Laser Capture Microdissection: Methods and Protocols, 2nd Edition (Methods in Molecular Biology Vol 755)

Methods

in

Molecular Biology™

Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfield, Hertfordshire, AL10 9AB, UK



For further volumes: http://www.springer.com/series/7651

wwwwwwwwwwwwwwwww

Laser Capture Microdissection Methods and Protocols Second Edition

Edited by

Graeme I. Murray Department of Pathology, University of Aberdeen, Aberdeen, UK

Editor Graeme I. Murray, MB ChB, PhD, DSc, FRCPath Department of Pathology University of Aberdeen Aberdeen, UK [email protected]

ISSN 1064-3745 e-ISSN 1940-6029 ISBN 978-1-61779-162-8 e-ISBN 978-1-61779-163-5 DOI 10.1007/978-1-61779-163-5 Springer New York Heidelberg London Dordrecht Library of Congress Control Number: 2011931522 © 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. Printed on acid-free paper Humana Press is part of Springer Science+Business Media (www.springer.com)

Preface Laser microdissection techniques have revolutionized the ability of researchers in general, and pathologists in particular, to carry out molecular analysis on specific types of normal and diseased cells and to fully utilize the power of current molecular technologies, including PCR, microarrays, and proteomics. The primary purpose of the second edition of this volume of Methods in Molecular Biology is to provide the reader with practical advice on how to carry out tissue-based laser microdissection successfully in their own laboratory using the different laser microdissection systems that are available and to apply a wide range of molecular technologies. The individual chapters encompass detailed descriptions of the individual laser-based microdissection systems. The downstream applications of the laser microdissected tissue described in the book include PCR in its many different forms as well as gene expression analysis, including the application to microarrays and proteomics. The editor is especially grateful to all the contributing authors for the time and effort they have put into the individual chapters. The series editor John Walker has provided expert guidance through the editorial process while colleagues at Springer have been very helpful in dealing with all the publication related issues. Aberdeen, UK

Graeme I. Murray

v

wwwwwwwwwwwwwwwww

Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

v xi

  1 Laser Capture Microdissection: Methods and Applications . . . . . . . . . . . . . . . . . . 1 Kristen DeCarlo, Andrew Emley, Ophelia E. Dadzie, and Meera Mahalingam   2 Laser Microdissection for Gene Expression Profiling . . . . . . . . . . . . . . . . . . . . . . 17 Lori A. Field, Brenda Deyarmin, Craig D. Shriver, Darrell L. Ellsworth, and Rachel E. Ellsworth   3 Gene Expression Using the PALM System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Jian-Xin Lu and Cheuk-Chun Szeto   4 Immunoguided Microdissection Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Michael A. Tangrea, Jeffrey C. Hanson, Robert F. Bonner, Thomas J. Pohida, Jaime Rodriguez-Canales, and Michael R. Emmert-Buck   5 Optimized RNA Extraction from Non-deparaffinized, Laser-Microdissected Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Danny Jonigk, Friedrich Modde, Clemens L. Bockmeyer, Jan Ulrich Becker, and Ulrich Lehmann   6 Laser Capture Microdissection for Analysis of Gene Expression in Formalin-Fixed Paraffin-Embedded Tissue . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Ru Jiang, Rona S. Scott, and Lindsey M. Hutt-Fletcher   7 MicroRNA Profiling Using RNA from Microdissected Immunostained Tissue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Clemens L. Bockmeyer, Danny Jonigk, Hans Kreipe, and Ulrich Lehmann   8 Profiling Solid Tumor Heterogeneity by LCM and Biological MS of Fresh-Frozen Tissue Sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Donald J. Johann, Sumana Mukherjee, DaRue A. Prieto, Timothy D. Veenstra, and Josip Blonder   9 Amplification Testing in Breast Cancer by Multiplex Ligation-Dependent Probe Amplification of Microdissected Tissue . . . . . . . . . . . 107 Cathy B. Moelans, Roel A. de Weger, and Paul J. van Diest 10 Detection and Quantification of MicroRNAs in Laser-Microdissected Formalin-Fixed Paraffin-Embedded Breast Cancer Tissues . . . . . . . . . . . . . . . . . . 119 Sarkawt M. Khoshnaw, Des G. Powe, Ian O. Ellis, and Andrew R. Green 11 Laser Capture Microdissection Applications in Breast Cancer Proteomics . . . . . . . 143 René B.H. Braakman, Theo M. Luider, John W.M. Martens, John A. Foekens, and Arzu Umar

vii

viii

Contents

12 Proteomic Analysis of Laser Microdissected Ovarian Cancer Tissue with SELDI-TOF MS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Isabelle Cadron, Toon Van Gorp, Philippe Moerman, Etienne Waelkens, and Ignace Vergote 13 LCM Assisted Biomarker Discovery from Archival Neoplastic Gastrointestinal Tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Patricia A. Meitner and Murray B. Resnick 14 Purification of Diseased Cells from Barrett’s Esophagus and Related Lesions by Laser Capture Microdissection . . . . . . . . . . . . . . . . . . . . . Masood A. Shammas and Manjula Y. Rao 15 Laser Microdissection of Intestinal Epithelial Cells and Downstream Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Benjamin Funke 16 Application of Laser Microdissection and Quantitative PCR to Assess the Response of Esophageal Cancer to Neoadjuvant Chemo-Radiotherapy . . . . . . Claus Hann von Weyhern and Björn L.D.M. Brücher 17 Oligonucleotide Microarray Expression Profiling of Contrasting Invasive Phenotypes in Colorectal Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Christopher C. Thorn, Deborah Williams, and Thomas C. Freeman 18 Evaluation of Gastrointestinal mtDNA Depletion in Mitochondrial Neurogastrointestinal Encephalomyopathy (MNGIE) . . . . . . . . . . . . . . . . . . . . . Carla Giordano and Giulia d’Amati 19 Laser Microdissection for Gene Expression Study of Hepatocellular Carcinomas Arising in Cirrhotic and Non-Cirrhotic Livers . . . . . . . . . . . . . . . . . . Maria Tretiakova and John Hart 20 Laser Capture Microdissection of Pancreatic Ductal Adeno-Carcinoma Cells to Analyze EzH2 by Western Blot Analysis . . . . . . . . . . . . . . . . . . . . . . . . . Aamer M. Qazi, Sita Aggarwal, Christopher S. Steffer, David L. Bouwman, Donald W. Weaver, Scott A. Gruber, and Ramesh B. Batchu 21 Laser-Capture Microdissection of Renal Tubule Cells and Linear Amplification of RNA for Microarray Profiling and Real-Time PCR . . . . . . . . . . . Susie-Jane Noppert, Susanne Eder, and Michael Rudnicki 22 Subcellular Renal Proximal Tubular Mitochondrial Toxicity with Tenofovir Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . James J. Kohler and Seyed H. Hosseini 23 Application of Laser-Capture Microdissection to Study Renal Carcinogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kerstin Stemmer and Daniel R. Dietrich 24 Laser-Capture Microdissection and Transcriptional Profiling in Archival FFPE Tissue in Prostate Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ajay Joseph and Vincent J. Gnanapragasam 25 Quantitative Analysis of the Enzymes Associated with 5-Fluorouracil Metabolism in Prostate Cancer Biopsies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tomoaki Tanaka 26 Microdissection of Gonadal Tissues for Gene Expression Analyses . . . . . . . . . . . . Anne Jørgensen, Marlene Danner Dalgaard, and Si Brask Sonne

155

165

181

189

197

203

223

233

245

257

267

279

291

301 307

Contents

27 Duplex Real-Time PCR Assay for Quantifying Mitochondrial DNA Deletions in Laser Microdissected Single Spiral Ganglion Cells . . . . . . . . . . . . . . . Adam Markaryan, Erik G. Nelson, and Raul Hinojosa 28 Neuronal Type-Specific Gene Expression Profiling and Laser-Capture Microdissection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Charmaine Y. Pietersen, Maribel P. Lim, Laurel Macey, Tsung-Ung W. Woo, and Kai C. Sonntag 29 Region-Specific In Situ Hybridization-Guided Laser-Capture Microdissection on Postmortem Human Brain Tissue Coupled with Gene Expression Quantification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . René Bernard, Sharon Burke, and Ilan A. Kerman 30 UV-Laser Microdissection and mRNA Expression Analysis of Individual Neurons from Postmortem Parkinson’s Disease Brains . . . . . . . . . . Jan Gründemann, Falk Schlaudraff, and Birgit Liss 31 Transcriptome Profiling of Murine Spinal Neurulation Using Laser Capture Microdissection and High-Density Oligonucleotide Microarrays . . . . . . . . . . . . . Shoufeng Cao, Boon-Huat Bay, and George W. Yip 32 Probing the CNS Microvascular Endothelium by Immune-Guided Laser-Capture Microdissection Coupled to Quantitative RT-PCR . . . . . . . . . . . . Nivetha Murugesan, Jennifer Macdonald, Shujun Ge, and Joel S. Pachter 33 Laser-Capture Microdissection for Factor VIII-Expressing Endothelial Cells in Cancer Tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tomoatsu Kaneko, Takashi Okiji, Reika Kaneko, Hideaki Suda, and Jacques E. Nör 34 Laser-Capture Microdissection and Analysis of Liver Endothelial Cells from Patients with Budd–Chiari Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Selcuk Sozer and Ronald Hoffman 35 Laser-Capture Microdissection of Hyperlipidemic/ApoE−/− Mouse Aorta Atherosclerosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Michael Beer, Sandra Doepping, Markus Hildner, Gabriele Weber, Rolf Grabner, Desheng Hu, Sarajo Kumar Mohanta, Prasad Srikakulapu, Falk Weih, and Andreas J.R. Habenicht 36 Gene Expression Profiling in Laser-Microdissected Bone Marrow Megakaryocytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kais Hussein 37 Specific RNA Collection from the Rat Endolymphatic Sac by Laser-Capture Microdissection (LCM): LCM of a Very Small Organ Surrounded by Bony Tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kosuke Akiyama, Takenori Miyashita, Ai Matsubara, and Nozomu Mori 38 The Use of Laser Capture Microdissection on Adult Human Articular Cartilage for Gene Expression Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Naoshi Fukui, Yasuko Ikeda, and Nobuho Tanaka 39 Laser-Capture Microdissection of Developing Barley Seeds and cDNA Array Analysis of Selected Tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . . Johannes Thiel, Diana Weier, and Winfriede Weschke

ix

315

327

345

363

375

385

395

405

417

429

441

449

461

x

Contents

40 Quantitative RT-PCR Gene Expression Analysis of a Laser Microdissected Placenta: An Approach to Study Preeclampsia . . . . . . . . 477 Yuditiya Purwosunu, Akihiko Sekizawa, Takashi Okai, and Tetsuhiko Tachikawa Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491

Contributors Sita Aggarwal  •  Pennington Biomedical Research Center, Louisiana State University System, Baton Rouge, LA, USA Kosuke Akiyama  •  Department of Otolaryngology, Faculty of Medicine, Kagawa University, Kagawa, Japan Giulia d’Amati  •  Department of Experimental Medicine, Sapienza University, Rome, Italy Ramesh B. Batchu  •  Laboratory of Surgical Oncology & Developmental Therapeutics, Department of Surgery, Wayne State University, Detroit, MI, USA; John D Dingell VA Medical Center, Detroit, MI, USA Boon-Huat Bay  •  Department of Anatomy, Yong Loo Lin School of Medicine, National University of Singapore, Singapore Jan Ulrich Becker  •  Institute of Pathology, Medizinische Hochschule Hannover, Hannover, Germany Michael Beer  •  Institute for Vascular Medicine, Friedrich Schiller University of Jena, Jena, Germany René Bernard  •  Charité Campus Mitte – Universitätsmedizin Berlin, Centrum für Anatomie, Institut für Integrative Neuroanatomie, Berlin, Germany Josip Blonder  •  Laboratory of Proteomics and Analytical Technologies, SAIC-Frederick, Inc., National Cancer Institute at Frederick, Frederick, MD, USA Clemens L. Bockmeyer  •  Institute of Pathology, Medizinische Hochschule Hannover, Hannover, Germany Robert F. Bonner  •  Pathogenetics Unit and Laser Microdissection Core, Laboratory of Pathology, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA David L. Bouwman  •  Department of Surgery, Wayne State University, Detroit, MI, USA René B.H. Braakman  •  Department of Medical Oncology, Center for Translational Molecular Medicine, and Cancer Genomics Centre, Erasmus MC Rotterdam, Rotterdam, The Netherlands Björn L.D.M. Brücher  •  Comprehensive Cancer Center, University of Tübingen, Tübingen, Germany Sharon Burke  •  Molecular and Behavioral Neuroscience Institute, Ann Arbor, MI, USA Isabelle Cadron  •  Division of Gynecological Oncology, Department of Obstetrics and Gynecology, University Hospitals Leuven, Katholieke Universiteit Leuven, Leuven, Belgium

xi

xii

Contributors

Shoufeng Cao  •  Department of Anatomy, Yong Loo Lin School of Medicine, National University of Singapore, Singapore Ophelia E. Dadzie  •  Dermatopathology Section, St John’s Institute of Dermatology, St. Thomas’ Hospital, London, UK Marlene Danner Dalgaard  •  Department of Growth and Reproduction, Rigshospitalet, Copenhagen, Denmark Kristen DeCarlo  •  Boston University School of Medicine, Boston, MA, USA Brenda Deyarmin  •  Windber Research Institute, Windber, PA, USA Paul J. van Diest   •  Department of Pathology, University Medical Centre Utrecht, Utrecht, The Netherlands Daniel R. Dietrich  •  Human and Environmental Toxicology, University of Konstanz, Konstanz, Germany Sandra Doepping  •  Institute for Vascular Medicine, Friedrich Schiller University of Jena, Jena, Germany Susanne Eder   •  Department of Internal Medicine IV (Nephrology and Hypertension), Functional Genomics Research Group, Center of Internal Medicine, Medical University Innsbruck, Innsbruck, Austria Ian O. Ellis  •  Department of Histopathology, School of Molecular Medical Sciences, University of Nottingham and Nottingham University Hospitals Trust, Nottingham, UK Darrell L. Ellsworth  •  Windber Research Institute, Windber, PA, USA Rachel E. Ellsworth  •  Translational Breast Research, Clinical Breast Care Project, Windber Research Institute, Windber, PA, USA Andrew Emley  •  Dermatopathology Section, Department of Dermatology, Boston University School of Medicine, Boston, MA, USA Michael R. Emmert-Buck  •  Pathogenetics Unit and Laser Microdissection Core, Laboratory of Pathology, Center for Cancer Research, Gaithersburg, MD, USA, Lori A. Field  •  Windber Research Institute, Windber, PA, USA John A. Foekens  •  Department of Medical Oncology, Center for Translational Molecular Medicine, and Cancer Genomics Centre, Erasmus MC Rotterdam, Rotterdam, The Netherlands Thomas C. Freeman  •  Roslin Institute, University of Edinburgh, Edinburgh, UK Naoshi Fukui  •  Clinical Research Center, National Hospital Organization, Sagamihara Hospital, Kanagawa, Japan Benjamin Funke  •  Institute of Pathology, University Hospital Heidelberg, Heidelberg, Germany; Department of Anaesthesiology, University Hospital Heidelberg, Heidelberg, Germany Shujun Ge  •  Blood-Brain Barrier Laboratory, Department of Cell Biology, University of Connecticut Health Center, Farmington, CT, USA Carla Giordano  •  Department of Experimental Medicine, Sapienza University, Rome, Italy Vincent J. Gnanapragasam  •  Translational Prostate Cancer Group, Hutchison MRC Research Centre, University of Cambridge, Cambridge, UK

Contributors

xiii

Rolf Grabner  •  Institute for Vascular Medicine, Friedrich Schiller University of Jena, Jena, Germany Andrew R. Green  •  Department of Histopathology, School of Molecular Medical Sciences, University of Nottingham and Nottingham University Hospitals Trust, Nottingham, UK Scott A. Gruber  •  John D Dingell VA Medical Center, Wayne State University, Detroit, MI, USA Jan Gründemann  •  Wolfson Institute for Biomedical Research, University College London, London, UK Andreas J.R. Habenicht  •  Institute for Vascular Medicine, Friedrich Schiller University of Jena, Jena, Germany Jeffrey C. Hanson  •  Pathogenetics Unit and Laser Microdissection Core, Laboratory of Pathology, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA John Hart  •  Department of Pathology, University of Chicago, Chicago, IL, USA Markus Hildner  •  Institute for Vascular Medicine, Friedrich Schiller University of Jena, Jena, Germany Raul Hinojosa  •  Section of Otolaryngology – Head and Neck Surgery, Department of Surgery, University of Chicago, Chicago, IL, USA Ronald Hoffman  •  Tisch Cancer Institute, Department of Medicine, Mount Sinai School of Medicine, New York, NY, USA; Myeloproliferative Disorder Research Consortium, New York, NY, USA Seyed H. Hosseini  •  Science Department, Georgia Perimeter College, Clarkston, GA, USA Desheng Hu  •  Institute for Vascular Medicine, Friedrich Schiller University of Jena, Jena, Germany Kais Hussein  •  Institute of Pathology, Hannover Medical School, Hannover, Germany Lindsey M. Hutt-Fletcher  •  Department of Microbiology and Immunology, Center for Molecular and Tumor Virology and Feist-Weiller Cancer Center, Louisiana State University Health Sciences Center, Shreveport, LA, USA Yasuko Ikeda  •  Clinical Research Center, National Hospital Organization, Sagamihara Hospital, Kanagawa, Japan Ru Jiang  •  Department of Microbiology and Immunology, Center for Molecular and Tumor Virology and Feist-Weiller Cancer Center, Louisiana State University Health Sciences Center, Shreveport, LA, USA Donald J. Johann  •  Medical Oncology Branch, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Danny Jonigk   •  Institute of Pathology, Medizinische Hochschule Hannover, Hannover, Germany Anne Jørgensen  •  Department of Growth and Reproduction, Rigshospitalet, Copenhagen, Denmark Ajay Joseph  •  Translational Prostate Cancer Group, Hutchison MRC Research Centre, University of Cambridge, Cambridge, UK Reika Kaneko  •  Applied Molecular Medicine, Niigata University Graduate School of Medical and Dental Sciences, Chuo-Ku, Niigata, Japan

xiv

Contributors

Tomoatsu Kaneko  •  Cariology, Operative Dentistry and Endodontics, Niigata University Graduate School of Medical and Dental Sciences, Niigata, Japan Ilan A. Kerman  •  University of Alabama at Birmingham, Department of Psychiatry and Behavioral Neurobiology, Birmingham, AL, USA Sarkawt M. Khoshnaw  •  Department of Histopathology, School of Molecular Medical Sciences, University of Nottingham and Nottingham University Hospitals Trust, Nottingham, UK James J. Kohler  •  Department of Pediatrics, Laboratory of Biochemical Pharmacology, Emory University School of Medicine, Decatur, GA, USA Hans Kreipe   •  Institute of Pathology, Medizinische Hochschule Hannover, Hannover, Germany Ulrich Lehmann  •  Institute of Pathology, Medizinische Hochschule Hannover, Hannover, Germany Maribel P. Lim  •  Laboratory of Cellular Neuropathology, Department of Psychiatry, McLean Hospital, Harvard Medical School, Belmont, MA, USA Birgit Liss  •  Institute of Applied Physiology, University of Ulm, Ulm, Germany Jian-Xin Lu   •  Department of Medicine & Therapeutics, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, Hong Kong, China Theo M. Luider  •  Department of Neurology and Laboratory of Clinical and Cancer Proteomics, Erasmus MC Rotterdam, Rotterdam, The Netherlands Jennifer Macdonald  •  Blood-Brain Barrier Laboratory, Department of Cell Biology, University of Connecticut Health Center, Farmington, CT, USA Laurel Macey  •  Department of Psychiatry, McLean Hospital, Harvard Medical School, Belmont, MA, USA Meera Mahalingam  •  Dermatopathology Section, Department of Dermatology, Boston University School of Medicine, Boston, MA, USA Adam Markaryan  •  Section of Otolaryngology – Head and Neck Surgery, Department of Surgery, University of Chicago, Chicago, IL, USA John W.M. Martens  •  Department of Medical Oncology, Center for Translational Molecular Medicine, and Cancer Genomics Centre, Erasmus MC Rotterdam, Rotterdam, The Netherlands Ai Matsubara  •  Department of Otolaryngology, Faculty of Medicine, Kagawa University, Kagawa, Japan Patricia A. Meitner  •  COBRE Center for Cancer Research Development, Rhode Island Hospital, Providence, RI, USA Takenori Miyashita  •  Department of Otolaryngology, Faculty of Medicine, Kagawa University, Kagawa, Japan Friedrich Modde  •  Institute of Pathology, Medizinische Hochschule Hannover, Hannover, Germany Cathy B. Moelans  •  Department of Pathology, University Medical Centre Utrecht, Utrecht, The Netherlands Philippe Moerman  •  Department of Pathology, University Hospitals Leuven, Katholieke Universiteit Leuven, Leuven, Belgium Sarajo Kumar Mohanta  •  Institute for Vascular Medicine, Friedrich Schiller University of Jena, Jena, Germany

Contributors

xv

Nozomu Mori  •  Department of Otolaryngology, Faculty of Medicine, Kagawa University, Kagawa, Japan Sumana Mukherjee  •  Medical Oncology Branch, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Nivetha Murugesan  •  Blood-Brain Barrier Laboratory, Department of Cell Biology, University of Connecticut Health Center, Farmington, CT, USA Erik G. Nelson  •  Section of Otolaryngology – Head and Neck Surgery, Department of Surgery, University of Chicago, Chicago, IL, USA Susie-Jane Noppert  •  Department of Internal Medicine IV (Nephrology and Hypertension), Functional Genomics Research Group, Center of Internal Medicine, Medical University Innsbruck, Innsbruck, Austria Jacques E. Nör  •  Cariology, Restorative Sciences, and Endodontics, School of Dentistry, University of Michigan, Ann Arbor, MI, USA; Department of Biomedical Engineering, College of Engineering, University of Michigan, Ann Arbor, MI, USA; Comprehensive Cancer Center, University of Michigan, Ann Arbor, MI, USA Takashi Okai  •  Department of Obstetrics and Gynecology, Showa University School of Medicine, Tokyo, Japan Takashi Okiji  •  Cariology, Operative Dentistry and Endodontics, Niigata University Graduate School of Medical and Dental Sciences, Niigata, Japan Joel S. Pachter  •  Blood-Brain Barrier Laboratory, Department of Cell Biology, University of Connecticut Health Center, Farmington, CT, USA Charmaine Y. Pietersen  •  Laboratory of Cellular Neuropathology, Department of Psychiatry, McLean Hospital, Harvard Medical School, Belmont, MA, USA Thomas J. Pohida  •  Pathogenetics Unit and Laser Microdissection Core, Laboratory of Pathology, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Des G. Powe  •  Department of Histopathology, School of Molecular Medical Sciences, University of Nottingham and Nottingham University Hospitals Trust, Nottingham, UK DaRue A. Prieto  •  Laboratory of Proteomics and Analytical Technologies, SAIC-Frederick, Inc., National Cancer Institute at Frederick, Frederick, MD, USA Yuditiya Purwosunu  •  Department of Obstetrics and Gynecology, Showa University School of Medicine, Tokyo, Japan; Department of Obstetrics and Gynecology, University of Indonesia, Cipto Mangunkusumo National Hospital, Jakarta, Indonesia Aamer M. Qazi  •  Department of Surgery, John D Dingell VA Medical Center, Wayne State University, Detroit, MI, USA Manjula Y. Rao  •  Department of Neurology, Center on Human Development and Disability, University of Washington, Seattle, WA, USA Murray B. Resnick  •  Department of Pathology, Rhode Island and The Miriam Hospital, Alpert Medical School, Brown University, Providence, RI, USA Jaime Rodriguez-Canales  •  Pathogenetics Unit and Laser Microdissection Core, Laboratory of Pathology, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA

xvi

Contributors

Michael Rudnicki  •  Department of Internal Medicine IV (Nephrology and Hypertension), Functional Genomics Research Group, Center of Internal Medicine, Medical University Innsbruck, Innsbruck, Austria Falk Schlaudraff  •  Institute of Applied Physiology, University of Ulm, Ulm, Germany Rona S. Scott  •  Department of Microbiology and Immunology, Center for Molecular and Tumor Virology and Feist-Weiller Cancer Center, Louisiana State University Health Sciences Center, Shreveport, LA, USA Akihiko Sekizawa  •  Department of Obstetrics and Gynecology, Showa University School of Medicine, Bunkyo-ku, Tokyo, Japan Masood A. Shammas  •  Department of Medical Oncology, Harvard (Dana Farber) Cancer Institute and VA Boston Healthcare System, Boston, MA, USA Craig D. Shriver   •  Walter Reed Army Medical Center, Washington, DC, USA Si Brask Sonne  •  Department of Biology, University of Copenhagen, Copenhagen, Denmark Kai C. Sonntag  •  Department of Psychiatry, McLean Hospital, Harvard Medical School, Belmont, MA, USA Selcuk Sozer  •  Research Institute for Experimental Medicine (DETAE), Istanbul University, Istanbul, Turkey; Tisch Cancer Institute, Department of Medicine, Mount Sinai School of Medicine, New York, NY, USA Prasad Srikakulapu  •  Institute for Vascular Medicine, Friedrich Schiller University of Jena, Jena, Germany Christopher S. Steffer  •  Department of Surgery, Wayne State University, Detroit, MI, USA Kerstin Stemmer  •  Human and Environmental Toxicology, University of Konstanz, Konstanz, Germany; Department of Internal Medicine, Metabolic Diseases Institute, University of Cincinnati, Cincinnati, OH, USA Hideaki Suda  •  Pulp Biology and Endodontics, Department of Restorative Sciences, Graduate School of Medical and Dental Sciences, Tokyo Medical and Dental University, Bunkyo-ku, Tokyo, Japan Cheuk-Chun Szeto  •  Department of Medicine & Therapeutics, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, Hong Kong, China Tetsuhiko Tachikawa  •  Department of Oral Pathology, Showa University School of Dentistry, Tokyo, Japan Nobuho Tanaka  •  Clinical Research Center, National Hospital Organization, Sagamihara Hospital, Kanagawa, Japan Tomoaki Tanaka  •  Department of Urology, Osaka City University Graduate School of Medicine, Osaka, Japan Michael A. Tangrea  •  Pathogenetics Unit and Laser Microdissection Core, Laboratory of Pathology, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Johannes Thiel  •  Leibniz-Institut für Pflanzengenetik und Kulturpflanzenforschung, Gatersleben, Germany

Contributors

xvii

Christopher C. Thorn  •  Department of Academic Surgery, St. James’s University Hospital, Leeds, UK Maria Tretiakova  •  Department of Pathology, University of Chicago, Chicago, IL, USA Arzu Umar  •  Netherlands Proteomics Center, Erasmus MC Rotterdam, Rotterdam, The Netherlands; Department of Medical Oncology, Center for Translational Molecular Medicine, and Cancer Genomics Centre, Erasmus MC Rotterdam, Rotterdam, The Netherlands Toon Van Gorp  •  Division of Gynecological Oncology, Department of Obstetrics and Gynecology, University Hospitals Leuven, Katholieke Universiteit Leuven, Leuven, Belgium; Division of Gynaecological Oncology, Department of Obstetrics and Gynaecology, MUMC+, GROW – School for Oncology and Developmental Biology, Maastricht, The Netherlands Timothy D. Veenstra  •  Laboratory of Proteomics and Analytical Technologies, SAIC-Frederick, Inc., National Cancer Institute at Frederick, Frederick, MD, USA Ignace Vergote  •  Division of Gynecological Oncology, Department of Obstetrics and Gynecology, University Hospitals Leuven, Katholieke Universiteit Leuven, Leuven, Belgium Etienne Waelkens  •  Department of Molecular Cell Biology, University Hospitals Leuven, Katholieke Universiteit Leuven, Leuven, Belgium Donald W. Weaver  •  Department of Surgery, Wayne State University, Detroit, MI, USA Gabriele Weber  •  Institute for Vascular Medicine, Friedrich Schiller University of Jena, Jena, Germany Roel A. de Weger  •  Department of Pathology, University Medical Centre Utrecht, Utrecht, The Netherlands Diana Weier  •  Leibniz-Institut für Pflanzengenetik und Kulturpflanzenforschung, Gatersleben, Germany Falk Weih  •  Institute for Vascular Medicine, Friedrich Schiller University of Jena, Jena, Germany Winfriede Weschke  •  Leibniz-Institut für Pflanzengenetik und Kulturpflanzenforschung, Gatersleben, Germany Claus Hann von Weyhern  •  Comprehensive Cancer Center, University of Tübingen, Tübingen, Germany Deborah Williams  •  MRC Harwell, Oxford, UK Tsung-Ung W. Woo  •  Laboratory of Cellular Neuropathology, Department of Psychiatry, McLean Hospital, Harvard Medical School, Belmont, MA, USA George W. Yip  •  Department of Anatomy, Yong Loo Lin School of Medicine, National University of Singapore, Singapore

wwwwwwwwwwwwwww

Chapter 1 Laser Capture Microdissection: Methods and Applications Kristen DeCarlo, Andrew Emley, Ophelia E. Dadzie, and Meera Mahalingam Abstract Laser microdissection is a nonmolecular, minimally disruptive method to obtain cytologically and/or phenotypically defined cells or groups of cells from heterogeneous tissues. It is a versatile technology and allows the preparation of homogenous isolates of specific subpopulations of cells from which RNA/DNA or protein can be extracted for RT-polymerase chain reaction (PCR), quantitative PCR, Western blot analyses, and mass spectrophotometry. Key words: DNA analysis, Laser capture microdissection, Melanoma, PCR, Proteomics, RNA analysis

1. Introduction The molecular analysis of DNA, RNA, and protein derived from diagnostic tissue, has revolutionized pathology and led to the identification of a broad range of diagnostic and prognostic markers (1). Analysis of critical gene expression and protein patterns in normal developing and diseased tissue progression requires the microdissection and extraction of a microscopic homogeneous cellular subpopulation from its complex tissue milieu (2). However, the reliability of tests based on tissue or cell extracts often depends crucially on the relative abundance of the cell population in question (1). Therefore, a prerequisite for modern molecular research is the capability of preparing pure samples without a large number of “contaminating” cells (1, 3). Laser capture microdissection (LCM) offers a simple, one-step process that provides scientists with a fast and dependable method of preserving and isolating single cells, or clusters of cells, from tissue sections by direct microscopic visualization (2, 4, 5). Graeme I. Murray (ed.), Laser Capture Microdissection: Methods and Protocols, Methods in Molecular Biology, vol. 755, DOI 10.1007/978-1-61779-163-5_1, © Springer Science+Business Media, LLC 2011

1

2

K. DeCarlo et al.

1.1. History

The need to isolate specific cells from complex tissues in order to carry out accurate molecular assays has been argued for decades (6). In the 1970s, Lowry and Passonneau pioneered a procedure for biochemical microanalysis, which utilized “freehand” microdissection of specific cell types under a microscope (6, 7). At the same time, several papers described different techniques that were also based on manual dissection (under microscope control) using razor blades, needles, or fine glass pipettes to isolate the cells of interest (6). An obvious shortcoming is that manual microdissection is time consuming, tedious, and does not allow for precise control of the material effectively selected (6, 7). A significant technological advance was proposed by Shibata in 1993 who suggested selective ultraviolet radiation fractionation, a procedure which utilized an ultraviolet laser beam to destroy the DNA of all undesired components of the tissue, while the cells of interest were protected by a specific dye (6–8). Unfortunately, this technique is only useful for analytes that are susceptible to degradation by UV-light, such as DNA (7). Subsequent improvements of this procedure led to the development of more sophisticated techniques that enabled isolation of single cells (6). The LCM system was developed during the mid-1990s by Dr. Emmert-Buck and colleagues at the National Institutes of Health (NIH), Bethesda, ML, USA (9). The system was initially developed for the analyses of solid tumors, and was later commercialized by Arcturus Engineering (Sunnyvale, CA, USA) as the PixCell system (6, 9). The PixCell series is currently the most widely used laser-based microdissection system, its development propelled by its integration into the “cancer genome anatomy project” (CGAP) sponsored by the National Cancer Institute (NCI) (1, 9). Multiple generations of this instrument (PixCell II; Arcturus Engineering, Mountain View, CA, USA) are currently on the market (1). Arcturus has also recently commercialized a new system (VeritasTM microdissection) that combines their LCM system, based on infrared laser, with UV laser cutting possibilities, the latter ideal for nonsoft tissues, and capturing large numbers of cells (6, 10).

2. Overview 2.1. Principle

The LCM system by Arcturus (PixCell II) is based on the selective adherence of visually targeted cells and tissue fragments to a special thermoplastic film made of an ethylene vinyl acetate (EVA) membrane activated by a low energy infrared laser pulse (1, 6). The system consists of an inverted microscope, a solid-state nearinfrared laser diode, a laser control unit, a joystick controlled microscope stage with a vacuum chuck for slide immobilization, a charge coupled device camera, and a color monitor. The LCM

1  Laser Capture Microdissection: Methods and Applications

3

microscope is usually connected to a personal computer for additional laser control and image archiving (1). The thermoplastic membrane used for transfer of selected cells is manufactured on the bottom surface of a plastic support cap, which acts as an optic for focusing the laser (1, 11). It has a diameter of approximately 6 mm and fits on standard 0.5 ml microcentrifuge tubes to facilitate further tissue processing (1). The cap is suspended on a mechanical transport arm and placed on the desired area of the mounted tissue sections (1). After visual selection of the desired cells, laser activation leads to focal melting of the EVA membrane, which has its absorption maximum near the wavelength of the laser (1). The polymer melts only in the vicinity of the laser, and expands into the section filling small hollow spaces present in the tissue (1, 11). Properly melted polymer spots have a dark outer ring and a clear center, indicating that the polymer has melted and is in direct contact with the slide (Fig. 1) (11). The polymer then resolidifies within milliseconds (ms) and forms a composite with the tissue (1). A dye incorporated into the polymer serves two purposes: first, it absorbs laser energy, preventing damage to the cellular constituents, and second, it aids in visualizing areas of melted polymer (11). The adherence of the tissue to the activated membrane exceeds the adhesion to the glass slide and allows for selective removal of the desired cells (1). Laser pulses between 0.5 and 5 ms in duration repeated multiple times across the cap surface, allow for rapid isolation of large numbers of cells (1). Lifting the cap then shears the selected cells from the heterogeneous tissue section (1, 11). The minimum diameter of the laser beam (7.5 mm) has been reduced in the newer generation machine. Under standard working

Fig. 1. LCM polymer bubbles. Properly melted polymer bubbles have a dark outer ring, indicating the polymer has melted and is in direct contact with the slide. (a) Larger spots can be created by increasing the power and spotsize of the laser to 100 mW and 30 mm, respectively. (b) Smaller spots can be created by decreasing the power and spotsize of the laser to 30 mW and 10 mm, respectively.

4

K. DeCarlo et al.

conditions, the area of the polymer melting corresponds exactly to the laser spot size. Also, since most of the energy is absorbed by the membrane, the maximum temperatures reached by the tissue upon laser activation are in the range of 90°C for several milliseconds, thus leaving biological macromolecules intact (1). The short laser pulse durations used (0.5–5.0 ms), the low laser power levels required (1–100 mW), the absorption of the laser pulse by the dye-impregnated polymer, and the long elapsed time (0.2 ms) between laser pulses combine to prevent any significant amount of heat deposition at the tissue surface which might compromise the quality of the tissue/cells utilized in later laboratory analyses (1, 9, 11). 2.2. Tissue Fixation, Sectioning, and Staining

Laser-based microdissection techniques have been applied to a wide range of tissues, prepared with a variety of methods, and utilizing a diverse range of biological samples (9). However, the procedures used in the preparation of tissue or cells for microdissection vary with the intended purposes and the analytes sought (7). Tissue specimens are typically either fixed in aldehyde-based fixatives (e.g., 10% formalin) or snap frozen (12). Formalin-fixation (10% buffered formaldehyde) is the standard for morphologic preservation of tissue, and has been used in histology laboratories for decades because of its low cost and rapid, complete penetration of tissue (7, 11). Although formalinfixed tissues are well preserved for histopathological evaluation, the quality of the macromolecules is severely compromised (12). It is an “additive” fixative that creates cross-links between itself and proteins, and between nucleic acids and proteins (6). This cross-linking interferes with recovery of nucleic acids and proteins, as well as the amplification of DNA and RNA by polymerase chain reaction (PCR) (6, 7). As a consequence of these crosslinks, the nucleic acids isolated from these specimens are highly fragmented, especially as fixation time is increased (6). This problem often occurs when using archival material, especially since pathology laboratories did not pay much attention to fixation times in the past (6). Fortunately, it has been shown that shorter lengths of DNA, up to approximately 200 base pairs, are recoverable by PCR after extraction from formalin fixed-paraffin embedded (FFPE) tissue (7). Ethanol-based fixatives offer the best RNA preservation by fixing tissues through dehydration without creating chemical links (6, 7). However, it has been found that sectioning with alcoholbased fixatives is more difficult (13). Therefore, the use of alcohol fixatives is only feasible if microdissection is considered as one of the possible options for processing the sample from the start (6). In the case of histological preparations, it is certainly better to utilize samples that have been snap-frozen and stored in liquid

1  Laser Capture Microdissection: Methods and Applications

5

nitrogen at 80°C or colder (6, 7). Frozen sections do not undergo cross-linking due to fixatives, and as a result, yield high quality messenger RNA (mRNA) and proteins (6, 12). However, freezing and cryostat sectioning can significantly disrupt the histological architecture of the tissue (12). This is a major problem since LCM is accomplished through identification of cells by morphological characteristics (11). The main goal of tissue preparation is to ensure that both the morphology of the tissue and molecules of interest are preserved (9). Recently, methods have been developed for the extraction and amplification of RNA from FFPE tissue sections (14). Like fresh tissue, mRNA amplification by nested RT-PCR (reversetranscriptase PCR) has been reported for single cells isolated from FFPE tissue through LCM (1, 15). Similarly, there has been a development of protocols which permit the extraction and mass spectrometric analysis of proteins from FFPE tissues (9). However, even though new technologies are being developed to reverse cross-linking for extraction of sufficient quantities of nucleic acids and proteins, high-quality yield of RNA and proteins is best achieved with frozen or ethanol-fixed tissue (11). The ability to effectively break the cross-links in nucleic acid caused by formalin could allow the utilization of a wealth of archived FFPE tissue for RNA expression and genomic analysis (7). Optimal LCM is achieved with tissue sections cut at a thickness of 2–15 mm (11). Tissue sections thinner than 5 mm may not provide full cell thickness, necessitating multiple microdissections in order to obtain an adequate number of cells for a given assay (11). Tissue sections thicker than 15  mm may not microdissect completely, leaving integral cellular components adhering to the slide (11). Ideally, staining should provide an acceptable morphology to allow the selection of target cells without interfering with the macromolecules of interest, or subsequent molecular techniques (6). Therefore, tissue sections should be exposed to the dye solution for the briefest period of time (9, 11). Minimal staining times limit potential protein alterations, and reduce the risk of chemical modification due to contact with reagents (9, 11). Sections can be stained satisfactorily by a few seconds exposure to the dye solution, followed by removal of excess dye with rapid washing (9). Examples of LCM-compatible stains are hematoxylin and eosin (most commonly used for examination of histologic sections), methylene blue, Wright-Giemsa, and toluidine blue (7, 11). In our experience, eosin staining is not necessary for visualization of cells. Specimens can also be stained immunohistochemically or with fluorescent labels, allowing the investigator to target cells based on the presence of specific antigens (7, 9). Stained sections are dehydrated and kept without a coverslip (6).

6

K. DeCarlo et al.

3. Protocols 3.1. Evolution of LCM Protocols

Due to the infancy of LCM technology, protocols have been constantly changing. Our own experience confirms this. At the outset, tissue slides were cut in 7–10-mm sections and mounted on uncharged slides. However, we found that 5 mm sections allowed for better procurement of cells, particularly in melanoma samples (Fig. 2). It is our contention that in this entity, thinner tissue sections allowed the melted polymer to more effectively penetrate tissue samples, thus enhancing yield. Similarly, modifications in LCM power and spotsize have led to more efficient tissue retrieval. Initially, the PixCell IIe LCM machine was used with a power ranging from 70 to 100 mW and a spotsize of 10 mm. However, after numerous trials, we found that a power of 80  mW and a spotsize of 7.5 mm were most effective in optimizing our yield.

3.2. Factors Affecting Yield of DNA, RNA, and Protein

These include quality of sample, time of preservation before microdissection, type of preservation, fixation method, and efficiency of microdissection (2). In our experience, fixation is the most critical step to ensure a high-quality yield of DNA, RNA, or protein (11). Quality of fixation is dependent on the length of time for fixative penetration in the tissue, temperature of fixation, and tissue size (2). In contrast to DNA, mRNA and protein are more sensitive to fixation, are quickly degraded, and require stringent RNase and proteinase-free conditions during specimen handling and preparation (1, 2). Therefore, the longer the fixative takes to penetrate the tissue, the greater the chance of RNA or protein degradation due to these ubiquitous RNases and proteinases (2). As a result, tissue microdissection is currently more widely employed in the analysis of DNA, as opposed to RNA and proteins, which are much more sensitive to degradation and fixation (6). In general, one set of microdissected cells is used for the downstream analysis of only one type of molecule (2). Each class of molecule requires different solubilization schemes, extraction buffers, and denaturing temperatures. For example, a population of 10,000 microdissected cells could be solubilized in denaturing buffer at 70°C for downstream protein analysis, while a second set of 100 cells could be treated with proteinase K at 65°C for downstream DNA analysis (2). Captured cells are detached from the cap membrane by proteinase digestion, and standard singlestep PCR protocols can be applied if enough cells have been collected (1). As can be seen, it is often necessary to microdissect many more cells than necessary based solely on DNA, RNA, or protein content of a cell (11). Examples of cellular yield required for DNA, RNA, and protein analyses are greatly varied, and range from 100 to 2,000 cells for DNA, 5,000–10,000 cells for RNA, and up to 4,000–200,000 cells for protein analyses (2).

1  Laser Capture Microdissection: Methods and Applications

7

Fig. 2. LCM of melanoma. (a) Melanoma nested in heterogeneous tissue section prior to LCM (40× magnification). (b) Melanoma after LCM. (c) Melted polymer bubble containing melanoma cells extracted from the heterogeneous tissue section.

3.3. Utility 3.3.1. Advantages

Perhaps the most relevant advantage of LCM is its speed while maintaining precision and versatility (1). LCM provides a reliable method to procure pure, precise populations of target cells from a wide range of cell and tissue preparations via microscopic visualization (16). The LCM system is applicable to normal glass slides (along with a wide range of other preparations), allowing routinely prepared material to be used after removal of the coverslip (6). Conventional techniques for molecular analysis are based on whole tissue dissociation and therefore introduce inherent contamination problems, thus reducing the specificity and sensitivity to subsequent molecular analysis, while requiring a high level of manual dexterity. LCM on the other hand, is a “no touch” technique that does not destroy adjacent tissues following initial microdissection. This allows several tissue components to be sampled sequentially from the same slide (e.g., normal and atypical cells) (1, 6, 16). LCM creates no chemical bonds to the target tissue so molecules in LCM-transferred cells are not degraded when compared to the original tissue slide (17). Furthermore, LCM isolates cells via firm adherence to the cap, reducing tissue loss, where other microdissection techniques require the removal of the isolated cells with the help of a needle tip or a microcapillary (1). The LCM technique is easily documented via a database program able to record images of both captured cells and residual tissue before and after microdissection. This diagnostic record is critical for maintaining an accurate record of each dissection, and for correlating histopathology with subsequent molecular analysis (6, 16). A final, critical advantage of LCM is its application to FFPE material, one of the most widely practiced methods for clinical sample preservation and archiving. Recent discoveries show promising advances in the use of FFPE tissues with LCM and subsequent molecular analysis. Collections of FFPE tissues comprise an invaluable resource for retrospective molecular studies of diseased tissues, including translational studies of cancer development (7, 14).

8

K. DeCarlo et al.

3.3.2. Disadvantages

The few limitations of LCM mostly reflect the difficulties of microdissection in general (1). Cell identification is performed in conjunction with a pathologist, and is based upon the morphological characteristics of the cells of interest (11). However, sections for microdissection are dehydrated and kept without a coverslip, making visualization of certain samples difficult due to decreased cellular detail (6, 18). This sometimes makes precise dissection of cells from complex tissues very difficult. However, this problem can be circumvented by special stains, in particular immunohistochemical stains, which help highlight cell populations to be isolated or avoided (1). Unfortunately, standard immunohistochemical staining protocols require several hours, which can lead to further degradation of RNA and protein by RNases and proteinases, respectively (1, 2, 6). Fixation, dehydration, and staining of tissue sections also makes “live-cell analysis” (18) impossible. Another problem occasionally encountered in LCM is failure to remove selected cells from the slide (1). This can result from a lack of adherence of the cells to the EVA membrane, usually because of incomplete dehydration or a laser setting that is too low for complete permeation of the melted polymer into the section (1, 6). On the other hand, increased adherence of the section to the slide can prevent the removal of the targeted cells (1). As a result, isolation of large numbers of cells (e.g., for protein analysis) from many sections can require considerable time (2, 18). Older machines face problems related to a minimum laser spot size of 7.5 mm, which imposes restrictions on the precision of LCM recovery and makes it difficult to isolate cells of interest without contamination. The more recent generation of LCM machines, capable of dissecting cells at single cell level, have overcome these limitations (1).

3.4. Clinical Applications

LCM has significantly enhanced the molecular analysis of pathological processes as it offers a simple and efficient technique for procuring a homogeneous population of cells from their native tissue via direct microscopic visualization. LCM makes it possible to analyze cellular function between neighboring, intermingling, and morphologically identifiable cells within complex tissues and organs (17). Overall, LCM is applicable to molecular profiling of tissue in normal and disease states; this includes correlations of cellular molecular signatures within specific cell populations and the comparison of different cellular elements within a single tissue microenvironment (11). LCM-based molecular analysis is being used in many fields of research, including the study of normal cell biology, as well as in vivo genomic and proteomic states such as the profiling of cultured intervertebral disc cells, molecular analysis of skeletal cell differentiation, and gene expression in testicular cell populations (16, 19–23). Other studies focusing on the molecular analysis of

3.4.1. An Overview

1  Laser Capture Microdissection: Methods and Applications

9

histopathological lesions and disease processes include mapping genetic alterations associated with the progression of premalignant cancer lesions (breast cancer and their lymph node metastases, ovarian cancer, and prostate cancer); analyses of gene expression patterns in multiple sclerosis, atherosclerosis, and Alzheimer’s disease plaques; diagnosis of infectious micro-organisms; and the analysis of genetic abnormalities in utero from selected fetal cells in maternal fluids (1, 2, 9, 16). LCM is currently being used in the Cancer Genome Anatomy Program (CGAP) and exemplifies the molecular advances that LCM offers, as it allows researchers to catalog genes that are expressed in human tissue as normal cells undergo premalignant changes and further develop into invasive and metastatic cancer. Changes in expressed genes or alterations in cellular DNA corresponding to a specific disease state can be compared within or between individual patients, as a large number of microdissected cDNA libraries (produced from microdissected normal and premalignant tissue RNA) have been produced and published on the CGAP web page. This catalog of gene expression patterns has the potential to provide clues to etiology and, hopefully, contribute to early diagnostic detection and more accurate diagnosis of disease, followed by therapies tailored to individual patients (11, 16, 17). LCM has been applied to genomic analyses such as studies of X-chromosome inactivation patterns to assess clonality, promoter hypermethylation, restriction fragment length polymorphisms, and single strand conformation polymorphism analysis for assessment of mutations in critical genes such as p53 and K-RAS. Novel uses include cancer chemoprevention, biomarker discovery, and live and rare cell isolation. LCM has been used for biomarker discovery in various human tissue types and organ systems. In these studies, LCM is used in combination with DNA transcriptome profiling to identify differentially expressed genes (24). Intermediate endpoint biomarkers, used to monitor the success of chemoprevention, have been successfully developed for prostate cancer, cervical carcinoma, and adenomas for colorectal cancer (24). Finally, LCM has been applied to the study of live and rare cell populations. Remarkably, LCM has no influence on the viability, metabolism, and proliferation rate of isolated living cells where even an entire living organism (such as the nematode Caenorhabditis elegans) can be successfully transferred without compromising the biological composition or viability of the organism. Live cell LCM isolation equipment is available from several manufacturers (25). Finally, LCM is being used to isolate rare cells. In this rapidly developing method, rarely occurring cells are identified with automated scanning software, immediately followed by computer-controlled LCM (25).

10

K. DeCarlo et al.

3.4.2. DNA Analysis

Microdissection is now an established technique used to collect homogeneous cell populations for the analysis of genetic alterations at the DNA level (1). With the advent of efficient analytical methods for small amounts of biological material, LCM is applied in pathological diagnosis, classification, and treatment of tumors. It even plays a major role in gene mutation studies where a homogenous tumor cell population is necessary for accurate genomic analysis (1).

3.4.3. RNA Analysis

LCM offers several other advantages for mRNA analysis as compared to other laboratory techniques such as mRNA in situ hybridization or immunohistochemistry. Microdissection of purified cells, in combination with methods such as real-time quantitative RT-PCR, allows for a precise determination of cell-specific gene expression (25). Furthermore, LCM is an efficient technique that allows sampling of large numbers of cells without significant RNA degradation where tissue dehydration may even inhibit the activity of tissue RNases, thereby maintaining the tissue integrity during specimen handling and preparation (1). Gene expression analysis is critical in uncovering the patterns related to neoplastic transformation, however, the simultaneous detection of multiple different messages is preferable over the examination of single or few expressed genes. Therefore, microdissected cells are used in conjunction with cDNA array hybridization or serial analysis of gene expression to reveal the differences in gene expression profiles of normal and neoplastic cells, or to show altered gene expression patterns at various stages of cancer progression. LCM is also an essential tool in this process, as mRNA from microdissected lesions is subsequently used as the precursor for creating cDNA and expression libraries from purified cell populations (1).

3.4.4. Proteomic Analyses

Proteomics aims to establish the complete set of proteins or the “proteome” that are important in normal cellular physiology. The normal proteome is compared to a disease state proteome such as cancer using a variety of analyses including western blotting, high-resolution two-dimensional polyacrylamide gel electrophoresis (2-D PAGE), and mass spectrometry and peptide sequencing. Proteomics is a complementary approach to gene expression studies and provides supplementary information not obtained through genome or transcriptome analysis (24, 25). Deciphering alterations in proteomic profiles using LCM techniques offers the advantage of studying physiological relationships unique to protein analysis, thereby offering the potential to identify novel diagnostic and therapeutic targets.

3.4.5. Singe Cell Analysis

LCM has been applied to the isolation of single cells for the analysis of specific targets such as the identification of point mutations

1  Laser Capture Microdissection: Methods and Applications

11

in oncogenes such as RAS and the amplification of expressed gene sequences by RT-PCR. Additionally, microdissected single cells can be used as a template for whole genome amplification, the generation of expression libraries, or probes for expression profiling with cDNA arrays (1). Saurez-Quian et al. has modified the LCM protocol specifically for single cell capturing. In this technique, a cylinder covered with EVA polymer membrane has replaced the large cap surface. This decreases the contact area with the tissue and increases the accuracy of procuring a homogenous cell population (1). 3.5. Specific Diagnostic Applications 3.5.1. Tumors

3.5.2. Clonality Studies

The identification of genetic mutations is paramount in the pathological diagnosis, classification, and treatment of tumors. Loss of heterozygosity (LOH) analysis has been pivotal in cancer research for mapping of tumor suppressor genes, localization of putative chromosomal “hot spots,” and the study of sequential genetic changes in preneoplastic lesions. Microdissection has become a key technique used in LOH studies, since pure populations of tumor cells are necessary, and contamination by even a few unwanted cells may result in inappropriate amplification (via PCR) of the “lost” second allele present in noncancer tissue. LOH studies preformed via microdissection have shown that the frequencies of genetic alterations have been largely under-­ estimated such that there may even be heterogeneity present within a single tumor where some genetic changes occur early in tumorigenesis (24). Furthermore, LCM has been applied to the study of protein alterations in preneoplastic lesions and their tumor counterparts in an effort to elucidate novel tumor-specific alterations in peptide products of cancer cells. From these proteomic studies, distinct protein expression patterns have successfully classified normal, premalignant and malignant cancer cells collected using LCM from human tissues (24). Recently, it has become possible to use smaller samples of cells (not more than 20–100 dissected cells per PCR) obtained via microdissection, allowing a more refined study of preneoplastic lesions in addition to neoplastic lesions. This has been made possible using a combination of microdissection with primer extension preamplification and whole genome amplification techniques, thereby opening a whole new frontier in cancer research (24). Assessing clonality via DNA analyses using LCM has played an instrumental role in identifying the multiple endocrine neoplasia type 1 gene (MEN1), and will hopefully uncover the genetic basis underlying other cancer types. In the case of MEN1, LOH analysis of 200 microdissected endocrine tumors narrowed the interval of the genetic aberration to 300 kb. This LOH information from LCM analysis was used in conjunction with haplotyping

12

K. DeCarlo et al.

and newly identified polymorphic markers, and led to the identification of a new tumor suppressor gene responsible for MEN1 (1, 24). LCM, in conjunction with DNA analyses, has the ability to distinguish the presence of two clonal populations in the same tumor site. Fend et  al. have demonstrated this in malignant nonHodgkin’s lymphoma, where two phenotypically and morphologically distinct cell populations were present in the same tumor. In this study, LCM was used to procure homogenous samples of the two populations from immunostained slides. Subsequent sequencing of rearranged immunoglobulin genes confirmed the presence of two unrelated clones in all cases. LCM played a pivotal role in this study, as PCR analysis of DNA obtained from whole sections was not able to detect the biclonal composition of the tumors (1). 3.5.3. Clinical Applications of LCM in Dermatopathology

LCM has broadened the role of dermatopathology in molecular diagnosis and has greatly enhanced the understanding of the pathogenesis of inherited skin diseases (9, 26, 27). The examination of precancerous lesions by LCM has been applied in the study of melanomagenesis, as it is widely believed that benign nevi undergo genetic alterations that progressively lead to melanoma development. LCM is used to assess the incidence of genetic gains and losses in tumors and preneoplastic lesions, and in doing so, has the potential to uncover the molecular events associated with the transformation of banal nevi into malignant melanoma formation (28–30).

Nevi Versus Melanoma

From a histopathological perspective, melanoma development is tracked by a series of melanocyte transitions from easily characterized precursors. However, from a genetic perspective, these transitions are poorly understood (30). LCM, therefore, has the potential to shed light on the genetic profiles of melanocytes as they undergo these morphological transitions, hopefully uncovering the molecular events that lead to melanomagenesis. Using LCM to dissect distinct populations of nevic aggregates in association with melanoma, we have been able to show that banal nevic aggregates might serve as precursor lesions (31).

Clonality in Cutaneous T-cell Lymphoma

Analysis of T-cell gene rearrangement in cutaneous T-cell lymphoma (CTCL) has led to the discovery that the earliest manifestation of CTCL may be “clonal dermatitis.” Clonal dermatitis is a chronic form of dermatitis that contains a dominant T-cell clone but does not show the typical histologic features diagnostic for CTCL. Significantly, approximately 25% of clonal dermatitis cases develop into CTCL within 5 years, where the same clone is present in both the clonal dermatitis and the CTCL lesions, indicating that the clonal dermatitis clone is a precursor to the CTCL. LCM is ideal for this type of study as the often-sparse lymphocytic

1  Laser Capture Microdissection: Methods and Applications

13

infiltrate can be specifically captured. Furthermore, these studies allow unprecedented investigations into the molecular pathogenesis of CTCL, which will hopefully lead to early disease detection and help guide gene therapy (32). LCM is also able to demonstrate genetically different clones or gene mutations limited to one specific neoplastic population. This has been an important tool in cutaneous lymphoma lesions containing a mixed B- and T-cell population. Using microdissection followed by genotypic analysis, Gallardo et  al. established that the lesion of interest in the case study was cutaneous B-cell lymphoma with a dual B- and T-cell genotype. Conventional methods were not able to make this distinction, therefore illustrating the usefulness of LCM in clinical diagnostics (33). Infectious Diseases

LCM allows for the isolation of pure cell populations which can be screened through PCR for infectious agents depending on the clinical and histological suspicion. LCM also plays an important role in routine histopathologic diagnostics and has been applied to the diagnosis of infectious diseases such as borreliosis, herpes simplex virus infection, herpes zoster, Epstein–Barr virus infection, Myobacterium tuberculosis, and many others (5, 34).

4. Conclusions Tissue-based laser microdissection is a powerful technique, which combines morphology, histochemistry, and sophisticated downstream molecular analysis (35). High speed, easy handling, and good control and documentation of dissected tissue make LCM an ideal tool for the rapid collection of larger amounts of tissue. Further technological advances such as touch-screen cell annotation, automated cell microdissection, and cell recognition software are leading to the next generation machines with enhanced microdissection capabilities. The ability of LCM to visualize and capture specific populations of cells has made LCM an important diagnostic tool, not just in dermatopathology. References 1. Fend F, Raffeld M (2000) Laser capture microdissection in pathology. J Clin Pathol 53, 666–672 2. Espina V, Heiby M, Pierobon M et al (2007) Laser capture microdissection technology. Expert Rev Mol Diagn 7, 647–657 3. Burgemeister R (2005) New aspects of laser capture microdissection in research and routine. J Histochem Cytochem 53, 409–412

4. Agar NS, Halliday GM, Barnetson RS et  al (2003) A novel technique for the examination of skin biopsies by laser capture microdissection. J Cutan Pathol 30, 265–270 5. Yazdi AS, Puchta U, Flaig MJ et  al (2004) Laser-capture microdissection: Applications in routine molecular dermatopathology. J Cutan Pathol 31, 465–470 6. Esposito G (2007) Complementary techniques: Laser capture microdissection—increasing

14

K. DeCarlo et al.

specificity of gene expression profiling of cancer specimens. Adv Exp Med Biol 593, 54–65 7. Eltoum IA, Siegal GP, Frost AR (2002) Microdissection of histologic sections: Past, present, and future. Adv Anat Pathol 9, 316–322 8. Shibata D (1993) Selective ultraviolet radiation fractionation and polymerase chain reaction analysis of genetic alterations. Am J Pathol 143, 1523–1526 9. Murray GI (2007) An overview of laser capture microdissection technologies. Acta Histochem 109, 171–176 10. Veritas Microdissection System. MDS Analytical Technologies. http://www.moleculardevices. com/pages/instr uments/veritas.html. Accessed 11 July 2008 11. Espina V, Wulfkuhle JD, Calvert VS et  al (2006) Laser-capture microdissection. Nat Protoc 1, 586–603 12. Ahram M, Flaig MJ, Gillespie JW, et al (2003) Evaluation of ethanol-fixed, paraffin-embedded tissues for proteomic applications. Proteomics 3, 413–421 13. Bostwick DG, al Annouf N, Choi C (1994) Establishment of the formalin-free surgical pathology laboratory. Utility of an alcoholbased fixative. Arch Pathol Lab Med 118, 298–302 14. Gianni L, Zambetti M, Clark K 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 15. Schutze K, Lahr G (1998) Identification of expressed genes by laser-mediated manipulation of single cells. Nat Biotechnol 16, 737–742 16. Bonner RF, Emmert-Buck M, Cole K et  al (1997) Laser capture microdissection: Molecular analysis of tissue. Science 278, 1481–1483 17. Simone NL, Bonner RF, Gillespie JW et  al (1998) Laser-capture microdissection: Opening the microscopic frontier to molecular analysis. Trends Genet 14, 272–276 18. Brignole E (2000) Laser-capture microdissection: Isolating individual cells for molecular analysis. Mod Drug Discovery 3, 69–70 19. Gruber HE, Mougeot JL, Hoelscher G et al (2007) Microarray analysis of laser capture microdissected-anulus cells from the human intervertebral disc. Spine 32, 1181–1187 20. Benayahu D, Socher R, Shur I (2008) Application of the laser capture microdissection technique for molecular definition of skeletal cell differentiation in  vivo. Methods Mol Biol 455, 191–201

21. Sluka P, O’Donnell L, McLachlan RI et  al (2008) Application of laser-capture microdissection to analysis of gene expression in the testis. Prog Histochem Cytochem 43, 173–201 22. Shukla CJ, Pennington CJ, Riddick AC et  al (2008) Laser-capture microdissection in prostate cancer research: establishment and validation of a powerful tool for the assessment of tumour-stroma interactions. BJU Int 101, 765–774 23. Harrell JC, Dye WW, Harvell DM et  al (2008) Contaminating cells alter gene signatures in whole organ versus laser capture microdissected tumors; a comparison of experimental breast cancers and their lymph node metastases. Clin Exp Metastasis 25, 81–88 24. Domazet B, MacLennan G, Lopez-Beltran A et al (2008) Laser capture microdissection in the genomic and proteomic era: targeting the genetic basis of cancer. Int J Clin Exp Pathol 1, 475–488 25. Ladanyi A, Sipos F, Szoke D et al (2006) Laser microdissection in translational and clinical research. Cytometry A 69A, 947-960 26. Bergman R (2008) Dermatopathology and molecular genetics. J Am Acad Dermatol 58, 452–457 27. What is a Dermatopathologist? The American Society of Dermatopathology. http://www. asdp.org/about/dermatopathologist.cfm. Accessed 1 March 2009 28. Boni R, Zhuang Z, Albuquerque A et  al (1998) Loss of heterozygosity detected on 1p and 9q in microdissected atypical nevi. Arch Dermatol 134, 882–883 29. Maitra A, Gasdar AF, Moore TO et al (2002) Loss of heterozygosity analysis of cutaneous melanoma and benign melanocytic nevi: laser capture microdissection demonstrates clonal genetic changes in acquired nevocellular nevi. Hum Pathol 33, 191–197 30. Hussein MR (2004) Genetic pathways to melanoma tumorigenesis. J Clin Pathol 57, 797–801 31. Dadzie OE, Yang S, Emley A et al (2009) RAS and RAF mutations in banal melanocytic aggregates contiguous with primary Cutaneous melanoma: clues to melanomagenesis. Br J Dermatol 160, 368–375 32. Woody GS (2001) Analysis of clonality in cutaneous T cell lymphoma and associated diseases. Ann NY Acad Sci 941, 26–30 33. Gallardo F, Pujol RM, Bellosillo D et al (2006) Primary cutaneous B-cell lymphoma (marginal zone) with prominent T-cell component

1  Laser Capture Microdissection: Methods and Applications and aberrant dual (T and B) genotype; diagnostic usefulness of laser-capture microdissection. Br J Dermatol 154, 162–166 34. Zhu G, Xiao H, Mohan VP et al (2003) Gene expression in the tuberculous granuloma:

15

analysis by laser capture microdissection and real-time PCR. Cell Microbiol 5, 445–453 35. Curran S, McKay JA, McLeod HL, et  al (2000). Laser capture microscopy. Mol Pathol 53, 64 –68

wwwwwwwwwwwwwww

Chapter 2 Laser Microdissection for Gene Expression Profiling Lori A. Field, Brenda Deyarmin, Craig D. Shriver, Darrell L. Ellsworth, and Rachel E. Ellsworth Abstract Microarray-based gene expression profiling is revolutionizing biomedical research by allowing expression profiles of thousands of genes to be interrogated in a single experiment. In cancer research, the use of laser microdissection (LM) to isolate RNA from tissues provides the ability to accurately identify molecular profiles from different cell types that comprise the tumor and its surrounding microenvironment. Because RNA is an unstable molecule, the quality of RNA extracted from tissues can be affected by sample preparation and processing. Thus, special protocols have been developed to isolate researchquality RNA after LM. This chapter provides detailed descriptions of protocols used to generate micro­ array data from high-quality frozen breast tissue specimens, as well as challenges associated with formalin-fixed paraffin-embedded specimens. Key words: Laser microdissection, Gene expression, Microarray, Frozen tissue, FFPE, Molecular signature, Breast cancer

1. Introduction Tumorigenesis is a complex process, involving structural changes at multiple chromosomal locations and altered expression of numerous genes and proteins. Early efforts to identify genes involved in cancer development evaluated single genes with known or putative roles in cellular processes such as growth, proliferation, angiogenesis, and apoptosis. While these efforts have resulted in the identification of more than 350 genes (1), additional genes of unknown or presumably unrelated function likely play critical roles in cancer development and progression (2). cDNA microarrays, which allow quantitative, large-scale analysis of gene expression, provide a global approach to identifying

Graeme I. Murray (ed.), Laser Capture Microdissection: Methods and Protocols, Methods in Molecular Biology, vol. 755, DOI 10.1007/978-1-61779-163-5_2, © Springer Science+Business Media, LLC 2011

17

18

L.A. Field et al.

genes involved in tumorigenesis and metastasis without a priori knowledge of the underlying molecular pathways (3). Microarrays have been used to develop molecular signatures that correlate with tumor characteristics or outcomes and are being used in clinical diagnostic tests to guide treatments for patients with breast cancer (4, 5). Despite the successful development of clinical assays and the publication of hundreds of microarray-based papers, the majority of microarray studies have used RNA isolated from tissue by homogenization or manual microdissection. Because the majority of human tumors are highly heterogeneous, with numerous cell types comprising the primary tumor and surrounding microenvironment, laser microdissection (LM) is necessary to isolate specific cells. For example, RNA isolated from laser-microdissected breast tumor cells will be free from contamination from normal epithelial, stromal, and vascular cells, which could compromise the accuracy of the resulting gene expression profiles. Because RNA is sensitive to degradation, isolation of RNA after LM requires a defined protocol that includes careful cleaning of all equipment with RNase inhibitors, special histological stains, and rapidity (less than 30 min) in cutting, mounting, and microdissecting the tissues. In this chapter, we present protocols for performing microarray analysis using RNA isolated after LM and describe alternate protocols for gene expression analysis of formalin-fixed paraffin-embedded (FFPE) archival specimens.

2. Materials 2.1. Tissue Sectioning, Staining, and Laser Microdissection

1. Membrane-based laser microdissection slides (W. Nuhsbaum, McHenry, IL). 2. Disposable microtome blades, HP35n, noncoated (Thermo Fisher Scientific, Pittsburgh, PA). 3. 0.5 ml PCR tubes (Eppendorf, Hauppauge, NY). 4. RNaseZap® (Applied Biosystems, Carlsbad, CA). 5. Nuclease-free water (Applied Biosystems). 6. LCM Staining Kit (Applied Biosystems) – store cresyl violet at 4°C. 7. 50% ethanol. 8. 75% ethanol. 9. 95% ethanol. 10. 100% ethanol. 11. Xylene (used only for FFPE samples).

2  Laser Microdissection for Gene Expression Profiling

19

12. Tissue-Tek® Cryomold® Standard, 25 × 20 × 5 mm (Electron Microscopy Sciences, Hatfield, PA). 13. Cryomatrix optimal cutting temperature (OCT) compound (Thermo Fisher Scientific). 2.2. RNA Isolation from Frozen Tissue

1. RNAqueous®-Micro kit (Applied Biosystems). 2. Nuclease-free water. 3. 100% ethanol. 4. Agilent RNA 6000 Pico kit (Agilent Technologies, Santa Clara, CA). 5. Agilent 2100 Bioanalyzer (Agilent Technologies). 6. RNaseZap®.

2.3. Amplification and Fragmentation of RNA from Frozen Tissue

1. MessageAmp™ Biosystems).

II

aRNA

Amplification

kit

(Applied

2. GeneChip® Eukaryotic Poly-A RNA Control kit (Affymetrix, Santa Clara, CA). 3. 75 mM Bio-11-UTP (Applied Biosystems). 4. Nuclease-free water. 5. 5× Fragmentation buffer, component of the GeneChip® Sample Cleanup Module (Affymetrix). 6. Agilent RNA 6000 Pico kit. 7. Agilent RNA 6000 Nano kit (Agilent Technologies, Santa Clara, CA). 8. Agilent 2100 Bioanalyzer. 9. NanoDrop ND-1000 Spectrophotometer (Thermo Fisher Scientific) – Note: the current model is the NanoDrop 2000.

2.4. Hybridization of aRNA to Microarrays

1. GeneChip® Expression 3¢ Amplification reagents containing 20× Eukaryotic Hybridization Controls and Control Oligo­ nucleotide B2 (Affymetrix). 2. Herring Sperm DNA (Promega, Madison, WI). 3. Bovine serum albumin (BSA) (Invitrogen, Carlsbad, CA). 4. MES hydrate (Sigma-Aldrich, St Louis, MO). 5. MES sodium salt (Sigma-Aldrich). 6. 5 M NaCl (Sigma-Aldrich). 7. 0.5 M EDTA (Sigma-Aldrich). 8. Tween 20 (Promega). 9. DMSO (Sigma-Aldrich). 10. Nuclease-free water.

20

L.A. Field et al.

11. GeneChip® Human Genome U133A 2.0 Arrays (HG U133A 2.0) (Affymetrix). 12. Hybridization oven (Affymetrix). 2.5. Washing, Staining, and Scanning Microarrays

1. Bovine serum albumin. 2. Streptavidin phycoerythrin (SAPE) (Invitrogen). 3. Goat IgG (Sigma-Aldrich). 4. Biotinylated antistreptavidin (Vector Laboratories, Burlingame, CA). 5. 20× SSPE (Sigma-Aldrich). 6. 5 M NaCl. 7. Tween 20. 8. Nuclease-free water. 9. Tough Spots (T-SPOTS; Diversified Biotech, Boston, MA). 10. Fluidics Station (Affymetrix). 11. Scanner (Affymetrix).

2.6. Quantitative Real-Time Polymerase Chain Reaction of RNA from Frozen Tissue or FFPE

1. High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). 2. TaqMan® Universal PCR Master Mix (Applied Biosystems). 3. TaqMan® Gene Expression Assays (Applied Biosystems). 4. FirstChoice® Human Brain Reference RNA (Applied Biosystems). 5. iCycler iQ™ PCR plates (Bio-Rad Laboratories, Hercules, CA). 6. iCycler iQ™ thermal seals (Bio-Rad Laboratories). 7. iCycler iQ™ real-time PCR detection system (Bio-Rad Laboratories).

2.7. RNA Isolation from Formalin-Fixed Paraffin Embedded Specimens

1. RecoverAll™ total nucleic acid isolation kit (Applied Biosystems).

2.8. Commercial Vendor Information

1. Affymetrix – http://www.affymetrix.com.

2. 100% ethanol.

2. Agilent Technologies – http://www.agilent.com. 3. Applied Biosystems – http://www.appliedbiosystems.com. 4. Bio-Rad Laboratories – http://www.bio-rad.com. 5. Diversified Biotech – http://divbio.com/. 6. Electron Microscopy Sciences – http://emsdiasum.com/ microscopy/. 7. Eppendorf – http://www.eppendorf.com.

2  Laser Microdissection for Gene Expression Profiling

21

8. Invitrogen – http://www.invitrogen.com. 9. Promega – http://www.promega.com. 10. Sigma-Aldrich – http://www.sigmaaldrich.com. 11. Thermo Fisher Scientific – http://www.thermofisher.com. 12. Vector Laboratories – http://vectorlabs.com/. 13. W. Nuhsbaum – http://www.nuhsbaum.com/.

3. Methods RNA is extremely susceptible to degradation by RNase enzymes in the environment. To generate high-quality microarray or quantitative real-time polymerase chain reaction (qRT-PCR) data, it is critical to obtain RNA of the highest possible quality by preventing RNase contamination during tissue collection and processing, RNA isolation, and downstream applications. Several general precautions should be taken when working with RNA in the laboratory. All equipment and laboratory benches should be thoroughly cleaned with RNaseZap® and then rinsed with nuclease-free or deionized water. All pipette tips, tubes, reagents, and other consumables must be RNase-free. Pipette tips should contain barriers and should be changed each time you pipette, even if you are pipetting the same reagent, to avoid potential cross-contamination between samples and to prevent RNase contamination. For most procedures, it is advisable to use nuclease-free, hydrophobic, nonstick tubes to minimize loss of sample that may otherwise adhere to the tube walls. Gloves should be worn at all times and changed frequently, especially after coming into contact with liquids or surfaces that may be contaminated with RNases. 3.1. Sectioning and Staining

3.1.1. Frozen Tissue

To prevent RNA degradation, tissue sectioning, staining, and LM must be performed as quickly as possible (typically within 30 min). In our laboratory, two individuals perform these steps and process one slide at a time. The LCM Staining Kit employs a novel staining procedure that avoids exposing the tissue sections to pure water at any step, thus minimizing the potential for RNA degradation. 1. In the bottles provided with the LCM Staining Kit, prepare 95, 75, and 50% ethanol solutions by diluting 100% ethanol with nuclease-free water. Add the dehydration beads to the bottle labeled 100% ethanol and add absolute ethanol. Do not use the ethanol in this container to make any of the diluted solutions.

22

L.A. Field et al.

2. Clean the staining containers included in the LCM Staining Kit with RNaseZap®. For FFPE samples, a glass staining dish should also be cleaned. Spray the containers generously with RNaseZap® and allow them to sit for 10  min. Rinse twice with distilled water and then perform a final rinse with nuclease-free water. Allow the containers to dry under a hood and then fill with the appropriate solutions. 3. Set the temperature of the cryostat to −30°C. 4. Clean the knife holder (not the knife blade itself) with 100% ethanol and treat the brushes that will be used to manipulate the tissue sections with RNaseZap®. 5. Cool the specimen and brushes in the cryostat. 6. Inside the cryostat, remove the frozen OCT-embedded tissue from its cryomold and mount securely to the metal specimen stage with OCT compound, orienting the tissue according to regions of interest (see Note 1). 7. Using a fresh disposable blade, shave OCT from the block until the tissue becomes visible. Set the cutting thickness to 8 mm. 8. Section the tissue and use a small brush to straighten out the newly cut sections. 9. Manipulate sections onto the foil slides (see Note 2). Perform staining under a hood used only for RNA procedures. Change all containers and blade surfaces between each patient sample. 10. Cut sections at 8  mm and mount onto a membrane-based laser microdissection slide. 11. Wash slide in 95% ethanol for 30 s. 12. Wash in 75% ethanol for 30 s. 13. Wash in 50% ethanol for 30 s (see Note 3). 14. Pipette cresyl violet (~50  ml) onto the slide to completely cover the tissue sections; allow the slide to sit for 15 s. 15. Rinse in 95% ethanol for 5 s. 16. Rinse in 100% ethanol for 5 s. 17. Rinse in a second container of 100% ethanol for 30  s (see Note 4). 18. Allow slide to air dry. 3.1.2. Formalin-Fixed Paraffin-Embedded Tissue

1. Fill the clean staining dish with nuclease-free water and warm on a hot plate to the desired temperature for the paraffin being used (typically 37–42°C). Change the water bath between each sample.

2  Laser Microdissection for Gene Expression Profiling

23

2. Cut sections at 8 mm and lay out ribbon onto the warm water bath. 3. Mount sections onto a membrane-based LM slide. 4. Place slides in an incubator set at 56°C for 15  min (see Note 5). 5. Wash in xylene for 1  min; repeat twice for a total of three washes. 6. Wash in 95% ethanol for 30 s. 7. Wash in 75% ethanol for 30 s. 8. Wash in 50% ethanol for 30 s. 9. Pipette Cresyl Violet stain onto the slide using enough volume to cover the sections; allow to sit for 15 s. 10. Rinse in 95% ethanol for 5 s. 11. Rinse in 100% ethanol for 5 s. 12. Rinse in 100% ethanol for 30 s. 3.2. Laser Microdissection

1. Use a cover-slipped H&E section to orient the tissue for microdissection. Estimate the number of cells – in our experience, ~10,000 cells usually yields sufficient RNA for downstream applications. 2. Locate the area on the cresyl violet-stained section to be microdissected (see Note 6). 3. Pipette 60 ml of Lysis solution for OCT-embedded tissues, or 60  ml of digestion buffer for FFPE-embedded tissues, into the cap of a clean 0.5 ml Eppendorf tube. Place the cap into the cap holder apparatus of the laser microdissection system. 4. Microdissect the area of interest (Fig. 1) and drop the sample into the buffer (see Note 7). 5. Add the remaining 40 ml of Lysis solution (OCT tissues) or 340  ml of digestion buffer (FFPE tissues) to the tube and carefully close the lid.

3.3. RNA Isolation from Frozen Tissue

1. Before first use, add 10.5 ml of 100% ethanol to Wash solution 1 and 22.4  ml of 100% ethanol to Wash solution 2/3 and mix well (see Note 8). 2. On first use, thaw the Pico Ladder on ice, centrifuge briefly, and transfer to an RNase-free tube. Heat-denature the ladder for 2 min at 70°C in a heat block, then immediately place on ice. Add 90 ml of nuclease-free water, pipette up and down several times, and flick the tube to mix. Briefly centrifuge the tube and aliquot 5–10 ml to RNase-free tubes. Store at −70°C (see Note 9).

24

L.A. Field et al.

Fig. 1. Staining and laser microdissection of formalin-fixed paraffin-embedded breast tissue containing a ductal carcinoma in situ (DCIS). (a) Standard hematoxylin and eosin (H&E) stain of the DCIS on a glass slide. (b) DCIS on a foil slide stained with Cresyl Violet. (c) Breast tissue after removal of the DCIS by laser microdissection.

3. Place a tube containing nuclease-free water (at least 50 ml per sample) in a heat block at 95°C. 4. Prewarm an air incubator to 42°C. 5. Thaw LCM Additive and 10× DNase I buffer on ice. 6. Flick the tube containing the microdissected sample several times and centrifuge briefly. Place the sample in the 42°C incubator for 30 min (see Note 10). 7. Approximately 6–7  min prior to completion of the 30-min incubation, prewet the Micro Filter by adding 30 ml of Lysis solution to the filter, which is placed in a Micro-Elution Tube (Micro Filter Cartridge Assembly). After 5  min, centrifuge the Micro Filter Cartridge Assembly for 30 s at 16,000 rcf to remove the Lysis solution from the filter. 8. Remove the microdissected sample from the 42°C incubator, vortex on maximum speed by pulsing three times, and centrifuge briefly. Add 3  ml of LCM Additive, mix by vortexing, and centrifuge briefly. 9. Add 52 ml of 100% ethanol and mix completely into the sample by pipetting up and down (see Note 11). Transfer the sample to the center of the filter in the Micro Filter Cartridge Assembly. Centrifuge for 1 min at 10,000 rcf (see Note 12). 10. Add 180 ml of Wash solution 1 to the filter and centrifuge for 1 min at 10,000 rcf. 11. Add 180 ml of Wash solution 2/3 and centrifuge for 30 s at 16,000 rcf. Repeat this step one time. 12. Remove the filter from the collection tube and discard the flow-through. Recap the assembly and centrifuge for 1 min to remove trace amounts of liquid. 13. Remove the filter containing the sample and place in a new Micro Elution Tube. 14. Add 10  ml of nuclease-free water heated to 95°C in step 1 above to the center of the filter (see Note 13). Incubate the

2  Laser Microdissection for Gene Expression Profiling

25

assembly for 5 min at room temperature, then centrifuge for 1 min at 16,000 rcf to elute the RNA. Repeat this step with a second 10 ml volume of 95°C nuclease-free water, incubate, and centrifuge. 15. Remove the filter and place the sample on ice. 16. Add 2 ml of 10× DNase I buffer and 1 ml of DNase I to the sample and mix by gently flicking the tube. Centrifuge briefly and incubate for 20 min in a heat block at 37°C. During the incubation, remove the DNase Inactivation Reagent from the freezer and thaw at room temperature. 17. Remove the sample from the heat block. Vigorously vortex the DNase Inactivation Reagent and add 2.3 ml to the sample. Gently tap the side of the tube to mix and incubate for 2 min at room temperature. After 1 min, vortex the sample, tap the tube to move all contents to the bottom, and continue the incubation for 1 min. 18. Centrifuge the sample for 1 min 30 s at 16,000 rcf to pellet the DNase Inactivation Reagent. Transfer the supernatant containing the RNA to a new tube without disturbing the pellet, then place the RNA on ice (see Note 14). 3.4. Assessing RNA Integrity

1. Remove an aliquot of the Pico Ladder from the freezer and thaw on ice. Remove the Pico Gel Matrix, Pico Dye Concentrate, Pico Conditioning Solution, and Pico Marker from 4°C and allow the reagents to warm to room temperature for at least 30  min. Ensure that the Dye Concentrate is shielded from light (see Note 15). 2. Add 550 ml of Gel Matrix to a Spin Filter and centrifuge for 10  min at 1,500 rcf. Aliquot 65  ml of filtered gel into the tubes provided with the kit (produces seven to eight tubes of filtered gel). The filtered gel may be stored at 4°C for up to 2 months. 3. Vortex the tube of Dye Concentrate for 10 s and then centrifuge briefly. Add 1 ml of Dye Concentrate to a tube of filtered gel (warmed to room temperature), vortex for 10 s, then centrifuge for 10 min at 16,000 rcf. One tube of gel-dye mix can be used to run two chips per day. 4. Transfer 1.25–1.5  ml of each RNA sample into a 0.65-ml tube. Heat the sample for 2 min in a heat block at 65–70°C. Place on ice for ~5 min to cool, then centrifuge briefly to collect the RNA at the bottom of the tube. 5. Start the 2100 Expert Software and turn on the Bioanalyzer. Place an electrode cleaner containing 350 ml of nuclease-free water in the instrument and close the lid (see Note 16). On the instrument menu, select “Assays,” “Electrophoresis,”

26

L.A. Field et al.

“RNA,” and finally “Eukaryotic Total RNA Pico Series.” Select the number of samples (from 1 to 11) to be assayed. Enter the sample information and any additional comments pertaining to that sample. 6. Place the Pico Chip on the chip priming station (ensuring that the base plate is on “C”) and pull the syringe back to 1  ml. Add 9  ml of gel-dye mix to the well labeled with an encircled “G” (see Note 17). Close the chip priming station until you hear a click, then press the syringe down until it is secured beneath the syringe clip. After 30 s, release the clip, wait 5 s, and pull the syringe back to the 1 ml mark. 7. Add 9 ml of gel-dye mix to the two remaining wells marked “G.” Add 9  ml of Pico Conditioning Solution to the well marked “CS.” Add 5 ml of Pico Marker to the ladder well and to each well that will contain an RNA sample. Add 6 ml of Pico Marker to any empty sample wells. 8. Add 1 ml of diluted Pico Ladder to the ladder well and 1 ml of sample to the appropriate sample well. After loading all wells, vortex the chip using the manufacturer-supplied vortex for 1 min at 2,400 rpm. During this time, remove the electrode cleaner from the instrument. Place the Pico Chip on the Agilent 2100 Bioanalyzer and begin the run by pressing “Start” (see Notes 18 and 19) (Fig. 2). 3.5. Amplification of RNA from Frozen Tissue

When using small amounts of RNA for gene expression analysis, it is often necessary to first amplify the RNA to generate sufficient material for hybridization to the microarray. For RNA isolated

Fig. 2. Electropherogram of total RNA isolated from frozen breast tissue collected via laser microdissection using the RNAqueous®-Micro kit. The RNA (RIN = 8.7) was assayed on the Bioanalyzer using a Pico Chip. The 18S rRNA and 28S rRNA peaks are visible near 2,000 and 4,000 nucleotides (nt), respectively.

2  Laser Microdissection for Gene Expression Profiling

27

from laser-microdissected tissues, two rounds of amplification are normally required. All frozen reagents for the amplification protocol should be thawed on ice; enzymes should be stored at −20°C immediately prior to and after use. All master mixes should be prepared in excess (generally ~5%) to avoid running short of master mix when working with large numbers of samples. 1. Completely thaw the Poly-A Control Stock on ice, then add 2 ml to a small tube. Add 38 ml of nuclease-free water and mix well by vortexing or flicking the tube. Centrifuge briefly to collect the liquid at the bottom of the tube. This is the first dilution and can be stored at −80°C for up to 6  weeks (or eight freeze–thaw cycles). 2. Remove 2 ml of the first dilution and place in a new tube. Add 98 ml of nuclease-free water to make the second dilution. Mix well and centrifuge briefly. 3. Combine 2 ml of the second dilution with 98 ml of nucleasefree water to make the third dilution. Mix well and centrifuge. 4. Combine 2 ml of the third dilution with 18 ml of nuclease-free water to prepare the fourth dilution. Mix well and centrifuge. 5. Combine 2 ml of the fourth dilution with 18 ml of nucleasefree water to prepare the fifth dilution. Mix well and centrifuge (see Note 20). 6. Using the estimated RNA concentration obtained from the Bioanalyzer, calculate the volume of sample containing 10 ng of RNA (see Note 21). Transfer this volume to a 0.2 ml PCR tube and adjust the total volume to 9 ml with nuclease-free water. If the volume needed for 10 ng of RNA is greater than 9 ml, transfer this amount to a hydrophobic, nonstick microcentrifuge tube, and centrifuge in a vacuum concentrator until the volume is £9 ml. Transfer the concentrated sample to a 0.2-ml PCR tube and adjust the volume to 9 ml with nuclease-free water. 7. Flick the tubes to mix and centrifuge briefly to collect the liquid at the bottom of the tube. 3.6. First Round Amplification

1. Add 2 ml of the fifth dilution of the Poly-A Controls to each sample containing 10  ng of RNA (see Note 22). Flick the tubes to mix and centrifuge briefly. 2. Add 1 ml of Oligo(dT) primer to each sample, flick the tubes to mix, and centrifuge briefly. Incubate samples for 10 min at 70°C in a thermal cycler. 3. Remove samples from the thermal cycler, centrifuge briefly, and place on ice.

28

L.A. Field et al.

4. In a small tube, prepare a master mix containing the following for each sample: (a) 2 ml 10× first strand buffer. (b) 4 ml dNTP mix. (c) 1 ml RNase inhibitor. (d) 1 ml ArrayScript™. Vortex the tube to mix and centrifuge briefly to collect the contents at the bottom of the tube. Add 8 ml of the master mix to each sample, flick the tubes to mix, and centrifuge. Incubate samples for 2  h at 42°C in an air incubator or hybridization oven, then centrifuge briefly, and place on ice. 5. Prepare a master mix on ice containing the following reagents for each sample: (a) 63 ml nuclease-free water. (b) 10 ml 10× second strand buffer. (c) 4 ml dNTP mix. (d) 2 ml DNA polymerase. (e) 1 ml of RNase H. Vortex to mix and centrifuge briefly to collect the master mix at the bottom of the tube. Add 80 ml of master mix to each sample, flick the samples to mix, and centrifuge briefly. Incubate the samples in a precooled thermal cycler for 2 h at 16°C (see Note 23), then centrifuge briefly and place on ice. 6. Place a tube containing at least 30 ml of nuclease-free water per sample in a heat block set to 50–55 ° C. For each sample, place a filter inside a cDNA Elution tube. Note: add 24 ml of 100% ethanol to the Wash buffer before using for the first time. 7. Transfer the samples from the 0.2 ml tubes to 1.5 ml microcentrifuge tubes. Add 250  ml of cDNA Binding buffer to each sample, mix by pipetting up and down and then flicking the tubes several times. Centrifuge samples briefly, then transfer each sample to the filter of a cDNA Filter Cartridge. Centrifuge samples for 1 min at 10,000 rcf, then discard the flow-through. 8. Add 500 ml of Wash buffer to each filter. Centrifuge for 1 min at 10,000 rcf and discard the flow-through. 9. Centrifuge the cDNA Filter Cartridges for 1 min at 10,000 rcf to remove any residual liquid from the filter. Transfer filters to new cDNA Elution tubes and discard the old tubes. 10. Add 10 ml of nuclease-free water warmed to 50–55°C to the center of each filter. Incubate for 2 min at room temperature. Elute samples by centrifuging for 1 min 30 s at 10,000 rcf. Repeat this step using a second 10 ml volume of warm nuclease-free water.

2  Laser Microdissection for Gene Expression Profiling

29

11. Discard filters and place tubes containing the eluted cDNA on ice. 12. Prepare the in vitro transcription (IVT) master mix at room temperature. Note that for the first round of amplification, the IVT reactions contain only unmodified dNTPs. For each sample include: (a) 4 ml T7 ATP. (b) 4 ml T7 CTP. (c) 4 ml T7 GTP. (d) 4 ml T7 UTP. (e) 4 ml T7 10× reaction buffer. (f) 4 ml T7 enzyme mix. Vortex the master mix and centrifuge briefly to collect the contents at the bottom of the tube. Aliquot 24 ml of master mix to each sample, flick the tubes to mix, and centrifuge briefly. Incubate samples for 14 h in an air incubator or hybridization oven at 37°C. 13. Place a tube containing nuclease-free water in a heat block at 50–55°C – we recommend heating at least 120 ml of nucleasefree water per sample. 14. For each sample, place an aRNA Filter Cartridge in an aRNA Collection Tube. 15. Remove the IVT reactions from the incubator. Add 60 ml of nuclease-free water to each sample, mix by flicking the tube, and centrifuge briefly. Add 350 ml of aRNA Binding buffer followed by 250 ml of 100% ethanol to each sample. Mix the samples by pipetting up and down at least five times, then transfer each sample to an aRNA Filter Cartridge. Centrifuge samples for 1 min at 10,000 rcf, then discard the flow-through and remount the filter on the collection tube. 16. Add 650 ml of wash buffer to each Filter Cartridge and centrifuge for 1 min at 10,000 rcf. Discard the flow-through and place the Filter Cartridge back inside the collection tube. Centrifuge samples for an additional 1 min at 10,000 rcf to remove residual wash buffer. Discard the flow-through and place the Filter Cartridge in a new collection tube. 17. Apply 100 ml of nuclease-free water warmed to 50–55°C to the center of each filter. Incubate at room temperature for 2 min, then centrifuge for 1 min at 10,000 rcf to elute the aRNA. 18. Remove 3 ml of the aRNA and transfer to a small tube. Heat the samples for 2 min in a heat block at 65–70°C. Place samples on ice to cool, then centrifuge the samples briefly, and return to ice.

30

L.A. Field et al.

Fig. 3. Electropherograms of RNA isolated from frozen breast tissue following one and two rounds of amplification. (a) Total RNA amplified using the MessageAmp™ II aRNA Amplification kit and assayed on the Bioanalyzer using a Pico Chip. (b) Second round aRNA assayed using a Nano Chip. The majority of the second round aRNA product should be >500 nucleotides (nt) in length.

19. Run 1 ml of the first round aRNA samples from step 18 on the Bioanalyzer using a Pico Chip following the instructions outlined above (Fig. 3a). Use 1.5 ml of the remaining aRNA to measure the concentration of each sample on a NanoDrop ND-1000 Spectrophotometer. 3.7. Second Round Amplification

1. Calculate the volume of first round aRNA (using the concentration obtained on the NanoDrop) needed to obtain 1 mg of starting material for the second round of amplification

2  Laser Microdissection for Gene Expression Profiling

31

(see Note 24). If this volume exceeds 10 ml for any sample, concentrate those samples in a vacuum concentrator to less than 10 ml. In a 0.2-ml PCR tube, adjust the volume of all samples to 10 ml using nuclease-free water. 2. Add 2  ml of second round primers to each aRNA sample. Flick the tubes to mix and centrifuge briefly. Place samples in a thermal cycler heated to 70°C for 10 min, then centrifuge briefly and place on ice. 3. Prepare a master mix containing the following for each sample: (a) 2 ml 10× first strand buffer. (b) 4 ml dNTP mix. (c) 1 ml RNase inhibitor. (d) 1 ml ArrayScript™. Vortex the master mix and centrifuge briefly. Add 8  ml of master mix to each sample and flick the tubes to mix. Centrifuge briefly and incubate for 2  h at 42°C in an air incubator or hybridization oven. 4. Following incubation, centrifuge the samples briefly and place on ice. Add 1 ml of RNase H to each sample, flick the tubes to mix, and centrifuge briefly to collect the contents at the bottom of the tube. Incubate samples for 30 min at 37°C in an air incubator or hybridization oven, then centrifuge briefly and place on ice. 5. Add 5 ml of the Oligo(dT) primer to each sample, flick the tubes to mix, and centrifuge briefly. Incubate samples for 10 min at 70°C in a thermal cycler, then centrifuge and place on ice. 6. Prepare a master mix on ice for the second strand synthesis that includes the following for each sample: (a) 58 ml nuclease-free water. (b) 10 ml 10× second strand buffer. (c) 4 ml dNTP mix. (d) 2 ml DNA polymerase. Vortex to mix and add 74 ml to each sample. Flick the tubes to mix and centrifuge briefly. Incubate samples for 2 h in a thermal cycler that has been precooled to 16°C. Remove samples from the thermal cycler, centrifuge briefly and place on ice. 7. Purify the cDNA following the exact procedure outlined above.

32

L.A. Field et al.

8. Prepare a master mix at room temperature that contains for each sample: (a) 4 ml T7 ATP. (b) 4 ml T7 CTP. (c) 4 ml T7 GTP. (d) 2.6 ml T7 UTP. (e) 1.4 ml biotin-11-UTP. (f) 4 ml T7 reaction buffer. (g) 4 ml T7 enzyme mix. Vortex and centrifuge briefly to collect the contents at the bottom of the tube. Add 24 ml of master mix to each sample, flick the tubes to mix, and centrifuge briefly. Incubate the samples for 14 h at 37°C in an air incubator or hybridization oven. 9. Purify the second round, labeled aRNA using the same procedure for aRNA purification outlined above. 10. Run the second round, labeled aRNA on the Bioanalyzer using the Agilent RNA 6000 Nano kit. Reagent and sample preparation for the Nano Chip is very similar to that for the Pico Chip with minor exceptions. Warm all refrigerated Nano reagents to room temperature. Prepare the filtered gel and gel-dye mix using the Nano Gel Matrix and Nano Dye Concentrate as outlined above. 11. Thaw the Nano Ladder on ice. Flick the tube several times and centrifuge briefly. Transfer 2.5 ml of the Nano Ladder to a new tube. Prepare 5–10 ml aliquots of the remaining ladder and store at −20°C for future use. Place 3 ml of each aRNA sample in a small tube. Heat the samples and ladder for 2 min in a heat block at 65–70°C. Place the samples on ice to cool, then centrifuge briefly to collect any condensation at the bottom of the tube. 12. Start the 2100 Expert Software and clean the electrodes as previously described. Select the “Eukaryotic Total RNA Nano Series Assay” and select the number of samples (from 1 to 12) that will be run. Enter the sample information. 13. Load the Nano Chip using the same procedure for loading the Pico Chip, but note that the Nano Chip does not use Conditioning Solution, allowing 12 samples to be run (Fig. 3b). 14. Measure the concentration of 1.5 ml of the remaining second round aRNA on the NanoDrop. 3.8. Fragmentation of Labeled aRNA

1. For each sample, transfer 15 mg of labeled aRNA to a 0.2-ml PCR tube. The aRNA yield after two rounds of amplification typically exceeds 1,500  ng/ml and often is ~2,000  ng/ml; therefore, you should not have to vacuum-concentrate the samples prior to fragmentation. Add nuclease-free water to

2  Laser Microdissection for Gene Expression Profiling

33

Fig. 4. Electropherogram of fragmented, second round aRNA assayed on the Bioanalyzer using a Nano Chip. The aRNA fragments should be ~35 to 200 nucleotides (nt) in length.

each sample to bring the total volume to 24 ml. Flick the tubes to mix and centrifuge briefly. 2. Add 6 ml of 5× Fragmentation buffer to each sample, flick the tubes to mix, and briefly centrifuge. Fragment the aRNA samples by incubating for 35 min at 94°C in a thermal cycler. Place the samples on ice to cool, then centrifuge briefly to collect condensation at the bottom of the tube. 3. Run the fragmented samples on a Bioanalyzer Nano Chip (see Note 25) (Fig. 4). 3.9. Hybridization of Fragmented aRNA to Microarrays

Microarrays for gene expression analysis are available from several commercial vendors, including Affymetrix, Agilent Technologies, and Illumina. In this chapter, we provide protocols for hybridization of labeled, fragmented aRNA to the Affymetrix HG U133A 2.0 arrays. The protocol is the same for all eukaryotic arrays manufactured by Affymetrix; however, the starting amount of fragmented aRNA and volume of reagents used in certain steps will vary across the different array formats. 1. Thaw the 20× Eukaryotic Hybridization Controls on ice. Mix well by flicking the tube or vortexing gently and then centrifuge briefly to collect the contents at the bottom of the tube. Prepare aliquots containing the volume used during a typical hybridization set-up (we usually hybridize eight samples at a time) and store at −20°C. 2. Prepare 1 l of 12× MES stock buffer by dissolving 64.61 g of MES hydrate and 193.3 g of MES sodium salt in 800 ml of nuclease-free water. When completely dissolved, adjust

34

L.A. Field et al.

the volume to 1,000  ml with nuclease-free water and filter through a 0.22-mm filter. The pH should be between 6.5 and 6.7. Store at 4°C wrapped in foil to shield from light. Discard the solution if it turns yellow. 3. Prepare 50 ml of 2× hybridization buffer by combining: (a) 8.3 ml 12× MES stock buffer. (b) 17.7 ml 5 M NaCl. (c) 4.0 ml 0.5 M EDTA. (d) 19.9 ml nuclease-free water. (e) 100 ml 10% Tween 20. Filter the solution through a 0.22-mm filter and store wrapped in foil at 4°C. 4. Approximately 1  h before setting up the hybridization, remove the HG U133A 2.0 arrays from the refrigerator and allow them to equilibrate to room temperature. Thaw the Control Oligonucleotide B2, 20× Eukaryotic Hybridization Controls, Herring Sperm DNA, and BSA on ice. Set the temperature of the hybridization oven to 45°C. 5. Transfer 20  ml (10  mg) of fragmented aRNA to a 1.5-ml microcentrifuge tube. 6. Prepare a 1× solution of hybridization buffer by combining equal volumes of 2× hybridization buffer and nuclease-free water; vortex to mix well. 7. Heat the 20× Eukaryotic Hybridization Controls for 5 min at 65°C in a heat block before adding to the hybridization master mix. 8. Prepare the hybridization master mix (at 5% excess) by combining for each sample: (a) 3.3 ml Control Oligonucleotide B2. (b) 10 ml 20× Eukaryotic Hybridization Controls. (c) 2 ml Herring Sperm DNA. (d) 2 ml BSA. (e) 100 ml 2× hybridization buffer. (f) 20 ml DMSO. (g) 42.7 ml nuclease-free water. Vortex to mix well, centrifuge briefly, and add 180 ml of the master mix to each fragmented sample (from step 2 above). Vortex the samples, centrifuge briefly, and incubate for 5 min at 99°C in a heat block. 9. Place the arrays face down on a laboratory tissue. Insert a pipette tip into the top right septum on the back of the array to permit air to vent when filling the array. Fill the arrays with 160 ml of 1× hybridization buffer, remove the pipette tip vent,

2  Laser Microdissection for Gene Expression Profiling

35

and incubate the arrays for 10 min at 60 rpm in the hybridization oven set to 45°C. 10. Transfer the samples from 99°C to a 45°C heat block. Incubate for 5 min, then centrifuge the samples for 5 min at 16,000 rcf. 11. Remove the arrays from the hybridization oven. Insert the pipette tip vent and remove the hybridization buffer. Load 130 ml of the sample into the array, making sure that a bubble is present that can freely move when the array is slowly tilted from side to side. Incubate the arrays for 16 h at 45°C in the hybridization oven while rotating at 60 rpm (see Notes 26 and 27). 3.10. Washing, Staining, and Scanning of Microarrays

1. Prepare 250 ml of 2× stain buffer by combining: (a) 41.7 ml 12× MES stock buffer. (b) 92.5 ml 5 M NaCl. (c) 2.5 ml 10% Tween 20. (d) 113.3 ml nuclease-free water. Filter through a 0.22-mm filter, wrap in foil to shield from light, and store at 4°C. 2. Reconstitute the goat IgG by adding 150 mM NaCl to make a 10  mg/ml solution. For example, add 1  ml of 150  mM NaCl to 10 mg of lyophilized goat IgG. Aliquots should be stored at −20°C, but once thawed for use, store at 4°C. 3. Add 1 ml of nuclease-free water to reconstitute the biotinylated antistreptavidin antibody to 0.5 mg/ml. Gently pipette up and down to mix, then store at 4°C. 4. Prepare 1 l of Wash A solution by combining: (a) 300 ml 20× SSPE. (b) 699 ml deionized water. (c) 1 ml 10% Tween 20. Filter through a 0.22-mm filter and store at room temperature. 5. Prepare 1 l of Wash B solution by combining: (a) 83.3 ml 12× MES stock buffer. (b) 5.2 ml 5 M NaCl. (c) 1 ml 10% Tween 20. (d) 910.5 ml deionized water. Filter through a 0.22-mm filter, cover with foil, and store at 4°C. 6. Prepare a master mix (in 5% excess) of the SAPE Solution (stains 1 and 3) by combining for each sample: (a) 600 ml 2× stain buffer. (b) 48 ml BSA. (c) 12 ml SAPE (vortex well before pipetting).

36

L.A. Field et al.

(d) 540 ml of deionized water. Vortex to mix. For each sample, aliquot 600 ml of the SAPE Solution into two 1.5 ml microcentrifuge tubes. 7. Prepare a master mix for the Antibody Solution (Stain 2) in 5% excess. For each sample, combine: (a) 300 ml 2× stain buffer. (b) 24 ml BSA. (c) 6 ml IgG. (d) 3.6 ml biotinylated antistreptavidin. (e) 266.4 ml deionized water. Mix well by vortexing and aliquot 600 ml of the Antibody Solution master mix into a 1.5-ml microcentrifuge tube for each sample. 8. After 16  h, remove the arrays from the hybridization oven. Insert the pipette tip vent and remove the sample solution from the arrays. Fill the arrays with 160 ml of Wash A solution. 9. Start the GeneChip® Operating System (GCOS) or Affymetrix® GeneChip® Command Console® (AGCC) and enter the sample and experiment information. 10. Replace the appropriate water bottles on the fluidics station with Wash A solution and Wash B solution. Prime all modules on the fluidics station. 11. Select the appropriate wash and stain protocol for your arrays and press “Run.” Load the arrays and staining tubes onto the fluidics station. The protocol will automatically run until staining of the arrays is complete (usually ~1 h and 15 min). 12. Remove the arrays from the fluidics station. Place Tough Spots over the septa on the back of the arrays to prevent leakage during scanning. Once the scanner has warmed up (10– 15 min), scan the arrays using the GCOS or AGCC software. When scanning is complete, zoom in on each scanned image and check the entire image for any abnormalities. The scanned images can now be analyzed to generate signal intensities for the probes on the arrays as well as a QC report. 13. Once the wash and stain protocol is complete, replace the Wash A solution and Wash B solution with deionized water. Run the Shutdown protocol on all modules, and when finished, turn off the fluidics station. 3.11. Quantitative Real-Time Polymerase Chain Reaction Assays Using RNA from Frozen Tissue

Gene expression differences identified by microarray analysis are frequently confirmed using qRT-PCR. Due to the low yield of RNA following laser microdissection of small tissue sections, RNA is amplified for a single round prior to performing qRTPCR. To amplify the RNA, follow the procedure detailed above; however, it is not necessary to add the Poly-A controls.

2  Laser Microdissection for Gene Expression Profiling

37

Note that there are several different chemistries available for performing qRT-PCR, including TaqMan® probes, Molecular Beacons, Scorpions®, or SYBR® Green. The following protocol uses TaqMan® Gene Expression Assays, which utilize the same PCR conditions so there is no need to design PCR primers or optimize PCR conditions. In order to achieve accurate results from qRT-PCR, be careful and precise when performing each step in the protocol, especially pipetting. 1. Reverse transcribe the aRNA (one round of amplification) using the High-Capacity cDNA Reverse Transcription Kit. Up to 2 mg of RNA (or aRNA) can be reverse transcribed (see Note 28). In addition, reverse transcribe an appropriate amount of FirstChoice® Human Brain Reference RNA or other reference RNA which will be used to calibrate the relative levels of gene expression in the samples of interest. 2. If necessary, vacuum concentrate the samples to £10 ml and increase the volume to 10  ml with nuclease-free water in a 0.2-ml PCR tube. 3. Prepare the reverse transcription master mix (in 5% excess) containing the following for each sample: (a) 2 ml 10× RT buffer. (b) 0.8 ml 25× dNTP mix. (c) 2 ml 10× RT random primers. (d) 1 ml MultiScribe Reverse Transcriptase. (e) 1 ml RNase inhibitor. (f) 3.2 ml nuclease-free water. Vortex to mix, centrifuge briefly, and add 10 ml of the master mix to each sample. 4. Flick the sample tubes to mix and centrifuge briefly to collect the contents at the bottom of the tube. Incubate the samples in a thermal cycler using the following program: 10 min at 25°C, 2 h at 37°C, 5 s at 85°C, and then hold at 4°C. 5. When the reverse transcription reaction is complete, add nuclease-free water to adjust the concentration of the samples and the reference to 5 ng/ml. For example, if 1 mg of aRNA was reverse transcribed in a 20-ml reaction, add 180 ml of water to bring the concentration to 1 mg/200 ml or 5 ng/ml. 6. Mix the diluted samples well and centrifuge briefly. Place on ice or store at −20°C. 7. Perform the qRT-PCR using the TaqMan® Gene Expression Assay of interest (see Note 29). These assays can be run on a variety of real-time instruments, including the Bio-Rad iCycler™. Turn the iCycler™ on and allow it to warm up for at

38

L.A. Field et al.

least 15 min. Enter the plate set-up sample information, select and load the appropriate fluorophore (FAM-490), and save the file. 8. qRT-PCR reactions should be performed in duplicate, but if sufficient starting material is available, triplicate or quadruplicate reactions are recommended. The following protocol is based on 10 ng of cDNA per reaction (see Note 30). Pipette 2  ml of cDNA (5  ng/ml) into the appropriate well of an iCycler™ PCR plate. 9. Prepare a master mix containing the following per reaction: (a) 2.5 ml TaqMan® Gene Expression Assay (20×). (b) 20.5 ml nuclease-free water. (c) 25 ml TaqMan® Universal PCR Master Mix. When running a full 96-well plate, make a master mix for 100 reactions (excess of four reactions). Vortex the master mix and centrifuge briefly to bring the contents to the bottom of the tube. Add 48 ml of master mix to each sample well. Place a thermal seal on the plate, centrifuge the plate for ~10 s at ~3,000 rcf, and then load the plate on the iCycler™. Set the reaction volume to 50  ml and run the following protocol: 95°C for 10 min, 50 cycles of 95°C for 15 s followed by 60°C for 1 min, then hold at 4°C. 10. Average the Ct values for each sample and the reference (see Note 31). Calculate the relative transcript levels for each sample using the Comparative Ct Method (see Note 32). 3.12. RNA Isolation from Formalin-Fixed Paraffin-Embedded Specimens

Archived FFPE tissues represent a valuable source of molecular information for the study of cancer (6–8). Although formalin fixation and paraffin embedding preserves tissue structure and cellular morphology, protein–protein and protein–nucleic acid cross links form during the preservation process, which chemically modify and damage the nucleic acids. RNA is particularly susceptible to fragmentation by FFPE, often making RNA isolated from FFPE tissues unusable for molecular analysis. The following protocol is specifically designed to recover RNA from FFPE samples that can be used for downstream applications such as qRT-PCR. 1. Add 42 ml of 100% ethanol to the Wash 1 concentrate and 48  ml of 100% ethanol to the Wash 2/3 concentrate. Mix well. 2. Place the laser microdissected FFPE samples in 400  ml of digestion buffer. Add 4 ml of protease to each sample and mix the tubes gently ensuring that the samples are entirely immersed in buffer. Incubate the samples in a heat block set to 50°C for 3 h (see Note 33).

2  Laser Microdissection for Gene Expression Profiling

39

3. Add 480 ml of isolation additive to each sample and vortex to mix. The solution should be white and cloudy (see Note 33). 4. Add 1.1 ml of 100% ethanol and mix by carefully pipetting the sample up and down. The sample should become clear after mixing (see Note 33). 5. Place a filter cartridge inside a collection tube provided with the kit. Apply 700 ml of the sample to the filter, centrifuge for 1 min at 10,000 rcf, then discard the flow-through and return the filter to the same collection tube. 6. Apply an additional 700 ml of sample to the filter and centrifuge for 1 min at 10,000 rcf. Continue this process until the entire sample has been passed through the filter. 7. Add 700  ml of Wash 1 solution to the Filter Cartridge and centrifuge for 30 s at 10,000 rcf. Discard the flow-through. 8. Add 500 ml of Wash 2/3 solution to the filter and centrifuge for 30 s at 10,000 rcf. Discard the flow-through. Centrifuge the Filter Cartridge assembly an additional 30 s at 10,000 rcf to remove residual liquid from the filter. 9. Treat the samples in the filters with DNase by combining for each sample: (a) 6 ml 10× DNase buffer. (b) 4 ml DNase. (c) 50 ml nuclease-free water. Add this solution to the center of each filter and incubate for 30 min at room temperature. 10. Add 700  ml of Wash 1 to the filters, incubate for 1  min at room temperature, then centrifuge for 30  s at 10,000 rcf. Discard the flow-through. 11. Add 500 ml of Wash 2/3 to the filters and centrifuge for 30 s at 10,000 rcf. Discard the flow-through. Add an additional 500  ml of Wash 2/3 to the filters, centrifuge for 1  min at 10,000 rcf, then transfer the filters to new collection tubes. 12. Add 30 ml of nuclease-free water heated to 95°C to the center of each filter (see Note 33). Incubate at room temperature for 1 min and then centrifuge for 1 min at 16,000 rcf. Repeat this step using a second 30  ml volume of heated nuclease-free water. 13. Read the concentration of the eluted samples on the NanoDrop and run the samples on a Bioanalyzer Pico Chip as described in Subheading 3.2.4 (Fig. 5). 14. Store samples at −80°C.

40

L.A. Field et al.

Fig. 5. Electropherogram of total RNA isolated from formalin-fixed paraffin-embedded breast tissue following laser microdissection. RNA was isolated using the RecoverAll™ Total Nucleic Acid Isolation Kit and assayed on the Bioanalyzer using a Pico Chip. Note that the 18S rRNA and 28S rRNA are completely degraded and thus not distinguishable.

3.13. Quantitative Real-Time Polymerase Chain Reaction Assays Using RNA from FFPE

RNA extracted from FFPE tissues is usually not conducive to amplification using an oligo(dT) primer due to fragmentation and degradation. Therefore, qRT-PCR is performed directly on RNA from FFPE tissues – there is no RNA amplification step. 1. Reverse transcribe the FFPE RNA as described. 2. Following reverse transcription of the FFPE RNA to cDNA, perform qRT-PCR as outlined for frozen tissue, but increase the number of PCR cycles to 60 cycles. 3. Average the Ct values for each sample and the reference as described above. Calculate the relative transcript levels for each sample using the Comparative Ct Method.

4. Notes 1. Flash-frozen tissues can only be sectioned after embedding in OCT. To embed a flash-frozen tissue sample in OCT, place the tissue and plastic cryomold in the cryostat set to −30°C. Cover the bottom of the cryomold with OCT, orient the tissue so that the region of interest is facing up, then fill the mold with additional OCT. Allow the compound to solidify – the block will turn white in color when completely frozen. 2. The foil slides should be at room temperature before applying a tissue section. Tissues will not properly adhere to cold slides. Note: If you decide to place more than one tissue section on

2  Laser Microdissection for Gene Expression Profiling

41

a slide to maximize resources, keep the slide in the cryostat while placing the additional tissue section(s) onto the slide. If the additional section does not adhere to the cold slide, warm the area of the slide where the tissue will be deposited by pressing your thumb to the reverse side. Work as quickly as possible. 3. If the OCT is not completely dissolved after the 50% ethanol wash, make sure that the solution is mixed properly. Shake the original container, add fresh solution, and wash the slide again for 30 s. 4. When staining multiple slides per patient sample, change the second 100% ethanol solution when it is noticeably purple in color, indicating that it is no longer a 100% solution. A 100% solution prevents leaching of stain from the slide and allows for faster drying. 5. Incubation at 56°C heat-fixes the section to the slide. The section will fall off the slide during staining if not properly fixed to the slide. 6. Because the slide has no cover-slip, histologic features may not be easily identified. Cellular details can be better visualized by pipetting 100% ethanol onto the tissue section. 7. Large microdissected areas can be lifted directly from the slide with a clean pair of forceps and placed into 100 ml of buffer. We generally perform LM at 4× magnification (after careful inspection at 10×) when working with frozen breast tissue. If the laser tends to drift from the user-defined cut lines, calibrate the laser periodically. With Laser Microdissection LMD software version 4.4, the laser control settings we typically use at 4× are aperture, 15; intensity, 44; speed, 6 for clean cuts; offset, 18; bridge, medium; aperture differential, 6. 8. The lysis solution and wash solutions may be stored at 4°C or room temperature, but be sure to warm the reagents to room temperature before beginning the protocol. 9. The Pico Ladder should only be heated when first diluted and aliquoted and may not run properly if heated multiple times. 10. Completely submerge the foil containing the microdissected tissue in the lysis solution to ensure complete cell lysis and inactivation of endogenous RNases. If this is not achieved by centrifugation, push the foil into the buffer using a pipette tip. 11. To isolate both large and small RNA species including tRNAs or microRNAs, add 129  ml of 100% ethanol rather than 52 ml. 12. When transferring the sample to the filter, be sure not to transfer large pieces of foil that may block the flow of liquid through the filter.

42

L.A. Field et al.

13. To increase accuracy when pipetting hot liquids, prewet the pipette tip before pipetting the 95°C water and applying it to the filter. 14. To easily remove the DNase Inactivation Reagent from the RNA and maximize RNA recovery, centrifuge the sample to pellet the Reagent. Transfer ~20 ml of RNA to a new tube, centrifuge the tube containing the Reagent for an additional 30 s to 1 min, then transfer the remaining RNA using a small bore pipette tip (10XL tips work well) without disturbing the DNase Inactivation Reagent pellet. 15. When running a Pico or Nano Chip, ensure that the reagents have not expired and that no more than 2 months have passed since the gel was filtered. Expired reagents can adversely affect the run and can cause aberrations to the baseline of the electropherogram. 16. When cleaning the Bioanalyzer electrodes prior to running a Pico or Nano Chip, fill the electrode cleaner with nucleasefree water, not RNaseZap®. The Pico Chip is very sensitive and even minute amounts of RNaseZap® can interfere with proper function. 17. When loading the Pico or Nano Chips, do not use the blowout function of the pipette as this may introduce air bubbles that may interfere with the chip running properly. 18. Most RNA samples isolated from frozen tissue via laser microdissection have a RIN (RNA Integrity Number) greater than 7, which is acceptable for downstream applications. Samples with a lower RIN (between 6 and 7) may be usable. Repeat the RNA isolation if significant degradation has occurred. Remove the sample from the study if subsequent isolations do not show improved RNA quality. 19. The pin set on the Bioanalyzer will need periodic maintenance – remove and clean thoroughly with RNaseZap®, then rinse thoroughly with nuclease-free water. 20. The final volume of this dilution may be adjusted based on the number of samples that will be amplified at one time. For example, if amplifying ten or more samples, prepare 40 ml of the fifth dilution by mixing 4 ml of the fourth dilution with 36 ml of nuclease-free water. 21. We typically use 10 ng of RNA as input for the first round of amplification; however, we have amplified as little as 2  ng with more than sufficient yields of second round aRNA. 22. If the starting amount for amplification is greater or less than 10 ng, adjust the volume of the Poly-A Controls proportionately. For example, if starting with 5 ng of RNA, add 1 ml of the fifth dilution to the sample.

2  Laser Microdissection for Gene Expression Profiling

43

23. Ensure that the thermal cycler lid is also cooled to 16°C. Exposure of the reaction to higher temperatures will lead to inefficient second-strand cDNA synthesis and will compromise aRNA yield. If the lid temperature cannot be adjusted to 16°C, turn the lid temperature off or incubate with the lid off. 24. We have started the second round of amplification with as little as 350 ng of first round aRNA with successful results. Any remaining aRNA from the first round of amplification can be stored at −80°C for future use. 25. Because the samples have just been heated to 94°C, it is not necessary to heat denature them at 70°C. The heat-denatured ladder can be used in this run without reheating. The same gel-dye mix that was prepared to run the Nano Chip for the second round aRNA can be used to run the Nano Chip for the fragmented aRNA if both chips are run on the same day. 26. Before placing the arrays in the hybridization oven overnight, cover each septum with a small piece of tape to prevent the sample from leaking out of the array. 27. Arrange the arrays in the hybridization oven with a balanced configuration to prevent undue stress to the motor. 28. We typically reverse transcribe sufficient aRNA to run all of the TaqMan® Gene Expression Assays that we have defined for the study so that additional reverse transcription reactions on the same sample are not necessary. 29. We recommend using TaqMan® Gene Expression Assays with the suffix “_ml” if possible, as these assays amplify regions spanning exon junctions and will not amplify genomic DNA that may contaminate the RNA sample. For “_g1” or “_s1” assays, include a control with no reverse transcriptase when performing the reverse transcription reaction and subsequent PCR to ensure that amplification is due to the presence of RNA transcripts and not genomic DNA contamination. 30. 10 ng of cDNA usually produces good results; however, when working with certain low abundance transcripts, the starting amount of cDNA may need to be increased and the volume of water in the master mix adjusted accordingly. 31. In some cases, the chosen reference RNA may not express a particular transcript of interest and an alternative reference RNA will need to be selected. 32. To calculate relative transcript levels using the Comparative Ct Method: (a) Determine the DCt as follows:

DC t = C t target - C t reference

44

L.A. Field et al.

where the target is the gene of interest and the reference represents an endogenous control such as actin or GAPDH. (b) Calculate the DDCt using the following formula:

DDC t = DC t test sample - DC t calibrator (c) The fold change relative to the calibrator is 2- DDC t . As an alternative to the Comparative Ct Method, relative levels of gene expression can be determined using the Relative Standard Curve Method. The following document describes both the Comparative Ct Method and the Relative Standard Curve Method: Applied Biosystems. Guide to Performing Relative Quantitation of Gene Expression Using Real-Time Quanti­tative PCR, available at: h t t p : // w w w 3 . a p p l i e d b i o s y s t e m s . c o m / c m s / g r o u p s / mcb_support/documents/generaldocuments/cms_042380.pdf.

33. This protocol has been modified by the manufacturer. The current RecoverAll™ protocol reduces the volume of digestion buffer from 400 to 100 ml, shortens the incubation from 3 h at 50°C to 15 min at 50°C followed by 15 min at 80°C, reduces the volume of isolation additive from 480 to 120 ml and the volume of 100% ethanol from 1.1 ml to 275 ml, and lowers the temperature of the eluant from 95°C to room temperature (22–25°C).

Acknowledgments This work was supported by the United States Department of Defense (Military Molecular Medicine Initiative MDA W81XWH05-2-0075, protocol #01-20006) and was performed under the auspices of the Clinical Breast Care Project, a joint effort of many investigators and staff members. The opinion and assertions contained herein are the private views of the authors and are not to be construed as official or as representing the views of the Department of the Army or the Department of Defense. References 1. Collins, F. S. and Barker, A. D. (2007) Mapping the cancer genome. Pinpointing the genes involved in cancer will help chart a new course across the complex landscape of human malignancies. Sci. Am. 296, 50–57. 2. Manolio, T. A., Brooks, L. D., and Collins, F. S. (2008) A HapMap harvest of insights into the

genetics of common disease. J. Clin. Invest. 118, 1590–1605. 3. Ellsworth, R. E., Seebach, J., Field, L. A., Heckman, C., Kane, J., Hooke, J. A., et  al. (2009) A gene expression signature that defines breast cancer metastases. Clin. Exp. Metastasis 26, 205–213.

2  Laser Microdissection for Gene Expression Profiling 4. Perou, C. M., Sorlie, T., Eisen, M. B., van de Rijn, M., Jeffrey, S. S., Rees, C. A., et  al. (2000) Molecular portraits of human breast tumours. Nature 406, 747–752. 5. Sørlie, 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. 6. Ellsworth, R. E., Ellsworth, D. L., Deyarmin, B., Hoffman, L. R., Love, B., Hooke, J. A., et  al. (2005) Timing of critical genetic changes in human breast disease. Ann. Surg. Oncol. 12, 1054–1060.

45

7. Becker, T. E., Ellsworth, R. E., Deyarmin, B., Patney, H. L., Jordan, R. M., Hooke, J. A., et  al. (2008) The genomic heritage of lymph node metastases: implications for clinical management of patients with breast cancer. Ann. Surg. Oncol. 15, 1056–1063. 8. Ellsworth, R. E., Hooke, J. A., Love, B., Kane, J. L., Patney, H. L., Ellsworth, D. L., et al. (2008) Correlation of levels and patterns of genomic instability with histological grading of invasive breast tumors. Breast Cancer Res. Treat. 107, 259–265.

wwwwwwwwwwwwwww

Chapter 3 Gene Expression Using the PALM System Jian-Xin Lu and Cheuk-Chun Szeto Abstract The PALM Robot MicroBeam laser microdissection system can isolate specified cells from complex tissues section, in a rapid and precise manner. Combined with other methods, PALM may be used for gene expression elucidating the role of specialized cell type in physiological and pathological activity. This chapter describes the application of the PALM MicroBeam system to isolate RNA from cells in a complex tissue for subsequent gene expression analysis. Protocols show the steps from preparation of tissue samples to the final quantitative results. The process is articulated in several steps, each of which requires optimal choices in order to obtain reliable data from a limited number of cells (500–10,000 cells). Furthermore, the notes regarding tissue preparation, microdissection of the interested cells, are also emphasized. Key words: Gene expression, PALM, Microdissection, RNA, Real-time quantitative PCR

1. Introduction The PALM® Robot MicroBeam laser microdissection system (P.A.L.M. Carl Zeiss, Bernried, Germany) provides a convenient, efficient, and precise method of selecting specified cell populations from complex tissues for subsequent analysis of their RNA, DNA, or protein content. Microdissection of cells defined under the microscope ensures to obtain pure samples of cells of interest for downstream molecular applications. Molecular biological techno­ logy developments (e.g., PCR) allow analyzing gene expression from only limited amounts of cells. Thereby, the combination of microdissection and PCR may allow the possibility of assessing the role of specialized cell type or tissue in the normal physiologic condition or under disease process (1–3). Thus, the details of the disease’s pathogenesis, diagnostic and prognostic accuracy, and ultimately targeted therapeutics may be elucidated much clearer.

Graeme I. Murray (ed.), Laser Capture Microdissection: Methods and Protocols, Methods in Molecular Biology, vol. 755, DOI 10.1007/978-1-61779-163-5_3, © Springer Science+Business Media, LLC 2011

47

48

J.-X. Lu and C.-C. Szeto

In principle, PALM system removes coherent cell fields by applying a pulsed UVA laser through an inverted microscope to allow laser ablation of cells and tissue on a tissue section, either manually or automatically (4). It belongs to laser cutting microdissection. Briefly, three steps are involved: first, the tissue is mounted on a membrane-cover slide and is viewed by a computer. Subsequently, the pulsed ultraviolet A (UVA) laser beam (focal spot 0.200 V and the spot size to 7.5 mm. The spot size and intensity of the laser can be adjusted by the computer software or more easily by the Controller box. The laser cannot be seen through the oculars of the microscope and has to be visualized and focused on the computer screen. Switch to the 20× objective, and focus the beam as a sharp white targeting spot surrounded by one or two sharp pink or purple halos (it may be necessary to reduce the level of illumination to do this). Once focused, set up the laser spot size to approximate the size of the cells to be captured (often, this is 7.5 mm). Put the computer mouse arrow over the laser to mark it. This will allow you to increase illumination to visualize cells, should it be necessary, without losing sight of the location of the laser beam. 7. The power and duration of the laser pulse need to be set and will depend on the type of CapSure Cap. Tables indicating appropriate power and duration for different Caps and different laser spot sizes are provided by the manufacturer. Then, still on an area of slide where there are no cells, test fire the laser. There should be an audible beep as it is fired. Move the stage slightly to see the melted spot of polymer that should have been produced. There should be a visible thin ring around a clear area marking where the laser has been fired and melted the polymer. The manufacturer’s manual provides pictures of what the spot should look like and suggestions about what adjustments to the laser duration and power might be made to correct one that is not optimal. The power and duration values are saved with the spot size and will be recalled when the machine is next used. It is also possible to save personalized settings with the Archiving Software. 8. Move the stage to find cells to be captured that are within the black ring of the CapSure Cap. Position a cell that is wanted under the laser and fire the laser. Repeat until all cells within

6  Laser Capture Microdissection for Analysis of Gene Expression¼

81

the black ring that are wanted have been captured. Up to 3,000 cells may be accommodated by one cap, depending on the size of the cells. However, the cap cannot be usefully moved to another position on the sample and used to collect more cells. 9. Lift the CapSure Cap from the slide with the Cap Placement Arm. A picture of the area that has been captured and the cells remaining can be taken following the instructions provided for the Image Archiving Software. Cells that have been lifted can also be visualized by turning off the vacuum pump, moving the slide to an area where there is no sample, putting the pump back on, and lowering the Cap again. Adjust the focus, and scan the cap by moving the stage. An image of the cells that have been captured can also be made with the Image Archiving Software. If cells have not been captured, refer to the manufacturer’s troubleshooting guide for suggestions. 10. Lift the CapSure Cap again and move it to the Unload Station, a platform on the right of the microscope. Remove the cap from the Upload Station with the Cap insertion tool provided, taking great care not to touch the polymer surface of the Cap and in turn remove the Cap from the Cap insertion tool using tweezers, and put it in the Alignment tray with the sample facing up. The tray will hold up to 28 Caps. 11. Position an ExtracSure Device over the Cap with its central fill port facing up and push down to engage. The device seals the perimeter of the Cap, but allows the addition of enzymes to its surface through the Device fill port. 12. Add proteinase K solution to the fill port, and put a 0.5-ml microcentrifuge tube over the top of the port. The volume depends on the number of cells captured, but about 25 ml is appropriate for 1,000–3,000 cells. Incubate in the incubation block at 50°C for 16–20  h (a hybridization oven is convenient for this). Invert the device and cap and centrifuge cell extract into the microcentrifuge tube. If there is considerable debris in the extract, transfer the supernatant to a new tube. Freeze at −80°C or proceed immediately to RNA isolation. 3.3. RNA Isolation and Reverse Transcription

1. Follow the detailed instructions provided with the Paradise Whole Transcript RT Reagent System, taking particular care to clean the work area with nuclease decontamination solution. Wear disposable gloves and ensure that solutions are nuclease-free, plasticware is nuclease-free, and all forceps are washed with detergent and well-rinsed with nuclease-free water. The steps involved are, briefly, as follows: 2. Precondition the purification column with conditioning buffer using the collection tube provided to centrifuge out the buffer. Prepare the master mix of binding buffer and ethanol

82

R. Jiang et al.

solution, add to the cell extract in the microcentrifuge tube and load in no more than 210 ml volumes onto the preconditioned column. Centrifuge to bind RNA to the column membrane and repeat until all extract is loaded. 3. Wash as instructed, making sure that all wash buffer is removed before elution of RNA with 12 ml of elution buffer into a new microcentrifuge collection tube that is provided. Discard the purification column. At this point, the RNA can be stored at −80°C or DNase treatment can be done immediately. 4. Add 2 ml sDNase reaction mix to the RNA. Incubate tube in a thermal cycler at 37°C for 15 min, chill, and add sDNase stop solution. Inactivate DNase at 70°C for 10 min. Chill and centrifuge. Remove 1–2 ml to measure RNA yield by spectrophotometry. Either store the remaining RNA at −80°C or proceed directly to reverse transcription. 5. Transfer 10–2,000 ng of the RNA to be reverse transcribed to a microcentrifuge tube, and add the reagents provided as instructed. These include random primers and First Strand Synthesis mix comprising an RT master mix, RT enzyme mix, enhancer, and the SuperScript III reverse transcriptase, which is not provided in the kit and has to be purchased separately. Incubate as described in the user guide. Reverse transcription is then followed with a final nuclease digestion to remove any remaining RNA. The cDNA can be frozen for storage or used immediately for RT-QPCR. 3.4. RT-QPCR

1. Primers and probes used for RT-QPCR can be designed by a software program, such as Beacon Designer 7.0 (Premier Biosoft International) or Primer Express 2.0 (Applied Biosystems). Check the specificity of the primers using cDNA obtained from a cell line known to express the target gene and SYBR Green reagents (see below). A dilution series should be prepared for primer validation. The dissociation curves of the amplified products, melting temperatures, and profiles should be similar. Also visualize the specificity of the PCR product on an agarose gel. 2. Set up RT-QPCR in duplicate in a MicroAmp Fast Optical 96-well plate. For SYBR green reactions, use 3  ml cDNA, 15 ml TaqMan Universal PCR master mix, and a final concentration of primers of 300 nM in a total volume of 30 ml. For TaqMan reactions, use 3 ml cDNA, 15 ml TaqMan Universal PCR master mix, primers at a final concentration of 300 nM, and probes at a final concentration of 200  nM in a final volume of 30  ml (see Note 2). Seal the plate with Optical Adhesive Film. 3. For TaqMan thermocycling, run 40 cycles of 95°C for 15  s and 60°C for 1 min with an initial cycle of 50°C for 2 min and

6  Laser Capture Microdissection for Analysis of Gene Expression¼

83

95°C for 10 min. For SYBR green, after the PCR cycles, add a dissociation step of 95°C for 15 s, 60°C for 1 min, 95°C for 15 s, and 60°C for 15 s to obtain a melting curve (see Note 3). 4. Set the background immediately prior to the exponential rise in amplification. Set the threshold cycle (Ct) of target and reference genes to ten standard deviations above the background. If the relative efficiencies of amplification of target and reference genes are similar (see Note 4), the relative amounts of the target gene can be calculated by the comparative Ct method (3) and, if desired, reactions can be multiplexed. DCt represents the difference in Ct values between the target and reference genes at the same condition points for each sample. DDCt is described by the equation: DDCt = DCt (target gene) − DCt (target gene expression in a cell line chosen as an arbitrary standard). The numerical value for DDCt is then used to calculate 2−DDCt, which represents the differential expression of the target gene in the sample relative to that in the chosen arbitrary standard. If the efficiencies of amplification of target and reference genes are not the same, then reactions for each are run separately and compared to a standard curve constructed from an RNA of known concentration (see Note 4).

4. Notes 1. Sections of up to 7.5 mm can be used, but those of 4–5 mm are easier to capture with the LCM system. 2. Primer concentrations may have to be optimized for each given set and scaled to the amount of template available. 3. Ct values for amplification of reverse-transcribed RNA obtained from FFPE sections may be high because of the difficulties in isolating sufficient quantities of good-quality RNA, even when multiple cells are captured. This is particularly true for sections from tissues more than 2 or 3 years old. To ensure that the Ct values do not represent artifacts, it is important to ensure that melting temperatures of individual cycles are identical. This can either be checked using SYBR Green and comparing the dissociation curves of the amplified products, which should be the same, or by determining if the Ct values fall in the linear range previously determined for the primer set in Subheading 3.4, step 1. 4. To make valid comparisons of the expression of different genes, it is important to verify that the relative efficiencies of amplification for each primer set are similar to those of the reference gene. The absolute value of the slope of a plot of

84

R. Jiang et al.

log cDNA concentration vs. DCt (Cttarget − CtGAPDH) should be less than 0.1. If it is not, then amplification efficiencies of each primer set will have to be taken into account in the quantitation method (4).

Acknowledgments This work was supported by grants DE 016669 and CA114416. References 1. Curran, S., McKay, J. A., McLeod, H. L., Murray, G. I. (2000) Laser capture microscopy. Journal of Clinical Pathology 53, 64–68. 2. Fend, F., Emmert-Buck, M. R., Chuaqui, R., Cole, K., Lee, J., Liotta, L. A., Raffeld, M. (1999) Immuno-LCM: laser capture microdissection of immunostained frozen sections for mRNA anal­ ysis. American Journal of Pathology 154, 61–66.

3. Livak, K. J., Schmittgen, T. D. (2001) Analysis of relative gene expression data using real-time quantitative PCR and the 2(−Delta Delta C(T)) Method. Methods 25, 402–408. 4. Pfaffl, M. W. (2001) A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Research 29, e45.

Chapter 7 MicroRNA Profiling Using RNA from Microdissected Immunostained Tissue Clemens L. Bockmeyer, Danny Jonigk, Hans Kreipe, and Ulrich Lehmann Abstract mRNA expression profiling has been used to define molecular subtypes of human breast cancer. Also microRNAs have been investigated in these breast cancer subtypes. However, little is known regarding the microRNA signature of healthy luminal and basal breast epithelial cells. Therefore, a method is described to isolate immunostained luminal and basal breast epithelial cells in formalin-fixed paraffin-embedded tissues by laser microdissection. Employing this new methodological approach, we could identify distinct microRNA profiles of luminal and basal breast epithelial cells by real-time PCR-based profiling. Key words: Breast, Laser microdissection, miRNA, Myoepithelial, Luminal

1. Introduction MiRNA gene expression studies are in the focus of many research areas (1). Not only in cancer, but also in inflammatory or vascular disease, miRNAs are important regulators of many pathways (2–4). Real-time PCR is suitable for quantitative studies of changes in the individual miRNA expression of small amounts of miRNAs (5). Precise quantification is achieved routinely with as little as 25 pg of total RNA for most miRNAs (6). Multiplex approaches are also developed to profile all known miRNAs similar to the microarray approach, but with higher sensitivity for the detection of small amounts of RNA (7). Breast epithelial cells have a specialized phenotype maintained by the expression of a unique set of mRNA genes (8). Furthermore, in situ hybridization studies performed on healthy breast tissues

Graeme I. Murray (ed.), Laser Capture Microdissection: Methods and Protocols, Methods in Molecular Biology, vol. 755, DOI 10.1007/978-1-61779-163-5_7, © Springer Science+Business Media, LLC 2011

85

86

C.L. Bockmeyer et al.

identified an alteration of miRNA expression (9). This method is excellent for direct visualization of cell-specific miRNA transcripts. However, in contrast to real-time PCR, in situ hybridization provides no quantification, has a low-throughput and a low specificity in discriminating between closely related microRNA transcripts. The MMI CellCut plus laser microdissection system is a robust and reliable tool for contamination-free isolation of single cells or small groups of cells. It allows sample inspection at high magnification, and cells of interest are easy to dissect due to solid-state laser technology and user-friendly operation software (10). This technique is especially useful for dissection of single cells like megakaryocytes (11). Combination of this laser microdissection system and real-time PCR allows appropriate gene expression profiling of specific cell types. In this chapter, a detailed protocol is provided for microdissection of luminal and basal breast epithelial cells and subsequent RT-PCR-based profiling of miRNA transcripts.

2. Materials Manufacturers or distributors are specified only if reagents or laboratory equipment might be important for the outcome or if a source might be difficult to identify. All chemicals were purchased in analytical grade quality from Merck, Roth, J.T Baker, or Sigma. 2.1. Tissue Preparation

1. 4% Phosphate-buffered formaldehyde. 2. Graded ethanols (100, 90, 70, and 50%). 3. Xylene. 4. Paraffin. 5. Membrane slides for laser microdissection, nuclease-free (Molecular Machines & Industries, Glattburg, Switzerland). 6. Hydrogen peroxide. 7. Double-distilled H2O. 8. 50 mM Tris–HCl, pH 7.4, 0.9%NaCl (Tris buffer). 9. ZytoChem Plus (HRP) Polymer Kit (Zytomed Systems). 10. Smooth muscle actin antibody, monoclonal anti-mouse, clone 1A4 (DAKO). 11. Antibody diluent (Zytomed Systems). 12. Diaminobenzidine (Zytomed Systems). 13. Hemalaun. 14. Tubes for laser microdissection (Molecular Machines & Industries, Glattburg, Switzerland).

7  MicroRNA Profiling Using RNA from Microdissected Immunostained Tissue

2.2. RNA Extraction

87

1. Proteinase K solution (20 mg/mL, Merck). 2. Guanidinium thiocyanate. 3. 1 M Tris–HCl pH 7.6. 4. ß-Mercaptoethanol. 5. 3 M Sodium acetate. 6. Roti-aqua-phenol (Roth, Karlsruhe). 7. Chloroform. 8. Isopropanol. 9. Glycogen (Roche, Basel). 10. Ethanol. 11. DEPC H2O. 12. Digestion solution: 50  μl 4.2  M guanidinium thiocyanate/ 30  mM Tris–HCl pH 7.6/2% Sodium-N-lauryl sarcosine, 50 ml proteinase K, and 0.5 ml ß-Mercaptoethanol.

2.3. cDNA Synthesis and Preamplification

1. TaqMan microRNA RT Kit (Applied Biosystems). 2. Megaplex RT primers, human Pool A and B (Applied Biosystems). 3. Megaplex Preamp primers, human Pool A and B (Applied Biosystems). 4. Preamp mastermix (Applied Biosystems).

2.4. Real-Time PCR-Based Profiling

1. Universal PCR Mastermix (Applied Biosystems).

2.5. Equipment

1. Automatic tissue-processing device.

2. TaqMan human microRNA Array A and B (Applied Biosystems).

2. Microtome. 3. MMI CellCut plus laser microdissection system (Molecular Machines & Industries, Glattburg, Switzerland). 4. Centrifuge (Thermos scientific, Fresco 17). 5. GeneAmp PCR system 9700 – 96-well (Applied Biosystems). 6. Centrifuge (Heraeus, Multicentrifuge 3SR). 7. Sealer (Applied Biosystems). 8. Array holder for centrifugation (Applied Biosystems). 9. 7900 HT Fast Real-Time PCR system (Applied Biosystems). 10. SDS 2.3 software (Applied Biosystems). 11. RQ manager 1.2 software (Applied Biosystems).

88

C.L. Bockmeyer et al.

3. Methods The protocols described below for microRNA profiling of immunostained laser-microdissected tissue concentrate on the following steps: 1. Specimen preparation and immunohistochemistry. 2. Laser microdissection. 3. Isolation of RNA. 4. Megaplex cDNA synthesis and preamplification. 5. MicroRNA profiling. 6. Data interpretation. 3.1. Specimen Preparation (see Note 1) and Immunohistochemistry

This work has to be done under nuclease-free conditions. 1. Serial 5 mm breast tissue sections are cut with a fresh knife, floated out on a hot water bath, and finally mounted on membrane slides. Tissue sections should be positioned in the center of the slide and on the bottom side. Slides are dried at 37°C overnight. 2. Slides are deparaffinised, incubated for 10 min in 3% hydrogen peroxide, and rehydrated in 50 mM Tris buffer for 5 min. 3. Sections are incubated for 5 min with a commercial blocking solution (Reagent 1, ZytoChem Plus (HRP) Polymer Kit) and washed in Tris buffer for 5 min. 4. Slides are incubated for 1  h with SM-actin antibody (room temperature, 1:25 diluted) and washed in 50 mM Tris buffer for 5 min (see Note 2). 5. For signal intensification, reagent 2 [ZytoChem Plus (HRP) Polymer Kit] is applied for 20 min and slides are washed in Tris buffer for 5 min. 6. Slides are incubated for 30 min with secondary anti-mousepolymer antibody (Reagent 3, ZytoChem Plus (HRP) Polymer Kit) and washed in Tris buffer for 5 min. 7. Slides are incubated for 5 min in distilled water and incubated for 8 min with DAB (1:20 diluted). 8. Slides are washed in distilled water for 5  min and counterstained with hemalaun for 2 min (see Note 3). 9. Slides are dried at 37°C overnight.

3.2. Laser Microdissection

This work has to be done under nuclease-free conditions. For stabilization, one glass slide is positioned under the membrane slide, so the tissue is located between a glass slide and the membrane. It is necessary to use the 40× objective for dissection

7  MicroRNA Profiling Using RNA from Microdissected Immunostained Tissue

89

Fig. 1. Laser microdissection of immunostained breast epithelial cells. Basal cells were stained for smooth muscle actin (horseradish peroxidase, dark grey ), whereas nuclei were counterstained by hemalaun (a). First luminal cells were cut (b) out and then retrieved using adhesive isolation caps (c). Afterward, basal cells were cut out and retrieved using a second cap (d) (original magnification: ×400).

of luminal and basal breast epithelial cells. First, all luminal cells were cut and collected by the adhesive lid (Fig. 1). In our experience, one can collect about 100 acini by the adhesive membrane of the lid. In the second step, all basal cells were cut and collected by the adhesive lid of a second tube. In total, for each sample, about 3,000 luminal cells and 1,600 basal cells were dissected. Using the 40× objective, cutting speed 4, laser focus 43, and power 71, the estimated line width is optimal for dissection of luminal and basal breast epithelial cells (about 0.8 mm; see Note 4). The MMI laser has the advantages of using low pulse energy and high repetition rate. 3.3. Isolation of RNA

This work was done under nuclease-free conditions (see Note 5). 1. Microdissection tubes are filled with 50 ml digestion solution, and then the tubes are turned upside down and incubated at 55°C overnight.

90

C.L. Bockmeyer et al.

2. The tubes have to be centrifuged briefly, and the lysate has to be transferred into a normal tube without adhesive lid. This step is necessary due to the fact that microdissection tubes cannot be centrifuged for a long time at high speed because the adhesive lid might drop down into the tube during prolonged centrifugation. 3. For phenol–chloroform extraction, 100  ml lysate is mixed with 10 ml 3 M sodium acetate, 63 ml Roti-aqua-phenol, and 27 ml chloroform strongly vortexed and incubated on ice for 20 min. It is necessary to maintain the order of adding first sodium acetate, then Roti-aqua-phenol, and finally chloroform. Subsequently, the samples are centrifuged for 30 min at 16,200 × g. 4. The supernatant (95  ml) is precipitated by incubation with 95 ml ice-cold isopropanol and 1.5 ml glycogen (20 mg/mL) at −20°C for at least one night. 5. After centrifugation for 30 min (16,200 × g), the supernatant is discarded and the pellet is washed in 180 ml ice-cold ethanol (70%). 6. After centrifugation for 5 min at 16,200 × g, the supernatant is again discarded and the pellet air-dried (see Note 6). 7. The RNA pellet is dissolved in 15 ml DEPC H2O and stored at −20°C. (Note 7). 3.4. Megaplex cDNA Synthesis and Preamplification

1. 3 ml RNA is mixed with 4.5 ml megaplex–mastermix (0.8 ml megaplex RT primer, 0.2 ml dNTPs (100 mM), 0.2 ml DEPC H2O, 0.8 Ml 10× RT buffer, 0.9  ml MgCl, 0.1  ml RNaseInhibitor, 1.5  ml multiscribe reverse transcriptase) (each for human pool A and B). 2. Cycling modus (ABI Gene Amp 9700 HT): 40 cycles at 16°C for 2 min, 42°C for 1 min, and 50°C for 1 s, and last 1 cycle at 85°C for 5 min. 3. 2.5  ml cDNA is preamplified: 12.5  ml preamp mastermix, 2.5  ml megaplex–preamp primers, 7.5  ml DEPC H2O, and cDNA are mixed (each for human pool A and B). 4. Cycling modus (ABI Gene Amp 9700 HT): 1 cycle at 95°C for 10 min, 1 cycle at 55°C for 2 min, 1 cycle at 72°C for 2 min, 12 cycles at 95°C for 15 s and 60°C for 4 min, 1 cycle at 99°C for 10 min.

3.5. MicroRNA Profiling (See User Bulletin Applied Biosystems TaqMan Low-Density Array)

1. For miRNA profiling, we use the so-called TaqMan low-density arrays. One array has 384 reaction chambers therein lyophilized primers are provided. 2. For one measurement, 24  ml preamplified cDNA, 388.5  ml DEPC H2O, and 412.5  ml universal PCR mastermix are

7  MicroRNA Profiling Using RNA from Microdissected Immunostained Tissue

91

mixed. 100 ml of this sample-specific reaction mix is dispensed in each fill port. 3. Arrays are placed into special array holders and centrifuged two times at 200 ´ g for 1 min. 4. After that, the microfluidic card has to be carefully placed in the sealer to seal each reaction chamber of the array. 5. Real-time PCR is performed on the 7900 HT PCR system. Cycling modus: 1 cycle at 50°C for 2 min, 1 cycle at 94.5°C for 10 min, 45 cycles at 97°C for 30 s and 59.7°C for 1 min. 3.6. Data Interpretation

Amplification curves and CT values are generated with the software RQ manager 1.2 (Applied Biosystems). Amplification curves for every reaction are inspected visually and underwent a stringent quality control procedure: 1. Transcripts at a detectable level are defined as those with a CT value below 38 and a regular sigmoid-shaped amplification curve. 2. “No meaningful amplification curve” is defined as atypical curve with double-sigmoid form, early flat slope, or no amplification plateau (Fig.  2). This curve is omitted from further analysis, even when CT values are below 38. All miRNA transcripts are classified into three groups: (1) no detectable expression, (2) “no meaningful amplification curve” considered as not analyzable, and (3) distinct gene expression patterns. Unfortunately, up to now, no universally accepted consensus of how to evaluate amplification plots has been established. 3. For the identification of suitable endogenous controls, eight different candidate microRNAs were analyzed for variance in gene expression according to two different algorithms: NormFinder (12) and geNorm (13). Both statistical methods ranked the candidate endogenous control genes with an excellent correlation in raw stability values (Fig. 3). The most stably expressed were snRNA U6 and RNU48.

4. Notes 1. It is difficult to get healthy mammary tissues with sufficient amounts of epithelial cells. Therefore, you have to sample about ten tissue blocks from one tissue specimen. Use the best block for microdissection. 2. It is necessary to use immunohistochemistry for clear separation of luminal from basal cells. In our experience, only hemealaun staining does not provide sufficient distinction between luminal

92

C.L. Bockmeyer et al.

Fig. 2. “No meaningful amplification curves”. Amplification curves of six luminal samples each for miR-224 and miR-551b are shown. Amplification curves were defined as “no meaningful amplification curves,” when curves demonstrated a double-sigmoid form, early flat slope, or no amplification plateau.

7  MicroRNA Profiling Using RNA from Microdissected Immunostained Tissue

93

NormFinder (stability value)

2.0 miR17

1.5 RNU44

1.0 RNU24

0.5

miR191 RNU48 snRNU6

2.0

p < 0.01 r = 0.91

RNU43

2.5

3.0

3.5

geNorm (M-value) Fig.  3. Normalization of different candidate genes for qPCR of miRNA transcripts. CT values of snRNU6, RNU6B, RNU24, RNU43, RNU44, RNU48, miR-17, and miR-191 were analyzed by calculating DCT values [CT (luminal sample–basal sample); n = 5]. Relative quantities (2−DCT) for each candidate gene were measured for stability by two different calculations (geNorm and NormFinder). snRNU6 and RNU48 were most stable as indicated by a low-stability value and M-value.

and basal cells probably due to the reduced optical quality. You have to keep in mind that tissue sections are dried and not coverslipped for microdissection. We stained SMA to visualize basal cells due to the fact that this antigen does not need to be unmasked. We have no experience with heat-induced or enzymatic antigen retrieval. Immunohistochemistry was performed under PCR contamination-free conditions. In general, you have to consider that all steps after RNA isolation have to be made under PCR contamination-free conditions. 3. Slides should be stained only for 1  min with hemealaun. Otherwise, staining results may be too dark and it could be difficult to differentiate between luminal and basal cells. In our experience, it is better to use tubes with a transparent lid for laser microdissection of brown and blue stained tissues. 4. The exact values might vary slightly depending on the original installation of the laser by the manufacturer. 5. We use the conventional guanidine-phenol-chloroform method with an ethanol precipitation step. Different methods for RNA extraction and consecutive mRNA analysis are described elsewhere (see, e.g., ref. 14). You have to keep in mind that not all published protocols or commercially available kits are suitable for the isolation of very short, mature microRNAs. 6. Avoid prolonged air drying because overdried RNA pellets cannot be dissolved again. 7. For short-term storage (up to a few weeks), aqueous RNA solutions might be stored at −20°C. However, for longer

94

C.L. Bockmeyer et al.

periods of time, storage at −80°C is recommended. For longterm storage, precipitation is the best way of preserving RNA integrity. References 1. Boyd SD. (2008) Everything you wanted to know about small RNA but were afraid to ask. Lab Invest. 88, 569–78. 2. Mirnezami AH, Pickard K, Zhang L, Primrose JN, Packham G. (2009) MicroRNAs: key players in carcinogenesis and novel therapeutic targets. Eur J Surg Oncol. 35, 339–347. 3. Suárez Y, Sessa WC. (2009) MicroRNAs as novel regulators of angiogenesis. Circ Res. 104, 442–54. 4. Urbich C, Kuehbacher A, Dimmeler S. (2008) Role of microRNAs in vascular diseases, inflammation, and angiogenesis. Cardiovasc Res. 79, 581–588. 5. Tang F, Hajkova P, Barton SC, Lao K, Surani MA. (2006) MicroRNA expression profiling of single whole embryonic stem cells. Nucleic Acids Res. 34, e9. 6. Chen C, Ridzon DA, Broomer AJ, Zhou Z, Lee DH, Nguyen JT, Barbisin M, Xu NL, Mahuvakar VR, Andersen MR, Lao KQ, Livak KJ, Guegler KJ. (2005) Real-time quantification of microRNAs by stem-loop RT-PCR. Nucleic Acids Res. 33, e179. 7. Lao K, Xu NL, Yeung V, Chen C, Livak KJ, Straus NA. (2006) Multiplexing RT-PCR for the detection of multiple miRNA species in small samples. Biochem Biophys Res Commun. 343, 85–89. 8. Jones C, Mackay A, Grigoriadis A, Cossu A, Reis-Filho JS, Fulford L, Dexter T, Davies S, Bulmer K, Ford E, Parry S, Budroni M, Palmieri G, Neville AM, O’Hare MJ, Lakhani SR. (2004) Expression profiling of purified normal human luminal and myoepithelial breast

cells: identification of novel prognostic markers for breast cancer. Cancer Res. 64, 3037–3045. 9. Sempere LF, Christensen M, Silahtaroglu A, Bak M, Heath CV, Schwartz G, Wells W, Kauppinen S, Cole CN. (2007) Altered MicroRNA expression confined to specific epithelial cell subpopulations in breast cancer. Cancer Res. 67, 11612–11620. 10. Palinauskas V, Dolnik O, Valkiūnas G, Bensch S. (2010) Laser microdissection microscopy and single cell PCR of avian haemosporidians. J Parasitol.:1. 11. Theophile K, Hussein K, Kreipe H, Bock O. (2008) Expression profiling of apoptosisrelated genes in megakaryocytes: BNIP3 is downregulated in primary myelofibrosis. Exp Hematol. 36, 1728–1738. 12. Andersen CL, Jensen JL, Ørntoft TF. (2004) Normalization of real-time quantitative reverse transcription-PCR data: a model-based variance estimation approach to identify genes suited for normalization, applied to bladder and colon cancer data sets. Cancer Res 64, 5245–5250. 13. Vandesompele J, De Preter K, Pattyn F, Poppe B, et al. (2002) Accurate normalization of realtime quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 3, 34.1–34.11. 14. Votavova H, Forsterova K, Stritesky J, Velenska Z, Trneny M. (2009) Optimized protocol for gene expression analysis in formalin-fixed, paraffin-embedded tissue using real-time quantitative polymerase chain reaction. Diagn Mol Pathol. 18, 176–82.

Chapter 8 Profiling Solid Tumor Heterogeneity by LCM and Biological MS of Fresh-Frozen Tissue Sections Donald J. Johann, Sumana Mukherjee, DaRue A. Prieto, Timothy D. Veenstra, and Josip Blonder Abstract The heterogeneous nature of solid tumors represents a common problem in mass spectrometry (MS)-based analysis of fresh-frozen tissue specimens. Here, we describe a method that relies on synergy between laser capture microdissection (LCM) and MS for enhanced molecular profiling of solid tumors. This method involves dissection of homogeneous histologic cell types from thin fresh-frozen tissue sections via LCM, coupled with liquid chromatography (LC)-MS analysis. Such an approach enables an in-depth molecular profiling of captured cells. This is a bottom-up proteomic approach, where proteins are identified through peptide sequencing and matching against a specific proteomic database. Sample losses are minimized, since lysis, solubilization, and digestion are carried out directly on LCM caps in buffered methanol using a single tube, thus reducing sample loss between these steps. The rationale for the LCM-MS coupling is that once the optimal method parameters are established for a solid tumor of interest, homogeneous histologic tumor/tissue cells (i.e., tumor proper, stroma, etc.) can be effectively studied for potential biomarkers, drug targets, pathway analysis, as well as enhanced understanding of the pathological process under study. Key words: Thin fresh-frozen tissue sections, Laser capture microdissection, Liquid chromatographymass spectrometry, Solid tumor heterogeneity, Biomarker, Cancer

1. Introduction Solid tumors have a heterogeneous cellular architecture. Critical functional units include cancer cells proper and stromal elements. The histology is often complex. For instance, an epithelial tumor may contain regions of: inflammation, neovascularity, carcinoma in situ, well to poorly differentiated carcinoma, nerves, hyperplasia, etc. The tumor microenvironment is composed of both normal and modified stromal cells that serve to nurture the malignant Graeme I. Murray (ed.), Laser Capture Microdissection: Methods and Protocols, Methods in Molecular Biology, vol. 755, DOI 10.1007/978-1-61779-163-5_8, © Springer Science+Business Media, LLC 2011

95

96

D.J. Johann et al.

process. The tumor stroma is now recognized as an important area in cancer therapy, and many new therapeutic strategies target the aspects of this functional region (1). Solid tumor heterogeneity is reflective of a diversity present at the molecular level that has profound biologic and therapeutic implications (2). For instance, breast cancer is actually many different diseases with the only common characteristic being the organ of origin. Hence, the ability to directly and effectively profile solid tumors at the proteome level is essential, since proteins are the final mediators of pathologic processes and proteomics in particular can begin to characterize molecular events, such as alternative protein splicing and posttranslational modifications, which are fundamental events in physiologic/pathologic processes. Additionally, although cell culture studies are quite important, they lack a true microenvironment and thus clinical translation can be limited. Therefore, methods to deconstruct solid tumors to better enable biological understanding and biomarker discovery are needed (3, 4). Laser capture microdissection (LCM) (5) and mass spectrometry (MS) (6) are powerful independent analytical technologies. Both have been commonly used for molecular profiling of formalinfixed paraffin-embedded tissue sections (7). We have shown that an LCM-MS platform can be effectively used for proteomic profiling of thin fresh-frozen tissue sections obtained from a solid tumor in conjunction with a simple methanol-aided solubilization and digestion (8) of captured cell proteomes (9). In this chapter, we further illustrate the method employed for the proteomic profiling of a solid tumor using LCM coupled to biological MS.

2. Materials 2.1. LCM

1. TISSUE-Tek O.C.T. cryostat-mounting medium (Sakura Finetek Inc., Torrance, CA). 2. Mayer’s hematoxylin solution (Sigma, St. Louis, MO). 3. Eosin Y solution (alcohol-based) (Sigma, St. Louis, MO). 4. Scott’s Tap Water Substitute Bluing Solution (magnesium sulfate buffered with sodium bicarbonate; Fisher Scientific, Hampton, NH). 5. 100% ethanol (ethyl alcohol, absolute, 200° proof for molecular biology). 6. 70% ethanol (v/v) and 95% (v/v) were established using Milli-Q-filter with purified H2O (Millipore, Billerica, MA).

8  Profiling Solid Tumor Heterogeneity by LCM and Biological¼

97

7. Xylene. 8. CapSure® Macro LCM Caps (MDS Analytical Technologies, Sunnyvale, CA). 9. PixCell IIe, Veritas, or ArcturisXT (Arcturus Molecular Devices now MDS Analytical Technologies, Sunnyvale, CA). 10. Leica Cryostat CM 1850 UV (Leica Microsystems, Wetzlar, Germany). 11. Precleaned glass microscope slides, 25 mm × 75 mm (Fisher Scientific, Hampton, NH). 12. Membrane slide options include: (a) Pen-membrane glass slide. (b) Pen-membrane frame slide; both options available from (Arcturus Molecular Devices, now MDS Analytical Technologies, Sunnyvale, CA). 2.2. Protein Extraction and Digestion

1. Ammonium bicarbonate. 2. Sequencing grade trypsin (Promega, Madison, WI). 3. Trifluoroacetic acid (TFA) and formic acid (FA, Fluka, Milwaukee, WI). 4. HPLC grade acetonitrile (ACN; CH3CN) and methanol (MeOH; CH3OH, EM Science, Darmstadt, Germany). 5. Tris[2-carboxyethyl] phosphine (TCEP; Pierce, Rockford, IL). 6. Barnstead Nanopure water purification system (Barnstead, Dubuque, IA). 7. ZipTips packed with C18 reversed-phase resin Millipore (Billerica, MA, USA). 8. MeOH lysis buffer [50 mM NH4HCO3 with 100% MeOH (v/v 40/60)].

2.3. RP-LC-MS

1. HPLC grade ACN (CH3CN, EM Science, Darmstadt, Germany).

2.4. Computational Support for CollisionInduced Dissociation Spectra Analysis

1. Single computer workstation or a cluster computer that follows a Beowulf design model (see Note 1). 2. Software for protein database search and match to experimental mass spectrometry data (see Note 2). 3. Nonredundant human proteome database. 4. Software for reverse database creation for the assessment of a false-positive rate. 5. Software to analyze the experimental data for biologic classification and implications (see Note 3).

98

D.J. Johann et al.

3. Methods To obtain reliable and reproducible results, it is important to handle the sample rapidly and effectively during the tissue acquisition step. During a surgical procedure, once a solid tumor is ligated from its blood supply, tissue degradation and possibly frank necrosis will eventually commence. Therefore, a few simple, but deliberate steps are recommended to minimize ischemic phenomena. Most succinctly, as soon as possible, the tissue should be snap frozen in liquid nitrogen and then placed in a freezer at −80°C. Subsequently, the tissue specimen will be embedded in cryostatmounting medium (TISSUE-Tek O.C.T.). Tissue sections, usually with a slice thickness range of 8–12 mm, are then serially cut with the cryostat from the frozen tissue block. As a convenient measurement or rule of thumb, the majority of cells will have a diameter either larger or within this range. Therefore, the recommended slice thickness will aid in the homogenization/ lysis procedure, since one of the key steps of the method being presented is effective liberation of proteins from captured cells. While MS is a critical component in any bottom-up proteomic analysis, sample handling factors, such as effective lysis and digestion, are the key requisites for effective large-scale protein identification. Effective digestion of small-size LCM specimens requires optimal buffering conditions. Such conditions maintain the proteins solubilized and denatured throughout the digestion process, without unnecessary manipulations and use of reagents that might interfere with LC-MS analysis. To simplify and improve the analysis of captured cells and avoid the deficiencies associated with traditional approaches (which use detergents and chaotropes), we developed a simple two-step methanol-assisted solubilization/digestion protocol. In the first solubilization step, 20% buffered methanol is used to facilitate denaturation and solubilization of cytosolic proteins. In the second step, the solubilization/digestion is carried out in a 60% methanol buffer, targeting more hydrophobic proteins, which are insoluble in 20% buffered methanol, thus resulting in enhanced proteome coverage. A schematic of this experimental workflow is shown in Fig. 1. 3.1. Initial Pathologic Analysis (Prior to LCM)

A formal H&E with coverslip should be performed with every tenth slide. Then, prior to LCM analysis, these slides should be reviewed with a pathologist in order to properly evaluate the histology, plan LCM sessions, and guard against potential bias in the z-dimension of the tumor tissue plane.

3.2. LCM Staining

The fresh-frozen tissue slide must melt before beginning the staining protocol below. Placing the slide on the palm of your glove works well. As soon as condensate forms on the entire slide,

8  Profiling Solid Tumor Heterogeneity by LCM and Biological¼

99

Fresh-frozen tissue acquisition

Pathology review LCM design

LCM slide prep

LCM

Homogenization Lysis

LCM quality control

MeOH-based twostep tryptic digestion

LC-MS/MS

Data Analysis

Desalting

Fig. 1. LCM-MS experimental design.

the protocol below may commence. In order to ensure good visualization and tissue capture, suggested times are provided for both membrane and glass slides (see Note 4). 1. 70% ethanol, fix tissue section to slide, 15 s (membrane slide), 30 s (glass slide). 2. H2O, remove OCT, rehydrate tissue, 30 s (membrane slide), 30 s (glass slide). 3. Hematoxylin, stain nuclei, 45 s (membrane slide), 30 s (glass slide). 4. H2O, remove excess hematoxylin, 15  s (membrane slide), 30 s (glass slide). 5. Bluing solution, change hematoxylin hue, 15  s (membrane slide), 30 s (glass slide). 6. 70% ethanol, start dehydration, 15 s (membrane slide), 30 s (glass slide). 7. Eosin, stain cytoplasm (1–2 quick dips), 1–2  s (membrane slide), 2 s (glass slide). 8. 95% ethanol, dehydration, 30  s (membrane slide), 1  min (glass slide).

100

D.J. Johann et al.

9. 95% ethanol, dehydration, 30  s (membrane slide), 1  min (glass slide). 10. 100% ethanol, dehydration, 30  s (membrane slide), 2  min (glass slide). 11. 100% ethanol, dehydration, 30  s (membrane slide), 2  min (glass slide). 12. Xylene, ethanol removal, 3  min (membrane slide), 3  min (glass slide). 3.3. LCM Procedure

LCM analysis may begin on the slide(s) once they are air-dried. Laser-based dissection systems allow for dissections approaching 100% purity. Staining with hematoxylin and eosin allows microscopic visualization during microdissection and does not diminish protein recovery. Generally, we have found that depending on the type of tissue under study, approximately 5,000–50,000 cells are required to produce mass spectrometry results with acceptable numbers of protein identifications, and species diversity (see Note 5). Figure  2a–c illustrates a stepwise approach for successful LCM tissue extraction by either a PixCell IIe or Veritas system. LCM tissue extraction involves: 1. Establishing a histology area of interest (Fig. 2a). 2. Manual filling of the pattern to enable the removal of a larger amount of cells (Fig. 2b). 3. LCM extraction of the cells from the selected region (Fig. 2c).

3.4. LCM Membrane Tissue/Cell Extraction and Lysis

The sample preparation protocol for protein extraction/digestion from LCM samples captured on polymer cap is now presented, initially as a brief overview and then with full details. 1. Carefully remove the LCM polymer membrane by peeling it off the cap and then place it in a siliconized tube (conical bottom). 2. Add 50  mL of 12.5  mM hypotonic lysis buffer containing 1 mM TCEP (final concentration). 3. Incubate on dry ice for 30 min. 4. Thaw the sample in ice-cold water for 10 min. 5. Incubate the sample in a water bath for 2 h at 70°C. 6. Cool the sample on ice for 20 min. 7. Adjust buffer to 50  mM by adding 1.65  mL of 1  M NH4HCO3.

8  Profiling Solid Tumor Heterogeneity by LCM and Biological¼

101

Fig. 2. LCM workflow.

3.5. First Trypsin Digestion

Trypsin dilution. The background to this protocol is based on single-cell protein content estimates that are in the range of 0.75 pg–0.5 ng (10) (see Note 6). Prepare the dehydrated and frozen trypsin, e.g., Promega Trypsin Gold, 20-mg vial. Mix with 20 mL 50 mM NH4HCO3, yielding a concentration of 1 mg/1 mL. The availability of some tissue samples is quite limited in quantity. Therefore, this section attempts to accommodate these circumstances as well as situations with more abundant tumor tissues. 1. Rehydrate trypsin by adding 20 mL of 50 mM NH4HCO3. 2. Dilute trypsin in accordance with sample cell count per the trypsin dilution protocol (Table 1). 3. Add the appropriate volume of trypsin solution and mix for 10 min. 4. Briefly vortex the sample.

102

D.J. Johann et al.

Table 1 Recommended amount of trypsin as a function of sample cell count Cell count

Protein estimate, mg

Trypsin:protein

Trypsin for sample, mg

50

0.025

1:50

0.0005

500

0.25

1:50

0.005

5,000

2.5

1:50

0.05

15,000

7.5

1:50

0.15

1:50

0.5

50,000

25

5. Place the sample in the water bath sonicator for 5 min. 6. Transfer the sample tube to a small centrifuge and spin for ~15 s. 7. Incubate the sample digest for 6 h at 37°C with good table motion. 3.6. Second Trypsin Digestion

1. Add 60% MeOH lysis buffer. This can be achieved by adding 50 mL of MeOH to each sample. 2. Add the appropriate volume of trypsin solution and mix for 10 min. 3. Briefly (~5 s) vortex the sample. 4. Place the sample in the water bath sonicator for 5 min. 5. Transfer the sample tube to a small centrifuge and spin for ~15 s. 6. Incubate the sample digest for 6 h at 37°C with good table motion. 7. Lyophilize all samples to dryness.

3.7. Desalting Using ZipTip Columns

1. Rehydrate peptides in 20 mL 0.1% TFA by sonication in water bath for 2 min. 2. Prepare 10-mL aliquots of elution buffer 60% ACN/0.1% TFA (v/v) for each sample before beginning (to avoid contamination). Avoid drawing air through the tip during the procedure (from equilibration to elution). If you find that you make bubbles in the tip, try pulling the buffers in more slowly. 3. Set the Pipetman to 10 mL and attach the ZipTip. 4. Activate the ZipTip column by pipetting up 20  mL of 60% ACN and then discard it in the waste. Repeat it three times. 5. Equilibrate the ZipTip column by pipetting up 20 mL of 0.1% TFA and then discard it in the waste. Repeat it three times. These steps act as a gradient for the minicolumn, which activates the resin and conditions it to bind peptides.

8  Profiling Solid Tumor Heterogeneity by LCM and Biological¼

103

6. Load the peptides by pipetting the sample up and down (discarding it back into its tube). Repeat it ten times. 7. Wash the ZipTip column using the 0.1 TFA. Pipette up the buffer and then discard it in the waste. Repeat it ten times. 8. Elute the sample by pipetting the ZipTip up and down in the elution buffer [back into its tube in the 4 mL 60% acetonitrile/0.1% TFA (already aliquoted)]. Repeat it ten times. The organic phase elutes the peptides off the resin into the buffer. Now, the sample is desalted as well as concentrated. 9. Lyophilize to dryness and dissolve the peptides in 10 mL of 0.1% TFA prior to LC-MS/MS analysis. 3.8. Guidelines for LC-MS Analysis of LCM Samples

Although there are a wide variety of mass spectrometer systems and liquid chromatography platforms, the linear ion trap coupled with a reversed-phase liquid chromatography separation system is widely used in the proteomics community, and will be illustrated in this section. For our LCM-based proteomic studies, a reversed-phase column is coupled with a linear ion trap mass spectrometer (LTQ ThermoElectron, San Jose, CA) for shotgun proteomic analysis. 1. In our configuration, the solvent system is delivered by an HP 1100 pump (Agilent Technologies, Palo Alto, CA). 2. A nanoelectrospray ionization source is employed applying a voltage of 1.7 kV and a capillary temperature of 160°C. 3. The LTQ is operated in a data-dependent mode. 4. The seven most abundant peptide molecular ions detected by each MS survey scan are dynamically selected for MS/MS using collision-induced dissociation (CID) facilitated by a normalized collision energy of 35%. 5. Dynamic exclusion is employed to avoid redundant acquisition of precursor ions previously selected for fragmentation. 6. Reversed-phase liquid chromatography separations are performed with a 75-mm i.d. × 10-cm long fused silica capillary column (Polymicro Technologies, Inc., Phoenix, AZ) with a flame-pulled tip (~5–7 mm orifice). 7. The column is slurry packed in-house with 5 mm, 300 Å pore size C-18 stationary phase (Vydac, Hercules, CA) using a slurrypacking pump (model 1666, Alltech Associates, Deerfield, IL). 8. Note: The total MS run time for each sample is 180 min. (a) After injecting 5 mL of sample, the column is washed for 30 min with 98% mobile phase A (0.1% FA in H2O). (b) Peptides are then eluted using a linear step gradient from 2 to 40% mobile phase B (0.1% FA in ACN) over 90 min. (c) Then, an elution gradient of 60–98% for mobile phase B over 10 min at a constant flow rate of 0.25  mL/min is performed.

104

D.J. Johann et al.

(d) Next, the column is washed for 20 min with 98% mobile phase B. (e) Finally, the column is re-equilibrated with 2% mobile phase B for 30  min prior to subsequent loading of the next sample. 3.9. Data Processing Guidelines

As previously stated, the searching and matching of experimentally obtained spectra against a nonredundant protein database are computationally intensive, but highly parallelizable, and therefore amendable to divide and conquer strategies employing cluster computers. 1. For our LTQ-derived data, the precursor ion tolerance is set to 1.5 Da and the fragment ion tolerance to 0.5 Da. These two values effectively serve as binning parameters during data acquisition concerning parent and child (fragment) ions. 2. We require candidate peptides to possess tryptic terminus at both ends, and generally will allow for a maximum of two missed tryptic cleavages. 3. The following SEQUEST thresholds are routinely used to filter experimental peptides: (a) Delta-correlation score (dCn) ³ 0.08 (b) Charge state cross-correlation scores as follows: ³2.1 for [M + H]1+ peptides ³2.3 for [M + H]2+ peptides ³3.5 for [M + H]3+ peptides 4. The final list of protein identifications is created using a parsimony principle, reporting a minimal number of protein identifications from a pool of uniquely identified peptides. 5. Resultant raw data are routinely subjected to a false-positive rate assessment via decoy (reverse) database analysis. 6. Lastly, data are analyzed for biologic implications by Ingenuity Pathway Analysis (IPA) and the Database for Annotation, Visualization, and Integrated Discovery (DAVID).

4. Notes 1. The processing of CID spectra is computationally intensive, but highly parallelizable, and follows a classic divide and conquer paradigm. Therefore, a cluster computer solution generally offers substantial time savings that follows a linear function, depending on the number of computational elements in the cluster configuration.

8  Profiling Solid Tumor Heterogeneity by LCM and Biological¼

105

2. Commercial products include MASCOT (Matrix Science, http://www.matrixscience.com) and SEQUEST (Thermo Scientific, http://www.thermo.com). Open source solutions include (1) the X!Tandem database search engine (http://www.thegpm.org/TANDEM/index.html), (2) the Trans Proteomic Pipeline (TPP, http://tools.proteomecenter. org/wiki/index.php?title=Software:TPP), and (3) the Open Mass Spectrometry Search Algorithm (OMSSA), http://pubchem.ncbi.nlm.nih.gov/omssa/. 3. Commercial products include IPA, http://www.ingenuity. com). Public domain tools include the DAVID, http://david. abcc.ncifcrf.gov). 4. For each step in the staining protocol, a different solution bath is recommended. Through experience, this has been found to make a significant difference. 5. We have found that tissues with a compact cellular density provide greater protein yields, and thus usually require a smaller quantity of cells. However, when encountering a new tumor tissue type, a few preliminary experiments are recommended for a general estimate of protein yield. 6. In addition to the provided reference, these estimates are also cited in the following text books: Molecular Biology of the Cell, 3rd Edition by Alberts et al. and Molecular Cell Biology, 4th Edition by Lodish et al.

Acknowledgments This project was funded in whole or in part with federal funds from the National Cancer Institute, National Institutes of Health, under contract HSN261200800001E. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does the mention of trade names, commercial products, or organizations implies endorsement by the US Government. References 1. Mbeunkui F, Johann DJ, Jr., (2009) Cancer and the tumor microenvironment: a review of an essential relationship. Cancer Chemother Pharmacol 63: 571–82. 2. Swanton C, Caldas C, (2009) Molecular classification of solid tumours: towards pathwaydriven therapeutics. Br J Cancer100: 1517–22.

3. Johann DJ, Jr., Blonder J, (2007) Biomarker discovery: tissues versus fluids versus both. Expert Rev Mol Diagn 7: 473–5. 4. Johann DJ, Wei BR, Prieto DA, Chan KC, Ye X, Valera VA, Simpson RM, Rudnick PA, Xiao Z, Issaq HJ, Linehan WM, Stein SE, Veenstra TD, Blonder J. Combined Blood/ Tissue Analysis for Cancer Biomarker

106

D.J. Johann et al.

Discovery: Application to Renal Cell Carcinoma. Anal Chem 2010. 5. Emmert-Buck MR, Bonner RF, Smith PD, Chuaqui RF, Zhuang Z, Goldstein SR, Weiss RA, Liotta LA. (1996) Laser capture microdissection. Science 274: 998–1001. 6. Aebersold R, Mann M, (2003) Mass spectrometry-based proteomics. Nature 422: 198–207. 7. Hwang SI, Thumar J, Lundgren DH, Rezaul K, Mayya V, Wu L, Eng J, Wright ME, Han DK. (2007) Direct cancer tissue proteomics: a method to identify candidate cancer biomarkers from formalin-fixed paraffinembedded archival tissues. Oncogene 26: 65–76.

8. Blonder J, Chan KC, Issaq HJ, Veenstra TD, (2006) Identification of membrane proteins from mammalian cell/tissue using methanol-facilitated solubilization and tryptic digestion coupled with 2D-LC-MS/MS. Nat Protoc 1: 2784–90. 9. Johann DJ, Rodriguez-Canales J, Mukherjee S, Prieto DA, Hanson JC, Emmert-Buck M, Blonder J. (2009) Approaching solid tumor heterogeneity on a cellular basis by tissue proteomics using laser capture microdissection and biological mass spectrometry. J Proteome Res 8: 2310–8. 10. Wibke H, Pelargus C, Keffhalm K, Ros A, Anselmetti D. (2005) Single cell manipulation, analytics, and label-free protein detection in microfluidic devices for systems nanobiology. Electrophoresis 26: 3689–96.

Chapter 9 Amplification Testing in Breast Cancer by Multiplex Ligation-Dependent Probe Amplification of Microdissected Tissue Cathy B. Moelans, Roel A. de Weger, and Paul J. van Diest Abstract This chapter describes a method for the rapid assessment of gene copy numbers in laser-microdissected materials using multiplex ligation-dependent probe amplification (MLPA). An MLPA is a powerful multiplex PCR technique that can identify gains, amplification, or losses of up to 50 genes in a single experiment, thereby requiring only minute quantities of DNA extracted from frozen or paraffin-embedded materials. A previous study in breast cancer has shown that MLPA can detect amplifications in cases with a tumor percentage lower than 10%, but still a low tumor percentage in the tissue tested could obscure low levels of amplification due to dilution of the tumor cell population by normal cells. Laser capture microdissection allows enrichment of tumor cells by eliminating background noise from normal and preinvasive cells, thereby increasing specificity and sensitivity. This chapter describes a method for MLPA analysis using invasive breast tumor cells acquired by laser capture microdissection. This protocol can also be applied to MLPA analysis of preinvasive lesions and metastases. Key words: Multiplex ligation-dependent probe amplification, MLPA, Coffalyser, Laser microdissection, Cancer

1. Introduction Several genes have been shown to be involved in the development, progression, and response to therapy of invasive breast cancer. Among these, HER-2/neu is likely the most important proto-oncogene. Amplification of the HER2 gene is present in about 15–30% of breast carcinomas and leads to protein overexpression (1, 2). Patients having this overexpression have an overall worse prognosis (3), but respond well to the treatment with

Graeme I. Murray (ed.), Laser Capture Microdissection: Methods and Protocols, Methods in Molecular Biology, vol. 755, DOI 10.1007/978-1-61779-163-5_9, © Springer Science+Business Media, LLC 2011

107

108

C.B. Moelans et al.

trastuzumab, a recombinant humanized monoclonal anti-HER2 antibody (4, 5). Furthermore, amplification of HER2 has also been shown to be associated with resistance to conventional adjuvant chemotherapy and tamoxifen (6, 7). HER2-targeted therapy in breast cancer can aid approximately 20% of women with HER2 overexpression, but no single gene copy number assessment seems to completely explain prognosis or response to therapy of individual breast cancer patients. A simultaneous analysis of copy number changes of a variety of genes involved in prognosis and therapy response may thus be very useful for molecular profiling of individual breast cancer patients. This can be achieved by an easy-to-perform, high-throughput PCR-based technique, called multiplex ligation-dependent probe amplification (MLPA). The MLPA technique was first described in 2002 by Schouten et al. (8) and is summarized in Fig. 1. Up to 50 probe sets can be run in one reaction. MLPA has been used to assess gene copy number changes (9–11), gene expression (12, 13), and methylation (14–16). Due to the short lengths of the target sequences of hemiprobes, MLPA can not only be applied to DNA isolated from fresh-frozen materials, but is also suitable for the more fragmented DNA from paraffin-embedded materials. Depending on the quality of the DNA, 50–200 ng of DNA suffice. The ability to carry out a multiplex copy number assessment on small amounts of paraffin-embedded materials makes MLPA a very attractive method in pathology. In previous studies using whole tissue sections, we obtained very promising results with MLPA for HER2 amplification detection in comparison with immunohistochemistry (IHC) (17), fluorescence in situ hybridization (FISH), and chromogenic in situ hybridization (CISH) (18). However, the dynamic range of MLPA copy number ratios was lower than that with FISH. Furthermore, although results showed that amplification could be detected in cases with a tumor percentage lower than 10%, the sensitivity of MLPA in these cases depend on the degree of amplification, so lower levels of amplifications can be missed in case of a low tumor percentage. Obviously, MLPA is a nonmorphological method that requires proper morphological control of the input materials. In cases with a low percentage of relevant materials, laser microdissection may be necessary (19).

2. Materials 2.1. Tissue Cutting, H&E, and Microdissection

1. Uncoated glass slides (Thermo Scientific, Menzel-Gläser, Germany). 2. Hematoxylin solution (Klinipath, Duiven, The Netherlands). 3. Eosin yellow solution (Klinipath). 4. Xylene.

9  Amplification Testing in Breast Cancer by MLPA

109

Fig. 1. Principle of multiplex ligation-dependent probe amplification (MLPA). MLPA uses a mixture of hemi-probe sets that consist of two oligonucleotides, both having PCR primer sequences (x /y) on the outer ends and a sequence complementary to a part of the target sequence on the inner ends. One of the primers has a spacer (stuffer sequence) of variable length in between the PCR primer sequence and the complementary target sequence. When the complementary target sequences of both hemi-probes hybridize adjacent to each other on the target sequence, they can be ligated to each other, and subsequently amplified using the PCR primer sequences. Because the PCR primers are the same for all hemi-probe sets, they can be amplified in a single PCR, which will provide amplicons of unique and defined lengths (120 – 500  bp) due to the specific stuffer length within each probe set.

5. Pertex-mounting medium (Histolab products AB, Göteborg, Sweden). 6. PALM Liquid Cover Glass N (P.A.L.M. Microlaser Techn., Bernried, Germany): Dilute 2 ml stock resin in 10 ml dilution buffer (1/6) (see Note 1). 7. MembraneSlide 1.0 PEN (Carl Zeiss, Munchen, Germany; 1 mm). 8. 70, 85, and 100% ethanol. 9. PALM Laser Microbeam System (P.A.L.M. AG, Bernried, Germany).

110

C.B. Moelans et al.

2.2. DNA Isolation

1. Proteinase K: 10 mg/ml. Dissolve 1 g proteinase K (Sigma Aldrich Chemie BV) in 100 ml distilled water, and store 1-ml aliquot at −20°C. 2. Lysis buffer: 50  mM Tris–HCl buffer, pH 8.0 with 0.5% Tween 20. Dissolve 0.61 g Tris (Roche, 708976) in 80 ml Milli-Q water, add 20 ml Milli-Q water, and adjust pH to 8.0 with HCl (37%). Add Tween 20 (Riedel de Haën) until final concentration of 0.5%.

2.3. Multiplex Ligation-Dependent Probe Amplification

1. MLPA kit (P004-B1 for HER2, MRC-Holland, Amsterdam, The Netherlands) (see Note 2). 2. TE buffer: Dissolve 1.21 g Tris (10 mM) and 0.37 g sodium– EDTA (1  mM) in 900  ml distilled water, adjust the pH between 7.5 and 8.0 with HCl, and add the other 100  ml distilled water. 3. Labeled size standard (Applied Biosystems, Foster City, CA). 4. Deionized formamide (Amresco, OH) (see Note 3). 5. 10× EDTA buffer (Applied Biosystems). 6. Performance-optimized polymer (POP, Applied Biosystems). 7. Thermocycler and capillary sequencer. 8. Fragment analysis software (Genescan, Applied Biosystems). 9. Coffalyser MLPA analysis software (freeware at http://www. mrc-holland.nl).

3. Methods The preparation of samples, laser microdissection, and the optimum reaction conditions for MLPA is described. A set of guidelines specific for software analysis of MLPA assays is included. 3.1. Preparation of Samples and Laser Microdissection

1. Cut 2–3 mm slides (uncoated) and perform a routine H&E staining for each tissue block to be microdissected. Mark the appropriate area to be microdissected. 2. Mount 8–10  mm thick sections on sequential PALM MembraneSlides (see Note 4). 3. Bake these sections at 56°C for 1 h, deparaffinize in xylene for 10 min, and rehydrate through graded alcohols (100, 85, and 70% for 1 min each) (see Note 5). 4. Stain slides with hematoxylin solution for 10 s, rinse in tap water for 1 min, and dip briefly in eosin solution (see Note 6). 5. Dehydrate in 100% ethanol for 1 min and air dry.

9  Amplification Testing in Breast Cancer by MLPA

111

6. To improve morphology and allow larger tissue areas to be laser pressure-catapulted (20), apply PALM Liquid Cover Glass by aerosol and air dry sections for at least 15 min. 7. A microdissection system with UV laser separates the marked invasive tumor groups from their surrounding tissues. Subsequently, these groups are catapulted by laser pressure catapulting into a cap of a common microfuge tube moistened with a drop of paraffin oil (see Note 7). 3.2. DNA Extraction from LaserMicrodissected Samples

1. Add 50 ml direct lysis buffer to the tube containing the microdissected material (see Note 8). 2. Add 10 ml proteinase K (10 mg/ml), incubate at 56°C for at least 1 h, and then boil (heat inactivation) for 10 min. Cool the lysate immediately on ice (see Note 9). 3. Centrifuge for 2 min at room temperature at 20,800 rcf and carefully pipet the DNA-containing supernatant into a clean tube (see Note 10).

3.3. MLPA Assay

1. Dilute the DNA sample (preferred range 50–200  ng) with TE (10 mM Tris–HCl pH 8.2; 1 mM EDTA) to 5 ml and add to PCR tubes. Take appropriate positive/negative controls along with your samples (see Notes 11 and 12). 2. Heat for 5 min at 98°C and cool to 25°C before opening the thermocycler. 3. Add a mixture of 1.5 ml SALSA Probemix (black cap) + 1.5 ml MLPA buffer (yellow cap) to each tube (see Note 13). 4. Mix with care. Incubate for 1 min at 95°C (denaturation), followed by 16  h at 60°C (hybridization overnight) (see Note 14). 5. Prepare the Ligase-65 mix (less than 1 h before use and stored on ice): Add 3 ml Ligase-65 buffer A (transparent cap) to 3 ml Ligase-65 buffer (white cap) and 25  ml MilliQ water, mix, add 1  ml Ligase-65 thermostable enzyme (green cap), and mix again. 6. Reduce the temperature of the thermocycler to 54°C. While at this temperature, add 32 ml Ligase mix to each sample and mix well. 7. Incubate for 15  min at 54°C, and then heat for 5  min at 98°C. Place the ligation product on ice (see Note 15). 8. Mix in new PCR tubes: 4  ml SALSA PCR buffer (red cap) + 26 ml MilliQ water + 10 ml MLPA ligation reaction. 9. Prepare polymerase mix (less than 1 h before use and stored on ice): Mix 2 ml SALSA PCR primers (brown cap) with 2 ml SALSA enzyme dilution buffer and 5.5  ml MilliQ water. Add 0.5  ml SALSA polymerase (orange cap). Mix well (see Note 16).

112

C.B. Moelans et al.

Table 1 Multiplex ligation-dependent probe amplification (MLPA) PCR conditions Hybridization reaction

98°C 25°C 95°C 60°C

5 min Hold 1 min Hold

Ligation reaction

54°C 54°C 98°C 4°C

Hold 15 min 5 min Hold

PCR

60°C 95°Ca 60°Ca 72°Ca 72°C 4°C

Hold 30 s 30 s 60 s 20 min Hold

35 cycles

a

10. While the tubes are in the thermocycler at 60°C, add 10 ml polymerase mix to each tube and immediately start the PCR. PCR conditions are 35 cycles of 30 s at 95°C, 30 s at 60°C followed by 60 s at 72°C. The PCR ends with 20 min incubation at 72°C (Table 1; see Note 17). 11. Separate the amplification products by a capillary sequencing system with fragment analysis software. In our case (ABI310), we mix 22 ml deionized formamide, 0.75 ml ROX-500, and 3 ml PCR product, denature at 80°C for 5 min, and keep the plate on ice until the start of the run (see Note 18). 3.4. Software Analysis and Interpretation of MLPA Data

1. To analyze MLPA data, we recommend using Coffalyser software developed by the manufacturer of MLPA kits, especially for large numbers of samples (see Note 19). 2. Open Coffalyser, select the probemix you used (for example, P004-B1), and import your raw fragment analysis files (see Notes 20 and 21). 3. Navigate to the “Data filtering” page, where you can filter your data with the automatically generated default bin set that needs to be checked. If necessary, adjust the set bin lengths (that Coffalyser determines based on your selected reference samples) and use these as default. Start filtering the reference and sample data.

9  Amplification Testing in Breast Cancer by MLPA

113

4. Navigate to the “Data analysis” page. A short summary of all filtered data is given, including name, DNA concentration test, number of signals/probes found, number of control probes found, ligation-dependent probes found, and sex determination (M/F). Reference runs should always contain all signals! Delete reference runs that do not contain all signals (see Note 22). 5. Choose an appropriate analysis method, for example “Tumor analysis.” In this case, Coffalyser uses all probes in the selected MLPA kit for slope correction (corrects for signal strength drop with increasing length of the probes), but normalizes the data using only the reference probes present in the kit (located in stable genomic areas) (see Note 23). 6. Analyze and explore results by navigating to “MLPA results.” You will be able to look at your reference results and sample results in different sheets. The gene copy number status is marked with a color. If the MLPA ratio is >1.3 (gain), the color is green. If the ratio is 2.0 as HER2amplified (see Note 24).

4. Notes 1. PALM Liquid Cover Glass, when diluted 1/6, can be kept at 4°C for 2 weeks. 2. In case MRC-Holland cannot offer an MLPA probemix for your application, you can design your own synthetic MLPA probes according to a protocol provided by the manufacturer. Synthetic probes differ from MRC-Holland probes in that the latter consist of one synthetic oligonucleotide and a clonederived one. This allows to make longer probes and to include up to 50 different probes in one MLPA reaction. As an extra control, MLPA kits often contain more than one probe per gene. 3. Currently, most capillary sequencers no longer use deionized formamide. In these cases, formamide is replaced by distilled water.

114

C.B. Moelans et al.

4. To overcome the hydrophobic nature of the membrane, it is advisable to irradiate with UV light at 254 nm for 20–30 min. The membrane gets more hydrophilic, therefore the sections obtain a better adherence. We did not observe any negative effects on MLPA results when using UV irradiation. When cutting the sections for microdissection and subsequent MLPA analysis, it is preferred not to have 0.2% bovine serum albumin (BSA) in the water bath, often used for releasing surface tension, since it can influence DNA quality and therefore MLPA results. 5. Be very careful during xylene and alcohol steps, as the section and membrane can be very fragile. 6. Hematoxylin stain can considerably inhibit PCR amplification (23). The dye may bind to DNA, and subsequently interfere with proteinase digestion, or it might influence divalent cation (Mg++) concentration that is important in maintaining polymerase activity. No deleterious effects have been described for eosin. We, therefore, advise to keep staining times as short as possible, especially for hematoxylin. 7. If DNA isolation is not performed immediately after laser microdissection, microfuge tubes can be kept at 4°C for 2 weeks, perhaps even longer (not tested). 8. Since the microdissected material is in the tip of the tube, it is advisable to add the lysis buffer to the tip and then tick hard so that the material drops to the bottom of the tube together with the lysis buffer. Another option is to use a pipette tip to move the laser-microdissected tissue to the bottom of the tube (or a new tube). We used 50 ml of lysis buffer, but it is also possible to use less or more depending on the amount of tissue you were able to microdissect. 9. If necessary, the incubation time can be increased to an overnight step; in most cases, this leads to a higher DNA concentration. Always use a lysis buffer/proteinase K proportion of 5/1. 10. The extracted DNA is tenable for 2–3 weeks at 4–10°C. This easy, fast, and cheap DNA isolation method leads to lower quality and fragmented DNA, but is sufficient for MLPA analysis due to its very short hybridizing probe sequences. We have tried different column precipitation isolation methods that also work well and generally lead to a higher DNA quality, but at the same time to a lower DNA quantity. 11. The EDTA concentration in the DNA sample should not exceed 1 mM, and the sample volume should not exceed 5 ml. The volume of the reaction is important for the hybridization speed, which is probe- and salt concentration-dependent. If the DNA concentration of a sample is very low, add 5 ml of

9  Amplification Testing in Breast Cancer by MLPA

115

DNA without TE buffer. We usually perform every sample in duplicate as an extra control. 12. Negative (and optional positive) control samples should be run simultaneously with the test samples. It is highly recommended to compare reference and tumor samples extracted by the same method and derived from the same source (e.g., blood versus blood). Use at least three reference samples in each MLPA run (we used four negative control samples per 21 samples). When using more than 21 samples, add one additional reference sample for each seven samples. Reference samples should be spread randomly over the sample plate to avoid bias, and thus minimize variation. It is also recommended to include a no-DNA (5 ml water or TE) reaction in each experiment, as it reveals contamination of water, MLPA reagents, electrophoresis reagents, or capillaries. 13. The SALSA MLPA buffer is usually frozen at −20°C, so thawing is necessary before use. Furthermore, the MLPA buffer is viscous and does not mix easily. Mix probemix and MLPA buffer by repeated pipetting just before use. 14. The hybridization time can be anywhere between 12 and 24 h (16–18 h is recommended, but hybridization should be nearly complete at 12 h). Be sure that there is no excessive evaporation, if so try a different brand of tubes or try using mineral oil on top of the ligation–PCR mix. 15. Following ligation inactivation at 98°C, samples can be stored at 4°C for up to 1 week. For longer periods, storage at −20°C is recommended. 16. Start the PCR as soon as possible after the addition of polymerase mix. 17. All volumes of the PCR can be reduced in order to save reagents. The recommended number of PCR cycles is 35; however, the number of cycles can be reduced to 30 and, in the case of small DNA amounts, the number of cycles can be increased up to 37. The sequence (5¢-3¢) of the labeled forward PCR primer is GGGTTCCCTAAGGGTTGGA and of the unlabeled reverse primer GTGCCAGCAAGATCC AATCTAGA. The PCR product can be stored in the dark at 4°C for at least 1 week. 18. The amount of the PCR product required for analysis by capillary electrophoresis depends on the instrument and fluorescent label used. The settings can be found at the MRCHolland Web site in the support section under technical MLPA protocols. 19. Coffalyser is an excel-based program, which runs on a Microsoft Office version 2003 or higher. Data normalization and correction for probe length-dependent decrease in peak area/height are built-in functions of this program.

116

C.B. Moelans et al.

20. Many more MLPA mixes are commercially available. There are kits for hereditary cancers, tumor characterization, genetic disorders, mental retardation/neurogenetics, pre- and postnatal defects, pharmacogenetics, methylation analysis, and mRNA analysis. 21. Manuals on how to create files for import can be found at the size calling and export manuals on the MRC-Holland Web site. ABIF (FSA) files coming from the ABI-310 and ABI3100 can directly be imported and will be size called during import. TXT files are also supported and can be exported from Genescan, Genemapper, Peak scanner software, LICOR software, Megabace software, and the Spectrophotometrix. After import, the electropherograms from raw runs can be visualized. 22. Coffalyser calculates the ratios of all analyzed sample runs. Several quality check points are also displayed: the number of found probes; the number of found reference probes; whether the ligation control peak (92  bp) was found; whether the sample was male or female (if a Y probe was present); whether there was enough DNA [by estimation of the relative signal of Q-fragments (60, 68, 74, and 80  bp) present in every MLPA mix]; whether the DNA was denatured completely [relative signals of denaturation fragments (88 and 96  bp) present in every kit]; the Pearson product moment correlation (PPMC) value (the correlation value found when slope correction is applied); and the median of absolute deviation (MAD) value (a measure for the stability of reference probes). When the MAD value becomes red, you may need to try a different normalization factor or adjust the reference probes for a more optimized normalization. 23. To best meet the requirements of the context of your experimental design, a number of pre-made settings are available (tumor analysis, direct analysis, control probe analysis, population analysis, methylation status analysis). Depending on expected aberrations, availability of reference probes and reference runs, and quality of the runs, a method needs to be chosen by the user. The manufacturer provides a guideline. 24. Probe signals for DNA sequences can be very high when the DNA in question is amplified at very high levels (e.g., in the case of HER2, EGFR, etc), causing other probe signals to be dwarfed. Probe signals can be reduced by the inclusion of a competitor oligo in the probemix. This competitor is identical to the left probe oligo (LPO) and to a small part (four nucleotides: TGGA) of the PCR primer sequence. The competitor competes with the LPO for the limiting number of binding sites on the DNA. It can be ligated to the right probe

9  Amplification Testing in Breast Cancer by MLPA

117

oligo (RPO), but the resulting comp-RPO ligation product cannot be amplified exponentially as the probe thus formed does not contain both primer sequences. The use of a 1:1 ratio of LPO and its corresponding competitor reduces the probe signal twofold. References 1. Ross JS, Fletcher JA, Bloom KJ et  al (2004) Targeted therapy in breast cancer: the HER-2/ neu gene and protein. Mol Cell Proteomics 3, 379–398. 2. Slamon DJ, Godolphin W, Jones LA et  al (1989) Studies of the HER-2/neu protooncogene in human breast and ovarian cancer. Science 244, 707–712. 3. Baak JPA, Chin D, Van Diest PJ et al (1991) Comparative long term prognostic value of quantitative Her2/Neu protein expression, DNA ploidy, morphometric and clinical features in paraffin-embedded invasive breast cancer. Lab Invest 64, 215–222. 4. Slamon DJ, Leyland-Jones B, Shak S et  al (2001) Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N Engl J Med 344, 783–792. 5. Hudis CA (2007) Trastuzumab, mechanism of action and use in clinical practice. N Engl J Med 357, 39–51. 6. Borg A, Baldetorp B, Ferno M et al (1994) ERBB2 amplification is associated with tamoxifen resistance in steroid-receptor positive breast cancer. Cancer Lett 81, 137–144. 7. Tetu B, Brisson J, Plante V et  al (1998) p53 and c-erbB-2 as markers of resistance to adjuvant chemotherapy in breast cancer. Mod Pathol 11, 823–830. 8. Schouten JP, McElgunn CJ, Waaijer R et  al (2002) Relative quantification of 40 nucleic acid sequences by multiplex ligation-dependent probe amplification. Nucleic Acids Res 30, e57. 9. Moelans CB, de Weger RA, van Blokland MT et al (2010) Simultaneous detection of TOP2A and HER2 gene amplification by multiplex ligation-dependent probe amplification in breast cancer. Mod Pathol 23, 62–70. 10. Moelans CB, de Weger RA, and van Diest PJ (2010) Absence of chromosome 17 polysomy in breast cancer: analysis by CEP17 chromogenic in situ hybridization and multiplex ligationdependent probe amplification. Breast Cancer Res Treat 120, 1–7.

11. Vorstman JA, Jalali GR, Rappaport EF et  al (2006) MLPA: a rapid, reliable, and sensitive method for detection and analysis of abnormalities of 22q. Hum Mutat 27, 814–821. 12. Eldering E, Spek CA, Aberson HL et al (2003) Expression profiling via novel multiplex assay allows rapid assessment of gene regulation in defined signalling pathways. Nucleic Acids Res 31, e153. 13. Hess CJ, Denkers F, Ossenkoppele GJ et  al (2004) Gene expression profiling of minimal residual disease in acute myeloid leukaemia by novel multiplex-PCR-based method. Leukemia 18, 1981–1988. 14. Dikow N, Nygren AO, Schouten JP et  al (2007) Quantification of the methylation status of the PWS/AS imprinted region: comparison of two approaches based on bisulfite sequencing and methylation-sensitive MLPA. Mol Cell Probes 21, 208–215. 15. Nygren AO, Ameziane N, Duarte HM et  al (2005) Methylation-specific MLPA (MS-MLPA): simultaneous detection of CpG methylation and copy number changes of up to 40 sequences. Nucleic Acids Res 33, e128. 16. Procter M, Chou LS, Tang W et  al (2006) Molecular diagnosis of Prader-Willi and Angelman syndromes by methylation-specific melting analysis and methylation-specific multiplex ligation-dependent probe amplification. Clin Chem 52, 1276–1283. 17. Purnomosari D, Aryandono T, Setiaji K et  al (2006) Comparison of multiplex ligation dependent probe amplification to immunohistochemistry for assessing HER-2/neu amplification in invasive breast cancer. Biotech Histochem 81, 79–85. 18. Moelans CB, de Weger RA, van Blokland MT et al (2009) HER-2/neu amplification testing in breast cancer by multiplex ligation-dependent probe amplification in comparison with immunohistochemistry and in situ hybridization. Cell Oncol 31, 1–10. 19. Moelans CB, de Weger RA, Ezendam C et al (2009) HER-2/neu amplification testing in breast cancer by Multiplex Ligation-dependent

118

C.B. Moelans et al.

Probe Amplification: influence of manual- and laser microdissection. BMC Cancer 9, 4. 20. Micke P, Bjornsen T, Scheidl S et al (2004) A fluid cover medium provides superior morphology and preserves RNA integrity in tissue sections for laser microdissection and pressure catapulting. J Pathol 202, 130–138. 21. Coffa J, van de Wiel MA, Diosdado B et  al (2008) MLPAnalyzer: data analysis tool for

reliable automated normalization of MLPA fragment data. Cell Oncol 30, 323–335. 22. Bunyan DJ, Eccles DM, Sillibourne J et  al (2004) Dosage analysis of cancer predisposition genes by multiplex ligation-dependent probe amplification. Br J Cancer 91, 1155–1159. 23. Murase T, Inagaki H, and Eimoto T (2000) Influence of histochemical and immunohistochemical stains on polymerase chain reaction. Mod Pathol 13, 147–151.

Chapter 10 Detection and Quantification of MicroRNAs in Laser-Microdissected Formalin-Fixed Paraffin-Embedded Breast Cancer Tissues Sarkawt M. Khoshnaw, Des G. Powe, Ian O. Ellis, and Andrew R. Green Abstract MicroRNAs (miRNAs) are a class of small endogenous non-coding RNAs that regulate gene expression post-transcriptionally through targeting protein-coding mRNAs for cleavage or translational repression, and thus play key roles in cellular fate-determinant pathways. Both profiling and functional studies demonstrated derangement of miRNA repertoire in many human cancers, including breast tumours. Discovery of miRNAs provided new insights into cancer pathogenesis and led the scientific community to approach novel diagnostic and therapeutic strategies in cancer management. Research in this field is increasing, and the potential for miRNAs being used in clinical settings emphasises the need for high-throughput and sensitive detection techniques. In this chapter, techniques for the analysis of miRNA expression in lasermicrodissected formalin-fixed paraffin-embedded breast cancer tissues are discussed. Key words: Breast cancer, MicroRNA, Laser capture microdissection, Formalin-fixed paraffinembedded, FFPE

1. Introduction 1.1. MicroRNAs’ Biogenesis and Functions

MicroRNAs (miRNAs) are a recently described class of nonprotein-coding RNA molecules (18–25 nucleotides (nt) in length) involved in critical cellular regulatory pathways (1, 2). Mature miRNAs are produced in the nucleus as primary (pri-) miRNAs transcribed by RNA polymerase II (3). Pri-miRNAs, which range between several hundred to a thousand nucleotides in length, are processed inside the nucleus to shorter (70–85  nt) precursor (pre-) miRNAs mediated by RNase III enzyme complex Drosha/ DGCR8 (4). Pre-miRNAs are exported to the cytoplasm (5) and

Graeme I. Murray (ed.), Laser Capture Microdissection: Methods and Protocols, Methods in Molecular Biology, vol. 755, DOI 10.1007/978-1-61779-163-5_10, © Springer Science+Business Media, LLC 2011

119

120

S.M. Khoshnaw et al.

cleaved by Dicer, a second RNase III enzyme, to produce a ~22 nt temporary miRNA duplex made up of a mature miRNA sequence and its complementary sequence (6). The strand with less stable hydrogen bonding at its 5¢ end is incorporated into the RNAinduced silencing complex (RISC) to form the mature miRNA, and the other strand is degraded (7). The main subunit of RISC is the Argonaute2 (Ago2) protein, which is the catalytic endonuclease of human RISC (8). It is thought that miRNA genes constitute more than 1% of human genome, (9) and approximately one third of human genes may be targets of miRNAs (10). Down-regulation of gene expression by miRNAs involves both mRNA degradation and translational repression which are complex processes occurring through multiple mechanisms, such as (1) miRNA (one of the components of RISC complex) can combine with the 3¢UTR of the mRNA, which requires imperfect complementarity only, resulting in translational repression (11). (2) miRNAs may also bind to the open reading frame (ORF) of target mRNAs, requiring perfect or near-perfect complementarity, leading to cleavage and degradation of target mRNAs by the action of Ago2 (12). miRNAs have key regulatory functions in various cellular processes, such as proliferation, differentiation, and apoptosis (2). They play significant roles in the regulation of metabolism, development, cell cycle, carcinogenesis, cancer progression, and cancer metastasis (13, 14), and their expression is under tight control both spatially and temporally. Hence, altered expression of miRNAs is coupled with abnormal cellular behaviour, causing a range of disorders including human cancers. Understanding overall gene expression regulation has become more challenging since the discovery of miRNAs. It is estimated that a normal miRNA may have between 100 and 200 targets (15). Depending on the functional activity of the protein products of their target mRNAs, miRNAs can play oncogenic or tumour suppressor roles (16, 17). Tumour suppressor miRNAs negatively inhibit oncogenes and are down-regulated in cancer (16) while oncomiRs represent oncogenic miRNAs which are up-regulated in cancer cells, and have a role in the process of carcinogenesis through down-regulation of tumour suppressor molecules and/or molecules which are important for cellular differentiation (17). 1.2. Involvement of miRNAs in Breast Cancer and Their Potential Future Applications in Clinical Settings

Many studies have demonstrated the deregulation of miRNA profiling in breast tumours (18). miRNA expression profiles have been shown to be apparently aberrant in breast tumours compared with normal breast tissues. Iorio et al. found the expression level of miR-10b, miR-125b, miR-145, miR-21, and miR-155 to be considerably deregulated in breast cancer tissues, with miR-10b, miR-125b, and miR-145 being down-regulated and miR-21 and miR-155 up-regulated (18). miRNA expression patterns could

10  Detection and Quantification of MicroRNAs in Laser-Microdissected…

121

be used to classify human breast tumours, predict prognosis, and distinguish cancer tissues from adjacent normal tissues (18). Interestingly, breast cancer biological subtypes are shown to have distinctive miRNA expression signatures (19). This intimate involvement of miRNAs in malignant transformation, cellular invasion, and metastases in breast cancer makes them significant targets for developing a modern breast cancer molecular classification and opening avenues for more tailored treatment strategies for breast cancer patients. Dissecting the complex molecular network linking miRNAs to well-known oncogenes and tumour suppressor genes (such as H-RAS, HMGA2, LIN28, PEBP1, mucin 1, SOX4, PTPRN2, MERTK, TNC, HER2, HER3, IRS-1, P85 Beta, Raf1, ERK5, BMI1, ZEB1, ZEB2, Beta-catanin, AKT1, HMGA2, P53, RHOA, BCL-2, TPM1, PDCD4, PTEN, MASPIN, HOXD10, MCL1) has been an area of intense research during the last few years in an attempt to provide molecular explanation of how miRNAs regulate gene expression and induce cancer formation (see Note 3). Studies performed on human breast cancer tissues suggest that miRNAs could serve as future breast cancer biomarkers for earlier detection and more accurate diagnosis and prognostication. Moreover, they could unravel the molecular mechanisms of cancer pathogenesis and be targeted using a novel therapeutic approach involving synthetic oligonucleotide technologies. A comprehensive study of miRNA expression profiling in the currently defined breast cancer subclasses (20) has the potential to generate a more accurate molecular classification of breast cancer, which could guide clinicians to a more precise prognostication and treatment plan for each individual patient (see Note 4). 1.3. miRNA Extraction from Sections of Frozen and FormalinFixed ParaffinEmbedded Tissues

Compared with fresh-frozen tissue samples, formalin-fixed paraffin-embedded (FFPE) materials are more readily available and archives represent a major resource for the study of clinical samples with possible long-term follow-up data. However, mRNA from FFPE tissue is subject to damage, degradation, and alteration during fixation and processing, and is thus of limited value for gene expression analysis (21). RNA undergoes chemical modification by formalin and additional degradation during storage which is thought to be due to the occurrence of methylol cross links between RNA and protein during tissue processing (22). Nonetheless, it has been demonstrated that miRNA levels detected in total RNA extracted from FFPE are higher than those extracted from frozen cells when the amount of total RNA is identical (23). This might be attributed to the fact that generation of equivalent amounts of total RNA entails the inclusion of larger numbers of FFPE cells than fresh-frozen tissues, which could be due to the presence of residual cross links in RNA molecules which are not eliminated by proteinase K digestion,

122

S.M. Khoshnaw et al.

and hence failure of these RNA molecules being extracted (23). The possibility of existence of cross links is higher in longer RNA molecules; therefore, small RNAs are extracted more readily because they are less affected by this phenomenon (23). 1.4. Laser Capture Microdissection

Tissues are 3D structures comprising an assortment of cell subpopulations with different molecular and functional signatures. Laser capture microdissection (LCM) is an efficient and effective method for the separation of morphologically and/or phenotypically defined cell groups from complex tissues to facilitate downstream molecular analyses. LCM enables precise isolation of a particular cell subpopulation without contamination from surrounding cells; hence, it can be used to harvest cells of interest or remove unwanted cells under direct microscopic visualisation from a complex tissue sample. An application of LCM involves comparison of gene expression in normal tissues against tissues representing different stages of cancer progression. Principles, technical aspects, and applications of LCM were reviewed by Fend and Raffeld (24) and Espina et al. (25).

1.5. Aim of This Study

The aim of this work is to develop and optimise the methodology for miRNA expression analysis from microdissected FFPE breast cancer tissues.

2. Materials 2.1. Laser Microdissection

1. Positioning and ablation with Laser Microbeams (PALM) non-contact Laser catapulsion instrument (P.A.L.M. Microlaser Technologies), Carl Zeiss Ltd (Welwyn Garden City, Hertfordshire, UK). 2. PALM MembraneSlide 1.0 PEN (P.A.L.M. Microlaser Technologies GmbH), Carl Zeiss Ltd. 3. AdhesiveCap 500 opaque, Carl Zeiss Ltd. 4. 0.025% RNase-free aqueous Toluidine blue (HD Supplies, Botolph Claydon, Buckinghamshire, UK).

2.2. Total RNA Extraction and Quantification

1. miRNeasy FFPE Kit (50), QIAGEN, Inc. (Crawley, UK). 2. RecoverAll Total RNA Isol Kit FFPE 40Rxn, Applied Biosystems UK (Warrington, UK). 3. TRIzol reagent, Invitrogen, supplied by Fisher Scientific UK Ltd. (Loughborough, Leicestershire, UK). 4. Xylene, Fisher AR, 2.5 l, Fisher Scientific UK Ltd. 5. Ethanol absolute, Sigma-Aldrich (Gillingham, Dorset, UK). 6. Proteinase K solution RNA, Fisher Scientific.

10  Detection and Quantification of MicroRNAs in Laser-Microdissected…

123

7. Proteinase K digestion buffer, a component in the RecoverAll Total RNA Isol Kit FFPE 40Rxn, Applied Biosystems UK. 8. Chloroform, contains amylenes as stabiliser, ³99% (SigmaAldrich). 9. 2-Propanol, BioReagent, for molecular biology, ³99%, SigmaAldrich. 10. Glycogen UltraPure, Fisher Scientific. 11. Water, molecular biology reagent, Sigma-Aldrich. 12. Pipettor tip Microman capillary pistons, pure translucent polypropylene 1–10 ml Gilson, Fisher Scientific. 13. DNase I Amp grade Invitrogen GIBCO, 100  units, Fisher Scientific. 14. Quant-iT RiboGreen RNA Quantitation Kit, Fisher Scientific UK Ltd. (a) Quant-iT™ RiboGreen® RNA, Reagent (Component A) (b) 20× TE buffer, RNase-free (Component B) (c) Ribosomal RNA standard, 16S and 23S rRNA from, Escherichia coli (Component C) 2.3. miRNA Detection by TaqMan Real-Time PCR

1. TaqMan® MicroRNA Assays, hsa-miR-21, Applied Biosystems UK. 2. TaqMan® MicroRNA Biosystems UK.

Assays,

hsa-miR-29c,

Applied

3. TaqMan® MicroRNA Biosystems UK.

Assays,

hsa-miR-127,

Applied

4. TaqMan® Universal PCR Master Mix, No AmpErase® UNG, 1-Pack (1 × 5 ml). 5. MultiScribe™ Reverse Transcriptase, Applied Biosystems UK. 6. TaqMan “Universal PCR Master Mix, No AmpErase” UNG, 1-Pack (1 × 5 ml). Applied Biosystems UK. 7. TaqMan MicroRNA RT Kit, 200 Rxn, Applied Biosystems UK. 8. Mx3005P™ QPCR System (Agilent Technologies, South Queensferry, Scotland).

3. Methods 3.1. Samples and Cases

1. In surgically resected FFPE primary breast tumour specimens from three female breast cancer patients, “optimisation cases” were collected retrospectively with appropriate ethical approval

124

S.M. Khoshnaw et al.

Table 1 Three cases form archives of the Histopathology department of Nottingham City Hospital for methodology optimization No

Age (years)

Tumour size (cm)

Tumour stage

Tumour grade

Histological type

ER

HER2

1

46

3.6

3

3

NST

Positive

Positive

2

32

2.4

3

3

NST

Positive

Negative

3

63

5.8

3

3

NST

Negative

Negative

NST invasive non-special type

Fig. 1. Invasive breast cancer tissue.

from the archival stores of Nottingham Breast Cancer Series (Nottingham City Hospital Tumour Bank, Nottingham, UK; Table 1). 2. The average age of FFPE samples was 14 years (12–16-yearold samples), where patients were operated upon in 1993, 1995, and 1997. Selection of these cases was random, with the proviso that tissue samples from each case contained large amounts of invasive breast cancer cells (Fig. 1). H&E slides were produced from the FFPE blocks, and the presence of invasive tissue in each case was confirmed by Dr Z Hodi, consultant pathologist at Nottingham City Hospital. 3.2. Methodology Optimisation

Since the amount of total RNA in microdissected FFPE tissues is tiny, extraction of an amount of RNA sufficient for reliable miRNA detection requires investigation and optimisation of three vital aspects of methodology:

10  Detection and Quantification of MicroRNAs in Laser-Microdissected…

125

1. Thickness of FFPE samples mounted on PALM membrane slides to undergo LCM: 10 and 20 mm thick FFPE sections were investigated (see Note 1). 2. Amount of tissue to be microdissected: After mounting 10 and 20 mm thick sections on PALM membrane slides, a range of different microdissected areas (1 × 103–1 × 107  mm2) were investigated. Tissue macrodissection was also performed in which 5 × 107 mm2 was dissected with a sterile needle from tissue sections (see Note 1). 3. Choosing a total RNA extraction kit among three commercially available kits: Three different kits (Qiagen, RecoverAll, and TRIzol) were used to extract total RNA from full-face and microdissected FFPE tissues (see Note 2). To answer each of the above three questions, the outcomes of two core aspects of the procedure were taken into consideration: 1. Quantity of total RNA extract. For downstream analysis of miRNAs, 100 ng total RNA input is required for application on the human microRNA Microarrays (Agilent Technologies), (Publication number: 5990-4944EN), which contain probes for 723 humans and 76 human viral miRNAs from the Sanger database v.10.1. This technique was developed by Wang et al. (26). The labelling, probe design, and hybridization procedure used in this technique are simple and sensitive. The dephosphorylation and labelling of total RNA in the same sample tube are easy, and very low RNA input is needed. For miRNA profiling, 100 ng of tissue total RNA is dephosphorylated for 30  min at 37°C. This miRNA profiling assay directly utilises total RNA, and thus minimises the predictable sample losses occurring in more complicated procedures which require RNA size fractionation or many purification steps. In an ideal quantitative assay, each step should proceed in reproducibly high yield and be unaffected by minute variations from the standard procedure. These objectives are made possible when no amplification or separation steps are applied that can bring in sample-dependent variations and if both labelling and hybridization attain stable end points close to equilibrium, negligibly reliant on reaction kinetics and concentrations. Labelling reaction in this technique adds precisely one fluorophore to each miRNA in reproducibly high yield under circumstances that are insensitive to tiny differences in pCp-dye or miRNA concentration. Hybridization is permitted to progress towards equilibrium, and the probe melting temperature (Tm) matching ensures that nearly all miRNAs are hybridised at equilibrium. The number of hybridised labelled targets can be accurately measured via a calibrated scanner (26).

126

S.M. Khoshnaw et al.

2. miRNA detection sensitivity limit. In order to determine limits of sensitivity of miRNA detection in laser-microdissected tissues, the methodology was optimised via testing the techniques’ sensitivity to detect different miRNAs with varying expression levels in breast cancer. Therefore, three miRNAs which are differentially expressed in invasive breast tumours (high, intermediate, and low expression) were selected for methodology optimisation. It was previously shown that among 157 miRNAs, miR-21 was the most abundantly expressed miRNA in a total of five pairs of matched advanced breast tumour tissue specimens, and the level of miR-21 was much higher in the tumour tissues than in matched normal tissues (27). Recently, Yan et al. performed an extensive miRNA expression profiling study on eight primary human breast cancer tissues along with the adjacent normal tissues. They used an miRNA microarray containing 435 mature human miRNA oligonucleotide probes, and observed that 9 miRNAs (miR-21, miR-365, miR-181b, let-7f, miR-155, miR-29b, miR-181d, miR-98, and miR-29c) were up-regulated more than twofold in breast cancer compared with normal adjacent tissues while seven miRNAs (miR-497, miR-31, miR-355, miR-320, rno-mir-140, miR-127, and miR-30a-3p) were down-regulated more than twofold (28). The median of original signal of miR-21, miR-29c, and miR-127 were 10,768.5, 5,126.0, and 1,377.0, respectively. For methodology optimisation, specific probes for miR-21 as highly expressed, miR-29c as intermediately expressed, and miR-127 as an miRNA with low expression were utilised. 3.3. Sample Preparation for RNA Extraction

1. Full-face (gross) sections: 10 mm thick full-face FFPE sections were cut from the three optimisation cases. This was followed by immediate RNA extraction. 2. Microdissected tissues: (a) Mounting FFPE sections on PALM membrane slides and preparing them for LCM: As paraffin sections adhere more readily to hydrophilic surfaces, PALM membrane slides were irradiated with UV light at 254 nm for 30 min to overcome the hydrophobic nature of the membrane. Moreover, UV treatment sterilises the membrane and destroys potentially contaminating nucleic acids. The slides were used immediately after the UV treatment. 10 and 20 mm thick sections were cut using an RNase-free microtome from case three and mounted on RNase-free PALM membrane slides. For each block, the topmost few sections were not used. Sections were deparaffinised with xylene, rehydrated in RNase-free 100% ethanol, rinsed in RNase-free water and stained with 0.025% RNase-free aqueous toluidine blue.

10  Detection and Quantification of MicroRNAs in Laser-Microdissected…

127

(b) LCM: PALM non-contact Laser catapulsion instrument was used according to the manufacturer’s instructions to microdissect ten different cellular areas ranging from 1 × 103 to 1 × 107 mm2 from breast tissue sections (Fig. 2).

Fig. 2. Invasive breast cancer tissues pre- and post-laser capture microdessection (LCM) from one sample (Master Index number: 4,063). a1, b1, c1, and d1 are Pre-LCM and a2, b2, c2, and d2 are Post-LCM.

128

S.M. Khoshnaw et al.

The microdissectates were collected in PALM adhesive caps. A total of 45.5 ml proteinase K and 155.5 ml proteinase K digestion buffer were added to each sample and incubated at 55°C overnight (29) prior to RNA extraction. 3. Macrodissected tissues (needle macrodissection): Macro­ dissection of invasive breast cancer tissues, as judged by a serial H&E section, was performed with a sterile needle and placed into sterile microfuge tubes. The area macrodissected was approximately 5 × 107 mm2. 3.4. Total RNA Extraction

Three different commercial RNA extraction kits were used according to manufacturer’s instructions. 1. miRNeasy FFPE Kit, Qiagen: (a) For gross sections, four 10 mm thick sections were deparaffinised in xylene and washed by 100% ethanol. The tissues/microdissectates were resuspended in 150 ml buffer PKD, and 10 ml proteinase K was added. Samples were incubated at 55°C for 15  min, and then at 80°C for 15 min. Subsequently, 320 ml buffer RBC was added to each sample. (b) The lysate was transferred to a gDNA Eliminator spin column and centrifuged for 30 s at ³8,000 × g. A total of 1,120 ml ethanol was added. The sample was transferred to an RNeasy MinElute® spin column (700 ml at a time) and centrifuged for 15 s at ³8,000 × g. (c) This was repeated until the entire sample had passed through the spin column. 500 ml buffer RPE was added to the RNeasy MinElute spin column and centrifuged for 15 s at ³8,000 × g. 500 ml buffer RPE was again added to the RNeasy MinElute spin column and centrifuged for 2 min at ³8,000 × g. (d) The spin columns were centrifuged at full speed for 5 min with their lids open. 20 ml RNase-free water was added directly to the spin column membrane and centrifuged for 1 min at full speed to elute the RNA. 2. RecoverAll Total RNA Isol Kit FFPE: (a) For gross sections, three 20  mm thick sections were deparaffinised in xylene and washed by 100% ethanol. 400 ml digestion buffer and 4 ml protease were added to each tissue/microdissectate sample and incubated for 3 h at 50°C. 480 ml Isolation Additive was added followed by 1.1 ml 100% ethanol. 700 ml of the sample/ethanol mixture was pipetted onto a Filter Cartridge and centrifuged at 10,000 × g for 30–60 s.

10  Detection and Quantification of MicroRNAs in Laser-Microdissected…

129

(b) Subsequently, 700 ml of Wash 1 was added to the Filter Cartridge and centrifuged for 30  s at 10,000 × g. This was followed by adding 500 ml of Wash 2/3 to the Filter Cartridge and centrifuging for 30 s at 10,000 × g. 60 ml DNase mix was added to each sample and incubated for 30 min at room temperature, and then 700 ml of Wash 1 was added to the Filter Cartridge. The Filter Cartridge was washed twice with 500 ml of Wash 2/3 and centrifuged for 1 min at 10,000 × g. RNA was eluted with 30 ml nuclease-free water and centrifuged for 1  min at maximum speed. The elution step was repeated, and the volume of collected eluate (which contained the RNA) was close to 60 ml. 3. TRIzol reagent: (a) For gross sections, three 10  mm thick sections were deparaffinised in xylene by incubation at 65°C for a total of 20 min substituting xylene twice. Tissue samples were washed twice by 100% ethanol and incubated in 45.5 ml proteinase K and 155.5 ml proteinase K digestion buffer at 55°C overnight (29). 1 ml Trizol was added to each sample, and the samples were vortexed. 200 ml of chloroform was added to each sample. Samples were shaken vigorously for 15  s and left at room temperature for 1 min to settle, and then spinned at 18,894 × g for 15 min at 4°C. (b) The aqueous layer was carefully taken without touching the interphase. This was added to 650  ml isopropanol with 0.5  ml RNase-free glycogen. Samples were left at room temperature for 30  min and centrifuged at 12,000 × g for 15 min at 4°C. (c) The medium was taken, and the pellet was washed twice by adding 1 ml of 75% ethanol and spinning at 12,000 × g for 15 min at 4°C. (d) Finally, ethanol was removed and RNA was dissolved in 12 ml of nuclease-free water. For microdissectates samples, 10 and 20  mm thick sections were mounted on PALM membrane slides. Sections were dried at room temperature overnight, deparaffinised with xylene (5 min twice) and 100% ethanol (5  min twice), rinsed in RNase-free water, and then stained with 0.025% aqueous Toluidine blue. Trizol, chloroform, and isopropanol were used in different volumes as compared to full-face sections, as follows: Trizol – 200 ml, chloroform – 40 ml, and isopropanol – 300 ml.

130

S.M. Khoshnaw et al.

3.5. DNase Treatment of RNA Extracts

3.6. Total RNA Quantification

DNase treatment is one of the steps in Qiagen and RecoverAll RNA extraction kit protocols, while this is not included in the TRIzol reagent procedure. Therefore, the quantification of total RNA was repeated after performing DNase treatment for the RNA extracted from the three optimisation cases via TRIzol reagent. Briefly, 6 ml total RNA was used and increased to 10 ml by adding 1 ml DNase I Amplification Grade (1 U/ml), 1 ml 10× DNase I buffer, and 2  ml DEPC-treated water. This was gently vortexed and spinned briefly, and incubated at room temperature for 15 min. After a brief centrifuge, 1 ml of 25 mM EDTA was added to each sample on ice. Gentle mixing and a brief spinning were performed, and samples were incubated at 65°C for 10 min. 1. Total RNA in microdissected tissues is very small; hence, a sensitive technique is required to accurately quantify RNA in these samples. NanoDrop® ND-1000 UV–Vis Spectro­ photometer instrument was initially used to measure the quantity and quality of extracted total RNA. Detection limit for this method is 2–3,000 ng/ml for RNA, but when used on the microdissected tissue samples, results were not reproducible and the method was deemed inaccurate. 2. Therefore, RiboGreen reagent was used according to manufacturer’s instructions to accurately quantify total RNA in microdissected tissues. Low- and high-range standard curves were initially used in this technique. Detection limit of lowrange standard curve is 0.001–0.05 ng/ml and of high-range standard curve is 0.02–1 ng/ml. Low- and high-range standard curves were produced to match the value of RNA in the samples (Fig. 3). TE buffer (10 mM Tris–HCl, 1 mM EDTA, pH 7.5) was used to dilute Quant-iT™ RiboGreen® reagent 200-fold for the high-range assay. RNA extracted from gross tissues was diluted 1:1,500 and that from microdissected tissues 1:25 to remain within detection limits of the highrange standard curve.

Fig. 3. High- and low-range standard curves for RiboGreen reagent produced in our laboratory.

10  Detection and Quantification of MicroRNAs in Laser-Microdissected…

131

3. To produce a high-range standard curve, 16S and 23S ribosomal RNAs were used to produce a 2 ng/ml RNA stock solution in TE. Using serial dilutions, three other RNA ­concentrations (1, 0.2 and 0.04  ng/ml) were prepared to ­produce the curve. 10 ml of the diluted RiboGreen reagent was added to 10 ml of the diluted RNA samples and 10 ml of each of the four ribosomal RNA concentrations (2, 1, 0.2 and 0.04 ng/ml). 4. Fluorescence signal was measured in samples using Mx3005P™ QPCR System (Agilent Technologies). 3.7. miRNA Detection by TaqMan Real-Time PCR

1. Three mature miRNAs (miR-21, miR-29c, and miR-127) were quantified using TaqMan® MicroRNA Assays, according to the manufacturer’s instructions. 2. TaqMan miRNA assays use the stem-loop method to detect the expression level of mature miRNAs (30, 31). Total RNA extract was transcribed into cDNA in a 15  ml volume reaction. All PCRs were performed in a final volume of 20 ml and were performed in duplicate for each cDNA sample. 3. The number of PCR cycles needed to reach the threshold cycle (Ct) was identified in duplicate for each cDNA and an average taken. Briefly, 10 ng total RNA (in 5 ml) was mixed with 3  ml RT primer and 7  ml master mix per each reverse transcription (RT) reaction (15 ml) which was carried out as follows: 16°C for 30 min; 42°C for 30 min; 85°C for 5 min; and then held at 4°C. Amplification of each target miRNA was performed in duplicate using a 1.33  ml aliquot of each first-strand cDNA reaction in a 20 ml total reaction volume along with 1 ml TaqMan miRNA assay (20×), 10 ml TaqMan 2× Universal PCR Master Mix-No AmpErase UNGa, and 7.67  ml nuclease-free water. An miRNA amplification was performed using Mx3005P® QPCR System (Agilent Technologies) following manufacturer’s instructions with the following incubation times: 95°C for 10 min followed by 40 cycles at 95°C for 15 s and 60°C for 60 s. 4. The real-time PCR results were analysed using MxPro™ QPCR Software. Two negative controls were run in duplicate for each sample, in which RNase-free water was substituted for cDNA which showed no amplification.

3.8. miRNA Profiling from Full-Face (Gross) FFPE Sections

1. Total RNA was extracted from full-face FFPE sections of the three optimisation cases using three RNA extraction kits (Qiagen, RecoverAll, and TRIzol). 2. Total RNA was quantified using RiboGreen reagent, a highly sensitive RNA detection method. Total RNA was highest using TRIzol reagent (608.38  ng/1  mm tissue thickness/

132

S.M. Khoshnaw et al.

Table 2 RiboGreen (RG) reagent was used to measure the quantity of total RNA extracted from the three optimization cases RNA (ng)/1 mm thick section/ whole volume Ct of elutea (miR-21)

Cases

Total section thickness (mm)

Qiagen (DT)

Case 1 Case 2 Case 3

40 40 40

186.30

18.41 18.73 19.47

24.53 25.09 25.34

26.22 27.36 27.56

Ambion (DT)

Case 1 Case 2 Case 3

60 60 60

292.52

20.07 19.76 20.73

26.59 26.14 27.18

28.21 28.24 28.67

Invitrogen (no DT)

Case 1 Case 2 Case 3

30 30 30

608.38

19.79 18.815 20.73

26.14 25.51 26.98

27.98 27.96 28.63

Invitrogen (with DT)

Case 1 Case 2 Case 3

30 30 30

91.86

19.48 18.52 19.32

25.14 24.9 24.96

27.63 26.96 27.65

RNA extraction kit

Ct Ct (miR-29c) (miR-127)

Mature miR-21, miR-29c, and miR-127 were quantified in the total RNA using TaqMan® MicroRNA Assays Ct threshold cycle value, DT DNase treatment a 18 ml was used to elute RNA in Qiagen kit, 60 ml in Ambion kit, and 12 ml in Invitrogen kit

whole elute), and lowest using the Qiagen kit (186.30 ng/1 mm tissue thickness/whole elute; Table 2). 3.9. miRNA Detection and Quantification

1. RNA samples were analysed in duplicate to quantify miR-21, miR-29c, and miR-127 using RT-PCR (Table  2). All the samples showed earliest amplification for miR-21 and latest amplification for miR-127. 2. The Ct values for miR-21, miR-29c, and miR-127 were comparable in RNAs extracted by Qiagen and TRIzol kits while Ct values in the RNA extracted with RecoverAll kit were slightly higher (Table 2). 3. DNase treatment reduced the quantity of RNA by about sevenfold; however, the three miRNAs were amplified in DNasetreated samples at almost the same Ct values compared to non-DNase-treated samples (Table 2). 4. To confirm reproducibility of our results, RNA extraction from gross tissues of the three optimisation cases and quantification of the three miRNAs from the RNA extracts were repeated, and the results were comparable (Table 2, Fig. 4).

10  Detection and Quantification of MicroRNAs in Laser-Microdissected…

133

Fig.  4. miRNA quantification from RNA extracted from full-face sections of the three optimisation cases. RNA was extracted using three different kits. Qiagen and Invitrogen showed comparable results while Ambion showed slightly later miRNA amplification.

134

S.M. Khoshnaw et al.

3.10. miRNA Profiling from LaserMicrodissected FFPE Tissues

For laser-microdissected tissues, RecoverAll RNA extraction kit was not investigated because RNA extracts by this kit from fullface sections showed later amplification of miRNAs compared to Qiagen and TRIzol kits (see Note 2).

3.11. Total RNA Extraction and Quantification

Ten different areas (1 × 103, 5 × 103, 5 × 104, 2 × 105, 5 × 105, 1 × 106, 3 × 106, 5 × 106, 7 × 106, and 1 × 107 mm2) were microdissected from case three of the optimisation cases and RNA extracted using two alternative commercially available kits (Qiagen and TRIzol). Qiagen kit was used to extract total RNA from 10 mm thick FFE sections of case three of the optimisation cases, where the RNA quantity in microdissectates was very low,