Basic Science

Techniques in
Clinical Practice
Basic Science
Techniques in
Clinical Practice

H.R.H. Patel, M. Arya, and

I.S. Shergill (Eds)
H.R.H. Patel, BMSc.Hons, BM, M. Arya, FRCS
BCh, MRCS, PhD, FRCS(Urol), Molecular Uro-Oncology Research
FRCS(Eng.Hons) Fellow and Specialist Registrar
Consultant Laparoscopic Surgeon Prostate Cancer Research Centre
Head University College London
Section of Laparoscopic Urology London, UK
University College Hospital
London, UK I.S. Shergill, BSc, MRCS
Clinical Research Fellow in
and
Molecular Uro-Pathology and
Visiting Professor in Laparoscopy Specialist Registrar in Urology
and Robotic Surgery Institute of Urology
University of Rochester University College London
Medical Center London, UK
Rochester, NY, USA
and
Visiting Professor in Advanced
Laparoscopic Onco-Surgery,
Gujarat Cancer and Research
Institute
Amedabad, Gujarat, India
and
Visiting Professor in
Laparoscopic
Urology
Hospital Clinic
University of Barcelona
Barcelona, Spain

British Library Cataloguing in Publication Data

Basic science techniques in clinical practice
1. Clinical medicine—Research—Methodology
I. Patel, Hitendra II. Arya, M. III. Shergill, I. S.
616′.007
ISBN-13: 9781846285462
Library of Congress Control Number: 2006940910
ISBN: 978-1-84628-546-2 e-ISBN: 978-1-84628-740-4
Printed on acid-free paper
© Springer-Verlag London Limited 2007
Apart from any fair dealing for the purposes of research or private study, or criti-
cism or review, as permitted under the Copyright, Designs and Patents Act 1988,
this publication may only be reproduced, stored or transmitted, in any form or by
any means, with the prior permission in writing of the publishers, or in the case of
reprographic reproduction in accordance with the terms of licences issued by the
Copyright Licensing Agency. Enquiries concerning reproduction outside those
terms should be sent to the publishers.
The use of registered names, trademarks, etc. in this publication does not imply,
even in the absence of a specific statement, that such names are exempt from the
relevant laws and regulations and therefore free for general use.
Product liability: The publisher can give no guarantee for information about drug
dosage and application thereof contained in this book. In every individual case the
respective user must check its accuracy by consulting other pharmaceutical
literature.
9 8 7 6 5 4 3 2 1
Springer Science+Business Media
springer.com
Foreword

As medical research evolves, the scientific methodology and prac-

tice becomes ever more complex. The editors of this book have
brought together a timely piece of work which will truly help
future health professionals, whether they embark on research or
just wish to understand what laboratory methods are available.
An important area of the book looks at statistics, study design,
and analysis. This is particularly helpful when students, re-
searchers, and clinically active people are trying to understand
evidence-based medicine.
As the editors have put it, simple but informative text should
inspire and give confidence to all people considering undertaking
high-quality research. I envisage this to be a key book for the
future of medical research and strongly recommend it.

Professor Sir Ara Darzi

Preface

Medical researchers including doctors, nurses, medical students,

and allied health professionals have traditionally undertaken a
period of research as part of their career pathway. Research is
poorly conceptualized by many and often never formally dis-
cussed during the training process. Interestingly, recruiting to
scientific and academic research is currently a major problem,
with the main perceived disadvantage being a financial loss of
earnings. However, the aims of clinical research are primarily to
allow an understanding of scientific methodology and practice,
such that the same principles can be applied to clinical medicine
with the significant advantage of enhanced patient care. Through
personal experience and feedback from others, one of the striking
features of basic scientific research is that the underlying phi-
losophy is invariably different to clinical practice and many clini-
cians can become easily disillusioned. Many health professionals
experience a “culture shock” as they move into a research envi-
ronment. This is due to several factors; some practical and some
more fundamental. Simply speaking, medical professionals are
usually thrown in at the deep end and expected to “swim”. It is
evident that the value of basic knowledge and support at the
beginning of the research period is therefore vital, to build a
suitable platform to carry out good quality research. In this book,
we have tried to cover the main areas in research, allowing
anyone to set up and complete research projects in clinical
research as well as in basic science research.
We have endeavored to advise on the common basic science
techniques that are currently popular in research, providing new
information, as well as updating areas that may be familiar to
most health professional. A host of international authors have
provided their unique perspective in their field of specialist inter-
est, representing institutions from around the world.
The book has three sections. Section A considers study design
and research governance which are core subjects to understand
viii PREFACE

before embarking on any project. Section B concentrates on basic

science techniques. Modern laboratory techniques have been
covered for laboratory-based research. These should provide
enough detail to allow readers to take up their technique with
confidence, as well as allowing a reference point to more experi-
enced researchers. The final section tackles data analysis and the
presentation of the results, both in the oral and written format.
We hope this simple but informative text inspires and gives
confidence to all people considering undertaking high-quality
research.

Enjoy!

Hiten Patel
Manit Arya
Iqbal Shergill
Contents

Foreword by Professor Sir Ara Darzi . . . . . . . . . . . . . . . . . v

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

1. Research Governance . . . . . . . . . . . . . . . . . . . . . . . . . . 1
S.J. Vyas, M. Arya, I.S. Shergill, and H.R.H. Patel

2. Designing Health Studies . . . . . . . . . . . . . . . . . . . . . . . 8

Rumana Z. Omar, Julie A. Barber, and Gareth Ambler

3. Immunohistochemistry . . . . . . . . . . . . . . . . . . . . . . . . 18
Philippa Munson

4. Cell Culturing: A Beginner’s Guide to Understanding

the Basics of Cell Culturing . . . . . . . . . . . . . . . . . . . . . 31
Khurshid Alam, Edwin T. Anthony, P.N. Vaiude,
Faruquz Zaman, and Harshad A. Navsaria

5. Flow Cytometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
P. Erotocritou, M. Arya, S.N. Shukla, and
H.R.H. Patel

6. Western, Northern, and Southern Blotting . . . . . . . . 48

Stephan Schlickeiser and Uwe Pleyer

7. Fluorescent In Situ Hybridization . . . . . . . . . . . . . . . 58

Fiona Campbell and John M.S. Bartlett

8. Quantitative Reverse Transcriptase Polymerase

Chain Reaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Lyndon M. Gommersall, M. Arya,
Prabhabhai S. Patel, and H.R.H. Patel

9. Proteonomics: High-Throughput Structural

Biology—Methods for Cloning, Protein Expression,
and Purification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
William K. Gillette and James L. Hartley
x CONTENTS

10. DNA and Tissue Microarrays . . . . . . . . . . . . . . . . . . . . 98

Maysa M. Abu-Khalaf, Lyndsay N. Harris, and
Gina G. Chung

11. Basic Scientific Techniques in Recording

Cellular Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
George Z. Mentis, Yoshiyasu Arai, and
Michael J. O’Donovan

12. Presenting and Publishing Research Data . . . . . . . . . 117

Howard A. Bird

13. Analyzing Health Studies . . . . . . . . . . . . . . . . . . . . . . . 126

Julie A. Barber, Gareth Ambler, and
Rumana Z. Omar

14. Future . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

H.R.H. Patel, I.S. Shergill, and M. Arya

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Contributors

Maysa M. Abu-Khalaf, MBBS Queen Mary University of

Internal Medicine – Medical London
Oncology London, UK
Yale University School of
Yoshiyasu Arai, MD, PhD
Medicine
Lab of Developmental
New Haven, CT, USA
Neurobiology Section
Khurshid Alam, MBBS, MRCS, National Institute of
MSc Neurological Diseases and
Centre for Cutaneous Stroke
Research National Institutes of Health
Institute of Cell and The Porter Neuroscience
Molecular Science Center
St. Bartholomew’s and The Bethesda, MD, USA
London School of Medicine M. Arya, FRCS
and Dentistry Molecular Uro-Oncology
Queen Mary University of Research
London Fellow and Specialist
London, UK Registrar
Gareth Ambler, PhD Prostate Cancer Research
Department of Statistical Centre
Science University College London
University College London London, UK
London, UK Julie A. Barber, PhD
Department of Statistical
Edwin T. Anthony, MRCS,
Science
MB, Bch, BA, BAO
University College London
Centre for Cutaneous
London, UK
Research
Institute of Cell and John M.S. Bartlett
Molecular Science Endocrine Cancer Research
St. Bartholomew’s and The Group
London School of Medicine Western General Hospital
and Dentistry Edinburgh, UK
xii CONTRIBUTORS

Howard A. Bird, MA, MD, Lyndsay N. Harris, MD

FRCP Yale Cancer Center – Medical
Clinical Pharmacology Oncology
Unit Yale University School of
University of Leeds Medicine
Chapel Allerton Hospital New Haven, CT, USA
Leeds, UK
James L. Hartley, PhD
Fiona Campbell Protein Expression
Endocrine Cancer Research Laboratory
Group Research Technology
Western General Hospital Program
Edinburgh, UK SAIC-Frederick, Inc.
NCI-Frederick
Frederick, MD, USA
Gina G. Chung, MD
Yale Cancer Center – Medical George Z. Mentis, PhD
Oncology National Institute of
Yale University School of Neurological Diseases and
Medicine Stroke
New Haven, CT, USA National Institutes of Health
The Porter Neuroscience
Center
P. Erotocritou
Bethesda, MD, USA
Institute of Urology
University College Hospital
London, UK Philippa Munson, BSc, PhD
Department of Histopathology
University College London
William K. Gillette, PhD London, UK
Protein Expression
Laboratory
Research Technology Harshad A. Navsaria, BSc,
Program MSc, PhD
SAIC-Frederick, Inc. Centre for Cutaneous
NCI-Frederick Research
Frederick, MD, USA Institute of Cell and
Molecular Science
St. Bartholomew’s and The
Lyndon M. Gommersall, London School of Medicine
MBBS, MRCS and Dentistry
Institute of Urology Queen Mary University of
University College Hospital London
London, UK London, UK
 CONTRIBUTORS xiii

Michael J. O’Donovan, BA, Institute of Pharmacy

MB, ChB, PhD Charite
Lab of Developmental Humboldt University
Neurobiology Section Berlin, Germany
National Institute of
Neurological Diseases and
Stephan Schlickeiser, PhD
Stroke
candidate
National Institutes of Health
Department of
The Porter Neuroscience
Ophthalmology
Center
Institute of Medical
Bethesda, MD, USA
Immunology
Berlin, Germany
Rumana Z. Omar, PhD
Department of Statistical
Science I.S. Shergill, BSc, MRCS
University College London Clinical Research Fellow in
London, UK Molecular Uro-Pathology
and Specialist Registrar in
H.R.H. Patel, BMSc.Hons, BM, Urology
BCh, MRCS, PhD, Institute of Urology
FRCS(Urol), FRCS(Eng.Hons) University College London
Consultant Laparoscopic London, UK
Surgeon
Head
S.N. Shukla, MBBS
Section of Laparoscopic
Deputy Director
Urology
Basic Science Research
University College Hospital
Gujarat Cancer and Research
London, UK;
Institute
University of Rochester
Civil Hospital
Medical Center
Amedabad
Rochester, NY, USA
Gujarat, India
Prabhabhai S. Patel, PhD
Head P.N. Vaiude, MBBS, MRCS
Basic Science Research Centre for Cutaneous
Gujarat Cancer and Research Research
Institute Institute of Cell and
Civil Hospital Molecular Science
Amedabad St. Bartholomew’s and The
Gujarat, India London School of Medicine
and Dentistry
Uwe Pleyer, MD Queen Mary University of
Department of London
Ophthalmology London, UK
xiv CONTRIBUTORS

S.J. Vyas, MS, FRCS(Edin), Institute of Cell and

DNB, MNAMS, Molecular Science
M. Med(Surg) St. Bartholomew’s and The
Upper GI/HPB Surgery London School of Medicine
University College Hospital and Dentistry
London, UK Queen Mary University of
London
Faruquz Zaman, MBBS, London, UK
MRCS
Centre for Cutaneous
Research
Chapter 1
Research Governance
S.J. Vyas, M. Arya, I.S. Shergill, and H.R.H. Patel

INTRODUCTION
The Department of Health (DOH), United Kingdom, regulates
the conduct of medical practice in the country. Its scope of action
extends beyond the same, as it also defines and formulates crite-
ria pertaining to performing research activity. Indeed, research
governance (RG) is more like research regulation, and the DOH
would largely assume the role of the regulation.
Research governance oversees a broad range of regulations,
principles and standards of good practice that exist to achieve,
and continuously improve, research quality across all aspects of
health care in the United Kingdom and worldwide.1
The DOH is responsible for setting health and social care
policy in England. The department’s work sets standards and
drives modernization across all areas of the National Health
Service (NHS), social care, and public health. As well as policy
documents, the DOH also issues guidance on implementation of
policies. The research governance framework (RGF) for health
and social care defines the broad principles of good RG and is
key to ensuring that health and social care research is conducted
to high scientific and ethical standards. The first issue of the RGF
was issued in March 2001 and a later updated in April 2005.2,3
These guidelines and principles of RG apply to everyone con-
nected to health-care research, whether a chief investigator, care
professional, researcher, their employer(s), or support staff. This
list includes any health-related research which involves humans,
their tissues, and/or data.1
Examples of such research include the following:

• Analysis of data from a patient’s medical notes

• Observations
• Conducting surveys
• Using noninvasive imaging
• Using blood or other tissue samples
2 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

• Inclusion in trials of drugs, devices, surgical procedures, or

other treatments

It is now important ethically and medicolegally that any indi-

vidual/group/department or institution involved in any kind of the
above mentioned research activity be aware of these guidelines
and follow any kind of healthcare research process. It therefore
becomes their responsibility and obligation to be well informed
of the protocols before embarking on any such exercise. There-
fore, institutions would implement their RG procedures which
would govern and monitor the research process and identify any
violations of the set system of procedure by audit cycles.
The Internet/World Wide Web has hundreds of examples of
RG and policy documents. Each university/academic institution
and government body (e.g., General Medical Council, Medical
Research Council, National Health Service) has its own rule sets
regarding the type of research activity that takes place within its
institution. While there will be subtle differences in each of these,
the basic principles of RG embedded in the constitution of each
of these institutions/organizations would be similar.

WHY WE NEED RESEARCH GOVERNANCE

Research governance is needed for the following purposes:1,2
• Safeguard participants in research
• Protect researchers/investigators (by providing a clear frame-
work to work within)
• Enhance ethical and scientific quality
• Minimize risk
• Monitor practice and performance
• Promote good practice and ensure lessons are learned
In broad terms RG ensures that health and social care
research is conducted to the highest scientific and ethical stan-
dards. The legal implications of the RGF would apply to everyone
involved in research, health, or social care setting using NHS
resources or involving NHS patients.

HISTORY OF DEVELOPMENT OF RESEARCH GOVERNANCE1,2

It was as early as the 1930s that the seeds of RG were sown. There
was a mistake in the formulation of a children’s syrup in the
United States, which caused a number of deaths. Waking up to
this call the US Food and Drug Administration (FDA) thought it
necessary to issue guidelines and bring about tight regulation of
the healthcare industry.
 1. RESEARCH GOVERNANCE 3

The Nuremburg Code (1947) was one of the first attempts to

regulate the ethics of medical research. It was written shortly
after World War II, following revelations at the Nuremberg Trials
that unethical research was carried out by certain members of
the medical profession during the Nazi period in Germany. The
code has 10 requirements and begins with the now widely rec-
ognized principle that voluntary consent of human participants
in research is paramount.
The code has since been superseded by documents such as
the Declaration of Helsinki (1964 ), Good Clinical Practice (1996–
1997), and the European Union Directive on Good Clinical Prac-
tice issued in 2005 (2005/28/EU ).4,5
There have since been a number of directives issued to regu-
late healthcare research such as the following:

2001: European Union Directive on Clinical Trials (2001/20/EC)6

2001: Research Governance Framework for Health and Social
Care (UK)
2004: Medicines for Human Use (Clinical Trails) Regulations (UK)
2004: Human Tissue Act 2004 (UK)
2005: Mental Capacity Act 2005 (UK)

All these acts and guidelines, as stated above, are laws by

themselves and need to be complied with as part of any research
being undertaken. Failure to do so may have medico-legal implica-
tions, and hence it is imperative that all participating bodies are
aware of the legal obligations within the existing framework.
The European Union Directive on Clinical Trials Act 2001
(2001/20/EC) is a good example.4,5,6 This is a legal document
which sets out how the system of procedure in the conduct of
clinical trials evaluating the safety and efficacy of a medicinal
product in humans should be performed. The directive was intro-
duced to simplify and harmonize the administration of clinical
trials across the European Union by establishing clear, transpar-
ent procedures. More importantly the directive ensures unifor-
mity in conduct of clinical trials across member states and hence
ensures uniformity in the conduct of these multicentred trials
across different nations. From this, RG requirements, in relation
to investigational medicinal products, have evolved in response
to a number of different factors:

1. Mistakes or problems with medicinal products historically

have increased the need for product regulation
4 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

2. Abuses of humans rights and incidents of fraud have increased

the need for tighter ethical controls (e.g., Tuskegee Syphilis
Study)2
3. A divergence in regulations and guidance in different coun-
tries has caused duplication of research, raising concerns
about the:
• Cost of clinical trials
• Ethical implications of repeating studies
• Need to rationalize and harmonize RG requirements
Obviously to safeguard research and take care of the ethical
issues involved in research a RGF has been set up. The RGF
outlines the principles of good governance for all research within
the remit of the Secretary of State for Health. This includes clini-
cal and nonclinical research (Table 1.1).3

TABLE 1.1. RGF basic principles and guidelines (not exhaustive)3

Standard for health

Personnel who are care covered under
directly or indirectly following five
Basic criteria affected by the RGF domains

Promotion and All personnel involved Ethics

protection of in social care and
public health public health at
primary, secondary,
and tertiary levels
of health and allied
care
Undertaken by NHS, Managers and staff of Science
nongovernmental all professional
health agencies groups irrespective
and Department of the level of
of Health seniority
Undertaken by, For all those involved Information
within, or on in clinical research,
behalf of social directly or indirectly
care agencies
Sets out principles, Health, safety, and
requirements, employment
and standards,
and defines
mechanisms to
deliver them
Improves research Finance and
and safeguards intellectual
the public property
 1. RESEARCH GOVERNANCE 5

SALIENT FEATURES OF THE RESEARCH

GOVERNANCE FRAMEWORK
The following summary depicts the RGF guidelines more
specifically: 3
Research involving patients in general and NHS patients in
particular should ensure that the ethical guidelines are met with
and the basic ethical issues are complied with. The research has
to be approved by the Hospital Ethical Review Committee.

1. Complete informed consent of the participants should be

obtained and this is the obligation of the Ethical Review
Committee to ensure that it is done in a just and transparent
manner.
2. Provisions should be available to the participants to
withdraw themselves from the research process, whenever they
want to.
3. Respect the multi-cultural nature and diversity of human
society and conditions.
4. Ensure patient confidentiality and the confidentiality of
data.
5. Take into account the risk to participants and provision
for adequate compensation wherever necessary.
6. Base current research in the light of the existing evidence.
Hence, peer reviews and the strength of the current evidence
should be taken into consideration when planning research.
7. Research should be peer reviewed by experts in the field
from time to time.
8. Findings and data should be made available to independent
bodies and individuals for review from time to time.
9. Freedom of information: findings should be made
available to people who may benefit from such research, general
public at large and the research should be communicated in
professional channels through appropriate channels and should
be open to criticism.
10. Health and safety of participants should be assured at all
times.
11. Financial and intellectual property rights must be
respected at all times.
12. Strict guidelines for the recording and reporting of
adverse outcomes exist and these must be maintained as per the
guidelines laid down in the standard operating procedures
(SOP).
13. Clinical trials involved in animal and human research
must be registered with a central registering authority and their
findings should be accessible to the public.
6 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

The RGF is incorporated into the constitution of the DOH.

They are a leader in the field of implementing standards of RG in
the field of health and social care. As of April 2004, the responsibil-
ity of ensuring that adequate standards are maintained and kept
up, were passed on to the Strategic Health Authority, who ensure
tight quality control and implementations of the research regula-
tions in the NHS.7 Thus RG has become a part of the NHS quality
agenda. The framework of the RG is closely allied with the frame-
work for clinical governance. Individual NHS bodies now report
directly to the DOH on research and clinical governance issues.
As a part of the national implementation plan for the United
Kingdom, a network of research management and governance
offices was established in April 2003.These offices ensure main-
taining and compliance of the research procedures in primary,
community, and social care. The responsibility for the national
RG standards and policy will remain with the United Kingdom
DOH. There is also a control assurance standard for RG that has
been commissioned to assure quality control in RG.8

CONCLUSION
Regulation of research has evolved considerably since medieval
ages, when most research and discovery was less controlled, but
the product of relentless persuasion and thinking of the human
mind. This applies mostly to medical discoveries such as admin-
istration of the first anesthetic (Lister), penicillin (Fleming), and
small pox vaccine. One wonders if such strict regulations existed,
whether it would have been possible to discover these landmark
medical innovations. We can believe that had mankind not taken
these risks then, we would not have seen so many advancements
in modern medicine as we see today. Alternatively, we have
become defensive as we protect ourselves from the speculated
and feared side effects of new discovery. It is more like a fear
and the insecurity of the unknown. However, regulation of
research is equally important so as to direct and organize it while
providing maximum protection to the participants and not com-
promise their safety in today’s day and age.

References
1. www3.imperial.ac.uk/clinicalresearchoffice/researchgovernance
2. www.dh.gov.uk/PublicationsAndStatistics/Publications/Publications
PolicyAndGuidance
3. Research Governance Framework for Health and Social Care, 2nd ed.
Publication of the Department of Health, United Kingdom.
4. Guidelines for Good Clinical Practice- ICH Guidelines.
 1. RESEARCH GOVERNANCE 7

5. EUCTD (2001/20/EC). Official Journal of European Communities L L

121/34–L121/44, European Union Directive on Clinical Trials
6. European Union Directive on Good Clinical Practice (2005/28/EC).
7. www.mrc.ac.uk/index/current-research
8. www.controlassurance.co.uk
Chapter 2
Designing Health Studies
Rumana Z. Omar, Julie A. Barber, and Gareth Ambler

Study design is a fundamental aspect of research. If the design

of a study is poor, no amount of clever analysis will provide reli-
able results. The results from poorly designed studies could be
meaningless, and many resources will have been wasted, not to
mention the possible risk to the subjects, which would be unethi-
cal. Therefore, it is essential that researchers invest adequate
time and effort in designing their studies appropriately. It would
be advisable to involve a medical statistician or an epidemiologist
at the design stage of a study.
This chapter is divided into two parts. In the first part, the
important aspects of study design are discussed. The second part
focuses on the types of studies that are commonly encountered
in health research.

ESSENTIALS OF A STUDY DESIGN

Specifying the Research Question

The first step should be to provide a clear research question, for
example, the hypothesis to be tested. The research question
should be justified on the basis of a thorough literature review
of relevant previous research.

Selection of Subjects
Study subjects should be selected to ensure that they are repre-
sentative of the population to which the results of the study will
be applied (target population). For example, studies including
only patients referred to one particular hospital or volunteers
may not always provide a representative sample of a more general
population. The source of subjects and the inclusion and exclu-
sion criteria need to be clearly defined.

Specifying the Primary Outcome

It should be decided in advance which outcome measure is of
major interest. The analysis of this outcome should be used to
 2. DESIGNING HEALTH STUDIES 9

provide the study’s main conclusions. Information on other out-

comes could be collected, but these should be considered to be
of secondary importance. Any interesting findings among the
secondary outcomes should be interpreted cautiously, possibly
as ideas for further research than as definitive results.

Inclusion of a Control Group

For studies investigating disease associations, it is not enough
simply to consider what happens to subjects who are exposed (for
example, to a treatment or risk factor); we need also to know what
happens to subjects who are not exposed. This comparative role
is undertaken by a control group of subjects. A control group is a
set of subjects who either have not been exposed to the risk factor
or treatment under investigation, or do not have the outcome of
interest, depending on the type of design used. Subjects could also
act as their own controls—for example, drawing comparisons
between what happens before and after an exposure.

Confounding
In a study population, there could be differences in the charac-
teristics of the subjects, such as age and sex, which may affect
outcome, and which may also be related to the exposure of inter-
est. For example, in comparing the operative mortality between
two surgical techniques, the differences between the outcomes
of the two operations could be due to the procedures. It could
also be due to differences in the preoperative patient character-
istics, which could also affect the choice of the technique. In
statistical terms, the effects of operation and patient character-
istics are said to be confounded (see Figure 2.1). It is important

Operation type

Mortality Patient
characteristics

FIGURE 2.1. Confounding parameters when comparing two operative

groups.
10 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

to identify potential confounders and make plans for dealing

with these at the design stage of a study.
Confounding may be controlled through study design or in
analysis. Two common methods for controlling confounding at
the design stage are matching and randomization. In matching
subjects in the groups under comparison are matched on factors
that are considered as potential confounders. Randomization
uses random allocation to assign patients to comparison groups
to ensure that patients in the different groups only differ in their
characteristics by chance. In analysis, one could adjust for con-
founding variables using statistical techniques such as stratifica-
tion or multiple regressions. In order to do this, it is important
to plan at the design stage what information should be collected
on potential confounders.

Sample Size
When planning a study it is important to estimate the number
of subjects required. Otherwise it could be impossible to tell if
a study has a good chance of producing worthwhile results.
In general, the larger a study is the greater its power and preci-
sion. Power is concerned with testing for effects (for example a
difference between two treatments) and is defined as the proba-
bility that a study will be able to detect a clinically important
effect if it really exists. Precision of a sample estimate is deter-
mined by the width of a confidence interval for the estimate.
Statistical formulae are available to calculate sample size for
a study.1

Bias
The presence of bias in a study may affect the validity of its find-
ings. Steps should be taken at the design stage to avoid bias. The
most commonly occurring biases are described below:

Selection bias: Stems from an absence of comparability between

groups being studied or nonresponse. This could be avoided
by selecting comparable groups, which are representative of
the target population and taking effective strategies to mini-
mize or handle missing data.
Assessor/response bias: If the assessor or respondent is aware of
the exposure and/or disease status of the subjects this may
influence their assessment/response. The exposure and disease
conditions should be concealed from the assessors and respond-
ers if possible to avoid this type of bias. This is known as
blinding.
 2. DESIGNING HEALTH STUDIES 11

Recall bias: For retrospective studies, subjects may find it difficult

to provide accurate information on past exposures. The ability
to recall may vary between subjects who have the disease and
those who do not. This is particularly the case for exposures
for which information is not usually recorded—for example,
dietary habits. It is important to phrase questions carefully to
avoid recall bias.

Writing a Protocol
A study design should start with the writing of a protocol. It
should lay out systematically the various stages of the study
described above and include strategies for data collection and
achieving completeness, data entry, storage and validation of
data, a broad statistical analysis plan, and the responsibilities of
the study personnel. An analysis plan ensures that analysis for
the study is performed in an objective way, thus avoiding data
dredging, which may show spurious relationships.

TWO TYPES OF STUDIES

Health research studies may be broadly divided into two types:
randomized controlled trials and observational studies.

Randomized Controlled Trials

In randomized controlled trials, a health intervention is planned
for a specific outcome, and its effectiveness for that particular
outcome is evaluated. Randomization is used to allocate patients
into the intervention and control groups. Further design consid-
erations for trials include the following factors.

Concealment of Random Allocation

When using randomization in trials, it is essential that the treat-
ment allocation system prevents the person entering patients
from knowing the next treatment allocation in advance. A
common way of doing this is to use a series of sealed opaque
envelopes, each containing a treatment specification. For drug
trials, the allocation may be carried out by the pharmacy, which
produces numbered bottles that do not indicate the treatment
contained. In many large multicenter studies, patients are
enrolled by telephoning a central office.

Blinding
Blinding is a technique used to minimize response bias. By blind-
ing treatment allocation from both the patient and assessors, it
is possible to eliminate response bias—this is known as double
12 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

blinding. Where it is only possible to blind either the assessor or

patient, the study is single blinded. It is important to achieve the
maximum degree of blindness when designing a trial.

Placebos
If there is no existing standard beneficial treatment, then it is
reasonable to give the control group placebos instead of any
active treatment. Placebos are identical in appearance to the
active treatment, but are pharmacologically inactive. Placebos
are used because the act of taking a treatment may itself have
some benefit to the patient, so that part of any benefit observed
in the treatment group could be due to the knowledge/belief that
they had taken a treatment. Moreover, for a study to be double
blind, it is necessary for the two treatments to be indistinguish-
able. When the control treatment is an alternative active treat-
ment rather than a placebo, the different treatments should still
be indistinguishable, if possible.

Protocol Violations
A common problem relates to patients who have not followed the
protocol—for example patients who receive the wrong treatment
or do not take their treatment, known as noncompliers. If this
does occur, it is advisable to keep all randomized patients in the
trial and analyze groups as randomized. This is known as an
intention-to-treat analysis.

Types of Trials

Parallel Group Trials

In the simplest type of trial design, one group of subjects receives
the planned intervention and is known as the intervention group.
A group of subjects does not receive the planned intervention and
is known as the control group. The outcome is compared between
the two groups. Subjects are allocated to the intervention and
control groups using randomization. This ensures that each subject
has an equal chance of being allocated to either group, and groups
differ only with respect to their intervention. An example is a trial
where patients with type 2 diabetes were randomized to the angio-
tensing converting enzyme (ACE) inhibitor Ramipril or placebo
(on top of standard treatment). The outcome of interest is the
occurrence of cardiovascular events in patients.2 This is the most
frequently used design for trials.

Crossover Trials
The most common alternative is the crossover trial, in which all
patients are given both the intervention and control treatments
 2. DESIGNING HEALTH STUDIES 13

in a sequence. Here randomization is used to determine the

sequence of the treatments. For example, a crossover trial was
carried out to evaluate the effectiveness of oral morphine for the
management of refractory dyspnoea.3 Patients were randomized
to receive four days of oral morphine, followed by four days of
identically formulated placebo, while the other half received
placebo followed by morphine.

Strengths and Limitations of Parallel and Crossover Trials

In a crossover trial, the variability is less as the comparison is
within subjects and hence a smaller sample size is needed.
However, this design is only suitable when:

• Treatment periods are fairly short to minimize the risk of drop

out for other reasons
• Conditions are chronic and cannot be cured
• There is no carry over of effect of intervention from one period
to the next. It may be possible to have a wash-out period between
the intervention periods, to reduce the risk of carry over

Observational Studies
In observational studies, associations between health outcomes
and exposure to risk factors or preventative measures are
observed in subjects without any planned intervention. Observa-
tional studies may be classified as 1) descriptive (includes case
report/series and cross-sectional) or 2) analytic (includes cohort
and case-control studies). Descriptive observational studies are
used to describe disease patterns and trends. Often these studies
are used to generate hypotheses and plan health programs. Ana-
lytic studies may be used to estimate or test relationships between
an exposure to a risk factor or a preventative measure, and a
health outcome.

Case Report/Series
A case report is a detailed profile of a single patient, reported by
one or more clinicians. For example, a report was published on
a 40-year-old woman who developed pulmonary embolism after
use of oral contraceptive.4
A case series is an expanded version of a case report that
includes a series of patients with a given health condition. For
example, a study was conducted on 12 children who had received
the measles-mumps-rubella (MMR) vaccine and were referred to
a gastroenterology unit and had a history of normal development
followed by a pervasive developmental disorder (PDD).5
14 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

An important point to note here is that case report/series

focuses on a single group of patients and crucially does not
include a control group.

Cross-Sectional Study
In this type of study design, all information on a group of sub-
jects is collected at a single time point. There is no typical format,
and each study may be designed to meet the need of the
researcher. This study design includes subjects who are exposed
to the risk factor or preventative measure under investigation, as
well as those who are not. The outcome of interest is compared
between the exposed and unexposed groups. This design is par-
ticularly suitable to estimate the prevalence of a disease and to
examine health trends. It may also be used to examine the asso-
ciation between a risk factor or a preventative measure and a
health outcome. An example is the Health Survey of England
conducted to monitor health trends, and estimate prevalence of
certain diseases and certain risk factors associated with this
outcomes.6

Strengths and Limitations of the Descriptive

Observational Studies
A case report/series does not include control groups and, hence,
cannot be used to investigate associations or causation between
exposure and outcome. These studies may only be used for
descriptive purposes. There are also issues about whether the
subjects included in a case report/series are representative of the
study target population. These types of studies could be used for
reporting or sharing experiences on rare health conditions.
A cross-sectional study has no rigid format and is therefore
more prone to bias. It is a simple design, which could be used
when resources are limited to give some idea about disease asso-
ciations. It is not suitable for studying disease causation, as it
provides no information about the order of events, i.e., whether
disease preceded or followed exposure.

Cohort Study
In a cohort study, a group of subjects is identified according to
the research objectives and followed over the study period, until
the subjects drop out, have the event of interest, or reach the end
of the study period. The event rate is compared between the
group of subjects exposed to the risk factor, or any preventative
measure under investigation, and the unexposed group. A cohort
study may be both prospective and retrospective. A retrospective
 2. DESIGNING HEALTH STUDIES 15

cohort study could be about patients with a past history of expo-

sure to certain factors, experiencing health conditions that were
also observed in the past.
An example of a cohort study is a study investigating the
association between stable partnership and development of AIDS
or death in HIV patients receiving highly active antiretroviral
therapy (HAART) treatment.7 In this study, patients were fol-
lowed from the first follow-up visit after receiving HAART to
their last follow-up visit within the study period or death or
development of AIDS. The conclusion from this study was that
stable partnership was associated with a slower rate of progres-
sion to AIDS or death in HIV patients receiving HAART.

Case-Control Study
Case-control studies are always retrospective. They examine how
exposure to retrospective factors contributes to current health
conditions. A group of subjects with the outcome of interest is
recruited according to some prespecified inclusion criteria
(cases). Another group of subjects who have not experienced the
outcome of interest is recruited as controls. Exposure to the risk
factor of interest is then compared between the cases and con-
trols. Cases and controls may be matched on important con-
founding characteristics. For example, a matched case-control
study was used to investigate the association between the MMR
vaccine and PDD.8 The exposure of interest was the MMR vaccine.
Subjects with a diagnosis of PDD, while registered with a general
practitioner (GP), were recruited as cases. Subjects with no diag-
nosis of PDD, matched on age, sex, and GP to the cases, were
recruited as controls. The conclusion was that there was no
association between MMR vaccine and PDD.

Strengths and Limitations of Analytic

Observational Studies
Cohort studies are particularly suitable for studying rare expo-
sure. For example, to investigate the association between long-
term chronic exposure to radiation and risk of cancer, a cohort
of workers from a nuclear plant may be recruited. It may also be
used to examine risk factors that change over time (temporal
relationships), estimate incidence rates of a disease, and study
multiple outcomes. However, this design may not be suitable for
rare outcomes, as it could be time consuming to observe all the
events required to achieve the desired sample size. Alternatively,
a large number of subjects may have to be recruited. Both of
these could prove to be expensive.
16 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

TABLE 2.1. Appropriateness of study design types

Study Objective Type of Study Design

Evaluating health intervention Randomized controlled trials

To study natural history of a Cohort and case-controls studies
disease
To examine disease associations
To estimate disease prevalence Cross-sectional studies
To examine health trends
To generate hypothesis
To report observation of rare Case-report/series
conditions for descriptive
purposes
To generate hypotheses

Case-control studies are particularly suitable for rare out-

comes. It is cheap and fast. However, they can be more prone to
bias than cohort studies, depending on the way the controls are
selected. For example, if volunteers are selected as controls, they
may not be representative of the study target population. Using
hospital controls could distort the effect of exposure if controls
are recruited from patients who are more likely to be exposed to
the risk factor of interest compared to the members of the target
population. For example, in a case-control study examining the
association between non-steroidal anti-inflammatories (NSAIDs)
and the development of gastric ulcer, patients in the gastroenter-
ology unit of a hospital are recruited as cases. Selecting controls
from the rheumatology unit of the same hospital would not be
appropriate, as they are likely to use more NSAIDs than the
general population.

Reliability of Different Study Designs

A properly conducted randomized trial is the most reliable study
design. Causal effect may only be directly inferred from such
studies. Drawing causal inference from analytic observational
studies is more difficult. Bradford Hills suggests several condi-
tions that need to be met before causal inferences could be drawn
from an analytical observational study.9 Cohort studies are
usually less prone to bias than case-control studies. The use of
case-report/series is very limited. The appropriateness of each
type of design for health studies is summarized in Table 2.1.

References
1. Machin D, Campbell M, Fayers P, Pinol A. Sample Size Tables for
Clinical Studies, 2nd ed. London: Blackwell Science; 1997.
 2. DESIGNING HEALTH STUDIES 17

2. Marre M, Lievre M, Chatellier G, Mann JFE, Passa P, Ménard J.

Effects of low dose ramipril on cardiovascular and renal outcomes in
patients with type 2 diabetes and raised excretion of urinary albumin:
randomised, double blind, placebo controlled trial (the DIABHYCAR
study). BMJ 2004;328:495.
3. Abernethy AP, Currow DC, Frith P, Fazekas BS, McHugh A, Bui C.
Randomised, double blind, placebo controlled crossover trial of sus-
tained release morphine for the management of refractory dyspnoea.
BMJ 2003;327:523–528.
4. Jordan WM. Pulmonary embolism. Lancet 1961;2:1146–1147.
5. Wakefield AJ, Murch SH, Linnell AAJ, et al. Ileal-lymphoid-nodular
hyperplasia, non-specific colitis, and pervasive developmental disor-
der in children. Lancet 1998;351:637–641.
6. Health Survey for England. http://www.dh.gov.uk/PublicationsAnd
Statistics/PublishedSurvey/HealthSurveyForEngland/fs/en.
7. Young J, Geest SD, Spirig R, et al. For the Swiss HIV Cohort Study
Group. Stable partnership and progression to AIDS or death in HIV
infected patients receiving highly active antiretroviral therapy: Swiss
HIV cohort study. BMJ 2004;328:15.
8. Smeeth L, Cook C, Fombonne E, et al. MMR vaccination and pervasive
developmental disorders: a case-control study. Lancet 2004;364:
963–969.
9. Rothman KJ, Greenland S. Causation and causal inference. In:
Modern Epidemiology. Rothman KJ, Greenland S, eds. Philadelphia:
Lippencott-Raven; 1998:7–28.
Chapter 3
Immunohistochemistry
Philippa Munson

INTRODUCTION
Immunohistochemistry describes the localization of antigens in
histological and cytological preparations using antibodies. It is
now recognized as an essential element and a major tool, both
in diagnostic and research-orientated cellular pathology. The
technique involves the detection of specific or highly selective
cellular epitopes with an antibody and appropriate labelling
system. Immunohistochemistry can be performed on cytological
preparations, frozen sections and paraffin-embedded histological
sections.
The first point of call for any researcher wishing to perform
immunohistochemistry should be the local immunohistoche-
mistry laboratory. Immunohistochemistry is a speciality that is
constantly changing and improving, and the local immunohisto-
chemistry laboratory will have more experience and more up-to-
date methods than most research laboratories. This should
therefore be done, even when in possession of an existing proto-
col from other researchers. It is also wise to refrain from pur-
chasing any reagents without first consulting the laboratory and
to ask the histology laboratory to perform the section cutting.
Although microtomy can be learnt in a few weeks, it is a skill
that takes years to perfect and poor section quality can severely
affect the interpretation of the immunohistochemistry.

BASIC IMMUNOHISTOCHEMISTRY
The basics of immunohistochemistry involve the use of an anti-
body to detect a specific epitope, which is then visualized using
a detection system and chromogen.
Fixation, be it with alcohol or formalin, will mask some anti-
gens to a certain extent. When this occurs, some form of antigen
retrieval will be needed to re-expose the antigen. This has to take
place before applying the primary antibody (Table 3.1).
 3. IMMUNOHISTOCHEMISTRY 19

TABLE 3.1. Glossary of immunohistochemical terms

Antigen A molecule that induces the formation of an

antibody
Epitope A single antigenic determinant (functionally it
is the portion of a antigen that combines
with antibody paratope)
Antibody A molecule produced in response to an antigen.
It has the property of combining specifically
with the antigen that induced its formation
Fixation The process of preservation that is necessary
for both cytological and histological
specimens
Taking sections The wax in the sections must be removed and
to water the sections brought through a gradient of
alcohols before immersing them in water
and performing immunohistochemistry.
Antigen retrieval Aka antigen unmasking epitope retrieval. This
is the use of enzymatic or heat-mediated
methods that “reverse” the effects of fixation,
enabling the antibody to combine with
the antigen
Peroxidase block If peroxidase is the enzyme used in the
detection system, any peroxidase in the
section must first be saturated with its
substrate (hydrogen peroxide) to prevent
endogenous staining
Primary antibody The first antibody to be applied to the section,
i.e., the antibody that identifies the antigen
under investigation
Secondary Aka bridge or link antibody. It attaches to the
antibody primary antibody. This may be labelled with
biotin, fluoroscein or another molecule or be
used in its unlabeled form
Tertiary layer The third (and usually final) layer that attaches
either directly to the secondary antibody (if
the secondary is unlabeled) or to the
molecule with which the secondary antibody
is labelled (e.g., biotin). The tertiary layer
will be labelled with an enzyme that works
with the chromogen to produce a colored
product
Chromogen The solution that enables a colored product to
be formed over the site of the antigen. The
chromogen is usually used in solution with
the substrate of the enzyme that has been
used in the tertiary layer NB. The secondary,
tertiary, and chromogen are collectively
known as the detection or labelling system.
(Continued)
20 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

TABLE 3.1. (Continued)

Counterstain The histochemical dye, usually hematoxylin,

that is applied to the section last, enabling
morphological identification of the tissue
components
Dehydrate, clear, The process by which sections are dehydrated
and mount through graded alcohols, cleared in xylene,
and mounted using a synthetic mountant

The basic steps of an immunohistochemistry protocol are:

fixation/processing/embedding
↓
section cutting/microtomy
↓
dewaxing sections and taking to water
↓
antigen retrieval
↓
peroxidase block
↓
primary antibody
↓
secondary antibody
↓
tertiary layer
↓
chromogen
↓
counterstain
↓
dehydrate, clear and mount sections

METHODOLOGY
Immunohistochemistry methodology starts with the process of
fixation. The length of time a sample spends in fixative is very
important as under- or over-fixation can lead to problems with
proteolytic antigen retrieval.
Samples should not be kept in fixative indefinitely (24 hours
is optimal for formalin). If a delay is experienced between sample
collections, they should be taken to the histology laboratory for
processing once they have spent 24 hours in fixative.
 3. IMMUNOHISTOCHEMISTRY 21

If the antibody of interest is going to be used on patient

samples, it is imperative to determine first that there is an avail-
able antibody that works on paraffin-embedded sections; other-
wise retrospective studies will be exceedingly difficult.

Fixation
Fixation is essential for tissue and antigen preservation. The
most important reactions that take place are those that stabilize
the proteins. The general principle is that the fixatives form
cross-links between proteins, thereby stabilizing the cytoskeletal
structure.
Formaldehyde (formalin) is the fixative of choice for routine
histology; therefore, any retrospective studies using patient
samples will involve the use of immunohistochemistry on
formalin-fixed, paraffin-embedded blocks. The aldehydes form
cross-links between protein molecules, the reaction being with
the basic amino acid lysine.
These cross-linking “methylene bridges” ensure that the
structures of intracytoplasmic proteins are not significantly
altered. However, they can also have the effect of “masking”
antigens, therefore, tissue that has been fixed in formalin will
generally require some form of antigen “unmasking,” i.e., antigen
retrieval.
Alcoholic fixatives are generally used for frozen sections or
cytological preparations, as they are poor penetrators of tissue.
They preserve most proteins in a relatively undenatured state,
although some antigen retrieval may be necessary. It is wise to
refrain from fixing frozen sections in glutaraldehyde or parafor-
maldehyde, as these will usually mask the antigen, and effective
antigen retrieval is very difficult to perform on frozen sections.

Antigen Retrieval
Antigen retrieval is the method by which antigens that have been
masked through fixation and processing are unmasked prior to
immunostaining. There are two main methods for antigen
retrieval: proteolytic and heat-mediated. Of these two, heat-
mediated antigen retrieval (HMAR) is the most effective on the
majority of antigens; however, the correct antigen retrieval
method must be determined for each antibody.
Proteolytic digestion generally utilizes trypsin, chymotrypsin,
protease, or proteinase K enzymes. The majority of these need
to be used at 37oC, with correctly prepared pH solutions. The
main pitfall of proteolytic digestion is that the digestion time has
to be tailored to the fixation time, i.e., the longer a tissue has
22 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

been in fixative, the more methylene bridges will have formed,

requiring a longer time in enzyme to break these down. If the
antigen under investigation is localized in the membrane, then
HMAR is generally preferable to enzyme digestion.
Heat-mediated antigen retrieval was first described by Shi
et al.1 (1991) and has proved to be a revolutionary technique as
many antigens previously thought to have been lost or destroyed
by fixation and processing can now be recovered and demon-
strated;1 however, the theory behind HMAR methods remains
unclear.
There are two major factors in HMAR: the antigen retrieval
solution and the cooking method used. A number of different
retrieval solutions with different pHs are available. The most
commonly used are citrate (pH 6.0) and some form of Tris-EDTA
(pH 9.0). These retrieval solutions can be used in microwave
ovens, pressure cookers, decloakers, or on some automated
immunostainers. Any of these cooking methods are acceptable,
provided they are performed and used in a consistent manner.
HMAR methods do require strong adhesives on the slide to
prevent section detachment. The slides must then be thoroughly
drained before being heated to at least 60oC for at least one
hour.
The main advantage of HMAR over proteolytic digestion is
that heating times to retrieve antigens tend to be uniform, regard-
less of the amount of time spent in fixative.2 This is in contrast
to the variability in digestion times required when using
enzymes.
The main pitfall with HMAR is that extreme care must be
taken not to allow the sections to dry, as this destroys antigenic-
ity. Sections are particularly susceptible to drying when being
removed from a hot solution, therefore, the solutions should be
flushed from the container with cold running tap water. The
slides can then be removed when the fluid is cool.

Antibodies
The vast majority of primary antibodies available for use on
human tissue are made in either rabbits or mice. It is essential
to check (before purchasing) that the required primary antibody
is available that works on formalin-fixed, paraffin-embedded
tissue. Again, liaison with the local immunohistochemistry labo-
ratory will be helpful. Generally speaking, monoclonal murine
antibodies are preferable to polyclonal antibodies, as they tend
to be more specific. See Animal Tissue section for more specific
guidelines on selecting antibodies for tissue other than human.
 3. IMMUNOHISTOCHEMISTRY 23

Antigens and epitopes: An antigen can be defined as a mole-

cule (protein, carbohydrate, or lipid) that binds with an antibody.
One antigen is composed of a number of epitopes or antigenic
determinant groups. An epitope consists of a small amino acid
sequence and which is what binds to the variable region of the
antibody. Due to differences in their manufacture, a monoclonal
antibody will only recognize one epitope on an antigen whereas
a polyclonal antibody will recognize many epitopes on an antigen.
One analogy is that of a Christmas tree and lights. If the antigen
is the tree and the epitopes are the lights, then a monoclonal
antibody will only recognize the red lights on a particular tree,
whereas a polyclonal antibody will recognize all the colors of
lights on that particular tree.

Staining Methods/Detection Systems

There is a wide variety of systems available today, but the best
and most reliable of these are the avidin-biotin and polymer-
based systems. Both of these methods are available in prediluted
forms, which are generally preferable where immunohistochem-
istry is not performed on a regular basis.

Avidin-Biotin Methods
These methods were first described by Heggeness and Ash3 (1977)
and utilize the high affinity of the glycoprotein avidin for biotin,
a low molecular weight vitamin.3 Avidin is present in egg white
and is composed of four subunits that form a tertiary structure
possessing four specific binding sites for biotin. Egg-white avidin
contains some oligosaccharide residues that possess an affinity
for some tissue components. As a result, a similar molecule,
streptavidin (extracted from the culture broth of the bacterium
Streptomyces avidinii), is generally used as this molecule does not
contain the oligosaccharide residues.
Biotin (vitamin H) can be conjugated to both antibody and
enzyme molecules. Up to 200 molecules of biotin can be conju-
gated to one antibody molecule, often with the aid of spacer
arms. By spacing the biotin molecules, streptavidin is given room
to bind and is able to maximize its strong affinity for biotin.
In a streptavidin-biotin complex, the streptavidin and biotinyl-
ated enzyme are supplied as two separate reagents, which need
to be added together 30 minutes before use. The streptavidin can
be added in slight excess so that the biotinylated enzyme does
not saturate all of the biotin-binding sites. Either peroxidase or
alkaline phosphatase can be used as the enzyme label.
24 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

STREPTAVIDIN + BIOTINYLATED
PEROXIDASE

BIOTINYLATED RABBIT ANTI -

MOUSE SECONDARY
ANTIBODY

PRIMARY ANTIBODY (MOUSE ANTI-HUMAN)

ANTIGEN

FIGURE 3.1. Primary streptavidin-biotin complexing.

In a labelled streptavidin technique, the streptavidin molecule

is directly labelled with the enzyme of choice. This is the tech-
nique favored by companies producing prediluted reagents, as
the streptavidin-biotin complex is not stable for long periods.
Avidin-biotin techniques provide high sensitivity as the high
affinity of streptavidin for biotin enables, firstly, a very stable
complex to be formed and, secondly, many enzyme molecules to
be deposited at the antigenic site (Figures 3.1 and 3.2).

PEROXIDASE-LABELLED STREPTAVIDIN

BIOTINYLATED RABBIT ANTI -

MOUSE SECONDARY
ANTIBODY

PRIMARY ANTIBODY (MOUSE ANTI-HUMAN)

ANTIGEN

FIGURE 3.2. Secondary streptavidin-biotin complex labeling.

 3. IMMUNOHISTOCHEMISTRY 25

Polymer-Based Methods
Polymer-based methods can be either a two- or three-layer
system. In the two-layer system, the secondary antibody is part
of the polymer molecule. The secondary antibody is conjugated
to the polymer as are a large number of enzyme molecules. The
EnVision kit, available from Dako, U.K., is an example of a two-
layer polymer system. Vyberg and Neilsen4 (1998) reported com-
parable sensitivity between the Dako kit and a three-stage
avidin-biotin system (Figure 3.3).4
In the three-layer system, the secondary antibody is applied
unconjugated, then the tertiary antibody which is conjugated to
the polymer along with enzyme molecules. The Novolink Polymer
kit, available from Novocastra Laboratories, U.K., is an example
of a three-layer polymer system (Figure 3.4).
The main advantage of polymer-based systems is that
they can be used on tissues containing a lot of endogenous
biotin without producing background staining. (See Background
section.)

GOAT ANTI-MOUSE ANTIBODIES AND PEROXIDASE MOLECULES

CONJUGATED TO DEXTRAN MOLECULE

PRIMARY ANTIBODY (MOUSE ANTI-

HUMAN)

ANTIGEN

FIGURE 3.3. Two layer polymer-based labeling.

26 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

POLYMER + PEROXIDASE
MOLECULES

GOAT ANTI-MOUSE ANTIBODY

PRIMARY ANTIBODY (MOUSE ANTI-

HUMAN)

ANTIGEN

FIGURE 3.4. Three layer polymer based labeling.

Enzyme Labels and Chromogens

Enzymes are the most widely used labels in immunohistochem-
istry, and incubation with a chromogen using a standard histo-
chemical method produces a stable, colored reaction end product
suitable for the light microscope. In addition, the variety of
enzymes and chromogens available allow the user a choice of
color for the reaction end product.
Horseradish peroxidase labelling + DAB (3,3′diaminobenzi-
dine tetrahydrochloride) were first described by Nakane and
Pierce in 1966,5 and this is still the most commonly used combi-
nation of enzyme and chromogen in immunocytochemistry. DAB
precipitates to a brown reaction end product when in the pres-
ence of peroxidase and hydrogen peroxide (hydrogen peroxide is
in solution with the DAB). A by-product of the reaction of hydro-
gen peroxide (the substrate) with peroxidase (the enzyme) is an
oxygen radical, which acts on DAB and precipitates it at the
antigenic site.
Some endogenous pigments (e.g., melanin, lipofuchsin, hae-
mosiderin, and formalin pigment) can mimic the appearance of
DAB, and in these situations, it may be preferable to use another
chromogen, e.g., 3-amino-9-ethylcarbazole (AEC), which pro-
duces a red end product.6,7
 3. IMMUNOHISTOCHEMISTRY 27

Alternatively, a different enzyme can be used in the detection

system. Alkaline phosphatase is the most widely used alternative
to peroxidase and can be developed using a number of chromo-
gens, most usually Fast Red, Fast Blue, or NBT/BCIP (nitro blue
tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate). Alkaline
phosphatase can also be useful when the target tissue contains
a lot of endogenous peroxidase, e.g. tissue containing red blood
cells, etc. (See Background section.)

Animal Tissue
When using animal tissue, the detection system components
need to be chosen with care. First, the primary antibody must be
able to recognize that particular species; this information will be
on the specification sheet. The secondary antibody then needs to
recognize the primary antibody without cross-reacting with the
host tissue. For example, if the host tissue is mouse, the primary
antibody should be raised in another animal, e.g., rat. The sec-
ondary antibody will then be raised in another animal, e.g.,
rabbit, and be anti-rat.
Occasionally, the only available primary antibody will be
raised in the same species as the host tissue. In these circum-
stances, there are commercial kits available to facilitate these
staining procedures.

TROUBLESHOOTING/OPTIMIZATION

Optimizing Primary Antibodies

The two main factors involved in optimizing a primary antibody
are antibody dilution and antigen retrieval. Guidelines may be
given on the antibody specification sheet, however, if none is
available, one approach is to pick a starting dilution (e.g., 1/50)
and test known positive tissue using a number of different
retrieval methods. A negative control must always be included
when optimizing a new antibody to ensure any staining seen is
appropriate.

Background
The major causes of background staining in immunocytochem-
istry are hydrophobic and ionic interactions and endogenous
enzyme activity. Hydrophobicity is a property shared by most
proteins and confers stability on the tertiary structure of pep-
tides. It may also take place between different protein molecules
and impart stability to immune complexes.
28 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

Proteins are rendered more hydrophobic by aldehyde fixa-

tion, and the extent of hydrophobic cross-linking of tissue pro-
teins is primarily a function of fixation. Therefore, factors such
as time, temperature, and pH of fixation can be optimized to
avoid excessive cross-linking.
Connective tissue (e.g., collagen, elastin, laminin), epithe-
lium, and adipocytes are especially prone to hydrophobic cross-
linking, as are immunoglobulins. Methods to decrease this type
of background include the addition of detergent/surfactant, e.g.,
Tween 20, to the buffer, or the use of a blocking protein applied
prior to the primary antibody. The blocking protein must be of
the type that can compete effectively with IgG for hydrophobic
binding sites in the tissue. Therefore, the blocking protein must
contain proteins identical to those in the link antibody in order
to prevent nonspecific binding of the secondary antibody, e.g.,
normal swine serum for polyclonal antibodies when the second-
ary antibody would be swine anti-rabbit.
Generally, most existing immunocytochemical protocols
will be optimized to reduce hydrophobic binding, most usually
through the addition of detergent to the wash buffer.
Ionic interactions result when proteins of opposite charges
meet. Most IgG class antibodies have a net negative surface
charge at a buffer pH of 7.0–7.8. Ionic interactions can be
expected if tissue proteins have a net positive surface charge.
These interactions can be reduced by using diluent buffers with
higher ionic strength. As a rule of thumb, buffers used during
immunocytochemical procedures should be between pH 7.0 and
pH 8.0, even during the DAB stage.

Endogenous Enzyme Activity

If enzymes similar to those used as the label are present in the
tissue, they may react with the substrate used to localize the label
and give rise to interpretation problems. False-positive reactions
produced in this way can be eliminated by inhibiting the endog-
enous enzyme activity prior to staining.
Peroxidase results in the decomposition of hydrogen peroxide
and is a common property of all hemoproteins, myoglobin, cyto-
chrome, and catalases. The most frequently used method is satu-
ration of the endogenous peroxidase with its substrate (hydrogen
peroxide), usually in the form of 0.5% hydrogen peroxide in a
diluent (water, buffer, or methanol).8
When staining specimens that contain large amounts of
endogenous peroxidase, it is often easier to use a different
enzyme, i.e., alkaline phosphatase, in the detection system.
 3. IMMUNOHISTOCHEMISTRY 29

As for endogenous peroxidase, endogenous alkaline phospha-

tase activity is quenched by the addition of its substrate (levami-
sole) in the chromogen solution. Levamisole will inhibit many
types of alkaline phosphatase activity. However, the alkaline
phosphatase used in the detection system is usually intestinal in
nature and remains unaffected by levamisole.

Endogenous Biotin
Biotin is a vitamin and coenzyme that is found in liver, kidney,
and a variety of other tissues. Biotin binds specifically and with
a very high affinity to avidin and streptavidin. Its endogenous
activity is most pronounced in frozen sections. It is possible to
block endogenous biotin with successive incubations of 0.1%
avidin and 0.01% biotin. The avidin blocks the endogenous
biotin, and the dilute biotin blocks any free sites left on the
avidin. Alternatively, if the tissue under investigation contains
large amounts of endogenous biotin, a different detection system
can be used, e.g., a polymer-based method, which does not use
biotin in its labelling system.

Washing
Thorough washing of slides in between the various steps of an
immunohistochemical technique is essential. Most protocols use
a tris buffered saline (TBS), although phosphate buffered saline
(PBS) can be used. As mentioned above, the addition of a deter-
gent to the buffer wash will help reduce some background stain-
ing; however, this should not be used when staining frozen
sections, as there is a higher risk of section detachment when
dealing with these specimens. It is worth remembering that it is
difficult to over wash sections whereas under washing them will
result in dirty, patchy staining.

References
1. Shi SR, Keye ME, Kalra KL. Antigen retrieval in formalin-fixed
paraffin-embedded tissues: an enhancement method for immun-
ohistochemical staining based on microwave oven heating of
sections. J Histochem Cytochem 1991;39:741–748.
2. Singh N, Wotherspoon AC, Miller KD, Isaacson PG. The effect of for-
malin fixation time on the immunocytochemical detection of antigen
using the microwave. J Pathol (Suppl.) 1993;382A.
3. Heggeness, MH Ash, JF. Use of the avidin-biotin complex for the
localization of actin and myosin with fluorescence microscopy. J Cell
Biol 1977;73:783.
4. Vyberg M, Neilsen S. Dextran polymer conjugate two-step visualisa-
tion system for immunocytochemistry: a comparison of EnVision+
30 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

with two three-step avidin-biotin techniques. Appl Immunohistochem

1998;6(1):3–10.
5. Nakane, PK, Pierce GB. Enzyme-labelled antibodies: preparation and
localisation of antigens. J Histochem Cytochem 1966;14:929–931.
6. Graham RC, Ladholm, U, Karnovsky, MJ. Cytochemical demonstra-
tion of perxidase activity by 3-amino-9-ethylcarbazole. J Histochem
Cytochem 1965;150–152.
7. Kaplow, LS. Substitute for benzidine in myeloperoxidase stains. Am
J Clin Pathol 1975;63:451.
8. Streefkerk, JG. Inhibition of erythrocyte pseudoperoxidase activity by
treatment with hydrogen peroxidase following methanol. J Histochem
Cytochem 1972;20:829.
Chapter 4
Cell Culturing: A Beginner’s Guide
to Understanding the Basics of
Cell Culturing
Khurshid Alam, Edwin T. Anthony, P.N. Vaiude, Faruquz Zaman,
and Harshad A. Navsaria

INTRODUCTION
The fundamental aspect of cell culturing is to understand the
type of cells that the investigator wishes to grow. There are many
differences between the cell types; however, the easiest method
of categorizing them is into primary cells and cell lines.
Two types of cultured cell types:

• Primary cells
• Cell lines

Primary Cells
These are cells derived directly from tissue samples/biopsies
which have heterogeneous nature of cells with variable growth
friction. The cells are extracted directly from the tissue and
grown directly in specified optimal cell culture Medias.

Cell Lines
These are cells subcultured from primary cells but now have been
manipulated in the laboratory so that they last longer and can
go through more cell passages with growth friction of 80% or
more, than the original primary cells before they change their
morphology. On the whole, they tend to be more resilient than
primary cells. However, due to their manipulation, they may not
mimic in-vivo cells as closely as the unmanipulated primary cells.
Subcultured cells separated from primary cells create homoge-
neous cell lineages. They ascertain specific properties by the
process of cloning or physical cell separation, thus leading to
so-called cell strains.
32 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

FUNDAMENTALS OF CELL CULTURING

The reason for undertaking cell culturing is that the growth of
the cells can be mimicked as closely as possible to its growth in
its normal host environment. By growing the cells in cell culture
flasks or cell culture dishes, the investigator may build up the
cell numbers and then use these cells to conduct experiments.
These in-vitro cell experiments may be performed, and later the
final effects on the cells may be investigated.
The cells are delicate and prone to bacterial and fungal
infections from the environment, so all precautions are used to
minimize contamination. Bacterial contamination is a difficult
problem. In this respect, mycoplasma-species contamination
is a more difficult problem to deal with than the growth of
other bacteria and fungi. To combat contamination, the single-
use disposable equipment and aseptic techniques are applied
(Figure 4.1).
Methods to prevent cell contamination:

• Gloves
• Sterilized equipment
• Laminar flow cabinet
• Singe-se pipettes
• Resterilization of equipment

What Happens When an Infection Occurs?

Usually a bacterial or fungal infection would mean that the cells
would need to be safely discarded to prevent cross contamina-
tion. However, when the cells are “precious,” an attempt to treat
the contamination may be made. Free-floating contaminants in
the media are physically removed, and the cells are washed
numerous times with phosphate buffered saline (PBS). Loose
lids or covers are replaced with new sterile ones. An antibiotic
mix is added to the media, and the antibiotic medium is added
to the cells. The cells are then cultured for 2 to 3 days. If the cells
remain uncontaminated, then they are returned back to a normal
media for another 2 to 3 days to confirm that no contamination
remains.
First-choice antibiotics/antifungal agents for cell culture:

• Gentamicin
• Streptomycin
• Penicillin G
• Amphotericin B (antifungal)
 4. CELL CULTURING 33

FIGURE 4.1. Cell culturing in a laminar flow cabinet. Note the scientist
is using a battery-operated “pipette boy” to draw the media up into the
pipette. (Courtesy of P.N. Vaiude, Honorary Clinical Fellow, Barts and
the London NHS Trust.)

Media: Their Function in Cell Culturing

Media provide the food and supplements that are needed for the
growth of cells. Over the years, scientists have searched for the
ideal media needed for the growth of various cell types. There is
now an exhaustive list of media which are suited for the different
cell types. So, when the investigator is growing a cell type, the
investigator must be careful to select the correct medium. Also,
as part of an investigator’s experiment, the investigator may add
or remove certain substances from the media and monitor the
cells’ response to the changes in their environment. A medium
may be a solid, broth, or solution. Figure 4.2 shows the Caplan-1
and Paca-3 cell strains growing in solution medium.
34 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

B
FIGURE 4.2. Caplan-1 (A) and Paca-3 (B) cell strains growing in solution
medium. (Courtesy of P.N. Vaiude, Honorary Clinical Fellow, Barts and
the London NHS Trust.)

Support Cells
Certain cells need “feeder” cells to support their growth, espe-
cially during the initial stages when their cell numbers are low.
Once the cells of interest have reached an adequate number and
are able to support themselves, the “feeder” cells are removed,
and the cells are allowed to grow on their own momentum. An
 4. CELL CULTURING 35

example of this is the use of “3T3” cells from mice for keratino-
cytes (skin epithelial cells). These cells are gamma irradiated
before use so that they can provide support for the keratinocytes
when they are initially extracted from the biopsies and are low
in number. The 3T3 cells are irradiated so that they can provide
support but then die off within a few days; otherwise they will
overtake the cells being cultured and kill them.

Incubation
Ideal conditions for the growth of cells are normally maintained
for maximal growth. This usually includes incubating the cells
at a suitable temperature, environmental moisture, and an ade-
quate environmental CO2 concentration. Humidified CO2 incuba-
tors with shelves are usually available with these preset conditions
for cell culture. They are regularly cleaned out with a detergent
or 70% alcohol, and fungicide/bactericide is placed in the humid-
ified water to prevent cell infections during incubation.

CELL EXTRACTION
There are many means of extracting cells, and the one that we
will describe is for fibroblasts from tissue biopsies.

Handling Tissue Biopsies

Tissue biopsies should be taken immediately from the time of
the operation/procedure and kept in a suitable transport media.
(An example is E4 [DMEM Eagle’s medium] medium with 4 mls
of glutamine and an antibiotic mix.) Ideally, the cell extraction
should start immediately, but it may be done up to a maximum
of 48 hours later if the samples are kept in the transport media
at a constant fridge temperature of 4°C–6°C.

Fibroblast Extraction
Primary fibroblasts are fairly resilient cells that proliferate quickly
and do not need support cells for their growth. However, like all
cells, they will need their ideal medium and environmental con-
ditions (Figures 4.3 and 4.4).1
Fibroblast extraction steps:

• Take tissue biopsy

• Place on a Petri dish
• Using a sharp scalpel cut the tissue into tiny pieces
• Moisten cut biopsy tissue
• Leave for 30 minutes
• Add medium and close Petri dish with its cover
36 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

FIGURE 4.3. Fibroblasts growing out from the explanted tissue sample
on a cell culture flask. The picture was taken with a camera mounted on
a light microscope. Magnified ×20. (Courtesy of P.N. Vaiude, Honorary
Clinical Fellow, Barts and the London NHS Trust.)

FIGURE 4.4. The investigator is using a light microscope to monitor the

progress of his cells, which are bathed in media and growing in cell
culture flasks. (Courtesy of P.N. Vaiude, Honorary Clinical Fellow, Barts
and the London NHS Trust.)
 4. CELL CULTURING 37

• Incubate and inspect at least twice-weekly

• When the fibroblast outgrowths from the biopsy are adequate,
remove the tissue
• Add Trypsin to remove the fibroblast from the plate
• Transfer the fibroblasts to a 25 cm2 tissue culture flask
• Add medium to the flask
• When the flask is confluent (full of cells), trypsinize again
• Transfer the fibroblasts to a larger 75 cm2 tissue culture flask
• Change the medium at least twice-weekly
• Cells can be grown for many passages before their morphology
changes
• The cells can also be stored in liquid nitrogen for use later.
(Use standard cell-freezing protocols

CONCLUSION
Cell culturing techniques have developed and significantly
improved over the decades and have enabled major scientific
leaps to important fields in cancer study, wound healing, and,
not least, stem cell technologies.2
Glossary:

• Cell lines—manipulated primary cells (engineered cells)

• In vitro—out of the body/natural environment; in a laboratory
vessel
• In vivo—in the body/natural environment
• Passage—to remove/detach the cells from the culture
disk/flask
• Primary cells—cells directly from the tissue (unmanipulated
cells)
• Trypsinize—to add trypsin

References
1. Jones, GE., Witkowski, JA. Reduced adhesiveness between skin fibro-
blasts from patients with Duchenne muscular dystrophy. J Neurol Sci
1979;43(3):465-470.
2. Navsaria HA, Myers S, Leigh IM, McKay IA. Culturing skin in vitro
for wound therapy. (Review). Trends Biotechnol 1995;13:91-100.
Chapter 5
Flow Cytometry
P. Erotocritou, M. Arya, S.N. Shukla, and H.R.H. Patel

WHAT IS FLOW CYTOMETRY?

The measurement of the physical and chemical characteristics
of cells is called cytometry. Flow cytometry is the technique where
these measurements are made individually of single particles
(cells, nuclei, chromosomes), suspended within a stream of
liquid, as they pass through a laser light source.
All parameters measured can be divided into two main
groups: 1) those related to light scattering, which mainly reflects
the size of the cell and its internal complexity and 2) those related
to fluorescence. These are associated with the presence of one or
more fluorochromes inside the cell or attached to the cell surface
membrane, either naturally (autofluorescence) or artificially
(e.g., using fluorochrome conjugated monoclonal antibodies).
The light scatter and emitted fluorescence of each particle is
collected, filtered and converted to digital values which are stored
on a computer. The essential feature of flow cytometry is that
individual measurements are made for each particle within the
suspension, rather than measuring an average property for the
entire population, thus allowing subpopulations to be detected
within the overall population. An extremely large number of
particles can be evaluated in a very short time; some systems can
run particles at rates approaching 100,000 particles per second
whilst simultaneously collecting 10–20 parameters from each
particle. Additionally, flow cytometry can assess particles of most
sizes—these range from particles below the resolution limits of
visible light, as they can be detected by their fluorescent signa-
tures; to those particles as large as several thousand microns.
There are many applications of flow cytometry in both basic
and clinical research. The technology permits rapid and accurate
measurements of multiple cell and intracellular characteristics
that include DNA/RNA cell content, mitochondria and chromo-
somes, enzyme activity, membrane potential, the measurement
of intracellular pH or ions such as calcium, the detection
 5. FLOW CYTOMETRY 39

and quantification of cell antigens, the analysis of multidrug

resistance.1,2,3,4

HISTORICAL ASPECTS OF FLOW CYTOMETRY AND

THE FLOW CYTOMETER
The foundations of flow cytometry began with the seminal work
by Caspersson and Schultz in 1938.5 They showed that the DNA
content of unstained cells doubled during the cell cycle, when
measured by ultraviolet and visible light absorption. Following
this, in the 1940s it was shown by Papanicolaou6 that cancerous
cells from cervical cancer could be identified by observing the
staining patterns obtained by staining tissues with specifically
designed dyes.6
Coons and Kaplan7 in 1950 demonstrated fluorescence anti-
body methods improved detection of antigens,7 which suggested
that it was more beneficial to measure fluorescence than absorp-
tion. Fluorescin has since remained the most common label for
quantitative immunofluorescence analysis, and now for routine
flow cytometric applications in immunology and hematology.
Moldovan initially established the principle of the automated
flow analysis of single cells by reporting a photoelectric method
for counting individual cells flowing through a capillary tube
mounted on a microscope stage.8 Coulter9 modified this system
and built the forerunner of modern day flow cytometers. In this
machine blood cells in saline suspensions passed one by one
through a small orifice being detected by changes of electrical
impedance at the orifice.9
In the 1960s, Kamentsky10 designed a cytometer that mea-
sured individual cell absorption and scatter. The apparatus was
used to measure the nucleic acid content and light scattering of
unstained mammalian cancer cells in a flow stream.10 Very soon
afterwards, Fulwyler11 described the first flow sorter.11 This
instrument allowed biological cells to be separated according to
their volume, which was determined as they passed through a
Coulter aperture.
Thus, flow cytometry was based mainly on fluorescence and
light scattering, and was originally used in the detection of
nuclear DNA and cell surface antigens. The applications of this
technique have since expanded rapidly.

FLUORESCENCE
This is a key concept in flow cytometry and is the property of
molecules to absorb light energy at one wavelength, called the
excitation wavelength, and then rapidly re-radiates some of that
40 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

energy as light of a longer or emission wavelength. Fluorescence

is always of a lower energy, and therefore longer wavelength,
than the exciting light.
When a compound absorbs light, electrons are initially raised
from the ground state to an excited state and then subsequently
return to the ground state via numerous routes; some routes do
not result in fluorescence, such as the loss of the energy by heat,
but certain molecules lose energy by a process of radiative trans-
fer, termed fluorescence. When stimulation of the fluorescent
compound is stopped by removing the exciting light source, fluo-
rescence emission immediately stops.
Immunofluorescence allows the visualization of cellular fea-
tures or structures by linking them to a fluorescent molecule.
These fluorescent molecules can either be dyes which bind
directly to structures within or on the cell, or fluorochromes
conjugated to a ligand, such as monoclonal antibodies. The fluo-
rochromes and dyes used for flow cytometry must be compatible
with the laser utilized by the cytometer.
The characteristic distribution of radiated energy for a fluo-
rochrome is called its emission spectrum. Each fluorochrome
has a clear, distinct peak in its emission spectrum represented
visually by a characteristic color for that fluorochrome (Table
5.1). This property allows flow cytometric assays to use multiple
fluorochromes in a single experiment thus allowing the simulta-
neous evaluation of different cell features, without having to be
concerned about significant overlap or interference from indi-
vidual emission spectra.

TABLE 5.1. Photomultiplier detection of light emission produced by

Common Fluorochromes/Dyes in the Becton Dickinson FACScan

Light emission
wavelength detected
Photomultiplier (color) Fluorochrome

FL1 515–545 nm (green) FITC

FL2 564–606 nm (yellow-orange) PE/PI
FL3 650–675 nm (red) PE-CY5

Key: FITC, fluorescein isothiocyanate; PE, phycoerythrin; PI, propidium

iodide;
PE–CY5, a tandem conjugate of two fluorochromes (PE and Cyanine-5)
in which the energy is transferred from the first to the second.
 5. FLOW CYTOMETRY 41

LIGHT SCATTER
The flow cytometer is able to detect and quantify light scatter
signals as well as immunofluorescence. The light scatter signals
are due to the laser light of the flow cytometer reflecting and
refracting off the cells. Two types of light scatter are quantified:
1) Orthogonal or side light scatter, which is light scatter
measured at a right angle (90 degrees) to the direction of the
laser beam. This is detected in the side channel, the intensity of
the reading correlating with the granularity of the cell.
2) Forward or low angle scatter, which is light scatter
measured along the same axis that the laser light is traveling or
near 180 degrees from the point at which the laser beam intersects
the cell. This measurement correlates with cell size and cell
surface properties, such as cell ruffling. Therefore not only can
it be used to distinguish cell size, but also live from dead cells.

COMPONENTS OF THE MODERN FLOW CYTOMETER

Fluidic System
The fluidic system focuses the cells/particles into a fine stream,
which are moved individually to intersect the laser light source.
Furthermore, the fluidic system is vital in cell sorting by flow
cytometry. Initially, the sample of fluorescently labelled suspen-
sion of single cells flows rapidly through a narrow channel in the
instrument, where a small amount of cell suspension joins a
larger amount of cell free buffer (sheath fluid). These two streams
do not mix, and consist of an inner sample stream surrounded
by an outer sheath stream (coaxial flow). The coaxial stream
ensures the cells are centered in the flowing stream, passing the
laser beam optimally centered, in addition to being spaced out
sequentially and passing the laser beam individually.
To ensure the sample fluid is flowing continuously, positive
air pressure is applied to the sample reservoir and sheath fluid.
A purge line is connected to the sheath inlet to allow a vacuum
to be applied for clearing blockages and air bubbles.

Optical System and Analysis

The optical system has two components:
1. A light source, most frequently a laser, which is necessary
to excite sufficiently the cells, in addition to the lens and mirrors
required to focus and direct the laser beam to the flow chamber
through which the cells are passing. As mentioned earlier, the
42 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

lasers, which emit light at specific peak emission wavelengths,

need to be matched with appropriate fluorochromes used for
immunofluorescent staining analyzed on the cytometer. The flow
channels through which the coaxial stream passes is positioned
vertically to the laser beam.
2. A variety of mirrors and filters that absorb certain
wavelengths while transmitting others are needed for the
collection of light emitted by the cells or scattered from them.
These include dichroic mirrors, which reflect light above a
specific wavelength, while permitting light below that wavelength
to pass through; long pass filters (permit only light above a
specified wavelength to pass through); short pass filters (permit
only light below a specified wavelength to pass through); band
pass filters (permit light within a specified wavelength range to
pass through).

These mirrors and filters also act to separate and direct emis-
sions of varying wavelengths to the corresponding detectors or
photomultiplier tubes (PMT). Within the PMTs the incoming
light is converted into electronic signals. Subsequent electronic
and computational processing of these signals results in graphic
display and statistical analysis of the measurements being
made.
There are cytometers now available that are capable of ana-
lyzing up to 13 parameters for each cell (forward scatter, side
scatter and up to 11 colors of immunofluorescence).12 This allows
for cell size, lipid content, protein content, DNA content, enzyme
activity to name a few characteristics for each cell to measured.
Thus, allowing for a multidimensional representation of a popu-
lation to be obtained. With most cytometers, it is almost always
possible for at least 1,000 cells to be analyzed per second, whilst
with appropriate specimens some cytometers are able to analyze
up to 100,000 cells per second, with no marked reduction in data
quality.13

Color Assignment
The signal emitted by a fluorochrome is detected by its corre-
sponding PMT and then converted to a parameter that can be
acquired. The series of optical filters and mirrors used ensures
that only specific regions of the spectrum reach each PMT. The
PMT detectors in the flow cytometer are labeled FL1, FL2, FL3
and onwards depending on how many are present, with light of
specific emission wavelength only been detected by each.
 5. FLOW CYTOMETRY 43

The Becton Dickinson FACScan contains an argon laser with

an emission at 488 nm, with three PMTs detecting fluorescence
specific wavelengths produced through excitation of several com-
monly used fluorochromes/dyes (Table 5.1) and two for detecting
side and forward scatter.
A wide variety of fluorochrome conjugated ligands and dyes
are available for directly estimating cellular parameters such as
DNA content. The simplest method utilizes a fluorescent dye that
binds preferentially to DNA, allowing for cellular DNA levels to
be detected. These DNA dyes, such as propidium iodide are
termed stoichiometric, as the number of molecules of the dye
bound is equal to the number of DNA molecules. Fluorochrome
conjugated ligands are used in measuring membrane potential,
enzyme activity, calcium flux, pH, the density and distribution
of cell-surface and cytoplasmic determinants and receptors
(Table 5.1).

DATA ANALYSIS

Histograms and Dot Plots

When light from a specific region of the spectrum is detected by
a PMT, it is converted via an amplifier to a voltage, this voltage
being proportional to the intensity of fluorescence. These volt-
ages are then processed through a series of linear and log
amplifiers. Using a logarithmic scale to measure fluorescence is
indicated in most biological situations, and allows for a distribu-
tion to be normalized. Also, a logarithmic scale is important
when there is a broad range of fluorescence, as is often encoun-
tered with biological distributions, because this type of amplifica-
tion allows for an expansion of the scale for weak signals with a
compression of the scale for strong or specific fluorescence
signals.
These voltages are a continuous distribution and are con-
verted to a discrete distribution by an analog to digital converter
(ADC) which positions each signal according to the level of fluo-
rescence into a specific channel. These channels correspond to
the original voltage generated by a specific light event detected
by the PMT detector. Consequently the ADC assigns a channel
number based on the pulse height for individual events; brighter
specific fluorescence events yield a higher pulse height and thus
a higher channel number when displayed as a histogram. The
greater the resolution of the ADC, the closer this reflects the
continuous distribution. The ADCs in the FACScan are 10-bit,
meaning they divide data into four decades across 1024 channels
44 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

with 256 channels per decade. The number of decades is fixed

for the FACScan.
The data gained through flow cytometry is most often repre-
sented as histograms or dot plots (Figure 5.1). The histogram
graph has fluorescent intensity on the x-axis with the number of
cells in each channel on the y-axis. With the FACScan, histo-
grams have 1024 channels displayed on the x-axis using a 4-log
decade logarithmic scale (256 channels per decade), with higher
channel number corresponding to a greater signal intensity of a
fluorescence/scatter after amplification (Figure 5.1). A histogram
can only represent the intensity of a single parameter (scatter or
fluorescence).
Dot plots (bivariate display, two parameter histograms, scat-
tergram, bitmap) unlike histograms show the correlated distri-
bution for two parameters, such as forward scatter versus side
scatter, with each cell being represented as a dot, positioned on
the x and y axes according to the intensities detected for that cell.
Dot plots can be divided into quadrants allowing for the number
of particles in each of the defined areas to be counted (Figure
5.1).
Histograms and dot plots both allow for discrete subpopula-
tions of cells with different intensities to be identified.

Gating and Negative Controls

Normally data only from single, living cells is wanted and data
from cell debris, dead cells, and clumps of two or more cells
needs to be eliminated. Single cells are distinguished from sub-
cellular debris and clumps of cells according to size. Living cells
are distinguished from dead cells by forward scatter (dead cells
have lower values) and side scatter (dead cells have higher
values). These differences remain even after formaldehyde fixa-
tion, despite the fact that after fixation, all the cells are dead. Due
to the above, the computer used can be configured to display the
fluorescence signals only from single living cells (Figure 5.1).
This is termed scatter-gated fluorescence analysis, where a speci-
fied set of scatter properties is set to identify fluorescence signals
from a desired group of particles only. It is possible to “gate” on
any set of signals in order to analyze the specific data of either
the main cell population or a subpopulation.
On dot plot graphs quadrant gates can be set using back-
ground levels of fluorescence of either the respective unstained
negative control (fluorochrome-conjugated secondary antibody
only) or fluorochrome isotype matched control population of
cells (when using a fluorochrome-conjugated primary antibody)
 5. FLOW CYTOMETRY 45

Histograms Dot plot - negative control

– contains secondary
Negative Control – contains secondary antibody only antibody only
Contains primary and secondary antibody

Dot plot -
contains primary and
secondary antibody

This figure shows representative flow cytometry histograms and dot plots. On both the histograms and
dot plots the x-axis shows fluorescence intensity (channel number) on 1024 channels encompassing
4-log decades (i.e. logarithmic scale).
On the histograms the y-axis shows cell counts in each channel, whereas on the dot plots the y-axis
shows side scatter intensity - SSC (channel number) on a linear scale.
Quadrant gates were set on the dot plots using the background levels of fluorescence of the respective
unstained negative control (contains only the fluorochrome-conjugated secondary antibody without the
primary antibody).
Data due to dead cells/cell debris and non-specific binding of the secondary antibody being have been
gated out of the final results

FIGURE 5.1. Histograms and dot plots.

46 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

(Figure 5.1). The fluorochrome isotype matched control primary

antibody has no specificity to any epitope on the cells being
analyzed, is the same isotype and species used at the same con-
centration and bound to the identical fluorochrome at the same
fluorochrome/protein ratio as the antigen specific primary anti-
body. Both of these negative control samples are needed to high-
light the amount of nonspecific binding that is present with the
antigen specific antibody.
When quadrant gates are used on the graph, each quadrant
then demonstrates the following:

1. Lower-left quadrant represents cells negative for the descrip-

tors on both the x- and y-axis.
2. Upper-right quadrant represents cells dual-positive for the
descriptors on both the x- and y-axis.
3. Upper-left quadrant represents cells positive for the y-axis
descriptor, but negative for the x-axis descriptor.
4. Lower-right quadrant represents cells positive for the x-axis
descriptor, but negative for the y-axis descriptor.

PRINCIPLES OF FLOW CYTOMETRY SAMPLE PREPARATION

The aim of sample preparation is to produce a monodisperse sus-
pension; this is a suspension of single cells with minimal aggrega-
tion. Furthermore, the sample should contain minimal cell debris,
dead cells and clumps as the clumps can cause disruption of fluid
flow or block the tubes within the flow cytometer. In addition, it
is important to ensure that preparative methods used will not bias
results. For instance, enzymatic preparative techniques can alter
cell-surface antigens and affect cell viability.
Once a monodisperse suspension is gained the cells are
labelled by incubation with a fluorescent tag under appropriate
conditions. This may be a fluorescent dye, fluorescent conjugated
antibody or ligand, which is specific for the “antigen” to be mea-
sured. The fluorescent probes and monoclonal antibodies may
bind nonspecifically and it is essential that care is taken to mini-
mize the chances of cross-reactions. Data acquisition can occur
with the cells either alive or fixed with substances such as form-
aldehyde. The advantage of formaldehyde fixed cells is that the
light scatter and fluorescent properties are maintained and data
can be acquired several weeks later.

References
1. Cram LS. Flow cytometry, an overview. Methods Cell Sci 2002;
24:1–9.
 5. FLOW CYTOMETRY 47

2. McCoy JP Jr. Basic principles of flow cytometry. Hematol Oncol Clin

North Am 2002;16:229–243.
3. Brown M, Wittwer C. Flow cytometry: principles and clinical applica-
tions in hematology. Clin Chem 2000;46:1221–1229.
4. Herzenberg LA, Parks D, Sahaf B, Perez O, Roederer M, Herzenberg
LA. The history and future of the fluorescence activated cell sorter
and flow cytometry: a view from Stanford. Clin Chem 2002;48:
1819–1827.
5. Caspersson T, Schultz J. Nucleic acid metabolism of the chromo-
somes in relation to gene reproduction. Nature 1938;142:294–297.
6. Papanicolaou GN, Traut R. The diagnostic value of vaginal smears
in carcinoma of the uterus.Am J Obstet Gynecol 1941;42:193.
7. Coons AH, Kaplan MH. Localization of antigen in tissue cells, II.
Improvements in a method for the detection of antigen by means of
fluorescent antibody. J Exp Med 1950;91:1–4.
8. Moldovan A. Photo-electric technique for the counting of microscopi-
cal cells. Science 1934;80:188–189.
9. Coulter WH. High speed automatic blood cell counter and cell size
analyzer. Proc Natl Electronics Conf 1956;12:1034–1040.
10. Kamentsky LA, Melamed MR, Derman H. Spectrophotometer: New
instrument for ultrarapid cell analysis. Science 1965;150:630–631.
11. Fulwyler MJ. Electronic separation of biological cells by volume.
Science 1965;150:910–911.
12. De Rosa SC, Herzenberg LA, Roederer M. 11-color, 13-parameter
flow cytometry: identification of human naive T-cells by phenotype,
function, and T-cell receptor diversity. Nat Med 2001;7:245–248.
13. Ashcroft RG, Lopez PA. Commercial high speed machines open new
opportunities in high throughput flow cytometry. J Immunol Methods
2000;243:13–24.
Chapter 6
Western, Northern, and
Southern Blotting
Stephan Schlickeiser and Uwe Pleyer

INTRODUCTION: THE HISTORY OF SOUTHERN, NORTHERN,

AND WESTERN BLOTTING
In the early 1970s, the possibility of mapping whole genomes
arose, due to the prior discovery of bacterial enzymes that cut
DNA at specific “restriction sites,” and the development of recom-
binant DNA technologies and gene cloning. Nevertheless, one
could not identify one single gene among thousands of fragments
of DNA—until Edward Southern1 introduced his eponymous
powerful DNA transfer and probing technique in 1975.1 He real-
ized that restriction fragments can first be separated electropho-
retically on an agarose gel and then be transferred to a nylon
membrane by capillary action—the same way that blotting paper
absorbs ink. Afterward, the blotted membrane can be incubated
with a radioactive probe specific for the gene fragments of inter-
est, which in turn become visible by placing an x-ray film on top
of the membrane.
In 1977, gene expression analysis ascended dramatically
when George Stark and colleagues replicated the configuration
of Southern’s transfer apparatus in an effort to transfer cellular
RNA to chemically activated cellulose paper.2 This technique was
named “Northern blotting”—as a pun on Southern’s name.
Two years later, Stark developed an early protein blotting
technique, relying on overnight capillary transfer from gel to
activated cellulose paper.3 Harry Towbin’s4 faster and simpler
approach for electroblotting proteins to nitrocellulose mem-
branes is clearly preferred today.4 The name “Western blot” was
first given to the technique by W. Neal Burnette5 in 1981,
though.5
Definitions:

Southern blot: DNA is detected with a hybridization DNA or RNA

probe.
 6. WESTERN, NORTHERN, AND SOUTHERN BLOTTING 49

Northern blot: RNA is detected with a hybridization DNA or RNA

probe.
Western blot: Protein is detected with a complementary
antibody.

BASIC PRINCIPLES AND METHODS: ELECTROPHORESIS,

GEL BLOTTING, AND DETECTION
All three blotting techniques use a very similar methodology.
Molecules separated by electrophoretic procedures are trans-
ferred to a membrane that is especially suited to support the
detection of fragments with a particular DNA sequence, single
species of RNA, or proteins.

Gel Electrophoresis: Size Separation

A simple way to identify one particular molecule amongst others,
that are related but not identical, is to distinguish them by their
physical characteristics such as size or mass. The gel electropho-
resis is a technique that permits a size based separation of mac-
romolecules by their movement in an electric field. The buffer
conditions, such as pH and ion composition, provide a negative
net charge of the nucleic acid fragments or proteins, which will
be moved from the cathode to the anode. The gel usually consists
of a cross-linked polymer matrix. Its sub-microscopic pores will
impede large molecules and let smaller fragments migrate faster
through the gel.

Sample Preparation: Southern and Northern

DNA is first isolated by common extraction protocols. Afterward,
the purified genomic or plasmid DNA is partially digested by
restriction endonucleases. These enzymes cut double-stranded
nucleic acids at certain sequences. The resulting mixtures of dif-
ferent-sized linear fragments are then loaded side-by-side into
“wells” formed in an agarose gel.
RNA is single-stranded. Due to their natural instability, work
with RNA molecules requires much more care. All procedures
need to be done with sterile, RNAse-free supplies. Furthermore,
RNA has to be pretreated with denaturing formaldehyde to
prevent formation of base-paired secondary structures.

Sample Preparation: Western

Samples from cell culture are lysed in an extraction buffer con-
taining proteinase inhibitors. Tissue samples are cooled or frozen
rapidly and homogenized using mechanical force. The cell debris
is removed by sharp centrifugation. Further centrifugation steps
50 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

can be used to isolate cytosolic or nuclear fractions. The resulting

lysates are then assayed for whole protein content, so that from
each sample equal amounts of protein can be taken for the most
commonly used sodium dodecylsulfate–polyacrylamide gel
electrophoresis (SDS-PAGE). Before applying the samples to the
gel, they have to be boiled one to five minutes in a denaturing
buffer solution (e.g., Laemmli’s buffer). This buffer contains a
tracking dye optionally; reducing agents, such as dithiothreitol
(DTT) or 2-mercaptoethanol; and, most importantly, SDS. Heat
and reducing agents destroy tertiary protein folding and quater-
nary protein structures by breaking up disulfide bonds and pre-
venting their reformation simultaneously. SDS is an anionic
detergent that denatures secondary and nondisulfide-linked ter-
tiary structures, thus unfolding and linearizing the proteins
completely. Furthermore, the anionic SDS binds to most of the
proteins in a uniform ratio of approximately 1.4 g SDS per 1.0 g
protein. Hence, it applies a negative charge to each protein in
proportion to its mass (number of amino acids) so that they may
be separated in an electric field strictly by their length or size,
respectively.

Agarose Electrophoresis: Southern and Northern

As mentioned above, DNA and RNA strands, which have a nega-
tively charged phosphate backbone, will move towards the anode
if an electric current is applied to the gel matrix.
Shorter fragments will move faster because they are able to
slip through the agarose mesh networks more easily. Agarose
concentrations of 0.5% to 2% are normally used for the separa-
tion of fragments of 100 base pairs (bp) up to several kilobases
(kb). For an ideal resolution of larger molecules (up to 750 kb),
the agarose concentration has to be lowered. Inconveniently,
these gels are very fragile and have long running times.
The voltage normally applied to a gel is 100 mV (about 5 to
8 V/cm). Higher voltages will shorten the running time to less
than one hour, but as well lead to a decrease in resolution.
Buffers used for agarose electrophoresis are in general tris acetate
EDTA (TAE) and sodium boride (SB). TAE has a relatively low
buffering capacity but provides the best resolution, whereas SB
has the highest buffering capacity, allowing shorter running
times for fragments smaller than 5 kb (voltages up to 350 mV).
After completion of the separation, bands corresponding to
nucleic acids fragments of different length can be visualized
using a fluorescent dye (e.g., ethidium bromide). Fragment-size
determination is typically done by comparison to strands of
 6. WESTERN, NORTHERN, AND SOUTHERN BLOTTING 51

known length. Size standards are commercially available (DNA

ladders).
For high resolution of short DNA fragments or RNA mole-
cules (<100 bp), polyacrylamide gels are commonly used.

Sodium Sodecylsulfate–Polyacrylamide Gel

Electrophoresis: Western
Samples with the denatured proteins are placed into the lanes of
the stacking site of a polyacrylamide gel submerged in a suitable
buffer. One lane is reserved for a molecular weight standard in
order to determine the size of unknown proteins, usually in
Dalton. An electric current of about 5 to 50 mA is commonly
applied across the gel. Movement through the stacking gel forces
all proteins to focus in one sharp front at the boundary to the
resolving gel. Thus, different sized proteins may resolve in sharp
and distinct bands at the end of the separation. To achieve longer
retention of proteins and higher resolutions respectively, one can
enhance the grade of polymerization of the resolving gel by
increasing the percentage of acrylamide. Typically, resolving gels
are made in 8% to 15%. The stacking gel (5%) is poured on top
of the readily polymerized resolving gel to form a sharp transi-
tion. A gel comb defines the sample lanes. The recipe for a 10 mL
10% polyacrylamide gel could be:

Resolving polyacrylamide Stacking polyacrylamide

mixture mixture (5%)
dH20 mL dH20 mL

30% acrylamide mix 4.0 30% acrylamide mix 4.0

1.5M Tris pH 8.8 3.3 1.0M Tris pH 6.8 3.3
10% SDS 2.5 10% SDS 2.5
10% ammonium persulfate 0.1 10% ammonium persulfate 0.1
TEMED 0.1 TEMED 0.1
(not to add before (not to add before
everything is ready) 0.004 everything is ready) 0.004

Note: Monomeric acrylamide is toxic!

Following electrophoresis, the gel is stained (e.g., with Coo-

massie Brilliant Blue or silver stain), to visualize the separated
proteins. The crucial step is to take an image of the gel before
transferring to the membrane, because the molecular weight
marker cannot be shown after blotting.
To better characterize one particular protein, it is possible to
use a two-dimensional gel electrophoresis, which separates the
52 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

proteins of a single sample according to their isoelectric point

(1st dimension) and their molecular weight (2nd dimension).

Blotting: Transfer to Solid Support

After the DNA fragments, RNA molecules, or proteins have been
separated according to their size, they have to be transferred to
a solid membrane in order to make these molecules accessible
to specific detection. The membranes or filters are made of nitro-
cellulose, nylon, diazo-modified cellulose, or polyvinylidene fluo-
ride (PVDF). These materials allow binding with high capacities
by either crosslinking with single-stranded nucleic acids or strong
hydrophobic and charged interactions with proteins. Therefore,
the relative positions of the molecules are not altered.
Since DNA fragments in the agarose gel are double stranded,
they have to be denatured by pretreatment with an alkaline solu-
tion (typically containing sodium hydroxide).
There are two methods to transfer the molecules. The first
method is to place the membrane on top of the gel that lies on
a stack of wet filter paper. Then pressure is applied by either
suction (vacuum blot), or by placing another stack of dry paper
towels and a weight on top of the membrane and gel (capillary
blot). Thus, molecules will move from the gel onto the membrane
with a flow of buffer by simple capillary action.
The second method takes advantage of the molecules’ nega-
tive charge (electro blot). In an electrophoretic chamber, the
membrane is placed face-to-face with the gel between two large
electrode plates (Figure 6.1). These electrodes can either be posi-
tioned in a tank filled with suitable buffer (tank blot), or mem-
brane, gel, two buffer soaked stacks of filter paper and electrodes
can be assembled akin to a sandwich (semidry blot). Pay attention
to the right polarization, because the negatively charged mole-
cules will move towards the anode!
Afterward, nitrocellulose membranes are baked, and nylon
membranes are exposed to ultraviolet light to covalently link
DNA or RNA to the solid support. As proteins usually stick very
tightly to the membrane, they don’t need to be crosslinked.
Efficiency and integrity of the blotting process may be
controlled by reversible staining of the membrane, e.g., with
Ponceau S.

Detection and Localization: Complementarity and Hybridization

The basic principle behind the specific detection of any target
macromolecule is the sequence-specific or shape-specific mole-
cular recognition that occurs when a complementary probe mol-
 6. WESTERN, NORTHERN, AND SOUTHERN BLOTTING 53

size
standard

gel
electrophoresis
A

blotting

direction of
B transfer
detection
h·ν
/ β, γ
probe
hybridization

filter gel
paper nylon
cathode membrane anode

FIGURE 6.1. Gel loading and blot electro/blotting.

ecule binds to the target. A high complementarity between both

molecules results in a highly stringent formation of a probe-
target complex. Hence, hybridization reactions will specifically
occur in the presence of a large number of similar but noncom-
plementary molecules. This hybrid complex can then be located
if the probe molecule bears a detectable function. The most com-
monly used probes in Southern and Northern blotting are short
radioactive (32P or 125I), fluorescent- or digoxigenin (DIG)–labeled
DNA or RNA single strands (100 to 500 bp) that have to be reverse
54 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

complement to the target sequence. In Western blotting, a two-

step hybridization with a primary antibody against the target
epitope and a secondary antibody directed at a species-specific
portion of the primary antibody is preferred. The secondary
antibody is typically tagged with radioactivity or linked to a
reporter enzyme, which drives a colorimetric or chemilumines-
cent signal.

Blocking
Because the used membranes are most suitable to bind protein
and nucleic acids, nonspecific binding between probes and the
material has to be blocked. These nonspecific interactions are
prevented by soaking the membranes in a solution containing a
high concentration of DNA (e.g., herring or salmon sperm DNA)
in case of Southern and Northern hybridization. For antibody-
based Western hybridization, protein is used for blocking,
typically 5% nonfat dry milk or 3% bovine serum albumin (BSA).
The presence of a detergent in the blocking solution, like Tween
20 at 0.05% is also important.

Hybridization
The labeled probe is added to the blocked Southern, Northern,
or Western blot membrane in a buffered saline solution, contain-
ing a small percentage of detergent and nonfat milk or BSA.
Under gentle agitation, the probe molecules are allowed to bind
for a period of hours. The recommended final DNA or RNA probe
concentration is 2 to 10 ng/ml. A primary antibody is generally
incubated with the filter in dilution between 0.5 and 5 µg/ml. The
secondary antibody can be added after the unbound primary
antibody is removed by washing the membrane.

Washing
After hybridization, the membrane is rinsed repeatedly in several
changes of buffer to wash off any unbound, excess antibody, or
nucleic acid probe, in order to avoid unspecific background
signals.

Detection
If the probe is radioactive, the pattern of hybridization is visual-
ized on x-ray film by autoradiography. The membrane is pressed
against the film, which in turn is exposed, for a few minutes up
to weeks, wherever the probe bound. Because nonradioactive
detecting methods are safer, quicker, and cheaper, nowadays
autoradiography is scarcely used.
 6. WESTERN, NORTHERN, AND SOUTHERN BLOTTING 55

The secondary antibody, which can be directed against

protein-bound primary antibodies as well as DIG-labeled DNA or
RNA probes, is usually linked to biotin, a fluorescence dye, or to
reporter enzymes. Alkaline phosphatase (colorimetric detection)
converts a chromogenic dye to a colored reaction product visible
on the membrane. Most frequently used is horseradish peroxi-
dase (chemiluminescence detection) which, if provided with
hydrogen peroxide and a chemiluminescent agent (e.g., luminol),
produces fluorescence in proportion to the amount of bound
protein. The light is detected by a photographic film, and more
recently by CCD cameras, which create a digital image of the
blot’s localization pattern.
The use of a polyclonal antibody, that binds severalfold to
one primary antibody, as well as the enzyme catalyzed chemilu-
minescent emission of light, provides strong signal amplification
(Figure 6.2). Therefore, secondary “enhanced chemilumines-
cence” (ECL) detection is considered to be one of the most sensi-
tive detection methods for blotting analysis. By this way, down

O
NH
luminol
NH
NH2 O

+ H2 O2

protein
HRP h·

O
O
FC O
NH2 O

+ H2 O + N2

nylon primary secondary chemi- hypersensitive

membrane antibody antibody luminescence film

FIGURE 6.2. Enhanced chemiluminescence (ECL).

56 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

to 10 ng of protein, as well as less than 10 nucleic acid molecules

are detectable.

Stripping
After detection with one particular probe, this can almost com-
pletely be removed by stripping the membrane. The membrane
has to be submerged with a special stripping buffer and, in case
of Southern and Northern blots, heated to 94°C. Then the mem-
brane is ready for the next hybridization with a different probe.
This process can be repeated up to 20 times.

KEY POINTS
• Southern blots are used to identify DNA, Northern blots for
RNA, and Western blots for protein, respectively
• Molecules of DNA cut with restriction enzymes, RNA dena-
tured with formaldehyde, and protein denatured with SDS are
separated according to their size by agarose gel electrophoresis
(Southern, Northern) or SDS-PAGE (Western)
• The separated molecules are transferred to a solid nylon or
nitrocellulose membrane by capillary action (Southern, North-
ern) or electrophoresis (Western)
• The blotted membranes are blocked with extensive DNA
(Southern, Northern) or protein (Western) to prevent unspe-
cific binding of probes
• The blocked membranes are hybridized with a target-specific
ssDNA or RNA probe (Southern, Northern), or antibody probe
(Western)
• The probes are labeled with radioactivity, fluorescence dyes,
DIG or reporter enzymes
• A labeled secondary antibody directed against DIG or primary
antibody is applied in a second hybridization step
• After hybridization, unbound probes are extensively washed
off
• Bound probes are detected by autoradiography, fluorescence,
or enzymatic chemiluminescent emission of light
• Southern blotting can be used for DNA fingerprinting and
genome mapping, northern blotting for gene expression profil-
ing, and Western blotting for protein characterization, identi-
fication as well as expression analysis

References
1. Southern EM. Detection of specific sequences among DNA fragments
separated by gel electrophoresis. J Mol Biol 1975;98(3):503–517.
 6. WESTERN, NORTHERN, AND SOUTHERN BLOTTING 57

2. Alwine JC, Kemp DJ, Stark GR. Method for detection of specific RNAs
in agarose gels by transfer to diazobenzyloxymethyl-paper and hybrid-
ization with DNA probes. Proc Natl Acad Sci USA 1977;74(12):
5350–5354.
3. Renart J, Reiser J, Stark GR. Transfer of proteins from gels to diazo-
benzyloxymethyl-paper and detection with antisera: a method for
studying antibody specificity and antigen structure. Proc Natl Acad Sci
USA 1979;76(7):3116–3120.
4. Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of proteins
from polyacrylamide gels to nitrocellulose sheets: procedure and some
applications. Proc Natl Acad Sci USA 1979;76(9):4350–4354.
5. Burnette WN. Western blotting: electrophoretic transfer of proteins
from sodium dodecyl sulfate-polyacrylamide gels to unmodified nitro-
cellulose and radiographic detection with antibody and radioiodinated
protein A. Anal Biochem 1981;112(2):195–203.
Chapter 7
Fluorescent In Situ Hybridization
Fiona Campbell and John M.S. Bartlett

INTRODUCTION
In situ hybridization (ISH) technique was introduced by Gall and
Pardue1 in 1969. At that time the technique was limited by the use
of radioactively labelled probes that were subsequently visualized
by autoradiography. The development of interphase cytogenetics
in the 1980s and fluorescent labels in 19862 has seen the technology
applied in a number of fields. Although fluorescent in situ hybrid-
ization (FISH) is a valuable research tool, it is now also a technique
employed in the diagnostic laboratory. It is currently a standard
tool in cytogenetics laboratories, where it is used for the diagnosis
of hereditary disorders, chromosomal aberrations, and hemato-
logic cancer markers. More recently, the technique has been
applied to formalin-fixed, paraffin-embedded cells and tissues. The
application of FISH to detect gene amplifications (HER2 in breast
cancer),3 gene rearrangements (BCR-ABL in leukaemias),4 micro-
deletions,5 chromosomal duplication, and viral infections (HPV)
highlights the importance of this methodology, not only in the
clinical setting but in wider research applications.
This chapter serves to highlight the key points in the applica-
tions of FISH technology.

BASIC PRINCIPLES
Fluorescent in situ hybridization is based on the ability of com-
plimentary single-stranded nucleic acid molecules to hybridize
to each other, and therefore allow the demonstration of specific
genes or chromosomes in their cellular environment. The tech-
niques are simple and involve the pretreatment of the tissue or
cellular preparation to unmask the target DNA and the hybridiza-
tion of a specific DNA probe, of complimentary base sequence,
to this target.
Fluorescent in situ hybridization can be performed on either
fresh or archival tissues and unstained cytologic preparations.
Standard FISH protocols can be divided into 5 basic steps:
 7. FLUORESCENT IN SITU HYBRIDIZATION 59

1. Probe and sample preparation

2. DNA unmasking (pretreatment and digestion)
3. Probe and target DNA denaturation
4. Probe and target DNA hybridization
5. Posthybridization wash and visualization

METHODOLOGICAL ASPECTS OF FISH

Probe and Sample Preparation

Probes
The main requirement for FISH is the development of a DNA
probe with homology to the target region. In most instances,
these probes are commercially available, and these are highly
recommended, as the “in-house” development of such probes
needs to be subject to rigorous quality control measures. The
probe DNA can be labelled by nick translation, polymerase chain
reaction (PCR), random priming, or chemical conjugation. All
probes must have Cot-1 DNA added, as this represents repetitive
sequences of the human genome and addition of it to the probe
will suppress nonspecific hybridization.6 Cot-1 DNA is usually
supplied with commercial probes.
All FISH procedures using commercial probes should be carried
out according to manufacturer’s instructions for optimal results, as
they may differ slightly from the method detailed below.

Sample Preparation
Fluorescent in situ hybridization methodologies are most
commonly applied to isolated cells from whole blood or other
bodily fluids, frozen-tissue sections, or formalin-fixed paraffin-
embedded (FFPE) tissues.1,2 Cytologic preparations and frozen-
tissue sections can be prepared on to silane-coated slides or
charged slides and fixed in an acid: alcohol mix, for example,
1 part acetic acid with 3 parts methanol.
Formalin-fixed paraffin-embedded tissue sections can also be
prepared by sectioning at 3–4 µm thickness and mounting onto
silane-coated or charged slides. Formalin fixation is used to
prevent tissue degradation and to preserve morphology but can
also affect the retention of nucleic acids. Most pathology labora-
tories use neutral buffered formalin to fix tissues, and this pro-
vides an optimal platform for tissue morphology assessment.
Neutral buffered formalin acts as a strong oxidizing agent, pro-
moting divalent protein-protein and protein-nucleic acid cross-
links in fixed tissues. This fixation is followed by dehydration and
60 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

heating (in paraffin wax) to “embed” tissue for sectioning. Each

of these effects can alter nucleic acids and intracellular proteins,
and result in a reduction in the ability for macromolecules (such
as FISH probes) to enter cells. The early steps of ISH/FISH pro-
tocols are designed to reverse this by breaking down cross-links
between macromolecules and rehydrating the tissue section.

DNA UNMASKING: PRETREATMENT AND DIGESTION

De-waxing and Rehydration

These steps are common to most histological staining methods,
including routine hematoxylin and eosin (H&E) staining, various
special stains, and many immuno-histochemical staining tech-
niques. After sectioning and incubation in a 56°C oven, the tissue
sections must have the residual paraffin wax removed and need
to be rehydrated through graded alcohols to distilled water.

Acid Permeabilization
Following de-waxing and rehydration a number of different steps
may be used to increase tissue permeabilization and to allow
probe access. Each of these steps has the potential to cause tissue
damage and result in the loss of tissue morphology, causing the
technique to fail. Thus, tissue permeabilization is a balance
between allowing probe access and retaining tissue architecture.
The most frequently used pretreatment protocols involve one or
more steps, using acid, detergents, and/or reducing agents with
the aim to permeabilize tissue prior to protease digestion steps,
which break down the cellular proteins to enhance probe pene-
tration, and to reduce autofluorescence.
Most commonly, incubation in 0.2 normal hydrochloric acid
is used to remove protein from the tissue and improve probe
access. This incubation step is thought to reverse some of the
effects of formalin fixation.

Reducing Agents
Sodium metabisulphite, sodium thiocyanate,3 and MES are used to
break disulphide bonds formed by formalin fixation and aid subse-
quent protease digestion, and therefore, increase probe access.
Using both acids and reducing agents prior to protease diges-
tion ultimately reduces the amount of exposure to proteolytic
enzymes and therefore minimizes the amount of tissue damage
and any loss of morphology.

Protease Digestion
As mentioned above, this step needs to be performed with
minimal tissue damage. This is achieved by varying the times the
 7. FLUORESCENT IN SITU HYBRIDIZATION 61

tissue is exposed to the proteolytic enzymes, and by assessing the

tissue section microscopically before proceeding to the hybrid-
ization steps. The digestion can be assessed by mounting the
slides using a nuclear stain known as DAPI (4,6-diamidino-2-
phenylindole-2-hydrochloride). Protease digestion can be per-
formed using proteinase K or pepsin. The protease digestion will
remove any proteinaceous material from the tissue section, and
the extent of this digestion will depend on the nature of the tissue
being treated. Hemorrhagic tissue and those tissues with large
amounts of adipose will show areas of sparse, widespread nuclei,
whereas those tissues with dense collagen will retain much of the
stroma even after substantial digestion times. The main impor-
tance with protease digestion is the nuclear appearance. The
nuclei should be clear and evenly spread, not touching, overlying,
or clumping together. If a tissue section is under-digested, the
nuclei will be hard to see underneath the protein. If a tissue
section is over-digested, the nuclear boundaries may be lost, or
the nuclei will clump tightly together so that they do not appear
as individual nuclei.
This step is the most critical step in the entire FISH proce-
dure. One of the most common reasons for a failed result is due
to inadequate or inappropriate digestion times.3 It is recom-
mended that various digestion times be evaluated for individual
tissue sections even when protocols from manufacturers do not
include this option; frequently, commercially derived protocols
underestimated the duration of protease digestion required. It is
also extremely important that tissue digestions are routinely
checked microscopically to ensure that optimal tissue digestion
is achieved before proceeding to the hybridization step.4–6

Probe and Target DNA Denaturation

Tissue DNA denaturation can be performed using heat, chemi-
cals, or a combination of both. Commonly tissue DNA is
denatured using a solution containing formamide at 72°C for 5
minutes on a flat-bed-heated stage.
Probe DNA are generally heated to 95°C for approximately 5
minutes, although there are some commercial probes that do not
require denaturation.

Probe and Target DNA Hybridization

Hybridization between the probe DNA and the target tissue DNA
is performed, usually on a flat-bed-heated stage at 37°C–45°C, for
at least 14 hours in most instances. Humidity control is critical
in some procedures. Hybridization temperatures are chosen to
ensure maximum binding of the probe to the target DNA.
62 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

POSTHYBRIDIZATION WASH AND VISUALIZATION

Posthybridization Wash
Posthybridization washes are essential to remove any excess
unbound probe and any nonspecifically bound probe. The strin-
gency of the posthybridization wash can be determined by
salt and formamide concentrations and by temperature.7 Increas-
ing temperature and formamide concentrations, and decreasing
salt concentrations, will decrease hybridization. Each of these
methods results in removing the hydrogen bonds required for
probe-target DNA binding, and, therefore, the removal of any
nonspecifically bound probe.

Visualization and Scoring/Interpretation of Results

The results of the FISH procedure are visualized using epifluo-
rescence microscopy.

QUALITY ASSURANCE
Examples of a FISH protocol for formalin-fixed, paraffin-
embedded tissue sections are listed below.
This pretreatment method has been adapted from the Path-
Vysion pretreatment protocol (Abbott Diagnostic Labs., UK). If
many slides are to be treated, steps 2–19 can be performed on a
VP2000 automated tissue processing station, substituting the
2xSSC washes with distilled water. The VP2000 can accommo-
date up to 50 slides per run. All glassware should be rinsed with
distilled water before use.

Slide Pretreatment
1. Cut 4-µm tissue sections, pick up on silane-coated slides,
and bake overnight in a 56°C oven. Store at room temperature
until required
2. Set up two water baths, one at 37°C and one at 80°C.
Place one Coplin jar per 5 slides to be treated in each water bath.
Fill those at 80°C with 8% sodium thiocyanate (pre-treatment
solution) pH 6.5–7.0 and those at 37°C with 0.2 N HCl pH 2.0 ±
0.2 for protease digestion, but do not add the protease at this
time
3. Immerse slides into xylene for 5 minutes at room
temperature to remove the paraffin wax
4. Repeat step 3 with fresh xylene
5. Transfer slides to 99% ethanol for 5 minutes at room
temperature to remove xylene from the tissue sections
6. Repeat step 5 with fresh 99% ethanol
 7. FLUORESCENT IN SITU HYBRIDIZATION 63

Acid Permeabilization
7. Remove slides and allow them to air-dry before immersing
in 0.2 N HCl for 20 minutes at room temperature
8. Remove slides and wash in distilled water for 3 minutes
9. Wash slides in 2xSSC buffer for 3 minutes at room
temperature

Pretreatment (Reducing Agent)

10. Carefully place slides into the sodium thiocyanate
solution at 80°C and incubate for 30 minutes
11. Remove slides and rinse in distilled water. Transfer to
2xSSC buffer for 5 minutes at room temperature
12. Wash in fresh 2xSSC buffer for a further 5 minutes and,
during this step, add the protease to the 0.2 N HCl at 37°C and
mix well

Protease Digestion
13. Incubate the slides in the protease at 37°C for the
recommended digestion time. This incubation time will vary
depending on the tissue type, tissue fixation method used, and
the concentration and activity of pepsin used
14. Remove slide and immerse in 2xSSC buffer for 5 minutes
at room temperature
15. Repeat step 14 with fresh 2xSSC buffer
16. Immerse slides in 10% neutral-buffered formalin for 10
minutes at room temperature
17. Place slides in 70% alcohol for 1 minute at room
temperature
18. Place slides in 85% alcohol for 1 minute at room
temperature
19. Place slides in 99% alcohol for 1 minute at room
temperature
21. Air-dry slides. Mount using a mountant containing DAPI
(4,6-diamidino-2-phenylindole-2-hydrochloride) and apply glass
cover slips
21. Assess the tissue digestion with a 100-W fluorescence
microscope that incorporates a filter specific for the excitation
and emission wavelengths of DAPI. If digestion is optimal, proceed
to step 22. If sections are under-digested, proceed to step
22. Repeat steps 13 to 21 and reassess the digestion. If
sections are over-digested, discard the slides and repeat the
procedure using new sections, reducing the incubation time in
protease
64 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

22. Place slides into 2xSSC buffer until the cover slips wash
off. Place slides into fresh 2xSSC buffer for 5 minutes
23. Completely dry slides in 56°C oven

Denaturation
24. In a fume hood, check the pH of the denaturing solution
(49 mls formamide, 7 mls 20xSSC, 14 mls distilled water) and
apply 100 µl to each slide. Cover with temporary cover slips
made from strips of Parafilm cut to size. Place slides on a flat-
bed-heated stage (e.g., Omnislide) and incubate at 72°C for 5
minutes
25. Remove the slides from the Omnislide and transfer a
fume hood. Remove the temporary cover slips
26. Within the fume hood, place the slides in 70% alcohol for
1 minute at room temperature
27. Place slides in 85% alcohol for 1 minute at room
temperature
28. Place slides in 99% alcohol for 1 minute at room
temperature
29. Remove the slides and leave to air dry at room
temperature.

Hybridization
30. Work in reduced light; apply 10 µl of the appropriate
probe to a 22 × 26 mm cover slip. Invert the slide gently on to the
cover slip, taking care to avoid air bubbles
31. Seal the edges of the cover slip with rubber cement
32. Incubate the slides on the Omnislide with a light-shielding
lid, overnight at 37°C

Posthybridization Wash
33. Place one Coplin jar containing 50 mls of posthybridization
wash buffer (2xSSC, 0.3% NP40) per 5 slides to be washed into
a water bath set at 72°C. Fill a Coplin jar or staining dish with
posthybridization wash buffer and keep at room temperature
34. Work in reduced light; remove the slides from the
Omnislide
35. Use forceps, remove the rubber cement from around the
edges of the cover slips and place the slides into the dish of
posthybridization wash buffer at room temperature until the
cover slips fall off
36. Check the temperature of the posthybridization wash
buffer is at 72 ± 1°C before proceeding
 7. FLUORESCENT IN SITU HYBRIDIZATION 65

37. Remove slide from the room temperature buffer and

drain off excess fluid. Place into the 72°C buffer for 2 minutes.
Do not put more than 5 slides into each Coplin jar
38. Remove slides from the posthybridization wash buffer
and allow to air dry in the dark
39. Mount slides in mounting media containing DAPI. Seal
the edges of the cover slip with nail varnish

Visualization and Scoring Slides

40. Use a fluorescent microscope, scan the slide and identify
areas for scoring
41. Score the slides using the preferred method. In most
cases, this will involve counting the appropriate signals within
a predefined number of nonoverlapping cell nuclei. This will
produce an overall ratio for the whole slide, a result for the
average signals per number of cells counted

References
1. Watters AD, Going JJ, Cooke TG, Bartlett JMS. Chromosome 17 aneu-
somy is associated with poor prognostic factors in invasive breast
carcinoma. Breast Cancer Res Treat 2003;77:109–114.
2. Bartlett JMS. Pharmacodiagnostic testing in breast cancer: focus on
HER2 and trastuzumab therapy. Am J Pharmacogenomics 2005;5:
303–315.
3. Watters AD, Bartlett JMS. Fluorescence in situ hybridization in paraf-
fin tissue sections: pretreatment protocol. Mol Biotechnol 2002;21:
217–220.
4. Watters AD, Ballantyne SA, Going JJ, Grigor KM, Bartlett JMS.
Aneusomy of chromosomes 7 and 17 predicts the recurrence of
transitional cell carcinoma of the urinary bladder. BJU International
2000;85:42–47.
5. Bartlett JM, Going JJ, Mallon EA, et al. Evaluating HER2 amplifica-
tion and overexpression in breast cancer. J Pathol 2001;195:
422–428.
6. Edwards J, Krishna NS, Mukherjee R, Watters AD, Underwood MA,
Bartlett JMS. Amplification of the androgen receptor may not explain
development of androgen independent prostate cancer. Br J Urol
2001;88:1–10.
Chapter 8
Quantitative Reverse Transcriptase
Polymerase Chain Reaction
Lyndon M. Gommersall, M. Arya, Prabhabhai S. Patel,
and H.R.H. Patel

INTRODUCTION
Since the first documentation of real-time polymerase chain reac-
tion (PCR),1 it has been used for an increasing and diverse number
of applications, including mRNA expression studies, DNA copy
number measurements in genomic or viral DNAs,2–7 allelic
discrimination assays,8,9 expression analysis of specific splice
variants of genes10–13 and gene expression in paraffin-embedded
tissues,14,15 and laser captured microdissected cells.13,16–19 There-
fore, quantitative reverse transcriptase polymerase chain reaction
(Q-RT-PCR) is now essential in molecular diagnostics to quanti-
tatively assess the level of RNA or DNA in a given specimen. Q-
RT-PCR enables the detection and quantification of very small
amounts of DNA, cDNA, or RNA, even down to a single copy. It
is based on the detection of fluorescence produced by reporter
probes, which varies with reaction cycle number. Only during the
exponential phase of the conventional PCR reaction is it possible
to extrapolate back in order to determine the quantity of initial
template sequence. The “real-time” nature of this technology per-
tains to the constant monitoring of fluorescence from specially
designed reporter probes during each cycle. Due to inhibitors of
the polymerase reaction found with the template, reagent limita-
tion or accumulation of pyrophosphate molecules, the PCR reac-
tion eventually ceases to generate template at an exponential rate
(i.e., the plateau phase), making the end point quantitation of
PCR products unreliable in all but the exponential phase. Exam-
ples of fluorescent reporter molecules include dyes that bind to
the double-stranded DNA (i.e., SYBR® Green) or sequence-
specific probes (i.e., TaqMan® products or Molecular Beacons
Probes). The automation of the reaction as a whole has enabled
Q-RT-PCR assays to be easy to perform with higher sensitivity
and more specificity.
 8. QUANTITATIVE REVERSE TRANSCRIPTASE PCR 67

This chapter introduces important aspects of Q-RT-PCR,

including comparison between conventional PCR and Q-RT-PCR
methodology, fluorogenic and sequence-specific probes, methods
for quantitation of Q-RT-PCR products, common terminology,
and a review of instrumentation.

UNDERSTANDING THE FUNDAMENTALS

Quantitative reverse transcriptase polymerase chain reaction is
the reliable detection and measurement of products generated
during each cycle of the PCR process, which are directly propor-
tional to the amount of template prior to the start of the PCR
process. Holland and co-workers20 demonstrated that the ther-
mostable enzyme Thermus aquaticus (i.e., Taq) DNA polymerase
had 5′ to 3′ exonuclease activity. This group also showed that
cleavage of a specifically designed target probe during PCR by
the 5′ nuclease activity of Taq polymerase can be used to detect
amplification of the amplified product.20 An oligonucleotide
probe, which was designed to hybridize within the target
sequence, was introduced into the PCR assay. This probe was
labeled with 32P at its 5′ end and was nonextendable at its 3′ end
to ensure it could not act as a primer. Annealing of probe to one
of the PCR product strands during the course of amplification
generated a substrate suitable for exonuclease activity. Also,
during amplification, the 5′ to 3′ exonuclease activity of Taq DNA
polymerase (when the enzyme extended from an upstream primer
into the region of the probe) degraded the probe into smaller
fragments that could be differentiated from undegraded probe.
This dependence on polymerization ensured that cleavage of the
probe occurred only if the target sequence was being amplified.
After PCR, cleavage of the probe was measured by using thin-
layer chromatography to separate cleavage fragments from intact
probe. The introduction of dual-labeled oligonucleotide fluoro-
genic probes allowed the elimination of post-PCR processing for
the analysis of probe degradation.21 The probe has a reporter
fluorescent dye at the 5′ end and a quencher dye attached to
the 3′ end. While the probe is intact, the close proximity of the
quencher significantly decreases the fluorescence emitted by the
reporter dye. A fluorescence signal is only emitted on cleavage of
the probe, based on the fluorescence resonance energy transfer
(FRET) principle.22
In the real-time quantitative TaqMan® assay, a fluorogenic
nonextendable probe, termed the “TaqMan” probe, is used
(Figure 8.1).23 The probe has a fluorescent reporter dye attached
to its 5′ end and a quencher dye at its 3′ terminus. If the target
68 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

When the TaqMan probe is intact, the reporter

and quencher stay close to each other, which
prevents the emission of any fluorescence

Reporter TaqMan Quencher

fluorophore probe
Primer
The primer and TaqMan 5′
probe anneal to the
complementary DNA strand 3′
following denaturation cDNA

Reporter
After hybridization and during Quencher
fluorophore
the extension phase, the 5′
endonuclease activity of the
Taq DNA polymerase cleaves
the probe which separates 5′
reporter and quencher dyes
and fluorescence is detected. 3′

FIGURE 8.1. Hydrolysis probes (e.g., TaqMan assay). (From Ref. 73)

sequence is present, the fluorogenic probe anneals downstream

from one of the primer sites and is cleaved by the 5′ nuclease
activity of the Taq polymerase enzyme during the extension
phase of the PCR. While the probe is intact, FRET occurs, and
the fluorescence emission of the reporter dye is absorbed by the
quenching dye. Cleavage of the probe by Taq polymerase during
PCR separates the reporter and quencher dyes, thereby increas-
ing the fluorescence from the former. Additionally, cleavage
removes the probe from the target strand, allowing primer exten-
sion to continue to the end of template strand, thereby not inter-
fering with the exponential accumulation of PCR product.
Additional reporter dye molecules are cleaved from their respec-
tive probes with each cycle, leading to an increase in fluorescence
intensity proportional to the amount of amplicon produced. The
various available chemistries for real-time PCR are described
later in this review.
Using any of the developed chemistries, the increase in fluo-
rescence emission during the PCR reaction can be detected in
real time by a modified conventional PCR thermocycler. The
computer software constructs amplification plots using the fluo-
 8. QUANTITATIVE REVERSE TRANSCRIPTASE PCR 69

rescence emission data that are collected during the PCR ampli-
fication (Figure 8.2). Figure 8.2 demonstrates a representative
amplification plot and defines the important terms associated
with it.

• Baseline: The baseline is defined as the PCR cycles in which a

reporter fluorescent signal is accumulating but is beneath the
limits of detection of the instrument. By default, the computer
software sets the baseline from cycles three to 15; however,
this often needs to be changed manually utilizing software
supplied with each particular thermocycler.
• ∆Rn: A computer software program calculates a ∆Rn using the
equation Rn = Rnf − Rnb, where Rnf is the fluorescence emis-
sion of the product at each time point and Rnb is the fluores-
cence emission of the baseline.23,24 The ∆Rn values are plotted

2,000,000

∆Rn Plateau
Sample
1,000,000
Log/exponential phase

Threshold Ct

0 No template
Baseline

0 20 40
PCR cycle number

∆Rn = Fluorescence emission of the product at each time point –

fluorescence emission of the baseline

Ct = Threshold cycle

FIGURE 8.2. Model of a single amplification plot illustrating the nomen-

clature commonly used in real-time quantitative PCR. (From Ref. 73)
70 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

versus the cycle number. During the early cycles of PCR ampli-
fication, ∆Rn values do not exceed the baseline.
• Threshold: An arbitrary threshold is chosen by the computers,
based on the variability of the baseline. It is calculated as ten
times the standard deviation of the average signal of the base-
line fluorescent signal between cycles three to 15. A fluorescent
signal that is detected above the threshold is considered a real
signal that can be used to define the threshold cycle (Ct) for a
sample. If required, the threshold can be manually changed for
each experiment so that it is in the region of exponential ampli-
fication across all amplification plots.
• Ct: This is defined as the fractional PCR cycle number at which
the reporter fluorescence is greater than the minimal detection
level (i.e., the threshold). The Ct is a basic principle of real-time
PCR and is an essential component in producing accurate and
reproducible data.1

The presence of more template at the start of the reaction

leads to a fewer number of cycles passing before reaching the
point at which the fluorescent signal is recorded as statistically
significant above background.24 This Ct value will always occur
during the exponential phase of target amplification, which
occurs during the early cycles of PCR. As reaction components
become limiting, the rate of target amplification decreases until
the PCR reaction is no longer generating template at an expo-
nential rate (plateau phase), and there is little or no increase in
PCR product. This is the main reason why Ct is a more reliable
measure of starting copy number than an endpoint measurement
of the amount of accumulated PCR product. During the expo-
nential phase, none of the reaction components is limiting, and
therefore, Ct values are very reproducible for replicate reactions
with the same starting copy number.

DISCUSSION OF METHODS TO QUANTIFY REAL-TIME

POLYMERASE CHAIN REACTION RESULTS
Standard-curve or absolute quantitation: As shown by Higuchi
and co-workers, the plot of the log of initial target copy number
for a set of known standards (five- or tenfold serial dilution)
versus Ct is a straight line (the standard curve).1 Quantitation of
the amount of target in the “unknown” samples of interest is
accomplished by measuring Ct and using the standard curve to
determine starting copy number. The most common source of a
known sample is a plasmid for the gene of interest, and the stan-
dard curve is generated based on a serial dilution of a starting
 8. QUANTITATIVE REVERSE TRANSCRIPTASE PCR 71

amount. Another option, and easier to generate if a plasmid

is unavailable, is the use of a synthetic single-stranded sense
oligonucleotide for the entire amplicon. The advantage of this
approach is that it significantly simplifies the process of obtain-
ing a standard curve for amplicons up to 100 bp, which encom-
passes most real-time PCR amplicons. Furthermore, it is also less
susceptible to bias when quantified by a spectrophotometer due
to the relative purity of the oligonucleotide. Together with the
greater precision of measurement of the standard and the possi-
bility of calculating the moles of oligonucleotide (hence, number
of copies); it is possible to approximate the number of copies of
a template in an unknown sample, although not in terms of
absolute copy number. One final option for a standard curve is
to use a cell line with a known copy number or expression level
of the gene of interest. The standard curve method is used in
circumstances when absolute quantitation is critical for the
investigator (e.g., when measuring a small number of genes in
either a few or many samples)25,26 or in quantitation of viral
load.27–29
Relative quantitation: Relative quantitation is also known as
the comparative threshold method (2-Ct method). This method
eliminates the need for standard curves, and mathematical equa-
tions are used to calculate the relative expression levels of a
target relative to a reference control or calibrator, such as a
nontreated sample or RNA from normal tissue. The amount of
target, normalized to an endogenous housekeeping gene and
relative to the calibrator, is then given by 2-Ct, where Ct = Ct
(sample) − Ct (calibrator), and Ct is the Ct of the target gene
subtracted from the Ct of the housekeeping gene. The equation
thus represents the normalized expression of the target gene in
the unknown sample, relative to the normalized expression of the
calibrator sample. For this calculation to be valid and in order
to obtain reliable results, it is imperative that the amplification
efficiencies of the housekeeping and target gene are approxi-
mately equal and at or above 90%. This can be established by
looking at how Ct (of both sample and calibrator) varies with
template dilution. If the plot of complementary DNA (cDNA)
dilution versus Ct is close to zero, it implies that the efficiencies
of the target and housekeeping genes are very similar. If a house-
keeping gene cannot be found whose amplification efficiency is
similar to the target, the standard curve method is then prefera-
ble. Alternatively, new primers can be designed and/or optimized
to achieve a similar efficiency for the target and housekeeping
gene amplicons.
72 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

CONTROLS IN QUANTITATIVE REVERSE TRANSCRIPTASE

POLYMERASE CHAIN REACTION: USE OF THE
“HOUSEKEEPING” GENE
In real-time quantitative PCR experiments specific errors will be
introduced due to minor differences in the starting amount of
RNA, quality of RNA, or differences in efficiency of cDNA syn-
thesis and PCR amplification. In order to minimize these errors
and correct for sample-to-sample variation, a cellular RNA is
simultaneously amplified with the target, which serves as an
internal reference against which other RNA values can be nor-
malized. The most common genes used for normalization, termed
“housekeeping” genes, are β-actin, a cytoskeletal protein, and
glceraldehyde-3-phosphate dehydrogenase (GAPDH), a glycolytic
enzyme,30 and ribosomal RNA (rRNA). These genes should theo-
retically be expressed at a constant level among different tissues
of an organism, at all stages of development, and their expression
levels should also remain relatively constant in different experi-
mental conditions.
However, none of these housekeeping genes are ideal. It has
been shown that GAPDH expression levels are altered by glucose,
insulin, heat shock, and cellular proliferation; and β-actin levels
may also be modified by experimental treatments.31–35 rRNA pro-
duction is less likely to vary under conditions affecting mRNA
transcription.36,37 However, it is not always a good representative
of total mRNA population in a cell as rRNA is expressed at a
much higher level than mRNA.
Other alternative housekeeping genes have been proposed,
but none have been entirely satisfactory, and no single unequivo-
cal reference gene has yet been identified. Some authors have
suggested the use of several housekeeping genes in a single exper-
iment and that the mean expression of these multiple housekeep-
ing genes can be used for normalization.38 Importantly, selection
of the housekeeping gene for each specific experiment should be
made very carefully as the reliability of the results depends on
the choice of the most relevant housekeeping gene according to
the cells of interest and specific experimental treatments.

AMPLICON DETECTION STRATEGIES

Two general chemistries are available. These include double-
stranded (ds) DNA-intercalating agents (DNA-binding dyes) and
fluorescent probes. The former includes SYBR Green I or ethid-
ium bromide and is the simplest and most cost-effective method,
as amplicon-specific labeled hybridization probes are not
required. SYBR Green I only fluoresces when intercalated into
 8. QUANTITATIVE REVERSE TRANSCRIPTASE PCR 73

dsDNA. The intensity of the fluorescence signal is therefore

dependent on the quantity of dsDNA present in the reaction. The
main disadvantage of this method is that it is not specific, because
the dye binds to all dsDNAs formed during the PCR reaction (i.e.,
nonspecific PCR products and primer-dimers).
With fluorogenic probes, nonspecific amplification due to
mispriming or primer-dimer artifact does not generate signal, as
specific hybridization between probe and template is necessary
for fluorescence emission. Also, fluorogenic probes can be labeled
with different and distinguishable reporter dyes, thus allowing
the detection of amplicons that may have been produced by one
or several primer pairs in a single PCR reaction—termed multi-
plex real-time PCR. However, different probes must be developed
to detect different sequences. The various chemistries are now
described in more detail.
Double-stranded DNA-intercalating agents (DNA-binding
dyes): SYBR Green I is a nonsequence-specific fluorogenic minor
groove DNA-binding dye that intercalates into dsDNA (it does
not bind to single-stranded DNA). SYBR Green 1 exhibits little
fluorescence when unbound in solution but emits a strong fluo-
rescent signal upon binding to dsDNA.39 An increase in the fluo-
rescence signal occurs during polymerization, and this decreases
when DNA is denatured. Fluorescent measurements are per-
formed at the end of the elongation step of each PCR cycle to
monitor the increasing amount of amplified DNA. The advantage
of this technique is that it is relatively cheap, as it can be used
with any pair of primers for any target. However, as the presence
of any dsDNA generates fluorescence, specificity of this assay is
greatly decreased due to amplification of nonspecific PCR prod-
ucts and primer-dimers.40 Generating and comparing melting
curves (plotting fluorescence as a function of temperature) using
the LightCyclerTM (Roche Molecular Diagnostics) (or RotorGene,
Smart Cycler, iCycler, Mx4000) is one method of increasing the
specificity of the reaction.40 A characteristic melting peak at the
melting temperatureTM of the amplicon will distinguish it from
amplification artifacts that melt at lower temperatures at broader
peaks. It is possible to set the software to acquire fluorescence
above the primer-dimers’ melting temperature but below that of
the target. Another controllable problem is that longer amplicons
create a stronger signal. Usually, SYBR Green is used in single-
plex reactions; however, when coupled with melting-point analy-
sis, it can be used for multiplex reactions. The SYBR Green I
reaction has been used for many applications (e.g., viral load
detection41 and cytokine quantifaction.42–44
74 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

Hydrolysis probes (e.g., TaqMan probes): This chemistry

has already been outlined earlier in this review (Figure 8.1). A
forward and reverse primer and a probe are used. The efficiency
of the assay is mainly dependent on 5′ to 3′ nuclease activity; the
most commonly used enzyme is Taq polymerase,20 but any
enzyme with 5′ nuclease activity can be used.45 The oligonucle-
otide probe has a covalently bonded fluorescent reporter dye and
quencher dye at the 5′ and 3′ ends, respectively. Various fluores-
cent reporter dyes are in use including 6-carboxyfluorescein
(FAM), tetrachloro-6-carboxyfluorescein (TET), hexacholoro-
6-carboxyfluorescein (HEX), or VIC. Quenchers include either
6-carboxytetramethylrhodamine (TAMRA) or 4-(dimethylamino-
azo) benzene-4-carboxylic acid (DABCYL). When the probe is
intact, the proximity of the reporter and quencher dyes permits
FRET, and fluorescence emission does not occur. During PCR
amplification, the probe anneals to the target, and Taq poly-
merase cleaves the probe, allowing an increase in fluorescence
emission. The increase in fluorescence intensity is directly pro-
portional to the amount of amplicon produced. The TaqMan
chemistry is the most widely used real-time PCR assay and has
been used for multiple purposes.32,46,47
TaqMan minor groove-binding probes have more recently
been developed. In this chemistry, the standard TAMRA quencher
at the 3′ end is replaced by a nonfluorescent quencher, and a
minor groove-binder molecule is also incorporated at the 3′ ter-
minus. The latter stabilizes the probe-target complex by folding
into the minor groove of the dsDNA. Additionally, the Tm of the
probes is increased, allowing the use of very short oligoprobes
(14 nucleotides in length) and providing more accurate allelic
discrimination. Thus, TaqMan minor groove-binding probes are
ideal for detecting single nucleotide polymorphisms48,49 and for
the quantitative analysis of methylated alleles.50
Dual hybridization probes: This method has been convinc-
ingly validated in studies using the LightCycler instrument. Two
hybridization probes are used—one carries a donor fluorophore
at its 3′ end, and the other is labeled with an acceptor fluorophore
at its 5′ end. After the denaturation step, both probes hybridize
to their target sequence in a head-to-tail arrangement during the
annealing step. This brings the two dyes in close proximity,
allowing FRET. The donor dye in one of the probes transfers
energy, allowing the other one to dissipate fluorescence at a dif-
ferent wavelength. The measured fluorescence is directly propor-
tional to the amount of DNA synthesized during the PCR reaction.
The specificity of this reaction is therefore increased as a fluo-
 8. QUANTITATIVE REVERSE TRANSCRIPTASE PCR 75

rescent signal is only detected when two independent probes

hybridize to their correct target sequence. This method has been
widely used for detection of minimal residual disease after
therapy51,52 and viral load quantification.53,54
Molecular beacons: Molecular beacons also contain cova-
lently bound fluorescent and quenching dyes at either end of a
single-stranded DNA molecule. However, they are also designed
to adopt a hairpin or stem-and-loop structure while free in solu-
tion to bring the fluorescent dye and the quencher in close prox-
imity for FRET to occur.55 The loop portion of the molecule is
complementary to the target nucleic acid molecule, and the stem
is formed by the annealing of complementary arm sequences on
the ends of the probe sequence. The close proximity of the fluo-
rophore and the quencher in this hairpin configuration sup-
presses reporter fluorescence. When the probe sequence in the
loop hybridizes to a complementary nucleic acid target sequence
during the annealing step, a conformational change occurs that
forces the stem apart. This results in a linear structure and thus
separation of the fluorophore from the quencher dye (FRET does
not occur) and an increase in fluorescence emission. A new
hybridization takes place in the annealing step of each cycle, and
the intensity of the resultant fluorescence indicates the amount
of accumulated amplicon at the end of the previous cycle. Molec-
ular beacons remain intact during PCR, and they must rehybrid-
ize to the target sequence each cycle for fluorescence emission.
Molecular beacons are especially suitable for identifying point
mutations.56–58
Scorpions: Similar to molecular beacons, scorpions adopt
a stem-and-loop configuration with a 5′ fluorophore and 3′
quencher. The specific probe sequence is held within the hairpin
loop, which is attached to the 5′ terminus of a PCR primer
sequence by a nonamplifiable monomer (termed the PCR
stopper). This chemical modification prevents PCR from copying
the stem-loop sequence of the scorpion primer. During PCR,
scorpion primers are extended to form amplicon. In the anneal-
ing phase, the specific probe sequence in the scorpion tail curls
back to hybridize to the complementary target sequence in the
amplicon, thus opening up the hairpin loop. This prevents the
fluorescence from being quenched, and a signal is observed.59 As
the tail of the scorpion and the amplicon are now part of the
same strand of DNA, the interaction is intramolecular.
The benefits of scorpions derive from the fact that the
probe element is physically coupled to the primer element,
which means that the reaction leading to signal generation is a
76 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

unimolecular event. This contrasts to the bimolecular collisions

required by other technologies such as TaqMan or molecular
beacons. The benefits of a unimolecular rearrangement are sig-
nificant in that the reaction is effectively instantaneous and the
fluorescence signal much stronger. Also better discrimination
and specificity are achieved using scorpions. Scorpion probes
have been used for viral load and mutation detection.60,61
Duplex scorpions are a modification of scorpions. However,
in contrast to scorpions (or molecular beacons), the fluorophore
and quencher dye is separated onto different and complementary
oligonucleotides. The advantage of duplex scorpions is the sig-
nificantly greater separation between the quencher and reporter
fluorophore, which decreases fluorophore quenching when the
probe is bound to the target, resulting in better signal intensity
compared with conventional scorpions.62

DESIGNING A PRIMER, PROBE, AND AMPLICON

Great care should go into the design of an assay. Primers, probes,
and amplicons are designed to very exacting specifications, and
the TaqMan system provides its own primer/probe design soft-
ware from Applied Biosystems known as Primer Express, which
is one of the most widely used oligonucleotide design programs
for developing real-time quantitative PCR assays. Primer3, a free
program from Massachusetts Institute of Technology (MA, USA),
can also be used to generate good real-time PCR assays, includ-
ing designs incorporating an internal hybridization probe. The
amplicon for the PCR product should be as small as reasonably
possible, usually 50–150 bp in length for designs using hybridiza-
tion probes (and less than 300 bp for SYBR Green assays). Shorter
amplicons amplify more efficiently and are more tolerant of reac-
tion conditions. The optimal length for single-stranded primers
is 15–20 bases with a G/C content of 20%–80%. Their Tm should
be in the range of 68°C–70°C for TaqMan primers. Molecular
beacon and hybridization probe-associated primers can have a
wider range of Tm, but the Tm of any one pair should be similar
(i.e., not differ by more than 1°C–2°C). Nonspecific priming is
minimized by selecting primers that have only one or two G/Cs
within the last five nucleotides at the 3′ end. If using a SYBR
Green I approach, the PCR primers must not form an appreciable
amount of primer-dimer bands. A melting curve analysis of each
product is needed to ensure that the fluorescent signal observed
is from the desired PCR product. In mRNA expression assays
using a hybridization probe, the probe sequence should span an
exon/exon boundary if possible. Having the probe Tm 8°C–10°C
 8. QUANTITATIVE REVERSE TRANSCRIPTASE PCR 77

higher than that of the primers ensures that the probe is

fully hybridized during primer extension. TaqMan probes should
not contain a G at their 5′ ends due to the quenching effect of a
G in this position on reporter fluorescence, even after probe
cleavage.

PUSHING THE TECHNOLOGY FURTHER: MULTIPLEX

QUANTITATIVE REVERSE TRANSCRIPTASE
POLYMERASE CHAIN REACTION
The term multiplex Q-RT-PCR is used to describe the use of
multiple fluorogenic probes for the discrimination of multiple
amplicons in a single tube. The main advantages of multiplexing
over single-target analysis are the ability to provide internal con-
trols, lower reagent costs and preservation of precious low quan-
tity samples. The main restrictions of this technique have been
the limited number of available fluorophores, fluorescence emis-
sion from quenching dyes, and the common use in real-time
instruments of a monochromatic light source. The introduction
of nonfluorescent quenchers, which have no inherent fluores-
cence, has been a breakthrough that has allowed an increase in
the number of spectrally discernable fluorogenic probes used per
reaction. Initial Q-RT-PCR instrumentation contained optimized
filters to minimize overlap of the emission spectra from the fluo-
rophores. Newer systems have used either multiple light-emitting
diodes, which span the whole visible spectrum, or a tungsten
lamp, which emits light over a broad range of wavelengths.
However, despite these advancements, only four-color multiplex
reactions are usually possible,63,64 of which one color may be used
for an internal control. One recent development is the introduc-
tion of combinatorial fluorescence energy transfer tags,65,66 which
will help to boost the development of multiplex real-time PCR.

EQUIPMENT REVIEW
There are a variety of instruments available on the market, each
of which has its own individual characteristics. Great care should
be taken when choosing which instrument to buy, and it is
important to match the instruments capabilities with laboratory
needs. Cost should not be the only factor when making a choice;
the cheaper models cannot compensate for the variance in the
optics and therefore are not capable of detecting smaller differ-
ences. The higher-throughput instrument may be more than is
needed. The ABI Prism® 7700 Sequence Detection System (SDS)
from Applied Biosystems was the first commercially available
thermocycler for real-time PCR, but has now been discontinued.
78 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

Continuous fluorescence wavelength laser light detection from

500–660 nm allowed multiplex PCR on this machine.
The ABI Prism 7700 has more recently been replaced by the
ABI Prism 7900HT, which has similar specifications to the 7700
SDS but is completely automated and designed especially for
very high-throughput applications (384 samples per run). Another
recent introduction is the less expensive ABI Prism 7000 SDS. It
retains the Peltier-based 96-well block thermal cycling format of
the ABI 7700, but replaces the laser with a tungsten–halogen
lamp that simultaneously illuminates all sample wells. The soft-
ware supplied with the instrument is much more user friendly
and is Microsoft Windows–based, which allows easy export of
data and amplification plots. One of the major advantages of the
ABI instruments is the collection of data from a passive reference
signal to normalize each reaction for variances in the optics of
the system. In addition, Applied Biosystems have launched the
Applied Biosystems 7300 and 7500 Real Time PCR systems,
which represent less expensive alternatives.
The low-priced LightCycler from Roche Molecular Biochemi-
cals induces fluorescence excitation by a blue light-emitting
diode that is read by three silicon photodiodes with different-
wavelength filters, allowing detection of spectrally distinct fluo-
rophores. Therefore, multiplex PCR can be performed. A complete
PCR run of 30–40 cycles is performed in 20–30 min, but only a
limited number of samples (maximum 32) can be analyzed
simultaneously. As the LightCycler analyzes the specificity of the
results by performing melting curves, it makes the use of dsDNA-
binding dyes such as SYBR Green I more reliable. However, as
samples need to be in capillaries rather than tubes, it is less
practical for the investigator.
The iCycler iQ from BioRad Instruments has a tungsten-
halogen lamp allowing excitation of a wide range of fluorophores
(400–700 nm). It is able to multiplex four different fluorophores
per sample tube. Also, it has an optical module, allowing fluores-
cence emission to be viewed during the course of PCR amplifica-
tion. Furthermore, the 96 samples are tracked simultaneously,
thereby providing a fast assay. A recently launched module allows
it to amplify 384 samples at any one time.
A new option is the Mx4000® Multiplex from Stratagene.
This sequence detector instrument is able to detect multiple fluo-
rescence PCR chemistries, including TaqMan and hybridization
probes, and molecular beacons. The light source for the Mx4000
system is a quartz tungsten–halogen lamp that generates a broad
excitation range of 350–750 nm, and there are four photomulti-
 8. QUANTITATIVE REVERSE TRANSCRIPTASE PCR 79

plier tubes with a detection range of 350–830 nm. The instrument

is ideal for performing multiplex PCR. Importantly, the system
contains an integrated personal computer that operates indepen-
dently from the instrument’s embedded microprocessor, which
gives some protection against data loss.
The Smart Cycler System has recently become available from
Cepheid. The system can be operated with molecular beacons,
scorpions, hybridization probes, TaqMan probes, or SYBR Green
I. An advantage of this system is its high flexibility, as it contains
16 different modules. Each module can be individually pro-
grammed and has its own optical subsystem, and can detect four
different fluorophores in one reaction. Different operators can
define the parameters for each reaction and different runs can be
carried out at the same time for individual experiments. A disad-
vantage of the basic system is the small sample number (maximum
16); however, this can now be increased to 96 wells per run.
The Rotor GeneTM 3000, designed by Corbett Research,
is a centrifugal thermal cycler comparable with the LightCycler.
It uses four separate light-emitting diode light sources that
excite at 470, 530, 585 and 625 nm. Excitation is detected using
six filters and photomultipliers at 510, 555, 610, 660, 580, and
610 nm. The design of this instrument is radically different from
all other instruments: the real-time reactions are carried out in
standard microfuge tubes inside a 36- or 72-well rotor that spins
at 500 rpm. This is meant to remove any temperature equilibra-
tion time and nonuniformity, and sample-to-sample variation of
less than 0.01°C is claimed.

CONCLUSION
The introduction of real-time PCR technology has revolutionized
the field of molecular diagnostics and has enabled the shift of
molecular diagnostics toward a high-throughput, automated with
lower turnaround times. It allows the sensitive, specific, and repro-
ducible quantification of mRNA. Real-time PCR assays are charac-
terized by a wide dynamic range of quantification of 7–8 logarithmic
decades, a high technical sensitivity (<5 copies) and a high preci-
sion (<2% standard deviation).32 Also, no post-PCR steps are
required, thus avoiding the possibility of cross contamination due
to PCR products. The disadvantages of real-time quantitative PCR
when compared with conventional PCR include the fact that:

• Amplicon size cannot be monitored without opening the

system
• The limited multiplex capabilities of existing instruments
80 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

• The incompatibility of several systems with some fluorogenic

chemistries

Real-time PCR technology is only as reliable as the accompa-

nying controls and associated quality assurance programs. This
includes the quality of standards and choice of housekeeping gene
(the search for the ideal housekeeping gene or protocol is ongoing),
the use of suitably controlled standard curves and the need to fully
optimize, validate, and evaluate each and every new assay against
previously standardized assays. Without such care, real-time PCR
will provide an enormous amount of fast but inaccurate data.
The contemporary competition for Q-RT-PCR technology is
microarray. However, due to current microarray technologies
requiring a large amount of starting material and displaying a
limited dynamic range for quantification expression levels of
selected genes, true quantification experiments will continue to
be conducted using real-time PCR methods.67,68 Therefore, a
combination of both technologies, in which the screening of the
involved genes is performed by microarrays and the precise
quantification and high throughput screening is performed
by real-time PCR, is the ideal method. Similarly, real-time
PCR technology will continue to be combined with advanced
microdissection techniques13,16–19 or nucleic acids obtained from
paraffin-fixed archival samples.14,15 The detection and analysis of
minimal residual disease51,69 and viral loads will remain an
important application. Also, it will be possible to measure gene
expression or DNA copy number in specific cells that are isolated
with difficulty and are present in only very small numbers. Com-
bining techniques for sorting fetal cells or DNA from the mater-
nal circulation with real-time PCR will enable early prenatal
diagnostics of numerous congenital disorders using minimally
invasive procedures.70–72 Real-time techniques will inevitably be
used in the analysis of clinical samples to aid clinicians in prog-
nosis and management of patients with a variety of diverse
disease states. Limiting this application at present is the lack of
universal agreement on basic issues such as quality and quantity
control of RNA, storage standards, guidelines for analysis and
reporting of results, and standardization of protocols. These
assays are likely to be increasingly utilized as an important area
of molecular diagnostics in the future.

References
1. Higuchi R, Fockler C, Dollinger G, Watson R. Kinetic PCR analysis:
real-time monitoring of DNA amplification reactions. Biotechnology
1993;11:1026–1030.
 8. QUANTITATIVE REVERSE TRANSCRIPTASE PCR 81

2. Kariyazono H, Ohno T, Ihara K, et al. Rapid detection of the 22q11.2

deletion with quantitative real-time PCR. Mol Cell Probes 2001;
15:71–73.
3. Nigro JM, Takahashi MA, Ginzinger DG, et al. Detection of 1p
and 19q loss in oligodendroglioma by quantitative microsatellite
analysis, a real-time quantitative PCR assay. Am J Pathol 2001;4:
1253–1262.
4. Ginzinger DG, Godfrey TE, Nigro J, et al. Measurement of DNA copy
number at microsatellite loci using quantitative PCR analysis. Cancer
Res 2000;60:5405–5409.
5. Ingham DJ. The study of transgene copy number and organization.
Methods Mol Biol 2005;286:273–290.
6. Bai RK, Perng CL, Hsu CH, Wong LJ. Quantitative PCR analysis of
mitochondrial DNA content in patients with mitochondrial disease.
Ann NY Acad Sci 2004;1011:304–309.
7. Desire N, Dehee A, Schneider V, et al. Quantification of human
immunodeficiency virus type 1 proviral load by a TaqMan real-time
PCR assay. J Clin Microbiol 2001;39:1303.
8. Johnson VJ, Yucesoy B, Luster MI. Genotyping of single nucleotide
polymorphisms in cytokine genes using real-time PCR allelic dis-
crimination technology. Cytokine 2004;27:135–141.
9. Petersen K, Vogel U, Rockenbauer E, et al. Short PNA molecular
beacons for real-time PCR allelic discrimination of single nucleotide
polymorphisms. Mol Cell Probes 2004;18:117–122.
10. Elson D, Thurston G, Huang E, et al. Quiescent angiogenesis in
transgenic mice expressing constitutively active hypoxiainducible
factor-1a. Genes Dev 2001;15:2520.
11. Schmittgen TD, Teske S, Vessella RL, True LD, Zakrajsek BA. Expres-
sion of prostate specific membrane antigen and three alternatively
spliced variants of PSMA in prostate cancer patients. Int J Cancer
2003;107:323–329.
12. Caberlotto L, Hurd YL, Murdock P, et al. Neurokinin 1 receptor and
relative abundance of the short and long isoforms in the human
brain. Eur J Neurosci 2003;17:1736–1746.
13. Sethi N, Palefsky J. Transcriptional profiling of dysplastic lesions in
K14-HPV16 transgenic mice using laser microdissection. FASEB J
2004;18:1243–1245.
14. Godfrey TE, Kim SH, Chavira M, et al. Quantitative mRNA expres-
sion analysis from formalin-fixed, paraffin-embedded tissues using 5′
nuclease quantitative reverse transcription-polymerase chain reac-
tion. J Mol Diagn 2000;2:84–91.
15. Andreassen CN, Sorensen FB, Overgaard J, Alsner J. Optimization
and validation of methods to assess single nucleotide polymorphisms
(SNPs) in archival histological material. Radiother Oncol 2004;72:
351–356.
16. Glockner S, Lehmann U, Wilke N, Kleeberger W, Langer F, Kreipe
H. Detection of gene amplification in intraductal and infiltrating
breast cancer by laser-assisted microdissection and quantitative real-
time PCR. Pathobiology 2000;68:173–179.
82 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

17. Ehrig T, Abdulkadir SA, Dintzis SM, Milbrandt J, Watson MA. Quan-
titative amplification of genomic DNA from histological tissue sec-
tions after staining with nuclear dyes and laser capture microdissection.
J Mol Diagn 2001;3:22–25.
18. Fink L, Seeger W, Ermert L, et al. Real-time quantitative RTPCR after
laser-assisted cell picking. Nat Med 1998;4:1329–1333.
19. Shieh DB, Chou WP, Wei YH, Wong TY, Jin YT. Mitochondrial DNA
4,977-bp deletion in paired oral cancer and precancerous lesions
revealed by laser microdissection and real-time quantitative PCR.
Ann NY Acad Sci 2004;1011:154.
20. Holland PM, Abramson RD, Watson R, Gelfand DH. Detection of
specific polymerase chain reaction product by utilizing the 5′–3′ exo-
nuclease activity of Thermus aquaticus DNA polymerase. Proc Natl
Acad Sci USA 1991;88:7276–7280.
21. Lee LG, Connell CR, Bloch W. Allelic discrimination by nick-translation
PCR with fluorogenic probes. Nucleic Acids Res 1993;21:3761–3766.
22. Cardullo RA, Agrawal S, Flores C, Zamecnick PC, Wolf DE. Detection
of nucleic acid hybridization by non-radiative fluorescence resonance
energy transfer. Proc Natl Acad Sci USA 1988;85:8790.
23. Heid CA, Stevens J, Livak KJ, Williams PM. Real time quantitative
PCR. Genome Res 1996;6:986–994.
24. Gibson UE, Heid CA, Williams PM. A novel method for real time
quantitative RT-PCR. Genome Res 1996;6:995–1001.
25. Dumur CI, Dechsukhum C, Wilkinson DS, Garrett CT, Ware JL,
Ferreira-Gonzalez A. Analytical validation of a real-time reverse
transcriptionpolymerase chain reaction quantitation of different
transcripts of the Wilms’ tumor suppressor gene (WT1). Anal Biochem
2002;309:127–136.
26. Jurado J, Prieto-Alamo MJ, Madrid-Risquez J, Pueyo C. Absolute
gene expression patterns of thioredoxin and glutaredoxin redox
systems in mouse. J Biol Chem 2003;278:45546.
27. Borg I, Rohde G, Loseke S, et al. Evaluation of a quantitative real-
time PCR for the detection of respiratory syncytial virus in pulmo-
nary diseases. Eur Respir J 2003;21:944–951.
28. Lin JC, Wang WY, Chen KY, et al. Quantification of plasma Epstein-
Barr virus DNA in patients with advanced nasopharyngeal carci-
noma. N Engl J Med 2004;350:2461–2470.
29. Castelain S, Descamps V, Thibault V, et al. TaqMan amplification
system with an internal positive control for HCV RNA quantitation.
J Clin Virol 2004;31:227–234.
30. Gilliland G, Perrin S, Bunn HF. Competitive PCR for quantitation
of mRNA. In: PCR Protocols: A Guide to Methods and Applications.
Innis, MA, ed. CA, USA: Academic Press, 1990, 60–69.
31. Suzuki T, Higgins PJ, Crawford DR. Control selection for RNA quan-
titation. BioTechniques 2000;29:332–337.
32. Bustin SA. Absolute quantification of mRNA using real-time reverse
transcription polymerase chain reaction assays. J Mol Endocrinol
2000;25:169–193.
 8. QUANTITATIVE REVERSE TRANSCRIPTASE PCR 83

33. Rhoads RP, McManaman C, Ingvartsen KL, Boisclair YR. The house-
keeping genes GAPDH and cyclophilin are regulated by metabolic
state in the liver of dairy cows. J Dairy Sci 2004;87:248.
34. Steele BK, Meyers C, Ozbun MA. Variable expression of some “house-
keeping” genes during human keratinocyte differentiation. Anal
Biochem 2002;307:341–347.
35. Yperman J, De Visscher G, Holvoet P, Flameng W. β-actin cannot be
used as a control for gene expression in ovine interstitial cells derived
from heart valves. J Heart Valve Dis 2004;13:848.
36. Dheda K, Huggett JF, Bustin SA, Johnson MA, Rook G, Zumla A.
Validation of housekeeping genes for normalizing RNA expression
in real-time PCR. BioTechniques 2004;37:112, 116, 118.
37. BasA, Forsberg G, Hammarstrom S, Hammarstrom ML. Utility of
the housekeeping genes 18S rRNA, β-actin and glyceraldehyde-3-
phosphate-dehydrogenase for normalization in real-time quantita-
tive reverse transcriptase-polymerase chain reaction analysis of
gene expression in human T lymphocytes. Scand J Immunol 2004;59:
566–573.
38. Vandesompele J, De Preter K, Pattyn F, et al. Accurate normalization
of real-time quantitative RT-PCR data by geometric averaging of
multiple internal control genes. Genome Biol 2002;3:0034.I.
39. Morrison TB, Weis JJ, Wittwer CT. Quantification of low-copy tran-
scripts by continuous SYBR Green I monitoring during amplifica-
tion. BioTechniques 1998;24:954–958, 960, 962.
40. Ririe KM, Rasmussen RP, Wittwer CT. Product differentiation by
analysis of DNA melting curves during the polymerase chain reac-
tion. Anal Biochem 1997;245:154–160.
41. Gibellini D, Vitone F, Schiavone P, Ponti C, La Placa M, Re MC.
Quantitative detection of human immunodeficiency virus type 1
(HIV-1) proviral DNA in peripheral blood mononuclear cells
by SYBR green real-time PCR technique. J Clin Virol 2004;29:
282–289.
42. Blaschke V, Reich K, Blaschke S, Zipprich S, Neumann CJ. Rapid
quantitation of proinflammatory and chemoattractant cytokine
expression in small tissue samples and monocyte-derived dendritic
cells: validation of a new real-time RT-PCR technology. Immunol
Methods 2000;246:79–90.
43. Ramos-Payan R, Aguilar-Medina M, Estrada-Parra S, et al. Quantifi-
cation of cytokine gene expression using an economical real-time
polymerase chain reaction method based on SYBR Green I. Scand J
Immunol 2003;57:439–445.
44. Nakamura T, Scorilas A, Stephan C, et al. Quantitative analysis of
macrophage inhibitory cytokine-1 (MIC-1) gene expression in human
prostatic tissues. Br J Cancer 2003;88:1101–1104.
45. Gut M, Leutenegger CM, Huder JB, Pedersen NC, Lutz H. One-tube
fluorogenic reverse transcriptionpolymerase chain reaction for
the quantitation of feline coronaviruses. J Virol Methods 1999;77:
37–46.
84 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

46. Giulietti A, Overbergh L, Valckx D, Decallonne B, Bouillon R, Mathieu

C. An overview of real-time quantitative PCR: applications to quan-
tify cytokine gene expression. Methods 2001;25:386–401.
47. Ginzinger DG. Gene quantification using real-time quantitative
PCR: an emerging technology hits the mainstream. Exp Hematol
2002;30:503–512.
48. van Hoeyveld E, Houtmeyers F, Massonet C, et al. Detection of single
nucleotide polymorphisms in the mannosebinding lectin gene using
minor groove binder-DNA probes. J Immunol Methods 2004;287:
227–230.
49. de Kok JB, Wiegerinck ET, Giesendorf BA, Swinkels DW. Rapid
genotyping of single nucleotide polymorphisms using novel minor
groove binding DNA oligonucleotides (MGB probes). Hum Mutat
2002;19:554–559.
50. Zeschnigk M, Bohringer S, Price EA, Onadim Z, Masshofer L,
Lohmann DR. A novel real-time PCR assay for quantitative analysis
of methylated alleles (QAMA): analysis of the retinoblastoma locus.
Nucleic Acids Res 2004;32:E125.
51. Emig M, Saussele S, Wittor H, et al. Accurate and rapid analysis of
residual disease in patients with CML using specific fluorescent
hybridization probes for real time quantitative RT-PCR. Leukemia
1999;13:1825–1832.
52. van der Velden VH, Hochhaus A, Cazzaniga G, Szczepanski T,
Gabert J, van Dongen JJ. Detection of minimal residual disease
in hematologic malignancies by real-time quantitative PCR: prin-
ciples, approaches, and laboratory aspects. Leukemia 2003;17:
1013–1034.
53. Schalasta G, Eggers M, Schmid M, Enders G. Analysis of human
cytomegalovirus DNA in urines of newborns and infants by means
of a new ultrarapid real-time PCR-system. J Clin Virol 2000;19:
175–185.
54. Aliyu SH, Aliyu MH, Salihu HM, Parmar S, Jalal H, Curran MD.
Rapid detection and quantitation of hepatitis B virus DNA by real-
time PCR using a new fluorescent (FRET) detection system. J Clin
Virol 2004;30:191–194.
55. Tyagi S, Kramer FR. Molecular beacons: probes that fluoresce upon
hybridization. Nature Biotechnol 1996;14:303–308.
56. Smit ML, Giesendorf BA, Vet JA, Trijbels FJ, Blom HJ. Semiauto-
mated DNA mutation analysis using a robotic workstation and
molecular beacons. Clin Chem 2001;47;739–744.
57. Abravaya K, Huff J, Marshall R, et al. Molecular beacons as diagnos-
tic tools: technology and applications. Clin Chem Lab Med 2003;41:
468–474.
58. Wabuyele MB, Farquar H, Stryjewski W, et al. Approaching real-time
molecular diagnostics: single-pair fluorescence resonance energy
transfer (spFRET) detection for the analysis of low abundant point
mutations in K-ras oncogenes. J Am Chem Soc 2003;125:6937–
6945.
 8. QUANTITATIVE REVERSE TRANSCRIPTASE PCR 85

59. Whitcombe D, Theaker J, Guy SP, Brown T, Little S. Detection of

PCR products using self-probing amplicons and flourescence. Nature
1999;17:804.
60. Hart KW, Williams OM, Thelwell N, et al. Novel method for detec-
tion, typing, and quantification of human papillomaviruses in clinical
samples. J Clin Microbiol 2001;39:3204–3212.
61. Thelwell N, Millington S, Solinas A, Booth J, Brown T. Mode of
action and application of Scorpion primers to mutation detection.
Nucleic Acids Res 2000;28:3752–3761.
62. Solinas A, Brown LJ, McKeen C, et al. Duplex Scorpion primers in
SNP analysis and FRET applications. Nucleic Acids Res 2001;29:
E96.
63. Ugozzoli LA, Hamby K. Four-color multiplex 5′ nuclease assay for
the simultaneous detection of the factor V Leiden and the prothrom-
bin G20210A mutations. Mol Cell Probes 2004;18:161–166.
64. Vet JA, Majithia AR, Marras SA, et al. Multiplex detection of four
pathogenic retroviruses using molecular beacons. Proc Natl Acad Sci
USA 1999;96:6394–6399.
65. Tong AK, Li Z, Jones GS, Russo JJ, Ju J. Combinatorial fluorescence
energy transfer tags for multiplex biological assays. Nature Biotech-
nol 2001;19:756–759.
66. Tong AK, Ju J. Single nucleotide polymorphism detection by combi-
natorial fluorescence energy transfer tags and biotinylated dideoxy-
nucleotides. Nucleic Acids Res 2002;30:E19.
67. Rickman D, Bobek MP, Misek DE, et al. Distinctive molecular pro-
files of high-grade and low-grade gliomas based on oligonucleotide
microarray analysis. Cancer Res 2001;65:6885–6891.
68. Miyazato A, Ueno S, Ohmine K, et al. Identification of myelodysplas-
tic syndrome-specific genes by DNA microarray analysis with puri-
fied hematopoietic stem cell fraction. Blood 2001;98:422–427.
69. Dolken G. Detection of minimal residual disease. Adv Cancer Res
2001;82:133.
70. Lo YM, Tein MS, Lau TK, et al. Quantitative analysis of fetal DNA
in maternal plasma and serum: implications for noninvasive prenatal
diagnosis. Am J Hum Genet 1998:62:768–775.
71. Hu Y, Zheng M, Xu Z, Wang X, Cui H. Quantitative real-time PCR
technique for rapid prenatal diagnosis of Down syndrome. Prenat
Diagn 2004;24:704–707.
72. Costa C, Pissard S, Girodon E, Huot D, Goossens M. A one-step real-
time PCR assay for rapid prenatal diagnosis of sickle cell disease and
detection of maternal contamination. Mol Diagn 2003;7:45–48.
73. Arya M, Shergill IS, Williamson M, Gommersall L, Arya N, Patel HR.
Basic principles of real-time quantitative PCR. Expert Rev Mol Diagn
2005;5:209–219.
Chapter 9
Proteonomics: High-Throughput
Structural Biology—Methods for
Cloning, Protein Expression,
and Purification
William K. Gillette and James L. Hartley

INTRODUCTION
The problems associated with expressing and purifying human
proteins, especially in Escherichia coli, the primary host or-
ganism for high-throughput (HTP) applications, are well-
documented and have plagued researchers for decades. Low
yields due to toxicity, recombinant protein insolubility, and poor
purification are just some of the problems that result1 in typical
success rates from 2%–20% when expressing eukaryotic proteins
in E. coli (Service). HTP structural genomic (SG) projects, such
as NIH’s protein structure initiative (PSI) begun in 2000. Ini-
tially, the PSI focused on technology development to provide
highly automated procedures for cloning, expression testing,
protein purification, and protein crystallization, thus addressing
production problems by increasing throughput. The develop-
ment of these techniques has allowed PSI centers and other
similar initiatives around the world to deposit over 2,000 novel
protein structures in the Protein Dada Bank (PBD) as of January
2006 (PSI, http://www.nigms.nih.gov/Initiatives/PSI/). Neverthe-
less, despite the expenditure of significant resources,2 the rate of
discovery is much less than hoped for at the beginning of the
initiative due to bottlenecks at every stage of the pipeline1 as the
problems mentioned above persist. Also, the citation rate of
structures from SG centers is significantly lower than that for the
top structural biology laboratories,3 suggesting that the current
HTP approaches are not as successful in determining the struc-
tures of more difficult, and perhaps, more significant proteins.
The phrase “picking the low hanging fruit” is often used as an
analogy to describe this situation. Accordingly, in the second
 9. PROTEONOMICS 87

phase of the PSI that began in mid-2005, four PSI centers are
focusing on high throughput production, while the remaining six
centers are specializing in specific areas, such as higher eukary-
otic proteins (especially human), membrane proteins, and pro-
teins relevant to disease.

CURRENT APPROACHES

HTP Cloning
The combination of open reading frame (ORF) identification by
genome sequencing projects and the availability of HTP cloning
methods such as the recombination-based Gateway cloning
system4 of Invitrogen and ligation-independent cloning or LIC,5
has enabled the creation of large numbers of “ORF clones,”
bypassing technical problems inherent in using pooled cDNA
libraries.6 A schematic of the HTP techniques used by SG centers
to proceed from cloning to protein purification is pictured in
Figure 9.1.
Once sequence-verified, ORF clones can be used to generate
a wide variety of expression clones by simple in vitro recombina-
tion techniques without the need for subsequent sequence valida-
tion. A variety of recombinational cloning systems (reviewed in
Marsischky and LaBaer)6 are used by HTP production facilities
as well as ligase-independent cloning (LIC) and the standard
cloning of PCR fragments by endonuclease/ligase cloning.6 A
major liability of the latter approach is the limitation on the
number of expression constructs that can be reasonably created.7
This is due to the effort required for their construction and the
introduction of sequence errors from faulty primers and the PCR
used in the creation of each clone. A downside to recombination
cloning is the addition of more non-native amino acids from the
translated recombination sites to the desired protein molecule.
Since more than 1,400 papers using Gateway cloning have been
published, any detrimental effects of these amino acids are prob-
ably minor.
Expression clones can be constructed to test many variables
that affect protein expression and purification. Chief among
these are the use of affinity tags, solubility tags, and vector
sequences required for the expression in a given expression
system (bacteria, mammalian, yeast, etc.). Although there are
trends,8–10 there is no way to predict a priori the best expression
construct and system for a given protein. The possible combina-
tions (e.g., a limited set might include 3 affinity tags, 3 solubility
fusions, N- and C-terminal locations, 3 expression hosts) to test
88 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

Target identification

Cloning into multiple vectors

Plasmid isolation/verification

Transformation/transfection/infection
of expression host

Growth/Temperature shift/Induction

Reclone to screen:
solubility tag
No tag position
Soluble? expression strain
expression conditions
expression system
Yes truncations

Small-scale purification

IMAC

IEX

SEC

FIGURE 9.1. Schematic of the typical workflow for HTP protein produc-
tion. IMAC (immobilized metal-ion affinity chromatography), IEX (ion
exchange chromatography), SEC (size exclusion chromatography).
 9. PROTEONOMICS 89

can become quite numerous and hence the need for automation-
friendly methods at all steps of HTP SG.
Hexa-histidine tags are by far the most commonly used
affinity tag in HTP applications due to the low cost, ease of
subsequent downstream purification, and the relatively modest
addition of non-native amino acids. However, this tag does not
enhance the solubility of the fused target protein, thus solubility
enhancing proteins are often cloned in-frame with the target
protein to improve solubility.11 Commonly used solubility partner
proteins are maltose binding protein (MBP), NusA, and thiore-
doxin.8 Reports of differential success with these and other fusion
proteins are widespread in the literature, however, MBP, NusA,
and thioredoxin are widely reported as the most useful.8,10,11

Expression Systems
Several expressions systems have been adopted by the majority
of HTP SG centers: E. coli, baculovirus, and mammalian cell
culture (HEK293 cells). Additional approaches are and will be
coming available to address particular problems of recombinant
protein expression.12 For example, a Rhodobacter expression
system shows promise for the expression of membrane proteins,13
and both Saccharomyces cerevisiae and Pichia pastoris have been
reported as amenable to HTP production of human proteins for
structural genomics.14
Of the available expression systems, E. coli is preferred for
HTP production, despite the limitations mentioned above, for its
ease of use and reduced cost. Advantages include a robust growth
rate, inexpensive media, a well-defined genetic system, a variety
of induction systems, and the availability of strains and tech-
niques for different applications. For example, isotopically
labeled amino acids important for NMR and x-ray crystallogra-
phy can be incorporated into the protein with the use of labeled
precursors in the media. Although E. coli has limited capability
for post-translational modification of proteins that can be impor-
tant for folding and activity,15 this can be beneficial for structural
studies due to the reduction in sample heterogeneity. Strains
with low protease levels (e.g., BL21) and modifications to express
tRNAs present at higher levels in eukaryotes than bacteria can
be the difference between no expression and high levels of soluble
protein.16 The major drawback to recombinant expression in E.
coli is the lack of the proper protein folding machinery, which
often leads to mis-folding and/or aggregation. However, new
solubility and affinity tags are frequently generated for protein
expression in E. coli,17–20 and the flexibility of the system allows
90 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

the co-expression of chaperones and potential binding partners

that might enhance proper folding and protein stability. The
robustness and flexibility of the system will likely prove impor-
tant when screening difficult proteins. The use of E. coli extracts
is also becoming a viable option and is particularly attractive for
the expression of toxic proteins and in the screening of additives
and co-factors to enhance solubility.21
The baculovirus/insect cell expression system is a robust
expression system capable of producing levels of protein similar
to that from E. coli, with the added benefit that the proteins are
more likely to be soluble.21,22 Additionally, insect cells are capable
of post-translational modifications that are sometime necessary
for activity.15 Although the pipeline for production is significantly
longer for the baculovirus system compared to E. coli, the tech-
niques needed for cloning, viral production, insect cell culturing,
infection, and harvest have all been readily adapted to HTP
methods and can lead to protein structure determination.21
Expression in mammalian systems is attractive due to the
possibility of obtaining correctly folded proteins. However, mam-
malian cell culture is costly; yields are considerably lower, and
generally not amenable to HTP techniques. These factors have
limited the use of mammalian expression system in HTP applica-
tions to the HEK293 cell line.23 As the difficult proteins are pro-
cessed through the SG pipeline, this system and others as yet to
be developed will likely play an important role.

Expression Conditions
The conditions under which a protein is expressed have some of
the most dramatic effects on the success of expression. Variables
included temperature, levels of inducer, media formulations,
aeration, time of induction, and time of harvest. The most dra-
matic of these variables is the temperature during induction. It
has long been known that lowering the temperature at the point
of induction can realize dramatic improvements in protein solu-
bility in E. coli (Schein and Noteborn).24 For this reason, HTP
protein expression in E. coli is routinely carried out at tempera-
tures ranging from 4°C–30°C.7,8,25 Temperature also can affect
solubility in the baculovirus/insect cell expression system (manu-
script in preparation).
The auto-induction system for E. coli recently described by
Studier26 obviates the need for monitoring culture ODs and the
addition of the expensive inducer IPTG when using the tradi-
tional T7 RNA polymerase transcription system widely favored
 9. PROTEONOMICS 91

in HTP applications. With this procedure, batch cultures of high

ODs, normally only obtained in controlled fermentation experi-
ments, can be achieved. The use of this system is being adopted
by several HTP applications (Protein Expression Meeting, CHI
PepTalk, San Diego, 2006). A limitation on autoinduction is the
difficulty in timing reduction in culture temperature to maximize
both yield and solubility (Protein Expression Meeting, CHI
PepTalk, San Diego, 2006).

Expression and Solubility Screening

Standard methods of detergent-based cell-lysis or sonication are
easily amenable to HTP techniques to produce lysates from the
harvested expression material. After the insoluble material is
removed, either by filtration or centrifugation, a variety of screen-
ing methods are used to determine the amount of soluble protein
in the lysate. One approach is to assay for the His-tagged protein,
using an antibody specific for the His tag. Antibodies against
many of the other commonly used affinity and solubility tags are
also available. By comparing the signal from whole cell lysates
to signal from clarified lysates, the amount of expressed and
soluble material can be determined. If the tag assayed is N-
terminally located, this approach detects all tagged species with
no differentiation between full-length proteins and truncated
species. Another approach is to assay the samples directly by
SDS-PAGE. However, this is a time-consuming technique, and
recently HTP separation of proteins by size using microfluidic
technology has been achieved.27,28 Alternatively, some groups
avoid the analysis of soluble protein altogether and proceed
directly to the first step in HTP purification.7

Purification
Immobilized metal-ion affinity chromatography (IMAC) is most
often the first step in the purification of His-tagged target pro-
teins. If expression optimization has been performed and the
target is expressed as a fusion to a solubility tag such as MBP, a
His tag is included in the construct for purification. HTP formats
for IMAC enable purification without the need for expensive
chromatography workstations in this first stage of screening.
Indeed, the entire protein production process from gene cloning
through expression and to purification can be performed in 96-
well format using IMAC.
Well-expressed, soluble proteins are often >90% pure after a
single IMAC step. For many uses, proteins expressed as fusions
92 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

must be cleaved with an appropriate protease to separate the

target from the solubility protein. An example of a commonly
used approach is to express the protein in the format, His6-MBP-
tev-POI, where an N-terminal His tag is fused to the solubility
protein MBP upstream of a TEV protease recognition site,
TEV,29,30 and POI is the protein of interest. After cleavage with
TEV (in this example the TEV protease is His-tagged to enable
removal after cleavage), the POI is released. TEV protease cleaves
the site ENLYFQ/X, where X can be any amino acid except
proline.31 Thus, in theory, a protein can be cloned in such a way
as to produce the wild type sequence after TEV cleavage. In
practice, this procedure can fail at several points in a protein
dependent, and as yet, unpredictable manner. Some proteins are
resistant to TEV protease cleavage, while others precipitate after
being released from the solubility partner and some proteins do
not separate well from the components of the TEV protease
digestion in the second IMAC step designed to remove all His-
tagged proteins from the sample. Still, the benefits afforded by
solubility tags (allowing the purification of intractable proteins)
warrants their continued use in HTP SG.29
At this point, some proteins are sufficiently pure, while other
proteins may require an additional chromatographic step to
remove contaminants (typically ion-exchange chromatography).
A final gel filtration step (size exclusion chromatography) to
remove soluble aggregates is also often employed. Successfully
purified proteins can be used for buffer optimization studies to
refine the chromatographic conditions and are scaled up to
provide the quantities necessary for structural studies, usually
10–100 mg.

ADDITIONAL APPROACHES
While HTP approaches have greatly improved the ability to
screen large numbers of samples and advances in cloning and
the introduction of new affinity and solubility tags have led to
the successful purification of proteins that were previously intrac-
table, the inherent problems of protein production in heterolo-
gous expression systems still remain. It is expected that as the
structure of the relatively tractable proteins are determined, what
remains are especially difficult proteins. Notable in this group
are membrane proteins, which are the target of 60%–70% of
pharmaceutical drugs.32
One lesson that has been learned repeatedly in the study of
proteins is that we do not know the rules that govern protein
 9. PROTEONOMICS 93

behavior. Thus, additional techniques and approaches will

be needed to address the problems associated with protein
production. We have developed such an approach named POET,
for Pooled ORF Expression Technology,33 that can improve
the efficiency of a majority of the processes discussed here
(Figure 9.2). The basic approach of POET is that working with
pools of known ORFs, instead of the individual clones, finesses
the complexities of protein expression by improving the effi-
ciency of any step in the protein production pipeline by a factor
of n, where n is the number of ORFs in a given pool. We have
applied POET to a pool of 688 C. elegans ORFs. A high percentage
of ORFs identified in this experiment yielded expressed, soluble,
purified proteins in agreement with POET predictions. Standard
techniques of recombinational Gateway cloning, protein expres-
sion, and purification are applied to the pool. The resulting puri-
fied proteins are analyzed by standard proteomic techniques
(two-dimensional gel electrophoresis (2DGE) and mass spec-
trometry) to identify and quantify the successfully purified pro-
teins. Recent results in our lab suggest that application of a new
mass spectrometry approach34 to a purified POET protein pool
can produce quantification without the need for 2DGE. If vali-
dated, this would allow the automation of all steps in the POET
process, further increasing the efficiency of the approach. The
ability to quantify protein abundance by mass spectrometry
would be an exceptional improvement in the many proteomic
fields.
The POET method is useful in identifying proteins that are
likely to be successfully expressed and purified in isolation under
a given set of experimental conditions. As proteins that are
relatively easily expressed and purified are characterized and the
remaining uncharacterized proteins are tested under more
conditions, the success rate will decrease, and techniques that
improve the efficiency of the process, such as POET, will be
needed.

Acknowledgment
This project has been funded in whole or in part with federal
funds from the National Cancer Institute, National Institutes of
Health, under contract N01-CO-12400. The content of this pub-
lication does not necessarily reflect the views or policies of the
Department of Health and Human Services, nor does mention
of trade names, commercial products, or organizations imply
endorsement by the U.S. government.
 94

Gateway-compatible Pool 250-

Subclone pool Express pool
ORFeome collection 500 ORFs

I
M
A
C
BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

Retreive positive clones

and scale up Identify proteins Separate purified proteins
Purify pool
expression/purification by mass spectrometry by 2DGE
FIGURE 9.2. Summary of POET method for screening an ORF clone collection for expressed, soluble proteins that can
be purified. IMAC (immobilized metal-ion affinity chromatography).
 9. PROTEONOMICS 95

References
1. Service RF. Structural genomics. Tapping DNA for structures pro-
duces a trickle. Science 2002;298:948–950.
2. Lattman E. The state of the Protein Structure Initiative. Proteins
2004;54:611–615.
3. Chandonia JM, Brenner SE. The impact of structural genomics:
expectations and outcomes. Science 2006;311:347–351.
4. Hartley JL, Temple GF, Brasch MA. DNA cloning using in vitro site-
specific recombination. Genome Res 2000;10:1788–1795.
5. Doyle SA. High-throughput cloning for proteomics research. Methods
Mol Biol 2005;310:107–113.
6. Marsischky G, LaBaer J. Many paths to many clones: a comparative
look at high-throughput cloning methods. Genome Res 2004;14:
2020–2028.
7. Acton TB, Gunsalus KC, Xiao R, et al. Robotic cloning and protein
production platform of the Northeast Structural Genomics Consor-
tium. Methods Enzymol 2005;394:210–243.
8. Dyson MR, Shadbolt SP, Vincent KJ, Perera RL, McCafferty J. Pro-
duction of soluble mammalian proteins in Escherichia coli: identifica-
tion of protein features that correlate with successful expression.
BMC Biotechnol 2004;4:32.
9. Holz C, Prinz B, Bolotina N, et al. Establishing the yeast Saccharo-
myces cerevisiae as a system for expression of human proteins on a
proteome-scale. J Struct Funct Genomics 2003;4:97–108.
10. Lichty JJ, Malecki JL, Agnew HD, Michelson-Horowitz DJ, Tan S.
Comparison of affinity tags for protein purification. Protein Expr
Purif 2005;41:98–105.
11. Waugh DS. Making the most of affinity tags. Trends Biotechnol 2005;
23:316–320.
12. Giomarelli B, Schumacher KM, Taylor TE, et al. Recombinant pro-
duction of anti-HIV protein, griffithsin, by auto-induction in a fer-
mentor culture. Protein Expr Purif. 2006;47(1):194–202.
13. Laible PD, Scott HN, Henry L, Hanson DK. Towards higher-
throughput membrane protein production for structural genomics
initiatives. J Struct Funct Genomics 2004;5:167–172.
14. Prinz B, Schultchen J, Rydzewski R, Holz C, Boettner M, Stahl U,
Lang C. Establishing a versatile fermentation and purification pro-
cedure for human proteins expressed in the yeasts Saccharomyces
cerevisiae and Pichia pastoris for structural genomics. J Struct Funct
Genomics 2004;5:29–44.
15. Jono S, Peinado C, Giachelli CM. Phosphorylation of osteopontin is
required for inhibition of vascular smooth muscle cell calcification.
J Biol Chem 2000;275:20197–20203.
16. Kane JF. Effects of rare codon clusters on high-level expression of
heterologous proteins in Escherichia coli. Curr Opin Biotechnol
1995;6:494–500.
17. Banki MR, Feng L, Wood DW. Simple bioseparations using self-
cleaving elastin-like polypeptide tags. Nat Methods 2005;2:659–661.
96 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

18. Chatterjee DK, Esposito D. Enhanced soluble protein expression

using two new fusion tags. Protein Expr Purif 2006;46(1):122–
129.
19. Lamla T, Erdmann VA. The Nano-tag, a streptavidin-binding peptide
for the purification and detection of recombinant proteins. Protein
Expr Purif 2004;33:39–47.
20. Zuo X, Li S, Hall J, et al. Enhanced expression and purification of
membrane proteins by SUMO fusion in Escherichia coli. J Struct
Funct Genomics 2005;6:103–111.
21. Albeck S, Burstein Y, Dym O, et al. Three-dimensional structure
determination of proteins related to human health in their functional
context at The Israel Structural Proteomics Center (ISPC). Acta Crys-
tallogr D Biol Crystallogr 2005;61:1364–1372.
22. Gao M, Brufatto N, Chen T, et al. Expression profiling of herpes virus
and vaccinia virus proteins using a high-throughput baculovirus
screening system. J Proteome Res 2005;4:2225–2235.
23. Thomas P, Smart TG. HEK293 cell line: a vehicle for the expression
of recombinant proteins. J Pharmacol Toxicol Methods 2005;51:
187–200.
24. Schein CH, Noteborn MHM. Formation of soluble recombinant pro-
teins in Escherichia coli is favored by lower growth temperatures.
Biotechnology (NY). 1988;6:291–294.
25. Goulding CW, Perry LJ. Protein production in Escherichia coli for
structural studies by X-ray crystallography. J Struct Biol 2003;142:
133–143.
26. Studier FW. Protein production by auto-induction in high density
shaking cultures. Protein Expr Purif 2005;41:207–234.
27. Goetz H, Kuschel M, Wulff T, et al. Comparison of selected analytical
techniques for protein sizing, quantitation and molecular weight
determination. J Biochem Biophys Meth 2004;60:281–293.
28. Mouradian S. Lab-on-a-chip: applications in proteomics. Curr Opin
Chem Biol 2002;6:51–56.
29. Jeon WB, Aceti DJ, Bingman CA, et al. High-throughput purifica-
tion and quality assurance of Arabidopsis thaliana proteins for
eukaryotic structural genomics. J Struct Funct Genomics 2005;6:
143–147.
30. Nallamsetty S, Austin BP, Penrose KJ, Waugh DS. Gateway vectors
for the production of combinatorially-tagged His6-MBP fusion pro-
teins in the cytoplasm and periplasm of Escherichia coli. Protein Sci
2005;14:2964–2971.
31. Kapust RB, Tozser J, Copeland TD, Waugh DS. The P1′ specificity of
tobacco etch virus protease. Biochem Biophys Res Commun 2002;294:
949–955.
32. Lundstrom K. Structural genomics on membrane proteins: mini
review. Comb Chem High Throughput Screen 2004;7:431–439.
33. Gillette WK, Esposito D, Frank PH, et al. Pooled ORF expression
technology (POET): using proteomics to screen pools of open reading
 9. PROTEONOMICS 97

frames for protein expression. Mol Cell Proteomics 2005;4:1647–

1652.
34. Ishihama Y, Oda Y, Tabata T, et al. Exponentially modified protein
abundance index (emPAI) for estimation of absolute protein amount
in proteomics by the number of sequenced peptides per protein.
Mol Cell Proteomics 2005;4:1265–1272.
Chapter 10
DNA and Tissue Microarrays
Maysa M. Abu-Khalaf, Lyndsay N. Harris, and Gina G. Chung

DNA MICROARRAY

Methodology
Although techniques such as RT-PCR and in situ hybridization
(ISH) can give information about gene expression, they are
limited in scope as typically one gene product is evaluated with
each assay. The advent of transcriptional profiling using DNA
microarray has revolutionized the field of molecular medicine as
measurement of thousands of genes simultaneously in a given
sample provide a vast amount of data for new disease classifica-
tions and biomarker discoveries. DNA microarray-based gene
expression profiling relies on nucleic acid polymers, immobilized
on a solid surface, which act as probes for complementary gene
sequences.1 Microarrays typically contain several thousand
single-stranded DNA sequences, which are “arrayed” at specific
locations on a synthetic “chip” through covalent linkage. These
DNA fragments provide a matrix of probes for fluorescently
labeled complementary RNA (cRNA) derived from the sample of
interest. The expression of each gene in the sample is quantified
by the intensity of fluorescence emitted from a specific location
on the array matrix which is proportional to the amount of that
gene product (Figure 10.1).2
Several different microarray systems have been developed,
using either 25-mer or 60-mer oligonucleotides or cDNA as
probes. It is important to note that technical differences between
various types of arrays can influence the subsets of genes de-
tected, especially when analyzing 12,500+ transcripts per slide.
Two main types of microarrays are used for evaluation of
clinical samples: cDNA or spotted arrays, and oligonucleotide
microarrays3

1. cDNA array: Spotted arrays are constructed using annotated

cDNA probes from commercial vendors, or by reverse transcription
 10. DNA AND TISSUE MICROARRAYS 99

Unfixed sample
of tumor tissue

Tumor RNA

Surgical removal
of tumor tissue

Labeled tumor
Labeled control cDNA or cRNA
cDNA or cRNA

Poor
prognosis

Comparative Molecular
analysis of gene signature
expression
Good
prognosis
DNA microarray

FIGURE 10.1. Reprinted from Sauter G, Simon R. Predictive molecular

pathology. N Engl J Med 2002;347:1995–1996. (Copyright 2002 Massa-
chusetts Medical Society)

of mRNA from a known source used to generate cDNA probes for

specific gene products which are then systematically spotted on
glass slides.3 For sample testing, human Universal RNA standards
(serial 10-fold dilutions) are extracted and reverse transcribed in
parallel along with test samples and labeled with a fluorescent
dye (e.g. Cy3). cRNA from test samples is also labeled with a
fluorescent dye (e.g., Cy5). Each Cy5-labeled test sample cRNA
together with the Cy3-labeled reference probe is hybridized to the
spotted microarray simultaneously. Fluorescence intensities of
the scanned images from the hybridizations are quantified,
normalized, and corrected to yield the transcript abundance of a
gene as an intensity ratio with respect to that of the reference
sample.4 Arrays are generally analyzed with a confocal laser
scanner, which allows quantification of gene expression as a
relative value at each coordinate on the slide.
2. Oligonucleotide microarrays: Oligonucleotide arrays use
25 base pair DNA oligomers to be printed onto the array matrix.5
100 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

These arrays, designed and patented by Affymetrix, use a

combination of photolithography and combinatorial chemistry,
which allows the simultaneous generation of thousands of probes.
Oligonucleotide probes for different genes can be deposited or
synthesized directly on the surface of a silicon wafer in a
predetermined order. Sequences are designed to minimize cross-
reactivity, however, some nonspecific hybridization will usually
occur. Therefore, the Perfect Match/Mismatch (PM/MM) probe
strategy is utilized, in which a second probe that is identical to
the first except for a mismatched base (the MM probe) is laid
adjacent to the PM probe. After sample hybridization, signal
from the MM probe is subtracted from the PM probe signal to
account for the nonspecific binding.

Data Analysis
Relative levels of expression on the microarrays are analyzed
using sophisticated statistical techniques. There are two main
types of multisample analyses: class discovery (creating new
classes based on differences in expression among samples) and
class prediction (using samples from known biologic classes to
identify a list of genes whose expression pattern can be used to
predict the class of a new sample).6
The first step of analysis is normalization of the raw data
which maximizes the likelihood that the measurements of dif-
ferential expression are not artifacts of the hybridization reac-
tion.6 Data are then filtered to select those genes with the largest
magnitude of differences in expression for further analysis.
To discover new subgroups based on the gene expression
patterns of biologically similar samples (class discovery), unsu-
pervised analysis is used. This technique uses clustering algo-
rithms to group specimens according to similarities in their
transcriptional profile. For class prediction, supervised analysis
is used, whereby the gene expression profile of one defined group
is compared to another and a list of genes is generated which
distinguishes the two groups. Once a set of genes is identified in
the “training set,” they are ranked by their power to predict the
group to which each sample belongs by cross-validation. Typi-
cally, one sample at a time is left out, the classifier is trained on
the remaining samples, and the sample is then classified based
on its correlation to a predictor set generated from the remaining
samples. The independent predictive ability of the gene list (“clas-
sifier”) is ideally performed on a separate set of samples. Often
this is achieved by dividing the original set into a training set and
a validation set.
 10. DNA AND TISSUE MICROARRAYS 101

The predictive success with either technique is used to calcu-

late the error rate of the predictive genes. It should be noted that
leave-one-out cross-validation generally overestimates classifica-
tion accuracy, and so ultimately, an independent validation set
is required.6

Sample Preparation Issues

For clinical applications of gene profiling, an important consid-
eration is the isolation of sufficient mRNA from the tumor
samples of interest. Adequate numbers of cells of interest must
be isolated from heterogeneous tissue extracts, and these speci-
mens must be preserved in a way that does not degrade the
quality of the RNA. With respect to preservation, most clinical
samples are preserved using aldehyde-based chemicals such as
formalin, which has been reported to degrade the quality of
RNA.6 Alternatively, flash-freezing samples by immersion in
liquid nitrogen can maintain the quality of mRNA. Samples,
however, must be frozen as soon as possible after excision to
avoid RNA degradation or changes due to ischemia.

Clinical Applications
Molecular phenotyping of human tumors using micorarray tech-
nology has provided insights into tumor subtypes and their atten-
dant clinical outcomes. For example, using microarray technology,
Sorlie et al. demonstrated that invasive breast cancers could be
divided into four main subtypes by unsupervised analysis.7 These
groups were distinguished by differences in their expression pro-
files: a “luminal cell-like” group expressing ER; a “basal cell-like”
group expressing keratins 5 and 17, integrin β4, and laminin, but
lacking ER; an Erb-B2 positive group; and a “normal” epithelial
group, typified by genes characteristic of basal epithelial cells
and adipose cells. A subsequent study by this group showed that
these subtypes have specific clinical outcomes, which are repro-
ducible in other data sets.
Other investigators have used supervised analysis to identify
a signature profile in breast cancer patients at very low risk for
distant relapse.8 This 70-gene prognostic signature was able to
delineate a group of patients with very low risk of distant recur-
rence in an independent dataset which appeared to perform
better than traditional prognostic tumor characteristics as defined
by St. Gallen or NIH criteria.9
While microarray profiling has provided important insights
into the biology of disease, applying this technology to the study
of response to therapy is likely to benefit patients in a more
102 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

immediate way. Recent studies suggest that subsets of genes

identified by microarray profiling can be used to predict response
to a chemotherapy used in breast cancer, the taxanes. Applica-
tion of this method has lead to the identification of tau10 and
apoptosis genes11 as potential markers associated for resistance
to taxane-based therapy. Furthermore, not only was high expres-
sion of tau correlated with a lack of response to paclitaxel in vivo,
down-regulation of tau via small interfering RNA caused an
increase in sensitivity to paclitaxel in vitro.10
In a preoperative trial of trastuzumab and vinorelbine, pre-
treatment tissue core biopsies were used to perform transcrip-
tional profiling using Affymetrix U133+2 Gene Chips (Harris et
al. Clin Can Res in press). Unsupervised analysis for class distinc-
tion revealed three top-level clusters; all tumors which achieved
a pathologic complete response to this combination therapy were
in one cluster whereas resistant tumors (defined as lack of
response or progression within 1 year) fell in another distinct
cluster. Larger T4 tumors were more frequent in the “nonre-
sponse cluster” (p = 0.02), and supervised analysis comparing
differential expression patterns between T4 vs. lower stage
tumors showed that both under and overexpressed genes of the
basal lineage were more frequent in T4 tumors (p < 0.00001).
Supervised analysis of nonresponding tumors showed higher
expression of several growth factors (HGF, IGF-1, PDGF, pleo-
tropin), growth factor receptors (c-met, leptin receptor), and the
PI3Kinase regulatory subunit p85 and MAP2.

TISSUE MICROARRAY

Construction
Tissue microarrays complement the large scale genomic/pro-
teomic discovery approach of DNA arrays by allowing the simul-
taneous analysis of DNA, RNA, or protein in large numbers of
samples per single experiment (as opposed to DNA arrays which
look at large numbers of gene products simultaneously in a test
sample). By linking these data to relevant outcome information,
e.g., survival, these analyses can give insight into the clinical
significance of a given biomarker.
Although the concept of standardizing and streamlining
immunohistochemistry (IHC) techniques have been previously
reported,12 Kononen et al.13 first described a device for the con-
struction of TMAs that could be feasibly accessible to many labs
(Figure 10.1). The bulk of the time spent in TMA construction is
the collection of the appropriate paraffin-embedded “donor”
 10. DNA AND TISSUE MICROARRAYS 103

tissue blocks and identification of the area of tissue of interest

(e.g., invasive tumor). The “recipient” or “master” arrays are then
assembled by taking a core tissue specimen from hundreds of
separate donor blocks (e.g., different patient tumor blocks) and
re-embedding them into the recipient block. Typically, cores are
0.6 mm in diameter spaced at 0.7–0.8 mm, which allows up to
1000 samples to be placed on a recipient block. Larger diameter
cores can also be taken in certain instances (for example, when
tissue heterogeneity is expected to be greater), although this
reduces the number of cores that can be taken from the donor
block and that can be placed into the recipient block. Depending
on the thickness of the samples, 100–200 5-µm sections can be
cut from the recipient block for transfer onto a standard glass
slide using an adhesive tape transfer method. The resultant slide
can then be analyzed for a variety of molecular targets at the
DNA, RNA, or protein level. Redundant arrays can also be con-
structed by obtaining multiple cores from the donor blocks and
placing them at identical coordinates in recipient blocks.
Because both cut slides as well as blocks may be subject to
antigen oxidation and degeneration, some facilities store recipi-
ent blocks in sealed nitrogen chambers and coat cut slides in
paraffin to minimize these effects.14 In addition, because tissue
blocks are three-dimensional structures that can change as more
sections are cut; most facilities employ a quality control monitor-
ing system (e.g., every 10th section stained with H&E to assess
tissue representatively).

Advantages and Criticisms of Tissue Microarrays

There are several advantages to TMAs. First and most significant
is the amplification of tissue resources. A conventional block
would be exhausted by 50–100 cuts, and analysis of 50 antibodies
on 250 specimens would require 12,500 slides. This approach to
tissue analysis is a prohibitive task that also very quickly exhausts
tissue resources. As an example using TMAs, up to 400 master
blocks can be made from a 1-cm tissue section. These can each
be cut as many as 200 times, allowing the evaluation of 80,000
unique reagents. As the TMA technology becomes more com-
monly utilized and incorporated into clinical trials, this may be
of particular importance. First, this allows the efficient organiza-
tion and storage of archived tissue blocks in many pathology
departments. At the Yale TMA Facility, radiofrequency identifica-
tion tags are used in the tissue blocks for this purpose. Second,
tissue collection for biologic correlative studies is now standard
in clinical trials, however, the samples are often small and obtained
104 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

by invasive procedure specifically for research purposes. TMA

technology significantly increases the number of experiments that
can be performed from these precious resources as well as assist
in the distribution of samples to other investigators.
Other advantages of TMAs relate to the high-throughput
format. Large numbers of different types of tissues (benign and
malignant), xenograft tissues, cell lines, or recombinant proteins
can be readily integrated into the arrays to serve as intra- and
interslide references. Only a limited amount of antibody and
other reagents, similar to what is used for a whole section, are
required. In addition, because hundreds of samples can be
studied in one experiment, common variables that can affect
reproducibility, such as antigen retrieval, reagent concentrations,
and washing times, can be standardized.
A common criticism of TMAs relates to the issue of tissue and
tumor heterogeneity—whether a small core is representative of
the entire tumor. Indeed, this argument can be expanded to
whole tissue sections and blocks of tumor, as large surgical resec-
tions are only semirandomly sampled by the pathologist. Never-
theless, many investigators have shown concordance rates of
approximately 95% between 2–4 0.6 mm TMA spots and whole
sections for common biomarkers such as estrogen receptor and
progesterone receptor in breast cancer.15,16 Furthermore, they
were able to reproduce known clinicopathologic correlations
with the TMA-based studies.16 Similar validation studies have
been performed in numerous other tumors, including those felt
to be more inherently heterogenous such as Hodgkin’s lym-
phoma,17 pancreatic carcinoma,18 and soft-tissue sarcomas.19
Although TMAs are best used as epidemiology-based research
tools to examine relative expression of molecular markers in
large cohorts of patients, these studies suggest that diagnostic
application to individual clinical patients may also be appropri-
ate if used judiciously.

Applications
Over 600 studies have been published using TMA technology
since Kononen’s initial description in 1998. Most of these studies
were performed on various malignancies to study different
molecular markers using IHC. When linked to a clinical end-
point, e.g., survival or response to a specific therapy, they have
the ability to assess rapidly the prognostic or predictive value
respectively of the marker of interest. TMAs have also been used
to validate genes discovered by genomic surveys such as DNA
microarrays. A variety of TMAs spanning tumor development
 10. DNA AND TISSUE MICROARRAYS 105

and progression has been described such as normal breast tissue,

atypical hyperplasia, in situ carcinoma, invasive carcinoma,
nodal metastases, and distant metastases. Similar arrays have
been described for prostate and for pancreatic carcinomas. Other
techniques such as fluorescence in situ hybridization (FISH) or
mRNA in situ hybridization (mRNA-ISH) can be readily adapted
to TMAs. Finally, TMAs may prove to be a useful technology
beyond cancer research. TMA spots can be arrayed with parental
and modified cell lines, genetically engineered animal tissues,
noncancerous diseased tissues, and multiorgan/multitumor
tissues. In the last model, a miniature animal system can be
represented for rapid and global assessment of a specific marker,
although optimization of antibody titers across such disparate
tissues can be problematic.

Quantitative Analysis
Whereas analysis or “scoring” of chromagen-linked IHC stains of
TMAs is relatively straightforward with a bright-field microscope,
it is time-consuming and laborious, often making this the rate-
limiting step. More importantly, because judgments of intensity
are subjective and limited, intra- and interobserver reproduc-
ibility is difficult. Many efforts to produce a more automated and
quantitative analysis of IHC stains have been developed, and
more recently, several commercially available programs have
become available.19–21 We have developed an automated quantita-
tive technology that uses modified IHC with immunofluores-
cence-based detection rather than optical density, which allows
increased sensitivity and dynamic range.22 Using molecular tags
to define tumors (i.e., cytokeratin for epithelium) and localize
subcellular compartments (i.e., DAPI for nucleus), protein expres-
sion is assessed on a continuous scale by co-localization algo-
rithms. This technology has been applied to the study of a variety
of biomarkers in numerous different cancers.23,24 In addition,
because a molecular tag is simply defined by a molecule with
specificity for a defined/localized antigen, one can envision study-
ing in situ quantitative co-localization to subcellular compart-
ments such as Golgi or mitochondria and stromal compartments
such as endothelial cells. As the TMA technology becomes
increasingly utilized, the need for more sophisticated biostatisti-
cal strategies to rigorously organize and analyze these data
becomes even critical. Simple spreadsheet and standard statisti-
cal software packages will most likely be inadequate. Clustering
algorithms analogous to those employed for gene expression
arrays may be further adapted and utilized in the future.
106 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

DNA and tissue microarray: key points

• Transcriptional profiling has substantially advanced molecular
medicine by producing new disease classifications and by serving
as a paradigmatic methodology for biomarker discoveries. DNA
arrays allow the simultaneous measurement of thousands of gene
products by hybridizing cRNA from a sample of interest to
immobilized nucleic acid polymers on a solid surface.
• Expression levels are analyzed mainly by class discovery (which
creates new biologic categories based on sample differences of
expression) or class prediction (which identifies new profiles of
expression that can predict known biologic classes).
• Molecular profiles of human tumors using micorarray technology
can provide insights into newly defined tumor subtypes and aid in
prediction of clinical outcomes and response to specific
treatments.
• TMA technology is a powerful tool for the study of biomarker
development and clinicopathologic correlations in large numbers
of tissue samples.
• The TMA platform allows the analysis of DNA, RNA, and protein
levels in morphologically intact tissue.
• The high-density placement of small 0.6-mm cores of tissues
amplifies tissue resources and allows biomarker analysis on large
numbers of samples rapidly and with standardized experimental
conditions.
• Several-fold TMA cores have been shown to be representative of
whole sections in a variety of cancers, but new tumor cohorts
should be carefully assessed for heterogeneity.
• A variety of automated image analysis technologies have been
described to produce readings of TMAs more rapidly and
reproducibly.
• TMAs can readily validate potential biomarkers identified in
DNA microarray experiments, assess prevalence of biomarkers
in commonly prevalent cancers, and can be incorporated into
clinical trials for retrospective investigations into the molecular
mechanisms of therapies.

References
1. Southern E, Mir K, Shchepinov M. Molecular interactions on micro-
arrays. Nat Genet 1999;21:5–9.
2. Sauter G, Simon R. Predictive molecular pathology. N Engl J Med
2002;347:1995–1996.
3. Ramaswamy S, Golub TR. DNA microarrays in clinical oncology.
J Clin Oncol 2002;20:1932–1941.
4. Cheung VG, Morley M, Aguilar F, et al. Making and reading micro-
arrays. Nat Genet 1999;21:15–19.
 10. DNA AND TISSUE MICROARRAYS 107

5. Lockhart DJ, Dong H, Byrne MC, et al. Expression monitoring by

hybridization to high-density oligonucleotide arrays. Nat Biotechnol
1996;14:1675–1680.
6. Goldsmith ZG, Dhanasekaran N. The microrevolution: applications
and impacts of microarray technology on molecular biology and
medicine (review). Int J Mol Med 2004;13:483–495.
7. Sorlie T, Perou CM, Tibshirani R, et al. Gene expression patterns of
breast carcinomas distinguish tumor subclasses with clinical impli-
cations. Proc Natl Acad Sci USA 2001;98:10869–10874.
8. van’t Veer LJ, Dai H, van de Vijver MJ, et al. Gene expression profiling
predicts clinical outcome of breast cancer. Nature 2002;415:
530–536.
9. Wang Y, Klijn JG, Zhang Y, et al. Gene-expression profiles to predict
distant metastasis of lymph-node-negative primary breast cancer.
Lancet 2005;365:671–679.
10. Rouzier R, Rajan R, Wagner P, et al. Microtubule-associated protein
tau: a marker of paclitaxel sensitivity in breast cancer. Proc Natl Acad
Sci USA 2005;102:8315–8320.
11. Chang JC, Wooten EC, Tsimelzon A, et al. Gene expression profiling
for the prediction of therapeutic response to docetaxel in patients
with breast cancer. Lancet 2003;362:362–369.
12. Battifora H. The multitumor (sausage) tissue block: novel method
for immunohistochemical antibody testing. Lab Invest 1986;55:
244–248.
13. Kononen J, Bubendorf L, Kallioniemi A, et al. Tissue microarrays for
high-throughput molecular profiling of tumor specimens. Nat Med
1998;4:844–847.
14. DiVito KA, Charette LA, Rimm DL, et al. Long-term preservation of
antigenicity on tissue microarrays. Lab Invest 2004;84:1071–1078.
15. Camp RL, Charette LA, Rimm DL. Validation of tissue microarray
technology in breast carcinoma. Lab Invest 2000;80:1943–1949.
16. Torhorst J, Bucher C, Kononen J, et al. Tissue microarrays for rapid
linking of molecular changes to clinical endpoints. Am J Pathol
2001;159:2249–2256.
17. Garcia JF, Camacho FI, Morente M, et al. Hodgkin and Reed-
Sternberg cells harbor alterations in the major tumor suppressor
pathways and cell-cycle checkpoints: analyses using tissue microar-
rays. Blood 2003;101:681–689.
18. Maitra A, Adsay NV, Argani P, et al. Multicomponent analysis of the
pancreatic adenocarcinoma progression model using a pancreatic
intraepithelial neoplasia tissue microarray. Mod Pathol 2003;16:
902–912.
19. Engellau J, Akerman M, Anderson H, et al. Tissue microarray tech-
nique in soft tissue sarcoma: immunohistochemical Ki-67 expression
in malignant fibrous histiocytoma. Appl Immunohistochem Mol
Morphol 2001;9:358–363.
20. Wang S, Saboorian MH, Frenkel EP, et al. Assessment of HER-2/
neu status in breast cancer. Automated Cellular Imaging System
108 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

(ACIS)–assisted quantitation of immunohistochemical assay achieves

high accuracy in comparison with fluorescence in situ hybridization
assay as the standard. Am J Clin Pathol 2001;116:495–503.
21. Sauter G, Simon R, Hillan K. Tissue microarrays in drug discovery.
Nat Rev Drug Discov 2003;2:962–972.
22. Camp RL, Chung GG, Rimm DL. Automated subcellular localization
and quantification of protein expression in tissue microarrays. Nat
Med 2002;8:1323–1327.
23. Camp RL, Dolled-Filhart M, King BL, et al. Quantitative analysis of
breast cancer tissue microarrays shows that both high and normal
levels of HER2 expression are associated with poor outcome. Cancer
Res 2003;63:1445–1448.
24. Psyrri A, Yu Z, Weinberger PM, et al. Quantitative determination of
nuclear and cytoplasmic epidermal growth factor receptor expres-
sion in oropharyngeal squamous cell cancer by using automated
quantitative analysis. Clin Cancer Res 2005;11:5856–5862.
Chapter 11
Basic Scientific Techniques in
Recording Cellular Data
George Z. Mentis, Yoshiyasu Arai, and Michael J. O’Donovan

Basic electrophysiological techniques have been used for years to

address fundamental biological questions in the peripheral and
central nervous system. Recently these have evolved to include
optical as well as electrical recording methods. This chapter, des-
cribes some of these methods and their application to the study of
neuronal organization in the neonatal mouse spinal cord.1,2
The major topics that are covered here include electrical
recording setup, extracellular recordings, intracellular record-
ings, morphology of single cells by intracellular staining, optical
recording, calcium-sensitive optical imaging, voltage sensitive
recordings, and intrinsic signal imaging.

ELECTRICAL RECORDING SETUP

To perform physiological experiments, investigators have to equip
the lab with many specialized pieces of equipment. These are used
to amplify, record, and analyze the data acquired from the animal
nervous system. To illustrate such techniques, we describe experi-
ments that are performed in our laboratory on the development
and operation of locomotor circuits in the spinal cord of the neo-
natal mouse. However, the basic principles we describe here also
apply to studies in other parts of the nervous system.
Setting up an electrophysiological laboratory, one has to keep
in mind that experiments should be reproducible and manage-
able. The tissue under investigation must be kept viable and as
close as possible to the in vivo physiological condition. Applica-
tion of modern recording methods usually necessitates removing
the tissue from the animal where it can be studied more conve-
niently than in situ. Once removed, the tissue must be maintained
alive in a bathing medium containing all the important ions and
buffers that are required for physiological function. In addition,
an energy supply must be provided—typically glucose—and the
bathing solution must be oxygenated. The tissue is then placed in
a chamber where physiological studies can be performed.
110 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

One great advantage of studying the spinal cord in isolation

is that reflexes and even complex behaviors are preserved. These
can then be examined without anesthesia or the constraints
imposed by whole animal in vivo studies. Indeed, it is possible
to activate locomotor-like behavior in which the motoneurons
supplying the limbs discharge in a manner closely resembling
overground locomotion in the intact animal. To understand how
these circuits function, it is necessary to define the properties of
individual neurons, their connections with each other and their
operation in the locomotor circuit.
Studies of connectivity between neurons are usually accom-
plished by electrically stimulating one class of neuron and record-
ing the responses in its presumed neuronal targets. For instance,
a muscle cell will contract, or a nerve fiber will initiate an impulse,
if an electric current of suitable size and duration is passed
through the tissue. The conductors which are used to deliver the
current to the tissue are called “electrodes.” The apparatus which
produces the electric current is called the “stimulator.” Stimuli
can be delivered singly or in trains of pulses depending on the
requirements of the experiment.
Specialized instruments are used to record the result of these
stimuli. The instruments vary both in their sensitivity and in their
ability to follow slowly or rapidly occurring events. The highest
demands are found in experiments where a combination of
steady and rapidly-changing potentials of small size must be
recorded simultaneously. This requires an “amplifier” and record-
ing system with a fl at frequency response from zero to several
tens of KHz and a maximum sensitivity sufficient to detect
changes of a few microvolts. Traditionally, the instrument of
choice to visualize these amplified voltages has been the oscillo-
scope but personal computers containing circuitry to convert
analog signals into digital form are now more commonly used
to display, process, store, and analyze the signals.

EXTRACELLULAR RECORDINGS
Extracellular recordings involve the acquisition of electrical
signals from a population of neuronal processes (e.g., axons) or
the electrical “field” generated in the space surrounding the cell
bodies or dendrites of a group of neurons. These electrical signals
are typically very small and require substantial amplification.
One method for acquiring and recording such small signals is by
the use of “suction” electrodes. Suction electrodes, comprise a
glass or plastic tipped capillary filled with a conductor (usually
saline) into which a nerve or a piece of central nervous system
is drawn. With such electrodes, it is possible to record two types
 11. BASIC SCIENTIFIC TECHNIQUES IN RECORDING CELLULAR DATA 111

of signal from a nerve or fiber tract. The first corresponds to the

electrical activity accompanying action potentials as they flow
along the nerve. In the spinal cord such signals are generated in
motoneurons by spinal circuits and ow fl out of the cord along
motor nerves where they can be recorded (Figure 11.1D, spikes).
A second type of record corresponds to slow synaptic activity at
the motoneuron cell body and dendrites that can be recorded as
an electrotonic signal at the tip of the suction electrode (Figure
11.1D, slow potential). Both types of signals are measured as the
voltage across the nerve/electrode junction resistance. A more
detailed explanation of the electrical model used to explain this
type of recording has been published.3
Among the important design considerations of these elec-
trodes are 1) the inner diameter of the capillary electrode must
be chosen carefully to accommodate the nerve or fiber tract under
study, 2) a flexible piece of plastic tubing must be attached to the
other end of the electrode and connected to a small syringe to
produce the negative pressure required to draw the nerve into the
capillary, and 3) two chlorided silver wires are used (one inside
the capillary and the other outside) to record the difference in
voltage across the two wires or electrodes. In this way, signals
that are common to both wires can be cancelled out, leaving only
the response from the wire detecting the neural signal. Suction
electrodes can also be used to deliver a stimulus to the tissue by
applying a brief current or voltage pulse across the electrodes.
Multiple electrodes are used to study complex behaviors such
as locomotor activity where at least four extracellular electrodes
are required to detect the alternating rhythmic activity between
the two sides of the spinal cord and between the rostral and
caudal segments.

INTRACELLULAR RECORDINGS
One of the most important signals that can be recorded from the
nervous system is the electrical potential across the membrane
of individual neurons. This measurement is accomplished either
by inserting a very fine glass electrode (sharp electrode) inside a
neuron or by sealing a blunter electrode (patch electrode) to the
membrane and then rupturing the membrane under the tip to
gain access the cell interior (Figure 11.1D). The electrodes are
connected to specialized amplifiers that can compensate for their
high capacitance and resistance which would otherwise filter the
electrical signals. These electrical signals provide information
about the electrical properties of the individual neurons, their
connections with other neurons and their activity during complex
behaviors such as locomotion.
112 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

A Alexa 568 hydrazide B 2D projection

Dendrites
Motoneuron
soma

Oregon Green BAPTA-1

(calcium sensitive indicator) Intracellular electrode
50µm
C 2 100
C1 C2 C3 C4
80
3 1
60
40
20
0
∆F/Fo(%) ∆F/Fo
0 20 40 60 80 100
(%)
D ROI:1
(Soma)
Optical

ROI: 2
RECORDINGS

(dendrite)
∆F/Fo
20%
ROI: 3
(dendrite)
40mV
Electrical

Vm motoneuron
(intracellular recording)
100µV
Right vr-L6
(extracellular recording)
Right vr-L5 electrical stimulation 2 sec

FIGURE 11.1. Combined electrical and optical recordings from neonatal

mouse spinal motoneurons. (A) Image of a neonatal motoneuron during
intracellular recording and stained with a uorescent
fl marker (Alexa 568
hydrazide). (B) Confocal image of the same motoneuron after with-
drawal of the intracellular micropipette. The dotted box indicates the
area shown in the panels below (C-C4). (C) The motoneuron was also
injected with a calcium-sensitive dye (Oregon Green BAPTA-1). An aver-
aged image of the intracellular uorescence
fl is illustrated together with
3 regions of interest (ROI) (blue: soma, green, and red: primary den-
drites). C1–4, confocal images showing the changes in fluorescence of
the calcium dye during spontaneous activity (C1), quiescence (C2 and
C4) and evoked bursting (C3). (D) Optical signals measured from the 3
ROIs shown in C (blue, red, and green), are shown together with the
intracellular membrane potential (Vm motoneuron) and spike discharge
and slow potential recorded extracellularly from the ventral root of the
same spinal segment (Right vr-L6). The box in green highlights the
increased somatic and dendritic optical signal corresponding to the burst
of action potentials recorded intracellularly. Electrical stimulation of an
adjacent ventral root (denoted by the horizontal violet bar) evoked a
similar episode of bursting and optical activity.

MORPHOLOGY OF SINGLE CELLS BY

INTRACELLULAR STAINING
The morphological identification of intracellularly recorded
neurons provides a powerful means of studying structure-func-
tion relationships in the CNS. Intracellular electrodes can be
 11. BASIC SCIENTIFIC TECHNIQUES IN RECORDING CELLULAR DATA 113

used to inject a dye inside the cell in order to reveal its cytoar-
chitecture. Ideally, the intracellular label should have the follow-
ing characteristics: Firstly, it should be easily ejected from the
intracellular micropipette without clogging; secondly, it should
be water soluble and diffuse quickly so that the cell can be filled
in a reasonable time; thirdly, it should not interfere with the
function of the neuron and finally, it must be easily visualized in
the light or fluorescence microscope.
Horseradish peroxidase (HRP) was one of the first successful
intracellular labels. Since then, low molecular weight markers
such as the biotin-lysine complex (biocytin; MW = 372) and neu-
robiotin (MW = 323) have been widely employed. In addition,
several uorescent
fl markers (Lucifer Yellow, Alexa hydrazides)
have been used to visualize the injected cells using a fluorescent
microscope (Figure 11.1A, B). The use of fl uorescent markers is
particularly useful when it is necessary to visualize the neuron
while obtaining intracellular recordings. Most of the intracellular
markers carry a positive or negative charge which is determines
how the marker is injected to the cell. For example, biocytin is
negatively charged and requires negative current pulses for injec-
tion whereas neurobiotin is positively charged and therefore
requires positive current.
OPTICAL RECORDINGS
Optical recording methods offer another set of powerful tools for
investigating neuronal and network function. They have the
advantage of being noninvasive and are capable of resolving the
activity of many cells simultaneously. Currently, optical tech-
niques fall into three categories. First, uorescent
fl probes of
intracellular ion concentration (e.g., calcium, sodium and chlo-
ride); second, direct measurement of membrane potential using
voltage-sensitive dyes; third, intrinsic signal imaging which mon-
itors the changes in tissue properties (light scattering, hemoglo-
bin oxygenation) that accompany neuronal activity. In the spinal
cord, we have used all three methods to monitor the activity and
spatio-temporal dynamics of individual neurons and neuronal
populations during several refl ex and complex behaviors. 4
CALCIUM-SENSITIVE OPTICAL IMAGING
Calcium-sensitive dyes exhibit the largest changes in fluorescence
on binding to their target ion.5 This is, in part, because calcium
ions undergo much larger changes in intracellular concentration
than other ions (often 100-fold), and so provide a very easily
detected indirect signal of neuronal activity. A critical component
of experiments involving ion-sensitive dyes is the loading of the
neurons under investigation. Several successful approaches have
114 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

been reported, ranging from direct injection into the tissue using
a membrane-permeable type of dye (AM),6 retrograde loading,7
and electroporation.8 The particular loading method employed
will be dictated by the requirements of the study.
Ion-indicator dyes change their uorescence
fl when the dye
binds the free ion in question. Of course, care must be taken
when using such dyes not to “buffer” the ions which can change
their dynamics and possibly the neuronal function being studied.
For this reason, the lowest useable concentration of the dye
should be employed. When using conventional epifluorescence
microscopy, these changes in uorescence
fl are monitored with
sensitive charged coupled device (CCD) or intensified video
cameras. Such devices usually operate at 30 frames/sec but spe-
cialized cameras can employ much higher frame rates.
Alternatively, the uorescent
fl signals can be detected using
confocal or multiphoton microscopy (Figure 11.1). Multiphoton
microscopy exploits the fact that uorophores
fl exposed to very
brief laser pulses can absorb two photons at a time instead of
the usual single photon. Each of the absorbed photons is approxi-
mately double the wavelength of the single photon that is nor-
mally absorbed. This has several major advantages. First, long
wavelengths penetrate biological tissue with less scattering than
shorter wavelengths and can visualize labeled cells up to several
hundred micrometers below the surface. Second, because the
probability of the uorophore
fl absorbing two photons is highest
at the focal plane, only a thin slice of tissue is fluorescent, thereby
resulting in reduced phototoxicity. Finally, 2-photon microscopy
allows the use of nondescanned detectors which do not use a
pin-hole to achieve confocality and, as a result, collect both the
direct and scattered light emitted from the fluorophore.9

VOLTAGE SENSITIVE RECORDINGS

Several classes of dye have been introduced that monitor neuro-
nal membrane potential. In general, these are lipophilic dyes that
diffuse into the neuronal membrane where their fluorescence
and orientation in the membrane is infl uenced by the transmem-
brane potential. They are usually bath-applied to the tissue in
question although they may also be introduced into defined neu-
ronal populations by retrograde loading.10 Highly specialized
equipment must be used to detect the changes in fluorescence
accompanying changes of membrane potential because these
changes are often extremely small (1/1000 of the resting light
level). As a result, the detectors must have great sensitivity and
a very wide dynamic range (up to 17 bits). Typically, this involves
the use of an array of photodiodes (128–464) that are coupled to
 A amplifiers computer
I-V converter

photodiode array
multiplexer

amplifiers
objective

Extracellular electrodes

Artificial cerebrospinal fluid “out-flow”

“in-flow” “in vitro” chamber
Tissue
condenser
interference filter
light source
mirror

DC power supply

B C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
A 1 2 3
cc

4 5 6 7
Optical

Right VR 8 9 10 11
(extracellular)

12 13 14 15
Left VR
(extracellular)

25 ms

FIGURE 11.2. Combined electrical and voltage-sensitive optical signals.

This figure shows the spread of activity at the beginning of an episode
initiated by a single stimulus to the ipsilateral (right) dorsal root in the
neonatal chick spinal cord slice in vitro. (A) diagram of the apparatus
used for optical and electrical recordings. The slice preparation was
continuously perfused with artificial cerebrospinal solution in an in vitro
chamber on the stage of a modified Nikon microscope (OPTIPHOT,
Japan) mounted on an air table. (B) Right and left ventral root (VR)
electrical responses together with the optical signals from three different
cord regions (ventral/lateral motor column: red; intermediate: green;
dorsal: blue). Each trace was averaged from 4–6 adjacent diodes (indi-
cated on panel A of part C) and was normalized to its peak amplitude.
Data were obtained at a sampling interval of 0.64 ms. The gray lines
indicate the timing of 10 frame averages (each 6.4 ms) corresponding to
the pseudocolored images shown in C. The dotted gray line corresponds
to the onset of the dorsal root stimulus (arrowhead). (C) Montage of
the pseudocolored optical signals from the array, superimposed on the
outline of the cord (dotted gray lines). The outlines of the motor nucleus
and the primary afferents are shown by dotted black lines. The first image
(A) shows a pseudocolored image of the cord following antidromic stimu-
lation of the ipsilateral ventral root to identify the location of motoneu-
rons (indicated by the red-outlined diodes). cc—central canal. The arrow
in panel 12 shows the activity propagating to the contralateral lateral
motor column. Data from an E11 chick spinal cord embryo.
116 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

individual amplifiers under computer control. An example of this

type of recording is shown in Figure 11.2.
INTRINSIC SIGNAL IMAGING
One may notice that when neural tissue becomes active, its light
absorption changes. Although these changes are not fully under-
stood, they are believed to originate from activity-dependent
alterations in neuronal birefringence and tissue volume. What-
ever the source of these changes, they can be exploited to provide
a noninvasive measure of neural activity in the absence of dye-
labeling and its attendant phototoxicity. One disadvantage of
these signals over those originating from voltage-sensitive dyes
is that they are significantly slower and cannot be used to detect
rapid events such as individual action potentials. Nevertheless,
we have successfully employed this technique to visualize the
patterns of activity accompanying locomotor-like behavior in the
isolated mouse cord.
References
1. Andrew BL, ed. Experimental Physiology, 8th ed. London: E&S Liv-
ingstone, 1969.
2. Ketterman H, Grantyn R, eds. Practical Electrophysiological Methods:
A Guide for in vitro Studies in Vertebrate Neurobiology. New York:
Wiley-Liss, 1992.
3. Stys PK, Ransom BR, Waxman SG. A compound action potential of
nerve recorded by suction electrode: a theoretical and experimental
analysis. Brain Res 1991;546:18–32.
4. O’Donovan MJ, Bonnot A, Wenner P, Mentis GZ. Calcium imaging
of network function in the developing spinal cord. Cell Calcium
2005;37:443–450.
5. Baker BJ, Kosmidis EK, Vucinic D, et al. Imaging brain activity with
voltage- and calcium-sensitive dyes. Cell Mol Neurobiol 2005;25:
245–282.
6. Tsien RY. A non-disruptive technique for loading calcium buffers and
indicators into cells. Nature 1981;290:527–528.
7. O’Donovan MJ, Ho S, Sholomenko G, Yee W. Real-time imaging of
neurons retrogradely and anterogradely labelled with calcium-sensi-
tive dyes. J Neurosci Meth 1993;46:91–106.
8. Bonnot A, Mentis GZ, Skoch J, O’Donovan MJ. Electroporation
loading of calcium-sensitive dyes into the CNS. J Neurophysiol
2005;93:1793–1808.
9. Denk W, Yuste R, Svoboda K, Tank DW. Imaging calcium dynamics
in dendritic spines. Curr Opin Neurobiol 1996;6:372–378.
10. Wenner P, Tsau, Y, Cohen LB, O’Donovan MJ, Loew LM. Voltage
sensitive dye recording using retrogradely transported dye in the
chicken spinal cord: staining and signal characteristics. J Neurosci
Meth 1996;70:111–120
Chapter 12
Presenting and Publishing
Research Data
Howard A. Bird

INTRODUCTION
Publication of research is an important though often neglected
aspect of the research pathway. Although often perceived as the
final link in the chain, an ever increasing emphasis on audit dic-
tates that the research ultimately with be judged on the quality
of publications emanating from it. In turn, this will influence the
availability of future research grants. Therefore, it is sensible to
plan likely outlets for any research findings at the very earliest
stage of an application for funding. If, on consideration, there
are unlikely to be any obvious outlets for the results, the would-
be researcher should give very careful thought to the value of
embarking on the research project in the first place.
In recent years, academia has been driven by the “Holy Grail”
of the Research Assessment Exercise with its dependence on
placing research publications in journals where they not only
attract attention but from which they will be regularly quoted by
other researchers. Essential to this is the concept of impact
factor, an index that is not without its critics.
Mercifully, an increasing number of funding bodies are start-
ing to attach equal importance to other important benefits from
research, including the development of a critical mind that leads
to writing and awarding of higher research degrees such as MD
and PhD, and the possible immediate benefits the research might
have to clinical practice. Quite clearly, the longer the period
devoted to research the greater the chances of higher quality
publication. Ultimately, this is still likely to be audited by the
grant awarding body and the quality of the research and the way
in which it is presented is likely to have impact on future funding
opportunities.
Sources of funding include major government or indepen-
dent bodies such as the Medical Research Council or the Well-
come Trust; at present arguably the most respected sources of
118 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

funding in United Kingdom universities. For National Health

Service (NHS) employees, funding may have been derived from
central or local NHS funds. Alternatively, funding may be derived
from the larger charities such as the British Heart Foundation,
Diabetes United Kingdom or the Arthritis Research Campaign,
or even the smaller “disease-based” charities of modest size (such
as the Scleroderma Association) or of very small size (such as
the Ehlers-Danlos Support Group), to give two examples from
my own speciality. Sometimes local university funds may be
tapped. A department may be the beneficiary of donations or
bequests that are intended to be used for research purposes.
Increasingly, applicants are looking towards Europe for largesse
amongst members of the European Community.
Industry still contributes to a substantial proportion of health-
care research worldwide. This type of research is less likely to be
directed to fundamental research and more likely to be linked to
industrial products that ultimately will be sold at a profit to
satisfy shareholders. This represents the easiest source of research
funding but arguably the one with the greatest number of strings
attached and one less likely to command the respect of university
or NHS research directly, however important this source might
be for the regular provision of “oft funds.”
Each funding body is likely to offer their own guidelines on
the publication and dissemination of results and the would be
researcher is strongly advised to check what is expected of them
and which outlet for the results will be judged most acceptable
by the funding body.

OUTLETS FOR RESULTS

A variety of outlets are available. Most respected and therefore
most desirable is to publish in a peer-reviewed journal, prefera-
bly one with high-impact factor. Although the choice of journal
is wide, in reality this may be quite restricted according to the
area in which the researcher is working. An editor, in selecting
material for publication, has obligations not only to the author
but also to the readership, the editorial board and, to a lesser
extent, the publisher. Journals with the highest impact factors
are likely to have achieved this because they are publishing high-
quality fundamental research in basic science, of appeal to a
large discriminating audience, or because they are “mainstream”
medical journals, taking articles only of the widest interest and
therefore those likely to be most quoted.
The majority of publications emanating from research will
be placed in a more specialized journal, inevitably with a lower
 12. PRESENTING AND PUBLISHING RESEARCH DATA 119

impact factor. If the speciality is a relatively small one, the impact

factor will always be significantly lower than in a more popular
speciality, even for the “best” journal. In general, the impact
factors for rheumatology journals are always higher than those
for journals of rehabilitation, though not as high as for journals
in gastroenterology or cardiology, both of which are larger
specialities worldwide.
Less satisfactory alternatives are to publish in a chapter
(though this normally requires an invitation), which is likely to
be less critically peer reviewed, or even to write a book, which
will be subject to very little peer review at all though the pub-
lisher, in commissioning it, will obviously take soundings.
A better ploy is to start by presenting work as a talk or poster
presentation to a learned society. Most speciality societies referee
submissions competitively. Poster presentations are particularly
valuable for young researchers since they enable the researcher
to meet other experts in the field and allow for discussion, which
may well lead to modification of the work. In turn, this acts as a
“dress rehearsal” before submitting a formal paper to a journal.
Indeed, the assessors who may later review it may even have
visited the poster!
Less satisfactory outlets are talks to industry-sponsored sym-
posia or even press releases. Sometimes the sponsors will require
a talk, perhaps as part of a review after one year, and then as a
prerequisite to further funding.

PUBLICATION IN PEER-REVIEWED JOURNALS

Ideally, research will be published as an original paper. If the
editor feels, on submission, that the work is a little thin for a
full-length paper, alternative outlets may be suggested. Several
journals publish “brief communications” or might suggest
revamping the work in the style of a “letter to the editor.” Although
this may come as a disappointment to the author, a letter to the
editor published in a high-quality journal may well enhance a
curriculum vitae more than a full-length paper in a journal of
low repute.
Sometimes, particularly for industry-sponsored work, publi-
cation may have to be in an industry-sponsored symposium pub-
lished as a supplement to the journal. These make a lot of money
for journals and although normally subjected that the peers
review them, this may be less searching than the conventional
review for an article in the main journal.
Editorials perhaps command most attention but these are
normally only likely to be written on the invitation of the editor,
120 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

though co-authorship of an editorial with a senior colleague is

always a good ploy.

THE SPECTRUM OF JOURNALS

Even when the research is being planned, it makes sense to
consider which journal will be targeted for publication in due
course.
For the best quality work of general interest, an international
mainstream independent journal such as the Lancet, British
Medical Journal, Journal of the American Medical Association, or
New England Journal of Medicine is likely to be selected.
For more specialized work, the researcher will focus on jour-
nals publishing research in that speciality. To take my own spe-
ciality, which is rheumatology, at the time of writing, the choice
is quite varied, with some 20 respected peer-reviewed journals
worldwide.
Inevitably, the major American journals have the widest read-
ership and, therefore, carry the highest impact factors. Arthritis
and Rheumatism publishes basic science and clinical research
applied to basic science, or the sister journal Arthritis Care and
Research with its more clinical flavor is likely to represent a first
choice. The second choice would be divided between the other
major North American journals, The Journal of Rheumatology
(published in Canada), or one of the two major general European
journals, Annals of the Rheumatic Diseases or Rheumatology.
That both of the last two are published from the United Kingdom
reflects the desirability of publishing in an English-language
journal, as well as the foresight shown in the establishment of
what is now the Arthritis Research Campaign years ago with the
encouragement that has been given to research in arthritis in the
United Kingdom over many decades.
If a paper bounces from this group, the author can choose
either to move to the house journal of one of the smaller national
societies (e.g., Clinical and Experimental Rheumatology from
Italy; Scandinavian Journal of Rheumatology or Clinical Rheuma-
tology, originally from Belgium) or to one of the disease-based
journals such as Lupus or an osteoarthritis journal.

IMPACT FACTOR
This is one of several indexes by which the quality of journal
is compared. Inevitably, it has been subject to some criticism
and there are alternative ways of measuring this. Table 12.1
provides the definitions for impact factor, immediacy index, and
cited half-life. Table 12.2 gives impact factors for a variety of
 12. PRESENTING AND PUBLISHING RESEARCH DATA 121

TABLE 12.1. Impact factor and variance

Impact Factor (“Impact”)

A measure of the frequency with which the “average article” in a
journal has been cited in a particular year. Thus, the impact factor
of journal X would be calculated by dividing the number of all
current citations to articles published in journal X during the
previous two years by the number of articles (“source items”)
journal X published in those two years.
Immediacy Index (“Immediacy”)
A measure of how quickly the “average article” in a specific journal
is cited. Thus, the immediacy index of journal X would be
calculated by dividing the number of all current citations of current
source items published in journal X by the total number of articles
journal X published in that year.
Cited half-life (“Cited Half”)
The number of publication years going back from 1990, which
account for 50% of the total citations received by the cited journal
in the current year. The “cited half-life” thus indicates for how long
articles in journal X stayed “alive,” i.e., were considered worthwhile
as citations.

general journals, rheumatological journals, and rehabilitation

journals.
The impact factor is a measure of the frequency with which
the “average article” in a journal has been cited in a particular
year. It is based on an analysis of the journal’s content in the
previous two years and that has sparked some criticism. Some

TABLE 12.2. Selected impact factors (2005)

General
New England Journal of Medicine 34.8
Nature 30.9
Science 29.7
Cell 26.6
Lancet 18.3
British Medical Journal 7.2

Rheumatology
Arthritis and Rheumatism (USA) 7.2
Annals of the Rheumatic Diseases 3.8
Rheumatology 3.7
The Journal of Rheumatology 2.9

Rehabilitation
Clinical Rehabilitation 1.0
122 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

have suggested that the immediacy index, based on a shorter

period of the most recent year, is a fairer index of the “cutting
edge” quality of research. Others have favoured the cited half-life,
which surveys a much longer period, as a better measure of
“durability.”
Although an index based on two years analysis probably rep-
resents the best compromise, other sorts of errors can creep in.
The change in name of a journal will cause it to drop temporarily
in the tables. A new journal will take some time to appear in the
tables. It is also theoretically possible for an editor to “fudge” the
analysis by the exclusive publication of articles that automati-
cally boost impact factors (such as those describing a new disease
for the first time or defining diagnostic criteria), however boring
to the readership such a journal might become.

WRITING THE PAPER

It is prudent to look through several papers previously published
in the journal to which you will be submitting, if only to capture
the style. A wider survey of recent issues of the journal will also
give an impression of whether the research paper you will finally
produce would be of interest to that particular journal or whether,
as a matter of policy, it would be directed elsewhere by the
editor.
Research is nowadays rarely done in isolation and the novice
is likely to have senior colleagues, probably as co-authors, to
whom he or she can turn when the paper is being written. In
general, the introduction is likely to have been already written
and referenced as part of a preceding grant application. If this
can be slanted slightly towards the style of the journal, perhaps
even quoting papers previously published in the same journal
that have prompted the study, that is obviously helpful. The
section “patients and methods” is also likely to be lifted almost
directly from the protocol. The “results” section should write
itself if the research has been well designed and executed. Again,
it is prudent to check previous issues of the journal to judge the
extent to which results should be presented as tables and dia-
grams and the depth of the statistical analysis that particular
journal will require. The “discussion” section is perhaps the
hardest part to write and should be well balanced and objective,
perhaps suggesting future research projects that the most recent
one has prompted. The “summary,” typically 150 words, is nor-
mally written last of all.
For the references, the majority of journals use Vancouver
style, but this also should be checked. Computer programs are
 12. PRESENTING AND PUBLISHING RESEARCH DATA 123

available to adapt references to a different style if the first journal

rejects it.

ASSESSMENT
Editors and journals vary in the style of assessment. A small
number use a single assessor and, typically, major international
general journals have a professional committee that considers
papers as well. The editors of speciality journals tend to send
papers to three assessors, one of whom may well be a member
of the editorial board. Since the composition of the board is
published in the journal, it is therefore normally possible to
predict whom one of the assessors might be! If a paper spans
several disciplines the group of referees is likely to be representa-
tive of each of the various disciplines. Sometimes papers are sent
routinely to a statistician; in other cases, this is only done if one
of the first groups of assessors specifically requests it.
Once a decision is made, this is relayed back to the author,
normally with copies of the different assessors comments, the
assessor usually also receiving copies of each others comments,
which not only educates assessors as well as authors but also
provides an element of internal audit.
Authors are sometimes disappointed if the editor’s final deci-
sion does not at first seem to be in line with the feedback pro-
vided for authors. This is because, in addition to providing
authors feedback, each assessor has the opportunity to make
comments to the editor in confidence that the authors will not
see. An editor may also decide to weight the opinion of a senior
and particularly trusted assessor more highly than that of a
trainee assessor to whom the paper has also been sent. Ulti-
mately, an author may also fall foul of the pages available to be
filled (which correlates with the cost of the subscription to
readers) and the particular balance of contents in recent issues
of the journal, the editor needing to keep the readership happy
as well as the authors. If a junior researcher has had work
rejected for one of these reasons, this should normally be
explained in a letter from the editor.
It is unusual for a paper to be accepted first time without any
change. If reconsideration of the work is offered subject to the
changes suggested by assessors, the author is well advised to
make these changes since it then becomes extremely hard for the
editor and the original assessors to reject it.
If the rejection seems unfair and the assessors’ comments
particularly critical, the author can always appeal to the editor
with a set of reasonable arguments supplied to him/her. In this
124 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

situation, the editor may sometimes arrange for reassessment or

may take the additional view of a further assessor, typically a
senior member of the board. Invariably, demand outstrips the
space available, particularly in a popular journal, and it is not
unusual for papers to do the rounds of several journals one after
another, the paper getting a little better each time as a result of
the multiple assessors’ comments! In general, papers are offered
to journals with a lower impact factor if they have been turned
down by one with a high-impact factor. Where a choice has to
be made between several journals of equal impact factor, priority
would normally be given to one that either specializes in publish-
ing in that particular area or one for which the author feels a
particular rapport with the editor, perhaps as a result of previous
publications in that journal.

FUTURE TRENDS
Publication of results is not just to enhance the curriculum vitae.
It is also seemed important by ethics committees since a failure
to disseminate the results of research might be considered uneth-
ical if patients had been exposed to risk by participation in a
study from which nothing was then learnt. A discussion of future
dissemination therefore forms an integral part of ethics commit-
tee submission and would also be of interest to the research
directorate of a trust since publicity through dissemination of
research can only enhance its reputation.
Funding charities will also be keen to ensure that their money
has been spent in optimal fashion. Here, a broader view may be
taken. Given that the use of impact factors can be justifiably
criticized, though may still represent the best compromise, are
stakeholders best satisfied with a large output of papers that are
only occasionally cited or is payback better with a smaller number
of papers that are cited more frequently? Many are now deeming
successful MD and PhD dissertations as almost important since
these emphasize the educational importance of research to the
researcher and seed research for future generations. Put another
way, the teaching component of research may be as important
as the research itself. By contrast, the NHS, as an organization
responsible for patient care, may attach most importance of all
to changes in clinical practice that might result from the research,
especially if this demonstrates options for saving the managers
budget by recommending less expensive clinical practice that is
equally as effective as more expensive older methods.
It should also be noted that within two to three years major
journals are going to insist on an international standard of ran-
 12. PRESENTING AND PUBLISHING RESEARCH DATA 125

domized controlled trial number allocation at the start of the

research project as a prerequisite for consideration for later pub-
lication when this is appropriate. This is perceived as an audit
of randomized controlled clinical trials, both ensuring their
quality and also encouraging the publication of results, if they
are good or bad to the financial interests of the funding body.

Ten key points

1. Plan publication from outset.
2. Aim for publication in a peer reviewed journal.
3. Select on the basis of impact factor amongst other things.
4. Benefits in care may be as important as advances in
fundamental knowledge.
5. A higher research degree is also a prestigious outcome.
6. Aim to present as a poster before submitting to journal.
7. Read the instructions to authors and follow them.
8. Concede the majority of points made by assessors when
revising.
9. Try several journals in turn if necessary.
10. Publication as a letter in good a journal may be better than as a
paper in a poor journal.
Chapter 13
Analyzing Health Studies
Julie A. Barber, Gareth Ambler, and Rumana Z. Omar

OVERVIEW
This chapter will consider commonly used methods for describ-
ing and analyzing data. We begin with an introduction to some
important basic statistical concepts and then focus on some of
the most well used methods of analysis for observational and
intervention studies. The two types of data we will discuss in
detail are continuous and binary data. For further reading, we
recommend consulting a medical statistics textbook.1,2,3,4

TYPES OF DATA

Numerical Data
Numerical data are either continuous or discrete. Continuous
variables can take any value within a given range; examples are
height, weight, and blood pressure. Discrete variables take only
certain values, usually integers. Number of GP visits per year and
other count variables are examples of discrete variables. Discrete
variables may be treated as continuous if they have a wide range
of values.

Categorical Data
A variable is categorical if individuals are assigned to one of a
number of categories. The simplest form is a binary variable with
two categories, for example, alive/dead, pregnant/not-pregnant,
and cancer/cancer-free. With more than two categories, the vari-
able is either nominal, where there is no ordering of the categories,
such as with blood group (A/B/AB/O), or ordinal, where there is
ordering such as with pain score (none/mild/moderate/severe).

Data Structure
Understanding the structure of data is essential, as it determines
the appropriate method of analysis. Most statistical methods
assume that subjects or observations are independent of one
 13. ANALYZING HEALTH STUDIES 127

another. In some cases, however, we may have data where the

observations are not independent, such as with paired data where
subjects are measured twice (e.g., crossover trials) or are matched
(e.g., matched case-control studies). In general, clustered data
occur where sets of subjects or observations are clustered within
larger units (e.g., patients within a general practice, or measure-
ments taken repeatedly over time for each patient).

DESCRIBING DATA
It is important to perform a descriptive analysis, as it allows us
to gain a basic understanding of the data. This helps to select the
appropriate method of analysis as well as identify problems such
as outliers (observations inconsistent with the rest of the data)
and missing values. We can examine and summarize the data
using graphs, tabulations, and descriptive statistics.

Continuous Data

Displaying and Assessing the Shape of the Data

Histograms are commonly used to display continuous data. The
range of the variable is typically divided into intervals of equal
width and the data displayed as a set of vertical bars with heights
proportional to the number of observations in each interval.
Often the histogram shows a symmetrical, “bell-shaped” normal
distribution for the data. This distribution is fundamental to sta-
tistics with many statistical methods requiring assumptions of
normality. A more reliable graph to assess normality is the normal
plot. This plots the data against the values you would expect to
observe if the data were actually Normally distributed; hence a
straight line of points indicates Normality.1

Descriptive Statistics
We summarize continuous data with measures of central ten-
dency (the location of the middle of the distribution) and spread
(the variability of the distribution). The appropriate measures
depend on the research question and the shape of the distribu-
tion of the data.

Mean and Standard Deviation

The mean (or average) and standard deviation (SD) are the most
commonly used measures for summarizing continuous data. The
standard deviation (and its square, the variance) quantifies the
average distance of the observations from the mean. These mea-
sures are sensitive to extreme observations in the data. When the
128 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

data have a Normal distribution, approximately 95% of the data

lie within 2 standard deviations of the mean.

Median and Interquartile Range

For non-normal distributions, it may be preferable to use quar-
tiles to summarize the data, because these are less affected by
extreme observations. The upper, median, and lower quartiles
are the values that divide the data into quarters after arranging
the data in ascending order. The median measures the central
tendency and splits the data in half. The difference between the
lower and upper quartiles is known as the interquartile range, and
measures variability.

Categorical Data
To summarize categorical data, we calculate the number and
proportion (or percentage) of subjects that fall into each cate-
gory. For a graphical illustration of categorical data, one could
use a bar chart, which is similar to a histogram but displays bars
for each category.

ANALYZING DATA
As mentioned in the chapter on study design, we usually collect
data from a representative sample of individuals in order to make
inferences about the target population of such individuals. For
example, to evaluate the effectiveness of a new treatment for
breast cancer, we might evaluate the treatment on a sample of
breast cancer patients from two UK centres, and then draw con-
clusions about the likely usefulness of this treatment for all
breast cancer patients.
However, statistical uncertainty arises because we have infor-
mation for only one of many potential samples that could have
been taken from the population. The two basic methods of quan-
tifying this uncertainty are estimation and hypothesis testing.

Estimation
We use our study sample to estimate population characteristics.
For example, the mean blood pressure amongst a sample of British
men might be used to estimate the true mean blood pressure for
all British men. We could be interested in descriptive measures
such as proportions or means, or comparative measures, such as
relative risks, or differences in means between two groups.

Sampling Distributions
A sample estimate is unlikely to be exactly equal to the true
population value and different samples will provide different
 13. ANALYZING HEALTH STUDIES 129

estimates. The distribution of estimates from different samples

is called the sampling distribution and is used to quantify statisti-
cal uncertainty. Because we only have one sample, we often have
to make assumptions about this distribution (e.g., Normality).
When estimating population characteristics using a sample, it is
essential that we report on the statistical uncertainty. This is
achieved using either standard errors or confidence intervals.

Standard Errors
The standard error (SE) is used to quantify how precisely the
population characteristic is estimated by the sample. It is the
estimated standard deviation of the sampling distribution with
smaller values of the standard error indicating higher precision
and less uncertainty. Larger sample sizes give smaller standard
errors and hence more precise estimates.

Confidence Intervals
It is often more useful to construct a range for the likely values
of the population characteristic. The sample estimate and its
standard error can be used to provide such a range in the form
of a confidence interval. Conventionally we calculate 95% confi-
dence intervals, which we can think of as the interval likely to
contain the true population value with a probability of 95%.
Because these intervals are obtained from the sampling distribu-
tion of the parameter, the formal interpretation of confidence
intervals is in terms of taking many samples from the population.
If we took 100 samples from the population and calculated a 95%
confidence interval for each, we would expect 95 of these inter-
vals to include the population value.
The upper and lower confidence limits are often calculated
as
estimate ± multiplier × SE
where the multiplier is derived from the sampling distribution.
For example, an estimate of the average age of patients undergo-
ing elective cardiac bypass graft surgery is 60.5 years (SE = 1.7
years) based on a sample of 31 such patients from a London
hospital. A 95% confidence interval showing the likely range for
the true average age is 57.0 to 64.0 (60.5 ± 2.0 × 1.7).

Hypothesis Testing
The sample data can be used to test a predefined statistical
hypothesis about a population characteristic. Typically this null
hypothesis describes situations of no effect or no difference. For
example, when investigating the possible link between respira-
130 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

tory infection and sudden infant death, we would test the null
hypothesis that there was no link. There are many statistical tests
suitable for particular hypotheses and types of data, and all
produce probability statements called P-values.

P-values
To test our hypothesis, we consider how much evidence there is
in the data against the null hypothesis. The amount of evidence
is quantified in the form of a P-value, which lies between 0 and
1. Large P-values suggest we have little evidence that the null
hypothesis is untrue, and so we do not reject the null hypothesis.
Large P-values do not tell us that the null hypothesis is true.
Small P-values tell us we have evidence against the null hypoth-
esis, and we would reject the null hypothesis.
Formally, P-values are the probability of observing similar or
more unlikely data than our own sample when the null hypoth-
esis is true.

Statistical Significance
Often we refer to small P-values as statistically significant and
large P-values as nonsignificant. A cut-off of 0.05 conventionally
defines statistical significance, but this is arbitrary and hence
should be used cautiously.

Over Interpretation of P-values

It is important to note that small P-values do not necessarily
imply clinically important effects, and large P-values do not tell
us there is no effect. By examining the estimate and confidence
interval, we can make a better interpretation. The size of the
estimate allows us to judge clinical significance. Wide confidence
intervals may indicate that a large P-value has occurred because
of low power (limited sample size) (see design chapter on study
design for definition of power).

EXAMPLES
Using examples, we illustrate the basic methods of analysis for
continuous and binary outcomes and look at the interpretation
of statistical results. We focus on the comparison between two
groups of data, both independent and paired. The results shown
can be obtained using any good statistical software.

Two Independent Groups of Continuous Data

To investigate the relationship between behavior and risk
of heart disease, we use the cholesterol levels from 60 men
 13. ANALYZING HEALTH STUDIES 131

categorized into two behavior groups: type A: urgent, aggressive

and ambitious; type B: relaxed, noncompetitive and less
hurried.1

Describing the Data

The data are shown in the format required for most statistical
software (Table 13.1). Cholesterol level is a continuous outcome
and its distribution in each behaviour group can be examined
using histograms and Normal plots (Figure 13.1). Cholesterol

TABLE 13.1. Cholesterol levels (mmol/l) by behaviour type. Format

required for analysis: id = patient identifier, cholesterol (mmol/l), behavior
group (0 = type A, 1 = type B)

Id Cholesterol Group id cholesterol Group

1 236 1 31 178 1
2 209 1 32 242 1
3 253 0 33 273 0
4 250 1 34 164 1
5 156 1 35 185 1
6 281 1 36 153 0
7 251 0 37 218 1
8 201 1 38 187 1
9 257 1 39 212 0
10 203 1 40 248 0
11 230 0 41 255 0
12 210 0 42 158 1
13 291 1 43 234 0
14 278 0 44 268 0
15 263 0 45 194 1
16 241 0 46 188 1
17 270 0 47 212 0
18 227 1 48 272 1
19 186 1 49 165 1
20 236 1 50 212 1
21 228 1 51 260 0
22 246 1 52 218 1
23 185 0 53 261 1
24 212 1 54 244 1
25 280 0 55 207 0
26 244 1 56 248 0
27 294 1 57 304 0
28 276 1 58 204 0
29 173 0 59 230 0
30 202 1 60 181 1
 132 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

Type A Type B
8

8
6

6
Frequency

Frequency
4

4
2

2
0

0
150 200 250 300 150 200 250 300
total cholesterol total cholesterol

300
300

250
250
total cholesterol

total cholesterol
200
200

150
150

150 200 250 300 150 200 250 300

Inverse Normal Inverse Normal

FIGURE 13.1. Histograms and Normal plots of cholesterol levels by

behaviour type.

levels in each group appear to be approximately Normally

distributed.
Summary statistics suggest that mean cholesterol for type A
men is higher than for type B (Table 13.2). The standard devia-
tions in the groups are reasonably similar.

Analysis: Two-Sample T-test

To formally compare sample means between two independent
groups, we use parametric methods based on the t-distribution

TABLE 13.2. Mean (SD) cholesterol level by behaviour type

Type A Type B

Mean (mmol/l) 237.7 220.1

SD (mmol/l) 36.2 38.6
N 25 35
 13. ANALYZING HEALTH STUDIES 133

(which is related to the Normal distribution). These assume that

the data in both groups come from Normal distributions and the
groups have equal population standard deviations. The first
assumption is important for small samples; we should check
Normality if group sizes are less than 30. Our sample size is
moderate, but the distributions appear approximately Normal
(Figure 13.1), and the standard deviations are similar (Table
13.2), so we can go ahead and use these methods.
We are interested in whether cholesterol differs by behavioral
type, so the parameter of interest is the difference in mean cho-
lesterol between the groups. This is estimated as 17.6 mmol/l with
a standard error of 9.9 mmol/l. The latter is relatively small, indi-
cating some precision. A 95% confidence interval for the differ-
ence is −2.1 to 37.4 mmol/l (approximately 17.6 ± 2 × 9.9). We
can think of this as the interval within which we are 95% confi-
dent that the true difference in mean cholesterol in the popula-
tion lies. The confidence interval suggests that type A personalities
may, on average, have higher cholesterol values (i.e., differences
greater than zero). However, the interval also includes the pos-
sibility of no difference in means (i.e., a difference of zero).
We use the two-sample t-test to test the null hypothesis that
mean cholesterol levels are the same in each group. The t-test
gives a P-value of 0.08. This indicates an 8 in 100 chance of
obtaining a sample like ours if there is really no relationship
between behavior and cholesterol. That is, there is some evidence
that cholesterol levels differ by behavior type.

Violation of Assumptions
If the assumptions of the t-test are violated, we may use a differ-
ent approach. Welch’s test is suitable if the data are approxi-
mately Normal but the variances are unequal.4 When there is
non-normality, transforming data to a scale where assumptions
are met may be useful; for example, analyzing log-transformed
data and reporting results in terms of geometric means. Alterna-
tively, we may use a nonparametric method, which does not
require Normality assumptions. The Mann-Whitney U (or Wil-
coxon Rank Sum test) is the nonparametric counterpart of the
two-sample t-test. The latter compares the medians of the two
groups, assuming that their distributions have identical shapes.

Two Paired Groups of Continuous Data

An example of paired data is a matched case-control study to
investigate risk factors for cardiovascular disease (CVD) in
134 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

TABLE 13.3. Mean (SD) triglyceride level for each CVD group, and the
group of differences

CVD no CVD CVD—no CVD

Mean (mmol/l) 1.77 1.29 0.49

SD (mmol/l) 0.19 0.14 0.23
N 22 22 22

patients with Systemic Lupus Erythematosus (SLE).2 Twenty-

two patients with SLE and CVD were individually matched by
age, sex, and ethnic group to patients with SLE but no CVD. We
consider the continuous outcome triglyceride concentration.

Describing the Data

Summary statistics suggest that mean triglyceride concentration
is higher in the CVD group (Table 13.3). However, because data
are paired, we need to summarize the differences between the
paired measurements. We create a new variable containing the
differences between the concentrations for CVD and matched
non-CVD patients. The mean (SD) of these differences is 0.49
(0.23) mmol/l (Table 13.3). A Normal plot of these differences
suggests approximate Normality (Figure 13.2).
3
2
CVD - no CVD
0 -1
-2 1

-1 0 1 2 3
Inverse Normal

FIGURE 13.2. Normal plot of the differences in triglyceride levels between

matched CVD groups.
 13. ANALYZING HEALTH STUDIES 135

Analysis: The Paired T-test

The usual parametric approach for paired data is the paired t-test
method. This assumes that the differences between paired
measurements have a Normal distribution. For small samples,
this should be checked using a Normal plot. The parameter of
interest is the mean of these differences. We can obtain an esti-
mate of this mean and its standard error and confidence
interval.
For the triglyceride data, the paired t-test methods seem
appropriate (Figure 13.2). The mean (SE) of these differences is
0.49 (0.23) mmol/l, indicating a reasonable level of precision. A
95% confidence interval for the mean of the differences is 0.01
to 0.96 mmol/l. This suggests the true mean difference is likely to
be positive, indicating that SLE patients with CVD have a higher
mean triglyceride concentration.
A Paired T-test formally tests the null hypothesis that the
mean difference is zero. For the triglyceride data, this gives a P-
value of 0.045. Thus, we have evidence to reject the null hypoth-
esis that there is no relationship between CVD and triglyceride
concentration (i.e., that the mean difference is zero). This result
is consistent with the interpretation of the 95% confidence
interval.

Violation of Assumptions
If the differences between paired measurements are severely
non-normal, it may be preferable to transform the original data
or use a nonparametric method such as the Wilcoxon signed rank
test.

5.3 Two Independent Groups of Binary Data

A trial was performed to compare neurocognitive impairment
after cardiac bypass graft surgery. Sixty patients were random-
ized to either off-pump or on-pump surgery, and the binary
outcome was neurocognitive impairment (yes/no) one week fol-
lowing surgery.7

Describing the Data

The complete dataset can be summarized as a 2 × 2 table (Table
13.4). The overall risk of impairment in the study was 27/60
(45%).

Analysis
When comparing groups of binary data, the parameter of interest
is the estimated risk difference. For this trial, the difference is
136 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

TABLE 13.4. 2 × 2 table summarising neurocognitive impairment in the

two groups

Impaired Not impaired Total

Off-pump 8 (26.7%) 22 30
On-pump 19 (63.3%) 11 30
Total 27 33 60

36.7% (63.3% minus 27.7%) in favor of the off-pump group with

a standard error of 3.7%. Assuming Normality, a 95% confidence
interval for the risk difference is 13.3% to 60.1%. The interval is
not close to zero, suggesting that off-pump surgery reduces the
risk of neurocognitive impairment. The Normality assumption is
considered acceptable if the observed values in the 2 × 2 table all
exceed 10; this is approximately the case here.
To test the null hypothesis that the risk difference is zero, we
use a chi-squared test. If the null hypothesis is true, we would
expect the risk of impairment to be 45% in both groups; thus we
would expect about 13.5 subjects in both groups to have neuro-
cognitive impairment (for details on expected values, see the
textbooks referenced). These expected numbers are very different
from the observed numbers, and the P-value for the chi-squared
test is 0.004, suggesting a clear difference between the two types
of surgery.
The chi-square test is only valid when the sample size is suf-
ficient. For a 2 × 2 table, all the expected values in the table
should ideally exceed 10. This is the case here (Table 13.5).

Violation of Assumptions
If the sample size is too small, we may use Fisher’s Exact Test
and exact methods to obtain confidence limits.

TABLE 13.5. 2 × 2 table summarising neurocognitive impairment in the

two groups: observed (expected) values

Impaired Not impaired Total

Off-pump 8 (45% × 30 = 13.5) 22 (16.5) 30

On-pump 19 (13.5) 11 (16.5) 30
Total 27 33 60
 13. ANALYZING HEALTH STUDIES 137

TABLE 13.6. Outcome pairs breast cancer stud

Recurrence?
(no hormone)
Recurrence?
(hormone) Yes No Total

Yes 9 6 15
No 18 14 32
Total 27 20 47

Paired Binary Data

A study to assess the effect of hormone treatment on patients
with node-positive breast cancer recruited 47 women. These
were all given hormone treatment and were matched (on age,
tumor grade, and number of nodes affected) to 47 women who
did not receive hormone treatment. The binary outcome was
tumor recurrence or death within 3 years.

Describing the Data

Since the data are paired, we arrange the data by pairs of out-
comes (Table 13.6). The risk of a recurrence after hormone treat-
ment is 31.9% (15/47), compared with 57.4% (27/47) for the no
hormone group.

Analysis
The risk difference is 25.5% in favor of the hormone treatment,
and the standard error is 9.7%. Assuming Normality, the 95%
confidence interval is 6.5% to 44.6%. To formally compare the
groups, we use McNemar’s test. This gives a P-value of 0.014,
showing evidence of a beneficial effect of hormone treatment.
This test is valid provided that both discordant pairs in the table
exceed 5. We have 6 and 18 discordant pairs.

Violation of Assumptions
If the sample size is too small, we can use exact methods for
testing and calculating confidence intervals.

CONCLUSION
We have described the most basic methods for summarizing and
analyzing two groups of continuous or binary outcomes. More
advanced methods exist for different types of data, more than
two groups, and other types of associations. Regression is used
138 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

to explore relationships between variables, or to adjust for con-

founders. Data clustered within larger units require more sophis-
ticated methods. When conducting a research study, we strongly
recommend that you consult a statistician regarding all aspects
of the study, including design, data collection, analysis, and
interpretation

References
1. Altman DG. Practical Statistics for Medical Research. Boca Raton, FL:
Chapman and Hall, 1991
2. Bland JM. An Introduction to Medical Statistics, 3rd edition. New
York: Oxford University Press, 2000
3. Kirkwood BR, Sterne JAC. Essential Medical Statistics, 2nd edition.
Oxford, UK: Blackwell Science, 2003
4. Armitage P, Berry G, Matthews JNS. Statistical Methods in Medical
Research, 4th edition, Oxford, UK: Blackwell Science, 2002
5. Ragland DR, Brand RJ. Coronary heart disease mortality in the
Western Collaborative Group study. Am J Epidemiol 1988;127:
462–475.
6. Bessant R, Duncan R, Ambler G, et al. Prevalence of conventional and
lupus-specific risk factors for cardiovascular disease in patients with
systemic lupus erythematosus: a case control study. Arthritis Rheum
2006;55(6):892–899.
7. Zamvar V, Williams D, Hall J, et al. Assessment of neurocognitive
impairment after off-pump and on-pump techniques for coronary
artery bypass graft surgery: prospective randomised controlled trial.
BMJ 2002;325:1268–1271.
Chapter 14
Future
H.R.H. Patel, I.S. Shergill, and M. Arya

Research is the future, and the future is research! We believe that

this mantra is the fundamental basis for successful basic science
and clinical research applicable to medical practice. As is very
much apparent, many characteristics of current medical practice
require medical staff to have at least some exposure to basic
science research. In the future, we anticipate many institutions
worldwide will provide specific courses on clinical research for
all health professionals, as is occurring in UK professional regu-
latory bodies. For the near future, well-established techniques
will continue to be used, refined and, combined, such as the use
of tissue microarrays, immunohistochemistry, and fluorescent in
situ hybridization. Moreover, newer techniques will be intro-
duced at the protein, RNA, and DNA levels, allowing the
researcher a wide array of tools to address important scientific
and clinical hypotheses. Notably, on the horizon is the wide-
spread introduction of nanotechnology. We believe the use of
nanotechnology in research has no bounds, if properly regulated,
and will allow rapid advancements in research as well as poten-
tially providing the scientific basis for cures of currently terminal
diseases.
Index

A Becton Dickinson FACScan,

Affymetrix, 100 40, 43–44
U133+2 Gene Chips, 102 Behavior type, relationship to
Alkaline phosphatase heart disease risk,
in colorimetric detection, 130–133
55 Bias, 10–11
in immunohistochemical Biotin, 23–24, 29, 55
detection systems, 28, Blinding, 10, 11–12
29 Blotting techniques, 48–57.
Annals of the Rheumatic See also Northern
Diseases, 120, 122 blotting; Southern
Antigens blotting; Western
definition of, 23 blotting
immunohistochemical basic principles and
localization of. See methods of, 49–56
Immunohistochemistry blocking, 54
Arthritis and Rheumatism, blotting process, 52
120, 122 detection and
Arthritis Care and Research, localization of target
120 macromolecules, 52–54
Arthritis Research Campaign, enhanced
118, 120 chemiluminescence,
Assessor/response bias, 10–11 55–56
Assumptions (statistical), 129 gel electrophoresis, 49
Autoradiography, in blotting hybridization, 52–54
technology, 54 sample preparation,
Avidin/biotin methods, in 49–50
immunohistochemistry, sodium sodecylsulfate-
23–24, 29 polyacrylamide gel
electrophoresis, 51–52
B stripping, 56
Bacterial infections, of cell washing, 54
cultures, 32 Breast cancer
Baculovirus protein effect of hormone treatment
expression systems, 89, on, 137
90 estrogen receptor
Bar charts, 128 biomarkers for, 101, 104
142 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

Breast cancer (cont.) intrinsic signal imaging in,

microarray analysis of, 101, 115
102, 104–105 optical recordings, 113
progestrogen receptor voltage-sensitive recordings,
biomarkers for, 104 115
British Heart Foundation, 118 Chi-squared test, 136
British Medical Journal, 120, Cholesterol levels, relationship
122 to behavior type,
130–133
C Chromatography,
Cardiovascular disease, immobilized metal-ion
systemic lupus affinity (IMAC), 91–92,
erythematosus as risk 94
factor for, 133–135 Clinical and Experimental
Case-control studies, 13, 15 Rheumatology, 120
Case reports, 13, 14, 16 Clinical Rehabilitation, 122
Case series, 13–14, 16 Clinical Rheumatology, 120
Cell, 122 Clinical trials
Cell culturing, 31–37 research governance
of cell lines, 31, 37 guidelines for, 3–4, 5
fundamentals of, 32–37 tissue microarray use in,
bacterial/fungal infection 103–104
control, 32 Cohort studies, 13, 14–15, 16
cell extraction, 35–37 Confidence intervals, 129
culture media, 33–34 Confidentiality, in research, 5
handling of tissue Confounding, 9–10
biopsies, 35 Control groups, 9
incubation, 35 Coronary artery bypass graft
support (“feeder”) cells, patients, neurocognitive
34–35 impairment in,
trypsinizing, 3537 135–136
of primary cells, 31, 37 Crossover trials, 12–13
Cellular data recording, Cross-sectional studies, 14,
109–116 16
calcium-sensitive optical Cytometry. See also Flow
imaging in, 113–114 cytometry
electrical recording setup definition of, 38
for, 109–110
extracellular recordings, D
110–111, 112 Data
intracellular recordings, binary, 126
111–114 paired, 137
intracellular staining of two independent groups
single cells for, 112, of, 137
113 categorical, 126, 128
 INDEX 143

clustered, 127 EnVision

continuous, 126, 127–128 immunohistochemistry
two independent groups kit, 25
of, 130–133 Epitopes, definition of, 23
two paired groups of, Escherichia coli protein
133–135 expression systems, 86,
description of, 127–128 89–91
discrete, 126 Estimated risk difference,
nominal, 126 135–136
normal distribution of, Ethical guidelines, for
127–128 research, 1–7
numerical, 126 European Community, 118
structure of, 126–127 European Union Directive on
Data analysis. See Statistical Clinical Trials Act
analysis (2001), 3
Declaration of Helsinki, European Union Directive on
3 Good Clinical Practice,
Diabetes United Kingdom, 3
118 Excitation wavelengths,
DNA (deoxyribonucleic acid ) 39–40
blotting. See Southern
blotting F
DNA (deoxyribonucleic acid) FACScan, 40, 43–44
dyes, 43 Fibroblasts, extraction from
cell cultures, 35–37
E Fisher’s Exact test, 136
Ehlers-Danos Support Group, Flow cytometers
118 components of, 41–43
Electrophysiological color assignment,
techniques, for 42–43
recording of cellular fluidic system, 41
data, 109–116 optical system and
calcium-sensitive optical analysis, 41–42
imaging, 113–114 photomultiplier tubes, 42,
electrical recording setup 43
for, 109–110 development of, 39
intracellular staining of Flow cytometry, 38–47
single cells, 112, 113 applications of, 38–39
intrinsic signal imaging, data analysis capabilities of,
115 43–46
optical recordings, 113 gating and negative
voltage-sensitive recordings, controls, 44, 46
115 histograms and dot plots,
Emission wavelengths, 43–46
39–40 definition of, 38
144 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

Flow cytometry (cont.) G

fluorescence-related Gene cloning, 48
parameters in, 38, Good Clinical Practice, 3
39–40
light scattering-related H
parameters in, 38, 41 Health studies, description
preparation of samples for, and analysis of data
46 from, 126–138
Fluorescence, 39–40 Health study design, 8–17, 12,
Fluorescence antibody 13, 14, 16
methods, 39 bias in, 10–11
Fluorescent in situ blinding in, 10, 11–12
hybridization (FISH), case reports, 13, 14, 16
58–65 case series, 13–14, 16
applications of, 58 confounding in, 9–10
basic principles of, 58–59 control groups in, 9
methodological aspects of, crossover trials, 12–13
59–60 cross-sectional studies, 14,
cytologic preparations, 59 16
DNA probes, 59, 61 essentials of, 8–11
DNA unmasking intention-to-treat analysis
processes, 60–61 in, 12
formalin-fixed paraffin- noncompliers in, 12
embedded tissue observational studies, 11,
sections, 59–60 13, 14
posthybridization wash analytic, 13, 14–16
and visualization, 62 case-control studies, 13,
preparation of samples, 15
59–60 cohort studies, 13, 14–15,
quality assurance in, 62–65 16
Fluorescin, 39 descriptive, 13
Fluorochromes placebos in, 12
emission spectrum of, 40 protocol violations in,
in flow cytometry, 40, 42 12
Food and Drug protocol writing for, 11
Administration (FDA), random allocation of
2 subjects in, 10
Formaldehyde (formalin), as randomized controlled
immunohistochemical trials, 10, 11–12, 16
preservative, 21 reliability of, 16 15
Freedom of information, in sample size in, 10
research, 5 selection of subjects in, 8
Fungal infections, of cell specification of primary
cultures, 32 outcome in, 8–9
 INDEX 145

specification of research selection of antibodies,

questions in, 8 22–23, 27
HEK293 cell protein staining methods and
expression systems, 89, detection systems, 23
90 in tissue microarray
High-throughput format. See analysis, 102, 104, 105
also Proteonomics, troubleshooting/
high-throughput (HTP) optimization in, 27–29
of tissue microarrays, 104 background staining,
Histograms, 127, 131, 132 27–28
Horseradish peroxidase endogenous biotin, 29
labelling with DAB, 26 endogenous enzyme
use in blotting technology, activity, 28–29
55 optimizing primary
Human Tissue Act (2004), 3 antibodies, 27
Hypothesis testing, 128, washing of slides, 29
129–130 In vitro, definition of, 37
In vivo, definition of, 37
I In situ hybridization. See also
Immunofluorescence, 40, 41 Fluorescent in situ
Immunoglobulin G hybridization
antibodies, net negative limitations to, 98
surface charge of, 28 Intention-to-treat analysis, 12
Immunohistochemistry, 18–30 Interquantrile range, 128
basic principles of, 18–20
definition of, 18 J
glossary of terms in, 19–20 Journal articles. See also
methodology of, 20–27 Research, presentation
antigen retrieval, 21–22 and publication of
avidin/biotin methods, assessment of, 123–124
23–24, 29 writing of, 122–123
enzyme labels and Journal of Rheumatology, 120,
chromogenes, 18, 122
26–27 Journal of the American
heat-mediated antigen Medical Association,
retrieval (HMAT), 21, 120
22 Journals. See also names of
horseradish peroxidase specific journals
labelling with DAB, 26 peer-reviewed, 5, 118,
polymer-based methods, 119–122
25–26
proteolytic digestion, L
21–22 Lancet, 120, 122
sample fixation, 18, 20–21 Levamisole, 29
146 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

LightCycler instrument, 74, N

78 National Health Service
Lupus, 120 employees, research
funding for, 118
M National Institutes of Health,
Mann-Whitney U (Wilcoxon Protein Structure
Rank Sum) test, 133 Initiative (PSI) of,
McNemar’s test, 137 86–87
Mean (average), 127 Nature, 122
Median, 128 Nazis, 3
Medical Research Council, New England Journal of
117–118 Medicine, 120, 122
Medicines for Human Use Northern blotting
(Clinical Trials) applications of, 48, 56
Regulations (UK), 3 basic principles and
Mental Capacity Act (2005), methods of
3 agarose electrophoresis,
Microarrays, 98–108 50–51
comparison with blocking, 54
quantitative reverse hybridization, 53–54
polymerase chain sample preparation, 49
reaction, 80 definition of, 48, 49
DNA, 98–102, 106 Novolink Polymer kit, 25
cDNA (spotted) arrays, Nuremberg Code, 3
98–99
clinical applications of, O
101–102 Observational studies, 11, 13,
data analysis techniques 14
in, 100–101 analytic, 13, 14–16
oligonucelotide arrays, case-control studies, 13,
98, 99–100 15
Perfect Match/Mismatch cohort studies, 13, 14–15,
(PM/MM) probe 16
strategy in, 100 descriptive, 13
sample preparation in,
101 P
tissue, 99, 102–105, 106 Paclitaxel, 102
advantages and criticisms Pancreatic carcinoma, tissue
of, 103–104 microarray analysis of,
applications of, 104–105 105
construction of, 99, Parametric analysis,
102–103 132–133
quantitative analysis of, Peer review, of research, 5,
105 118, 119–122
 INDEX 147

Peroxidase, endogenous, in protein expression

immunohistochemical systems, 89–90
analysis, 28–29 protein purification,
Pichia pastoris protein 91–92
expression systems, 89 Protocol violations, 12
Placebos, 12 Protocol writing, 11
POET (Pooled ORF Publication, of research. See
Expression Research data,
Technology), 93, 94 presentation and
Polymerase chain reaction publication of
(PCR), real-time. See P-value, 130
also Quantitative for chi-squared test, 136
reverse transcriptase overinterpretation of, 130
polymerase chain in paired t-test, 135
reaction (Q-RT-PCR) in t-test, 133
applications of, 66
Polymer-based methods, in Q
immunohistochemistry, Quantitative-reverse
25–26 transcriptase
Primer Express, 76 polymerase chain
Prostate carcinoma, tissue reaction (Q-RT-PCR)
microarray analysis of, amplicon detection
105 strategies in, 72–76
Protein blotting. See Western double-stranded (ds)
blotting DNA-intercalating
Protein Data Bank (PDB), 86 agents, 72, 73–76
Proteins. See also dual hybridization
Proteonomics, probes, 74–75
high-throughput (HTP) fluorescent probes, 72–73
hydrophobicity of, 27–28 hydrolysis probes, 74
Protein Structure Initiative molecular beacons, 75, 79
(PSI), 86–87 “scorpions,” 75–76
Proteonomics, high- SYBR Green I, 72–73, 76,
throughput (HTP), 78, 79
86–97 TaqMan minor groove-
additional approaches in, binding probes, 74
92–93 comparison with
applications of conventional PCR, 79–80
cloning, 87–89 microarray technologies,
protein expression and 80
solubility screening, controls (“housekeeping”
91 genes) in, 72
protein expression definition of, 67
conditions, 90–91 equipment for, 77–79
148 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

Quantitative-reverse future trends in, 124–125

transcriptase outlets for, 118, 119–122
polymerase chain in peer-reviewed journals,
reaction (Q-RT-PCR) 118, 119–122
(cont.) Research governance, 1–7
ABI Prism 7900 HT, 78 in the United Kingdom, 1–7
iCycler iQ, 78 in the United States, 2
LightCylcer, 74, 78 Research Governance
Mx4000®, 78–79 Framework for Health
Rotor GeneTM, 79 and Social Care, 3
Smart Cycler System, 79 Research questions,
fluorescence emission data specification of, 8
from, 68–70 Rheumatology, 120, 122
fluorescence resonance RNA (ribonucleic acid)
energy transfer (FRET) blotting. See Northern
principle of, 67, 68, 74, blotting
75
fundamentals of, 67–70 S
limitations to, 98 Saccharomyces cerevisiae
multiplex, 77 protein expression
primer, probe, and systems, 89
amplicon design for, Sample size, 10
76–77 Scandinavian Journal of
quantification methods in, Rheumatology, 120
70–71 Science, 122
TaqMan probes/system for, Scleroderma Association, 118
67–68, 74, 76–77, 79 Selection bias, 10
threshold cycle (Ct) value Southern, Edward, 48
of, 69, 70, 71 Southern blotting
applications of, 48, 56
R basic principles and
Randomized controlled trials, methods of, 49–56
10, 11–12, 16 agarose electrophoresis,
Recall bias, 11 50–51
Recombinant DNA blocking, 54
technologies, 48 hybridization, 53–54, 54
Regression analysis, 137–138 sample preparation, 49
Representative samples, 128 definition of, 48
Research, funding for, Standard deviation, 127
117–118 Standard errors, 129
Research Assessment Stark, George, 48
Exercise, 117 Statistical analysis, 128–138
Research data, presentation assumptions, 129
and publication of, of binary data
117–125 of paired data, 137
 INDEX 149

of two independent violations of assumptions,

groups of data, 133, 135, 136
135–136 Welch’s test, 133
chi-squared test, 136 Wilcoxon signed rank test,
confidence intervals, 129, 135
133 Statistical significance,
of continuous data 130
of two independent Streptavidin-biotin complexes,
groups, 130–133 23–24
of two paired groups, Systemic lupus
133–135 erythematosus. as
descriptive, 127 risk factor for
of DNA microarray gene cardiovascular disease,
expression results, 133–135
100–101
estimated risk difference, T
135–136 TaqMan system, 67–68, 74,
estimation, 128 76–77, 79
examples of, 130–138 Taxanes, 102
Fisher’s Exact test, 136 t-distribution, 132–133
histograms, 127, 131, 132 3T3 cells, 34–35
hypothesis testing, 128, Towbin, Harry, 48
129–130 Trastuzumab, 102
Mann-Whitney U (Wilcoxon t-test
Rank Sum) test, 133 paired, 135
McNemar’s test, 137 two-sample, 132–133
nonparametric analysis, 133 Tumors. See also specific types
normal distribution, of cancer
132–133 microarray-based molecular
normal plots, 127, 131–133, phenotyping of, 101
134 with DNA microarrays,
P-value, 130, 135, 136 102–102
as research publication with tissue microarrays,
component, 123 102–105, 106
sampling distributions, 2x2 tables, 135, 136
128–129
standard errors, 129 U
statistical significance in, United Kingdom, research
130 governance in, 1–7
statistical uncertainty, 128, Department of Health and,
129 1, 6
t-distribution, 132–133 National Health Service
t-test and, 1, 2, 5–6
paired, 135 United States Food and Drug
two-sample, 132–133 Administration (FDA), 2
150 BASIC SCIENCE TECHNIQUES IN CLINICAL PRACTICE

V agarose electrophoresis,
Variance, 127 51–52
Vinorelbine, 102 blocking, 54
hybridization, 54
W sample preparation,
Welch’s test, 133 49–50
Wellcome Trust, 117–118 definition of, 48, 49
Western blotting Wilcoxon Rank Sum test,
applications of, 48, 56 133
basic principles and Wilcoxon signed rank test,
methods of 135