Methods
in
Molecular Biology™
Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfield, Hertfordshire, AL10 9AB, UK
For other titles published in this series, go to www.springer.com/series/7651
wwwwwwwwwwwwwww
Formalin-Fixed Paraffin-Embedded Tissues Methods and Protocols Edited by
Fahd Al-Mulla Molecular Pathology Unit, Health Sciences Center, Kuwait University, Safat, Kuwait
Editor Fahd Al-Mulla Molecular Pathology Unit Department of Pathology Faculty of Medicine Kuwait University Safat, Kuwait
[email protected]
ISSN 1064-3745 e-ISSN 1940-6029 ISBN 978-1-61779-054-6 e-ISBN 978-1-61779-055-3 DOI 10.1007/978-1-61779-055-3 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2011922256 © Springer Science+Business Media, LLC 2011 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the advice and information in this book are believed to be true and accurate at the date of going to press, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Printed on acid-free paper Humana Press is part of Springer Science+Business Media (www.springer.com)
Dedication I dedicate this book to two women in my life; my mother who brought me to this world, nourished and taught me, and my wife who maintained such nourishment and ethics.
wwwwwwwwwwwwwww
Preface The wide availability of formalin-fixed, paraffin-embedded tissues (FFPET) makes them an attractive source of material to study and classify diseases at molecular level. It is now inconceivable to offer a diagnostic report without the application of modern techniques to FFPET. The application of state-of-the-art molecular techniques has revolutionized not only diagnostic skills of pathologists, but also allowed them to be more active members in tailored disease treatment and prognostication than ever before. However, working with FFPET remains a great challenge, especially when the aim is to decipher diseases at mole cular level. The purpose of Formalin-Fixed, Paraffin-Embedded Tissues: Methods and Protocols is to provide an up-to-date methodological information pertaining to the utilization of genomic, transcriptomic, and proteomic data in diagnosis, prognosis, and tailored therapy. Many molecular-flavored protocols dealing with FFPET exist, and some offer conflicting advice. This book brings forward to scientists and clinicians working with FFPET, wellestablished and tested protocols focused on genomics, epigenetics, proteomics, and cellular biology. The book starts with a chapter on Ethics. This, I believe is appropriate given the importance of the topic to any study and the lack of single regulatory or bioethical standard that covers research with FFPET, which introduces another complexity in the design and execution of studies requiring such specimens. There is scanty information in the scientific literature that covers this subject, and this chapter, I hope, fills this niche. The remaining chapters are closely interconnected, yet at the same time, they cluster into well-organized themes. Chapter 2 deals with the construction and uses of tissue arrays, a technique that brought not only much needed power in the number of patients that can be analyzed at once, but also introduced substantial cost-savings to laboratories. This theme that deals with cellular structures and content continues in Chapters 3–6, which describe detailed protocols in immunohistochemistry, immunofluorescence, fluorescent, and chromogenic in situ hybridization. The next theme, introduces wellestablished protocols for FFPET microdissection and nucleic acids extraction for their utilization in advanced techniques such as microarray CGH, DNA methylation, and pyrosequencing. Chapters 14–18 describe methods that unlock expression-related information stored as RNA, miRNA, and proteins in FFPET. The book ends with a thought provoking chapter, which describes a novel tissue fixative. The decision to use such a fixative, of course, will be left to you. The book is aimed at the pathologist, molecular pathologist, geneticist, and the clinician who, today more than ever, is required to understand how technology is impacting health care. Also, the book is aimed at the more experienced molecular biologist who wishes to apply sophisticated techniques to FFPET in order to decipher disease-associated molecules. Students working with FFPET may find this book a valuable source of practical and theoretical information that can save them both time and effort throughout their
vii
viii
Preface
journey of optimization because the protocols came directly from people who invented them and use them on a daily bases for patients’ care. As always, an Editor has a lot of people to thank. I would like to thank the authors, who contributed to this work. I thank them for their enthusiasm, effort, and patience. I am indeed very grateful for all the staff at the pathology Department at Kuwait University for their support during the preparation of this book. They had to tolerate my grumpiness and few days! of absence to complete this book. To them I apologize. Also, I am very grateful for Milad Bitar for his support during the write-up of the project. Of course, the price of success in something is a failure in another. A saying that my family repeatedly hears from me! I thank them all, my mother, wife, sons, and siblings for tolerating my long hours at work.
Safat, Kuwait
Fahd Al-Mulla
Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
vii xi
1 Regulatory and Ethical Issues on the Utilization of FFPE Tissues in Research . . . Catherine M. With, David L. Evers, and Jeffrey T. Mason 2 Tissue Microarrays: Construction and Uses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Carol B. Fowler, Yan-Gao Man, Shimin Zhang, Timothy J. O’Leary, Jeffrey T. Mason, and Robert E. Cunningham 3 Standardization in Immunohistology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anthony S.-Y. Leong and Trishe Y.-M. Leong 4 Multiple Immunofluorescence Labeling of Formalin-Fixed Paraffin-Embedded Tissue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . David Robertson and Clare M. Isacke 5 Microwaves for Chromogenic In Situ Hybridization . . . . . . . . . . . . . . . . . . . . . . Anthony S.-Y. Leong and Zenobia Haffajee 6 Automated Analysis of FISH-Stained HER2/neu Samples with Metafer . . . . . . . Christian Schunck and Eiman Mohammad 7 Laser Capture Microdissection of FFPE Tissue Sections Bridging the Gap Between Microscopy and Molecular Analysis . . . . . . . . . . . . . . . . . . . . . . Renate Burgemeister 8 Nucleic Acids Extraction from Laser Microdissected FFPE Tissue Sections . . . . . . Renate Burgemeister 9 Microarray-Based CGH and Copy Number Analysis of FFPE Samples . . . . . . . . . Fahd Al-Mulla 10 Microarray Profiling of DNA Extracted from FFPE Tissues Using SNP 6.0 Affymetrix Platform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Marianne Tuefferd, An de Bondt, Ilse Van den Wyngaert, Willem Talloen, and Hinrich Göhlmann 11 Whole Genome Amplification of DNA Extracted from FFPE Tissues . . . . . . . . . . Mira Bosso and Fahd Al-Mulla 12 Pyrosequencing of DNA Extracted from Formalin-Fixed Paraffin-Embedded Tissue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Brendan Doyle, Ciarán O’Riain, and Kim Appleton 13 Analysis of DNA Methylation in FFPE Tissues Using the MethyLight Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ashraf Dallol, Waleed Al-Ali, Amina Al-Shaibani, and Fahd Al-Mulla
1
ix
23
37
69 79 91
105 117 131
147
161
181
191
x
Contents
14 RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues Prepared Using a Range of Different Fixatives and Conditions . . . . . . . . Mei-Lan Liu, Jennie Jeong, Ranjana Ambannavar, Carl Millward, Frederick Baehner, Chithra Sangli, Debjani Dutta, Mylan Pho, Anhthu Nguyen, and Maureen T. Cronin 15 RT-PCR-Based Gene Expression Profiling for Cancer Biomarker Discovery from Fixed, Paraffin-Embedded Tissues . . . . . . . . . . . . . . . . . . . . . . . Aaron Scott, Ranjana Ambannavar, Jennie Jeong, Mei-Lan Liu, and Maureen T. Cronin 16 MicroRNA Isolation from Formalin-Fixed, Paraffin-Embedded Tissues . . . . . . . . Aihua Liu and Xiaowei Xu 17 Gene Expression Profiling of RNA Extracted from FFPE Tissues: NuGEN Technologies’ Whole-Transcriptome Amplification System . . . . . . . . . . . Leah Turner, Joe Don Heath, and Nurith Kurn 18 Protein Mass Spectrometry Applications on FFPE Tissue Sections . . . . . . . . . . . . Carol B. Fowler, Timothy J. O’Leary, and Jeffrey T. Mason 19 An Alternative Fixative to Formalin Fixation for Molecular Applications: The RCL2®-CS100 Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Amélie Denouël, Florence Boissière-Michot, Philippe Rochaix, Frédéric Bibeau, and Nathalie Boulle
205
239
259
269 281
297
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Contributors Waleed Al-Ali • Molecular Pathology Unit, Health Sciences Center, Safat, Kuwait Fahd Al-Mulla • Molecular Pathology Unit, Health Sciences Center, Safat, Kuwait Amina Al-Shaibani • Molecular Pathology Unit, Health Sciences Center, Safat, Kuwait Ranjana Ambannavar • Genomic Health, Inc., Redwood City, CA, USA Kim Appleton • Centre for Oncology and Applied Pharmacology, University of Glasgow, Glasgow, UK Frederick Baehner • Genomic Health, Inc., Redwood City, CA, USA Frédéric Bibeau • CRLC Val d’Aurelle, Montpellier, France Florence Boissière-Michot • CRLC Val d’Aurelle, Montpellier, France Mira Bosso • Faculty of Medicine, Department of Pathology, Kuwait University, Safat, Kuwait Nathalie Boulle • Hospital Arnaud de Villeneuve, Montpellier, France Renate Burgemeister • Carl Zeiss MicroImaging, München, Germany Maureen T. Cronin • Genomic Health, Inc., Redwood City, CA, USA Robert E. Cunningham • Armed Forces Institute of Pathology, Washington, DC, USA Ashraf Dallol • King Fahad Medical Research Center, King Abdulaziz University, Jeddah, Saudi Arabia An de Bondt • Johnson & Johnson Pharmaceutical Research & Development, Beerse, Belgium Amélie Denouël • CRLC Val d’Aurelle, Montpellier, France Brendan Doyle • Beatson Institute for Cancer Research, Glasgow, UK Debjani Dutta • Genomic Health, Inc., Redwood City, CA, USA David L. Evers • Department of Biophysics, Armed Forces Institute of Pathology, Washington DC, USA Carol B. Fowler • Armed Forces Institute of Pathology, Washington, DC, USA Hinrich Göhlmann • Johnson & Johnson Pharmaceutical Research & Development, Beerse, Belgium Zenobia Haffajee • Immunohistology Unit, Hunter Area Pathology Service, Newcastle, NSW, Australia Joe Don Heath • NuGEN Technologies, Inc., San Carlos, CA, USA Clare M. Isacke • Breakthrough Centre, The Institute of Cancer Research, Royal Cancer Hospital, London, UK Jennie Jeong • Genomic Health, Inc., Redwood City, CA, USA Nurith Kurn • NuGEN Technologies, Inc., San Carlos, CA, USA Anthony S.-Y. Leong • University of Newcastle, Newcastle, NSW, Australia and Peking University, Beijing, China Trishe Y.-M. Leong • Victorian Cytology Service, Melbourne, VIC, Australia Aihua Liu • University of Pennsylvania School of Medicine, Philadelphia, PA, USA
xi
xii
Contributors
Mei-Lan Liu • Genomic Health, Inc., Redwood City, CA, USA Yan-Gao Man • Armed Forces Institute of Pathology, Washington, DC, USA Jeffrey T. Mason • Armed Forces Institute of Pathology, Washington, DC, USA Carl Millward • Genomic Health, Inc., Redwood City, CA, USA Eiman Mohammad • Faculty of Medicine, Molecular Pathology Unit, Kuwait University, Safat, Kuwait Anhthu Nguyen • Genomic Health, Inc., Redwood City, CA, USA Timothy J. O’Leary • Veterans Health Administration, Washington, DC, USA Ciarán ÓRiain • John Vane Science Centre, Charterhouse Square, London, UK Mylan Pho • Genomic Health, Inc., Redwood City, CA, USA David Robertson • Breakthrough Centre, The Institute of Cancer Research, Royal Cancer Hospital, London, UK Philippe Rochaix • CRLC Claudius Regaud, Toulouse, France Chithra Sangli • Genomic Health, Inc., Redwood City, CA, USA Christian Schunck • MetaSystems, Altlussheim, Germany Aaron Scott • Genomic Health, Inc., Redwood City, CA, USA Willem Talloen • Johnson & Johnson Pharmaceutical Research & Development, Beerse, Belgium Marianne Tuefferd • Johnson & Johnson Pharmaceutical Research & Development, Beerse, Belgium Leah Turner • NuGEN Technologies, Inc., San Carlos, CA, USA Ilse Van den Wyngaert • Johnson & Johnson Pharmaceutical Research & Development, Beerse, Belgium Catherine M. With • Legal Counsel, Armed Forces Institute of Pathology, Washington, DC, USA Xiaowei Xu • University of Pennsylvania School of Medicine, Philadelphia, PA, USA Shimin Zhang • Armed Forces Institute of Pathology, Washington, DC, USA
wwwwwwwwwwwwwww
Chapter 1 Regulatory and Ethical Issues on the Utilization of FFPE Tissues in Research Catherine M. With, David L. Evers, and Jeffrey T. Mason Abstract Formalin-fixed, paraffin-embedded (FFPE) archival tissues and their associated diagnostic records represent an invaluable source of information on diseases where the patient outcomes are already known. Older archives contain many unique FFPE tissue specimens that would be impossible to replicate today due to changes in medical practice and technology. Unfortunately, there is no single regulatory or bioethical standard that covers research with FFPE tissue specimens. This makes it difficult for researchers to prepare protocols involving FFPE tissues and equally difficult for Institutional Review Boards to evaluate them. In this review, focused on US regulatory policy, the application of the Common Rule and the Privacy Rule of the Health Insurance Portability and Accountability Act to research involving FFPE tissue specimens will be discussed. It will be shown that the difficulty in applying regulatory and ethical standards to FFPE tissues results not from the tissues themselves, but from the personally identifiable health information associated with the tissue specimens. Key words: Regulatory bioethics, formalin-fixed paraffin-embedded, FFPE, Biorepository, Common Rule, OHRP, IRB, HIPAA, Privacy rule, PHI, Protected health information, Personally identifiable information, De-identification, Anonymized specimen, coded specimen, Informed consent, Surgical consent, Privacy rule authorization, Usage agreement, Honest broker, Human subjects research, Deceased persons
1. Introduction The modern principles of human subjects research protection – the requirement for voluntary and informed consent, a favorable benefit to risk assessment, and the right to withdraw from research – were first articulated in 1948 in the Nuremberg Code as part of the proceedings of the Nuremberg Trials (1). In 1964, the World Medical Association augmented the Nuremberg Code by adding two important concepts that were incorporated into the
Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_1, © Springer Science+Business Media, LLC 2011
1
2
With, Evers, and Mason
Declaration of Helsinki (2). These concepts were that the interests of research subjects supersede those of society and that every clinical research subject should receive the best treatment available. In the same year, in the United States (US), the National Institutes of Health (NIH) published the first policy requiring an ethics committee review of research, in this case research funded by the Public Health Service. In 1974, the US Congress passed the National Research Act (3), which established two modern pillars of human subjects research oversight. The first was the Institutional Review Board (IRB) whose function is to evaluate and approve most kinds of research involving human subjects. The second was the National Commission for Protection of Human Subjects of Biomedical and Behavioral Research. This commission was influential in establishing an ethical framework for the protection of vulnerable populations. In 1978, this commission issued the Belmont Report (4) that articulated three main principles of bioethics upon which IRB review of human subjects research is based (5). The first principle is respect for persons, which established the tenants of voluntary participation, informed consent, and the protection of privacy. The second principle is beneficence, which established the tenants that studies should be designed to minimize risk, that the risks of research must be justified by the potential benefits, and that conflicts of interest are managed equitably. The third principle is justice, which established the tenants of protecting vulnerable subjects and populations, and insuring that the people likely to benefit from research were not systematically excluded. Modern oversight of human subjects research in the US is governed by three primary federal regulations; however, these regulations clearly do not answer all the ethical questions posed by the topic. The first regulatory guidance is the Federal Policy for the Protection of Human Subjects. This regulation was codified by the Department of Health and Human Services (DHHS) in 1981 as Title 45, Part 46 of the Code of Federal Regulations (45 CFR 46). In 1991, 17 federal agencies adopted Title 45 Part 46, which subsequently became known as the “Common Rule,” although it is codified as a different CFR within each agency. The Federal Office of Human Research Protection (OHRP) is responsible for interpreting and providing guidance on DHHS codification of the Common Rule for conducting human subjects research funded by DHHS. The Food and Drug Administration (FDA) did not adopt the Common Rule and has its own regulations covering human subjects research. Therefore, the second set of regulatory guidance is found in the FDA Protection of Human Subjects (Title 21 CFR Parts 50, 56, and 812). The third source of regulatory guidance is found in the Privacy Rule of the Health Insurance Portability and Accountability Act (HIPAA) (Title 45 CFR Part 160, and Subparts A and E of Part 164). In addition to
Regulatory and Ethical Issues on the Utilization of FFPE Tissues in Research
3
these federal regulations, there are also state regulations governing privacy issues for research using medical records, but these vary widely among the jurisdictions (6–8). This chapter will focus on the regulatory measures and the ethical issues associated with the use of formalin-fixed, paraffinembedded (FFPE) tissues in research. FFPE tissues are considered biospecimens for regulatory purposes and, as such, the regulatory and ethical issues deal primarily with personally identifiable information associated with the biospecimens rather than the biospecimens themselves. Unfortunately, there is no unified regulatory framework that covers the spectrum of research activities associated with FFPE tissues. The three federal regulations described above differ in regulatory scope, and use varying and sometimes contradictory terminology when describing personally identifiable information, informed consent, and regulatory exemp tions. Even the Common Rule allows individual agencies to develop their own interpretation of certain terms and regulations. Guidelines available through the OHRP are generally considered the “gold standard” for human subject research oversight (http:// www.hhs.gov/ohrp/). Discussion will focus mainly on the application of the Common Rule and the HIPAA Privacy Rule to research using FFPE tissues and their associated personally identifiable information. Sources for Website information on tissue repository policies, additional regulations, privacy policies, and international ethical guidelines are listed in Table 1.
2. Biorepositories and Identifiability of Tissue
Biorepositories (archives) focus on the collection of FFPE tissue biospecimens and fresh frozen biospecimens from a broad range of diseases, including matching nondiseased adjacent tissues and normal tissue controls (see Note 1). Collections typically include biospecimens from an ethnically diverse population of men and women of all ages to ensure demographic representation of the population. To enhance the usefulness of the biospecimens, clinical data, such as the individual’s diagnosis, drug regimen, outcomes, age, race, and gender, are also collected. Privacy issues arise when the personally identifiable information associated with the FFPE tissue biospecimens render the donor identifiable. Fresh frozen biospecimens are collected from donors for the express purpose of research and are generally broadly consented. In contrast, FFPE tissues from biorepositories are more likely to have been collected for diagnostic or surgical purposes and lack a specific consent for research. More often a surgical consent was obtained with, at most, a vague reference to research. In the case of older FFPE tissues, there may be no consent at all because the
4
With, Evers, and Mason
Table 1 Web sites providing information on various aspects of regulation and ethical issues regarding biospecimens Topic
Organization
Website URL
Basic regulations and organizations Common rule
DHHS/OHRP http://www.hhs.gov/ohrp/humansubjects/ guidance/45cfr46.html
FDA regulations
FDA
HIPAA regulations
DHHS/OHRP http://www.hhs.gov/ocr/hipaa
OHRP
DHHS/OHRP http://www.hhs.gov/ohrp
OHRP guidance by topic
DHHS/OHRP http://www.hhs.gov/ohrp/policy/index.html#topics
Expedited review criteria
DHHS/OHRP http://www.hhs.gov/ohrp/policy/exprev.html
Investigational device exemption
FDA
http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/ cfcfr/CFRsearch.cfm?CFRPart=812
Office of research integrity
DHHS
http://ori.dhhs.gov
NIH bioethics resource
DHHS/NIH
http://www.nih.gov/sigs/bioethics and http://bioethics.od.nih.gov/humantissue.html
Informed consent, practicability criteria
CIHR
http://dev.cihr.ca/e/documents/et_pbp_nov05_ sept2005_e.pdf
Ethics resource
VA
http://www.ethics.va.gov/resources/siteindex.asp
HIPAA privacy rule and research
DHHS/NIH
http://privacyruleandresearch.nih.gov
IRBs and the HIPAA privacy rule
DHHS/NIH
http://privacyruleandresearch.nih.gov/ irbandprivacyrule.asp
Coded privacy information
DHHS/OHRP http://www.hhs.gov/ohrp/policy/cdebiol.pdf
Repositories and the HIPAA privacy rule
DHHS/NIH
http://privacyruleandresearch.nih.gov/ research_repositories.asp
US genome research
NHGRI
http://www.ornl.gov/sci/techresources/ Human_Genome/elsi/elsi.shtml
http://www.fda.gov/ScienceResearch/SpecialTopics/ RunningClinicalTrials/ucm155713.htm http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/ cfcfr/CFRsearch.cfm?CFRPart=50 http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/ cfcfr/CFRsearch.cfm?CFRPart=56 http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/ cfcfr/CFRsearch.cfm?CFRPart=812
Privacy policy
(continued)
Regulatory and Ethical Issues on the Utilization of FFPE Tissues in Research
5
Table 1 (continued) Topic
Organization
Website URL
Repositories and general information OHRP, research using stored tissues and data
DHHS/OHRP http://www.hhs.gov/ohrp/policy/reposit.html
NCI best practices for biospecimen resources
DHHS/NCI
http://biospecimens.cancer.gov/practices/ http://biospecimens.cancer.gov/global/pdfs/ NCI_Best_Practices_060507.pdf
Biorepositories
RAND
http://www.rand.org/pubs/monographs/2004/ RAND_MG120.pdf
Tissue/specimen banking
PRIM&R
http://www.primr.org/ResourceCenter.aspx?id=268
Best practices for repositories
ISBER
http://www.isber.org/Pubs/BestPractices2008.pdf
Pathology laboratory guidelines
CAP
http://www.cap.org
Bioethics
NBAC
http://bioethics.georgetown.edu/nbac
Clinical practice guidelines
ICH
http://www.ich.org/fileadmin/Public_Web_Site/ ICH_Products/Guidelines/Efficacy/E2A/Step4/ E2A_Guideline.pdf
International ethical guidelines
CIOMS
http://www.cioms.ch/index.html
Research policy and cooperation
WHO
http://www.who.int/rpc/en/
European biospecimens protections
COE/COM
https://wcd.coe.int/ViewDoc.jsp?id=977859
International human subjects research protection compilation
DHHS/OHRP http://www.hhs.gov/ohrp/international/ intlcompilation/hspcompilation-v20101130.pdf
International guidelines
European Commission
http://ec.europa.eu
UK
http://www.mrc.ac.uk/Ourresearch/ Ethicsresearchguidance/Useofhumantissue/index.htm
Australia
http://www.monash.edu.au/researchoffice/ethics. php/human/ethics.php
Research laws: USA vs. Europe
http://pharmalicensing.com/public/articles/ view/1064164853_3f6dddf5630a1
DHHS Department of Health and Human Services, OHRP Office of Human Research Protection, FDA Food and Drug Administration, HIPAA Health Insurance Portability and Accountability Act, NIH National Institutes of Health, CIHR Canadian Institute of Health Research, VA Veterans Administration, IRB Institutional Review Board, HGO Human Genome Organization, NCI National Cancer Institute, RAND Research and Development Corporation, PRIM&R Public Responsibility in Medicine and Research, ISBER International Society for Biological and Environmental Repositories, CAP College of American Pathologists, NBAC National Bioethics Advisory Commission, ICH International Committee on Harmonization, CIOMS Council for International Organizations of Medical Sciences, WHO World Health Organi zation, COE/COM Council of Europe/Committee of Ministries, NHGRI National Human Genome Research Institute
6
With, Evers, and Mason
specimen was obtained prior to the current regulatory guidance. Consequently, FFPE tissues from human tissue repositories typically lack documentation of informed consent for research activities and lack permission to collect prospective patient medical data. In the US, biorepositories should have established policies for access to FFPE tissues and their associated clinical data that are consistent with federal regulations, ethical principles, state regulations (where applicable), and the nature of the informed consent (if any) associated with the FFPE tissues. Virtually all repositories will require local IRB approval (or documentation of IRB exemption) of the protocols associated with requests for repository biospecimens. The researcher will need to sign a “usage agreement” that establishes provisions for protecting personally identifiable information, and for the use, disposition, and security of the biospecimens and their associated data. The usage agreement may also specify terms for the publication of study results and the proprietary rights associated with the biospecimen. The usage agreement may also be called a material transfer agreement (MTA) or a data usage agreement (9). Some repositories will also evaluate the scientific merit of the proposed study, the qualifications of the researcher, the level of institutional support, and, in some cases, that an appropriate level of funding is available to support the research. The identifiability of personal information associated with FFPE tissues falls into one of four categories as defined by the National Bioethics Advisory Commission (10). The first category is unidentified biospecimens, which are biospecimens that lack associated personally identifiable information that can be retrieved by the repository. The second category is identified biospecimens, which are biospecimens linked to personally identifiable information in such a way that the donors could be identified by name, patient number, or clear family relationship by the repository. The third and fourth categories refer to the way in which identified biospecimens are provided to researchers by the repository. The first of these is unlinked biospecimens (more commonly known as anonymized biospecimens), which are biospecimens that have been stripped of any personally identifiable information that would allow them to be traced back to the original donors. The second is coded biospecimens, which are identified biospecimens where all patient identifiers are removed and replaced by a code (11) prior to being provided to the researchers. Although the researchers do not have access to the “code key,” the person or organization holding the code key can link the biospecimens back to the donors. The code key holder can be the repository itself or a third party known as an “honest broker.” An honest broker is a neutral third party who collects and collates personally identifiable information associated with repository biospecimens, replaces the information identifiers
Regulatory and Ethical Issues on the Utilization of FFPE Tissues in Research
7
with a code, and releases the coded information to the recipient while retaining the private information and the code key. An honest broker can act as an intermediary between the repository and the researcher.
3. Informed Consent Informed consent is a process designed to educate potential human subjects as to the nature of a research project, including its procedures, duration, benefits, and potential risks, in language they can understand. If the research includes genetic sequencing, the potential risks to the donor should also be addressed. This process enables individuals to make an informed decision to parti cipate in the research study. An affirmative decision is documented by an IRB-approved and signed informed consent document. FFPE tissues obtained from biorepositories will generally require informed consent if accompanying personally identifiable information is requested (identified biospecimens) or if the researcher wishes to contact the donors for additional personally identifiable information and/or desires access to their medical records as part of a longitudinal study. Informed consent is generally not required for unidentified, anonymized, or coded FFPE biospecimens when no additional personally identifiable information is, or will be, requested (see Subheading 4). When tissues are obtained from biopsies or surgical procedures, a general surgical consent form is typically obtained indicating that the biospecimens may be used for research. However, this is generally not regarded as informed consent. The informed consent process should occur separately from surgical consent or, at the very least, should be a separate section of the surgical consent form that requires a separate signature. Informed consent is typically obtained for a specific research study. However, biorepositories generally obtain broader informed consent specifying that the donated tissue can be used for unspecified future research studies. This practice is consistent with the Common Rule as interpreted by the OHRP. A tiered informed consent process may be used to allow the donor maximum latitude in specifying how their donated tissues can be used for research. At a minimum, a tiered consent form should allow donors to chose the type of biospecimen(s) they wish to donate, the type of research the biospecimens can be used for (specific or global), and whether their personally identifiable information can be accessed. For a more complete listing of informed consent options, see Note 2. The informed consent document must also clearly communicate the donor’s option to withdraw from the study. If the donors were to exercise this option, the data collected up to the time the subject withdraws are generally kept to maintain
8
With, Evers, and Mason
study integrity. With regard to the remainder of the study the researcher could take the following actions: (1) stop using the donor’s individually identifiable biospecimens and personally identifiable information in the study, (2) anonymize the donor’s tissue biospecimen by removing all identifiable information and eliminating the donor’s personally identifiable information, or (3) destroy the donor’s biospecimens and all associated personally identifiable information. The Common Rule states that human subjects research is presumed to require informed consent, but that this requirement can be waived or altered, with IRB approval, if all four of the following conditions set forth in 45 CFR Part 46.116(d) are met. These requirements are (1) the research involves no more than minimal risk to participants, (2) the waiver or alteration of consent will not adversely affect the rights and welfare of participants, (3) the research could not practicably be carried out without the waiver or alteration, and (4) whenever appropriate, participants will be provided with additional pertinent information after participation. Several aspects of this regulation require further comment. Minimal risk is interpreted to mean that the probability and magnitude of harm or discomfort associated with the research is not greater than that ordinarily encountered in daily life or during the performance of routine physical or psychological examinations or tests (45 CFR Parts 46.102(i), also see Note 3). Another ambiguous concept in the waiver requirements is practicability, which is not defined in the Common Rule or by OHRP. Guidelines for determining when obtaining informed consent for the use of existing biospecimens is impractical has been put forth by the Canadian Institutes of Health Research (see Note 4). Certain research subjects belong to protected populations and are not eligible for a waiver of consent. These groups include (1) minor children less than 18 years of age, (2) prisoners, (3) pregnant women, (4) fetuses and products of labor and delivery, (5) people with diminished capacity to give consent, including socially vulnerable populations, and (6) mentally or physically challenged individuals.
4. The Common Rule and the IRB The IRB system was established in 1974 as part of the National Research Act and is codified in the Common Rule (45 CFR Part 46). The function of the IRB is to serve as a human subjects oversight committee to ensure that the rights and welfare of research subjects are protected. This responsibility includes approval of informed consent documents and may include assurance that all HIPAA requirements are met, although the latter is not an IRB
Regulatory and Ethical Issues on the Utilization of FFPE Tissues in Research
9
requirement (see Note 5). Technically, only research supported by federal funding is subject to the requirement of IRB review. However, most institutions with an OHRP-Approved FederalWide Assurance agreement apply the Common Rule and IRB oversight to all human subjects research supported by their institution, regardless of the funding source. The IRB is only concerned with human subjects research, thus prior to review of a protocol the IRB will determine (1) if the research involves human subjects, and, if so, (2) if the research is exempt from the Common Rule (see Note 6). The IRB is only responsible for review of research protocols if question 1 is answered in the affirmative and question 2 is answered in the negative. In order to make this decision, the meaning of several key terms, as defined under the Common Rule, must be understood. Under the Common Rule, research is defined as a systematic investigation, including research development, testing, and evaluation, designed to develop or contribute to generalizable knowledge. Thus, to be considered research, the activity must be a systematic investigation and the primary goal must be to develop or contribute to generalizable knowledge. Unfortunately, the Common Rule does not provide a specific definition of generalizable knowledge, and the OHRP has left it up to individual agencies to develop their own standards. The Belmont Report (4) refers to generalizable knowledge as that which is expressed in theories, principles, and statements of relationships. Another perspective is that generalizable knowledge is information that has the potential to be expanded beyond the specific circumstances in which it was acquired to any broader context. In a medical context, this means that the conclusions of a research study should be applicable to individuals beyond those who participated in the study. It is important to appreciate that publication of, or the intent to publish, the results of a research study does not in itself constitute generalizable knowledge. The Common Rule defines human subject as a living individual from whom an investigator conducting research obtains either (1) data through intervention or interaction with the individual and/or (2) the individual’s personally identifiable information. In this context, intervention includes data gathered by physical means, by manipulation of the individual, or by manipulation of the individual’s environment undertaken for research purposes, and interaction means communication or interpersonal contact between the individual and the researcher. The term researcher refers to anyone involved in conducting research including information gathering, interpretation, data analysis, and authorship of research results. Under the Common Rule, the act of coding personally identifiable information associated with FFPE tissues is not considered a research activity unless the person performing the coding is also involved in other aspects of the study that fall under the definition of researcher as delineated above.
10
With, Evers, and Mason
The Common Rule defines personally identifiable information under 45 CFR 46.102(f) using the concepts of private information and identifiable private information. Private information includes information about behavior that occurs in a context in which an individual can reasonably expect that no observation is taking place (such as in a public restroom) and that information which has been provided for specific purposes by an individual and which the individual can reasonably expect will not be made public (such as a healthcare record). Private information is further characterized as identifiable private information if the identity of the subject is, or could be, ascertained by the researchers, or if the identity of the subject is contained within the private information. Identifiable private information is also information collected specifically for the proposed project through intervention or interaction with living individuals where the investigator can readily ascertain the identity of the individuals. Thus, the Common Rule defines personal information in the context of behavior, information, and documentation. This differs from the concept of personal information as defined under HIPAA (see Subheading 5). There are three levels of IRB review based primarily upon the characteristics of the research subjects and the risks to which they will be exposed. These levels are full review, expedited review, or exempt from review. A full IRB review is required when more than minimal risk is involved or if the research subjects are members of a vulnerable population. Expedited review is applicable to research protocols that involve no more than minimal risk or review of minor changes to previously approved protocols. A protocol is eligible for expedited review if certain specific requirements are satisfied (see Note 7) as determined by the IRB chairperson or a designated board member. The Common Rule establishes that research is exempt from IRB review if the pro tocol satisfies any one of six specific requirements (see Note 8). This determination cannot be made by the researcher and the final determination should be made by the IRB. The exemption must be documented and indicate the specific criteria upon which the exemption is based (see Note 8). Finally, the Common Rule requires all IRB-approved protocols to undergo annual review (continuing review) to determine if a change in status is required due to a change in research direction, a change in the number of study subjects, a change in informed consent, reported safety issues, changes in the risks or benefits to the study participants, or a change in research personnel. The exemption most relevant to FFPE tissues is that concerning publicly available or unidentified biospecimens [45 CFR 46.101(b)(4)] (see Note 8). However, the applicability of this exemption is closely tied to the design of the research protocol. First, the FFPE tissue must be publicly available. This term was first introduced to apply to data, such as birth records, that
Regulatory and Ethical Issues on the Utilization of FFPE Tissues in Research
11
were available in the public domain. In the case of biospecimens, publicly available could be interpreted as indicating that they must come from repositories that make them available to researchers beyond those directly affiliated with the repository. However, there is currently no consensus on the interpretation of publicly available as it relates to biospecimens. The second key term is existing biospecimens, meaning that the FFPE tissues must exist prior to the protocol being submitted to the IRB. This precludes prospective collection or even identification of the set of biospecimens to be used in the protocol. The use of coded biospecimens is consistent with this exemption, but only if no member of the research team was involved in the coding and had access to patient identifiable private information. Also the researchers cannot have access to the code key or have any other means of matching the biospecimens back to the donors. To be exempt, the FFPE tissues must have been collected for purposes other than research, such as for diagnosis or surgical procedures, because biospecimens collected specifically for research would require informed consent. Further, most IRBs will not consider a protocol exempt if the collecting pathologist is a member of the research team as they will have knowledge of the patient’s personally identifiable information. Finally, even if a protocol is deemed not to be human subjects research and is exempt from IRB review, HIPAA privacy regulations may still apply as discussed in Subheading 5.
5. The HIPAA Privacy Rule Arguably, the most confusing regulation relevant to research with FFPE tissue biospecimens is the issue of personally identifiable information. The Common Rule provides a conceptual definition (see Subheading 4), whereas the HIPAA Privacy Rule provides a much more prescriptive definition of personally identifiable information. Again, as under the Common Rule, the HIPAA Privacy Rule does not cover biospecimens themselves, but rather the personally identifiable information linked to the biospecimens. Under the Privacy Rule, individually identifiable health information is defined as a subset of health information, created or received by a covered entity or employer and related to past, present, or future physical or mental health or condition of an individual; the provision of health care to an individual; or the past, present, or future payment for provision of health, and that identifies the individual or there is a reasonable basis to believe the information can be used to identify the individual (45 CFR Part 160 and Subparts A and E of Part 164). A further delineation is the concept of Protected Health Information (PHI), which is defined as individually identifiable health information transmitted by electronic
12
With, Evers, and Mason
media, maintained in electronic media, or transmitted or maintained in any other form or medium. PHI specifically excludes educational records. A covered entity is defined as a health plan, a health care clearinghouse, or health care provider who transmits health information in electronic form in connection with a transaction for which the DHHS has adopted a standard. The covered entity definition is fairly restrictive, leaving most entities that conduct research not covered by HIPAA. However, recent recommendations by the Institute of Medicine (IOM) argue for broader coverage that would include most entities that conduct research involving PHI (12). According to the Privacy Rule, de-identified PHI is health information that does not identify an individual and for which there is no reasonable basis to believe that the information can be used to identify an individual (45 CFR Part 164.514(a)–(c)). Under HIPAA, PHI can be de-identified either by removing 18 key identifiers or by using a method for statistical verification of de-identification. In the first method, 18 identifiers (see Note 9) that could be used to identify the individual or their employers, relatives, or household members are removed from the individual’s record so that the remaining information, alone or in combination with other information, cannot be used to identify the individual. Under the Privacy Rule, coded information is considered de-identified if the 18 identifiers have been removed provided that the code itself is not derived from any of the 18 identifiers, that the code cannot be used to identify the donor, and that the covered entity does not use or disclose the code or the mechanism for re-identification. In the second method, a person with an appropriate knowledge of and experience with generally accepted statistical and scientific principles and methods for rendering information not individually identifiable may certify that there is a very small risk that the information be used by the recipient to identify the donor. A covered entity can use or disclose a donor’s PHI for research by obtaining a privacy rule authorization. A privacy rule authorization is an individual’s signed permission that allows a covered entity to use or disclose the donor’s PHI for the purposes, and to the recipient or recipients, as stated in the authorization. An authorization differs from an informed consent in that the authorization focuses on privacy risks and states how, why, and to whom the PHI will be used and/or disclosed for research. An informed consent provides research subjects with a description of the study and its anticipated risks and benefits, and a description of how the confidentiality of records will be protected. A privacy rule authorization can be combined with an informed consent document to participate in research (see Note 10). The HIPAA Privacy Rule requires authorization for the use or disclosure of PHI; however, authorization is not required if the
Regulatory and Ethical Issues on the Utilization of FFPE Tissues in Research
13
biospecimen has been de-identified (as described above) or if one of the following four conditions is met: (1) documentation that an IRB or equivalent Privacy Board has waived the authorization requirement in accordance with the conditions specified in 45 CFR Part 164.512(i) (see Note 11); (2) the use of PHI is solely to prepare a research protocol or for similar purposes preparatory to research, that the researcher will not remove any PHI from the covered entity, and that access to the requested PHI is required for preparing of the research protocol as specified in 45 CFR Part 164.512(i)(1)(ii); (3) the use or disclosure of PHI is solely for research on decedents, that the requested PHI is necessary for the research, and at the request of the covered entity documentation of death of the individuals in question is provided as specified in 45 CFR Part 164.512(i)(1)(iii); (4) a data use agreement is entered into by the researcher and the covered entity, pursuant to which the covered entity may disclose a limited data set to the researcher for research purposes as specified in 45 CFR Part 164.514(e). A limited data set excludes specific direct identifiers of the individual or of relatives, employers, or household members of the individual (see Note 12).
6. Concluding Thoughts Regarding FFPE Tissues
From the preceding material, it should be clear that in the US there is no single regulation that covers FFPE tissue specimens, and that the most relevant regulations, the Common Rule and the HIPAA Privacy Rule, provide apparent conflicting guidance. This circumstance has developed, in large part, from the fact that advances in research, medicine, and biorepository sciences have outpaced legislation. However, this makes it difficult for researchers to prepare protocols involving FFPE tissues and equally difficult for IRBs to review them. The OHRP (http://www.hhs.gov/ ohrp) and the DHHS Office of Civil Rights (OCRs) for HIPAArelated issues (http://www.hhs.gov/ocr/hipaa) are the sources most often consulted for guidance. OHRP has attempted to clarify research involving biospecimens, including FFPE tissues, in regards to what constitutes human subjects research. The OHRP Guidance on Research Involving Coded Private Informa tion or Biological Specimens (http://www.hhs.gov/ohrp/policy/ cdebiol.pdf) concludes that research on FFPE tissue biospecimens is not considered human subjects research if two criteria are met (see Note 13). Research using FFPE tissues from deceased persons is another area of confusion. The Common Rule is silent on research involving biospecimens and associated personally identifiable information obtained from deceased individuals. However, the Common
14
With, Evers, and Mason
Rule definition of human subjects as “living individuals” is generally interpreted to mean that such biospecimens are not covered. The HIPAA Privacy Rule does specifically cover deceased individuals and requires that certain assurances be obtained from the researcher in regards to the use or disclosure of PHI from deceased individuals. These assurances are (1) that the use and disclosure of PHI is solely for research, (2) that the PHI is necessary for the research, and (3) that documentation of death of the individuals in the study be provided, if requested by the covered entity. There are two additional areas of conflict between the Common Rule and the Privacy Rule. The first involves deidentification of personally identifiable information associated with biospecimens. Both regulations exempt de-identified and coded information for research purposes, but the specific requirements for de-identification differ. Thus, to be fully compliant it is necessary to follow both the Privacy Rule requirements for PHI and the Common Rule requirements for identifiable private information when de-identifying documentation associated with biospecimens. The second conflict centers on the difference between informed consent and authorization. The Common Rule is interpreted by OHRP as allowing the informed consent document to specify that donated tissue may be used for research not anticipated at the time of collection. In contrast, the Privacy Rule requires that an authorization be tied to a specific research protocol and does not authorize future unspecified use of the requested PHI. The Clinical Laboratory Improvement Amendments of 1988 (CLIA), 42 CFR 493.1105 (B) provides that laboratory facilities retain pathology specimen blocks for at least 2 years from the date of examination. In contrast, the College of American Pathologists (CAP) mandates the retention of FFPE blocks for a minimum of 10 years (13) and there are state regulations that require even longer retention times. These regulations are not intended to discourage research using FFPE tissues, but rather to ensure that patient material is available to confirm diagnosis, apply future analytical tests, support patient participation in clinical trials requiring access to original case material, and for legal purposes. However, there are no specific recommendations on how much case material should be retained. The key point is to retain sufficient FFPE tissue to allow a pathologist to confirm or alter the original diagnosis, and to analyze the tissue for diagnostic or prognostic markers that may become known in the future. If several blocks are available, it is certainly advisable to save one representative block in pristine condition. If a single block is available, retaining one-third to one-half of the block containing the lesion in question would seem prudent. An additional consideration concerns the use of FFPE blocks to prepare tissue microarrays (TMAs), for which a core punch is taken from the block rather
Regulatory and Ethical Issues on the Utilization of FFPE Tissues in Research
15
than a whole section. In this case, it is important not only to retain a sufficient fraction of the lesion for future analysis, but one should also be careful that the core punch(es) does not obscure the boundary of the lesion or compromise the pathologists’ ability to appreciate the relationship between the lesion and the adjacent tissue. A final comment is in regard to archival FFPE tissue collected decades ago before the modern era requiring informed consent and protection of personally identifiable information. Again, there are no specific regulations addressing research with archival tissues that fall in this category. A practical suggestion of attempting to obtain retroactive consent for identified specimens that are £10 years old has been proposed in the literature (14). Perhaps the best perspective for older archival specimens is summarized in a quote from Richard Ashcroft of the Imperial College of Medicine, London, UK (15), with additionally comments by William Stempsey of Loyola University, Chicago, USA (16), “in areas of moral complexity and change, we gain nothing by judging our past actions with our new-found wisdom of hindsight and in the light of our hard-won consensus. In particular, the re-use of existing archives gathered in the past is essential, and must be managed effectively, as a matter of respect to past patients and minimizing the burden on current patients. But, at the same time, having made our decision about the ethical standards we now wish to apply to pathological research, we must stick to it in collecting new samples and constructing new archives.”
7. Notes 1. The lack of uniformity in the terminology used to refer to the different types of collections of tissue specimens – tissue repository, biobank, specimen repository, biospecimen repository, biorepository, tissue bank, human tissue repository – in addition to being confusing, has led to many challenges in determining the applicability of some of the regulatory guidance and procedures (Human Tissue Banking White Paper, Part I Assessment and Recommendations (2007), PRIM&R Human Tissue/Specimen Banking Working Group and Partners HealthCare Systems Inc., available at: http://www. primr.org/uploadedFiles/PRIMR_Site_Home/Public_ Policy/Recently_Files_Comments/Tissue%20Banking%20 White%20Paper%203-7-07%20final%20combined.pdf). Specimens and associated health information can be collected prospectively or obtained from existing archives. Specimens can be collected specifically for research purposes or obtained during the course of routine medical care (e.g., residual material
16
With, Evers, and Mason
remaining after the removal of a tumor). Therefore, in order to determine which regulations apply to the tissue, one must determine the manner and purpose by which the tissue was obtained. For the purposes of the discussions in this chapter, the terms biorepository and human tissue repository will be used. 2. A tiered consent form may contain the following list of options as described in the NCI Best Practices for Biospecimen Resources document (see Table 1): (a) the type of tissue biospecimen(s) the donor wishes to contribute, (b) the type(s) of research the biospecimens may be used for (specific or global), (c) an option to specify that tissues may not be collected for certain research activities (such as genetic sequencing), (d) if the donors personally identifiable information may be accessed or if the donation is to be anonymized, (e) if the donor, a surviving relative, or the donor’s physician can be contacted in the future about the use of their biospecimens or personally identifiable information, and (f) the right of the donor to withdraw from the research study or repository and what is to be done with their biospecimens and personally identifiable information if this option is exercised. 3. Examples of procedures that are considered minimal risk are (1) clinical studies of new applications of drugs and medical devices already approved for marketing, (2) collection of blood samples by finger stick, heel stick, ear stick, or venipuncture, (3) prospective collection of biological specimens for research purposes by noninvasive means, (4) collection of data through noninvasive procedures (not involving general anesthesia or sedation) routinely employed in clinical practice, excluding procedures involving X-rays or microwaves, (5) research involving materials (data, documents, records, or specimens) that have been collected, or will be collected, solely for nonresearch purposes (such as for medical treatment or diagnosis), (6) collection of data from voice, video, digital, or image recording made for research purposes, (7) research on individual or group characteristics or behavior (including, but not limited to, research on perception, cognition, motivation, identity, language, communication, culture beliefs or practices, and social behavior) or research employing survey, interview, oral history, focus group, program evaluation, human factors evaluation, or quality assurance methodologies. There is a great deal of controversy in assessing risk associated with genetic research, particularly germline sequencing studies, and how this might compromise privacy for the individual or their family members. The National Bioethics Advisory Commission has published guidelines on this issue and concluded that the majority of human subjects genetic research studies should be considered
Regulatory and Ethical Issues on the Utilization of FFPE Tissues in Research
17
minimal risk (see http://bioethics.georgetown.edu/pcbe/ r epor ts/past_commissions/nbac_biological1.pdf). However, this may change if, in the future, genetic sequen cing and associated data banking of genetic information becomes so commonplace that DNA itself falls in the category of personally identifiable information. 4. Obtaining informed consent for existing data sets is considered impractical under the following conditions: (a) the size of the population being researched is extremely large, (b) a large proportion of individuals have likely relocated or died since the personally identifiable information was originally collected, (c) obtaining informed consent is likely to introduce bias into the research, (d) obtaining informed consent is likely to create threats to privacy by having to link otherwise de-identified data with nominal identifiers, (e) there is a risk of inflicting psychological or social harm by contacting individuals or their families, (f) it would be difficult to contact individuals directly if the relationship with the researcher or repository no longer exists, (g) it would be difficult to contact the individuals through public means, such as advertisements or notices, and (h) if in any of the above circumstances the requirement for additional financial, material, human, or other resources to obtain consent would impose an undue hardship on the researcher or organization (see http://dev.cihr.ca/e/ documents/et_pbp_nov05_sept2005_e.pdf, pages 6 and 7). 5. The IRB should, at a minimum, assess the following aspects of a research protocol and request revisions or alterations if any aspects of the protocol are not in compliance: (1) there is appropriate prior informed consent for the protocol, an informed consent document appropriate for the protocol has been prepared, or that a waiver of informed consent has been approved for the protocol, (2) the privacy of personally identifiable information is protected, (3) the research protocol is scientifically and statistically sound, (4) that safeguards exist to ensure the safety of the research subjects and appropriate data safety monitoring is provided, (5) that the research subjects have been equitably selected, and (6) that vulnerable populations are protected. 6. A more detailed list of questions to be asked by the IRB when deciding the status of a protocol: (1) is the activity research, (2) if so, does the research involve human subjects, (3) if so, is the research supported in whole or in part by federal funds or an OHRP-Approved Federal-Wide Assurance agreement, (4) if so, is the research subject to exemption, and (5) if not exempt, is it entitled to expedited review by the IRB. 7. The following requirements must be met for a protocol to be eligible for expedited review: (1) the protocol presents minimal
18
With, Evers, and Mason
risk to the research subjects, (2) the risks to the research subjects are reasonable in relation to the anticipated benefits, (3) the selection of the research subjects was/will be equitable and noncoercive, (4) appropriate informed consent will be obtained from the perspective research subjects unless informed consent is waived, (5) informed consent will be properly documented, (6) when appropriate the research plan provides for adequate monitoring to ensure safety for the subjects, and (7) there are provisions to adequately protect the privacy of the research subjects and to maintain the confidentiality of the subjects. 8. The following activities are exempt from IRB review: (1) Exemption 45 CFR 46.101(b)(1): research on common educational practices in educational settings, (2) Exemption 45 CFR 46.101(b)(2): research involving educational tests (cognitive, diagnostic, aptitude, achievement), surveys or observations of public behavior that is not recorded in an identifiable format and could not place subjects at risk for criminal or civil liability or damage the subjects’ reputation, employability, or financial standing, (3) Exemption 45 CFR 46.101(b)(3): research involving educational tests (cognitive, diagnostic, aptitude, achievement), surveys, or observations of public behavior involving elected or appointed public officials or candidates, or if the information is required under federal statute to be kept confidential throughout the research and thereafter, (4) Exemption 45 CFR 46.101(b) (4): research involving existing data, documents, records, pathological biospecimens, or diagnostic biospecimens, if these sources are publicly available or if the information is recorded by the investigator in such as manner that subjects cannot be identified, directly or through identifiers linked to the subjects, (5) Exemption 45 CFR 46.101(b)(5): research conducted by agency/department heads to evaluate public benefit or service programs; procedures for obtaining benefits or services; possible changes or alternatives to the programs or procedures; possible changes in payment levels; and methods for services under these programs. The research must be conducted pursuant to specific Federal statutory authority, and (6) Exemption 45 CFR 46.101(b)(6): taste and food quality examinations and consumer acceptance studies. 9. The 18 identifiers under the Privacy Rule are (1) names, (2) geographic subdivisions smaller than a state, including street address, city, county, precinct, zip code, and their equivalent geocodes, except for the initial three digits of the zip code, (3) all elements of dates (except year) for dates directly related to an individual (e.g., data of birth, admission, diagnosis,
Regulatory and Ethical Issues on the Utilization of FFPE Tissues in Research
19
and age when ³89), (4) telephone numbers, (5) fax numbers, (6) electronic mail address, (7) social security numbers, (8) medical record numbers, (9) health plan beneficiary numbers, (10) account numbers, (11) certificate/license num bers, (12) vehicle identifiers and serial numbers, including license plate numbers, (13) device identifiers and serial numbers, (14) web universal resource locators (URLs), (15) internet protocol (IP) address numbers, (16) biometric identifiers, including finger and voiceprints, (17) full-face photographic images and any comparable image, and (18) any other unique identifying number, characteristic, or code. 10. For examples of combined Common Rule informed consent and HIPAA privacy authorization documents see (http:// privacyruleandresearch.nih.gov/pr_02.asp) and (https://rcr. gradsch.wisc.edu/cafwizard/start.asp?wisc#). 11. Approval of a waiver or modification of authorization to use or disclose PHI may be approved by an IRB or equivalent Privacy Board if all of the following conditions are met: (1) the use or disclosure of PHI involves no more than a minimal risk based upon the following elements: (a) an adequate plan to protect identifiers from improper use or disclosure, (b) an adequate plan to destroy the identifiers at the earliest opportunity consistent with the research study unless there is a justification for retaining the identifiers, and (c) adequate written assurances that the PHI will not be reused or disclosed to other parties except as required by law or oversight; (2) the research could not practicably be conducted without the waiver or alteration; and (3) the research could not practicably be conducted without access to and use of the PHI. 12. The limited data set excludes all of the 18 individual identifiers required for de-identification of PHI (see Note 9) with the acceptation of the following: dates, postal address information limited to city state and zip code, and any other unique identifying number, characteristic, or code (45 CFR Part 164.514(e)(3)(i)). Further requirements for use of a limited data set with data use agreement are: (1) the covered entity must establish the permitted uses and disclosures of the limited data set consistent with the purpose of the research, and which may not include any use or disclosure that would violate the Privacy Rule by the covered entity; (2) limit who can use and receive the data; and (3) require the researcher to agree to the following: (a) not to use or disclose the PHI other than permitted by the data use agreement or otherwise as required by law, (b) use appropriate safeguards to prevent disclosure of the PHI other than as provided by the data usage agreement, (c) report to the covered entity any improper use or disclosure of the PHI of which the recipient becomes
20
With, Evers, and Mason
aware, (d) ensure that any third party working with or for the researcher who has access to the PHI follows the same terms and restrictions as the researcher, and (e) not to identify the PHI or contact any individual. 13. Research involving only coded private information or specimens does not involve human subjects research if both the following conditions are met: (1) the private information or specimens were not collected specifically for the currently proposed research project through an interaction or intervention with living individuals, and (2) the investigator(s) cannot readily ascertain the identity of the individual(s) to whom the coded information or specimens pertain because, for example: (a) a key to decipher the code is destroyed before the research begins, (b) the investigators and the holder of the individual identifiers enter into an agreement prohibiting the release of individual identifiers to the investigators under any circumstances until the individuals are deceased, (c) there are IRB-approved written policies and operating procedures for a repository or data management center that prohibit the release of individual identifiers to the investigators under any circumstances until the individuals are deceased, or (d) there are other legal requirements prohibiting the release of individual identifiers to the investigators, until the individuals are deceased. References 1. Text of the Nuremberg Code can be found at: http://www.hhs.gov/ohrp/archive/nurcode.html. 2. Text of the Declaration of Helsinki can be found at: http://www.wma.net/en/30publications/ 10policies/b3/index.html. 3. Bankert, E., and Amdur, R. J., eds. (2006) Institutional Review Board: Management and Function. Jones and Bartlett Publishers, Sudbury, MA. 4. Text of the Belmont Report can be found at: http://www.hhs.gov/ohrp/humansubjects/ guidance/belmont.html. 5. Kapp, M. B. (2006) Ethical and legal issues in research involving human subjects: do you want a piece of me? J Clin Pathol. 59, 335–339. 6. Schwartz, A. J. (2001) Oversight of Human Subject Research: The Role of the States, in National Bioethics Advisory Commission, Ethical and Policy Issues in Research Involving Human Participants, Vol. II, pp. M-1 through M-20. Available at: http://bioethics.georgetown.edu/ nbac/human/overvol2.html.
7. Pritts, J., Goldman, J., Hudson, Z., Berenson, A., and Hadley, E. (1999) The State of Health Privacy: An Uneven Terrain – A Comprehensive Survey of State Health Privacy Statutes. Health Privacy Project; Institute for Health Care Research and Policy, Georgetown University. Available at: https://gushare.georgetown. edu/jlp/1999%20State%20Report/State%20 Report%201999.pdf. 8. Weir, R. F., and Olick, R. S. (2004) The Stored Tissue Issue: Biomedical Research, Ethics, and Law in the Era of Genomic Medicine. Oxford University Press, New York, NY. 9. An excellent sample MTA can be found at: http://privacyruleandresearch.nih.gov/ pr_02.asp and https://rcr.gradsch.wisc.edu/ cafwizard/start.asp?wisc#. 10. National Bioethics Advisory Commission (NBAC) August 1999. Available at: http:// bioethics.georgetown.edu/pcbe/reports/ past_commissions/; and, http://bioethics. georgetown.edu/pcbe/reports/past_commissions/nbac_biological1.pdf; and http:// bioethics.georgetown.edu/pcbe/reports/ past_commissions/nbac_biological2.pdf.
Regulatory and Ethical Issues on the Utilization of FFPE Tissues in Research 11. Merz, J. F., Sankar, P., Taube, S. E., and Livolski, V. (1997) Use of human tissues in research: clarifying clinician and research roles and information flows. J Invest Med. 45, 252–257. 12. A copy of the report brief “Beyond the HIPAA privacy rule: enhancing privacy, improving health through research”. Available at: http:// www.iom.edu/hipaa. 13. College of American Pathologists Laboratory Accreditation Program Inspection Checklists. Available at: http://www.cap.org and http://www.cap.org/apps/cap.por tal?_ nfpb=true&cntvwrPtlt_actionOverride=%2Fp ortlets%2FcontentViewer%2Fshow&_windo
21
wLabel=cntvwrPtlt&cntvwrPtlt%7BactionForm. contentReference%7D=cap_links%2Flap_ linkschar ts.html&_state=maximized&_ pageLabel=cntvwr. 14. Vermeulen, E., Schmidt, M. K., Aaronson, N. K., Kuenen, M., and van Leeuwen, F. E. (2009) Obtaining ‘fresh’ consent for genetic research with biological samples archived 10 years ago. Eur J Cancer. 45, 1168–1174. 15. Ashcroft, R. (2000) The ethics of reusing archived tissue for research. Neuropathol Appl Neurobiol. 26, 408–411. 16. Stempsey, W. E. (1989) The virtuous pathologist. An ethical basis for laboratory medicine. Am J Pathol. 91, 730–738.
wwwwwwwwwwwwwww
Chapter 2 Tissue Microarrays: Construction and Uses Carol B. Fowler, Yan-Gao Man, Shimin Zhang, Timothy J. O’Leary, Jeffrey T. Mason, and Robert E. Cunningham Abstract Tissue microarrays (TMAs) are produced by taking small punches from a series of paraffin-embedded (donor) tissue blocks and transferring these tissue cores into a positionally encoded array in a recipient paraffin block. Though TMAs are not used for clinical diagnosis, they have several advantages over using conventional whole histological sections for research. Tissue from multiple patients or blocks can be examined on the same slide, and only a very small amount of reagent is required to stain or label an entire array. Multiple sections (100–300) can be cut from a single array block, allowing for hundreds of analyses per microarray. These advantages allow the use of TMAs in high-throughput procedures, such as screening antibodies for diagnostics and validating prognostic markers that are impractical using conventional whole tissue sections. TMAs can be used for immunohistochemistry, immunofluorescence, in situ hybridization, and conventional histochemical staining. Finally, several tissue cores may be taken without consuming the tissue block, allowing the donor block to be returned to its archive for any additional studies. Key words: Fluorescence in situ hybridization, Formalin-fixed, paraffin-embedded, Hematoxylin and eosin, HER-2 gene, Tissue microarray, TMA
1. Introduction Tissue microarrays (TMAs) were first described by Wan et al. (1), and are constructed by transferring small tissue punches from formalin-fixed, paraffin-embedded (FFPE) tissue blocks to spatially fixed positions in a recipient block. The method was further developed by Kononen et al. (2) and Camp et al (3). In recent years, high-density TMAs have become a standard laboratory tool for identifying and validating diagnostic and prognostic biomarkers for a variety of diseases, such as breast cancer (3, 4), gastrointestinal tumors (5), prostate cancer (6, 7), and lung cancer (8).
Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_2, © Springer Science+Business Media, LLC 2011
23
24
Fowler et al.
Fig. 1. A general method for tissue microarray construction: (a) a blank recipient block is prepared and cored. (b) A 0.6-mm tissue punch is extracted from the donor block, then (c) relocated to the prepared hole in the recipient block. (d) To section the TMA, a tape window is applied to the TMA, and a 5-mm section is cut using a conventional microtome. (e) The tape window is applied to an adhesive-coated slide, and the adhesive is cured under UV light (not shown). (f) An H&E-stained TMA section (2.5× magnification).
For a comprehensive review of tissue array techniques and applications, see Eguíluz et al. (9). A general scheme for constructing TMAs is shown in Fig. 1. First, the region of interest is identified on a donor tissue block. A hematoxylin and eosin (H&E)-stained slide serves as a useful guide for selecting the area to be sampled. Next, a small (0.6–3 mm) punch is taken from the donor tissue block and seated into a recipient paraffin block in a positionally encoded array format. The tissue punches can be taken with a manual or automatic tissue arrayer, such as those manufactured by Beecher Instruments (Sun Prairie, WI). Once the TMA is constructed, it
Tissue Microarrays: Construction and Uses
25
can be sectioned, and mounted on a microscope slide. The TMA can then be stained using a standard histochemical method (i.e., H&E), or more commonly used for immunohistochemical, immunofluorescence, or fluorescence in situ hybridization (FISH) studies. An example of using FISH with TMAs is for the analysis of HER-2 amplification in breast tumors. The HER-2 gene, which maps to 17q21.1, is amplified in ~20% of breast cancers (10) leading to an overexpression of its protein product, human epithelial growth factor receptor 2, a cell surface tyrosine kinase receptor. Such cancers are receptive to a targeted therapy by the monoclonal antibody trastuzumab (Herceptin). Accordingly, the accurate assessment of HER-2 status is critical to predict the responsiveness of breast cancer to adjuvant treatment with trastuzumab. Determination of HER-2 status is usually carried out by FISH with whole tissue sections using hybridization probes for the HER-2 gene and a reference hybridization probe for the centromere of chromosome 17 (11). A tumor with a HER-2/chromosome 17 ratio of >2.2 is considered to be positive for HER-2 amplification (10). Several recent studies have compared whole sections and TMAs for the assessment of HER-2 status using FISH (12, 13). These studies concluded that the use of TMAs to assess HER-2 status gave comparable results to the current standard methodology using whole sections. In addition, these studies demonstrated that TMA technology offered improvements in cost, time, and quality control. A procedure for performing FISH for HER-2 status using a TMA will be described later in the chapter. Once stained, TMAs may be analyzed in one of two ways. Biomarker expression can be manually assessed using an ordinal grading scale. Though this method has been used for a number of microarray studies, it is time-consuming, semiquantitative, and requires an experienced observer. Dedicated TMA readers and analysis software can improve and simplify the evaluation of TMAs, while streamlining data archiving. Automated analysis protocols can select the region of interest and normalize it so that expression levels can be compared, both between different tissue cores on the same array slide and between different array slides. Some systems use immunofluorescent substrates, while others use chromogenic substrates to quantify biomarkers. Localization of disease biomarkers is usually achieved by counterstaining the TMA with H&E, then the automated systems use morphometric analyses to distinguish the tumor from normal cellular features. A partial list of available TMA analysis software and slide readers is shown in Table 1. Some of these systems have dedicated applications for HER-2, estrogen receptor, or progesterone receptor [i.e., Automated Cellular Imaging System (ACIS III), AQUA (14), and Pathological Image Analysis and Management (PATHIAM)].
26
Fowler et al.
Table 1 Available tissue microarray analysis tools TMA analysis system
Source/reference
Automated Cellular Imaging System (ACIS®III)
Dako North America, Inc., Carpinteria, CA http://www.dakousa.com
Automated Quantitative Analysis Software AQUAnalysis™ and AQUAsition™ (14)
HistoRx, New Haven, CT http://www.historx.com/
TMA Lab and ScanScope slide scanning systems
Aperio Technologies, Inc., Vista, CA http://www.aperio.com/
Pathological Image Analysis and Management (PATHIAM)
Bioimagene, Inc., San Jose, CA http://www.bioimagene.com/
Alpha Scan microarray scanner and Image Acquisition and Analysis software
Alpha Innotech Corp., San Leandro, CA http://www.cellbiosciences.com/
TMA deconvoluter/Stainfinder (15)
http://genome-www.stanford.edu/TMA/
2. Materials 2.1. Construction of the Tissue Microarray
1. Paraplast Tissue Embedding Media (Oxford Labware, St. Louis, MO) or similar. Hold the paraffin at 60°C prior to use. 2. Slotted tissue cassettes (Tissue-Tek, Sakura Finetek, Torrance, CA). 3. Base molds up to 37 × 24 × 5 mm (Tissue-Tek). 4. “Donor” blocks with paraffin-embedded tissue of interest (see Note 1). 5. H&E-stained slides corresponding to the paraffin blocks of interest (see Note 1).
2.2. Sectioning of the Tissue Microarray
1. Adhesive-coated microscope slides (Paraffin-Tape Transfer Slides, Instrumedics, Inc., St. Louis, MO). 2. Tape windows for section transfer (Instrumedics, Inc.). 3. TPC SOLVENT TM for removing the tape window and adhesive (Instrumedics, Inc.).
2.3. Clearing and Rehydrating Slides for Staining
1. Coplin jars for slide incubation (Fisher Scientific, Pittsburgh, PA). 2. Xylene (Sigma-Aldrich, St. Louis, MO) (see Note 2). 3. 100% Ethanol (Sigma-Aldrich) used to prepare the graded series of alcohols (100, 95, 70, and 30%) for tissue rehydration. 4. Distilled water.
Tissue Microarrays: Construction and Uses
2.4. Hematoxylin and Eosin Staining
27
1. Coplin jars for slide incubation (Fisher). 2. Mayer’s hematoxylin solution (Sigma-Aldrich). Protect from light (see Note 3). 3. EosinY with phloxine solution (Sigma-Aldrich; see Note 4). 4. Xylene (Sigma-Aldrich; see Note 2). 5. 100% Ethanol (Sigma-Aldrich) used to prepare the 95 and 80% ethanol solutions. 6. Distilled water. 7. Permount mounting media (Fisher). 8. Glass coverslips (Fisher).
2.5. Fluorescence In Situ Hybridization
1. Coplin jars for slide incubation during the pretreatment, protease, and rinse steps (Fisher). 2. Pretreatment solution: 1 M sodium thiocyanate (SigmaAldrich), preheated to 97°C in a water bath. 3. Protease buffer: 0.9% (w/v) sodium chloride (Sigma-Aldrich). Adjust to pH 2 with HCl and preheat to 37°C in a water bath. 4. Pepsin, porcine (50 mg/mL), from Sigma-Aldrich (3,200– 4,500 AU/mg protein). 5. Sodium chloride/sodium citrate buffer (SSC, 2×): 17.5 g sodium chloride, 8.8 g SSC (Sigma-Aldrich) in 1 L distilled water. Adjust the pH to 7.0. 6. Distilled water. 7. Fluorescent DNA probes [Vysis LSI HER-2/neu (red) and CEP 17 (green), Abbot Molecular, Inc., Des Plaines, IL]. Protect from light. 8. Posthybridization buffer I: 2× SSC with 0.3% (v/v) NP-40 (Sigma-Aldrich). 9. Posthybridization buffer II: 2× SSC with 0.1% (v/v) NP-40. 10. Glass coverslips (Fisher). 11. Rubber cement (Fisher). 12. Clear nail polish (Fisher). 13. Vectashield mounting media with DAPI (4´, 6-diamidino2-phenylindole, Vector Laboratories, Burlingame, CA). Store at 4°C in the dark.
3. Methods 3.1. Construction of the Tissue Microarray
1. This protocol assumes the use of a manual tissue arrayer such as the MTA I from Beecher Instruments (Sun Prairie, WI), shown in Fig. 1. Other arrayers may be configured differently and their manual should be consulted prior to use (see Note 5).
28
Fowler et al.
2. Prepare an empty, “recipient,” paraffin block by pouring liquid paraffin into a stainless steel base mold. Cover with a slotted tissue cassette and allow to cool. Remove the recipient block from the mold and inspect the paraffin surface for any bubbles or holes. Trim any excess paraffin from the plastic cassette. 3. Identify regions of interest from the donor tissue blocks (see Note 1). 4. Determine the desired layout of the TMA and prepare a corresponding positional chart (TMA block summary) to record all of the pertinent information about each tissue core in the array (see Note 6). 5. Using an empty paraffin block, check the alignment of the two punches of the MTA I arrayer. The circular indents made by the small (recipient) and large (donor) punches on the paraffin block surface should have identical centers if they are properly aligned. If necessary, adjust the alignment of the punches as described in the manual. 6. Place the recipient block in the block holder. Adjust the depth stop by tightening the adjustment nut until the punch stops at the desired depth within the paraffin block, typically 0.5–1 mm above the base of the plastic tissue cassette. 7. Using the smaller punch, make a hole in the first position of the array (i.e., position A1). All other array positions will be in reference to this first spot. Accordingly, set the X and Y micrometers of the MTA I to zero. When the depth stop blocks the downward motion, slowly release the tissue punch. Eject the paraffin core. 8. Place the donor block bridge over the array block holder, and move the larger punch into the sampling position. Manually hold the donor block in position on top of the donor block bridge while positioning the area to be sampled directly underneath the sample punch (see Note 7). 9. Push downward on the sample punch to retrieve the tissue core. Note that the depth stop will not block the punch motion at the proper position for the donor block, so care must be taken to prevent the punch from entering too deeply into the block (see Note 8). 10. Remove the donor tissue block and bridge and push the punch downward until its tip reaches the top of the hole in the recipient array block. Use the large punch stylet to inject the tissue core into the hole created by the smaller punch. 11. Adjust the micrometers to move the tissue punch to the next X, Y position. A spacing of 0.8–1 mm between sample centers is customary (see Note 5).
Tissue Microarrays: Construction and Uses
29
12. Repeat steps 7–11, using the remaining donor tissue blocks, to construct the entire TMA. 13. Once the microarray is constructed, press the TMA block, paraffin side down, lightly on a flat surface to seat the tissue cores. 3.2. Sectioning of the Tissue Microarray
1. This protocol assumes the use of the Paraffin-Tape Transfer System from Instrumedics, Inc. (see Note 9). 2. To section the TMA, apply a tape window, adhesive side down, onto the surface of the TMA. Gently roll or press the tape window to remove any wrinkles or air bubbles. 3. Place the tissue block on a standard microtome, positioning the leading edge of tape window over the microtome blade. Cut a 5-mm section. A microarray block may yield 100–300 sections, depending upon core depth (see Note 10). 4. Press the tape window, with the TMA section facing downward, onto the adhesive-coated microscope slide. Use the roller to gently remove any air bubbles. 5. Place the tape-covered slide under a UV lamp for 30–60 s to cure the adhesive. 6. Place the slide in the TPC SOLVENT (TM) for 3 min. Carefully peel away the tape window. The TMA slide is now ready for clearing and staining.
3.3. Clearing and Rehydrating Slides for Staining
1. Incubate the TMA slide through three changes of xylene, 2 min each, to clear the paraffin. 2. Rehydrate the slide through two changes each of 100, 95, 70, and 30% ethanol, 2 min each. 3. Place the slide in distilled water prior to staining (see Subheading 3.4 or 3.5).
3.4. Hematoxylin and Eosin Staining
Adapted from the AFIP Laboratory Methods in Histotech nology (16). 1. Once the TMA slide is cleared and rehydrated, stain for 4–15 min in the Mayer’s hematoxylin solution. 2. Remove the slide from the stain and wash in lukewarm running tap water for 15 min. 3. Transfer the slide to distilled water until ready for eosin staining. 4. Rinse the slide in 80% ethanol for 1–2 min. 5. Counterstain in the eosin–phloxine solution for 1–2 min. 6. Dehydrate through two changes each of 95% ethanol, 100% ethanol, and xylene. Each incubation is for 2 min. 7. Mount with a xylene-based mounting media, such as Permount.
30
Fowler et al.
3.5. Fluorescence In Situ Hybridization
1. This protocol is for detecting the amplification of the HER-2 gene by FISH in FFPE breast cancer tissues, but can be adapted for other genes. Individual probe concentrations will need to be optimized. For an example of a TMA slide stained positive for HER-2 gene amplification, see Fig. 2. 2. Remove the TMA slide from the distilled water (Sub heading 3.3, step 3) and incubate in the pretreatment solution for 20 min at 97°C.
Fig. 2. Dual-color FISH showing that the HER-2 gene is locally amplified in a breast carcinoma. (a) A single H&E-stained TMA core, 40× magnification. (b) and (d) Selected areas at 400× magnification, and their corresponding FISH images (c) and (e). The Vysis HER-2 probe system was purchased from Abbot Molecular, Inc. The red HER-2 probe detects the HER-2 gene and the green Cep 17 probe is an internal control for chromosome 17. In normal breast cells, there are two green and two red probe signals per cell. The ratio between the red and green probe signals determines the HER-2 status of the tissue.
Tissue Microarrays: Construction and Uses
31
3. Remove the slide from the pretreatment solution and allow it to cool for 2 min. Rinse with distilled water. 4. Add 600 mL of the pepsin solution to 55 mL of prewarmed protease buffer. 5. Incubate the slide in the protease solution for 25 min at 37°C. 6. Wash the slide in 2× SSC, two times, 5 min each. Briefly dip the slide in distilled water. 7. Air-dry the slide or dry on a 42°C hot plate. 8. Care should be taken to protect the slide from light for steps 9–19. 9. Denature the amount of Her-2/Cep17 probe solution equivalent to that required for a standard whole tissue section (8–10 mL) in a 74°C water bath for 5 min. 10. Add the probe solution to a glass coverslip. Place the pretreated slide, tissue side down, on top of the coverslip and press to remove any air bubbles. Seal the edges of the coverslip with rubber cement. 11. Place the slide in a 37°C incubator. Incubate for 6–20 h (see Note 11 for an alternative method). 12. Remove the rubber cement from the coverslip by hand. 13. Soak the slide in posthybridization buffer I at room temperature to release the coverslip. 14. Incubate the slide in posthybridization buffer I for 2 min at 74°C. 15. Wash the slide in posthybridization buffer II at room temperature for 30 s. 16. Rinse the slide in distilled water. 17. Air-dry the slide or dry on a 42°C hot plate. 18. Add 1–2 drops of Vectashield mounting media with DAPI to a coverslip. Place the slide on top, tissue side down, and press to remove any air bubbles. Seal the edges of the coverslip with clear nail polish. 19. Place the slide in a slide folder and store at −20°C in the dark for at least 30 min prior to examination using a fluorescence microscope.
4. Notes 1. Choosing the correct sampling site from the donor blocks is critical and the most time-consuming step in TMA construction. A fresh H&E-stained slide from each block should be used as a guide to select the regions of interest for tissue sampling.
32
Fowler et al.
It may be helpful to mark the tissue sampling site on its corresponding slide prior to constructing the microarray. The following criteria for region of interest selection may be helpful: (a) The tissue block is typically examined by a pathologist to confirm the initial diagnosis and select the most appropriate area(s) in the tissue section for the extraction of the punch(es). (b) It is important that the tissue punches should not compromise the diagnostic value of the tissue block. Sufficient material should remain so that a pathologist can confirm the diagnosis and perform additional tests on the tissue. Also, the punches should not compromise the pathologists’ ability to appreciate the anatomic relationship between the lesion and the surrounding tissue. (c) Regions of interest within the tissue block must be ³2 mm in depth to be used in a TMA. (d) Regions of interest should not have excessive amounts of necrotic or poorly fixed tissue. 2. Short-term exposure to xylene can cause irritation of the skin, eyes, nose, and throat. Xylene should be handled in a certified chemical fume hood. Alternately, an orange-oil based product, such as Histo-Clear (National Diagnostics; Atlanta, GA) or Hemo-De (Scientific Safety Solvents; Keller, TX) may be used to remove excess paraffin from FFPE tissue sections. 3. Mayer’s hematoxylin solution stains the nuclei only. The blue color is enhanced by washing the slide in running tap water. The solution is commercially available, but may be prepared as follows: dissolve 50 g of ammonium or potassium alum in 1 L of distilled water, then add 1 g of hematoxylin crystals. When all of the hematoxylin has been dissolved, add 0.2 g of sodium iodate and stir for 10 min before adding 1 g of citric acid. Stir for an additional 10 min before adding 50 g of chloral hydrate. The resulting solution should be a deep “wine red” color. 4. Eosin Y is a conventional counterstain for hematoxylin and gives a wide range of contrast from pink to bright red. For example, cytoplasm stains pink and collagen and muscle stain bright red. The solution is commercially available, but may be prepared as follows: combine 100 mL of eosin stock solution (1 g Eosin Y in 100 mL in distilled water) with 10 mL of phloxine stock solution (1 g phloxine B in 100 mL distilled water). Add 780 mL of 95% ethanol and 4 mL of glacial acetic acid. The working solution is good for 1 week. 5. Manual tissue arrayers such as the MTA I will allow the average user to take 30–70 tissue punches per hour. Depending upon
Tissue Microarrays: Construction and Uses
33
tissue core size and density, recipient blocks can accommodate up to 1,000 samples. Using 0.6 mm punches, with a spacing of 0.8 mm between sample centers, the typical array size is 300–500 specimens per block using regular tissue cassettes. Using 2 mm punches allows construction of tissue arrays with about 50–100 specimens. An automated tissue arrayer can take up to 180 punches per hour (0.6–3 mm in size) and can produce up to 26 replicate array blocks in a single run. 6. For array orientation, it may be useful to leave the first position empty, or to substitute with a core taken from a FFPE 3% agarose-India ink plug. India ink will survive all stages of histology. Figures 3 and 4 show a slide diagram of an 80-core TMA and its corresponding TMA block summary, respectively.
1
2
3
4
5
6
7
8
9
10
Slide Label
A B C D E F G H Fig. 3. A slide diagram showing the physical layout of a TMA array on a 1″ × 3″. icroscope slide. The rows are labeled with letters and the columns are labeled with m numbers. The slide diagram shown is for a TMA consisting of an 8 × 10 array of tissue cores.
Block ID
Cases
Cores
Layout
Diameter
Section Thickness
######
##
80
80
0.6 mm
5 µm
Position A A A A A A A A
No. 1 2 3 4 5 6 7 8
Case No. xxxxx xxxxx xxxxx xxxxx xxxxx xxxxx xxxxx xxxxx
Age 50 58 60 54 52 45 62 53
Sex F F F F F F F F
Organ/Tissue Breast Breast Breast Breast Breast Breast Breast Breast
Diagnosis/Type Her2/neu + Her2/neu + Her2/neu − Her2/neu + Her2/neu − Her2/neu + Her2/neu + Her2/neu −
Fig. 4. An example of a format for a TMA block summary, which links the tissue core array position (columns 1 and 2) with the diagnostic details of the corresponding specimen (columns 3–7).
34
Fowler et al.
7. Superimposing the corresponding marked H&E-stained slide over the tissue block will assist in positioning the area to be sampled underneath the tissue punch. Move the H&E slide out of the way prior to taking the tissue core. 8. The number of tissue cores sampled per block depends on the tissue type. For example, normal breast epithelium, and other cancers with a high degree of regional heterogeneity will typically require >2 punches per block, while most other cancers are well represented with 1–2 punches (3). 9. TMA sections are most commonly captured onto a tape window and transferred to an adhesive-coated slide. The ParaffinTape Transfer system from Instrumedics minimizes tissue loss and floating of the tissue cores on the slide. It also eliminates the need for a water bath and a 60°C incubation step. Sections are immediately ready for deparaffinization. For examples of the tape-transfer system used in the literature, see Camp et al. (3), Kononen et al. (2), or Sauter et al. (17). If conventional microtome sectioning techniques are applied, section quality may be improved by heating the TMA block at 37°C for 30 min and allowing the block to cool to room temperature before sectioning. This process will soften the paraffin helping to anneal the tissue cores to the recipient paraffin block. 10. The first few TMA sections may have missing tissue cores or tears due to uneven seating of the tissue punches in the donor block. 11. Alternatively, undenatured probe may be applied to the pretreated slide (Subheading 3.5, step 10). Once the coverslip is sealed with rubber cement, the slide is incubated in a 74°C oven for 5 min, followed by 37°C for 6–24 h before proceeding to the next step (Subheading 3.5, step 12). An automated slide incubator, such as the ThermoBrite® System (Abbot Molecular, Inc.), may also be used for the slide drying and probe hybridization steps. References 1. Wan, W. H., Fortuna, M. B., and Furmanski, P. (1987) A rapid and efficient method for testing immunohistochemical reactivity of monoclonal antibodies against multiple tissue samples simultaneously. J Immunol Methods 103, 121–9. 2. Kononen, J., Bubendorf, L., Kallioniemi, A., Barlund, M., Schraml, P., Leighton, S., et al. (1998) Tissue microarrays for high-throughput molecular profiling of tumor specimens. Nat Med. 4, 844–7. 3. Camp, R. L., Charette, L. A., and Rimm, D. L. (2000) Validation of tissue microarray
technology in breast carcinoma. Lab Invest. 80, 1943–9. 4. Cryan, J., O’Grady, A., Allen, D., and Kay, E. (2009) Tissue microarray technology in the routine assessment of HER2 status in invasive breast cancer. Histopathology 54, 901. 5. Cunningham, R. E., Abbondanzo, S. L., Chu, W. S., Emory, T. S., Sobin, L. H., and O’Leary, T. J. (2001) Apoptosis, bcl-2 expression, and p53 expression in gastrointestinal stromal/ smooth muscle tumors. Appl Immunohistochem Mol Morphol. 9, 19–23.
Tissue Microarrays: Construction and Uses 6. Dhanasekaran, S. M., Barrette, T. R., Ghosh, D., Shah, R., Varambally, S., Kurachi, K., et al. (2001) Delineation of prognostic biomarkers in prostate cancer. Nature 412, 822–6. 7. Schlomm, T., Iwers, L., Kirstein, P., Jessen, B., Kollermann, J., Minner, S., et al. (2008) Clinical significance of p53 alterations in surgically treated prostate cancers. Mod Pathol. 21, 1371–8. 8. Fernandes, A. P., Capitanio, A., Selenius, M., Brodin, O., Rundlof, A. K., and Bjornstedt, M. (2009) Expression profiles of thioredoxin family proteins in human lung cancer tissue: correlation with proliferation and differentiation. Histopathology 55, 313–20. 9. Eguíluz, C., Viguera, E., Millán, L., and Pérez, J. (2006) Multitissue array review: a chronological description of tissue array techniques, applications and procedures. Pathol Res Pract. 202, 561–8. 10. Carlson, R. W., Moench, S. J., Hammond, M. E., Perez, E. A., Burstein, H. J., Allred, D. C., et al. (2006) HER2 testing in breast cancer: NCCN Task Force report and recommendations. J Natl Compr Canc Netw. 4 Suppl 3, S1–22. 11. Dandachi, N., Dietze, O., and HauserKronberger, C. (2002) Chromogenic in situ hybridization: a novel approach to a practical and sensitive method for the detection of HER2 oncogene in archival human breast carcinoma. Lab Invest. 82, 1007–14.
35
12. Graham, A. D., Faratian, D., Rae, F., and Thomas, J. S. (2008) Tissue microarray technology in the routine assessment of HER-2 status in invasive breast cancer: a prospective study of the use of immunohistochemistry and fluorescence in situ hybridization. Histopathology 52, 847–55. 13. Faratian, D., Graham, A., Rae, F., and Thomas, J. (2009) Rapid screening of tissue microarrays for Her-2 fluorescence in situ hybridization testing is an accurate, efficient and economic method of providing an entirely in situ hybridization-based Her-2 testing service. Histopathology 54, 428–32. 14. Camp, R. L., Chung, G. G., and Rimm, D. L. (2002) Automated subcellular localization and quantification of protein expression in tissue microarrays. Nat Med. 8, 1323–7. 15. Liu, C. L., Prapong, W., Natkunam, Y., Alizadeh, A., Montgomery, K., Gilks, C. B., et al. (2002) Software tools for high-throughput analysis and archiving of immunohistochemistry staining data obtained with tissue microarrays. Am J Pathol. 161, 1557–65. 16. Allen, T. C. (1992) Hematoxylin and Eosin. In E. B. Prophet, B. Mills, J. B. Arrington, and L. H. Sobin, ed. AIFP Laboratory Methods in Histotechnology, Armed Forces Institute of Pathology, American Registry of Pathology, Washington, DC, 53–58. 17. Sauter, G., Simon, R., and Hillan, K. (2003) Tissue microarrays in drug discovery. Nat Rev Drug Discov. 2, 962–72.
wwwwwwwwwwwwwww
Chapter 3 Standardization in Immunohistology Anthony S.-Y. Leong and Trishe Y.-M. Leong Abstract The rapid acceptance of immunohistology as an invaluable adjunct to morphologic diagnosis has been possible because of the development of new and more sensitive antibodies and detection systems that allow its application to formalin-fixed, paraffin-embedded tissue (FFPT). More importantly, antigenretrieval techniques have resulted in some degree of consistency allowing immunohistology to be used reliably as a diagnostic tool. The advent of prognostic and predictive biomarkers, and the desire for individualized therapy has resulted in mounting pressure to employ the immunohistological assay in a quantitative manner. While it was not a major issue when the technique was employed in a qualitative manner, the numerous variables in the preanalytical and analytical phases of the test procedure that influence the immunoexpression of proteins in FFPT become critical to standardization. Tissue fixation is pivotal to antigen preservation but exposure to fixative prior to accessioning by the laboratory is not controlled. Antigen retrieval, crucial in the analytical phase, continues to be employed in an empirical manner with the actual mechanism of action remaining elusive. There is great variation in reagents, methodology, and duration of tissue processing and immunostaining procedure, and the detection systems employed are not standardized between laboratories. While many of these variables are offset by the application of antigen retrieval, which enables the detection of a wide range of antigens in FFPT, the method itself is not standardized. This myriad of variables makes it inappropriate to provide meaningful comparisons of results obtained in different laboratories and even in the same laboratory, as in current practice, each specimen experiences different preanalytical variables. Furthermore, variables in interpretation exist and cutoff thresholds for positivity differ. Failure to recognize false-positive and false-negative stains leads to further errors of quantitative measurement. Many of the problems relating to the technology and interpretation of immunostaining originate from failure to recognize that this procedure is different from other histological stains and involves many more steps that cannot be monitored until the end result is attained. While several remedial measures can be suggested to address some of these problems, accurate and reproducible quantitative assessment of immunostains presently remains elusive as important variables that impact on antigen preservation in the paraffin-embedded biopsy cannot be standardized. Key words: Immunohistochemistry, Variables, Antibodies, Controls, Quantitation, Pitfalls, Validation
Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_3, © Springer Science+Business Media, LLC 2011
37
38
Leong and Leong
1. Introduction Fluorescence-based immunohistochemistry was first introduced in the mid-1940s but it was only with the development of the peroxidase–antiperoxidase and avidin–biotin peroxidase techniques that the procedure could be applied to formalin-fixed, paraffin-embedded tissues (FFPTs), facilitating its usefulness in tissue diagnosis. With the development of hybridoma technology, the technique rapidly became entrenched as an invaluable adjunct to morphologic diagnosis and through the introduction of antigenretrieval techniques; an ever-increasing range of diagnostic proteins could be identified in FFPT. The ability to work with FFPT, the most common medium in which histological diagnosis is rendered, and the increasing sensitivity of the technique enabled exquisite localization of staining to specific cell structures and organelles. As such, we proposed “immunohistology” as a more appropriate term for this morphologybased investigation in order to emphasize this fundamental attribute (1). Alternative appellations like “immunohistopathology” (2) and “immunomicroscopy” (3) were subsequently proffered in recognition of the importance of correlating morphologic features with the immunological assay. Through the identification of proteins expressed in the cell and connective tissues, the assay has been employed for a myriad of diagnostic and research purposes (4, 5) that include tumor diagnosis (1, 3, 6), identification of infective organisms, phenotyping of lymphomas and leukemias, identification of hormones and peptides, and demonstration of diagnostic morphological patterns and structures such as basal lamina for the identification of soft tissue tumors (7) and microvilli (8, 9) that are normally not visible in routinely stained sections. Immunohistological markers have been employed for prognostication in a wide range of neoplasms (10, 11). The immunohistological identification of cellular proteins serves as a surrogate technique for molecular analysis in the identification of genetic alterations, normal expression, overexpression, aberrant expression, and loss of expression of genes. For example, the absence of E-cadherin distinguishes infiltrating lobular breast carcinoma from infiltrating ductal carcinoma (12, 13), and the loss of mismatch repair gene proteins is useful for the screening of microsatellite instability (14, 15). There have even been proposals that some tumors such as gastrointestinal stromal tumors and lobular carcinomas of the breast be defined by their immunohistological phenotype (16, 17) and the assay can also be employed for the identification of carrier states. Antibodies have been raised to chimeric proteins that result from the translocation of certain genes that are recognized to be specific for certain tumors. For example, the immunoexpression of
Standardization in Immunohistology
39
the chimeric NPM–ALK protein that represents the fusion of the nucleophosmin (NPM) gene on chromosome 5q35 to the anaplastic lymphoma kinase (ALK) gene on chromosome 2p23 is diagnostic of anaplastic large cell lymphoma (18). When immunohistological assays are employed in this manner it is a cheaper and more rapid test that serves as a useful screening procedure for several genetic abnormalities. The foregoing applications of immunohistology are largely qualitative but individualized cancer treatment has resulted in attempts at quantitation of a number of immunohistochemical biomarkers. A contemporary trend in cancer therapy is the targeting of specific molecules expressed by cancer cells, commonly molecules that are involved in the regulation of growth and proliferation (19). These molecules include Her2/neu in breast cancer, CD117 in gastrointestinal stromal tumors, CD20 in lymphomas, CD33 in acute myeloid leukemia, epidermal growth factor receptor (EGFR) in colorectal carcinoma and nonsmall cell carcinoma of the lung and somatostatin receptor in neuroendocrine carcinomas of the pancreas (19). Humanized or chimeric monoclonal antibodies have been produced to these target molecules and the best therapeutic response generally occurs in those tumors expressing large amounts of the target molecules. As such, there has been an increasing push for the quantitation of such expressed molecules as detected by immunohistochemical stains. 1.1. The Immuno histological Assay: A Total Test Approach
Earlier in the development of immunohistology when the assay lacked sensitivity and specificity, the problems of reproducibility and consistency were very evident because many stains were capricious. With the advent of sensitive antibodies, enhanced detection systems, and antigen-retrieval procedures, it became possible to stain for a large variety of antigens in FFPT in a consistent manner so that the initial problems of reproducibility were forgotten or relegated to the background. However, the proliferation of reagents and procedures resulted in a wide variation in adopted practice and it was soon realized that this diversity could be a potential issue and standardization was required (20–23). The concern largely focused on reagents and procedures, without much attention paid to fixation and other variables that influence antigen preservation. The drive for quantitation of therapeutic and prognostic markers in a variety of tumors has resulted in a renewed necessity to address the problems of reproducibility and standardization. Immunohistology should not be regarded as an empirical procedure similar to other histological stains. It entails many more steps to perform and success cannot be monitored until completion of the entire procedure, akin to a biochemical assay. As such, this procedure is more appropriately viewed as a total test comprising preanalytical, analytical, and postanalytical phases.
40
Leong and Leong
Unfortunately, unlike biochemical assays that are largely standardized across most of the world, the same is not true for the immunohistological assay. To attain standardization of the immunohistochemical assay there must be a detailed understanding of the different phases of the entire test.
2. Materials 2.1. Preanalytical Variables: Tissue Preservation, Transportation, Fixation, and Tissue Processing
With biochemical assays, the blood sample is collected according to standardized protocols that dictate the quantity, site where the blood is drawn, state of the patient, fixative or preservative, if any, and the duration of storage before analysis. With the histological sample, however, the preanalytical variables are particularly difficult or currently impossible to control because tissue samples are obtained in different ways by different clinicians, and in diverse hospital and clinical settings. Tissue samples may be obtained as fine needle aspirates, needle cores, endoscopic biopsies, superficial biopsies of skin or mucosa, or by more extensive excisional procedures. The state of the biopsy, which may be intrinsically necrotic, or ischemic, and the duration between removal of the tissue and placement into fixative will determine the quantity of detectable antigen. Immediately after tissue death or removal of the biopsy sample, degradation and autolysis commence, and delays in fixation affect the preservation of tissue antigens. Some antigens appear to withstand adverse conditions, e.g., blood group antigens can still be immunohistochemically detected in minute spots of dried blood, but many other antigens do not. Degradation rates differ between organs and tissues (24) and biochemical changes occur very rapidly within seconds after death (25). Biopsy specimens should be immersed in formalin soon after removal. Large specimens such as mastectomies are often immersed whole in fixative, delaying fixation to the deeper tissues and adversely affecting antigen preservation.
2.2. Fixation and Fixatives
Fixatives vary in type and composition. Ten percent formalin (4% formaldehyde) is the universal fixative, producing morphological detail that pathologists are most accustomed. However, formalin penetrates tissues slowly and it is a commonly held belief that formalin preserves antigens poorly. Furthermore, formalin deteriorates with storage. Aqueous solutions of formaldehyde contain equilibrium of its monohydrate methylene glycol, formaldehyde, and water. Methylene glycol, in turn, forms various oligomers including low molecular weight polymeric hydrates or polyoxymethylene glycols. Formaldehyde appears to be an active
Standardization in Immunohistology
41
component, and effective fixation results from the formation of cross-links between protein end groups that include amino, imino, amido, peptide, guanidyl, hydroxyl, carboxyl, SH, and aromatic rings. The formation of such methylene bridges between amino groups on adjacent molecules is thought to be the basis of formaldehyde-based fixation (25). It has been clearly demonstrated that the ability to demonstrate a wide range of tissue antigens varies inversely with the duration of exposure to formaldehyde (26). Tissues fixed in formaldehyde displayed a distinct and progressive loss of staining of many antigens, frequently proportional to the duration of exposure to the fixative. There was an appreciable loss of staining of some antigens after 3 days, many antigens were lost after 7 days, and most were not demonstrable after 14 days of fixation. Interestingly, enzymatic predigestion per se was useful for the unmasking of only some antigens, ineffective and even deleterious in others (26). Unfortunately, formalin fixation is not a single uniform procedure but varies from laboratory to laboratory with respect to concentration, pH and buffers, and temperature and duration of exposure. All these parameters strongly impact on the outcome of immunostaining. Furthermore, exposure to formalin occurs in several stages, after removal of the specimen and before receipt by the laboratory, during the interim period in the laboratory before tissue processing, and again during tissue processing. All these durations vary between laboratories, and for different specimens accessioned by the same laboratory. It is also likely that antigens have different optimal durations of fixation, e.g., immunohistological demonstration of estrogen receptor requires a minimum of 6–8 h in formalin for consistent results (27). Capricious epitopes of antigens that cannot be demonstrated following routine formalin fixation can be successfully stained if exposed to the fixative for only a short duration (28), or by using a physical fixation agent such as microwaves (MWs) (29, 30). Antigen preservation in tissues fixed by MW irradiation in normal saline was clearly superior to formalin fixation and some antigens not demonstrable in formalin-fixed tissues were readily labeled in MW-fixed sections (31). Similarly, immunolabeling of cytological preparations has been shown to be optimal with 0.1% formal saline as the fixative and air-drying as the best method of preparation of such material (32). There are a variety of fixation and tissue preparation methods for immunohistochemistry (33) but immersion in 10% buffered formalin is the most common method of routine fixation, with coagulant fixatives less frequently employed (34). Alcoholbased fixatives appear not to react covalently with amino acids and therefore leave the primary protein structure unaltered although aspects of the tertiary structure may be changed (25).
42
Leong and Leong
Methacarn (a mixture of methanol and acetic acid) (35) and a variety of other fixatives including mixtures of alcohol with other reagents have been used (33), but shrinkage of tissue and extraction of small antigens remain a problem with alcoholic fixatives so that formalin has remained the universal fixative. The dehydration process may also affect antigens and various special dehydration procedures have claimed improved tissue antigenicity. Similarly, certain types of paraffin, celloidin, and polyethylene glycol have been shown to be useful for immunohistology (33). Other aspects of the fixation procedure, including pH and temperature, can impact on antigen preservation, more severely for some antigens compared to others. The pH of formalin affects the type of cross-linking that occurs. With 10% neutralbuffered formalin, hydrogen sites in peptide molecules are available for linkage as they are in the uncharged state. Lowering of the pH induces formation of charged amino groups (NH3+) that lack reactive hydrogen sites and favors interactions with amide (CO–NH2) groups. The formation of methylol bridges can theoretically alter the native conformation of a protein substantially and the configuration of covalent cross-links may alter the structure of important epitopes (34). In addition to inducing crosslinking, formalin also disrupts hydrogen bonds and other electrostatic interactions that affect the configuration of proteins, further increasing the possibility of important alterations to epitopic targets. As antibodies are mostly raised to proteins in their native conformation state, they may not bind as effectively to target polypeptides that have such structural transmogrification, the so-called “antigen masking.” The extent of this antigen masking is proportional to the duration of fixation (26, 34). The antigen masking effects of formalin and concerns about its toxicity has prompted the use of formalin substitutes, which are generally either alcohol- or water-based. Several proprietary fixatives are available, including Histochoice (Amresco, Solon, OH, USA), FineFix (Milestone, Bergamo, Italy), NoToX (Earth Sate Technologies, Lumberton, NC, USA), and Ominfix (An-Con Genetics Inc., Melville, NY, USA), but their efficacy in antigen preservation remains to be proven particularly as alcohol-based fixative have been shown to be detrimental to some predictive and prognostic factors such as hormone receptors and HER2/neu. Fresh acetone-fixed tissue sections have traditionally been held as the “gold standard” for reference purposes in immunohistochemistry as it was assumed that fresh tissue had unadulterated nuclear, membranous, and cytoplasmic epitopes, which, when exposed to fixative, were altered or lost forever. However, it was recently shown that more than half of the 26 antibodies tested showed better immunohistological signals following fixation in
Standardization in Immunohistology
43
neutral-buffered formalin and antigen retrieval with only two antibodies displaying better results in fresh acetone-fixed sections (36), in particular, nuclear antigens showed better staining in FFPT. Neutral-buffered formalin thus performs well; not only as a convenient and cheap universal fixative for cytomorphology but also for antigen preservation and a better alternative is yet to be found. Heat-induced antigen retrieval appeared to improve immunohistochemical staining of unfixed frozen sections and dot-blot protein extracts suggesting that natural steric barriers exist even in the fresh state (37).
3. Methods 3.1. T issue Processing
While it is possible to control the duration of exposure to fixative after accessioning by the laboratory, it is difficult or presently impossible to control the events preceding. Furthermore, the exact duration between excision of tissue and placement in formalin is invariably not known, this factor being a major obstacle to standardization of fixation. There is another period of exposure to formalin when the specimen is in transit in the laboratory awaiting examination, dissection, and sampling before final tissue processing. This duration varies between different specimens accessioned in the same laboratory. There is also formalin in the tissue processor, and tissue-processing cycles can vary considerably in reagents and duration, with newer tissue processors employing chemicals different to the conventional formalin, ethanol, and xylene (38, 39). There are differences in tissue-processing protocols with variations in exposure to formalin and alcohols. These variables can be standardized for the individual laboratory but in the case of tissue blocks prepared in other laboratories, such variations may significantly affect the staining of tissue antigens. Tissues previously subjected to freezing or decalcification show poor and inconsistent preservation of some antigens.
3.2. Storage of Tissue Sections
Exposure to the elements can affect tissue antigens. Tissue sections that are left on the bench at room temperature have shown deterioration of antigenicity (40–42) and stored tissue blocks displayed degradation of target antigens (43). Heat, drying, and exposure to ultra-violet light are significant factors contributing to this loss (43). It is a common practice to have control sections cut well in advance of immunostaining and unless optimally stored, such controls will affect the optimization of new antibodies and other reagents. FFPT sections should be wrapped in aluminum foil and stored at −20°C for optimal antigen preservation.
44
Leong and Leong
3.3. Analytical Variables: Antigen Retrieval, Antibodies, Detection Systems, and Controls
Unlike the preanalytical factors, analytical factors are more readily controlled within the individual laboratory. Such analytical variables include antibody specificity and sensitivity, dilution, detection system, and antigen retrieval. The immunohistological test is the application of an antibody to label a protein in fixed or unfixed tissues or cells based on antibody–antigen recognition. Its basic aim is the distinct localization of signal to specific cell and tissue components, at the same time retaining good morphological visualization. Amplifying the signal without increasing the nonspecific background staining or noise is a major strategy to allow application to FFPT. The direct immunofluorescence technique is a simple one-step procedure but the insensitivity of its 1:1 antigen-to-signal ratio, and the poor morphologic visualization and preservation of frozen sections restricted its application. The development of two- and three-step techniques to amplify the antigen-to-signal ratio, and permanent chromogen systems such as horseradish peroxidase and diaminobenzidine (DAB) allowed the assay to be applied to FFPT (1, 44). Additional developments such as the use of polymers (45) and tyramide (46) further amplified the antigen-to-signal ratios, and antigen-retrieval methods including enzyme digestion and postfixation in heavy metal solutions improved the detection of antigens in fixed tissues (1, 34). However, it was the introduction of the so-called heat-induced antigen-retrieval method that had the greatest impact (29, 47, 48). Pari passu with these developments, the range of antibodies and their specificity and sensitivity continued to improve, all contributing to consolidate the pivotal role of the immunohistological assay in both research and diagnosis.
3.4. Substrate and Chromogen Systems
Visualization of antibody molecules is done with a variety of labels including fluorescent compounds and their active enzymes that have the property of inducing the formation of a colored reaction product from a suitable substrate system for visualization. A number of chromogenic systems are available. They include DAB, 3-amino-9-ethyl-carbazole (AEC), Hanker–Yates reagent, alpha-naphthol pyronin used with peroxidase as substrate; fast blue, fast red, 5-bromo-4-chloro-3-indolyl phosphate (BCIP)nitroblue tetrazolium (NBT) used with alkaline phosphatase as substrate; tetrazolium, tetranitroblue tetrazolium used with glucose oxidase as substrate; and immunogold with silver enhancement (1, 44). Their use varies between laboratories. The horseradish peroxidase–DAB system is the most widely favored. Osmification can produce a more intense dark brown-black color and a similar effect is achieved by posttreatment with nickel sulfate or cobalt chloride.
3.5. A ntigen Retrieval
One of the earliest methods of antigen retrieval was proteolytic digestion employed prior to the application of the primary antibodies. A number of proteolytic enzymes served this purpose,
Standardization in Immunohistology
45
including trypsin, proteinase K, pronase, pepsin, ficin, DNase, and others (1, 44, 49). Not only the enzymes used are different, but there is also variation in concentration, duration, and temperature of digestion. Furthermore, not all antigens benefit from proteolytic digestion, and some show deleterious effects with loss of staining (26). Inappropriate protocols result in tissue breakdown and loss of morphology with high levels of background and false-positive staining. Heat-induced antigen retrieval was a major milestone, greatly enhancing the ability to demonstrate antigens in FFPT (29, 30, 47, 48). The initial technique was achieved with MWs, which has remained the most convenient, but a variety of other methods of generating heat have since been spawned, including water baths, hot plates, wet autoclaves, pressure cookers, and vegetable steamers. Shi et al. (47) described MW heating of FFPT in the presence of heavy metal solutions such as lead thiocyanate, up to temperatures of 100°C to “unmask” a wide variety of antigens for immunostaining. It was subsequently shown that MW irradiation of deparaffinized–rehydrated sections in 10 mmol citrate buffer pH 6.0 produced, with few exceptions, increased intensity and extent of immunostaining of a wide variety of tissue antigens (29, 47, 48). The use of citrate buffer eliminated the need to employ heavy metal solutions which, when heated, generate toxic fumes. Several commercial antigen-retrieval reagents are available but they mostly do not produce any better results than that obtained with citrate buffer (50). All methods of heat generation listed above suffer from problems with accurate temperature and time control. These two variables have been shown to be critical to the process of heat-induced antigen retrieval. They are inversely related so that antigen retrieval at lower temperatures requires longer durations to achieve the same results as that obtained with higher temperatures (51). The time taken to attain the desired temperature from variable starting temperatures, time required to cool to room temperature, and actual temperature attained are variables that cannot be controlled with most methods of heating. Furthermore, there is a problem of unevenness of heating within microwave oven cavities, making the entire process impossible to standardize with inconsistencies in methodology. Computerized control of time and temperature now available with some commercial MW instruments takes the guesswork out of heat-induced antigen retrieval. Accurate time and temperature control not only produces superior antigen retrieval across the spectrum of diagnostic antigens, but accurate heating to 120°C or “superheating” has also proven to produce notably better antigen staining (52). Our understanding of the effects of formaldehyde on proteins dates back to work from the 1940s (53–56). The amino acid side chain of proteins includes many groups that may react with
46
Leong and Leong
aldehydes that contribute to the stabilization of proteins. However, there is no clear consensus as to which is the predominant molecular species resulting from this method of fixation. There is no doubt that some of the cross-linked adducts are very stable and remain irreversibly changed even after extensive washing, while others revert under varying conditions to free formaldehyde and the amino acid (25). Without a complete understanding of the actions of formaldehyde on proteins, it is not surprising that we do not fully understand the mechanisms of antigen retrieval. Heat appears to be a common denominator in antigen retrieval produced by a variety of methods including MWs. Heat is hypothesized to cause protein denaturation based on the observation that some antigens or endogenous enzymatic activities may be lost after heat treatment (57). Heat also induces reversal of various chemical modifications of the protein structure that result from formalin fixation. Additional actions that produce antigen retrieval include the loosening or breakage of the cross-linkages caused by formalin fixation, hydrolysis of Schiff bases, and multiple other pathways, including extraction of diffusible blocking proteins, precipitation of proteins, and dehydration, all of which allow better penetration of antibody and increased accessibility to epitopes (58). All or some of these mechanisms may be achieved by other methods of retrieval including enzyme digestion and changes in pH. MW energy may itself mobilize the last traces of paraffin that may not have been extractable by standard techniques, thereby improving antibody penetration (47). Another hypothesis to explain antigen retrieval is the masking by calcium complex formation that occurs with formalin fixation. The release of calcium from this cage-like complex may require a considerable amount of energy such as high-temperature heating or calcium chelation by citrate (59). While this mechanism may be operative in some, it is not sufficient to explain the loss of immunoreactivity for many other antigens (56). The role of kinetics in antigen retrieval is also not known. While the focus has been on heat as the responsible factor in MW retrieval, the rapidly oscillating electromagnetic field of MWs may itself have an effect. While heat or thermal energy will increase molecular kinetics and hasten chemical reactions, the rapid rotation of molecules directly induced by the MWs will give rise to greatly increased collision of molecules, in turn accelerating chemical reactions. The heat generated may represent only an epiphenomenon secondary to the kinetics. One study that examined MW stimulation of CEA/antiCEA reaction in an enzyme-linked immunosorbent assay system found that despite continuous cooling by ice, MW stimulation increased reaction rates by a factor of 1,000, allowing the investigators to conclude that such rate increases were far too large to be explained solely by the modest increase in temperature (60). Another study further elucidated the existence of a
Standardization in Immunohistology
47
“microwave effect” (61). The rate of droplet temperature increase obtained in a thermal cycler is similar to that achieved by MW irradiation. However, the immunostaining obtained from 3-min incubation at 37°C in a thermal cycler followed by 2-min incubation without heating was much weaker than that obtained with MWs. Similarly, it was demonstrated that 7-s MW irradiation followed by 5-min room temperature incubation for each step of the avidin–biotin peroxidase complex procedure produced good immunolabeling (62). The droplet temperature rose no more than 5°C following the 7-s irradiation at 100% power in an 850-W oven so that the temperature was deemed not to be a significant component of the accelerated reaction (61). Others have argued that there is no significant MW effect and the accelerated reactions are a function of heat. It has been concluded that MW irradiation did not produce cleavage or polymerization of proteins and irradiation resulted in an electrophoretic pattern that was similar to that obtained when lysozyme and hemoglobin was heated in formaldehyde to 60°C for 30 min (63). Interestingly, results to the contrary have been shown in a study of S-adenosylhomocysteine hydrolase and 5¢-methylthioadenosine phosphorylase, two thermophilic and thermostable enzymes, where exposure to MWs caused a nonthermal, irreversible, and time-dependent inactivation of both enzymes (64). In a model immunostaining system using short synthetic peptides to mimic the antibody-binding site of common diagnostic protein targets, Sompuram et al. (65) found that not all of the peptides studied exhibited the formalin-fixation and antigenretrieval phenomenon. One group of peptides was recognized by antibody even after prolonged exposure to formalin while another group exhibited the formalin-fixation and antigen-retrieval phenomenon only after another irrelevant protein was mixed with the peptide before fixation. Amino acid sequence analysis indicated that fixation and antigen retrieval were associated with a tyrosine in or near the antibody-binding site bound covalently to a nearby arginine implicating the Mannich reaction as an important factor in the process. These findings concurred with those of Fraenkel-Conrat et al. (53–55) who had indicated that of all the protein cross-linking reactions that occur as a result of formalin fixation, the Mannich reaction is different, in that the cross-linkages can be hydrolyzed with heat or alkaline treatment. Antibodies appear to recognize linear protein epitopes in FFPT and antigen retrieval may simply remove cross-linked proteins that are sterically interfering with antibody binding (65, 66). The recent demonstration that antigen retrieval produces immunohistological staining results in FFPTs that are comparable or better than that in acetone-fixed fresh frozen section (36) and that heat-induced antigen retrieval enhances immunostaining in unfixed fresh frozen sections, and dot-blot protein extracts (37) further support
48
Leong and Leong
the concept that intrinsic natural steric barriers exist and interfere with antibody binding. The demonstration that MWs can also be employed to enhance the demonstration of HER2/neu gene in chromogenic in situ hybridization (CISH) is particularly interesting and it suggests that similar mechanisms may be operative in the “masking” of DNA (67). While these observations provide some insights into the action of antigen retrieval with some peptides the answers to the majority still remain unknown. The demonstration that ultrasound (68) can significantly increase antibody–antigen reaction in immunostaining lends further support to the relevance of molecular movement as an accelerant of the chemical reaction as the heat generated by this physical modality is negligible (69). A number of other hypothetical physical mechanisms may also play a role in the actions of MWs and include alteration of the integrity of noncovalent secondary bonding, including hydrophobic interactions, hydrogen bonds, and van der Waal’s interactions that make up the precise steric interactions at the cell membrane. The combination of heat retrieval with enzymatic digestion allowed enhanced demonstration and localization of a number of antigens including the immunoglobulins (70). A number of enzymes can be used, at varying concentrations and for varying durations. It can precede or follow heat-induced antigen retrieval with different results. For optimal outcomes, it is necessary to explore all possible combinations and permutations of these variables including antibody dilutions with the realization that excesses result in increased background, and loss of antigen and cell morphology. The chemical composition of the retrieval solution may affect the efficacy of the retrieval and a number of solutions have been advocated including citrate buffer, Tris buffer, glycine– hydrochloric acid, EDTA, urea, heavy metal solutions, and proprietary reagents. The molarity of the solution may also influence immunostaining (71). The pH of the retrieval solution is one of the most important factors. Three patterns of staining reflect the influence of pH. Some antigens [CD20, AE1, EMA, NSE, and proliferating cell nuclear antigen (PCNA)] showed no variation at pH values ranging from 1.0 to 10.0, other antigens (MIB1, ER) displayed a dramatic decrease in staining intensity at middle pH values (pH 3.0–6.0) with strong staining above and below the range, and a third pattern was demonstrated by other antigens (CD43, HMB45) which were weakly stained at low pH (1.0– 2.0) and demonstrated a sharp rise in intensity with increasing pH (72). MWs have been applied between sequential rounds of a threelayer immunoenzyme staining (mouse Mab, goat antimouse IgG, and mouse PAP or mouse APAAP) and color development technique for multiple antigen detection (73). MWs denatured bound
Standardization in Immunohistology
49
antibody molecules resulting in the blocking of cross-reactivity between the sequential staining steps, allowing the use of primary and other antibodies raised in the same species. Besides serving a role in antigen retrieval, MWs also inactivated peroxidase and alkaline phosphatase enzymes present in PAP and APAAP complexes, which would otherwise have led to inappropriate color development (73, 74). 3.6. Antibodies and Detection Systems
While the ever-increasing range of antibodies is a boon to diagnostic immunohistology, it can also be confusing, as specificity and sensitivity of each reagent differs. There is also an increasing choice of antibodies that are purported to detect the same antigen but, in fact, detect different epitopes, albeit of the same antigen. For example, in the case of two antibodies to the estrogen receptor (ER), H222 labels ER only in frozen sections and not in FFPT, whereas, ID5 displays the reverse properties, suggesting that they detect different epitopes on the same antigen. The sensitivity of different antibody clones directed to the same antigen may also vary significantly. Two antibodies to PCNA, namely, 19A2 and PC10, show vastly different sensitivities (75), and proliferation indices obtained with these two markers are significantly different to other markers of cycling cells such as MIB1, Ki-67, KiS1, and KiS5 (76). While the majority of monoclonal antibodies are from mice, recently produced rabbit monoclonal antibodies to the same antigens have been shown to have greater sensitivity with staining results affecting clinical outcomes (19). The manufacturer’s antibody concentration and methodology serve as useful guides but each new antibody has to be carefully optimized as the vagaries of preanalytical and analytical variables differ between laboratories. Antibodies produce a positive reaction over a range of concentrations and selection of the optimal concentration is largely one of individual choice. Many laboratories favor an intense dark brown to almost black DAB reaction product while others prefer a golden brown color, which does not obscure the cytomorphological features, an all-important attribute in immunohistology. Antibodies should be stored at the appropriate temperature. After aliquoting working quantities of primary antibody for dilution, the concentrate should be stored at –20°C or preferably –70°C. Antibody concentrates stored at the latter temperature can remain effective indefinitely despite their “use by date.” When stored at 4°C, antibodies have a shorter shelf life. The process of antibody optimization also requires that different temperatures of antigen retrieval be explored, and although citrate buffer, 10 mmol/l is a good universal retrieval solution, a higher pH may be required for more capricious antigens (72). In addition, the synergistic action of proteolytic digestion should be routinely explored as a pre-MW and post-MW procedure.
50
Leong and Leong
Antigen detection systems vary significantly in sensitivity. The ability to detect small amounts of antigen not detectable by less sensitive techniques requires that we must continually adjust our diagnostic criteria to incorporate new immunohistological information. In some situations, the markedly increased sensitivity of the detection system can result in a high background and accurate localization of the signal can be lost. This is demonstrated in the staining for C4d where localization to peritubular capillaries is not discernable with ultrasensitive polymer detection systems such as the EPOS (DakoCytomation, Sydney, Australia) because of extension of staining beyond the capillaries into the intertubular interstitium. 3.7. Automated Immunostaining
With the adoption of immunohistochemistry as an integral component of morphologic diagnosis, there has been a proliferation of automated immunostaining devices. These devices serve to produce timely results and consistency of staining within the individual laboratories. Importantly, they perform many slow and repetitive steps of application of reagents and antibodies, monitoring of incubation times, and washing and wiping of slides after each step that are otherwise operator-dependent in the manual procedure and may be prone to error. While automation has been embraced with enthusiasm, it comes with its own price in that each device employs a different system, especially in the antibody concentration, duration of incubation, and method of antibody incubation. Capillary gap stainers employ surface tension for the antibodies and reagents to ascend to immerse the tissue section, whereas other devices “blow” a fixed aliquot of reagent over the tissue section or drop the reagent over the section. Some devices are “closed” systems that require specific proprietary reagents, whereas others are “open” and can be used with reagents from other sources although the detection system remains fixed. Importantly, some systems employ polymer detection systems, whereas the majority of systems use variations of the conventional avidin–biotin peroxidase system. The temperature at which incubation is performed can vary with each device. Thus, automated immunostaining devices introduce a further set of variables that will differ between laboratories.
3.8. C ontrols
As with all laboratory procedures positive and negative controls must be employed. Negative controls take the form of tissues that are known not to contain the antigen of interest. Another form of negative control includes the substitution of the primary antibody with antibody diluent, nonimmune immunoglobulin, or an antibody of irrelevant specificity derived from the same species and at the same dilution. It was thought that absorbing the primary antibody with highly purified protein or the peptide antigen employed to
Standardization in Immunohistology
51
enerate the primary antibody produced an ideal negative control. g This eliminated the binding of the antibody to the protein in the section. However, it has been shown that the absorption control may not bind to the same protein that was used to generate the primary antibody. Furthermore, the latter may recognize a similar epitope of unrelated protein, especially after tissue fixation (77, 78). The most appropriate control for any immunostain would be an internal control because it would have been subjected to identical preanalytical and analytical variables as the test tissue. Such controls are invariably nonlesional or benign cells. However, they express the antigen of interest at levels different to the tumor cells and are thus not ideal. Nonetheless, they are currently the best controls available. The alternative, an external control of similar tumor tissue known to express the antigen of interest, would have been subjected to an entirely different set of preanalytical variables. It is inappropriate to use benign tissue as external controls when examining tumor cells. Reference standards for quality control of reagents and tests in the clinical laboratory are well established. Such standards can be obtained from pooled serum but the development of reference standard controls for immunohistology is subjected to many more obstacles. Unlike serum samples, pathological tissues cannot be pooled and their supply is not limitless. Furthermore, morphologically similar tumors are not necessarily antigenically identical. The use of multitissue blocks provides a solution to some of these problems. Multitissue blocks are prepared by binding together many slivers of a wide range of different tumors to serve as both positive and negative controls. Similar blocks can also be made of nontumor tissue, however, but all such blocks contain a limited amount of material. When testing for tumor antigen, it is more appropriate to employ controls from tumor tissue as the level of antigen is more likely to correspond to that in the test section. Microtissue arrays are a possible solution to the limited supply of control tissue. Microarray blocks allow the incorporation of 200–300 fine tissue cores that can be used for controls against a wide variety of antibodies and as the cores are small (0.5–1.5 mm diameter) much of the original tissue block remains preserved. Microtissue arrays should be used with the recognition that each core of tissue has been subjected to different fixation and processing so that the level of antigen preservation in each of the 200– 300 tissue samples are different and by no means standardized. Recognition of this deficiency in controls led to the development of the “Quicgel” control which was an artificial tissue control block using a breast cancer cell line which was added to the tissue cassette containing the test sample (79). This method requires the availability of suitable cell lines expressing the antigen in question, which needs to display consistent behavior under
52
Leong and Leong
cell culture and storage. An extension of the “Quicgel” method is “histoids” in which three-dimensional pellets are grown in centrifugal cell culture to produce, in theory, an unlimited supply of “faux tissue” controls. If three or more cell lines are cocultured, the faux tissue can be employed as controls for many of the commonly employed antibodies including controls for fixation and processing. Alternatively, it may be possible to develop preparations of purified protein that can be diluted to produce a series of known reference standards for both Western blotting and immunohistochemistry (80). In quantitative procedures, a validated control expressing the range of scores is included alongside the test section so that it is subjected to the identical staining procedure. Closer examination of this common practice reveals that such “controls” are not optimal simply because they have been subjected to different preanalytical variables, e.g., fixation and tissue processing may be quite different to the test tissue. As such, titrating the staining procedure to such “controls” can be misleading and inappropriate. Ideally, both controls and test must be subjected to identical preanalytical and analytical conditions so that the state of antigen preservation in both tissues is identical. The question of appropriate controls thus has not been satisfactorily answered at this time.
4. Notes 4.1. Test Validation
As the immunohistological assay becomes employed as a prognostic as well as predictive tool in cancer, there is a gradual realization that the test requires careful validation. All too often, an antibody is purchased that is claimed by the vendor to be specific for a cancer and this is run against a few such cancers that are high expressors of the protein. Titration and optimization of the antibody may be performed, although even these procedures may not be carried out and the manufacturers’ recommendations are simply followed without regard for the variations in preanalytical and analytical factors. Such testing is not the same as validation, which requires that the testing be done against a large number of both positive and negative specimens. While all the criteria for clinical validation that include a definitive clinical study with a sample size that is adequate for statistical analysis, methodological validation, and optimized cutoff value cannot be adhered to, immunohistological prognostic and predictive markers must reliably predict outcomes or response to treatment in the patient samples used for validation (81). Indeed, the arguments for validation of prognostic and predictive markers can also be extended to all other diagnostic markers as identification and specific
Standardization in Immunohistology
53
typing of a neoplasm can rise to the level of a predictive test as specific chemotherapeutic agents and regimens have been developed to many tumors and treatment becomes highly individualized. This requirement for appropriate validation was first imposed in the recent guidelines for HER2 testing in breast cancer (82). Both technical and clinical validation should be performed for the immunohistological assay. Ideally, a valid method should show substantial equivalence with protein expression or a clinically relevant surrogate or methodological identity to the original clinical validation assay. This is easier said than done. Validation against the original clinical validation assay is difficult or even impossible and the alternative is validation against a recognized “gold standard.” In the case of HER2, there is in reality no such “gold standard,” neither fluorescence in situ hybridization (FISH) nor immunohistology is accurate in predicting outcome in 100% of cases (83, 84) and we have to settle for substantial methodological or analytical equivalence to the original study using a validated sample set or cross-validation with an alternative validated method, in this case either FISH or immunohistology. Another method of validation is through interlaboratory comparison or comparison with a reference laboratory whose testing has been validated. In the latter situation, the question arises as to “what constitutes a reference laboratory?” Even large volume, the so-called central laboratories, can fall short of required standards. Concordance between two such laboratories for FISH was 92% and between FISH and IHC was 82% (85), figures that fall short of the 95% recommended by the recent HER2 Testing Guidelines (83). When results from a peripheral laboratory were compared with those obtained in a central reference laboratory the concordance for FISH and IHC were 88.1 and 81.6% respectively (86). Reproducibility and concordance between laboratories are clearly a problem that extends to other prognostic and predictive markers as exemplified by interlaboratory comparisons of estrogen and progesterone receptor assays where reliable assays were found in only 36% of participating laboratories in Europe (87). It has been suggested that validation exercises may not be necessary if a validated commercial method is employed and the vendor’s protocol is carefully followed. This clearly is not true as preanalytical factors are so different between laboratories that standardization of the staining method is no guarantee of uniformity of sensitivity. Another method of validation is the use of standard samples from an approved source but such sources and tissue are not currently available. The use of consensus positive and negative tissue in the form of tissue microarrays is a possible substitute (88). Alternatively, tissue samples from cases accessioned by your own
54
Leong and Leong
laboratory known to harbor the target protein by non-IHC means can be used, but in all these situations it has to be remembered that there is no fixation or processing standard so that agreement between laboratories and between samples is subject to preanalytical variables discussed. Clearly, the ideal validation procedure would be against patient outcome but this is a costly exercise and often not practical as they require appropriate numbers of patients and a prospective study. The recent HER2 testing guidelines suggest 25–100 samples as sufficient numbers for validation. However, 25 samples may not be sufficient to achieve concordance between laboratories or methods in any validation exercise as the probability of making the 95% concordance standard is significantly less than 0.5 according to the Table A8 of the Guidelines (83). The setting of rigorous performance standards in prognostic and predictive testing (with potential extension to all IHC staining) will have consequences. If proficiency is regulated, those laboratories not meeting the required 95% concordance benchmark may have to cease testing in the USA and elsewhere where proficiency is regulated. This raises again the previous question of “what is a validated method?” Currently, there is no objective arbiter of what is valid. 4.2. Postanalytical Phase
While the postanalytical phase also includes generation and delivery of the results/reports, these aspects are not relevant to the present discussion. Some of the problems encountered in this area result from the lack of well-defined standards of what constitutes a positive result and if there are grades or degrees of positivity. There is no consensus of what is an adequate threshold or cutoff. What percentage of cells displaying immunoexpression is required to for a lesion to be positive? Undoubtedly, this figure varies among observers and is very much influenced by the quality and sensitivity of the staining procedure in different laboratories. If cytomorphologically atypical cells that correspond to the tumor population express the antigen, then no more than a few definitely stained tumor cells are necessary for a positive result. Often, large numbers of positive tumor cells can be demonstrated in this situation by simply increasing the primary antibody concentration or increasing the duration and/or temperature of the antigenretrieval process. It should be borne in mind that when the cutoff is set at 10% the highest degree of interobserver concordance is achieved. When set above this level concordance falls to unacceptable levels (see below). In a way, the selection of a cutoff level for any immunostain represents an excursion into some form of quantitation. Pathologists often report immunohistological stains in some semiquantitative manner by grading the intensity of staining as
Standardization in Immunohistology
55
“negative,” “weak,” “moderate,” or “strong”; or numerically as “0,” “1+,” “2+,” or “3+,” and would assess the extent of staining in the tumor as “focal” or “extensive,” sometimes based on a percentage of tumor cell staining. Such methods have not been standardized. Cutoff values, intensities, and extent of staining carry little relevance when immunohistology is employed in a qualitative manner and the question asked “is the tumor positive?” (for a specific antigen) and not “how positive is the tumor?” However, standardization becomes immensely relevant when some quantitative value has to be assigned to the results. Problems associated with quantitative immunohistology have been discussed elsewhere (89). Essentially, because it is virtually impossible to control or standardize the preanalytical factors that influence the preservation of antigens in FFPT, quantitation becomes farcical. Besides counting the number of positive cells, many scoring methods include an assessment of intensity, often on a three- or four-tiered scale. Such methods of colorific scoring are done visually by eyeballing and without calibration. There is a lesson to be learnt from static cytometric analysis of FFPT for tumor ploidy. The technique lost popularity because of conflicting results. A 5-mm section generally will not include entire nuclei, many being only partially cut. If a nucleus is not sectioned through its center, there is less nuclear material and hence less of the nuclear antigen of interest. Staining for the antigen will therefore reveal varying intensities, not because of variation in expression, but a reflection of differing thickness of nuclei in the section. Unless, this inequity is corrected, colorimetric scoring is not valid (90). In the ideal situation, clinicopathological validation for each laboratory’s staining procedure and results should be conducted but even so, it has been argued that such retrospective studies are usually conducted on selected material from one institution where fixation and tissue processing are relatively uniform and staining is batched to a single run, further minimizing variation. This is very different to the situation in routine diagnostic practice. Simple variation in chromogen incubation times can produce vast differences in stain intensity that can severely affect visual quantitation. Indeed, one acknowledged expert has observed that even with automated immunostaining there can be a “daily variation in optical density of as much as 30% for estrogen receptor, when the same block of tissue was used as a daily control” (91). There have been recommendations that quantitative immunohistology be conducted only after validation procedures are carried out by comparing results with those achieved by another technology such as FISH, or by comparing with an external control such as cell lines calibrated by immunohistology or FISH. Even if reference cell lines are stained alongside the test material, it must be remembered that the ability to stain the
56
Leong and Leong
c ontrol tissues appropriately only serves to control the staining procedure. Only when controls and test materials are subjected to the same preanalytical and analytical variables are they truly comparable. Until we are able to ensure this, results obtained by visual quantitation of immunohistological stains are, at best, approximations (92). Other variables that are relevant to quantitative immunohistology include the area of the section to assess: random versus peripheral versus central versus invasive tumor versus in situ tumor. With markers such as cell proliferation, should areas of highest activity be counted or should counts be conducted randomly? With counts for vessels, should values be based on the relative area (hence volume) occupied by the vessels or density? Interpretation of positive staining can be particularly difficult in the case of some nuclear antigens such as Ki-67, p53, and ER due to nucleolar staining, resulting in significant interobserver variation. While stain enhancement techniques using metallic ions or organic compounds have been employed to assist visualization, many only change the color of the chromogen and do not truly enhance sensitivity. 4.3. False-Positive, False-Negative, Cross-Reactivity, and Aberrant Expression
Familiarity with the characteristics and specificity of the antibody used will avert incorrect interpretation. Antigens often show specific organelle localization. For example, cyclin D1 is a nuclear antigen and cytoplasmic or membrane staining should not be read as a positive result. On the other hand, the ALK protein is commonly located in the nucleus and cytoplasm of anaplastic large cell lymphoma as a result of t(2;5). However, in as many as 28% of cases where there is a variant translocation, staining is confined to the cytoplasm and/or cell membrane. This cytoplasmic and membrane localization should not be interpreted as a false-positive reaction as it represents the true distribution of a variant form of the ALK fusion protein (18). A large number of cell types and their corresponding tumors may display positive staining for proteins that are not anticipated. In this context, the staining is called false-positive, aberrant, or cross-reactive and can potentially lead to incorrect diagnosis. Cross-reactivity may occur with some antibodies. Polyclonal CEA cross reacts with nonspecific cross-reactive antigens in granulocytes, whereas the monoclonal version of this antibody does not. Some antibodies may cross react with common epitopes on different intermediate filaments such as cytokeratin antibodies with GFAP-expressing glial cells. When cells express proteins that are not expected, the phenomenon has been described as “aberrant” expression. In many cases, the expression has been shown to be true expression by molecular analysis, such as the staining of cytokeratin in some mesenchymal cells and their corresponding tumors, such as leiomyosarcoma, rhabdomyosarcoma, and angiosarcoma.
Standardization in Immunohistology
57
True false-positive staining may occur in a variety of situations. In the assessment of immunostains, an “edge effect” may occur whereby cells in the periphery of the tissue section stain more strongly than the rest of the section. Often this is seen as a distinct peripheral band and may reflect better fixation of tissue in the periphery of the tissue block. It may also be the result of reagents seeping beneath the section at the edges so that both surfaces of the section are stained. In any event, this edge effect is more often than not true staining, albeit, enhanced in comparison to the rest of the section. Cracks and spaces in the tissue section can show nonspecific entrapment of reagents producing false-positive staining of cells alongside such spaces. Necrotic and apoptotic cells often show false staining because of increased oxidative enzymes, and areas of necrosis, in particular, should be avoided when assessing immunostains or when attempting to quantitate. The stratum granulosum of the epidermis may show nonspecific staining as do RNA-rich cells. Interestingly, cells rich in mitochondria may also stain nonspecifically and in oncocytomas, the abundance of mitochondria can pose a problem in the assessment of immunostaining. The same occurs with cells rich in lysosomes with abundant oxidative enzymes such as in granular cell tumors. Certain tissues are rich in endogenous biotin and staining systems that employ avidin will produce nonspecific staining unless the endogenous biotin is blocked. Renal, liver, and large bowel tissues contain high levels of endogenous biotin. Endogenous biotin in gestational endometrial glands may produce prominent false-positive intranuclear inclusions when stained for viral antigens. In such situations, irrelevant antibodies are likely to produce staining of the biotin that can be reduced with biotin blocking procedures or more conveniently eliminated by employing a system that does not use avidin such as the alkaline- phosphatase-peroxidase or EPOS (Dako, Santa Barbara, CA, USA) (93). Lastly, antigens may be phagocytosed by macrophages. Myoglobin released from necrotic skeletal muscle in the vicinity of soft tissue tumors may be phagocytosed by macrophages and should not be mistaken for myoglobin immunoexpressing tumor cells. 4.4. Receptor Detection for Targeted Therapies
The drive for quantitation of immunohistological stains has largely escalated because of target therapies in an increasing number of cancers, as there is evidence to suggest that the response to such treatments is dependent on the amount of receptors expressed by the tumor cells. Currently, the number of tumors that can be treated in this manner is small but this form of therapy has potential to increase. Targeted therapy with humanized antibodies that have been validated include those to HER2/neu (Trastuzumab) in breast cancer, CD117 (Imatinib, Glivec, STI571) in chronic myeloid
58
Leong and Leong
leukemia and gastrointestinal stromal tumor, CD20 (Rituximab, Rituxan) in aggressive B cell lymphomas including mantle cell lymphoma and diffuse large cell lymphoma, CD33 (Gemtuzumab ozogamicin, Mylotarg) in myeloblastic leukemia, EGFR in colorectal cancer (Cetuximab, Erbitux) and nonsmall cell lung cancer (NSCLC) (Gefitinib, Iressa, Erlotinib, Tarceva), and the somatostatin receptors (Sandostatin, Octreotide) in pituitary and gastroenteropancreatic endocrine tumors. CD20 and CD33 staining are enumerated in the flow cytometer so that scores are based on the percentage of positive cells. Cutoff values are generally set at 90% and above. With immunohistological methods of receptor detection, a variety of methods of scoring have been devised based on the percentage of positivity and intensity of staining of tumor cells. For the immunohistological scoring of EGFR in colorectal cancer and NSCLC, at least three Food and Drug Administration (FDA) approved EGFR kits have been employed in clinical trials of monoclonal antibody-based targeted treatment. These have included the Dako EGFR pharmaDx, Zymed EGFR kit, and Ventana EGFR 3C6 antibody. One study comparing the sensitivity of these kits for EGFR detection in metastatic colorectal carcinoma found Zymed and Ventana kits to be more sensitive but a high concordance was observed for all three kits in the evaluation of intensely stained tumor cells (94). Interestingly, when the authors examined scoring systems that combined the percentage of positive cells and staining intensity they found it to be not useful as staining intensity correlated with the percentage of positive cells. They also found that fixatives and the nature of the specimen did not influence staining results. Other studies have suggested that the Dako kit may be more sensitive especially for the prediction of survival with Gefitinib in NSCLC (95) and that the percentage of positive tumor cells predicted benefit from gefitinib and not the intensity of staining (95, 96). Cutoff values have been chosen arbitrarily with some trials employing a cutoff of 1% or more positivity irrespective of membrane staining being complete or incomplete (97, 98), while others have adopted different values, reflecting the lack of a consensus concerning scoring for EGFR immunostaining (99, 100). The conflicting results obtained with different reagents and cutoff values has raised question as to the reliability of immunohistochemical assessment of EGFR (101). 4.5. Hormone Receptors and HER2/ neu in Breast Cancer
One of the earliest immunohistological stains to be subjected to some form of quantitative evaluation was the staining of hormone receptors in breast cancer. In the case of estrogen and progesterone receptors, when the initial reluctance to accept immunohistological staining in place of cytosolic assays was overcome, there was pressure to adopt some method of quantitation to replace the
Standardization in Immunohistology
59
quantitative results obtained from cytosolic assays. In our study, we found that 10% positivity corresponded to the 10 fmol cutoff adopted for the cytosolic assay of estrogen receptor (28). Because of pressure to provide numbers to the stains we reported on two parameters for estrogen and progesterone immunohistological stains, namely intensity of stain, i.e., negative, weak, moderate, and strong (or 0, 1+, 2+, and 3+, respectively) and the extent of staining, i.e., <10%, 11–25%, 26–50%, 51–75%, and >75%. The latter were chosen for convenience as they could, with experience, be assessed by examination with a low-power objective lens. Several methods of quantitation for hormone receptors have been proposed, and one of the more widely used methods employs both the intensity and extent of staining in a combined fashion (102–104). The scores are employed clinically to predict response to hormone therapy and for prognosis, despite the observation that a clinical response was obtained with tumors displaying as low as 1% of estrogen receptor positivity (104) which led to the National Institutes of Health recommendation that any positive staining for estrogen receptor is considered to be a definitive result and indication for antiestrogen therapy. More recently, it was demonstrated in a large number of patients that the distribution of estrogen and progesterone receptors is bimodal as is the distribution of Allred scores (105). This finding implies that these parameters predictive of response to hormonal therapy are generally either negative or positive, with only a small number falling in between (106). The overexpression of Her2/neu protein has been confirmed to be an independent prognostic marker in node-positive and more recently, also in node-negative patients. More importantly, HER2/neu positive status predicts positive response to adriamycinbased therapies and poor response to tamoxifen, even in estrogen receptor-positive tumors. The recommended scoring for HER2 is a four-tiered system, where 0 = no staining; 1+ = faint/barely perceptible membrane staining in more than 10% of tumor cells, which only stain in part of their membrane; 2+ = weak to moderate complete membrane staining in >10% of tumor cells; 3+ = strong complete membrane staining in >10% of tumor cells (107, 108). As with other quantitative scores in immunohistology these cutoffs have been arbitrarily set and the 10% value was recently revised to 30%, again arbitrarily (82). Scores 0 and 3+ are easy to identify, but it is more difficult to discriminate between 1+ and 2+, and between 2+ and 3+ as the distinction is largely based on subjective perception of the intensity of staining. Another issue of contention is the use of the 10% cutoff for a positive result (84). This value is entirely arbitrary and did not take into account the significance of heterogeneity in Her2 staining. When the cutoff was arbitrary raised to 60% in a multicenter study, a concordance of 95% was obtained with FISH
60
Leong and Leong
(109). In the National Surgical Adjuvant Breast and Bowel Project studies, when 3+ cases were retested in a central laboratory up to 26% were found to be negative (110, 111). In one quality assurance program, it was found that when using an approved test kit for HER2, only 56% of participating laboratories attained acceptable staining (112). In view of the subjectivity involved in assessing stain intensity and therefore percentages of positive-stained cells, it would be more accurate to associate immunohistochemical labeling with specific cell structures. For example, when Her2 immunoexpression was linked to a specific pattern of membrane localization whereby staining of the entire thickness of cell membrane present in the section produced a band-like pattern; a significantly greater degree of concordance with FISH was achieved (113). The HER2 protein is synthesized in the cytoplasm and transported to the cell membrane; so cytoplasmic staining is not an artifact. At the membrane in normal breast epithelium, HER2 aggregates in clusters, located predominantly in the basolateral aspect of the cells (84, 114). When HER2 is upregulated, these aggregates become larger and eventually coalesce to give the appearance of linear membrane expression (114). As the tissue section includes 5-mm thick slices of cell membrane, the appearance of positive staining for HER2 should be visualized as a band (84, 113). This pattern is also seen with other cell-membrane localized antigens such as epithelial membrane antigen, E-cadherin, CD79a, CD20, and CD3. 4.6. Interobserver Variability in Quantitative Immunohistology
In one study of interobserver reliability in the scoring of four markers in colorectal cancer it was suggested that there was substantial agreement between six observers for p53 and VEGF when employing a “positive” score with a predetermined cutoff (115). Closer examination of the results reveal that the interclass correlation coefficient was “strong” (>0.75%) only for p53 (a nuclear antigen) when a cutoff of 10% was applied and “excellent” (Kappa coefficient 0.831) when no positivity versus any positivity was evaluated but fell significantly when other cutoff values were used. With the cytoplasmic antigens VEGF, Bcl-2, and APAF-1, Kappa coefficients were all <0.50 or “poor” when values above 10% cutoff were used. In other words, correlation was good for a “yes” or “no” result or at a 10% cutoff, and only in the case of a nuclear antigen. When cutoffs were based on different values, correlation was weak suggesting that quantitation of immunohistology is unreliable and exposing the subjectivity of quantitative scoring. Interestingly, in subsequent publications from the same group, it was recommended that a “scoring method based on percentage of positive tumor cells, rather than on staining intensity” be adopted (116). Furthermore, results from such studies reveal that nuclear staining is easier to assess compared to cytoplasmic
Standardization in Immunohistology
61
s taining where weak staining can pose difficulties in assessment (117). It is for these reasons that we advocate localization of staining to specific anatomical structures as the most accurate method of assessing immunostains.
5. Conclusions From the preceeding, it is clear that many variables exist in the different analytical phases of the immunohistological assay (118). Some of these variables are critical to the preservation of antigens in FFPT. To attain reproducible quantitative results that are meaningful, these variables need to be standardized between laboratories and between different specimens accessioned in the same laboratory. The variables, particularly those in the preanalytical phase are not within the control of the laboratory and currently pose insurmountable obstacles to standardization. Antigen retrieval has been a major milestone in alleviating many of the problems related to the identification of proteins in FFPT but careful consideration must be given to all the variables discussed and these need to be standardized. For this to happen, international and national consensus groups can play an important role in their recommendations of optimal methods and procedures. The recent guidelines from the College of American Pathologists and American Society of Clinical Oncology for human EGFR 2 testing in breast cancer is one such example (82), although it is still a long way from enforcing standardization of the technology and methodology. Standardization of the significant variables that occur before accessioning by the laboratory requires co-operation of the clinicians responsible for obtaining the specimens. One way of achieving this is through some form of enforcement by peer groups. Delays in fixation should be minimized and the time placed into formalin should be recorded allowing the total duration of fixation to be controlled. Specimens that do not conform to recommended optimal conditions cannot be regarded to produce acceptable quantitative results especially if all other laboratory procedures are standardized. For consistency, the laboratory should, as far as possible, adopt standard procedures for fixation, tissue processing, sectioning, and staining. Automated tissue processing requires standardization and newer methods of processing that employ proprietary reagents require validation. The availability of several automated staining devices with different protocols presents yet another variable that requires validation. As it is not possible to standardize the many variables that can influence antigen preservation in the sample, the best alternative at present is to optimize the staining procedure in order to obtain the best results,
62
Leong and Leong
and this is done against known positive controls of appropriate tissues. Quality assurance programs can be particularly helpful in identifying faults in laboratory protocols and providing a medium to compare performance. New antibodies require appropriate validation. All too often, the claims of the antibody distributor/manufacturer regarding sensitivity and specificity are accepted without in-house validation, sometimes leading to inappropriate interpretation of the staining results. The antibody should be validated and optimized to known positive controls of similar tissue that have been fixed and processed in a similar if not identical manner to the test tissue. In addition, the antigen-retrieval method and staining protocol should be similar. To ascertain the most optimal method of antigen retrieval the following procedures should be routinely employed for each new antibody. Antigen retrieval should be performed with citrate buffer, 10 mmol/l at pH 6.0, and citrate/ EDTA at pH 8.0 as retrieval solutions. This should be employed at both 98 and 110°C, with varying concentrations of primary antibody. If satisfactory results are not obtained, enzyme digestion (with protease or similar) should be introduced as an additional step before and after the heat-induced antigen retrieval. Assessment is based on the signal-to-noise ratio and the integrity of tissue morphology. It may be necessary, with less sensitive antibodies, to experiment with buffers of varying pH as retrieval solutions. Another variable that may require manipulation is the duration of heating. Optimization of the antigen-retrieval procedure for each antibody is, at present, the most important step toward standardization. The optimization process merely ensures that the procedure is at its greatest sensitivity but cannot verify proper tissue fixation or processing, and does not equate to standardization. Validation of the antibody and methodology against benchmark standards is a more complex issue that has been discussed. Appropriate scoring methods have to be developed. Scoring methods need to be evidence-based; the best form of validation being patient outcome. Such validation will require collaborative or international studies. It is clear that scores based on intensity produce variable results that are often not reproducible. When the target antigen is confined to a known anatomical structure, scoring becomes easier with greater concordance and reproducibility (115, 116) but only when it is applied as percentage of staining cells without concern for intensity. Cytoplasmic staining is more difficult to score as the target antigen can be localized in several organelles. If possible, cytoplasmic staining should be assessed on the basis of specific organelle localization. Similarly, with cell membrane staining as with Her2/neu antigen, the entire thickness of the cell membrane in the tissue section should be stained before being assigned a positive score.
Standardization in Immunohistology
63
Other factors in the postanalytical phase are dependent on the experience of the reporting pathologist. Selection of the appropriate antibodies to be used and the appropriate area in the section for assessment is dependent on pathologists’ expertise. Cutoff numbers are mostly arbitrary and when scoring, should assessment be made in the areas of highest intensity or is a mean value more appropriate? Some of these issues can be standardized through the use of automated cellular imaging systems that can evaluate the target cells using a large number of different morphologic parameters. Despite all these measures it is clear that there will remain a number of variables that can significantly influence antigen preservation that are beyond the control of the laboratory. Even among specimens accessioned by the same laboratory, many preanalytical factors are different and no two specimens are subjected to identical conditions. The recognition of these variables and the acknowledgement of the need for standardization, however, is a large step closer to attaining the goal of quantitative immunohistology. In the interim, all quantitative measures in immu nohistology, should be viewed with caution, especially if therapeutic decisions are based on assigned scores (89, 92). References 1. Leong, A.S.-Y., Wick, M.R., and Swanson, P.E. (1997) Immunohistology and Electron Microscopy of Anaplastic and Pleomorphic Tumours. Cambridge: Cambridge University Press; 2–35. 2. Elias, J.M. (2003) Immunohistopathology. A Practical Approach to Diagnosis. Chicago: ASCP Press. 3. Taylor, C.R. and Cote, R.J. (2005) Immunomicroscopy, A Tool for the Surgical Pathology, 3rd Edition. Edinburgh: Elsevier. 4. Leong, A.S.-Y. and Wright, J. (1987) The contributions of immunohistochemical staining in tumour diagnosis. Histopathology 11, 1295–1305. 5. Leong, A.S.-Y. and Leong, F.J. (1997) Immunohistochemistry in the Diagnosis of Solid Tumours. IN: Nakamura, R., ed. Manual of Clinical Laboratory Immunology, 5th Edition. Washington, DC: ASM Press; 380–387. 6. Leong, A.S.-Y. (1992) New vistas in the histopathological assessment of cancer. Med J Aust. 157, 699–701. 7. Leong, A.S.-Y., Vinyuvat, S., Suthipintawong, C., and Leong, F.J. (1997) Patterns of basal lamina immunostaining in soft tissue tumours. Appl Immunohistochem. 5, 1–7. 8. Leong, A.S.-Y., Parkinson, R., and Milios, J. (1990) ‘Thick’ cell membranes revealed by immunocytochemical staining. A clue to the
diagnosis of malignant mesothelioma. Diagn Cytopathol. 6, 9–13. 9. Leong, A.S.-Y., Stevens, M.W., and Mukherjee, T.M. (1992) Malignant mesothelioma: cytologic diagnosis with histologic, immunohistochemical and ultrastructural correlation. Semin Diagn Pathol. 9, 141–150. 10. Leong, A.S.-Y. and Lee, A.K.C. (1995) Biological indices in the assessment of breast cancer. Clin Mol Pathol. 48, M221–M238. 11. Leong, A.S.-Y. (2001) Immunohistological markers for tumor prognostication. Curr Diagn Pathol. 7, 176–186 12. Bratthauer, G.L., Moinfar, F., Stamatakos, M.D., et al. (2002) Combined E-cadherin and high molecular weight cytokeratin immunoprofile differentiates lobular, ductal, and hybrid mammary intraepithelial neoplasias. Hum Pathol. 33, 620–627. 13. Wood, B. and Leong, A.S.-Y. (2003) Cell adhesion proteins – biology, detection and applications. Pathology 35, 101–105. 14. Ribic, C.M., Sargent, D.J., Moore, M.J., et al. (2003) Tumor microsatellite-instability status as a predictor of benefit from fluorouracil-based adjuvant chemotherapy for colon cancer. N Engl J Med. 349, 247–257. 15. Popat, S., Hubner, R., and Houlston, R.S. (2005) Systematic review of microsatellite
64
Leong and Leong
instability and colorectal cancer prognosis. J Clin Oncol. 23, 609–618. 16. Acs, G., Lawton, T.J., Rebbeck, T.R., et al. (2001) Differential expression of E-cadherin in lobular and ductal neoplasms of the breast and its biologic and diagnostic implications. Am J Clin Pathol. 115, 85–98. 17. Fletcher, C.D., Berman, J.J., Corless, C., et al. (2003) Diagnosis of gastrointestinal stromal tumors: a consensus approach. Hum Pathol. 33, 459–465. 18. Stein, H., Foss H.D., Durkop, H., et al. (2000) CD30+ anaplastic large cell lymphoma: a review of its histopathologic, genetic, and clinical features. Blood 96, 3681–3695. 19. Leong, A.S.-Y. and Leong, T.Y.-M. (2006) Invited review: newer developments in immunohistology. J Clin Pathol. 59, 1117–1126. 20. Taylor, C.R. (1994) An exaltation of experts: concerted efforts in the standardization of immunohistochemistry. Hum Pathol. 25, 2–4. 21. O’Leary, T.J. (2001) Standardization in immunohistochemistry. Appl Immunohis tochem Mol Morphol. 9, 3–8. 22. Taylor, C.R. and Levenson, R.M. (2006) Quantification of immunohistochemical – issues concerning methods, utility and semiquantitative assessment II. Histopathology 49, 411–424. 23. Goldstein, N.S., Hewitt, S.M., Taylor, C.R., et al. (2007) Recommendations for improved standardization of immunohistochemistry. Appl Immunohistochem Mol Morphol. 15, 124–133. 24. Pelstring, R.J., Allred, D.C., Esther, R.J., et al. (1991) Differential antigen preservation during autolysis. Hum Pathol. 22, 237–241. 25. Pearse, A.G.E. (1980) Histochemistry. Theoretical and Applied, 4th Edition, vol 1. Edinburgh: Churchill Livingstone; 95. 26. Leong, A.S.-Y. and Gilham, P.N. (1989) The effects of progressive formaldehyde fixation on the preservation of tissue antigens. Pathology 21, 81–89. 27. Goldstein, N.S., Ferkowicz, M., Odish, E., et al. (2003) Minimum formalin fixation time for consistent estrogen receptor immunohistochemical staining of invasive breast carcinoma. Am J Clin Pathol. 120, 86–92. 28. Raymond, W. and Leong, A.S.-Y. (1990) Oestrogen receptor staining of paraffinembedded breast carcinomas following short fixation in formalin: a comparison with
c ytosolic and frozen section receptor analyses. J Pathol. 160, 295–303. 29. Leong, A.S.-Y. and Milios, J. (1993) An assessment of the efficacy of the microwaveantigen retrieval procedure on a range of tissue antigens. Appl Immunohistochem. 1, 267–274. 30. Leong, A.S.-Y. and Milios, J. (1993) Comparison of antibodies to oestrogen and progesterone receptors and the influence on microwave-antigen retrieval. Appl Immunohistochem. 1, 2–88. 31. Leong, A.S.-Y., Milios, J., and Duncis, C.G. (1988) Antigen preservation in microwaveirradiated tissues. A comparison with routine formalin fixation. J Pathol. 156, 275–282. 32. Suthipintawong, C., Vinyuvat, S., and Leong, A.S.-Y. (1996) Immunostaining of cell preparations: a comparative evaluation of common fixatives and protocols. Diagn Cytopathol. 15, 167–174. 33. Larsson, L. (1993) Tissue preparation methods for light microscopic immunohistochemistry. Appl Immunohistochem. 1, 2–16. 34. Leong, A.S.-Y. (1994) Fixation and Fixatives. IN: Woods, A.E., Ellis, R.C., eds. Laboratory Histopathology – A Complete Reference. London: Churchill Livingstone; 1–26. 35. Gown, A.M. and Vogel, A.M. (1984) Monoclonal antibodies to human intermediate filament proteins. II. Distribution of filament proteins in normal human tissues. Am J Pathol. 114, 309–321. 36. Shi, S.R., Liu, C., Pootrakul, L., et al. (2006) Evaluation of the value of frozen tissue section used as ‘gold standard’ for immunohistochemistry. Am J Clin Pathol. 129, 358–366. 37. Kakimoto, K., Takekoshi, S., Miyajima, K., and Osamura, R.Y. (2008) Hypothesis for the mechanism for heat-induced antigen retrieval occurring on fresh frozen sections without formalin fixation in immunohistochemistry. J Mol Histol. 39, 389–399. 38. Visinoni, F., Milios, J., Leong, A.S.-Y., et al. (1998) Ultra-rapid microwave/variable pressure induced histoprocessing: description of a new tissue processor. J Histotechnol. 21, 219–224. 39. Morales, A.R., Nassiri, M., Kanhoush, R., et al. (2004) Experience with an automated microwave assisted rapid tissue processing method: effect on histology and timeliness of diagnostic surgical pathology. Am J Clin Pathol. 121, 528–536. 40. Jacobs, T.W., Prioleau, J.E., Stillman, I.E., et al. (1996) Loss of tumor markerimmunostaining intensity on stored paraffin
Standardization in Immunohistology slides of breast cancer. J Natl Cancer Inst. 88, 1054–1060. 41. Wester, K., Wahlund, E., Sundstrom, C., et al. (2000) Paraffin section storage and immunohistochemistry. Effects of time, temperature, fixation, and retrieval protocol with emphasis on p53 protein and MIB1 antigen. Appl Immunohistochem Mol Morphol. 8, 61–70. 42. DiVito, K.A., Charette, L.A., Rimm, D.L., and Camp, R.L. (2004) Long-term preservation of antigenicity on tissue microarrays. Lab Invest. 84, 1071–1078. 43. Blind, C., Koepenik, A., Pacyna-Gengelbach, M., et al. (2008) Antigenicity testing by immunohistochemistry after tissue oxidation. J Clin Pathol. 61, 79–83. 44. Leong, A.S.-Y. (1993) Immunohistochemistry – Theoretical and Practical Aspects. IN: Leong, A.S.-Y., ed. Applied Immunohis tochemistry for the Surgical Pathologist. London: Edward Arnold; 1–22. 45. Sabattini, E., Bisgaard, K., Ascani, S., et al. (1998) The EnVision system. A new immunohistochemical method for diagnostics and research: critical comparison with the APAAP, ChemMate, CSA, LABC, and SABC techniques. J Clin Pathol. 51, 506–510. 46. King, G., Payne, S., Walker, F., et al. (1998) A highly sensitive detection method for immunohistochemistry using biotinylated tyramine. J Pathol. 183, 237–241. 47. Shi, S.R., Key, M.E., and Kalra, K.L. (1991) Antigen retrieval in formalin-fixed, paraffinembedded tissues: an enhancement method for immunohistochemical staining based on microwave oven heating of tissue sections. J Histochem Cytochem. 39, 741–748. 48. Gown, A.M., de Wever, N., and Battifora, H. (1993) Microwave-based antigenic unmasking: a revolutionary new technique for routine immunohistochemistry. Appl Immunohistochem. 1, 256–266. 49. Miller, R.T., Swanson, P.E., and Wick, M.R. (2000) Fixation and epitope retrieval in diagnostic immunohistochemistry: a concise review with practical considerations. Appl Immunohistochem Mol Morphol. 8, 228–235. 50. Leong, A.S-Y., Milios, J., and Leong, F.J. (1996) Epitope retrieval with microwaves: a comparison of citrate buffer and EDTA with three commercial retrieval solutions. Appl Immunohistochem. 4, 201–207. 51. Chaiwun, B., Shi, S.R., Cote, R.J., and Taylor, C.R. (2000) Major Factors Influencing the Effectiveness of Antigen Retrieval Immunohistochemistry. IN: Shi, S.R., Gu, J., Taylor, C.R., eds. Antigen
65
Retrieval Techniques. Natick, MA: Eaton Publishing; 41–53. 52. Leong, A.S.-Y., Lee, E.S., Yin, H., et al. (2002) Superheating antigen retrieval. Appl Immunohistochem. 10, 263–268. 53. Fraenkel-Conrat, H., Brandon, B., and Olcott, H. (1947) The reaction of formaldehyde with proteins. IV: Participation of indole groups: gramicidin. J Biol Chem. 168, 99–118. 54. Fraenkel-Conrat, H. and Olcott, H. (1948) Reaction of formaldehyde with proteins. VI: Crosslinking between amino groups with phenol, imidazole, or indole groups. J Biol Chem. 174, 827–843. 55. Fraenkel-Conrat, H. and Olcott, H. (1948) The reaction of formaldehyde with proteins. V: Crosslinking between amino and primary amide or guanidyl groups. J Am Chem Soc. 70, 2673–2684. 56. Shi, S.-R., Gu, J., Turrens, J., et al. (2000) Development of the Antigen Retrieval Technique: Philosophical and Theoretical Bases. IN: Shi, S.-R., Gu, J., Taylor, C.R., eds. Antigen Retrieval Techniques: Immuno histochemical and Molecular Morphology. Natick, MA: Eaton Publishing; 17–40. 57. Cattoretti, G., Peleri, S., Parravicini, C., et al. (1993) Antigen unmasking on formalinfixed, paraffin-embedded tissue sections. J Pathol. 171, 83–98. 58. Suurmeijer, A.J.H. and Boon, M.E. (1993) Notes on the application of microwaves for antigen retrieval in paraffin and plastic tissue sections. Eur J Morphol. 31, 144–150. 59. Morgan, J.M., Navabi, H., and Jasani, B. (1997) Role of calcium chelation in hightemperature antigen retrieval at different pH values. J Pathol. 182, 233–237. 60. Hjerpe, A., Boon, M.E., and Kok, L.P. (1988) Microwave stimulation of an immunological reaction (CEA/anti-CEA) and its use in immunohistochemistry. Histochem J. 20, 388–396. 61. Choi, T.-S., Whittlesey, M., Slap, S.E., et al. (1997) Microwave Immunocytochemistry: Advances in Temperature Control. IN: Gu, J., ed. Analytical Morphology: Theory, Applica tions, and Protocols. Natick, MA: Eaton Publishing; 91–114. 62. Takes, P.A., Kohrs, J., Krug, R., and Kewley, S. (1999) Microwave technology in immunohistochemistry: application to avidinbiotin staining of diverse antigens. J Histotechnol. 12, 95–98. 63. Hopwood, D., Yeaman, G., and Milne, G. (1988) Differentiating the effects of microwave and heat on tissue proteins and their
66
Leong and Leong
cross linking by formaldehyde. Histochem J. 20, 341–346. 64. Porcelli, M., Cacciapuoti, G., Fusco, S., et al. (1997) Non-thermal effects of microwaves on proteins: thermophilic enzymes as model system. FEBS Lett. 402, 102–106. 65. Sompuram, S.R., Vani, K., and Bogen, S.A. (2006) A molecular model of antigen retrieval using a peptide array. Am J Clin Pathol. 25, 91–98. 66. Sompuram, S.R., Vani, K., Messana, E., and Bogen, S.A. (2004) A molecular mechanism of formalin fixation and antigen retrieval. Am J Clin Pathol. 121, 190–199. 67. Leong, A.S.-Y., Haffajee, Z., and Clark, M. (2007) Microwave enhancement of CISH for Her2 oncogene. Appl Immunohistochem Mol Morphol. 15, 88–93. 68. Portiansky, E.L. and Gimeno, E.J. (1996) A new epitope retrieval method for the detection of structural cytokeratins in the bovine prostate tissue. Appl Immunohistochem. 4, 208–214. 69. Leong, A.S.-Y. (2005) Microwave Technology for Light Microscopy and Ultrastructural Studies. Bangkok: Amarin Printing and Publishing Company Ltd. 70. Leong, A.S.-Y., Yin, H., and Haffajee, Z. (2002) Patterns of immunostaining of immunoglobulin in formalin-fixed, paraffinembedded sections. Appl Immunohistochem Mol Morphol. 10, 110–114. 71. Taylor, C.R., Shi, S.R., Chen, C., et al. (1996) Comparative study of antigen retrieval heating methods: microwave, microwave and pressure cooker, autoclave and steamer. Biotech Histochem. 71, 263–270. 72. Shi, S.R., Imam, S.A., Young, L., Cote, R.J., and Taylor, C.R. (1995) Antigen retrieval immunohistochemistry under the influence of pH using monoclonal antibodies. J Histochem Cytochem. 43, 193–201. 73. Lan, H.Y., Mu, W., Nikolic-Paterson, D.J., and Atkins, R.C. (1995) A novel, simple, reliable, and sensitive method for multiple immunoenzyme staining: use of microwave oven heating to block antibody cross reactivity and retrieve antigens. J Histochem Cytochem. 43, 97–102. 74. Gao, C., Wang, A.Y., and Han, Y.J. (2008) Microwave antigen retrieval blocks endogenous peroxidase activity in immunohistochemistry. Appl Immunohistochem Mol Morphol. 16, 393–399. 75. Leong, A.S.-Y., Milios, J., and Tang, S.K. (1993) Is immunolocalisation of proliferating cell nuclear antigen (PCNA) in paraffin
sections a valid index of cell proliferation? Appl Immunohistochem. 1, 27–35. 76. Leong, A.S.-Y., Vinyuvat, S., Suthipintawong, C., and Milios, J. (1995) A comparative study of cell proliferation markers in breast carcinomas. Clin Mol Pathol. 48, M83–M87. 77. Willingham, M.C. (1999) Conditional epitopes: is your antibody always specific? J Histochem Cytochem. 47, 1233–1239. 78. Burry, R.W. (2000) Specificity controls for immunocytochemical methods. J Histochem Cytochem. 48, 163–168. 79. Riera, J., Simpson, J.F., Tamayo, R., et al. (1999) Use of cultured cells as a control for quantitative immunocytochemical analysis of estrogen receptor in breast cancer. The Quicgel method. Am J Clin Pathol. 111, 329–332. 80. Taylor, C.R. (2006) Personal communication. 81. McGuire, W.L. (1991) Breast cancer prognostic factors: evaluation guidelines. J Natl Cancer Inst. 83, 154–155. 82. Wolff, A.C., Hammond, M.E., Schwartz, J.N., et al. (2007) American Society of Clinical Oncology/College of American Pathologists guideline recommendations for human epidermal growth factor receptor 2 testing in breast cancer. J Clin Oncol. 25, 118–145. 83. Vogel, C.L., Cobleigh, M.A., Tripathy, D., et al. (2002) Efficacy and safety of trastuzumab as a single agent in first-line treatment of HER2-overexpressing metastatic breast cancer. J Clin Oncol. 20, 719–726. 84. Leong, T.Y.-M. and Leong, A.S.-Y. (2006) Controversies in the assessment of HER-2. More questions than answers. Adv Anat Pathol. 13, 263–269. 85. Dybdal, N., Leiberman, G., Anderson, S., et al. (2005) Determination of HER2 gene amplification by fluorescence in situ hybridization and concordance with the clinical trials immunohistochemical assay in women with metastatic breast cancer evaluated for treatment with trastuzumab. Breast Cancer Res Treat. 93, 3–11. 86. Perez, E.A., Suman, V.J., Davidson, N.E., et al. (2006) HER2 testing by local, central, and reference laboratories in specimens from the North Central Cancer Treatment Group N9831 intergroup adjuvant trial. J Clin Oncol. 24, 3032–3038. 87. Rhodes, A., Jasani, B., Balaton, A.J., et al. (2001) Study of interlaboratory reliability and reproducibility of estrogen and progesterone receptor assays in Europe. Documentation of
Standardization in Immunohistology poor reliability and identification of insufficient microwave antigen retrieval time as a major contributory element of unreliable results. Am J Clin Pathol. 115, 44–58. 88. Fitzgibbons, P.L., Murphy, D.A., Dorfman, D.M., et al. (2006) Interlaboratory comparison of immunohistochemical testing for HER2: results of the 2004 and 2005 College of American Pathologists HER2 Immunohistochemistry Tissue Microarray Survey. Arch Pathol Lab Med. 130, 1440–1445. 89. Leong, A.S.-Y. (2004) Pitfalls in diagnostic immunohistology. Adv Anat Pathol. 11, 86–93. 90. Smith, P.S., Parkinson, I.H., and Leong, A.S.-Y. (1996) Principles of ploidy analysis by static cytometry. Clin Mol Pathol. 49, M104–M111. 91. Seidal, T., Balaton, A.J., and Battifora, H. (2001) Interpretation and quantification of immunostains. Am J Surg Pathol. 25, 1204–1207. 92. Leong, A.S.-Y. (2004) Quantitation in immunohistology: fact or fiction? A discussion of variables that influence results. Appl Immunohistochem Mol Morphol. 12, 1–7. 93. Cooper, K., Haffajee, Z., and Taylor, L. (1997) Comparative analysis of biotin intranuclear inclusions of gestational endometrium using the APAAP, ABC and PAP immunodetection systems. J Clin Pathol. 50, 153–156. 94. Penault-Llorca, F., Cayre, A., Arnould, L., et al. (2006) Is there an immunohistochemical technique definitely valid in epidermal growth factor assessment? Oncol Rep. 16, 1173–1179. 95. Hirsch, F.R., Dziadziuszko, R., Thacher, N., et al. (2008) Epidermal growth factor receptor immunohistochemistry: comparison of antibodies and cutoff points to predict benefit from gefitinib in a phase 3 placebocontrolled study in advanced nonsmall-cell lung cancer. Cancer 112, 1114–1121. 96. Cappuzzo, F., Finocchiaro, G., Rossi, E., et al. (2008) EGFR FISH assay predicts for response to cetuximab in chemotherapy refractory colorectal cancer patients. Ann Oncol. 19, 717–723. 97. Cunningham, D., Humblet, Y., Siena, S., et al. (2004) Cetuximab monotherapy and cetuximab plus irinotecan in irinotecanrefractory metastatic colorectal cancer. N Engl J Med. 351, 337–345. 98. Saltz, L.B., Meropol, N.J., Loehrer, P.J., et al. (2004) Phase II trial of cetuximab in patients with refractory colorectal cancer that
67
expresses the epidermal growth factor receptor. J Clin Oncol. 22, 1201–1208. 99. Goldstein, N.S. and Amin, M. (2001) Epidermal growth factor receptor immunohistochemical reactivity in patients with American Joint Committee on Cancer Stage IV colon adenocarcinoma: implications for a standardized scoring system. Cancer 92, 1331–1346. 100. Tos, A.P. and Ellis, I. (2005) Assessing epidermal growth factor receptor expression in tumours: what is the value of current test methods? Eur J Cancer. 41, 1383–1392. 101. Sabourin, J.C., Cayre, A., Arnould, L., et al. (2005) Comparison of three commercially available immunohistochemical tests for EGFR expression in clorectal cancers. Is immunohistochemistry (IHC) reliable? J Clin Oncol. 23(June Suppl), 9705–9710. 102. Barnes, D.M., Harris, W.H., Smith, P., et al. (1996) Immunohistochemical determination of oestrogen receptor: comparison of different methods of assessment of staining and correlation with clinical outcome of breast cancer patients. Br J Cancer 74, 1445–1451. 103. Allred, D.C., Harvey, J.M., Berardo, M., et al. (1998) Prognostic and predictive factors in breast cancer by immunohistochemical analysis. Mod Pathol. 11, 155–168. 104. Hervey, J.M., Clark, G.M., Osborne, C.K., et al. (1999) Estrogen receptor status by immunohistochemistry is superior to the ligand-binding assay for predicting response to adjuvant endocrine therapy in breast cancer. J Clin Oncol. 17, 1474–1481. 105. Collins, L.C., Botero, M.L., and Schnitt, S.J. (2005) Bimodal frequency distribution of estrogen receptor immunohistochemical staining results in breast cancer: an analysis of 825 cases. Am J Clin Pathol. 123, 16–20. 106. Nadji, M. (2006) Quantitative immunohistochemistry of estrogen receptor in breast cancer. Much ado about nothing. Appl Immunohistochem Mol Morphol. 16, 105–107. 107. Jacobs, T.W., Gown, A.M., Yazji, H., et al. (1999) Specificity of HercepTest in determining HER-2/neu status of breast cancers using the United States Food and Drug Administration-approved scoring system. J Clin Oncol. 17, 1983–1987. 108. Allred, D.C. and Swanson, P.E. (2000) Testing for erbB-2 by immunohistochemistry in breast cancer. Am J Clin Pathol. 113, 171–175. 109. Vincent-Salomon, A., MacGrogan, G., Couturier, J., et al. (2003) Calibration of
68
Leong and Leong
immunohistochemistry for assessment of Her2/neu in breast cancer: results of the French Multicentre GEFPICS Study. Histopathology 42, 337–347. 110. Paik, S., Bryant, J., and Tan-Chiu, E., et al. (2002) Real-world performance of HER2 testing – National Surgical Adjuvant Breast and Bowel Project experience. J Natl Cancer Inst. 94, 852–854. 111. Roche, P.C., Suman, V.J., Jenkins, R.B., et al. (2002) Concordance between local and central laboratory HER2 testing in the Breast Intergroup Trial N9831. J Natl Cancer Inst. 94, 855–857. 112. Miller, K. and Ibrahim, M. (2004) The breast HER-2 module. Immunocytochemistry 3, 147–150. 113. Leong, A.S.-Y., Formby, M., Haffajee, Z., and Morey, A. (2006) Refinement of immunohistological parameters for Her2/neu scoring. Validation by FISH and CISH. Appl Immunohistochem Mol Morphol. 14, 384–389.
114. Nagy, P., Jenei, A., Kirsch, A.K., et al. (1999) Activation-dependent clustering of the erbB2 receptor tyrosine kinase detected by scanning near-field optical microscopy. J Cell Sci. 112, 1733–1741. 115. Zlobec, I., Steele, R., Michel, R.P., et al. (2006) Scoring of p53, VEGF, Bcl-2 and APAF-1 immunohistochemistry and interobserver reliability in colorectal cancer. Mod Pathol. 19, 1236–1242. 116. Zlobec, I., Terracciano, L., Jass, J.R., and Lugli, A. (2007) Value of staining intensity in the interpretation of immunohistochemistry for tumor markers in colorectal cancer. Virchows Arch. 451, 763–769. 117. Kay, E.W., Walsh, C.J., Whelan, D., et al. (1996) Interobserver variation of p53 immunohistochemistry – an assessment of a practical problem and comparison with other studies. Br J Biomed Sci. 53, 101–107. 118. Leong, T.Y.-M. and Leong, A.S.-Y. (2007) Variables that influence outcomes in immunohistology. Aust J Med Sci. 28, 47–59.
Chapter 4 Multiple Immunofluorescence Labeling of Formalin-Fixed Paraffin-Embedded Tissue David Robertson and Clare M. Isacke Abstract Multiple immunofluorescent labeling of formalin-fixed paraffin-embedded (FFPE) tissue is not a routinely used method. At least in part, this is due to the perception that the innate autofluorescence of the FFPE material forbids the use of immunofluorescent labeling. As a result, immunohistochemical (immunoperoxidase) staining of FFPE material or cryosectioning methods is used instead. In this chapter, we describe a robust optimized method for high-resolution immunofluorescence labeling of FFPE tissue that involves the combination of antigen retrieval, indirect immunofluorescence, and confocal laser scanning microscopy. Once such samples have been prepared and imaged by confocal microscopy, they can be stored at −20°C for extensive periods (>250 days) and reexamined with minimal loss of quality. As a consequence, this method has the potential to open up the large archival sample collections to multiple immunofluorescent investigations. Key words: Immunofluorescent labeling, Formalin-fixed paraffin-embedded, FFPE, Confocal microscopy, Antigen retrieval, Pathology
1. Introduction Immunohistochemistry (IHC) is one of the pillars of modern diagnostic pathology and a fundamental research tool in both pathology and translational research laboratories. Currently, the most commonly used method is to stain the section with a primary antibody followed by a peroxidase-conjugated second layer antibody and development with a chromogenic substrate. While this method is robust and reliable, it is mainly limited to only revealing one protein at a time. There is, therefore, a need for a method that would allow multiple labeling of proteins in FFPE material. Recently, we described an optimized method for multiple
Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_4, © Springer Science+Business Media, LLC 2011
69
70
Robertson and Isacke
immunofluorescent labeling of FFPE human tissue (1). Subsequently, we have extended this method for mouse tissue and in addition have tested out some variations to the original method. These additions to the method are highlighted in this chapter together with additional notes to help those using this method for the first time. We, and others, now routinely use this method for examining human tissue samples (2, 3) and experimental samples from in vivo studies (4, 5), and as a consequence we believe that this method will be of wide use to both pathologists and research scientists.
2. Materials Unless otherwise stated, materials were stored at room temperature. 2.1. Tissue Fixation, Embedding, and Sectioning
1. 10% Neutral-buffered formalin (Bios Europe, Lancashire, UK). 2. 4% Paraformaldehyde (see Note 1). Prepared by heating the appropriate amount of paraformaldehyde in PBS on a stirring hotplate set at 150°C until the paraformaldehyde depolymerizes and goes into solution (2–3 h depending on volume being prepared). Once prepared, store the solution at room temperature. 3. Paraffin wax (Tissue-Tek; Sakura, Finland). 4. Tissue-Tek VIP automatic tissue processor (Sakura, Finland). 5. Superfrost plus glass slides (Cat No 631-0108, VWR International, Lutterworth, Leicestershire, UK). 6. Oxygen-free nitrogen gas (BOC gases, Guildford, Surrey, UK).
2.2. Antigen Retrieval
1. Xylene (Fisher Scientific, Loughborough, Leicestershire, UK). 2. Histoclear (Agar Scientific, Stanstead, Essex UK). 3. Ethanol (BDH, Poole, Dorset, UK). 4. Plastic slide rack and holder (Slide Staining System Easy Dip™ 720-0791VWR International Lutterworth Leicestershire, UK). 5. Target retrieval solution S1699 (Dako, Ely, Cambridgeshire, UK).
2.3. Immunofluorescent Labeling
1. ImmEdge pen (H-4000, Vector Laboratories, Peterborough, UK). 2. Phosphate-buffered saline (PBS, 10 mM phosphate buffer, 2.7 mM KCl, 137 mM NaCl, pH 7.4 Sigma-P4417-100TAB, Poole Dorset, UK).
Multiple Immunofluorescence Labeling of Formalin-Fixed Paraffin-Embedded Tissue
71
3. Immunofluorescence buffer (IFF). PBS plus 1% BSA (A3059, Sigma, Poole, Dorset, UK) and 2% fetal calf serum (FCS). IFF is filtered through a 0.2 mm filter and stored in 20 ml aliquots at −20°C. 4. Primary antibodies (see Note 2). Details of the primary antibodies used in the figures are described in Figs. 1 and 2 and Table 1. 5. Alexa Fluor conjugated-secondary antibodies (Invitrogen, Paisley, UK) (see Note 3). Details of secondary antibodies used here are provided in Figs. 1 and 2 and Table 1. 6. DAPI (4¢,6-diamidino-2-phenylindole; D21490, Invitrogen). 7. Vectashield (H-1000, Vector Laboratories). 8. Coverslips [22 × 40 mm, 0.155–0.185 mm thickness (VWR International, Lutterworth Leicestershire, UK)]. 9. Immersion oil (518F, Hertfordshire, UK).
Zeiss,
Welwyn
Garden
City,
Fig. 1. Expression of cytokeratin, a-smooth muscle actin and endosialin in normal human breast. 3 mm FFPE sections of human adult breast were stained for (a) DAPI (nuclear stain); (b) a-smooth muscle actin (aSMA); (c) endosialin; (d) wide spectrum cytokeratin; (e) merged image. aSMA is expressed by the myoepithelial cells and the pericytes closely associated with the endothelial cells in the vasculature. Luminal epithelial cells and non-pericyte stromal cells are aSMA negative. Endosialin expression is restricted to the stromal fibroblasts. The wide spectrum cytokeratin antibody detects the luminal cell cytokeratins and less efficiently the myoepithelial cell cytokeratins. Stromal cells are cytokeratin negative. Scale bar, 50 mm.
72
Fig. 2. Expression of E-cadherin, muscle actin, and caveolin-1 in normal mouse small intestine. 3 mm FFPE sections of mouse adult intestine were stained for (a) DAPI (nuclear stain); (b) E-cadherin; (c) muscle actin (MAct); (d) caveolin-1 staining; (e) merged image. E-cadherin expression is restricted to the lateral membranes of the intestinal epithelial cells. Muscle actin and caveolin-1 are expressed by distinct stromal cell populations. Scale bar, 25 mm.
Table 1 Details of primary and secondary antibodies Antibody
Species/Isotype
Concentration
Supplier
Endosialin: B1/35.1
Mouse IgG1
4.0 mg/ml
Isacke laboratory (6)
a-Smooth muscle actin (aSMA)
Mouse IgG2A
0.88 mg/ml
Sigma
Cytokeratin: wide spectrum screening (WSS)
Rabbit polyclonal
1:500 dilution
Dako
Muscle actin (MAct)
Mouse IgG1
1.0 mg/ml
Dako
E-cadherin
Mouse IgG2A
0.25 mg/ml
BD Biosciences
Caveolin-1
Rabbit polyclonal
1.0 mg/ml
Santa Cruz
Alexa Fluor 555 anti-mouse IgG1
Goat polyclonal
2.0 mg/ml
Invitrogen
Alexa Fluor 488 anti-mouse IgG2A
Goat polyclonal
2.0 mg/ml
Invitrogen
Alexa Fluor 633 anti-rabbit Ig
Goat polyclonal
2.0 mg/ml
Invitrogen
Multiple Immunofluorescence Labeling of Formalin-Fixed Paraffin-Embedded Tissue
73
3. Methods 3.1. Starting Material
Starting material is FFPE tissues. In our studies, these samples are typically (a) archival FFPE pathology specimens. Typically, these would have fixed overnight at room temperature in 10% neutralbuffered formalin before paraffin embedding or (b) human or rodent tissue samples that we have prepared. For this, small pieces of tissue 2–10 mm thick are fixed in 4% paraformaldehyde at room temperature with gentle tumbling for 1–16 h. Fixed material is then processed using a Tissue-Tek VIP automatic tissue processor with a standard 14 h protocol and embedded into paraffin wax.
3.2. Sectioning, Mounting onto Slides, and Slide Storage
1. Cut 3–4 mm sections from the embedded blocks and float onto a warm (42°C) water bath. 2. Pick sections from the water bath and place onto Superfrost plus slides. 3. Place slides into a vertical rack and dry overnight at 37°C in a fan-assisted cabinet. 4. The next day, label the slides and either use immediately or place into a container that is then purged with oxygen-free nitrogen gas. Store the slide containers at 4°C (see Note 4). 5. To carry out a labeling experiment, remove the required slides. Purge the container containing the remaining slides with oxygen-free nitrogen gas before replacing at 4°C.
3.3. Dewaxing and Antigen Retrieval
1. Load slides into glass or plastic slide racks. 2. Place the slide racks into a glass or plastic slide holder (see Note 5) and dewax by incubating for 2 × 10 min in 100% xylene or Histoclear (see Note 6) with agitation about once a minute. 3. Rehydrate the dewaxed slides in ethanol as follows: 2 × 10–20 s agitation in 100% ethanol, 1 × 10–20 s agitation in 90% ethanol, 1 × 10–20 s agitation in 70% ethanol, and 2 × 10–20 s agitation in tap water. 4. Transfer the slide rack into a fresh slide holder containing 150 ml of Dako target retrieval solution (15 ml of stock into 135 ml H2O) that has been prewarmed in a 95°C water bath. Leave in the water bath for 30 min. 5. Remove the slide rack and holder from the water bath and leave on the bench at room temperature for 20 min. 6. Place slide rack under running cold tap water for 5 min.
3.4. Controls and Work-Up for Immuno fluorescence Labeling
As for any multiple immunofluorescent labeling experiments, it is important to conduct the control and work-up experiments to ensure optimal immunolabeling.
74
Robertson and Isacke
1. Ascertain whether your primary antibody works on FFPE material (see Note 7). A good source of antibodies that will work in this method are ones that have already been worked up for IHC. 2. Titrate each primary antibody individually to ascertain the minimal working concentration. If the antibody has been titrated out for immunohistochemical staining of FFPE material, this will be a useful starting point for the immunofluorescence titration. 3. Wherever possible we combine the primary antibodies together for the first incubation and combine the fluorescentconjugated secondary antibodies in the second incubation. However, it is essential to test that there is no cross reactivity of the antibodies (see step 4 below). In addition, if the experiment includes a directly conjugated primary antibody, then these are incubated with the section after the indirect labeling has been completed. 4. Ensure that there is no antibody cross reactivity and minimal nonspecific labeling by undertaking the following controls (a) perform the staining reaction omitting each primary antibody one at a time but retaining all the fluorescent-conjugated secondary antibodies and (b) perform the staining reaction replacing the primary antibodies with the same concentration of isotype-matched Ig. 3.5. Immuno fluorescence Labeling
1. Remove slides from slide rack, wipe around the section with a tissue to create an “island.” Carefully pipette 100–200 ml of PBS onto the section to prevent it from drying out. 2. Using an ImmEdge pen, draw a ring or rectangle around the “island.” 3. Shake the PBS off the slide and replace with 100–200 ml of IFF. Ensure that the whole island is wetted. If not, add an extra 200 ml of IFF and gently rock the slide until the whole island is wetted. Once the island has been wetted, shake off the IFF, and replace with 100–200 ml of IFF. 4. Place slides into a moist chamber at room temperature (see Note 8). Unless otherwise stated, all following incubations are at room temperature with gentle mixing on a rocking platform. 5. Incubate in 100–200 ml primary antibodies diluted in IFF for 60 min at room temperature or overnight at 4°C (see Notes 9 and 10). 6. Shake off the primary antibodies and replace with 100–200 ml PBS for 3 × 5 min. 7. Shake of the PBS and incubate with 100–200 ml secondary antibodies diluted to 2 mg/ml in IFF for 60 min (see Note 10).
Multiple Immunofluorescence Labeling of Formalin-Fixed Paraffin-Embedded Tissue
75
8. Wash 3 × 5 min as described above using 100–200 ml PBS containing 1.43 mM DAPI (1:10,000 dilution of stock solution). 9. Rinse slides 1 × in PBS. 10. Shake off the PBS and mount the sections in Vectashield by adding 8–10 ml of Vectashield and lowering a coverslip onto the slide. Gently squeeze out the excess mountant and seal with clear nail varnish. 11. Once slides have been stained they can be examined immediately or stored at 4°C for up to 2 weeks. If it is envisaged that the slide will be examined a number of times over a long period, then storage at −20°C is highly recommended. We have reexamined slide that have been stored at −20°C for >250 days with minimal loss of quality (1). 3.6. Confocal Microscopy
To collect immunofluorescent images from FFPE sections is essential to use a confocal microscope. Details provided below are for the Leica SP2 confocal microscope that we have in our laboratory but this method should be adaptable for any confocal microscope that can collect images in sequential mode. 1. Visualize the stained slides using a suitable confocal microscope. In this study, a Leica SP2 confocal scanning microscope with the laser outputs controlled via the Acousto-Optical Tunable Filter (AOTF) and the four collection windows set using the Acousto-Optical Beam Splitter (AOBS) was used. 2. The settings used were as follows: 403 nm Laser (25%) window 410–483 nm 488 nm Laser (25%) window 493–538 nm 543 nm Laser (100%) window 548–628 nm 633 nm Laser (25%) window 638–700 nm. 3. We routinely collect images using a 20× dry lens (lens specification, HCPLAPOCS NA 0.7; Leica) at 1× zoom or with a 40× oil immersion lens (lens specification, HCXPLAPO NA 1.25; Leica) at 1× or 2× zoom. For the oil immersion lens, a small drop of immersion oil is placed in the center of the coverslip prior to imaging (see Note 11). 4. Collect confocal images using the microscope in sequential mode with a line average of 4 and a format of 1,024 × 1,024 pixels. 5. The confocal microscope collects images in black and white and then assigns them a false color. Within the confocal software these colors can be reassigned (see Note 12). 6. In our laboratory, images are exported from the Leica confocal software into Adobe Photoshop CS2 v.9. In Figs. 1 and 2, no image manipulation was performed.
76
Robertson and Isacke
4. Notes 1. Although this fixative is usually referred to as paraformaldehyde, it is in fact formaldehyde prepared by depolymerizing paraformaldehyde. 2. Antibodies should be aliquoted and stored according to manufacturer’s instructions. 3. In our laboratory, we use almost exclusively Alexa Fluorconjugated secondary antibodies but other commercially available conjugates are similar in performance. For example, FITC conjugates perform similar to Alex Fluor 488 conjugates. In our laboratory, we aliquot the secondary antibodies into 50 ml lots in 0.5 ml Eppendorf tubes, seal with parafilm and store at 4°C. A sensible precaution is to date the conjugates on arrival so that you know how long you have had them. 4. We have used cut sections that have been stored for 140 days in this way with no obvious loss of staining. 5. In our original description of this method (1), we used glass slide racks and holders. Subsequently, we have been using the plastic Easy Dip slide staining system that works extremely well. 6. In our original description of this method (1), we used xylene to dewax the slides. Subsequently, we have switched to using Histoclear. Both work equally well. 7. Antibodies which work by western blotting and immunofluorescence on paraformaldehyde-fixed material do not necessarily work on FFPE material. If testing out antibodies, for the first time, for their ability to work on FFPE material, we recommend identifying a suitable control tissue containing cell types that do and do not express the antigen. If such control tissue is not available, in our laboratory, we generate FFPE pellets of paired cell lines that do and do not express the antigen for antibody testing. 8. Moist chambers are dark or foil-covered plastic boxes with moist filter or chromatography papers in the base. 9. For staining overnight at 4°C the slides are not rocked. In our laboratory, we usually stain with the primary antibodies overnight at 4°C, but for some antibodies this incubation increases the level of nonspecific background staining. It is important to determine the optimal labeling conditions for each individual antibody. 10. As shown in this chapter, it is possible to stain with two monoclonal antibodies provided they are of different subclasses. Here we show examples of staining simultaneously with a mouse monoclonal IgG1 antibody, a mouse monoclonal IgG2A
Multiple Immunofluorescence Labeling of Formalin-Fixed Paraffin-Embedded Tissue
77
antibody, and a rabbit polyclonal antibody and detecting these with Alexa Fluor 555-conjugated anti-mouse IgG1, Alex Fluor 488-conjugated anti-mouse IgG2A, and Alexa Fluor 633-conjugated anti-rabbit Ig. 11. If slides are being stored at −20°C, any immersion oil remaining on the slide has to be carefully wiped away as it tends to creep over the slide during storage. 12. In double antibody labeling experiments, we normally depict the nuclear DAPI staining in blue. In triple antibody labeling experiments, we would normally depict the DAPI staining in gray. In cases where a series of experiments are being reported, the colors can be reassigned within the confocal software to maintain continuity.
Acknowledgements This work was funded by Breakthrough Breast Cancer. We thank Jorge Reis-Filho, Kay Savage and Suzanne Parry for their help in developing this method. References 1. Robertson, D., Savage, K., Reis-Filho, J. S. and Isacke, C. M. (2008) Multiple immunofluorescence labelling of formalin-fixed paraffinembedded (FFPE) tissue. BMC Cell Biol. 9, 13. 2. Simonavicius, N., Robertson, D., Bax, D. A., Jones, C., Huijbers, I. J. and Isacke, C. M. (2008) Endosialin (CD248) is a marker of tumor-associated pericytes in high-grade glioma. Mod Pathol. 21, 308–315. 3. Pastrana, D. V., Tolstov, Y. L., Becker, J. C., Moore, P. S., Chang, Y. and Buck, C. B. (2009) Quantitation of human seroresponsiveness to Merkel cell polyomavirus. PLoS Pathog. 5, e1000578. 4. Kendrick, H., Regan, J. L., Magnay, F. A., Grigoriadis, A., Mitsopoulos, C., Zvelebil,
M., et al. (2008) Transcriptome analysis of mammary epithelial subpopulations identifies novel determinants of lineage commitment and cell fate. BMC Genomics 9, 591. 5. Kosaka, N., Ogawa, M., Choyke, P. L., Karassina, N., Corona, C., McDougall, M., et al. (2009) In vivo stable tumor-specific painting in various colors using dehalogenasebased protein-tag fluorescent ligands. Bioconjug Chem. 20, 1367–1374. 6. MacFadyen, J. R., Haworth, O., Roberston, D., Hardie, D., Webster, M. T., Morris, H. R., et al. (2005) Endosialin (TEM1, CD248) is a marker of stromal fibroblasts and is not selectively expressed on tumour endothelium. FEBS Lett. 579, 2569–2575.
wwwwwwwwwwwwwww
Chapter 5 Microwaves for Chromogenic In Situ Hybridization Anthony S.-Y. Leong and Zenobia Haffajee Abstract In situ hybridization can be employed in formalin-fixed, paraffin-embedded tissue sections (FFPT) and allows direct visualization of amplified genes and chromosomes in individual cell nuclei. Fluorescence in situ hybridization (FISH) is the most widely employed method, but the fluorescence preparations suffer from the main disadvantages of fading over time and poor visualization, the latter making it difficult to accurately separate invasive from in situ cancer cells. Chromogenic in situ hybridization (CISH) is a viable alternative to FISH in FFPT as it employs a peroxidase reaction to visualize the chromogen thus allowing the convenience of bright field microscopy and the correlation of the visualized gene amplification with cytomorphology. It is relatively less expensive and allows a permanent record, with several studies attesting to its validity. As with FISH, heat pretreatment and enzyme digestion are two critical components of the protocol. We describe a protocol for CISH in which a microwave-induced target retrieval step is introduced as a replacement for heat pretreatment. The same procedure is performed following enzyme digestion to produce consistent signals in amplified and nonamplified cells that are both larger in size and numbers when compared with those produced by the conventional protocol. Key words: Fluorescence in situ hybridization, Chromogenic in situ hybridization, Microwaves, Formalin-fixed, Paraffin-embedded tissue, Genes
1. Introduction Fluorescence in situ hybridization (FISH) allows the identification of amplifications and translocations of genomic components in human neoplasms and has been employed extensively in the diagnosis of hematolymphoid malignancies, soft tissue sarcomas, and childhood tumors. The technique has several advantages including its application in formalin-fixed, paraffin-embedded tissue sections (FFPT) and direct visualization of amplified genes and chromosomes in individual cell nuclei. Also, several disadvantages are inherent in the technique. Fluorescence preparations are not permanent and there is the requirement for specialized fluorescence Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_5, © Springer Science+Business Media, LLC 2011
79
80
Leong and Haffajee
microscope and filters, and the technique is relatively costly (about US$100 per test for reagents alone, excluding probes). There is also the difficulty of accurately separating invasive from in situ cancer cells in fluorescence microscopy that make chromogenic in situ hybridization (CISH) a viable alternative for this technique. As CISH employs a peroxidase reaction to visualize the chromogen, it allows the convenience of bright field microscopy and the direct visualization of gene amplification and corresponding tissue cytomorphology. It is relatively less expensive and provides a permanent record. Several publications attest to the validity of CISH. We describe our method for CISH in which microwave irradiation is employed to produce consistently enhanced signals (1). While the protocol is described for the Zymed Spot-Light HER2 CISH kit (84-0146) (Zymed/Invitrogen Laboratories, San Francisco, CA, USA), it is applicable with any other probe. The procedure extends over 2 days.
2. Materials and Methods Tissue samples should be fixed for at least 8 h in 10% buffered formalin and processed in the routine manner. Five micron-thick FFPT sections are employed for this technique. 2.1. MicrowaveEnhanced Chromogenic In Situ Hybridization for FFPT 2.1.1. Day 1
1. Dry 5-mm thick FFPT sections on adhesive-coated slides overnight at 75°C. 2. Dewax in xylene two times, 5 min each. 3. Hydrate through graded ethanol (100, 80, and 70%), 3 min each. 4. Wash in distilled water three times, 2 min each. 5. Place in citrate buffer 10 mM/L pH 6.0 and microwave at 98°C for 10 min. 6. Allow sections to cool in buffer for 10 min. 7. Wash in distilled water three times, 2 min each. 8. Apply two drops of enzyme (Reagent B) to section and cover with glass coverslip, 7 min at 22°C. 9. Wash in distilled water three times, 2 min each. 10. Place in citrate buffer 10 mM/L pH 6.0 and microwave at 98°C for 10 min. 11. Retain sections in container of buffer and cool with running tap water for 10 min. 12. Wash in distilled water three times, 2 min each. 13. Dehydrate in 70% ethanol, 2 min.
Microwaves for Chromogenic In Situ Hybridization
81
14. Dehydrate in 85% ethanol, 2 min. 15. Dehydrate in 95% ethanol, 2 min. 16. Dehydrate in 100% ethanol, 2 min. 17. Air dry sections at 22°C for 30 min. 18. Layer 15 mL of probe (Reagent C) to cover the entire section and apply coverslip. Carefully spread the reagent evenly by depressing the coverslip with a cotton tip stick. Seal the coverslip with rubber cement (Vulcanizing Rubber Solution, Weldtite Products Ltd, Barton-on-Humber, UK). 19. Place the sections in a humidified hybridizer chamber and incubate for 5 min at 95°C. The automated process will continue incubation at 37°C for 16 h. Steps 10–18 can be automated in the Vysis Processor (VP 2000, Vysis, Santa Clara, CA, USA). We employ a computer-controlled microwave oven that allows accurate time and temperature settings. This can be substituted by a domestic microwave oven provided the time taken to heat a fixed volume of buffer and slides in the Coplin jar to the required temperature is known. Care should also be taken to place the jar in the same position within the oven cavity as heating is notoriously uneven in domestic ovens (2). 2.1.2. Day 2
1. Fill three Coplin jars with SCC reagent (Reagent D, sodium chloride–sodium citrate buffer). Place one jar into a waterbath and heat to 75°C, leaving the remaining two jars at room temperature. 2. Carefully peel off the rubber cement. Place slide in the jar containing Reagent D at room temperature. After a few seconds, the coverslip can be removed without damage to the section. Place slide in buffer in the other jar at room temperature for 5 min before transferring to the jar with buffer at 75°C for 5 min. 3. Wash slide in distilled water three times, 2 min each. 4. Immerse slides in 3% hydrogen peroxide in 100% methanol for 10 min. 5. Wash in phosphate-buffered saline/Tween 20 (0.01%) buffer three times, 2 min each. 6. Cover section with two to three drops of Reagent F (CASblock). Leave for 10 min. 7. Blot off reagent (do not rinse off). 8. Apply mouse anti-Dig antibody (Reagent G), two drops/ slide. Apply coverslip and incubate for 30 min at 22°C. 9. Wash in PBS/Tween 20 (0.01%) buffer three times, 2 min each.
82
Leong and Haffajee
10. Apply two drops/slide of polymerized HRP anti-mouse (Reagent H) to cover the entire section. Apply coverslip and incubate at 22°C for 30 min. 11. Prepare DAB solution immediately before use. This is done by adding one drop of each reagent (I1, I2, and I3) to 1 mL distilled water. Mix well. 12. Wash slides in PBS/Tween 20 (0.01%) buffer three times, 2 min each. 13. Apply DAB solution to cover the entire section (two to three drops/slide), apply coverslip, and incubate for 45 min. 14. Wash slides in running tap water, 2 min. 15. Counterstain with Mayer’s hematoxylin, 20 s. 16. Wash slides in running tap water, 2 min. 17. Dehydrate through graded ethanol (70, 85, 95, and 100%), 2 min each. 18. Apply mounting solution and coverslip. Gene copies visualized by CISH are clearly identifiable through a 40× objective in sections counterstained with Mayer’s hematoxylin. Individual gene copies appear as a small-rounded single dot. In the absence of amplification, the cells show typically one to two dots per nucleus when diploid (Fig. 1) and three to five dots in polysomy (Fig. 2). When more than five dots are present, it signifies gene amplification (Fig. 3), and when there is high amplification, there are multiple dots which tend to fuse to form large and small clusters (Fig. 4).
Fig. 1. Benign breast ductal epithelium showing two dots per cell (anti-Her2 probe).
Microwaves for Chromogenic In Situ Hybridization
83
Fig. 2. Low-grade ductal carcinoma displaying cells with three to four signals per nucleus (anti-Her2 probe). Confirmed to be polysomic by CISH for chromosome 17 and by FISH (data not shown).
Fig. 3. Grade 3 infiltrating ductal carcinoma showing low amplification for Her2. There are increased numbers of single dots in the nuclei; the mean signal count was 6.5 dots per cell (anti-Her2 probe).
Fig. 4. Adjacent sections showing high amplification in grade 3 infiltrating ductal carcinoma with a mean count of 9.8 signals per cell. Both preparations, (a) microwave protocol and (b) conventional manufacturer’s protocol, show fusion of signals into small and large clusters, but many more signals are seen in (a)
84
Leong and Haffajee
3. Notes The above protocol differs from that recommended by the manufacturer by the introduction of two steps in Day 1. Step 5 replaces the heat pretreatment recommended by the manufacturer (3). This step is described as “the most critical step for successful CISH performance.” The manufacturer recommends boiling the slides at ³98°C over a hotplate for 15 min in a proprietary reagent (Reagent A) provided with the kit. We have substituted this step with microwave irradiation of the slide in 10 mM/L citrate buffer, pH 6.0 at 98°C for 10 min. The same procedure is introduced in Step 10 after enzyme digestion and washing. The introduction of the two microwave target retrieval steps resulted in consistent enhancement of signals compared with the protocol recommended by the manufacturer. When consecutive sections stained according to the manufacturer’s protocol and our modification with microwave exposure were compared, our preparations showed significantly larger and more dots in similar areas of adjacent tissue sections (Figs. 4 and 5). One explanation for the larger dots may be swelling of the cells that occurs following microwave irradiation. Besides the increase in size of the signals (dots), there is concomitant increase in nuclear size. The explanation for the greater sensitivity of the microwave protocol, however, remains speculative. The enhancement of immunostaining following microwave antigen retrieval has been hailed as a “revolutionary” (4, 5), and there are a number of proposals to explain its mechanism of action. Basic to the understanding of antigen retrieval is the concept that fixation in formaldehyde results in crosslinking of amino acid side chains of proteins which contribute to their stabilization (6–9). Heat, the common denominator in antigen retrieval, produced by a variety of
Fig. 5. Comparison of mean signal count in 30 cells from nine randomly selected cases obtained by the microwave protocol (white ribbon) and the conventional manufacturer’s protocol (gray ribbon).
Microwaves for Chromogenic In Situ Hybridization
85
methods including microwaves, is hypothesized to cause protein denaturation. This is based on the observation that some antigens and endogenous enzymatic activities may be lost after heat treatment (10). Heat is known to induce reversal of the various chemical modifications of the protein structure produced by interaction with formaldehyde. These include the loosening or breakage of cross-linkages, hydrolysis of Schiff bases, as well as multiple other actions including extraction of diffusible blocking proteins, precipitation of proteins, and dehydration of the tissue sections to allow better penetration of antibody and increased accessibility to target epitopes (11). All or some of these mechanisms may also be achieved by other antigen retrieval methods including enzyme digestion and changes in pH. The last traces of paraffin may also be mobilized, thereby improving antibody penetration. Formalin fixation also results in the formation of a cage-like complex of proteins with calcium ions and the release of calcium may require energy in the form of high temperature heating or through calcium chelation by citrate (12). While this latter explanation may be operative for some proteins, it does not account for the loss of immunoreactivity for many other antigens (13). The role played by kinetics is also not understood. Microwaves generate heat and the focus has been on heat as the effective factor in antigen retrieval. However, the electromagnetic field of microwaves oscillates through 180° at 2.45 billion cycles per second. The very rapid rotation of protein molecules in the field of exposure can itself hasten chemical reactions, with heat being generated only as an epiphenomenon. Support for a “microwave effect” comes from several reported experiments that have demonstrated that antibody–antigen reactions are greatly accelerated by exposure to microwaves in the absence of significant rise in temperature (14–17). Ultrasound has also been demonstrated to significantly accelerate antibody–antigen reaction and this physical modality produces negligible amounts of heat (18). A more recent suggestion is that antibodies appear to recognize linear protein epitopes in FFPT, and antigen retrieval may simply remove crosslinked proteins that are sterically interfering with antibody binding (19, 20). In contrast to its extensive application for antigen retrieval in diagnostic and research immunohistochemistry, microwaves have had comparatively little use in the demonstration of RNA and DNA. The exposure of FFPT sections in citrate buffer to microwaves in a manner similar to that applied for antigen retrieval produces enhanced signal detection of both mRNA (21) and DNA (22, 23). By computer-assisted quantification of the radioactive signals, the enhancement was evaluated to be 60–120% above that obtained with the conventional method (21). The combination of microwave irradiation followed by short proteolytic
86
Leong and Haffajee
digestion produced a cumulative effect on tissue and target sequences that results in a significantly improved ISH signal detection compared with enzyme digestion alone or microwave retrieval alone (24–26). Unlike proteolytic digestion which needs to be applied for sufficient durations to produce the desired staining intensity, the combined method of microwaves and shortened periods of enzyme digestion resulted in morphology that was significantly better than the proteolytic method alone as the latter tended to result in over digestion and destruction of morphology (24). Microwave treatment generally decreased the amount of background staining simply by reducing the time required for enzymatic digestion. Prolonged enzyme digestion disrupts cellular integrity, allowing target molecules to migrate into the background thereby increasing nonspecific background staining and decreasing signal specificity. Sperry et al. (25) examined the effects of microwaves, enzyme digestion, and simple heating in sodium chloride–sodium citrate buffer for the detection of RNA and DNA in FFPT. They found that a combination of microwave treatment for 15–20 min in 10-mM citrate buffer at pH 6.0 with a shortened digestion with proteinase K produced the best results. Not only were the positive signals enhanced but the number of positive cases detected was also increased and nucleotide sequences were detected with probe concentrations that were ineffective with other methods of retrieval. There was a tenfold difference in the minimum concentration of albumin probe using the microwaves compared with the other two methods studied. Enhanced signals were obtained irrespective of the order in which digestion and microwave irradiation were carried out (24). Using a nonradioactive in situ hybridization technique, it was found that microwave pretreatment in conjunction with enzyme digestion gave positive results in all cases for which in situ hybridization without the microwave pretreatment was not successful (27). In addition, these workers also experimented with various buffer solutions for retrieval, enzyme digestion, and durations of microwave exposure. They found that the optimal procedure varied with the target RNA and the tissue with different combinations of buffer/duration/power. Their results suggested that microwaves may facilitate the combination of in situ hybridization and immunohistochemical labeling on the same slide. Others have obtained similar results for mRNA in human infant brain tissue following 12 min of microwave pretreatment in citrate and Tris/EDTA buffers (28). This method produced optimum signal to background ratio, preserved tissue morphology, and was suitable across a broad range of fixatives when compared with protease digestion and autoclaving in citrate and Tris/EDTA buffers. Lan et al. (29) demonstrated that microwaves could be employed in two steps: as microwave pretreatment of the tissue
Microwaves for Chromogenic In Situ Hybridization
87
and microwave heating of the probe during the hybridization step, both substantially enhancing the hybridization signal obtained as well as accelerating the procedure without compromise to tissue morphology and background precipitation compared with the conventional method. The same workers have pointed out that microwave can replace proteinase K digestion for frozen sections, enhance proteinase K digestion in paraffin sections, denature mRNA structure to enable better probe access, preserve tissue architecture, and inactivate endogenous alkaline phosphatase within the tissue sections (30). The same retrieval method was equally effective for the demonstration of Epstein–Barr virus EBER RNA with quantitative confirmation of the increased sensitivity render by microwave pretreatment (31). Importantly, microwave irradiation renders RNA-ISH, a more consistent and reliable procedure (32). Microwave irradiation has also been applied for the in situ hybridization demonstration of chick Sox 11 and Sox 12 gene mRNA in semithin plastic sections (33). Three methods of retrieval were examined including microwave irradiation in 10 mM citrate buffer at pH 6.0 heated for 20 min at 450 W, enzyme digestion with proteinase K at 10 mg/mL at 37°C for 15 min, and superheating at 121°C in a pressure cooker in 10 mM citrate buffer at pH 6.0 for 3 min. Superheating proved to be the most effective method of enhancing the target signals even when reactivity appeared to be lost in tissue blocks prepared some months previously. Although publications describing in situ hybridization procedures in plastic sections are few, it appears possible to attain good results if the tissue is embedded in methyl methacrylate and pretreated by superheating in a microwave oven (26). Interestingly, the exposure of serum to microwaves facilitated the detection of hepatitis B virus DNA by the polymerase chain reaction (PCR) (34) and the direct irradiation of whole blood and hair shafts allowed sensitive genomic amplification by PCR (35). Microwave irradiation allowed DNA extraction from paraffinembedded tissues (36), including genomic DNA from Aspergillus fumigatus (37). Microwave denaturation of metaphase chromosome preparations resulted in reproducible comparative genomic hybridization analysis with a potential application in paint and DNA probe hybridization to chromosome spreads and to RNA in tissue sections (38). More recently, it was shown that microwave irradiation of the sample before incubation with the DNA probe allowed the detection of estrogen receptor and cyclic adenosine monophosphate-responsive element binding protein (CREB) by Southwestern histochemistry, whereas no signal was detected in the absence of the microwave treatment (39). Clearly, the application of microwaves to enhance the demonstration of both RNA and DNA in FFPT is not without precedent. However, like its use to enhance the demonstration of antigens in
88
Leong and Haffajee
FFPT, the mechanism of action in the labeling of DNA with the CISH technique can only be speculated. Chromosomes are tightly bound around histones, which are simple proteins with many basic groups. It is possible that a similar reaction may occur between DNA and aldehydes as occurs with aldehydes and other cellular proteins, rendering the target DNA inaccessible to the probe. Microwaves thus may have an action similar to that with antigen retrieval by producing “unmasking” of target molecules in DNA. References 1. Leong, A.S-Y., Haffajee, Z., and Clark, M. (2007) Microwave enhancement of CISH for Her2 oncogene. Appl Immunohistochem Mol Morphol. 15, 88–93. 2. Leong, A.S-Y. (2005) Microwave Technology for Light Microscopy and Ultrastructural Studies. Bangkok, Thailand; Amarin Printing and Publishing Company Ltd. 3. Zymed/Invitrogen Laboratories. Zymed Spot-Light HER2 CISH Kit (84-0146) Product Insert 2005. 4. Gown, A.M., de Wever, N., and Battifora, H. (1993) Microwave-based antigenic unmasking: A revolutionary new technique for routine immunohistochemistry. Appl Immunohistochem. 1, 256–266. 5. Leong, A.S-Y., and Milios, J. (1993) An assessment of the efficacy of the microwaveantigen retrieval procedure on a range of tissue antigens. Appl Immunohistochem. 1, 267–274. 6. Pearse, A.G.E. (1980) Histochemistry. Theoretical and Applied, 4th ed., vol 1. Edinburgh; Churchill Livingstone, 95. 7. Fraenkel-Conrat, H., Brandon, B., and Olcott, H. (1947) The reaction of formaldehyde with proteins. IV: Participation of indole groups: Gramicidin. J Biol Chem. 168, 99–118. 8. Fraenkel-Conrat, H., and Olcott, H. (1948) Reaction of formaldehyde with proteins. VI: Crosslinking between amino groups with phenol, imidazole, or indole groups. J Biol Chem. 174, 827–843. 9. Fraenkel-Conrat, H., and Olcott, H. (1948) The reaction of formaldehyde with proteins. V: Crosslinking between amino and primary amide or guanidyl groups. J Am Chem Soc. 70, 2673–2684. 10. Cattoretti, G., Peleri, S., Parravicini, C., et al. (1993) Antigen unmasking on formalin-fixed, paraffin-embedded tissue sections. J Pathol. 171, 83–98.
11. Suurmeijer, A.J.H., and Boon, M.E. (1993) Notes on the application of microwaves for antigen retrieval in paraffin and plastic tissue sections. Eur J Morphol. 31, 144–150. 12. Morgan, J.M., Navabi, H., and Jasani, B. (1997) Role of calcium chelation in hightemperature antigen retrieval at different pH values. J Pathol. 182, 233–237. 13. Shi, S-R., Gu, J., Turrens, J., et al. (2000) Development of the antigen retrieval technique: Philosophical and theoretical bases. In: Shi S-R, Gu J, Taylor CR. Eds. Antigen Retrieval Techniques: Immunohistochemical and Molecular Morphology. Natick, MA; Eaton Publishing, 17–40. 14. Hjerpe, A., Boon, M.E., and Kok, L.P. (1988) Microwave stimulation of an immunological reaction (CEA/anti-CEA) and its use in immunohistochemistry. Histochem J. 20, 388–396. 15. Choi, T-S., Whittlesey, M., Slap, S.E., et al. (1997) Microwave immunocytochemistry: Advances in temperature control. In: Gu J. Ed. Analytical Morphology: Theory, Applications, and Protocols. Natick, MA; Eaton Publishing, 91–114. 16. Takes, P.A., Kohrs, J., Krug, R., and Kewley, S. (1999) Microwave technology in immunohistochemistry: Application to avidin–biotin staining of diverse antigens. J Histotechnol. 12, 95–98. 17. Porcelli, M., Cacciapuoti, G., Fusco, S., et al. (1997) Non-thermal effects of microwaves on proteins: Thermophilic enzymes as model system. FEBS Lett. 402, 102–106. 18. Portiansky, E.L., and Gimeno, E.J. (1996) A new epitope retrieval method for the detection of structural cytokeratins in the bovine prostate tissue. Appl Immunohistochem. 4, 208–214. 19. Sompuram, S.R., Vani, K., Messana, E., and Bogen, S.A. (2004) A molecular mechanism
Microwaves for Chromogenic In Situ Hybridization
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
of formalin fixation and antigen retrieval. Am J Clin Pathol. 121, 190–199. Sompuram, S.R., Vani, K., and Bogen, S.A. (2006) A molecular model of antigen retrieval using a peptide array. Am J Clin Pathol. 125, 91–98. Sibony, M., Commo, F., Callard, P., and Gasc, J.M. (1995) Enhancement of mRNA in situ hybridization signal by microwave heating. Lab Invest. 73, 586–591. Bourinbaiar, A.S., Zacharopoulos, V.R., and Phillips, D.M. (1991) Microwave irradiationaccelerated in situ hybridization technique for HIV detection. J Virol Methods. 35, 49–58. Allan, G.M., Smyth, J.A., Todd, D., and McNulty, M.S. (1993) In situ hybridization for the detection of chicken anemia virus in formalin fixed, paraffin-embedded sections. Avian Dis. 37, 177–182. McMahon, J., and McQuaid, S. (1986) The use of microwave irradiation as a pre-treatment to in situ hybridisation for the detection of measles virus and chicken anaemia virus in formalin-fixed paraffin-embedded tissue. Histochem J. 28, 157–164. Sperry, A., Jin, L., and Lloyd, R.V. (1996) Microwave treatment enhances detection of RNA and DNA by in situ hybridization. Diagn Mol Pathol. 5, 291–296. Gu, J., Farley, R., Shi, S-R., and Taylor, C.R. (2000) Target retrieval for in situ hybridisation. In: Shi S-R, Gu J, Taylor CR. Eds. Antigen Retrieval Techniques: Immunohistochemical and Molecular Morphology. Natick, MA; Eaton Publishing, 115–128. Haas, C.J., Hirschmann, A., Sendelhofert, A., et al. (1998) Improvement of nonradioactive DNA in situ hybridization. Biotech Histochem. 73, 228–232. Relf, B.L., Machaalani, R., and Waters, K.A. (2002) Retrieval of mRNA from praffinembedded human infant brain tissue for radioactive in situ hybridisation using oligonucleotides. J Neurosci Methods. 115, 129–136. Lan, H.Y., Mu, W., Ng, Y.Y., et al. (1996) A simple, reliable, and sensitive method for nonradioactive in situ hybridization: Use of microwave heating to improve hybridization
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
89
efficiency and preserve tissue morphology. J Histochem Cytochem. 44, 281–287. Tesch, G.H., Lan, H.Y., and Nikolic-Paterson, D.J. (2006) Treatment of tissue sections for in situ hybridization. Methods Mol Biol. 326, 1–7. Oliver, K.R., Heavens, R.P., and Sirinathsinghji, D.J. (1997) Quantitative comparison of pretreatment regimens used to sensitise in situ hybridization using oligonucleotide probes on paraffin-embedded brain tissue. J Histochem Cytochem. 45, 1707–1713. Wilkens, L., von Wasielewski, R., Werner, M., et al. (1996) Microwave pretreatment improves RNA-ISH in various formalin-fixed tissues using a uniform protocol. Pathol Res Pract. 192, 588–594. Church, R.J., Hand, N.M., Rex, M., and Scotting, P.J. (1997) Non-isotopic in situ hybridisation to detect chick Sox gene mRNA in plastic-embedded tissue sections using microwave irradiation. Histochem J. 29, 625–629. Costa, J., Lopez-Labrador, F.X., SanchezTapias, J.M., et al. (1995) Microwave treatment of serum facilitates detection of hepatitis B virus DNA by the polymerase chain reaction. Results of a study in anti-HBe positive chronic hepatitis B. J Hepatol. 22, 35–42. Ohhara, M., Kurosu, Y., and Esumi, M. (1994) Direct PCR of whole blood and hair shafts by microwave treatment. Biotechniques. 17, 726. Banerjee, S.K., Makdisi, W.F., Weston, A.P., et al. (1995) Microwave-based DNA extraction from paraffin-embedded tissue for PCR amplification. Biotechniques. 18, 768–770. Bir, N., Paliwal, A., Muralidhar, K., et al. (1995) A rapid method for the isolation of genomic DNA from Aspergillus fumigatus. Prep Biochem. 25, 171–181. de Meulemeester, M., Vink, A., Jakobs, M., et al. (1996) The application of microwave denaturation in comparative genomic hybridization. Genet Anal. 13, 129–133. Shin, M., Hishikawa, Y., Izumi, S., and Koji, T. (2002) Southwestern histochemistry as a molecular histochemical tool for analysis of expression of transcription factors: application to paraffin-embedded tissue sections. Med Electron Microsc. 35, 217–224.
wwwwwwwwwwwwwww
Chapter 6 Automated Analysis of FISH-Stained HER2/neu Samples with Metafer Christian Schunck and Eiman Mohammad Abstract The HER2/neu gene (also known as ERBB2) is located on chromosome 17 (q11.2–q12) and encodes a glycoprotein known to be a member of the epidermal growth factor receptor family. Clinically, the determination of its amplification status is of utmost importance, as 10–35% of invasive human breast carcinomas come along with HER2/neu overexpression, and treatment has to be adjusted accordingly. Here a method to analyze HER2 FISH samples with digital microscopy, using the slide scanning platform Metafer PV (MetaSystems, Altlussheim, Germany), is presented. Metafer PV is a system for the automation of HER2/neu FISH assay analysis of samples hybridized with the Abbott™ PathVysion® probe kit. Key words: HER2/neu, Breast cancer, Metafer PV, Digital microscopy, PathVysion®, Tile sampling, Slide scanning
1. Introduction The HER2/neu gene (also known as ERBB2) is located on chromosome 17 (q11.2–q12) and encodes a glycoprotein known to be a member of the epidermal growth factor receptor family. Clinically, the determination of its amplification status is of utmost importance, as 10–35% of invasive human breast carcinomas come along with HER2/neu overexpression (1), and treatment has to be adjusted accordingly. The status of the HER2/neu gene and its product can be obtained in breast tissue sections either based on immunohistochemistry (IHC) assays (detecting the overexpression of the HER2 protein) or using a respective fluorescence in situ (FISH) probe kit that directly detects the amplification status of the HER2/neu gene. Metastudies have shown that the FISH
Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_6, © Springer Science+Business Media, LLC 2011
91
92
Schunck and Mohammad
method is superior to IHC in terms of preciseness and cost effectiveness (2). Usually the FISH assay is based on quantification of the overall ratio of HER2 spot signals and centromeric signals of chromosome 17 (CEP®-17) in 20 tumor cells. Analysis can be performed using a conventional fluorescence microscope. However, routine analysis of quantitative assays such as the HER2 FISH test bears two major disadvantages, which are (1) the time that has to be invested for the microscopic analysis and (2) the likeliness to generate false or imprecise results due to misinterpretation of signals and scoring biases (3, 4). Here a method to analyze HER2 FISH samples with digital microscopy, using the slide scanning platform Metafer PV (MetaSystems, Altlussheim, Germany), is presented. Metafer PV is a system for the automation of HER2 FISH assay analysis of samples hybridized with the Abbott™ PathVysion® probe kit (Abbott Molecular, Abbott Park, IL, USA).
2. Materials 2.1. S amples
For digital analyses of HER2/neu amplification with the Metafer PV system, preparations of formalin-fixed, paraffinembedded human breast cancer tissue specimens are required. Slides should be labeled with the Abbott PathVysion® FISH probe kit for detecting amplification of the HER2/neu gene. The kit consists of two-labeled DNA probes, in detail (1) the HER2 probe spans the entire HER2 gene, which is labeled with a red fluorochrome and (2) the centromeric probe of chromosome 17 (17p11.1–q11.1), which is labeled with a green fluorochrome. For use with Metafer PV, the hybridization of samples should be precisely done following manufacturers’ recommendations. The following materials are required: 1. Paraffin Pretreatment Reagent Kit II (Abbott Molecular, Abbott Park, IL, USA): pretreatment solution sodium thiocyanate NaSCN (5 × 50 ml; store at 2–25°C), protease (Pepsin 2,500–3,000 U/mg protease, lyophilized 5 × 250 mg; store at −20°C), protease buffer (NaCl solution, pH 2, 5 × 50 ml; store at 2–25°C), and wash buffer (2× SSC, pH 7.0, 2 × 250 ml; store at 2–25°C). 2. Wash buffer 20× SSC (3 M sodium chloride and 0.3 M sodium citrate): 66 g 20× SSC dissolved in 250 ml purified water (pH 5.3; store at room temperature for up to 6 months). 3. Posthybridization wash buffer 2× SSC/0.3% NP-40: 100 ml 20× SSC (pH adjusted to 5.3 with 1.0 N HCl or 1.0 N
Automated Analysis of FISH-Stained HER2/neu Samples with Metafer
93
NaOH) mixed with 847 ml deionized water and 3 ml of 0.3% NP-40 (pH adjusted to 7.0–7.5 with 1 N NaOH), brought to a total volume of 1 l with purified water, and filtered (store at room temperature up to 6 months). 4. 10% buffered formalin: 4% formaldehyde in PBS (store at room temperature). 5. Denaturing solution (70% formamide/2× SSC): 5 ml 20× SSC washing solution with 10 ml deionized water and 35 ml formamide (pH 7.0–8.0; store at 8°C). Use solution for up to 1 week and discard if it becomes cloudy (store at 2–8°C in a tightly capped container when not in use). 6. Abbott PathVysion® HER2/neu DNA Probe Kit (see Note 1): multicolor DNA FISH probes 200 ml (store at −20°C in the dark), DAPI counterstain 300 ml (store at −20°C in the dark), 20× SCC (66 g; store at −20 to 25°C), and 0.3% NP-40 (3 ml; store at −20°C). 7. 0.2 N HCl, xylene (scientific safety solvents; store at room temperature). 8. Ethanol solutions: v/v 70, 85, and 100% using 100% ethanol and purified water (see Note 2; store at room temperature in tightly capped containers when not in use). 9. Coated glass slides: slides are immersed in mix solution of 4 ml of 3-aminopropyltriethoxysiline (EPAC) and 224 ml of acetone for 25 s, immersed twice in acetone for 25 s, and dipped in deionized water (dried at 37°C overnight). 2.2. S ystem
Metafer PV is a slide scanning system based on the hardware platform Metafer (Fig. 1a). The central unit of any Metafer system is a microcomputer (DELL, Langen, Germany). For image acquisition, a high-resolution monochrome megapixel chargecoupled device (CCD) camera (CoolCube 1m, MetaSystems, Germany) with a resolution of 1,360 × 1,024 pixels (2/3″ CCD; Pixel size 6.45 mm × 6.45 mm) is used. Metafer is connected to the motorized microscope AxioImager Z2 (Carl Zeiss, Göttingen, Germany) and takes full advantage of the microscope components for automated focusing, light source adjustment (for bright-field imaging), and fluorescence filter change. Movement of slides in X- and Y-direction is done using a motorized 8-slides microscope stage (Maerzhaeuser, Wetzlar, Germany), which is controlled by a TANGO controller unit (also Maerzhaeuser). Optionally the 8-slides capacity can be extended to 80 slides per scanning session with the external slide feeder, which loads up to ten 8-slide frames to the scanning stage automatically and unattended.
94
Schunck and Mohammad
Fig. 1. (a) The Metafer PV system with SlideFeeder. (b) Detail of a results gallery with 8 tiles showing DAPI counterstain, HER2- and CEP®-17 FISH signals. The ratio between the two signal counts is displayed in the top right of each gallery image. (c) Correlation of manual and automated analyses of 212 tissue section samples labeled with the PathVysion® HER2 FISH probe kit. All samples were visually scored following the protocol described in the PathVysion® package insert (x-axis). For automated analysis, tumor regions suitable for analysis were interactively selected using the respective function of a Metafer PV system, and subsequently analyzed automatically at a magnification of 40× (y-axis). Symbols indicate the single ratios between HER2 and CEP®-17 FISH signals for each sample.
3. Methods 3.1. Preparation of Slides from Formalin-Fixed, Paraffin-Embedded Tissue (see Note 3)
1. H&E staining can be conducted prior to performing FISH assay to identify target areas (cancer cells). 2. Paraffin sections are cut into 4–6-mm thickness using microtome. The sections are floated on a protein-free bath set at 40°C. 3. Paraffin sections are mounted on the positive side of an organosaline-coated slide to minimize detachment of tissue from the slide during FISH Assay and allowed to dry at 37°C overnight. 4. The slides are baked in oven overnight at 56°C.
Automated Analysis of FISH-Stained HER2/neu Samples with Metafer
95
3.2. Deparaffinizing of Formalin-Fixed, Paraffin-Embedded Tissue Sections (see Note 4)
1. The slides are immersed in xylene for 10 min.
3.3. Preparation of the Paraffin Pretreatment Reagents
1. One bottle (50 ml) of pretreatment solution (NaSCN) is added into a Coplin jar. Then the jar is placed in 80°± 1°C water bath prior to deparaffinizing the slides. Discard after use.
2. Step 1 is repeated twice using fresh xylene each time. 3. The slides then dehydrated using 100% ethanol for 5 min. This step is repeated once using fresh 100% ethanol. 4. The slides are allowed to air dry or placed on a 45–50°C slide warmer for 2–5 min.
2. One bottle (50 ml) of protease buffer (NaCl) is poured into a Coplin jar, and the jar is placed in 37 ± 1°C water bath. The protease solution is prepared by adding 250 mg (one tube) of protease powder to the 37°± 1°C Protease buffer. Discard after use. Adjust pH to 2.0 using 1.0 N HCl or 1.0 N NaOH. 3. Two Coplin jars containing 70 ml of wash buffer 2× SCC, pH 7 are prepared and used at room temperature. 4. Two Coplin jars, one containing 70 ml deionized water and the other containing 70 ml 0.2 N HCl are prepared and used at room temperature. 5. Denaturing solution is put in a Coplin jar and placed in 73 ± 1°C water bath for at least 30 min prior to use.
3.4. Slides Pretreatment Procedure
1. The slides are immersed in 0.2 N HCl for 20 min and then dipped in deionized water for 3 min at room temperature. 2. Slides are removed from deionized water and immersed in wash buffer for 2× SCC for 3 min at room temperature. 3. Slides are immersed in the pretreatment solution jar at 80 ± 1°C for 20 min (see Notes 5 and 6). 4. Slides are immersed in purified water for 1 min. 5. Slides are immersed in wash buffer for 5 min. This step is repeated once using fresh wash buffer. 6. Once the excessive buffer has been removed, the slides are immersed in protease solution jar at 37 ± 1°C for 25 min (see Note 7). 7. Step 5 is repeated. 8. Slides are air dried for 2–5 min and then immersed in 10% buffered formalin jar for 10 min at room temperature (see Note 8). 9. Step 5 is repeated. 10. Slides are air dried or placed on a 45–50°C slide warmer for 2–5 min.
96
Schunck and Mohammad
3.5. Fluorescence In Situ Hybridization Procedure
1. The humidity chamber (e.g., HYBRADE, OmniSlide) is switched on and set to a temperature of 37°C to allow it to heat equilibrate.
3.5.1. Denaturation of Specimen DNA
2. The slides are immersed in denaturing solution jar at 72 ± 1°C for 5 min. Not more than six slides at one time per Coplin jar should be denatured (see Note 9). 3. Slides are immediately dehydrated into 70, 85, and 100% ethanol wash solution for 1 min each at room temperature. 4. The slides are air dried for 2–5 min.
3.5.2. Hybridization
1. The probe mixture is allowed to warm to room temperature to ensure that the viscosity decreases sufficiently for allowing accurate pipetting. 2. 10 ml of probe mixture is applied to a target area, and a 22mm × 22mm glass coverslip is carefully inverted on a slide (see Note 10). 3. Rubber cement is used to seal the coverslip. 4. Slides are incubated in the hybridizer (humidity chamber) with a tight lid at 37°C for 16 h in the dark (see Note 11).
3.5.3. Posthybridization Washes (Dark Area)
1. Three Coplin jars containing 70 ml of 2× SCC/0.3% NP-40 posthybridization wash buffer are prepared. One Coplin jar is placed in 72 ± 1°C water bath at least 30 min or until solution temperature has reached 72 ± 1°C (see Note 12). The other two Coplin jars are placed at room temperature. All wash solutions should be discarded after 1 day use. 2. The slides are removed from the hybridizer, and the rubber cement is removed by gently pulling up the sealant with forceps (see Note 13). 3. Slides are immersed in the 2× SCC/0.3% NP-40 posthybridization wash buffer jar at room temperature until the coverslip floats off. 4. Slides are removed from wash buffer after 5 min of floated coverslip. 5. The excess liquid is removed from the slides by wicking off the edge of the slides. The slides are then immersed in the prewarmed posthybridization wash buffer jar at 72 ± 1°C for exactly 2 min. (see Note 14). 6. Slides are removed from the prewarmed wash buffer and immersed in the final room temperature 2× SCC/0.3% NP-40 posthybridization wash buffer jar for 1 min.
3.5.4. Finishing the Hybridization (Dark Area)
1. Slides are removed from the final wash buffer jar and air dried for 2–5 min in the dark at room temperature in an upright position (a closed drawer or shelf inside cabinet is sufficient as dark place).
Automated Analysis of FISH-Stained HER2/neu Samples with Metafer
97
2. The DAPI counterstain of 10 ml (1,000 ng/ml DAPI in phenylenediamine dihydrochloride, glycerol, and buffer) is applied to the target area, then the coverslip is carefully inverted on a slide (see Note 15). 3. The coverslip is carefully sealed with nail varnish. 4. Slides are stored in the dark until analysis (see Note 16). 3.6. Analysis 3.6.1. Preparing the Search
1. The Metafer PV system is started, usually by double-clicking the respective icon on the desktop of the computer. After the start, the system performs a sequence of hardware initialization steps. To guarantee proper condition of system and peripherals, a number of system service reminders may be shown after start, usually asking for the following procedures: a. Fluorescence lamp adjustment b. Data backup to removable media c. Checking the focus of the collector lens d. Running the Stage End Switch test and Stage Movements test. Each reminder can be confirmed using the button [DONE] or skipped using the button [CLOSE]. In the first case, the reminder will be displayed as soon as it is due again (after the period of time specified in the configuration). In the latter case, it will be displayed on the next start of Metafer PV or when the command CHECK REMINDERS from the TOOLS submenu is selected. 2. Slides are loaded to the stage, with the frosted end toward the front. It is helpful to lower the stage using the respective microscope buttons during this step (usually the buttons on the lower left side of the microscope; refer to the microscope manual for details). Inserting slides is even easier if you move the stage away horizontally: select the command Move to ... Slide Center 1 from the Stage submenu to insert slides in the positions 5–8. Then use Move to ... Slide Center 8 for access to slide positions 1–4. 3. For PathVysion® HER2 analysis, the operating mode MetaCyte has to be selected from the mode menu. The mode which is currently active is indicated in the title bar of the window. 4. The command Setup is started by clicking on the respective command button. It opens a dialog window to define the search parameters. 5. The slide positions can be selected/deselected by clicking on the corresponding slide number. 6. The Data Path input box specifies where the slide files resulting from the search will be stored. Use the […] button besides
98
Schunck and Mohammad
this field to select an existing path, or type in a new location which will then be created automatically after confirmation. 7. Each slide must be given a unique Slide ID (which will be the file name of the results file) before a search can be started. Follow the naming conventions of the operating system for filenames when choosing the slide ID. 8. In the classifier column, the assay to be performed on the selected slide is entered. This is usually PathVysion VX- (‘X’ stands for the version number of the classifier) for clinical specimen hybridized with the Abbott™ PathVysion® probe kit. A classifier is the file that precisely describes the scanning and analysis procedure. 9. A slide comment can be entered if required. The comment will be stored in the slide file and printed in the report. 3.6.2. Selecting Scan Areas
Usually HER2/neu FISH tissue samples are inhomogeneous, with interweaving clusters of normal cells and tumor tissue. Therefore, it is required to generate a position list for scanning prior to the automated analysis. 1. Open the command Mark Fields in the MetaCyte menu. 2. The slide position(s) is/are selected from the slide designation fields at the lower part of the screen. To cross-check the correctness of the slide ID originally given in the slide setup, it is not visible at this stage and has to be entered again after the corresponding slide info field is selected. To be able to read the slide ID from the slide label area, the stage is moved toward the front. The system only accepts the slide if the same name as in the Setup dialog is entered. 3. The new slide file will be created and the system will select the microscope lens (if enabled in the specified classifier; usually a 40×/0.75 dry objective is used) and the fluorescence filter for DAPI counterstain, and the fluorescence light path shutter is opened. 4. The scanning stage can now be moved manually using the trackball to target regions on the sample. The positions of target regions for later relocation can be recorded either by pressing the right button of the trackball or by using the [Record] button. For HER2/neu amplification analysis, a minimum of ten positions have to be defined. Target regions should have a certain distance to each other; if the distance between two positions is too close, the system will not allow recording the second coordinate. For each stored position, a gallery image will be automatically captured. 5. While selecting the target positions, it is possible to check the signal quality by switching to the signal channel
Automated Analysis of FISH-Stained HER2/neu Samples with Metafer
99
uorescence filters, using the respective buttons at the fl microscope. 6. The system generates a list of coordinates reflecting the selected target regions. The coordinates are saved automatically into the slide file. 3.6.3. Running the Search
After recording the coordinates the automated part of the analysis can be started. The system will refer to the list of positions recorded in the step before and automatically analyze these positions for the presence and status of HER2/neu and CEP®-17 FISH signals. 1. Start the search by clicking the [SEARCH] button. 2. After search start, the system will move to the position of the stored reference object (which usually is the first object position that was recorded in the Mark Fields function). On screen there will be the stored reference image right of the live image. Both images should now be aligned with the trackball, using the cross-hair mark in the live image. 3. After reference object adjustment of all active slides, the user is asked to check if the microscope is setup correctly for scanning (magnification; light path). This should be checked carefully and subsequently be confirmed with the [OK] button. 4. The search starts automatically and will be performed until all positions on all active slides were scanned.
3.6.4. Results Review
1. After the search, the gallery of all detected tiles is shown (see Note 17 and Fig. 1b). For each tile, the signal count ratio is calculated, and the result is displayed in the gallery image. Certain gallery images may display the entry “→FOV X,” indicating that the fields of view (FOV) with the number X has been rejected by the field rejection algorithm (see Note 18). If the number of successfully evaluated fields is too low, it may not be possible to generate a complete results report (see Note 19). Then it might be required to add additional coordinates to the position list and repeat the scanning procedure. 2. Regions corresponding to a tile can be relocated under the microscope for further inspection by clicking onto the gallery image. The current tile is highlighted by a red frame (see Note 20). 3. The feature value diagram shows the distribution of HER2/ neu and CEP®-17 signal count ratios of all cells. It is continuously updated during the scan. After the search, it can be used for selecting subpopulations. 4. It is possible to export detailed cell data using the command PathVysion in the menu MetaCyte.
100
Schunck and Mohammad
3.6.5. Locking Slides and Printing Reports
1. After review, the slide has to be locked before the results can be reported. This can be done by clicking on the [Lock Slide] button. This procedure ensures that slide results have been reviewed and approved. 2. After reviewing and locking a slide, a report can be printed using a predefined report template. Based on the assay (PathVysion® or ProbeChek™), the correct report template will automatically be selected (for details on user responsibilities, interpretation of results, and the use of ProbeChek™ slides, see Notes 17–22).
4. Notes 1. Store the unopened PathVysion Kit as a unit at −20°C, protected from light and humidity. Exposure to light, heat, or humidity may affect the shelf life of some of the kit components and should be avoided. Components stored underconditions other than those stated on the labels may not perform properly and may affect the assay results. 2. Dilutions of 70, 85, and 100% may be used for 1 week unless evaporation occurs or the solution becomes diluted due to excessive use. Store at room temperature. 3. All biological specimens should be treated as if capable of transmitting infectious agents. The specimens should have been fixed in 10% formalin before preparing the slide. 4. Deparaffinizing the samples is very critical to ensure a complete removal of paraffin. Incomplete removal of paraffin can cause poor signaling. 5. The pretreatment procedure has been optimized for use with certain multicolor DNA FISH Probes. Quantities given are optimized for preparing five slides with one 22 × 22 mm2 coverslip. 6. Ensure that the temperature of pretreatment solution is 80 ± 1°C. If necessary, two slides may be placed back to back in each slot of the Coplin jar, with one slide placed in each end slot. For the end slide, the side of the slide with the tissue section must face the centre of the jar. 7. Ensure the temperature of the protease solution is 37 ± 1°C with pH 2.0. 8. Fixation of the sample is performed to minimize tissue loss during sample denaturation. This procedure is highly recommended when processing samples in a denaturation bath. 9. Ensure the temperature of the denaturing solution is 73°C before immersing the slides. This step is important to
Automated Analysis of FISH-Stained HER2/neu Samples with Metafer
101
denature the specimen DNA and be ready for hybridization with the probe. 10. The probe is applied using a cut pipette tip. Ensure that the probe is spread evenly under the coverslip. Air bubbles should be avoided to prevent any interference with the hybridization step. 11. Ensure that the probe is hybridized for at least 16 h to avoid poor signal quality after the probe hybridization step. The hybridizer should be ensured to be airtight, and the blackpainted hybridization box with horizontal slide found in chamber should be filled with a minimum amount of water to the side of the box to keep the chamber humid. 12. Ensure that the temperature of the posthybridization washing buffer is 72 ± 1°C for each single washing step. 13. Rubber cement should be removed very gently to avoid tissue damage. 14. Only six slides per Coplin jar should be washed in each batch. 15. The nuclei are counterstained with DAPI (4,6 diamidino-2phenylindole), a DNA-specific stain with blue fluorescence. 16. The hybridized slides are stored (with coverslip) at −20°C in the dark. After removing from −20°C storage, allow slide(s) to reach room temperature prior to analysis. 17. Metafer PV uses a number of dedicated strategies to increase the precision of automated HER2/neu amplification analyses. The goal of these procedures is to generate valid results from HER2 FISH slides hybridized with PathVysion® and to cope with problems arising from typical sample features. Major issues are (1) the segmentation of analysis regions in dense tissue sections, (2) the analysis of quality and the rejection of regions that do not match the required quality criteria, and (3) distinguishing between samples that show clearly separated, countable signals in the HER2 channel, and those where the HER2 signals are clustered. The methods to handle these issues are described in detail in ref. 3, but are briefly summarized below. 18. Segmenting single cells in tissue sections with high density using digital microscopy is often difficult, if not impossible. Therefore, Metafer PV uses a method named Tile Sampling to generate single measurement objects. Tiles are nonoverlapping, equi-sized square regions placed in the counterstain image, maximizing the total fluorescence intensity covered. This procedure stops when the total intensity within the next potential tile drops below a preset limit. This way as much cellular material and as little empty space as possible is included. For HER2/neu analysis, each single tile receives its signal count ratio, which is also displayed in the gallery.
102
Schunck and Mohammad
19. The system automatically rejects FOV, if the quality seems to be insufficient for successful analysis. The quality analysis is done separately for each color channel. A detailed description of parameters leading to a field rejection can be found in ref. 3. Rejection of a field results in a respective indication in the gallery. 20. Large clusters of FISH signals in the HER2 channel, indicating the presence of HER2 gene amplification in tandem repeats, are referred to as homogeneously staining regions (HSR). HSR cannot be analyzed using standard FISH spot counting algorithms. Therefore, the entire sample is automatically classified as HSR or non-HSR prior to spot counting, and the analysis is done with different approaches. Whereas signals in non-HSR regions are basically analyzed by counting the number of spots in a processed image at a relative intensity level of 40%, HSR samples are analyzed using area measurement, and the estimation of HER2 signal count from the area by quadratic regression without constant term. 21. In addition to the rejection of entire fields of view, it is sometimes also necessary to rejection single tiles due to poor quality. Analysis problems most often arise due to hybridization failures of the HER2 and/or the CEP®-17 probes. A total failure of hybridization would lead to a spot count of zero, and such situation is not distinguishable from the loss of signals due to the complete loss of the respective DNA or chromosome region. Therefore, tiles where one of the signal counts in the red or green channel is zero are automatically excluded from the analysis. Large artifacts, e.g., variations of the intra-cell background in the HER2 channel, may cause patterns that are not easily distinguishable from large HSRs. The system uses a concept named Area Confidence to overcome this problem. In addition to the area measurement at a relative intensity level used for the signal count estimation, the system also measures the area at relative intensity levels which are above and below. Ratios of the results are calculated and are close to 1.0, if the difference between the measurements is low, indicating a clear morphology and, thus, a high area confidence. If the ratio between two of the measurement results differs significantly from 1.0, it can be concluded that the morphology of the object is diffuse and that the area confidence is low. Tiles where the area confidence of objects in the HER2 channel is below a preset threshold value are also rejected. 22. Metafer PV is a tool to ease the analysis of HER2/neu amplification samples. The system does not have FDA approval for automated PathVysion® HER2 analysis and is not intended for clinical use in the USA. An FDA approved system offering the same functionality is available under the name AutoVysion™. The Metafer PV HER2 analysis has been developed by
Automated Analysis of FISH-Stained HER2/neu Samples with Metafer
103
MetaSystems in collaboration with Vysis® (now Abbott Molecular), the manufacturer of the PathVysion® probe kit. It is based on a large training data set of slides of varying quality from a number of different laboratories and has been extensively tested and validated (Fig. 1c). This does, however, not release the user from the responsibility to carefully control the operation of the system and the validity of results data. The user’s responsibilities include but are not limited to: (a) Validation of the system, using a sample of sufficient size, consisting of slides with typical quality, for which the correct result has been determined manually. (b) Checking the plausibility of each result by reviewing the tile gallery on screen and the specimen under the microscope. (c) Manually analyzing borderline cases in the range around the cutoff ratio of 2 (recommended: 1.5–3). (d) Checking the correct functioning of the system regularly by ProbeChek™ scans. To ensure the correct analysis of the PathVysion® clinical samples, it is necessary to run a scan of a set of ProbeChek™ slides (normal and amplified) at least once a week. These have to be hybridized together with the clinical samples to ensure the same treatment. If the analysis of these slides fails, the results of clinical samples are probably incorrect and have to be checked carefully. (e) Following all procedures and recommendations for keeping the system within the specifications. (f) After the analysis and when reviewing the results (1) checking that appropriate regions have been selected for analysis, (2) making sure that the captured images are valid and that no system component has failed during the scan, and (3) confirming that the results are consistent with the visual information displayed on screen and with the details seen through the microscope oculars. References 1. Stevens, R., Almanaseer, I., Gonzalez, M., Caglar, D., Knudson, R.A., Ketterling, R.P., Schrock, D.S., Seemayer, T.A., and Bridge, J.A. (2007) Analysis of HER2 gene amplification using an automated fluorescence in situ hybridization signal enumeration system. Journal of Molecular Diagnostics 9, 144–50. 2. Dendukuri, N., Khetani, K., McIsaac, M., and Brophy, J. (2007) Testing for HER2-positive breast cancer: a systematic review and costeffectiveness analysis. CMAJ 176, 1429–34. 3. Lörch, T., Piper, J., and Tomisek, J.D. (2002) “Tile Sampling”: a new method for the auto-
mated quantitative analysis of samples with high cell density and its application to HER2 scanning. In: Proceedings of the Third Euroconference on Quantitative Molecular Cytogenetics. Rosenön, Stockholm, Sweden. 4. Piper, J., Lörch, T., Poole, I., and Tomisek, J.D. (2002) Computing the HER2:CEP-17 ratio of tumour cells in breast cancer tissue sections by analysois of the FISH spot counts of a tiles sampling. In: Proceedings of the Third Euroconference on Quantitative Molecular Cytogenetics. Rosenön, Stockholm, Sweden.
wwwwwwwwwwwwwww
Chapter 7 Laser Capture Microdissection of FFPE Tissue Sections Bridging the Gap Between Microscopy and Molecular Analysis Renate Burgemeister Abstract Laser capture microdissection (LCM) enables researchers to combine structure identification by microscopy with structure investigation by modern molecular techniques. The main question in modern biomedical research is the understanding of cellular and molecular mechanisms. The methods to investigate pathological changes on a molecular, cellular, or tissue level become more and more exact, whereas at the same time the sample amounts available become smaller and smaller. The challenge in microscopy is the identification of structures or molecules. Today, scientists are no longer satisfied with just observing tissues and cells. They demand the ability to get access to the identified structures to bring their observations to the subcellular and genetic level. Downstream to microscopy the full toolbox of molecular biology for DNA, RNA, and protein analysis has to be applied. Key words: Laser microdissection, LCM, MicroBeam, Single-cell analysis, Quantitative RT-PCR, Image processing
1. Introduction The isolation and characterization of homogeneous cell populations are of great importance for the analysis of gene expression. For decades, tissue heterogeneity represented a challenge for scientists wishing to study isolated tissues or cells. Normal tissues contain various types of cells, and the use of heterogeneous tissue for subsequent analysis increases the variability of results. Different approaches have been attempted to develop methods for the isolation of pure samples from various sources. But traditional methods of selective purification, such as flow cytometry, needle
Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_7, © Springer Science+Business Media, LLC 2011
105
106
Burgemeister
extraction, or application of adhesive transfer membrane require mechanical interaction with the specimen. This often causes tissue artifacts and contamination. Moreover, they mostly are limited by a minimum tissue requirement volume. The recent emergence of noncontact laser-assisted microdissection has revolutionized the isolation of desired tissue and cell areas, down to single-cell populations from solid tissues (1). With the help of a cutting laser, laser microdissection can isolate tissues or cells of interest completely contact-free and therefore without contamination from the surrounding. The technique is simple, practical, and accurate.
2. Sample Preparation: Workflow
From the moment the tissue is excised, one must pay great attention to how the sample is stored, processed, extracted, and analyzed. Most tissues in routine work are processed to paraffin blocks. The process involves preservation of the tissue structure followed by the removal of all water and fats, replacing them with paraffin wax, which is then hardened and sectioned.
2.1. F ixation
To ensure the preservation and stabilization of tissue architecture and cell morphology, prompt and adequate fixation is essential. The most common fixative in routine pathology is buffered formaldehyde.
2.2. E mbedding
Paraffin wax has remained the most widely used embedding medium, which converts the tissue into a solid form to be sectioned for diagnostic histopathology in routine histological laboratories. The largest proportion of material is formalin fixed and paraffin embedded.
2.3. Sectioning (Slicing)
Thickness of microtome sections routinely varies between 5 and 15 mm. With PALM MicroBeam almost every kind of biological material can be microdissected and collected, even sections of more than 30 mm can be handled easily. The limit usually is set not by the technique, but mostly by the morphology checked by having a look in the microscope. Sections can be prepared on routine glass slides or on MembraneSlides. MembraneSlides are special slides covered with a membrane on one side. This membrane is easily laser cut together with the sample and acts as a stabilizing backbone during lifting up. This is especially important for the isolation of single cells or chromosomes as well as for live cells and small organisms. This way even large areas can be collected by a single laser pulse without affecting morphological integrity.
Laser Capture Microdissection
107
Two kinds of membrane are available: PEN (polyethylene naphthalate) membrane is highly absorptive in the UV-A range and can be used for all kind of applications. PET (polyethylene teraphtalate) membrane is especially recommendable for the isolation of chromosomes or some fluorescence applications. To overcome the hydrophobic nature of the membrane, it is recommended to irradiate with UV light at 254 nm for 30 min. The membrane becomes more hydrophilic, therefore adherence of the sections is improved. Positive side effects are sterilization and disintegration of potentially contaminating nucleic acids. To ensure easy lifting, additional adhesive substances on the slides or “Superfrost” or “charged slides” should only be applied when absolutely necessary for the attachment of poorly adhering special tissue such as some brain sections or blood vessel rings. To afford laser capture microdissection (LCM), a coverslip and standard mounting medium must not be applied. On the other hand, it is simply feasible to use old-archived sections for LCM after removing the coverslip. 2.4. S taining
3. Optional Workflow Step: Digitization of Slides
After deparaffinization, the standard staining procedures can be used for FFPE sections. Refer staining protocols for Hematoxylin/ Eosin and Cresyl Violet in Chapter 8.
Digital slides are highly resolved digital images of a whole histological specimen. The slide is loaded into a digital slide scanner. A preview camera captures an overview image of the whole sample; if desired even with labels or barcode. From this image, the software determines the scanning areas. With a high-resolution fast camera, the subsequent scan produces images in rapid sequence, which altogether composes the virtual slide (Fig. 1). One of the most important advantages of digital slides is the ability to efficiently and securely archive the information. Digital slides provide greater long-term stability than conventional slides: images retain their original quality, perfectly preserving brightfield, and fluorescent stains without any bleaching. This allows keeping the same image quality over time. Moreover, slides cannot be lost or broken. They automatically comprise various magnifications and are quickly available via data networks and are therefore available worldwide at any time. Regarding laser microdissection slide scanning has a vital benefit: The information of the slide stays stable even if parts of the tissue have been cut out and molecularly investigated. So one can combine complete slide information kept forever with results of molecular analysis from the same slide.
108
Burgemeister
Fig. 1. Digital slides are highly resolved digital images of a whole histological specimen.
An additional benefit is that annotations made on the stored image can be transferred directly to the microdissection system via a specific software (“VisDat”).
4. Microdissection with PALM MicroBeam
To obtain reliable and reproducible results, samples that will be investigated have to be well defined, pure, and free of any contamination. So the tissue areas or selected cells of interest have to be separated from any unselected surrounding material with defined spatial resolution on a microscopic level (Fig. 2a, b). The selected material then is transported into appropriate collection devices for further downstream analyses. PALM MicroBeam from Carl Zeiss works with cutting and transportation performed only by laser light without any mechanical contact (2). As the direction of sample transport is against gravity, any potential contamination is completely avoided (Fig. 2c). Many of the molecular approaches need amplification of the microdissected material, so it is indispensable to decrease the background noise level in downstream analysis this way (3). The system is equipped with a pulsed solid state laser of 355 nm. The laser is coupled into a routine research microscope
Laser Capture Microdissection
109
Fig. 2. An example of localized microdissection. (a) Outlining the required area. (b) LCM and (c) view inside the collection device after capture.
Fig. 3. The laser is focused via the objective. After the cutting process, a laser pulse transports the outlined area into a collection device.
via the epifluorescence pathway and focused via the objective lenses to a micron-sized spot diameter (Fig. 3). The width of the high precision cutting line depends on the objective used. Using a 100× oil immersion objective a line width of about 600 nm can be achieved, and this way allows even cutting of chromosomes or chromosome parts. Within the narrow laser focal spot, forces are generated that allow ablation of material (laser microdissection), while the surrounding tissue remains fully intact.
110
Burgemeister
Fig. 4. Workflow from LCM over cell or tissue lysis to molecular analysis.
As this cutting is a physical and fast process without heat transfer, adjacent biological matter or biomolecules such as DNA, RNA, or proteins are not affected. Therefore, these molecules can routinely be isolated from the specimen for downstream analyses. Using the same laser, the separated sample is lifted up and captured in a collection device that usually is the cap of a microfuge tube. The captured material subsequently will be simply spun down from the cap into the tube, analyzed directly, or used for further processing or various experiments (Fig. 4).
5. Selection of Applications Any kind of application, e.g., in cell biology, pathology, forensics, and cytogenetics, it is all about selection – isolating only which one is interested in and screening out the rest (4). LCM allows researchers to be clearly selective about the piece of tissue, cell type, or chromosome that they want to study. 5.1. Molecular Analyses
Cells isolated by LCM have been characterized by a wide variety of qualitative molecular assays, e.g., detection of loss of heterozygosity (LOH), point mutations, clonality, and lineage origin. PCR and RT-PCR are best investigated for captured specimen. Realtime PCR technology renders the reliable quantification of very small amounts of nucleic acids possible. These techniques were successfully applied for the quantification of DNA and RNA isolated from microdissected tissue sections or even single cells. The exact analysis of quantitative changes of nucleic acids during the course of pathological alterations has thus become possible (5). Unwanted cell contamination will dramatically reduce the detection level of, e.g., genetic alterations. For example, if one “unaltered” cell of a tumor section is mixed with one tumor cell
Laser Capture Microdissection
111
carrying a heterozygous gene mutation (that means a mutation in one of the two alleles), the gene mutation signal is only up to 25%. It is therefore of outstanding importance to obtain target cell samples as pure as possible for molecular genetic analysis. 5.2. Oncology/ Pathology
The exact quantitative detection of gene amplification, one of the most important mechanisms leading to deregulated gene expression in cancer, is often hampered by an admixture of nonneoplastic bystander cells and stroma. Especially in small-sized lesions, such as preinvasive precursor lesions, the admixture of unaltered cells may lead to a misinterpreted result. Laser microdissection is the state-of-the-art technology for the preparation of tumor cells without any contamination with nontumor cells. A suitable assay that efficiently recognizes invasive tumor cells appearing in the blood circulation is the filtration of peripheral blood through a Teflon filter. With this assay, it is possible to detect one single tumor cell in 1 ml of peripheral blood. Subsequent collection of the tumor cells allows DNA amplification and screening for genetic abnormalities in target sequences (6). This is a powerful tool for molecular analyses in diagnostic and experimental tumor pathology, which may help to provide new insights into the molecular basis of neoplasia, in particular of carcinogenesis and tumor heterogeneity (7, 8).
5.3. Brain Research
Many of the current methods for analyzing the genome, proteome, or most frequently, the transcriptome, rely on homogenization and extraction of the elements of interest. The complexity of the brain makes the investigation of anatomically defined regions using manual dissection techniques problematic. It analyzes only the average of many different cell types, but effects specific for certain cell types are obscured. Laser microdissection allows the efficient isolation of single cells or cell groups, e.g., from patients of neurodegenerative diseases, e.g., Alzheimer’s disease, Creutzfeld–Jakob disease, Parkinson, Multiple Sclerosis, stroke, trauma, and so on avoiding any contamination of surrounding tissue components, simultaneously leaving the intracellular structure and molecules intact (9).
5.4. Single-Cell Analysis
A fundamental perspective can be achieved by targeting single cells for analysis. Many biological disciplines have the goal to elucidate the causes of cellular differentiation on the single-cell level. Human maturation, regeneration, and genetic diseases – all lie hidden in a single cell that was originally part of the genetically clonal, multicellular organism. To explore epigenetics, proteomics, and cell signaling, the investigation of single cells is a good way. However, single-cell studies have their own difficulties, such as making sure to really investigate only one single cell without any contaminating cell or cell fragment.
112
Burgemeister
Several studies suggest that pooled cell samples thought to be homogeneous, often are composed of cells with quite different phenotypes. To be sure to really investigate pure single cells one has to apply LCM (10). Specifically cancer and drug discovery research within the field of genomics shows a great potential (11). Chromosomal aberrations, often found in cancer, could be better investigated when comparing a single cancerous cell with its normal counterpart. This provides a basis for searching for the relevant genes in specific stages. These genetic alterations may permit the discrimination between normal, premalignant, and tumor cells. In addition, comparing single cells from the treated population to the untreated population to evaluate genomic effects can be used to screen drug candidates. 5.5. Cytogenetics
Due to the precision of laser microdissection with PALM MicroBeam from Carl Zeiss, it is possible to cut and catapult single chromosomes or chromosome fragments without contamination of unwanted chromosomal material. Chromosome-specific DNA probes can be generated by isolating DNA from whole chromosomes or chromosomal subregions and its subsequent universal amplification by degenerate oligo-primed (DOP)-PCR. DNA of only one single-captured chromosome is sufficient to generate a painting probe. The availability of such paint probes has become an important tool for cytogenetic analysis with wide applications ranging from research studies to diagnostic use in clinical genetics (12–14).
5.6. Live Cells from Cell Cultures
An innovation in the field of laser microdissection is the laserbased isolation of live cells out of a cell culture. Individual or small groups of cultured cells, even from primary cultures or stem cell preparations, can be used for direct molecular analysis or recultivation. The reculturing method demonstrates a new and easy way for clonally expansion of cells. As the viability of catapulted cells is not affected, different cell types, discriminable by morphology, fluorescence or transfection markers can be isolated fast and reliably by LCM. The work with selected live cells is extremely facilitated with this new approach and opens a wide field of new applications and research possibilities in molecular biology and medicine as well as cell biology. Usually, a preparation of stem cells or selective elimination of specific cells from a culture is not easy to perform, but simple and fast with LCM. In some cases, sterility is an important point. The whole LCM process can be performed sterile in the closed dish with the help of PALM LiveCell Collector as well as with an on-stage incubator. Proliferation rates are very good, as no additions, e.g., poly-l-lysine have to be added. As well trypsination is not necessary. This is of immense importance as even single cells can be
Laser Capture Microdissection
113
captured and cloned this way. Yet as sensitive cells, such as stem cells, are easily to handle in this vein. LCM has no effect on the phenotype: After LCM and clonally expansion the cells keep their stem cell character. It has also been proven that even the genotype is unaffected (15). Additionally, cells in the original PALM DuplexDish used for LCM can be long-term cultivated and ensure reproducible experiments. Outstanding optics and contrast methods such as PlasDIC are unique tools for monitoring cell morphology (16, 17).
6. Automated Cell Recognition: Image Processing
The requirements of biomedical sciences are growing ever more complex. Software-based detection methods therefore are of high importance for either high throughput work or finding rare events in a justifiable expenditure of time. If tumor areas have to be found in large areas of nontumor areas or if single cells have to be detected on a preparation slide, time-saving methods are encouraged and sometimes are even prerequisites for specific projects. Modern software-based detection methods allow finding the wanted structures in a fast and reliable way. The high degree of automation realized in the latest generation of PALM MicroBeam systems can be augmented by image analyzing software modules allowing automated fast scanning for specimen identification and image processing. Coupled with any of these software modules, PALM MicroBeam in an automated manner is able to scan, detect, isolate, and finally capture the specimen of interest, e.g., prespecified tissue areas such as tumor and nontumor areas, fluorescent-labeled rare cells, metaphases, or FISH-treated cells. Auto-marked areas can subsequently be extracted automatically by the appropriate laser functions. These versatile-automated scanning software modules also comprise the advantage of fast and reliable detection and auto-evaluation of particular cells, cell components, or chromosomes based on optimized classifiers by means of morphological phenotypes. The efficient detection algorithms are trained by specialists together with the user to achieve integrated, interactive classifiers for optimized recognition and accurate results (18, 19).
7. Notes Laser microdissection has established itself as a powerful technology essential to a range of downstream applications in different fields of medicine and biology.
114
Burgemeister
As LCM takes place in a completely touch-free manner and the material is collected against gravity, only selected material will be conveyed for subsequent investigation – and nothing else. The motto “What you see is what you get” provides purest samples for high-quality experiments. PALM MicroBeam bridges the gap between microscopy and molecular genetics. It is a tool that is very flexible in handling and in applications. From archival material to live cells, from individual experiments to automation, all the different applications can be operated easily. References 1. Von Eggeling, F., and Ernst, G. (2007) Microdissected tissue: an underestimated source for biomarker discovery? Biomark Med 1, 217–219. 2. Schütze, K., Niyaz, Y., Stich, M., and Buchstaller, A. (2007) Noncontact laser microdissection and catapulting for pure sample capture. Methods Cell Biol 82, 649–673. 3. George, M.D., Wehkamp, J., Kays, R.J., Leutenegger, C.M, Sabir, S., Grishina, I., et al. (2008) In vivo gene expression profilin of human intestinal epithelial cells: analysis by laser microdissection of formalin fixed tissues. BMC Genomics 9, 209–213. 4. Burgemeister, R. (2005) New aspects of laser microdissection in research and routine. J Histochem Cytochem 53, 409–412. 5. Hoffmann, A.-C., Danenberg, K.D., Taubert, H., Danenberg, P.V., and Wuerl, P. (2009) A three-gene signature for outcome in soft tissue sarcoma. Clin Cancer Res 15, 5191–5198. 6. Vona, G., Sabile, A., Louha, M., Sitruk, V., Romana, S., Schütze, K., et al. (2000) Isolation by size of epithelial tumor cells : a new method for the immunomorphological and molecular characterization of circulating tumor cells. Am J Pathol 156, 57–63. 7. Kreft, A., Springer, E., Lipka, D.B., and Kirkpatrick, Ch.J. (2009) Wild-type JAK2 secondary acute erythroleukemia developing after JAK2-V617F-mutated primary myelofibrosis. Acta Haematol 122, 36–38. 8. Rödder, S., Scherer, A., Raulf, F., Bertier, C.C., Hertig, A., Couzi, L., et al. (2009) Renal allografts with IF/TA display distinct expression profiles of metzincins and related genes. Am J Transplant 9, 517–526. 9. Churchill, M.J., Wesseling, S.L., Cowley, D., Pardo, C.A., McArthur, J.C., Brew, B.J., et al. (2009) Extensive astrocyte infection is prominent in human immunodeficiency virus-associated dementia. Ann Neurol 66, 253–258.
10. Langer, S., Geigl, B., Gangnus, R., and Speicher, M.R. (2005) Sequential application of interphase-FISH and CGH to single cells. Lab Invest 85, 582–592. 11. Sotlar, K., Bache, A., Stellmacher, F., Bültmann, B., Valent, P., and Horny, H.-P. (2008) Systemic mastocytosis associated with chronic idiopathic myelofibrosis. J Mol Diagn 10, 58–66. 12. Langer, S., Geigl, J.B., Ehnle, S., Gangnus, R., and Speicher, M. (2005) Live cell catapulting and recultivation does not change the karyotype of HCT116 tumor cells. Cancer Genet Cytogenet 161, 174–177. 13. Thalhammer, S., Langer, S., Speicher, M.R., Heckl, W., and Geigl, J.B. (2004) Generation of chromosome painting probes from single chromosomes by laser microdissection and linker-adaptor PCR. Chromosome Res 12, 337–343. 14. Fiegler, H., Geigl, J.B., Langer, S., Rigler, D., Porter, K., Unger, K., et al. (2007) High resolution array-CGH analysis of single cells. Nucleic Acids Res 35, e15. 15. Terstegge, S., Rath, B.H., Laufenberg, I., Limbach, N., Buchstaller, A., Schütze, K., et al. (2009) Laser assisted selection and passaging of human pluripotent stem cell colonies. J Biotechnol 10, 224–230. 16. Chaudhary, K.W., Barrezueta, N.X., Bauchmann, M.B., Milici, A.J., Beckius, G., Stedman, D.B., et al. (2006) Embryonic stem cells in predictive cardiotoxicity: laser capture microscopy enables assay development. Toxicol Sci 90, 149–158. 17. Duan, Y., Catana, A., Meng, Y., Yamamoto, N., He, S., Gupta, S., et al. (2007) Differentiation and enrichment of hepatocyte-like cells from human embryonic stem cells in vitro and in vivo. Stem Cells 25, 3058–3068. 18. Vandewoestyne, M., van Hoofstat, D., van Nieuwerburgh, F., and Deforce, D. (2009)
Laser Capture Microdissection Suspension fluorescence in situ hybridization (S-FISH) combined with automatic detection and laser microdissection for STR profiling of male cells in male/female mixtures. Int J Legal Med 123, 169–175.
115
19. Seitz, G., Warmann, S.W., Fuchs, J., Heitmann, H., Mahrt, J., Busse, A.-C., et al. (2008) Imaging of cell trafficking and metastases of paediatric rhabdomyosarcoma. Cell Prolif 41, 365–374.
wwwwwwwwwwwwwww
Chapter 8 Nucleic Acids Extraction from Laser Microdissected FFPE Tissue Sections Renate Burgemeister Abstract Tissue heterogeneity is a common source of unsuccessful experiments. Laser capture microdissection is a tool to prepare homogeneous tissue and cell areas as starting material for reliable and reproducible results as it allows the defined investigation of spatially different tissue areas. Nearly all samples allow the extraction of DNA. Fresh or fresh frozen samples are an ideal source for getting access to high-quality RNA. But also the large archives of formalin-fixed, paraffin-embedded (FFPE) tissue specimens are a valuable source of sample material for RNA extraction. Optimized protocols may help to make the RNA from FFPE material suitable for expression studies. Key words: Laser microdissection, Defined tissue areas, Homogeneous tissue, Biomarker identification, High-quality RNA, FFPE tissue, Expression studies, Archived biopsies
1. Introduction Genomics and proteomics techniques have become increasingly sophisticated; however, accuracy and reliability of results are strongly dependent from the purity of the sample. Recent scientific and medical applications depend on the selective procurement of defined cell or tissue populations (1). In the characterization of the early molecular genetic events of tumor development, it is of immense importance to getting access to histopathologically accurately defined tissue. To obtain high-quality DNA, mRNA, and proteins from these (often small) tissue samples and even from single cells, laser microdissection is one of the most useful techniques. Microdissected tissue material or single cells, free of contaminating and unwanted tissue components, are extremely important
Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_8, © Springer Science+Business Media, LLC 2011
117
118
Burgemeister
for producing clean data, e.g., biomarker identification in cancer diagnostics and in elucidating biological variance within tumors (2–4). Thus, these technologies will have an enormous impact on molecular pathology with the potential to improve diagnosis, estimation of prognosis, and treatment decisions in individual patients (5). In most instances, RNA can be extracted from fresh frozen material in high quality. Unfortunately sometimes the morphology of this material is inadequate. While formalin-fixation and paraffin-embedding efficiently preserve tissues for morphological analysis, the effects of the fixation process on nucleic acids make molecular analyses difficult. Nucleic acids in FFPE tissues are cross-linked and often irreversibly damaged, becoming increasingly fragmented with prolonged storage. Nevertheless, FFPE tissue samples represent a large source of unused biological material for basic clinical research. In every hospital, there are large archives of morphologically defined biopsies that exist as fixed and embedded samples. Retrospective analysis of this archived material derived from normal and pathologically altered tissues (6, 7), for which clinical data are available, could enable the correlation of molecular findings with the effect of treatment and the clinical outcome (8). Expression studies in these biopsies offer a promising extension of current methods to study the pathogenesis of many different diseases. Laser microdissection allows for the selective isolation of specific tissue or cell areas and is the best option to quantitate mRNA levels from such kind of archival material (6–8).
2. Materials 2.1. Removing Coverslips of Archived Samples 2.2. MembraneSlides
Xylene or warm water (30–50°C).
1. MembraneSlide 1.0 PEN (Carl Zeiss). 2. MembraneSlide 0.17 PEN (Carl Zeiss). 3. MembraneSlide 1.0 PET (Carl Zeiss). 4. MembraneSlide 0.17 PET (Carl Zeiss).
2.3. Heat Inactivation to Remove Nucleases
Heating oven, 180°C.
2.4. UV Treatment of the Membrane
UV light source, 254 nm.
Nucleic Acids Extraction from Laser Microdissected FFPE Tissue Sections
2.5. Poly-l-Lysine Treatment
119
Poly-l-lysine 0.1% (w/v).
2.6. Mounting FFPE Sections onto MembraneSlides
1. Drying oven, 56°C.
2.7. Staining
1. 70% (v/v) Ethanol.
2. Ethanol 100, 96, 70%.
2. RNase-free distilled water. 3. Cresyl violet acetate solution: Dissolve solid Cresyl violet acetate at a concentration of 1% (w/v) in 50% (v/v) ethanol at room temperature with agitation/stirring for several hours or overnight. Filter the staining solution before use to remove unsolubilized powder (see Note 1). For an enhancement of the staining (see Note 2). 4. Mayer’s hematoxylin solution (ready to use). 5. DEPC-treated tap water: 50% (v/v) DEPC in Aqua dest (see Note 3). 6. Eosin Y solution aqueous (ready to use). 7. Increasing ethanol series: 70% (v/v) ethanol, 96% (v/v) ethanol, 100% (v/v) ethanol. 2.8. Dry Collection of Laser Microdissected Material (Recommended for Subsequent RNA Extraction)
1. AdhesiveCap 200 (Carl Zeiss), AdhesiveCap 500 (Carl Zeiss) (see Note 3).
2.9. Wet Collection of Laser Microdissected Material
Capturing buffer: 20 ml of 0.05 M EDTA, pH 8.0, 200 ml of 1 M Tris–HCl, pH 8.0, 50 ml Igepal CA-630, and 100 ml of 20 mg/ml Proteinase K (see Notes 5–7).
2.10. RNA Extraction
2. Removal of RNases from regular tubes with 0.1% (v/v) DEPC. Add 100 ml DEPC to 100 ml of double-distilled water (see Note 4). Stir for 5–6 h at room temperature to dissolve the DEPC. Dump the reaction tubes in the DEPC solution, take care that the tubes are completely covered with liquid (not blistered!) and incubate overnight at room temperature. Autoclave the tubes together with the solution for 20 min at 121°C to inactivate the DEPC. Discard the liquid carefully and thoroughly. Dry the tubes at 50–80°C. Use the tubes as usual.
1. Incubator, 55°C. 2. Qiagen RNeasy FFPE Kit. 3. Digestion buffer containing Proteinase K: 150 mM NaCl, 100 mM Tris–HCl, pH 7.5, 0.5% Igepal, 0.5 mg/ml Proteinase K.
120
Burgemeister
2.11. DNA Extraction
1. Qiagen QIAamp Micro Kit. 2. Digestion buffer containing Proteinase K: 150 mM NaCl, 100 mM Tris–HCl, pH 7.5, 0.5% Igepal, 0.5 mg/ml Proteinase K.
3. Methods 1. To allow laser capture microdissection (LCM) (cutting and lifting) a coverslip and mounting medium must not be applied. 2. To ensure RNase-free conditions use only RNase-free solutions and materials. 3. For FFPE samples, a Proteinase K digestion step is essential. The time necessary for optimal Proteinase K digestion depends on many factors such as tissue type, fixation procedure, or element size of lifted material. An overnight digestion (12–18 h) is a good starting point for optimization, but shorter digestion times may be tested as well. To our experience, digestion of at least 3 h should be applied with any extraction procedure. 4. A forecast of the extractable amount of RNA from FFPE tissue is very difficult since many factors such as species, cell/ tissue fixation, staining, fragmentation, modification, and others will strongly influence the outcome. Any FFPE tissue block should therefore be tested individually. 3.1. Removing the Coverslip of Archived Samples
1. To allow LCM (cutting and lifting), a coverslip and mounting medium must not be applied. 2. Depending on the applied mounting medium of archived samples (whether it was xylene based or water soluble), the whole slide should be completely submerged in the respective solvent. 3. Stand up slide in a glass jar filled with either pure xylene or warm water (30–50°C). 4. Let the coverslip swim off (see Note 8). 5. Air-dry the slide.
3.2. MembraneSlides
1. MembraneSlides for LCM are used as regular glass slides. They are glass slides covered with a membrane on one side (Fig. 1). This membrane is easily cut together with the tissue and acts as a stabilizing backbone during lifting. Therefore, even large areas are captured by a single laser pulse without affecting the morphological integrity.
Nucleic Acids Extraction from Laser Microdissected FFPE Tissue Sections
121
Fig. 1. MembraneSlides are glass slides covered with a thin biochemically inert membrane. This warrants excellent DNA, RNA, or protein recovery.
2. Use of MembraneSlide is especially important for isolating single cells, chromosomes as well as live cells, or small organisms. There are slides available in different thickness (1 and 0.17 mm) covered with different membranes [polyethylene naphthalate (PEN)- or polyethylene teraphthalate (PET) membrane] (Carl Zeiss). These membranes are highly absorptive in the UV-A range, which facilitates laser cutting. 3. When working with low magnifying objectives such as 5× or 10×, both regular 1- and 0.17-mm thick slides can be used. To keep this flexibility for higher magnifications (20×, 40×, or 63×), we recommend using long-distance objectives. With those, it is possible to adapt the working distance to the different slides by moving the correction collar on the objective. Due to the short working distance of the 100× magnifying objectives, only 0.17-mm thin slides can be used for that lens. 4. Besides MembraneSlides, regular glass slides are applicable for laser microdissection. Freshly prepared slides must not be coverslipped. Archived “old” slides have to be de-coverslipped (see Subheading 3.1). 5. With the laser microdissection system PALM MicroBeam almost every kind of biological material can be microdissected and lifted directly from slides into collection devices. To facilitate easy lifting, additional adhesive substances or “Superfrost + charged slides” should only be applied when absolutely necessary for the attachment of poorly adhering special material (e.g., some brain sections or blood vessel rings). In those cases, higher laser energy is needed for lifting.
122
Burgemeister
3.3. Heat Inactivation to Remove Nucleases
1. To make MembraneSlides or glass slides RNase free, heat slides at 180°C for 4 h. Nucleases are inactivated this way. 2. MembraneSlide NF (nuclease free) is certified to be free of DNase, RNase, and human DNA. Treatments to remove nucleases are therefore not necessary.
3.4. UV Treatment of the Membrane
1. To overcome the hydrophobic nature of the membrane, it for some samples may be advisable to irradiate the MembraneSlide with UV light at 254 nm for 30 min (e.g., in a cell culture hood). The membrane gets more hydrophilic, therefore the adhesion of the sections (paraffin- and cryosections) is improved. 2. Positive side effects are sterilization and destruction of potentially contaminating nucleic acids.
3.5. Poly-l-Lysine Treatment
1. Additional coating of the slide with poly-l-lysine could be beneficial for poorly adhering materials (e.g., brain sections) and should be performed after UV treatment. Distribute a drop of the solution on top of the slide. 2. Let air-dry at room temperature for 2–3 min. Avoid any leakage of the membrane, as this might result in the impairment of LCM.
3.6. Mounting FFPE Sections onto MembraneSlides
1. Sections are mounted onto MembraneSlides the same way as routinely done using glass slides. Floating the section on warm water as well as hot plate techniques can be applied. After mounting, let dry the slides overnight in a drying oven at 56°C. To allow laser cutting and lifting, a coverslip and standard mounting medium must not be applied. 2. Paraffin will reduce the efficiency of the laser, sometimes completely inhibiting cutting and lifting. For using unstained sections, it is therefore very important to remove the paraffin before laser cutting and lifting. If applying standard staining procedures deparaffinization is routinely included in any protocol. 3. Deparaffinization of unstained sections: xylene 2 min, two times, ethanol 100% 1 min, ethanol 96% 1 min, and ethanol 70% 1 min.
3.7. Staining
1. After deparaffinization continue with the staining procedure of your choice. Most standard staining procedures can be used for FFPE sections. Most standard histological stains (e.g., hematoxylin/eosin (HE), Methyl Green, Cresyl Violet, and Nuclear Fast Red) are compatible with subsequent RNA isolation. For best RNA results from stained samples Cresyl Violet or HE staining is recommended. This short staining procedure colors the nuclei violet and the cytoplasm weak violet.
Nucleic Acids Extraction from Laser Microdissected FFPE Tissue Sections
123
It is particularly recommended for RNase-rich tissues since all solutions contain high ethanol. 2. Staining for subsequent RNA extraction: use only freshly prepared and precooled staining solutions. Use RNase-free water and solutions for all steps. All required reagents should be kept on ice. 3. Cresyl violet stain. After fixation (2 min 70% ethanol), dip slide for 30 s into 1% Cresyl violet acetate solution. Remove excess stain on absorbent surface. Dip into 70% of ethanol. Dip into 100% of ethanol. Air-dry briefly for 1–2 min (see Note 2). 4. HE stain. HE staining is used routinely in most histological laboratories and does not interfere with good RNA preparation, if intrinsic RNase activity is low. After fixation, quickly dip slide five to six times in RNase-free distilled water. Stain 1–2 min in Mayer’s Hematoxylin. Rinse 1 min in DEPC-treated tap water. Stain 10 s in Eosin Y. Perform a quick increasing ethanol series (70, 96, and 100%). Airdry shortly. 5. Storage. Stained slides can be used immediately or stored at −80°C before LCM (see Note 9). To avoid excess condensation of moisture during thawing, the slides should be frozen in a tightly sealed container (e.g., two slides back to back in a 50 ml Falcon tube). 3.8. Laser Microdissection: Capture in AdhesiveCaps: Dry Collection
1. An AdhesiveCap (Fig. 2) is filled with a kind of silicon that is completely dry. The intention of AdhesiveCap is to allow LCM without applying any capturing liquid into the caps prior to LCM. Beside the quick relocation of the lifted samples in the cap due to instant immobilization, there is no risk
Fig. 2. AdhesiveCap is a collection device filled with an adhesive material. It is especially adapted for a buffer-free sampling of microdissected specimens.
124
Burgemeister
of evaporation and crystal formation of a capturing buffer during extended specimen harvesting (see Note 10). 2. Other commercially available RNase-free plasticware can be used too (e.g., ABgene #AB-0350; 0.5 ml tubes). If there are no RNase-free tubes available, RNases from regular tubes have to be removed. 3.9. Laser Microdissection: Capture in Routine Caps: Wet Collection
When using unfilled regular microfuge tubes, it is recommended to fill a liquid into the cap to improve the adhesion of the captured cells. 1. Pipette 20 ml of lysis buffer into the cap. 2. Perform LCM. 3. The captured cells or cell areas will stick onto the wet inner surface of the cap and will not fall down after the lifting procedure (see Note 7). 4. When using glass-mounted samples (without membrane), it may be advisory to put more liquid (up to 40 ml) into the cap.
3.10. RNA Extraction: Remarks
1. In our hands, the QIAGEN RNeasy® FFPE Kit with some specific modifications is most useful. This procedure is very effective and allows a high final concentration of RNA due to a small elution volume. Genomic DNA contamination is minimized by a special DNA removal column (gDNA Eliminator spin column). 2. Since normally only stained tissue sections are used for microdissection, the deparaffinization and staining are done according to standard procedures for slides. Furthermore, the incubation with Proteinase K in our protocols is prolonged significantly compared with the QIAGEN protocol, because all our tests with laser microdissected material from various tissues showed higher RNA yields applying extended digestion times. 3. For formalin-fixed samples, a Proteinase K digestion step is essential. The time necessary for optimal Proteinase K digestion depends on many factors such as tissue type, fixation procedure, or element size of lifted material (see Note 11). The RNA solution may be stored at −20°C or used directly for reverse transcription. 4. Quality control by direct analyses such as the Agilent Bioanalyzer is very limited and only possible with large microdissected samples (some 4 mm2 from tissue sections of 5–10 mm thickness). We normally use 5–10 ml of the final RNA solution in an RT reaction of 20 ml using random oligomers (instead of oligoT) as primers for the cDNA synthesis (see Note 12).
Nucleic Acids Extraction from Laser Microdissected FFPE Tissue Sections
3.11. RNA Extraction by Using Components of the QIAGEN RNeasy ® FFPE Kit
125
1. Add 150 ml of Buffer PKD and 10 ml of Proteinase K to the tube, containing the LCM elements in the AdhesiveCap, and vortex in an “upside down” position. 2. Use an incubator to digest the samples in an “upside down” position at 55°C overnight (or for at least 3 h, see Note 11), then vortex and heat at 80°C for 15 min in a heating block. 3. Add 320 ml of Buffer RBC to adjust binding conditions. 4. Mix the lysate thoroughly and transfer it to a gDNA Eliminator spin column placed in a 2 ml collection tube. Centrifuge for 30 s at ³8,000 × g. Discard the column and save the flow through. 5. Add 720 ml of 100% ethanol to the flow through and mix well by pipetting. Do not centrifuge. Proceed immediately to the next step. 6. Transfer 700 ml of the sample to an RNeasy MinElute spin column placed in a 2 ml collection tube. Close the lid gently and centrifuge for 15 s at ³8,000 × g. Discard the flow through. Reuse the collection tube in step 7. 7. Repeat step 6 until the entire sample has passed through the RNeasy MinElute spin column. Reuse the collection tube in step 8. 8. Add 500 ml of Buffer RPE to the RNeasy MinElute spin column. (Buffer RPE is supplied as a concentrate. Ensure that ethanol is added to Buffer RPE before use.) Close the lid gently and centrifuge for 15 s at ³8,000 × g to wash the spin column membrane. Discard the flow through. Reuse the collection tube in step 9. 9. Add 500 ml of Buffer RPE to the RNeasy MinElute spin column. Close the lid gently and centrifuge for 15 s at ³8,000 × g to wash the spin column. After centrifugation carefully remove the spin column from the collection tube so that the column does not contact the flow through. 10. Place the RNeasy MinElute spin column in a new 2 ml collection tube, and discard the old collection tube with the flow through. Open the lid of the spin column and centrifuge at full speed for 5 min. Discard the collection tube with the flow through. It is important to dry the spin column membrane, since residual ethanol may interfere with downstream reactions. 11. Place the RNeasy MinElute spin column in a new 1.5 ml collection tube. Add 14–30 ml RNase-free water directly to the spin column membrane. Close the lid gently and centrifuge for 1 min at full speed to elute the RNA. The dead volume of the RNeasy MinElute spin column is 2 ml: elution with 14 ml of RNase-free water results in a 12 ml eluate. 12. The RNA solution may be stored at −20°C or used directly for reverse transcription.
126
Burgemeister
3.12. Other RNA Extraction Methods
1. Apart from the QIAGEN Kit, there are many other possibilities and kits to extract RNA from FFPE material. Depending on the material and the experience of the user, even simple procedures such as homemade AGTC methods or Trizol can be quite efficient. 2. If the original extraction protocol does not contain any Proteinase K digestion step, we recommend to apply the following simple procedure. 3. Add 20 ml of digestion buffer containing Proteinase K to the tube containing the LCM elements in the AdhesiveCap. 4. Use an incubator to digest the samples in an “upside down” position at 55°C overnight. 5. Spin down the lysate in a microcentrifuge (13,400 rcf). 6. Inactivate Proteinase K by heating to 90°C for 10 min. 7. Add the appropriate lysis buffer and mix by intense vortexing; if not proceeding immediately, store the digested samples at −20 or −80°C. 8. Continue with your preferred extraction procedure.
3.13. Quality Control of RNA
1. The most common method used for assessing the integrity of total RNA is to analyze the RNA sample on an agarose gel. In general, at least 200 ng of RNA must be loaded onto the gel. This usually is not practicable with low amounts as usually obtained during laser microdissection. 2. To analyze RNA samples with concentrations down to 50 pg/ml, the Agilent 2100 Bioanalyzer is an alternative to traditional gel-based analysis and provides information about RNA quality (degradation and purity) and quantity. 3. A prognosis of the expected amount of RNA in a tissue is difficult, since many factors such as species, cell/tissue type, fixation, staining, fragmentation, extraction procedure, and others will influence the outcome.
3.14. DNA Extraction
1. Proteinase K digestion step is essential for formalin-fixed samples. The time necessary for optimal Proteinase K digestion depends on many factors such as tissue type, fixation procedure, or element size of lifted material. An overnight digestion (12–18 h) is a good starting point for optimization but shorter digestion times may be tested as well. To our experience at least 3 h digestion should be applied with any extraction procedure and material. 2. Add 15 ml ATL of the QIAamp Micro Kit for the isolation of genomic DNA to the microdissected sample in the AdhesiveCap.
Nucleic Acids Extraction from Laser Microdissected FFPE Tissue Sections
127
3. Add 10 ml Proteinase K and mix by pulse-vortexing for 15 s. 4. Place the 0.2 ml tube in an “upside down” position at 56°C in an incubator for 2–18 h by occasional agitation (see Note 13). 5. Add 25 ml Buffer ATL and 50 ml Buffer AL, close the lid and mix by pulse-vortexing for 15 s. To ensure efficient lysis, it is essential that the sample and Buffer AL are thoroughly mixed to a homogeneous solution. 6. Add 50 ml of ethanol (96–100%), close the lid, and mix thoroughly by pulse-vortexing for 15 s. Incubate for 5 min at room temperature (15–25°C). If room temperature exceeds 25°C, cool the ethanol on ice before adding to the tube. 7. Briefly centrifuge the 0.2 ml tube to remove drops from the lid. 8. Carefully transfer the entire lysate to the column without wetting the rim, close the lid, and centrifuge at 6,000 × g for 1 min. Place the column in a clean 2 ml collection tube, and discard the collection tube containing the flow through. If the lysate has not completely passed through the column after centrifugation, centrifuge again at a higher speed until the column is empty. 9. Carefully open the column and add 500 ml of Buffer AW1 without wetting the rim. Close the lid and centrifuge at 6,000 × g for 1 min. Place the column in a clean 2 ml collection tube, and discard the collection tube containing flow through. 10. Repeat step 9 (see Note 14). 11. Centrifuge at full speed (20,000 × g) for 3 min to dry the membrane completely. This step is necessary, since ethanol carryover into the eluate may interfere with some downstream applications. 12. Place the column in a clean 1.5 ml microcentrifuge tube and discard the collection tube containing the flow through. Carefully open the lid of the column and apply 20 ml of distilled water to the center of the membrane. Ensure that distilled water is equilibrated to room temperature (15– 25°C). Dispense distilled water onto the center of the membrane to ensure complete elution of bound DNA (see Note 15). 13. Close the lid and incubate at room temperature (15–25°C) for 1 min. Centrifuge at full speed (20,000 × g) for 1 min.
128
Burgemeister
4. Notes 1. Sometimes lot-to-lot variations in the purchased Cresyl violet powder can lead to weaker staining results if the dye content is below 75%. 2. In most cases, the Cresyl violet staining procedure will be sufficient for cell identification. If an enhancement of the staining intensity is desired, two additional steps of 50% (v/v) ethanol are possible: one step before staining in Cresyl violet, one step after staining in Cresyl violet. Additional intensification can be obtained by increasing the working temperature of all solutions to room temperature. 3. We recommend AdhesiveCap as a collection device for all RNA experiments. After LCM add lysis buffer of your own choice and incubate “upside down” for 30 min. Subsequently centrifuge the lysate and then apply the further steps of the experiments. 4. DEPC is toxic and should be used under a hood. 5. Always prepare a fresh mixture of capturing buffer and Proteinase K. 6. The detergent Igepal CA-630 in the capturing buffer allows to smear out a small amount of liquid in the whole cap area. 7. Please keep in mind that all kinds of aqueous solutions will run dry during extended working time. 8. It is very important not to use any force to push off the coverslip because this might damage the section. Wait until it falls off by itself. The necessary time depends on the age of the sample and the dryness of the mounting medium and may range from hours to days. Fresh slides (only days old) can be de-coverslipped much faster. From the dry glass slide, sample material can be captured directly by the “AutoLPC” function of PALM RoboSoftware. 9. For rethawing, the container should not be opened before it is completely warmed up again to ambient temperature. 10. We recommend AdhesiveCap as a collection device for most experiments. After LCM add lysis buffer of your own choice and incubate “upside down” for 30 min. Subsequently centrifuge the lysate and then apply the further steps of the experiments. 11. The time necessary for optimal Proteinase K digestion depends on many factors such as tissue type, fixation procedure, or element size of lifted material. An overnight digestion (12–18 h) is a good starting point for optimization but shorter digestion times may be tested as well. To our experience at least 3 h
Nucleic Acids Extraction from Laser Microdissected FFPE Tissue Sections
129
digestion should be applied with any extraction procedure and material. 12. The use of random or gene-specific primers is important. Reverse transcription of formalin-fixed RNA with standard oligoT primers is inefficient and strongly 3¢-biased due to the numerous strand breaks and modifications inflicted by the formalin fixation and paraffin-embedding procedure. 13. The time necessary for complete Proteinase K digestion depends on the kind and the amount of collected material. 14. Contact between the column and the flow through should be avoided. Some centrifuge rotors may vibrate upon deceleration, resulting in the flow through, which contains ethanol coming into contact with the column. Take care when removing the column and collection tube from the rotor, so that flow through does not come into contact with the column. 15. QIAamp MinElute Columns provide flexibility in the choice of elution volume. Choose a volume according to the requirements of the downstream application. Remember that the volume of eluate will be up to 5 ml less than the volume of elution solution applied to the column. References 1. Lehmann, U., and Kreipe, H. (2009) Tissue procurement for molecular studies using laserassisted microdissection. Methods Mol Biol. 506, 299–310. 2. Von Eggeling, F., and Ernst, G. (2007) Microdissected tissue: an underestimated source for biomarker discovery? Biomark Med. 1, 217–219. 3. Alvarez, H., Corvalan, A., Roa, J.C., Argani P., Murillo F., Edwards J., et al. (2008) Serial analysis of gene expression identifies connective tissue growth factor expression as a prognostic biomarker in gallbladder cancer. Clin Cancer Res. 14, 2631–2638. 4. Melle, C., Ernst, G., Schimmel, B., Bleul, A., Koscielny, S., Wiesner, A., et al. (2004) A technical trade for proteomic identification and characterization of cancer biomarkers. Cancer Res. 64, 4099–4104. 5. Zhang, Y., Ye, Y., Shen, D., Jiang, K., Zhang, H., Sun, W., et al. (2010) Identification of transgelin-2 as a biomarker of colorectal cancer
by laser capture microdissection and quantitative proteome analysis. Cancer Sci. 101, 523–529. 6. Mori, R., Wang, Q., Danenberg, K.D., Pinski, J.K., and Danenberg, P.V. (2008) Both b-actin and GAPDH are useful reference genes for normalization of quantitative RT-PCR in human FFPE tissue samples of prostate cancer. Prostate. 68, 1555–1560. 7. Theophile, K., Jonigk, D., Kreipe, H., and Bock, O. (2008) Amplification of mRNA from laser microdissected single or clustered cells in formalin-fixed and paraffin-embedded tissues for application in quantitative real-time PCR. Diagn Mol Pathol. 17, 101–106. 8. Schlomm, T., Luebke, A.M., Sultmann, H., Hellwinkel, O.J., Sauer, U., Poustka, A., et al. (2005) Extraction and processing of high quality RNA from impalpable and macroscopically invisible prostate cancer for microarray gene expression analysis. Int J Oncol. 27, 713–720.
wwwwwwwwwwwwwww
Chapter 9 Microarray-Based CGH and Copy Number Analysis of FFPE Samples Fahd Al-Mulla Abstract Over the past decade, utilization of microarray technology has flourished in biomedical research. It has evolved rapidly into a revolutionary tool that offers deeper insight into the molecular basis associated with complex diseases, especially in the field of cancer. Specifically, array-based Comparative Genomic Hybridization (aCGH) permits the detection of genome-wide copy number alterations with high resolution. Microarray application to DNA extracted from formalin-fixed paraffin-embedded tissue (FFPE), in particular, poses a challenge due to the partially degraded nature and compromised quality of the DNA. This chapter gives a description of the several CGH-microarray platforms currently available and offers practical steps that guide you through optimal handling and superior aCGH data acquisition of DNA extracted from FFPE tissues. Key words: Microarray, Array-based comparative genomic hybridization, Formalin-fixed paraffin embedded, Copy number, Microarray platforms, Copy number variation
1. Introduction Microarray-based comparative genomic hybridization or aCGH is an advanced cytogenetic tool that has been widely used for over a decade (1). The application of microarray to clinical samples, and especially to formalin-fixed paraffin-embedded (FFPE) tissues has recently been shown to dramatically change the standard of patients’ care (2). Array CGH technology enables comprehensive genome mapping and detects changes in copy number of target DNA sequences (3). These targets could be cloned DNA segments, e.g., BAC clones, cDNA, oligonucleotides, or PCR generated sequences. Typically they are immobilized (spotted) on a solid surface (4). Differentially labeled test and reference samples are hybridized to their complementary target sequences Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_9, © Springer Science+Business Media, LLC 2011
131
132
Al-Mulla
spotted on the array. Ratio of fluorescent measurements of both label dyes at each hybridized arrayed spot is calculated to estimate relative copy number variation at specified loci. The resolution of the array CGH is determined by the map distance between target DNA markers and by the length of the clones used. Microarray technology has opened the gates to a new era in research and translational medicine through allowing the simultaneous high resolution analysis of thousands of different genes (5). The obtained microarray data of genetic copy number changes associated with diseases provides rich diagnostic and prognostic valuable information in terms of: tumor classification, tumor biological staging, risk assessment of premalignant lesions, prediction of responses to chemotherapeutic or hormonal agents, and detection of microorganisms (6–8). Furthermore, microarray data obtained from tumors at different stages of progression have paved the way to a more comprehensive understanding of tumor development and have illuminated new opportunities for theranostics. Different competing microarray platforms have emerged (5). Each introduced platform is designed to work with a dual- or single-color detection system. Affymetrix® GeneChips, in particular, gained large acceptance in obtaining transcription profiles. GeneChip technology offers many advantages like the availability of ready for use chips, large genome coverage and high reproducibility. The chip consists of 25-mers short oligonucleotides built in orderly fashion and synthesized either by chemical or by lightdirected synthesis. High-density arrays are formed using the lightdirected synthesis of oligonucleotides using a combination of photolithography and solid-phase DNA synthesis. The fact that GeneChips are designed in silico eliminates the need for managing clone libraries. In addition, probe redundancy feature improves signal-to-noise ratio, minimizes cross hybridization effect, and enhances the range of detection (5). Agilent platforms are also popular and supplied into two forms: a whole genome array and high definition targeted oligo aCGH microarrays. Oligonucleotide probes are synthesized in situ by inkjet printing using phosphoramidite chemistry. Agilent Oligonucleotide platforms consist of 60 mers. Only one 60 mer per gene or transcript is required in Agilent platforms. A newer 1 M chip has been recently released by Agilent. However, its application onto FFPE tissue requires more intense standardization (9). CodeLink™ Bioarray platform from Amersham Biosciences, which consists of 30 mer Oligonucleotides immobilized through covalent bonds to active functional groups on three dimensional polyacramide gel matrix slide surface. Oligonucleotides are synthesized using standard Phosphoramidite chemistry. Like Agilent platforms, CodeLink Bioarray platforms have a single oligonucleotide for each interrogated gene (5).
Microarray-Based CGH and Copy Number Analysis of FFPE Samples
133
NimbleGen systems offer two types of platforms: whole genome and custom-targeted arrays. NimbleGen arrays are of ultrahigh density with long Oligonucleotide probes. The arrays design could hold up to 1.2 million Oligonucleotide probes offering up to ~5–10 kb resolution. NimbleGen offers multiplex platforms that enable simultaneous multiple sample pairs to hybridize on a single slide (5). Choosing the most appropriate microarray platform in practice depends on the nature of the intended study and length of targeted genome sequences for analysis. In addition, factors such as cost effectiveness, dedicated hardware and software requirements for the chosen microarray system are considered when planning a microarray experiment (5). This chapter describes the analysis of FFPE-extracted DNA using Agilent oligonucleotide array-based CGH platform (244 K) and BAC-based arrays (21 K) with practical guide steps toward optimal handling of FFPE-extracted DNA.
2. Materials 2.1. DNA Extraction from FFPE Tissue
1. Gentra Puregene Tissue Kit 4 g (Qiagen). 2. Nuclease-free water (Ambion). 3. TBE buffer (Accugene). 4. Gel loading solution (Sigma). 5. 100 bp Ladder (NEB). 6. Proteinase K 20 mg/ml (Invitrogen). 7. Glycogen (Qiagen). 8. O-ring tube. 9. Isopropanol. 10. Absolute Ethanol. 11. Agarose. 12. Xylene. 13. P10, P20, P200, and P1000 pipettes. 14. Sterile, nuclease-free pipette tips. 15. Powder-free Gloves.
2.2. DNA Labeling and Hybridization with Agilent Oligonucleotide Array-Based CGH
1. Human Genome CGH Microarray Kit 244 A (Aglient). 2. Hybridization Chamber, stainless steel (Agilent). 3. Hybridization Chamber gasket slides (Agilent). 4. Oligo aCGH Labeling Kit for FFPE Samples contains 10× CGH block (Agilent).
134
Al-Mulla
5. Agilent Oligo aCGH Hybridization Kit (Agilent). 6. Agilent Oligo aCGH Wash Buffer 1 and 2 set (Agilent). 7. Stabilization and Drying Solution, 500 mL (Agilent). 8. Hybridization oven; temperature set at 65°C (Agilent). 9. Human Cot-1 DNA (Invitrogen). 10. Acetonitrile (Sigma). 11. Ethanol 95–100% molecular biology Grade (Sigma). 12. Elmasonic water bath sonicator (ELMA). 13. Slide staining dish (Wheaton). 14. Nuclease-free water. 15. P10, P20, P200, and P1000 pipettes. 16. Sterile, nuclease-free aerosol barrier pipette tips. 17. Nuclease-free 0.2 ml PCR tubes, thin-walled (sterile). 18. Nuclease-free 1.5 ml microfuge tubes (sterile). 19. Magnetic stir bar and stirrer plate. 20. PCR machine with heated lid. 21. Agilent scanner. 22. Micro centrifuge. 23. UV Transilluminator. 24. Thermomixer. 25. Speed-vac. 26. Powder-free gloves. 27. Vortex mixer. 28. Agarose. 2.3. Wash Buffers Composition
1. Wash Buffer 1 (0.1× SSC, 0.1% SDS) (a) 20× SSC (5 ml) (b) 10% SDS (10 ml) (c) Distilled water (985 ml) Final volume 1,000 ml 2. Wash Buffer 2 (0.1× SSC) (a) 20× SSC (5 ml) (b) Distilled water (995 ml)
2.4. DNA Labeling and Hybridization Using the 21 K BAC Array CGH System
1. Ultrahigh Resolution Pangenomic Tiling 21 K BAC-CGH Array (Array Genomics). 2. ULS arrayCGH Labelling Kit for BAC array contains 10× CGH block (Kreatech, EA-005).
Microarray-Based CGH and Copy Number Analysis of FFPE Samples
135
3. Machery Nagel Kit, NucleoSpin Extract II (Catalog # 740 609.50). 4. Human or Mouse Cot –1 (Invitrogen, Cat #’s 15279-011 and 18440-016, respectively). 5. Yeast tRNA (Invitrogen, Cat #15401-011). 6. 20× SSC (Invitrogen, Catalog # 15557-044). 7. 10% SDS (Invitrogen). 8. Slide staining dish (Wheaton). 9. 25 × 60 inch. Coverslip (Erie Scientific). 10. Hybridization chamber (Corning). 11. Elmasonic water bath sonicator (ELMA). 12. Nuclease-free water (Ambion). 13. Micro-centrifuge. 14. Refrigerated Micro-centrifuge. 15. 96–100% Ethanol. 16. Sterile, nuclease-free 0.2 ml PCR tubes, thin-walled. 17. PCR machine with heated lid. 18. Thermocycler. 19. Thermomixer.
3. Methods 3.1. DNA Extraction from FFPE Tissue
The following technique describes DNA extraction from 50 mm tissue sections or approximately 5–10 mg of tissue. Depending on the amount of tissue available for DNA extraction, the reagents and solutions used can be scaled up or down (10) and Chapters 7, 8, and 11).
3.1.1. Deparaffinization of FFPE Sections
Five finely cut FFPE sections of 10-mm thick each (total 5–10 mg) are transferred into an O-ring tube to be deparaffinized. The cut sections are then incubated in 300 ml of xylene at room temperature for 15 min, centrifuged for 5 min at 13,000–16,000× g. The supernatant is discarded leaving the precipitant tissue at the bottom of the tube. This step is repeated three times (see Note 1). The tissue pellet is then washed twice with 300 ml of absolute ethanol, mixed thoroughly and centrifuged for 5 min. The supernatant is discarded each time leaving the precipitated pellet without disruption. A final wash with 300 ml of 70% Ethanol is completed accompanied with thorough mixing and centrifugation at 13,000–16,000× g for 5 min. Supernatant is discarded without disrupting the precipitant tissue. The pellet is left to dry in speed-vac for 3–5 min.
136
Al-Mulla
3.1.2. Cell Lysis
Cell lysis solution (300 ml) is being added to the previously dried tissue pellet. Closed tubes are inverted 50 times to ensure proper mixing and homogenization of the lysis solution with the tissue. A volume of 15 ml of proteinase K (20 mg/ml) is added to the pellet in tubes and mixed well. The tightly covered and secured tubes are incubated at 55–58°C with agitation for 24–48 h for proper tissue digestion. Subsequent addition of extra 10 ml of proteinase K with longer incubation period might be needed in the case of residues of undigested tissue. Proceeding to the next step is recommended only when the tissue is fully digested.
3.1.3. DNA Extraction
Tubes containing the digested tissue are centrifuged for 30 s then incubated with 3 ml of RNase A solution (17,500 U) for 30–40 min at 37°C. Tubes are then cooled at room temperature for 15–20 min. Protein precipitation solution (200 ml) is added to the tubes, mixed thoroughly for 20 s and centrifuged for 5 min at 13,000–16,000× g. The precipitated proteins will form a tight pellet that is easily visible at the bottom of the tube. If the pellet is not clearly visibly, leave the tubes on ice for 5 min. In new sterile O-ring tubes, 600 ml of isopropanol and 1 ml of glycogen are added. Supernatant from precipitation step is carefully added to the newly prepared tubes (see Note 2). Tubes are inverted 50 times, and then centrifuged for 5 min. Discard the supernatant carefully making sure the pellet does not dislodge and the excess is blotted out by inverting the tube on an absorbent paper. The DNA pellet is washed with 300 ml of 70% ethanol and the tubes are inverted several times to ensure proper washing. The tubes are centrifuged for 1 min followed by careful clearance of the supernatant without dislodging the DNA pellet. Excess supernatant is blotted out on an absorbent paper and the DNA pellet is left to dry for a few minutes at 37°C. Care should be taken not to over dry the pellet. The dried DNA pellet is rehydrated in 25–70 ml of nuclease-free Water or DNA Hydration solution. The tubes can be kept at room temperature for 1–2 h to allow the pellet to completely dissolve. The concentrations of extracted DNA are estimated by absorbance measurement at 260, 280, and 230 nm wavelengths using Nanodrop® ND-1000 spectrophotometer (Nanodrop® Technologies Inc., Wilmington, USA) (see Note 3). To further assess the extracted DNA is to run product on 1.5% Agarose gel electrophoresis to check the integrity of the DNA (see Note 4). A 50 bp ladder is loaded as a reference for DNA size
3.2. Agilent CGH Methodology
Using the Agilent aCGH labeling Kit, DNA samples can be differentially labeled with fluorescent dyes (Cy3 and Cy5). The reference or control DNA is usually labeled with Cy3 and the experimental or test DNA with Cy5. Equal amounts of test and control DNA are used in this protocol (see Notes 5 and 6).
3.2.1. DNA Preparation, Fragmentation, and Labeling
Microarray-Based CGH and Copy Number Analysis of FFPE Samples
137
A quantity of 300–400 ng of the test DNA is run on a 1.5% agarose gel electorphoresis to check the integrity and level degradation of each sample (see Note 7). In separate 0.2 ml sterile PCR tubes, 5 mg of the genomic DNA, test and control are added and completed to a final volume of 20 ml with nuclease-free water (see Note 8). The DNA samples are then fragmented by sonication for 20–30 s in water bath sonicator (10). The fragmented DNA (500–750 ng) is then run on 1.5% Agarose gel electrophoresis. Ideally the fragmented DNA appears in the form of a smear ranging in size between 500 and 2,000 bp with the majority of the fragments migrating between 500 and 1,000 bp. In new separate 0.2 ml sterile PCR tubes, 2 ml of fragmented DNA from each control or test DNA sample is added and the volume is completed till 16 ml with nuclease-free water. A total of 2 ml of labeling reagent (ULS-Cy3 for the control DNA, ULS-Cy5 for the test DNA) is added to each sample and mixed well with gentle pipetting, then the tubes are spun for 10 s at 12,000× g in microcentrifuge. The tubes are transferred to a PCR machine with a heated lid and incubated for 30 min at 85°C (see Note 9). When incubation is over, the tubes are spun in the microcentrifuge for 10 s. 3.2.2. Free-Dye Removal Using KREA-pure Columns
This step is essential to remove free ULS-Cy3 and ULS-Cy5, as it will interfere with the subsequent steps and increase background noise. Here, the Agilent KREApure columns are used according to manufacturer instructions to remove free ULS-Cy3 and ULS-Cy5 in the samples. The columns are spun at maximum speed for 1 min in a micro centrifuge. The flow through is discarded and the columns are placed again in the collection tube. A wash of 300 ml of nuclease-free water is added into columns and tubes with the columns are spun for 1 min at maximum speed. The collection tubes are discarded and the columns are placed in new sterile 1.5 ml micro centrifuge tube. The labeled samples are pipetted into the columns assembled with the new tubes (see Note 10). After centrifugation for 1 min, the purified DNA is collected in the tubes. It is necessary at this point to measure the degree of labeling (DOL) of the samples using Nanodrop® spectrophotometer. The following formula is used to calculate the DOL. Degree of labeling =
340 * pmol/m l of dye ´ 100%. ng/m l of GenomicDNA*1, 000
The average optimal DOL for ULS-Cy5-Labeled samples is 0.75–2.5%, while for ULS-Cy3 is 1.75–3.5%.
138
Al-Mulla
Before proceeding to hybridization, Combine the ULS-Cy3labeled control DNA to its paired ULS-Cy5-labeled test DNA for a total volume of around 40 ml (see Note 11). 3.2.3. Hybridization 3.2.3.1. Preparation of Hybridization Cocktail
A 10× CGH Blocking Agent is prepared according to manufacturer’s instructions and kept at room temperature for 60 min (see Note 12). In advance, set heat blocks (thermal cycler) to 95 and 37°C, respectively. Then, the volume of combined ULS-Cy3–Cy5-labeled DNA is concentrated down to 28 ml using a speed-vac. A volume of 50 ml of Cot I DNA (1 mg/ml) is added to the labeled DNA, followed by the addition of 52 ml of 10× CGH Blocking Agent and finally the addition of 260 ml of 2× CGH Hybridization Buffer to the above Hybridization cocktail. The Cocktail is mixed a few times by inverting the tube carefully, avoiding the formation of bubble and centrifuged briefly in a micro centrifuge. The tubes are incubated at 95°C for 3 min in a heat block, then immediately transferred to the heat block set at 37°C for 30 min after which the tubes are removed and spun for 10 s in microcentrifuge. This is followed by the addition of 130 ml of Agilent CGH Block to the hybridization cocktail, mixing and avoiding bubbles in the mix. Finally, the tubes are centrifuged for 10 s to collect the cocktail at the bottom.
3.2.3.2. Microarray Hybridization
An Agilent microarray slide has two sides. The side on which the microarray is printed is referred to as the active side and is usually distinguishable by the “Agilent” labeled barcode. The side on which there is no microarray is called the inactive side and is usually identified by the presence of a numeric barcode.
3.2.3.3. Hybridization Assembly
A new gasket slide is loaded into the Agilent SureHyb chamber base, with the gasket well and Barcode facing up, making sure that the gasket is aligned properly with the base and not ajar. Carefully dispense 490 ml of the sample hybridization cocktail onto the gasket slide in a “drag and dispense” manner. Then, carefully place the microarray slide onto the gasket slide making sure that the active side of the slide is facing down and the numeric barcode is facing up, forming a gasket-array sandwich. Ensure that the sandwich is aligned properly. The SureHyb chamber cover is placed on the sandwich and the clamp is moved over and hand-tightened but not too firmly to avoid damaging the slides. The assembly is rotated making sure to wet the slides completely. This is also done to make sure that all the bubbles, if present, are moving freely. Any stationary bubbles can be removed by gently tapping the gasket side of the assembly with forceps. Assembled chambers are placed in the rotator racks within the hybridization oven at 65°C for 40 h at 20 rpm (see Note 13).
Microarray-Based CGH and Copy Number Analysis of FFPE Samples 3.2.3.4. Washing the Microarray Slides
139
Cy5 has been shown to be adversely affected by the presence of ozone when it is present in levels greater than 5 ppb in the atmosphere. To avoid ozone-induced degradation in high-ozone environments, the Stabilization and Drying solution is used. The ozone scavenging compound in the Stabilization and Drying solution is usually present in saturating amounts and will sometimes precipitate. It is very important to warm the solution if it shows visible precipitation. A volume of 500 ml of the buffer II is added to a bottle dedicated to the use of this buffer only and leave it overnight in an incubator at 37–38°C. This is because Oligo aCGH Wash buffer 2 needs to be at 37°C for its optimal performance. Then, the stabilization solution is checked for any visible precipitation (see Note 14). If present, the bottle is placed in an incubator set at 37–40°C with gentle shaking overnight. The stabilization solution is taken out of the incubator and left to equilibrate to room temperature on the day of the wash. A 500 ml of the Oligo aCGH Wash Buffer 1 is added at room temperature into a small slide staining dish. The slide rack is immersed into the Oligo aCGH Wash Buffer 1 at room temperature into slide staining dish no. 2. The dish is placed on a magnetic stir plate to stir the buffer solution using a magnetic stir bar. A 500 ml of the prewarmed Oligo aCGH Wash Buffer 2 is added at 37°C to slide staining dish no. 3 and stirred in the same way. A volume of 500 ml of Acetonitrile is poured at room temperature into slide staining dish no. 4 and stirred using the magnetic stir bar (see Note 15). For staining dish no. 5, add 500 ml of the Stabilization and Drying solution, which has been allowed to equilibrate to room temperature, and stir using a magnetic stir bar while the dish is placed on a magnetic stir plate. At this point, the Hybridization chamber is removed from the oven to be disassembled. The hybridization chamber is then placed on a flat surface to loosen the thumbscrew. Slide off the clamp and remove the chamber cover. Now the array–gasket sandwich is removed by holding the edges making sure the array is at the top. The sandwich is immersed completely in the small slide staining dish containing Oligo aCGH Wash Buffer 1 without letting go of the sandwich. While keeping the slides completely submerged, the forceps are inserted gently between the slides to separate them by gently twisting the forceps. The gasket is left to drop to the bottom without letting go of the microarray slides. The slides are transferred quickly to slide rack that is submerged in slide staining dish no. 2 containing Oligo aCGH Wash Buffer 1 at room temperature (see Note 16). After all the slides are placed on the rack in slide staining dish no. 2, stir for 5 min. The slide rack is then transferred quickly into slide staining dish no. 3 containing Oligo aCGH Buffer II at 37°C for 1 min with stirring set at 4 min.
140
Al-Mulla
Fig. 1. Agilent 244K CGH microarray slide showing at the top the full view of the slide and at the bottom a zoomed view of the representative square.
The slide rack is transferred to staining dish number 4 containing acetonitrile at room temperature for 1 min with gentle stirring. Then the slide rack is transferred into Stabilization and drying solution at room temperature for 30 s with gentle stirring. The rack is removed slowly to minimize the number of droplets on the slide (see Note 17). The slides are best scanned immediately to minimize environmental effects on the signals (see Note 18 and 19). 3.2.4. Microarray Scanning and Feature Extraction of Data
It is recommended to scan the Agilent 244K slides at 5 um using the Agilent Scanner Control v7.0 (see Note 20). Other scanners that support this array platform are also available. Scanning is done according to the scanner manufacturer’s instructions. Feature Extraction (FE) Software is used to extract data from the TIFF (.tif) images (Fig. 1) generated after scanning the Agilent aCGH microarrays with the Agilent Scanner (see Note 21).
3.3. 21 K BAC Array CGH Methodology on FFPE Samples
To check the integrity and level of degradation of the test DNA, 300–400 ng DNA amount is run on a 1.5% Agarose gel electrophoresis. In a sterile 0.2 ml PCR tube, 5 mg of genomic DNA is diluted in a total volume of 20 ml. The water bath sonicator is used to sonicate and randomly fragment the DNA sample for 20–30 s (see Note 22). The tubes
3.3.1. DNA Sample Preparation and Fragmentation
Microarray-Based CGH and Copy Number Analysis of FFPE Samples
141
are transferred to ice for 2 min and centrifuged briefly to collect the sample at bottom of the tube. A quantity of 500–750 ng of the fragmented DNA is run on Agarose gel electrophoresis to make sure that there is no native DNA left in the wells. The fragmented DNA appears as a smear, with the ideal size range between 500 and 2,000 bp. 3.3.2. DNA Labeling
In two sterile separate 0.2 PCR tubes, 2 mg of fragmented control and test DNA is diluted in a final volume of 17 ml of nuclease-free water. Then, 2 ml of the 10× Labeling solution is added to each tube, followed by the addition of 1 ml of ULS-Cy3 to the control DNA and 1 ml of ULS-Cy5 to the test DNA (see Note 23). The Tubes are mixed gently and spun for 10 s in a micro centrifuge. The tubes are transferred to a PCR machine with a heated lid and incubate for 30 min at 85°C (see Note 24). The tubes are then removed and centrifuged for 10 s (see Note 25).
3.3.3. Free- Dye Removal Using KREA-pure Columns
This step is essential to remove free ULS-Cy3 and ULS-Cy5 as it will interfere with the subsequent steps and increase background noise (see Note 23). Using the labeling quantification measurements obtained by Nanodrop® spectrophotometer, the degree of labeling (DOL) is calculated using the following equation:
Degree of labeling =
340 * pmol / m l of dye ´ 100%. ng / m l of genomic DNA*1, 000
Optimally the DOL for Labeled samples should be between 1.0 and 2.5%. The ULS-Cy3-labeled control DNA with the ULS-Cy5labeled test DNA for a total volume of around 40 ml. 3.3.4. DNA Precipitation
A volume of 50 ml of DNA is added to each tube containing the combined Cy3-Cy5-labeled DNA, followed by the addition of 90 ml (or 1 volume) of 0.3 M sodium acetate (pH 5–8). Then, 225 ml of ice cold ethanol is added to each tube and the tubes are mixed well and incubated at −80°C for 30 min (see Note 23). The tubes are centrifuged for 15 min in a refrigerated micro centrifuge at 4°C for 15 min. The supernatant is discarded while taking care not to disturb the DNA pellet. Dry for 10 min at room temperature in the dark.
3.3.5. DNA Hybridization
To each tube containing combined samples, 6.9 ml of the KREAblock buffer, 8.1 ml of 10% SDS, and 12 ml of Yeast tRNA (50 mg/ml) are added. The tubes are mixed gently by pipetting and avoid bubble formation and incubate at room temperature for 10 min. A volume of 43 ml of the KREA-Hyb CGH solution is added to each tube to make a final volume of 70 ml and mixed.
142
Al-Mulla
The samples are transferred to a heat block at 70°C and incubated for 15 min. Then the tubes are immediately transferred to a heat block at 37°C and incubated for 30min. At this stage, the 21 K BAC Microarray slide is prepared for hybridization. The slide is removed from storage in a dessicator and is put in a UV crosslinker at 350 mJ to crosslink slides. The crosslinker is set to spin down the tubes quickly in a micro centrifuge. The microarray slide is put on a flat surface. The probe is added to the active side (bar-coded side) of the microarray in a drag and dispense manner. Carefully, a clean 24 × 60 in. coverslip is placed on top of the slide and the hybridization chamber is placed on a flat surface. A 10–15 ml of sterile nuclease-free water is added to humidity ports in the chamber. The Microarray coverslip sandwich is placed carefully into the chamber. Make sure the slide is not ajar. The chamber is assembled and transferred to an oven at 42°C for 16 h. 3.3.6. Posthybridization Washing
A volume of 300 ml of the Wash Buffer 1 is added at 50°C into three slide staining dishes (1, 2 and 3). A volume of 300 ml of Wash Buffer 2 at 50°C to another two slide staining dishes (4 and 5) and 300 ml of the wash Buffer 1 into slide staining dish 6. Finally, 300 ml of 96–100% ethanol is poured into slide staining dish 7. The hybridization chamber is removed from the oven, placed on a flat surface, and then disassembled. The sandwich pair is removed and quickly immersed into slide staining dish 1. Remove the coverslip by agitating the slide gently. The slide is transferred to the slide rack in slide staining dish 2, quickly without letting the microarray slide to dry out and the slides are washed for 45 s with agitation followed by incubation for 45 s without agitation. The rack is transferred into slide staining dish 3 at 50°C and same washing process is repeated. The rack is then to be transferred to slide staining dish 4 containing Wash Buffer 2 at 50°C and washing is done in the same manner as before. The rack is then moved to slide staining dish 5 containing at 50°C to repeat washing step same as before. Quickly the rack is plunged and removed into slide staining dish 6 at room temperature once. The rack is plunged into slide staining dish 7 containing ethanol once. The rack is removed carefully to minimize droplets being formed on the Microarray slides. The slides are finally transferred to a slide spinner and spin for a few seconds. The slides are to be scanned immediately.
3.3.7. Microarray Scanning Using Agilent Scanner
It is recommended to scan the 21 K BAC slides at 5 mm using the Agilent Scanner Control v7.0 (see Note 20). Other scanners that support this array can also be used. For microarray data analysis, readers are advised to explore various softwares available in the market (11). Typically, using our protocol, we have been generating very high-resolution profiles of genomic DNA extracted from FFPE cancer tissues (Fig. 2).
Microarray-Based CGH and Copy Number Analysis of FFPE Samples
p15.1 p14 p13
Genes Exons CNVs miRNA
q12 q13.1
q13.3
q24
q25
q26
q28.3
143
q32.1
hsa-mir-95 hsa-mir-218-1
2.50 2.25 2.00 1.75 1.50 1.25 1.00 .75 .50 .25 0 −.25 −.50 −.75 −1.00 −1.25 −1.50 −1.75 −2.00 −2.25 −2.50
0
20 Mb
40 Mb
60 Mb
80 Mb
100 Mb
120 Mb
140 Mb
160 Mb
180 Mb
Fig. 2. Log2 ratio along chromosome 4 from a more than 5-year old case of FFPE colorectal cancer showing at the chromosome ideogram at the top with copy number gains above the chromosome and copy number losses below the chromosome. The lower panel depicts probes aligned along chromosome 4 with clear deviation from 0-line (this line indicates normal copy number ratio). Probes with ratios above 1.8 are considered amplified (lines above chromosome ideogram) and below −1.8 (lines below the chromosome ideograms) are considered as deletions (p = 1 × 10−6).
4. Notes 1. Xylene is a hazardous chemical and care must be taken to wear gloves, safety goggles, and a laboratory coat. Work in a fume hood. 2. If the DNA yield is expected to be low, then the addition of 0.5 ml of glycogen will improve the DNA yield. 3. The A260/A280 ratio, which indicates the absence of protein contamination, should ideally lie between 1.8 and 2.0. The A260/A230 ratio should be greater than 2.0 and indicates the absence of organic compounds such as phenol, alcohol, and other carbohydrates. 4. DNA from FFPE samples are often degraded and will be visualized as a smear.
144
Al-Mulla
5. DNA should preferably be suspended in nuclease-free water or a buffer (e.g., Low-concentration TE buffer) that will not interfere with downstream labeling procedures. 6. Use Powder-free gloves throughout the protocol. 7. The minimum amount of DNA required for this protocol is 2.5–3.0 mg. Control and test DNA should be suspended, in separate tubes, in a total volume of 16 l for fragmentation. Use 500 ng to run on a gel to confirm fragmentation. 8. Genomic DNA obtained from FFPE samples are often degraded and does not require fragmentation. The need for fragmentation is determined by running 500–750 ng of DNA on 1.5% gel. 9. ULS-Cy3 has better labeling efficiency than ULS-Cy5. Therefore, samples being labeled with ULS-Cy5 are allowed to incubate for 100–120 min at 85°C. The labeled DNA samples are stored on ice until purification. 10. If the total volume of the labeled DNA is less than 20 ml, then make it up with nuclease-free water. 11. The labeled DNA can be stored at −20°C in the dark overnight. 12. After reconstitution, the 10× CGH blocking agent should be stored at −20°C. 13. Make sure to balance the assembly chamber within the rotator rack. 14. The stabilization solution can be reused for washing up to 20 slides. 15. Acetonitrile is highly flammable and extremely toxic. It is advisable to warm the Stabilization solution in the original bottle. Do not filter the Stabilization solution. 16. It is recommended to wash up to five slides at a time to facilitate uniform washing. 17. Discard the used Oligo aCGH buffer I and II, if there are additional slides that need to be washed. 18. Minimize exposure to air. Do not let the slide dry out in between any of the washing steps 19. The slides can be stored at room temperature in the dark and rescanned if needed (for up to 1 month). 20. To scan multiple slides, place the slides in adjacent slots to facilitate continuous scanning. 21. The feature extraction (FE) program will automatically assign a default Grid template and protocol. 22. DNA obtained from FFPE tissue is often degraded and will not require additional fragmentation.
Microarray-Based CGH and Copy Number Analysis of FFPE Samples
145
23. ULS-Cy3 has better labeling efficiency than ULS-Cy5. Therefore, samples being labeled with ULS-Cy5 are allowed to incubate for 100–120 min at 85°C. 24. If the total volume of the labeled DNA is less than 20 ml, then make it up with nuclease-free water. 25. The labeled DNA can be left overnight at −80°C.
Acknowledgements This work is supported by Kuwait Foundation for the Advancement of Sciences grant number 2006-1302-07, Kuwait University Grant number MG02/08 and Research Core Facility (RCF) grant number GM 01/01 and GM 01/05. Thanks go to Dian Ann Thomas for her excellent technical skills. References 1. Nielsen, C., Cantor, M., Dubchak, I., Gordon, D. and Wang, T. (2010) Visualizing genomes: techniques and challenges. Nature methods supplement 7, 5–15. 2. Hagenkord, J.M., Monzon, F.A., Kash, S.F., Lilleberg, S., Xie, Q., Kant, J.A. (2010) Arraybased karyotyping for prognostic assessment in chronic lymphocytic leukemia: performance comparison of affymetrix 10K2.0, 250K Nsp, and SNP6.0 arrays. J Mol Diagn. 12, 184–96. 3. Snijders, A.M., Meijer, G.A., Brakenhoff, R.H., van den Brule, A.J.C., and van Diest, P.V. (2000) Microarray technology in pathology: tool or toy?. Clin Pathol: Mol Pathol. 53, 289–294. 4. Brown, P.O. and Botstein, D. (1999) Exploring the new world of the genome with DNA microarrays. Nature genetics supplement 21, 33–37. 5. Hardiman, G. (2004) Microarray platforms: comparisons and contrasts. Pharmacogenetics 5, 487–502. 6. Al-Mulla, F., Behbehani, A.I., Bitar, M.S., Varadharaj, G., and Going, J.J. (2006) Genetic
7.
8.
9. 10.
11.
profiling of stage I and II colorectal cancer may predict metastatic relapse. Mod Pathol. 19, 648–58. Albertson, D.G. and Pinkel, D. (2003) Genomic microarrays in human genetic disease and cancer. Human Molecular Genetics 12, 145–152. Veltman, J.A. and De Vries, B.A. (2006) Diagnostic Genome Profiling: Unbiased Whole Genome or Targeted Analysis? Journal of Molecular Diagnostics, 8, 534–537. http://www.agilent.com User Guide. © Agilent Technologies, Inc. 2007. Hostetter, G., Kim, S. Y., Savage, S., Gooden, G.C., Barrett, M., Zhang, J., et al. (2010) Random DNA fragmentation allows detection of single-copy, single-exon alterations of copy number by oligonucleotide array CGH in clinical FFPE samples. Nucleic Acids Research 38, e9. De Bruyne, V., Al-Mulla, F., and Pot, B. (2007). Methods for microarray data analysis. Methods Mol Biol. 382, 373–391.
wwwwwwwwwwwwwww
Chapter 10 Microarray Profiling of DNA Extracted from FFPE Tissues Using SNP 6.0 Affymetrix Platform Marianne Tuefferd, An de Bondt, Ilse Van den Wyngaert, Willem Talloen, and Hinrich Göhlmann Abstract High-density oligonucleotide microarrays are commonly used for GWAS studies as well as for tumor genome alteration identifications. The recent Affymetrix Genome-Wide SNP 6.0 microarray generation has two major advantages: (1) showing high genome coverage and (2) starting with very small amount of DNA material. The hybridization protocol needs to be standardized and highly reproducible, as DNA is first digested by restriction enzymes and then PCR-amplified to reduce genome complexity. Especially the restriction digestion step is highly sensitive to degradation of the initial material. The stronger the sample is degraded, the lower the number of restriction sites still present in the genome, and hence the less-efficient amplification step. Paraffin-embedded material generally only allows to extract partially degraded DNA, and therefore is difficult to analyze using SNP array technology. We and others (Jacobs et al., Cancer Res 67:2544–2551, 2007; Tuefferd et al., Genes Chromosomes Cancer 47:957–964, 2008) have shown that target preparation protocol can be adjusted to improve hybridization performances. The final in silico data analysis procedure should be modified accordingly to extract most of the biological information from the signal measured. By optimizing these crucial steps, it is possible to use Affymetrix SNP array 6.0 technology in the context of genome variation, even for FFPE partially degraded material. This opens a lot of potential for large retrospective series of samples. Key words: Oligonucleotide microarrays, Single-nucleotide polymorphisms, Copy number alterations, PCR amplification, Restriction enzyme, Fragment length, Data analysis
1. Introduction Most of the available high-content technologies for genomic studies have been developed for high-quality DNA. The use of DNA extracted from FFPE material for genomic studies is particularly challenging. The Affymetrix® platform based on SNP microarrays is one of the most standardized technologies available. Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_10, © Springer Science+Business Media, LLC 2011
147
148
Tuefferd et al.
The last generation SNP 6.0 array features on a single array more than 906,600 single-nucleotide polymorphisms (SNPs) and 946,000 copy number variants (CNVs) markers, with a median marker spacing of less than 700 bases. Starting with as few as 500 ng of genomic DNA, the target synthesis is performed using whole genome sampling analysis (1, 2), where DNA is first digested by specific restriction enzymes, and corresponding adaptors are ligated to the sticky ends of the restrictions sites. After amplification, PCR products are purified and fragmented to reach a maximum size of 200 bases long. Starting with partially degraded DNA such as from FFPE samples, the amplification and fragmentation steps might not give optimal products both in size and yield. Protocol steps need to be adjusted according to the DNA quality to reach sufficient amount of target. Knowing that the hybridization quality is dependent on the DNA integrity, genomic polymorphisms and rearrangements can still be evaluated from FFPE samples as long as potential technical artifacts such as fragment length and composition are taken into account during the analysis. The examples presented here are taken from a small lung cancer series described previously (3), where both fresh-frozen (FF) and FFPE material were available from the same tumor samples.
2. Materials 2.1. FFPE Samples
As a very small amount of DNA is necessary for hybridization on Affymetrix SNP 6.0 arrays (250 ng for each restriction enzyme digestion), starting from around five slices of 10 mm thick allows to reach up to 40 mg of DNA material using QIAGEN QIAmp minikit (QIAGEN, Valencia, CA, USA). DNA extracted from FFPE samples should be diluted with low concentration of salt or chelating agent (50 ng/mL in reduced EDTA TE-buffer: 10 mM Tris HCL, 0.1 mM EDTA, pH 8.0) to prevent enzymatic activity inhibition.
2.2. Target Preparation
All steps except the PCR product purification include the reagents and consumables described in the Affymetrix® Genome-Wide Human SNP 6.0 manual. The amounts listed are sufficient for processing 48 samples.
2.2.1. Sty Restriction Enzyme Digestion
1. One vial BSA (100×; 10 mg/mL). 2. One vial NE Buffer 3 (10×). 3. One vial Sty I (10 U/mL; New England Biolabs). 4. 2.5 mL AccuGENE® Water, molecular biology grade.
Microarray Profiling of DNA Extracted from FFPE Tissues 2.2.2. Sty Ligation
149
1. One vial T4 DNA Ligase (400 U/mL; New England Biolabs). 2. One vial T4 DNA Ligase Buffer (10×). 3. One vial Adaptor, Sty (50 mM). 4. 10 mL AccuGENE water, molecular biology grade.
2.2.3. Sty PCR Amplification
In addition to the Sty ligated samples, the following reagents are necessary: 1. 15 mL AccuGENE water, molecular biology grade. 2. One vial PCR Primer 002 (100 mM). 3. Reagents from Clontech TITANIUM™ DNA Amplification Kit [dNTPs (2.5 mM each), GC-Melt (5 M), TITANIUM™ Taq DNA Polymerase (50×), TITANIUM™ Taq PCR Buffer (10×)].
2.2.4. Nsp Restriction Enzyme Digestion
1. One vial BSA (100×; 10 mg/mL). 2. One vial NE Buffer 2 (10×). 3. One vial Nsp I (10 U/mL; New England Biolabs). 4. 2.5 mL AccuGENE® Water, molecular biology grade.
2.2.5. Nsp Ligation
1. One vial T4 DNA Ligase (400 U/mL; New England Biolabs). 2. One vial T4 DNA Ligase Buffer (10×). 3. One vial Adaptor, Nsp (50 mM). 4. 10 mL AccuGENE water, molecular biology grade.
2.2.6. Nsp PCR Amplification
In addition to the Sty ligated samples, the following reagents are necessary: 1. 15 mL AccuGENE water, molecular biology grade. 2. One vial PCR Primer 002 (100 mM). 3. Reagents from the Clontech TITANIUM™ DNA Amplification Kit [dNTPs (2.5 mM each), GC-Melt (5 M), TITANIUM™ Taq DNA Polymerase (50×), TITANIUM™ Taq PCR Buffer (10×)].
2.2.7. PCR Product Purification and Elution
For this step, the following material is used instead of the Millipore purification system, as described on the Affymetrix® GenomeWide Human SNP 6.0 manual: 1. DNA Amplification Clean-Up Kit from Clontech (Mountain View, CA, USA) containing RB buffer and 1–4 Clean-Up Plates (Clontech). 2. BioRobot Universal System (QIAGEN, Valencia, CA, USA) Vacuum manifold.
150
Tuefferd et al.
3. 3 mL EDTA, diluted to 0.1 M (working stock is 0.5 M, pH 8.0). 4. 5 mL RB Buffer; 75 mL AccuGENE® water molecular biology grade. 2.2.8. Fragmentation
1. One vial Fragmentation Buffer (10×). 2. One vial Fragmentation Reagent (DNase I). 3. 1 mL AccuGENE® water, molecular biology grade.
2.2.9. Labeling
1. One vial DNA Labeling Reagent (30 mM). 2. One vial Terminal Deoxynucleotidyl Transferase (TdT; 30 U/mL). 3. One vial Terminal Deoxynucleotidyl Transferase Buffer (TdT Buffer; 5×).
2.2.10. Hybridization
1. 5 mL Denhardt’s Solution (50×). 2. 1.5 mL DMSO (100%). 3. 0.5 mL EDTA (0.5 M). 4. 1 mL Herring Sperm DNA (HSDNA; 10 mg/mL). 5. 500 mL Human Cot-1 DNA® (1 mg/mL). 6. 80 g MES Hydrate SigmaUltra; 200 g MES Sodium Salt. 7. 16 mL Tetramethyl Ammonium Chloride (TMACL; 5 M). 8. 10 mL Tween-20, 10%. 9. 250 mL Oligo Control Reagent (OCR).
3. Methods The main limitations of using FFPE samples for Affymetrix platform are the level of DNA degradation and the potential inhibitory effects associated with formalin cross-linking of nucleic acids. Different steps of the Affymetrix® Genome-Wide Human SNP 6.0 protocol can be adjusted to improve the target preparation. 3.1. Evaluation of Genomic DNA Quality
Genomic DNA integrity should be evaluated on a 1 or 2% TBE agarose gel. FFPE samples are expected to show various levels of degradation, represented by a smear during DNA migration on the gel (Fig. 1a). The larger the size of the genomic DNA (and higher the smear detected on the gel), the better the DNA quality is, and more biological information can be detected from the sample. From our experience, FFPE-extracted genomic DNA length varies between 100 and 700 bp.
Microarray Profiling of DNA Extracted from FFPE Tissues
151
Fig. 1. DNA quality comparing high-quality fresh-frozen (FF) tumor samples to paraffin-embedded material (FFPE). (a) Example of genomic DNA extracted using QIAGEN QIAmp minikit (run on 1.2% TBE agarose gel. Smart ladder Eurogentec) (b) Example of PCR amplification products. Left panel: from DNA of seven FF samples, run on 2% TBE agarose gel (with Invitrogen 25-bp ladder). Average product distribution is between 200 and 1,000 bp. Right panel: example of 14 FFPE samples run on 1.2% TBE agarose gel (with Smart ladder Eurogentec on the left of the gel and Invitrogen 25-bp ladder on the right). Average product distribution is between 100 and 400 bp.
3.2. T arget Preparation
The PCR amplification step and the fragmentation protocol need to be adjusted for FFPE samples. The other steps of target preparation are described in details in Affymetrix® Genome-Wide Human SNP Nsp/Sty 6.0 User Guide.
3.3. Sty and Nsp Digestion
Aliquot 5 mL of each DNA to the corresponding wells of two 96-well reaction plates. Two replicates of each sample are required for this protocol: one for Nsp and one for processing Sty. The exact methodology described in Affymetrix® Genome-Wide Human SNP 6.0 manual is applied.
3.4. Sty and Nsp Ligation
The exact methodology described in Affymetrix® Genome-Wide Human SNP 6.0 manual is applied.
3.5. Sty and Nsp PCR Amplification
PCR conditions have been optimized to preferentially amplify fragments in the 200–1,100 bp size range. Since genomic DNA from FFPE samples are degraded (Fig. 1b), less binding sites for adaptor ligation are generated, and less PCR primer can bind, leading to less-efficient PCR and lower PCR yields (Fig. 2).
152
Tuefferd et al.
High quality DNA
Degraded DNA
I. Restriction enzyme digestion
partial digestion II. Adaptor ligation
less binding sites for adaptor III.PCR Amplification
partial amplification Fig. 2. Why PCR is less efficient for FFPE samples showing degraded DNA StyI and Nsp I enzymes recognize specific genomic sites. When DNA is partially degraded, few restriction sites are generated and only few adaptors can bind. As a consequence, few PCR primers bind the adaptor reducing the PCR amplification efficiency. The larger the fragment generated by restriction enzyme digestion is, the higher risk it has to be degraded and not amplified properly.
After ligation step, 100 mL of Sty as well as 100 mL of Nsp digested and ligated products are available for amplification. Only 10 mL are needed for each PCR, so in practice, it is possible to run up to nine independent PCRs for each Sty and Nsp digested and ligated products. Conditions and timing applied for PCR amplification are the same as described on the Affymetrix® Genome-Wide Human SNP 6.0 manual. Only the number of aliquot changes. To reach sufficient PCR yield from degraded samples, nine PCRs are set up by default per sample. For some samples showing limited DNA degradation, less reactions will be necessary to reach the expected yield of products for fragmentation, but the PCR efficiency is difficult to predict. For those, after evaluating PCR product yield on a gel, it is more optimal to keep the same proportion between Sty and Nsp PCR products (as Nsp I enzyme is less efficient than Sty I in the Affymetrix® protocol, it is recommended to pool three Sty with four Nsp PCR-amplified products). 3.6. PCR Product Purification and Elution
Several methods are available for PCR cleanup, all being timeconsuming. From our experience, the Clontech DNA Amplification Clean-Up Kit, suggested in earlier Affymetrix® SNP arrays protocols such as 500K array set (for a full description see GeneChip® Mapping 500K Assay Manual (4)), is easy to perform. To each PCR product, 8 mL of diluted EDTA (to a concentration of.0.1 M) is added. It is recommended to vortex the center of each plate at high speed for 3 s. Then, spin down each plate at 2,000 rpm (943 g) for 30 s.
Microarray Profiling of DNA Extracted from FFPE Tissues
153
3.6.1. Purify the PCR Products
The Clean-Up Plates are prepared according to the Clontech Clean-Up Plate Handbook. Samples from the same row and well of each PCR product plate (as standard from three first plates for Sty or Nsp) are pooled to the corresponding row and well of the Clean-Up Plate to reach a volume of about 320 mL by well. Unused wells should be covered. The Clean-Up Plate is transferred onto a vacuum manifold. The PCR products are concentrated by a constant vacuum of 600 mbar until the wells are dry (approximately 1.5–2 h). The wells are reloaded until all single PCRs are added and concentrated by constant vacuum of 600 mbar. When the wells are completely dried, the PCR products are washed by adding 50 mL of Biology-Grade water to each well and dried completely, keeping the vacuum throughout. The wells are dried for around 20 min. This step has to be repeated two additional times for a total of three washes. After the third wash, samples need to be dried completely (may take more than an hour).
3.6.2. Elute the PCR Products
Subsequently, 40 mL of RB buffer is added to each completely dried well and shaken moderately on a plate shaker for 10 min at room temperature. The purified DNA is transferred from the Clean-Up Plate to fresh tubes. Finally, the concentration is measured with the Nanodrop spectrophotometer (see Note 1).
3.7. Fragmentation
DNase I enzyme is used for unspecific fragmentation of PCR products into small fragments of around 200 base large. Since FFPE PCR products are smaller than expected from high-quality material, the fragmentation time needs to be shortened (Fig. 3).
Fig. 3. Enzymatic fragmentation performances (run on 4% TBE agarose gel. Right panel Smart ladder Eurogentec. Left panel E-gel® low range quantitative DNA ladder, Invitrogen). Left panel: Example of PCR product fragmentation from 12 fresh-frozen samples following Affymetrix® protocol (1 × 35 min): the fragmentation resulted in a smear of desired length (below 200 bp). Right panel: PCR product fragmentation from eight FFPE samples after 10 min fragmentation. Fragmentation is incomplete after 10 min on FFPE samples and efficiency varies from one sample to the other. A stronger band could be observed around 25 bp, corresponding to primer dimers caused by inefficient PCR.
154
Tuefferd et al.
Depending on the level degradation of the samples, the fragmentation time necessary can vary. A stepwise fragmentation is then more accurate. We modified the protocol to reach the expected size by performing the reaction during 10 min (instead of 35 min as recommended in the Affymetrix® protocol). The reaction is stopped by increasing the temperature to 95°C for 15 min, and the samples are kept at 4°C. The fragmentation efficiency is checked on agarose gel. If the expected 200 bp smear is not observed on the gel, fresh DNase I (5 mL from the fragmentation master mix) is added, and the reaction is performed again for 10 min. The reaction should be repeated until the expected band around 200 bp is observed on the gel. 3.8. Labeling
The exact methodology described in Affymetrix® Genome-Wide Human SNP 6.0 manual is applied.
3.9. Hybridization
The exact methodology described in Affymetrix® Genome-Wide Human SNP 6.0 manual is applied.
3.10. Data Analysis
The scanned image of the array is overall less bright for FFPE samples as compared to fresh-frozen samples, and the distinction between copy number probes and SNP probes is more difficult to make (Fig. 4), suggesting an hybridization of lower quality. It is
3.10.1. Quality Control
Fig. 4. Scanned Images of Genome-Wide Human SNP Array 6.0 (.DAT files). (a) hybridized with high-quality DNA material. The four quadrants pattern is recognizable (corresponding to SNP probes), and limited by a central cross containing the copy number probes. (b) FFPE sample hybridized on SNP array 6.0. The overall chip is less bright, especially the SNP probes. The background intensity here is very bright, as the contrast and overall intensities had to be adjusted.
Microarray Profiling of DNA Extracted from FFPE Tissues
155
often necessary to adjust the image settings and increase the overall brightness to identify the copy number probes. Quality controls commonly used to assess hybridization performance using Affymetrix® SNP 6.0 (5, 6), usually show lower values when applied to FFPE material than those to highquality material and have to be considered as indicative only. The Overall SNP call rate associated with hybridization of a sample is equal to the number of SNPs receiving an AA, AB, or BB genotype call using Birdseed genotyping algorithm (5) divided by the total number of SNPs on the chip. Starting with highquality material, the expected overall call rate is above 90%. Starting with FFPE samples, the observed SNP call rates are always lower and generally vary between 70 and 95%. From our experience, arrays showing call rates below 65% are particularly difficult to use for copy number analysis, because of the high level of noise in the signal measured. 3.10.2. Preprocessing
Preprocessing steps take into account undesired sources of variations by standardizing the measured signal. Quite a large number of algorithms are available for preprocessing SNP 6.0 arrays, and new methodologies are proposed regularly. Some of them have implemented some interesting features for FFPE samples analysis. We describe the potential confounding factors to take into account for the analysis. Nannya et al. (7) have highlighted the importance of taking into account the length and GC content of the hybridized PCR products to improve the signal-to-noise ratio. The shorter the hybridizing fragment, the stronger the signal measured. The signal also increased with percentage of GC nucleotides in the fragment. Correcting for these two confounding effects is of high importance, especially when analyzing data from FFPE samples. This can be understood as hybridization quality is directly linked to DNA integrity and therefore to the size of fragments generated. Jacobs et al. (8) suggested to filter out probeset signals associated with large fragments. This improvement of performance is associated with a reduction in resolution, as only part of the probesets are considered for the analysis. We observed the same trend as Nannya et al. (7) when comparing fresh-frozen and FFPE material from the same tumor sample (Fig. 5) using the preprocessing algorithms dChip (9) (no correction) or oligo (10) (correcting for fragment length and GC content of PCR products). The linear effect of fragment could be highlighted in FFPE samples only (the shorter the fragment, the higher the intensity measured). The GC effect was observed on both high-quality and FFPE material. The importance of sequence composition seems to be accentuated on degraded samples, where generated fragments are smaller.
156
Tuefferd et al.
Fig. 5. Intensity values for fresh-frozen and FFPE samples from the same tumor depending on fragment length (a) or GC content (b). The fitted line shows the fitted intensity values based on kernel regression smoother. Above panel: using dChip algorithm for preprocessing (not taking into account GC content and fragment length). Below panel: using oligo algorithm for preprocessing (accounting for PCR). Reproduced from Tuefferd, 2008 (3). Courtesy from Wiley.
Ignoring the GC content dependency would classify FC poor regions falsely as deletions and GC-rich bands as amplifications (Fig. 6) for both fresh-frozen and FFPE material. This would result in a copy number status that would be in concordance with the isochore map of the genome (11). 3.10.3. Smoothing the Signal
In order to increase signal-to-noise ratio, consecutive signal intensities are often summarized in one value. This step allows to reduce dispersion of the signal, keeping a reasonable coverage of the genome. Averaging the intensities of 50 consecutive probesets using SNP 6.0 platform allows to keep a coverage in the order of BAC arrays. If this smoothing method might be associated with a loss of sensitivity in detection of microamplifications/microdeletions, from our experience, most of the “high-level” amplifications (defined as being smaller than 2 Mb but present at more than eight copies) could still be detected on FFPE material. Focal amplifications are difficult to identify without smoothing the signal (Fig. 7). They become detectable after grouping 10–20 consecutive probesets. If this smoothing step reduces precision detection, local variations are still identified.
3.10.4. Comparison of SNP vs. CN Probesets
In the SNP 6.0 system, probe sequences are organized in either SNP probesets (four exact probe replicates targeting allele A and four targeting allele B of one SNP) or CN probesets (one probe targeting either a copy number polymorphic region or a
Microarray Profiling of DNA Extracted from FFPE Tissues
157
Fig. 6. Importance of GC content correction for copy number analysis. FFPE and FF material from the same tumor showing 1q amplification are compared. Preprocessed signal intensities are represented along the physical position. GC content from PCR products is represented by color gradient. The horizontal bold lines represent copy number analysis results (centered = nonmodified, −1 = deletion; +1 = amplification). GC effect is slightly higher in FFPE than FF material. Not correcting for this potential confounding factor (upper panel ) leads to misclassification of the first part of the 1p arm, detected as amplified. Reproduced from Tuefferd, 2008 (3). Courtesy from Wiley.
onpolymorphic region). For copy number analysis, it is theoretin cally possible to combine both types of probesets, increasing the genome coverage. Considering probe signal before any preprocessing method is applied, the distribution of the signal differs for the two kinds of probes (see Note 2). It is important to note that the variability of signal detection is different between SNP and CN probesets. After preprocessing step, variance is often twice more important for CN probesets than SNP probesets for FFPE-extracted material. This is not only because the genomic regions targeted are different (punctual variations vs. nonpolymorphic regions) but probably also because SNP probesets include more probes than CN probesets. It seems that SNP probesets are more adapted to degraded material than CN probesets and that probe replicates increase the signal robustness (Fig. 8).
Fig. 7. Effect of different smoothing windows on focal amplification detection. CCND1 locus example (where an amplification of more than ten copies could be detected by FISH); comparing FF and FFPE material from the same tumor sample.
Fig. 8. Comparing SNP and CN probesets considering all autosomes of a tumor sample where both FF and FFPE materials are available. Oligo preprocessing was applied independently to SNP probes and CN probes (correcting for fragment length and GC content of PCR product). Fifty consecutive probesets were averaged to smooth the signal. Reproduced from Tuefferd, 2008 (3). Courtesy from Wiley.
Microarray Profiling of DNA Extracted from FFPE Tissues
159
4. Notes 1. Following the Affymetrix® protocol, a final PCR product concentration of 4.5–6.0 mg/mL in 55 mL of buffer is needed. Depending on the PCR performance, it might be necessary to adjust the RB buffer dilution (to 40 mL) and therefore to downscale the fragmentation and labeling steps. Working this way, a minimum of 180 mg of PCR product (Nsp I + Sty I) is needed instead of 225 mg as recommended by Affymetrix®. 2. SNP probes vs. CN probes show before any preprocessing a different behavior. The intensity distribution varies (Fig. 9a), especially for values between 6 and 8. It is interesting to note that for high-quality material, variability is twice as high for SNP probes than for CN probes. For corresponding FFPE material, variability of SNP probes and CN probes is equally high. This observation suggests that the two kinds of probes should be preprocessed independently and that CN probes might be more sensitive to degradation.
a CN probes signal
QQplot : Signal intensities are ordered from the lowest to the highest
If the signal distribution is the same for both probe types, it should be aligned along first diagonal
SNP probes signal
b Variance
SNP probes
CN probes
FF material
1.40
0.7
FFPE material
1.55
1.53
Fig. 9. Intensity distribution comparing probe types. (a) QQPlot: log2 signal distribution of CN probes against SNP probes ordered from the lowest to the highest intensity for the same tumor sample, illustrating that, overall CN probes result in higher intensities (b) Variance of the signal of 15 tumor samples where both FF and FFPE materials were available. SNP probes and CN probes are considered separately.
160
Tuefferd et al.
References 1. Kennedy, G.C., Matsuzaki, H., Dong, S., Liu, W.M., Huang, J., Liu, G., et al. (2003) Largescale genotyping of complex DNA. Nat. Biotechnol. 21, 1233–1237. 2. Affymetrix® Genome-Wide Human SNP Nsp/ Sty 6.0 User Guide, www.affymetrix.com. 3. Tuefferd, M., De Bondt, A., Van Den Wyngaert, I., Talloen, W., Verbeke, T., Carvalho, B., et al. (2008) Genome-wide copy number alterations detection in fresh frozen and matched FFPE samples using SNP 6.0 arrays. Genes Chromosomes Cancer 47, 957–964. 4. GeneChip® Mapping 500K Assay Manual, www.affymetrix.com. 5. Korn, J.M., Kuruvilla, F.G., McCarroll, S.A., Wysoker, A., Nemesh, J., Cawley, S., et al. (2008) Integrated genotype calling and association analysis of SNPs, common copy number polymorphisms and rare CNVs. Nat. Genet. 40, 1253–1260. 6. Carvalho, B., Louis, T.A., and Irizarry, R.A. (2010) Quantifying uncertainty in genotype calls. Bioinformatics 26, 242–249. 7. Nannya, Y.M., Sanada, K., Nakazaki, N., Hosoya, L., Wang, A., Hangaishi, M., et al.
(2005) A robust algorithm for copy number detection using high-density oligonucleotide single nucleotide polymorphism genotyping arrays. Cancer Res. 65, 6071–6079. 8. Jacobs, S., Thompson, E.R., Nannya, Y., Yamamoto, G., Pillai, R., Ogawa, S., et al. (2007) Genome-wide, high-resolution detection of copy number, loss of heterozygosity, and genotypes from formalin-fixed, paraffinembedded tumour tissue using microarrays. Cancer Res. 67, 2544–2551. 9. Lin, M., Wei, L.J., Sellers, W.R., Lieberfarb, M., Wong, W.H., and Li, C. (2004) dChipSNP: significance curve and clustering of SNP-array-based loss-of-heterozygosity data. Bioinformatics 20, 1233–1240. 10. Carvalho, B., Bengtsson, H., Speed, T.P, and Irizarry R.A. (2007) Exploration, normalization, and genotype calls of high-density oligonucleotide SNP array data. Biostatistics 8, 485–499. 11. Costantini, M., Clay, O., Auletta, F., and Bernardi, G. (2006) An isochore map of human chromosomes. Genome Res. 16, 536–541.
Chapter 11 Whole Genome Amplification of DNA Extracted from FFPE Tissues Mira Bosso and Fahd Al-Mulla Abstract Whole genome amplification systems were developed to meet the increasing research demands on DNA resources and to avoid DNA shortage. The technology enables amplification of nanogram amounts of DNA into microgram quantities and is increasingly used in the amplification of DNA from multiple origins such as blood, fresh frozen tissue, formalin-fixed paraffin-embedded tissues, saliva, buccal swabs, bacteria, and plant and animal sources. This chapter focuses on the use of GenomePlex® tissue Whole Genome Amplification Kit, to amplify DNA directly from archived tissue. In addition, this chapter documents our unique experience with the utilization of GenomePlex® amplified DNA using several molecular techniques including metaphase Comparative Genomic Hybridization, array Comparative Genomic Hybridization, and real-time quantitative polymerase chain reaction assays. GenomePlex® is a registered trademark of Rubicon Genomics Incorporation. Key words: Formalin-fixed, Paraffin-embedded tissue, Whole genome amplification, Archived, DNA, Comparative genomic hybridization, Array CGH, Real-time PCR, GenomePlex®, PCR
1. Introduction The basic principle behind whole genome amplification (WGA) is to copy and amplify whole DNA molecules into an ample amount, which is different from the conventional polymerase chain reaction (PCR) technique, which facilitates amplification of only a specific segment within a DNA sequence. This feature provides the immense advantage of providing sufficient DNA amounts for large-scale multigenetic analysis, hence the importance of WGA systems. WGA can be applied to any human disease research in which DNA analysis is required and its paucity represents an impediment to any researcher. Furthermore, it permits the analysis of Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_11, © Springer Science+Business Media, LLC 2011
161
162
Bosso and Al-Mulla
minute DNA content of microdissected individual cells by means of individually amplifying and analyzing DNA from different microdissected areas of tumor, hence giving the opportunity to study the extent and influence of tumor heterogeneity in cancer progression and metastasis (1). WGA techniques have been successfully utilized in several clinical applications (1, 2), such as preimplantation genetic diagnosis (3–5), noninvasive prenatal diagnosis (1), archival DNA samples (6), microbial studies, and forensics (1). In addition, several WGA techniques have also been integrated with other genotyping technologies such as DNA microarrays and SNP GeneChip® platforms. Since formalin-fixed, paraffin-embedded (FFPE) tissue or “archival” samples provide a rich source of genetic material that can be preserved for decades, several WGA techniques have been used to amplify DNA of FFPE tissue origin but with variable success. Different types of WGA techniques, utilizing a variety of mechanisms that are PCR or non-PCR-based, have been developed. The choice to utilize any specific WGA method depends on multiple parameters, of which the most important is DNA quantity and quality. For example, poor-quality fragmented DNA is not compatible with multiple displacement amplification (MDA) technique as it requires long DNA fragments for optimal amplification (2). On the other hand, poor-quality DNA is successfully amplified by the degenerate oligonucleotide primed-PCR (DOP-PCR), GenomePlex®, primer extension preamplification (PEP) and T7-based linear amplification of DNA (TLAD) techniques (2). However, previous work has shown that DOP-PCR and PEPPCR may introduce significant bias in the amplified DNA (3). GenomePlex® WGA was introduced to markets by Rubicon Genomics Inc. in 2002, and since then, researchers have demonstrated promising results and performance in terms of representative DNA amplification as well as complete genome coverage with minimal bias introduction (7, 8). This technique has been chosen by the National Cancer Institute to amplify DNA samples for The Cancer Genome Atlas Project (http://www.cancergenome.nih.gov). Other institutions such as Sanger Institute and John’s Hopkins University have reported excellent performance of GenomePlex® amplified DNA in different genotyping analysis (http://www.rubicongenomics. com/technology). GenomePlex® WGA technique is an adaptor ligation-based PCR method (6), which uses universal adaptor primers to anneal to randomly fragmented DNA. The first set of primers used to anneal to the fragmented DNA is enhanced oligonucleotide adaptor primers with degenerate 3′ end and a constant fixed 5′ end. The degenerate 3′ end design allows annealing of the primers at millions of sites along the DNA fragments. After PCR with low annealing
Whole Genome Amplification of DNA Extracted from FFPE Tissues
163
and extension temperatures, a library of amplifiable DNA molecules that is representative of the original fragmented DNA is generated (9–11). The adaptor primers provide sites at the 5′ fixed end to which adaptor-specific primers can anneal and facilitate amplification of the library of molecules (9). GenomePlex® tissue WGA kit allows DNA amplification directly from FFPE tissue as well as frozen, RNA later™-preserved, and fresh tissues. This kit removes the necessity for tiresome organic extractions from the tissue and DNA purification before amplification procedure (8, 9). The method involves three steps: random fragmentation, OmniPlex® library preparation, and PCR amplification. The steps can be performed in a single reaction tube within 3 h. GenomePlex® technology is able to multiply the starting amount of DNA to 400– 600 folds more, thus generating 4–6 mg of amplified DNA product from as little as 0.1 mg of FFPE tissue (9). Here, we describe an optimized protocol for utilizing GenomePlex® on DNA extracted from FFPE sections (8) and validate the technology using a variety of techniques. We show that GenomePlex® is suitable to amplify partially degraded samples for genome-wide analysis including metaphase and array-based CGH. However, caution must be exercised when the technology is applied for the analysis of gene or exon dosage using quantitative PCR.
2. Materials 2.1. Deparaffinization of FFPE Tissue Sections Mounted on Microscopic Slides
1. Fresh xylene. 2. Absolute ethanol. 3. Deionized water. 4. Copland jars or compartments. 5. Microscopic slides (75 × 25 mm).
2.2. Microdissection of Deparaffinized Tissue and Tissue Lysate Formation
1. Conventional light microscope (to guide tissue microdissection). 2. 27-Gage needle (microdissecting tool). 3. Hematoxylin and eosin (H&E) preprepared slides from the same corresponding sample as a reference to guide selection of target tissue area to be microdissected. 4. 0.2-ml PCR tubes. 5. Microcentrifuge (Microfuge I8 Centrifuge, Beckman Coulter, Inc., Harbor Boulevard, USA). 6. GeneAmp® PCR System Framingham, MA).
9700
(Applied
Biosystems,
164
Bosso and Al-Mulla
7. CelLytic Y lysis solution (Sigma–Aldrich, Inc.). 8. Proteinase K solution (Sigma–Aldrich, Inc.). 2.3. GenomePlex ® Tissue WGA Procedure
GenoemPlex® Tissue Whole Genome Kit (Sigma–Aldrich, Inc., Saint Louis, MO, USA). The nature and design of each component of the kit is proprietary of Sigma–Aldrich Inc. WGA5 kit contents: 1. CelLytic Y Lysis solution. 2. Proteinase K solution from Tritirachium album for molecular biology >800 U/mL. 3. 1× Library preparation buffer. 4. Library stabilization solution. 5. Library preparation enzyme. 6. 10× Amplification master mix. 7. DNA polymerase. 8. Control Human Genomic DNA. 9. Nuclease-free water.
2.4. Purification of Amplified DNA Product
QIAquick PCR Purification Kit (QIAGEN Group Inc., Germantown, MD, USA). Kit components: 1. QIAquick spin columns. 2. Buffer PB. 3. Buffer PE. 4. Collection tubes (2 ml).
2.5. Agarose Gel Electrophoresis (1.5% Agarose)
1. Agarose. 2. Tris–EDTA buffer solution 100× for molecular biology (1.0 M Tris–HCl, pH approximately 8.0, containing 0.1 M EDTA 10× and 1×). 3. Directload™ Step Ladder 50 bp. 4. Deionized water.
2.6. NanoDrop Measurements
Nanodrop® ND-1000 spectrophotometer (Nanorop® Technologies, Inc., Wilmington, USA).
3. Methods 3.1. Deparaffinization and Microdissection of FFPE Microscopic Tissue Sections
1. Deparaffinizing FFPE tissue sections (see Note 1) is achieved by serial immersion of 5-mm thick tissue, mounted on microscopic slide, into three xylene compartments consecutively for 10 min each. This is followed by immersion in 100%, 100%, 95%, 95%, and 70% prepared ethanol jars successively
Whole Genome Amplification of DNA Extracted from FFPE Tissues
165
Fig. 1. Image of a microscopic slide with three tissue sections mounted on it. Both dotted circles demarcate areas where tissue was microdissected.
for 10 min each. Finally, the slides are transferred to a deionized water compartment. 2. Selection of cell population of interest to investigate. For instance, tumor tissue areas, is guided by corresponding H&Estained slides. Microdissection is performed by simple manual gentle scraping of chosen area using the tip of a fine needle (see Note 2). Thick tissues (5–10 mm) can be microdissected for WGA process (Fig. 1). A weight range of 0.1–0.4 mg of microdissected tissue was successfully amplified in our laboratory (see Note 3). Kit brochure shows that successful amplification is obtainable from as small as 0.1–1 mg of tissue (see Note 4). 3.2. Tissue Lysate Formation
The dissected tissue detaches from the slide and forms small dark clumps of tissue that can be collected on the tip of the needle. The collected dissected tissue on the tip of the needle is placed carefully into a 0.2-ml PCR tube containing 24 ml of CelLysis solution and 6 ml of Proteinase K (Sigma–Adrich, Inc., Saint Louis, MO, USA). Tubes are mixed thoroughly for 30 s and centrifuged briefly at 18,000 × g. The PCR tubes are then placed in thermal cycler at 60°C for 60 min, then at 99°C for 4 min. Afterward, tubes are removed from thermal cycler (see Notes 5–7).
3.3. OmniPlex® Library Preparation
Aliquot 1 ml of tissue lysate (see Note 8) in a new 0.2-ml PCR tube to which 2 ml of library preparation buffer and 1 ml of library stabilization solution are added (see Note 9). Tubes are mixed thoroughly for 30 s and centrifuged briefly, then placed in GeneAmp® PCR System 9700 set at 95°C for 2 min (see Note 10). When incubation time is finished, tubes are removed from PCR system and placed instantly on ice. After that, 1 ml of library preparation enzyme is added to the tubes to be reinserted in PCR system previously set at 16°C for 20 min, 24°C for 20 min, 37°C for 20 min, 74°C for 5 min, and finally 4°C hold (see Notes 11 and 12). It is strongly recommended that the library preparation is done in, at least, triplicates for each sample (see Note 13).
166
Bosso and Al-Mulla
3.4. P CR Amplification
To amplify the prepared library of DNA, 7.5 ml of 10× amplification master mix, 48.5 ml of nuclease-free water and 5 ml of WGA DNA polymerase are added (all supplied with WGA5 kit) to the PCR tubes from the previous step (8). The tubes are then incubated in PCR system at 95°C for 2 min (denaturation), followed by 20 cycles of 94°C for 15 s (denature) and 65°C (anneal/extend) for 4 min and final hold step at 4°C (see Note 14). After completion of PCR amplification step, the PCR tubes are removed to be purified in the final step (8).
3.5. Purification of Amplified Products
Purification of amplified DNA is done using QIAquick PCR purification kit (PN 28104, QIAGEN group, Inc.,) following the manufacturer’s instructions (see Notes 15 and 16).
3.6. Assessment of GenomePlex ® Amplified DNA
1. Run product on 1.5% Agarose gel electrophoresis. A 50-bp ladder is loaded as a reference for DNA smear size. A 5-ml per purified amplified product is mixed with 1 ml of loading dye and loaded through gel holes. Ideal amplified DNA product shows as a smear that ranges in size between 150 and 1,000 bp (Fig. 2). 2. The concentrations of amplified DNA products are estimated by absorbance measurement at 260, 280 nm and 230 nm wavelengths using Nanodrop® ND-1000 spectrophotometer (Nanorop® Technologies, Inc., Wilmington, USA). The ratio 260/280 reveals the degree of purity of the DNA samples from protein contaminants. A ratio of less than 1.5 indicates poor DNA quality and significant protein contamination. 260/230 ratio demonstrates the degree of purity of DNA from other organic contaminants such as phenolate ion and thiocyanates. The yield of WGA-amplified DNA is 4–9 mg suspended in 30 ml of elution buffer (EB). High 260/280 and 260/230 ratios obtained in the range of 1.8–2 indicate high purity of the sample.
3.7. WGA5 Optimization for Tissue Lysate Formation
Different weights (<0.25, 0.25, 0.5, and 1 mg) and areas of tissue sections (10 mm thick) were microdissected from the same tissue sample to form tissue lysates in separate 0.2-ml PCR tubes. The microdissected areas corresponded to less than 1/4 (6 × 5 mm), 1/4 (10 × 6.5 mm), 1/2 (11 × 10 mm), and whole tissue section (20 × 12 mm) (Fig. 3). WGA was performed individually on 1 ml obtained from each prepared tissue lysate. WGA-amplified DNA obtained from a lysate of 0.25 mg of starting tissue demonstrated the best amplified smear on 1.5% agarose gel electrophoresis (Fig. 4).
3.8. WGA5 Verification of Representative DNA Amplification
All WGA techniques have demonstrated some sort of bias and varying degrees of success depending on the type and quality of DNA used, the genotyping methodology performed, and the WGA technique itself.
Whole Genome Amplification of DNA Extracted from FFPE Tissues
167
Fig. 2. Image of gel electrophoresis results done for purified GenomePlex® amplified DNA samples along with ladder DNA.
Fig. 3. Image of a microscopic slide with three-mounted tissue sections (10 mm thick) from the same sample. Dotted circles depict the amount of tissue areas microdissected.
We examined the reliability of the GenomePlex® WGA5 kit in providing complete and accurate genome coverage in copying DNA templates from partially degraded, minute DNA quantities using different molecular analyses. 3.8.1. Comparative Genomic Hybridization
Comparative genomic hybridization (CGH) technique enables detection of genome-wide chromosomal copy number alterations (gains/losses) in a target DNA sample compared to a reference normal sample. The target DNA and the reference normal DNA are separately labeled with different florescent dyes (green “target” and red “normal”) and then co-hybridized to normal metaphase chromosomal preparations on microscopic slides. This technique
168
Bosso and Al-Mulla
Fig. 4. GenomePlex® WGA5 amplified DNA products from different tissue lysates of the same sample. Lane 1; product from <1/4 section lysate, lane 2; product from 1/4 section lysate, lane 3; product from 1/2 section lysate, and lane 4; product from whole tissue section lysate. Lane 2 demonstrated the best amplified product with a strong white smear.
can detect a single copy loss only if the region is at least 5–10 Mb in length (12). However, this technique is more sensitive to copy number gains, as it can detect a tenfold amplification of at least 1 Mb region length. To investigate the reliability and ability of GenomePlex® WGA5 technique to maintain genome-wide copy number status, we tested seven pairs of DNA samples in the before (unamplified) and after (amplified) WGA process using CGH technique and compared CGH analysis results (13). Statistical analysis was performed to detect if there were any significant copy number changes induced by WGA5. Table 1 summarizes the statistical results that demonstrate excellent genome coverage and good overall maintenance of genome-wide copy number states. Kappa statistic (14, 15) was used to assess level of agreement between qualitative data of before and after GenoemPlex® WGA (Table 2). Error rate was also calculated by analyzing copy number variations in 39 chromosomal segments including the p and q arms of all nonacrocentric autosomes and the q arms of all acrocentric autosomes. The X and Y chromosomes were excluded
Whole Genome Amplification of DNA Extracted from FFPE Tissues
169
from the analysis. An error was called when a CNA (Copy Number Alteration) involving two or more bands was detected in the karyogram of the amplified sample and was not detected in the karyogram of the unamplified sample or vice versa (13) (Fig. 5). The percentage of error rate was calculated in each case according to the following equation: Error rate% = 3.8.2. Array CGH
Number of errors ´ 100. Total number of segments
Like metaphase CGH, array CGH offers the ability to view genomewide copy number changes but on a higher resolution. Instead of a metaphase spread substrate, an array of DNA fragments of known
Table 1 Statistical analysis results for pairs of metaphase-CGH profiles of all seven cases of colorectal cancer in the amplified and unamplified forms Sample pair ID
Kappa (degree of agreement) between before and after WGA
Error rate (%)
1
k = 0.505 P-value 0.0005
17
2
k = 0.27 P-value 0.01
25
3
k = 0.629 P-value 0.0001
7
4
k = 0.827 P-value 0.0001
5
5
k = 0.549 P-value 0.0001
10
6
k = 0.489 P-value 0.001
23
7
k = 0.526 P-value 0.0001
17
Table 2 Kappa (k) values interpretation chart as proposed by Landis and Koch in 1977 k
Interpretation
0.81–1
Almost perfect agreement
0.61–0.8
Substantial agreement
0.41–0.6
Moderate agreement
0.21–0.4
Fair agreement
0–0.2
Slight agreement
<0
Poor agreement
170
Bosso and Al-Mulla
Fig. 5. Scoring of errors in CGH profiles. (a) Part of a CGH profile of a GenomePlex® amplified sample, (b) part of a CGH profile of the unamplified sample. Black arrows on the amplified DNA CGH profile represent errors. An error was marked when a copy number alteration difference affecting two or more chromosomal bands in the amplified compared to the unamplified CGH profile. Numbers below ideograms are chromosomal numbers, and parentheses indicate the number of chromosomes counted.
chromosomal locus is positioned on an array slide. The genomic resolution depends on the map distance between target DNA markers positioned on the array or the length of the clones used. Individually labeled target and reference DNA are co-hybridized to the array slide (16, 17). Three pairs of GenoemPlex® WGA5 amplified and unamplified DNA samples were tested using NimbleGen® HG18 CGH WG Tiling v1.0 array (Cat. no. B4366-00-01). NimbelGen® hybridization kit (Cat. no. 05223474001) and NimbleGen® wash buffer kit (Cat. no. 05223504001) were used (Roche NimbleGen, Inc., Madison, USA). Results were analyzed, compared, and statistically studied to detect any significant copy number differences induced by WGA. A comprehensive and comparative visual assessment of array CGH single panel segmentation profiles of each of the three samples of the before and after WGA amplification, showed no significant differences found in gene copy number changes throughout all the chromosomes (Fig. 6). Another comparative visual assessment of array CGH multipanel profiles allowed a more detailed and comprehensive comparison on a chromosome-by-chromosome basis. The assessment shows no significant differences in copy number throughout all the chromosomes between before and after WGA treatment among all the three samples (Fig. 7).
Whole Genome Amplification of DNA Extracted from FFPE Tissues
171
Fig. 6. CGH copy number (CN) single panel plot profiles, profile A corresponds to the sample after WGA amplification, profile B corresponds to the same sample before WGA. The chromosome number (X-axis) is plotted against Log2 ratio values (Y-axis). A horizontal line above zero threshold represents an increase in CN. Line drawn below zero threshold represent a decrease in CN. Comparison between the two profiles showed no significant differences introduced in CN by the WGA method.
To investigate the degree of agreement statistically between each pair (before and after WGA) of array-CGH data, Bland– Altman Plots (18) were generated (see Fig. 8). Difference in log2 ratios between before and after WGA is plotted on the Y-axis and the means of both log2 ratios (ratio 1 + ratio 2)/2 are plotted on the X-axis (18, 19). The boundaries of agreement were determined by the following equation: [Mean ± 2 (standard deviation)]. Data dots plotted within these limits were considered in agreement, while dots plotted outside the limits were outliers (Table 3) and represent poor agreement (19). Moreover, MAD (median absolute deviation) statistic was calculated to measure variability of array-CGH data for each sample
172
Bosso and Al-Mulla
Fig. 7. Example of multipanel array-CGH profiles in which a plot of the normalized averaged Log2 ratios (Y-axis) for each chromosome is displayed independently. (a) Represents a sample’s a-CGH data of chromosomes 1, 2, and 3 in the WGAamplified state of the sample. (b) Represents a-CGH data of chromosomes 1, 2, and 3 of the same sample in its unamplified form. The X-axis shows the relevant position of the data point along the chromosome. Horizontal lines above or below the zero level indicate an increased or decreased copy number, respectively.
(Table 4). The MAD of corrected log2 ratios of all the chromosomes in each sample was calculated. The higher the MAD value, the lower the precision level. The MAD values were compared between before and after WGA states for each sample. There are three calculations to get the MAD value; first, the median of the actual data values, second, calculating the absolute value of the difference between each data value and the median, and finally computing the median of the set of difference.
Whole Genome Amplification of DNA Extracted from FFPE Tissues
173
Fig. 8. Bland–Altman plots of the calculated difference between log2 corrected ratios of a-CGH data of the WGA-amplified and -unamplified sample and the mean of both before and after WGA ratios.
Table 3 Percentage of differences between before/ after that fall outside/inside the acceptable boundaries of agreement on the Bland– Altman plot Sample pair Percent of data out Percent of data ID of agreement in agreement
3.8.3. Real-Time Relative Quantitation Assay
1
4.5
95.5
2
4.4
95.6
3
4.5
95.5
Dosage alterations at HER-2/neu (exon B) in 29 breast cancer cases that had been previously analyzed were investigated using Taqman customized primers (forward primer sequence: 5′-GCC AGG GTA TGT GGC ATC ATG-3′, reverse primer sequence: TTT ACT AAC CTG TGC CCT TGG, probe sequence: 6-FAMATG AGA TGA GCA GTG GCA). These breast cancer cases were previously tested for the presence of an amplification mutation (increased copy number) of HER-2/neu gene using fluorescent in situ hybridization (FISH) technique.
174
Bosso and Al-Mulla
WGA5-amplified DNA samples were tested using TaqMan® gene copy number assays with 7500 Fast Real-Time PCR System (20, 21) (Applied Biosystems, Foster city, CA, USA). This technique quantitates exon/intron dosage changes in tested samples comparative to exon/intron dosage of a reference sample and uses delta delta Ct (DDCt) mathematical model to calculate RQ (relative quantitation) values of the tested loci. Topo-A2 primer (forward sequence: AGT TAC CTG GGC TTT TCT CTT TTA, reverse sequence: AGG AGC CAC AGC TGA GTC AAA, probe sequence: 6-FAM-AGT GAT GAC TTC CAT ATG G TAMRA) was used as an endogenous control. Results showed significant differences in exon/Intron dosage after GenoemPlex® (Table 5).
Table 4 Comparison of calculated MAD (median absolute deviation) values of three DNA samples in the before and after WGA Sample MAD ID unamplified
MAD amplified
1
0.1616
0.1954
2
0.1474
0.19
3
0.15
0.17
Table 5 Comparison between data obtained on HER-2/neu gene copy number by FISH analysis and real-time PCR-relative quantitation assay done on breast cancer cases Serial
HER-2/neu gene status by FISH
Results obtained by real-time PCR on WGA
1
Amplified
Amplified
2
Amplified
Amplified
3
Amplified
Amplified
4
Amplified
Amplified
5
Amplified
Amplified
6
Normal
Normal
7
Normal
Normal (continued)
Whole Genome Amplification of DNA Extracted from FFPE Tissues
175
Table 5 (continued) Serial
HER-2/neu gene status by FISH
Results obtained by real-time PCR on WGA
8
Normal
Amplifieda
9
Normal
Amplifieda
10
Amplified
Amplified
11
Amplified
Normala
12
Amplified
Amplified
13
Amplified
Amplified
14
Amplified
Amplified
15
Amplified
Amplified
16
Normal
Normal
17
Normal
Normal
18
Normal
Normal
19
Amplified
Amplified
20
Amplified
Amplified
21
Normal
Amplifieda
22
Normal
Amplifieda
23
Normal
Normal
24
Amplified
Amplified
25
Amplified
Amplified
26
Amplified
Amplified
27
Amplified
Amplified
28
Amplified
Amplified
29
Normal
Normal
FISH fluorescence in situ hybridization, PCR polymerase chain reaction, WGA whole genome amplification a Indicates discordant results between the two methods
Real-time relative quantitation analysis results showed that the sensitivity in detecting cases with positive amplification mutation of HER-2/neu was 94.4%. The specificity was 63.63% in detecting cases with normal copy number of HER-2/neu gene. Positive predictive value and negative predictive value were 80.9 and 87.5%, respectively. The error rate was 17.5%, while the percentage of samples that maintained their HER-2/neu gene status after GenomePlex® WGA was 82.75%. Although the Kappa statistic (14, 15) value of
176
Bosso and Al-Mulla
Table 6 Summary of statistical analysis results of real-time allelic discrimination assay
RKIP–SNP ID
Gene/region
Percent Percent of concor- of disconcordant samples dant samples
dbsnp no.: rs2088702
RKIP/Intronic
92.8
dbsnp no.: rs904661
RKIP/Intronic
82.14
7.14 17.8
0.6133 reflects substantial agreement in data between GenomePlex® amplified and unamplified cases, the low specificity of 63.63% shows that an obvious number of samples that did not have the mutation were identified as positive for the mutation after WGA by real-time quantitative analysis. This indicates significant false-positive occurrences. 3.8.4. Real-Time Allelic Discrimination (SNP) Assay
3.8.5. Conclusions and Recommendations
Two intronic SNP markers targeting RKIP gene (Table .6) were tested in 30 colon cancer cases in the pre- and post-WGA forms using real-time PCR to investigate the fidelity of GenomePlex® to maintain allelic ratio. TaqMan® SNP Genotyping assays using 7500 Fast Real-Time PCR System (20) (Applied Biosystems, Inc., Foster city, CA, USA) were applied to investigate the ability of the GenomePlex® WGA technique considering the selected SNPs. The percentages of samples that gave concordant and disconcordant SNP allele calls after WGA compared to the allele calls before WGA are shown in Table 6. Statistically significant number of samples exhibited allele bias introduced after GenomePlex® WGA5 of DNA. 1. The GenomePlex® WGA5 kit proved to be able to amplify DNA extracted from FFPE tissue starting from minute amount (1 ml) of tissue lysate (consisting at least of 10 ng of DNA). 2. GenomePlex® is able to generate amplified DNA products sufficient in yield (4–8 mg) to perform several molecular procedures. 3. GenomePlex® provided superior genome coverage, preserved copy number status, and produced negligible representational bias evident by CGH and array-CGH analysis. 4. GenomePlex®-amplified DNA experienced at the exonic/ intronic level gene dosage bias shown by real-time quantitative PCR. It is worth noting that exon dosage results of HER-2/neu in the WGA-amplified cases were compared to
Whole Genome Amplification of DNA Extracted from FFPE Tissues
177
the FISH copy number results of the whole HER-2/neu gene in WGA-unamplified cases. It is possible that some of the detected bias is due to different techniques being compared, in which FISH technique that enables detecting copy number alteration over about 450,000 bp, while the real-time quantitative assay enables detection of copy number change within much smaller region in a gene (around 70 bp in this case). So, it is possible that the bias observed is not only entirely due to faulty WGA, but also due to variation between the detection perimeter of both techniques compared. 5. Significant number of partially degraded DNA samples exhibited allele bias after GenomePlex® WGA in regard to two targeted SNPs analyzed by Real-Time Allelic discrimination assay. 6. In conclusion, GenomePlex® WGA can be used to amplify partially degraded samples for genome-wide analysis including metaphase and array-based CGH. 7. Further investigation is needed to test the efficiency of GenomePlex® WGA technique in amplifying partially degraded DNA at the exonic and SNP level. 8. Our data suggest that GenomePlex® WGA technique is to be used with caution if used in analyzing specific small targeted areas (exons/introns) of the genome (20 bp or less) as with real-time or single-nucleotide polymorphisms (SNPs) assays. 9. The quality and degree of fragmentation of the FFPE sample greatly influences the performance of their amplified GenomePlex® products in downstream applications; it is therefore highly recommended that early considerations of proper tissue preservation and fixation take place.
4. Notes 1. The quality of DNA present in FFPE tissues is variable and compromised due to the earlier tissue fixation and preservation procedures as well as the aging of the tissue sample itself. The lower the quality of the DNA sample, the lower is its performance with downstream applications including WGA technique. Therefore, optimal initial handling of the tissue under controlled fixation conditions (fixative type, pH, concentration, fixation time, and temperature) is recommended to minimize tissue damage as far as possible. 2. If tumor tissue is to be amplified for further analysis, it is recommended that microdissection is performed and guided by an H&E-stained slide of the tissue to avoid including normal tissue from the same sample, which might obscure and compromise tumor genotyping results.
178
Bosso and Al-Mulla
3. It is recommended that the microdissected tissue be free from necrotic and fatty residues, which if present would further limit the amount of obtainable amplifiable DNA. 4. The amount of starting microdissected tissue to form tissue lysate could range from 0.1 to 1 mg. It is expected that microscopic tissue sections vary in their cross section area and cellular content, and thus the weight of microdissected tissue will vary accordingly. 5. WGA5 kit reagents are to be stored at −20°C. This is particularly important for the stability of the library preparation enzyme and the Taq polymerase enzyme provided with the kit. 6. Throughout the practical procedures, keep the WGA reagents when thawed on ice, especially the enzyme preparations. 7. Do not directly open reagents when thawed for use. Centrifuge reagent tubes briefly ahead to collect all solution at the bottom of the tubes. 8. The remaining tissue lysates obtained for each sample can be aliquoted and stored at −20°C as a reservoir for future needed WGA of the same sample. In our laboratories, tissue lysates were reused successfully to perform WGA after over 3 months of storage. This is understandable, since the microdissected tissue is being mildly treated and randomly fragmented by the CelLytic Y Lysis solution provided with the kit. 9. Negative and positive controls should always be included. Since this is a method that is highly sensitive to DNA contamination, inclusion of a negative control is a must to ensure contaminant free WGA amplified product. Otherwise, if there was any trace of DNA in the negative control apparent from gel electrophoresis imaging, the experiment is considered unsuccessful, even though target samples were amplified and displayed a white smear on gel. 10. Utterly important consideration when starting thermal cycling is to place the 0.2-ml PCR tubes in PCR system only when it has reached and held at the required temperature. Never place the sample tube in PCR system while its heating up, as precision in both time and temperature is critical for the success of WGA procedure. 11. When Library preparation step is completed, the samples can either be amplified in the final PCR step immediately or be stored at −20°C up to 3 days. 12. To avoid contamination of WGA experiment, preparation of experiment environment is highly recommended. Using fresh filter tips and aliquoted reagents, wearing facial masks to prevent contamination in case of sneezing or talking, and working in sterile tissue culture hood are all beneficial precautionary requirements.
Whole Genome Amplification of DNA Extracted from FFPE Tissues
179
13. To diminish any bias or errors that could be introduced during WGA of DNA in the final product, it is recommended to perform three WGAs for each target sample in three different tubes and then to pool the amplified products together after purification of the products separately. 14. In the PCR amplification step, 20 cycles are recommended for optimal amplification. If exceeded, the method could overrepresent nondominant DNA fragments. 15. The last purification step of DNA amplified products, the purified DNA can either be suspended in EB (provided with the kit) or dehydration solution or nuclease-free water. Depending on which molecular application you intend to use your purified WGA-amplified DNA with, select the appropriate suspension solution described for the required protocol. 16. Since the amplified DNA products are of partially degraded nature, it is recommended for long-term storage to keep the DNA products at −20°C in aliquots. Otherwise, for short-term analysis demands, amplified DNA samples can be stored at 4°C, to avoid frequent freezing and thawing, which worsens the quality and enhances further degradation of the samples.
Acknowledgments This work is supported by Kuwait Foundation for the Advancement of Sciences grant number 2006-1302-07, Kuwait University Grant numbers YM10/07 and MG02/08, and Research Core Facility (RCF) grant number GM 01/01 and GM 01/05. References 1. Hawkins, T., Detter, J., and Richardson, P. (2002) Whole genome amplification applications and advances. Current Opinion in Biotechnology 13, 65–67. 2. Hughes, S., Arneson, N., Done, S., and Squire, J. (2005) The use of whole genome amplification in the Study of human disease. Progress in Biophysics & Molecular Biology 88, 173–189. 3. Wells, D., Sherlick, J., Handyside, A., and Delhanty, J. (1999) Detailed chromosomal and molecular genetic analysis of single cells by whole genome amplification and comparative genomic hybridization. Nucleic Acids Research 27, 1214–1218. 4. Sermon, K., Lissens, W., Joris, H., Steirteghem, A., and Liebaers, I. (1996) Adaptation of the primer extension preamplification (PEP) reac-
tion for preimplantation diagnosis: single blastomere analysis using short PEP protocols. Molecular Human Reproduction 2, 209–212. 5. Kristjansson, K., Chong, S., Veyver, I., Subramanian, S., Snabes, M., and Hughes, M. (1994) Preimplantation single cell analyses of dystrophin gene deletions using whole genome amplification. Nature Genetics 6, 19–23. 6. Tanabe, C., Aoyagi, K., Sakiyama, T., Kohno, T., Yanagitani, N., Akimoto, S., et al. (2003) Evaluation of a whole-genome amplification method based on adaptor-ligation PCR of randomly sheared genomic DNA. Genes, Chromosomes & Cancer 38, 168–176. 7. Mueller, E. (2007) Genomic Analysis of Formalin-Fixed Paraffin Embedded (FFPE)
180
8. 9.
10. 11. 12.
13.
14.
Bosso and Al-Mulla Tissues Through the Use of Whole Genome Amplification (WGA). http://www.sigmaaldrich.com/etc/medialib/docs/Sigma/ General_Infor mation/f fpewhitepaper. Par.0001.File.tmp/ffpewhitepaper.pdf. WGA5 Bulletin. http://www.sigmaaldrich. com/etc/medialib/docs/Sigma/Bulletin/ wga5bul.Par.0001.File.tmp/wga5bul.pdf. Whole Genome Amplification Advisor. http://www.sigmaaldrich.com/life-science/ molecular-biology/whole-genome-amplification/promo/wga-advisor-download.html. http://www.sigmaaldrich.com. http://www.biocompare.com/console/ sigma/genomeplex_video.asp. Weiss, M., Hermsen, M., Meijer, G., van Grieken, N., Baak, J., Kuipers, E., and Van Diest, P. (1999) Comparative genomic hybridization. Journal of Clinical Pathology 52, 243–251. Ng, G., Roberts, I., and Coleman, N. (2005) Evaluation of 3 methods of whole-genome amplification for subsequent metaphase comparative genomic hybridization. Diagnostic Molecular Pathology 14, 203–212. Viera, A. and Garrett, J. (2005) Understanding interobserver agreement: the kappa statistic. Family Medicine 37, 360–363.
15. Chan, Y. (2003) Biostatistics 104: correlational analysis. Singapore Medical Journal 44, 614–619. 16. Snijders, A.M., Meijer, G.A., Brakenhoff, R.H., van den Brule, A.J.C., and van Diest, P.J. (2000) Microarray techniques in pathology: tool or toy? Journal of Clinical Pathology 53, 289–294. 17. Shaffer, G.L. and Bejjani, A.B. (2004) A cytogeneticist’s perspective on genomic microarrays. Human Reproduction Update 10, 221–226. 18. Myles, P. and Cui, J. (2007) Using the Bland– Altman method to measure agreement with repeated measures. British Journal of Anaesthesia 99, 309–311. 19. Kirkwood, B. and Sterne, J. (2003) Essential medical statistics, 2nd edition, Blackwell Publishing Ltd, Oxford, UK, Chapter 36, 440–442. 20. Boulay, J.L., Reuter, J., Ritschard, R., Terracciano, L., Herrmann, R., and Rochlitz, C. (1999) Gene dosage by quantitative realtime PCR. BioTechniques 27, 228–232. 21. Allelic Discrimination Getting Started Guide for 7300/7500/7500 Fast Systems. © Copyright 2006, Applied Biosystems. All rights reserved.
Chapter 12 Pyrosequencing of DNA Extracted from Formalin-Fixed Paraffin-Embedded Tissue Brendan Doyle, Ciarán O’Riain, and Kim Appleton Abstract Gene promoter hypermethylation is recognised as an important mechanism by which genes may be silenced both physiologically and in disease states. This mechanism of gene silencing has been shown to play a role in many common human tumours. A number of methods are available for the detection of promoter hypermethylation, including the methylation-specific polymerase chain reaction (PCR), bisulphite sequencing, and pyrosequencing. Pyrosequencing is a reproducible method for obtaining data on the methylation status of DNA. It also has the advantage of providing quantitative data regarding the amount of methylation present in multiple CpGs in a given sample. The technique is based on the bisulphite conversion of unmethylated cytosine to uracil and subsequent amplification by PCR. The technique is also appropriate for use on DNA extracted from formalin-fixed paraffin-embedded tissue. Key words: Pyrosequencing, Methylation, Bisulphite modification
1. Introduction DNA methylation involves the addition of a methyl group to the hydrogen bond in the 5¢ position of cytosine to form methylcytosine. The reaction is catalysed by a family of enzymes known as DNA methyltransferases using S-adenosyl-methionine as a donor. Typically, this occurs within CpG dinucleotides in differentiated cells, but may occur at other sites in embryonic stem cells (1). In general, CpG dinucleotides are underrepresented in the genome as a whole, but are common in the promoter regions of many genes. CpG-rich regions are termed CpG islands and methylation of these has been associated with silencing of the corresponding genes. The accurate and quantitative measurement of the DNA methylation state has become important owing to the role played
Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_12, © Springer Science+Business Media, LLC 2011
181
182
Doyle, O’Riain, and Appleton
by DNA methylation both in physiology and in disease. Gene silencing by promoter methylation plays an important role in a number of physiological processes, including silencing of the second X-chromosome in females and in imprinting. Consequently, methylation abnormalities have been implicated in imprinting disorders, such as Prader–Willi syndrome, Angelman syndrome (2), and Beckwith–Weidemann syndrome (3). DNA promoter methylation has been implicated in other disease states, particularly in cancer. In breast cancer, DNA hypermethylation has been reported in important tumour suppressor genes such as RASSF-1; moreover, changes have been detected in preneoplastic lesions, suggesting that it may be an early event in tumourogenesis (4). In colorectal carcinoma (CRC), a new subgroup of tumours has been described in which widespread hypermethylation occurs (5). In this group, which has been termed the CpG island methylator phenotype (CIMP), promoter hypermethylation has been shown to lead to the silencing of tumour suppressor genes, including MLH1 (6). Promoter methylation has been shown to affect a range of haematological malignancies (7), with the germinal centre-related lymphomas such as follicular lymphoma most susceptible, and it has been suggested that the hypermethylation is an early event in lymphomagenesis (8). A number of techniques exist for the detection of promoter methylation, including methylation-specific polymerase chain reaction (MSP), bisulphite sequencing, and pyrosequencing. Pyrosequencing is a method of analysis which is both reproducible and provides quantitative data. The workflow involves the initial bisulphite conversion of DNA. During this process, unmethylated cytosine is converted to uracil, while methylated cytosine is resistant to the conversion. Following conversion the DNA is amplified by PCR, before being analysed by the pyrosequencer. When a nucleotide base is incorporated into an elongating oligonucleotide, a pyrophosphate molecule (PPi) is released. The pyrosequencing assay makes use of this by converting the released PPi, in a quantitative fashion, to a light signal which can be detected by a camera. The PPi molecule is converted to ATP by the enzyme ATP sulphurylase. ATP then provides the energy for the enzyme luciferase to convert luciferin to oxyluciferin. This reaction releases one photon of light which can be detected. It is this which gives pyrosequencing its quantitative power, as the incorporation of each base is directly proportional to the light emitted. Nucleotides are added sequentially to the reaction mixture in a predetermined sequence and if they are incorporated into the DNA oligonucleotide, this is detected as detailed above. The technology can be used in a high-throughput environment analysing the same DNA sequence in up to 96 samples simultaneously. Alternatively, multiple different analyses may be performed on the samples in a 96-well plate.
Pyrosequencing of DNA Extracted from Formalin-Fixed Paraffin-Embedded Tissue
183
A number of groups have shown that pyrosequencing can be successfully applied to DNA extracted from formalin-fixed paraffinembedded tissue (9–13), and it has been shown to compare favourably with other methods of methylation detection in this setting (12).
2. Materials 2.1. Bisulphite Modification
1. Bisulphite modification kit (e.g., Epitect Bisulfite, Qiagen). The kit involves the following steps: ●●
●●
Bisulphite modification of the DNA. Binding of the modified, single-stranded DNA to the membrane of a spin column.
●●
Desulphonation of the DNA.
●●
Washing to remove the desulphonation agent.
●●
Elution of the modified DNA.
2. Thermocycler. 3. Bench top centrifuge. 2.2. Polymerase Chain Reaction
1. PyroMark Assay Design Software (Qiagen). 2. Thermocycler. 3. PCR primers at 10 mM one of which must be biotinylated (Invitrogen). 4. Taq DNA polymerase (e.g., FastStart Taq, Roche). 5. dNTPs. 6. Thermofast semi-skirted 96-well plate (Thermo Scientific). 7. Microseal A film (Bio Rad).
2.3. Preparation of Template for Sequencing
1. Pyrosequencing vacuum prep workstation (Qiagen). 2. Vacuum source. 3. Orbital shaker. 4. PyroMark 96 Plate Low (Qiagen). 5. Pyrosequencing Thermoplate (Qiagen). 6. Heating block. 7. Troughs (Qiagen). 8. Streptavidin sepharose HP beads (GE Healthcare). 9. 1× Annealing buffer (Qiagen). 10. 1× Binding buffer (Qiagen). 11. Denaturation solution (0.2 M NaOH).
184
Doyle, O’Riain, and Appleton
12. 10× Washing buffer (10 mM Tris–acetate, Qiagen). 13. 70% Ethanol. 14. High purity water (Milli-Q water or equivalent). 15. Sequencing primer at 10 mM. 2.4. Pyrosequencing Analysis
1. PyroMark Q 96 Reagent Cartridge (Qiagen). 2. Pyro Q CpG (Pyrosequencing analysis) software (Qiagen). 3. PyroMark Gold Q 96 Reagent Kit (Qiagen).
3. Methods Obtaining quantitative methylation values using pyrosequencing from DNA is outlined here in the following five steps: 1. Bisulphite modification of the genomic DNA 2. Pyrosequencing primer design 3. PCR 4. Preparation and purification of the PCR product 5. Running the pyrosequencing analysis. 3.1. Bisulphite Modification of Genomic DNA
The process of bisulphite modification of the DNA allows a sequence difference to be created between methylated and unmethylated DNA facilitating subsequent examination of methylation status. The reaction results in the conversion of unmethylated cytosine (C) to uracil (U). However, methylated Cs are resistant to the conversion and remain unchanged. The subsequent PCR reaction matches thymine (T) to the converted U. Therefore, in the PCR product, the methylation status of a given CpG dinucleotide may be determined by assessing the ratio of C (methylated CpG) to T (unmethylated CpG). Complete conversion can be assessed by amplifying a sequence of DNA containing multiple non-CpG Cs by PCR, using primers complementary to both the modified and the unmodified sequence (see Note 1). 1. A number of commercial kits are available for the process of bisulphite modification. We have successfully used the Epitect Bisulphite Kit (Qiagen). The kit is convenient and effective using a starting amount of between 1 ng and 2 mg of DNA and can be used on DNA extracted from formalin-fixed paraffin-embedded tissue. 2. Modified DNA may be stored at 4°C for a period of around 24 h or for up to 6 months at −20°C.
Pyrosequencing of DNA Extracted from Formalin-Fixed Paraffin-Embedded Tissue
3.2. Primer Design
185
Three primers are required for the process. Forward and reverse primers are required for the initial PCR step and a sequencing primer is needed for the pyrosequencing analysis. One of the PCR primers must be biotinylated to allow binding of PCR product to streptavidin-coated beads in subsequent steps. The sequencing primer will be complementary to the sequence incorporating the biotinylated primer. Although the PCR primers and sequencing primers may be designed separately, Pyrosequencing Assay design software is available from Qiagen, which allows for the design of both PCR and sequencing primers that can be used in the analysis. 1. The bisulphite-modified sequence of interest should be pasted into the sequence input screen. The CpG islands of interest should be entered using the IUPAC codes (i.e., YG to indicate a C/TG if analysing the forward sequence and GR to indicate a CG/A if analysing the reverse sequence). 2. The software will generate forward, reverse, and sequencing primers as well as a report and a score indicating potential problems with the primers. As the DNA strands are no longer complementary following bisulphite conversion, both the forward and the reverse sequences can be used for analysis. The direction of the sequencing primer will determine which of the primers needs to be biotinylated. 3. While the software will generate primers and an associated score with a list of potential problems, a number of factors in primer design should be kept in mind. ●●
●●
●●
●●
●●
The amplification product should be less than 350 bp. Longer-amplification products may lead to the formation of secondary structures and decrease the efficiency of binding to the streptavidin beads (14). Non-CpG Cs should be included in the “sequence to analyse” to act as a control for complete bisulphite conversion (see Note 1). Complementary sequences within or between primers should be avoided to prevent secondary structure or dimer formation. Primers should ideally not be complementary to CpGs as this could lead to preferential amplification of either methylated or unmethylated DNA (see Note 2). The sequencing primer should only anneal to one sequence within the sequence to be analysed to avoid mispriming.
4. The primers and sequence to analyse information generated using this software can now be saved and then be imported directly into the pyrosequencing analysis software for the analysis step.
186
Doyle, O’Riain, and Appleton
3.3. Polymerase Chain Reaction
The PCR step is required to amplify the sequence of interest for analysis. It should be optimised to provide a single, reproducible, clear, and strong band when 5 ml of the product is loaded on an agarose gel. The PCR reaction mixture should be made up to a final volume of 50 ml. 1. We have used FastStart Taq (Roche) but other authors have reported good results using other Taq enzymes, including Platinum Taq (Invitrogen) and StarTaq (Qiagen). 2. 1 ml of a 10 mM solution of each of the PCR primers should be used. It is important not to use an excess of primers as residual biotinylated primer may impede the binding of the amplification product to the streptavidin beads in the sample purification step. A high number of cycles (40–50) are used to exhaust any excess biotinylated primer. 3. The starting amount of template DNA should be above 10 ng as the potential for nonspecific amplification increases with decreasing concentrations of template DNA (15). 4. Due to the high number of cycles which are necessary to exhaust the biotinylated primers, it is very important to run negative controls (water in place of template DNA) to ensure that there is no contamination and subsequent nonspecific amplification in the reaction mixture. 5. At least in initial experiments it is important to run controls of known methylation status. This is to ensure that there is no preferential amplification of either methylated or unmethylated DNA as it has been shown that this can occur as a result of the base changes induced by the bisulphite modification (16). Unmethylated and 100% in vitro methylated DNA is available commercially and we run controls consisting of 100, 75, 50, 25, and 0% in vitro methylated DNA, respectively (see Note 2).
3.4. Sample Preparation and Purification
In order for the analysis to be carried out, the amplicon must be single stranded and any residual dNTPs should be removed from the reaction mixture. This procedure may be carried out using the pyrosequencing vacuum prep tool and workstation. The biotinylated amplicons adhere to streptavidin beads and are maintained on the membrane, while contaminants are washed through. 1. Set heating block to 80°C and place the pyrosequencing thermoplate on it to heat. 2. To each well of a 96-well Thermofast PCR plate add: 40 ml of the PCR product, 3 ml of the streptavidin sepharose beads, and 37 ml of binding buffer and seal the plate (the beads and Binding buffer can be mixed prior to the addition of the PCR product).
Pyrosequencing of DNA Extracted from Formalin-Fixed Paraffin-Embedded Tissue
187
3. Prepare a master mix consisting of 1.5 ml of the sequencing primer (10 mM solution) and 43.5 ml of annealing buffer for each sample. Add 45 ml of this mix to each well in the PyroMark 96 Low Plate. 4. The PyroMark 96 Low Plate can now be placed in the PyroMark plate slot on the vacuum prep workstation. 5. Fill each of the four troughs in the following order: 70% ethanol, denaturing solution (0.2 M NaOH), wash buffer, and high-purity water. The troughs should be filled to the mark with the exception of the denaturing solution which should be filled to slightly below the mark. This is to ensure that in the subsequent washing step no residual NaOH is left on the vacuum prep tool. Fill the trough at the parking position of the vacuum prep workstation with high-purity water. 6. Vortex the PCR plate for 5 min. The speed should be high enough to agitate the liquid and ensure that the beads are resuspended but not so high that there is excessive spray onto the seal. This may lead to contamination upon opening the seal (this step may be performed while preparing the PyroMark 96 Low plate and the vacuum prep workstation). 7. Place the plate in its slot in the vacuum prep workstation and carefully remove the seal. 8. Start the vacuum pump and leave the vacuum prep tool in the parked position until approximately half of the water has been aspirated. 9. Place the vacuum tool into the PCR plate, ensuring that it reaches the bottom of the wells. Allow the tool to aspirate for 5 s. Before moving on ensure that all of the liquid has been aspirated. 10. Move the vacuum tool to the 70% ethanol and aspirate for 5 s. 11. Move the vacuum tool to the denaturing solution and aspirate for 5 s. 12. Move the vacuum tool to the wash buffer and aspirate for 5 s. 13. Remove the vacuum tool from the wash buffer, turn off the vacuum, and invert the tool. 14. Place the vacuum tool into the PyroMark 96 Low plate and shake vigorously to dislodge the beads. 15. Replace the vacuum tool to the parking slot and transfer the PyroMark 96 Low plate to the heating block at 80°C for 2–3 min. 16. Transfer the PyroMark 96 Low plate to the pyrosequencer. 17. The vacuum prep tool should be washed through with water and should then be allowed to aspirate air to allow it to dry.
188
Doyle, O’Riain, and Appleton
3.5. Pyrosequencing Analysis
1. Open the Pyro Q-CpG™ software. 2. From the “File” menu, select “Import” and “Assay Design File.” 3. Select the Assay Design File saved earlier. 4. The sequence to analyse will appear in the window, beside this click on the “Generate Dispensation Order” button and the software will generate a list of the nucleotides that will be dispensed by the machine. 5. From the “Tools” menu, select “Volume Information,” a new window will appear giving the appropriate volumes of reagents (substrate, enzymes, and dNTPs) to be loaded into the PyroMark 96 Reagent Cartridge. 6. Allow the reagents (substrate, enzymes, and dNTPs) to come to room temperature and make up the enzyme and substrate with the appropriate volume of high-purity water (prior to reconstitution the enzyme, substrate, and nucleotides are stored at 4°C, after reconstitution the enzyme and substrate solutions are stable at 4°C for up to 5 days. For long-storage periods, they may be stored at −20°C, no more than two freeze/thaw cycles is recommended). 7. Load the reagents into the appropriate segments of the PyroMark 96 Reagent Cartridge (Fig. 1) (see Note 3). 8. Insert the PyroMark 96 Cartridge into the pyrosequencer and click “Run” on the computer. 9. When the run is complete, the individual wells can be analysed using the software. The software will analyse each individual well and assign it a colour, which flags up potential problems (i.e., blue indicates no problem, yellow a minor problem, and red a more serious problem). Any potential problems in the pyrograms which have been flagged by the software should be checked manually.
Fig. 1. Diagram indicating how the enzyme, substrate, and nucleotides should be loaded into the reagent cartridge (C cytosine, G guanine, T thymine, A adenosine, S substrate, and E enzyme).
Pyrosequencing of DNA Extracted from Formalin-Fixed Paraffin-Embedded Tissue
189
4. Notes 1. A potential confounding factor in the analysis is incomplete bisulphite modification of the DNA. Incomplete bisulphite modification can lead to an overestimate of DNA methylation, as unmodified Cs will be misinterpreted as methylated. A number of controls should be included to overcome this and to check for complete conversion. Firstly, a section of the converted DNA containing multiple non-CpG Cs should be amplified using a primer sets complimentary to both converted and unconverted DNA. Complete conversion of the DNA should result in amplification only using the primers complimentary to the converted DNA. Secondly, a number of non-CpG Cs should be included in the sequence to be analysed. If the conversion reaction has resulted in complete conversion, then all of these Cs should have been converted to Ts and this should be seen on the resulting pyrogram. 2. It has been shown that the base changes brought about by the bisulphite conversion process can lead to preferential amplification of either methylated or unmethylated DNA (16, 17), this can lead to an over- or underestimation of DNA methylation in the final analysis. It is important, therefore, to run appropriate controls to ensure equal amplification of both methylated and unmethylated DNA. Serial dilutions of in vitro methylated DNA should be prepared (e.g., 100, 75, 50, 25, and 0%). These samples should be amplified by PCR and 5 ml of the reaction mixture run on an agarose gel, to ensure that bands are of approximately equal intensities. Following this the samples should be analysed using the pyrosequencing protocol to ensure that the methylation levels measured for the samples are that expected from the dilutions prepared. We and others have B6 : T AT AACT ACCRAATTCTCRAAAACAACRACTCRACCTACAAAACTATTATTTCAAAAA 9% 5% 9% 6% 60 50 40 30 20 10 0 ES C T A T A C T A T CA GA T C T AC AA GA A CA TC AG AC T AC AG AC TA 5
10
15
20
25
30
35
40
Fig. 2. Example of a pyrogram, showing low levels of methylation. Four CpG dinucleotides have been analysed. In this case, the lower DNA sequence has been analysed and thus the amount of methylation is measured by the ratio of A (unmethylated) to G (methylated). In each of the four-shaded CpGs, there is an “A” peak but with little or no “G” peak, indicating low levels of methylation.
190
Doyle, O’Riain, and Appleton
found that careful optimisation of the annealing temperature of the PCR reaction allows for unbiased amplification of both methylated and unmethylated DNA (17). 3. The PyroMark 96 Reagent Cartridge should be washed carefully after use with high-purity water, ensuring that the stream of water coming from the dispensing needle is straight. The cartridge should last approximately 20 runs with appropriate care. References 1. Lister, R., Pelizzola, M., Dowen, R. H., Hawkins, R. D., Hon, G., Tonti-Filippini, J., et al. (2009) Human DNA methylomes at base resolution show widespread epigenomic differences. Nature 462, 315–22. 2. White, H. E., Hall, V. J., and Cross, N. C. (2007) Methylation-sensitive high-resolution melting-curve analysis of the SNRPN gene as a diagnostic screen for Prader–Willi and Angelman syndromes. Clin Chem. 53, 1960–2. 3. Bliek, J., Verde, G., Callaway, J., Maas, S. M., De Crescenzo, A., Sparago, A., et al. (2009) Hypomethylation at multiple maternally methylated imprinted regions including PLAGL1 and GNAS loci in Beckwith– Weidemann syndrome. Eur J Hum Genet. 17, 611–9. 4. Zhu, W., Qin, W., Hewett, J. E., and Sauter, E. R. (2010) Quantitative evaluation of DNA hypermethylation in malignant and benign breast tissue and fluids. Int J Cancer 126, 474–82. 5. Toyota, M., Ahuja, N., Ohe-Toyota, M., Herman, J. G., Baylin, S. B., and Issa, J. P. (1999) CpG island methylator phenotype in colorectal cancer. Proc Natl Acad Sci USA 96, 8681–6. 6. Issa, J. P. (2008) Colon cancer: it’s CIN or CIMP. Clin Cancer Res. 14, 5939–40. 7. Martin-Subero, J. I., Ammerpohl, O., Bibikova, M., Wickham-Garcia, E., Agirre, X., Alvarez, S., et al. (2009) A comprehensive microarray-based DNA methylation study of 367 hematological neoplasms. PLoS One 4, e6986. 8. O’Riain, C., O’Shea, D. M., Yang, Y., Le Dieu, R., Gribben, J. G., Summers, K., et al. (2009) Array-based DNA methylation profiling in follicular lymphoma. Leukemia 23, 1858–66. 9. Buckingham, L., Faber, L. P., Kim, A., Liptay, M., Barger, C., Basu, S., et al. (2010) PTEN, RASSF1 and DAPK site-specific hypermethylation and
10.
11.
12.
13.
14. 15.
16.
17.
outcome in surgically treated stage I and II non small cell lung cancer patients. Int J Cancer 126, 1630–39. Dufort, S., Richard, M. J., and de Fraipont, F. (2009) Pyrosequencing method to detect KRAS mutation in formalin-fixed and paraffinembedded tumor tissues. Anal Biochem. 391, 166–8. Kure, S., Nosho, K., Baba, Y., Irahara, N., Shima, K., Ng, K., et al. (2009) Vitamin D receptor expression is associated with PIK3CA and KRAS mutations in colorectal cancer. Cancer Epidemiol Biomarkers Prev. 18, 2765–72. Mikeska, T., Bock, C., El-Maarri, O., Hubner, A., Ehrentraut, D., Schramm, J., et al. (2007) Optimization of quantitative MGMT promoter methylation analysis using pyrosequencing and combined bisulfite restriction analysis. J Mol Diagn. 9, 368–81. Nosho, K., Shima, K., Irahara, N., Kure, S., Baba, Y., Kirkner, G. J., et al. (2009) DNMT3B expression might contribute to CpG island methylator phenotype in colorectal cancer. Clin Cancer Res. 15, 3663–71. Tost, J., and Gut, I. G. (2007) DNA methylation analysis by pyrosequencing. Nat Protoc. 2, 2265–75. Dupont, J. M., Tost, J., Jammes, H., and Gut, I. G. (2004) De novo quantitative bisulfite sequencing using the pyrosequencing technology. Anal Biochem. 333, 119–27. Warnecke, P. M., Stirzaker, C., Melki, J. R., Millar, D. S., Paul, C. L., and Clark, S. J. (1997) Detection and measurement of PCR bias in quantitative methylation analysis of bisulphite-treated DNA. Nucleic Acids Res. 25, 4422–6. Shen, L., Guo, Y., Chen, X., Ahmed, S., and Issa, J. P. (2007) Optimizing annealing temperature overcomes bias in bisulfite PCR methylation analysis. Biotechniques 42, 48, 50, 52 passim.
Chapter 13 Analysis of DNA Methylation in FFPE Tissues Using the MethyLight Technology Ashraf Dallol, Waleed Al-Ali, Amina Al-Shaibani, and Fahd Al-Mulla Abstract Novel biomarkers are sought after by mining DNA extracted from formalin-fixed, paraffin-embedded (FFPE) tissues. Such tissues offer the great advantage of often having complete clinical data (including survival), as well as the tissues are amenable for laser microdissection targeting specific tissue areas. Downstream analysis of such DNA includes mutational screens and methylation profiling. Screening for mutations by sequencing requires a significant amount of DNA for PCR and cycle sequencing. This is self-inhibitory if the gene screened has a large number of exons. Profiling DNA methylation using the MethyLight technology circumvents this problem and allows for the mining of several biomarkers from DNA extracted from a single microscope slide of the tissue of interest. We describe in this chapter a detailed protocol for MethyLight and its use in the determination of CpG Island Methylator Phenotype status in FFPE colorectal cancer samples. Key words: PCR, MethyLight, FFPE, DNA, Methylation, CIMP, PMR
1. Introduction Profiling of DNA methylation in clinical samples is fast becoming a way to discover novel biomarkers for any given disease. Information obtained from DNA methylation analysis can complement gene expression analysis especially where RNA cannot be retrieved in sufficient quantity or quality. This is indeed the case with formalin-fixed, paraffin-embedded (FFPE) tissues. Such tissues are currently successfully used in tissue microarray analysis. However, their use in gene expression microarrays has been severely hindered by the poor quality of RNA retrieved by traditional methods (1). Since DNA methylation at a given loci more often than not reflects silencing of the associated gene, DNA methylation analysis can potentially reflect expression analysis. DNA from FFPE tissues is an extremely valuable resource for Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_13, © Springer Science+Business Media, LLC 2011
191
192
Dallol et al.
cancer studies as Laser microdissection from FFPE tissue followed by DNA extraction and analysis is now becoming a standard tool for scientists to understand the biological processes involved in carcinogenesis (2, 3). Analysis of DNA methylation is centered on the detection of 5-methyl-cytosine residue in the context of a CpG dinucleotide concentrated in an island, CpG island. Detection of methylation events benefited enormously from the development of the bisulfiteDNA treatment (4). This treatment leads to the conversion of all unmethylated cytosines into uracil/thymidine residues leaving the methylated cytosine residues intact. PCR-based techniques can then be used to differentiate the two species. Bisulfite-DNA sequencing utilizes the methylation-independent PCR amplification of a region of interest followed by sequencing. However, this method suffers from undesirable background interference caused by unincorporated labeled nucleotides, and therefore PCR cloning is often required to resolve the methylated alleles. Combined Bisulfite and Restriction Analysis (COBRA) is a slight variation on the sequencing method where the methylated alleles are identified by restriction analysis with enzymes that digest the sequences such as CG^CG (BstUI) or T^CGA (TaqI) (5). Despite being informative and reliable techniques, their use is limited in DNA extracted from FFPE tissues. In such samples, the PCR is limited to amplifying short fragments, and nested or seminested PCR is often required to obtain a product with sufficient quantities to sequence and perform restriction digest on. An alternative approach to COBRA and bisulfite-DNA sequencing is the use of MethylationSpecific PCR (6). In this PCR, the primers are designed to anneal to a specific stretch of DNA sequence that contains presumably methylated CpG dinucleotides. Successful amplification reaction would indicate a positive methylation result while failure reflects lack of methylation. To control for DNA quality, quantity, and conversion efficiency, a control primer is also made to amplify the unmethylated variant. This approach has been the most commonly used technique in analyzing methylation in DNA extracted from FFPE tissues. However, the inherent problem of lack of quantification as well as the high rate of false-positive results hinders the development of biomarker sets based from methylation studies on archival material. A major advance to the MSP techniques is offered in the form of real-time detection of methylation using qPCR technology. One such protocol that uses this approach is MethyLight (7). This technique uses the TaqMan approach to amplify and quantify methylation levels in any sample. In MethyLight, two methylation-specific primers are used to amplify a region of interest from bisulfite-converted DNA. The primers span a region targeted with a fluorescence-labeled oligonucleotide probe targeting the methylated CpG within that sequence. This probe is synthesized with a 5¢-fluorochrome and a 3¢ quencher. The 5¢–3¢ exonuclease activity of the Taq polymerase would cleave
Analysis of DNA Methylation in FFPE Tissues Using the MethyLight Technology
193
the probe, releasing the fluorochrome from the proximity of the quencher moiety and freeing to emit light upon excitation. MethyLight is highly specific, offers a high degree of sensitivity and reproducibility. Furthermore, it is the technique of choice when faced with very small amounts of template DNA (as in the case of DNA extracted from laser-microdissected FFPE tissues) or with the need to perform rapid analysis of multiple gene loci as the hunt for cancer-specific informative biomarkers require. MethyLight is utilized in our research into colorectal cancer to identify the tumor-subtypes that exhibits the phenomenon of CpG Island Methylator Phenotype (CIMP) (8). Tumors are classified as CIMP-negative if all of the evaluated genes and loci are unmethylated, CIMP-low if one to two of the evaluated sites are methylated, and CIMP-high if three or more of the evaluated sites are methylated. In order to classify our cohort of colorectal cancers into CIMP-high, low, or negative, we use MethyLight to scrutinize the methylation status of a panel of seven genes; the methylation of which were shown to correlate tightly with the CIMP status of the tumor studied (IGF2, CDKN2A, SOCS1, CRAPB1, NEUROG1, CACNA1G, and RUNX3) (9–11). The CIMP-high colorectal cancers have specific clinical, pathologic, and molecular profile. For example, they are associated with proximal tumor location, female sex, mucinous and poor tumor differentiation, microsatellite instability, and high BRAF and low TP53 mutation rates (11). Therefore, cancers with high degrees of methylation represent a clinically and etiologically distinct group that is characterized by epigenetic instability. Furthermore, CIMP-associated cancers seem to have distinct molecular features (8). Therefore, DNA methylation is proving to be a useful marker of disease risk, activity, and prognosis in various malignancies (12).
2. Materials 2.1. M.SssI Modification
The MethyLight assay relies heavily on the use of in vitro methylated DNA control. In order to achieve this, a CpG methyltransferase, M.SssI, is used to methylate all the cytosine residues in the genomic DNA. The reagents for this procedure include the following: 1. M.SssI CpG Methyltransferase (New England Biolabs, MA, USA). The enzyme comes with 10× NEBuffer 2 and S-adenosylmethionine (SAM) that acts as the methyl donor for the reaction. 2. Genomic DNA (50–100 mg). Any genomic sample can be methylated in this reaction. This can include genomic DNA isolated from peripheral blood leukocytes (or human cell line DNA, see Note 1).
194
Dallol et al.
2.2. Bisulfite-DNA Conversion
1. 3 M Sodium hydroxide (NaOH), made fresh just prior to modification. This can be either diluted in double-distilled, nuclease-free water from 10 M stock or made directly from reagent (Sigma–Aldrich, MO, USA). 2. 1 M Hydroquinone (Sigma–Aldrich, MO, USA), freshly prepared. Once dissolved, the solution should be wrapped in foil (see Note 2). 3. 3.12 M Sodium bisulfate solution, at pH 5.0: this solution should always be prepared fresh. To make 10 ml of this solution, mix 5.93 g of sodium metabisulfite (Sigma–Aldrich, MO, USA, see Note 3) and 8 ml of boiled nuclease-free water. To achieve the desired pH, add 1 ml of 3 M NaOH and continue vortexing until the solution is clear and the Sodium Metabisulfite reagent has completely dissolved. Add 50 ml of 1 M hydroquinone that will work as a stabilizer and top up the solution to 10 ml. 4. Wizard® DNA cleanup kit from Promega (WI, USA) for the removal of the bisulfite reagent (see Note 4). 5. A thermal cycler that can take 0.5-ml Eppendorf tubes. 6. PCR-grade mineral oil (see Note 5).
2.3. M ethyLight Assay
1. TaqMan Probes and amplification primers for target of choice (see Note 6). 2. In vitro methylated and bisulfite-converted genomic DNA to work as positive control and generate a standard curve to calculate the PMR (Percentage of Methylated Ratio) of the test sample. Dilute this DNA to 10 ng/ml. As a negative control, unmethylated and bisulfite-converted genomic DNA should be used especially when new primer–probe combinations are made (see Note 7). Both DNA sets can be purchased from Qiagen as part of a PCR control DNA set (Qiagen, CA, USA). 3. TaqMan reagents: TaqMan® Fast Universal PCR Master Mix (2×), No AmpErase® UNG (Applied Biosystems, CA, USA, see Note 8). 4. MicroAmp® Fast 96-Well Reaction Plate, 0.1 ml, or the MicroAmp® Fast Optical 96-Well Reaction Plate with Barcode, 0.1 ml, with the MicroAmp® Optical Adhesive Film (Applied Biosystems, CA, USA). 5. Real-time PCR system: Applied Biosystems; 7500 Fast RealTime PCR System (Applied Biosystems, CA, USA) or the Applied Biosystem StepOnePlus systems or any other system that is compatible with the dyes and plates used.
Analysis of DNA Methylation in FFPE Tissues Using the MethyLight Technology
195
3. Methods 3.1. M.SssI CpG Methyltransferase Modification
1. In order to modify 50 mg of genomic DNA in a total volume of 1,000 ml, add the following components in order: 795 ml of Nuclease-free water, 100 ml of 10× NEBuffer 2, 5 ml of 32 mM SAM, 50 ml of 1 mg/ml Genomic DNA, 50 ml of 4 U/ml M.SssI Methyltransferase. Mix the reaction by pipetting several times (see Note 9). 2. Incubate the reaction overnight at 37°C to give the reaction better chance to reach completeness. 3. The next morning, add 2.5 ml of 32 mM SAM and 5 ml of 4 U/ml M.SssI Methyltransferase and incubate overnight again. 4. Stop the reaction by heating at 65°C for 20 min and purify the DNA by using phenol–chloroform extraction or the Promega Wizard DNA cleanup system as recommended by the manufacturer’s protocol. 5. Proceed to bisulfite modification (see Note 10).
3.2. Sodium Bisulfite Modification
1. Denaturation: denature the genomic DNA by boiling for 10 min at 100°C and incubating in 0.3 M NaOH for 15 min at 37°C. Set up this reaction in 10-ml final volume. A typical amount of template DNA used is 500–1,000 ng. However, this may not always be possible with FFPE tissues. Therefore, 100–200 ng of genomic DNA extracted from FFPE tissues can also be used (see Note 11). 2. Sulfonation: add 500 ml of the freshly prepared sodium bisulfite solution to the denatured DNA. Overlay the reaction with a drop of mineral oil and incubate the reaction in a thermocycler for 20 cycles of 30 s at 99°C/15 min at 50°C (see Note 12). 3. Cleanup: following sulfonation, cleanup the samples by using the Wizard DNA cleanup modules according to the manufacturer’s protocol (see Note 13). 4. Desulfonation: incubate the eluted DNA from the step above with 0.3 M NaOH for 10 min at room temperature. 5. DNA precipitation: precipitate the DNA using 2.5 volume of absolute ethanol and one-tenth volume of 3 M sodium acetate pH 5.2. This mix is incubated at −20°C. When FFPE DNA was used, we routinely left the precipitation reaction at −20°C overnight to maximize the yield of recovery and added up to 40 mg as carrier DNA. However, if the genomic DNA was obtained from fresh tissues or cell lines, a shorter incubation time is used. Following centrifugation and cleanup with 70% ethanol, resuspend the air-dried bisulfate-converted DNA pellet in 20 ml of nuclease-free water. It is possible to quantify this DNA using a Nanodrop spectrophotometer.
196
Dallol et al.
3.3. M ethyLight Assay
The MethyLight assay utilizes the TaqMan PCR principle, which requires forward and reverse primers, as well as an oligomeric probe that emits fluorescence only after it is degraded by the 5¢–3¢ exonuclease activity of Taq polymerase. Each PCR uses the same basic reaction setup. The choice of primer/probe sets is the only variable in these reactions. All primer/probe sets used are diluted to the same stock concentrations to standardize the PCR setup as well as the running of the PCR program.
3.3.1. Primer and Probe Design
In order to design a MethyLight assay for any given gene, the CpG island residing at the promoter region of that gene should be predicted and determined (Fig. 1). There are several programs that can be utilized to achieve such aim in addition to design the MethyLight probes and primers. An example of such program can be the Beacon Designer (Premier BioSoft) or the online BiSearch program (http://bisearch.enzim.hu/). However, for those who do not have access to a dedicated primer design software, the procedure can be done in a “semimanual” fashion. We routinely use the CpG island prediction track from the UCSC genome browser (http://genome.ucsc.edu). The DNA sequence of the required CpG island can be processed in any word processing software like Microsoft Word. Simple macros can then be utilized to search for the CpG dinucleotides and convert all the other cytosine nucleotides into thymidine. Following sequence conversion and determination of the region of interest, the methylation-specific primers to be used for amplification could be designed so that they amplify short products (60–150 bp). Once modified, the sense and antisense strands are no longer complementary; therefore, a decision has to be made regarding what strand to be used for the analysis. Generally, the regular PCR primer designing rules should be followed especially in terms of Tm for both primers being in the same range and checking the specificity of annealing by performing BLAST searches (using the original unmodified sequence). Additional rules are applied such as the Tm of the MSP primers being significantly higher than that
Fig. 1. Designing the primer/probes for the MethyLight assay for the IGF2 promoter.
Analysis of DNA Methylation in FFPE Tissues Using the MethyLight Technology
197
for the primers targeting the unmethylated alleles. In addition, it is better to stop at a C (of a CpG dinucleotide) at 3¢ of the oligonucleotide. Furthermore, differences with unmodified DNA sequence can be increased by including as many as possible of thymidine residues that were originally cytosines. 3.3.2. Preparation of Probe/Primer Mix (Oligo-Mix)
1. Resuspend the lyophilized oligonucleotides (primers and probe) in sterile nuclease-free water. Leave on ice for 10 min before vortexing briefly and centrifugation to collect any liquid underneath the lid. Add enough water to achieve a stock solution of 200 mM. Avoid repeated freeze–thaw cycles by making aliquots. Store at −20°C. The final concentration of each oligonucleotide (primers and probe) is 0.4 mM (forward and reverse primers) and 0.2 mM (probe) in a final volume of 25 ml. 2. It is recommended to prepare a 10× mixture of the oligonucleotides (10× Oligo-Mix) prior to starting the procedure. To make a 100 ml of the 10× Oligo-Mix, combine 2 ml of the forward primer, 2 ml of the reverse primer, and 1 ml of the probe with 95 ml of nuclease-free water.
3.3.3. Preparation of DNA Samples
1. Measure the concentration of the bisulfite-treated DNA samples required using the nanodrop spectrophotometer and record the measurements. Dilute each bisulfite-treated DNA sample to have a final concentration of 5 ng/ml using nuclease-free water. 2. The prepared new tubes must be labeled carefully according to their original undiluted DNA sample number. The remainder original DNA samples are returned to the freezer to be preserved under −70°C. The diluted samples are kept on ice during the procedure (see Note 14).
3.3.4. MethyLight Reaction Setup
1. The components of the MethyLight reaction setup are bisulfite-converted DNA samples, positive and negative control DNA, Oligo-Mix, and the commercially available PCR master mix. The steps involved include setting up a standard curve and test wells. It is recommended to set up the standard curve and the methylation controls in duplicate. However, the test samples can be done in single wells. 2. To use a five-points standard curve, make a serial 1:5 dilution in nuclease-free water of the bisulfite-converted and in vitro methylated DNA. This will generate solutions with 5, 1, 0.2, 0.04, and 0.008 ng/ml concentrations (representing 10, 2, 0.4, 0.08, and 0.016 ng/well). Add 2 ml of each dilution into a well of the 96-well optical plate. Repeat this once in the next set of wells (see Note 15 and PMR calculation). 3. In order to normalize the methylation value of the test gene to the amount of input DNA, it is important to always include
198
Dallol et al.
the amplification of COL2A1 from test DNA in the reaction setup (see Note 16). Prepare the PCR master mix as follows per sample/well: 12.5 ml TaqMan 2× Universal PCR Master Mix, 2.5 ml Oligo-Mix for test gene, 2.5 ml for COL2A1 control gene, 5.5 ml nuclease-free water. 4. For the COL2A1 control master mix, prepare enough solution for 12 samples (eight standard curve points in addition to a water-only control in duplicate). The master mix should therefore include 150 ml TaqMan 2× Universal PCR Master Mix, 30 ml Oligo-Mix for COL2A1 control gene, 96 ml nuclease-free water. 5. The CIMP status of a given colorectal cancer sample can be queried by analyzing the methylation levels of seven genes (IGF2, CDKN2A, SOCS1, CRAPB1, NEUROG1, CACNA1G, and RUNX3). To do this, ten colorectal samples can be tested per plate (ten wells each row per gene plus the M.SssI-methylated control in duplicate). Therefore, prepare a different PCR master mix for each individual gene for 14 samples (accounting for pipetting errors) by combining 175 ml TaqMan 2× Universal PCR Master Mix, 35 ml OligoMix for test gene, 35 ml for COL2A1 control gene, and 77 ml nuclease-free water (Table 1). 6. Prepare the standard curve and test DNA as well as the bisulfite-converted unmethylated DNA control by aliquoting 2 ml of each DNA (10 ng) into the required wells of the MicroAmp Fast Optical 96-well reaction plate (0.1 ml). Do not forget to include nuclease-free water only to DNA negative control. 7. To each well, add 23 ml of the relevant PCR master mix. This will make up the reaction volume to 25 ml per well (see Note 17). 8. The reaction plate is then sealed with MicroAmp Optical Adhesive cover. Hold the plate from the sides and do not try to extend the cover on the plate with your gloved fingers, instead use the edge of a plastic extender instrument supplied with the adhesive covers. 9. In a plate centrifuge, the covered plate is placed carefully in the machine and centrifuged for 1 min at 3,600 × g. This will serve to collect all the solution at the bottom of the well and remove any bubbles introduced during the pipetting procedure. 10. The reaction plate is kept on ice until setting up the 7500 Fast System SDS Software. The PCR cycling conditions should be set up according to the following steps: 95°C for 10 min followed by 45 cycles of 95°C for 15 s and 60°C for 1 min (Fig. 2).
AACAACGCCCGC AACTCCT
CCGAAACCATCTTCAC GCTAA TATCCGTACCTACCG CCGC CGATAATTACGAACACACT CCGAAT
TGGAGTTTTCGGTTGATTGGTT
GCGTCGAGTTCGTGGGTATTT
TCGAAATTTTTGTTGCGT
CGTGTAGCGTTCGGGTATTTGTA
CDKN2A Cyclin-dependent kinase inhibitor 2A (p16/INK4A)
SOCS1 Suppressor of cytokine signaling 1
CRAPB1 Cellular retinoic acid binding protein 1
NEUROG1 Neurogenin 1
GGGAAGATGGGATAG AAGGGAATAT
TCTAACAATTATAAACTCCAACC ACCAA
COL2A1* Collagen 2A1
* control probe sequence.
GACGAACAACGTCTT ATTACAACGC
CGTTCGATGGTGGACGTGT
RUNX3 Runt-related transcription factor 3
CTCGAAACGACTT CGCCG
CCAACTCGATTTAAA CCGACG
GAGCGGTTTCGGTGTCGTTA¢
IGF2 Insulin-like growth factor 2
TTTTTTCGTTTCGCGTTTAGGT CACNA1G Calcium channel, voltage dependent, T-type alpha-1G subunit
Reverse primer (5¢–3¢)
Forward primer (5¢–3¢)
Gene
(10)
(10)
(10)
(11)
(10)
(9)
(10)
References
VIC/6FAM-CCTTCATTCTAACCCAATA (9) CCTATCCCACCTCTAAA-TAMARA
VIC-CGCACGAACTCGCCTACGT AATCCG-TAMARA
6FAM-AAATAACGCCGAATCCGA CAACCGA-TAMARA
6FAM-CGATAACGACCTCCCGCG AACATAAA-TAMARA
VIC-ACCATACCCAACTTCGCCG ACACCTAA-TAMARA
6FAM-ACAATTCCGCTAACGAC TATCGCGCA-TAMARA
VIC-ACCCGACCCCGAACCGCGTAMARA
VIC-CCCTCTACCGTCGCGAACCCGATAMARA
Probe oligo sequence (5¢–3¢)
Table 1 Primer and probe sequences used for the determination of the CIMP status in colorectal cancer
200
Dallol et al.
1.000e+001
1.000e+000
1.000e−001
Delta Rn
1.000e−002
1.000e−003
1.000e−004
1.000e−005
1.000e−006
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Cycle Number
Fig. 2. MethyLight amplification plot of DNA extracted from paraffin-embedded colon cancer tissues using IGF-2 primer (IGF-2 as an example). X-axis represents the number of cycles in which the samples started the amplification, and Y-axis represents delta Rn (delta Rn is an increment of fluorescent signal at each time point).
3.4. C alculating PMR
The plate setup is important to get an accurate representation of methylated ratio. The first ten wells of the first row are reserved for a standard curve (Subheading 3.3.4). Thus, wells A1–A2 have 10 ng, wells A3–A4 contain 2 ng, wells A5–A6 have 0.4 ng, wells A7–A8 have 0.08 ng, and A9–A10 contains 0.016 ng of M.SssItreated and bisulfite-modified DNA. Reserve six additional wells for the 100% M.SssI-methylated control and no-template negative controls, two wells for 100% M.SssI-methylated and bisulfitetreated control DNA containing your test Oligo-Mix (call this as B after averaging the concentration for simplicity), and two wells for 100% M.SssI-methylated and bisulfite-treated control DNA containing COL2A1 Oligo-Mix (call this as D). All the standard curve wells should contain the master mixes and primers and probes (Oligo-Mix) for the COL2A1 housekeeping/control gene as described above. This standard curve is used to convert the number of cycles (CT) to concentrations for both the control gene and the test gene (using the same COL2A1 standard curve). There is no need to construct another standard curve for the test gene (Fig. 3).
Analysis of DNA Methylation in FFPE Tissues Using the MethyLight Technology Standard Curve
44.500 42.500
Omit Wells Show all standard points for selected detector(s)
40.500
Detector: MLH1
38.500 Ct
201
Key MLH1 Std Unknown
36.500 34.500 32.500 30.500 29.514 −3.724
Slope: −2.957464 Intercept: 33.543736 R2: 0.977687
−3.000
−2.000
−1.000
0.000
1.000 1.301
Log C0
Fig. 3. MethyLight amplification plot of the standard curve used to calculate the quantity of percentage of methylated ratio (PMR) value for MLH1 gene.
Export the CT values and corresponding concentrations for the standard curve to Excel. Plot a graph with the CT values at the X-axis and the concentrations as Y-values (after converting the Y-axis to log10 scale). Generate a logarithmic trend line and get the formula of the straight line in the form of Y = mX + c (m is the slope and c is the Y-intercept where the trend line crosses the Y-axis). Use the following formula to calculate the quantity of the sample (Q):
Q = 2.71828(CT -c /m). For each test sample, you should get the two quantities or concentrations; the first is for the test gene (call this as A for simplicity) and the second is for the control gene COL2A1 (call this as C). PMR is calculated by dividing the quantity of the test gene in the sample by the quantity of test gene in the M.SssI-methylated and bisulfite-treated control DNA (ratio 1) over the quantity of the COL2A1 in the sample by the quantity of COL2A1 in the M.SssI-methylated and bisulfite-treated control DNA (ratio 2) and multiplying by 100. More simply, as depicted with the hypothetical numbers we called 2, 4, 1, and 3 above, divide the ratios A/B by C/D and multiply by 100 according to the following formula:
æ A /Bö PMR = ç ´ 100. è C / D ÷ø
By convention, a PMR value of 10 and above is considered as evidence for specific gene methylation. Samples with PMR below 10 are considered unmethylated. We believe that this division is arbitrary and is in need of further examination.
202
Dallol et al.
4. Notes 1. The use of DNA from established cell lines in this procedure can be advantageous or disadvantageous depending on the degree of polyploidy of the cell line used. The degree of ploidy in the methylated control may affect the calculation of the PMR in the test sample, and changing the origin of the control DNA mid-analysis may introduce undesirable reproducibility issues. 2. 1 M Hydroquinone should be colorless. However, do not use if the color turns pink. 3. Sodium metabisulfite is light-sensitive, so it should always be kept in the dark during the preparation procedure. 4. We found this kit to give far superior cleanup efficiency compared to kits provided by other suppliers or traditional methods. 5. As an alternative to this method, we used the EpiTect Bisulfite Conversion kit from Qiagen (CA, USA) with excellent results especially from DNA extracted from FFPE tissues. 6. The amplification primers should aim to amplify the shortest possible sequence (60–150 bp) since the template DNA obtained from FFPE tissue will be mostly fragmented and in poor quality. TaqMan probes are labeled with 5¢ reporter dye and 3¢ nonfluorescent quencher. The probe can be tagged at the 5¢ with any fluorescent reporter molecule; however, this choice should take into consideration the capabilities of the Real-time PCR system available. A commonly used reporter dye is 6-FAM [6-Carboxyfluorescein] with absorbance/emission range of 495/517 nm. The nonfluorescent quencher we routinely use is the Black-Hole Quencher 1 (BHQ1). The BHQ1 quencher has a maximum absorbance up to 534 nm, which should be able to quench light emitted from 6-FAM. For amplification of very short regions, the amplification primers and probe can be shortened in length. Naturally, the specificity of the PCR will become a concern. However, the use of the 3¢-Minor groove binder-DNA probes (MGB probes) as supplied by Applied Biosciences can overcome this problem. The MGB Moiety stabilizes the hybridized probe and effectively raises its melting temperature (Tm). TaqMan probes and primers are sensitive to light and should be kept in dark (covered by foil) at −20°C. 7. Both control DNA sets can be purchased from Qiagen as part of a PCR control DNA set.
Analysis of DNA Methylation in FFPE Tissues Using the MethyLight Technology
203
8. This master mix is optimized for 5¢ nuclease assay using TaqMan® probes and contains a hot-start DNA polymerase that does not require an activation step. In addition, it lacks AmpErase® Uracil N-glycosylase (UNG), which otherwise would cleave the bisulfite-converted DNA that contains uracil for the unmethylated cytosines. We also used the EpiTect MethyLight Master Mix for methylation-specific real-time PCR analysis provided by Qiagen. 9. NEB recommends modifying 1 mg in 20 ml reaction. Therefore, the reaction is scaled up to prepare a larger batch of in vitro methylated DNA and thus reduce the variations in modification efficiencies that can happen from lot to lot. This means that the reaction volume must be scaled up as well to reduce the inhibitory effects of salts and other contaminants that can remain in the genomic DNA sample. 10. In order to guarantee complete methylation with a high level of certainty, the control DNA should be subjected to several rounds of M.SssI modifications and purification prior to bisulfite conversion. 11. Using too much starting material may interfere with the bisulfite treatment and leaves a considerable amount of unconverted DNA in the sample. This would potentially lead to false-positive results. 12. The reaction will not work efficiently unless the entire DNA has been denatured and the strands are separated. Any nondenatured DNA left in the sample will be resistant to conversion and may yield to false-positive results. 13. This is achieved by transferring the reaction mix to a 1.5-ml tube that contains prewarmed 1 ml of the resin provided in the kit. Mix by inversion and transfer to the filter attached to a syringe barrel. The syringe plunger can be used or a vacuum manifold to push through the DNA–resin mix through the filter. Wash the DNA with 2 ml of 80% isopropanol made in nuclease-free water. Centrifuge to remove any isopropanol traces and elute the DNA from the column with 50 ml of nuclease-free water. 14. Bisulfite-converted DNA is susceptible to degradation since it is single-stranded. Avoid repeated freeze–thaw cycles and aliquot the converted DNA to minimize degradation. 15. For example, wells A1–A4 can be used for the standard curve and wells A5–A9 can be used for its replicate. The bisulfitespecific amplification of the house-keeping gene COL2A1 or the ALU repeats can be used to convert the number of cycles to concentration for all samples amplified with control and test primers (for alternative plate design, see PMR calculation notes).
204
Dallol et al.
16. This can be done in separate wells in replicates, but it can also be included as part of a multiplex PCR with the test gene. To do this, make sure that the probes for the control and test genes are labeled with different fluorophores (e.g., 6FAM and VIC with BHQ and TAMRA as quenchers, respectively). 17. Total volume as low as 10 ml was successfully used in this setup.
Acknowledgments This work is supported by Kuwait Foundation for the Advancement of Sciences grant number 2006-1302-07, Kuwait University Grant number MG02/08, and Research Core Facility (RCF) grant number GM 01/01and GM 01/05. References 1. Al-Mulla, F. (2007) Utilization of microarray platforms in clinical practice: an insight on the preparation and amplification of nucleic acids from frozen and fixed tissues. Methods Mol Biol. 382, 115–136. 2. Gagnon, J. F., Sanschagrin, F., Jacob, S., Tremblay, A. A., Provencher, L., Robert, J., et al. (2009) Quantitative DNA methylation analysis of laser capture microdissected formalin-fixed and paraffin-embedded tissues. Exp Mol Pathol. 8, 8. 3. Dietrich, D., Lesche, R., Tetzner, R., Krispin, M., Dietrich, J., Haedicke, W., et al. (2009) Analysis of DNA methylation of multiple genes in microdissected cells from formalinfixed and paraffin-embedded tissues. J Histochem Cytochem. 57, 477–489. 4. Clark, S. J., Harrison, J., Paul, C. L., and Frommer, M. (1994) High sensitivity mapping of methylated cytosines. Nucleic Acids Res. 22, 2990–2997. 5. Eads, C. A. and Laird, P. W. (2002) Combined bisulfite restriction analysis (COBRA). Methods Mol Biol. 200, 71–85. 6. Herman, J. G., Graff, J. R., Myohanen, S., Nelkin, B. D., and Baylin, S. B. (1996) Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands. Proc Natl Acad Sci USA. 93, 9821–9826.
7. Eads, C. A., Danenberg, K. D., Kawakami, K., Saltz, L. B., Blake, C., Shibata, D., et al. (2000) MethyLight: a high-throughput assay to measure DNA methylation. Nucleic Acids Res. 28, E32. 8. Issa, J. P. (2004) CpG island methylator phenotype in cancer. Nat Rev Cancer. 4, 988–993. 9. Widschwendter, M., Siegmund, K. D., Muller, H. M., Fiegl, H., Marth, C., Muller-Holzner, E., et al. (2004) Association of breast cancer DNA methylation profiles with hormone receptor status and response to tamoxifen. Cancer Res. 64, 3807–3813. 10. Weisenberger, D. J., Siegmund, K. D., Campan, M., Young, J., Long, T. I., Faasse, M. A., et al. (2006) CpG island methylator phenotype underlies sporadic microsatellite instability and is tightly associated with BRAF mutation in colorectal cancer. Nat Genet. 38, 787–793. 11. Ogino, S., Brahmandam, M., Kawasaki, T., Kirkner, G. J., Loda, M., and Fuchs, C. S. (2006) Epigenetic profiling of synchronous colorectal neoplasias by quantitative DNA methylation analysis. Mod Pathol. 19, 1083–1090. 12. Esteller, M. (2003) Relevance of DNA methylation in the management of cancer. Lancet Oncol. 4, 351–358.
Chapter 14 RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues Prepared Using a Range of Different Fixatives and Conditions Mei-Lan Liu, Jennie Jeong, Ranjana Ambannavar, Carl Millward, Frederick Baehner, Chithra Sangli, Debjani Dutta, Mylan Pho, Anhthu Nguyen, and Maureen T. Cronin Abstract Although RNA is isolated from archival fixed tissues routinely for reverse transcription polymerase chain reaction (RT-PCR) and microarray analyses to identify biomarkers of cancer prognosis and therapeutic response prediction, the sensitivity of these molecular profiling methods to variability in pathology tissue processing has not been described in depth. As increasing numbers of expression analysis studies using fixed archival tumor specimens are reported, it is important to examine how dependent these results are on tissue-processing methods. We carried out a series of studies to systematically evaluate the effects of various tissue-fixation reagents and protocols on RNA quality and RT-PCR gene expression profiles. Human placenta was selected as a model specimen for these studies since it is relatively easily obtained and has proliferative and invasive qualities similar to solid tumors. In addition, each specimen is relatively homogeneous and large enough to provide sufficient tissue to systematically compare a range of fixation conditions and reagents, thereby avoiding the variability inherent in studying collections of tumor tissue specimens. Since anatomical pathology laboratories generally offer hundreds of different tissue-fixation protocols, we focused on fixation reagents and conditions used to process the most common solid tumors for primary cancer diagnosis. Fresh placentas donated under an IRB-approved protocol were collected at delivery and immediately submerged in cold saline for transport to a central pathology laboratory for processing. RNA was extracted from each specimen, quantified, and analyzed for size distribution and analytical performance using a panel of 24 RT-PCR gene expression assays. We found that different tissue-fixation reagents and tissue-processing conditions resulted in widely varying RNA extraction yields and extents of RNA fragmentation. However, the RNA extraction method and RT-PCR assays could be optimized to achieve successful gene expression analysis for nearly all fixation conditions represented in these studies. Key words: Breast cancer, Formalin-fixed, paraffin-embedded tissue, Quantitative RT-PCR, Gene expression profiling, Molecular biomarkers, Prognosis, Prediction
Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_14, © Springer Science+Business Media, LLC 2011
205
206
Liu et al.
1. Introduction It is well known that pathology processes for tissue preservation are not standard and diverge widely among anatomical pathology laboratories. Fixation protocols vary with the tissue being fixed, individual laboratory standards, and the range of diagnostic tests planned for the tissue specimen. Even within laboratories, protocols and reagents may vary over time, particularly when measured over decades. Therefore, clinical study protocols requiring molecular analysis of paraffin-embedded tissues from pathology archives are challenged to accommodate inherently heterogeneous sample sets. As technologies for extracting and analyzing nucleic acids from fixed, paraffin-embedded tissues are now broadly available, it has become important to assess whether variations in fixation conditions and reagents affect the success of downstream molecular assays and therefore impact the interpretation of clinical study results. Reports of successful RNA recovery from archival fixed, paraffin-embedded tissues and its subsequent analysis in molecular assays for biomarker discovery have appeared regularly during the last 20 years and currently appear relatively frequently (1–14). However, these reports generally describe small-scale studies designed to test modest numbers of markers and are generally limited to demonstrating the feasibility of accessing RNA from fixed tissues for molecular profiling. More recently, large-scale studies where gene expression profiling has been done for hundreds of candidate biomarkers from hundreds of preserved clinical specimens have been reported, bringing the routine use of molecular markers for personalized medicine to reality (15–26). Two key advances have made these clinical biomarker discovery programs possible: the development of extraction methods for isolating RNA from fixed tissues and the optimization of expression profiling assays for the fragmented RNA recovered from these tissues. Expression profiling assays for RNA from FFPE tissues are now frequently reported for both RT-PCR assay methods and microarray assay methods (11, 27–36). With molecular profiling in archival samples being done widely, it is important to investigate how much impact pathology fixation reagents and tissue-processing conditions have on molecular expression profiling assay results. Understanding tissue-fixation effects is equally important to characterize for the commercial reference laboratory setting since diagnostic tests carried out on fixed tissues received from many different laboratories will represent a wide range of protocols and conditions. Fixed paraffin-embedded tissues (FPET) are the products of essentially all surgical pathology procedures and therefore are ideal substrates for molecular analysis. Using these tissues for molecular diagnostic tests offers the added advantages that they are stable over a wide temperature range, are nonbiohazardous,
RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues
207
and the molecular data from a genomic assay can be directly correlated with standard anatomic pathology data for the same fixed tissue. Furthermore, a number of studies have shown the close similarity between RNA expression profiles in samples where paired FPET specimens and frozen specimens are available (37–40). The ultimate benefit of using gene expression biomarkers in FPET specimens is the ability to develop and validate clinical diagnostic assays based on prospective analysis of archival samples from completed clinical studies. Therefore, using human placenta as an experimental model, we systematically studied the effect of a range of typical pathology processes on the quality of RNA recovered from fixed tissue and its performance in RT-PCRbased gene expression profiling assays.
2. Materials 2.1. Sample Preparation
1. Leica RM2235 Microtome (Leica Biosystems, Nussloch, Germany). 2. VWR Microscope sides, Superfrost Plus (VWR International, LLC, West Chester, PA). 3. Hematoxylin and eosin stain, standard reagent-grade hematoxylin, eosin, alcohol, and xylene. 4. Polypropylene 1.5-ml microcentrifuge tubes, nuclease-free grade.
2.2. R NA Extraction
1. MasterPure™ RNA Purification Kit (Epicentre Bio technologies, Madison, WI) containing Tissue and Cell Lysis Solution (1×), Proteinase K (50 mg/ml), MPC Protein Precipitation Reagent, RNase-Free DNase I enzyme (1 U/ml), 1× DNase Buffer, 2× T and C Lysis Solution (see Note 1). 2. 75% Ethyl alcohol, molecular biology grade (prepared from ethyl alcohol, 200 proof). 3. Ethyl alcohol, 200 proof, molecular biology grade (Sigma– Aldrich, St. Louis, MO). 4. Isopropanol (Omnisolv, EM Science, Gibbstown, NJ). 5. Nuclease-free water (Ambion, Austin, TX). 6. Xylene. 7. Nutator, Clay Adams (Becton, Dickinson & Co, Sparks, MD). 8. Thermomixer R (Eppendorf North America, Hauppauge, NY). 9. TRIzol® Reagents (Invitrogen Corporation, Carlsbad, CA). 10. Chloroform.
208
Liu et al.
2.3. R NA Quantitation
1. NanoDrop ND-1000 Sepctrophotometer Technologies, Willmington, DE).
(NanoDrop
2. Quant-iT™ RiboGreen® RNA Reagent and Kit (Invitrogen, Molecular Probes, Eugene, OR). 2.4. R NA Sizing
1. Novex® TBE-Urea Gels (15%, 1.0 mm) (Invitrogen Corporation, Carlsbad, CA). 2. RNA Century Marker (Ambion, Austin, TX). 3. XCell SureLock™ Carlsbad, CA).
Mini-Cell
(Invitrogen
Corporation,
4. 5× Novex® TBE Running buffer (Invitrogen Corporation, Carlsbad, CA). 5. Nuclease-Free Water (Ambion, Austin, TX). 6. SYBR® Gold nucleic acid gel stain (10,000× concentrate in DMSO) (Invitrogen Corporation, Carlsbad, CA). 7. Agilent 2100 BioAnalyzer, RNA 6000 Nano LabChip® kits (Agilent Technologies, Inc., Santa Clara, CA). 8. IKA vortexer (Agilent Technologies, Inc., Santa Clara, CA). 9. Chip Priming Station (Agilent Technologies, Inc., Santa Clara, CA). 2.5. Reverse Transcription
1. Hard-Shell thin-wall 96-well microplate or 96-well Twin-tec PCR, Skirt-C (MJ Research, Inc., Waltham, MA). 2. Nuclease-free water (Ambion, Austin, TX). 3. PCR strip tubes with cap (Phenix Research Products, Road Candler, NC). 4. Polarseal foil (E&K Scientific, Santa Clara, CA). 5. Universal human reference RNA (Stratagene, San Diego, CA). 6. Gene-specific oligonucleotide primers. 7. Omniscript RT Kit containing RNase-free water, 5 mM dNTP Mix, 10× RT Buffer, 4 U/ml Omniscript RT (Qiagen, Valencia, CA). 8. RNase inhibitor (20 U/ml) (Applied Biosystems, Foster City, CA).
2.6. Polymerase Chain Reaction
1. 384-Well clear optical reaction plate or 384-well clear Twin-tec plate (VWR International). 2. ABI PRISM 7900HT sequence detection system (Applied Biosystems, Foster City, CA). 3. Hard-Shell thin-wall 96-Well microplate or 96-well Twin-tec PCR, Skirt-C (MJ Research, Inc., Waltham, MA). 4. Nuclease-free water (Ambion, Austin, TX).
RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues
209
5. Optical adhesive covers (Applied Biosystems, Foster City, CA). 6. Polar seal foil (E&K Scientific, Santa Clara, CA). 7. Sealing tape for 96-well plates (E&K Scientific, Santa Clara, CA). 8. Oligonucleotide primers and dual-labeled TaqMan probe for each gene assay. 9. 2× TaqMan® Universal PCR Master Mix, no AmpErase UNG (Applied Biosystems, Foster City, CA).
3. Methods A placenta was obtained from a consenting donor, and immediately after delivery it was chilled to 4°C in normal saline for transporting to a pathology laboratory for further processing. A relatively homogeneous portion of the placenta was selected to be cut into similar size pieces (1 × 1 × 0.5 cm) for fixation by a series of standardized tissue-processing protocols. Additional tissue was prepared using two fresh tissue-processing protocols; OCT frozen tissue preparation and RNAlater™ preparation (see Table 1).
Table 1 Tissue-fixation methods and reagents tested in this study Fixing condition
Fixative composition
pH
Buffer system
RNAlater® (Ambion, Austin, TX)
NA
NA
Proprietary
OCT fresh frozen (Sakura Finetek, Torrance, CA)
Polyvinyl alcohol, polyethyleneglycol
NA
NA
Neutral buffered formalin (Richard Allan Scientific, Kalamazoo, MI)
4% Formaldehyde, sodium phosphate
6.8–7.2
Phosphate
Ethanol (Allegiance Cardinal, Dublin, OH)
70% Ethanol
NA
NA
B5 fixative (AMTS, Inc., Modesto, CA)
4% Formaldehyde, mercuric chloride, sodium acetate
~6.0
Acetate
Bouin’s fixative (AMTS, Inc., Modesto, CA)
9.5% Formaldehyde, picric acid, 1.0–2.0 methanol, glacial acetic acid
NA
Buffered zinc formalin (Anatech, Battle Creek, MI)
3.7% Formaldehyde, ionized zinc, buffer
5.6–5.8
Acetate
Pen-Fix (Richard-Allan Scientific, Kalamazoo, MI)
10% Formaldehyde, ethanol, isopropanol, methanol
6.9–7.1
Acetate
Prefer-Fix (Anatech, Battle Creek, MI)
Glyoxal, ethanol, buffer
3.75–4.25
Proprietary
210
Liu et al.
A second placenta, collected in the same way as described above, was used to test a range of neutral buffered formalin processing conditions. Portions of this placenta were also immediately preserved in RNAlater™ reagent to provide a fresh tissue baseline for reference. The remaining placenta tissue was stored in normal saline at 4°C for approximately 72 h before the OCT frozen and neutral buffered formalin fixation protocols were carried out following a schema designed to represent a full range of common fixation scenarios, including short (underfixation), standard and long (overfixation) formalin processing times; underfixation followed by reprocessing (reverse processing) and a standard fixation protocol applied to very small tissue samples (overfixation) (see Fig. 1). 3.1. Preparation of Samples for Anatomic Pathology Review 3.2. Preparation of Samples for RNA Extraction and Deparaffinization
One hematoxylin and eosin (H&E) stained slide was prepared from each paraffin and OCT block included in this study to assess the fixation and the morphology of the tissue blocks. After this review, the most representative, well-fixed blocks were used for the molecular portion of the study. 1. Preparing formalin-fixed, paraffin-embedded samples for RNA extraction requires using a microtome to cut sections from paraffin blocks in which the fixed tissues have been embedded (see Note 2). One H&E-stained slide was prepared for each OCT fresh-frozen block and each fixed paraffin-embedded block for histopathology review. 2. Although the dimensions of embedded tissues vary widely, three 10-mm FPE sections cut from one paraffin block from each patient case is a good rule of thumb for obtaining microgram amounts of RNA. Cut the three sections sequentially, allowing them to curl as the microtome blade moves over the block surface.
RNAlater ® OCT Fresh Frozen Neutral Buffered Formalin (NBF) 1.5 to 2-Hour NBF Fixation 1.5 to 2-Hour NBF Fixation with Reverse Processing 3 to 4-Hour NBF Fixation 6 to 8-Hour NBF Fixation 6 to 8-Hour NBF Fixation (Extra Small Tissue) 48 to 72-Hour NBF Fixation
Fig. 1. Processing schema comparing a range of neutral buffered formalin tissue-fixation protocols with fresh (RNAlater) and OCT frozen tissues.
RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues
211
3. Using a fine brush and/or tissue forceps, place all three 10-mm sections into one 1.5-ml polypropylene microcentrifuge tube for extraction. 4. Centrifuge the sample tubes in a microcentrifuge set between +18 and +25°C for 30 s at approximately 6,000 rpm (3824 × g). 5. Add 1 ml of xylene to each tube containing the tissue sample. Invert the tube until the tissue sample is dislodged from the bottom of the tube. Mix the samples on the Nutator for 5 min. Centrifuge the samples in a microcentrifuge set between +18 and +25°C for 5 min at approximately 14,000 rpm (20,817 × g). Carefully aspirate and discard the supernatant without disturbing the tissue at the bottom of the tube. 6. Repeat the steps for a total of two xylene extractions. Visually inspect the sample for evidence of paraffin. If residual paraffin exists, repeat the xylene extraction process until the paraffin is completely removed. 7. Add 1 ml of 100% ethanol to each sample tube. Invert and shake each tube until the tissue sample is dislodged from the bottom of the tube. Loosening the tissue increases the surface area exposed to the ethanol. Mix the samples on the Nutator for 5 min to ensure adequate mixing. Centrifuge the sample tubes in a microcentrifuge set between +18 and +25°C for 5 min at approximately 14,000 rpm (20,817 × g). Repeat for a total of two ethanol washes. 3.3. RNA Extraction from Fresh Tissue Preserved in RNAlater™ Using the TRIzol® Reagent
1. Prior to homogenization with a polytron homogenizer, each sample was weighed and cut into appropriate size pieces to ensure that no more than 100 mg of tissue was homogenized in 1 ml of TRIzol® Reagent (see Note 3). 2. Incubate the homogenized samples for 5 min at room temperature to permit the complete dissociation of nucleoprotein complexes. 3. Add 200 ml of chloroform per 1 ml of TRIzol® Reagent. Shake tubes vigorously by hand for 15 s and incubate the samples for 3 min at room temperature. 4. Centrifuge the samples at 14,000 rpm (20,817 × g) for 15 min at 2–8°C. RNA remains exclusively in the aqueous phase. 5. Transfer the upper aqueous phase (clear solution) into a fresh tube. 6. Add 500 ml of isopropyl alcohol per 1 ml of TRIzol® Reagent used for the initial homogenization and mix well for 1 min on the Nutator. 7. Incubate samples at room temperature for 10 min and centrifuge the samples at 14,000 rpm (20,817 × g) for 15 min at 2–8°C. 8. The RNA precipitate forms a gel-like pellet on the side and bottom of the tube.
212
Liu et al.
9. Remove the supernatant and discard. 10. Wash the RNA pellet in 1 ml of 75% ethanol and centrifuge the sample at 8,750 rpm (8,225 × g) for 5 min. 11. Repeat the 75% ethanol wash one more time. 12. Briefly dry the RNA pellet for 1–2 min in a 37°C heat block to remove the ethanol. 13. Resuspend the pellet in 50 ml of nuclease-free water. 14. The RNA extracts were treated with 20 units of RNase-free DNase I (Ambion, Austin, TX) for 1 h at 37°C to remove residual genomic DNA. 15. RNA pellets were resuspended in 100 ml of nuclease-free water. 3.4. RNA Extraction from FPE and OCT Tissues Using the MasterPure™ RNA Purification Kit (Epicentre Biotechnologies, Madison, WI)
1. RNA was isolated from the OCT fresh-frozen and fixed paraffin-embedded samples using the MasterPure™ Purification Kit according to the manufacturer’s protocol, as described below. 2. After the deparaffinization step, the samples are ready for tissue digestion. 3. Prepare Proteinase K/Tissue and Cell Lysis Solution: Calculate the amount of Tissue and Cell Lysis Solution (1×) needed including a 10% overage when samples will be processed as a batch: (No. of samples × 1.1) × 298 ml. Calculate the amount of Proteinase K (50 mg/ml) needed includinga 10% overage when samples will be processed as a batch: (No. of samples × 1.1) × 2 ml. Preparation of Proteinase K/Tissue and Cell Lysis Solution Reagent
1 reaction (ml/reaction)
Tissue and Cell Lysis Solution (1×)
298
Proteinase K
2
Dispense the required volume of Tissue and Cell Lysis Solution (1×) prepared into a microcentrifuge tube. Add the required volume of Proteinase K to the centrifuge tube containing the 1× Tissue & Cell Lysis Solution. Important: Do not store the Proteinase K/Tissue and Cell Lysis Solution on ice; rather keep the solution at room temperature. 4. Add 300 ml of Proteinase K/Tissue and Cell Lysis Solution to each sample tube. Tightly cap each tube. Vortex each tube for approximately 5–10 s. 5. Incubate the samples at +65°C for 2 h. After the 2-h incubation in a thermomixer at 850 rpm, shake down or centrifuge the tubes in a microcentrifuge set at +20°C for 5–30 s at 3,000 rpm (956 × g). Place the samples on ice for 3–5 min.
RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues
213
6. Add 150 ml of MPC Protein Precipitation Reagent to each sample and vortex the samples for approximately 10 s. Centrifuge the samples in a microcentrifuge set at +20°C for 10 min at approximately 14,000 rpm (20,817 × g). 7. Obtain a new 1.5-ml microcentrifuge tube for each sample. Dispense 500 ml of isopropanol (at +4°C) into each tube. Aspirate the supernatant from the sample tube and transfer it to the matching 1.5-ml microcentrifuge tube containing the isopropanol. 8. Mix the sample for approximately 3 min on the Nutator (or manually invert the tubes 30–40 times). Centrifuge the samples in a microcentrifuge at +4°C for 10 min at 14,000 rpm (20,817 × g). Discard the supernatant without disturbing the precipitated nucleic acid pellet (DNA and RNA). 9. Prepare DNase I Solution: Calculate the amount of 1× DNase Buffer needed as follows: (No. of samples × 1.1) × 180 ml. Calculate the amount of RNase-free DNase I enzyme (1 U/ml) needed as follows: (No. of samples × 1.1) × 20 ml. Dispense the required volume of 1× DNase Buffer into a centrifuge tube. Add the required volume of RNase-free DNase I enzyme to the test tube containing the 1× DNase Buffer. Important: Prepare the DNase I Solution immediately before dispensing it into the samples. Preparation of DNase I Solution Reagent
1 reaction (ml/reaction)
1× DNase Buffer
180
RNase-free DNase I enzyme
20
10. Add 200 ml of the DNase I Solution to each sample. Flick the sample tubes until the pellet is completely resuspended in the solution. Centrifuge the samples in a microcentrifuge set between +18 and +25°C for 30–60 s at 3,000 rpm (956 × g). 11. Incubate the samples at +37°C for 1 h. 12. Add 200 ml of 2× Tissue and Cell Lysis Solution to each sample. Mix the samples by vortexing for approximately 5 s. Centrifuge the tubes in a microcentrifuge set at +20°C for 5–30 s at 14,000 rpm (20,817 × g). 13. Add 200 ml of MPC Protein Precipitation Reagent to each sample. Mix the samples by vortexing for approximately 5 s. Incubate the samples on ice for 3–5 min. Centrifuge the samples in a microcentrifuge set at +20°C for 10 min at 14,000 rpm (20,817 × g) to pellet the protein precipitate.
214
Liu et al.
14. Label a new 1.5-ml microcentrifuge tube for each sample. Dispense 500 ml of isopropanol (at +4°C) into each tube. Aspirate the supernatant from the sample tube (from the step above) and transfer it to the matching 1.5-ml microcentrifuge tube containing the isopropanol. Invert the tube for approximately 1–2 s to ensure that the supernatant and isopropanol are mixed. Then, mix the sample for approximately 3 min on the Nutator. 15. Centrifuge the samples in a microcentrifuge set at +4°C for 10 min at 14,000 rpm (20,817 × g). The precipitated RNA will form a pellet at the bottom of the tube. Discard the supernatant. 16. Add 1 ml of 75% ethanol (at +4°C) to each sample tube. Invert the sample tubes for 1–2 s to mix. Centrifuge the samples in a microcentrifuge set at +4°C for 2 min at approximately 14,000 rpm (20,817 × g). Discard the ethanol without disturbing the pellet at the bottom of the tube. 17. Repeat the steps for a total of two ethanol washes. Incubate the samples at +37°C for 1–2 min until the pellet is dry. 18. Resuspend the RNA in 30 ml of nuclease-free water. 19. Store the RNA at −20°C for periods of less than 1 day, or at −80°C for periods greater than a day. 3.5. RNA Quantification
There are reports in the literature regarding the differential effects of various fixatives and other variables such as delay until fixation on characteristics of the macromolecules in the fixed tissue (41–44). Both total RNA yield from the extraction and total RNA size distribution were used to evaluate the quality of the RNA recovered from each fixation condition tested: 1. Quantify each RNA sample using the ND-1000 spectrophotometer. The instrument should first be initialized using water and then the background blank reading is taken with the sample diluents, in this case also water. Add 1–1.5 ml of the RNA solution to the sample pedestal and take the optical density 260 nm reading. Calculate the sample concentration as follows: For nucleic acids, the solution’s absorbance at 260 nm is used to calculate the concentration as follows:
C=
A ´e , b
where C is the nucleic acid concentration (in ng/ml), A is the absorbance (in absorbance units, or AU), e is the wavelengthdependent extinction coefficient (in ng cm/ml; 40 for RNA, 50 for dsDNA, 33 for ssDNA), and b is the cm path length. The default NanoDrop path length = 1 mm; adjustable to 2 mm for highly concentrated samples.
RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues
215
2. Alternatively, RNA may be quantified using the Quant-iT™ RiboGreen® RNA Reagent and Kit (Invitrogen, Molecular Probes Inc., Eugene, OR). The kit includes a ribosomal RNA standard used to generate a standard curve against which both high and low concentration RNA extract samples may be quantified. Plates are read on a Spectramax GeminiXS spectrophotometer (Molecular Devices, Menlo Park, CA). Samples can be assayed in triplicate in multiwell plates. The linearity of the RiboGreen assay is maintained in the presence of several compounds commonly found to contaminate nucleic acid preparations, including nucleotides, salts, urea, ethanol, chloroform, detergents, proteins, and agarose. In addition, although the Quant-iT™ RiboGreen® reagent also binds to DNA, pretreatment of mixed samples with DNase can be used to generate an RNA-selective assay. The method outlined below gives an example for assaying 24 samples and one positive control, using a four-point standard curve, in triplicates in a 96-well plate. Prepare 1:200 dilution of the Quant-iT™ RiboGreen® reagent in 10 mM Tris–HCl, 1 mM EDTA, pH 7.5 (1× TE buffer) to make up a total volume of 10 ml, and once prepared keep protected from light by placing in a drawer or covering by foil. Prepare the ribosomal RNA standard curve and ribosomal RNA positive control working solutions. Ribosomal RNA standard solution: Dispense 490 ml of 1× TE Buffer into a 1.5-ml microcentrifuge tube and add 10 ml of ribosomal RNA (100 mg/ml) stock solution into the tube containing the 1× TE to make a working stock solution for the ribosomal RNA standards (2 mg/ml). Ribosomal RNA Positive control: Dispense 8 ml of 1× TE into a 1.5-ml microcentrifuge tube and add 2 ml of Ribosomal RNA (100 mg/ml) stock solution into the tube containing the 1× TE to make a working stock solution for the Ribosomal RNA Positive control (20 mg/ml). Prepare RNA Samples by one of the three methods based on expected yield of samples: For standard assay: 1,000× dilution Dispense 8 ml of 1× TE into a 1.5-ml microcentrifuge tube and add 2 ml of RNA sample. For low-yield samples: no dilution No dilution of RNA sample is required before adding to the 96-well plate. For very-high yield samples: 2,000× dilution Dispense 9 ml of 1× TE into a 1.5-ml microcentrifuge tube and add 1 ml of RNA sample.
216
Liu et al.
Layout of the 96-well RiboGreen Plate. Position the RNA samples, ribosomal RNA standards, and ribosomal RNA positive control on the 96-well plate as illustrated in the figure above. Positions on the plate marked as 1–24, represents each of 24 RNA samples in triplicate, indicating sample 1 in positions A4, A5, A6 and so on till sample 24 in positions H10, H11, H12. Positions on the plate marked as 25 (positions F1, F2, F3), represents the ribosomal RNA positive control. Positions on the plate marked as 26 (columns 1–3 from A to E), represent the ribosomal RNA standards. Add the prepared ribosomal RNA standard (2 mg/ml) and 1× TE Buffer to the designated standard curve wells as follows:
Wells
2 mg/ml rRNA (ml)
1× TE (ml)
Final RNA concentration
A1, A2, A3
100
0
1 mg/ml
B1, B2, B3
50
50
500 ng/ml
C1, C2, C3
10
90
100 ng/ml
D1, D2, D3
2
98
20 ng/ml
E1, E2, E3
0
100
Blank
Dispense 99 ml of 1× TE Buffer to all other wells except for the ribosomal RNA standards. Next, dispense 1 ml of the diluted/ undiluted (as the case may be, refer to the 3 dilution methods for sample preparation) RNA sample to the corresponding positions of each sample in the plate. Add 1 ml of the ribosomal RNA positive control working stock (20 mg/ml) to the corresponding positions (position 25 in the layout above) on the plate. Add 100 ml of the 1:200 RiboGreen reagent to all wells containing RNA samples, the ribosomal RNA Standards, and
RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues
217
ribosomal RNA positive control. Pipette the solution up and down once to mix. Incubate the plate in the dark for 3–5 min. Immediately read the plate on the Spectramax Gemini spectrofluorometer. Samples are excited at 485 ± 10 nm, and the fluorescence emission intensity is measured at 530 ± 12.5 nm using the fluorescence microplate reader. Fluorescence emission intensity is then plotted against the standard ribosomal RNA concentrations as prepared above, and this is used to extrapolate the RNA concentration of the samples after multiplying by the dilution factor, to obtain the actual concentration of the RNA sample. The final concentration of the ribosomal RNA positive control should be in the range of 100 ng/ml. 3.6. RNA Yield and Size Distributions
RNA size distribution was assessed by electrophoresis on 15% polyacrylamide gels (Invitrogen Corporation, Carlsbad, CA) or Agilent 2100 Bioanalyzer using RNA 6000 Nano LabChip® electrophoresis gels (Agilent Technologies, Inc., Santa Clara, CA): 1. Instructions are provided below for electrophoresis of Novex® TBE-Urea Gels using the XCell SureLock™ Mini-Cell. Load only about 200 ng of each RNA sample with RNA size standards (Ambion, Austin, TX) to estimate the average size of total RNA isolated from each tissue preparation. Reagent
Sample (ml)
200 ng RNA sample
x
Loading buffer (2×)
5
Nuclease-free water
(5 − x)
Total volume
10
Prepare the reagent accordingly following the recipe above and heat samples at 70°C for 3 min prior to loading. Prepare 1× TBE running buffer and load samples. Run conditions were 180 V constant, 13 mA for current, and run time was approximately 75–90 min. For staining the gels after electrophoresis, SYBR Gold stain (Invitrogen Corporation, Carlsbad, CA) was used according to the manufacturer’s instructions. 2. Instructions for using the Agilent 2100 Bioanalyzer RNA 6000 Nano LabChip electrophoresis gels are included in the manufacturer’s protocol as described below. Prepare the gel–dye mix by mixing 130 ml of filtered RNA gel matrix with 2 ml of RNA dye concentrate in a RNasefree 1.5-ml microcentrifuge tube. Put a new RNA 6000 chip on the Chip Priming Station. Pipette 9 ml of gel–dye mix in the well marked “G” and close Chip Priming Station. Press
218
Liu et al.
plunger until it is held by the chip. Wait representative of each fixation method for exactly 30 s, then release the chip. Check chip for air bubbles and pipette 9 ml of gel–dye mix in the wells marked G. Load 5 ml of RNA 6000 Nano Marker in all 12 sample wells. Pipette 5 ml of RNA 6000 Nano Marker into the well marked with “ladder” and pipette 1 ml of RNA 6000 ladder. Pipette 1 ml of sample in each of the 12 sample wells. Do not leave any wells empty. Add an additional 1 ml of RNA 6000 Nano Marker in each unused sample well. Put the chip in the adapter and vortex for 1 min at the set point of the IKA vortexer. Run the chip in the Agilent 2100 within 5 min. 3. RNA yield and size distribution observed for differently fixed tissues. Duplicate extractions were prepared from each of four samples chosen to represent each fixation method for a total of eight extracts for each fixative tested. Average RNA yields were calculated for each fixative and normalized to the amount of tissue extracted (micrograms of RNA per cubic millimeters of tissue) (see Fig. 2a). Although essentially equivalent amounts of tissue were processed for each extraction and replicate extractions were done for each block to minimize random process variation, widely different RNA yields were recovered from among differently fixed tissues. Tissues prepared using ethanol, neutral buffered formalin (with or without Zn2+), and Pen-Fix resulted in higher average RNA yields than those prepared using B5, Bouin’s, Prefer, or OCT freezing. The trends in RNA recovery among these samples were paralleled by the trends seen in RNA size distribution (see Fig. 2b). RNA extracted from formalin-fixed, ethanol-fixed, and Pen-Fix processed tissues showed more high-molecularweight species than RNA extracted from B5, Bouin’s, or Preferfixed tissues. All samples, including OCT frozen tissue, show some degree of RNA degradation compared with RNA later processed samples. RNA recoveries tended to parallel the degree of RNA degradation, indicating a more efficient recovery of the higher molecular weight RNA species and a greater loss of the lower-molecular-weight RNA. 4. RNA yield and size distribution observed for formalin-fixed tissues: RNA extracted from tissue blocks all fixed in neutral buffered formalin but using varying conditions showed narrower ranges in extraction yield and size distribution. RNA yield from OCT frozen tissue was very similar to that seen in the first set of experiments (see Fig. 3a). RNA yields from alternatively processed formalin-fixed tissues were within about a twofold range, and the size
RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues
219
Total RNA Yield (µg per mm3 of tissue)
a 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0
NBF
Ethanol
B5
Bouin
Zn-NBF Pen-Fix
Prefer
OCT
b
Fig. 2. (a) Average RNA yield from eight extractions (30 mm of tissue sections for each extraction) expressed as micrograms per cubic millimeters of tissue ± SD from placenta tissue prepared using different fixing agents including neutral buffered formalin (NBF), 75% ethanol, B5 Fixative (B5), Bouin’s Fixative (Bouin), buffered zinc formalin (Zn-NBF), Pen-Fix, Prefer-Fix (Prefer), and OCT frozen. (b) RNA extract (200 ng) and RNA size markers on 15% polyacrylamide gel stained with SYBR® Gold. M1, RNA 6000 Ladder; M2, RNA Century™ Marker; other lanes labeled by fixing agents.
distributions seen in the Agilent BioAnalyzer images are all very similar (see Fig. 3b). 3.7. Gene Selection and Assay Design for Quantitative RT-PCR
1. For each gene candidate to be used in the study, identify the appropriate mRNA reference sequence using Entrez Gene and the correct accession number/sequence ID for the gene of interest. Access and download the consensus sequence through the NCBI Entrez nucleotide database. When selecting candidate genes, both biomarker and reference normalization genes should be considered. 2. Design RT-PCR primers and probes using any assay design method: Primer Express® (Applied Biosystems, Foster City, CA) or the publicly available Primer3 (http://www.frodo. wi.mit.edu/primer3/input.htm) both generally work well (45) (see Note 4).
220
Liu et al.
Total RNA Yield (µg per mm3 of tissue)
a 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0
b
OCT fresh frozen M 1
2
6-8 hr NBF
3
4
5
6
48-72 hr NBF 7
3-4 hr NBF
1.5-2 hr NBF
1.5-2 hr NBF& Rev Proc
6-8 hr NBF (extra small)
8
bases
4000
2000 1000 500 200
Fig. 3. (a) Average total RNA yield from five extractions (30 mm of tissue sections for each extraction) per formalin fixation condition expressed as micrograms per cubic millimeters of tissue ± SD. (b) RNA extract (200 ng/lane) and RNA size marker analyzed on Agilent 2100 Bioanalyzer using RNA 6000 Nano LabChip®. M, RNA Century Marker; (1) RNAlater; (2) OCT frozen; (3) 6–8 h NBF; (4) 48–72 h NBF; (5) 3–4 h NBF; (6) 1.5–2 h NBF; (7) 1.5–2 h NBF with reverse processing; (8) 6–8 h NBF with extra small tissue specimen. The gel image combines representative lanes from several LabChip gels run using the same conditions.
3. Order the TaqMan assays, i.e., the oligonucleotide primers and dual-labeled TaqMan probes, with 5¢-FAM as a reporter and a 3¢ quencher molecule (see Note 5), from a commercial oligonucleotide supplier.
RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues
3.8. Reverse Transcription
221
Reverse transcription (RT) is done in a 96-well plate as a batch process scalable from one sample up to 94 samples along with one positive control sample and one negative control sample. Any RNA source known to express the candidate genes to be used in the study can act as a positive control. Commercially available standard RNA mixtures made from cell-line extracts such as the Universal Human Reference RNA (Stratagene, San Diego, CA) can serve the purpose of a positive control. The negative control uses water in place of the RNA template and tests for laboratory contamination of reagents with target sequences: 1. Create the gene-specific primer (GSP) pool for specifically priming the genes of interest during the RT reaction or, alternatively, use random primers. Random hexamers up to random nonamers have been shown to be successful in random primed RT. 2. To prepare a GSP pool, remove the appropriate reverse PCR primer stock tubes from −20°C storage. Allow the tubes to remain at room temperature until fully thawed, then vortex the reagent tubes for approximately 5 s until thoroughly mixed. Centrifuge in a microcentrifuge set at +4°C for 5–30 s at no more than 14,000 rpm (20,817 × g). Store the thawed tubes on ice while in use. Using the concentration on the reverse primer vendor certificate of analysis, calculate the volume of reverse primer to be added using the following equation: (see Note 6).
C1V1 = C 2V 2 , where C1 is starting concentration of each reverse primer stock solution, V1 is volume of each reverse primer stock solution to be added, C2 is final concentration of each reverse primer (1 mM), V2 is total volume of the GSP pool. Example: To create 1 ml of the GSP pool (at 1 mM final concentration for each primer) with reverse primer each having a starting concentration of 200 mM:
(200mM )(V1 ) = (1mM )´ (1, 000ml ), é(1mM )´ (1, 000 ml )ùû V1 = ë , [200mM ]
V1 = 5ml.
Calculate the volume of nuclease-free water to be added using the following equation:
Vol. of nuclease ¯ free water = Total vol. of GSP pool (No. of reverse primers ´ vol. of each reverse primer).
222
Liu et al.
GSP priming pool preparation Number of genes
Total pool volume
Vol. per reverse primer
Nucleasefree water
24
1,000 ml
5 ml
880 ml
Dispense the required volume of nuclease-free water into a microcentrifuge tube. Pipette the required volume of each reverse primer into the tube containing the water. Vortex the tube for approximately 5 s until thoroughly mixed. Centrifuge for approximately 5 s to remove any reagent from the tube lid. 3. Obtain the following reagents from −20°C storage: Omniscript RT Kit, including: dNTP Mix (5 mM), 10× RT Buffer, RNase-free Water. 4. Allow the reagents listed above to thaw at room temperature until no crystals are visible. Vortex all the tubes containing 10× buffer, dNTP mix, RNase-free water, and GSP pool for approximately 5 s at a medium-high speed until thoroughly mixed. Shake down or spin down each tube to remove condensation from the tube lid, if necessary. Store the reagents on ice if the reagents will not be used immediately after it is thawed. 5. Calculate number of RT reactions to be run as follows:
No. of RT reactions = (No. of samples + 1 positive control + 1 negative control) + 5 sample reactions as overage. Based on the number of RT reactions, use the table below to c alculate the volume of each reagent required to make the Master Mix.
Reagents
1 reaction (ml/ reaction)
94 samples (101 reactions including 5 sample overage)
10× Buffer RT
4
404
dNTP mix, 5 mM each dNTP
4
404
ABI Random Hexamer, 50 mM
1
101
GSP pool, 1 mM
2
202
ABI RNase inhibitor, 20 U/ml
1
101
Omniscript RT, 4 U/ml
1
101
Nuclease-free water
3
303
16
1,616
Total volume
RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues
223
Master Mix preparation for RT reaction Master Mix/reaction
16
RNA ± water
24
Total volume
40
6. Pipette the reagents listed above into a labeled tube. Return the component reagents to the −20°C storage after use. Vortex the Master Mix tube at medium-high speed for approximately 2–5 s until thoroughly mixed. Shake down or centrifuge the tubes in the minicentrifuge for approximately 5 s to remove any reagent from the tube lids. Place the Master Mix tube on ice. 7. Obtain the sample RNA and positive RT control (e.g., Universal Human Reference RNA; genomic DNA-free at concentration of 100 ng/ml) from −20°C storage. Thaw the sample RNA and positive control RNA completely. Flick the tubes for 2–5 s to mix the RNA. Centrifuge the tubes in a microcentrifuge set at +4°C for 5–30 s at 14,000 rpm (20,817 × g). 8. Calculate the amount of RNA to use in each reaction using the table below. The positive RT control will be the same RNA input as the sample RNA. The standard qPCR assay will be 10 ml with the equivalent of 2 ng of cDNA per reaction well. RNA required = [(No. of genes × No. of replicate wells) + overage wells] × 2 ng/well. The RNA must be at a concentration that the required amount is contained in 24 ml, since the standard reaction volume is 40 and 16 ml of reagent mix will be added. Sample RNA requirements for duplicate wells of 10 ml qPCR No. of genes
No. of replicates
Overage wells
RNA required
Volume
24
2
22
140 ng/sample
24 ml
The qPCR assay can be reduced to 5 ml with the equivalent of 1 ng of cDNA per reaction well. RNA required = [(No. of genes × No. of replicate wells) + overage wells] × 1 ng/well. Sample RNA requirements for triplicate wells of 5 ml qPCR No. of genes
No. of replicates
Overage wells
RNA required
Volume
24
3
28
100 ng/sample
24 ml
Note that the overage wells could be scaled down if manual process is used for qPCR plate assembly. 9. Assemble the RT reactions in a hard-shell thin-wall 96-well microplate or strip tubes (see Note 7).
224
Liu et al.
10. To assemble the positive control for duplicate wells in 10-ml qPCR assay, dispense 22.6 ml of nuclease-free water into the positive control well or tube. Dispense 1.4 ml of the positive control RNA (at 100 ng/ml) into the positive control well. 11. To assemble the negative control (no-template control), dispense 24 ml of nuclease-free water into the negative control well. 12. To assemble sample RT reactions, dispense appropriate volume of nuclease-free water into the sample well or tube. Dispense appropriate volume of each normalized RNA sample into a reaction well. Repeat the steps for the remaining RNA samples (see Note 8). 13. Add 16 ml of RT Master Mix to each RNA sample well. Total reaction volume for each reaction is 40 ml. Mix the contents of the plate or strip tubes. When using plates, pipette the solution up and down two to three times for well mixing. When using strip tubes, flick the tubes for 2–5 s for well mixing. Seal plates with polar seal film or cap strip tubes with the attached caps. Centrifuge the plate in a plate centrifuge set at +10°C for approximately 2 min at approximately 2,000 rpm (827 × g) or pulse-spin the strip tubes in a mini centrifuge for approximately 5 s (see Note 9). 14. Thermal-cycle the RT reactions at +37°C for 60 min followed by +93°C for 5 min, followed by +10°C until the instrument is stopped. Reaction volume is 40 ml. 15. If the samples will be processed through quantitative PCR the same day, store them on ice. If the samples will be processed at a later time, store the samples at −20°C. 3.9. Quantitative RT-PCR Assay
The Real-Time PCR method used was the TaqMan method run on the ABI PRISM 7900HT real-time detection system (Applied Biosystems, Foster City, CA). Gene expression was analyzed using TaqMan® quantitative RT-PCR as described below. RT was performed using the Omniscript™ RT kit (Qiagen, Valencia, CA) in a final volume of 40 ml. Quantitative PCR (qPCR) reactions were performed in 384-well plates using PRISM® 7900HT instruments (Applied Biosystems, Foster City, CA). Expression for 24 genes was measured in triplicate 5-ml reactions using cDNA derived from 1 ng of total RNA per reaction well or in duplicate 10 ml reactions using cDNA derived from 2 ng of total RNA per reaction well (see Table 2). Preparation of the qPCR Assays: PCR assay reagent preparation includes combining the Master Mix Reagent–cDNA mixture with the primer and probes mixtures. The protocol given here assumes 10 ml reactions for 24 genes with an equivalent of 2 ng of RNA (cDNA) per reaction run in a 384-well plate. Nevertheless, this protocol is very scalable and can be adapted to different numbers of candidate genes,
RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues
225
Table 2 List of genes used for expression profiling in these studies Gene
Official symbol
Accession ID
Gene
Official symbol
Accession ID
B-actin
ACTBa
NM_001101
HER2
ERBB2
NM_004448
BAG1
BAG1
NM_004323
HIF1A
HIF1A
NM_001530
CD68
CD68
NM_001251
IGFBP2
IGFBP2
NM_000597
cMet
MET
NM_000245
IL6
IL6
NM_000600
COX2
PTGS2
NM_000963
ITGA7
ITGA7
NM_002206
CYP
PPIH
NM_006347
p27
CDKN1B
NM_004064
DPYD
DPYD
NM_000110
RPLPO
RPLP0a
NM_001002
DR5
TNFRSF10B
NM_003842
TBP
TBP
NM_003194
EGFR
EGFR
NM_005228
TFRC
TFRC
EREG
EREG
NM_001432
TIMP1
TIMP1
NM_003254
GAPDH
GAPDHa
NM_002046
TUBB
TUBB2A
NM_001069
GUS
GUSB
NM_000181
WISP1
WISP1
NM_003882
a
a
NM_003234
ACTB, GAPDH, GUSB, RPLP0, and TFRC were included as reference genes
a
modified for reactions run at different volumes, or reactions run with different amounts of RNA in each reaction. Similarly, the process can be adapted for using 96-well reaction plates rather than 384-well reaction plates. One qPCR Master Mix is prepared for each of the following: each test sample, the RT positive control sample, and the RT negative (no-template) control sample. In this process, 2× TaqMan® Universal PCR Master Mix solution is combined with a cDNA sample or nuclease-free water. All steps described here can be done using multichannel pipettors or with automated liquid-handling robots such as the Tecan Aquarius robot or Beckman Biomek workstation. 1. Preparation of the primer and probe (P3) pools Thaw the primer and probe stock reagent tubes completely. Vortex the primer and probe tubes for approximately 30 s until thoroughly mixed. Centrifuge at +4°C for 5–30 s at no more than 14,000 rpm (20,817 × g). While in use, store the thawed primer and probe tubes on ice. Obtain an amber microcentrifuge tube, centrifuge tube, or other equivalent container and label it with the primer and probe set name. Using the concentration provided on the certificate of analysis for the primers, calculate the volume of each primer to be added as follows (see Note 10): Volume of forward or reverse primer = [(Final concentration of primer) × (Total volume of P3 pool)]/Initial concentration of primer.
226
Liu et al.
Example: To create 600 ml of the P3 pool with each forward and reverse primers having a starting and final concentration of 200 and 4.5 mM, respectively: Volume of forward or reverse primers = [(4.5 mM) × (Total volume of P3 pool)]/200 mM. Using the concentration provided on the certificate of analysis for the probes, calculate the volume of probe to be added as follows (see Note 11): Volume of probe = [(Final concentration of probe) × (Total volume of P3 pool)]/Initial concentration of probe. Example: To create 600 ml of the P3 pool with each probe having a starting and final concentration of 100 and 1 mM, respectively: Volume of probe = [(1 mM) × (Total volume of P3 pool)]/100 mM. Determine the volume of nuclease-free water as follows: Volume of nuclease-free water = (Total vol. of P3 mix) − [(Vol. of forward Primer) + (Vol. of reverse Primer) + (Vol. of probe)].
Reagent
Volume (ml)
Volume of forward Primer
13.5
Volume of reverse Primer
13.5
Volume of probe
6
Volume of nuclease-free water
567
Total volume
600
Calculation for preparing 600 ml of P3 pool with Primer and probe final concentration at 4.5 and 1 mM, respectively. Pipette the calculated volumes of the required reagents into a microcentrifuge tube. Vortex the P3 pool for approximately 30 s until thoroughly mixed. Centrifuge in a microcentrifuge set at +4°C for 5–30 s at no more than 14,000 rpm (20,817 × g). Store the P3 pool tube on ice while in use. Optional: If using automated liquid-handling robots with 96-channel head for qPCR plate assembly, prepare P3 plates in 96-well plate format by distributing the P3 pools into predetermined well position (see table below for an example of P3 plate layout of a 24-gene panel). Seal the P3 plates with aluminum foil seals (e.g., PolarSeal Foil Sealing Tape, E&K Scientific) and store at +4°C for using within 3 days. Otherwise, store the P3 plates at −20°C until ready for qPCR plate assembly.
RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues
227
P3 plate layout for a 24-gene panel 1
2
3
4
5
6
A Gene 1 Gene 9 Gene 17 Gene 1 Gene 9
7
8
Gene 17 Gene 1 Gene 9
9
10
11
Gene 17 Gene 1 Gene 9
12 Gene 17
B Gene 2 Gene 10 Gene 18 Gene 2 Gene 10 Gene 18 Gene 2 Gene 10 Gene 18 Gene 2 Gene 10 Gene 18 C Gene 3 Gene 11 Gene 19 Gene 3 Gene 11 Gene 19 Gene 3 Gene 11 Gene 19 Gene 3 Gene 11 Gene 19 D Gene 4 Gene 12 Gene 20 Gene 4 Gene 12 Gene 20 Gene 4 Gene 12 Gene 20 Gene 4 Gene 12 Gene 20 E Gene 5 Gene 13 Gene 21 Gene 5 Gene 13 Gene 21 Gene 5 Gene 13 Gene 21 Gene 5 Gene 13 Gene 21 F Gene 6 Gene 14 Gene 22 Gene 6 Gene 14 Gene 22 Gene 6 Gene 14 Gene 22 Gene 6 Gene 14 Gene 22 G Gene 7 Gene 15 Gene 23 Gene 7 Gene 15 Gene 23 Gene 7 Gene 15 Gene 23 Gene 7 Gene 15 Gene 23 H Gene 8 Gene 16 Gene 24 Gene 8 Gene 16 Gene 24 Gene 8 Gene 16 Gene 24 Gene 8 Gene 16 Gene 24
2. Preparation of qPCR Master Mix for test samples and RT positive and negative controls: Obtain a bottle of 2× TaqMan Universal PCR Master Mix reagent from +4°C storage. Store the 2× TaqMan Universal PCR Master Mix on ice, protected from light, when in use. Obtain the samples from +4°C storage or −20°C storage. If stored in the −20°C, store the samples on ice until thawed. If the samples are in plates, spin down using the plate centrifuge set at +10°C for approximately 2 min at approximately 2,000 rpm (827 × g). Centrifuge strip tubes in the minicentrifuge for approximately 5 s to remove any reagent from the tube lid. Store the thawed sample on ice while in use. Obtain an appropriate tube for preparing the sample Master Mix for one sample to be tested for 24 genes in duplicate wells (140 ng for RNA input in RT reaction; 22 wells to allow for an overage) of 10 ml reactions. Pipette the reagents listed in the table, qPCR Master Mix preparation for one sample. qPCR Master Mix preparation for one sample for 10 ml qPCR assay in duplicate wells Reagent
Volume (ml)
2× Universal PCR Master Mix
350
RT reaction mix (sample)
40
Nuclease-free water
170
Total volume
560
Obtain an appropriate tube for preparing the sample Master Mix for one sample to be tested for 24 genes in triplicate wells (100 ng for RNA input in RT reaction; 28 wells to allow for an overage) of 5 ml reactions. Pipette the reagents listed in the table, qPCR Master Mix preparation for one Sample.
228
Liu et al.
qPCR Master Mix preparation for one sample for 5 ml qPCR assay in triplicate wells Reagent
Volume (ml)
2× Universal PCR Master Mix
250
RT reaction mix (sample)
40
Nuclease-free water
110
Total Volume
400
Cap the Master Mix tube and invert for 2–5 s until fully mixed. Centrifuge the tube for approximately 5 s to remove any reagent from the lid, if necessary. Place the Master Mix tube on ice. To prepare the Master Mix containing the RT Positive and Negative control samples, label an appropriate tube for each and prepare each exactly as a test sample above (see Note 12). Optional: If using automated liquid-handling robots with 96-channel head for qPCR plate assembly, arrange the qPCR Master Mix in a 96-well plate format by distributing the qPCR Master Mix containing test or control samples into predetermined well position (Note: for 10-ml qPCR assay in duplicate and 5-ml qPCR assay in triplicates, dispensing no less than 20 and 15 ml, respectively, of the qPCR Master Mix in each of the 24 wells for a sample). See below for an example of a Sample Plate layout for a 24-gene panel (refer to the above step for the corresponding P3 Plate layout): Sample plate layout for testing a 24-gene panel 1
2
3
4
5
6
7
8
9
10
11
12
A Sample 1 Sample 1 Sample 1 Sample 2 Sample 2 Sample 2 Sample 3 Sample 3 Sample 3 Sample 4 Sample 4 Sample 4 B Sample 1 Sample 1 Sample 1 Sample 2 Sample 2 Sample 2 Sample 3 Sample 3 Sample 3 Sample 4 Sample 4 Sample 4 C Sample 1 Sample 1 Sample 1 Sample 2 Sample 2 Sample 2 Sample 3 Sample 3 Sample 3 Sample 4 Sample 4 Sample 4 D Sample 1 Sample 1 Sample 1 Sample 2 Sample 2 Sample 2 Sample 3 Sample 3 Sample 3 Sample 4 Sample 4 Sample 4 E Sample 1 Sample 1 Sample 1 Sample 2 Sample 2 Sample 2 Sample 3 Sample 3 Sample 3 Sample 4 Sample 4 Sample 4 F Sample 1 Sample 1 Sample 1 Sample 2 Sample 2 Sample 2 Sample 3 Sample 3 Sample 3 Sample 4 Sample 4 Sample 4 G Sample 1 Sample 1 Sample 1 Sample 2 Sample 2 Sample 2 Sample 3 Sample 3 Sample 3 Sample 4 Sample 4 Sample 4 H Sample 1 Sample 1 Sample 1 Sample 2 Sample 2 Sample 2 Sample 3 Sample 3 Sample 3 Sample 4 Sample 4 Sample 4
3. qPCR plate assembly To set up qPCR assays, a plate layout should be predetermined where each P3 pool is deposited into corresponding wells. Then, combine the P3 pool and qPCR Master Mix containing test or control samples as shown below in 96-well or 384-well plates:
RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues
229
qPCR volume (ml reaction)
P3 pool (ml)
qPCR Master Mix (ml)
10
2
8
5
1
4
The P3 plate and sample plate described above allow for qPCR plate assembly using robotic system with a 96-channel head. For the 10-ml assay in duplicate wells, a total of eight samples can be assembled in a 384-well qPCR plate using quadrant 1 and 2 for four samples (sample plate no. 1) and quadrant 3 and 4 for another four samples (sample plate no. 2). For the 5-ml assay in triplicate wells, a total of four samples (sample plate no. 1) can be assembled in a 384-well qPCR plate using quadrants 1–3. Hand-seal the plate with optical adhesive film (e.g., ABI Prism Optical Adhesive Covers). Spin down the plate in a plate centrifuge set at +10°C for approximately 2 min at approximately 2,000 rpm. Ensure that no air bubbles are trapped at the bottom of the plate. If air bubbles are seen, spin down the plate in the plate centrifuge set at +10°C for approximately 2 min at 2,000 rpm (827 × g). Repeat until no air bubbles are trapped at the bottom of the plate. Run the Plate on the PRISM 7900 HT Sequence Detection System using the following selections: Assay: Absolute Quantitation; Container: 384 Wells Clear Plate; Template: TaqMan 10 ml; Cycling Conditions: +95°C, 10 min; (+95°C, 20 s, +60°C, 45 s) for 40 cycles; Assay volume: 10 ml (for the 10 ml reaction) or 5 ml (for the 5 ml reaction). 3.10. Data Analysis
For each of the 24 genes, eight RNA extracts (4 FPET blocks x 2 RNA extracts per block) were assayed in duplicate wells for each fixative for a total of 384 assay wells for each condition. RNA extracts from the same block were assayed on the same qPCR plate. Samples were randomized on the qPCR plates to avoid any systematic process bias. A random effects model was used to separate total variance in nonnormalized and reference-normalized expression (CT) measurements for each gene at each condition into components of variance among replicate blocks, among replicate RNA extract tubes within a block, and among replicate assay wells. Unsupervised cluster and principal component analyses were performed to further measure variance and correlation structures of gene expression profiles among conditions, blocks within a condition, and tubes within a block. For the study comparing RNA from formalin-fixed samples processed using varying fixation conditions, each gene was assayed in triplicate from five replicate samples from each process for a total of 15 data points for each condition. No formal statistical comparisons were done among the fixation conditions; however, correlative analyses and descriptive statistics of gene expression were done for each set of conditions.
230
Liu et al.
3.10.1. Quantitative RT-PCR Results for Different Fixatives
Expression analysis using quantitative RT-PCR assays for 24 genes was done in duplicate for each of the eight RNA extracts prepared from each FPET block for each fixative tested. Expression values are given on the CT (threshold cycle) scale where one unit of change reflects approximately a twofold change in expression. Expression distribution ranges, as well as median and mean values with interquartile ranges and outliers were calculated for each fixation condition investigated with the exception of the Bouin’s processed samples (see Fig. 4a). RNA prepared from tissues preserved in Bouin’s fixative was excluded from further analysis since RT-PCR for these samples resulted in average CT values near the limit of quantitation for the assay (mean = 37.2, SD = 3.00) and many data points fell outside of the quantitative range of the assay. The lowest mean raw CT value (28.2) for the 24 test genes was seen in the RNAlater™ preserved tissue which is consistent with the high yield and high molecular weight of these samples. The mean CT values obtained with RNA extracted from tissues preserved using OCT freezing, Pen-Fix, neutral buffered formalin, ethanol, and buffered zinc formalin were slightly higher and very similar: 28.6, 28.9, 29.7, 29.7, and 30.1, respectively. RNA extracted from tissue sections preserved in B5 and Prefer-Fix fixative showed somewhat poorer RT-PCR performance with mean raw CT values for the 24 genes of 31.7 and 33.0, roughly a four- to eightfold loss in raw signal.
3.10.2. Quantitative RT-PCR Results for Tissues Fixed in Formalin Under Varying Conditions
RT-PCR performance was also assessed for RNA isolated from the tissues fixed in neutral buffered formalin using variations in fixation protocol. In this case, RT-PCR was done using 22 of the 24 genes listed in Table 2 (HIF1A and TUBB were not included). The distribution of the mean raw CT values for the 22 test genes was strikingly narrow among these formalin-fixed tissues, ranging between 30.1 and 30.8 (see Fig. 5a) and was very similar to the values obtained for the formalin-fixed tissues in the prior placenta experiment (29.7 and 30.1) (see Fig. 4a). However, for tissue preserved in RNAlater™ and by the OCT freezing process, increases in mean CT value of 2.0 and 1.3, respectively, were observed. Since these studies were carried out on two different placentas, some of the study-to-study variation in average raw CT values between RNAlater™ and OCT frozen samples may be explained by different expression levels for the study genes between the two placentas.
3.10.3. Reference Gene Normalization Corrects for Fixation Differences
A set of reference genes, ACTB, GAPDH, GUSB, RPLP0, and TFRC, was included in the gene panel. The average of the CT values for these five reference genes was used to normalize the raw CT values obtained from the differently fixed tissues. The results show that normalization very effectively corrects for the variances in CT values that result from using different fixation reagents (see Fig. 4b). Although the average CT values were quite similar among
RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues
a
231
40 38 36 34
CT Value
32 30 28 26 24 22 20 RNAlater
NBF
Ethanol
B5
Zn–NBF
Pen–Fix
Prefer
OCT
Pen–Fix
Prefer
OCT
Fixation Method
b 17.5
Reference Normalized CT Value
15.0
12.5
10.0
7.5
5.0
2.5
0 RNAlater
NBF
Ethanol
B5
Zn–NBF
Fixation Method
Fig. 4. Distribution of raw CT values (a) and reference-normalized CT values (b) for 24 genes analyzed in RNA isolated from human placenta tissue processed with different fixing agents. Both median (−) and mean (+) CT values for 384 wells of data are displayed in the box plot for each fixation condition. The box itself contained the middle 50% of the data. The upper and lower edge of the box indicated the 75th and 25th percentile of the dataset. The ends of the vertical lines indicated the minimum and maximum data values, unless outliers are present in which case the lines extend to a maximum of 1.5 times the interquartile range. For raw CT value, the higher the CT value, the lower the expression level detected; for reference-normalized CT value, the higher the CT value, the higher the expression level detected.
232
Liu et al.
all of the samples fixed in neutral buffered formalin, reference gene normalization still corrected for the small differences that did occur with the variations in sample processing (see Fig. 5b). In the studies described here, the only samples where interpretable RT-PCR expression profile results could not be obtained were the RNA samples isolated from Bouin’s fixed tissue. This set of samples showed the overall poorest RNA extraction yield and greatest degree of RNA fragmentation of all the samples evaluated in both studies. As RNA fragmentation increases, the efficiency of RNA extraction recovery decreases, and the average raw CT values of the RT-PCR assays increase, causing an apparent decrease in gene expression. However, this effect is corrected for by reference gene normalization since all of the genes are affected to the same degree. Therefore, it is evident that gene expression profiling can be done successfully in fixed tissues up to the point where RNA fragmentation diminishes the number of detectable template molecules below the assay limit of quantitation. Furthermore, it can be inferred that gene expression information is faithfully preserved in fixed tissues and that with reference gene normalization; results are comparable among samples despite differences in fixation processing methods used and the resulting RNA fragmentation.
4. Notes 1. It is important to store and handle the components of the Epicentre MasterPure Kit appropriately. It contains 1× Tissue and Cell Lysis Solution, Proteinase K (50 mg/ml), MPC Protein Precipitation Reagent, RNase-free DNase I enzyme (1 U/ml), 1× DNase I buffer, 2× Tissue and Cell Lysis Solution. Proteinase K and DNase I enzymes must be kept at −20°C. Store the remaining components at +18 to +25°C. Before using any buffers in the kit, ensure that they have been completely thawed (i.e., no particles or crystals are visible in the solutions) and are well mixed. 2. Pathology fixation procedures are widely variable depending on tissue type and each laboratory’s optimized practices. As shown here, RNA can be successfully extracted from nearly all types and ages of fixed tissues; however, yield will vary with sample quality and age. The quality of the extracted RNA (degree of fragmentation) will also vary with sample age and fixation method. If tissue is available, a test extraction can be very helpful for estimating how much RNA to expect from a sample.
RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues
a
233
40 38 36 34
CT Value
32 30 28 26 24 22 20 RNAlater
OCT
6–8 hr
48–72 3–4 hr hr Fixation Protocol
1.5–2 hr
1.5–2 hr 6–8 hr (Rev. P) (extra small)
RNAlater
OCT
6–8 hr
48–72 3–4 hr hr Fixation Protocol
1.5–2 hr
1.5–2 hr 6–8 hr (Rev. P) (extra small)
b 17.5
Reference Normalized CT Value
15.0
12.5
10.0
7.5
5.0
2.5
0
Fig. 5. Distribution of raw CT values (a) and reference-normalized CT values (b) for the 22 genes analyzed in RNA isolated from human placenta tissue processed by different neutral buffered formalin fixing protocols. Both median (−) and mean (+) CT values are displayed in the box plots. Refer to Fig. 4 for box plot interpretation and the relationship between raw or normalized CT value and expression level.
234
Liu et al.
3. It is very important for successful TRIzol® nucleic acid extraction not to exceed the recommended tissue-to-volume ratio recommended by the manufacturer’s package insert. 4. Amplicon sizes should be constrained to less than 100 bases total in length, including primer sequences. This design parameter should be adjusted in the Primer Express or Primer3 settings since the default usually permits amplicons as long as 300 bases to be designed. It is generally easier to get a design that meets all necessary criteria by not demanding the amplicon span an exon junction. DNase treatment of the RNA extracts and inclusion of the no-template controls eliminate the necessity to design RNA specific assays. 5. The ABI PRISM 7900HT system is capable of detecting a number of different fluorescent reporter dyes, but the fluorescein-based dye FAM is the most robust and widely available. There are also multiple different quencher molecules that can be paired with the FAM dye on the dual-labeled probes. Although TAMARA dye was often used in the original form of the TaqMan assay, the newer “dark” (nonfluorescent) quenchers give much better assay results due to better quenching and lower assay backgrounds. Commercial oligonucleotide vendors offer a variety of TaqMan probe configurations. 6. The concentration of each reverse primer stock solution is typically 100 or 200 mM. The reverse primers are mixed with nuclease-free water to bring the final concentration of each reverse primer in the GSP pool to 1 mM. 7. Keep all RT reaction components cold throughout the qPCR assembly process, including the plate or strip tubes by keeping them on ice, in a cooler, or equivalent. 8. Assembling the RT reactions can be done more easily if at the time each RNA sample is quantified, it is also normalized to a standard concentration of 200 ng/ml using nuclease-free water. Alternatively, each RT reaction can be individually calculated and the final volumes normalized with nuclease-free water to compensate for any differences in added RNA volume. 9. When carrying out temperature-sensitive reactions in microtiter plates, it is very important to ensure that no air bubbles are trapped at the bottom of the plate so that enzymatic mixtures are exposed uniformly to correct temperature conditions. If air bubbles are seen, tap the plate to dislodge the bubbles. Spin down the plate in the plate centrifuge set at +10°C for approximately 2 min at approximately 2,000 rpm (827 × g). Repeat the previous step until no air bubbles are trapped at the bottom of the plate. 10. The concentration of each forward and reverse primer stock solution is typically 100 or 200 mM. The primers are mixed with
RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues
235
probe and nuclease-free water to bring the final concentration of the primer in the P3 pool to 4.5 mM. Final primer concentrations in the PCR are 900 nM. 11. The concentration of each probe stock solution is typically 100 mM. The probes are mixed with primer and nuclease-free water to bring the final concentration of the probe in the P3 pool to 1 mM. The final PCR probe concentration is 200 nM. 12. This section describes the steps used to prepare Master Mix for one sample. One Master Mix is prepared for each cDNA test sample, the RT positive control sample, and the RT negative control sample. It is recommended that the RT positive and negative samples be assembled and processed through PCR prior to assembling the test sample PCR plates to ensure that the RT reaction was successful and that no contaminations occurred. References 1. Rupp, G.M. and Locker, J. (1988) Purification and analysis of RNA from paraffin-embedded tissues. Biotechniques 6, 56–60. 2. Finke, J., Fritzen, R., Ternes, P., Lange, W., and Dolken, G. (1993) An improved strategy and a useful housekeeping gene for RNA analysis from formalin-fixed, paraffin-embedded tissues by PCR. Biotechniques 14, 448–453. 3. Mies, C. (1994) Molecular biological analysis of paraffin-embedded tissues. Hum. Pathol. 25, 555–560. 4. Krafft, A.E., Duncan, B.W., Bijwaard, K.E., Taubenberger, J.K., and Lichy, J.H. (1997) Optimization of the isolation and amplification of RNA from formalin-fixed, paraffinembedded tissue: the armed forces institute of pathology experience and literature review. Mol. Diagn. 2, 217–230. 5. Stanta, G. and Bonin, S. (1998) RNA quantitative analysis from fixed and paraffinembedded tissues: membrane hybridization and capillary electrophoresis. Biotechniques 24, 271–276. 6. Chang, J., Powles, T.J., Allred, D.C., Ashley, S.E., Clark, G.M., Makris, A., et al. (1999) Biologic markers as predictors of clinical outcome from systemic therapy for primary operable breast cancer. J. Clin. Oncol. 17, 3058–3063. 7. Sheils, O.M. and Sweeney, E.C. (1999) TSH receptor status of thyroid neoplasms-TaqMan RT-PCR analysis of archival material. J. Pathol. 188, 87–92. 8. Godfrey, T.E., Kim, S.H., Chavira, M., Ruff, D.W., Warren, R.S., Gray, J.W., et al. (2000)
9.
10.
11.
12.
13.
Quantitative mRNA expression analysis from formalin-fixed, paraffin-embedded tissues using 5¢ nuclease quantitative reverse transcription-polymerase chain reaction. J. Mol. Diagn. 2, 84–91. Specht, K., Richter, T., Muller, U., Walch, A., Werner, M., and Hofler, H. (2001) Quantitative gene expression analysis in microdissected archival formalin-fixed and paraffin-embedded tumor tissue. Am. J. Pathol. 158, 419–429. Abrahamsen, H.N., Steiniche, T., Nexo, E., Hamilton-Dutoit, S.J., and Sorensen, B.S. (2003) Towards quantitative mRNA analysis in paraffin-embedded tissues using real-time reverse transcriptase-polymerase chain reaction: a methodological study on lymph nodes from melanoma patients. J. Mol. Diagn. 5, 34–41. Cronin, M., Pho, M., Dutta, D., Stephans, J.C., Shak, S., Kiefer, M.C., et al. (2004) Measurement of gene expression in archival paraffin-embedded tissues: development and performance of a 92-gene reverse transcriptase-polymerase chain reaction assay. Am. J. Pathol. 164, 35–42. Haque, T., Faury, D., Albrecht, S., LopezAguilar, E., Hauser, P., Garami, M., et al. (2007) Gene expression profiling from formalin-fixed paraffin-embedded tumors of pediatric glioblastoma. Clin. Cancer Res. 13, 6284–6292. Linton, K.M., Hey, Y., Saunders, E., Jeziorska, M., Denton, J., Wilson, C.L., et al. (2008) Acquisition of biologically relevant gene
236
Liu et al.
expression data by Affymetrix microarray analysis of archival formalin-fixed paraffinembedded tumours. Br. J. Cancer 98, 1403–1414. 14. Oberli, A., Popovici, V., Delorenzi, M., Baltzer, A., Antonov, J., Matthey, S., et al. (2008) Expression profiling with RNA from formalin-fixed, paraffin-embedded material. BMC Med. Genomics 1, 1–15. 15. Golub, T.R., Slonim, D.K., Tamayo, P., Huard, C., Gaasenbeek, M., Mesirov, J.P., et al. (1999) Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science 286, 531–537. 16. Sorlie, T., Perou, C.M., Tibshirani, R., Aas, T., Geisler, S., Johnsen, H., et al. (2001) Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc. Natl Acad. Sci. U. S. A. 98, 10869–10874. 17. van de Vijver, M.J., He, Y.D., van’t Veer, L.J., Dai, H., Hart, A.A., Voskuil, D.W., et al. (2002) A gene-expression signature as a predictor of survival in breast cancer. N. Engl. J. Med. 347, 1999–2009. 18. Ramaswamy, S., Ross, K.N., Lander, E.S., and Golub, T.R. (2003) A molecular signature of metastasis in primary solid tumors. Nat.Genet. 33, 49–54. 19. Rosenwald, A., Wright, G., Wiestner, A., Chan, W.C., Connors, J.M., Campo, E., et al. (2003) The proliferation gene expression signature is a quantitative integrator of oncogenic events that predicts survival in mantle cell lymphoma. Cancer Cell 3, 185–197. 20. Paik, S., Shak, S., Tang, G., Kim, C., Baker, J., Cronin, M., et al. (2004) A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer. N. Engl. J. Med. 351, 2817–2826. 21. Gianni, L., Zambetti, M., Clark, K., Baker, J., Cronin, M., Wu, J., et al. (2005) Gene expression profiles in paraffin-embedded core biopsy tissue predict response to chemotherapy in women with locally advanced breast cancer. J. Clin. Oncol. 23, 7265–7277. 22. Cobleigh, M.A., Tabesh, B., Bitterman, P., Baker, J., Cronin, M., Liu, M.-L., et al. (2005) Tumor gene expression and prognosis in breast cancer patients with 10 or more positive lymph nodes. Clin. Cancer Res. 11, 8623–8631. 23. Habel, L.A., Shak, S., Jacobs, M.K., Capra, A., Alexander, C., Pho, M., et al. (2006) A population-based study of tumor gene expression and risk of breast cancer death among lymph node-negative patients. Breast Cancer Res. 8, R25.
24. Paik, S., Tang, G., Shak, S., Kim, C., Baker, J., Kim, W., et al. (2006) Gene expression and benefit of chemotherapy in women with nodenegative, estrogen receptor-positive breast cancer. J. Clin. Oncol. 24, 3726–3734. 25. Talantov, D., Baden, J., Jatkoe, T., Hahn, K., Yu, J., Rajpurohit, Y., et al. (2006) A quantitative reverse transcriptase-polymerase chain reaction assay to identify metastatic carcinoma tissue of origin. J. Mol. Diagn. 8, 320–329. 26. Clark-Langone, K.M., Wu, J.Y., Sangli, C., Chen, A., Snable, J.L., Nguyen, A., et al. (2007) Biomarker discovery for colon cancer using a 761 gene RT-PCR assay. BMC Genomics 8, 279. 27. Chomczynski, P. and Sacchi, N. (1987) Single step method of RNA isolation by acid guanidinium thiocyanate–phenol–chloroform extraction. Anal. Biochem. 162, 156–159. 28. Jackson, D.P., Lewis, F.A., Taylor, G.R., Boylston, A.W., and Quirke, P. (1990) Tissue extraction of DNA and RNA and analysis by the polymerase chain reaction. J. Clin. Pathol. 43, 499–504. 29. Lewis, F., Maughan, N.J., Smith, V., Hillan, K., and Quirke, P. (2001) Unlocking the archivegene expression in paraffin-embedded tissue. J. Pathol. 195, 66–71. 30. Beqaj, S.H., Flesher, R., Walker, G.R., and Smith, S.A. (2007) Use of the real-time PCR assay in conjunction with MagNA Pure for the detection of mycobacterial DNA from fixed specimens. Diagn. Mol. Pathol. 16, 169–173. 31. Ribeiro-Silva, A., Zhang, H., and Jeffrey, S.S. (2007) RNA extraction from ten year old formalin-fixed paraffin-embedded breast cancer samples: a comparison of column purification and magnetic bead-based technologies. BMC Mol. Biol. 8, 118. 32. Bohmann, K., Hennig, G, Rogel, U., Poremba, C., Mueller, B.M., Fritz, P., et al. (2009) RNA extraction from archival formalin-fixed paraffin-embedded tissue: a comparison of manual, semiautomated, and fully automated purification methods. Clin. Chem. 55, 1719–1727. 33. Cronin, M., Sangli, C., Liu, M.L., Pho, M., Dutta, D., Nguyen, A., et al. (2007) Analytical validation of the oncotype DX genomic diagnostic test for recurrence prognosis and therapeutic response prediction in node-negative, estrogen receptor-positive breast cancer. Clin. Chem. 53, 1084–1091. 34. Mullins, M., Perreard, L., Quackenbush, J.F., Gauthier, N., Bayer, S., Ellis, M., et al. (2007) Agreement in breast cancer classification between microarray and quantitative reverse transcription PCR from fresh-frozen and for-
RT-PCR Gene Expression Profiling of RNA from Paraffin-Embedded Tissues malin-fixed, paraffin-embedded tissues. Clin. Chem. 53, 1273–1279. 35. Scicchitano, M.S., Dalmas, D.A., Bertiaux, M.A., Anderson, S.M., Turner, L.R., Thomas, R.A., et al. (2006) Preliminary comparison of quantity, quality, and microarray performance of RNA extractedfromformalin-fixed,paraffin-embedded, and unfixed frozen tissue samples. J. Histochem. Cytochem. 54, 1229–1237. 36. Coudry, R.A., Meireles, S.I., Stoyanova, R., Cooper, H.S., Carpino, A., Wang, X., et al. (2007) Successful application of microarray technology to microdissected formalin-fixed, paraffin-embedded tissue. J. Mol. Diagn. 9, 70–79. 37. Macabeo-Ong, M., Ginzinger, D.G., Dekker, N., McMillan, A., Regezi, J.A., Wong, D.T.W., et al. (2001) Effect of duration of fixation on quantitative reverse transcription polymerase chain reaction analyses. Mod. Pathol. 15, 979–987. 38. Cox, M.L., Schray, C.L., Luster, C.N., Stewart, Z.S., Korytko, P.J., Khan, K.N.M., et al. (2006) Assessment of fixatives, fixation, and tissue processing on morphology and RNA integrity. Exp. Mol. Pathol. 80, 183–191. 39. Penland, S.K., Keku, T.O., Torrice, C., He, X., Krishnamurthy, J., Hoadley, K.A., et al. (2007) RNA expression analysis of formalinfixed paraffin-embedded tumors. Lab. Invest. 87, 383–391.
237
40. Roberts, L., Bowers, J., Sensinger, K., Lisowski, A., Getts, R., and Anderson, M.G. (2009) Identification of methods for use of formalin-fixed, paraffin-embedded tissue samples in RNA expression profiling. Genomics 94, 341–348. 41. Gloghini, A., Canal, B., Klein, U., Dal Maso, L., Perin, T., Dalla-Favera, R., et al. (2004) RT-PCR analysis of RNA extracted from Bouin-fixed and paraffin-embedded lymphoid tissues. J. Mol. Diagn. 6, 290–296. 42. Lee, J., Hever, A., Willhite, D., Zlotnik, A., and Hevezi, P. (2005) Effects of RNA degradation on gene expression analysis of human postmortem tissues. FASEB J. 19, 1356–1358. 43. Lee, K.Y., Shibutani, M., Inoue, K., Kuroiwa, K., U, M., Woo, G.H., and Hirose, M. (2006) Methacarn fixation – effects of tissue processing and storage conditions on detection of mRNAs and proteins in paraffin-embedded tissues. Anal. Biochem. 351, 36–43. 44. von Ahlfen, S., Missel, A., Bendrat, K., and Schlumpberger, M. (2007) Determinants of RNA quality from FFPE samples. PlosOne 12, e1261. 45. Rozen, S. and Skaletsky, H. (2000) Primer3 on the WWW for general users and for biologist programmers. Methods Mol Biol. 132, 365–386.
wwwwwwwwwwwwwww
Chapter 15 RT-PCR-Based Gene Expression Profiling for Cancer Biomarker Discovery from Fixed, Paraffin-Embedded Tissues Aaron Scott, Ranjana Ambannavar, Jennie Jeong, Mei-Lan Liu, and Maureen T. Cronin Abstract A molecular test providing clear identification of individuals at highest risk for developing metastatic disease from among early stage breast cancer patients has proven to be of great benefit in breast cancer treatment planning and therapeutic management. Patients with high risk of disease recurrence can also get an estimate of the magnitude of benefit to be gained by adding chemotherapy to surgery and hormonal therapy. Developing this clinical test was made possible by the availability of technologies capable of identifying molecular biomarkers from the gene expression profiles of preserved surgical specimens. Molecular tests such as the Oncotype DX® breast cancer test are proving to be more effective tools for individualized patient stratification and treatment planning than traditional methods such as patient demographic variables and histopathology indicators. Molecular biomarkers must be clinically validated before they can be effectively applied toward patient management in clinical practice. The most effective and efficient means of clinical validation is to use archived surgical specimens annotated with well-characterized clinical outcomes. However, carrying out this type of clinical study requires optimization of traditional molecular expression profiling techniques to analyze RNA from fixed, paraffin-embedded (FPE) tissues. In order to develop our clinically validated breast cancer assay, we modified molecular methods for RNA extraction, RNA quantitation, reverse transcription, and quantitative PCR to work optimally in archived clinical samples. Here, we present an updated description of current best practices for isolating both mRNA and microRNA from FPE tissues for RT-PCR-based expression profiling. Key words: Breast cancer, Formalin-fixed, paraffin-embedded tissue, Quantitative RT-PCR, Expression profiling, Molecular biomarkers, Prognosis, Prediction
Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_15, © Springer Science+Business Media, LLC 2011
239
240
Scott et al.
1. Introduction Many different laboratories have demonstrated that mRNA and microRNA (miRNA) levels can be measured in fixed, paraffinembedded (FPE) tumor specimens, despite the fact that RNA extracted from FPE tissue is often present as fragments less than ~300 bases in length (1–13). This methodology is now routinely applied to tumor marker discovery for clinical diagnostic assays. The focus on FPE specimens has arisen for two reasons: first, the prognostic/predictive potential of gene expression profiling in FPE specimens has been clearly demonstrated (14–24); and second, FPE tissue represents an abundant supply of tissue specimens with well-annotated clinical records of a quality sufficient for developing clinically useful tests (25). Essentially, all clinical biopsy and surgical specimens are treated in a fixing agent such as formalin and then embedded in paraffin. Surgical specimens processed in this manner are stable at room temperature over long periods of time (decades) and non-biohazardous. They are easily sectioned and stained for histopathological diagnostic analyses, immunohistochemistry assays, and fluorescent in situ hybridization analysis or prepared for DNA and RNA isolation. RT-PCR (reverse transcription, polymerase chain reaction) assay of FPE RNA can be applied to perform high-throughput gene expression analysis of specimens from completed clinical trials in order to identify and validate diagnostic biomarkers. It is possible to create a standardized set of methods, including all process steps from tissue extraction through PCR, to support sensitive, precise panels of RT-PCR assays for this type of FPE specimen analysis. Applying these methods to clinical study specimens can be an effective way to generate clinically validated diagnostic tests (26). Although variables such as tissue fixation reagent and fixation protocol selection may affect the extent of RNA fragmentation in FPE tissue as much as archive storage time, these effects can be corrected for with proper assay design and data normalization strategies. Probe and primer sets for RT-PCR assays should be constrained to short sequences of homogeneous length to minimize assay efficiency variability and allow for effective reference-gene-based data normalization (10). Extracting RNA from fixed breast cancer specimens using a proteinase K-based tissue digestion method allows for maximum recovery of short RNA species. Once RNA has been released from the tissue, precipitation-based nucleic acid recovery can be applied to these digests. However, excellent membrane-based and magnetic-bead-based binding methods are also commercially available and lend themselves more easily to process automation (27–29). Normalizing the amount of RNA used in multiplexed, genespecific reverse transcription helps mitigate specimen-to-specimen
RT-PCR-Based Gene Expression Profiling for Cancer Biomarker
241
variability at the PCR step of the analysis. Distributing the cDNA products of a single reverse transcription reaction into parallel quantitative PCR assays allows maximum sensitivity and straight forward analysis of gene expression differences among sets of clinical specimens.
2. Materials 2.1. Sample Preparation
1. Leica RM2235 Microtome (Leica Biosystems, Nussloch, Germany). 2. VWR Microscope slides, Superfrost Plus (VWR International, LLC, West Chester, PA). 3. Polypropylene 1.5 mL microcentrifuge tubes, nuclease-free grade.
2.2. RNA Extraction
1. High Pure miRNA Isolation Kit (Cat. No. 05080576001, Roche Applied Sciences, Indianapolis, IN) containing Paraffin Tissue Lysis buffer, Proteinase K, Binding Buffer, Binding Enhancer, Wash Buffer, Elution Buffer, and High Pure Filter and Collection tubes (see Note 1). 2. 10% Sodium dodecyl sulfate (SDS) solution. 3. DNase I recombinant, RNase-free kit containing 10 U/mL DNase I enzyme and 10× incubation buffer (Cat. No. 04716728001, Roche Applied Sciences, Indianapolis, IN). 4. Ethyl alcohol, 200 proof, molecular biology grade. 5. Nuclease-free water. 6. Xylene. 7. Thermomixer R from Eppendorf.
2.3. RNA Quantitation
1. NanoDrop ND-1000 Sepctrophotometer Technologies, Willmington, DE).
(NanoDrop
2. Quant-iT™ RiboGreen® RNA Assay Kit (Invitrogen, Molecular Probes, Eugene, OR). 2.4. Reverse Transcription
1. Hard-shell thin-wall 96-well microplate or 96-well twin-tec PCR, Skirt-C. 2. Nuclease-free water. 3. PCR strip tubes with cap. 4. Polarseal foil. 5. Universal Human Reference RNA (Stratagene, San Diego, CA). 6. Gene-specific oligonucleotide primers.
242
Scott et al.
7. Omniscript RT Kit containing RNase-free water, 5 mM dNTP mix, 10× RT Buffer, 4 U/mL Omniscript RT (Qiagen, Valencia, CA). 8. RNase Inhibitor (20 U/mL) (Applied Biosystems, Foster City, CA). 2.5. Quantitative PCR
1. 384-Well clear optical reaction plate or 384-well clear twintec plate. 2. LightCycler® 480 Real-Time PCR system (Roche Applied Sciences, Indianapolis, IN). 3. Hard-shell thin-wall 96-well microplate or 96-well twin-tec PCR, Skirt-C. 4. Nuclease-free water. 5. Optical adhesive covers. 6. Polarseal foil. 7. Sealing tape for 96-well plates. 8. Oligonucleotide primers and dual-labeled TaqMan probe for each gene assay (TaqMan assays). 9. 2× TaqMan® EagleTaq PCR Master Mix, (Roche Applied Sciences, Indianapolis, IN).
3. Methods Expression profiling of RNA from FPE samples is challenging due to the generally fragmented state of the RNA. However, fixed specimens are easily and safely handled at room temperature since the fixation process stabilizes any biomolecules and destroys pathogens. Historically, most protocols for isolating RNA and preparing it for molecular analysis were developed and optimized for RNA isolated from fresh tissue or cultured cells. However, due to the appeal of FPE tissue for its ease of storage and associated clinical information, good commercial kits for isolating RNA from fixed specimens are now widely available. RNA from fresh tissue or cultured cell sources is typically greater than a kilobase in length with intact polyadenylated tails suitable for oligo-dT primed cDNA synthesis. However, this approach is typically unsuccessful when applied to RNA extracted from fixed tissues. Successful extraction from fixed tissues requires that protocols be optimized to recover a smaller class of RNA than typically obtained from fresh tissue and cell lines. Priming for cDNA synthesis must be based on random or gene-specific oligonucleotide primers rather than oligo-dT priming. In addition, amplicon length for RT-PCR targets must be constrained to match the
RT-PCR-Based Gene Expression Profiling for Cancer Biomarker
243
small size of RNA recovered from the sample population under study. Often, for FPE tissues, these RT-PCR targets are less than 100 nucleotides in length. Until recently, an RNA extraction method based on proteinase K tissue digestion, followed by nucleic acid precipitation was recommended (e.g., MasterPure Kit from Epicentre Biotechnologies Madison, WI). Now, improved alternative methods have become commercially available such as the High Pure kits from Roche Applied Sciences (Indianapolis, IN). These kits utilize a proteinase K digestion procedure followed by membrane binding to wash away impurities and selective elution of the nucleic acids. Then the eluate goes through in-solution DNase treatment to remove residual genomic DNA which could interfere with accurate RT-PCR expression analysis. Then the digestion mixture is rebound to a new membrane and DNase reagents are washed away to finally enable selective release of purified RNA. This membrane-binding method is effective in FPE tissue because it efficiently recovers large quantities of RNA over a wide size range, including small size RNA molecules such as microRNA. The RNA is quantified so that a standard amount of RNA can be added to the cDNA synthesis and consequently the qPCR assays. Optical density reading using the ND-1000 microspectrophotometer (NanoDrop Technologies, Wilmington, DE) is an accurate and effective way to measure RNA concentration for small numbers of samples. However, when processing large number of samples, an assay based on using RiboGreen fluorescent dye with a standard curve of ribosomal RNA (RiboGreen Assay, Life Technologies, Carlsbad, CA) can be assembled in 96-well plates and analyzed on a plate reader (SpectraMax, Molecular Devices, Sunnyvale, CA). Quantitation of RNA by direct measurement of optical density on a spectrophotometer will be described here. The RT-PCR method given in this protocol is carried out using TaqMan assays run on the LightCycler® 480 Real-Time PCR System (Roche Applied Sciences, Indianapolis, IN). This platform permits flexibility in scaling to handle samples of widely varying characteristics, including low RNA yield specimens, since the instrument can be used in the 384-well format with assay volumes as small as 3 mL. The assay plates can be assembled by hand using multichannel pipettes or on a larger scale using robotic liquid handing systems such as the Beckman Coulter Bio-Mek (Fullerton, CA) or the Tecan Aquarius and Genesis (Männedorf, Switzerland) robots. The protocols described here are manual methods that can be applied in any laboratory without relying on robotic systems. 3.1. Gene Selection and Assay Design
1. For each gene candidate to be used in the study, identify the appropriate mRNA reference sequence using Entrez Gene and the correct accession number/sequence ID for the gene of interest. Access and download the consensus
244
Scott et al.
sequence through the NCBI Entrez nucleotide database. When selecting candidate genes, both biomarker and reference normalization genes should be considered. 2. Design RT-PCR primers and probes using any assay design method. Primer Express® (Applied Biosystems, Foster City, CA) or the publicly available Primer3 (http://frodo.wi.mit. edu/primer3/input.htm) both generally work well (30) (see Note 2). 3. Order the TaqMan assays, i.e., the oligonucleotide primers and dual-labeled TaqMan probes, with 5¢-FAM as a reporter and a 3¢ quencher molecule (see Note 3), from a commercial oligonucleotide supplier. 3.2. Sample Preparation
1. Preparing formalin-fixed, paraffin-embedded samples for RNA extraction requires a microtome to cut sections from paraffin blocks in which the fixed tissues have been embedded (see Note 4). Although the dimensions of embedded tissues can vary widely, cutting three 10 mm thick FPE sections from each paraffin block is a good standard procedure and will often provide microgram amounts of RNA for each sample. For easy handling, cut the three sections sequentially, allowing them to curl as the microtome blade moves over the block surface. 2. Using a fine brush and/or tissue forceps, place all three 10 mm sections into one 1.5 mL polypropylene microcentrifuge tube for extraction.
3.3. RNA Extraction Using the High Pure miRNA Purification Kit
1. Centrifuge the samples in a microcentrifuge set between +18 and +25°C for 30 s at approximately 6,000 rpm (3,824 ´ g). 2. Add 1 mL of xylene to each tube containing a tissue sample. Invert the tube until the tissue sample is dislodged from the bottom. Mix the samples for 5 min at +50°C in a thermomixer at 850 rpm. Centrifuge the samples in a microcentrifuge set between +18 and +25°C for 5 min at approximately 14,000 rpm (20,817 ´ g). Carefully aspirate and discard the supernatant without disturbing the tissue at the bottom of the tube. 3. Repeat the steps for a total of two xylene extractions. Visually inspect the sample for evidence of paraffin. If residual paraffin exists, repeat the xylene extraction process until the paraffin is completely removed. 4. Add 1 mL of 100% ethanol to each sample tube. Invert and shake each tube until the tissue sample is dislodged from the bottom of the tube. Loosening the tissue increases the surface area exposed to the ethanol. Invert the sample tubes 10–15 times to ensure adequate mixing. Centrifuge the sample tubes in a microcentrifuge set between +18 and +25°C for 5 min at approximately 14,000 rpm (20,817 ´ g).
RT-PCR-Based Gene Expression Profiling for Cancer Biomarker
245
Carefully aspirate and discard all the ethanol without disturbing the tissue at the bottom of the tube. 5. Repeat for a total of two clean-up washes. 6. Add 100 mL Paraffin Tissue Lysis Buffer, 16 mL 10% SDS, and 40 mL Proteinase K working solution to each deparaffinized sample pellet. Vortex briefly and centrifuge samples for 20 s at 1,000 rpm (106 ´ g). 7. Incubate the samples at 55°C for 2–3 h in a thermomixer at 850 rpm. Concluding the incubation, centrifuge samples for 1 min at 3,000 rpm (956 ´ g). 8. Add 325 mL of Binding Buffer to each sample tube. Vortex briefly. 9. Add 325 mL of Binding Enhancer to each sample tube. Vortex briefly. 10. Combine the High Pure filter tube with a collection tube and pipette the whole mixture from the previous step into the upper reservoir. Note: Transfer a maximum of 650 mL to avoid overloading the column. 11. Centrifuge for 30 s at 11,000 rpm, discard the flow-through. The above steps can be repeated in order to load the column with additional sample material (do not overload the column). 12. Add 500 mL Wash Buffer working solution. Centrifuge for 30 s at 11,000 rpm (12,851 ´ g), discard the flow-through. 13. Add 300 mL Wash Buffer working solution. Centrifuge for 30 s at 11,000 rpm (12,851 ´ g), discard the flow-through. 14. Centrifuge at 11,000 rpm (12,851 ´ g) for 1 min, in order to dry the filter fleece completely. 15. Place the High Pure filter tube into a fresh, labeled 1.5 mL microcentrifuge tube, add 50 mL Elution Buffer (preheated at 75°C), and incubate for 1 min at +15 to +25°C. Centrifuge for 1 min at 11,000 rpm (12,851 ´ g). 16. Once again, add 50 mL Elution Buffer (preheated at 75°C) and incubate for 1 min at +15 to +25°C. Centrifuge for 1 min at 11,000 rpm (12,851 ´ g). The final elution volume will be 100 mL. 17. For DNase treatment, it is recommended to elute the sample with 100 mL Elution Buffer to minimize the loss of sample material. 18. To approximately 100 mL eluate, add 50 mL DNase solution prepared as shown below. Vortex briefly and centrifuge for 20 s at 1,000 rpm (106 ´ g). Incubate for 30 min at 37°C without shaking. 19. Preparation of DNase master mix (see Note 5)
246
Scott et al.
Reagents
Each Rxn. (mL) Calculation
DNase master mix for 30 samples (mL)
10× DNase I buffer
15
(Number of samples (30 × 1.12) × 15 = 504 × 1.12) × 15
DNase I enzyme (10 U/mL)
2
(Number of samples (30 × 1.12) × 2 = 67.2 × 1.12) × 2
Nuclease-free water
33
(Number of samples (30 × 1.12) × × 1.12) × 33 33 = 1,108.8
Total
50
1,680
20. Add 325 mL Binding Buffer to the DNase-treated eluate. Vortex briefly. 21. Add 210 mL Binding Enhancer to the DNase-treated eluate/ Binding Buffer mixture. Vortex briefly. 22. Combine the High Pure filter tube with a 2 mL collection tube and pipette the DNase-treated sample/Binding Buffer/ Binding Enhancer mixture into the upper reservoir. 23. Centrifuge for 30 s at 11,000 rpm (12,851 ´ g) in a microcentrifuge and discard the flow-through. 24. Add 500 mL Wash Buffer working solution. Centrifuge for 30 s at 11,000 rpm (12,851 ´ g), discard the flow-through. 25. Add 300 mL Wash Buffer working solution. Centrifuge for 30 s at 11,000 rpm (12,851 ´ g), discard the flow-through. 26. Centrifuge at 11,000 rpm (12,851 ´ g) for 1 min, in order to dry the filter fleece completely. 27. Place the High Pure filter tube into a fresh, labeled 1.5 mL microcentrifuge tube, add 50–100 mL (final elution volume depends on the expected yield for your samples) Elution Buffer, preheated at 75°C (bottle 6, colorless cap), and incubate for 1 min at +15 to +25°C. Centrifuge for 1 min at 11,000 rpm (12,851 ´ g). 28. The microcentrifuge tube now contains the eluted total RNA. Store purified RNA at −20 ± 5°C for periods of less than 1 day or at −80 ± 10°C for periods greater than a day, up to 6 months. 3.4. RNA Quantification
The concentration of RNA samples may be quantified using RiboGreen fluorescence assay or NanoDrop spectrophotometer as described in Subheading 3.5 of Chapter 14 (see Note 6)
3.5. Reverse Transcription
Reverse transcription (RT) is performed in a 96-well plate as a batch process scalable from 1 to 94 samples plus one positive control sample and one negative control sample. Any RNA source
RT-PCR-Based Gene Expression Profiling for Cancer Biomarker
247
known to express the candidate genes used in the study can act as a positive control. Commercially available standard RNA mixtures made from cell line extracts such as Universal Human Reference RNA (Stratagene, San Diego, CA) are a convenient positive control option. The negative control is simply an RT reaction that replaces the RNA template with water in order to test for laboratory contamination of reagents with target sequences as well as false-positive signals due to nonspecific priming. 1. Create the gene-specific primer (GSP) pool for specifically priming the genes of interest during the reverse transcription reaction or, alternatively, use random primers. Random primers from hexamers up to nonamers have proven successful in randomly primed reverse transcription. 2. To prepare a GSP pool, remove the appropriate PCR reverse primer stock tubes from −20°C storage. Allow the tubes to remain at room temperature until fully thawed then vortex the reagent tubes for approximately 5 s until thoroughly mixed. Centrifuge in a microcentrifuge set at +4°C for 5–30 s at no more than 14,000 rpm (20,817 ´ g). Store the thawed tubes on ice while in use. 3. Using the concentration on the vendor certificate of analysis, calculate the volume of reverse primer to be added using the following equation: (see Note 7) C1V1 = C2V2, where C1 is the starting concentration of each reverse primer stock solution; V1 the volume of each reverse primer stock solution to be added; C2 the final concentration of each reverse primer (1 mM); and V2 the total volume of the GSP pool. Example: To create 1 mL of the GSP pool with reverse primers each having a starting concentration of 200 mM: (200 mM)(V1) = (1 mM) × (1,000 mL) V1 = [(1 mM) × (1,000 mL)]/[200 mM] V1 = 5 mL 4. Calculate the volume of nuclease-free water to be added using the following equation: Volume of nuclease-free water = Total volume of GSP pool − (Number of reverse primers × Volume of each reverse primer). GSP priming pool preparation Number of genes
Total pool volume (mL)
Volume per reverse primer (mL)
Nuclease-free water (mL)
192
1,000
5
40
248
Scott et al.
5. Dispense the required volume of nuclease-free water into a microcentrifuge tube. Pipette the required volume of each reverse primer into the same tube containing water. Vortex the tube for approximately 5 s until thoroughly mixed. Centrifuge for approximately 5 s to remove any residual reagent from the tube lid. 6. Obtain the following reagents from −20°C storage: GSP pool, RNase Inhibitor, Omniscript RT Kit, including: 5 mM dNTP mix, 10× RT Buffer, RNase-free water, and 4 U/mL Omniscript RT. 7. Flick the RNase Inhibitor and Omniscript RT tubes to mix, then spin them down, and place them on ice. Allow the remaining reagents listed above to thaw at room temperature until no crystals are visible. Vortex the tubes containing 10× buffer, dNTP mix, RNase-free water, and GSP pool for approximately 5 s at a medium-high speed until thoroughly mixed. Shake down or spin down each tube to remove condensation from the tube lid, if necessary. Store the reagents on ice while in use. 8. Calculate the number of RT reactions as follows: Total number of RT reactions in RT master mix = (Number of samples + 1 Positive control + 1 Negative control) + An additional ~10% sample reactions as overage Based on the number of RT reactions, use the table below to calculate the volume of each reagent required to make the Master Mix. Master Mix preparation for RT reaction 94 samples + 2 controls (106 reactions includes ~10% sample overage)
Reagents
1 reaction (mL/reaction)
10× Buffer RT
8
848
dNTP mix, 5 mM each dNTP
8
848
Nuclease-free water
0.2
GSP pool, 1 mM
4
424
ABI RNase Inhibitor, 20 U/mL
2
212
Omniscript RT, 4 U/mL
3
318
Total volume (mL)
25.2
21.2
2,671.2
RT-PCR-Based Gene Expression Profiling for Cancer Biomarker
249
9. Pipette the reagents listed above into a labeled tube. Return the component reagents to the −20°C storage after use. Vortex the Master Mix tube at medium-high speed for approximately 2–5 s until thoroughly mixed. Shake down or centrifuge the tube in a mini centrifuge for approximately 5 s to remove any residual reagent from the lid. Place the Master Mix tube on ice. 10. Obtain the RNA samples and positive RT control (e.g., Universal Human Reference RNA) from −20°C storage and thaw completely on ice. Flick the tubes for 2–5 s to mix the RNA. Centrifuge the tubes in a microcentrifuge set at +4°C for 5–30 s at no more than 14,000 rpm (20,817 ´ g). 11. Calculate the amount of RNA to use in each reaction using the table below. The standard PCR volume is 5 mL with 1 ng of cDNA per reaction well. RNA required = [Number of genes × Number of replicates + Overage wells] × 1 ng/well. The RNA must be at a high enough concentration so that the required amount is contained in £ 54.8 mL since 25.2 mL of reagent Master Mix will be added to achieve a standard RT reaction volume of 80 mL. Sample RNA requirements Numbter of genes
Number of replicates
Overage wells
RNA required
Volume
192
1
40 (~20%)
232 ng/sample
54.8 mL
12. Assemble the reverse transcription reactions in a hard-shell thinwall 96-well microplate or a set of strip tubes (see Note 8). 13. To assemble the positive control, dispense 52.48 mL of nuclease-free water into the positive control well or tube along with 2.32 mL of the positive control RNA that is normalized to (100 ng/mL). 14. To assemble the negative control (no template control), dispense 54.8 mL of nuclease-free water into the negative control well. 15. To assemble the sample reverse transcription reactions, dispense 31.6 mL of nuclease-free water into each reaction well. Dispense 23.2 mL of each normalized RNA sample (10 ng/mL) into one of the reaction wells. Repeat the steps for the remaining RNA samples (see Note 9). 16. Add 25.2 mL of RT Master Mix to each reaction well containing diluted RNA samples, positive and negative RT controls. Total reaction volume for each reaction is 80 mL. Mix the contents of the plate or strip tubes as follows. When using plates, pipette the solution up and down two to three times to mix. When using strip tubes, flick the tubes for 2–5 s to mix.
250
Scott et al.
Seal plates with polarseal film or cap strip tubes with the attached caps. Centrifuge the plate in a plate centrifuge set at +10°C for approximately 2 min at approximately 2,000 rpm (827 ´ g) or pulse-spin the strip tubes in a mini centrifuge for approximately 5 s (see Note 10). 17. Thermal cycle the RT reactions at +37°C for 60 min followed by +93°C for 5 min, and then hold the temperature at +10°C until the instrument is stopped. Reaction volume is 80 mL. 18. If the samples will be processed through quantitative PCR the same day, store them on ice. If the samples will be processed at a later time, store the samples in −20°C storage. 3.6. Preparation of the qPCR Assays
PCR assay preparation includes combining the cDNA mixture from the RT reaction with PCR Master Mix reagent and a primer–probe mixture. The protocol given here assumes 5 mL reactions for 192 genes with 1 ng of RNA (cDNA) per reaction in a 384-well plate. This setup allows three possible scenarios on a single qPCR plate: (1) a sample can be tested against a negative PCR control (two reactions for each of 192 genes, one for the sample and one for the control); (2) a sample can be tested in duplicate (two identical reactions for each gene for a single sample); and (3) two samples can be tested together (each sample without replication). In addition, this protocol is very scalable and can be adapted to different numbers of candidate genes, different reaction volumes, or different quantities of RNA in a reaction. One PCR Master Mix is prepared for each of the following: each experimental sample, the PCR negative (no template) control for each test sample (optional), the reverse transcription positive control sample, and the reverse transcription negative control sample. In this process, 2× TaqMan® EagleTaq PCR Master Mix solution is combined with a cDNA sample and nuclease-free water. All steps described here can be performed using multichannel pipettors or automated liquid handling robots such as the Tecan Aquarius or Genesis robots or the Beckman Biomek workstation. 1. Obtain a bottle of 2× TaqMan® EagleTaq PCR Master Mix reagent from +4°C storage. Store the 2× TaqMan® EagleTaq PCR Master Mix on ice, protected from light, when in use (see Note 11). 2. Obtain the samples (cDNA from the RT reactions) from +4°C storage or −20°C storage. If stored at −20°C, place the samples on ice until thawed. If the samples were assembled in plates, spin down using a plate centrifuge set at +10°C for approximately 2 min at approximately 2,000 rpm (827 ´ g). Centrifuge strip tubes in a mini centrifuge for approximately 5 s to remove any reagent from the tube lid. Store the thawed samples on ice while in use. 3. Obtain an appropriate tube for preparing the sample Master Mix. Each sample will be tested for 192 genes requiring
RT-PCR-Based Gene Expression Profiling for Cancer Biomarker
251
192 individual 5 mL reactions (prepare enough Master Mix for 232 reactions to allow for ~20% overage). Using the table below, pipette the reagents and quantities listed into the qPCR Master Mix tube. Four microliters of the PCR Master Mix will be used to assemble a 5 mL qPCR for each gene, i.e., 4 mL of qPCR Master Mix + 1 mL of P3 pool (see step 8). Quantitative PCR Master Mix preparation for one sample 2× EagleTaq PCR Master Mix
580 mL
RT Reaction Mix (sample)
80 mL
Nuclease-free water
268 mL
Total volume
928 mL
4. Cap the Master Mix tube and invert for 2–5 s until fully mixed. Centrifuge the tube for approximately 5 s to remove any reagent from the lid. Place the Master Mix tube on ice. 5. To prepare the Master Mix for the reverse transcription positive and negative control samples, label an appropriate tube for each and follow the instructions for test samples above. 6. To prepare the optional PCR negative control (no template) for each sample, obtain an appropriate tube for preparing the Master Mix. Prepare enough reagent for 192 reactions plus a ~20% pipetting overage as shown in the table below. EagleTaq PCR Master Mix preparation for one negative control 2× EagleTaq PCR Master Mix
580 mL
Nuclease-free water
348 mL
Total volume
928 mL
7. Cap the Master Mix tube and invert for 2–5 s until fully mixed. Centrifuge the tube for approximately 5 s to remove any reagent from the lid. Place the Master Mix tube on ice. 8. To prepare the primer and probe (P3) pools, thaw the primer and probe stock reagent tubes completely. Vortex the primer and probe tubes for approximately 30 s until thoroughly mixed. Centrifuge at +4°C for 5–30 s at no more than 14,000 rpm (20,817 ´ g). While in use, store the thawed primer and probe tubes on ice. 9. Using the concentration provided on the certificate of analysis for the primers, calculate the volume of each primer to be added as follows: (see Note 12)
252
Scott et al.
Volume of forward or reverse primer = [(Final concentration of primer) × (Total volume of P3 pool)]/(Initial concentration of primer) Example: To create 600 mL of the P3 pool with each forward and reverse primer having a starting concentration of 200 mM: Volume of forward or reverse primers = [(4.5 mM) × (600 mL)]/ [(200 mM)] 10. Using the concentration provided on the certificate of analysis for the probes, calculate the volume of probe to be added as follows: (see Note 13) Volume of probe = [(Final concentration of probe) × (Total volume of P3 pool)]/(Initial concentration of probe) Example: To create 600 mL of the P3 pool with each probe having a starting concentration of 100 mM: Volume of Probe = [(1 mM) × (600 mL)]/[(100 mM)] 11. Determine the volume of nuclease-free water as follows: Volume of nuclease-free water = (Total volume of P3 mix)− [(Volume of forward primer) + (Volume of reverse primer)+ (Volume of probe)] Sample calculation for preparing 600 mL of P3 pool Reagent
Volume (mL)
Volume of forward primer
13.5
Volume of reverse primer
13.5
Volume of probe
6
Volume of nuclease-free water
567
Total volume
600
12. Pipette the calculated volumes of the required reagents into an amber microcentrifuge tube labeled with primer and probe set name. Vortex the P3 pool for approximately 30 s until thoroughly mixed. Centrifuge in a microcentrifuge set at +4°C for 5–30 s at no more than 14,000 rpm (20,817 ´ g). Store the P3 pool tube on ice while in use. 13. Distribute the P3 mixtures into a 384-well plate using 1 mL/ well for each 5 mL reaction. A plate layout should be predetermined where each primer and probe mix is deposited into the appropriate reaction wells for each test sample or control on the plate. Once the P3 mixtures have been dispensed, the qPCR Master Mix containing RT Reaction Mix from samples
RT-PCR-Based Gene Expression Profiling for Cancer Biomarker
253
or controls (refer to steps 3–5 above) should be distributed into the appropriate wells at 4 mL/well. 14. Seal the plate by hand with optical adhesive film. Spin down the plate in a plate centrifuge set at +10°C for approximately 2 min at approximately 2,000 rpm (827 ´ g). Ensure no air bubbles are trapped at the bottom of the plate. If air bubbles are seen, spin the plate down again. Repeat until no air bubbles are trapped at the bottom of the plate. 15. Run the plate on the LightCycler® 480 Real-Time PCR System (Roche Applied Sciences, Indianapolis, IN) using the following settings: assay: absolute quantitation; container: 384-well clear plate; template: TaqMan 5 mL; cycling conditions: +95°C, 155 min (+95°C, 20 s, +60°C, 45 s) for 40 cycles; and assay volume: 5 mL. 3.7. Data Analysis
At the conclusion of the run, use the Roche LightCycler 480 System User Guide to export and process the data following the instructions for real-time analysis for absolute quantification.
4. Notes 1. It is important to store and handle the components of the High Pure miRNA Isolation Kit appropriately. It contains: Paraffin Tissue Lysis Buffer, Proteinase K (100 mg lyophilized), Binding Buffer, Binding Enhancer, Wash Buffer, Elution Buffer, and High Pure Filter and Collection tubes. Also, the High Pure miRNA kit does not come with a DNase I enzyme and associated DNase Buffer, so it must be ordered separately from Roche. The Proteinase K must be dissolved in 4.5 mL Elution Buffer and aliquots of this must be stored at −20°C. The DNase I enzyme and buffer kit must also be kept at −20°C. Store the remaining components at +18 to +25°C. The Wash Buffer reagents must be reconstituted using absolute ethanol as indicated in the Roche High Pure miRNA manual. Ensure that ethanol has been added and the solution thoroughly mixed. 2. Amplicon sizes should be constrained to less than 100 bases total in length, including primer sequences. This design parameter should be adjusted in the Primer Express or Primer3 settings since the default usually permits amplicons as long as 300 bases. It is generally easier to get a design that meets all necessary criteria by allowing amplicons that do not span an exon junction. DNase treatment of the RNA extracts and inclusion of no template controls eliminate the necessity to design RNA-specific assays.
254
Scott et al.
3. The LightCycler® 480 Real-Time PCR System (Roche Applied Sciences, Indianapolis, IN) is capable of detecting a number of different fluorescent reporter dyes, but the fluorescein-based dye FAM is the most robust and widely available. There are also multiple different quencher molecules that can be paired with FAM on the dual-labeled probes. Although TAMARA dye was often used in the original form of the TaqMan assay, the newer “dark” (nonfluorescent) quenchers give much better assay results due to superior quenching and lower background fluorescence. Commercial oligonucleotide vendors offer a variety of TaqMan probe configurations. 4. Pathology fixation procedures vary widely depending on tissue type and each laboratory’s optimized practices. RNA can be successfully extracted from nearly all types and ages of fixed tissue; however, yield will vary with sample quality and age. The quality of the extracted RNA (degree of fragmentation) will also vary with sample age and fixation method. If tissue is available, a test extraction can be very helpful for estimating how much RNA to expect from a sample. 5. Note that there may be a wide range of possible concentrations (amounts) of genomic DNA remaining in a nucleic acid extract after DNase treatment. As a result, a single standard DNase treatment condition may prove to be insufficient to completely remove interfering DNA. It is therefore important to test for residual DNA and treat more aggressively with DNase if the result shows persistent genomic DNA in the nucleic acid extract. 6. It is important to keep in mind that when determining the RNA concentration spectrophotometrically for samples extracted using filter-based methods, glass fibers might get released along with RNA in the eluate and may interfere with optical density measurement, thereby giving a false increase in concentration. Hence before determining RNA concentration spectrophotometrically, centrifuge the sample at 11,000 rpm (12,851 ´ g) for 1 min and transfer supernatant to a fresh 1.5 mL reaction tube being careful not to pipette the glass fibers that have settled to the bottom of the tube. This step must be included if RNA isolation is carried out using a filterbased method. The alternative method for RNA quantification using the Quant-iT™ RiboGreen® RNA Assay Kit (Invitrogen, Molecular Probes, Eugene, OR) is not subject to this complication. This fluorescence-based kit uses the RiboGreen dye to detect RNA even in the presence of DNA relative to a standard curve generated using a riboRNA standard included in the kit. This is a much higher throughput method for quantifying large numbers of samples. The output
RT-PCR-Based Gene Expression Profiling for Cancer Biomarker
255
is read on a fluorescent plate reader, the Spectramax GeminiXS spectrophotometer (Molecular Devices, Menlo Park, CA). 7. The concentration of each reverse primer stock solution is typically 100 mM or 200 mM. The reverse primers will be mixed with nuclease-free water to bring the final concentration of each reverse primer in the GSP pool to 1 mM. 8. Keep all reverse transcription reaction components cold throughout the PCR assembly process, including after being added to the plate or strip tubes, by keeping them on ice, in a cooler, or equivalent. 9. Assembling all of the reverse transcription reactions is much easier if each RNA sample is normalized to a standard concentration of 10 ng/mL using nuclease-free water. Alternatively, each reverse transcription reaction can be individually calculated and the final volumes normalized with nuclease-free water to compensate for any differences in added RNA volume. 10. When carrying out temperature-sensitive reactions in microtiter plates, it is very important to ensure no air bubbles are trapped at the bottom of the plate so that enzymatic mixtures are exposed uniformly to correct temperature conditions. If air bubbles are seen, tap the plate to dislodge the bubbles. Spin down the plate in a plate centrifuge set at +10°C for approximately 2 min at approximately 2,000 rpm (827 ´ g). Repeat until no air bubbles are trapped at the bottom of the plate. 11. This section describes the steps used to prepare Master Mix for one sample. One Master Mix is prepared for each cDNA test sample, the RT positive control sample, and the RT negative control sample. Similarly, one negative control Master Mix can be prepared as a no template control for each sample where nuclease-free water is substituted for the cDNA products from the RT reaction. It is recommended that the positive and negative reverse transcription samples be assembled and processed through PCR prior to assembling the test sample PCR plates to ensure that the reverse transcription reaction was successful and no contaminations occurred. 12. The concentration of each forward and reverse primer stock solution is typically 100 or 200 mM. The primers will be mixed with probe and nuclease-free water to bring the final concentration of the primer in the P3 pool to 4.5 mM. Final primer concentrations in the PCR are 900 nM. 13. The concentration of each probe stock solution is typically 100 mM. The probes will be mixed with primer and nucleasefree water to bring the final concentration of the probe in the P3 pool to 1 mM. The final PCR probe concentration is 200 nM.
256
Scott et al.
References 1. Rupp, G.M. and Locker, J. (1988) Purification and analysis of RNA from paraffin-embedded tissues. Biotechniques 6, 56–60. 2. Finke, J., Fritzen, R., Ternes, P., Lange, W. and Dolken, G. (1993) An improved strategy and a useful housekeeping gene for RNA analysis from formalin-fixed, paraffinembedded tissues by PCR. Biotechniques 14, 448–453. 3. Mies, C. (1994) Molecular biological analysis of paraffin-embedded tissues. Hum. Pathol. 25, 555–560. 4. Krafft, A.E., Duncan, B.W., Bijwaard, K.E., Taubenberger, J.K. and Lichy, J.H. (1997) Optimization of the isolation and amplification of RNA from formalin-fixed, paraffinembedded tissue: the armed forces institute of pathology experience and literature review. Mol. Diagn. 2, 217–230. 5. Stanta, G. and Bonin, S. (1998) RNA quantitative analysis from fixed and paraffin-embedded tissues: membrane hybridization and capillary electrophoresis. Biotechniques 24, 271–276. 6. Sheils, O.M. and Sweeney, E.C. (1999) TSH receptor status of thyroid neoplasms – TaqMan RT-PCR analysis of archival material. J. Pathol. 188, 87–92. 7. Godfrey, T.E., Kim, S.H., Chavira, M., Ruff, D.W., Warren, R.S., Gray, J.W. et al. (2000) Quantitative mRNA expression analysis from formalin-fixed, paraffin-embedded tissues using 5¢ nuclease quantitative reverse transcription-polymerase chain reaction. J. Mol. Diagn. 2, 84–91. 8. Specht, K., Richter, T., Muller, U., Walch, A., Werner, M. and Hofler, H. (2001) Quantitative gene expression analysis in microdissected archival formalin-fixed and paraffin-embedded tumor tissue. Am. J. Pathol. 158, 419–429. 9. Abrahamsen, H.N., Steiniche, T., Nexo, E., Hamilton-Dutoit, S.J. and Sorensen, B.S. (2003) Towards quantitative mRNA analysis in paraffin-embedded tissues using real-time reverse transcriptase-polymerase chain reaction: a methodological study on lymph nodes from melanoma patients. J. Mol. Diagn. 5, 34–41. 10. Cronin, M., Pho, M., Dutta, D., Stephans, J.C., Shak, S., Kiefer, M.C., et al. (2004) Measurement of gene expression in archival paraffin-embedded tissues: development and performance of a 92-gene reverse transcriptase-polymerase chain reaction assay. Am. J. Pathol. 164, 35–42. 11. Haque, T., Faury, D., Albrecht, S., LopezAguilar, E., Hauser, P., Garami, M., et al. (2007)
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
Gene expression profiling from formalin-fixed paraffin-embedded tumors of pediatric glioblastoma. Clin. Cancer Res. 13, 6284–6292. Linton, K.M., Hey, Y., Saunders, E., Jeziorska, M., Denton, J., Wilson, C.L., et al. (2008) Acquisition of biologically relevant gene expression data by Affymetrix microarray analysis of archival formalin-fixed paraffin-embedded tumours. Br. J. Cancer 98, 1403–1414. Oberli, A., Popovici, V., Delorenzi, M., Baltzer, A., Antonov, J., Matthey, S., et al. (2008) Expression profiling with RNA from formalin-fixed, paraffin-embedded material. BMC Med. Genet. 1, 1–15. Chang, J., Powles, T.J., Allred, D.C., Ashley, S.E., Clark, G.M., Makris, A., et al. (1999) Biologic markers as predictors of clinical outcome from systemic therapy for primary operable breast cancer. J. Clin. Oncol. 17, 3058–3063. Davis, R.E. and Staudt, L.M. (2002) Molecular diagnosis of lymphoid malignancies by gene expression profiling. Curr. Opin. Hematol. 9, 333–338. Gianni, L., Zambetti, M., Clark, K., Baker, J., Cronin, M., Wu, J., et al. (2005) Gene expression profiles in paraffin-embedded core biopsy tissue predict response to chemotherapy in women with locally advanced breast cancer. J. Clin. Oncol. 23, 7265–7277. Golub, T.R., Slonim, D.K., Tamayo, P., Huard, C., Gaasenbeek, M., Mesirov, J.P., et al. (1999) Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science 286, 531–537. Habel, L.A., Shak, S., Jacobs, M.K., Capra, A., Alexander, C., Pho, M., et al. (2006) A population-based study of tumor gene expression and risk of breast cancer death among lymph node-negative patients. Breast Cancer Res. 8, R25. Paik, S., Shak, S., Tang, G., Kim, C., Baker, J., Cronin, M., et al. (2004) A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer. N. Engl. J. Med. 351, 2817–2826. Paik, S., Tang, G., Shak, S., Kim, C., Baker, J., Kim, W., et al. (2006) Gene expression and benefit of chemotherapy in women with nodenegative, estrogen receptor-positive breast cancer. J. Clin. Oncol. 24, 3726–3734. Ramaswamy, S., Ross, K.N., Lander, E.S. and Golub, T.R. (2003) A molecular signature of metastasis in primary solid tumors. Nat. Genet. 33, 49–54.
RT-PCR-Based Gene Expression Profiling for Cancer Biomarker 22. Rosenwald, A., Wright, G., Wiestner, A., Chan, W.C., Connors, J.M., Campo, E., et al. (2003) The proliferation gene expression signature is a quantitative integrator of oncogenic events that predicts survival in mantle cell lymphoma. Cancer Cell 3, 185–197. 23. Sorlie, T., Perou, C.M., Tibshirani, R., Aas, T., Geisler, S., Johnsen, H., et al. (2001) Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc. Natl Acad. Sci. USA 98, 10869–10874. 24. van de Vijver, M.J., He, Y.D., Veer, L.J., Dai, H., Hart, A.A., Voskuil, D.W., et al. (2002) A gene-expression signature as a predictor of survival in breast cancer. N. Engl. J. Med. 347, 1999–2009. 25. Lewis, F., Maughan, N.J., Smith, V., Hillan, K. and Quirke, P. (2001) Unlocking the archive – gene expression in paraffin-embedded tissue. J. Pathol. 195, 66–71. 26. Cronin, M., Sangli, C., Liu, M.L., Pho, M., Dutta, D., Nguyen, A., et al. (2007) Analy tical validation of the Oncotype DX geno mic diagnostic test for recurrence prognosis
27.
28.
29.
30.
257
and therapeutic response prediction in node-negative, estrogen receptor-positive breast cancer. Clin. Chem. 53, 1084–1091. Beqaj, S.H., Flesher, R., Walker, G.R. and Smith, S.A. (2007) Use of the real-time PCR assay in conjunction with MagNA Pure for the detection of mycobacterial DNA from fixed specimens. Diagn. Mol. Pathol. 16, 169–173. Ribeiro-Silva, A., Zhang, H. and Jeffrey, S.S. (2007) RNA extraction from ten year old formalin-fixed paraffin-embedded breast cancer samples: a comparison of column purification and magnetic bead-based technologies. BMC Mol. Biol. 8, 118. Bohmann, K., Hennig, G., Rogel, U., Poremba, C., Mueller, B.M., Fritz, P., et al. (2009) RNA extraction from archival formalinfixed paraffin-embedded tissue: a comparison of manual, semiautomated, and fully automated purification methods. Clin. Chem. 55, 1719–1727. Rozen, S. and Skaletsky, H. (2000) Primer3 on the WWW for general users and for bio logist programmers. Meth. Mol. Biol. 132, 365–386.
wwwwwwwwwwwwwww
Chapter 16 MicroRNA Isolation from Formalin-Fixed, Paraffin-Embedded Tissues Aihua Liu and Xiaowei Xu Abstract MicroRNAs (miRNAs) are small (19–23nt), highly conserved noncoding RNAs that posttranscriptionally regulate target gene expression. Altered expression of miRNAs has been demonstrated in many different human diseases, including cancer. The large archives of formalin-fixed, paraffin-embedded (FFPE) tissue specimens with clinical follow-up information that exist are a highly valuable source of tissue to study human diseases. However, RNA in the FFPE tissue is fragmented and may be chemically modified. In this study, we prepared miRNA preserving total RNA from matched pairs of FFPE and respective freshfrozen clinical samples, and used that in microarray experiments to compare miRNA expression profiles between FFPE and fresh-frozen tissue from the same tissue samples. We demonstrate that miRNA expression profile from FFPE tissues closely resembles that from fresh tissues. These results underscore the suitability of FFPE tissues as appropriate resources for miRNA expression analyses. Key words: MicroRNA, Formalin-fixed paraffin-embedded (FFPE), Microarray profiling, qPCR
1. Introduction Formalin-fixed paraffin-embedded (FFPE) is the most common method of collection and storage of surgical specimens; these archived tissues are of significant value to retrospective studies of human diseases since they may have in-depth clinical follow-up information. The ability to isolate nucleic acid that is suitable for molecular analysis from these archived tissue samples provides a powerful tool to reveal the mechanism of disease at both the genomic and gene expression level (1). Historically, the nucleic acids isolated from FFPE tissues are typically fragmented and chemically modified to a degree that renders it incompatible with most molecular analysis. Recently, it was found that fragmentation of FFPE RNA seems to have a size endpoint of around 80nt, and modifications affect only about 1% of the nucleotides. Since Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_16, © Springer Science+Business Media, LLC 2011
259
260
Liu and Xu
miRNAs are small (19–23nt), they are less affected by the fixation/ embedding process (2, 3). With miRNAs recently taking center stage as a key regulator of mRNA and protein expression (4, 5), many endeavors were made to profile miRNA from FFPE tissues (6, 7). Several miRNA isolation kits for FFPE tissues are commercially available (8, 9), and there are a number of platforms that are commercially available for miRNAs profiling (8, 10, 11).
2. Materials 1. Tissues: FFPE tissues were sectioned at 20 mm using a fresh microtome blade for each block, four thick sections were used for RNA isolation. We found that it is very helpful to macrodissect the thick sections using a razor blade to trim off excessive surround tissue, which may help reduce signals from contaminating tissue. 2. RecoverALL total nucleic acid isolation kit (Ambion, Austin, Texas): This kit contains reagents for 40 isolations of total RNA from paraffin-embedded tissue. Amount Component
Storage
16 mL
Digestion buffer
Room temperature
60 mL
Wash 1 concentrate Room temperature Add 42 mL 100% ethanol before use
60 mL
Wash 2/3 concentrate Room temperature Add 48 mL 100% ethanol before use
80
Collection tubes
Room temperature
40
Filter cartridges
Room temperature
19.2 mL Isolation additive
Room temperature
5 mL
Elution Solution
Room temperature
160 mL
Protease
−20°C
240 mL
10× DNase buffer
−20°C
160 mL
DNase
−20°C
400 mL
RNase A
−20°C
3. 100% xylene, ACS grade or higher quality. 4. 100% ethanol for molecular biology. 5. Microtome for tissue sectioning. 6. Nuclease-free 1.5 mL microcentrifuge tubes. 7. Adjustable pipettors and RNase-free tips. 8. Microcentrifuge capable of at least 10,000 × g.
MicroRNA Isolation from Formalin-Fixed, Paraffin-Embedded Tissues
261
9. Incubators or heat blocks (deep well preferred) set at 50 and 85°C. 10. (Optional) Nonstick tubes for long-term storage of recovered nucleic acid. 11. (Optional) Centrifugal vacuum concentrator (e.g., SpeedVac) for drying deparaffinized tissue samples. 12. Spectrophotometer, such as the NanoDrop 1000 Spectro photometer. 13. Agilent 2100 bioanalyzer or reagents and apparatus for preparation and electrophoresis of agarose gels.
3. Methods miRNA containing total RNA Isolation. 3.1. D eparaffinization
1. Cut 20 mm sections from FFPE tissue blocks using a microtome. 2. Place the equivalent of £80 mm of tissue slices (i.e., a maximum of 4 × 20 mm) in a 1.5 mL microcentrifuge tube (see Note 1). 3. Add 1 mL of 100% xylene to the sample (see Note 2). 4. Vortex briefly to mix and centrifuge briefly to bring down the tissue. 5. Heat the sample for 3 min at 50°C to melt the paraffin. 6. Centrifuge the sample for 2 min at room temperature at a maximum speed to pellet the tissue. 7. Remove the xylene without disturbing the pellet. Discard the xylene. If the pellet is loose, leave some xylene in the tube to avoid removing any tissue pieces. 8. Add 1 mL of RT 100% ethanol to the sample and vortex to mix. 9. Centrifuge the sample for 2 min at a maximum speed to pellet tissue. 10. Remove and discard the ethanol without disturbing the pellet. The ethanol will contain trace amounts of xylene and must be discarded accordingly (see Note 3). 11. Repeat steps 7–10 above to wash second time with 1 mL of 100% ethanol. 12. Briefly centrifuge again to collect any remaining drops of ethanol in the bottom of the tube. Remove as much residual ethanol as possible without disturbing the pellet. 13. Air dry the pellet 15–30 min to remove residual ethanol.
262
Liu and Xu
3.2. P rotease Digestion
1. Add 400 mL digestion buffer to each sample. 2. Add 4 mL protease to each sample. 3. Swirl the tube gently to mix and to immerse the tissue. If tissue sticks to the sides of the tube, use a pipette tip to push it into the solution. 4. Incubate the sample in a heat block or water bath for 3 h at 50°C (see Note 4). Most samples will clarify within 3 h. If the sample mixture does not clarify after 3 h, avoid removing undigested tissue pieces when applying to the filter cartridge. Stopping Point Samples can be stored at −20°C, then thawed on ice before proceeding to total RNA isolation.
3.3. Total RNA Isolation
1. Add 480 mL isolation additive to each sample. 2. Vortex to mix. The solution should appear white and cloudy after vortexing. 3. Add 1.1 mL 100% ethanol to each sample. 4. Mix by pipetting up and down. The solution should become clear at this point. 5. For each sample, place a filter cartridge in one of the collection tubes supplied. 6. Pipet 700 mL of the sample/ethanol mixture (from step 3) onto the filter cartridge and close the lid. To prevent clogging of the filter, avoid pipetting large pieces of undigested tissue onto the filter cartridge. 7. Centrifuge at 10,000 × g for 30 s to pass the mixture through the filter (see Note 5). 8. Discard the flow-through, and re-insert the filter cartridge in the same . 9. Repeat steps 6–8 until all the sample mixture has passed through the filter. 10. Add 700 mL of Wash 1 to the filter cartridge. 11. Centrifuge for 30 s at 10,000 × g to pass the mixture through the filter. 12. Discard the flow-through, and re-insert the filter cartridge in the same collection tube. 13. Add 500 mL of Wash 2/3 to the filter cartridge. 14. Centrifuge for 30 s at 10,000 × g to pass the mixture through the filter.
MicroRNA Isolation from Formalin-Fixed, Paraffin-Embedded Tissues
263
15. Discard the flow-through, and re-insert the filter cartridge in the same collection tube. 16. Spin the assembly for an additional 30 s to remove residual fluid from the filter. 3.4. DNase Digestion and Final RNA Purification
Preheat Elution Solution or nuclease-free water to 95°C in a heat block before you start this step. 1. Combine the following solutions to make the DNase mix (a master DNase mix can be used if there is more than one sample). Amount (per reaction) (mL)
Component
6
10× DNase buffer
4
DNase
50
Nuclease-free water
2. Add 60 mL of the DNase mix to the center of each filter cartridge. 3. Cap the tube and incubate for 30 min at room temperature (22–25°C). 4. Add 700 mL of Wash 1 to the filter cartridge. 5. Incubate for 30–60 s at RT. 6. Centrifuge for 30 s at 10,000 × g. 7. Discard the flow-through and re-insert the filter cartridge in the same collection tube. 8. Add 500 mL of Wash 2/3 to the filter cartridge. 9. Centrifuge for 30 s at 10,000 × g. 10. Discard the flow-through and re-insert the filter cartridge in the same collection tube. 11. Repeat steps 8–10 to wash second time with 500 mL of Wash 2/3. 12. Centrifuge the assembly for 1 min at 10,000 × g to remove residual fluid from the filter. 13. Transfer the filter cartridge to a fresh collection tube. 14. Apply 30 mL of Elution Solution or nuclease-free water, preheated to 95°C, to the center of the filter, and close the cap (see Note 6). 15. Allow the sample to sit at room temperature for 1 min. 16. Centrifuge for 1 min at a maximum speed to pass the mixture through the filter.
264
Liu and Xu
17. Repeat steps 14–16 with a second 30 mL eluant to the same collection tube. The final volume should be close to 60 mL. 18. Store the nucleic acid at −20°C or colder (see Note 7). 3.5. (Optional) Further Purify the Recovered RNA.
Ethanol precipitation can further purify the recovered RNA if needed. 1. Add ammonium acetate to a final concentration of 2–2.5 M and mix well. 2. Add 4 volumes of ethanol and mix again. 3. Chill at −20°C or lower from 12 h to overnight. 4. Centrifuge at 16,000 × g for 20–30 min to pellet the RNA. 5. Wash the pellet twice with 80% ethanol. 6. Resuspend the RNA in nuclease-free water or TE (10 mM Tris–HCl, pH 8, 1 mM EDTA).
3.6. Assessing RNA Yield and Integrity 3.6.1. RNA Yield
The expected RNA yield varies greatly depending on the amount of tissue used, the tissue type, and the method used for fixing and storing the sample. 1. Spectrophotometry The concentration of an RNA solution can be determined by measuring its absorbance at 260 nm. We recommend using the NanoDrop 1000 Spectrophotometer because it is very quick and easy to use by measuring 1.5 mL of the RNA sample directly. Alternatively, the RNA concentration can be determined by diluting an aliquot of the preparation in TE (10 mM Tris–HCl, pH 8, 1 mM EDTA) and reading the absorbance in a traditional spectrophotometer at 260 nm. Calculate the RNA concentration (mg/mL) as follows: A260 × dilution factor × 40 mg/mL = mg RNA/mL. 2. Fluorometry If a fluorometer or a fluorescence microplate reader is available, Molecular Probe’s RiboGreen fluorescence-based assay for RNA quantitation is a convenient and sensitive way to measure RNA concentration.
3.6.2. RNA Quality
1. Spectrophotometry The A260/A280 ratio and A260/A230 of the RNA are indications of its purity. A260/A280 should be ³1.8–2.1 and A260/ A230 should be ³1.8 (see Note 8). 2. Agarose gel electrophoresis RNA recovered from FFPE tissues will typically appear smeared. In higher quality preparations, two broad bands representing 18S and 28S rRNA will be visible but most likely only a smear will be seen in most cases.
MicroRNA Isolation from Formalin-Fixed, Paraffin-Embedded Tissues
265
Fig. 1. Representative total RNA integrity analysis using Agilent 2100 bioanalyzer. Left upper panel: fresh-frozen human melanoma tissue; Left lower panel: matched FFPE tissue; Right panel: an agarose gel-like mimic. L ladder, F frozen tissue, P FFPE tissue.
3. Microfluidics analysis Agilent’s 2100 bioanalyzer used in conjunction with an RNA LabChip kit provides a powerful and sensitive method to assess RNA integrity. The data will mimic that seen on agarose gels. An example result is shown in Fig. 1. 3.7. miRNA Profiling Platforms
1. miRNA microarray There are a number of miRNA microarray platforms commercially available: miRCURY LNATM microRNA Array from Exiqon (7); Agilent miRNA array platform (11); NCodeTM miRNA Arrays from Invitrogen; Illumina miRNA expression panel and Affymetrix miRNA arrays. The amount of input RNA required is different when using these platforms. An example of our miRNA microarry results using miRCURY LNATM microRNA Array from Exiqon is shown in Fig. 2. 2. miRNA real-time PCR The PCR Array performs miRNA expression analysis with realtime RT-PCR sensitivity and the multisequence profiling capabilities of a microarray. TaqMan MicroRNA Arrays v2.0 and Megaplex™ Primer Pools from ABI enable a comprehensive miRNA expression profiling and its individual TaqMan® MicroRNA Assays provide the ideal tool for further confirmation and validation for miRNA profiling; The miRNA PCR Arrays from SABiosciences is a set of optimized real-time PCR assays, in 96-well or 384-well plates, for a miRNA panel as well as appropriate housekeeping assays and RNA quality controls.
266
Liu and Xu
Fig. 2. Pairwise Pearson’s correlations of miRNA expression between FFPE samples and matched fresh frozen. RNA from fresh frozen and matched FFPE tissues was profiled using a commercially available microarray platform. Mean miRNA intensity correlation between FFPE samples (indicated by P1, P2, P3, P4, P5) and fresh-frozen tissue samples (indicated by F1, F2, F3, F4, F5) is shown in the figure.
4. Notes 1. Overloading the tissue >80 mm will result in lower RNA yield and poor quality from incomplete protease digestion. 2. Using slices that are ³10 mm thickness and minimizing time in the xylene and ethanol washes will minimize the loss of miRNA. 3. Xylene is a toxic substance, handle it only in a well ventilated area using personal protection equipment. Dispose xylene waste according to applicable regulations. 4. Extending the incubation at 70°C for additional 15 min after 3 h at 50°C helps to release longer mRNA form. 5. Do not centrifuge filter cartridges at relative centrifugal forces greater than 10,000 × g to avoid damaging the filters. 6. The Elution Solution from the kit contains salts which may affect downstream applications if concentrated. If you intend to vacuum dry the sample, elute in nuclease-free water instead.
MicroRNA Isolation from Formalin-Fixed, Paraffin-Embedded Tissues
267
7. If a very small amount of RNA was recovered or to store your RNA sample for an extended period of time, transfer the eluate to a nonstick tube (e.g., P/N 12450) to prevent the loss of RNA. 8. Contaminating genomic DNA in the recovered RNA can be removed by further digestion of DNase.
Acknowledgment This work is supported by Specialized Program of Research Excellence (SPORE) on Skin Cancer (CA-093372). References 1. Cronin, M., Pho, M., Dutta, D., Stephans, J.C., Shak, S., Kiefer, M.C., et al. (2004) Measurement of gene expression in archival paraffin-embedded tissues: development and performance of a 92-gene reverse transcriptase-polymerase chain reaction assay. Am. J. Pathol. 164, 35–42. 2. Xi, Y., Nakajima, G., Gavin, E., Morris, C.G., Kudo, K., Hayashi, K., Ju, J. (2007) Systematic analysis of microRNA expression of RNA extracted from fresh frozen and formalinfixed paraffin-embedded samples. RNA. 13:1668–74. Epub 2007 Aug 13. 3. Li, J., Smyth, P., Flavin, R., Cahill, S., Denning, K., Aherne, S., et al. (2007) Comparison of miRNA expression patterns using total RNA extracted from matched samples of formalinfixed paraffin-embedded (FFPE) cells and snap frozen cells. BMC Biotechnol. 7, 36. 4. Calin, G.A. and Croce, C.M. (2006) MicroRNA signatures in human cancers. Nat. Rev. Cancer 6, 857–866. 5. Lu, J., Getz, G., Miska, E.A., AlvarezSaavedra, E., Lamb, J., Peck, D., et al. (2005) MicroRNA expression profiles classify human cancers. Nature 435, 834–838. 6. Nelson, P.T., Baldwin, D.A., Scearce, L.M., Oberholtzer, J.C., Tobias, J.W., and Mourelatos, Z. ( 2004) Microarray-based, high-throughput gene expression profiling of microRNAs. Nat. Methods 1, 155–161.
7. Liu, A., Tetzlaff, M., VanBelle, P., Elder, D., Feldman, M., Tobias, J., et al. (2009) MicroRNA expression profiling outperforms mRNA expression profiling in formalin-fixed paraffin-embedded tissues. Int. J. Clin. Exp. Pathol. 2, 519–527. 8. Tetzlaff, M.T., Liu, A., Xu, X., Master, S.R., Baldwin, D.A., Tobias, J.W., et al. (2007) Differential expression of miRNAs in papillary thyroid carcinoma compared to multinodular goiter using formalin fixed paraffin embedded tissues. Endocr. Pathol. 18,163–173. 9. Siebolts, U., Varnholt, H., Drebber, U., Dienes, H-P., Wickenhauser, C., and Odenthal, M. (2009) Tissues from routine pathology archives are suitable for microRNA analyses by quantitative PCR. J. Clin. Pathol. 62, 84–88. 10. Doleshal, M., Magotra, A.A., Choudhury, B., Cannon, B.D., Labourier, E., and Szafranska, A.E. (2008) Evaluation and validation of total RNA extraction methods for microRNA expression analyses in formalin-fixed, paraffinembedded tissues. J. Mol. Diagn. 10, 203–211. 11. Zhang, X., Chen, J., Radcliffe, T., LeBrun, D.P., Tron, V.A., and Feilotter, H. (2008) An array-based analysis of microRNA expression comparing matched frozen and formalin-fixed paraffin-embedded human tissue samples. J. Mol. Diagn. 10, 513–519.
wwwwwwwwwwwwwww
Chapter 17 Gene Expression Profiling of RNA Extracted from FFPE Tissues: NuGEN Technologies’ Whole-Transcriptome Amplification System Leah Turner, Joe Don Heath, and Nurith Kurn Abstract Gene expression profiling of RNA isolated from formalin fixed, paraffin-embedded (FFPE) tissue samples has been historically challenging. Yet FFPE samples are sought-after because of the in-depth retrospective records typically associated with them rendering these samples a valuable resource for translational medicine studies. Extensive degradation, chemical modifications, and cross-linking have made it difficult to isolate RNA of sufficient quality required for large-scale gene expression profiling studies. NuGEN Technologies’ WT-Ovation™ FFPE System linearly amplifies RNA from FFPE samples through a robust and simple whole-transcriptome approach using as little as 50 ng total RNA isolated from FFPE samples. The amplified material may be labeled with validated kits and/or protocols from NuGEN for analysis on any of the major gene expression microarray platforms, including: Affymetrix, Agilent, and Illumina gene expression arrays. Results compare well with those obtained using RNA from fresh-frozen samples. RNA quality from FFPE samples varies significantly and neither sample age nor sample size analysis via gel electrophoresis or the Agilent Bioanalyzer system accurately predict materials suitable for amplification. Therefore, NuGEN has validated a correlative qPCR-based analytical method for the RNA derived from FFPE samples which effectively predicts array results. The NuGEN approach enables fast and successful analysis of samples previously thought to be too degraded for gene expression analysis. Key words: FFPE, Gene expression, Isothermal linear amplification
1. Introduction Clinical tissues, and especially tumor biopsies, are typically processed and stored as formalin fixed, paraffin-embedded (FFPE) tissue blocks. While the tissue structure of such samples is maintained sufficiently for histological analysis, RNA samples isolated from these samples are frequently partially degraded and/or cross-linked with other molecules. Yet they have tremendous Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_17, © Springer Science+Business Media, LLC 2011
269
270
Turner, Heath, and Kurn
potential value because they are typically associated with complete patient histories of diagnoses, treatments, and outcomes, and as such are valuable for retrospective diagnostic, prognostic, and predictive molecular marker development. Gene expression profiles, or signatures, employing microarray platforms as genome-wide analytical tools, have been established in recent years as biomarkers for various disease states (1). Although gene expression analysis of RNA extracted from clinical FFPE samples has been shown to be suitable for RT-PCR analysis of selected gene panels (2, 3), genome-wide interrogation of gene expression profiles of FFPE-derived RNA has been a challenge due to the poor quality of the purified RNA from these samples. The NuGEN WT-Ovation FFPE System, based on Ribo-SPIA® linear whole-transcriptome amplification technology (4), overcomes these challenges by providing a robust linear amplification of total RNA from limited amounts of significantly degraded RNA, yielding amplified cDNA sufficient for genome-wide gene expression profiling using various microarray platforms (5). This system addresses the technical gap that has persisted for global gene expression analysis of these valuable sample sources, enabling researchers to take advantage of large FFPE archived sample sets and their associated clinical data for research studies. The performance of this whole-transcriptome linear amplification system for gene expression profiling from precious FFPE samples will be described. NuGEN Technologies’ Ribo-SPIA approach allows linear, isothermal-transcriptome amplification from total RNA. A key feature of the amplification technology is the replication of only the original template and not the replication of the copies. It is a simple, three-step process that can be completed in less than 6 hrs. The RNA templates are primed from both the poly-A tail of cellular mRNA and randomly primed across the transcripts. Therefore, the system can accommodate RNA samples that are quite degraded, including the loss of the poly-A tail. The WT-Ovation FFPE RNA Amplification solution starts with as little as 50 ng of total RNA and produces sufficient cDNA for analysis on any of the major gene expression platforms. Utilizing the appropriate NuGEN fragmentation and labeling modules and protocols, the amplified cDNA generated using the WT-Ovation FFPE System V2 can be used for analysis on most major microarray platforms for gene expression including: Affymetrix 3’, Gene-ST or Exon GeneChip arrays, Agilent Gene Expression microarrays, and Illumina Genome-Wide Expression BeadChips. This cDNA also enables the detection of gene transcripts using real-time quantitative PCR (qPCR). NuGEN’s WT-Ovation FFPE System and accompanying Encore Biotin Module and Ovation Exon Module provide useful tools for genome-wide microarray-based gene expression analysis
Gene Expression Profiling of RNA Extracted from FFPE Tissues
271
and biomarker discovery and development from FFPE oncology tissues (6–8) and allow the vast archives of valuable FFPE samples to be accessed for clinical research projects.
2. Materials 2.1. Equipment
1. Microcentrifuge for individual 1.5 and 0.5 mL tubes. 2. 0.5–10 mL pipette, 2–20 mL pipette, 20–200 mL pipette, and 200–1,000 mL pipette. 3. Vortexer. 4. Thermal cycler with 0.2 mL tube heat block, heated lid, and 100 mL reaction capacity. 5. Appropriate spectrophotometer and cuvettes, or Nanodrop® ND-1000 UV–Vis (NanoDrop Tech, Wilmington, DE). Spectrophotometer.
2.2. Reagents
1. NuGEN’s WT-Ovation FFPE System V2. 2. Ethanol for purification steps. 3. A variety of RNA isolation methods from FFPE samples are suitable for whole-transcriptome amplification using the NuGEN WT-Ovation FFPE System, dependent on sample source. Some of the RNA isolation procedures and products successfully used are Formapure™ Kits (Agencourt), and the High Pure FFPE RNA Micro Kit (Roche). 4. Multiple options exist for the purification of the final SPIA cDNA product. These include: RNAClean Beads (Agencourt), MinElute® Reaction Cleanup Kit (QIAGEN), QIAquick® PCR Purification Kit (QIAGEN), and DNA Clean and Concentrator™-25 (Zymo Research).
2.3. Supplies and Labware
1. Nuclease-free pipette tips. 2. 1.5 and 0.5 mL RNase-free microcentrifuge tubes. 3. 0.2 mL individual thin-wall PCR tubes or 8 × 0.2 mL strip PCR tubes or 0.2 mL thin-wall PCR plates. 4. Agencourt SPRIPlate® 96R, Ring Magnet Plate (Agencourt), or Agencourt SPRIPlate® Ring Super. 5. Magnet plate. Other magnetic stands may be used as well, although their performance has not been validated by NuGEN. 6. Disposable gloves. 7. Ice bucket. 8. Decontamination solutions such as RNaseZap® and DNAOFF™ (MP Biomedicals).
272
Turner, Heath, and Kurn
3. Methods 3.1. Ribo-SPIA Technology
NuGEN Technologies’ Ribo-SPIA approach allows linear, isothermal-transcriptome amplification from total RNA. A key feature of the amplification technology is the replication of the original template only and not the replication of the copies. It is a simple, three-step process that can be completed in less than 6 hrs. The WT-Ovation FFPE RNA Amplification solution starts with as little as 50 ng of total RNA and produces ample cDNA for analysis on any of the major gene expression platforms. For gene expression analysis on microarrays, the RNA must be isolated, amplified, and labeled appropriately for the platform of choice. Several methods are available for isolating RNA from the FFPE blocks. Once isolated, the RNA must be stored at −80°C and should ideally be amplified as soon as possible to prevent further degradation. In the first step (Fig. 1), first-strand cDNA is prepared from total RNA using a unique first-strand DNA/RNA chimeric primer mix and reverse transcriptase (RT). The primers have a DNA portion that hybridizes either to the 3′ portion of the poly-(A) sequence or randomly across the transcript. RT extends the 3′end of each primer generating first-strand cDNA. The resulting cDNA/RNA hybrid molecule contains a unique RNA sequence at the 5′-end of the cDNA strand. The second step involves the fragmentation of the RNA within the cDNA/RNA heteroduplex; thus, creating priming sites for DNA polymerase to synthesize a second strand. In this step, DNA complementary to the RNA portion of the chimeric primers, which is present at the 5′-end of the first-strand cDNA, is synthesized. The result is a double-stranded cDNA with a unique DNA/RNA heteroduplex at one end. The third step is a linear isothermal DNA amplification process, SPIA (single primer isothermal amplification), developed by NuGEN. It uses a SPIA DNA/RNA chimeric primer, DNA polymerase, and RNase H in a homogeneous isothermal reaction that provides highly efficient amplification of DNA sequences. RNase H degrades RNA in the DNA/RNA heteroduplex at the end of the dscDNA, which expose the DNA sequence at the 3′-end of the second-strand cDNA. This region is now accessible for binding a SPIA DNA/RNA chimeric primer. DNA polymerase then initiates synthesis at the 3′-end of the primer, displacing the existing forward strand. The RNA portion at the 5′-end of the newly synthesized strand, which hybridized to the template cDNA to be replicated, is again hydrolyzed by RNase H, exposing part of the unique priming site for initiation of the next round of cDNA synthesis. The process of SPIA DNA/RNA primer binding, DNA replication, strand displacement, and RNA cleavage is
Gene Expression Profiling of RNA Extracted from FFPE Tissues
273
Fig. 1. The Ribo-SPIA RNA amplification process used in the WT-Ovation™ FFPE System V2.
repeated, resulting in rapid accumulation of cDNA with sequence complementary to the original RNA transcripts. The WT-Ovation FFPE System V2 synthesizes 6 mg of amplified cDNA or more starting with total cellular RNA extracted from FFPE samples in input amounts of 50–100 ng in approximately 6 hrs. The yield of amplified cDNA from the total RNA derived from the FFPE sample is dependent on the overall quality of the RNA, which is highly variable. The efficiency of this amplification, as reflected by the yield of amplified cDNA is shown in Fig. 2, which was obtained from a study of a various FFPE RNA samples obtained from FFPE tissue sample of varying age (http://www.nugeninc.com/tasks/sites/nugen/assets/File/ technical_documents/techdoc_wt_ov_ffpe_rep_02.pdf). The size of the amplified cDNA products is directly proportional to the size of the degraded RNA used for the amplification reactions.
274
Turner, Heath, and Kurn Yields from 2-15 year old Ovarian FFPE RNA FFPE samples
10
8
Array minimum
cDNA yields (µ) µg
9
7 6 5 4 3 2 1
3
5
7
9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41
Sample # Fig. 2. Yield of WT-Ovation FFPE-amplified cDNA from total RNA (50 ng) derived from FFPE tissue samples. A total of 41 FFPE Ovarian tumor samples are shown. A minimum of 5.8 mg is required for gene expression analysis on HG-U133A 2.0 GeneChip arrays (shown by the horizontal bar ).
With the whole-transcriptome amplification approach, the size distribution of the starting RNA is far less important compared with a 3′ amplification strategy, since it results in densely over lapping cDNA fragments representing the entire transcriptome. The amplified cDNA obtained is suitable for gene expression profiling using a variety of platforms, including genome-wide expression analysis using microarrays. Target preparation for ana lysis on GeneChip arrays is performed using NuGEN’s Encore Biotin Module in combination with Ovation Exon Module when testing on GeneChip Exon or Gene-ST arrays. Gene expression analysis using the WT-Ovation FFPE-amplified cDNA has been demonstrated to be highly reproducible and maintain the integrity of the biological data (Fig. 3). 3.2. RNA Preparations
A variety of RNA isolation methods from FFPE samples are suitable for whole-transcriptome amplification using the NuGEN WT-Ovation FFPE System, dependent on the sample source. Formalin fixation of tissues results in a high degree of cross-linking of the RNA. It is essential to reverse these cross-links for successful cDNA synthesis and amplification. Many commercial RNA isolation kits designed for the extraction of RNA from FFPE tissues include a demodification step.
3.3. WholeTranscriptome Amplification of Total RNA from FFPE Samples
NuGEN Technologies’ WT-Ovation FFPE System contains all reagents necessary for first-strand cDNA synthesis, second-strand synthesis, and amplification in tubes with color-coded lids. Nuclease-free Water and Agencourt RNAClean Beads are also provided. Whole-transcriptome amplification of RNA from FFPE samples using the WT-Ovation FFPE System is carried out according to the instructions supplemented in the corresponding user
Gene Expression Profiling of RNA Extracted from FFPE Tissues
275
Fig. 3. Principle component analysis (PCA) of gene expression profiling of RNA derived from colon tumor and normal adjacent tissue (NAT) FFPE, and fresh-frozen samples amplified with WT-Ovation FFPE System and fragmented and biotin labeled using Encore™ Biotin Module. Targets were prepared from RNA extracted from formalin fixed, paraffin-embedded (FFPE) tissue obtained from one donor (left spheres) and freshfrozen tissue obtained from a second donor (spheres on the right). Each sample was amplified in quadruplicate, and expression profiles were analyzed using Affymetrix HG-U133A 2.0 GeneChip arrays. PCA was performed using Partek Genomics Suite software. The upper and lower ellipses (NAT and Tumor, respectively) define the boundary of 2 standard deviations from the centroid of each cluster indicating a statistically significant separation of samples based on the disease state of the tissue. This demonstrates that the amplification system maintains the integrity of the biological data.
guide (http://www.nugeninc.com/tasks/sites/nugen/assets/File/ user_guides/userguide_wt_ov_ffpe.pdf). The WT-Ovation FFPE System provides optimized reagents and methods for robust whole-transcriptome amplification of RNA purified from FFPE samples which was validated on numerous FFPE RNA samples of varying age and tissue sources. The Ribo-SPIA amplification process used in the WT-Ovation FFPE System V2 is performed in three stages: first-strand cDNA synthesis (1.2 h), second-strand cDNA synthesis and purification (2 h), SPIA isothermal linear amplification 1 (0.5 h), and SPIA isothermal linear amplification 2 and purification (2 h). We do not recommend stopping at any intermediate stage of the protocol. Recommendations for preventing cross-over contamination are provided in Subheading 4 (see Note 1).
276
Turner, Heath, and Kurn
WT-Ovation FFPE System V2 components are color-coded, with each color linked to a specific stage of the process. Performing each stage requires making a master mix then adding it to the reaction, followed by incubation. Master mixes are prepared by mixing components provided for that stage. 3.4. A Summary of the WT-Ovation FFPE Protocol Is as Follows 3.4.1. First-Strand cDNA Synthesis
1. Obtain Nuclease-Free Water D1 (green cap vial) from −20°C and leave at room temperature. 2. For each assay, place 2 mL of First-Strand Primer Mix A1 into a 0.2 mL PCR tube and place on ice. 3. Add 5 mL of total RNA (50 ng) to the primer, flick tubes to mix and spin. 4. Place the tubes in a thermal cycler running Program 1 (65°C – 2 min, 4°C – forever). When cycler reaches 4°C, spin and place tubes on ice. 5. Make First-Strand Master Mix. Per sample combine: 2.5 mL Buffer Mix A2 + 0.5 mL Enzyme Mix A3. Mix the First-Strand Master Mix, spin, and place on ice. 6. Add 3 mL of the First-Strand Master Mix to each tube, mix, and spin. 7. Place the tubes in a thermal cycler running Program 2 (4°C – 2 min, 25°C – 30 min, 42°C – 15 min, 70°C – 15 min, 4°C – forever). Once the thermal cycler reaches 4°C, spin and place tubes on ice.
3.4.2. Second-Strand cDNA Synthesis
1. Resuspend the RNAClean® beads provided with the WT-Ovation kit and leave at room temperature. Thaw the Second-Strand Reagents (Set B, yellow cap vials). Mix each reagent, spin, and place on ice. 2. Make Second-Strand Master Mix. Per sample combine: 9.75 mL Buffer Mix B1 + 0.25 mL Enzyme Mix B2. Mix the Second-Strand Master Mix, spin, and place on ice. 3. Add 10 mL of Second-Strand Master Mix to each first-strand reaction tube, mix, and spin. Place the tubes in a thermal cycler running Program 3 (4°C – 1 min, 25°C – 10 min, 50°C – 30 min, 70°C – 5 min, 4°C – forever). Once the thermal cycler reaches 4°C, spin and place tubes on ice.
3.4.3. Purification of Double-Stranded cDNA
1. Ensure the magnetic bead suspension has reached room temperature. Mix the bead suspension by inverting several times. 2. At room temperature, add 32 mL of the bead suspension to each tube, mix, and incubate at room temperature for 10 min.
Gene Expression Profiling of RNA Extracted from FFPE Tissues
277
3. Transfer samples to magnet, let stand for 5 min, remove 45 mL of the binding buffer. Minimize bead loss at this. 4. Add 200 mL of freshly prepared 80% ethanol, let stand for 30 s, then remove the ethanol using a pipette. 5. Repeat the ethanol wash two times, remove all excess ethanol, and let beads air dry for 15–20 min. Ensure all residual ethanol is removed from the sample. 6. Proceed immediately with SPIA® amplification, with the cDNA bound to the dry beads. 3.4.4. SPIA Amplification
1. Thaw the SPIA Amplification Reagents (Set C, red cap vials). Vortex C1, C2, and C5. Gently pipette the mix C3 and C6. Spin and place on ice. 2. Make SPIA Master Mix 1. Per sample combine: 50 mL C2 + 2 0 mL C1 + 0.7 mL C6, mix, and then add 10 mL C3. Mix well, spin, and place on ice. 3. On ice, add 80 mL of SPIA Master Mix 1 to each secondstrand reaction tube, mix, and spin. 4. Place tubes in a thermal cycler running Program 4 (4°C – 1 min, 47°C – 30 min, 4°C – forever). Once the thermal cycler reaches 4°C, spin and place tubes on ice. 5. Make SPIA Master Mix 2. Per sample combine: 30 mL C2 + 2 0 mL C5 + 2.3 mL C6, mix, and then add 30 mL C3. Mix well, spin, and place on ice. 6. On ice, add 80 mL of SPIA Master Mix 2 to each reaction tube and mix, transfer 80 mL to a second reaction tube. 7. Place tubes in a thermal cycler running Program 5 (4°C – 1 min, 47°C – 60 min, 95°C – 5 min, 4°C – forever). Once the thermal cycler reaches 4°C, spin and place tubes on ice. 8. Proceed immediately to purification step or store SPIA cDNA at −20°C.
3.5. Purification of Amplified cDNA Protocol
Amplified cDNA products can be purified using various methods. Purification is required if the amplified cDNA is intended for use in fragmentation and labeling reactions. Selection of the optimum purification option can depend on many factors. NuGEN recommends that the amplified cDNA products be purified prior to qPCR analysis. There are four currently supported alternatives for carrying out the final purification of amplified cDNA. Listed alphabetically, they are: (1) Agencourt RNAClean Magnetic Beads, (2) the Qiagen MinElute Reaction Cleanup Kit, (3) the Qiagen QIAQuick PCR Purification Kit, and (4) the Zymo Clean and Concentrator-25. These procedures are specifically adapted for use with NuGEN products and may differ significantly from the protocols published
278
Turner, Heath, and Kurn
by the manufacturers. Failure to follow the purification procedures as given may negatively impact your results. Detailed description of the purification protocols is provided in the WT-Ovation FFPE System user guide (http://www.nugeninc.com/tasks/sites/ nugen/assets/File/user_guides/userguide_wt_ov_ffpe.pdf). 3.6. Measuring cDNA Product Yield and Purity
1. Mix your sample by brief vortexing and spinning prior to checking the concentration. 2. Measure the absorbance at 260, 280, and 320 nm of your amplified cDNA product. You may need to make a 1:20 dilution of the cDNA in water prior to measuring the absorbance. 3. Purity: Subtract the Abs320 value from both Abs260 and Abs280 values. The adjusted (Abs260 − Abs320/Abs280 − Abs320) ratio should be > 1.8. 4. Yield: Assume one absorbance unit at 260 nm of singlestranded DNA = 33 mg/mL. To calculate: (Abs260 − Abs320 of diluted sample) × (dilution factor) × 33 (concentration in mg/mL of a one absorbance unit solution) × 0.03 (final volume in mL) = total yield in micrograms. 5. Alternatively, you may measure the concentration and purity of cDNA with a Nanodrop, using one absorbance unit at 260 nm of single-stranded DNA = 33 mg/mL as the constant. 6. The purified cDNA may be stored at −20°C.
3.7. RNA Amplification and Target Preparation for Transcriptome Analysis Using Affymetrix GeneChip Arrays
Target preparation for Affymetrix GeneChip arrays is performed using 5 mg of amplified cDNA as an input to the NuGEN Encore™ Biotin Module following this product’s user guide (http://www. nugeninc.com/tasks/sites/nugen/assets/File/user_guides/ userguide_encore_biotin.pdf). The fragmented and biotin-labeled cDNA targets are hybridized to HG-U133A 2.0 or HG-U133A Plus 2.0 Affymetrix GeneChip arrays, stained with streptavidin-phycoerythrin with antibody amplification, and scanned (with GeneChip Fluidics Station 450 and GeneChip Scanner 3000) following manu facturers’ protocols (Affymetirx Expression Analysis Technical Manual). Array data can be analyzed using Expression Console (Affymetrix), BioConductor 1.9 package, and Partek Genomics Suite. Pathway and functional analysis can be performed using Ingenuity Pathway Analysis software.
Gene Expression Profiling of RNA Extracted from FFPE Tissues
279
3.8. RNA Amplification and Target Preparation for Transcriptome Analysis Using Affymetrix GeneChip Exon and Gene-ST Arrays
The WT-Ovation Exon Module enables global gene expression analysis of amplified cDNA from FFPE RNA samples on Affymetrix GeneChip Exon and Gene-ST arrays. Following purification of amplified cDNA (as above), 3 mg is used as input to the WT-Ovation Exon Module for the generation of sense-strand products. The WT-Ovation Exon Module provides optimized reagent mixes and protocols for the generation of sense-strand targets when preformed according to the manufacturer’s user guide (http://www.nugeninc.com/ tasks/sites/nugen/assets/File/user_guides/userguide_wt_ ov_exon.pdf). Fragmentation and biotin labeling of product of the WT-Ovation Exon Module for the preparation of targets for hybridization and analysis of expression profiling using GeneChip Exon and Gene-ST arrays are carried out using NuGEN’s Encore Biotin Module as described in the manu facturer’s user guide (http://www.nugeninc.com/tasks/sites/ nugen/assets/File/user_guides/userguide_encore_biotin.pdf). The fragmented and biotin-labeled ST-cDNA generated by the combination of the WT-Ovation FFPE System, the WT-Ovation Exon Module, and the Encore Biotin Module are hybridized to the GeneChip Exon or Gene-ST arrays, and the arrays are scanned according to the manufacturer’s instructions. Array analysis is carried out using the Expression Console analysis package.
3.9. Quantitative PCR on WT-Ovation FFPE System Amplified cDNA
The WT-Ovation FFPE System may be used as a method of preamplification prior to qPCR. Although it is not absolutely necessary for qPCR applications, if quantification of the amplified cDNA product is desired, the cDNA must be purified prior to spectrophotometric quantification. Furthermore, different amplified cDNA samples may be variable in concentration; the purified products can be quantified and mass normalized to ensure that the cDNA inputs to qPCR are equal for all samples. Purified amplified cDNA produced with the WT-Ovation FFPE System has been successfully used as templates for qPCR systems including TaqMan® and SYBR® Green. Various aspects of quantification of amplified cDNA product using qPCR are described in Subheading 4 (see Note 2). 1. Real-time PCR master mixes containing the enzyme uracil N-glycosylase (UNG) are not compatible with the WT-Ovation FFPE System. Reagents for qPCR: (a) TaqMan: ABsolute qPCR Mix plus ROX (ABgene) or Fast Universal PCR Master Mix 2× (Applied Biosystems). (b) SYBR: QuantiTect SYBR Green PCR Kit (QIAGEN) or iQ SYBR Green Supermix (BioRad) or FastStart SYBR Green Master (ROX) (Roche).
280
Turner, Heath, and Kurn
Dilution of the amplified product After purification and quantification of amplified cDNA, it can be diluted to an appropriate concentration for qPCR. We recommend using 20 ng of cDNA in a 20 mL TaqMan reaction and 2 ng of cDNA for a 20 mL SYBR Green reaction. Depending on the abundance of the transcripts of interest, you may wish to use more or less cDNA. Primer design Designing multiple assays across the length of the transcript is recommended since the starting FFPE RNA is likely to be highly degraded. Using primers and probes designed with as small an amplicon size as possible is also recommended, due to the degraded nature of the input RNA. Primers may be designed at any position along a transcript since the WT-Ovation amplification covers the entire length of transcripts. References 1. van’t Veer, J.L., Paik, S., and Hayes, D.F. (2005) Gene expression profiling of breast cancer: a new tumor marker. J. Clin. Oncol. 23, 1631–1635. 2. Specht, K., Richter, T., Müller, U., Walch, A., Werner, M., and Höfler, H. (2001) Quantitative gene expression analysis in microdissected archival formalin-fixed and paraffinembedded tumor tissue. Am. J. Pathol. 158, 419–429. 3. Cronin, M., Pho, M., Dutta, D., Stephans, J.C., Shak, S., Kiefer, M.C., et al. (2004) Measurement of gene expression in archival paraffin-embedded tissues: development and performance of a 92-gene reverse transcriptasepolymerase chain reaction assay. Am. J. Pathol. 164, 35–42. 4. Kurn, N., Chen, P., Heath, J.D., Kopf-Sill, A., Stephens, K.M., and Wang, S. (2005) Isothermal, linear nucleic acid amplification systems for highly multiplexed applications. Clin. Chem. 51, 10. 5. Scicchitano, M.S., Dalmas, D.A., Bertiaux, M.A., Anderson, S.M., Turner, L.R., Thomas, R.A., et al. ( 2006) Preliminary comparison of
quantity, quality, and microarray performance of RNA extracted from formalin-fixed, paraffin-embedded, and unfixed frozen tissue samples. J. Histochem. Cytochem. 54,1229–1237. 6. Linton, K.M., Hey, Y., Dibben, S., Miller, C.J., Freemont, A.J., Radford, J.A., et al. (2009) Methods comparison for high-resolution transcriptional analysis of archival material on Affymetrix Plus 2.0 and Exon 1.0 microarrays. BioTechniques 47, 587–596. 7. Kennedy, R.D., Winter, A., Davison, T., McDyer, F., Proutski, V., Wilson, C., et al. (2009) Molecular subtyping of colon cancer from FFPE tissue. ASCO, http://www.almacgroup.com/papers/Papers/Molecular%20 Subtyping%20of%20Colon%20Cancer%20 ASCO%202009.pdf. 8. Chung, C.H., Aulino, J., Muldowney, N.J., Hatakeyama, H., Baumann, J., Burkey, B., et al. (2010) Nuclear factor-kappa B pathway and response in a phase II trial of bortezomib and docetaxel in patients with recurrent and/or metastatic head and neck squamous cell carcinoma. Ann. Oncol. 21, 864–870.
Chapter 18 Protein Mass Spectrometry Applications on FFPE Tissue Sections Carol B. Fowler, Timothy J. O’Leary, and Jeffrey T. Mason Abstract Formalin-fixed, paraffin-embedded (FFPE) tissue archives and their associated diagnostic records represent an invaluable source of proteomic information on diseases where the patient outcomes are already known. Over the last few years, advances in methodology have made it possible to recover peptides from FFPE tissues that yield a reasonable representation of the proteins recovered from identical fresh or frozen specimens. These new methods, based largely upon heat-induced antigen retrieval techniques borrowed from immunohistochemistry, have developed sufficiently to allow at least a qualitative analysis of the proteome of FFPE archival tissues. This chapter describes the approaches for performing proteomic analysis on FFPE tissues by liquid chromatography and mass spectrometry. Key words: Formalin-fixed, paraffin-embedded, Proteolytic digestion, Proteomics, Liquid chromatography, Tandem mass spectrometry, Protein extraction, Gel electrophoresis, Antigen retrieval, Histology
1. Introduction Efforts at proteomic analysis of formalin-fixed, paraffin-embedded (FFPE) tissues using whole proteins have proven largely unsuccessful (1). This arises from the extensive protein cross-linking produced when tissues are treated with formaldehyde. Intact proteins can be identified in FFPE tissue sections by immunohistochemistry (2), but this requires the availability of a specific antibody and an a priori knowledge of the protein of interest and thus is an inadequate approach for biomarker discovery. However, methods adapted from immunohistochemistry, specifically heatinduced antigen retrieval, have enabled qualitative analysis of archival tissue proteomes by facilitating the proteolytic release of peptides from FFPE tissues (3–13). This approach is successful Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_18, © Springer Science+Business Media, LLC 2011
281
282
Fowler, O’Leary, and Mason
because only a small percentage of the formaldehyde-reactive amino acids in a given protein form irreversible chemical modifications. Further, the amino acids involved in these irreversible modifications will vary for the individual members of a given protein population. Thus, the entire unmodified sequence of a given protein is likely to be present in the proteolytic digest of an FFPE tissue specimen. A systematic comparison of published extraction techniques reiterates the importance of heat, detergent, a protein denaturant, and physical agitation for efficient protein extraction from FFPE tissues (11, 14). The work flow for many published protocols has a number of common elements and is represented graphically in Fig. 1. Typically, sections cut from a FFPE tissue block are cleared of paraffin, rehydrated, and placed in an appropriate extraction buffer. The tissue sections are briefly homogenized by sonication or with a pestle, and heated. The crude protein extracts are separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Individual gel bands are then excised and prepared for mass spectrometry by in-gel digestion with trypsin. Alternately, the protein extracts may be directly digested in solution with trypsin for downstream analysis by liquid chromatography/tandem mass spectrometry (LC/MS/MS). Other separation and analysis techniques may also be used, such as 2-D gel electrophoresis and matrix-assisted laser-desorption ionization (MALDI) imaging mass spectrometry (15, 16); however, a complete review is beyond the scope of this chapter. Therefore, this chapter focuses on the unique methods for extracting proteins and peptides from FFPE tissue.
Fig. 1. A generalized proteomics work flow for the extraction and identification of proteins in FFPE tissue. Methods marked with an asterisk are not described in the text.
Protein Mass Spectrometry Applications on FFPE Tissue Sections
283
2. Materials All reagents may be purchased from Sigma-Aldrich (St. Louis, MO) unless otherwise noted. 2.1. Tissue Sectioning and Dewaxing
1. Polyethylene microcentrifuge tubes (1.5 mL) from Fisher Scientific (Pittsburgh, PA). 2. Xylene (Fisher) (see Note 1). 3. Ethanol (100%, Fisher) used to prepare the graded series of alcohols (100, 85, 70%) for tissue rehydration. 4. Deionized water.
2.2. Protocol for Extracting Proteins from FFPE Tissue Sections
1. Tris extraction buffer: 10 mM Tris–HCl, pH 7, with 2% (w/v) sodium dodecyl sulfate (SDS) (8).
2.3. SDS-PAGE of Extracted Proteins
1. Tris buffer for separating gel: 1.5 M Tris–HCl, pH 8.8. Store at room temperature.
2. Alternate extraction buffers (12) (see Note 2): (a) 6 M guanidine–HCl; or (b) 20 ng/mL sequencing-grade modified trypsin (Promega, Madison, WI) in 50 mM ammonium bicarbonate, pH 8.
2. Buffer for stacking gel: 1 M Tris–HCl, pH 6.8. Store at room temperature. 3. 10% (w/v) Sodium dodecyl sulfate (SDS) solution. Store at room temperature. 4. 30% Acrylamide/bis acrylamide solution. Prepare a stock solution with 29% (w/v) acrylamide and 1% (w/v) N,N¢methylene-bis-acrylamide in deionized water. Acrylamide is a neurotoxin when unpolymerized, so care should be taken to minimize exposure (see Note 3). 5. N,N,N,N¢-Tetramethylethylenediamine (TEMED, Bio-Rad, Hercules, CA; see Note 4). 6. Ammonium persulfate: prepare a 10% (w/v) solution in water and immediately freeze in single-use (200 mL) aliquots at −20°C. 7. Water or t-amyl alcohol for overlaying the separating gel. 8. Running buffer (5×): 125 mM Tris–HCl, 960 mM glycine, 0.5% (w/v) SDS. Store at room temperature. 9. SDS gel-loading buffer (4×): 200 mM Tris–HCl, pH 6.8, 8% (w/v) SDS, 0.4% (w/v) bromophenol blue, 40% (v/v) glycerol. Store at room temperature. Add 10× reducing buffer just before use. 10. Reducing buffer (10×): 1 M dithiothreitol (DTT).
284
Fowler, O’Leary, and Mason
11. Molecular-weight standards: Benchmark Protein Ladder (Invitrogen, Carlsbad, CA). 12. Methanol:acetic acid solution: 500 mL of methanol, 400 mL of deionized water, 100 mL of glacial acetic acid (Fisher). 13. Coomassie staining solution: Prepare by dissolving 0.25 g of Coomassie Brilliant Blue R-250 (Sigma-Aldrich) per 100 mL of the methanol:acetic acid solution (see Note 5). 2.4. In-Gel Tryptic Digestion of SDS-PAGE Separated Samples
1. Destaining solution: 50:5:45 methanol:acetic acid:deionized water (v/v/v). 2. 100 mM Ammonium bicarbonate in distilled water: 0.158 g in 20 mL of water. 3. 50 mM Ammonium bicarbonate in distilled water. 4. Mass spectrometry or HPLC-grade acetonitrile. 5. 10 mM Dithiothreitol (DTT): 1.5 mg/mL in 100 mM ammonium bicarbonate. 6. 50 mM Iodoacetamide (IAA): 10 mg/mL in 100 mM ammonium bicarbonate (see Note 6). 7. Trypsin solution (kept on ice): 20 ng/mL sequencing-grade modified trypsin (Promega, Madison, WI) in 50 mM ammonium bicarbonate, pH 8.1. 8. Extraction solution: 5:50:45 formic acid:acetonitrile:water (v/v/v). 9. Polyethylene microcentrifuge tubes (0.5 and 1.5 mL, Fisher). Rinse all tubes with water and then with ethanol, repeating at least once. Thorough washing removes surface contaminants, including acetonitrile-soluble material, which forms a layer on aqueous solutions and interferes with evaporation.
2.5. In-Solution Digestion of Protein Extracts
1. 100 mM Ammonium bicarbonate in distilled water. 2. 10 mM Dithiothreitol (DTT) in 100 mM ammonium bicarbonate. 3. 50 mM Iodoacetamide (IAA) in 100 mM ammonium bicarbonate (see Note 6). 4. Sequencing-grade modified trypsin (Promega, Madison, WI). 5. Microcon YM-3 spin filters (Millipore, Billerica MA, optional).
2.6. Liquid Chromatography/ Tandem Mass Spectrometry Analysis
1. Solvent A: 0.1% formic acid in HPLC-grade water (Fisher). 2. Solvent B: 0.1% formic acid in HPLC-grade acetonitrile (Fisher).
Protein Mass Spectrometry Applications on FFPE Tissue Sections
285
3. Methods 3.1. Tissue Sectioning and Dewaxing
1. Cut four 10-mm tissue sections from a FFPE tissue block using a microtome. Any excess paraffin is trimmed away and the sections are transferred to a microcentrifuge tube (see Note 7). A fresh microtome blade should be used to prevent carryover between samples. 2. The sections are incubated through two changes of 1 mL of xylene (3 min each) to clear any remaining paraffin. 3. The tissue is then rehydrated by passing the sections through a series of graded alcohols for 3 min each: 100 (´2), 85, and 70% ethanol (1 mL per wash). 4. The sections are rinsed thoroughly in deionized water and then transferred to fresh 1.5-mL microcentrifuge tubes.
3.2. Protocol for Extracting Proteins from FFPE Tissue Sections
1. Deparaffinized and rehydrated tissue sections are lysed in extraction buffer [10 mM Tris–HCl, pH 7, with 2% (w/v) SDS] using a 1:10 sample-to-buffer ratio (see Note 8). 2. The tissue is homogenized briefly on ice with a disposable pellet pestle (Kontes Scientific, Vineland, NJ) and then sonicated briefly (5–10 s on ice) (see Note 9). 3. The samples are incubated at 100°C on a heating block for 20 min and then at 60°C for 2 h. Clarify the samples by centrifugation prior to further processing and separation. For alternate extraction methods (see Note 2).
3.3. SDS-PAGE of Extracted Proteins
These instructions assume the use of a Bio-Rad mini-PROTEAN 3 gel system (Bio-Rad, Hercules, CA). They are easily adaptable to other formats. It is critical that the glass plates for the gels are scrubbed clean with a rinsable detergent (e.g., Alconox; New York, NY). Rinse extensively with distilled water and 100% isopropanol. Air-dry prior to use and assemble the gel-casting station according to the manufacturer’s instructions. 1. This recipe should be sufficient for most minigel systems with total volumes of 5–6 mL. Prepare a 1.0-mm thick, 10% gel by mixing 1.3 mL of 1.5 M Tris–HCl, pH 8.8, with 1.7 mL of acrylamide/bis solution, 1.9 mL of deionized water, 0.1 mL of 10% (w/v) SDS, 50 mL of ammonium persulfate solution, and 2 mL of TEMED. Pour the gel, leaving space for a stacking gel (the length of the gel comb plus 3–5 mm). Overlay with t-amyl alcohol or water. If water is used, add it slowly and evenly to prevent mixing. The gel should polymerize in about 30 min. 2. Pour off the t-amyl alcohol or water and rinse the top of the gel twice with water. Drain the water well and wick any excess water with a piece of filter paper.
286
Fowler, O’Leary, and Mason
3. Prepare the stacking gel by mixing 0.25 mL of 1.0 M Tris–HCl, pH 6.8, with 0.33 mL of acrylamide/bis acrylamide solution, 1.4 mL of water, 20 mL of 10% (w/v) SDS, 20 mL of ammonium persulfate solution, and 2 mL of TEMED. Use about 0.5 mL of this mixture to quickly rinse the top of the gel and then pour the stacking gel and insert the comb. The stacking gel should polymerize within 30 min (see Note 10). 4. Prepare the running buffer by diluting 200 mL of the 5× running buffer with 800 mL of water in a measuring cylinder. Cover and invert to mix. 5. Once the stacking gel has set, carefully remove the comb. Remove the gel from the casting station and assemble the gel unit. Add the running buffer to the upper and lower chambers of the gel unit, and use a 3-mL syringe fitted with a 22-gage needle to wash the gel wells with running buffer. 6. Use a standard protein assay (see Note 11) to determine the total protein content of the sample. Add 4× sample buffer and 10× reducing buffer to 20–60 mg of each protein extract and adjust the volume to 20 mL with water. Load each sample into a well. Use one well for the molecular-weight marker (5–10 mL). 7. Complete the assembly of the gel unit and connect it to a power supply. Apply power and begin electrophoresis; 200 V constant for 35 min is recommended for SDS-PAGE. The dye front (blue) can be run off the gel if desired. 8. Once electrophoresis is complete, remove the gel from the gel cassette and rinse twice in deionized water and once in the methanol:acetic acid solution. Add enough coomassie staining solution to cover the gel and incubate on an orbital shaker for 2–4 h. Once bands appear, decant the staining solution and destain the gel in the methanol:acetic acid solution. This will require several changes of destaining solution over 2–16 h. Placing a lint-free laboratory wipe, such as a Kimwipe, in the destaining bath helps absorb excess dye. 3.4. In-Gel Tryptic Digestion of SDS-PAGE Separated Samples
Adapted from Shevchenko et al. (17). 1. Using a clean razor blade or scalpel, cut each vertical gel lane into individual strips. Then, cut each of these gel lanes into horizontal bands ~2 mm wide. Mince each band into small pieces (~1 × 1 mm) and place all of the pieces from each lane into separate 1.5-mL polyethylene microcentrifuge tubes (see Note 12). Alternately, you may select individual gel bands of interest and place them in separate tubes. 2. Incubate the gel pieces in 500 mL of destaining solution overnight at room temperature.
Protein Mass Spectrometry Applications on FFPE Tissue Sections
287
3. Remove the destaining solution and replace it with 200 mL of fresh destaining solution for 2–3 h. 4. Discard the destaining solution and dehydrate the gel slices in 200 mL of acetonitrile. The gel pieces should turn opaquewhite within 5 min. If they do not, remove the acetonitrile and add 200 mL of fresh acetonitrile. 5. Discard the acetonitrile and evaporate any residual acetonitrile under vacuum (2–3 min) (see Note 13). 6. Reduce the gel pieces in 30–50 mL of 10 mM DTT for 30 min at room temperature. Remove the DTT solution. 7. Alkylate in 30–50 mL of 50 mM IAA for 30 min at room temp in the dark. Remove the IAA solution. 8. Wash with 100 mL of 100 mM ammonium bicarbonate for 10 min. Remove the wash and briefly invert the tube on a clean piece of filter paper to dry the gel pieces. 9. Dehydrate the gel slices in 200 mL of acetonitrile. If the gel pieces do not turn opaque-white within 5 min, remove the acetonitrile and add 200 mL of fresh acetonitrile. 10. Remove the acetonitrile and rehydrate the gel pieces by swelling in 100 mL of 100 mM ammonium bicarbonate for 10 min. Remove the ammonium bicarbonate. 11. Dehydrate the gel slices in 200 mL of acetonitrile. Remove the acetonitrile and repeat the dehydration step. 12. Dry the gel pieces under vacuum (2–3 min). 13. Prepare the trypsin solution: 20 mg of Promega trypsin in 1 mL of ice-cold 50 mM ammonium bicarbonate (trypsin concentration = 20 ng/mL). Keep the trypsin solution on ice. 14. Add enough trypsin solution (30–50 mL) to cover the gel pieces and incubate for 5–10 min on ice. The gel pieces should swell to their original size prior to dehydration. If not, extend the incubation time or add additional trypsin solution. (The idea is to allow the trypsin to penetrate into the gel, but not to begin digestion.) 15. Remove any excess trypsin solution and add 5–20 mL of 50 mM ammonium bicarbonate (enough to cover the gel pieces). Cap the microcentrifuge tubes and allow the samples to incubate overnight at 37°C. 16. Extract the peptides from the digested gel pieces by adding 30 mL of 100 mM ammonium bicarbonate. Vortex well and incubate for 10 min at room temperature. Centrifuge at 14,000 × g for 10 min and then transfer the supernatant to a clean 0.5-mL microcentrifuge tube using a clean pipette tip.
288
Fowler, O’Leary, and Mason
17. Repeat the extraction two more times (with 30 mL of ammonium bicarbonate buffer each time) and add the additional two supernatants to the first. 18. Reduce the total sample volume to about 20 mL by evaporation under vacuum (see Note 13). 3.5. In-Solution Digestion of Protein Extracts (see Note 14)
1. Determine the protein concentration in your sample using a standard protein assay (see Note 11). The amount of sample needed for MS analysis varies by technique, but generally, 2–5 mg total protein per analysis is sufficient for capillary liquid chromatography/tandem mass spectrometry (LC/MS/ MS) and 1–2 mg of total protein per analysis is sufficient for nano-flow LC/MS/MS. 2. Reduce the sample by adding an aliquot of the 10 mM DTT solution equal to 1/10 of the sample volume. Incubate for 30 min at 95°C. Remove the sample from the heating block and allow it to cool to room temperature. 3. Add an aliquot of the 10 mM IAA solution equal to 1/10 of the sample volume. Incubate the sample for 20 min in the dark. 4. If desired, the reduced and alkylated sample may be washed using a Microcon YM-3 spin filter according to the manufacturer’s instructions. Wash the protein sample three times using 300 mL of 100 mM ammonium bicarbonate for each wash. 5. Add sequencing-grade trypsin so that the ratio of protein:trypsin is 20:1, and incubate overnight at 37°C.
3.6. LC/MS/MS analysis
1. This protocol assumes the use of an Agilent 1100 series capillary liquid chromatography (LC) system coupled to an Agilent 6300 series linear ion trap mass spectrometer (MS) (Palo Alto, CA). The tryptic peptides are separated on a 0.3-mm (inner diameter) × 15-cm long Zorbax stable bond capillary column packed with 3.5-mm, 300-Å pore-size C8 media (Agilent). 2. A binary gradient consisting of 0.1% formic acid in water (A) and 0.1% formic acid in acetonitrile (B) is used as the mobile phase. After injecting 5–8 mL of sample (2–5 mg of total protein), wash the column with 2% solvent B for 30 min at a flow rate 5 mL/min (see Note 15). 3. The peptides are then eluted at 10 mL/min using the following discontinuous gradient program: 2–60% solvent B over 100 min, 60–98% solvent B over 20 min, and 98% solvent B for 20 min. 4. Equilibrate the column with 2% solvent B for 30 min prior to subsequent sample loading using a flow rate of 10 mL/min. 5. The MS is operated in data-dependent positive ion mode with precursor ion scans acquired from 50 to 2,200 m/z. The 3–7
Protein Mass Spectrometry Applications on FFPE Tissue Sections
289
most intense ions detected in each precursor scan are selected for tandem MS in the linear ion trap. The threshold for tandem MS is set at ~5% signal intensity relative to the instrument background. Helium is used as the collision gas, and the optimum collision energy is selected automatically based upon the charge and mass of the precursor ion. Tandem MS spectra are accumulated from 200 to 2,200 m/z for 150 ms, and three such scans are averaged to improve the quality of the final spectrum. A 90-s dynamic exclusion is used to reduce redundant peptide selection (see Note 16). 3.7. Data Analysis and Database Searching
1. The raw MS/MS data is searched using Sequest (Thermo Fisher, Pittsburg, PA), which is a software program for protein identification using tandem MS data. Precursor ion tolerance is set to 1.5 Da and fragment ion tolerance is set to 0.5 Da. Peptides are considered legitimate identifications only if they exhibit the following charge state-dependent cross correlation (Xcorr) criteria: ³1.9 for [M+H]+1 peptides, ³2.2 for [M+2H]+2 peptides, and ³3.5 for [M+3H]+3 peptides (see Note 17). 2. Only peptides exhibiting tryptic termini are accepted as legitimate identifications, and data can be analyzed allowing for 1–3 internal missed protein cleavage sites (see Note 18). Cysteine alkylation is treated as a standard modification, and oxidized methionine and cysteine acrylamide modification are chosen as variable modifications. 3. Protein identifications are considered legitimate only if a minimum of two constituent peptides are identified for a given protein during the database search (see Note 19). 4. All identified proteins are entered into an Excel spreadsheet and used as input to the free-access software program UniProt (18) [Swiss Institute of Bioinformatics; (http://www.uniprot. org)] to generate their corresponding UniProt accession numbers. The UniProt accession numbers are then analyzed using GOMiner (19), a Web, free-access software application based on Gene Ontology (GO) (discover.nci.nih.gov/ gominer). This program maps identified proteins to molecular function and subcellular location.
4. Notes 1. Short-term exposure to xylene can cause irritation of the skin, eyes, nose, and throat, difficulty in breathing, dizziness, and headache. Xylene should be handled in a certified chemical fume hood. Alternately, an orange-oil based product, such as Histo-Clear (National Diagnostics; Atlanta, GA) or Hemo-De
290
Fowler, O’Leary, and Mason
(Scientific Safety Solvents; Keller, TX) may be used to remove excess paraffin from FFPE tissue sections. 2. Detergent-based buffers are widely reported in the literature (8–12) and generally give the highest extraction efficiencies for FFPE tissues. However, detergents such as SDS may inhibit proteases used in proteomics and can dominate mass spectra relative to the individual peptides. Because of this, SDS-PAGE, followed by in-gel digestion of individual bands of interest, (Subheadings 3.3 and 3.4) is usually required to prepare SDS-solubilized protein extracts for MS analysis. A second work flow for proteomic analysis utilizes a detergentfree extraction buffer, followed by in-solution tryptic digestion (Subheading 3.5). Though the total protein extraction efficiencies may be reduced, this method requires fewer sample handling steps. (a) Alternate extraction A (12): suspend the deparaffinized tissue sections in 6 M guanidine–HCl, sonicating briefly on ice to lyse any DNA. Heat at 100°C for 30 min. Clear the supernatant by centrifugation before proceeding with in-solution digestion (Subheading 3.5). (b) Alternate extraction B (12): Suspend the deparaffinized tissue section in 50 mM ammonium bicarbonate, sonicating briefly on ice. Add sequencing-grade trypsin (Promega, Madison, WI) to a final concentration of 20 ng/mL and incubate at 37°C overnight. Clear the supernatant by centrifugation before proceeding with LC/MS/MS (Subheading 3.6). The results from several detergent- and nondetergentbased extraction techniques reported in the literature for FFPE tissues are summarized in Table 1. 3. When working with acrylamide, use gloves, a laboratory coat, and eye protection. It is best to prepare acrylamide solutions in a chemical hood. 4. TEMED is best stored at room temperature in a desiccator. Purchase small bottles as it may degrade after opening. 5. A commercially available stain such as SimplyBlue SafeStain (Invitrogen, Carlsbad, CA) may also be used to visualize the protein bands. These colloidal coomassie® G-250-based stains require much less destaining, thus reducing total staining time. Many do not require methanol or acetic acid fixatives and destaining solutions. 6. Iodoacetamide (IAA) is light sensitive, so the solution should be freshly prepared and protected from light. 7. In addition to whole tissue sections, regions of interest within a tissue section (such as a cancerous lesion) can be isolated by
Human renal cancer
Human renal cancer
Glioblastoma
Canalplasty
Mouse liver
Colon adenoma
Shi et al. (8)
Shi et al. (8)
Guo et al. (9)
Palmer-Toy et al. (11)
Jiang et al. (12)
Sprung et al. (13)
Ammonium bicarbonate/80°C, 1 h; trifluoroethanol added; 60°C, 1 h
6 M guanidine–HCl/no heating 6 M guanidine/100°C, 30 min CNBr treatment of cell pellet from #2 Tris + SDS, pH 8.2a Direct tryptic digest
2%SDS/ammonium bicarbonate/70°C, 1 h
Tris + SDS, pH 9a
No AR b
Tris + SDS, pH 7a
FPPE tissue extraction method
b
a
Samples were heated at 100°C for 20 min, followed by 60°C for 2 h No heat-induced protein extraction step performed c Fresh-frozen soluble fraction d Fresh-frozen cell pellet fraction
FFPE tissue source
Authors
12265
3207
266
2554
480
94
2380c/3110d
2404
3305 12517c/16023d
2404
3305
2302
395 106
2540 589 10349
57 470 202
123
2733
1883
3263
Protein IDs
352 3005 1129
412
14748
1714
3336
Distinct peptide IDs
Distinct peptide IDs
Protein IDs
FFPE tissue
Fresh tissue
Table 1 Protein extraction protocols from FFPE tissues documented in literature: comparison with fresh or frozen tissue Protein Mass Spectrometry Applications on FFPE Tissue Sections 291
292
Fowler, O’Leary, and Mason
laser-capture microdissection (3). Generally, at least 20,000 cells, or a minimum of 1–2 mg protein, should be used to ensure an adequate protein yield for nano-flow LC/MS/MS. Using smaller FFPE specimens generally results in better protein recovery and superior MS results. However, such specimens cover a smaller fraction of the tissue proteome. Thus, a balance must be struck between adequate coverage of the tissue proteome and the quality of the MS results. 8. Protein extraction from FFPE tissue specimens may be pHdependent, leading to preferential extraction of certain proteins at a particular pH (14). Thus, if sufficient sample exists, carry out extractions at pH 4, 7, and 9 and pool the individual pH fractions. 9. Some form of physical disruption is essential to adequately hydrate the FFPE specimens, which facilitates reversal of the formaldehyde–protein adducts and access of the proteolytic enzyme to the protein backbone (20). This physical disruption can take the form of grinding (mortar and pestle), ultrasound (21), or elevated pressure (22). 10. Commercially available precast gels may be used in place of home-made acrylamide gels. For example Bis–Tris, Tris– Acetate, and Tris–Glycine gels (available from Invitrogen, Carlsbad, CA) with optimized sample preparation and running buffers are a good alternative. These gels are available with a fixed acrylamide concentration or as acrylamide gradient gels. The choice of gel depends upon the molecularweight range of the proteins to be separated. 11. The selection of an assay for total protein determination is dependent upon sample condition and extraction buffer used. The Bradford protein assay (23), for example, is not compatible with detergents and so will not give accurate results for samples extracted with SDS. The bicinchoninic acid (BCA) assay (24) is detergent-compatible and tolerant to a wide range of buffers, making it a suitable protein assay for most applications. 12. The most frequent contaminant in MS-based proteomic analysis is human keratin from dust, hair, and fingerprints. Always wear gloves and rinse all equipment (gloves, gel apparatus, staining dishes, tubes, etc.) with water and ethanol or isopropanol prior to use. 13. The gel slices may be dried in a vacuum oven, but a superior method is to use a centrifugal dryer (Fisher), which prevents sample loss by reducing solvent foaming. 14. Samples extracted using the methods in Subheading 3.2 may also be digested in solution with trypsin and then directly analyzed by LC/MS/MS in lieu of SDS-PAGE and in-gel digestion.
Protein Mass Spectrometry Applications on FFPE Tissue Sections
293
However, if a detergent-containing buffer has been used for extraction, it is necessary to first remove excess detergent by dialysis or by multiple filter-aided washes against a MS-compatible buffer (such as 8 M urea or ammonium bicarbonate) (25). The tryptic digests may be desalted using ZipTips (Millipore, Billerica, MA) prior to MS analysis. 15. These separation conditions are typical for a capillary-flow LC/MS/MS system. For a nano-flow LC/MS/MS system, flow rates are typically 0.25–0.5 mL/min, and the total amount of protein loaded is 1–2 mg in 2–5 mL of solvent. A typical column for nano-flow LC is 75-mm (inner diameter) × 360-mm (outer diameter) × 10-cm long packed with 5-mm, 300-Å pore-size C18 media (Vydac, Hysperia, CA). 16. Liner ion trap MS instruments from other vendors may have different optimal settings for peptide analysis, and hence, the user should consult with the manufacturer and the instrument manual. A number of MS instrument formats can be used for protein analysis, each with their own strengths and weaknesses. However, a discussion of these attributes is beyond the scope of this chapter. 17. Protein identification databases can generally be purchased with the MS instrument and are accessed through the instruments proprietary analysis software (such as Spectrum Mill MS Proteomics Workbench for the Agilent 6300 series ion trap MS). There are a number of proprietary and public access proteomics databases in addition to Sequest including MASCOT (Matrix Science, Boston, MA), SwissProt (public access), and RefSeq (available through the National Center for Biotechnology Information, Bethesda, MD). 18. When FFPE tissue is heated at temperatures ³65°C in buffers with a pH £ 6, peptide cleavage can occur on the C-terminal side of aspartic acid residues due to the formation of an internal anhydride with the aspartic acid side chain (22). Since these cleavage sites are specific for the aspartic acid n+1 peptide bond, they do not interfere with protein analysis. For the identification of aspartic cleavage sites, raw MS or tandem MS data are searched without enzyme constraint. Peptides possessing either two aspartic acid cleavage sites or one aspartic acid cleavage site and one tryptic cleavage site are considered positive identifications if their Xcorr probability scores are acceptable. 19. The presence of formaldehyde modifications in peptide tryptic fragments derived from FFPE tissues can lead to peptide and protein misidentification. Thus, increased stringency in the analysis of the MS data should be used. This includes allowing for two or three missed internal tryptic cleavage
294
Fowler, O’Leary, and Mason
sites, analyzing for aspartic acid cleavages, using high Xcorr values for deciding positive identification, and basing proteins calls on the presence of at least two constituent peptides for each protein. Also, estimates of misidentification frequency can be determined from false-positive rates using the method of Elias et al. (26) and by the reversed protein database method of Palmer-Toy et al. (11). References 1. Blonder, J. and Veenstra, T. D. (2009) Clinical proteomic applications of formalin-fixed paraffin-embedded tissues. Clin Lab Med. 29, 101–13. 2. Wilkinson, E. J. and Hendricks, J. B. (1995) Role of the pathologist in biomarker studies. J Cell Biochem Suppl. 23, 10–8. 3. Patel, V., Hood, B. L., Molinolo, A. A., Lee, N. H., Conrads, T. P., Braisted, J. C., et al. (2008) Proteomic analysis of laser-captured paraffin-embedded tissues: a molecular portrait of head and neck cancer progression. Clin Cancer Res. 14, 1002–14. 4. Negishi, A., Masuda, M., Ono, M., Honda, K., Shitashige, M., Satow, R., et al. (2009) Quantitative proteomics using formalinfixed paraffin-embedded tissues of oral squamous cell carcinoma. Cancer Sci. 100, 1605–11. 5. Jain, M. R., Liu, T., Hu, J., Darfler, M., Fitzhugh, V., Rinaggio, J., et al. (2008) Quantitative proteomic analysis of formalin fixed paraffin embedded oral HPV lesions from HIV patients. Open Proteomics J. 1, 40–5. 6. Hood, B. L., Darfler, M. M., Guiel, T. G., Furusato, B., Lucas, D. A., Ringeisen, B. R., et al. (2005) Proteomic analysis of formalinfixed prostate cancer tissue. Mol Cell Proteomics 4, 1741–53. 7. Prieto, D. A., Hood, B. L., Darfler, M. M., Guiel, T. G., Lucas, D. A., Conrads, T. P., et al. (2005) Liquid tissue: proteomic profiling of formalin-fixed tissues. Biotechniques 38(Suppl), 32–5. 8. Shi, S.-R., Liu, C., Balgley, B. M., Lee, C., and Taylor, C. R. (2006) Protein extraction from formalin-fixed, paraffin-embedded tissue sections: quality evaluation by mass spectrometry. J Histochem Cytochem. 54, 739–43. 9. Guo, T., Wang, W., Rudnick, P. A., Song, T., Li, J., Zhuang, Z., et al. (2007) Proteome analysis of microdissected formalin-fixed and paraffin-embedded tissue specimens. J Histochem Cytochem. 55, 763–72.
10. Xu, H., Yang, L., Wang, W., Shi, S. R., Liu, C., Liu, Y., et al. (2008) Antigen retrieval for proteomic characterization of formalin-fixed and paraffin-embedded tissues. J Proteome Res. 7, 1098–108. 11. Palmer-Toy, D. E., Krastins, B., Sarracino, D. A., Nadol, J. B., Jr., and Merchant, S. N. (2005) Efficient method for the proteomic analysis of fixed and embedded tissues. J Proteome Res. 4, 2404–11. 12. Jiang, X., Jiang, X., Feng, S., Tian, R., Ye, M., and Zou, H. (2007) Development of efficient protein extraction methods for shotgun proteome analysis of formalin-fixed tissues. J Proteome Res. 6, 1038–47. 13. Sprung, R. W., Jr., Brock, J. W., Tanksley, J. P., Li, M., Washington, M. K., Slebos, R. J., et al. (2009) Equivalence of protein inventories obtained from formalin-fixed paraffinembedded and frozen tissue in multidimensional liquid chromatography-tandem mass spectrometry shotgun proteomic analysis. Mol Cell Proteomics 8, 1988–98. 14. Fowler, C. B., Cunningham, R. E., O’Leary, T. J., and Mason, J. T. (2007) “Tissue surrogates” as a model for archival formalin-fixed paraffin-embedded tissues. Lab Invest. 87, 836–46. 15. Lemaire, R., Desmons, A., Tabet, J. C., Day, R., Salzet, M., and Fournier, I. (2007) Direct analysis and MALDI imaging of formalinfixed, paraffin-embedded tissue sections. J Proteome Res. 6, 1295–305. 16. Ronci, M., Bonanno, E., Colantoni, A., Pieroni, L., Di, I. C., Spagnoli, L. G., et al. (2008) Protein unlocking procedures of formalin-fixed paraffin-embedded tissues: application to MALDI-TOF imaging MS investigations. Proteomics 8, 3702–14. 17. Shevchenko, A., Wilm, M., Vorm, O., and Mann, M. (1996) Mass spectrometric sequencing of proteins silver-stained polyacrylamide gels. Anal Chem. 68, 850–8. 18. Bairoch, A., Apweiler, R., Wu, C. H., Barker, W. C., Boeckmann, B., Ferro, S., et al. (2005)
Protein Mass Spectrometry Applications on FFPE Tissue Sections
19.
20.
21.
22.
The universal protein resource (UniProt). Nucleic Acids Res. 33, D154–9. Zeeberg, B. R., Feng, W., Wang, G., Wang, M. D., Fojo, A. T., Sunshine, M., et al. (2003) GoMiner: a resource for biological interpretation of genomic and proteomic data. Genome Biol. 4, R28. Fowler, C. B., O’Leary, T. J., and Mason, J. T. (2008) Modeling formalin fixation and histological processing with bovine ribonuclease A: effects of ethanol dehydration on reversal of formaldehyde-induced cross-links. Lab Invest. 88, 785–91. Chu, W. S., Liang, Q., Liu, J., Wei, M. Q., Winters, M., Liotta, L., et al. (2005) A nondestructive molecule extraction method allowing morphological and molecular analyses using a single tissue section. Lab Invest. 85, 1416–28. Fowler, C. B., Cunningham, R. E., Waybright, T. J., Blonder, J., Veenstra, T. D., O’Leary, T. J., et al. (2008) Elevated hydrostatic pressure
23.
24.
25.
26.
295
promotes protein recovery from formalin-fixed, paraffin-embedded tissue surrogates. Lab Invest. 88, 185–95. Bradford, M. M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem. 72, 248–54. Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., et al. (1985) Measurement of protein using bicinchoninic acid. Anal Biochem. 150, 76–85. Wisniewski, J. R., Zougman, A., Nagaraj, N., and Mann, M. (2009) Universal sample preparation method for proteome analysis. Nat Methods 6, 359–62. Elias, J. E., Haas, W., Faherty, B. K., and Gygi, S. P. (2005) Comparative evaluation of mass spectrometry platforms used in largescale proteomics investigations. Nat Methods 2, 667–75.
wwwwwwwwwwwwwww
Chapter 19 An Alternative Fixative to Formalin Fixation for Molecular Applications: The RCL2®-CS100 Approach Amélie Denouël, Florence Boissière-Michot, Philippe Rochaix, Frédéric Bibeau, and Nathalie Boulle Abstract
1
2 3
4 5
6
Molecular analysis of tissue lesions is increasingly used in laboratories to identify new prognostic and therapeutic markers. Formalin has long been the tissue fixative of choice in the laboratories of pathology, as it preserves tissue morphology allowing accurate histological diagnosis. However, formalin is highly toxic and alters nucleic acids and protein integrity, so that new fixatives are critically needed that would allow both morphological and molecular analysis on the same tissue specimen. Recently, we found RCL2®-CS100, a noncross-linking fixative, to display interesting performances regarding tissue morphology and DNA, RNA, and protein quality. We adapted RCL2 tissue fixation protocol so it could be used on a routine and automated laboratory basis, still preserving its good performances. This protocol will be described in detail in the following review. Key words: Tissue fixation, Formalin, RCL2®-CS100, Immunohistochemistry, Nucleic acids, DNA, RNA
1. Introduction
7 8 9 10 11 12 13 14 15 16 17
18
In the last few years, gene and protein expression profiling have been widely developed in pathological tissues to better understand the molecular events leading to diseases and to identify new therapeutic markers for treatment of patients. Formalin-fixed paraffin-embedded tissues represent the most abundant supply of archival material for clinical and molecular ana lyses. Although formalin is adapted to morphological examination of tissues, it is a cross-linking agent which is known to alter and fragment nucleic acids and proteins, thus impairing extraction efficiency and quality of DNA and more strikingly RNA and proteins.
Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3_19, © Springer Science+Business Media, LLC 2011
297
19 20 21 22 23 24 25 26 27 28
298
Denouël et al.
Recent studies have reported improved protocols for molecular analysis from formalin-fixed tissues (see Chapters 11–18 of this review). However, these approaches may have technical limitations when working with genes or proteins harboring weak expression levels or with very small amounts of RNA, such as RNA extracted from few microdissected cells. The gold standard for molecular analyses thus remains unfixed fresh or snap-frozen tissues. Unfortunately, these treatments cannot be used for routine laboratory samples as they do not provide accurate morphological details and may impair histological diagnosis. Searching for new fixatives, we found that RCL2®-CS100 (ALPHELYS, Plaisir, France), a new noncross-linking and nontoxic fixative, displayed good performances regarding tissue morphology and nucleic acid quality when evaluated on breast tumor specimen, even after several months of tissue storage (1). A following study from our laboratory also confirmed preservation of protein quality in RCL2® fixed tissues (2). These first studies have been performed using manual protocols for fixation and paraffin-embedding. We next evaluated RCL2®-CS100 fixation on various tumor specimen using automated steps to assess if this new fixative could be used on a routine laboratory basis for pathological diagnosis and molecular analysis. In these latter conditions, we found that RCL2®-CS100 still presented with high performances. However, to obtain best results using RCL2®-CS100 for routine laboratory, a few precautions in fixative handling and in some tissue processing steps are required. These precautions will be described in detail in the following paragraphs.
29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55
56
2. Materials
64
1. RCL2®-CS100 (ALPHELYS, Plaisir, France). Store at room temperature. Working solutions are prepared by diluting three volumes of RCL2®-CS100 (concentrated) with five volumes of absolute ethanol (see Note 1). The diluted solution should be stored at +4°C and is stable for 1 month in these conditions. Carefully control the absence of crystallization before use. In case of crystallization, discard the solution.
65
2. Absolute ethanol and xylene (see Note 2).
57 58 59 60 61 62 63
66 67 68 69 70
2.1. Tissue Sample Fixation and ParaffinEmbedding
3. PARAPLAST X-TRA tissue embedding medium (McCormick, St. Louis, MO, USA). The melting point of this paraffin wax is 52°C but for better embedding, use it at 56°C. 4. Disposable plastic histology cassettes (Labonord, Templemars, France).
An Alternative Fixative to Formalin Fixation for Molecular Applications
2.2. Sectioning and Histology Analysis
299
1. RNase AWAY (Sigma–Aldrich, St. Louis, MO, USA).
71
2. Precleaned and ready to use microscope SuperFrost slides (MENZEL-GLÄSER, Braunschweig, Germany). ®
3. Precleaned and ready to use microscope SuperFrost Plus slides (MENZEL-GLÄSER, Braunschweig, Germany). ®
2.3. DNA and RNA Extractions
1. Xylene, molecular biology grade (VWR International, Strasbourg, France). 2. Absolute ethanol, molecular biology grade (Carlo-Erba, Val de Reuil, France). 3. All Prep DNA/RNA Mini Kit, 50 spin columns (QIAGEN, Courtaboeuf, France) (http://www1.qiagen.com/Products/ RnaStabilizationPurification/RnaDnaSystem/AllPrep DNARNAMiniKit.aspx#Tabs=t2), but other nucleic acids extraction systems can be used. 4. b2-Mercaptoethanol (BME) electrophoresis reagent ³98% (Sigma–Aldrich, St. Louis, MO, USA) must be added to Buffer RLT Plus (included in All Prep DNA/RNA Mini Kit) before use. Add 10 ml BME per 1 ml Buffer RLT Plus. Dispense in a fume hood and wear appropriate protective gloves and clothing. Buffer RLT Plus is stable at room temperature for 1 month after addition of BME. 5. RNase-free DNase I Set (QIAGEN, Courtaboeuf, France): The RNase-free DNase I Set is shipped at room temperature and should be stored immediately upon receipt at 2–8°C. When stored at 2–8°C and handled correctly, the buffer and lyophilized enzyme can be kept for at least 9 months without any reduction in performance (see Note 3). DNase I is dissolved in 550 ml of RNase-free water (included in the RNasefree DNase I Set) and aliquots are stored at −20°C. The aliquots are stable for 9 months after reconstitution. After thawing, aliquots must not be frozen again but stored at +4°C, being stable for 6 weeks.
2.4. DNA Analysis
1. Tris–borate EDTA 10× buffer (TBE, Sigma–Aldrich, St. Louis, MO, USA). Dilute 100 ml TBE 10× with 900 ml water for use (TBE 1×). Store at room temperature. 2. Agarose of molecular biology grade (Eurobio, France). 3. Ethidium bromide 0.7 mg/ml (BET, Eurobio, France). Because BET is highly mutagenic and therefore potentially carcinogenic, it should be handled with stringent security precautions (under a fume hood and with appropriate protective gloves and clothing). We suggest the use of prediluted solution packaged in drops counter that allows a more secure use by avoiding pipetting.
72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113
300
Denouël et al.
4. Demineralized and deionized water as, for example, Milli-Q® water (Millipore, Molsheim, France).
114 115
5. Orange G for molecular biology (Sigma–Aldrich, St. Louis, MO, USA). Ficoll® PM 400 (Sigma–Aldrich, St. Louis, MO, USA). To prepare a 100-ml solution, dissolve 15 g of Ficoll® in 100 ml of Milli-Q® water. Vortex and heat the solution on a stirring hot plate at 37°C until the Ficoll® is dissolved. Add a pinch of Orange G to color the buffer. Filter the solution and store in aliquots at room temperature.
116 117 118 119 120 121 122
6. 1 kb DNA ladder at 1 mg/ml (Invitrogen, Carlsbad, CA, USA). Stored in aliquots at −20°C. For agarose gels, prepare extemporaneously 2 ml of ladder with 2 ml of Orange G solution, before loading into the gel.
123 124 125 126 127
2.5. RNA Analysis
2. RNase-free water or sterile water (B. Braun, Melsungen, Germany).
128 129
3. RNA 6000 Nano Chip Kit (© Agilent Technologies, Waldbronn, Germany) contains Nano Chips, Ladder, Dye Concentrate, Marker, Gel Matrix, and spin filters. (http:// www.chem.agilent.com/Librar y/usermanuals/Public/ G2938-90034_KitRNA6000Nano_ebook.pdf).
130 131 132 133 134
135
3. Methods RNA extracted from tissues is inherently labile owing to RNase activity. Achieving high-quality RNA from human samples requires a succession of processes and a multidisciplinary organization that involves surgeons, pathologists, and molecular biologists. Sample-handling protocols need to be standardized. First, a particular attention must be paid to the time between tissues surgical excision and fixation. Importantly, whatever the experimental step, the delay of tissue handling at room temperature must be minimized. By using flash-frozen tissues as references, we tested some critical steps in tissue harvesting, sample-handling, and nucleic acid extraction. Hereafter are described the optimized procedures for RCL2®-CS100 fixation (see Note 4).
136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154
1. RNase AWAY (Sigma–Aldrich, St. Louis, MO, USA).
3.1. Tissue Sample Collection and Paraffin-Embedding
1. Fresh tissues are obtained directly after surgical resection. When possible, the time of vessels ligation is recorded by the surgical team in order to assess the warm ischemia time which may impact on RNA degradation (3). A pathologist assistant transfers the surgical specimen as quickly as possible to the pathology department. A macroscopic examination is then performed, including localization and measurement of the
An Alternative Fixative to Formalin Fixation for Molecular Applications
301
lesion. The selected tissue (avoiding necrotic areas) is cut in pieces of approximately 5 × 5 × 5 mm. Larger samples are allowed but the maximal thickness for optimal fixation with RCL2®-CS100 is ~5 mm. The tissue sample is immersed in cold RCL2®-CS100 solution (see Notes 1 and 5) and kept overnight at +4°C. In our experience, the optimal time for RCL2®-CS100 fixation is comprised between 4 and 24 h. A RCL2®-CS100 fixation longer than 24 h may hamper the RNA quality (see Note 6). 2. After overnight fixation at +4°C, put the sample in a disposable plastic histology cassette. Each cassette must be labeled with a unique identifying code. Cassettes must be labeled with a dark pencil, since the ink contained in most laboratory markers is soluble in alcohol or xylene. Immerse the cassette in cold (+4°C) absolute ethanol for 5 min. 3. Fixed tissue pieces are dehydrated through a series of alcohol baths before clearing with xylenes. For manual inclusion (without any processor), the optimal sequence is composed of six successive baths (50 min long) in absolute ethanol at room temperature, followed by two baths (45 min long) in xylene. The cassettes are then incubated in an oven at 56°C with molten paraffin (see Note 7) in three successive baths for 30 min each. 4. If a tissue processor is to be used, be sure that there is no formalin contamination and that all alcohol and xylene baths have been changed after formalin use (see Note 8). A careful control of the baths’ temperature during the various steps of the automated procedure is also critical to obtain the highest quality for nucleic acids and protein. The tissue processor is a vacuum infiltration processor allowing automated reagents transfer. As for manual processing, the samples are dipped successively in six baths of absolute ethanol, 30 min each, at 37°C, followed by four baths of xylene for 5, 10, 30, and 45 min at 37°C. The final step in the processing is the replacement of the xylene by molten paraffin, using four baths at 56°C for 5, 5, 20, and 30 min, respectively. In our experience, manual inclusion gave slightly better nucleic acids recovery than automated inclusion. However, used with the optimized procedures, the tissue processor is a good compromise allowing RCL2®-CS100 use in a routine setting along with the recovery of nucleic acids suitable for molecular analysis. 5. Whatever the processing (manual or processor-assisted), paraffin steps are critical. To avoid RNA degradation, it is of major importance that tissues are immediately removed from the last paraffin bath and covered with molten paraffin
155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200
302
Denouël et al.
(56°C) subsequently. Tissues are oriented and allowed to cool and harden into a block at room temperature.
201 202
6. For optimal recovery of nucleic acids, store the tissue blocks at −20°C until use.
203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229
3.2. Histology Analyses
1. Put gloves and place the paraffin blocks on a refrigerated plate (−14°C). 2. 4-mm thick sections of paraffin-embedded tissues are cut on a microtome workstation. The sections are laid on a bain-marie set at 48°C to unfold them. They are then picked up on glass slides (see Note 9). To ensure optimal adhesion of tissue sections, the slides are baked in a ventilated oven for at least 30 min at 56°C. 3. After sectioning, coat the tissue blocks with paraffin to ensure nucleic acids preservation for further experiments. 4. Use the sections as for classically formalin-fixed tissue sections (histological staining, immunohistochemistry, and in situ hybridization). 5. For histological staining, the hematein–eosin–saffron (HES) staining usually employed in the routine setting for formalinfixed samples can be used. Although the morphology could be slightly impacted by the RCL2®-CS100 fixation (tissue retractions were sometimes observed), RCL2®-CS100 preserves tissue integrity as well as cellular details, allowing accurate pathological diagnosis (Fig. 1). 6. For immunohistochemical studies, some antibodies may require some adjustments of the staining procedure. In our experience, adjusting target retrieval solution or antibody dilutions always allowed us to obtain similar staining quality as observed for formalin-fixed tissues (Fig. 2). Of note, a lack
Fig. 1. Morphology of a colon adenocarcinoma fixed either in formalin (a) or RCL2®-CS100 (b). A representative view of hematein–eosin–saffron-stained sections is presented. Original magnification: ×200.
An Alternative Fixative to Formalin Fixation for Molecular Applications
303
Fig. 2. Immunohistochemical staining of cytokeratin 20 (CK20). A colon carcinoma sample was fixed either in formalin (a) or RCL2®-CS100 (b, c) and processed for CK20 immunohistochemistry. This figure shows that some antibodies, such as CK20, require an adjustment of the staining procedure. In (b), similar experimental conditions were used as for formalin-fixed tissues, leading to a weaker signal. Targeting retrieval solution by shifting from citrate pH 6 to EDTA pH 9 allowed to obtain similar staining as observed for formalin-fixed tissues (c). Original magnification: ×200.
of staining of the tissue periphery when using RCL2®-CS100, can also be observed for some antibodies; this artifact must be taken into account for signal interpretation. 7. For both immunohistochemical and in situ hybridization studies, we suggest to increase the time of counterstaining with hematoxylin as the contrast is often lower than that observed for formalin-fixed tissue. 3.3. Tissue Processing for Molecular Biology
1. It is critical that all the equipment is RNase-free. In particular, the knife holder of the microtome must be scrubbed clean with RNase AWAY before tissue sectioning. For 5 × 5 mm samples, 10-mm thick sections of paraffin-embedded are collected in two RNase-free vials (Eppendorf), five sections in each vial. Such tissue sections can be shipped at room temperature for 24–48 h without dramatic impact on nucleic acids quality. 2. Paraffin removal is critical to allow efficient nucleic acids recovery. For optimal results, pour 700 ml of xylene (molecular biology grade) in each vial, vortex vigorously for at least 30 s and centrifuge for 5 min at 18,000 g at room temperature. Discard the supernatant. Repeat once with 700 ml of xylene. Add 700 ml of absolute ethanol, gently turn the vial upside down to resuspend the pellet and centrifuge for 5 min at 18,000 g at room temperature. Carefully remove the supernatant solution with a pipette and discard it. Repeat this procedure once with 700 ml of absolute ethanol. Allow the alcohol to evaporate for 30 min at room temperature under a fume hood. 3. Efficient disruption and homogenization of the tissues is an absolute requirement for nucleic acid purification procedures.
230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258
304
Denouël et al.
Indeed, incomplete disruption results in a significant reduction in nucleic acids yields. From this step, DNA, RNA, or both can be extracted. We selected the AllPrep DNA/RNA Mini Kit (QIAGEN, Courtaboeuf, France) that allows simultaneous purification of genomic DNA and total RNA. All the procedures are described in the AllPrep DNA/RNA Mini Handbook. (http:// www1.qiagen.com/Products/RnaStabilizationPurification/ RnaDnaSystem/AllPrepDNARNAMiniKit.aspx#Tabs=t2). According to the manufacturer’s recommendations, a DNase I step was performed for RNA extraction to avoid genomic DNA contamination. Of note, for purification of DNA or RNA alone, other kits from Qiagen have been successfully used, leading to better nucleic acids yield as compared to those obtained after concomitant DNA/RNA extraction.
259 260 261 262 263 264 265 266 267 268 269 270 271 272
4. The tissue processing procedure described in steps 1 and 2 of Subheading 3.3 has also been successfully used for protein extraction (2).
273 274 275
5. If Laser Capture Microdissection (LCM) needs to be performed, 7-mm thick sections of RCL2®-CS100-fixed paraffinembedded tissues are cut on a microtome workstation (see Subheading 3.3, step 1). Sections are then deparaffinized prior to staining in 100% RNase-free ethanol for 5 min at 60°C in a water bath. Tissue sections are then processed for LCM as previously described (1).
276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303
3.4. Quantitative and Qualitative Assessment of DNA
1. After DNA extraction, quantitative and qualitative analyses are necessary to evaluate the yield of DNA extracted and the suitability of the DNA sample for further molecular analyses. 2. DNA quantification is commonly assessed by spectrophotometric analysis, measuring the optical density at 260 nm (1 DO260 = 50 mg/ml DNA). 3. Agarose gel electrophoresis is the easiest and the most common way of analyzing DNA quality. Extracted DNA is run in parallel with a DNA ladder on 0.8% agarose gel which allows good separation of large DNA fragments (Fig. 3). 4. For each experiment, prepare a 0.8% (w/v) agarose gel in TBE 1× (200 ml for a 10 × 10 cm gel). The solution needs to be carefully heated (using, for instance, a microwave oven at maximal power for 2.5 min or a stirring hot plate) until the agarose melts completely. In a fume hood, immediately dispense three drops of ready-to-use BET solution in 200 ml of the 0.8% agarose solution. Allow the solution to cool (~65– 70°C) under the fume hood and pour it carefully on a clean casting plate with an appropriate comb. If bubbles appear, push them away to the sides with a disposable tip. Allow the gel to solidify for about 30 min before using (see Note 10).
An Alternative Fixative to Formalin Fixation for Molecular Applications
305
Fig. 3. Quality of extracted DNA from paired frozen (F) and RCL2®-CS100-fixed (R) paraffin-embedded tissues following electrophoresis. DNA was extracted from whole tissue sections of colon carcinomas as described in Subheading 3.3. Two hundred nanograms of total DNA were submitted to electrophoresis on a 0.8% agarose gel and run in parallel with a DNA ladder (L). DNA quality is shown for three paired samples. Note the apparition of a weak smear for RCL2®-CS100 samples (R1–R3) that did not impair the quality of the DNA molecular analyses.
5. Native DNA migrates as a tight band of high molecular weight (³40 kb) and the presence of degraded/sheared DNA suppress “and/or RNA”, if any, can be visualized on the gel using an UV transilluminator. 3.5. Quantitative and Qualitative Assessment of RNA
1. Careful evaluation of RNA integrity is of major importance as it significantly impacts relative gene expression (4–7). Indeed, expression profiling by real-time reverse transcription polymerase chain reaction and microarray-based assays is a powerful tool for molecular classification of diseases and finding of new prognostic and predictive factors. 2. Both qualitative and quantitative assessments of extracted RNA are performed using Agilent technology (Agilent RNA 6000 Nano Kit Guide). This technology, through a specific algorithm, provides an accurate evaluation of RNA quality by measuring the RNA integrity number (RIN) based on the ratio of the ribosomal bands and on the presence or absence of degradation products visualized after capillary electrophoretic separation. In our experience, DNase I treatment of RNA extracted with All Prep DNA/RNA Mini Kit allows an easier analysis of the Agilent profiles. To date, a RIN ³ 6 is described to provide more reproducible microarray results (8). In our experience, the Agilent RNA 6000 Nano kit also provides a good estimation of the RNA concentration as
304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326
306
Denouël et al.
Fig. 4. Quality of extracted RNA from paired frozen (a) and RCL2®-CS100-fixed paraffin-embedded (b) tissues following Bioanalyser electrophoresis. Total RNA was extracted from whole tissue sections as described in Subheading 3.3. One microliter of extracted RNA was run on Agilent RNA 6000 Nanochip. The 18S and 28S ribosomal RNAs are indicated by arrow heads and the RNA integrity number (RIN) is indicated for each sample.
compared to other methods of RNA quantification (i.e., spectrophotometry measurements at 260 nm).
327 328
3. An example of an electrophoretic profile of the same sample, either flash frozen or after fixation with RCL2®-CS100, is shown in Fig. 4.
329 330 331
332
333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351
4. Notes 1. The volume of RCL2-CS100 required for consistent fixation is at least tenfold the volume of the tissue to be fixed. 2. Reagents with molecular biology grade are not required at this step. 3. In some cases, the vial of DNase may appear to be empty. This is due to lyophilized enzyme sticking to the septum. To avoid loss of DNase, do not open the vial. Instead, inject RNase-free water into the vial using a needle and syringe, invert the vial to dissolve the DNase, and remove the dissolved DNase using the syringe and needle. 4. To compare various experimental conditions, one piece from each sample can be placed in CryoTube™ vials (Nunc™) and fresh-frozen in liquid nitrogen prior being stored in a lowtemperature freezer (−80°C). 5. The diluted RCL2®-CS100 solution is stable for 1 month at +4°C. Carefully control the lack of crystallization before use. In case of crystallization, discard the solution. Indeed, in our experience, using crystallized diluted RCL2®-CS100 resulted in still preserved tissue morphology but degraded RNA.
An Alternative Fixative to Formalin Fixation for Molecular Applications
307
356
6. In case the sample cannot be processed after 24 h immersion in RCL2®-CS100 (for a week-end, for example), shorten the time of tissue fixation (minimum of 4 h) and store the sample in cold absolute ethanol (up to 48 h) before dehydration and paraffin-embedding.
357
7. The paraffin used has a melting point of 52°C.
352 353 354 355
8. Formalin is a cross-linking reagent that induces RNA chemical modifications and fragmentation. Thus, it is critical to ensure that all baths are formalin-free. Experiments from our laboratory have indeed shown that trace amounts of formalin in the tissue processor may alter RNA quality.
358 359 360 361 362
9. For immunohistochemistry or in situ hybridization, the use of silanized slides is recommended. They have a permanent positive charge associated with the glass surface that allows greater adherence of tissue sections.
363 364 365 366
10. Once the agarose gel is polymerized, pour 1× TBE buffer into the gel tank to submerge it to 2–5 mm depth. For optimal results, we suggest to load 200 ng of DNA per well. Transfer the appropriate amount of each DNA sample to a fresh microfuge tube; add ~0.5 volume of Orange G solution (e.g., 5 ml into a 10-ml sample). Load the first well with a molecular weight marker. If possible, avoid using the outer wells, because the samples often run aberrantly. Close the gel tank, switch on the power-source and run the gel at 5 V/cm. Stop the gel when the Orange G has run three-fourth the length of the gel.
367 368 369 370 371 372 373 374 375 376
377
References
378 379 380 381 382 383
1. Delfour, C., Roger, P., Bret, C., Berthe, M. L., Rochaix, P., Kalfa, N., et al. (2006) RCL2, a new fixative, preserves morphology and nucleic acid integrity in paraffin-embedded breast carcinoma and microdissected breast tumor cells. J. Mol. Diagn. 8, 157–169. 2. Bellet, V., Boissière, F., Bibeau, F., Desmetz, C., Berthe, M. L., Rochaix, P., et al. (2008) Pro teomic analysis of RCL2 paraffin-embedded tissues. J. Cell Mol. Med. 12, 2027–2036. 3. Almeida, A., Paul Thiery J., Magdelénat, H., and Radvanyi, F. (2004) Gene expression analysis by real-time reverse transcription polymerase chain reaction: influence of tissue handling. Anal. Biochem. 328, 101–108. 4. Bustin, S. A. and Nolan, T. (2004) Pitfalls of quantitative real-time reverse-transcription polymerase chain reaction. J. Biomol. Tech. 15, 155–166. 5. Copois, V., Bibeau, F., Bascoul-Mollevi, C., Salvetat, N., Chalbos, P., Bareil, C., et al. (2007) Impact of RNA degradation on gene
384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399
expression profiles: assessment of different methods to reliably determine RNA quality. J. Biotechnol. 127, 549–559. 6. Fleige, S., Walf, V., Huch, S., Prgomet, C., Sehm, J., and Pfaffl, M. W. (2006) Comparison of relative mRNA quantification models and the impact of RNA integrity in quantitative real-time RT-PCR. Biotechnol. Lett. 28, 1601–1613. 7. Ho-Pun-Cheung, A., Bascoul-Mollevi, C., Assenat, E., Boissiere-Michot, F., Bibeau, F., Cellier, D., et al. (2009) Reverse transcription-quantitative polymerase chain reaction: description of a RIN-based algorithm for accurate data normalization. BMC Mol. Biol. 15, 10–31. 8. Strand, C., Enell, J., Hedenfalk, I., and Ferno, M. (2007) RNA quality in frozen breast cancer samples and the influence on gene expression analysis – a comparison of three evaluation methods using microcapillary electrophoresis traces. BMC Mol. Biol. 22, 8–38.
400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421
wwwwwwwwwwwwwww
Index A aCGH. See Array-based comparative genomic hybridization Acute myeloid leukemia................................................... 39 Alexa Fluor......................................................71, 72, 76, 77 Anaplastic large cell lymphoma.................................. 39, 56 Anaplastic lymphoma kinase (ALK)................................ 39 Anonymized specimen................................................... 6–8 Antibodies...................25, 38, 39, 42–44, 46–52, 54, 56–58, 62, 63, 69, 71, 72, 74, 76, 77, 81, 85, 278, 302, 303 Antigen...................39–46, 48–51, 54–57, 60–63, 76, 85, 87 Antigen retrieval........................... 38, 39, 43–49, 54, 61, 62, 70, 73, 84, 85, 88, 281 Archive DNA................................................................. 161 Archives......................................... 3, 15, 107, 118, 206, 271 Area confidence.............................................................. 102 Array-based comparative genomic hybridization (aCGH)....................... 131–134, 136, 139, 140, 144 Array CGH......131, 132, 134–135, 140–143, 169–173, 176
B BAC clones..................................................................... 131 Beneficence........................................................................ 2 Bioethics................................................................... 2, 5, 16 Biomarker.............. 23, 25, 39, 118, 191–193, 205–207, 219, 239–255, 270, 271, 281 Biomarker identification................................................. 118 Biopsy........................................... 7, 40, 117, 118, 240, 269 Biorespository................................................3–7, 13, 15, 16 Bisulphite modification...........................183, 184, 186, 189 Breast cancer..................................23, 25, 30, 39, 51, 53, 57, 58, 61, 92, 173, 174, 182, 240
C Cancer/carcinoma................5, 23, 25, 30, 34, 38, 39, 51–53, 57, 58, 60, 61, 80, 83, 91, 92, 94, 111, 112, 118, 142, 143, 148, 162, 169, 173, 174, 176, 182, 192, 193, 198–200, 239–255, 302, 303, 305 CD20...............................................................39, 48, 58, 60 CD33.......................................................................... 39, 58 CD117....................................................................... 39, 57
cDNA.............................. 124, 131, 223–225, 234, 241–243, 249, 250, 255, 270–280 CEA...............................................................................46, 56 Cetuximab........................................................................ 58 CGH. See Comparative genomic hybridization Chimeric monoclonal antibodies...................................... 39 Chromogen.....................................................44, 55, 56, 80 Chromogenic in situ hybridization (CISH)......... 48, 79–88 Chromosome............. 25, 30, 39, 79, 83, 87, 88, 91, 92, 102, 106, 107, 109, 110, 112, 121, 143, 168–172, 182 Chromosome paint......................................................... 112 Clinical Laboratory Improvement Amendments (CLIA)................................................................. 14 Clonality......................................................................... 110 Coded specimen....................................................... 6, 7, 11 Colorectal cancer................................ 58, 60, 143, 169, 193, 198, 199 Colorectal carcinoma (CRC)...............................39, 58, 182 Common Rule..................................... 2, 3, 7–11, 13, 14, 19 Comparative genomic hybridization (CGH)............87, 131–145, 163, 167–173, 176, 177 Confocal microscopy........................................................ 75 Controls......................... 3, 43, 44, 50–52, 56, 62, 73–74, 76, 137, 140, 142, 150, 155, 178, 185, 186, 189, 193, 194, 197, 198, 200, 201, 215–217, 221–225, 227, 228, 234, 235, 246–253, 255 Copy number alterations (CNA).............167, 169, 170, 177 Copy number analysis..................... 131–145, 155, 157, 177 Copy number variation (CNV).......................132, 148, 168 CpG.........................181, 182, 184, 185, 188, 189, 192–197 CpG island methylator phenotype (CIMP)...........182, 193, 198, 199 CpG methyltransferase........................................... 193, 195 Cy3.................................................. 136–138, 141, 144, 145 Cy5.................................................. 136–139, 141, 144, 145 Cyclin D1......................................................................... 55 Cytogenetics....................................................110, 112, 131
D Data analysis............................... 9, 142, 154–158, 229–232, 253, 289 Deceased persons...................................................13, 14, 20
Fahd Al-Mulla (ed.), Formalin-Fixed Paraffin-Embedded Tissues: Methods and Protocols, Methods in Molecular Biology, vol. 724, DOI 10.1007/978-1-61779-055-3, © Springer Science+Business Media, LLC 2011
309
Formalin-Fixed Paraffin-Embedded Tissues 310 Index
Declaration of Helsinki...................................................... 2 Defined tissue areas................................................ 108, 113 Degradation.........................40, 43, 126, 137, 139, 140, 150, 152, 154, 159, 179, 203, 218, 272, 300, 301, 305 De-identification...................................................12, 14, 19 Department of Health and Human Services (DHHS)..................................................2, 5, 12, 13 Digital microscopy.................................................... 92, 101 DNA..................................17, 27, 48, 85, 92, 110, 117, 131, 147, 161, 181, 191, 212, 240, 271, 297 amplification.......111, 149, 152, 162, 163, 166–177, 272 extraction......87, 120, 126–127, 133, 135–136, 192, 304 labeling..............................................133–135, 141, 150 quality.........................148, 150–151, 166, 192, 304, 305 quantity.....................................................162, 167, 304 sonication..................................................137, 140, 290
E Epidermal growth factor receptor (EGFR).....39, 58, 61, 91 Epitope.....................................41, 42, 46, 47, 49, 51, 56, 85 Erbitux............................................................................. 58 Erlotinib........................................................................... 58 Estrogen receptor (ER)............25, 41, 48, 49, 55, 56, 59, 87 Ethics....................................................................2, 5, 6, 16 Exon microarray..............................................270, 274, 279 Expression profiling........................ 205–235, 239–255, 265, 269–280, 297, 305 Expression studies.......................................................... 118
F FFPE. See Formalin-fixed paraffin embedded tissue FFPT. See Formalin-fixed paraffin embedded tissue FISH. See Fluorescence in situ hybridization Fixative, fixation......................39–44, 46, 47, 51, 52, 54, 55, 57, 58, 61, 62, 70, 76, 84, 86, 100, 106, 118, 120, 123, 124, 126, 128, 129, 177, 205–235, 240, 242, 254, 260, 274, 290, 297–307 Fluorescence in situ hybridization (FISH)................ 25, 27, 30–31, 53, 55, 59, 60, 79, 83, 91–103, 113, 158, 173, 174, 177 Food and Drug Administration (FDA).............2, 5, 58, 102 Forensics................................................................. 110, 162 Formalin........................ 3, 23, 38, 40–43, 46, 47, 61, 69–77, 79, 80, 85, 92–95, 100, 106, 118, 124, 126, 129, 131, 150, 162, 181–191, 210, 218–220, 229, 230, 232, 233, 240, 244, 259–267, 269, 274, 279, 281, 297–307 Formalin-fixed paraffin embedded tissue (FFPT)....................1–20, 23, 30, 32, 33, 38, 69–77, 79, 94, 95, 105–114, 117–129, 131–145, 147–159, 161–179, 181–204, 206, 259–267, 269–294 Fragmentation.................120, 126, 136–137, 140–141, 144, 148, 150–154, 159, 163, 177, 232, 240, 254, 259, 270, 272, 277, 279, 307
Fragment length......................................148, 155, 156, 158 Frozen sections.......................................... 43, 44, 47, 49, 87
G Gastrointestinal cancer....................................23, 38, 39, 58 GC content............................................................ 155–158 Gefitinib........................................................................... 58 Gel electrophoresis......................... 136, 137, 140, 141, 164, 166, 167, 178, 217, 261, 264, 282, 304, 305 Gemtuzumab.................................................................... 58 Gene amplification........................ 30, 80, 82, 102, 111, 176 GeneChip.................132, 152, 162, 270, 274, 275, 278, 279 Genes..............25, 30, 38, 39, 48, 79, 80, 82, 87, 91, 92, 102, 105, 111, 112, 129, 132, 163, 170, 173–177, 181, 182, 191, 193, 196–198, 200, 201, 203–235, 239–255, 259, 269–280, 289, 297, 298, 305 dosage............................................................... 163, 176 expression................................. 105, 111, 191, 205–235, 239–255, 259, 269–280, 305 expression profiling............ 205–235, 239–255, 269–280 silencing.................................................................... 182 ST microarray................................................... 270, 274 Genetic alterations............................................38, 110, 112 GenomePlex®.................. 162–164, 166–168, 170, 175–177 Genome-wide SNP 6.0...................................148–152, 154 Glivec...................................................................................57
H Health Insurance Portability and Accountability Act (HIPAA)............................................................ 2, 5 Hematoxylin and eosin (H&E)......................24–27, 29–31, 33, 34, 94, 163, 165, 177, 207, 210 Herceptin......................................................................... 25 HER-2 gene............................................................... 25, 30 Her2/neu................................39, 42, 48, 57, 59, 62, 91–103 Histology........................................... 33, 298, 299, 301–303 Homogeneously staining regions (HSR)........................ 102 Homogeneous tissue....................................................... 117 Honest broker................................................................. 6, 7 Hormone receptor.................................................42, 58, 59 Horseradish peroxidase-DAB.......................................... 44 Human subjects research............1–3, 7–9, 11, 13, 16, 17, 20 Hybridization....................25, 27, 30–31, 34, 48, 53, 79–88, 92, 96–97, 101, 102, 131–135, 138–142, 148, 150, 154, 155, 167–170, 173, 240, 279, 302, 303, 307
I Image processing............................................................ 113 Imatinib............................................................................ 57 Immunofluorescent labeling....................................... 69–77 Immunohistochemistry (IHC).......................38, 41, 42, 50, 52–54, 69, 74, 85, 91, 92, 240, 281, 302, 303, 307 Immunohistology....................................................... 37–63 Individualized therapy (Targeted therapies)............... 25, 57
Formalin-Fixed Paraffin-Embedded Tissues 311 Index
Informed consent.......................................1–3, 6–8, 10–12, 14, 15, 17–19 In situ hybridization........................... 25, 27, 30–31, 48, 53, 79–88, 96–97, 173, 240, 302, 303, 307 Institutional Review Board (IRB)...........2, 5–11, 13, 17–20 Iressa................................................................................. 58 Isothermal linear amplification............................... 272, 275
N
Ki-67..............................................................................49, 56
National Bioethics Advisory Commission (NBAC)........................................................ 5, 6, 16 National Commission for Protection of Human Subjects of Biomedical and Behavioral Research.................. 2 Neuroendocrine carcinomas............................................. 38 Nonsmall cell carcinoma................................................... 39 Nucleic acids........................... 107, 110, 117–129, 150, 206, 208, 213–215, 234, 240, 243, 254, 259–261, 264, 297–304 Nucleophosmin (NPM)................................................... 39 Nuremberg Code................................................................ 1 Nuremberg Trials................................................................ 1
L
O
Laser capture microdissection (LCM)................... 105–114, 117–129, 192, 292, 304 LCM. See Laser capture microdissection Liquid chromatography...................................282, 284, 288 Live cell imaging/microdissection.................................. 112 Loss of heterozygosity (LOH)....................................... 110 Lung cancer.........................................................23, 57, 148
Office of Civil Rights (OCR)........................................... 13 Office of Human Research Protection (OHRP)................................ 2, 3, 5, 7–9, 13, 14, 17 Oligo-dT primer............................................................ 242 Oligonucleotide microarrays....................133–134, 147–159 Oligonucleotides............................ 131–134, 162, 182, 192, 197, 208, 209, 220, 234, 241, 242, 244, 254
M
P
Material transfer agreement (MTA)..................6, 27, 28, 32 Matrix-assisted laser-desorption ionization (MALDI)........................................................... 282 Median absolute deviation (MAD).................171, 172, 174 Medical records........................................................ 3, 7, 19 MembraneSlide...............................................106, 118–122 Metafer PV............................................92–94, 97, 101, 102 Methylation....................................................181–184, 186, 189, 191–204 Methylation specific polymerase chain reaction (MSP)..................................................182, 192, 196 Methylight.............................................................. 191–204 Microarray.................14, 23–34, 51, 53, 131–145, 147–159, 162, 191, 206, 265, 266, 270, 272, 274, 305 data analysis.............................................................. 142 expression profiling............................191, 270, 272, 274 platforms...................................................132, 133, 265, 266, 270 profiling............................................................ 147–159 MicroBeam.............................. 106, 108–110, 112–114, 121 Microdissection...................... 105–114, 117–129, 163–165, 177, 192, 292, 304 MicroRNA (miRNA).............................240, 241, 243–246, 253, 259–267 Microtome...............................24, 29, 34, 94, 106, 207, 210, 241, 244, 260, 261, 285, 302–304 Microwaves.................................. 16, 41, 45, 47, 79–88, 304 Morphology..............................38, 45, 48, 86, 87, 102, 106, 112, 113, 118, 210, 298, 302, 306
P53.................................................................................56, 60 Paraffin blocks..................24, 26, 28, 34, 106, 210, 244, 302 Pathologists............ 5, 11, 14, 15, 32, 40, 54, 61, 63, 70, 300 Pathology...............................14, 69, 73, 106, 110, 111, 118, 206, 207, 209, 210, 232, 254, 300 PathVysion®..................................... 92–94, 97, 98, 100–103 PCNA. See Proliferating cell nuclear antigen PCR. See Polymerase chain reaction Percentage methylated ratio (PMR).......................194, 197, 200–203 Personalized medicine.................................................... 206 Personally identifiable information................3, 6–11, 13–17 Pitfalls............................................................................... 37 Polymerase chain reaction (PCR)............. 87, 110, 112, 131, 134, 135, 137, 140, 141, 148, 149, 151–153, 155–159, 161–166, 174–176, 178, 179, 182–187, 189, 190, 192, 194, 196–198, 202–235, 239–255, 265, 270, 271, 276, 277, 279–280 amplification.....................................149, 151–152, 163, 166, 179, 192 Polysomy.................................................................... 82, 83 Prediction..........................................................58, 132, 196 Primer design..................................................180, 184–186 Privacy Rule............................................. 2, 3, 11–14, 18, 19 Privacy rule authorization................................................. 12 Prognosis...................................................59, 118, 126, 193 Proliferating cell nuclear antigen (PCNA)................. 48, 49 Protected health information (PHI)................11–14, 19, 20 Protein extraction....................................282, 290–292, 304
J Justice.....................................................................................2
K
Formalin-Fixed Paraffin-Embedded Tissues 312 Index
Proteins.............................25, 27, 38, 39, 41–47, 50–54, 56, 59–61, 69, 84–88, 91, 94, 110, 117, 121, 136, 143, 166, 207, 211, 213, 215, 232, 281–294, 297, 298, 301, 304 Proteolytic digestion.......................... 44, 45, 49, 85, 86, 282 Proteomics........................111, 117, 281, 282, 290, 292, 293 Pyrosequencing....................................................... 181–190
Q qPCR...............................192, 223–229, 234, 243, 250–253, 270, 277, 279, 280 Quantitation......................... 39, 54–60, 173–176, 208, 229, 230, 232, 241, 243, 253, 264 Quantitative RT-PCR.............................219–220, 224–230
R RASSF-1........................................................................ 182 RCL2®-CS100....................................................... 297–307 Real-time PCR....................... 174–176, 194, 202, 203, 224, 242, 243, 253, 254, 265, 279 Regulatory.................................................................... 1–20 Restriction enzyme..................................148, 149, 152, 192 Rituximab......................................................................... 58 RNA..............................57, 85–87, 110, 118–126, 128, 129, 163, 191, 205–235, 240–247, 249, 250, 253–255, 259–267, 269–280, 297–301, 304–307 amplification.............................. 270, 272, 273, 278, 279 extraction.......................... 119, 123–126, 207, 210–214, 232, 241, 243–246, 299, 304 isolation...... 122, 240, 241, 253, 254, 259–267, 271, 274 quality................................ 126, 264–265, 301, 305, 307 quantification......208, 214–217, 241, 246, 254, 264, 306 quantity.................................................................... 126 sizing.................................................208, 214, 217–220 RNeasy............................................................119, 124, 125 RT-PCR.......................... 110, 205–235, 239–255, 265, 270
Sequencing..............................7, 16, 17, 182–185, 187, 192, 283, 284, 288, 290 Single cell analysis.................................................. 111–112 Single nucleotide polymorphisms (SNP).............. 147–159, 162, 176, 177 Slide scanning......................................................92, 93, 107 Specificity......................................39, 44, 49, 50, 56, 62, 86, 175, 176, 196, 202 Storage........................ 40, 43, 52, 73, 75, 77, 101, 118, 123, 142, 178, 179, 188, 221–223, 227, 240, 242, 247–250, 259, 261, 298 Surgical consent.............................................................. 3, 7
T Tandem mass spectrometry.............................282, 284, 288 TaqMan...........................173, 174, 176, 192, 194, 196, 198, 202, 203, 209, 220, 224, 225, 227, 229, 234, 242–244, 250, 253, 254, 265, 279, 280 Tarceva............................................................................. 57 Tile sampling.................................................................. 101 Tissue banking................................................................. 15 Tissue fixation.............................. 51, 62, 70, 120, 177, 206, 209, 210, 240, 307 Tissue microarrays (TMA)......................14, 23–34, 53, 191 Trastuzumab............................................................... 25, 57 Tumor heterogeneity.............................................. 111, 162 Tyramide amplification.................................................... 44
U Usage agreement.......................................................... 6, 19
V Validation.............................................................52–55, 61, 62, 103, 265 Variables......................................51, 52, 54, 56, 61–63, 162, 177, 196, 214, 232, 240, 273, 279, 289
S
W
Semiquantitative......................................................... 25, 54 Sensitivity................................38, 39, 44, 49, 50, 53, 54, 56, 58, 62, 84, 87, 156, 175, 193, 241, 265
Whole genome amplification (WGA)................... 161–179 Whole transcriptome amplification........................ 269–280 World Medical Association................................................ 1
