer Vision
Computer vision can be used for two tasks during somatic embryo production:
(1) calculation of the proportions of embryos at different developmental stages
for estimating the state of the fermentation process, and (2) recognition of
embryos at a predefined developmental stage followed by automated separation
of the embryos from suspension culture.
The suspension in a Petri dish is illuminated by diffuse backlighting from
above so that the silhouettes of the embryos can be viewed through the bottom of
the dish. Figure 5 shows an image of a torpedo stage embryo. Other objects
consisting of undifferentiated cell clusters appear as irregular objects in the
image.
A computer vision system has been developed for the classification of
somatic birch embryos according to their developmental stage (Hamalainen
et al. 1992, 1993). Figure 6 shows the boundaries of globular, heart, and torpedo
stage embryos. It should be noted that the variation in shape of the embryos is
considerable within all classes.
The classification algorithm is based on discarding objects that are clearly
not embryos by minimum calculations and then classifying the remaining objects
as nonembryos, globular stage embryos, and heart and torpedo stage embryos.
An index of the developmental stage is calculated for embryos at the heart and
Fig. 5. A torpedo stage embryo and clusters of undifferentiated cells appear as silhouettes in the image.
(HamaHiinen et al. 1993)
254 A.M. Nuutila et a!.
"
, ,:
\ .... "f.
/
........ -..;. - "/

/",
a b c
Fig. 6a~. Boundaries of embryos at different developmental stages: a globular; b heart; and c torpedo.
Radius in mm and angle marked at intervals of 30° (1.1. Hiimiiliiinen, unpub!.)
torpedo stage. The index describes the relation of embryo breadth to the length
of the root.
Table 3 shows the result obtained when an independent testing set, not used
in the development of the classification algorithm, was classified. None of the
globular embryos were classified as a heart or torpedo and 17% were discarded by
the algorithm. None of the heart or torpedo stage embryos were classified as
being at the globular stage. Of the heart and torpedo stage embryos 12% was
discarded. The embryos at the heart and torpedo stage were classified into three
developmental classes by the above-mentioned index: HTI, HT2, and HT3.
Table 3 shows that none of the heart stage embryos were classified into class HT3
and none of the torpedo stage embryos into class HTl. Thus, the index describes
the true morphological change due to embryo development.
A system for the automated monitoring of samples from the bioreactor was
also described by Hamalainen et al. (1993). This system is based on the use of a
line scan camera with a linear drive unit capable of simultaneously transferring
several Petri dishes. Into the automated classification system it is possible to
include a separation unit for the selection of embryos at a predefined develop-
mental stage.
2.4 Embryo Conversion
Birch somatic embryos are converted to plantlets on hormone-free medium and

subsequently potted in soil and grown under greenhouse conditions. The
Table 3. Classification of an independent testing set by computer vision (Hiimiiliiinen et a!. 1993)
Stage" N G(%) HTI ('Yo) HT2(%) HT3 (%) Discarded
Globular 72 83 0 0 0 17
Heart 32 0 38 50 0 12
Torpedo 42 0 0 45 43 12
G, Globular; HTI, heart-torpedo stage I; HI2, heart-torpedo stage 2; HI3, heart-torpedo stage 3.
aClassified by visual observation
conversion rate of suspension-derived as well as callus-derived embryos is low.

Only 3-5% of the embryos derived on N73 medium (Table 1) form a distinct root
and a shoot apex. With careful selection and by handpicking individual early
cotyledonary stage embryos, the conversion rate can be improved.
Precocious germination of N73 medium (Table 1) produces abnormal
plants with fused, vitrified cotyledons and swollen hypocotyls. Embryos devel-
oping multiple shoots are also frequently observed. These shoots can be rooted
and subsequently transferred to soil (Kurten et al. 1990).
Preliminary maturation studies of selected torpedo stage somatic embryos
indicate the beneficial effect of a 2-week maturation phase prior to germination,
under high humidity on filter paper disks (Boulay et al. 1988) moistened with N7
basal medium supplemented with 0.1 /lM abscisic acid (unpubl.).
The vigor and subsequent conversion of birch somatic embryos is also
affected by callus age and the hormone ratio during embryo production on
solidified medium. Conversion on WPMO (Table 1) is favored by a high
2,4-D/kinetin ratio during embryo formation. Birch cell cultures may retain their
embryogenic capacity for well over 1 year, but with increasing callus age the
embryo quality declines (unpuhl.).
2.5 Molecular Biology of Somatic Embryogenesis in Birch
Somatic embryogenesis is known to be delicate, highly plastic, and sensitive to

stresses induced by culturing conditions (Carman 1990). The efficient application
of somatic embryogenesis to vegetative multiplication requires that the embryo-
genically competent cells can be identified from a heterogeneous cell population
as early as possible. For this purpose, early indicators for embryogenically
competent cells would be very important. So far, only a few genes (Aleith and
Richter 1990) have been reported to be specific for the early, preglobular stages
of somatic embryogenesis. To study gene expression during somatic em-
bryogenesis of birch, a gene named BP8, was isolated from a birch genomic
library using a carrot embryo-specific cDNA clone (Choi et al. 1987), kindly
provided by Dr. Z.R. Sung (University of California, Berkeley) as a probe. The
BP8 gene was sequenced and compared with the corresponding carrot gene, DC8
(Franz et al. 1989). On the basis of DNA sequence analysis, the BP8 gene codes
for a protein of 53 kDa molecular weight and most probably consists of three
exons similar to DC8. The BP8 sequence shows 52% identity with the DC8
sequence at the amino acid level. The central part of the BP8 protein is structured
into 20 repeats of 11 amino acids. According to its structure and homology at the
repeat region, the BP8 belongs to the group of LEA proteins isolated from, e.g.,
carrot, cotton, and wheat. Expression of the BP8 gene was studied in cell cultures
of birch, representing different developmental stages of somatic embryogenesis
and also in nonembryogenic material. Northern hybridization with mRNA did
not give any signals, probably due to a very low expression level of the gene. This
interpretation was supported by low number of signals obtained from the cDNA
library screening. The BP8 gene was shown to be expressed at a low level in
proembryogenic cell aggregates but not in visible embryos (Puupponen-Pimia
et al. 1993). These results suggest that although the expression of the BP8 gene
indicates embryogenesis, its usefulness as a marker gene for embryogenesis in
birch cell cultures is hindered by its low expression level.
2.6 Long-Term Preservation of Embryogenic Cell Lines
Cryopreservation and preservation under mineral oil was done as described by

Mannonen et al. (1990). Embryogenic callus at the exponential growth phase was
transferred in a tube on to solid medium and covered with thick layer of mineral
oil. The tubes were incubated at 25°C for 18 months. The embryogenic capacity
was tested as described by Nuutila et al. (1991) every 2 months after recovery
under normal growth conditions for two passages.
An embryogenic cell line of birch (BP 6/3) was successfully preserved in
liquid nitrogen and under mineral oil. However, the embryogenic capacity was
reduced significantly during prolonged subculture as well as during incubation
under mineral oil. The results from preservation experiments are shown in Table
4 (unpubl.). Similar results were achieved with cryopreservation when other cell
lines were tested (data not shown).
3 Summary
Some factors affecting somatic embryogenesis in cell cultures of birch (B.

pendula) were studied. Induction of embryogenic callus is strongly genotype-
dependent. Explant tissue is not solely confined to juvenile reproductive tissue,
but mature leaf tissue can also be used. This is important if somatic em-
bryogenesis is to be used for cloning of superior genotypes.
Somatic embryos of birch can be produced on solidified medium as well as in
liquid culture. The effect of aggregate size and cell population on embryo
production indicates the importance of identifying embryogenic cells or clusters
in order to obtain a homogeneous cell population to be used as an inoculum. For
Table 4. Comparison of the embryogenic properties of an embryogenic birch cell line (BP 6f3) during
continuous subculture, during storage under mineral oil, and after cryopreservation. (1. Mannonen,
unpub\.)
Starting Continuous subculture Oil preservation Cryopreservation

material
months months months
2 8 12 2 6 8 12 2 12
*** *** * *f- *** *- *f- *f- **** ****

*f-, Less than 10 cmbryosfg fr.wt.; *, 10-100 embryosfg fr.wt.; **, 100-500 embryosfg fr.wt.; ***,
500-1000 embryosfg fr.wt.; ****, over 1000 embryosfg fr.wt.
this purpose, early indicators of embryogenic cells at the level of gene expression
were studied.
The embryogenic potential in cell cultures eventually declines during
prolonged subculture. Cryopreservation was successfully used not only for pre-
serving but also for enriching the embryogenic properties of birch cell cultures.
Somatic birch embryos can also be produced in bioreactors. An automated
classification system, which can be used to monitor embryo development during
culture, was developed. It is also possible to include a separation unit in the
device.
Somatic embryos were converted to seedlings which survived transfer to
greenhouse conditions. Rooted cuttings from embryos forming multiple shoots
wintered in a nursery over three subsequent years. The conversion rate is still low
due to abnormal shoot and root development and must be improved before
somatic embryogenesis can be used for mass propagation.
4 Protocol
4.1 Somatic Embryo Production
Approximately 75 g of 3 week -old embryogenic birch callus is suspended in 250 ml ofliquid N7 growth
medium (Table I) in a 500-ml Erlenmeyer flask. The flasks are incubated on a horizontal rotary shaker
(133 rpm, stroke radius 2.5 cm) for I to 4 days, after which the cell suspension is sieved. The cells
passing through the sieve (343 IJlll pore diameter) are collected by centrifugation (300 g, 4 min). For
embryo production in suspension culture, the cells are suspended in N73 medium (Table!) using I ml
PCV (packed cell volume, 300 g, 4 min) per 30 ml of medium in 100-ml Erlenmeyer flasks. For embryo
production on solidified medium, the centrifuged cell mass is weighed (100 mg! plate) onto embryo
production media plates (40 ml of medium in a (I) 9-cm Perti dish).
All cultures are grown at 23°C. The suspension cultures are grown in the dark on a horizontal
rotary shaker (Infors AG)(133 rpm, stroke radius 2.5 cm) and the callus cultures in a 16-8 h light regime
(light intensity 125IJlllols-Im-2, white fluorescent lamp 30 W AiramlFluora 36 W Osram, I : I).
4.2 Cryopreservation of Embryogenic Cultures
Embryogenic callus cultures of birch are suspended in normal growth medium without agar. After 24 h
of culture on a rotary shaker the suspension is sieved through a 1000-1Jlll screen and adjusted to 20%
PCV. The suspension is precultured in 0.2 M sorbitol for 24 hand 0.4 M sorbitol for another 24 h. For
cryoprotection 5% DMSO is added and the PCV is adjusted to 40%. Freezing is at -0.5 °C/min to
-35°C whereafter the ampoules are plunged in liquid nitrogen. Thawing is performed rapidly in a
+40 °C water bath.
4.3 Embryo Conversion
Early torpedo stage embryos are placed on hormone-free medium N70 or WPMO (Table I) gelled with
Gelrite. For vigorous root growth the plantlets are transferred to 1/ 5 WPMO medium (Table!) with
a reduced salt concentration. Plantlets are potted in a mixture of soil: peat: perlite (I : I : 2) and
acclimatized for 2-3 weeks in high humidity under plastic hoods. Plants are grown under greenhouse
conditions (22°C, 1251!Mol S-l m-2, 16-8 light-dark regime).
References
Ammirato P (1983) Embryogenesis. In: Evans DA, Sharp WR, Ammirato PV, Yamada Y (eds)
Handbook of plant cell culture, vol I. Macmillan, New York, pp 82-123
Boulay MP, Gupta PK, Krogstrup P, Durzan DJ (1988) Development of somatic embryos from cell
suspension cultures of Norway spruce (Picea abies Karst.). Plant Cell Rep 7: 134-137
Box BEP, Wilson KB (1951) On the experimental attainment of optimum conditions, J R Soc 13: 1-45
Carman JG (1990) Embryogenic cells in plant tissue cultures: occurrence and behaviour. In Vitro Cell
Dev Bioi 26: 746-753
Cazzullino DL, Pedersen H, Chin C-K, Styer D (1990) Kinetics of carrot somatic embryos
development in suspension culture. Biotechnol Bioeng 35: 781-786
Choi JH, Liu L, Borkid C, Sung ZR (1987) Cloning of genes developmentally regulated during plant
embryogenesis. Proc Nat! Acad Sci USA 84: 1906-1910
Cochran WG, Cox GM (1957) Experimental designs, 2nd edn. John Wiley, New York
embryogenic gene, DC8, from Daucus carota L. Mol Gen Genet 218: 143-151
Gleddie S, Keller W, Setterfield G (1983) Somatic embryogenesis and plant regeneration from leaf
explants and cell suspensions of Solanum melongena (eggplant). Can J Bot 61: 656-666
Hiimiiliiinen II, Kurten U, Kauppinen V (1992) Automated classification of somatic plant embryos.
Acta Hortic 319: 601-606
Hiimiiliiinen II, Kurten U, Kauppinen V (1993) 'Classification of somatic plant embryos by computer
vision. Biotechnol Bioeng41: 35-42
Jury SL (1978) Betulaceae. In: Heywood VH (ed) Flowering plants of the world. Mayflower,
New York, pp 59-60
Kurten U, Nuutila AM, Kauppinen V, Rousi M (1990) Somatic embryogenesis in cell cultures of birch
(Betulapendula Roth.) Plant Cell Tissue Organ Cult 23: 101-105
Mannonen L, Toivonen L, Kauppinen V (1990) Effects of long-term preservation on growth and
productivity of Panax ginseng and Catharanthus roseus cell cultures. Plant Cell Rep 9: 173-177
McDonald KA, Jackman AP (1989) Bioreactor studies of growth and nutrient utilization in alfalfa
Meijer WGM, Brown DCW (1987) Role of exogenous reduced nitrogen and sucrose in rapid high
frequency somatic embryogenesis in Medicago sativa. Plant Cell Tissue Organ Cult 10: 11-19
Mott RL (1981) Trees. In: BV Conger (ed) Cloning agricultural plants via in vitro techniques. CRC,
Boca Raton, pp 217-254
Nuutila AM, Kauppinen V (1992) Nutrient uptake and growth of an embryogenic and a non-
embryogenic cell line of birch (Betula pendula Roth.) in suspension culture. Plant Cell Tissue Organ
Cult 30: 7-13
Nuutila AM, Kurten U, Kauppinen V (1991) Optimization and inorganic nitrogen concentrations for
somatic embryogenesis of birch (Betula pendula Roth.) callus cultures: a statistical approach. Plant
Cell Tissue Organ Cult 24: 73-77
Puupponen-Pimiii R, Saloheimo M, Vasara T, Ra R, Gaugezc J, Kurten U, Knowles J, Keriinen S,
Kauppinen V (1993) Characterization of a birch (Betula pendula Roth.) embryogenic gene, BP8.
Plant Mol Bioi 23: 423-428
Sagishima K, Kubota K, Ashihara H (1989) Uptake and metabolism of sugars by suspension-cultured
Catharanthus roseus cells. Ann Bot 64: 185-193
Simola LK (1985) Propagation of plant!ets from leaf callus of Betula pendula f. purpurea. Scient Hortic
26: 77-85
Smith MAL, McCown BH (1982/83) A comparison of source tissue for protoplast isolation from three
woody plant species. Plant Sci Lett 28: 149-156
Srivastava PS, Steinhauer A (1981) Regeneration of birch plants from catkin tissue cultures. Plant Sci
Lett 22: 379-386
Toivonen L, Ojala M, Kauppinen V (1991) Studies on the optimization of growth and indole alkaloid
production by hairy root cultures of Catharanthus roseus. Biotechnol Bioeng 37: 673-680
Tormiilii T (1990) Genotype-phenotype interplay in micropropagation. In: Nijkamp HJJ, Van der
Plas LHW, Van Aartrijk J (eds) Progress in plant cellular and molecular biology. Kluwer
Treat WJ, Engler CR, Soltes EJ (1990) Culture ofphotomixotrophic soybean and pine in a modified
fermenter using novel impeller. Biotechnol Bioeng 34: 1191-1202
Tulecke W (1987) Somatic embryogenesis in woody perennials. In: Bonga JM, Durzan DJ (eds) Cell
and tissue culture in forestry, vol 2. Martinus Nijhoff, Dordrecht, pp 61-91
Tuominen U, Toivonen L, Kauppinen V, Markkanen P, Bjork L (1990) Studies on growth and
cardenolide production of Digitalis lanata tissue cultures. Biotechnol Bioeng 33: 558-562
111.3 Somatic Embryogenesis in Papaya
(Carica papaya L.)
M.M.M. FITCH 1
1 General Account
1.1 Importance and Distribution of Papaya
Papaya belongs to the Caricaceae, a family of dicotyledonous, tree-like plants

that is made up of four genera, three of which, Carica, Jacaratia, and Jarilla are
native to tropical America, while Cylicomorpha is native to Africa (Badillo 1967,
1971). The genus Carica consists of about 21 species, but only C. papaya is of
economic importance (Litz 1986a). The ripe fruits of other species are edible, for
example, C. pubescens Lenne et Koch ("candamarcensis"), C. monoica Desf.,
C. erythrocarpa Heilborn., C. x heilbornii nm pentagona (babaco), C. quercifolia
St. Hil (Heiron.), and C. goudotiana Solms-Lauback, but they are more often
consumed in preserves (Purseglove 1968; Jordan et a1. 1982; Litz 1984).
The center of origin of papayas is believed to be southern Mexico or Central
America, and it appears to be a highly domesticated crop because no wild
papayas have been found (Purseglove 1968). The worldwide distribution of
papayas is likely due to the ease with which its numerous seeds can be transported.
Papayas are among the fastest growing fruit trees with flowers produced as
early as 4 months after germination from seed. The annual production rate of
papaya worldwide is 4.43 million metric tons (F AO 1990). More than half of the
crop is grown in South America and about one-fourth is grown in Asia.
Ripe papayas are rich in vitamins A and C (Arriola et a1. 1980). The enzyme
papain is produced in laticifers that are found throughout the plant, but the
richest source of papain is the mature green fruit (Thomas and Beckley 1923).
Papain content decreases as the fruit matures. Fresh fruit as well as papain, the
basic component of meat tenderizer (Poulter and Caygill 1985), are the major
products, while pulp as a juice additive and for preserves and condiments has
some economic importance (Chan and Cavaletto 1982).
I Department of Horticulture, University of Hawaii, Honolulu, HI 96822, USA

Present address: US Department of Agriculture, Agricultural Research Service, Experiment Station
HSPA, P.O. Box 1057, Aiea, HI 96701, USA

Somatic Embryogenesis in Papaya (Carica papaya L.) 261
1.2 Significance of Papaya Somatic Embryos
Worldwide, the papaya crop is beset by several pathogen problems, the most
important being papaya ringspot virus, PRY (Purcifull et al. 1985), the Mediter-
ranean fruit fly (Namba and Kawanishi 1966), and the fungus, Phytophthora
palmivora Butl. (Alvarez and Nelson 1982). Introgression of virus resistance
genes into the commercial germplasm by conventional means has been difficult;
therefore, new methods for crop protection based on genetic engineering and
gene transfer have been sought. A widely successful protocol is coat protein gene-
mediated protection (reviewed in Beachy et al. 1990) whereby a plant is trans-
formed with the coat protein gene of a virus and the resulting transgenic plant is
protected from the virus by mechanisms still not understood. An integral part of
gene transfer methods is efficient plant regeneration. Gene transfer to embryo-
genic tissues can be expected to result in transgenic plants from a large number of
species. Transgenic papaya calli were first demonstrated by Pang and Sanford
(1988) after leaf disk cocultivation with Agrobacterium tumefaciens. Further-
more, Fitch et al. (1990) reported transgenic papaya plants via microprojectile
bombardment of embryogenic tissues from zygotic embryos (Fitch and Manshardt
1990) and hypocotyl sections (Fitch 1993). Some of the transgenic papaya plants
contained the coat protein gene of PR V and were resistant to PR V infection in
greenhouse testing (Fitch 1991; Fitch et al. 1991).
1.3 Review of Somatic Embryogenesis in Papaya and Its Relatives
Protocols for inducing somatic embryos in papaya have been developed for a
variety of reasons, ranging from interest in methods for mass micropropagation
(De Bruijne et al. 1974; Yie and Liaw 1977; Chen et al. 1987; Chen 1988a,b) to a
requirement for recipient tissues for gene transfer technology (Fitch and
Manshardt 1990; Fitch 1993). Ironically, somatic embryos were produced in
cultures of papaya hybrid embryos resulting from crosses between PRY-resistant
wild species and commercial papaya cultivars that were recovered by ovule or
ovary culture or embryo rescue (Litz and Conover 1981a, 1982, 1983; Moore and
Litz 1984; Chen and Kuo 1988; Chen 1988a; Manshardt and Wenslaff 1989a,b;
Chen et al. 1991). Cultures of other Carica explants, C. stipulata peduncles (Litz
and Conover 1980), C. pubescens hypocotyls (Jordan et al. 1982), and C. x
heilbornii nm. pentagona ovules (Vega de Rojas and Kitto 1991) were also highly
embryogenic.
The earliest report of successful somatic embryogenesis in papaya was that of
De Bruijne et al. (1974) who induced somatic embryos from petiole sections
cultured on Murashige and Skoog MS (1962) and White (1963) media (Table 1)
in a multistep protocol. They obtained somatic embryos but were not able to
regenerate plants. Yie and Liaw (1977), Mehdi and Hogan (1979), Chen et al.
(1987), Chen (1988b), and Fitch and Manshardt (1990) subsequently reported on
papaya somatic embryogenesis and plant regeneration. An alternative method of
regenerating papaya plants by shoot organogenesis has been documented or
suggested in a few reports (Arora and Singh 1978; Litz et al. 1983).
Table 1. Media used for somatic embryogenesis and micropropagation of papaya. (Fitch 1991)
N
0..
MEDl Growth MED2 Growth Results N
Species Explant Reference
REG1,mgll REG2,mgll
Somatic embryogenesis
Carica
papaya Petiole MS NAAO.2 W BA 0.002 Embryogenic callus De Bruijne et al. (1974)
2iP2 NAAO.02
or
W ZEA 0.0002 Embryogenic callus
BAO.002
and
O.IMS BAO.002 Somatic embryos,
NAAO.02 no plants
papaya Seedling stem MS NAAI MS IAAO.05 Somatic embryos, Yie and Liaw (1977)
internodes KO.1 K 1-2 plants, micropropa-
gation, roots
papaya Seedling shoot tips, MS IBA,IAA Somatic embryos, Mehdi and Hogan (1979)
mature shoots,internodes, K,NAA plants, micropropa-
seedling internodes MS CW gation, roots
papaya Seedling, 4-6 weeks 'hMS NAAI Somatic embryos Chen et al. (1987);
old: stem, cotyledon, KO.5,GAI and plants from root Chen (1988b)
root, leaf, shoot tip, or 'h MS NAAI callus
3-5mm or 'h MS NAA I,GA I
papaya Seedling, I-cm LS 2,4-D 2 LS 2,4-DO.02 Somatic embryos, Yamamoto and Tabata
pieces KO.2 plants not sought (1989)
papaya Seedling, I-cm pieces LS 2,4-D 2 Callus Yamamoto et al. (1986)
papaya Immature zygotic 'hMS 2,4-DO.5-25 '12MS Somatic embryos, Fitch and Manshardt
embryos, 75-105 days andMS plants (1990)
papaya Ovules, 20-120 days W CW 20% or Somatic embryos, Litz and Conover
x caulifiora BAO.l-l no plants (l98Ib) ~
papaya Ovules, 20-40 days W CW20% W NAAO.1-2 Somatic embryos, Litz and Conover
~
x caulifiora BAO.05-0.2 plants (1982)
papaya Ovules, 20-140 days 'hMS CW20% '12MS 2,4-D 1-2 Somatic embryos, Litz and Conover ~
'Tj
x caulifiora plants (1983) 8"
papaya Ovules, 65 days '12MS CW20% 'hMS Somatic embryos, Moore and Litz ::r
x caulifiora plants (1984)
papaya Ovules, 6S days Y:.MS CW200/0 Y:.MS 2,4-D2 Somatic embryos Litz(1986a)
x cauliflora en
0
papaya Ovules, ovaries Y2MS Y:.MS NAAO.S Somatic embryos, ChenandKuo 8
III
x cauliflora IS-30days plants (1988) g.
papaya Ovules, 60-90 days Y:.MS BAO.OOI Y2MS BAO.OOI Somatic embryos, Chen (1988a) tT1
x cauliflora plants 8
tr
papaya Ovules, 90-180 days MSorW CW200/0 MS NAAO.S Somatic embryos, Manshardt and
x cauliflora BAO.2 plants Wenslaff(1989a) ~
~
papaya Ovules, 60-90 days Y:.MS BAO.OOI Y2MS ±ABAI Somatic embryos, Chen et al. (1991) ::s
<1l
x cauliflora plants f!J.
papaya x pubescens, Ovules, 90-180 days MS orW CW200/0 MS NAAO.S Somatic embryos, Manshardt and '"S·
quercifolia, stipulata BAO.2 plants Wenslaff(1989b) '"CI
III
and pubescens x '0
III
'<
papaya x III
heilbornii nm. Ovules, 23-140 days MS, Dicamba2 Y2MS IAAO.S Somatic embryos, Vega de Rojas g
pentagona Y2MS NAA 0.2S-0.S GAO.I plants and Kitto (1991) ~.
orW BA I Charcoal ~
stipulata Peduncle MS NAAO.2 MS NAAO.2 Somatic embryos, Litz and Conover "'-§"
BAO.4 BAO.4 plants (1980) ~
~
or ±charcoal r
MS BAO.4 -:..."
or
MS
or
MS NAAO.2
±charcoal
pubescens Seedling hypocotyls, NNor NAAI-S Somatic embryos Jordan et al.
2-months old MS KI plants (1982)
Micropropagation
Carica
papaya Seedling apices, deF NAAO.2 MS Micropropagation, Drew and Smith
2-months old, BAO.2 few roots (1986)
apices, 4-months old
papaya Stem sections, MS KI MS IBAS Micropropagation Mehdi and Hogan (1976)
shoot tips KI
papaya Field-grown axillary DS NAAO.2 DS NAAO.OS Micropropagation, Drew (1988) tv
0-,
rooted cuttings or buds, BAO.2 BAO.2 rooted plants to field w
6-months old DS NAAO.I
BAO.4
Table1. (Contd.)
N
Species Explant MEDl Growth MED2 Growth Results Reference ~
REGl,mgll REG2,mgll
Carica
papaya Shoots, deF NAAO.2 DS NAAO.S Micropropagation, Miller and Drew (1990)
6-months old BAO.2 BA2 rooted plants to field
papaya Lateral shoots of MS NAAO.1 Micropropagation, Reuveni et al. (1990)
hedged, rooted BAO.S rooted plants
cuttings or main-
stem buds
papaya Apices MS NAAO.I-2 MS NAAO.l Micropropagation Litz and Conover
K 1-10 or BASor (1977)
IAAO.S-2 NAA2
BAO.04-2 KIO
papaya Apices, mature MS NAA2 MS NAAO.2 Micropropagation Litz and Conover
field-grown KI0 BAO.S (1978a)
papaya Apices MS NAA2 MS NAAO.I Micropropagation Litz and Conover (1978b)
KI0 BAO.S
papaya Shoot tips, mature, MS NAA2 MS NAAO.2 Micropropagation, Litz and Conover (1981 a)
field-grown KI0 BAO.S rooted plants
papaya Apices, field, nursery YoMT NAAO.2 YoMT NAAO.I Micropropagation, De Winnaar (1988)
plants BAO.S BAO.S rooted plants
papaya Seedling, 6-weeks old: MS NAA2 MS NAAO.l Micropropagation Pandey and Rajeevan
stem, petiole, shoot, BS KO.S BAO.5 (1983)
Smm MS NAAI
KIO
papaya Field-grown lateral MS NAAO.1-2 MSor BAO.S Micropropagation, Rajeevan and Pandey
shoots KIO MSor ZEAl fast (1986)
2iP2
MS NAAO.l Short shoots
BAO.S ~
~
Media abbreviations: MS = Murashige and Skoog (1962); LS = Linsmaier and Skoog (1965); W = White (1963); BS = Gamborg et al. (1968); NN = Nitsch and
Nitsch (1969); DS = Drew and Smith (1986); deF = de Fossardetal. (1974); MT = Murashige and Tucker (1969).
~
Other abbreviations: MED 1 = initiation media; MED2 = other media for further development, proliferation, etc.; Growth REG 1 =concentrations of growth ~
regulators for culture initiation; Growth REG2 =concentrations of growth regulators used in other media; NAA = 1-naphthaleneacetic acid; 2iP = 6-(1, 1- ::r
dimethylallyloamino)purine; K = kinetin; IBA = indole-3-butyric acid; IAA = indole-3-acetic acid; CW = coconut water; GA = gibberellic acid; BA =
6-benzylaminopurine; ZEA =zeatin; 2,4-D =2,4-dichlorophenoxyacetic acid; dicamba = 3,6-dichloro-2-methoxybenzoic acid; ABA = abscisic acid.
Somatic Embryogenesis in Papaya ( Carica papaya L.) 265
1.3.1 Somatic Embryogenesis from Ovary, Ovule,

and Immature Embryo Cultures
In an early investigation to breed for PRV resistance, Horovitz and Jimenez

(1958) observed that cultured embryos from crosses between the virus-resistant
C. cauliflora and C. papaya tended to be polyembryonic. They crossed other
Carica species differing in PR V resistance to determine heritability and type of
resistance and observed that F[ progenies from crosses of susceptible C. monoica
with resistant C. pubescens or C. cauliflora were all PRV resistant and when
backcrossed and test-crossed, segregated for PRY resistance in a manner that
suggested that resistance is controlled by a single dominant gene (Horovitz and
Jimenez 1967). Provvidenti and Robinson (1977) and Provvidenti and Gonsalves
(1982) observed similar results in crosses between wild cucurbits.
Several groups have sought to extend the introgression studies by crossing
resistant species with papaya. Ovaries, ovules, or immature zygotic embryos
resulting from interspecific crosses were cultured with the intention of rescuing
the hybrids, screening them for virus resistance, and backcrossing them to papaya
(Litz and Conover 1981b, 1982, 1983; Moore and Litz 1984; Chen 1988a; Chen
and Kuo 1988; Manshardt and Wenslaff 1989a,b; Chen et al. 1991). After
hybridization, fruits 20 to 182 days old, were bisected to free the ovules. Ovary,
ovule, and immature embryo cultures gave rise to large numbers of somatic
embryos on White or 1/2 MS medium devoid of growth regulators but
supplemented with 6% sucrose, 20% CW, and 400 mg/l glutamine. The higher
sucrose concentration, 6% vs. the normal 2 or 3%, and coconut water contributed
to increased osmotic potential, and together with the additional source of reduced
nitrogen provided by glutamine, may have effected the increased somatic embryo
induction.
1.3.2 Somatic Embryogenesis from Seedling Tissues
Seedlings grown in vitro represent a particularly convenient source of explants for

somatic embryo induction (Yie and Liaw 1977; Mehdi and Hogan 1979;
Yamamoto et al. 1986; Chen et al. 1987; Chen 1988a,b; Yamamoto and Tabata
1989), although greenhouse-grown seedlings have also been cultured following
decontamination (Jordan et al. 1982). Seeds were germinated aseptically on
media, sterile wet filter papers, or water agar. Explants were cultured on 112 MS,
MS, or LS medium containing either NAA or 2,4-D. Somatic embryos developed
directly from calli or after the embryogenic calli were transferred to media
containing lower concentrations of growth regulators (Table 1).
1.3.3 Somatic Embryogenesis from Mature Tissues
Tissues from mature papayas and its relatives have yielded embryogenic calli as
well, although fewer reports exist (De Bruijne et al. 1974; Mehdi and Hogan 1979;
266 M.M.M. Fitch
Litz and Conover 1980). Explants from petioles, peduncles, stems, and shoots
were cultured on MS medium with NAA, IBA, IAA, 2iP, kinetin, BA, and/or
CWo De Bruijne et al. (1974) subcultured calli to more dilute W or MS medium
containing lower concentrations of growth regulators to stimulate embryo
development, while Litz and Conover (1980) added charcoal to the MS medium
(Table 1). As with the seedling-derived cultures, somatic embryos from mature
explants developed from an intermediate callus phase in all three studies.
1.3.4 Media
Table 1 shows the range of nutrient media that was used for various types of
papaya cultures along with the results in the different reports.
1.3.5 Regeneration of Plants: Somatic Embryo Maturation

and Germination
In most cases, somatic embryos of papaya, its hybrids, and its relatives germi-
nated readily in the absence of auxins or on media containing relatively low
concentrations of auxins and cytokinins (Table 1). De Bruijne et al. (1974) were
not able to culture embryos beyond the mature cotyledonary stage and may have
needed to treat embryos with maturation-inducing compounds like cytokinins or
ABA (Ammirato 1983).
Media devoid of growth regulators were widely used for germination (Litz
and Conover 1980, 1983; Litz 1986b; Manshardt and Wenslaff 1989a). In other
reports, one or more growth regulators were supplemented in the media to
enhance germination (Yie and Liaw 1977; Litz and Conover 1982; Jordan et al.
1982; Chen et al. 1987; Chen 1988a,b; Table 1). Normal embryo development was
reported in many cases, although higher concentrations of BA caused excessive
swelling of hypocotyls and inhibited shoot and leaf development.
1.3.6 Papaya Micropropagation
Papaya micropropagation is well established (Mehdi and Hogan 1976, 1979; Yie
and Liaw 1977; Litz and Conover 1978a,b; Pandey and Rajeevan 1983; Litz
1986a; Rajeevan and Pandey 1986; De Winnaar 1988; Drew 1988; Reuveni et al.
1990; Drew et al. 1991). If plants regenerated from selected somatic embryos
require mass propagation, micropropagation methods are useful to enhance
culture establishment. Cytokinins have been used for the proliferation of shoots
(Table 1). A generally useful medium is MS supplemented with relatively low
concentrations of cytokinins, for example, 0.2mg/1 kinetin or BA.
1.3.7 Rooting, Acclimatization, and Transfer of Papayas to Greenhouse

and Field Conditions
Papaya cuttings were most efficiently rooted in MS medium containing 1 to 4 mg/l

IBA (Litz and Conover 1978a, 1981a; Rajeevan and Pandey 1986; De Winnaar
1988; Miller and Drew 1990; Reuveni et al. 1990) or 5 mg/l IAA (Yie and Liaw
1977) if cuttings were 1 to 2 em tall (Miller and Drew 1990). Drewet al. (1991)
enhanced rooting by the addition of 0.4 mg/l riboflavin to culture medium
containing 2 mg/l IBA. Apparently, riboflavin enhanced the photodegradation of
IBA which was important for further root development. In most cases, once roots
were established and plants acclimated, plants were easily grown in the green-
house and field.
2 Somatic Embryogenesis in Papaya

(Fitch and Manshardt 1990; Fitch 1991, 1993)
2.1 Zygotic Embryos
Using the protocols of Litz and Conover (1981b, 1982, 1983) and Manshardt and
Wenslaff (1989a,b) as guidelines, Fitch and Manshardt (1990) produced somatic
embryo cultures for transformation studies. Green, self- or open-pollinated
fruit, about 24 to 105 days old, were rinsed with water and 95% ethanol and
soaked for 1 h in 1.05% sodium hypochlorite containing two drops of Tween 20
per liter of solution. Immature ovules between 24 and 105 days old and zygotic
embryos between 75 and 105 days old (Fig. lA) were cultured on 112 MS medium
containing 6% sucrose, 400 mg/l glutamine, 20% CW, with or without 1 mg/l
2,4-D or 0.5 mg/l picloram. The ovule cultures did not give rise to embryogenic
cultures except for one Sunset culture in which several 36-day-old ovules
produced somatic embryos after 11 months in culture. Quicker in responding
were excised embryos on CW- or auxin-containing medium. One 87-day-old
Sunrise zygotic embryo on medium containing CW produced multiple somatic
embryos at the transition zone between the radicle and hypocotyl after 10 days in
culture. In 2,4-D- and picloram-containing media, somatic embryos budded from
tissues at the shoot apices and radicles of zygotic embryos after about 6 weeks in
culture.
A range of 2,4-D concentrations was tested for determining optimum levels
to induce embryogenesis in four Hawaiian papaya cultivars (Fitch and
Manshardt 1990). Embryos were excised from 90- to 120-day-old ovules of
Kapoho, Sunrise, Sunset, and Waimanalo and plated on 100 xiS mm Petri dishes
containing 112 MS medium supplemented with 6% sucrose, 400 mg/l glutamine,
MS vitamins, 0 to 25 mg/12,4-D, 0.5% agar, pH 5.7 (induction medium) with or
without 20% CWo Ten zygotic embryos were plated in each dish. The cultures
were incubated at 27°C in the dark and monitored for 6 weeks. Somatic embryos
on shoot apices were observed as early as 3 weeks after culture initiation (Fig. IB).
Fig. lA-D. Papaya regeneration
via 2,4-D-induced somatic em-
bryogenesis in immature zygotic
embryos. 112 MS induction medi-
um contained 20% CW and 5 mgt
12,4-D. Bar = I mm. A Immature
Waimanalo zygotic embryo ex-
cised from 75-day-old ovule. B
Embryogenic apex of Sunset
zygotic embryo cultured on in-
duction medium for 17 days. C
Zygotic embryo of Sunrise
cultured for 3 months, showing
about 20 somatic embryos in the
apical region between the two
cotyledons. D Large number of
somatic embryos produced from
one Sunset zygotic embryo after 5
months on induction medium
o Kapoho
III
0
>. 100 • Sunrise
Waimanalo
.a"" Sunset
El
Q)
80
.....
()
+I
0
Ill)
>. 60
N
.....C)
r:1 40
Q)
Ill)
0
>.
20
.a""
aQ)
0
----~
~
o 5 10 15 20 25 30
-1
2.4-D concentration ( mg 1 )
Fig.2. Percentage of zygotic embryos that developed embryogenic growth in four Hawaiian papaya
cultivars after 5 to 6 weeks of culture on induction medium containing 20% coconut water and 0 to
25 mg/I 2,4-D. Error bars represent ± I SE
Up to 20 somatic embryos were observed after 6 weeks in culture (Fig. 1C).

Continued culture on induction medium resulted in large numbers of somatic
embryos, for example, 400 per zygotic embryo after 4 months of culture (Fig. 1D).
CuItivar differences were observed in the numbers of zygotic embryos that
became embryogenic over time (Fig. 2). A 2,4-D concentration range of 1 to 10
mgll was suitable for somatic embryo induction in the four cultivars.
Brown, wet calli were produced on zygotic embryos at higher 2,4-
D concentrations, and embryogenesis appeared to be inhibited (Fig. 2).
Germinated embryos were micropropagated in MPH liquid medium (Lee
1987) rather than potted directly into vermiculite or soil because the hypocotyls
were often too long and slender to support strong plant growth. Multiple shoots
developed in MPH and were cut when stems were 1 cm long. The shoots were
rooted on MS medium containing 1 mg/l IBA in 7 to 30 days and were potted in
50% vermiculite/50% 1/2 MS (v/v) that had been autoclaved. When the plants
were well rooted, they were transplanted to moistened, autoclaved commercial
potting soil and bagged under clear plastic. Holes were cut in the bags after 1 day
and were gradually enlarged over a week to acclimate the plants to greenhouse
conditions (Fig. 3). The plants were grown in the greenhouse until they had four
or five leaves and were transferred to the field.
270 M.M.M. Fitch
Fig.3. Papaya regeneration from somatic embryos. Bar = I cm. Sunset and Kapoho
2.2 Hypocotyl Sections
Highly embryogenic calli developed from sections of papaya hypocotyls cultured

on media containing 2,4-D (Fitch 1993). Seeds of Kapoho, Sunrise, Sunset, and
Waimanalo were surface sterilized, in batches of 50, by suspending them in 50 ml
of 1.05% sodium hypochlorite for 1 h. Seeds with dark, unbleached spots or holes
were discarded. The seeds were rinsed in sterile distilled water and shaken for
24 h in 50 ml of 1 M KN0 3 (Nagao and Furutani 1986). Floating seeds were
discarded, and the remaining seeds were suspended in 100 ml of sterile water at
32.2 °C until the testae cracked. The germinating seeds were sown on 1% water
agar and grown under cool white fluorescent lights (photosynthetically active
radiation, PAR = 35 J.lmollm2/s.)
In about 10 days, seedlings with expanded cotyledons and one 2- to 4-mm-
long trilobed leaf (Fig. 4A) were explanted. The 3- to 12-cm-long hypocotyls were
sectioned into 2- to 3-mm lengths, and the sections were plated on 112 MS
medium containing 6% sucrose, 400 mg/I glutamine, MS vitamins, and 0 to 25
mg/l 2,4-D (Fig 4B). The cultures were incubated in the dark at 27 °C for 6 to 8
weeks, after which time hypocotyl calli were observed to be embryogenic (Fig. 4C,
D). The Hawaiian cultivar Kapoho was the most reliable in producing embryo-
genic calli, but Sunrise, Sunset, and Waimanalo also became embryogenic on some
2,4-D-containing media (Table 2). Kapoho hypocotyls on induction medium
containing 10 mg/I 2,4-D consistently produced high frequencies of embryogenic
calli. A sucrose concentration of 7% was determined to be optimal (Fig. 5).
Varying both the 2,4-D and sucrose concentrations in the medium may increase
the frequency of embryogenesis in cultivars other than Kapoho. Addition of the
growth regulator abscisic acid (ABA) may also increase the frequency of
embryogenesis (Ammirato 1983). Removal of 2,4-D by transfer of calli to MS
medium enhanced embryo maturation and development. Somatic embryos from
Fig. 4 A-D. Papaya regeneration via
2,4-D-induced somatic embryogenesis
in hypocotyl-derived callus. Bar = I em.
A IO-day-old papaya seedlings germi-
nated on 1% water agar. Hypocotyls
and other tissues were explanted from
seedlings at this stage of development.
B Freshly explanted hypocotyl sections,
2 mm long. C Hypocotyl sections on
induction medium containing 10 mg/I
2,4-D after 2 months in culture. All
sections had developed into highly
embryogenic calli. DSomatic embryos
of Kapoho that developed directly on
induction medium containing 10 mg/I
2,4-D, two months after culture
initiation. Bar = I mm
272 M.M.M. Fitch
Table 2. Embryogenic callus production by seedling tissues from four Hawaiian papaya
cultivars grown on induction medium containing various concentrations of2,4-D. (Fitch
1991)
CV %S 2,4-D Hypocotyls Cotyledons Shoots
K 3 0 0 nd nd
K 3 0.5 0 nd nd
K 3 1 0 nd +
K 3 10 +++ nd +
K 6 0 0 0 0
K 6 0.5 +++ 0 0
K 6 1 +++ ++ +
K 6 10 +++ +++ ++
SR 3 0 0 0 0
SR 3 0.5 0 0 0
SR 3 1 nd nd nd
SR 3 10 0 0 ++
SR 6 0 0 0 0
SR 6 0.5 +++ + +
SR 6 1 +++ 0 nd
SR 6 10 +++ 0 0
SS 3 0 0 0 0
SS 3 0.5 0 0 ++
SS 3 1 0 0 0
SS 3 10 nd nd nd
SS 6 0 0 0 0
SS 6 0.5 +++ ++ 0
SS 6 1 +++ 0 +
SS 6 10 +++ 0 nd
W 3 0 0 0 0
W 3 0.5 +++ ++ 0
W 3 1 nd ++ nd
W 3 10 0 nd 0
W 6 0 0 0 0
W 6 0.5 0 0 nd
W 6 1 nd 0 nd
W 6 10 0 nd nd
CV =cultivar; %S =sucrose concentration; 2,4-D = concentration in mg/I; K =Kapoho;

SR =Sunrise; SS =Sunset; W =Waimanalo; nd =no data.
+ =>0 to <20% ofthe explants became embryogenic; ++ =>20% to <40% of the explants
became embryogenic; +++ =>40% of the explants became embryogenic.
hypocotyl-derived calli were germinated, propagated, and rooted as described for

somatic embryos from zygotic embryos.
2.3 Embryogenic Calli from Other Seedling Explants
Cotyledons, roots, and shoots from lO-day-old seedlings grown in 1% water agar
were also cultured and sometimes became embryogenic. The cultivar Kapoho
80
III
.......
~
0
60
0
CD
III
...
0
~
CD 40
tID
0
.c'"""'
ElCD
20
N
o
2 3 4 5 6 7 8 9 10
sucrose concentration, "

Fig.5. Effect ofsucrose concentration on induction of somatic embryogenesis in Kapoho hypocotyl
sections after 2 months of culture on induction media containing 10 mg/12,4-D. Data are means of
three replicates. Error bars represent 1 SE. (Fitch 1993)
produced the largest number of embryogenic cultures but mainly because larger
numbers of this cultivar were cultured (Table 2).
2.4 Cotyledons
In early studies, a small number of cotyledon slices cultured on 112 MS induction

medium containing 0.002 to 0.02 mg/l picloram and 0.2 mg/l BA produced a very
low frequency of somatic embryos. Highly embryogenic calli were initiated with
the substitution of 2,4-D in 112 MS induction medium. Cotyledons, whether
sliced into 2-mm-wide sections or whole, sometimes became embryogenic in 6 to
8 weeks, although the frequency of induction was not as high as observed with
hypocotyl sections (Table 2). The data is limited, but it is clear that cotyledons as
well as hypocotyls of Kapoho seedlings are capable of forming embryogenic
callus within a wide range of 2,4-D concentrations. Sunset, Sunrise, and
Waimanalo cotyledons were less responsive (Table 2).
274 M.M.M. Fitch
2.5 Seedling Shoots
The culture media suitable for hypocotyls and cotyledons were also suitable for
inducing embryogenic calli from seedling shoots. Kapoho shoots produced
embryogenic calli in media containing 1 or 10 mg/l 2,4-0 (Table 2), reflecting
again the tendency of this cultivar toward embryogenesis. Embryogenic calli
from Sunrise and Sunset developed on 2,4-0-containing media, but no pattern of
response was obvious. Although only a small number of shoots were plated on
different media, the data suggest that most of the tissues of 10-day-old Kapoho,
Sunrise, and Sunset seedlings can be induced to become embryogenic on
induction media containing 2,4-0.
2.6 Root Tips
Some of the root masses removed from lO-day-old Kapoho and Sunset seedlings
became embryogenic. The intact seedling root mass was cultured, in contrast to
Chen et al. (1987), who cultured l-cm-Iong root sections with root tips removed.
The root masses were placed on 112 MS induction medium containing 1 mg/l
2,4-D with 0 or 0.005 to O.oI mg/l thidiazuron. After a month, a small number of
root tips became embryogenic on media containing I mg/l 2,4-0 with or without
thidiazuron. Embryogenic calli were produced only on root apices. The root
masses developed into brown, wet-looking calli, and shiny, light green, fan-like
structures developed at the tips of roots in cultures containing thidiazuron. Upon
removal of the tissues to maturation medium, the structures differentiated into
shoots. Somatic embryos produced on induction media containing 2,4-0 alone
resembled those initiated from hypocotyl sections and from immature zygotic
embryos (Figs. 10 and 5). Various other 112 MS media containing picloram,
NAA, 2iP, and/or thidiazuron did not induce embryogenesis in roots.
The results from seedling tissues show that cotyledons, shoot apices, and root
masses from lO-day-old seedlings can develop embryogenic calli on the same
media that induced embryogenesis in hypocotyl sections and zygotic embryos.
Hypocotyls were, by far, the most efficient producers of embryogenic calli among
the seedling tissues. However, the occurrence of somatic embryogenesis on
papaya zygotic embryos was even more consistent. More than 40% of the zygotic
embryos of all four Hawaiian cultivars, Kapoho, Sunrise, Sunset, and
Waimanalo, became embryogenic on induction medium containing 5 mg/12,4-D
in one experiment, whereas hypocotyls of the four Hawaiian cultivars showed
different medium requirements for optimal embryogenic response.
3 Conclusions
In our work, somatic embryos of papaya have been readily produced from
immature embryo and seedling tissues cultured on media containing 2,4-0.
Somatic embryos were induced most rapidly on the apical domes and radicles of
immature zygotic embryos that were about 90 to 120 days old when these tissues
are plated on 112 MS medium containing 2,4-D. Somatic embryos were observed
as quickly as 3 weeks after culture initiation. The short time interval in which the
papaya tissues were exposed to 2,4-D prior to observation of embryogenic tissues
may have important consequences for later use of these cultures for propagation
or for genetic transformation. Aberrant genotypes due to somaclonal variation
may result from tissue cultures, particularly those exposed to 2,4-D for extended
periods of time (Larkin and Scowcroft 1981). Somatic embryos were readily
induced, in 6 weeks, on embryogenic calli from hypocotyl sections of lO-day-old
seedlings cultured on 112 MS induction medium containing 2,4-D. Other explants
from lO-day-old seedlings produced embryogenic calli after 6 weeks in culture on
induction medium but at lower frequencies.
In these experiments, embryogenic papaya calli from zygotic embryos and
seedling tissues cultured on media containing 2,4-D were produced in shorter
time compared to earlier reports for papaya, its hybrids, and other Carica species.
Cultured root sections from papaya seedlings of unspecified age produced
embryogenic calli on 112 MS medium containing NAA and BA but after culture
period of 3 months (Chen et al. 1987; Chen 1988a). Papaya seedling and mature
tissue sections gave rise to somatic embryos in multistep protocols in other
reports in which neither frequency nor the time required were reported (De
Bruijne et al. 1974; Yie and Liaw 1977; Mehdi and Hogan 1979; Yamamoto and
Tabata 1989). Therefore, papaya tissues most amenable to somatic embryo
formation are immature zygotic embryos and young seedling tissues cultured on
112 MS media containing either 0.5 to 10 mg/12,4-D or 1 mg/I NAA and 0.5 mg/I
kinetin. A range in sucrose concentrations from 3 to 7% in simple factorials
with 2,4-D concentrations can be used to optimize embryogenesis for specific
cultivars.
The usefulness of embryogenic cultures for genetic transformation is now
well known. Some papayas transformed with vectors containing genes for trans-
genic cell selection (neomycin phospho transferase II), detection (b-glucuroni-
dase), and the coat protein gene of papaya ringspot virus (Fitch et al. 1990) have
been characterized as having intermediate or complete resistance to papaya
ringspot virus (Fitch et al. 1992). Plants like these should help increase the scale of
papaya production worldwide, allowing growers to utilize areas now abandoned
due to virus infestation.
4 Protocols
Two papaya tissues, immature zygotic embryos and hypocotyl sections from I O-day-old, aseptically
grown seedlings, are well suited for the induction of somatic embryos. Zygotic embryos are the most
reliable, because all four Hawaiian genotypes have become embryogenic under similar conditions.
Papaya seeds are probably more readily available than immature fruit; thus, the hypocotyl protocol
may be more useful.
276 M.M.M. Fitch
4.1 Immature Zygotic Embryos
1. The Hawaiian papaya cultivars Kapoho, Sunrise, Sunset, and Waimanalo have been successfully
cultured with this protocol. Rinse green fruits, 90 to 120 days old, with water and 95% ethanol.
Soak fruits for 1 h in 1.05% sodium hypochlorite containing one drop of Tween 20 or other
surfactant per liter of solution. Using an aseptic technique, bisect the fruit, remove the ovules, and
excise the immature zygotic embryos.
2. Plate the zygotic embryos on 100 x 15-mm Petri dishes containing 112 MS medium (Murashige
and Skoog 1962) supplemented with 6% sucrose, 400 mg/I glutamine, MS vitamins, 10 mg/I
2,4-D, and 0.8% Difco Bacto-agar, pH 5.7 (induction medium). Ten to 30 zygotic embryos can be
plated in each dish.
3. Incubate the cultures at 27°C in the dark for 4 weeks, after which the apical domes and radic1es
will show evidence of2 to 20 somatic embryos. Continued culture on this medium results in large
numbers of somatic embryos, for example, 400 per zygotic embryo after 4 months of culture.
4. Allow the somatic embryos to mature and germinate them on MS medium containing 3% sucrose
and 0.8% Difco Bacto-agar.
5. Micropropagate the germinated embryos by growing them in MPH liquid medium (Lee 1987).
6. Remove 1- to 2-cm-tall shoots from the micropropagated masses and root them on MS agar
medium containing 3% sucrose, 1 mg/I IBA, and 0.8% Difco Bacto-agar.
7. Pot rooted plants in autoc1aved vermiculite/liquid 1/2 MS medium (50/50, by volume) for 1 to 2
months to increase the size of the plant and root system. Transfer the plants to moistened,
autoc1aved commercial potting soil or vermiculite, and cover with a plastic bag. Enlarge holes in
the bags gradually over a week to acclimatize the plants to greenhouse conditions.
4.2 Hypocotyl Sections
1. Soak mature, dry papaya seeds in 1.05% sodium hypochlorite (20% Clorox or other commercial
bleaching agent) for I h with agitation, 100 rpm, 27°C. Remove seeds that are not entirely
bleached or ones that are broken.
2. Agitate overnight in sterile 1 M KNO J • Transfer submerged seeds to sterile water, agitate at 32.2
°C for 3 to 5 days or until the testae crack.
3. Plant seeds with cracked testae on 1% water agar and grow under lighted conditions, about
35 /-Im01lm'/s of photosynthetically active radiation (PAR).
4. Explant hypocotyl sections from seedlings that are about 10 days old or that have unfolded
cotyledons and one true leaf, 2 to 4 mm long.
5. Slice hypocotyls into 2-mm sections, and plate sections on 112 MS medium containing 400 mg/I
glutamine, 6% sucrose, MS vitamins, 0.8% Difco Bacto-agar, and 1 to 10 mg/12,4-D.
6. Incubate the cultures in the dark at 27°C for 6 weeks, after which some embryogenic calli may
be observed. With increased time in culture, more hypocotyl sections become embryogenic.
The Hawaiian cultivar Kapoho is the most reliable in producing embryogenic calli, but Sunrise,
Sunset, and Waimanalo can become embryogenic on some 2,4-D media. It may be necessary to
vary both the 2,4-D and sucrose concentrations in the medium to induce embryogenesis in
cultivars other than Kapoho. Addition of the growth regulator ABA may be beneficial
(Ammirato 1983). Germinate and grow the somatic embryos as described in points 4-7 (Sect. 4.1)
for greenhouse culture.
References
Alvarez AM, Nelson MG (1982) Control of Phytophthora palmivora in papaya orchards with weekly
sprays of chlorothalonil. Plant Dis 66: 37-39
Ammirato PV (1983) Embryogenesis. In: Evans DA, Sharp WR, Ammirato PV , Yamada Y (eds)
Handbook of plant cell culture, vol I. Techniques for propagation and breeding. Macmillan,
New York, pp 82-123
Arora IK, Singh RN (1978) In vitro plant regeneration in papaya. Curr Sci 47: 867-868
Arriola M, Calzada J, Menchu J, Rolz C, Garcia R, Cabrera S (1980) Papaya. In: Nagy S, Shaw P
(eds) Tropical and subtropical fruits. A VI, Westport, pp 316-340
Badillo VM (1967) Esquema de las Caricaceae. Agron Trop 17: 245-272
Badillo VM (1971) Monografia de la Familia Caricaceae. Asociacion de Profesores. Maracay,
Venezuela
Beachy RN, Loesch-Fries S, Turner NE (1990) Coat protein-mediated resistance against virus
infection. Annu Rev Phytopathol28: 451-474
Chan HT, Cavaletto CG (1982) Aseptically packaged papaya and guava puree: changes in chemical
and sensory quality during processing and storage. J Food Sci 47: 1164-1169,1174
Chen FC, Kuo MH (1988) Ovule and ovary culture in papaya and somatic embryogenesis. In: Ma
S-S, Shii C-T, Chang W-C, Chang T-L (eds) Proc Symp Tissue culture of horticultural crops,
National Taiwan University, Taipei, March 8-9,1988. Department of Horticulture, National
Taiwan University, 1991, pp 50-61 (In Chinese)
Chen MH (1988a) Tissue culture and ringspot resistant breeding of papaya. In: Ma S-S, Shii C-T,
Chang W -C, Chang T -L (eds) Proc Symp Tissue culture of horticultural crops, National Taiwan
University, Taipei, March 8-9, 1988. Department of Horticulture, National Taiwan University,
1991, pp 62-72 (In Chinese)
Chen MH (1988b) Somatic embryogenesis and plant regeneration in Carica papaya L. tissue culture
derived from root explants. In: Ma S-S, Shii C-T, Chang W -C, Chang T -L (eds) Proc Symp Tissue
culture of horticultural crops, National Taiwan University, Taipei, March 8-9, 1988.
Department of Horticulture, National Taiwan University, 1991, pp 230-233
Chen MH, Wang PJ, Maeda E (1987) Somatic embryogenesis and plant regeneration in Carica
papaya L. tissue cultures derived from root explants. Plant Cell Rep 6: 348-351
Chen MH, Chen CC, Wang DN, Chen FC (1991) Somatic embryogenesis and plant regeneration
from immature embryos of Carica papaya x Carica cauliflora cultured in vitro. Can J Bot 69:
1913-1918
De Bruijne E, De Langhe E, van Rijck R (1974) Actions of hormones and embryoid formation in
callus cultures of Carica papaya. In: Int Symp Crop protection, Fytopharmacie en Fytiatrie.
Rijkslandsbouwhoogeschool Medelelingen 39: 637-645
De Fossard RA, MyintA, Lee ECM (1974) A broad spectrum tissue culture experiment with tobacco
(Nicotiana tabacum) pith tissue callus. Physiol Plant 30: 125-130
De Winnaar W (1988) Clonal propagation of papaya in vitro. Plant Cell Tissue Organ Cult 12:
305-310
Drew RA (1988) Rapid clonal propagation of papaya in vitro from mature field-grown trees.
HortScience23: 609-611
Drew RA, Smith NG (1986) Growth of apical and lateral buds of papaw (Carica papaya L.) as
affected by nutritional and hormonal factors. J Hortic Sci 61: 535-543
Drew RA, Simpson BW, Osborne WJ (1991) Degradation of exogenous indole -3-butyric acid and
riboflavin and their influence on rooting response of papaya in vitro. Plant Cell Tissue Organ Cult
26:29-34
F AO (1990) Production yearbook, vol 44. F AO Statistics Series #94. F AO, Rome
Fitch MMM (1991) Development of genetic transformation systems in papaya. PhD Thesis,
University of Hawaii at Manoa, Honolulu, Hawaii, August, 1991
Fitch MMM (1993) High frequency somatic embryogenesis and plant regeneration from papaya
hypocotyl callus. Plant Cell Tissue Organ Cult 32: 205-212
Fitch MMM, Manshardt RM (1990) Somatic embryogenesis and plant regeneration from immature
zygotic embryos of papaya (Carica papaya L.). Plant Cell Rep 9: 320-324
Fitch MMM, Manshardt RM, Gonsalves D, Slightom JL, Sanford JC (1990) Stable transformation
of papaya via microprojectile bombardment. Plant Cell Rep 9: 189-194
Fitch MMM, Manshardt RM, Gonsalves D, Slightom JL, Sanford JC (1991) Characterization of
transgenic papaya plants. In: Hallick RB (ed) 3rd Int Congr Plant molecular biology. Molecular
biology of plant growth and development. Department ofBiochemistry, University of Arizona,
6-11 October 1991, Tucson, p607
278 M.M.M. Fitch
Fitch MMM, Manshardt RM, Gonsalves D, SlightomJL, SanfordJC (1992) Virus resistant papaya
plants derived from tissues bombarded with the coat protein gene of papaya ringspot virus.
Bioffechnology 10: 1466-1472
Gamborg 0 L, Miller RA, Ojima K (1968) Nutrient requirements of suspension cultures of soybean
root cells. ExpCell Res 50: 148-151
Horovitz S, Jimenez H (1958) Cruzibilidad entre species de Carica. Agron Trop (Maracay) 7:
207-215
Horovitz S, Jimenez H (1967) Cruzamientos interspecificos e intergenericos en caricaceas y sus
implicaciones fitotecnicas. Agron Trop (Maracay) 17: 323-343
Jordan M, Cortes I, Montenegro G (1982) Regeneration ofplantlets by embryogenesis from callus
cultures of Carica candamarcensis. Plant Sci Lett 28: 321-326
Larkin PJ, Scowcroft JM (1981) Somaclonal variation - a novel source of variability from cell
cultures for plant improvement. Theor Appl Genet 60: 197-214
Lee TSG (1987) Micropropagation of sugarcane (Saccharum spp.). Plant Cell Tissue Organ Cult 10:
47-55
Linsmaier EM, Skoog F (1965) Organic growth factor requirements of tobacco tissue culture.
Physiol Plant 18: 100-127
Litz RE(1984) Papaya. In: Sharp W, EvansD,Ammirato P, Yamada Y (eds) Handbook of plant cell
culture, vol2. Crop species. Macmillan, London, pp 349-368
Litz RE (1986a) Papaya (Carica papaya L.). In: Bajaj YPS (ed) Biotechnology in agriculture and
forestry, vol I. Trees I. Springer, Berlin Heidelberg New York, pp 220-232
Litz RE (1986b) Effect of osmotic stress on somatic embryogenesis in Carica suspension cultures. J
Am Soc Hortic Sci 11: 969-972
Litz RE, Conover RA (1977) Tissue culture propagation of papaya. Proc Fla State Hortic Soc 90:
245-246
Litz RE, Conover RA (1978a) In vitro propagation of papaya. HortScience 13: 241-242
Litz RE, Conover RA (1978b) Recent advances in papaya tissue culture. Proc Fla State Hortic Soc
91:180-182
Litz RE, Conover RA (1980) Somatic embryogenesis in cell cultures of Carica stipulata. HortScience
15:733-735
Litz RE, Conover RA (1981 a) Effect of sex type, season, and other factors on in vitro establishment
and culture of Carica papaya L. explants. J Am Soc Hortic Sci 106: 792-794
Litz RE, Conover RA (1981 b) In vitro polyembryony in Carica papaya L. ovules. Z Pflanzenphysiol
104:285-288
Litz RE, Conover RA (1982) In vitro somatic embryogenesis and plant regeneration from Carica
papaya L. ovular callus. Plant Sci Lett 26: 153-158
Litz RE, Conover RA (1983) High-frequency somatic embryogenesis from Carica suspension
cultures. Ann Bot 51: 683-686
Litz RE, O'Hair SK, Conover RA (1983) In vitro growth of Carica papaya L. cotyledons. Sci Hortic
19:287-293
Manshardt RM, Wenslaff TF (1989a) Zygotic polyembryony in interspecific hybrids of Carica
papaya and C. caulifiora. J Am Soc Hortic Sci 114: 684-689
Manshardt RM, Wenslaff TF (1989b) Interspecific hybridization of papaya with other Carica
species. J Am Soc Hortic Sci 114: 689-694
Mehdi AA, Hogan L (1976) Tissue culture of Carica papaya. HortScience (Abstr) 11: 311
Mehdi AA, Hogan L (1979) In vitro growth and development of papaya (Carica papaya L.) and date
palm (Phoenix dactylifera L.). HortScience (Abstr) 14: 46
Miller RM, Drew RA (1990) Effect of explant type on proliferation of Carica papaya L. in vitro.
Moore GA, Litz RE (1984) Biochemical markers of Carica papaya. C. caulifiora, and plants from
somatic embryos of their hybrid. J Am Soc Hortic Sci 109: 213-218
Murashige T, Tucker DPH (1969) Growth factor requirements of citrus tissue culture. Proc 1st Int
CitrusSymp3: 1155
Nagao MA, Furutani SC (1986) Improving germination of papaya seed by density separation,
potassium nitrate, and gibberellic acid. HortScience21: 1439-1440
Namba R, Kawanishi C (1966) Transmission of papaya mosaic virus by the green peach aphid. J
Econ Entomol59: 669-671
Nitsch JP, Nitsch C (1969) Haploid plants from pollen grains. Science 163: 85-87
Pandey RM, Rajeevan MS (1983) Callus initiation and regeneration in tissue culture of papaya. In
Sen SK, Giles KL (eds) Plant cell culture in crop improvement. Plenum Press, New York, pp
427-430
Pang SZ, Sanford JC (1988) Agrobacterium-mediated gene transfer in papaya. J Am Soc Hortic Sci
113: 287-291
Poulter NH, Caygill JC (1985) Production and utilization of papain - a proteolytic enzyme from
Carica papaya L. Trop Sci 25: 123-137
Provvidenti R, Gonsalves D (1982) Resistance to papaya ringspot virus in Cucumis metuliferus and
its relationship to resistance to watermelon mosaic virus I. J Hered 73: 239-240
Provvidenti R, Robinson RW (1977) Inheritance of resistance to watermelon mosaic virus I in
Cucumis metuliferus. J Hered 68: 56--57
Purcifull DE, Edwardson J, Hiebert E, Gonsalves D (1985) Papaya ringspot virus (revised). No 84
In: Descriptions of plant viruses. Commonw Mycol Inst!Assoc Appl Bioi Kew, Surrey
Purse gloveJW (1968) Papayas. In: Tropical crops, vol I. Dicotyledons. John Wiley, New York
Rajeevan MS, Pandey RM (1986) Lateral bud culture of papaya (Carica papaya L.) for clonal
propagation. Plant Cell Tissue Organ Cult 6: 181-188
Reuveni 0, Shlesinger DR, Lavi U (1990) In vitro clonal propagation of dioecious Carica papaya.
Thomas E, Beckley VA (1923) Papaw and papain. J Dep Agric S Afr 6: 357-360
Vega de Rojas R, Kitto SL (1991) Regeneration ofbabaco (Carica pentagona) from ovular callus.
JAm Soc Hortic Sci 116: 747-752
White PR (1963) A handbook of plant and animal tissue culture. Jaques Cattel, Lancaster,
Pennsylvania
Yamamoto H, Tabata M (1989) Correlation of papain-like enzyme production with laticifer
formation in somatic embryos of papaya. Plant Cell Rep 8: 251-254
Yamamoto H, Tanaka S, Fukui H, Tabata M (1986) Enzymatic difference between laticifers and
cultured cells of papaya. Plant Cell Rep 5: 269-272
Yie S, Liaw SI (1977) Plant regeneration from shoot tips and callus of papaya. In Vitro 13: 564--568
111.4 Somatic Embryogenesis in Citrus Species
H. KUNITAKE l,2 and M, MIll
1 Introduction
Citrus fruits are one of the most important fruit crops in the world, beyond
grapes in total production, World Citrus production has been continuously
increasing and it exceeded 78 million tons by 1991, whereas it was only about 22
million tons in 1960, Citrus fruits are widely produced throughout the world
both in tropical and subtropical countries with latitudes between 40 0 Nand
40 0 S. At present, there are approximately 90 Citrus-producing countries, such
as Brazil, USA, Mexico, Spain, Italy, China, Japan etc, (FAO 1992; Fortucci-
Marongiu 1988),
The genus Citrus and its relatives are members of the family Rutaceae,
subfamily Aurantioideae, Economically important Citrus crops include sweet,
orange, grapefruit, mandarin, lemon, lime, citron, pummelo, and their hybrids.
In the related genera, fruits of Fortunella have a limited market and Poncirus is
used as rootstock.
Somatic embryogenesis has a great potential for rapid clonal propagation of
horticultural crops, and embryogenic cell culture is a promising material for
protoplast culture, somatic hybridization, and genetic transformation. However,
somatic embryos are not a prerequisite for clonal propagation of Citrus cultivars,
because all Citrus cultivars could be easily propagated by grafting on rootstock
plants such as trifoliate orange (Poncirus trifoliata Raf.), sour orange (Citrus
aurantium L.), and cit range (C sinensis x P. trifoliata). Important applications
of somatic embryogenesis to the Citrus industry can be classified into two
categories: (1) application of embryogenic cells for protoplast culture, somatic
hybridization, and genetic transformation, and (2) application of somatic
embryos and embryogenic callus for in vitro preservation of germplasm.
Embryogenic callus culture in Citrus was first produced from the nucellar
tissue ofShamouti sweet orange (Citrus sinensis Osb.) by immature ovule culture
1 Laboratory of Plant Cell Technology, Faculty of Horticulture, Chiba University, 648 Matsudo,
Chiba 271, Japan
2 Present address: Laboratory of Plant Biotechnology, Saga Prefectural Agricultural Research Center,
1088 Nanri, Kawasoe-cho, Saga-gun, Saga 840-23, Japan

Somatic Embryogenesis and Synthetic Seed I (ed, by y.P,S, Bajaj)
Table 1. Published conditions for somatic embryogenesis and plant regeneration of Citrus and its relatives
Species Cultivars Explants Basal Growth regulators and other elements Plant regeneration Other culture References
medium conditions
Callus induction Embryoid induction
C. sinensis Shamouti Ovule MP 500 mg/I malt extract I mg/IGA] 16 h daylight Kochba et al.
(oranges) 1000 Ix, 26 ± I °C (\972)
Trovita Ovule MT 10 mg/I BA 5% galactose I mg/IGA] 16 h daylight Kobayashi et al.
or lactose 2000 Ix, 25°C (\ 984)
Trovita Anther MS 0.02 mg/IIAA + Growth regulator-free Dark, 28°C (only Hidaka (\984)
0.02 mg/l kinetin embryoid induction)
Hamlin Undeveloped MS 500 mg/I malt extract I mg/IGA] 16 h daylight Gmitter and
ovule or 0.01 mg/I 2,4-D + 76 J.lmol m 's 1 Moore (1986)
0.1 mg!1 daminozide 27°C
Washington Embryo MS 0.5 g/I malt extract I J.lM zeatin or IJ.lMGA] 16 h daylight Hidaka and
navel + 50 J.lM kinetin growth regulator-free 4.7-8.6 Wm-2 Kajiura (1988)
25 ± I °C
Ridge Endosperm MT 5 mg/I BA+ (2MT) 500 mg/I CH + (2MT) Dark, 24-28 °C Gmitter et al.
pineapple 2 mg/12,4-D 2mg/IGA, + 5mg!1 GA, (only callus (1990)
1000 mg/I CH b + 0.25 mg/I BA induction)
5 mg/I kinetin
500 mg/I malt extract 2 mg!1 adenine
C.limon Femminello Undeveloped MS 500 mgll malt extract 20 mg!1 16 h daylight Starrantino and
(lemons) ovule 6-Methylaminopurine 1000 Ix, 27 °C Russo (1980)
C. paradisi Duncan Ovule MT 500 mg/I malt extract Growth regulator-free 10 mg/I GA, 16 h daylight Vardi etal. (1982)
(grapefruits) 26 ± I °C
C. grandi.l· Bei-pei Endosperm MT 2 mg/I 2,4-D + 5 mg/I (2MT) (2MT) Dark (only Wang and Chang
(pumelos) pumelo BA + 1000 mg/I CH 2-15 mg/I GA, 2-15 mg/I GA] callus induction) (1978)
C. aurantium Anther MS 0.02 mg/I kinetin Growth regulator-free Dark, 28°C (only Hidaka (1982)
(sour orange) callus induction)
C. reticulata Dancy Ovule MT 500 mg!1 malt extract Growth regulator-free 10 mg!1GA, 16 h daylight Vardi and Spiegel-
(mandarins) 26 ± I °C Roy (1982)
Ponkan Embryo MT 500 mg/I malt extract I J.lM zeatin or IJ.lMGA] 16 h daylight Hidaka and Kajiura
+ 50 J.lM kinetin growth regulator-free 4.7-8.6 Wm-2 (\988)
25 ± I °C
Table 1. (con/d)
Species CuItivars Explants Basal Growth regulators and other elements Plant regeneration Other culture References
medium conditions
Callus induction Embryoid induction
Satsuma Undeveloped MT 185 11M adenine + 5% lactose 2.8IlMGA, Dark, 25 ± I °C Ling et al. (1990)
Kusumoto- ovule 500 mg/l malt extract (only callus
wase induction)
Satsuma Undeveloped MT 40 11M adenine + 5% lactose I mg/lGA, 38 Ilffiol m-'s-l, Kunitake et al.
Miyagawa- ovule Img/lGA, + continuous iIIumi- (199Ia)
wase 500 mg/l malt extract nation, 25°C
Satsuma Juice vesicle MT Img/INAA I mg/lGA, 16 h daylight, Nito and Iwamasa
Hayashi- 17.71lmol m-'s-l, (1990)
unshiu 25 ± I ·C
C. madurensis Hypocotyl of MT I mg/lGA, Growth regulator-free Growth regulator-free 16 h daylight Ling et al. (1989)
anther-derived 35.3 Ilmol/m-'s-l
plantlet 25 ± I ·C
Poncirus Embryo MT 2 mg/l2,4-D + I mg/lNAA+ I mg/lGA,or Dark, 25 ± 2 ·C Beloualy (1991)
trifoliata 5 mg/l BA + 5 mg/IBA I mg/I GA, + I mg/I NAA (only callus
(trifoliate 1000 mg/l CH induction)
oranges) Anther MS 0.2mg/l IAA Growth regulator-free Dark, 27 ± I ·C Hidaka et al.
(only callus (1979)
induction)
Fortunella Ovule MT 500 mg/I malt extract 500 mg/I malt extract 0.1 mg/I NAA Nagata et al.
crassfolia 0.1 M lactose (1992)
Microcitrus sp. Ovule MT 500 mg/I malt extract Growth regulator-free Growth regulator-free Dark, 25 ± 3 °C Vardi et al.
(only callus (1986)
induction)
Citropsis Ovule MT 0.05 mg/I 2,4-D + 5% lactose 0.02 mg/l NAA 16 h daylight Ling et al.
gabunensis (zygotic 0.05 mg/I BA + 35.4 Ilffiol m-'s-l (1991)
embryo) 400 mg/l malt extract 25 ± I °C
Severinia Ovule MT 0.05 mg/I 2,4-D + 5% lactose 0.02 mg/l NAA 16 h daylight Ling et al.
buxifolia (zygotic 0.05 mg/I BA + 35.4 Ilffiol m-'s-l (1992)
embryo) 400 mg/I malt extract 25 ± I °C
• Murashige and Tucker (1969).

b Casein hydrolysate.
Somatic Embryogenesis in Citrus Species 283
(Kochba et al. 1972). Plant regeneration from protoplasts was also achieved
using embryogenic callus cultures as the source of protoplasts by Vardi (1977).
Since then, a protoplast culture system has been established for numerous Citrus
species and cultivars (Vardi et al.1982; Kobayashi et al. 1983, 1985; Hidaka and
Kajiura 1988). However, in Satsuma mandarin, induction of embryogenic
calli by culturing immature ovules has been difficult. Ling et al. (1990) and
Kunitake et al. (1991a) recently succeeded in inducing embryogenic calli by
culturing undeveloped ovules excised from mature fruits of Satsuma mandarin,
and in regenerating whole plants from protoplasts isolated from the embryogenic
calli.
Ohgawara et al. (1985) fused embryogenic protoplasts of Trovita orange
with those of Poncirus trifoliata and obtained somatic hybrids. To date, more
than ten interspecific and intergeneric somatic hybrids have been obtained in
Citrus and its related genera (Grosser and Gmitter 1990). Furthermore,
transgenic plants have been obtained in Washington navel orange (Hidaka et al.
1990), Trovita orange and rough lemon (Vardi et al. 1990). These successful
reports were based on the use of embryogenic calli, which has a high potential to
regenerate plants.
Recently, cryopreservation of plant cells and meristems has become an
important tool for the long-term preservation of valuable germplasm and
important experimental materials without genetic alteration (Bajaj 1979, 1991;
Sakai 1985). In Citrus cryopreservation has great potential in germplasm preser-
vation (Bajaj 1984). Hidaka and Omura (1989) established methods for con-
trolling somatic embryogenesis in several Citrus species and proposed a
"callus bank" for the in vitro preservation of citrus germplasm. Furthermore,
Kobayashi et al. (1990) and Sakai et al. (1991) succeeded in regenerating
whole plants from cryopreserved embryogenic cells.
This chapter describes several factors for inducing somatic embryogenesis
and discusses the application of somatic embryogenesis for the genetic improve-
ment and preservation of Citrus germplasm.
2 Induction of Embryogenic Calli
2.1 Species and Cultivars
Since the first report on the successful induction of embryogenic calli from
nucellar tissues of Shamouti sweet orange by immature ovule culture (Kochba
et al. 1972), somatic embryogenesis of many cultivars of poly embryonicspecies
such as C. sinensis, C. aurantium, C. paradisi, C. reticulata, C. limon, and C.
madurensis has been reported by numerous workers (Table 1). Our results on the
induction of embryogenic calli from immature ovules in other Citrus cultivars are
shown in Table 2. Among the Citrus cultivars tested, high frequency embryogenic
callus formation was shown in sweet oranges such as Yoshida navel and Murcott
(Fig. lA-C). Eureka lemon and sour orange produced nonembryonic calli
284 H. Kunitake and M. Mii
Table 2. Induction of embryogenic calli on immature ovule culture of polyembryogenic Citrus

cultivars
Class Cultivar Embryogenic callus formation (%)"
Culture medium
MT + 10 mgtl BA MT + 500 mgtl MEb
Mandarin Okitsu-wase o o
Shiikuwasha o 13.8
Orange Joppa 1.3 2.1
White Silletta 15.3 o
Yoshida navel 29.4 13.5
Maltese blood 2.5 o
Sour orange o o
Tangor Murcott 3.3 23.7
Tangelo Mineola o 1.5
Grapefruit Marsh seedless o o
Lemon Eureka o o
Acid fruits Yuzu 1.9 o
Kabosu 1.6 o
a Immature ovules 2-4 weeks after anthesis were used throughout this experiment. Frequencies of
embryogenic callus formation from immature ovules were recorded after 3 months of culture.
b Malt extract.
derived from integument tissues (Fig. 2.). No reactions were shown in immature
ovules of Satsuma mandarin (c. reticulata cv. Okitsu-wase). However, embryo-
genic calli of Satsuma mandarin were produced from undeveloped ovules of
mature fruits (Ling et al. 1990; Kunitake et al. 1991a). In the related genera,
trifoliate orange (Poncirus trifoliata Raf.) and kumquat (Fortunella crassifolia
Swingle), whole plants were regenerated from embryogenic calli derived from
zygotic embryos and nucellar tissue, respectively (Beloualy 1991; Nagata et al.
1992).
Previous attempts to induce embryogenic calli from nucellar tissue of
monoembryonic cultivars were not successful (Button and Kochba 1977;
Kobayashi et al. 1984; Moore 1985). The difficulty was considered to be due to a
lack of nucellar embryos or their primordium cells in monoembryonic cultivars
(Kobyashi et al. 1981). However, recently, Ling (1991) has reported that
embryogenic calli could be induced from zygotic embryos of monoembryonic
Citrus cultivars. Furthermore, Vardi et al. (1986) and Ling (1991) showed that
embryogenic calli or monoembryonic wild relatives such as Microcitrus sp.
Citropsis gabunensis, and Severinia buxifolia were produced from immature,
ovule-derived globular embryos by caulogenesis. Although these embryogenic
calli derived from zygotic embryos are not true-to-type and differ genotypically
from line to line, they could be useful materials for in vitro germplasm preser-
vation of Citrus and other related wild species.
Fig. lA- C. Induction of embryogenic

calli from immature ovule of Yoshida
navel (Citrus sinensis Osb.). A Cross
section of immature ovule 100 days after
culture. I Integument; C nucellar-
derived callus; bar = 2 mm. B Nucellar
callus and embryoid formation from
immature ovule on MT medium
containing 10 mg/I BA. C Callus; E
=
embryoids; bar 3 mm. C Habituated
callus with white, friable, and nodular
appearance subcultured on MT medium
containing 10 mg/l BA; bar = 10 mm
Fig. 2. Non-embryogenic calli

induced from integument tissues of
sour orange (Citrus aurantium)
and Eureka lemon (Citrus lemon).
=
Bar 10mm
2.2 Explants
The most suitable explant to induce embryogenic calli of Citrus is immature

ovule which can be isolated from young fruits several weeks after anthesis.
Embryogenic calli have been efficiently obtained from nucellar tissues in ovule
culture of many sweet orange cultivars. Embryogenic calli were also obtained
from embryos, hypocotyls, anthers, endosperm, juice vesicles of several species,
and cultivars (Tablel). However, the use of these explants resulted in low
frequency and unstable induction of embryogenic calli. Although other explants
Fig. 3. A Adventitious shoot formation from hypocotyl segment of Valencia sweet orange (Citrus
sinensis Osb.) 30 days after culture; bar = 10 mm. 8 Cross section of adventitious shoot formation
from trifoliate orange (Poncirus trifoliata) cotyledon-derived callus. C Cotyledon; CA callus; S
adventitious shoot formation. Bar = 2 mm
such as shoot tips, stem nodes, internodes, epicotyls, cotyledons, and root tissues
were also used as explants, embryo ids were not observed, but shoots and whole
plants were regenerated through organogenesis (Grinblat 1972; Chaturvedi and
Mitra 1974; Einset 1978; Burger and Hackett 1981; Kitto and Young 1981;
Barlass and Skene 1982; Edriss and Burger 1984; Duran-Vila et al. 1989). In our
experiments, organ formation was also observed in hypocotyl culture of sweet
orange and cotyledon tissue culture of trifoliate orange (Fig. 3A,B).
2.3 Culture Media
Generally, synthetic auxins such as 2,4-D, picioram, and dicamba have been used
for the induction of embryogenic calli in many higher plants. However, Citrus
species show unique characteristics in response to growth regulators in the
induction of embryogenic calli. Induction of embryogenic calli of Citrus can be
achieved by cytokinins, however several substances such as casein hydrolysate,
malt extract, and auxins have no important role in the induction of embryogenic
calli.
Culture media previously used for inducing embryogenic calli of Citrus are
shown in Table I. Embryogenic calli from nucellar tissue could be induced by
ovule culture on MT medium containing 500 mg/\ malt extract (Kochba et al.
1972; Vardi et al. 1982), or on MT medium containing a high concentration of
cytokinin (Kobayashi et al. 1984). Kobayashi et al. (1984) reported that the
addition of auxin to the induction medium promoted callus formation from
integument tissues, but inhibited embryogenic callus formation from nucellar
tissue. Furthermore, they showed that the addition of 10 mg/l BA increased the
induction frequency of embryogenic callus and promoted callus growth .
Hidaka and Kajiura (\988) reported that (MS) medium (Murashige and
Skoog 1962) containing 500 mg/I malt extract and 50 f-lM kinetin was most
effective for callus production from young, seed-derived embryos. They
described that the addition ofNAA and GA3 resulted in the formation of only
compact yellowish calli, and these calli did not differentiate into embryoids after
subculture.
Gmitter et al. (1990) reported that embryogenic calli were induced from
endosperm on MT medium containing 5 mg/1 BA, 2 mg/I 2,4-D, 1000 mg/I
casein hydrolysate, 5 mg/I kinetin, and 500 mg/I malt extract. Particularly casein
hydrolysate and malt extract influenced the frequency of callus induction from
"Pineapple" sweet orange endosperm.
3 Induction of Somatic Embryos (Embryoids)
3.1 Saccharides
For embryoid induction from embryogenic calli of Citrus, perhaps the most
important factor is the response of callus to several saccharides. Button (1978),
Kochba et al. (1982) and Ben-Hayyim and Neumann (1983) reported that
somatic embryogenesis of nucellar calli could be stimulated when the sucrose
in the medium was replaced by several saccharides such as galactose, lactose,
Table 3. Effect of saccharides on embryoid formation of embryogenic calli of several Citrus cultivars
Class Cultivar Average number of embryoids "
Saccharides b
Sucrose Lactose Glucose Fructose Galactose
Mandarin Ponkan 9 132 5 89 227

Okitsu-wase 0 67 0 0 0
Orange Shamouti 74 305 104 0 90
Jaffa 47 205 140 0 0
Salustiana 139 243 44 165 316
Trovita 89 99 193 49 152
Valencia 93 165 40 38 39
Washington navel 101 397 0 15 53
Yoshida navel 0 74 0 I 0
Morita navel 18 477 0 80 0
Adachi navel 108 453 0 26 0
Torocco 0 281 127 0 61
Calamondin' II 895 2 0 615
Tangor Murcott 19 219 7 12 127
Acid Fruits Yuzu 7 66 0 0 0
Lemon Meyerd 73 1139 64 58 466
a Embryoids that developed into globular or more advanced stages were recorded after 2 months of
culture. The data indicate the means of one experiment, each with 10 or 15 replicates.
b Concentration of each saccharide was 0.2M.
C Callus was induced from hypocotyl of anther-derived plantlets by Ling et al. (1989).
d Embryogenic callus of Meyer lemon was induced from zygotic embryos by culturing immature
ovules.
Fig.4. A Formation of numerous embryoids of Meyer lemon on MT medium containing 5% lactose

60 days after culture; har = 10 mm. B Various stages of embryoid formation of Meyer lemon (Citrus
lemon) on MT medium containing 5% lactose 60 days after culture; bar = 10 mm
and glycerol. Spiegel-Roy et al. (1978) proposed that this phenomenon can
be ascribed to a reduction in exogenous auxin in embryogenic calli. The
effectiveness of these saccharides has also been reported by Kobayashi et al.
(1984), Hidaka and Omura (1989), and Kunitake et al. (l991a). Furthermore,
Hidaka and Omura (1989) reported that embryoid formation was stimulated not
only by galactose and lactose, which yield galactose, but also by maltose and
cellobiose, which yield glucose. However, the physiological mechanism involved
in the effects of glucose-yielding saccharides on Citrus embryogenic calli is
unclear.
Differences in the response of embryogenic calli to several saccharides were
observed among the species and cultivars of Citrus (Vardi et a1.l982; Kochba
et aI.1982). In our experiment, high frequency embryoid formation was shown in
all of the Citrus cultivars examined by replacing sucrose with galactose or lactose;
these reactions are obviously different among the cultivars (Table 3). Particularly
embryoid formation of zygotic embryo-derived embryogenic calli in Meyer
lemon was higher than that of other embryogenic calli (Fig. 4A,B). Embryogenic
calli of Satsuma mandarin produced embryoids only when lactose was supplied
and no embryoid formation was observed with other saccharides.
3.2 Growth Regulators
The effect of growth regulators on the induction of embryeids from nucellar calli
has been extensively studied. Gibberellic acid (Kochba et al. 1978), cytokinin
(Kochba and Spiegel-Roy 1977; Kobayashi et al. 1984), and auxin (Kochba and
Spiegel-Roy 1977) marked inhibited embryoid formation. In contrast, malt
extract, adenine (Kochba and Spiegel-Roy 1973), ethephon, and ABA (Kochba
and Spiegel-Roy 1977) considerably promoted somatic embryogenesis. Further-
more, inhibitors of GA3 biosynthesis such as CCC (2-chloroethyl trimethyl-
ammonium) and Alar (succinic acid 2,2-methylhydrazide), antiauxins
(7-aza-indole and 5-hydroxynitrobenzylbromide), and anticytokinin (azagua-
Fig. SA-E. Somatic embryogenesis and plant regeneration from nucellar calli of Valencia sweet
orange (Citrus sinensis Osb.). AGlobular embryoid; bar = I mm. B Heart-shaped embryoid; bar = I
mm. C Cotyledonary embryoid; bar = 2 mm. D Plantlet formation; bar = 20 mm. E Plant
regeneration; bar = 100 mm
nine) promoted embryoid development in nucellar calli (Kochba and Spiegel-

Roy 1977; Kochba et al. 1978).
3.3 Other Factors
Kochba and Button (1974) reported that aging of the callus prior to subculture
and sucrose starvation of embryogenic calli for several days enhanced embryoid
development.
Spiegel-Roy and Kochba (1973) exposed embryogenic calli of Shamouti

orange to gamma irradiation at a dosage of 0.5 to 32 Kiloroentgen (kR) and
showed that irradiation at a dosage of 0.5 to 8 kR stimulated callus growth,
whereas 4 to 20 kR stimulated embryoid development. They assumed that
irradiation caused a decrease in endogenous auxin, enchanced utilization of
galactose or galactose-yielding saccharides, and gave rise to changes in callus
growth. It was assumed that these were the main reasons for the stimulation of
4 Plant Regeneration from Emhryoids
Generally, proembryoids (approximately I mm in diameter; Fig 5 A) induced

from nucellar calli were transferred to MT or MS medium containing gibberellic
acid to stimulate germination (Kobayashi et al. 1984; Hidaka and Omura 1989;
Ling et al. 1990; Kunitake et al. 1991a). In the presence of gibberellic acid, they
developed first into heart-shaped (Fig. 5B), torpedo, and cotyledonary
embryoids (Fig. 5C), and then into plantlets (Fig. 50). Vardi and Galun (1989)
and Beloualy (1991) reported that the addition of NAA to MT medium
Fig. 6A-D. Cross section of somatic embryogenesis of Okitsu-wase, Satsuma mandarin (Citrus
reticulata) nucellar calli. A Single cell or groups of cells with thickened cell walls on MT medium
=
containing 5% lactose I week after culture; bar 50 ).1m. B Early stage of proembryoid formation;
bar = 50 ).1m. C Globular embryo ids formation 4 weeks after culture; bar = 1 mm. D Globular
embryoid with a vascular bundle; bar = 1 mm
promoted root formation and axis elongation in embryo ids which had expanded
cotyledonary leaves. Plant lets with two or three mature leaves that developed
from embryoids could be transplanted to pots in a greenhouse after accli-
matization (Fig. 5E) .
5 Histological Observation of Somatic Embryogenesis
A morphological and histological process of somatic embryogenesis from

nucellar calli was first reported in Shamouti orange (Button et al. 1974).
Histological observation of somatic embryogenesis from nucellar calli was also
reported in Satsuma mandarin (Kunitake et al. 1991 b). In Satsuma mandarin, 1
week after transfer to MT medium containing 5% lactose, single cells or groups
\ ~.
I
m 0 fl
~ ~.
J \.
~
t'~ J )
"(, ~ \ ~ til Ii] rn
t'
~114
( ' "I \~
(P 11
I I
~ iii rJ
,~ ~,
r ~,~ ,
!J ,~
B iii ~ Ii iii ~ Ii
Fig. 7A,B. Comparison ofleaf morphology among the plants derived from different sources. A Leaves
of embryoid-derived plants (Pl-P6) and nucellar seedlings (N) of Valencia sweet orange (Citrus
sinensis Osb.). B Leaves of embryoids-derived plants (Pl-P6), anther-derived plants (A) , nucellar
seedlings (NS), and mature trees (PA) of Calamondin (Citrus madurensis Lour.)
of cells with thickened cell walls were observed (Fig. 6A). These cell continued to
divide (Fig. 6B) and became globular pro embryo ids after 3-4 weeks of culture
(Fig. 6C). The globular embryoid clearly showed a development of the vascular
bundle (Fig. 60). This process closely resembled Citrus nucellar embryony. One
to 2 months after transplanting these embryo ids to the regeneration medium, 5%
of them showed normal germination and regenerated plantIets via cotyledonary
embryoids. Fifty to 60%, however, showed abnormal growth with the pro-
duction of secondary embryoids on their surfaces. Factors affecting normal
germination of embryoids in Satsuma mandarin are still unclear and should also
be clarified with special reference to the relationship between dormancy and
developmental stage of somatic embryoids.
p N 1 2 3 456
B
P N A 1 2 3 4 5 6
Fig. SA,B. Comparison of zymograms among the plants derived from different sources. A Isoelectric
focusing profiles of GOT (glutamate oxaloacetate transaminase) from embryoid-derived plants (1-6),
mature plants (P), and nucellar seedling (N) of Valencia sweet orange (Citrus sinensis Osb.) B
Isoelectric focusing profiles of GOT (glutamate oxaloacetate transaminase) from embryoid-derived
plants (1-6), mature plants (P) , nucellar seedling (N), and anther-derived plants (A) of Calamondin
(Citrus madurensis Lour.)
Fig.9. Chromosomes in a root tip cell of an embryoid-derived plant of Valencia sweet orange (Citrus
sinensis Osb.); 2n = 18
6 Stability of Embryoid-Derived Plants
Somatic embryoid-derived plants in several Citrus species have been reported to

be genetically stable with respect to morphology, cytology and physiology (Vardi
1981; Vardi et al. 1982; Kobayashi et al. 1984; Kobayashi 1987; Hidaka and
Omura 1989). We also found no variations in the examined leaf shape index
(Fig. 7A,B, Tables 4,5), the isozyme banding pattern of three enzymes
(glutamate oxaloacetate transaminase, peroxidase, phospho glucose mutase,
Fig. 8A,B), the gas chromatographic pattern ofleaf oil, or chromosome number
(Fig. 9) in embryogenic calli-derived plants of Valencia orange and Calamondin
(c. madurensis Lour.). Furthermore, no differences were noted among the
regenerated plants, nucellar seedlings, or mature trees.
So far, embryogenic calli have been induced from nucellar tissues of many
polyembryonic cultivars and zygotic embryos of monoembryogenic cultivars
and from wild relatives. Generally, a high concentration of cytokinin such as BA
is effective in increasing the induction frequency of embryogenic calli and in
stimulating callus growth. It is also interesting to note that relatively uncommon

saccharides such as galactose, lactose, and glycerol stimulate embryoid
formation from Citrus embryogenic calli, and reactions of the embryogenic calli
showed differences among Citrus species and cultivars. These reactions to
cytokinin, and saccharides and complexed natural products such as casein
hydrolysate and malt extract seem to be novel and unique characteristics of
Citrus species compared with somatic embryogenesis of other plants.
Furthermore, normality in regard to morphology, cytology and physiology
observed in embryoid-derived plants, is important for Citrus breeding using cell
and genetic engineering and for the preservation of Citrus germ plasm. On the
basis of these results, somatic hybrids, hybrids, and transformants have been
produced in Citrus and its relatives for the purpose of Citrus breeding. It can
be expected that, in the near future, novel Citrus cultivars will be produced
with disease and stress resistance which cannot be produced by conventional
breeding. Moreover, a "callus bank" could be successfully established for the
preservation of Citrus germplasm by utilizing cryopreservation of embryogenic
calli or cells.
Table 4. Comparison of the leaf shape index among

embryoid-derived plants, nucellar seedlings and mature
plants of Valencia sweet orange (Citrus sinensis Osb.)
Plant Leaf shape indexb
Embryoid-derived
plant" I 2.08 ± 0.15
2 2.04 ± 0.16
3 2.04 ± 0.10
4 2.11±0.13
5 2.15±0.17
6 2.08 ± 0.18
7 2.18 ± 0.19
8 2.07 ± 0.17
9 2.16 ± 0.17
10 2.11 ± 0.12
Nucellar seedling I 2.11 ± 0.20

2 2.09 ± 0.21
3 2.ll±O.15
Mature plant 2.04 ± 0.10
"Embryoid-derived plants year after acclimatization

were used throughout this experiment.
b Length/ width of the leaf blade. Each value represents the
mean± SE.
TableS. Comparison of the leaf shape index among embryoid-derived plants, anther·derived plants,
nucellar seedlings and mature plants of Calamondin (Citrus madurensis Lour.)
Plant Leaf shape Plant Leaf shape

indexb indexb
Embryoid·derived Embryoid-derived
plant" 2.17 ± 0.14 plant" 19 2.23 ± 0.18
2 2.17 ± 0.06 20 2.24 ± 0.10
3 2.30 ± 0.12 21 2.17 ± 0.15
4 2.19 ± 0.09 22 2.20 ± 0.10
5 2.17 ± 0.16 23 2.31 ±0.15
6 2.23 ± 0.09 24 2.14 ± 0.17
7 2.26 ± 0.14 25 2.24 ± 0.10
8 2.23 ± 0.17 Anther-derived
9 2.16±0.11 plant I 2.01 ± 0.08
10 2.14 ± 0.11 2 1.96 ± 0.16
II 2.23 ± 0.07 3 1.99 ± 0.07
12 2.22 ± 0.18 4 2.02 ± 0.09
13 2.26 ± 0.08
14 2.13 ± 0.05 Nucellar seedling I 2.23 ± 0.20
IS 2.14 ± 0.16 2 2.25 ± 0.13
16 2.28 ± 0.10
17 2.24 ± 0.19 Mature seedling I 2.03 ± 0.08
18 2.21 ± 0.20 2 1.98 ± 0.23
"Embryoid-derived plants and anther-derived plants I year after acclimatization were used
throughout this experiment. Embryogenic calli of Cal amon din were induced from hypocotyl of anther·
derived plantlets by Ling et al. (1989).
bLength/width of the leaf blade. Each value represents the mean ± SE.
8 Protocol
8.1 Somatic Embryogenesis and Plant Regeneration from Nucellar Calli

of Satsuma Mandarin (Citrus unshiu Marc.)
8.1.1 Induction of Nucellar Calli from Undeveloped Ovules
I. Harvest open· pollinated mature fruits from trees of Satsuma mandarin.

2. Sterilize mature fruits in 70% ethanol for 30 s, cut transversely, and then extract undeveloped ovules
(approximately 2 mm in length) aseptically.
3. Place the undeveloped ovules into test tubes, each containing 20ml ofMTmedium (Murashige and
Tucker 1969) with 500 mgl1 malt extract, 40 mgtl adenine, I mgtl gibberellic acid, 5% sucrose, and
0.2% Gellan gum.
4. Keep at 25 °C under continuous illumination at 38 !lmol m-'s-i.
5. When several embryo ids or white, friable, and nodular calli are induced from the undeveloped
ovules, transplant these cultures into the same medium.
6. After proliferation of the calli subculture them at I-month intervals under the same culture
conditions.
8.1.2 Somatic Embryogenesis and Plant Regeneration
I. Transplant the calli 3-4 weeks after subculture on MT medium containing 5% lactose and 0.2%
Gellan gum. After approximately 8 weeks of culture, various stages of green embryoids should
have formed.
2. Isolate the embryoids (approximately I mm in diameter) individually and transfer them to a test
tube containing MT medium with I mg/I gibberellic acid, 3% sucrose, and 0.2% Gellan gum to
stimulate germination.
3. When 2~3 mature leaves develop and roots are formed, transplant the plantlets to pots containing
vermiculite and peat moss (I: I mix) and cover the plantlets with a plastic container.
4. Place in an incubator to acclimatize for 3-4 weeks. During this period gradually open the plastic
container in order to lower the humidity until the plants can adopt to greenhouse conditions.
Acknowledgments. The authors are grateful to Dr. M. Iwamasa (Faculty of Agriculture, Kyushu
Tokai University), Dr. N. Nito (Faculty of Agriculture, Saga University), and Dr. J.T. Ling (School of
Agriculture and Home Economics, Tennessee State University), for their valuable comments and
advice regarding the experiments.
References
Bajaj YPS (1979) Establishment of germplasm banks through freeze-storage of plant tissue culture
and their implications in agriculture. In: Sharp WR, Larsen PO, Paddock EF, Raghavan V (eds)
Plant tissue culture ~ principles and applications. Ohio State University Press, Columbus, pp
745~774
Bajaj YPS (1984) Induction of growth in frozen embryos of coconut and ovules of citrus. Curr Sci 53:
1215~1216
Bajaj YPS (1991) Storage and cryopreservation of in vitro cultures. In: Bajaj YPS (ed) Biotechnology
in agriculture and forestry, vol 17 , High-tech and micropropagation I. Springer, Berlin Heidelberg
New York, pp 361~381
Barlass M, Skene KGM (1982) In vitro plantlet formation from Citrus species and hybrids. Sci Hortic
17: 333-341
Beloualy N (1991) Plant regeneration from callus culture of three Citrus rootstocks. Plant Cell Tissue
Organ Cult 24: 29~34
Ben-Hayyim G, Neumann H (1983) Stimulatory effect of glycerol on growth and somatic
embryogenesis in Citrus callus culture. Z PflanzenphysiolllO: 331~337
Burger DW, Hackett WP (1981) Regeneration of buds and roots from several citrus tissues. Proc Soc
Int Citricult I: 161~163
Button J (1978) The effect of some carbohydrates on the growth and organization of Citrus ovular
callus. Z Pflanzenphysiol88: 61-68
Button J, Kochba J (1977) Tissue culture in the citrus industry. In: Reinert J, Bajaj YPS (eds) Applied
and fundamental aspects of plant cell, tissue, and organ culture. Springer, Berlin Heidelberg
New York, pp 70-92
Button J, Kochba J, Bornman CH (1974) Fine structure of and embryoid development from
embryogenic ovular callus of Shamouti orange (Citrus sinensis Osb.). J Exp Bot 25: 446-457
Chaturvedi HC, Mitra G (1974) Clonal propagation of Citrus from somatic callus cultures.
HortScience 9: 118~ 120
Duran-Vila N, Ortega V, Navarro L (1989) Morphogenesis and tissue cultures of three Citrus species.
Plant Cell Tissue Organ Cult 16: 123~133
Edriss MH, Burgar DW (1984) In vitro propagation of Troyer citrange from epicotyl segments. Sci
Hort 23: 159~162
Einset JW (1978) Citrus tissue culture. Stimulation offruit explant cultures with orange juice. Plant
Physiol 62: 885~888
FAO (1992) Production yearbook. FAO, Rome

Fortucci-Marongiu (1988) Three decades of the world citrus economy. In: Goren R, Mendel K (eds)
Proc 6th Int Citrus Congr. Balaban, PhiladelphialRehovot, pp 19-31
Gmiter FG, Moore GA (1986) Plant regeneration from undeveloped ovules and embryogenic calli of
Citrus: embryo production, germination, and plant survival. Plant Cell Tissue Organ Cult 6:
139-147
Gmitter FG, Ling XB, Deng XX (1990) Induction of triploid Citrus plants from endosperm calli in
vitro. Theor Appl Genet 80: 785-790
Grinblat V (1972) Differentiation of citrus stem in vitro. J Am Soc Hortic Sci 97: 599-603
Grosser JW, Gmitter FG (1990) Protoplast fusion and citrus improvement. Plant Breed Rev 8:
339-374
Hidaka T (1984) Induction ofplantlets from anthers of "Trovita" orange (Citrus sinensis Osb.) J Jpn
Hort Sci 53: 1-5
Hidaka T, Kajiura M (1988) Plantlet differentiation from callus protoplasts induced from citrus
embryo. Sci Hortic 34: 85-92
Hidaka T, Omura M (1989) Control of embryogenesis in Citrus cell culture. Regeneration from
protoplasts and attempts to establish a callus bank. Bull Fruit Tree Res Stn Ser B (Okitsu) 16: 1-17
(in Japanese)
Hidaka T, Yamada Y, Shichijo T (1979) In vitro differentiation of haploid plants by anther culture in
Poncirus trifoliata (L) Raf. Jpn J Breed 29: 248-254
Hidaka T, Yamada Y, Shichijo T (1982) Plantlet formation from anthers of Citrus aurantium L. Jpn
J Breed 32: 247-252
Hidaka T, Omura M, Ugaki M, Tomiyama M (1990) Agrobacterium mediated transformation and
regeneration of Citrus spp. from suspension cells. Jpn J Breed 40: 199-207
Kitto SL, Young MJ (1981) In vitro propagation of Carrizo citrange. HortScience 16:305-306
Kobayashi S (1987) Uniformity of plants regenerated from orange (Citrus sinensis Osb.) protoplasts.
Theor Appl Genet 74: 10--14
Kobayashi S, Ikeda I, Nakatani M (1981) Role of the primordium cells in nucellar embryogenesis in
citrus. Proc Int Soc Citricult 1: 44-48
Kobayashi S, Uchimiya H, Ikeda I (1983) Plant regeneration from Trovita orange protoplasts. Jpn J
Breed 33: 119-122
Kobayashi S, Ikeda J, Nakatani M (1984) Induction of nucellar callus from orange (Citrus sinensis
Osb.) ovules, and uniformity of the regenerated plants. Bull Fruit Tree Res Stn Ser E 5: 43-54
(in Japanese)
Kobayashi S, Ikeda I, Uchimiya H (1985) Conditions for high frequency embryogenesis from orange
(Citrus sinensis Osb.) protoplasts. Plant Cell Tissue Organ Cult 4: 249-259
Kobayashi S, Sakai A, Oiyama 1(1990) Cryopreservation in liquid nitrogen of cultured navel orange
(Citrus sinensis Osb.) nucellar cells and subsequent plant regeneration. Plant Cell Tissue Organ
Cult 23: 15-20
KochbaJ, Button J (1974) The stimulation of embryogenesis and embryoid development in habituated
ovular callus from the Shamouti orange (Citrus Sinensis) as affected by tissue age and sucrose
concentration. Z Pflanzenphysiol81: 283-288
Kochba J, Speigel-Roy P (1973) Effect of culture media on embryoid formation from ovular callus of
Shamouti orange (Citrus sinensis). Z Pflanzenzucht 69: 159--162
Kochba J, Spiegel-Roy P (1977) The effect of auxins, cytokinins and inhibitors on embryogenesis in
habituated ovular callus of the Shamouti orange (Citrus sinensis). Z Pflanzenphysiol81: 283-288
Kochba J, Spiegel-Roy P, Saad S (1972) Adventive plants from ovules and nucelli in citrus. Planta
106:237-245
Kochba J, Spiegel-Roy P, Neumann H, Saad S (1978) Stimulation of embryogenesis in citrus ovular
callus by ABA, ethephon, CCC and Alar and its suppression by GA3. Z Pflanzenphysiol 89:
427-432
Kochba J, Spiegel-Roy P, Neumann H, Saad S (1982) Effect of carbohydrates on somatic
embryogenesis in subcultured nucellar callus of Citrus cultivars. Z Pflanzenphysioll05: 359-368
Kunitake H, Kagami H, Mii M (1991a) Somatic embryogenesis and plant regeneration from
protoplasts of "Satsuma" mandarin (Citrus unshiu Marc.). Sci Hortic 47: 27-33
Kunitake H, Kagami H, Mii M (199Ib) Plant regeneration from protoplasts of Satsuma mandarin
(Citrus unshiu Marc.) via somatic embryogenesis. JARC 24: 287-291
298 H. Kunitake and M. Mii: Somatic Embryogenesis in Citrus Species
Ling JT (1991) Studies on tissue, cell and protoplast culture in Citrus. PhD Thesis, Kagoshima
University, Japan
Ling JT, Nito N, Iwamasa M (1989) Plant regeneration from protoplasts of Calamondin (Citrus
madurensis LouL). Sci Hortic 40: 325-333
Ling JT, Nito N, Iwamasa M, Kunitake H (1990) Plant regeneration from protoplasts isolated from
embryogenic callus of Satsuma. HortScience 25(8): 970-972
Moore GA (1985) Factors affecting in vitro embryogenesis from undeveloped ovules of mature Citrus
fruit. J Am Soc Hortic Sci 110: 66-70
Murashige T, Tucker DPH (1969) Growth factor requirement of citrus tissue culture. Proc 1st Int
Citrus Symp 3: 1155-1161
Nagata R, Kashino T, Todoroki A (1992) Isolation of protoplasts and regeneration of plantlets from
kumquat nucellar callus. J Jpn Soc Hortic Sci 61 (Suppll): 772 (in Japanese)
Nito N, Iwamasa M (1990) In vitro plantlet formation from juice vesicle callus of satsuma (Citrus
unshiu Marc.). Plant Cell Tissue Organ Cult 20:137-140
Ohgawara T, Kobayashi S, Ohgawara E, Uchmiya H, Ishii S (1985) Somatic hybrid plants obtained
by protoplast fusion between Citrus sinensis and Poncirus trifoliata. Theor Appl Genet 71: 1-4
Sakai A (1985) Cryopreservation of shoot-tips offruit trees and herbaceous plants. In: Kartha KK
(ed) Cryopreservation of plant cells and organs. CRC, Boca Raton, pp 243-257
Sakai A, Kobayashi S, Oiyama I (1991) Cryopreservation of nucellar cells of navel orange (Citrus
sinensis Osb.) by a simple freezing method. Plant Sci 74: 243-248
Spiegel-Roy P, Kochba J (1973) Stimulation of differentiation in orange (Citrus sinensis) ovular callus
in relation to irradiation of the media. Radiat Bot 13: 97-103
Spiegel-Roy P, Saad S, Neumann H (1978) Stimulation of embryogenesis in Citrus tissue culture by
galactose. Naturwissenschaften 65: 261-262
Starrantino A, Russo F (1980) Seedling from undeveloped ovules of ripe fruits of polyembryonic
Citrus cultivars. HortScience 15(3): 296-297
Vardi A (1977) Isolation of pro toplasts in Citrus. Proc Int Soc Citric 1977,2: 575-578
Vardi A (1981 ) Protoplast-derived plants from different Citrus species and cultivars. Proc Int Soc
Citricult I: 149-152
Vardi A, Galun E (1989) Isolation and culture of Citrus protoplasts. In: Bajaj YPS (ed) Biotechnology
in agriculture and forestry, vol 8. Plant protoplasts and genetic engineering I. Springer, Berlin
Vardi A, Spiegel-Roy P, Galun E (1982) Plant regeneration from Citrus protoplasts: variability in
methodological requirements amongcultivars and species. Theor Appl Genet 62: 171-176
Vardi A, Hutchison DJ, Galun E (1986) A protoplast-to-tree system in Microcitrus based on
protoplasts derived from a sustained embryogenic callus. Plant Cell Rep 5: 412-414
Vardi A, Bleichman S, Aviv D (1990) Genetic transformation of citrus protoplasts and regeneration of
transgenic plants. Plant Sci 69: 199-206
Wang D, Chang CJ (1978) Triploid citrus plantlet from endosperm culture. Sci Sin 21: 822-827
(in Chinese)
111.5 Somatic Embryogenesis in Coconut
(Cocos nucifera L.)
J.L. VERDEIL 1 and J. BUFFARD-MoREL2
1 General Account
1.1 Distribution and Socioeconomic Impact
The coconut palm, (Cocos nucifera L.), also given the name "tree of life" or the
"tree of a hundred uses", plays an important socioeconomic role. It is a
heliophilous plant widely distributed throughout the humid intertropical zone,
where it is grown on more than 10 million ha in 80 countries (Pannetier and
Buffard-MorelI986).
Coconut holds the seventh position among oil-bearing plants. Industrially,
oil is extracted from the dried endosperm (6% moisture), i.e., copra. An efficient
oil mill can extract up to 62% oil and produces 32% press cake from copra. The
main copra producers are the Philippines, Indonesia, and India (Das 1985).
Copra oil is an important commercial product worldwide since, along with oil
palm kernel oil, it is the only source of short-chain fatty acids (from 8 to 14
carbon atoms) (Berger and Ong 1985); it is greatly sought after by industry for
its richness in lauric acid (48% on average); it has good lathering properties and
is used in soap manufacture, and by the cosmetics industry. The manufacture of
alcohol esters, particularly methyl esters, used primarily as a diesel fuel in
producer countries, involves glyceride transformation by alcoholysis in the
presence of catalysts (coconut shell ash) (Graille et aI. 1985). Fatty acid produc-
tion through glyceride splitting is involved in the manufacture of a wide range of
derivatives with numerous applications: production of detergents, emulsifiers
and plasticizers used as accelerators in paint drying and in rubber manufacture.
Soap stocks, which are important by-products in the oil and fat refining
industries, can be valorized by yeasts and their bioconversion makes them a
possible source of proteins in human and animal diets (Montet et aI., 1985).
The food industry uses coconut oil in margarines, synthetic creams, and confec-
tionery.
IResearch worker of CIRAD/CP

2Research worker of ORSTOM
Present address: ORSTOM, Laboratoire des Ressources Genetiques et Amelioration des Plantes
Tropicales, 911 avenue d'Agropolis, BP 5045, 34032 Montpellier Cedex, France

300 J.L. Verdeil and J. Buffard-Morel
While coconut is an industrially oriented oil crop (Gurtler and Dronne

1992), it also fits perfectly well in a smallholder economy given its myriad uses
(Hyman 1990). Every part of the plant can be used. Oil can be extracted from the
fresh nuts (Boutin 1990). The woody stem is used as a building material and in
cabinet-making. Its leaves are used in local handicrafts (shoes, hats) and as
roofing material. The processed sap is used to make sugar, syrup, and vinegar
(Reyes-JaelI985). The fibers from the husk surrounding the nut can be used to
manufacture esparto-type goods, after retting (brushes, brooms, mats, ropes).
Sawmill waste can be used to fuel wood-fired ovens.
1.2 Morphology
Coconut is a perennial monocotyledon characterized by the existence of a single

terminal vegetative bud. The species is divided into two groups of varieties,
Dwarfs and Talls, in accordance with their morphology and reproduction
biology. "Tall" coconuts make up most of the plantations; they are allogamous,
with a strong development, are late yielders (taking 6 to 10 years to flower), and
produce a small quantity of large nuts. The vegetative development of Dwarf
coconuts is less substantial and they are usually autogamous, precocious (taking
3 to 4 years to flower), and bear numerous small nuts. Their genetic variability is
much lower than that of the TalIs and they play an important role in breeding
programs.
Coconut palms are characterized by their slow vegetative development and
their sexual reproduction. An adult tree has virtually as many unopened (20 to
30) as opened leaves. Each leaf bears a flower primordium in its axil. The seed,
which is the largest in the plant kingdom, is characterized by its lack of dor-
mancy, and the time necessary for development from embryo plantlet depends
on environmental conditions.
1.3 Significance of Somatic Embryogenesis
Although coconut has been undergoing more or less empirical mass selection for
some time, genetic progress has been only possible since the 1960s, when the first
hybrids between ecotypes, which were superior to the conventional ecotypes,
were obtained. The IRHO-CIRAD breeding scheme, which has now been
adopted by all breeders, is based on reciprocal, recurrent selection. The method
is based on the search for combining ability between ecotypes and on
phenotypical choice of heritable characters.
Once the first hybrids were obtained, it proved possible to double yields
within the space of20 years. Improvement of the best hybrids led to a further 20
to 30% gain in a single generation. However, conventional breeding techniques
come up against constraints linked to the plant's morphological and biological
characters:
Somatic Embryogenesis in Coconut (Cocos nucifera L.) 301
- frequent allogamy responsible for substantial variability within progenies,

leading to considerable heterogeneity in plantations;
- very long breeding cycle (12 to 16 years);
- low multiplication coefficient, as the coconut palm is propagated by seed alone
and the number of nuts per tree remains relatively low (100 to 200 nuts per tree
per year);
- vegetative propagation does not exist under natural conditions (single termi-
nal growing point, hence no outgrowths).
Vegetative multiplication of high performance individuals remains a prom-
ising possibility for the production of homogeneous planting material and for a
substantial improvement in plantation productivity and homogeneity (Rillo,
1993). Cloning would also give rise to more rapid valorization of breeding results.
In vitro culture techniques remain the only way open for vegetative propagation
of coconut. All the teams working for more than 15 years on coconut cloning are
now concentrating on somatic embryogenesis, which appears to be the more
promising technique. Mastery of coconut regeneration by somatic embryo-
genesis also opens up the way for genetic transformation through the introduc-
tion of genes resistant to certain diseases.
1.4 Brief Review of Work by Others
The first results obtained by the main teams currently working on this subject
were promising (embryo-like formations were obtained as early as 1982); they
were reported in numerous publications. The scarcity of articles published
between 1985 and 1988 illustrates the difficulties encountered in obtaining whole
embryos with shoot development. From 1984 to 1988, whole plant regeneration
from somatic tissues was achieved, (Branton and Blake 1984; Raju et al. 1984;
Bhalla-Sarin et al. 1986; Smith 1986; Buffard-Morel et al. 1988), but the ramets
were apparently obtained from a limited number of cultures. To our knowledge,
somatic embryogenesis has still not been completely mastered and results are not
reproducible.
The reasons for the incapacity of coconut to undergo embryogenesis remain
unclear and there are probably many causes. Various hypotheses have been put
forward regarding the low reactivity of coconut tissue in vitro. Jesty and Francis
(1992) observed an important decrease in the nucleocytoplasmic ratio on imma-
ture leaffragments after culturing, resulting from a rapid increase in cell volume
with no increase in nucleus size. According to these authors, this cell imbalance
may cause coconut tissue recalcitrance. Eeuwens (1976) tried to improve callus
formation and growth by optimizing the mineral composition of the culture
media through factorial experiments. The mineral medium developed by this
author is particularly rich in potassium, chloride, and iodine (ten times the
concentration ofMurashige and Skoog's 1962 medium). At the tissue level, these
mineral requirements would seem to reflect the adaptation of coconut to its
natural environment (sea spray, salt water). While Eeuwens' mineral medium is
now very widely used for coconut, however, the results obtained show that there
are other factors limiting the regeneration of the plant. The hypothesis that a
hormone imbalance may be one of the reasons for obtaining incomplete or
abnormal embryogenesis has often been put forward in order to explain the
frequent formation of teratogenic structures (Branton and Blake 1984; Blake
1989; Dublin et al. 1991). Despite the multitude of experiments conducted, no
precise argument has been put forward to support this hypothesis.
The few ramets obtained have the merit of proving that coconut regenera-
tion by somatic embryogenesis remains a possibility that has simply not yet been
mastered.
2 Achieving Embryogenesis in Coconut. The Different Stages
2.1 Difficulties Encountered and Research Strategies
The difficulties encountered in coconut regeneration by somatic embryogenesis

are:
- intense browning of tissue linked to the oxidation of polyphenolic compounds;
- considerable heterogeneity in tissue reaction;
- a strong rooting ability and low embryogenesis and caulogenesis capability;
- an extremely long response time in vitro.
The browning observed is linked to the tissues' high sensitivity to synthetic
auxins like 2,4-D, which are essential for callus formation, and activated charcoal
is indispensable for maintaining the tissues under culturing conditions. Charcoal
is often used in plant tissue culture, especially for palms because of its promotive
effect on somatic embryogenesis. It also has powerful adsorption properties,
taking up the growth inhibitors secreted by the tissues, such as phenolic com-
pounds (Georges and Sherrington 1984). Fridborg and Eriksson (1975) sug-
gested that activated charcoal might also adsorb certain growth regulators,
particularly auxins. A kinetic study showed that 2,4-D adsorption by activated
charcoal is a gradual process. Ebert and Taylor (1990) quantified 2,4-D adsorp-
tion by activated charcoal, using molecules labeled with carbon 14 and showed
that 99.5% of the initiaI2,4-D (10-4 M) was retained by activated charcoal (2.5
gil) in a liquid medium 5 days after medium preparation. In a semiliquid medium
the time taken for stabilization was longer (10 to 20 days). They also showed that
a low pH (3.8 instead of 7.8) and medium storage at 20-30 °C instead of 5 to
10 °C accelerated 2,4-D fixation on the activated charcoal.
Measurements were taken in the laboratory by HPLC using a solid medium
with 2,4-D detection in UV; under our culture conditions powerful 2,4-D
adsorption by activated charcoal was confirmed and the time taken for the
free 2,4-D level to stabilize depended on the initial 2,4-D concentration and
the amount of activated charcoal (J.L. Verdeil, unpubl.). Under our medium
preparation conditions, in the presence of 3 gil activated charcoal and 100 mgll
2,4-D, the free 2,4-D/adsorbed 2,4-D balance is reached 21 days after media
preparation.
Similar to that observed in our laboratory, Ebert and Taylor reported

marked differences in tissue growth response, depending on the age of the
medium used and the associated variations in free 2,4-D concentrations. The
heterogeneous performance of tissues in callogenesis and the nonrepetitiveness of
results could be attributed to charcoal under certain medium preparation condi-
tions and utilization. Thus, it is now clear that activated charcoal adds a source
of variability in cultures.
Coconut calli have a low embryogenic potential; furthermore, embryogen-
esis is often incompletely expressed, either producing fused embryos, frequently
having a foliaceous appearance, or incomplete embryos with a highly developed
cotyledonary part, with or without a root (Branton and Blake 1983; Blake 1990).
There are never any shoots on such formations, however, the rooting potential is
considerable, as with many monocotyledons (Hunault 1979).
The coconut is characterized by the slowness of different morphogenic
events. At least 2 or 3 years elapse between primary explant culturing and the first
ramets. This slowness makes in vitro experimentation difficult, insofar as a trial
is usually set up before all the results of previous trials are available.
In vitro control of somatic embryogenesis in palms is also made extremely
difficult by the effect of the highly determined tissue component, which prompted
Tisserat (1990) to say "I believe that the inherent explant character rather than
the medium composition is the determining factor in obtaining growth or
organogenesis from palm tissue culture". With coconut, the effect of the tissue
component is particularly marked; it complicates the search for medium factors
that are essential for the induction and expression of somatic embryogenesis.
In order to circumvent these difficulties, we created homogeneous callus
strains (Buffard-Morel et al. 1992; Verdeil et al. 1994); this strategy enabled us to
obtain a homogeneous tissue component for each strain, so that we could define
embryogenesis induction and expression conditions and adopt an approach
enabling the analysis of the problems encountered during somatic embryo-
genesis.
Systematic, histological checks detect the first signs of embryogenesis; they
are essential for monitoring tissue development on embryogenesis-inducing
media. In fact, an embryo is a well-defined structural entity, characterized by the
existence of a bipolar axis (opposite cauline and root meristems), with no
vascular connection to the mother tissues (Haccius and Philip 1979). It is also
defined by its ontogeny and it develops from embryogenic cells in accordance
with a precise sequence of events, revealing numerous similarities to zygotic
embryos. Given these definitions, only histological arguments can provide proof
that somatic embryogenesis has been achieved. The lack of a histological
approach at this stage of the process casts doubt on the organogenic or embryo-
genic origin of the formations described in certain publications on coconut
somatic embryogenesis. The formation of structures revealing morphological
similarities to a cotyledon is an indication, but it is not sufficient for stating that
a somatic embryo has been obtained (Blake 1990).
An important stage was reached by the ORSTOM-CIRAD team (Verdeil
et al. 1992a) which obtained reproducible ramets from different genotypes. These
encouraging results were obtained through an analytical approach (histological
study, study of culture medium composition) undertaken to provide a better

understanding of the causes behind halts or deviations in somatic embryogenesis
which are often seen in coconut.
2.2 Callogenesis
Coconut somatic embryogenesis is usually indirect and has to pass through the
callogenesis stage. In fact, direct embryogenesis on explants has only been
reported once (Raju et al. 1984) and can be considered exceptional. In this report,
embryoids are proposed to arise from vascular tissue but, as noted by Blake
(1989), this would be unusual as this is the area where root primordia normally
originate.
2.2.1 Conditions for Callus Induction
With coconut palm, like most monocotyledons the use of immature explants
containing meristematic cells is essential for callus induction. The second factor
determining successful callogenesis is the presence of a growth regulator with
high auxinic activity. This hormone is essential for the activation and division of
nondifferentiated cells from which the calli derive.
The most commonly used auxin is 2,4-D in varying concentrations, depend-
ing on the type of explant and the concentration of the activated charcoal added
to the culture medium. With 2.5 g/l charcoal, the 2,4-D concentrations used by
different workers are as follows: 22 mg/l to obtain nodular meristematic struc-
tures (calloids) formed from the extemallayers of floral meristems (Branton and
Blake 1983; Blake 1990); 26 to 44 mg/l to obtain calli at the cotyledonary
separation of immature zygotic embryos (Karunaratne and Periyapperuma
1989).
Other auxins have been tested, e.g., 2,4,5-T, which led to the formation of
nodular calli on inflorescence explants (Buffard-Morel et al. 1988). NAA and
IAA led to direct embryogenesis on leaf explants in the presence of a specific
Mg/K ratio (Raju et al. 1984).
Cytokinins can be added to the auxin during callogenesis: addition of 2 mg/l
kinetin to 50 mg/l 2,4-D and 1 g/l charcoal to obtain calli from endosperm
(Prakash Kumar et al. 1985); addition of 1 mg/12iP and BAP to 22 mg/12,4-D
and 2.5 g/l charcoal to obtain calli from immature inflorescences (Branton and
Blake 1984). Auxins (IAA, NAA, 2,4-D) and cytokinins (BAP, kinetin) were
tested alone or in combinations on coconut zygotic embryos, at concentrations
varying from 0.5 to 5 mg/l (Bhalla-Sarin et al. 1986). In particular, these authors
showed that IAA combined with alanine or aspartic acid, or a mixture of two
compounds favored callus formation (50%). The considerable activity of com-
bined IAA could be attributed to its stability with respect to peroxidases (Cohen
and Bandurski 1982).
2.2.2 Procedure for Obtaining Callogenesis in the Laboratory
The ORSTOM-CIRAD team receives material from the Ivory Coast (Marc
Delorme Station, IDEFOR, Cote d'Ivoire). This material is taken from adult
individuals (20-25 years old) selected for their agronomic value. The planting
material comes mainly from the Malayan Yellow Dwarf and Malayan Yellow
Dwarf x West African Tall (PB 121) and Cameroon Red Dwarf x West African
Tall (PB 111), hybrids created by IRHO-CIRAD.
Samples can be taken without causing injury to the apex, so that regrowth is
possible and has actually been observed. Two types of explants are used in the
laboratory: young non-chlorophyllous leaves and immature inflorescences.
Young leaves are protected by the petiole bases of older leaves; the surface of the
latter is disinfected with a sodium hypochlorite solution at 6% active chlorine.
The l-cm-Iong leaf fragments are placed abaxial side down on the culture
medium. Explants are taken from several young inflorescences; the outer spathe,
between 10 and 45 cm is disinfected for 10 min in the same disinfectant solution,
then rinsed in deionized sterile water. The two spathes are then removed asepti-
cally under a laminar flow hood prior to splitting the spikelets into 1- to 1.5-mm-
long fragments.
The standard medium consisted of Eeuwens Y3 mineral solution (Eeuwens
1976) following tests carried out by this author to develop a medium advanta-
geous to the growth of explants and calli ofleaf origin. We added the following
elements to the mineral solution: Morel and Wetmore's vitamins (1951), sucrose
at 40 gil and agar at 7.5 gil; 2,4-D was added at a rate between 30 and 80 mgll, due
to the great variability in sensitivity from one individual to another. The pH was
adjusted to 4.5 before autoclaving. The cultures were placed in the dark at 27 °C
± 1°C, as callus initiation and development take place without light (Buffard-
Morel et a11992; Verdeil et aI1994).
2.2.3 Callus Origin Sites and Histology of Calli
It takes a long time for calli to appear, depending on the type of explant used: 3
to 4 months on leaf explants, 5 to 6 months from floral meristems, the optimum
being between 7 and 9 months in both cases. The percentage of callus obtained
varied significantly according to the physiological state of the explant and the
auxin concentration in the culture medium. Histological studies pinpointed the
callus origin sites on both explant sources and provided a better understanding of
the mechanisms involved in their growth and transformation during transfers.
Neoformed tissues have two possible origins:
- A deep origin for calli obtained from leaf explants(Buffard-Morel et al. 1992)
and for those formed on the rachilla cross section of inflorescence fragments
(Verdeil et al. 1992b). In both cases, they were formed on undifferentiated
perivascular cells. This origin was identical to that described in oil palm
(Schwendiman et aI1988).
306 J.L. VerdeiI and J. Buffard-Morel
- A superficial origin observed for calli forming on floral areas; these calli
appeared later (more than 5 months after culturing). They were derived from
preexisting male floral meristems and were similar to the calloids described by
Branton and Blake (1984). According to these authors, such calloids con-
tained small cells with dense cytoplasm and, as we verified, comprised few
differentiated cells.
Formations of internal origin can easily lead to root formation if the auxin
concentration is too low. In fact, in coconut as in many monocots (Hunault
1979), tissues are prone to form organogenic primary meristems of root
meristem type. Root expression can be inhibited, especially on leaf explants, by
using high 2,4-D concentrations.
Regardless of the origin of tissue both types of calli gave rise to compact,
slow-growing nodular calli; their growth and multiplication were ensured by a
peripheral meristematic zone organized throughout as a cambium like meristem-
atic zone (Buffard-Morel et al. 1992). This layer was similar to an enlarged
meristem, which ensures thickness growth throughout the entire plant, and it had
a unidirectional function, leading to inward differentiation of parenchyma and
vascular elements (tracheids) in both the callus and the stem.
2.2.4 Callus Subculturing and Obtaining Homogeneous Strains
Calli were isolated from the sixth month of culturing onwards. They were placed
on media containing activated charcoal and 2,4-D. The 2,4-D concentration was
so chosen as to optimize the callus growth of each strain. On both explant
sources, callus growth and multiplication, ensured by the peripheral meristem-
atic zone, may lead to the formation of calli with a granular texture. By
multiplying them in the presence of an optimum 2,4-D concentration, we were
able to select homogeneous callus strains. These calli are being used to study
different factors that might favor the induction of embryogenesis particularly the
search for a hormone balance leading to the formation of whole embryos with a
cauline apex.
Element consumption from the callus growth medium was determined for
different strains. For all the strains studied, ammonium, calcium, and magne-
sium consumption was regular throughout the entire culturing period. However,
sucrose and sulfate were consumed during the latent phase and at the start of the
exponential phase of growth; phosphate and nitrate were only used in the second
part of the exponential phase. The ionic balance was maintained by potassium
and chlorine throughout the culturing period (Verdeil et al. 1993). This study
showed the homogeneity from one strain to the other and our interest in using
these calli to study the induction of embryogenesis.
2.3 Induction and Expression of Embryogenesis
2.3.1 Induction 0/ Embryogenesis/rom Calli
With coconut palm, as with most species, the auxin concentration is the deter-
mining factor for the induction of somatic embryogenesis. In 1990, Blake
reported that a high auxin concentration was required to obtain proembryoids:
"The calloids form nodular meristematic structures in the superficial layers when
sub-cultured on the high level of auxin and early embryoids can be identified."
The presence of activated charcoal is considered to be an essential element
in the induction of embryogenesis (Blake and Eeuwens 1982; Pannetier and
Buft'ard-Morel 1986; Tisserat 1990) and in initial embryo development (N wanko
and Krikorian 1983; Brackpool et al. 1986). Once again, its beneficial action
would seem to be due to reversible adsorption of the auxin and its slow and
gradual release (Brackpool et al. 1986). On the other hand, its use makes it
difficult to control the free 2,4-D level in the medium, and it is unfortunate that
none of the teams working on this subject have produced precise data. This may
explain the difficulties encountered in obtaining embryogenesis and the lack of
reproducible responses.
Embryogenesis on calli isolated from explants is nonsynchronous (from 2 to
24 months after callus isolation from the explant) and it is induced on less than
10% of calli. These characteristics complicate the study of the phenomenon and
explain the lack of precise information on the induction of embryogenesis in
coconut.
2.3.2 Induction 0/ Embryogenesis/rom Homogeneous Callus Strains
In order to bypass the above-mentioned difficulties, the ORSTOM -CIRAD team

worked on homogeneous calli strains previously described. Callus strains were
screened on the basis of their embryogenic competence. This work led to
synchronous induction of embryogenesis in some selected strains.
A histological study of embryogenic calli revealed two embryogenesis path-
ways:
1. Multicellular pathway (Fig. 1): the embryogenic capacity of calli can be recog-
nized due to the existence of meristematic and epidermized structures ob-
tained with 2 gil charcoal and low 2,4-D concentrations (40 to 60 mgll)
(Verdeil et al. 1992b, 1994). Histologically, the first stage of development of
these structures is characterized by the breaking up of the cambium-like zone
and very active cell divisions in this discontinuous zone (Fig. 2) led to the
formation ofmeristematic nodules followed by their epidermization (Fig. 3).
They give rise to embryo-type structures. When the auxin concentration is too
low, they often lead to the formation of incomplete or abnormal embryonic
structures (haustorium only with or without a root pole, foliaceous type of
embryo) (Branton and Blake 1983; Brackpool et al. 1986).
308 1.L. Verdeil and 1. Buffard-Morel
Fig. I. Embryogenic epidermized structures obtained by a multicellular pathway. Scale = 2.5 mm

(Buffard-Morel et al. 1992)
Fig. 2. Histological cross section in an embryogenic callus giving rise to an embryogenic structure after
cambium-like zone fragmentation. Scale = 250 ~m. BCZ Break up of the cambium-like zone; CZ
cambium-like zone; MA meristematic areas giving rise to embryogenic structures. (Buffard-Morel
et a1.I992)
2. Unicellular origin (Fig. 4): this leads to the appearance and individualization
of embryogenic cells (Fig. 5) identical to those described for oil palm
(Schwendiman et al. 1988). This second pathway, obtained in the presence of
2 to 3 gil charcoal and a high 2,4-D concentration (80 to 120 mgll, depending
on the strain) (Verdeil et al. 1992b) leads to the formation of typical
proembryos (Fig.6) which have the characteristics of the initial stages of
zygotic embryogenesis as described by Haccius and Philip (1979). These cells
Fig. 3. Histological cross section of an embryogenic structure initiated from a multicellular pathway.
Scale = 100 ~m. CZ Cambium-like zone, ES embryogenic structure (Verdeil et al. 1992b)
Fig. 4. Induction of embryogenic structures on granular callus from unicellular origin. S cale = 2.6 mm.
ES Embryogenic structure (Verdeil et al. 1994)
are characterized by dense cytoplasm, a high nucleocytoplasmic ratio, the

existence of a single, voluminous nucleolus, signifying substantial RNA
synthesis, and the presence of starch and protein reserves (Fig.S). They
become individualized from the rest of the callus through thickening of the cell
wall (Lu and Vasil 1985; Smith 1986; Williams and Maheswaran 1986;
Schwendiman et al. 1988).
Fig. 5. Histological cross section of an embryogenic callus obtained by a unicellular pathway.

Scale = 251lm. PE Proembryo; ECembryogenic cell (Verdeil et al. 1992b)
Fig. 6. Individualization of a proembryo in a callus through external cell wall thickening. Scale =
251lm. N Nucleus; S starch. N nucleolus; TWthickened wall (Verdeil et al. 1994)
F or each embryogenesis pathway, a kinetic comparison of the callus nutrient

requirements of the multiplication and embryogenesis media (the two medium
formulations differ only in 2,4-D concentration) revealed greater ammonium,
calcium, and magnesium absorption per gram dry matter for embryogenesis,
accompanied by an increase in protein synthesis, leading to protein accumulation
in embryogenic cells (Verdeil et al. 1993).
This kind of study has largely contributed to a better understanding of

somatic embryogenesis factors in coconut. It therefore provides an opportunity
to control it better by acting upon medium composition and culture transfer
frequency.
2.3.3 Expression and Maturation of Embryogenesis
According to most authors, the essential factor in the expression of embryogen-

esis is the gradual drop in 2,4-0 concentration (Karunaratne 1989; Blake 1990).
In our laboratory, trials have shown that, although the drop in 2,4-0 is a critical
factor (5 to 10 mg every 4 months according to Buffard-Morel et a1.l992), the
initial concentration from which this gradual reduction occurs nevertheless
remains the essential factor, it ensures a good start of embryogenesis which will
likely result in the formation of whole embryos.
If the drop in the auxin concentration is too rapid, the maturation of
embryogenic structures usually leads to incomplete or abnormal forms: incom-
plete embryogenesis characterized by the development of the haustorium and
sometimes the root; multiplication and formation of a clump of fused embryoids
whose development is inhibited.
Frequent deviations in embryonic morphogenesis could be linked to a
hormone imbalance (Dublin et al. 1991). The effect of cytokinins has been
studied by various workers, but their role has not been clearly determined
(Karunaratne and Periyapperuma 1989; Blake 1990). In our laboratory, embryo
maturation began when the 2,4-0 concentration was gradually reduced, and it is
clear that further maturation was only achieved on a medium containing BAP
Fig.7. Embryo clump with emission of the first leaf scale on one of them. Scale = 10 J.lm. EEmbryo;
S shoot (Buffard-Morel et al. I 992)
312 J.L. Verdeil and J. Buffard-More1
Fig. 8. Conversion of an isolated somatic embryo. Scale =4 mm. S shoot, R root; H haustorium
(Verdeil unpub!.)
(Verdeil et al. 1994), (Figs.7 and 8). Thus, a strictly defined sequence of events is
required to obtain complete embryogenesis:
- haustorium elongation obtained throughout a decrease in 2,4-D concentra-
tion;
- cauline pole differentiation in the presence of a cytokinin;
- root pole differentiation.
2.3.4 Embryoid Conversion Leading to Ramet Formation
Conversion of these embryos, with emerging shoots is carried out on an auxin-

free medium; a cotyledonary split forms at the base of the embryo, then a shoot
appears. Establishment of the cauline meristem precedes root pole differenti-
ation, the root may develop, indicating the presence of a functional bipolar axis
in the embryonic structures (Fig. 8). However, in most cases, addition ofNAA is
necessary to obtain roots, as observed in zygotic embryos (Assy Bah et al. 1989).
Histological cross sections along the cotyledonary split revealed a well-structured
cauline meristem with a meristematic dome surrounded by leaf primordia
(Fig. 9). The haustoria I tissue located opposite the stem meristem, contains
starch reserves. The location of starch grains, which are a clear indicator of
embryo polarity, is identical to that found in a zygotic embryo. The frequent
absence of a cauline meristem could be linked to the prolonged use of auxin
without cytokinin treatments. This phenomenon has also been observed on
somatic embryos of carrot and soybean in particular (Lazzeri et al. 1987) after
exposure to prolonged auxin treatments.
Fig. 9. Histological cross section of a somatic embryo (just below the notch) showing the existence of
=
a well-structured ca uline apex. Scale 500 f.Im . LP Leaf primordia; MD meristematic dome; eN
cotyledonary notch; S starch (Verdeil et al. 1994)
3 Discussion and Conclusion
F or the coconut, as for many palms, somatic embryogenesis is indirect and has to
pass through a callogenesis phase. Initiated calli cannot be considered to be true
calli in the strict meaning of the word, in that the meristematic cells which they
contain are organized in a cambium-like zone. This layer, as with any cambium-
type layer, is essentially histogenetic. Its perfectly determined function (mitosis
orientation) gives rise to the inward development of a parenchyma and vascular
tissue in the callus. Cellular differentiation makes it difficult to induce embryo-
genesis in coconut calli. Acquisition of embryogenesis competence requires prior
disorganization of the structure, and functioning of this meristematic zone,
under the effect of a critical auxin concentration. This new level of disorganiza-
tion allowed changes in the morphogenic capacity of surviving cells, which can
acquire the histological characters of embryogenic cells and lead to either
unicellular or multicellular embryogenesis initiation.
Embryogenic cells require appropriate conditions conducive to their devel-
opment into complete embryos. Nascent embryonic structures have not yet been
irreversibly determined. Because of their plasticity, they are extremely vulnerable
to physical and chemical stresses leading to abnormal or incomplete embryos
(Tulecke 1987; Carman 1990; Dublin et al. 1991). Auxin concentration also plays
a paramount role in embryo development; a hormone imbalance leads to
abnormal or incomplete structures, whose evolution is irreversible. Our findings
on somatic embryogenesis in coconut show that a decrease in 2,4-0 con-
centration followed by addition of a cytokinin like BAP is a key to complete
differentiation of bipolar embryo.
Fig. 10. Shoots obtained from isolated somatic embryos of different genotypes. Scale = 10 mm
(Verdeil et al. 1994)
PlantIets have been obtained in our laboratory with several genotypes;

however, the number of ramets (Fig. 10) regenerated per clone remains low, due
to the lack of an intense mUltiplication phase. The main goal of our future work
will be to determine the conditions required to achieve adventitious embryo-
genesis. Successful multiplication of the somatic embryos obtained should
make it possible to satisfy ramet mass production requirements.
4 Protocol
The procedure developed by the ORSTOM/CIRAO team comprises the following steps:
I. Callogenesis from immature leaves and inflorescence explants collected without Injury to
the caulinar apex of the mature trees. The standard medium consisted of Eeuwens (1976),
mineral solution. Morel and Wetmore (1951) vitamins, 40 gil sucrose, and 7.5 gil agar agar, to
which 30 to 80 mg/I, 2.4-0 was added, along with 2 or 3 gil activated charcoal, in order to limit
tissue browning.
2. Induction of embryogenesis from calli isolated from the sixth month of culturing onward. Calli
were placed on media containing 2 to 3 g/I activated charcoal and 2,4-0 at low concentrations (40
to 60 mg/I), which can lead to embryogenesis of multicellular origin, and at high concentrations
(80 to 120 mg/I), which are used for initiation of embryogenesis of unicellular origin.
3. Embryo maturation/conversion and shoot formation. Embryo maturation was obtained by
gradually reducing the 2,4-0 concentration and adding BAP. Shoot development from isolated
embryos takes place under light on a hormone-free medium with activated charcoal. In the majority
of cases, root development requires treatment with an auxin: 20 mgll NAA and 2gll activated
charcoal.
Acknowledgments. This work is supported by CIRAD/CP and ORSTOM. We would like to thank the
M. Delorme research station (Port Bouet, IDEFOR Cote D'lvoire) for supplying the plant material
and for the valuable help. We are glad to acknowledge the help of those who contributed to the work
described: C. Pannetier, S. Dussert, C. Magnaval, C. Huet, and F. Grosdemange.
References
Assy Bah B, Durand-Gasselin T, Engelmann F, Pannetier C (1989) Culture in vitro d'embryons

zygotiques de cocotier (Cocos nucifera L.). Methode, revisee et simplifiee, d' obtention de plants de
cocotiers transferables au champ. Oleagineux 44,11: 515-523
Berger KG, Ong SH (1985) The industrial uses of palm and coconut oils. Oleagineux 40,12: 6l3-624
Bhalla-Sarin N, Bagga S, Sopory SK, Guha-Mukherjee S (1986) Induction and differentiation of callus
from embryo of Cocos nucifera L. by IAA-conjugates. Plant Cell Rep 5: 322-324
Blake J (1989) Coconut tissue culture research at London University's unit for advanced propagation
systems, Wye College. COCOTECH, 26th Meet Bangkok, Thailand.
Blake J (1990) Coconut (Cocos nucifera L.): Micropropagation. In: Bajaj YPS (ed) Biotechnology in
agriculture and forestry, vol 10. Legumes and oilseed crops I. Springer, Berlin Heidelberg, New
York, pp 538-554
Blake J, Eeuwens CJ (1978) Inflorescence tissue as source material for vegetative propagation of
coconut palm. In: Proc Int Conf Cocoa/coconuts, Kuala Lumpur, pp 549-556
Blake J, Eeuwens CJ (1982) Culture of coconut palm tissues with a view to vegetative propagation. In:
Rao AN (ed) Tissue culture of economically important plants. Proc Int Symp Costed ANBS,
Singapore, 145-148
Boutin D (1990) Huileries de coco villageoises - utilisation de I'amande fraiche. Oleagineux 45, 7:
327-331
Brackpool AL, Branton RL, Blake J (1986) Regeneration in palms. In: Vasil IK (ed) Cell culture and
somatic cell genetics of plants, vol 3. Academic Press, Orlando, pp 207-222
Branton AL, Blake J (1983) Development of organized structures in callus derived from explants of
Cocos nucifera L. Ann Bot 52: 673-678
Branton RL, Blake J (1984) Clonal propagation of coconut palm. Proc Int Conf Cocoa/coconuts,
Kuala Lumpur, Paper 46, pp 1-9
Buffard-Morel J, Verdeil JL, Pannetier C (1988) Vegetative propagation of coconut palm through
somatic embryogenesis, obtention ofplantIet from leaf explant. 8th Int Symp Biotechnology, Paris.
Abstr, 177 pp
Buffard-Morel J, Verdeil JL, Pannetier C (1992) Embryogenese somatique du cocotier (Cocos nucifera
L.) Ii partir de tissus foliaires: etude histologique. Can J Bot 70: 735-741
Carman JG (1990) Embryogenic cells in plant tissue cultures: occurrence and behavior. In Vitro Cell
Dev Bioi 26: 746--753
Cohen JD, Bandurski RS (1982) Chemistry and physiology of the bound auxins. Ann Rev Plant
Physiol 33: 403--430
Das PK (1985) Tendances de la production et du commerce de I'huile de coco Ii travers Ie monde.
Oleagineux 40, 2: 85-93
Dublin P, Enjalric F, Lardet L, Carron MP, Trolinder N, Pannetier C (1991) Estate crops. In: Debergh
PC, Zimmerman RH (eds) Micropropagation technology and application. Kluwer Dordrecht, pp
337-361
Ebert A, Taylor HF (1990) Assessment of the changes of2,4-dichlorophenoxyacetic acid concentra-
tions in plant tissue culture media in the presence of activated charcoal. Plant Cell Tissue Organ
Cult 20: 165-172
Eeuwens CJ (1976) Mineral requirements for growth and callus initiation of tissue explants excised
from mature coconut palms (Cocos nucifera) and date palms (Phoenix dactylifera) cultured in vitro.
Physiol Plant 36: 23-28
Fridborg G, Eriksson T (1975) Effects of activated charcoal on growth and morphogenesis in cell
cultures. Physiol Plant 34: 306--308
316 1.L. Verdeil and 1. Buffard-Morel
George EF, Sherrington PD (eds) (1984) Plant propagation by tissue culture. Handbook of directory
and commerciallahoratories. Exegetics Eversley, Busingstoke, pp 387-389
Graille 1, Lozano P, Pioch D, Geneste P (1985) Essais d'alcoolyse d'huiles vegetales avec des
catalyseurs naturels pour la production de carburants diesels. Oleagineux 40,5: 271-276
Gupta PK, Kendurkar SV, Kulkarni VM, Shigurkar MV, Mascarenhas AF (1984) Somatic embryo-
genesis and plants from zygotic embryos of coconut (Cocos nucifera L.). Plant Cell Rep 3, (6):
222-225
Gurtler lL, Dronne Y (1992) Evolution du marche mondial des Oleagineux. Oleagineux 47,3: 151-153
Haccius B, Philip Vl (1979) Embryo development in Cocos nucifera L.: a critical contribution to a
general understanding of palm embryogenesis. Plant Syst Evol132: 91-106
Hunault G (1979) Recherches sur Ie comportement des fragments d'organes et des tis sus de
Monocotyledones cultives in vitro. V. Discussion. Rev Cytol BioI Veget Bot 2: 259-287
Hyman EL (1990) The choice of technology and scale in coconut processing in the Philippines,
Oleagineux 45, (6): 279-295
lesty lHF, Francis D (1992) Cellular responses ofleaf explants of Cocos nucifera L. in vitro. Plant Cell
Tissue Organ Cult 28: 235-244
Karunaratne SM (1989) Applications of tissue culture techniques for vegetative propagation of
coconut. COCOTECH, 24th Meet, Bangkok (Thailand)
Karunaratne S, Periyapperuma K (1989) Culture of immature embryos of coconut, Cocos nucifera L.:
callus proliferation and somatic embryogenesis. Plant Sci 62: 247-253
Lazzeri PA, Hildebrand DR, Collins GB (1987) Soybean somatic embryogenesis: effects of hormones
and culture manipulations. Plant Cell Tissue Organ Cult 10: 197-208
Lu CY, Vasil IK (1985) Histology of somatic embryogenesis in Panicum maximum (guinea grass). Am
1 Bot 72: 1908-1913
Montet D, Ratomahenina R, Laborde 1M, Pina M, Graille 1, Galzy P (1985) Production de proteines
d'organismes unicellulaire it partir de pates de neutralisation d'origine industrielle. Oleagineux 40
(10): 505-509
Morel G, Wetmore RH (1951) Tissue culture ofmonocotyledones. Am 1 Bot 38: 138-140
Nwanko BA, Krikorian AD (1986) Morphogenetic potential of embryo and seedling derived callus of
Elaeis guineensis lacq var pisifera Becc. Ann Bot 51: 65-76
Pannetier C, Buffard-Morel 1 (1982) Production of somatic embryos from leaf tissues of coconut,
Cocos nucifera L. In: Fujiwara A (ed) Plant Tissue and Cell Culture 1982, Maruzen, Tokyo, pp
755-756
Pannetier C, Buffard-Morel 1 (1986) Coconut palm (Cocos nucifera L.). In: Bajaj YPS (ed) Biotechnol-
ogy in agriculture and forestry, vol I. Trees I. Springer, Berlin Heidelberg New York, pp 430-450
Prakash Kumar P, Raju CR, Chandramohan M, Iyer RD (1985) Induction and maintenance of friable
callus from the cellular endosperm of Cocos nucifera L. Plant Sci 40: 203-207
Raju CR, Prakash Kumar P, Chandramohan M, Iyer RD (1984) Coconut plantlets from leaf tissue
cultures. Plantation Crops 12(1): 75-81
Reyes-lael M (1985) Coconut varieties for sugar production. Coconuts Today 3: 30-31
Reynolds IF, Murashige T (1979) Asexual embryogenesis in callus cultures of palms. In Vitro 15:
383-387
Rillo EP (1993) Current status of coconut embryo and tissue culture in the Philippines. In: Nair MK
et al. (eds) Advances in coconut research and development. Kasaragod, India, 26-2911111991, pp
161-167
Schwendiman 1, Pannetier C, Michaux-Ferriere N (1988) Histology of somatic embryogenesis from
leaf explants of the oil palm Elaeis guineensis. Ann Bot 62: 43-52
Smith BG (1986) Tissue culture of Cocos nucifera - biochemical changes preceding embryogenesis. In:
Pushparajah E, Chew Poh Soon (eds), Cocoa and coconuts: progress and outlook. Kuala Lumpur
Inc Soc Plant, Kuala Lumpur, pp 781-786
Tisserat B (1990) Palm tissue culture. US Department of Agriculture, Agricultural Research Service,
Pasadena, pp 55-60
Tulecke W (1987) Somatic embryogenesis in woody perennials. In: Bonga 1M, Durzan Dl (eds) Cell
and tissue culture in forestry, vol 2. Specific principles and methods: Growth and developments,
Martinus Nijhoff, Dordrecht, pp 61-85
Verdeil JL, BufIard-Morel J, Pannetier C (1990) Somatic embryogenesis of coconut from leaf and
inflorescence tissue. Difficulties encountered and prospects. COCOTECH 24th Meet Bangkok,
Thailand
Verdeil JL, Huet C, Grosdemange F, Rival A, BufIard-Morel J (1992a) Somatic embryogenesis in
coconut (Cocos nucifera L.): obtention of several ramet clones. Oleagineux 47,7: 465-469
Verdeil JL, Dussert S, Grosdemange F, Huet C, Rival A, BufIard-Morel J (1992) Initiation of somatic
embryogenesis in coconut (Cocos nucifera L.): histological, hormonal and nutritional aspects. In
Proc 13th EUCARPIA Congr, Reproductive and plant breeding, Angers, France, 6-10/7/92 pp
401-402
Verdeil JL, BufIard-Morel J, Rival A, Grosdemange F, Huet C, Pannetier C (1993) Coconut clones
through somatic embryogenesis. In: Nair MK et al. (eds) Advances in coconut research and
development, Kasaragod, India, 26-29/1111991, pp 173-179
Verdeil JL, Huet C, Grosdemange F, BufIard-More1 J (1994) Plant regeneration from cultured
immature inflorescences of coconut (Cocos nucifera L.): evidence for somatic embryogenesis. Plant
Cell Rep 13: 218-221
Williams E, Maheswaran G (1986) Somatic embryogenesis: factors influencing coordinated behavior
of cells as an embryogenic group. Ann Bot 57: 443-462
111.6 Somatic Embryogenesis in Hazelnut
(Corylus Species)
B. BERROS, M. REY, C. DiAz-SALA, M. ALBUERNE, and R. RODRiGUEZ l
1 Introduction
The common names of hazelnut or filbert include species belonging to the genus
Corylus (tribe Coryleae, family Betulaceae). The importance and distribution of
this plant have been dealt with in detail by Rodriguez et al. (1989) in Volume 5
of this Series (Bajaj 1989). The genus Corylus comprises about 15 species.
Filbert multiplies through rooting suckers, since it is one of the fruit tree
species most prone to suckering. It produces several suckers every year which
develop from the roots in the vicinity of the trunk (Alvarez Requejo 1965;
Tombesi and Standardi 1970; Paglietta 1978; Tombesi 1985). However, more
desirable characteristics ate maintained by means of cuttings (Howard 1968;
Roversi 1969; Germain et al. 1974; Piskornik et al. 1982a,b; Rodriguez 1983).
Furthermore, micropropagation of filbert is an attractive method of alternative
propagation, and it will enable the production of cultivars at a commercial scale
(Diaz-Sala et al. 1990). For this purpose, induction of somatic embryogenesis
would be valuable.
2.1 Material and Methods of Sterilization
Induction of embryogenesis was attempted with plant materials at different

developmental phases:
1. Embryogenic phase: embryogenic and related tissues; endosperm, immature
and mature embryos, cotyledon and carpellary portions.
2. Juvenile phase: cotyledon nodes, proximal cotyledon portions, shoot tips, and
intact seedlings.
3. Mature phase: leaves and isolated shoot tips from in vitro mature tissues
during the shoot-bud proliferation step.
I Laboratorio de Fisiologia Vegetal, Departamento de Biologia, Organismos y Sistemas, Facultad de

Biologia, Universidad de Oviedo, 33006 Oviedo, Spain

Somatic Embryogenesis in Hazelnut (Corylus Species) 319
Surface sterilization is done with ethanol (50-90%) immersion for 1-5 min.
Explants were then repeatedly washed with distilled sterile water and afterwards
tissues were immersed in sodium or calcium hypochlorite at several concentra-
tions (1-2.5%) for various times (5-20 min). Modifications norma1ly depend on
the developmental and phytopathological state of the tissues.
A few drops of detergent (Tween-20 or Alconox) can be added to the
hypochlorite solution in order to increase the contact between the sterilizing
agent and the plant tissues, any surplus is further eliminated by washing the seeds
several times with sterile distilled water. Afterwards, the seeds are ready to
culture, having in a1l cases, less than 5-20% contamination.
When embryos are used, indirect surface sterilization may be applied, as
reported by Radojevic et al. (1975). Immature seeds are immersed in a 5%
calcium hypochlorite solution with the subsequent dissection of embryos with no
risk of contamination.
2.2 Phytohormones and Culture Media
The regulation of somatic embryogenesis in hazelnut cotyledon nodes was

investigated (Perez et al. 1986), showing that IBA (3-indolbutyric acid) seems to
promote embryo initiation. On the other hand, (6-benzylaminopurine), which
promoted development during the first subculture, acts as the main agent in
restoring the polarized development. Thus, the addition of BAP and IBA in
various concentrations and combinations in two consecutive cultures facilitates
embryogenic induction in the primary explants. Based on these results, the
induction and proliferative media, named P, are based on two consecutive steps:
first, 20 days in the presence of 5JlM IBA plus 0.5JlM BAP, fo1lowed by transfer
to 5JlM BAP plus 0.5 JlM IBA.
It is important to emphasize that the most suitable auxin is IBA. Although
2,4-D (2,4-dichlorophenoxyacetic acid) may be used to induce embryogenesis, a
high concentration of this phytohormone appears to inhibit subsequent develop-
mental steps at the globular and heart stages (Radojevic et al. 1975). However,
the use of a lower 2,4-D concentration may enhance the development of these
stages (Radojevic et al. 1983).
In regard to cytokinin effects and the fact that BAP is the most effective
cytokinin, embryo induction was also obtained in the presence of Ad (adenine,
0.7-7 JlM).
Despite the variability of the induction media, when 2,4-D is applied, ca1lus-
mediated embryogenesis can be induced. Indirect induction of embryogenesis is
readily achieved when primary explants are cultured for 20 days on R media
composed of 2,4-D (4.5 JlM) and kinetin (Kin, 4.5 JlM).
The evolution of embryogenesis can be determined by the time of culture in
the presence of 2,4-D, which is affected principa1ly by the establishment of the
polarity. Conversely, when filbert embryogenic tissues were transferred to a 2,4-
D-free medium, very low plantlet regeneration was obtained. On the other hand,
it can be concluded that direct embryogenesis can be achieved in the presence of
320 B. Berros et al.
cytokinins which reduced the manifestation time when other regulators were
combined.
Alterna tively, embryogenesis can be induced by using T media, developed by
Tulecke and McGranahan (1985) for the induction of somatic embryogenesis
from walnut cotyledons. MS medium plus 5 IlM BAP, 9 IlM Kin, and 0.05 IlM
IBA also gave rise to direct embryogenesis on several hazelnut explants (see
below).
In all the experiments, 3% sucrose was added as the carbon source and the
medium pH was adjusted to 5.8 ± 0.2 with NaOH or HCI; the media were
solidified with 0.7 % bacteriological agar. Cultures were maintained in a growth
room at 25 ± 2°C under a 16-h photoperiod or in complete darkness. The
light source was fluorescent tubes of cool white light with an irradiance of
15.6 IlE m-2s- 1•
2.3 Various Approaches to Somatic Embryogenesis
Somatic embryogenesis was first induced on callus tissue cultures of Corylus

avellana L. isolated from immature embryos (Radojevic et al. 1975). After several
months of culture on a modified MS medium supplemented with Kin (6-furfuryl-
aminopurine) and 2,4-D (4.5 IlM each), proembryogenic structures emerged.
However, further steps after the globular stages were not observed. Radojevic
et al. (1983) also reported the effect of a low concentration of 2,4-D (0.05 IlM)
which increased both the fresh weight and number of embryos. A decreased
concentration or complete omission of auxins during two passages and the
addition of GA3 induced plantlet development.
Recently (Table 1) efforts have been made in our laboratory to verify the
embryogenic ability of differently aged tissues. All the experiments were carried
out using P (Perez et al. 1983), T (Tulecke and McGranahan 1985), or R
(Radojevic et al. 1975) media during the induction and proliferation phases.
Details of further development and embryo germination will be given later,
depending on the responses achieved (Fig. 1).
Embryogenic Plant Material. Desirable embryogenesis induction was easily

obtained with tissues taken from the first stages of fruit development, 150-240
days after pollination. The most plastic tissues were those derived from immature
fruits (240-270 days old).
Results obtained with tissues collected during 1991 and 1992 indicated that
induction was not constant, probably due to physiological changes and deriva-
tives ontogenic differences.
Endosperm, after a short induction period (15 or 20 days on P or T media),
gave rise to direct embryogenesis in about 5% of the tissues; longer culture
periods increased the response to 20%. During embryogenesis, a cluster of
globular and cotyledonary leaves emerges. In endosperm, which is a triploid
tissue, there were no visible macromorphological differences compared with
diploid tissues.
Table 1. Somatic embryogenesis in hazelnut CI'l
0
S
Cultivar Age Explant Medium Plant growth Culture Period Success Reference n'
-""
regulators ()lM) condition passages tTl
(days) S
c:r
~
0
(JQ
Inmature Embryos MS Kin 4.5, 2,4-D 4.5 Agar 0.8-1% 42 Globular and Radojevic (1)
p
(2mm) 12 h light cotyledonary et al. (1975)
25 ± 1°C stages ~.
Mature, Cotyledonary K(h) IBA 5/0.5 Agar 0.7% 20 Embryogenesis Perez et al. S·
5 week old nodes from BAP 0.5/5 18 h light 3 and plantlets (J 983, 1986) ::c:
~
plantlets stratified seeds 25 ± 2°C, pH 5.6 !!.
p
Mature Seedlings from Vz K(h) Ad 0.7-70 Agar 0.7% 45 Embryogenesis Rodriguez S
unstratified seeds 18 hlight and plantlets et al. (1985)
25 ± 2°C, pH 5.6
g
~
Mature Seedlings from y, K(h) None Agar 0.7% 45 Embryogenesis Rodriguez ~
unstratified 18 h light et al. (1985) '"CI'l
'0
seeds 25 ± 2°C, pH 4.5 (1)
(")
Mature Deembryonated y, K(h) IBA50 Agar 0.7% Globular stages Gonzalez

cotyledons Kin 4 18 h light (1990)
!
25 ± 2°C, pH 5.5
Juvenile Meristems MS BAP25 Agar 0.7% 50 Cotyledonary Fernandez
AlA 0.5 18 h light stages et al. (1990)
GA 3 O.02 25 ± 2°C, pH 5.8
Gironell Mature, Embryos or MS Kin 4.5, 2,4-D 4.5,
Negret from pruned proximal IBA 5/0.5, BAP Agar 0.8% 30 Primary embryo- Reyetal.
or un pruned cotyledonary 0.5/5, BAP 5, 18 h light 3 genesis, embryo- (1991)
trees portions Kin 9, IBA 0.05 25 ± 2°C, pH5.8 genic clusters
and plantlets
Gironell Immature, Embryos or MS Kin 4.5, 2,4-D 4.5, Agar 0.8% 30 Primary embryo- Berros
Negret from pruned proximal IBA 5/0.5, BAP 18 h light 3 genesis, embryo- et al. (1992)
orunpruned cotyledonary 0.5/5, BAP 5, 25 ± 2°C, pH 5.8 genic clusters
trees portions Kin 9, IBA 0.05 and plantlets
<.;.>
~
embryonic axes or cotyedon portions
~ ;"du,I;",,;" P " T m.d;,

after 20 or 40 days
b.-O-M-A-TI-C-E-MS-R-Y-os----..I
/
isolated embryos
development in 6 0
...
day~
cotyledonary leaves
proliferation in P and T media
ISOMA TIC EMBRYOS

development in 20 days
/ reiterative embryogenesis
/. development in proliferation media
treatment with ABA and MS

IPLANTLET I
Fig. 1. Proliferation and germination scheme
Immature, isolated embryos are easily induced to yield an embryogenic

cluster in 40% of the explants; most of the responses are obtained after short
culture periods. Conversely, when isolated cotyledon portions are used, only the
proximal ones react to the phytohormonal combinations of P or T media after
longer culture periods. Proximal portions after 20 days or distal portions after
longer periods of induction never give rise to the desired responses.
Direct embryogenesis starts with the swelling of the tissues. First, cotyledon-
ary leaves emerge at separate points; later, multistage clusters are observed. It is
important to emphasize that during embryogenic cluster development and
proliferation, other kinds of organogenic responses, such as adventitious root
development or precocious germination, take place.
When R medium is used, only very few, indirect responses were induced in all
the assayed tissues.
During maturation of the fruit, an important decrease in embryogenic
responses is observed, as discussed below (Rey et al. 1991; Berros et al. 1992).
Juvenile Plant Material. It was observed that even mature seeds stored without
stratification over 6 months showed spontaneous embryogenic responses during
germination (Rodriguez et al. 1985). Since this type of embryogenic manifesta-
tion is very low, no more than 5% of the seedlings, posterior proliferation, or
plantlet development was as efficient as in other systems. Higher responses (20%)
were induced in the presence of Ad (0.7 J.lM) or on germination media with a pH
of 4.5. Loci of embryo induction are mainly at the cotyledon nodes and at the
apical bud levels, which are the most responsive tissues at this stage. The main
Fig. 2. Embryogenic cell cluster located on the cotyledons of a young plantlet (x 450)
macromorphological features of this response are clusters bipolar structures in

which their own autonomy was histologically confirmed. The isolation of these
structures and further culture on fresh Ad-containing media allow the initiation
of an embryogenic program mediated by embryogenic callus formation. Trans-
ferring the embryogenic clumps onto a basal medium furthers the development
into plantlets.
High embryogenic ability was also confirmed in juvenile tissues derived from
5-week-old plantlets. Proximal cotyledon portions were cultured on half-
strength K(h) medium containing 50 11M IBA plus 4.5 11M Kin (Fig. 2), giving
rise to internal embryogenic groups of cells; externally, a large number of roots
developed (Gonzalez 1990).
Cotyledon nodes excised from young seedling grown for 5 weeks on K(h)
medium were also used to induce embryo formation. Embryogenesis was
achieved in 60% of the inocula over two 20-day subculture steps in the presence
of P medium (Perez et al. 1983). Subsequent proliferation was successfully
maintained during five subcultures on K(h) basal medium and in an unlimited
number of subcultures in the presence of 0.5 J.lM BAP. Plant regeneration
(50-55%) was easily achieved by transferring to a basal K(h) medium after
10- 20 days. These results emphasize the great morphological heterogeneity
of the somatic embryos obtained during different steps of the process (Perez
et al. 1986).
Cotyledonary stages were also induced on isolated shoot tips of juvenile
plantlets after 20 to 50 days on MS medium supplemented with 25 11M BAP plus
0.5 J.lM IAA, and 0.02 11M GA 3.
w
Table 2. Determinative macromorphological characteristics of the ontogenic stage of hazelnut fruit. (Berros 1991)
~
Date 29 May 1991 22 June 1991 16 July 1991 2 August 1991 20 August 1991 8 September 1991
Fruit size (mm) 3-6 6-8 8-10 10-12 12 12

Bracts Greenish, joined up to the middle Basally attached to the fruit Easily separable from the pericarp
of the prefruit
Carpellary Carpellary Not lignified Lignified Lignified and easily separable from the pericarp
tissue/peri carp tissue
Carpellary Carpellary Spongy Reabsorbed
tissue/mesocarp tissue
Carpellary Carpellary Pulpy and thick Thin and attached to the cotyledon
tissue/testa tissue
Embryonic axis Proembryogenic Cellular endosperm Individualization and growth Final size of the two organs
and cotyledons stages with milky aspect
~
<I>
..,I:C
:3en
~
~
2.4 Effect of Seed Maturity on Somatic Embryo Induction
The maturation state of seed greatly affects plantlet regeneration through so-
matic embryogenesis. The macromorphological characterization performed
during seed development (Table 2) may be correlated with the determination of
embryogenic ability. It is then possible to define the isolated embryonic axes
derived from immature fruits as the explant which presents higher embryogenic
capacity. This is in accordance also with other studies which related embryogenic
induction to the most active phase of growth (Me et al. 1989), as reported for
other species (Cornu 1988; Sotack et al. 1991).
The decrease in embryogenesis could be due to the reduced effectiveness of
BAP at the assayed concentrations during these maturation stages. Improve-
ment of the embryogenic level could be achieved by increasing BAP concentra-
tions or by using a more effective cytokinin regulator, as previously reported for
walnut (Newman et al. 1992).
The experiments carried out with inocula derived from immature fruits
reveal that three tissue types are suitable for the induction of somatic embryogen-
esis: endosperm, embryonic axes and cotyledon portions; however, different
percentages and times of response are observed (Table 3). In both cases, embryo-
genesis could be directly induced using P or T media. In the presence of the
cited phytohormonal combinations, embryonic axes and proximal cotyledon
portions, derived from seeds collected in August, yield a higher percentage of
responses after 20 or 40 days of culture. The same tissues gave rise to embryo-
genic callus in the presence of 2,4-D (Radojevic et al. 1975). Although no
embryos developed, histological sections confirmed the existence of the first
stages of embryogenic induction, mainly embryos with a globular appearance.
Once fruits reach maturity, embryos are able to germinate. This process was
established in September (Fig. 3), then embryogenesis decreased. When these
primary inocula are observed on the induction media, only the distal cotyledon
portions show a very low embryo induction ability.
Table 3. Percentages of embryogenesis induced in tissues of different developmental stage
Development Type of Culture Induction Embryogenesis

stage tissues media period (days) percentage (%)
Embryonic and carpellary Tor P 0

related tissues portions Tor P 15-20 5
endosperm 60-100 20
Juvenile tissues embryonic axes Tor P 20-40 25-45
cotyledon Tor P 20-40 5-15
portions P 20-40 60
cotyledon BAP25 20-40 10
nodes shoot AlA 0.5
tips GAl 0.02
Adult tissues shoot tips Tor P 20-40 0
IMMATURE FRUIT MATURE FRUIT

60.------------------------,,-----------,
55
50
45
~ 40
Ul
.~ 35
5j 30
01
~ 25
~ 20
OJ 15
10
5
OJ---~~----~~~--_.~~-L_.--~------~
17-July 04-August 20-August OB-September

culture initiation date
a ~ embryonic axes _ proximal portions 0 distal portions
IMMATURE FRUIT MATURE FRUIT

60
55
50
45
'i
~40
Ul
.~ 35
5j30 [>
01 r<
~ 25
~20
OJ 15
I II
10
~
r<
5
o
17-July 04-August 2O-August OS-September
culture initiation date
b ~ embryonic axes _ proximal portions 0 distal portions
Fig. 3a,b. Effect of seed maturation on the induction of somatic embryogenesis on the indicated
tissues. Cultures were done on: a T or b P media
2.5 Secondary Embryogenesis and Proliferation
Once somatic embryos appear, further proliferation normally occurs from the
newly formed tissues which experience the same pattern of differentiation. First,
proliferations are easily maintained on the same induction media, while on fresh
media deprived of hormones embryogenesis declines. Unfortunately, callus
70,-------------------------
60
50
20
10 I
OL---~r_------_,----------r_-------_.____J
5 10 20 40
days of subculture
Fig. 4. Effect of time of culture on secondary embryogenesis
formation is progressively induced and interacts with embryoid proliferation.

Furthermore, since somatic embryogenesis in hazelnut is not a synchronous
response, different stages of embryo organization, embryo germination, shoot
development, and root manifestation are visible.
To avoid the great decrease observed, cotyledonary somatic embryos or
isolated leaves derived from the proliferating cluster may be used. The induction
phase may be newly induced by transferring to fresh media. New induction takes
place in shorter periods; after 5 or 10 days, rosettes composed of cotyledonary
leaves and globular embryos appear in almost 60% of subcultured tissues. Every
40 days new transfers must be carried out (Fig. 4).
2.6 Embryo Germination and Plantlet Formation
Regeneration of plantlets is difficult mainly due to anastomosis and nonsyn-

chronous development. Embryonic shoots and plantlets can develop during
proliferation, yielding one to four clusters.
Apart from the experiments carried out on different media containing ABA,
better results were easily obtained by subculturing isolated embryos or embryo-
nic clusters on to basal fresh media or onto media having low BAP concentra-
tions (0.5 ~M) (Perez et al. 1983). Isolated embryos germinated, giving rise to
plantlets, [at least almost one plantlet may be obtained from each of the 40%
clusters with well-developed shoots (Fig. 5).] Nevertheless, due to the culture
conditions and embryogenic potential, it is possible to observe precocious,
germinated plantlets with structures that are, similar to embryos, distributed
along the developed shoots.
Fig. 5. Shoot development achieved by embryogenesis
Both shoots and plantlets, originating from somatic embryos, are easily
multiplied. Derived shoot segments can be used in a micro propagation program
again. Thus the presence of a reduced amount of BAP (2.5 f..l.M) added to an MS
medium is the only factor necessary to maintain a cycloclonal proliferation line.
Rooting of the tissues is readily accomplished by base dipping in a high concen-
tration IBA solution for a few seconds.
3 Histological Studies
3.1 Somatic Embryogenesis Through Callus Formation
Indirect responses are mainly obtained when embryonic axes or cotyledon

portions are cultured on 2,4-D medium. Macromorphological characteristics
appear to be represented mainly by callus proliferation without external manifes-
tation of somatic embryogenesis.
Calli are composed oflarge cells with wavy walls forming vast, disorganized
masses. Internally two types of organization may be observed: embryogenic
structures and vascular, meristematic organizations.
From zygotic embryos, globular stages with a spherical appearance (Fig. 6a)
and heart-shaped formations are easily distinguished. Globular formations are
composed of small cells with a dense cytoplasm. Using a staining procedure, it
Fig.6a-d. Indirect embryogenesis induced on isolated embryonic axes. a Globular stages (Radojevic
et al. 1975); b further stages, note differentiation of the protoderm {x 300); c globular embryo with a
visible suspensor (x \50); d heart-shape stage (x60)
could be seen that these cells are bounded by walls, indicating a certain indepen-
dence. Sometimes the presence of elongated structures, e.g., suspensors
(Fig. 6b,c), is observed. Here, the terminal cells do not stain so strongly. On the
other hand, the heart-shaped structures show a bipolarity not only through
differential dyeing but also through the organization, which seems to be
characteristic of more developed embryogenic stages (Fig. 6d).
When cotyledon portions are cultured, vascular strands are induced.
Meristematic centers consist of small cells that may occasionally coexist with
prevascular elements located in the central zone that will later develop into the
radicular primordium. This result is confirmed by the manifestation of external
roots when 2,4-D is used.
Fig.7a-d. Stages of direct embryogenesis from embryonic axes. a Localization of the proembryogenic
cells (x300); b organization of embryogenic cells (x450); cglobular, external somatic embryo (x300);
d different stages of external embryogenesis development (x60)
3.2 Somatic Embryogenesis Without Callus Formation
Macromorphologically, all the responses induced on media deprived of 2,4-D

correspond to the formation of roots and somatic embryos without callus
manifestation.
Induction occurs on the most superficial cell layers of the explant used
originating from the parenchyma cells (Fig. 7a,b). These could give rise to
continuous bands of cells with embryogenic characteristics. However, it cannot
be excluded that these sources could be either the origin of embryos or cotyledon-
ary leaves, but the constitution of diads, tetrads and cellular groups with cellular
walls stained more dense, may show that the real source of a percentage of the
somatic embryos constituted is unicellular.
Structures at the globular stage are characterized by spherical formations
joined to the original tissue by a band of cells not differentiable from the rest of
cells constituting the embryo. These and other characteristics, such as the
differentiation of a monolayer of slightly flatter cells by way of a dermatogen
(Fig.7c).
The subsequent evolution towards other stages, i.e., torpedo, heart, etc., has
been demonstrated by the macromorphological investigation of the clusters and
cross sections. Elongated structures that have not completely developed into
individual initial cotyledonary stages (Fig. 7d), but that are in agreement with the
external characteristics, can be observed.
Thus, the histology of embryos at the cotyledonary stage shows, as a
characteristic of this phase, the differentiation ofthe pole surrounded by scarce,
vascular, cotyledonary leaves. Histological sections of the cotyledonary leaves
show that differentiation gave rise to a mesophyll without a clear pallisade layer
or spongy parenchyma. A completely independent ovoid bundle is located
longitudinally and transversally in the future embryonal axis, in which the
radicular pole is as yet not differentiated. According to the organization of the
vascular system, the individuality of the embryos induced can be emphasized. In
the less developed axes, the bundle is composed of small cells with meristematic
characteristics, which would explain the growth in all directions typical of this
stage. As they develop, the leaves become connected by means of a continuous,
vascular bundle that joins with the procambium of the embryo axis.
The development and the complete individualization of these embryos occur
simultaneously or subsequent to the differentiation of the radicular pole. At this
point, cells of the vascular system, representing helicoid enlargements typical of
this stage, can be observed. The parenchymatous cells of the somatic embryo
present a superior diameter with regard to cotyledonary stages less developed.
Being located the meristematic cell, only in the cauline and radicular meristem;
that would explain the dominion of the growth in the longitudinal way observed.
Although most of the embryos give rise to a cluster which is characterized by
a developmental stage, it has been shown that the four described structures can
coexist, giving a certain non synchronous character to this type of induction.
The histological data reveal the embryogenic ability of the tissues used.
Furthermore, a similar evolution corresponding to somatic embryo formation
has been demonstrated, as reported for filbert (Me et al. 1989).
3.3 Histology of Reiterative Embryogenesis
According to the macroscopic organization, in which the induction of embryo-

genesis in the absence of callus can be observed, abundant bands and superficial
formations would mean that the embryos originated from parenchyma cells.
Cotyledonary stages, usually characterized by the absence of differentiation
from the cauline meristem, and ovoid vascularization, composed of cells with
Fig.8a-d. Histological sections during secondary embryogenesis. aTransverse section of an embryo-

genic cluster (x 150); b junction area showing different types of calli (x300); c precocious germination
(x60); d abnormal embryo development by anastomosis of cotyledonary leaves; note two root
primordia (x60)
Somatic Embryogenesis in Hazelnut (Cory/us Species) 333
more or less embryogenic characteristics, from which new immature, vascular

systems develop, would explain the anastomosis and the small number of stems
that rooted from the embryonal cluster (Fig. 8a,b). Occasionally, structures
corresponding to the first precocious germination have been observed. These are
characterized by an abnormal development of the pole, corresponding to the
hypocotyl (Fig. 8c). In this area, a spherical structure with its own vascular
bundle develops. Here, there is no connection with the vascular tissue of the
initial embryo, and a typical ovoid structure of the somatic embryos does not
occur. The constitution of the procambial tissue, in structures with spherical
shapes and without developed cotyledonary leaves, may show that this is not a
globular stage typical of embryogenic development (Fig. 8d).
Somatic embryogenesis is induced on immature embryonic axes or cotyledons

using various phytohormonal combinations. However, the induction on mature
tissues is more complicated and depends on rejuvenation processes. Plantlet
regeneration could be induced, but the percentage of plantlets is still low with
respect to the induction of embryogenesis and proliferation. Cotyledon nodes
which gave rise to 60% embryogenic induction, and embryonic axes with 45%
induction represent two alternative approaches to plantlet formation. Possible
differences may be due to genetic variability, depending on the cultivars used.
Acknowledgments. Partial support by CAICYT grant 860-84 and CICYT (PB 89-0531) is acknowl-
edged.
References
Alvarez Requejo S (1965) EI avellano. Publ Minist agric, Madrid

Bajaj YPS (ed) (1989) Biotechnology in agriculture and forestry, vol 5. Trees II. Springer, Berlin
Heidelberg New York
Berros B (1991) Embryogenesis Somittica en avellano. Seminario de investigaci6n. Servicio de Publi-
cations. U niversidad de Oviedo, Oviedo
Berros B, Rey M, Albuerne M, Rodriguez R (1992) Hazelnut (Corylus avellana L cvs. Gironella and
Negretta) somatic embryogenesis. In: Romisondo P (ed) Res III ISHS Int Congr Hazelnut. ISHS
Working Group Nut Production. Alba, Italy, pp 41
Cheng TY (1975) Adventitious bud formation in culture of Douglas fir (Pseudotsuga menziensii Mirb
Franco.). Plant Sci Lett 5: 97-102
Cornu D (1988) Walnut somatic embryogenesis: Physiological and histological aspects. In: Dreyer E,
Aussenac G, Bonnet-Masimbert M, Disengremel P, Favre JM, Garrec JP, Le Tacon F, Martin F
(eds) Forest tree physiology. Elsevier, Nancy, pp 133-135
Diaz-Sala C, Rey M, Rodriguez R (1990) "In vitro" establishment of a cycloclonal chain from nodal
segments and apical buds of adult hazel Cory/us avellana L. Plant Cell Tissue Organ Cult 23:
151-157
334 B. Berros et al.: Somatic Embryogenesis in Hazelnut (Corylus Species)
Fernandez MT, Rey M, Diaz-Sala C (1990) Age and meristem "in vitro" culture behaviour in filbert.
In: Rodriguez R, Sanchez-Tames R, Durzan DJ (eds) Plant aging: basic and applied approaches.
Plenum Press, New York, pp 351-355
Germain E, Leglise P, Froidefond G (1974) Le bouturage du noisetier (Corylus avellana L.). CTIFL-
Doc 43: 1-8
Gonzalez A (1990) Rizogenesis "in vitro" de Corylus avellana L.: analisis histol6gico y bioquimico
Tesis Doctoral, Servicio de Publications, Universidad de Oviedo, Oviedo
GonzaIezA, Casares A, Sanchez-Tames R, Rodriguez R (1991) Adventitious root induction in Cory/us
avellana L. cotyledon slices, In Vitro Cell Dev Bioi 27: 125-131
Howard BH (1968) Hazel propagation by hardwood cuttings. Rep Exp Malling Stn, pp 93-95
Me G, Enrica E, Botta R, Vallania R (1989) Embryo development in "Tonda Gentile delle Langhe"
hazelnut. Hortiscience 24 (I): 122-125
Neuman MC, Preece JE, Van Sambeek JW, Gaffney (1992) Somatic embryogenesis and callus
production from cotyledon explants of eastern black walnut (Jug/ans nigra L.). Plant Cell Tissue
Organ Cult 32,1 (Jan. 1993): 9-18
Paglietta R (1978) II nocciuolo e la sua coltura. In: Publ a cura del comitato Difesa e Valorizzazione del
Nocciuolo (eds) Metodi di propagazione del Nocciuolo. Alba, Italy pp 1-19
Perez C, Fernandez B, Rodriguez R (1983) In vitro plantlet regeneration through asexual embryogen-
esis in cotyledonary segments of Cory/us avellana L. Plant Cell Rep 2: 226-228
Perez C, Rodriguez R, Sanchez-Tames R (1986) Regulation of asexual embryogenesis in filbert
cotyledonary nodes. Morphological variability. Plant Sci 45: 59-64
Piskornik Z, Poskornik M, Goz F (1982a) The influence of indolebutyric acid, sample-dates and
juvenility on the rooting of filbert (Cory/us sp.) cuttings. Acta Agrobot 35 (2): 167-174
Piskornik Z, Rosek S, Radwan-Kaznica M (1982b) Relation between the ATP content and the viability
of filbert Cory/us avellana pollen stored at 20 ° and 2 °C. Acta Physiol Plant 3 (4): 181-186
Radojevic L, Vujicic R, Neskovic M (1975) Embryogenesis in tissue culture of Cory/us avellana L. Z
Pflanzenphysiol 77: 33-41
Radojevic LJ, Neskovic M, Vujicic R, Kovoor A (1983) In vitro vegetative propagation of Cory/us
avellana L. In: Elenco delle comunizazioni delle Conveg. Int sui Nocciuolo. Conveg Int sui
Nocciuolo, Avellino, pp 283-295
Rey M, Berros B, Alburene M, Rodriguez R (1991) Filbert seed, maturity effect on somatic embryo
induction. Section somatic cell genetics. In: Trends in the biotechnology of woody Plants. 3rd Int
Worksh IUFRO, Dehra Dun, pp 4-7
Rodriguez A (1983) Estudio de la rizogenesis y sustancias reguladoras del crecimiento en Cory/us
avellana L. PhD Thesis, Servicio de publicationes Universidad de Oviedo, Oviedo
Rodriguez R, Vazquez MA, Diaz-Sala C, Sanchez-Tames R (1985) Embryoid induction in filbert
plantlets derived from unstratified seeds. In: Abstr Posters for Conf on Tissue culture as a plant
production system for horticultural crops, Beltsville, pp 16
Rodriguez R, Rodriguez A, Gonzalez A, Perez C (1989) Hazelnut (Cory/us avellana) In: Bajaj YPS
(eds) Biotechnology in agriculture and forestry, vol 5. Trees II. Springer, Berlin Heidelberg
Rodriguez R, Sanchez-Tames R, Durzan DJ (1990) Plant aging: basic and applied approaches. Plenum
Press, New York
Roversi A (1969) Propagazione per talea del nocciuolo con i1 resicaldamento basale e con la forzatura
insachetti di polietilene. Inf Agr Verona, Verona, pp 45
Sotack RJ, Sommer HE, Merkle SA (1991) Relation of the development stage of zygotic embryos of
yellow-poplar to their somatic embryogenic potential. Plant Cell Rep 10: 175-178
Tombesi A (1985) II nocciuolo. REDA (edn Agric)
Tombesi A, Standardi A (1970) La rispondenza alia propagazione delle cultivar de nocciuolo. Tuscia
Econ 9-10: 1-12
Tulecke W, McGranahan G (1985) Somatic embryogenesis and plant regeneration from cotyledons of
walnut, Juglans regia L. Plant Sci 40: 57-63
Wann SR (1988) Somatic embryogenesis in woody species Hortic Rev 10: 153-181
111.7 Somatic Embryogenesis in Oil Palm
(Elaeis guineensis Jacq.)
Y. DUVALl, F. ENGELMANN2, and T. DURAND-GASSELIN3
1 Introduction
1.1 Morphology, Importance, and Distribution
Oil palm (Elaeis guineensis Jacq.) belongs to the family Aracaceae. At the adult
stage, this arborescent monocotyledon shows typical features of palms with a
crown consisting of 40-50 opened, palmate leaves and a cone with 40-50 leaves
at various development stages at the end of an unbranched stem. Growth is
ensured by a single terminal meristem which produces an average of two leaves
per month.
The oil palm is a temporal, dioecious species (Cruden 1988) which presents
alternate male and female flowering cycles during the life time of the individual.
The 50-cm-long inflorescence consists of a peduncle and a spadix enclosed in two
spathes. It emerges from the base of the petiole after total leaf expansion. The
male inflorescence has finger-like, spineless spikes with 400 to 1500 flowers
composed of 6 perianth pieces and an androecium with 6 stamens. The female
inflorescence is formed to shorter spikes bearing 10 to 30 trifloral groups with two
nonfunctional, aborted male flowers and one female flower comprising a
tricarpellary ovary and a trilobed stigma (Beirnaert 1935).
The fruit bunches are harvested 5 to 6 months after flowering. The fruit is a
sessile drupe, black to orange depending on the stage of ripeness, 2 to 5 cm long,
and weighs 10 to 30 g. The mesocarp consists of30 to 60% palm oil. The endocarp
(or shell) surrounds the seed (or kernel) which contains kernel oil.
Oil palm is the second source after soybean for edible vegetable oil. It is
mainly cultivated in industrial plantations between 7° Sand 7° N, at an altitude
lower than 400 m, in areas with abundant rainfall (500 to 3000 mm), well
distributed over the year (Surre and Ziller 1963). Spontaneous or subspontane-
ous plantations exist in Africa, between 10° and 11 ° N (Guinea) and 8° and 10° S
(Angola) (Hartley 1988), which is now considered as the zone of origin of the
I ORSTOM, Laboratoire de Ressources Genetiques et Amelioration des Plantes Tropicales, BP 5045,
34032 Montpellier Cedex, France

2 IPGRI, Via delle Sette Chiese, 142,00145 Rome, Italy
3 CIRAD/CP, Station Principale IDEFORIDPO de La Me, 13 BP 989, Abibjan 13, Ivory Coast

336 Y. Duval et al.
Table 1. Evolution of world production (106 t) of the major vegetable oils. (Anonymous 1991)
Oil 1909-1913 1958-1962 1983-1987 1991-1992" 1988-2002
Palm oil 0.28 1.30 6.77 12.3 17.95

Kernel oil 0.15 0.43 0.90 1.57 2.11
Copra 0.75 1.85 2.69 3.00 4.03
Groundnut 0.64 2.46 3.25 3.98 2.83
Soybean 0.30 3.29 14.15 16.11 22.25
Sunflower 0.12 1.90 6.55 7.90 7.96
Rapeseed 1.08 1.18 6.00 8.96 7.68
Cotton 0.98 2.29 3.40 4.11 4.12
Total 4.3 14.7 43.71 57.93 68.93
a October-September.
Table 2. Evolution and world distribution of harvested areas (1000 ha) of oil palm between 1987 and
1991. (Anonymous 1991)
1987 1988 1989 1990 1991
Cameroon 40 41 41 41 42
Ghana 23 25 28 30 31
Ivory Coast 105 109 115 125 135
Nigeria 230 250 260 270 265
Zaire 68 71 74 75 77
Costa Rica 16 18 19 20 21
Honduras 28 29 29 30 30
Brazil 20 25 29 32 37
Colombia 52 58 70 81 90
Ecuador 45 48 51 53 55
Peru 8 9 7 8 9
China, PR 4 4 5 6 6
Indonesia 399 450 513 605 715
Malaysia 1264 1370 1495 1599 1713
Philippines 23 23 24 25 26
Thailand 64 76 86 94 104
Solomon Islands 3 4 6 6 6
Papua/N. Guinea 31 32 34 37 42
Other countries 131 131 133 135 138
World 2554 2773 3018 3217 3561
species (Zeven 1965). Industrial exploitation of oil palm started at the beginning
of the 20th century, first in Southeast Asia (Malaysia, Indonesia), then in Africa
on the coasts of the Gulf of Guinea and in Brazil in the 1920s. Since the 1960s, its
area of cultivation has been extended to Latin America, mainly in the Amazon
basin (Equator, Colombia, Peru) and to the Pacific coast (Costa Rica). During
this period, considerable planting programs were developed in Southeast Asia
and to a lower extent in West Africa. This led to an eightfold increase in the world
Somatic Embryogenesis in Oil Palm (Elaeis guineensis Jacq.) 337
production of palm oil between 1960 and 1970 (Table 1). Production is presently
over 11 million tons/year and is still increasing, due to a continuous increase in
the planted surface which has now reached 3.5 millions ha (Table 2). During the
next decade, the consumption of oleaginous products should continue to rise
(Table 1). According to Oil World (Anonymous 1991), this increase in the
consumption of oil products will affect mainly soybean and oil palm.
1.2 Significance of Somatic Embryogenesis for Oil Palm
Due to the increasing demand for palm oil and the foreseeable rises in exploita-
tion costs, material with high genetic potential must be made available to the
planters. Planting material consists solely of tenera hybrids (fruits with a shell of
intermediate thickness), originating from crosses between dura (thick shell) and
pisifera (thin shell) types, the thickness being controlled by a monofactorial gene
(Beirnaert and Vanderweyen 1941). The breeding schemes developed by seed
companies aim at producing dura x pisifera hybrids with a high production of oil
with a high proportion of unsaturated fatty acids, a low growth rate and, in
particular cases, resistance to diseases such as vascular wilt caused by Fusarium
oxysporum elaeidis in Africa (Hard on et al. 1976). Each selection cycle lasts for
around 10 years, and genetic improvement is very slow, even if highly significant
progress has been achieved over the last 40 years. A very high heterogeneity is still
observed among hybrids, some palms producing 60% more oil than the average
cross (Noiret 1981). When these characteristics are combined with the low
planting density (generally 143 palms/ha) and the necessity of establishing seed
orchards for the production of commercial material, it appears that oil palm
improvement is labor-intensive, time-consuming, and therefore expensive. All
the constraints of oil palm breeding call for the development of a vegetative
propagation technique which would allow one:
1. To exploit the variability existing among the ten era hybrids by cloning elite
trees;
2. To increase the production of high quality seeds by cloning the best male
genitors (pisifera), since pollen production can be a limiting factor (Hartley
1988; Krikorian 1989);
3. To exploit the interspecific hybrids E. guineensis x E. oleifera, a limited
number of which are fertile, but which can show good tolerance to pests and
diseases, notably in South America (Meunier 1975).
The biological characteristics of the oil palm do not allow its vegetative
propagation by conventional means, in spite of the works carried out on the
reversion offIoral buds or the study of viviparous palms (Chevalier 1910; Henry
1948; Davis 1980). The in vitro culture of apices was investigated in juvenile
palms (Staritsky 1970; Abdullah 1990), but they did not produce exploitable
results. Moreover, the excision of the apex leads to the death of the mother tree.
Therefore, the best way to clonalIy propagate elite oil palms is by means of
338 Y. Duval et al.
Table 3. Progress in oil palm somatic embryogenesis in tropical countries
Country Reference Progress
Malaysia Paranjothy and Othman (1982) Field plantation

P.R. China Yuang et al. (1984) In vitro plantlets
India Thomas and Rao (1985) Prenursery
Brazil Cid (1987) Prenursery
Indonesia Tahardi (1989) Field plantation
Thailand Te-Chato et al. (1991) Somatic embryos
1.3 Work Done by Others
Many detailed reviews have been published on oil palm (Blake 1983; Paranjothy
1984; Jones and Hughes 1989; Krikorian 1989; Wooi 1990). For more than 20
years, two teams have carried out research programs: in the UK and in Malaysia,
the U nilever Plantations and Harrisson and Crossfield Plantations group (Smith
and Jones 1970; Corley et al. 1977) and in France and the Ivory Coast, the
ORSTOM/IRHO group (Rabechault et al. 1970; Noiret 1981). The programs,
initiated by various companies, aimed at multiplying the elite germplasm avail-
able in their plantations for commercial use. Although the results obtained led to
potentially important commercial applications, only limited information was
published concerning the techniques used, as underlined by various authors
(Krikorian 1989; Blake 1990). During the 1980s several teams from tropical
countries started their own research programs (Table 3). Due to the dynamics of
the oil palm industry in Malaysia, several local teams developed large-scale
research programs, notably at PORIM (Paranjothy and Othman 1982). In 1990,
Wooi estimated that, in Malaysia around ten commercial laboratories were
carrying out research on in vitro propagation of the oil palm through somatic
embryogenesis.
2 Somatic Embryogenesis and Regeneration in Oil Palm
The regeneration of oil palm through somatic embryogenesis comprises four

steps (callogenesis, embryogenesis, shoot development, and rooting), which are
summarized in Fig. 1. In vitro culture of this plant is characterized by extended
delays (up to 2 years between sampling of the primary explants and acclimatiza-
tion of the regenerated plantlets), the low efficiency of some steps of the process,
and the production of polyphenols, the influence of which on the success of
regeneration is poorly known.
POLYEMBRYONIC
CULTURE
Fig. 1. Schematic representation of the various pathways for oil palm micropropagation through
somatic embryogenesis
2.1 Callogenesis
Various types of explants can be used for callus production but they do not offer
the same potentialities. Juvenile material, embryos, roots, and leaves of asepti-
cally germinated seedlings (Smith and Thomas 1973; Jones 1974) or leaves from
nursery palms (RaMchault and Martin 1976; Pannetier et al. 1981; Thomas and
Rao 1985; Cid 1987) have been used because tissues of juvenile origin are more
reactive than those from adult material. With adult palms, callus is initiated from
immature leaves (Noiret et al. 1985), roots, and possibly young inflorescences
(Wooi et al. 1982). Each type of explant presents different advantages. Immature
340 Y. Duval et al.
leaves sampled from the spike of an adult palm can provide several thousand
explants. The leaves are well protected in the unopened spike and, therefore,
disinfection is not necessary if special precautions are taken during their dissec-
tion in the laboratory. Sampling of the immature leaves should not be carried out
too close to the apex, in order to avoid damaging it, thus ensuring the survival of
the donor palm. Depending on the genetic origin of the palm, up to 60% of the
leaf fragments can produce callus. It has been observed that the total absence of
callus production is extremely rare, and can be considered an experimental
artifact.
Roots have the drawback of requiring rigorous disinfection, usually with
mercuric chloride which is likely to damage the tissues. The large root system
allows frequent sampling without stressing the ortet, in contrast to the sampling
of immature leaves which can be repeated only every 2 years after the regrowth
of the crown. However, special precautions have to be taken in order to sample
roots from the palm selected. Indeed, the roots of trees growing next to another
are deeply entangled and mistakes can easily be made (Jones 1984), which is not
the case when immature leaves or inflorescences are used.
In spite ofthe limited information available, it is recognized that callogenesis
occurs in the presence of auxins in a full-strength Murashige and Skoog (1962)
derived basal medium (see W ooi 1990) or sometimes half-strength (Nwanko and
Krikorian 1983), supplemented with vitamins, casein hydrolysate (0.5 to 1 gil)
and sucrose or glucose (20 to 30 g). In contrast to coconut, which requires a
special mineral formulation (Eeuwens 1976), it seems that oil palm is less
demanding regarding mineral elements. The pH ranges between 4.5 and 5.7, but
it seems to have little influence on the success of callogenesis (Krikorian and
Kann 1986). Various auxins are employed, e.g., naphthalene acetic acid (NAA),
2,4-dichlorophenoxyacetic acid (2,4-D), and less frequently trichlorophen-
oxyacetic acid (2,4,5-T) or trichlorophenoxypropionic acid (2,4,5-TP). Cytoki-
nins [Kinetin, benzylaminopurine (BAP) , isopentenyladenine (2-iP)] are
sometimes added at very low concentrations. Therefore, the auxin/cytokinin
balance is very high (around 20) (Rabechault and Martin 1976), and according to
some authors (Nwanko and Krikorian 1983), cytokinins are not necessary. The
auxin concentration varies between 10-6 and 10-4 M. The highest auxin concen-
trations are always used in the presence of activated charcoal which adsorbs up
to 99% of the growth regulators present in the medium (Ebert and Taylor 1990).
The addition of activated charcoal also limits the diffusion of phenolic com-
pounds and the darkening of both media and explants during the callogenesis
phase, although browning is not a negative factor with oil palm (Wooi 1990). The
time necessary for calli to appear varies between 8 and 16 weeks, depending on
the type of explant and the genetic origin of the material.
The origin of the callus has been studied on all types of explants, but
apparently not on inflorescences. All the observations indicate that callus origi-
nates from cells connected with the vascular system. In embryos, a longitudinal
section performed after 3 weeks of culturing shows disruption and branching of
vascular strands (Nwanko and Krikorian 1983). Multiplication of perivascular
cells occurs leading to the formation of centers with high meristematic activity
from which callus originates. In root explants, cells close to the pericycle and the
Fig. 2. Cross section of a nodular callus showing the cambium-like zone (CL) and the sclerenchyma
(SC); bar = 100 ~M
cortical cells multiply to form a callus (Jones 1983). In young leaves, the callus
originates at the level of the veins (Ahee et al. 1981). This latter observation was
established by an extensive histological analysis of callogenesis (Schwendiman
et al. 1988). A transverse section in the primary explant revealed that callus
originates from the multiplication of perivascular cells, resulting in a layer of
meristematic tissue, four or five cells thick, close to the xylem. This layer evolves
into a cambium-like zone which differentiates rapidly into a sclerenchyma
towards the inside of the nodule and into quiescent cells towards the outside
(Fig. 2.). The activity of the pseudocambium ensures callus growth. Invagina-
tions of the pseudocambium lead to the formation of individualized nodules.
2.2 Embryogenesis
In oil palm, embryogenesis is always indirect, with an intermediate callogenesis

phase. The calli are isolated and transferred to a medium inducing embryogen-
esis, generally under light conditions. Auxin concentration is generally lower
than that used during the callogenesis phase. 2,4-D and NAA can be employed
separately at 2.5 x 10-6-2 x IO-4M. Cytokinins (BAP, kinetin, 2-iP) may be added
to the medium at low concentrations (10-7 to 0.5 10-5 M) (see Wooi 1990 for a
review). However, due to the presence of activated charcoal in most of the media
used, the exact quantity of growth regulators available to the cultures cannot be
342 Y. Duval et al.
determined precisely. The difference in the growth regulator concentration

between the callogenesis and embryogenesis phases seems to be more important
for the production of embryos than the exact concentration of the growth
hormones used during each step. The production of somatic embryos occurs over
three pathways depending on the type of callus from which they originate (Fig. 1).
2.2.1 Embryogenesis on Nodular Callus
The mode of appearance of embryos on nodular calli has been extensively studied
by Schwendiman et al. (1988, 1990). The cambium-like zone undergoes an
important modification. The meristematic cells of the nodular callus evolve into
cells displaying histochemical characteristics typical of embryogenic cells: high
nucleocytoplasmic ratio, large central nucleus with a single nucleolus, numerous
lipidic droplets, dense cytoplasm intensely stained by naphthol blue black,
gelification of the median lamella which individualizes isolated cells and cell
clusters (Fig. 3). This differentiation of the cambium-like zone, which is not
visible to the naked eye, seems to be the first phase of embryogenesis in nodular
calli. This phenomenon is followed by intense multiplication of the embryogenic
cells. The cell clusters grow, develop an epidermis, and give rise to structures
termed proembryos, embryoids, or somatic embryos, depending on the authors
and certainly on the developmental stage considered. Somatic embryos appear
Fig. 3. Cross section showing an embryogenic zone in a nodular callus. PE Proembryo; bar = 100 11M
Fig.4. Evolution with time of the cumulative 100

percentage of callus cultures producing
somatic embryos (-) and polyembryonic
cultures (---) 8, 80
I
.l2
60
~
~ 40
::.
§
o 20
o 10 20 30 40 50
Culture duration (months)
with almost all clones, but generally at a low frequency. Experiments performed
at the La Me Research Station (Ivory Coast) showed that after 50 months in
culture, 95% ofthe calli produced at least one embryo and that 50% ofthem gave
rise to polyembryonic cultures which could be used for mass production of
plantlets (Fig. 4).
2.2.2 Embryogenesis ofFast-Growing Callus
When nodular calli are cultivated on media with a high auxin content, a new type
of callus can appear, termed fast-growing callus (FGC) (Smith and Thomas
1973; Ahee et al. 1981; Hanower and Pannetier 1982). FGCs have a very high
growth rate, they are whitish, soft and friable, and are composed of dispersed
clumps of meristematic cells, large cells, and lacunae (Ahee et al. 1981). They
originate from a disorganization of the procambial zone, in which the meristem-
atic cells show a disorganized proliferation (Fig. 5). When FGCs are transferred
on to media with a reduced auxin concentration and/or supplemented with
cytokinin, intense embryogenesis is observed. This phenomenon can be enhanced
by the addition of antiauxins [2-0-chlorophenoxyisobutyric acid (10-6 M) or 7-
azaindol (10-5 M)] or phenolic compounds (phloridzin or phloroglucinol at
1O-3 M) to the medium (Hanower and Hanower 1984). The meristematic cells
become embryogenic, multiply, and give rise to embryogenic formations.
2.2.3 Embryogenesis on Granular Callus
Granular callus is rarely observed, but appears on the surface of nodular callus
cultivated under standard conditions. Granular callus is composed of small « 1
mm) meristematic nodules. In contrast to the embryogenic structures observed
344 Y. Duval et al.
Fig. 5. Cross section ofa fast-growing callus. bar = 100 J.IM
Fig. 6. Cross section of a granular callus. N Meristematic nodule; bar = 100 J.IM
on nodular callus which have an internal origin, these nodules are located at the
periphery of the parenchymatous callus to which they are only slightly connected
(Fig. 6; de Touchet 1991). The histological examination allows the characteriza-
tion of meristematic nodules of variable sizes, from two to several tens of cells.
The most advanced developmental stage that can be attained by these structure
is the production of a dermatogen even though culture conditions are identical to
those used for the induction of somatic embryos on nodular callus. When these
nodules are transferred to liquid medium containing 2,4-D (5 x 10-4 M) and
activated charcoal (2g/1), they can give rise to embryogenic cell cluster suspen-
sion (Fig. 7). These suspensions allow the production of individualized somatic
embryos after one subculture in liquid, hormone-free medium and plating on
solid medium (de Touchet et al. 1991). Thus, embryogenesis on granular callus
corresponds to the third pathway of embryogenesis in oil palm.
2.3 Proliferation and Regeneration of Plantlets
The embryogenesis phenomenon observed on nodular calli and FGCs leads to

a complex system termed polyembryonic culture. It consists of a group of
proembryonic formations composed of meristematic cells and somatic
embryos at various developmental stages. On hormone-free medium, these
polyembryonic cultures produce new adventitious somatic embryos and, simul-
Fig. 7. Suspension of embryogenic clusters in proliferation; bar = I cm (de Touchet et al. 1991)
346 y. Duval et al.
Fig. 8. Two-year old clonal palm in field trial
taneously, the most advanced embryos develop into shoots. Depending on the
culture conditions and genotype, the phenomenon of adventitious embryogen-
esis is rare and leads to the production of a limited number of shoots only. In
contrast, in the most favorable cases, polyembryonic cultures show continuous
proliferation and ensure mass propagation.
When the shoots have reached a sufficient size (5 to 7 cm), they are separated
manually and rooted on a medium containing 2.5 x 10-6 to 10-4 M NAA and
sometimes GA3(Nwanko and Krikorian 1983). Plantlets are acclimatized under
controlled conditions in order to avoid hydric stress which is common for all
plants produced in vitro. Wooi et al. (1982) mention an acclimatization proce-
dure ensuring 90% success with well-rooted plantlets placed under a plastic
frame. The observations carried out by our team show that it is necessary to
develop the root system during the in vitro phase. Indeed, rooting during
hardening could never be obtained, even after auxin treatment of the shoots.
After 4 weeks of acclimatization, the plantlets are transferred to the prenursery
and treated like seedlings before their transfer to the field (Fig. 8).
2.4 Trueness to Type of Regenerated Palms
The first observations carried out on clonal palms at the prenursery and nursery
stages (Corley 1977; Corley et al. 1977) did not reveal any special problems. The
further vegetative development of clonal material revealed a high homogeneity

for all characters observed (Corley 1981; Wooi et al. 1982). Vegetative abnor-
malities of clonal material and/or seedlings (curled leaves, erected palms, thick
leaflets) were rare. Therefore, the culling percentage is identical for clonal and
sexual material (T. Durand-Gasselin, unpubl.).
The occurrence of a floral morphogenetic abnormality in clonal palms was
mentioned first in 1986 by Corley et al. in Malaysia and then by Duval et al.
(1988) in the Ivory Coast. This irregularity called a mantled abnormality (Corley
et al. 1986), is induced by in vitro culture, and consists of a feminization of the
male parts of female flowers. The staminoids exhibit too vigorous growth,
leading to supernumerary carpel formation. This leads to the formation of
abnormal (mantled) fruits or to the partial or complete sterility of the palms,
depending on the extent of the abnormality. In the most severe cases, besides the
sterility of female flowers, male flowers are affected, stamens developing into
pseudocarpels. This abnormality is reversible since, in many cases, a definitive
recovery has been observed after several months or years in the field (Durand-
Gasselin et al. 1990). The information available indicates that this phenomenon
is probably a consequence of the in vitro culture technique employed. According
to Corley et al. (1986), successive plantings performed between 1981 and 1983
showed that the frequency of mantled palms increased in parallel with the
duration of proliferation of the polyembryonic cultures, up to 95% being affected
in 1983. In the Ivory Coast, all the palms produced through the FGC pathway
exhibited severe abnormality (Duval et al. 1988). By contrast, abnormality was
observ~d in a limited number of plants originating from somatic embryos
obtained on nodular calli. Indeed, an average of only 3.1 % of severely mantled
palms was observed in 1991 on 11000 plants coming from 54 clones. However, a
very high variability could be observed among clones, the most affected one
displaying 67% abnormal palms (T. Durand-Gasselin, unpubl.). No correlation
could be made between the duration of proliferation of polyembryonic structure
and the percentage of abnormality. Various hypotheses have been proposed in
the case of oil palm, unfortunately without any experimental basis (Soh 1987;
Paranjothy et al. 1990).
Numerous works have underlined the importance of growth regulators in
the process of sexual differentiation in plants (Chailakhyan 1979; Durand and
Durand 1984), notably in the case of oil palm (Corley 1976). The analyses of
Table 4. Endogenous cytokinin levels (fmol/mg dry weight) in normal and abnormal inflorescences of
palms from two clones'
Z [9R]-Z iP [9R]-iP
Normal <10 430 ± 15 < 10 330 ± 8

CloneE
Abnormal <10 155 ± 21 < 10 57 ± 5
Normal < 10 398 ± 30 <10 437 ± 44
CloneF
Abnormal < 10 212 ± 31 <10 91 ± 7
• Z: Zeatin; [9R]-Z: 9-b--D-ribofuranosylzeatin; iP: N6-(,\2-isopentenyl)adenine; [9R]-iP: 9-b-

ribofuranosyl-iP.
348 y. Duval et al.
Fig. 9. Encapsulated somatic embryos derived from an embryogenic suspension; bar = 5 mm
endogenous growth regulators performed on immature inflorescences sampled

on normal and abnormal palms revealed a deficiency in cytokinins in the
abnormal inflorescences (Table 4; Besse et al. 1991). The observations made in
the Ivory Coast, which stressed the importance of the type of callus used with
regard to the regeneration of true-to-type plants, were confirmed by hormone
analyses. Indeed, they showed that the metabolism of cytokinins of the zeatin
group was disturbed in FGCs (Besse et al. 1992). This was linked with the
concentration of2,4-D used in the embryogenesis media (Besse 1992). Although
no hormonal dysfunctioning could be detected in the embryos and plantlets
(Jones and Hanke 1990), it is likely that this disruption in cytokinin metabolism
is responsible for the occurrence of the mantled abnormality in adult palms.
These recent results, which allow one to determine the origin of mantled
abnormality, give hope for the future that it will be possible to control this
phenomenon.
2.5 Embryogenic Suspensions and Encapsulated Single Embryos
The cloning technique through somatic embryogenesis on solid medium is still

based on classical micro propagation procedures. The growth rate of poly-
embryonic cultures hardly allows the production of tens of thousands of plantlets
per clone, notably due to the delays necessary to complete each phase of the
process and the inability to master adventitious embryogenesis. Therefore,
research on the use of embryogenic suspensions was initiated in our laboratory
(de Touchet el al. 1990, 1991) in order to develop new methods of automation
and scale-up of plant propagation in vitro (see Bajaj 1991). Embryogenic
suspensions have already been obtained from 17 different clones. With this
technique the first encapsulated single embryos of oil palm (Fig. 9) were pro-
duced which developed into plantlets under in vitro conditions. The concentra-
tion of the embryogenic suspensions available in our laboratory is around 105 cell
clusters/l with a multiplication factor of 4 per month. This will allow mass
propagation. Thus, the culture of embryogenic suspensions in bioreactors must
be developed, this has already been successfully attempted with oil palm
(SondahI1991). It will be necessary to set up new field trials to test the trueness to
type of the material produced through this new process. With this aim, palms of
the first clone multiplied in the form of embryogenic suspensions were planted in
1991 at La Me Research Station in the Ivory Coast. These promising results,
coupled with the cryopreservation technique which has been used successfully
with oil palm (Engelmann et al. 1985; Dumet et al. 1992), should allow the drastic
modification of the production strategy in the coming years.
3 Conclusions and Prospects
Somatic embryogenesis appears to be the only possible way to clonally propagate

oil palm. With this species, somatic embryogenesis is an indirect process, having
a callogenesis phase. Calli can be obtained from various explants, leaves, roots,
and inflorescences. Three pathways have been identified which differ according to
the type of callus employed for the production of somatic embryos. This has a
dramatic influence regarding both the trueness to type of the plants regenerated
and the possibilities of large-scale development of the process. The use of fast-
growing callus (FGC) leads to the production of 100% abnormal plants due to a
deregulation of the metabolism of the endogenous growth substances in this type
of callus. The production of embryos from nodular callus is the pathway
routinely employed by ORSTOM/IRHO. Several million clonal palms have
already been produced and in the agronomical field trials under observation
(over 2000 ha), the level of floral abnormality is very low (3.1%). However, two
problems are encountered when using nodular callus as a source for somatic
embryos: large-scale propagation is impossible to achieve with all the clones and
the numerous manual operations necessary at all stages of the process (regular
transfers of calli and polyembryonic cultures, isolation and rooting ofplantlets)
cause very high labor costs. The use of the third type of callus, i.e. granular callus,
is still at the experimental stage, but should be extensively developed in the
coming years due to its potential interest. Indeed, the culture of embryogenic
suspensions in bioreactors should make the large-scale production of somatic
embryos with all clones possible. For this new strategy new technologies must be
developed in order to ensure the production and storage of encapsulated, single
embryos.
Acknowledgments. This work is supported by ORSTOM (Institut Francais de Recherche Scientifique

pour Ie Developpement en Cooperation) and CIRAD/CP (Centre de Cooperation Intemationale en
Recherche Agronomique pour Ie DeveloppementlDepartement Cultures Perennes). The authors wish
350 Y. Duval et al.
to thank the personnel of the tissue culture laboratories ofIDEFOR/La Me Research Station (Ivory
Coast) and ORSTOM/Montpellier as well as those involved in the oil palm vegetative propagation
program.
References
Abdullah R (1990) Clonal micropropagation of oil palm through apical meristem culture. In: Abstr
VIlth Int Cong Plant tissue and cell culture, Amsterdam, June 24-29, p 85
Ahee J, Arthuis P, Cas G, Duval Y, Guenin G, Hanower J, Hanower P, Lievoux D, Lioret C,
Malaurie B, Pannetier C, Raillot D, Varechon C, Zuckermann L (1981) La multiplication
vegetative in vitro de palmier it huile par embryo genese somatique. Oleagineux 36: 113-118
Anonymous (1991) Oil world annual. ISTA, Hamburg
Bajaj YPS (ed) (1991) Biotechnology in agriculture and forestry, vol 17. High-Tech and Micropro-
pagation 1. Springer, Berlin Heidelberg New York
Beirnaert A (1935) Introduction it la biologie florale du palmier it huile (Elaeis guineensis Jacq.). Publ
Inst Nat! Etudes Agron, Congo Beige, Ser Sci 5: 42
Beirnaert A, Vanderweyen R (1941) Contribution it I'etude genetique et biometrique des varietes
d' Elaeis guineensis Jacq. Publ NEAC Ser Sci no 27
Besse I (1992) Recherche du determinisme hormonal de l'anomalie de la morphogenese florale liee it
I'embryogenese somatique chez Ie palmier it huile (Elaeis guineensis Jacq.). These d'Universite,
Paris 6, 11112/92, 152 p
Bessel, VerdeilJL, Duval Y, SottaB, Maldiney R, Miginiac E(1991)Oil palm (ElaeisguineensisJacq.)
clonal fidelity; endogenous cytokinins in callus cultures. In: Proc 14th Int Conf Plant growth
substances, Amsterdam, 21-26 June 1991,95 pp
Besse I, VerdeiJJL, Duval Y, Sotta B, Maldiney R, Miginiac E (1992) Oil palm (Elaeis guineensis Jacq.)
clonal fidelity: endogenous cytokinins and indoleacetic acid in embryogenic callus cultures. J Exp
Bot 43: 983-989
Blake J (1983) Tissue culture propagation of coconut, date and oil palm. In: Dodds JH (ed) Tissue
culture of trees. Croom Helm, London, pp 29-50
Blake J (1990) Coconut (Cocos nucifera 1.) micropropagation. In: Bajaj YPS (ed) Biotechnology in
agriculture and forestry, vol 10. Legumes and oilseed crops 1. Springer, Berlin Heidelberg New
York, pp 538-554
Chailakhyan MK (1979) Genetic and hormonal regulation of growth, flowering and sex expression in
plants. Am J Bot 66: 717-736
Chevalier A (1910) Les vegetaux utiles de I' Afrique tropicale francaise. Doc Palmier Huile, Publisher
Challanel, Paris, 127 pp
Cid LPB (1987) Regeneracao de plantas de Elaeis oleifera e du seu hibrido com Elaeis guineensis via
embriogenese somatica. Pesq Agropec Bras, Brasilia 22: 109-113
Corley RHV (1976) Inflorescence abortion and sex differentiation. In: Corley RHV, Hardon JJ, Wood
BJ (eds) Oil palm research. Elsevier Amsterdam, pp 37-54
Corley RHV (1977) First clonal oil palms planted in the field. Planter 616: 331-332
Corley RHV (1981) Clonal planting material for the oil palm industry. Perak Planter's Assoc J, pp
35-49
Corley RHV, Barrett IN, Jones LH (1977) Vegetative propagation of oil palm via tissue culture. Oil
Palm News 22: 2-8
Corley RHV, Lee CH, Law LH, Wong CY (1986) Abnormal flower development in oil palm clones.
Planter, Kuala Lumpur 62: 233-240
Cruden RW (1988) Temporal dioecism: systematic breadth, associated traits, and temporal patterns.
Bot Gaz 149: 1-15
Davis TA (1990) Double reversal of spadices in an African oil palm. Planter, Kuala Lumpur 56: 212
de Touchet B (1991) Micropropagation du palmier it huile (Elaeis guineensis Jacq.) en milieu liquide.
These Doctorat en Sciences, Univ Paris Sud-Centre d'Orsay, 13/12/1991, 153 pp
de Touchet B, Duval Y, Pannetier C (1990) Oil palm (Elaeis guineensis Jacq.) regeneration from
embryogenic suspension culture. In: Abstr VIIth Int Congr Plant tissue and cell culture,
Amsterdam, June 24-29, p 249
de Touchet B, Duval Y, Pannetier C (1991) Plant regeneration from embryogenic suspension cultures
of oil palm (Elaeis guineensis Jacq.). Plant Cell Rep 10: 529-532
Dumet D. Engelmann F, Duval Y (1992) Use of standard somatic embryos for the cryopreservation of
oil palm (Elaeis guineensis Jacq.) cultures. In: Proc XIIIth EUCARPIA Cong Reproductive
biology and plant breeding, Angers, France, 6-10 July 1992, pp 413-414
Durand R, Durand B (1984) Sexual differentiation in higher plants. Physiol Plant 60: 267-274
Durand-Gasselin T, Le Guen V, Konan K, Duval Y (1990) Oil palm (Elaeis guineensis Jacq.)
plantations in Cote d'Ivoire obtained through in vitro culture. First results. Oleagineux 45: I-II
Duval Y, Durand-Gasselin T, Konan K, Pannetier C (1988) Multiplication vegetative du palmier Ii
huile par culture in vitro, strategie et resultats. Oleagineux 43: 39-47
Ebert A, Taylor HF (1990) Assessment of the changes of 2,4-dichlorophenoxyacetic acid concentra-
tions in plant tissue culture media in the presence of activated charcoal. Plant Cell Tissue Organ
Cult 20: 165-172
Eeuwens CJ (1976) Mineral requirements for growth and callus initiation of tissue explants excised
from mature coconut palms (Cocos nucifera) cultured in vitro. Physiol Plant 36: 23-28
Engelmann F, Duval Y, Dereuddre J (1985) Survie et proliferation d'embryons somatiques de palmier
Ii huile (Elaeis guineensis Jacq.) apres congelation dans l'azote liquide. C R Acad Sci Paris 30 I, Ser
III: 111-116
Hanower J, Hanower P (1984) Inhibition et stimulation, en culture in vitro, de l'embryogenese de
souches issues d'explants foliaires de palmiers Ii huile. C R Acad Sci Paris Ser III, 298: 45-48
Hanower J, Pannetier C (1982) In vitro vegetative propagation of the oil palm, Elaeis guineensis Jacq.
In: Fujiwara A (ed) Proc 5th Int Cong Plant tissue culture 1982. Maruzen, Tokyo, pp 745-746
Hardon JJ, Gascon JP, Noiret JM, Meunier J, Tan GY, Tam TK (1976) Major oil palm breeding
programmes. In: Corley RHV, Hardon JJ , Wood BJ (eds) Development in crop science I. Oil palm
research. Elsevier, Amsterdam, pp 109-126
Hartley CWS (1988) The oil palm - tropical agriculture series. Longman, London, 761 pp
Henry P (1948) Un Elaeis remarquable: Ie palmier Ii huile vivipare. Rev Int Bot Appl Agric Trop 28:
422
Jones LH (1983) The oil palm and its clonal propagation by tissue culture. Biologist 30: 181-188
Jones LH (1984) Novel palm oils from cloned palms. 1 Am Oil Chern Soc 61: 1717-1719
Jones LH, Hanke DE (1990) The endogenous cytokinins of oil palm cultures and the effect of kinetin
on cytokinin metabolism. Plant Cell Tissue Organ Cult 20: 1-5
Jones LH, Hughues WA (1989) Oil palm (Elaeis guineensis Jacq.). In: Bajaj YPS (ed) Biotechnology in
agriculture and forestry, vol 5. Trees II. Springer, Berlin Heidelberg New York, pp 176-202
Krikorian AD (1989) The context and strategies for tissue culture of date, African oil and coconut
palms. In: Vibha Dhawan (ed) Application of biotechnology in forestry and horticulture. Plenum
Press, New York, 119-144
Krikorian AD, Kann RP (1986) Oil palm improvement via tissue culture. Plant Breed 57: 175-202
Meunier J (1975) Le palmier a huile americain Elaeis melanococca. Oleagineux 30: 51-61
Noiret 1M (1981) Application de la culture in vitro Ii I'amelioration et a la production de material
clonal chez Ie palmier Ii huile. Oleagineux 36: 123-126
Noiret JM, Gascon JP, Pannetier C (1985) La production de palmier Ii huile par culture in vitro.
Oleagineux 40: 365-372
Nwanko BA, Krikorian AD (1983) Morphogenetic potential of embryo and seedling derived callus of
Elaeis guineensis Jacq. var pisifera Becc. Ann Bot 51: 65-76
Pannetier C, Arthuis P, Lievoux D (1981) Neoformation de jeunes plantes d' Elaeis guineensis Ii partir
de cals primaires obtenus sur fragments foliaires cultives in vitro. Oleagineux 36: 119-122
Paranjothy K (1984) Oil palm. In: Ammirato PV, Evans DA, Sharp WR, Yamada Y (eds) Handbook
of plant cell culture, vol 3. Crop species. Macmillan, New York, pp 591-605
Paranjothy K, Othman R (1982) In vitro propagation of oil palm. In: Fujiwara A (ed) Proc 5th Int
Cong Plant tissue culture 1982. Maruzen, Tokyo, pp 755-756
352 Y. Duval et al.: Somatic Embryogenesis in Oil Palm (Elaeis guineensis Jacq.)
Paranjothy K, Saxena S, Banerjee M, Jaiswal VS, Bhojwani SS (1990) Clonal multiplication of woody
perennials. In: Bhojwani SS (ed) Plant tissue culture applications and limitations. Elsevier,
Amsterdam, pp 190-200
Rabechault H, Martin JP (1976) Multiplication vegetative du palmier it huile (Elaeis guineensis Jacq.)
it l'aide de culture de tissues foliaires. C R Acad Sci, Paris Ser D 283: 1735-1737
Rabechault H, Guenin G, Ahee J (1970) Recherches sur la culture in vitro des embryons de palmier it
huile (Elaeis guineensis Jacq. var. dura Becc.). Comparaison de divers milieux mineraux.
Schwendiman J, Pannetier C, Michaux-Ferriere N (1988) Histology of somatic embryogenesis from
leaf explants of the oil palm Elaeis guineensis. Ann Bot 62: 43-52
Schwendiman J, Pannetier C, Michaux-Ferriere N (1990) Histology of embryogenic formations
during in vitro culture of oil palm Elaeis guineensis Jacq. Oleagineux 45: 409-418
Smith WK, Jones LH (1970) Plant propagation through cell culture. Chern Ind 44: 1399-1401
Smith WK, Thomas JA (1973) The isolation and in vitro cultivation of cells of Elaeis guineensis.
Soh AC (1987) Abnormal oil palm clones. Possible causes and implications: further discussions.
Planter, Kuala Lumpur 63: 59--65
Sondahl M (1991) Tissue culture of cacao, coffee and oil palm. In: Proc 4th IPBNet Conf, Biotechnol-
ogy for tropical crop improvement in Latin America. San Jose, Costa Rica, 14-18 January 1991, pp
98-99
Staritsky G (1970) Tissue culture of the oil palm (Elaeis guineensis Jacq.) as a tool instrument for its
vegetative propagation. Euphytica 19: 228-292
Surre C, Ziller R (1963) Le palmier it huile. Maisonneuve et Larose, Paris, 243 pp
Tahardi S (1989) In vitro approaches to oil palm improvement. Proc Worksh Biotechnology for the
improvement of oil- and steroid-producing plants. Jakarta, 5--6 April 1988, pp 49-55
Te-Chato S, Salee D, Arunee MK (1991) Origin of somatic embryos in oil palm tissue culture. Plant
Biotechnol Newslett, Thailand 19: 8-9
Thomas V, Rao PS (1985) In vitro propagation of oil palm (Elaeis guineensis Jacq. var. tenera) through
somatic embryogenesis in leaf-derived callus. Curr Sci 54: 184-185
Wooi KC (1990) Oil palm (Elaeis guineensis Jacq.): tissue culture and micropropagation. In: Bajaj YPS
(ed) Biotechnology in agriculture and forestry, vol 10. Legumes and oilseed crops I. Springer,
Wooi KC, Wong CY, Corley RHV (1982) Tissue culture of palm: a review. In: Rao AN (ed) Proc Symp
Tissue culture of economically important plants COSTED, Singapore, pp 138-144
Yuang C, Zheng G, Ping F, Xinghua C, Xingdi S (1984) Induction of embryoids and plantlets from oil
palm callus. In: Proc 3rd Int Symp Genetic manipulation in crops, Beijing, China, 22-26 Oct 1984
Zeven AC (1965) The origin of the oil palm (Elaeis guineensis Jacq.). A Summary. JW Afr Inst Oil Palm
Res 4: 218-225
I1l8 Somatic Embryogenesis in Rubber Tree
(Hevea brasiliensis MOO. Arg.)
M.P. CARRON, H. ETIENNE, N. MrcHAuX-FERRIERE, and P. MONTORO l
1 Introduction
Hevea brasiliensis Miill. Arg. (Family Euphorbiaceae), usually called rubber tree,
may reach 25 m in height and I m in trunk diameter at I m above ground.
Plantations for the production of natural rubber cover about 7 million ha mainly
in Southeast Asia. As such, it represents an artificial forest as large as the
Amazonian forest (Compagnon 1986). Ninety-three percent ofthe total produc-
tion comes from Asia; Indonesia, Thailand, and Malaysia each produces about
25% of the total annual production (RSB 1992). The consumption (70% for tires)
is distributed in industrial countries.
F or more than 50 years plantations were made with grafted clones. Propaga-
tion by grafting onto seedling rootstock led to low vigor, compared to elite
seedlings and maintained intraclonal heterogeneity which decreased the field
productivity. Agronomical features and problems, together with the possible
contributions of in vitro culture, were discussed earlier (Carron et al. 1989). As
cutting does not work for selected rubber clones, somatic embryogenesis appears
the best way for micropropagation.
Somatic embryogenesis in Hevea has been reported from severallaborato-
ries (Wanget al. 1980; Wan et al. 1982; Carron and Enjalric 1985). However, the
technique is still not sufficiently controlled to be utilized by breeders or growers:
reliable somatic embryo formation was limited to only a few genotypes, embryo-
genic capacity was short-lived, and there was a very low rate of conversion of
embryos into plantlets.
Since 1984--1985 our studies have been aimed at improving the understand-
ing of the characteristics of plant material in culture and the effect of different
culture medium parameters on its development. Separate use of the data ob-
tained in these studies have already helped in controlling the somatic embryogen-
esis process. This review plans to integrate these specific findings in such a way as
to increase the overall understanding of callus, induction, development, matura-
tion, and germination of somatic embryos, thus allowing the production of
plantlets from suspensions of embryogenic cell clusters.
I CIRAD-Cultures Perennes, Programme Hevea, Laboratoires BIOTROP, BP 5035, 34032

Montpellier Cedex, France

©Springer-VerJag Berlin Heidelberg 1995
354 M.P. Carron et al.
F
Somatic Embryogenesis in Rubber Tree (Hevea brasiliensis Miill. Arg.) 355
2.1 Plant Material and Basic Conditions for Culture Studies
The explants consist of slices of internal integument of seed from immature fruit
(about 8-10 weeks after anthesis, Fig. 1A). This a priori sterile tissue does not
require surface sterilization (Carron and Enjalric 1985). Experiments were
carried out with clones PB 260, PR 107, PB 235, or RRIM 600. Callus was
initiated on MH1 medium (Carron and Enjalric 1985) supplemented with
80 gil sucrose, 9 J.lM each of 3,4-D and benzyladenine (BA), 10 J.lM AgN0 3
(Auboiron et al. 1990) and solidified with 2 gil Gelrite. Three phases were
distinguished: (1) the initiation of embryogenesis (day 0 to 25); (2) a phase of
embryogenesis expression (day 26 to 50), in which the standard medium was
supplemented with 50 IlM spermidin (El Hadrami et al. 1989a); and (3) a phase
of embryo development during which AgN0 3 was eliminated and the
concentrations of BA and 3,4-D were reduced to 0.9 J.lM each. Callus
development was carried out in the dark at 27°C, in glass tubes (150 x 25 mm)
closed with polycarbonate caps.
Methodology of analyses has been described for histology (Michaux-
Ferriere and Carron 1989; Michaux-Ferriere et al. 1992), atmospheric gases
in culture vessels (Auboiron et al. 1990), growth regulators (Nangah 1986;
El Hadrami et al. 1991; Etienne et al. 1991a), polyamines (El Hadrami et al.
1989a,b), water parameters (Etienne et al. 1991a,b), mineral nutrition (Etienne
et al. 1991b), and for callus browning (Housti et al. 1991, 1992).
2.2 Primary Culture
Internal integument is genetically identical to the mother clone; it is a parenchy-

matous and highly vascularized tissue. After the start of culture, the callus is
initiated by proliferation of perivascular cells close to the excision zone, on the
inner surface of the internal integument, near the nucellus (Michaux-Ferriere and
Carron 1989; Fig. 1A,B). From these data we are able to adjust explant cutting
precisely in order to obtain fragments of about 5 x 3 x 0.5 mm with a fringe of
nucellus along one edge.
~--------------------------------------------------
Fig. lA-F. A Longitudinal section in the immature fruit of rubber tree at the right stage for the
initiation of the culture. NU Nucellus; IT inner integument; ET external integument; bar = 8 mm
(Carron 1992). B Histological structure of the inner integument of the seed and callus initiation from
the multiplication of perivascular cells (PV) and layers close to the nucellus (NCl), 10th day after the
start of the culture; bar = 150 11m (Escoute 1992). C Embryos (E) and (PE) on compact callus from
clone PB 260. Day 50. Bar = 1.25 mm (Etienne, 1991). D Anatomical aspect of compact callus.
Proembryos originate from such lobes of embryonic cells (arrows). Day 30. Bar = 1 mm (Escoute
1992). E Friable callus from clone PR 107 bearing embryos, at day 75; bar = 3 mm (Montoro 1992).
F Histological structure of young proembryos originating from first divisions of embryogenic cells in
a friable callus from clone PB 260; day 30; bar = 25 11m. (Photograph by J Escoute 1992). Periodic acid
Schiff-naphthol blue black was used in B, D, F
94 -0.4
, , 'I' callus
92 .. , ' \"'" ,, -0.5
~
, ::c:
'<
~ WCcallus \ -0.6 §:
'-'
"E 90 \"
,, ()
....r::
II)
, -0.7
"d
0
~
0
()
.... 88
',J; __ _ g.
'f -_ e:..
~
....
II)
--- --- -0.8 s=

~
'"0
po
'l'medium
86 --- -0.9
'-'
84 -1
1000 100
""" .................. ..
~
~ 800 .... 80
Q
...... X, "" s=
{JQ
~
~I "" tv
+I
'0
a , 600 , ".. \" ..
60
:::l
'-' z "EJ----------- ~
,,
"E , npo
....r::
~
II)
400 40 tv
0
() +I
,,
*
"a
.... +
II) ::.::
r:: 200 .............. . ,., 20
~
0 0
o 5 10 15 20 25
Time of culture (days)
Fig. 2. Related evolutions of the water status of the callus, i.e., water content WC and hydric potential
('1'), and the variations in potassium (K+), magnesium (Mg2+), soluble nitrogen (NwJ )' and calcium
(CA'+) contents during the first culture (days 0 to 23). Values are means ± SE (n = 4) for water
parameters. (Etienne et al. 1991 b)
After initiation of the culture, there is generally a 5-10 day latency phase
before the first cell divisions are observed; some parenchyma cells already begin
loading oxidized polyphenols.
Water parameter analyses reveal that explants with a high water potential
(-0.4 MPa) had been placed on culture medium with a lower water potential
(-0.9 MPa); several days are required to establish a water balance, which is
crucial for nutrient exchange between the culture medium and the explant.
During the 10 days after initiation of the culture, the water content of the explant
drops from 92 to 86% and the water potential from -0.4 to -0.75 MPa. This
water flow is accompanied by desorption of minerals (notably Mg, Ca, K, and N)
from the explant into the culture medium (Fig. 2; Etienne et al. 1991 b). A rapid
pretreatment (10 min) to partially dry the explant reduces the delay in callus
initiation, improves the embryogenesis rate, and reduces browning.
2.3 Initiation of Embryogenic Callus
After the latency phase which immediately follows the beginning of the culture,
the explant enters a highly active phase and forms a callus of essentially
meristematic cells. Pre-embryogenic cells begin to appear after day 15 (Michaux-
Ferriere and Carron 1989); they show one or several of the typical characteristics
of embryogenic cells.
Analysis of atmospheric gases during this period shows a high rate of CO 2
synthesis (thus indicating intense cell activity) along with a significant rate of
C 2 H 4 synthesis. This latter factor was found to have an adverse effect on the
development and maintenance of callogenesis. The elimination of atmospheric
ethylene, using either culture vessels or chemical traps such as KMn04' is
efficient. Much more remarkable results were obtained with the inhibition of
endogenous synthesis by aminooxyacetic acid (AOA), or by using a competitive
inhibitor, such as silver ion, active at the ethylene receptor sites (Auboiron et al.
1990; Housti et al. 1992).
Callus initiation is accompanied by highly stimulated endogenous synthesis
of polyamines (putrescine, spermidine, and spermine). The intensity of this
stimulation is related to the subsequent embryogenesis (El Hadrami et al. 1989b);
clonal differences have been observed in this respect, whereas the initial
polyamine content in the explant was similar and generally low in all of the ten
clones investigated (El Hadrami et al. 1989a).
These results on ethylene and polyamines are in line with data on the
interaction between the two metabolisms reported earlier by Ben-Arie et al.
(1982) and Fuhrer et al. (1982). These authors explained the action of polyamines
on membrane stabilization as being due to tissue inhibition of ethylene synthesis.
More recent studies (Kende 1989) favor the hypothesis that polyamines interfere
directly with the key enzyme of ethylene biosynthesis, l-amino-cyclo-propane
acid (ACC) synthase, whose activity depends on the disturbance of membrane
integrity.
Growth regulator types and concentrations are important parameters for
callus development. In several successive studies, still aimed at obtaining calli
that are more active and less subject to browning, we were able to progress from
using 9 J.1M 2,4-dichlorophenoxy-acetic acid (2,4-D) to an association of 1.3 J.1M
2,4-D + 5.7 J.1M indolyl-acetic acid (lAA) (Carron and Enjalric 1985) and then to
9 J.1M 3,4-dichlorophenoxy-acetic acid (3,4-D), a "weak" auxin (Michaux-
Ferriere and Carron 1989); auxin was used along with 9 J.1M BA. More recently,
the overall growth regulator concentration has been reduced to 4.44 or 2.22 J.1M
(El Hadrami etal. 1991; Etienne et al. 1991a) and we have demonstrated thatthe
use of kinetin instead of BA in this first culture medium strongly promotes the
induction of somatic embryogenesis and the subsequent conversion of
the embryo into plantlets (Montoro et al. 1992). Variations in the response
occurred at various concentrations according to the clone and the season, in
other words, according to the physiological state of the fruits. It almost seems
that steady improvements in culture conditions make it increasingly "easier" to
control cell reactivity. Recent studies in carrots, performed without any growth
regulators, suggest that this is indeed the case (Smith and Krikorian 1990).
The type of medium support is also an important culture parameter. We
determined that the use of different supports (e.g., Labosi agar, Difco agar,
agarose, Gelrite or cellulose plugs soaked in a liquid medium) led to clearly
different results with respect to the structure, browning tendency, and embryo-
genic potential of the callus (El Hadrami et al. 1992). Gelrite is a beneficial
support for Hevea; bound polyphenol oxidase activity and phenolic compounds
levels were markedly lower and superoxide dismutase activity was higher
between days 20 and 40, stage of somatic embryos formation, than with the other
supports. These activities thus appear to be related to reduced callus browning
and enhanced embryogenic potential obtained with Gelrite, especially when used
with a low hormone concentration. Agars, which are less efficient, have irregular
compositions and are thought to contain inhibitory substances (Kohlenbach and
Wernicke 1978), which might explain why they promoted browning in our
cultures.
Around day 25 the callus state is optimal; with no subculturing or after
subculturing on an unsuitable medium, areas of parenchymatous cells that are
more or less loaded with polyphenols develop, meristematic activity decreases,
the pre-embryogenic cells degenerate, and the calli become brown and necrotic
(Michaux-Ferricre and Carron, 1989). With no subculturing, culture aging leads
to very high C2H 4 production, which as previously mentioned, is detrimental to
tissue development. If the culture vessel is not airtight, which is preferable,
ethylene does not accumulate, however, evaporation of the medium causes a
drastic reduction in its water potential. The callus thus becomes dehydrated,
which is incompatible with embryogenic development. Subculturing on an
unsatisfactory medium, with a very different water potential or poor growth
regulators for instance, leads to the same type of degeneration, but takes a few
weeks longer. Histological analysis of the callus shows that even with a good
culture medium subculturing is stressful for the callus: the beneficial effects of
renewing the medium are only observed 4-5 days after subculturing, with a
resumption ofmeristematic activity.
Most stresses of the culture led to browning of the callus. Disorganization of
the oxidative metabolism has been correlated with browning of the callus (Housti
et al. 1991). The original explant displays high catalase (CAT) and NADH-
quinone-reductase (NQR) activity, whereas superoxide dismutase (SOD) activ-
ity is low, and peroxidase (PO) and polyphenol oxidase (PPO) activities are
almost nil. During the first culture, CAT and NQR activities concurrently
decrease, while SOD, PO, and PPO activities rise slightly: the overall balance is
maintained: NQR activity produces O;-ions converted by SOD into HP2' which
is then transformed by CAT into HP2' In contrast beyond day 25, callus
browning sharply increases with no subculturing or after subculturing on an
unsatisfactory medium, CAT activity is rapidly stalled, SOD activity remains
steady, and NQR activity increases. HP2 accumulates and reacts with 0;- to
form OH ions, which are particularly aggressive against unsaturated fatty acids
that form the phospholipid membrane. This leads to cellular decompartmenta-
tion; thus, the oxidative systems (essentially cytoplasmic) and their substrates
(generally vacuolar) come in contact with one another. There is a consequent
sharp increase in PO and PPO activity. It is therefore quite clear why the use of
antioxidants, although sometimes effective in reducing callus browning, never
restores cell activity.
25
,-..
~ 20
0
.....
I
O!l
0 15
E
-=-
Q) 10
>
~
~
p:) 5
~
0
0 I
,-..
«:S -0.4
0..
6
~ -0.8
C
Q)
(5
c.. -1.2
.~
....
"0
>.
=: - 1.6
Fig. 3. Comparison between hydric poten-

_2~--~----~--~----~---L
tial and ABA content in embryogenic vs
nonembryogeniccallus (day 50). Values are
onembryogenic Embryogenic
means ± SE (n = 4). (Etienne et al. 1991a) callus callu.
360 M.P. Carron et aI.
2.4 Embryogenic Expression
For somatic embryogenesis to occur, the callus must be given the means to
continue its cellular program initiated during the first culture. Water now
appears to be a crucial parameter for in vitro culture in Hevea: a close
relationship between callus embryogenic potential and a specific water status has
been recently demonstrated (Etienne et al. 1991a). Several effective techniques
have been developed to favor this water status. Water stress must be limited as
much as possible by reducing water loss, by controlling the relative humidity of
the culture room, and by regular subcultures onto media with a high water
potential (--0.8 MPa). Reduction in auxin and cytokinin concentrations relative
to the starting medium, and addition of abscisic acid (ABA) or polyamines, are
further possible means of promoting the embryogenesis process. We determined
that stimulation of embryogenesis is always accompanied by increased cell
turgidity and intense cell activity. It is associated, when ABA is added, with
increased absorption of minerals (particularly N, P, and K). These results are in
full agreement with the known role of polyamines in cell membrane protection,
of ABA in water stress resistance and, conversely, of auxin in enhancing cell
elasticity and permeability.
Embryogenic calli, in contrast to the nonembryogenic form, are able to
control their turgidity. Thus, irrespective of the treatment used to acquire
embryogenic capability, embryogenic calli are characterized by a high water
potential (--0.9 to -1.0 MPa) and high relative water content (93-94%; Fig. 3).
This difference may be observed between two treatments in which the embryo-
genic rates are very different and also within a single treatment between speci-
mens of nonembryogenic and embryogenic calli. Active absorption of minerals
has been shown in embryogenic calli. The water potential value of the callus,
which was equal to or slightly higher than that of the medium, indicated that the
flow of water and minerals takes place against osmotic laws which should highly
favor nonembryogenic calli.
The enhancement of the expression of somatic embryogenesis either by
lowering 3,4-D and BA concentrations or by adding ABA always resulted in a
decrease in endogenous ABA and an increase in IAA in callus (Etienne et al.
1992). Thus, low endogenous ABA (2-4 nmollg dry wt.) and high IAA (1-2
nmollg dry wt.) appear to be necessary for the acquisition and conservation of
an embryogenic state. The measurement of a high level of ABA (25-40 nmollg
dry wt.) in nonembryogenic callus confirms the hypothesis that this Hevea callus
is water stressed.
Embryogenic calli are also characterized by higher polyamines (El Hadrami
et al. 1989b) than the nonembryogenic ones. Browning is only a secondary
characteristic which is not antagonistic to somatic embryogenesis, but it does
hinder tissue maintenance after a certain stage.
2.5 Origin of Somatic Embryogenesis
Histological monitoring of calli has shown that embryogenic potential is

expressed at the end of the initial culture days 20-25 in two different forms
(Michaux-Ferriere et al. 1992):
1. The first form involves isolated clumps of embryogenic cells; some of these
cells subdivide to form small globular proembryos. This is embryogenesis
having a unicellular origin.
2. The second form embryogenesis having a multicellular origin, starts from
lobes on the surface of the callus. These parenchymatous lobes progressively
acquire specific embryonic characteristics and show intense metabolic activity,
expressed cytologically as very dense cytoplasm, highly stained with basic
dyes, and a very large nucleole. These sites, which can be termed globular
proembryos, considering their subsequent development, are always bound by
two or three layers of starch-rich vacuolated cells; these cells burst and are cut
off as soon as growth and epidermization of the proembryos begin and the
pro cambial bundles establish.
Our first success in obtaining somatic embryogenesis using internal integu-
ment of immature seed from clone PB 260 revealed that the resulting embryos
were of unicellular origin. At present using the same clone and the same type of
explant, the embryogenic rate can be enhanced by modifying certain culture
conditions while favoring the multicellular route; the unicellular process is
inhibited by the beginning of the second culture. The same culture conditions
were found to be effective for the induction and expression of somatic embryo-
genesis in clones PB 235, RRIM 600, and PR 107. However, considering the
morphological characteristics of the calli and embryos, only the multicellular
route seemed to be favored in clones PB 260 and PB 235, and only the unicellular
route in clones RRIM 600 and PR 107.
The simultaneous presence of these two forms of somatic embryogenesis in
a single callus has also been demonstrated in the oil palm (Schwendiman et al.
1990). Moreover, our results show that preferential development of either form
of embryogenic expression is dependent on a genotype/medium interaction and
thus can be controlled. If taken into account, these results could have an
important effect on the conformity of plant material produced for micropropa-
gation and on genetic transformation projects.
2.6 Proliferation of Embryogenic Cultures
Friability of the callus is a prerequisite for the multiplication of embryogenic

callus through subculture. Compact embryogenic calli were spontaneously
formed in three clones (PB 260, PB 235, and GTl; Fig. I C,D) on the basic
conditions of culture, while friable and embryogenic calli were formed in two
other rubber tree clones (PRI07 and RRIM 600; Fig. IE). Callus friability was
induced in clone PB 260 when the concentration of one growth factor (3,4-D or
362 M .P. Carron et al.
'"
~
E
<> 0.5
'0
,....., c..
~
0..
0 0
6en ::s
... Cl)
.:!
B
Q)
-0.5
E El
....0::10::1 -I
c.. .S!
'0
u E
·c '"
0 - 1.5
'0
;>,
:r: 0
,~ -2
"Cl
>.
:I:
-2.5
Sucrose level (mM) 234 351 234
Calcium level (mM) 3 3 12
Compact Friable callus
Fig. 4. Water status of two types of callus (day 50) depending on sucrose and calcium supply in the
medium. Values are means ± SE (n = 4)
kinetin) was reduced from 4.5 to 0.45IlM during the first culture, or when high
sucrose (351 mM) or calcium (12 mM) levels were sustained during subcultures
(Montoro et al. 1993a). However, friable calli obtained by modifying the
auxin/cytokinin balance lost their embryogenic potential. In contrast, those
obtained on a high sucrose level were mainly composed of embryogenic cells and
globular proembryos embedded in a mucilaginous matrix (Fig. IE,F). They
displayed high turgor pressure and low osmotic potential which are favorable
characteristics for somatic embryogenesis (Fig. 4). Moreover, this treatment
stimulated the accumulation of starch in embryonic structures. The reliable
Fig. 5. A Suspension of embryogenic cell clumps from clone PB 260. MicrocaIli give rise to a final
suspension of multicellular aggregates; bar = 9 mm (Photograph by M Lartaud 1993). B Embryos
from clone PB 260 at the cotyledonary stage of development (day 80); bar = I mm (Photograph by
Etienne 1992). CHistological structure of immature embryo from clone PR 107. Apical meristem is not
yet fonned and cells are highly vacuolized with few reserves; bar = 0.5 mm (Photograph by Escoute
1992). DHistological structure of mature embryo from clone PR 107; note profuse starch reserve (red);
bar = 0.5 mm (Photograph by J Escoute 1992). EYoung plantlet from somatic embryo of clone PR 107;
bar = 7 mm (Photograph by J Etienne 1992). F Field trial of somaplants from clone PR 107, 4 months
after planting (Photograph by Leconte 1992). Periodic acid Schiff-naphthol blue black used in C, D. C
cotyledon, RM root meristem, A M apical meristem
Somatic Embryogenesis in Rubber Tree (Hevea brasiliensis Mull. Arg.) 363
B o
F
induction of friability with a high calcium supply has been analyzed (Montoro
et al. 1995). This appeared like callus from a high sucrose supply according to
embryogenic cells and mucilaginous matrix, but water status and starch accumu-
lation were not so favorable for somatic embryogenesis. Increased calcium led to
proportional enhancement of calcium nutrition which directly stimulated the
excretion of polysaccharides and could explain the gelification of the middle
lamella and the decrease in turgor pressure. It also led to a drop in nitrate and
potassium absorption. These modifications strongly affect amino acid and
organic acid synthesis and also osmotic regulation.
Friable callus is easily disaggregated from a high calcium medium in agitated
liquid medium. Suspension cultures were then established from three clones: PB
260, PR 107, GTl (Fig. 5A). The embryogenic capacity of these cultures has been
sustained for more than 1 year. The expression of embryogenesis needs to
consider the transfer of cell clumps on solid medium. Embryogenic rates were 6,
12, and 7%, respectively, for clones PB 260, PR 107, and GTI. Plantlets were
obtained from clones PB 260 and PR 107 with high germination rates (respec-
tively 83 and 59%), but not with GTl whose embryos do not grow (Montoro
et al. 1994).
2.7 Development, Maturation, and Germination
Ontogenic, hormonal, water, and histological parameters of somatic embryos

were systematically compared to those of the zygotic embryos used as reference.
Two successive phases - development and maturation - were thus defined during
ontogenesis.
The development phase led the embryos on the callus to the cotyledonary
stage (Fig. 5). They were then markedly different to their zygotic counterparts
(Etienne et al. 1993a). The dry weight of these somatic embryos was much
smaller than that of the zygotic embryo of the same clone (PB 260), mainly due
to the atrophy of the cotyledons. They were distinctly more turgid with high
turgor pressure and relative water content. Histological observation showed that
root and shoot meristems were not yet formed. Somatic embryos at this stage
mainly consisted of extremely differentiated vacuolized cells which did not
contain reserves (Fig. 5C). These characteristics confirm that poor germination
and conversion into plantlets are related to immaturity. Moreover, the hormonal
content of somatic embryos did not display the kinetics shown in zygotic
embryos, i.e., a peak ofIAA (20 nmol/g dry wt.) during the development phase
and a peak of ABA, 25 nmol/g dry wt.) at the end of this phase. In comparison,
IAA and ABA contents in somatic embryos were low (0.6- 2 nmol/g dry wt.) and
varied little during ontogenesis (Etienne et al. 1993b; Fig. 6). These low hormonal
contents might indicate why additional 3,4-D during expression and develop-
ment of embryogenesis enhances the number of well-developed embryos and
why ABA supplementation in maturation medium stimulates both embryo
germination and conversion into plantlets.
For the maturation phase, slow desiccation or 351 mM sucrose supple-
mented with 1 f..lM ABA strongly improved germination and conversion of
Fig. 6. Comparison of the ratio ABN Development Maturation

IAA at different stages of ontogenesis of 8
somatic and zygotic embryos
7
6
5
Zygotic embryo
. . ...... .
. __ ... __
4
3
2
0
...
.~
<t: 0
<t: 8
~
IX) 7
<t: 6
5
Somatic embryo
4
3
2
III IV V VI VII
Ontogenic stages
embryos into plants (Etienne et al. 1993a). The second treatment was the most
effective, increasing the germination frequency by 3.7 (77%) and plant conversion
by 6.6 (33%) in clone PR 107. Each of these two treatments increased the vigor of
somatic embryos and stimulated the formation of root and shoot meristems and
the accumulation of starch and protein reserves. At the end of the maturation the
Hevea somatic embryos looked like the mature zygotic embryos (Fig. 5D).
Likewise, the two treatments brought the water status of somatic embryos closer
to that of the mature zygotic embryos, but without achieving a perfect match.
Optimization of the successful conversion into plants may require the full
acquisition of this water status (Fig. 5E).
Biochemical and histological characteristics of calli and the formed embryos

during somatic embryogenesis in Hevea were studied at various stages. These
results supplement data from the literature and will thus be beneficial in
Table 1. Somatic embryogenesis in Hevea
Short method Long method
1st culture: induction of Explant = slice of internal 1st culture: initiation of friable
embryogenesis, days 0 to 25 integument of imma- embryogenic callus, days 0 to 25
ture seed (8-10 weeks
after anthesis)
Explant is placed on mineral ------0> Explant is cultivated on mineral

salts and vitamins of MH, 30 salts and vitamins of MH but 9 mM
JlM AgN0 3, 234 mM sucrose, CaCI2, 292 mM sucrose, 30 JlM
4.44 JlM 3,4-D and kinetin, AgN03, 4.44 JlM 3,4-D and kinetin,
2 gil Gelrite, pH 5.8, 27 ·C 2g1l Gelrite, pH 5.8, 27 ·C and
and darkness. darkness.
2nd culture: expression of Compact Friable 2nd culture: setting up of the

embryogenesis, days 26 to 50 callus callus suspension of cell clusters, days 26
......- ------0> to 50
Callus is subcultured on the Callus (300-500 mg) is dispersed in

same basic medium but 1.35 liquid medium (5 ml) with the same
JlM 3,4-D and BA and supple- composition as the previous culture
mented with 50 JlM spermidine but 3 mM CaCI2 and no AgN03.
and 5 x 10-3 JlM ABA. Halfliquid medium is renewed 3-4
every days.
3rd culture: development of the Callus bearing

proembryos, days 51 to 80 visible
proembryos
Callus is cultivated on the same ......-
medium as induction but 3,4-D
(1.8 JlM), BA (0.9 JlM) in place
of kinetin and no further AgN03.
4th culture: maturation of the Morpholo- Suspen- N subcultures: proliferation of the

embryos, days 81 to 105 gically normal sionof embryogenic aggregates suspension,
embryo iso- cell several months.
Embryos are placed in multiwellS lated at clusters
dishes with medium consisting of cotyledon ------0> 10 ml "fraction pipetable" (less
30% macroelements and 200% stage than 1.5 mm) of the previous suspen-
microelements of MS mineral ......- sion is subcultured in 10 ml of fresh
medium, vitamins MH, 351 mM medium. Halfliquid medium is
sucrose, 500 mgll activated renewed each week.
charcoal, ABA 1 JlM,
2 gil Gelrite pH 5.8.
5th culture: germination of the Isolated, mature Repetition of this process every 15
embryos, days 106 to 130 embryo days leads to maintenance and pro-
......- liferation of the suspension.
Embryo is cultivated in glass
tube with the same medium as
maturation but supplemented
with sucrose (146 mM) and
growth regulators (8.7 JlM GA3
only).
Somatic Embryogenesis in Rubber Tree (Hevea brasiliensis MiiJl. Arg.) 367
Table 1. (Contd.)
Short method Long method
6th culture: development of the Rooted Micro- For regeneration, microcalli are iso-
plantlet, days 131 to 156 embryo caJlus, lated from the suspension and
~ 2-4mm undergo the "short" process.
in
diameter
Embryos are transferred to the
light (12h112h, 50 f.lmollm2/s)
and sub-cultured on the same
medium with a lower sucrose
content (73 mM) and an increase Plantlets
in GA3 (28.9 f.lM). ~
The basic medium caJled MH, is composed of the major elements: NH4N0 3(20 mM), KN03(20 mM),
NaH2P04,HP (2 mM), CaCI2,2HP (I mM), MgS04,7H20 (3 mM); minor elements: H 3B03(150f.lM),
MnS04,Hp (100 f.lM), ZnS04,7HP (40 f.lM), CuS04,5H 20 (1,5 f.lM), Na2Mo04 ,2HP (I f.lM),
KI (5 f.lM), CoCI2,6HP (I f.lM), Na2S04 (1300 f.lM), FeS04 (100 f.lM), Na2EDTA (100 f.lM); organic
compounds: myo-inositol (300 f.lM), nicotinic acid (20 f.lM), pyridoxine HCI (3 J.lM), thiamine HCI
(2 J.lM), biotin (0,2 f.lM), Ca D( +)-pantothenate (1 J.lM), ascorbic acid (1 J.lM), choline chloride (1 f.lM),
L-cysteine (60 f.lM), glycine (5 f.lM), riboflavin (1 f.lM).
specifying any modifications required in the culture conditions to overcome

eventual deadlines. Somatic embryogenesis, which was problematic until
recently, can now be obtained systematically in most of the clones studied
(particularly PB 260, PB 235, PR 107, RRIM 600, GTl), with a high frequency
of embryogenic calli. Moreover, knowledge of the plant material will be useful in
perfecting a procedure to promote the development of embryos into complete
plantlets (the germination rate has recently been upgraded from 111000 to about
30%) and to stabilize the embryogenic potential, either by maintaining calli on
semisolid media or by obtaining cell cluster suspensions.
About 100 somaplants from clones PR 107, PB 260, and RRIM 600 were
planted in June 1992 for field trials in order to compare them with classical
budded clones (Fig. SF). This will give information on the positive value of
somaplants with regard to the vigor and yield.
In addition to the objective of developing a reliable and effective procedure,
understanding the characteristics of the culture material should help in dealing
with one of the major problems of this technique, Le., the uniformity in the plants
produced.
4 New Protocol for Somatic Embryogenesis in Hevea
Regeneration of somatic embryos from rubber can now be obtained according to different methods:
(1) a short method, which uses the rapid but fugacious formation- of embryos on compact
callus; (2) a longer method which involves the obtention and multiplication of friable, embryogenic
caJlus. The two processes are described in Table I.
References
Auboiron E, Carron MP, Michaux-Ferriere N (1990) Influence of atmospheric gases particularly

ethylene, on somatic embryogenesis of Hevea brasiliensis. Plant Cell Tissue Organ Cult 21: 31-37
Ben-Arie R, Lurie S, Mattoo AK (1982) Temperature-dependent inhibitory effects of calcium and
spermine on ethylene biosynthesis in apple discs correlated with changes in microsomal membrane
microviscosity. Plant Sci Lett 24: 239-247
Carron MP, Enjalric F (1985) Somatic embryogenesis from inner integument of the seed of Hevea
brasiliensis (Kunth) Mull.Arg. CR Acad Sci Paris 300: 653-658
Carron MP, Enjalric F, Lardet L, Deschamps A (1989) Rubber (Hevea brasiliensis Mull.Arg.). In:
Bajaj YPS (eds) Biotechnology in agriculture and forestry, vol 5. Trees II. Springer, Berlin
Compagnon P (1986) Le caoutchouc naturel. Biologie, Culture, Production. Maisonneuve et Larose,
Paris
EI Hadrami I, Carron MP, d' Auzac J (1989a) Clonal variability of the embryogenic potential of Hevea
brasiliensis; relations with polyamines and peroxidases in calli. CR Acad Sci Paris 308: 299-305
EI Hadrami I, Michaux-Ferriere N, Carron MP, d'Auzac J (1989b) Polyamines, a possible limiting
factor in somatic embryogenesis of Hevea brasiliensis. CR Acad Sci Paris 308: 205-211
EI Hadrami I, Carron MP, d'Auzac J (1991) Influence of exogenous hormones on somatic
embryogenesis in Hevea brasiliensis. Ann Bot 67: 511-515
EI Hadrami I, Housti F, Michaux-Ferriere N, Carron MP, d'Auzac J (1992) Effects of gelling agents
and liquid medium on embryogenic potential, polyamines and enzymatic factors in browning in
Hevea brasiliensis calli. J Plant Physiol141: 230--233
Etienne H, Berger A, Carron MP (1991a) Water status of caJlus from Hevea brasiliensis during
induction of somatic embryogenesis. Physiol Plant 82: 213-218
Etienne H, Montoro P, Carron MP (1991b) The effect of water parameters on the development of
Hevea brasiliensis calli in in vitro culture. Ann Sci For 48: 253-265
Etienne H, Sotta B, Montoro P, Miginiac E, Carron MP (1992) Relations between exogenous growth
regulators and endogenous indole-acetic acid and abscisic acid in the expression of somatic
embryogenesis in Hevea brasiliensis. Plant Sci 88: 91-96
Etienne H, Montoro P, Michaux-Ferriere N, Carron MP (1993a) Water and histological parameters
during the maturation of Hevea brasiliensis somatic embryos: effects of desiccation, medium
osmolarity and abscisic acid. J Exp Bot 44(267): 1613-1619
Etienne H, Sotta B, Montoro P, Miginiac E, Carron MP (1993b) Comparison of endogenous ABA and
IAA contents in somatic and zygotic embryos of Hevea brasiliensis Mull.Arg. during ontogenesis.
Plant Sci 92: 111-119
Fuhrer J, Kaur-Sawhney R, Shih LM, Galston AW (1982) Effects of exogenous 1,3-diaminopropane
and spermidine on senescence of oat leaves. II. Inhibition of ethylene biosynthesis and possible
mode of action. Plant Physiol 70: 1597-1600
Housti F, Coupe M, d'Auzac J (1991) Enzymatic factors in browning in vitro and the embryogenic
capacity of Hevea brasiliensis caJlus. CR Acad Sci Paris 313: 461-466
Housti F, Coupe M, d'Auzac J (1992) Effect of ethylene on enzymatic activities involved in the
browning of Hevea brasiliensis callus. Physiol Plant 86: 445-450
Kende H (1989) Enzyme of ethylene biosynthesis. Plant Physiol 91: 1-4
Kohlenbach HW, Wernicke W (1978) Investigations of the inhibitory effect of agar and the function of
active carbon in anther culture. Z Pflanzenphysiol 86: 463-72
Michaux-Ferriere N, Carron MP (1989) Histology of early somatic embryogenesis in Hevea
brasiliensis: the importance of the timing of subculturing. Plant Cell Tissue Organ Culture 19:
243-256
Michaux-Ferriere N, Grout H, Carron MP (1992) Origin and ontogenesis of somatic embryos in
Hevea brasiliensis. Am J Bot 79: 174-180
Montoro P, Etienne H, Carron MP, Nougarede A (1992) Effect of cytokinins on the induction of
embryogenesis and the quality of somatic embryos in Hevea brasiliensis MuII.Arg .. CR Acad Sci
Paris 315: 567-574
Somatic Embryogenesis in Rubber Tree (Hevea brasiliensis Miil1. Arg.) 369
Montoro P, Etienne H, Carron MP (1993a) Callus friability and somatic embryogenesis in Hevea
brasiliensis. Plant Cell Tissue Organ Cult 33: 331-338
Montoro P, Etienne H, Michaux-Ferriere N, Carron MP (1994) Embryogenese somatique entretenue
chez Hevea brasiliensis. Rev Cytol Bioi veget Bot 17: 113-119
Montoro P, Etienne H, Carron MP (1995) The role of calcium in callus friability and somatic
embryogenesis in Hevea brasiliensis Miill.Arg. Relations with mineral nutrition and water status of
the callus. J Exp Bot 46(282)
Nangah JC (1986) Etude de la callogenese chez l' Hevea brasiliensis Miill.Arg. These de doctorat,
Universite des Sciences et Techniques du Languedoc, France
RSB (1992) Production of natural rubber. Int Rubber Study Group (ed) Rubber Statistical Bulletin
London 47(2): 9-10
Schwendiman JC, Pannetier C, Michaux-Ferriere N (1990) Histology of embryogenic formations
during in vitro culture of oil palm Elaeis guineensis Jacq .. Oleagineux 45: 409-418
Smith DL, Krikorian AD (1990) pH control of carrot somatic embryogenesis. In: Nijkamp HJJ, Van
Der Plas LHW, Van Aartrijk J (eds) Progress in plant cellular and molecular biology. Kluwer,
Wan AR, Ghandimathi WY, Rohani H, Paranjothy K (1982) Recent developments in tissue culture of
Hevea. In: Rao AN (ed) Tissue culture of economically important plants. COSTED and ANBS,
Singapore, pp 152-158
Wang Z, Zeng X, Chen C, Wu H, Li Q, Fan G, Lu W (1980) Induction of rubber plantlets from anther
of Hevea brasiliensis (Miill.Arg.) in vitro. Chin J Trop Crops 1: 25-26
111.9 Somatic Embryogenesis in Walnut
(Iuglans Species)
W. TULEcKE 1, G.H. MCGRANAHAN 2 , and c.A. LESLIE 2
1 Introduction
As one of the world's most important nut crops, walnuts are produced in
amounts exceeding 800 000 metric tons annually and the wood from walnut trees
is valued for furniture, veneer, and gunstocks. For these products, walnuts are
cultivated in the United States, China, Turkey, France, and many other coun-
tries. The English or Persian walnut, Jugians regia, is native to the mountains of
central Asia. It is a member of the luglandaceae which includes black walnuts
(J. nigra, J. hindsii, J. major), butternut (J. cinerea), pecans and hickories (Carya
spp.), and wingnuts (Pterocarya spp.) (McGranahan and Leslie 1990).
Somatic embryogenesis and micropropagation are two relatively new tech-
niques which augment the methods of cutting, budding, and grafting tradition-
ally used to produce clones of walnut trees. These techniques are being used for
the selection of trees with disease resistance, the introduction of elite bioengi-
neered traits, special wood characteristics, and improved nut quality. Somatic
embryos from walnut are being used for clonal propagation (Tulecke and
McGranahan 1985), production of interspecific hybrids (McGranahan et al.
1986), triploids from endosperm (Tulecke et al. 1988), and introduction of
specific genes (McGranahan et al. 1988). These approaches serve to complement
conventional breeding programs for nut production, timber (Neuman et al.
1992), disease resistance, or other purposes such as walnut propagation
(McGranahan et al. 1987; lay-Allemand et al. 1991).
A dependable system for inducing somatic embryogenesis from the cotyledons of

immature embryos of several varieties of English walnut (J. regia) and California
black walnut (1. hindsii) was first reported in 1985 (Tulecke and McGranahan).
The system included repetitive embryogenesis (Figs. 1-4) from the cotyledons,
I Antioch College, Yellow Spring, OH 45387, USA

2 Department ofPomology, University of California, Davis, CA 95616, USA

Fig. 1. Globular stage somatic embryo showing the origin of a somatic embryo from the superficial
tissue of a somatic embryo
Fig. 2. Repetitive somatic embryogenesis: somatic embryos on the surface of the cotyledons of a
somatic embryo of cv. Scharsch-Franquette
Fig. 3. Somatic embryos and callus tissue originating from brown aging tissue derived from a somatic
embryo; this is another form of repetitive somatic embryogenesis
Fig. 4. A longitudinal section of a mature somatic embryo of cv Scharsch-Franquette showing
bipolarity, the vascular cylinder, leaf traces, leaf primordia, the primary root, and a portion of the
cotyledon
Fig. 5. Walnut trees derived from transformed walnut somatic embryos. The scions of somatic
embryos were micropropagated, then grafted on Paradox rootstocks and later planted in the field
Fig. 6. This walnut pistillate flower originated from cuttings grown in micropropagation. The cuttings
came from a somatic embryo from a zygotic embryo of a cross between two precocious cultivars from
China, 85- 8 x 85- 10
372 W. Tulecke et al.
hypocotyls, and roots of somatic embryos. Some of these embryogenic lines

continue to produce somatic embryos after more than 9 years. Trees from some
of these lines are more than 5 years old (Fig. 5).
2.1 Media
The culture medium used by Tulecke and McGranahan (1985) was based on a
medium devised by Driver and Kuniyuki (1984) for micropropagation. A
conditioning medium with I-glutamine, 6-benzylaminopurine, and indole-3-
butyric acid added to the basal medium was used for the initial explants of
cotyledon tissue. After 4-6 weeks the explants were transferred to basal medium
and thereafter maintained on it. The Driver Kuniyuki walnut medium (DKW)
was subsequently used for somatic embryogenesis of walnut by Cornu (1988,
1989), Deng and Cornu (1992), Jay-Allemand and Cornu (1986), Lee et al.
(1988), Liu and Han (1989) and their co-investigators. However, Neuman et al.
(1992) used the woody plant medium (WPM) of Lloyd and McCown (1980) as
their basal medium to initiate somatic embryogenesis from cotyledon tissue of
J. nigra, the eastern black walnut. They reported little success with the procedure
of Tulecke and McGranaham (1985) when applied to J. nigra. However, Cornu
(1988), and Deng and Cornu (1992) were able to use this procedure to obtain
somatic embryogenesis in J. nigra. In other studies, Yuan et al. (1990) used a
modified DKW medium and protocol to induce somatic embryos from imma-
ture embryos. They added I-cysteine, calcium pantothenate, biotin, and pyrid-
oxine to the medium and used somewhat different concentrations of other
media components. Also, some researchers used Gelrite and others used agar as
the solidifying agent for media.
Label and Cornu (1988) used an ELISA technique to analyze the liquid
endosperm of walnut obtained at the time when zygotic embryos were very
young. This was the same stage of endosperm development that gave somatic
embryos from zygotic embryos. They reported low levels of cytokinin and high
levels of IAA and abscisic acid. Their study provided a more rational basis for
medium design by giving the endogenous levels of plant growth substances in the
endosperm.
Deng and Cornu (1992) reported somatic embryos from hybrid zygotic
embryos of the cross J. regia x J. nigra and from J. major and J. nigra (Cornu
1988, 1989) using DKW as their basal medium. Leslie (unpubl. data) was also
successful in using DKW basal medium to obtain somatic embryogenesis from
zygotic embryo tissue as early as 3-4 weeks post-pollination without prior
treatment with conditioning medium.
Neuman et al. (1992) induced somatic embryos from cotyledons of J. nigra
by using WPM rather than DKW medium. They observed that the cotyledon
explants of J. nigra responded better to WPM than to DKW, similar to results
reported for pecan (Merkle et al. 1987). They used thidiazuron in the condition-
ing medium and noted that higher concentrations of thidiazuron produced
somatic embryos from somewhat older cotyledonary tissue, thus widening the
developmental window for the induction of embryogenesis for J. nigra.
Somatic Embryogenesis in Walnut (Jug/ans Species) 373
2.2 The Source of Explants
Cotyledons of zygotic embryos were routinely used in the original experiments

on inducing somatic embryogenesis in walnut (Tulecke and McGranahan 1985).
The cotyledons which were embryogenic varied from 10-44%, depending on the
cultivar or species. The optimum stage of cotyledon development for the induc-
tion of somatic embryos was 6-11 weeks post-pollination for the cultivars
Scharsch-Franquette, Payne, Early Ehrhardt, Plant Introduction 18256
(Manregian), J. hindsii, and Pterocarya (wingnut). Neuman et al. (1992) also
reported the greatest number of somatic embryos from cotyledon tissue of
1. nigra taken 8-12 weeks after pollination. Since the original publication
(Tulecke and McGranahan 1985), embryogenic lines were obtained from open
pollinated seeds of other cultivars of J. regia; these include Eureka, Sharkey,
Chandler, Sunland, Hartley, Cisco, and Vina. In addition, ten somatic embryo
clones were obtained from controlled crosses of Pedro X56-224, two precocious
accessions from China (85-8 x 85-10), Chandler x 85-8, Cisco x Sunland, and
Chandler selfs.
Liu and Han (1989) and Liu et al. (1992) (pers. comm) reported on the
formation of embryoids from petiolar callus but these did not reach full embryo
development or growth into plants.
Considering the Juglandaceae as a group, there has been fairly widespread
success with the induction of somatic embryogenesis in cotyledon tissue. Several
species of Jug/ans (J. regia, J. nigra, J. major, J. hindsii), several hybrids among
these, many cultivars of J. regia and Pterocarya have been induced to form
somatic embryos. In addition, somatic embryogenesis has been achieved for
Carya illinoensis, the pecan (Merkle et al. 1987; Corte Olivares et al. 1990; Yates
and Reilly 1990).
2.3 Origin
Somatic embryos were reported to originate primarily from epidermal cells

(Polito et al. 1989; Fig. 1). Globular, heart-shaped, and cotyledonary somatic
embryos were routinely observed and other types of abnormal development were
common, including hypertrophy, vitreous embryos, those with double apices,
fused, neomorphs and reversion to callus (Tulecke and McGranahan 1985).
These abnormalities were assumed to & related to the different stages of
development of surrounding embryos or the effect of culture conditions and
media. These observations were made on somatic embryos derived from the
cotyledons or other embryogenic tissues of zygotic embryos. The development of
triploid somatic embryos from endosperm came from endosperm-derived callus
(Tulecke et al. 1988).
In repetitive somatic embryogenesis the somatic embryos, or the tissues
derived from them, frequently gave rise to other somatic embryos, thus forming
an embryogenic line. Thus far, somatic embryos of walnut have been obtained
from cotyledon tissue, hypocotyl, primary root, root cap, and endosperm callus
(Tulecke et al. '1988). Brown aging tissue masses which develop from some of
these organs or tissues also produce vigorous white somatic embryos or callus
(Fig. 3). Some of these embryogenic cultures, such as the original one from
J. regia cv. Scharsch-Franquette which was started in 1983, remain embryogenic
after 9 years in culture. These embryogenic lines are maintained as a source of
somatic embryos by culturing continuously in the dark; the cultures are exposed
to light only during transfer to fresh media or when embryos are selected.
2.4 Germination
Successful maturation and germination of somatic embryos requires that they

be carefully selected and nurtured. Somatic embryos showing typical walnut
embryos morphology (Fig. 4), an ivory white appearance which indicates an
accumulation of starch, and a size of 0.5-1 cm in length are selected and
transferred to basal medium for germination. A cold treatment of 2-4 °C for
8-10 weeks in the dark helps to overcome dormancy. With or without this
treatment some embryos will germinate when transferred to the light. The
cotyledons expand and become green, the primary root elongates and epicotyl
growth produces leaves. After 6 weeks, 5-8 cm long seedlings are transferred to
sterile peat plugs in half-strength liquid basal medium. After 1-2 months young
rooted plants are removed from the sterile environment, rinsed thoroughly in tap
water to remove all of the medium, especially sucrose, and placed in a mixture of
equal parts of sterile peat, vermiculite, sand, and soil, first in plastic cups and later
in plastic pots. The seedlings are watered three times per week with O.I-strength
basal salts and later with tap water. The plants are covered with plastic bags
which are gradually perforated to aid acclimatization ofthe developing seedlings,
including cuticle formation and hardiness to ambient fluctuations in humidity,
wind, and temperature. After 4 months of growth in soil the seedlings develop
into healthy plants with mature leaves, buds, bark, and an extensive root system,
resembling normal walnut seedlings. They are further acclimatized by trans-
planting to soil in the greenhouse, shade house, and then into the field, receiving
only water and soil nutrients. A more convenient method of somatic embryo-
derived plants from culture into the field is to micropropagate the shoots from a
germinated somatic embryo and then to graft this shoot on a seedling in the
greenhouse. The percent take of the grafts is not high, but the method moves
plants into the field more quickly and in a hardier condition (Leslie and
McGranahan 1992; see also Cornu and Jay-Allemand 1989).
Deng and Cornu (1992) developed a desiccation procedure to replace
the 2-month cold treatment to overcome dormancy. They also germinated
somatic embryos on cotton compresses saturated with liquid medium; they
obtained 45% germination on the liquid medium and 10% on the solid
medium.
Somatic Embryogenesis in Walnut (Jug/ans Species) 375
2.5 Transformation
Somatic embryos of repetitive embryogenic lines of walnut were used in the

process of gene transfer (McGranahan et al. 1988; Jay-Allemand et al. 1991). The
first report of the genetic transformation and regeneration of transgenic plants of
a woody crop species was in walnut (McGranahan et al. 1988). Genes of interest
that have been inserted into walnuts using this approach include genes encoding
the cryIA(c) endotoxin of Bacillus thuringiensis (Dandekar et al. 1994) and the
chalcone synthase antisense gene (Jay-Allemand et al. 1991). Walnut plants
expressing inserted genes have been developed by germinating transformed
embryos, micro propagating shoots from these embryos and then grafting the
shoots to seedling rootstocks in the greenhouse. Trees containing inserted genes
have been in the field for 2 years.
3 Summary and Conclusion
It is important to call attention to earlier work on zygotic embryos of walnut

by Cossio and Minotta (1983), Rodriguez and Sanchez-Tames (1981), and
Rodriguez (1982), as well related recent publications by Jay-Allemand and
Cornu (1986), Penuela et al. (1988), Rodriquez et al. (1989), and the work on
apomixis in walnut by Deryuzhkin et al. (1985). These studies contribute to the
general advance of our understanding of walnut embryogenic development.
Nevertheless, significant problems remain in achieving the objective of using
somatic embryogenesis as a procedure for cloning walnut cultivars and introduc-
ing specific genes into specific cultivars. Foremost among these problems is the
need to devise a protocol for obtaining somatic embryogenesis from such organs
as leaves or roots or from tissues such as cambium. Success in this approach has
been achieved in several woody plant systems (Tulecke 1987) such as coffee
(Hatanaka et al. 1991), so the appropriate media, sequences, and stages of
development need to be worked out for walnut and for other members of the
Juglandaceae. Experiments in this laboratory with nucellus, catkins, leaves,
roots, and unfertilized ovules cultured on several media have yielded no somatic
embryos (Aly et al. 1992). Driver Kuniyuki walnut medium (McGranahan et al.
1987) with several combinations of zeatin, 6-benzylaminopurine, thidiazuron,
kinetin, indole-3-acetic acid, indole-3-butyric acid, gibberellic acid, and
abscisic acid were tested with different cultivars. Callus, shoots and roots
were obtained in some experiments, but no somatic embryos. To date, no
reliable protocol is available for the induction of somatic embryos from
cultivar tissue.
References
Aly MAM, Fjellstrom, RG, McGranahan GM, Parfitt DE (1992) Origin of walnut somatic embryos
determined by RFLP and isozyme analysis. HortScience 27: 61-63
Cornu D (1988) Somatic embryogenesis in tissue cultures of walnut Jug/ans nigra, J. major and hybrids
J. nigra x J. regia. In: Ahuja MR (ed) Somatic cell genetics of woody plant. Kluwer, Dordrecht, pp
45-59
Cornu D (1989) Walnut somatic embryogenesis. Physiological and histological aspects. Ann Sci For
46: 133-135
Cornu D, Jay-Allemand C (1989) Micropropagation of hybrid walnut trees (Jug/ans nigra x Jug/ans
regia) through culture and multiplication of embryos. Ann Sci For 46: 113-116
Corte-Olivares J, Phillips GC, Butler-Nance SA (1990) Somatic embryogenesis from pecan zygotic
embryo explants. HortSci 25: 983
Cossio F, Minotta G (1983) Prove preliminari de colture "in vitro" di embrione isolati de noce (Jug/ans
regia L.) e confronto tra differenti combinizioni de sali minerali. Riv Ortoflorofrutt 67:
287-298
Dandekar AM, McGranahan GH, Vail PV, Uratsu SL, Leslie C, Tebbets JS (1994) Low levels of
expression of wild type Bacillus thuringiensis var. kurstaki cry IA (c) sequences in transgenic walnut
somatic embryos. Plant Sci 96: 151-162
Deng Ming De, Cornu C (1992) Maturation and germination of walnut somatic embryos. Plant Cell
Deryuzhkin RI, Ulykina MK, Khazova II (1985) Apomixis in nuts of the genus Jug/ans in Voronezh
province. BioI. osnovy selektsii rastenii Sbornuk nauchnykh trudov. 41--47
Driver JA, Kuniyuki AH (1984) In vitro propagation of Paradox walnut rootstock. HortScience 19:
507-509
Hatanaka T, Arakawa 0, Yasuda UchidaN, Yamaguchi T(l991)Effectofplantgrowthregulators on
somatic embryogenesis in leaf cultures of Coffea canephora. Plant Cell Rep 10: 179-182
Jay-Allemand C, Cornu D (1986) Culture in vitro d'embryons isoles denoyer commun (Jug/ans regia
L.) Ann Sci For 43: 189-198
Jay-Allemand C, Jouanin L, Deng MC, Claudot AC, Drouet A, Cornu D (1991) Transfer of chalcone
synthase antisense gene: new strategy for studying polyphenols involved in walnut rhizogenesis. In:
Dekouchkovsky Y (ed) Plant sciences today. INRA, Paris, 305 pp
Lable P, Cornu D (1988) Determination of plant growth substances in liquid endosperm of immature
walnut (Jug/ans nigra) nuts by an ELISA technique. Plant Growth Regul 7: 209-215
Lee BC, Shim SY, Lee SK (1988) Mass propagation and germination of somatic embryos in Jug/ans
regia L. (English walnut). Res Pep Inst For Gen Korea 24: 99-106
Leslie C, McGranahan G (1992) Micropropagation of Persian Walnut (Jug/ans regia L.). In: Bajaj
YPS (ed) Biotchnology in agriculture and forestry, vol 18. High-tech and micropropagation II.
Liu Shu Lan, Bi Wen (1989) Plant regeneration from excised embryos of Jug/ans regia. Acta Physiol
Sin 15: 98-100
Lloyd G, McCown B (1980) Commercially feasible micropropagation of mountain laurel, Ka/mia
/atifolia, by use of shoot tip culture. Proc Int Plant Prop Soc 30: 421--427
McGranahan GH, Leslie CA (1990) Walnut. In: Moore IN, Bellington JR (eds) Genetic resources
of fruit and nut crops, vol 2. ISHS, Wageningen, pp 907-951
McGranahan GH, Tulecke W, Arulsekar S, Hansen 11 (1986) Intergeneric hybridization in the
Juglandaceae: Pterocarya sp. x Jug/ans regia. J Am Soc Hort Sci III: 627-630
McGranahan GH, Driver JA, Tulecke W (1987) Tissue culture ofJuglans. In: Bonga JM, Durzan DH
(eds) Cell and tissue culture in forestry, vol 3. Martinus Nijhoff, Boston, pp 261-271
McGranahan GH, Leslie CA, Uratsu SL, Martin LA, Dandekar AM (1988) Agrobacterium-mediated
transformation of walnut somatic embryos and regeneration of transgenic plants. Biotechnology
6: 800-804
Merkle SA, Wetzstein HY, Sommer HE (1987) Somatic embryogenesis in tissue cultures of pecan.
Somatic Embryogenesis in Walnut (Juglans Species) 377
Neuman MC, Preece lE, Van Sambeek lW, Gaffney GR (1993) Somatic embryogenesis and callus
production from cotyledon explants of eastern black walnut (Juglans nigra L). Plant Cell Tissue
Organ Cult 32: 9-18
Penuela R, Garavito C, Sanchez-Tames R, Rodriguez R (1988) Multiple shoot-bud stimulation and
rhizogenesis induction of embryogenic and juvenile explants of walnut. Acta Hortic 227: 457-459
Polito VS, McGranahan G, Pinney K, Leslie C (1989) Origin of somatic embryos from repetitively
embryogenic cultures of walnut (Juglans regia L): Implications for Agrobacterium-mediated
transformation. Plant Cell Rep 8: 219-221
Rodriguez R (1982) Callus intiation and root formation from in vitro callus of walnut cotyledons.
Rodriguez R, Sanchez-Tames R (1981) Cultivo de tejidos y differenciacionen nogal. Rev Fac Cienc
Univ Oviedo 22: 21-28
Rodriguez R, Revilla A, Albuerne M, Perez C (1989) Walnut (Juglans spp.). In: Bajaj YPS (ed)
Biotechnology in agriculture and forestry, vol 5. Trees II. Springer, Berlin Heidelberg New York,
pp 101-125
Tulecke W (1987) Somatic embryogenesis in woody perennials. In: Bonga 1M, Durzan DH (eds) Cell
and tissue culture in forestry, vol 2. Martinus Nijhoff, Boston, pp 61-91
Tulecke W, McGranahan G (1985) Somatic embryogenesis and plant regeneration from cotyledons of
walnut, Juglans regia L Plant Sci 40: 57-63
Tulecke W, McGranahan G, Ahmadi H (1988) Regeneration by somatic embryogenesis of triploid
plants from endosperm of walnut, Juglans regia L cv. Manregian. Plant Cell Rep 7: 301-304
Wetzstein H Y, Ault 1 R, Merkle, SA (1989) Further characterization of somatic embryogenesis and
plantlet regeneration in pecan (Carya illinoensis). Plant Sci 64: 193-20 I
Yates IE, Reilly C (1990) Somatic embryogenesis and plant development in eight cultivars of pecan.
Yuan QP, Dong MS, lay-Allemand C, Huang QC Liu S (1990) Organogenesis and somatic embryo-
genesis from culture of walnut. Sci Silvae Sin 26
111.10 Somatic Embryogenesis in Western Larch
(Larix occidentalis)
P. VON ADERKAS1, R.G. THOMPSON2, M. ZAKI 3, and L. BENKRIMA4
1 Introduction
Western larch (Larix occidentalis Nutt.) is the tallest species in the genus Larix
(Harrison and Dallimore 1966). It produces a valuable lumber which is rated for
its mechanical stress capabilities (Balatinecz 1983). The species is fast-growing.
Its range extends from south-central British Columbia to Oregon and from
western Washington to Montana (Owens and Molder 1979). Throughout its
range it is a successional species, dependent on periodic fires for establishment. It
is restricted to altitudes between 600 and 1500 m in the northern parts of its range
and up to 2200 m in the southern portions of its distribution. Generally found
on north-facing slopes, it is also common in valley bottoms and in quite a variety
of forest types (Fowells 1965).
Although generally considered a good seed producer, western larch varies
greatly in consistency from year to year in particular locations, so much so that
there are almost constant seed shortages. In these areas a good seed crop may
only occur once a decade, which is a bottleneck to local reforestation efforts.
Somatic embryogenesis may provide a method of propagation which could
alleviate shortfalls in seed supply. Orchards for seed production and breeding
have only recently been implemented (Jaquish 1987).
Somatic embryos have been produced in various larch species from immature
and mature zygotic embryos, as well as from mature somatic embryos and
needles of plants produced by somatic embryogenesis (Table 1).
1 Centre for Forest Biology, Department of Biology, University of Victoria, Victoria, British
Columbia, Canada, V8W 2Y2
2Crop Development Centre, University of Saskatchewan, Saskatoon, Saskatchewan, Canada,
S7NOWO
3Department of Plant Sciences, Institute for Efficient Productivity, Zagazig University, Zagazig,
Egypt
4Celex Laboratories Inc., Suite 1300,409 Granville Street, Vancouver, British Columbia, Canada,
V6C lT2

©Springer-VerJag Berlin Heidelberg 1995
Somatic Embryogenesis in Western Larch (Larix occidentalis) 379
Table 1. Summary of work on somatic embryogenesis in the genus Larix
Species Stage Medium PGRs References
L. decidua Immature Modified 2,4-D Cornu and Geoffrion

zygotic embryo BMS BAP,KIN (1990)
L. decidua Immature 112 LM 2,4-D von Aderkas et al. (1990)

zygotic embryo BAP
L. decidua Immature Modified 2,4-D Klimaszewska (1989)

x L. leptolepis zygotic embryos MS BAP
L. leptolepis Immature 112 LM 2,4-D von Aderkas et al. (1990)

x L. decidua zygotic embryos BAP
L. leptolepis Immature 1I2LM 2,4-D von Aderkas et al. (1990)

zygotic embryos BAP
L. occidentalis Immature 1/2LM 2,4-D Thompson and von

zygotic embryos BAP Aderkas (1992)
L. occidentalis Mature zygotic 1I2LM 2,4-D von Aderkas et al.

embryos BAP (unpubl.)
L. decidua Mature zygotic MSG 2,4-D Lelu et al. (1994)

embryos BAP
L. decidua Immature MSG 2,4-D Lelu et al. (1994)

x L. /epto/epis zygotic embryos BAP
L. /eptolepis Immature MSG 2,4-D Lelu et al. (1994)

x L. decidua zygotic embryos BAP
L. lepto/epis Mature zygotic MSG 2,4-D Lelu et al. (1994)

x L. decidua embryos BAP
L. lepto/epis Cotyledons of MSG BAP (pretreat) Lelu et al. (1994)

x L. decidua somatic embryo BAP,2,4-D
L. lepto/epis Needles of MSG BAP (pretreat) Lelu et al. (1994)

x L. decidua emblings BAP,2,4-D
The technical details of somatic embryogenesis in western larch share many

features with other larch species. Induction medium used is very similar to that
used in other larch systems - a modified half-strength Litvay's medium (Litvay
et al. 1981). The other medium frequently used is MSG (Becwar et al. 1987).
Plant growth regulating substances include an auxin, 2,4-dichlorophenoxyacetic
acid (2,4-D), in combination with a cytokinin such as benzyl adenine (BA) or
kinetin (KIN). During maintenance KIN is used instead of BA, as long-term
growth on BA was deleterious to mature embryo formation. Pretreatment with
380 P. von Aderkas et al.
BA has been used to induce somatic embryogenesis from somatic embryo-

derived plants (Lelu et al. 1991). As with other species of conifers, results appear
to be strongly variable according to line.
Initially, we reported the induction of somatic embryogenesis from
immature embryos dissected from seed a few weeks after fertilization (Thompson
and von Aderkas 1992). In this review, embryogenesis from mature seed embryos
(supplied from 3- and 4-year-old frozen seed stores of the British Columbia
Ministry of Forests) is also discussed.
2.1 Induction from Immature Embryos
Thompson and von Aderkas (1992) reported that embryogenic tissue could be
induced from embryos dissected and prepared in a number of different ways. The
simplest method was to isolate the embryo. Coculture with the longitudinal half
of the megagametophyte was also described. Not mentioned in that paper, but
also attempted was whole megagametophyte culture as well as coculture with
only the micropylar half of the megagametophyte, as had been done previously
by Klimaszewska (1989). The only explant preparation which had a completely
inhibitory influence was whole megagametophyte culture; embryogenic tissue
never multiplied from the embryo inside. Single, isolated embryos placed directly
on the medium initiated proportionately more cultures (32%, n = 41) than
explants consisting of longitudinally opened megagametophytes containing
embryos (21%, n = 42). The greater initiation rate with excised embryos of
western larch was not statistically significant (Chi-square test of independence,
r} =1.13). Cornu and Geoffrion (1990) initiated embryogenic tissue from isolated
embryos of L. decidua. In previous work with L. decidua, L. leptolepis and their
reciprocal hybrids, embryogenic cultures were initiated from immature embryos
cultured with the micropylar half of the megagametophyte (Klimszewska 1989;
von Aderkas et al. 1990).
Somatic embryogenesis in western larch was obtained from embryos
cultured with the micropylar half of the megagametophyte. This method is
technically difficult, does not give higher initiation rates, and requires that the
ploidy of resulting cultures be established, as the haploid megagametophyte of
other larch species is known to produce embryogenic tissue (Nagmani and Bonga
=
1985; von Aderkas et al. 1990). Diploid lines were routinely 2n 24. As the
presence of the megagametophyte is not necessary for western larch, the excised
embryo technique is preferred.
Initiation of cultures (Fig. 1, 2) from immature embryos was apparent 5 to 7
days. Although the explants were observed regularly for 2 months, no new
cultures were initiated after the first week. After 1 month most cultures were 5 to
8 mm in diameter and the original explant was obscured by translucent tissue
consisting of embryonal masses and suspensors. In some lines distinct somatic
embryos were observed as early as 4 weeks. After 2 to 4 months of culture and
subculture many lines turned brown and ceased to grow. Others developed into
a pale green nonembryogenic callus which was hard in texture. Most of these
died, with the notable exception of two green calli which produced white
Fig. 1. Isolated western larch embryo on basal medium with 2,4-D and BA showing proliferation of
tissue from the cotyledons and upper portion of the hypocotyl; x370. (Fig. 1- 7are from Thompson and
von Aderkas 1992)
Fig. 2. Isolated embryo with embryogenic tissue originating from the cotyledons (arrow head) ,
hypocotyl and suspensor end of the embryo; x 192
Fig. 3. After several months of subculture on basal medium with 0.9 mM 2,4-D, green callus (GC)
derived from an isolated embryo produced white embryogenic tissue; x 90
Fig. 4. A somatic embryo at the early cotyledonary stage. Aring of cotyledons crowns the hypocotyl.
The darker region at the base of the hypocotyl is the site of the primary root meristem. A winding
suspensor is visible; x320
Fig. 5. A somatic embryo at the late cotyledonary stage. At the base of the embryo is portion of the
thinner suspensor. The darker point at the junction of the embryo and suspensor is the tip of the root
meristem; x 138
embryogenic cultures after subculturing on basal medium supplemented with

0.9 J.lM 2,4-D for several months (Fig. 3). The remaining sustainable embryo-
genic cultures have been maintained on medium containing 0.9 J.lM 2,4-D for
more than I year.
The sustainable cultures obtained from this work represent somatic embryo-
genesis from less than 5% of the total number of explants in the combined
initiation experiments. The very significant losses in the first few weeks and
months of culture underscore the importance of developing methods to maintain
the viability of embryogenic cultures in somatic embryogenesis is to become a
useful technique in this species.
2.2 Influence of Developmental Stage on Initiation
Embryos in the explant preparation study were obtained from cones collected
early in the growing season. Early cotyledonary embryos gave better results
(Thompson and von Aderkas 1992). This contrasts with previous findings in L.
decidua and Larix hybrids, where embryogenic cultures could be obtained only
from precotyledonary embryos (Klimaszewska 1989; Cornu and Geoffrion 1990;
von Aderkas et al. 1990).
The optimal developmental stage for initiation of somatic embryogenesis in
conifers has often been described in terms of cone collection date and/or weeks of
post-fertilization. However, in western larch cones collected 27 June and 5 July,
developmental stages ranging from very immature embryos to embryos bearing
small distinct cotyledons were found in the same cone. Explants from the 13 July
collection were all cotyledonary, but differences in the size and maturity of
embryos within cones were noted. To increase the frequency of somatic embryo-
genesis, selection of individual embryos on the basis of their observed develop-
mental stage would be a more appropriate criterion than collection date.
The site of initiation on the embryo explants was variable. White or translu-
cent cell masses originated from the suspensor/radical end, the hypocotyl from
between the cotyledons. In one case, initiation from all three sites was observed
on a single explant (Fig. 2). There was no apparent correlation between initiation
site and developmental stage ofthe explant.
2.3 Initiation from Mature Embryos
Using 11 different seed sources supplied by the Ministry of Forests of British

Columbia, isolated mature embryos were cultured on 112 LM supplemented with
9 f.LM 2,4-D and 2.2 f.LM BA, 3.4 f.LM glutamine, 1 gil casein hydrolysate, 0.4%
Gelrite. After 6 months of subculture, only 1% of embryos initiated embryogenic
tissue which could be developed to maturity. This percentage was considerably
higher after only a few months in culture, but most of the promising lines turned
brown and died. On unsuitable medium, such as MSG, there was no embryo-
genic initiation at all (n =400). Somatic embryogenesis also occurred when KIN
was substituted for BA. Seeds which were allowed to imbibe water for 4 h, and
were then dissected, never responded. Seed dissected immediately after surface
sterilization did respond; however, stratification for 1 week at 4°C was also
completely inhibitory.
2.4 Maturation of Embryos
The effect of ABA on maturation was investigated using seven lines that pro-
duced immature somatic embryos. Both the effect of the concentration and the
duration of exposure to ABA were tested. Cultures of each line were grown on
basal medium without plant growth regulators and on media supplemented with
various concentrations of ABA (0.025, 0.10, 0.25,1,10,20,40,80, and 100/lM)
for 1,2, 3, or 4 weeks. After the treatment period all cultures were transferred to
basal medium without plant growth regulators for a further 4 weeks. Cultures
were maintained at 25°C in the dark for the duration of the experiment.
Differentiated structures were selected and cultured individually on basal
medium without plant growth regulators at 21°C (using a 16-h photoperiod, 40
/lmol m-2s- 1; Sylvania Cool White).
The results in Table 2 show that of the seven lines, four produced mature
somatic embryos bearing cotyledons and root apices (Fig. 4, 5). Three lines did
not respond to any of the experimental conditions. No mature embryos were
observed on media containing 10,60, or 80 /lM ABA; for clarity, these variables
have been omitted from the table.
Line 06 produced the greatest number of mature embryos; best results were
obtained with 0.025/lM ABA for 1 or 2 weeks, although maturation in this line
was observed less frequently in several other treatments. Dramatically different
conditions (4-week exposure to 40 /lM ABA) were optimal for line 16.
Line 2177 only produced mature embryos on medium without ABA. This
line and two others (2159 and 2165) sporadically produced mature embryos
when subcultured on medium with 0.9 /lM 2,4-0 or medium without plant
growth regulators; however, 2159 and 2165 were among those that did not
respond to any of the experimental treatments. Low production of mature
embryos in the absence of ABA also occurred in lines 06 and 2168. Maturation
of somatic embryos (usually at low frequency) upon removal of growth regu-
lators has been noted previously in Larix (Nagmani and Bonga 1985;
Klimaszewska 1989; von Aderkas et al. 1990; Lelu et al. 1994).
Table 2. Number of mature somatic embryos produced in response to various ABA treatments. Seven
cell lines were used for each ABA concentration and each duration. (Thompson and von Aderkas
1992).
Duration of ABA ABA cone." (j..tM)

treatment (weeks) 0 0.025 0.10 0.25 20 40 100
4+ 19+ 14+ 0 3+ 0 0 0
1* 1*
2 1* 30+ 1+ 0 1+ 3+ 0 0
19**
3 6+ 0 0 0 0 0 0 1#
4 0 3+ 1+ 1* 0 0 17# 0
3* 1*
Symbols represent the lines that produced mature embryos: +, D6; #, 16; *, 2168; **, 2177.
Research on the role of ABA in embryo development and its prior use in
angiosperm somatic embryogenesis led to the use of ABA in conifers to increase
the frequency of maturation and improve the morphology of somatic embryos.
A wide range of ABA concentrations, exposure times on ABA, and culture
conditions have been reported. For several species the optimal concentrations of
ABA are in the micromolar range (7.6-60 J..LM) (Tautorus et al. 1991), although
lower concentrations have proved effective for Pseudotsuga menziesii (0.5 J..LM)
(Durzan and Gupta 1987) and for hybrid larch (0.38 J..LM) (Klimaszewska 1989).
In the present study, the highest frequency of maturation occurred on medium
containing 0.025 J..LM ABA, an order of magnitude less than has been reported
previously for maturation of conifer somatic embryos.
It is apparent that optimal ABA treatments are influenced by many factors
including species, genotype within species, the stage of the somatic embryos at the
time of ABA treatment (von Arnold and Hakman 1988), the presence of other
growth regulators such as IBA (Becwar et al. 1987; Roberts et al. 1990) or kinetin
(Klimaszewska 1989), and possibly endogenous levels of ABA in the somatic
embryos.
A number of other variations have been tried. These included a charcoal
treatment (1 % charcoal in 112 LM with no plant growth regulators for I week)
followed by an ABA maturation step. The response was line-specific, with only
five often lines responding. However, for the five lines which did respond, there
was no significant difference attributable to the 0.1 J..LM ABA compared with the
ABA-free control (x =5.1 + 3.7 embryos/g vs 5.7 + 5.2 embryos/g, respectively).
The best result was in line 2159 (19 embryos/g fresh weight), a line which in earlier
experiments had not responded to any level of ABA. This result was achieved on
medium which had not been supplemented with an osmoticum, such as a hexose
sugar or polyol. As the results were so variable between ABA treatments, it was
decided to raise the osmoticum level in an attempt to get an improved response.
Using 20 J..Lm ABA in combination with either polyethylene glycol 1450 or
melibiose (both are compounds which do not enter the cell), the plant material"
was exposed for an 8-week period. Such treatments, when successful, did not lead
to higher yields of somatic embryos than had previously been observed.
However, there did appear to be a qualitative difference. In the cultures with the
most embryos, ABA- and PEG-treated embryos germinated better than ABA
and melibiose-treated embryos, even though morphologically both appeared to
be of high quality (Benkrima and von Aderkas 1993). Some lines responded quite
well to ABA and 6% sucrose, germinating well when put on a medium free of
ABA and containing less sucrose (2%). Later experiments have shown that lines
may require, as they age, changes in the amount of ABA required to trigger
maturation.
2.5 Plantlet Regeneration
Germination was only successful from line D6, which produced plantlets from 61
of 85 (72%) of the cotyledonary embryos identified after 6 weeks of maturation
on media containing several concentrations of ABA. Line D6 cultures left on
Fig. 6. A germinating somatic embryo. The root is expanding straight down from the base of the
hypocotyl. A remnant of suspensor trails to the bottom left of the photograph. The cotyledons are
reflexed and the shoot meristem is apparent at the center of the ring of cotyledons; x80
Fig. 7. A young larch plant germinated in vitro. Juvenile leaves are numerous. A portion of the root
can be seen in the lower right of the photograph; x70
basal medium, produced additional plantlets after 2 or 3 months: 11 from the l-

or 2-week treatments on 0.025 /lM ABA, and 11 more from the 4-week 40 /lM
ABA treatment, which had not previously produced mature embryos in this line.
In total, 83 plantlets were regenerated via ABA treatments from line D6 embryo-
genic cultures. The conversion rate from immature to mature embryos was low,
ranging from 1.1 to 32 embryos/g.
At the time of transfer to sterile peat, the plantlets from the I-week 0.25/lM
ABA treatment group had the best and most uniform development. They
consisted of shoots with well-developed needles and roots (1.5 - 2 cm) with root
hairs (Figs. 6, 7). Root development is frequently the limiting factor in plant
regeneration via somatic embryogenesis in conifers. All viable plants obtained in
this study were derived from line D6. Later improvements using short ABA
exposure and alteration of osmoticum levels led to germination in lines 2159 and
2169. Plants ofD6 continue to grow in our greenhouse, and will be transferred to
the field as 2-year-old stock in the autumn of 1993.
2.6 Cell Suspensions
2.6.1 Embryogenesis Via Cell Suspension Culture
Establishment was done by adding approximately 3 g of embryogenic tissue to

100 ml of 112 LM liquid medium, which was supplemented with 3.4 mM L-
glutamine, 3% sucrose, 0.9 /lM 2,4-D, O.22/lM KIN. This phase could take as
386 P. von Aderkas et aI.
little as 2 weeks, but usually took 4 weeks. Subcultures were carried out every 2
weeks. Cell densities were established by digesting a sample in Cr03 and counting
the singularized cells in a hemocytometer.
The results indicated that taking cultures (eight different lines) through
suspension was not a neutral step. Suspension affected competence. No line
raised at 20°C in suspension was able to form mature embryos when transferred
to maturation medium. Lines raised at 25 °C in suspension retained this capacity.
There was a strong clonal difference in growth rates of the different lines in
suspension culture, high density lines produced the most mature embryos.
Lines can be increased for a few months in cell suspension, but it was difficult
to maintain most lines at high densities beyond 6 months.
Western larch somatic embryos may be induced from both immature and mature
embryos. The optimal stage appears to be the early cotyledonary stage. General
embryogenic response was of the order of 3 -5%. Induction was achieved using
media and combinations of growth regulating substances commonly used by
others working on Larix. Isolation ofthe embryos was the most efficient method
for induction, which occurred from various tissues, such as the base of the
embryo, the hypocotyl, and cotyledons. Lines were maintained on semisolid
medium in which the growth regulator level had been lowered to one-tenth of the
original level. However, the cytokinin source was changed to kinetin. Cell
suspensions could be used to effectively multiply lines at a rate higher than they
would grow on semisolid medium, but this was not a neutral step. Later,
maturation was influenced by the temperature at which the cells were raised in
suspension. Maturation was achieved on a variety of media. The influence of
ABA was line-dependent. Some lines responded to very low levels of ABA,
whereas required relatively high amounts. There were even a few lines in which
ABA showed no significant influence on embryo development. ABA treatments
were optimal when they were short, ofthe order of2 weeks. ABA in combination
with PEG, or hexose sugars did not increase the embryos yield, but there was an
improvement in the quality. Embryos were germinated and the plantlets con-
tinue to grow.
4 Protocol for the Induction of Somatic Embryos
Sterilized seed was dissected, the embryo at either the late embryo or mature stage removed and placed
on a semisolid medium. Litvay's medium (Litvay et al. 1981) was used at 112 strength and modified to
include 1 gil casein hydrolysate, 3.4 mM L-glutamine, 2% sucrose, and 0.4% Ge1rite, 9!lM 2,4-D, and
2.2!lM BA. Maturation was achieved by transferring material onto 112 LM containing 3% sucrose, 40
!lM ABA, and either 8% PEG or 6% sucrose or nonosmoticum (depending on line). After a 2-week
exposure period, embryos were plated onto plant growth regulator-free medium and allowed to
continue maturation for another 4 weeks. Plants were then allowed to develop on 112 LM 2% sucrose
(no growth regulators) and were transferred peat and grown in the greenhouse.
Acknowledgments. The authors would like to acknowledge the assistance ofP. Anderson, V. Barron,
G. Catalano, M. Dawkins, T. Gore, N. Luck, J. Paltiel, D. Parker, G. Parmer, and K. Thorlacius.
Seeds were kindly supplied by Barry Jaquish of the BC Ministry of Forests, as well as by Pope and
Talbot Limited of Midway, BC. A grant from the Science Council of British Columbia to the senior
author is gratefully acknowledged.
References
Balatinecz JJ (1983) Properties and utilization oflarch grown in Canada; an overview. In: Graham
CM, Farintosh HL, Graham BJ (eds) Proc Larch Symp Potential for the future, University of
Toronto, Toronto, Ontario, pp 65-81
Becwar MR, Noland TL, Wann SR (1987) Somatic embryo development and plant regeneration from
embryogenic Norway spruce callus. Tappi 70: 155-160
Benkrima L, von Aderkas P (1993) In vitro embryogenesis in larch. In: Schmidt W, McDonald K (eds)
Proc Int Symp on Ecology and management of Larix Forests, Oct 5-9, 1992, Whitefish, MT, US
Dept Agric Forest Service Rep No Washington
Cornu D, Geoffrion C (1990) Aspects de l'embryogenese chez Ie meU:ze . Bull Soc Bot Fr Actual Bot
137: 25-34
Durzan DJ, Gupta PK (1987) Somatic embryogenesis and polyembryogenesis in Douglas-fir cell
suspension cultures. Plant Sci 52: 229-235
Fowells HA (1965) Silvics of forest trees of the United States. US Department of Agriculture
Handbook No 271, Washington
Harrison SG, Dallimore EA (1996) A handbook of Coniferae and Ginkgoceae. Edward Arnold,
London
Jaquish BC (1987) A breeding plan for western larch in British Columbia. BC Ministry of Forests
Report. Kalamalka Res Stn, Kalamalka, BC, Canada
Klimaszewska K (1989) PlantIet development from immature zygotic embryos of hybrid larch through
somatic embryogenesis. Plant Sci 63: 95-103
Lelu M-A, Klimaszewska K, Charest PG (1994) Somatic embryogenesis from immature and mature
zygotic embryos and from cotyledons and needles of somatic plantIets of Larix. Can J For Res (in
press)
Litvay JD, Johnson MA, Verma D, Einspahr D, Weyrauch K (1981) Conifer suspension culture
medium development using analytical data from developing seeds. Tech pap Ser Inst Pap Chern
115: 1-17
Nagmani R, Bonga JM (1985) Embryogenesis in subcultured callus of Larix decidua. Can J For Res
IS: 1088-1091
Owens IN, Molder M (1979) Sexual reproduction of Larix occidentalis. Can J Bot 57: 2673-2690
Roberts D, Flinn BS, Webb DT, Webster FB, Sutton BCS (1990) Abscisic acid and indole-3-butyric
acid regulation of maturation and accumulation of storage proteins in somatic embryos of interior
spruce. Physiol Plant 78: 355-360
Tautorus T, Fowke LC, Dunstan DI (1991) Somatic embryogenesis in conifers. Can J Bot 69:
1873-1899
Thompson RG, von Aderkas P (1992) Somatic embryogenesis and plant regeneration from immature
embryos of western larch. Plant Cell Rep 11: 379-385
von Aderkas P, Klimaszewska K, Bonga JM (1990) Diploid and haploid embryogenesis in Larix
leptolepis, L. decidua and their reciprocal hybrids. Can J For Res 20: 9-14
von Arnold S, Hakman I (1988) Regulation of somatic embryo development in Picea abies by abscisic
acid (ABA). J Plant Physiol132: 164-169
111.11 Somatic Embryogenesis in Magnoliaceae
(Liriodendron and Magnolia)
S.A. MERKLEl
1 Introduction
The family Magnoliaceae comprises 8 to 12 genera, including some 200 species of

trees and shrubs distributed in southeast Asia, the eastern United States through
Central America, and from the West Indies to eastern Brazil (Harlow et al. 1991).
It is divided into two subfamilies or tribes, Magnolieae and Liriodendreae
(Gardiner 1989). Within the Magnolieae, the well-known genera are Magnolia,
Manglietia, Michelia, and Talauma, while the only known genus in the
Liriodendreae is Liriodendron.
Fossil remains of Liriodendron indicate that the genus once contained
multiple forms widely distributed over North America, Europe, and Asia
(Harlow et al. 1991). Currently, however, it includes only two species, L.
tulipifera L. (yellow poplar, tulip poplar, tulip tree) of eastern North America,
and L. chinense (Chinese tulip tree), native to central mainland China. Yellow
poplar is one of the most distinctive and valuable hardwoods in the eastern
United States. The tree is characterized by rapid height growth on good sites,
reaching up to 40 m in 50 years, and by its self-pruning, straight trunk, and
conical crown. Large volumes of yellow poplar wood are used for furniture,
plywood, corestock, millwork, siding, and other light construction lumber. It is
also used for pulping and for products manufactured from chips or flakes
(Russell 1977). Yellow poplar is known as an excellent honey species and as a
highly desirable ornamental, for which a number of horticultural cultivars have
been described (Santamour and McArdle 1984).
Chinese tulip tree is a medium to large tree found across a wide geographical
area covering most of the Chinese provinces of Anhwei, Kiangsi, Fukien, Hupeh,
Szechuan, Kweichow, Kwangsi and Yunnan. It is also found in Thapa, Vietnam
(Gardiner 1989). The Chinese species is morphologically very similar to the
North American one. Allozyme data, sequence divergence in the plastid genomes
of the two species, and paleobotanical evidence have been used to estimate the
time since divergence from a common ancestor at 10-16 million years (Parks and
Wendel 1990). Hybrids between the two species have been synthesized which are
I Daniel B. Warnell School afForest Resources, University of Georgia, Athens, GA 30602, 2152, USA

Somatic Embryogenesis in MagnoJiaceae (Liriodendron and Magnolia) 389
male fertile, although most are female sterile (Santamour 1972; He and
Santamour 1983; Parks et al. 1983). Highly heterotic growth rates have been
reported for the hybrid trees.
Magnolias, of which there are approximately 80 species today, are distri-
buted mainly in two distinct temperate and tropical regions of the world, eastern
America and eastern Asia. According to Gardiner (1989), the majority are native
to eastern Asia, from Manchuria, Korea and Japan, south through China and
the eastern Himalayas to Java and Malacca in Malaysia. The American magno-
lias are found from southern Canada south through the eastern United States,
the West Indies, Mexico, and Central America to Venezuela. Magnolias are
divided into two subgenera. Subgenus Yulania includes magnolias that flower
before or concurrently with the appearance ofleaves; all plants in this subgenus
are deciduous. Members of the subgenus Magnolia, which contains both
deciduous and evergreen species, flower subsequent to leaf appearance.
American and Asian species are found in both subgenera.
Although the wood of a few magnolias is valued for products such as
furniture and utensils, the trees and shrubs included in this group are particularly
well known on account of their flowers, which vary in size (up to 36 cm in
diameter in some species), in fragrance, and in color, from a pure white to royal
purple. The floral and fruiting displays put on by members of this genus, in
combination with the wide variety of habits and foliage, have made them one of
the most widely employed groups of ornamental woody perennials in the world.
A recently published checklist of the known cultivars of the genus, including
hybrids, contains over 700 entries (Langford 1990). M. grandiflora (southern
magnolia), native to the southeastern United States, is probably more widely
cultivated than any other evergreen ornamental tree (Treseder 1978). However,
there are a number of deciduous magnolias with highly desirable flowering and
foliage characteristics in the same region. M. virginiana (sweetbay magnolia) is
found in the coastal plain of the eastern United States, where it occupies low, wet
woodlands. It varies in habit from a large, multistemmed shrub to a medium-
sized tree (Brown and Kirkman 1990). The species bears fragrant, white flowers
and crimson fruit. M.fraseri (fraser magnolia) is a small to medium tree native
to the rich woods of the southern Appalachians (Brown and Kirkman 1990). It
has leaves up to 30 cm long characterized by auriculate lobes at the base and
fragrant, milky white to pale yellow flowers. M. cordata (yellow cucumber tree)
ranges in habit from a large, spreading shrub to a medium-sized tree. It is a rare
species, restricted mainly to several counties in Georgia and scattered locations in
Alabama, Florida, and the Carolinas (Harrar and Harrar 1962). Trees produce
canary yellow or, rarely, orange tulip-shaped blossoms.
1.2 Significance of Somatic Embryogenesis in Liriodendron and Magnolia
Members of Magnolia can be propagated by seeds, rooted stem cuttings, layer-

ing, grafting, and budding. As magnolias have been the object of intense
horticultural interest for centuries, methods for breeding and vegetative propa-
gation for this group are highly developed and have been well documented.
390 S.A. Merkle
Books by Treseder (1978) and Gardiner (1989) include instructional chapters on

propagation of these species. As is the case with the magnolias, a number of
vegetative propagation methods have been reported for Liriodendron, including
rooted stem cuttings, root cuttings, and grafting. However, since members of this
genus are known more as forest tree species than as horticultural ones, develop-
ment of breeding and vegetative propagation methodology has been less exten-
sive than in Magnolia. Few breeding orchards of yellow poplar have been
established in the eastern United States, and the only report of the performance
of improved seedlings from a breeding program was a study of heritability
estimates for height growth by full-sib, open-pollinated, and selfed progeny
from two Tennessee orchards (Farmer et al. 1983). The ability to root stem
cuttings of yellow poplar was very poor until it was discovered that over 70%
rooting could be obtained by using cuttings taken from stump sprouts or
epicormic shoots (McAlpine 1964; Kormanick and Porterfield 1966; McAlpine
and Kormanick 1971).
Given the broad range of techniques already available for propagating
members of the Magnoliaceae, development of embryogenic regeneration sys-
tems may not be considered a high priority. However, the reproductive charac-
teristics ofindividual species in this group and the potential for mass propagation
of horticulturally desirable hybrids make embryogenic systems a useful goal. In
the case of yellow poplar, the low filled seed percentage of open-pollinated seeds
makes them an unreliable method for propagation, while artificial pollination is
labor-intensive and therefore expensive. Propagation via rooted cuttings requires
either sacrificing the tree to obtain stump sprouts or partially girdling the tree
to produce epicormic shoots for rooting. Although the use of rooted cuttings
is much more highly developed in magnolias, some species are apparently
recalcitrant to rooting, and the number of propagules produced by this
method is ultimately limited by the availability of suitable stem material from the
mother tree.
A major drawback of the embryogenic systems described here is the inability
to initiate embryogenic cultures from mature, genetically proven material. Thus,
it would be misleading to conclude that application of these embryogenic
regeneration systems will overcome the problems listed above for propagation of
these two genera via seed or vegetative means. However, used in conjunction with
a well-planned breeding program, the existing embryogenic systems can make a
major contribution to propagation and improvement in the Magnoliaceae.
1.3 Brief Review of Work Done by Others
Considering the intense interest in breeding and propagation of the

Magnoliaceae, and in particular, members of Magnolia, there has been
surprisingly little information published on propagation of this group via any
tissue or cell culture technique. Le Page-Degivry (1970) reported on in vitro
germination of M. grandiflora and M. x souiangiana embryos. Biedermann
(1987) attempted micropropagation of a group of yellow-flowering magnolia
hybrids from axillary buds of mature trees, but apparently obtained only callus
Somatic Embryogenesis in Magno1iaceae (Liriodendron and Magnolia) 391
proliferation. A preliminary study of micropropagation of M. grandiflora from

shoot tip cuttings also failed to produce plantlets (Tobe 1990). In L. tulipifera,
Furmanowa and Rzedowski (1977) initiated callus cultures from 1-3-year-old
stems and shoot tips. Subsequently, Rzedowski (1981) reported that this callus
was capable of forming roots and shoots, but no plantlets capable of surviving
transfer to soil were produced. Stefaniak and Wozny (1983) cultured fragments
of internodes, leaf petioles, and leaf blades of yellow poplar, which produced
nonmorphogenic callus.
2.1 Yellow Poplar Somatic Embryogenesis and Plantlet Regeneration
Somatic embryogenesis in yellow-poplar tissue cultures was first reported by

Merkle and Sommer (1986) and the status of our research with this system was
updated in an earlier volume of this series (Merkle and Sommer 1991). In the
preliminary study (Merkle and Sommer 1986), immature and mature aggregates
of samaras were collected from a single tree and dissected into individual samaras
using a grafting knife. Samaras were dewinged and surface-sterilized using
the following sequence: 70% ethanol (20 s), 10% Roccal (National Laboratories;
2 min), repeat ethanol and Roccal steps, 100% Clorox (5.25% sodium
hypochlorite; 5 min), water rinse (3 min), 0.01 N HCI rinse (3 min), and three
additional water rinses (3 min each). Following surface sterilization, dewinged
samaras were dissected aseptically and embryos and endosperm were placed on
a semisolid induction medium consisting of Blaydes' (Witham et al. 1971) major
salts, Brown's minor salts (Sommer and Brown 1980), Murashige and Skoog's
(1962) iron, Gresshoff and Doy's (1972) vitamins, 40 gil sucrose, 2 mglI2,4-D,
0.25 mg/l BAP, 8 gil Phytagar (Gibco), and either 1 gil casein hydrolysate
(enzymatic; CH) or 2.5 gil yeast extract (YE). The medium was dispensed in
60-mm plastic Petri dishes. Petri dishes were sealed with Parafi1m and incubated
in the dark at about 22°C. Explants were transferred to fresh induction medium
monthly.
In the preliminary study, out of a total of 50 immature and 41 mature
embryo explants, 3 immature embryos produced an embryogenic "callus", which
we have since reclassified as proembryogenic masses (PEMs), since the prolifer-
ating cell masses have a nodular structure. Although two of the embryogenic
cultures were originally produced on induction medium with YE and one on
induction medium with CH, all subsequent experiments employed 1 gil CH in the
induction medium and no YEo Proliferation of PEMs could be maintained by
monthly passage to fresh induction medium.
To initiate development of somatic embryos, PEMs were transferred to
basal medium (induction medium lacking growth regulators). Globular embryos
were produced about 1 month following transfer to basal medium. Less than 1%
of these embryos matured to a stage resembling mature zygotic embryos, with
392 S.A. Merkle
two distinct cotyledons. Instead, most grew into thick, fleshy structures, with
partially or wholly fused cotyledons. Although many of these embryos germi-
nated, apical development was absent. Some of these malformed embryos
produced secondary embryos along their hypocotyls. Normal-appearing em-
bryos developed into seedling-like plantlets following transfer to test tubes
containing 20 ml of plantlet development medium, which was modified Risser
and White's (1964) medium with 2% sucrose and no CH. Plantlets with three to
four leaves were transferred to potting mix and acclimatized in a humidifying
chamber, gradually lowering the relative humidity from 100% to ambient condi-
tions over 6-8 weeks. During this time plantlets were fertilized weekly with 1 ml
of a modified Hoagland's solution. Acclimatized plantlets were grown in the
greenhouse for 4--5 months and planted in raised concrete block nursery beds. A
small demonstration field planting of over 200 somatic embryo-derived trees was
installed at the University of Georgia School of Forest Resources' Whitehall
Forest in March 1988.
Since the 1986 report (and the update published in 1991), the yellow-poplar
embryogenic system has been optimized for reliable initiation of embryogenic
cultures and high frequency embryo and plantlet production. Sotak et al. (1991)
characterized the optimal developmental stage of zygotic embryo explants for
initiation of embryogenic cultures. Controlled pollinations were performed in a
University of Tennessee yellow-poplar breeding orchard on three yellow-poplar
trees in a diallel mating design with no selfs, resulting in six full-sibling seed
families. Cultures were initiated as described above, using embryos resulting
from the pollinations, collected at 2 week intervals from 4 weeks post-pollination
to 18 weeks post-pollination. At the end ofthe second month in culture, plates
were scored for the presence ofPEMs. A subsample of embryos from each of the
last five sampling dates was measured (total length, cotyledon length, and
hypocotyl thickness) and analyzed for soluble polypeptides using SDS polyacryl-
amide gel electrophoresis (SDS-PAGE). Embryogenic potential, defined as the
mean percentage of explanted zygotic embryos giving rise to PEMs, rose from
near 0% for cultures initiated 4 weeks post-pollination to a peak of 28% for
cultures initiated 8 weeks post-pollination. Embryogenic potential gradually fell
with sampling date to 0% for cultures initiated from mature embryos. The stage
of embryo development corresponding to the peak of embryogenic potential was
the globular stage, 0.2 mm or less in diameter. As the zygotic embryo elongated
to the torpedo stage, embryogenic potential dropped. SDS-PAGE analyses of
zygotic embryo extracts from the last five sampling dates (10--18 weeks post-
pollination) revealed some semiquantitative differences in polypeptide patterns
among embryos from different dates. The most conspicuous of these involved a
polypeptide of approximately 55 kDa, the relative amount of which appeared to
increase with the later sampling dates, rising from approximately 12% of total
soluble protein at 10 weeks post-pollination to 25% at 18 weeks. Thus, the 55-
kDa polypeptide appeared to accumulate with maturation and to be correlated
with a decrease in the potential of an embryo to initiate an embryogenic culture.
Sotak et al. (1991) speculated that the 55-kDa polypeptide was a storage protein.
Based on these results, we now routinely employ 7-9-weeks post-pollination
zygotic embryos as explants to initiate embryogenic cultures.
Somatic Embryogenesis in Magnoliaceae (Liriodendron and Magnolia) 393
... . 140"'"
A
b
,~.-.~
<f9"
@. eo 0 'Ill
~~
..
B & Ib
~
~
0
0
{/J
8"0" <b t9
381l m
, 8
'.
II
~
~
...
I
F
Fig. lA-F. Synchronization by size fractionation of embryogenic yellow-poplar suspension cultures.

A Proembryogenic masses are grown in liquid induction medium. B Fractionating suspension on
stainless steel screens isolates PEMs between 38 and 140 11m in diameter. C Isolated PEMs are collected
on filter paper and plated with the filter on semisolid basal medium. D Embryos develop synchro-
nously, reaching maturity in about 14 days. E Mature somatic embryos are transferred from filter to
semisolid induction medium lacking casein hydrolysate to promote germination. F Germinants are
transferred to plantlet development medium in GA 7 vessels for further growth and are ready for
transplanting to potting mix within 8 weeks. (Reproduced with permission of CAB International from
Parrott et al. 1991)
394 S.A. Merkle
The other aspect of the yellow-poplar embryogenic system that required

optimization was the large-scale production of mature somatic embryos capable
of converting to plantlets at a high frequency. Soon after the first embryogenic
cultures were initiated we found that yellow-poplar PEMs proliferated rapidly as
suspension cultures. Embryogenic suspension cultures were grown in 125-ml
Erlenmeyer flasks containing 40 ml of liquid induction medium on a gyratory
shaker at 90 rpm and maintained by transferring approximately 1 g ofPEMs to
fresh medium every 3 weeks. Protocols involving size fractionation of embryo-
genic suspension cultures were tested for their potential to produce synchronous
populations of mature somatic embryos (Merkle et al. 1990). In one protocol
tested, approximately 1 g of suspension cultured PEMs was transferred to liquid
basal medium and sieved on a 140-J.Ull stainless steel screen; the fraction passing
through was then sieved on a 38-llm screen. The fraction remaining on the 38-llm
screen was cultured for 1 week in basal medium. Developing embryos derived
from the cultured PEM fraction were sieved on a 230-llm screen; the fraction
passing through was resieved on a 140-J.Ull screen. The fraction remaining on this
screen was cultured in liquid basal medium. Although this size fractionation
procedure generated very uniform populations of globular somatic embryos
within 1 week following the second fractionation, development of the embryos
did not proceed normally from this point, with cotyledons and radicles elongat-
ing prematurely. Embryo development in the suspension cultures could be
modified by replacing the liquid basal medium with basal medium supplemented
with abscisic acid (ABA). ABA (5 x 10-7 M) appeared to promote more normal
maturation of somatic embryos, so that in some cultures up to 50% of the
embryos continued deVelopment to heart or early torpedo stages. However,
conversion rates of these suspension-cultured embryos remained below 1%.
Low conversion rates of somatic embryos prompted us to modify the size
fractionation procedure to avoid prolonged culture of developing embryos in
liquid medium. In this alternative protocol (Fig. 1),2 weeks following transfer to
fresh induction medium, 1 g of PEMs was sieved on a 140-llm stainless steel
screen and the fraction passing through was resieved on a 38-llm screen. The
fraction remaining on the 38-llm screen was rinsed with basal medium and then
backwashed from the screen onto a single layer of filter paper in a Buchner
funnel. PEMs were again rinsed with basal medium while under mild vacuum,
which served to spread PEMs into a single layer. When excess liquid medium had
been drawn off, the filter paper with PEMs was placed on semisolid basal medium
and incubated under fluorescent light (16 h/day) at 30°C. As PEMs developed
into somatic embryos on the filter paper, mature torpedo-stage embryos with
well-developed cotyledons were selected and transferred to Petri plates contain-
ing germination medium (basal medium lacking CH). This modification of the
basal medium was made following our discovery that CH inhibited somatic
embryo germination, probably by raising the osmotic potential of the medium,
and that eliminating it promoted vigorous germination and cotyledon greening
within 1 week (Merkle et al. 1990). Using this method, a 60--70% synchronous
population of embryos could be produced within 2 weeks following plating of
PEMs (Fig. 2). Following transfer to plantlet development medium in GA7
vessels (Magenta Corp.), an average of32% of these somatic embryos converted
Fig. 2. Roughly synchronous population of yellow-poplar somatic embryos, approximately 12 days

=
following fractionation and plating, bar 200 J.lm
to plantlets (Merkle et al. 1990). Recently, this protocol was applied to nine
embryogenic lines to produce over 5500 somatic embryo-derived plantlets for
field testing (Merkle et al. 1991).
2.2 Magnolia Somatic Embryogenesis and Plant Regeneration
Using our experience with yellow poplar as a basis, experiments were initiated
to establish embryogenic cultures of three magnolia species, M. virginiana
(sweetbay magnolia), M. fraseri (fraser magnolia), and M. cordata (yellow
cucumber tree) (Merkle and Wiecko 1990). Developing fruits were collected from
ten ornamental sweetbay trees growing on the University of Georgia campus, at
weekly intervals from 2 to 7 weeks post-anthesis. Because of their inaccessibility,
immature fruits from three fraser magnolia trees growing in the North Georgia
mountains were collected only once, approximately 5 weeks post-anthesis. Simi-
larly, immature fruits from a single yellow cucumber tree were collected only
twice approximately 3 and 9 weeks post-anthesis. Aggregates of follicles were
dissected using a grafting knife, and seeds were surface-sterilized using the same
sequence as was used with yellow-poplar samaras. Following surface-steriliza-
396 S.A. Merkle
Fig. 3A-D. Somatic embryogenesis in sweetbay magnolia. A Somatic embryos arising by direct
embryogenesis from end of an immature seed. BMature somatic embryos arising from proembryo-
genic masses. C Germinating somatic embryo. DSomatic embryo-derived plantlets. Bar in all photos
= 500 J.Im. (Merkle and Wiecko 1990)
tion, seeds of all three species were bisected longitudinally with a scalpel and the
halves were placed cut surface downward in 60-mm plastic Petri dishes contain-
ing semisolid yellow-poplar induction medium. Cultures were maintained in
darkness at 22°C and transferred to fresh induction medium monthly.
Each magnolia species responded somewhat differently to culture condi-
tions. Somatic embryogenesis in sweetbay magnolia, which first was apparent 5
weeks following explanting, took two forms that appeared to depend on explant-
ing date. Somatic embryos from seeds explanted 2 weeks post-anthesis appeared
to be similar to PEMs described above for yellow poplar, while embryos from
cultures initiated at later dates were produced directly, with no intermediate
callus or PEM formation (Fig. 3A). Three weeks post-anthesis seed explants of
sweetbay magnolia appeared to have the highest potential for initiation of
embryogenic cultures, with 12 ofthe 45 seeds producing somatic embryos. Fraser
magnolia explants responded rapidly to culture, with PEMs appearing after only
3 weeks on induction medium. Within 5 weeks, 11 of the 90 explanted seeds had
produced PEMs. Only one explanted seed of yellow cucumber tree, collected 3
weeks post-anthesis, produced an embryogenic culture, 6 weeks following
explanting. Embryogenic material of this species first appeared as a rapidly
proliferating mixture ofPEMs and globular stage somatic embryos on induction
medium. With the exception of the single yellow cucumber tree culture, embryo-
genic magnolia cultures differed from yellow-poplar cultures in that they could
not be maintained as suspensions ofPEMs. Both sweetbay magnolia and fraser
magnolia cultures tended to form clumps and darken within a few weeks after
being inoculated into liquid induction medium. Thus, prospects for large-scale
propagation of these magnolias via synchronization of embryo development are
not promising currently.
As in the yellow-poplar cultures, PEMs (or early-stage embryos in the case of
sweetbay magnolia) produced clumps of developing somatic embryos usually
within 1 month following transfer to yellow-poplar basal medium (Fig. 3B).
Embryogenic sweetbay cultures were easier to maintain on basal medium than
on induction medium, since on basal medium they tended to continue production
of new embryos via repetitive embryogenesis, whereas the same cultures declined
with extended culture on induction medium. Germination of mature somatic
embryos of all three magnolia species was stimulated by transferring them to
basal medium lacking CH (Fig. 3C). Germinants transferred to yellow-poplar
plantlet development medium in test tubes or GA7 vessels continued root
elongation, but apical development generally lagged behind that of yellow-
poplar plantlets, with only one new leaf produced every 2-3 weeks (Fig 3D).
Percent conversion of germinants to plantlets was also lower for the magnolia
species than for yellow poplar, ranging from a high of 25% for sweetbay
magnolia to less than 10% for fraser magnolia. Plantlets of all three species
transferred to potting mix and grown in the humidifying chamber were
acclimatized to greenhouse conditions with very low mortality. Although no
organized field testing of magnolia somatic embryo-derived plantlets is planned,
those trees which have been planted out resemble seedling-derived trees in form,
growth rate and bud phenology (Fig. 4).
398 SA Merkle
Fig. 4. Somatic embryo-derived plantlets of fraser magnolia (foreground) and yellow cucumber tree
(background) following one season of growth in the field, (rod is I m)
2.3 Protoplast Isolation and Culture from Embryogenic Suspensions
Populations of highly regenerative protoplasts have been isolated from embryo-

genic yellow-poplar suspensions (Merkle and Sommer 1987). In this study,
protoplasts were isolated from suspension cultures grown under 16-h photo-
periods, 6 days following subculture (Merkle and Sommer 1991). Yields of 1- 2 x
106 protoplasts/g were obtained. Suspension cultures grown in the dark failed to
yield significant numbers of protoplasts. Wilde et al. (1989), however, found that
the highest protoplast yields (1 .7 x 107 protoplasts/g) were obtained from 20-day-
old suspensions. The isolation and culture procedure has been detailed in an
earlier volume of this series (Merkle and Sommer 1991). Briefly, approximately
1 g ofPEMs was placed in 10 ml of filter-sterilized digestion medium, which was
of the same formulation as yellow-poplar basal medium, supplemented with 500
mgll CaCl2·2Hp, 600 mglI2(N-morpholino) ethane sulfonic acid (MES), 109 gil
mannitol as osmoticum, 10 gil Cellulysin (Calbiochem-Behring Corp.), 5 gil
Macerase (Calbiochem-Behring Corp.), and 1 gil bovine serum albumin. PEMs
were incubated in digestion medium for 12 or 24 h at 30 °C on a gyratory shaker
at 50 rpm. Following digestion, the suspension was dripped through two layers of
Mirac1oth, then through a 25-llm pore size stainless steel sieve. The filtrate was
centrifuged at 100 x g for 5 min and the supernatant was discarded and the pellet
resuspended in 5 ml of filter-sterilized wash medium (digestion medium lacking
enzymes and bovine albumin). Centrifugation and resuspension in fresh wash
medium were repeated twice more. After the final centrifugation, the supernatant
was removed and the protoplast pellet was resuspended in regeneration medium,
which was a modified yellow-poplar induction medium supplemented with 500
mgll CaCl2 ·2H20, 250 mgll each of xylose, fructose, and sucrose, and 0.5 M
glucose as osmoticum. Protoplasts were plated by mixing drops of protoplast
suspension with approximately equal amounts of autoc1aved regeneration medi-
um containing 2.5% SeaPlaque (FMC Corp.) low melting point agarose, which
had been cooled to 38 °C. The mixture was pipetted up and redistributed as
droplets in an empty plastic Petri plate. Once the droplets solidified, the plates
were flooded with liquid regeneration medium.
Cultured yellow-poplar protoplasts regenerated cell walls within 3 days and
the first cell divisions generally occurred within 4 days. Cell colonies of eight or
more cells were observed within 2 weeks and calli were visible without a micro-
scope within 3 weeks. Plating efficiency (percent of protoplasts forming cell
colonies) ranged as high as 30%. Suspension cultures could easily be established
from the protoplast-derived calli by placing the agarose droplets containing the
calli into liquid induction medium in flasks on a gyratory shaker. Within a few
weeks, calli expanded out of the droplets and PEMs proliferated rapidly in
suspension. PEMs from suspensions regenerated from the protoplast-derived
calli produced somatic embryos following transfer to basal medium, and plant-
lets regenerated from these embryos appeared normal.
2.4 Use of Embryogenic Suspensions for Gene Transfer
Yellow-poplar protoplasts derived from embryogenic suspension cultured cells

were employed in genetic transformation experiments. In these experiments,
electroporation and polyethylene glycol (PEG) were used to transfer plasmid
DNA (pBI221; Jefferson et al. 1987) containing the reporter gene encoding
~-glucuronidase (GUS) into the protoplasts (Wilde et al. 1989). Protoplasts were
electroporated with a Promega Model 450 in buffer containing 100 mM NaCl,
4 mM CaCl2, 8% PEG, and 0.5 M mannitol. A field strength of 700 V/cm and a
capacitance of 450 microfarads were found to be optimal to give the desired pulse
400 S.A. Merkle
decay constant of 12 ms and protoplast viability of 53%. Transient expression of

the GUS gene showed that DNA uptake by PEG treatment alone was as effective
as electroporation when the PEG concentration was raised to 20%. Physiological
age of the protoplasts influenced the amount of GUS activity observed after the
plasmid uptake, with the highest level of transient expression in protoplasts
isolated from suspensions 20 days following subculture. Although transient
GUS expression could be optimized in the protoplasts, transformation with a
plasmid (pBI121) carrying a neomycin phospho transferase (NPT II) gene in
addition to GUS did not result in the production of stably transformed proto-
plast-derived cell colonies following selection on medium containing 200 ).lg/ml
kanamycin.
The development of microprojectile-mediated gene transfer (Klein et al.
1988) eliminated the need to use protoplasts for transformation of embryogenic
yellow-poplar cells, since foreign DNA could be delivered directly to intact cells.
Using a DuPont Biolistics PDS-lOOO particle gun, Wilde et al. (1992) stably
transformed embryogenic cells of yellow poplar, from which transgenic plantlets
were regenerated via somatic embryogenesis. Plasmid DNA (pBI121.1) was
precipitated onto 1.1-).lffi tungsten particles which were used to bombard individ-
ual cells and small cell clusters that had been isolated by size fractionation of an
embryogenic yellow-poplar suspension, and collected on a filter paper disk. After
2 days of incubation on semisolid induction medium, filters were transferred to
induction medium containing 100 ).lg/ml kanamycin. After 5-6 weeks, kanamy-
cin-resistant microcalli, approximately 1 mm in diameter, were transferred
individually to fresh antibiotic-containing plates. Suspension cultures were initi-
ated from the resulting lines by transferring PEMs to liquid induction medium
containing 50 ).lg/ml kanamycin, which was sufficient to inhibit growth of
nontransformed yellow-poplar suspensions. The ability to grow these cultures as
suspensions was particularly desirable as it allowed stringent selection pressure to
be applied, thereby reducing the likelihood of regenerating chimeras. Southern
analysis indicated that independently transformed embryogenic sublines carried
from approximately 3 to 30 full-length copies of the GUS gene. A histochemical
assay for GUS activity employing the substrate 5-bromo-4-chloro-3-indolyl-~
D-glucuronic acid (X-Gluc) revealed a heterogeneous staining pattern in PEMs,
which seemed to indicate that they were chimeric for the GUS gene. However,
extracts from cell clusters reacting positively (blue) or negatively (white) with this
substrate both demonstrated GUS activity when a fluorometric assay was
performed using the substrate 4-methylumbelliferyl-~-D-glucuronide. Further-
more, somatic embryos regenerated from transformed sublines were found to be
uniformly GUS positive using the histochemical assay. Plantlets derived from
transformed somatic embryos and grown in the greenhouse expressed GUS and
NPT II in leaves and roots.
Members of the Magnoliaceae have proven to be very amenable to the induction

of somatic embryogenesis via a straightforward procedure of culturing immature
zygotic embryos on a medium containing the auxin 2,4-D. The optimal develop-
mental stage for explanting of zygotic embryos of yellow poplar and at least one
magnolia appears to be the globular to early heart stage. Although somatic
embryos and plantlets of both Liriodendron and Magnolia have been produced
following 2,4-D induction and subsequent transfer of induced material to basal
medium, only L. tulipifera cultures have proven to be amenable to long-term
culture as suspensions ofPEMs in liquid induction medium. The ability to grow
embryogenic yellow-poplar cultures has enabled us to employ size fractionation
of the cultures for synchronization of embryo development and generation of
optimal target tissue for gene transfer via microprojectile bombardment. In
addition, the ability to grow yellow-poplar PEMs as suspension cultures has
provided a source of highly regenerable protoplasts and allowed stringent
antibiotic selection pressure to be applied following gene transfer. While none of
the Magnolia species can be maintained for long periods in suspension culture,
cultures of these species maintained on semisolid induction medium have pro-
duced somatic embryos and plantlets at high frequencies. Plantlets of both yellow
poplar and the magnolias have survived acclimatization, greenhouse culture, and
field planting, and a large-scale field test of somatic embryo-derived yellow-
poplar plantlets is currently being established. Microprojectile bombardment of
embryogenic cultures has proven to be an effective means by which to generate
transgenic yellow-poplar trees.
4 Protocol for the Induction of Somatic Embryos
Although the optimal treatment for induction of somatic embryogenesis differs between yellow poplar
and the magnolias, and among the magnolia species themselves, a general protocol for production of
embryogenic cultures, somatic embryos, and plantlets of all of these species can be prescribed. For
establishment of embryogenic cultures (PEMs or repetitive somatic embryos), immature zygotic
embryos (approximately 8 weeks post-pollination for yellow poplar and 3-5 weeks post-pollination
for magnolias) are explanted onto a modified Blayde's (Witham et al. 1971) induction medium,
supplemented with 2 mgll 2,4-D, 0.25 mg/I, BAP, I gil CH, and 40 mgll sucrose and solidified with
8 gil Phytagar. Cultures are incubated in 60-mm plastic Petri dishes in the dark at 22°C. For
production of mature somatic embryos, PEMs or clumps of immature repetitive embryos are trans-
ferred to basal Blayde's medium (induction medium lacking plant growth regulators), and grown in the
light (16 h/day) at 22°C. In the case of suspension cultures of yellow-poplar PEMs, optimal embryo
production is obtained by size-fractionating the PEMs using a 140-~ pore size stainless steel sieve,
followed by a 38-~ pore size sieve. The fraction passing the 140-~ sieve but retained on the 38-/!m
sieve is backwashed from the screen onto a disk of filter paper in a Buchner funnel and washed with
basal medium under mild vacuum. Then the filter with PEMs is transferred to a plate of semisolid basal
medium for production of mature embryos.
Acknowledgments. The author wishes to thank Dr. Roberta K. Merkle and Dr. H. Dayton Wilde for
advice in preparing the manuscript.
402 S.A. Merkle
References
Biedermann lEG (1987) Factors affecting establishment and development of magnolia hybrids in
vitro. Acta Hortic 212: 625-629
Brown CL, Kirkman LK (1990) Trees of Georgia and adjacent states. Timber Press, Portland, 292 pp
Chu CD (1979) Liriodendron. In: Technical consultation on fast growing plantation broadleaved trees
for Mediterranean and temperate zones. FAO Publ FO: FGB-79-8/14 pp 381-383
Farmer RE Jr, Barnett PE, Thor E, Rennie JC (1993) Heritability estimates for height growth of
Tennessee yellow-poplar. Silvae Genet 32(1-2): 15-18
Furmanowa M, Rzedowski M (1979) Liriodendron tulipifera L. in tissue culture. Acta Pol Pharm 36
(2): 261-262
Gardiner JM (1989) Magnolias: their care and cultivation. Cassell, London, 143 pp
GresshoffPM, Doy CH (1972) Development and differentiation of haploid Lycopersicon esculentum
(tomato). Planta 107: 161-170
Harlow WM, Harrar ES, Hardin JW, White FM (1991) Textbook of dendrology, 7th edn. McGraw-
Hill, New York, 510 pp
Harrar ES, Harrar JG (1962) Guide to southern trees. Dover publications, New York, 709 pp
He SA, Santamour FS Jr (1983) Isoenzyme verification of American-Chinese hybrids of Liquidambar
and Liriodendron. Ann Mo Bot Gard 70: 748-749
Jefferson RA, Kavanaugh TA, Bevan MW (1987) GUS fusions: p-glucuronidase as a sensitive and
versatile gene fusion marker in higher plants. EMBO J 6: 3901-3907
Klein TM, Gradziel T, Fromm ME, Sanford JC (1988) Factors influencing gene delivery into Zea mays
cells by high-velocity microprojectiles. Biorrechnology 6: 440--444
Kormanick PP, Porterfield EJ (1996) Rooting yellow-poplar cuttings. For Farmer 26: 24, 41-42
Langford L (1990) Check list revision. J Am Magnolia Soc 25(1): 22-28
Le Page-Degivry MT (1970) Seed dormancy associated with embryo immaturity: contribution for the
study of Magnolia soulangeana Soul. Bod. and Magnolia grandijiora L. by means of in vitro
culture. Planta 90: 267-271
McAlpine RG (1964) A method for producing clones of yellow-poplar. J For 62: 115-116
McAlpine RG, Kormanick PP (1971) Rooted cuttings from select yellow-poplar trees. In: Proc 11th
ConfSouthern forest tree improvement, Atlanta, Georgia, pp 241-243
Merkle SA, Sommer HE (1986) Somatic embryogenesis in tissue cultures of Liriodendron tulipifera.
Can J For Res 16: 420-422
Merkle SA, Sommer HE (1987) Regeneration of Liriodendron tulipifera (family Magnoliaceae) from
protoplast culture. Am J Bot 74(8): 1317-1321
Merkle SA, Sommer HE (1991) Yellow-poplar (Liriodendron spp.).ln: Bajaj YPS (ed) Biotechnology
in agriculture and forestry, vol 16. Trees III. Springer, Berlin Heidelberg New York, pp 94-110
Merkle SA, Wiecko AT (1990) Somatic embryogenesis in three magnolia species. J Am Soc Hortic Sci
115(5): 858-860
Merkle SA, Wiecko AT, Sotak RJ, Sommer HE (1990) Maturation and conversion of Liriodendron
tulipifera somatic embryos. In Vitro Cell Dev Bioi 26: 1086-1093
derived plantlets of yellow-poplar for field testing. In: Proc 21 st Southern forest tree improvement
Conf, June 17-20,1991, Knoxville, TN pp 56-68
Parks CR, Miller NG, Wendel JF, McDougall KM (1983) Genetic divergence within the genus
Liriodendron (Magnoliaceae). Ann Mo Bot Gard 70: 658-666
Parks CR, Wendel JF (1990) Molecular divergence between Asian and North American species of
Liriodendron (Magnoliaceae) with implications for interpretation of fossil floras. Am J Bot 77 (10):
1243-1256
Parrott W A, Merkle SA, Williams EG (1991) Somatic embryogenesis: potential for use in propagation
and gene transfer systems. In: Murray DR (ed) Advanced methods in plant breeding and
biotechnology. CAB International, Wallingford, Oxon, UK, pp 158-200
Risser PG, White PR (1964) Nutritional requirements of spruce tumor cells in vitro. Physiol Plant 15:
620-635
Russell TE (1977) Planting yellow-poplar - where we stand today. USDA Forest Service Gen Tech
Rep SO-17. South For Exp Stn, New Orleans LA, 8 pp
Rzedowski M (1981) Alkaloidy w kulturze tkankowej Liriodendron tulipifera L. PhD Thesis, Medical
Academy, Warsaw
Santamour FS Jr (1972) Interspecific hybrids in Liriodendron and their chemical verification. For Sci
18(3): 233-236
Santamour FS Jr, McArdle AI (1984) Cultivar checklist for Liquidambar and Liriodendron. J Arboric
10(11): 309-312
Sommer HE, Brown CL (1980) Embryogenesis in tissue cultures of sweetgum. For Sci 26(2): 257-260
Sotak RJ, Sommer HE, Merkle SA (1991) Relation of the developmental stage of zygotic embryos of
yellow-poplar to their somatic embryogenic potential. Plant Cell Rep 10: 175-178
Stefaniak B, Wozny A (1983) Investigations on the ultrastructure of the callus tissue of Liriodendron
tulipifera L. developed in vitro. Acta Soc Bot Pol 82(1): 3-8
Tobe JD (1990) In vitro growth of Magnolia grandiflora L. cv. Bracken's Brown Beauty. Magnolia
26(1): 4--8
Treseder NG (1978) Magnolias. Faber and Faber, London 243 pp
Wilde HD, Meagher RB, Merkle SA (1989) Introduction of foreign DNA into yellow-poplar
protoplasts by electroporation. In: Proc 20th Southern forest tree improvement Conf, Charleston,
SC, pp 389-394
plants. Plant Physiol 98: 114--120
Witham FH, Blaydes DF, Devlin RM (1971) Experiments in plant physiology. Van Nostrand-
Reinhold, New York, 245 pp
111.12 Somatic Embryogenesis in Olive
( Olea europaea L.)
E. RUGINI, A. PEZZA, M. MUGANU, and G. CARICATO i
1 Introduction
In many Mediterranean countries olive is one of the oldest and most important
crops. It belongs to the family Oleaceae and has approximately 30 genera with
600 species (Cronquist 1981) which are distributed on every continent and
comprise a large number of garden plants and trees cultivated for economical
purposes. According to Taylor (1945), the basic chromosome number of the
genus Olea is x =23 (2n =46). Cultivated olive originated in Asia Minor and then
subsequently spread to the Mediterranean basin. Today, olive is present even in
western and South Africa, the USA, Australia, and China. Most cultivated olives
belong to Olea europaea L., with more than 2600 different cultivars, although
many of these might be ecotypes. Olea europaea L. does not seem to be a species,
but rather a group of forms which originated by mutation and hybridization
(Roselli and Scaramuzzi 1974; Chevalier 1948).
There are two types of Euro-Mediterranean olives: wild (oleaster), and
cultivated (sativa). The oleaster is not found in many areas and it is often
mistaken for ole vaster which is a cultivated ecotype of Olea europaea. Olevaster
grows in the wild in a continuous juvenile stage and never produces flowers, but
can flower when transferred to standard cultivation (Rugini and Lavee 1992).
About 96% of the olive production is concentrated in the Mediterranean
countries, mostly for oil extraction; only about 7-8% are table olives. The use of
olive oil has greatly increased in recent years due to its appetizing taste and
nutritional quality. It has a high oleic acid content which has been found to be of
an important dietary value and is the only fluid oil which may be used in the crude
form. Considering the growing importance of the species and the present limits of
both cultivars and traditional genetic improvement methods, it seems that
biotechnological techniques are necessary; thus, an efficient in vitro regeneration
method is fundamental.
1 Dipartimento di Produzione Vegetale sez. Ortofloroarboricoltura, Universitit della Tuscia, Facoltit
di Agraria, via S. Camillo de Lellis, 01100 Viterbo, Italy

Somatic Embryogenesis in Olive (Olea europaea L.) 405
1.2 Brief Review of Work Done In Vitro
The main objectives of the in vitro techniques are rapid propagation of difficult-
to-root cultivars by cutting, production of disease-free plants, cryopreservation
of valuable germplasm, and genetic improvement. The details on in vitro cultures
have been reported in previous reviews (Rugini and Fedeli 1990; Rugini and
Lavee 1992).
1.2.1 Micropropagation
Different attempts to establish sterile cultures with meristems or shoot tips from
field-grown or greenhouse plants were unsuccessful due to the rapid oxidation of
tissues after collection; even the use of active antioxidants gave poor results.
Vigorous nodal apical herbaceous twigs, vigorous node explants and inflo-
rescences from mixed buds are recommended for use. In our laboratory, sterile
shoots have also been recently efficiently produced from subepidermal vegetative
buds located in inflorescences which sprouted from mixed buds. One advantage
is that these explants can be easily disinfested and buds can easily develop into
shoots on a low concentration of mineral salt agar medium plus 2.5 !lM zeatin
(Rugini 1984). A CaiN ratio between I: 7-11 is important to improve the shoot
quality in proliferation medium (Rugini 1984; Fiorino and Leva 1986). Among
cytokinins only zeatin allows satisfactory olive growth; a low agar concentration
and well-aerated jars are essential conditions for rapid growth. The shoot growth
is enhanced by adding filter sterile 50 !lm GA3 to the medium without affecting
the rooting in the subsequent phase. Olive presents a strong apical dominance
that up to now could not be reduced; from one bud only one shoot can develop
and for each shoot 8-9 nodes can be obtained in 35-40 days. Some cultivars
prefer mannitol in place of sucrose as carbon source (Leva et al. 1992). We
observed that this substance decreases the basal callus formation without chang-
ing shoot development. The inconstant rooting often observed among the
explants was efficiently reduced by two techniques: rooting with basal etiolation
(Rugini et al. 1988) or the addition of 1 mM putrescine to the medium (Ruginin
and Fedeli 1990).
1.2.2 Callus Culture
Callus cultures of olives were first established by Levee and Messer (1969).
Friable callus can be initiated from any part of the in vitro germinated seedlings
in the dark using various concentrations of NAA or 2,4-D or IBA and zeatin.
After initiation, olive callus can be kept for a long time in cultures under long day
conditions (Rugini 1986; Canas and Benbadis 1988). In our laboratory, a healthy
semihard and white callus was obtained from petioles of in vitro proliferated
shoots of several cultivars using TDZ [thidiazuron (N-phenyl-N'-1 ,2,3-thidiazol-
5-ylurea)], in the dark at 28°C. A good quality of callus was also recently
obtained from the vascular tissues connecting the peduncle with a seed of
406 E. Rugini et al.
1-month-old fruitlets. Lavee and Adiri (1974) found that ABA increased callus
growth; in contrast, GA3 had an inhibitory effect on olive bark.
1.2.3 Shoot Organogenesisfrom Juvenile and Mature Explants
Organogenesis was observed under different conditions and in different tissues.

Direct shoot organogenesis was obtained from olive hypocotyl sections in
White's medium with O.OSIlM NAA and 2.SIlM BAP (Bao et al. 1980). High
shoot regeneration has also been shown by callus derived from mature cotyledon
fragments of Tanche and Picual cultivars. The regeneration was higher in calli
from cotyledon segments proximal to the embryo axes than distal ones (Canas
and Benbadis 1988).
Adventitious shoots were induced only in the dark in petioles of in vitro
grown shoots of cultivars Moraiolo, Dolce Agogia, and Halkidikis (Fig. 1). The
petioles forming shoots ranged from 10 to 40% according to leaf position on the
shoot and the medium used (Menucuccini and Rugini 1993).
1.2.4 Protoplast Isolation and Cultures
Viable protoplasts from hypocotyls, cotyledons, or leaves of micropropagated

shoots were obtained by using I.S% Driselase for 10 h after incubation with 0.6
M mannitol, SOO mgll CaCI2" 2HP, and SOO mg/l KCl. In BN basal medium
supplemented with SIlM NAA, 2.Sllm 2,4-D, and 2.SIlM zeatin riboside, some
divisions were observed (Rugini 1986). Using a similar protocol, with only the
addition of 0.03 mM ornithine in the medium, Canas et al. (1987) produced
rnicrocalli from protoplast-derived cells of mature cotyledons.
Olive has not yet been genetically improved, due mainly to its very long juvenile
period and its prevalent self-sterility which makes conventional breeding diffi-
cult. The nonconventional methods, such as transformation, somaclonal vari-
ation, and protoplast manipulation, could be useful in genetic for olive
improvement. In this context, somatic embryogenesis is an important tool for
plant regeneration from manipulated cells. In addition, given the difficulty in
obtaining homozygous plants by traditional methods, somatic embryogenesis
could be useful in obtaining dihaploid homozygous plants from anthers, pollen,
or ovules. Furthermore, considering the difficulty of rooting in several cultivars
and new genotypes, somatic embryogenesis could also represent an efficient
method of propagation. Up to now somatic embryogenesis has been obtained
from both immature and mature zygotic embryo tissue callus, and recently also
from mature tissues of cultivars.
2 3
Fig. I. Shoot regenerated from leaf petiole callus of cuItivar Dolce Agogia on medium containing
10 J.lM TDZ and 2.5 J.lM NAA
Fig.2. Somatic embryogenesis from callus derived from 75-day-old zygotic embryos on MS medium
containing 0.5 J.lM zeatin
Fig. 3. Somatic embryogenesis which appeared from taproot seedling callus on MS medium contain-
ing 0.5 J.lM NAA and 0.5 J.lM BAP. Callus was initiated from injured rootlets with 25 J.lM NAA and
2.5J.1M BAP
2.1 Somatic Embryogenesis from Immature Zygotic Embryos
Somatic embryogenesis from immature zygotic embryos of cvs. Dolce Agogia,

Leccino, Frantoio, and Moraiolo was obtained first by Rugini (1988). Date of
embryo collection, growth regulator, and light are critical factors in embryo
formation. The zygotic embryos were tested for their embryogenic capacity at 50,
75,90, and 105 days after full bloom by using two basal media: (1) half-strength
MS medium (Murashige and Skoog 1962); (2) half-strength modified OM (olive
medium; Rugini 1984), in which glutamine was replaced by 0.3 mM Ca(N03)2'
4HzO. Both basal media were solidified with 0.8% w/v Bacto-agar (Difco) and
supplemented with 2% w/v sucrose, and the pH was adjusted at 5.7 before
autoclaving. Different concentrations of NAA and 2,4-D (0, 0.5, 2.5 and 12.5
J..lM) were separately combined with four concentrations ofBAP (0,0.5, 2.S, and
12.S J..lM) and then tested.
Embryogenesis occurred only from 7S-day-old zygotic embryos, indepen-
dent of cultivars, while those collected in the other period produced only in some
treatments callus with globular structures which never developed into distinct
embryos. Somatic embryos began to appear after 6 weeks in culture from some
parts of the zygotic embryos tissues, i.e, 20% from calli and 10% from uncertain
origin. The highest percentage of somatic embryogenesis was obtained in the
absence of auxins and with BAP at 0.5 J..lM (40%) or with BAP at 2.5 J..lM (30%)
(Fig. 2). At a higher concentration BAP (12.S J..lM) inhibited embryogenesis
which was reestablished by adding NAA (Table I). The number of somatic
embryos derived from each zygotic embryos was 2 to IS, often arranged in a
crown-like position. MS medium supported higher embryogenic capacity than
modified OM medium, but in the latter medium embryos developed more rapidly
into plantlets. Even 2,4-D at a low concentration stimulated more yellow callus
growth than NAA, but at 12.S J..lM reduced callus formation. The increase in
osmolarity by sorbitol did not affect embryogenesis but inhibited rooting and
shoot development. Both light and 2,4-D completely inhibited somatic embryo-
genesis (Table I).
Table 1. Percentage of zygotic embryos harvested 75 days after full bloom which gave rise to somatic
embryos in half-strength MS medium with different growth regulators. 2,4-D, even at a very low
concentration, alone, or in combination with cytokinins completely inhibits somatic embryogenesis
NAA
J.lM 0 0.5 5 12.5
BAP 0 10 0 0 0
BAP 0.5 40 10 10 10
BAP 2.5 30 10 10 10
BAP 12.5 0 14 12 10
Zeatin' 0.5 5 8
'Zeatin was used only with embryos from fruitlets harvested 75 days after full bloom and stored 45
days at 14-15 DC. The calli maintained their embryogenic capacity for at least I year in half-strength-
MS medium with 0.5J.1M zeatin and 0.5 11M NAA.
It was observed that the capacity of the immature zygotic embryos to form
somatic embryos may be maintained for at least 2 months by collecting the
fruitlets at 75 days after full bloom and then stored at 14 to 15°C. Under these
conditions the small embryos continue to grow, but do not lose their embryo-
genetic capacity.
2.2 Somatic Embryogenesis from Mature Zygotic Embryo Tissues
Embryogenesis was obtained from: (1) callus of the taproot of seedlings, (2)
dissected radicles, and to a lesser extent, from cotyledon callus.
1. From Taproot Callus. Drupes of cv. S. Agostino were harvested at maturity

and deprived of mesocarp; the stones were washed in a solution of2% NaOH for
about 0.5 and stored in the sand at 3-4 °C for 4 to 5 months. Before use, the seeds
were freed from endocarp, surface-sterilized with alcohol for 20-30 s, then
immersed in a solution containing 6% sodium hypochlorite for 20 min before
washing them in distilled water and left imbibed in Petri dishes on a wet filter
paper for 24 h. The embryos were then aseptically extracted and germinated as
previously reported (Rugini 1986). When the seedlings developed into two to
three nodes, several superficial wounds were inflicted by a knife on the taproot,
and the whole plantlets were laid on the surface of the solid MS basal medium
plus 25 ~M NAA and 0.5 ~M BAP and 3% sucrose. The medium was contained
in 1OO-mljars and submitted to a 16-h photoperiod at 23°C. After about 4 weeks,
when callus developed from the wounds the whole plantlets were subcultured
according to the above modalities, in half-strength MS medium plus 0.5 ~M
NAA and 0.5 ~M BAP, 3% sucrose, and 0.6% agar and subcultured in the dark.
After 35 days the calli started to turn brown and some embryos simultaneously
appeared (Fig. 3). These were transferred to half-strength MS solid medium plus
2.5 ~M zeatin and 2% sucrose for germination and the calli quickly developed
into plantlets (Rugini and Tarini 1986).
2. From Dissected Radicle Callus. Similar procedures were adopted by other

workers by using segments of nongerminated mature embryos of both wild
(Orinos and Mitrakos 1991) and cultivated olive (Mitrakos et al. 1992). They
dissected the embryos into three parts (radicle, proximal and distal cotyledons)
and these segments were first placed on OMc basal medium (Canas and Bendadis
1988) with 5 mg/l IBA and 0.5 mg/12iP under a 16-h photoperiod at 24°C to
induce callus. (OMc is olive medium in which macro elements are replaced by
those of Bourgin and Nitsch 1967 with addition of Ig/l caseine hydrolysate.)
Subsequently, the derived calli were transferred to the regeneration medium
(OMc with or without 2.5 ~M IBA), at 7-day intervals (0, 7, 14,21). Somatic
embryogenesis was induced in both wild and cultivated olive embryo callus after
35 to 75 days in the regeneration medium. The difference in the amount of
embryogenesis depends on the origin of the explant and on the time of
permanence of explant on callus induction medium. Calli derived from radicles
showed high somatic embryogenesis (even 40%); in contrast, both cotyledonary
segments showed a low percentage of embryogenesis. Somatic embryogenesis

occurred only in calli placed on induction medium for 14 and 21 days. The
cytokinin (2iP) in regeneration medium inhibited somatic embryogenesis in all
concentrations tested as well as IBA, at a concentration higher than Img/l.
Somatic embryogenesis was not influenced by the salt concentration or by light
or dark conditions (Orinos and Mitrakos 1991).
2.3 Somatic Embryogenesis from Mature Tissues of Cultivar
Somatic embryogenesis has been obtained from whole, unexpanded leaves and
petioles of shoots regenerated from petioles of micropropagated olive cultivars,
Canino and Moraiolo. First, shoot organogenesis was induced by culturing
petioles of shoots proliferated on OM medium subcultured on two media: (1)
half-strength MS supplemented with 30 IJM TDZ (thidiazuron), 0.541JM NAA,
3.4% sucrose, and 0.6% Difco Bacto agar; (2) OMc plus 10 IJM 2iP, 2.21JM BAP,
and 3.4% sucrose and 0.6% Difco Bacto agar. In the first medium, the first shoots
regenerated from callus and showed expanded leaves, while in the second one
they originated directly from petiole tissue showing unexpanded leaf lamina if
not transferred to fresh medium. When both explants, petioles of the expanded
leaves and whole leaflets were transferred to OMc medium supplemented with
O.5IJM 2iP, O.44IJM BAP, and 0.251JM IBA, 3.4% sucrose and 0.6% agar, some
of them produced morphogenetic masses in the proximal part of the petiole.
These masses gave rise to normal embryos and teratomas and were able to form
continuous cycles of successive embryos for over 2 years in the dark; light
inhibited embryo induction. The rejuvenation acquired by the shoots through
organogenesis seems to be essential for the subsequent induction of somatic
embryogenesis.
2.4 Transformation with Rol Genes

and Regeneration Via Somatic Embryogenesis
A preliminary transformation experiment, aimed at reducing plant size and

increasing rooting ability (Rugini et al. 1991), was carried out by using Agro-
bacterium tumefaciens strain LBA 4404 carrying three rol A, B, and C genes
[Open Reading Frames (ORFs) 10, 11, 12] of Agrobacterium rhizogenes and a
gene for kanamycin resistance. The zygotic embryos were collected from seeds
harvested 75 days after full bloom and stored at 14°C for 45 days before using
them. In order to favor the bacterium attack the naked embryos were subjected
to various treatments before dipping them in a suspension of bacteria for 20 min:
(1) surface damaged with carborundum powder; (2) washed with ethylic alcohol
at 30% to eliminate mucilage on the surface; and (3) the infection was done later
on derived callus by dipping it into a bacterium suspension. The neoformed
somatic embryos and morphogenetic calli were selected in 50 mg/l of kanamycin
in half-strength MS basal medium plus 2.5 IJM zeatin and 1.5 IJM NAA. Some
Somatic Embryogenesis in Olive ( Olea europaea L.) 411
embryos and some calli from each of the above treatments were resistant to 50
mg/l kanamycin.
3 Summary and Conclusion
In the last decade significant progress has been made in developing techniques for
cloning olive, but there are still some difficulties: sterile cultures cannot be
initiated from shoot tips or meristem, buds require natural cytokinin (zeatin) to
develop into shoots, shoots or cells often require particular exogenous substances
such as polyamines to improve rooting ability and to delay senescence. In
addition, neither in vitro shoots nor callus grow easily in all cultivars; the apical
dominance is strong. Significant progress has been made on callus and cell
suspension cultures: a responsive tissue, such as petioles of several cultivars and
hormone combinations, induces and supports callus of good quality. Significant
advances have also been made in shoot regeneration from petioles of in vitro
grown shoots of several cultivars, but the regeneration capacity is still low for use
in biotechnological applications.
Some progress has also been made in somatic embryogenesis by using
immature zygotic embryos and radicles of mature ones and recently also from
mature tissues of cultivars. Correlations among embryogenesis potential and
tissue type, period of seed harvesting and storage of fruit and the effect of
exogenous growth regulators have been observed. The observation of Mitrakos
et al. (1992) on cotyledon segments demonstrated how auxin can inhibit or
stimulate embryogenesis according to the tissue used. Research should be con-
centrated on maternal tissue as well as particular petiole tissues which have been
shown to possess morphogenetic capacity after rejuvenation.
4 Protocol for the Induction of Somatic Embryogenesis
4.1 From Immature Zygotic Embryo
1. Collect the immature zygotic embryos from fruitlets of the open-poJlinated flowers 75 days after
full bloom and use immediately or store them at 14-15 °C for 2-3 months before use.
2. Surface-sterilize the young drupes with a mixture of350 mgll HgCI2 and 12% sodium hypochlorite
(with 6% active chlorine) plus 7-8 drops/l of Tween 80, for 5 min under vacuum with the use of a
waterpump or use only 12% sodium hypochloride for 20 min. Then rinse in sterile water.
3. Aseptically cut the drupes transversally into two pieces at 113 of the distance from the proximal
portion and excise the embryos with a scalpel and forceps. Place one embryo horizontally on each
of 25-hole multiwell plates containing 3 ml solid medium per hole.
4. Medium composition: Half-strength MS medium supplemented with 0.5 or 2.5 J.lM BAP and no
more than 0.5 J.lM NAA, 2% sucrose, 0.7% agar (Difco Bacto). Autoclave medium at 121°C for
10 min after pH adjustment to 5.7.
5. Place the multiwells in a growth chamber at a constant temperature of 23°C in the dark.
6. After 6 weeks, subculture the embryos with their callus to the same fresh basal medium with 0.5 to
2.5 ~m zeatin and 0.5 ~M NAA, in order to induce further embryo formation.
7. Subculture the neoformed embryos in the original OM medium plus 2.5 ~M zeatin, 3% sucrose,
and 0.7% agar to induce embryo germination.
8. Transfer germinated embryos to substrate in pots.
9. Eventual dormancy can be prevented by spraying with 400 mg/l GA3 water solution.
4.2 From Radicle Callus of Mature Olive Zygotic Embryos
I. Harvest mature drupes, free the stones and wash them in a solution of 2% NaOH. After mixing
them with sand, store for 3-4 months before use. Extract the seeds from mechanically broken
endocarp and then surface-sterilize first with ethanol (70% for 1 min) and then with commercial
bleach (6% active chlorine), plus a few drops of Tween 20 per liter for 20 min). Rinse three times
with sterile-distilled water and imbibe for 24 h in darkness at 24°C.
2. Excise embryos using a scalpel or forceps and dissect radicles. Place the radicles on OM modified
medium (OMc) (Canas and Benbadis 1988) supplemented with 2.5 ~M, 2iP, 25 ~M IBA and add
tenfold more concentration of folic acid. Use 30 ml medium in each 9-cm Petri dish with 30-35
dissected radicles. Place them under a 16-h photoperiod at 4 W/m' at 24°C.
3. Transfer the radicle calli after 3 weeks to OMc supplemented with or without 0.5 or 2.5 ~M IBA
either under a 16-h photoperiod or in the dark.
4. Transfer the easily separable embryos from the callus, which will appear after 35 days in culture,
to half-strength MS plus 50 mg/l myo-inositol, 3 g/I sucrose, and 0.8% w/v agar, or better to OM
plus 2.5 ~M zeatin for germination.
4.3 From Mature Tissues of Cultivars
1. Collect the petioles from growing shoots on OM medium with 18 to 23 ~M zeatin, not later than
30 days after the last subculture; place 20 petioles in Petri dishes with 20 ml of one of the following
media:
a) for indirect regeneration: half-strength MS supplemented with 30 ~M TDZ (thidiazuron),
0.54 ~M NAA, 3.4% sucrose; and 0.6% Difco Bacto agar;
b) for direct regeneration: OMc plus 10 ~M 2iP, 2.2 ~M BAP, 3.4% sucrose, and 0.6'/'0 Difco
Bacto agar;
Autoclave both media for 10 min at 121°C after pH adjustment to 5.7.
2. Place the cultures in the dark at 23 to 28°C until the production of regenerated shoot and
development oftwo or three nodes for those derived from indirect regeneration and only one node
for those derived from direct regeneration.
3. Collect the petioles or the whole, unexpanded leaves from the regenerated shoots, and place them
in Petri dishes with OMc medium containing 0.5 ~M 2iP, 0.44 ~M BAP, 0.25 ~M IBA, 3.4%
sucrose, and 0.6% agar to induce proembryo masses.
4. Place the proembryo masses on filter paper embedded with OMc medium with the same growth
regulators reported in point No.2 to induce embryo formation from their epidermal layer.
5. Subculture the neoformed embryos or teratomas in the same liquid medium or add monthly an
equivalent volume of fresh medium aftcr removing the spent one by pipetting to induce cyclic
embryogenesis.
6. Germinate the normal embryos in abundant OMc medium in a shaker at 80 rpm for about I week
in the light photoperiod, then transfer the germinated embryos to Jiffy pots in a box sealed with
polyethylene film, under a 19-h light photoperiod at 22-23 dc.
Acknowledgments. The authors thank Kim Manzi for her assistance in preparing this paper in English
and Taratufolo Claudio and Camilli Mariano for technical assistance.
References
Bao Z-H, Ma Y-F, Liu I-F, Wang K-J, Zhang P-F, Ni D-X, Yang W-Q (1980) Induction ofplantlets
from the hypocotyl of Olea europaea L. in vitro. Acta Bot Sin 2: 96-97
Bourgin IP, Nitsch IP (1967) Obtention de Nicotiana haploides a partir d'etamines cultivees in vitro.
Ann Physiol Veg 9: 377-383
Canas LA, Benbadis A (1988) Plant regeneration from cotyledon fragments of the olive tree (Olea
europaea L.). Plant Sci 54: 65-74
Canas LA, Wyssmann AM, Bendabis MC (1987) Isolation, culture and division of olive (Olea
europaea L.) protoplasts. Plant Cell Rep 6: 369-371
Chevalier A (1948) L'origine de l'olivier cultive et ses variations. Rev Bot AppI303-304, 1-25
Cronquist A (1981) An integrated system of classification of flowering plants. Columbia University
Press, New York
Fiorino F, Leva AR (1986) Investigation on the micropropagation of the olive (Olea europaea L.).
Influence of some mineral elements on the proliferation and rooting of explants. Olea 17:
101-104
Lavee S, Adiri N (1974) The effects ofabscisic acid and gibberellic acid on the development of apple and
olive callus. In: Plant growth substances, part VI. Tissue Culture Proc 18th Int Conf, Hirokawa,
Tokyo, pp 1141-1148
Lavee S, Messer G (1969) The effect of growth-regulating substances and light on olive callus growth
in vitro. J Exp Bot 20: 604-614
Leva AR, Petruccelli R, Goretti R, Panicucci M (1992) Ruolo di alcuni microelementi e carboidrati
nella proliferazione "in vitro" di cv di olivo (Olea europaea L.). In: Atti Qualita Olio Extravergine
di Oliva, Regione Toscana, Giunta Regionale. Firenze 1-3 Dec, pp 333-334
Mencuccini M, Rugini E (1993) In vitro shoot regeneration from Olive cultivar tissues. Plant Cell
Miller CO, Skoog F (1953) Chemical control of bud formation in tobacco stem segments. Am J Bot 40:
768-773
Mitrakos K, Alexaki A, Papadimitriou P (1992) Dependance of olive morphogenesis on callus origin
and age. J Plant Physiol 139: 269-273
Orinos T, Mitrakos K (1991) Rhizogenesis and somatic embryogenesis in calli from wild olive [Olea
europaea var. sylvestris (Miller) LehrJ mature zygotic embryos. Plant Cell Tissue Organ Cult 27:
183-187
Roselli G, Scaramuzzi F (1974) Definizione dei termini di cultivar e clone. Atti Accad Agric Bologna 2:
119
Rugini E (1984) In vitro propagation of some olive (Olea europaea L.) cultivars with different root-
ability, and medium development using analytical data from developing shoots and embryos.
Sci Hortic 24: 123-134
Rugini E (1986) Olive (Olea europaea L.). In: Bajaj YPS (ed) Biotechnology in agriculture and forestry,
vol I. Trees I. Springer, Berlin Heidelberg New York, pp 253-267
Rugini E (1988) Somatic embryogenesis and plant regeneration in olive (Olea europaea L.). Plant Cell
Rugini E, Fedeli E (1990) Olive (Olea europaea L.) as an oilseed crop. In: Bajaj YPS (ed) Biotechnology
in agriculture and forestry, vol 10. Legumes and oilseed crops I. Springer, Berlin Heidelberg
Rugini E, Lavee S (1992) Olive biotechnology. In: Hammerschlag FA, Litz R (eds) Biotechnology of
perennial fruit crops. CAB Int, Cambridge, UK, pp 371-382
414 E. Rugini et al.: Somatic Embryogenesis in Olive (Olea europaea L.)
Rugini E, Bazzoffia A, Jacoboni A (1988) A simple in vitro method to avoid the initial dark period and
to increase rooting in woody species. Acta Hortic 227: 438-440
Rugini E, Tarini P (1986) Somatic embryogenesis in olive (Olea europaea L.). Moet-Hennessy Conf
Fruit tree biotechnology, Paris, p 62 (Abstr)
R ugini E, Pellegrineschi A, Mencuccini M, Mariotti D (1991) Increase of rooting ability in the woody
species kiwi (Actinidia deliciosa A. Chev.) by transformation with Agrobacterium rhizogenes
T-DNA rol genes. Plant Cell Rep 10: 291-295
Taylor H (1945) Cyto-taxonomy and phylogeny of the Oleaceae. Brittonia 5: 337-367
111.13 Somatic Embryogenesis in Norway Spruce
(Picea abies)
S. VON ARNOLD, D. CLAPHAM, U. EGERTSDOTTER,
I. EKBERG, H. Mo, and H. YIBRAH 1
1 Introduction
1.1 Picea abies - an Important Tree Species
The genus Picea (family Pinaceae) consists of36 species and about 100 subspecies
and varieties. Picea is widely distributed in the northern hemisphere. Picea abies
(L.) Karst, Norway spruce, is widespread in the central and northern parts of
Europe and it is the most common tree in Sweden. It thrives in all soils, except
heavy clays and badly waterlogged soils. The trees are generally intolerant of
pollution. The wood is light in weight, soft, long-fibered, elastic, slightly resinous
with a few and scattered and cream-colored or white resin ducts. Picea abies is
one of the most important coniferous species in Europe for timber, pulp and
paper production. It also has numerous other applications, including general
joinery and carpentry, packing cases, boxes, musical instruments, fuel wood, etc.
Picea abies plays a vital role in the international trade balance, especially for the
Nordic countries.
Natural regeneration of P. abies is by seeds. However, it can also propagate
vegetatively through the lowest branches which, when pressed to the soil, can
differentiate roots and subsequently new shoots. Vegetative propagation occurs
in the alpine regions where seed production frequently fails.
In Sweden flowering occurs in spring, usually in May. The male flowers are
formed predominantly in the lower part of the crown, while the female flowers
are concentrated on the apical branches and exclusively in the terminal buds.
Flowering occurs infrequently, on average every 4th year in the southern part of
Sweden but only every 10th year in the northern part. The seeds are dispersed the
year after flowering. The viability of the seeds is usually very high (about 95%).
The seedlings have six to nine cotyledons. Growth is slow during the first 15
years, then it accelerates until the trees reach a height of about 20 m, after which
the growth rate slows down. In the southern parts of Sweden the trees are
harvested after 70 to 80 years and in the northern part after 120 to 150 years. The
trees can reach a height of 40 m and an age of 500 years. The reason why it can
become so tall is that the apical shoot continues to grow during the whole life
I Swedish University of Agricultural Sciences, Uppsala Genetic Center, Department of Forest
Genetics. Box 7027, S-750 07 Uppsala, Sweden

416 s. von Arnold et al.
span. The root system is shallow and the trees are easily felled by storms. The
shallow rooting also makes the trees sensitive to water deficiency. Picea abies has
only long shoots; the needles, which are spirally arranged along the branches,
remain for 5 to 8 years. For more detailed information, see reviews by Schmidt-
Vogt (1977, 1986).
The tree morphology varies greatly among different varieties, perhaps the
best known being Picea abies var. virgata which almost lacks branches.
From many tree species young plants can be propagated vegetatively through
cuttings. However, this technique is at present too inefficient for the production
of plants for progeny testing or for reforestation. The main limitations are that
only a limited number of shoots can be produced from each plant and that
rooting of the shoots is inadequate. To some extent the problems can be
overcome by using tissue culture techniques. Plant regeneration in vitro can be
achieved in three different ways: (1) regeneration from existing meristems; (2)
regeneration from adventitious meristems; (3) regeneration via somatic embryos.
Somatic embryogenesis can be very efficient for micropropagation and in some
respects it is preferable to the other methods based on bud differentiation. The
advantages with somatic embryogenesis are: the somatic embryos can be propa-
gated on a large scale in bioreactors, a high yield of plants can be obtained in a
short time, the embryos already have a taproot, the embryos can be encapsulated
and treated like seeds, true rejuvenation can be obtained if the somatic embryos
are regenerated from mature trees, and somatic embryos can be cryopreserved
for germplasm storage.
Picea abies was the first conifer from which somatic embryos were obtained.
Since then, the number of coniferous species from which somatic embryos can be
achieved has continuously increased and today we have no reason to believe that
there are species from which somatic embryos cannot be obtained. In many
laboratories Picea abies has been chosen as a model system. Since somatic
embryos were obtained from immature zygotic embryos of Picea abies in 1985,
there have been many reports (Table 1).
2.1 From a Cell to a Plant
Plant regeneration via somatic embryogenesis includes four different procedural

steps: (1) initiation of embryogenic cultures (Fig. 1A,B); (2) proliferation of
embryogenic cultures (Fig. 1C); (3) maturation of somatic embryos (Fig. 1D);
Somatic Embryogenesis in Norway Spruce (Picea abies) 417
Table 1. Publications on somatic embryogenesis of Picea abies
Authors Year Content
Hakmanetal. 1985 Initiation of embryogenic cultures from

immature zygotic embryos
Hakman and von Arnold 1985 Plant regeneration from somatic embryos
from immature zygotic embryos
Krogstrup 1986 Initiation of embryogenic
Gupta and Durzan 1986 cultures from mature zygotic
von Arnold and Hakman 1986 embryos
von Arnold 1987
Becwar et al. 1987 Quantification of the number of somatic
embryos in an embryogenic culture
Le1u et al. 1987 Initiation of embryogenic cultures from
germinated zygotic embryos
Nagmani et al. 1987 Origin and development of somatic
embryos in embryogenic cultures from
mature zygotic embryos
Wann et al. 1987 Biochemical differences between embryo-
genic and nonembryogenic callus
von Arnold and Hakman 1988a Description of the embryogenic
Becwar et al. 1988 system
von Arnold and Hakman 1988b Maturation of somatic embryos after
treatment with ABA
Boulay et al. 1988 Plant regeneration from embryogenic
suspension cultures
Jain et al. 1988 Enhanced initiation of embryogenic
cultures from mature zygotic embryos
Mo etal. 1989 Genetic stability in long-term embryogenic
cultures
Becwar et al. 1989 Maturation and germination of somatic
embryos
Feirer et al. 1989 Triglycerides in somatic and zygotic
embryos
Verhagen and Wann 1989 Somatic embryogenesis from liquid-
grown mature zygotic embryos
Berchetche et al. 1990 Cryopreservation of embryogenic cultures
Lelu et al. 1990 Somatic embryogenesis from cotyledons
from precultured germinated zygotic
embryos
Hakmanet al. 1990 Storage protein accumulation in somatic
embryos
Simola and Santanen 1990 Initiation of embryogenic cultures from the
megagametophyte
Jalonen and von Arnold 1991 Characterization of embryogenic cell lines
Kvaalen and von Arnold 1991 Effects of oxygen and carbon dioxide on
somatic embryogenesis
Mo and von Arnold 1991 Initiation and origin of embryogenic
cultures from seedlings
Bozhkov et al. 1992 Maturation of somatic embryos after
treatment with ABA and BA
Bellarosa et al. 1992 Hormonal regulation of the morphology of
embryogenic cell lines
418 S. von Arnold et al.
Table 1. (Contd.)
Authors Year Content
Newtonetal. 1992 Expression of an ABA-responsive promo-

tor in embryogenic cultures
Egertsdotter and von Arnold 1993 Isolation and regeneration of protoplasts
Ekberg and von Arnold 1993 Absence of association between embryo-
genic capacity and phenological characters
Hakman 1993 Storage proteins in somatic and zygotic
embryos
Norgaard et al. 1993 Cryopreservation of embryogenic cultures
Egertsdotter et al. 1993 Extracellular proteins
Moet al. 1994 in embryogenic cell lines
Yibrahetal. 1994 Transient transformation of embryogenic
cell lines
and (4) "germination" of somatic embryos (Fig. IE). The culture requirements
for each step are presented in Table 2. The following sections summarize our
knowledge about the process in Picea abies. Although similar results have
been reported in several coniferous species, we here only refer to the work
done with Picea abies, information about other conifers being presented in
other chapter.
2.1.1 Initiation of Embryogenic Cultures
In the beginning it was possible to initiate embryogenic tissue only from imma-
ture zygotic embryos. However, the age of the initial explant has successively
increased from immature to mature zygotic embryos and further via germinated
embryos to I-month-old seedlings (see Table 1). When mature zygotic embryos
(von Arnold 1987) or seedlings (Mo and von Arnold 1991) are cultured under the
conditions described in Table 2, 40 to 50% of the embryos/seedlings produce
embryogenic tissue. If immature embryos are cultured under identical condi-
tions, close to 100% form embryogenic tissue. A bud-inductive treatment before
transfer to callus induction medium can sometimes stimulate the induction of
embryogenic tissue (Lelu et al. 1990).
When the primary explants are cultured on a medium containing both auxin
(2,4-D) and cytokinin (BA) three different kinds of tissues are formed: (l) a green,
nodulated tissue with meristemoids which under suitable conditions develop
further into adventitious buds; (2) a nonembryogenic callus composed of small,
rounded cells which continue to grow in the same way for years; (3) an
embryogenic tissue which is translucent, mucilagenous, and composed of many
small somatic embryos. Attempts have been made to characterize embryogenic
and nonembryogenic callus cultures by using biochemcial methods (Wann et al.
1987). Differences have been found in ethylene production, in the amount of
gluthathione, and in the reducing power. However, it is not clear whether these
Fig. lA-F. Various developmental stages during the process of plant regeneration via somatic
embryogenesis from seedlings of Picea abies. A Embryogenic structure protruding from an 18-day-old
seedling cultured for 2 months on medium containing auxin and cytokinin (c cotyledons; h hypocotyl).
8 Higher magnification of similar embryogenic structures as shown in A. C Embryogenic culture
containing a large number of undeveloped somatic embryos. D Somatic embryos in various develop-
mental stages, 1month after transferring the embryogenic culture to medium containing abscisic acid.
E Germinating somatic embryo. FA Picea abies plant derived from a somatic embryo. The plant has
been grown in the field for 5 years. Scale bars represent 2 mm in A; 0.2 mm in 8 ; 1.5 mm in C and D;
1 mm in E; and 7 cm in F. (Mo et al. 1989; Mo and von Arnold 1991)
substances play a causal role or are merely associated with embryogenic

conditions.
The culture conditions for initiation of embryogenic tissue are slightly
different in several of the studies listed in Table 1. However, a striking feature is
that fairly similar results are obtained in different laboratories throughout the
world when using the same protocol. In general, the culture conditions for
obtaining embryogenic tissue from immature zygotic embryos are not very
critical as long as the medium contains both auxin and cytokinin. In contrast,
many factors have to be considered when older plant materials are used as
explants. Important factors to consider are: light regime; gas phase; strength of
basal medium; concentrations of sucrose and NH4N0 3; type and concentrations

of auxin and cytokinin; pH of the medium; medium hardness. However, it is
difficult to stress the importance of any single factor, since they all depend upon
one another. For example, the partial pressure of oxygen affects the initiation
frequency differently depending on the strength of the basal medium and/or
on the concentration of NH 4N0 3 (Kvaalen and von Arnold 1991). Low p02
stimulated the formation of embryogenic tissue on zygotic embryos when they
were incubated on full-strength medium, but was inhibitory when half-strength
medium was used.
Which cells in the explant give rise to somatic embryos has been much
discussed. In most cases it seems that in zygotic embryos the cells in the hypocotyl
and in the cotyledons have the capacity to regenerate somatic embryos (Lelu
et al. 1987; Nagmani et al. 1987; von Arnold and Hakman 1988a). However, it
has also been suggested that cells in the radicle have this ability (Gupta and
Durzan 1986). In seedlings, embryogenic structures differentiate from the epi-
cotyl, hypocotyl, and cotyledons (Mo and von Arnold 1991). The primary event
during initiation from seedlings is the process by which a few mesophyll cells
(epidermal and subepidermal or cortical) become meristematic and develop into
nodules. Irrespective of whether the nodules are associated or disassociated from
the primary explant, they continue to produce somatic embryos. The nodules
which differentiate from epidermal and subepidermal cells have a similar appear-
ance of those formed during initiation of adventitious buds. Therefore, it is very
likely that the earliest events during differentiation of adventitious buds and
somatic embryos are similar. However, after the formation of nodules the
developmental patterns are completely different.
Table 2. Culture requirements during various stages of somatic embryogenesis in Picea abies
Initiation Proliferation Maturation Germination
Medium LP x l/2+extra LP x l/2+extra LP xl LP xl

NH4N0 3 NH4N0 3
sucrose 1% 1% 3% 2%
agar/gellen 0.7/0.2% 0.7/0.2% 0.7/0.2% 0.7/0.2%
pH 6.1 6.1 6.1 5.8
Hormones 2,4-D 2,4-D ABA 0
BA BA
Light Darkness Darkness Darkness 1 week in darkness
1 week in light
Temperature 20-25°C 20-25°C 20-25% 20-25°C
Gas phase High High Low O2 and

O2 and CO 2 O2 and CO2 high CO2
Time 3-8 weeks Years 4-8 weeks

2.1.2 Proliferation of Embryogenic Tissue
Although embryogenic tissue is frequently initiated, fast-growing embryogenic

cultures from which embryogenic cell lines can be established are only formed
from about 5% of the initial explants. Somatic embryo proliferation frequency
can be significantly increased if the primary explants are treated carefully, but still
it is not known if embryogenic cell lines can be established from all genotypes.
The culture requirements are similar for proliferation of embryogenic tissue and
for initiation (Table 2).
Fig. 2A, B. Somatic embryos from different types of embryogenic cell lines of Picea abies. A Somatic
embryos belonging to group A. Note the densely packed cells in the embryonic region and the long,
vacuolated suspensor cells. B Somatic embryos belonging to group B. Note the loose aggregated cells
in the embryonic region. Scale bars represent 0.3 mm (Bellarosa et al. 1992)
Proliferating embryogenic cultures are always mucilagenous and translucent

and they consist of numerous somatic embryos. However, the texture of the
cultures as well as the organization ofthe somatic embryos varies. Embryogenic
cultures can be divided into two main categories (Fig. 2): (A) those consisting of
somatic embryos with a distinct embryonic region which is clearly separated
from the suspensor region, the embryos being either polarized with suspensor
cells protruding in one direction or solar shaped with suspensor cells protruding
all around the embryonic region; and (B) those consisting of somatic embryos
where the suspensor cells are intermingled with the cells in the embryonic region.
The cell lines continue to grow in the same way for years and a group A cell line
is not transformed to a group B cell line or vice versa. Only somatic embryos
belonging to group A undergo maturation under normal culture conditions. Our
hypothesis is that somatic embryos belonging to group B are blocked in their
development, and therefore are not able to respond to ABA (see further Sect.
2.1.3). In addition to the morphological differences between the two groups of
cell lines, they also differ in their secretion of proteins. That the secreted proteins
influence the development of the embryos was shown by adding secreted proteins
from one group of cell lines to another. Extracellular proteins which vary
between group A and group B embryos, and therefore have putative regulatory
effects on embryo development, include arabinogalactan proteins, zeamitia-like
proteins, chitinases and peroxidases (Egertsdotter et al. 1993; Mo et al. 1994).
Embryogenic cultures grown on solidified medium can also be transferred
to liquid media. Embryogenic cell suspensions maintained for several years
are today routinely used in many laboratories. We have found that cell lines
belonging to group B are easy to run as suspension cultures, while those
belonging to group A are more difficult. In general, it seems that the group A
cultures need to be subcultured regularly and each cell line has its specific
requirements. The reason for these differences are not clear. Embryogenic
cultures can also be cryopreserved (Norgaard et al. 1993).
When cultured on a medium containing both 2,4-D and BA, new somatic
embryos are continuously formed. A drastic change in the development and
morphology of the somatic embryos takes place when they are cultured without
one or both of the growth regulators and no new somatic embryos are formed
(Bellarosa et al.1992). The pattern of extracellular proteins reflects the morpho-
logy of the embryos (Mo et al. 1994). On medium containing only BA the already
existing somatic embryos increase in size and develop an extremely large embry-
onic region. On medium containing only 2,4-D the already existing embryos
disorganize into loose aggregates. When transferred to medium containing both
2,4-D and BA new normal-looking embryos directly differentiate from the
nodules on BA-containing medium and from the single cells on the 2,4-D-
containing medium. Our conclusions from these studies are that both auxin and
cytokinin are required for proliferation of new somatic embryos, that cytokinin
is required for retaining the organization of the somatic embryos, and that auxin
is required for the differentiation of new somatic embryos.
How the embryogenic cultures proliferate is not clear. It seems that prolifera-
tion occurs in several ways (Nagmani et al. 1987; von Arnold and Hakman
1988a; Bellarosa et al. 1992): (1) new somatic embryos differentiate from cells in
the embryonic region; (2) new somatic embryos differentiate from small cells
within the suspensor region; and (3) new somatic embryos differentiate from
single cells or small cell aggregates. The acetocarmine-Evan's blue staining
(Gupta and Durzan 1987) allows a distinction between embryonic cells and
suspensor cells (Bellarosa et al. 1992). The vacuolated cells of the suspensor
always have strong blue nuclei and weak blue cytoplasm. In the embryonic region
two types of cells can be recognized: (1) cells with a red nucleus and blue
cytoplasm, located in the superficial layer, and (2) completely red cells, constitut-
ing the inner part of the embryonic region. From our studies it seems that the
blue-red stained cells, located superficially in the embryogenic region, are those
which differentiate new somatic embryos.
2.1.3 Maturation of Somatic Embryos
Somatic embryos can complete development into plantlets without first going
through a maturation process. However, the yield of plantlets is much higher if
the embryos are first stimulated to go through a maturation process in which the
embryos stop proliferating, increase in size and accumulate storage material,
especially lipids but also carbohydrates and proteins (Feirer et al. 1989; Hakman
et al. 1990; Hakman 1993). A mature somatic embryo is characterized by a firm
structure and a glossy surface. It resembles a mature zygotic embryo (Fig. ID).
Somatic embryos are stimulated to mature by treatment with ABA (von Arnold
and Hakman 1988b; Becwar et al. 1989). The optimal ABA treatment varies
from 7 to 60 J.1M for 1 to 3 months dt"Jending on the cell line. On medium with
low ABA concentrations maturation can be stimulated by lowering the pressure
of oxygen and increasing the pressure of the carbon dioxide (Kvaalen and von
Arnold 1991).
The ability of somatic embryos to respond to ABA by maturation varies
significantly among different cell lines. This maturation response is one of the
main problems which have to be solved before the somatic embryos will be
valuable for clonal propagation. Normal mature somatic embryos are today
only obtained from cell lines classified as group A (Fig. 2A). The best cell lines
give hundreds of mature somatic embryos per gram tissue.
2.1.4 "Germination" of Somatic Embryos
Mature somatic embryos can develop further when transferred to medium

lacking ABA (Table 2; Fig. IE). Desiccation treatment of mature somatic
embryos stimulates the development of plants. Most of the regenerated plants
have a normal appearance. Plant regeneration from somatic embryos is routine
in most laboratories working with embryogenic cultures of Picea abies,
although maturation and regeneration are usually restricted to a limited number
of cell lines.
A significant variation in growth pattern exists between embryos from
different cell lines (Jalonen and von Arnold 1991). The majority of the cell lines
give rise to plants with a single taproot, which usually is covered by root hairs.
However, plants from some cell lines also form a large number of secondary
roots. Shoots from many cell lines develop roots only after prolonged culture in
vitro. In general, the shoots, either with or without roots, have a tendency to form
a resting bud. However, in a few cell lines the epicotyl continues to develop
without a rest period.
Plants regenerated from somatic embryos can be transferred to nonsterile
conditions. Often they produce a resting bud, either during tissue culture or after
transfer to nonsterile conditions, and they require a budbreak treatment before
they start to grow fast and increase in size. Regenerated plants can be transferred
to the field where they grow normally (Fig. IF).
2.2 Absence of Associations Between Embryogenic Capacity

and Phenological Traits
An important question is whether directional selection occurs when a standard

method for initiation of embryogenic cultures is used, in other words, whether
there is an associated between embryogenic traits and traits of physiological
importance such as time of bud flushing and growth cessation. A well-adapted
growth rhythm is of great significance for avoiding injuries caused by late spring
frosts and early autumn frosts, or for the ability to attain adequate winter
hardiness.
Embryogenic cultures were established from seeds collected from individual
trees for which growth rhythm data had been recorded in a nursery trial with
single-tree progenies from these trees. The seed collections were made from
randomly selected trees in two Swedish Picea abies populations. There was a
large variation in the ability to initiate somatic embryos and to produce mature
somatic embryos between the single tree families in both popUlations (Ekberg et
al. 1993). No associations between embryogenic traits in vitro and the growth
rhythm trait assessed in the nursery could be demonstrated.
2.3 Genetic and Physiological Stability of Regenerated Plants
The growth characteristics, morphogenic capacity, and DNA content of

embryogenic cultures after long-term culture in vitro have been analyzed (Mo
et al. 1989). Somatic embryos in embryogenic cultures were first stimulated to
mature and then either to develop further into plants or to differentiate new
embryogenic cultures. The procedure was repeated three times over 2 years. The
ability to give rise to new embryogenic cultures or to develop into plants was
similar for all somatic embryos irrespective of how long they had been cultured
in vitro. The nuclear DNA content, measured in a flow cytometer, was estimated
at 32 pg/G 1 nuclei in seedlings developed from zygotic embryos. Nuclei isolated
from embryogenic cultures and from plantlets regenerated from somatic
embryos had the same DNA content as those isolated from seedlings. These
results show that no change occurs in growth characteristics after prolonged
culture and that the cultures remain at the diploid stage. However, smaller
quantitative or qualitative changes at the gene level cannot be observed by this
technique.
If plants regenerated from somatic embryos are to be used in reforestation
programs, it is very important that they grow like seedlings of the same genotype.
The growth capacity must be the same as expected. Large field trails run at
Weyerhauser have shown that plants regenerated from somatic embryos grow
normally in the field (Gupta, pers. comm.). Furthermore, the growth pattern
must be as expected, e.g., bud flushing must occur after a certain temperature
sum and bud setting after a certain number of long nights. The growth pattern
must also change in the right way when the plants get older. Normally, the period
of growth shortens successively each year, but the shortening varies depending on
the genotype. If the growth pattern has changed for the plants regenerated from
somatic embryos, there is a significant risk that the plants will be damaged by late
spring frost or early autumn frost. At present, we are running experiments in a
phytotron in order to test whether plants regenerated from somatic embryos
retain the same growth characteristics as expected for plants cultivated from
seeds. The preliminary data obtained so far indicate that the somatic embryos
have the same growth characteristics as isolated zygotic embryos from the same
open-pollinated single tree family. However, the growth characteristics of
isolated zygotic embryos differ from those of seedlings. Isolated zygotic embryos,
as well as somatic embryos, had more mature growth characteristics. These
results indicate that the culture conditions during germination of embryos in
vitro must be improved. But the results also show that the procedure for
producing somatic embryos does not induce aberrations. However, it is impor-
tant to stress that the results we obtain are based on our standard methods for
induction, proliferation, and maturation of somatic embryos. If the routines are
changed, aberrations might be induced.
2.4 Gene Transfer to Somatic Embryos
The greatest advantages of the gene transfer technique in forest trees will be to
study in more detail the genetic regulation of specific traits. However, in the
future, the technique might also become useful for certain breeding purposes. At
present, we are developing techniques for producing transgenic plants.
The 1.5-kilobase promoter sequence upstream of De8, a late embryo-
abundant gene of Daucus carota, fused to the reporter J3-glucoronidase gene was
introduced into several tissues of Picea abies via a custom-made electric-dis-
charge particle accelerator (Newton et al. 1992). Transient expression was
measured histochemically as spot number, 2 days after bombardment (Fig. 3A).
Often the expression units, the blue spots showing GUS activity, represented an
area larger than one cell. However, observation of serial microtome sections of
plastic-embedded somatic embryos indicated that spots were usually comprised
a single, darkly stained cell with the substrate diffusing to surrounding cells (Fig.
3B). In some cases, it was possible to identify a gold particle in the stained cells.
These gold particles were associated with the nuclei.
Fig. 3. Transient expression of the GUS reporter gene after bombardment of an embryogenic
suspension (see also legend to Fig. 4). A An overview of suspension cells collected on filter paper.
Several cells are expressingp-glucuronidase (arrows); bar = 0.5 mm. 8 Section through an embryonic
region showing a transformed cell with a gold particle in the nucleus (arrow); many other gold particles
are visible; bar = 20 !lm. (Newton et al. 1992)
Embryogenic suspension cultures gave higher levels of expression depending

on the cell line than embryogenic callus cultures or zygotic embryos. Expression
was enhanced when cultures were treated with ABA for 3 days before bombard-
ment. A mean and maximum of 17 and 34 spots/disk, respectively, were observed
with the best cell line, which was comparable with the level of expression driven
by an enhanced 35S promoter. Subsequently, the expression of the reporter gene
was related to the time of bombardment after subculture of the embryogenic
suspensions (Yibrah et al. 1994). Interestingly, with cell line E86: 17, the expres-
sion driven by De8 in the presence of ABA increased sixfold when the interval
between subculture and bombardment was increased from 3 to 6 days, and then
declined (Fig. 4). The expression driven by the enhanced 35S promoter increased
less than 50% over the same period. The difference in the time course of
expression was similar for all tested cell lines. This indicates that transient
expression varies with the time of bombardment after subculture because cells
change in their competence to transcribe from individual promoters, rather than
because they change in their general competence for transformation. Transient
gene expression varies significantly depending on the promoter.
---tIt-- 70S
III
:!:: 60 --+- DeB (+ABA)
c ---- DeB (-ABA)
~
c 40
0
.
'iii
III
CD
Q. 20
>C
W
0
3 5 7 9 11 13
Days
Fig. 4. Effect of the presence of ABA in the culture medium, and time of bombardment after
subculturing, on GUS expression driven by an ABA-responsive promoter, De8. Embryogenic suspen-
sion cultures (line E86: 17) were bombarded with gold particles coated either with 2x35S-GUS (the
plasmid pJIT65 from Dr. F. Guerineau containing the structural gene for ~-glucuronidase driven by
an enhanced 35S promoter), or with De8-GUS (a plasmid from Dr R Sung containing the GUS
gene driven by the carrot De8 promoter) at 3,6,9, or 12 days after subculture. One day before
bombardment, ABA (15IJM) was added to half of the cultures that were subsequently bombarded with
De8-GUS. Two days after bombardment substrate was added to the cultures and GUS activity was
measured as the number of expression units developing after 48 h. Each point represents the mean of
10 cultures bombarded; bar indicates standard error. (Yibrah et al. 1994)
2.5 Isolation and Regeneration of Somatic Embryos from Protoplast
Protoplast can be isolated from all embryogenic cell lines tested; we have been
working with 20 different cell lines belonging to both groups A and B (Fig. 2).
However, yield, viability, division frequencies, and aggregate formation vary
significantly among different cell lines (Egertsdotter and von Arnold 1992).
Newly isolated pro top lasts can be divided into two groups based on their
size, 30 and 50 IJ.m (Fig. 5A). Cell-wall regeneration takes place after 1 day and
division starts after 2 days (Fig. 5B). The division frequency varies from 7 to 48%
depending on cell line and treatment. Formation of cell clusters resembling
somatic embryos takes place after about 1 week (Fig. 5C). Protoplasts from most
cell lines can form small clusters when cultured in beads of 112 LP medium
solidified with agarose. However, regeneration of new embryogenic cultures only
takes place if the pro top lasts are cultured in liquid medium in a tissue culture
insert (Fig. 5D).
Only protoplasts derived from group B suspension cultures can establish
new embryogenic cultures from protoplasts. Similar result have been obtained
with other coniferous species, namely that regeneration of somatic embryos
only occurs from protoplasts isolated from cell suspensions of the lowest
embryogenic potential. An obvious limitation is that regeneration of plants is
very restricted from this type of cell line.
428 S. von Arnold et a\.
Fig. SA-D. Culturing of protoplasts. A Newly isolated protoplasts; B small cluster after 5 days;
C polarized cluster 10 days after preparation; D regenerated embryogenic culture after I month
in culture. Scale bars represent 50 IJ.Ill in A and B; 20 IJ.Ill in C; 100 J.1M in D (Egertsdotter and
von Arnold 1992)
3 Conclusions
It is possible to stimulate the formation of somatic embryos that can develop

further into plantlets from juvenile tissue (zygotic embryos and seedlings) of
Picea abies. The somatic embryos arise from nodules which differentiate from
epidermal/subepidermal cells or from cortical cells. This shows that differentiated
cells have the capacity to dedifferentiate to such a stage that they can develop
somatic embryos. The initiated somatic embryos can proliferate and give rise to
embryogenic cultures. Although all embryogenic cultures consist of somatic
embryos, the structure and the growth characteristics of the somatic embryos
vary in different cell lines. Embryogenic cell lines can be divided into two main
groups based on : (l) morphology; (2) ability to grow as suspension cultures; (3)
capacity to regenerate new embryogenic cultures from protoplasts; (4) pattern of
extracellular proteins; and (5) ability to develop mature somatic embryos. New
genes can be introduced and expressed in the somatic embryos. Mature somatic
embryos, developed after treatment with ABA, can develop further into plants.
Plants transferred to the field survive and grow normally. Initiation of embryo-
genic cultures and the subsequent development of plants is restricted to a few cell
lines when using one standard protocol. However, so far we have not observed
that the embryogenic trait is associated with any growth characteristics of plants
from the same single-tree family.
References
Becwar M, Noland T, Wann S (1987) A method for quantification of the level of somatic
embryogenesis among Norway spruce cell lines. Plant Cell Rep 6: 35-38
Becwar M, Wann S, Johnson M, Verhagen S, Feirer R, Nagmani R (1988) Development and
characterization of in vitro embryogenic systems in conifers. In: Ahuja MR (ed) Somatic cell
genetics of woody plants. Kluwer, Dordrecht, pp 1-18
Becwar M, Noland T, Wyckoff J (1989) Maturation, germination and conversion of Norway spruce
(Picea abies L.) somatic embryos to plants. In Vitro Cell Dev Bioi 25 (6): 525-580
Bellarosa R, Mo LH, von Arnold S (1992) The influence of auxin and cytokinin on proliferation and
morphology of somatic embryos of Picea abies. Ann Bot 70: 199-206
Berchetche J, Galerne M, Dereuddre J (1990) Augmentation des capacites de regeneration de cals
embrygogenes de Picea abies (L) Karst. apres congelation dans I'azote liquide. CR Acad Sci Paris
310: 357-363
Bozhkov PV, Lebedenko LA, Shiryaeva (1992) A pronounced synergistic effect of abscisic acid and
6-benzyladenin on Norway spruce (Picea abies) somatic embryo maturation. Plant Cell Rep II:
386-389
Boulay MP, Gupta PK, Krogstrup P, Durzan DJ (1988) Development of somatic embryos from cell
suspension cultures of Norway spruce. Plant Cell Rep 7: 134-137
Egertsdotter U, von Arnold S (1993) Classification of embryogenic cell lines of Picea abies as regards
protoplast isolation and culture. J Plant Physiol 141: 222-229
Egertsdotter U, Mo H, von Arnold S (1993) Extracellular proteins secreted to the culture medium
of embryogenic suspension cultures of Norway spruce (Picea abies). Physiol Plant 88: 315-321
Ekberg I, Norell L, von Arnold S (1993) Are there any associations between embryogenic capacity and
phenological characteristics in two populations of Picea abies. Can J For Res 23: 731-737
Feirer R, Conkey J, Verhagen S (1989) Trygiycerides in embryogenic conifer calli: a comparison with
zygotic embryos. Plant Cell Rep 8: 207-209
Gupta PK, Durzan DJ (1986) Plantlet regeneration via somatic embryogenesis from subcultured callus
of mature embryos of Picea abies (Norway spruce). In Vitro Cell Dev Bioi 22: 685-688
Gupta PK, Durzan DJ (1987) Biotechnology of somatic polyembryogenesis and plantlet regeneration
in loblolly pine. Bio/Technology 5:147-151
Hakman J (1993) Embryology in Norway spruce (Picea abies). An analysis of the composition of seed
storage protein and deposition of storage reserves, dummy seed development and somatic
embryogenesis. Physiol Plant 87: 148-159
Hakman I, von Arnold S (1985) Plantlet regeneration through somatic embryogenesis in Picea abies
(Norway spruce). J Plant Physiol121: 149-158
Hakman I, Fowke LC, von Arnold S, Eriksson T (1985) The developmentof somatic embryos in
tissue cultures initiated from immature embryos of Picea abies (Norway spruce). Plant Sci Lett 38:
53-59
Hakman I, Stable P, Engstrom P, Eriksson T (1990) Storage protein accumulation during zygotic and
somatic embryo development in Norway spruce. Physiol Plant 80: 441-445
Jain SM, Newton RJ, Soltes EJ (1988) Enhancement of somatic embryogenesis in Norway spruce
(Picea abies). Theor Appl Genet 76: 501-506
Jalonen P, von Arnold S (1991) Characterisation of embryogenic cell lines of Picea abies in relation to
their competence for maturation. Plant Cell 10(8): 384-387
430 S. von Arnold et a1.: Somatic Embryogenesis in Norway Spruce (Picea abies)
Krogstrup P (1986) Embryo-like structures from cotyledons and ripe embryos of Norway spruce
(Picea abies). Can J For Res 16: 664-668
K vaalen H, von Arnold S (1991) Effects of various partial pressures of oxygen and carbon dioxide
on different stages of somatic embryogenesis in Picea abies. Plant Cell Tissue Organ Cult 27 (1):
49-57
Lelu A-M, Boulay M, Armand I (1987) Obtention de cals embryogenes it partir de cotyledons de Picea
abies (L.) Karst. preieves sur des jeunes plantes agees de 3 it 7 jours apres germination. CR Acad
Sci Paris 305: 105
Lelu A-M, Boulay M, Bornman C (1990) Somatic embryogenesis in cotyledons of Picea abies is
enchanced by an adventitious bud-inducing treatment. New For 4: 125-135
Mo LH, von Arnold S (1991) Origin and development of embryogenic cultures from seedlings of
Norway spruce (Picea abies). J Plant Physiol138: 223-230
Mo LH, von Arnold S, Lagercrantz U (1989) Morphogenic and genetic stability in long-term
embryogenic cultures and somatic embryos of Norway spruce [Picea abies (L.) Karst.]. Plant Cell
Rep 8: 375-378
Mo H, Egertsdotter U, von Arnold S (1994) Secretion of specific extracellular proteins by somatic
embryos of Picea abies is dependent on embryo morphology. Plant Sci (in press)
Nagmani A, Becwar MR, Wann SR (1987) Single-cell origin and development of somatic embryos in
Picea abies (L.) Karst. (Norway spruce) and Picea glauca (Moench) Voss (White spruce). Plant Cell
Rep 6: 157-159
Newton RJ, Yibrah HS, Dong N, Clapham D, von Arnold S (1992) Expression of an abscisic acid
reponsive promotor in Picea abies following bombardment with electrical discharge. Plant Cell
Rep 11: 188-191
Norgaard JV, Durus V, Johnson 0, Verogstrup P, Baldursson S, von Arnold S (1993) Variation in
cryotolerance of embryogenic Picea abies cultures and the relation to genotype, family, culture
morphology, maturation competence and frost hardiness. Can J For Res 23: 2560-2567
Schmidt-Vogt H (1977, 1986) Die Fichte. Ein Handbuch in zwei Biinden. Paul Parey, Hamburg
Simola L, Santanen A (1990) Improvement of nutrient medium for growth and embryogenesis of
megagametophyte and embryo callus lines of Picea abies. Physiol Plant 80: 27-35
Verhagen S, Wann SR (1989) Norway spruce somatic embryogenesis: higher frequency initiation
from light cultured mature embryos. Plant Cell Tissue Organ Cult 16: 103-111
von Arnold S (1987) Improved efficiency of somatic embryogenesis in mature embryos of Picea abies.
J Plant Physiol 128: 233-244
von Arnold S, Hakman I (1986) Effect of sucrose on initiation of embryogenic callus cultures from
mature zygotic embryos of Picea abies (L.) Karst. (Norway spruce). J Plant Physiol122: 261-265
von Arnold S, Hakman I (1988a) PlantIet regeneration in vitro via adventitious buds and somatic
embryos in Norway spruce (Picea abies). In: Hanover JW, Keathley DE (eds) Genetic
manipulation of woody plants. Plenum Press, New York, pp 199-215
von Arnold S, Hakman I (1988b) Regulation of somatic embryo development in Picea abies by abscisic
Wann SR, Johnson MA, Noland TL, Carlson JA (1987) Biochemical differences between embryogenic
and nonembryogenic callus of Picea abies (L.) Karst. Plant Cell Rep 6: 39-42
Yibrah H, Manders G, Clapham D, von Arnold S (1994) Biological factors affecting transient
transformation in embryogenic suspension cultures of Picea abies. J Plant Physiol (in press)
111.14 Somatic Embryogenesis in Black Spruce
[Picea mariana (Mill.) B.S.P'] and Red Spruce
(P. ruhens Sarg.)
L. TREMBLAY and F.M. TREMBLAY!
1 Introduction
1.1 Distribution, Economic Importance, and Morphology
Black spruce [Picea mariana (Mill) B.S.P.] is the most abundant conifer in North
America, covering the continent from Alaska to Newfoundland (Wright 1955).
It can grow in different soil types, varying from fiat, low, peaty areas to well-
drained rocky uplands (Hatcher 1963). Red spruce (Picea rubens Sarg.), a
moderately wide-ranging species, grows from the southern Appalachians to the
maritime provinces of eastern Canada (Wright 1955). Although it can occasion-
ally form pure stands, it is more often mixed with other species and grows in
soil types intermediate between well-drained uplands and bogs. Hybrids can
be found between black spruce and red spruce where they grow together
(Hosie 1980).
Spruce species represent 7.1 billion m 3 of wood volume or 31 % of the total
merchantable stock in Canada. The contribution of forest industry to the
Canadian economy in 1989 was approximately C$20 billion supporting the
employment of 888 000 people.
The wood properties of the spruce species Picea giauca, P. mariana, P.
rubens, P. engeimannii, and P. sitchensis are so similar that they are not separated
during harvesting and marketing (Hosie 1980). Their natural light color, low
resin content, and fiber characteristics make spruces much favored for the
production of pulpwood. Spruce timber is also used for general construction,
mill work, interior finishing, plywood, boxes, and crating.
Spruces have long straight trunks with a scaly bark, and dense narrow
branches that can extend to the ground in open-grown trees (Hosie 1990). The
root system is shallow and the tree not usually wind-firm. Black spruce can grow
to approximately 9 to 15 m in height and 15 to 25 cm in diameter. Trees as young
as 10 years old can bear cones. Heavy crops occur every 4 years on average, with
lighter crops almost every year. Red spruce reaches 21 to 24 m in height and 30
I Centre de Recherche en Biologie Forestiere, Faculte de Foresterie et de Geomatique, Universite

Laval, Quebec, Canada, GIK 7P4

~Springer-Verlag Berlin Heidelberg 1995
432 L. Tremblay and F.M. Tremblay
to 60 cm in diameter. Good cone production in red spruce usually begins after the
tree is 30-year-old, with good seed crops every 3 to 8 years (Fowells 1965).
Efficient methods for selecting and cloning spruce trees suitable for reforestation
programs represent a major asset to the forest industry. For example, utilization
of black spruce selected clones should result in productivity gains estimated to
20-25% within a few years and 30-35% for second generation seeds (Davidson
1990).
Cryopreservation (see Chapter II.8, this Vol.) represents another unique
feature of conifer somatic embryogenesis, allowing 10-15 years of clonal testing
to be performed while maintaining the original cells in a juvenile state. Clonal
materials also provide valuable tools for basic research on conifer physiology,
morphology, and genetics.
1.3 Brief Review of Work Done on Other Species
Somatic embryogenesis in conifers was first reported for Picea abies, using
immature zygotic embryos (Hakman et al. 1985). Several species from genera
such as Abies, Larix, Picea, Pinus, Pseudotsuga, and Sequoia can now be
regenerated via somatic embryogenesis (reviewed by Attree and Fowke 1991;
Tautorus et al. 1991). Conifer embryogenic tissues have a similar morphology,
irrespective of species or type of explant. The embryogenic tissue is white,
translucent, and composed of long cells more or less organized as suspensor
structures terminated by an immature embryo. Despite their similar morpho-
logy, differences among different embryogenic cell lines of Picea abies, Pinus
caribaea, and Picea glauca-P. engelmannii complexes have been observed in
relation to their competence for maturation (Laine and David 1990; Roberts
et al. 1990a; Jalonen and von Arnold 1991).
2.1 Induction
Red spruce embryogenic tissue was first reported using mature zygotic embryos
dissected from stored seeds (Tremblay and Tremblay 1991a). An induction
frequency of29% was obtained using HLM-l medium, modified from Litvay's
medium (Litvay et al. 1985), and conditions developed on P. glauca (Tremblay
1990a). A 20% induction frequency was reported for the same species from
mature embryos using LP salt formulation (von Arnold and Eriksson 1981)
supplemented with 1 gil of casein hydrolysate, 250 mg/l ofL-glutamine, 10 11M of
either naphthalene acetic acid (NAA) or 2,4-dichlorophenoxyacetic acid (2,4-D),
Somatic Embryogenesis in Black Spruce and Red Spruce 433
10 11M benzylaminopurine (BAP), 2% sucrose, and 3% Gelrite gellan gum (Harry

and Thorpe 1991).
Black spruce somatic embryogenesis was first reported using immature
zygotic embryos as explant (Hakman and Fowke 1987). Later, use of mature
zygotic embryos (Tautorus et al. 1990) and 12- to 30-day-old germinated
seedlings (Attree et al. 1990a) gave 8-10 and 12-18% induction rates, respec-
tively. Addition of 0.5 gil casein hydrolysate to an LP-based induction medium
was found to reduce the induction frequency of immature zygotic embryos from
black spruce. In the same report, the highest induction frequency from mature
zygotic embryos was obtained on a Litvay-based medium including 1 gil casein
hydrolysate (Tautorus et al. 1990). In our laboratory, no difference for induction
requirement between immature and mature zygotic embryos has been observed
when using HLM-1 medium (unpubl.).
Induction of embryogenic tissue was found to be more difficult when
germinated seedlings were used as source of explant. In such a case, preculture
on a medium containing only a cytokinin was shown to stimulate the induction
of embryogenic tissue on Norway spruce (Krogstrup 1986; Lelu et al. 1987,
1990; Mo and von Arnold 1991), white spruce, and black spruce (Lelu and
Bornman 1990).
Recently, embryogenic tissues were obtained for 38 black spruce genotypes
originating from six control-pollinated full-sib families using a technique de-
scribed by Tremblay (l990a). All seeds were stored for 1 year before use, since
yield of embryogenic tissue was found to be higher with stored seeds than with
fresh seeds (F .M. Tremblay, unpubl.). The frequency of induction varied among
crosses (Table 1; Fig. 1).It was possible to relate the frequency of induction to the
genetics of the tree used as the male parent. For instance, induction frequencies
of 43 and 48% were obtained in the crosses 416 (d') x 433 (~) and 416 (d') x 422
(~), respectively, while an average of 4--5% induction was obtained from the
crosses 421 (d') x 81 (~), 421 (d') x 397 (~), and 421 (d') x 83 (~) (unpub1.). For
crosses involving 421 (0'), 50% of the seeds were empty. In the filled seeds, the
zygotic embryos were not well developed and often presented a folded form. The
percentage of germination of these seedlots was very low, confirming previous
results on the importance of the seed quality for the induction process (Tremblay
1990a).
Table 1. Induction frequency of embryogenic tissue originating from six control-pollinated

full-sib families, after 7 weeks on HLM -1 medium
Parental cross No. of dissected No. of zygotic Induction

zygotic embryogenic embryos with frequency
embryos tissue (%)
416 x 433 49 21 43
416 x 422 29 14 48
416 x 430 38 7 18
421 x 81 100 4 4
421 x 397 100 4 4
421 x 83 100 5 5
Fig. I. Induction of black spruce embryogenic tissue after 4 weeks in culture on HLM-J medium;
bar = 0.08 cm
2.2 Maintenance
The maintenance stage per se has not been studied well. Maintenance is generally
considered the necessary intermediate - tissue multiplication - stage between the
induction of embryogenic tissue and the maturation of somatic embryos. What
is known is that embryogenic tissue can be maintained by serial subcultures for a
long time (Fig. 2). In our laboratory (F.M. Tremblay, unpubl.), the oldest black
spruce cell line (MEE-201) has been subcultured every 2 weeks since 1986, mature
embryos have been regularly produced since 1987, and the regenerated plants
grown in soil. Six years after induction, the cell line MEE-201 still produces
mature embryos, but in lower numbers than during the first 4 years of mainte-
nance. Although embryos can still be produced, they become very difficult to
germinate. Presently, a conversion rate of less than I % (number of plants in
soil/number of mature embryos) is obtained. After acclimatization, the somatic
embryo-derived plants obtained after 6 years maintenance exhibit low vigor
under normal greenhouse conditions, while plants produced after 3 years had
much higher vigor. More research on the maintenance stage is necessary to deal
with the genetic or epigenetic effects that can result from long-term maintenance.
A common phenomenon, observed in several species in different laborato-
ries, is a modification of the tissues in maintenance. The tissues take on an opaque
wet appearance and become more difficult to maintain than when fluffy and
translucent. In our laboratory, the wet type of culture produces little or no
mature embryos. Similar observations were reported on black spruce and white
spruce by Attree et al. (l990b). We observed that this undesirable tissue morpho-
logy is often a consequence of handling. The size of the embryogenic tissue for the
Fig. 2. Embryogenic tissue of black spruce, 2 weeks after transfer to maintenance medium;
bar = 0.16 cm
subculture has a significant effect on long-term maintenance of healthy, that is,

fluffy and translucent embryogenic tissue, retaining its capacity to produce
somatic embryos. Transfer oflarge pieces of embryogenic tissue (> 50 mg) results
in a decreasing growth rate of the tissue, which rapidly becomes wet and opaque
white (F.M . Tremblay, unpubl.). When we get this type of culture, dishes are
incubated unsealed for up to 6 months; also, agar concentration can be routinely
increased to 1% without growth inhibition.
Various incubation conditions can be used during the maintenance of
embryogenic tissue. Although tissue can be maintained under darkness or light,
it was shown that light (16 h/day photoperiod) or continuous darkness influence
the type of somatic embryos obtained after maturation. Significantly higher
numbers of precociously germinating embryos are produced on embryogenic
tissue maintained in darkness than in light, even when the maturation stage was
under light (Tremblay and Tremblay 1991a).
2.3 Embryogenic Suspension Cultures
Embryogenic tissue has been used to establish cell suspension cultures of black
spruce (Tau torus et al. 1990), red spruce (Tremblay 1990b), and other coniferous
species (reviewed by Tautorus et al. 1991). Although the maintenance of
embryogenic tissue can be realized in liquid medium, no embryo maturation has
been obtained yet in liquid culture. Up to now, embryogenic tissue has been
easily maintained in liquid culture, but solid medium is necessary for embryo
maturation.
35 120 6
:;
.s
30
25
20 {
100
80
,,
.'
t
....
...1· . ·. · ~- t -~l- . . . .
-- - A'~~t
.... \.
",
........
'1:
5
0'"
60 /i-~~ ... ,'.z \
3 ~
>
l>
Il.
15
.s 40 ' f' .•1' . 2 -
co
10 Il. .,,,' ......... .1 ...... -1 ......
1·····-
0
5 20
OP FW --~-- PCV
0 0
~ 200 •. -1 •••• -1 .... - .... , ................ _
~ 100 ....................... .
O~r_,_-.--~~--.__r~r_,_-.--~_r_
2 3 4 5 6 7 8 9 10 11 12
Days
Fig.3. Growth curves for embryogenic Picea rubens cell suspension (line RS61.1) on HLM·I medium,
showing the PCV, FW, DW, and OP of the medium over 12 days. Vertical bars = SD. (Tremblay
1990b)
Growth curves were determined using biomass parameters (PCV, FW, and
DW) and osmotic pressure (OP) of the medium over 12- to 14-day periods
(Tremblay 1990b). Embryogenic suspension cultures were established with
embryogenic lines of white spruce (W-9) and red spruce (RS61.1), and compared
to nonembryogenic white spruce (PG-434), Alnus incana (European grey alder),
and Betula papyrifera (paper birch) cell suspensions. For RS61.1, the linear
growth phase was obtained after 4 days in culture and the doubling time
evaluated as 72 h between days 4 and 7 (Fig. 3). The W-9 cell suspension reached
a linear growth phase on day 3, with a doubling time of24 h between days 3 and
4 (data not shown). Becwar et al. (1988) reported a 48-h doubling time for a Picea
30 250
25 200
f
20 8
:; 150
6~
i
.§. 15
>
l>
100
Il. 10
5
& 50 2
0 0 o
1·......·-l---------§--------.
f : . . __. . . . . . _.....-.. .
500
o 2 4 6 8 10 12 14
Days
Fig. 4. Growth curves for nonembryogenic Picea glauca cell suspension (line PG-434) on Litvay's
medium, showing the PCV, FW, DW, and OP of the medium over 14 days. Vertical bars = SD.
(Tremblay 1990b)
abies suspension, with a lag phase of approximately 5 days in culture. In our case,
no typical lag phase was observed for both species, as later on confirmed on
interior spruce and black spruce (Lulsdorf et al. 1992). The nonembryogenic
white spruce suspension culture, PG-434, showed a lag phase during the first
days in culture and a stationary growth phase after 10 days (Fig. 4). For similar
PCV values, FW and DW were double for the nonembryogenic PG-434 cell
suspension compared to the embryogenic cell suspensions.
PCV, FW, and DW were compared to a technique based on the measure-
ment of OP of the medium. Correlations between OP of the medium and the
biomass parameters were obtained with embryogenic and non embryogenic
suspension cultures. For all the suspension cultures tested, Pearson's correlation
test showed significant direct correlations (a:s 0.0001) among the parameters
measuring biomass (DW, FW, and PCV) and significant inverse correlations
(a =0.0001) between the OP of the medium and the biomass parameters. The
OP measurement was shown to be as reliable as the biomass parameters to
monitor the growth of suspension cultures (Tremblay 1990b). Similar
correlations between OP and the biomass parameters were obtained for interior
spruce and black spruce (Lulsdorf et al. 1992).
2.4 Maturation
During the maturation stage, abscisic acid (ABA) was shown to promote
accumulation of storage proteins and lipids in conifer somatic embryos (Feirer
et al. 1989; Hakman et al. 1990; Roberts et al. 1990a). Addition of8-16 /lM ABA
Fig. 5. Black spruce somatic embryos after 4 weeks of maturation; bar = 0.2 cm
to the maturation medium was beneficial for black spruce embryo maturation
(Attree et al. 1990b), as previously reported for Norway spruce (von Arnold and
Hakman 1988) and Pinus strobus (white pine) (Finer et al. 1989). Higher ABA
concentrations of 20--60 IlM were found to promote normal development of
embryos into plants for interior spruce (Roberts et al. 1990a; Webster et al. 1990),
white spruce (Dunstan et al. 1991), and red spruce (Harry and Thorpe 1991). In
our laboratory, normal black spruce and red spruce plants have been regularly
produced on media containing 7.51lM ABA (Tremblay and Tremblay 1991a) to
45 IlM ABA, in the presence of 6% sucrose (Fig. 5). In fact, optimal ABA
concentration varies among black spruce cell lines and, although 25 IlM could
constitute a good compromise for most cell lines, lower or higher ABA concen-
trations can be used without subsequent abnormal development of the plants
(unpubl.).
Modification of basic components of the culture medium was shown to
improve embryo maturation for black spruce and red spruce. The best embryo
maturation of both species was obtained with a maturation medium containing
3.4-10 IlM ammonium nitrate (Tremblay and Tremblay 1991a). In the same
report, the gelling agent type and concentration were shown to strongly influence
embryo maturation, 0.4% of Gelrite gellan gum increasing the number of
somatic embryos compared to Difco Bacto-agar. In a detailed study of the
carbohydrate requirements during maturation of both species, it was shown that
175 mM (corresponding to 6% sucrose) of either sucrose, fructose, glucose,
maltose, or cellobiose supported embryo development (Tremblay and Tremblay
1991b). For black spruce, sucrose gave a significantly higher number of mature
embryos than the monosaccharides and dissaccharides tested. However, an
important difference between autoc1aved and filter-sterilized sucrose was ob-
tained, the latter producing twice as many embryos for black spruce. For red
spruce, no significant difference was observed among the carbohydrates tested
but it was interesting to note that the highest number of embryos was obtained
with cellobiose and maltose in the maturation medium. It was also shown that
6% sucrose could be replaced by either 6% fructose, 6% glucose, or 3% sucrose
plus 3% osmoticum for black spruce embryo development but that the embryos
produced on the different treatments were not all similar in their later develop-
ment into plantlets. For red spruce, 6% fructose gave a significantly higher
number of mature embryos than 6% glucose or 6% sucrose (Tremblay and
Tremblay 1991b).
Recently, the carbohydrate in the maturation medium was more thoroughly
studied through a sugar analysis, in parallel with a follow-up of the OP of the
medium (Tremblay and Tremblay, submitted). These data underlined the impor-
tance of an increment of the OP of the medium for the maturation process of
black spruce and showed that sucrose in the medium, in this case 6% sucrose
initially, is totally hydrolyzed into fructose and glucose after 2 weeks in the
presence of embryogenic tissue (Fig. 6).
F or embryo maturation, a beneficial effect of a pretreatment with activated
charcoal, 7 days before the maturation, was reported for red spruce, Norway
spruce, and interior spruce (Becwar et al. 1989; Roberts et al 1990a; Harry
and Thorpe 1991). For black spruce, pretreatments with activated charcoal
A~
J~_l~L
I '
I I
5 10
Retention time (min)
Fig. 6. Chromatograph illustrating carbohydrate peaks obtained in a maturation medium prepared

with 6% w/v sucrose A before inoculation with embryogenic tissue, B after 2 weeks of incubation with
black spruce embryogenic tissue, and C after 2 weeks of incubation without embryogenic tissue. The
peaks were 1 phenyl-j3-D-glucoside; 2 sucrose; 3 and 4 fructose; 5 and 6 glucose
at 0.01, 0.1, and 1 % w/v were tested, but they did not improve embryo
maturation (unpubl.).
2.5 Germination
Following ABA treatment, mature embryos are transferred to phytohormone-

free medium for germination. Germination for red spruce and interior spruce
somatic embryos is improved following partial drying at high relative humidity
(HRH) (Roberts et al. 1990b; Harry and Thorpe 1991). This technique was tested
for black spruce somatic embryos from several cell lines compared to control
somatic embryos germinated directly onto Sorbarod plugs (Baumgartner
Papers, 30X 16-B I-PVA l-CAF2). The conversion rate of somatic embryos into
plantlets was 23.3% for the controls transferred directly onto Sorbarod, while
2.8% germination was obtained following HRH treatment. In fact, somatic
embryos immersed in sterile water for 1 week under the same conditions as the
HRH treatment showed a better germination frequency than embryos partially
dehydrated (F. M. Tremblay, unpubl.). Cold treatment at 5 °C for 1 to 3 weeks
before germination was shown to improve germination of red spruce somatic
embryos (Harry and Thorpe 1991).
440 L. Tremblay and F .M. Tremblay
The conditions used for the development of somatic embryos into plantlets
playa predominant role in the conversion rate. For example, use of 114 SH
(Schenk and Hildebrandt 1972) medium was found to be extremely limiting on
somatic embryo germination compared to a 1/2CD-based (Campbell and
Durzan 1975) medium (Tremblay and Tremblay 1991b). On the other hand,
1I2CD medium solidified with Difco Bacto-agar gave 16% germination com-
pared to the 40% obtained using liquid 1l2CD-imbibing Sorbarod plugs
(Tremblay and Tremblay 1991 b), confirming Bercetche's (1988) results on Nor-
way spruce. Furthermore, we found that the Sorbarod plugs have to be prevented
from sinking tathe bottom of the test tube, in order to prevent saturation of the
plug with the medium (Fig. 7). It is probable that a higher oxygen level around
the roots accounts for the increased germination rate. Using this method, most
somatic embryos developed further when transferred to germination conditions
and the time necessary to obtain a transplantable plantlet was shortened from
4-5 months (Attree et al. 1990b) to 2- 3 months (unpubl.).
Fig. 7. Somatic embryo-derived plantlets of black

spruce after 2 months of germination of Sorbarod
plugs (P) supported half-way in the test tube,
here by disposable pipette tips. Using this
method, plantlets exhibit good epicotyl and
roo! development and most survive transfer to soil;
bar = I.Scm
Fig. 8. Somatic embryo-derived black spruce plantlets, clone MA-3, growing under greenhouse
conditions, 5 months after transfer to soil; bar = 5 cm
2.6 Acclimatization
More than 1500 black spruce plantlets, well established in soil (Fig. 8), have been
recovered from 26 genotypes (unpubl.). For all the black spruce genotypes, as
well as for four white spruce genotypes, survival after soil transfer is presently
higher than 95% (unpubl.). Previous work reported a survival frequency of 60%
(160/265) for black spruce and 18% (29/160) for white spruce (Attree et al.
1990b). Following transfer to soil, plantlets exhibit continuous and rapid devel-
opment under greenhouse conditions, with the exception of plants recovered
from long-term maintained embryogenic tissue (see Sect. 2.2).
2.7 Protoplast Culture
For conifers, somatic embryogenesis has made genetic transformation possible

because of the ability of protoplasts isolated from embryogenic tissue to regener-
ate into plants. Protoplasts isolated from embryogenic suspension culture of
black spruce and red spruce were obtained and regenerated into somatic embryos
(Tautorus et al. 1990; Tremblay 1990b). Black spruce protoplasts obtained from
suspension cultures were electroporated and showed transient gene expression
(Tautorus et al. 1989; Bekkaoui et al. 1990) Recently, embryogenic tissue
exposed to particle acceleration and transiently expressing introduced genes have
been reported for white spruce (Ellis et al. 1991) and black spruce (Duchesne and
Charest 1991).
2.8 Somaclonal Variation
Somatic embryogenesis can supply clones for breeding and reforestation pro-
grams. However, the development of testing methods have to be elaborated to
certify the genetic integrity of the somatic embryo-derived plantlets. No
somaclonal variation was observed during conifer somatic embryogenesis (Mo
et al. 1989; Eastman et al. 1991), while a variation in chromosome number was
observed for two immature somatic embryos of Norway spruce (Lelu 1987). In
our laboratory, the usefulness of RAPD markers (randomly amplified polymor-
phic DNA) was evaluated to assess the genetic stability of in vitro-derived clonal
material of black spruce. Twenty-five embryogenic cell lines and additional
zygotic embryos and their megagametophytes obtained from three controlled
crosses were used for segregation analysis ofRAPD variants. Using ten geneti-
cally characterized markers and three cell lines, no genetic variation was found
among somatic embryos for each cell line (Isabel et al. 1993).
Although the genetic integrity of the plants produced by somatic embryo-
genesis has to be verified, it is essential to start looking for epigenetic effects of
somatic embryogenesis. In any system involving tissue culture techniques, a
maximum length of time, or number of subcultures, has to be defined in terms of
epigenetic effects on the in vitro-derived plants. With conifers, it is urgent to start
thinking in terms of possible epigenetic variation, particularly when considering
their long life cycle. For example, in other plant species, flowering is a process
which is often affected by in vitro techniques. With somatic embryo-derived trees,
could the flower production start earlier or later? These considerations might
have significant consequences if embryo-derived clones are used to establish seed
orchards or plantations for fiber production.
In the last years, the conversion rate of black spruce and red spruce somatic
embryos into plantlets has been greatly improved through the development of
better maturation and germination conditions, applicable to a wide range of
genotypes.
At present, somatic embryogenesis can be induced only from juvenile tissues,
which limits the potential of the technique. It would be desirable to obtain
material from more mature tissue to be able to test as early as possible the genetic
potential of the tree. Despite this limitation, somatic embryogenesis can produce
large quantities of plantlets derived from full-sib seeds, leading to improved
productivity of spruce plantations. Also, somatic embryogenesis provides a
regeneration system that permits gene transfer studies.
4 Protocol
Black spruce and red spruce zygotic embryos are induced and maintained on HLM-l, according to
Tremblay (1990a). Portions of embryogenic tissue are transferred every 2 weeks onto fresh medium,
incubated at 22°C, and a 16-h photoperiod given by Gro-Lux WS (Sylvania) fluorescent lamps (5-10
~mol m-'s-I). For embryo development, 75-100 mg portions of embryogenic tissues are taken, 7 days
after subculture on maintenance medium, and placed on a basal maturation medium. The basal
maturation medium consists in half-Litvay's salts (Litvay et al. 1985) with 1 gil glutamine, 1 gil casein
hydrolysate, 6% sucrose, and 7.5-30 ~M abscisic acid (depending on the cell lines). The medium is
solidified with 0.4% Gelrite gellan gum and the pH adjusted to 5.7 before autoc1aving at 121°C.
Sucrose, abscisic acid, and glutamine are filter-sterilized and added to the warm medium. Embryogenic
tissues are incubated at 22°C with a 16-h photoperiod given by Vita-Lite (Duro-Test) fluorescent
lamps (10-15 ~mol m-'s-I). For plantlet development, liquid 1I2CD medium containing 1.5% sucrose
saturates 30 x 16 mm Sorbarod plugs at the rate of 5 ml medium per plug. The plug is prevented from
sinking by the use of pipette tips. After 2-3 months, plantlets with a I-cm epicotyl are transferred to soil
under mist conditions with a gradual decrease in relative humidity from 95 to 70% over 2 weeks, at
which time plantlets are transferred to regular greenhouse conditions.
Acknowledgments. This research was supported by the Ministere des Forets, Quebec (Grant No.
032540 to F .M.T.) and made possible by the Interchange Canada Programme through the assignment
of F.M.T. to Laval University.
References
Attree SM, Fowke LC (1991) Micropropagation through somatic embryogenesis in conifers. In: Bajaj
YPS (ed) Biotechnology in agriculture and forestry, vol 17. High-tech and micropropagation I.
Attree SM, Budimir S, Fowke LC (1990a) Somatic embryogenesis and plantlet regeneration from
cultured shoots and cotyledons of seedlings from stored seeds of black spruce and white spruce
(Picea mariana and Picea glauca). Can J Bot 68: 30-34
Attree SM, Tautorus TE, Dunstan DI, Fowke LC (I 990b) Somatic embryo maturation, germination,
and soil establishment of plants of black spruce and white spruce (Picea mariana and Picea glauca).
Can J Bot 68: 2583-2589
Becwar MR, Wann SR, Johnson MA, Verhagen SA, Feirer RP, Nagmani R (1988) Development and
characterization of in vitro embryogenic system in conifers. In: Ahuja MR (eds) Somatic cell
genetics of woody plants. Kluwer, Dordrecht, pp 1-18
Becwar MR, Noland TL, Wyckoff JL (1989) Maturation, germination, and conversion of Norway
spruce (Picea abies L.) somatic embryos to plants. In Vitro Cell Dev Bioi 25: 575-580
Bekkaoui F, Dalta RSS, Pilon M, Tautorus TE, Crosby WL, Dunstan DI (1990) The effects of
promoter on transient expression in conifer cell lines. Theor Appl Genet 79: 353-359
Bercetche J (1988) Optimisation des conditions d' obtention de plantules it partir de cals embryogenes
chez Picea abies. AFOCEL, Direction des Recherches, Paris, pp 97-115
Campbell RA, Durzan DJ (1975) Induction of multiple buds and needles in tissue cultures of Picea
glauca. Can J Bot 53: 1652-1657
Davidson JG (1990) La necessite est la mere de I'invention. InfoFon!ts 9:2
Duchesne LC, Charest PJ (1991) Transient expression of the p-glucuronidase gene in embryogenic
callus of Picea mariana following microprojection. Plant Cell Rep \0: 191-194
Dunstan DI, Bethune TD, Abrams SR (1991) Racemic abscisic acid and abscisyl alcohol promote
maturation of white spruce (Picea glauca) somatic embryos. Plant Sci 76: 219-228
Eastmen PAK, Webster FB, Pitel JA, Roberts DR (1991) Evaluation ofsomac1onal variation during
somatic embryogenesis of interior spruce (Picea glauca-engelmannii complex) using culture mor-
phology and isozyme analysis. Plant Cell Rep 10: 425-430
Ellis DD, McCabe D, Russell D, Martinell B, McCown BH (1991) Expression ofinducible angiosperm
promoters in a gymnosperm, Picea glauca (white spruce). Plant Mol Bioi 17: 19-27
Feirer RP, Conkey JH, Verhagen SA (1989) Triglycerides in embryogenic conifer calli: a comparison
with zygotic embryos. Plant Cell Rep 8: 207-209
Finer JJ, Kriebel HB, Becwar MR (1989) Initiation of embryogenic callus and suspension cultures of
eastern white pine (Pinus strobus L.). Plant Cell Rep 8: 203-206
Fowells HA (1965) Silvics offorest trees of the United States No 2 Agric Handb USDA, For Serv, US
Department of Agriculture, Washington, DC 20250
Hakman I, Fowke LC (1987) Somatic embryogenesis in Picea glauca (white spruce) and Picea mariana
(black spruce). Can J Bot 65: 656-659
Hakman I, F owke LC, von Arnold S, Eriksson T (1985) The development of somatic embryos in tissue
cultures initiated from immature embryos of Picea abies (Norway spruce). Plant Sci 38: 53--60
Hakman I, Stabel P, Engstrom P, Eriksson T (1990) Storage protein accumulation during zygotic and
somatic embryo development in Picea abies (Norway spruce). Physiol Plant 80: 441-445
Harry IS, Thorpe TA (1991) Somatic embryogenesis and plant regeneration from mature zygotic
embryos of red spruce. Bot Gaz 152: 446-452
Hatcher RJ (1963) A study of black spruce forests in northern Quebec. Department of Forestry Pub!.
no 1018, Forest Research Branch, Quebec, Canada
Hosie RC (1980) Native trees of Canada. 8th Can For Serv, Dep Environ Editions Fides, Canada,
380pp
Isabel I, Tremblay L, Michaud M, Tremblay FM, Bousquet J (1993) RAPDs as an aid to evaluate the
genetic integrity of somatic embryogenesis-derived populations of Picea mariana (Mil!.) B.S.P.
Theor Appl Gent 86: 81-87
J alonen P, von Arnold S (1991) Characterization of embryogenic cell lines of Picea abies in relation to
their competence for maturation. Plant Cell Rep 10: 384--387
Krogstrup P (1986) Embryolike structures from cotyledons and ripe embryos ofN orway spruce (Picea
abies). Can J For Res 16: 664--668
Laine E, David A (1990) Somatic embryogenesis in immature embryos and protoplasts of Pinus
caribaea. Plant Sci 69: 215-224
Lelu M-A (1987) Etude preliminaire des variations morphologiques et genetiques obtenues apres
embryogenese somatique chez Picea abies. In: AFOCEL Annales de Recherches Sylvicoles, Direc-
tion des Recherches, Paris, pp 35-47
Lelu M-A, Bornman CH (1990) Induction of somatic embryogenesis in excised cotyledons of Picea
glauca and Picea mariana. Plant Physiol Biochem 28: 785-791
Lelu M-A, Boulay M, Arnaud Y (1987) Obtention de cals embryogenes Ii partir de cotyledons de Picea
abies (L.) Karst. Preleves sur de jeunes plantes agees de 3 Ii 7 jours apres germination. C R Acad Sci
Paris 305: 105-109
Lelu M-A, Boulay MP, Bornman CH (1990) Somatic embryogenesis in cotyledons of Picea abies is
enhanced by an adventitious bud-inducing treatment. New For 4: 125-135
Litvay JD, Verma DC, Johnson MA (1985) Influence ofa loblolly pine (Pinus taeda L.) culture medium
and its components on growth and somatic embryogenesis of the wild carrot (Daucus carota L.).
Plant Cell Rep 4: 325-328
Lulsdorf MM, Tautorus TE, Kikcio SI, Dunstan DI (1992) Growth parameters of embryogenic
suspension cultures of interior spruce (Picea glauca-engelmannii complex) and black spruce (Picea
mariana Mill). Plant Sci 82: 227-234
Mo LH, von Arnold S (1991) Origin and development of embryogenic cultures from seedlings of
Norway spruce (Picea abies). J Plant Physiol138: 223-230
Mo LH, von Arnold S, Lagercrantz U (1989) Morphogenic and genetic stability in long-term
embryogenic cultures and somatic embryos on Norway spruce Picea abies L. Karst. Plant Cell Rep
8: 375-378
Roberts DR, Flinn BS, Webb DT, Webster FB, Sutton BCS (I 990a) Abscisic acid andindole-3-butyric
acid regulation of maturation and accumulation of storage proteins in somatic embryos of interior
spruce. Physiol Plant 78: 355-360
Roberts DR, Sutton BCS, Flinn BS (1990b) Synchronous and high frequency germination of interior
spruce somatic embryos following partial drying at high relative humidity. Can J Bot 68:
1086-1090
Schenk RU, Hildebrandt AC (1972) Medium and techniques for induction and growth ofmonocoty-
ledonous and dicotyledonous plant cell cultures. Can J Bot 50: 199-204
Tautorus TE, Bekkaoui F, Pilon M, Dalta RSS, Crosby WL, Fowke LC, Dunstan DI (1989) Factors
affecting transient gene expression in electroporated black spruce (Picea mariana) and jack pine
(Pinus banksiana) protoplasts. Theor Appl Genet 78: 531-536
Tautorus TE, Attree SM, F owke LC, Dunstan Dr (1990) Somatic embryogenesis from immature and
mature zygotic embryos, and embryo regeneration from protoplasts in black spruce (Picea mariana
Mill.). Plant Sci 67: 115-124
Tautorus TE, Fowke LC, Dunstan DI (1991) Somatic embryogenesis in conifers. Can J Bot 69:
1873-1899
Tremblay FM (1990a) Somatic embryogenesis and plantlet regeneration from embryos isolated from
stored seeds of Picea glauca. Can J Bot 68: 236-242
Tremblay L (1990b) Etudes sur I'embryogenese somatique chez Picea spp. MSc, Dept Forestry, Laval
University, Quebec, Canada
Tremblay L, Tremblay FM (199Ia) Effects of gelling agents, ammonium nitrate, and light on the
development of Picea mariana (Mill.) B.S.P. (black spruce) and P. rubens Sarg. (red spruce) somatic
embryos. Plant Sci 77: 233-242
Tremblay L, Tremblay FM (199Ib) Carbohydrate requirements for the maturation of black spruce
[Picea mariana (Mill.) B.S.P.j and red spruce (P. rubens Sarg.) somatic embryos. Plant Cell Tissue
Organ Cult 27: 95-103
Tremblay L, Tremblay FM (submitted) Maturation of black spruce somatic embryos: sucrose
hydrolysis and resulting osmotic pressure of the medium. Plant Cell Tissue Organ Cult
von Arnold S, Eriksson T (1981) In vitro studies of adventitious shoot formation in Pinus contorta.
Can J Bot 59: 870-874
von Arnold S, Hakman I (1988) Regulation of somatic embryo development in Picea abies by abscisic
Webster FB, Roberts DR, McInnis SM, Sutton BCS (1990) Propagation of interior spruce by somatic
embryogenesis. Can J For Res 20: 1759-1765
Wright JW (1955) Species crossability in spruce in relation to distribution and taxonomy. For Sci I:
319-348
111.15 Somatic Embryogenesis in Poplars
Y.G. PARK l and S.H. SON2
1 Introduction
1.1 General Description of Poplars

Poplars (genus Populus, family Salicaceae) consist of more than 30 species widely
distributed in temperate and cold regions of the northern hemisphere (FAO
1980). The varied uses of Populus include pulpwood, plywood, and match
production. Due to the rapid growth rate, this species has gained importance as
an energy plantation (Hall et al. 1989), this makes Populus a valuable tree species
for further breeding. Some cultivars have been selected and tested. Among the
poplars, the hybrid, Populus nigra L. x P. maximowiczii Henry, is rarely
described in spite of its remarkable traits. The male Asian balsam poplar, Populus
maximowiczii has its natural habitats in Korea which is frequently associated
with both broad-leaved species and conifers. The female black poplar, Populus
nigra, distributed in Europe, Asia, and North Africa, was introduced to Korea in
1958. Morphologically, Populus nigra has slender and round shoots with small
leaves. The female tree typically shows short catkins with round, crowned
capsules. These species were crossed by Hyun and Hong (1959) to create a new
clone, Rochester (P. maximowiczii x P. nigra plantierensis). From the provenance
test, the hybrid poplar (Populus nigra x P. maximowiczii) is characterized by its
fast growth rate, high resistance potential to cold, disease, and acidic soil
conditions as well as its suitability for a wide variety for forest lands (Noh et al.
1984).
1.2 Macro- and Micropropagation of Populus Species
Various species of Populus have been macropropagated from stem cutting, single
node cutting, stump sprout, grafting, and root sprout (Snow 1938; Hall et al.
I Department of Forestry, College of Agriculture, Kyungpook National University, Daegu 702~701,
Republic of Korea
2 Laboratory of Biotechnology, Forest Genetics Research Institute, Forestry Administration, PO Box
24, Suwon, Kyonggido 440-350, Republic of Korea

Somatic Embryogenesis in Poplars (Populus nigra L. x P. maximowiczii Henry) 447
1989). These conventional methods provide advantages in establishing a clonal

bank of elite trees for reforestation. However, genotypic variation, limited
availability of elite genotypes, cost-increasing factors such as vast area require-
ments and intensive labor, often observed self-incompatibility of grafting, and
the aging of stock plants make it difficult to develop a practical system of large-
scale macro propagation to meet the demand for reforestation which increases
year after year (Ahuja 1987).
Current advances in plant cell and tissue culture technology emphasize the
use of in vitro approaches as part of tree breeding and commercial-scale micro-
propagation (Son 1991). Mass propagation via a tissue culture system has been
extensively studied in Populus species (Ahuja 1987; Sellmer et al. 1989; Son
and Hall 1990a,b). To obtain rapid true-to-type micropropagules, the axillary
branching method has been most commonly used (Whitehead and Giles 1977;
Chun et al. 1986; Son et al. 1991). Even though a large-scale propagation system
was successfully demonstrated in some genotypes for commercial purposes there
are still limitations, such as lack of reliable systems for increasing the rate of
multiplication, for simplifying procedures, and for easy handling. To overcome
these obstacles, somatic embryogenesis has been suggested as a potential tool for
commercial-scale micropropagation and molecular biological application. Vari-
ous aspects of micro propagation and genetic transformation in poplars have
been reviewed (Douglas 1989; Seller and McCown 1989; Lubrano 1992).
Leaves have been used as explants for the direct and indirect induction of somatic
embryogenesis. In poplars, in vitro grown leaf explants were frequently used for
studies on morphogenesis. For example, Cheema (1989), Park and Son (1988a,
1992), and Michler and Bauer (1991) obtained direct and/or indirect somatic
embryogenesis from cell suspension and callus cultures derived from leaf ex-
plants of Himalayan poplar, hybrid poplar, and hybrid aspen. Direct and
indirect somatic embryogenesis using leaf explants of the hybrid poplar (Populus
nigra L. x P. maximowiczii Henry), which has a great regeneration capacity
among the poplars, is described here.
2.1 Establishment of Aseptic Shoot Culture
Stem nodes (3-4 cm) containing an axillary bud were collected from the upper
parts of actively growing branches of 20-year-old Populus nigra x P. maximo-
wiczii at the Forest Genetics Research Institute, Suwon, Korea. For surface-
sterilization, leaves were excised from the stem node explants and immediately
washed thoroughly in running tap water for 30 min. They were soaked in 70%
ethanol for min, subsequently sterilized with 2% sodium hypochlorite containing
one drop of Tween 80 for 10 min, and following each step, rinsed at least five
448 Y.G. Park and S.H. Son
Fig. 1. Mature leaf used for culture
times with sterile distilled water. The disinfested stem node explants were individ-
ually transferred to test tubes (2.4 x 15 cm) containing lO ml of MS medium
(Murashige and Skoog 1962) without plant growth regulators (PGR). When the
axillary buds sprouted, they were isolated from the stem node explants and
placed on the same fresh medium for 4 to 6 weeks. To obtain a sufficient number
of shoots, the shoot apex from each shoot was removed and then lO bi- or
trinodal shoots were cultured in Magenta GA-7 vessels (7.6 x 7.6 x 10.2 cm;
Magenta Co., Chicago, IL) containing 50 ml of the medium supplemented with
0.88 11M 6-benzylaminopurine (BAP). After five subcultures with a 4-week
interval on the same medium mentioned above, shoot cultures excised from the
multiplied axillary branches were subcultured on the PGR-free MS media for
more than 6 weeks. Plants of8 to lO cm in height with fully expanded leaves were
used as in vitro source materials.
2.2 Direct Embryogenesis from Punctured Leaf Culture
Fully expanded leaves (Fig. 1) of approximately 16- 21 mm diameter were

prepared by puncturing 40- 50 times with a pin (for a detailed description, see
Park and Son 1988a). The leafexplants were placed with the abaxial side (adaxial
contact) on MS medium supplemented with 10 mIll coconut milk, lO mill malt
extract, 30 gil sucrose, 7.5 gil Difco Bacto-agar, and various plant growth
regulators. The pH of the media was adjusted to 5.8 prior to autoclaving at
121°C for 15 min. Cultures were maintained at 26 ± 1 °C under 20-40 IlE m- 2s- 1
irradiation by fluorescent light for a 16-h photoperiod. Two explants were used
per treatment, which were repeated five times in three separate experiments.
Cultures were observed periodically under a stereo microscope. Somatic embry-
Table 1. Effect of different plant growth regulators on somatic embryo formation from punctured leaf
of Populus nigra L.x P. maximowiczii Henry. (Park and Son 1988a)
Plant growth regulators (11M) Shape of somatic embryos'
Globular Heart Torpedo
BAP 0.44 + NAA 2.69 + +

BAP 0.44 + 2,4-D 2.26 ++ +++ ++++
BAP 0.88 + NAA 0.54 + ++ +
BAP 0.88 + 2,4-D 2.26 + ++ +
• No response; +, 2-3 embryosJIeaf; ++, 3-5 embryos/leaf; +++,5-10 embryoslleaf; ++++, more than
10 embryoslleaf. Plant growth regulators were selected from a total of 70 combinations.
Table 2. Effect of BAP plus 2,4-D on embryogenic callus induction in punctured

leaves of Populus nigra L. x P. maximowiczii Henry. (Park and Son 1988a)
Plant growth regulators (11M) Mean fresh weight of embryogenic callus

SE (mg)a.b
BAP 2,4-D
0.00 0.00
0.05 15.3 7.3
0.23 1136.8 38.3
0.45 2405.6 176.5
2.26 1299.4 37.8
4.54 1166.4 408.5
0.44 0.00
0.05
0.23 1185.4 53.3
0.45 2326.4 207.7
2.26 3107.2 48.9
4.52 2529.3 693.2
0.88 0.00
0.05
0.23
0.45 770.2 28.9
2.26 474.0 394.9
4.52 1175.4 617.6
1.78 0.00
0.05
0.23
0.45 545.8 28.9
2.26 830.2 242.9
4.52 602.0 573.8
3.55 0.00
0.05
0.23 113.6 3.7
0.45 308.2 7.8
2.26 267.8 85.4
4.52 254.0 155.6
• Each value represents the mean ± SE of 5 replications (after 6 weeks in culture).

b_, No response.
oids, 0.5 to 1 mm in diameter at the globular to torpedo-shaped stage, occurred

sporadically on the surface of punctured leaf explants. Adventitious shoots also
developed simultaneously with the embryoids. The highest frequency of induced
somatic embryos occurred when the PGR combination was 0.44 IlM BAP and
2.261lM 2,4-dichlorophenoxyacetic acid (2,4-D) (Table 1).
2.3 Initiation of Cell Suspension Culture
Primary callus was induced from punctured leaves of the fully expanded shoot
cultures by the transfer to MS medium supplemented with 0.44 IlM BAP and
4.521lM 2,4-D. Cell suspension cultures were established by inoculating approxi-
mately 1 g (fresh weight) of embryogenic callus into 100 ml Erlenmeyer flasks
containing 20 ml of MS liquid medium with 0.441lM BAP and 4.521lM 2,4-D.
The morphology of embryogenic callus showed an uneven surface and a pale
greenish color (Table 2). Cultures were agitated at 121 rpm/min on a gyratory
shaker, and maintained in a culture room at 26 ± 1 °C in complete darkness.
Suspension cultures were subcultured at 2-week intervals.
2.4 Somatic Embryogenesis from Induced Embryogenic Cells
Two ml of agar (0.75% w/v) media (45°C) with no PGR was quickly poured into
60 x 15 mm sterile disposable plastic Petri dishes. For the induction of somatic
embryos, 1 rnl aliquots of cell suspension culture (l x 104 to 195 cells) were washed
with PGR-free MS liquid media and then added to the plastic Petri dishes prior
to solidification of the media. After 6 weeks, an early stage of somatic embryos
started to appear when the cultures were exposed to light. Ingredients such as
coconut milk and malt extract were observed to be the most effective additives for
somatic embryogenesis in this species when combined (Table 3).
Table 3. Effect of some ingredients on somatic embryogenesis from suspension cultured

cells of Populus nigra L. x P. maximowiczii Henry. (Y.G. Park and S.H. Son, unpubl.)
Type of ingredient Mean number of somatic embryos

(m1l1) per plate •. b
Without CM and ME 6 4
Coconut milk 10 26 6
(CM) 30 18 8
50 19 13
Malt extract 10 12 5
(ME) 30 7 2
50 9 3
CMIO+ME 10 67 23
a Somatic embryo counts represent mean ± SE of 3 replicated plates (after 8 weeks in
culture).
b All media were PGR-free MS with an initial cell density of 104 to lOS per plate.
Fig.2. a Globular and heart-shaped embryos (bar =0.2 mm); b elongation of somatic embryo (bar =
0.19 mm); c root initiation from embryos on half-strength MS medium lacking plant growth regulator.
(Park and Son 1988a)
2.5 Maturation of Somatic Embryos
Recent work has pointed out many factors affecting maturation of early stage
embryos, i.e., various types ofPGRs, evolved gases, concentration of carbohy-
drate, and reduced levels of nitrogen. Since the effect of these treatments usually
specifically responded to certain types of species, it might be necessary to test
all possible ways to determine the best conditions for given explants. The timing
of application of some factors seems to be important for somatic embryo
maturation.
Maturation of the early stage poplar embryoids was achieved by isolating
periodically and reculturing the embryos onto fresh half-strength MS media
without PGR. Although an increased level of sucrose had a slightly better effect
on somatic embryo maturation (Fig. 2), it is not clear whether the key factors for
the maturation are due to the high level of carbohydrate or elevated osmolarity.
In vitro grown leaf explants gave relatively high numbers of embryos. The
morphogenetic responses in poplar leaf culture were affected by factors, such as
type and composition of the media, amino acid, sugar, and PGR. Michler and
Bauer (1991) reported the effect of glutamine which stimulated the number of
somatic embryos from the leaf disk cultures when added at 20 to 40 /lM. In our
work, the critical factors affecting organogenesis and somatic embryogenesis are
ingredients such as coconut milk, malt extract, and PGR, especially the level of
2,4-D. Best results of direct embryogenesis from punctured leaf explants were
obtained using MS media supplemented with 0.44/lM BAP and 2.26/lM 2,4-D.
In the present study, most tested leaves produced simultaneously three different
types of morphogenesis, such as somatic embryogenesis, adventitious shoot bud
induction, and embryogenic and nonembryogenic callus initiation.
The growth responses of cell suspensions with BAP vs. 2,4-D and naphtha-
lene acetic acid (NAA) combinations were similar to previously reported results
(Park and Son 1988b). After 10 days of culture, the best yield of a fine suspension
of single cells and small cell aggregates was obtained on media containing 0.44
)lM BAP and 4.52 /lM 2,4-D. When the suspension cultures were exposed to
light, the cells eventually lost their capacity to undergo somatic embryogenesis.
Cell suspension cultures were, therefore, maintained in the complete darkness. In
addition, nodule-like bodies were also observed in cell suspension cultures. Most
of the somatic embryos derived from cell suspension cultures showed abnormal
morphology including small size, abnormally developed cotyledons with a long
root, a shortened length between cotyledon and radicle, and multiple clusters of
embryos with radicles (data not shown).
The timing for the isolation of somatic embryos and embryogenic calli
seems to be important for obtaining consistent somatic embryogenesis. The
protocols for direct andlor indirect somatic embryogenesis of Populus nigra x
P. maximowiczii are described in (Fig. 3).
ICollection of stem nodes I

~ Surface disinfection
I Culture of aseptically prepared stem node explants I
I - - Culture in test tube containing PGR-free
MS medium for 4 to 6 weeks
I - - Isolation and subculture of the bud
sprouts derived from stem node
explants
I Further growth of the newly propagated shoots I

~ Culture in PGR-free MS medium for 4 weeks
rShoot proliferation*l
I - - Excision of the shoot apex from the
elongated shoots and then inoculation of 10
bi- or trinodal shoots on Magenta Box
containing 50 ml of MS medium containing
0.88 11M BAP
I Single shoot elongation I
I - - Culture of multiplied shoots on
Magenta Box containing 50 ml of PGR-free MS
medium
IHarvest of healthy leaves I
r
r- Puncturing of the leaves with a sharp pin
I
IDirect somatic embryogenesis I ICallus induction I
f-- Culture of punctured leaves
I-Culture in plastic Petri dish
onto MS media containing O. 44
containing 20 ml of MS medium
containing 10 mill coconut milk, 11M BAP and 4.52 11M 2,4-0
10 mill malt extract, 30 gil
sucrose, 0.44 11M BAP, and 2.26 11M
IEstablishment of cell suspension I
culture*
2,4-0
f-- Inoculation of 1 g of embryo genic
callus into a conical
flask containing MS liquid
media supplemented with O. 44
11M BAP and 4.52 11M 2,4-0
ISomatic embryogenesis from I

embryogenic determined cells
- Plating of single cells and small
cell aggregates onto Petri
dishes containing PGR-free MS
IMaturation of somatic embryos I media with coconut milk
I T
and malt extract with a cell
density of 104 to 105 per mi.
Fig. 3. Schematic outlines of somatic embryogenesis from punctured leaf-derived callus and suspen-
sion cultured cells of hybrid poplar (Populus nigra L. x P. maximowiczii Henry). * denotes in vitro
stock culture
454 Y.G. Park and S.H. Son: Somatic Embryogenesis in Poplars
References
Ahuja MR (1987) In vitro propagation of poplar and aspen. In: Bonga JM, Durzan DJ (eds) Cell and
tissue culture in forestry, vol 3. Martinus Nijhoff, Dordrecht, pp 207-223
Cheema GS (1989) Somatic embryogenesis and plant regeneration from cell suspension and tissue
cultures of mature Himalayan poplar (Populus ciliata). Plant Cell Rep 8: 124-127'
Chun YW, Hall RB, Stephens LC (1986) Influences of medium consistency and shoot density on in
vitro shoot proliferation of Populus alba x P. grandidentata. Plant Cell Tissue Organ Cult 5:
179-185
Douglas GC (1989) Poplar (Populus spp.). In: Bajaj YPS (ed) Biotechnology in agriculture and
forestry, vol 5. Trees II. Springer, Berlin Heidelberg New York, pp 300-323
FAO (1980) Poplar and willows in wood production and land use, No 10, Rome
Hall RB, Colletti JP, Schultz RS, Faltonson RR, Kolison SH, Hanna RD, Hillson TD, Morrison JW
(1989) Commercial-scale vegetative propagation of aspen. Proc Aspen Symp, 25-27 July, Duluth,
MN, pp 211-219
Hyun SK, Hong SC (1959) Inter- and intraspecies hybridization in poplars. (I) List of poplar hybrids
produced by the Institute of Forest Genetics in Suwon. Res Rep Inst For Gen Korea I: 61-73
Lubrano L (1992) Micropropagation of poplars (Populus spp.). In: Bajaj YPS (ed) Biotechnology in
agriculture and forestry, vol 18. High-tech and micropropagation II. Springer, Berlin Heidelberg
Michler CH, Bauer EO (1991) High frequency somatic embryogenesis from leaf tissue of Populus spp.
Plant Sci 77: 111-118
Murashige T, Skoog F (1962) A revised medium for rapid growth and bio-assays with tobacco tissue
Noh ER, Ahn JK, Hyun SK (1984) Growth and adequate sites for the hybrid poplar Populus nigra x
P. maximowiczii F J clones in Korea. Res Rep Inst For Gen Korea 20: 46-51
Park YG, Son SH (1988a) In vitro organogenesis and somatic embryogenesis from punctured leaf of
Populus nigra x P. maximowiczii. Plant Cell Tissue Organ Cult 15: 95-195
Park YG, Son SH (1988b) Regeneration of plantlets from cell suspension culture derived callus of white
poplar (Populus alba L.). Plant Cell Rep 7: 567-570
Park YG, Son SH (1992) In vitro shoot regeneration from leaf mesophyll protoplasts of hybrid poplar
(Populus nigra x P. maximowiczil). Plant Cell Rep 11: 2-6
Sellmer JC, McCown BH (1989) Transformation in Populus spp. In: Bajaj YPS (ed) Biotechnology in
agriculture and forestry, vol 9. Plant protoplasts and genetic engineering II. Springer, Berlin
Sellmer JC, McCown BH, Haissing EE (1989) Shoot culture dynamics of six Populus clones. Tree
Physiol 5: 219-227
Snow J rAG (1938) Use of indolebutyric acid to stimulate the rooting of dormant aspen cuttings. J For
36: 582-587
Son SH (1991) In vitro culture system of hybrid aspen as tool for tree improvement programs and
commercial applications. PhD Dissertation, Iowa State University, Ames, Iowa
Son SH, Hall RB (1990a) Multiple shoot regeneration from root organ cultures of Populus alba x P.
grandidentata. Plant Cell Tissue Organ Cult 20: 53-57
Son SH, Hall RB (l990b) Plant regeneration capacity of callus derived from leaf, stem, and root
segment of hybrid poplar (Populus alba L. x P. grandidentata Michx.). Plant Cell Rep 9: 344-347
Son SH, Chun YW, Hall RB (1991) Cold storage of in vitro cultures of hybrid poplar shoots (Populus
alba L. x P. grandidentata Michx.). Plant Cell Tissue Organ Cult 27: 161-168
Whitehead HCM, Giles KL (1977) Rapid propagation of poplar by tissue culture methods. N Z J For
Sci 7: 40-43
111.16 Somatic Embryogenesis in Cacao
(Theobroma cacao)
V.c. PENCE l
1 Introduction
The seeds of the cacao tree, Theobroma cacao L. (Sterculiaceae), provide the basis
of the multibillion dollar chocolate industry. A native of the Neotropics, cacao
grows as an understory tree in the rainforest. Flowers and fruit are produced
directly from the stem, with pods attaining lengths of approximately 15-25 cm,
each containing 30-40 seeds (Pence 1989).
Cacao seed cotyledons contain the precursors of chocolate, but must un-
dergo fermentation, drying, and roasting to fully develop the cocoa or chocolate
flavor. They also contain up to 50% fat, which, when expressed from the cacao
seed, is known as cocoa butter. Cocoa and cocoa butter are used in various ways
to formulate the many types of chocolate available, while cocoa butter also has
use in the pharmaceutical industry. The major cacao producers are Cote d'I voire,
Brazil, Ghana, and Malaysia. Cacao is a tropical crop and must be cultivated
within 20° of the equator.
Because of the short-lived nature of cacao seeds, cacao germplasm must be
maintained as trees in plantations or "field gene banks". Although cacao is
maintained in a number of field gene banks worldwide, the genetic base of cacao
in cultivation is relatively narrow. Further collections from the wild could expand
the collection, while materials in hand need to be more fully characterized to
further define lines of interest for breeding. Somatic embryogenesis is one
technique which could help improve the efficiency of clonal propagation of
cultivated and wild lines of interest to breeders and which might also be used for
the transport and storage of cacao germplasm. Recent developments in somatic
embryo technology with cacao offer the potential for the application of these
methods to the improvement of this species.
I Center for Reproduction of Endangered Wildlife, Cincinnati Zoo and Botanical Garden, 3400 Vine
Street, Cincinnati, OH 45220, USA

456 V.C.Pence
2.1 Embryo Initiation
The first reports of somatic embryogenesis in cacao described the initiation

of somatic embryos from immature zygotic embryos (Esan 1977; Pence et al.
1979, 1980). This system has been used by several labs with some variations
(Table 1).
Immature zygotic cacao embryos are aseptically removed from the ovules of
surface-sterilized pods and placed in culture. Media for somatic embryo initi-
ation have been based on that of Murashige and Skoog (1962) (MS) and most
often contain casein hydrolysate. Initiation has been shown to be stimulated by
the presence of auxin, coconut milk, or peptone (Pence et al. 1980; Adu-
Ampomah et al. 1988). In some cases, White's (1963) organics (Wen 1989) and a
medium with cytokinin and no auxin (Dos Santos and Machado 1989) have been
Table 1. Summary of research on the initiation of somatic embryos of Theobroma cacao
References Explant Results
I. Esan (1977, 1992) Immature embryos Direct embryogenesis

2. Pence et al. (1979, 1980) Immature embryos Direct embryogenesis
3. Kononowicz et al. (1984) Somatic embryos Embryogenic callus
4. Kononowicz and Janick (1984) Somatic embryos Embryogenic callus
5. Novak et aI. (1986) Immature embryos Direct and indirect
embryogenesis
6. Litz (1986) Leaf tissue Developmentally arrested
somatic embryos
7. Elhag et al. (1987, 1988) Somatic embryos Embryogenic callus
8. Adu-Ampomah et al. (1988) Immature embryos Direct and indirect
embryogenesis
9. Wen (1989) Immature embryos Direct embryogenesis
10. Duham et al. (1989) Immature embryos Direct embryogenesis
II. Dos Santos and Machado (1989) Immature embryos Direct embryogenesis
12. Chatelet and Dufour (1990) Immature embryos Direct embryogenesis
13. Sondahl et al. (1989) Nucellus, immature Somatic embryos, multistep
flower petals procedure
14. Aguilar et al. (1992) Mature cotyledons Somatic embryos, two-step
procedure
15. Chatelet et al. (1992) Nucellus, inner Developmentally arrested
integument somatic embryos
16. Figueira and Janick (1993) Nucellus Somatic embryos, multistep
procedure
17. Lopez-Baez et al. (1993) Immature flower Somatic embryos, multistep
procedure
Somatic Embryogenesis in Cacao (Theobroma cacao) 457
used for initiation. Culturing the zygotic embryos in liquid rather than on
semisolid medium can also improve direct embryogenesis (Pence et al. 1980;
Adu-Ampomah et al. 1988).
The genetic background of the tissue is important for the success of these
procedures, as zygotic embryos from different varieties have shown a wide range
in the embryogenic response (Pence et al. 1980; Adu-Ampomah et al. 1988; Wen
1989). The stage of the immature embryo used is also a factor. Embryos of
approximately 4-10 mm in length gave the highest frequency of somatic em-
bryos, compared with earlier or later stages, on medium containing auxin and
coconut water (Pence et al. 1980). No somatic embryogenesis was observed from
mature embryos on this medium or in a similar study on medium containing
auxin alone (Duhem et al. 1989).
Somatic embryos can be initiated from the cotyledons of mature zygotic
embryos using a two-step protocol (Aguilar et al. 1992). Slices of mature
cotyledons are incubated on a medium with both cytokinin and auxin in the dark
for 3 months, during which globular and torpedo stage embryos are formed.
These are then transferred to a medium without growth regulators in the light for
an additional month for further development.
In addition to direct embryogenesis, immature zygotic embryos have also
been used to initiate embryogenic callus (Kononowicz et al. 1984). The addition
of 2,4-dichlorophenoxyacetic acid (2,4-D) stimulated the production of em-
bryos, as did coconut water or replacing sucrose with glucose or fructose
(Kononowicz and Janick 1984; Elhag et al. 1987, 1988). In a particular callus line,
gibberellic acid (GA3) also stimulated embryogenesis (Kononowicz and Janick
1984). Embryogenic cacao callus has been used to study changes in nuclear
DNA, RNA, and protein which accompany embryo initiation (e.g. Kononowicz
and Janick 1988a,b).
Although successful in producing somatic embryos, these methods depend
on the zygotic embryo as the explant source and are not useful for clonal
propagation. Other tissues of cacao have not been as responsive in initiating
somatic embryos, but some procedures have been successful. Somatic embryo-
genesis has been reported from leaf tissue, stimulated by very high levels of auxin
and cytokinin (Litz 1986). These embryos were developmentally arrested, how-
ever, and did not grow beyond the heart-shaped stage.
A multistep procedure for initiating somatic embryos from nucellar and
floral tissues was developed by Sondahl et al. (1989). This consisted of separate
semisolid media for regeneration, embryo development, embryo maturation,
and germination. Nucellus was inoculated on a basal medium of reduced
strength MS salts supplemented with auxin, cytokinin, polyvinylpyrrolidone
(PVP), and various organic addenda, including casein hydrolysate, cysteine, malt
extract, and coconut water, and the cultures were incubated in the dark. The
small embryoids which were produced were then transferred to another complex
medium with cytokinin and auxin, as well as GA3 and abscisic acid (ABA) in the
light. After formation of the cotyledons, the embryos were moved to a matura-
tion medium with cytokinin, auxin, GA3 and ABA with increased sucrose and
charcoal. This allowed development of a root and shoot pole in preparation for
germination. Somatic embryos were also obtained from immature flower petals
458 V.C.Pence
cultured on a medium with auxin and cytokinin. Once the embryos were
initiated, they were transferred to development medium, etc., as for nucellus-
derived embryos.
Figueira and Janick (1993) were also able to initiate nucellar embryos using
the same initiation medium in liquid form. After 2 months in the dark, callus was
transferred to a semisolid medium with cytokinin, malt extract, coconut water,
and PVP in the light for a further 2 months. The embryogenic callus was then
transferred to a maintenance medium of MS salts and casein hydrolysate, but
lacking growth regulators and other complex addenda. This allowed the develop-
ment of globular and torpedo stage embryos. These could be matured further in
preparation for germination by transferring to a liquid medium containing a low
(1 %) level of sucrose supplemented with 4.4% sorbitol.
Proembryos have been reported from nucellus and inner integument tissue
using a simpler medium containing auxin and cytokinin and dark incubation, but
these embryos did not develop further, possibly due to the presence of phenolics
which accumulated in the embryos (Chatelet et al. 1992). The addition of silver
nitrate to the medium did not overcome thi~ problem.
Somatic embryos have also been obtained from immature flower explants
using a multistep protocol (Lopez-Baez et al. 1993). A critical step in this
procedure was the removal of auxin and cytokinin after 2-3 weeks on an
initiation medium containing amino acids and coconut water. After another 6-8
weeks, globular embryos were then moved to a third medium containing half-
strength salts, auxin, GA 3, adenine sulfate and maltose for maturation.
2.2 Development of Somatic Embryos
Direct development of somatic embryos from zygotic embryo tissues occurs

through two apparently distinct processes. The first is a "budding" process,
whereby glandular hair-like structures on the surface of the zygotic embryo give
rise to somatic embryos which progress through the normal stages of embryo
development (Fig. la; Pence et al. 1980; Adu-Ampomahet al. 1988). These have
been examined with scanning electron microscopy, revealing a wrinkled surface
on the globular stage embryos, which becomes smooth at the heart stage (Dos
Santos and Machado 1989). Alternatively, a "nonbudding" process occurs, in
which somatic embryos develop from internal cotyledonary tissue (Fig. I b; Pence
et al. 1980). In some cases, it appears that nonbudded embryos are arrested
before cotyledon development and initiate budded embryos from their surfaces.
Budded embryos and embryos developing from callus often have a suspen-
sor-like structure (Esan 1977; Pence et al. 1980; Kononowicz et al. 1984; Adu-
Ampomah et al. 1988; Dos Santos and Machado 1989) similar to that reported
for zygotic embryos of T. cacao (Bouharmont 1960) and generally appear
normal in morphology. However, abnormalities among somatic cacao embryos,
such as polycotyledony and fasciation, have also been reported (Dos Santos and
Machado 1989; Duhem et al. 1989).
Histological studies have also been made on cultured nucellus and inner
integument tissue, revealing the development of embryogenic cells and proem-
Fig. L a A cluster of budded somatic cacao embryos of various developmental stages. bNonbudded
embryos developing from an immature zygotic cacao embryo. (Pence et al. 1980)
bryos both from the edges and from internal tissues (Chatelet et al. 1992). Under
the conditions used, development was arrested at this stage, with the embryonic
cells becoming vacuolized and accumulating phenolic substances.
As well as being morphologically similar to zygotic embryos, somatic
embryos of cacao also have normal biosynthetic capabilities. Anthocyanins,
fatty acids, triglycerides, and alkaloids all can be stimulated in somatic embryos
d1,lring in vitro maturation in a manner similar to zygotic embryos in vitro,
although levds are generally less than those observed in zygotic embryos matur-
ing in vivo (Pence et al. 1981a,b; Janick et al. 1982; Paiva and Janick 1983).
2.3 Conversion of Embryos to Plants
Immature embryos of cacao, whether zygotic or somatic, do not readily undergo

precocious germination to normal plants. Several protocols have, however, been
reported for the conversion of somatic embryos to plantlets.
460 V.C.Pence
Radicle extension of cacao somatic embryos was obtained by lowering the

concentration of MS salts from full- to half-strength and by transferring the
embryos to fresh medium (Wang and Janick 1984). The latter result suggested
the presence of a germination inhibitor leached from the embryos into the
medium, and tests with leachates from somatic embryos indicated that they were
capable of inhibiting lettuce seed germination (Wang and Janick 1985). The
inhibitor was not identified, although it was shown not to co-chromatograph
with abscisic acid.
Somatic cacao embryos cultured on a medium with zeatin and naphthalene
acetic acid (NAA) germinated into plants when the cotyledons were removed
(Novak et al. 1986). This also suggested the presence of a germination inhibitor
in the cotyledons. Germination of decotyledonated embryos was also obtained
on a medium containing charcoal or in a liquid medium with NAA and GA3
(Adu-Ampomah et al. 1988). The addition of charcoal, 2-isopentenyl adenine
(2iP) or zeatin, a reduction of MS salts to half-strength, and an increase in gas
exchange in the culture tube also promoted the germination of somatic cacao
embryos axes (Duhem et al. 1989). Embryos did not germinate ifbenzylamino-
purine (BAP) was used as the cytokinin. In one report, cellulose supports have
been used in place of a gelled medium for germination (Chatelet and Dufour
1990). Decotyledonated somatic embryos have also been grown to plants by
micro grafting the embryos onto 3-week-old seedling rootstock (Aguilar et al.
1992).
Sandahl et al. (1989) used a medium containing coconut water, BAP, IAA,
GA 3, ABA, and charcoal for the germination of somatic embryos which had
been brought through the multistep initiation, development, and maturation
protocol for cacao embryogenesis. Figueira and Janick (1993) similarly trans-
ferred matured somatic embryos to a semisolid woody plant medium (Lloyd and
McCown 1980) with fructose. Embryos were then moved to chambers with
20 000 ppm CO 2 , which stimulated seedling development.
In their multistep procedure, Lopez-Beaz et al. (1993) transferred matured
somatic embryos to half-strength MS medium with amino acids, NAA, GA3,
2iP, ABA, adenine sulfate, activated charcoal, and glucose or maltose to obtain
germination. Embryos longer than 1 cm were then transferred to a conversion
medium lacking growth regulators and with reduced levels of charcoal and
glucose, on which further growth and leaf development occurred.
2.4 Germplasm Preservation
Long-term preservation of the genetic diversity of cacao and of related wild

species is problematic, because of the short-lived nature of cacao seeds. Unlike
many species which have seeds which can survive drying and can be maintained
in cool, dry storage for extended periods of time, cacao seeds are both
desiccation- and cold-sensitive. They remain viable for only a few days to a few
weeks and must either germinate or deteriorate.
Methods for the cryopreservation of tissues in liquid nitrogen (LN) (see
Bajaj 1991) could provide alternatives for long-term storage. Immature zygotic
Fig. 2. Immature cacao zygotic embryos cryopreserved without precuIture, limp and translucent,
producing a somatic embryos, and b callus; and immature cacao zygotic embryos cryopreserved with
preculture, firm and brown, producing c somatic embryos, and d callus
embryos of cacao have been successfully cryopreserved using both slow freezing
and desiccated LN exposure (Pence 1991). For both procedures, embryos of
approximately 110- 120 days development were used. Some embryos were cryo-
protected immediately, while others were precultured on a medium increasing in
sucrose concentration to 21 % through successive media changes or on a medium
containing the basal level of 3% sucrose, also changed successively but main-
tained at 3%. Media were tested with and without ABA.
For slow freezing, the embryos were cryoprotected in a liquid medium
containing 0.5 M sucrose and 10% dimethylsufoxide (DMSO) for 45 min and
then frozen at a rate of 0.4 °C/min down to - 35°C, at which point the embryos
were transferred to LN. Thawing was done rapidly in a 40°C water bath and the
embryos transferred to a recovery medium containing coconut water, charcoal,
andNAA.
After 3 weeks of incubation, callus growth and somatic embryos were
produced by the recovering cryopreserved embryos. In all cases the original
embryo structure was damaged such that it could not resume growth. Embryos
which were cryopreserved without preculturing were translucent and limp upon
recovery, whereas embryos which were precultured were more rigid in structure
and became dark brown. In both cases, although germination of the original
462 V.C.Pence
Table 2. Callus and somatic embryo production from hydrated immature zygotic cacao embryos
pretreated and slow frozen. (Pence 1991)
Pretreatment Total No. with No. with somatic

callus embryos
Experiment I
Unfrozen control 8 6 1
Frozen control, no 11 7 0
preculture
3% sucrose 15 6 3
3% sucrose + ABA 16 2 6
21% sucrose 8 4 0
21 % sucrose + ABA 13 11 I
Summary:
All 3% sucrose 31 8 9
All 21% sucrose 21 IS I
Experiment II
Frozen control, no 11 5
preculture
3% sucrose 36 2 0
21% sucrose 44 IS 3
somatic embryo did not occur, some cells did survive and grew either as callus or
somatic embryos (Fig. 2). This growth often occurred from the area of the shoot
meristem.
The proportion of callus produced was highest from embryos pretreated
with 21% sucrose (Table 2). S<l>matic embryogenesis occurred from non-precul-
tured embryos and embryos tJrecultured on 21% sucrose at a very low rate,
whereas on 3% sucrose, embryogenesis occurred either at a higher rate or did not
occur at all. This suggests that pretreating embryo~ on the 3% sucrose medium
holds potential for stimulating embryogenesis from cryopresenred zygotic em-
bryos, but that other factors are also importantand need to be defined.
A desiccated cryostorage procedure was also used in which the immature
zygotic embryos were dried under the air flow of a laminar flow hood overnight
and then fast frozen by immersion into LN. Callus and somatic embryogenesis
occurred from these embryos, but only from those precultured with 21 % sucrose
and ABA (Table 3). The recovery of somatic embryos was low, but these
experiments demonstrate the possibility of using desiccation, with further refine-
ments, for cryopreserving cacao embryos without the need for cryoprotectants
and a programmable freezer. Vitrification, a technique which has been used to
cryopreserve cell cultures and shoot tips (Yamaqa et al. 1991), might similarly
provide a simpler method for freezing cacao embryos in the hydrated state.
Somatic embryos arising from frozen-thawed zygotic embryos in these
experiments germinated into plants using the procedure of Duhem et al. (1989).
Although the frozen zygotic embryo was damaged by the freezing and was
unable to undergo normal germination, germination of clonal embryos pro-
duced from surviving cells allowed the retrieval of the cryopreserved germplasm.
Table 3. The effect of preculture on callus initiation and somatic embryogenesis from immature
zygotic cacao embryos desiccated and fast frozen in liquid nitrogen. (Pence 1991)
Experiment pretreatment Total No. No. with No. with somatic

explants callus embryos
21% sucrose + ABA 14 6 o

(dried, not frozen)
21 % sucrose + ABA 15 9
II None 70 0 o
III 3% sucrose 6 0 o
21% sucrose 11 0 o
Cryopreservation of zygotic embryos provides the genetic equivalent of a

seed bank. Germplasm preservation of characterized cacao trees requires that
clonal material be frozen. Somatic embryos originating from nucellar, floral or
other maternal tissue could provide such clonal tissue for storage. Since somatic
embryos are capable of secondary somatic embryogenesis (Pence et al. 1979) and
are similar morphologically and physiologically to zygotic embryos, it is likely
that they will respond to freezing and thawing as the zygotic embryos. Prelimi-
nary experiments in this laboratory suggest that this is the case.
Somatic embryogenesis is well documented in cacao. It has until relatively

recently, however, been confined to embryogenesis from zygotic embryos. This
has limited the usefulness of the technology because of the need for clonal
propagation systems in cacao.
The deVelopment of protocols which have extended the ability to initiate
somatic embryogenesis to previously unresponsive tissues (e.g., mature cotyle-
dons, nucellus, floral explants) has depended on the use of two or more successive
culture regimes and relatively long culture periods. This may reflect the
presence of inhibitors which might be lacking in the immature zygotic embryos,
or the lack of some stimulating factor present in the immature embryos. An
understanding of the differences between embryogenesis from immature
embryos and from other cacao tissues should be instructive in more clearly
defining the physiology of this process.
The conversion of somatic embryos to plants has traditionally been difficult
in cacao. Several procedures have been reported, and these have included the
partial or full removal of the cotyledons or the maturation of the embryos.
464 V.C.Pence
Recent studies suggest that plant development can also be stimulated by an

elevated level of CO2 , If somatic embryos from clonal material can be converted
to plants in high percentage, the application of somatic embryogenesis tech-
nology to real problems in cacao improvement can be undertaken.
Somatic embryos from clonal material would also be good candidates for the
development of cryostorage protocols for cacao germplasm. The proven embry-
ogenic capacity of this tissue allows that if damage occurs to the cryopreserved
embryo - as is likely with a structure this large - surviving cells will have the
capacity to regenerate somatic embryos, thereby preserving the line. This tech-
nology is feasible, although optimization of the cryopreservation precedures is
yet to be done. It could, however, provide a method for greatly reducing the cost
and labor of maintaining and transporting cacao germplasm.
4 Protocols
4.1 Initiation of Somatic Embryos from Immature Zygotic Embryos

(Pence et al. 1979, 1980)
I. Surface sterilize an immature cacao fruit, about 110-120 days post-fertilization, by spraying with
70% ethanol.
2. Aseptically dissect out the small white embryos and transfer directly to medium.
3. Culture embryos on a semisolid medium ofMS salts and organics, I gil casein hydrolysate, 1.5 mgll
NAA, 10% coconut water, 3% sucrose, 0.2% Phytage1 or 1% agar, in a 16:8 h light:dark cycle, at
approximately 26°C.
4. As tissues producing somatic embryos develop, transfer these to medium lacking auxin and coconut
water for proliferation and maintenance.
4.2 Initiation of Somatic Embryos from Mature Zygotic Embryos

(Aguilar et al. 1992)
I. Surface sterilize mature fruit by washing with detergent, spraying with 70% ethanol and flaming.
2. Aseptically remove embryos and make transverse slices of the cotyledons.
3. Culture the slices on a semisolid medium containing MS salts and organics, 3% sucrose, 3 mgll
BAP, I mg/l NAA, 0.8% agar, in the dark for 90 days.
4. Globular and torpedo stage embryos which form are then transferred to a medium ofMS salts and
organics, 0.5 gil casein hydrolysate, 5% sucrose, in a 16:8 h light:dark photoperiod for 30 days for
further development.
4.3 Initiation of Somatic Embryos from Nucellar Tissue

(Sondahl et al. 1989; Figueira and Janick 1993)
I. Surface sterilize immature pods and aseptically remove ovules, approximately 12 mm in length.
2. Discard the basal end of the ovule and slice the rest of the tissue for culture.
3. Culture ovule slices in a liquid medium of half-strength MS salts, full-strength MS organics, 0.5 g/l
malt extract, 0.5 gil casein hydrolysate. 0.2%PVP 10000, 0.9mgll2,4-D, 0.1 mgll2iP, 10% coconut
water, 4% sucrose on a gyratory shaker for 30 days in the dark.
4. Transfer cultures to fresh medium and incubate for another 30 days.
5. Transfer cultures to semisolid medium of half-strength MS salts, full-strength MS organics, 0.1 gil
malt extract, 0.2% PVP 10000,0.5 mgll2iP, 10% coconut water, 4% sucrose, and 0.8% agar, in the
light for 60 days.
6. Transfer cultures to semisolid maintenance medium of half-strength MS salts, full-strength MS
organics, I gil casein hydrolysate, 3% sucrose, and 0.8% agar.
4.4 Maturation and Conversion of Matured Somatic Embryos to Plants

(Lopez-Baez et al. 1993)
1. Transfer globular embryos to maturation medium, consisting of half-strength MS salts, 004 mgll
L-leucine,OA mgll L-lysine, 0.2 mgll L- tryptophan, 004 mgll L- arginine, 0.05 mgll IAA, 0.05 mgll
IBA, 0.02 mgll GA3 , 0.5 mgll adenine sulfate, 40 gil maltose, 2 gIl gelrite. Incubate in a 12/12 h light!
dark cycle at 26°C for 4-5 weeks.
2. Transfer matured embryos to a germination medium, consisting of half-strength MS salts, the same
amino acids, 0.01 mgll NAA, 0.02 mgll GA 3, 0.2 mgll2iP, I mgll ABA, 0.5 mgll adenine sulfate,
I gil activated charcoal, 80 gil maltose or 40 gil glucose, and 3 gil gelrite. Maintain cultures for
4-5 weeks in a 12112 h light!dark cycle, with a 31/26 °C temperature cycle.
3. Transfer embryos with hypocotyls longer than 1 em to half-strength MS medium with the same
amino acids, 0.15 gil activated charcoal, 5 gil glucose, and 3 gil gelrite in a larger container, such as
a baby food jar, for growth and leaf development. Maintain cultures for 6-8 weeks under the same
light and temperature regime.
4. When plants have 2-3 leaves, transfer them to a 212/1 mixture ofvermiculite/sand/perlite, under the
same light and temperature conditions. Water with one-quarter strength MS salts, without sugar,
and maintain the relative humidity at 90--95%.
5. After 5-6 weeks, transfer plants to soil.
4.5 Initiation of Somatic Embryos from Cryopreserved Embryo Tissue

(Pence 1991)
I. Aseptically excise 110--120-day small white zygotic embryos.

2. Cryoprotect embryos directly or preculture on medium of MS salts and organics, 1 gil casein
hydrolysate, 10% coconut water, 3% sucrose, 0.2% Phytagel for 1-4 weeks.
3. For cryoprotection, incubate embryos in a liquid medium of MS salts and organics, 3% sucrose,
and over a period of 45 min add, in 3 aliquots, an equal volume of medium ofMS salts and organics,
with I M sucrose and 20"10 DMSO (giving a final volume of 0.5 M sucrose and 10"10 DMSO).
4. Cool embryos for an additional 30 min and transfer to 2-ml cryovials.
5. Freeze at a rate of 004 °C/min down -35°C and quickly transfer vials to liquid nitrogen.
6. For thawing, remove vials from liquid nitrogen and immediately transfer to a water bath at 40°C
for 2 min.
7. Transfer thawed embryos to recovery medium ofMS salts and organics, IgIl casein hydrolysate,
10% coconut water, 3 mgll NAA, 0.05% charcoal, 0.2% Phytagel to initiate growth of
surviving cells.
466 V.C.Pence
References
Adu-Ampomah Y, Novak FJ, Afza R, Van Duren M, Perea Dallos M (1988) Initiation and growth of
somatic embryos in cocoa (Theobroma cacao L.). Cafe Cacao The 32: 187-200
Aguilar ME, Villalobos VM, Vasquez N (1992) Production of cocoa plant (Theobroma cacao L.) via
micrografting of somatic embryos. In Vitro Cell Dev BioI 28P: 15-19
Bajaj YPS (1991) Storage and cryopreservation of in vitro cultures. In: Bajaj YPS (ed) Biotechnology
in agriculture and forestry, vol 17. High-tech and micropropagation 1. Springer, Berlin Heidelberg
Bouharmont J (1960) Recherches cytologiques sur la fructification et I'incompatibilite chez Theobroma
cacao L. Publ Institut National pour l'Etude Agronomique du Congo (INEAC), Ser Sci no 89
Chatelet P, Dufour M (1990) Culture in vitro du cacaoyer (Theobroma cacao L.). Progres recent en
embryogenese somatique et en micropropagation. In: Cinquantenaire de la culture in vitro.
Versailles (France). Les colloques de I'INRA, No 51 INRA, Paris
Chatelet P, Michauxferriere N, Dublin P (1992) Embryogenic potential in nucellus and inner
integument tissue cultures of immature cacao seeds. C R Acad Sci Ser III 315: 55-62
Dos Santos AVP, Machado RD (1989) A scanning electron microscope study of Theobroma cacao
somatic embryogenesis. Ann Bot 64: 293-296
Duhem K, Le Mercier N, Boxus P (1989) Donnees nouvelles sur I'induction et Ie developpement
d'embryons somatiques chez Theobroma cacao L. Cafe Cacao The 33: 9-14
Elhag HM, Whipkey A, Janick J (1987) Induction of somatic embryogenesis from callus in Theobroma
cacao in response to carbon source and concentration. Rev Theobroma 17: 153-162
Elhag H, Whipkey A, Janick J (1988) Factors affecting asexual embryogenesis via callus in Theobroma
cacao L. Arab Gulf J Sci Res Agric Bio Sci B6: 31-43
Esan EB (1977) Tissue culture studies on cacao (Theobroma Cacao L.) A supplementation of current
research. In: Proc 5th Int Cacao Res Conf, 1975, Cacao Res Inst 1'l"igeria, Ibadan, pp 116-125
Esan EB (1992) Micropropagation of cocoa (Theobroma cacao L.). In: Bajaj YPS (ed) Biotechnology
in agriculture and forestry, vol 18. High-tech and micropropagation II. Springer, Berlin Heidelberg
New York, pp 96-122
Figueira A, Janick J (1993) Development of nucellar somatic embryos of Theobroma cacao. Acta
Hortic 336: 231-236
Janick J, Wright DC, Hasegawa PM (1982) In vitro production of cacao seed lipids. J Am Soc Hortic
Sci 107: 919-922
Kononowicz H, Janick J (1984) Response of embryogenic callus of Theobroma cacao L. to gibberellic
acid and inhibitors of gibberellic acid synthesis. Z Pflanzenphysioll13: 359-366
Kononowicz H, Janick J (1988a) Changes in nucleus, nucleolus and cell size accompanying somatic
embryogenesis of Theobroma cacao L. 1. Relationship between DNA and total protein content and
size of nucleus, nucleolus and cell. Folia Histochem Cytobiol26: 237-248
Kononowicz H, Janick J (I 988b) Somatic embryogenesis via callus of Theobroma cacao L. I. Cell cycle,
DNA content, RNA synthesis and DNA template activity. Acta Physiol Plant 10: 93-106
Kononowicz H, Kononowicz AK, Janick J (1984) Asexual embryogenesis via callus of Theobroma
cacao L. Z Pflanzenphysioll13: 347-358
Litz RE (1986) Tissue culture studies with Theobroma cacao. In: Dimick PS (ed) Proc Cacao
Biotechnol Symp. Pennsylvania State Univ, University Park, pp 111-120
Lloyd G, McCown B (1980) Commercially feasible micropropagation of mountain laurel, Kalmia
latifolia by use of shoot-tip culture. Comb Proc Int Plant Propagators Soc 30: 421-427
Lopez-Baez 0, Bollon H, Eskes A, Petiard V (1993) Embryogenese somatique de cacaoyer Theobroma
cacao L. it partir de pieces florales. C R Acad Sci Paris, Life Sci 316: 579-584
Novak FJ, Donini B, Owusu G (1986) Somatic embryogenesis and in vitro plant development of cocoa
(Theobroma cacao). In: Proc Int Symp Nuclear techniques and in vitro culture for plant
improvement. IAEA, Vienna, pp 443-449
Paiva M, Janick J (1983) In vivo and in vitro production of alkaloids in Theobroma cacao L. Acta
Hortic 131: 265-273
Pence VC (1989) Cacao (TheobYoma cacao L.). In: Bajaj YPS (ed) Biotechnology in agriculture and
forestry, vol 5. Trees II. Springer, Berlin Heidelberg New York, pp 203-221
Pence VC (1991) Cryopreservation of immature embryos of Theobroma cacao. Plant Cell Rep 10:
144-147
Pence VC, Hasegawa PM, Janick J (1979) Asexual embryogenesis in Theobroma cacao L. J Am Soc
Hortic Sci 104: 145-148
Pence VC, Hasegawa PM, Janick J (1980) Initiation and development of asexual embryos of
Theobroma cacao L. in vitro. Z Pfianzenphysiol 98: 1-14
Pence VC, Hasegawa PM, Janick J (1981a) In vitro cotyledonary development and anthocyanin
synthesis in zygotic and asexual embryos of Theobroma cacao. J Am Soc Hortic Sci 106: 381-385
Pence VC, Hasegawa PM, Janick J (198 I b) Sucrose-mediated regulation of fatty acid composition in
asexual embryos of Theobroma cacao. Physiol Plant 53: 378-384
Sondahl M, Sereduk TB, Chen Z, Bellato CM (1989) Somatic embryogenesis and plant regeneration
of cacao patent, patent no. 88/3078, Jan 25, 1989. Republic of South Africa
Wang Y-C, Janick J (1984) Inducing precocious germination in asexual embryos of cacao. HortScience
19: 839-841
Wang Y-C, Janick J (1985) Characterizing the germination inhibitor from asexual embryo leachate of
Theobroma cacao. J Am Soc Hortic Sci 110: 113-117
Wen M-C (1989) Cacao tissue culture. PhD Thesis, Cornell University, Ithaca, New York
White PR (1963) The cultivation of animal and plant cells, 2nd edn. Ronald Press, New York
Yamada T, Sakai A, Matsumura T, Higuchi S (1991) Cryopreservation of apical meristems of white
clover (Trifolium repens L.) by vitrification. Plant Sci 78: 81-87
SUbject Index
i\Bi\ 134,266,288,327,383,438,441 black spruce 431-445

accessory embryos 105 black walnut 370
acclimatization 267,441 blot hybridization 9
activated charcoal 302, 345, 438 Brassica napus 210
adventitious embryogenesis 346
- shoots 173 cacao 221,455-467
adventive embryos 71,105 calcium alginate 134
aescin 233 callogenesis 304, 339
Aesculus hippocastanum 233-245 call us cult ure 405
Agrobacterium rhizogenes 170 -induction 304
A. tumefaciens 114,210,212,261,410 - tissue 77
airlift bioreactor 178, 199 Carica papaya 114,212,260-279
Ajuga rep tans 94 Carica species 262-264
alfalfa 130, 154, 195 carrot 4,30,46,58,99,183
alginate coating 223 Carya illinoensis 214
anabasine 54 casein hydrolysate 287,456
anatomy 71-86 cassava 222
androgenesis 3 cavaderine 110
androgenic embryo ids 234 celery 133
anthocyanins 456 cell density 252
applications of somatic embryos 105-125 - suspension 156, 385
arabinogalactan protein 6 cell's commitment 3-19
i\racaceae 335 cellular competence 13
Armoracia rusticana 171 charcoal 384
artificial seed 153, 170 chestnut 222
asexual embryos 105 chinese tulip tree 388
asparagus 223 chilling 186
Atropa belladonna 107 chitosan 117
automation 97 chocolate 384
- of cell culture 138 Citrus sinensis 107, 292
- of somatic embryo 139 Citrus species 280--298
- of synthetic seed 138 clonal propagation 423
auxin 31,63,175,195 clover 56
auxin binding protein 25 coat protein 261
coconut 299-317
Bi\ 130,358,379,406 - water 269
Bacillus thuringiensis 375 Cocos nucifera 299-317
Bi\P 327 commercial outlook 193
barley 131 - scale-up 97
bead sorter 99 computer vision 97, 253
Betula species 246-259 conifers 57, 143
Betulaceae 318 copra 299
biomass measurement 91 Corylus species 318-334
bioreactor 89,105,115,178,199,221,252,349 cotton 144
birches 246-259 crop improvement 105
470 Subject Index
cryopreservation 221-229,256,283,349 filbert 318

cryoprotectant 226 fluid drilling 117, 135, 181-192
cytokinins 175,197,234,312,347,405,457 forage crops 143
Fortunella 280
2,4-D 155,173,265,286,302,340,406 fraser magnolia 389
dedifferentiation 194 fruit crops 144
dehydration 130 fundamental aspects 280
dehydrogenase activity 108
delivery methods 133, 181 GA 131
Dendrophthoefalcata 108 gametogenesis 107
desiccation 225 gel electrophoresis III
- of somatic embryos 152-169 -type 135
-tolerance 128,161 gelatin 116
dicamba 286 gelrite 116, 358
differentiation 30,201, 217 gene expression 35,38,41-52, III
direct embryogenesis 194,330,456 - transfer 15,114,207,399,425
DMSO 223,461 genetic diversity 460
DNA 207 - engineering 209
-hybridization 37 - stability 170, 424
- measurement 13 - transformation 207-220
dormancy - variability 13, 300, 333
drop culture 195 genotype 225
dry artificial seed 152 germination 346, 374, 423, 439
germplasm preservation 153,460
Elaeis guineensis 335-352 - storage 132
electron microscopy 458 grape 214
electrophoresis 34 growth regulators 288, 347, 449
ELISA 372 GUS 210,214,400,425
embryogenic callus 283, 359
- cell culture 248, 392 hairy roots 170
- expression 360 hazelnut 318-334
- suspensions 348,435 Hevea brasiliensis 353-369
embryo conversion 254, 256 Hippocastanaceae 233
-development 157 histochemical assay 400
- germination 327 histology 107,237,291,308,327,458
- induction 155 hormones 252
- initiation 456 horse chestnut 233-245
-maturation 158 -radish 171
-synchronization 156 HPLC 202, 242, 302
embryogenesis 341 hybrid cereals 145
embryoids 105,290 hybridization analysis 37
encapsulation 175
endogenous auxin 32 IAA 173,276
endoplasmic reticulum 240 IBA 173
epigenetic events 10,442 image analysis 113
estimated cost 141 -data 90
estimation of viability 227 immunofluorescence 6
ethephon 11 0, 288 indirect embryogenesis 194,329
ethylene 62 induction medium 397
Euphorbiaceae 353 interspecific hybrids 370
exogenous auxin 32 intracellular freezing 226
explants 373
expression of embryogenesis 307 Juglans regia 114,211
Juglans species 370
fatty acid 299 juvenile material 322
field gene bank 455
- trial 346 kanamycin resistance 210
Subject Index 471
kinetin 176, 358,379 papaya 133,211,212,260-279

parthenocarpy 3
larch 378-387 particle bombardment 210
Larix species 379 -gun 38
Larix occidentalis 378-387 pecan 214,370
lea gene 35,45, 112 PEG 183, 199, 399
lettuce 136 Pennisetum americanum 109
liquid culture 195 peroxidases 293, 304
Liriodendron 115, 388-403 physiological stability 424
phytohormones 319
machine vision analysis 87-101 Phytophthora palmivora 261
Magnolia 388-403 Picea abies 415-430
Magnoliaceae 388-403 P. glauca 114, 128,215
maize 210 P. mariana 431-445
malformed embryos 392 P. rubens 431-445
mandarin 290 picioram 274, 286
Mangifera indica 56, 114,210 plant regeneration 199,290,384
mango 221 plantlet formation 327
mantled palm 347 plastid D~A 399
mass propagation 22, 346 plating density 93
media 248,264,281,319,327,372 poinsettia 131
medicagin 161 polarity 38, 75
Medicago sativa 56, 154 pollen embryogenesis 72
melon 222 -mitosis 72
microcomputer 91 polyamines 53-70, 109,355
micrografting 115 polyembryonic cultures 347
microinjectile 261 polyembryony 71
microinjection 38, 114 polyethyleneoxide 133
micropropagation 193-206,266,372,405 polypeptide 34, III
molecular basis 30, III Poncirus 280
- biology 255 poplars 446-454
- markers 34, III Populus nigra x P. maximowiczii 446
monosaccharides 203 proembryoids 76
morphogenetic capacity 20 proline 156
multicellular pathway 307 protoplast culture 427,441
mUltiple embryos 140 - isolation 398, 406
pseudocarpels 347
~AJ\ 173,265,340,346 pseudothallus 73
neomorphs 373 putrescine 53, 109
nicotine 54
non-quiescent somatic embryos 116 quiescence 128
~orthern blot analysis 47
~orway spruce 415-430 rapeseed 210
nucellarcallus 295 recalcitrant seed 221
nucleic acid hybridization 112 red spruce 431-445
nut crops 144 regeneration medium 399
- of plants 345
oil palm 132, 335-352 repetitive embryogenesis 211,370,397
Olea europaea 404-414 reporter gene 399
Oleaceae 404 RFLP 10
olive 404-414 robotics 97
ontogenesis 364 robots 115
orchard grass 109 rubber 221, 353-369
organogenesis 23,71,214,261,303
ornamental crops 142 saccharides 287
Salicaceae 455
papain 260 scanning electron microscopy 73,171,174
472 Subject Index
secondary embryos 105, 158, 213, 223, 292, transformation 375

326,463 transformed somatic embryos 370
seed maturity 325 transgenic cell lines 61
shake culture 222 - plants 44, 114, 170,207
shoot organogenesis 410 transient expression 426
- tips 170 tree oflife 299
silica gel 223 trees 231-463
Sinapis alba 109 trehalose 165
sodium alginate 171 tulip poplar 388
soft gel capsules 134 -tree 388
Solanum melongena 56
somaclonal variation 27, 131, 442 ultrastructural studies 80, 107, 234, 237
somaplants 367 unicellular origin 308
somatic meiosis 8, 10, 107
Sorghum bicolor 109 vegetable crops 143
Southern blot analysis 212, 400 vegetative propagation 301
soybean 130, 144 video image analysis 87
spermidine 53, 109,355 Vilis rupestris 114,214
spermine 53, 109, 357 vitrification 224
Sterculiaceae 455
sterilization 318 walnut 211,222,370-377
storage protein 160,392,417 water binding 162
-temperature 226 western larch 378-397
suspension culture 46, 385
sweet potato 98 yellow cucumber tree 395
sweetbay magnolia 395 yellow poplar 388
tenera hybrids 337 Zea mays 130

Theobroma cacao 455-467 zeatin 405, 460
thidiazuron 274, 405 zygotic embryo 72,127
thioproline 156 zymogram 292
totipotency 3,20, 105
Spri nger-Verlag
and the Environment
We at Springer-Verlag firmly believe that an

international science publisher has a special
obligation to the environment, and our corpo-
rate pol icies consistently reflect th is conviction.
We also expect our busi-

ness partners - paper mills, printers, packag-
ing manufacturers, etc. - to commit themselves
to using environmentally friendly materials and
production processes.
The paper in this book is made from

low- or no-chlorine pulp and is acid free, in
conformance with international standards for
paper permanency.

(Biotechnology in Agriculture and Forestry 30) V. Nuti Ronchi, L. Giorgetti (auth.), Professor Dr. Y. P. S. Bajaj (eds.)-Somatic Embryogenesis and Synthetic Seed I-Springer-Verlag Berlin Heidelberg (1.pdf

Uploaded by

Copyright:

Available Formats

(Biotechnology in Agriculture and Forestry 30) V. Nuti Ronchi, L. Giorgetti (auth.), Professor Dr. Y. P. S. Bajaj (eds.)-Somatic Embryogenesis and Synthetic Seed I-Springer-Verlag Berlin Heidelberg (1.pdf

Uploaded by

Document Information

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

(Biotechnology in Agriculture and Forestry 30) V. Nuti Ronchi, L. Giorgetti (auth.), Professor Dr. Y. P. S. Bajaj (eds.)-Somatic Embryogenesis and Synthetic Seed I-Springer-Verlag Berlin Heidelberg (1.pdf

Uploaded by

Copyright:

Available Formats

Volumes already published

Volume 1: Trees I (1986)

Edited by YP.S. Bajaj

With 164 Figures and 54 Tables

Springer-Verlag Berlin Heidelberg GmbH

ISBN 978-3-642-08183-5 ISBN 978-3-662-03091-2 (eBook)

CIP data applied for

Somatic Embryogenesis and Synthetic Seed I comprises 31 chapters,

Section ICommitment of the cell to somatic embryogenesis; early

Section II Vegetables and Fruits - asparagus, chicory, cucurbits,

These books will be of interest to students, teachers, and research

New Delhi, January 1995 Professor y.P.S. BAJAJ

Section I Basic and Fundamental Aspects

1.1 The Cell's Commitment to Somatic Embryogenesis

1.2 Early Events in Embryogenesis

I.3 Molecular Basis of Somatic Embryogenesis

1.4 Gene Expression in Somatic Embryos

1.5 Role ofPolyamines in Somatic Embryogenesis

1.6 Anatomy of Somatic Embryogenesis

1. 7 Machine Vision Analysis of Plant Cells

Section II Applications of Somatic Embryos; Technology

11.1 Somatic Embryogenesis and Its Applications

11.2 Somatic Embryogenesis and the Technology

11.3 Role of Maturation and Desiccation of Somatic Embryos

11.4 Artificial Seed Production Through Encapsulation

4 Discussion .......................................... 178

II.5 Fluid Drilling as a Delivery System

II.6 Micropropagation Through Somatic Embryos

II. 7 Genetic Transformation of Somatic Embryos

II.8 Cryopreservation of Somatic Embryos

Section III Somatic Embryogenesis in Trees

III. 1 Somatic Embryogenesis in Horse Chestnut

III.2 Somatic Embryogenesis in Birches (Betula spp.)

IIL3 Somatic Embryogenesis in Papaya (Carica papaya L.)

III.4 Somatic Embryogenesis in Citrus Species

III. 5 Somatic Embryogenesis in Coconut (Cocos nucifera L.)

3 Discussion and Conclusion ............................. 313

I1L6 Somatic Embryogenesis in Hazelnut (Corylus Species)

III.7 Somatic Embryogenesis in Oil Palm (Elaeis guineensis Jacq.)

I1L8 Somatic Embryogenesis in Rubber Tree

I1L9 Somatic Embryogenesis in Walnut (Juglans Species)

III. 10 Somatic Embryogenesis in Western Larch

3 Summary and Conclusions ............................. 386

111.11 Somatic Embryogenesis in Magnoliaceae

111.12 Somatic Embryogenesis in Olive (Olea europaea L.)

111.13 Somatic Embryogenesis in Norway Spruce (Picea abies)

III.14 Somatic Embryogenesis in Black Spruce

111.15 Somatic Embryogenesis in Poplars

3 Summary and Conclusions ............................. 452

111.16 Somatic Embryogenesis in Cacao (Theobroma cacao)

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 469

ALBUERNE, M., Laboratorio de Fisiologia Vegetal, Departamento de

ATANAssov, A.I., Institute of Genetic Engineering,

BAJAJ, y'P.S., Former Professor of Tissue Culture,

BALDAN, B., Department of Biology, University of Padova,

BENKRIMA, L., Celex Laboratories Inc., 409 Granville Street,

BERROS, B., Laboratorio de Fisiologia Vegetal,

BUFFARD-MoREL, l., ORSTOM, Laboratoire des Ressources Genetiques

CANTLIFFE, D.l., Department of Horticultural Sciences,

CARICATO, G., Dipartimento di Produzione Vegetale sez.