Single topic volume
Cancer Genomics
Editors
L.A. Cannizzaro, Bronx, N.Y. K.H. Ramesh, Bronx, N.Y.
88 figures, 48 in color, and 35 tables, 2007
Basel • Freiburg • Paris • London • New York • Bangalore • Bangkok • Singapore • Tokyo • Sydney
Cover illustration Partial ideograms showing the abnormal chromosomes (only) which are involved in what is known as the ‘classic translocations’ that are diagnostic of some hematologic and soft tissue malignancies. 1st Row The t(9;22)(q34;q11.2) diagnostic of chronic myelogenous leukemia (CML). The t(8;21)(q22;q22) diagnostic of acute myeloid leukemia (AML). The t(15;17)(q22;q12) diagnostic of acute promyelocytic leukemia (APL). 2nd Row The t(2;5)(p23;q25) diagnostic of anaplastic large cell lymphoma (ALCL). The t(11;14)(q13;q32) diagnostic of mantle cell lymphoma (MCL). The t(8;14)(q24;q32) diagnostic of Burkitt lymphoma (BL). 3rd Row The t(1;13)(p36;q14) diagnostic of alveolar rhabdomyosarcoma. The t(X;18)(p11.2;q11.2) diagnostic of synovial sarcoma. The t(11;22)(q24;q12) diagnostic of Ewing sarcoma. Cover illustration from K.H. Ramesh and Linda A. Cannizzaro, Cytogenetics Laboratory, Montefiore Medical Center, Bronx, NY.
S. Karger Medical and Scientific Publishers Basel • Freiburg • Paris • London New York • Bangalore • Bangkok Singapore • Tokyo • Sydney
Disclaimer The statements, options and data contained in this publication are solely those of the individual authors and contributors and not of the publisher and the editor(s). The appearance of advertisements in the journal is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements. Drug Dosage The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any change in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.
Fax +41 61 306 12 34 E-Mail
[email protected] www.karger.com
All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher or, in the case of photocopying, direct payment of a specified fee to the Copyright Clearance Center (see ‘General Information’). © Copyright 2007 by S. Karger AG, P.O. Box, CH–4009 Basel (Switzerland) Printed in Switzerland on acid-free and non-aging paper (ISO 9706) by Reinhardt Druck, Basel
Vol. 118, No. 2–4, 2007
Contents
91 Preface Cannizzaro, L.A.; Ramesh, K.H. (Bronx, NY) 92 Head and neck cancer: reduce and integrate for
optimal outcome Belbin, T.J.; Bergman, A.; Brandwein-Gensler, M.; Chen, Q.; Childs, G.; Garg, M.; Haigentz, M.; Hogue-Angeletti, R.; Moadel, R.; Negassa, A.; Owen, R.; Prystowsky, M.B.; Schiff, B.; Schlecht, N.F.; Shifteh, K.; Smith, R.V.; Zheng, X. (Bronx, NY) 110 Origin and functional significance of large-scale
chromosomal imbalances in neuroblastoma Stallings, R.L. (San Antonio, TX) 116 Genomic signatures of breast cancer metastasis Urquidi, V.; Goodison, S. (Jacksonville, FL) 130 High-resolution array comparative genomic
hybridization of chromosome 8q: evaluation of putative progression markers for gastroesophageal junction adenocarcinomas van Duin, M.; van Marion, R.; Vissers, K.J.; Hop, W.C.J.; Dinjens, W.N.M.; Tilanus, H.W.; Siersema, P.D.; van Dekken, H. (Rotterdam) 138 Pathology and genetics of adipocytic tumors Hameed, M. (Newark, NJ) 148 Molecular cytogenetic characterization of
pancreas cancer cell lines reveals high complexity chromosomal alterations Griffin, C.A.; Morsberger, L.; Hawkins, A.L.; Haddadin, M.; Patel, A. (Baltimore, MD); Ried, T.; Schrock, E. (Bethesda, MD); Perlman, E.J.; Jaffee, E. (Baltimore, MD) 157 Molecular mechanisms underlying the MiT
translocation subgroup of renal cell carcinomas Medendorp, K.; van Groningen, J.J.M.; Schepens, M.; Vreede, L.; Thijssen, J.; Schoenmakers, E.F.P.M.; van den Hurk, W.H.; Geurts van Kessel, A.; Kuiper, R.P. (Nijmegen)
Contents
166 Disruption of the FA/BRCA pathway in bladder cancer Neveling, K.; Kalb, R. (Würzburg); Florl, A.R. (Düsseldorf); Herterich, S.; Friedl, R.; Hoehn, H. (Würzburg); Hader, C.; Hartmann, F.H. (Düsseldorf); Nanda, I.; Steinlein, C.; Schmid, M. (Würzburg); Tönnies, H. (Berlin); Hurst, C.D.; Knowles, M.A. (Leeds); Hanenberg, H. (Düsseldorf); Schulz, W.A.; Schindler, D. (Würzburg) 177 The genetics of bladder cancer: a cytogeneticist’s
perspective Wolff, D.J. (Charleston, SC) 182 The cytogenetics and molecular biology of
endometrial stromal sarcoma Sandberg, A.A. (Phoenix, AZ) 190 Pathogenetic mechanisms in endometrial stromal
sarcoma Micci, F.; Heim, S. (Oslo) 196 Influence of a nonfragile FHIT transgene on murine
tumor susceptibility McCorkell, K.A. (Philadelphia, PA); Mancini, R. (Rome); Siprashvili, Z. (Stanford, CA); Barnoski, B.L. (Philadelphia, PA); Iliopoulos, D. (Columbus, OH); Siracusa, L.D. (Philadelphia, PA); Zanesi, N.; Croce, C.M.; Fong, L.Y.Y.; Druck, T.; Huebner, K. (Columbus, OH) 204 Exogenous mycoplasmal p37 protein alters gene
expression, growth and morphology of prostate cancer cells Goodison, S. (Jacksonville, FL); Nakamura, K.; Iczkowski, K.A.; Anai, S.; Boehlein, S.K.; Rosser, C.J. (Gainesville, FL) 214 SNP-Array genotyping and spectral karyotyping
reveal uniparental disomy as early mutational event in MSS- and MSI-colorectal cancer cell lines Melcher, R. (Würzburg); Al-Taie, O. (Aschaffenburg); Kudlich, T.; Hartmann, E.; Maisch, S.; Steinlein, C.; Schmid, M.; Rosenwald, A.; Menzel, T.; Scheppach, W.; Lührs, H. (Würzburg)
Cytogenet Genome Res Vol. 118, 2007
89
310 Genetic alteration associated with chronic
222 Familial cancer syndromes: catalog with comments Hecht, F. (Scottsdale, AZ)
lymphocytic leukemia Cotter, F.E.; Auer, R.L. (London)
229 Interphase FISH as a new tool in tumor pathology Tibiletti, M.G. (Varese)
320 Genetics and epigenetics of 1q rearrangements in
hematological malignancies
237 The dynamics of cancer chromosomes and genomes Ye, C.J.; Liu, G.; Bremer, S.W.; Heng, H.H.Q. (Detroit, MI)
Fournier, A.; Florin, A.; Lefebvre, C.; Solly, F.; Leroux, D.; Callanan, M.B. (Grenoble)
247 The mystery of chromosomal translocations in cancer Koss, L.G. (Bronx, NY)
328 Cell migration patterns and ongoing somatic
mutations in the progression of follicular lymphoma Adam, P.; Schoof, J.; Hartmann, M. (Würzburg); Schwarz, S. (Regensburg); Puppe, B.; Ott, M.; Rosenwald, A.; Ott, G.; Müller-Hermelink, H.K. (Würzburg)
252 MicroRNAs in carcinogenesis Hagan, J.P.; Croce, C.M. (Columbus, OH) 260 Non-random inactivation of large common fragile site
genes in different cancers
337 Molecular cytogenetic analysis of follicular
lymphoma (FL) provides detailed characterization of chromosomal instability associated with the t(14;18)(q32;q21) positive and negative subsets and histologic progression
McAvoy, S.; Ganapathiraju, S.C.; Ducharme-Smith, A.L.; Pritchett, J.R.; Kosari, F.; Perez, D.S.; Zhu, Y. (Rochester, MN); James, C.D. (San Francisco, CA); Smith, D.I. (Rochester, MN) 270 Telomere dysfunction and telomerase activation in
Nanjangud, G. (New York, NY); Rao, P.H. (Houston, TX); Teruya-Feldstein, J.; Donnelly, G.; Qin, J.; Mehra, S.; Jhanwar, S.C.; Zelenetz, A.D.; Chaganti, R.S.K. (New York, NY)
cancer – a pathological paradox? Calcagnile, O.; Gisselsson, D. (Lund)
345 Molecular cytogenetics of IGH rearrangements in
277 Etiology of specific molecular alterations in human
malignancies
non-Hodgkin B-cell lymphoma
Vauhkonen, H.; Heino, S.; Myllykangas, S.; Lindholm, P.M.; Savola, S.; Knuutila, S. (Helsinki)
Bernicot, I.; Douet-Guilbert, N.; Le Bris, M.-J.; Herry, A.; Morel, F.; De Braekeleer, M. (Brest) 353 Molecular cytogenetics in the study of cutaneous
284 FISH panels for hematologic malignancies Sreekantaiah, C. (Stratford, CT)
T-cell lymphomas (CTCL)
297 Genetic and epigenetic alterations in myelodysplastic
Karenko, L.; Hahtola, S.; Ranki, A. (Helsinki)
syndrome Mihara, K.; Takihara, Y.; Kimura, A. (Hiroshima) 304 DNA profiling by arrayCGH in acute myeloid
leukemia and myelodysplastic syndromes
362 Author Index Vol. 118, No. 2–4, 2007 363 Author Index Vol. 118, 2007 after 364 Contents Vol. 118, 2007
Suela, J.; Alvarez, S.; Cigudosa, J.C. (Madrid)
90
Cytogenet Genome Res Vol. 118, 2007
Contents
Cytogenet Genome Res 118:91 (2007) DOI: 10.1159/000108289
Preface
This special themed issue is dedicated to the founder of this Journal, Dr. Harold P. Klinger (1929– 2004). Dr. Klinger devoted most of his research career to the study of the genetics of cancer. He was instrumental to the creation of this Journal and rounded out his very interesting and illustrious life by being an avid salmon fisherman. Many of us who were privileged to know him benefitted from both of his lifes’ passions. His faithful and tireless devotion to his research and this Journal along with the multitude of professional contributions have been enumerated in another issue of this Journal (Vol. 109(4), 2005). The efforts and contributions within our present volume are an added way to cherish and honor Dr. Klinger’s many achievements in the area of genetic research. We would like to acknowledge and thank Dr. Michael Schmid for his sage advice and constant support for the duration of the preparation of this issue. We also wish to thank the many investigators and clinicians who participated in this endeavor for their inspiring ideas and innovative contributions. Advances in the area of cancer genomics have blossomed in the last decade as a result of an enormous diversity of technological advances. Momentous strides made in genome sequencing efforts have transformed this area of investigation in a significant fashion. The papers contained in this volume represent the enormous progress made in recent years in the study of leukemias, lymphomas and solid tumors. We consider ourselves privileged to consolidate such a pivotal volume in cancer genomics as we feel it contains a compendium of information and knowledge that will impact the study of cancer for many years to come. We have organized the papers to elicit the advances made in solid tumor exploration by placing these investigations at the beginning of the volume. In the past, solid tumor genomics has been a difficult area for investigations to yield much in the way of new information due to the difficulty in obtaining premalignant and early stage tumor samples from patients. Now with the availability of more sensitive diagnostic expertise and state of the art techniques, the number of samples obtainable at earlier and hence, more treatable stages, has significantly increased the success rate of tumor diagnoses. This needs to be emphasized in our approach to presenting this information to our readers. The middle of the volume is devoted to investigations which have utilized some of the more refined and sensitive technologies now available to study cancer. And the latter part of the volume represents advances specifically in the investigation of leukemias and lymphomas. All of the work presented in this volume represents just a small smattering of the continually expanding area of cancer genomic research, i.e., the collaboration of physicians, scientists, and everyone who is responsible for detecting, diagnosing and treating malignancies in patients with cancer. We also celebrate this work represented by a much improved quality of life and survival for patients with cancers of all kinds. Linda A. Cannizzaro K. H. Ramesh Bronx, NY, July 2007
Fax +41 61 306 12 34 E-Mail
[email protected] www.karger.com
© 2007 S. Karger AG, Basel 1424–8581/07/1184–0091$23.50/0
Accessible online at: www.karger.com/cgr
Etiology Environmental factors Infectious agents
Cancer
Clinical visit
Anatomic site LN+
LN–
Mutations
Diagnostic evaluation Imaging
Surgery Radiotherapy Chemotherapy Combination therapy Palliative therapy
History (exposure, risk factors, etc.) Physical exam
Susceptibility Individual’s genome
Treatment selection
Pathology Programmed cell death
Fig. 1. Factors affecting development and behavior of HNSCC. Etiologies include exposure to environmental factors that can vary geographically and by individual choice and to infectious agents. Genetic determinants of risk will influence whether cells accumulate DNA mutations in response to carcinogens or undergo programmed cell death. Once a cancer has developed in an individual, genetics will determine the range of responses to the cancer and therapy. For small (T1 and T2) tumors, there is a small chance of lymph node (LN) metastasis at initial presentation when the tumor is located in the oral cavity (⬃10%) as compared to a relatively frequent rate of metastasis for the same sized tumor in the oropharynx (⬃70%).
have a predilection for developing cancers in particular locations, e.g., nasopharynx in Southern China and oral cavity in India. The vast majority of head and neck cancers are squamous cell carcinomas arising from the surface epithelium of the upper aerodigestive tract mucosa. Although tobacco exposure and alcohol consumption are the most common environmental exposures related to this disease, other factors associated with this disease are HPV or EBV infection, chronic irritation, Betel quid exposure, chewing tobacco and a variety of other agents. Thus, the complexity of the anatomy, the variety of etiologies and the heterogeneity of the populations affected create a significant challenge in treatment selection and management of patients with HNSCC. Despite the global importance of this disease, the fiveyear survival rate for HNSCC (ca. 50%) has remained relatively stable over the past three to four decades. Meta-analyses report a marginal improvement in survival of approximately 8% over the past decade when using combination and concurrent therapies (Pignon et al., 2000; Khuri and Jain, 2004; Pignon and Burdett, 2006). Some improvements in survival are observed for certain head and neck sites when compared to others, notably a decline in survival for oral cavity and laryngeal tumors compared to pharyngeal cancers (Carvalho et al., 2005). As a result, it is estimated that 45,641 cases and 11,210 deaths will occur in 2007 in the United States from HNSCC (Jemal et al., 2007). HNSCC, as noted, constitute an anatomically heterogeneous group arising most often from the oral cavity, oropharynx, hypopharynx, larynx and nasopharynx. Certain anatomic sites have a predilection for early metastasis (Lindberg, 1972; Schwartz et al., 1998, 2000; Barnes et al., 2001) and treatment planning depends entirely upon anatomical staging of the disease at presentation. At this point, staging and treatment decisions rely solely upon the anatomic location of the primary cancer and the presence, or absence, of metastatic disease. Physical examination plays a pivotal role in local-
Biomarkers?
Fig. 2. Diagnosis and treatment decision for HNSCC. The transition from a patient’s complaint to the development of a treatment plan currently consists of an assessment of an individual’s risk factors, a determination of the anatomic location of the tumor (physical exam and imaging) and a description of the tumor type (histology). An appropriate treatment plan is then developed based upon a clinician’s cumulative understanding of the disease from their experience and medical knowledge. A better way to develop an appropriate treatment plan would be to incorporate clinically meaningful biomarker analyses which would assist in predicting response or resistance to particular therapies and tumor aggressiveness for early stage (I and II) tumors. Such biomarkers could potentially enhance survival or minimize toxicities and improve quality of life in these patients.
izing the tumor and regional spread to the lymph nodes. Imaging techniques, such as computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography (PET) are all used to augment the physical exam and identify occult tumor extension into non-visualized, adjacent structures or lymph nodes. It is also important to assess the possibility of distant metastasis to the lungs for advanced stage disease. Pathologic features may also have an important impact in staging particular tumors of the head and neck and are critical in the most accurate assessment of the lymph nodes in patients who undergo surgical management of the neck (Fig. 2). Imaging techniques have improved our ability to diagnose occult disease and more accurately stage tumors prior to therapy. Standard imaging is composed predominantly of CT scanning, though MRI may be superior for some tumor sites such as the tongue base. These anatomic imaging techniques rely upon certain characteristics to predict involvement of structures with a tumor. Thus far, however, with the exception of the anatomical structures involved, a function of the standard clinical TNM (Tumor, Node, Metastasis) stage, imaging has not been predictive of outcome or response. Functional imaging with PET scanning has become commonplace in the initial evaluation of HNSCC patients, particularly those with advanced disease, as well as in the follow-up of patients after completion of therapy. This functional technique relies upon the transport of interstitial glucose into cells, mediated by a family of homologous glucose transport proteins (GLUTs), which differ in their tissue distribution and physiological properties (Joost and Thorens, 2001). Of these isoforms, GLUT1, an insulin independent transporter, in particular, is over-expressed in a variety of
Cytogenet Genome Res 118:92–109 (2007)
93
cancers including HNSCC (Mellanen et al., 1994; Reisser et al., 1999), and is associated with poorer prognosis (Mineta et al., 2002; Kunkel et al., 2003; Oliver et al., 2004; De Schutter et al., 2005). High 18F-fluordeoxyglucose (FDG) uptake is seen in PET scans of HNSCC patients with overexpression of GLUT1. Unfortunately, to date, the relative uptake of PET isotopes has not shown a consistent relationship with prognosis. Efforts to correlate nuclear imaging with response have been attempted in the past without significant impact. Multi-drug resistance P-glycoprotein (MDR) is a significant factor in HNSCC resistance to chemotherapy (Rabkin et al., 1995) and expression is induced by external beam radiotherapy (Ng et al., 1998). Imaging and uptake of 99mTc-SetaMIBI (MIBI), a substrate for MDR, is a surrogate for MDR activity in many cancers. However, there was no correlation between MIBI uptake and MDR expression pathologically in patients with head and neck cancer (Leitha et al., 1998) and modulation of MDR expression in a head and neck cancer xenograft model was not accurately assessed with MIBI scanning (Mubashar et al., 2004). Additional functional imaging techniques are emerging which may hold promise in helping predict treatment response or other tumor characteristics. Diffusion weighted MRI (DWI) is a new, emerging functional technique which provides information about tissue cellularity and integrity of cellular membranes (Koh and Padhani, 2006). Diffusion imaging detects movement of water molecules. While in a freely diffusible solute, diffusion is random (Brownian motion), in biological tissues, diffusion is restricted by interaction of water molecules with hydrophobic cellular membranes and macromolecules. The ‘apparent diffusion coefficient’ (ADC) quantifies the magnitude of diffusion in tissues. Different tissues have different apparent diffusion coefficients, producing contrast between the tissues. Wang et al. (2001) found that malignant tumors had lower ADC compared to benign lesions. They attributed the relatively restricted diffusion in malignant tumors to their greater cellular density as shown by histology. Vandecaveye et al. (2006) found diffusion measurements could differentiate between recurrent tumor and necrosis after radiation therapy. Loss of cellular structure due to necrosis is thought to lead to the increase in diffusion following radiotherapy. Ultimately, DWI MRI may be used in conjunction with conventional imaging techniques to differentiate between normal and abnormal tissues as well as to ascertain the response to therapy. Treatment for any given tumor depends upon the disease site, the extent of the primary tumor and the presence of regional or distant metastases. Conventional treatment may employ surgery, radiation therapy and chemotherapy either alone or in combination with any, or all, of the other modalities. Any of these therapies can, and do, produce morbidities affecting speech, swallowing and overall quality of life. These morbidities are often increased by the routine use of combined therapy in patients with advanced disease (Calais et al., 1999; Brizel et al., 2000; Epstein et al., 2001; Gal et al., 2003; Cooper et al., 2004). This is done in an effort to improve survival and locoregional control. Despite these
94
Cytogenet Genome Res 118:92–109 (2007)
interventions, recurrence of the disease is observed in about 50% of patients, locally, regionally or at a distant site with high rates of associated mortality (Takes et al., 1997). In fact, with the modest improvement in overall survival over the past decades, one must question whether a large percentage of patients are overtreated, incurring additional morbidity. Strategies must be developed to improve the selection of the most appropriate treatment for any given patient, and clinicians may need to reconsider the strategy of treatment escalation in all patients. Clinical experience has shown us, as cited above, that such escalation increases the morbidity of the therapy but does not necessarily translate into improved survival. In contrast, it may significantly reduce the quality of life in these patients and impact their ability to socialize and interact with family and friends (List and Bilir, 2004). While difficult to predict, treatment failures for HNSCC can be attributed to multiple factors. Some of the factors causing treatment failure are unknown and addition of isolated biomarkers has not improved our ability to predict outcome. Currently, the best predictors are the clinical findings of the TNM staging system (Tumor, Node, Metastasis). These factors, in conjunction with newer data supporting the use of aggressive chemotherapy and radiotherapy regimens (Forastiere et al., 2003; Bernier et al., 2004; Cooper et al., 2004; Bonner et al., 2006), argue for the need to develop new treatment paradigms. Concurrently, it would be useful to identify those patients who would benefit from multimodality therapies versus those patients who may receive less aggressive and less toxic therapy without impacting survival expectations (Smith RV et al., 2004; Beitler et al., 2007); the latter group might receive single modality therapy, or, more likely, the introduction of biologic response modifiers with potentially less morbidity. This paradigm has been recently employed through the use of the EGFR antagonist cetuximab in combination with radiation therapy and shown to improve locoregional control, median duration of overall survival and progression-free survival when compared to radiation therapy alone (Bonner et al., 2006). Efforts to identify biomarkers for HNSCC have been numerous and none have proven to be useful clinically; i.e., no biomarker is used to guide treatment selection at initial diagnosis. Studies on p53, EGFR, BCL2, MMPs, cyclins and molecular markers have demonstrated inconsistent, and at times contradictory, results (Kyzas et al., 2005; McShane et al., 2005a; Lothaire et al., 2006). Overall, identification of a single genetic or protein abnormality is likely to have a limited clinical benefit due to the inherent heterogeneity of these tumors and patients. In squamous cell carcinomas, the multistep carcinogenesis pathway, as proposed by Califano et al. (1996), accounts for many events in the transition from normal to cancer cells. These events are likely to vary with different etiologic agents in individual patients. One might expect a spectrum of genetic changes to occur, with some common themes, throughout the progression from normal tissue to carcinoma. In fact, early gene expression changes can be seen in oral mucosa after relatively limited exposure to tobacco smoke (Smith RV et al., 2006). Recent data support such differences, particularly with respect to
HPV exposure. Therefore, global changes in gene expression, DNA methylation status, proteomic expression or other measures are more likely to provide us with prognostic information rather than individual mutations. In such a model, each specific change is not as important as the overall pattern of gene expression. In fact, expression changes predictive of tumor behavior may represent changes in the host’s ability to control and respond to a tumor, i.e. immune response, angiogenesis induction, apoptosis induction, rather than tumor specific factors such as alterations in EGFR expression or p53 expression. The complexity of the tumor-host interaction and the tumor microenvironment, almost by necessity, limits the potential utility of single biomarkers (Condeelis and Pollard, 2006). This is also demonstrated by the experience of predictive gene expression arrays in breast cancer, where patterns of expression have been demonstrated to be useful in outcome prediction and treatment determination (Perou et al., 2000; van de Vijver et al., 2002; van ‘t Veer et al., 2002; Potti et al., 2006). Molecular characterization of this genetically complex disease has provided some insight into individual genetic abnormalities that contribute to tumor progression (Forastiere et al., 2001). Gene profiling can predict differences in tumor behavior not discernable by histopathology (Belbin et al., 2002). Our ability to provide effective treatment will depend upon our ability to determine that surgical margins are adequate, and our ability to predict tumor potential for metastasis and local recurrence and response to therapy. Thus, profiling of primary tumors using global RNA expression and global proteomic analyses coupled with sophisticated analyses identifying combinatorial and modifier effects will likely yield new diagnostics with high clinical impact that will predict aggressive growth, metastatic potential and responsiveness to several types of therapy. Selection, collection and management of complex data
Because of the challenge we face in understanding HNSCC and in improving the treatment for this cancer, we designed our program to collect as much information as possible on each patient and piece of tissue. We considered the platforms available with the idea that each platform should provide unique data that could be analyzed in the context of the clinical data and could be integrated with data from other platforms to complement our understanding of the pathophysiology and the responses of individual patients. For example, results from RNA expression profiling are commonly used to classify tumors; however, integration with results from CpG island methylation data will provide better insight into gene silencing and potential new therapeutic targets. Alternatively, expression of RNA may not lead to protein expression if miRNA is present. In addition, the activity of certain proteins is regulated posttranslationally without major changes in RNA expression. Thus having genomic and proteomic data for each cancer will enable us to identify aberrant pathways in the cancer and, in the context of the genome of each patient (e.g., prev-
alence of single nucleotide polymorphisms), will allow us to integrate data across platforms to identify key factors in disease progression and patient response. The sections below give our considerations and strategy in setting up infrastructure and our experience in data collection with respect to the field. Design considerations for data and sample collection Data management challenges. Two areas where epidemiological and biostatistical expertise are indispensable are in experimental design and data analysis requiring interaction with the investigators in all stages of the research, beginning with the formulation of the research question, through the experimental design stage and data collection stage and quality control and data management (including regular reports to monitor missing data, out-of-range values, inconsistent answers, etc.), to the generation of analytical datasets and data analysis and interpretation and dissemination of results. This also extends to the collection and integration of clinical, laboratory and epidemiological data necessary to conduct comprehensive statistical analyses on microarray platform data and account for the heterogeneity inherent in head and neck cancer risk data. In order to handle such complex and multi-source data, existing epidemiological studies now often develop and depend on extensive web-based data collection systems to recruit and collect participant study data. The advantage is that all data collected through this systems-based approach can be tracked, queried and even analyzed through automated processes, often in real time, while also allowing interactive and flexible interfaces online for investigators (Fig. 3). The head and neck cancer program website at Albert Einstein College of Medicine (AECOM) was developed for this purpose (http://www.aecom.yu.edu/headandneck/). The development of such a system, however, requires a truly multi-disciplinary team including data managers, network administrators and database programmers, in addition to the requisite clinical and scientific investigators. Further incorporating microarray platform data like that being generated at AECOM, demands the additional expertise of bioinformatics professionals experienced in manipulating and tracking multiple data forms including microarray, proteomic and genomic data. Finally, the success of an integrated research program, such as ours, necessitates the establishment of quality control procedures and standardized protocols by experts in all fields and effective communication between all participants. Quality control and validity analysis. In population studies involving microarray hybridization techniques, a second assay is routinely employed to verify the results obtained from this high-throughput approach. For example, methylation-specific PCR (MS-PCR) is one of the most trusted means of determining the methylation status of a given CpG island region. The choice of MS-PCR is also an alternative to the use of methylation-specific restriction enzymes, as used in our MSRE microarray assay, that is known to be less prone to false-positive results due to the inclusion of primer sets that serve as a control for the efficiency of bisulfite mod-
Cytogenet Genome Res 118:92–109 (2007)
95
ORL-HNS
Pathology Clinical database De-identification Hospital Firewall
a
b
H&N information system Relational database and web application c
e
High throughput Genomic data Proteomic data Systems biologist
e
d
Project specific data e
Biomedical scientist
Epidemiologist Biostatistician
f Integrated data analysis and model development
Fig. 3. Head and Neck Cancer Project Information System. a, b: Clinical data collected by Otorhinolaryngology-Head and Neck Surgery (ORL-HNS) including tumor staging, treatment, pathology, tissue inventory and follow-up visits are collected and entered into the project database behind the hospital firewall. These data are available to healthcare providers caring for the patients. Relevant data is de-identified and downloaded onto the relational database; c: raw datasets from genomics, proteomics and tissue microarray platforms are uploaded into the database and linked by a unique (de-identified) number to the clinical data; d: project specific data is uploaded into the database and linked by a unique (de-identified) number to the clinical data; e: data entry, queries and reports can be accessed in real time through a web portal that is password protected and customized for the specific project, also annotated and integrated datasets are generated for correlative and prognostic biostatistical analyses; f: integrated data analysis and model development in collaboration with experts covering various biomedical disciplines. Blue arrows indicate data flow and green arrows indicate communication.
ification (Liu and Maekawa, 2003). The sensitivity of MSPCR would allow the detection of hypermethylated CpG islands even if they contribute relatively little to the overall DNA sample. Furthermore, as it is also known that hypermethylation of CpG island promoter elements is often associated with the transcriptional silencing of associated genes, the examination of the effects of specific CpG island methylation on expression of neighboring genes offers an opportunity to add to the list of genes transcriptionally silenced by DNA methylation, and to identify new tumor suppressor genes in head and neck cancer. Using online software at NCBI, such as Map Viewer, island boundaries can be positioned by aligning partial sequences from these islands against the genomic sequence (Dombrowski and Maglott, 2003). Assessing the reliability of microarray data. Additional quality control measures employed in systemic approaches to population studies like ours include: (i) the collection of paired samples from the same patient (i.e., biopsy and resections or tumor and normal), (ii) the collection of tissue through multiple sources (i.e., fresh frozen and paraffinembedded tissue) and (iii) multiple samples from the same tissue (i.e., collecting multiple cores from the same tumor for constructing tissue arrays) or running multiple microarray assays on the same tumor specimen.
96
Cytogenet Genome Res 118:92–109 (2007)
For example, when validating our MSRE microarray assay for DNA methylation, where the quantity of amplified PCR product depends upon complete digestion of DNA by the restriction enzyme, we estimated the reproducibility of the MSRE assay by intraclass correlation (ICC) for three tumor samples, with four replicate experiments each, using normalized log-transformed ratio values (Adrien et al., 2006). Samples represented stage IV squamous cell carcinomas derived from the hypopharynx (tumor 1), larynx (tumor 2) and oropharynx (tumor 3). The ICC assesses reproducibility in a given assay by comparing the variability of different ratings of the same clone to the total variation across all ratings and all clones (Shrout and Fleiss, 1979). As such, the ICC decreases in response to both lower correlation between repeated assays and larger mean differences between assays. We also evaluated the gain in reliability achieved by filtering the array results based on fluorescence intensity and used these results to select optimal cutoff values for filtering. The overall reproducibility of the MSRE assay was high, with baseline (unfiltered) ICCs across the three tumor specimens of 0.68 to 0.85. ICC values close to 1.0 indicate that a larger proportion of the variance observed in fluorescence ratios is due to variability between rather than within CpG islands. We also assessed the change in ICC under different filtering conditions based on fluorescence intensity. Interestingly, the largest change in ICC occurred after restricting observations to results with a fluorescence intensity of 200 or more. We also evaluated whether the reliability of the fluorescence ratios was a function of their location on the microarray slide. Sites were restricted to individual grid blocks, each containing 525 clones, which can be identified on the microarray grid. Overall, differences in ICC between blocks were small (!0.05) and showed no pattern by location on the grid. We attempted to estimate the number of repeated assays that would be needed to achieve a given level of reliability without filtering. Using a variant of the Spearman-Browne prophecy formula (Shrout and Fleiss, 1979), we calculated that we would need to repeat our assay twice at most to obtain an average reliability of 0.8 and five times to achieve a reliability of 0.9. Pathology and tissue procurement Tissue procurement. Optimal procurement, processing and distribution of solid human tissue requires close coordination among surgeons, pathologists and clinical and laboratory investigators. Once IRB approval has been obtained, the pathologist works with investigators and technical staff to develop a project-specific protocol for tissue procurement and distribution (Fig. 4). The coordinated process insures that sufficient specimen is available for diagnosis and that tissue delivered to the investigator has a histopathologic diagnosis (Neiderhuber, 2006). In addition, appropriate staff familiar with each protocol are aware of special requirements for handling, processing and distribution of tissue. The research specimen is snap frozen in liquid nitrogen and stored in a –80 ° C freezer. In selected situations where there is clearly ample tissue for a diagnostic speci-
men, fresh tissue may also be taken for cell culture or transplantation into immunodeficient mice. A separate portion of the selected research specimen is prepared for standard histology. This is a quality assurance measure that documents and insures the nature of the specimen procured. No tissue is released to a research laboratory until a final pathologic diagnosis is rendered. In the event of a discrepancy (e.g., there is a clinical suspicion of cancer, and the biopsies designated for diagnosis are negative), the snap frozen tissue is retrieved and processed for histopathology. The full integration of clinical and translational research operations yields the highest quality of tissue data. Tissue microarray. Tissue microarray (TMA) is the process of utilizing needle cores of paraffin embedded pathology specimens to construct paraffin microarray blocks (Rimm et al., 2001). These blocks are then used for immunohistochemical (IHC) analysis. The advantage of this technique is that one block can be designed to hold multiple cores from scores of patients, thus compressing the informative potential to a single TMA slide, facilitating high throughput analyses while preserving the limited resource of patient samples. The microarrayer holds the pathology specimen block (donor block), the TMA block in preparation (recipient block) and is coupled to a microscope holding the corresponding pathology slide. Selecting the exact areas of interest on the pathology slide (tumor versus non-tumor areas) allows donor cores to be harvested from the corresponding area in the paraffin block. Three or more cores are harvested from the areas of interest. TMA IHC analysis is an important validation arm of molecular profiling. It allows us to (i) correlate RNA expression and gene silencing results with resultant proteomes, (ii) identify the cells with significant expression alterations (tumor versus stromal cells), and (iii) localize subcellular expression (nuclear vs. cytoplasmic vs. membranous staining patterns). The latter two issues can only be addressed at the in-situ tissue level. Protein expression may be discordant with respect to RNA expression due to reasons stated above. Alternatively, intracellular localization may affect function and tumor behavior as has been shown for ERM proteins (Kobayashi et al., 2004; Madan et al., 2006). An ultimate goal of biomarker discovery is to apply a panel as a routine diagnostic test. IHC analysis is a simple analytic tool for biomarker detection that can be used daily in surgical pathology reporting. Thus biomarker discovery, and its validation by TMA IHC, can lead back to its application as a routine IHC diagnostic test. Histopathology. Histopathology has a crucial role in initiating treatment decisions as it confirms the diagnosis of malignancy. There are many elements of the histopathologic analysis that predict tumor behavior (described below); however, since the initial analysis is performed on a small biopsy specimen, there is limited potential for gleaning further significant information (e.g., perineural invasion which would foretell the need for adjuvant therapy). If the decision is made to treat the patient surgically, then the histopathologic findings from the resection have greater informative potential. The pathology report establishes the
Hypothesis
Study design
Biopsy/resection
PRC and IRB approval
Brush biopsy
Pathology 1-Routine pathology 2-Tissue distribution 3-Tissue microarray
Tissue processing 1-Separation of macromolecules 2-Storage of samples
Viable tissue 1-Implantation 2-Cell culture
Viable tissue bank
Clinical database Pathology database
In vitro drug testing
Genomics 1-RNA expression 2-DNA hypermethylation analysis Proteomics 1-Global proteomics 2-Targeted proteomics
Merged data for analysis
Genomics database Proteomics database
Fig. 4. Tissue procurement and distribution for head and neck cancer program. The highly interactive, multidisciplinary group of investigators works together to develop study design and documents for regulatory approval. Tissue procurement protocols are developed and coordinated between Surgery, Pathology and individual investigators. Tissue processing and banking are performed in pathology laboratories. High throughput genomic and proteomic analyses are performed in shared resource laboratories at Einstein Medical School. Data collection and analysis is carried out as described in Fig. 3.
AJCC tumor stage with respect to primary site and nodal status. The reporting of inadequate resection margins or perineural invasion will lead to adjuvant therapy. The finding of extranodal tumor extension will lead to increased radiation dosing to the neck. But what further insight into tumor biology can be derived from studying a resection specimen? Single histological variables such as tumor grade, mitotic rate and degree of keratinization have been disappointing with respect to providing additional prognostic information. Tumor thickness for T1 tongue SCC has strongly associated with propensity for lymph node metastasis, and survival (Martinez-Gimeno et al., 1995), but is not a predictor for local recurrence (Kim et al., 1993; Po Wing et al., 2002). Multiparameter prognostic histological assessments have been developed and refined over the last two decades, based on variables that include pattern of invasion, degree of keratinization, nuclear pleomorphism, lymphocytic response and mitotic rate (Anneroth et al., 1987; Bryne et al., 1989, 1992, 1995; Kim et al., 1993; Spiro et al., 1999; Bundgaard et al., 2002). Pattern of tumor invasion (POI) refers to the manner in which cancer infiltrates tissue at the tumor/ host interface. It is intuitive that neoplasia infiltrating in a widely dispersed manner is more aggressive than those growing in a bulky pushing fashion. POI alone, and as part of weighted scoring systems, has been demonstrated to predict local recurrence and overall survival. We have further refined the histological definitions of POI to recognize a subgroup with greater aggressive potential and we devel-
Cytogenet Genome Res 118:92–109 (2007)
97
oped a risk score model that is highly predictive of local disease-free and overall survival when corrected for resection margin status and adjuvant treatment (BrandweinGensler et al., 2005). This risk model can be assessed by examining the resection specimen and quantifying three significant histological variables: 1) tumor pattern of invasion (POI) at the advancing tumor edge, 2) perineural invasion, and 3) lymphocytic host response at the patient interface. Importantly, this risk assessment identified a subset of patients who benefited from post-surgery adjuvant therapy (‘high-risk’ patients) despite the absence of traditional indications; these were patients with Stage I or II cancers, negative resection margins and no perineural invasion. Patients with high-risk scores including perineural spread would have received adjuvant therapy based on the finding of perineural tumor invasion by usual treatment paradigms. But patients with an intermediate degree of lymphocytic host response at the interface and either a type 4 or 5 POI represent an example of a newly defined high-risk category. We are prospectively validating this ‘risk assessment’ model on a new patient cohort. Our histological risk assessment is derived from reading a patient’s primary resection specimen; it cannot be reliably determined from preoperative biopsies. Is this valuable and simple predictive tool limited to patients who are treated with primary surgery? A future direction will be to use the risk scoring to identify surrogate biomarkers. Filtering RNA expression data by the risk grouping of the corresponding resection specimens will make possible the identification of biomarkers that associate with low-, intermediate- and high-risk groups. We believe that a panel of biomarkers derived in this manner may predict treatment response with respect to the initial decision bifurcation of primary surgery versus organ-sparing protocol. This panel of surrogate biomarkers will reflect risk score but can be determined at initial patient workup. Thus we return to the issue of limited informative potential gleaned from the initial diagnostic biopsy. Our goal is to push this envelope and identify and validate a prognostic biomarker panel that has informative added value to traditional TNM prognostication. Epigenetic alterations and head and neck carcinogenesis Whole genome approaches using high-throughput methodologies such as DNA microarrays have permitted the classification of many tumor types, including HNSCC, based on parameters such as the patterns of global gene expression or chromosomal aberrations. The results of these studies have revealed many new molecular ‘signatures’ for previously indistinguishable subtypes of these diseases. These and other technologies have now been extended to include so-called ‘epigenetic’ changes, as it is now well known that such alterations also contribute significantly to the onset of human malignancies. By far, the most studied epigenetic event has been the addition of a methyl group to the carbon-5 position of cytosine nucleotides. Of particular interest are the aberrant DNA methylation events associated with CpG-rich sequences, known as ‘CpG islands’, which
98
Cytogenet Genome Res 118:92–109 (2007)
range in length from several hundred to several thousand nucleotides and are frequently associated with gene promoters or first exons. In certain diseases, including cancer, aberrant hypermethylation of these CpG islands is associated with the inappropriate transcriptional silencing of critical genes. Such transcriptional repression can therefore act as one of the hits in the Knudsen two-hit model for tumor generation by abrogating gene function in a manner similar to a mutation or gene deletion but without any change in the nucleotide sequence (Robertson, 2001). As for other cancers, epigenetic markers are of prognostic value to HNSCC. Numerous studies have identified promoter methylation of CDKN2A (p16), DAP kinase (DAPK), and DNA repair genes MGMT and MLH1 (Rosas et al., 2001; Maruya et al., 2004). In HNSCC patients, methylation of the promoter region of MGMT was associated with decreased expression of MGMT, as well as increased tumor recurrence and decreased patient survival independent of other factors (Zuo et al., 2004). In studies by Ogi et al. (2002) with oral SCC, methylation of the DCC gene was significantly associated with bone invasion by gingival tumors, aggressive invasiveness of tumors of the tongue and reduced survival. Methylation of two CpG islands (MINT1 and MINT31) also correlated with poor prognosis in these patients, whereas methylation of p14ARF actually correlated with a good prognosis. Invasion and metastasis of oral SCC cells have recently been shown to be dependent on methylation of the E-cadherin promoter with associated reduction of E-cadherin expression (Kudo et al., 2004). In addition to the value of individual epigenetic events, there is also the potential that overall patterns of genomic DNA methylation may play a critical role in the molecular characteristics of HNSCC. For example, it is known that a subset of colorectal tumors with widespread DNA hypermethylation (the so-called ‘CpG island methylator phenotype’, CIMP) have distinctive clinicopathological and molecular characteristics (Bariol et al., 2003). Recent studies have now shown that this phenotype is also an independent predictor of response to 5-fluorouracil (5FU) and survival for patients with stage III colorectal cancer (van Rijnsoever et al., 2003). It has been suggested that a CIMP may also exist for HNSCC (Rosas et al., 2001) but more studies will be required to confirm its existence for HNSCC and any clinicopathological consequences for the patient. On a global ‘epigenomic’ scale, new technologies to study global promoter hypermethylation events in human malignancies have also revealed a great deal of information. For example, technologies such as restriction landmark genome scanning (RLGS) have demonstrated that overall patterns of aberrant CpG island methylation are tumor-type specific (Costello et al., 2000). Similarly, microarrays containing thousands of CpG island clones have been utilized to generate profiles based not on gene expression, but on the DNA methylation status of each CpG island clone. Studies by our group using primary HNSCC tumor DNA have revealed that such aberrant DNA methylation events affect hundreds of CpG island clones. Furthermore, these DNA methylation profiles of primary HNSCC tumors are tumor-specific and can be uti-
lized to distinguish individual tumors from one another in a reproducible manner (Adrien et al., 2006). And finally, classification of HNSCC tumors based on patterns of DNA methylation appeared to reveal three distinct epigenomic subtypes of the disease, with clustering predominantly influenced by the number (high, medium, low) of methylated CpG island clones observed for each cluster of tumors. Additional statistical approaches are now underway to evaluate the clinical implications of these DNA methylation profiles for HNSCC. These and other findings have laid the groundwork for population-based studies to examine DNA methylation patterns in human malignancies and helped to identify associations between specific epigenetic signatures and clinical parameters. Already, epigenomic profiling techniques have revealed that CpG island methylation is associated with the histological grades of breast tumors (Yan et al., 2000). A similar study of late-stage ovarian carcinomas with CpG island microarrays revealed that a higher degree of CpG island methylation was significantly associated with early disease recurrence following chemotherapy (Wei et al., 2002). Most importantly, this study also identified a select group of CpG island loci that could be used as epigenetic markers for predicting outcome in ovarian cancer patients. The integration of these events with other systematic data (chromosomal deletion maps, microRNA) is a key to understanding DNA methylation profiles and elucidating transcriptional silencing mechanisms as a convergence of epigenetic-silencing, RNA interference and chromosomal events. Aberrant DNA methylation events remain an exciting molecular marker for future investigations, due to the fact that it represents a stable tumor-specific marker that often occurs early in tumor progression and can be easily detected by PCR-based methods. Such PCR-based detection methods for diagnostic and prognostic biomarkers would be minimally invasive to the patient and would require trace amounts of genetic material. Such approaches have already been utilized to detect aberrant DNA methylation in the saliva of HNSCC patients (Rosas et al., 2001). Gene expression profiling and the evolution of HNSCC tumor classification The vast majority of genomic profiling and molecular classification of all tumor types, including HNSCC, has involved high throughput methodologies to measure differential gene expression. One of the best known of these technologies is the DNA microarray, known also as the ‘DNA chip’ or ‘gene chip.’ Many of the early investigations in HNSCC research used microarrays for a comparison of two biological states. In the case of cDNA microarrays, RNA was extracted to compare tissues of two different states (e.g., a pair of normal and tumor tissues) and labeled with distinct fluorescent dyes. By combining and simultaneously hybridizing these labeled products to their corresponding cDNAs on the microarray, one could assess the relative abundance of thousands of genes in parallel simply by measuring the fluorescence ratio of the two dyes.
Many of the studies involving DNA microarrays have focused on the identification of alterations in gene expression that are associated with HNSCC carcinogenesis, specifically the identification of genes whose expression has changed in HNSCC tissue samples compared to normal tissue. For example, in an initial study Villaret et al. (2000) using microarrays containing 985 clones, examined sixteen HNSCC cases and identified nine over-expressed genes. A later study, utilizing laser capture microdissection of HNSCC cells to measure the expression of 588 known cancer-related genes, demonstrated increased expression of genes related to the Wnt and Notch signaling pathways, as well as a decrease in expression of differentiation markers such as cytokeratins (Leethanakul et al., 2000). Using a similar study design with cDNA expression microarrays (Belbin et al., 2005), our group examined gene expression changes in primary HNSCC versus adjacent mucosa, as well as primary HNSCC versus metastatic carcinoma. One of the genes identified as consistently increasing in expression during tumor progression was moesin, a member of the ERM (ezrin; radixin; moesin) family of closely related membrane cytoskeletal linkers known to regulate cell adhesion (reviewed in Mangeat et al., 1999). In addition to gene discovery and mechanistic studies elucidating the roles of altered gene expression in carcinogenesis, the use of microarrays permits the identification of distinct ‘molecular signatures’ of tumor cells that are of prognostic value. This approach has been applied successfully for HNSCC, demonstrating that tumors originating from different anatomic sites within the head and neck can exhibit varying behavior that is not predictable by histopathology of the primary tumor but is discernable by gene profiling (Belbin et al., 2002). In oral cancer, 23 differentially expressed genes were identified that correlated with tumor stage (III–IV) and metastasis (Warner et al., 2004). In one of the early studies correlating gene expression profiling and treatment response, Hanna et al. (2001) identified 60 tumor-related genes from a cDNA microarray containing 1,187 genes that could successfully predict tumor response to radiation. Gene expression signatures have now been identified that are associated with recurrence of HNSCC disease (Ginos et al., 2004). While these initial findings are promising, they have not been accepted widely for use in managing patients with HNSCC. Microarray experiments designed to examine differential gene expression or molecular tumor classification each face unique issues, both with experimental design and with measurement quality (Cheung et al., 1999; Novoradovskaya et al., 2004). Overall, extensive data collection and analysis by our group confirm that cDNA microarrays represent a reproducible and accurate high throughput measurement tool for gene expression profiling. Nevertheless, in the case of direct differential gene expression profiling, validation using technologies such as real time PCR, Northern blots and immunohistochemistry have become common. It is noteworthy that differential gene expression profiling is complicated by the inclusion of non-tumor cells in tumor specimens and precancerous changes in so-called ‘normal’
Cytogenet Genome Res 118:92–109 (2007)
99
mucosa. Differential expression profiling also does not reveal information related to post-translational modification or subcellular localization that may be critical to the specific functioning of the translated protein. In fact, the combination of elevations in expression of certain genes in cells, with their translation into proteins and the localization of these gene products, may confer additional information that reflects pathogenesis. A systems approach is therefore critical to integrating diverse forms of collected data and understanding the mechanisms at work. For example, in the identification of upregulation of moesin described previously, it was the cytoplasmic localization of moesin that was subsequently shown to be significantly associated with metastasis and poor survival in HNSCC patients (Kobayashi et al., 2004). Similarly, our group has shown that cytoplasmic ezrin, another member of the ERM family of proteins, may have prognostic significance in this disease (Madan et al., 2006). Overall, the identification of ERM family members provides an example of the rapid appraisal of molecular markers that is possible with direct comparisons using new genomics-based technologies such as DNA microarrays coupled with validation studies using tissue microarrays. For molecular classification studies with a population of tumor samples, it is becoming increasingly clear that analysis of tumor gene expression patterns has a valuable predictive and prognostic capacity. In fact, one of its most promising aspects is the possibility of isolating smaller subsets of genes whose expression (or lack thereof) correlate significantly with clinical parameters, and the ability to construct a disease-specific microarray chip that can be used to predict clinical outcome in the future. However, in addition to issues mentioned above, care must be taken in experimental design to maximize the ability to discriminate between members of the tumor population. While direct comparisons are commonly used in simple microarray hybridization experiments, this approach quickly becomes impractical when trying to compare differential gene expression in a larger population of samples. Under these circumstances, the most popular approach is to co-hybridize each sample with a common ‘reference’ sample and then infer differences in gene expression among the sample population via an ‘indirect’ ratio measurement (Yang and Speed, 2002). In some studies, such as the comparison of gene expression profiles in a population of tumor samples, the choice of a reference RNA to use is not immediately obvious. One of the more common approaches in use today is a reference pool of RNA produced from a collection of tissues or established cell lines representing a variety of cell types (Puskas et al., 2002). Studies by our group indicated a strong overall correlation between direct and indirect fluorescence ratio measurements with cDNA microarrays, but also a high degree of false positives in our indirect measurements (Belbin et al., 2004). These results indicated that the applications of more stringent ratio filters are required when assessing differential gene expression utilizing a common RNA reference in classification studies. The choice of a reference sample close in overall gene expression profile to those samples within the tumor population also maximizes the probability of dis-
100
Cytogenet Genome Res 118:92–109 (2007)
criminating subtle gene expression differences among the sample population that may have potential as a prognostic predictive gene expression signature (Cao et al., 2004). Relating gene expression to HPV: an example of impact of etiology Epidemiological and laboratory evidence now warrant the conclusion that, in addition to tobacco and alcohol, human papillomaviruses (HPV) play an etiologic role in some HNSCC (Gillison and Shah, 2001; Schlecht, 2005). To characterize the molecular profiles of HPV-positive head and neck cancer, we compared differences in gene expression patterns between HPV-negative and -positive HNSCC tumors using cDNA microarrays. Tumor samples were collected from histological confirmed HNSCC patients undergoing treatment at Montefiore Medical Center and AECOM affiliated medical center in the Bronx. HPV detection and genotyping was performed by PCR using several primer protocols to specifically detect multiple types of HPV. Total RNA was extracted and purified from frozen tumor samples and gene expression levels were assessed using our 27,323 cDNA microarray chip compared to a universal human reference library of RNA (http://microarray1k.aecom.yu.edu) (Cheung et al., 1999; Novoradovskaya et al., 2004). To identify genes that were differentially expressed between HPV+ and HPV– tumors, statistical analysis of tumor samples based on gene expression profiles was performed (Eisen et al., 1998; Tusher et al., 2001; Tibshirani et al., 2002; Slebos et al., 2006). Genes with differing expression profiles between HPV+ and HPV– tumors were then identified. While some authors have reported a higher prevalence of HPV in tumors of non-smokers (Lindel et al., 2001; Herrero et al., 2003), as we observed in our population, others have reported that individuals who are positive for HPV and who smoke are at greater risk of HNSCC than individuals presenting with either exposure alone (Schwartz et al., 1998; Smith EM et al., 2006). However, both additive and competing relationships have been reported between tobacco smoking and HPV in HNSCC (Smith EM et al., 2004; Strati et al., 2006). Therefore, given the putative association between smoking and anatomic site, and the reported disparate etiologic pathways for tobacco carcinogens and HPV (Califano et al., 1996; Gillison and Shah, 2001; Ragin et al., 2006), we assessed whether the observed gene expression profiles also differed between head and neck cancer tumors as a result of either tobacco exposure or HPV infection. To separate the tobacco- and HPV-induced effects on HNSCC pathogenesis, we segregated patients on their smoking history, reducing the analyses to smokers and never-smokers, and therefore removing potential bias from our analysis of HPV expression signatures. Among the most prominently up-regulated genes in HPV16+ tumors of nonsmokers were cyclin-dependent kinase inhibitor 2C (CDKN2C) and the retinoblastoma (RB1) gene. CDKN2C encodes the p18 tumor suppressor protein, a cyclin-dependent kinase inhibitor of the retinoblastoma tumor suppressor protein (pRb). Others have also found CDKN2C to be
up-regulated in HPV16+ HNSCC derived cell lines when compared to normal oral tissue (Martinez et al., 2006). In addition, we observed differential expression of several genes in our analysis of HPV16+ tumors of smokers and nonsmokers combined that are known to be regulated by p53 proteins or E2F transcription factors including replication factor C4 (RFC4), cell division cycle (CDC7), and transcription factor E2F dimerization partner 2 (TFDP2). RFC4 encodes an ATP-binding domain of the heteropentamer RFC (also called activator 1), an accessory protein required for the elongation of primed DNA templates by DNA polymerase (Lee et al., 1991) that is not responsive to E2F signaling (Angus et al., 2004). It was particularly interesting that restricting analyses to never-smokers revealed the most overlap with previously identified HPV-induced gene profiles from cervical SCC. A role for microRNAs in head and neck squamous cell carcinoma There has been a great deal of evidence to indicate that microRNAs have an important role in the regulation of cell division and differentiation (Lee et al., 1993; Lagos-Quintana et al., 2001; Lau et al., 2001; Lee and Ambros, 2001) and through these key processes play a part in numerous cancers (Lu et al., 2005; Calin and Croce, 2006a; Cummins and Velculescu, 2006; Zhang et al., 2006). MicroRNAs are a recently discovered group of non-protein coding short 21–23 nucleotide RNA molecules that are expressed in relative abundant quantities in a tissue specific and stage specific manner. They are highly conserved throughout the animal and plant kingdoms and there are several hundred to possibly thousands of miRNA genes predicted from the human genomic sequence. Individual miRNAs regulate the stability and translation of target genes that have homologous sequences within their transcribed mRNAs using a cellular mechanism similar or identical to siRNA inhibition (Couzin, 2002). Several groups have looked at the global patterns of expression of miRNAs in a number of different cancer models (Lu et al., 2005; Calin and Croce, 2006a; Cummins and Velculescu, 2006; Pallante et al., 2006; Yanaihara et al., 2006; Zhang et al., 2006). Malignant cells have been shown to express altered patterns of miRNAs relative to their normal cellular counterparts and some of these differences can be used to predict clinical properties of individual samples (Lu et al., 2005; Calin and Croce, 2006b; Pallante et al., 2006; Volinia et al., 2006; Yanaihara et al., 2006). In fact, in one example where a direct comparison between miRNA expression and mRNA expression in tumor samples was possible, the miRNA data was more reliable at predicting the clinical data (Lu et al., 2005). We speculate that this might be due to the fact that each miRNA is a regulatory molecule that specifically impacts the downstream expression of many proteins and will make a larger impact on the phenotype of a cell than a typical mRNA encoding a single protein that may not itself be regulatory in function. We have developed a unique microarray to measure the global expression patterns of miRNAs in HNSCC. The ar-
ray consists of triplicate spotting of most of the known human miRNAs from Ambion Inc. as well as the complete set described by the Croce laboratory which using our assay system allows measurement of both sense and anti-sense expression. In addition, a control oligonucleotide was spotted that allows measurement of labeling efficiencies and recovery of fluorescent probes. We have collected miRNA expression data from about a dozen tumor-normal pairs. These same clinical samples were used to generate global mRNA expression and CpG methylation data (see other sections of this review). We can conclude that like other tumor models, HNSCC samples each express altered patterns of miRNAs when compared to matched normal tissue. In addition, it is clear that there are distinct differences among the different tumor samples as well. It is these differences that we are currently using in an attempt to correlate with clinical parameters of the tumor samples. Additional samples will be studied to obtain statistically significant results. We are also investigating potential interactions between methylation and chromatin structure and miRNA expression. Jones and colleagues have observed that tumor cell lines treated with 5-azacytidine alone or in combination with the histone deacetylase inhibitor 4-phenylbutyric acid activate a limited subset of miRNAs and one in particular, miR-127, can inhibit the expression of the proto-oncogene BCL6 (Saito et al., 2006). We have treated the HNSCC cell line FaDu with 5-azacytidine and collected samples 24 and 48 hours post treatment for both mRNA expression microarrays and miRNA microarray analysis. Preliminary results demonstrate that two miRNAs previously shown to be underexpressed in breast cancer or having anti-apoptotic effects are activated in this HNSCC cell line. It is not uncommon to have important regulatory mechanisms interact with one another to fine-tune or feedback the regulation of downstream pathways. The ability to measure miRNAs, methylation patterns, mRNA levels and proteomics from the same samples should allow us to better understand the differences between HNSCC and its normal counterparts on a mechanistic level. Global proteomics The expression of many proteins has been studied and several of them have been proposed to have potential prognostic significance because of differential expression associated with clinical outcomes such as metastasis (van de Vijver et al., 2002) and overall survival (Villaret et al., 2000). In recent years, sophisticated analytic techniques have been applied not only to identify protein biomarkers that can discriminate tumor behavior (invasiveness, metastatic potential) but also to predict response to therapy and to determine the presence or recurrence of disease. The application of high throughput approaches for assessing tumor behavior permits the analysis of many proteins from a single cancer using antibody arrays (Xiao et al., 2002) or enables the measure of a potential biomarker in multiple cancers using tissue microarrays (Yonemura et al., 1998; Yarbrough et al., 2006). Perhaps the most elegant approach for identifying
Cytogenet Genome Res 118:92–109 (2007)
101
new prognostic biomarkers is the profiling and imaging of proteins directly in fresh frozen human cancer tissue by mass spectrometry (Yonemura et al., 1999). Alternatively, data from extracts of tissue analyzed by SELDI-TOF-MS (surface-enhanced laser desorption/ionization time-offlight mass spectrometry) have revealed differentially expressed proteins between healthy mucosa and HNSCC (Roesch-Ely et al., 2006). These new approaches should facilitate the identification of prognostic biomarkers. Currently no protein biomarker has been identified that can reliably detect the presence of HNSCC in serum or saliva. However, the recent application of high throughput mass spectrometry approaches has attempted to identify prognostic and predictive signatures/patterns in the serum to be used as a screen for individuals developing disease or recurrent disease (Yoo et al., 2002). While this approach has been fraught with problems in general (Yu et al., 2002), when properly applied, it holds great promise for identifying peptides/proteins that can distinguish tumor behavior and potentially response to therapy or early detection. Analysis and integration of data sets
Standard approaches Normalization of microarray data. Approaches to preparation and normalization of microarray hybridization data have been largely standardized now. One common approach, applied to RNA expression microarray data at our facility, is to compare gene expression patterns between epithelial mucosa (Cyanine 5-labeled sample) and a Universal Human Reference Sample (UHR) (Stratagene, CA, USA) (Cyanine 3-labeled sample) (van Rijnsoever et al., 2003; Novoradovskaya et al., 2004). The ratio of the fluorescence intensities of the two dyes therefore represents a measure of differential gene expression between the two samples of interest. Red (Cy5) and green (Cy3) signal intensities for each element on the array were calculated using GenePix Pro 3.0 software. This software gives an integrated intensity per spot for each channel in addition to an integrated background count. In order to establish consistent ratios, we first compute an intensity dependent normalization factor for each microarray experiment by finding the rank invariant subset of the spots (the spots that have equal or almost equal ranks in the two channels) (Tseng et al., 2001). Once this core of spots has been identified, a robust curve is fitted using the ‘lowess’ function from the R statistical package (Yang et al., 2002). We then compute the intensity dependent normalization factor where A is the geometric average intensity. Data designated to be of bad quality, which did not achieve a signal to noise ratio of at least two-fold, are discarded from subsequent analysis. Overview of microarray data analysis approaches and planning. One of the primary goals of the informatics part of the studies is the identification of epigenetic, proteomic or expression events that can serve as markers or predictors of various clinical subtypes. For initial statistical analysis of
102
Cytogenet Genome Res 118:92–109 (2007)
the microarray data, we use clinical features including histopathologic factors (i.e. advanced stage, lymph node metastasis, etc.) at diagnosis and clinical outcome (i.e., recurrence, progression-free and overall survival) as clinical indicators of disease aggressiveness. However, given the shear volume of data generated for each patient and cancer, it is not unusual to anticipate that a subset of genetic or epigenetic events will be significant with respect to patient outcome. The size of this subset is unknown; it may be the case that a single genetic event or hundreds of events may need to be simultaneously considered. Therefore, the first phase of supervised learning consists of training/learning of the statistical prediction model from a data set that has been classified; in other words, the class labels have been assigned to the samples. This data set is called a ‘training’ set. Once the model has been trained, it can then be used to predict (classify) data that has not been assigned to labels. In principle, one can first collect a training set, train the model and then collect additional data to be used as a ‘test’ set to see how well the model works. As a result, there are two very different error rates that can be calculated: the training error and the testing error (more often called the generalization error). The generalization error is the performance of the model on data that was not used in the training phase. It is this error that is the more important in that it indicates how well the model will work in prediction (classification) of data. One of the problems in supervised learning techniques, however, is the risk of overtraining or overfitting the model. The result would be a model that is overdetermined in terms of the parameters it has and, in a sense, overfits the specific data on which the algorithm was trained. In extreme cases, one can actually perfectly predict all of the training examples. What one then has learned is not some generic mapping of the input data (i.e., DNA methylation) to the target data (recurrence and survival) but rather specific features of a particular input dataset. Such overtrained models can have very poor ‘generalization’ error. The problem of overtraining or overfitting is a special concern in the supervised analysis of microarray data because of the relatively large number of input variables compared to the number of clinical samples that are practically available. Therefore, prior to training the model, one should first handle the problem of overfitting using a ‘feature selection process’ that will select on a subset of CpG island clones, expressed genes or proteins and peptides to be assessed by alternate methods for confirmation. Support vector machines, linear discriminant analysis, classification and regression trees (CART) and principal components analysis (PCA) are some approaches to achieve this. A review of some of these methods is provided in the next section. Finally, we validate the results of the supervised learning using cross validation procedures that enable us to estimate the generalization error of the selected genes. In fact, it is also possible to iterate over all of these steps (feature selection, training, cross validation) several times in order to improve the overall performance of the model.
Planning of studies employing microarray data. There are currently no widely accepted and comprehensive methods for planning sample sizes of microarray experiments. However, for the microarray analysis, a common approach is to employ a sample size determination method by Simon et al. (2002) for testing whether a particular gene is differentially expressed between two pre-defined classes. The basic approach is to apply the usual sample size formulas for comparing two means from normally distributed data, but with Type I (␣) and Type II () error rates that do not exceed 0.001 and 0.05, respectively. This ensures that if 1,000 CpG island clones are equivalent in two groups, the number of false positives will be at most one. Similarly, if 100 CpG island clones are differentially methylated, then the procedure will result in at most five false negative results. Other approaches for addressing the inflation in Type I error rate due to multiple testing will also be considered, including controlling for the false discovery rate (Hochberg and Benjamini, 1990; Reiner et al., 2003; Benjamini and Yekutieli, 2005). Alternative and advanced statistical approaches
Statistical approaches Statistical analysis of expression data might start with establishing the reliability of the approach employed. For instance, in our proteomics study, one of our objectives was to establish the reliability of the protein analysis. For this purpose we employed two HNSCC tumor cell lines (SCC25 and FADU). They provided a scale for the expected biological signal to be measured. This enabled us to determine whether the biological variability was much larger than the extraction and analytical variability, hence enabling us to test the reliability of the protein analysis. This experiment also helped us to identify those peptide fragments that are most robustly measured and that we would consider for further detailed analysis. In this experiment, three extractions were prepared from each cell line, the protein fraction from each of these extracts was fragmented into peptides and analyzed in triplicate by two-dimensional liquid chromatography-mass spectrometry (2D LC-MS) for global proteomics. Use of internal and external protein and peptide standards provided relative quantitation. Each peptide in the final pattern derived from 2D LC-MS had a characteristic mass, elution time from a reversed phase column and a salt concentration from which it is eluted from a strong cation exchange column. In order to quantify the reliability of our approach we employed a random effects ANOVA model. Based on the random effects model, we computed the ratio of the variation due to differences between cell lines for a given peptide to the total variation. We refer to this variation as the percentage biologic variation (Shrout and Fleiss, 1979). A peptide with 690% biologic variation has high reliability in differentiating the two cell lines. We identified peptides that were highly reliable, i.e., showing large biological variation hence discriminating the two cell lines (Shrout and Fleiss, 1979). The ratio of the vari-
ation due to differences between cell lines for a given peptide to the total variation is what we refer to as the biologic variation. A peptide with 690% biologic variation has high reliability. We found 150 peptides of 1 40,000 analyzed with a reliability index (intra class correlation coefficient) 690%. For these 150 peptides the median coefficient of variation was 0.165 (with inter-quartile range 0.14–0.20). This analysis of cell lines showed that the 2D LC-MS protocol could be used reliably to identify biologic variation in HNSCC. In assessing the predictive ability of expression data (microarray or proteomic) standard approaches need to account for the fact that hundreds and even thousands of molecular markers are being considered simultaneously. This leads to the problem of multiplicity of hypothesis which in turn leads to inflation of the Type I error rate, i.e., the probability of rejecting the null hypothesis when it is really true (also called a false positive). For instance, if we were simply to perform uncorrected hypothesis tests at the traditional 5% level, then 10,000 ! 0.05 = 500 false positives would be expected under the full null hypothesis if we were testing 10,000 genes or peptides. This is clearly unacceptable, so some procedure must be performed to control the false positive rate in a reasonable manner. For a single test, the false positive rate can be useful for measuring how likely it is for a true null case to be as significant as what has been observed, as shown in the above example. However, in this setting knowing what percentage of the significant tests is actually false is more informative and this is called the false discovery rate. For instance, if we are willing to incur a false discovery rate of 5%, then this means that among all tests we call significant, about 5% of them will be false positives. If there are 500 significant tests then this results in about 25 false positives. There are approaches that effectively control the false discovery rate (Storey, 2002). Equally important is ensuring reasonable study power, i.e., rejecting the null hypothesis when the alternative hypothesis is really true. This could be achieved by employing an appropriate test statistic as well as recruiting sufficient sample. In a preliminary proteomic analysis of HNSCC, we employed the q-value statistic, i.e., the lowest positive false discovery rate at which a peptide is called significant (Storey, 2002) in assessing whether a given peptide is related to clinical outcomes such as recurrence. It is also possible to refine this approach by taking into account not only statistical significance but also biological importance as assessed by magnitude of effect, i.e., larger fold change or odds ratio or hazard ratios. One may expect for a number of reasons that there may exist correlations between various groups of target biomarkers that collectively will be better predictors/discriminators of clinical outcome. For this reason, other standard approaches such as logistic regression and Cox proportional hazards model could be employed as appropriate in assessing simultaneously the predictive ability of such biomarkers. However, such analysis needs to be accompanied by cross-validation in order to establish the robustness of the results (Simon et al., 2002). Furthermore, such multivariable approaches also allow for the assessment of effect modification by other traditional risk factors and adjust-
Cytogenet Genome Res 118:92–109 (2007)
103
ment for potential confounding factors such as smoking. For instance, smoking itself may be correlated with stage and directly influence length of survival or disease-free survival. It becomes important to rule out confounding by smoking exposure when evaluating molecular studies in the larynx and pharynx in particular, and in many cases, may even require stratification on cumulative smoking status to assess accurately potential differential (interactive) associations with tobacco smoking (Koch et al., 1999). Similar arguments can also be extended to other etiological risk factors for head and neck cancers in general, e.g., alcohol, age and gender. While the evidence for alcohol in oral and pharyngeal cancers varies with respect to sub-site, increased risk associations with alcohol for laryngeal cancers also cannot be ruled out and have been documented by our group and others especially for supraglottic cancers (Schlecht et al., 1999; Kapil et al., 2005). Moreover, increased consumption has been shown to have an effect on the relationship between risk factors like HPV and disease risk (Smith EM et al., 2004). Furthermore, since age and gender can be correlated with stage and independently influence length of survival, it becomes important to see if there is a mediating effect or rule out confounding by such demographic factors in predicting prognosis. Although one may feel reassured that recent large biomarker studies have found independent associations with head and neck cancer after controlling for tobacco, alcohol, age and gender (Gillison et al., 2000; Ritchie et al., 2003; Eriksen et al., 2005), and in some cases among non-smokers (Koch et al., 1999), a concern for residual confounding may remain since these factors are also associated with co-morbid conditions and treatment regimen, which are not ascertained in most studies. As a large number of such potential biomarkers are expected, one might want to restrict the results to the top 5–15 biomarkers on the basis of a bivariate analysis that appropriately takes multiplicity into account. Then one might employ standard model building approaches such as all subset regression. The all subsets regression approach computes all possible models and allows examination of the multivariable effects of a set of peptides. In the multivariable model, one may want to restrict the model building to the best five subsets of candidate peptides depending on the sample size. This is because if the sample size is small, which is usually the case, then increasing the number of candidate peptides in the model would lead to model instability. Using this approach, it is probable that more than one model will be considered as a potential final model due to multicollinearity between the predictors. Dimensionality (a result of having a number of variables much larger than the number of observations) is an issue in this type of analysis and as a result there might be a need to consider other dimension reduction approaches such as principal component analysis (PCA) to determine whether the set of peptides may be reduced to several principal components that describe most of the variation in the data. A potential drawback of PCA is that the summary variables (i.e., principal components) do not necessarily have a clear biological interpretation.
104
Cytogenet Genome Res 118:92–109 (2007)
As an alternative solution to both the problems of dimensionality and biologic interpretation, one can employ classification and regression trees (CART) (Breiman et al., 1984) and their extensions (Ciampi et al., 1995). These approaches recursively partition the space of candidate peptide profiles into subgroups that are highly predictive of the outcome of interest such as metastasis or recurrence. These modeling approaches are more robust than the traditional log-linear models. Moreover, a large number of peptides can be considered simultaneously for searching for and isolating complex patterns and relationships; hence dimensionality is not a problem. These computationally intensive approaches are prone to overfitting. Therefore, it is important to carefully consider model selection approaches (Negassa et al., 2000). Other approaches are also possible, in fact, a wide variety of class prediction algorithms, including complex new algorithms, are being applied in the analysis of expression data. However, in settings where the number of candidate predictors is an order of magnitude greater than the number of cases, complex methods with many parameters often do not perform well when properly evaluated. Comparisons of class predictors varying in degree of complexity demonstrate this finding specifically for DNA microarray data (Dudoit and Fridlyand, 2002). In addition, a Bayesian modeling approach could also be employed in order to deal with the large number of candidate predictors and the uncertainty regarding the choice of the ‘best’ model (George and McCulloch, 1993). Since there are various approaches to the analysis of expression data, it is important to state from the outset whether the study is exploratory or confirmatory and interpret results accordingly. For exploratory studies, use of complete cross-validation of all findings derived from exploratory models is essential. Otherwise, results could be overly optimistic. Various statistical approaches could be interesting for exploration, however, reporting of such results needs to explicitly state this exploratory nature. For confirmatory studies, independent validation, sufficiently large sample size and adjustment for other classic risk factors of the disease of interest are essential. In addition, discussion of the successful adoption of biomarkers into clinical practice should also consider reproducibility in routine practice setting, clinical effects on patients’ outcome, and cost-effectiveness (McShane et al., 2005b). Alternative approaches Another alternative approach is to make use of the observation that most species maintain abundant genetic variation, yet the phenotypic variation of most wildtype complex traits is low, which does not reflect the underlying genetic variation present in a population. That is, development is robust to changes in genotype and environment. In the context of complex traits such as the head and neck cancer, the phenotype can be considered as the gene expression pattern; more specifically one would expect to observe a low variation in genes that are associated with the disease among healthy individuals. In the past we have shown (Siegal and Bergman, 2002) that such robust behavior among wildtype
Within group variance by node status 10 9 8 Node positive
population is the result of the complexity of the developmental process and selection, though important, might not be the primary force behind the observed phenomenon. We further demonstrated that the capacity to harbor genetic variation depends on the complexity of the genotype-phenotype mapping, and that higher mapping complexity can evolve to harbor larger variation, thus higher robustness. With this we have demonstrated that robustness may be an inevitable consequence of complex developmental-genetic processes. In a subsequent study (Bergman and Siegal, 2003), we demonstrated, using numerical simulations of complex gene networks, as well as genome-scale expression data from yeast single-gene deletion strains, that most, if not all, genes involved in the production of complex phenotypic traits reveal phenotypic variation when functionally compromised. This is in contrast to the more widely held hypothesis that there exist specific classes of genes, such as the Hsp90, that are unique and have been selected for by natural selection to act as evolutionary capacitors, that is, when functionality of such factors is impaired, or overwhelmed, the underlying genotypic variation is revealed phenotypically. These findings led us to assume that some genes may exist solely to stabilize the function of a developmental program and, when compromised, the underling genotypic variation is revealed which may result in higher variation in both expression level of genes in the same pathway as well as the complex phenotype. The existence in model organisms of a large number of gene knockouts without apparent phenotypic effect lends plausibility to this idea. Here we have applied this hypothesis to study revealed variation when comparing the variation among individuals at two different stages of the disease. We assess for indicators of cellular instability that may be indicative of metastatic potential by comparing the relative variance in microarray profiles between tumor tissues within node class and tumor stage categories with respect to presence of lymph node metastases at diagnosis or tumor stage. Genes can be ranked by degree of relative difference in variance between comparison groups and compared by F-test statistic. Absolute differences in variation exceeding two or even three standard deviations may therefore single out potential outlying genes of concern. The robustness of a wildtype organism can be extended to the study of any complex trait which is the result of a complex genotype to phenotype mapping. Here we assume that robustness is lost when a cell is transformed from its wildtype functionality to that of cancer, furthermore, as the tumor progresses loss of robustness is increased. In a preliminary study we took expression arrays from 55 patients, each of which contains the expression level of 28,700 genes. Data were obtained from primary tumors that were characterized as lymph-node positive and lymph-node negative. We hypothesize that genes in lymph-node positive tumors will exhibit higher variation between the different patients, given their ‘cell regulatory systems’ are disturbed as the stage of the disease advances. More specifically, genes directly involved in the pathways, both those that are novel as well as those that have already been implicated in the de-
MYL2
7 6
SPRR2A Unknown KLK10 4 TNNT1 5
UPK1B
3 2 1 0 0
1
2
3
a PLEKHF1
4 5 6 Node negative TNNI2
7
8
10
9
TNNI3K PKD2L1 TNNI3 PKD2
TNNI1
KIAA1279
PAPPA KCTD13
MRVI1 RGS2 RAF1
ZMYND19 HPCL2
PRKG1
PPP1R12A
DC284352 MGC2749
NPR1 GTF2I
TNNT1 TNNT2 TPM1ACTB TPM2 PSMC4 RORA PSMC3 PSMC5 KRT15 THRB ERCC3 SP1
KIAA0980
TNNC1
DIPA FXR2
b
PPFIA1
Fig. 5. Variant gene expression as a means to create protein interaction networks in HNSCC. (a) The within group variances for expression of each RNA (blue spot) for node positive tumors are plotted on the y-axis and for node negative tumors on the x-axis. Guidelines indicating no difference in variability, ± 2-times difference and ± 3-times difference in variability between groups are drawn on the diagonal. (b) Literature-based protein-protein interaction networks using Grid database and Osprey Network Visualization Systems. The network was originated from one of the proteins identified by the variance analysis described in part a.
velopment of the disease, are hypothesized to exhibit such behavior. When comparing the variation between the two tumor types, lymph-node positive and negative, we have identified five such genes. Four of the five genes that have been detected outside the two-fold range have already been implicated as being relevant for the development of the head and neck cancer (assuming there are approximately 100 genes involved in the disease, the probability for such an event to occur at random is excitingly low, P = 6 ! 10 –10). Figure 5a shows the comparison between variations observed within 14 patients with lymph-node negative to that of 41 patients with lymph-node positive disease. Further investigation can take place following the identification of the outliers. Investigation can be experimental
Cytogenet Genome Res 118:92–109 (2007)
105
Fig. 6. Systems biology approach to the study of complex traits. Our proposed approach integrates the different aspects and elements of clinical and experimental sciences involved in the study of HNSCC while applying system sciences to improve our understanding of this complex disease.
in nature as well as more theoretical. Experimental approaches will confirm the relationship of gene expression to clinical outcome and the dysregulation of particular pathways. Using theoretical approaches our understanding of the biological process can also be enhanced. For example, one of the genes identified has already been implicated in HNSCC. Using protein-protein interaction databases, an interaction network can be created. Here we focus on the gene to generate such a network (see Fig. 5b). As can be seen this gene appears to serve as a hub, i.e., a gene that is highly connected. Thus, it may have a pleiotropic effect when disrupted causing large phenotypic changes such as inducing the head and neck cancer phenotype. The protein interactions themselves may be obtained either from databases such as Osprey, a database based on reported interaction, or from databases such as DIP which are based on predicted interaction using computational approaches. Once such a network is in place, it provides a mechanistic model for the underlying genetic system of the trait in question. Such a model can in turn be manipulated to further predict its behavior under different genetic and environmental conditions. Integration of data sets In the future, we plan to make use of multiple types of data available for patients with head and neck cancer at different stages of the disease and its different types. The available data include: RNA expression, proteomic, epigenetic
106
Cytogenet Genome Res 118:92–109 (2007)
hypermethylation of CpG island promoter elements and the transcriptional silencing of corresponding genes, together with other biological and clinical information. An example of integration among the different modalities can be envisioned by the following example. In silico digestion of all proteins discovered using our increased variation analysis as well as their targets, discovered by the protein-protein interaction prediction, can be used to generate computationally derived mass spectrometry of the relevant proteins. This computationally derived signature can suggest particular proteins for detailed analysis using high resolution mass spectrometry, then be compared with that derived from cells at different stages of the disease and from different individuals. If a similar pattern of variation is observed in both modalities for the set of identified genes, our predictive power will dramatically increase. In addition, the increase in mRNA variation may be the result of transcriptional silencing in a subset of our patient population, thus, data obtained from epigenetic hypermethylation of CpG island promoter elements can be used not only to provide validation for our original observation, but may also provide a biological mechanism to explain the observed variation. Theoretical and systems biological approaches can be used to generate biological signatures and provide diagnostic tools, but more importantly, these approaches can also serve as strategies to generate hypotheses and plausible mechanisms for combinatorial and modifier effects that determine clinical outcome (Fig. 6). Acknowledgements The authors would like to acknowledge Ms. Annie D’Alauro for her editorial assistance.
References
Adrien LR, Schlecht NF, Kawachi N, Smith RV, Brandwein-Gensler M, et al: Classification of DNA methylation patterns in tumor cell genomes using a CpG island microarray. Cytogenet Genome Res 114: 16–23 (2006). Angus SP, Mayhew CN, Solomon DA, Braden WA, Markey MP, et al: RB reversibly inhibits DNA replication via two temporally distinct mechanisms. Mol Cell Biol 24: 5404–5420 (2004). Anneroth G, Batsakis J, Luna M: Review of the literature and a recommended system of malignancy grading in oral squamous cell carcinomas. Scand J Dent Res 95: 229–249 (1987). Bariol C, Suter C, Cheong K, Ku SL, Meagher A, et al: The relationship between hypomethylation and CpG island methylation in colorectal neoplasia. Am J Pathol 162: 1361–1371 (2003). Barnes L, Verbin R, Guggenheimer J: Cancer of the oral cavity and oropharynx, in Barnes L (ed): Surgical Pathology of the Head and Neck, 2nd ed, pp 370–438 (Marcel Dekker, New York 2001). Beitler JJ, Smith RV, Owen RP, Silver CE, Mazumdar M, Wadler S: Phase II clinical trial of parenteral hydroxyurea and hyperfractionated, accelerated external beam radiation therapy in patients with advanced squamous cell carcinoma of the head and neck: Toxicity and efficacy with continuous ribonucleoside reductase inhibition. Head Neck 29: 18–25 (2007).
Belbin TJ, Singh B, Barber I, Socci N, Wenig B, et al: Molecular classification of head and neck squamous cell carcinoma using cDNA microarrays. Cancer Res 62:1184–1190 (2002). Belbin TJ, Gaspar J, Haigentz M, Perez-Soler R, Keller SM, et al: Indirect measurements of differential gene expression with cDNA microarrays. Biotechniques 36: 310–314 (2004). Belbin TJ, Singh B, Smith RV, Socci ND, Wreesmann VB, et al: Molecular profiling of tumor progression in head and neck cancer. Arch Otolaryngol Head Neck Surg 131: 10–18 (2005). Benjamini Y, Yekutieli D: Quantitative trait loci analysis using the false discovery rate. Genetics 171:783–790 (2005). Bergman A, Siegal ML: Evolutionary capacitance as a general feature of complex gene networks. Nature 424:549–552 (2003). Bernier J, Domenge C, Ozsahin M, Matuszewska K, Lefebvre JL, et al: Postoperative irradiation with or without concomitant chemotherapy for locally advanced head and neck cancer. N Engl J Med 350:1945–1952 (2004). Bonner JA, Harari PM, Giralt J, Azarnia N, Shin DM, et al: Radiotherapy plus cetuximab for squamous-cell carcinoma of the head and neck. N Engl J Med 354:567–578 (2006). Brandwein-Gensler M, Teixeira MS, Lewis CM, Lee B, Rolnitzky L, et al: Oral squamous cell carcinoma: histologic risk assessment, but not margin status, is strongly predictive of local disease-free and overall survival. Am J Surg Pathol 29: 167–178 (2005). Breiman L, Friedman J, Olshen R, Stone C: Classification and Regression Trees (Wadsworth International Group, Belmont 1984). Brizel DM, Wasserman TH, Henke M, Strnad V, Rudat V, et al: Phase III randomized trial of amifostine as a radioprotector in head and neck cancer. J Clin Oncol 18: 3339–3345 (2000). Bryne M, Koppang HS, Lilleng R, Stene T, Bang G, Dabelsteen E: New malignancy grading is a better prognostic indicator than Broders’ grading in oral squamous cell carcinomas. J Oral Pathol Med 18:432–437 (1989). Bryne M, Koppang HS, Lilleng R, Kjaerheim A: Malignancy grading of the deep invasive margins of oral squamous cell carcinomas has high prognostic value. J Pathol 166: 375–381 (1992). Bryne M, Jenssen N, Boysen M: Histological grading in the deep invasive front of T1 and T2 glottic squamous cell carcinomas has high prognostic value. Virchows Arch 427: 277–281 (1995). Bundgaard T, Rossen K, Henriksen SD, Charabi S, Sogaard H, Grau C: Histopathologic parameters in the evaluation of T1 squamous cell carcinomas of the oral cavity. Head Neck 24: 656– 660 (2002). Calais G, Alfonsi M, Bardet E, Sire C, Germain T, et al: Randomized trial of radiation therapy versus concomitant chemotherapy and radiation therapy for advanced-stage oropharynx carcinoma. J Natl Cancer Inst 91: 2081–2086 (1999). Califano J, van der Riet P, Westra W, Nawroz H, Clayman G, et al: Genetic progression model for head and neck cancer: implications for field cancerization. Cancer Res 56: 2488–2492 (1996). Calin GA, Croce CM: MicroRNAs and chromosomal abnormalities in cancer cells. Oncogene 25:6202–6210 (2006a). Calin GA, Croce CM: MicroRNA signatures in human cancers. Nat Rev Cancer 6: 857–866 (2006b).
Cao QJ, Belbin T, Socci N, Balan R, Prystowsky MB, et al: Distinctive gene expression profiles by cDNA microarrays in endometrioid and serous carcinomas of the endometrium. Int J Gynecol Pathol 23: 321–329 (2004). Carvalho AL, Nishimoto IN, Califano JA, Kowalski LP: Trends in incidence and prognosis for head and neck cancer in the United States: a site-specific analysis of the SEER database. Int J Cancer 114:806–816 (2005). Cheung VG, Morley M, Aguilar F, Massimi A, Kucherlapati R, Childs G: Making and reading microarrays. Nat Genet 21: 15–19 (1999). Ciampi A, Negassa A, Lou Z: Tree-structured prediction for censored survival data and the Cox model. J Clin Epidemiol 48: 675–689 (1995). Condeelis J, Pollard JW: Macrophages: obligate partners for tumor cell migration, invasion, and metastasis. Cell 124:263–266 (2006). Cooper JS, Pajak TF, Forastiere AA, Jacobs J, Campbell BH, et al: Postoperative concurrent radiotherapy and chemotherapy for high-risk squamous-cell carcinoma of the head and neck. N Engl J Med 350:1937–1944 (2004). Costello JF, Fruhwald MC, Smiraglia DJ, Rush LJ, Robertson GP, et al: Aberrant CpG-island methylation has non-random and tumourtype-specific patterns. Nat Genet 24: 132–138 (2000). Couzin J: Breakthrough of the year. Small RNAs make big splash. Science 298: 2296–2297 (2002). Cummins JM, Velculescu VE: Implications of micro-RNA profiling for cancer diagnosis. Oncogene 25: 6220–6227 (2006). De Schutter H, Landuyt W, Verbeken E, Goethals L, Hermans R, Nuyts S: The prognostic value of the hypoxia markers CA IX and GLUT 1 and the cytokines VEGF and IL 6 in head and neck squamous cell carcinoma treated by radiotherapy +/– chemotherapy. BMC Cancer 5: 42 (2005). Dombrowski SM, Maglott D: Using the map viewer to explore genomes, in: The NCBI handbook [Internet], Chapter 19 (National Library of Medicine (US), National Center for Biotechnology Information, Bethesda 2003). Dudoit S, Fridlyand J: A prediction-based resampling method for estimating the number of clusters in a dataset. Genome Biol 3:RESEARCH0036 (2002). Eisen MB, Spellman PT, Brown PO, Botstein D: Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA 95: 14863–14868 (1998). Epstein JB, Robertson M, Emerton S, Phillips N, Stevenson-Moore P: Quality of life and oral function in patients treated with radiation therapy for head and neck cancer. Head Neck 23: 389–398 (2001). Eriksen JG, Steiniche T, Overgaard J: The role of epidermal growth factor receptor and E-cadherin for the outcome of reduction in the overall treatment time of radiotherapy of supraglottic larynx squamous cell carcinoma. Acta Oncol 44:50–58 (2005). Ferlay J, Bray F, Pissani P, Parkin DM: GLOBOCAN 2000: Cancer Incidence, Mortality and Prevalence Worldwide, Version 1.0. IARC CancerBase No. 5 (IARC Press, Lyon 2001). Forastiere A, Koch W, Trotti A, Sidransky D: Head and neck cancer. N Engl J Med 345: 1890–1900 (2001). Forastiere AA, Goepfert H, Maor M, Pajak TF, Weber R, et al: Concurrent chemotherapy and radiotherapy for organ preservation in advanced laryngeal cancer. N Engl J Med 349: 2091–2098 (2003).
Gal TJ, Yueh B, Futran ND: Influence of prior hyperbaric oxygen therapy in complications following microvascular reconstruction for advanced osteoradionecrosis. Arch Otolaryngol Head Neck Surg 129: 72–76 (2003). George E, McCulloch R: Variable selection via Gibbs sampling. JASA 88: 881–889 (1993). Gillison ML, Shah KV: Human papillomavirus-associated head and neck squamous cell carcinoma: mounting evidence for an etiologic role for human papillomavirus in a subset of head and neck cancers. Curr Opin Oncol 13: 183–188 (2001). Gillison ML, Koch WM, Capone RB, Spafford M, Westra WH, et al: Evidence for a causal association between human papillomavirus and a subset of head and neck cancers. J Natl Cancer Inst 92:709–720 (2000). Ginos MA, Page GP, Michalowicz BS, Patel KJ, Volker SE, et al: Identification of a gene expression signature associated with recurrent disease in squamous cell carcinoma of the head and neck. Cancer Res 64: 55–63 (2004). Hanna E, Shrieve DC, Ratanatharathorn V, Xia X, Breau R, et al: A novel alternative approach for prediction of radiation response of squamous cell carcinoma of head and neck. Cancer Res 61: 2376–2380 (2001). Herrero R, Castellsague X, Pawlita M, Lissowska J, Kee F, et al: Human papillomavirus and oral cancer: the International Agency for Research on Cancer multicenter study. J Natl Cancer Inst 95: 1772–1783 (2003). Hochberg Y, Benjamini Y: More powerful procedures for multiple significance testing. Stat Med 9:811–818 (1990). Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ: Cancer statistics, 2007. CA Cancer J Clin 57: 43–66 (2007). Joost HG, Thorens B: The extended GLUT-family of sugar/polyol transport facilitators: nomenclature, sequence characteristics, and potential function of its novel members (review). Mol Membr Biol 18: 247–256 (2001). Kapil U, Singh P, Bahadur S, Dwivedi SN, Singh R, Shukla N: Assessment of risk factors in laryngeal cancer in India: a case-control study. Asian Pac J Cancer Prev 6: 202–207 (2005). Khuri FR, Jain SR: Novel agents and incremental advances in the treatment of head and neck cancer. Semin Oncol 31:3–10 (2004). Kim HC, Kusukawa J, Kameyama T: Clinicopathologic parameters in predicting cervical nodal metastasis in early squamous cell carcinoma of the oral cavity. Kurume Med J 40: 183–192 (1993). Kobayashi H, Sagara J, Kurita H, Morifuji M, Ohishi M, et al: Clinical significance of cellular distribution of moesin in patients with oral squamous cell carcinoma. Clin Cancer Res 10: 572–580 (2004). Koch WM, Lango M, Sewell D, Zahurak M, Sidransky D: Head and neck cancer in nonsmokers: a distinct clinical and molecular entity. Laryngoscope 109:1544–1551 (1999). Koh DM, Padhani AR: Diffusion-weighted MRI: a new functional clinical technique for tumour imaging. Br J Radiol 79: 633–635 (2006). Kudo Y, Kitajima S, Ogawa I, Hiraoka M, Sargolzaei S, et al: Invasion and metastasis of oral cancer cells require methylation of E-cadherin and/or degradation of membranous betacatenin. Clin Cancer Res 10:5455–5463 (2004). Kunkel M, Reichert TE, Benz P, Lehr HA, Jeong JH, et al: Overexpression of Glut-1 and increased glucose metabolism in tumors are associated with a poor prognosis in patients with oral squamous cell carcinoma. Cancer 97: 1015– 1024 (2003).
Cytogenet Genome Res 118:92–109 (2007)
107
Kyzas PA, Loizou KT, Ioannidis JP: Selective reporting biases in cancer prognostic factor studies. J Natl Cancer Inst 97: 1043–1055 (2005). Lagos-Quintana M, Rauhut R, Lendeckel W, Tuschl T: Identification of novel genes coding for small expressed RNAs. Science 294:853–858 (2001). Lau NC, Lim LP, Weinstein EG, Bartel DP: An abundant class of tiny RNAs with probable regulatory roles in Caenorhabditis elegans. Science 294:858–862 (2001). Lee RC, Ambros V: An extensive class of small RNAs in Caenorhabditis elegans. Science 294: 862–864 (2001). Lee RC, Feinbaum RL, Ambros V: The C. elegans heterochronic gene lin-4 encodes small RNAs with antisense complementarity to lin-14. Cell 75:843–854 (1993). Lee SH, Kwong AD, Pan ZQ, Hurwitz J: Studies on the activator 1 protein complex, an accessory factor for proliferating cell nuclear antigen-dependent DNA polymerase delta. J Biol Chem 266:594–602 (1991). Leethanakul C, Patel V, Gillespie J, Pallente M, Ensley JF, et al: Distinct pattern of expression of differentiation and growth-related genes in squamous cell carcinomas of the head and neck revealed by the use of laser capture microdissection and cDNA arrays. Oncogene 19: 3220– 3224 (2000). Leitha T, Glaser C, Lang S: Is early sestamibi imaging in head and neck cancer affected by MDR status, p53 expression, or cell proliferation? Nucl Med Biol 25: 539–541 (1998). Lindberg R: Distribution of cervical lymph node metastases from squamous cell carcinoma of the upper respiratory and digestive tracts. Cancer 29: 1446–1449 (1972). Lindel K, Beer KT, Laissue J, Greiner RH, Aebersold DM: Human papillomavirus positive squamous cell carcinoma of the oropharynx: a radiosensitive subgroup of head and neck carcinoma. Cancer 92: 805–813 (2001). List MA, Bilir SP: Evaluations of quality of life and organ function. Semin Oncol 31: 827–835 (2004). Liu ZJ, Maekawa M: Polymerase chain reactionbased methods of DNA methylation analysis. Anal Biochem 317: 259–265 (2003). Lothaire P, de Azambuja E, Dequanter D, Lalami Y, Sotiriou C, et al: Molecular markers of head and neck squamous cell carcinoma: promising signs in need of prospective evaluation. Head Neck 28:256–269 (2006). Lu J, Getz G, Miska EA, varez-Saavedra E, Lamb J, et al: MicroRNA expression profiles classify human cancers. Nature 435: 834–838 (2005). Madan R, Brandwein-Gensler M, Schlecht NF, Elias K, Gorbovitsky E, et al: Differential tissue and subcellular expression of ERM proteins in normal and malignant tissues: cytoplasmic ezrin expression has prognostic signficance for head and neck squamous cell carcinoma. Head Neck 28:1018–1027 (2006). Mangeat P, Roy C, Martin M: ERM proteins in cell adhesion and membrane dynamics. Trends Cell Biol 9: 187–192 (1999). Martinez I, Wang J, Hobson KF, Ferris RL, Khan SA: Identification of differentially expressed genes in HPV-positive and HPV-negative oropharyngeal squamous cell carcinomas. Eur J Cancer 43: 415–432 (2006). Martinez-Gimeno C, Rodriguez EM, Vila CN, Varela CL: Squamous cell carcinoma of the oral cavity: a clinicopathologic scoring system for evaluating risk of cervical lymph node metastasis. Laryngoscope 105:728–733 (1995).
108
Maruya S, Issa JP, Weber RS, Rosenthal DI, Haviland JC, et al: Differential methylation status of tumor-associated genes in head and neck squamous carcinoma: incidence and potential implications. Clin Cancer Res 10: 3825–3830 (2004). McShane LM, Altman DG, Sauerbrei W: Identification of clinically useful cancer prognostic factors: what are we missing? J Natl Cancer Inst 97: 1023–1025 (2005a). McShane LM, Altman DG, Sauerbrei W, Taube SE, Gion M, Clark GM: Reporting recommendations for tumor marker prognostic studies (REMARK). J Natl Cancer Inst 97: 1180–1184 (2005b). Mellanen P, Minn H, Grenman R, Harkonen P: Expression of glucose transporters in head-andneck tumors. Int J Cancer 56: 622–629 (1994). Mineta H, Miura K, Takebayashi S, Misawa K, Araki K, et al: Prognostic value of glucose transporter 1 expression in patients with hypopharyngeal carcinoma. Anticancer Res 22: 3489– 3494 (2002). Mubashar M, Harrington KJ, Chaudhary KS, Lalani el-N, Stamp GW, Peters AM: Differential effects of toremifene on doxorubicin, vinblastine and Tc-99m-sestamibi in P-glycoproteinexpressing breast and head and neck cancer cell lines. Acta Oncol 43: 443–452 (2004). Negassa A, Ciampi A, Abrahamowicz M, Shapiro S, Boivin J: Validation of tree structured prediction for censored survival data: Its application to prognstic classification. J Statist Comput Simul 67:289–317 (2000). Neiderhuber J: First-Generation Guidelines for NCI-Supported Biorepositories. Federal Register 71:25184–25203 (2006). Ng IO, Lam KY, Ng M, Kwong DL, Sham JS: Expression of P-glycoprotein, a multidrug-resistance gene product, is induced by radiotherapy in patients with oral squamous cell carcinoma. Cancer 83:851–857 (1998). Novoradovskaya N, Whitfield ML, Basehore LS, Novoradovsky A, Pesich R, et al: Universal Reference RNA as a standard for microarray experiments. BMC Genomics 5: 20 (2004). Ogi K, Toyota M, Ohe-Toyota M, Tanaka N, Noguchi M, et al: Aberrant methylation of multiple genes and clinicopathological features in oral squamous cell carcinoma. Clin Cancer Res 8: 3164–3171 (2002). Oliver RJ, Woodwards RT, Sloan P, Thakker NS, Stratford IJ, Airley RE: Prognostic value of facilitative glucose transporter Glut-1 in oral squamous cell carcinomas treated by surgical resection; results of EORTC Translational Research Fund studies. Eur J Cancer 40: 503–507 (2004). Pallante P, Visone R, Ferracin M, Ferraro A, Berlingieri MT, et al: MicroRNA deregulation in human thyroid papillary carcinomas. Endocr Relat Cancer 13: 497–508 (2006). Perou CM, Sorlie T, Eisen MB, van de Rijn M, Jeffrey SS, et al: Molecular portraits of human breast tumours. Nature 406:747–752 (2000). Pignon JP, Burdett S: Comment on ‘Survival improvement in resectable non-small cell lung cancer with (neo)adjuvant chemotherapy: results of a meta-analysis of the literature’ by T. Berghmans, M. Paesmans, A.P. Meert, C. Mascaux, P. Lothaire, J.J. Lafitte, et al. (Lung Cancer 49 (2005) 13–23.) Lung Cancer 51: 261–262 (2006).
Cytogenet Genome Res 118:92–109 (2007)
Pignon JP, Bourhis J, Domenge C, Designe L: Chemotherapy added to locoregional treatment for head and neck squamous-cell carcinoma: three meta-analyses of updated individual data. MACH-NC Collaborative Group. Meta-Analysis of Chemotherapy on Head and Neck Cancer. Lancet 355:949–955 (2000). Po Wing YA, Lam KY, Lam LK, Ho CM, Wong A, et al: Prognostic factors of clinically stage I and II oral tongue carcinoma-A comparative study of stage, thickness, shape, growth pattern, invasive front malignancy grading, Martinez-Gimeno score, and pathologic features. Head Neck 24:513–520 (2002). Potti A, Dressman HK, Bild A, Riedel RF, Chan G, et al: Genomic signatures to guide the use of chemotherapeutics. Nat Med 12: 1294–1300 (2006). Puskas LG, Zvara A, Hackler L, Jr., Micsik T, van HP: Production of bulk amounts of universal RNA for DNA microarrays. Biotechniques 33: 898–900, 902, 904 (2002). Rabkin D, Chhieng DC, Miller MB, Jennings T, Feustel P, et al: P-glycoprotein expression in the squamous cell carcinoma of the tongue base. Laryngoscope 105:1294–1299 (1995). Ragin CC, Taioli E, Weissfeld JL, White JS, Rossie KM, et al: 11q13 amplification status and human papillomavirus in relation to p16 expression defines two distinct etiologies of head and neck tumours. Br J Cancer 95: 1432–1438 (2006). Reiner A, Yekutieli D, Benjamini Y: Identifying differentially expressed genes using false discovery rate controlling procedures. Bioinformatics 19:368–375 (2003). Reisser C, Eichhorn K, Herold-Mende C, Born AI, Bannasch P: Expression of facilitative glucose transport proteins during development of squamous cell carcinomas of the head and neck. Int J Cancer 80: 194–198 (1999). Rimm DL, Camp RL, Charette LA, Costa J, Olsen DA, Reiss M: Tissue microarray: a new technology for amplification of tissue resources. Cancer J 7:24–31 (2001). Ritchie JM, Smith EM, Summersgill KF, Hoffman HT, Wang D, et al: Human papillomavirus infection as a prognostic factor in carcinomas of the oral cavity and oropharynx. Int J Cancer 104:336–344 (2003). Robertson KD: DNA methylation, methyltransferases, and cancer. Oncogene 20: 3139–3155 (2001). Roesch-Ely M, Nees M, Karsai S, Ruess A, Bogumil R, et al: Proteomic analysis reveals successive aberrations in protein expression from healthy mucosa to invasive head and neck cancer. Oncogene 26: 54–64 (2006). Romond EH, Perez EA, Bryant J, Suman VJ, Geyer CE, Jr, et al: Trastuzumab plus adjuvant chemotherapy for operable HER2-positive breast cancer. N Engl J Med 353: 1673–1684 (2005). Rosas SL, Koch W, da Costa Carvalho MG, Wu L, Califano J, et al: Promoter hypermethylation patterns of p16, O6-methylguanine-DNAmethyltransferase, and death-associated protein kinase in tumors and saliva of head and neck cancer patients. Cancer Res 61: 939–942 (2001). Saito Y, Liang G, Egger G, Friedman JM, Chuang JC, et al: Specific activation of microRNA-127 with downregulation of the proto-oncogene BCL6 by chromatin-modifying drugs in human cancer cells. Cancer Cell 9: 435–443 (2006). Schlecht NF: Prognostic value of human papillomavirus in the survival of head and neck cancer patients: an overview of the evidence. Oncol Rep 14:1239–1247 (2005).
Schlecht NF, Franco EL, Pintos J, Negassa A, Kowalski LP, et al: Interaction between tobacco and alcohol consumption and the risk of cancers of the upper aero-digestive tract in Brazil. Am J Epidemiol 150: 1129–1137 (1999). Schwartz GJ, Mehta RH, Wenig BL, Shaligram C, Portugal LG: Salvage treatment for recurrent squamous cell carcinoma of the oral cavity. Head Neck 22:34–41 (2000). Schwartz SM, Daling JR, Doody DR, Wipf GC, Carter JJ, et al: Oral cancer risk in relation to sexual history and evidence of human papillomavirus infection. J Natl Cancer Inst 90: 1626– 1636 (1998). Shrout P, Fleiss J: Intraclass correlation: Uses in assessing rater reliability. Psychol Bull 86: 420– 428 (1979). Siegal ML, Bergman A: Waddington’s canalization revisited: developmental stability and evolution. Proc Natl Acad Sci USA 99: 10528–10532 (2002). Simon R, Radmacher MD, Dobbin K: Design of studies using DNA microarrays. Genet Epidemiol 23: 21–36 (2002). Slamon DJ, Clark GM, Wong SG, Levin WJ, Ullrich A, McGuire WL: Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 235: 177– 182 (1987). Slamon DJ, Leyland-Jones B, Shak S, Fuchs H, Paton V, et al: Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N Engl J Med 344:783–792 (2001). Slebos RJ, Yi Y, Ely K, Carter J, Evjen A, et al: Gene expression differences associated with human papillomavirus status in head and neck squamous cell carcinoma. Clin Cancer Res 12: 701– 709 (2006). Smith EM, Ritchie JM, Summersgill KF, Hoffman HT, Wang DH, et al: Human papillomavirus in oral exfoliated cells and risk of head and neck cancer. J Natl Cancer Inst 96: 449–455 (2004). Smith EM, Ritchie JM, Pawlita M, Rubenstein LM, Haugen TH, et al: Human papillomavirus seropositivity and risks of head and neck cancer. Int J Cancer 120:825–832 (2006). Smith RV, Goldman SY, Beitler JJ, Wadler SS: Decreased short- and long-term swallowing problems with altered radiotherapy dosing used in an organ-sparing protocol for advanced pharyngeal carcinoma. Arch Otolaryngol Head Neck Surg 130:831–836 (2004). Smith RV, Schlecht NF, Childs G, Prystowsky MB, Belbin TJ: Pilot study of mucosal genetic differences in early smokers and nonsmokers. Laryngoscope 116:1375–1379 (2006). Spiro RH, Guillamondegui O, Jr, Paulino AF, Huvos AG: Pattern of invasion and margin assessment in patients with oral tongue cancer. Head Neck 21:408–413 (1999).
Storey JD: A direct approach to false discovery rates. J R Stat Soc B 64: 479–498 (2002). Strati K, Pitot HC, Lambert PF: Identification of biomarkers that distinguish human papillomavirus (HPV)-positive versus HPV-negative head and neck cancers in a mouse model. Proc Natl Acad Sci USA 103: 14152–14157 (2006). Takes RP, Baatenburg de Jong RJ, Schuuring E, Hermans J, Vis AA, et al: Markers for assessment of nodal metastasis in laryngeal carcinoma. Arch Otolaryngol Head Neck Surg 123: 412–419 (1997). Tibshirani R, Hastie T, Narasimhan B, Chu G: Diagnosis of multiple cancer types by shrunken centroids of gene expression. Proc Natl Acad Sci USA 99:6567–6572 (2002). Tseng GC, Oh MK, Rohlin L, Liao JC, Wong WH: Issues in cDNA microarray analysis: quality filtering, channel normalization, models of variations and assessment of gene effects. Nucleic Acids Res 29: 2549–2557 (2001). Tusher VG, Tibshirani R, Chu G: Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci USA 98: 5116–5121 (2001). Vandecaveye V, de Keyzer F, Vander Poorten V, Deraedt K, Alaerts H, et al: Evaluation of the larynx for tumour recurrence by diffusion-weighted MRI after radiotherapy: initial experience in four cases. Br J Radiol 79: 681–687 (2006). van de Vijver M, He YD, van ‘t Veer LJ, Dai H, Hart AA, et al: A gene-expression signature as a predictor of survival in breast cancer. N Engl J Med 347:1999–2009 (2002). van Rijnsoever M, Elsaleh H, Joseph D, McCaul K, Iacopetta B: CpG island methylator phenotype is an independent predictor of survival benefit from 5-fluorouracil in stage III colorectal cancer. Clin Cancer Res 9: 2898–2903 (2003). van ‘t Veer L, Dai H, van de Vijver MJ, He YD, Hart AA, et al: Gene expression profiling predicts clinical outcome of breast cancer. Nature 415: 530–536 (2002). Villaret DB, Wang T, Dillon D, Xu J, Sivam D, et al: Identification of genes overexpressed in head and neck squamous cell carcinoma using a combination of complementary DNA subtraction and microarray analysis. Laryngoscope 110:374–381 (2000). Volinia S, Calin GA, Liu CG, Ambs S, Cimmino A, et al: A microRNA expression signature of human solid tumors defines cancer gene targets. Proc Natl Acad Sci USA 103: 2257–2261 (2006). Wang J, Takashima S, Takayama F, Kawakami S, Saito A, et al: Head and neck lesions: characterization with diffusion-weighted echo-planar MR imaging. Radiology 220: 621–630 (2001). Warner GD, Reis PP, Jurisica I, Sultan M, Arora S, et al: Molecular classification of oral cancer by cDNA microarrays identifies overexpressed genes correlated with nodal metastasis. Int J Cancer 110:857–868 (2004).
Wei SH, Chen CM, Strathdee G, Harnsomburana J, Shyu CR, et al: Methylation microarray analysis of late-stage ovarian carcinomas distinguishes progression-free survival in patients and identifies candidate epigenetic markers. Clin Cancer Res 8:2246–2252 (2002). Xiao GH, Beeser A, Chernoff J, Testa JR: p21-activated kinase links Rac/Cdc42 signaling to merlin. J Biol Chem 277: 883–886 (2002). Yan PS, Perry MR, Laux DE, Asare AL, Caldwell CW, Huang TH: CpG island arrays: an application toward deciphering epigenetic signatures of breast cancer. Clin Cancer Res 6: 1432–1438 (2000). Yanaihara N, Caplen N, Bowman E, Seike M, Kumamoto K, et al: Unique microRNA molecular profiles in lung cancer diagnosis and prognosis. Cancer Cell 9: 189–198 (2006). Yang YH, Speed T: Design issues for cDNA microarray experiments. Nat Rev Genet 3: 579–588 (2002). Yang YH, Dudoit S, Luu P, Lin DM, Peng V, et al: Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation. Nucleic Acids Res 30:e15 (2002). Yarbrough WG, Slebos RJ, Liebler D: Proteomics: clinical applications for head and neck squamous cell carcinoma. Head Neck 28: 549–558 (2006). Yonemura S, Hirao M, Doi Y, Takahashi N, Kondo T, et al: Ezrin/radixin/moesin (ERM) proteins bind to a positively charged amino acid cluster in the juxta-membrane cytoplasmic domain of CD44, CD43, and ICAM-2. J Cell Biol 140:885– 895 (1998). Yonemura S, Tsukita S, Tsukita S: Direct involvement of ezrin/radixin/moesin (ERM)-binding membrane proteins in the organization of microvilli in collaboration with activated ERM proteins. J Cell Biol 145: 1497–1509 (1999). Yoo GH, Piechocki MP, Ensley JF, Nguyen T, Oliver J, et al: Docetaxel induced gene expression patterns in head and neck squamous cell carcinoma using cDNA microarray and PowerBlot. Clin Cancer Res 8: 3910–3921 (2002). Yu T, Robb VA, Singh V, Gutmann DH, Newsham IF: The 4.1/ezrin/radixin/moesin domain of the DAL-1/Protein 4.1B tumour suppressor interacts with 14–3-3 proteins. Biochem J 365: 783– 789 (2002). Zhang B, Pan X, Cobb GP, Anderson TA: microRNAs as oncogenes and tumor suppressors. Dev Biol 302:1–12 (2006). Zuo C, Ai L, Ratliff P, Suen JY, Hanna E, et al: O6methylguanine-DNA methyltransferase gene: epigenetic silencing and prognostic value in head and neck squamous cell carcinoma. Cancer Epidemiol Biomarkers Prev 13: 967–975 (2004).
Cytogenet Genome Res 118:92–109 (2007)
109
(MNA), loss of 1p, and gain of 17q represent a distinct subgroup, while loss of 11q, 3p and gain of 17q characterize another subgroup (Vandesompele et al., 1998; Guo et al., 1999; Breen et al., 2000; Plantaz et al., 2001). Each of these abnormalities have highly unfavorable prognoses, implying that they play important roles in the biology of the tumors (Brodeur et al., 1984; Luttikhuis et al., 2001; Lastowska et al., 2002; Spitz et al., 2003; Attiyeh et al., 2005; Vandesompele et al., 2005). In addition, there are many other recurrent, as well as random, imbalances found in neuroblastoma. Large-scale chromosomal imbalances alter the transcriptome of cancer cells
DNA copy number changes substantially alter the gene expression profiles of cancer cells, as determined by gene expression microarray analyses (Pollack et al., 2002; Wolf et al., 2004). Certainly, one might expect that the expression of the majority of genes mapping within a region of imbalance will be altered by the change in gene dosage. However, in addition to simple dosage effects, some chromosomal imbalances lead to the altered expression of genes mapping to other chromosomes, perhaps by altering the dosage of transcriptional activators and repressors. The global dysregulation of the transcriptome was evident following experimentally induced aneuploidy in cell culture models (Upender et al., 2004). Loss of 11q in neuroblastoma is also an example of an imbalance that can be associated with significant up or down regulation of genes located throughout the genome (McArdle et al., 2004). Enormous effort has gone into the search for tumor suppressor genes that are affected by large-scale genomic loss since the seminal discovery of the ‘two-hit’ model of allelic inactivation of a tumor suppressor gene in retinoblastoma (Knudson, 1971). However, the fact that some TSGs can convey a tumorigenic effect by haploinsufficiency has made the search for tumor suppressor genes more complicated in neuroblastoma and in many other forms of cancer. More recently, there is also an emerging view, particularly for neuroblastoma, that multiple genes in the region of imbalance may be contributing to tumor pathogenesis. Wang et al. (2006) have shown that expression of many candidate tumor suppressor genes in the region of 1p and 11q loss are altered in neuroblastoma, leading them to suggest that multiple genes are important in the pathogenesis of these tumors. It is also likely that chromosomal imbalances also are causing the dysregulation of nontranslated RNA sequences such as microRNAs that have recently been demonstrated to have major roles in tumorigenesis (reviewed by Esquela-Kerscher and Slack, 2006). Many microRNA loci, and other nontranslated RNAs that are involved with cancer remain to be discovered. Finally, it is very important to mechanistically connect imbalances on different chromosomes by identifying genes that operate together in common genetic pathways. An excellent example of connecting the genetic effects of multiple imbalances was described for the MYCN amplified subtype of neuroblastoma that exhibit 1p loss and low level gain of
17q (Valentijn et al., 2005). MYCN transcriptionally represses a gene, CDC4 that maps to chromosome 1p, which is involved with the induction of differentiation of neuroblastoma cell lines. MYCN also transcriptionally activates the NME1 (alias nm23-H1) gene mapping to the 17q region which undergoes low level gain. Most significantly, the nm23-H1 protein interacts with CDC42 protein to block differentiation. Thus, there are multiple ways of transcriptionally and post-translationally down-regulating CDC42 activity in neuroblastoma, along with multiple mechanisms for increasing the activity of NME1 (Valentijn et al., 2005). Many more complex pathways affected by multiple genomic abnormalities remain to be identified, such as the genetic pathway(s) affected by loss of 3p, 11q and gain of 17q in the 11q– neuroblastoma subtype. High resolution dissection of the neuroblastoma genome using oligonucleotide array CGH (oaCGH)
The development of comparative genomic hybridization (CGH) analysis in the early 1990’s by Kallioniemi et al. (1992) allowed for the genome-wide detection of chromosomal imbalances in virtually every form of cancer, which have been summarized at the site: http://www.progenetix. de/pgscripts/progenetix/Aboutprogenetix.html. Initially, CGH analysis involved competitive hybridization of tumor and normal DNA labeled with different fluorochromes to normal metaphase chromosomes on a microscope slide and had a five to ten megabase resolution. The metaphase CGH studies carried out on neuroblastoma are numerous and will not be individually cited here. The application of CGH to target sequences such as BACs or cDNAs on microarrays has led to dramatic increases in resolution. For example, the use of complete tiling-path BAC arrays containing 32,000 clones has allowed for an 80 to 100 kb resolution (Ishkanian et al., 2004). Only a very limited number of CGH analyses, however, using BAC or cDNA arrays have been carried out on neuroblastoma (Chen et al., 2004; Mosse et al., 2005). More recently, a number of commercially available CGH platforms utilizing long (60 mer) oligonucleotides as arrayed target molecules have been developed which provide substantially higher resolution than the lower density BAC or cDNA arrays (Barrett et al., 2004; Selzer et al., 2005). We have used a very recently developed high resolution oligonucleotide array comparative genomic hybridization (oaCGH) protocol (Selzer et al., 2005) to map the chromosomal breakpoints leading to imbalances in neuroblastoma tumors and cell lines (Stallings et al., 2006). The whole genome oaCGH platform from NimbleGen Systems contains 390,000 probes, allowing for a median probe spacing of 6 kb throughout the genome. The purpose of these analyses was to determine the highest resolution achievable by CGH, to provide high resolution mapping of breakpoints leading to large-scale chromosomal imbalances, and to identify cryptic abnormalities that were too subtle to be detected by other CGH array analyses of neuroblastoma tumors.
Cytogenet Genome Res 118:110–115 (2007)
111
Fig. 1. Map of breakpoints on chromosome 11q that generated loss of 11q material. Each breakpoint is the most distal position of the breakpoint interval in megabase (Mb) pairs on the DNA sequence map. The highest degree of clustering of breakpoint positions is indicated in the rectangular box, between coordinates 68.7 and 71.8 Mb.
Mapping breakpoint positions of large-scale chromosomal imbalances High resolution breakpoint mapping is important because it allows us to define a minimal region of deletion that can be used to further pinpoint potential tumor suppressor genes. CGH analysis of 56 tumors and cell lines with the whole genome oligonucleotide arrays led to the identification of 467 chromosomal breakpoints that generated imbalances greater than 2 Mb in length (Stallings et al., 2006). Although a matched constitutional DNA sample from the patient was not used as the reference DNA sample for the majority of experiments, imbalances in excess of 2 Mb should be somatically acquired abnormalities in the tumors, otherwise one might expect patients to have phenotypic abnormalities. Breakpoints were identified using both window averaging and non-window averaging methods, as described in Selzer et al. (2005). Larger window lengths provide greater accuracy in breakpoint assignment, but at the sacrifice of resolution. We have found that 25 kb window averaging allows for assignment of breakpoints to 50 kb intervals (25 kb 8 12.5 kb), providing an accurate assignment in a high proportion of cases (190%). For approximately 10% of breakpoint assignments, larger 50 kb window averaging is required. The quality of the results is directly proportional to the quality of the DNA. As loss of chromosome 11q is associated with a poor clinical outcome, and defines a major genetic subtype of neuro-
112
Cytogenet Genome Res 118:110–115 (2007)
blastoma, we will review our findings on the mapping of 11q breakpoints which were identified in 33 tumors and cell lines. As illustrated in Fig. 1, the chromosome 11q breakpoints were non-randomly distributed. Approximately 60% of these breakpoints clustered to a 3 Mb region between positions 68.8 to 71.8 Mb on the DNA sequence map. Within this region, five breakpoints mapped to a very narrow 225 kb region (positions 71.037 to 71.262) that spanned a region that was duplicated on many chromosomes. This region on chromosome 11q is clearly a hotspot for double strand DNA breaks. Other hotspots for breakage on chromosome 11q occurred around positions 73.4 Mb, 82 to 83 Mb and 98 to 99 Mb. Overall, the breakpoints on chromosome 11q occurred at a significantly higher than expected frequency in regions that were segmentally duplicated, suggesting that some segmentally duplicated regions behave like fragile sites (Stallings et al., 2006). There was, however, no statistically significant association between breakpoints on other chromosomes and segmental duplications. It will be interesting to determine if any other recurrent chromosome abnormalities in other types of cancer occur preferentially in segmentally duplicated regions. Mapping of chromosomal breakpoint intervals using focused tiling arrays The chromosomal mechanism generating loss of 11q and gain of 17q in neuroblastoma is usually an unbalanced t(11;
Fig. 2. Detection of a t(11;17) in neuroblastoma nuclei by FISH analysis. The red labeled probe is from a region on chromosome 17q mapping distal of the 17q breakpoint while the green labeled probe comes from a position mapping proximal of the 11q breakpoint. Arrows point to the translocation breakpoint, where red and green signals come in close proximity. Three copies of chromosome 17q are evident. Two signals for the 11q are present because the probe maps proximal of the region that was deleted.
Table 1. Breakpoints on chromosomes 11q and 17q in cases of unbalanced t(11;17)
Tumor/Cell line
11q Breakpointsa
17q Breakpointsa
SK-N-AS 10 17 31 37 40 41 45 48
71.262/71.312 77.662/77.687 82.187/82.212 70.287/70.337 71.262/71.312 73.437/73.462 80.112/80.137 69.612/69.637 69.837/69.862
38.262/38.287 30.087/30.112 30.287/30.312 27.712/27.737 27.637/27.662 40.887/40.912 31.487/31.512 37.837/37.862 34.237/34.262
a
The figures on either side of the slash represent the proximal/distal boundaries of the breakpoint positions on the DNA sequence map (Mb), 8 12.5 kb.
17), as observed by many different labs (Avet-Loiseau et al., 1995; Stark, 2003; Stallings et al., 2004). The chromosomes segregate so that a der(11) chromosome, along with a normal chromosome 11 and two normal chromosomes 17 are retained in the tumor cells. Due to the importance of this abnormality in neuroblastoma, we were interested in determining if there was any DNA sequence homology between the 11q and 17q breakpoint regions that might be mechanistically involved. Nine 11q– tumors or cell lines were determined to possess an unbalanced t(11;17) using FISH analysis (Stallings et al., 2004), as illustrated in Fig. 2 and Table 1.
As shown in Table 1, breakpoint positions on the 11q and on the 17q ranged over megabase sized regions. The breakpoint interval on the 11q was identical in only two tumors, and the breakpoint intervals for the 17q were unique in all tumors (although some mapped very close together). There was no sequence homology between the 11q and 17q regions for each translocation, other than high abundance repeats such as Alu, L1 etc. Thus, homlogous recombination between low abundance repeats or segmental duplications is apparently not involved with the formation of the t(11;17). Although low abundance copy number repeats play major roles in the generation of constitutional chromosome abnormalities (Shaffer and Lupski, 2000), very few examples for the role of low copy number repeats (LCR) in the generation of chromosome abnormalities acquired in cancer can be cited. LCRs may be implicated in the formation of an unusual Philadelphia chromosome found in CML (Saglio et al., 2002) and in the isochromosome 17q which is common to many cancers (Barbouti et al., 2004). Mapping breakpoints using higher density tiling arrays Could recombination between high abundance repeats play a role in generating the t(11;17), as has been reported for Alu repeats (Kolomietz et al., 2002; Elliot et al., 2005)? This question is not possible to answer with 50 kb breakpoint intervals because an interval of this size will have many of the same repeats on both chromosomes. CGH analysis of higher density tiling arrays focused on the 11q and 17q breakpoint regions (median probe spacing of 50 bp) led to more refined mapping of the breakpoints for three independent t(11; 17) (Selzer et al., 2005). Breakpoints for 11q could be refined to a 10 kb interval, while 17q breakpoints could be refined to 1 kb intervals. These intervals could not be further mapped because of the presence of high abundance repeats. The 11q and 17q breakpoint intervals did not contain the same repetitive elements, however, indicating that illegitimate homologous recombination is not involved with the formation of the t(11;17). Exactly why a recurrent t(11;17) takes place in neuroblast cells is unknown, but it seems likely that nonhomologous end joining of double strand breaks is involved. There is increasing evidence that non-random spatial proximity of certain translocation prone loci in the nucleus may be one cause of why translocations occur repeatedly (Roix et al., 2003), so it is tempting to speculate that chromosomes 11 and 17 are positioned in close proximity in the neuroblast nucleus – a possibility which requires further study. Identification of subtle abnormalities and CNPs using oaCGH analysis In addition to providing high resolution maps of the breakpoints generating large-scale chromosomal imbalances, oaCGH analysis has enormous potential for detecting subtle chromosomal imbalances that are less than 2 Mb in length. It has been our experience that imbalances that are as small as 100 kb can be routinely detected by oaCGH using the NimbleGen platform (Stallings et al., 2006). However, the high resolution achievable by oaCGH is also a ma-
Cytogenet Genome Res 118:110–115 (2007)
113
Fig. 3. Representative oaCGH plot of chromosome 9p illustrating microdeletion of the PTPRD locus. The vertical axis represents the fluorescence ratio between the tumor and reference DNA on a log2 scale, while the horizontal axis is genome position in Mb. Plots for both the primary tumor (upper) and metastatic cells from bone marrow (lower) are displayed. The deletion of PTPRD is only detected in cells from the metastatic site, consistent with the notion that PTPRD is involved with metastasis suppression.
jor weakness because large numbers of constitutional copy number variants, as described by Sebat et al. (2004) are detected. This is particularly true if a matched constitutional DNA sample from the patient is unavailable. Distinguishing CNPs from somatically acquired deletions and duplications then becomes a major challenge. Although databases cataloguing CNPs are not yet complete, we have nevertheless found that the majority of CNP variants can be identified using a bioinformatics approach (Stallings et al., 2006). Our approach involves determining if an imbalance is in a segmentally duplicated region of the genome using the UCSC Genome Browser, or is listed as a known polymorphic region in the Database of Genomic Variants (Toronto). If either is the case, we consider the imbalance detected in the tumor to be a constitutional variant, as opposed to a somatically acquired DNA copy number alteration. After identifying hundreds of probable CNP variants following the analysis of only 56 samples, we were left with 100 subtle imbalances that were likely somatically acquired in the neuroblastomas. In our studies, the most frequently occurring (six out of 56 cases) subtle abnormalities involved microdeletions of the protein tyrosine phosphatase receptor D locus (PTPRD), as illustrated in Fig. 3. In some instances, the deletions were homozygous, indicating a two-hit mechanism of allele inactivation. Interestingly, large-scale loss of chromosome 9p is a recurrent imbalance in neuroblastoma (Vandesompele et al., 2005), and the detection of these microdeletions suggest that PTPRD is at least one of the genes targeted by the larger-scale deletions. Analysis of constitutional DNA samples from some patients confirmed that these deletions were somatically acquired. This locus appears to play a rather general role in cancer, having been identified as being one of the most frequent homozygous deletions in lung cancer cell lines (Sato et al., 2005). PTPRD maybe more of a metastasis suppressor gene,
114
Cytogenet Genome Res 118:110–115 (2007)
rather than a tumor suppressor gene, given that PTPRD protein interacts with MTSS1 (alias MIM missing in metastasis) (Woodings et al., 2003), a likely metastasis suppressor gene. This is consistent with loss of 9p being more of a secondary abnormality in neuroblastoma. Summary
There is an emerging body of evidence that large-scale chromosomal imbalances significantly contribute to tumor pathogenesis – this is particularly true for neuroblastoma where several large-scale imbalances are associated with poor clinical outcome. The importance of these genomic gains and losses in neuroblastoma have led us to map large numbers of breakpoints using CGH analysis of both whole genome (median probe spacing 6 kb) and focused tiling (median probe spacing of 50 bp in breakpoint regions) oligonucleotide arrays. These arrays permit the mapping of breakpoints, such as the unbalanced t(11;17) generating 11q loss and 17q gain, to intervals ranging in length from 1 to 10 kb, dependent upon local genomic architectural features. Although breakpoints on chromosome 11q are preferentially associated with regions of segmental duplication, indicating that such regions can act as fragile sites, the breakpoints on other chromosomes were not preferentially associated with segmental duplications. No evidence for homologous recombination in the generation of the recurrent t(11;17) was obtained, and nonhomologous end joining of DNA double strand breaks appears to be a more likely mechanism, based on the wide distribution of 11q and 17q breakpoints. Finally, the identification of recurrent microdeletions at the same genetic locus (PTPRD) demonstrates the utility of oaCGH analysis for pinpointing genes that may have been targeted by large-scale chromosomal imbalances.
References Attiyeh EF, London WB, Mosse YP, Wang Q, Winter C, et al: Chromosome 1p and 11q deletions and outcome in neuroblastoma. N Engl J Med 353:2243–2253 (2005). Avet-Loiseau H, Venuat AM, Benard J, Leibovitch MP, Hartmann O, Bernheim A: Morphologic and molecular cytogenetics in neuroblastoma. Cancer 75:1694–1699 (1995). Barbouti A, Stankiewicz P, Nusbaum C, Cuomo C, Cook A, et al: The breakpoint region of the most common isochromosome, i(17q), in human neoplasia is characterized by a complex genomic architecture with large, palindromic, lowcopy repeats. Am J Hum Genet 74: 1–10 (2004). Barrett MT, Scheffer A, Ben-Dor A, Sampas N, Lipson D, et al: Comparative genomic hybridization using oligonucleotide microarrays and total genomic DNA. Proc Natl Acad Sci USA 101: 17765–17770 (2004). Breen CJ, O’Meara A, McDermott M, Mullarkey M, Stallings RL: Co-ordinate deletion of chromosome 3p and 11q in neuroblastoma detected by comparative genomic hybridization. Cancer Genet Cytogenet 120:44–49 (2000). Brodeur GM: Neuroblastoma: biological insights into a clinical enigma. Nat Rev Cancer 3: 203– 216 (2003). Brodeur GM, Seeger RC, Schwab M, Varmus HE, Bishop JM: Amplification of N-myc in untreated human neuroblastomas correlates with advanced disease stage. Science 224: 1121–1124 (1984). Chen QR, Bilke S, Wei JS, Whiteford CC, Cenacchi N: cDNA array-CGH profiling identifies genomic alterations specific to stage and MYCNamplification in neuroblastoma. BMC Genomics 5:1–13 (2004). Elliott B, Richardson C, Jasin M: Chromosomal translocation mechanisms at intronic Alu elements in mammalian cells. Mol Cell 17: 885– 894 (2005). Esquela-Kerscher A, Slack FJ: Oncomirs – microRNAs with a role in cancer. Nature Rev Cancer 6:259–269 (2006). Guo C, White PS, Weiss MJ, Hogaerty MD, Thompson PM, et al: Allelic deletion at 11q23 is common in MYCN single copy neuroblastomas. Oncogene 18:4948–4957 (1999). Ishkanian AS, Malloff CA, Watson SK, DeLeeuw RJ, Chi B, et al: A tiling resolution DNA microarray with complete coverage of the human genome. Nat Genet 36: 299–303 (2004). Kallioniemi A, Kallioniemi OP, Sudar D, Rutovitz D, Gray JW, et al: Comparative genomic hybridization: a rapid new method for detecting and mapping DNA amplification in tumors. Science 258:818–821 (1992). Knudson AG: Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci USA 68:820–823 (1971).
Kolomietz E, Meyn MS, Pandita A, Squire JA: The role of Alu repeat clusters as mediators of recurrent chromosomal aberrations in tumors. Genes Chromosomes Cancer 35: 97–112 (2002). Lastowska M, Cotterill S, Bown N, Cullinane C, Variend S, et al: Breakpoint position on 17q identifies the most aggressive neuroblastoma tumors. Genes Chromosomes Cancer 34: 428– 436 (2002). Luttikhuis M, Powell JE, Rees SA, Genus T, Chughtai S, et al: Neuroblastomas with chromosome 11q loss and single copy MYCN comprise a biologically distinct group of tumors with adverse prognosis. Brit J Cancer 85: 531–537 (2001). McArdle L, McDermott M, Purcell R, Grehan D, O’Meara A, et al: DNA microarray analysis of gene expression in neuroblastoma displaying loss of 11q. Carcinogenesis 25: 1599–1609 (2004). Mosse YP, Greshock J, Margolin A, Naylor T, Cole K, et al: High resolution detection and mapping of genomic DNA alterations in neuroblastoma. Genes Chromosomes Cancer 43: 390–403 (2005). Plantaz D, Vandesompele J, Van Roy N, Lastowska M, Bown N, et al: Comparative genomic hybridization (CGH) analysis of stage 4 neuroblastoma reveals high frequency of 11q deletion in tumors lacking MYCN amplification. Int J Cancer 91:680–686 (2001). Pollack JR, Sorlie T, Perou CM, Rees CA, Jeffrey SS, et al: Microarray analysis reveals a major direct role of copy number alteration in the transcriptional program of human breast tumors. Proc Natl Acad Sci USA 99: 12963–12968 (2002). Roix JJ, McQueen PG, Munson PJ, Parada LA, Misteli T: Spatial proximity of translocationprone gene loci in human lymphomas. Nat Genet 34:287–291 (2003). Saglio G, Storlazzi CT, Giugliano E, Surace C, Anelli L, et al: A 76-kb duplicon maps close to the BCR gene on chromosome 22 and the ABL gene on chromosome 9: possible involvement in the genesis of the Philadelphia chromosome translocation. Proc Natl Acad Sci USA 99:9882–9887 (2002). Sato M, Takahashi K, Nagayama K, Arai Y, Ito N, et al: Identification of chromosome 9p as the most frequent target of homozygous deletions in lung cancer. Genes Chromosomes Cancer 44: 405–414 (2005). Sebat J, Lakshmi B, Troge J, Alexander J, Young J, et al: Large-scale copy number polymorphism in the human genome. Science 305: 525–528 (2004). Selzer RR, Richmond TA, Pofahl NJ, Green RD, Eis PS, et al: Analysis of chromosome breakpoints in neuroblastoma at sub-kilobase resolution using fine tiling oligonucleotide array CGH. Genes Chromosomes Cancer 44: 305–319 (2005).
Shaffer LG, Lupski JR: Molecular mechanisms for constitutional chromosomal rearrangements in humans. Annu Rev Genet 34: 297–329 (2000). Spitz R, Hero B, Ernestus K, Berthold F: Deletions in chromosome arms 3p and 11q are new prognostic markers in localized and 4s neuroblastoma. Clin Can Res 9:52–58 (2003). Stallings RL, Carty P, McArdle L, Mullarkey M, McDermott M, et al: Molecular cytogenetic analysis of recurrent unbalanced t(11; 17) in neuroblastoma. Cancer Genet Cytogenet 154: 44–51 (2004). Stallings RL, Nair P, Maris JM, Catchpoole D, McDermot M, et al: High resolution analysis of chromosome breakpoints and genomic instability identifies PTPRD as a candidate tumor suppressor gene in neuroblastoma. Cancer Res 66:3673–3680 (2006). Stark B: der(11)t(11;17): a distinct cytogenetic pathway of advanced stage neuroblastoma (NBL) – detected by spectral karyotyping (SKY). Cancer Lett 197:75–79 (2003). Upender MB, Habermann JK, McShane LM, Korn EL, Barrett JC, et al: Chromosome transfer induced aneuploidy results in complex dysregulation of the cellular transcriptome in immortalized and cancer cells. Cancer Res 64:6941–6949 (2004). Valentijn LJ, Koppen A, van Asperen R, Root HA, Haneveld F, Versteeg R: Inhibition of a new differentiation pathway in neuroblastoma by copy number defects of N-myc, Cdc42, and nm23 genes. Cancer Res 65: 3136–345 (2005). Vandesompele J, Van Roy N, Van Gele M, Laureys G, Ambros P, et al: Genetic heterogeneity of neuroblastoma studied by comparative genomic hybridization. Genes Chromosomes Cancer 23:141–152 (1998). Vandesompele J, Baudis M, De Preter K, Van Roy N, Ambros P, et al: Unequivocal delineation of clinicogenetic subgroups and development of a new model for improved outcome prediction in neuroblastoma. J Clin Oncol 23: 2280–2299 (2005). Wang Q, Diskin S, Rappaport E, Attiyeh E, Mosse Y, et al: Integrative genomics identifies distinct molecular classes of neuroblastoma and shows that multiple genes are targeted by regional alterations in DNA copy number. Cancer Res 66: 6050–6062 (2006). Wolf M, Mousses S, Hautaniemi S, Karhu R, Huusko P, et al: High-resolution analysis of gene copy number alterations in human prostate cancer using CGH on cDNA microarrays: impact of copy number on gene expression. Neoplasia 6: 240–247 (2004). Woodings JA, Sharp SJ, Machesky LM: MIM-B, a putative metastasis suppressor protein, binds to actin and to protein tyrosine phosphatase . Biochem J 371: 463–471 (2003).
Cytogenet Genome Res 118:110–115 (2007)
115
Gross chromosomal aberrations
The development of cancer is generally believed to require the accumulation of multiple genetic aberrations. These aberrations range from single nucleotide mutations to cytogenetically detectable numerical and structural chromosomal alterations (Devilee and Cornelisse, 1994; Bieche and Lidereau, 1995). While human tissue specimens and immortal breast cancer cell lines represent an excellent resource for genomic studies, chromosomal alterations have not been well characterized until relatively recently, largely because of limitations inherent to conventional chromosome banding techniques. The development of new techniques which enable the accurate, high throughput analysis of specimens has enabled substantial progress in this field. Comparative genomic hybridization (CGH), introduced in 1992, has been the technique of choice for mapping DNA copy number changes in human tumors. The technique uses hybridization to compare abundance of specific genome sequences in tumor cell DNA relative to normal reference genomes (Kallioniemi et al., 1994). This opened new avenues in genomic investigation because it obviated the need to culture cells before their chromosomes could be analyzed. The latest generation of CGH analysis, array-CGH or matrix-CGH, uses ordered arrays of genomic DNA sequences and further increases the potential of CGH to provide insight into chromosomal aberrations present in cancer by enabling single-gene resolution. Recently developed multi-color chromosome imaging techniques such as spectral karyotyping (SKY) (Schrock et al., 1996) and multiplexFISH (M-FISH) (Speicher et al., 1996) utilize simultaneous visualization of each human chromosome with specific fluorochrome combinations and enable a far more rapid and detailed karyotypic analysis of solid tumors and cell lines (Macville et al., 1997; Kawai et al., 2002). The chromosomal aberrations in the most commonly used human breast carcinoma cell lines have recently been investigated using these techniques (Kytola et al., 2000; Xie et al., 2002; Watson et al., 2004; Goodison et al., 2005a; Shadeo and Lam, 2006). Karyotypic analysis has shown breast tumor cell lines to be either near diploid with simple rearrangements or highly aneuploid with multiple, complex rearrangements (Morris et al., 1997; Kytola et al., 2000). The most frequent gains detected by CGH in breast tumor cell lines are 1q, 8q, 20q, 7, 11q13, 17q, 9q, and 16p, whereas losses were most common at 8p, 11q14]qter, 18q, and Xq (Kytola et al., 2000). The comparison of CGH data from cell lines to CGH studies from primary breast tumors (Nishizaki et al., 1997; Tirkkonen et al., 1998) reveals that the most common gains and losses are the same. In breast tumor specimens, the most common chromosomal imbalances detected by CGH are gains of 1q, 8q, 16p, 17q, and 20q and losses involving 8p, 13q, 18q, and 16q (Isola et al., 1995; Ried et al., 1995). Although literally hundreds of articles have been published describing the pattern of copy number alterations in cancer, very few of the genes affected have been identified. Consequently, investigators are currently attempting to correlate CGH data with gene expression. Hy-
man et al. (2002) combined CGH with cDNA arrays to analyze breast tumor cell lines and found that a significant correlation does exist between gene amplification and gene expression across the genome. The same group has focused on the 17q21]q23 amplicon, which includes ERBB2 and is a common region of amplification in breast cancers with poor prognosis (Kallioniemi et al., 1994; Kauraniemi et al., 2001). These analyses are examples of the power achieved by combining genomic and expression approaches, and have identified several genes that are consistently overexpressed in breast cancer cell lines and advanced breast cancer. SKY analysis of breast tumor cell line karyotypes reveals that the chromosomes most frequently involved in translocations are 8, 1, 17, 16, and 20. These chromosomes contain gene amplicons, for example, at 8q24, 17q11]q12, and 20q12]q13, that are present in up to 30% of breast carcinomas (Devilee and Cornelisse, 1994; Bieche and Lidereau, 1995; Tanner et al., 1996; Barlund et al., 1997). Furthermore, studies using conventional cytogenetics report that primary and metastatic breast carcinomas contain aberrations of chromosome 8 in up to 40% of cases (Adeyinka et al., 2000; Popescu and Zimonjic, 2002). Databases that make data from SKY/M-FISH and CGH studies in cancer available have been created by the NCI (Knutsen et al., 2005). The SKY/M-FISH and CGH Database (http://www.ncbi.nlm. nih.gov/projects/sky/) enables investigators to submit and analyze clinical and research cytogenetic data. The Cancer Chromosomes database integrates the SKY/M-FISH & CGH Database with the Mitelman Database of Chromosome Aberrations in Cancer (http://cgap.nci.nih.gov/Chromosomes/ Mitelman) and the Recurrent Chromosome Aberrations in Cancer database (http://cgap.nci.nih.gov/Chromosomes/ RecurrentAberrations). Relatively few studies have explored whether specific cytogenetic abnormalities can be used to stratify breast tumors with clinical course, namely metastatic relapse. Blegen et al. (2003) performed CGH on tumors from patients who had early relapse and from patients who remained free from distant metastases for more than ten years. Tumors in relapse patients showed a higher average number of chromosomal copy alterations compared to the long-term survivors, including gains of chromosome 3q, 9p, 11p and 11q and loss of 17p (Blegen et al., 2003). This aligns well with a CGH study that evaluated 76 LN-negative breast carcinomas (median follow-up 46 months) and found that a gain of 3q is a stronger predictor of recurrence than grade, tumor size, and estrogen receptor status (Janssen et al., 2003). In a CGH study of 39 invasive breast carcinomas with a mean follow-up period of 99 months, Aubele et al. (2002) identified an independent prognostic value for chromosomal gains on 11q13, 12q24, 17 and 18p. More recently, an arrayCGH study of ER-positive breast cancer tissues reported that the most significant chromosomal alterations found more often in the group with metastatic recurrence within five years were loss of 11p15.5]p15.4, 1p36.33, 11q13.1, and 11p11.2 (Han et al., 2006). Array-CGH was also used by Yao et al. (2006) to show that the overall frequency of copy num-
Cytogenet Genome Res 118:116–129 (2007)
117
ber alterations in regions 1q, 8q24, 11q13, 12p13, 17q21]q23, 16p13 and 20q13, correlated with the advancing nature of the tumors. Amplification of candidate loci was confirmed by quantitative PCR, and expression level analysis of genes present in these loci identified many putative target genes (Yao et al., 2006). These CGH results have significant overlap and indicate that malignant breast adenocarcinomas are characterized by specific chromosomal copy number changes. The high divergence of karyotype data makes it difficult to define more specific aberrations involved in breast cancer etiology or progression, however, the study of copy number aberrations has proved to lead to the development of useful diagnostic markers. The tight correlation of amplification of ERBB2 with increased expression of the gene product provides the basis for the tests that measure the DNA copy number of ERBB2 for predicting response to Herceptin (Pauletti et al., 1996). Furthermore, combinations of FISH probes for regions of recurrent copy number aberration in other tumor types have proved useful for monitoring disease status (Sokolova et al., 2000) and for distinguishing benign from malignant skin lesions (Bastian et al., 1999). Tissue-based signatures of breast cancer progression
Large-scale microarray analyses of human tissue specimens are building consensus gene expression profiles of various tumors, including breast (Sorlie et al., 2001, 2003, 2006; van ‘t Veer et al., 2002; van de Vijver et al., 2002; Ma et al., 2003, 2004; Weigelt et al., 2003; Hu et al., 2006; Sotiriou et al., 2006). Breast cancer is clearly a heterogeneous disease and microarray analysis has been successfully used to derive distinct patterns of gene expression that correlate with molecular subtypes of breast cancers. Expression profiles that are associated with estrogen-receptor status, HER2 (ERBB2) expression, BRCA1 or BRCA2 mutations and with ‘basal type’ or ‘luminal types’ have been identified using unsupervised analyses (Sorlie et al., 2001; Hedenfalk et al., 2003; Mackay et al., 2003; Kristensen et al., 2005). These breast cancer subtypes also represent clinically distinct subgroups of patients, with differences in disease progression and overall survival. For example, ER-positive tumors tend to have the best outcome, whereas HER2+ tumors have a bad prognosis. The genes which show high variance across different tumors have been termed ‘intrinsic genes’ and a set of 1300 such genes has been reported to accurately subdivide the molecular subtypes across multiple array platforms and independent microarray studies (Hu et al., 2006). Studies using reduced lists of genes are beginning to be used in independent validation studies, and as few as 40 genes have been used to stratify subtypes of breast cancer using quantitative PCR assays (Perreard et al., 2006). In a study conducted in 2002, a supervised classification analysis of DNA-microarray data predicted prognosis better than clinical prognostic indicators such as grade, stage, and nodal status. The investigators queried samples from a co-
118
Cytogenet Genome Res 118:116–129 (2007)
hort of young (!53) breast cancer patients with LN-negative tumors. Reduction of the data identified a 70-gene signature that distinguished patient groups with good or poor prognosis with respect to the likelihood of the later development of clinical metastases (van ‘t Veer et al., 2002). Subsequent testing of the 70-gene prognosis gene set in an independent cohort of 295 patients confirmed the accuracy of the signature, regardless of LN status, in predicting the occurrence of distant metastases within five years of treatment (van de Vijver et al., 2002). The resulting gene-expression profile was a far more powerful predictor of the outcome of disease in young patients with breast cancer than the currently used St. Gallen or NIH consensus criteria based on clinical and histological characteristics. Importantly, the predictive power of the 70-gene signature was based upon metastasis to non-lymphatic tissues. More recently, a similar approach identified a 76-gene signature that was also successfully used to predict distant metastasis in patients with LN-negative primary breast cancer (Wang et al., 2005). The signature showed 93% sensitivity and 48% specificity when tested in an independent set of 171 LN-negative patients. As with the 70-gene signature described above, this signature outperformed the St Gallen criteria. Another study that included 159 samples obtained from both treated and untreated patients, with and without LN involvement, derived a 64gene signature set that identified genes associated with the occurrence of distant metastasis or death within five years (Pawitan et al., 2005). Once again, the signature was validated in an independent set of 289 patients and was found to outperform clinical criteria in the stratification of risk and overall survival. A number of studies have investigated the hypothesis that molecular programs of wound healing might be reactivated in cancer metastasis (Bissell and Radisky, 2001). Chang et al. identified consistent features in the transcriptional response of normal fibroblasts to serum, and used this so called ‘wound-response signature’ to reveal links between this phenomena and cancer progression in a variety of common epithelial tumors, including breast (Chang et al., 2004). The same group subsequently tested the accuracy of the wound-response signature in the same 295 patients with early-stage breast cancer used to identify and validate the 70-gene prognostic signature (van ‘t Veer et al., 2002). It was found that this signature could identify approximately 90% of patients who developed metastases, independently of clinical or pathological risk factors (Chang et al., 2005). The suggestion that the risk of metastasis for breast cancer patients can be predicted by the gene expression profile of its heterogeneous primary tumor has come as something of a surprise. The prevailing idea has been that metastatic potential is acquired relatively late in the multistep process of tumor progression (Fidler and Kripke, 1977), however, the recent microarray findings have reopened the debate on this topic by suggesting that the ability to metastasize to distant sites may be an early and inherent genetic property of breast tumors. An early study reported the derivation of a gene expression signature that distinguished pri-
mary from metastatic adenocarcinomas. Ramaswamy et al. (2003) compared expression profiles of a range of 64 primary adenocarcinomas (breast, prostate, lung, colon, uterus and ovary) to 12 unmatched metastases resulting from adenocarcinomas from the same spectrum of sites, but resected from a variety of end-organs. This comparison identified an expression pattern of 17 genes that best distinguished primary and metastatic adenocarcinomas. Notably, components of the protein translation apparatus were heavily represented in the 17-gene pan-metastasis signature. By re-applying this metastasis-associated gene expression pattern to data on 279 primary solid tumors of diverse types, including 78 stage I primary breast carcinomas, it was found that primary tumors carrying the 17gene expression signature were more likely to be associated with metastasis and poor clinical outcome (Ramaswamy et al., 2003). Conversely, in more recent studies utilizing sets of matched tissue specimens, Weigelt et al. (2003, 2005a) found that gene expression profiles in human primary breast carcinoma are preserved in the associated distant metastasis, with respect to both subtype profile and poor prognosis signature, even if metastases develop after a long interval. In breast cancer, it is the axillary LNs that are most often the first sites in which disseminated tumor cells can be detected (Stacker et al., 2002) and their presence or absence is currently one of the most important factors for disease course prediction for breast cancer patients (McGuire, 1987). While prognosis signatures for distant metastasis occurrence have been achieved, the same research groups have not been able to identify a classifier predicting the LN status of primary breast tumors. The comparison of gene expression profiles of 15 primary breast carcinomas and their matching LN metastases revealed no common subset of differentially expressed genes, and in the analysis of the Dutch 295 primary breast tumor profile dataset, no classifier predicting LN metastasis could be developed (Weigelt et al., 2005b). However, a study by Huang et al. (2003) did identify patterns of expression that were associated with the LN status of 89 Taipei breast cancer patients. These analyses were based on biopsy material and resulted in a predictive accuracy of approximately 90% for LN metastasis and relapse. None of the genes implicated in the overall recurrence of disease in this study were common to the 70-gene prognosis signature of van ‘t Veer et al. (2002), and cross-validation with a previous US study did not identify significant numbers of overlapping genes, presumably due to the different racial genetic backgrounds, but the authors concluded that LN metastasis and disease recurrence are distinct biologically. Collectively, these data suggest that proximal LN metastasis occurs independently of distant metastasis (Weigelt et al., 2005b), but it remains unclear whether metastasis to more distant sites proceeds sequentially from LN metastasis or in parallel via a hematogenous route (Chambers et al., 2000).
Current limitations and future prospects
Improved prognostic markers are clearly needed in order to better stratify patients with respect to their risk for developing metastases, and important advances towards this have been made. However, many limitations with microarray-based tumor classifications exist. The subtype and prognosis gene sets remain relatively broad and no quantitative information is available. Subsequent tests based on the qualitative presence or absence of a single gene, or even multiple genes are unlikely to be accurate enough for clinical assay development. Due to economic considerations, the number of samples tested to date remains very small relative to the feature dimensionality present on microarray chips. Consequently, as more data are added the current signatures can change dramatically, and another round of adjustment to tissue-based signatures will occur now that arrays that provide complete genome coverage have become available. Furthermore, studies that assay grossly homogenized tissues define an average gene expression signature that does not account for variations in tumor complexity, heterogeneity or non-tumor cell contributions. Gene expression signatures may reflect genetic and/or epigenetic phenomena, effects of growing in different tissue environments, or simply the composition of cell types. These problems have caused some debate regarding the timing of the next phase of current gene signature clinical testing (Brenton et al., 2005; Loi et al., 2005), but it is clear that independent validation of these signatures is the next step. Indeed, multiinstitutional studies are being organized (TransBIG in Europe (Tuma, 2004) and NCI PACCT in the US) to facilitate the evaluation of breast cancer genomic signatures in far greater numbers of clinical specimens. At this stage, microarray data in itself does not provide definitive answers. The inclusion of a gene in a prognostic list that is determined by a supervised classification method does not indicate the importance of that gene in cancer biology, nor does it provide functional insight into the underlying mechanisms of disease. However, these analyses have provided a wealth of candidate genes and pathways for further study. In order to investigate the role of these genes in the mechanisms of metastasis animal models are needed. Animal models of metastasis
Signatures of breast cancer metastasis The multistep nature of metastasis poses difficulties in both design and interpretation of experiments to unveil the mechanisms causing the process. Studies on excised human tissues are complicated by the variance of genetic background between individuals and by the cellular heterogeneity of a complex tissue mass. Critical to the experimental analysis of metastasis has been the isolation of human tumor cell lines and the ability to study their behavior in vivo by inoculation into immune-compromised mice. Several established human breast cancer cell lines with varying documented abilities of invasiveness and/or migration in
Cytogenet Genome Res 118:116–129 (2007)
119
vitro are available, and some are capable of spontaneous metastasis in vivo, i.e. dissemination from a primary tumor and proliferation in a distal site (Price et al., 1990). This xenograft model represents an experimental system in which the role of specific genes can be screened and tested. However, many breast cell lines, especially those isolated from pleural effusions, are polyclonal in nature and composed of cell populations that are heterogeneous in metastatic phenotype, thus in order to use cell lines as models in studies seeking to define genes causing metastasis it is optimal to isolate single cell progenies from the parental cell line source. Through in vivo selection of monoclonal cultures of the MDA-MB-435 breast tumor cell line we were able to characterize a pair of single cell progenies (M4A4 and NM2C5) which differ markedly in their ability to complete the metastatic process (Urquidi et al., 2002). When orthotopically inoculated into athymic mice, both cell lines form primary tumors, but only M4A4 is capable of metastasis to the lungs and lymph nodes (Urquidi et al., 2002; Goodison et al., 2003). These isogenic cell lines of opposite metastatic propensity constitute a stable and accessible model for the identification of genes involved in the process of tumor metastasis. We have performed multiple comparative analyses of these paired cell lines, including cytogenetic analyses and evaluation of the expression of a number of gene products previously implicated in cellular transformation and metastasis (Urquidi et al., 2002; Agarwal et al., 2003; Goodison et al., 2003, 2005a). To further elucidate the extent of the molecular changes associated with acquisition of the metastatic phenotype in this model, we recently employed a genomewide expression profiling approach. Intensity modeling and hierarchical clustering analysis revealed a subset of 85 genes (12-fold change) whose expression was statistically correlated with metastatic phenotype (Goodison et al., 2005b). Some genes in this group have been implicated in invasion, tumor cell proliferation and/or metastasis previously, but GTPase signaling components were one of the most-wellrepresented functional groups. Restoration of the expression of deleted in liver cancer-1 (DLC-1), a Rho-GTPase-activating protein, in metastatic M4A4 cells resulted in the inhibition of migration and invasion in vitro and a significant reduction in the ability of these cells to form pulmonary metastases in athymic mice. DLC-1 has specific GTPase activating protein functions for RhoA and Cdc42 (Wong et al., 2003), members of the Rho family that are consistently overexpressed in breast tumors (Fritz et al., 1999). The finding that DLC1 can act as a ‘metastasis-suppressor gene’ supports an influential role for GTPase signaling in tumor progression. Metastasis suppressor genes are potential candidates for marker development because, by definition, their loss should be associated with the acquisition of metastatic potential (Shevde and Welch, 2003). A similar methodological approach was used by Kang et al. (2003) to study breast tumor cell homing to specific organ sites. The investigators were able to derive monoclonal lines from the MDA-MB-231 breast tumor cell line that had differing degrees of ability to form tumor deposits in murine bone. The experimental system was necessarily differ-
120
Cytogenet Genome Res 118:116–129 (2007)
ent to that used in a lung metastasis model, in that cells need to be inoculated directly into the heart in order to get sufficient cell numbers to the arterial side of the murine body and to the skeleton. Using comparative microarray analyses of the parental MDA-MB-231 cell line and a variant selected for bone colonization, the investigators identified a bone metastasis gene expression signature of 102 genes. Interestingly, in line with our lung metastasis model, DLC-1 was found to be downregulated in breast cell populations which were highly metastatic to bone (Kang et al., 2003) suggesting that some signaling pathways may be pivotal to metastatic efficiency regardless of the target organ. Transfection of combinations of genes confirmed some of these genes (IL11, OPN/SPP1, CTGF and CXCR4) as being functionally involved in the efficiency of MDA-MB-231 cell bone colonization. Both IL11 and CTGF are known to be activated by TGF, suggesting a potential prometastatic role for this cytokine in bone. The same group continued to derive monoclonal lines from MDA-MB-231 and test their propensity to colonize other organs. Comparison of transcriptional profiles of cell populations highly or weakly metastatic to the murine lung resulted in a 54-gene lung metastasis signature (Minn et al., 2005). This gene set was distinct from the bone metastasis signature overall (Kang et al., 2003), but some genes were common to both MDA-MB-231 models, including MMP1 and CXCR4 (Minn et al., 2005). The MDA-MB231 lung metastasis signature gene set was also compared to expression profiles obtained from a cohort of 82 therapeutically excised primary breast carcinomas from patients with known metastatic status. This revealed that expression profiles in the primary tumors from patients with subsequent lung metastases, but not bone metastases, correlated to some extent with the lung metastasis signature (Minn et al., 2005). Another variation on clonal models of metastasis has been characterized by Lev et al. (2003). This model consists of a more aggressive metastatic clone (GILM2) derived from the weakly metastatic GI101A human breast cancer cell line which metastasizes to the lung and lymph nodes when inoculated orthotopically (Hurst et al., 1993). Microarraybased comparison with the parental line identified a list of 106 genes that were differentially expressed (12.5 fold) in the highly metastatic GILM2 variant (Kluger et al., 2005). Immunohistochemical confirmation in human breast tissue specimens of the expression of three markers, heat shock protein 70 (HSP70), chemokine ligand 1 (CXCL1), and secretory leukocyte protease inhibitor (SLPI), revealed that the expression of all three genes was correlated with lymph node involvement, and the expression of HSP70 and CXCL1 was associated with decreased overall patient survival. Interestingly, we have identified and functionally proven the role of SLPI in an invasion-independent model of metastasis (Sugino et al., 2004). A wholly murine model of spontaneous breast cancer metastasis to multiple sites has been characterized and used to identify genes involved in metastatic progression. Several syngeneic tumor lines with a spectrum of metastatic phenotypes were isolated from a spontaneous mammary tumor in a BALB/cfC3H mouse (Lelekakis et al.,
1999). When inoculated orthotopically, the resulting primary tumors are either nonmetastatic or produce spontaneous metastases to lymph node, lung and/or bone (Eckhardt et al., 2005). To identify metastasis-related genes, the investigators grouped the expression profiles of a weakly metastatic group and compared these with the profiles of a highly metastatic group using mouse cDNA arrays. A metastasis signature of 216 genes was derived, of which 125 were known genes. A significant proportion of genes belonging to extracellular matrix protein families had elevated expression in the highly metastatic tumors. The role of one of these genes, POEM (nephronectin), was further investigated using RNA interference technology, and decreased POEM expression in 4T1 tumors significantly inhibited spontaneous metastasis to the lung, kidney and bone. POEM is a secreted ECM molecule that has been reported to be involved in kidney morphogenesis and the development of bone (Brandenberger et al., 2001; Morimura et al., 2001), and so it is logical to assume that the expression of POEM by the tumor cells in the murine model is critical for the establishment of metastases in these organs. While genetic studies of clinical specimens will continue to be informative, they provide only a snapshot of a complicated disease state, and there are few experimental opportunities in such analyses. Thus, the study of breast cancer progression requires experimental models for the investigation of the links between genetic profiles and a more aggressive tumor phenotype. Animal models provide a powerful resource for the identification and investigation of genes essential in distinct steps of the metastatic cascade, in sitespecific homing, in complex tumor-host interactions, and enable the identification of targets that are optimal for therapeutic perturbation. The unique advantage of the clonal metastasis models described above is the ability to profile cells of opposing metastatic phenotype that originate from a common genetic background. The human origin of the cells and the spontaneous acquisition of the distinct phenotypes make them optimal for investigations into genetic changes that correlate with metastatic sufficiency. Investigators can alter the expression or activity of single, or multiplexes of candidate genes in the clonal lines with known metastatic characteristics and monitor which specific mechanisms are perturbed by comparison with isogenic controls. Many of these candidate genes will come from tissue-based analyses, such as those described above. There are certainly limitations regarding extrapolation from studies in a murine host to the human clinical situation, perhaps particularly acute when investigating tumor-stroma interactions, but in vivo functional analyses can distinguish which genes are essential in tumor progression, and better delineate pathway and signaling network interactions in tumor cells, and thus may in turn aid the prioritization of genes in prognostic signatures. In this way the two fields can overlap and mutually benefit. Signature overlap Given the differences in experimental design used in the production of the nine signatures described in this review,
the presence of specific genes in multiple tissue or model signatures, or better yet, in both, might implicate pathways or mechanisms for further research focus. The nine signatures contain a total of 683 non-redundant genes for which a unique gene identifier (Entrez Gene ID) is available to all signatures, and of these, 47 genes were found to appear in at least two of the nine signatures. The details of the distribution of these signature intersect genes is depicted in Table 1. Grouping genes by cellular location revealed that the most prevalent (34%) localization was extracellular. This is a common theme revealed in metastasis studies due to the complex interactions that need to occur with the secondary tissue environment and the requirement for extracellular matrix remodeling during the development of metastasis. This was reinforced by the fact that 23% of common genes were localized to the plasma membrane. The majority of genes were common to only two signatures, just a few appeared in three signatures. The overlap between tissue-based studies was weak (signature gene range 17–83). Only five genes appeared in two signatures resulting from tissue-based microarray studies (Table 1). There was no overlap between the 17-gene metastasis signature of Ramaswamy et al. (2003) with any other tissue-based study. Of the microarray studies discussed in this review, those performed by van ‘t Veer et al. (2002) and Wang et al. (2005) are the most aligned in many ways. While there were many differences between the studies with respect to population, clinical information availability, platform, and statistical treatment, both studies evaluated lymph-node negative samples and the platforms had thousands of genes in common, yet only one gene, cyclin E2 (CCNE2), was common to both signatures. Three genes (LOC51203, PRC1 and L2DTL) of the 64 genes in the study by Pawitan et al. (2005) were present in the 70-gene prognostic signature identified by van ‘t Veer and colleagues, and one gene (MLF1lP) was present in both the Pawitan and the Wang signatures. Although there are clear differences between studies, including selection of patients according to different inclusion and exclusion criteria, different gene expression platforms, and mathematical analysis of the data, thousands of genes were common to all studies and the overall goals were similar. The lack of convergence of breast cancer prognostic signatures to date makes it difficult to discern which signatures are the most accurate and optimal for prognosis and/ or potential therapeutic development going forward. There was significantly more overlap between model signatures (signature gene range 85–143) than that observed between tissue signatures. A total of 42 genes were found to be present in at least two model-based signatures. Although each model was derived from a different original tumor source, the clonality of the models represents a far simpler situation than that found in excised tumors consisting of complex mixtures of tumor and non-tumor cell components. The largest overlap (ten genes) was between the MDAMB-231 bone and lung metastasis signatures. This is likely due to the fact that both signatures are derived from the same parental cell line, MDA-MB-231. However, there was
Cytogenet Genome Res 118:116–129 (2007)
121
Table 1. Genes common to at least two breast cancer metastasis signatures amongst five animal models and four tissue-based expression microarray studies
ANXA2 CLU COL1A1 COL1A2 COL5A1 COL6A1 CTGF FN1 CXCL1 CYR61 LAMB1 LOXL2 LTBP1 MGP MMP1 SERPINE2 ANXA1 HLA-DPA1 HLA-DPB1 HLA-DRB1 LY6E PPL PTK7 CXCR4 ABCC3 TNFSF10 GPR56 EPHX1 ARHGDIB KRT86 TPM2 DLC1 CRYAB ID1 SOX4 HIST1H2AC PRC1 CCNE2 ARNT2 NUSAP1 GTSE1 DTL SPANX MLF1IP C10orf116 C14orf139 ANXA2P1
15q21–q22 8p21–p12 17q21.33 7q22.1 9q34.2–q34.3 21q22.3 6q23.1 2q34 4q21 1p31–p22 7q22 8p21.3–p21.2 2p22–p21 12p13.1–p12.3 11q22.3 2q33–q35 9q12–q21.2 6p21.3 6p21.3 6p21.3 8q24.3 16p13.3 6p21.1–p12.2 2q21 17q22 3q26 16q12.2–q21 1q42.1 12p12.3 12q13 9p13.2–p13.1 8p22 11q22.3–q23.1 20q11 6p22.3 6p21.3 15q26.1 8q22.1 15q24 15q15.1 22q13.2–q13.3 1q32.1–32.2 Xq27.1 4q35.1 10q23.2 14q32.13 4q21–q31
302 1191 1277 1278 1289 1291 1490 2335 2919 3491 3912 4017 4052 4256 4312 5270 301 3113 3115 3123 4061 5493 5754 7852 8714 8743 9289 2052 397 3887 7169 10395 1410 3397 6659 8334 9055 9134 9915 51203 51512 51514 *** 79682 10974 79686 303
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 3 4 4 4 4 5 5 5 5 5 5 5 5 5 5 5 5 6 6 NA
+
+
+
+ + +
+
+
E
F
+
+ +
+
+
+ +
+
+
+
+ + +
+
+ +
+ +
+
+ + +
+ + + +
+ + +
+ + +
+
+
+
+
+ + + +
+ +
I
+ + +
+ +
+ + + +
H
Human tumors
+
+ +
G
MDA-231-LUNG
Sub- A B C D cell. loca- Animal models liz.b
GILM2
Entrez genea
MDA-435
Chromosomal location
MURINE-4T1c
Annexin A2 Clusterin Collagen, type I, alpha 1 Collagen, type I, alpha 2 Collagen, type V, alpha 1 Collagen, type VI, alpha 1 Connective tissue growth factor Fibronectin 1 Chemokine (C-X-C motif) ligand 1 Cysteine-rich, angiogenic inducer, 61 Laminin, beta 1 Lysyl oxidase-like 2 Latent transforming growth factor beta binding protein 1 Matrix Gla protein Matrix metallopeptidase 1 (interstitial collagenase) Serpin peptidase inhibitor Annexin A1 Major histocompatibility complex, class II, DP alpha 1 Major histocompatibility complex, class II, DP beta 1 Major histocompatibility complex, class II, DR beta 1 Lymphocyte antigen 6 complex, locus E Periplakin PTK7 protein tyrosine kinase 7 Chemokine (C-X-C motif) receptor 4 ATP-binding cassette, sub-family C (CFTR/MRP), member 3 Tumor necrosis factor (ligand) superfamily, member 10 G protein-coupled receptor 56 Epoxide hydrolase 1, microsomal (xenobiotic) Rho GDP dissociation inhibitor (GDI) beta Keratin, hair, basic, 1 Tropomyosin 2 (beta) Deleted in liver cancer 1 Crystallin, alpha B Inhibitor of DNA binding 1 SRY (sex determining region Y)-box 4 Histone 1, H2ac Protein regulator of cytokinesis 1 Cyclin E2 Aryl-hydrocarbon receptor nuclear translocator 2 Nucleolar and spindle associated protein 1 G-2 and S-phase expressed 1 Denticleless homolog (Drosophila) SPANX family, members A1, A2, B1, B2, C MLF1 interacting protein Chromosome 10 open reading frame 116 Chromosome 14 open reading frame 139 Annexin A2 pseudogene 1
Gene symbol
MDA-231-BONE
Gene name
+ +
+ +
+ +
+ + +
+
+ +
+ +
+ +
+
+
+
+
+ +
+
+
+
+
+
+ +
+ +
+
+
+ +
+
Studies in References: A: Kang et al. (2003); B: Eckhardt et al. (2005); C: Goodison et al. (2005); D: Kluger et al. (2005); E: Minn et al. (2005); F: Wang et al. (2005); G: Ramaswamy et al. (2002); H: Pawitan et al. (2005); I: Van ‘t Veer et al. (2002). a *** = Various SPANX family, members, Entrez Gene IDs: 30014; 64649; 64650; 64663; 64694. b Subcellular localization of gene products, 1: Extracellular space; 2: Plasma membrane; 3: Other membrane; 4: Cytoplasm; 5: Nucleus; 6: Other. c Homo sapiens homologous genes were considered for comparison (HomoloGene release 50.1).
122
Cytogenet Genome Res 118:116–129 (2007)
also significant overlap (seven genes) between the MDAMB-435 model and the GILM2 model. There was also considerable overlap (five genes) between the MDA-MB-435 model and the MDA-MB-231-bone signature, and this is in line with our recent observations that the M4A4 metastatic counterpart of the MDA-MB-435 model is also aggressively metastatic to bone when inoculated via an intra-cardiac route (article in preparation). Even though there was a relatively high intersection between the model signatures, there are major differences between the five models described in this review. The MDAMB-231 model is a xenograft model of ‘experimental’ metastasis involving the injection of human breast tumor cells directly into the circulatory system of immunocompromised mice, resulting in metastases in specific organs. This model has been valuable for analyzing the final stages of metastasis, and can identify genes that regulate the colonization of specific tissues (Kang et al., 2003; Minn et al., 2005), but does not encompass the initial stages of metastasis. Conversely, the MDA-MB-435 and the GILM2 model are xenograft models of ‘spontaneous’ metastasis involving the formation of a primary tumor in an orthotopic site, in this case breast, from which the cells must disseminate naturally. These spontaneous metastasis models recapitulate all of the steps of metastatic efficiency, from escape from the primary tumor to the establishment of distant metastases, and perhaps mimic more accurately the progression of breast cancer observed clinically. However, the MDA-MB231, MDA-MB-435 and the GILM2 models involve human tumor cells forming tumors in a mouse host and, therefore, may not best mimic important tumor-host interactions. The 4T1 mouse model has the advantage of being syngeneic and so may provide more insight into such interactions, but on the other hand, the genes identified as being associated with the tumor cell phenotype may not be functionally equivalent in human cells. Rather than being a hindrance to forming converging hypotheses of metastasis, the testing of candidate metastasis genes across these models with their inherent differences will provide powerful validation of specific gene functionality. A total of seven genes were found to be common to at least one tissue-based signature and one model-based signature (Table 1). Three to four such intersect genes were present in each of the tissue-based signatures, except for the 70-gene signature of van ‘t Veer et al. which had no intersect with any model signature. All model signatures were represented in the tissue vs. model intersect. A total of four genes were present in at least three signatures. Three of these (HLA-DPB1, ARHGDIB and C14orf139) were present in two model signatures and one tissue signature, whereas the CTGF gene was present in three model signatures but no tissue signature. Chromosomal location and alignment with CGH studies Alignment of the genes in the signature intersect list with chromosomal location reveals that they are randomly spread across the genome with extremes being an absence of any genes on chromosome 5, and seven genes being located on
chromosome 6. Yet, per chromosome the intersect genes appear to be tightly clustered in many cases. For example, five of the seven genes on chromosome 6 are located within 6p22]p21, both genes on chromosome 7 are located at 7q22, and all three genes on chromosome 11 are located at 11q22]q23. The NUSAP1 gene located at 15q15.1 is one of the few genes that were present in two of the tissue-based analyses. Interestingly, alignment of our 171-gene metastasis signature using ‘Genome View’ revealed that three of the genes more highly expressed in metastatic M4A4 cells were grouped together at 15q15 (Goodison et al., 2005b). Moreover, this region was revealed to be involved in cell-specific chromosomal rearrangements in the molecular cytogenetic analysis of MDA-MB-435 sub-lines (Goodison et al., 2005a). Although specific genes may not make the overlap lists, future meta-analyses of gene expression profiles with respect to chromosomal location may reveal metastasis-associated DNA hot-spots. Notably, many of the intersect genes are located at loci revealed by CGH to be perturbed in breast cancer. The one gene on chromosome 3 that was present in at least two signatures is TNFSF10, a member of the tumor necrosis factor ligand superfamily, located at 3q26. This region was found to be a strong predictor of breast cancer recurrence by CGH (Janssen et al., 2003). Furthermore, four clusters of intersect genes align very closely with recent reports of copy number alterations which correlate with the advancing nature of breast tumors. These include two genes located at 8q22]q24, two genes at 12p13.1]p12.3, one gene at 16p13, and the two genes on chromosome 17 at 17q21]q22. These four regions were all identified through the use of array-CGH to be amplified in late-stage breast cancer (Yao et al., 2006). The considerable alignment between the CGH and the gene expression microarray data suggests that the combination of DNAand RNA-based approaches can inform and guide the search for essential cancer-associated genetic events. Chemokines A common theme throughout the cell line model metastasis signatures are components of the human chemokine system. The chemokines and their receptors are a family of small secreted molecules that regulate the migration around the body of cells in the lymphoid system, a process that shares many characteristics with the successful dissemination of tumor cells in the body. Chemokine receptors are G-protein-coupled cell surface proteins expressed on leukocytes and many nonhematopoietic cells. These receptors bind various chemokines that are constitutively expressed in distinct tissue microenvironments. The chemokine CXCL1 and the chemokine receptor CXCR4 are present in the signature intersect gene list (Table 1). CXCR4 is the most widely expressed chemokine receptor in many different cancers. Primary breast tumors express the CXCR4 receptor, and target sites of breast cancer metastases, such as lung and bone, express more CXCL12, the CXCR4 ligand, than other organs (Muller et al., 2001). Antibody, or peptide-mediated blocking of the CXCR4 receptor decreased breast cancer cell invasiveness in vitro and also the bulk of metas-
Cytogenet Genome Res 118:116–129 (2007)
123
tases in the MDA-MB-231 xenograft mouse model (Muller et al., 2001; Liang et al., 2004; Lapteva et al., 2005). Smith et al. (2004) used stable RNAi to reduce expression of CXCR4 in murine 4T1 cells, a highly metastatic mammary cancer cell line that is a model for advanced human breast cancer. The knockdown of CXCR4 significantly limited the growth of orthotopically transplanted breast cancer cells in all mice and even prevented primary tumor formation in some. Most importantly, all mice transplanted with CXCR4 RNAi cells survived without developing macroscopic metastases (Smith et al., 2004). Recent studies have also identified a role for CXCL12 as a proliferation factor for breast tumor cells and in recruiting progenitor endothelial cells required for angiogenesis (Orimo et al., 2005). CXCR4 expression is also a prognostic marker in various types of cancer, including breast carcinoma. In a recent histological study of over 2000 breast carcinoma cases, CXCR4 was found to be expressed in the majority of cancers, and cytoplasmic expression was associated with parameters of tumor aggression, including grade and lymph node status, and had prognostic value (Salvucci et al., 2006). G protein-coupled receptors are considered among the most desirable targets for drug development (Li et al., 2005), thus, CXCR4 antagonists may become effective agents for the treatment of various malignancies. CCN family The connective tissue growth factor gene (CTGF) was present in three signatures, and the closely related CYR61 gene was present in two signatures. These genes encode proteins that belong to the CCN family of secreted regulatory factors involved in angiogenesis, chondrogenesis, and wound healing (Brigstock, 1999; Perbal, 2004). CTGF and CYR61 also promote endothelial cell growth, migration, adhesion, and survival in vitro, and many of these actions are mediated at least partly through interactions with integrins (Lau and Lam, 1999). The proangiogenic activity of CTGF and CYR61 supports a role in the establishment and functioning of the vasculature in metastasis. Recent studies have revealed an influence of CTGF in osteolytic bone metastases and its expression shown to be regulated by PTHrP (Shimo et al., 2006). In a large immunochemical study of human breast tumor specimens, high levels of CYR61, but low levels of CTGF were associated with poor prognosis and metastatic disease (Jiang et al., 2004). Interestingly, both genes are transcriptionally regulated by TGF (Bartholin et al., 2007), a multifunctional growth factor, which is thought to promote breast tumor metastasis and invasiveness (Beisner et al., 2006). ARHGD1B Another gene that was present in three metastasis signatures was the Rho GDP dissociation inhibitor (GDI) beta (ARHGD1B) gene, a member of the Rho GDI family which are major regulators of Rho GTPases. Evidence for the influence of Rho GTPase signaling in tumor metastasis is growing. The expression of the RhoC molecule was identified as being correlated with metastatic propensity in an increasingly metastatic series of the melanoma A375 cell
124
Cytogenet Genome Res 118:116–129 (2007)
line and the Rho-GTPases, RhoGD12 and Rac1, have been associated with aggressive phenotypes in experimental metastasis assays when testing bladder and breast cell line models respectively (Bourguignon et al., 2000; Seraj et al., 2000). In our own work, restoration of one of these GTPases, deleted in liver cancer-1 (DLC-1), in metastatic M4A4 cells resulted in the inhibition of migration and invasion in vitro and a significant reduction in the ability of these cells to form pulmonary metastases in athymic mice. Although RhoGDIs are known to inhibit Rho activities, recent studies indicate that RhoGDIs can also act as positive regulators through their ability to target Rho GTPases to specific subcellular membranes or to protect the GTPases from degradation by caspases. This dual role makes them promising cellular targets for novel anticancer drugs (Fritz and Kaina, 2006). While the signatures obtained from different microarray platforms might reveal different gene sets, they may be reporting related biological processes. Annotation of the 47 signature intersect genes using GenMapp (http://www. genmapp.org/) and KEGG (http://www.genome.jp/kegg/) pathway information summaries (Doniger et al., 2003; Kanehisa et al., 2006) revealed only one significant grouping. Three genes (FN1, LAMB1 and COL1A1) belong to the inflammatory response pathway. This finding perhaps reinforces the concept that gene expression programs related to normal physiological responses, as described above with respect to wound healing, may confer an increased risk of metastasis (Chang et al., 2004). Investigation of potential interaction between the gene products of the 47 signature intersect genes was queried using the STRING (Search Tool for the Retrieval of Interacting Genes/Proteins) database (http://string.embl.de/) (von Mering et al., 2005). Evidence for protein associations derived from experimental data and from the mining of databases and the literature was sought using the Protein Mode for maximum specificity. With the exception of C14orf139, DTL, and MLF1IP genes, and the pseudogene ANXA2P1, all gene products had annotations in the STRING database. Without entering into the detailed specific interactions here, it is worth noting that the analysis revealed that the transcription factor Sp1 may be a unifying factor for many of the proteins present in the signature intersect list. Sp1 was present in the node that connected a group that included CTGF, CYR61, MMP1, and several collagens with the chemokines CXCR4 and CXCL1. Sp1 protein expression is known to be elevated in breast carcinomas (Zannetti et al., 2000), and several studies show that Sp1 transcriptionally regulates many genes previously implicated in cancer progression, including VEGF (Ryuto et al., 1996; Abdelrahim et al., 2004), CCND1, FOS, and the antiapoptosis BCL2 gene (Safe and Abdelrahim, 2005). Furthermore, Sp proteins have been shown to regulate the expression of TGF receptors (Ji et al., 1997; Jennings et al., 2001) and thus will impact TGF signaling, which in turn regulates the CCN proteins discussed above. While additional research is required to determine direct linkages between the expression of Sp1 and its downstream factors in various tumor types, as a protein high in the regulatory hierarchy,
Sp1 may be an important prognostic factor and therapeutic target for metastatic disease. If a comparison of the intersect of just the nine signatures described in this review can implicate some common themes, a comprehensive meta-analysis of all available, unprocessed, metastasis-associated gene expression microarray data with more of a focus on biochemical and signaling pathways than gene identities may reveal much about the mechanisms of metastatic efficiency. Future directions
Genes revealed in analysis of multiple models, and those that are part of any overlap between models and tissuebased studies, will be a logical focus for functional investigation. As described above, relevant models for such investigations do exist, but it would be beneficial to develop more such models in order to identify those genes and pathways that are implicated across a number of platforms and are thus likely to be pivotal to metastatic efficiency. Likewise, more tissue-based studies are needed, both for discovery and validation of existing candidate genes. As discussed above, little overlap between tissue-based studies has been evident to date, presumably due to differences in tissue sources and processing and availability of consistent clinical information. The initiation of large-scale consortia such as TransBig and PACT will greatly facilitate this through standardization of all parameters. While there are economical and ease-of-use reasons to reduce the number of genes in diagnostic and prognostic signatures for routine clinical use, global gene expression profiling can have more specific roles. For example, based upon the expression of drug targets, and/or the expression of genes that metabolize these drugs, gene expression profiling may guide adjuvant therapy for individuals. Two studies that highlight this approach have reported an association in breast cancer patients between gene expression signatures and drug sensitivity to docetaxel or to a combination regimen containing paclitaxel, fluorouracil, doxorubicin and cyclophosphamide (Brown et al., 2004; Villeneuve et al., 2006). This use of gene expression profiling is likely to have a more rapid role in tailoring individual patient treatment. Prognostic signatures gleaned from gene expression data alone have been achieved through normalized microarray data comparison across literally hundreds of specimens. However, the heterogeneity of breast cancer as a disease, a fact confirmed by gene expression studies, plus the variation in patient genetic background makes the goal of predicting the outcome for an individual patient far from achievable at this stage. One way to improve upon this situation is to further stratify patients using clinical measures and current understanding of disease pathology. For example, the expression of the estrogen receptor (ER) and/or the Her2-neu gene at diagnosis is currently used to subdivide patients. ER and Her2-neu expression have a marked influence on the expression of many of the genes associated with breast cancer, and are thought to have an important impact
on survival and are used as independent prognostic factors (Hu et al., 2006; Nicolini et al., 2006). Investigators have begun to use clinical parameters to subdivide patients in order to further refine the prognostic classifiers for specific circumstances. For example, by combining expression data with ER status and age as clinical variables, Dai et al. (2005) identified a subgroup of patients in which a set of proliferation-associated genes was a strong predictor of poor outcome and occurrence of metastasis. The majority of expression studies to date have attempted to develop genetic marker-based prognostic systems that might replace the existing clinical criteria, rather than incorporating the valuable clinical information contained in established clinical markers. Given the complexity of breast cancer prognosis, a more promising strategy may be to combine both clinical and genetic marker information that may be complementary (Brenton et al., 2005). Along these lines, Pittman et al. (2004) have described a comprehensive modeling approach based on statistical classification tree models that evaluate the contributions of multiple forms of data, both clinical and genomic, in order to define multiple risk factors that associate with clinical outcome. A study of primary breast cancer recurrence demonstrated that models using ‘metagenes’ derived from microarray expression data combined with traditional clinical risk factors improved the accuracy of prediction on an individual basis over genomic or clinical data alone (Pittman et al., 2004). Furthermore, we have recently performed a computational study using the Dutch 70-gene prognosis signature and associated clinical information. The recently proposed I-RELIEF algorithm was used to identify a ‘hybrid signature’ through the combination of both genetic and clinical markers. The hybrid signature performed significantly better, with respect to specificity and odds-ratio, than the 70-gene signature, clinical markers alone and the St Gallen and NIH consensus criteria (Sun et al., 2007). One of the major differences between gene expressionprofile studies performed to date is in data processing. The advent of microarray technology has spawned a whole new field of bioinformatics which has the common goal of optimizing the accuracy and the applicability of the information gleaned from the wealth of data now available, but individual studies have often used quite different approaches which do not always facilitate inter-study comparisons. A number of groups have tackled this by designing algorithms that enable meta-analysis of publicly available data, including analysis across multiple array platforms (Bammler et al., 2005; Rhodes et al., 2005; Segal et al., 2005). These analyses seek to not only identify inter-study overlap but also to identify multidimensional interaction networks which in turn may implicate molecular hierarchies and regulatory mechanisms involved in cancer gene expression. Rhodes et al. (2004) analyzed compiled multiple data sets from some 3,700 cancer samples and developed an analytical strategy to assess the intersection between profiles. This approach identified components of transcriptional profiles which are similar across many cancer types, including breast. Specific to breast cancer, Shen et al. (2004) applied a Bayesian mod-
Cytogenet Genome Res 118:116–129 (2007)
125
eling strategy to analyze four independent microarray studies to derive a ‘meta-signature’ associated with breast cancer prognosis. The meta-signature had better prognostic performance than any of the classifiers of survival in each study, and which had minimal overlap with each other. These studies reveal that there are many legitimate ways to analyze array data sets, and it is hoped that the formation of consortia to monitor large-scale tissue-based studies will employ a standardized arsenal of bioinformatic and statistical analyses. Ongoing computational developments that enable inter-study comparison and the incorporation of distinct forms of data, including genomic and clinical data, will provide platforms for the refinement of cancer-related gene signatures and lead to more economical and accurate prognostic systems that may facilitate personalized patient evaluation and treatment decisions. These integrated analyses also have the potential to highlight pivotal genes and pathways that are likely part of the biological driving force of metastasis. Another way to significantly improve the quality of data resulting from tissue profile analyses is to provide more accurate histopathological details of each sample upfront. Concordance between two pathologists has been investigated and found to range from 50% to 85% (Robbins et al., 1995), and inaccuracy in sample evaluation may account for considerable error when array data is obtained from partial tissue specimens and subsequently related to parameters such as disease staging and histological grade. Furthermore, solid tumors are complex entities composed of malignant cells mixed and interacting with nonmalignant cells such as normal or benign epithelia, stromal counterparts and lymphocytic infiltrate. Thus, molecular analyses by standard gene expression profiling are limited when tissues are crudely homogenized causing loss of information on non-tumor cell types in sample preparation. Gene expression differences derived from such tumor samples may primarily reflect varying proportions of the non-neoplastic tumor components. In order to try to overcome these limitations we have previously employed a regression-based informatics approach for the identification of cell-type-specific patterns of gene expression in prostate cancer (Stuart et al., 2004). Through intense histological evaluation, we scored 88 prostate specimens for relative content of neoplastic and non-neoplastic components. The proportions of these cell types were then linked in silico to gene expression levels determined by microarray analysis, revealing unique cell-specific profiles. Gene expression differences for malignant and nonmalignant epithelial cells could be identified without being confounded by contributions from stroma that dominate many samples. Validation of selected cellspecific expression patterns confirmed that this analysis allowed segregation of molecular markers into more discrete and informative groups. This investigative approach is applicable to tumor marker discovery in any solid tumor and is an example of how more detailed evaluation of the tissue specimen can further refine expression profile data. Another way to separate the cells of a complex tissue from each other is through microdissection. This technique
126
Cytogenet Genome Res 118:116–129 (2007)
has most often been used to evaluate only several genes per sample (Wang-Rodriguez et al., 2003), but the advent of semi-automated, laser capture microdissection (LCM) systems now makes it feasible to use this procedure to obtain enough material for microarray analysis (Nishidate et al., 2004; Dahl et al., 2006; Schuetz et al., 2006; Yang et al., 2006). The combination of laser microdissection and gene expression profiling has been used to explore the gene expression changes that are associated with the early stages of breast cancer progression and ER-regulated genes (Ma et al., 2003; Yang et al., 2006). In a study by Dahl et al. the investigation of matched pairs of invasive ductal breast cancer and corresponding benign breast tissue by LCM and cDNA array profiling ultimately led to the identification of karyopherin alpha2 (KPNA2) expression as being associated with shorter overall survival and recurrence-free survival (Dahl et al., 2006). Finally, it has to be stated that genomic approaches are unlikely to be adequate as a sole prognostic and predictive platform in breast cancer. Most transcriptome array analyses implicitly treat mRNA expression as a surrogate for protein activity level, an assumption that does not account for processes such as mRNA stability, protein degradation and post-translational modifications. The key proteins driving tumor progression may very well undergo abnormal posttranslational modifications. While the animal models of metastasis described above lend themselves to proteomic analysis (Kreunin et al., 2004), proteomic analysis of complex tumor tissues is far more difficult. Proteomic techniques are far more limited with respect to high-throughput screening platforms. Current techniques cannot yet achieve high proteome coverage rates and they need large amounts of material given that there is no method for amplification. Due to the vastly superior gene screening advantages of microarrays, it is logical that transcriptome data should guide proteomics, for example, validation of array data should be performed at the protein level wherever possible.
References
Abdelrahim M, Smith R 3rd, Burghardt R, Safe S: Role of Sp proteins in regulation of vascular endothelial growth factor expression and proliferation of pancreatic cancer cells. Cancer Res 64:6740–6749 (2004). Adeyinka A, Kytola S, Mertens F, Pandis N, Larsson C: Spectral karyotyping and chromosome banding studies of primary breast carcinomas and their lymph node metastases. Int J Mol Med 5:235–240 (2000). Agarwal D, Goodison S, Nicholson B, Tarin D, Urquidi V: Expression of matrix metalloproteinase 8 (MMP-8) and tyrosinase-related protein-1 (TYRP-1) correlates with the absence of metastasis in an isogenic human breast cancer model. Differentiation 71: 114–125 (2003). Albertson DG: Profiling breast cancer by array CGH. Breast Cancer Res Treat 78: 289–298 (2003). Aubele M, Auer G, Braselmann H, Nahrig J, Zitzelsberger H, et al: Chromosomal imbalances are associated with metastasis-free survival in breast cancer patients. Anal Cell Pathol 24:77– 87 (2002).
Bammler T, Beyer RP, Bhattacharya S, Boorman GA, Boyles A, et al: Standardizing global gene expression analysis between laboratories and across platforms. Nat Methods 2: 351–356 (2005). Barlund M, Tirkkonen M, Forozan F, Tanner MM, Kallioniemi O, Kallioniemi A: Increased copy number at 17q22]q24 by CGH in breast cancer is due to high-level amplification of two separate regions. Genes Chromosomes Cancer 20: 372–376 (1997). Bartholin L, Wessner LL, Chirgwin JM, Guise TA: The human CYR61 gene is a transcriptional target of transforming growth factor beta in cancer cells. Cancer Lett 246:230–236 (2007). Bastian BC, Wesselmann U, Pinkel D, Leboit PE: Molecular cytogenetic analysis of spitz nevi shows clear differences to melanoma. J Invest Dermatol 113:1065–1069 (1999). Beisner J, Buck MB, Fritz P, Dippon J, Schwab M, et al: A novel functional polymorphism in the transforming growth factor-beta2 gene promoter and tumor progression in breast cancer. Cancer Res 66: 7554–7561 (2006). Bieche I, Lidereau R: Genetic alterations in breast cancer. Genes Chromosomes Cancer 14: 227– 251 (1995). Bissell MJ, Radisky D: Putting tumours in context. Nat Rev Cancer 1: 46–54 (2001). Blegen H, Will JS, Ghadimi BM, Nash HP, Zetterberg A, et al: DNA amplifications and aneuploidy, high proliferative activity and impaired cell cycle control characterize breast carcinomas with poor prognosis. Anal Cell Pathol 25: 103– 114 (2003). Bourguignon LY, Zhu H, Shao L, Chen YW: Ankyrin-tiam1 interaction promotes Rac1 signaling and metastatic breast tumor cell invasion and migration. J Cell Biol 150: 177–191 (2000). Brandenberger R, Schmidt A, Linton J, Wang D, Backus C, et al: Identification and characterization of a novel extracellular matrix protein nephronectin that is associated with integrin alpha8beta1 in the embryonic kidney. J Cell Biol 154: 447–458 (2001). Brenton JD, Carey LA, Ahmed AA, Caldas C: Molecular classification and molecular forecasting of breast cancer: Ready for clinical application? J Clin Oncol 23:7350–7360 (2005). Brigstock DR: The connective tissue growth factor/cysteine-rich 61/nephroblastoma overexpressed (CCN) family. Endocr Rev 20: 189–206 (1999). Brown I, Shalli K, McDonald SL, Moir SE, Hutcheon AW, et al: Reduced expression of p27 is a novel mechanism of docetaxel resistance in breast cancer cells. Breast Cancer Res 6:R601– 607 (2004). Chambers AF, Naumov GN, Vantyghem SA, Tuck AB: Molecular biology of breast cancer metastasis. Clinical implications of experimental studies on metastatic inefficiency. Breast Cancer Res 2:400–407 (2000). Chang HY, Sneddon JB, Alizadeh AA, Sood R, West RB, et al: Gene expression signature of fibroblast serum response predicts human cancer progression: Similarities between tumors and wounds. PLoS Biol 2:E7 (2004). Chang HY, Nuyten DS, Sneddon JB, Hastie T, Tibshirani R, et al: Robustness, scalability, and integration of a wound-response gene expression signature in predicting breast cancer survival. Proc Natl Acad Sci USA 102: 3738–3743 (2005). Dahl E, Kristiansen G, Gottlob K, Klaman I, Ebner E, et al: Molecular profiling of laser-microdissected matched tumor and normal breast tissue identifies karyopherin alpha2 as a potential novel prognostic marker in breast cancer. Clin Cancer Res 12: 3950–3960 (2006).
Dai H, van ‘t Veer L, Lamb J, He YD, Mao M, et al: A cell proliferation signature is a marker of extremely poor outcome in a subpopulation of breast cancer patients. Cancer Res 65: 4059– 4066 (2005). Devilee P, Cornelisse CJ: Somatic genetic changes in human breast cancer. Biochim Biophys Acta 1198:113–130 (1994). Doniger SW, Salomonis N, Dahlquist KD, Vranizan K, Lawlor SC, Conklin BR: Mappfinder: Using gene ontology and genmapp to create a global gene-expression profile from microarray data. Genome Biol 4:R7 (2003). Eckhardt BL, Parker BS, van Laar RK, Restall CM, Natoli AL, et al: Genomic analysis of a spontaneous model of breast cancer metastasis to bone reveals a role for the extracellular matrix. Mol Cancer Res 3: 1–13 (2005). Fidler IJ, Kripke ML: Metastasis results from preexisting variant cells within a malignant tumor. Science 197:893–895 (1977). Fritz G, Kaina B: Rho GTPases: Promising cellular targets for novel anticancer drugs. Curr Cancer Drug Targets 6:1–14 (2006). Fritz G, Just I, Kaina B: Rho GTPases are over-expressed in human tumors. Int J Cancer 81:682– 687 (1999). Goodison S, Kawai K, Hihara J, Jiang P, Yang M, et al: Prolonged dormancy and site-specific growth potential of cancer cells spontaneously disseminated from nonmetastatic breast tumors as revealed by labeling with green fluorescent protein. Clin Cancer Res 9: 3808–3814 (2003). Goodison S, Viars C, Urquidi V: Molecular cytogenetic analysis of a human breast metastasis model: Identification of phenotype-specific chromosomal rearrangements. Cancer Genet Cytogenet 156:37–48 (2005a). Goodison S, Yuan J, Sloan D, Kim R, Li C, et al: The RhoGAP protein DLC-1 functions as a metastasis suppressor in breast cancer cells. Cancer Res 65:6042–6053 (2005b). Han W, Han MR, Kang JJ, Bae JY, Lee JH, et al: Genomic alterations identified by array comparative genomic hybridization as prognostic markers in tamoxifen-treated estrogen receptor-positive breast cancer. BMC Cancer 6: 92 (2006). Hedenfalk I, Ringner M, Ben-Dor A, Yakhini Z, Chen Y, et al: Molecular classification of familial non-BRCA1/BRCA2 breast cancer. Proc Natl Acad Sci USA 100: 2532–2537 (2003). Hu Z, Fan C, Oh DS, Marron JS, He X, et al: The molecular portraits of breast tumors are conserved across microarray platforms. BMC Genomics 7: 96 (2006). Huang E, Cheng SH, Dressman H, Pittman J, Tsou MH, et al: Gene expression predictors of breast cancer outcomes. Lancet 361: 1590–1596 (2003). Hurst J, Maniar N, Tombarkiewicz J, Lucas F, Roberson C, et al: A novel model of a metastatic human breast tumour xenograft line. Br J Cancer 68:274–276 (1993). Hyman E, Kauraniemi P, Hautaniemi S, Wolf M, Mousses S, et al: Impact of DNA amplification on gene expression patterns in breast cancer. Cancer Res 62:6240–6245 (2002). Isola JJ, Kallioniemi OP, Chu LW, Fuqua SA, Hilsenbeck SG, et al: Genetic aberrations detected by comparative genomic hybridization predict outcome in node-negative breast cancer. Am J Pathol 147: 905–911 (1995).
Janssen EA, Baak JP, Guervos MA, van Diest PJ, Jiwa M, Hermsen MA: In lymph node-negative invasive breast carcinomas, specific chromosomal aberrations are strongly associated with high mitotic activity and predict outcome more accurately than grade, tumour diameter, and oestrogen receptor. J Pathol 201: 555–561 (2003). Jennings R, Alsarraj M, Wright KL, Munoz-Antonia T: Regulation of the human transforming growth factor beta type II receptor gene promoter by novel Sp1 sites. Oncogene 20: 6899– 6909 (2001). Ji C, Casinghino S, McCarthy TL, Centrella M: Multiple and essential Sp1 binding sites in the promoter for transforming growth factor-beta type I receptor. J Biol Chem 272: 21260–21267 (1997). Jiang WG, Watkins G, Fodstad O, Douglas-Jones A, Mokbel K, Mansel RE: Differential expression of the CCN family members Cyr61, CTGF and Nov in human breast cancer. Endocr Relat Cancer 11:781–791 (2004). Kallioniemi A, Kallioniemi OP, Piper J, Tanner M, Stokke T, et al: Detection and mapping of amplified DNA sequences in breast cancer by comparative genomic hybridization. Proc Natl Acad Sci USA 91:2156–2160 (1994). Kanehisa M, Goto S, Hattori M, Aoki-Kinoshita KF, Itoh M, et al: From genomics to chemical genomics: New developments in KEGG. Nucleic Acids Res 34:D354–357 (2006). Kang Y, Siegel PM, Shu W, Drobnjak M, Kakonen SM, et al: A multigenic program mediating breast cancer metastasis to bone. Cancer Cell 3: 537–549 (2003). Kauraniemi P, Barlund M, Monni O, Kallioniemi A: New amplified and highly expressed genes discovered in the ERBB2 amplicon in breast cancer by cDNA microarrays. Cancer Res 61: 8235– 8240 (2001). Kawai K, Viars C, Arden K, Tarin D, Urquidi V, Goodison S: Comprehensive karyotyping of the HT-29 colon adenocarcinoma cell line. Genes Chromosomes Cancer 34: 1–8 (2002). Kluger HM, Chelouche Lev D, Kluger Y, McCarthy MM, Kiriakova G, et al: Using a xenograft model of human breast cancer metastasis to find genes associated with clinically aggressive disease. Cancer Res 65: 5578–5587 (2005). Knutsen T, Gobu V, Knaus R, Padilla-Nash H, Augustus M, et al: The interactive online SKY/mFISH & CGH database and the Entrez cancer chromosomes search database: Linkage of chromosomal aberrations with the genome sequence. Genes Chromosomes Cancer 44:52–64 (2005). Kreunin P, Urquidi V, Lubman DM, Goodison S: Identification of metastasis-associated proteins in a human tumor metastasis model using the mass-mapping technique. Proteomics 4: 2754– 2765 (2004). Kristensen VN, Sorlie T, Geisler J, Langerod A, Yoshimura N, et al: Gene expression profiling of breast cancer in relation to estrogen receptor status and estrogen-metabolizing enzymes: Clinical implications. Clin Cancer Res 11:878s883s (2005). Kytola S, Rummukainen J, Nordgren A, Karhu R, Farnebo F, et al: Chromosomal alterations in 15 breast cancer cell lines by comparative genomic hybridization and spectral karyotyping. Genes Chromosomes Cancer 28: 308–317 (2000). Lapteva N, Yang AG, Sanders DE, Strube RW, Chen SY: CXCR4 knockdown by small interfering RNA abrogates breast tumor growth in vivo. Cancer Gene Ther 12:84–89 (2005).
Cytogenet Genome Res 118:116–129 (2007)
127
Lau LF, Lam SC: The CCN family of angiogenic regulators: The integrin connection. Exp Cell Res 248:44–57 (1999). Lelekakis M, Moseley JM, Martin TJ, Hards D, Williams E, et al: A novel orthotopic model of breast cancer metastasis to bone. Clin Exp Metastasis 17:163–170 (1999). Lev DC, Kiriakova G, Price JE: Selection of more aggressive variants of the gi101a human breast cancer cell line: A model for analyzing the metastatic phenotype of breast cancer. Clin Exp Metastasis 20:515–523 (2003). Li S, Huang S, Peng SB: Overexpression of G protein-coupled receptors in cancer cells: Involvement in tumor progression. Int J Oncol 27: 1329–1339 (2005). Liang Z, Wu T, Lou H, Yu X, Taichman RS, et al: Inhibition of breast cancer metastasis by selective synthetic polypeptide against CXCR4. Cancer Res 64: 4302–4308 (2004). Loi S, Desmedt C, Cardoso F, Piccart M, Sotiriou C: Breast cancer gene expression profiling: Clinical trial and practice implications. Pharmacogenomics 6:49–58 (2005). Ma XJ, Salunga R, Tuggle JT, Gaudet J, Enright E, et al: Gene expression profiles of human breast cancer progression. Proc Natl Acad Sci USA 100:5974–5979 (2003). Ma XJ, Wang Z, Ryan PD, Isakoff SJ, Barmettler A, et al: A two-gene expression ratio predicts clinical outcome in breast cancer patients treated with tamoxifen. Cancer Cell 5: 607–616 (2004). Mackay A, Jones C, Dexter T, Silva RL, Bulmer K, et al: cDNA microarray analysis of genes associated with ERBB2 (HER2/neu) overexpression in human mammary luminal epithelial cells. Oncogene 22: 2680–2688 (2003). Macville M, Veldman T, Padilla-Nash H, Wangsa D, O’Brien P, et al: Spectral karyotyping, a 24colour FISH technique for the identification of chromosomal rearrangements. Histochem Cell Biol 108: 299–305 (1997). McGuire WL: Prognostic factors for recurrence and survival in human breast cancer. Breast Cancer Res Treat 10:5–9 (1987). Minn AJ, Gupta GP, Siegel PM, Bos PD, Shu W, et al: Genes that mediate breast cancer metastasis to lung. Nature 436: 518–524 (2005). Morimura N, Tezuka Y, Watanabe N, Yasuda M, Miyatani S, et al: Molecular cloning of POEM: A novel adhesion molecule that interacts with alpha8beta1 integrin. J Biol Chem 276: 42172– 42181 (2001). Morris JS, Carter NP, Ferguson-Smith MA, Edwards PA: Cytogenetic analysis of three breast carcinoma cell lines using reverse chromosome painting. Genes Chromosomes Cancer 20:120– 139 (1997). Muller A, Homey B, Soto H, Ge N, Catron D, et al: Involvement of chemokine receptors in breast cancer metastasis. Nature 410: 50–56 (2001). Nicolini A, Carpi A, Tarro G: Biomolecular markers of breast cancer. Front Biosci 11: 1818–1843 (2006). Nishidate T, Katagiri T, Lin ML, Mano Y, Miki Y, et al: Genome-wide gene-expression profiles of breast-cancer cells purified with laser microbeam microdissection: Identification of genes associated with progression and metastasis. Int J Oncol 25:797–819 (2004). Nishizaki T, DeVries S, Chew K, Goodson WH 3rd, Ljung BM, et al: Genetic alterations in primary breast cancers and their metastases: Direct comparison using modified comparative genomic hybridization. Genes Chromosomes Cancer 19: 267–272 (1997).
128
Orimo A, Gupta PB, Sgroi DC, Arenzana-Seisdedos F, Delaunay T, et al: Stromal fibroblasts present in invasive human breast carcinomas promote tumor growth and angiogenesis through elevated SDF-1/CXCL12 secretion. Cell 121: 335– 348 (2005). Pauletti G, Godolphin W, Press MF, Slamon DJ: Detection and quantitation of HER-2/neu gene amplification in human breast cancer archival material using fluorescence in situ hybridization. Oncogene 13: 63–72 (1996). Pawitan Y, Bjohle J, Amler L, Borg AL, Egyhazi S, et al: Gene expression profiling spares early breast cancer patients from adjuvant therapy: Derived and validated in two population-based cohorts. Breast Cancer Res 7:R953–964 (2005). Perbal B: CCN proteins: Multifunctional signalling regulators. Lancet 363: 62–64 (2004). Perreard L, Fan C, Quackenbush JF, Mullins M, Gauthier NP, et al: Classification and risk stratification of invasive breast carcinomas using a real-time quantitative RT-PCR assay. Breast Cancer Res 8:R23 (2006). Pittman J, Huang E, Dressman H, Horng CF, Cheng SH, et al: Integrated modeling of clinical and gene expression information for personalized prediction of disease outcomes. Proc Natl Acad Sci USA 101:8431–8436 (2004). Popescu NC, Zimonjic DB: Chromosome and gene alterations in breast cancer as markers for diagnosis and prognosis as well as pathogenetic targets for therapy. Am J Med Genet 115: 142–149 (2002). Price JE, Polyzos A, Zhang RD, Daniels LM: Tumorigenicity and metastasis of human breast carcinoma cell lines in nude mice. Cancer Res 50:717–721 (1990). Ramaswamy S, Ross KN, Lander ES, Golub TR: A molecular signature of metastasis in primary solid tumors. Nat Genet 33: 49–54 (2003). Rhodes DR, Yu J, Shanker K, Deshpande N, Varambally R, et al: Large-scale meta-analysis of cancer microarray data identifies common transcriptional profiles of neoplastic transformation and progression. Proc Natl Acad Sci USA 101: 9309–9314 (2004). Rhodes DR, Kalyana-Sundaram S, Mahavisno V, Barrette TR, Ghosh D, Chinnaiyan AM: Mining for regulatory programs in the cancer transcriptome. Nat Genet 37: 579–583 (2005). Ried T, Just KE, Holtgreve-Grez H, du Manoir S, Speicher MR, et al: Comparative genomic hybridization of formalin-fixed, paraffin-embedded breast tumors reveals different patterns of chromosomal gains and losses in fibroadenomas and diploid and aneuploid carcinomas. Cancer Res 55: 5415–5423 (1995). Robbins P, Pinder S, de Klerk N, Dawkins H, Harvey J, et al: Histological grading of breast carcinomas: A study of interobserver agreement. Hum Pathol 26: 873–879 (1995). Ryuto M, Ono M, Izumi H, Yoshida S, Weich HA, et al: Induction of vascular endothelial growth factor by tumor necrosis factor alpha in human glioma cells. Possible roles of Sp-1. J Biol Chem 271:28220–28228 (1996). Safe S, Abdelrahim M: Sp transcription factor family and its role in cancer. Eur J Cancer 41: 2438– 2448 (2005). Salvucci O, Bouchard A, Baccarelli A, Deschenes J, Sauter G, et al: The role of CXCR4 receptor expression in breast cancer: A large tissue microarray study. Breast Cancer Res Treat 97: 275– 283 (2006). Schrock E, du Manoir S, Veldman T, Schoell B, Wienberg J, et al: Multicolor spectral karyotyping of human chromosomes. Science 273: 494– 497 (1996).
Cytogenet Genome Res 118:116–129 (2007)
Schuetz CS, Bonin M, Clare SE, Nieselt K, Sotlar K, et al: Progression-specific genes identified by expression profiling of matched ductal carcinomas in situ and invasive breast tumors, combining laser capture microdissection and oligonucleotide microarray analysis. Cancer Res 66: 5278–5286 (2006). Segal E, Friedman N, Kaminski N, Regev A, Koller D: From signatures to models: Understanding cancer using microarrays. Nat Genet 37 Suppl: S38–45 (2005). Seraj MJ, Harding MA, Gildea JJ, Welch DR, Theodorescu D: The relationship of BRMS1 and RHOGDI2 gene expression to metastatic potential in lineage related human bladder cancer cell lines. Clin Exp Metastasis 18: 519–525 (2000). Shadeo A, Lam WL: Comprehensive copy number profiles of breast cancer cell model genomes. Breast Cancer Res 8:R9 (2006). Shen R, Ghosh D, Chinnaiyan AM: Prognostic meta-signature of breast cancer developed by two-stage mixture modeling of microarray data. BMC Genomics 5: 94 (2004). Shevde LA, Welch DR: Metastasis suppressor pathways–an evolving paradigm. Cancer Lett 198: 1–20 (2003). Shimo T, Kubota S, Yoshioka N, Ibaragi S, Isowa S, et al: Pathogenic role of connective tissue growth factor (CTGF/CCN2) in osteolytic metastasis of breast cancer. J Bone Miner Res 21: 1045–1059 (2006). Smith MC, Luker KE, Garbow JR, Prior JL, Jackson E, et al: CXCR4 regulates growth of both primary and metastatic breast cancer. Cancer Res 64:8604–8612 (2004). Sokolova IA, Halling KC, Jenkins RB, Burkhardt HM, Meyer RG, et al: The development of a multitarget, multicolor fluorescence in situ hybridization assay for the detection of urothelial carcinoma in urine. J Mol Diagn 2: 116–123 (2000). Sorlie T, Perou CM, Tibshirani R, Aas T, Geisler S, et al: Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci USA 98: 10869–10874 (2001). Sorlie T, Tibshirani R, Parker J, Hastie T, Marron JS, et al: Repeated observation of breast tumor subtypes in independent gene expression data sets. Proc Natl Acad Sci USA 100: 8418–8423 (2003). Sorlie T, Wang Y, Xiao C, Johnsen H, Naume B, et al: Distinct molecular mechanisms underlying clinically relevant subtypes of breast cancer: Gene expression analyses across three different platforms. BMC Genomics 7: 127 (2006). Sotiriou C, Wirapati P, Loi S, Harris A, Fox S, et al: Gene expression profiling in breast cancer: Understanding the molecular basis of histologic grade to improve prognosis. J Natl Cancer Inst 98:262–272 (2006). Speicher MR, Gwyn Ballard S, Ward DC: Karyotyping human chromosomes by combinatorial multi-fluor FISH. Nat Genet 12: 368–375 (1996). Stacker SA, Achen MG, Jussila L, Baldwin ME, Alitalo K: Lymphangiogenesis and cancer metastasis. Nat Rev Cancer 2: 573–583 (2002). Stuart RO, Wachsman W, Berry CC, Wang-Rodriguez J, Wasserman L, et al: In silico dissection of cell-type-associated patterns of gene expression in prostate cancer. Proc Natl Acad Sci USA 101:615–620 (2004). Sugino T, Yamaguchi T, Ogura G, Saito A, Hashimoto T, et al: Morphological evidence for an invasion-independent metastasis pathway exists in multiple human cancers. BMC Med 2: 9 (2004).
Sun Y, Goodison S, Li J, Liu L, Farmerie W: Improved breast cancer prognosis through the combination of clinical and genetic markers. Bioinformatics 23: 30–37 (2007). Tanner MM, Tirkkonen M, Kallioniemi A, Isola J, Kuukasjarvi T, et al: Independent amplification and frequent co-amplification of three nonsyntenic regions on the long arm of chromosome 20 in human breast cancer. Cancer Res 56: 3441–3445 (1996). Tirkkonen M, Tanner M, Karhu R, Kallioniemi A, Isola J, Kallioniemi OP: Molecular cytogenetics of primary breast cancer by CGH. Genes Chromosomes Cancer 21: 177–184 (1998). Tuma RS: A big trial for a new technology: TransBIG project takes microarrays into clinical trials. J Natl Cancer Inst 96: 648–649 (2004). Urquidi V, Sloan D, Kawai K, Agarwal D, Woodman AC, et al: Contrasting expression of thrombospondin-1 and osteopontin correlates with absence or presence of metastatic phenotype in an isogenic model of spontaneous human breast cancer metastasis. Clin Cancer Res 8:61–74 (2002). van ‘t Veer LJ, Dai H, van de Vijver MJ, He YD, Hart AA, et al: Gene expression profiling predicts clinical outcome of breast cancer. Nature 415: 530–536 (2002). van de Vijver MJ, He YD, van ‘t Veer LJ, Dai H, Hart AA, et al: A gene-expression signature as a predictor of survival in breast cancer. N Engl J Med 347:1999–2009 (2002).
Villeneuve DJ, Hembruff SL, Veitch Z, Cecchetto M, Dew WA, Parissenti AM: cDNA microarray analysis of isogenic paclitaxel- and doxorubicin-resistant breast tumor cell lines reveals distinct drug-specific genetic signatures of resistance. Breast Cancer Res Treat 96: 17–39 (2006). von Mering C, Jensen LJ, Snel B, Hooper SD, Krupp M, et al: String: Known and predicted proteinprotein associations, integrated and transferred across organisms. Nucleic Acids Res 33:D433– 437 (2005). Wang Y, Klijn JG, Zhang Y, Sieuwerts AM, Look MP, et al: Gene-expression profiles to predict distant metastasis of lymph-node-negative primary breast cancer. Lancet 365: 671–679 (2005). Wang-Rodriguez J, Urquidi V, Rivard A, Goodison S: Elevated osteopontin and thrombospondin expression identifies malignant human breast carcinoma but is not indicative of metastatic status. Breast Cancer Res 5:R136–143 (2003). Watson MB, Bahia H, Ashman JN, Berrieman HK, Drew P, et al: Chromosomal alterations in breast cancer revealed by multicolour fluorescence in situ hybridization. Int J Oncol 25:277– 283 (2004). Weigelt B, Glas AM, Wessels LF, Witteveen AT, Peterse JL, van ‘t Veer LJ: Gene expression profiles of primary breast tumors maintained in distant metastases. Proc Natl Acad Sci USA 100:15901– 15905 (2003). Weigelt B, Hu Z, He X, Livasy C, Carey LA, et al: Molecular portraits and 70-gene prognosis signature are preserved throughout the metastatic process of breast cancer. Cancer Res 65: 9155– 9158 (2005a).
Weigelt B, Wessels LF, Bosma AJ, Glas AM, Nuyten DS, et al: No common denominator for breast cancer lymph node metastasis. Br J Cancer 93: 924–932 (2005b). Wong CM, Lee JM, Ching YP, Jin DY, Ng IO: Genetic and epigenetic alterations of DLC-1 gene in hepatocellular carcinoma. Cancer Res 63: 7646–7651 (2003). Xie D, Jauch A, Miller CW, Bartram CR, Koeffler HP: Discovery of over-expressed genes and genetic alterations in breast cancer cells using a combination of suppression subtractive hybridization, multiplex fish and comparative genomic hybridization. Int J Oncol 21: 499–507 (2002). Yang F, Foekens JA, Yu J, Sieuwerts AM, Timmermans M, et al: Laser microdissection and microarray analysis of breast tumors reveal ERalpha related genes and pathways. Oncogene 25:1413–1419 (2006). Yao J, Weremowicz S, Feng B, Gentleman RC, Marks JR, et al: Combined cDNA array comparative genomic hybridization and serial analysis of gene expression analysis of breast tumor progression. Cancer Res 66: 4065–4078 (2006). Zannetti A, Del Vecchio S, Carriero MV, Fonti R, Franco P, et al: Coordinate up-regulation of Sp1 DNA-binding activity and urokinase receptor expression in breast carcinoma. Cancer Res 60: 1546–1551 (2000).
Cytogenet Genome Res 118:116–129 (2007)
129
GEJ adenocarcinomas were characterized by several comparative genomic hybridization studies. Gains were detected most frequently at 8q and 20q, and less frequently at 1q, 3q, 6p, 7pq, 15q and 17q. Recurrent losses were found at 18q and also, less frequently, at 4pq, 5q, 9p, 14q and 17p. 8q gain was found in 47–79% of GEJ adenocarcinomas, with high-level amplifications occurring in 3–32%. The minimal overlapping region for gain at 8q was 8q23.3]q24.22 (Moskaluk et al., 1998; van Dekken et al., 1999, 2001, 2006; Walch et al., 2000; El-Rifai et al., 2001; Riegman et al., 2001; Stocks et al., 2001). Gain of 8q has been shown to be a progression marker in other cancer types, such as prostate and breast adenocarcinomas (Isola et al., 1995; Alers et al., 2000). Furthermore, gain of 8q has also been detected in precursor stages of GEJ adenocarcinoma, suggesting that this genetic aberration may also play a role in progressive behavior of this type of cancer (Walch et al., 2000; Riegman et al., 2001; van Dekken et al., 2001). In GEJ cancer, as in other types of cancer, copy number increase of 8q24 has often been related to the oncogene MYC. It has, however, been proposed that other target genes may exist on 8q (Cher et al., 1996; Garnis et al., 2004; van Duin et al., 2005a). To address the question whether genomic regions in addition to the MYC region are involved, array comparative genomic hybridization (aCGH) was performed using a tile-path array specific for the whole long arm of chromosome 8 (van Duin et al., 2005b). To this end, 37 GEJ cancer specimens were analysed, including 22 primary cancer samples, 13 cell lines and two xenografts. Subsequently, candidate genes were selected of the most frequently gained regions, and real-time RT-PCR was performed to study the role of these genes in GEJ adenocarcinomas. Materials and methods Tissue specimens GEJ adenocarcinoma specimens were obtained from 22 patients (Barrett’s adenocarcinoma, n = 11; gastric cardia adenocarcinoma, n = 11), of which 19 were male and three female (mean patient age 66, range 50–85 years). For fourteen cases, DNA samples were obtained from fresh-frozen tissue; for the remaining eight cases formalin-fixed, paraffin-embedded tissue was used. In addition, cell lines OEC19, OEC33, BE-3, BIC-1, ESO-26, ESO-51, FLO-1, KYAE-1, P4, SEG-1, OACM5.1C, SK-GT-4 and SK-GT-5, and xenografts OACM2.1X and OACM4.1X were included (Shimada et al., 1992; Altorki et al., 1993; Hughes et al., 1997; Rockett et al., 1997; de Both et al., 2001; Stiles et al., 2003). The cell lines and xenografts were derived from gastric cardia and oesophageal adenocarcinoma in five and ten cases, respectively. As a reference for RT-PCR, nine normal stomach samples were available. Arrays Genomic target DNA was isolated from bacterial cultures and arrayed as described previously (Watson et al., 2004; van Duin et al., 2005b). At the time of construction, clones and clone positions were derived from the UCSC genome browser June 2002 freeze. For this study, an update based on the May 2004 freeze was used. This resulted in the exclusion of 86 clones, which were not present in the new freeze. Hybridization to microarrays Isolation of DNA from the formalin-fixed, paraffin-embedded tumor material was performed as described previously (van Dekken et al., 2006), using the Puregene DNA isolation kit (Gentra systems, Min-
neapolis, MN). For DNA isolation from the fresh-frozen tissue, the DNA extraction kit (Qiagen, Valencia, CA, USA) was used. 400 ng of test DNA (xenograft or tumor) and reference genomic DNA was labeled with Cy3 and Cy5 (Amersham Pharmacia Biotech, Piscataway, NJ, USA), respectively, according to a modified random priming protocol from the Bio-Prime labeling system (Invitrogen, Carlsbad, CA, USA). Briefly, random primer (random DNA octamers) was added to the DNA to a final concentration of 300 ng/l. After denaturation for 10 min at 100 ° C, the sample was put on ice and dNTPs (dATP, dGTP and dCTP: final concentration 200 M; dTTP: final concentration 50 M), cyanine dye-labeled dUTP (either Cy3-dUTP or Cy5-dUTP: final concentration 40 M) and 40 units of Klenow fragment were added to a final volume of 25 l. The reaction mixture was incubated overnight at 37 ° C. Unincorporated nucleotides were removed using microspin columns according to the recommendations of the manufacturer (Amersham Pharmacia Biotech, Piscataway, NJ, USA). Reference and test DNA for each array were then pooled. The mixture of labelled reference and test DNA was precipitated in the presence of 60 g COT human DNA (Roche, Basel, Switzerland) and resuspended in 50 l hybridization solution containing 50% formamide, 10% dextran sulfate, 2! SSC, 4% SDS and 10 g/l yeast tRNA. The probes were denatured for 10 min at 74 ° C, upon which pre-annealing of the COT human DNA took place for 60 min at 37 ° C. Preparation of the slides consisted of cross-linking the slides (GS Gene Linker UV chamber, BioRad laboratories, Hercules, CA; 260 mJ) and placing a dam of rubber cement around each array, at least 60 min before use to allow the rubber cement to set. The hybridization mixture (50 l) was added to the array. A rubber gasket and a glass microscope slide fastened to the slide provided an enclosed chamber for the hybridization. Hybridization was carried out for 48 h at 37 ° C on a unidirectional tilting platform (8 3 rpm) placed within an incubator (Robbins Scientific, Sunnyvale, CA, USA). After hybridization, slides were washed for 15 min in 50% formamide, 2! SSC, pH 7.0 at 50 ° C, and then for an additional 20 min in 2! SSC, 0.1% SDS at 50 ° C. Finally, slides were washed in PN buffer (0.1 M sodium phosphate buffer, 0.1% NP40, pH 8.0) for 10 min at room temperature and rubber cement dams were removed.
Image and data analysis A ScanArray Express HT (Perkin Elmer Life Sciences, Boston, MA, USA) was used to collect 16-bit TIF images through Cy3 and Cy5 filter sets. Images were analysed using custom software as described (Snijders et al., 2001; Jain et al., 2002). Thus, a ratio of Cy3 and Cy5 intensities, averaged for the quadruplicate spots, was obtained per clone. Clones were excluded for analysis if the standard deviation of the replicate spots exceeded 0.2 or if none or a single spot had contributed to the mean of the replicate spots. For analysis and presentation log2 values of fluorescence ratios were used. Two types of analysis were used for this data set. Firstly, for depiction of gains and losses, aberrant log2 values were defined by an algorithm using flexible thresholds based on the standard deviation (SD) of the data sets of the specimens. SD’s over windows of five consecutive BACs were averaged, sliding along chromosome arm 8q one BAC at a time. The resulting SD was multiplied by + or –2.5 to create the threshold value for gains and losses, respectively. The factor of 2.5 was chosen on the basis of prior experience (van Duin et al., 2005a). Samples with less than 20 altered clones gained or lost were excluded from Fig. 3. A high-level amplification was defined as a distinct peak with a log2 fluorescence ratio 11. Secondly, to identify common regions of amplification the average was calculated for each clone for the full set of gastroesophageal specimens tested (n = 37). Subsequently, a simple smoothing procedure was applied by calculating the average over windows of nine consecutive BACs, sliding along 8q one BAC at a time. Quantitative RT-PCR RNA was isolated using the RNAbee reagent (Campro Scientific, Veenendaal, The Netherlands). Assays-On-Demand were obtained from Applied Biosystems. cDNA was generated using oligo(dT) primers and random hexamer primers (Amersham Pharmacia Biotech, Piscataway, NJ, USA). The primers were annealed to 5–10 g of
Cytogenet Genome Res 118:130–137 (2007)
131
Table 1. Genes within three regions of frequent amplification in a series of gastroesophageal specimens and EXT1 found within a specific amplification of cell line SK-GT-5
Region
Chromosome band
Genomic distance
Gene
I
8q24.13 8q24.13
125 Mb 126 Mb
ANXA13 MTSS1
Annexin A13 isoform b Missing in metastasis (MIM)
II
8q24.21 8q24.21
128 Mb 129 Mb
FAM84B MYC
Breast cancer membrane protein 101 Myelocytomatosis viral oncogene
III
8q24.3 8q24.3 8q24.11
141 Mb 142 Mb 119 Mb
C8orf17 PTK2 EXT1
Molt-4 sequence tag-1 Focal adhesion kinase 1 Exostosin 1
total RNA in a total volume of 25 l by incubation at 65 ° C for 5 min and cooling down to 0 ° C. Reverse transcription was performed by the addition of 20 l RT mix such that the final concentration was 1! buffer, 0.5 mM dNTP, 2.5 mM DTT, and 200 U Superscript II (Invitrogen, Carlsbad, CA, USA). The mixture was incubated at 25 ° C, subsequently incubated at 42 ° C for 50 min, heated to 70 ° C for 15 min, and cooled to 4 ° C. Assays were carried out according to the instructions of the manufacturer using an ABI PRISM 7700 Sequence Detection System (Applied Biosystems, Foster City, CA, USA). Briefly, PCR was performed using the Taqman Universal PCR master mix, which contains AmpliTaq Gold DNA polymerase. The final PCR mixture contained 1! master mix, 5 l of 20! diluted cDNA and assay mix, containing unlabeled primers (900 nM final reaction concentration) and Taqman probe (250 nM final reaction concentration) in a volume of 25 l. The HPRT, HMBS , GAPDH and B2M genes were used to generate a normalization factor for each sample using the Genorm software (Vandesompele et al., 2002). Relative quantitation of gene expression was done according to recommendations of ABI, using the standard curve method based on serial dilution of a pool of gastroesophageal cell lines, i.e. SK-GT-4, FLO-1, SEG-1 and OEC33.
Fluorescent in situ hybridisation (FISH) FISH of fresh cell preparations was accomplished as described by us before (van Duin et al., 2005b). DNA probes for bicolor FISH of the chromosome 8 centromere and 8q BAC DNA clones were labeled with biotin and digoxigenin, respectively, using a Nick Translation Reagent Kit (Vysis, Downers Grove IL, USA). The biotin-labeled probes were detected with FITC-conjugated avidin (Vector Laboratories, Burlingame CA, USA), the digoxigenin-labeled probes were visualized with rhodamin-conjugated, anti-digoxigenin antibody (Roche, Basel, Switzerland). For validation purposes, the following 8q BAC clones were chosen: RP11-728K22 (EXT1), RP11-580G9 (ANXA13), RP11-746D17 (MTSS1), RP11-177G8 (FAM84B), RP11-440N18 (MYC), RP11-910I20 (C8orf17) and RP11-1126O23 (PTK2). Briefly, after 48 h hybridization the cells were counterstained with DAPI in antifade solution (Vectashield; Vector, Burlingame CA, USA). Two investigators scored up to 100 interphase cells per specimen for the 8q BAC probe signals. Images of each of the three fluorochromes were collected using an epifluorescence microscope (Leica DM, Rijswijk, The Netherlands) equipped with appropriate excitation and emission filter sets (Leica), and a cooled CCD camera (Photometrics, Tucson AZ, USA). The green, red and blue images were collected sequentially by changing the excitation filter using CW4000 FISH software (Leica). Statistics Expression levels in normal and cancer specimens were compared using the Mann-Whitney U test. The average correlation of FISH with aCGH over the seven genomic locations was calculated (Pearson correlation coefficient, r) and the significance of the correlation was assessed using mixed model ANOVA (SAS PROC MIXED). A P value of !0.05 was considered significant.
132
Cytogenet Genome Res 118:130–137 (2007)
Results
Genomic array analysis We have genomically analyzed a series of 37 GEJ adenocarcinoma specimens, of which 21 originated from the esophagus and 16 from the gastric cardia. An 8q contig array was employed to define the region(s) of common gain. Previously, samples were shown to have 8q gain by chromosomal CGH (n = 15), genome wide aCGH (n = 16) or MYC amplification as demonstrated by FISH analysis (n = 6). Using the 8q contig array, 8q gain was confirmed in all these samples with substantially higher resolution. Single copy gain of the whole arm was found in three of 37 cases (8%). Partial single copy gain was found in essentially all cases, i.e. in 36 out of 37 (97%). Amplifications, defined by a clear, sharp increase in copy number with a log2 value of more than 1, were detected in 11 cases (30%). Furthermore, deletions were found in ten cases (27%). Specific MYC amplification was found in ten cases (27%). Representative examples are shown in Figs. 1 and 2. To determine common regions of genomic gain, a cumulative 8q profile of all 37 GEJ adenocarcinoma specimens was constructed (Fig. 3). Using this analysis, three commonly overrepresented regions were identified. The most distinct region of amplification encompasses the MYC gene at 8q24.21 (127–128 Mb), indicated as region II in Fig. 3. Regions of frequent gain were further found proximal to MYC at 124–125 Mb (8q24.13) and distal to MYC at 141–142 Mb (8q24.3). mRNA expression analysis Seven known genes were chosen from the UCSC Genome Browser, on the basis of relevance to cancer for expression analysis by RT-PCR (Table 1). In region II, in addition to MYC, the gene FAM84B was included. Known genes in the regions I and III include ANXA13 (125 Mb; 8q24.13), MTSS1 (126 Mb; 8q24.13), C8orf17 (141 Mb; 8q24.3) and PTK2 (142 Mb; 8q24.3). In addition to these genes, EXT1 (119 Mb; 8q24.11) was included since it was found in a specific amplification in both cell line SK-GT-5 (Fig. 2C) and a primary adenocarcinoma (data not shown). For all of these gene locations, FISH was performed to confirm aCGH results (Fig. 4). A significant relationship was found between aCGH and FISH (average Pearson correlation factor = 0.66; p = 0.02). Quantitative RT-PCR analysis of these seven genes
Fluorescence ratio (log2)
2
A
1
0
–1 46
56
66
76
86
96
106
116
126
136
146
56
66
76
86
96
106
116
126
136
146
56
66
76
86
96
106
116
126
136
146
B Fluorescence ratio (log2)
2
1
0
–1 46
C Fluorescence ratio (log2)
3 2 1 0 –1 46
24.3
24.23
24.22
24.21
24.13
24.11 24.12
23.3
23.2
23.1
22.3
22.2
22.1
21.3
21.2
21.13
21.12
21.11
13.3
13.2
13.1
12.3
12.2
12.1
11.21 11.22 11.23
11.1
Genomic distance (Mb)
Fig. 1. 8q aCGH profiles of primary GEJ adenocarcinoma specimens. The genomic distance is plotted on the X-axis, whereas the Y-axis shows the log2 fluorescence ratio of test and reference DNA. (A) Log2 ratios of a paraffin-embedded cardia adenocarcinoma sample showing a single copy gain from 8q24.12]qter (120–146 Mb). (B) Log2 ratios of a fresh-frozen cardia cancer specimen demonstrating a complex pattern of single copy gains (8q11.21]q21.12 (52–79 Mb), 8q21.13 (83–84 Mb) and 8q24.13]q24.22 (123–132 Mb)). Two high-level am-
plifications were found at 8q24.21 (129 Mb and 130 Mb). In addition, a deletion was detected at 97 Mb (8q22.1). (C) Log2 ratios of a fresh-frozen Barrett’s adenocarcinoma sample showing various single copy gains between 8q21.3 and 8q24.3 (89–146 Mb) and four high-level amplifications, i.e. 8q21.13 (82 Mb), 8q24.13 (125 Mb), 8q24.21 (129 Mb) and 8q24.22 (132 Mb). A chromosomal ideogram of 8q can be found at the bottom of the figure.
was performed on a panel of 24 gastroesophageal samples, including 13 cell lines, two xenografts and nine normal stomach controls. The mRNA expression data are shown in Fig. 5. For two of the seven genes, MYC (8q24.21) and EXT1 (8q24.11), a significantly increased expression was found in
GEJ adenocarcinoma cell lines and xenografts compared to normal controls (both P = 0.02). Expression of the genes MTSS1 (P ! 0.001; 8q24.13), FAM84B (P = 0.02; 8q24.21) and C8orf17 (P ! 0.001; 8q24.3) was found to be significantly decreased in this set of cell lines and xenografts.
Cytogenet Genome Res 118:130–137 (2007)
133
Fluorescence ratio (log2)
2
A
1
0
–1 46
56
66
76
86
96
106
116
126
136
146
56
66
76
86
96
106
116
126
136
146
56
66
76
86
96
106
116
126
136
146
B Fluorescence ratio (log2)
2
1
0
–1 46
C Fluorescence ratio (log2)
5 4 3 2 1 0 –1 46
Fig. 2. 8q aCGH profile of GEJ adenocarcinoma cell lines. The genomic distance is plotted on the X-axis, whereas the Y-axis shows the log2 fluorescence ratio of test and reference DNA. (A) Log2 ratios of cell line FLO-1 showing distal single copy gain (8q24.21]qter; 128–146 Mb); breakpoint is separated by approximately 700 kb from MYC.
Discussion
In this study, three regions of frequent gain on distal chromosome 8q in GEJ adenocarcinoma were found. Conventional chromosomal CGH of GEJ adenocarcinomas revealed a minimal overlapping region of gain on 8q23.3] q24.22 (van Dekken et al., 1999; Walch et al., 2000; Riegman et al., 2001). In addition, a study of gastric cardia adeno-
134
Cytogenet Genome Res 118:130–137 (2007)
24.3
24.23
24.22
24.21
24.13
24.11 24.12
23.3
23.2
23.1
22.3
22.2
22.1
21.3
21.2
21.13
21.12
21.11
13.3
13.2
13.1
12.3
12.2
12.1
11.21 11.22 11.23
11.1
Genomic distance (Mb)
(B) Log2 ratios of xenograft OACM2.1X showing genomic loss at 8q21.11]q21.12 (77–79 Mb), single copy gains at 8q21.2]q21.3 (86–88 Mb) and at 8q22.1]qter (95–146 Mb). (C) Log2 ratios of cell line SKGT-5. Two high-level amplifications were found, at 8q24.12 (119 Mb; EXT1) and at 8q24.21 (129 Mb; MYC).
carcinomas demonstrated a frequently gained region on 8q21.2]q24.13 (van Dekken et al., 2001). In our study, these amplified regions are worked out in more detail and illustrate the increased resolution of aCGH compared to chromosomal CGH. Moreover, separate cumulative plots of gastric cardia and esophageal adenocarcinoma specimens did not show significant differences of the region between 8q24.13 and 24.3 (data not shown).
0.75 I
II
III
0.5
0.25
0
–0.25 146
24.3
24.23
24.22
136
24.21
126
24.13
116
23.1 23.2 23.3 24.11 24.12
22.2 22.3
86 96 106 Genomic distance (Mb)
22.1
76
21.3
66
21.11 21.12 21.13 21.2
56
12.1 12.2 12.3 13.1 13.2 13.3
46
11.1 11.21 11.22 11.23
Fig. 3. High-resolution definition of amplified regions by average log2 ratios of the GEJ adenocarcinoma samples and cell lines. The mean fluorescence ratio for each BAC clone calculated over 37 samples is shown. Frequently amplified regions are indicated by the Roman numerals I to III. The first region is positioned at 8q24.13 (124–127 Mb), the second region at 8q24.21]q22 (128–132 Mb) and the third region at 8q24.3 (142–146 Mb). A chromosomal ideogram of 8q can be found at the bottom of the figure.
Fig. 4. Examples of bicolor FISH with a centromeric chromosome 8-specific DNA probe (green) and 8q-specific BAC DNA clones (red) to interphase nuclei of GEJ adenocarcinoma cell lines. (A) FISH with BAC clone RP11-440N18 (8q24.21; 129 Mb) to FLO-1 interphase cells illustrating single copy gain of 8q. In most cell nuclei five red BAC-related spots and four green centromeric spots are seen. (B) FISH with BAC clone RP11-1126O23 (8q24.3; 142 Mb) to OACM2.1X interphase cells. Two extra BAC-related red spots are seen relative to the green centromeric spots. (C) High-level amplification of SKGT-5 at 8q24.21 (129 Mb), visualized with BAC clone RP11-440N18. Clusters of red FISH signal represent highlevel amplification.
Copy number increase of MYC was found frequently in this study. Specific MYC amplification was detected in ten cases. Consequently, MYC was found to be overexpressed in GEJ adenocarcinoma cell lines and xenografts. Although there was no clear correlation between increased MYC expression and copy number in all cases, the amplification in SK-GT-5 did correspond to the highly increased mRNA ex-
pression, i.e. 30-fold higher than the median level in normal controls (data not shown). In line with our results, both amplification and overexpression of MYC have been reported to occur in Barrett’s adenocarcinoma and in precursor lesions of this disease (Walch et al., 2000; Tselepis et al., 2003). mRNA expression analyses of region II also included FAM84B, which was found to be downregulated in this
Cytogenet Genome Res 118:130–137 (2007)
135
Fig. 5. Quantitative expression analysis of candidate genes. Relative expression levels are shown for each candidate gene in normal (left of each panel) and cancer specimens (right of each panel). Gene names are shown below the panel. Asterisks denote significantly increased expression, i.e. overexpression of EXT1 and MYC.
Relative expression levels
100,000
10,000
*
*
1,000
100
10
1
study. FAM84B (alias BCMP101) was identified in a proteomics study of breast cancer membranes, and, in contrast to our findings, found to be overexpressed in this type of cancer (Adam et al., 2003). Tissue-specific expression patterns may underlie this difference. Two other regions of genomic overrepresentation were found, although none of the genes within these regions were found to be upregulated. In contrast, significant downregulation was found for MTSS1 and C8orf17. MTSS1, also known as missing in metastasis (MIM), was identified by differential display using bladder cell lines and was found not to be expressed in metastatic cell lines (Lee et al., 2002). In the present study, only metastatic cell line OACM5.1C demonstrated absence of MTSS1 expression. The MTSS1 gene product is an actin-binding protein and has been suggested to be involved in control of cellular proliferation (Lee et al., 2002; Woodings et al., 2003). No putative function has been described for C8orf17, but overexpression of this gene has been reported in prostate and breast cancer (Tan et al., 2003). EXT1 was found to be overexpressed in this set of GEJ cancer specimens as compared to normal controls. The gene product of EXT1 is a glycosyltransferase, involved in heparan sulfate proteoglycan (HSPG) biosynthesis. HSPGs are involved in several aspects of cancer biology, including
EXT1
ANXA13
MTSS1
FAM84B
MYC
C8orf17
PTK2
tumorigenesis, tumor progression and metastasis (Sasisekharan et al., 2002). Moreover, mutations of the EXT1) gene have been shown to be the cause of the disease of multiple osteochrondromas (Ahn et al., 1995; Wells et al., 1997). This suggests that EXT1 is a tumor suppressor at least in some types of cancer. However, EXT1 was found to be overexpressed in cancer of the salivary gland and head and neck squamous cell carcinoma (www.oncomine.org; Frierson et al., 2002; Cromer et al., 2004; Rhodes et al., 2004). In conclusion, we have evaluated chromosomal and expression patterns of chromosome 8q, a genomic region often described to be amplified in progressive cancer. We found that besides the oncogene MYC also other genes, such as EXT1, contribute to this genetic phenomenon. These findings do not only enlarge our understanding of biological mechanisms underlying GEJ adenocarcinomas. They might also prove useful in the surveillance of patients with Barrett’s esophagus, since 8q amplification has also been detected in pre-neoplastic lesions. Acknowledgements We would like to acknowledge L. Roque and P. Chaves for the possibility to include cell lines ESO-26 and ESO-51 in our study. We further thank A. Sieuwerts for excellent advice.
References Adam PJ, Boyd R, Tyson KL, Fletcher GC, Stamps A, et al: Comprehensive proteomic analysis of breast cancer cell membranes reveals unique proteins with potential roles in clinical cancer. J Biol Chem 278: 6482–6489 (2003). Ahn J, Ludecke HJ, Lindow S, Horton WA, Lee B, et al: Cloning of the putative tumour suppressor gene for hereditary multiple exostoses (EXT1). Nat Genet 11:137–143 (1995). Alers JC, Rochat J, Krijtenburg PJ, Hop WC, Kranse R, et al: Identification of genetic markers for prostatic cancer progression. Lab Invest 80: 931–942 (2000).
136
Altorki N, Schwartz GK, Blundell M, Davis BM, Kelsen DP, Albino AP: Characterization of cell lines established from human gastric-esophageal adenocarcinomas. Biologic phenotype and invasion potential. Cancer 72: 649–657 (1993). Botterweck AA, Schouten LJ, Volovics A, Dorant E, van Den Brandt PA: Trends in incidence of adenocarcinoma of the oesophagus and gastric cardia in ten European countries. Int J Epidemiol 29:645–654 (2000). Cameron AJ, Ott BJ, Payne WS: The incidence of adenocarcinoma in columnar-lined (Barrett’s) esophagus. N Engl J Med 313: 857–859 (1985).
Cytogenet Genome Res 118:130–137 (2007)
Cher ML, Bova GS, Moore DH, Small EJ, Carroll PR, et al: Genetic alterations in untreated metastases and androgen-independent prostate cancer detected by comparative genomic hybridization and allelotyping. Cancer Res 56: 3091–3102 (1996). Cromer A, Carles A, Millon R, Ganguli G, Chalmel F, et al: Identification of genes associated with tumorigenesis and metastatic potential of hypopharyngeal cancer by microarray analysis. Oncogene 23:2484–2498 (2004).
de Both NJ, Wijnhoven BP, Sleddens HF, Tilanus HW, Dinjens WN: Establishment of cell lines from adenocarcinomas of the esophagus and gastric cardia growing in vivo and in vitro. Virchows Arch 438: 451–456 (2001). El-Rifai W, Frierson HF Jr, Moskaluk CA, Harper JC, Petroni GR, et al: Genetic differences between adenocarcinomas arising in Barrett’s esophagus and gastric mucosa. Gastroenterology 121:592–598 (2001). Frierson HF Jr, El-Naggar AK, Welsh JB, Sapinoso LM, Su AI, et al: Large scale molecular analysis identifies genes with altered expression in salivary adenoid cystic carcinoma. Am J Pathol 161:1315–1323 (2002). Garnis C, Coe BP, Ishkanian A, Zhang L, Rosin MP, Lam WL: Novel regions of amplification on 8q distinct from the MYC locus and frequently altered in oral dysplasia and cancer. Genes Chromosomes Cancer 39: 93–98 (2004). Hughes SJ, Nambu Y, Soldes OS, Hamstra D, Rehemtulla A, et al: Fas/APO-1 (CD95) is not translocated to the cell membrane in esophageal adenocarcinoma. Cancer Res 57:5571–5578 (1997). Isola JJ, Kallioniemi OP, Chu LW, Fuqua SA, Hilsenbeck SG, et al: Genetic aberrations detected by comparative genomic hybridization predict outcome in node-negative breast cancer. Am J Pathol 147: 905–911 (1995). Jain AN, Tokuyasu TA, Snijders AM, Segraves R, Albertson DG, Pinkel D: Fully automatic quantification of microarray image data. Genome Res 12:325–332 (2002). Lee YG, Macoska JA, Korenchuk S, Pienta KJ: MIM, a potential metastasis suppressor gene in bladder cancer. Neoplasia 4: 291–294 (2002). Menke-Pluymers MB, Schoute NW, Mulder AH, Hop WC, van Blankenstein M, Tilanus HW: Outcome of surgical treatment of adenocarcinoma in Barrett’s oesophagus. Gut 33: 1454– 1458 (1992). Morales TG, Sampliner RE, Bhattacharyya A: Intestinal metaplasia of the gastric cardia. Am J Gastroenterol 92:414–418 (1997). Moskaluk CA, Hu J, Perlman EJ: Comparative genomic hybridization of esophageal and gastroesophageal adenocarcinomas shows consensus areas of DNA gain and loss. Genes Chromosomes Cancer 22: 305–311 (1998). Rhodes DR, Yu J, Shanker K, Deshpande N, Varambally R, et al: ONCOMINE: a cancer microarray database and integrated data-mining platform. Neoplasia 6: 1–6 (2004).
Riegman PH, Vissers KJ, Alers JC, Geelen E, Hop WC, et al: Genomic alterations in malignant transformation of Barrett’s esophagus. Cancer Res 61:3164–3170 (2001). Rockett JC, Larkin K, Darnton SJ, Morris AG, Matthews HR: Five newly established oesophageal carcinoma cell lines: phenotypic and immunological characterization. Br J Cancer 75: 258– 263 (1997). Sasisekharan R, Shriver Z, Venkataraman G, Narayanasami U: Roles of heparan-sulphate glycosaminoglycans in cancer. Nat Rev Cancer 2: 521–528 (2002). Shimada Y, Imamura M, Wagata T, Yamaguchi N, Tobe T: Characterization of 21 newly established esophageal cancer cell lines. Cancer 69: 277–284 (1992). Snijders AM, Nowak N, Segraves R, Blackwood S, Brown N, et al: Assembly of microarrays for genome-wide measurement of DNA copy number. Nat Genet 29: 263–264 (2001). Spechler SJ, Zeroogian JM, Antonioli DA, Wang HH, Goyal RK: Prevalence of metaplasia at the gastro-oesophageal junction. Lancet 344:1533– 1536 (1994). Stiles BM, Bhargava A, Adusumilli PS, Stanziale SF, Kim TH, et al: The replication-competent oncolytic herpes simplex mutant virus NV1066 is effective in the treatment of esophageal cancer. Surgery 134:357–364 (2003). Stocks SC, Pratt N, Sales M, Johnston DA, Thompson AM, et al: Chromosomal imbalances in gastric and esophageal adenocarcinoma: specific comparative genomic hybridization-detected abnormalities segregate with junctional adenocarcinomas. Genes Chromosomes Cancer 32:50–58 (2001). Tan JM, Tock EP, Chow VT: The novel human C8ORF17 (C8orf17 ) gene exhibits tissue specific expression, maps to chromosome 8q24.2, and is overexpressed/amplified in high grade cancers of the breast and prostate. Mol Pathol 56: 109–115 (2003). Tselepis C, Morris CD, Wakelin D, Hardy R, Perry I, et al: Upregulation of the oncogene c-myc in Barrett’s adenocarcinoma: induction of c-myc by acidified bile acid in vitro. Gut 52: 174–180 (2003). van Dekken H, Geelen E, Dinjens WN, Wijnhoven BP, Tilanus HW, et al: Comparative genomic hybridization of cancer of the gastroesophageal junction: deletion of 14q31]q32.1 discriminates between esophageal (Barrett’s) and gastric cardia adenocarcinomas. Cancer Res 59: 748–752 (1999).
van Dekken H, Alers JC, Riegman PH, Rosenberg C, Tilanus HW, Vissers K: Molecular cytogenetic evaluation of gastric cardia adenocarcinoma and precursor lesions. Am J Pathol 158: 1961–1967 (2001). van Dekken H, Wink JC, Vissers KJ, van Marion R, Koppert LB, et al: Genomic analysis of early adenocarcinoma of the esophagus or gastroesophageal junction: Tumor progression is associated with alteration of 1q and 8p sequences. Genes Chromosomes Cancer 45: 516–525 (2006). van Duin M, van Marion R, Vissers K, Watson JE, van Weerden WM, et al: High-resolution array comparative genomic hybridization of chromosome arm 8q: evaluation of genetic progression markers for prostate cancer. Genes Chromosomes Cancer 44: 438–449 (2005a). van Duin M, van Marion R, Watson JE, Paris PL, Lapuk A, et al: Construction and application of a full-coverage, high-resolution, human chromosome 8q genomic microarray for comparative genomic hybridization. Cytometry A 63: 10–19 (2005b). Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, et al: Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 3:RESEARCH0034 (2002). Walch AK, Zitzelsberger HF, Bink K, Hutzler P, Bruch J, et al: Molecular genetic changes in metastatic primary Barrett’s adenocarcinoma and related lymph node metastases: comparison with nonmetastatic Barrett’s adenocarcinoma. Mod Pathol 13: 814–824 (2000). Watson JE, Doggett NA, Albertson DG, Andaya A, Chinnaiyan A, et al: Integration of high-resolution array comparative genomic hybridization analysis of chromosome 16q with expression array data refines common regions of loss at 16q23]qter and identifies underlying candidate tumor suppressor genes in prostate cancer. Oncogene 23:3487–3494 (2004). Wells DE, Hill A, Lin X, Ahn J, Brown N, Wagner MJ: Identification of novel mutations in the human EXT1 tumor suppressor gene. Hum Genet 99:612–615 (1997). Woodings JA, Sharp SJ, Machesky LM: MIM-B, a putative metastasis suppressor protein, binds to actin and to protein tyrosine phosphatase delta. Biochem J 371: 463–471 (2003).
Cytogenet Genome Res 118:130–137 (2007)
137
t(3;12) 2
1
7
6
3
8
4
9
5
10
11
12
t(3;12)
13
19
14
20
15
16
21
17
22
18
X
Y
A
t(3;12) B
Fig. 2. (A , B) G-banded karyotype of lipoma showing t(3;12)(q28;q15).
Cytogenetics. Nearly 60% of lipomas have been reported to have clonal chromosomal aberrations of which translocations and rearrangements affecting 12q13]q15 are the most common (Sandberg and Bridge, 1994). The interacting partner chromosomes to this region are several and include 1p36, 1p34]p32, 2p24]p22, 5q33, 2q35]q37, 5q33, 11q13, 12p12]p11, 12q24, 13q12]q14, 17q23]q25 and 21q21] q22. However, the most common translocation partner is 3q leading to a t(3; 12)(q27–q28;q13–q15) (Fig. 2A, B) seen in almost 25% of lipomas (Rubin and Dal Cin, 2001). About 15–20% of lipomas show abnormalities of chromosome 13q in the form of breakpoints and deletions at the 13q12]q22 region (Dahlen et al., 2003). Dahlen et al. confirmed a nonrandom distribution of 13q rearrangements in 27 out of 62 adipocytic tumors by fluorescent in situ hybridization (FISH). The results of this FISH study indicated that chromosome 13 is involved in a variety of rearrangements and deletions that cover a limited segment (⬃2.5 Mb) of chromosome band 13q14, distal to the RB1 gene. Rearrangements and abnormalities of 6p23]p21, and changes in chromosome 7 are seen in about 20% of lipomas. Overlapping changes involving 6p23]p21 and 13q deletion can also be seen in some tumors; however, the majority of the time these are seen as sole abnormalities. 6p23]p21 abnormalities are usually balanced translocations, with 3q27]q28 being the recurrent translocation partner (Fletcher et al., 2002). It is to be borne in mind that 12q13]q15 rearrangements are also seen in other tumors such as uterine leio-
myoma, pulmonary hamartoma, pleomorphic adenoma of salivary glands and hemangiopericytomas (Nilbert and Heim, 1990; Mandahl et al., 1993; Fletcher et al., 1995; Wanschura et al., 1996). Molecular genetics. Lipomas associated with 12q13]q15 aberrations show rearrangement of high mobility group protein gene HMGA2 (previously HMGIC) (Schoenmakers et al., 1995). The HMGI family of proteins alters gene expression globally by changing conformation of large chromosomal segments leading to activation/repression of transcription of various genes involved in cell proliferation (Hess, 1998; Tallini and Dal Cin, 1999). HMGA2 is a transcription factor containing three DNA binding AT hooks encoded by exons 1–3 and is expressed primarily during embryonic development (Borrmann et al., 2003). In lipomas, this chromosomal region can recombine with many different chromosome bands, on all chromosomes except Y (http://cgap.nci.nih.gov/Chromosomes/Mitelman, 2006). Some translocations have resulted in chimeric genes. Identified gene fusion partners include LPP (lipoma preferred partner gene, 3q27]q28), LHFP (Lipoma HMGIC Fusion Partner 13q12) and CXCR7 (CMKORI) (2q37) (Petit et al., 1996, 1999; Borberg et al., 2002). In addition, either truncation or transcriptional activation of HMGA2 or creation of HMGA2/EBF fusion transcript has been reported in a series of eight lipomas with rearrangements of 5q32]q33 and 12q14]q15 reflecting the molecular heterogeneity of these tumors (Nilsson et al., 2006). Lipomas characterized by re-
Cytogenet Genome Res 118:138–147 (2007)
139
arrangements of 6p22]p21 involve another HMGI family, HMGA1 (HMGIY) gene. The rearrangements involving HMGA1 are located outside the coding region (usually downstream) in contrast to the HMGA2 rearrangements on chromosome 12q13]q15 (Kazmierczak et al., 1998). Both HMGA2 and HMGA1 have been detected by immunohistochemical analysis in lipomas and other benign mesenchymal tumors such as pulmonary hamartomas, uterine leiomyomas and endometrial polyps which are known to show similar cytogenetic aberrations (Tallini et al., 2000). Some lipomas with 13q12]q14 deletions or break points show translocations involving LHFP gene at 13q12 and HMGA2 at 12q13 as mentioned above. Another HMG family gene, HMGB1 (HMG1), has been located at the 13q12 region; however, no rearrangements in this gene were identified in a series of eight lipomas (Kazmierczak et al., 1999). Hibernoma Hibernoma is a rare benign adipose tumor (1.6%) composed of brown fat occurring predominantly in young adults and in a wide variety of locations including thigh, trunk, upper extremity and head and neck. The tumors are usually subcutaneous, and can also be seen in deeper locations such as intramuscular areas. Histologically, they contain multivacuolated brown fat cells and small capillary proliferations with variations including admixture of mature adipocytes (Fletcher et al., 2002). Cytogenetics and molecular genetics. The karyotypes of hibernoma are complex and usually near or pseudodiploid. The only reported recurrent rearrangement is involvement of 11q13]q21, most often 11q13 in the form of structural rearrangements with translocations involving three or more chromosomes (Sandberg et al., 1986). Metaphase FISH analyses have shown that the aberrations are more complex and that the loss of 11q13 was present also in the apparently ‘normal’ homologue with homozygous and heterozygous deletions comprising segments up to 4 Mb. Homozygous deletion of the multiple endocrine neoplasia type 1 (MEN1) tumor suppressor gene (11q13) and heterozygous loss of PPP1CA at 11q13 has been reported in four of five tumors and all five tumors respectively (Gisselsson et al., 1999b). Maire et al. (2003) in a study of cytogenetic and FISH analysis of two hibernomas have reported that the altered region at 11q13 was larger and the break points were located at 11q13.5, one of which is in the immediate vicinity of the CNGB (GARP) gene, implying a role for this gene in hibernoma pathogenesis. Chondroid lipoma This is a distinctive and uncommon benign fatty neoplasm containing lipoblasts, mature fat and chondroid matrix. This tumor is usually seen in middle aged adults, mostly women involving the proximal extremities or limb girdles, located predominantly in deep soft tissues, but can also be seen in in subcutis (Meis and Enzinger, 1993). They are well-circumscribed yellow neoplasms which show a lobular growth pattern and are composed of large univacuolated or multivacuolated cells which resemble lipoblasts which are
140
Cytogenet Genome Res 118:138–147 (2007)
arranged in cords or nests in a prominent hyaline or myxoid matrix. The histological pattern can be easily confused with malignant tumors such as myxoid liposarcoma and extraskeletal myxoid chondrosarcoma (Meis and Enzinger, 1993; Guillou and Coindre, 2001). Cytogenetics. Two cases with reciprocal t(11;16)(q13;p13) have been reported (Gisselsson et al., 1999a; Thomson et al., 1999). The translocation break point at chromosome 11 was located between CD5 and PLCB3, at least 1 Mb proximal to the region rearranged in hibernoma (Gisselsson et al., 1999a). This translocation is not seen in myxoid liposarcoma and extraskeletal myxoid chondrosarcoma and can serve as an additional diagnostic tool in difficult cases. Spindle cell and pleomorphic lipoma Spindle cell and pleomorphic lipomas are histological ends of a spectrum of a single clinicopathological entity and supported by cytogenetic evidence. They are uncommon benign encapsulated adipocytic tumors (1.5%) arising in elderly men located in the subcutis of the back of the neck, shoulders or upper back (Enzinger and Harvey, 1975). Histologically the tumor shows an admixture of mature adipocytes, small undifferentiated spindle cells with small nuclei without atypia and bundles of thick and ropy collagen. In the other end of the spectrum, pleomorphic lipomas are characterized by small hyperchromatic spindle cells, multinucleated giant cells with radially arranged nuclei in a ‘floret’ like pattern (Shmookler and Enzinger, 1981; Hawley et al., 1994). The spindle cells in both tumors stain positively for CD34 by immunohistochemical analysis (Suster and Fisher, 1997). Cytogenetics. Both tumors show similar cytogenetic aberrations which are generally more complex than conventional lipomas. Most tumors are hypodiploid often with multiple partial losses, no gain of sequences and a few balanced rearrangements. Aberrations involving chromosome 16 with loss of 16q13]qter, monosomy 13 or partial loss of 13q have been described (Mandahl et al., 1994). The structural change seen in 13q is similar to that seen in ordinary lipomas, whether these reflect the same changes at the molecular level is not known. Morphologically these two tumors are distinct entities and do not cause diagnostic confusion. Lipoblastoma Lipoblastoma is a lobulated childhood (first three years) neoplasm resembling fetal adipose tissue which presents as localized (lipoblastoma) or diffuse (lipoblastomatosis) form, most commonly in the extremities. The tumor can be present at birth or in older children and has a male predilection (Collins and Chatten, 1997). Lipoblastomas are confined to subcutis and lipoblastomatosis infiltrate deep muscle. Histologically, they are lobulated tumors showing an admixture of immature and mature adipocytes in varying stages of development. The matrix can be myxoid with plexiform vascular pattern and can mimic myxoid liposarcoma (Mentzel et al., 1993). Cytogenetics and molecular genetics. Lipoblastomas have simple, pseudodiploid karyotypes showing structural chro-
mosomal aberrations. Most cases demonstrate rearrangement of chromsome 8q11]q13 and recurrent recombinations of this segment are seen with 3q12]q13, 7p22 and 8q24 (Sandberg et al., 1986). Two different fusion genes have been reported, HAS2 (8q24)/PLAG1(8q12) in three cases and COL1A2 (7q22)/PLAG1 in one case (Hibbard et al., 2000). PLAG1 involvement is also seen in pleomorphic adenoma of salivary gland with a t(3;8)(p21;q12) resulting in promoter swapping between PLAG1 and the gene for beta catenin (CTNNB1) (Kas et al., 1997). A similar mechanism of promoter swapping has been proposed for lipoblastoma (Astrom et al., 2000; Hibbard et al., 2000). Additional copies of chromosome 8 which are seen in some lipoblastomas may serve as an alternate mechanism of tumorigenesis. In situ hybridization studies have shown split signals of the PLAG1 gene in classical, myxoid and lipoma-like lipoblastomas as well as other mesenchymal cell components indicating the genetic change occurs in a progenitor cell which then differentiates into lipoblastoma (Gisselsson et al., 2001). Miscellaneous Angiolipomas are subcutaneous painful neoplasms containing mature fat and thin walled vessels with thrombi. There has been only a single case report of an angiolipoma with t(X;12)(p22;p12) (Sciot et al., 1997). Angiomyolipomas which occur predominantly in the kidney and contain mature lipomatous, smooth muscle and vascular component have shown chromosomal aberrations involving chromomes 7 and 8 (trisomies) and rearrangements of 12q13] q15; however, these changes have not been observed in a consistent manner (Dal Cin et al., 1997). A CGH study has shown that genomic changes are common and indicated that the region at 5q33]q34 may contain a tumor suppressor gene in some renal angiomyolipomas (Kattar et al., 1999). Malignant neoplasms
Atypical lipomatous tumors Atypical lipomatous tumor (ALT)/well differentiated liposarcoma (WD) accounts for about 40–45% of liposarcomas occurring in middle aged and older individuals. It is an intermediate (locally aggressive) malignant neoplasm involving retroperitoneum, limbs followed by paratesticular region and mediastinum (Laurino et al., 2001). There are four subtypes based on morphological appearance; adipocytic (lipoma-like), sclerosing, inflammatory and spindle cell WD (Mentzel and Fletcher, 1995). Histologically the tumor is composed entirely or partially of mature adipocytic proliferation showing significant variation in cell size and at least focal nuclear atypia of both adipocytes and stromal cells (Fig. 3). In addition different subtypes show morphological variations as indicated by their names. These tumors are fully capable of transformation to a higher grade dedifferentiated liposarcoma (Laurino et al., 2001). Cytogenetics and molecular genetics. A consistent karyotypic change observed in these tumors is the presence of
Lipoblast
Fig. 3. Photomicrograph of ALT/WD showing occasional atypical lipoblasts (400!).
supernumerary ring and/or marker chromosomes (Fig. 4), which are negative on C-banding and lack alpha satellite sequences; however, they have functioning centromeres and are associated with a few other structural and numerical abnormalities (http://cgap.nci.nih.gov/Chromosomes/ Mitelman). Metaphase cells are often near-diploid or neartetraploid. Random and non-random telomere associations can be seen (Mandahl et al., 1998). The rings and marker chromosomes are shown to contain amplified material from the 12q13]q15 region which overlaps with the 12q14]q15 region rearranged in ordinary lipomas. This region contains several genes including MDM2, TSPAN31, CDK4 and HMGA2 which are expressed and amplified in a subset of ALTs (Berner et al., 1997; Pilotti et al., 1998; Dei Tos et al., 2000). Both MDM2 and CDK4 are involved in the regulation of cell cycle, MDM2 by binding to and inhibiting p53 and CDK4 by promoting RB phosphorylation. The involvement of HMGA2 in lipomas and ALTs has led to the hypothesis that there may be a molecular genetic continuum between lipoma and classical ALTs (Dei Tos et al., 2000). The amplified material at the 12q region often contains chromosome 1 sequences involving 1q21]q22 in these tumors. Two putative oncogenes, COAS (chromosome 1 amplified sequence) and PRUNE have been implicated in the pathogenesis (Arrigoni and Doglioni, 2004). Co-amplification of 12q21]q22 has also been reported (Dal Cin et al., 1993). Dedifferentiated liposarcoma Dedifferentiated liposarcoma is a malignant neoplasm showing transition from atypical lipomatous tumor (ALT) to a non-lipogenic high grade sarcoma either in the primary neoplasm or in a recurrence (Evans, 1979; Henricks et al., 1997). Dedifferentiation is thought to be a time dependent phenomenon, most (90%) tumors occurring ‘de novo’ while 10% occur after multiple recurrences of ALTs (Henricks et al., 1997). The tumor occurs in elderly indi-
Cytogenet Genome Res 118:138–147 (2007)
141
2
1
7
6
20
21
5
10
15
14
19
9
8
13
4
3
17
16
22
12
11
X
18
Y
rings A
A
A
A
Fig. 4. G-banded karyotype of ALT/WD with ring chromosomes.
DD
WD
Fig. 5. Dedifferentiated liposarcoma showing both WD and DD components.
142
Cytogenet Genome Res 118:138–147 (2007)
viduals, retroperitoneum being the common site and can also affect extremities, trunk and rarely subcutaneous tissue (McCormick et al., 1994; Henricks et al., 1997). The histological hallmark of this tumor is the transition of ALT/ WD liposarcoma to high grade non-lipogenic sarcoma, most often in an abrupt fashion (Fig. 5). The high grade component exhibits variable histological patterns; however, most frequently they resemble malignant fibrous histiocytoma (MFH) or myxofibrosarcoma (Nascimento, 2001). Low-grade dedifferentiation has also been described (Henricks et al., 1997). These tumors have a local recurrence rate of 41–52% and metastasis rate of 15–20% (Nascimento, 2001). Cytogenetics. Similar to ALT, dedifferentiated liposarcomas are characterized by the presence of ring or marker chromosomes (Mertens et al., 1998; Pilotti et al., 2000). In addition, their karyotype is relatively simple (Fig. 6) unlike their other histological mimics such as malignant fibrous histiocytoma (MFH) and pleomorphic liposarcoma (Mertens et al., 1998). Molecular genetics. CGH and FISH analyses have revealed amplification of the 12q13]q21 region associated with co-amplification of other regions. Southern blot anal-
1
2
7
6
13
19
3
4
9
8
dm
dm
12
dic 11;19
14
15
20
21
ring 2
11
10
17
16
22
dic 22;22 dm
5
X
18
Y
der (X)
ring 3
Fig. 6. Karytotype of DD showing multiple aberrations including ring chromosomes and double minutes (dm).
ysis showed MDM2 amplification in 5/5 retroperitoneal and not in 4/4 non retroperitoneal tumors (Pilotti et al., 1997). The majority of MDM2 positive dedifferentiated liposarcomas also do not reveal p53 mutations unlike other high grade sarcomas such as MFH (Dei Tos et al., 1997). Microarray analysis of 45 WD and five DD liposarcomas showed a distinct gene expression pattern in the WD component of the DD cases as compared to other WD cases. In this unsupervised hierarchical clustering analysis, 1,687 genes including 487 known genes were found to be discriminatory between the two groups (Shimoji et al., 2004). Recently, the first new human cell line of a dedifferentiated liposarcoma (FU-DDLS-1) has been established (Nishio et al., 2003). Despite high grade morphology, the clinical course of DD is less aggressive than other high grade sarcomas. Most likely this is related to the relatively simple karyotypic abnormalities and preservation of wild type p53 gene in most cases (Dei Tos, 2000). Myxoid liposarcoma Myxoid liposarcomas (MLS) represent about 30–35% of all liposarcomas and are malignant tumors composed of primitive mesenchymal cells, signet ring lipoblasts in a myxoid background with characteristic plexiform vascular
network (Antonescu et al., 2001). A histological spectrum exists in these tumors with myxoid and round cells morphological patterns (round cell liposarcoma) representing well and poorly differentiated components supported by clinical, morphologic and cytogenetic evidence (Orndal et al., 1990). The tumor typically occurs in adults (4th to 5th decade) and predominantly in extremities and only very rarely occurs in the retroperitoneum. Grossly this tumor has a gelatinous appearance and microscopically is composed of uniform round to oval cells and signet ring cells in a myxoid background with a delicate arborizing capillary network (Fig. 7A). When there is histological progression, the tumor shows transition towards hypercellular round cell morphology (Fig. 7B) representing dedifferentiation (Orvieto et al., 2001). Cytogenetics and molecular genetics. Myxoid and round cell liposarcomas show a characteristic t(12; 16)(q13;p11) (Fig. 8) in 90% of cases and t(12;22)(q13;q12) in the remaining 10% of cases (Sreekantaiah et al., 1992). These translocations involve fusion of DDIT3 (CHOP) on chromosome 12q13 with either FUS (TLS) on chromosome 16p11 or EWS on 22q12 (Aman et al., 1992; Panagopoulos et al., 1996). DDIT3 encodes a transcription factor belonging to the C/EBP group of basic leucine zipper family of transcription
Cytogenet Genome Res 118:138–147 (2007)
143
A
B Fig. 7. (A) Photomicrograph of myxoid liposarcoma showing primitive lipoblasts with delicate capillary network (400!). (B) Myxoid liposarcoma showing round cell dedifferentiation.
2
1
6
7
3
9
8
5
4
10
11
12
t(12;16)
13
19
14
20
15
21
16
t(12;16)
22
17
X
18
Y
Fig. 8. Karyotype of myxoid liposarcoma showing characteristic t(12;16)(q13;p11.1).
factors involved in adipogenesis (Ron and Habener, 1992). A dual color break apart rearrangement probe (LSI CHOPVysis쏐) is available which can be used in interphase/metaphase cells and paraffin embedded tissue sections. FUS and EWS encode proteins with RNA/DNA binding activities and act as transcriptional activators (Rabbitts et al., 1993).
144
Cytogenet Genome Res 118:138–147 (2007)
The chimeric proteins are characterized by retention of the transcriptional activation domain of FUS and EWS and replacement of their RNA/DNA binding region by the basic leucine zipper domain of DDIT3. This promotes proteinprotein interaction through dimerization and leads to blockage of adipogenesis with cell proliferation and resul-
Table 1. Summary of the various cytogenetic abnormalities in the neoplasms
Neoplasms
Cytogenetics
Molecular genetics
Benign neoplasms Lipoma
Rearrangements of 12q13–q15 Rearrangements of 13q (deletions 13q12–q22) Rearrangements of 6p23–p21 and 7 Complex karyotype Hibernoma Recurrent rearrangement 11q13 Rearrangement 11q13 t(11;16)(q13;p12–p13) Chondroid lipoma Spindle cell lipoma/pleomorphic lipoma Complex karyotype Deletions of 13q12 and 13q14–q22 Deletions of 16q13–qter Rearrangement of 8q11–q13 Lipoblastoma
Malignant neoplasms Atypical lipomatous tumor (ALT)/well differentiated liposarcoma Dedifferentiated liposarcoma Myxoid/round cell liposarcoma Pleomorphic liposarcoma
Supernumerary ring/giant marker chromosomes and dmin (amplification of 12q13–q15) Same as ALT can be more complex t(12;16)(q13;p11) t(12;22)(q13;q12) Complex karyotype (numerical and structural rearrangements)
tant tumor formation. The expression of the chimeric proteins in transgenic mice has led to the development of myxoid liposarcoma in these animals suggesting that this cytogenetic and molecular change alone is sufficient for tumor formation (Perez-Losada et al., 2000). Other karyotypic abnormalities in these tumors include +8, del (6q), and der (16) t(1; 16) and others (http://cgap.nci.nih.gov/Chromosomes/ Mitelman). Unfavorable outcome is predicted by high grade histology (round cell component), necrosis and p53 overexpression (Antonescu et al., 2001). Among sarcomas, MLS tend to metastasize in unusual sites such as retroperitoneum, opposite extremity, axilla, bone, abdominal cavity, etc. (Antonescu and Ladayni, 2002). Pleomorphic liposarcoma Pleomorphic liposarcoma is a rare high grade sarcoma containing a variable number of pleomorphic multivacuolated lipoblasts (Fletcher, 1992). The majority occur in elderly patients (6th to 7th decade) and affect lower extremities followed by upper extremities and retroperitoneum (Oliveira and Nascimento, 2001). These tumors are grossly multinodular with yellow to tan cut surfaces. Microscopically, the tumor is composed of pleomorphic multivacuolated lipoblasts admixed with spindled tumor cells and multinucleated giant cells (Mentzel and Pedeutour, 2002). Cytogenetics and molecular genetics. Only a handful of these tumors have been cytogenetically analyzed and show aneuploid karyotypes with complex structural aberrations. The chromosome counts can exceed 200 and the complexity is represented by numerous unidentifiable marker chromosomes, non-clonal alterations and intercellular heterogeneity making it difficult to detect specific rearrangements (Mertens et al., 1998). Some reported tumors have ring, marker or double minute chromosomes (Sreekantaiah et al., 1992). These tumors most likely represent dedifferentiated liposarcoma
HMGA2/LPP, HMGA2/LHFP HMGA2/others not known HMGA1 MEN1 and PPPICA Not known Not known Not known HAS2/PLAG1, COL1A2/PLAG1 Amplification of MDM2, TSPAN31, CDK4, HMGA2 Same as ALT DDIT3/TLS DDIT3/EWS Not known (possibly MDM2, TP53)
without a well-differentiated component. It is necessary to revisit histology in cases of pleomorphic liposarcoma where karyotypes reveal ring or marker chromosomes without complex structural and numerical arrangements. A CGH and cDNA based expression array analysis of dedifferentiated and pleomorphic liposarcoma showed a number of genes differentially expressed between the two tumor types (Fritz et al., 2002). CGH analysis of six pleomorphic liposarcomas showed predominantly gain of chromosomes 1, 5p, 13q and 22q (Schmidt et al., 2005). In this study, gain of 13q21]q32 correlated with poor prognosis in pleomorphic, myxoid/round cell and dedifferentiated liposarcomas. Conclusions
Adipocytic tumors are diverse neoplasms which have been classified traditionally based on clinicopathological criteria. Table 1 summarizes the various cytogenetic abnormalities in the above mentioned neoplasms. In the past 20 years, considerable strides have been achieved in the field of tumor cytogenetics and molecular genetics with significant increase in knowledge of the pathogenesis of these neoplasms. The consistent presence of ring and/or marker chromosomes in well-differentiated liposarcomas has allowed pathologists to accurately classify these tumors from their morphological mimics. The presence of t(12;16)(q13;p11) or t(12; 22)(q13;q12) in nearly 100% of myxoid liposarcomas has not only tremendous diagnostic value, but also has opened many doors towards exploration of the molecular genetic pathways and testing possible new therapeutic targets. Correlation between morphology and cytogenetic/ molecular genetic findings is essential in understanding the pathogenesis and biology of these very complex neoplastic processes.
Cytogenet Genome Res 118:138–147 (2007)
145
References Aman P, Ron D, Mandahl N, Fioretos T, Heim S, et al: Rearrangement of the transcription factor gene CHOP in myxoid liposarcomas with t(12;16)(q13;p11). Genes Chromosomes Cancer 5:278–285 (1992). Antonescu C, Ladayni M: Myxoid Liposarcoma. World Health Organization Classification of Tumors, Pathology and Genetics of Tumors of Soft Tissue and Bone (IARC Press, Lyon 2002). Antonescu C, Tschernyavsky SJ, Decuseara R, Leung DH, Woodruff JM, et al: Prognostic impact of p53 status, TLS-CHOP fusion transcript structure and histological grade in myxoid liposarcoma: A molecular and clinicopathologic study of 82 cases. Clin Can Res 7: 3977–3987 (2001). Arrigoni G, Doglioni C: Atypical lipomatous tumor: molecular characterization. Curr Opin Oncol 16:355–358 (2004). Astrom A, D’Amore ESG, Sainati L, Panarello C, Moreiro C, et al: Evidence of involvement of the PLAG1 gene in lipoblastomas. Int J Oncol 16: 1107–1110 (2000). Berner JM, Meza-Zepeda LA, Kools PF, Forus A, Schoenmakers EF, et al: HMGIC, the gene for an architectural transcription factor, is amplified and rearranged in a subset of human sarcomas. Oncogene 14: 2935–2941 (1997). Borberg K, Zhang M, Strombeck B, Isaksson M, Nilsson M, et al: Fusion of RDC1 and HMGA2 in lipomas as the result of chromosome aberrations involving 2q35–q37 and 12q13–q15. Int J Oncol 21:321–326 (2002). Borrmann L, Seebeck B, Rogalla P, Bullerdiek J: Human HMGA2 promotor is coregulated by a polymorphic dinucleotide (TC)-repeat. Oncogene 22: 756–760 (2003). Collins MH, Chatten J: Lipoblastoma/lipoblastomatosis: A clinicopathological study of 25 cases. Am J Surg Pathol 10: 1131–1137 (1997). Dahlen A, Debiec-Rychter M, Pedeutour F, Domanski HA, Hoglund M, et al: Clustering of deletions on chromosome 13 in benign and low malignant lipomatous tumors. Int J Cancer 103: 616–623 (2003). Dal Cin P, Kools P, Sciot R, De Wever L, Van Damme B, et al: Cytogenetic and fluorescence in situ hybridization investigation of ring chromosomes characterizing a specific pathologic subgroup of adipose tissue tumors. Cancer Genet Cytogenet 68: 85–90 (1993). Dal Cin P, Sciot R, Van Poppel H, Baert L, Van Damme B, et al: Chromosome analysis in angiomyolipoma. Cancer Genet Cytogenet 99: 132–134 (1997). Dei Tos AP: Liposarcoma: new entities and evolving concepts. Ann Diag Pathol 4: 252–266 (2000). Dei Tos AP, Doglioni C, Piccinin S, Maestro R, Mentzel T, et al: Molecular abnormalities of the p53 pathway in dedifferentiated liposarcoma. J Pathol 181: 8–13 (1997). Dei Tos AP, Doglioni C, Piccinin S, Sciot R, Furlanetto A, et al: Coordinated expression and amplification of the MDM2, CDK4 and HMGI-C genes in atypical lipomatous tumours. J Pathol 190:531–536 (2000). Enzinger FM, Harvey DA: Spindle cell lipoma. Cancer 36:1852–1859 (1975). Enzinger FM, Weiss SW: Soft Tissue Tumors. 2nd ed (C.V. Mosby, St Louis 2001). Evans HL: Liposarcoma: A study of 55 cases with reassessment of its classification. Am J Surg Pathol 3: 507–523 (1979). Fletcher CD: Pleomorphic malignant fibrous histiocytoma: Fact or fiction? A critical appraisal based on 159 tumors diagnosed as pleomorphic sarcoma. Am J Surg Pathol 16: 213–228 (1992).
146
Fletcher JA, Longtine J, Wallace K, Mentzer SJ, Sugarbaker DJ: Cytogenetic and histologic findings in 17 pulmonary chondroid hamartomas: evidence for a pathogenetic relationship with lipomas and leiomyomas. Genes Chromosomes Cancer 12: 220–223 (1995). Fletcher CDM, Unni KK, Mertens F (eds): World Health Organization Classification of Tumors, Pathology and Genetics of Tumors of Soft Tissue and Bone (IARC Press, Lyon 2002). Fritz B, Schubert F, Wrobel G, Schwaenen C, Wessendorf S, et al: Microarray based copy number and expression profiling in dedifferentiated and pleomorphic liposarcoma. Cancer Res 62: 2993–2998 (2002). Gisselsson D, Domanski HA, Hoglund M, Carlen B, Mertens F, et al: Unique cytological features and chromosome aberrations in chondroid lipoma: A case report based on fine needle aspiration cytology, histopathology, electron microscopy, chromosome banding and molecular cytogenetics. Am J Surg Pathol 23: 1300–1304 (1999a). Gisselsson D, Hoglund M, Mertens F, Dal Cin P, Mandahl N: Hibernomas are charaterized by homozygous deletions in the multiple endocrine neoplasia type 1 region. Metaphase fluorescence in situ hybridization reveals complex rearrangements not detected by conventional cytogenetics. Am J Surg Pathol 155: 61–66 (1999b). Gisselsson D, Hibbard MK, Dal Cin P, Sciot R, His B-L, et al: PLAG1 alterations in lipoblastomas: involvement in varied mesenchymal cell types and evidence for alternative oncogenic mechanisms. Am J Pathol 159: 955–962 (2001). Guillou L, Coindre JM: Newly described adipocytic lesions. Sem Diagn Pathol 18: 238–249 (2001). Hawley IC, Krausz T, Evans DJ, Fletcher CD: Spindle cell lipoma – a pseudoangiomatous variant. Histopathology 24: 565–569 (1994). Henricks WH, Chu YC, Goldblum JR, Weiss SW: Dedifferentiated liposarcoma: A clinicopathological study of 155 cases with a proposal for an expanded definition of dedifferentiation. Am J Surg Pathol 21:271–288 (1997). Hess JL: Chromosomal translocations in benign tumors: The HMGI proteins. Am J Clin Pathol 109:251–261 (1998). Hibbard MK, Kozakewich HP, Dal Cin P, Sciot R, Tan X, et al: PLAG1 fusion genes in lipoblastoma. Cancer Res 60: 4869–4872 (2000). Kas K, Voz ML, Roijer E, Astrom AK, Meyen E, et al: Promoter swapping between the genes for a novel zinc finger protein and beta catenin in pleomorphic adenomas with t(3; 8) (p21;q12) translocations. Nat Genet 15: 170–174 (1997). Kattar MM, Grignon DJ, Eble JN, Hurley PM, Lewis PE, et al: Chromosomal analysis of renal angiomyolipomas by comparative genomic hybridization: evidence for clonal origin. Hum Pathol 30: 295–299 (1999). Kazmierczak B, Dal Cin P, Wanschura S, Borrmann L, Fusco A, et al: HMGIY is the target of 6p21.3 rearrangements in various benign mesenchymal tumors. Genes Chromosomes Cancer 23: 279–285 (1998). Kazmierczak B, Dal Cin P, Meyer-Bolte K, Van den Berghe H, Bullerdiek J: HMGI is not rearranged by 13q12 aberrations in lipomas. Genes Chromosomes Cancer 24: 290–292 (1999). Laurino L, Furlanetto A, Orvieto E, Dei Tos A: Well differentiated liposarcoma (atypical lipomatous tumors). Sem Diag Pathol 18: 258–262 (2001).
Cytogenet Genome Res 118:138–147 (2007)
Maire G, Forus A, Bjerkehagen B, Mainguene C, Kresse SH, et al: 11q13 alterations in two cases of hibernoma; large heterozygous deletions and rearrangement breakpoints near GARP in 11q13.5. Genes Chromosomes Cancer 37: 389– 395 (2003). Mandahl N, Orndal C, Heim S, Willen H, Rydholm A, et al: Aberrations of chromosome segment 12q13–q15 characterize a subgroup of hemangiopericytoma. Cancer 71:3009–3013 (1993). Mandahl N, Mertens F, Willen H, Rydholm A, Brosjo O, Mitelman F: A new cytogenetic subgroup in lipomas: Loss of chromosome 16 material in spindle cell and pleomorphic lipomas. J Cancer Res Clin Oncol 120:707–711 (1994). Mandahl N, Mertens F, Willen H, Rydholm A, Kreicbergs A, Mitelman F: Nonrandom pattern of telomeric associations in atypical lipomatous tumors with ring and giant marker chromosomes. Cancer Genet Cytogenet 103: 25–34 (1998). McCormick D, Mentzel T, Beham A, Fletcher CD: Dedifferentiated liposarcoma: Clinicopathologic analysis of 32 cases suggesting a better prognosis subgroup among pleomorphic sarcomas. Am J Surg Pathol 18: 1213–1223 (1994). Meis JM, Enzinger FM: Chondroid lipoma, a unique tumor simulating liposarcoma and myxoid chondrosarcoma. Am J Surg Pathol 17: 1103– 1112 (1993). Mentzel T, Fletcher CDM: Lipomatous tumors of soft tissue; an update. Virchows Arch 427:353– 363 (1995). Mentzel T, Pedeutour F: Pleomorphic Liposarcoma. World Health Organization Classification of Tumors, Pathology and Genetics of Tumors of Soft Tissue and Bone (IARC Press, Lyon 2002). Mentzel T, Calonje E, Fletcher CD: Lipoblastoma and lipoblastomatosis: A clinicopathological study of 14 cases. Histopathology 23: 527–533 (1993). Mertens F, Fletcher CD, Dal Cin P, de Wever I, Mandahl N, et al: Cytogenetic analysis of 46 pleomorphic soft tissue sarcomas and correlation with morphologic and clinical features: a report of the CHAMP study group. Genes Chromosomes Cancer 22: 16–25 (1998). Mitelman F, Johansson B, Mertens F (eds): Mitelman Database of Chromosome Aberrations in Cancer. http://cgap.nci.nih.gov/Chromosomes/ Mitelman (2006). Nascimento AG: Dedifferentiated liposarcoma. Sem Diag Pathol 18:263–266 (2001). Nilbert M, Heim S: Uterine leiomyoma cytogenetics. Genes Chromosomes Cancer 2: 3–13 (1990). Nilsson M, Mertens F, Hoglund M, Mandahl N, Panagopoulos: Truncation and fusion of HMGA2 in lipomas with rearrangements of 5q32]q33 and 12q14]q15. Cytogenet Genome Res 112: 60–66 (2006). Nishio J, Iwasaki H, Ishiguro M, Ohjmi Y, Fuita C, et al: Establishment of a novel human dedifferentiated liposarcoma cell line, FU-DDLS-1: Conventional and molecular cytogenetic characterization. Inter J Oncol 22:535–542 (2003). Oliviera AM, Nascimento AG: Pleomorphic liposarcoma. Sem Diag Pathol 18: 274–285 (2001). Orndal C, Mandahl N, Rydholm A, Nilbert M, Heim S, et al: Chromosomal evolution and tumor progression in a myxoid liposarcoma. Acta Orthop Scand 61: 99–105 (1990). Orvieto E, Furlanetto A, Laurino L, Dei Tos AP: Myxoid and round cell liposarcoma: A spectrum of myxoid adipocytic neoplasia. Semin Diag Pathol 18: 267–283 (2001).
Panagopoulos I, Hoglund M, Mertens F, Mandahl N, Mitelman F, Aman P: Fusion of EWS and CHOP genes in myxoid liposarcoma. Oncogene 12:489–494 (1996). Perez-Losada J, Pintado B, Ggutierrez-Adan A, Rodriguez-Garcia MA, Perez Mancera PA, et al: The chimeric FUS/TLS-CHOP fusion protein specifically induces liposarcomas in transgenic mice. Oncogene 19: 2413–2422 (2000). Petit MMR, Mols R, Schoenmakers EFPM, Mandahl N, Van de Ven WJM: LPP, the preferred fusion partner gene of HMGIC in lipomas, is a novel member of the LIM family. Genomics 36: 118–126 (1996). Petit MM, Schoenmakers EF, Huysmans C, Geurts JM, Mandahl N, Van de Ven WJ: LHFP, a novel translocation partner gene of HMGIC in a lipoma, is a member of a new family of LHFP-like genes. Genomics 57: 438-441 (1999). Pilotti S, Della Torre G, Lavarino C, di Palma S, Sozzi G, et al: Distinct mdm2/p53 expression patterns in liposarcoma subgroups: implications for different pathogenetic mechanisms. J Pathol 181: 14–24 (1997). Pilotti S, Della Torre G, Lavarino C, Sozzi G, Minoletti F, et al: Molecular abnormalities in liposarcoma: Role of MDM2 and CDK4-containing amplicons at 12q13–q22. J Pathol 185: 188–190 (1998). Pilotti S, Mezzelani A, Vergani B, Minoletti F, Cristofori E, et al: Morphologic-cytogenetic analysis of dedifferentiated liposarcomas with an extensive misleading leiomyosarcomatous component. App Immunohistochem Mol Morphol 8:216-221 (2000).
Rabbitts TH, Forster A, Larson R, Nathan P: Fusion of the dominant negative transcription regulator CHOP with a novel gene FUS by translocation t(12; 16) in malignant liposarcoma. Nat Genet 4:175–180 (1993). Ron D, Habener JF: CHOP, a novel developmentally regulated nuclear protein that dimerizes with transcription factors C/EBP and LAP and functions as a dominant-negative inhibitor of gene transcription. Genes Dev 6: 439–453 (1992). Rubin BP, Dal Cin P: The genetics of lipomatous tumors. Sem Diag Pathol 18: 286–293 (2001). Sandberg AA, Bridge JA: The cytogenetics of bone and soft tissue tumors (RG Landers, Austin 1994). Sandberg AA, Gibas Z, Saren E, Li FP, Tebbi CK: Chromosome abnormalities in two benign adipose tumors. Cancer Genet Cytogenet 22: 55– 61 (1986). Schmidt H, Bartel F, Kappler M, Wurl P, Lange H, et al: Gains of 13q are correlated with a poor prognosis in liposarcoma. Mod Pathol 18: 638– 644 (2005). Schoenmakers EFPM, Wanschura S, Bullerdiek J, Van denBerghe H, Van de Ven WJM: Recurrent rearrangements in the high mobility group protein gene, HMGI-C, in benign mesenchymal tumors. Nat Genet 10: 436–443 (1995). Sciot R, Akerman M, Dal Cin P, De Wever I, Fletcher CDM, et al: Cytogenetic analysis of subcutaneous angiolipoma: further evidence supporting its difference from ordinary pure lipomas: a report of the CHAMP study group. Am J Surg Pathol 21: 441–444 (1997).
Shimoji T, Kanda H, Kitagawa T, Kadota K, Asai R, et al: Clinico-molecular study of dedifferentiation in well-differentiated liposarcoma. Biochem Biophy Res Comm 315: 1133–1140 (2004). Shmookler BM, Enzinger FM: Pleomorphic lipoma: A benign tumor simulating liposarcoma. A clinicopathologic analysis of 48 cases. Cancer 47:126–133 (1981). Sreekantaiah C, Karakousis CP, Leong SP, Sandberg AA: Cytogenetic findings in liposarcoma correlate with histopathologic subtypes. Cancer 69: 2484–2495 (1992). Suster S, Fisher C: Immunoreactivity for the human hematopoetic progenitor cell antigen (CD34) in lipomatous tumors. Am J Surg Pathol 21: 195– 200 (1997). Tallini G, Dal Cin P: HMGI(Y) and HMGI-C dysregulation: A common occurrence in human tumors. Adv Anat Pathol 6: 237–246 (1999). Tallini G, Vanni R, Manfioletti G, Kazmierczak B, Faa G, et al: HMGI-C and HMGI(Y) immunoreactivity correlates with cytogenetic abnormalities in lipomas, pulmonary chondroid hamartomas, endometrial polyps and uterine leiomyomas and is compatible with rearrangement of HMGI-C and HMGI(Y) genes. Lab Invest 80: 359–369 (2000). Thomson TA, Horsman D, Bainbridge TC: Cytogenetic and cytologic features of chondroid lipoma of soft tissue. Mod Pathol 12: 88–91 (1999). Wanschura S, Belge C, Stenman G, Kools P, Dal Cin P, et al: Mapping of the translocation breakpoints of primary pleomorphic adenoma and lipomas within a common region of chromosome 12. Cancer Genet Cytogenet 86: 39–45 (1996).
Cytogenet Genome Res 118:138–147 (2007)
147
Table 1. Summary of CGH gains and losses
Materials and methods Cell lines The methods for establishment and characterization of 11 cell lines (Panc 1.28, 2.03, 2.13, 3.27, 4.03, 4.14, 4.21, 6.03, 8.13, 9.06, 10.05) are described in detail elsewhere (Jaffee et al., 1998). Four additional cell lines from patients undergoing resection of primary adenocarcinoma of the pancreas (cell lines 2.02, 2.05, 2.08, 2.43), were established using the same conditions. All specimens were cultured in media formulated for optimal epithelial cell growth with the addition of insulin growth factors. Cell lines were grown in culture for 3–4 months and then frozen in liquid nitrogen at passage 4 to 17. When thawed for analysis, cultures were passaged an additional 4–5 times before harvest. Karyotypes G-band cytogenetic analysis was performed using standard techniques (Griffin et al., 1995). At least 20 metaphases were analyzed for each cell line. The G-banded karyotypes of the 11 previously described lines, with the exception of cell line 4.14, were reported previously (Jaffee et al., 1998). Karyotypes were described using guidelines established by International System for Human Cytogenetic Nomenclature (ISCN 2005). Metaphase comparative genomic hybridization DNA was isolated from cultured cells using standard protocols. Preparation of probes, hybridization, and image analysis was performed as previously described (Riopel et al., 1998). Briefly, reference and tumor DNA were labeled with Texas Red and FITC-dUTP (DuPont, Boston, MA) or with Spectrum-Red and Spectrum-Green-dUTP (Vysis, Downers Grove, IL) and cohybridized in equimolar amounts onto normal male metaphases. Hybridization proceeded for three days at 37 ° C in a moist chamber. Following washes in 2! SSC at 75 ° C, 37 ° C and room temperature, the chromosomes were counterstained with DAPI in antifade. A minimum of fifteen metaphases were captured and analyzed to generate a ratio profile. Analysis was performed utilizing dedicated software and hardware of the Cytovision system (Applied Imaging Corp., San Jose CA). Overrepresentations were interpreted from ratios 11.25; highly amplified 11.5, underrepresentations !0.75. Spectral Karyotyping Spectral Karyotyping 쏐 (SKY) analysis was performed on air-dried slides, made from standard cytogenetic harvests, hybridized according to the protocol supplied by the probe manufacturer (Applied Spectral Imaging, Inc., Carlsbad, CA). Slides were incubated with the separately denatured probe mix for two days, then washed and detected. The SKY probe is a mixture of whole-chromosome paint probes for each chromosome, combinatorially labeled with five fluorochromes. Metaphase images were acquired on a Zeiss Axiophot microscope with the ASI SpectraCube SD200, and DAPI counterstained images inverted by SkyView software (ASI) to provide enhanced banding. Five cell lines were analyzed at NHGRI using a Leica microsope with SpectraCube SD200 and SkyView software. Ten metaphases were captured and analyzed for each cell line. Chromosome abnormalities were described according to ISCN (2005) guidelines whenever possible. Targetted FISH YACs localized to the 8p21 and 18q11.2 regions were used (Fondation Jean Dausset-CEPH, Paris, France). DNA was prepared, labeled with Spectrum Green and Spectrum Orange fluorochromes and cohybridized with CEP 8 or 18 centromere probes (Vysis) to metaphases from normal cells to verify location and lack of chimerism, and to selected cell lines using standard FISH protocols. DAPI counterstained images were used to further identify derivative and normal chromosomes.
Chromosome region
No. of lines (% of all lines)
Through gain/ loss of entire chromosome
GAIN 11q 8q 20q 3q 1q 7p 7q 14q 15q 2q 12p 19q
13 (86.6) 11 (73.3) 10 (66.6) 9 (60.0) 8 (53.3) 7 (46.6) 7 (46.6) 7 (46.6) 7 (46.6) 6 (40.0) 5 (33.3) 5 (33.3)
4 3 6 2 0 2 2 3 2 3 0 3
9 8 4 7 8 5 5 4 5 3 5 2
LOSS 18q 17p 6q 8p 10q Y
14 (93.3) 11 (73.3) 9 (60.0) 9 (60.0) 6 (40.0) 5 (71.4)a
5 0 3 0 3 5
9 11 6 9 3 0
a
Through gain/loss of less than entire chromosome
Seven cell lines were from males; 5/7 = 71.4%.
Results and discussion
CGH CGH identified extensive copy number gains and losses in all 15 cell lines, with an average of 19.5 copy changes per line. Results are summarized in Table 1 and Fig. 1. The number of genomic gains per cell line ranged from 8 to 15, with an average of 11.1. Gain of 11q was the most common, found in 13 lines. Gain of 8q was observed in 11 lines, followed by 17q (10 lines), 20q (10 lines), 3q (9 lines), 1q (8 lines), 14q (7 lines), 15q (7 lines), 2q (6 lines), 12p (5 lines), and 19q (5 lines ). Gains occurred both through gain of an entire chromosome and of a portion of a chromosome. High level gains were found most often on chromosome 8q, including 8q24.2]q24.3 in two cell lines, and at 8q21]qter and all of 8q in one line each. Other areas of high level gain were found at 3q23]qter, 14q11.2]qter, and 12p in one cell line each. Genomic losses per cell line ranged from 2 to 15 with an average of 8.3. Loss of chromosome 18 material was the most common finding. This occurred by loss of a copy of chromosome 18 in five lines, loss of 18q in seven, and loss of a portion of 18q in two lines. Other frequent genomic losses included 17p (11 lines), 6q (nine lines), 8p (nine lines), 10q (six lines), and the Y chromosome (five of seven cell lines from males). Similar to what we observed for gains, loss occurred both through loss of an entire chromosome or portions of the chromosome. These findings are similar to those in reports of chromosomal CGH analysis of other pancreas cancer cell lines and
Cytogenet Genome Res 118:148–156 (2007)
149
1
2
7
13
19
3
4
8
14
20
9
15
10
16
21
5
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
X
Y
6
11
12
17
22
1
18
X
Y
Fig. 1. Metaphase CGH analysis of 15 pancreatic carcinoma cell lines. Lines indicate areas of loss (to the left of chromosome) and gain (to the right).
Fig. 2. Chromosomal breakpoints derived by SKY and G-band karyotypes. ( SKY indicated a balanced rearrangement; $ breakpoint derived from G-banded karyotype (SKY not performed).
primary tumors (Solinas-Toldo et al., 1996; Fukushige et al., 1997; Mahlamaki et al., 1997, 2002; Curtis et al., 1998; Ghadimi et al., 1999; Schleger et al., 2000; Shiraishi et al., 2001; Harada et al., 2002; Kitoh et al., 2005). As noted in a recent review (Karhu et al., 2006), 14–27 genetic changes per cell line were observed in published reports. As summarized in that review, the most commonly reported losses have been 6q (30–50% of cases), 9p (30–89%), 13q (15–67%), and 18q (42–89%). These series have found gains at 7q (56–67%), 8q (24–67%), 11q (56–67%), 17q (15–58%), and 20q (15–83%).
mal chromosomes were investigated, and SKY provided additional information in 67. The cell line karyotypes are listed in Table 2, and the breakpoints of clonal structural chromosomal aberrations are summarized in Fig. 2. A representative SKY metaphase is shown in Fig. 3. Rearrangements included 11 whole arm translocations, only one of which was seen twice: der(1;7)(p10;q10) in lines 2.3 and 6.3. 14q10 was involved in translocations with 5q10, 19q10, and 22q10 one time each; 15q10 was involved in translocations with 5p10 and 20q10 one time each. Ten isochromosomes were observed: i(22)(q10) and i(13)(q10) were found in three cell lines each, while i(1)(q10), i(5)(q10), i(8)(q10), and i(14)(q10) were observed once. Translocations that appeared to be balanced were rare. These included t(3;14)(p21;q22)del(14)(q22q32) in line 1.28, t(8;11)(q13;p15) in line 1.28, t(7;16)(q36;q11.2) in line 2.43,
Spectral karyotyping (SKY) The complexity of the G-band karyotypes of pancreatic cancer cell lines precludes full interpretation of structural rearrangements. We used SKY to further characterize chromosomal abnormalities in 11 lines. 93 structurally abnor-
150
Cytogenet Genome Res 118:148–156 (2007)
Table 2. Karyotypes of pancreatic carcinoma cell lines interpreted using SKY (except Line 4.14)
Cell Line
Karyotype
Panc 6.03
70⬃73<4n>,XX,–Y,–Y,+der(1)t(1;20;8)(?;?;?)dup(20)(?),ins(4;1)(q26;q42q44)x2,–7,der(7)t(7;11)(q31;q22),del(8)(p21)!2, der(8)t(8;22)(p22;q13)x2,–9,–9,–10,–10,–10,der(12)t(12;18)(p11.2;q12)!2,–13,–13,der(14;19)(q10;q10), der(14;22)(q10;q10),–17,–17,–18,–18,der(18)t(18;22)(q11.2;q11.2),–19,–22,i(22)(q10),1⬃3dmin[cp6]
Panc 2.03
71⬃77<4n>,XX,der(X)t(X;1)(p22.1;q25),der(X;4)(q10;p10),del(1)(q11),der(1;3)(q10;q10)!2,+der(1;7)(p10;q10), der(5;14)(q10;q10)!2,–6,–6,dup(8)(q23q24.3)!2,der(8)t(8;18)(p21;q23)!2,–9,–9,–10,–10, der(11)t(11;17)(p15;q21)dup(11)(p14p15)!2,–13,–13,der(17)t(9;17)(q21;p11.2)!2, der(18)t(18;22)(q11.2;q11.2)dup(22)(q?)!2,der(19)t(2;19)(q36;q13.4)!2,–21,–22,–22,1⬃2dmin [cp5]
Panc 2.13
42⬃43,XX,der(1;21)(p10;p10)del(1)(p34.1p36.1),i(1)(q10),der(5;15)(p10;q10),+i(5)(q10),–6,der(6)t(3;6)(q21;p21.3), der(7)t(3;7)(q21;q32),del(8)(p21),der(9)t(9;13)(p21;q12),r(10)del(10)(?),del(16)(p12),der(17)t(17;21)(p11.2;q22), der(18)ins(6;18)(p12;p11.3q12)del(6)(p12q25),–21,–22 [cp7]/80<4n>,idemx2,–X,–X,–i(5),–i(5),–r(10),–r(10),–21,–21 [1]
Panc 9.06
70⬃73<4n>,XX,–X,–X,–6,–6,der(8)t(8;14)(p12;q13)!2,r(9)del(9)(?),–14,–14,der(15;20)(q10;q10)!2, der(17)t(17;22)(p11.2;q12)!2,–18,–18,der(18)t(18;22)(q11.2;q11.2),–22,–22[cp5]
Panc 8.13
48⬃53,X,–Y,+ider(1)(q10)t(1;8)(q21;q13),+2,+3,+del(6)(p21.3),del(8)(p12),+der(8)t(8;16)(q22;q11.1),+9,+11, +del(11)(p11.2),+12,–13,i(13)(q10),–14,ider(14)(q10)dup(14)(q11.2q31),der(15)t(14;15)(q?;p11.1)del(14)(q?)!2, +der(15)t(15;22)(p11.1;q12),–16,der(17)t(16;17)(p11.1;p11.1),+del(17)(q23),–18,+20, der(20)t(18;20)(p11.2;p11.1)!2,del(22)(q11.2)[cp4]
Panc 4.14
48⬃50,XX,del(6)(p21.3),+del(7)(p21),+i(8)(q10),+del(11)(p15)[cp7]a
Panc 10.05
39,X,–Y,der(1)t(1;22)(p12;q12)ins(1;14)(p12;q32q12),–3,der(4)t(4;10)(q12;q21),der(6)t(6;17)(q14;q12)del(17)(q22q24), i(8)(q10),–10,–13,–14,–17,der(18)t(2;18)(p24;q11.2)ins(18;17)(q11.2;q12q25),–21, der(22)t(3;22)(q13.2;p11.2)t(4;22)(q31.2;q13)ins(22;1)(q13;p32p36.1),del(22)(q11.2)[cp12]
Panc 1.28
76⬃105<4n>,XX,–X,–X,+3,t(3;14)(p21;q22)del(14)(q22q32)!3,+5,+5,der(5)t(5;17)(p15.1;q21)!2, del(6)(p22)!2,t(8;11)(q13;p15),–9,+der(11)t(8;11)(q13;p15)!2,+12,+i(13)(q10),der(14;15)(q10;q10)!3, –17,–18,–19,+20,+20,+20,1⬃3 dmin[cp4]/46,XX[3]
Panc 2.05
118⬃123<4n>,XXXX,del(1)(q25)!2,+ider(1)(?)del(1)(?)ins(1;16)(?;?),+2,+2,del(3)(p21),+del(4)(q21), der(4)t(4;12)(p12;p11.2)!2,+der(5)t(1;5)(q25;q33)!2,+der(6)t(3;6)(q21;q15)!2,–7,+del(7)(q11.2)!2, +del(8)(q21.2)!2,dup(8)(q?),+11,+11,+del(12)(q14)!2,+13,+13,+i(13)(q10)!2,–14,+i(14)(q10)!2,+der(17;18)(p10;q10)!2, del(18)(q21)!2,+der(19)t(8;19)(q22;p12)!2,+20,+20,+20,+20,+21,+21,i(22)(q10)!2[cp5]
Panc 2.43
39⬃67<2n>,X,+X,der(Y)t(Y;13)(q11.2;q13),+1,+dup(1)(q12),+2,+del(3)(p12p21)x2,+4,+t(7;16)(q36;q11.2)!2, +inv(11)(q13p15),+del(15)(q21)!2,+20,+20,–22,i(22)(q10)[cp2]/91⬃120<4n>,idem!2[cp4]
Panc 2.02
65⬃71<3n>,XX,–X,–2,+5,+7,r(8)del(8)(?),–9,–10,der(10)t(8;10)(q12;p15)x2,+del(12)(q15),+der(14)t(3;14)(?;p11.2), ider(15)(q10)del(15)(q22),psu dic(15;18)(p11.2;q22)!2,der(17)t(2;17)(?;q25)del(17)(p11.2),–18, der(18)t(10;18)(q11.2;q11.1)!2,+der(20)t(8;20)(?;q11.2),+21,der(22)t(10;22)(?;p11)[cp6]
Panc 2.08
87⬃96,X,der(X)t(X;11)(p11.2;?),–Y,–Y,–1,der(1)t(1;14)(p11;q11.1)!2,–2,der(6)t(6;15)(q15;q24)x2,+9,+9,–10,+11, –14,–14,der(17)t(11;17)(q13;p11.2)!2,del(18)(q21)!2,+20,+20,+20,–21,–22[cp5]
a
Karyotype established by G-banding only.
and inv(11)(q13p15) in line 2.43. Unbalanced translocations were much more common. One apparently recurrent aberration was observed, der(18)t(18; 22)(q11.2;q11.2) in three cell lines. We further characterized these derivative chromosomes 18 using YACs (see below). SKY has been utilized previously to characterize only a few pancreatic cancer cell lines. SKY analysis of cell lines AsPC1, BxPC3, Capan 2, MiaPaCa2, PANC1, and CFPAC have been studied by two groups (Ghadimi et al., 1999; Sirivatanauksorn et al., 2001). Cell lines Hs766t, A18.1, FA6, MDA Panc3, PaTu1, PaTuII, QGP1, RO, RWP, SUIT2, SW979, and T3M have been analyzed by Sirivatanauksorn et al. (2001); Capan1, and Su86.86 have been reported by Ghadimi et al. (1999). All have found multiple complexly rearranged chromosomes.
Of 144 chromosomal aberrations identified in the six cell lines that Ghadimi et al. (1999) analyzed by both SKY and CGH, only six were balanced aberrations. The only recurrently involved bands were 7q21 (in two translocations) and 7q31 (in three). Of 344 chromosomal aberrations identified in the 20 cell lines analyzed by SKY by Sirivatanauksorn et al. (2001), 15 recurrent aberrations were found, all unbalanced. Eight of these were isochromosomes, including i(5)(p10) (in six cell lines), i(12)(p10) (n = 4), i(1)(q10) (n = 3), i(14)(q10) (n = 2), i(8)(p10) (n = 2), i(18)(p10) (n = 2), i(19)(q10) (n = 2), i(21)(q10) (n = 2), one was a Robertsonian translocation der(13; 15) (q10;q10) (n = 2), and the remainder were interpreted as terminal deletions, including del(11)(q23) (n = 5), del(7)(q22) (n = 3), del(10)(p11) (n = 3), del(1)(p22) (n = 2), del(17)(q21) (n = 2), del(18)(q21) (n = 2).
Cytogenet Genome Res 118:148–156 (2007)
151
Fig. 3. SKY metaphase illustrating multiple unbalanced rearrangements in cell line 2.08. (a) der(X)t(X;11)(p11.2;?); (b) der(1)t(1; 14) (p11;q11.2)!2; (c) der(6)t(6; 15)(q15;q24)!2; (d) der(17)t(11; 17)(q13; p11.2)!2; (e) del(18)(q21)!2.
Chromosome region 8p in pancreatic cancer Loss of regions of the short arm of chromosome 8 is common in epithelial tumors (Birnbaum et al., 2003). Metaphase cytogenetics has identified unbalanced translocations with loss of material distal to 8p12, and CGH studies in breast and other cancers have also shown loss of distal 8p material (reviewed in Pole et al., 2006). We found loss of 8p in nine cell lines by CGH. SKY identified seven derivative chromosomes involving 8p; these were formed as a result of an unbalanced translocation at band p21 or p22 in three lines, as an unbalanced translocation involving an unidentified portion of chromosome 8 in one line, and as an apparent terminal deletion in three lines. We used YACs to cover band 8p21 (between approximately 15 and 33 Mb from pter) (see Fig. 4a) to begin to investigate if a common breakpoint was involved. These results are summarized in Table 3. Two of the derivative chromosomes 8 retained 888d12, (the most proximal YAC at 33 Mb), while the remainder lacked signal from all four YACs tested, suggesting that the breakpoint was proximal to 888d12. Localization of the 8p breakpoints will require further mapping with additional probes. Adelaide et al. (2003) found recurrent chromosome translocation breakpoints involving the NRG1 gene at 8p12 in 4/34 breast and 2/9 pancreas cancer cell lines. The pancreas lines were PaTu1 and SUIT2. Recently, analysis of 8p rearrangements in 48 breast, pancreatic and colon cancer cell lines using FISH and array CGH with a tiling path of 0.2 Mb resolution over 8p12 and 1 Mb resolution over chromosome 8 was reported (Pole et al., 2006). Included in this study were nine rearrangements of 8p in seven pancreas cancer cell lines. They showed breakage and loss between 20–30 Mb from pter in two lines, at approximately 32 Mb in
152
Cytogenet Genome Res 118:148–156 (2007)
Fig. 4. Location of the YACs used to characterize breakpoints. ISCN G-bands listed to left of the ideogram, and megabases from pter to the right. YAC locations are approximate, based on information from STS markers described in CEPH/Genathon, as located on UCSD Build March 2006. (a) chromosome 8; (b) chromosome 18.
two lines, between 29 and 42 Mb in two lines, and between 40–45 Mb in three lines. From their overall study, they concluded that the complexity of 8p rearrangements includes various genes proximal to 31 Mb, involving both an amplicon of ZNF703/FLJ14299, and NRG1. Chromosome region 18q in pancreatic cancer Monosomy of chromosome 18 has been found YACarepeatedly in pancreas cancer. Metaphase CGH studies have
Cytogenet Genome Res 118:148–156 (2007)
153
–
ND
– –
8p
der(8)t(8;14) (p12;q13)!2
Panc 9.06
– – +
ND
8p12–qter
der(10)t(8;10) (q12;p15)!2, der(20)t(8;20)(?;q11.2)
Panc 2.02
– – + + – – – – +
+ ND + + + + + ND +
18q10–q22
+ + + + + + + + +
18q
der(18)t(18;22) der(18)ins(6;18) (q11.2;q11.2) (p12;p11.3q12) dup(22)(q?)!2 del(6)(p12q25)
Panc 2.13
– ND + + ND ND + ND ND
18
der(18)t(18;22) (q11.2;q11.2)
Panc 9.06
+ = YAC signal seen; – = no YAC signal seen on derivative chromosome; ND = not determined.
+ + + + + + + + +
18q
Region of 18 loss (CGH)
YACa 881h1 815a8 883b8 874b12 762b8 908a5 946c11 750a5 770c4
der(12)t(12;18) der(18)t(18;22) (p11.2 ;q12)!2 (q11.2;q11.2)
Partial karyotype (18s)
Panc 2.03
+ + – – – + – – –
18q
– – – – – – – – –
del(18)(q21)!2 der(17;18) (p10;q10)!2 (note: FISH suggests this may be iso18p)
Panc 2.05
– + + + + + + + +
18
der(18)t(10;18) (q11.2;q11.1)!2
Panc 2.02
–
Panc 8.13
ND
– –
8p
del(8)(p12), +der(8)t(8;16) (q22;q11.1)
Panc 8.13
+ + + + + + – + +
18q
+ + – – – + + + –
18q
+ + + – – – – – –
18q
del(18) der(18)t(2;18) der(20) (q21)!2 (p24;q11.2) t(18;20) ins(18;17) (p11.2;p11.1)!2 (q11.2;q12q25)
Panc 2.08 Panc 10.05
Table 4. Summary of YAC analysis of derivative chromosomes 18 in nine cell lines. All YACs are located in 18q11.2 except 881h1 in 18p11.3 (at 7 Mb).
Panc 6.03
a
– – – +
8p21
del(8)(p21)
Panc 2.13
+ = YAC signal seen; – = no YAC signal seen on derivative chromosome; ND = not determined.
Chromosome description
a
ND
–
ND
– –
8p
– –
8p21–pter
Region of 8p loss (CGH)
der(8)t(8;18) (p21;q23)!2, dup(8)(q23q24.3)!2
Panc 2.03
ND
del(8)(p21)!2, der(8)t(8;22) (p22;q13)
Partial karyotype (8s)
YACa 946c9 755b1 872e4 888d12
Panc 6.03
Chromosome description
Table 3. Summary of YAC analysis of derivative chromosomes 8 in six cell lines. All YACs are located in 8p21.
a
b
Fig. 5. The extreme complexity of a metaphase from line 2.43. (a) SKY, (b) reverse DAPI banding of same cell.
shown that partial losses of 18q are also common (Fukushige et al., 1997; Mahlamaki et al., 1997, 2002; Curtis et al., 1998; Gahdimi et al., 1999; Schleger et al., 2000; Harada et al., 2002). This has included loss of the distal half of 18q, including genes DCC and SMAD4. Importance of 18q in the biology of pancreas cancer has been demonstrated repeatedly since early G-banded cytogenetic studies first indicated its frequent loss in this neoplasm (Johansson et al., 1992; Bardi et al., 1993; Griffin et al., 1994, 1995). Identification of DPC4/SMAD4, at 18q21.1 as homozygously deleted in pancreas cancer was first found in 64% of pancreas cancers studied (Hahn et al., 1996), and mutations and loss of that gene have been reported numerous times since then (reviewed in Furukawa et al., 2006). DPC4/SMAD4 is located at 46.8 Mb in NCBI Build 36.1. CGH analysis of the 15 cell lines reported here showed loss of chromosome 18 material to be frequent. Defining breakpoints by metaphase CGH, however, is imprecise. Our G-banding and SKY studies identified eleven derivative chromosomes 18 in nine cell lines. We used seven YACS to further investigate the possibility of involvement of a specific region in 18q11.2. These YACs map between approximately 18 and 24 Mb from 18pter (see Fig. 4b). As can be seen in Table 4, presence of specific YACs on the different derivative chromosomes 18 varied, suggesting that rearrangements in this region are complex. However, our data do not exclude the possibility of one or more specific breakpoints which might be identified using more sensitive mapping techniques in that area. Several other investigators have described 18q breakpoint mapping in pancreas cancer. Hoglund et al. (1998) investigated breaks in 18q in 13 primary specimens of pancreatic carcinoma studied after only a few cell passages. For proximal 18q, they used FISH with YAC 766f9 (localized to 18q11.2) and a partial chromosome paint involving 18q11.1 to show that loss of 18q often was proximal to YAC 766f9. Using current databases we estimate 766f9 is located about 26.4 Mb from pter, approximately 2 Mb centromeric to the
154
Cytogenet Genome Res 118:148–156 (2007)
most proximal YAC we used. Alsop et al. (2006) studied breakpoints on chromosome 18 in nine pancreas cancer cell lines using BACs from an RPC11 library. They found breakpoints in the centromere region in four lines. The breakpoint was in proximal q11 in one line, within the region bounded by BACs 296E23 and 459H24 (18.7–20.4 Mb). In two other lines, breaks were between BACs 459H24 and 5G23 (20.4–23.2 Mb). One line had breaks within 289A1 (28.3–28.5 Mb) and another between 289A1 and 459B18 (29.0–29.2 Mb). Our results, in combination with these, suggest that breakpoints within the proximal 18q region are common in pancreas cancer. Very recently, since we completed our data, array CGH has been reported on a number of pancreas cancer cell lines (Aguirre et al., 2004; Heidenblad et al., 2004; Holzmann et al., 2004; Mahlamaki et al., 2004; Bashyam et al., 2005; Gysin et al., 2005; Nowak et al., 2005). Some of the studies have included some of the cell lines we report here. As expected, in addition to confirming regions of gain and loss identified by metaphase CGH analyses, the higher level of resolution attained by BAC and cDNA arrays has identified small regions of genomic gain and loss not previously detected. Regarding 18q, loss of both proximal and distal 18q have been found. However, none used a tiling path array on 18q. Results also emphasize those deletions which are most frequent and usually homozygous. Aguirre et al. (2004) using a cDNA array with average coverage of 1 Mb, found deletion on 18q. The peak boundary of the most proximal deletion locus was 34.95–40.58 Mb, but is listed as extending from 18.51 to 46.28; they also detected deletion at 18q22.1]q23 from 60.4 to 77.63, peak at 74.45–76.84. Bashyam et al. (2005), using a cDNA array with average resolution of 60 kb, found the most proximal 18q deletion at 18q21 (46.7–46.8 Mb) with suggested candidate gene WWOX. Gysin et al. (2005) used a BAC array with average of 1.4 Mb coverage and identified loss at 18q21.1, including DPC4. Heidenblad et al. (2004) using cDNA and BAC arrays with 1 Mb coverage also found homozygous de-
letion in SMAD4. Nowak et al. (2005) used a BAC array with average 420 kb coverage; deletion of 18q11.21]q23 was found as a recurrent region of loss. Interestingly, the 18q11 region has also been identified as amplified in two related cell lines PaTu8988T and PaTu8998S in two studies. Heidenblad et al. (2004) found the amplicon at 18–20 Mb using a 1 Mb BAC array. Holzmann et al. ( 2004), using a 15 Mb BAC array, also found amplification in PATU 8998 at 18q11.2 and suggested LAMA3 as a candidate gene in that amplicon. Amplification at 18q11.1]q11.2 at 16.98–18.86 Mb was found in cell line LPC6 by Heidenblad et al. (2004). This is similar to the rearrangements of 8p12 studied by Pole et al. (2006), who noted both deletions and an amplicon in 8p12. There is reason to think that proximal 18q may harbor a tumor suppressor gene. Lefter et al. (2002) used microcellmediated transfer of a normal copy of chromosome 18 into pancreas carcinoma cell lines. They observed suppressed growth of hybrid cells in culture and in nude mice, compared to the parental cells, regardless of the initial DPC4/ SMAD status of the cells, leading them to conclude that SMAD4 was not the only tumor suppressor involved on
chromosome 18. Sunamura et al. (2004) repeated this finding and used an expression array to determine that four genes related to apoptosis (not named) were upregulated in the hybrid cells. One of the difficulties in studying pancreatic cancer cell lines is the possibility of continuing genetic instability and development of subclones. During our SKY analyses we occasionally encountered metaphases with significant genomic disarray (Fig. 5) and in evaluating 18q breakpoints we not infrequently found cells that did not contain the derivative chromosome being analyzed. Nonetheless, most described experiences with these lines have found them to be relatively stable. Indeed, of the five cell lines studied by aCGH by Nowak et al. (2005) for which we had performed metaphase CGH several years earlier, there was strong concurrence of measurement of 18q deletions. In summary, these pancreas cancer cell lines contain significant chromosomal complexity. Data derived from CGH and SKY analyses confirm and extend the genomic gains and losses first identified by metaphase cytogenetics, and help to elucidate the chromosomal structural alterations which result in these changes.
References Adelaide J, Huang HE, Murati A, Alsop AE, Orsetti B, et al: A recurrent chromosome translocation breakpoint in breast and pancreatic cancer cell lines targets the neuregulin/NRG1 gene. Genes Chromosomes Cancer 37: 333–345 (2003). Aguirre AJ, Brennan C, Bailey G, Sinha R, Feng B, et al: High-resolution characterization of the pancreatic adenocarcinoma genome. Proc Natl Acad Sci USA 101:9067–9072 (2004). Alsop AE, Teschendorff AE, Edwards PA: Distribution of breakpoints on chromosome 18 in breast, colorectal, and pancreatic carcinoma cell lines. Cancer Genet Cytogenet 164: 97–109 (2006). Bardi G, Johansson B, Pandis N, Mandahl N, BakJensen E, et al: Karyotypic abnormalities in tumours of the pancreas. Br J Cancer 67: 1106– 1112 (1993). Bashyam MD, Bair R, Kim YH, Wang P, Hernandez-Boussard T, et al: Array-based comparative genomic hybridization identifies localized DNA amplifications and homozygous deletions in pancreatic cancer. Neoplasia 7:556–562 (2005). Birnbaum D, Adelaide J, Popovici C, Charafe-Jauffret E, Mozziconacci MJ, Chaffanet M: Chromosome arm 8p and cancer: a fragile hypothesis. Lancet Oncol 4: 639–642 (2003). Curtis LJ, Li Y, Gerbault-Seureau M, Kuick R, Dutrillaux AM, et al: Amplification of DNA sequences from chromosome 19q13.1 in human pancreatic cell lines. Genomics 53: 42–55 (1998). Fukushige S, Waldman FM, Kimura M, Abe T, Furukawa T, et al: Frequent gain of copy number on the long arm of chromosome 20 in human pancreatic adenocarcinoma. Genes Chromosomes Cancer 19: 161–169 (1997).
Furukawa T, Sunamura M, Horii A: Molecular mechanisms of pancreatic carcinogenesis. Cancer Sci 97:1–7 (2006). Ghadimi BM, Schrock E, Walker RL, Wangsa D, Jauho A, et al: Specific chromosomal aberrations and amplification of the AIB1 nuclear receptor coactivator gene in pancreatic carcinomas. Am J Pathol 154: 525–536 (1999). Griffin CA, Hruban RH, Long PP, Morsberger LA, Douna-Issa F, Yeo CJ: Chromosome abnormalities in pancreatic adenocarcinoma. Genes Chromosomes Cancer 9: 93–100 (1994). Griffin CA, Hruban RH, Morsberger LA, Ellingham T, Long PP, et al: Consistent chromosome abnormalities in adenocarcinoma of the pancreas. Cancer Res 55: 2394–2399 (1995). Gysin S, Rickert P, Kastury K, McMahon M: Analysis of genomic DNA alterations and mRNA expression patterns in a panel of human pancreatic cancer cell lines. Genes Chromosomes Cancer 44: 37–51 (2005). Hahn SA, Hoque AT, Moskaluk CA, da Costa LT, Schutte M, et al: Homozygous deletion map at 18q21.1 in pancreatic cancer. Cancer Res 56: 490–494 (1996). Harada T, Okita K, Shiraishi K, Kusano N, Furuya T, et al: Detection of genetic alterations in pancreatic cancers by comparative genomic hybridization coupled with tissue microdissection and degenerate oligonucleotide primed polymerase chain reaction. Oncology 62: 251– 258 (2002). Harada T, Okita K, Shiraishi K, Kusano N, Kondoh S, Sasaki K: Interglandular cytogenetic heterogeneity detected by comparative genomic hybridization in pancreatic cancer. Cancer Res 62: 835–839 (2002). Heidenblad M, Schoenmakers EF, Jonson T, Gorunova L, Veltman JA, et al: Genome-wide array-based comparative genomic hybridization reveals multiple amplification targets and novel homozygous deletions in pancreatic carcinoma cell lines. Cancer Res 64: 3052–3059 (2004).
Hoglund M, Gorunova L, Jonson T, Dawiskiba S, Andren-Sandberg A, et al: Cytogenetic and FISH analyses of pancreatic carcinoma reveal breaks in 18q11 with consistent loss of 18q12– qter and frequent gain of 18p. Br J Cancer 77: 1893–1899 (1998). Holzmann K, Kohlhammer H, Schwaenen C, Wessendorf S, Kestler HA, et al: Genomic DNAchip hybridization reveals a higher incidence of genomic amplifications in pancreatic cancer than conventional comparative genomic hybridization and leads to the identification of novel candidate genes. Cancer Res 64: 4428– 4433 (2004). ISCN (2005): An International System for Human Cytogenetic Nomenclature. Shaffer LG, Tommerup N (eds) (S Karger, Basel 2005). Jaffee EM, Schutte M, Gossett J, Morsberger LA, Adler AJ, et al: Development and characterization of a cytokine-secreting pancreatic adenocarcinoma vaccine from primary tumors for use in clinical trials. Cancer J Sci Am 4:194–203 (1998). Johansson B, Bardi G, Heim S, Mandahl N, Mertens F, et al: Nonrandom chromosomal rearrangements in pancreatic carcinomas. Cancer 69: 1674–1681 (1992). Karhu R, Mahlamaki E, Kallioniemi A: Pancreatic adenocarcinoma – genetic portrait from chromosomes to microarrays. Genes Chromosomes Cancer 45: 721–730 (2006). Kitoh H, Ryozawa S, Harada T, Kondoh S, Furuya T, et al: Comparative genomic hybridization analysis for pancreatic cancer specimens obtained by endoscopic ultrasonography-guided fine-needle aspiration. J Gastroenterol 40: 511– 517 (2005). Lefter LP, Furukawa T, Sunamura M, Duda DG, Takeda K, et al: Suppression of the tumorigenic phenotype by chromosome 18 transfer into pancreatic cancer cell lines. Genes Chromosomes Cancer 34: 234–242 (2002).
Cytogenet Genome Res 118:148–156 (2007)
155
Mahlamaki EH, Hoglund M, Gorunova L, Karhu R, Dawiskiba S, et al: Comparative genomic hybridization reveals frequent gains of 20q, 8q, 11q, 12p, and 17q, and losses of 18q, 9p, and 15q in pancreatic cancer. Genes Chromosomes Cancer 20: 383–391 (1997). Mahlamaki EH, Barlund M, Tanner M, Gorunova L, Hoglund M, et al: Frequent amplification of 8q24, 11q, 17q, and 20q-specific genes in pancreatic cancer. Genes Chromosomes Cancer 35: 353–358 (2002). Mahlamaki EH, Kauraniemi P, Monni O, Wolf M, Hautaniemi S, Kallioniemi A: High-resolution genomic and expression profiling reveals 105 putative amplification target genes in pancreatic cancer. Neoplasia 6: 432–439 (2004). Nowak NJ, Gaile D, Conroy JM, McQuaid D, Cowell J, et al: Genome-wide aberrations in pancreatic adenocarcinoma. Cancer Genet Cytogenet 61:36–50 (2005).
156
Pole JC, Courtay-Cahen C, Garcia MJ, Blood KA, Cooke SL, et al: High-resolution analysis of chromosome rearrangements on 8p in breast, colon and pancreatic cancer reveals a complex pattern of loss, gain and translocation. Oncogene 25: 5693-5706 (2006). Riopel MA, Spellerberg A, Griffin CA, Perlman EJ: Genetic analysis of ovarian germ cell tumors by comparative genomic hybridization. Cancer Res 58:3105–3110 (1998). Schleger C, Arens N, Zentgraf H, Bleyl U, Verbeke C: Identification of frequent chromosomal aberrations in ductal adenocarcinoma of the pancreas by comparative genomic hybridization (CGH). J Pathol 91: 27–32 (2000).
Cytogenet Genome Res 118:148–156 (2007)
Shiraishi K, Okita K, Kusano N, Harada T, Kondoh S, et al: A comparison of DNA copy number changes detected by comparative genomic hybridization in malignancies of the liver, biliary tract and pancreas. Oncology 60: 151–161 (2001). Sirivatanauksorn V, Sirivatanauksorn Y, Gorman PA, Davidson JM, Sheer D, et al: Non-random chromosomal rearrangements in pancreatic cancer cell lines identified by spectral karyotyping. Int J Cancer 91: 350–358 (2001). Solinas-Toldo S, Wallrapp C, Muller-Pillasch F, Bentz M, Gress T, Lichter P: Mapping of chromosomal imbalances in pancreatic carcinoma by comparative genomic hybridization. Cancer Res 56:3803–3807 (1996). Sunamura M, Lefter LP, Duda DG, Morita R, Inoue H, et al: The role of chromosome 18 abnormalities in the progression of pancreatic adenocarcinoma. Pancreas 28: 311–316 (2004).
n ai Se
r
-A
D
m do
Pr o
H HL
LZ
b-
D AA N–
–C
MiTF
–C
N–
TFE3
–C TFEB
N– –C
N–
TFEC
Fig. 1. Schematic representation of the MiT family of transcription factors. Depicted are the acidic activation domain (AAD), the basic helix-loop-helix/leucine zipper domain (bHLH-LZ), the proline-rich activation domain (Pro-AD) and the serine-rich domain (Ser).
tures. In addition, these tumors are characterized by the occurrence of recurrent balanced chromosome translocations, which result in disruption and/or fusion of members of the MiT family of basic helix-loop-helix/leucine-zipper transcription factor genes. Hence this subgroup has been denoted the MiT translocation subgroup of RCCs (Argani et al., 2005; Ramphal et al., 2006). The MiT translocation subgroup of RCCs
A first case of RCC and a t(X;1)(p11;q21) occurring in a two-year-old child was described by de Jong et al. (1986). Subsequently, several reports of additional RCC translocation cases appeared in the literature (Meloni et al., 1993; Shipley et al., 1995; Tonk et al., 1995; Dal Cin et al., 1998; Dijkhuizen et al., 1998; Kardas et al., 1998; Yenamandra et al., 1998; Desangles et al., 1999; Perot et al., 1999; ZattaraCannoni et al., 2000; Ramphal et al., 2006), including variants (Tomlinson et al., 1991; Ohjimi et al., 1993; Dijkhuizen et al., 1995; Hernandez-Marti et al., 1995; Zhao et al., 1995; Clark et al., 1997; Ramphal et al., 2006). We and others found that, as a result of the t(X;1)(p11;q21) translocation, the transcription factor TFE3 gene on the X chromosome is fused to the PRCC gene on chromosome 1 (Sidhar et al., 1996; Weterman et al., 1996a, b). Subsequently, it was found that in Xp11-associated variant translocations the TFE3 gene is disrupted and fused to either the PSF (SFPQ) gene on chromosome 1 (Clark et al., 1997), the NONO gene on the X chromosome (Clark et al., 1997), or the ASPL (ASPSCR1) gene on chromosome 17 (Heimann et al., 2001; Argani et al., 2001a). The latter gene fusion has also been observed in alveolar soft part sarcomas carrying a t(X;17)(p11;q25) translocation (Ladanyi et al., 2001). In addition, fusion of the TFE3 gene to the clathrin heavy chain gene (CLTC) gene on chromosome 17 has been reported in a case of pediatric RCC and a t(X;17)(p11;q23) translocation (Argani et al., 2003). Another recurrent chromosomal translocation, t(6; 11) (p21;q13), has been encountered in a clinically and histologically similar subgroup of early-onset RCCs (Dijkhuizen et al., 1996; van Asseldonk et al., 2000; Argani et al., 2001b).
158
Cytogenet Genome Res 118:157–165 (2007)
We, and others, found that through this translocation the transcription factor TFEB gene on chromosome 6 is fused to the intron-less non-protein-coding Alpha gene on chromosome 11 (Davis et al., 2003; Kuiper et al., 2003). Except for a case report of an 8-year-old patient with RCC and a t(6; 17)(p21;q24⬃q25) translocation (Dal Cin et al., 1991), a likely candidate for a TFEB fusion, as yet no variant 6p21 translocations and/or gene fusions have been described. TFE3 and TFEB are two closely related members of the so-called MiT family of basic helix-loop-helix/leucine-zipper (bHLH-LZ) transcription factors (also referred to as the MiTF/TFE family) (Fig. 1), which also includes TFEC and MiTF (Hemesath et al., 1994). The four members of this family can bind to the consensus hexanucleotide E-box sequence CA[C/T]GTG either as homo- or heterodimers. The MiT family members can form these dimers with each other but not with other basic helix-loop-helix (bHLH) transcription factors (Hemesath et al., 1994). In addition, TFE3 and TFEB share several downstream target genes. In accordance with this notion, TFE3 and TFEB were found to be functionally redundant in tissues in which they are co-expressed (Steingrimsson et al., 2002; Huan et al., 2006). This functional redundancy among MiT members was further supported by the recent finding that TFE3 and TFEB are able to rescue MiTF deficiency in clear cell sarcomas and, in reverse, that MiTF is able to rescue TFE3 deficiency in pediatric RCCs (Davis et al., 2006). It is assumed that anomalous expression of either TFE3 or TFEB, as observed in both Xp11- and 6p21-associated translocation RCCs, may result in the deregulation of similar oncogenic pathways. Also clinically and histologically, the RCCs of this subgroup share several characteristics, including poorly developed papillary structures and nests of pleomorphic granular clear to eosinophilic cells. These cells exhibit a markedly low expression of cytokeratins and, in contrast, a clearly up-regulated expression of the melanocytic markers HMB45 and Melan A (Argani and Ladanyi, 2003; Argani, 2006), both of which are well-known targets of MiTF (Du et al., 2003). Although these RCCs are sometimes encountered in elderly patients as well, in the majority of patients they present before the age of 40, i.e., the median age of patients with a t(X;1)(p11;q21)-positive RCC was found to be 22 years, and all patients with a t(X;17)(p11;q25)-positive RCC were found to be children with a median age of 7 years (Perot et al., 2003). Also the t(6; 11)(p21;q13)-positive RCCs present at a relatively young age, i.e., ten cases have been reported in predominantly adolescents with a median age of 18 years (Dijkhuizen et al., 1996; van Asseldonk et al., 2000; Argani et al., 2001b, 2006; Kuiper et al., 2003; Davis et al., 2003; Ramphal et al., 2006) (Table 1). Role of TFE3 gene fusions in RCC development
The t(X;1)(p11;q21) translocation results in a fusion of the TFE3 gene on the X chromosome with the PRCC gene on chromosome 1 (Sidhar et al., 1996; Weterman et al., 1996a, b). PRCC is a ubiquitously expressed nuclear protein,
Gene fusion
No. of cases
Median age
Referencesa
t(X;1)(p11;q21) t(X;1)(p11;p34) inv(X)(p11;q12) t(X;17)(p11;q25) t(X;17)(p11;q23) t(6;11)(p21;q13)
PRCC-TFE3 PSF-TFE3 NONO-TFE3 ASPL-TFE3 CLTC-TFE3 Alpha-TFEB
20 6 1 8 1 11b
22 31 39 7 14 18
1–14 13, 15–18 17 13, 19–23 24 25–27
a
1: de Jong et al., 1986; 2: Meloni et al., 1993; 3: Tonk et al., 1995; 4: Sidhar et al., 1996; 5: Shipley et al., 1995; 6: Yenamandra et al., 1998; 7: Kardas et al., 1998; 8: Dal Cin et al., 1998; 9: Dijkhuizen et al., 1998; 10: Perot et al., 1999; 11: Desangles et al., 1999; 12: Zattara-Cannoni et al., 2000; 13: Perot et al., 2003; 14: Onder et al., 2006; 15: Kovacs et al., 1987; 16: Dijkhuizen et al., 1995; 17: Clark et al., 1997; 18: Yoshida et al., 1985; 19: Tomlinson et al., 1991; 20: Hernandez-Marti et al., 1995; 21: Heimann et al., 2001; 22: Argani et al., 2001a; 23: Carcao et al., 1998; 24: Argani et al., 2003; 25: Argani et al., 2001b; 26: Argani et al., 2005; 27: Kuiper et al., 2003. b One additional case is described in this review.
which has a relatively high proline content (Sidhar et al., 1996; Weterman et al., 1996a). The fusion of TFE3 and PRCC results in two reciprocal in-frame fusion genes, TFE3-PRCC and PRCC-TFE3. These fusion genes are both expressed in t(X;1)(p11;q21)-positive tumor cells (Weterman et al., 1996a). The fusion protein PRCC-TFE3, which contains both the acidic activation domain (AAD) and the C-terminal proline-rich activation domain (pro-AD) of TFE3 (Fig. 2), has been shown to be a more potent transcriptional activator than wild-type TFE3 (Weterman et al., 2000). PRCC-TFE3 was also found to exhibit transformation-associated properties, including growth under low-serum conditions, anchorage-independent growth in soft agar, and tumor formation in athymic nude mice, after exogenous expression in non-tumorigenic NIH3T3 and/or Rat1 cells (Weterman et al., 2001a). In addition, we found that PRCC-TFE3 can overcome growth arrest and differentiation induction in conditionally immortalized kidney cells derived from the proximal tubules of H-2KbtsA58 transgenic mice (Weterman et al., 2001a). Additionally, it was found that PRCC can interact with the mitotic arrest deficient protein Mad2B (also referred to as Mad2L2) and, through this interaction, shuttle this protein to the nucleus where it can exert its normal physiological function (Weterman et al., 2001b). Mad2B is a member of the MAD family of proteins and shows 48% similarity to Mad2 (Cahill et al., 1999; van den Hurk et al., 2004), a protein that is known to play a critical role in mitotic spindle checkpoint control (Li and Murray, 1991). The PRCC-TFE3 fusion protein, however, has lost the capacity to bind Mad2B and, consequently, the ability to shuttle Mad2B to the nucleus, despite retention of the PRCC-Mad2B interaction domain (Weterman et al., 2001b). We speculate that this loss of function may impose a (loss of) cell cycle control upon t(X;1)(p11;q21)-positive RCCs.
D
H LZ LH do m ai Pr n oAD Se r
AA
Translocation
b-
Table 1. Recurrent translocations and MiT fusion genes in 45 RCC cases reported in the literature N–
N– N– N– N– Pro-Glu
Pro
–C
TFE3
–C
PRCCTFE3
–C
ASPLTFE3
–C
CLTCTFE3
–C
PSFTFE3
–C
NonOTFE3
RBD
N– Pro-His RBD
Fig. 2. Schematic representation of TFE3 fusion products. Depicted are the acidic activation domain (AAD), the basic helix-loop-helix/leucine zipper domain (bHLH-LZ), the proline-rich activation domain (Pro-AD) and the serine-rich domain (Ser) of TFE3, and the prolineglutamine-rich domain (Pro-Glu), the glutamine-histidine-rich domain (Glu-His) and the RNA-binding domain (RBD) of PSF and NonO, respectively.
As a result of the t(X;1)(p11;p34) translocation, the TFE3 gene on the X chromosome and the PSF gene on chromosome 1 are fused (Fig. 2) (Clark et al., 1997). Unlike wildtype TFE3 or PSF , which are both nuclear proteins, PSFTFE3 is targeted to the endosomal compartment. Although PSF-TFE3 appears to have no dominant effect on the nuclear localization of wild-type PSF it sequesters TFE3 and p53 in the extranuclear compartment, which leads to functionally null p53 and TFE3 cells (Mathur and Samuels, 2007). The splicing factor PSF is a component of the spliceosome complex which catalyses exon joining and has been shown to be required for catalytic step II of the splicing reaction (Patton et al., 1993; Gozani et al., 1994). The functional loss of p53 and/or TFE3, most likely, contributes to the transformed phenotype through interference with cell cycle control. An X chromosomal inversion, inv(X)(p11;q12), results in fusion of the TFE3 gene on the p-arm to the NONO gene on the q-arm (Clark et al., 1997). Functionally, NONO is closely related to PSF, both of which are members of a family of DBHS (Drosophila behavior and human splicing) proteins (Dong et al., 1993; Yang et al., 1993). These proteins are thought to play an important role(s) in the proper splicing of pre-mRNAs. As yet, its putative role in RCC development remains to be established. The NONO-TFE3 fusion transcript encodes a protein in which almost the entire NONO protein is fused to the C-terminal moiety of TFE3, including its basic DNA-binding domain and the C-terminal prolinerich activation domain (Clark et al., 1997) thus enabling this fusion protein to act, like PRCC-TFE3, as a transcriptional activator. As a result of the t(X;17)(p11;q25) translocation, the TFE3 gene on the X chromosome is fused to the ASPL gene on chromosome 17 (Fig. 2) (Argani et al., 2001a; Heimann et
Cytogenet Genome Res 118:157–165 (2007)
159
al., 2001). This translocation and its concomitant gene fusion are also encountered in alveolar soft part sarcomas, but in an unbalanced form (Ladanyi et al., 2001). Two mutually exclusive ASPL-TFE3 gene fusions have been observed in RCCs, resulting in two distinct ASPL-TFE3 fusion transcripts in which ASPL is fused to either exon 3 or exon 4 of the TFE3 gene (Argani et al., 2001a; Ladanyi et al., 2001). Interestingly, the acidic activation domain (AAD) of TFE3 is encoded by exon 3, thus suggesting that in at least a subset of the tumors anomalous transcriptional activation is brought about by the C-terminal activation domain. In a pediatric RCC with a t(X;17)(p11;q23) translocation the TFE3 gene on the X chromosome was found to be fused to the clathrin heavy chain gene CLTC on chromosome 17 (Fig. 2) (Argani et al., 2003). The predicted CLTC-TFE3 fusion protein retains the DNA-binding and C-terminal transactivation domains of TFE3, but lacks the multimerization domain of CLTC. Structurally, clathrin is a triskelion-shaped protein complex that is composed of a trimer of heavy chains each bound to a single light chain (Ybe et al., 1998, 1999). Clathrin is the major protein constituent of the coat that surrounds organelles to mediate selective protein transport (Smith and Pearse, 1999; Kirchhausen, 2000), including receptor-mediated endocytosis, intracellular trafficking and recycling. Recently, however, clathrin has also been found to stabilize fibers of the mitotic spindle thus facilitating the congression of chromosomes (Royle et al., 2005). As such, clathrin is thought to play an important role in cell cycle control. In conformity with this notion, it has been found that absence of clathrin at the mitotic spindle results in an increased frequency of misaligned chromosomes (Royle et al., 2005), which may lead to genetic instability and, ultimately, cancer. Interestingly, CLTC is also known as a recurrent fusion partner of the ALK tyrosine kinase gene in anaplastic large-cell lymphomas and inflammatory myofibroblastic tumors (Pulford et al., 2004). Role of Alpha-TFEB gene fusion in RCC development
Thus far, ten cases of t(6;11)(p21;q13) translocation-positive RCCs have been described in the literature, from seven of which the breakpoints have been molecularly cloned (Davis et al., 2003; Kuiper et al., 2003; Argani et al., 2005). In all these cases a fusion between the non protein-encoding Alpha gene on chromosome 11 and the TFEB gene on chromosome 6 was identified, with the breakpoints within the TFEB gene clustering in a 167-bp region within intron 1, just upstream of the translation initiation codon-containing exon 2 (Davis et al., 2003; Kuiper et al., 2003; Argani et al., 2005). As a result of promoter swapping, the Alpha-TFEB gene fusions were found to elicit a dramatic up-regulation of TFEB mRNA and protein levels, whereas the expression of the other MiT family members remained unchanged in the RCCs (Kuiper et al., 2003). In agreement with this notion, Argani et al. (2005) found a moderate to strong nuclear TFEB immunoreactivity in paraffin-embedded tissues of all tested RCCs carrying a t(6; 11)(p21;q13) translocation,
160
Cytogenet Genome Res 118:157–165 (2007)
whereas in 1,089 unrelated tumors no nuclear TFEB labeling could be observed. Therefore, strong nuclear expression of TFEB can be considered as a hallmark of this tumor subtype (Argani et al., 2005). Since in most t(6; 11)(p21;q13)positive RCCs this translocation is present as the sole cytogenetic anomaly, it is assumed that the up-regulation of TFEB expression directly triggers tumorigenesis (Davis et al., 2003; Kuiper et al., 2003). Interestingly, two renal oncocytomas with a similar t(6;11)(p21;q13) translocation were recently reported in the literature (Jhang et al., 2004). Using Western blot analysis, we failed to observe up-regulation of TFEB expression in these two unrelated tumors (unpublished data), whereas in a thus far unreported RCC case carrying a t(6;11)(p21;q13) translocation (all cases kindly provided by Dr. V. Murty, Columbia University, N.Y., USA), a strong TFEB immunoreactivity was found. Indeed, using RT-PCR we were able to demonstrate the presence of an Alpha-TFEB gene fusion in the RCC but not in the two unrelated oncocytomas. The latter translocation is, therefore, likely to involve different genes thus adding to the specificity of the Alpha-TFEB gene fusion for the subgroup of MiT translocation RCCs. Role of MiTF in tumor development
As for TFE3 and TFEB, accumulating evidence indicates that the MiT family member MiTF is also involved in tumor formation, i.e., MiTF deregulation has been shown to play a role in the development of human melanomas (McGill et al., 2002; Selzer et al., 2002; Widlund et al., 2002) and amplification of its gene has been observed in 20% of these tumors. As such, MITF may function as a melanoma-associated oncogene (Garraway et al., 2005). In addition, it was found that MITF amplification is most prevalent in melanoma patients with metastatic disease and that this amplification correlates with a poor prognosis (Garraway et al., 2005). Several MITF target genes have been identified which may be critical to melanoma development, including the cyclin-dependent kinase gene CDK2 (Du et al., 2004). Since CDK2 is known to play an important role in the regulation of cell cycle progression, it is assumed that also MiTF, through CDK2, may play such a role (Du et al., 2004). More recently, the hypoxia-inducible factor HIF1A was identified as a MiTF target gene in melanoma. Stimulation of HIF1A expression by MiTF facilitates melanoma survival, neo-vascularization and metabolic adaptation, thus contributing to the development of these tumors (Busca et al., 2005). Interestingly, HIF1alpha is normally bound to and destabilized by the VHL protein, a process that is known to be hampered in VHL-defective RCCs (Maxwell et al., 1999). Clear cell sarcoma, a soft tissue malignancy that expresses melanocytic markers, may harbor a t(12; 22)(q13;q12) translocation, which fuses the Ewing’s sarcoma gene (EWS) on chromosome 22 to the CREB family transcription factor ATF1 gene on chromosome 12, resulting in cAMP-independent transcriptional activation (Balaban et al., 1984; Zucman et al., 1993; Langezaal et al., 2001). It has been shown
that the EWSATF1 fusion protein occupies the MITF promoter and, by doing so, induces its expression and concomitant tumor growth (Davis et al., 2006). Tissue-specific regulation of MiT
It is generally believed that complex gene regulation in higher organisms is achieved through the generation of multiple proteins by single genes, generally realized by alternative pre-mRNA splicing (Lander et al., 2001; Modrek et al., 2001). Furthermore, recent studies have shown the frequent occurrence of alternative promoters and, at present, it is estimated that about 18% of all human genes exhibit alternative promoter usage (Trinklein et al., 2003). The MITF gene is known to make use of such alternative promoters, i.e., it encodes multiple transcripts that are each controlled by different, tissue-specific, promoters. The resulting MiTF isoforms (at least eight) do share the most important functional domains but differ in their first exons. In addition, alternative splicing of internal exons has been found (Steingrimsson et al., 1994; Hallsson et al., 2000; Udono et al., 2000; Shibahara et al., 2001). As for the TFEB and TFEC genes, new information on their genomic organization has recently become available, and it was found that also these two MiT genes contain multiple first exons (Kuiper et al., 2004). TFEB has at least seven distinct alternative first exons, which are all expressed in a tissuespecific manner. TFEC has at least three alternative first exons, which are also expressed in a tissue-specific manner. In the TFE3 and TFEB translocation tumors the first exons of these genes are replaced by other sequences, resulting in promoter swaps that abrogate the fine-tuned control of gene expression through alternative promoter usage. Most 5ⴕ exons of the TFEB variants are non-coding, and the resulting transcripts thus contain identical open reading frames that start in the translation initiation codon-containing exon 2. However, its mRNA variants may differ in their transcriptional and translational performance, as has previously been reported for other mRNAs (Phelps et al., 1998; Wang et al., 1999; Bonham et al., 2000; Medstrand et al., 2001; Kamat et al., 2002; Landry and Mager, 2002; Larsen et al., 2002; Saleh et al., 2002; Landry et al., 2003). In case of the TFEC gene, alternative 5ⴕ exon usage does result in the generation of different protein isoforms, i.e., a TFEC-A isoform which contains an extra N-terminal sequence and a TFEC-C isoform which is N-terminally truncated (Kuiper et al., 2004). Such N-terminally truncated isoforms may act as dominant-negative regulators (Yang et al., 1998; Pozniak et al., 2000). Furthermore, the TFEC gene may give rise to protein isoforms that result from alternative splicing of one or more exons encoding functional domains, including the Nterminal acidic activation domain encoded by exon 3 and/or the basic domain encoded by exon 5. In addition, several transcripts were identified in which intron 7 was retained, resulting in the formation of a C-terminally truncated protein that lacks the proline-rich activation domain and the C-terminal serine-rich domain (Kuiper et al., 2004).
Coexpression of MiT isoforms lacking functional domains has been observed for all members of the MiT family (Roman et al., 1991; Artandi et al., 1995; Kuiper et al., 2003), and this may drive the regulation of target gene expression, i.e., proteins lacking functional domains may associate with their full-length counterparts to form heterodimers, thereby imposing a dominant-negative effect upon transcription. A clear and well-investigated example is TFE3 of which a splice variant, that encodes a protein lacking the acidic activation domain AAD, can form a dimer with normal TFE3, thereby reducing its transactivating capacity (Roman et al., 1991; Artandi et al., 1995). A similar scenario has been reported for isoforms of TFEC and MiTF (Kuiper et al., 2004). The basic domain is essential for binding of the MiT members to their target DNA sequences. MiT isoforms lacking this basic domain may dimerize with normal isoforms and thus prevent DNA binding of the heterodimer, thereby functioning as negative regulators in a manner that has also been observed for the Id protein (Benezra et al., 1990). As a consequence of the genomic organization of the MiT family of genes, its (de)regulation may occur not only at the level of transcription, but also at the level of premRNA processing, resulting in functionally distinct isoforms. Indeed, we observed that in t(6;11)(p21;q13)-positive RCCs, next to the up-regulation of TFEB, the TFEC variant A was expressed, whereas the kidney-specific TFEC variant C lacking the N-terminal AAD domain (Kuiper et al., 2004), was not expressed. As yet, the exact functional consequences of these latter (de)regulatory expression phenomena remain to be established. MiT in growth and development
The MiT genes encode closely related proteins with overlapping functions in various growth and developmental processes (Steingrimsson et al., 2002, 2004). MiTF is the best characterized member of this family and is, among others, essential for mast cell, melanocyte and eye development (Steingrimsson et al., 1994). In addition, it has been found that MiTF regulates different target genes in osteoclasts and melanocytes (Mansky et al., 2002). Although osteoclasts seem normal in MITF or TFE3 null mice, the combined loss of these two genes results in severe osteopetrosis, possibly via deregulation of cathepsin-K (Motyckova et al., 2001). Through microarray-based expression profiling BCL2 has been identified as a putative MiTF transcriptional target, and it has been suggested that this functional relationship may explain the frequently observed treatment resistance in melanoma (McGill et al., 2002). MiTF also regulates BCL2 expression in osteoclasts and, in accordance with this notion, both Mitf mi/mi (dominant-negative mutant) and Bcl2–/– mice exhibit severe osteopetrosis (McGill et al., 2002). In addition, MITF has been found to act in concert with TFE3 and TFEC to efficiently regulate the expression of other target genes such as the tartrate-resistant acid phosphatase gene in osteoclasts (Mansky et al., 2002).
Cytogenet Genome Res 118:157–165 (2007)
161
A
MiT heterodimer bound to E-box sequence
MiT
MiT
CA[C/T]GTG Transcription start site
E-box
B
MiT homodimer bound to E-box sequence
MiT
MiT
CA[C/T]GTG E-box
C
Transcription start site
MiT (fusion) heterodimer bound to E-box sequence
MiT
PRCCTFE3
CA[C/T]GTG E-box
Transcription start site
Fig. 3. Model for MiT transcription (de)regulation in normal cells and tumor cells. In normal cells, MiT (A) hetero- or (B) homodimers can bind to E-boxes to regulate downstream target genes. In tumor cells, the ratio of hetero- or homodimers may be disturbed, resulting in a deregulation of downstream target genes. In case of tumor-associated MiT gene fusions (C), hetero- or homodimeric fusion proteins such as PRCCTFE3 can bind to E-boxes and deregulate downstream target genes.
TFE3 was first identified as a protein binding to the E-box of the immunoglobulin heavy chain enhancer (Beckmann et al., 1990). Using electrophoretic mobility shift assays it was found that TFE3 can, in fact, bind to several gene-specific regulatory elements (Kido et al., 1999; Verastegui et al., 2000). In addition, it has been shown that TFE3 and SMAD3 can act synergistically to activate the PE2 promoter, an E-box containing fragment of the plasminogen activator inhibitor PAI-1 (SERPINE1) gene (Hua et al., 1998). TFE3 plays a prominent role in the process of osteoclast formation (see above), in particular during the transition from mono-nucleated to multi-nucleated cells. Since mono-nucleated osteoclasts are derived from macrophages, TFE3 is thought to play a role during early hematopoietic differentiation as well (Balaban et al., 1984; Zanocco-Marani et al., 2006). Very recently, it was found that TFE3 may confer resistance to an RB-induced cell cycle arrest (Nijman et al., 2006) and that it can override such an arrest in Rat1 cells. In addition, it has been shown that TFE3 can block the anti-mitogenic effects of TGF- in rodent and human cells. Since TFE3 is also known to play a role in the direct regulation of cyclin E expression in an E2F3-dependent manner, TFE3 is considered as a functional regulator of proliferation (Nijman et al., 2006).
162
Cytogenet Genome Res 118:157–165 (2007)
Through the injection of TFE3-deficient ES cells into RAG2 (recombinase activating gene 2)-deficient blastocysts, chimeric animals were generated (Merrell et al., 1997). Lymphocytes of these animals showed lack of recombination of their immunoglobulin and T cell receptor genes. In other somatic tissues, however, no abnormalities were observed. Based on these observations, it was postulated that also here related members of the MiT family may have taken over the function of TFE3 (Merrell et al., 1997). TFEB was originally isolated by virtue of its promoter binding capacity (Fisher et al., 1991). Targeted disruption of the Tcfeb gene in mice revealed intra-uterine death between 9.5 and 10.5 days of gestation in conjunction with severe defects in placental vascularization (Steingrimsson et al., 1998). These defects could be attributed to a failure of labyrinthine cells to express VEGF, a potent mitogen required for normal vascularization of embryonic and extra-embryonic tissues. Therefore, it has been suggested that TFEB may play a critical role in signal transduction processes required for vascularization. Remarkably, TFEB expression in the placenta was found to be driven by a promoter that appears to be active exclusively in this tissue (Kuiper et al., 2004). Since none of the other MiT mutant mice showed defects in placental vascularization, TFEB may act as a homodimer in this process (Steingrimsson et al., 2002). In case of activation of Cd40lg expression in activated CD4+ T cells, TFEB and TFE3 were found to be able to rescue each others phenotype and thus, again, to be mutually redundant (Huan et al., 2006). TFEC was first identified in cells of the mononuclear phagocyte lineage, coexpressed with other members of the MiT family. Within this lineage, TFEC expression was found to be restricted to macrophages. No TFEC expression was observed in a variety of other cell types, including fibroblasts, myoblasts, chondrosarcoma and myeloma cells (Steingrimsson et al., 2002). The functional importance of the restricted expression pattern of TFEC in macrophages is likely related to the critical role of the MiT family in this particular hematopoietic lineage (Rehli et al., 1999). Initially TFEC was thought to act as a transcriptional repressor (Yang et al., 1998; Pozniak et al., 2000). More recently, however, it was found that TFEC can act equally well as a transcriptional activator when fused to the GAL4 DNA-binding domain in a yeast one hybrid type assay (Mansky et al., 2002). The tissue-specific expression of the various MiT isoforms very likely results in cell type-specific homo- and/or heterodimerization patterns, thus enabling the MiT family of transcription factors to regulate a broad spectrum of cellular growth and differentiation processes. This regulation appears to be tightly controlled at multiple levels (Kuiper et al., 2004). As a result of tumor-associated MiT deregulation, however, this tight control is lost. This loss, in turn, results in a considerable overlap in the deregulation of MiT downstream target genes in RCCs, as well as in melanomas and clear cell sarcomas (Du et al., 2003; Argani and Ladanyi, 2005; Argani, 2006; Davis et al., 2006).
Epilogue
Based on clinical, histological and molecular-cytogenetic data, a novel subgroup of RCCs has been defined, i.e. the MiT translocation subgroup of RCCs. As yet, in these RCCs two members of the MiT family of basic helix-loop-helix/ leucine zipper transcription factor genes have been shown to be disrupted and fused to a set of other genes, i.e., fusion of the TFE3 gene to the PRCC, PSF, NONO, ASPL and CLTC genes, and fusion of the TFEB gene to the Alpha gene. These gene fusions lead to activation and/or upregulation of the respective MiT genes. In accordance with this notion, a strong immunoreactivity for TFEB was observed in t(6; 11)(p21;q13)-positive RCCs. Similarly, a strong TFE3 immunoreactivity was observed in RCCs with Xp11 translocations. In these Xp11-associated RCCs also MiTF was found to be up-regulated. At the clinical level, these findings have yielded novel tools for the differential diagnosis of RCC. In addition, they have suggested that altered MiT activity, irrespective of the fusion partners, may contribute to tumorigenesis. Indeed, functional assays have revealed that the PRCCTFE3 fusion protein, such as the activated/up-regu-
lated forms of the TFEB, TFE3 and MiTF proteins, may act as oncogenes. The MiT family members are thought to act in similar signaling pathways. In RCC these pathways may be affected through altered transcription regulation and concomitant altered downstream target gene regulation (Fig. 3), or through interference with interacting proteins. These interacting proteins may include p53 or the RB1 target protein E2F3, both of which regulate the G1 to S phase transition of the cell cycle, or CLTC and the PRCC interactor Mad2B, both of which are involved in regulating the control over the mitotic spindle checkpoint. Together, the current data suggest that (i) deregulation of downstream target gene expression and (ii) loss of control over the cell cycle may play prominent roles in the development of the MiT translocation subgroup of RCCs. Acknowledgement The authors thank Dr. Murty for sharing patient material.
References Argani P: The evolving story of renal translocation carcinomas. Am J Clin Pathol 126: 332–334 (2006). Argani P, Ladanyi M: Distinctive neoplasms characterised by specific chromosomal translocations comprise a significant proportion of paediatric renal cell carcinomas. Pathology 35: 492–498 (2003). Argani P, Ladanyi M: Translocation carcinomas of the kidney. Clin Lab Med 25:363–378 (2005). Argani P, Antonescu CR, Illei PB, Lui MY, Timmons CF, et al: Primary renal neoplasms with the ASPL-TFE3 gene fusion of alveolar soft part sarcoma: a distinctive tumor entity previously included among renal cell carcinomas of children and adolescents. Am J Pathol 159:179–192 (2001a). Argani P, Hawkins A, Griffin CA, Goldstein JD, Haas M, et al: A distinctive pediatric renal neoplasm characterized by epithelioid morphology, basement membrane production, focal HMB45 immunoreactivity, and t(6; 11)(p21.1; q12) chromosome translocation. Am J Pathol 158:2089–2096 (2001b). Argani P, Lui MY, Couturier J, Bouvier R, Fournet JC, et al: A novel CLTC-TFE3 gene fusion in pediatric renal adenocarcinoma with t(X;17) (p11.2;q23). Oncogene 22: 5374–5378 (2003). Argani P, Lae M, Hutchinson B, Reuter VE, Collins MH, et al: Renal carcinomas with the t(6; 11)(p21;q12): clinicopathologic features and demonstration of the specific alpha-TFEB gene fusion by immunohistochemistry, RT-PCR, and DNA PCR. Am J Surg Pathol 29: 230–240 (2005). Argani P, Lae M, Ballard ET, Amin M, Manivel C, et al: Translocation carcinomas of the kidney after chemotherapy in childhood. J Clin Oncol 24:1529–1534 (2006). Artandi SE, Merrell K, Avitahl N, Wong KK, Calame K: TFE3 contains two activation domains, one acidic and the other proline-rich, that synergistically activate transcription. Nucleic Acids Res 23: 3865–3871 (1995).
Balaban G, Herlyn M, Guerry D, Bartolo R, Koprowski H, et al: Cytogenetics of human malignant melanoma and premalignant lesions. Cancer Genet Cytogenet 11: 429–439 (1984). Beckmann H, Su LK, Kadesch T: TFE3: a helixloop-helix protein that activates transcription through the immunoglobulin enhancer muE3 motif. Genes Dev 4: 167–179 (1990). Benezra R, Davis RL, Lockshon D, Turner DL, Weintraub H: The protein Id: a negative regulator of helix-loop-helix DNA binding proteins. Cell 61:49–59 (1990). Bodmer D, van den Hurk WH, van Groningen JJM, Eleveld MJ, Martens GJ, et al: Understanding familial and non-familial renal cell cancer. Hum Mol Genet 11: 2489–2498 (2002). Bonham K, Ritchie SA, Dehm SM, Snyder K, Boyd FM: An alternative, human SRC promoter and its regulation by hepatic nuclear factor-1alpha. J Biol Chem 275: 37604–37611 (2000). Busca R, Berra E, Gaggioli C, Khaled M, Bille K, et al: Hypoxia-inducible factor 1(alpha) is a new target of microphthalmia-associated transcription factor (MITF) in melanoma cells. J Cell Biol 170: 49–59 (2005). Cahill DP, da Costa LT, Carson-Walter EB, Kinzler KW, Vogelstein B, et al: Characterization of MAD2B and other mitotic spindle checkpoint genes. Genomics 58: 181–187 (1999). Carcao MD, Taylor GP, Greenberg ML, Bernstein ML, Champagne M, et al: Renal-cell carcinoma in children: a different disorder from its adult counterpart? Med Pediatr Oncol 31: 153–158 (1998). Clark J, Lu YJ, Sidhar SK, Parker C, Gill S, et al: Fusion of splicing factor genes PSF and NonO (p54nrb) to the TFE3 gene in papillary renal cell carcinoma. Oncogene 15: 2233–2239 (1997). Cohen HT, McGovern FJ: Renal-cell carcinoma. N Engl J Med 353: 2477–2490 (2005). Dal Cin P, Van Gool S, Brock P, Proesmans W, Casteels-Van Daele M, et al: Renal cell carcinoma in a child. Cancer Genet Cytogenet 57: 137–138 (1991).
Dal Cin P, Stas M, Sciot R, de Wever I, Van Damme B, et al: Translocation (X;1) reveals metastasis 31 years after renal cell carcinoma. Cancer Genet Cytogenet 101:58–61 (1998). Davis IJ, Hsi BL, Arroyo JD, Vargas SO, Yeh YA, et al: Cloning of an Alpha-TFEB fusion in renal tumors harboring the t(6;11)(p21;q13) chromosome translocation. Proc Natl Acad Sci USA 100:6051–6056 (2003). Davis IJ, Kim JJ, Ozsolak F, Widlund HR, Rozenblatt-Rosen O, et al: Oncogenic MITF dysregulation in clear cell sarcoma: defining the MiT family of human cancers. Cancer Cell 9: 473– 484 (2006). de Jong B, Molenaar IM, Leeuw JA, Idenberg VJ, Oosterhuis JW: Cytogenetics of a renal adenocarcinoma in a 2-year-old child. Cancer Genet Cytogenet 21: 165–169 (1986). Desangles F, Camparo P, Fouet C, Houlgatte A, Arborio M: Translocation (X;1) associated with a nonpapillary carcinoma in a young woman: a new definition for an Xp11.2 RCC subtype. Cancer Genet Cytogenet 113: 141–144 (1999). Dijkhuizen T, van den Berg E, Wilbrink M, Weterman M, Geurts van Kessel A, et al: Distinct Xp11.2 breakpoints in two renal cell carcinomas exhibiting X;autosome translocations. Genes Chromosomes Cancer 14: 43–50 (1995). Dijkhuizen T, van den Berg E, Storkel S, Geurts van Kessel A, Janssen B, et al: Two cases of renal cell carcinoma, clear cell type, revealling a t(6; 11)(p21;q13). Cancer Genet Cytogenet 91: 141 (1996). Dijkhuizen T, van den Berg E, Storkel S, Terpe HI, Burger H, et al: Distinct features for chromophilic renal cell cancer with Xp11.2 breakpoints. Cancer Genet Cytogenet 104: 74–76 (1998). Dong B, Horowitz DS, Kobayashi R, Krainer AR: Purification and cDNA cloning of HeLa cell p54nrb, a nuclear protein with two RNA recognition motifs and extensive homology to human splicing factor PSF and Drosophila NONA/ BJ6. Nucleic Acids Res 21: 4085–4092 (1993).
Cytogenet Genome Res 118:157–165 (2007)
163
Dreijerink K, Braga E, Kuzmin I, Geil L, Duh FM, et al: The candidate tumor suppressor gene, RASSF1A, from human chromosome 3p21.3 is involved in kidney tumorigenesis. Proc Natl Acad Sci USA 98:7504–7509 (2001). Du J, Miller AJ, Widlund HR, Horstmann MA, Ramaswamy S, et al: MLANA/MART1 and SILV/ PMEL17/GP100 are transcriptionally regulated by MITF in melanocytes and melanoma. Am J Pathol 163: 333–343 (2003). Du J, Widlund HR, Horstmann MA, Ramaswamy S, Ross K, et al: Critical role of CDK2 for melanoma growth linked to its melanocyte-specific transcriptional regulation by MITF. Cancer Cell 6:565–576 (2004). Eble JN, Sauter G, Epstein JI, Sesterhenn IA: Tumors of the kidney, in World Health Organization Classification of Tumors. Pathology and Genetics of Tumours of the Urinary System and Male Genital Organs, pp 37–38 (IARC Press, Lyon 2004). Fisher DE, Carr CS, Parent LA, Sharp PA: TFEB has DNA-binding and oligomerization properties of a unique helix-loop-helix/leucine-zipper family. Genes Dev 5:2342–2352 (1991). Garraway LA, Widlund HR, Rubin MA, Getz G, Berger AJ, et al: Integrative genomic analyses identify MITF as a lineage survival oncogene amplified in malignant melanoma. Nature 436: 117–122 (2005). Gnarra JR, Tory K, Weng Y, Schmidt L, Wei MH, et al: Mutations of the VHL tumour suppressor gene in renal carcinoma. Nat Genet 7: 85–90 (1994). Gozani O, Patton JG, Reed R: A novel set of spliceosome-associated proteins and the essential splicing factor PSF bind stably to pre-mRNA prior to catalytic step II of the splicing reaction. EMBO J 13:3356–3367 (1994). Hallsson JH, Favor J, Hodgkinson CA, Glaser T, Lamoreux ML, et al: Genomic, transcriptional and mutational analysis of the mouse microphthalmia locus. Genetics 155: 291–300 (2000). Heimann P, El Housni H, Ogur G, Weterman MA, Petty EM, et al: Fusion of a novel gene, RCC17, to the TFE3 gene in t(X;17)(p11.2;q25.3)-bearing papillary renal cell carcinomas. Cancer Res 61:4130–4135 (2001). Hemesath TJ, Steingrimsson E, McGill G, Hansen MJ, Vaught J, et al: Microphthalmia, a critical factor in melanocyte development, defines a discrete transcription factor family. Genes Dev 8:2770–2780 (1994). Herman JG, Latif F, Weng Y, Lerman MI, Zbar B, et al: Silencing of the VHL tumor-suppressor gene by DNA methylation in renal carcinoma. Proc Natl Acad Sci USA 91: 9700–9704 (1994). Hernandez-Marti MJ, Orellana-Alonso C, BadiaGarrabou L, Verdeguer MA, Paradis-Alos A: Renal adenocarcinoma in an 8-year-old child, with a t(X;17)(p11.2;q25). Cancer Genet Cytogenet 83: 82–83 (1995). Hua X, Liu X, Ansari DO, Lodish HF: Synergistic cooperation of TFE3 and smad proteins in TGF-beta-induced transcription of the plasminogen activator inhibitor-1 gene. Genes Dev 12:3084–3095 (1998). Huan C, Kelly ML, Steele R, Shapira I, Gottesman SR, et al: Transcription factors TFE3 and TFEB are critical for CD40 ligand expression and thymus-dependent humoral immunity. Nat Immunol 7: 1082–1091 (2006). Jhang JS, Narayan G, Murty VV, Mansukhani MM: Renal oncocytomas with 11q13 rearrangements: cytogenetic, molecular, and immunohistochemical analysis of cyclin D1. Cancer Genet Cytogenet 149:114–119 (2004).
164
Kamat A, Hinshelwood MM, Murry BA, Mendelson CR: Mechanisms in tissue-specific regulation of estrogen biosynthesis in humans. Trends Endocrinol Metab 13: 122–128 (2002). Kardas I, Denis A, Babinska M, Gronwald J, Podolski J, et al: Translocation (X;1)(p11.2;q21) in a papillary renal cell carcinoma in a 14-yearold girl. Cancer Genet Cytogenet 101: 159–161 (1998). Kido S, Miyamoto K, Mizobuchi H, Taketani Y, Ohkido I, et al: Identification of regulatory sequences and binding proteins in the type II sodium/phosphate cotransporter NPT2 gene responsive to dietary phosphate. J Biol Chem 274:28256–28263 (1999). Kirchhausen T: Clathrin. Annu Rev Biochem 69: 699–727 (2000). Kovacs G, Brusa P: Recurrent genomic rearrangements are not at the fragile sites on chromosomes 3 and 5 in human renal cell carcinomas. Hum Genet 80: 99–101 (1988). Kovacs G, Szucs S, de Riese W, Baumgartel H: Specific chromosome aberration in human renal cell carcinoma. Int J Cancer 40:171–178 (1987). Kovacs G, Akhtar M, Beckwith BJ, Bugert P, Cooper CS, et al: The Heidelberg classification of renal cell tumours. J Pathol 183: 131–133 (1997). Kuiper RP, Schepens M, Thijssen J, van Asseldonk M, van den Berg E, et al: Upregulation of the transcription factor TFEB in t(6; 11)(p21;q13)positive renal cell carcinomas due to promoter substitution. Hum Mol Genet 12: 1661–1669 (2003). Kuiper RP, Schepens M, Thijssen J, Schoenmakers EFPM, Geurts van Kessel A: Regulation of the MiTF/TFE bHLH-LZ transcription factors through restricted spatial expression and alternative splicing of functional domains. Nucleic Acids Res 32:2315–2322 (2004). Ladanyi M, Lui MY, Antonescu CR, Krause-Boehm A, Meindl A, et al: The der(17)t(X;17)(p11;q25) of human alveolar soft part sarcoma fuses the TFE3 transcription factor gene to ASPL , a novel gene at 17q25. Oncogene 20: 48–57 (2001). Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, et al: Initial sequencing and analysis of the human genome. Nature 409: 860–921 (2001). Landry JR, Mager DL: Widely spaced alternative promoters, conserved between human and rodent, control expression of the Opitz syndrome gene MID1. Genomics 80: 499–508 (2002). Landry JR, Mager DL, Wilhelm BT: Complex controls: the role of alternative promoters in mammalian genomes. Trends Genet 19: 640–648 (2003). Langezaal SM, Graadt van Roggen JF, Cleton-Jansen AM, Baelde JJ, Hogendoorn PC: Malignant melanoma is genetically distinct from clear cell sarcoma of tendons and aponeurosis (malignant melanoma of soft parts). Br J Cancer 84: 535–538 (2001). Larsen LK, Amri EZ, Mandrup S, Pacot C, Kristiansen K: Genomic organization of the mouse peroxisome proliferator-activated receptor beta/ delta gene: alternative promoter usage and splicing yield transcripts exhibiting differential translational efficiency. Biochem J 366:767–775 (2002). Li R, Murray AW: Feedback control of mitosis in budding yeast. Cell 66:519–531 (1991). Mansky KC, Sulzbacher S, Purdom G, Nelsen L, Hume DA, et al: The microphthalmia transcription factor and the related helix-loop-helix zipper factors TFE-3 and TFE-C collaborate to activate the tartrate-resistant acid phosphatase promoter. J Leukoc Biol 71: 304–310 (2002). Mathur M, Samuels HH: Role of PSF-TFE3 oncoprotein in the development of papillary renal cell carcinomas. Oncogene 26:277–283 (2007).
Cytogenet Genome Res 118:157–165 (2007)
Maxwell PH, Wiesener MS, Chang GW, Clifford SC, Vaux EC, et al: The tumour suppressor protein VHL targets hypoxia-inducible factors for oxygen-dependent proteolysis. Nature 399: 271–275 (1999). McGill GG, Horstmann M, Widlund HR, Du J, Motyckova G, et al: Bcl2 regulation by the melanocyte master regulator Mitf modulates lineage survival and melanoma cell viability. Cell 109: 707–718 (2002). Medstrand P, Landry JR, Mager DL: Long terminal repeats are used as alternative promoters for the endothelin B receptor and apolipoprotein C-I genes in humans. J Biol Chem 276: 1896–1903 (2001). Meloni AM, Dobbs RM, Pontes JE, Sandberg AA: Translocation (X;1) in papillary renal cell carcinoma. A new cytogenetic subtype. Cancer Genet Cytogenet 65: 1–6 (1993). Merrell K, Wells S, Henderson A, Gorman J, Alt F, et al: The absence of the transcription activator TFE3 impairs activation of B cells in vivo. Mol Cell Biol 17: 3335–3344 (1997). Modrek B, Resch A, Grasso C, Lee C: Genome-wide detection of alternative splicing in expressed sequences of human genes. Nucleic Acids Res 29: 2850–2859 (2001). Morrissey C, Martinez A, Zatyka M, Agathanggelou A, Honorio S, et al: Epigenetic inactivation of the RASSF1A 3p21.3 tumor suppressor gene in both clear cell and papillary renal cell carcinoma. Cancer Res 61: 7277–7281 (2001). Motyckova G, Weilbaecher KN, Horstmann M, Rieman DJ, Fisher DZ, et al: Linking osteopetrosis and pycnodysostosis: regulation of cathepsin K expression by the microphthalmia transcription factor family. Proc Natl Acad Sci USA 98: 5798–5803 (2001). Motzer RJ, Bander NH, Nanus DM: Renal-cell carcinoma. N Engl J Med 335: 865–875 (1996). Nijman SM, Hijmans EM, El Messaoudi S, van Dongen MM, Sardet C, et al: A functional genetic screen identifies TFE3 as a gene that confers resistance to the anti-proliferative effects of the retinoblastoma protein and transforming growth factor-beta. J Biol Chem 281: 21582– 21587 (2006). Ohjimi Y, Iwasaki H, Ishiguro M, Hara H, Ohgami A, et al: Deletion (X)(p11): another case of renal adenocarcinoma with involvement of Xp11. Cancer Genet Cytogenet 70: 77–78 (1993). Onder AM, Teomete U, Argani P, Toledano S, Zilleruelo G, et al: PRCC-TFE3 Renal cell carcinoma in a boy with a history of contralateral mesoblastic nephroma. Pediatr Nephrol 21: 1471–1475 (2006). Patton JG, Porro EB, Galceran J, Tempst P, NadalGinard B: Cloning and characterization of PSF, a novel pre-mRNA splicing factor. Genes Dev 7: 393–406 (1993). Perot C, Bougaran J, Boccon-Gibod L, Storkel S, Leverger G, et al: Two new cases of papillary renal cell carcinoma with t(X;1)(p11;q21) in females. Cancer Genet Cytogenet 110: 54–56 (1999). Perot C, Boccon-Gibod L, Bouvier R, Doz F, Fournet JC, et al: Five new cases of juvenile renal cell carcinoma with translocations involving Xp11.2: a cytogenetic and morphologic study. Cancer Genet Cytogenet 143: 93–99 (2003). Phelps DE, Hsiao KM, Li Y, Hu N, Franklin DS, et al: Coupled transcriptional and translational control of cyclin-dependent kinase inhibitor p18INK4c expression during myogenesis. Mol Cell Biol 18: 2334–2343 (1998). Pozniak CD, Radinovic S, Yang A, McKeon F, Kaplan DR, et al: An anti-apoptotic role for the p53 family member, p73, during developmental neuron death. Science 289:304–306 (2000).
Pulford K, Morris SW, Turturro F: Anaplastic lymphoma kinase proteins in growth control and cancer. J Cell Physiol 199:330–358 (2004). Ramp U, Caliskan E, Ebert T, Karagiannidis C, Willers R, et al: FHIT expression in clear cell renal carcinomas: versatility of protein levels and correlation with survival. J Pathol 196: 430–436 (2002). Ramphal R, Pappo A, Zielenska M, Grant R, Ngan BY: Pediatric renal cell carcinoma: clinical, pathologic, and molecular abnormalities associated with the members of the MiT transcription factor family. Am J Clin Pathol 126: 349– 364 (2006). Rehli M, Lichanska A, Cassady AI, Ostrowski MC, Hume DA: TFEC is a macrophage-restricted member of the microphthalmia-TFE subfamily of basic helix-loop-helix leucine zipper transcription factors. J Immunol 162: 1559–1565 (1999). Roman C, Cohn L, Calame K: A dominant negative form of transcription activator mTFE3 created by differential splicing. Science 254: 94–97 (1991). Royle SJ, Bright NA, Lagnado L: Clathrin is required for the function of the mitotic spindle. Nature 434:1152–1157 (2005). Saleh A, Makrigiannis AP, Hodge DL, Anderson SK: Identification of a novel Ly49 promoter that is active in bone marrow and fetal thymus. J Immunol 168: 5163–5169 (2002). Selzer E, Wacheck V, Lucas T, Heere-Ress E, Wu M, et al: The melanocyte-specific isoform of the microphthalmia transcription factor affects the phenotype of human melanoma. Cancer Res 62:2098–2103 (2002). Shibahara S, Takeda K, Yasumoto K, Udono T, Watanabe K, et al: Microphthalmia-associated transcription factor (MITF): multiplicity in structure, function, and regulation. J Investig Dermatol Symp Proc 6: 99–104 (2001). Shipley JM, Birdsall S, Clark J, Crew J, Gill S, et al: Mapping the X chromosome breakpoint in two papillary renal cell carcinoma cell lines with a t(X;1)(p11.2;q21.2) and the first report of a female case. Cytogenet Cell Genet 71: 280–284 (1995). Sidhar SK, Clark J, Gill S, Hamoudi R, Crew AJ, et al: The t(X;1)(p11.2;q21.2) translocation in papillary renal cell carcinoma fuses a novel gene PRCC to the TFE3 transcription factor gene. Hum Mol Genet 5:1333–1338 (1996). Smith CJ, Pearse BM: Clathrin: anatomy of a coat protein. Trends Cell Biol 9: 335–338 (1999). Steingrimsson E, Moore KJ, Lamoreux ML, FerreD’Amare AR, Burley SK, et al: Molecular basis of mouse microphthalmia (mi) mutations helps explain their developmental and phenotypic consequences. Nat Genet 8:256–263 (1994). Steingrimsson E, Tessarollo L, Reid SW, Jenkins NA, Copeland NG: The bHLH-Zip transcription factor Tfeb is essential for placental vascularization. Development 125: 4607–4616 (1998).
Steingrimsson E, Tessarollo L, Pathak B, Hou L, Arnheiter H, et al: Mitf and Tfe3, two members of the Mitf-Tfe family of bHLH-Zip transcription factors, have important but functionally redundant roles in osteoclast development. Proc Natl Acad Sci USA 99: 4477–4482 (2002). Steingrimsson E, Copeland NG, Jenkins NA: Melanocytes and the microphthalmia transcription factor network. Annu Rev Genet 38: 365–411 (2004). Tomlinson GE, Nisen PD, Timmons CF, Schneider NR: Cytogenetics of a renal cell carcinoma in a 17-month-old child. Evidence for Xp11.2 as a recurring breakpoint. Cancer Genet Cytogenet 57:11–17 (1991). Tonk V, Wilson KS, Timmons CF, Schneider NR, Tomlinson GE: Renal cell carcinoma with translocation (X;1). Further evidence for a cytogenetically defined subtype. Cancer Genet Cytogenet 81: 72–75 (1995). Trinklein ND, Aldred SJ, Saldanha AJ, Myers RM: Identification and functional analysis of human transcriptional promoters. Genome Res 13:308–312 (2003). Udono T, Yasumoto K, Takeda K, Amae S, Watanabe K, et al: Structural organization of the human microphthalmia-associated transcription factor gene containing four alternative promoters. Biochim Biophys Acta 1491: 205–219 (2000). van Asseldonk M, Schepens M, de Bruijn DR, Janssen B, Merkx G, et al: Construction of a 350-kb sequence-ready 11q13 cosmid contig encompassing the markers D11S4933 and D11S546: mapping of 11 genes and 3 tumor-associated translocation breakpoints. Genomics 66:35–42 (2000). van den Hurk WH, Martens GJ, Geurts van Kessel A, van Groningen JJM: Isolation and characterization of the Xenopus laevis orthologs of the human papillary renal cell carcinoma-associated genes PRCC and MAD2L2 (MAD2B). Cytogenet Genome Res 106: 68–73 (2004). Verastegui C, Bertolotto C, Bille K, Abbe P, Ortonne JP, et al: TFE3, a transcription factor homologous to microphthalmia, is a potential transcriptional activator of tyrosinase and TyrpI genes. Mol Endocrinol 14: 449–456 (2000). Wang Y, Newton DC, Robb GB, Kau CL, Miller TL, et al: RNA diversity has profound effects on the translation of neuronal nitric oxide synthase. Proc Natl Acad Sci USA 96: 12150–12155 (1999). Weterman MA, Wilbrink M, Geurts van Kessel A: Fusion of the transcription factor TFE3 gene to a novel gene, PRCC, in t(X;1)(p11;q21)-positive papillary renal cell carcinomas. Proc Natl Acad Sci USA 93: 15294–15298 (1996a). Weterman MA, Wilbrink M, Janssen I, Janssen HA, van den Berg E et al: Molecular cloning of the papillary renal cell carcinoma-associated translocation (X;1)(p11;q21) breakpoint. Cytogenet Cell Genet 75: 2–6 (1996b).
Weterman MJ, van Groningen JJM, Jansen A, Geurts van Kessel A: Nuclear localization and transactivating capacities of the papillary renal cell carcinoma-associated TFE3 and PRCC (fusion) proteins. Oncogene 19: 69–74 (2000). Weterman MA, van Groningen JJM, den Hartog A, Geurts van Kessel A: Transformation capacities of the papillary renal cell carcinoma-associated PRCCTFE3 and TFE3PRCC fusion genes. Oncogene 20: 1414–1424 (2001a). Weterman MA, van Groningen JJM, Tertoolen L, Geurts van Kessel A: Impairment of MAD2BPRCC interaction in mitotic checkpoint defective t(X;1)-positive renal cell carcinomas. Proc Natl Acad Sci USA 98: 13808–13813 (2001b). Widlund HR, Horstmann MA, Price ER, Cui J, Lessnick SL, et al: Beta-catenin-induced melanoma growth requires the downstream target microphthalmia-associated transcription factor. J Cell Biol 158: 1079–1087 (2002). Yang A, Kaghad M, Wang Y, Gillett E, Fleming MD, et al: p63, a p53 homolog at 3q27–29, encodes multiple products with transactivating, deathinducing, and dominant-negative activities. Mol Cell 2: 305–316 (1998). Yang YS, Hanke JH, Carayannopoulos L, Craft CM, Capra JD, et al: NonO, a non-POU-domaincontaining, octamer-binding protein, is the mammalian homolog of Drosophila nonAdiss. Mol Cell Biol 13: 5593–5603 (1993). Ybe JA, Greene B, Liu SH, Pley U, Parham P, et al: Clathrin self-assembly is regulated by three light-chain residues controlling the formation of critical salt bridges. EMBO J 17: 1297–1303 (1998). Ybe JA, Brodsky FM, Hofmann K, Lin K, Liu SH, et al: Clathrin self-assembly is mediated by a tandemly repeated superhelix. Nature 399:371–375 (1999). Yenamandra A, Zhou X, Trinchitella L, Susin M, Sastry S, et al: Renal cell carcinoma with X;1 translocation in a child with Klinefelter syndrome. Am J Med Genet 77: 281–284 (1998). Yoshida MA, Ochi-Takeuchi H, Gibas Z, Sandberg AA: Updating of chromosome changes in renal cell carcinoma. Proc AACR 28: 31 (1985). Zanocco-Marani T, Vignudelli T, Gemelli C, Pirondi S, Testa A, et al: Tfe3 expression is closely associated to macrophage terminal differentiation of human hematopoietic myeloid precursors. Exp Cell Res 312:4079–4089 (2006). Zattara-Cannoni H, Daniel L, Roll P, Coulange C, Vagner-Capodano A: Molecular cytogenetics of t(X;1)(p11.2;q21) with complex rearrangements in a renal cell carcinoma. Cancer Genet Cytogenet 123: 61–64 (2000). Zhao WP, Gnarra JR, Liu S, Knutsen T, Linehan WM, et al: Renal cell carcinoma. Cytogenetic analysis of tumors and cell lines. Cancer Genet Cytogenet 82: 128–139 (1995). Zucman J, Delattre O, Desmaze C, Epstein AL, Stenman G, et al: EWS and ATF-1 gene fusion induced by t(12; 22) translocation in malignant melanoma of soft parts. Nat Genet 4: 341–345 (1993).
Cytogenet Genome Res 118:157–165 (2007)
165
with cisplatin-based regimes (Knowles, 2001; Borden et al., 2005). A number of chromosomal changes have been associated with bladder tumor development and tumor progression. While oncogenic mutations of FGFR3 and loss of heterozygosity (LOH) at chromosome 9 are the most common alterations in non-invasive tumors, numerous genetic alterations including TP53 mutations and LOH at 3p, 8p, 9p/q, 13q, 14q and 17p are detectable in invasive urothelial carcinomas (UC) (Knowles, 2006; Schulz, 2006). Loss of genetic material on chromosome 9 is the most frequent alteration in UC and is found in more than 50% of all bladder tumors (Knowles, 1999; Kimura et al., 2001; Williams et al., 2002). Frequently deleted regions include 9p13]p12, 9p21, 9q12, 9q13]q31, 9q22.3, 9q32]33, 9q33]34 and 9q34 (Keen and Knowles, 1994; Knowles, 1999; Kimura et al., 2001; Florl and Schulz, 2003). Since deletions on chromosome 9 are found throughout all stages and grades of bladder cancer, it is believed that inactivation of genes on chromosome 9 represents an early event in tumor progression. One established target of deletions is CDKN2A at 9p21. However, as several other regions on 9p and 9q are consistently affected by LOH, the involvement of further, as yet unidentified tumor suppressor genes must be considered. Two prominent human caretaker genes are located in the frequently deleted regions of chromosome 9. These genes, FANCC (9q22.3) and FANCG (9p13), belong to a family of at least 13 genes (FANCA, -B, -C, -D1/BRCA2, -D2, -E, -F, -G, -I, -J/BRIP1, -L, -M/HEF and -N/PALB2) which act in the FA/BRCA pathway of DNA damage recognition and repair (Reid et al., 2007; Xia et al., 2007). The products of eight of these genes (FANCA, -B, -C, -E, -F, -G, -L and -M) form a nuclear complex required for monoubiquitination of FANCD2 during S-phase and in response to replication fork-stalling DNA damage. Monoubiquitinated FANCD2 is targeted to chromatin and assembles in nuclear foci representing regions of DNA repair. FANCD2 directly interacts with BRCA2, and co-localizes with DNA repair proteins such as BRCA1 and RAD51 involved in translesion synthesis and homologous recombination (Joenje and Patel, 2001; Kennedy and D’Andrea, 2005; Niedernhofer et al., 2005; Thompson et al., 2005). Three of the FA proteins, FANCD1, FANCJ, and FANCN, are not required for FANCD2 monoubiquitination (working downstream), and these three proteins are thought to be involved in tumorigenesis in heterozygous state: FANCD1 is identical to the breast cancer susceptibility gene BRCA2 (Howlett et al., 2002), FANCJ is also known as BRCA1-interacting protein (BRIP1) (Levran et al., 2005) and FANCN is a partner and localizer of BRCA2 (Xia et al., 2006). Cells deficient in any of the FA proteins show chromosomal instability, increased sensitivity towards DNA damaging agents such as cisplatin, MMC or diepoxybutane (DEB) and hypersensitivity towards oxygen (Sasaki and Tonomura, 1973; Joenje et al., 1981). Biallelic, or hemizygous in case of FANCB, germline mutations in any one of the FA genes cause the recessive hereditary disease FA which is associated with a greatly increased risk for neoplasia (Alter, 2003).
There are three main reasons for investigating a putative connection between FA genes and bladder cancer. First, the cellular and clinical phenotypes of FA indicate a connection between the FA proteins, DNA repair and tumorigenesis. Bladder cancer cells from invasive tumors are known to be genomically unstable. Second, cisplatin is a widespread component of all current chemotherapy regimes for invasive bladder cancer, and MMC is the most commonly used drug for adjuvant treatment of papillary bladder cancers. FA cells are highly sensitive towards these two DNA crosslinking agents. Third, deletions of chromosome 9 often include the loci of two of the FA genes (FANCC and FANCG). In order to establish whether disruption of FA genes might be involved in bladder cancer tumorigenesis or tumor progression, we investigated DNA from ten bladder cancers for LOH at FANCC and FANCG. Further 41 tumor tissues were investigated for methylation of the FANCF promoter region and 23 bladder cancer cell lines for defects in the FA/BRCA pathway. Materials and methods Cancer tissues DNA from 41 bladder carcinoma tissues was used which had been analyzed for methylation of multiple other genes in a previous study (Neuhausen et al., 2006). Of the 41 carcinomas, two were staged as pTa, five as pT1, 13 as pT2, 16 as pT3 and five as pT4. One case was graded as G1, two as G2, 37 as G3, and one as G4, respectively. Use of patient tissues was approved by the Institutional Review Board at the faculty of medicine of the Heinrich Heine University Düsseldorf. Cell lines and cell culture The following cell lines derived from human bladder carcinomas were investigated in this study (see Table 1 for references): 5637, BFTC905, HT1376, SD, SW1710, VMCub1 and VMCub2 were obtained from the DSMZ (Braunschweig, Germany), the cell lines 253J, 639V, 647V, BFTC909, EJ, J82, RT112, RT4, T24, Umuc3 and VMCub3 were from Dr. J. Fogh (New York, NY). HIA was a gift of H. Hameister (Ulm, Germany) and MGHU4 of W. Beecken (Frankfurt, Germany). DSH1 was established by one of the authors (M.A.K., reported in Williams et al., 2002). In addition, we investigated the only known cell line derived from a squamous cell carcinoma of the bladder (Scaber) and a novel cell line derived from an invasive bladder carcinoma (BC44; Seifert et al., unpublished). Primary urothelial strains (UP124, UP125) were used as controls. Bladder cancer cell lines were cultured in Eagle’s minimal essential medium (MEM) (Gibco, Karlsruhe, Germany) supplemented with 16% fetal bovine serum (FBS) (Sigma, Taufkirchen, Germany). All cell cultures were kept in high humidity incubators equipped with CO2 sensors in an atmosphere of 5% (v/v) CO2 by replacing ambient air with nitrogen. Primary urothelial cells were cultured with the specifics described (Swiatkowski et al., 2003). Cell cycle analysis Approximately 5 ! 105 cells each were cultured for 48 h without or with 5 or 10 ng/ml MMC, respectively. Cells were harvested by trypsinization and centrifuged. The cell pellets were resuspended in 1 ml staining buffer/1 million cells (staining buffer: 154 mM NaCl, 0.1 M Tris pH 7.4, 1 mM CaCl 2, 0.5 mM MgCl 2, 0.2% BSA, 0.1% NP40) and 1 g/ml DAPI. Following incubation for 30 min at 4 ° C, DNA histograms were recorded using an analytical flow cytometer of conventional design (LSR, Becton Dickinson, Heidelberg, Germany).
Cytogenet Genome Res 118:166–176 (2007)
167
Table 1. Bladder cancer cell lines used in this study, modified from Williams et al. (2005)
Cell line
Grade/stage of tumor
Sex
Other information
Reference
5637 253J 639V 647V BFTC905 BFTC909 DSH1 EJ HIA HT1376 J82 MGHU4 RT4 RT112 SD SW1710 T24 Umuc3 VMCub1 VMCub2 VMCub3 Scaber BC44 UP124 UP125
UC UC, G4 T4 UC, G3 UC, G2 UC, G3 papillary UC, Tx G3 UC, T1a G2 UC, G3 UC UC, G3 UC, poorly differentiated, papillary UC UC, G1 T2 UC, G2 papillary UC UC, G3 papillary UC, G3 UC UC UC UC Squamous cell carcinoma UC, T4 G3 Normal urothelium, control Normal urothelium, control
M M M M F M M F not recorded F M M M F not recorded F F M M M M
Primary tumor, bladder Metastatic tumor, lymph node Primary tumor, ureter Primary tumor, bladder Primary tumor, bladder Primary tumor, renal pelvis Recurrence, untreated, bladder Primary tumor, untreated, bladder not recorded Untreated, bladder Primary tumor, treated, bladder Urothelial atypia Recurrence, treated, bladder Primary tumor, untreated, bladder Primary tumor, bladder Bladder Primary tumor, untreated, bladder Bladder Primary tumor, bladder Metastatic tumor, lymph nodes Primary tumor, bladder Bladder Bladder Ureter Ureter
Williams, 1980 Elliott et al., 1974; Masters, 2000 Elliot et al., 1976 Williams, 1980 Tzeng et al., 1996 Tzeng et al., 1996 Williams et al., 2002 Williams, 1980 Bruch et al., 1999 Rasheed et al., 1977 O’Toole et al., 1978 Lin et al., 1985 Rigby and Franks, 1970 Masters, 2000 Paulie et al., 1983 Kyriazis et al., 1984 Williams, 1980 Grossmann et al., 1986 Williams, 1980 Williams, 1980 Williams, 1980 O’Toole et al., 1976 Seifert et al., unpublished Swiatkowski et al., 2003 Swiatkowski et al., 2003
F M F
Complementation One of the bladder cancer cell lines was transduced with retroviral constructs containing full-length cDNAs of FANCA, -C, -E, -F, -G or -L and analyzed for MMC sensitivity using cell cycle analysis as described (Hanenberg et al., 2002). FANCD2 immunoblotting FANCD2 immunoblotting was performed as described (GarciaHiguera et al., 2001) with minor modifications: cells were left untreated or were treated with 1 g/ml cisplatin for 6 or 16 h. Lysis was achieved with 1! lysis buffer (50 mM Tris-Cl, pH 7.4, 150 mM NaCl, 2 mM EDTA, 2 mM EGTA, 25 mM NaF, 25 mM -glycerophosphate, 0.1 mM Na3VO4, 0.3% NP40, 0.2% Triton X-100, proteinase inhibitors (Roche Diagnostics GmbH, Mannheim, Germany)). 50 g of protein extract was loaded on 7% Tris-acetate gels (Invitrogen, Karlsruhe, Germany) and electrophoresis was performed at 120 V for 6 h. Protein transfer was overnight at 4 ° C and 20 V onto PVDF membranes (HybondP, Amersham Biosciences, Little Chalfont, UK). Immunoblots were blocked with 5% non-fat dry milk (Hu et al., 2002) in 1! PBS, 0.05% Tween (PBS-T). As primary antibody, the mouse monoclonal FANCD2 antibody FL-17 (Santa Cruz Biotechnology Inc., Heidelberg, Germany) was used at a dilution of 1: 800. As secondary antibody we used the antimouse IgG horseradish-peroxidase-linked F(abⴕ)2 from sheep (Amersham Biosciences, Little Chalfont, UK, dilution 1:2000). For chemiluminescence detection, a standard ECL reagent (Amersham Biosciences, Little Chalfont, UK) was employed. Fluorescence in situ hybridization (FISH) FISH was performed as described elsewhere (Lichter and Cremer, 1992). The bacterial artificial chromosome (BAC) clone RP11-139M16 containing the entire FANCF gene was obtained from the Wellcome Trust Sanger Institute (Cambridge, UK). 1 g of BAC DNA was nicktranslated using biotin-16-dUTP and used as a probe on chromosome spreads prepared from the BFTC909 cell line. 100 ng of labeled probe dissolved in hybridization mixture per slide was hybridized to dena-
168
Cytogenet Genome Res 118:166–176 (2007)
tured metaphase chromosomes for 48 h at 37 ° C. Prior to hybridization, the labeled probe was annealed with a 100-fold excess of Cot-1 DNA at 37 ° C. Post-hybridization washing, detection of the probe with fluorescein isothiocyanate (FITC)-conjugated avidin, biotinylated anti-avidin antibody and FITC-avidin and counterstaining with 4ⴕ,6-diamidino-2-phenylindole (DAPI) were accomplished according to a standard protocol (Lichter and Cremer, 1992). Images of FITC and DAPI fluorescence were recorded separately using a Zeiss epifluorescence microscope equipped with appropriate excitation filters and a CCD camera. Digitalized images of the FITC and DAPI signals were overlaid using the Easy FISH 1.0 software (Applied Spectral Imaging, Edingen-Neckarhausen, Germany). A minimum of 15 metaphases per hybridization were examined to assess the location of the probe on specific chromosomes. In addition, human chromosome 11 paint obtained from a commercial source (HPW Diagnostics, Rabenau, Germany) was used to hybridize on metaphases of the BFTC909 cell line according to the manufacturer’s protocol.
PCR and sequencing High molecular weight genomic DNA was prepared using a saltingout technique. Amplification of FANCC, FANCG and FANCF exons was performed using Taq polymerase and, where available, published primer sets (FANCC: Gibson et al., 1993; FANCG: Auerbach et al., 2003). For FANCF primers see Table 2. PCR products were purified using the GFX PCR DNA and Gel Band Purification kit (Amersham Biosciences, Little Chalfont, UK). DNA sequencing of PCR products was performed using ABI-PRISM big-dye terminator chemistry on the ABI 310 instrument (Applied Biosystems, Darmstadt, Germany). Denaturing HPLC (dHPLC) Ten different DNAs from blood cells and the corresponding ten tumor-derived DNA preparations were analyzed. Heteroduplex formation of FANCC and FANCG PCR products was achieved by denaturating for 2 min at 96 ° C and cooling down to 4 ° C at 2 ° C/s. dHPLC
Table 2. Oligonucleotide primers used for genomic amplification and sequencing of FANCF
Primers for
Designation
Sequence (5ⴕ]3ⴕ)
Amplification
FA-F, –50 for FA-F, +50 rev
GCG GAT GTT CCA ATC AGT ACG CAC GAA GGC ATA TAT TTG GTG AGA
Sequencing
FA-F, –50 for FA-F, 123 for FA-F, 681 rev FA-F, 944 rev FA-F, +50 rev
GCG GAT GTT CCA ATC AGT ACG GCG CCA CAT CCA TCG GCG GTG GAT GCC GGG TTC CAA CTC CAG AGG CTT TGA AAC CTA TTG TGC CAC GAA GGC ATA TAT TTG GTG AGA
analysis was performed on a Transgenomic WAVE system using the DNASepTM column (Transgenomic, Flein, Germany). The melting characteristics and separation modus of the DNA fragments were predicted by use of the wavemakerTM software (version 4.1.44). 5 l of the PCR products were injected and separated. If PCR products of DNAs from patients’ blood cells suggested heteroduplex formation, the corresponding tumor-derived DNAs were also analyzed.
DNA methylation analysis The HpaII-restriction assay, as well as methylation-specific PCR (MS-PCR) were performed as described previously (Taniguchi et al., 2003). Bisulfite modification of genomic DNA employed the CpGenomeTM DNA Modification Kit (Q-Biogene, Heidelberg, Germany). For bisulfite sequencing, bisulfite-treated DNA was amplified as previously described (Florl et al., 2004) with primers FF345B (nt –345 to –315; 5ⴕ-GTTTAGAAAATTTTTATTTAAGGATA-3ⴕ) and FR27B (nt –6 to +24; 5ⴕ-ATCCAAATACTACAAAAAAAATTCCATAAA-3ⴕ) for 37 cycles at 52 ° C annealing temperature. PCR products were separated by agarose gel electrophoresis and cloned into the TA-vector pCR4TOPO (Invitrogen, Karlsruhe, Germany). Several clones from each sample were sequenced by standard methods. Comparative genomic hybridization (CGH) CGH was performed as described previously (Tönnies et al., 2003) with slight modifications. In brief, test and control DNA were differently labeled by nick translation using SpectrumGreen쏐-dUTP and SpectrumOrange쏐-dUTP (Vysis, Wiesbaden, Germany). 200 ng of labeled test DNA, 200 ng reference DNA, and 12.5 g Cot-1 DNA were co-precipitated, denatured, and hybridized to denatured normal male metaphase spreads. After incubation for 3 days at 37 ° C, standard posthybridization washes were performed. Metaphase images were evaluated as with the FISH studies. Image analysis and karyotyping were performed using the ISIS analysis system (Metasystems, Altlussheim, Germany). Spectral karyotyping (SKY) analysis SKY was performed on chromosome preparations prepared from the BFTC909 cell line. The human SKY-Paint DNA kit from Applied Spectral Imaging (Migdal Ha’Eemek, Israel) was hybridized to denatured metaphase spreads. After hybridization for 2 days at 37 ° C, slides were washed and haptenized probe sequences were detected following the manufacturer’s protocol. Slides were counterstained with DAPI and embedded in an antifade reagent (para-phenylenediamine). The multicolor hybridizations were visualized with the Spectral Cube system (SD200). Spectral analysis and classification was done with the SKY-View 1.6 software (Applied Spectral Imaging).
Results
Bladder carcinomas often show deletions of chromosome 9, apparently including regions that harbor the FA genes FANCC and FANCG. High sensitivity of these tumors to
MMC and cisplatin raised the questions whether these FA genes are indeed subject to allelic loss in such cases and if so, whether the remaining alleles contain inactivating mutations. According to a previous study using microsatellite markers between D9S168 and D9S158 (Kimura et al., 2001), DNA from ten bladder carcinoma tissues showed LOH involving the regions of interest of chromosome 9 in the vicinity of FANCC and FANCG. In our initial experiments, this tumor DNA and DNA from corresponding peripheral blood mononuclear cells was investigated for LOH at FANCC and FANCG by comparing germline and tumor DNA on direct sequencing (see Table 3). All base substitutions observed within FANCG were polymorphisms rather than bona fide mutations. These polymorphisms included IVS4 –18T/G, IVS5 +58C/T (rs17885726) and IVS12 +7A/G (rs17882272). Our study revealed LOH at these sites in five tumors (T1, T2, T5, T6 and T7), while one case showed retention of heterozygosity (T10). No LOH at FANCG was detectable in the other four tumors: two were non-informative cases (T8, T9), and in the two remaining cases (T3, T4), the markers could not be analyzed because of inadequate amounts of tumor DNA (Table 3). FANCG is located between the 9p STR markers D9S171 and D9S1862. All five tumors with LOH at FANCG showed also LOH of D9S171. Two of them (T2, T7) also exhibited LOH of D9S1862; in the three remaining (T1, T5, T6) this marker was not determined. Sequencing of FANCC in the DNA from the ten patients showed one case with retention of heterozygosity (IVS11 – 25A/C in T2), while all the remaining tumors were non-informative cases (Table 3). FANCC is located between markers D9S196 and D9S287 on 9q. Two tumors (T2, T7) displayed LOH involving both these markers, while a single tumor had LOH involving D9S196 but was non-informative for D9S287 (T4). Since our pilot study confirmed frequent LOH on chromosome 9 in bladder cancer involving at least the FANCG gene locus, as a next step functional testing of the FA/BRCA pathway in bladder cancer cell lines was performed via FANCD2 immunoblotting. Defects in any FA core complex gene (to which both FANCC and FANCG belong) result in lack of FANCD2 monoubiquitination after exposure to MMC or cisplatin. All 23 bladder cancer cell lines tested are known to carry extensive chromosomal aberrations which in the majority of cases involve chromosome 9 (Williams et al., 2002, 2005; Florl and Schulz, 2003). With the exception
Cytogenet Genome Res 118:166–176 (2007)
169
Table 3. LOH analysis in bladder carcinoma tissues
Marker
Position
T1
T2
T3
T4
T5
T6
T7
T8
T9
T10
D9S168 D9S157 D9S162 D9S942 D9S171 FANCG D9S1862 D9S15 D9S273 D9S153 D9S283 D9S196 FANCC D9S287 D9S180 D9S176 D9S53 D9S1872 D9S63 D9S1847 D9S158
10.578K 17.618K 19.669K 21.980K 24.524K 35.063K 68.373K 69.571K 69.768K 78.810K 89.643K 93.553K 94.490K 95.545K 97.729K 99.137K 104.641K 118.869K 129.734K 132.466K 136.325K
LOH LOH LOH RET LOH LOH n.d. LOH LOH n.d. LOH n.d. n.i. n.d. n.d. n.i. LOH LOH n.d. n.d. LOH
LOH RET LOH n.i. LOH LOH LOH LOH LOH LOH n.d. LOH RET LOH n.d. LOH n.d. LOH n.i. n.d. n.i.
LOH LOH LOH RET LOH n.d. n.d. n.i. LOH n.d. LOH n.d. n.i. n.d. n.d. LOH n.i. RET n.d. n.d. RET
LOH n.i. n.i. LOH LOH n.d. LOH n.i. n.d. LOH LOH LOH n.i. n.i. LOH n.i. n.i. LOH LOH LOH n.i.
LOH LOH LOH RET LOH LOH n.d. n.i. LOH n.d. LOH n.d. n.i. n.d. n.d. LOH LOH LOH n.d. n.d. LOH
LOH LOH LOH RET LOH LOH n.d. LOH LOH n.d. LOH n.d. n.i. n.d. n.d. LOH n.i. LOH n.d. n.d. LOH
n.i. n.i. RET RET LOH LOH LOH n.i. LOH LOH n.i. LOH n.i. LOH LOH n.i. LOH n.i. LOH n.i. n.i.
LOH n.i. LOH RET LOH n.i. n.d. LOH LOH n.d. LOH n.d. n.i. n.d. n.d. LOH LOH LOH n.d. n.d. LOH
MI RET n.i. LOH LOH n.i. n.d. n.i. LOH n.d. LOH n.d. n.i. n.d. n.d. n.i. n.i. LOH n.d. n.d. LOH
LOH n.i. LOH RET LOH RET n.d. LOH LOH n.d. LOH n.d. n.i. n.d. n.d. n.i. n.i. LOH n.d. n.d. LOH
Microsatellite markers were from Kimura et al. (2001). LOH: loss of heterozygosity; n.d.: not determined; n.i.: not informative; RET: retention of heterozygosity; MI: microsatellite instability.
of a single cell line (BFTC909), all cell lines tested responded normally to cisplatin treatment with the appearance of monoubiquitinated FANCD2 (FANCD2-L) in addition to the native isoform (FANCD2-S), indicating a fully functional FA core complex (Fig. 1A). Furthermore, prolonged cisplatin incubation times (16 h vs. 6 h) resulted in an increase of the monoubiquitinated FANCD2 isoform, suggesting regular function. In these cell lines, abrogation of the FA/BRCA pathway by mutations in FANCD1/BRCA2 or FANCJ/BRIP1, located downstream of FANCD2 (and therefore not necessary for FANCD2 monoubiquitination), were excluded via immunoblotting (data not shown). FANCN mutations were not excluded since the gene had not been reported at the time of this study. The single exceptional cell line BFTC909 displayed only the non-ubiquitinated FANCD2-S isoform suggesting a defect in one of the FA core complex genes. Since defective FA genes cause cellular sensitivity towards DNA damaging agents, the FANCD2 monoubiquitination-deficient cell line BFTC909 and six randomly chosen proficient bladder cancer cell lines were tested for MMC sensitivity resulting in G2 phase arrest. Figure 1B shows the cell cycle distribution of BFTC909 in comparison to the two proficient cell lines BFTC905 and RT112. BFTC909 proved higher sensitivity towards MMC as evidenced by a disproportionate elevation of the G2 phase cell cycle fraction. A paired t-test indicated statistical significance of this hypersensitivity, when the ratios of percentages of cells in G2 phase relative to cells in S phase were compared with and without MMC treatment (P = 0.014). In contrast to BFTC909,
170
Cytogenet Genome Res 118:166–176 (2007)
all other cell lines tested showed only minor increases of the G2:S ratio in response to MMC (Fig. 1C). Lack of the FANCD2-L isoform combined with increased sensitivity towards MMC is characteristic of a defect in one of the FA core complex genes. In order to determine which FA core complex gene might disrupt the pathway in BFTC909, the cell line was transduced with six known FA core complex genes (FANCA, -C, -E, -F, -G and -L). Surprisingly, transduction with the vector expressing FANCF in BFTC909 rather than FANCG or FANCC restored a normal MMC response on cell cycle analysis (Fig. 2A). Complementation of BFTC909 exclusively by FANCF was confirmed by FANCD2 immunoblotting (Fig. 2B). Only the FANCF-transduced cells showed both, the native and the monoubiquitinated isoforms of FANCD2, while transduction with all other FA genes failed to restore FANCD2 monoubiquitination. Sequencing of FANCF in BFTC909 failed to detect any mutation. Since silencing of FANCF via hypermethylation of the promoter region had been reported in other types of neoplasias (reviewed in Taniguchi and D’Andrea, 2006), we investigated the methylation status of FANCF in BFTC909, initially using a standard HpaII-restriction assay (Fig. 3A). In this line, FANCF was found to be resistant to digestion by HpaII, but sensitive to MspI. This suggested at least partial hypermethylation of FANCF. Methylation-specific PCR confirmed methylation of the BFTC909 FANCF promoter which turned out to be almost complete (Fig. 3B). With the same technique, two further cell lines, VMCub1 and 639V, and one out of 41 surgical bladder tumor specimens (TCC8),
Fig. 2. Retroviral complementation of BFTC909. (A) Flow cytometric analysis (DNA histograms). Cells were native or transduced with retroviral vectors separately expressing the FA genes FANCA, -C, -E, -F, -G or -L. Subsequently, cells were left untreated or treated with 10 ng/ml MMC for 16 h. Treatment of BFTC909 with MMC results in a prominent G2 phase arrest, indicating that this cell line is hypersensitive towards MMC. Only the transfer of FANCF was able to restore the G2 phase arrest to normal. (B) FANCD2 immunoblotting of the cell line BFTC909 after transduction with retroviral vectors separately expressing six FA genes of the nuclear core complex. Only FANCF was able to complement the defective FANCD2 monoubiquination as evidenced by the appearance of the FANCD2-L band. Fig. 1. Functional analysis of the FA/BRCA pathway. (A) FANCD2 immunoblotting of a primary urothelial cell culture (UP124) and the bladder carcinoma cell lines SD, BFTC909 and HT1376. Cells were treated with 1 g/ml cisplatin for 6 or 16 h. The blots reveal both nonubiquitinated (S) and monoubiquitinated (L) FANCD2 protein in the control cells UP124 as well as in bladder carcinoma cell lines SD and HT1376, with increasing levels of FANCD2-L upon longer treatment. In contrast, BFTC909 displays only the FANCD2-S band. (B) Flow cytometric analysis (DNA histograms) of three bladder cancer cell lines (BFTC905, BFTC909 and RT112) following incubation with 5 ng/ml MMC. In contrast to the cell lines BFTC905 and RT112, proficient for FANCD2 monoubiquitination, BFTC909 responded to MMC with a pronounced G2 phase arrest indicating hypersensitivity towards the crosslinking agent. (C) Bar graph showing the sensitivity of seven bladder cancer cell lines towards MMC, represented by the ratios of percentages of cells in G2 phase to cells in S phase after treatment with 5 ng/ml MMC (48 h).
revealed partial hypermethylation. However, there was no concomitant evidence for defective FANCD2 monoubiquitination in the two cell lines (data not shown). The hypermethylated BFTC909 cell line, one of the partially methylated cell lines (639V) and a non-methylated cell line (J82) were chosen for bisulfite sequencing in addition to normal peripheral blood mononuclear cells as control. As shown in Fig. 3C, nearly all CpG sites of the promoter region of FANCF in cell line BFTC909 were methylated. Partial hypermethylation of 639V and the unmethylated status of J82 were confirmed.
Cytogenet Genome Res 118:166–176 (2007)
171
Although there is no obvious breakpoint in 11p15, we additionally excluded microdeletions affecting FANCF at 11p15 by FISH using a FANCF probe. As shown in Fig. 4D, two signals each were detected for FANCF in control cells as well as in BFTC909 cells, indicating disomy for FANCF in BFTC909. In the NCBI single nucleotide polymorphism (SNP) database, four SNPs are described for the 1125-nt coding region of FANCF (rs11556562, rs11026706, rs7103674, rs7103293). In BFTC909, all of these are homozygous, suggesting loss of one allele early in tumor progression or evolution of the cell line and subsequent reduplication of the other allele. In order to investigate the prevalence of FANCF promoter hypermethylation in native bladder cancer tissues, DNA from 41 bladder cancer tissues previously studied for a range of methylation alterations (Neuhausen et al., 2006) was investigated by methylation-specific PCR of FANCF. While all samples yielded strong bands using primers specific for the unmethylated sequence, only a single case was weakly positive with primers specific for the methylated sequence (TCC8, see above). In this particular tumor, hypermethylation in several other genes together with pronounced global hypomethylation had been observed in the previous study (Neuhausen et al., 2006).
Fig. 3. Methylation analysis. (A) HpaII restriction assay of BFTC909 DNA. The HpaII restriction assay shows that the fragment +280 to +432 of FANCF is not restricted by HpaII in BFTC909, while it is restricted in a control DNA. U: untreated DNA; N: negative control. (B) MS-PCR analysis of FANCF. Bladder carcinoma cell lines and tumor tissues were analyzed for FANCF hypermethylation. Among the cell lines, only BFTC909 appeared almost completely methylated, whereas VMCub1 and 639V showed partial methylation. BFTC905, T24, HT1376, SW1710 and J82 are unmethylated. Leukocytes (Blood) and normal urothelial cells (UEC) served as controls. Among the three tumor tissues shown here (TCC8, 9 and 10), one (TCC8) showed weak methylation. Note that the figure is composed from three different gel runs. U: unmethylated, M: methylated, SssI: Blood DNA methylated in vitro with CpG-methyltransferase from Spiroplasma species. (C) Bisulfite sequencing analysis of the CpG island in the FANCF promoter (from –345 to +24; 21 CpGs). The methylation status in different bladder cancer cell lines was investigated. A control DNA derived from normal human blood showed no methylation in the FANCF CpG island, whereas the FA-deficient cell line BFTC909 was hypermethylated. Consistent with the results of methylation-specific PCR, the bladder cancer cell line 639V showed partial methylation and J82 essentially lacked methylated sites. Open circle: non-methylated CpG, filled circle: methylated CpG.
Chromosome 11, where FANCF is located, is represented in BFTC909 by one intact copy and one copy involved in a translocation, as demonstrated by chromosome 11 painting (Fig. 4A). Spectral karyotyping of BFTC909 revealed a hypo-tetraploid karyotype with a large number of chromosomal aberrations including a translocation between chromosomes 11 and 3 (Fig. 4B). CGH revealed an underrepresentation of 11p and of distal 11q relative to overall ploidy status, but retention of proximal 11q (Fig. 4C).
172
Cytogenet Genome Res 118:166–176 (2007)
Discussion
This is the first study of the FA/BRCA pathway in bladder cancer cells. Crosslinking compounds such as cisplatin or mitomycin C are frequent components of current chemotherapy regimens for bladder cancer. Their use is based on empirical clinical data, while an understanding of the biological basis of their efficacy is lacking (Lehmann et al., 2003). Since many bladder carcinomas display extensive deletions of chromosome 9 carrying the FA genes FANCC and FANCG, loss of these genes might have been responsible for hypersensitivity towards DNA crosslinking agents. In particular, FANCC is located at 9q22.3 within one of the most frequently deleted regions in bladder cancer (Knowles, 1999; Kimura et al., 2001). We therefore expected to detect genetic alterations of FANCC or FANCG in bladder carcinomas. Contrary to our expectation, neither LOH analysis and sequencing of the two genes in tumor tissues with LOH, nor FANCD2 immunoblotting in cell lines revealed consistent defects in FANCC or FANCG. Following treatment with DNA damaging agents, cells with defects in any of the FA core complex genes, including FANCC and FANCG, typically show prominent G2 phase accumulation and lack of monoubiquitination of FANCD2 (Garcia-Higuera et al., 2001; Shimamura et al., 2002). When tested for these FA-typical cell cycle and immunoblotting phenotypes, only a single cell line out of 23 bladder carcinoma cell lines displayed such a typical FA cellular feature. Again contrary to our expectation, neither FANCC nor FANCG were found to be defective in this cell line BFTC909. Instead, its FA-like phenotype was caused by silencing of the FANCF gene due to promoter hypermethyl-
A
C
B
D Fig. 4. Cytogenetic analysis of BFTC909. (A) Chromosome 11 painting. Chromosome 11 painting of the cell line BFTC909 showed only one intact chromosome 11. A second chromosome 11 was involved in a translocation. (B) Spectral karyotyping of BFTC909. The cell line revealed a hypo-tetraploid karyotype with a large number of chromosomal aberrations including a translocation between chromosomes 11 and 3. For chromosome 11, the cell line is disomic with no structural aberrations of 11p15, the gene locus of FANCF. (C) CGH ratio profile of BFTC909. CGH analysis shows underrepresentation of 11p and distal 11q relative to hypo-tetraploid status of the bladder cancer cell line with multiple structural aberrations. (D) Fluorescence in situ hybridization (FISH). In situ hybridization using a probe containing FANCF resulted in two specific signals on the short arms of the two chromosome 11 homologs (red arrows), showing that BFTC909 is disomic for FANCF.
Cytogenet Genome Res 118:166–176 (2007)
173
ation. Hypermethylation of FANCF has been described for several kinds of tumors, including cervical cancer (Narayan et al., 2004), ovarian cancer (Olopade and Wei, 2003; Taniguchi et al., 2003; Dhillon et al., 2004; Teodoridis et al., 2005; Wang et al., 2006), AML (Tischkowitz et al., 2003), head and neck squamous cell carcinoma (HNSCC) and non-small-cell lung cancer (NSCLC) (Marsit et al., 2004). Even though FANCF hypermethylation has been observed in a large variety of tumors, the proportion of cases affected by hypermethylation is relatively low in each of these tumors. FANCF is located on chromosome 11p15 adjacent to a region known to represent a hot-spot for hypermethylation (de Bustros et al., 1988; Feinberg, 1999). Dysregulation of imprinted genes in this region can be found in several tumors. A prominent example is the altered expression of CDKN1C resulting from promoter hypermethylation in the Beckwith-Wiedemann syndrome and in several other human cancers, including bladder cancer (Hoffmann et al., 2005). In this context, it is interesting that the only instance of FANCF hypermethylation in a bladder cancer tissue found in this study occurred in a tumor with multiple other hypermethylation events. These considerations raise the question whether hypermethylation of FANCF reflects a ‘bystander’ effect resulting from spreading of epigenetic modification of genes located in such hot spot regions, or whether the disruption of the FA/BRCA pathway via FANCF silencing is a specific event during tumorigenesis. Inactivation of FANCF causes genetic instability and might therefore be an early step in cancerogenesis, preferentially in FA-typical cancers that occur in non-FA patients (Tischkowitz et al., 2003). This concept has received strong support by the recent observation of frequent epigenetic silencing of another caretaker gene, WRN, in human tumors (Agrelo et al., 2006). However, other than squamous cell carcinoma of the upper digestive tract or the genital organs, bladder cancer is not a common type of cancer in FA (Alter, 2003). Our results show that FANCF hypermethylation is a rare event in bladder tumors since it was found in only one out of 23 established tumor cell lines, and rarely in native tumor tissues. Taniguchi et al. (2003) suggested that FANCF methylation may be found less than expected because secondary demethylation may render tumor cells resistant to DNA damaging agents. This hypothesis implies that hypermethylation accompanies tumor initiation. More typically, however, partial hypermethylation accompanies the early stages of tumor formation and methylation increases as the tumor progresses. This is particularly well documented in bladder cancer, where aberrant but weak hypermethylation is common in morphologically normal urothelial tissue of bladders carrying tumors (Dhawan et al., 2006; Neuhausen et al., 2006). The partial methylation of FANCF in two bladder carcinoma cell lines (VMCub1 and 639V) might reflect this standard situation. Secondary demethylation appears to be an unlikely explanation for the findings in bladder cancer, since the tumor samples studied were typically acquired prior to the initiation of chemotherapy.
174
Cytogenet Genome Res 118:166–176 (2007)
The physical and functional preservation of FANCC and FANCG in bladder cancer suggests that these two genes might be essential for survival of the tumor. This notion would fit with the general observation that both constitutional and somatic FA gene mutations are relatively rare in human neoplasia (reviewed in Lyakhovich and Surralles, 2006). Only a small minority of non-FA tumors were found to harbor mutations in FA genes, apart from BRCA2/FANCD1 and the low-penetrance breast cancer genes BRIP1/FANCJ and PALB2/FANCN (Seal et al., 2006; Rahman et al., 2007). Possible exceptions are AML with described mutations in FANCA and FANCC as well as pancreatic cancer where also several sequence changes in FANCA, FANCC and FANCG have been described (van der Heijden et al., 2003; Couch et al., 2005). However, other researchers found no convincing evidence for frequent occurrence of FA gene mutations in non-FA AML (reviewed in Neveling et al., 2007). In addition, it seems that germline FANCA mutations do not contribute to familial pancreatic cancer susceptibility and that FANCC and FANCG mutations may overall have a comparatively low penetrance, if any, for the pancreatic cancer phenotype. The paucity of involvement of FA genes in human neoplastic disease suggests that an intact FA pathway might be important for tumor cell survival. Our data shows that although tumors typically exhibit a high rate of mutations, in all but a single bladder cancer cell line the FA genes were found to be functional, including those FA genes that are located within the frequently deleted regions of chromosome 9. Moreover, if disruption of the FA pathway were a first and general step during tumor development, we would expect a much wider range of tumor types in FA patients as is actually observed. Conversely, if disruption of the FA core complex should favor the development of rather specific tumor types, hypermethylation of FANCF should be restricted to these particular tumors. However, hypermethylation of FANCF has been observed in a wide variety of tumors other than those typically seen in FA (Taniguchi and D’Andrea, 2006). Taken together, these arguments lead to the conclusion that although reduction or elimination of a DNA repair system of a given cell may facilitate the generation of mutations in tumor suppressor genes, disruption of the FA/BRCA pathway may be disadvantageous for tumor cells by rendering them vulnerable to the adverse effects of DNA damaging agents and oxidative stress. In contrast, retention of a functional FA/BRCA pathway might be important for tumor cells to retain a certain human cell character despite chromosomal rearrangements, gains and losses and the resulting consequences. We suggest that silencing of the FANCF gene via hypermethylation is a secondary and noncausal event in tumorigenesis, resulting primarily from the location of FANCF within a known hotspot region for methylation at 11p15. This would also imply that the therapeutic effectiveness of crosslinking agents used for bladder cancer chemotherapy does not result from defects in the FA DNA repair pathway in most of these tumors. However, if disrup-
tion of the FA pathway is present even though in a low proportion of bladder carcinomas, future clinical studies must show if this would influence prognosis or open additional therapeutic options.
Acknowledgements We thank Birgit Gottwald, Würzburg, for excellent cell culture work, Alessandra Baumer, Zurich, for helpful comments and assistance concerning methylation analysis and Dr. Irit Bar-Am, Applied Spectral Imaging, Migdal Ha’Eemek, Israel for excellent help and cooperation with the spectral karyotyping analysis.
References Agrelo R, Cheng WH, Setien F, Ropero S, Espada J, et al: Epigenetic inactivation of the premature aging Werner syndrome gene in human cancer. Proc Natl Acad Sci USA 103: 8822–8827 (2006). Alter BP: Cancer in Fanconi anemia, 1927–2001. Cancer 97: 425–440 (2003). Auerbach AD, Greenbaum J, Pujara K, Batish SD, Bitencourt MA, et al: Spectrum of sequence variation in the FANCG gene: an International Fanconi Anemia Registry (IFAR) study. Hum Mutat 21:158–168 (2003). Borden LS Jr, Clark PE, Hall MC: Bladder cancer. Curr Opin Oncol 17:275–280 (2005). Bruch J, Wohr G, Bruderlein S, Barbi G, Wolter H, et al: Detailed marker chromosome analysis in cell line U-BLC1, established from transitionalcell carcinoma of the bladder. Int J Cancer 80: 903–910 (1999). Couch FJ, Johnson MR, Rabe K, Boardman L, McWilliams R, et al: Germ line Fanconi anemia complementation group C mutations and pancreatic cancer. Cancer Res 65: 383–386 (2005). de Bustros A, Nelkin BD, Silverman A, Ehrlich G, Poiesz B, Baylin SB: The short arm of chromosome 11 is a ‘hot spot’ for hypermethylation in human neoplasia. Proc Natl Acad Sci USA 85: 5693–5697 (1988). Dhawan D, Hamdy F, Rehman I, Patterson J, Cross SS, et al: Evidence for the early onset of aberrant promoter methylation in urothelial carcinoma. J Pathol 209: 336–343 (2006). Dhillon VS, Shahid M, Husain SA: CpG methylation of the FHIT, FANCF, cyclin-D2, BRCA2 and RUNX3 genes in granulosa cell tumors (GCTs) of ovarian origin. Mol Cancer 3: 33 (2004). Elliott AY, Cleveland P, Cervenka J, Castro AE, Stein N, et al: Characterization of a cell line from human transitional cell cancer of the urinary tract. J Natl Cancer Inst 53: 1341–1349 (1974). Elliott AY, Bronson DL, Stein N, Fraley EE: In vitro cultivation of epithelial cells derived from tumors of the human urinary tract. Cancer Res 36:365–369 (1976). Feinberg AP: Imprinting of a genomic domain of 11p15 and loss of imprinting in cancer: an introduction. Cancer Res 59:1743–1746 (1999). Florl AR, Schulz WA: Peculiar structure and location of 9p21 homozygous deletion breakpoints in human cancer cells. Genes Chromosomes Cancer 37: 141–148 (2003). Florl AR, Steinhoff C, Muller M, Seifert HH, Hader C, Engers R, et al: Coordinate hypermethylation at specific genes in prostate carcinoma precedes LINE-1 hypomethylation. Br J Cancer 91:985–994 (2004). Garcia-Higuera I, Taniguchi T, Ganesan S, Meyn MS, Timmers C, et al: Interaction of the Fanconi anemia proteins and BRCA1 in a common pathway. Mol Cell 7: 249–262 (2001). Gibson RA, Buchwald M, Roberts RG, Mathew CG: Characterisation of the exon structure of the Fanconi anaemia group C gene by vectorette PCR. Hum Mol Genet 2:35–38 (1993).
Grossman HB, Wedemeyer G, Ren L, Wilson GN, Cox B: Improved growth of human urothelial carcinoma cell cultures. J Urol 136: 953–959 (1986). Hanenberg H, Batish SD, Pollok KE, Vieten L, Verlander PC, et al: Phenotypic correction of primary Fanconi anemia T cells with retroviral vectors as a diagnostic tool. Exp Hematol 30: 410–420 (2002). Hoffmann MJ, Florl AR, Seifert HH, Schulz WA: Multiple mechanisms downregulate CDKN1C in human bladder cancer. Int J Cancer 114:406– 413 (2005). Howlett NG, Taniguchi T, Olson S, Cox B, Waisfisz Q, et al: Biallelic inactivation of BRCA2 in Fanconi anemia. Science 297: 606–609 (2002). Hu N, Li G, Li WJ, Wang C, Goldstein AM, et al: Infrequent mutation in the BRCA2 gene in esophageal squamous cell carcinoma. Clin Cancer Res 8:1121–1126 (2002). Joenje H, Patel KJ: The emerging genetic and molecular basis of Fanconi anaemia. Nat Rev Genet 2:446–457 (2001). Joenje H, Arwert F, Eriksson AW, de Koning H, Oostra AB: Oxygen-dependence of chromosomal aberrations in Fanconi’s anaemia. Nature 290:142–143 (1981). Keen AJ, Knowles MA: Definition of two regions of deletion on chromosome 9 in carcinoma of the bladder. Oncogene 9: 2083–2088 (1994). Kennedy RD, D’Andrea AD: The Fanconi anemia/ BRCA pathway: new faces in the crowd. Genes Dev 19:2925–2940 (2005). Kimura F, Florl AR, Seifert HH, Louhelainen J, Maas S, et al: Destabilization of chromosome 9 in transitional cell carcinoma of the urinary bladder. Br J Cancer 85: 1887–1893 (2001). Knowles MA: The genetics of transitional cell carcinoma: progress and potential clinical application. BJU Int 84:412–427 (1999). Knowles MA: What we could do now: molecular pathology of bladder cancer. Mol Pathol 54: 215–221 (2001). Knowles MA: Molecular subtypes of bladder cancer: Jekyll and Hyde or chalk and cheese? Carcinogenesis 27: 361–373 (2006). Kyriazis AA, Kyriazis AP, McCombs WB, 3rd, Peterson WD Jr: Morphological, biological, and biochemical characteristics of human bladder transitional cell carcinomas grown in tissue culture and in nude mice. Cancer Res 44:3997– 4005 (1984). Lehmann J, Retz M, Stockle M: Is there standard chemotherapy for metastatic bladder cancer? Quality of life and medical resources utilization based on largest to date randomized trial. Crit Rev Oncol Hematol 47: 171–179 (2003). Levran O, Attwooll C, Henry RT, Milton KL, Neveling K, et al: The BRCA1-interacting helicase BRIP1 is deficient in Fanconi anemia. Nat Genet 37:931–933 (2005). Lichter P, Cremer T: Chromosome analysis by nonisotopic in situ hybridization, in Rooney DE, Czepulkowski BH (eds): Human Cytogenetics, A Practical Approach, vol I, pp 157–192 (IRL Press, Oxford 1992).
Lin CW, Lin JC, Prout GR: Establishment and characterization of four human bladder tumor cell lines and sublines with different degrees of malignancy. Cancer Res 45: 5070–5079 (1985). Lyakhovich A, Surralles J: Disruption of the Fanconi anemia/BRCA pathway in sporadic cancer. Cancer Lett 232: 99–106 (2006). Marsit CJ, Liu M, Nelson HH, Posner M, Suzuki M, Kelsey KT: Inactivation of the Fanconi anemia/ BRCA pathway in lung and oral cancers: implications for treatment and survival. Oncogene 23:1000–1004 (2004). Masters JR: Human cancer cell lines: fact and fantasy. Nat Rev Mol Cell Biol 1: 233–236 (2000). Narayan G, Arias-Pulido H, Nandula SV, Basso K, Sugirtharaj DD, et al: Promoter hypermethylation of FANCF: disruption of Fanconi anemiaBRCA pathway in cervical cancer. Cancer Res 64:2994–2997 (2004). Neuhausen A, Florl AR, Grimm MO, Schulz WA: DNA methylation alterations in urothelial carcinoma. Cancer Biol Ther 5: 993–1001 (2006). Neveling K, Kalb R, Schindler D: Cancer in Fanconi anemia and Fanconi anemia genes in cancer, in Schindler D, Hoehn H (eds): Monographs in Human Genetics, vol. 15, pp 59–78 (S. Karger, Basel 2007). Niedernhofer LJ, Lalai AS, Hoeijmakers JH: Fanconi anemia (cross)linked to DNA repair. Cell 123: 1191–1198 (2005). Olopade OI, Wei M: FANCF methylation contributes to chemoselectivity in ovarian cancer. Cancer Cell 3: 417–420 (2003). O’Toole C, Nayak S, Price Z, Gilbert WH, Waisman J: A cell line (SCABER) derived from squamous cell carcinoma of the human urinary bladder. Int J Cancer 17: 707–714 (1976). O’Toole C, Price ZH, Ohnuki Y, Unsgaard B: Ultrastructure, karyology and immunology of a cell line originated from a human transitional-cell carcinoma. Br J Cancer 38:64–76 (1978). Paulie S, Hansson Y, Lundblad ML, Perlmann P: Lectins as probes for identification of tumorassociated antigens on urothelial and colonic carcinoma cell lines. Int J Cancer 31: 297–303 (1983). Rahman N, Seal S, Thompson D, Kelly P, Renwick A, et al: PALB2, which encodes a BRCA2-interacting protein, is a breast cancer susceptibility gene. Nat Genet 39: 165–167 (2007). Rasheed S, Gardner MB, Rongey RW, Nelson-Rees WA, Arnstein P: Human bladder carcinoma: characterization of two new tumor cell lines and search for tumor viruses. J Natl Cancer Inst 58:881–890 (1977). Reid S, Schindler D, Hanenberg H, Barker K, Hanks S, et al: Biallelic mutations in PALB2 cause Fanconi anemia subtype FA-N and predispose to childhood cancer. Nat Genet 39: 162–164 (2007). Rigby CC, Franks LM: A human tissue culture cell line from a transitional cell tumour of the urinary bladder: growth, chromosome pattern and ultrastructure. Br J Cancer 24: 746–754 (1970).
Cytogenet Genome Res 118:166–176 (2007)
175
Sasaki MS, Tonomura A: A high susceptibility of Fanconi’s anemia to chromosome breakage by DNA cross-linking agents. Cancer Res 33: 1829–1836 (1973). Schulz WA: Understanding urothelial carcinoma through cancer pathways. Int J Cancer 119: 1513–1518 (2006). Seal S, Thompson D, Renwick A, Elliott A, Kelly P, et al: Truncating mutations in the Fanconi anemia J gene BRIP1 are low-penetrance breast cancer susceptibility alleles. Nat Genet 38: 1239–1241 (2006). Shimamura A, de Oca RM, Svenson JL, Haining N, Moreau LA, et al: A novel diagnostic screen for defects in the Fanconi anemia pathway. Blood 100:4649–4654 (2002). Swiatkowski S, Seifert HH, Steinhoff C, Prior A, Thievessen I, et al: Activities of MAP-kinase pathways in normal uroepithelial cells and urothelial carcinoma cell lines. Exp Cell Res 282: 48–57 (2003). Taniguchi T, D’Andrea AD: The molecular pathogenesis of Fanconi anemia: recent progress. Blood 107:4223–4233 (2006). Taniguchi T, Tischkowitz M, Ameziane N, Hodgson SV, Mathew CG, et al: Disruption of the Fanconi anemia-BRCA pathway in cisplatinsensitive ovarian tumors. Nat Med 9: 568–574 (2003).
176
Teodoridis JM, Hall J, Marsh S, Kannall HD, Smyth C, et al: CpG island methylation of DNA damage response genes in advanced ovarian cancer. Cancer Res 65: 8961–8967 (2005). Thompson LH, Hinz JM, Yamada NA, Jones NJ: How Fanconi anemia proteins promote the four Rs: replication, recombination, repair, and recovery. Environ Mol Mutagen 45: 128–142 (2005). Tischkowitz M, Ameziane N, Waisfisz Q, De Winter JP, Harris R, et al: Bi-allelic silencing of the Fanconi anaemia gene FANCF in acute myeloid leukaemia. Br J Haematol 123: 469–471 (2003). Tönnies H, Poland J, Sinha P, Lage H: Association of genomic imbalances with drug resistance and thermoresistance in human gastric carcinoma cells. Int J Cancer 103: 752–758 (2003). Tzeng CC, Liu HS, Li C, Jin YT, Chen RM, et al: Characterization of two urothelium cancer cell lines derived from a blackfoot disease endemic area in Taiwan. Anticancer Res 16: 179–1804 (1996).
Cytogenet Genome Res 118:166–176 (2007)
van der Heijden MS, Yeo CJ, Hruban RH, Kern SE: Fanconi anemia gene mutations in young-onset pancreatic cancer. Cancer Res 63: 2585–2588 (2003). Wang Z, Li M, Lu S, Zhang Y, Wang H: Promoter hypermethylation of FANCF plays an important role in the occurrence of ovarian cancer through disrupting Fanconi anemia-BRCA pathway. Cancer Biol Ther 5: 256–260 (2006). Williams RD: Human urologic cancer cell lines. Invest Urol 17: 359–363 (1980). Williams SV, Sibley KD, Davies AM, Nishiyama H, Hornigold N, et al: Molecular genetic analysis of chromosome 9 candidate tumor-suppressor loci in bladder cancer cell lines. Genes Chromosomes Cancer 34: 86–96 (2002). Williams SV, Adams J, Coulter J, Summersgill BM, Shipley J, Knowles MA: Assessment by M-FISH of karyotypic complexity and cytogenetic evolution in bladder cancer in vitro. Genes Chromosomes Cancer 43: 315–328 (2005). Xia B, Sheng Q, Nakanishi K, Ohashi A, Wu J, et al: Control of BRCA2 cellular and clinical functions by a nuclear partner, PALB2. Mol Cell 22: 719–729 (2006). Xia B, Dorsman JC, Ameziane N, de Vries Y, Rooimans MA, et al: Fanconi anemia is associated with a defect in the BRCA2 partner PALB2. Nat Genet 39:159–161 (2007).
2005). Non-random chromosomal changes are likely to have biologic significance and several of the genetic findings are discussed in more detail below. Early cytogenetic changes
Fig. 1. Diagram showing the various stages of bladder cancer. pTa lesions are superficial papillary lesions; CIS are flat superficial tumors; pT1 tumors invade into the subepithelial connective tissue; pT2 invade muscle tissue; pT3 lesions invade perivesicular tissue; pT4 tumors break through the bladder wall and invade surrounding tissues.
vival. The cancers are further classified based upon the extent of invasion into the surrounding tissues (Fig. 1). Papillary tumors are often non- or minimally invasive and are associated with regions of adjacent hyperplasia. These tumors rarely invade into the lamina propria, but are often multifocal. Carcinoma in situ (CIS) is also a superficial lesion; however, typically there is an associated high grade dysplasia and an increased risk for progression. Many flat tumors present clinically when invasion has already occurred (pT2–pT3). Patients with invasive tumors have a high risk for metastasis and death (Reuter, 2006). While the subclassification is useful for describing the tumors in general, the specific prognosis of each tumor cannot be predicted based solely upon the grading system. Thus, it is important to understand the underlying genetic aberrations that define the subtypes of TCC. Cytogenetic analyses have been useful in identifying genetic aberrations involved in the initiation and progression of bladder cancer. Non random changes occurring in the various subtypes have guided molecular studies that have uncovered pertinent tumor suppressor genes and oncogenes (Cordon-Cardo et al., 2000; Knowles, 2006; MhawechFauceglia et al., 2006). Approximately 200 cases of urothelial cancer have been studied using routine cytogenetic banding analysis. In general, superficial tumors tend to be near-diploid with relatively few chromosomal changes, while muscle-invasive tumors are polyploid with multiple rearrangements and complex karyotypes (Fadl-Elmula, 2005). The most common autosomal imbalances reported in karyotypic analyses of bladder cancer were losses of entire chromosomes 4, 9, 10, 14, 15, 16, 18, 21, 22 and gains of chromosomes 7, 16, 19 and 20 (Fadl-Elmula, 2005). In particular, losses of 1p, 5q, 8p, 9p, 11p, 17p and gains of 1q, 3q, 5p, 8q, 13q and 17q were frequently noted (Fadl-Elmula,
178
Cytogenet Genome Res 118:177–181 (2007)
Karyotypic assessment of early superficial tumors is relatively rare due to the difficulty in obtaining and culturing such tumors (Meloni et al., 1993). The limited studies reveal surprisingly few visible chromosomal aberrations in these solid tumors (Sandberg, 1992). Loss of the entire chromosome 9 or portions of chromosome 9 is the most common cytogenetic aberration detected in early cancers and is often the sole cytogenetic abnormality (Fadl-Elmula, 2005). These findings suggest that loss of tumor suppressor genes on chromosome 9 are early events, important in the initiation of tumorigenesis (Cairns et al., 1993; Simoneau et al., 2000). Loss of heterozygosity (LOH) of loci on both 9p and 9q in urothelial hyperplasia and dyplasia, both presumed precursors to malignancy, provide further support that loss of tumor suppressors on chromosome 9 are important in the development of urothelial neoplasia (Hartmann et al., 1999; van Oers et al., 2006). Mapping studies have localized the smallest regions of chromosome 9 loss to 9p21, 9q22, 9q31]q32 and 9q34 (Habuchi et al., 1995; Packenham et al., 1995; Williamson et al., 1995; Simoneau et al., 1999). While the associated tumor suppressor genes have not been definitively identified, several candidate genes have been suggested, including p15 (CDKN2B) and p16 (CDKN2A) at 9p21 (Packenham et al., 1995; Williamson et al., 1995) and the Gorlin syndrome (PTCH) and the tuberous sclerosis (TSC1) genes at 9q22 and 9q34, respectively (Habuchi et al., 1995). The fact that there are so few visible genetic aberrations suggests that submicroscopic mutations and/or epigenetic events are involved in initiation of papillary TCC. The most significant submicroscopic finding in superficial bladder cancer is mutation of the fibroblast growth factor receptor 3 (FGFR3) gene. Cappellen and colleagues (1999) studied the expression pattern of FGFR3 in bladder and cervical cancer specimens and found a striking number of cancers with mutant transcripts. Subsequent studies confirmed that FGFR3 mutations were common, particularly in superficial bladder tumors with a low risk for recurrence (Billerey et al., 2001; Van Rhijn et al., 2001, 2003; Lindgren et al., 2006). FGFR3 mutations were also reported in urothelial papillomas, suggesting that these mutations may predispose to papillary growth (Chow et al., 2000; Van Rhijn et al., 2002). Deletion of chromosome 9p and 9q and mutations of FGFR3 are the earliest and most common genetic aberrations known for superficial, low grade papillary tumors (pTa) tumors. LOH and comparative genomic hybridization (CGH) studies have also suggested that losses of 10q and 11p, and gains of 1q, 17 and 20q may be seen in pTa bladder cancers (Kallioniemi et al., 1995; Zhao et al., 1999). However, the additional findings are more common in higher grade carcinomas (pTaG2–3) (Zhao et al., 1999).
Cytogenetics of advanced stage urothelial cancer
Understanding the genetic events leading to bladder cancer progression is important for decreasing overall morbidity and mortality associated with the disease. Cytogenetic analyses have shown that invasive tumors typically have more genetic aberrations than superficial cancers (Richter et al., 1999; Fadl-Elmula, 2005). The presence of multiple tumors is common in patients and, although advanced stage cancer within the same person often exhibits divergent chromosomal abnormalities, the tumors can be traced to an original clone (Sidransky et al., 1992; van Tilborg et al., 2000). The genetic diversity seen with advanced disease usually originates from chromosomal instability (Takahashi et al., 1998; Kawamura et al., 2003; Yamamoto et al., 2006), although microsatellite instability has also been identified in a minority of cases (Yamamoto et al., 2006). Definition of the genetic pathways that lead to more aggressive disease is needed to identify markers that might serve as predictors of outcome and/or targets for therapy. While pTa and CIS-pT1 tumors (Fig. 1) have been described as superficial lesions, genetic studies have revealed considerable differences between these types of tumors (Richter et al., 1997, 1998; Simon et al., 1998; Placer et al., 2005). The abnormalities that are more frequently seen in CIS and pT1 tumors compared with pTa tumors are also more common in pT2–pT4 carcinomas, while only a few cytogenetic aberrations are less frequent in pT1 than in invasive pT2–T4 lesions (Richter et al., 1999). These findings suggest that, although pT1 tumors are superficial, these types of tumors have accumulated many of the genetic aberrations that are associated with progression to deeply invasive TCC. The chromosomal areas that are commonly altered in pT1–pT4 lesions include losses of 2q, 3p, 5q, 8p, 10q, 11q, 13q, 17p, 18q and gains of 1q, 5p, 6p, 8q, 10p, 17q, 20q (Richter et al., 1998, 1999; Simon et al., 2000; Hoglund et al., 2001; Yamamoto et al., 2006). Only a subset of the abnormalities have been associated with progression of disease and are reportedly overrepresented in invasive pT2–pT4 tumors (3p+, 3q+, 4p–, 5q–, 5p+, 6q–, 7p+, 10q–, –15, –18, 18p+ and –22) (Richter et al., 1998, 1999; Simon et al., 2000; Yamamoto et al., 2006). The finding of non-random cytogenetic aberrations in advanced stage disease is consistent with the multi step process of carcinogenesis and suggests that the areas discussed above play a role in cancer progression. Cytogenetic studies of advanced stage disease have been important for identifying potential tumor suppressor genes and oncogenes. Oncogenes that have been suggested to play a role in the pathogenesis of advanced stage cancer include TRIO (5p15), E2F3 (6p22), MYC (8q24), ERBB2 (17q21), and AURKA (20q13); while potential tumor suppressor genes are FHIT (3p11), RAF1 (3p25), SFRP1 (8p12), and ING1 (13q34) (reviewed in Mhawech-Fauceglia et al., 2006). Additional oncogenes and tumor suppressor genes that have been associated with disease progression include the delta TP73L (3q27) and EGFR genes (7p12) and the RB1 and TP53 genes, respectively (reviewed in Cordon-Cardo et al., 2000;
Mhawech-Fauceglia et al., 2006). Many of the genes that have been linked to progression of bladder cancer (e.g. epidermal growth factor receptors, cell cycle regulators) have also been reported to be associated with other epithelial carcinomas, including breast and colon cancers. Suggested pathways for bladder cancer
One of the major contributions that cytogenetics has made in the study of bladder cancer is the definition of the two clinically important subtypes. Tumors that are relatively benign, but often recur, have few cytogenetic changes and a stable karyotype, while tumors that display an aggressive phenotype with a propensity to progress exhibit genetic instability and have certain non-random cytogenetic changes. Numerous pathways have been proposed to link the genetic aberrations with the specific cancer type and to establish the relationship between the two subtypes (Sandberg, 2002). Based on the current genetic information, it would appear that loss of chromosome 9 and mutation in FGFR3 are early changes that are associated with the transition of normal urothelium to hyperplasia to low grade papillary tumors (Knowles, 2006). Alternatively, dysplastic urothelium transforms to superficial CIS flat bladder cancers that have many cytogenetic aberrations. These cancers have a significant risk to progress to invasive carcinomas with changes in important cancer genes such as RB1 and TP53 (Cordon-Cardo et al., 2000). However, from the cytogenetic studies, it is difficult to extrapolate the initial genetic changes associated with both subtypes of TCC. In an attempt to computer model clonal relationships and order the accumulation of changes, Hoglund et al. (2001) used principal component statistical analysis of 200 bladder cancers. The study found that there were two distinct pathways for TCC: one pathway that was correlated with pTa–pT2 involved loss of chromosome 9 as the initial event, followed by 11p– and 1q+, and a second pathway that was associated with advanced stage disease (pT1–pT3) showed +7 as the initiating event, followed by 8p– and 8q+. While this study provided compelling evidence for cytogenetic aberrations being associated with various stages of disease, the authors acknowledged that not all of the steps in the multi-step process could be accounted for in the noted visible cytogenetic alterations (Hoglund et al., 2001). The majority of pathways that have been proposed for bladder cancer do not completely explain the genetic basis of tumor initiation and progression (see Sandberg, 2002 for review). Using cytogenetics to predict TCC
While it is useful to study the genetics of TCC to understand the etiology and to provide information on potential treatments, currently cytogenetic studies are used to detect cases of bladder cancer, both for diagnosis and for recurrences. Given the propensity of TCC to progress to higher stages of disease it is vital to detect new cases, as well as re-
Cytogenet Genome Res 118:177–181 (2007)
179
Fig. 2. An example of FISH studies (UroVysionTM, Abbott Molecular/Abbott Laboratories, Inc., Des Plaines, IL) on a urine sample from a patient with a history of bladder cancer being monitored for disease recurrence. The three smaller cells on the left show a normal signal pattern for the four probes: 2 red (chromosome 3 centromere probe), 2 green (chromosome 7 centromere probe), 2 gold (9p21 locus specific probe) and 2 aqua (chromosome 17 centromere probe). The large cell on the right has an abnormal signal pattern consistent with polysomy (3 red, 4 green, 3 aqua) and homozygous loss of 9p21 (no gold signals).
currences, as early as possible. The current standard methods used to diagnose and monitor for bladder cancer are urine cytology and cystoscopy. Cystoscopy has a specificity of 190% but is an invasive procedure that can miss high grade flat lesions (Walker et al., 1993; Saad et al., 2002). Cytology provides a noninvasive alternative, however, urine cytology lacks diagnostic sensitivity, particularly for low grade papillary neoplasms (Bastacky et al., 1999; Halling et al., 2000). The study of genetic aberrations commonly associated with TCC provides an objective, noninvasive assessment of urine samples or bladder washings that can be used for diagnosing and detecting recurrent disease. Sandberg (1992) originally proposed that the use of fluorescence in situ hybridization (FISH) studies using probes for –9, +7, –10 and loss of the Y would be useful to establish a diagnosis or recurrence of early bladder cancers and his subsequent pilot study validated the utility of the approach (Meloni et al., 1993). A more recent study of nine chromosomal markers (Sokolova et al., 2000) showed probes that predicted polysomy of chromosomes 3, 7 and 17, and deletion of 9p21 provided a sensitive and specific test, detecting 95% of recurrent TCCs (Fig. 2). Subsequently, Halling et al. (2000) established that a threshold of five or more cells with polysomy was 84% sensitive and 92% specific for detecting recurrent urothelial cancer.
Comparisons of urine cytology with FISH for detecting bladder cancer recurrence showed a greater sensitivity for FISH (Halling et al., 2000; Bubendorf et al., 2001; Veeramachaneni et al., 2003; Laudadio et al., 2005). FISH was 42– 83% sensitive for detecting pTa and pT1 lesions and 92–100% sensitive for pT2–4 invasive lesions in patients with known bladder cancer, while urine cytology had sensitivities of 24– 50% for pTa and pT1 lesions, and 78–85% for pT2–4 invasive lesions (Halling et al., 2000; Bubendorf et al., 2001; Skacel et al., 2003; Laudadio et al., 2005). For suspected new cases of TCC, cytology had a reported diagnostic sensitivity of 48% and FISH’s sensitivity was suggested to be as high as 100% (Saad et al., 2002; Laudadio et al., 2005). Conclusion
In conclusion, the cytogenetic studies of bladder cancer have played an important role in understanding the various subtypes of TCC and have also provided a clinical assessment tool for monitoring patients. However, to fully understand the genetics of TCC, evaluation of expression arrays and gene control mechanisms (methylation, acetylation, etc) are needed to augment the chromosomal-based analyses. The chromosomal analyses have laid the foundation on which the further studies may be built.
References American Cancer Society: Cancer Facts and Figures 2006. (American Cancer Society, Atlanta 2006) (www.cancer.org/docroot. Last accessed Sept 7, 2006). Bastacky S, Ibrahim S, Wilczynski S, Murphy W: The accuracy of urinary cytology in daily practice. Cancer 87: 118–128 (1999). Billerey C, Chopin D, Aubriot-Lorton MH, Ricol D, Gil Diez de Medina S, et al: Frequent FGFR3 mutations in papillary non-invasive bladder (pTa) tumors. Am J Pathol 158: 1955–1959 (2001).
180
Bubendorf L, Grilli B, Sauter G, et al: Multiprobe FISH for enhanced detection of bladder cancer in voided urine specimens and bladder washings. Am J Clin Path 116:79–86 (2001). Cairns P, Shaw ME, Knowles MA: Initiation of bladder cancer may involve deletion of a tumour-suppressor gene on chromosome 9. Oncogene 8:1083–1085 (1993). Cappellen D, De Oliveria C, Ricol D, de Medina SGD, Bourdin J, et al: Frequent activating mutations of FGFR3 in human bladder and cervix carcinomas. Nat Genet 23: 18–20 (1999). Chow NH, Cairns P, Eisenberger CF, Schoenberg MP, Taylor DC, et al: Papillary urothelial hyperplasia is a clonal precursor to papillary transitional cell bladder cancer. Int J Cancer 89: 514–518 (2000).
Cytogenet Genome Res 118:177–181 (2007)
Cordon-Cardo C, Cote R, Sauter G: Genetic and molecular markers of urothelial premalignancy and malignancy. Scand J Urol Nephrol Suppl 205:82–93 (2000). Fadl-Elmula I: Chromosomal changes in uroepithelial carcinomas. Cell Chromosome 4: 1–9 (2005). Habuchi T, Devlin J, Elder PA, Knowles MA: Detailed deletion mapping of chromosome 9q in bladder cancer: evidence for two tumour suppressor loci. Oncogene 11: 1671–1674 (1995). Halling KC, King W, Sokolova I, et al: A comparison of cytology and fluorescence in situ hybridization for the detection of urothelial carcinoma. J Urol 164:1768–1775 (2000).
Hartmann A, Schlake G, Zaak, D, Hungerhuber E, Hofstetter A, et al: Occurrence of chromosome 9 and p53 alterations in multifocal dysplasia and carcinoma in situ of human urinary bladder. Cancer Res 62:809–818 (1999). Hoglund M, Sall T, Heim S, Mitelman F, Mandahl N, Fadl-Elmula I: Identification of cytogenetic subgroups and karyotypic pathways in transitional cell carcinoma. Cancer Res 61:8241–8246 (2001). Kallioniemi A, Kallioniemi OP, Citro G, Sauter G, DeVries S, et al: Identification of gains and losses of DNA sequences in primary bladder cancer by comparative genomic hybridization. Genes Chromosomes Cancer 12: 213–219 (1995). Kawamura K, Moriyama M, Shiba N, et al: Centrosome hyperamplification and chromosomal instability in bladder cancer. Eur Urol 43:505–515 (2003). Knowles M: Molecular subtypes of bladder cancer: Jekyll and Hyde or chalk and cheese? Carcinogenesis 27: 361–373 (2006). Laudadio J, Keane T, Reeves HM, Savage SJ, Hoda RS, et al: Fluorescence in situ hybridization for detecting transitional cell carcinoma: implications for clinical practice. Br J Urol Int 96:1280– 1285 (2005). Lindgren D, Liedberg F, Andersson A, Chebil G, Gudjonsson S, et al: Molecular characterization of early-stage bladder carcinomas by expression profiles, FGFR3 mutation status, and loss of 9q. Oncogene 25:2685–2696 (2006). Meloni AM, Peier AM, Haddad FS, Powell IJ, Block AW, et al: A new approach in the diagnosis and follow-up of bladder cancer. Cancer Genet Cytogenet 71: 105–118 (1993). Mhawech-Fauceglia P, Cheney R, Schwaller J: Genetic alterations in urothelial bladder carcinoma. Cancer 106:1205–1216 (2006). Nowell PC, Hungerford DA: A minute chromosome in human chronic granulocytic leukemia. Science 132:1497 (1960). Packenham JP, Taylor JA, Anna CH, White CM, Devereux TR: Homozygous deletions but no sequence mutations in coding regions of p15 and p16 in human primary bladder tumors. Mol Carcinog 14:147–151 (1995). Parkin DM, Bray F, Ferlay J, Pisani P: Global cancer statistics, 2002. CA Cancer J Clin 55: 74–108 (2005). Placer J, Espinet B, Salido M, Sole F, Gelabert-Mas A: Correlation between histologic findings and cytogenetic abnormalities in bladder carcinoma: a FISH study. Urology 65: 913–918 (2005). Reuter VE: The pathology of bladder cancer. Urology 67 (Suppl 3A):11–18 (2006).
Richter J, Jiang F, Görög JP, Sartorius G, Egenter C, et al: Marked genetic differences between stage pTa and stage pT1 papillary bladder cancer detected by comparative genomic hybridization. Cancer Res 47: 2860–2864 (1997). Richter J, Beffa L, Wagner U, et al: Patterns of chromosomal imbalances in advanced urinary bladder cancer detected by comparative genomic hybridization. Am J Pathol 153: 1615– 1621 (1998). Richter J, Wagner U, Schraml P, Maurer R, Alund G, et al: Chromosomal imbalances are associated with a high risk of progression in early invasive (pT1) urinary bladder cancer. Cancer Res 59:5687–5691 (1999). Rowley J: A new consistent chromosomal abnormality in chronic myelogenous leukaemia identified by quinacrine fluorescence and Giemsa staining. Nature 243: 290–293 (1973). Saad A, Handbury D, McNicholas T, et al: A study comparing various noninvasive methods of detecting bladder cancer in urine. Br J Urol Int 89: 369–373 (2002). Sandberg AA: Chromosome changes in early bladder neoplasms. J Cell Biochem 161(Suppl):76– 79 (1992). Sandberg AA: Cytogenetics and molecular genetics of bladder cancer: A personal view. Am J Med Genet 115:173–182 (2002). Sidransky D, Frost P, Von Eschenbach A, Oyasu R, Preisinger AC, Vogelstein B: Clonal origin of bladder cancer. N Engl J Med 326: 737–740 (1992). Simon R, Bürger H, Brinkschmidt C, et al: Chromosomal aberrations associated with invasion in papillary superficial bladder cancer. J Pathol 185:345–351 (1998). Simon R, Bürger H, Semjonow A, Hertie L, Terpe HJ, Bocker W: Patterns of chromosomal imbalances in muscle invasive bladder cancer. Int J Oncol 17:1025–1029 (2000). Simoneau M, Aboulkassim TO, LaRue H, Fradet Y: Four tumor suppressor loci on chromosome 9q in bladder cancer: evidence for two novel candidate regions at 9q22.3 and 9q31. Oncogene 18: 157–163 (1999). Simoneau M, LaRue H, Aboulkassim TO, Meyer F, Moore L, Fradet Y: Chromosome 9 deletions and recurrence of superficial bladder cancer: Identification of four regions of prognostic interest. Oncogene 19: 6317–6323 (2000). Skacel M, Fahmy M, Brainard J, et al: Multitarget fluorescence in situ hybridization assay detects transitional cell carcinoma in the majority of patients with bladder cancer and atypical or negative urine cytology. J Urol 169: 2101–2105 (2003).
Sokolova I, Halling K, Jenkins R, et al: The development of a multitarget, multicolor fluorescence in situ hybridization assay for the detection of urothelial carcinoma in urine. J Mol Diagn 2: 116–123 (2000). Takahashi T, Habuchi T, Kakehi Y, Mtsumori K, Akao T, et al: Clonal and chronological genetic analysis of multifocal cancers of the bladder and upper urinary tract. Cancer Res 58: 5835– 5841 (1998). van Oers JMM, Adam C, Denzinger S, Stoehr R, Bertz S, et al: Chromosome 9 deletions are more frequent than FGFR3 mutations in flat urothelial hyperplasias of the bladder. Int J Cancer 119:1212–1215 (2006). Van Rhijn BWG, Lurkin I, Radvanyi F, Kirkels WJ, van der Kwast TH, Zwarthoff EC: The fibroblast growth factor receptor 3 (FGFR3) mutation is a strong indicator of superficial bladder cancer with low recurrence rate. Cancer Res 61: 1265–1268 (2001). Van Rhijn BWG, Montironi BW, Zwarthoff EC, Jobsis AC, van der Kwast TH: Frequent FGFR3 mutations in urothelial papilloma. J Pathol 198: 245–251 (2002). Van Rhijn BWG, Vis AN, Van der Kwast TH, Kirkels WJ, Radvanyi F, et al: Molecular grading of urothelial cell carcinoma with fibroblast growth factor receptor 3 and MIB-1 is superior to pathological grade for the prediction of clinical course. J Clin Oncol 21:1912–1921 (2003). van Tilborg AAG, de Vries A, de Bont M, Groenfeld LE, van der Kwast TH, Zwarthoff EC: Molecular evolution of multiple recurrent cancers of the bladder. Hum Mol Genet 9: 2973–2980 (2000). Veeramachaneni R, Nordberg M, Shi R, et al: Evaluation of fluorescence in situ hybridization as an ancillary tool to urine cytology in diagnosing urothelial carcinoma. Diagnostic Cytopathol 28:301–307 (2003). Walker L, Liston T, Loyd-Davies R: Does flexible cystoscopy miss more tumours than rod-lens examination? Br J Urol 72: 449–450 (1993). Williamson MP, Elder PA, Shaw ME, Devlin J, Knowles MA: p16 (CDKN2) is a major deletion target at 9p21 in bladder cancer. Hum Mol Genet 4:1569–1577 (1995). Yamamoto Y, Matsuyama H, Kawauchi S, Furuya T, Liu XP, et al: Biological characteristics in bladder cancer depend on the type of genetic instability. Clin Cancer Res 12: 2752–2758 (2006). Zhao J, Richter J, Wagner U, Roth B, Schraml P, et al: Chromosomal imbalances in noninvasive papillary bladder neoplasms (pTa). Cancer Res 59:4658–4661 (1999).
Cytogenet Genome Res 118:177–181 (2007)
181
show destructive myometrial invasion rather than the permeating invasion of low-grade ESS. Moreover, they demonstrate marked cellular pleomorphism and brisk mitotic activity. Endometrial sarcomas without recognizable evidence of a definite endometrial stromal phenotype – designated poorly differentiated endometrial sarcomas – are almost invariably high grade and termed poorly differentiated or undifferentiated uterine sarcomas. The use of a panel for the immunohistochemical staining to differentiate ESS from other tumors, particularly those with a ‘hemangiopericytomatous’ element has been advocated (Oliva et al., 2002; Bhargava et al., 2005). Cytogenetics of ESS
The cytogenetic findings in ESS reported to-date are primarily on low-grade tumors. The first cytogenetic findings in ESS were published almost 20 years ago (Dal Cin et al., 1988), followed by the report of a t(7;17)(p21;q12) in an ESS a few years later (Sreekantaiah et al., 1991). A recurrent t(7; 17)(p15–p21;q11) (Fig. 2), characteristic of ESS, or variants of this translocation, have been found in about half of the nearly 30 ESS with karyotypic findings (Table 1). A small number of tumors was associated with changes at 7p21]13, suggesting the presence of a cryptic t(7;17). Chromosomal breakpoints also tended to cluster at 6p23]p11 (5 cases), at 7q11]q22 (5 cases) and at 16q13]q22 (4 cases). Partial deletions of 6q were seen in 4 ESS. Derivative chromosomes, usually due to translocations, were seen in a significant number of cases. None of these changes appeared to be recurrent or specific. The occurrence of a t(7; 17)(p15;q21) as the sole cytogenetic anomaly in two ESS cases (Koontz et al., 2001) is indicative of the primary role played by this karyotypic change in the pathogenesis of ESS. Two other ESS with translocations as sole anomalies, i.e. t(X;17)(p21-p11;q23) (Amant et al., 2003) and t(10;17)(q22; p13) (Leunen et al., 2003) have also been reported. In addition to the balanced t(7;17), chromosome 7 is also frequently involved in rearrangements with other chromosomal partners in ESS. The short arm of chromosome 7, with breakpoints in 7p21⬃p15, was found rearranged without leading to the above-mentioned t(7;17) in five of the 25 ESS available for assessment. The 7p21⬃p15 segment recombined with 6p21 in one tumor reported (Laxman et al., 1993) and with 6q13 in another (Hrynchak et al., 1994). In one tumor (Iliszko et al., 1998), there were three different 7p rearrangements, described as der(2)t(2;7), del(7)(p15p13) and der(7), resulting from an unbalanced t(7;17). Another tumor with an add(7)(p21) (Gil-Benso et al., 1999) and a case in which both chromosomes 7 were involved in complex rearrangements at 7p15 have been described. The involvement of the same bands 7p21⬃p15 in six cases without a t(7; 17) (p21⬃p15;q12⬃q21), suggests that this region may be pathogenetically involved in ESS tumorigenesis in more than one way. Whether it is the JAZF1 locus that is involved in rearrangements with partners other than SUZ12 (see below) or with another gene in the same area, is presently unknown.
25 24 23
22
22 21
17q
21
p
15
12
14
14 13
13 12 11.2 11.1 11.1
13 12 11.2 11.1 11.1
p 11.2 11.1 11.1 11.2 12
11.2
11.2
21
q 21
q
22
22 23 24 25
13
17p
12 11.2 11.1 11.1 11.2 15
7p
21 21
7q
22
22
17 31
31
32 33 34 35
33 34 35
36
36
32
7
t(7;17)(p15;q11)
Fig. 2. Schematic presentation of the translocation (7;17)(p15;q11) seen in a significant number of ESS. This translocation leads to the genesis of an abnormal fusion gene JAZF1/SUZ12.
The short arm of chromosome 6, in particular 6p21, has been found rearranged in seven ESS in which the t(7;17) was not present. The translocation partners involved in the 6p21 rearrangements differed among the tumors, however. Chromosome 7 was involved in five rearrangements affecting bands 7q34 and 7q11 in two different abnormalities in one case (Micci et al., 2003); with 7p22, 7p21 and 7q21 in other tumors (Fresia et al., 1992; Laxman et al., 1993; Hrynchak et al., 1994) and with chromosome bands 3p13 and 3q29 in two tumors (Hrynchak et al., 1994; Füzesi et al., 1995). The 6p21 rearrangement was part of a der(15)t(6;15)(p21;p12) in one tumor and an add(6)(p21) in another case. The clearly nonrandom involvement of chromosomal band 6p21 in ESS suggests that one or more genes reside there that may be of importance in ESS tumorigenesis, but their identity is as yet unknown. From the cytogenetic findings described it would appear that ESS may be more heterogeneous pathogenetically than has been hitherto appreciated. So far, no phenotypic differences seem to be associated with the presence or absence of a t(7;17), a 6p rearrangement or other cytogenetic subgroups (Fig. 3), but one cannot rule out such correlations until they are specifically looked for in large tumor series. In a limited interphase FISH study with probes for chromosomes X, 11, 12 and 17 in eight low-grade and three high-grade ESS, it was shown that four of the former tumors showed gain of 1–3 chromosomes, whereas all three of the latter tumors showed polysomy of all four chromosomes tested (Cheung et al., 1996 a, b). The one CGH study performed on ESS showed gains of chromosomes 1 and 19, 6q, 9q, 16p, 20q and 22q in lowgrade ESS and gains of 2q, 4q, 6q, 7p, 9q and 20q in highgrade ESS. Losses were observed on chromosomes 2, 6, 7
Cytogenet Genome Res 118:182–189 (2007)
183
Table 1. Karyotypes in endometrial stromal sarcoma
Karyotype
Reference
46,XX,del(7)(q22q32),del(12)(q14q22)/45,XX,idem,–4/45,XX,idem,–16/44,XX,idem,–5,–11/44,XX,idem,–16, –19/43,XX,idem,–4,–18,–21/43,XX,idem,–13,–14,–17/48,XX,+7,+12
Havel et al., 1989
46,XX,t(7;13)(q11.1;p13),t(7;17)(p21;q12),del(11)(q13q21)
Sreekantaiah et al., 1991
46,XX,t(7;17)(p15⬃p21;q12⬃q21)/46,XX,idem,–7,+der(?)t(?;7)(?;q11)/45,XX,idem,–7,dic(15;22)(p11;p11), +der(?)t(?;7)(?;q11)/46,XX
Dal Cin et al., 1992
46,XX,del(5)(q31.1),der(7)t(6;7)(p21;p22)/46,XX
Fresia et al., 1992
46,XX,del(5)(q33),+der(7)t(6;7)(p21;p21)
Laxman et al., 1993
49,XX,+7,+8,+9,der(14)t(14;22)(p13;q12)/49,XX,+3,+7,+8,der(14)t(14;22)(p13;q12)
Laxman et al., 1993
80,XX,?i(1p),del(1)(p11),del(6)(q12),del(12)(p11),der(16)t(16;?)(q12;?),der(19)t(19;?)(q13;?),inc.
a
Laxman et al., 1993
46,XX,der(3)t(3;7)(p12;p12),der(6)t(3;6)(p13;p21)t(6;7)(q13;p21),der(7)t(6;7)(q13;p13)t(7;6)(q21;p21), inv(17)(p12q11)/46,XX,inv(17)(p12q11)
Hrynchak et al., 1994
47,XX,der(3)t(3;6)(q29;p21.1),der(6)t(3;6)(q21;q27),+19
Füzesi et al., 1995
46,XX,der(6)t(6;11)(p21;q11),del(6)(q15),add(7)(p21),t(7;17)(p15⬃p21;q12⬃q21),+9,-11
b
Pauwels et al., 1996
46,XX,t(7;17)(p21–p14;q11.2–q21),der(7)t(7;16)(p14⬃p15;q22)t(7;9)(q22;q22),der(9)t(7;9)(q22;q22), del(16)(q22)/47,idem,del(3)(p13p23),+mar
Hennig et al., 1997
47,XX,der(3)t(3;6)(q29;p21.1),der(6)t(3;6)(q21;q27),+i(19)(q10) c
Gunawan et al., 1998
42⬃44,X,–X,der(2)t(2;7)(p23;p15)t(2;15)(q35;q15),add(4)(p16),del(7)(p13p15),der(7)(7;17)(p14;q12), add(8)(q24),–10,del(11)(p11),t(11;13)(p15;q14);del(15)(q15),–16,del(17)(q12),der(18)t(16;18)(p11;p11), add(19)(p13),–20,add(21)(q22),–22,+2⬃3mar/78⬃83,idem×2 d
Iliszko et al., 1998
46,XX,del(6)(q21),del(12)(p13)
Iliszko et al., 1998
46,XX,+t(1;3)(p13;p25),i(8)(q10),dic(15;16)(p11;q13),–16
Iliszko et al., 1998
48,XX,+2,+7/50,XX,+der(1;7)(q10;q10),+2,+7,+10 b
Iliszko et al., 1998
46,XX,del(6)(q22),add(20)(q13)
Gil-Benso et al., 1999
53⬃55,X,–X,del(1)(p32)+del(1)(p21),+der(1)t(1;3)(p32;p21),del(3)(p21),+der(3)t(3;15)(q10;q10),–5,+6,+der(6), add(6)(p11)add(6)(q27),add(7)(p11),+add(7)(p21),+8,+8,–11,–13,der(13;21)(q10;q10),add(14)(p11), der(15)t(6;15)(p21;p12),der(17)t(3;17)(p21;p13)!2,+18,+18,add(19)(q13),–20,der(21;21)(q10;q10),+mar, dmin a
Gil-Benso et al., 1999
38,XX,–1,del(1)(q11),–2,add(2)(p13),–3,der(4;14)psu dic(4;14)(q35;q11.2)add(4)(p12),add(6)(p21.3),add(7)(q22), del(7)(p11.2p13),–8,–9,add(9)(q34),–10,add(10)(q24),–11,–11,ins(12;?)(q13;?),–14,–14,–15,ins(15;?)(q22;?), add(16)(q22),add(17)(q11.2),–18,der(18)(7;18)(q11.2;p11.2),–19,add(20)(p13),add(21)(p11.2),– 22,add(22)(p11.2),+6mar
Sonobe et al., 1999
45,XX,–7,t(7;17)(p15;q21)
Koontz et al., 2001
46,XX,t(7;13)(p15;p13),t(7;17)(p15;q21)
Koontz et al., 2001
46,XX,t(7;17)(p15;q21)
Koontz et al., 2001
46,XX,t(7;17)(p15;q21)
Koontz et al., 2001
46,X,t(X;17)(p21]p11;q23)
Amant et al., 2003
45,XX,add(3)(p11),der(9)t(7;9)(q11;p24),del(10)(q11q26),t(10;17)(q22;p13),der(11)t(3;11)(p22;q22),–13 b 46,XX,der(6)(6qter]6p21::7q34]7q11:),der(7)(:7p15]7p22::7p22]7qter),der(7)(:7p15]7q11::6p21]6pter) 46,XX,der(7)t(7;21)(p21⬃p11;q11⬃q21),t(7;17)(p15;q12),r(8)der(13)del(13)(?q12q14)del(13)(?q22) 46,XX,inv(2)(p21q37),der(6)t(6;7)(q21;p15)del(6)(p21),der(7)t(6;7)(?q12;p15) 46,XX,t(10;17)(q22;p13)
b
b
e
46,XX,t(7;17)(p15;q11),del(9)(q22),add(19)(q13)
Micci et al., 2003 b
Micci et al., 2003 Micci et al., 2003 Micci et al., 2003 Leunen et al., 2003 Satoh et al., 2003
46,XX,inv(2)(p21q37),der(6)del(6)(p21)t(6;7)(q21;p15)der(7)t(6;7)(p21;p15)del(6)(q21)
Micci et al., 2006
46XX,t(6;10;10)(p21;q22;p11)
Micci et al., 2006
47,X,der(X)t(X;16)(?;?),+2,add(2)(q11),ins(2;22)(q31;q11q13),der(3)ins(3;13)(p24:q?22q32)t(3;6)(q28;q22), der(6)t(3;6)(q28;q22),del(6)(p11),der(7)(7qter]7p15::6?::15?::3?::13?::15?::6?::X?::6?::3?)der(14)t(1;14)(q25;q32)
Micci et al., 2006
a
184
High grade. b Metastatic. c Cell-line. d Recurrence. e Metastatic and aggressive, high mitotic index.
Cytogenet Genome Res 118:182–189 (2007)
1
2
3
4
A
X
X
6
7
5 B
8
9
10
11
12
17
18
C
14
13
Fig. 3. Karyotype of an ESS with a t(X;17)(p11;q23) as the only change (Amant et al., 2003).
Y
19
and X and 4q, 11q, 13q, 15q, 16q and 20q in the former tumors, and 3q, 10p and 14q in the latter tumors (Halbwedl et al., 2005). The chromosomal aberrations in ESS are heterogeneous and do not clearly correlate with the histologic grade. However, the common deletion of 7p (55.6% of the cases) in low-grade ESS may play a role in tumor development and progression (Halbwedl et al., 2005). Molecular studies in ESS
The demonstration (Koontz et al., 2001) of a fusion gene, JAZF1/SUZ12, resulting from the t(7; 17) in five low-grade ESS involving the genes JAZF1 (at 7p15) and SUZ12 (at 17q21), was complicated by the presence of this fusion gene in all three endometrial stromal nodules examined, but only in three of seven high-grade ESS. Normal endometrium was negative for the fusion gene. A subsequent report (Micci et al., 2003) described the fusion gene in a rather unusual case with ESS. The patient was a 78-year-old woman with a history of intestinal endometriosis who underwent surgery for multiple intra-abdominal metastases from a formerly removed ESS of the colon and thought to arise from an endometriotic lesion. The tumor was shown to have a t(7;17) (p15;q12) and a JAZF1/SUZ12 fusion gene (Fig. 4). Two other low-grade ESS studied (Micci et al., 2003) did not contain the fusion gene or the translocation, but did have involvement of 7p. In another study (Huang et al., 2004) JAZF1/SUZ12 fusion transcripts were found in three of four low-grade ESS, in one endometrial stromal nodule, as well as in a nodule labeled a mixed smooth muscle variant, but in none of the nine low-grade ESS labeled ‘variant’ by the authors. The lowgrade ESS and an ESS cell line negative for the fusion transcripts apparently did not show a t(7;17), which may account
15
16
D
E
20
21
F
22 G
M
1
2
B
Fig. 4. Chimeric JAZF1/SUZ12 fusion transcript in an ESS with a t(7; 17) shown in lane 2 (Micci et al., 2003). Lane 1 contained the cDNA from a gastric cancer and lane B a control with no RNA in the cDNA synthesis. Lane M is a 100-bp DNA ladder.
for the failure to see the fusion transcripts. An advanced stage ESS and its metastasis with a t(X;17)(p11;q23) as the sole chromosomal change was not examined for the fusion transcript (Amant et al., 2003). Thus, the JAZF1/SUZ12 fusion gene is frequently but not consistently present in typical ESS and less often in histologically variant cases, such as the mixed smooth muscle variants (Huang et al., 2004). The fusion gene data suggest that ESS are genetically heterogeneous and that the fusion gene may be confined to the tumors with ‘classic’ histology. The exact mechanisms by which the JAZF1/SUZ12 fusion products are involved in the pathogenetic processes in the tumors in which these fusions occur has not been readily explained (Koontz et al., 2001). The specific functions of JAZF1 and SUZ12 and the reason for their involvement in a gene fusion associated with neoplasia are not directly apparent from the sequences of the cDNAs (Koontz et al., 2001). The only recognizable regions within the two cDNAs resembling sequences with-
Cytogenet Genome Res 118:182–189 (2007)
185
Table 2. Molecular and other aspects of ESS
References DNA ploidy and prognosis in ESS
Hitchcock and Norris, 1992; Nordal et al., 1996
Although p53 mutations or overexpression are common in uterine sarcomas, these have not been found in the few ESS examined.
Liu et al., 1994
FISH studies of an ESS with mixed histology showed diploidy in the low-grade portion of the tumor, whereas the high-grade part had polysomy for chromosomes tested (X, 11, 12 and 17).
Cheung et al., 1996a, b
No chromosomal changes were seen in an ESS cell line whose cells closely resembled proliferative endometrial stromal cells not only morphologically but also functionally.
Nasu et al., 1998
Expression of metalloproteinases in ESS
Liokumovich et al., 1999
The detection of CD10 in ESS can be a useful marker for differentiating it from other tumors, such as uterine sarcomas. However, the diversity of CD10 expression should be taken into consideration when interpreting results based on CD10 expression.
Chu and Arber, 2000; Chu et al., 2001; McCluggage et al., 2001; Toki et al., 2001
Microsatellite instability was not present in ESS, nor was there loss of alleles.
Amant et al., 2001
h-Caldesmon, a smooth muscle-specific antibody, can be used to distinguish ESS from other tumors.
Nucci et al., 2001; Rush et al., 2001
Studies to define the breakpoint at 6p12.3 were performed using a BAC clone.
Gunawan et al., 2003
Oxytocin receptor expression as a means of distinguishing ESS from uterine sarcoma.
Loddenkemper et al., 2003
Utility of MIB-1 mitotic index in ESS as a possible predictor of recurrence and for the differential diagnosis has been reported.
Popiolek et al., 2003; Kir et al., 2005
Changes of the ERBB-2 gene are not found in ESS, in contrast to uterine adenocarcinoma.
Amant et al., 2004b
Activation of Wnt-signaling pathway occurs in ESS, supported by the translocation of beta-catenin to the nucleus. An inverse relation was found with the expression of frizzled-related protein 4.
Hrzenjak et al., 2004
Immunohistochemical features of ESS may be useful in the differential diagnosis from other tumors.
Zhu et al., 2004
The useful information supplied by RT-PCR in establishing the presence of the JAZF1/SUZ12 fusion gene has been stressed and optimized for paraffin-embedded specimens.
Hrzenjak et al., 2005
ESS frequently express epidermal growth factor receptor (EGFR, HER-1) which may serve as a possible therapeutic target.
Moinfar et al., 2005
Possible genetic background for ESS
Reich and Regauer, 2005
A case of ESS positive for KIT by staining for CD117 responded to imatinib and surgical therapy. Since ESS have been reported to express KIT protein but lack mutations at exons 11 and 17, the above mentioned case may have benefited more from the surgery than the imatinib therapy.
Salvatierra et al., 2006; Rushing et al., 2003
in known genes in humans and other species encode zinc finger motifs, as often found in DNA binding proteins. Another finding possibly relevant to the role of JAZF1 in oncogenesis was the lack of JAZF1 mRNA in some ESS (Koontz et al., 2001), one of which was monosomic for chromosome 7 and another which contained unrearranged DNA for JAZF1 on the allele not affected by the t(7;17), as demonstrated by Southern blot analysis (Koontz et al., 2001). Failure to detect mRNA in the latter case may therefore be due to transcriptional silencing of the normal copy of JAZF1. Loss of expression for normal versions of JAZF1 in multiple tumors suggests a possible role of this gene as a tumor suppressor. The distribution of the JAZF1/SUZ12 fusion gene among stromal tumors carries a number of implications for the biology of these neoplasms (Koontz et al., 2001). For example, the finding of the fusion in stromal nodules rais-
186
Cytogenet Genome Res 118:182–189 (2007)
es the possibility that ESS can develop from stromal proliferations that are initially benign. Additionally, of the seven high-grade ESS studied, only three showed evidence of the fusion, suggesting that some high-grade ESS may arise by a different pathogenetic mechanism than that of low-grade ESS. It will be important to correlate possible differences in the clinical aspects and biologic behavior of ESS with the presence or absence of this genetic lesion. Making the results difficult to interpret and correlate is the absence of accompanying cytogenetic data for some of the tumors, particularly stromal nodules. Publications in the past have indicated the involvement of 6p in these lesions, including reports of a t(6; 20)(p21;q13) and a t(1; 6;4) (q21;p21;q13), though other changes may be seen (Walter et al., 1989; Dal Cin et al., 1991; Speleman et al., 1991; Fletcher et al., 1992). Future combined cytogenetic and molecular
studies on the same tumors should resolve some of the enigma. The rearrangement at 6p21 in ESS was further investigated in three tumors, two with der(6) due to unbalanced t(6p;7p) and one due to t(6;10;10)(p21;q22;p11) (Micci et al., 2006). All three tumors showed a specific rearrangement of the PHD finger protein 1 (PHF1) gene located at 6p21. In the two tumors with t(6;7), PHF1 was combined with the JAZF1 gene located at 7p15, resulting in the formation of a JAZF1/ PHF1 fusion gene. In the third tumor with t(6p;10q;10p) as the sole karyotypic change, fusion of PHF1 with the enhancer of polycomb (EPC1) gene at 10p11 occurred. The PHF1 gene encodes a protein with two zinc finger motifs whose involvement in ESS appears to be unique to-date (Micci et al., 2006). Though LOH of tumor suppressor genes in ESS, and not microsatellite instability, may play a role in tumor development, the presence of similar findings in the microenvironment, including myometrium, is difficult to interpret (Moinfar et al., 2004a). The expression of KIT in some ESS, both low- and highgrade, may represent a therapeutic target (Geller et al., 2004). Low-grade ESS may resemble endometriosis (Brunisholz et al., 2004). Aromatase inhibitor was quite successful in a low-grade ESS (Leunen et al., 2004). This response may be related to the high expression levels of this enzyme found in low-grade ESS (Reich and Regauer, 2004). -Catenin is a crucial part of the Wnt and E-cadherin signaling pathways, which are involved in tumorigenesis. Dysregulation of these pathways allows -catenin to accumulate and translocate to the nucleus, where it may activate oncogenes. High-level nuclear staining for -catenin was seen in 40% of ESS and may be used as a diagnostic tool (Ng et al., 2005).
Steroid receptors
A study on the presence of estrogen receptors (ER) and progesterone receptors (PR) in ESS revealed the former to be present in 7/9 and the latter in 2/9 tumors (Wade et al., 1990). The amounts of these receptors were shown to be much higher than in normal endometrium during the proliferative phase (Navarro et al., 1992). Antiestrogenic therapy was suggested as possibly effective in these cases (Navarro et al., 1992). The lack of the beta ER expression in ESS may be a marker for malignancy (Chu et al., 2003). The expression of PR isoforms in ESS closely resembles the pattern of expression of these proteins in normal endometrial stroma. However, in cases with additional or aberrant features of differentiation the usual ratio of PRA or PRB is disrupted. Furthermore, recurrent ESS can show a change in the ratio of PRA to PRB compared with that of the primary tumors. This finding may reflect an effect of disease progression or intervening treatment on tumor differentiation (Balleine et al., 2004). Because it is likely that the relative expression of PR isoforms influences progesterone signaling, alterations in this feature may influence the hormonal responsiveness of ESS. Androgen receptors (AR) are expressed in nearly 50% of ESS (Moinfar et al., 2004b). The presence of aromatase in low-grade ESS has been reported (Reich and Regauer, 2004) and inhibition of this enzyme as an approach to therapy has been described (Leunen et al., 2004). A listing of cytogenetic, molecular and other aspects of ESS, including differential diagnosis, is presented in Table 1 and 2.
References Amant F, Dorfling CM, Dreyer L, Vergote I, Lindeque BG, Van Rensburg EJ: Microsatellite instability in uterine sarcomas. Int J Gynecol Cancer 11: 218–223 (2001). Amant F, Moerman P, Cadron I, Hagemeijer A, Vergote I, Debiec-Rychter M: Endometrial stromal sarcoma with a sole t(X;17) chromosome change: report of a case and review of the literature. Gynecol Oncol 88: 459–462 (2003). Amant F, Vergote I, Moerman P: The classification of a uterine sarcoma as ‘high-grade endometrial stromal sarcoma’ should be abandoned. Gynecol Oncol 95:412–413 (2004a). Amant F, Vloeberghs V, Woestenborghs H, DebiecRychter M, Verbist L, et al: ERBB-2 gene overexpression and amplification in uterine sarcomas. Gynecol Oncol 95: 583–587 (2004b). Balleine RL, Earls PJ, Webster LR, Mote PA, deFazio A, et al: Expression of progesterone receptor A and B isoforms in low-grade endometrial stromal sarcoma. Int J Gynecol Pathol 23: 138–144 (2004). Bhargava R, Shia J, Hummer AJ, Thaler HT, Tornos C, Soslow RA: Distinction of endometrial stromal sarcoma from ‘hemangiopericytomatous’ tumors using a panel of immunohistochemical stains. Mod Pathol 18: 40–47 (2005).
Brunisholz Y, Miller J, Scurry J, Proietto A: Endometrial stromal sarcoma resembling adenomyosis and menstrual-phase endometrium. Gynecol Oncol 95: 256–259 (2004). Cheung ANY, Ng WF, Chung LP, Khoo, US: Mixed low grade and high grade endometrial stromal sarcoma of uterus: differences on immunohistochemistry and chromosome in situ hybridisation. J Clin Pathol 49: 604–607 (1996a). Cheung ANY, Tin VPC, Ngan HYS, Chung LP, Khoo US: Interphase cytogenetic study of endometrial stromal sarcoma by chromosome in situ hybridization. Mod Pathol 9: 910–918 (1996b). Chu MC, Mor G, Lim C, Zheng W, Parkash V, Schwartz PE: Low-grade endometrial stromal sarcoma: hormonal aspects. Gynecol Oncol 90: 170–176 (2003). Chu P, Arber DA: Paraffin-section detection of CD10 in 505 nonhematopoietic neoplasms. Frequent expression in renal cell carcinoma and endometrial stromal sarcoma. Am J Clin Pathol 113:374–382 (2000). Chu PG, Arber DA, Weiss LM, Chang KL: Utility of CD10 in distinguishing between endometrial stromal sarcoma and uterine smooth muscle tumors: An immunohistochemical comparison of 34 cases. Mod Pathol 14: 465–471 (2001).
Dal Cin P, Talcott J, Abrams J, Li FP, Sandberg AA: Ins(10; 19) in an endometrial stromal sarcoma. Cancer Genet Cytogenet 36: 1–5 (1988). Dal Cin P, Van den Berghe H, Brosens I: Involvement of 6p in an endometrial polyp. Cancer Genet Cytogenet 51: 279–280 (1991). Dal Cin P, Sayed Aly M, De Wever I, Moerman P, Van den Berghe H: Endometrial stromal sarcoma t(7; 17)(p15–21;q12–21) is a nonrandom chromosome change. Cancer Genet Cytogenet 63:43–46 (1992). Evans H: Endometrial stromal sarcoma and poorly differentiated endometrial sarcoma. Cancer 50: 2170–2182 (1982). Fletcher JA, Pinkus JL, Lage JM, Morton CC, Pinkus GS: Clonal 6p21 rearrangement is restricted to the mesenchymal component of an endometrial polyp. Genes Chromosomes Cancer 5: 260– 263 (1992). Fresia AE, Currie JL, Farrington JE, Laxman R, Griffin CA: Uterine stromal sarcoma cell line. A cytogenetic and electron microscopic study. Cancer Genet Cytogenet 60: 60–66 (1992). Füzesi L, Gunawan B, Braun S, Karl MC: Endometrial stromal sarcoma with clonal chromosomal aberrations and mixed phenotype. Cancer Genet Cytogenet 84: 85–88 (1995).
Cytogenet Genome Res 118:182–189 (2007)
187
Geller MA, Argenta P, Bradley W, Dusenbery KE, Brooker D, et al: Treatment and recurrence patterns in endometrial stromal sarcomas and the relation to c-kit expression. Gynecol Oncol 95: 632–636 (2004). Gil-Benso R, López-Ginés C, Navarro S, Carda C, Llombart-Bosch A: Endometrial stromal sarcomas: immunohistochemical, electron microscopical and cytogenetic findings in two cases. Virchows Arch 434: 307–314 (1999). Gunawan B, Braun S, Cortés MJ, Bergmann F, Kari C, Füzesi L: Characterization of a newly established endometrial stromal sarcoma cell line. Int J Cancer 77: 424–428 (1998). Gunawan B, Schulten HJ, Füzesi L: Identification of a BAC clone overlappig the t(6p12.3) breakpoint in the cell line ESS-1 derived from an endometrial stromal sarcoma. Cancer Genet Cytogenet 147:84–86 (2003). Halbwedl I, Ullmann R, Kremser ML, Man YG, Isadi-Moud N, et al: Chromosomal alterations in low-grade endometrial stromal sarcoma and undifferentiated endometrial sarcoma as detected by comparative genomic hybridization. Gynecol Oncol 97: 582–587 (2005). Havel G, Dahlenfors F, Wedell B, Mark J: Similar chromosomal evolution in a uterine stromomyosarcoma and in one of two leiomyomas from the same patient. APMIS 97: 143–146 (1989). Hennig Y, Caselitz J, Bartnitzke S, Bullerdiek J: A third case of a low-grade endometrial stromal sarcoma with a t(7;17)(p14⬃21;q11.2⬃21). Cancer Genet Cytogenet 98: 84–86 (1997). Hitchcock CL, Norris HJ: Flow cytometric analysis of endometrial stromal sarcoma. Am J Clin Pathol 97: 267–271 (1992). Hrynchak M, Horsman D, Salski C, Berean K, Benedet JL: Complex karyotypic alterations in an endometrial stromal sarcoma. Cancer Genet Cytogenet 77: 45–49 (1994). Hrzenjak A, Tippl M, Kremser ML, Strohmeier B, Guelly C, et al: Inverse correlation of secreted frizzled-related protein 4 and beta-catenin expression in endometrial stromal sarcoma. J Pathol 204: 19–27 (2004). Hrzenjak A, Moinfar F, Tavassoli FA, Strohmeier B, Kremser ML, et al: JAZF1/JJAZ1 gene fusion in endometrial stromal sarcomas: molecular analysis by reverse transcriptase-polymerase chain reaction optimized for paraffin-embedded tissue. J Mol Diagn 7: 388–395 (2005). Huang HY, Ladanyi M, Soslow RA: Molecular detection of JAZF1-JJAZ1 gene fusion in endometrial stromal neoplasms with classic and variant histology. Evidence for genetic heterogeneity. Am J Surg Pathol 28: 224–232 (2004). Iliszko M, Mandahl N, Mrózek K, Denis A, Pandis N, et al: Cytogenetics of uterine sarcomas: Presentation of eight new cases and review of the literature. Gynecol Oncol 71: 172–176 (1998). Khalifa MA, Hansen CH, Moore JL Jr, Rusnock EJ, Lage JM: Endometrial stromal sarcoma with focal smooth muscle differentiation: recurrence after 17 years: a follow-up report with discussion of the nomenclature. Int J Gynecol Pathol 15:171–176 (1996). Kir G, Cetiner H, Karateke A, Gurbuz A, Bulbul D: Utility of MIB-1 and estrogen and progesterone receptor in distinguishing between endometrial stromal sarcomas and endometrial stromal nodules, highly cellular leiomyomas. Int J Gynecol Cancer 15: 337–342 (2005). Koontz JI, Soreng AL, Nucci M, Kuo FC, Pauwels P, et al: Frequent fusion of the JAZF1 and JJAZ1 genes in endometrial stromal tumors. Proc Natl Acad Sci USA 98:6348–6353 (2001).
188
Laxman R, Currie JL, Kurman RJ, Dudzinski M, Griffin CA: Cytogenetic profile of uterine sarcomas. Cancer 71: 1283–1288 (1993). Leunen K, Amant F, Debiec-Rychter M, Croes R, Hagemeijer A, et al: Endometrial stromal sarcoma presenting as postpartum haemorrhage: report of a case with a sole t(10; 17)(q22;p13) translocation. Gynecol Oncol 91: 265–271 (2003). Leunen M, Breugelmans M, De Sutter P, Bourgain C, Amy JJ: Low-grade endometrial stromal sarcoma treated with the aromatase inhibitor letrozole. Gynecol Oncol 95: 769–771 (2004). Liokumovich P, Goldberg I, Davidson B, Gotlieb WH, Zahavi T, et al: Expression of metalloproteinases endometrial stromal sarcoma: immunohistochemical study using image analysis. J Clin Pathol 52: 198–202 (1999). Liu FS, Kohler MF, Marks JR, Bast RC Jr, et al: Mutation and overexpression of the p53 tumor suppressor gene frequently occurs in uterine and ovarian sarcomas. Obstet Gynecol 83: 118–124 (1994). Loddenkemper C, Mechsner S, Foss HD, Dallenbach FE, Anagnostopoulos I, et al: Use of oxytocin receptor expression in distinguishing between uterine smooth muscle tumors and endometrial stromal sarcoma. Am J Surg Pathol 27:1458–1462 (2003). McCluggage WG, Sumathi VP, Maxwell P: CD10 is a sensitive and diagnostically useful immunohistochemical marker of normal endometrial stroma and of endometrial stromal neoplasms. Histopathology 39: 273–278 (2001). Micci F, Walter CU, Teixeira MR, Panagopoulos I, Bjerkehagen B, et al: Cytogenetic and molecular genetic analyses of endometrial stromal sarcoma: nonrandom involvement of chromosome arms 6p and 7p and confirmation of JAZF1JJAZ1 gene fusion in t(7; 17). Cancer Genet Cytogenet 144:119–124 (2003). Micci F, Panagopoulos I, Bjerkehagen B, Heim S: Consistent rearrangement of chromosomal band 6p21 with generation of fusion genes JAZF1/PHF1 and EPC1/PHF1 in endometrial stromal sarcoma. Cancer Res 66: 107–112 (2006). Miettinen M: Diagnostic Soft Tissue Pathology. (Churchill Livingstone, New York 2003). Moinfar F, Kremser ML, Man YG, Zatloukal K, Tavassoli FA, Denk H: Allelic imbalances in endometrial stromal neoplasms: frequent genetic alterations in the nontumorous normal-appearing endometrial and myometrial tissues. Gynecol Oncol 95: 662–671 (2004a). Moinfar F, Regitnig P, Dastranj Tabrizi A, Denk H, Tavassoli FA: Expression of androgen receptors in benign and malignant endometrial stromal neoplasms. Virchows Arch 444: 410–414 (2004b). Moinfar F, Gogg-Kamerer M, Sommersacher A, Regitnig P, Man YG, et al: Endometrial stromal sarcomas frequently express epidermal growth factor receptor (EGFR, HER-1). Potential basis for a new therapeutic approach. Am J Surg Pathol 29: 485–489 (2005). Nasu K, Matsui N, Narahara H, Tanaka Y, Takai N, et al: MaMi, a human endometrial stromal sarcoma cell line that constitutively produces interleukin-6, interleukin-8, and monocyte chemoattractant protein 1. Arch Pathol Lab Med 122:836–841 (1998). Navarro D, Cabrera JJ, León L, Chirino R, Fernández L: Endometrial stromal sarcoma expression of estrogen receptors, progesterone receptors and estrogen-induced srp27 (24K) suggests hormone responsiveness. J Steroid Biochem Mol Biol 41:589–596 (1992).
Cytogenet Genome Res 118:182–189 (2007)
Ng TL, Gown AM, Barry TS, Cheang MCU, Chan AKW, et al: Nuclear beta-catenin in mesenchymal tumors. Mod Pathol 18: 68–74 (2005). Nordal RR, Kristensen GB, Kaern J, Stenwig AE, Pettersen EO, Trope CG: The prognostic significance of surgery, tumor size, malignancy grade, menopausal status, and DNA ploidy in endometrial stromal sarcoma. Gynecol Oncol 62:254–259 (1996). Norris HJ, Taylor HB: Mesenchymal tumors of the uterus. I. A clinical and pathological study of 53 endometrial stromal tumors. Cancer 19: 755– 766 (1966). Norris HJ, Zaloudek CJ: Mesenchymal tumors of the uterus, in Blauestein A (ed): Pathology of the female genital tract, 2nd ed., pp 352–392 (Springer-Verlag, New York 1982). Nucci MR, O’Connell JT, Huettner PC, Cviko A, Sun D, Quade BJ: h-Caldesmon expression effectively distinguishes endometrial stromal tumors from uterine smooth muscle tumors. Am J Surg Pathol 25: 455–463 (2001). Oliva E, Clement P, Young R: Endometrial stromal tumors: an update on a group of tumors with a protean phenotype. Adv Anat Pathol 7:257–281 (2000). Oliva E, Young RH, Amin MB, Clement PB: An immunohistochemical analysis of endometrial stromal and smooth muscle tumors of the uterus. A study of 54 cases emphasizing the importance of using a panel because of overlap in immunoreactivity for individual antibodies. Am J Surg Pathol 26:403–412 (2002). Pauwels P, Dal Cin P, Van de Moosdijk CNF, Vrints L, Sciot R, et al: Cytogenetics revealing the diagnosis in a metastatic endometrial stromal sarcoma. Histopathology 29: 84–87 (1996). Popiolek D, Yee H, Levine P, Vamvakas E, Demopoulos RI: MIB1 as a possible predictor of recurrence in low-grade endometrial stromal sarcoma of the uterus. Gynecol Oncol 90: 353–357 (2003). Reich O, Regauer S: Aromatase expression in lowgrade endometrial stromal sarcomas: an immunohistochemical study. Mod Pathol 17:104– 108 (2004). Reich O, Regauer S: Endometrial stromal sarcoma – observational evidence of a genetic background? Eur J Gynaecol Oncol Pathol 26: 288– 290 (2005). Rush DS, Tan JY, Baergen RN, Soslow RA: h-Caldesmon, a novel smooth muscle-specific antibody, distinguishes between cellular leiomyoma and endometrial stromal sarcoma. Am J Surg Pathol 25:253–258 (2001). Rushing RS, Shajahan S, Chendil D, Wilder JL, Pulliam J, et al: Uterine sarcomas express KIT protein but lack mutation(s) in exon 11 or 17 of cKIT. Gynecol Oncol 91: 9–14 (2003). Salvatierra A, Tarrats A, Gomez C, Sastre JM, Balaña C: A case of c-kit positive high-grade stromal endometrial sarcoma responding to Imatinib Mesylate. Gynecol Oncol 101:545–547 (2006). Satoh Y, Ishikawa Y, Miyoshi T, Mukai H, Okumura S, Nakagawa K: Pulmonary metastases from a low-grade endometrial stromal sarcoma confirmed by chromosome aberration and fluorescence in situ hybridization approaches: a case of recurrence 15 years after hysterectomy. Virchows Arch 422:173–178 (2003). Sonobe H, Iwata J, Furihata M, Ohtsuki Y, Taguchi T, Shimizu K: Endometrial stromal sarcoma with clonal complex chromosome abnormalities. Report of a case and review of the literature. Cancer Genet Cytogenet 112: 34–37 (1999).
Speleman F, Dal Cin P, Van Roy N, Van Marck E, Buytaert P, et al: Is t(6;20)(p21;q13) a characteristic chromosome change in endometrial polyps? Genes Chromosomes Cancer 3: 318–319 (1991). Sreekantaiah C, Li FP, Weidner N, Sandberg AA: An endometrial stromal sarcoma with clonal cytogenetic abnormalities. Cancer Genet Cytogenet 55: 163–166 (1991). Toki T, Shimizu M, Takagi Y, Ashida T, Konishi I: CD10 is a marker for normal and neoplastic endometrial stromal cells. Int J Gynecol Pathol 21: 41–47 (2001).
Wade K, Quinn MA, Hammond I, Williams K, Cauchi M: Uterine sarcoma: Steroid receptors and response of hormonal therapy. Gynecol Oncol 39:364–367 (1990). Walter TA, Fan SX, Medchill MT, Berger CS, Decker HJH, Sandberg AA: Inv(12)(p11.2q13) in an endometrial polyp. Cancer Genet Cytogenet 41: 99–103 (1989). Yilmaz A, Rush DS, Soslow RA: Endometrial stromal sarcomas with unusual histologic features. A report of 24 primary and metastatic tumors emphasizing fibroblastic and smooth muscle differentiation. Am J Surg Pathol 26:1142–1150 (2002).
Zaloudek C, Hendrickson M: Mesenchymal tumors of the uterus, in Kurman RJ (ed): Blaustein’s Pathology of the Female Genital Tract, pp 561– 615 (Springer-Verlag, New York 2002). Zhu XQ, Shi YF, Cheng XD, Zhao CL, Wu YZ: Immunohistochemical markers in differential diagnosis of endometrial stromal sarcoma and cellular leiomyoma. Gynecol Oncol 92: 71–79 (2004).
Cytogenet Genome Res 118:182–189 (2007)
189
serve as useful diagnostic markers and may eventually even contribute to the development of rational forms of cancer therapy (Druker et al., 2001). With this consideration in mind, we have reviewed our own research as well as the available literature on possible pathogenetic mechanisms involved in the genesis of ESS. As the development of a malignant tumor may occur via different and at least partially complementary genetic pathways, i.e., rearrangement of oncogenes located at the breakpoints of chromosomal aberrations, loss of tumor suppressor genes or activation (gain) of oncogenes, and mutations of microsatellite instability genes, we shall overview the topic along the same lines. Rearrangements of genes due to chromosomal aberrations
Cytogenetic studies of ESS are few with only 34 tumors having been karyotyped and scientifically reported (Dal Cin et al., 1988, 1992; Havel et al., 1989; Fletcher et al., 1991; Sreekantaiah et al., 1991; Fresia et al., 1992; Laxman et al., 1993; Hrynchak et al., 1994; Fuzesi et al., 1995; Pauwels et al., 1996; Hennig et al., 1997; Iliszko et al., 1998; Gil-Benso et al., 1999; Sonobe et al., 1999; Koontz et al., 2001; Amant et al., 2003a; Leunen et al., 2003; Micci et al., 2003, 2006; Satoh et al., 2003). Although a variety of different chromosomal aberrations have been described, the pattern of their occurrence is nevertheless clearly nonrandom with particularly frequent involvement of chromosome arms 6p, 7p, and 17q (Micci et al., 2003). Chromosomes 7 and 17 are recombined in the first genetic hallmark to be discovered in ESS, namely the translocation t(7;17)(p15;q21) (Fig. 1) which has been described in altogether 12 tumors; in 11 of them as a balanced rearrangement (Fletcher et al., 1991; Sreekantaiah et al., 1991; Dal Cin et al., 1992; Pauwels et al., 1996; Hennig et al., 1997; Koontz et al., 2001; Micci et al., 2003; Satoh et al., 2003) and in one as only a der(7)t(7;17) (Iliszko et al., 1998). Koontz et al. (2001) demonstrated that two zinc finger genes were recombined by this translocation, that juxtaposed with another zinc f inger (JAZF1) gene from chromosomal band 7p15 and the joined to JAZF1 (JJAZ1, current gene symbol SUZ12) gene from 17q21. The chimeric transcript contains 5 sequences from the JAZF1 and 3 sequences from JJAZ1 but retains the zinc finger motifs from both genes. Because wild-type JAZF1 is expressed in normal endometrium, it has been suggested that the JAZF1/JJAZ1 fusion gene creates a chimeric protein that disrupts transcription in a lineage-specific manner (Koontz et al., 2001). The presence of the JAZF1/JJAZ1 chimeric transcript has been tested for in several different tumor entities and tissue types. Fusion of these genes appears to be frequent, although certainly not ubiquitous, in ESS of classic histology, but has also been found in other types of EST. Of a total of 58 EST tested for JAZF1/JJAZ1 gene fusion by polymerase chain reaction (PCR) based techniques, 37 were shown to have it (Koontz et al., 2001; Micci et al., 2003; Huang et al., 2004; Hrzenjak et al., 2005). Among the tumors that were found positive for this rearrangement were 29 ESS of classic histology (Koontz
Fig. 1. (a) Ideograms and G-banded images of the most commonly rearranged chromosomes in EST and the genes known to be involved, i.e., PHF1 on 6p21, JAZF1 on 7p15, and JJAZ1 on 17q21. (b) Ideograms and G-banded images of the most common chromosomal aberrations in EST and the genes involved. Arrows indicate breakpoint positions. The complete karyotypic descriptions of the aberrations are as follows: der(7)t(6; 7)(p21;p15)del(6)(q21) (left), der(7)t(7; 17)(p15;q21) (center), and der(17)t(7;17)(p15;q21) (right).
et al., 2001; Micci et al., 2003; Huang et al., 2004; Hrzenjak et al., 2005), one fibromyxoid ESS, one ESS with sex-cord like histology (Hrzenjak et al., 2005), four ESN of classic histology (Koontz et al., 2001; Huang et al., 2004), one ESN of the mixed smooth muscle variant, and the last tumor was a UES (Koontz et al., 2001). Positive cases were not only found among primary uterine tumors but also in metastatic lesions and in a primary extrauterine ESS (Huang et al., 2004). The same intragenic breakpoints have so far been seen in all JAZF1/JJAZ1-positive tumors, with G-435 from the JAZF1 sequence being followed by A-468 from the JJAZ1 sequence, suggesting a highly consistent breakage mechanism. By way of comparison, Hrzenjak et al. (2005) tested also normal endometrium (ten cases), leiomyomas (five cases), leiomyosarcomas (five cases), lung carcinomas (three cases), gastric carcinomas (three cases), and hepatic carcinomas (three cases) for occurrence of the JAZF1/JJAZ1 fusion gene but found no chimeric transcript. It should be noted that cytogenetic data were not available for any of these specimens. Hence, we do not know whether a 7;17-translocation was present whenever a JAZF1/JJAZ1 fusion was seen by PCR, although we assume this to be the case. Nor do we know with certainty that the JAZF1/JJAZ1 fusion is present
Cytogenet Genome Res 118:190–195 (2007)
191
only in EST; a larger series of tumors with cytogenetic rearrangements of 7p15 and/or 17q21 has to be investigated to be absolutely certain that the JAZF1/JJAZ1 is indeed tumor specific. In addition to the 7;17-translocation, chromosomal band 7p21]p15 was found rearranged in seven ESS with other partners than chromosome 17 (Laxman et al., 1993; Hrynchak et al., 1994; Iliszko et al., 1998; Gil-Benso et al., 1999; Micci et al., 2003, 2006) suggesting that alternative, pathogenetically equivalent variant translocations exist in this tumor type. Indeed, the first such variant, a t(6; 7), was recently described in two ESS in which a der(7)t(6; 7)(p21;p15)del(6)(q21) and a der(7)(7qter]7p15::6?::15?::3?:: 13?::15?::6?::X?::6?::3?:) were seen (Micci et al., 2006). The JAZF1 gene was found rearranged with the PHD f inger protein 1 (PHF1) gene from chromosomal band 6p21 (Fig. 1) in these tumors (Micci et al., 2006). The putative proteins in both cases retained one zinc finger domain from the JAZF1 gene and both zinc finger domains from PHF1. The specific functions of the JAZF1 and PHF1 and why they are involved in a neoplasia-specific gene fusion are not directly apparent, but the fact that regions in their sequences encode zinc finger motifs suggests that deregulation of normal transcription processes may be the crucial event. JAZF1 has been shown to interact physically and specifically with TAK1, which is a regulator of transcription, and JAZF1 acts as a strong repressor of DR1-dependent transcriptional activation by TAK1 (Nakajima et al., 2004). The PHF1 gene encodes a protein with significant sequence similarity to the protein encoded by the Drosophila polycomblike (PCL) gene. The Drosophila polycomb group of genes (PcG) has been shown to be required for the maintenance of repression of a number of key developmental regulatory genes, including the homeotic genes (Kennison, 1995; Simon, 1995; Coulson et al., 1998). At the moment, no studies on the human PHF1 gene have been published that may hint to its function in human neoplasia. The PHF1 gene is located in chromosomal band 6p21, which is the third most commonly rearranged band in ESS, involved as it is in ten of altogether 34 reported tumors and with different translocation partners (Fresia et al., 1992; Laxman et al., 1993; Hrynchak et al., 1994; Fuzesi et al., 1995; Gil-Benso et al., 1999; Sonobe et al., 1999; Micci et al., 2003, 2006). In four cases, it was recombined with chromosomal region 7pp22]p15 and in two of these, the already mentioned JAZF1/PHF1 fusion gene was identified (Micci et al., 2006). It is not known if the same fusion or involvement of other gene(s) from 7p occurred in the other tumors; however, the involvement of PHF1 unquestionably defines a new pathogenetic subgroup of ESS. The consistent involvement of PHF1 in ESS is further underscored by the demonstration of another seemingly ESS-specific fusion between PHF1 and the enhancer of polycomb (EPC1) gene from 10p11.2 in an ESS with a three-way t(6;10;10)(p21;q22; p11.2) translocation (Micci et al., 2006). It is important to mention that the three tumors in which PHF1 was found rearranged, were all ESS of classic histology. More samples from different types of EST need to be investigated to see
192
Cytogenet Genome Res 118:190–195 (2007)
how widely this gene is involved in not only ESS, but possibly also ESN and UES. The remaining 12 scientifically reported tumors (Dal Cin et al., 1988; Havel et al., 1989; Fletcher et al., 1991; Laxman et al., 1993; Iliszko et al., 1998; Gil Benso et al., 1999; Amant et al., 2003; Leunen et al., 2003; Micci et al., 2003) showed no direct involvement of chromosomal bands 6p21 and/or 7p15 at the cytogenetic level. We would like to call particular attention to the reports by Amant et al. (2003) and Leunen et al. (2003) of two ESS with a t(X;17)(p11.2;q23) and a t(10;17)(q22;p13), respectively, as the sole karyotypic aberration in the tumor cells. No molecular investigations had been performed on those tumors to detect possible cryptic rearrangements of chromosomes 6, 7, and 17 and the genes PHF1, JAZF1, and JJAZ1 already known to be recurrently involved in ESS. However, the identification of such karyotypes raises the suspicion that even more alternative pathogenetic translocations and, by inference, gene fusions, exist. Genomic imbalances leading to loss of tumor suppressor genes or gain of oncogenes
A different approach to detect pathogenetic mechanisms operative in EST was taken by Halbwedl et al. (2005) who tested for genomic imbalances in nine ESS and three UES using comparative genomic hybridization (CGH; Kallioniemi et al., 1992). They found a variety of gains and losses that apparently did not correlate with histologic grade. Nor was there any clear cut increase in copy number changes from ESS to UES, as the average of copy number alterations per case was 3.6 in ESS and 3.0 in UES. The authors reached these numbers by dividing the total number of recorded alterations by the total number of tumors, i.e., also including those tumors in the denominator that showed no alterations (tumors in which the generation of fusion genes or other balanced rearrangements were most likely the crucial pathogenetic event). If we recalculate the numbers dividing only by those tumors that actually showed imbalances, the average number of copy alterations (ANCA) index becomes 4.1 for ESS and 4.5 for UES. These results are in agreement with the idea that the number of imbalances increases with the aggressiveness of the tumor (Ried et al., 1999). The only consistent imbalance seen was loss of chromosomal arm 7p in five cases, with a common overlapping region corresponding to chromosomal band 7p21. Halbwedl et al. (2005) therefore suggested that, as in Wilms tumor, loss of heterozygosity on 7p could be associated with tumor progression also in EST. We think it may be of relevance to note that 7p21 is the chromosomal band telomeric to 7p15, where the JAZF1 gene is located. It is well known that in CGH analysis a blurring of observed imbalance margins may result from the relatively arbitrary thresholds used to score significant deviations. It is therefore possible that the common rearranged region identified by Halbwedl et al. (2005) is slightly more proximal with involvement of the JAZF1 gene also in unbalanced rearrangements. As no cytogenetic data were
available on the tumors examined by CGH, one can only speculate that fusion involving the JAZF1 gene may have occurred in the same tumors that also showed 7p loss. Anyway, the suggestion that loss from chromosomal arm 7p leads to tumor progression in EST merits more detailed studies. It is important to stress at this point that one gains much more information about tumor genomes from a multimodal approach using different techniques, and that ideally chromosomal banding analysis should be included to provide a screening background against which the various molecular-level investigations can be assessed. Microsatellite instability (MSI) and loss of heterozygosity (LOH)
Amant et al. (2001) and Moinfar et al. (2004) studied a total of 32 EST, including 25 ESS, four ESN, and three UES, and showed that the tumors were microsatellite stable. The same tumors were investigated also for LOH but the results were then less clear-cut. Amant et al. (2001) found no LOH in the five tumors they examined. Moinfar et al. (2004) studied a total of 27 tumors (20 ESS, four ESN, and three UES) as well as samples of non-tumorous surrounding tissue. LOH was found to be a frequent event in EST as 10 ESS, all three UES, and two ESN showed LOH with at least one polymorphic DNA marker. Moreover, LOH was identified also in seemingly tumor-free surrounding myometrial and endometrial tissues not only close to, but even distant from the tumors. Surprisingly, not only did ten of the 15 EST that were associated with LOH in their study show loss at the PTEN locus in 10q23, but PTEN alterations were identified even more frequently in adjacent and seemingly normal tissues. Moinfar et al. (2004) therefore suggested the intriguing possibility that EST is not just a localized genetic disease of the endometrium but rather the result of a micro-environmental genetic alteration occurring in different tissue components of the uterus. However, one should also keep in mind that false positive scoring of LOH in normal tissues may occur both from the inherent imperfection of this methodology and from contamination of normal tissues by tumor samples and cells. The use of repeated experiments and several polymorphic markers has been advocated to overcome these methodological problems (Tomlinson et al., 2002). Concluding remarks
A neoplastic phenotype may be brought about in somatic cells by the stepwise accumulation of genetic as well as epigenetic changes (Nowell, 1976; Hanahan and Weinberg, 2000). The initiating event often seems to be either a mutation (be it at the genic, chromosomal or genomic level) leading to loss of function of a tumor suppressor gene, or deregulation or fusion of an oncogene as a consequence of a cytogenetic rearrangement (Rabbitts, 1994; Rowley, 2001). This distinction is not only conceptually important but also
seems to correlate broadly with major tumor types: whereas acquired genetic imbalances, possibly facilitated by genomic instability and leading to suppressor gene inactivation or loss, predominate in epithelial tumors, quantitative or qualitative (gene fusion) activation of oncogenes play a substantial role in hematological malignancies and mesenchymal tumors (Lengauer, 2001; Albertson et al., 2003). The detection of disease-specific chromosomal translocations and gene fusions in ESS is in this sense typical of what has been previously found also in many other sarcomas as well as benign connective tissue tumors. It is equally obvious, however, that not all ESS have the cytogenetic rearrangements that we now know to be typical of endometrial stromal tumors, so we here seem to face a situation that is commonly encountered in cancer cytogenetics: several genotypes may lead to the same phenotype. It is possible, therefore, that some of the many genomic imbalances observed may act as initiating events in tumorigenesis, or at the very least they may contribute to tumor progression. For example, one may envisage that an already neoplastic clone harboring a 7;17translocation may acquire a more aggressive phenotype by losing one of the two derivative chromosomes or, for that matter, by acquiring any number of other genomic gains and losses. The mechanisms whereby this may occur and the molecular consequences of such secondary changes are completely unknown. Also quite intriguing is the fact that different tumor types within the spectrum of EST (ESN, ESS, and UES) have been found to carry the same JAZF1/JJAZ1 gene fusion, at least by PCR analysis. Does this indicate that the benign ESN may transform into the malignant ESS, which in its turn may change into UES? At this point in time such a scenario remains highly speculative, albeit not illogical, given the available genetic data. If such transformation occurs, we would assume that genomic changes beyond the primary translocation account for or at least contribute to it. Since some ESS have been shown karyotypically to have no other cytogenetic change than t(7;17)(p15;q21) corresponding to the JAZF1/JJAZ (Koontz et al., 2001), at least in these tumors the putative additional changes making the tumor an ESS and not an ESN must have been submicroscopic. Be that as it may, the new genetic knowledge indicates that EST with 7p15 rearrangements constitute a distinct pathogenetic entity, whether the chromosomal changes correspond to a JAZF1/JJAZ1, a PHF1/JAZF1 or other, as yet unidentified fusion genes. Such a pathogenetic classification of tumors rather than the current phenotypic one is likely to become mandatory once selective tailor-made therapies against the specific molecular consequences of the pathogenetic rearrangement become available. It should be noted, however, that our understanding of the molecular pathology of the 7p15 rearrangements is still in its infancy and that as of yet no therapeutic options are available based on this putative pathogenetic scheme; hence, the genetic grouping of these tumors under the term EST is still only of indirect importance. In future analyses of the genomic alterations that characterize ESS and, in general, EST, one should increasingly
Cytogenet Genome Res 118:190–195 (2007)
193
pay careful attention not only to the accurate description of all observed chromosomal abnormalities, but also to their functional outcome at the molecular level as well as their phenotypic correlates. In this context it is relevant to mention the attempt by Hrzenjak et al. (2004) to identify genes differentially expressed in ESS and UES in comparison with normal endometrium using gene expression profiling and ESS- and UES-specific cDNA libraries. They identified differences corresponding to several genes, some of which are involved in transcriptional activation, cell-
cycle progression, and cell-cell adhesion. Among the genes found to be significantly down-regulated in ESS and UES were SFRP1 and SFRP4, whose proteins are important members of the Wnt-signalling pathway, a complex cascade of heterogeneous molecules playing an important role in organ development as well as in the development of several tumor types (Hrzenjak et al., 2004). Whether interference with the Wnt system and the JAZF1 occurs in the same tumors or if these two are alternative pathogenetic mechanisms, is unknown.
References Albertson DG, Collins C, McCormick F, Gray JW: Chromosome aberrations in solid tumors. Nat Genet 34:369–376 (2003). Åman P: Fusion genes in solid tumors. Semin Cancer Biol 9: 303–318 (1999). Amant F, Dorfling CM, Dreyer L, Vergote I, Lindeque BG, Van Rensburg EJ: Microsatellite instability in uterine sarcomas. Int J Gynecol Cancer 11:218–223 (2001). Amant F, Moerman P, Cadron I, Hagemeijer A, Vergote I, Debiec-Rychter M: Endometrial stromal sarcoma with a sole t(X;17) chromosome change: report of a case and review of the literature. Gynecol Oncol 88: 459–462 (2003a). Amant F, Moerman P, Cadron I, Neven P, Berteloot P, Vergote I: The diagnostic problem of endometrial stromal sarcoma: report on six cases. Gynecol Oncol 90: 37–43 (2003b). Bennicelli JL, Barr FG: Chromosomal translocations and sarcomas. Curr Opin Oncol 14: 412– 419 (2002). Berchuck A, Rubin SC, Hoskins WJ, Saigo PE, Pierce VK, Lewis JL Jr: Treatment of endometrial stromal tumors. Gynecol Oncol 36: 60–65 (1990). Chu P, Arber DA: Paraffin-section detection of CD10 in 505 nonhematopoietic neoplasms. Frequent expression in renal cell carcinoma and endometrial stromal sarcoma. Am J Clin Pathol 113:374–382 (2000). Chu PG, Arber DA, Weiss LM, Chang KL: Utility of CD10 in distinguishing between endometrial stromal sarcoma and uterine smooth muscle tumors: an immunohistochemical comparison of 34 cases. Mod Pathol 14: 465–471 (2001). Coulson M, Robert S, Eyre HJ, Saint R: The identification and localization of a human gene with sequence similarity to Polycomblike of Drosophila melanogaster. Genomics 48: 381–383 (1998). Dal Cin P, Talcott J, Abrams J, Li FP, Sandberg AA: Ins(10; 19) in an endometrial stromal sarcoma. Cancer Genet Cytogenet 36: 1–5 (1988). Dal Cin P, Aly MS, De Wever I, Moerman P, Van Den Berghe H: Endometrial stromal sarcoma t(7; 17)(p15–21;q12–21) is a nonrandom chromosome change. Cancer Genet Cytogenet 63: 43–46 (1992). Druker BJ, Talpaz M, Resta DJ, Peng B, Buchdunger E, et al: Efficacy and safety of a specific inhibitor of the BCR-ABL tyrosine kinase in chronic myeloid leukemia. N Engl J Med 344:1031–1037 (2001). Evans HL: Endometrial stromal sarcoma and poorly differentiated endometrial sarcoma. Cancer 50:2170–2182 (1982). Fletcher JA, Kozakewich HP, Hoffer FA, Lage JM, Weidner N, et al: Diagnostic relevance of clonal cytogenetic aberrations in malignant soft-tissue tumors. N Engl J Med 324: 436–442 (1991).
194
Fresia AE, Currie JL, Farrington JE, Laxman R, Griffin CA: Uterine stromal sarcoma cell line. A cytogenetic and electron microscopic study. Cancer Genet Cytogenet 60: 60–66 (1992). Fuzesi L, Gunawan B, Braun S, Karl MC: Endometrial stromal sarcoma with clonal chromosomal aberrations and mixed phenotype. Cancer Genet Cytogenet 84:85–88 (1995). Gil-Benso R, Lopez-Gines C, Navarro S, Carda C, Llombart-Bosch A: Endometrial stromal sarcomas: immunohistochemical, electron microscopical and cytogenetic findings in two cases. Virchows Arch 434:307–314 (1999). Halbwedl I, Ullmann R, Kremser ML, Man YG, Isadi-Moud N, et al: Chromosomal alterations in low-grade endometrial stromal sarcoma and undifferentiated endometrial sarcoma as detected by comparative genomic hybridization. Gynecol Oncol 97:582–587 (2005). Hanahan D, Weinberg RA: The hallmarks of cancer. Cell 100:57–70 (2000). Havel G, Dahlenfors R, Wedell B, Mark J: Similar chromosomal evolution in a uterine stromomyosarcoma and in one of two leiomyomas from the same patient. APMIS 97: 143–146 (1989). Hennig Y, Caselitz J, Bartnitzke S, Bullerdiek J: A third case of a low-grade endometrial stromal sarcoma with a t(7;17)(p1421;q11.221). Cancer Genet Cytogenet 98: 84–86 (1997). Hrynchak M, Horsman D, Salski C, Berean K, Benedet JL: Complex karyotypic alterations in an endometrial stromal sarcoma. Cancer Genet Cytogenet 77:45–49 (1994). Hrzenjak A, Tippl M, Kremser ML, Strohmeier B, Guelly C, et al: Inverse correlation of secreted frizzled-related protein 4 and beta-catenin expression in endometrial stromal sarcomas. J Pathol 204: 19–27 (2004). Hrzenjak A, Moinfar F, Tavassoli FA, Strohmeier B, Kremser ML, et al: JAZF1/JJAZ1 gene fusion in endometrial stromal sarcomas: molecular analysis by reverse transcriptase-polymerase chain reaction optimized for paraffin-embedded tissue. J Mol Diagn 7: 388–395 (2005). Huang HY, Ladanyi M, Soslow RA: Molecular detection of JAZF1-JJAZ1 gene fusion in endometrial stromal neoplasms with classic and variant histology: evidence for genetic heterogeneity. Am J Surg Pathol 28: 224–232 (2004). Iliszko M, Mandahl N, Mrozek K, Denis A, Pandis N, et al: Cytogenetics of uterine sarcomas: presentation of eight new cases and review of the literature. Gynecol Oncol 71:172–176 (1998). Kallioniemi A, Kallioniemi OP, Sudar D, Rutovitz D, Gray JW, et al: Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science 258: 818–821 (1992).
Cytogenet Genome Res 118:190–195 (2007)
Kempson RL, Hendrickson MR: Smooth muscle, endometrial stromal, and mixed Mullerian tumors of the uterus. Mod Pathol 13: 328–342 (2000). Kennison JA: The Polycomb and trithorax group proteins of Drosophila: trans-regulators of homeotic gene function. Annu Rev Genet 29: 289–303 (1995). Koontz JI, Soreng AL, Nucci M, Kuo FC, Pauwels P, et al: Frequent fusion of the JAZF1 and JJAZ1 genes in endometrial stromal tumors. Proc Natl Acad Sci USA 98:6348–6353 (2001). Laxman R, Currie JL, Kurman RJ, Dudzinski M, Griffin CA: Cytogenetic profile of uterine sarcomas. Cancer 71: 1283–1288 (1993). Lengauer C: How do tumors make ends meet? Proc Natl Acad Sci USA 98: 12331–12333 (2001). Leunen K, Amant F, Debiec-Rychter M, Croes R, Hagemeijer A, et al: Endometrial stromal sarcoma presenting as postpartum haemorrhage: report of a case with a sole t(10; 17)(q22;p13) translocation. Gynecol Oncol 91: 265–271 (2003). Lloreta J, Prat J: Endometrial stromal nodule with smooth and skeletal muscle components simulating stromal sarcoma. Int J Gynecol Pathol 11: 293–298 (1992). Major FJ, Blessing JA, Silverberg SG, Morrow CP, Creasman WT, et al: Prognostic factors in early-stage uterine sarcoma. A Gynecologic Oncology Group study. Cancer 71: 1702–1709 (1993). McCluggage WG, Sumathi VP, Maxwell P: CD10 is a sensitive and diagnostically useful immunohistochemical marker of normal endometrial stroma and of endometrial stromal neoplasms. Histopathology 39: 273–278 (2001). Micci F, Walter CU, Teixeira MR, Panagopoulos I, Bjerkehagen B, et al: Cytogenetic and molecular genetic analyses of endometrial stromal sarcoma: nonrandom involvement of chromosome arms 6p and 7p and confirmation of JAZF1/ JJAZ1 gene fusion in t(7; 17). Cancer Genet Cytogenet 144: 119–124 (2003). Micci F, Panagopoulos I, Bjerkehagen B, Heim S: Consistent rearrangement of chromosomal band 6p21 with generation of fusion genes JAZF1/PHF1 and EPC1/PHF1 in endometrial stromal sarcoma. Cancer Res 66: 107–112 (2006). Moinfar F, Kremser ML, Man YG, Zatloukal K, Tavassoli FA, Denk H: Allelic imbalances in endometrial stromal neoplasms: frequent genetic alterations in the nontumorous normal-appearing endometrial and myometrial tissues. Gynecol Oncol 95:662–671 (2004). Nakajima T, Fujino S, Nakanishi G, Kim YS, Jetten AM: TIP27: a novel repressor of the nuclear orphan receptor TAK1/TR4. Nucleic Acids Res 32:4194–4204 (2004).
Norris HJ, Taylor HB: Mesenchymal tumors of the uterus. I. A clinical and pathological study of 53 endometrial stromal tumors. Cancer 19: 755– 766 (1966). Nowell PC: The clonal evolution of tumor cell populations. Science 194:23–28 (1976). Nucci MR, O’Connell JT, Huettner PC, Cviko A, Sun D, Quade BJ: h-Caldesmon expression effectively distinguishes endometrial stromal tumors from uterine smooth muscle tumors. Am J Surg Pathol 25: 455–463 (2001). Oliva E, Young RH, Clement PB, Bhan AK, Scully RE: Cellular benign mesenchymal tumors of the uterus. A comparative morphologic and immunohistochemical analysis of 33 highly cellular leiomyomas and six endometrial stromal nodules, two frequently confused tumors. Am J Surg Pathol 19: 757–768 (1995). Oliva E, Clement PB, Young RH: Endometrial stromal tumors: an update on a group of tumors with a protean phenotype. Adv Anat Pathol 7: 257–281 (2000). Pauwels P, Dal Cin P, Van de Moosdijk CN, Vrints L, Sciot R, Van Den Berghe H: Cytogenetics revealing the diagnosis in a metastatic endometrial stromal sarcoma. Histopathology 29: 84– 87 (1996).
Rabbitts TH: Chromosomal translocations in human cancer. Nature 372: 143–149 (1994). Ried T, Heselmeyer-Haddad K, Blegen H, Schrock E, Auer G: Genomic changes defining the genesis, progression, and malignancy potential in solid human tumors: a phenotype/genotype correlation. Genes Chromosomes Cancer 25: 195–204 (1999). Rowley JD: Chromosome translocations: dangerous liaisons revisited. Nat Rev Cancer 1: 245– 250 (2001). Rush DS, Tan J, Baergen RN, Soslow RA: h-Caldesmon, a novel smooth muscle-specific antibody, distinguishes between cellular leiomyoma and endometrial stromal sarcoma. Am J Surg Pathol 25:253–258 (2001). Satoh Y, Ishikawa Y, Miyoshi T, Mukai H, Okumura S, Nakagawa K: Pulmonary metastases from a low-grade endometrial stromal sarcoma confirmed by chromosome aberration and fluorescence in-situ hybridization approaches: a case of recurrence 13 years after hysterectomy. Virchows Arch 442:173–178 (2003). Scully RE, Bonfiglio TA, Kurman RJ, Silverberg SG, Wilkinson E: Uterine corpus: World Health Organization: Histological Typing of Female Genital Tract Tumors, pp 13–31 (Springer-Verlag, New York 1994).
Simon J: Locking in stable states of gene expression: transcriptional control during Drosophila development. Curr Opin Cell Biol 7: 376–385 (1995). Sonobe H, Iwata J, Furihata M, Ohtsuki Y, Taguchi T, Shimizu K: Endometrial stromal sarcoma with clonal complex chromosome abnormalities. Report of a case and review of the literature. Cancer Genet Cytogenet 112: 34–37 (1999). Sreekantaiah C, Li FP, Weidner N, Sandberg AA: An endometrial stromal sarcoma with clonal cytogenetic abnormalities. Cancer Genet Cytogenet 55:163–166 (1991). Taki T, Taniwaki M: Chromosomal translocations in cancer and their relevance for therapy. Curr Opin Oncol 18:62–68 (2006). Tavassoli FA, Devilee P: World Health Organization classification of tumors: Pathology and genetics of tumors of the breast and female genital organs. (IARC Press, Lyon 2003). Tomlinson IP, Lambros MB, Roylance RR: Loss of heterozygosity analysis: practically and conceptually flawed? Genes Chromosomes Cancer 34:349–353 (2002).
Cytogenet Genome Res 118:190–195 (2007)
195
ma cell lines derived from the esophagus, stomach, colon, lung, breast, kidney, uterine cervix and other organs, and Fhit protein expression is reduced or absent in the majority of primary tumors (reviewed in Huebner and Croce, 2003). Exogenous expression of Fhit, by vector transfection or adenoviral transduction, into cancer cell lines negative for Fhit expression, inhibited cell growth and reduced tumorigenicity in nude mice (reviewed in Ishii et al., 2004). Like the human FHIT gene, mouse Fhit spans a common fragile site on mouse chromosome 14, FRA14A2 (Glover et al., 1998), and is altered in mouse cancer cells. The murine and human Fhit protein sequences share 91% identity (Pekarsky et al., 1998). To investigate the role of Fhit in carcinogenesis, Fhit knockout mice were created and used in the established N-nitrosomethylbenzylamine (NMBA)-induced esophageal/gastric cancer model (Fong and Magee, 1999; Fong et al., 2000). The epithelial linings of the mouse forestomach and the squamocolumnar junction (SCJ) are comparable to the distal esophagus and esophago-gastric junction in humans, respectively, and provide an animal model to study esophageal tumorigenesis (Blot et al., 1991). Results of such studies indicated a significant difference between the wildtype mice, 8–25% of which exhibited tumors, compared to 80–100% of Fhit heterozygous mice, depending on NMBA dosage. Inactivation of one or both Fhit alleles increased susceptibility of mice to carcinogen-induced forestomach tumor formation (Fong et al., 2000), suggesting that Fhit is a haplo-insufficient tumor suppressor. In later studies, it was shown that infection with recombinant FHIT viruses significantly protected against tumor formation in Fhit heterozygous mice following NMBA treatment (Dumon et al., 2001; Ishii et al., 2003, 2004). To determine if a nonfragile, cDNA version of an FHIT allele, expressed from a chromosomal region outside the fragile site, could reduce susceptibility of mice to carcinogen-induced tumorigenicity, we generated FHIT transgenic mice on an Fhit+/+ and +/– background, treated them with NMBA, and assessed their tumor burden relative to wildtype and Fhit+/– mice.
Materials and methods The FHIT transgene The FHIT recombinant transgene vector was constructed by replacing the luciferase-coding region of the pGL2-Basic plasmid (Promega, Madison, WI) with the 444-bp fragment of human FHIT cDNA containing coding exons 5 through 9 flanked by HindIII /ClaI sites. The 4.4-kb DNA fragment proximal to exon 1 and containing the FHIT promoter was cloned into the SacI /XhoI sites of the modified pGL2-Basic vector upstream of FHIT cDNA. Lastly, the 5 UTR from FHIT cDNA was amplified and cloned into the XhoI site of the pGL-2 Basic plasmid polylinker. The resulting plasmid, designated pFHITPro4.4, was sequenced using Taq Dye Deoxy-Terminator Cycle Sequencing kit (Applied Biosystems Inc., Foster City, CA). The full length FHIT expression cassette was excised from the pFHITPro 4.4 plasmid by SmaI /Sal I digestion, purified, and injected into the male pronuclei of fertilized eggs from C3H/HeJ mice. The microinjected eggs were re-implanted into pseudopregnant C57BL/6J female mice. Pups were screened for transgene integration by Southern blot analysis. Two founder lines, C3H/HeJ-Tg(FHIT )664Kcc and C3H/
HeJ-Tg(FHIT )665Kcc, hereafter referred to as 664 or 665, were established and maintained according to Thomas Jefferson University Institutional Animal Care and Use Committee and NIH guidelines. These C3H transgenic strains were crossed with a) C57BL/6J wildtype mice to produce C3H/HeJ.B6-Tg(FHIT )664Kcc and 665Kcc strains with two endogenous murine Fhit alleles (referred to hereafter as Fhit+/+tg) or b) with the N10 generation of C57BL/6JX129Fhit–/– to produce C3H/HeJ.B6-Tg(FHIT )664Kcc hemizygous for murine endogenous Fhit (hereafter referred to as Fhit+/–tg), respectively. The parental transgene lines were maintained by crossing to isogenic C3H/HeJ.
Genotyping SstI-digested mouse tail DNA was screened for presence of the transgene by Southern blot analysis using labeled full-length FHIT expression cassette as a probe to identify transgenic founder mice. DNA was extracted from tail biopsies, digested in 50 mM Tris (pH 8.0), 100 mM EDTA (pH 8.0), 100 mM NaCl, 1% SDS and 500 g Proteinase K (Invitrogen, Carlsbad, CA). DNA was digested by SstI (Roche Applied Science, Indianapolis, IN), and electrophoresed. Blots were hybridized to random-primed [-P32]dCTP (50 Ci) labeled full-length FHIT cDNA. Subsequent genotyping was by PCR amplification. Tail DNA was amplified in a PCR reaction containing PCR Buffer, 0.2 mM dNTPs, 0.2 U Taq and 20–40 ng of primers: UR4 (5CTGCTCTGTCCGGTCACA3) and 642R (5AACATGGACGTGAACGTG3). These primers specifically amplify a 650-bp human cDNA fragment. The reaction mixture was denatured at 94 ° C for 5 min, followed by 30 cycles of 94 ° C for 30 s, 58 ° C for 30 s and 72° for 30 s, and extended for an additional 5 min at 72 ° C in a GeneAmp 9600 Thermocycler (Perkin Elmer Cetus). Cytogenetics and FISH analysis Mouse transgenic strains, 664 and 665, were evaluated cytogenetically by routine G-banding analysis. Bone marrow was extracted from the femur of anesthetized, sacrificed mice and cultured briefly in Giant Cell Tumor (GCT)-conditioned medium containing 20% FBS. The cultured bone marrow cells were arrested in metaphase by addition of colcemid (Gibco, Rockville, MD), treated with hypotonic solution (buffered 0.075 M KCl) and fixed in 3: 1 methanol:acetic acid (Seabright, 1971). Fixed cells in suspension were subsequently dropped on glass slides and examined for metaphase spreads. Prepared metaphase chromosomes were GTW-banded (Giemsa-Trypsin banding technique using Wright’s stain) and karyotyped using an automated capture and imaging system (Applied Imaging, San Jose, CA). To determine chromosomal localization of the FHIT transgene in 664 and 665 cells, FISH (fluorescent in situ hybridization) was performed using a modification of previously published techniques (Pinkel et al., 1986; Shi et al., 1994). The pFHITPro4.4 plasmid (9.3 kb) was direct-labeled with Spectrum Green by nick translation (Vysis Inc., Downers Grove, IL) and hybridized to metaphase spreads prepared from bone marrow of 664 and 665 transgenic mice. After overnight hybridization, the slides were washed in 0.4! SSC/0.3% NP-40 at 73 ° C for 2 min, followed by a 1 min wash in 2! SSC/0.1% NP-40 at ambient temperature. The slides were counterstained with DAPI (4,6-diamidino-2-phenylindole) and analyzed using a Leitz microscope. Color images of metaphase spreads showing hybridization with the FHIT transgene probe were obtained with Cytovision (Applied Imaging, San Jose, CA). RNA extraction and RT-PCR Total RNA was extracted from mouse tissues by lysis in RNA-Bee (Tel-Test, Friendswood, TX) according to manufacturer’s instructions, followed by DNaseI-treatment (Fisher Scientific, Pittsburgh, PA). Reverse Transcription was performed on 2 g of RNA incubated with first strand buffer, 600 U M-MLV reverse transcriptase, 40 U of RNasin (Promega, Madison, WI), 0.2 mM of dNTPs, and human FHIT 745R primer, located in exon 9, (5GCCATTTCCTCCTCTGAT3) at 37 °C for 90 min. Reverse-transcribed products were subjected to nested-PCR amplification, as previously described (Ohta et al., 1996). Briefly, 1–2 l of human cDNA was first amplified by primers UR4 and X8 (5TTTCAGAGGACTGCTACCTCTTT3) and the product (1 l)
Cytogenet Genome Res 118:196–203 (2007)
197
was subsequently amplified by primers UR5 (5CTGTAAAGGGTCCGTAGTC3) and X7/8 (5TCTTTGGAGTCCTCAGTG3), nested between the original primers.
Carcinogen treatment To examine the effect of NMBA on FHIT transgenic (+/+tg) and +/+ mice, cohorts of pups (4–6 wk old) were placed on a zinc-deficient diet, a custom-formulated, egg white-based diet containing 1.5 ppm of zinc, prepared by Teklad (Madison, WI). Wildtype mice are much less susceptible to NMBA-induced tumors and zinc deficiency increases their susceptibility (Fong and Magee, 1999), so that control Fhit+/+ mice show a significant tumor burden. After 4 weeks, mice received 4 intragastric doses of NMBA (Ash Stevens, Inc., Detroit, MI) at a concentration of 2 mg/kg body weight, administered over 2 weeks. Mice were autopsied 14 weeks following the first dose. Whole stomachs were excised, opened longitudinally, and forestomach and SCJ assessed for tumor formation. Tissues were fixed in buffered formalin and embedded in paraffin. Serial cross-sections (4 m) were prepared and stained with hematoxylin and eosin (H&E) for histological analysis. To examine the effect of NMBA on transgenic Fhit+/–tg mice, 4–6-week-old pups received 4 intragastric doses of NMBA, as described above, and were sacrificed 14 weeks following the first administered dose. Forestomach and squamocolumnar junction were scored for tumor formation as described above. Statistical analysis Statistical analysis of tumor incidence differences was performed by two-tailed Fisher’s exact test (Biostat, www.matforsk.no/ola/fisher. htm). The statistical significance of results of tumor burden data was calculated by the Mann-Whitney test using GraphPad InStat 3.0. Cell line establishment and maintenance TgG1-4 and TgG1-9 murine kidney cell lines were established by culturing disaggregated kidney tissue dissected from a female and male 664 Fhit–/–tg mouse, respectively, in MEM supplemented with 10% FBS and 100 g/ml gentamycin, as previously described (Ottey et al., 2004). Cells were incubated at 37 ° C in 5% CO2 and maintained in sterile environments. DNA methylation analysis Genomic DNA was isolated from kidney tissues of male and female mice of the Tg(FHIT )664 and 665 strains. 20 g of genomic DNA was first subjected to SacI (Promega, Madison, WI) digestion overnight, followed by digestion with either HpaII or MspI enzymes (Roche Diagnostics, Indianapolis, IN). The digested DNAs were electrophoresed and analyzed by Southern blot, as described above; and FHIT transgene restriction fragments were detected using labeled SacI-digested pFHITPro4.4 as probe. Inhibition of DNA methyltransferase and histone deacetylases TgG1-4 and TgG1-9 cells were seeded at a density of 5 ! 105 cells/ 100-mm dish and after 24 h, treated with either 10 M 5-Aza-2-deoxycytidine (5-Aza-dC, Sigma Aldrich, St. Louis, MO) for 6 days, 1 M Trichostatin A (TSA, Sigma Aldrich, St. Louis, MO) for 2 days, or a combination of 5-Aza-dC for 6 days and TSA for 1–2 days. Medium and reagents were replaced daily. Immunoblot analysis Protein lysates were prepared from homogenized mouse tissues or murine kidney cells lines in lysis buffer (150 mM NaCl, 30 mM Tris (pH 7.5), 10% glycerol, 1% NP40, 10 g/ml of chymostatin, leupeptin, aprotinin, pepstatin, and 1 mM phenylmethylsufonyl fluoride). Total protein was electrophoresed in Tris-HCl polyacrylamide gels, transferred to Hybond N+ nylon membrane (Amersham Biosciences, Piscataway, NJ) and blocked in 5% non-fat dry milk in PBS with 0.1% Tween20 for 1 h. Membranes were incubated with rabbit polyclonal anti-Fhit serum at 1: 5000 dilutions, followed by incubation with horseradish peroxidase-labeled goat anti-rabbit immunoglobin at 1: 10000 dilutions. Signal was detected by SuperSignal Chemiluscent substrate (Pierce, Rockford, IL).
198
Cytogenet Genome Res 118:196–203 (2007)
Results
Generation and characterization of FHIT transgenic mice To evaluate the effect of expression of an exogenous FHIT allele, transgenic mice were produced carrying human FHIT cDNA under control of the natural FHIT promoter (Fig. 1A). Detection of a 4.4-kb product in Southern blot screening of SacI-digested tail genomic DNA identified several founders with an integrated transgene, three of which transmitted the transgene through the germline. Two of the founders were used to establish the Tg(FHIT )664 and Tg(FHIT )665 transgenic mouse lines (Fig. 1B, left). Further genotyping of weanlings were carried out by allele-specific PCR amplification, in which the presence of the integrated transgene was determined by amplification of a PCR product from primers located to exons 1 and 3 of the human FHIT gene (Fig. 1B, right). To evaluate expression of the FHIT transgene under the control of FHIT promoter sequence, RNA, isolated from various tissues of transgenic mice, was used for reverse transcription (RT)-PCR analyses. In human tissues, endogenous FHIT is ubiquitously expressed at low levels, therefore RT-PCR analysis was performed as previously described (Druck et al., 1997), by nested PCR reactions. Expression of the transgene was found in all tissues examined, as shown for kidney, liver, heart and muscle, tissues known to express endogenous Fhit protein in mouse and human (Fig. 1C), suggesting that the transgene under the control of its natural FHIT promoter leads to ubiquitous expression of the transgene. The transgene was expressed at low levels, confirming the results of immunohistochemical analysis of histological sections of transgenic mouse tissues, showing very weak staining (data not shown). Tissues were also examined for expression of Fhit protein by immunoblot analysis but results were difficult to interpret due to cross-reaction of mouse Fhit protein with human Fhit antiserum (data not shown). To determine if FHIT transgene integration sites were different in the two mouse strains, transgene integration sites were examined by FISH and cytogenetics analysis. Metaphase spreads made from bone marrow cells collected from the 664 and 665 mice were stained by Wright’s stain for routine G-banding analysis. Evaluation of the karyotypes, from several metaphase spreads, revealed that both the 664 and 665 cell lines had normal complements. FISH analysis confirmed that the transgene had integrated at single distinct chromosome sites in the two established transgenic strains. The transgene was integrated into chromosome 9 in the 664 transgenic strain (Fig. 1D, left), and into chromosome 2 in the 665 transgenic strain (Fig. 1D, right). The two transgenic strains carry 5–10 tandem transgene copies, as determined by Southern blot and FISH analyses, by comparison to the two copies of a single FHIT exon in normal human DNA (data not shown).
444 bp
Sall 3' UTR, PolyA
800 bp 600 bp 500 bp
Muscle
Heart
C
Liver
1.3 kb
Kidney
7004
7001
664
664E-5
361 bp 14 bp
664E-4
4.4 kb
664E-3
FHIT
664
665
Sacl Clal
5' UTR
664E-2
B
Hindill
FHIT PRO 4.4
664E-1
–294 bp 0
Xhol
Hpall
Pvul Smal
685
Xhol
Xbal
Hpall
BspHl
Hpall Hpall
Sacl
Hpall
A
200 bp
D
Fig. 1. Generation and characterization of Tg(FHIT ) mice. (A) Schematic of construct used to generate FHIT transgenic mice. The drawing shows the three FHIT regions, FHIT coding exons 5 to 9 and 3 UTR, downstream of FHIT 5 UTR (exons 1 to 4) and FHIT promoter region. Several enzyme sites are shown. (B) Transgene identification and transmission. Left: Southern blot identification of 665 and 664 founder mice by a 5.0-kb product, 685 being negative for transgene integration and transmission of transgene to offspring, 7001 and 7004, offspring of 664 founder. Right: The 650-bp PCR product distinguish-
ing transgenic mice from wildtype littermates of the 664 mouse strain. (C) RT-PCR analysis of transgene expression. DNase I treated RNAs isolated from indicated tissues of a 664 transgenic mouse were subjected to reverse-transcription followed by nested PCR amplification of a 200-bp product of FHIT cDNA. (D) Transgene integration analysis by FISH. DAPI-stained metaphase spreads from bone marrow cells collected from a transgenic mouse of the 664 line (left) and the 665 line (right).
Carcinogen induction of tumor formation Fhit+/+tg mice. NMBA induces forestomach tumors in only 8–25% of wildtype C57BL/6J mice, depending on number of doses (Fong et al., 2000; Zanesi et al., 2001). Therefore to study protection of NMBA induced tumors by the transgene, zinc-deficient mice were used in the initial experiments. Zinc deficiency induces cell proliferation in esophageal and forestomach cells of mice, strengthening the impact of NMBA (Fong and Magee, 1999) and inducing forestomach tumors in up to 100% of normal mice. Groups of Fhit+/+tg and Fhit+/+ littermates from the 664 transgenic mouse line were placed on a zinc-deficient diet for 4 weeks, followed by NMBA treatments; 14 weeks following administration of the first NMBA dose, the mice were autopsied for tumor scoring. The resulting data showed a 2-fold higher tumor burden for male +/+ mice
than male +/+tg mice; 80% of the male +/+ mice had tumors, a 32% higher incidence (P = 0.03) than observed in male +/+tg mice. Additionally, the average tumor burden per mouse was nearly two-fold higher (1.88 tumors/mouse) in the +/+ males than in the +/+tg males (1 tumor/mouse; P = 0.03). For the female wildtype and transgenic mice, there was no difference in the fractions of tumor bearing mice, or in tumor multiplicity. Thus, male +/+tg mice were significantly protected from NMBA-induced forestomach tumor development, but the female +/+tg mice were not (Table 1). Results from experiments using the 665 mouse line showed similar reduction in tumor burden for male transgenic pups compared to the wildtype littermates (data not shown). Fhit+/–tg mice. Crosses between pups of the 664 transgenic strain and Fhit homozygous knockout lines produced +/–tg and +/– pups that were used to determine the effect of
Cytogenet Genome Res 118:196–203 (2007)
199
Genotype (n)
Tumor incidence in forestomachs and SCJ Tumor-bearing micea, c
Mice harboring multiple tumorsb, c
Tumors/ moused
M, +/+ (25) M, Tg (23) P value
47 23
20/25 (80%) 11/23 (48%) 0.03
14/25 (56%) 7/23 (30%) 0.09
1.88 1 0.03
F, +/+ (22) F, Tg (25) P value
65 72
19/22 (86%) 25/25 (100%) 0.09
17/22 (77%) 21/25 (84%) 0.71
2.29 2.88 0.39
M, +/– (22) M, Tg +/– (18) P value
52 27
19/22 (86%) 15/18 (83%) 1
18/22 (81%) 9/18 (50%) 0.02
2.36 1.5 0.05
F, +/– (18) F, Tg +/– (20) P value
37 34
15/18 (83%) 16/20 (80%) 0.77
11/18 (61%) 11/20 (55%) 1
2 1.7 0.56
Gender differences in transgene expression To investigate the molecular basis for lack of protection of the female mice, expression levels of the transgene were examined in male and female kidney tissues by RT-PCR analysis. RNA was isolated from kidney tissues excised from female and male mice of the 664 and 665 strains, and subjected to reverse-transcription followed by a nested PCR reaction. The FHIT transgene was amplified in both male and female tissues of both transgenic mouse strains, showing the transgene expression was not completely silenced in female or male mice. Since the transgene is expressed at low levels in both female and male mouse tissues and requires nested PCR amplification to be detected it was not possible to determine quantitatively a difference in transgene expression in female and male mice (Fig. 2).
Cytogenet Genome Res 118:196–203 (2007)
665
Positive control Negative control
664
Male
665
Female
664
Male
a non-fragile Fhit allele on the NMBA tumor burden of Fhit+/– mice; these mice are highly susceptible to NMBAinduced forestomach tumorigenesis, so zinc-deficient diets were unnecessary. Male and female littermates of these crosses were treated with NMBA and again autopsied 14 weeks after the first dose. No significant difference in the percentage of +/– or +/–tg mice bearing tumors was found for either sex but there was a 31% higher incidence of multiple forestomach tumors in male +/– than male +/–tg mice (P = 0.02). The male +/– mice also had a significantly higher tumor burden than the male +/–tg mice (P = 0.05), although there was no difference between the female groups (Table 1). Thus, perhaps surprisingly in view of the low level of expression of the FHIT transgene, the transgenic male mice were significantly protected from NMBA-induced forestomach tumors.
Female
d
Male
c
Tumor-bearing mice (number of mice with tumors/total mice). Mice harboring multiple tumors (number of mice with >1 tumor/total mice). P values calculated by two-tailed Fisher’s exact test. P values calculated by Mann-Whitney test.
Female
b
Male
a
200
Total tumors
Female
Table 1. NMBA-induced tumor formation in FHIT transgenic mice. Total tumors (>0.5 mm in diameter) were counted in the forestomach and SCJ.
Fig. 2. FHIT transgene expression in female and male mice. DNaseI-treated RNA isolated from kidney tissues of female and male mice of the 664 and 665 transgenic strains, was subjected to reversetranscription followed by nested PCR amplification of a final product of 150 bp corresponding to expressed FHIT transgene. Mouse strain and sex of each sample are indicated. Positive control is represented as RT-PCR product amplified from RNA isolated from 293 human kidney cell line. A negative RT sample, lacking RNA, was included in nested PCR amplification, represented by negative control lane.
Methylation of the transgene The significant protection of male FHIT transgenic mice but not female transgenic mice raises a question concerning the difference in activity of the transgenes in male and female mice, possibly due to the differential epigenetic modifications of the transgenes in the two sexes. To determine if this is the case, methylation of CpG islands of the 5 regulatory region of the FHIT transgene was examined by characterization of digestion patterns of transgene DNA cleaved with methylation sensitive and insensitive restriction enzymes. Transgene DNA, extracted from kidney tissue of female and male Fhit+/+tg mice, was excised from surrounding genomic DNA by SacI digestion, followed by digestion with either HpaII (methylation sensitive) or MspI (methylation insensitive). Southern blot analysis of digested DNA
Sac I
Msp I
Male Hpa II
Msp I
Female Hpa II
Msp I
Male Hpa II
Msp I
Hpa II
Sac I
Female
TSA 5-Aza-2-dC
– –
TgG1-4 – + + –
+ +
– –
TgG1-g – + + + – +
FHIT
664
665
GAPDH
Fig. 3. Methylation status of CpG sites in 5 regulatory regions of FHIT transgene. SacI-digested DNAs isolated from kidney tissues of female and male 664 and 665 transgenic mice were digested at CCGG sites by methylation-sensitive (HpaII) and -insensitive (MspI) restriction enzymes. Digested products were used in Southern blot analysis and detected by a radio-labeled transgene targeting vector. SacI-digested products used as control for fragment size determinations are shown.
Fig. 4. Epigenetic control of FHIT transgene expression. Example of immunoblot analysis of human Fhit expression in murine kidneyderived cells, established from kidney tissue of female (TgG1-4) and male (TgG1-9) Fhit–/–tg mice, following treatment with inhibitors of DNA methyltransferase and histone deacetylase. Both cell lines were treated with either 10 M 5-Aza-2-dC (6 days), 1 M TSA (2 days), a combination of 5-Aza-2-dC and TSA or left untreated, and lysed for analysis of total protein.
products, using the transgene targeting vector as probe, revealed different fragment patterns created by digestion with MspI or HpaII (Fig. 3; see Fig. 1A for map of HpaII sites); however the resulting fragments of DNA from female and male transgenic mice were similar or equal in length. As expected, digestion of transgene DNA with MspI resulted in a single fragment of about 4.3 kb (Fig. 3). The fragment corresponds to the larger transgene fragment resulting from the digestion of the most upstream CCGG recognition site, located at about –108 relative to the start site of exon 1. Digestion at any additional downstream CCGG sites by MspI would have resulted in fragments too small to be detected on this gel. Two transgene fragments were detected following digestion with HpaII, a robust 5.1-kb fragment and a smaller fragment of 4.4 kb (Fig. 3). The presence of a 4.4-kb fragment produced by HpaII, rather than a 4.3-kb fragment reveals that the first CCGG site at –108 of transgene DNA is methylated and therefore could not be recognized by HpaII. An unmethylated CCGG site located downstream in exon 1 must be present in some transgene DNA, allowing for HpaII digestion and production of the smaller fragment. The larger 5.1-kb fragment is equal in size to the transgene fragment produced by SacI digestion alone, suggesting that some transgene DNA is methylated at all CCGG sites and was not digested by the methylation-sensitive HpaII. The presence of digested and undigested fragments of the transgene DNA following HpaII digestion would suggest that the multiple copies of the integrated transgene are differentially methylated, with the majority of the copies being methylated at all CCGG recognition sites within the 5 regulatory region of the FHIT transgene. DNA methylation and histone deacetylation were further investigated in vitro for a possible role in the reduced expression of the FHIT transgene in female and male mice. Previous studies of murine cancer cell lines containing high percentages of methylated CpG sites within the promoter, exon 1 and intron 1 regions of the Fhit gene, were shown to re-express Fhit following treatment with inhibitors of DNA methyltransferase and histone deacetylases (Han et al., 2004). TgG1–4 and TgG1–9 murine kidney cell lines, established from kidney tissue of a female and male Fhit–/–tg
mouse, respectively were treated with a DNA methyltransferase inhibitor, 5-Aza-dC, and TSA, a histone deacetylase inhibitor, either individually or in combination and examined by immunoblot analysis. As mentioned earlier, protein expression from the FHIT transgene in both male and female mice was not sufficient for detection by immunoblot analysis. Treatment of cells with 5-Aza-dC did not result in a detectable difference relative to untreated cells in either cell line. TSA treatments increased transgene expression to detectable levels in both TgG1-4 and TgG1-9 cell lines. Increased transgene expression was also found in both cell lines following a combination treatment of 5-Aza-dC and TSA (Fig. 4). The lack of detectable protein expression in cells treated with the inhibitor of DNA methyltransferase alone suggests that the methylation of the transgene is not the sole cause for reduction of transgene expression. The elevated transgene expression resulting from inhibition of the histone deacetylation suggests that deacetylation of chromatin in the region of the transgene has contributed to silencing of the FHIT transgene. It is possible that the epigenetic silencing of the transgene is more complete in the female than the male transgenic mice. Discussion
The inactivation of one Fhit allele in mice was previously shown to result in significantly increased susceptibility of mice to spontaneous and carcinogen-induced tumor development (Fong et al., 2000). To explore the importance of FHIT locus fragility in tumor susceptibility, a non-fragile FHIT cDNA was introduced as a constitutively expressed transgene to test for effects on tumor susceptibility. To achieve natural expression levels of transgenes, some groups have successfully attempted bacterial artificial chromosome (BAC) or yeast artificial chromosome (YAC) transgenesis, thus introducing the gene of interest surrounded by regulatory sequences. The use of such large genomic fragments protects the transgene from position effects associated with transgene integration sites (Giraldo and Montoliu, 2001). Although FHIT is expressed as a 1.1-kb mRNA
Cytogenet Genome Res 118:196–203 (2007)
201
transcript, the FHIT gene itself is 1.7 Mb in length (Matsuyama et al., 2003). Therefore, to achieve a level of FHIT cDNA expression, similar to endogenous Fhit expression, a 4.4-kb region of the FHIT promoter and exons 1 through 4 of the 5 UTR of FHIT cDNA were included in the targeting vector used to generate FHIT transgenic mice. By cloning the FHIT cDNA under control of its natural promoter, the expression of the FHIT transgene allele was intended to be similar to endogenous Fhit expression in human tissues. Analysis of protein levels revealed the expression of the transgene was even lower than expected. There was very weak staining of human Fhit in tissue sections and barely detectable levels of human Fhit protein in lysates from tissues of mice from either the 664 or the 665 transgenic strains, as determined by immunohistochemical and immunoblot analysis. Evaluation of Fhit protein level is complicated due to the absence of antisera allowing selective identification of the human Fhit protein in mouse tissue. Therefore, the low level of FHIT transgene expression was confirmed at the mRNA level, by RT-PCR analysis. FHIT cDNA from transgenic mice was amplified by nested PCR reactions, as performed for FHIT cDNA in human tissues (Druck et al., 1997). FISH analysis showed that the transgene was integrated into different chromosomes in the 664 and 665 mouse strains, suggesting that the low level of transgene expression was not due to a specific integration site. FHIT transgenic and wildtype mice were used in the NMBA-induced esophageal and gastric cancer model to determine the effect of the nonfragile FHIT alleles on carcinogen-induced tumor development, using two approaches. In one experiment, mice were made zinc-deficient prior to NMBA treatment, augmenting susceptibility of mice to the carcinogen. Treatment with NMBA resulted in the formation of twice as many tumors per mouse in male wildtype mice relative to male transgenic mice, showing that the FHIT transgene, although expressed at a low level, significantly reduced susceptibility of male mice to NMBA-induced tumor development. The protection was not observed in female transgenic mice. The second approach took advantage of the susceptibility of Fhit deficient mice to tumor development (Fong et al., 2000; Zanesi et al., 2001) to determine the effect of the nonfragile FHIT transgene on tumor induction in Fhit heterozygous mice. Fhit+/– and Fhit+/–tg mice, for which there is no need to enhance susceptibility by zinc deficiency, were used in NMBA experiments. Again, there was a significant reduction in the tumor burden of male +/–tg mice compared to male +/– mice, yet no difference found between female +/–tg and +/– mice. There was also a significant reduction in the tumor multiplicity in male +/–tg mice. The significant difference in the effect of the FHIT transgene in tumor suppression in male and female mice was unexpected. Previous experiments using the NMBA carcinogen did not show a difference between male and female mice within experimental groups; thus there is no evidence for a sex-dependent response to the NMBA carcinogen. There was also no difference in tumor burdens of the con-
202
Cytogenet Genome Res 118:196–203 (2007)
trol male and female wildtype mice, confirming that female and male C3H/HeJ mice are equally susceptible to NMBA. Similar results were found in the analysis of mice from the 665 transgenic strain, following NMBA treatments, suggesting that the difference in tumor suppression in male and female transgenic mice was not the result of chromosomal position. Thus, the difference in NMBA response must be related to differences in level of transgene expression, a difference that has been difficult to demonstrate. RT-PCR analysis of transgene expression in female and male tissues revealed that the transgene was expressed in both female and male mice, but levels of expression detected by nested RT-PCR are difficult to quantify. The differences in NMBA susceptibilities of male and female transgenic mice suggests that either the low expression of the transgene was sufficient for protection in male forestomach tissue but not in female tissues, or there was actually a difference in the level of expression in the male and female mice. To examine further the cause of lack of protection of female mice by the FHIT transgene, we hypothesized differences in the male and female epigenome in the region of the transgene. Reduced or silenced expression of transgenes is often associated with epigenetic controls, such as methylation and histone deacetylation. Repetitive DNA sequences have been suggested to be attractive templates for methylation protein complexes, possibly explaining the targeting of transgene sequences, usually integrated as multiple copies, for methylation-induced silencing (Bird, 2002). Further supporting the association of CpG methylation content with repetitive DNA sequence, a direct correlation between methylation content of transgenes and transgene copy number was reported (Pena et al., 2004). Although initial analysis by methylation-sensitive and -insensitive enzyme digestions of the transgenes showed that some FHIT regulatory CpGs were methylated, treatment of transgenic male- and female-derived cell lines with a demethylating agent did not suggest transgene methylation differences that could explain the differences in protection of male and female mice. Based on the results of the inhibitor treatments of male and female cell lines, deacetylation of transgene chromatin was involved in control of transgene expression level, although the experimental results did not explain differences in protection of male and female mice. In previous studies using Fhit heterozygous and wildtype mice in the NMBA-induced cancer model, Fhit protein was absent in some tumors that developed in both wildtype and Fhit+/– mice following NMBA treatments, suggesting carcinogen-induced DNA damage to wildtype fragile Fhit alleles (Fong et al., 2000). The purpose of the transgene study was to examine effects of expressing a nonfragile FHIT allele from a site outside of the FRA14A2 site, on NMBA-induced tumor formation. By expressing a nonfragile FHIT allele, there would be a predicted protective effect by increasing Fhit expression from an allele that would be less prone to alteration by DNA damaging agents. The FHIT transgene did result in suppression of the carcinogen-induced tumor burden of wildtype and Fhit heterozygous male mice, confirming the original hypothesis.
The apparent control of FHIT transgene expression by histone deacetylation suggests that the FHIT transgenic strain may be a useful model to further study the role of epigenetic control of tumor suppressor genes in vivo. Hypermethylation of CpG DNA is often accompanied by histone deacetylation and chromatin remodeling; although which mechanism is the initiating factor has not been determined. Opposing models suggest either DNA methylation occurs and recruits histone deacetylase complexes (Bird and Wolffe, 1999) or DNA methylation occurs after the histones are deacetylated and chromatin compacted (Bird, 2002). Therefore, although methylation of CpG islands in the 5 regulatory region of the FHIT gene has been reported in several primary tumors and cancer cell lines associated with reduced or silenced Fhit expression (Zochbauer-Muller et al., 2001), the reduced expression may be related to histone deacetylation followed by CpG methylation. The results of the FHIT transgenic study suggests a
need to further examine specific roles of epigenetic mechanisms in gene silencing. The FHIT transgenic mice also provide a model for testing efficacy of the deacetylase inhibitor (TSA) in vivo, as a pharmacological approach for increasing expression of Fhit from the FHIT transgene allele. Successful increase in tumor suppression by the FHIT transgene allele, following TSA treatments of male and female FHIT transgenic mice, could lead to further studies of deacetylase inhibitors as chemopreventive agents using these mice. Acknowledgements We acknowledge the excellent service of the Kimmel Cancer Center Transgenic and Gene Targeting Facility for excellent advice in planning and construction of the transgene and for production of the transgenic founder mice, and we thank Maria Skorski for outstanding help with the mapping of transgene integration sites.
References Bird AP: DNA methylation patterns and epigenetic memory. Genes Dev 16: 6–21 (2002). Bird AP, Wolffe AP: Methylation-induced repression-belts, braces, and chromatin. Cell 99: 451– 454 (1999). Blot WJ, Devesa SS, Kneller RW, Fraumeni JF Jr: Rising incidence of adenocarcinoma of the esophagus and gastric cardia. J Am Med Assoc 265:1287–1289 (1991). Druck T, Hadaczek P, Fu T-B, Ohta M, Siprashvili Z, et al: Structure and expression of the human FHIT gene in normal and tumor cells. Cancer Res 61:4827–4836 (1997). Dumon KR, Ishii H, Fong LYY, Zanesi N, Fidanza V, et al: FHIT gene therapy prevents tumor development in Fhit-deficient mice. Proc Natl Acad Sci USA 98:3346–3351 (2001). Fong LYY, Magee PN: Dietary zinc deficiency enhances esophageal cell proliferation and Nnitrosomethylbenzylamine (NMBA)-induced esophageal tumor incidence in C57BL/6J mouse. Cancer Lett 143: 63–69 (1999). Fong LYY, Fidanza V, Zanesi N, Lock LF, Siracusa LD, et al: Muir-Torre-like syndrome in Fhit-deficient mice. Proc Natl Acad Sci USA 97: 4742– 4747 (2000). Giraldo P, Montoliu L: Size matters: use of YACs, BACs, and PACs in transgenic animals. Transgenic Res 10:83–103 (2001). Glover TW, Hoge AW, Miller DE, Ascara-Wilke JE, Adam AN, et al: The murine Fhit gene is highly similar to its human orthologue and maps to a common fragile site region. Cancer Res 58: 3409–3414 (1998).
Han SY, Iliopolous D, Druck T, Guler G, Grubbs CJ, et al: CpG methylation in the Fhit regulatory region: relation to Fhit expression in murine tumors. Oncogene 23: 3990–3998 (2004). Huebner K, Croce CM: Cancer and the FRA3B/ FHIT fragile locus: it’s a Hit. Brit J Cancer 88: 1501–1506 (2003). Ishii H, Zanesi N, Vecchione A, Trapasso F, Yendamuri S, et al: Regression of upper gastric cancer in mice by FHIT gene delivery. FASEB J 17: 1768–1770 (2003). Ishii H, Vecchione A, Fong LY, Zanesi N, Trapasso F, et al: Cancer prevention and therapy in a preclinical mouse model: impact of FHIT viruses. Curr Gene Ther 4:53–63 (2004). Matsuyama A, Shiraishi T, Trapasso F, Kuroki T, Alder H, et al: Fragile site orthologs FHIT/ FRA3B and Fhit/Fra14A2: evolutionarily conserved but highly recombinogenic. Proc Natl Acad Sci USA 100:14988–14993 (2003). Ohta M, Inoue H, Cotticelli MG, Kastury K, Baffa R, et al: The FHIT gene, spanning the chromosome 3p14.2 fragile site and renal carcinomaassociated t(3;8) breakpoint, is abnormal in digestive tract cancers. Cell 84: 587–597 (1996). Ottey M, Han SY, Druck T, Barnoski BL, McCorkell KA, et al: Fhit-deficient normal and cancer cells are mitomycin C and UVC resistant. Br J Cancer 91:1669–1677 (2004).
Paradee W, Wilke CM, Wang L, Shridhar R, Mullins CM, et al: A 350-kb cosmid contig in 3p14.2 that crosses the t(3; 8) hereditary renal cell carcinoma translocation breakpoint and 17 aphidicolin-induced FRA3B breakpoints. Genomics 35:87–93 (1996). Pekarsky Y, Druck T, Cotticelli MG, Ohta M, Shou J, et al: The murine Fhit locus: isolation, characterization, and expression in normal and tumor cells. Cancer Res 58: 3401–3408 (1998). Pena RN, Webster J, Kwan S, Korbel J, Whitelaw BA: Transgene methylation in mice reflects copy number but not expression level. Mol Biotech 26:215–219 (2004). Pinkel D, Straume T, Gray JW: Cytogenetic analysis using quantitative, high-sensitivity, fluorescence hybridization. Proc Natl Acad Sci USA 83:2934–2938 (1986). Seabright M: A rapid banding technique for human chromosomes. Lancet 2: 971–972 (1971). Shi YP, Huang TT, Carlson EJ, Epstein CJ: The mapping of transgenes by fluorescence in situ hybridization on G-banded mouse chromosomes. Mamm Genome 5: 337–341 (1994). Yunis JJ, Sorenga AL: Constitutive fragile sites and cancer. Science 226:1199–1204 (1994). Zanesi N, Fidanza V, Fong LYY, Mancini R, Druck T, et al: The tumor spectrum in Fhit-deficient mice. Proc Natl Acad Sci USA 98: 10250–10255 (2001). Zochbauer-Muller S, Fong KM, Maitra A, Lam S, Goradts J, et al: 5CpG island methylation of the FHIT gene is correlated with loss of gene expression in lung and breast cancer. Cancer Res 61:3581–3585 (2001).
Cytogenet Genome Res 118:196–203 (2007)
203
Mycoplasmas are among the few prokaryotes that can grow in close relationship with mammalian cells, often without any apparent pathology for extended periods of time (Maniloff et al., 1992). Previous studies have revealed that chronic and persistent infection with apparently lowvirulence mycoplasmas can gradually but significantly affect many important biological characteristics of mammalian cells and even lead to malignant transformation of murine cells (Tsai et al., 1995; Feng et al., 1999). Previously, we have reported that exogenous p37 protein enhances the invasiveness of two prostate cancer and two melanoma cell lines in a dose-dependent manner (Ketcham et al., 2005). The p37 protein from Mycoplasma hyorhinis was inadvertently found during attempts to identify human cell antigens that elicit tumor-specific antibodies. Sera from patients who were in a state of tumor regression had measurable antibody titers against several presumed tumor antigens, including a 38 kDa protein (Fareed et al., 1988) later designated as p37. When the p37 protein was isolated and sequenced, it turned out to be of M. hyorhinis origin (Gilson et al., 1988). Independently, two research groups showed that p37 on the surface of FS9 mouse fibrosarcoma cells was associated with a highly invasive phenotype. Antibodies against p37 inhibited the invasive potential of infected FS9 cells in an in vitro assay (Dudler et al., 1988) and reduced the lung metastasis of colon cancer in nude mouse models (Schmidhauser et al., 1990). The strongest link between M. hyorhinis and human cancer was reported recently by Huang et al. (2001) who used a monoclonal antibody that recognizes the unique M. hyorhinis-specific protein p37 to detect mycoplasma infection in over 600 cancer tissues from a variety of organs. The study indicated that up to 56% of gastric cancer and 55% of colon cancer biopsies were positive for M. hyorhinis (Huang et al., 2001). In this study, we further investigated the effects of p37 treatment on two prostate cell lines, PC-3 and DU145, and identified factors that may play a role in the biological potential by employing a genome-wide expression profiling approach. Oligonucleotide microarray technology (Affymetrix U133Plus2.0 GeneChips) was used to create a gene expression database of over 54,000 transcripts containing profiles of the prostate cell lines pre- and post-treatment with p37 protein. Our studies show significant changes in the transcription of numerous genes and the data implicate specific signaling pathways in p37-induced cellular changes. Methods Cell lines and culture Human prostate cancer cell lines, PC-3 and DU145, were obtained from the American Type Cell Culture (Manassas, VA, USA). Cells were maintained in RPMI medium supplemented with 5% fetal calf serum, 4.5 g/l glucose, 4 mM L-glutamine, 100 units/ml penicillin and 100 g/ml streptomycin. All cells were incubated at 37 ° C in a humidified atmosphere of 5% CO2 in air. All culture media were purchased from Invitrogen (Carlsbad, CA, USA).
Expression and purification of p37 Production of p37 followed our previously published method (Ketcham et al., 2005). Briefly, plasmid pMH38–113, which contained the entire coding sequence (the leader sequence) for p37 whereby all of the TGA codons (mycoplasmal codons for Trp) were changed to TGG to optimize its expression in E. coli, was ligated into a pET vector and transformed into BL21DE3PlysS for expression. Freshly transformed cells were grown to an OD of 0.7–1 and induced with 1 mM isopropyl -thiogalactoside (IPTG) for 2.5 h. Cells were broken by vortexing in 1/10 the original volume of 20 mM phosphate buffer (pH 7.8) followed by a sonication for three 15-second cycles, and clarified by centrifugation. Ion exchange chromatography (High Q resin, BioRad, Hercules, CA, USA) was then performed to purify the p37 from the rest of the bacterial components. The eluted sample was then concentrated using a Centriprep 10 spin column (Millipore, Bedford MA, USA). The yield of p37 protein was estimated to be ⬃30 mg of purified p37 from 1 liter of E. coli culture or 10% w/w. Following purification, p37 protein was loaded onto an Endotoxin Removal Gel (detoxi-Gel, Pierce) for removal of endotoxins (lipopolysaccharides or LPS) from the purified preps, since even low levels of endotoxins can be toxic to cells. Proliferation assay Briefly, prostate cancer cell lines, PC-3 and DU145, were seeded in 96-well plates at a density of 2.5!103 cells per well and treated with 25 g/ml p37 or vehicle only (Ketcham et al., 2005). After 1 and 4 days, 100 l of 1 mg/ml MTT (Sigma-Aldrich, St. Louis, MO, USA) solution was added to appropriate plates and allowed to incubate at 37 ° C for 2.5 h. The reaction was stopped with lysis buffer (200 mg/ml SDS, 50% N,N-Dimethylformamide, pH 4) at room temperature for 1 h, and the optical density was read on a microplate autoreader (Bio-Tek Instruments, Winooski, VT, USA) at 560 nm. Absorbance values were normalized to the values obtained for the vehicle only-treated cells to determine the survival percentage. Each assay was performed in triplicate, and an intra-experiment average was calculated. Cellular viability was confirmed with Trypan Blue exclusion test. In vitro cell invasion and migration assays Tumor cell migration and invasion was assessed using the membrane invasion system purchased from the Hendrix Laboratory (University of Iowa, Iowa City, IA, USA). Polycarbonate membranes of 8 m pore size were purchased from Poretics (Livermore, CA, USA). Growth factor reduced Matrigel was purchased from Becton Dickinson (Franklin Lakes, NJ, USA). Polycarbonate membranes were coated with 4 mg/ml growth factor reduced Matrigel as described (Hendrix et al., 1987) for invasion assays; control inserts (migration only) contained no coating. Cells were added to each insert at a density of 100,000 cells/ml/well in DFCI medium (Band and Sager, 1989). The lower chamber contained DFCI medium with 10% FBS as chemoattractant. In half of the wells, p37 at a final concentration of 25 g/ml was added. After incubation in a humidified incubator with 5% CO2 at 37 ° C for 24 h, the coating and the cells on the top of the polycarbonate membrane were removed. The cells attached to the bottom of the membrane were fixed in 100% methanol, stained with LeukoStat Staining kit from Fisher, and counted using a Carl Zeiss microscope. Each assay was performed in triplicate. Comparisons between group means were assessed with the paired Student t-test. Hematoxylin and eosin (H& E) staining and image analysis of nuclear area Both prostate cancer cell lines, PC-3 and DU145, were seeded in 100mm3 dishes at a density of 5 ! 105 cells per well and treated with or without 25 g/ml of p37. After 24 h, cells were trypsinized, washed with 1! PBS, immediately embedded in OCT (Tissue-Tek쏐, Sakura, CA, USA) and snap-frozen in liquid nitrogen. The OCT sections (10 m) were cut by using a cryostat and stained with H&E. Briefly, as previously described, the mean nuclear area was determined in each of the above conditions for PC-3 and DU145 cells (Iczkowski et al., 2005). At 400! magnification, the areas of 120 nuclei (4 nuclei from 30 separate high power fields) were measured by digital tracing utilizing Zeiss AxioCam HRc image analysis system (Carl Zeiss, Oberkochen, Germany).
Cytogenet Genome Res 118:204–213 (2007)
205
Sample collection and preparation of labeled RNA Duplicate cultures of prostate cancer cell lines, PC-3 and DU145, were seeded in 100-mm3 dishes at a density of 5 ! 105 cells per well and treated with or without 25 g/ml of p37 for 4 h. Total RNA was extracted from each of the cell line samples and prepared for hybridization according to the GeneChip Expression Analysis Technical Manual (Affymetrix, Santa Clara, CA, USA). Briefly, samples were homogenized in lysis buffer and the RNA extracted (RNeasy Mini kit; Qiagen, Inc., Valencia, CA, USA). The quality of each RNA sample was assessed after running 200-ng aliquots through a microchannel RNA analysis chip on a bioanalyzer (RNA 6000 Nano Chip; Agilent Technologies, Palo Alto, CA, USA) and evaluating the relative amounts of 28S and 18S ribosomal peaks. A 5-g aliquot of RNA was used as a template for complementary DNA (cDNA) synthesis (Superscript Choice System kit; Invitrogen Life Technologies, Gaithersburg, MD, USA). Firststrand synthesis was primed with a T7-(dT)24 oligonucleotide primer containing a T7 RNA polymerase promoter sequence on the 5ⴕ end (Genset Oligos, La Jolla, CA, USA). Second-strand products were cleaned (GeneChip Sample Cleanup Module; Affymetrix) and used as a template for in vitro transcription (IVT) with biotin-labeled nucleotides (Bioarray High Yield RNA Transcript Labeling Kit; Enzo Diagnostics, Farmingdale, NY, USA). The copy RNA (cRNA) product was cleaned with the cleanup module and a 20-g aliquot was heated at 94 ° C for 35 min in fragmentation buffer provided with the cleanup module (Affymetrix). Gene expression microarray hybridization Fifteen micrograms of adjusted cRNA from each sample was hybridized for 16 h at 45 ° C to an Affymetrix U133 Plus 2.0 GeneChip array, which offers coverage of 47,000 transcripts. After hybridization, each array was stained with a streptavidin-phycoerythrin conjugate (Molecular Probes, Eugene, OR, USA), washed and visualized with a microarray scanner (Genearray Scanner; Agilent Technologies, Santa Clara, CA, USA). Images were inspected visually for hybridization artifacts. In addition, quality assessment metrics were generated for each scanned image and evaluated based on empiric data from previous hybridizations and the signal intensity of internal standards present in the hybridization cocktail. Samples that did not pass quality assessment were eliminated from further analysis.
seemed heavily populated. In this way we were able to evaluate changes in gene expression at a more global level by identifying pathways and processes that may be interconnected. All microarray data obtained within this project is available at http://genomics.biotech.ufl.edu/people/goodison/.
Results and discussion
p37 treatment increases cellular proliferation in DU145 prostate cancer cells but not in PC-3 prostate cancer cells To determine the effects of p37 on cell proliferation and viability, PC-3 and DU145 prostate cancer cell lines were treated with or without 25 g/ml of purified p37 protein for 96 h. Proliferation and viability were determined at 24 h and 96 h. At 24 h, PC-3 cells treated with p37 were not noted to have a change in proliferation or viability when compared with untreated PC-3 cells. However, DU145 cells treated with p37 were observed to have a modest, yet significant increase in proliferation at 24 h post-incubation (data not shown). No change in DU145-cell viability was evident when compared to untreated cells.
Generation of expression values Microarray Suite, version 5 (Affymetrix), was used to generate *.cel files, and a computer program (Probe Profiler, ver. 1.3.11; Corimbia Inc., Berkeley, CA, USA) developed specifically for the GeneChip system (Affymetrix) was used to convert intensity data into quantitative estimates of gene expression for each probe set. A probability statistic was generated for each probe set (gene). The probability is associated with the null hypothesis that the expression level of the probe set is equal to zero (background). Genes not significantly expressed above background in at least two samples (P 1 0.05) were considered absent. Absent genes were removed from the data set and not included in further analyses.
In vitro invasion and migration assay We have previously published that in a dose-dependent manner, p37 increases tumor cell invasion through Matrigel in both PC-3 and DU145 cancer cells. To investigate whether this phenomenon was due specifically to invasion, i.e. the proteolytic breakdown of the extracellular matrix, we evaluated the invasive potential of PC-3 and DU145 of prostate cancer cell lines using an established invasion assay (Ketcham et al., 2005), with and without Matrigel present and in the presence or absence of exogenous p37 protein. As shown in Fig. 1, a statistically significant increase in cell numbers traversing the Matrigel-coated invasion assay membranes was evident for both cell lines treated with p37 protein (P = 0.03 and 0.05, respectively). However, under the same conditions, but without Matrigel, a significant increase in transferred PC-3 and DU145 cell numbers was still evident (P = 0.01 and !0.01, respectively). This data suggests that p37 does not increase the intrinsic ability of the prostate tumor cells tested to degrade extracellular matrix, but p37 does increase cellular migration rates.
Gene expression data analysis Gene expression levels were subjected to a two-way analysis of variance (ANOVA) for two treatments (+/– p37) and two cell lines (PC-3 and DU145) using AnalyzeIt Tools (http://genomics3.biotech.ufl.edu/ AnalyzeIt/AnalyzeIt.html), a custom software program developed by the Interdisciplinary Center for Biotechnology Research (ICBR; University of Florida) for the analysis of microarray data. In this software, the statistical package, R, serves as the back end for ANOVA. Signal values for genes significantly (P ! 0.01) affected by treatment (+/– p37), significantly (P ^ 0.001) affected by cell line, or for which there was a significant interaction (P ! 0.01) were normalized by performing a Ztransformation, thereby generating a distribution with a mean of 0 and SD of 1 for each gene. Hierarchical and K-means were performed on normalized values using Cluster and TreeView (Eisen et al., 1998). Genes identified as being affected by treatment, cell line, or having a significant interaction were placed in KEGG and GenMapp pathways using AnalyzeIt Tools. Of particular interest were pathways that
Analysis of cell nuclear area Light microscopy revealed several distinct changes in prostate cancer cells treated with p37. First, p37 treated cells (both PC-3 and DU145) were noted to have more vacuolation within the cytoplasm compared to non-p37 treated cells. Cytoplasmic vacuolation is associated with increased cellular activity. Secondly, p37 treated cells had evident macronucleoli and double nucleoli. Nucleoli increase in number and enlarge when the cell is stimulated to produce copious protein. This is a sign that the cell has been stimulated or is actively involved in protein synthesis. Lastly, the nuclei of p37 treated cells are enlarged compared to non-p37 treated cells. To objectively illustrate the increase in nuclear area in p37 treated cells we utilized digital image analysis. The mean
206
Cytogenet Genome Res 118:204–213 (2007)
P > 0.05 P > 0.05
Cell number
1,600
P = 0.01
P = 0.03
1,200 800 400 0 Untreated PC-3
1,600
p37 treated PC-3 with matrigel
P > 0.05 P < 0.01
Cell number
1,200
P = 0.05
800 400 0 Untreated DU145
Untreated PC-3
p37 treated DU145
p37 treated PC-3 P < 0.0001
400 350 300 250 200 150 100 50 0 Untreated PC-3
nuclear area of PC-3 cells treated with 25 g/ml of p37 increased from 143 to 261 m2 (P ! 0.0001). Similarly, the mean nuclear area of DU145 cells treated with 25 g/ml of p37 increased from 112 to 203 m2 (P ! 0.0001) (Fig. 2). An increase in nuclear area or change in nuclear to cytoplasmic ratio is a prominent feature of anaplasia where cellular organization is disrupted in favor of uncontrolled proliferation. Differential gene expression in prostate cell lines PC-3 and DU145 The gene expression profile of PC-3 and DU145 prostate cancer cell lines was evaluated using microarray technol-
p37 treated PC-3
Untreated DU145 with matrigel
Untreated DU145 Mean nuclear area (μm2)
Fig. 2. Light microscopy noted distinct changes in mean nuclear area in prostate cancer cells treated with p37. PC-3 and DU145 cells were seeded in 100-mm2 dishes and treated with 25 g/ml p37 protein for 24 h. Cells were fixed and stained with hematoxylin and eosin. Light microscopy revealed increased cytoplasmic vacuolation and macronucleoli/double nucleoli in p37 treated cells. Utilizing digital image analysis, we illustrated that the mean nuclear area of PC-3 cells treated with p37 increased from 143 to 261 m2 (P ! 0.0001). Similarly the mean nuclear area in DU145 cells treated with 25 g/ml of p37 increased from 112 to 203 m2 (P ! 0.0001).
Untreated PC-3 with matrigel P > 0.05
Mean nuclear area (μm2)
Fig. 1. Influence of p37 on the migration and invasion of prostate cell lines PC-3 and DU145. Cells in serum-free medium supplemented with or without p37 (25 g/ml) were seeded onto 8-m pore membranes, with or without a pre-coating of Matrigel, in a modified Boyden chamber assay. The rate of cell migration was assessed by induction of movement toward serum through an uncoated membrane. Cell invasiveness was assessed by the movement toward serum through Matrigel-coated membranes. Cells that had migrated through the membranes after 24 h were detached by trypsin and counted. Each value represents the mean of the triplicate results, and the error bars represent the standard error of the mean.
p37 treated PC-3
p37 treated DU145 with matrigel
p37 treated DU145 P < 0.0001
400 350 300 250 200 150 100 50 0
Untreated DU145
p37 treated DU145
ogy that can now query the entire transcriptome. RNA extracted from duplicate cultures of both cell lines, and with or without p37 treatment, was hybridized to a total of eight Affymetrix U133 Plus 2.0 arrays. Of the 54,613 targets (covering 47,000 transcripts) present on the Affymetrix U133 Plus 2.0 array, 34,707 were detected above background on at least one of the arrays. Prior to identifying the genes affected by p37 treatment it was of interest to compare the baseline expression of the two prostate cell lines used in this study. The PC-3 and DU145 cell lines are both androgen independent, and although isolated from different organs (PC-3 from bone and DU145 from brain) both
Cytogenet Genome Res 118:204–213 (2007)
207
DU145 1 DU145 2 DU145 p37 1 DU145 p37 2 PC-3 2 PC-3 1 PC-3 p37 1 PC-3 p37 2 Cluster
1
2
3
4
5
6
7
8
9
10
Fig. 3. K-means cluster of normalized signal values for genes with a significant (P ! 0.001) treatment effect (incubation with 20 g/ml p37 protein). Each row represents a sample and each column a gene. The scale represents standard deviations from the mean after a Z-transfor-
mation of signal values of a gene across all samples. Red represents a higher level of gene expression and green a lower level, relative to the mean across all samples for each gene. Individual clusters are indicated by numbers on the right.
originate from metastases (Stone et al., 1978; Kaighn et al., 1979). To our knowledge, data from the complete genome coverage chip has not been available to date in these cell lines. As expected, there were considerable differences between the gene expression profiles derived from the two cell lines. Of the over 54,000 targets present on the Affymetrix U133 2.0 Plus chips employed, there were some 6,000 genes that were differentially expressed at a significance level of P ! 0.001. We chose to focus on those genes that were present in one cell line and absent in the other, a total of 615 genes. The complete datasets and GenMapp pathway files are available as supplemental data at http:// genomics.biotech.ufl.edu/people/goodison/. With respect to the design of this study, we were particularly interested in differences in cell surface receptors, secreted molecules and signaling pathways that are involved in responses to extracellular stimuli. Many receptors were differentially expressed between the cell lines. Genes expressed only in PC-3 cells included prostaglandin receptors, multiple G-protein coupled receptors, several interleukin receptors and PPARA. Unique to DU145 cells was the expression of retinoic acid receptors and multiple TNF receptor superfamily members. Studies have convincingly demonstrated that matrix metalloproteinases (MMPs) are directly involved in a wide range of diseases, including cancer progression, through aberrant tissue remodeling (Mott and Werb, 2004). Current evidence suggests that the MMP family plays a pivotal role in the pathological proteolysis of the extracellular matrix and subsequent tissue remodeling. In addition, MMPs modulate the activity of growth factors and cytokines or their receptors and process certain adhesion and signaling receptor targets (Flannery, 2006). In neoplasms, the elevated expression of MMPs is associated with the invasion of malignant cells, the metastasis and neovascularization of tumors (Chetty et al., 2006). Remarkably, seven MMP family members (MMP2, MMP10, MMP13, MMP14, MMP16, MMP23 and MMP28) expressed in PC-3 cells were entirely absent from DU145 cells. Given their role in tissue remodeling, this MMP-profile likely contributes to the less aggressive and metastatic nature of DU145 cells relative to PC-3 as demonstrated by their capacity to invade extracellular matrix (Pulukuri et al., 2005).
The dysregulated behavior of epithelial cells in prostate cancer is associated with an increased capacity for autocrine expression of epidermal growth factor (EGF) and related family members such as transforming growth factor (TGF). The expression of these factors may circumvent a paracrine dependence on stromal cell-derived EGF (Djakiew, 2000). Increased expression of EGF and TGF proteins has been linked to the development of benign prostatic hyperplasia and prostate cancer (Harper et al., 1993; Yang et al., 1993), and it has also been observed that the EGF receptor (EGFR) is coexpressed in concert with the upregulated EGF and TGF proteins (Cohen et al., 1994). DU145 cells do express EGFR, but unlike PC-3 cells they do not express EGF. In addition to being mitogenic (Jones et al., 1997), EGF has also been shown to stimulate the invasiveness of prostate cancer cells (Jarrard et al., 1994), so this difference may also contribute to the overall lower biological potential of DU145 cells relative to PC-3.
208
Cytogenet Genome Res 118:204–213 (2007)
Gene expression is affected by p37 treatment The expression of 1,979 transcripts was significantly affected (P ! 0.001) by treatment with p37 protein (25 g/ml for 4 h) in at least one of the prostate cell lines. The 1,979 probe sets identified as being significantly affected by treatment were divided into 1,594 genes that had a functional annotation, 18 that had a strong similarity to an annotated gene, 112 that are designated as coding hypothetical proteins and 49 that had no annotation. In line with our cell behavior analyses described earlier, many of the genes induced by p37 treatment are associated with DNA synthesis enzymes, protein translation factors and factors involved in energy production (Table 1); all functions that are linked to increased cellular proliferation. K-means clustering was performed on the 1,979 genes in order to group together genes with expression patterns that were similar in response to treatment. Based on trend and magnitude of effect, we identified ten different K-means clusters (Fig. 3) but these fell into one of four more basic expression patterns. The four basic expression patterns were distinguished by their relative change in expression after treatment and are characterized by genes that increased or decreased similarly in both cell lines or showed differential change between cell lines. The first group, which consisted of clusters 1, 4 and 8 and contained 28.6% (566) of the genes was characterized by an
Table 1. Genes regulated by p37 in prostate tumor cells DU145 and PC-3
Functional groupa
Affymetrix ID
Krebs-TCA cycle isocitrate dehydrogenase 3 (NAD+) beta pyruvate dehydrogenase (lipoamide) alpha 1 pyruvate dehydrogenase (lipoamide) alpha 1 succinate-CoA ligase, GDP-forming, alpha subunit WD repeat domain 50 Wnt signaling protein kinase C, alpha frizzled homolog 7 frizzled homolog 2 Protein phosphatase 2, regulatory subunit B56 Mapk cascade mitogen-activated protein kinase kinase 3 v-raf murine sarcoma viral oncogene homolog v-raf-1 murine leukemia viral oncogene homolog 1 Phosphatidylinositol signaling system diacylglycerol kinase, zeta diacylglycerol kinase, epsilon CDC-like kinase 4 inositol 1,4,5-trisphosphate 3-kinase B phosphoinositide-3-kinase, alpha peptide phosphoinositide-3-kinase, beta polypeptide protein kinase C, alpha v-raf-1 murine leukemia viral oncogene homolog 1 serine/threonine kinase casein kinase 2, alpha 1 polypeptide inositol polyphosphate-4-phosphatase, type II Inositol phosphate metabolism inositol 1,4,5-trisphosphate 3-kinase B phosphoinositide-3-kinase, alpha polypeptide phosphoinositide-3-kinase, beta polypeptide inositol polyphosphate-4-phosphatase, type II Nuclear receptors peroxisome proliferative activated receptor, gam nuclear receptor subfamily 2, group F, member 1 Matrix metalloproteinases v-raf murine sarcoma matrix metalloproteinase 13 matrix metalloproteinase 9 basigin (OK blood group) Prostaglandin and leukotriene metabolism peroxiredoxin 6 prostaglandin-endoperoxide synthase 2 thromboxane A synthase 1 Interleukins interleukin 6 interleukin 8
Direction of regulationb DU145
PC-3
210418_s_at 1555864_s_at 200980_s_at 217874_at 203721_s_at
nr UP nr nr UP
UP UP UP UP UP
213093_at 203706_s_at 210220_at 228070_at
DOWN DOWN DOWN DOWN
DOWN DOWN DOWN DOWN
215499_at 230652_at 244373_at
UP UP UP
UP UP UP
207556_s_at 238694_at 210346_s_at 232499_at 204369_at 204484_at 213093_at 244373_at 204292_x_at 212075_s_at 223878_at
nr nr DOWN nr DOWN DOWN DOWN UP DOWN UP DOWN
UP UP DOWN DOWN DOWN DOWN DOWN UP DOWN UP UP
232499_at 204369_at 204484_at 223878_at
nr DOWN DOWN DOWN
DOWN DOWN DOWN UP
208510_s_at 209506_s_at
nr nr
DOWN DOWN
230652_at 205959_at 203936_s_at 208677_s_at
UP nr nr nr
UP DOWN DOWN UP
200845_s_at 204748_at 208130_s_at
nr UP UP
UP DOWN DOWN
205207_at 211506_s_at
nr nr
UP UP
a
Functional groups of genes included in Fig. 4 were assigned using gene-association files from the GO Consortium and by GenMapp and Kegg pathway analysis. Full gene lists and pathway information are available in supplemental data. b nr: not regulated by p37.
increase in expression in both cell lines after p37 treatment. The second group consisted of clusters 3, 5, 6 and 10 and contained 21.3% (423) of the genes. Genes in this group had an expression pattern opposite to that of the first group, characterized by a decrease in expression in both cell lines
after p37 treatment. The third group consisted of genes that were increased in PC-3 cells by p37 treatment, but did not change in DU145 cells. This group was represented by cluster 2 and contained 16.5% (326) of the 1,979 differentially expressed genes. In the fourth group (clusters 7 and 9),
Cytogenet Genome Res 118:204–213 (2007)
209
DNA damage checkpoint
Growth Growth factor factor withdrawal
ARF
SMC1L1 Condensin EP300
TGFB1
MDM2
e
+p
SCF
SMAD3 SMAD4
MAPK signaling pathway
Apoptosis
SKP2
CDKN2A
+u
CDKN1B
TP53
RB1
e
p16,15,18,19 p27,57
BUB1 BUB3 MPEG1
ATR ATM
PRKDC
GSK3B
e
+u p21
e
+p
GADD45A
CDKN1A
+p
e YWHAG
+p
+p CDC25A
+p CCND3
SCF SKP2
Both p37 up DU145 p37 up PC-3 p37 up Both p37 down DU145 p37 down PC-3 p37 down No criteria met Not found
CCND2 CDK4 CDK6
+u
ABL1 +p RB1
CCNE2 CCNE1 CDK2
CCNA1 CCNA2 CDK2
CCNH CDK2
CDC6
+p RBL1
ORC
HDAC
+p
CDC45L MCM
CCNA1 CCNA2 CDC2
+p+p RB1
+p +p
PTTG3 PTTG2 Securin PTTG1
BUB1B MAD1L1 MAD2L1 MAD2L2
CHEK1 CHEK2
PCNA
R-point (start)
ESPL1 Separin
CCNB1 CCNB2 CCNB3 CDC2
+p
WEE1 PKMYT1 +p
+u APC/C CDC20 14-3-3
–p +p PLK1 +u
DNA
APC/C CDH1
–p CDC14A CDC14B
+p E2F TFDP1
Ubiquitin mediated proteolysis
CDC25C CDC25B
CDC7 ASK TBC1D8
MEN
DNA biosynthesis
DNA S-phase proteins
Fig. 4. Expression changes in genes of the cell cycle pathway in PC-3 and DU145 cells exposed to p37 treatment, as measured by oligonucleotide microarray. Changes in gene expression were derived from Fig. 3. Using gene-association files from the KEGG Consortium, genes significantly regulated by p37 are depicted in a functional MAPP created by MAPPFinder (http://www.genmapp.org/). See color legend on figure for directional and cell-specific information.
33.6% (664) of the genes were decreased only in PC-3 cells by p37 treatment, while the expression levels in DU145 cells remained unchanged from the levels observed in untreated cells. Only a few genes were revealed to change in DU145 cells alone in response to p37 treatment. To further evaluate the microarray data, GenMAPP and MAPPFinder were used to organize gene expression data into MAPPs (microarray pathway profiles) that represent specific biological pathways and functionally grouped genes. MAPPFinder analysis identified several biological processes influenced by p37 (GenMapp and Kegg pathways and individual MAPPs are available as supplemental data at http://genomics.biotech.ufl.edu/people/goodison/). Consistent with our initial assessment, MAPPFinder confirmed that many genes in proliferative and metabolic pathways were altered by p37 treatment. These included the induction of several enzymes in the glycolytic and Krebs cycle pathways and in purine and pyrimidine metabolism, the latter of which can provide energy and substrates for increased DNA synthesis. Deregulation of the normally tightly controlled cell cycle is an essential component in the evolution of cancer. The cyclin D1 cyclin-dependent kinase inhibitors CDKN1B, CDC25B and CDC25C were down-regulated in both cell lines by p37 treatment (Fig. 4), and this would likely cause a shift in cyclin activities thus moving cells through the cell
210
Cytogenet Genome Res 118:204–213 (2007)
cycle and increasing proliferation. The observed dual increase of cyclins E2 and H, both known to be increased in prostate cancer (Amanatullah et al., 2000), in both cell lines would also support increased proliferation. However, PC-3 cells were not overall more proliferative when treated with p37 protein. The only cell cycle factor for which upregulation was restricted to DU145 cells was cyclin-dependent kinase 2 (CDK2 or CDC2) (Fig. 4). Given that activation of the cyclin B1/CDC2 kinase complex is a rate-limiting, regulatory step for cells to enter the mitotic phase of the cell cycle (Blagosklonny and Pardee, 2002) the differential induction of CDC2 expression may explain why a p37-induced proliferative response was restricted to DU145 cells. Signal transduction pathways Overall, the observed p37-induced changes in gene expression were remarkably similar in both prostate tumor cell lines. Grouping of genes into several functional categories using gene ontology databases revealed that only the signal transduction category was markedly different between the two cell lines (17.2% and 10.7% in PC-3 and DU145, respectively). Signal transduction in cancer cells frequently involves the activation of receptor tyrosine kinases that in turn trigger multiple cytoplasmic kinases (Massie and Mills, 2006). Induction of such cellular signaling pathways can operate independently or in parallel to
promote cancer development. Three major signaling pathways that have been identified to be important in cancer include the phosphatidylinositol 3-kinase (PI3K)/AKT cascade (Larue and Bellacosa, 2005), the protein kinase C (PKC) family (Gschwendt et al., 1998) and the mitogenactivated protein kinase (MAPK)/Ras signaling cascades (Fang and Richardson, 2005). Each of these pathways was perturbed to some extent in PC-3 and DU145 prostate cells by treatment with p37 protein (Table 1). The PI3K/AKT pathway has a pivotal role in cancer cell metabolism (Vivanco and Sawyers, 2002). Phosphorylated receptor tyrosine kinases interact with PI3K to generate phosphatidylinositol-3,4,5-trisphosphate (PIP3) from PIP2. PIP3 serves as a ligand to recruit AKT to the plasma membrane where AKT becomes phosphorylated and activated. Activated ATK, itself a kinase, promotes cell proliferation, growth, survival and other processes involved in cancer development by phosphorylating various intracellular proteins. The expression of eleven genes encoding different components of the phosphatidylinositol signaling pathway was altered by p37 treatment in at least one of the prostate cell lines. Subunits of the pivotal PI3K enzyme were actually down-regulated in both cell lines, but the opposing phosphatase, inositol polyphosphate-4-phosphatase (IP4Pase), which dephosphorylates PIP3 back to PIP2, was expressed differentially in the two cell lines. In PC-3 cells, the PI3K was decreased but the phosphatase was increased, so the PIP3 levels would more likely shift downwards. This may, in part, explain why p37 treatment did not increase proliferation in PC-3 cells. It is also possible that PC-3 cells are maximally stimulated with regard to proliferation rate and so external stimuli may have no additive effect. In DU145 cells, both PI3K and IP4Pase were down-regulated, so the overall balance of PIP3 could shift upwards. Components of other receptor tyrosine kinase pathways were also perturbed by p37 treatment (Table 1). The expression of several members of the MAPK pathway and of the Ras superfamily of proteins was altered significantly. The MAPK pathway, also known as the extracellular signal-regulated kinase (ERK) pathway, is activated in a variety of cell types by diverse extracellular stimuli (Fang and Richardson, 2005). Activation of the ERK pathway involves the guanosine triphosphate (GTP) loading of Ras at the plasma membrane and the sequential activation of a series of protein kinases. Ras recruits MAPK-kinase kinases which in turn activate MAPK-kinases and then MAPK by serine phosphorylation. MAPK then catalyzes the phosphorylation of ERK1 and ERK2, which in turn phosphorylate various downstream substrates involved in a multitude of cellular responses such as cell proliferation, cell survival and cell motility (Troppmair et al., 1994). MAPK-associated protein 1, MAPKKK-14 and MAPK kinase-3 were increased in both cell lines. Conversely, no MAPK phosphatases were upregulated by p37. The ras genes give rise to a family of related GTP-binding proteins that exhibit potent transforming potential by disturbing a multitude of cellular processes. Aberrant activation of the Ras proteins, either by mutation of the gene or overexpression of the protein, can deregulate
growth-factor receptor signaling or impinge on virtually all aspects of a malignant phenotype, including proliferation, invasion and metastasis (Campbell and Der, 2004). In humans, the Ras superfamily comprises over 100 related small GTP-binding proteins and one common classification places them into six subfamilies: Ras, Rho, Arf, Rab, Ran and Rad. Three members of the Ras subfamily were upregulated in both cell lines by p37, but more impressively, the expression of ten members of the Rab family was altered by p37 treatment. Dysregulation of RAB gene expression may be a generalized component of human tumors and many RAB proteins exhibit elevated expression at the RNA level (Cheng et al., 2005). RNA microarray analyses in ovarian cancer show that 45–50% of known RAB or RAB-associated genes display increases in mRNA expression. In this study, regulation of the RAB proteins by p37 was consistent in that eight of the ten genes were regulated in the same way in both prostate cell lines (Table 1). Four Rab genes were upregulated (RAB, RAB18, RAB 32 and RAB35), and four downregulated (RAB6A, RAB7, RAB8B and RAB14). RAB22a and RAB27B were regulated oppositely, both being upregulated in DU145. With so many positive and negative signaling factors being overexpressed in response to p37 it is difficult to be certain of the overall activity of the pathway, but the analyses performed in this study clearly suggest a shift to a more stimulated, aggressive phenotype indicative of the increased activity of these pathways. While we have shown that exogenous p37 binds to proteins on the cell membrane of the prostate cells used in this study (Ketcham et al., 2005), this binding interaction may exert cellular effects by a number of possible mechanisms. The p37 protein may operate through molecular mimicry, i.e. conformationally. p37 may by chance mimic human proteins that bind specific receptors and stimulate the cells through an intrinsic mechanism in which there is no recognition of any pathogen. The epithelial cells then respond to the activating stimulus as they would in normal, healthy processes such as wound healing. Conversely, the human cells might actually recognize a mycoplasmal insult through detection of mycoplasmal surface proteins and p37 may be part of such a recognition system. In this case, the cell may respond defensively in order to protect itself or to recruit a local immune response. A combination of these mechanisms may also exist. Given their limited biosynthetic capabilities, most mycoplasmas are obligate parasites so they need to adhere to, fuse with or invade the host cells. Perhaps through parallel evolutionary mechanisms, mycoplasmas have learned to stimulate the host cells in a way that facilitates mycoplasmal entry into the cell and/or co-existence with it. An example of this is the increased invasive capacity of M. fermentans arising from the potential of this organism to bind host plasminogen and activate it via the urokinase-type plasminogen activator (uPA) to plasmin (Yavlovich et al., 2001). Plasmin, a protease with broad substrate specificity, may alter M. fermentans cell surface proteins and thereby promotes its internalization. Interestingly, the uPA receptor was one of the few receptors to be upregulated in both cell lines by p37 in this study.
Cytogenet Genome Res 118:204–213 (2007)
211
It is increasingly recognized that for many bacteria and mycoplasmas, induction of cytokines is a major virulence mechanism. The first reports on the cytokine-inducing ability of mycoplasmal proteins showed that a lipoprotein from M. fermentans or M. arginini is capable of stimulating the release of proinflammatory cytokines such as tumor necrosis factor (TNF), interleukins IL1 and IL6 and chemokines such as IL8 (Razin et al., 1998). In this study, p37 was able to induce the expression of IL6 and IL8, but only in PC3 cells. The TNF gene family was one of the groups most affected by p37 treatment in PC-3 cells. Of eleven TNF-associated genes affected overall, seven were down-regulated in PC-3 cells but were unaffected in DU145 cells. Four other genes, including two TNF receptors, were regulated similarly in both cell lines. The TNF pattern fits the TGF pattern of expression in that TGFA and TGFB2 were both upregulated by p37, but only in PC-3 cells. TGF and TNF family expression often have a reciprocal relationship as they have largely opposing effects on cellular activity. Finally, it has been shown that invasion of HeLa cells by M. penetrans is associated with receptor mediated influence of the phosphatidylinositol signaling pathway and changes in host cell lipid turnover occur as a result of M. penetrans adherence to Molt-3 lymphocytes (Salman et al., 1998). One of the major effects of p37 on the two prostate cell lines was the perturbation of phosphoinositol signaling. The expression of key enzymes in the production and metabolism of PIP3 and diacylglycerol (DAG), the two major signaling metabolites in the pathway, was significantly altered. Furthermore, five pivotal enzymes in fatty acid metabolism were markedly affected by p37 treatment. These observations suggest that p37 stimulates host phospholipid turnover, thereby initiating signal transduction cascades. Also, M. hyorhinis p37 lipoprotein can induce effects similar to those observed during association of M. penetrans with human cells.
Conclusions
Microarray technology offers the unique ability to assess major trends and broad patterns of gene expression. Gene array studies yield large amounts of unbiased data, identifying the differential expression of genes across biological processes. Our study was designed within this context and gives an initial broad overview of a biologic response. We highlighted several groups of mediators whose expression levels are significantly altered by the mycoplasmal p37 protein in two prostate cancer cell lines. Despite our choice of a stringent significance cutoff (P ! 0.001) for changes in gene expression, minimizing the number of false positives, 1,979 genes were implicated as affected by p37. Conversely, the vast majority (134,000) of genes expressed in the prostate cell lines did not change after p37 treatment. The overall response to p37 was increased aggressiveness, as evidenced by nuclear and nucleolar enlargement, motility and concomitant changes in gene expression associated with higher biologic potential. Identification of key mediators that induce prostate cancer proliferation and invasion is needed for the development of specific treatments that could improve cancer therapy and disease management. Acknowledgements The authors thank Mick Popp and Marina Telonis-Scott (University of Florida’s joint Shands Cancer Center/Interdisciplinary Center for Biotechnology Research Center Microarray Core) for assistance with microarray hybridization, scanning, image analysis and data analysis.
References Amanatullah DF, Reutens AT, Zafonte BT, Fu M, Mani S, Pestell RG: Cell-cycle dysregulation and the molecular mechanisms of prostate cancer. Front Biosci 5:D372–390 (2000). Band C, Sager R: Distinctive traits of normal and tumor-derived human mammary epithelial cells expressed in a medium that supports longterm growth of both cell types. Proc Natl Acad Sci USA 86:1249–1253 (1989). Blagosklonny MV, Pardee AB: The restriction point of the cell cycle. Cell Cycle 1: 103–110 (2002). Blaser MJ, Parsonnet J: Parasitism by the ‘slow’ bacterium Helicobacter pylori leads to altered gastric homeostasis and neoplasia. J Clin Invest 94: 4–8 (1994). Campbell PM, Der CJ: Oncogenic Ras and its role in tumor cell invasion and metastasis. Semin Cancer Biol 14: 105–114 (2004). Cheng KW, Lahad JP, Gray JW, Mills GB: Emerging role of RAB GTPases in cancer and human disease. Cancer Res 65: 2516–2519 (2005). Chetty C, Bhoopathi P, Joseph P, Chittivelu S, Rao JS, Lakka S: Adenovirus-mediated small interfering RNA against matrix metalloproteinase-2 suppresses tumor growth and lung metastasis in mice. Mol Cancer Ther 5: 2289–2299 (2006).
212
Cohen DW, Simak R, Fair WR, Melamed J, Scher HI, Cordon-Cardo C: Expression of transforming growth factor-alpha and the epidermal growth factor receptor in human prostate tissues. J Urol 152: 2120–2124 (1994). Djakiew D: Dysregulated expression of growth factors and their receptors in the development of prostate cancer. Prostate 42: 150–160 (2000). Dudler R, Schmidhauser C, Parish RW, Wettenhall REH, Schmidt T: A mycoplasma high-affinity transport system and the in vitro invasiveness of mouse sarcoma cells. EMBO J 7: 3971–3974 (1988). Eisen MB, Spellman PT, Brown PO, Botstein D: Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA 95: 14863–14868 (1998). Fang JY, Richardson BC: The MAPK signalling pathways and colorectal cancer. Lancet Oncol 6:322–327 (2005). Fareed GC, Mendiaz E, Sen A, Juillare GJF, Weisenburger TH, Totanes TJ: Novel antigenic markers of human tumor regression. Biol Res Mod 7: 11–23 (1988).
Cytogenet Genome Res 118:204–213 (2007)
Feng SH, Tsai S, Rodriguez J, Lo SC: Mycoplasmal infections prevent apoptosis and induce malignant transformation of interleukin-3-dependent 32D hematopoietic cells. Mol Cell Biol 19: 7995–8002 (1999). Fidler IJ: The organ microenvironment and cancer metastasis. Differentiation 70: 498–505 (2002). Flannery CR: MMPs and ADAMTSs: functional studies. Front Biosci 11: 544–569 (2006). Gilson E, Alloing G, Schmidt T, Claverys JP, Dudler R, Hofnung M: Evidence for a high affinity binding-protein dependent transport system in Gram-positive bacteria and in Mycoplasma. EMBO J 7:3971–3974 (1988). Gschwendt M, Kittstein W, Johannes FJ: Differential effects of suramin on protein kinase C isoenzymes. A novel tool for discriminating protein kinase C activities. FEBS Lett 421: 165–68 (1998). Hanto DW, Frizzera G, Gajl-Peczalska KJ, Sakamoto K, Purtilo DT: Epstein-Barr virus induced B cell lymphoma after renal transplantation: acyclovir therapy and transition from polyclonal to monoclonal B-cell proliferation. N Engl J Med 306:913–918 (1982).
Harper ME, Goddard L, Glynne-Jones E, Wilson DW, Price-Thomas M: An immunocytochemical analysis of TGF alpha expression in benign and malignant prostatic tumors. Prostate 23: 9–23 (1993). Hendrix MJ, Seftor EA, Seftor RE, Fidler IJ: A simple quantitative assay for studying the invasive potential of high and low human metastatic variants. Cancer Lett 38: 137–147 (1987). Huang S, Li JY, Wu J, Meng L, Shou CC: Mycoplasma infections and different human carcinomas. World J Gastroenterol 7: 266–269 (2001). Iczkowski KA, Qiu J, Qian J, Somerville MC, Rittmaster RS, et al: The dual 5-alpha-reductase inhibitor dutasteride induces atrophic changes and decreases relative cancer volume in human prostate. Urology 65: 76–82 (2005). Jarrard DF, Blitz BF, Smith RC, Patai BL, Rukstalis DB: Effect of epidermal growth factor on prostate cancer cell line PC-3 growth and invasion. Prostate 24:46–53 (1994). Jemal A, Siegel R, Ward E, Murray T, Xu J, et al: Cancer statistics, 2006. CA Cancer J Clin 56: 106–130 (2006). Jones HE, Eaton CL, Barrow D, Dutkowski CM, Gee JM, Griffiths K: Comparative studies of the mitogenic effects of epidermal growth factor and transforming growth factor-alpha and the expression of various growth factors in neoplastic and non-neoplastic prostatic cell lines. Prostate 30:219–231 (1997). Kaighn ME, Narayan KS, Ohnuki Y, Lechner JF, Jones LW: Establishment and characterization of a human prostatic carcinoma cell line (PC-3). Invest Urol 17:16–23 (1979). Ketcham CM, Anai S, Reutzel R, Sheng S, Schuster SM, et al: p37 induces tumor invasiveness. Mol Cancer Ther 4:1031–1038 (2005).
Koraitim MM, Metwalli NE, Atta MA, el-Sadr AA: Changing age incidence and pathological types of schistosoma-associated bladder carcinoma. J Urol 154:1714–1716 (1995). Larue L, Bellacosa A: Epithelial-mesenchymal transition in development and cancer: role of phosphatidylinositol 3ⴕ kinase/AKT pathways. Oncogene 24:7443–54 (2005). Maniloff J, McElhaney RN, Finch LR, Baseman JB (eds): Mycoplasmas: Molecular Biology and Pathogenesis. (American Society of Microbiology, Washington, DC, 1992). Massie C, Mills IG: The developing role of receptors and adaptors. Nat Rev Cancer 6: 403–409 (2006). Mott JD, Werb Z: Regulation of matrix biology by matrix metalloproteinases. Curr Opin Cell Biol 16:558–564 (2004). Pulukuri SM, Gondi CS, Lakka SS, Jutla A, Estes N, et al: RNA interference-directed knockdown of urokinase plasminogen activator and urokinase plasminogen activator receptor inhibits prostate cancer cell invasion, survival and tumorigenicity in vivo. J Biol Chem 280: 36529– 36540 (2005). Razin S, Yogev D, Naot Y: Molecular biology and pathogenicity of mycoplasmas. Microbiol Mol Biol Rev 62: 1094–1156 (1998). Reznikoff CA, Brankow DW, Heidelberger C: Establishment and characterization of a cloned line of C3H mouse embryo cells sensitive to postconfluence inhibition of division. Cancer Res 33:3231–3238 (1973).
Salman M, Borovsky ZP, Rottem S: Mycoplasma penetrans infection of Molt-3 lymphocytes induces changes in the lipid composition of host cells. Microbiology 144: 3447–3454 (1998). Schmidhauser C, Dudler R, Schmidt T, Parish RWA: Mycoplasmal protein influences tumour cell invasiveness and contact inhibition in vitro. J Cell Sci 95: 499–506 (1990). Stone KR, Mickey DD, Wunderli H, Mickey GH, Paulson DF: Isolation of a human prostate carcinoma cell line (DU 145). Int J Cancer 21: 274– 281 (1978). Tannock IF, de Wit R, Berry WR, Horti J, Pluzanska A, et al: Docetaxel plus prednisone or mitoxantrone plus prednisone for advanced prostate cancer. N Engl J Med 351:1502–1512 (2004). Troppmair J, Bruder JT, Munoz H, Lloyd PA, Kyriakis J, et al: Mitogen-activated protein kinase/ extracellular signal-regulated protein kinase activation by oncogenes, serum, and 12-O-tetradecanoylphorbol-13-acetate requires Raf and is necessary for transformation. J Biol Chem 269:7030–7035 (1994). Tsai S, Wear DJ, Shih JW, Lo SC: Mycoplasmas and oncogenesis: persistent infection and multistage malignant transformation. Proc Natl Acad Sci USA 92:10197–10201 (1995). Vivanco I, Sawyers CL: The phosphatidylinositol 3Kinase AKT pathway in human cancer. Nat Rev Cancer 2:489–501 (2002). Yang Y, Chisholm GD, Habib FK: Epidermal growth factor and transforming growth factor alpha concentrations in BPH and cancer of the prostate: their relationships with tissue androgen levels. Br J Cancer 67: 152–155 (1993). Yavlovich A, Higazi AA, Rottem S: Plasminogen binding and activation by Mycoplasma fermentans. Infect Immun 69: 1977–1982 (2001).
Cytogenet Genome Res 118:204–213 (2007)
213
MSI-carcinomas and mutations of TP53 or SMAD4 are highly uncommon. However, frameshift mutations affecting genes such as TGFR2, IGF2R, MSH3, MSH6 and BAX are commonly found (Goel et al., 2003). While mutation, hypermethylation and loss of heterozygosity constitute well characterized mechanisms of inactivation, uniparental disomy has only been recently decribed as an alternative mechanism in breast cancer (Murthy et al., 2002; Zhao et al., 2004), basal cell carcinoma (Teh et al., 2005), bronchial cancer (Laframboise et al., 2005) and in leukemia (Raghavan et al., 2005). In the studies mentioned above tumor suppressor genes have been inactivated by deletion of the wildtype allele followed by amplification of the mutated allele. The implementation of the SNP-array analysis was a prerequisite for the large-scale analysis of uniparental disomy. The use of this assay permits us to analyse up to 500,000 SNPs with a single chip and generates detailed information about the number of copies of an allele. Furthermore, homozygosity or heterozygosity of an SNP can be determined. Uniparental disomy is confirmed when chromosomal regions are heterozygous in normal mucous membrane and homozygous in the corresponding carcinoma mucous membrane provided that the copy number in both specimens is identical. The aim of this study is to scrutinize colorectal cancer for corresponding changes and for differences between microsatellite stable and instable carcinomas. Apart from 10KSNP-arrays, spectral karyotyping of nine colorectal carcinoma cell lines (six microsatellite stable, three microsatellite instable) was carried out to correlate changes found in the SNP-array with the corresponding karyotype.
Materials and methods Cells and culture conditions The human colon cancer cell lines SW480, SW620, Caco-2, HCT15, HCT-116 and LoVo were obtained from the American Type Culture collection (ATCC CCL-228, CCl-227, HTB-37, CCL-225, CCL-247 and CCL-229). HT29 and Colo-206 were obtained from the German Collection of Microorganisms and Cell Cultures (DSMZ; Heidelberg, Germany) (ACC299, ACC129, ACC21). Mismatch repair status was retrieved from the literature (Lengauer et al., 1997; Cahill et al., 1998). Cells were grown in standard culture medium (Life Technologies, Eggenstein) supplemented with 10% (v/v) fetal bovine serum (Sigma Chemie, Deisenhofen), penicillin (100 U/ml) and streptomycin (100 g/ml) and were subcultured every four days using trypsin-EDTA detachment. Geki4 was established from a microsatellite stable T3N2M1-tumor of a 79-year-old female (unpublished results) and cultured in final culture medium (DMEM containing 10% FBS, EGF 1 nM/ml (Promega, Mannheim, Germany), TGF- ␣ 200 g/ml (Gibco), gentamicin 125 pg/ml, penicillin 200 U/ml, streptomycin 200 pg/ml, essential amino acids 1%, fungizone 6.6 g/ml, MITO plus serum-expander 1 ml/l (Collaborative Biomedicals, Bedford, MA). Spectral karyotyping (SKY) For spectral karyotyping (SKY) experiments, conventional airdried chromosome slides were prepared from log phase cultures after 3 h of colcemid treatment at a final concentration of 0.3–0.5 g/ml. Slides were stored for 3–14 days at room temperature before hybridization. If necessary, chromosome preparations were subjected to RNase
treatment, pepsin treatment and formaldehyde fixation. Then, the samples were denatured in 70% formamide, 2! SSC for 1–2 min at 70 ° C and quickly dehydrated in a cold ethanol series (70%, 80%, 90%, 100%). Hybridization and detection of the SKY probe (Applied Spectral Imaging Inc., San Diego, CA) was done according to the manufacturer’s instructions. Metaphases were counterstained with DAPI (150 ng/ml DAPI in 2! SSC), and covered with antifade solution (Vectashield mounting medium; Vector Laboratories, Burlingham, CA). Image acquisition was achieved with the SpectraCube system (Applied Spectral Imaging, Inc. (San Diego, CA), and analysed with SKYview imaging software.
Single-nucleotide-polymorphism (SNP-) array analysis 10K-SNP arrays – 10032 SNP’s, Affymetrix. First, 250 ng of either amplified DNA or unamplified DNA was digested with XbaI and ligated to XbaI adaptor before subsequent PCR amplification using AmpliTaq Gold (Applied Biosystems, Foster City, CA). To obtain a sufficient amount of PCR products, four 100 l PCRs were set up for each XbaI adaptor-ligated DNA sample, respectively. The PCR products were pooled, purified, fragmented with DNase I and visualized on a 4% TBE agarose gel to confirm that the sizes ranged from 50 to 100 bp. Subsequently, the fragmented PCR products were end-labeled with biotin and hybridized to the array. Detection was performed with an Affymetrix Fluidics and a GeneArray Scanner. SNP-data analysis SNP-array data were analysed with the CNAT (‘copy number analysis’) tool (Affymetrix). For each cell line information is provided on zygosity (homo- or heterozygote: AA, BB, or AB), copy number (SPA_ CN) and the significance of the copy number variation (SPA_pval: +/– log10 (P value)). Furthermore, the ‘meta analysis significance’ (CPA_ pval: +/–log10 (meta P value)), and the –log10 of the probability for long homozygous stretches happen in random (LOH) was assessed. In this study, the level of significance for LOH was set to 10 (P ! 10 –10). Mean values of SPA_pval of the identified significant stretches of homozygous calls (‘monoallelic regions’) were calculated and the cut off value for amplification or deletion was set to 8 1.3 (P ! 0.05). With this approach monoallelic fragments and the corresponding copy numbers could be detected in MSI and MSS-cell lines. We assumed that monoallelic fragments accompanied by a loss of copy number are the consequence of a hemizygous deletion. Monoallelic fragments with no change of copy number emerge from copy-neutral events such as mitotic nondisjunction followed by duplication of one of the parental chromosomes. However, a monoallelic fragment accompanied by significant increase in copy numbers, strongly suggests that preferential amplification of one parental allele that may be masking the presence of the other allele has occurred. On the other hand, the increase in copy number may suggest a loss of one allele followed by amplification of the remaining allele (Zhou et al., 2005). In another approach, amplifications and deletions, detected by SNP-array and chromosomal aberrations found in SKY-karyotypes are correlated. Stretches of significant SPA_pval and CPA_pval values (P ! 0.05) were compared to marker chromosomes irrespectively of their homozygous or heterozygous status. Chromosomal aberrations resulting in copy number differences were also detected by SNP-array, thus allowing an exact mapping of the corresponding breakpoints.
Results
Spectral karyotyping (SKY) and SNP-array analysis SKY results are based on the analysis of eight (Geki4) to 20 (HCT15) metaphases. Deleted and amplified chromosomal segments were identified by SNP-array analysis. In combination, SKY-analysis, DAPI-staining, and SNP-array results allowed a detailed characterization of nearly all marker chromosomes (Table 1). Furthermore, SNP-array
Cytogenet Genome Res 118:214–221 (2007)
215
Table 1. SKY-karyotypes of MSS- and MSI-cell lines
Cell linea
SKY-Karyotypeb
HT29 [11] px+128
69–74,XX,del(X)(p11.2),del(3)(p11.1),del(4)(q32.3),del(5)(q11),t(6;14)(q21;q13), +7,del(7)(p?;q?),–8,der(8)i(8)(qter]q10::q10]q24::hsr::q24]qter),+11, –13,der(13)i(13)(q10)del(q14),der(13)t(5;13)(p13;p11.1),i(13)(q10),–14,+15, –16,der(17)t(17;19)(p11;p11),i(18)(p10),–19,i(19)(q10)dup(19)(q13.1q13.4),+20,–21,–22
SW480 [10] px+11
52–58,XY,–Y,+X,t(1;9)(q10;q10),+der(2)t(2;12)(q36.1;q11),+der(3)del(3) (p24.2),t(5;20)(q21.1;p12),+der(7)inv(7)(q34q22)t(7;14)(q22;q22),+der(7) t(7;13)(q22;q14.11),–8,der(8)t(8;19)(p10;q10),+der(9)t(8;9)(q11.2;p13.3), der(10)t(3;10)(q11.1;p11.2),+11,i(12)(q10),+13,+17,del(18)(q11)!2,der(19)t(19;8;19;5) (pter]q11::q24.13]q24.21::?]?::pter]p14.1),+der(20)t(5;20)(q21.1;p12),+21
SW620 [10] px+18
46–52,XY,–Y,+X,der(2)t(2;12)(p24;p11),del(3)(p13),del(4)(q28),t(5;20)(q21.1;p12), +der(5)t(5;20)(q21.1;p12),der(6)t(6;7)(q25.3;q33),+der(7)del(7)(p13)del(7)(q21), der(8)t(8;17)(p22;p12),der(10)t(10;13)(q24;q14.11),+11,–13,der(3;16;1;16;8;16;1;16;10) (pter]p13::p13.3]q21::?]?::q21]q23.3::?]?::q23.3]q21::?]?::?]?::q24]qter),der(18) t(18;15;17)(pter]q11.2::?]?::p12]pter),der(18)t(15;18)(p11;q21)t(15;17)(p13;q22)
CaCo2 [11] px+23
89,XX,–X,–X,der(1)t(1;19)(p10;q13.11),del(1)(p13.11),der(2)t(2;9)(q42.2;p21.1)!2, +der(7)del(p?)del(q?),–8,–9!3,+der(10)t(10;16)(q22;q12.1)!2,+11,del(12)(q13.21),+13, –14,der(15;20)(p11;q11.21)!2,–16!2,–17!2,der(17)t(11;17)(q13.3;p12)!3, –18,del(18)(p11.21)!2,+20,–21,der(21)t(20;21)(q11.21;?)!2
Colo206 [10] px+6
66–71,XX,der(Y)t(3;Y;Y;3)(qter]q13.32::qter]pter::pter]qter::q13.32]qter), i(1)(q10),der(1)t(1;20)(q36;q12),der(1)t(1;10)(p11;q21.3),der(3)del(p?)del(q12.3), +der(5)t(5;21)t(p13;q11)[10],–6,+der(6)t(13;6;13)(qter]q11::q12.3]q13::q11]qter)!3, +7,der(8)t(8;21)(p11.21;q11),del(10)(p11),der(10)t(6;10)del(6)del(10)(p24.3]p12.3::p11 ]q21.3),+del(11)(q11),–13!2,–14,+del(17)(q11),der(18)t(17;18)(q11;q21.1), der(19)t(19;22;15)(?;?;?),+20!2,+der(20)t(20;22)(q12;?),–21!2
Geki4 [8] p8
75,XXX,+X,+der(1)t(1;17)(q21.1;q11),+3,der(4)t(4;6)(?;?),+5,+del(7)(q11), +i(8)(q13),der(9)t(9;11)(q33.2;q12),+16[3],–17,–18,+20!2,–22
LoVo [11] px+6
48–49,XY,t(2;12),+5,+7,+12,+i(15)(q10)
HCT15 [20] px+2
46,XY,der(8)t(8;17),inv(11)(p15.3q13.1)
HCT116 [10] px+8
45,XY,–Y,der(10)dup(q23.31]q26.3)t(10;16)(q26.3;?),der(16)t(8;16) (q11.23;pter),der(18)t(17;18)(q21.31;p11.2)
a b
Number of analyzed metaphases in brackets. Marker chromosomes that were identified and further characterized by SNP-array analysis are in bold.
analysis enabled the identification of monoallelic chromosomal regions in MSS- and MSI-cell lines and their appropriate copy numbers (Fig. 1A, B). Monoallelic chromosomal fragments in MSI-cell lines Monoallelic regions with unchanged copy number (uniparental disomy) were identified in the region 2pter]qter (LOVO), 3pter]p14.3 (HCT116), 5pter]q11.2 (HCT116), 5q14.1]qter (HCT15), 7pter]p15.2 (HCT116), 9pter] p13.3 (LOVO/HCT15), and 14q13.1]qter (HCT15) (Fig. 1A). In these regions genes relevant for colon carcinogenesis could be assigned to 2p21]p22 (MSH2), 3p21.3 (MLH1), 5q21]q22 (APC), and 9p21 (CDKN2A). A biallelic deletion of MSH2 is described in LOVO (Wheeler et al., 1999), a homozygous point mutation of MLH1 in HCT116 (Wheeler et al., 1999), a homozygous mutation of APC in HCT15 (Rowan et al., 2000), and a biallelic promotor methylation of CDKN2A in LOVO and HCT15 (Zheng et al., 2000). Identified biallelic mutations fit well to the monoallelic regions of the respective cell lines. The results are in accordance with a
216
Cytogenet Genome Res 118:214–221 (2007)
mutation followed by a deletion of the nonmutated allele and an amplification of the mutated allele resulting in a uniparental disomy and an apparent biallelic mutation (Table 2, 3). Monoallelic chromosomal fragments in MSS-cell lines Due to the high degree of chromosomal instability only common aberrations which could be detected in more than 50% of the cell lines (four of six) were considered. Monoallelic regions without a copy number alteration (uniparental disomy) or with copy number reduction were located in 4pter]qter, 8pter]p22, 17pter]p12, 18q11.2]qter. Genes relevant for colon carcinogenesis are located in 17p13.1 (TP53) and 18q21.2 (SMAD4). Mutations in SMAD4 were described in HT29, SW480, SW620 and Caco2 (WoodfordRichens et al., 2001). The copy number of 18q21.2 was reduced in all cell lines, suggesting classical loss of heterozygositiy (LOH). Mutations in TP53 were identified in HT29, SW480, SW620 and Caco2 (Woodford-Richens et al., 2001). Whereas copy number of 17p13.1 was reduced in HT29 and
LOVO HCT15 HCT116
LOVO HCT15 HCT116
LOVO HCT15 HCT116
LOVO HCT15 HCT116
LOVO HCT15 HCT116
3
5
9
14
LOVO HCT15 HCT116
LOVO HCT15 HCT116
LOVO HCT15 HCT116
2
A
7
10
17
Fig. 1. (A) Alterations of MSI-colon cancer cell lines identified by SNP-array analysis. Column 1: Lovo, column 2: HCT15, column 3: HCT116. Green – decreased copy number, yellow – no copy number alteration/monoallelic fragment, red – increased copy number.
Cytogenet Genome Res 118:214–221 (2007)
217
17
HT29 SW480 SW620 Caco2 Colo206F Geki4
HT29 SW480 SW620 Caco2 Colo206F Geki4
HT29 SW480 SW620 Caco2 Colo206F Geki4
HT29 SW480 SW620 Caco2 Colo206F Geki4
8
18
HT29 SW480 SW620 Caco2 Colo206F Geki4
HT29 SW480 SW620 Caco2 Colo206F Geki4
4
B
11
13 Fig. 1. (B) Alterations of MSS-colon cancer cell lines identified by SNP-array analysis. Alterations shown were found in at least four of the six MSS-cell lines. Column 1: HT29, column 2: SW480, column 3: SW620, column 4: Caco2, column 5: Colo206, column 6: Geki4. Green – decreased copy number, yellow striped green – decreased copy number/monoallelic fragment, yellow – no copy number alteration/monoallelic fragment, yellow striped red – increased copy number/ monoallelic fragment, red – increased copy number.
218
Cytogenet Genome Res 118:214–221 (2007)
Table 2. Mutations of tumor suppressor genes CDKN2A, MLH1, MSH2 in MSI- and MSS-cell lines
Cell line
MSH2 2p21–22a
Gain/ UPDc MLH1 lossb 3p21.3a
Gain/ lossb
UPDc
CDKN2A Gain/ UPDc 9p21.3a lossb
HT29 SW480 SW620 Caco2 LOVO HCT15 HCT116
– – – – D/D – –
– gain – gain – – –
loss gain – – – – –
– + – – – – +
M/M M/M M/M M/M M/M M/M M/23
a b c
Table 3. Mutations of tumor suppressor genes APC, TP53, SMAD4 in MSI- and MSS-cell lines
– – – – + – –
M/U – – – – – 252/252
– – – – – – –
– + + + + + –
Mutation: ‘XXX’ = codon of mutation; D = allele deleted; M = allele methylated. Copy number determined by SNP-array: ‘–’ = no copy number alteration. UPD = uniparental disomy.
Cell line
APC 5q21–q22a
Gain/ UPDc TP53 lossb 17p13.1a
Gain/ UPDc SMAD4 lossb 18q21.1a
Gain/ UPDc lossb
HT29
853 1555 1338 1338 1367 1114 1430 1417 –
–
–
273
loss
+
311
loss
+
gain – – –
+ + + –
273 273 204 –
gain gain loss –
+ + + –
IVS7 +5G>C IVS7 +5G>C 351 –
loss loss loss –
+ + + –
– –
+ –
145/WT –
– –
– –
– –
– –
– –
SW480 SW620 Caco2 LOVO HCT15 HCT116 a b c
Mutation: ‘XXX’ = codon of mutation. Copy number determined by SNP-array: ‘–’ = no copy number alteration. UPD = uniparental disomy.
Caco2, a copy number gain of 17p13.1 was detected in SW480/SW620 (Table 2, 3). An increased copy number was also found in the monoallelic regions 11p11.2]q13.4, 11q14.1]q22.2, and 13pter]q13.3. The proto-oncogene Cyclin D1 (CCND1) is located at 11q13.3. In 1998, Hosokawa and Arnold demonstrated an allele-specific overexpression of CCND1 in HT29, localized at chromosomal band 11q13.3. To date, no published data is available on the TP53 and SMAD4 mutational status of Colo206 and Geki4 nor on potential tumor suppressor genes in the chromosomal regions 4pter]qter and 8pter]p22 or potential oncogenes in 13pter]q13.3.
N2A is localized at 9p21. One methylated and one unmethylated allele, respectively were detected in HT29 and HCT116. No corresponding monoallelic region was identified. A biallelic promotor methylation of CDKN2A was detected in SW480, SW620 and Caco2, LOVO, HCT15 (Zheng et al., 2000). A corresponding monoallelic region without copy number alterations was found in 9p22 of these cell lines. To date, no data on the APC and CDKN2A mutational status of Colo206 and Geki4 is available.
Monoallelic chromosomal fragments in MSI- and MSS-cell lines Common monoallelic chromosomal fragments were found in 5q21]q22 and 9p21. APC, which is located in 5q21]q22, is frequently inactivated early in colon carcinogenesis (Fearon and Vogelstein, 1990). In HT29 and LOVO, both alleles harbour different mutations (Rowan et al., 2000) and no monoallelic regions are found in the SNP array. Only one mutation was identified in SW480/SW620, Caco2 and HCT15 (Rowan et al., 2000) and a monoallelic region without copy number alteration was found in the 5q21]q22 region of these cell lines. The tumor suppressor gene CDK-
Spectral karyotyping (SKY) and SNP-array analysis The application of SKY enabled a detailed analysis of marker chromosomes in nine colorectal colon cancer cell lines. The limitations of SKY are breakpoint analysis and the identification of hidden alterations like uniparental disomy (UPD). Breakpoint analysis can effectively be supported by SNP-array technology. By measuring relative intensities of SNPs in defined chromosomal regions, amplified and deleted regions were identified. Some aberrations were not identified by SNP-array analysis. This may be explained by the existence of subclones that do not contain the respective marker chromosome affecting the intensities of
Discussion
Cytogenet Genome Res 118:214–221 (2007)
219
the respective SNPs. Abdel-Rahman et al. (2001) analysed 17 colorectal cancer cell lines by SKY and CGH including HT29, SW480, SW620, HCT116 and LoVo. Some marker chromosomes were differently described by this group. Reasons for this discrepancy may be a different passage number that may lead to an overgrowth of different subclones. SNPArray and CGH-analysis may also lead to some different results because SNP-array is more precise than CGH in characterizing small chromosomal alterations and has to be compared to Matrix-CGH. The advantage of SNP-array compared to Matrix-CGH is the possibility of detecting monoallelic regions even without the analysis of the corresponding normal mucosa. These monoallelic regions originated either by deletions, amplifications or uniparental disomy. Interestingly we found uniparental disomy in microsatellite stable and instable colorectal cell lines that may be involved in inactivation of tumor suppressor genes. Inactivation of tumor suppressor MLH1, MSH2, APC and CDKN2A genes by uniparental disomy in MSI-cell lines In the MSI-cell lines monoallelic chromosomal fragments were always accompanied with a balanced copy number and a diploid karyotype (uniparental disomy) (Fig. 1A). Four of seven identified regions were correlated with homozygous mutations in tumor relevant suppressor genes (MSH2, MLH1, APC, CDKN2A) (Table 2, 3). Obviously, inactivation of these genes was achieved by a mutation of one allele, loss of the remainig wildtype allele and a duplication of the mutated allele. Interestingly, MSH2, MLH1, APC and CDKN2A are genes that are mutated early in the adenomacarcinoma-sequence. Inactivation of MSH2 or MLH1 leads to MSI-phenotype. Frameshift-mutations in genes such as TGFRB2 (3p22), IGF2R (6q26), BAX (19q13.3]q13.4) that are a consequence of MSI, could not be correlated with monoallelic regions (data not shown). Thus, uniparental disomy seems to occur early in the carcinogenesis of MSI-tumors. Further studies are warranted to detect potentially mutated genes in the regions 14q13.1]qter, 5pter]q11.2, 7pter]p15.2. Inactivation of tumor suppressor genes by LOH (TP53 and SMAD4) and uniparental disomy (APC, CDKN2A) in MSS-cell lines In contrast to the MSI-cell lines, MSS-cell lines are known to harbour multiple chromosomal gains and losses that represent activated oncogenes and inactivated tumor suppressor genes. Common alterations were identified in at least four of six MSS-cell lines (Fig. 1B). As expected, mutations of TP53/SMAD4 were described in all MSS-cell lines (Table 3). Monoallelic regions containing TP53 and SMAD4 displayed a reduced copy number (representing allelic loss) in almost all MSS-cell lines analysed, except for TP53 in SW480/ SW620 (Fig. 1B). In accordance with our results, Rochette et al. (2005) showed a hyperdiploid (52–58) karyotype of SW480 with a double mutation of TP53 in SW480, which was detected in each of three copies of the gene. In comparison to LOVO, TP53 is 20! over-expressed in
220
Cytogenet Genome Res 118:214–221 (2007)
SW480 and retained some of its function (Rochette et al., 2005). The relevance of this finding has to be discussed further. However, it was shown previously that over-expression of TP53 may be correlated with metastatic disease and impaired prognosis. Interestingly, genes that are mutated early in the adenoma-carcinoma sequence (APC, CDKN2A) were not inactivated by LOH. Uniparental disomy was found to be the inactivating mechanism for APC in SW620/ Caco2 and for CDKN2A in SW480/SW620/Caco2. The frequency of UPD in these chromosomal regions seems to be very similar in MSI- and MSS cell lines (Table 2, 3). Inactivation of tumor suppressor genes by UPD in other studies Uniparental disomy in colorectal cancer was also described recently by Andersen et al. (2007). They investigated 15 microdissected adenocarcinomas by 10K SNP-arrays and found multiple regions with copy number alterations and loss of heterozygosity (LOH). Half of the LOH regions identified showed no evidence of a reduced copy number, indicating the presence of uniparental structures. The distribution of these structures was non-random primarily involving 8q, 13q, and 20q. A transcriptional analysis revealed an unchanged expression level in areas with intact copy number. Karyotyping and MSI-analysis were not performed. The results are partially in accordance with our findings in MSS-cell lines but differ in the distribution of the affected regions. Raghavan et al. (2005) detected UPD in 20% of 64 examined AML specimens. Most frequently (n = 5) a UPD was detected in the chromosomal region 11pter]p13. Also affected were 6pter]p21, 6pter]p11, 9p, 9q33]qter, 13, 19q12]qter and 21q21.1. Gorletta et al. (2005) found similar results in 32 analysed NK-AML-specimens. UPD was found in 9pter]p13.1, 10pter]p11.21, 10q21.3]qter, 13q12.11]qter and 16q24.1. UPD of 9p13.3 was demonstrated in polycythemia vera by Kralovics et al. (2002). Zhao et al. (2004) reported UPD of chromosome 9 and 9q in two mamma and lung cancer cell lines. Recently, Teh et al. (2005) detected a monoallelic region 9q22.3 in 13 of 14 examined basal cell carcinomas (38% UPD). The tumor suppressor gene PTCH which is located in this region was mutated in 68% of the specimens. Activation of oncogenes (CCND1) by specific amplification of the mutated allele The monoallelic regions 11p11.2]q13.4, 11q14.1]q22.2 and 13pter]q13.3 are accompanied by a significant increase in copy number. This suggests a preferential amplification of one parental allele that may be masking the presence of the other allele. Alternatively, the increase in copy number may suggest a loss of one allele followed by amplification of the remaining allele (Zhou et al., 2005). CCND1 is localized at 11q13.3 and it has been shown previously that only the mutated allele is expressed in HT29 (Hosokawa and Arnold, 1998). Nothing is known yet about allele-specific expression of a mutated CCND1 in SW480/SW620 and Caco2. Our results suggest similar mechanisms that have to be confirmed by further studies.
Activation of oncogenes by specific amplification of the mutated allele in other studies A monoallelic amplification of EGFR (7p11.2) was demonstrated by Laframboise et al. (2005) in lung cancer. Bianchi et al. (1990) reported a nonrandom duplication of the chromosome bearing a mutated Ha-ras-1 allele in mouse skin tumors. Zhuang et al. (1998) described a trisomy 7-harbouring non-random duplication of the mutant MET allele in hereditary papillary renal carcinomas. Conclusion
This study further characterized complex chromosomal alterations by SNP-array analysis and correlated monoallelic regions found by SNP-array analysis with inactivated tumor suppressor genes and activated oncogenes. Some of the genes relevant for colon carcinogenesis are inactivated
by LOH (TP53, SMAD4 in MSS-cell lines). Monoallelic regions without copy number alterations represent uniparental disomy (UPD). For the first time it was shown that UPD seems to be involved in activation of early-acting tumor suppressor genes in MSS- (APC, CDKN2A) and MSI(MLH1, MSH2, APC, CDKN2A) colorectal cancer cell lines. Monoallelic regions with increased copy number may represent oncogenes activated by allele-specific amplification (CCND1 in MSS-cell lines). Further studies are warranted to identify new tumor suppressor genes in UPD-regions of MSS- (4pter]qter and 8pter]p11) and MSI-colorectal cancer cells (5pter]q11.2, 7pter]p15.2, 14q13.1]qter) and a potential oncogene in MSS-colorectal cancer cells (13pter]q13.3). Furthermore, additional SNP-array analysis with a higher number of tumors and more SNPs (100,000– 500,000) will lead to a greater accuracy in detecting candidate regions.
References Abdel-Rahman WM, Katsura K, Rens W, Gorman PA, Sheer D, et al: Spectral karyotyping suggests additional subsets of colorectal cancers characterized by pattern of chromosome rearrangement. Proc Natl Acad Sci USA 98: 2538– 2543 (2001). Andersen CL, Wiuf C, Kruhoffer M, Korsgaard M, Laurberg S, Orntoft TF: Frequent occurrence of uniparental disomy in colorectal cancer. Carcinogenesis 28: 38–48 (2007). Bianchi AB, Aldaz CM, Conti CJ: Nonrandom duplication of the chromosome bearing a mutated Ha-ras-1 allele in mouse skin tumors. Proc Natl Acad Sci USA 87: 6902–6906 (1990). Cahill DP, Lengauer C, Yu J, Riggins GJ, Willson JK, et al: Mutations of mitotic checkpoint genes in human cancers. Nature 392: 300–303 (1998). Fearon ER, Vogelstein B: A genetic model for colorectal tumorigenesis: Cell 61: 759–767 (1990). Goel A, Arnold CN, Niedzwiecki D, Chang DK, Ricciardiello L, et al: Characterization of sporadic colon cancer by patterns of genomic instability. Cancer Res 63: 1608–1614 (2003). Gorletta TA, Gasparini P, D’Elios MM, Trubia M, Pelicci PG, Di Fiore PP: Frequent loss of heterozygosity without loss of genetic material in acute myeloid leukemia with a normal karyotype. Genes Chromosomes Cancer 44: 334–337 (2005). Hosokawa Y, Arnold A: Mechanism of CCND1 (CCND1, PRAD1) overexpression in human cancer cells: analysis of allele-specific expression. Genes Chromosomes Cancer 22: 66–71 (1998). Jallepalli PV, Lengauer C: Chromosome segregation and cancer: cutting through the mystery. Nat Rev Cancer 1: 109–117 (2001).
Kane MF, Loda M, Gaida GM, Lipman J, Mishra R, et al: Methylation of the hMLH1 promoter correlates with lack of expression of hMLH1 in sporadic colon tumors and mismatch repairdefective human tumor cell lines. Cancer Res 57:808–811 (1997). Knudson AG: Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci USA 68: 820–823 (1971). Kralovics R, Guan Y, Prchal JT: Acquired uniparental disomy of chromosome 9p is a frequent stem cell defect in polycythemia vera. Exp Hematol 30:229–236 (2002). Laframboise T, Weir BA, Zhao X, Beroukhim R, Li C, et al: Allele-specific amplification in cancer revealed by SNP array analysis. PLoS Comput Biol 1:e65 (2005). Lengauer C, Kinzler KW, Vogelstein B: Genetic instability in colorectal cancer. Nature 386: 623– 627 (1997). Moolgavkar SH, Knudson AG: Mutation and cancer: a model for human carcinogenesis. J Natl Cancer Inst 66:1037–1052 (1981). Murthy SK, DiFrancesco LM, Ogilvie RT, Demetrick DJ: Loss of heterozygosity associated with uniparental disomy in breast carcinoma. Mod Pathol 15: 1241–1250 (2002). Raghavan M, Lillington DM, Skoulakis S, Debernardi S, Chaplin T, et al: Genome-wide single nucleotide polymorphism analysis reveals frequent partial uniparental disomy due to somatic recombination in acute myeloid leukemias. Cancer Res 65: 375–378 (2005). Rochette PJ, Bastien N, Lavoie J, Guerin SL, Drouin R: SW480, a p53 double-mutant cell line retains proficiency for some p53 functions. J Mol Biol 352:44–57 (2005). Rowan AJ, Lamlum H, Ilyas M, Wheeler J, Straub J, et al: APC mutations in sporadic colorectal tumors: A mutational ‘hotspot’ and interdependence of the ‘two hits’. Proc Natl Acad Sci USA 97:3352–3357 (2000).
Teh MT, Blaydon D, Chaplin T, Foot NJ, Skoulakis S, et al: Genomewide single nucleotide polymorphism microarray mapping in basal cell carcinomas unveils uniparental disomy as a key somatic event. Cancer Res 65: 8597–8603 (2005). Wheeler JM, Beck NE, Kim HC, Tomlinson IP, Mortensen NJ, Bodmer WF: Mechanisms of inactivation of mismatch repair genes in human colorectal cancer cell lines: the predominant role of hMLH1. Proc Natl Acad Sci USA 96: 10296–10301 (1999). Woodford-Richens KL, Rowan AJ, Gorman P, Halford S, Bicknell DC, et al: SMAD4 mutations in colorectal cancer probably occur before chromosomal instability, but after divergence of the microsatellite instability pathway. Proc Natl Acad Sci USA 98:9719–9723 (2001). Zhao X, Li C, Paez JG, Chin K, Janne PA, et al: An integrated view of copy number and allelic alterations in the cancer genome using single nucleotide polymorphism arrays. Cancer Res 64: 3060–3071 (2004). Zheng S, Chen P, McMillan A, Lafuente A, Lafuente MJ, et al: Correlations of partial and extensive methylation at the p14(ARF) locus with reduced mRNA expression in colorectal cancer cell lines and clinicopathological features in primary tumors. Carcinogenesis 21: 2057–2064 (2000). Zhou X, Rao NP, Cole SW, Mok SC, Chen Z, Wong DT: Progress in concurrent analysis of loss of heterozygosity and comparative genomic hybridization utilizing high density single nucleotide polymorphism arrays. Cancer Genet Cytogenet 159:53–57 (2005). Zhuang Z, Park WS, Pack S, Schmidt L, Vortmeyer AO, et al: Trisomy 7-harbouring non-random duplication of the mutant MET allele in hereditary papillary renal carcinomas. Nat Genet 20: 66–69 (1998).
Cytogenet Genome Res 118:214–221 (2007)
221
Acoustic neurinoma, bilateral See: Neurofibromatosis 2 Acute intermittent porphyria See: Porphyria, acute in-
termittent Adenomatous polyposis, familial See: Familial adenomatous polyposis (FAP) Aicardi syndrome (OMIM No. 304050) This enigmatic syndrome is clinically characterized by agenesis of the corpus callosum, chorioretinal lacunae and infantile spasms. The syndrome is associated with an increased risk of tumors, most commonly papillomas of the choroid plexus. Other tumors reported in Aicardi syndrome include lipomas, angiosarcomas, hepatoblastomas, intestinal polyposis, and embryonal carcinomas. The inheritance of Aicardi syndrome is not entirely clear. Since the syndrome occurs exclusively in girls and in boys with an XXY sex chromosomal complement, it is thought to be due to an X-linked dominant mutation lethal in XY males. The putative AIC gene is located in chromosome band Xp22. Albinism See: Oculocutaneous albinism type 1 Alpha-1-antitrypsin deficiency (OMIM No. 107400) This disorder carries an elevated risk of hepatocellular carcinoma (as does acute intermittent porphyria, porphyria cutanea tarda and tyrosinemia type I). Alpha-1-antitrypsin deficiency is inherited in an autosomal recessive manner and is caused by mutations in the protease inhibitor 1 gene which is located in chromosome band 14q32.1. APC-associated polyposis is a term that has been applied to polyposis, the development of multiple polyps in the colon, caused by mutation of the APC (adenomatous polyposis coli) gene. The term ‘APC-associated polyposis’ brings things together on the molecular level, but it tends to blur the clinical entities contained within APC-associated polyposis. These include: (1) familial adenomatous polyposis (FAP); (2) Gardner syndrome and Turcot syndrome, and (3) attenuated FAP. See: Attenuated FAP; Familial adenomatous polyposis (FAP). Ataxia-telangiectasia (OMIM Nos. 208900, 251260) This autosomal recessive disease was first recognized by the clinical concurrence of progressive cerebellar ataxia, oculocutaneous telangiectasia, and frequent sinopulmonary infections. On the laboratory level, ataxia-telangiectasia (A-T) was found associated with immunodeficiency and t(7;14) chromosome translocations in peripheral lymphocytes. A-T carries an increased risk of malignancy, particularly of leukemia and lymphoma. The leukemia in A-T is usually of T-cell origin, while lymphoma tends paradoxically to be B-cell. Adults with A-T may also be at elevated risk for solid tumors including ovarian, breast, gastric and pancreatic cancers. The gene responsible for A-T has been termed ATM (ataxia-telangiectasia mutated) and is located in chromosome band 11q22.3. ATM encodes a serine-protein kinase. Attenuated FAP (attenuated familial adenomatous polyposis or AFAP) is an autosomal dominant condition associated with the formation of fewer colon polyps than in classical FAP and a decidedly less prominent association with extraintestinal tumors. Attenuated FAP, like FAP, is caused by mutations in the APC gene. The APC gene, which en-
codes the adenomatous polyposis coli protein, is situated in chromosomal region 5q21]q22. Barrett esophagus (OMIM No. 109350) Barrett esophagus is chronic ulcerating esophagitis secondary to gastroesophageal reflux disease (GERD). Barrett esophagus carries an elevated risk of cancer of the esophagus. The pediatric form of this disorder is inherited as an autosomal dominant trait and is due to mutation of the GER gene in chromosome 13q14. (GER stands for GastroEsophageal Reflux.). Bannayan-Riley-Ruvalcaba syndrome See: Macrocephaly, multiple lipomas, hemangiomata Basal cell nevus syndrome See: Nevoid basal cell carcinoma syndrome Beckwith-Wiedemann syndrome (OMIM No. 130650) This syndrome is a disorder of growth associated with an increased risk of embryonal tumors. These include Wilms tumor, hepatoblastoma, neuroblastoma and rhabdomyosarcoma. The increased risk of embryonal neoplasia appears confined to the first eight years of life. A minority of cases of Beckwith-Wiedemann syndrome are inherited in an autosomal dominant manner. Abnormalities in chromosome band 11p15.5 underlie the syndrome. Candidate imprinted genes mapping to band 11p15.5 include the H19 gene, the CDKN1C (cyclin-dependent kinase inhibitor 1C) gene, and the KCNQ1OT1 gene (encoding the KCNQ1-overlapping transcript 1). Birt-Hogg-Dubé syndrome (OMIM No. 135150) This autosomal dominant disorder is characterized by skin abnormalities, pulmonary cysts, and renal tumors of the oncocytoma and chromophobe types. Clear cell renal cell carcinoma and papillary renal carcinoma may also be increased in frequency. Mutation of the FLCN (folliculin) gene in chromosome band 17p11.2 is responsible for the syndrome. Bloom syndrome (OMIM No. 210900) This autosomal recessive disorder is associated with a broad distribution of malignancies (epithelial, hematopoietic, lymphoid, connective tissue, germ cell, nervous system, and kidney) resembling the distribution of cancers in the general population but tending to occur at an earlier age. The BLM gene involved in Bloom syndrome has been mapped to chromosome band 15q26.1. BLM is a member of the RecQ family of DNA helicases. Mutations in other RecQ helicase genes cause Rothmund-Thomson syndrome and Werner syndrome. BRCA1 (OMIM No. 113705) Mutations in BRCA1 are associated with a predisposition to breast and ovarian cancer. There may also be an increase in prostate cancer and colon cancer in families with BRCA1. Cancer-susceptibility mutations in BRCA1 are transmitted in an autosomal dominant manner. BRCA1 is in chromosome band 17q21. It encodes the breast cancer type 1 susceptibility protein which may act as a tumor suppressor. BRCA2 (OMIM No. 600185) Mutations in BRCA2 predispose to breast and ovarian cancer as well as to prostatic and pancreatic cancer. Numerous other cancers have been reported in BRCA2 families, including male breast cancer and
Cytogenet Genome Res 118:222–228 (2007
223
pancreatic cancer. BRCA2 mutations are inherited in an autosomal dominant manner. The BRCA2 gene, which is in chromosome band 13q12.3, encodes the breast cancer type 2 susceptibility protein. This protein may function similarly to the breast cancer type 1 susceptibility protein produced by BRCA1, and the two proteins may interact. Carcinoid, familial intestinal (OMIM No. 114900) Families with carcinoid in several persons in one or two generations have been reported. It is thought that these represent the transmission in an autosomal dominant manner of a mutant SDHD gene in chromosome band 11q23. Note that familial carcinoid also occurs in multiple endocrine neoplasia (MEM1 and MEM2). Carney complex (OMIM Nos. 160980, 605244) This autosomal dominant disorder also known as the Name or Lamb syndrome, is characterized by skin pigmentary anomalies (lentigines and nevi), atrial and mucocutaneous myxomas, endocrine tumors or hyperactivity and schwannomas. Associated malignancies include Sertoli cell tumor and Leydig cell tumor. The Carney complex results from mutations in the PRKAR1A gene in chromosome region 17q23] q24 or the CNC2 gene in chromosome band 2p16. The PRKAR1A gene appears to function as a tumor suppressor. Congenital central hypoventilation syndrome See: Central hypoventilation syndrome, congenital Congenital neutropenia See: Neutropenia, congenital Central hypoventilation syndrome, congenital (OMIM Nos. 209880, 603851) Congenital central hypoventilation syndrome is associated with tumors of neural crest origin including neuroblastoma, ganglioneuroma, and ganglioneuroblastoma. The syndrome is inherited in an autosomal dominant fashion and is also known as Haddad syndrome and Ondine’s curse. The major mutation for the syndrome is in the PHOX2B gene in chromosome band 4p12. The syndrome can also be caused by mutation in several other genes, including RET in chromosome band 10q11.2, GDNF in 5p13.1]p12, EDN3 in 20q13.2]q13.3, BDNF in 11p13, and ASCL1 in 12q22]q23. Chordoma, familial (OMIM No. 215400) Chordomas are tumors derived from the embryonic notochord. Multigeneration families with chordoma (and an instance of maleto-male transmission) are consistent with the autosomal dominant inheritance of a gene (CHDM) in chromosome band 7q33. Colon cancer, hereditary non-polyposis See: Hereditary non-polyposis colon cancer Congenital neutropenia See: Neutropenia, congenital Cowden syndrome (OMIM No. 158350) This autosomal dominant syndrome is also known as the gingival multiple hamartoma syndrome or PTEN hamartoma tumor syndrome. It is a multiple hamartoma condition which carries a high risk of benign and malignant tumors of the breast and thyroid. Cancer of the endometrium may also be increased. Mutations in the PTEN gene in 10q23.31 are responsible for the disorder which is allelic with the syndrome of macrocephaly, multiple lipomas, hemangiomata.
224
Cytogenet Genome Res 118:222–228 (2007
Costello syndrome (OMIM No. 218040) Costello syndrome is an autosomal dominant disorder caused by mutation in the HRAS gene in chromosome band 11p15.5. The syndrome predisposes to neoplasms including epithelioma, bladder carcinoma, rhabdomyosarcoma and vestibular schwannoma. The disorder is also known as the faciocutaneoskeletal or FCS syndrome. ELA-related neutropenia See: Neutropenia, congenital Exostoses, multiple (OMIM Nos. 133700, 133701, 600209, 608177, 608210) This autosomal dominant disorder is characterized by abnormal prominences at the ends of long bones. The prominences consist of bone capped with cartilage and present an increased risk of chondrosarcoma, particularly in adulthood. The syndrome is caused by mutation in one of several genes – the EXT1 gene encoding exostosin1 in chromosome region 8q24.11]q24.13, the EXT2 gene encoding exostosin-2 in chromosome region 11p12]p11, and the EXT3 gene on chromosome 19p. The disorder has been correspondingly divided into multiple exostoses types I, 2 and 3. Faciocutaneoskeletal syndrome See: Costello syndrome Familial adenomatous polyposis (OMIM Nos. 175100, 135290) Familial adenomatous polyposis (FAP) overlaps with the Gardner and Turcot syndromes. All of these conditions cause the development of numerous colonic polyps. Unless a patient has a colectomy, the risk of colon cancer is essentially 100%. FAP, Gardner and Turcot syndromes are all inherited in an autosomal dominant manner and are caused by mutations in the APC (adenomatous polyposis coli) gene which is in chromosome 5q21]q22. In FAP there are variable extracolonic problems including polyps of the stomach and duodenum, osteomas, soft tissue and desmoid tumors. Aside from colonic polyposis, there are osteomas and soft tissue tumors in Gardner syndrome and tumors of the central nervous system in Turcot syndrome. Familial chordoma See: Chordoma, familial Familial intestinal carcinoid See: Carcinoid, familial intestinal Familial medullary thyroid carcinoma See: Multiple endocrine neoplasia type 2 Fanconi anemia (OMIM Nos. 227650, 227645, 227646, 300514, 300515, 600185, 600901, 603467, 605724, 607139, 609053, 609054) Fanconi anemia carries an elevated risk of hematologic malignancies and solid tumors. The hematologic malignancies are most often acute myeloid leukemia and myelodysplastic syndrome. The solid tumors tend to be squamous cell carcinomas of the head and neck, skin and gastrointestinal and genital tracts. Hepatic adenoma and, especially after anabolic steroid therapy, hepatocellular carcinoma are added risks. There are numerous complementation groups in Fanconi anemia. The complementation group A gene is at 16q24.3; complementation group B at Xp22.31; complementation group C at 9q22.3; complementation group D1 (BRCA2) at 13q12.3; complementation group D2 at 3p25.3; complementation group E at 6p22]p21; complementation group F at 11p15; complementation group G at 9p13; complementation group J at 17q22; complementation
group L at 2p16.1, and complementation group M at 14q21.3. Fanconi anemia is inherited in an autosomal recessive fashion except in the case of complementation B mutations which are transmitted as X-linked traits. FCS syndrome See: Costello syndrome Gardner syndrome See: Familial adenomatous polyposis Gastric cancer, hereditary diffuse (OMIM Nos. 137215, 192090) This autosomal dominant disorder predisposes to diffuse adenocarcinoma of the stomach. Women with the disorder also have an elevated risk for lobular breast cancer. Mutations in the CDH1 (epithelial-cadherin) gene in 16q22.1 cause the disorder. Gastroesophageal reflux disease (GERD) See: Barrett esophagus Gingival multiple hamartoma syndrome See: Cowden syndrome Gorlin syndrome See: Nevoid basal cell carcinoma syndrome Haddad syndrome See: Central hypoventilation syndrome, congenital Hereditary diffuse gastric cancer See: Gastric cancer, hereditary diffuse Hereditary leiomyomatosis and renal cell cancer See: Leiomyomatosis and renal cell carcinoma, hereditary Hereditary multiple exostoses See: Exostoses, multiple Hereditary non-polyposis colon cancer (OMIM Nos. 120435, 120436, 158320, 600259, 600258, 600678) Hereditary non-polyposis colon cancer (HNPCC) is an autosomal dominant disorder and one that is too narrowly named. HNPCC does carry a high risk of cancer in the colon, particularly the proximal colon. But HNPCC is associated with other significant cancer risks as well. Women with HNPCC face a high risk of endometrial cancer which may precede the diagnosis of colon cancer. In HNPCC there is also an elevated risk of ovarian cancer, gastric adenocarcinoma and duodenal and jejunal cancer, transitional carcinoma of the renal pelvis and ureter, glioblastoma and other tumors. HNPCC is associated with mutations of four genes in the mismatch repair pathway: MLH1 in 3p21.3, MSH2 in 2p22p21, MSH6 in 2p16 and PMS2 in 7p22. Hodgkin lymphoma (OMIM No. 236000) The existence of family aggregations of Hodgkin disease suggests that there may be a hereditary susceptibility to the disease, perhaps inherited in an autosomal recessive manner, but this is not yet certain. Juvenile polyposis syndrome (OMIM Nos. 174900, 600993, 601299) Juvenile polyposis syndrome is an autosomal dominant disorder associated with hamartomatous (‘juvenile’) polyps in the intestinal tract and an increased risk of their transformation into cancer of the gastrointestinal tract, particularly the colon. Juvenile polyposis syndrome is caused by mutations in the SMAD4 (mothers against decapentaplegic homolog 4) gene in chromosome band 18q21.1 or the BMPR1A ( bone morphogenetic protein receptor type IA) gene in 10q22.3. Keratosis palmaris et plantaris with esophageal cancer See: Tylosis with esophageal cancer
LAMB syndrome See: Carney complex type 1 and 2 Leiomyomatosis and renal cell cancer, hereditary
(OMIM Nos. 138650, 605839) Hereditary leiomyomatosis and renal cell cancer is an autosomal dominant disorder. While both sexes are subject to cutaneous leiomyomata and renal cell carcinomas, women also have uterine leiomyomata (fibroids). The disorder is due to mutation of the FH gene in chromosome band 1q42.1 encoding fumarate hydratase, an enzyme in the tricarboxylic acid cycle. Li-Fraumeni syndrome (OMIM No. 151623) This autosomal dominant syndrome predisposes to a remarkable range of cancers and nothing else. The syndrome was first characterized by the familial aggregation of six types of neoplasms: soft-tissue sarcoma, osteosarcoma, breast cancer, brain tumors, adrenal cortical tumors and acute leukemia. There also appears to be an excess risk of melanoma; cancer of the stomach, colon, pancreas, and esophagus; and gonadal germ cell tumors. The syndrome is due to a mutation of the TP53 gene in chromosome band 17p13.1 encoding the cellular tumor antigenic p53 or, less often, a mutation of CHEK2 (checkpoint kinase 2) in chromosome band 22q12.1 encoding serine/theonine-protein kinase Chk 2. Lymphoproliferative disease, X-linked See: X-linked lymphoproliferative disease Macrocephaly, multiple lipomas, hemangiomata (OMIM No. 153480) This overgrowth syndrome carries a number of neoplastic risks including multiple lipomas, hemangiomas, hamartomatous intestinal polyps, lymphangiomas, meningioma and thyroid follicular cell tumor. The syndrome is due to mutation of the PTEN gene in chromosome subband 10q23.3 and is allelic with Cowden disease. Melanoma-astrocytoma syndrome (OMIM No.155755) This singular syndrome is associated with malignancies in two sites – the skin and brain – and with no other known features. The skin tumor is consistently a melanoma, while the nervous system tumor may or may not be an astrocytoma. Affected individuals develop melanoma or neural system tumor, or both. The syndrome is transmitted in an autosomal dominant manner and appears due to mutation of the CDKN2A gene in chromosome band 9p21. CDKN2A, also known as p16, is a suppressor for multiple types of tumors. A number of CDKN2A mutations have been identified in malignant melanoma as well as in the familial atypical multiple mole melanoma (FAMMM) syndrome and in the melanoma-pancreatic cancer syndrome. Melanoma-pancreatic cancer syndrome (OMIM No. 606719) A more explicit name for this syndrome is familial atypical multiple mole melanoma (FAMMM)-pancreatic carcinoma syndrome. It is eerily similar to the melanomaastrocytoma syndrome in its association with two types of malignancy, one being melanoma, and its genetic causation by mutation of the CDKN2A gene in 9p21. The FAMMM syndrome, the melanoma-astrocytoma syndrome, the melanoma-pancreatic cancer syndrome, and a significant proportion of melanoma cases may all be variations on one mutant gene theme. MEN1 See: Multiple endocrine neoplasia type 1 MEN2 See: Multiple endocrine neoplasia type 2
Cytogenet Genome Res 118:222–228 (2007
225
Multiple endocrine neoplasia type 1 (OMIM No. 131100) Multiple endocrine neoplasia type 1 (MEN1) is a remarkable autosomal dominant disorder that predisposes to over 20 types of benign endocrine and non-endocrine tumors. The endocrine tumors may arise in various sites including the parathyroids, anterior pituitary and the gastro-entero-pancreatic tract. The non-endocrine tumors in MEN1 include angiofibromas, collagenomas, lipomas, meningiomas, ependymomas and leimyomas. The disease is caused by mutation of the MEN1 gene in 11q13. Multiple endocrine neoplasia type 2 (OMIM No. 171400) Multiple endocrine neoplasia type 2 (MEN2) carries a high risk of medullary thyroid carcinoma, and in certain subtypes of MEN2 there is also an increased risk of pheochromocytoma and adrenal adenoma. MEN2 is an autosomal dominant disorder due to mutation of the RET gene which is located in chromosome subband 10q11.2. Multiple exostoses See: Exostoses, multiple Multiple osteochondromatosis See: Exostoses, multiple NAME syndrome See: Carney complex type 1 and 2 Neutropenia, congenital (OMIM Nos. 130130, 202700) Severe congenital neutropenia is an autosomal dominant disorder associated with an elevated risk of myelodysplasia and acute myeloid leukemia. The ELA2 (leukocyte elastase) gene responsible for congenital neutropenia maps to 19p13.3. Cyclic neutropenia is also an ELA2-related disorder but is not associated with a known elevated risk of malignancy. Neurofibromatosis 1 (OMIM No. 162200) Neurofibromatosis type 1 (NF1) is an autosomal dominant disorder with multiple dermal neurofibromas and a predisposition to the development of gliomas of the optic nerve and brain and malignant tumors of the nerve sheath. Leukemia is more common in children with NF1 than in the general population. This applies particularly to juvenile CML (chronic myelogenous leukemia) and MDS (myelodysplastic syndrome). NF1 is caused by mutation of the NF1 gene in chromosome band 17q11.2. The NF1 gene encodes the protein neurofibromin whose function is not fully understood. Neurofibromatosis 2 (OMIM No. 10100) Neurofibromatosis type 2 (NF2) is an autosomal dominant disorder characterized by bilateral acoustic neuromas (vestibular schwannomas). Individuals with NF2 may also develop schwannomas of other nerves, meningiomas and, upon occasion, ependymomas and astrocytomas. NF2 is due to mutation of the NF2 gene in chromosome band 22q12.2 encoding the protein merlin (Moezin-Ezrin-Radixin-Like proteIN). Nevoid basal cell carcinoma syndrome (OMIM Nos. 109400, 601309) As the name indicates, this syndrome predisposes to basal cell carcinoma. In addition, it predisposes to cardiac and ovarian fibromas and medulloblastoma. It is an autosomal dominant disorder caused by mutation of the PTCH gene in 9q22.3. NF1 See: Neurofibromatosis 1 NF2 See: Neurofibromatosis 2 Nijmegen breakage syndrome (OMIM Nos. 251260, 602667) This rare autosomal recessive disorder carries a high risk of malignancy, especially B-cell lymphoma. The
226
Cytogenet Genome Res 118:222–228 (2007
NBN gene responsible for the disorder is in chromosome band 8q21. Non-polyposis colon cancer, hereditary See: Hereditary non-polyposis colon cancer Oculocutaneous albinism type 1 (OMIM Nos. 203100, 606933) This form of albinism predisposes to premalignant solar keratoses and basal cell and squamous cell carcinomas of the skin. An autosomal recessive disorder, it is due to tyrosinase deficiency resulting from mutation of the TYR gene in chromosome region 11q14]q21. Ondine’s curse See: Central hypoventilation syndrome, congenital Osteochondromatosis, multiple See: Exostoses, multiple Palmoplantar keratosis See: Tylosis with esophageal cancer Peutz-Jeghers syndrome (OMIM Nos. 175200, 602216) The Peutz-Jeghers syndrome is characterized by gastrointestinal polyps which do not become malignant. PeutzJeghers syndrome also raises the risk of benign tumors in the ovaries and testes. Women with the syndrome may have an aggressive malignancy called adenoma malignum of the cervix. Peutz-Jeghers syndrome is inherited in an autosomal dominant manner and is due to mutation of the STK11 gene in chromosome band 19p13.3. Polyposis coli, juvenile See: Juvenile polyposis syndrome Porphyria, acute intermittent (OMIM Nos. 176000, 609806) Acute intermittent porphyria is an autosomal dominant disorder that carries an elevated risk of hepatocellular carcinoma. Mutation of the HMBS gene in chromosome band 11q23.3 is responsible for this form of porphyria. HMBS encodes the enzyme hydroxymethylbilane synthase. There is also a predisposition to hepatocellular carcinoma in porphyria cutanea tarda. Porphyria cutanea tarda (OMIM No. 176100) This form of porphyria is an autosomal dominant disorder characterized by a light-sensitive dermatitis and the excretion of large amounts of uroporphyrin in the urine. There is a predisposition to hepatocellular carcinoma. Porphyria cutanea tarda is due to mutation of the UROD gene in 1p34 encoding the enzyme uroporphyrinogen decarboxylase. Another form of porphyria that also predisposes to hepatocellular carcinoma is acute intermittent porphyria. Protease inhibitor 1 See: Alpha-1-antitrypsin deficiency. PTEN hamartoma tumor syndrome See Cowden syndrome Retinoblastoma (OMIM No. 180200) This malignant tumor of the retina is due to mutation of the RB1 gene in 13q13.1]q14.2. Hereditary retinoblastoma is caused by a germline mutation of RB1 and describes an autosomal dominant pattern of inheritance. Individuals with hereditary retinoblastoma are also at risk for pineolomas and second primary tumors including osteosarcoma, soft tissue sarcomas, and melanomas. Rothmund-Thomson syndrome (OMIM Nos. 268400, 603780) This autosomal recessive disorder carries an in-
creased risk of bone cancer (osteosarcoma) and skin cancer (basal cell and squamous cell carcinoma). The syndrome is due to mutation of the RECQL4 gene in 8q24.3, a member of the RecQ family of DNA helicases. Mutations in other RecQ helicases cause Bloom syndrome and Werner syndrome. Thyroid carcinoma, familial medullary See: Multiple endocrine neoplasia type 2 Tuberous sclerosis (OMIM Nos. 191100, 191092) Tuberous sclerosis is an autosomal dominant disorder that carries a number of neoplastic risks. These neoplastic risks are remarkably varied. They include renal angiomyolipoma (common in childhood); renal carcinoma (in adults); cardiac rhabdomyoma (often congenital); brain tumors (mainly subependymal giant cell astrocytoma or ependymoma), hamartomas (in liver, spleen, and elsewhere), and pulmonary lymphangiomyomatosis. Two genes, TSC1 and TSC2, are individually capable of causing tuberous sclerosis. TSC1 is in chromosome band 9q34 and encodes a protein called hamartin, while TSC2 is in chromosome band 16p13.3 and encodes a protein called tuberin. Although hamartin and tuberin are known to cohybridize in vivo, TSC1 mutations cause a milder form of tuberous sclerosis than do TSC2 mutations. Turcot syndrome See: Familial adenomatous polyposis. Tylosis with esophageal cancer (OMIM No. 148500) This disorder is also known as palmoplantar keratosis and as keratosis palmaris et plantaris with esophageal cancer. It is characterized by thickening of the palms and soles and a high risk of squamous cell carcinoma of the esophagus. The condition is inherited in an autosomal dominant manner and is due to mutation of the TOC gene in chromosome region 17q25. Tyrosinemia type 1 (OMIM No. 276700) Untreated tyrosinemia carries a high risk of hepatocellular carcinoma. Tyrosinemia type 1 is an autosomal recessive disorder due to deficiency of the enzyme fumarylacetoacetate hydrolase (FAH) encoded by the gene FAH in chromosome region 15q23]q25. Von Hippel-Lindau syndrome (OMIM Nos. 193300, 608537) Von Hippel-Lindau syndrome predisposes to a number of tumors including renal cell carcinoma, pheochromocytoma, and hemangioblastomas of the brain, retina and spinal cord. An autosomal dominant disorder, it is due to mutation of the VHL gene in 3p26]p25. Von Recklinghausen disease See: Neurofibromatosis 1 Wermer syndrome See: Multiple endocrine neoplasia type 1 Werner syndrome (OMIM Nos. 277700, 604611) This autosomal recessive condition is characterized by premature aging and a predisposition to cancer, particularly sarcomas and other unusual types of cancer including soft tissue sarcoma, osteosarcoma, acral lentiginous melanoma, and thyroid carcinoma. Werner syndrome results from mutation of the WRN gene which is in chromosome region 8p12]p11.2 and encodes a member of the RecQ family of DNA helicases. Mutations in other RecQ helicases cause Bloom syndrome and Rothmund-Thomson syndrome.
Wilms tumor 1 (OMIM No.194070) Wilms tumor of the kidney may occur as an isolated finding or as part of a syndrome (e.g., the WAGR syndrome: Wilms tumor-aniridiagenitourinary abnormalities-mental retardation; or Beckwith-Wiedemann syndrome). As an isolated finding, the predisposition to Wilms tumor is inherited as an autosomal dominant trait. The WT1 gene is in chromosome band 11p13. Wiskott-Aldrich syndrome (OMIM Nos. 300392, 301000) The Wiskott-Aldrich syndrome (WAS) is an X-linked recessive disorder that carries a high risk of lymphoma. The WAS gene is in Xp11.23–p11.22. WAS is part of a spectrum that includes X-linked thrombocytopenia which is also due to mutation of the WAS gene. X-linked lymphoproliferative disease (OMIM Nos. 300490, 308240) This disease is an X-linked recessive condition caused by mutation of the SH2D1A gene in chromosome band Xq25. It predisposes to non-Hodgkin lymphomas, typically high-grade B cell lymphomas. Xeroderma pigmentosum (OMIM Nos. 278700, 133510, 278720, 278730, 278740, 278760, 278780) Xeroderma pigmentosum (XP), an autosomal recessive disorder, carries a very high risk of skin and eye tumors. The skin tumors are basal cell and squamous cell carcinomas and melanomas. The eye tumors are epitheliomas, squamous cell carcinomas and melanomas. The genetics of XP are complex. XP is associated with mutation of XPA in 9q22.3, ERCC3 in 2q21, XPC in 3p25, ERCC2 in 19q13.2]q13.3, DDB2 in 11p12]p11, ERCC4 in 16p13.3]p13.13, ERCC5 in 13q33, and POLH in 6p21.1]p12. Zollinger-Ellison syndrome See: Multiple endocrine neoplasia type 1
Discussion
This catalog of familial cancer syndromes is incomplete. A number of other entities might (or might not) be added to it. For instance, one could consider adding Alagille syndrome, Bruton-type agammaglobulinemia, Cockayne syndrome, diaphyseal medullary stenosis with malignant fibrous histiocytoma, Dubowitz syndrome, dyskeratosis congenita, familial paraganglioma, familial prostate cancer, familial testicular cancer, etc. The present catalog is just a start. Modes of inheritance of family cancer syndromes All four Mendelian modes of inheritance are represented in this catalog of familial cancer syndromes: autosomal dominant, autosomal recessive, X-linked dominant and Xlinked recessive. The predominant mode of inheritance in this sample of family cancer syndromes is autosomal dominant accounting for more than half of the syndromes. The next most common mode of inheritance among these family cancer syndromes is autosomal recessive. About a quarter of the syndromes are autosomal recessive traits. X-linked inheritance is represented, for examples, by the Wiskott-Aldrich syndrome and X-linked lymphoprolifera-
Cytogenet Genome Res 118:222–228 (2007
227
tive syndrome. The chromosome breakage/cancer syndromes – ataxia-telangiectasia, Bloom syndrome, Fanconi anemia and Nijmegen breakage syndrome – are all heritable as autosomal recessive traits. Fanconi anemia is unusual in being inherited in two different ways – as an autosomal recessive trait and as an X-linked trait. Chromosomes involved in family cancer syndromes This catalog of family cancer syndromes involves some 77 genes. They are on chromosomes 1–20, 22 and the X. Thus, all of the human chromosomes are implicated here except number 21 and the Y. Some chromosome bands contain two or more genes for family cancer syndromes in this sample. These chromosome bands which may possibly be ‘cancer-prone’ (or random sites of coincidence) including the following: 2p16: Carney complex, Fanconi anemia, and hereditary non-polyposis colon cancer 3p25: Von Hippel-Lindau syndrome, Fanconi anemia and xeroderma pigmentosum 6p21: Fanconi anemia and xeroderma pigmentosum 9q22.3: Nevoid basal cell carcinoma, Fanconi anemia and xeroderma pigmentosum 10q11.2: Central hypoventilation syndrome and MEN2 10q23.31: Cowden syndrome and macrocephaly, lipomas, hemangiomata (allelic) 11p15: Beckwith-Wiedemann syndrome, Costello syndrome and Fanconi anemia 11p13: Central hypoventilation syndrome and Wilms tumor 11p12]p11: Multiple exostoses and xeroderma pigmentosum 13q14: Barrett esophagus and retinoblastoma 19p13.3: Congenital neutropenia and Peutz-Jeghers syndrome Genes involved in family cancer syndromes A gene that is mutated and causes a family cancer syndrome may not necessarily be the same gene as causes the tumor. The historical sequence of events surrounding the clinical syndrome and the tumor may matter. Different family cancer syndromes have followed different scenarios. For example, ataxia-telangiectasia was first recognized because of its salient non-neoplastic features. Later the associations of ataxia-telangiectasia with lymphoma and leukemia were discovered. Although mutation of the ATM gene has been found to cause ataxia-telangiectasia, it is still not clear whether mutation of the ATM gene, in fact, causes the malignancies and if so, how it does so. By contrast, in retinoblastoma the cancer and its recurrence in families led to the syndrome. Identifying the mutations in the RB gene might therefore logically be expected to be close to (or identical with) the genetic changes underlying the cancer. Similarly, familial adenomatous polyposis (FAP) was first detected through the familial occurrence of polyps and cancer of the colon. The finding that mutations in the APC gene underlie FAP might similarly be expected
228
Cytogenet Genome Res 118:222–228 (2007
to place us close to the genetic changes underlying the colon cancer itself. Certain types of genes tend to be involved in family cancer syndromes. For example, the BLM gene involved in Bloom syndrome is a member of the RecQ family of DNA helicases. Mutations in other RecQ helicase genes cause Rothmund-Thomson syndrome and Werner syndrome. Interestingly, several of the overgrowth syndromes such as Beckwith-Wiedemann syndrome and Bannayan-RileyRuvalcaba syndrome (macrocephaly, multiple lipomas, hemangiomata) are among the family cancer syndromes. It would seem that excessive somatic overgrowth early in development may increase the chance of malignancy. Familial cancer syndrome recognition Family cancer syndromes are in a state of flux. Some disorders such as multiple exostoses are being divided into multiple types on the basis of new genetic information while other disorders such as familial adematous polyposis, Gardner syndrome, and Turcot syndrome, once thought to be separate disorders, have been consolidated as APC-associated polyposis. The process of fission and fusion of familial cancer syndromes is part and parcel of the process that has been going on in clinical genetics as a whole since the 1950s. It is symptomatic of a field of investigation that is undergoing rapid growth and development, as is currently the case with family cancer syndromes. A key concern is whether family cancer syndromes are being regularly spotted in clinical practice. A study by Tyler and Snyder (2006) bears on this issue. In a study of the ambulatory records of 734 patients at the Cleveland Clinic they found that the presence or absence of a family history of cancer was documented in almost all (98%) of cases. However, in over two-thirds of patients with a positive family history of cancer there was insufficient information to assess the risk. Primary care clinicians must obtain enough information from their patients to identify an increased risk of cancer. The taking of an adequate family history of cancer is essential to spotting a family cancer syndrome. Acknowledgement I wish to thank my intrafamilial colleague Barbara K. Hecht for all kinds of assistance.
References
Lindor NM, Greene MH: The concise handbook of family cancer syndromes. Mayo Familial Cancer Program. J Natl Cancer Inst 90: 1039–1071 (1998). Tyler CV Jr, Snyder CW: Cancer risk assessment: examining the family physician’s role. J Am Board Fam Med 19:468-477 (2006).
Technical aspects
In situ hybridization of DNA to interphase nuclei requires several sequential procedures (Pinkel et al., 1986; Tibiletti et al., 1999). First of all, pre-hybridization procedures are performed using RNase to strip native RNA, which can interfere with hybridization. Occasionally, the retention of cytoplasm around nuclei or chromosomes may interfere with optimal hybridization and proteolysis treatments such as proteinase K or pepsin, or treatments with detergents to permeabilize the cells may be required. When the target of FISH are interphase nuclei from paraffin-embedded sections the pre-hybridization procedures have to be appropriate: the paraffin must be completely removed with successive washes of sections in xylol and the proteolysis treatments have to be more aggressive in order to remove autofluorescence and to let the probe penetrate and interact with the target DNA (Thompson et al., 1994). Due to these prerequisites IFISH on archival material requires time-consuming and complex protocols with difficult optimization. Before hybridization the target and probe DNAs are denatured usually using formamide at high temperature. A variety of probes may be prepared for the in situ hybridization procedures (see below), labeled with digoxigenin, biotin or directly with different fluorochromes by the use of nick translation or random priming procedures. With both protocols the labeling results in the incorporation of one modified nucleotide at every 20–25th position in the newly synthesized DNA probe. Very recently Peptide Nucleic Acid (PNA) probes have become commercially available in which the sugar phosphate backbone of DNA has been replaced by a synthetic peptide backbone, keeping the distances between bases exactly the same as in DNA. The PNA probes may be labeled with fluorescein, Texas-red or other fluorochromes (Nielsen and Egholm, 1999). The probe-DNA and the target-DNA reallocation (hybridization step) occurs some 25 ° C below the melting point of the corresponding native duplex and stringent conditions will favor accurate base pairing. Usually the hybridization time is overnight, but depending on the type of probes may require two or more days. After hybridization subsequent washes may be performed in order to remove the background. Different stringency conditions are required for different types of probes. An essential feature of FISH is the requirement of high quality DNA probes to ensure efficient labeling. With respect to this, cloned DNA must be purified and cleaned and the labeling must be performed accurately in order to obtain fragments of DNA with an average length of 300 base pairs (large fragments would interfere with proper probe penetration and also produce high background signals on the slide). For this reason, nick translation is a particularly suitable method for probe labeling. Signal detection, when the probes are biotin or digoxigenin labeled, can be achieved using different combinations of fluorochrome-conjugated antibodies with a sandwich amplification technique. Usually the more employed fluorochromes are FITC (green), or Texas red, TRITC, CY-3
230
Cytogenet Genome Res 118:229–236 (2007)
(red). More recently other fluorochromes such as Aqua (blue), and yellow have became available. Counterstaining may be performed with propidium iodide when the spot is green or with 4ⴕ,6-diamidino-2-phenylindole (DAPI) when red or dual colors (red and green) or multicolor (red, green, aqua and yellow) probes are hybridized. A dual color detection protocol can be used to simultaneously identify two probes. This approach employs a probe labeled with biotin, and the other labeled with digoxigenin. The probes are visualised with green and red fluorochromes. Probes directly labeled with fluorochromes (red, green, aqua and yellow) are preferentially used in order to minimize background and to save time and costs. Using directly labeled probes the signals can be directly visualized at the standard fluorescence microscope using appropriate filters. Many of the basic applications of in situ hybridization can be performed with a standard fluorescence microscope equipped with a suitable filter set. Dual color FISH may be visualized directly using double or triple band filters. FISH images may be archived using a CCD camera connected to the microscope and commercially available specific FISH software. Probes
Different types of probes are available for hybridization purposes (Fig. 1), and they can be broadly subdivided into two groups: 1) probes with repeated sequences and 2) probes with unique sequences. The first group of probes (including alphoid centromere repeats, heterochromatic repeats, telomeric repeats and -satellite repeats), are useful for tagging a particular chromosome of interest, for rapid sexing or scoring for chromosome aneuploidy. The second group of probes characterized by unique sequences include whole chromosome paintings, partial chromosome paintings (that cover large chromosome regions), YACs, BACs, PACs and cosmids that cover regions of 200–2000, 300, 130–150, and 30–50 kilobases in length, respectively. This group of probes includes BACs and PACs located at subtelomeres that may be used to identify short and long arm telomeres of specific chromosomes. The unique sequence probes avoid the problem of background hybridization because they lack interspersed repeats, which are likely to recognize sites for hybridization on many other chromosomes. However such hybridization can be suppressed in a pre-hybridization reassociation with either total human DNA or DNA used as a competitor. This approach gives the possibility to map a specific DNA region to any given chromosome or to identify specific small or large regions involved in chromosome anomalies. Different kinds of ready-to-use probes, and in particular, centromeric, telomeric, painting and locus specific probes are now commercially available. All the other types of cloned sequences useful as probes are available through human genome databases, i.e. http// bioserver.uniba.it/fish/Cytogenetics/welcome.html; http:// sgiweb.ncbi.nlm.nih.gov:80/Zjing/yac.html.
Centromeric
(alpha satellite se uences)
Whole chromosome painting (unique sequences of the whole chromosome)
Locus specific
(unique sequences at specific locus)
Fig. 2. (a) IFISH using centromeric FITC-labeled probe on a histological section of borderline ovarian tumor. (b) Break-apart IFISH using BCL6 specific probe on histological section of lymphoma: the overlapping and sectioning of nuclei can influence the evaluation of spots.
Sub-Telomeric
(unique sequences mapping at subtelomere)
Fig. 1. Types of probes.
Targets
IFISH may be performed on different types of nuclei such as imprinted nuclei (IM), obtained by touching of tumor samples directly on glass slides (Kontogeorgos et al., 1999), nuclei obtained from conventional cytogenetic procedures (PC nuclei), by harvesting of tumor cell lines, short term cultures and of direct preparations of tumor samples (Tibiletti et al., 1999, 2001); nuclei obtained by tumor cell isolation from frozen tissues (FN nuclei) and finally, nuclei processed directly from paraffin sections (PE nuclei), or extracted from paraffin-embedded sections (EX nuclei) (Thompson et al., 1994; Liehr et al., 1995). Although PC nuclei are the most suitable nuclei for IFISH experiments because of their homogeneous distribution and adequate DNA fixation, the requirement for resources of fresh samples is a limiting factor. IFISH can be applied to paraffin-embedded, formalinfixed tissue sections and to tissue arrays, enabling retrospective analysis. It is conceptually intriguing because it allows the simultaneous assessment of chromosomal aberrations, cellular phenotype, and tissue morphology. Approaches using PE and EX are often limited by necrosis as well as the presence of overlapping cells and sectioning procedures (Figs. 2, 3).
Freshly fixed imprints represent excellent material for FISH studies because they can be easily obtained and provide a large number of intact cells with good morphology and accessible chromatin. Nevertheless, the imprinted nuclei are difficult to keep archived for an extended time. Sensitivity and specificity
IFISH results are obviously influenced by hybridization efficiency and generally, experiments showing spots in less than 90% of nuclei are not suitable for the definition of chromosome aberrations. The spot counts should be performed according to the I.S.C.N. recommendations (ISCN 2005) on IFISH experiments with clear spots detectable either with single band filters or dual and triple band filters. Since IFISH is an indirect cytogenetic procedure, a control of probe mapping on a normal metaphase should be performed in order to avoid mistakes in the chromosome assessment evaluation. The evidence of chromosome mapping of the probes is particularly useful when specific IFISH strategies are used for cytogenetic rearrangement detection. In addition, to obtain adequate sensitivity and specificity for the IFISH test, controls on normal nuclei may be performed in order to establish adequate cut-offs for trisomies and monosomies of whole chromosomes or deletions and duplications of specific cytogenetic regions. The cutoff values depend on both the types of probes and on the types of target nuclei (Tibiletti, 2004), and for this reason the use of the same threshold level for the detection of different chromosome abnormalities is not recommended. The omission of the cut-off point for specific types of nuclei
Cytogenet Genome Res 118:229–236 (2007)
231
Fig. 3. Technical problems of hybridized nuclei: (a, b) IFISH on histological sections showing the fragility of nuclei due to the presence of necrosis that prevents the spot evaluation. (c) Autofluorescence not completely removed on histological section.
introduces a bias in the evaluation of chromosomal abnormalities, particularly under- and over-estimation of monosomies. IFISH studies with more than 10% of cells lacking fluorescent spots are usually not acceptable for evaluation of chromosomal abnormalities because they lead to overestimation of monosomies and underestimation of trisomies. IFISH on PE and EX nuclei need particular attention: the sectioning of the nuclei may give an underestimation of trisomies and overestimation of monosomies if the dimensions of the analyzable nuclei are not correctly evaluated. Concerning this, the choice of the panel of control nuclei may be made on the basis of nuclear dimensions, and archival sections should be adequately cut. Parallel problems arise when the shapes of the control nuclei are not comparable to those of the analyzable nuclei. For example IFISH on stromal nuclei is an inadequate control for evaluation of IFISH results of epithelial nuclei because the differences in shape, dimension and hybridization efficiencies introduce a bias in monosomy and trisomy detection (Tibiletti, 2004). The evaluation of chromosome aneuploidies, duplications and deletions using accurate controls is particularly relevant for the analysis of chromosomes obtained from tumor cells. The majority of tumors are heterogeneous and contain different cell populations with different chromosome complements and the lack of accurate IFISH controls can result in the misdiagnosis of some abnormalities.
232
Cytogenet Genome Res 118:229–236 (2007)
Strategy for identification of specific chromosome abnormalities
The value of interphase cytogenetic analyses of nuclei was greatly enhanced by the availability of several types of commercial probes. These probes directly labeled with different fluorochromes provide several advantages including specific high-intensity signals without amplification and detection steps, low background and rapid easy-to-use assays. Since IFISH is an indirect cytogenetic approach, specific chromosome aberrations may be identified using only appropriate molecular strategies. Centromere-specific probes are used for the enumeration of chromosome copy number changes, although these probes are certainly useful for monosomies or trisomies of a given chromosome, copy number changes affecting only chromosomal arms remain elusive. Interestingly, several locus- and gene-specific probes are now commercially available and innumerable structural chromosome abnormalities can be diagnosed on interphase nuclei. Detection of deletions/allelic losses, duplications/allelic gains, amplifications of specific chromosome regions or genes may be carried out on different types of target nuclei by the simultaneous use of a specific probe for the investigated region and a reference probe mapping on the same chromosome. Usually the investigative and the reference probes are labeled with different fluorochromes and visualized as dual color FISH using a dual or triple band filter. The reference probe employed in this type of experiment is a centromeric probe or a locus-specific probe presumably not implicated in the aberrations.
Fig. 4. IFISH application. Scheme showing a translocation between IGH and BCL2 genes using a dual color-single fusion IFISH strategy. Two large probes labeled respectively, in green and in red are located on one side of each of the two genetic breakpoints of IGH and BCL2 genes. When a specific translocation involving both genes is present one fusion signal (derivative chromosome), in addition to one green and one red signal (normal chromosomes), is observed.
Fig. 5. IFISH application. Scheme showing a translocation involving the BCL2 gene using a break-apart IFISH strategy. Two probes labeled respectively in green and in red hybridize on opposite sides of a breakpoint of the BCL2 gene. When a specific translocation involving BCL2 is present one fusion signal (normal chromosome), in addition to one green and one red signal (derivative chromosome), are observed. With this strategy only one of the two homologues involved in chromosome translocation may be identified.
Recently several probes designed for chromosome translocations have been provided by different companies. These probes include: Dual color single-fusion probe, the DNA probe hybridization targets are located on one side of each of two genetic breakpoints (Fig. 4). When the specific translocation is present a specific fusion signal (generally yellow due to green + red signals overlapping), in addition to single color signals (red and green) corresponding to the normal alleles are observed. This strategy is useful for detecting samples with a high percentage of cells possessing this translocation. Usually the size of probes in this case is about 300 kb. Dual color break-apart: two differently labeled probes (250–600 kb) hybridize to targets on opposite sides of a
breakpoint in one gene (Fig. 5). This probe is useful in cases where multiple translocation partners may be associated with known genetic breakpoints. Dual color dual-fusion: differently labeled large probes (300 kb–1 Mb), span two breakpoints on different chromosomes (Fig. 6). This strategy offers advantages in detecting low numbers of nuclei possessing a simple balanced translocation. All these strategies may be applied to different nuclei, nevertheless the PC nuclei are the better targets for interphase chromosome translocation detection for the following two reasons: 1) low cut-off level of cells showing fusion signals due to few random co-localizations of probe signals in a normal nucleus and 2) no bias in the choice of hybrid-
Cytogenet Genome Res 118:229–236 (2007)
233
Fig. 6. IFISH application. Scheme showing a translocation between IGH and BCL2 genes using a dual color-dual fusion IFISH strategy. Two large probes labeled respectively in green and in red span breakpoints of IGH and BCL2 genes. When a specific translocation involving both genes is present two fusion signals (derivative chromosomes), in addition to one green and one red signal (normal chromosomes), are observed.
Table 1. Correlation between IFISH results on paraffin embedded sections of nine malignant lymphomas using break-apart and dual color-dual fusion strategies. Two cases are normal and seven cases showed a BCL2 translocation. The break-apart strategy is more efficient for identifying nuclei showing the translocation.
Case
1 2 3 4 5 6 7 8 9
BCL2 break-apart
dual color-dual fusion
0/150 (0%) 0/215 (0%) 40/150 (27%) 1/200 (1%) 40/130 (31%) 30/150 (20%) 30/200 (15%) 30/160 (19%) 53/263 (20%)
0/400 (0%) 4/400 (1%) 14/200 (7%) 0/200 (0%) 5/306 (2%) 15/400 (4%) 4/104 (4%) 4/100 (4%) 7/127 (6%)
ized nuclei for the diagnosis. On the contrary PE sections showing overlapping and sectioning of nuclei may suffer from false negative results. Regarding this, our group performed several experiments in order to establish the sensitivities of different IFISH strategies for translocation detection. When the break apart and dual color fusion strategies for the same gene translocation were compared in a panel of PC and PE nuclei from the same neoplastic samples, the break-apart strategy was the more sensitive test to detect gene translocation on histological archive sections (Table 1). From a technical point of view, the fusion signals are sometimes not so clear and instead of a yellow fusion signal
234
Cytogenet Genome Res 118:229–236 (2007)
due to green and red fusion, two closely located red and green signals are visible. This characteristic is probably due to DNA conservation and treatment is more frequent in nuclei from exfoliative cytological specimens. With this in mind, a control IFISH on normal and pathologic nuclei of the same origin of the investigated nuclei is suggested when routine testing will be performed. Applications
The great variety of available probes and the possibility of using combined fluorochromes makes IFISH a valued method for the detection of genetic aberrations in solid tumors and in hematopoietic disorders. A large number of IFISH applications are now available for diagnostic and prognostic purposes in tumor pathology. Recent data have demonstrated that for different types of solid tumors the presence of a specific genetic defect is strictly related to therapy response: the presence of ERBB2 (alias HER2/neu) gene (cell surface-receptor) amplification in breast cancers (Fig. 7) correlates with the response to Trastuzumab monoclonal antibody therapy (Bast et al., 2001; Di Leo et al., 2002; Fornier et al., 2002), an increasing copy number of EGFR (epidermal growth factor receptor), in colon and lung cancers confers a positive response to Cetuximab or Panitumumab (Moroni et al., 2005; Sequist et al., 2005). Because of this interesting positive relationship between genetic defects and therapy response, it is relevant to define the gene status in the cancers of patients which could benefit from new monoclonal therapies. Currently the gene status is determined by the assessment of protein levels using immunohistochemistry (IHC). However, IHC
has significant shortcomings, the most important of which is the loss of sensitivity resulting from antigenic alterations caused by standard fixation procedures. A second issue with IHC is the lack of consistent interpretation of results (Lebeau et al., 2001; Kobayashi et al., 2002; Hammond et al., 2003). New detection technologies using IFISH offer the option to determine the genetic status directly at the level of gene amplification instead of at the protein level. At the moment the definition of the ERBB2 gene status with IFISH in specific breast cancers represents a critical point for the clinician in selecting a therapy of choice. Likewise EGFR gene amplification defined by IFISH is beginning to represent the elective test to identify patients with colorectal and lung cancers who are likely to benefit from monoclonal antibodies (Moroni et al., 2005; Sequist et al., 2005). The relevance of cytogenetic findings for assessment of hematological disorders and lymphomas has been known for a long time (Look, 1998). Such analyses may be of immediate relevance to patient management. This is so when such investigations are performed as an aid to the diagnosis or classification of a hematological disorder or when cytogenetic analysis is expected to yield information of relevance to prognosis, treatment, or both. The relevance of cytogenetic data has been recently highlighted by the incorporation of specific genetic markers in the publication of the recent WHO classification of tumors of hematopoietic and lymphoid tissues (Jaffe et al., 2002). Hematological diseases and lymphomas are characterized by different and specific cytogenetic structural rearrangements; of these, the most frequent type are reciprocal translocations disrupting oncogenes that play a relevant role in the pathogenesis of disease. IFISH, using specific strategies as reported above, is the technology of choice for detecting the presence of a specific translocation independently of the presence of metaphases. Locus-specific IFISH techniques which are applicable to either fresh or paraffin-embedded tissues are demonstrated as the most sensitive method in contrast to routine PCR strategies to detect specific rearrangements in lymphomas and in other hematological disease (Paternoster et al., 2002). Chromosome anomalies also characterize 80–90% of solid tumors (Mitelman database http://cgap.nci.nih.gov/ chromosomes/Mitelman). Nevertheless, in the majority of solid cancers chromosome abnormalities are not as specific as in lymphomas and hematological diseases. Currently, information on more that 47,000 neoplasms are available (Mitelman, 1994) and for several solid tumors the anomalies more frequently observed and involved in tumor progression (Dal Cin and Van den Berghe, 1997; Gibas and Gibas, 1997; Hruban et al., 2000; Tibiletti et al., 2000, 2003) are well known. These data gave the idea to identify, for specific solid tumors, a panel of anomalies that may be investigated for both diagnostic and prognostic purposes. At present, commercial kits using multi-probe IFISH are available for detection of chromosome anomalies in cytological samples. An example of this approach is the detection in urine specimens of aneuploidies of chromosomes 3, 7 and 17 and allelic loss of the p16 gene, which more efficiently detects blad-
Fig. 7. IFISH application. IFISH with a dual color HER2/neu (ERBBZ) (red), and -satellite for chromosome 17 (green), probes on PE of breast cancer. The gene amplification is present.
der cancers. Like other solid tumors, bladder cancer cells are characterized by a high frequency of chromosome abnormalities and bladder cancer cells in particular readily exfoliate into urine. Recent studies (Junker et al., 1999; Ishiwata et al., 2001) have demonstrated that the IFISH analysis for specific abnormalities may be useful to screen urine or bladder washings for early tumor detection or recurrence. This approach may be particularly useful because routine cytology analysis shows low sensitivity for bladder cancer. The multi-color IFISH strategy may be applied to many other types of cell samples to more efficiently identify cancer cells. For this reason, cytogenetic studies on solid tumors and the identification of cytogenetic abnormalities involved in tumor progression should be stimulated. In conclusion, IFISH analysis is a versatile approach to investigate chromosome abnormalities at a cell-to-cell level and thus, is an effective technique to detect chromosome abnormalities in tumors. Because of the increasing evidence concerning the relationship between therapy and genetic defects, IFISH technology will become an essential method to improve diagnosis, prognosis and survival of cancer patients.
Cytogenet Genome Res 118:229–236 (2007)
235
References Bast CR, Ravdin P, Hayes DF, Bates S, Fritsche H, et al: 2000 update of recommendations for the use of tumor markers in breast and colorectal cancer: clinical practice guidelines of the American Society of Clinical Oncology. J Clin Oncol 196:1865–1878 (2001). Botti C, Pescatore B, Mottolese M, Sciarretta F, Greco C, et al: Incidence of chromosome 1 and 17 aneusomy in breast cancer and adjacent tissue: an interphase cytogenetic study. J Am Coll Surg 190:530–539 (2000). Dal Cin P, Van den Berghe H: Bone and soft tissue tumors. Cancer Genet Cytogenet 95: 59–66 (1997). Di Leo A, Dowsett M, Horten B, Penault-Llorca F: Current status of HER2 testing. Oncology 63: 25–32 (2002). Farabegoli F, Santini D, Maccarelli C, Taffurelli M, Marrano D, Baldini N: Clone heterogeneity in diploid and aneuploid breast carcinomas as detected by FISH. Cytometry 46: 50–56 (2001). Fornier M, Risio M, Van Poznak C, Seidman A: HER2 testing and correlation with efficacy of trastuzumab therapy. Oncology 16: 1340–1358 (2002). Gibas Z, Gibas L: Cytogenetics of bladder cancer. Cancer Genet Cytogenet 95: 108–115 (1997). Hammond MEH, Barker P, Taube S, Gutman S: Standard reference material for Her2 testing. Appl Immunohistochem Mol Morphol 112: 103–106 (2003). Hruban RH, Goggins M, Pearson J, Kern SE: Progression model for pancreatic cancer. Clin Cancer Res 6:2969–2972 (2000). ISCN (2005): An International System for Human Cytogenetic Nomenclature. Shaffer LG, Tommerup N (eds) (S Karger, Basel 2005). Ishiwata S, Takahahashi S, Homma Y, Tanaka Y, Kameyama S, et al: Noninvasive detection and prediction of bladder cancer by fluorescence in situ hybridization, analysis of exfoliated urothelial cells in voided urine. Urology 574: 811– 815 (2001). Jaffe ES, Harris NL, Stein H, Vardiman JW (eds): WHO Classification. Tumours of Haematopoietic and Lymphoid Tissues (IARC Press, Lyon 2002).
236
Junker K, Werner W, Mueller C, Ebert W, Schubert J, Claussen U: Interphase cytogenetic diagnosis of bladder cancer on cells from urine and bladder washing. Int J Oncol 142:309–313 (1999). Kobayashi M, Ooi A, Oda Y, Nakanishi I: Protein overexpression and gene amplification of c-erb-2 in breast carcinomas: a comparative study of immunohistochemistry and fluorescence in situ hybridization of formalin-fixed paraffin-embedded tissues. Hum Pathol 33:21– 28 (2002). Kontogeorgos G, Kapranos N, Orphanidis G, Rologis D, Kokka E: Molecular cytogenetics of chromosome 11 in pituitary adenomas: a comparison of fluorescence in situ hybridization and DNA ploidy study. Human Pathol 30: 1377– 1382 (1999). Lebeau A, Deimling D, Kaltz C, Sendelhofert A, Iff A, et al: HER2/neu analysis in archival tissue samples of human breast cancer: comparison of immunohistochemistry and fluorescence in situ hybridization. J Clin Oncol 19: 354–363 (2001). Liehr T, Grehel H, Rautenstraus B: FISH analyses of interphase nuclei extracted from paraffinembedded tissue. Trends Genet 11: 377–378 (1995). Look AT: Genes altered by chromosomal translocations in leukaemia and lymphomas, in Vogelstein B, Kinzler KV (eds): The Genetic Basis of Human Cancer (McGraw Hill, Columbus 1998). Mitelman F: Catalog of Chromosome Aberrations in Cancer. 7th Ed (Wiley-Liss, New York 1994). Moroni M, Veronese S, Benvenuti S, Marrapese G, Sartore-Bianchi A, et al: Gene copy number of epidermal growth factor receptor (EGFR) and clinical response to anti EGFR treatment in colorectal cancer: a cohort study. Lancet Oncol 6:279–86 (2005). Nielsen PE, Egholm M (eds): Peptide Nucleic Acids Protocols and Applications (Horizon Scientific Press, Norfolk 1999).
Cytogenet Genome Res 118:229–236 (2007)
Paternoster SF, Brockman SR, McClure RF, Remstein ED, Kurtin PJ, Dewald GW: A new method to extract nuclei from paraffin-embedded tissue to study lymphomas using interphase fluorescence in situ hybridization. Am J Pathol 160:1967–1972 (2002). Pinkel D, Straume T, Gray JW: Cytogenetic analysis using quantitative high-sensitivity fluorescence hybridization. Proc Natl Acad Sci USA 83:2934–2938 (1986). Sequist LV, Haber DA, Lynch TJ: Epidermal growth factor receptor mutations in non-small cell lung cancer: predicting clinical response to kinase inhibitors. Clin Cancer Res 11: 5668–5670 (2005). Thompson CT, LeBoit PE, Nederlof PM, Gray JW: Thick section fluorescence in situ hybridization on formalin-fixed paraffin-embedded archival tissue provides a histogenetic profile. Am J Pathol 144: 237–243 (1994). Tibiletti MG: Specificity of interphase fluorescence in situ hybridization for detection of chromosome aberrations in tumor pathology. Cancer Genet Cytogenet 155:143–148 (2004). Tibiletti MG, Bernasconi B, Dionigi A, Riva C: The applications of FISH in tumor pathology. Adv Clin Path 3:111–118 (1999). Tibiletti MG, Sessa F, Bernasconi B, Cerutti R, Brogli B, et al: A large 6q deletion is a common cytogenetic alteration in fibroadenomas pre-malignant lesions and carcinomas of the breast. Clin Cancer Res 6: 1422–1431 (2000). Tibiletti MG, Bernasconi B, Furlan D, Bressan P, Cerutti R, et al: Chromosome 6 abnormalities in ovarian surface epithelial tumors of borderline malignancy suggest a genetic continuum in the progression model of ovarian neoplasms. Clin Cancer Res 7: 3404–3409 (2001). Tibiletti MG, Bernasconi B, Taborelli M, Facco C, Riva C, et al: Genetic and cytogenetic observations among different types of ovarian tumors are compatible with progression model underlying ovarian tumorigenesis. Cancer Genet Cytogenet 146: 145–153 (2003).
Traditionally, recurrent types of clonal chromosome aberrations have been the focus of cancer research based on the assumption that the same types of cancer share the same pathways of progression and can be traced by identifying recurrent genetic alterations, such as the same karyotypic aberrations and their patterns of evolution. This approach has been successfully illustrated in some types of blood cancers and the continued search for signature types of aberrations has represented a major effort in cancer cytogenetics. In addition to revealing the mechanism of cancer formation by cloning various fusion genes, recurrent chromosome aberrations have played a significant role in both diagnostic and prognostic applications (Rowley, 1998, 2001). A primary example of this is chronic myelogenous leukemia or CML (Johansson et al., 2002). Initially linked to the Ph chromosome, the BCR/ABL fusion gene has been identified as a primary recurrent genetic event. Together, molecular and cytogenetic methods have been routinely used in clinical diagnosis. Recently, the successful story of the therapeutic results of Gleevec has been widely praised as an example of targeting a specific cancer pathway (Druker et al., 2001), and has raised a great deal of hope that this approach can be used when dealing with other types of cancer. On the other hand, chromosomal aberration patterns have become increasingly complicated in a given type of cancer. As greater numbers of tumor types and subtypes are examined by using more sensitive methods, major patterns have been diluted with increasing variations documented that are considered to be exceptions to the main patterns. This high level of variation has presented a major challenge both in basic research and in clinical applications (Losi et al., 2005; Heng, 2007a). These challenges, in particular, hamper the study of the clinical usage of chromosomal aberrations in the majority of solid tumors, where recurrent chromosomal aberrations have been difficult to find. Recently, recurrent types of fusion genes have been reported from a certain proportion of prostate cancer patients (Tomlins et al., 2005). As there are extensive karyotypic changes coupled with these fusion genes, the significance of these fusion genes in prostate cancer progression needs further investigation, as the genome context is fundamentally different with a variety of chromosomal compositions even when recurrent chromosomal aberrations or specific gene mutations are detected (Heng et al., 2006a, Heng et al., in preparation). It is also known that the evolutionary pattern of solid tumors and blood cancers are clearly different and it is likely that these fusion genes detected in solid tumors represent late products of cancer evolution (Heng et al., 2006b, Heng et al., unpublished data). In addition, there are many clinical cases and typical prostate cancer cell lines that are negative for these specific fusion pairs (in particular, all benign tumors examined are fusion gene negative) (Tomlins et al., 2005). This indicates that there are many cases that are not related to these two fusion genes but may be caused by other fusion events or distinct chromosomal rearrangements leading to the late stage cancer phenotype. To reconcile the fact that it is difficult to identify patterns in clinical samples, there has been an effort to develop statisti-
238
Cytogenet Genome Res 118:237–246 (2007)
cal analysis to group these diverse karyotypes, based on the notion that karyotypic convergence patterns will be found (Hoglund et al., 2001, 2005). This type of analysis will reveal some overall trend (such as, the degree of convergence or divergence), but is of much less practical value when dealing with a specific clinical case as it would be difficult to predict one individual case based on a divergent overall trend. Fundamental questions still remain, such as what causes high degrees of genetic heterogeneity? Are high levels of karyotypic heterogeneity a difficult issue to solve in terms of clinical implications? What is the relationship between genetic convergence and divergence in karyotypic evolution? To answer these questions, one needs to understand the mechanism of karyotypic heterogeneity. An effective method would be to study the dynamic process of karyotypic evolution and to compare the patterns as well as various end products of cancer evolution in different types of cancer and in the same types of cancer. Following this thought, the types of chromosomal aberrations have been systematically compared using various model systems that include the dysfunction of genes responsible for genome stability, the expression of oncoproteins and the treatment of carcinogens. In particular, the dynamics of karyotypic progression have been studied using an in vitro immortalization model. As a result of years of research, several previously ignored forms of chromosomal aberration, such as defective mitotic figures (DMFs) and chromosome fragmentations as well as other forms of nonclonal chromosome aberrations or NCCAs, have been re-investigated in the context of system instability, dynamic genome aberration and cancer evolution (Heng et al., 2004a, 2006a, b, c; Heng, 2007a; Stevens et al., 2007). These new discoveries indicate that the dynamic process occurs through instability mediated NCCAs and there are different patterns of karyotypic evolution in the process of cancer progression. These discoveries have shed new light on the mechanism of karyotypic heterogeneity as the central piece of carcinogenesis. At the same time, various genome projects have brought important new ideas and perspectives to support and explain some of these important findings. First, vertebrate evolution has been described as having ‘a shortage of genes’, as human and other species such as mice and rats share the same essential genes, while genome reconstruction seems to play an important role in evolution, indicating the importance of genetic variation at the genome level (Marcus, 2004; Murphy et al., 2005; Heng et al., 2006a; Kohn et al., 2006). It is thus possible that changing the genome context rather than gene content is a key for organismal evolution (Heng, 2007b; Heng et al., in preparation); second, as revealed by cutting edge genomics and molecular cytogenetic approaches, such as high resolution genomic arrays and large scale sequencing, there is a shortage of known cancer specific mutations in many cancer types while there is a high degree of sub-chromosomal aberrations (including large scale regional duplications/deletions at the subchromosome level) (Garnis et al., 2004; Baldwin et al., 2005; Stephens et al., 2005; Bignell et al., 2006; Feuk et al., 2006; Sjoblom et al., 2006). These facts coupled with high levels of stochastic mu-
tations that do not belong to the category of typical oncogenes or tumor suppressor genes, indicate that non-recurrent genetic aberrations occurring at various genetic organization levels might be more important than a handful of recurrent cancer genes (if they ever existed) (Heng, 2007a). All evidence points in the direction that stochastic genome aberrations coupled with stochastic mutations (rather than mutations themselves) play a significant role in cancer formation and natural speciation. To understand and appreciate this new perspective, we need to integrate this knowledge into a new conceptual framework for current cancer research. We thus will discuss the following aspects of genome dynamics in the context of cancer evolution, and the importance of system stability. We hope that this integrated model will initiate discussion within the field and will influence the field for years to come. The highly dynamic karyotypic patterns of cancer progression
The views on the importance of karyotypic heterogeneity are drastically different among investigators. It has been generally appreciated that karyotypic heterogeneity represents a common feature of a majority of cancer types (Heppner and Miller, 1998). However, the heterogeneous nature of chromosomal aberrations has been viewed as incidental background noise and recurrent karyotypes have been considered the signature of a particular cancer type (Mitelman, 2000; Albertson et al., 2003; Heng et al., 2004b). The search to identify specific clonal karyotypes has been a major decade-long cytogenetic effort. As evident from and influenced by blood cancers, the importance of inter-tumor and intra-tumor heterogeneity has not been considered at the conceptual level to study the discontinuity of clonal evolution. To reconcile the fact that the recurrent karyotypic signatures are so difficult to identify in solid tumors, the common explanation has been used that if we try harder, if we have more advanced technologies, and if we could reduce the stochastic genetic background ‘noise’, the long expected patterns will be found in all cancer types. Alternative viewpoints that focus on the non-recurrent type of chromosome aberrations, even if only accepted by a minority of researchers, does persistently emerge from time to time due to the fact that they are constantly observed in both laboratory and clinical settings. During the past seven years, our group had initially focused on the identification of chromosomal aberrations using a number of cutting edge molecular cytogenetic technologies. By applying advanced FISH technologies including multiple color SKY detection (Schrock et al., 1996; Heng et al., 1997, 2001b, 2003; Ye et al., 2001, 2006), we were not able to find cancer type specific aberrations in large numbers of examined primary tumors (Heng et al., 2004a, 2006a). This experience has triggered a number of important yet seemingly unpopular questions, such as, what if, there are no dominant recurrent karyotypes shared by the majority of cancer cases or solid tumors? What if high levels of heterogeneity are a key feature of sol-
id tumors and genetic divergence occurs in most cancer types (in this case, recurrent karyotypes only represent a small portion of all cases)? These alternative questions are of great interest from the perspective of systems biology as system dynamics is a key property of a system (Kitano, 2002). When a given system is unstable, both stochastic and specific changes can be observed and the importance of stochastic changes cannot be ignored. If we consider the genome as a system, then stochastic chromosomal changes are an important dynamic feature of a given genome as opposed to being just genetic ‘noise’. We have experimentally tested this by correlating the unstable status of the genome and the levels of stochastic chromosomal aberrations. Whenever the genome is unstable, caused by a dysfunction of gene maintenance or genome integrity or arising from environmental factors, it is the elevated levels of NCCAs that represent a universal feature of an increased level of system dynamics. Even if only a certain portion of the cells displays NCCAs, they in fact contribute to the population diversity that promotes cancer evolution. Following a demonstration that the level of stochastic chromosomal aberrations reflects the instability of the genome, our studies have further demonstrated that immortalization, cell transformation and the drug resistant process are stochastic events that occur along with dynamic interactions between various NCCAs and CCAs (Heng et al., 2006a–c; Heng et al., unpublished data). Significantly, the two phases of karyotypic evolution are a key to understanding the patterns of cancer evolution, as NCCAs represent the mechanism to produce a survival advantage while CCAs represent the mechanism producing growth advantages. These two aspects of cancer progression can be mutually exclusive or work in collaboration, depending on the stage of cancer progression and the level of genome or population stability. The dynamic relationship between NCCAs and CCAs can also explain why some tumor cell lines display relatively stable abnormal karyotypes, while others display a greater degree of variation. In general, however, cancer cell lines often have relatively stable karyotypes after evolutionary selection. The given karyotype would resist additional changes. Under new selection pressure, however, new clonal karyotypes will emerge though given NCCAs. Such phase interactions occur multiple times during this process and are responsible for dynamic cancer evolution and are the basis for high levels of genetic heterogeneity (Fig. 1). In addition to our demonstration that there are multiple cycles occurring at the karyotypic level, our recent data also supports that these dynamics occurred and are detectable at the sub-chromosomal and gene mutation levels. Even though further studies are needed to test the relationship between CIN and gene mutations in tumorigenesis (Michor et al., 2005), combined with published data on other key events in cancer progression including metastasis (Gao et al., 2005), the message is loud and clear: multiple cycles of NCCAs/CCAs are necessary for cancer initiation and progression to occur (Heng et al., submitted). Our recent analysis also suggested that the high level of genetic shuffling that we observed can effec-
Cytogenet Genome Res 118:237–246 (2007)
239
Fig. 1. A diagram illustrating the relationship between non-clonal genomic aberrations, such as non-clonal chromosome aberrations (NCCAs) and recurrent genomic aberrations, such as clonal chromosome aberrations (CCAs) during cancer evolution. The key element to determine which phase will be dominant is genome system stability, which can be influenced by various factors including internal genome structure and dynamics as well as environmental influences, such as cellcell interactions and selection pressure. For normal cells to transform into cancer cells, multiple cycles are required (Heng et al., 2006a–c).
System stable: Growth advantage
Unstable NCCA Phase
“Stepwise” Genomic Evolution
Stochastic Discontinuous Genomic Evolution
tively bring large numbers of mutations together, which is extremely difficult to achieve by a chronological series of mutations through growth and stepwise accumulation (Fearon and Vogelstein, 1990). Another important aspect of genome dynamics in cancer cells has been illustrated by a wave of genomic array technologies (Garnis et al., 2004; Mantripragada et al., 2004; Davies et al., 2005). Even though cryptic rearrangements have long been linked to human disease conditions through the close examination of subtelomeric regions and by conventional CGH, the appreciation of high levels of genetic aberration at the sub-chromosomal level was recently firmly established using a variety of platforms of CGH arrays. In particular, with the use of high resolution arrays with tiling resolution of the whole genome, high levels of sub-chromosomal aberrations have been documented in various cancer types. In conjunction with both interphase FISH and high resolution fiber FISH (Heng et al., 1992; Heng and Tsui, 1998), CGH arrays play an increasingly important role both to understand the mechanism of cancer formation and in clinical diagnosis applications (Iafrate et al., 2004; Feuk et al., 2006). The high levels of genomic changes detected above the gene level have in fact challenged the traditional thinking of the dominant role of defined cancer genes. It should be noted that the current profiling methods of genetic aberrations by these genomic technologies is mainly based on a mixture of cell populations isolated from tumor samples. By recording the average profile of a population containing variations of mixed CCAs and NCCAs, the detected degree of genetic heterogeneity will tend to be significantly lower than the actual degree of heterogeneity that exists. In addition, there is also a selection bias if the late tumor stages are used for profiling, as there is a certain degree of genetic convergence caused either by tissue specific selection or representing some common features of ‘out of control’ growth. Clearly, single cell CGH or expression profiling will provide more accurate information. Combined with other data sets targeting the gene level, and the subchromosomal level, our studies further indicated that the high level of genetic dynamics can be traced at multiple levels and in particular at the genome level. The
240
Stable CCA Phase
Cytogenet Genome Res 118:237–246 (2007)
System unstable: Survival advantage
genome system is the highest level of genetic organization that contains lower levels consisting of genes and regulation elements. The dynamic changes of the higher level will likely contain significant changes at the lower levels. In contrast, the lower level changes may not reflect a high level of change. Following this line of thinking, monitoring the overall instability from the genome level makes sense; even through targeting specific pathways at the gene level could provide more detailed information at the lower level of the system. Thus, comprehensive approaches are urgently needed to compare and to combine the information of genetic aberrations at various levels, and in particular at the higher organization levels. However, detecting genetic aberrations at the gene level has been a main focus in the field of molecular genetics and current genomics. Chromosomal aberrations have been traditionally thought of as ‘resolution too low’ compared to molecular analysis at the gene level. With the emergence of large scale genomics/proteomics and systems biology, the dynamic nature of protein-protein or gene-gene interactions represents a future direction. Unfortunately, even in some systems biologist’s minds, the importance of higher levels of genome organization has been more or less ignored as major efforts have been focused on interaction at the molecular level. To change this situation, it is necessary to define genome context, and to illustrate molecular interactions within the perspective of genome context. Following the establishment of a definition of genome context and typical levels of genetic organization, a systems approach can be effectively applied. For example, focusing on the sub-chromosomal level, concentrating on the patterns of co-deletions or duplication or amplifications within large genomic regions rather than the detection of single gene aberrations deserves more attention (Albertson, 2006). The concept of ‘chromatin domains’ needs to be integrated into the deletion and amplification data sets, as the chromatin loop domain is the unit of function that contains not only coding sequences but the sequences responsible for organization and gene regulation (Heng et al., 1996, 2001a, 2004b). Further, the functional contributions from interactions defined by sub-chromosomal compart-
ments (Guasconi et al., 2005; Albiez et al., 2006), as well as the functional impact of the entire genome generated by any chromosomal abnormalities needs to be closely analyzed (Heng et al., 2004a; Mai and Garini, 2006; Heng et al., submitted). Our surprising observation that the most significant genetic aberrations are detected mainly at the genome level (specifically when genomes are unstable) will change the viewpoint regarding the importance of genome aberrations in cancer progression. Previously, chromosomal aberrations were often thought to be secondary effects of gene function and the meaning of chromosomal aberrations is illustrated by linkages with specific oncogenes or tumor suppressor genes. Furthermore, variation can be found at the cell population level when scoring NCCAs (scoring high frequencies of CCAs is not as significant as scoring NCCAs). Most interesting, immediately prior to the crisis stage, none of the cells were chromosomally the same (the genome content was not the same) (Heng et al., 2006c)! In contrast, it is known that there is no detectable microsatellite instability in these cell lines indicating that the general mutation rate may not be significantly increased at least from the average cell population (Nahhas et al., unpublished data). In fact, p53 deficiency does not affect the accumulation of point mutations in a transgene target (Sands et al., 1995). Additional experiments are now under way to prove our hypothesis that cancer progression at large is driven by genome aberrations rather than the accumulation of gene mutations, and that stochastic gene mutations could be passive contributors through the mechanism of stochastic genome aberrations. Since multiple aberrations exist at such a high degree in cancer it is no wonder that cancer evolution can easily progress, even though it may take long periods of time to reach a cancer phenotype through the selection of successful combinations of genetic aberrations. It should be pointed out that, even though the concept of cancer evolution is generally accepted in the field, the pattern of cancer evolution has not been understood. In fact, there is a great deal of confusion in the field when dealing with core evolutionary questions. For example, there is confusion regarding the evolutionary process and the products of evolution as well as confusion regarding the dynamic nature of evolution as the main focus of research is on the continuity aspect of evolution. Accordingly, many key features of cancer evolution have not been vigorously tested. These questions include: is cancer a clonal disease? What is the importance of clonal discontinuity during cancer evolution? Do gene mutations or genome aberrations drive cancer evolution? What is the relationship of genetic/epigenetic variations between the gene and genome levels? The recent discoveries on patterns of cancer evolution will certainly shed new light on these issues. Obviously, understanding genome dynamics and the interactions among various genetic organization levels holds answers to these questions.
The source of multiple levels of genomic dynamics
In general, genomic dynamics are caused by system adaptability and instability. If one examines a given system, the degree of dynamics reflects on the status of this system. As we discussed, genome instability can be classified into two main forms: 1) Internal instability caused by defective genes or the unstable status of the genome context; and 2) induced instability caused by environmental factors, including virus infection, oncoprotein expression, carcinogen treatment, inflammation and many other factors. All these factors can trigger a similar response, as if the genome were internally unstable, at least in the initial stages. Induced instability is clearly related to both the nature of the inducers, and the status and structures of the internal stability of the system. System dynamics is one of the most important properties for a given system, which can be caused by instability and can be measured through levels of instability. Such dynamics are necessary for a system to function by providing adaptability but also could cause damage to the system itself. The dynamics of a system can be regulated and/or influenced by both normal physiological processes and by abnormal variations. For example, genome instability can be generated from metabolism products of normal processes, such as the impact of reactive oxygen species, as well as the normal aging process. Instability can also be generated by variations in regulation including the stochastic feature of regulation and epigenetic effects that occur during biological adaptation. In addition, instability can occur through, as we will discuss later, the accumulation of errors and further mistakes generated from ‘imperfect biological correction’ mechanisms. Another important source of system dynamics may come from the microenvironment of cancer cells. Our recent data demonstrated that selective pressures from the environment can destabilize the genome system and speed up the cancer evolutionary process (Heng et al., unpublished data), which is in agreement with the viewpoint/observation that selection plays an important role in cancer formation (Rubin, 2005). From a cell biology viewpoint, all elements that are involved in various aspects of a cell’s life should potentially contribute to the system’s stability. Frequently involved elements are linked to all stages of the cell cycle where DNA is replicated and condensed into chromosomes, segregated and then decondensed ready for the next cycle of DNA replication. During these main phases, there are a number of checkpoint mechanisms to ensure that mistakes are corrected and environmental stress can be defused or adapted. Traditionally, attention has been focused on replication errors and their repair pathways. Various cytogenetic assays have been developed to quantify the damage during the replication stage including scoring chromatin breaks, measuring SCE, and counting the dots of interphase nuclei (ISCN, 2005; Heng et al., 1997; Wu et al., 2004). It has long been suggested that to monitor each stage of the cell cycle it is important to score the abnormalities that contribute to overall chromosomal aberrations, as suggested in the ex-
Cytogenet Genome Res 118:237–246 (2007)
241
ample of the monitoring of condensation errors (Heng et al., 1988). Unfortunately, this suggestion has generally been ignored. With the exception of the monitoring of chromosome segregation in recent years due to the establishment of the linkage between errors of chromosomal segregation and aneuploidy, the idea of monitoring the entire process of the cell cycle has so far not been appreciated. A new phenotype for chromosome condensation defects, called defective mitotic figures or DMFs, was previously considered to be a slide preparation artifact, despite the original description having occurred over 20 years ago and its powerful applications have been used in the establishment of high resolution fiber FISH technologies (Heng et al., 1988, 1992; Heng and Tsui, 1998). We have recently finished the characterization of this new phenotype for this chromosomal aberration. Our studies demonstrated that DMFs are not a slide preparation artifact but an ignored form of chromosomal aberration caused by a combination of defects of condensation and G2-M checkpoints. Interestingly, DMFs are directly linked to cancer formation by causing the further destabilization of the genome and by generating new types of chromosomal aberrations including aneuploidy (Heng et al., 2004a, 2006a, and in preparation). Another example of genome dynamics that is related to mitosis is the chromosome fragmentation phenomenon. Characterized as a major form of mitotic cell death, chromosome fragmentation is significantly different from typical apoptosis or mitotic catastrophe. It occurs spontaneously or can be induced by treatment with chemotherapeutics and is easily observable within cell lines and tumor samples. Interestingly, chromosome fragmentation is linked with genomic instability, serving as a method to eliminate genomically unstable cells. Paradoxically this process could result in genome aberrations common in cancer (Heng et al., 2004a; Stevens et al., 2004, 2007). The acknowledgement that various phases of the chromosomal cycle can all contribute to genetic aberrations is an important first step. Further integration of genetic defects and environmental influences is another essential step in understanding how genome aberrations occur. The mechanism generating DMFs and chromosome fragmentation is an excellent example illustrating this issue. Indeed, if there are some defective genes involved, as in cell lines with ATM–/– genes and other defective genes responsible for the G2-M checkpoint, it is much easier to induce the DMF phenotype by interfering with the condensation process using inhibitors of Topo II. However, even without these known defective genes, DMFs can still be induced in lower frequency. The situation is similar in chromosomal fragmentation. For some cell lines or individuals, there is a high spontaneous frequency of chromosomal fragmentation reflected by an unstable genome. For cell lines with low frequencies, many drugs can effectively induce them in higher frequencies. Clearly, both internal instability and induced instability can influence the frequency level. In the majority of the human population there are no significant inherited bad genes, and system instability is more likely generated from environmentally influenced sources. This is
242
Cytogenet Genome Res 118:237–246 (2007)
in agreement with the epidemiological studies that an individual’s risk of developing cancer is approximately onethird inherited and two-third environmentally determined (Czene et al., 2002). We have thus hypothesized that inherent low levels of genome variation are an important basis of cancer evolution. These low levels of variation can stochastically trigger further instability. The results of elevated gene aberrations can generate new combinations that can either grow, or die off, or keep a similar pace, until the next stochastic events occur. Another aspect of the environmental impact on genome aberration comes from the ‘imperfect execution’ of some of the self repair or correction processes. When there is some damaged DNA detected by the cell system, in collaboration with checkpoint mechanisms, the DNA repair process will be initiated. In an ideal world, these mistakes would be corrected (and in most cases, they are). However, in the real world, some of the repairs will not be done as designed and further mistakes could be introduced. In these cases, the new introduced damage is more severe than the previous damage. A new series of repairs are then initiated. As a result, tiny mistakes in the repair events will eventually result in disaster to the genome stability maintenance process. The cell death process is an example of this situation. In a majority of cells that have been programmed for cell death, the elimination process gets rid of these cells. However, as we have observed, in some examined mitotic figures, the altered genome sometimes can survive when there are only 1–2 chromosomes fragmented in a given mitotic figure, these cells if they survive, will likely introduce further genome aberrations in the cell population (Stevens et al., in preparation). In conclusion, large quantities of genetic aberrations are generated through various normal and abnormal biological processes occurring at multiple genetic organization levels. These stochastic aberrations are necessary for cancer evolution to occur. Even though there are many types of aberrations (from gene mutations to genome aberrations), it is the combined effect of all of these aberrations that actually defines genome context. From a detection viewpoint, all genomic aberrations can be classified into a gene level, an epigenetic level and the chromosome or genome level. From a cell cycle point of view, aberrations can be generated from the phases of DNA replication, chromosomal condensation, chromosome segregation, de-condensation and their corresponding checkpoints. From an instability standpoint, aberrations can be classified according to the level of internal instability or induced instability. Among all these factors, factors that involve system instability favor cancer evolution. By contrast, typical oncogenes tend to be activated late as they are usually favored during the late stage of cancer evolution (Heng et al., in preparation). Evolutionary significance of genome dynamics
For normal cells to turn into cancer cells, evolutionary selection plays a key role. According to our model, the cellular metabolism and adaptability as well as environmental
stress associated with system instability mediated genome aberrations represents the initiation factors, while any other factors following the initiation could serve as promoting factors to speed up or slow down the evolutionary process. Even though the patterns of evolution can be drastically different as the internal and/or external environment are different and can constantly change, the genetic variations are preconditions for evolution to take place. Thus, the mechanism of effectively generating genetic variations will be favored by cancer evolution. Based on this idea, it is not difficult to understand why genome level aberrations are more effective compared to variation at the gene level, and thus would be expected to be, and actually are more frequently detected from clinical samples in human populations. Our model is different from the traditional viewpoint, that states that environmental stress activates cancer genes and then triggers genome instability coupled with chromosomal aberrations. This traditional viewpoint focuses more on specific pathways rather than on the system, and does not represent the majority of cases where genome aberrations are the initial event. The discovery that high levels of chromosomal dynamics are the key feature in cancer evolution is of significance. The high level of genomic heterogeneity that is illustrated by the observation that there are distinctively different patterns of karyotypic evolution, and every cell is chromosomally different prior to the crisis stage sheds new light on evolution in general and cancer evolution in particular. Combined with the fact that it is difficult to detect the time window when high rates of mutations occur, we hypothesize that stochastic genome aberrations are also responsible for assembling great numbers of gene mutations, in addition to genome level aberrations. Based on the concept of genome context and its essential role in cancer and organismal evolution, we agree with the viewpoint that various cancer cells represent a ‘new species’ when compared to normal cells from where cancer cells are derived from. Strong support comes from the fact that most of the cancer cells display altered karyotypes coupled with many variable abnormal phenotypes. As it is known that cancer progression can be monitored by population replacement as illustrated by our recent experiments, it is logical to suggest that the process of cancer progression in fact represents the generation of a series of new species. Interestingly, the stochastic process can generate a variety of different cell populations or ‘species’ with similar ‘out of control growth’ phenotypes. This viewpoint has gained experimental support as all stages of cancer progression tested so far including immortalization, transformation and drug resistance are always associated with novel karyotypes (Heng et al., 2006c; Heng, 2007a; Heng et al., in preparation). Following this lead, cancer should be considered as a genome based disease rather than a gene based disease, as in the majority of cancers it is the altered karyotype that is the universal feature where no universal mutations have been detected so far (Heng, 2007a). It is also possible that the progression of cancers mainly depends on karyotype evolution. In this sense, gene mutation alone will not be able to generate new ‘species’ of cancer.
This viewpoint is consistent with the systems biology point of view. The genome is the highest level of the genetic system. Within a given genome, genes and their interaction form a genetic network. The form and the boundary of the network are determined by the specific karyotypes we have named as genome context (Heng et al., 2006a; Heng, 2007b). These karyotypes are stochastically formed during cancer progression. As soon as a given karyotype is formed, the network context is determined with a specific pattern, even though the evolutionary future of this new genome context is not clear. Some of them might survive as NCCAs, with a majority of them being eliminated and a small number of them will grow to form CCAs with dominant gene context. The rationale of introducing the concept of genome context is to emphasize the importance of the higher order structure of the genome and the interactions within a given system in addition to its gene content. Obviously, cells with similar genome content could have drastically different genome context, as the spatial variation of the relative location of the same gene content could significantly change the system behavior. The concept of genome context is important to cancer evolution and gains support from similar evidence in natural evolution. By comparing the karyotypic patterns of a large number of species, distinctive karyotypes are the most important features among various species (King, 1993). Ample evidence shows that the karyotypic changes are crucial for species evolution. The mammalian karyotypic evolution is one good example (Navarro and Barton, 2003; Murphy et al., 2005; Kohn et al., 2006). Even though traditionally, the molecular evolution studies have focused on DNA or protein sequences rather than karyotypes, increasing evidence points to the importance of karyotypic evolution and its role in evolution, especially in light of the genome sequencing projects where there is a shortage of species-specific genes among various species and genome stochastic re-organization could play an important function. On the other hand, our novel observation that cancer evolution is mainly driven by karyotypic evolution will provide a new perspective on natural evolution. The whole set of chromosomes defines the genome context and determines the speciation (Heng et al., 2006a), while gene mutations modify the genome context to make it fit better (Heng et al., in preparation). Based on this conceptual thinking, we have recently revisited the long standing issue of why sexual reproduction exists despite its high cost as compared to asexual reproduction (the so called paradox of sex). It turns out, the generally accepted assumption regarding the degree of genetic diversity between sexual and asexual reproduction is not correct at all, and it is actually asexual reproduction that generates a much higher degree of genome diversity! The main function of sexual reproduction is not to provide genetic diversity, but rather to reduce diversity at the genome level and maintain genome integrity (Heng, 2007b). It should be pointed out that defining genome context according to chromosome based karyotypes rather than using the similarity of genes represents a non-traditional but
Cytogenet Genome Res 118:237–246 (2007)
243
correct viewpoint. The genome is not simply a bag of all the genes from a given species. The genome is the highest level of genetic organization and is very important. The genome arranges the composition of genes (both normal and mutations) and regulation elements. It is at the genome level that special rearrangements of genes along the chromosomes occur and where the chromosomal specific environment is defined by various non-coding sequences and serves as anchors for chromatin loops and architecture for domain interaction, as well as the relative position of the chromosomes within the nucleus. All these issues (rather than genes alone) are critical in terms of defining the system. For any chromosomal translocation and aneuploidy event, all the above relationships will be changed and in fact redefine a new genome system. Most significantly, our new discovery that the function of the genome and genes is opposite under certain circumstances (such as reducing or increasing the genetic diversity during the sexual process) illustrates the importance of departing from the gene centric view (Heng, 2007b). We believe that the conflict between the gene and genome needs immediate attention when considering studying the mechanism of natural evolution (Heng, unpublished data). Caution is thus needed when we deal with human cancer cell lines when there are drastic karyotype changes occurring. Are these lines still representative of human cells from the original generated tissue? Conclusions and applications
Understanding the dynamic nature of the cancer genome is a key to understanding the mechanism of genetic heterogeneity (system complexity) and population diversity, which is the genetic basis for cancer formation. Knowing that distinctive patterns of karyotypic evolution are defined by the genome system instability is particularly useful to study the pattern of cancer evolution and to characterize cell lines and clinical samples during the right time window. The pattern of karyotypic evolution can also be used for diagnostic and treatment purposes. For example, the level of NCCAs and CCAs could be used in clinical diagnosis as a ‘stochastic index’. To date, the majority of NCCAs are still considered to be non-significant background noise, and have not been used for clinical purposes. Based on our preliminary mouse data and clinical data, the level of NCCAs is well correlated with the overall instability of an individual, which could potentially be used as a biomarker to monitor system instability. More interestingly, the frequencies of the most complicated types of NCCAs that we have named ‘chaotic karyotypes’ seems closely linked to the status of cancer genome instability (Heng et al., in preparation). Accepting the fact that cancer evolution is driven by stochastic karyotypic progression can also help us to redefine our strategies in the fight against cancer. Traditional concepts suggest that cancer is caused by a stepwise accumulation of a handful of gene mutations. This concept has driven attempts to identify early mutations and their corre-
244
Cytogenet Genome Res 118:237–246 (2007)
sponding molecular pathways in the field for decades. If stochastic genome aberrations are the universal feature of cancer, then, focusing on specific pathways would be much less productive as there are potentially infinite pathways. Any given pathway representing only a small portion of patients coupled with a high level of heterogeneity, would be less practical for clinical usage. The right approach then should be to focus on monitoring overall genome stability and the patterns of evolution. As demonstrated (Heng et al., 2006, and in preparation), the elevation of NCCAs can be detected prior to the formation of dominant CCAs (often marked at the stage of carcinoma in situ). By monitoring the level of NCCAs we can move the clinical detection window to a much earlier period. The true appreciation of genome dynamics will once again put cytogenetics (chromosomal composition and patterns of rearrangement) into an important position in cancer research. In basic research, in vitro cell culture models play an important role in understanding the molecular basis of cancer, as a majority of researchers actively use in vitro models that can be relatively easier for experimentation. A key flaw is assuming that all cells in the same cell line have the same genome context when used by different investigators. Considering our previous discussion, most of the cell lines are, in fact, drastically different when cultured in various laboratories worldwide, especially when a particular cell line is within the dynamic phase, or is in a relatively stable phase but is under challenging culture conditions thus driving the genome into the dynamic phase (Heng et al., 2006). This serious situation has provided a great deal of misleading information and needs to be acknowledged and corrected. To compare biochemical or molecular genetic profiles, the key is using cells with the same genome context. Specifically, we need to make certain that sublines share the same or at least similar karyotypes. If there are numerous chromosomal aberrations detectable, they no longer have the same genome context, and should be considered to be different genetic identities. Thus, before we initiate any new research, the first consideration is a detailed karyotyping of our starting cell lines and their overall pattern of stability. During the course of research, it is very important to constantly monitor genome context changes. When the genome context is changed, this may represent a change in identity or ‘species’, and caution is needed. Another necessary consideration is using a combination of technologies to monitor various levels of genomic changes in the same cell line. Clearly, genetic variation at various levels will have different biological consequences. It is an important task to systematically compare the genome dynamics reflected at the overall karyotypic level (such as CCAs and NCCAs) and at the sub-chromosomal level (regional deletions and duplications monitored by CGH arrays). Of course, both clonal and non-clonal events should be investigated at these genetic levels using both single cell and mixed cell settings. This type of comparison will yield the following important information: a) if the overall aberrations detected from the higher system level contain aberration messages detected at the lower levels, and b) if various
levels of aberration can be simultaneously detected in a complimentary fashion. Our recent comparative studies have illustrated that there is a strong linkage between the overall karyotypic variation (measured by SKY) and the sub-chromosomal variations (measured by SNP analysis) (Heng et al., unpublished data). Based on this analysis, we anticipate that there will be a strong linkage between genome variations and gene mutations as well, even though a higher degree of dynamics is expected at the lower gene mutation level. Additional types of chromosomal aberrations need to be identified and their biological meanings need to be illustrated. Similar to the case of DMFs and chromosome fragmentation, there are still many types of uncharacterized chromosome aberrations frequently observed in cancer samples and efforts are needed to further characterize them and study their potential value to clinical diagnostic applications. These seemingly random aberrations may hold significant information on genome instability and the potential response to drug treatments. It is possible, upon our systematic characterization of these chromosomal aberrations, that the level of genome dynamics can someday be
used to monitor the overall genome instability of individuals, and to predict the probability of cancer as well as provide clues to early diagnosis and to classify patients into different groups based on overall genomic instability and likely corresponding drug response. Similar concepts and approaches could also be applied to other types of human disease conditions where system complexity and genome variations that reflect dynamics occurring at higher genetic levels play important roles. More importantly, the time has come to apply the genome centric concept to examine and reinterpret our current biological and medical research. Obviously, the concept of dynamic chromosomes and the genome will play a key role. Acknowledgements We would like to thank Dr. Gloria Heppner and Peter Moens for their encouragement. Thanks also go to Drs. Markku Kurkinen, Anton-Scott Goustin and other members of Dr. Heng’s Laboratory for their comments. Due to space limitations, we regret that we were unable to cite all the notable references deserving acknowledgement.
References Albertson DG: Gene amplification in cancer. Trends Genet 22: 447–55 (2006). Albertson DG, Collins C, McCormick F, Gray JW: Chromosome aberrations in solid tumors. Nat Genet 34: 369–376 (2003). Albiez H, Cremer M, Tiberi C, Vecchio L, Schermelleh L, et al: Chromatin domains and the interchromatin compartment form structurally defined and functionally interacting nuclear networks. Chromosome Res 14: 707–733 (2006). Atkin NB, Baker MC: Are human cancers ever diploid – or often trisomic? Conflicting evidence from direct preparation and cultures. Cytogenet Cell Genet 53: 58–60 (1990). Baldwin C, Garnis C, Zhang L, Rosin MP, Lam WL: Multiple microalterations detected at high frequency in oral cancer. Cancer Res 65:7561–7567 (2005). Bignell G, Smith R, Hunter C, Stephens P, Davies H, et al: Sequence analysis of the protein kinase gene family in human testicular germ-cell tumors of adolescents and adults. Genes Chromosomes Cancer 45: 42–46 (2006). Czene K, Lichtenstein P, Hemminki K: Environmental and heritable causes of cancer among 9.6 million individuals in the Swedish FamilyCancer Database. Int J Cancer 99: 260–266 (2002). Davies JJ, Wilson IM, Lam WL: Array CGH technologies and their applications to cancer genomes. Chromosome Res 13: 237–248 (2005). Druker BJ, Sawyers CL, Kantarjian H, Resta DJ, Reese SF, et al: Activity of a specific inhibitor of the BCR-ABL tyrosine kinase in the blast crisis of chronic myeloid leukemia and acute lymphoblastic leukemia with the Philadelphia chromosome. N Engl J Med 344:1038–1042 (2001). Duesberg P, Rasnick D: Aneuploidy, the somatic mutation that makes cancer a species of its own. Cell Motil Cytoskeleton 47: 81–107 (2000).
Duker NJ: Chromosome breakage syndromes and cancer. Am J Med Genet 115:125–129 (2002). Fearon ER, Vogelstein B: A genetic model for colorectal tumorigenesis. Cell 61: 759–767 (1990). Feuk L, Carson AR, Scherer SW: Structural variation in the human genome. Nat Rev Genet 7: 85–97 (2006). Foijer F, te Riele H: Check, double check: the G2 barrier to cancer. Cell Cycle 5: 831–836 (2006). Gao CF, Xie Q, Su YL, Koeman J, Khoo SK, Gustafson M, et al: Proliferation and invasion: plasticity in tumor cells. Proc Natl Acad Sci USA 102:10528–10533 (2005). Garnis C, Buys T, Lam W: Genetic alteration and gene expression modulation during cancer progression. Mol Cancer 3: 9 (2004). Guasconi V, Souidi M, Ait-Si-Ali S: Nuclear positioning, gene activity and cancer. Cancer Biol Ther 4:134–138 (2005). Heim S, Mitelman F: Cancer Cytogenetics. 2nd ed. (Wiley-Liss, New York 1995). Heng HH: Cancer genome sequencing: the challenges ahead. BioEssays 29: 783–794 (2007a). Heng HH: Elimination of altered karyotypes by sexual reproduction preserves species identity. Genome 50: 517–524 (2007b). Heng HH, Tsui LC: High resolution free chromatin/DNA fiber fluorescent in situ hybridization. J Chromatogr A 806:219–229 (1998). Heng H, Chen WY, Wang YC: Effects of pingyanymycin on chromosomes: a possible structural basis for chromosome aberration. Mutat Res 199:199–205 (1988). Heng HH, Squire J, Tsui LC: High-resolution mapping of mammalian genes by in situ hybridization to free chromatin. Proc Natl Acad Sci USA 89:9509–9513 (1992). Heng HH, Chamberlain J, Shi XM, Spyropoulos B, Tsui L-C, Moens PB: Regulation of meiotic chromatin loop size by chromosomal position. Proc Natl Acad Sci USA 93: 2795–2800 (1996).
Heng HH, Spyropoulos B, Moens PB: FISH technology in chromosome and genome research. Bioessays 19: 75–84 (1997). Heng HH, Krawetz SK, Lu W, Bremer S, Liu G, Ye CJ: Re-defining the chromatin loop domain. Cytogenet Cell Genet 93: 155–161 (2001a). Heng HH, Liu G, Lu W, Bremer S, Ye CJ, Hughes M, Moens P: Spectral karyotyping (SKY) of mouse meiotic chromosomes. Genome 44: 293–298 (2001b). Heng HH, Ye CJ, Yang F, Ebrahim S, Liu G, Bremer SW, et al: Analysis of marker or complex chromosomal rearrangements present in pre- and post-natal karyotypes utilizing a combination of G-banding, spectral karyotyping and fluorescence in situ hybridization. Clin Genet 63: 358–367 (2003). Heng HH, Goetze S, Ye CJ, Liu G, Stevens JB, Bremer SW, et al: Chromatin loops are selectively anchored using scaffold/matrix-attachment regions. J Cell Sci 117: 999–1008 (2004a). Heng HH, Stevens J, Liu G, Bremer SW, Ye CJ: Imaging genome abnormalities in cancer research. Cell Chromosomes 3: 1 (2004b). Heng HH, Liu G, Bremer S, Ye JK, Stevens J, Ye CJ: Clonal and non clonal chromosome aberrations and genome variation/aberration. Genome 49: 195–204 (2006a). Heng HH, Stevens JB, Liu G, Bremer SW, Ye KJ, Reddy PV, et al: Stochastic cancer progression driven by non-clonal chromosome aberrations. J Cell Physiol 208: 461–472 (2006b). Heng HH, Bremer SW, Stevens J, Ye KJ, Miller F, Liu G, Ye CJ: Cancer progression by non-clonal chromosome aberrations. J Cell Biochem 98: 1424–1435 (2006c). Heppner GH, Miller FR: The cellular basis of tumor progression. Intl Rev Cytol 177: 1–56 (1998). Hoglund, M, Sall T, Heim S, Mitelman F, Mandahl N, Fadl-Elmula I: Identification of cytogenetic subgroups and karyotypic pathways in transitional cell carcinoma. Cancer Res 61:8241–8246 (2001).
Cytogenet Genome Res 118:237–246 (2007)
245
Hoglund M, Frigyesi A, Sall T, Gisselsson D, Mitelman F: Statistical behavior of complex cancer karyotypes. Genes Chromosomes Cancer 42: 327–341 (2005). Iafrate AJ, Feuk L, Rivera MN, Listewnik ML, Donahoe PK, et al: Detection of large-scale variation in the human genome. Nat Genet 36: 949– 951 (2004). ISCN (1995): An International System for Human Cytogenetic Nomenclature, Mitelman F (ed) (S Karger, Basel 1995). ISCN (2005): An International System for Human Cytogenetic Nomenclature, Shaffer LG, Tommerup N (eds). (S Karger, Basel 2005). Johansson B, Fioretos T, Mitelman F: Cytogenetic and molecular genetic evolution of chronic myeloid leukemia. Acta Haematol 107: 76–94 (2002). Khanna KK, Jackson SP: DNA double-strand breaks: signaling, repair and the cancer connection. Nat Genet 27: 247–254 (2001). King M: Species Evolution. The Role of Chromosome Change (Cambridge University Press, Cambridge 1993). Kitano H: Systems biology: a brief overview. Science 295:1662–1664 (2002). Kohn M, Hogel J, Vogel W, Minich P, KehrerSawatzki H, et al: Reconstruction of a 450My-old ancestral vertebrate protokaryotype. Trends Genet 22: 203–210 (2006). Losi L, Baisse B, Bouzourene H, Benhattar J: Evolution of intratumoral genetic heterogeneity during colorectal cancer progression. Carcinogenesis 26:916–922 (2005). Mai S, Garini Y: The significance of telomeric aggregates in the interphase nuclei of tumor cells. J Cell Biochem 97: 904–915 (2006).
246
Mantripragada KK, Buckley PG, de Stahl TD, Dumanski JP: Genomic microarrays in the spotlight. Trends Genet 20: 87–94 (2004). Marcus GF: The Birth of the Mind. (Basic Books, New York 2004). McFadden D, Friedman J: Chromosome abnormalities in human beings. Mutat Res 396: 129–140 (1997). Michor F, Iwasa Y, Vogelstein B, Lengauer C, Nowak MA: Can chromosomal instability initiate tumorigenesis? Semin Cancer Biol 15: 43–49 (2005). Miller OJ, Therman E: Human Chromosomes. (Springer, New York 2003). Mitelman F: Recurrent chromosome aberrations in cancer. Mutat Res 462:247–253 (2000). Murphy WJ, Larkin DM, Everts-van der Wind A, Bourque G, Tesler G, et al: Dynamics of mammalian chromosome evolution inferred from multispecies comparative maps. Science 309: 613–617 (2005). Navarro A, Barton NH: Chromosomal speciation and molecular divergence–accelerated evolution in rearranged chromosomes. Science 300: 321–324 (2003). Rowley JD: The critical role of chromosome translocations in human leukemias. Annu Rev Genet 32:495–519 (1998). Rowley JD: Chromosome translocations: dangerous liaisons revisited. Nat Rev Cancer 1: 245– 250 (2001). Rubin H: Degree and kinds of selection in spontaneous neoplastic transformation: an operational analysis. Proc Natl Acad Sci USA 102: 9276– 9281 (2005). Sandberg AA: Cancer cytogenetics for clinicians. CA: Cancer J Clinicians 44: 136–159 (1994). Sands AT, Suraokar MB, Sanchez A, Marth JE, Donehower LA, Bradley A: p53 deficiency does not affect the accumulation of point mutations in a transgene target. Proc Natl Acad Sci USA 92:8517–8521 (1995).
Cytogenet Genome Res 118:237–246 (2007)
Schrock E, du Manoir S, Velman T, Schoell B, Wienberg J, et al: Multicolor spectral karyotyping of human chromosomes. Science 273: 494–497 (1996). Sjoblom T, Jones S, Wood LD, Parsons DW, Lin J, Barber T, et al: The consensus coding sequences of human breast and colorectal cancers. Science 314:268–274 (2006). Stephens P, Edkins S, Davies H, Greenman C, Cox C, et al: A screen of the complete protein kinase gene family identifies diverse patterns of somatic mutations in human breast cancer. Nat Genet 37:590–592 (2005). Stevens JB, Savasan S, Liu G, Bremer S, Atanasovski M, Xu W, et al: Cell death by chromosome elimination: Characterization of drug induced chromosomal fragmentation. ASHG 2004 Annual Meeting, program number 326 (2004). Stevens JB, Liu G, Bremer SW, Ye KJ, Xu W, et al: Mitotic cell death by chromosome fragmentation. Cancer Res 67:7686–7694 (2007). Tomlins SA, Rhodes DR, Perner S, Dhanasekaran SM, Mehra R, et al: Recurrent fusion of TMPRSS2 and ETS transcription factor genes in prostate cancer. Science 310:644–648 (2005). Wu X, Avni D, Chiba T, Yan F, Zhao Q, et al: SV40 T antigen interacts with Nbs1 to disrupt DNA replication control. Genes Dev 18: 1305–1316 (2004). Ye CJ, Lu W, Liu G, Bremer SW, Wang YA, et al: The combination of SKY and specific loci detection with FISH or immunostaining. Cytogenet Cell Genet 93:195–202 (2001). Ye CJ, Stevens J, Liu G, Ye J, Yang F, Bremer SW, Heng HH: Combined multicolor-FISH and immunostaining. Cytogenet Genome Res 114: 227–234 (2006).
balanced translocation of t(11; 22) and t(X;18) but also a large number of secondary changes (Hoglund et al., 2005; Lazar et al., 2006; Nakagawa et al., 2006). The presence of these translocations is often helpful in precise classification of soft part tumors. However, not all soft part tumors show such translocations and their absence does not rule out the diagnosis. Further, it is not known at this time whether those translocations can be neutralized or abolished by specific drugs, as is the case with CML and GIST. These observations raise a number of fundamental questions about the causes and mechanisms of chromosomal translocations in human tumors, the subject of this essay. ‘Crossing over’ in first meiotic division
Although the analogy is probably not perfect, the most common form of chromosomal translocation occurs during the ‘crossing over’ of chromosomal segments occurring during the normal first meiotic division of gametes. The purpose of this cell division, which occurs without prior duplication of DNA, is to reduce the number of chromosomes by 50%. The ‘crossing over’ is an exchange of segments between maternal and paternal homologous chromosomes. To achieve this goal, the two morphologically identical homologues are closely intertwined. In an analogy suggested by Miles (1979), each homologous chromosome may be conceived as a line of soldiers in blue uniform facing another line of soldiers in red uniform. Small groups of these soldiers exchange places, resulting in several crossing over or translocations of chromosomal segments. In normal meiosis, the translocation is balanced, i.e. the segments of DNA exchanged are of the same size but differ by their genetic content. Although meiosis is an essential event in all eukaryotic cells, its mechanisms are not well understood. For a crossing over to occur, a number of events must take place: 1. The homologous chromosomes must be adjacent to each other; 2. The two identical homologues must be able to recognize each other in order to avoid pairing with an inappropriate chromosome; 3. The surface of the chromosomes must be ‘sticky’ to allow them to adhere to each other; 4. The segments to be exchanged must be simultaneously identified on both homologues; 5. An enzyme, presumably an endonuclease, must precisely excise the segments to be exchanged on both homologues; 6. The segments must move from one homologue to another; 7. The newly translocated segments of DNA must be integrated into the new host homologue. Obviously, this process is extremely complex. Most data pertaining to mechanisms of crossing over in meiosis have been derived from study of fungi. An excellent summary of the current knowledge is found in Lorenz and Whitby (2006). In particular, these authors review the model by
248
Cytogenet Genome Res 118:247–251 (2007)
Szostak et al. (1983) who proposed that the recombination is initiated by single-strand nicks of DNA, followed by double-strand breaks and repair. Further support for this thesis was offered by Orr-Weaver and Szostak (1983) using plasmids containing double-strand gaps. Neale and Keeney (2006) suggested that the first step in crossing over is a homologous recombination of maternal and paternal genes before the genes change position. A meiosis-specific DNA strand exchange protein (DMC 1) may be the key to these events but its precise origin and mechanism are still unknown. Page and Hawley (2004) described a ‘synaptonemal complex’, a protein lattice that connects paired chromosomes in most meiotic systems. The complex resembles railroad tracks, known as ‘lateral elements,’ connected to each other by ‘transverse elements’. The lateral elements appear to play a key role in the recombination pathway, ensuring orderly exchange of chromosomal segments during the crossover. Sandovici et al. (2006) speculated that imprinted chromosomal regions are the parts of chromosomes that are the ‘hot spots’ and most likely to participate in recombination. Still, even these authors admit that ‘characterization of factors that influence the position and frequency of crossovers remain a challenge’. To complicate the issue still further, Housworth and Stahl (2003), as well as Whitby (2005), suggested that there are at least two pathways for generating crossover, involving as yet unidentified endonucleases. It is evident from this brief summary that we are still very far from fully understanding the mechanisms of crossover in normal meiosis. Failure of meiosis It has been repeatedly pointed out that a failure in any one of the steps of meiosis may lead to aneuploidy (Champion and Hawley, 2002; Oliver-Bonet et al., 2005; CodinaPascual et al., 2006; Lorenz and Whitby, 2006; Rajesh and Pittman, 2006). Thus, an error in the crossover sequence, whatever its nature, may lead to a life-threatening situation. The accuracy of the mechanism of crossover during meiosis must be very tightly controlled because so few such errors become of clinical significance. Translocations
If the mechanism of the common crossover is still shrouded in mystery, the mechanism of chromosomal translocation is still more difficult to understand. A few basic events must occur before a translocation of chromosomal segments may take place. Some of these events are similar to meiosis. 1. The segments of chromosomes to be exchanged must be adjacent to each other. 2. The surface of the adjacent segments must be modified to allow them to adhere to each other. 3. The segments must be excised and change location. a) If the translocation is balanced, two segments must change positions;
Fig. 1. Presumed territories of chromosomes X and Y in ciliated human bronchial cells. Unpublished data.
Fig. 2. Synchronous FISH staining of chromosomes 9 (red) and 22 (green) in normal human lymphocytes. Yellow areas indicate overlapping of the two chromosomes. A translocation t(9; 22) is common in chronic myelogenous leukemia, as discussed in the text. Unpublished data.
b) If the translocation is not balanced, a segment of one chromosome may adhere to or become incorporated into the second chromosome. Spatial proximity of chromosomes Spatial proximity of chromosomes exchanging segments was studied by Nikiforova et al. (2000) in reference to the position of RET H4 genes, common in thyroid carcinomas induced by radiation. Juxtaposition of the two genes was observed in 35% of normal human thyroid cells but in only 6% of normal mammary epithelial cells. Spatial proximity was also studied by Ashley et al. (2006) in reference to translocation (11;22) (q23;q11), the only known recurrent human non-Robertsonian translocation. Because this translocation occurs during meiosis, the spatial proximity of chromosomes 11 and 22 was observed in oocytes and spermatocytes. Proximity of the two translocating genes was observed. However, the same authors reported that proximity was also observed between sites 21p11 and 11q23, although this translocation is extremely rare. Thus, besides proximity, other factors must play a role in the translocation. The search for double-strand breaks and protein MRE11, involved in crossing over, did not shed a critical light on the mechanism of this translocation. Chromosomal territories and loss of heterozygosity
In a study of distribution of chromosomal territories in terminally differentiated ciliated human bronchial cells, I attempted to determine whether the position of chromosomal territories was constant vis-à-vis the poles of the cells (Koss, 1998). One of the purposes of the study was to deter-
Fig. 3. Differences in the size of territories occupied by the two homologues of chromosome 1 (left) and the territories occupied by the two homologues of chromosome 7 (right). Note the difference in the configuration of the two homologues, one compact and the other loosely structured. The differences were statistically significant. FISH preparations on bronchial cells in female patients. From Koss (1998).
mine whether chromosomes involved in translocations were adjacent to each other. Measuring the angles formed by chromosomes with the vertical and horizontal lines transecting the columnar cells disclosed that the territories of chromosomes X, 1, and 7 appeared to be well defined. The same was subsequently shown for territories of chromosomes X and Y in similar cells (Fig. 1). The proximity of the translocating chromosomes could not be established in this experimental setting. However, studies of painted chromosomes 9 and 22 in normal human lymphocytes documented an occasional single overlap (Fig. 2).
Cytogenet Genome Res 118:247–251 (2007)
249
Perhaps the greater significance of these studies was the observation that the two homologues of the same autosomes were not identical, one being compact and the other loosely structured. This observation strongly suggested that translocations involved only one of the two homologues, presumably resulting in the common phenomenon of loss of heterozygozity (LOH) (Fig. 3). LOH implies a genetic aberration on only one of the two homologues. The observation that the two normal homologues are not identical, with one dense and the other loosely structured, hence more likely to participate in DNA exchanges, may prove to be an important factor in this phenomenon (Koss, 1998). LOH in translocation between chromosomes 1 and 19 in acute lymphoblastic leukemias was discussed by Paulsson et al. (2005). These authors proposed complex mechanisms for this translocation but did not consider the simple possibility that only one of the two homologues was subject to translocation. Difilippantonio et al. (2002) suggested that translocations in mice are initiated by DNA cleavage caused by a recombination activating gene. A highly artificial intronbased system to induce chromosomal translocations was proposed by Elliot et al. (2005). Fenton et al. (2004) studied the t(11;14) translocation in multiple myeloma and concluded that the mechanism is very complex, not based on simple reciprocal events mediated by errors in class-switch recombination. Similar conclusions were reached by Weinstock et al. (2006). Aplan (2006) also concluded that the mechanisms of translocation remain poorly understood. Studies of a specific gene MYC in Burkitt’s lymphoma has not led to any conclusive results in reference to the function of the gene or the mechanisms of its origin (Hecht and Aster, 2000; Karlsson et al., 2003). Breger et al. (2005) proposed that delay in replication timing and mitotic condensation of chromosomes (DRT/DMC) may represent a common mechanism responsible for genetic instability in cancer cells. This concept requires additional proof and most likely will not solve any of the mysteries of translocation. Translocations in the absence of cancer
Mori et al. (2002) and Janz et al. (2003) observed translocations usually associated with lymphoma and leukemia in normal individuals without evidence of cancer. Mori ob-
served such changes in cord blood of monozygotic twins at a rate 100 times greater than the calculated frequency of leukemia. Janz et al. (2003) summarized existing evidence of cancerous translocations occurring in benign cells in healthy individuals starting with the paper by Limpens et al. (1991). There is no evidence at the time of this writing (2007) to determine whether such translocations have any prognostic value. Janz et al. (2003) conclude that translocations as such are probably insufficient to cause malignant transformation. Genetics of advanced tumors
Many years ago, I stated that advanced cancer was a genetic chaos (Koss, 1989). Recent genetic analyses of cancer have shown that besides the key genetic changes, most caused by translocations, innumerable other chromosomal abnormalities may be observed. Even in the much-studied neuroblastoma, Selzer et al. (2005) reported 58 chromosomal breakpoints that generated 45 large-scale partial chromosomal imbalances. Similarly, Sjoblom et al. (2006) who have analyzed only 11 breast and 11 colorectal cancers, reported 90 mutant genes. Also, 189 additional genes (11 per tumor) were mutated with significant frequency. In spite of the optimistic conclusions of this paper that the source and mechanism of these abnormalities will be conquered, it is not likely that answers of clinical value will become available in the near future. Conclusions
Translocations of chromosomes are a very common event in cancer. Although some products of the translocation may lead to helpful drugs as in CMM of GIST, neither the mechanism nor the significance of translocations is understood at this time. It may even be that translocations are a secondary or perhaps even tertiary event in cancer. The first step necessary in translocation could be a mechanism that allows the affected chromosomes to stick together. The selection of the segment of chromosome to be translocated is shrouded in total mystery. What happens to cells before translocation? We do not know but should try to find out.
References Aplan PD: Causes of oncogenic chromosomal translocation. Trends Genet 22: 46–55 (2006). Ashley T, Gaeth AP, Inagaki H, et al: Meiotic recombination and spatial proximity in the etiology of the recurrent t(11; 22). Am J Hum Genet 79: 524–538 (2006). Breger KS, Smith L, Thayer MJ: Engineering translocations with delayed replication: evidence for cis control of chromosome replication timing. Hum Mol Genet 14: 2813–2827 (2005).
250
Cannizzaro L: Principles of cytogenetics, in Koss LG, Melamed MR (eds): Diagnostic Cytology and its Histopathologic Bases, 5th Ed (Lippincott Williams and Wilkins, Philadelphia and Baltimore 2006). Champion MD, Hawley RS: Playing for half the deck: the molecular biology of meiosis. Nat Cell Biol 4(Suppl):s50-s56 (2002). Codina-Pascual M, Campillo M, Kraus J, et al: Crossover frequency and synaptonemal complex length: their variability and effects on human male meiosis. Mol Hum Reprod 12: 123– 133 (2006).
Cytogenet Genome Res 118:247–251 (2007)
Difilippantonio MJ, Petersen S, Chen HT, et al: Evidence for replicative repair of DNA doublestrand breaks leading to oncogenic translocation and gene amplification. J Exp Med 196: 469–480 (2002). Druker BJ, Talpaz M, Resta DJ, et al: Efficacy and safety of a specific inhibitor of the BCR-ABL tyrosine kinase in chronic myeloid leukemia. N Engl J Med 344:1031–1037 (2001). Druker BJ, Guilhot F, O’Brien SG, et al: Five-year follow-up of patients receiving imatinib for chronic myeloid leukemia. N Engl J Med 355: 2408–2417 (2006).
Elliott B, Richardson C, Jasin M: Chromosomal translocation mechanisms at intronic Alu elements in mammalian cells. Mol Cell 17: 885– 894 (2005). Fenton JA, Pratt G, Rothwell DG, et al: Translocation t(11; 14) in multiple myeloma: analysis of translocation breakpoints on der(11) and der(14) chromosomes suggests complex molecular mechanisms of recombination. Genes Chromosomes Cancer 39: 151–155 (2004). Gold JS, Dematteo RP: Combined surgical and molecular therapy: the gastrointestinal stromal tumor model. Ann Surg 244: 176–184 (2006). Groffen J, Stephenson JR, Heistercamp N, et al: Philadelphia chromosomal breakpoints are clustered within a limited region, bcr, on chromosome 22. Cell 36: 93–99 (1984). Haller F, Detken S, Schulten HJ, et al: Surgical management after neoadjuvant imatinib therapy in gastrointestinal strom tumours (GISTs) with respect to imatinib resistance caused by secondary KIT mutations. Ann Surg Oncol 14:52632 (2006). Hecht JL, Aster JC: Molecular biology of Burkitt’s lymphoma. J Clin Oncol 18: 3707–3721 (2000). Hoglund M, Gisselsson D, Mandahl N, Mitelman F: Ewing tumours and synovial sarcomas have critical features of karyotyping evolution in common with epithelial tumours. Int J Cancer 116:401–406 (2005). Housworth EA, Stahl FW: Crossover interference in humans. Am J Hum Genet 73: 188–197 (2003). Janz S, Potter M, Rabkin CS: Lymphoma- and leukemia-associated chromosomal translocations in healthy individuals. Genes Chromosomes Cancer 36: 211–223 (2003). Karlsson A, Deb-Basu D, Cherry A, et al: Defective double-strand DNA break repair and chromosomal translocations by MYC overexpression. Proc Natl Acad Sci USA 100: 9974–9979 (2003). Koss LG: From koilocytosis to molecular biology: The impact of cytology on concepts of early human cancer. The 32nd Maude Abbott Lecture delivered on March 7, 1989 in San Francisco, CA. Modern Pathol 2: 526–535 (1989).
Koss LG: Characteristics of chromosomes in polarized normal human bronchial cells provide a blueprint for nuclear organization. Cytogenet Cell Genet 82: 230–237 (1998). Lazar A, Abruzzo LV, Pollock RE, et al: Molecular diagnosis of sarcomas. Chromosomal translocations in sarcomas. Arch Pathol Lab Med 130: 1199–1207 (2006). Limpens J, de Jong D, van Krieken JH, et al: Bcl-2/ JH rearrangements in benign lymphoid tissues with follicular hyperplasia. Oncogene 6: 2271– 2276 (1991). Lorenz A, Whitby MC: Crossover promotion and prevention. Biochem Soc Trans 34: 537–541 (2006). Miles CP: The chromosomes and the mitotic cycle, in Koss LG (ed): Diagnostic Cytology and its Histopathologic Bases, 3rd Ed (JB Lippincott, Philadelphia 1979). Mitelman F: Recurrent chromosome aberrations in cancer. Mutat Res 462:247–253 (2000). Mori H, Colman SM, Xiao Z, et al: Chromosome translocations and covert leukemic clones are generated during normal fetal development. Proc Natl Acad Sci USA 99: 8242- 8247 (2002). Nakagawa Y, Numoto K, Yoshida A, et al: Chromosomal and genetic imbalances in synovial sarcoma detected by conventional and microarray comparative genomic hybridization. Cancer Res Clin Oncol 132:444–450 (2006). Neale MJ, Keeney S: Clarifying the mechanics of DNA strand exchange in meiotic recombination. Nature 442: 153–158 (2006). Nikiforova MN, Stringer JR, Blough R, et al: Proximity of chromosomal loci that participate in radiation-induced rearrangements in human cells. Science 290: 138–141 (2000). Nowell PC, Hungerford DA: A minute chromosome in human chronic granulocytic leukemia. Science 132:1497 (1960). Nowell PC: Progress with chronic myelogenous leukemia: a personal perspective over four decades. Annu Rev Med 53:1–13 (2002). Oliver-Bonet M, Benet J, Sun F, et al: Meiotic studies in two human reciprocal translocations and their association with spermatogenic failure. Hum Reprod 20:683–688 (2005).
Orr-Weaver TL, Szostak JW: Yeast recombination: the association between double-strand gap repair and crossing-over. Proc Natl Acad Sci USA 80:4417–4421 (1983). Page SL, Hawley RS: The genetics and molecular biology of the synaptonemal complex. Annu Rev Cell Dev Biol 20: 525–558 (2004). Paulsson K, Horvat A, Fioretos T, et al: Formation of der(19)t(1;19)(q23;p13) in acute lymphoblastic leukemia. Genes Chromosomes Cancer 42: 144–148 (2005). Rajesh C, Pittman DL: Cell cycle regulation in mammalian germ cells. Results Probl Cell Differ 42:343–367 (2006). Randolph TR: Chronic myelocytic leukemia–Part I: History, clinical presentation, and molecular biology. Clin Lab Sci 18: 38–48 (2005). Rowley JD: A new consistent chromosomal abnormality in chronic myelogenous leukemia identified by quinacrine fluorescence and Giemsa staining. Nature 243: 290–293 (1973). Sandovici I, Kassovska-Bratinova S, Vaughan JE, et al: Human imprinted chromosomal regions are historical hot-spots of recombination. PLOS Genet 2:944–954 (2006). Schnadig ID, Blanke CD: Gastrointestinal stromal tumors: imatinib and beyond. Curr Treat Options Oncol 7:427–437 (2006). Selzer RR, Richmond TA, Pofahl NJ, et al: Analysis of chromosome breakpoints in neuroblastoma at sub-kilobase resolution using fine-tiling oligonucleotide array CGH. Genes Chromosomes Cancer 44: 305–319 (2005). Sjoblom T, Jones S, Wood LD, et al: The consensus coding sequences of human breast and colorectal cancers. Science 314: 268–274 (2006). Szostak JW, Orr-Weaver TL, Rothstein RJ, Stahl FW: The double-strand-break repair model for recombination. Cell 33: 25–35 (1983). Weinstock DM, Elliott B, Jasin M: A model of oncogenic rearrangements: differences between chromosomal translocation mechanisms and simple double-strand break repair. Blood 107: 777–780 (2006). Whitby MC: Making crossovers during meiosis. Biochem Soc Trans 33: 1451–1455 (2005).
Cytogenet Genome Res 118:247–251 (2007)
251
Fig. 1. MicroRNA biogenesis. MicroRNAs are processed via a two-step mechanism, requiring the action of both nuclear and cytoplasmic members of the RNase III enzyme family (Drosha and Dicer, respectively). Ultimately, there is assembly of the strand selected mature microRNA into the microRNA RISC (miRISC) complex, where it performs its regulatory activity by targeting specific mRNAs for degradation or translational inhibition. * Indicates that the nuclear pri-miRNA in some cases is a substrate for adenosine to inosine dsRNA editing that may lead to differential processing and alterations in the mature microRNA sequence.
drimada et al., 2005). This final cleavage step is coupled to the assembly of the strand-selected mature microRNA into the RNA-induced silencing complex (RISC) where ultimately the microRNA performs its gene regulatory function (Gregory et al., 2005). The global importance of appropriate microRNA biogenesis in mammalian development has been strongly suggested by obligatory and conditional mouse knockouts of the Dicer gene that is required for microRNA processing. Genetic ablation of Dicer results in mouse embryonic lethality at day 7.5 during gastrulation (Bernstein et al., 2003) and mouse embryonic stem cells lacking Dicer are defective in cell proliferation and in vitro differentiation (Kanellopoulou et al., 2005; Murchison et al., 2005). Using conditional Dicer alleles to investigate later developmental events, Dicer has been demonstrated to be critical for hair follicle development (Andl et al., 2006), limb outgrowth (Harfe et al., 2005), lung epithelial morphogenesis (Harris et al., 2006), and T cell differentiation (Muljo et al., 2005), suggesting a possible role for microRNAs in these quite dissimilar processes. In general, microRNA-mRNA interactions are characterized by a duplex with thermal stability of at least 15–20 kcal/mole and by perfect or nearly perfect Watson-Crick base-pairing involving the 5ⴕ miRNA seed region (typically bases 2–8) (Doench and Sharp, 2004). Given the limited required sequence complementarity, each microRNA may
have hundreds of putative mRNA targets, although the available bioinformatic algorithms differ substantially in their predictions. The identification of mRNA targets for specific microRNAs has proved quite challenging. To date, roughly seventy mammalian mRNAs have been demonstrated to be direct targets of an individual microRNA, often by the use of luciferase reporter assays (see Sethupathy et al., 2006 and references therein). Even though there is a paucity of experimentally verified targets, current estimates suggest that 30% of all mRNAs may be regulated by miRNAs. In accordance with this hypothesis, transfection of miR-1 or miR-124 reduced the amounts of 96 or 174 mRNAs, respectively (Lim et al., 2005). In each case, twothirds of the downregulated mRNAs have a common seven nucleotide sequence motif in their 3ⴕ UTR. This motif is perfectly complementary to the seed region of the corresponding transfected microRNA, confirming that large numbers of mRNAs are miRNA targets. The mechanism of microRNA-directed mRNA destabilization in the cases of imperfect complementarity is still unclear but may involve rapid deadenylation (Wu et al., 2006). When perfect or near perfect complementarity exists, microRNAs act in a manner resembling siRNA. For example, several imprinted microRNAs that are generated from anti-peg11 bind with perfect complementarity to the RTL1/PEG11 ORF, leading to the endonucleolytic cleavage of RTL1/PEG11 mRNA (Davis
Cytogenet Genome Res 118:252–259 (2007)
253
et al., 2005). Alternatively, microRNA may regulate its mRNA target at the level of translational repression. The affected step in translation appears to be cap and polyA-tail dependent and occurs during initiation (Humphreys et al., 2005; Pillai et al., 2005). Given their regulatory role in gene expression, microRNAs have had a profound influence in the evolution of mRNA 3ⴕ UTRs, as tissues where a given microRNA is highly expressed tend to lack mRNAs that contain putative binding sites (Farh et al., 2005; Stark et al., 2005). Ultimately, mRNA expression profiling in conjunction with proteomic techniques will be necessary to determine the precise rules of microRNA-mRNA interaction that govern target specificity. Control of microRNA biogenesis
Pri-miRNAs are generated from either a separate gene or more commonly from a fragment, usually intronic, of a gene. Roughly, two-thirds of microRNAs investigated overlap known transcription units in the same orientation (Rodriguez et al., 2004; Weber, 2005). MicroRNAs in noncoding host genes map with similar frequencies to exons and introns. In contrast, microRNAs in protein coding genes are almost exclusively intronic. This finding makes sense, because the processing of an exonic microRNA would result in a cleaved, nonfunctional mRNA. Expression studies have demonstrated that microRNAs are coexpressed with neighboring microRNAs and their host gene (Baskerville and Bartel, 2005; Deo et al., 2007), strongly suggesting that the host gene promoter drives microRNA expression. Therefore, an important caveat for studying differentially expressed microRNAs that lie in protein coding genes is determining the significant molecule be it the protein, microRNA, or both, due to the coupled nature of their expression. Although microRNA expression is controlled primarily at the level of transcription, several noteworthy exceptions exist. Regulation of microRNA processing is one such mechanism. Interestingly, reduced Dicer expression is associated with poor prognosis in lung cancers (Karube et al., 2005), connecting levels of the microRNA processing machinery to clinical outcomes in cancer. Likewise, genes implicated in microRNA biogenesis such as RNASEN (Drosha), DGCR8, XPO5 (Exportin-5), and DICER are often located in regions of copy number change in three studied cancer types (Zhang et al., 2006). Even though it is unclear how exactly alterations in the microRNA processing machinery may contribute to cancer, differential microRNA processing occurs under normal physiological conditions. For example, in vivo processing of pre-miR-138-2 into the mature product is tissue specific, as determined by Northern and in situ hybridization analyses (Obernosterer et al., 2006). Two groups also have reported that the expression of several microRNAs are controlled epigenetically, as revealed by microRNA profiling of cells treated with either a histone deacetylase inhibitor and/or a DNA demethylating agent (Saito et al., 2006; Scott et al., 2006). In addition, al-
254
Cytogenet Genome Res 118:252–259 (2007)
ternative pre-miRNA processing may determine levels of a few microRNAs. As an example, miR-21, a commonly overexpressed microRNA in several solid tumors (Chan et al., 2005; Ciafre et al., 2005; Volinia et al., 2006), maps between alternative polyadenylation sites in the last exon of TMEM49. Finally, pri-miRNAs are subject to tissue specific adenosine to inosine RNA editing, leading to differential nuclear primiRNA processing (Yang et al., 2006) or alterations in the mature microRNA sequence that change target specificity (Blow et al., 2006). Thus, a multitude of mechanisms dictate functional microRNA levels; however, their specific involvement in carcinogenesis remains an open question. It is worth mentioning the significance of copy number changes on microRNA expression in cancer. A recent genome-wide screen was undertaken to explore this issue in breast cancer, ovarian cancer, and melanoma (Zhang et al., 2006). Even though microRNAs are frequently located within regions of genomic instability, microRNA expression and genomic copy number changes were correlated for only a subset of microRNAs investigated, indicating that other mechanisms are responsible for the majority of cancer-related microRNA dysregulation. Modulation of microRNA:mRNA interactions
MicroRNA sequence variants Mutations play an important role in oncogene activation and tumor suppressor gene inactivation. Consequently, an intriguing question is whether sequence variations in microRNAs are prevalent in and functionally relevant to cancer. Three studies have been conducted to address this issue. First, 173 pre-miRNAs were sequenced in 96 individuals, leading to the identification of nine polymorphisms in different pre-miRNAs and one in a mature sequence (Iwai and Naraba, 2005). As the nine polymorphisms in pre-miRNAs occur typically in the loop or in bulged/unpaired base within the stem, these polymorphisms were thought to be neutral. The sole polymorphism in a mature microRNA sequence was found in miR-30c-2. As expected, this polymorphism was predicted to alter mRNA target avidity and potentially specificity. This polymorphism is exceptionally rare in the Japanese population (3620 C/C: 11 C/A: 0 A/A) and its functional significance is unknown. In the second study, Calin et al. (2005) sequenced 42 microRNAs in seventy-five chronic lymphocytic leukemia (CLL) samples, revealing seven putative mutations (six in pri-miRNAs and one in a pre-miRNA) that were absent in 160 controls (Calin et al., 2005). Interestingly, mutations in either pri-miRNA16-1 or pre-miRNA-206 were associated with reduced expression of the corresponding microRNA in the affected patients, as determined by microarray hybridization. Of note, both patients that harbored the mutated pri-miRNA16-1 had a deletion of the second miR-15a/miR-16-1 allele, indicating that a mutation in the vicinity of precursor microRNA may be important in carcinogenesis. Finally, fifteen potentially tumorigenic miRNAs were investigated in 91 cancer cell lines for sequence variations (Diederichs and
Haber, 2006). In total, 22 known polymorphisms and 14 novel sequence variants were identified in pri-miRNAs outside of the pre-miRNA region. A novel sequence variant in pre-miR-26a variant was also found. Transfection experiments of 18 different variants revealed that none appeared to significantly affect the levels of the corresponding mature microRNA when compared to wild-type controls. As discussed earlier, RNA editing of microRNAs is an additional source of sequence variations. Ultimately, further studies will be necessary to determine the contribution of all types of microRNA sequence variants in carcinogenesis. mRNA sequence variants An underappreciated but obvious determinant for a factor modifying microRNA:mRNA interactions is the role of alternative pre-mRNA processing, either splicing or polyadenylation, in the generation of mRNA with unique 3ⴕ UTRs. For example, mRNA isoforms that contained a binding site for miR-1 or miR-104 were underrepresented in tissues where the corresponding microRNA was expressed (Legendre et al., 2006). Also, polymorphisms in the mRNA sequence may result in the creation or destruction of microRNA binding sites that have important phenotypic consequences. Consistent with this idea, quantitative trait mapping revealed that the Texel sheep locus containing the myostatin gene is responsible for the extreme meatiness observed in these sheep (Laville et al., 2004). Subsequent work revealed that a single nucleotide transition in the 3ⴕ UTR of the myostatin gene is causative, as it generates an illegitimate binding site for miR-1 and miR-206, leading to muscular hypertrophy by reducing myostatin protein levels (Clop et al., 2006). In papillary thyroid carcinomas, polymorphisms in the 3ⴕ UTR of KIT are thought to have functional significance in cancers as it alters the free energy by which the highly overexpressed miR-222 and miR-146 bind (He et al., 2005a). Hence, sequence variants in the 3ⴕ UTR of an oncogene may relieve microRNA-mediated repression, while changes in 3ⴕ UTR of a tumor suppressor may create an inappropriate microRNA binding site, resulting in gene silencing. Role of RNA binding proteins The regulatory activity of microRNAs is dependent on the complementarity and accessibility to its mRNA targets. Since the repertoire of RNA binding proteins is tissue specific and can be regulated in response to various stimuli, these proteins may be important in modulating microRNA: mRNA interactions by at least three distinct mechanisms. RNA binding proteins may directly compete for binding to the same site as a given microRNA. Alternatively, RNA binding proteins may have an indirect effect by modulating mRNA secondary structure, either increasing or decreasing the availability of mRNA regions complementary to a microRNA. Lastly, RNA binding proteins may affect mRNA subcellular localization. Evidence supporting this idea is the finding that the binding of HuR in the 3ⴕ UTR of the cationic amino acid transporter 1 (SLC7A1) overcomes miR122 mediated repression by altering mRNA subcellular lo-
calization from translational repressed cytoplasmic processing bodies to polysomes where it is actively transcribed (Bhattacharyya et al., 2006). Expression studies of microRNAs in cancer
A wide role for microRNAs in carcinogenesis was suggested by the fact that the majority (98 of 186) of known microRNAs at the time mapped to regions implicated in cancer susceptibility and in fragile sites (Calin et al., 2004b). Experimental support for this idea has been generated by high throughput microRNA expression profiling of multiple cancer types including breast (Iorio et al., 2005; Mattie et al., 2006), chronic lymphocytic leukemia (CLL) (Calin et al., 2004a, 2005; Pekarsky et al., 2006), colon (Michael et al., 2003; Bandres et al., 2006; Cummins et al., 2006), lung (Hayashita et al., 2005; Yanaihara et al., 2006), pancreatic endocrine (Roldo et al., 2006), pancreatic adenocarcinoma (Bloomston et al., 2007), prostate (Mattie et al., 2006), stomach (Volinia et al., 2006), and glioblastomas (Chan et al., 2005; Ciafre et al., 2005). In the majority of these studies, microRNA microarrays were utilized, demonstrating that numerous miRNAs are dysregulated in cancers. Furthermore, miRNAs often had predictive value in classifying the tumors relative to specific clinicopathological characteristics. For brevity’s sake, only three studies will be discussed here. The first published work that investigated global microRNA expression utilized a custom microarray to evaluate the expression of 190 human microRNAs in CLL and control CD5+ B cells (Calin et al., 2004a). The resulting pattern of microRNA expression could easily distinguish CLL from CD5+ cells and found that certain microRNAs had predictive value in distinguishing CLL classes based on clinically relevant criteria of ZAP70 expression, 13q14 deletion, or IgVH mutation. Later, Lu et al. (2005) utilized a beadbased flow cytometric method to profile the expression levels of 217 miRNAs in 334 samples derived from many different tumor types. They confirmed that microRNAs were particularly informative for classifying tumors, even those that are poorly differentiated. They also reported that microRNAs were generally downregulated in cancers. Lastly, a comprehensive study on 540 samples was performed interrogating microRNA expression signatures associated with six solid tumor types (lung, breast, stomach, prostate, colon, and pancreatic endocrine) using microRNA microarrays (Volinia et al., 2006). These authors concluded that certain microRNAs were commonly misregulated among several cancers, suggesting that they might play a central role in oncogenesis. In general, microRNAs were found to be upregulated in cancers relative to their normal controls, in contrast to the work of Lu et al. (2005). Notably, miR-21, a microRNA implicated as being anti-apoptotic in glioblastoma cells (Chan et al., 2005) and a negative regulator of PTEN expression (Meng et al., 2006), was overexpressed in all six cancer types investigated. Also, they identified important targets for three of the identified microRNAs, TGFBR2, PLAG1, and RB1 for miR-20a, miR-26a, and miR-106a,
Cytogenet Genome Res 118:252–259 (2007)
255
respectively. These results, when taken together, demonstrate the unequivocal significance of miRNA expression profiling of cancers and provide the intellectual framework for the selection of microRNAs for further cancer studies. MicroRNA and cancer
Many microRNAs have been suggested to play a key role in cancer due to their expression profiles. Unfortunately, their numbers exclude the possibility of discussing each in detail. For example, let-7 negatively regulates RAS (Johnson et al., 2005; Akao et al., 2006) and its low expression is correlated in lung cancer with poor survival (Takamizawa et al., 2004; Yanaihara et al., 2006). MiR-125b that is overexpressed in breast cancer (Iorio et al., 2005; Mattie et al., 2006) has been found to be inserted into a rearranged IgH allele in a patient with acute lymphoblastic leukemia (Sonoki et al., 2005). A translocation involving miR-142 and MYC was identified in an aggressive B-cell leukemia (Gauwerky et al., 1989). To date, the most extensively studied microRNAs that play a definitive role in cancer are miR-155 and two microRNA clusters, miR-15a/miR-16-1 and miR17/miR-18a/miR-19a/miR-20a/miR-19b-1/miR-92-1. CLL and miR-15a/miR-16-1 The first experimental evidence that microRNAs were involved in mammalian carcinogenesis was found in studies of CLL, the most prevalent adult leukemia in the Western world. Approximately half of all B-CLL cases are characterized by heterozygous or homozygous deletions of chromosome 13q14.3. Despite considerable efforts to identify a tumor suppressor gene in this region, the causative gene remained elusive. In 2002, analysis of ⬃30 kb deletion at 13q14.3 and chromosomal breakpoint mapping of a translocation t(2;13)(q32;q14) led to the identification of two tightly linked microRNAs, miR-15a and miR-16-1 as possible candidates (Calin et al., 2002). Although highly expressed in normal CD5+ cells, these microRNAs exhibited a dramatic reduction in expression in the majority of CLL cases studied. Importantly, the chromosomal position of these microRNAs mapped within all reported 13q14 deletions in B-CLL, including one that was a mere 10 kb in length (Liu et al., 1997). Subsequent work demonstrated that both miR-15a and miR16-1 regulate posttranscriptionally Bcl2, a potent cell survival factor, and that the enforced expression of miR-15a/ miR-16-1 induced apoptotic cell death in the Bcl2+ MEG-01 leukemia cell line (Cimmino et al., 2005). Intriguingly, the 3ⴕ UTR of BCL2 has three binding sites for miR-15a/miR-16 that are evolutionarily conserved in chickens and many mammalian species (Hagan, unpublished observations). Finally, miR-16 is important for mRNA turnover of transcripts containing AU-rich instability elements in their 3ⴕ UTR, a feature present in many growth regulatory mRNAs, suggesting that other mRNA targets for miR-16 may be important in the pathogenesis of CLL (Jing et al., 2005). The future generation of mice deficient for this important microRNA cluster may serve as a new mouse model for B-CLL.
256
Cytogenet Genome Res 118:252–259 (2007)
miR-17/92 cluster in cancer Human 13q31]q32 is a chromosomal region that is commonly amplified in several lymphomas. Previously, delineation of the minimal amplicon led to the discovery that it contained only two genes of which C13orf25 was the only one misexpressed (Ota et al., 2004). Surprisingly, C13orf25 is a noncoding RNA in which only an intronic region is evolutionarily conserved. This intronic region contains the polycistronic microRNA cluster, miR-17/miR-18a/miR-19a/ miR-20a/miR-19b-1/miR-92-1 (hereafter, referred to as miR17/92), whose overexpression was confirmed by microRNA microarray analysis in several B-cell lymphomas (He et al., 2005b). Using mice with B-cell specific overexpression of cmyc, He et al. (2005b) isolated hematopoietic stem cells that were subsequently retrovirally transduced with either miR17/92 or an empty vector. Upon transplantation into lethally irradiated recipients, they observed that overexpression of the miR-17/92 decreased lymphoma latency and increased tumor frequency. Interestingly, introduction of any single microRNA in this cluster did not enhance tumorigenesis, arguing that at least two microRNAs within this cluster act synergistically in disease development. In addition to its confirmed role in lymphoma development, this microRNA cluster may have wider significance in carcinogenesis. Members of this microRNA cluster are overexpressed in a wide range of tumors such as breast, colon, lung, prostate and pancreatic endocrine (Volinia et al., 2006). In lung cancer cells, miR-17/92 overexpression enhances cellular proliferation (Hayashita et al., 2005). The transcription factor c-myc can induce miR-17/92 expression and two of its members, miR-17-3 and miR-20a, negatively regulate E2F1 expression (O’Donnell et al., 2005). Furthermore, miR-20a negatively regulates Transforming Growth Factor  Receptor 2 (Volinia et al., 2006). Future mouse studies will be necessary to confirm the direct involvement of miR-17/92 in other cancers. BIC and miR-155 in cancer In birds and mammals, the BIC gene and more specifically miR-155 that it encodes plays a direct role in carcinogenesis. BIC was initially identified in chickens as a common proviral DNA integration site for avian leukosis virus (ALV) in ALV-induced lymphomas (Tam et al., 1997). BIC is a noncoding RNA characterized by limited sequence homology across species, except for an evolutionarily conserved hairpin and its immediate flanking sequence (Tam, 2001). Tam et al. (2002) reported that the retroviral introduction in chickens of a 491-bp BIC fragment containing the conserved hairpin region led to an increase versus controls in neoplastic disease, primarily lymphomas and erythroblastosis (Tam et al., 2002). Moreover, BIC was able to cooperate with c-myc in this process. The characterized BIC-related tumors have several noteworthy features. They lack light chain rearrangements indicative of an early B or T cell origin, often contain lymphocytes of different sizes, and lastly, exhibit oligo- or monoclonality. Although not known at the time, the avian BIC fragment that was utilized in this study contained miR-155.
Fig. 2. Postulated roles for microRNAs in cancer. * Shown in human gallbladder cancer cell line Mz-ChA-2. # Demonstrated in the human megakaryocytic cell line MEG-01.
In humans, miR-155 upregulation occurs in several solid tumors, including breast, colon, and lung (Iorio et al., 2005; Volinia et al., 2006; Yanaihara et al., 2006). In particular, high expression levels of miR-155 are statistically correlated in lung cancer with shortened length of postoperative survival in three cohorts (Yanaihara et al., 2006). Angiotensin II Type 1 Receptor is a direct target for miR155 in the lung (Martin et al., 2006). BIC and miR-155 are also overexpressed in several human lymphomas including Hodgkin, primary mediastinal and diffuse large B cell (Metzler et al., 2004; Eis et al., 2005; Kluiver et al., 2005). To investigate miR-155 in B-cell malignancies, transgenic mice were generated that overexpressed this microRNA under the control of a VH promoter-Ig heavy chain E enhancer (Costinean et al., 2006). Transgene expression in B cells began during the late pro-B cell stage and led to a preB cell proliferation defect and neoplasia. Specifically, B cell expansion was most pronounced for B220low/CD10low/ IgM–/TCR–/CD43– lymphoid cells that are also of prominence in human acute lymphoblastic leukemia and lymphoblastic lymphoma. Future directions in microRNA research
Although the microRNA field has grown exponentially in the last decade, much remains to be discovered, especially with respect to the developmental role of mammalian microRNAs and how these key regulators contribute to carcinogenesis. Several microRNAs are clearly implicated in carcinogenesis at the level of initiation and progression (Fig. 2). Moreover, microRNA expression studies, not only of cancer samples but for normal tissues as well, will be of
continued importance as the number of known and predicted microRNAs is constantly increasing. These studies will help in defining microRNAs that may be functionally significant. To date, no mouse microRNA gene knockout has been published, so the developmental role of individual microRNAs in mammals still remains to be elucidated. In particular, conditional knockout mouse alleles of putative tumor suppressor microRNAs will be helpful in determining how microRNA functions in normal development as well as in cancer. Likewise, tissue-specific overexpression of putative oncogenic microRNAs can be utilized to establish their precise role in cancer. Another avenue for future research is the development of therapies based on microRNA, due to their small size. For example, one can envisage using antisense oligonucleotides to inhibit the activity of an overexpressed, pathogenic microRNA. For example, intravenous injection in mice of cholesterol conjugated antago-miRs has been shown to decrease endogenous levels of the corresponding microRNA (Krutzfeldt et al., 2005). Moreover, various chemical modifications of antisense microRNA inhibitors have been studied to determine the ones that are most effective (Davis et al., 2006; Orom et al., 2006). Conversely, gene therapy vectors can be generated for tumor suppressor microRNAs that are lost in a cancer. Although it is unclear if these therapies will prove effective, future microRNA studies will clearly be informative for understanding the function of microRNAs in physiology and disease.
Cytogenet Genome Res 118:252–259 (2007)
257
References Akao Y, Nakagawa Y, Naoe T: Let-7 microRNA functions as a potential growth suppressor in human colon cancer cells. Biol Pharm Bull 29: 903–906 (2006). Alvarez-Garcia I, Miska EA: MicroRNA functions in animal development and human disease. Development 132: 4653–4662 (2005). Andl T, Murchison EP, Liu F, Zhang Y, Yunta-Gonzalez M, et al: The miRNA-processing enzyme Dicer is essential for the morphogenesis and maintenance of hair follicles. Curr Biol 16: 1041–1049 (2006). Bandres E, Cubedo E, Agirre X, Malumbres R, Zarate R, et al: Identification by real-time PCR of 13 mature microRNAs differentially expressed in colorectal cancer and non-tumoral tissues. Mol Cancer 5: 29 (2006). Baskerville S, Bartel DP: Microarray profiling of microRNAs reveals frequent coexpression with neighboring miRNAs and host genes. RNA 11: 241–247 (2005). Bernstein E, Kim SY, Carmell MA, Murchison EP, Alcorn H, et al: Dicer is essential for mouse development. Nat Genet 35: 215–217 (2003). Bhattacharyya SN, Habermacher R, Martine U, Closs EI, Filipowicz W: Relief of microRNAmediated translational repression in human cells subjected to stress. Cell 125: 1111–1124 (2006). Bloomston M, Frankel WL, Petrocca F, Volinia S, Alder H, et al: MicroRNA expression patterns to differentiate pancreatic adenocarcinoma from normal pancreas and chronic pancreatitis. J Am Med Assoc 297:1901–1908 (2007). Blow MJ, Grocock RJ, van Dongen S, Enright AJ, Dicks E, et al: RNA editing of human microRNAs. Genome Biol 7:R27 (2006). Cai X, Hagedorn CH, Cullen BR: Human microRNAs are processed from capped, polyadenylated transcripts that can also function as mRNAs. RNA 10:1957–1966 (2004). Calin GA, Dumitru CD, Shimizu M, Bichi R, Zupo S, et al: Frequent deletions and down-regulation of microRNA genes mir15 and mir16 at 13q14 in chronic lymphocytic leukemia. Proc Natl Acad Sci USA 99: 15524–15529 (2002). Calin GA, Liu CG, Sevignani C, Ferracin M, Felli N, et al: MicroRNA profiling reveals distinct signatures in B cell chronic lymphocytic leukemias. Proc Natl Acad Sci USA 101:11755–11760 (2004a). Calin GA, Sevignani C, Dumitru CD, Hyslop T, Noch E, et al: Human microRNA genes are frequently located at fragile sites and genomic regions involved in cancers. Proc Natl Acad Sci USA 101:2999–3004 (2004b). Calin GA, Ferracin M, Cimmino A, Di Leva G, Shimizu M, et al: A microRNA signature associated with prognosis and progression in chronic lymphocytic leukemia. N Engl J Med 353:1793– 1801 (2005). Carthew RW: Gene regulation by microRNAs. Curr Opin Genet Dev 16: 203–208 (2006). Chan JA, Krichevsky AM, Kosik KS: MicroRNA-21 is an antiapoptotic factor in human glioblastoma cells. Cancer Res 65: 6029–6033 (2005). Chendrimada TP, Gregory RI, Kumaraswamy E, Norman J, Cooch N, et al: TRBP recruits the Dicer complex to Ago2 for microRNA processing and gene silencing. Nature 436: 740–744 (2005). Ciafre SA, Galardi S, Mangiola A, Ferracin M, Liu CG, et al: Extensive modulation of a set of microRNAs in primary glioblastoma. Biochem Biophys Res Commun 334: 1351–1358 (2005).
258
Cimmino A, Calin GA, Fabbri M, Iorio MV, Ferracin M, et al: Mir-15 and mir-16 induce apoptosis by targeting BCL2. Proc Natl Acad Sci USA 102: 13944–13949 (2005). Clop A, Marcq F, Takeda H, Pirottin D, Tordoir X, et al: A mutation creating a potential illegitimate microRNA target site in the myostatin gene affects muscularity in sheep. Nat Genet 38: 813–818 (2006). Costinean S, Zanesi N, Pekarsky Y, Tili E, Volinia S, et al: Pre-B cell proliferation and lymphoblastic leukemia/high-grade lymphoma in E(mu)miR155 transgenic mice. Proc Natl Acad Sci USA 103: 7024–7029 (2006). Cummins JM, He Y, Leary RJ, Pagliarini R, Diaz LA, Jr, et al: The colorectal microRNAome. Proc Natl Acad Sci USA 103:3687–3692 (2006). Davis E, Caiment F, Tordoir X, Cavaille J, Ferguson-Smith A, et al: RNAi-mediated allelic trans-interaction at the imprinted Rtl1/Peg11 locus. Curr Biol 15: 743–749 (2005). Davis S, Lollo B, Freier S, Esau C: Improved targeting of miRNA with antisense oligonucleotides. Nucleic Acids Res 34:2294–2304 (2006). Deo M, Yu JY, Chung KH, Tippens M, Turner DL: Detection of mammalian microRNA expression by in situ hybridization with RNA oligonucleotides. Dev Dyn 236:912 (2007). Diederichs S, Haber DA: Sequence variations of microRNAs in human cancer: Alterations in predicted secondary structure do not affect processing. Cancer Res 66: 6097–6104 (2006). Doench JG, Sharp PA: Specificity of microRNA target selection in translational repression. Genes Dev 18:504–511 (2004). Eis PS, Tam W, Sun L, Chadburn A, Li Z, et al: Accumulation of miR-155 and BIC RNA in human B cell lymphomas. Proc Natl Acad Sci USA 102: 3627–3632 (2005). Farh KK, Grimson A, Jan C, Lewis BP, Johnston WK, et al: The widespread impact of mammalian microRNAs on mRNA repression and evolution. Science 310:1817–1821 (2005). Gauwerky CE, Huebner K, Isobe M, Nowell PC, Croce CM: Activation of MYC in a masked t(8; 17) translocation results in an aggressive B-cell leukemia. Proc Natl Acad Sci USA 86: 8867– 8871 (1989). Gregory RI, Chendrimada TP, Cooch N, Shiekhattar R: Human RISC couples microRNA biogenesis and posttranscriptional gene silencing. Cell 123:631–640 (2005). Han J, Lee Y, Yeom KH, Kim YK, Jin H, Kim VN: The Drosha-DGCR8 complex in primary microRNA processing. Genes Dev 18:3016–3027 (2004). Harfe BD, McManus MT, Mansfield JH, Hornstein E, Tabin CJ: The RNAse III enzyme Dicer is required for morphogenesis but not patterning of the vertebrate limb. Proc Natl Acad Sci USA 102:10898–10903 (2005). Harris KS, Zhang Z, McManus MT, Harfe BD, Sun X: Dicer function is essential for lung epithelium morphogenesis. Proc Natl Acad Sci USA 103:2208–2213 (2006). Hayashita Y, Osada H, Tatematsu Y, Yamada H, Yanagisawa K, et al: A polycistronic microRNA cluster, miR-17-92, is overexpressed in human lung cancers and enhances cell proliferation. Cancer Res 65: 9628–9632 (2005). He H, Jazdzewski K, Li W, Liyanarachchi S, Nagy R, et al: The role of microRNA genes in papillary thyroid carcinoma. Proc Natl Acad Sci USA 102: 19075–19080 (2005a). He L, Thomson JM, Hemann MT, HernandoMonge E, Mu D, et al: A microRNA polycistron as a potential human oncogene. Nature 435: 828–833 (2005b).
Cytogenet Genome Res 118:252–259 (2007)
Humphreys DT, Westman BJ, Martin DI, Preiss T: MicroRNAs control translation initiation by inhibiting eukaryotic initiation factor 4E/cap and poly(A) tail function. Proc Natl Acad Sci USA 102: 16961–16966 (2005). Iorio MV, Ferracin M, Liu CG, Veronese A, Spizzo R, et al: MicroRNA gene expression deregulation in human breast cancer. Cancer Res 65: 7065–7070 (2005). Iwai N, Naraba H: Polymorphisms in human premiRNAs. Biochem Biophys Res Commun 331: 1439–1444 (2005). Jing Q, Huang S, Guth S, Zarubin T, Motoyama A, et al: Involvement of microRNA in AU-rich element-mediated mRNA instability. Cell 120: 623–634 (2005). Johnson SM, Grosshans H, Shingara J, Byrom M, Jarvis R, et al: RAS is regulated by the let-7 microRNA family. Cell 120:635–647 (2005). Kanellopoulou C, Muljo SA, Kung AL, Ganesan S, Drapkin R, et al: Dicer-deficient mouse embryonic stem cells are defective in differentiation and centromeric silencing. Genes Dev 19: 489– 501 (2005). Karube Y, Tanaka H, Osada H, Tomida S, Tatematsu Y, et al: Reduced expression of Dicer associated with poor prognosis in lung cancer patients. Cancer Sci 96: 111–115 (2005). Kluiver J, Poppema S, de Jong D, Blokzijl T, Harms G, et al: BIC and miR-155 are highly expressed in Hodgkin, primary mediastinal and diffuse large B cell lymphomas. J Pathol 207: 243–249 (2005). Krutzfeldt J, Rajewsky N, Braich R, Rajeev KG, Tuschl T, et al: Silencing of microRNAs in vivo with ‘antagomirs’. Nature 438: 685–689 (2005). Laville E, Bouix J, Sayd T, Bibe B, Elsen JM, et al: Effects of a quantitative trait locus for muscle hypertrophy from Belgian Texel sheep on carcass conformation and muscularity. J Anim Sci 82:3128–3137 (2004). Lee Y, Ahn C, Han J, Choi H, Kim J, et al: The nuclear RNAse III Drosha initiates microRNA processing. Nature 425: 415–419 (2003). Lee Y, Kim M, Han J, Yeom KH, Lee S, et al: MicroRNA genes are transcribed by RNA polymerase II. EMBO J 23: 4051–4060 (2004). Legendre M, Ritchie W, Lopez F, Gautheret D: Differential repression of alternative transcripts: a screen for miRNA targets. PLoS Comput Biol 2: e43 (2006). Lim LP, Lau NC, Garrett-Engele P, Grimson A, Schelter JM, et al: Microarray analysis shows that some microRNAs downregulate large numbers of target mRNAs. Nature 433:769–773 (2005). Liu Y, Corcoran M, Rasool O, Ivanova G, Ibbotson R, et al: Cloning of two candidate tumor suppressor genes within a 10 kb region on chromosome 13q14, frequently deleted in chronic lymphocytic leukemia. Oncogene 15: 2463–2473 (1997). Lu J, Getz G, Miska EA, Alvarez-Saavedra E, Lamb J, et al: MicroRNA expression profiles classify human cancers. Nature 435: 834–838 (2005). Lum AM, Wang BB, Li L, Channa N, Bartha G, et al: Retroviral activation of the mir-106a microRNA cistron in T lymphoma. Retrovirology 4:5 (2007). Lund E, Guttinger S, Calado A, Dahlberg JE, Kutay U: Nuclear export of microRNA precursors. Science 303: 95–98 (2004). Martin MM, Lee EJ, Buckenberger JA, Schmittgen TD, Elton TS: MicroRNA-155 regulates human angiotensin II type 1 receptor expression in fibroblasts. J Biol Chem 281: 18277–18284 (2006).
Mattie MD, Benz CC, Bowers J, Sensinger K, Wong L, et al: Optimized high-throughput microRNA expression profiling provides novel biomarker assessment of clinical prostate and breast cancer biopsies. Mol Cancer 5: 24 (2006). Meng F, Henson R, Lang M, Wehbe H, Maheshwari S, et al: Involvement of human microRNA in growth and response to chemotherapy in human cholangiocarcinoma cell lines. Gastroenterology 130: 2113–2129 (2006). Metzler M, Wilda M, Busch K, Viehmann S, Borkhardt A: High expression of precursor microRNA-155/BIC RNA in children with Burkitt lymphoma. Genes Chromosomes Cancer 39: 167–169 (2004). Michael MZ, O’Connor SM, van Holst Pellekaan NG, Young GP, James RJ: Reduced accumulation of specific microRNAs in colorectal neoplasia. Mol Cancer Res 1:882–891 (2003). Muljo SA, Ansel KM, Kanellopoulou C, Livingston DM, Rao A, Rajewsky K: Aberrant T cell differentiation in the absence of Dicer. J Exp Med 202:261–269 (2005). Murchison EP, Partridge JF, Tam OH, Cheloufi S, Hannon GJ: Characterization of Dicer-deficient murine embryonic stem cells. Proc Natl Acad Sci USA 102:12135–12140 (2005). O’Donnell KA, Wentzel EA, Zeller KI, Dang CV, Mendell JT: C-myc-regulated microRNAs modulate E2F1 expression. Nature 435: 839– 843 (2005). Obernosterer G, Leuschner PJ, Alenius M, Martinez J: Post-transcriptional regulation of microRNA expression. RNA 12: 1161–1167 (2006). Orom UA, Kauppinen S, Lund AH: Lna-modified oligonucleotides mediate specific inhibition of microRNA function. Gene 372: 137–141 (2006). Ota A, Tagawa H, Karnan S, Tsuzuki S, Karpas A, et al: Identification and characterization of a novel gene, C13orf25, as a target for 13q31]q32 amplification in malignant lymphoma. Cancer Res 64:3087–3095 (2004). Pekarsky Y, Santanam U, Cimmino A, Palamarchuk A, Efanov A, et al: Tcl1 expression in chronic lymphocytic leukemia is regulated by mir-29 and mir-181. Cancer Res 66: 11590–11593 (2006). Pillai RS, Bhattacharyya SN, Artus CG, Zoller T, Cougot N, et al: Inhibition of translational initiation by let-7 microRNA in human cells. Science 309:1573–1576 (2005). Rodriguez A, Griffiths-Jones S, Ashurst JL, Bradley A: Identification of mammalian microRNA host genes and transcription units. Genome Res 14:1902–1910 (2004). Rodriguez A, Vigorito E, Clare S, Warren MV, Couttet P, et al: Requirement of bic/microRNA155 for normal immune function. Science 316: 608–611 (2007).
Roldo C, Missiaglia E, Hagan JP, Falconi M, Capelli P, et al: MicroRNA expression abnormalities in pancreatic endocrine and acinar tumors are associated with distinctive pathologic features and clinical behavior. J Clin Oncol 24: 4677– 4684 (2006). Saito Y, Liang G, Egger G, Friedman JM, Chuang JC, et al: Specific activation of microRNA-127 with downregulation of the proto-oncogene BCL6 by chromatin-modifying drugs in human cancer cells. Cancer Cell 9: 435–443 (2006). Scott GK, Mattie MD, Berger CE, Benz SC, Benz CC: Rapid alteration of microRNA levels by histone deacetylase inhibition. Cancer Res 66: 1277–1281 (2006). Scott GK, Goga A, Bhaumik D, Berger CE, Sullivan CS, et al: Coordinate suppression of ERBB2 and ERBB3 by enforced expression of micro-RNA miR-125a or miR-125b. J Biol Chem 282: 1479– 1486 (2007). Sethupathy P, Corda B, Hatzigeorgiou AG: Tarbase: A comprehensive database of experimentally supported animal microRNA targets. RNA 12: 192–197 (2006). Sonoki T, Iwanaga E, Mitsuya H, Asou N: Insertion of microRNA-125b-1, a human homologue of lin-4, into a rearranged immunoglobulin heavy chain gene locus in a patient with precursor Bcell acute lymphoblastic leukemia. Leukemia 19:2009–2010 (2005). Stark A, Brennecke J, Bushati N, Russell RB, Cohen SM: Animal microRNAs confer robustness to gene expression and have a significant impact on 3ⴕ UTR evolution. Cell 123: 1133–1146 (2005). Takamizawa J, Konishi H, Yanagisawa K, Tomida S, Osada H, et al: Reduced expression of the let-7 microRNAs in human lung cancers in association with shortened postoperative survival. Cancer Res 64: 3753–3756 (2004). Tam W: Identification and characterization of human BIC, a gene on chromosome 21 that encodes a noncoding RNA. Gene 274: 157–167 (2001). Tam W, Ben-Yehuda D, Hayward WS: BIC, a novel gene activated by proviral insertions in avian leukosis virus-induced lymphomas, is likely to function through its noncoding RNA. Mol Cell Biol 17: 1490–1502 (1997). Tam W, Hughes SH, Hayward WS, Besmer P: Avian BIC, a gene isolated from a common retroviral site in avian leukosis virus-induced lymphomas that encodes a noncoding RNA, cooperates with c-myc in lymphomagenesis and erythroleukemogenesis. J Virol 76: 4275–4286 (2002). Thai TH, Calado DP, Casola S, Ansel KM, Xiao C, et al: Regulation of the germinal center response by microRNA-155. Science 316:604–608 (2007). Volinia S, Calin GA, Liu CG, Ambs S, Cimmino A, et al: A microRNA expression signature of human solid tumors defines cancer gene targets. Proc Natl Acad Sci USA 103: 2257–2261 (2006).
Wang CL, Wang BB, Bartha G, Li L, Channa N, et al: Activation of an oncogenic microRNA cistron by provirus integration. Proc Natl Acad Sci USA 103: 18680–18684 (2006). Weber MJ: New human and mouse microRNA genes found by homology search. FEBS J 272: 59–73 (2005). Wienholds E, Plasterk RH: MicroRNA function in animal development. FEBS Lett 579:5911–5922 (2005). Wu L, Fan J, Belasco JG: MicroRNAs direct rapid deadenylation of mRNA. Proc Natl Acad Sci USA 103: 4034–4039 (2006). Yanaihara N, Caplen N, Bowman E, Seike M, Kumamoto K, et al: Unique microRNA molecular profiles in lung cancer diagnosis and prognosis. Cancer Cell 9: 189–198 (2006). Yang W, Chendrimada TP, Wang Q, Higuchi M, Seeburg PH, et al: Modulation of microRNA processing and expression through RNA editing by ADAR deaminases. Nat Struct Mol Biol 13:13–21 (2006). Zhang L, Huang J, Yang N, Greshock J, Megraw MS, et al: MicroRNAs exhibit high frequency genomic alterations in human cancer. Proc Natl Acad Sci USA 103: 9136–9141 (2006). Zhao Y, Ransom JF, Li A, Vedantham V, von Drehle M, et al: Dysregulation of cardiogenesis, cardiac conduction, and cell cycle in mice lacking miRNA-1-2. Cell 129: 303–317 (2007).
Note added in proof Since the acceptance of this review, many important papers have been published in relation to mammalian microRNA biology. For brevity’s sake, a few noteworthy studies in the context of this review will be mentioned. MicroRNA-125a/b whose expression is frequently lost or reduced in breast cancer has been reported to regulate the important oncogenes ERBB2 and ERBB3 (Scott et al., 2007). Two research groups have published a mouse knockout of miR-155, demonstrating that this microRNA is required for normal immune function (Rodriguez et al., 2007; Thai et al., 2007). Similarly, a mouse knockout of miR-1-2 is characterized by cardiac defects, leading to 50% lethality by weaning age (Zhao et al., 2007). Lastly, mouse screens that involve mapping of retroviral integration sites that lead to T-cell lymphoma development have identified the activation of two separate microRNA clusters, miR-17/18/19a/ 20a/19b-1/92-1 or miR-106a/20b/19b-2/92-2/ 363, as the causative event (Wang et al., 2006; Lum et al., 2007). These results confirm the significance of microRNAs in normal mammalian development and in cancer.
Cytogenet Genome Res 118:252–259 (2007)
259
FISH-based assay with lymphocytes cultured in the presence of aphidicolin to determine if a specific BAC is localized proximal, distal or within a CFS region of instability (Smith et al., 1998). However, many of the CFS regions were localized by determining that 46% of HPV16 (Thorland et al., 2003) and 65% of HPV18 positive cervical cancers (Ferber et al., 2003) were found to have the viral genome integrated within a different CFS region. A total of 41 CFS regions have been localized and nine have been fully characterized to determine precisely their full region of instability (Callahan et al., 2003; Huang et al., 1998a, b; Morelli et al., 2002; Rozier et al., 2004; Smith et al., 2006). These analyses have revealed that the size of different CFS’s region of instability can vary from under 1 Mb to over 10 Mb in size. FRA3B, the most frequently expressed CFS, spans 4.5 Mb (Becker et al., 2002). The highly expressed FRAXB CFS region is less than 500 kb (Arlt et al., 2002). However, the relatively infrequently expressed FRA9E CFS spans over 9 Mb (Callahan et al., 2003). Thus, there is no relationship between the size of a specific CFS region and the frequency that that region is expressed. Three of the four most frequently expressed CFS regions (FRA3B, FRA16D and FRA6E) were found to span genes that were encoded by large chromosomal regions. These were the 1.5 Mb FHIT gene (Zimonjic et al., 1997), the 1.0 Mb WWOX gene (Bednarak et al., 2000), and the 1.36 Mb PARK2 gene (Denison et al., 2003a). In spite of the fact that these genes span large chromosomal regions, their final processed transcripts are relatively small: FHIT – 1,095 bp, WWOX – 2,160 bp and PARK2 – 2,960 bp. These genes are frequently inactivated in multiple tumor types (Baffa et al., 1998; Huebner et al., 1998; Pekarsky et al., 2002; Denison et al., 2003a, b; Ludes-Meyer et al., 2003; Finnis et al., 2005). Two of these genes, FHIT and WWOX, have been demonstrated to function as tumor suppressor genes, both in vitro and in vivo (Siprashvili et al., 1997; Dumon et al., 2001; Sevignani et al., 2003; Fabbri et al., 2005). It is still not known if PARK2 functions as a tumor suppressor, but it is inactivated in multiple different tumor types (Cesari et al., 2003; Wang et al., 2004) and over-expression of PARK2 suppresses tumor growth in vitro (Denison et al., 2003a). In addition, the inactivation of expression of FHIT and/ or WWOX is associated with a worse clinical outcome (Lee et al., 2001; Toledo et al., 2004; Arun et al., 2005; Nunez et al., 2005a, b; Guerin et al., 2006), and these two genes are coordinately inactivated in breast cancers (Guler et al., 2004). The 730 kb RORA gene was found to be another large CFS gene and it was also inactivated in several different cancers (Zhu et al., 2006). In addition, over-expression of RORA resulted in decreased growth of MCF12 cells (Zhu et al., 2006), similar to what was observed with PARK2. Thus other large CFS genes may also function as tumor suppressors. Our group previously examined a number of extremely large genes to determine if they spanned CFS regions and this analysis revealed six new large CFS genes (Smith et al., 2006). Other groups subsequently identified a couple of other large CFS genes (Morelli et al., 2002; Savalyeva et al.,
2006) and by combining all this data it appears that about half of the CFS regions span other very large CFS genes. Twenty large genes have already been determined to reside within CFS regions, and we estimate that there may be 40– 50 of these large CFS genes throughout the genome (Smith et al., 2006). Some of the newly identified large CFS genes include DAB1 (the human homolog of the disabled locus from Drosophila), DLG2 (one of the human homologues of the large discs tumor suppressor from Drosophila), GRID2 (the delta2 glutamate receptor) (Rozier et al., 2004), DMD (the Duchenne Muscular Dystrophy gene), IL1RAPL1 (interleukin 1 receptor accessory protein Toll/Il-1 receptor), and CNTNAP2 (the contactin-associated protein-like 2 gene and the largest known gene spanning 2.3 Mb). In addition to being large CFS genes that may be important tumor suppressor genes, a number of the large CFS genes have been found to be important for normal neurological development (Smith et al., 2006). In this report three very large genes were tested as potential CFS genes, DCC, A2BP1 and RAD51, but none of these were localized within a CFS region. The expression of 13 of the 20 known large CFS genes were examined in breast, ovarian, endometrial, and brain cancers to determine whether there was random inactivation of expression of these genes due to genomic instability within cancer cells. One goal was to examine whether inactivation of expression of these genes was associated with cancers that generally have a worse overall clinical outcome. Real-time RT-PCR analysis with primers derived from within the most 3ⴕ exon of each large gene was utilized and the expression of these genes was compared against that of two control genes, actin and GAPDH. We demonstrate that inactivation of expression of these genes is not due to random instability and that specific large CFS genes are inactivated in different cancers. In addition there was much more loss of expression of these genes in the highly lethal brain cancers. Materials and methods Fluorescence in situ hybridization (FISH) FISH can be performed to detect the expression of CFSs and to determine if a candidate gene is actually derived from within that region of instability. To determine if several of the largest human genes were derived from within CFS regions we took BAC clones that spanned the center of each candidate gene and fluorescently labeled these clones to utilize them as FISH-based probes to test whether they were derived from within a CFS region. The genes that we tested as possible large CFS genes were the 1.0 Mb deleted in colorectal cancer (DCC) gene (Fearon et al., 1990), the 1.8 Mb ataxin-2 binding protein (A2BP1) (Shibata et al., 2000), and the 1.4 Mb RAD51 gene (Thacker, 2005). Cell culture and FISH protocols followed previously published methods (Smith et al., 2006). Briefly, metaphase cell preparations were prepared from mitogen-stimulated peripheral blood cultures obtained from normal individuals. Cultures were established with 9.5 ml RPMI 1640, 10% fetal bovine serum, 100 U/ml penicillin, 100 g/ml streptomycin, 0.5 ml lymphocyte-rich blood, and 10 g/ml PHA (Irvine Scientific, Santa Anna, CA, USA) and incubated at 37 ° C in 5% CO2 for 72 h. Twenty-four hours prior to harvest, select cultures were inoculated with 0.4 M APC (Sigma, St Louis, MO, USA). Cell harvest and slide preparation followed routine cytogenetic techniques. Purified BAC DNA was biotin labeled using a nick translation kit (Invitrogen)
Cytogenet Genome Res 118:260–269 (2007)
261
according to the manufacturer’s protocols. The probe is precipitated and then hybridized to the APC treated metaphase chromosomes. Following incubation and a series of wash steps, detection of the probe signal was performed by applying 75 l of an avidin, 5% bovine serum albumin mixture to each slide and incubating. After washing the slides, chromosomes were counterstained with DAPI (Vector Laboratories, Burlingame, CA). Photomicroscopy was performed using a Zeiss Axioplan fluorescence microscope equipped with MacProbe software (Applied Imaging, San Jose, CA).
Ovarian cancer panel We examined the expression of 13 large CFS genes in a panel of ovarian cancer cell lines and primary tumors. Patient samples were obtained after Institutional Review Board approval. Each primary ovarian tumor sample was cut, hematoxylin and eosin (H&E) stained and examined by qualified pathologists to obtain samples that were at least 80% tumor. Total RNA was isolated from each cell line using Gentra VersaGene Cell kits and from each primary tumor using Genetra VersaGene Tissue kits. The RNA was converted into cDNA using oligo-dT priming and then diluted 1/10 or more for real-time RT-PCR analysis. We analyzed six ovarian cancer cell lines (SKOV3, OV167, OV177, OV202, OV207 and OVCAR5) and 13 primary ovarian tumors (all serous ovarian cancer). The expression in these cancer-derived cell lines or primary tumors was compared to the expression of four short term cultures of normal ovarian surface epithelium. Breast cancer panel We also examined the expression of the CFS genes in a panel of breast cancer cell lines and primary tumors. We analyzed seven breast cancer cell lines (BT20, HCC70, MDA157, MDA468, UACC893, BT474 and T70) and 12 primary breast tumors. Primary fresh frozen breast tumors were obtained after Institutional Review Board approval. Each primary breast cancer was cut, H&E stained and examined by pathologists to obtain samples that were at least 80% tumor. The expression in these cell lines and primary tumors was compared to the expression of five normal breast epithelium samples from different individuals. Endometrial cancer panel We examined the expression of the CFS genes in a panel of primary endometrial cancers. There were a total of 20 primary endometrioid endometrial cancers that were examined. The primary endometrial cancers examined were fresh-frozen and obtained after Institutional Review Board approval. Each primary endometrial tumor was cut, H&E stained and examined by pathologists to obtain samples that were at least 80% tumor. To compare the expression in these primary tumors we also obtained normal endometrium from individuals without endometrial cancer representing atrophic, proliferative and secretory endometrium. We compared the expression in the primary endometrial cancers and considered a specific CFS gene to be aberrantly expressed in a specific tumor sample if the expression in that sample was at least four-fold lower than the lowest expression observed in the different normal endometrial samples. Brain cancer panel We finally examined the expression of the CFS genes in a panel of brain cancer-derived cell lines and brain cancer xenografts obtained from an invasive intracranial xenograft model of glioblastoma multiforme (Giannini et al., 2005). There were a total of eight brain cancer cell lines (T98H, H4, U87, TP365, U231, U138, A172, G32) and 13 xenografts derived from glioblastoma multiforme (Giannini et al., 2005). We compared the expression in these samples to the expression in four normal brain samples. RT-PCR and real-time RT-PCR analysis of expression of the CFS genes We first constructed PCR primers for regular RT-PCR analysis by targeting either the 3ⴕ exon of each gene or PCR primers which flanked several of the last exons (to control for possible DNA contamination). We constructed primers for each of the large CFS genes. These were then used to amplify cDNA constructed from oligo dT primed cDNA
262
Cytogenet Genome Res 118:260–269 (2007)
synthesis from total RNA isolated from each normal sample, cancerderived cell line or primary tumor to be characterized. Primers were optimized for best annealing temperature and then used to amplify DNA from the various samples before running PCR products on 0.8– 2% agarose gels (depending upon the size of the PCR products in each reaction). To determine the precise expression of the large CFS genes we constructed PCR primers optimized for real-time RT-PCR (amplicons of 120–130 bp in size and primers constructed from the most 3ⴕ exon of each large gene) and used -actin and GAPDH as control genes. RORA has two expressed isoforms (RORA-1 and RORA-4). PCR primers were constructed to test the level of each isoform individually. The real-time RT-PCR primers that were constructed for each large CFS gene are shown in Table 1. We then determined the optimal concentration of primers for each gene that gave a linear curve relating the input concentration of cDNA and performed real-time RT-PCR analysis in the ABI 7900HT Fast Real-Time PCR system. Each sample was run in triplicate with each tested gene or isoform. To determine the expression of each gene we determined the threshold cycle (CT) for each gene for each sample and then subtracted the CT for the more abundant control -actin gene from the CT obtained for each CFS gene. We also included cDNA obtained from RNA isolated from four normal brains and then compared the differential CT measurements in the normal brain samples to those obtained in the brain cancer cell lines and brain xenografts. For the ovarian cancer panel we compared the expression in the ovarian cancer cell lines and primary tumors to several short term cultures of normal ovarian surface epithelium (REF). For the breast cancer panel we compared the expression in the breast cancer cell lines and primary tumors to that of normal breast epithelium. For the endometrial cancers we compared the expression in primary endometrial tumors to normal endometrium. We considered any change of greater than four-fold (a delta CT of more than 2 greater or less than the range of delta CT’s observed in the normal samples) as a significant change in expression.
Results
Not all large genes are derived from within CFS regions Examination of the lists of the human genes that spanned the largest genomic regions revealed that many of these genes were derived from chromosomal regions known to contain a CFS (Smith et al., 2006). Two very interesting large genes that are derived from a chromosomal region known to contain a CFS are the 1.0 Mb deleted in colorectal cancer (DCC) gene (Fearon et al., 1990) which is localized within chromosomal band 18q21.2 which also contains the FRA18C CFS and the 750 kb RAD51 gene (the human homolog of the recA gene) which is localized within chromosomal band 14q24.1 that also contains FRA14A. BAC clones which spanned the center of these genes were hybridized to metaphase chromosomes isolated from cells exposed to aphidicolin and we analyzed over 20 metaphases with good discernible decondensation/breakage (D/C) within FRA18C and over 20 with D/C within FRA14A. However, these two BAC clones always hybridized proximal to the D/C within these two CFS regions; hence these genes are not localized within their respective CFS regions. The third gene that was analyzed was the 1.8 Mb ataxin2 binding protein gene (A2BP1) (Shibata et al., 2000). This gene is localized within chromosomal band 16p13.2. However, there are no known CFSs on the short arm of chromo-
Table 1. List of the primers used for real-time RT-PCR. Included on this table are the various genes that were analyzed including -actin (ACTB) and GAPDH (the two control genes) and the various large common fragile site genes. There are two expressed isoforms for RORA (RORA 1 and RORA 4) and the primers were chosen to differentiate between these two.
Gene
Primer sequences
pmol/l
ACTB
L: TCC TCT CCC AAG TCC ACA CA R: GCA CGA AGG CTC ATC ATT CA L: CAT GGC CTC CAA AGG AGT AAG AC R: TCT CTT CCT CTT GTG CTC TTG CT L: TGA TGA AGT GGC CGA TTT R: GGA AAG AAC ATG GAC GTG AAC L: GGG CTG GCC TTC TCC TAC TT R: TGG CTC AGT CTT GGT CAG GAT L: AGG AGC CTT TCC AGA ACT ACC A R: CCT GGG ACT TGC AGA TTG G L: CCC TCT GGA CAT TGA CAC TTT G R: TCT GGA TCT GCT CTG GGA TAA A L: GGT AGG GCT ATG AGG CAT GTT AC R: TTT CTG GTG CAA AGT GCC AAT L: CAG CTC TAT CCC ATT GCC ATC T R: TTC TTG TTC TAG CTT AAT TGC TCG AT L: TTC AGA ATT TGA GCC AGC AA R: TCT GTG CAG GGC CAT ATA AA L: GCA GCG ATG AAA GCT CAA AAT R: GGA CAG GAG TAG GTG GCA TTG L: TGC CAG TGA TCC TAC CAC AGA R: ACT GGG CTC ACC AAA TGG AT L: CAA GTG TTC AGG TTG TGT CAA GAA R: AGC AGG GTA TAA TGT GGT GAG GA L: TGA GAG CTT TAT TGC TGC ATT T R: CAT GCC ATG TGA TGT TTA TGC L: ACG GTC ATT AAA TGG CAT GG R: GCC CTT GCT CAC TGA CAT CT L: TGC CCA TGC CAC ACA TAT C R: CAA CTC AGG TAG ATT CAT GCA GTC AT L: CCT CTG GTT CTT GCC TGC TT R: CTG TGT GAA GGT GAA GGT GGT T
0.075
GAPDH FHIT WWOX PARK2 GRID2 NBEA DLG2 RORA1 RORA4 DAB1 CNTNAP2 DMRTA1/DMO IL1RAPL1 IMMP2L LARGE
some 16. We decided to analyze thousands of metaphases from lymphocytes cultured in the presence of 0.4 M aphidicolin to see if we could identify a CFS region from this chromosome arm that was expressed at very low frequencies. This analysis did not detect any reproducible CFS breakage on this chromosomal arm; hence A2BP1 is another large gene that is not a CFS gene. Several other investigators have now identified a number of other very large genes that are localized within CFS regions including GRID2 localized in FRA4G (Rozier et al., 2004), IMMP2L , localized in FRA7K (Helmrich et al., 2007) and neurobeachin localized within FRA13A (Savalyeva et al., 2006). Combining this with the already known CFS genes including FHIT, WWOX, and PARK2 there are now a total of 20 known large CFS genes. Thus six of the ten of the largest human genes (CNTNAP2, DMD, LRP1B, CTNNA3, DAB1, and FHIT) were derived from within CFS regions (Smith et al., 2006). We also found that 14/31 of the partially characterized CFS regions were associated with one or more very large genes. Table 2 lists the 20 genes that are now known to be large CFS genes, their genomic organization (number of exons and size of their final processed transcript), and the CFS region that spans them.
0.15 0.15 0.075 0.15 0.15 0.15 0.25 0.15 0.25 0.15 0.075 0.15 0.075 0.25 0.25
Table 2. Large common fragile site genes
Gene
Chromosome Size (bp)
Exons/FPTa CFS
CNTNAP2 DMD LRP1B CTNNA3 DAB1 FHIT KIAA1680 GRID2 DLG2 PARK2 IL1RAPL1 WWOX PDGFRA/FIPL IMMP2L RORA PTPRG NBEA LARGE ARHGAP15 ATXN1/SCA1
7q35 Xp21.1 2q22.1 10q21.3 1p32.3 3p14.2 4q22.1 4q22.3 11q14.1 6q26 Xp21.1 16q23.2 4q12 7q31.1 15q22.2 3p14.2 13q13.3 22q12.3 2q22.3 6p22.3
25/8107 79/13957 91/16556 18/3024 21/2683 9/1095 11/5833 16/3024 23/3071 12/2960 11/2722 9/2264 24/2550 6/1540 11/1816 30/4707 58/10812 16/4323 14/1706 11/10926
a
2304258 2092287 1900275 1775996 1548827 1499181 1474315 1467842 1463760 1379130 1368379 1113013 917434 899238 731967 731390 730417 647480 638958 462345
FRA7I FRAXC FRA2F FRA10D FRA1B FRA3B FRA4? FRA4? FRA11F FRA6E FRAXC FRA16D FRA4B FRA7K FRA15A FRA3B FRA13A FRA22B FRA2F FRA6C
FPT: size of final processed transcript (bp).
Cytogenet Genome Res 118:260–269 (2007)
263
14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Real-time RT-PCR analysis of ovarian tumors The expression of the 13 large CFS genes was examined in a panel of ovarian cancer cell lines and primary tumors. For a control RNA was obtained from brushings of ovarian surface epithelium from normal ovaries. The expression in these normal samples was compared to that in several ovarian cancer cell lines and from 13 primary ovarian cancers. The expression of each of these large CFS genes was compared to that of the abundantly expressed actin gene as well as the lower expressed GAPDH gene. Several very interesting features were found in this analysis. The first is that some ovarian tumors had inactivation of expression of just a few of the tested genes, while others had inactivation of expression of multiple large CFS genes. Some of the large CFS genes were not inactivated in any of the ovarian tumors tested, while others were inactivated in many of the ovarian tumors. The two large CFS genes that were most frequently under-expressed in ovarian cancers were DLG2 and RORA 4 (which were under-expressed in 16/21 and 12/21 of the ovarian cancers tested, respectively). CFS genes from the most highly expressed CFSs, FHIT (FRA3B), WWOX (FRA16D), and PARK2 (FRA6E) were inactivated only in 1/21, 6/21 and 0/21 of the ovarian cancers tested, respectively. This demonstrates that it is not merely the instability within the CFS region that promotes gene inactivation as DLG2 and RORA are both derived from CFS regions that are expressed at much lower levels than FRA3B, FRA16D, and FRA6E. Figure 1 shows results we obtained when we analyzed the expression of these genes in ovarian cancers utilizing real-time RT-PCR. From this figure it is readily observable that one tumor had normal expression of all the tested CFS genes, some had decreased expression of
264
LARGE IMMP2L IL1RAPL1 DMD CNTNAP2 Dab1 RORA-4 RORA-1 DLG2 Nbea Grid2 Parkin WWOX FHIT
SK OV OV 3 16 OV 7 1 OV 77 20 OV 2 OV 2 07 CA *O R 5 V1 *O 34 V1 *O 85 V1 *O 92 V3 *O 96 V4 *O 25 V4 *O 61 V5 *O 26 V6 *O 32 V *O 946 V1 * O 00 6 V1 * O 03 9 V1 * O 26 9 V1 34 8
Fig. 1. Decreased expression of the 14 large CFS genes in ovarian cancer cell lines and primary tumors. The cell lines are shown on the left and the primary tumors (shown with a star) are on the right. One primary tumor, OV1269, had normal expression of all the tested genes. In contrast, several of the cell lines and primary tumors had decreased expression of four, or more, of the tested genes.
Number of Large CFS Genes Under-Expressed
Ovarian
Cytogenet Genome Res 118:260–269 (2007)
Specimen
one or two of the tested genes, and a third group had altered expression of multiple large CFS genes. While the majority of the changes in expression observed in ovarian tumors corresponded to decreased expression of one of the large CFS genes, there were a few instances where we observed increased expression of one or more of the large CFS genes. DMD was over-expressed in four ovarian tumor samples, and IL1RAPL1 and IMMP2L were over-expressed in two of the tested samples. Real-time RT-PCR analysis of breast cancer We next examined the expression of the 13 large CFS genes in a panel of breast cancer cell lines and primary tumors. The results in the breast tumors were very similar to what was observed in the ovarian tumors. None of the breast tumors had normal expression for each of the tested genes. There were two breast tumors with decreased expression of one gene, three with decreased expression of two genes, and five with decreased expression of three genes. There were also a number of breast tumors that had decreased expression of multiple large CFS genes. The data obtained with breast cancer is summarized in Fig. 2. Interestingly, the genes with the greatest changes in expression in breast cancer were DMD (decreased expression in 15/19 breast tumors examined), IL1RAPL1 (loss in 13/19 of the breast tumors), and RORA isoform 4 (decreased expression in 12/19 of the breast tumors). This is distinct from what was observed in ovarian cancers. All three genes are derived from CFS regions that are not frequently expressed (as compared to FRA3B, FRA16D and FRA6E); once again there was no correlation between the frequency of expression of a specific CFS region and the frequency that the large
Number of Large CFS Genes Under-Expressed
BREAST LARGE IMMP2L IL1RAPL1 DMD CNTNAP2 Dab1 RORA-4 RORA-1 DLG2 Nbea Grid2 Parkin WWOX FHIT
BT H C 20 C MD 70 A1 MD 57 UA A468 CC 89 BT 3 47 4 T7 0 *B C9 83 * 98 * 10 6 10 * 10 00 * 10 01 * 10 03 * 10 05 * 10 06 * 10 07 * 10 09 * 10 11 * 10 17
Fig. 2. Decreased expression of the 14 large CFS genes in breast cancer cell lines and primary tumors. The cell lines are shown on the left and the primary tumors (shown with a star) are on the right.
14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Specimen
ENDOMETRIAL 14 13
LARGE IMMP2L
11
IL1RAPL1
10
DMD
9
CNTNAP2
8
Dab1
7
RORA-4
6
RORA-1
5
DLG2
4
Nbea
3
Grid2
2
Parkin
1
WWOX
CFS gene from that region was altered in ovarian or breast cancers. We did not detect over-expression of any of the tested genes in this panel of breast cancer cell lines and primary tumors. Real-time RT-PCR analysis of endometrial cancer The third cancer examined for the expression of the large CFS genes was endometrial cancer. We examined a panel consisting of eight normal endometrial specimens and 20 primary endometrial cancers. The normal endometrial specimens were derived from women at different stages in
34
31
21
18
15
13
8
6
3
0 1
Fig. 3. Decreased expression of the 14 large CFS genes in primary endometrial tumors. Six of the endometrial tumors, EC1, 17, 18, 20, 21, 31 and 32, had normal expression of all the tested genes. In contrast, seven endometrial tumors, EC2, 3, 6, 13, 22, 34 and 35, had decreased expression of four, or more, of the tested genes.
Number of Large CFS Genes Under-Expressed
12
FHIT
Specimen
their menstrual cycle (as well as two samples from postmenopausal women) to provide epithelial cells representing secretary, proliferative and atrophic normal endometrial cells. In contrast to the ovarian cancers analyzed which consist of many late stage tumors, the majority of the endometrial cancers were early stage. The expression of the large CFS genes was examined first in the normal endometrial specimens and then in the primary tumors. The data summarizing the results observed with the endometrial cancers is shown in Fig. 3. Six of the endometrial samples had normal expression for all the test-
Cytogenet Genome Res 118:260–269 (2007)
265
IL1RAPL1 DMD CNTNAP2 Dab1 RORA-4 RORA-1 DLG2 Nbea Grid2 Parkin
Real-time RT-PCR analysis of brain cancer For brain cancers we had a panel consisting of RNA derived from four normal brains, eight brain cancer cell lines, and 13 brain cancer xenografts grown in mice. All of the tumors corresponded to glioblastoma multiforme, a highly aggressive and malignant tumor that is generally associated with a very poor clinical outcome. The expression of -actin, and GAPDH were also used here as controls and compared to the expression of the 13 large CFS genes and there was a very distinct pattern of expression when compared to
Cytogenet Genome Res 118:260–269 (2007)
*X 38 *X 44
*X 34 *X 36
*X 16 *X 22 *X 26
*X 14 *X 15
*X 5 *X 6 *X 8 *X 10
U23 1 U13 8 A17 2 G 32
H4 U87 TP3 65
WWOX
ed CFS genes. Seven of the tested samples showed decreased expression of one or two of the tested genes. However, there was a third group that had alterations in expression (generally inactivation) of multiple large CFS genes. In contrast to ovarian and breast tumors, there was no large CFS gene that was inactivated in the majority of the tumor specimens tested. Instead, there were multiple large CFS genes that had alterations in five to eight of the 20 tested endometrial tumors. Decreased expression of GRID2 and IL1RAPL1 was found in eight of 20 of the tested endometrial tumors. There was also decreased expression of DAB1, DLG2 and CNTNAP2 in six of 20 of the tested endometrial tumors, and decreased expression of CNTNAP2, NBEA and PARK2 in five of 20 of the tested endometrial tumors. The RORA 4 isoform which had decreased expression in a majority of the tested ovarian and breast tumors was normally expressed in most of the endometrial tumors. Only two tumors had decreased expression of RORA 4. In endometrial cancers there was also no relationship between instability in different CFS regions and loss of expression of the large gene from that region, as FHIT and WWOX had decreased expression in only two of the 20 tested endometrial specimens. None of the large CFS genes were over-expressed in any of the tested endometrial samples.
266
LARGE IMMP2L
T98H
Fig. 4. Decreased expression of the 14 large CFS genes in brain cancer cell lines and brain cancer xenografts. The cell lines are shown on the left and the xenografts (shown with a star) are on the right. Note that the cell lines had decreased expression of more of the tested genes than the xenografts and that each of the xenografts had decreased expression of seven, or more of the tested genes.
Number of Large CFS Genes Under-Expressed
BRAIN 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FHIT
Specimen
the panels of ovarian, breast, or endometrial cancers. Many of the tested large CFS genes were inactivated in most of the brain tumors examined. This is shown in Fig. 4. In the brain tumors examined there were three xenografts that had decreased expression of only two or three of the tested large CFS genes, but most of the brain cancer cell lines and xenografts examined had decreased expression of five or more of the tested genes. It is also interesting to note that the brain cancer cell lines had decreased expression of more of the tested CFS genes than the xenografts. The genes that had decreased expression in the majority of the brain cancers examined were DMD and IL1RAPL1 (which had decreased or no detectable expression in any of the tested brain tumors examined), CNTNAP2 and DLG2 (which had decreased expression in 17 of the 21 brain tumors examined), and LARGE (which had decreased expression in 14 of 21 of the brain tumors examined). None of these genes were over-expressed in any of the brain cancer cells lines or xenografts. Discussion
Many of the CFSs are derived from chromosomal regions that are consistently deleted during the development of many different cancer types (Smith et al., 1998; Sutherland et al., 1998). We were particularly interested in these consistently deleted regions of instability as attractive candidates for containing important tumor suppressor genes that could be involved in cancer development. The cloning and characterization of the most frequently expressed CFS region, FRA3B (3p14.2), revealed the extremely large 1.4 Mb FHIT gene which spanned the center of the 4.5 Mb FRA3B region of instability (Ohta et al., 1996; Sozzi et al., 1996; Becker et al., 2002). Subsequent studies revealed that this
gene was frequently altered during the development of many different cancers (Baffa et al., 1998; Huebner et al., 1998; Pekarsky et al., 2002) and that it functioned as a tumor suppressor, both in vitro and in vivo (Dumon et al., 2001; Fabbri et al., 2005; Sevignani et al., 2003). In addition, the inactivation of expression of this gene was associated with a worse clinical phenotype in different cancers (Arun et al., 2005; Guerin et al., 2006; Toledo et al., 2004). The second most frequently expressed CFS region, FRA16D (16q23.2), was also found to be a large region of genomic instability (2.0 Mb) (Krummel et al., 2000) and this region spanned another large gene, the 1.0 Mb WWOX gene (Bednarak et al., 2000). This gene was also a target for alterations in multiple tumor types (Finnis et al., 2005; Ludes-Meyer et al., 2003) and has been demonstrated to function as a tumor suppressor, both in vitro and in vivo (Fabbri et al., 2005). Recently targeted deletion of the Wox1 gene in mice revealed that these mice developed osteosarcomas and lung papillary carcinomas further demonstrating that this gene does function as a tumor suppressor (Aqeilan et al., 2007). FHIT and WWOX are coordinately inactivated in different cancers (Guler et al., 2004) and inactivation of expression of these genes is associated with a worse clinical phenotype (Toledo et al., 2004; Arun et al., 2005; Nunez et al., 2005a, b; Guerin et al., 2006). Several other CFS regions were also found to be associated with extremely large genes, including FRA6E (6q26, which spans the 1.36 Mb PARK2 gene) (Denison et al., 2003a), FRA13A (13q13.3, which spans the 750 kb NBEA gene) (Savalyeva et al., 2006), and FRA15A (spanning the 730 kb RORA gene) (Zhu et al., 2006). These genes are also targets of alteration during the development of several different cancer types (Denison et al., 2003a, b; Zhu et al., 2006), but they have not yet been demonstrated to function as tumor suppressors. Thus, there appears to be an association between the highly unstable CFS regions and the presence of genes that span very large genomic segments. However, this is not the case for all the CFS regions. The highly expressed FRAXB CFS is not associated with any very large genes, although several smaller genes are spanned by this region of instability (Arlt et al., 2002). FRA7H is also highly unstable, and this region is not associated with any known genes, large or small (Mishmar et al., 1998). FRA7G is expressed at much lower frequencies and this 500 kb region of instability only spans a couple of relatively small genes, including caveolin1 and -2 (Engelman et al., 1998). Lists of the human genes that spanned the largest genomic regions were examined and there were more large genes than anticipated; indeed there are 40 genes that span over 1 Mb and 240 that span greater than 500 kb of genomic sequence. Many of these genes were localized to chromosomal regions known to contain a CFS. BAC clones that spanned the central portion of each gene were used as FISHbased probes to determine if the large gene was proximal, distal or was actually spanned by the CFS region. A number of genes were found not to reside within the CFS regions. In this report we demonstrate that three additional large genes,
DCC, RAD52 and A2BP1, are not localized within CFS regions. However, many of the previously tested large genes were within CFS regions. This includes DAB1, DLG2, DMD, IL1RAPL1 and CTNNA3. In addition, we found that six of the ten largest genes were derived from within CFS regions (Smith et al., 2006). What proportion of the CFSs are associated with large genes? Through a variety of different strategies we, and others, have localized 31 of the 90 known CFS regions. An examination of several Mb surrounding each of these regions revealed that about one-half of the CFS regions are associated with extremely large genes. A number of the CFS regions span two very large genes. This is observed in FRA3B (which spans the 1.36 Mb FHIT gene and the 700 kb PTPRG gene) (Becker et al., 2002), and FRAXC (which spans the 1.8 Mb IL1RAPL1 gene and the 2.0 Mb DMD gene) (Smith et al., 2006). We therefore estimate that there may be 40– 50 large CFS genes distributed throughout the human genome. A number of the very large CFS genes function in normal neurological development. This is certainly the case with RORA, as RORA-deficient mice have fewer ectopically localized Purkinje cells and display the cerebellar defects of staggerer (Dussault et al., 1998). Mutation of DAB1 in mice results in another neurological mutant called scrambler (Sheldon et al., 1997), and these mice have cerebellar hyperplasia with Purkinje cell ectopia. CNTNAP2 was found to be disrupted in a family with Gilles de la Tourette syndrome (Verkerk et al., 2003), and LRP1B and WWOX have possible roles in Alzheimer’s disease (Cam et al., 2004; Sze et al., 2004). Therefore is there a potential link between large CFS genes that are involved in normal neurological development and the development of cancer? Many of the large CFS genes are found to be inactivated in different cancers. However, how many of them function as important tumor suppressors? Only FHIT and WWOX have been proven to function as tumor suppressors both in vitro and in vivo (Siprashvili et al., 1997; Dumon et al., 2001; Sevignani et al., 2003; Fabbri et al., 2005). We have analyzed the effect of re-introducing PARK2 and RORA into cancerderived cell lines that do not produce them and have shown this is associated with decreased growth and increased sensitivity to chemotherapeutic agents (Denison et al., 2003b; Zhu et al., 2006). We are currently analyzing several other large genes to determine if they could also be potential tumor suppressor genes. An important question to ask is whether or not the alterations in the expression of the large CFS genes are merely due to their location within the highly unstable CFS regions. Our data suggests that this is not the case, as genes localized within the most frequently expressed CFS regions (FHIT, WWOX and PARK2) were not the most frequently altered in the cancers studied. Instead it was genes including RORA and DLG2, which are localized within CFS regions that are expressed at much lower frequencies that had the greatest inactivation in expression in the cancers studied. We have also demonstrated that different cancers have different CFS genes inactivated. In ovarian tumors DLG2
Cytogenet Genome Res 118:260–269 (2007)
267
and RORA 4 were inactivated in a majority of the tumors examined. However, in breast cancer it was DMD and IL1RAPL1 that were inactivated in a majority of the tested tumors. In endometrial cancer, none of the genes were inactivated in a majority of the tumors tested; however, a number of the genes were inactivated in a proportion of the tumors. In brain cancers, there was inactivation of expression of many of the CFS genes. The two genes which showed the greatest inactivation were DMD and IL1RAPL1 which were completely inactivated in every brain cancer cell line and xenograft examined. It is interesting to note that these two X-linked genes are inactivated in all tested brain cancers as the frequency of brain cancers in men is 1.5 times higher than in women (Deorah et al., 2006). Could these two genes be playing a very important role in the development of brain cancer and possibly explain why men are more susceptible to developing brain cancer than women? It has already been demonstrated that the inactivation of FHIT and WWOX is associated with a worse clinical outcome in several different tumors (Toledo et al., 2004; Arun et al., 2005; Nunez et al., 2005a, b; Guerin et al., 2006). Is it also possible that the inactivation of other large CFS genes is also associated with a poor clinical outcome? Our results in the different tumors examined would tend to support this. For example, there is much more inactivation of ex-
pression of the large CFS genes in the highly lethal brain cancers than in any of the other tested cancers. In addition, the brain cancer cell lines had greater inactivation than the brain cancer xenografts. We are currently analyzing cancers of the breast and ovary with good versus poor clinical outcome to determine if cancers that have more of these genes inactivated are associated with a poorer clinical outcome. This could potentially develop into a powerful diagnostic tool to analyze these types of cancer. We have therefore found that many of the human genes which span the largest genomic regions are associated with CFS regions of instability. A number of these genes appear to play important roles in normal neurological development. We have also found that there is frequent inactivation of the expression of these large CFS genes in different cancers, but that there is no association between the frequency that a particular CFS region is expressed and the inactivation of expression of the large CFS gene from that region. Instead, there appears to be selection for the inactivation of different CFS genes in different cancers. The inactivation of expression of multiple large CFS genes may be associated with a poorer clinical outcome, thus characterizing different cancers for the expression of these genes may help to differentiate cancers with a good versus a poor overall prognosis.
References Aqeilan RI, Trapasso F, Hussain S, Costinean S, Marshall D, et al: Targeted deletion of Wwox reveals a tumor suppressor function. Proc Natl Acad Sci USA 104:3949–3954 (2007). Arlt MF, Miller DE, Beer DG, Glover TW: Molecular characterization of FRAXB and comparative common fragile site instability in cancer cells. Genes Chromosomes Cancer 33: 82–92 (2002). Arun B, Kilic G, Yen C, Foster B, Yardley DA, et al: Loss of FHIT expression in breast cancer is correlated with poor prognostic markers. Cancer Epidemiol Biomarkers Prev 14: 1681–1685 (2005). Baffa R, Veronese ML, Santoro R, Mandes B, Palazzo JP, et al: Loss of FHIT expression in gastric carcinoma. Cancer Res 58: 4708–4714 (1998). Becker NA, Thorland EC, Denison SR, Phillips LA, Smith DI: Evidence that instability within the FRA3B region extends four megabases. Oncogene 21: 8713–8722 (2002). Bednarek AK, Laflin KJ, Daniel RL, Liao Q, Hawkins KA, Aldaz CM: WWOX, a novel WW domain-containing protein mapping to human chromosome 16q23.3–q24.1, a region frequently affected in breast cancer. Cancer Res 60: 2140–2145 (2000). Callahan G, Denison SR, Phillips LA, Shridhar V, Smith DI: Characterization of the common fragile site FRA9E and its potential role in ovarian cancer. Oncogene 22:590–601 (2003). Cam JA, Zerbinatti CV, Knisely JM, Hecimovic S, Li Y, Bu G: The low density lipoprotein receptor-related protein 1B retains beta-amyloid precursor protein at the cell surface and reduces amyloid-beta peptide production. J Biol Chem 279:29639–29646 (2004).
268
Cesari R, Martin ES, Calin GA, Pentimalli F, Bichi R, et al: Parkin, a gene implicated in autosomal recessive juvenile parkinsonism, is a candidate tumor suppressor gene on chromosome 6q25– q27. Proc Natl Acad Sci USA 100: 5956–5961 (2003). Denison SR, Callahan G, Becker NA, Phillips LA, Smith DI: Characterization of FRA6E and its potential role in autosomal recessive juvenile parkinsonism and ovarian cancer. Genes Chromosomes Cancer 38: 40–52 (2003a). Denison SR, Wang F, Becker NA, Schule B, Kock N, et al: Alterations in the common fragile site gene Parkin in ovarian and other cancers. Oncogene 22: 8370–8378 (2003b). Deorah S, Lynch CF, Bibenaller ZA, Ryken TC: Trends in brain cancer incidence and survival in the United States: Surveillance, epidemiology, and end results program, 1973 to 2001. Neurosurgical Focus 20: 1 (2006). Dumon KR, Ishii H, Fong LY, Zanesi N, Fidanza V, Mancini R, et al: FHIT gene therapy prevents tumor development in Fhit-deficient mice. Proc Natl Acad Sci USA 98: 3346–3351 (2001). Dussault I, Fawcett D, Matthyssen A, Bader JA, Giguere V: Orphan nuclear receptor ROR alpha-deficient mice display the cerebellar defects of staggerer. Mech Dev 70: 147–153 (1998). Engelman JA, Zhang XL, Lisanti MP: Genes encoding human caveolin-1 and -2 are co-localized to the D7S522 locus (7q31.1), a known fragile site (FRA7G) that is frequently deleted in human cancers. FEBS Lett 436: 403–410 (1998). Fabbri M, Iliopoulos D, Trapasso F, Aqeilan FI, Cimmino A, et al: WWOX gene restoration prevents lung cancer growth in vitro and in vivo. Proc Natl Acad Sci USA 102: 15611–15616 (2005).
Cytogenet Genome Res 118:260–269 (2007)
Fearon ER, Cho KR, Nigro JM, Kern SE, Simons JW, et al: Identification of a chromosome 18q gene that is altered in colorectal cancers. Science 247: 49–56 (1990). Ferber MJ, Thorland EC, Brink AA, Rapp AK, Phillips LA, et al: Preferential integration of human papillomavirus type 18 near the c-myc locus in cervical carcinoma. Oncogene 22:7233–7242 (2003). Finnis M, Dayan S, Hobson L, Chenevix-Trench G, Friend K, et al: Common chromosomal fragile site FRA16D mutation in cancer cells. Hum Mol Genet 14: 1341–1349 (2005). Giannini C, Sarkaria JN, Saito A, Uhn JH, Galanis E, et al: Patient tumor EGFR and PDGFRA gene amplifications retained in an invasive intracranial xenograft model of glioblastoma multiforme. Neuro-oncology 7: 164–176 (2005). Glover TW, Berger C, Coyle J, Echo B: DNA polymerase alpha inhibition by aphidicolin induces gaps and breaks at common fragile sites in human chromosomes. Hum Genet 67: 136–142 (1984). Guerin LA, Hoffman HT, Zimmerman MB, Robinson RA: Decreased fragile histidine gene protein expression is associated with worse prognosis in oral squamous carcinoma. Arch Pathol Lab Med 130:158–164 (2006). Guler G, Uner A, Guler N, Han SY, Iliopoulos D, et al: The fragile genes FHIT and WWOX are inactivated coordinately in invasive breast carcinoma. Cancer 100: 1605–1614 (2004). Helmrich A, Stout-Weider K, Matthaei A, Hermann K, Heiden T, Schrock E: Identification of the human-mouse syntenic common fragile site FRA7K/Fra12C1 – relation of FRA7K and other human common fragile sites on chromosome 7 to evolutionary breakpoints. Int J Cancer 120: 48–54 (2007).
Huang H, Qian C, Jenkins RB, Smith DI: FISH mapping of YAC clones at human chromosomal band 7q31.2: identification of YACs spanning FRA7G within the common region of LOH in breast and prostate cancer. Genes Chromosomes Cancer 21: 152–159 (1998a). Huang H, Qian J, Proffit J, Wilber K, Jenkins R, Smith DI: FRA7G extends over a broad region: coincidence of human endogenous retroviral sequences (HERV-H) and small polydispersed circular DNAs (spcDNA) and fragile sites. Oncogene 16: 2311–2319 (1998b). Huebner K, Garrison PN, Barnes LD, Croce CM: The role of the FHIT/FRA3B locus in cancer. Ann Rev Genet 32:7–31 (1998). Krummel KA, Roberts LR, Kawakami M, Glover TW, Smith DI: The characterization of the common fragile site FRA16D and its involvement in multiple myeloma translocations. Genomics 69: 37–46 (2000). Lee JI, Soria JC, Hassan K, Liu D, Tang X, et al: Loss of Fhit expression is a predictor of poor outcome in tongue cancer. Cancer Res 61: 837–841 (2001). Ludes-Meyer JH, Bednarek AK, Popescu NC, Bedford M, Aldaz CM: WWOX, the common chromosomal fragile site, FRA16D, cancer gene. Cytogenet Genome Res 100: 101–110 (2003). Mishmar D, Rahat A, Scherer SW, Nyakatura G, Hinzmannn B, et al: Molecular characterization of a common fragile site (FRA7H) on human chromosome 7 by the cloning of a simian virus 40 integration site. Proc Natl Acad Sci USA 95: 8141–8146 (1998). Morelli C, Karayianni E, Magnanini C, Mungall AJ, Thorland E, et al: Cloning and characterization of the common fragile site FRA6F harboring a replicative senescence gene and frequently deleted in human tumors. Oncogene 21: 7266– 7276 (2002). Nunez MI, Ludes-Meyers J, Abba MC, Kil H, Abbey NW, et al: Frequent loss of WWOX expression in breast cancer: correlation with estrogen receptor status. Breast Cancer Res Treat 89: 99– 105 (2005a).
Nunez MI, Rosen DG, Ludes-Meyers JH, Abba MC, Kil H, et al: WWOX protein expression varies among ovarian carcinoma histotypes and correlates with less favorable outcome. BMC Cancer 5:64 (2005b). Ohta M, Inoue H, Cotticelli MG, Kasstury K, Baffa R, et al: The FHIT gene, spanning the 3p14.2 fragile site and renal carcinoma-associated t(3;8) breakpoint, is abnormal in digestive tract cancers. Cell 84: 587–597 (1996). Pekarsky Y, Zanesi N, Palamarchuk A, Huebner K, Croce CM: FHIT: from gene discovery to cancer treatment and prevention. Lancet Oncol 3:748– 754 (2002). Rozier L, El-Achkar E, Apiou F, Debatisse M: Characterization of a conserved aphidicolin-sensitive common fragile site at human 4q22 and mouse 6C1: possible association with an inherited disease and cancer. Oncogene 23: 6872– 6880 (2004). Savelyeva L, Sagulenko E, Schmitt JG, Schwab M: The neurobeachin gene spans the common fragile site FRA13A. Hum Genet 118: 551–558 (2006). Sevignani C, Calin GA, Cesari R, Sarti M, Ishii H, et al: Restoration of fragile histidine triad (FHIT) expression induces apoptosis and suppresses tumorigenicity in breast cancer cell lines. Cancer Res 63: 1183–1187 (2003). Sheldon M, Rice DS, D’Arcangelo G, Yneshima H, Nakajima K, et al: Scrambler and yotari disrupt the disabled gene and produce a reeler-like phenotype in mice. Nature 389: 730–733 (1997). Shibata H, Huynh DP, Pulst SM: A novel protein with RNA-binding motifs interacts with ataxin-2. Hum Mol Genet 9:1303–1313 (2000). Siprashvili Z, Sozzi G, Barnes LD, McCue P, Robinson AK, et al: Replacement of Fhit in cancer cells suppresses tumorigenicity. Proc Natl Acad Sci USA 94:13771–13776 (1997). Smith DI, Huang H, Wang L: Common fragile sites and cancer (review). Int J Oncol 12: 187–196 (1998). Smith DI, Zhu Y, McAvoy S, Kuhn R: Common fragile sites, extremely large genes, neural development and cancer. Cancer Lett 232: 48–57 (2006).
Sozzi G, Veronese ML, Negrini M, Baffa R, Cotticelli MG, et al: The FHIT gene is abnormal in lung cancer. Cell 85: 17–26 (1996). Sutherland GR, Parslow MI, Baker E: New classes of common fragile sites induced by 5-azacytidine and bromodeoxyuridine. Hum Genet 69: 233– 237 (1985). Sutherland GR, Baker E, Richards RI: Fragile sites still breaking. Trends Genet 14: 501–506 (1998). Sze CI, Su M, Pugazhenthi S, Jambal P, Hsu LJ, et al: Down-regulation of WW domain-containing oxidoreductase induces Tau phosphorylation in vitro: A potential role in Alzheimer’s disease. J Biol Chem 279: 30498–30506 (2004). Thacker J: The RAD51 gene family: genetic instability and cancer. Cancer Lett 219: 125–135 (2005). Thorland EC, Myers SL, Gostout BS, Smith DI: Common fragile sites are preferential targets for HPV16 integrations in cervical tumors. Oncogene 22: 1225–1237 (2003). Toledo G, Sola JJ, Lozano MD, Soria E, Pardo J: Loss of FHIT protein expression is related to high proliferation, low apoptosis and worse prognosis in non-small-cell lung cancer. Mod Pathol 17:440–448 (2004). Verkerk AJ, Mathews CA, Hoosse M, Eussen BH, Heutink P, Oostra BA: Tourette Syndrome Association International Consortium for Genetics. CNTNAP2 is disrupted in a family with Gilles de la Tourette syndrome and obsessive compulsive disorder. Genomics 82: 1–9 (2003). Wang F, Denison S, Lai JP, Phillips LA, Montoya D, et al: Parkin gene alterations in hepatocellular carcinoma. Genes Chromosomes Cancer 40: 85–96 (2004). Zhu Y, McAvoy S, Kuhn R, Smith DI: RORA, a large common fragile site gene, is involved in cellular stress response. Oncogene 25: 2901–2908 (2006). Zimonjic DB, Druck T, Ohta M, Kastury K, Croce CM, et al: Positions of chromosome 3p14.2 fragile sites (FRA3B) within the FHIT gene. Cancer Res 57:1166–1170 (1997).
Cytogenet Genome Res 118:260–269 (2007)
269
shelterin. Besides determining the structure of telomeric sequences (the T-loop), shelterin controls the synthesis of telomeric DNA and has a remodelling activity that changes the telomeric structure in order to protect the chromosome ends from being recognized as double stranded DNA breaks (de Lange, 2005). A recent study has also shown that telomeres recruit MRE11, phosphorylated NBS1 and ATM in every G2 phase of the cell cycle (Verdun et al., 2005). This protein recruitment coincides with a partial release of telomeric POT1. Based on these, and complementary data, it has been suggested that localized DNA damage response at telomeres after replication is essential for recruiting the processing machinery promoting formation of the chromosome end protection complex. Physiological telomere length regulation
DNA polymerases are unable to completely replicate terminal DNA and telomeres in somatic cells progressively lose approximately 50–200 nucleotides during each mitotic replication. The TTAGGG repeats act as a buffer zone to prevent erosion of coding sequences due to this so-called end replication problem. Once telomeres become too short to maintain T-loop structure, they may be recognised as double strand breaks and thereby trigger cellular senescence or apoptosis (Karlseder et al., 2002). Telomere shortening thus protects the cell from gene loss but at the same time limits cell division to a finite number of cycles in most cells. However, the physiological role of some cell lineages such as stem cells, activated lymphocytes, and germ cells, is dependent on continuous cell proliferation. The most common mechanism to maintain telomere length under continuously proliferative conditions is by the expression of telomerase (Cao et al., 2002). Telomerase is a large ribonucleoprotein complex containing two core subunits, the human telomerase reverse transcriptase (hTERT) and the telomerase RNA template (TER). TER binds to the 3 overhang of telomeric DNA with a complementary sequence, providing a template for reverse transcription and the synthesis of new telomeric repeats (Blackburn, 2001). In human tissues, telomerase activity changes through life, going from a peak of activity during the first trimester in utero, where virtually all the tissues have active telomerase (Forsyth et al., 2002), to undetectable levels after birth in most somatic tissues. However, telomerase activity is present after birth in highly proliferative cells such as primary germ line cells, activated lymphocytes, some smooth muscle cells and fibroblasts, and in reproductive organs such as endometrium, testis, and ovary (Lehner et al., 2002). The interest has been high in defining mechanisms that regulate telomerase activity, with a special focus on the regulation of hTERT, the active concentration of which is the rate-limiting step of the telomerase enzyme. Characterisation of the TERT gene promoter has identified several potential regulatory factors. For example, an estrogen response element was found and estrogen modulates telomerase ac-
tivity by increasing TERT transcription in estrogen-receptor-positive cells (Misiti et al., 2000). Progesterone, on the other hand, is implicated in the negative regulation of telomerase activity in the endometrium, and may arrest the cell cycle and depress TERT expression via the tumour suppressor CDKN1A (Wang et al., 2000). Androgens negatively regulate telomerase activity in normal prostate, while prostate cancer cells have high levels of telomerase activity even in the presence of androgens (Bouchal et al., 2002). Also paracrine and autocrine factors have essential roles in regulating telomerase expression. For example, transforming growth factor- (TGF-) has profound anti-proliferate effects in a variety of cells by up-regulation of CDKN1A and CDKN2B/INK4B, and down-regulation of cell cycle stimulating kinases. Moreover it inhibits telomerase activity by transcriptional repression of the TERT promoter (Yang et al., 2001). Epidermal growth factor (EGF) promotes cell proliferation in different cell types. There is also a rapid increase of hTERT production under EGF stimulation in cancer cells, and it is considered to have an activating effect on TERT transcription (Budiyanto et al., 2003). Insulin-like growth factors (IGFs) have complex effects on the regulation of telomerase activity. IGF-I is unable alone to up-regulate telomerase activity in non-neoplastic cells, while it stimulates TERT expression in prostate cancer cell lines independent of androgen response (Wetterau et al., 2003). IGF binding protein 2 (IGFBP2) stimulates telomerase activity in prostate cancer cells, although it has an inhibitory effect on normal prostate epithelial cells (Moore et al., 2003). Telomerase activity is thus regulated both by systemic and local signalling networks. Under normal conditions, these systems serve to limit telomerase activity to cells with longterm proliferation requirements. Telomere erosion and genomic instability in cancer
In tumour cells, a close link between telomeres and telomerase, on the one hand, and genomic instability, on the other hand, has been established. The most common type of instability so far described in solid tumours is characterised by chromosomal breakage-fusion-bridge (BFB) cycles (Fig. 1). In this process, ring-shaped or dicentric chromosomes fail to undergo normal chromatid separation at anaphase. Instead they form chromatin bridges that may subsequently break due to the tension forces of the mitotic spindle (McClintock, 1938; Gisselsson et al., 2000; Saunders et al., 2000). The broken ends in each daughter cell then fuse and form novel dicentric and ring chromosomes, which may enter another cycle of bridging, breakage and fusion. BFB cycles thus offer a mechanism for continuous rearrangement of chromosome structure, contributing to oncogene amplification and tumour suppressor gene deletion (Coquelle et al., 1997; Gisselsson et al., 2001). Anaphase bridging may also lead to loss of whole chromosomes, failure of cytokinesis, centrosomal amplification, and mitotic multipolarity (Gisselsson et al., 2005; Stewénius et al., 2005).
Cytogenet Genome Res 118:270–276 (2007)
271
Fig. 1. (a) Fluorescence in situ hybridisation detection of TTAGGG sequences (green) in a urothelial cancer metaphase cell with several signal-negative chromosome termini (arrows) implying loss or severe reduction (!500 bp) of telomeric repeats. (b) Loss of telomere capping function leads to fusion of chromosome arms through non-homologous end-joining; centromeres are indicated by white circles, intact T-loop telomeres by green circles, and disrupted telomeres by green lines. (c) Fusion of chromosome arms results in a functionally dicentric chromosome that may form a bridge at anaphase when the two centromeres in each chromatid are pulled in different directions by the mitotic machinery; centrosomes are indicated by yellow circles.
Studies of cultured human tumour cells have shown that the vast majority of tumours with BFB instability exhibit critical telomere shortening below the limits for T-loop formation. Fusion of destabilised, unprotected termini readily explains the formation of rings and dicentrics and the ensuing BFB events (McClintock, 1940). In fact, concurrent telomere shortening and anaphase bridging has been observed in the majority of investigated epithelial tumours, including breast (Meeker et al., 2004a), prostate (Meeker, 2006), colorectal (Rudolph et al., 2001; Stewénius et al., 2005), head and neck (Saunders et al., 2000; Gisselsson et al., 2002), pancreatic (Gisselsson et al., 2001), and ovarian cancers (Gisselsson et al., 2005). Many of these tumours show TP53 mutations or other types of cell cycle checkpoint disruptions conferring increased tolerance to DNA breakage (Artandi et al., 2000; Gisselsson et al., 2000). However, TP53 deficiency does not single-handedly bring about genomic instability (Bunz et al., 2002). Therefore, a combination of telomere deficiency and cell cycle checkpoint disruption may be a prerequisite for sustainable BFB instability, possibly because a functional checkpoint machinery would quickly eliminate cells undergoing anaphase bridging and chromosome breakage.
272
Cytogenet Genome Res 118:270–276 (2007)
(d) Anaphase bridges can be observed in vivo, here in a section from a high-grade urothelial cancer stained by haematoxylin and eosin. (e) Anaphase bridging may result in breakage of chromatin; here DNA double strand breaks are visualised by terminal transferase detection (green) in a remnant of an anaphase bridge protruding from an interphase nucleus (blue), where centromeric DNA is labelled by an alphasatellite probe (red). (f ) If the centromeres in the bridge completely detach from one of the anaphase poles, bridging may result in loss of chromosomes from one daughter cell; the other daughter cell could gain a dicentric chromosome, which may again participate in anaphase bridging.
Telomerase activation in cancer
It has been shown that deregulated TERT expression contributes to neoplastic transformation (Hahn et al., 1999) and it is generally believed that the main contribution of telomerase in carcinogenesis is to prevent telomere shortening and crisis during prolonged neoplastic growth. According to this model, telomerase counteracts the senescence, apoptosis, and/or genomic instability triggered by dysfunctional telomeres. Indeed, telomerase activity and/or TERT expression has been detected by the functional telomere repeat amplification protocol (TRAP) and/or reverse transcriptase PCR in 190% of malignant tumours (Kallakury et al., 1997; Yano et al., 2002; Sanz-Casla et al., 2005; Ernst et al., 2006). Among these cancers, chromosomal BFB instability triggered by telomere dysfunction is still a common phenomenon. The co-existence of genomic instability and telomerase expression is difficult to reconcile with the theory that the main function of telomerase in cancer is genomic stabilisation. Is it possible that telomere-dependent genomic instability is confined to a sub-population of tumour cells without telomerase activity? Immunohistochemical detection of hTERT has shown that its expression often exhibits extensive intercellular variability within tu-
Fig. 2. (a) Temporal development of the percentage of anaphase cells with bridges (AB; red circles) and telomerase expression (grey shading) in immortalised human ovarian surface epithelial cells during successive in vitro passages (p); values for normal fibroblasts (N) were used as the baseline for AB (Gisselsson et al., 2005). (b) The respective proportions of urothelial cancers showing anaphase bridges (red circles) and telomerase activity by TRAP assay (blue circles) at preinvasive (Ta–T1) and invasive (T2–T4) stages (De Kok et al., 2000;
Longchampt et al., 2003; Gisselsson et al., unpublished data). (c) Mean frequencies of anaphase bridges (AB; red circles) and mean telomerase activity (TA) by TRAP assay (blue circles) during colorectal carcinogenesis; N = normal epithelium, LGA = low grade dysplastic adenoma, HGA = high-grade dysplastic adenoma, CIS = carcinoma in situ; ICA = invasive carcinoma; MET = metastasis (Yoshida et al., 1999; Rudolph et al., 2001; Kanamaru et al., 2002).
mours (Yan et al., 2004). However, telomere shortening is typically a generalised phenomenon in epithelial tumour cells (Stewénius et al., 2005). Thus, topographical differences do not convincingly explain the telomere-telomerase discrepancy in cancer. It is important to note that some tumours lacking detectable telomerase activity maintain telomeres through a telomerase-independent route (Bryan et al., 1997), referred to as alternative lengthening of telomeres (ALT). The ALT mechanism is still not delineated in detail, but it appears to be common particularly in bone and soft tissue sarcomas which are typically characterised by very heterogeneous telomere lengths within individual tumours (Montgomery et al., 2004). Nevertheless, most common cancers show telomerase activity concurrent with some degree of telomere-dependent chromosomal instability. The hypothesis of telomerase being capable of stabilising telomeres in cancer cells should therefore be re-examined.
dysplasias) from the urinary bladder, oesophagus, large intestine, oral cavity, and uterine cervix by a recently developed technique for direct in situ telomere length assessment in formalin-fixed tissue (Meeker et al., 2004b). Furthermore, an in vitro model of cytogenetic evolution in ovarian carcinoma, using E6–E7-transfected ovarian surface epithelial cells, has shown that excessive telomere shortening and chromosomal instability typically occur prior to telomerase expression and immortalisation (Fig. 2a). In these cells, telomerase expression was associated with a decrease in genetic heterogeneity and an increased proliferation rate (Gisselsson et al., 2005). Telomerase expression further led to a reduction in the number of critically short telomeres. However, neither dysfunctional telomeres, nor anaphase bridging was completely eliminated by telomerase expression in this model. Rather, the genomic instability was maintained at a steady-state level allowing efficient cellular proliferation while retaining some potential for cytogenetic evolution. In cancer biopsies, an analogous situation has been observed. In urothelial cancer, anaphase bridging is present in the great majority of early-stage (Ta and T1) cancers. The proportion of these tumours showing telomerase activity by TRAP is considerably lower (Fig. 2b). In comparison, a majority of tumours at later (T2–T4) stages show both anaphase bridging and telomerase expression (De Kok et al., 2000; Longchampt et al., 2003). A similar scenario (Fig. 2c) is seen when comparing the mean frequencies of anaphase bridging to telomerase activity in colorectal neoplasms (Yoshida et al., 1999; Kanamaru et al., 2002). Here, the rate of anaphase bridging is elevated compared to normal tissue already in low-grade dysplastic polyps, is further elevated in high-grade pre-invasive lesions and invasive carcinomas, and finally decreases in carcinoma metastases (Rudolph et al., 2001). Telomerase activity, on the other hand continu-
Is telomerase activation too late?
In germ cells and stem cells, telomerase continuously counteracts telomere shortening during cellular proliferation. In tumour development, the timing of telomerase activation and telomere shortening may not be coordinated in a similar fashion. The risk for a genome to enter into BFB cycles is not as dependent on the average telomere length as on the length of the shortest telomere present. Elegant fluorescence in situ hybridisation (FISH) studies have shown recently that one dysfunctional telomere is sufficient to trigger a series of unbalanced translocations, leading to chromosomal rearrangements similar to those seen in tumour cells (Sabatier et al., 2005). Significant shortening of individual telomeric repeat sequences has been detected in almost 90% of cancer precursor lesions (low- and high-grade
Cytogenet Genome Res 118:270–276 (2007)
273
ously increases during neoplastic progression, reaching maximum values in metastatic tumours. Taken together these studies indicate that telomere shortening occurs early in tumorigenesis, and reaches high levels before telomerase activation through up-regulation of TERT reaches maximum levels. Once telomerase is active, telomere sequences are stabilised to some extent but some remain unstable. The molecular composition of the reactive chromosomal termini remaining after telomerase expression is up-regulated is not known. Combined telomere sequence detection and chromosome banding have demonstrated that chromosome arms may lack TTAGGG signals by FISH without being affected by any other rearrangement than a tendency of telomeric fusion with other chromosome arms (Gisselsson et al., 2002). One could hypothesise that the remaining TTAGGG repeats are too few for functional interaction with telomerase. Alternatively, the chromosome termini shown to be TTAGGG-negative by FISH could represent terminal deletions with a complete elimination of telomeric repeats. In fact, single-telomere length analysis has shown that additional, sporadic, and rare mutational events resulting in large-scale changes in telomere length are superimposed on end-replication losses in normal somatic cells (Britt-Compton et al., 2006) and male germ cells (Baird et al., 2006). Furthermore, it has been suggested that subtelomeric sequences exhibit endogenous plasticity, resulting in structural diversity of these regions in somatic cells due to mitotic rearrangements, and polymorphisms in the population due to germ line rearrangements (van Overveld et al., 2000). Whether similar large-scale telomeric or subtelomeric rearrangements contribute to telomere deletion in neoplastic cells has not been tested. Addition of telomeric TTAGGG sequences to terminal deletions is known to be very rare (Cristofari and Lingner, 2003). Extensive deletions of TTAGGG repeats during the telomerase-negative phase of tumour development could thus, at least in theory, explain the persistence of breakage-fusion-bridge events in telomerase-positive tumours. Evidently, further investigations of the molecular composition of chromosome termini in cancer would be of great interest. Is telomere shortening too fast or telomerase too slow?
Several proteins are attached to telomeric DNA. Some of these, such as XRCC5, MRE11, NBS and BLM, are also involved in DNA damage response pathways. Germ-line mutations in the genes encoding these proteins cause rare genetic syndromes characterised by genetic instability in somatic cells and cancer predisposition. In these diseases, disrupted protein functions cause a defective telomere protection, resulting in accelerated telomere shortening, lack of end-capping function, and/or end-to-end chromosome fusions. The presence of acquired somatic mutations in such genes regulating telomere length has been little investigated in tumour cells. Moreover, few studies have systematically addressed the function of the telosome protein complex in tumours. It can therefore not be excluded that telomeric
274
Cytogenet Genome Res 118:270–276 (2007)
protein deregulation has a role in cancer-associated telomeric instability. A new mouse model with critically short telomeres has been generated by over-expressing the TRF2 telomere-binding protein (Munoz et al., 2005). These mice show short telomeres in the presence of telomerase activity, leading to premature aging and increased cancer incidence. Furthermore, recent studies of liver neoplasms have suggested that up-regulation of the TRF1, TRF2, and TIN2 proteins might be involved in multi-step carcinogenesis by promoting telomere shortening (Oh et al., 2005). The expression of TRF1, TRF2, and TIN2 mRNAs, as well as the TRF1 protein was gradually increased through progression to hepatocellular carcinoma. In parallel, there was a gradual shortening of telomere length, with a significant telomere reduction in dysplastic nodules compared to adjacent foci of cirrhosis and chronic hepatitis. Abnormal telomere length regulation could thus result in telomere shortening at a rate which cannot be compensated for by telomerase. Alternatively, it can be envisioned that dysfunction of telomere-associated proteins leads directly to uncapping of telomeres, thus triggering chromosomal instability independently of telomere length. TRF2 inhibition has been shown to promote mutations that increase the ability of immortalised cells to grow in soft agar, although the capacity of tumour growth in nude mice was not similarly increased (Brunori et al., 2006). This transforming activity was associated with a burst of telomere instability, independent of telomere length reduction. It has also been shown that expression of inducible mutations in the Nijmegen breakage syndrome gene, NBS1, in telomerase-positive human tumour cells leads to an increased rate of telomere loss without any changes in average telomere length and it was suggested that this phenomenon could be due to failure of telomere capping (Bai and Murnane, 2003). Considering the efficacy by which disruption of the telomeric protein complex can cause genomic instability, the corresponding genes may be plausible targets for inactivating mutations in neoplastic cells (Karlseder et al., 2002). However, it is also possible that such inactivation would be unfavourable for tumour growth, if irreversible. While telomeric instability promotes carcinogenesis by triggering genomic diversity at early stages of neoplastic growth (Artandi and DePinho, 2000), subsequent stabilisation of the tumour genome could be crucial for the fixation of favourable, oncogenic mutations. Telomere shortening, followed by relative telomeric stabilisation by TERT up-regulation would thus provide an ideal mechanism for tumour progression. A direct and irreversible loss of telomere capping may not be equally efficient to promote neoplasia in human cells. Telomere-independent functions of telomerase
Considering that ALT may also provide relative stabilisation of telomeres in cancer cells, it can be questioned whether telomere maintenance is the sole mechanism by which telomerase promotes tumorigenesis. Recent studies have shown that telomerase may enhance tumour development
by regulating genes promoting cellular growth (Chung et al., 2005). Telomerase expression can regulate the transcription of basic fibroblast growth factor and the EGF receptor (Smith et al., 2003; Geserick et al., 2006). Telomerase expression may also promote carcinogenesis by diminishing the rate of cell death. Retroviral expression of TERT in cytokine-dependent human hematopoietic progenitor cells resulted in survival of these cells also at withdrawal of cytokine stimulation (Li et al., 2006). Furthermore, expression of TERT under the -myosin heavy chain promoter conferred protection from apoptosis to cardiac myocytes (Oh et al., 2001) and TERT expression in neuronal cells prevented apoptosis by suppressing cell death at a pre-mitochondrial
step (Lu et al., 2001). Taken together, these data suggest that the function of telomerase in carcinogenesis could be more diverse than previously believed. Regardless of its role for telomere protection in cancer, this strongly supports the notion that inhibition of telomerase is a feasible strategy for targeted tumour therapy. During recent years, several inhibitors of telomerase have been developed and have been shown to reduce the growth of tumour cells in vitro as well as in vivo (Seimiya et al., 2005; Hochreiter et al., 2006; Taetz et al., 2006; Zhang et al., 2006). The results of the first clinical trial with a telomerase inhibitor, initiated last year, are eagerly anticipated (Shay and Wright, 2006).
References Artandi SE, DePinho RA: A critical role for telomeres in suppressing and facilitating carcinogenesis. Curr Opin Genet Dev 10: 39–46 (2000). Artandi SE, Chang S, Lee SL, Alson S, Gottlieb GJ, et al: Telomere dysfunction promotes non-reciprocal translocations and epithelial cancers in mice. Nature 406: 641–645 (2000). Bai Y, Murnane JP: Telomere instability in a human tumor cell line expressing NBS1 with mutations at sites phosphorylated by ATM. Mol Cancer Res 1:1058–1069 (2003). Bailey SM, Murnane JP: Telomeres, chromosome instability and cancer. Nucleic Acids Res 34: 2408–2417 (2006). Baird DM, Britt-Compton B, Rowson J, Amso NN, Gregory L, Kipling D: Telomere instability in the male germline. Hum Mol Genet 15: 45–51 (2006). Blackburn EH: Switching and signaling at the telomere. Cell 106: 661–673 (2001). Blasco MA: Telomeres and human disease: ageing, cancer and beyond. Nat Rev Genet 6: 611–622 (2005). Bouchal J, Kolar Z, Mad’arova J, Hlobilkova A, von Angerer E: The effects of natural ligands of hormone receptors and their antagonists on telomerase activity in the androgen sensitive prostatic cancer cell line LNCaP. Biochem Pharmacol 63:1177–1181 (2002). Britt-Compton B, Rowson J, Locke M, Mackenzie I, Kipling D, Baird DM: Structural stability and chromosome-specific telomere length is governed by cis-acting determinants in humans. Hum Mol Genet 15:725–733 (2006). Brunori M, Mathieu N, Ricoul M, Bauwens S, Koering CE, et al: TRF2 inhibition promotes anchorage-independent growth of telomerasepositive human fibroblasts. Oncogene 25: 990–997 (2006). Bryan TM, Englezou A, Dalla-Pozza L, Dunham MA, Reddel RR: Evidence for an alternative mechanism for maintaining telomere length in human tumors and tumor-derived cell lines. Nat Med 3: 1271–1274 (1997). Budiyanto A, Bito T, Kunisada M, Ashida M, Ichihashi M, Ueda M: Inhibition of the epidermal growth factor receptor suppresses telomerase activity in HSC-1 human cutaneous squamous cell carcinoma cells. J Invest Dermatol 121: 1088–1094 (2003). Bunz F, Fauth C, Speicher MR, Dutriaux A, Sedivy JM, et al: Targeted inactivation of p53 in human cells does not result in aneuploidy. Cancer Res 62:1129–1133 (2002).
Cao Y, Li H, Deb S, Liu JP: TERT regulates cell survival independent of telomerase enzymatic activity. Oncogene 21: 3130–3138 (2002). Chung HK, Cheong C, Song J, Lee HW: Extratelomeric functions of telomerase. Curr Mol Med 5: 233–241 (2005). Coquelle A, Pipiras E, Toledo F, Buttin G, Debatisse M: Expression of fragile sites triggers intrachromosomal mammalian gene amplification and sets boundaries to early amplicons. Cell 89: 215–225 (1997). Cristofari G, Lingner J: Fingering the ends: how to make new telomeres. Cell 113: 552–554 (2003). De Kok JB, Schalken JA, Aalders TW, Ruers TJ, Willems HL, Swinkels DW: Quantitative measurement of telomerase reverse transcriptase (hTERT) mRNA in urothelial cell carcinomas. Int J Cancer 87: 217–220 (2000). de Lange T: Shelterin: the protein complex that shapes and safeguards human telomeres. Genes Dev 19:2100–2110 (2005). Ernst A, Okkels H, Grove A, Svarrer T, Knudsen A, et al: Telomerase activity determined by conventional telomeric repeat amplification protocol and reverse transcriptase-polymerase chain reaction assay in ovarian lesions: A comparison of assays. Scand J Clin Lab Invest 66: 287–298 (2006). Forsyth NR, Wright WE, Shay JW: Telomerase and differentiation in multicellular organisms: turn it off, turn it on, and turn it off again. Differentiation 69:188–197 (2002). Geserick C, Tejera A, Gonzalez-Suarez E, Klatt P, Blasco MA: Expression of mTert in primary murine cells links the growth-promoting effects of telomerase to transforming growth factor-beta signaling. Oncogene 25: 4310–4319 (2006). Gisselsson D, Pettersson L, Höglund M, Heidenblad M, Gorunova L, et al: Chromosomal breakagefusion-bridge events cause genetic intratumor heterogeneity. Proc Natl Acad Sci USA 97:5357– 5362 (2000). Gisselsson D, Jonson T, Petersén A, Strömbeck B, Dal Cin P, et al: Telomere dysfunction triggers extensive DNA fragmentation and evolution of complex chromosome abnormalities in human malignant tumors. Proc Natl Acad Sci USA 98: 12683–12688 (2001). Gisselsson D, Jonson T, Yu C, Martins C, Jin Y, et al: Centrosome abnormalities, multipolar mitoses, and chromosomal instability in head and neck tumours with dysfunctional telomeres. Br J Cancer 87: 202–207 (2002).
Gisselsson D, Lv M, Tsao SW, Man C, Jin C, et al: Telomere-mediated mitotic disturbances in immortalized ovarian epithelial cells reproduce chromosomal losses and breakpoints from ovarian carcinoma. Genes Chromosomes Cancer 42:22–33 (2005). Hahn WC, Counter CM, Lundberg AS, Beijersbergen RL, Brooks MW, Weinberg RA: Creation of human tumour cells with defined genetic elements. Nature 400:464–468 (1999). Hochreiter AE, Xiao H, Goldblatt EM, Gryaznov SM, Miller KD, et al: Telomerase template antagonist GRN163L disrupts telomere maintenance, tumor growth, and metastasis of breast cancer. Clin Cancer Res 12: 3184–3192 (2006). Kallakury BV, Brien TP, Lowry CV, Muraca PJ, Fisher HA, et al: Telomerase activity in human benign prostate tissue and prostatic adenocarcinomas. Diagn Mol Pathol 6: 192–198 (1997). Kanamaru T, Tanaka K, Kotani J, Ueno K, Yamamoto M, et al: Telomerase activity and hTERT mRNA in development and progression of adenoma to colorectal cancer. Int J Mol Med 10: 205–210 (2002). Karlseder J, Smogorzewska A, de Lange T: Senescence induced by altered telomere state, not telomere loss. Science 295: 2446–2449 (2002). Lehner R, Enomoto T, McGregor JA, Shroyer AL, Haugen BR, et al: Quantitative analysis of telomerase hTERT mRNA and telomerase activity in endometrioid adenocarcinoma and in normal endometrium. Gynecol Oncol 84: 120– 125 (2002). Li S, Ferguson MJ, Hawkins CJ, Smith C, Elwood NJ: Human telomerase reverse transcriptase protects hematopoietic progenitor TF-1 cells from death and quiescence induced by cytokine withdrawal. Leukemia 20: 1270–1278 (2006). Longchampt E, Lebret T, Molinie V, Bieche I, Botto H, Lidereau R: Detection of telomerase status by semiquantitative and in situ assays, and by real-time reverse transcription-polymerase chain reaction (telomerase reverse transcriptase) assay in bladder carcinomas. BJU Int 91: 567–572 (2003). Lu C, Fu W, Mattson MP: Telomerase protects developing neurons against DNA damage-induced cell death. Brain Res Dev Brain Res 131: 167–171 (2001). McClintock B: The production of homozygous deficient tissues with mutant characteristics by means of the aberrant behavior of ring-shaped chromosomes. Genetics 23: 215–376 (1938).
Cytogenet Genome Res 118:270–276 (2007)
275
McClintock B: The stability of broken ends of chromosomes in Zea mays. Genetics 26: 234–282 (1940). Meeker AK: Telomeres and telomerase in prostatic intraepithelial neoplasia and prostate cancer biology. Urol Oncol 24: 122–130 (2006). Meeker AK, Hicks JL, Gabrielson E, Strauss WM, De Marzo AM, Argani P: Telomere shortening occurs in subsets of normal breast epithelium as well as in situ and invasive carcinoma. Am J Pathol 164: 925–935 (2004a). Meeker AK, Hicks JL, Iacobuzio-Donahue CA, Montgomery EA, Westra WH, et al: Telomere length abnormalities occur early in the initiation of epithelial carcinogenesis. Clin Cancer Res 10:3317–3326 (2004b). Misiti S, Nanni S, Fontemaggi G, Cong YS, Wen J, et al: Induction of hTERT expression and telomerase activity by estrogens in human ovary epithelium cells. Mol Cell Biol 20: 3764–3771 (2000). Montgomery E, Argani P, Hicks JL, DeMarzo AM, Meeker AK: Telomere lengths of translocationassociated and nontranslocation-associated sarcomas differ dramatically. Am J Pathol 164: 1523–1529 (2004). Moore MG, Wetterau LA, Francis MJ, Peehl DM, Cohen P: Novel stimulatory role for insulinlike growth factor binding protein-2 in prostate cancer cells. Int J Cancer 105: 14–19 (2003). Munoz P, Blanco R, Flores JM, Blasco MA: XPF nuclease-dependent telomere loss and increased DNA damage in mice overexpressing TRF2 result in premature aging and cancer. Nat Genet 37:1063–1071 (2005). Oh BK, Kim YJ, Park C, Park YN: Up-regulation of telomere-binding proteins, TRF1, TRF2, and TIN2 is related to telomere shortening during human multistep hepatocarcinogenesis. Am J Pathol 166: 73–80 (2005).
276
Oh H, Taffet GE, Youker KA, Entman ML, Overbeek PA, et al: Telomerase reverse transcriptase promotes cardiac muscle cell proliferation, hypertrophy, and survival. Proc Natl Acad Sci USA 98: 10308–10313 (2001). Rudolph KL, Millard M, Bosenberg MW, DePinho RA: Telomere dysfunction and evolution of intestinal carcinoma in mice and humans. Nat Genet 28:155–159 (2001). Sabatier L, Ricoul M, Pottier G, Murnane JP: The loss of a single telomere can result in instability of multiple chromosomes in a human tumor cell line. Mol Cancer Res 3: 139–150 (2005). Sanz-Casla MT, Vidaurreta M, Sanchez-Rueda D, Maestro ML, Arroyo M, Cerdan FJ: Telomerase activity as a prognostic factor in colorectal cancer. Onkologie 28: 553–557 (2005). Saunders WS, Shuster M, Huang X, Gharaibeh B, Enyenihi AH, et al: Chromosomal instability and cytoskeletal defects in oral cancer cells. Proc Natl Acad Sci USA 97: 303–308 (2000). Seimiya H, Muramatsu Y, Ohishi T, Tsuruo T: Tankyrase 1 as a target for telomere-directed molecular cancer therapeutics. Cancer Cell 7: 25– 37 (2005). Shay JW, Wright WE: Telomerase therapeutics for cancer: challenges and new directions. Nat Rev Drug Discov 5:577–584 (2006). Shin JS, Hong A, Solomon MJ, Lee CS: The role of telomeres and telomerase in the pathology of human cancer and aging. Pathology 38: 103– 113 (2006). Smith LL, Coller HA, Roberts JM: Telomerase modulates expression of growth-controlling genes and enhances cell proliferation. Nat Cell Biol 5: 474–479 (2003). Stewénius Y, Gorunova L, Jonson T, Larsson N, Höglund M, et al: Structural and numerical chromosome changes in colon cancer develop through telomere-mediated anaphase bridges, not through mitotic multipolarity. Proc Natl Acad Sci USA 102:5541–5546 (2005). Taetz S, Baldes C, Murdter TE, Kleideiter E, Piotrowska K, et al: Biopharmaceutical characterization of the telomerase inhibitor BRACO19. Pharm Res 23:1031–1037 (2006).
Cytogenet Genome Res 118:270–276 (2007)
van Overveld PG, Lemmers RJ, Deidda G, Sandkuijl L, Padberg GW, et al: Interchromosomal repeat array interactions between chromosomes 4 and 10: a model for subtelomeric plasticity. Hum Mol Genet 9:2879–2884 (2000). Verdun RE, Crabbe L, Haggblom C, Karlseder J: Functional human telomeres are recognized as DNA damage in G2 of the cell cycle. Mol Cell 20:551–561 (2005). Wang Z, Kyo S, Takakura M, Tanaka M, Yatabe N, et al: Progesterone regulates human telomerase reverse transcriptase gene expression via activation of mitogen-activated protein kinase signaling pathway. Cancer Res 60: 5376–5381 (2000). Wetterau LA, Francis MJ, Ma L, Cohen P: Insulinlike growth factor I stimulates telomerase activity in prostate cancer cells. J Clin Endocrinol Metab 88:3354–3359 (2003). Yan P, Benhattar J, Seelentag W, Stehle JC, Bosman FT: Immunohistochemical localization of hTERT protein in human tissues. Histochem Cell Biol 121:391–397 (2004). Yang H, Kyo S, Takatura M, Sun L: Autocrine transforming growth factor beta suppresses telomerase activity and transcription of human telomerase reverse transcriptase in human cancer cells. Cell Growth Differ 12:119–127 (2001). Yano Y, Yoshida K, Osaki A, Toge T, Tahara H, et al: Expression and distribution of human telomerase catalytic component, hTERT, in human breast tissues. Anticancer Res 22: 4101–4107 (2002). Yoshida R, Kiyozuka Y, Ichiyoshi H, Senzaki H, Takada H, et al: Change in telomerase activity during human colorectal carcinogenesis. Anticancer Res 19: 2167–2172 (1999). Zhang L, Tamura K, Shin-ya K, Takahashi H: The telomerase inhibitor telomestatin induces telomere shortening and cell death in Arabidopsis. Biochim Biophys Acta 1763: 39–44 (2006).
Table 1. Examples of morphological findings, and clinical and therapeutic features associated with cytogenetic and molecular genetic changes in various cancersa Abnormality/Gene(s)
Cancer type
Subtype, clinical and therapeutic features
del(1p) and/or del(19q)/? t(2;5)(p23;q35)/ ALK;NPM1 t(12;13)(q35;q14)/ PAX;FOXO1A t(2p12)/ IGK@ amp(2p)/ MYCN t(3;8)(p21;q12)/ ? inv(3)(q21q26)/ ROPN1;EVI1 t(4;11)(q21;q23) /MLL T2(AF-4);MLL(ALL-1)
Glial tumors NHL Alveolar rhabdomyosarcoma NHL Neuroblastoma Salivary gland AML ALL
–5/? del(5q)/?
AML MDS
t(5;14)(q31;q32)/ IL3;IGH@ t(6;9)(p23;q34)/ DEK;CAN t(6;14)(q21;q24)/ ? t(7q34) and t(7p15)/ TRB@(TCRB); TRG@(TCRG) t(7;10)(q35;q21)/ ? –7/ ?
ALL AML Ovary NHL Thyroid AML
–5, del(5q), –7 and/or +8/?
AML
amp(7p)/ EGFR t(8;14)(q24;q32)/ MYC;IGH@ t(8;16)(p11;p13)/ MYST3(MOZ); CREBBP(CBP) t(8;21)(q22;q22)/ CBFA2T1(ETO); RUNX2(CBFA1), RUNX1(AML1) t(9;11)(p21;q23)/ MLL T3(AF9); MLL(ALL-1) t(9;14)(p13;q32)/ ?;IGH@ t(9;16)(q22;p13)/ ? t(9;22)(q34;q11)/ ABL1;BCR
Astrocytotic tumor ALL-L3, NHL AML M5b AML M2
PCV-sensitive Anaplastic large T-cell Small round cell tumor B-cell Poor prognosis Pleomorphic adenoma Frequent presence of thrombocytosis, Resistant to chemotherapy Leukemic cell bears myeloid and B lymphatic features; High leukocyte count; Occurs mainly in infants, age below 24 months; Poor prognosis Resistant to chemotherapy 5q– syndrome: elderly female, macrocytic anemia, normal thrombocyte count, hypolobulated megakaryocytes Reactive eosinophilia/ basophilia Frequent presence of basophilia Adenocarcinoma T-cell Papillary carcinoma Repeated hemorrhages. Fever, infections. High white cell count. Poor prognosis Frequent in patients with chemotherapy-related secondary leukemia or occupational mutagen exposure. Poor prognosis Poor prognosis Mostly Burkitt’s lymphoma Erythrophagocytosis Leukemic tumors more common than typically
dim(10,14)/ ? t(11;14)(q13;q32)/ CCND1(BCL1,PRAD1);IGH@ t(11;18)(q21;q21)/ BIRC3;MALT1 t(11;22)(q24;q12)/ FL1;EWSR1 del/t(11q23)/ MLL(ALL-1)
Mesothelioma NHL NHL Ewing’s sarcoma AML
del(11q22)/ ? amp(12q)/ numerous del (13q14)/ ? t(14q11)/ TRD@(TRCD) t(14q32)/ IGH@ t(14;18)(q32;q23)/ BCL2;IGH@ t(15;17)(q22;q11)/ NPM;RARA inv(16p)(p13q22)/ MYH11;CBFB2
CLL Liposarcoma CLL NHL NHL NHL AML M3 AML M4EO
amp(17q12)/ ERBB2 amp(17q21)/ ? (del(17p13)/ TP53) t(17;22)(q22;q13)/ COL1A1;PDGFB t(22q11)/ IGL@ t(X;18)(p11;q11)/ SYT1;SSX1,SSX2 KIT-mutation
Breast carcinoma Neuroblastoma CLL Dermatofibrosarcoma NHL Synovial sarcoma GIST
AML M5 NHL Skin CML, ALL
a Detailed listing in Mitelman Database of Chromosome Aberrations in Cancer (2005) (http://cgap.nci.nih.gov/Chromosomes/Mitelman). Online access to detailed data of recurrent chromosomal aberrations in various malignancies at http://www.infobiogen.fr/services/chromcancer/Anomalies/ and http://infobiogen.fr/services/chromcancer/. ALL = acute lymphoblastic leukaemia; AML = acute myeloid leukaemia; amp = amplification; ATRA = all-trans retinoic acid; CLL = chronic lymphocytic leukaemia; CML = chronic myeloid leukaemia; DIC = disseminated intravascular coagulation; del = deletion (chromosomal material missing); dim = dimin-
278
Cytogenet Genome Res 118:277–283 (2007)
Small lymphocytic Basal cell carcinoma Leukemic cells bears myeloid and B lymphaticfeatures in ALL. Imatinib-sensitive Mesothelioma vs. adenocarcinoma of the lung Mantle cell: poor prognosis MALT: resistant to eradication treatment of Helicobacter pylori Differential diagnosis in small round cell tumors Frequent in patients who have been exposed to epipodophyllotoxins Poor prognosis Liposarcoma versus lipoma Good prognosis T-cell B-cell Mostly follicular DIC; ATRA-sensitive. Good prognosis Eosinophilia: metastizing to central nervous system more commonly than typically Herceptin-sensitive Poor prognosis Poor prognosis Dignosis B-cell Diagnosis Imanitib-sensitive
ished (DNA copy number loss); inv = inversion (chromosomal); EO = eosinophilia; GIST = gastrointestinal stromal tumor; L = morphologic subtype of ALL; M = morphologic subtype of AML; MALT = mucosa associated lymphatic tumor; MDS = myelodysplastic syndrome; t = translocation (chromosomal); (–) = monosomy (whole chromosome missing); PCV = procarbazine, cyclophosphamide, vincristine. Reproduced from Annals of Medicine 36:162–171 (2004) by permission of Francis and Taylor Journals.
has been shown to reside in the FRA7G fragile site (Miller et al., 2005). The FHIT gene located in FRA3B has been shown to have deletions in a variety of cancers, e.g., gastric and lung cancers (Ohta et al., 1996). FRA16D has been shown to be involved in multiple myeloma translocations (Krummel et al., 2000). In all, a DNA sequence is not evenly durable but specific genomic regions are more prone to break than others. Recently, Myllykangas et al. (2006) studied the resemblance of the amplification patterns of different tumor types to each other in a bibliomics survey. The comparative genomic hybridization (CGH) results of 838 published studies were included in the study. The results of the amplification-based clustering of different tumors showed that cancers with similar origins or behavior tended to cluster together, and the majority of the amplified regions were found to correlate with genomic fragile sites (Myllykangas et al., 2006). Numerous agents are known to cause DNA doublestranded breaks and induce genetic aberrations. In vitro experiments have shown that radiation, cigarette smoking, aphidicolin, ethanol, caffeine and FUdR induce DNA breaks and facilitate the formation of chromosome aberrations (Kuwano et al., 1987; Rao et al., 1988; Ban et al., 1995). Viruses are capable of increasing the carcinogenic effect of clastogenic chemicals (reviewed by Haverkos, 2004). Previous studies have shown that specific chromosomal changes co-occur with malignant transformation of cells and are caused by viruses or ethylnitrosourea in vitro (Nystrom et al., 1985). For example, Simian virus 40 (SV40) has been implicated as a cofactor in asbestos inflicted malignant mesothelioma, but its role in the pathogenesis of mesothelioma is still unclear (Ascoli et al., 2001). The association between human papilloma virus (HPV) infection and risk for developing cervical cancer has been established by the International Agency for Research on Cancer (IARC, 1995). The virus infection alone is not sufficient for cancer induction (reviewed by zur Hausen, 2000). Smoking is the most significant environmental risk factor (Kjellberg et al., 2000). The susceptibility for cancer and its progression can be considered to be a sum of different factors, induced by both extraneous and intrinsic factors, which so far have not been fully elucidated, as reviewed by Wogan et al. (2004). Analysis of genomic DNA and gene expression in cancer diagnostics and research
DNA copy number changes in the tumor cell genome (gains, losses, high-level amplifications and deletions) can be assessed using CGH (Kallioniemi et al., 1992). In conventional CGH, DNAs extracted from tumor and normal tissue (reference) are labeled using specific fluorescent dyes (Kallioniemi et al., 1992). Labeled DNAs (probes) are hybridized on a glass slide containing spread metaphase chromosomes (targets). Copy number changes in the tumor sample appear as an increase or decrease in the fluorescence intensity in comparison to normal tissue (Forozan et al., 1997). How-
ever, as the resolution of conventional chromosomal CGH is restricted to chromosome bands (10–20 Mb), the method fails to reveal gene-specific amplifications and deletions. Array CGH (aCGH) has been developed to overcome this constraint. The aCGH method is similar to conventional CGH with the exception that hybridization is carried out using DNA sequence targets instead of metaphase chromosomes. CGH microarrays consist of BAC-, cDNA- or oligonucleotide targets spotted on solid platforms (Pinkel et al., 1998; Pollack et al., 1999; Brennan et al., 2004). Figure 1 illustrates the CGH methods. Fresh, fresh-frozen or paraffinembedded and fixed tumor material can be used for CGH analysis (Zielenska et al., 2004). Figure 2 shows a genomewide DNA copy number profiling result obtained using cDNA-based aCGH in osteosarcoma (Atiye et al., 2005). Alterations in gene expression can be assessed using array-based technologies (Schena et al., 1996). Methodogically, gene expression microarrays are similar to aCGH but use mRNA instead of DNA (Fig. 1). Tumor mRNA (sample) and a corresponding normal tissue reference are labeled using different fluorescent dyes (probe) and hybridized on a glass slide with spotted cDNA- or oligonucleotide targets (Schena et al., 1996). The fluorescence intensity ratios of sample and reference represent differences in gene expression. Using microarrays, genes have been shown to be differentially expressed in tumor-specific clusters. Gene expression patterns can distinguish specific cancers from normal tissue and other tumors, which suggest that gene expression profiling might have important diagnostic usage. For instance, the expression patterns observed in breast cancer may be used for predicting the outcome of the disease (van’t Veer et al., 1996; Sorlie et al., 2001). In toxicology research expression microarrays have been utilized for determination of gene expression effects of hazardous substances (reviewed by Pennie et al., 2004) and in mutagenicity testing (Bartosiewicz et al., 2001). Helicobacter pylori and asbestos as carcinogens leading to gastric cancer and mesothelioma
Helicobacter pylori infection and asbestos exposure are known to predispose to gastric cancer and mesothelioma, respectively (Graham, 2000; Robinson et al., 2005). Apart from mesothelioma, asbestos exposure is also known to be involved in lung cancer development (Selikoff et al., 1968; Robinson et al., 2005) and together with smoking, to cause an additive effect in asbestos-related lung cancer (Selikoff et al., 1968; Vainio and Boffetta, 1994). Gastric and mesothelial carcinogenesis share many features: These diseases are caused by exposure to a chronic inflammation inducing agent (H. pylori and asbestos), the onset takes place long after exposure, the tumors are heterogeneous, and show multiple chromosomal aberrations (Wu et al., 2001; Krissmann et al., 2002). H. pylori infection and asbestos exposure cause similar cellular responses; activation of nuclear factor kB and mitogen activated protein kinase pathways, and up-regulation of pro-inflammatory cytokines (such as JUN and
Cytogenet Genome Res 118:277–283 (2007)
279
Fig. 1. Comparative genomic hybridization (CGH) and gene expression analysis. DNA or RNA is extracted from tumor tissue and labeled with fluorescent dye. Prior to labeling RNA is converted to cDNA. A reference DNA or RNA with normal DNA copy number or expression level is labeled with another dye. The samples are combined and hybridized on a metaphase spread (chromosomal CGH) or arrayed cDNA clones (array CGH, expression array) and the resulting images
are analyzed using compatible software. Over- or under-representation of tumor DNA or RNA as compared to the control sample are determined and the results are interpreted as gains or losses of genomic regions (DNA analysis) or over- or underexpression of particular genes (expression analysis). Reproduced from Annals of Medicine 36: 162– 171 (2004) by permission of Francis and Taylor Journals.
FOS) as well as various stress response genes (Zanella et al., 1996; Manning et al., 2002; Myllykangas et al., 2004). An association with viruses (SV40 in mesothelioma and EBV in gastric cancer) has been observed in both cancer types. Even though associations between asbestos and mesothelioma as well as H. pylori and gastric cancer are undisputable, the primary events in gastric and mesothelial carcinogenesis still remain largely unknown. Using cell lines and in vitro exposure experiments, carcinogenic mechanisms of asbestos have been shown to instigate chromosomal aberrations (Sincock et al., 1975), disrupt mitosis (Ault et al., 1995) and block chromosome segregation (Jensen et al., 1996). Specific asbestos-related regions of DNA copy number alteration have recently been published by Nymark et al. (2006), and the regions have been shown to be associated with genomic fragile sites. DNA damage induced by H. pylori is suggested to result from the action of reactive oxygen species (Obst et al., 2000). An object of interest is to resolve whether the primary gene expression changes induced by H. pylori infection in the stomach and asbestos burden in the lung are associated with the chromosomal changes found in later stages of the diseases.
The open chromatin hypothesis – linking expression and copy number alterations
280
Cytogenet Genome Res 118:277–283 (2007)
Unraveling of the human genome sequence and subsequent transcriptome analyses have revealed that highly expressed genes are clustered in distinct chromosomal domains (Caron et al., 2001). According to various studies, the up-regulated genes in cancers cluster to distinct chromosomal regions, which show preferential amplification during tumorigenesis. The clustering has been shown in breast cancer (Monni et al., 2001; Hyman et al., 2002), pancreatic cancer (Heidenblad et al., 2005), lung cancer (Tonon et al., 2005) and ovarian cancer (Israeli et al., 2005). In pancreatic carcinoma the proportion of overexpressed genes within the amplified region has been reported to be as high as 60% (Heidenblad et al., 2005). It is not known whether the increased expression is a consequence of the amplification of the corresponding genes or whether the amplifications tend to occur in transcriptionally active regions. Myllykangas et al. (2004) showed that in H. pylori-exposed cells the chromosomal locations of the up-regulated genes corresponded to the regions that are frequently amplified in gastric cancer. The results are shown in Fig. 3. It
Fig. 2. cDNA array CGH result in osteosarcoma (Atiye et al., 2004) The cumulative base pair locations of the genome (1pter–Xqter) are indicated on the X-axis and the relative fluorescence ratios (tumor vs. normal tissue) are indicated on the Y-axis. In addition to a well-known amplification at 17p13]p11 (arrow), a novel region of amplification was found to occur at 17q25 (arrow).
Fig. 3. H. pylori-regulated genes in AGS cell line are frequently co-localized with amplified chromosome loci in gastric adenocarcinoma. (A) Number of genes differentially expressed in H. pylori-infected cell line by genome location (Myllykangas et al., 2004). (B) Frequently amplified chromosome bands in gastric adenocarcinoma (data obtained from http://www.helsinki.fi/cmg/cgh_data.html).
Cytogenet Genome Res 118:277–283 (2007)
281
was suggested that prolonged activation of H. pylori target genes reduces physical protection of DNA by untying the histone coiling during transcription. Such DNA structure changes might facilitate the formation of DNA doublestranded breaks and alterations in chromosomal integrity. Transcription activation induced amplification mechanisms are discussed in a review by Myllykangas and Knuutila (2006). Previous studies show that H. pylori-related and non-related gastric cancers do not differ from each other in respect to chromosomal aberrations (van Grieken et al., 2000). This might be expected because H. pylori-induced gastric adenocarcinomas are not supposed to be any different from gastric cancers, which develop without H. pylori involvement. Still, H. pylori infection might promote the formation of gastric cancer-specific DNA alterations.
Conclusions
The application of microarray methods and systems biology approach gives us for the first time a possibility to design a model of the carcinogenetic progression of tumors associated with exposure to extraneous agents (Myllykangas and Knuutila, 2006). The described cytogenetic and molecular genetic methods may also be used in investigation of mutagenic processes. Acknowledgements The authors thank Taylor and Francis Journals for permission to republish Table 1 and Fig. 1.
References Ascoli V, Aalto Y, Carnovale-Scalzo C, Nardi F, Falzetti D, et al: DNA copy number changes in familial malignant mesothelioma. Cancer Genet Cytogenet 127:80–82 (2001). Atiye J, Wolf M, Kaur S, Monni O, Böhling T, et al: Gene amplifications in osteosarcoma-CGH microarray analysis. Genes Chromosomes Cancer 42:158–163 (2005). Ault JG, Cole RW, Jensen CG, Jensen LC, Bachert LA, et al: Behavior of crocidolite asbestos during mitosis in living vertebrate lung epithelial cells. Cancer Res 55:792–798 (1995). Ban S, Cologne JB, Neriishi K: Effect of radiation and cigarette smoking on expression of FUdRinducible common fragile sites in human peripheral lymphocytes. Mut Res 334: 197–203 (1995). Bartosiewicz M, Penn S, Buckpitt A: Applications of gene arrays in environmental toxicology: Fingerprints of gene regulation associated with cadmium chloride, benzo(a)pyrene, and trichloroethylene. Envir Health Perspect 109: 71– 74 (2001). Brennan C, Zhang Y, Leo C, Feng B, Cauwels C, et al: High-resolution global profiling of genomic alterations with long oligonucleotide microarray. Cancer Res 64: 4744–4748 (2004). Caron H, van Schaik B, van der Mee M, Baas F, Riggins G, et al: The human transcriptome map: Clustering of highly expressed genes in chromosomal domains. Science 291: 1289–1292 (2001). Delattre O, Zucman J, Plougastel B, Desmaze C, Melot T, et al: Gene fusion with an ETS DNAbinding domain caused by chromosome translocation in human tumors. Nature 359:162–165 (1992). Forozan F, Karhu R, Kononen J, Kallioniemi A, Kallioniemi OP: Genome screening by comparative genomic hybridization. Trends Genet 13: 405–409 (1997). Glover TW: Common fragile sites. Cancer Lett 232: 4–12 (2006). Graham DY: Helicobacter pylori infection is the primary cause of gastric cancer. J Gastroenterol 35 S12:90–97 (2000). Hanahan D, Weinberg RA: The hallmarks of cancer. Cell 100:57–70 (2000). Haverkos HW: Viruses, chemicals and co-carcinogenesis. Oncogene 23: 6492–6499 (2004).
282
Heidenblad M, Lindgren D, Veltman JA, Jonson T, Mahlamaki EH, et al: Microarray analyses reveal strong influence of DNA copy number alterations on the transcriptional patterns in pancreatic cancer: Implications for the interpretation of genomic amplifications. Oncogene 24:1794–1801 (2005). Hyman E, Kauraniemi P, Hautaniemi S, Wolf M, Mousses S, et al: Impact of DNA amplification on gene expression patterns in breast cancer. Cancer Res 62: 6240–6245 (2002). International Agency for Research on Cancer: IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Volume 64 (WHO, Lyon 1995). Israeli O, Goldring-Aviram A, Rienstein S, Ben-Baruch G, Korach J, et al: In silico chromosomal clustering of genes displaying altered expression patterns in ovarian cancer. Cancer Genet Cytogenet 160:35–42 (2005). Jensen CG, Jensen LC, Rieder CL, Cole RW, Ault JG: Long crocidolite asbestos fibers cause polyploidy by sterically blocking cytokinesis. Carcinogenesis 17:2013–2021 (1996). Kallioniemi A, Kallioniemi OP, Sudar D, Rutovitz D, Gray JW, et al: Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science 258: 818–821 (1992). Kjellberg L, Hallmans G, Ahren AM, Johansson R, Bergman F, et al: Smoking, diet, pregnancy and oral contraceptive use as risk factors for cervical intra-epithelial neoplasia in relation to human papillomavirus infection. Br J Cancer 82: 1332–1338 (2000). Knuutila S: Cytogenetics and molecular pathology in cancer diagnostics. Ann Med 36: 162–171 (2004). Knuutila S, Autio K, Aalto Y: Online access to CGH data of DNA sequence copy number changes. Am J Pathol 157: 689 (2000). Krismann M, Muller KM, Jaworska M, Johnen G: Molecular cytogenetic differences between histological subtypes of malignant mesotheliomas: DNA cytometry and comparative genomic hybridization of 90 cases. J Pathol 197: 363–371 (2002). Krummel KA, Roberts LR, Kawakami M, Glover TW, Smith DI: The characterization of the common fragile site FRA16D and its involvement in multiple myeloma translocations. Genomics 69: 37–46 (2000).
Cytogenet Genome Res 118:277–283 (2007)
Kuwano A, Kajii T: Synergistic effect of aphidicolin and ethanol on the induction of common fragile sites. Hum Genet 75:75–78 (1987). Lengauer C, Kinzler KW, Vogelstein B: Genetic instabilities in human cancers. Nature 396: 643– 649 (1998). Manning CB, Cummins AB, Jung MW, Berlanger I, Timblin CR, et al: A mutant epidermal growth factor receptor targeted to lung epithelium inhibits asbestos-induced proliferation and proto-oncogene expression. Cancer Res 62: 4169– 4175 (2002). Miller CT, Lin L, Casper AM, Lim J, Thomas DG, et al: Genomic amplification of MET with boundaries within fragile site FRA7G and upregulation of MET pathways in esophageal adenocarcinoma. Oncogene 25:409–418 (2005). Mitelman F, Johansson B, Mertens F (eds): Mitelman Database of Chromosome Aberrations in Cancer, http://cgap.nci.nih.gov/Chromosomes/Mitelman (2005). Monni O, Barlund M, Mousses S, Kononen J, Sauter G, et al: Comprehensive copy number and gene expression profiling of the 17q23 amplicon in human breast cancer. Proc Natl Acad Sci USA 98:5711–5716 (2001). Myllykangas S, Knuutila S: Manifestation, mechanisms and mysteries of gene amplifications. Cancer Lett 232:79–89 (2006). Myllykangas S, Monni O, Nagy B, Rautelin H, Knuutila S: Helicobacter pylori infection activates FOS and stress-response genes and alters expression of genes in gastric cancer-specific loci. Genes Chromosomes Cancer 40: 334–341 (2004). Myllykangas S, Himberg J, Bohling T, Nagy B, Hollmen J, Knuutila S: DNA copy number amplification profiling of human neoplasms. Oncogene 25:7324–7332 (2006). Nymark P, Wikman H, Ruosaari S, Hollmen J, Vanhala E, et al: Identification of specific gene copy number changes in asbestos-related lung cancer. Cancer Res 66: 5737–5743 (2006). Nystrom M, Knuutila S, Rapp UR, Keski-Oja J: MMC-E cells – origin and changes in karyotype accompanying malignant transformation. Cancer Genet Cytogenet 15: 243–251 (1985). Obst B, Wagner S, Sewing KF, Beil W: Helicobacter pylori causes DNA damage in gastric epithelial cells. Carcinogenesis 21:1111–1115 (2000).
Ohta M, Inoue H, Cotticelli MG, Kastury K, Baffa R, et al: The FHIT gene, spanning the chromosome 3p14.2 fragile site and renal carcinomaassociated t(3;8) breakpoint, is abnormal in digestive tract cancers. Cell 84: 587–597 (1996). Pennie W, Pettit SD, Lord PG: Toxicogenomics in risk assessment: An overview of an HESI collaborative research program. Envir Health Perspect 112:417–419 (2004). Pinkel D, Sewgraves R, Sudar D, Clark S, Poole I, et al: High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays. Nat Genet 20: 207– 211 (1998). Pollack JR, Perou CM, Alizadeh AA, Eisen MB, Pergemenhikov A, et al: Genome-wide analysis of DNA copy-number changes using cDNA microarrays. Nat Genet 23: 41–46 (1999). Rao PN, Heerema NA, Palmer CG: Fragile sites induced by FUdR, caffeine, and aphidicolin. Their frequency, distribution, and analysis. Hum Genet 78:21–26 (1988). Robinson BW, Musk AW, Lake RA: Malignant mesothelioma. Lancet 366: 397–408 (2005). Schena M, Shalon D, Heller R, Chai A, Brown PO, Davis RW: Parallel human genome analysis: Microarray-based expression monitoring of 1000 genes. Proc Natl Acad Sci USA 93: 10614– 10619 (1996).
Selikoff I, Hammond E, Churg J: Asbestos exposure, smoking, and neoplasia. J Am Med Assoc 204:106–112 (1968). Sincock A, Seabright M: Induction of chromosome changes in Chinese hamster cells by exposure to asbestos fibres. Nature 257: 56–58 (1975). Sorlie T, Perou CM, Tibshirani R, Aas T, Geisler S, et al: Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci USA 98: 10869–10874 (2001). Tonon G, Wong KK, Maulik G, Brennan C, Feng B, et al: High-resolution genomic profiles of human lung cancer. Proc Natl Acad Sci USA 102: 9625–9630 (2005). Vainio H, Boffetta P: Mechanisms of the combined effect of asbestos and smoking in the etiology of lung cancer. Scand J Work Environ Health 20: 235–242 (1994). van Grieken NC, Weiss MM, Meijer GA, Hermsen MA, Scholte GH, et al: Helicobacter pylori-related and -non-related gastric cancers do not differ with respect to chromosomal aberrations. J Pathology 192: 301–306 (2000).
van ‘t Veer LJ, Dai H, van de Vijver MJ, He YD, Hart AA, et al: Gene expression profiling predicts clinical outcome of breast cancer. Nature 415: 530–536 (2002). Wogan GN, Hecht SS, Felton JS, Conney AH, Loeb LA: Environmental and chemical carcinogenesis. Semin Cancer Biol 14: 473–486 (2004). Wu MS, Chang MC, Huang SP, Tseng CC, Sheu JC, et al: Correlation of histologic subtypes and replication error phenotype with comparative genomic hybridization in gastric cancer. Genes Chromosomes Cancer 30: 80–86 (2001). Zanella CL, Posada J, Tritton TR, Mossman BT: Asbestos causes stimulation of the extracellular signal-regulated kinase 1 mitogen-activated protein kinase cascade after phosphorylation of the epidermal growth factor receptor. Cancer Res 56:5334–5338 (1996). Zielenska M, Marrano P, Thorner P, Pei J, Beheshti B, et al: High-resolution cDNA microarray CGH mapping of genomic imbalances in osteosarcoma using formalin-fixed paraffin-embedded tissue. Cytogenet Genome Res 107:77–82 (2004). zur Hausen H: Papillomaviruses causing cancer: Evasion from host-cell control in early events in carcinogenesis. J Natl Cancer Instit 92:690–698 (2000).
Cytogenet Genome Res 118:277–283 (2007)
283
sequence probes, fusion probes or break-apart probes. The probes are hybridized either singly or more commonly in multi-color combinations allowing simultaneous detection of multiple chromosomes and/or chromosome regions. The number of interphase nuclei scored range from 100 to 500 per probe in different laboratories and, when the percentage of cells with an abnormality exceeds a cut-off value established for each probe following internal validation, is reported as a clonal abnormality. The incidence of relevant cytogenetic abnormalities using specifically designed probe sets in chronic lymphocytic leukemia (CLL), multiple myeloma (MM), myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML), and their clinical relevance including prognostic implications is reviewed. Chronic lymphocytic leukemia (CLL)
Chronic lymphocytic leukemia (CLL) is the most common adult leukemia in the United States and Europe accounting for approximately 30% of all leukemia (Ries et al., 2006). The majority of patients are over 50 years old with a median age of 65 and an excess of males. The disease is characterized by the clonal proliferation and accumulation of neoplastic B lymphocytes in the blood, bone marrow, lymph nodes and spleen (Jaffe et al., 2001). The immunophenotype of this monoclonal population includes the pathognomic coexpression of CD5 and CD19, as well as positivity for CD20, CD21, CD23, and CD24. CLL may also present as a lymph node-based disease, called small lymphocytic lymphoma (SLL). The clinical course is heterogeneous with survival times ranging from within a few months of diagnosis, despite aggressive therapy, to many years without therapy. Clinical staging systems devised to classify patients based on the extent of the disease at initial diagnosis and provide the best predictors of survival (Rai et al., 1975; Binet et al., 1981), predicted survival duration for patients with the most advanced stage as 1–2 years, and a median survival time of over 10 years for patients with the lowest stage of disease. The exception to this system includes patients with early stage CLL. More recently, immunophenotype and genetic changes have also been shown to contribute prognostic information. CD38 positivity is associated with significantly shorter overall survival and progression-free survival times (Jelinek et al., 2001; Ghia et al., 2003). Absence of hypermutation of the immunoglobulin heavy chain variable region (IGHV) genes present in about 50% of CLL cases is also associated with a high risk for early progression (Hamblin et al., 1999; Krober et al., 2002). Expression of the intracellular signaling molecule ZAP70 has been identified in a majority of CLL cases without IGHV mutations, and is highly predictive of patients with more rapid disease progression and death (Crespo et al., 2003; Wiestner et al., 2003). Chromosomal abnormalities provide independent prognostic information in newly diagnosed CLL patients. The presence of clonal aberrations, specific chromosomal abnormalities, the percentage of abnormal cells and the complexity of the
aberrations are all indicators of disease progression and survival, and have resulted in the delineation of prognostic subgroups in CLL with implications for the design of riskadapted therapeutic strategies (Juliusson and Gahrton, 1990; Juliusson et al., 1990; Dohner et al., 2000; Thornton et al., 2004). Conventional cytogenetic analysis of CLL has been hampered by low spontaneous mitotic activity and the lack of growth of leukemic cells in culture, despite appropriate stimulation with polyclonal B-cell mitogens. The incidence of clonal cytogenetic abnormalities in reported studies ranges from 28.5% (Reddy, 2006) to over 80% (Dicker et al., 2006; Schoch et al., 2006) with an average of 40–50% of cases (Juliusson and Gahrton, 1990; Escudier et al., 1993; Hernandez et al., 1995; Finn et al., 1996; Bigoni et al., 1997; Athanasiadou et al., 2006). About 10% of these have complex karyotypic changes. The most common cytogenetic abnormalities account for more than 60% of all abnormal cases and include trisomy of chromosome 12, translocation or deletion of 13q, rearrangements involving 14q32, and deletions of 6q, 11q and 17p. Less frequent are trisomies of chromosomes 3 and 18. Because of its greater sensitivity, FISH analysis was often used initially as an adjunct to cytogenetic analysis in the detection of specific abnormalities in cases with a normal karyotype or those with insufficient or no metaphases. These analyses showed not only an increase in the detection rate of abnormalities to about 80% (Dohner et al., 1999, 2000; Dewald et al., 2003), but also a difference in the spectrum of chromosomal abnormalities detected. Abnormalities that were cytogenetically rare were observed in greater frequency by FISH, such as deletion of 13q, 11q and 17p, and those that were common by routine chromosome analysis were observed to be not as frequent, e.g. trisomy 12. FISH panels were therefore designed to reliably detect genetic abnormalities of proven clinical significance. The DNA probe set generally includes probes for chromosome 12 centromere for detection of trisomy 12, ATM (11q22.3), D13S25 (13q14.3), D13S319 (13q14.3), RB1 (13q14), LAMP1 (13q34) and TP53 (17p13.1) for deletions of 11q, 13q and 17p, respectively, and IGH (14q32.3) for rearrangements at that locus. MYB probe to determine deletion of 6q has also been included by some laboratories. Deletion 13q Deletion of the long arm of chromosome 13 was reported in 10–20% of cases by cytogenetics (Juliusson et al., 1990; Escudier et al., 1993; Athanasiadou et al., 2006). By FISH analysis it is the most common abnormality and is present in about 31–55% of cases (Dohner et al., 2000; Dewald et al., 2003; Aoun et al., 2004; Goorha et al., 2004; Gozzetti et al., 2004; Glassman and Hayes, 2005; Sindelarova et al., 2005; Reddy, 2006). The deletion is seen as the sole abnormality or in association with other aberrations, and in either the hemizygous state (60–70%), or the homozygous and mosaic states in equal proportions (10–20% each) (Dohner et al., 2000; Dewald et al., 2003; Aoun et al., 2004; Fink et al., 2004; Reddy, 2006). Prognostically, patients with deletion 13q
Cytogenet Genome Res 118:284–296 (2007)
285
have the overall best disease free survival times, akin to those with normal karyotypes (Juliusson and Gahrton, 1990; Dohner et al., 2000). However, even within this subgroup of CLL the clinical course of the disease may be heterogeneous. Patients with a homozygous deletion fare worse than the hemizygous cases. Also, patients with deletion 13q and other abnormalities have an unfavorable prognosis (Dewald et al., 2003; Fink et al., 2004). Trisomy 12 Trisomy 12 was the most common abnormality observed cytogenetically with an incidence of 8–20% of all cases (Juliusson et al., 1990; Escudier et al., 1993; Que et al., 1993; Athanasiadou et al., 2006), but by FISH it is the second most common abnormality and has been observed in 7–35% of the cases (Anastasi et al., 1992; Cuneo et al., 1992; Escudier et al., 1993; Que et al., 1993; Matutes et al., 1996; Aoun et al., 2004; Goorha et al., 2004; Gozzetti et al., 2004; Glassman and Hayes, 2005; Sindelarova et al., 2005; Reddy, 2006). Trisomy 12 is strongly associated with atypical lymphocyte morphology, atypical marker expression, advanced stage of the disease and resistance to chemotherapy (Cuneo et al., 1992; Coignet et al., 1993; Escudier et al., 1993; Que et al., 1993). Expression of FMC7, CD38, CD20 and surface immunoglobulin light chain was significantly higher (Finn et al., 1996; Matutes et al., 1996; Ghia et al., 2003; Athanasiadou et al., 2006; Reddy, 2006). One study found that patients with trisomy 12 tended to present with more splenomegaly than those without and this was consistent with the more advanced stage seen in trisomy 12 patients (Escudier et al., 1993). They were also more likely to be previously treated. Prognostically, trisomy 12 patients had significantly shorter median survival times compared to those with deletion of 13q as the sole aberration (Dohner et al., 2000) and those with normal karyotypes in some studies (Juliusson et al., 1990; Escudier et al., 1993; Criel et al., 1997), while other analyses found no significant difference in survival between patients with trisomy 12 and normal karyotypes (Athanasiadou et al., 2006). Based on studies combining FISH and immunophenotyping and the observation of trisomy 12 in low numbers of trisomic cells in patients with other abnormalities, it has been suggested to be a secondary event in leukemogenesis in some patients (Escudier et al., 1993).
This was independent of the stage of disease and may therefore represent an early marker for aggressive disease. Deletion 17p Deletions of 17p have been observed in 7–15% of patients by FISH and correlate with advanced disease, resistance to treatment and poor survival (el Rouby et al., 1993; Dohner et al., 1995, 2000; Geisler et al., 1997). Also, the development of 17p deletion is associated with mutations or deletions of the tumor suppressor gene TP53 and is more likely to occur as a result of clonal evolution after initial diagnosis. Deletion 6q Deletions of 6q are commonly observed nonrandom chromosomal aberrations in lymphoid malignancies. In CLL, they have been reported cytogenetically in about 4– 21% of cases, with an average of 7% by FISH analysis (Amiel et al., 1999; Stilgenbauer et al., 1999; Dohner et al., 2000; Cuneo et al., 2004; Reddy, 2006). Two regions of minimal deletion have been established and include 6q21–q23 and 6q25–q27. Patients with deletion 6q had higher white blood cell counts, frequent splenomegaly, more extensive lymphadenopathy, atypical morphology, CD38 positivity and short to intermediate survival (Bigoni et al., 1997; Stilgenbauer et al., 1999; Cuneo et al., 2004; Reddy, 2006). This deletion is not routinely screened for, however, because of its association with poor prognosis it might be a good candidate for inclusion in CLL FISH panels. Rearrangements of 14q32 Earlier studies of chromosomal abnormalities in CLL found 14q32 rearrangements in a high percentage of cases and an association with poorer survival compared to abnormalities of 13q (Juliusson et al., 1990). However, with the current reclassification of t(11; 14) with CCND1/IGH rearrangement as mantle cell lymphoma as well as other lymphoma associated translocations, fewer cases of CLL, about 4–5%, have abnormalities of 14q32 (Dohner et al., 2000; Aoun et al., 2004; Mayr et al., 2006). Patients with a rearrangement involving 14q32 have an unfavorable prognosis. Multiple myeloma (MM)
Deletion 11q Deletion of 11q was not identified as a frequent aberration by routine chromosome analysis, but by FISH analysis it was determined to be the third most frequent abnormality, observed in up to 23% of the cases (Dohner et al., 1997, 2000; Neilson et al., 1997; Dewald et al., 2003; Aoun et al., 2004; Goorha et al., 2004; Gozzetti et al., 2004; Glassman and Hayes, 2005; Sindelarova et al., 2005; Reddy, 2006). Clinically, patients with deletion 11q were younger, had more advanced clinical stages and were characterized by extensive peripheral, abdominal and mediastinal lymphadenopathy. The younger patients (age group less than 55 years) showed a more rapid disease progression with shorter survival times (Dohner et al., 1997; Neilson et al., 1997).
286
Cytogenet Genome Res 118:284–296 (2007)
Multiple myeloma (MM) is a heterogeneous mature Bcell lymphoid disorder characterized by a serum monoclonal protein and skeletal destruction with osteolytic lesions, pathological fractures, bone pain, hypercalcemia and anemia (Jaffe et al., 2001). It is the most common lymphoid malignancy in Blacks and the second most common in Whites, representing 15% of all hematological malignancies. The median age of diagnosis is 68 years in males and 70 years in females. The male to female ratio is approximately 1:1 (Jaffe et al., 2001). The clinical course of MM patients is highly variable and survival times range from a few months to several years (Kyle et al., 2003). Prognostic factors include clinical staging (Durie and Salmon, 1975), plasma-
cell proliferation (Greipp et al., 1993), plasma-cell morphology (Fritz et al., 1984), serum levels of beta-2 microglobulin, C-reactive protein (CRP) and lactate dehydrogenase (LDH) (Dimopoulos et al., 1991; Bataille et al., 1992). Advanced stages of the disease, plasmablastic morphology, 2 microglobulin, and high serum CRP and LDH levels are associated with a poor prognosis (Rajkumar and Greipp, 1999). More recently, cytogenetics has emerged as another important independent prognostic indicator helping to define a high-risk population that would benefit from intensive therapy. Abnormal karyotypes were consistently associated with a rapidly fatal outcome, and fewer than 10% of patients with these abnormalities survived longer than five years. Konigsberg et al. (2000) found that in the presence of adverse cytogenetic aberrations other prognostic factors of clinical importance in univariate analysis no longer have independent significance. In particular, deletions of 13q and 17p, and 11q rearrangements, as well as cases with a t(11; 14), were determined to be unfavorable cytogenetic abnormalities (Drach et al., 1998; Fonseca et al., 1999a; Konigsberg et al., 2000; Zojer et al., 2000; Facon et al., 2001; Chang et al., 2005). Extensive cytogenetic analyses have been carried out on MM patients, but the generally low mitotic activity resulted in an underestimation of the true incidence of abnormalities. In most published series, cases with abnormal karyotypes ranged from approximately 30 to 50% (Dewald et al., 1985; Weh et al., 1993; Lai et al., 1995a; Sawyer et al., 1995; Debes-Marun et al., 2003; Pantou et al., 2005). Complex karyotypes with multiple chromosomal abnormalities are the rule. Most MM cases are aneuploid and recurrent numerical aberrations include trisomies of chromosomes 3, 5, 7, 9, 11, 15, 19 and 21, and losses of the sex chromosomes, 8, 13, 14, 16 and 22. Of the structural abnormalities, the most consistent ones include translocations involving 14q32 with various chromosome partners, most commonly 4, 11, 16 and 6. Rearrangements of chromosome 1 are also frequent; however, the aberrations are diverse involving both the short and long arms, and their variable nature suggests they are secondary events. Deletion of the long arm of chromosome 13 is also consistently noted and together with monosomy of chromosome 13 suggests that total or partial loss of chromosome 13 is a recurrent event. Other recurrent abnormalities include deletions of 11q, 17p13 and 19q13. With the use of FISH, identification of specific deletions and rearrangements, not possible by conventional cytogenetics, raised the percentage of cases with abnormalities to up to 90% (Drach et al., 1995; Zandecki et al., 1996; SchmidtWolf et al., 2006; Avet-Loiseau et al., 2007), particularly when analysis was performed on purified bone marrow plasma cells. It also became clear that the delineation of specific cytogenetic abnormalities had clinical and biological implications and MM patients could be divided into subgroups based on their karyotypic profile (Tricot et al., 1995, 1997; Fonseca et al., 1999b, 2002a, b; Konigsberg et al., 2000; Zojer et al., 2000; Facon et al., 2001; Avet-Loiseau et al., 2002; Moreau et al., 2002; Dewald et al., 2005). The most frequent abnormalities detected by FISH analyses were de-
letion of the RB1 gene on 13q14, translocations involving the heavy chain gene on 14q32, most commonly t(11;14), t(4;14) and t(14;16), gain of 11q and deletion of 17p. Because of the increased detection rate of genetic abnormalities by FISH and its immense prognostic value, the current practice in cytogenetic laboratories is to test for genetic abnormalities in MM and its precursor condition, monoclonal gammopathy of undetermined significance (MGUS) by FISH. The FISH panel for MM in most laboratories has been designed to include probes for the detection of deletions of 11q (MLL at 11q23), 13q (RB1 at 13q14 and LAMP1 at 13q34) and 17p (TP53), and rearrangement of the IGH gene at 14q32 particularly with CCND1 at 11q13. Some laboratories also include probes for 4p16.3 (FGFR3) and 16q23 (MAF) for detection of translocations with the IGH gene. In addition to this, studies have recommended inclusion of probes to determine gains of regions 1q, 9q and 11q (Liebisch et al., 2003). The most frequent cytogenetic abnormalities and their clinical attributes are discussed in detail. Aneuploidy Aneuploidy is a common characteristic of MM (Dewald et al., 1985; Drach et al., 1995; Lai et al., 1995a; Sawyer et al., 1995; Zandecki et al., 1996; Debes-Marun et al., 2003) and is independent of clinical stage. Karyotypes may be hypodiploid, hyperdiploid, or pseudodiploid with balanced structural rearrangements. Ploidy category has been shown to have a significant effect on prognosis. Patients with a normal karyotype have a better prognosis than those with chromosomal abnormalities. Hypodiploidy has been associated with a poor prognosis (Smadja et al., 2001; Fassas et al., 2002; Debes-Marun et al., 2003). This, however, may not be independent of other adverse cytogenetic abnormalities such as 13q deletions and 14q translocations that are predominant in hypodiploid MMs (Fonseca et al., 2003). Hyperdiploidy was seen in 39% of patients and was observed to have marginal prognostic impact (Avet-Loiseau et al., 2007). Deletion 13q Deletion of 13q or monosomy of chromosome 13 was reported in 10–20% of MM cases by routine cytogenetic analysis (Weh et al.,1993; Lai et al., 1995a; Sawyer et al., 1995; Debes-Marun et al., 2003). By FISH analysis partial deletion of 13q or loss of chromosome 13 is one of the most common abnormalities in MM and has been reported in 30–50% of patients (Zojer et al., 2000; Facon et al., 2001; Fonseca et al., 2002b, 2003; Chang et al., 2004a; Avet-Loiseau et al., 2007). About 85% of deletion 13q cases detected by interphase FISH analysis represent monosomy 13 and only 15% are interstitial deletions (Avet-Loiseau et al., 2000; Fonseca et al., 2001). The minimal area of deletion, although not precisely mapped, appears to predominantly involve the RB1 gene and the D13S319 locus at 13q14.3 (Shaughnessy et al., 2000; Zojer et al., 2000). The deletions are mostly monoallelic. Deletion of 13q/monosomy 13 is an independent prognostic variable on multivariate analysis (Zojer et al., 2000; Facon et al., 2001; Fonseca et al., 2002b). It is associated with short-
Cytogenet Genome Res 118:284–296 (2007)
287
er survival, a lower likelihood of response to therapy and advanced stages of the disease (Desikan et al., 2000; Zojer et al., 2000; Facon et al., 2001; Fonseca et al., 2002b; Shaughnessy et al., 2003). A recent study, however, determined that the prognostic power of deletion 13q was related to t(4;14) and deletion of 17p which are often associated with deletion of 13q (Avet-Loiseau et al., 2007). Clinically, 13q loss is also associated with lambda-light chain type, low serum monoclonal concentration of ^1 q/dl and higher PC labeling index (Facon et al., 2001; Fonseca et al., 2002b). 14q32 translocations Cytogenetic analysis revealed that 14q32 was one of the most frequent chromosomal breakpoints. Rearrangements occurred at this band in about 10–60% of patients (Lai et al., 1995a; Sawyer et al., 1995; Pantou et al., 2005) with various chromosomal partners, most commonly chromosomes 11 and 8. FISH analysis revealed that the translocations involve the immunoglobulin heavy chain (IGH) locus at 14q32, are mostly cryptic and present in 40–75% of patients. The most common partners are 11, 4, 16 and 6 (Bergsagel et al., 1996; Nishida et al., 1997; Fonseca et al., 1999a; Avet-Loiseau et al., 1999, 2002; Chang et al., 2004a; Schmidt-Wolf et al., 2006), although other chromosomal regions may be involved. All the translocations lead to upregulation of the partner gene. The translocations represent a significant prognostic factor in MM with the rate of 14q32 rearrangements increasing with disease progression and reaching 90% in advanced cancers. These translocations are also present in MGUS and are therefore considered to be early pathogenetic events (Fonseca et al., 2002c). In MGUS, however, the translocations have no effect on prognosis. t(11;14)(q13;q32) This translocation has been identified in 15–20% of patients with MM (Avet-Loiseau et al., 2002, 2007; Fonseca et al., 2002a, 2003; Gertz et al., 2005) and 15–30% of patients with MGUS (Avet-Loiseau et al., 2002; Fonseca et al., 2002a). The t(11;14) correlates with a lymphoplasmacytic, mature morphology of plasma cells, CD20 expression and the oligo-/asecretory MM subtype (Hoyer et al., 2000; Fonseca et al., 2002a; Moreau et al., 2002; Avet-Loiseau et al., 2003; Robillard et al., 2003). The translocation results in overexpression of cyclin D1 akin to the situation in mantle cell lymphoma. The t(11; 14) has been associated with good prognosis in MM patients receiving high-dose chemotherapy and stem cell transplant (Moreau et al., 2002; Soverini et al., 2003). Avet-Loiseau et al. (2007), however, reported that it does not influence prognosis. t(4;14)(p16.3;q32) This cryptic translocation present in about 13–20% of cases and detectable by FISH, leads to the upregulation of the oncogenes Multiple Myeloma SET (MMSET or WHSC1) on 4p and FGFR3 on 14q, respectively (Chesi et al., 1998; Fonseca et al., 2001, 2003; Avet-Loiseau et al., 2007). Patients with the translocation commonly have an IgA isotype, aggressive clinical features and are associated with poor prog-
288
Cytogenet Genome Res 118:284–296 (2007)
nosis after high dose therapy (Moreau et al., 2002; Fonseca et al., 2003; Keats et al., 2003; Chang et al., 2004b; Gertz et al., 2005; Avet-Loiseau et al., 2007). When evaluated together with 2 microglobulin levels, however, patients with t(4;14) and a low 2 microglobulin had longer survival, close to that of patients without the translocation but with a high 2 microglobulin level (Avet-Loiseau et al., 2007). A strong correlation between t(4;14) and deletion 13q was noted and appeared to influence the prognostic impact of deletion 13q conferring an especially aggressive phenotype (Fonseca et al., 2001, 2003; Moreau et al., 2002). t(14;16)(q32;q23) This translocation is also more reliably detected by FISH and is present in about 2–10% of cases (Chesi et al., 1998; Avet-Loiseau et al., 2002; Fonseca et al., 2003). It results in the upregulation of the transcription factor c-MAF and cyclin D2. The prognostic correlation is not clear; however, Fonseca et al. (2003) noted a shorter survival and features of aggressiveness in patients with the translocation. A significant positive correlation appears to exist between t(14;16) and deletion of 17p13.1 (Fonseca et al., 2003). t(6;14) This translocation is relatively rare and has been observed in 3–4% of MM patients (Sawyer et al., 2001). It is associated with the upregulation of cyclin D3 (Shaughnessy et al., 2001). There is no known clinical or prognostic information for this translocation yet. Deletion 17p Deletions of the TP53 tumor suppressor gene at 17p13.1, as detected by FISH analysis, are predominantly monoallelic and range from about 5 to 33% in newly diagnosed MM patients (Drach et al., 1998; Fonseca et al., 2003; Chang et al., 2005; Avet-Loiseau et al., 2007). In patients with relapsed MM the incidence was about 55% (Drach et al., 1998). Loss of 17p occurs either due to an unbalanced rearrangement, an interstitial deletion or a translocation involving 17p, or monosomy of chromosome 17. Deletion of TP53 appears to be a marker of disease progression, functional loss of the gene by deletion or mutation was present in about 40% of patients with advanced MM (Neri et al., 1993). Deletion of 17p is associated with significantly worse prognosis (Drach et al., 1998; Fonseca et al., 2003; Chang et al., 2005; Gertz et al., 2005) and the patients also have other features of aggressiveness such as plasmacytoma and hypercalcemia (Fonseca et al., 2003). Other abnormalities Chromosome 1 rearrangements. Structural abnormalities of chromosome 1 involve both the short arm and the long arm and have been observed cytogenetically in 15–50% of patients (Lai et al., 1995a; Sawyer et al., 1995; Segeren et al., 2003). A strong association between chromosome 1 abnormalities and those of chromosome 13 was also noted (Segeren et al., 2003). Chromosome 1 abnormalities appeared to define a high-risk population with advanced dis-
ease, shorter event-free survival, time to progression and overall survival (Sawyer et al., 1995; Segeren et al., 2003; Bang et al., 2006). Amplification and overexpression of the cell cycle regulator gene CKS1B at 1q21 was shown to predict an aggressive clinical course (Shaughnessy, 2005) and this is consistent with the adverse prognostic features observed by Bang et al. (2006) in their cases with trisomy 1q. 9q rearrangements. Gain of 9q is also associated with very poor prognosis, particularly in patients with deletion of 13q also (Liebisch et al., 2005). Trisomy 11. Trisomy of chromosome 11 has been reported in 25–66% of cases (Harrison et al., 2003; Cremer et al., 2005; Bang et al., 2006; Guglielmelli et al., 2007). The prognostic impact of this gain is disputed; while Konigsberg et al. (2000) determined that it had no prognostic value, Gutierrez et al. (2004) reported shorter survival in patients with gains of 11q. Deletion of 6q27. Deletion of the long arm of chromosome 6 is associated with various lymphoproliferative disorders. In MM loss of 6q, particularly band 6q21, was one of the most frequent losses and it was detected in 10–28% of patients (Cigudosa et al., 1998; Amiel et al., 1999; Gutierrez et al., 2004). Coexistence of abnormalities. Patients with more than one abnormality often have a worse prognosis. Fonseca et al. (2003) observed a single chromosome abnormality in 40%, two abnormalities in 22% and three abnormalities in 3% of MM patients. Certain abnormalities are found in association with many other abnormalities, for instance deletion of 13q has been reported in 36% of hyperdiploid patients, in 39% with t(11; 14), 85% with t(4; 14), 92% with t(14; 16) and 78% with deletion 17p. Loss of 17p occurred with t(14;16) in 33% of patients, with t(4;14) in 20% and a low incidence was seen with t(11;14). Few patients were seen where t(4;14) and t(11;14) occurred together with hyperdiploidy. No association of t(4;14) and t(11;14) has been reported (Avet-Loiseau et al., 2002, 2007; Dewald, 2005). Other genetic mutations have been identified in MM and involve RAS, MYC, TP53, PTEN and various components of the RB pathway (Fonseca et al., 2004). All of these factors also have an impact on survival (Avet-Loiseau et al., 2002, 2007; Fonseca et al., 2003; Dewald, 2005). Acute myeloid leukemia
Acute myeloid leukemia (AML) is a heterogeneous clonal disease of hematologic progenitor cells. Its salient feature is the excessive accumulation of myeloid blasts in bone marrow, peripheral blood and other tissues. It is the most common malignant myeloid disease in adults with a median age at presentation of 70 years and a male to female ratio of 3:2. Etiologic factors include viruses and exposure to ionizing radiation, benzene and cytotoxic chemotherapy (Jaffe et al., 2001; Estey and Dohner, 2006). Up to 10–15% of patients with AML are therapy-related and develop the disorder after treatment with cytotoxic agents, some 5–10 years after exposure to alkylating agents and others 1–5 years after
treatment with topoisomerase II inhibitor drugs such as doxorubicin and etoposide (Estey and Dohner, 2006). Each of these types is associated with specific chromosomal aberrations. Prognostic factors in AML include age, immunophenotype, leukocyte count, previous history of MDS, and levels of LDH, serum albumin, bilirubin and creatinine (Chang et al., 2004c; Gupta et al., 2005); however, cytogenetic and molecular genetic findings at diagnosis constitute one of the most important prognostic determinants (Grimwade et al., 1998; Mrozek et al., 2000; Slovak et al., 2000; Schoch et al., 2001; Byrd et al., 2002). Cytogenetic abnormalities are present in an average of 55% (range 50–80%) of adults at diagnosis (Stasi et al., 1993; Heim and Mitelman, 1995; Mrozek et al., 2001; Estey and Dohner, 2006). A great majority of them are recurrent and in addition to being a critical component of the armamentarium of the diagnostic workup of AML, cytogenetic characterization is a basis for formulating therapeutic strategies (Tallman et al., 1997; Bloomfield et al., 1998). It is also a critical independent predictive factor of prognosis, response to chemotherapy, risk of relapse and outcome (Grimwade et al., 1998). Some of the abnormalities are associated with remarkable specificity with morphological and cytochemical hematologic subtypes as defined by the FAB classification (Bennett et al., 1976; Heim and Mitelman, 1995) and currently the WHO classification incorporates cytogenetics and molecular genetics in an attempt to define entities that are biologically homogeneous and prognostically relevant (Vardiman et al., 2002). Four abnormalities, t(15;17), t(8;21), inv(16) and t(16;16), are seen in nearly 30% of patients with AML (Grimwade et al., 1998; Mrozek et al., 2000; Slovak et al., 2000) and have a strong correlation with morphology, as well as distinctive clinical findings and favorable response to therapy. These are included in the subgroup AML with recurrent genetic abnormalities, and emerging new information may result in the expansion of this subgroup in the future. The other major WHO categories include AML with multilineage dysplasia, therapy-related AML and MDS, and AML not otherwise categorized. The therapy-related AMLs and MDSs are also characterized by specific chromosomal abnormalities. Alkylating agent/radiation-related t-AML and t-MDS show a higher incidence of abnormalities of chromosomes 5 and 7 and have a worse clinical outcome. Topoisomerase II inhibitor-related AML does not have a preceding MDS phase and is associated with balanced translocations involving 11q23 or 21q22, as well as inv(16) and t(15;17) (Vardiman et al., 2002). Numerous other recurrent structural and numerical abnormalities are present in AML (Heim and Mitelman, 1995; Mrozek et al., 2001) and some of the more common ones include monosomy or deletion of 5/5q, monosomy or deletion of 7/7q, translocations involving 11q23 with multiple chromosome partners, t(6;9), deletions of 11q and 20q, rearrangements of 3q36, and trisomies of chromosomes 8, 9, 11, 13, 19, 21 and 22. Secondary aberrations often accompany the primary change and can cause substantial variability in a patient’s outcome. Molecular characterization of the specific rearrangements has implicated genes important in cel-
Cytogenet Genome Res 118:284–296 (2007)
289
lular proliferation and differentiation (Cline, 1994) and demonstrated the creation of fusion genes that encode chimeric proteins responsible for leukemogenesis. The AML FISH panel is extensive and is used to detect known losses and gains, or specific translocations and rearrangements of chromosomes. The probes commonly used are EGR1 at 5q31 and D5S23, D5S721 at 5p15.2 for detection of monosomy/deletion of 5/5q, CEP7 and D7S486 for loss/ deletion of 7/7q, CEP8 for gain of chromosome 8, ETO/ AML1 (RUNX1T1/RUN X1, 8q22/21q22) for the t(8;21) rearrangement, RARA at 17q21 for the t(15; 17) (PML/RARA) rearrangement, CBFB at 16q22 for the inv(16) rearrangement, and MLL for rearrangements of 11q. Prognostically, based on cytogenetic and molecular genetic findings at diagnosis, AML patients can be broadly divided into those with favorable, intermediate or adverse outcomes (Grimwade et al., 1998, 2001; Slovak et al., 2000; Byrd et al., 2002). Within each group, particularly the favorable and intermediate groups, there is variability in outcome depending on secondary chromosomal abnormalities, gene mutations and deregulated gene expression (Estey and Dohner, 2006). Prognostically favorable group This group is characterized by low rates of primary drug resistance and superior overall survival associated with a reduced recurrence risk. Additional cytogenetic aberrations in this group did not in general have a deleterious effect on outcome (Grimwade et al., 1998). t(8; 21). This translocation is present in approximately 30% of AML with maturation. The AML1 (RUNX1) gene on 21q22 is juxtaposed with the ETO (RUNX1T1) gene on 8q22 resulting in a novel chimeric gene AML1/ETO (RUNX1/ RUNX1T1). Loss of the Y chromosome, when present in association with t(8;21), is associated with shorter survival in men (Schlenk et al., 2004). Also, deletion 9q as an additional change has been reported as a poor risk indicator requiring more aggressive treatment (Schoch et al., 1996). t(15;17). This translocation is characteristic of acute promyelocytic leukemia (APL) in which abnormal promyelocytes predominate. The retinoic acid receptor alpha (RARA) gene on 17q12 fuses with the PML gene on 15q22 to produce a PML/RARA gene fusion product. The translocation is detectable by FISH with the PML-RARA probe set. A RARA break-apart probe can also be used to detect t(15;17), as well as variant translocations associated with APL involving the RARA gene with genes other than PML. Patients with t(15; 17) are sensitive to treatment with all trans retinoic acid with significant improvement in outcome (Fenaux et al., 1997; Tallman et al., 1997). inv(16), t(16; 16). These rearrangements are seen in 10– 12% of AML. Both rearrangements result in the fusion of the CBFB gene at 16q22 with the MYH11 gene at 16p13 giving rise to a chimeric protein. The abnormality is detectable by FISH using a CBFB break-apart probe. Trisomy of chromosome 22, when present, improves relapse-free survival (Schlenk et al., 2004; Marcucci et al., 2005).
290
Cytogenet Genome Res 118:284–296 (2007)
Intermediate prognosis group This subgroup is very heterogeneous and patients with a normal karyotype constitute the largest proportion of patients (⬃88%). The remainder includes gains of chromosomes 6, 8, 11, 13, 21 and 22, loss of the Y chromosome, deletions of 7q, 9q, 12p and 20q, and abnormalities other than those noted in the other two categories (Grimwade et al., 1998; Slovak et al., 2000; Estey and Dohner, 2006). Their complete remission rate, risk of relapse and outcome are worse than those of treated patients in the prognostically favorable group described above but better than that of patients belonging to the adverse group with unfavorable cytogenetic findings. Additional chromosome abnormalities in this group, however, had a deleterious effect on outcome. Patients with a normal karyotype represent about 45% of patients with AML. At the molecular level, however, these patients are very heterogeneous and contain several prognostically significant gene mutations and changes in gene expression (Estey and Dohner, 2006; Mrozek et al., 2007). Group with a poor prognosis Patients in this group had a significantly poorer outcome than patients with a normal karyotype. They had resistant disease and were less likely to achieve complete remission and had a poorer overall survival reflecting increased risk of death on induction and/or relapse. Additional chromosomal abnormalities in this group did not affect the outcome. This group included patients with a complex karyotype, i.e. three or more abnormalities, monosomies of chromosomes 5 and 7, deletion of 5q, abnormalities of 3q (inv(3), t(3;3)), 11q (t(9;11) and t(11;19)), 20q, 21q, deletion of 9q, t(9; 9) and t(9; 22) (Grimwade et al., 1998, 2001; Slovak et al., 2000; Schoch et al., 2001; Byrd et al., 2002; Estey and Dohner, 2006; Chen et al., 2007). A few rearrangements, t(6;9), deletion 7q and 11q23 rearrangements, were classified as unfavorable by some groups (Buchner et al., 1999; Slovak et al., 2000; Byrd et al., 2002) and intermediate by others (Grimwade et al., 1998) reflecting in part the heterogeneity of the 11q23 abnormalities. Myelodysplastic syndrome (MDS)
The myelodysplastic syndromes (MDS) are a clinically heterogeneous group of hematologic diseases occurring predominantly in older adults (median age 76 years) with a significantly higher incidence in men (Ma et al., 2007). They are characterized by bone marrow dysplasia in one or more myeloid cell lineages, peripheral blood cytopenia, most commonly anemia and less frequently neutropenia and/or thrombocytopenia, and frequent progression to acute myeloid leukemia (AML). The number of myeloblasts in the blood or bone marrow is less than the 20% requisite for a diagnosis of AML (Vardiman et al., 2002). MDS may occur de novo or as a result of therapy with either alkylating agents or radiotherapy. They have an overall short survival, death being generally due to the consequences of cytopenia or progression to AML (Jaffe et al., 2001). Survival has been
correlated with age at diagnosis, gender, clinical subtype, bone marrow morphology and the percentage of bone marrow blast cells. Cytogenetic analysis is an accepted independent predictor of clinical outcome and overall survival, and of the likelihood of progression to AML. Several studies have demonstrated its value in determining prognosis and as a basis for targeted therapies (Aul et al., 1992; Morel et al., 1993; Toyama et al., 1993; Greenberg et al., 1997; Nevill et al., 1998; Pfeilstocker et al., 1999; Sole et al., 2000; Malcovati et al., 2005). An International MDS Risk Analysis Workshop classified untreated primary MDS into cytogenetic subgroups with good, intermediate and poor prognoses and these were incorporated into an International Prognostic Scoring System (IPSS) (Greenberg et al., 1997). Routine cytogenetic analysis has demonstrated chromosomal aberrations in approximately 40–70% of patients with primary MDS at diagnosis and in 95% of patients with tMDS (Heim and Mitelman, 1995; Fenaux et al., 1996; Vallespí et al., 1998; Olney and Le Beau, 2001; Mauritzson et al., 2002). Numerous recurrent structural and numerical chromosomal abnormalities have also been identified in MDS and the most commonly observed are losses of chromosomes 5 and/or 7, deletions of the long arm of chromosomes 5, 7, and/or 20, and trisomy of chromosome 8. Less frequently observed abnormalities include structural abnormalities of chromosome 1, the long arm of chromosome 3, deletions of the long arms of chromosomes 11 and 13, loss of the short arm of chromosome 17 due to monosomy or an unbalanced rearrangement, and trisomy of chromosomes 9 and 21. The abnormalities may occur as a sole aberration or as part of complex karyotypes with multiple abnormalities. Many of these changes also characterize AML and the remarkable karyotypic similarity between MDS and AML emphasizes the pathobiologic similarity between the two disorders. FISH panel testing in MDS has most commonly used probes specific for chromosomes 5 (EGR1 at 5q31 and D5S23, D5S721 at 5p15.2), 7 (D7S486 at 7q31 and CEP7, the centromere probe), 8 (also the centromere probe), and 20 (D20S108 at 20q12), and in some reports chromosomes 11 (MLL at 11q23) and 13 (RB1 at 13q14) as well. FISH analyses in MDS have for the most part shown that it is nearly as sensitive as routine cytogenetic analysis when used independently, with a slight increase in detection rate of abnormalities when both methods are used in combination since some abnormalities go undetected by either approach (Romeo et al., 2002). Studies favoring the use of FISH have shown that about 15–20% of MDS patients with a normal karyotype have minor clones of cells (about 15– 30% of interphase nuclei) with abnormalities or submicroscopic rearrangements that are clinically relevant and influence therapy (Bernasconi et al., 2003, 2006). Such studies have also shown that FISH abnormalities were observed more frequently among patients with increased bone marrow blasts, a higher rate of progression to AML and were predictive of worse prognosis (Rigolin et al., 2001). Others have indicated that FISH panel testing has limited utility in MDS except in cases where cytogenetic analysis is not possible (Ketterling et al., 2002; Cherry et al, 2003; Sun et al.,
2004), such as in specimens with low or no mitotic cells, poor chromosome morphology, or specimens delayed in transit. They conclude that in contrast to its application in lymphoid malignancies, FISH panel testing for MDS appears to not be an efficient and cost-effective screening method in the diagnosis of MDS and that routine cytogenetic analysis should remain the method of choice for studying bone marrow aspirates since it provides an overall picture. Based on the cytogenetic pattern, according to the IPSS for MDS (Greenberg et al., 1997), patients were separated for both survival and AML evolution into three prognostic subgroups, good, intermediate or poor. Good prognosis group Patients with favorable outcomes constituted about 50– 70% of the patients and had a normal karyotype, loss of the Y chromosome, deletion 5q, or deletion 20q as the sole changes. Normal karyotype. About 30–60% of patients with MDS have a normal karyotype. This subgroup is heterogeneous and contains abnormalities not detected by routine cytogenetic analysis. By FISH analysis up to 18% of such cases have been found to carry common chromosomal aberrations (–5/5q–, –7/7q–, +8, +11q, 17p–) in 15–32% of cells analyzed (Rigolin et al., 2001; Bernasconi et al., 2003, 2006). These clones of abnormal cells may be clinically relevant because they identify a subset of patients within this group that may have an inferior prognosis. Deletion 5q. Deletion of 5q has been reported in about 10–20% of de novo MDS patients and about 40% of t-MDS (Vallespí et al., 1998). Patients with deletion of 5q have a relatively good prognosis when it is present as the sole karyotypic abnormality. However, when additional abnormalities are present the prognosis is poor with early progression to leukemia, resistance to treatment and short survival (Jacobs et al., 1986; Giagounidis et al., 2004). Transfusion-dependent, low or intermediate risk MDS patients with deletion 5q receive targeted therapy with lenalidomide resulting in durable cytogenetic remission and hematological responses in the majority of patients (List et al., 2006; Nimer, 2006). Deletion 20q. Deletion of 20q is present in about 4–5% of patients with MDS and 7% of t-MDS (Fenaux et al., 1996; Vallespí et al., 1998). It is associated with low-risk disease, low rate of progression to AML, and when present as the sole abnormality, patients with deletion of 20q have a favorable prognosis (Wattel et al., 1993; Greenberg et al., 1997). However, when deletion of 20q is part of a complex karyotype, the prognosis is less favorable with a high rate of transformation to acute leukemia (Campbell and Garson, 1994; Brezinova et al., 2005). The most frequent additional abnormalities were deletions of 5q and/or 7q. Loss of the Y chromosome. Y chromosome loss has been described in hematologic malignancies as well as in bone marrows of hematologically normal elderly men (UKCCG, 1992). Thus, although this finding does not indicate the presence of MDS, when present in patients diagnosed with MDS the prognosis is improved (Greenberg et al., 1997).
Cytogenet Genome Res 118:284–296 (2007)
291
Poor prognosis group About 16–32% of patients belonged to this group. Patients with abnormal cytogenetic findings carried a higher risk of progression to AML than those with a normal karyotype and those with complex karyotypes (three or more abnormalities) or abnormalities of chromosome 7 progressed more often to AML (Morel et al., 1993; Greenberg et al., 1997). Complex karyotype. Complex karyotypes are variably defined and include multiple structural rearrangements (more than three), the presence of multiple clones or complex rearrangements. They are observed in 10–20% of patients with primary MDS and up to 90% of patients with tMDS (Olney and Le Beau, 2001). Most patients with complex karyotypes have abnormalities of chromosomes 5 and/or 7, together with other abnormalities. Complex karyotypic abnormalities are associated with poor prognoses. Abnormalities of chromosome 7. Monosomy or deletion of the long arm of chromosome 7 is present in about 5% of de novo MDS and 55% of t-MDS (Heim and Mitelman, 1995; Sole et al., 2000). It is characterized by disease progression and poor prognosis. Intermediate prognosis group The remaining 14–17% of patients with various single and double abnormalities not included in the other two groups were included in this group. Trisomy 8. About 10% of patients with MDS have trisomy of chromosome 8 (Morel et al., 1993; Vallespí et al., 1998; Sole et al., 2000). It is also associated with other recurring abnormalities of known prognostic significance. Other abnormalities Rearrangements of 11q23. About 5% of MDS patients reveal abnormalities involving band 11q23. It frequently occurs as part of a complex karyotype and is associated with other abnormalities, usually monosomy 7 or deletion of 7q. Prognosis is poor with progression to AML in 20–30% of cases (Bain et al., 1998). Deletion of 17p. Rearrangements resulting in loss of the short arm of chromosome 17 have been seen in about 5% of MDS (Lai et al., 1995b; Soenen et al., 1998). Loss of 17p occurred through monosomy, deletion, unbalanced translo-
cations between 17p and another chromosome or isochromosome formation. Most of the patients had additional complex cytogenetic findings. Abnormalities involving chromosomes 5 and 7 were frequently present. Clinically, patients with 17p deletion had aggressive disease with poor response to therapy and short survival. Conclusion
Cytogenetic and FISH analyses have elucidated the basic genetic and clinical heterogeneity of the various leukemias and lymphomas. Recurrent cytogenetic abnormalities have been shown to be associated with specific clinical characteristics and treatment outcomes in most hematological malignancies and their identification has had critical prognostic value by directing therapeutic decisions. FISH panels, with their improved detection rate of the recurrent aberrations, have become established both as a means of identifying specific chromosomal abnormalities, and as genetic indicators of prognostic outcome and are regularly utilized in the initial assessment of a patient, particularly when the cytogenetic results are normal. They are also used to differentiate the heterogeneous nature of the leukemias, manifested by the different genetic subtypes. Evaluation of these and the monitoring of evolving cytogenetic abnormalities throughout the course of the disease allows clinicians to treat patients with more effective and targeted therapeutic regimens and assess their responses more effectively. The significance of many abnormalities, particularly the less frequent ones, is currently unknown and most have been combined into single prognostic categories. Characterization of these and new genetic abnormalities that are still being discovered will further contribute to the development of therapeutic agents designed to target the unique genes involved and evaluation of their clinical outcome will help define their prognostic significance. Routine cytogenetic analysis, however, still remains an irreplaceable procedure in the initial workup of a patient, because it provides a comprehensive representation of the spectrum of abnormalities. The strategy of using both procedures together is highly informative and sensitive, and recommended.
References Amiel A, Mulchanov I, Elis A, Gaber E, Manor Y, et al: Deletion of 6q27 in chronic lymphocytic leukemia and multiple myeloma detected by fluorescence in situ hybridization. Cancer Genet Cytogenet 112:53–56 (1999). Anastasi J, Le Beau MM, Vardiman JW, Fernald AA, Larson RA, Rowley JD: Detection of trisomy 12 in chronic lymphocytic leukemia by fluorescence in situ hybridization to interphase cells: a simple and sensitive method. Blood 79: 1796–1801 (1992).
292
Aoun P, Blair HE, Smith LM, Dave BJ, Lynch J, et al: Fluorescence in situ hybridisation detection of cytogenetic abnormalities in B-cell chronic lymphocytic leukemia/small lymphocytic lymphoma. Leuk Lymphoma 45:1595–1603 (2004). Athanasiadou A, Stamatopoulos K, Tsompanakou A, Gaitatzi M, Kalogiannidis P, et al: Clinical, immunophenotypic, and molecular profiling of trisomy 12 in chronic lymphocytic leukemia and comparison with other karyotypic subgroups defined by cytogenetic analysis. Cancer Genet Cytogenet 168:109–119 (2006).
Cytogenet Genome Res 118:284–296 (2007)
Aul C, Gattermann N, Heyll A, Germing U, Derigs G, Schneider W: Primary myelodysplastic syndromes: analysis of prognostic factors in 235 patients and proposals for an improved scoring system. Leukemia 6:52–59 (1992). Avet-Loiseau H, Brigaudeau C, Morineau N, Talmant P, Lai JL, et al: High incidence of cryptic translocations involving the Ig heavy chain gene in multiple myeloma, as shown by fluorescence in situ hybridization. Genes Chromosomes Cancer 24: 9–15 (1999).
Avet-Loiseau H, Daviet A, Sauner S, Bataille R; Intergroupe Francophone du Myelome: Chromosome 13 abnormalities in multiple myeloma are mostly monosomy 13. Br J Haematol 111: 1116– 1117 (2000). Avet-Loiseau H, Facon T, Grosbois B, Magrangeas F, Rapp MJ, et al; Intergroupe Francophone du Myelome: Oncogenesis of multiple myeloma: 14q32 and 13q chromosomal abnormalities are not randomly distributed, but correlate with natural history, immunological features, and clinical presentation. Blood 99: 2185–2191 (2002). Avet-Loiseau H, Garand R, Lode L, Harousseau J-L, Bataille R: Translocation t(11;14)(q13;q32) is the hallmark of IgM, IgE, and nonsecretory multiple myeloma variants. Blood 101: 1570–1571 (2003). Avet-Loiseau H, Attal M, Moreau P, Charbonnel C, Garban F, et al: Genetic abnormalities and survival in multiple myeloma: the experience of the Intergroupe Francophone du Myelome. Blood 109:3489–3495 (2007). Bain BJ, Moorman AV, Johansson B, Mehta AB, Secker-Walker LM: Myelodysplastic syndromes associated with 11q23 abnormalities. European 11q23 Workshop participants. Leukemia 12: 834–839 (1998). Bang SM, Kim YR, Cho HI, Chi HS, Seo EJ, et al: Identification of 13q deletion, trisomy 1q, and IgH rearrangement as the most frequent chromosomal changes found in Korean patients with multiple myeloma. Cancer Genet Cytogenet 16: 124–132 (2006). Bataille R, Boccadoro M, Klein B, Durie B, Pileri A: C-reactive protein and beta-2 microglobulin produce a simple and powerful myeloma staging system. Blood 80:733–737 (1992). Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton DA, et al: Proposals for the classification of the acute leukaemias. French-AmericanBritish (FAB) co-operative group. Br J Haematol 33:451–458 (1976). Bergsagel PL, Chesi M, Nardini E, Brents LA, Kirby SL, Kuehl WM: Promiscuous translocations into immunoglobulin heavy chain switch regions in multiple myeloma. Proc Natl Acad Sci USA 93: 13931–13936 (1996). Bernasconi P, Cavigliano PM, Boni M, Calatroni S, Klersy C, et al: Is FISH a relevant prognostic tool in myelodysplastic syndromes with a normal chromosome pattern on conventional cytogenetics? A study on 57 patients. Leukemia 17:2107–2112 (2003). Bernasconi P, Boni M, Cavigliano PM, Calatroni S, Giardini I, et al: Clinical relevance of cytogenetics in myelodysplastic syndromes. Ann NY Acad Sci 1089:395–410 (2006). Bigoni R, Cuneo A, Roberti MG, Bardi A, Rigolin GM, et al: Chromosome aberrations in atypical chronic lymphocytic leukemia: a cytogenetic and interphase cytogenetic study. Leukemia 11: 1933–1940 (1997). Binet JL, Auquier A, Dighiero G, Chastang C, Piguet H, et al: A new prognostic classification of chronic lymphocytic leukemia derived from a multivariate survival analysis. Cancer 48: 198–206 (1981). Bloomfield CD, Lawrence D, Byrd JC, Carroll A, Pettenati MJ, et al: Frequency of prolonged remission duration after high-dose cytarabine intensification in acute myeloid leukemia varies by cytogenetic subtype. Cancer Res 58: 4173–4179 (1998). Brezinova J, Zemanova Z, Ransdorfova S, Sindelarova L, Siskova M, et al: Prognostic significance of del(20q) in patients with hematological malignancies. Cancer Genet Cytogenet 160: 188–192 (2005).
Buchner T, Hiddemann W, Wormann B, Loffler H, Gassmann W, et al: Double induction strategy for acute myeloid leukemia: the effect of highdose cytarabine with mitoxantrone instead of standard-dose cytarabine with daunorubicin and 6-thioguanine: a randomized trial by the German AML Cooperative Group. Blood 3: 4116–4124 (1999). Byrd JC, Mrozek K, Dodge RK, Carroll AJ, Edwards CG, et al; Cancer and Leukemia Group B (CALGB 8461): Pretreatment cytogenetic abnormalities are predictive of induction success, cumulative incidence of relapse, and overall survival in adult patients with de novo acute myeloid leukemia: results from Cancer and Leukemia Group B (CALGB 8461). Blood 100: 4325–4336 (2002). Campbell LJ, Garson OM: The prognostic significance of deletion of the long arm of chromosome 20 in myeloid disorders. Leukemia 8: 67– 71 (1994). Chang H, Li D, Zhuang L, Nie E, Bouman D, et al: Detection of chromosome 13q deletions and IgH translocations in patients with multiple myeloma by FISH: comparison with karyotype analysis. Leuk Lymphoma 45: 965–969 (2004a). Chang H, Sloan S, Li D, Zhuang L, Yi QL, et al: The t(4;14) is associated with poor prognosis in myeloma patients undergoing autologous stem cell transplant. Br J Haematol 125: 64–68 (2004b). Chang H, Salma F, Yi QL, Patterson B, Brien B, Minden MD: Prognostic relevance of immunophenotyping in 379 patients with acute myeloid leukemia. Leuk Res 28:43–48 (2004c). Chang H, Qi C, Yi Q-L, Reece D, Stewart AK: p53 gene deletion detected by fluorescence in situ hybridization is an adverse prognostic factor for patients with multiple myeloma following autologous stem cell transplantation. Blood 105:358–360 (2005). Chen CC, Yang CF, Lee KD, You JY, Yu YB, et al: Complex karyotypes confer a poor survival in adult acute myeloid leukemia with unfavorable cytogenetic abnormalities. Cancer Genet Cytogenet 174:138–146 (2007). Cherry AM, Brockman SR, Paternoster SF, Hicks GA, Neuberg D, et al: Comparison of interphase FISH and metaphase cytogenetics to study myelodysplastic syndrome: an Eastern Cooperative Oncology Group (ECOG) study. Leuk Res 27:1085–1090 (2003). Chesi M, Nardini E, Lim RS, Smith KD, Kuehl WM, Bergsagel PL: The t(4; 14) translocation in myeloma dysregulates both FGFR3 and a novel gene, MMSET, resulting in IgH/MMSET hybrid transcripts. Blood 92: 3025–3034 (1998). Cigudosa JC, Rao PH, Calasanz MJ, Odero MD, Michaeli J, et al: Characterization of nonrandom chromosomal gains and losses in multiple myeloma by comparative genomic hybridization. Blood 91:3007–3010 (1998). Cline MJ: The molecular basis of leukemia. N Engl J Med 330:328–336 (1994). Coignet L, Bertheas MF, Vasselon C, Jaubert J, Reynaud J, et al: Trisomy 12 in B-cell chronic lymphocytic leukemia: interphase study by in situ hybridization in 75 patients. Nouv Rev Fr Hematol 35:129–134 (1993). Cremer FW, Kartal M, Hose D, Bila J, Buck I, et al: High incidence and intraclonal heterogeneity of chromosome 11 aberrations in patients with newly diagnosed multiple myeloma detected by multiprobe interphase FISH. Cancer Genet Cytogenet 161:116–124 (2005).
Crespo M, Bosch F, Villamor N, Bellosillo B, Colomer D, et al: ZAP-70 expression as a surrogate for immunoglobulin-variable-region mutations in chronic lymphocytic leukemia. N Engl J Med 348:1764–1775 (2003). Criel A, Verhoef G, Vlietinck R, Mecucci C, Billiet J, et al: Further characterization of morphologically defined typical and atypical CLL: a clinical, immunophenotypic, cytogenetic and prognostic study on 390 cases. Br J Haematol 97: 383–391 (1997). Cuneo A, Wlodarska I, Sayed Aly M, Piva N, Carli MG, et al: Non-radioactive in situ hybridization for the detection and monitoring of trisomy 12 in B-cell chronic lymphocytic leukaemia. Br J Haematol 81: 192–196 (1992). Cuneo A, Rigolin GM, Bigoni R, De Angeli C, Veronese A, et al: Chronic lymphocytic leukemia with 6q– shows distinct hematological features and intermediate prognosis. Leukemia 18:476– 483 (2004). Debes-Marun CS, Dewald GW, Bryant S, Picken E, Santana-Davila R, et al: Chromosome abnormalities clustering and its implications for pathogenesis and prognosis in myeloma. Leukemia 17:427–436 (2003). Desikan R, Barlogie B, Sawyer J, Ayers D, Tricot G, et al: Results of high-dose therapy for 1000 patients with multiple myeloma: durable complete remissions and superior survival in the absence of chromosome 13 abnormalities. Blood 95:4008–4010 (2000). Dewald GW, Kyle RA, Hicks GA, Greipp PR: The clinical significance of cytogenetic studies in 100 patients with multiple myeloma, plasma cell leukemia, or amyloidosis. Blood 66:380–390 (1985). Dewald GW, Brockman SR, Paternoster SF, Bone ND, O’Fallon JR, et al: Chromosome anomalies detected by interphase fluorescence in situ hybridization: correlation with significant biological features of B-cell chronic lymphocytic leukaemia. Br J Haematol 121: 287–295 (2003). Dewald GW, Therneau T, Larson D, Lee YK, Fink S, et al: Relationship of patient survival and chromosome anomalies detected in metaphase and/ or interphase cells at diagnosis of myeloma. Blood 106:3553–3558 (2005). Dicker F, Schnittger S, Haferlach T, Kern W, Schoch C: Immunostimulatory oligonucleotide-induced metaphase cytogenetics detect chromosomal aberrations in 80% of CLL patients: A study of 132 CLL cases with correlation to FISH, IgVH status, and CD38 expression. Blood 108: 3152–3160 (2006). Dimopoulos MA, Barlogie B, Smith TL, Alexanian R: High serum lactate dehydrogenase level as a marker for drug resistance and short survival in multiple myeloma. Ann Intern Med 115: 931– 935 (1991). Dohner H, Fischer K, Bentz M, Hansen K, Benner A, et al: p53 gene deletion predicts for poor survival and non-response to therapy with purine analogs in chronic B-cell leukemias. Blood 85: 1580–1589 (1995). Dohner H, Stilgenbauer S, James MR, Benner A, Weilguni T, et al: 11q deletions identify a new subset of B-cell chronic lymphocytic leukemia characterized by extensive nodal involvement and inferior prognosis. Blood 89: 2516–2522 (1997). Dohner H, Stilgenbauer S, Dohner K, Bentz, M, Lichter P: Chromosome aberrations in B-cell chronic lymphocytic leukemia: reassessment based on molecular cytogenetic analysis. J Mol Med 77:266–281 (1999). Dohner H, Stilgenbauer S, Benner A, Leupolt E, Krober A, et al: Genomic aberrations and survival in chronic lymphocytic leukemia. N Engl J Med 343: 1910–1916 (2000).
Cytogenet Genome Res 118:284–296 (2007)
293
Drach J, Schuster J, Nowotny H, Angerler J, Rosenthal F, et al: Multiple myeloma: high incidence of chromosomal aneuploidy as detected by interphase fluorescence in situ hybridization. Cancer Res 55: 3854–3859 (1995). Drach J, Ackermann J, Fritz E, Kromer E, Schuster R, et al: Presence of a p53 gene deletion in patients with multiple myeloma predicts for short survival after conventional-dose chemotherapy. Blood 92:802–809 (1998). Durie BG, Salmon SE: A clinical staging system for multiple myeloma. Correlation of measured myeloma cell mass with presenting clinical features, response to treatment and survival. Cancer 36:842–854 (1975). el Rouby S, Thomas A, Costin D, Rosenberg CR, Potmesil M, et al: p53 gene mutation in B-cell chronic lymphocytic leukemia is associated with drug resistance and is independent of MDR1/MDR3 gene expression. Blood 82: 3452– 3459 (1993). Escudier SM, Pereira-Leahy JM, Drach JW, Weier HU, Goodacre AM, et al: Fluorescent in situ hybridization and cytogenetic studies of trisomy 12 in chronic lymphocytic leukemia. Blood 81: 2702–2707 (1993). Estey E, Dohner H: Acute myeloid leukaemia. Lancet 368:1894–1907 (2006). Facon T, Avet-Loiseau H, Guillerm G, Moreau P, Genevieve F, et al; Intergroupe Francophone du Myelome: Chromosome 13 abnormalities identified by FISH analysis and serum beta2-microglobulin produce a powerful myeloma staging system for patients receiving high-dose therapy. Blood 97:1566–1571 (2001). Fassas AB, Spencer T, Sawyer J, Zangari M, Lee CK, et al: Both hypodiploidy and deletion of chromosome 13 independently confer poor prognosis in multiple myeloma. Br J Haematol 118: 1041–1047 (2002). Fenaux P, Morel P, Lai JL: Cytogenetics of myelodysplastic syndromes. Semin Hematol 33: 127– 138 (1996). Fenaux P, Chomienne C, Degos L: Acute promyelocytic leukemia: biology and treatment. Semin Oncol 24:92–102 (1997). Fink S, Geyer S, Shanafelt T, Smoley S, Paternoster S, et al: Clinical significance of homozygous D13S319 deletion in B-cell chronic lymphocytic leukemia (B-CLL). Blood 104: 2799a (2004). Finn WG, Thangavelu M, Yelavarthi KK, Goolsby CL, Tallman MS, et al: Karyotype correlates with peripheral blood morphology and immunophenotype in chronic lymphocytic leukemia. Am J Clin Pathol 105: 458–467 (1996). Fonseca R, Hoyer JD, Aguayo P, Jalal SM, Ahmann GJ, et al: Clinical significance of the translocation (11; 14)(q13;q32) in multiple myeloma. Leuk Lymphoma 35: 599–605 (1999a). Fonseca R, Coignet LJ, Dewald GW: Cytogenetic abnormalities in multiple myeloma. Hematol Oncol Clin North Am 13: 1169–1180, viii (1999b). Fonseca R, Oken M, Greipp P: The t(4; 14)(p16.3; q32) is strongly associated with chromosome 13 abnormalities in both multiple myeloma and monoclonal gammopathies of undetermined significance. Blood 98: 1271–1272 (2001). Fonseca R, Blood EA, Oken MM, Kyle RA, Dewald GW, et al: Myeloma and the t(11; 14)(q13;q32); evidence for a biologically defined unique subset of patients. Blood 99: 3735–3741 (2002a). Fonseca R, Harrington D, Oken MM, Dewald GW, Bailey RJ, et al: Biological and prognostic significance of interphase fluorescence in situ hybridization detection of chromosome 13 abnormalities (delta13) in multiple myeloma: an eastern cooperative oncology group study. Cancer Res 62:715–720 (2002b).
294
Fonseca R, Bailey RJ, Ahmann GJ, Rajkumar SV, Hoyer JD, et al: Genomic abnormalities in monoclonal gammopathy of undetermined significance. Blood 100: 1417–1424 (2002c). Fonseca R, Blood E, Rue M, Harrington D, Oken MM, et al: Clinical and biologic implications of recurrent genomic aberrations in myeloma. Blood 101:4569–4575 (2003). Fonseca R, Barlogie B, Bataille R, Bastard C, Bergsagel PL, et al: Genetics and cytogenetics of multiple myeloma: a workshop report. Cancer Res 64:1546–1558 (2004). Fritz E, Ludwig H, Kundi M: Prognostic relevance of cellular morphology in multiple myeloma. Blood 63:1072–1079 (1984). Geisler CH, Philip P, Christensen BE, Hou-Jensen K, Pedersen NT, et al: In B-cell chronic lymphocytic leukaemia chromosome 17 abnormalities and not trisomy 12 are the single most important cytogenetic abnormalities for the prognosis: a cytogenetic and immunophenotypic study of 480 unselected newly diagnosed patients. Leuk Res 21:1011–1023 (1997). Gertz MA, Lacy MQ, Dispenzieri A, Greipp PR, Litzow MR, et al: Clinical implications of t(11; 14)(q13;q32), t(4; 14)(p16.3;q32), and –17p13 in myeloma patients treated with high-dose therapy. Blood 106:2837–2840 (2005). Ghia P, Guida G, Stella S, Gottardi D, Geuna M, et al: The pattern of CD38 expression defines a distinct subset of chronic lymphocytic leukemia (CLL) patients at risk of disease progression. Blood 101:1262–1269 (2003). Giagounidis AA, Germing U, Haase S, Hildebrandt B, Schlegelberger B, et al: Clinical, morphological, cytogenetic, and prognostic features of patients with myelodysplastic syndromes and del(5q) including band q31. Leukemia 18: 113– 119 (2004). Glassman AB, Hayes KJ: The value of fluorescence in situ hybridization in the diagnosis and prognosis of chronic lymphocytic leukemia. Cancer Genet Cytogenet 158: 88–91 (2005). Goorha S, Glenn MJ, Drozd-Borysiuk E, Chen Z: A set of commercially available fluorescent insitu hybridization probes efficiently detects cytogenetic abnormalities in patients with chronic lymphocytic leukemia. Genet Med 6: 48–53 (2004). Gozzetti A, Crupi R, Tozzuoli D, Raspadori D, Forconi F, Lauria F: Molecular cytogenetic analysis of B-CLL patients with aggressive disease. Hematology 9: 383–385 (2004). Greenberg P, Cox C, LeBeau MM, Fenaux P, Morel P, et al: International scoring system for evaluating prognosis in myelodysplastic syndromes. Blood 89:2079–2088 (1997). Greipp PR, Lust JA, O’Fallon WM, Katzmann JA, Witzig TE, Kyle RA: Plasma cell labeling index and beta 2-microglobulin predict survival independent of thymidine kinase and C-reactive protein in multiple myeloma. Blood 81: 3382– 3387 (1993). Grimwade D, Walker H, Oliver F, Wheatley K, Harrison C, et al; The Medical Research Council Adult and Children’s Leukaemia Working Parties: The importance of diagnostic cytogenetics on outcome in AML: analysis of 1,612 patients entered into the MRC AML 10 trial. Blood 92: 2322–2333 (1998). Grimwade D, Walker H, Harrison G, Oliver F, Chatters S, et al; Medical Research Council Adult Leukemia Working Party: The predictive value of hierarchical cytogenetic classification in older adults with acute myeloid leukemia (AML): analysis of 1065 patients entered into the United Kingdom Medical Research Council AML11 trial. Blood 98:1312–1320 (2001).
Cytogenet Genome Res 118:284–296 (2007)
Guglielmelli T, Giugliano E, Cappia S, Papotti M, Saglio G: Frequency and distribution of trisomy 11 in multiple myeloma patients: relation with overexpression of CCND1 and t(11;14). Cancer Genet Cytogenet 173: 51–56 (2007). Gupta V, Chun K, Yi QL, Minden M, Schuh A, et al: Disease biology rather than age is the most important determinant of survival of patients 1 or = 60 years with acute myeloid leukemia treated with uniform intensive therapy. Cancer 103: 2082–2090 (2005). Gutierrez NC, Garcia JL, Hernandez JM, Lumbreras E, Castellanos M, et al: Prognostic and biologic significance of chromosomal imbalances assessed by comparative genomic hybridization in multiple myeloma. Blood 104: 2661– 2666 (2004). Hamblin TJ, Davis Z, Gardiner A, Oscier DG, Stevenson FK: Unmutated Ig V(H) genes are associated with a more aggressive form of chronic lymphocytic leukemia. Blood 94: 1848–1854 (1999). Harrison CJ, Mazzullo H, Cheung KL, Gerrard G, Jalali GR, et al: Cytogenetics of multiple myeloma: interpretation of fluorescence in situ hybridization results. Br J Haematol 120:944–952 (2003). Heim S, Mitelman F: Cancer Cytogenetics (WileyLiss, New York 1995). Hernandez JM, Mecucci C, Criel A, Meeus P, Michaux I, et al: Cytogenetic analysis of B cell chronic lymphoid leukemias classified according to morphologic and immunophenotypic (FAB) criteria. Leukemia 9:2140–2146 (1995). Hoyer JD, Hanson CA, Fonseca R, Greipp PR, Dewald GW, Kurtin PJ: The (11;14)(q13;q32) translocation in multiple myeloma. A morphologic and immunohistochemical study. Am J Clin Pathol 113: 831–837 (2000). Jacobs RH, Cornbleet MA, Vardiman JW, Larson RA, Le Beau MM, Rowley JD: Prognostic implications of morphology and karyotype in primary myelodysplastic syndromes. Blood 67: 1765–1772 (1986). Jaffe ES, Harris NL, Stein H, Vardiman JW (eds): World Health Organization Classification of Tumors. Pathology and Genetics of Tumors of Haematomoietic and Lymphoid Tissues (IARC Press, Lyon 2001). Jelinek DF, Tschumper RC, Geyer SM, Bone ND, Dewald GW, et al: Analysis of clonal B-cell CD38 and immunoglobulin variable region sequence status in relation to clinical outcome for B-chronic lymphocytic leukaemia. Br J Haematol 115:854–861 (2001). Juliusson G, Gahrton G: Chromosome aberrations in B-cell chronic lymphocytic leukemia. Pathogenetic and clinical implications. Cancer Genet Cytogenet 45: 143–160 (1990). Juliusson G, Oscier DG, Fitchett M, Ross FM, Stockdill G, et al: Prognostic subgroups in B-cell chronic lymphocytic leukemia defined by specific chromosomal abnormalities. N Engl J Med 323:720–724 (1990). Keats JJ, Reiman T, Maxwell CA, Taylor BJ, Larratt LM, et al: In multiple myeloma, t(4;14)(p16;q32) is an adverse prognostic factor irrespective of FGFR3 expression. Blood 101: 1520–1529 (2003). Ketterling RP, Wyatt WA, VanWier SA, Law M, Hodnefield JM, et al: Primary myelodysplastic syndrome with normal cytogenetics: utility of ‘FISH panel testing’ and M-FISH. Leuk Res 26: 235–240 (2002). Konigsberg R, Zojer N, Ackermann J, Kromer E, Kittler H, et al: Predictive role of interphase cytogenetics for survival of patients with multiple myeloma. J Clin Oncol 18:804–812 (2000).
Krober A, Seiler T, Benner A, Bullinger L, Bruckle E, et al: V(H) mutation status, CD38 expression level, genomic aberrations, and survival in chronic lymphocytic leukemia. Blood 100: 1410–1416 (2002). Kyle RA, Gertz MA, Witzig TE, Lust JA, Lacy MQ, et al: Review of 1027 patients with newly diagnosed multiple myeloma. Mayo Clin Proc 78: 21–33 (2003). Lai JL, Zandecki M, Mary JY, Bernardi F, Izydorczyk V, et al: Improved cytogenetics in multiple myeloma: a study of 151 patients including 117 patients at diagnosis. Blood 85: 2490–2497 (1995a). Lai JL, Preudhomme C, Zandecki M, Flactif M, Vanrumbeke M, et al: Myelodysplastic syndromes and acute myeloid leukemia with 17p deletion. An entity characterized by specific dysgranulopoiesis and a high incidence of P53 mutations. Leukemia 9:370–381 (1995b). Liebisch P, Wendl C, Wellmann A, Krober A, Schilling G, et al: High incidence of trisomies 1q, 9q, and 11q in multiple myeloma: results from a comprehensive molecular cytogenetic analysis. Leukemia 17:2535–2537 (2003). Liebisch P, Scheck D, Erne SA, Wellmann A, Wendl C, et al: Duplication of chromosome arms 9q and 11q: evidence for a novel, 14q32 translocation-independent pathogenetic pathway in multiple myeloma. Genes Chromosomes Cancer 42:78–81 (2005). List A, Dewald G, Bennett J, Giagounidis A, Raza A, et al; Myelodysplastic Syndrome-003 Study Investigators: Lenalidomide in the myelodysplastic syndrome with chromosome 5q deletion. N Engl J Med 355:1456–1465 (2006). Ma X, Does M, Raza A, Mayne ST: Myelodysplastic syndromes: incidence and survival in the United States. Cancer 109: 1536–1542 (2007). Malcovati L, Porta MG, Pascutto C, Invernizzi R, Boni M, et al: Prognostic factors and life expectancy in myelodysplastic syndromes classified according to WHO criteria: a basis for clinical decision making. J Clin Oncol 23: 7594–7603 (2005). Marcucci G, Mrozek K, Ruppert AS, Maharry K, Kolitz JE, et al: Prognostic factors and outcome of core binding factor acute myeloid leukemia patients with t(8; 21) differ from those of patients with inv(16): a Cancer and Leukemia Group B study. J Clin Oncol 23: 5705–5717 (2005). Matutes E, Oscier D, Garcia-Marco J, Ellis J, Copplestone A, et al: Trisomy 12 defines a group of CLL with atypical morphology: correlation between cytogenetic, clinical and laboratory features in 544 patients. Br J Haematol 92:382–388 (1996). Mauritzson N, Albin M, Rylander L, Billström R, Ahlgren T, et al: Pooled analysis of clinical and cytogenetic features in treatment-related and de novo adult acute myeloid leukemia and myelodysplastic syndromes based on a consecutive series of 761 patients analyzed 1976–1993 and on 5098 unselected cases reported in the literature 1974–2001. Leukemia 16: 2366–2378 (2002). Mayr C, Speicher MR, Kofler DM, Buhmann R, Strehl J, et al: Chromosomal translocations are associated with poor prognosis in chronic lymphocytic leukemia. Blood 107:742–751 (2006). Moreau P, Facon T, Leleu X, Morineau N, Huyghe P, et al; for the Intergroupe Francophone du Myélome: Recurrent 14q32 translocations determine the prognosis of multiple myeloma, especially in patients receiving intensive chemotherapy. Blood 100:1579–1583 (2002).
Morel P, Hebbar M, Lai JL, Duhamel A, Preudhomme C, et al: Cytogenetic analysis has strong independent prognostic value in de novo myelodysplastic syndromes and can be incorporated in a new scoring system: a report on 408 cases. Leukemia 7:1315–1323 (1993). Mrozek K, Heinonen K, Bloomfield CD: Prognostic value of cytogenetic findings in adults with acute myeloid leukemia. Int J Hematol 72: 261– 271 (2000). Mrozek K, Heinonen K, Bloomfield CD: Clinical importance of cytogenetics in acute myeloid leukaemia. Best Pract Res Clin Haematol 14: 19–47 (2001). Mrozek K, Marcucci G, Paschka P, Whitman SP, Bloomfield CD: Clinical relevance of mutations and gene-expression changes in adult acute myeloid leukemia with normal cytogenetics: are we ready for a prognostically prioritized molecular classification? Blood 109: 431–448 (2007). Neilson JR, Auer R, White D, Bienz N, Waters JJ, et al: Deletions at 11q identify a subset of patients with typical CLL who show consistent disease progression and reduced survival. Leukemia 11:1929–1932 (1997). Neri A, Baldini L, Trecca D, Cro L, Polli E, Maiolo AT: p53 gene mutations in multiple myeloma are associated with advanced forms of malignancy. Blood 81: 128–135 (1993). Nevill TJ, Fung HC, Shepherd JD, Horsman DE, Nantel SH, et al: Cytogenetic abnormalities in primary myelodysplastic syndrome are highly predictive of outcome after allogeneic bone marrow transplantation. Blood 92: 1910–1917 (1998). Nimer SD: Clinical management of myelodysplastic syndromes with interstitial deletion of chromosome 5q. J Clin Oncol 24: 2576–2582 (2006). Nishida K, Tamura A, Nakazawa N, Ueda Y, Abe T, et al: The Ig heavy chain gene is frequently involved in chromosomal translocations in multiple myeloma and plasma cell leukemia as detected by in situ hybridization. Blood 90: 526–534 (1997). Olney HJ, Le Beau MM: The cytogenetics of myelodysplastic syndromes. Best Pract Res Clin Haematol 14:479–495 (2001). Pantou D, Rizou H, Tsarouha H, Pouli A, Papanastasiou K, et al: Cytogenetic manifestations of multiple myeloma heterogeneity. Genes Chromosomes Cancer 42: 44–57 (2005). Pfeilstocker M, Reisner R, Nosslinger T, Gruner H, Nowotny H, et al: Cross-validation of prognostic scores in myelodysplastic syndromes on 386 patients from a single institution confirms importance of cytogenetics. Br J Haematol 106: 455–463 (1999). Que TH, Marco JG, Ellis J, Matutes E, Babapulle VB, et al: Trisomy 12 in chronic lymphocytic leukemia detected by fluorescence in situ hybridization: analysis by stage, immunophenotype, and morphology. Blood 82: 571–575 (1993). Rai KR, Sawitsky A, Cronkite EP, Chanana AD, Levy RN, Pasternack BS: Clinical staging of chronic lymphocytic leukemia. Blood 46: 219– 234 (1975). Rajkumar SV, Greipp PR: Prognostic factors in multiple myeloma. Hematol Oncol Clin North Am 13:1295–314 (1999). Reddy KS: Chronic lymphocytic leukaemia profiled for prognosis using a fluorescence in situ hybridisation panel. Br J Haematol 132: 705– 722 (2006).
Ries LAG, Harkins D, Krapcho M, Mariotto A, Miller BA, et al (eds): SEER Cancer Statistics Review, 1975–2003. (National Cancer Institute, Bethesda, MD 2006). Rigolin GM, Bigoni R, Milani R, Cavazzini F, Roberti MG, et al: Clinical importance of interphase cytogenetics detecting occult chromosome lesions in myelodysplastic syndromes with normal karyotype. Leukemia 15: 1841– 1847 (2001). Robillard N, Avet-Loiseau H, Garand R, Moreau P, Pineau D, et al: CD20 is associated with a small mature plasma cell morphology and t(11; 14) in multiple myeloma. Blood 102: 1070–1071 (2003). Romeo M, Chauffaille Mde L, Silva MR, Bahia DM, Kerbauy J: Comparison of cytogenetics with FISH in 40 myelodysplastic syndrome patients. Leuk Res 26:993–996 (2002). Sawyer JR, Waldron JA, Jagannath S, Barlogie B: Cytogenetic findings in 200 patients with multiple myeloma. Cancer Genet Cytogenet 82: 41– 49 (1995). Sawyer JR, Lukacs JL, Thomas EL, Swanson CM, Goosen LS, et al: Multicolour spectral karyotyping identifies new translocations and a recurring pathway for chromosome loss in multiple myeloma. Br J Haematol 112:167–174 (2001). Schlenk RF, Frohling S, Hartmann F, Fischer JT, Glasmacher A, et al; AML Study Group Ulm: Individual patient data-based meta-analysis of patients aged 16 to 60 years with core binding factor acute myeloid leukemia: a survey of the German Acute Myeloid Leukemia Intergroup. J Clin Oncol 22:3741–3750 (2004). Schmidt-Wolf IG, Glasmacher A, Hahn-Ast C, Juttner A, Schnurr T, et al: Chromosomal aberrations in 130 patients with multiple myeloma studied by interphase FISH: diagnostic and prognostic relevance. Cancer Genet Cytogenet 167:20–25 (2006). Schoch C, Haase D, Haferlach T, Gudat H, Buchner T, et al: Fifty-one patients with acute myeloid leukemia and translocation t(8;21)(q22;q22): an additional deletion in 9q is an adverse prognostic factor. Leukemia 10:1288–1295 (1996). Schoch C, Haferlach T, Haase D, Fonatsch C, Loffler H, et al; German AML Cooperative Study Group: Patients with de novo acute myeloid leukaemia and complex karyotype aberrations show a poor prognosis despite intensive treatment: a study of 90 patients. Br J Haematol 112: 118–126 (2001). Schoch C, Dicker F, Schnittger S, Kern W, Haferlach T: Chromosome banding analysis defines subclasses of CLL with 13q14 deletion and identifies a new mechanism of submicroscopic 13q14 deletions occurring in the breakpoint region of reciprocal translocations. Blood 108: 585a (2006). Segeren CM, Sonneveld P, van der Holt B, Vellenga E, Croockewit AJ, et al; Dutch-Belgian Hemato-Oncology Cooperative Study Group: Overall and event-free survival are not improved by the use of myeloablative therapy following intensified chemotherapy in previously untreated patients with multiple myeloma: a prospective randomized phase 3 study. Blood 101: 2144– 2151 (2003). Shaughnessy J, Tian E, Sawyer J, Bumm K, Landes R, et al: High incidence of chromosome 13 deletion in multiple myeloma detected by multiprobe interphase FISH. Blood 96: 1505–1511 (2000). Shaughnessy JD, Gabrea A, Qi Y, Brents L, Zhan F, et al: Cyclin D3 at 6p21 is dysregulated by recurrent chromosomal translocations to immunoglobulin loci in multiple myeloma. Blood 98: 217–223 (2001).
Cytogenet Genome Res 118:284–296 (2007)
295
Shaughnessy J Jr: Amplification and overexpression of CKS1B at chromosome band 1q21 is associated with reduced levels of p27Kip1 and an aggressive clinical course in multiple myeloma. Hematology 1: 117–126 (2005). Shaughnessy J Jr, Tian E, Sawyer J, McCoy J, Tricot G, et al: Prognostic impact of cytogenetic and interphase fluorescence in situ hybridizationdefined chromosome 13 deletion in multiple myeloma: early results of total therapy II. Br J Haematol 120: 44–52 (2003). Sindelarova L, Michalova K, Zemanova Z, Ransdorfova S, Brezinova J, et al: Incidence of chromosomal anomalies detected with FISH and their clinical correlations in B-chronic lymphocytic leukemia. Cancer Genet Cytogenet 160:27–34 (2005). Slovak ML, Kopecky KJ, Cassileth PA, Harrington DH, Theil KS, et al: Karyotypic analysis predicts outcome of preremission and postremission therapy in adult acute myeloid leukemia: a Southwest Oncology Group/Eastern Cooperative Oncology Group Study. Blood 96: 4075– 4083 (2000). Smadja NV, Bastard C, Brigaudeau C, Leroux D, Fruchart C; Groupe Français de Cytogénétique Hématologique: Hypodiploidy is a major prognostic factor in multiple myeloma. Blood 98: 2229–2238 (2001). Soenen V, Preudhomme C, Roumier C, Daudignon A, Lai JL, Fenaux P: 17p deletion in acute myeloid leukemia and myelodysplastic syndrome. Analysis of breakpoints and deleted segments by fluorescence in situ. Blood 91: 1008–1015 (1998). Sole F, Espinet B, Sanz GF, Cervera J, Calasanz MJ, et al: Incidence, characterization and prognostic significance of chromosomal abnormalities in 640 patients with primary myelodysplastic syndromes. Grupo Cooperativo Espanol de Citogenetica Hematologica. Br J Haematol 108: 346–356 (2000).
296
Soverini S, Cavo M, Cellini C, Terragna C, Zamagni E, et al: Cyclin D1 overexpression is a favourable prognostic variable for newly diagnosed multiple myeloma patients treated with high-dose chemotherapy and single or double autologous transplantation. Blood 102: 1588– 1594 (2003). Stasi R, Del Poeta G, Masi M, Tribalto M, Venditti A, et al: Incidence of chromosome abnormalities and clinical significance of karyotype in de novo acute myeloid leukemia. Cancer Genet Cytogenet 67: 28–34 (1993). Stilgenbauer S, Bullinger L, Benner A, Wildenberger K, Bentz M, et al: Incidence and clinical significance of 6q deletions in B cell chronic lymphocytic leukemia. Leukemia 13: 1331–1334 (1999). Sun G, Lorenzo M, Tripodi J, Chrysostomou C, Stephenson CF, McGill JR: Valuation of cytogenetics and FISH in evaluation of myelodysplastic syndromes: a cancer cytogenetics reference lab’s experience. Blood 104:404a (2004). Tallman MS, Andersen JW, Schiffer CA, Appelbaum FR, Feusner JH, et al: All-trans-retinoic acid in acute promyelocytic leukemia. N Engl J Med 337:1021–1028 (1997). Thornton PD, Gruszka-Westwood AM, Hamoudi RA, Atkinson S, Kaczmarek P, et al: Characterisation of TP53 abnormalities in chronic lymphocytic leukaemia. Hematol 5: 47–54 (2004). Toyama K, Ohyashiki K, Yoshida Y, Abe T, Asano S, et al: Clinical implications of chromosomal abnormalities in 401 patients with myelodysplastic syndromes: a multicentric study in Japan. Leukemia 7:499–508 (1993). Tricot G, Barlogie B, Jagannath S, Bracy D, Mattox S, et al: Poor prognosis in multiple myeloma is associated only with partial or complete deletions of chromosome 13 or abnormalities involving 11q and not with other karyotype abnormalities. Blood 86: 4250–4256 (1995).
Cytogenet Genome Res 118:284–296 (2007)
Tricot G, Sawyer JR, Jagannath S, Desikan KR, Siegel D, et al: Unique role of cytogenetics in the prognosis of patients with myeloma receiving high-dose therapy and autotransplants. J Clin Oncol 15:2659–2666 (1997). UKCCG, United Kingdom Cancer Cytogenetics Group: Loss of the Y chromosome from normal and neoplastic bone marrows. Genes Chromosomes Cancer 5: 83–88 (1992). Vallespí T, Imbert M, Mecucci C, Preudhomme C, Fenaux P: Diagnosis, classification, and cytogenetics of myelodysplastic syndromes. Haematologica 83: 258–275 (1998). Vardiman JW, Harris NL, Brunning RD: The World Health Organization (WHO) classification of the myeloid neoplasms. Blood 100: 2292–2302 (2002). Wattel E, Lai JL, Hebbar M, Preudhomme C, Grahek D, et al: De novo myelodysplastic syndrome (MDS) with deletion of the long arm of chromosome 20: a subtype of MDS with distinct hematological and prognostic features? Leuk Res 17:921–926 (1993). Weh HJ, Gutensohn K, Selbach J, Kruse R, WackerBackhaus G, et al: Karyotype in multiple myeloma and plasma cell leukaemia. Eur J Cancer 29A:1269–1273 (1993). Wiestner A, Rosenwald A, Barry TS, Wright G, Davis RE, et al: ZAP-70 expression identifies a chronic lymphocytic leukemia subtype with unmutated immunoglobulin genes, inferior clinical outcome, and distinct gene expression profile. Blood 101:4944–4951 (2003). Zandecki M, Lai JL, Facon T: Multiple myeloma: almost all patients are cytogenetically abnormal. Br J Haematol 94: 217–227 (1996). Zojer N, Konigsberg R, Ackermann J, Fritz E, Dallinger S, et al: Deletion of 13q14 remains an independent adverse prognostic variable in multiple myeloma despite its frequent detection by interphase fluorescence in situ hybridization. Blood 95:1925–1930 (2000).
anemia (RA), RA with ringed sideroblasts (RARS), RA with excess blasts (RAEB), RAEB in transformation (RAEB-T) and chronic myelomonocytic leukemia (CMMoL). RA and RARS subgroups have more favorable prognosis than RAEB and RAEB-T. Although FAB classification roughly predicts prognosis of the patients at diagnosis, the classification is not sufficient as a marker for progression of the disease stages and patient prognosis. This is because the classification is based mainly on the burden of blast, not on its quality. Consequently, the addition of cytogenetic data, which could indirectly reflect the quality of blast, provided more reliable prognosis of patient outcome. The WHO introduced the new classification for MDS requiring cytogenetic investigation to properly evaluate the prognosis. Furthermore, the International Prognostic Scoring System (IPSS) subclassifies MDS into four groups with respect to survival expectations based on blast ratio, number of cytopenias and chromosomal abnormalities, which seem to be among the independent variables associated with prognosis of MDS (Sanz et al., 1998; Maes et al., 1999). Taken together, cytogenetic analysis is indispensable for evaluating prognosis of the MDS patients, and cytogenetic aberrations reflect the indirect characterization of blast. Cytogenetic alterations
More than 30% of patients with MDS have cytogenetic alterations and moreover, advanced MDS patients are prone to show complex abnormal chromosomes (Mauritzson et al., 2002). However, no cytogenetic abnormality specific to the FAB subtype is different from AML. This might show the multistep process of the transformation and progression of leukemia from MDS. 5q– The 5q– chromosomal abnormality is the most common in MDS (120% of MDS patients) (Van den Berghe et al., 1985). This chromosomal abnormality is predominantly included in the MDS-RA. The WHO puts forward this novel disease entity as the 5q– syndrome, which is associated with a favorable prognosis. At present, 5q31 is a probable region involved in MDS pathogenesis, which includes NPM (Nucleophosmin) and CTNNA1 (␣-catenin) as candidate genes (Horrigan et al., 2000; Look., 2005; Grisendi et al., 2006). Very recently, the FDA approved lenalidomide, a derivative of thalidomide, for 5q– syndrome patients with transfusion-dependent anemia (Giagounidis et al., 2005). Monosomy 7 Monosomy 7 is one of the frequent chromosomal aberrations. This abnormality is associated with aggressive disease progression and short survival (Le Beau et al., 1996; Fischer et al., 1997; Tosi et al., 1999). Several genes are located in chromosome 7q, such as T-cell receptor  (TRB@) and erythropoietin (EPO).
298
Cytogenet Genome Res 118:297–303 (2007)
20q– A chromosome 20q deletion is seen in about 5% of MDS patients. Isolated del(20q) confers a relatively favorable prognosis. Alternatively, the patients with del(20q) in conjunction with complex chromosomal abnormalities have unfavorable prognosis (Asimakopoulos et al., 1994; Kurtin et al., 1996). 12p– Deletion of 12p accounts for 5–10% of chromosomal abnormalities of patients with CMMoL, RAEB and RAEB-T. The TEL gene (ETV6), which is a putative tumor suppressor gene, is located in this region. Isolated del(12p) is involved in intermediate cytogenetic risk (Sole et al., 2000). Other cytogenetic aberrations In addition, a number of other cytogenetic aberrations have been reported including –Y, +8, –17q and translocations described below. The significance of –Y in the pathogenesis of MDS is still controversial. However, isolated –Y suggests a favorable prognosis for the MDS patient (Wiktor et al., 2000). The incidence of +8 is observed in about 10% of MDS patients and a male predominance accounts for two-thirds of patients with +8 (McClure et al., 1999). A variety of evidence indicates that this abnormality is involved in leukemic transformation and short survival (Pfeilstocker et al., 1999; de Souza Fernandez et al., 2000; Sole et al., 2000). Translocation Some rare chromosomal translocations have been described in patients with MDS and provide a useful resource in identifying genes involved in pathogenesis. t(5; 12)(q33; p13) is a rare chromosomal abnormality, which is detected in patients with CMMoL. This translocation of t(5; 12) (q33;p13) gives rise to a fusion between the platelet derived growth factor receptor  (PDGFRB) gene on chromosome 5 and ETV6 on chromosome 12. The ETV6-PDGFRB fusion is shown to be an oncogene, and it can be a molecular target for imatinib methylate, which is clinically used for the treatment of chronic myeloid leukemia, to inhibit its kinase activity (Golub et al., 1994). A t(3;5)(q25;q34) is also detectable in patients with MDS. The consequence of the translocation between the NPM gene on 3q25 and MLF1 on 5q34 may affect DNA replication and deregulate the cell cycle (Yoneda-Kato et al., 1996). Genetic alterations
AML1/RUNX1 gene A high incidence of point mutations was identified in the RUNX1 gene of patients with MDS, especially in RAEB, RAEB-T and MDS-AML (Harada et al., 2004). The RUNX1 gene encodes a transcription factor with the Runt homology domain, which provides a DNA binding activity. Missense mutations account for about 50% of all mutations, leading to diminished DNA binding potential. This suggests that
Table 1. Representative PcG genes
Drosophila (abbreviation) PcG complex 1 Polycomb (Pc) polyhomeotic (ph) Posterior sex comb (Psc) Sex comb on midleg (Scm) dRING1 (dRING1) PcG complex 2 Extra sex comb (esc) Enhancer of zeste (E(z)) Pleiohomeotic (pho) Suppressor of Zeste 12 (Su(z)12)
Mouse (human) M33/Cbx2 (HPC1/CBX2), MPc2/Cbx4 (HPC2/CBX4), Pc3/Cbx8 (HPC3/CBX8) rae28/mph1/edr1/Phc1 (RAE28/HPH1/PHC1), mph2/Phc2 (HPH2/PHC2), Phc3 (HPH3/PHC3) bmi1 (BMI1), mel18/Zfp144/ Pcgf2 (MEL18) Scmh1 (SCMH1), Scml2 (SCML2), (SCML1) Ring1A (RING1), Ring1B/Rnf2 (RING2/RNF2) eed (EED) Enx1/Ezh2 (ENX1/EZH2), Enx2/Ezh1 (ENX2/EZH1) YY1 (YY1) Suz12 (JJAZ1/SUZ12)
loss of the function of the RUNX1 gene is involved in the pathogenesis of MDS and leukemic transformation. FLT3 gene The FLT3 gene, encoding receptor type tyrosine kinase, is mainly associated with cellular proliferation. Somatic mutations were found in the internal tandem duplication of the FLT3 gene in about 5% of patients with MDS (Yokota et al., 1997; Au et al., 2004; Shih et al., 2004). These FLT3 mutations were frequently detected in the course of disease progression of MDS, suggesting that it plays an important role in leukemic transformation. N-RAS gene N-Ras is a signal transducer mediated in response to cytokines. An N-Ras gene mutation is most frequently described among NRAS genes in 10–15% of MDS patients, especially in those with poor prognosis (Hirai et at., 1988; Paquette et al., 1993; de Souza Fernandez et al., 1998; Shih et al., 2004). The point mutations of NRAS genes were most commonly identified at codon 12 N-Ras in MDS (Hirai et al., 1987). It is widely thought that this gene is associated with transformation to leukemia. Epigenetic alterations
The emergence of MDS is further associated with epigenetic alterations, which involve DNA methylation and chromatin regulation. The DNA methylation of the promoter lesion of CDKN2B (alias p15/INK4B), the cyclin-dependent kinase inhibitor gene, may be involved in the occurrence or progression to leukemia (Kamb et al., 1994; Nobori et al., 1994). The transcriptional inactivation of CDKN2B by DNA methylation is observed in 30–50% of MDS patients, especially in high risk MDS compared with low risk MDS, suggesting that CDKN2B plays a critical role in the disease progression of MDS. A demethylating agent, 5-aza-2ⴕ deoxycytidine (decitabine) has been recently emphasized. The response rate of decitabine is about 50% in
high risk MDS (Wijermans et al., 2000) and CDKN2B was shown to be one of the most important target genes of the demethylation agent. On the other hand, BMI1, belonging to PcG, was discovered to be an excellent molecular marker for reflecting disease progression and prognosis of MDS (Mihara et al., 2006). Thus, we highlight PcG genes and overview their molecular roles and possible pathological involvement to MDS. PcG genes were originally identified in Drosophila mutants with posterior homeotic transformations, which result from ectopic HOMC gene expression in the mutants. Spatial regulation of HOMC genes in these mutants is correctly initiated, but is adversely affected several hours later, demonstrating that PcG is required, not for its initial establishment, but for the maintenance of repression (Takihara and Hara, 2000). PcG displays direct antagonistic functions on chromatin to trithorax-group (trxG) genes, which create either a repressed or activated chromatin state. The mammalian homologue of trxG, MLL (All-1/HRX/Htrx1), was discovered by virtue of its involvement in childhood acute leukemia (Takihara and Hara, 2000). Members of the trxG genes, trxlike, brm, and Mll, are directly involved in chromatin regulation. Recent molecular analysis showed that the SET domain in Mll encodes histone methyltransferase targeting histone H3 lysine 4 (H3K4) (Milne et al., 2002; Nakamura et al., 2002). H3K4 methylated by Mll is recognized by WDR5 to further recruit Mll and to expand transcriptionally active chromatin domain regions (Dou et al., 2005; Wysocka et al., 2005). The histone code is presumed to be recognized by the hitherto unidentified effector molecules to maintain transcriptionally active chromatin states. The effector molecules may be involved in regulation of chromatin remodeling and may represent the molecular rationale underlying the trxG function antagonistic to PcG genes (Takihara and Hara, 2000). There is estimated to be more than 30–40 PcG genes in flies, and 20 of these have been characterized. PcG genes are conserved. The representative PcG genes and their mammalian homologues are summarized in Table 1 (Takihara
Cytogenet Genome Res 118:297–303 (2007)
299
chromatin
s AC D H YY1 eed Ezh 2
DNA PRE : CXGCCATXXXXGX
(PcG complex 2) DNA methylation
(PcG complex 1) rae28
M
CpG
: SET domain
histone methylation methylated H3
: chromodomain
Ring1A,B M33
Scmh1
DN
Ts
methylated H3
Recognition of methylated H3 through chromodomain
bmi1, mel18
Recruitment of PcG complex 1
SWI/SNF
inhibition of chromatin remodeling
Ring1B
H2A
ubiquitination
transcription silencing
Fig. 1. Model for PcG complex 1 and 2 mediated gene silencing.
and Hara, 2000). Biochemical studies showed that PcG members form complexes in the chromatin through domains that are conserved in their vertebrate counterparts. Since PcG complexes maintain the chromatin states once they have been heritably determined by early events in the cellular descendants, these complexes are known to establish a cell memory system through epigenetic mechanisms (Takihara and Hara, 2000). Although the molecular nature underlying the cell memory system has long been unclear, recent molecular characterizations of PcG complexes provide a clue to unveil the enigma. Two distinct PcG complexes, PcG complex 1 and 2, have been identified. PcG complex 1 and 2 are mammalian homologues of Drosophila Polycomb repressive complex 1 and 2, respectively. Based on molecular functions of Drosophila PcG complexes, molecular mechanisms underlying PcG complex-mediated transcriptional repression in mammals are presumed as follows; PcG complex 2 recruits histone deacetylases and plays a role in the initiation of transcriptional silencing (van der Vlag and Otte, 1999; Tie et al., 2001). Evolutionarily conserved SET domain of Ezh2, a component of PcG complex 2, methylates histone H3 lysine 9 and 27, providing a histone code for silencing transcription in the chromatin region (Cao et al., 2002; Czermin et al., 2002; Muller et al., 2002). PcG complex 2 including distinct isoforms of EED, which was also designated as PcG
300
Cytogenet Genome Res 118:297–303 (2007)
repressive complex 3, was shown to methylate histone H1 lysine 26, although its role in transcriptional regulation remains to be elucidated (Kuzmichev et al., 2004). PcG complex 1 is recruited to the chromatin domain by recognizing methylated histone H3 lysine 27 through chromodomain in the complex (Fischle et al., 2003; Min et al., 2003). Biochemical evidence indicates that PcG complex 1 inhibits chromatin remodeling activity of SWI/SNF (Shao et al., 1999). PcG complex 1 was further demonstrated to have ubiquitin E3 ligase activity for ubiquitination of histone H2A lysine 119 (Wang et al., 2004; Cao et al., 2005). Ubiquitination of histone H2A was found to be involved in silencing HOX gene expression. Since PcG complex 2 interacts with DNA methyltransferases through Ezh2, these complexes may also silence the chromatin domain through CpG methylation in DNA (Vire et al., 2006). Based on these findings, a molecular model underlying PcG-mediated gene silencing is shown in Fig. 1. Recent genome-wide location analysis of PcG complexes 1, 2 and methylated histone at lysine 27 in embryonic stem (ES) cells clarified the potential Polycomb target genes in the genome, indicating that PcG complexes maintain ES cell pluripotency and plasticity during embryonic development through the control of major developmental regulators including transcription factors with cell fate specification (Boyer et al., 2006; Lee et al., 2006).
(PcG complex 1) Scmh1 Ring1A,B
rae28
M33
bmi1 Stem cells
p19 ARF
P53
?
ink4a Fig. 2. Bmi-1 maintains stem cell function through the repression of the Ink4A locus inducing cellular senescence.
The expression of PcG genes in hematopoietic cells has been characterized. CBX2, PCGF2, BMI1 and PHC1 show high expression levels in primitive hematopoietic cells, and this expression is reduced during hematopoietic cell differentiation. On the other hand, Enx1/Ezh2 and Eed are expressed in both primitive and mature murine hematopoietic cells (Iwama et al., 2004). Although the number of myeloid and lymphoid cells and their progenitors undergoes almost normal expansion in early embryonic stages of Bmi1- and Rae28-deficient mice, it is progressively reduced in these cells with age. Furthermore, the number and proliferative activity of bone marrow primitive myeloid and lymphoid cells progressively decrease with age (van der Lugt et al., 1994; Ohta et al., 2002). Detailed characterization of HSCs in these mutant mice clearly demonstrated impaired HSC functions including the self-renewal activity (Ohta et al., 2002; Kim et al., 2004). Knockout mice deficient in PcG genes have further highlighted a crucial role of PcG in cellular senescence and stemness. Neural stem cells, neural crest cells, cerebellar progenitors, embryonic stem cells as well as hematopoietic stem cells are defective in Bmi1, Rae28 and Ezh2-deficient mice, indicating that PcG genes are essential for sustaining stemness in adult stem cells as well as in embryonic stem cells (O’Carroll et al., 2001; Ohta et al., 2002; Molofsky et al., 2003, 2005; Park et al., 2003). Murine embryonic fibroblasts from mice deficient in Bmi1, Pcgf2 and Cbx2 undergo premature senescence due to early transcription induction of the Ink4A locus, encoding two kinds of proteins, p16INK4A and p19ARF which mediate apoptosis through regulation of the tumor suppressor p53 (Jacobs et al., 1999; Core et al., 2004). Early induction of the Ink4A locus and resultant impairment of hematopoietic and neural stem cells was demonstrated, indicating that PcG genes sustain stem cell function through repression of the Ink4A locus (Fig. 2), although it remains unclear whether repression of the Ink4A locus by PcG genes is a direct effect or not. Furthermore, genetic deletion of the Ink4A locus did not completely complement the stem cell activity in neural stem and neural crest cells (Molofsky et al.,
Cellular senescence p16 CDK inhibitor
2003, 2005; Bruggeman et al., 2005). Overexpression of Bmi1 augmented the activity of HSCs even from Ink4A-deficient mice (Iwama et al., 2004). These findings suggest PcG sustains stem cells further through a molecular role other than the repression of the Ink4A locus. Recently PcG complex 1 was shown to interact with Geminin, an inhibitor of DNA replication, and may provide a clue for a novel molecular role of PcG in stem cells (Luo and Kessel, 2004). Evidence further shows that bmi1 is also essential for sustaining the activity of leukemic stem cells (Lessard and Sauvageau, 2003). Thus PcG genes may play a crucial role in sustaining cancer stem cells as well as physiological stem cells. PcG as a novel biomarker for prognosis and progression of MDS
The International Prognostic Scoring System (IPSS) has achieved international acceptance and assigns scores according to marrow blast cell ratio, karyotype and cytopenias to estimate prognosis in patients with MDS. However, the IPSS score is not necessarily enough to evaluate the patient for the purpose of choosing intervention therapies including antineoplastic drugs and cytokines. Recently, we reported expression of BMI1 in CD34+ cells as a novel biomarker for the disease progression and prognosis of patients with MDS (Mihara et al., 2006). We stress the importance of PcG in stem cells and the usefulness of BMI1 as a novel biomarker for clinically monitoring the disease. A higher positivity rate of BMI1 in CD34+ cells was preferentially seen in RAEB, RAEBT and MDS-AML compared with RA and RARS. The IPSS score was positively correlated with the percentage of BMI1 expression. RA and RARS patients with a higher percentage of BMI1+ cells showed disease progression to RAEB. The increased BMI1 expression could reflect a qualitative difference of CD34+ cells. The detection and follow-up of this marker may provide important information about when to initiate chemotherapy and/or stem cell transplantation.
Cytogenet Genome Res 118:297–303 (2007)
301
Perspectives
As discussed above, although a variety of genetic alterations have been reported to be involved in molecular pathogenesis of MDS, recent findings indicate the involvement of epigenetic alterations. Since the epigenetic alterations may be reversible, molecular dissection of epigenetic involvement in MDS is expected to provide a new therapeutic strategy such as Decitabine, which has attracted attention for treatment of patients with MDS. The expression level of BMI1 was proven to be sensitively correlated with disease progression and prognosis of patients with MDS. Although a molecular role of increased expression of BMI1 in disease
progression of MDS remains to be elucidated, there is emerging evidence indicating that the PcG complexes are indispensable for sustaining stem cell activity and cancer stem cells through repression of the Ink4A locus. And recent genome-wide analyses showed that major transcriptional regulators governing development are under the regulation of PcG complexes. Thus PcG not only provides a molecular marker for monitoring disease progression of MDS, but also provides a clue for elucidating a molecular mechanism underlying the disease progression, which may help in the development of a new therapeutic strategy against MDS.
References Asimakopoulos FA, White NJ, Nacheva E, Green AR: Molecular analysis of chromosome 20q deletions associated with myeloproliferative disorders and myelodysplastic syndromes. Blood 84:3086–3094 (1994). Au WY, Fung AT, Ma ES, Liang RH, Kwong YL: Low frequency of FLT3 gene internal tandem duplication and activating loop mutation in therapyrelated acute myelocytic leukemia and myelodysplastic syndrome. Cancer Genet Cytogenet 149:169–172 (2004). Boyer LA, Plath K, Zeitlinger J, Brambrink T, Medeiros LA, et al: Polycomb complexes repress developmental regulators in murine embryonic stem cells. Nature 441:349–353 (2006). Bruggeman SWM, Valk-Lingbeek ME, van der Stoop PPM, Jacobs JJL, Kieboom K, et al: Ink4A and Arf differentially affect cell proliferation and neural stem cell self-renewal in Bmi1-deficient mice. Genes Dev 19:1438–1443 (2005). Cao R, Wang L, Wang H, Xia L, Erdjument-Bromage H, et al: Role of histone H3 lysine 27 methylation in Polycomb-group silencing. Science 298:1039–1043 (2002). Cao R, Tsukada Y, Zhang Y: Role of Bmi-1 and Ring1A in H2A ubiquitylation and Hox gene silencing. Mol Cell 20: 845–854 (2005). Core N, Joly F, Boned A, Djabali M: Disruption of E2F signaling suppresses the Ink4A-induced proliferative defect in M33-deficient mice. Oncogene 23: 7660–7668 (2004). Czermin B, Melfi R, McCabe D, Seitz V, Imhof A, Pirrotta V: Drosophila enhancer of Zeste/ESC complexes have a histone H3 methyltransferase activity that marks chromosomal Polycomb sites. Cell 111:185–196 (2002). de Souza Fernandez T, Menezes de Souza J, Macedo Silva ML, Tabak D, Abdelhay E: Correlation of N-ras point mutations with specific chromosomal abnormalities in primary myelodysplastic syndrome. Leuk Res 22: 125–134 (1998). de Souza Fernandez T, Ornellas MH, Otero de Carvalho L, Tabak D, Abdelhay E: Chromosomal alterations associated with evolution from myelodysplastic syndrome to acute myeloid leukemia. Leuk Res 24:839–848 (2000). Dou Y, Moilne TA, Tackett AJ, Smith ER, Fukuda A, et al: Physical association and coordinate function of the H3 K4 methyltransferase MlLL1 and H4 K16 acetyltransferase MOF. Cell 121: 873–885 (2005). Fischer K, Frohling S, Scherer SW, McAllister Brown J, Scholl C, et al: Molecular cytogenetic delineation of deletions and translocations involving chromosome band 7q22 in myeloid leukemias. Blood 89:2036–2041 (1997).
302
Fischle W, Wang Y, Jacobs SA, Kim Y, Allis CD, Khorasanizadeh S: Molecular basis for the discrimination of repressive methyl-lysine marks in histone H3 by polycomb and HP1 chromodomains. Genes Dev 17: 1870–1881 (2003). Giagounidis AA, Germing U, Strupp C, Hildebrandt B, Heinsch M, Aul C: Prognosis of patients with del(5q) MDS and complex karyotype and the possible role of lenalidomide in this patient subgroup. Ann Hematol 84: 569– 571 (2005). Golub TR, Barker GF, Lovett M, Gilliland DG: Fusion of PDGF receptor beta to a novel ets-like gene, tel, in chronic myelomonocytic leukemia with t(5; 12) chromosomal translocation. Cell 77:307–316 (1994). Grisendi S, Mecucci C, Falini B, Pandolfi PP: Nucleophosmin and cancer. Nat Rev Cancer 6: 493–505 (2006) Harada H, Harada Y, Niimi H, Kyo T, Kimura A, Inaba T: High incidence of somatic mutations in the AML1/RUNX1 gene in myelodysplastic syndrome and low blast percentage myeloid leukemia with myelodysplasia. Blood 103: 2316–2324 (2004). Heaney ML, Golde DW: Myelodysplasia. N Engl J Med 340:1649–1660 (1999). Hirai H, Kobayashi Y, Mano H, Hagiwara K, Maru Y, et al: A point mutation at codon 13 of the Nras oncogene in myelodysplastic syndrome. Nature 327: 430–432 (1987). Hirai H, Okada M, Mizoguchi H, Mano H, Kobayashi Y, et al: Relationship between an activated N-ras oncogene and chromosomal abnormality during leukemic progression from myelodysplastic syndrome. Blood 71: 256–258 (1988). Horrigan SK, Arbieva ZH, Xie HY, Kravarusic J, Fulton NC, et al: Delineation of a minimal interval and identification of 9 candidates for a tumor suppressor gene in malignant myeloid disorders on 5q31. Blood 95:2372–2377 (2000). Iwama A, Oguro H, Negishi M, Kato Y, Morita Y, et al: Enhanced self-renewal of hematopoietic stem cells mediated by the Polycomb gene product Bmi-1. Immunity 21: 843–851 (2004). Jacobs JJ, Kieboom K, Marino S, DePinho RA, van Lohuizen M: The oncogene and Polycombgroup gene bmi-1 regulates cell proliferation and senescence through the Ink4A locus. Nature 397:164–168 (1999). Kamb A, Gruis NA, Weaver-Feldhaus J, Liu Q, Harshman K, et al: A cell cycle regulator potentially involved in genesis of many tumor types. Science 264:436–440 (1994).
Cytogenet Genome Res 118:297–303 (2007)
Kim JY, Sawada A, Tokimasa S, Endo H, Ozono K, et al: Defective long-term repopulating ability in hematopoietic stem cells lacking the Polycomb-group gene rae28. Eur J Hematol 73: 75– 84 (2004). Kurtin PJ, Dewald GW, Shields DJ, Hanson CA: Hematologic disorders associated with deletions of chromosome 20q: a clinicopathologic study of 107 patients. Am J Clin Pathol 106: 680–688 (1996). Kuzmichev A, Jenuwein T, Tempst P, Reinberg D: Different EZH2-containing complexes target mehylation of histone H1 or nucleosomal histone H3. Mol Cell 14: 183–193 (2004). Le Beau MM, Espinosa R III, Davis EM, Eisenbart JD, Larson RA, Green ED: Cytogenetic and molecular delineation of a region of chromosome 7 commonly deleted in malignant myeloid diseases. Blood 88:1930–1935 (1996). Lee TI, Jenner RG, Boyer LA, Guenther MG, Levine SS, et al: Control of developmental regulators by Polycomb in human embryonic stem cells. Cell 125:301–313 (2006). Lessard J, Sauvageau G: Bmi-1 determines the proliferative capacity of normal and leukemic stem cells. Nature 423: 255–260 (2003). Look AT: Molecular Pathogenesis of MDS. Hematology Am Soc Hematol Educ Program 156–160 (2005). Lowenthal RM, Marsden KA: Myelodysplastic syndromes. Int J Hematol 65: 319–338 (1997). Luo L, Kessel M: Geminin coordinates cell cycle and developmental control. Cell Cycle 3: 711– 714 (2004). Maes B, Meeus P, Michaux L, Bijnens L, Boogaerts M, et al: Application of the International Prognostic Scoring System for myelodysplastic syndromes. Ann Oncol 10: 825–829 (1999). Mauritzson N, Albin M, Rylander L, Billstrom R, Ahlgren T, et al: Pooled analysis of clinical and cytogenetic features in treatment-related and de novo adult acute myeloid leukemia and myelodysplastic syndromes based on a consecutive series of 761 patients analyzed 1976–1993 and on 5098 unselected cases reported in the literature 1974–2001. Leukemia 16:2366–2378 (2002). McClure RF, Dewald GW, Hoyer JD, Hanson CA: Isolated isochromosome 17q: a distinct type of mixed myeloproliferative disorder/myelodysplastic syndrome with an aggressive clinical course. Br J Haematol 106: 445–454 (1999). Mihara K, Chowdhury M, Nakaju N, Hidani S, Ihara A, et al: Bmi-1 is useful as a novel molecular marker for predicting progression of myelodysplastic syndrome and patient prognosis. Blood 107:305–308 (2006).
Milne TA, Briggs SD, Brock HW, Martin ME, Gibbs D, Hess JL: MLL targets SET domain methyltransferase activity to Hox gene promoters. Mol Cell 10:1107–1117 (2002). Min J, Zhang Y, Xu RM: Structure basis for specific binding of Polycomb chromodomain to histone H3 methylated at Lys 27. Genes Dev 17: 1823– 1828 (2003). Molofsky AV, Pardal R, Iwashita T, Park I, Clarke MF, Morrison SJ: Bmi-1 dependence distinguishes neural stem cell self-renewal from progenitor proliferation. Nature 425: 962–967 (2003). Molofsky AV, He S, Bydon M, Morrison SJ, Pardal R: Bmi-1 promotes neural stem cell self-renewal and neural development but not mouse growth and survival by repressing the p16Ink4A and p19Arf senescence pathways. Genes Dev 19:432–1437 (2005). Muller J, Hart CM, Francis NJ, Vargas ML, Sengupta A, et al: Histone methyltransferase activity of a Drosophila Polycomb group repressor complex. Cell 111: 197–208 (2002). Nakamura T, Mori T, Tada S, Krajewski W, Rozovskaia T, et al: ALL-1 is a histone methyltransferase that assembles a supercomplex of proteins involved in transcriptional regulation. Mol Cell 10:1119–1128 (2002). Nobori T, Miura K, Wu DJ, Lois A, Takabayashi K, Carson DA: Deletions of the cyclin-dependent kinase-4 inhibitor gene in multiple human cancers. Nature 368: 753–756 (1994). O’Carroll D, Erhardt S, Pagani M, Barton SC, Surani MA, Jenuwein T: The polycomb-group gene Ezh2 is required for early mouse development. Mol Cell Biol 21: 4330–4336 (2001). Ohta H, Sawada A, Kim J-Y, Tokimasa S, Nishiguchi S, et al: Polycomb-group gene rae28 is required for sustaining activity of hematopoietic stem cells. J Exp Med 195:759–770 (2002). Paquette RL, Landaw EM, Pierre RV, Kahan J, Lubbert M, et al: N-ras mutations are associated with poor prognosis and increased risk of leukemia in myelodysplastic syndrome. Blood 82: 590–599 (1993).
Park I, Qian D, Kiel M, Becker MW, Pihalja M, et al: Bmi-1 is required for maintenance of adult selfrenewing haematopoietic stem cells. Nature 423:302–305 (2003). Pfeilstocker M, Reisner R, Nosslinger T, Gruner H, Nowotny H, et al: Cross-validation of prognostic scores in myelodysplastic syndromes on 386 patients from a single institution confirms importance of cytogenetics. Br J Haematol 106: 455–463 (1999). Sanz GF, Sanz MA, Greenberg PL: Prognostic factors and scoring systems in myelodysplastic syndromes. Haematologica 83: 358–368 (1998). Shao Z, Raible F, Mollaaghababa R, Guyon JR, Wu CT, et al: Stabilization of chromatin structure by PRC1, a Polycomb complex. Cell 98: 37–46 (1999). Shih LY, Huang CF, Wang PN, Wu JH, Lin TL, et al: Acquisition of FLT3 or N-ras mutations is frequently associated with progression of myelodysplastic syndrome to acute myeloid leukemia. Leukemia 18:466–475 (2004). Sole F, Espinet B, Sanz GF, Cervera J, Calasanz MJ, et al: Incidence, characterization and prognostic significance of chromosomal abnormalities in 640 patients with primary myelodysplastic syndromes. Br J Haematol 108: 346–356 (2000). Takihara Y, Hara J: The Polycomb-group genes and hematopoiesis. Int J Hematol 72: 5–172 (2000). Tie F, Furuyama T, Prasad-Sinha J, Jane E, Harte PJ: The Drosophila Polycomb group proteins ESC and E(Z) are present in a complex containing the histone-binding protein p55 and the histone deacetylase RPD3. Development 128: 275– 286 (2001). Tosi S, Scherer SW, Giudici G, Czepulkowski B, Biondi A, Kearney L: Delineation of multiple deleted regions in 7q in myeloid disorders. Genes Chromosomes Cancer 25: 384–392 (1999). Van den Berghe H, Vermaelen K, Mecucci C, Barbieri D, Tricot G: The 5q-anomaly. Cancer Genet Cytogenet 17: 189–255 (1985).
van der Lugt MT, Domen J, Linders K, Roon M, Robanus-Maandag E, et al: Posterior transformation, neurological abnormalities, and severe hematopoietic defects in mice with a targeted deletion of the bmi-1 proto-oncogene. Genes Dev 8:757–769 (1994). van der Vlag J, Otte AP: Transcriptional repression mediated by the human polycomb-group protein EED involves histone deacetylation. Nat Genet 23:474–478 (1999). Vire E, Brenner C, Deplus R, Blanchon L, Fraga M, et al: The Polycomb group protein EZH2 directly controls DNA methylation. Nature 439: 871–874 (2006). Wang H, Wang L, Erdjument-Bromage H, Vidal M, Tempst P, et al: Role of histone H2A ubiquitination in Polycomb silencing. Nature 431: 873– 878 (2004). Wijermans P, Lubbert M, Verhoef G, Bosly A, Ravoet C, et al: Low-dose 5-aza-2ⴕ-deoxycytidine, a DNA hypomethylating agent, for the treatment of high-risk myelodysplastic syndrome: a multicenter phase II study in elderly patients. J Clin Oncol 18:956–962 (2000). Wiktor A, Rybicki BA, Piao ZS, Shurafa M, Barthel B, et al: Clinical significance of Y chromosome loss in hematologic disease. Genes Chromosomes Cancer 27: 11–16 (2000). Wysocka J, Swigut T, Milne TA, Dou Y, Zhang X, et al: WDR5 associates with histone H3 methylated at K4 and is essential for H3 K4 methylation and vertebrate development. Cell 121:859– 872 (2005). Yokota S, Kiyoi H, Nakao M, Iwai T, Misawa S, et al: Internal tandem duplication of the FLT3 gene is preferentially seen in acute myeloid leukemia and myelodysplastic syndrome among various hematological malignancies. A study on a large series of patients and cell lines. Leukemia 11: 1605–1609 (1997). Yoneda-Kato N, Look AT, Kirstein MN, Valentine MB, Raimondi SC, et al: The t(3;5)(q25.1;q34) of myelodysplastic syndrome and acute myeloid leukemia produces a novel fusion gene, NPMMLF1. Oncogene 12: 265–275 (1996).
Cytogenet Genome Res 118:297–303 (2007)
303
primary balanced chromosome aberration (approximately 20% of all AML cases, and some therapy-related MDS), (2) cases with normal karyotypes and (3) cases with unbalanced karyotype abnormalities, which are characterized by gains and/or losses of usually large regions of the genome and no known primary balanced abnormality (35–40% of AML, 50% of the novo MDS, and more than 80% of the therapy related MDS/AML) (Grimwade et al., 2001, 2004; Schoch et al., 2005). Focusing in these last two groups of patients (normal and unbalanced karyotypes), this review will show a brief summary of what is known about conventional CGH in myeloid disorders and will try to present the current state of the arrayCGH advances in this type of leukemia. Myelodysplastic syndromes and acute myeloid leukemia under the light of chromosome based CGH (cCGH)
In addition to providing insights into the molecular pathophysiology of myeloid disorders, cytogenetic analysis has been the main provider of the prognostic information that influences therapy and outcome of AML and MDS (Alvarez et al., 2001). Still, individuals displaying a normal karyotype in chromosome-banding analysis represent 40– 45% of MDS and AML patients, precluding any advance that may derive from classical cytogenetics in terms of molecular mechanisms and/or prognosis. Lately, several molecular defects, such as length mutations of the FLT3 gene (Bullinger and Valk, 2005) and NPM genes (Dohner et al., 2005; Falini et al., 2005), have been described in AML with normal karyotypes. On the other hand, myeloid disorders with unbalanced karyotype abnormalities constitute a subtype that has been classified mostly based on its cytogenetic/molecular genetic profile, although these patients may truly represent a distinct biological entity (Schoch et al., 2005). The characterization and the understanding of specific roles of these rearrangements have dramatically improved through the application of a spectrum of cytogenetic and molecular diagnostic techniques. These techniques include multicolor karyotyping, conventional comparative genomic hybridization (CGH), loss of heterozygosity analysis, CGH arrays, and expression arrays. Taking into account only published series that included more than ten patients studied with cCGH, nearly 120 MDS and 260 cases of AML have been reported (Bentz et al., 1995; El-Rifai et al., 1997; Huhta et al., 1999; Wilkens et al., 1999; Castuma et al., 2000; Heller et al., 2000; Kim et al., 2001; Lindvall et al., 2001; Dalley et al., 2002; Casas et al., 2004; Babicz et al., 2005; Karst et al., 2005; Martinez-Ramirez et al., 2005). Compared with data obtained with classic cytogenetic methods, cCGH analysis yielded a higher number of well defined structural changes. This increased rate of chromosome aberrations was nurtured by the fact that CGH is a DNA based technique so those cases where cell culture was unsuccessful (no metaphase chromosomes could be analyzed) were however suitable for the comparative hybrid-
ization analysis. On the other hand, another well known source of discrepancy comes from the analysis of different clonal cell populations that may be present in the tumor samples. This fact remains as a problem for CGH, while classic cytogenetics has the advantage of allowing the detection and characterization of such clonal populations. cCGH studies of MDS with unbalanced karyotype, previously detected by conventional cytogenetics, established that genomic losses are the most prevalent findings, involving 5q in around 75% of cases, 7q in 30%, and 20q in 24%. Most common gains involved total or partial chromosome 8 (about 40%), 11q (35%) and 4q (27%) (Wilkens et al., 1999; Castuma et al., 2000; Martinez-Ramirez et al., 2005). Amplifications were also detected and confirmed in one third of the patients involving MLL and RUNX1, among others (Martinez-Ramirez et al., 2005). AML has been more extensively studied (Gebhart, 2005). With nearly 300 cases, including individual reports, the main conclusions confirmed that genomic losses affected 5q and 7q (with high frequency), followed by 1p, 3pq2, 12p, 13q, and 17p. Frequent gains involved chromosomes 6, 8q and 21q (with high frequency), followed by 3q26]q27 and 22q. Amplifications were not common but were demonstrated for well known genes as MYC and MLL (Kim et al., 2001; Casas et al., 2004), as well as for some other rare locations on 13q without a clear candidate gene (Heller et al., 2000). ArrayCGH approach: development and platforms
The sensitivity of CGH technique is dependent on several factors. The main intrinsic factors are the degree of condensation of the chromosomes and power of resolution of the microscope hardware and software. Extrinsically, the sensitivity is also limited by the size of the chromosomal aberration (if it is large enough to be visualized under the microscope) as well as by the minimal portion of a cell population carrying a specific change that needs to be present in the test sample to be detected. Adding up all these limitations, it is widely accepted that copy number changes (deletions, gains or amplifications) that affect genomic regions smaller than 3–5 Mb of size whose presence in a cell population is under 30% are not readable or efficiently detected by chromosome based CGH (Gebhart et al., 2000; Gebhart, 2005). The change from chromosome to genome based approaches has had an immediate effect also in chromosome CGH. Metaphase spreads are being replaced as target for hybridization by microarrays. Therefore, one of the limitations mentioned above, the power of resolution of the chromosomal CGH, has been easily overcome by substituting chromosomes with small fragments of DNA arrayed onto a solid support e.g. large insert clones (BAC/PAC clones), cDNA clones, or oligonucleotides (Solinas-Toldo et al., 1997; Pinkel et al., 1998; Pollack et al., 1999; Lucito et al., 2000; Mei et al., 2000). Resolution is now limited by the type, amount, and distribution through the genome of the clones
Cytogenet Genome Res 118:304–309 (2007)
305
that are included in the array. Typically, most studies utilize microarrays comprising large insert clones (100 kb) spaced at approximately one clone per Mb (Fiegler et al., 2003), but higher resolution arrays comprising overlapping clone sets from specific regions are being employed. There are published experiments with array platforms that include from a few hundred clones to over 30,000 clones covering the complete human genome (Ishkanian et al., 2004). The panel of platforms can be initially divided into three groups regarding the type and size of clones arrayed: large BAC/ PAC clones of a mean size of 150 kb, cDNA clones with sizes of around hundreds of bp, and oligonucleotides with sizes around 40 to 70 bp. Each type of platform has its own advantages and limitations. In recent years we have observed in the literature and in scientific forums a competition among users and providers of different platforms in order to demonstrate which one was the most powerful or efficient, or both, to detect DNA gains and losses. BAC based arrays are the most commonly used because of their robustness and good signal in hybridization experiments. Because of this good hybridization yield, they are especially useful in the detection of genomic losses (deletions). This type of platform has been used in six reported works dealing with myeloid disorders (Martinez-Ramirez et al., 2003, 2005; Tchinda et al., 2004; Paulsson et al., 2006; Rucker et al., 2006a, b). cDNA arrays to be used as CGH arrays is another option. In this type of platform the clones are expressed sequences that are been obtained from a previously defined library. cDNA clones have been extensively used for expression profiles and they are not a first choice for studying copy number changes because of their poor hybridization proficiency. They are mostly indicated for in certain experiments such as localization of genes that are simultaneously overexpressed and amplified. However, the cDNA approach is clearly not recommended if the aim is the detection of low copy number changes (simple gains or deletions) because the available bioinformatics (that includes the running average of multiple clones, typically five to ten, along the genome) often fails to provide reproducible data (Alvarez and Cigudosa, 2005). Finally, oligonucleotide based CGH arrays are the most recent approach. It implies the use of short sequences of newly synthesized fragments of DNA (oligonucleotides) of 40 or 60 base pair length (40-mer or 60-mer oligos) as targets for hybridization on the slides. This type of array, characterized by the high density of clones that may be easily arrayed (it is rather normal that they contain 40,000 clones), have been used in six reports dealing with myeloid disorders (Fitzgibbon et al., 2005; Gorletta et al., 2005; Raghavan et al., 2005; Tyybakinoja et al., 2006, 2007; Suela et al., 2007). Myeloid leukemias and arrayCGH
It is reasonable to expect that genomics may give different approaches to unveil leukemogenesis, aggressiveness and other clinical factors in AML and MDS. In fact, some
306
Cytogenet Genome Res 118:304–309 (2007)
genomic approaches have been already taken to explain leukemia, mostly in the form of expression arrays (see Bullinger and Valk, 2005; Valk et al., 2005; Mano, 2006 for recent reviews). However, it is somehow striking that, while a large number of arrayCGH studies have been reported in solid tumors (mostly breast), prenatal disorders or other diseases, there are few articles yet published about copy number aberrations on myeloid disorders using this approach. Only ten works have been published so far in which arrayCGH has been applied to AML or MDS samples. We initiated the research in this field first confirming the suitability of the arrayCGH for studying copy number aberrations in MDS (Martinez-Ramirez et al., 2003). Some published reports have chosen the arrayCGH approach in a small number of cases in order to provide a refined characterization of specific chromosome aberrations previously detected using other methods. Tchinda et al. (2004) reported two cases of de novo AML with translocations involving the breakpoint 6p22 first detected at relapse. By using arrayCGH they found that translocations were truly unbalanced showing gains and losses at the breakpoints. Tyybäkinoja et al. (2006) reported another two cases of AML characterized by the presence of amplicons in the 11q23]q25 cytobands. They described the limits and physical genomic structure of the amplicon, concluding that, unexpectedly, MLL was not affected by this genetic event. They proposed a common region of amplification which included 14 known genes. As mentioned, an important advantage of the high resolution of arrayCGH is the possibility to detect DNA copy number changes that may be simply invisible to conventional karyotyping. Along this line, an interesting study was recently published to disclose such hidden aberrations in ten cases of AML/MDS where trisomy 8 was the only detected chromosome abnormality (Paulsson et al., 2006). By using a tiling Path BAC array that includes 32,000 human clones they showed that new undetected CNA could be observed in all analyzed cases. However, due to the design of the platform, that includes in its coverage polymorphic noncoding genomic regions, it was indicated that most of these aberrations could represent genetic polymorphisms. After removing these predicted or known polymorphisms, they observed that four out of ten cases showed new CNA, therefore suggesting that trisomy 8 alone may not be the unique genetic event that is needed for the onset of this kind of myeloid leukemia. Regarding unbalanced and complex karyotypes in myeloid disorders, two more articles, using a combination of 1 Mb BAC arrayCGH and expression profiling on a panel of 17 cell lines (Rucker et al., 2006b) and a large series of patients (Rucker et al., 2006a), have been published. The series of primary samples included 60 cases with newly diagnosed complex karyotype leukemias (de novo AML, post MDS and therapy related leukemias). Both the recurrence and the nature of genomic changes were precisely delineated in this work. Genomic losses were more frequent than gains. While the recurrences of 5q and 7q deletions were 77% and 45%, similar to that observed with conventional CGH in a similar
group of patients (Martinez-Ramirez et al., 2005), the incidence of losses at 17p (55%), and those at 18q, 16q, and 17q (35% each) were higher than previous reports. Additionally, a large set of smaller recurrent aberrations, including other losses and gains of 11q and 8q, were observed. Of interest they described the presence of 41 high level DNA amplifications with a noticeable recurrence of those located at 11q23.3]q24 (seven cases), 21q22 (six cases), 11q23.3 (five cases). These findings have been useful to point to potential candidate genes in several regions. Although global expression profiling was performed in the same set of samples, no significant gene expression signatures correlated with major arrayCGH aberrations as losses of 5q, 17p, and 7q, or gains of 11q and 8q. With all collected evidence of DNA copy number changes and expression studies, it is becoming accepted that myeloid leukemias with complex karyotypes constitute a distinct genetic and probably biological entity (Alvarez and Cigudosa, 2005; Schoch et al., 2005). However, the extent and nature of CNA in myeloid leukemia samples with normal karyotype at diagnosis still remain to be clarified. Conventional CGH has not been successful in finding recurrent aberrations in this type of sample (Dalley et al., 2002; Casas et al., 2004), and we will probably see in the near future the results of currently ongoing research dealing with this issue. Our group and others are actively working on this matter (Suela et al., 2007; Tyybakinoja et al., 2007). CNA are frequent in AML but our findings show that, except for those seen in complex karyotypes, recurrent aberrations are not a common feature (some aberrations as seen by arrayCGH that were detected in NK cases are represented in Table 1 and Fig. 1). Myeloid leukemias and new genomics strategies: array-SNP
Single Nucleotide Polymorphisms (SNP) can be also arrayed and used for studying simultaneously variations in the DNA copy number content and the presence of genomic regions with loss of heterozygosity (LOH). Although the
Fig. 1. ArrayCGH representative ideograms of deletions in a case of AML with normal karyotype. Human Genome CGH 44k microarrays (Agilent Technologies Inc., Palo Alto, CA), versions A and B, were used. Hybridizations were done as previously described according to the manufacturer’s protocols. Graphics are modified output from CGHAnalyitics v3.2.1.(also from Agilent Technologies Inc.). Moving average (0.5 Mb) log2 ratio values along the chromosome are represented by a blue line. Balanced ratio (no copy number changes) falls around the 0! vertical lines. Displacement of this blue line to the left or right represents genomic losses or gains, respectively. Aberrations that were statistically significant are represented as blue dots on the left (deletion) or right (gain) of the moving average. Features of the regions are highlighted and zoomed in to a gene view on the right of the chromosome view. Gene view displays oligo probes as dots. The color of each dot represents losses (green), normal (black) or gains (red). Large (6p24, size 10 Mb) and small deletions (2q13, size 115 kb) are represented.
Cytogenet Genome Res 118:304–309 (2007)
307
Table 1. Detection of new copy number aberrations in normal karyotype cases of AML using arrayCGH platform
Case
Karyotype
Array-CGH
1 2 3 4 5
46,XX 46,XX 46,XX 46,XX 46,XY
46,XX,del(2)(q13),del(6)(p24) 46,XX,del(X)(p32.21) 46,XX,del(2)(q23.3) 46,XX,del(1)(p36),del(2)(p24),del(3)(q27q28),del(7)(q31),+18q21 46,XY,del(2)(q12),del(4)(q24),del(5)(q35)
use of array-SNP for detecting CNA may be redundant with other platforms, their ability to detect LOH regions without net changes in DNA content (diploid LOH) represents a substantial change in genomic research. Diploid LOH is thought to arise in somatic cells through the genetic process known as uniparental disomy (UPD), and it was found in nearly 20% of AML cases with normal karyotype studied by a 10K array-SNP platform (Raghavan et al., 2005). Mutation analysis of several genes included in specific diploid LOH regions observed by this group demonstrated that those regions may include the presence of homozygous mutations in cancer related genes, like CEBPA, FLT3 or RUNX1 (Fitzgibbon et al., 2005). Additionally, a LOH study of a panel of 30 myeloid leukemias with normal karyotype confirmed these previous results (Gorletta et al., 2005). They proposed the existence of two types of LOH: terminal, involving large genomic regions (30–90 Mb), produced by a UPD phenomenon and interstitial, affecting smaller regions (2–8 Mb) usually caused by a microdeletion event (resulting in loss of one DNA copy or haploid LOH). They also found that 10% of the normal karyotype cases showed interstitial microdeletions, similar to our results (Table 1) obtained with the arrayCGH platform.
Conclusion
In conclusion, arrayCGH, confirming previous data based on chromosome based CGH, provides useful information regarding the nature of genomic aberrations that take place in cases with complex karyotypes. The high resolution and the use of several platforms on complex karyotype cases have produced an enormous amount of data with interesting candidate genes whose involvement has to be explored. These genetic aberrations are probably related to late stages of myeloid disease, transformation of MDS to AML, or exposure to environmental- or therapy-related clastogenic agents. However, the role of CNA in myeloid leukemias with normal karyotypes or single numerical changes remains unclear, mostly due to the lack of recurrence of the so far observed aberrations. Obviously more genomic data have to be collected in as many patients as possible. It is of major interest to direct the genomic research of AML and MDS towards other emerging approaches. The role and extent of UPD, the large-scale detection of mutations on key proliferation-related genes, and the epigenetics profiling of this disease are the next unavoidable steps which need to be taken.
References Alvarez S, Cigudosa JC: Gains, losses and complex karyotypes in myeloid disorders: A light at the end of the tunnel. Hematol Oncol 23: 18–25 (2005). Alvarez S, MacGrogan D, Calasanz MJ, Nimer SD, Jhanwar SC: Frequent gain of chromosome 19 in megakaryoblastic leukemias detected by comparative genomic hybridization. Genes Chromosomes Cancer 32: 285–293 (2001). Babicz M, Kowalczyk JR, Winnicka D, Gaworczyk A, Lejman M, et al: The effectiveness of highresolution-comparative genomic hybridization in detecting the most common chromosomal abnormalities in pediatric myelodysplastic syndromes. Cancer Genet Cytogenet 158: 49– 54 (2005). Bentz M, Dohner H, Huck K, Schutz B, Ganser A, et al: Comparative genomic hybridization in the investigation of myeloid leukemias. Genes Chromosomes Cancer 12: 193–200 (1995). Bullinger L, Valk PJ: Gene expression profiling in acute myeloid leukemia. J Clin Oncol 23:6296– 6305 (2005).
308
Casas S, Aventin A, Fuentes F, Vallespi T, Granada I, et al: Genetic diagnosis by comparative genomic hybridization in adult de novo acute myelocytic leukemia. Cancer Genet Cytogenet 153:16–25 (2004). Castuma MV, Rao PH, Acevedo SH, Larripa IB: Comparative genomic hybridization study of de novo myeloid neoplasia. Acta Haematol 104: 25–30 (2000). Dalley CD, Neat MJ, Foot NJ, Burridge M, Byrne L, et al: Comparative genomic hybridization and multiplex-fluorescence in situ hybridization: An appraisal in elderly patients with acute myelogenous leukemia. Hematol J 3: 290–298 (2002). Dohner K, Schlenk RF, Habdank M, Scholl C, Rucker FG, et al: Mutant nucleophosmin (NPM1) predicts favorable prognosis in younger adults with acute myeloid leukemia and normal cytogenetics: Interaction with other gene mutations. Blood 106:3740–3746 (2005). El-Rifai W, Elonen E, Larramendy M, Ruutu T, Knuutila S: Chromosomal breakpoints and changes in DNA copy number in refractory acute myeloid leukemia. Leukemia 11: 958–963 (1997).
Cytogenet Genome Res 118:304–309 (2007)
Falini B, Mecucci C, Tiacci E, Alcalay M, Rosati R, et al: Cytoplasmic nucleophosmin in acute myelogenous leukemia with a normal karyotype. N Engl J Med 352:254–266 (2005). Fiegler H, Carr P, Douglas EJ, Burford DC, Hunt S, et al: DNA microarrays for comparative genomic hybridization based on DOP-PCR amplification of BAC and PAC clones. Genes Chromosomes Cancer 36: 361–374 (2003). Fitzgibbon J, Smith LL, Raghavan M, Smith ML, Debernardi S, et al: Association between acquired uniparental disomy and homozygous gene mutation in acute myeloid leukemias. Cancer Res 65: 9152–9154 (2005). Gebhart E: Genomic imbalances in human leukemia and lymphoma detected by comparative genomic hybridization (review). Int J Oncol 27: 593–606 (2005). Gebhart E, Verdorfer I, Saul W, Trautmann U, Brecevic L: Delimiting the use of comparative genomic hybridization in human myeloid neoplastic disorders. Int J Oncol 16: 1099–1105 (2000).
Gorletta TA, Gasparini P, D’Elios MM, Trubia M, Pelicci PG, Di Fiore PP: Frequent loss of heterozygosity without loss of genetic material in acute myeloid leukemia with a normal karyotype. Genes Chromosomes Cancer 44: 334–337 (2005). Grimwade D, Walker H, Harrison G, Oliver F, Chatters S, et al: The predictive value of hierarchical cytogenetic classification in older adults with acute myeloid leukemia (AML): Analysis of 1065 patients entered into the United Kingdom Medical Research Council AML11 trial. Blood 98:1312–1320 (2001). Grimwade D, Moorman A, Hills R, Wheatley K, Walker H, et al: Impact of karyotype on treatment outcome in acute myeloid leukemia. Ann Hematol 83 Suppl 1:S45–48 (2004). Harris NL, Jaffe ES, Diebold J, Flandrin G, MullerHermelink HK, et al: The World Health Organization classification of hematological malignancies report of the clinical advisory committee meeting, Airlie House, Virginia, November 1997. Mod Pathol 13: 193–207 (2000). Heller A, Chudoba I, Bleck C, Senger G, Claussen U, Liehr T: Microdissection based comparative genomic hybridization analysis (micro-CGH) of secondary acute myelogenous leukemias. Int J Oncol 16:461–468 (2000). Huhta T, Vettenranta K, Heinonen K, Kanerva J, Larramendy ML, et al: Comparative genomic hybridization and conventional cytogenetic analyses in childhood acute myeloid leukemia. Leuk Lymphoma 35: 311–315 (1999). Ishkanian AS, Malloff CA, Watson SK, DeLeeuw RJ, Chi B, et al: A tiling resolution DNA microarray with complete coverage of the human genome. Nat Genet 36: 299–303 (2004). Kallioniemi A, Kallioniemi OP, Sudar D, Rutovitz D, Gray JW, et al: Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science 258: 818–821 (1992). Kallioniemi OP, Kallioniemi A, Piper J, Isola J, Waldman FM, et al: Optimizing comparative genomic hybridization for analysis of DNA sequence copy number changes in solid tumors. Genes Chromosomes Cancer 10: 231–243 (1994). Karst C, Heller A, Claussen U, Gebhart E, Liehr T: Detection of cryptic chromosomal aberrations in the in vitro non-proliferating cells of acute myeloid leukemia. Int J Oncol 27: 355–359 (2005).
Kim MH, Stewart J, Devlin C, Kim YT, Boyd E, Connor M: The application of comparative genomic hybridization as an additional tool in the chromosome analysis of acute myeloid leukemia and myelodysplastic syndromes. Cancer Genet Cytogenet 126:26–33 (2001). Lindvall C, Nordenskjold M, Porwit A, Bjorkholm M, Blennow E: Molecular cytogenetic characterization of acute myeloid leukemia and myelodysplastic syndromes with multiple chromosome rearrangements. Haematologica 86: 1158–1164 (2001). Lucito R, West J, Reiner A, Alexander J, Esposito D, et al: Detecting gene copy number fluctuations in tumor cells by microarray analysis of genomic representations. Genome Res 10: 1726–1736 (2000). Mano H: DNA micro-array analysis of myelodysplastic syndrome. Leuk Lymphoma 47: 9–14 (2006). Martinez-Ramirez A, Urioste M, Calasanz MJ, Cigudosa JC, Benitez J: Array comparative genomic hybridization analysis of myelodysplastic syndromes with complex karyotypes. A technical evaluation. Cancer Genet Cytogenet 144:87–89 (2003). Martinez-Ramirez A, Urioste M, Melchor L, Blesa D, Valle L, et al: Analysis of myelodysplastic syndromes with complex karyotypes by highresolution comparative genomic hybridization and subtelomeric CGH array. Genes Chromosomes Cancer 42: 287–298 (2005). Mei R, Galipeau PC, Prass C, Berno A, Ghandour G, et al: Genome-wide detection of allelic imbalance using human SNPs and high-density DNA arrays. Genome Res 10: 1126–1137 (2000). Paulsson K, Heidenblad M, Strombeck B, Staaf J, Jonsson G, et al: High-resolution genome-wide array-based comparative genome hybridization reveals cryptic chromosome changes in AML and MDS cases with trisomy 8 as the sole cytogenetic aberration. Leukemia 20: 840–846 (2006). Pinkel D, Segraves R, Sudar D, Clark S, Poole I, et al: High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays. Nat Genet 20: 207–211 (1998). Pollack JR, Perou CM, Alizadeh AA, Eisen MB, Pergamenschikov A, et al: Genome-wide analysis of DNA copy-number changes using cDNA microarrays. Nat Genet 23: 41–46 (1999). Raghavan M, Lillington DM, Skoulakis S, Debernardi S, Chaplin T, et al: Genome-wide single nucleotide polymorphism analysis reveals frequent partial uniparental disomy due to somatic recombination in acute myeloid leukemias. Cancer Res 65: 375–378 (2005).
Rucker FG, Bullinger L, Schwaenen C, Lipka DB, Wessendorf S, et al: Disclosure of candidate genes in acute myeloid leukemia with complex karyotypes using microarray-based molecular characterization. J Clin Oncol 24:3887–3894 (2006a). Rucker FG, Sander S, Dohner K, Dohner H, Pollack JR, Bullinger L: Molecular profiling reveals myeloid leukemia cell lines to be faithful model systems characterized by distinct genomic aberrations. Leukemia 20:994–1001 (2006b). Schoch C, Kern W, Kohlmann A, Hiddemann W, Schnittger S, Haferlach T: Acute myeloid leukemia with a complex aberrant karyotype is a distinct biological entity characterized by genomic imbalances and a specific gene expression profile. Genes Chromosomes Cancer 43: 227– 238 (2005). Solinas-Toldo S, Lampel S, Stilgenbauer S, Nickolenko J, Benner A, et al: Matrix-based comparative genomic hybridization: Biochips to screen for genomic imbalances. Genes Chromosomes Cancer 20:399–407 (1997). Suela J, Alvarez S, Cifuentes F, Largo C, Ferreira BI, et al: DNA profiling analysis of 100 consecutive de novo acute myeloid leukemia cases reveals patterns of genomic instability that affect all cytogenetic risk groups. Leukemia 21: 12241231 (2007). Tchinda J, Dijkhuizen T, Vlies Pv P, Kok K, Horst J: Translocations involving 6p22 in acute myeloid leukaemia at relapse: Breakpoint characterization using microarray-based comparative genomic hybridization. Br J Haematol 126: 495– 500 (2004). Tyybakinoja A, Saarinen-Pihkala U, Elonen E, Knuutila S: Amplified, lost, and fused genes in 11q23]q25 amplicon in acute myeloid leukemia, an array-CGH study. Genes Chromosomes Cancer 45:257–264 (2006). Tyybakinoja A, Elonen E, Piippo K, Porkka K, Knuutila S: Oligonucleotide array-CGH reveals cryptic gene copy number alterations in karyotypically normal acute myeloid leukemia: Leukemia 21:571–574 (2007). Valk PJ, Delwel R, Lowenberg B: Gene expression profiling in acute myeloid leukemia. Curr Opin Hematol 12:76–81 (2005). Wilkens L, Burkhardt D, Tchinda J, Busche G, Werner M, et al: Cytogenetic aberrations in myelodysplastic syndrome detected by comparative genomic hybridization and fluorescence in situ hybridization. Diagn Mol Pathol 8: 47–53 (1999).
Cytogenet Genome Res 118:304–309 (2007)
309
Table 1. Cytogenetic abnormalities and potential candidate genes in B-CLL
Genetic abnormality
Candidate genes
13q14
ARL11; C13orf1; DLEU1; DLEU2; RCBTB1; SETDB2; TRIM13 Non-coding RNAs BCMS; BSMCUN microRNAs mir-15a; mir-16-1 ATM; ARHGAP20; BTG4; POU2AF1; FDX1; RDX TP53 CDK4; CDKN1B; CLLU1; MDM2 None described
11q22–q23 17p13 Trisomy 12 6q
in 54% of patients and multiple rearrangements occur in up to 29% (Table 1) (Dohner et al., 2000). The most frequent 13q14 deletion is found in 55% of patients, particularly as the sole genetic abnormality, and confers a more favourable prognosis (Dohner et al., 2000). Typically it occurs in patients with highly stable and indolent disease often requiring no treatment (Guarini et al., 2003). In contrast, high-risk genetic defects are defined by the presence of an 11q or 17p deletion, both being independent markers of poor prognosis (Krober et al., 2002) often in association with specific clinical features. 11q deletions occur in up to 20% of patients and are predominantly associated with extensive lymphadenopathy and advanced disease (Dohner et al., 1997; Neilson et al., 1997; Dickinson et al., 2006a). In addition, 11q deletions are almost invariably associated with unmutated immunoglobulin VH genes (Krober et al., 2002), itself an independent poor prognostic marker. About 7% of patients have deletions involving the tumor suppressor gene TP53 at 17p and this is the strongest predictor of poor survival (el Rouby et al., 1993; Geisler et al., 1997; Cordone et al., 1998; Oscier et al., 2002) together with a failure of disease response to either alkylating agents (el Rouby et al., 1993) or fludarabine (Dohner et al., 1995). There is evidence emerging to justify risk adapted therapy in patients with high-risk-genetic features (Lozanski et al., 2004; Caballero et al., 2005; Byrd et al., 2006, 2007; Dreger et al., 2007) and ongoing prospective trials are addressing this. Trisomy 12 occurs in approximately 16% of patients with B-CLL (Dohner et al., 2000), but its effect on prognosis has been controversial because of the association with atypical morphology (Knauf et al., 1995; Matutes et al., 1996; Su’ut et al., 1998). In a study of chromosomal abnormalities in atypical CLL, trisomy 12 was found to be the commonest abnormality (Bigoni et al., 1997; Criel et al., 1997). However, patients with B-CLL and trisomy 12 are reported to have shorter survival times (Dohner et al., 2000). Deletions at 6q are less common and occur in 6% of patients (Dohner et al., 2000). Patients with this karyotype tend to have higher white blood cell counts and more extensive lymphadenopathy at presentation (Stilgenbauer et al., 1999; Cuneo et al., 2004). The prognostic significance associated with the 6q abnormality is unclear and studies assessing risk have generally had only small patient numbers. The largely higher stage of disease at presentation with the 6q deletion may be considered an indicator of poorer outcome itself. It is most
likely that the prognostic significance lies somewhere between no adverse risk (Stilgenbauer et al., 1999) to intermediate-risk disease (Cuneo et al., 2004). Relevance of B-CLL genetics to pathogenesis
Mechanisms of gene inactivation Classic tumor suppressor genetics. The existence of tumor suppressor genes was first suggested by Dr Alfred Knudson in 1971 and led to the ‘two-hit’ model of tumorigenesis (Knudson, 1971; Comings, 1973). From his analysis of patients with retinoblastoma, Knudson first proposed that as few as one somatic event (the ‘second hit’) in addition to the germ-line mutation (the ‘first hit’) is required to induce tumors in patients with the hereditary form of retinoblastoma (Knudson, 1971) and the non-hereditary form requires two somatic events to occur in the same cell. The Knudson model combines the concept of dominant inheritance of a susceptibility gene with a recessive mechanism of cancer development at the cellular level. The cellular event is recessive because both copies of the susceptibility allele need to be ‘knocked out’ for the phenotype to be expressed. In the case of retinoblastoma, the gene suppressed cell cycle progression and, when removed from the retinal cells, permitted tumor formation by the process of uncontrolled cell proliferation. The concept arose that the normal function of these susceptibility alleles was likely to be of ‘cancer preventing genes’ thus predicting the existence of tumor suppressor genes. It later emerged that normal cells could suppress the malignant phenotype when mixed with cancer cells of the same species in fusion experiments (Stanbridge et al., 1982) concluding that ‘suppressor’ genes were defective in cancerous cells. This became the prototype in cancer genetics where loss of function of both copies of a tumor suppressor gene was required to induce a cancer phenotype. While the Knudson two-hit theory has provided a useful framework for interpreting the kinetics of tumor suppressor gene inactivation, it may be considered an oversimplification. Cancer is a complex, multi-gene process involving both oncogenes and tumor suppressors. Individual tumor suppressor genes may follow the two event paradigm described by Knudson. Haploinsufficient tumor suppressor model. Haploinsufficiency describes a process where a single functional allele of a gene is insufficient to accomplish the normal activity of that gene product in the cell. A change in gene dosage is
Cytogenet Genome Res 118:310–319 (2007)
311
produced by loss of one allele without mutation in the second remaining allele. Many genes are pleiotropic and function in multiple cellular processes. It may be considered that an individual gene can differ in its function according to the level of expression (ie. the ‘dose’). TP53 is a good example where low levels of expression lead to cell cycle arrest, while high levels induce apoptosis. Although not all genes have this dual dose dependent response, those very sensitive to gene dosage levels may have potential for carcinogenesis if expressed at lower levels, in a similar manner to the Knudson retinoblastoma model. Moreover, gene dosage-dependence of a tumor suppressor can vary not only with the genetic background of an organism but also between tissues, adding further complexity. One of the first examples of a haploinsufficient tumor suppressor gene was the cyclin-dependent kinase inhibitor CDKN1B (p27/Kip1) where loss of a single allele in a CDKN1B knock-out murine model demonstrated a strong selective advantage for tumor development (Fero et al., 1998). Since these first descriptions of haploinsufficient tumor suppressor genes in 1998, there have been numerous publications involving other genes and a tumor phenotype (reviewed in Payne and Kemp, 2005). This concept is becoming an increasingly attractive pathomechanism for the deletions of B-CLL (Mertens et al., 2002; Haslinger et al., 2004; Kienle et al., 2005; Dickinson et al., 2006b), particularly in light of the failure by multiple groups to detect mutations in the retained allele on the normal chromosome of patients with a single chromosome loss (Mabuchi et al., 2001; Migliazza et al., 2001; Wolf et al., 2001). Thus, B-CLL can develop without needing to fulfil the 2-hit model for tumor suppressor genes. Epigenetic suppression. The phenomenon of epigenetic supression involves alteration in the level of gene expression without a change in nucleotide sequence. Epigenetic mechanisms principally include DNA methylation and a variety of histone modifications, the best characterized of which is acetylation. DNA hypermethylation and histone hypoacetylation are hallmarks of gene silencing. In contrast, DNA hypomethylation and acetylated histones promote active transcription. Epigenetic alterations are reversible, opening the way for therapeutic manipulation. One of the most frequently observed epigenetic means of gene silencing in malignancy is that of methylation. Whilst DNA methylation is globally decreased, regional hypermethylation of gene promoters leads to gene silencing. The role of DNA methylation in inactivating tumor suppressor genes in cancer has been widely studied. Hypermethylation of the promoter region has been identified in more than half the genes causing familial forms of cancer (reviewed in Baylin et al., 2001). Methylation of these genes is associated with lack of gene transcription in the absence of coding region mutations (Herman et al., 1995; Merlo et al., 1995). Such alterations are also observed in non-familial cancers, and have been frequently described in hemopoietic neoplasms (reviewed in Baylin et al., 1998). The epigenetic mechanisms affecting gene expression are, however, complex and inter-related. DNA promoter methylation will also impact on the degree of histone acetylation and methylation (Bachman et al., 2003).
312
Cytogenet Genome Res 118:310–319 (2007)
Chromosomal defects in B-CLL and implicated genes
13q14 deletions i) Minimal deleted region. A consensus region of deletion with an approximate size of 600 kb between the chromosomal locations D13S273 and D13S25 has been described (Kitamura et al., 2000; Bullrich et al., 2001; Mabuchi et al., 2001; Migliazza et al., 2001; Wolf et al., 2001), although various reports describe smaller internal regions based largely on observations from one or two cases (Corcoran et al., 1998; Kalachikov et al., 1997; Liu et al., 1997; Stilgenbauer et al., 1998). Various candidate gene(s) have been identified from extensive sequencing analysis of the minimal regions of deletion and include TRIM13 (CAR/LEU5/DLEU5/RFP2), DLEU1 and 2 (Liu et al., 1997; Migliazza et al., 2001), C13orf1 (CLLD6), RCBTB1 (CLLD7) and SETDB2 (CLLD8) (Mabuchi et al., 2001), yet studies have consistently failed to demonstrate the presence of mutations in any of the genes, even in the unmutated remaining allele of patients with hemizygous 13q14 deletions (Liu et al., 1997; Mabuchi et al., 2001; Migliazza et al., 2001). No alteration in gene expression has been found for TRIM13 and DLEU1, even in patients with monoallelic 13q14 loss, whereas DLEU2 has no detectable expression in B-CLL samples regardless of 13q14 status (Migliazza et al., 2001). An alternative splice variant of the DLEU2 gene, ALT1, is ubiquitously expressed but remains unaltered in leukemia cases (Bullrich et al., 2001). BCMS/ DLEU1 and BCMSUN/DLEU2/RFP2OS are non-coding RNAs and span the entire deletion (Wolf et al., 2001; Baranova et al., 2003). Little is currently known about the role of such non-coding RNAs, both in normal cell function and in malignancy, but considerable interest is being generated currently by functional analysis of these non-coding regions (see below). Although multiple splicing of BCMS/ DLEU1 occurs with tissue specific expression of the splice variants, no variant is specific for B-CLL and no mutations have been detected (Wolf et al., 2001). ARL11 (ARLTS1), a member of the Ras superfamily involved in apoptotic signalling, resides within this deleted chromosomal region and has been proposed as a tumor-suppressor gene (Calin et al., 2005). The polymorphism G446A was suggested to represent a susceptibility allele for familial cancer, including B-CLL. However, these findings have not been confirmed in a large series including familial cases (Sellick et al., 2006). Over expression of AKT1 has been associated with the 13q deletion and may promote a survival advantage in clones with this genetic abnormality (Kienle et al., 2005). The absence of pathogenetic mutations within candidate genes from analysis of patients with monoallelic loss at 13q14 suggests that either haploinsufficiency or epigenetic suppression of the retained allele is the favoured pathomechanism and there are recently published data to support these hypotheses for all the genetically altered regions in B-CLL (Mertens et al., 2002, 2006; Haslinger et al., 2004; Kienle et al., 2005; Dickinson et al., 2006b). Studies have consistently revealed deregulation of genes from the critical genomic regions of deletion, implicating gene dosage in molecular
pathogenesis (Mertens et al., 2002; Haslinger et al., 2004; Kienle et al., 2005). Although the majority of the genes within the 13q14 region are down-regulated (Mertens et al., 2002; Dickinson et al., 2006b), transcriptional silencing through methylation has been excluded as one of the potential mechanisms (Mertens et al., 2002). Mertens et al. (2006) have recently identified asynchronous replication as an epigenetic mechanism of monoallelic inactivation at 13q14. They investigated whether the candidate genes at the 13q14 tumor suppressor locus are monoallelically expressed and epigenetic features of the critical region. In their model they demonstrated asynchronous replication of the 13q14 critical region and suggest that this results in differential chromatin packaging of the two copies, with subsequent monoallelic expression of specific genes at 13q14, evident in the B- and T-cells of healthy control individuals. This epigenetic silencing of one copy could be sufficient to completely abolish gene function in B-CLL patients with monoallelic 13q deletion and the active gene copy was lost from the Bcells of such patients in this study. These results would also explain the almost complete down regulation of candidate 13q14 genes and microRNAs observed in CLL tumors (Calin et al., 2002; Mertens et al., 2002) rather than down-regulation by a factor of 2 which would be expected if the mechanism was purely one of gene dosage. ii) microRNAs. MicroRNAs (miRNAs), non-coding RNAs with the ability to ‘silence genes’, are increasingly being implicated in the pathogenesis of B-CLL and many cancers. miRNAs are short evolutionarily conserved non-protein-coding sequences of between 19 and 23 nucleotides that control gene expression at the post-transcriptional level, by degrading or repressing target messenger RNA through sequence specific base pairing with their targets (Miska, 2005). They are found in intergenic regions as well as introns, constitute about 1–5% of the predicted genes in worms, mice and humans (Chen, 2005) and have a role in the control of a wide range of biological processes. Vertebrates generate miRNAs from a primary transcript (primiRNA) of at least a few hundred base pairs through sequential processing by nucleases of the ribonuclease III family (reviewed in Miska, 2005). Cleavage of the primiRNA by the enzymes Drosha and Pasha produces a stemloop precursor (pre-miRNA) and further processing by the enzyme Dicer generates small RNA duplexes, the mature miRNA. One strand of the processed duplex is incorporated into a silencing complex and guided to target mRNA sequences with cleavage and/or inhibition of translation, thus ‘silencing’ the gene. Each miRNA has the potential to regulate a large number of target genes. Evidence is emerging that miRNAs can act as oncogenes or tumor suppressors (He et al., 2005; O’Donnell et al., 2005). The mir-17 complex at 13q31]q32 is highly expressed in B-cell lymphomas and when co-expressed with MYC can act in concert to accelerate tumor development in a mouse B-cell lymphoma model (He et al., 2005). The transcription factor E2F1 is a target of MYC and if the mir-17 cluster is up-regulated by MYC, E2F1 is negatively regulated by this miRNA cluster (O’Donnell et al., 2005). In B-CLL,
two miRNAs at 13q14, mir-15a and mir-16-1, are down-regulated (Calin et al., 2002) to remove the suppressor influence on the BCL2 gene (Cimmino et al., 2005). In addition, a miRNA signature distinguishes B-CLL cells from normal B-cells (Calin et al., 2004) and is associated with prognosis and disease progression (Calin et al., 2005). Further evidence implicating miRNAs in B-CLL pathogenesis occurs with the regulation of the oncogene TCL1A. This gene was discovered as the target of rearrangements at 14q31.2 in mature T-cell leukemias and is preferentially expressed early in T- and B-lymphocyte differentiation (Virgilio et al., 1994). The TCL1A gene is over-expressed in B-CLL and correlates with unmutated VH genes, ZAP70 positivity (Herling et al., 2006) and 11q deletions (Pekarsky et al., 2006). A causal role in disease pathogenesis has been confirmed by the development of tumors resembling B-CLL in TCL1 transgenic mice (Bichi et al., 2002). Further data demonstrate that TCL1A expression in B-CLL is regulated by two miRNAs, mir-29 and mir-181 (Pekarsky et al., 2006). Co-expression of TCL1A with mir-29 and mir-181 significantly reduced TCL1A expression. Another miRNA, mir-155, is over-expressed in BCLL (Calin et al., 2004) as well as diffuse large B-cell lymphoma (Eis et al., 2005). Transgenic mice carrying a mir-155 transgene, with expression targeted to B-cells, develop a preleukemic pre-B cell proliferation in the spleen and bone marrow followed by frank B-cell malignancy (Costinean et al., 2006). Little is known yet about the individual roles of most mammalian miRNAs and for the majority function remains unknown (Chen et al., 2004). Based on the high degree of miRNA sequence conservation, miRNAs are thought to have essential functions during development and are now known to regulate mammalian hemopoiesis. Three miRNAs (mir-181, mir-223, mir-142s) have been found to be specifically expressed in hemopoietic cells and are involved in the control of lineage differentiation. Ectopic expression of mir-181 in murine bone marrow cells leads to B-lymphoid proliferation in vitro and in vivo (Chen et al., 2004). Thus, further validation of target genes of miRNAs in BCLL is important and could reveal new targets for therapy. 11q22]q23 deletions i) Minimal deleted region. The region of deletion at 11q22]q23 is much larger and less well defined than that of the 13q14 deletion. A 2–3 Mb region has been reported and is covered by three yeast artificial chromosome (YAC) clones, y801e11, y975h6 and y755b11, containing the ATM, RDX and FDX1 genes, all of which are potential candidate tumor suppressors (Stilgenbauer et al., 1996). Work from two other groups suggests a slightly more telomeric but overlapping region that does not include ATM (Zhu et al., 1999; Auer et al., 2001). Various groups have analyzed the ATM gene and described mutations in patients with B-CLL, together with loss of protein expression (Starostik et al., 1998; Bullrich et al., 1999; Schaffner et al., 1999; Stankovic et al., 1999; Austen et al., 2005). However, not all patients with deletions of 11q have evidence of an ATM mutation (and vice versa) (Bullrich et al., 1999; Schaffner et al., 1999;
Cytogenet Genome Res 118:310–319 (2007)
313
Austen et al., 2005) and mutations are only found in up to 20% of patients with B-CLL (Schaffner et al., 1999; Stankovic et al., 1999; Austen et al., 2005). This suggests that although ATM is likely to be important in the pathogenesis of B-CLL, it is unlikely to be the sole cause of the 11q abnormality. Other tumor suppressor genes are likely to exist within the minimal region of deletion. B-CLL cells bearing 11q deletions have altered expression profiles some of which are pathway specific and may aid identification of further candidate genes (Aalto et al., 2001; Haslinger et al., 2004; Dickinson et al., 2005). B-CLL cells with an 11q22]q23 deletion, for example, exhibit over-expression of genes associated with cell signalling for which the serine/threonine kinase CDC2 is an illustration (Dickinson et al., 2005). The cloning of two rare translocation breakpoints within the minimal region of deletion from patients with B-CLL has identified additional candidate genes (Auer et al., 2005; Kalla et al., 2005). POU2AF1 is a B-cell-specific transcriptional coactivator and is differentially expressed in B-CLL cells compared to normal B lymphocytes. BTG4 is a member of the BTG family of negative regulators of the cell cycle, but mutations were not detected in patients with monoallelic 11q deletions (Auer et al., 2005). Similarly Kalla et al. identified the ARHGAP20 gene as a result of a translocation breakpoint. Again no mutations could be found. ARHGAP20, encoding a protein predicted to be involved in the regulation of Rho family GTPases, was, however, significantly up-regulated in CLL B-cells (Kalla et al., 2005). ii) ATM and B-CLL. The ataxia telangiectasia mutated (ATM) protein is the principal activator of the p53 protein in the response to DNA double-strand breaks. Recently, ATM mutations have been shown to have independent prognostic information on multivariate analysis, irrespective of 11q23 status (Austen et al., 2005). The majority of patients with ATM mutations are refractory to DNA damaging chemotherapeutic drugs. In patients with bulky lymphadenopathy ATM is consistently under-expressed, including those without 11q deletion (Joshi et al., 2007). Mutant ATM CLL tumor samples generally lack somatic VH hypermutation (Stankovic et al., 2002; Austen et al., 2005) suggesting a common pathogenesis for these tumours. In the same sample population, TP53 mutations, on the whole, are only present in tumors without an ATM mutation or loss of ATM protein (Stankovic et al., 2002). How ATM contributes to the development of B-CLL is not yet fully understood. ATM mutations in B-CLL are mainly missense substitutions outside the kinase domain distributed in different areas of the gene (Bullrich et al., 1999; Schaffner et al., 1999; Stankovic et al., 1999) and result in impaired in vitro DNA damage responses. There is no clustering of mutations along the protein to suggest an association between a particular domain function inactivation and B-CLL leukemogenesis. ATM inactivation is thought to be an early event in B-CLL pathogenesis; mutations are generally present at diagnosis (Austen et al., 2005) and experimental data have demonstrated mutations in a progenitor cell at a stage prior to lymphoid commitment (Stankovic et al., 2002). It can be speculated that because the
314
Cytogenet Genome Res 118:310–319 (2007)
ATM protein is involved in mediating the p53 response during cell cycle arrest and apoptosis, abnormal function or loss of this protein may render the B-CLL cells resistant to apoptosis, a characteristic feature of this disease. Studies have analysed p53 and ATM function in B-CLL to address this question. ATM mutant tumors exhibit a deficient ATMdependent-p53 response to ionising radiation (IR). These tumors are defective in TRAIL-R2 up-regulation, a downstream target that links irradiation-induced p53 response with apoptosis, and explains why neither ATM nor TP53 mutant B-CLLs can undergo apoptosis after exposure to gamma rays (Stankovic et al., 2002). Inactivation of ATM has been shown to cause p53 dysfunction in CLL via a mechanism independent of TP53 mutation (Pettitt et al., 2001). Tumors with p53 dysfunction (measured by impairment of p21 up-regulation) caused by mutations were completely resistant to IR-induced apoptosis. Other B-CLL tumors, without TP53 mutations, also failed to show up-regulation of p53 and apoptosis in response to IR and were found to be partially resistant to IR-induced killing but, in contrast, were found to have mutations of ATM and a reduction in ATM protein (Pettitt et al., 2001). With respect to genetic predisposition to B-CLL and mutations in the ATM gene, it remains difficult to identify a clear connection. ATM homozygotes develop various malignancies, particularly leukemias and lymphomas (reviewed in Taylor et al., 1996). ATM heterozygotes are already known to be at an increased risk of breast cancer, although the risk for other cancers remains less clear (Taylor, 1992). Germline mutations have been detected in a handful of patients with B-CLL suggesting a link between ATM heterozygosity and genetic predisposition to B-CLL (Bullrich et al., 1999; Stankovic et al., 1999). In both studies the carrier frequency was significantly higher amongst B-CLL patients than that of the general population. Of the mutations detected, some had been previously described in families with ataxia telangiectasia (A-T) from the British Isles (Stankovic et al., 1998) suggesting a link between a previously known germline ATM mutation and development of sporadic B-CLL. However, there are descriptions of these germline mutations as rare ATM polymorphisms rather than a cancer predisposing allele (Vorechovsky et al., 1996) and no differences in the frequency of this allele were observed between controls and cancer patients (Vorechovsky et al., 1999). Linkage analysis with markers around the ATM locus of patients with familial B-CLL has failed to demonstrate a role for ATM in the development of the familial form (Bevan et al., 1999) and mutations are not increased in families with the disease (Yuille et al., 2002). A slightly increased number of patients with B-CLL among relatives of A-T individuals has been described by some (Morrell et al., 1990) but not others (Geoffroy-Perez et al., 2001; Olsen et al., 2001). Most recently, single nucleotide polymorphisms within the ATM-BRCA2-CHEK2 DNA damage-response axis have been associated with an increased likelihood of developing the disease (Rudd et al., 2006). In general, however, there remains doubt about the role of ATM mutations in predisposition to B-CLL. Overall the penetrance of the
ATM gene in promoting a tumor phenotype is likely to be low and additional genetic events are almost certainly required.
MDM2, PSMD9/p27, Smac/DIABLO and STAT6) has also been analyzed but is independent of trisomy 12 status (Winkler et al., 2005).
17p13 deletions Deletions at 17p13 are associated with loss and/or mutation of the TP53 tumor suppressor gene and become more frequent as the disease progresses (Thornton et al., 2004). Disruption of the p53 pathway can occur through mechanisms other than TP53 mutation (see above also). Basal steady-state mRNA levels of TP53 are significantly decreased in B-CLL cells with respect to normal B lymphocytes, although the 17p status of these patients was not clear in this study (Secchiero et al., 2006). Additional forms of p53 dysfunction are evident when B-CLL lymphocytes are exposed to ultraviolet radiation, a known activator of the ATR-p53 pathway. Jones et al. (2004) demonstrated that the ATR-p53 pathway is suppressed in noncycling lymphocytes via ATR downregulation. Growing evidence has documented hypermethylation in the promoter region of TP53 and a decrease in the transcription of this gene in a subset of BCLL samples without TP53 deletions and/or mutations (Valganon et al., 2005). In addition, the possibility that BCLL cells can acquire p53 dysfunction through extrinsic factors in the microenvironment has been explored. Basic fibroblast growth factor inhibited p53 activation in response to ionising radiation in the majority of B-CLL cells analysed (Romanov et al., 2005).
6q deletions A minimal region of deletion has been assigned to 6q21 although 6q27 (Amiel et al., 1999) is also deleted in a smaller number of patients. These regions are both commonly deleted in many lymphoid malignancies (Merup et al., 1998; Takeuchi et al., 1998; Zhang et al., 2000). No definitive gene(s) have been attributed to this abnormality.
Trisomy 12 The pathogenic role of trisomy 12 in B-CLL remains unresolved. Molecular analysis of cases with amplifications at 12q13]q24 defined a smaller region of duplication at 12q13]q15 which includes the oncogene murine double minute 2 (MDM2) (Dierlamm et al., 1997; Merup et al., 1997). The MDM2 oncoprotein is involved in an autoregulatory feedback loop with p53 which it binds to and inactivates. Over-expression of MDM2 RNA has been detected in B-CLL when compared to normal lymphocytes (Watanabe et al., 1994) together with over-expression of various MDM2 proteins (p57, p59, p67 and p90) in various combinations (Haidar et al., 1997). A novel gene, CLL up-regulated gene 1 (CLLU1), is located on chromosome 12q22 and was identified as a result of differential expression amongst B-CLL samples with mutated and unmutated VH genes. Expression is increased in B-CLL but this is not specific for cells bearing the trisomy 12 abnormality (Buhl et al., 2006). CLLU1 was not detected in normal tissue or other hematologic tumors and this may, therefore, represent a B-CLL specific gene encoding a peptide structurally very similar to human interleukin 4. A number of other genes have been suggested and include E2F1, BAX, CDKN1B and CDK4 which are differentially upregulated in trisomy 12 B-CLL cells. Both CDK4 and CDKN1B lie on chromosome 12, so over-expression may suggest a gene dosage phenomenon (Kienle et al., 2005). Protein expression of chromosome 12 candidate genes (ANKS1B, APAF1, ARF3, CCND2, CDK2, CKD4, GLI,
Familial CLL
The term familial B-CLL is used to describe family pedigrees where B-CLL is present in multiple generations and transmitted vertically in a manner consistent with the expression of an autosomal dominant gene (Catovsky, 1997). The molecular pathogenesis may or may not differ from sporadic B-CLL and some of the hereditary factors could be the same. Identification of the genes that are inherited in a mutated form in B-CLL families could indicate those responsible for sporadic cases of B-CLL. A number of studies suggested that families with B-CLL display the phenomenon of anticipation (Horwitz et al., 1996; Horwitz, 1997; Yuille et al., 1998, 2000; Goldin et al., 1999; Wiernik et al., 2001), an increase in severity or earlier age of onset of a disease occurring with each subsequent generation, in an autosomal dominant manner (Sutherland and Richards, 1995). Analysis of family pedigrees with B-CLL showed a significant reduction in the age of onset, approximately 20 years between generations (Horwitz et al., 1996; Yuille et al., 1998; Goldin et al., 1999; Wiernik et al., 2001) with evidence of more severe disease in successive generations (Goldin et al., 1999; Yuille et al., 2000). These descriptions of anticipation were not a new phenomenon and had previously been reported in other familial cancer syndromes (Horwitz et al., 1996; Fraser, 1997; De Lord et al., 1998; Deshpande et al., 1998; Wiernik et al., 2000), but the scientific basis was unclear. The original reports of anticipation were for the inherited neurodegenerative disorders and the pathogenetic mechanism was subsequently found to be due to the extensive expansion of unstable trinucleotide repeat sequences (Sutherland and Richards, 1995). Epigenetic change (Petronis et al., 1997) and telomere abnormalities (Horwitz et al., 1996) may also contribute to the mechanism in complex disorders. B-CLL families have been examined for evidence of transmission of dynamic mutations involving these triplet repeats, but no potentially pathological expansion of CAG- or CCG-repeats has been detected (Benzow et al., 2002; Auer et al., 2007a). It has long been debated whether anticipation truly exists in B-CLL or if the phenomenon arises as a result of bias. The conclusion of the most recent study of a large, population-based database evaluating anticipation in lymphoproliferative disorders (Daugherty et al., 2005) does not suggest that anticipation occurs. Potential biases include heightened reporting of other family mem-
Cytogenet Genome Res 118:310–319 (2007)
315
bers, reduced fertility of affected individuals, selection of cases with simultaneous onset in parent and child as well as secular disease trends. Such sampling biases are inherent in detecting anticipation, being difficult to exclude without a large population registry. Although several studies attempted to address these issues in their design and continued to conclude that anticipation does occur in familial B-CLL (Goldin et al., 1999; Wiernik et al., 2001), the results of the molecular analyses do support Daugherty et al. (2005) in their lack of clinical evidence for anticipation in B-CLL. The genetic basis of familial B-CLL remains poorly understood and predisposition genes have not yet been identified. Various genome-wide analyses have been performed in an attempt to identify potential susceptibility loci (Summersgill et al., 2002; Goldin et al., 2003; Sellick et al., 2005; Ng et al., 2006). Comparative genomic hybridization analysis of 24 familial B-CLL cases demonstrated that the common regions of chromosomal loss observed in sporadic BCLL occur at a comparable frequency in familial B-CLL. In addition, novel regions of gain and loss were identified at Xp21]p11.2 and Xq21]qter, 2p14]p12 and 4q11]q21 suggesting the presence of predisposition gene(s) at these loci (Summersgill et al., 2002). Linkage analysis of 18 families suggests a region of interest in band 13q22.1 (Goldin et al., 2003) and detailed mapping of this region identified a 3.68 Mb region containing 13 candidate genes (Ng et al., 2007). Again no mutations were detected amongst the families studied, although 11 of the 13 genes exhibit immune restricted expression. When expression levels of ten of the genes were examined in sporadic B-CLL cells, down-regulation in seven was observed. Interestingly, two unaffected family members had evidence of monoclonal B-cell lymphocytosis by flow cytometry and these cells shared the 13q21.33]q22.2 haplotype (Ng et al., 2007). Using high density SNP arrays, the pericentric region of chromosome 11 was of suggestive linkage, but no other regions of significant linkage were identified (Sellick et al., 2005). Linkage analysis of the ATM region in familial B-CLL cases has failed to identify any predisposing loci on 11q23 (Bevan et al., 1999) (see above). Inheritance of a loss of function mutation of the P2RX7 receptor gene has been proposed as a candidate predisposition gene potentially contributing to B-CLL pathogenesis. Activation of the P2X7 receptor is an important mechanism of apoptosis in cells of the immune system, including B-
lymphocytes. This mutation has been observed in the germline of two families with B-CLL together with pronounced loss of P2X7 function (Wiley et al., 2002). The prevalence of the mutation is also increased in cases of sporadic B-CLL (Wiley et al., 2002) and affects survival (Thunberg et al., 2002). However, larger studies by several different groups have failed to corroborate these initial findings (Starczynski et al., 2003; Zhang et al., 2003; Nuckel et al., 2004; Sellick et al., 2004) and question whether the P2RX7-A1513C polymorphism affects the risk of CLL. This controversial association highlights the problems in the search for susceptibility alleles if sample size is too small. Analysis of VH gene usage fails to provide any further information about the etiology of familial predisposition. One study demonstrated that the pattern of immunoglobulin gene usage and frequency of somatic mutation of familial B-CLL is indistinguishable from sporadic B-CLL (Sakai et al., 2000), but another similar study indicated an increase in VH gene hypermutation (Pritsch et al., 1999). In both there is concordance for IgVH hypermutation status between affected siblings. There are many potential indicators for the pathogenesis of familial B-CLL largely leading to confusion rather than clarity. The definitive mechanism/ genes await identification. Summary
The last decade has seen much progress in delineating the chromosome defects associated with B-CLL, relating them to prognosis, mutational status and gene expression profiles. Surprisingly the genetics of familial B-CLL lags behind that of sporadic B-CLL. It has become clear that the thinking behind the possible contribution of chromosomal deletions to the development of malignancy has been too simplistic. We may now combine some of the gene signatures together with the genetics and biomarkers of mutational status to consider stratification of treatment. The recent miRNA research has opened up potential new understanding of the disease process and novel therapy, although considerably more investigation is required. Possibly real emphasis should be placed on developing in vivo models. The next decade now has a firm platform of definition for the chromosomal alterations to move towards true functional understanding of this genetically complex leukemia.
References Aalto Y, El-Rifa W, Vilpo L, Ollila J, Nagy B, et al: Distinct gene expression profiling in chronic lymphocytic leukemia with 11q23 deletion. Leukemia 15:721–1728 (2001). Amiel A, Mulchanov I, Elis A, Gaber E, Manor Y, et al: Deletion of 6q27 in chronic lymphocytic leukemia and multiple myeloma detected by fluorescence in situ hybridization. Cancer Genet Cytogenet 112:53–56 (1999).
316
Auer RL, Jones C, Mullenbach RA, SyndercombeCourt D, Milligan DW, et al: Role for CCG-trinucleotide repeats in the pathogenesis of chronic lymphocytic leukemia. Blood 97: 509–515 (2001). Auer RL, Starczynski J, McElwaine S, Bertoni F, Newland AC, et al: Identification of a potential role for POU2AF1 and BTG4 in the deletion of 11q23 in chronic lymphocytic leukemia. Genes Chromosomes Cancer 43: 1–10 (2005).
Cytogenet Genome Res 118:310–319 (2007)
Auer RL, Dighiero G, Goldin LR, SyndercombeCourt D, Jones C, et al: Trinucleotide repeat dynamic mutation identifying susceptibility in familial and sporadic chronic lymphocytic leukaemia. Br J Haematol 136: 73–79 (2007a). Auer RL, Riaz S, Cotter FE: The 13q and 11q B cell chronic lymphocytic leukaemia associated regions derive from a common ancestral region in the zebrafish. Brit J Haematol 137: 443–453 (2007b).
Austen B, Powell JE, Alvi A, Edwards I, Hooper L, et al: Mutations in the ATM gene lead to impaired overall and treatment-free survival that is independent of IGVH mutation status in patients with B-CLL. Blood 106: 3175–3182 (2005). Bachman KE, Park BH, Rhee I, Rajagopalan H, Herman JG, et al: Histone modifications and silencing prior to DNA methylation of a tumor suppressor gene. Cancer Cell 3: 89–95 (2003). Baranova A, Hammarsund M, Ivanov D, Skoblov M, Sangfelt O, et al: Distinct organization of the candidate tumor suppressor gene RFP2 in human and mouse: multiple mRNA isoforms in both species- and human-specific antisense transcript RFP2OS. Gene 321: 103–112 (2003). Baylin SB, Herman JG, Graff JR, Vertino PM, Issa JP: Alterations in DNA methylation: a fundamental aspect of neoplasia. Adv Cancer Res 72: 141–196 (1998). Baylin SB, Esteller M, Rountree MR, Bachman KE, Schuebel K, Herman JG: Aberrant patterns of DNA methylation, chromatin formation and gene expression in cancer. Hum Mol Genet 10: 687–692 (2001). Benzow KA, Koob MD, Condie A, Catovsky D, Matutes E, et al: Instability of CAG-trinucleotide repeats in chronic lymphocytic leukemia. Leuk Lymphoma 43: 1987–1990 (2002). Bevan S, Catovsky D, Marossy A, Matutes E, Popat S, et al: Linkage analysis for ATM in familial B cell chronic lymphocytic leukaemia. Leukemia 13:1497–1500 (1999). Bichi R, Shinton SA, Martin ES, Koval A, Calin GA, et al: Human chronic lymphocytic leukemia modeled in mouse by targeted TCL1 expression. Proc Natl Acad Sci USA 99: 6955–6960 (2002). Bigoni R, Cuneo A, Roberti MG, Bardi A, Rigolin GM, et al: Chromosome aberrations in atypical chronic lymphocytic leukemia: a cytogenetic and interphase cytogenetic study. Leukemia 11: 1933–1940 (1997). Buhl AM, Jurlander J, Jorgensen FS, Ottesen AM, Cowland JB, et al: Identification of a gene on chromosome 12q22 uniquely overexpressed in chronic lymphocytic leukemia. Blood 107: 2904–2911 (2006). Bullrich F, Rasio D, Kitada S, Starostik P, Kipps T, et al: ATM mutations in B-cell chronic lymphocytic leukemia. Cancer Res 59: 24–27 (1999). Bullrich F, Fujii H, Calin G, Mabuchi H, Negrini M, et al: Characterization of the 13q14 tumor suppressor locus in CLL: identification of ALT1, an alternative splice variant of the LEU2 gene. Cancer Res 61: 6640–6648 (2001). Byrd JC, Gribben JG, Peterson BL, Grever MR, Lozanski G, et al: Select high-risk genetic features predict earlier progression following chemoimmunotherapy with fludarabine and rituximab in chronic lymphocytic leukemia: justification for risk-adapted therapy. J Clin Oncol 24: 437– 443 (2006). Byrd JC, Lin TS, Dalton JT, Wu D, Phelps MA, et al: Flavopiridol administered using a pharmacologically derived schedule is associated with marked clinical efficacy in refractory, genetically high-risk chronic lymphocytic leukemia. Blood 109:399–404 (2007). Caballero D, Garcia-Marco JA, Martino R, Mateos V, Ribera JM, et al: Allogeneic transplant with reduced intensity conditioning regimens may overcome the poor prognosis of B-cell chronic lymphocytic leukemia with unmutated immunoglobulin variable heavy-chain gene and chromosomal abnormalities (11q– and 17p–). Clin Cancer Res 11: 7757–7763 (2005).
Calin GA, Dumitru CD, Shimizu M, Bichi R, Zupo S, et al: Frequent deletions and down-regulation of micro-RNA genes miR15 and miR16 at 13q14 in chronic lymphocytic leukemia. Proc Natl Acad Sci USA 99: 15524–15529 (2002). Calin GA, Liu CG, Sevignani C, Ferracin M, Felli N, et al: MicroRNA profiling reveals distinct signatures in B cell chronic lymphocytic leukemias. Proc Natl Acad Sci USA 101:11755–11760 (2004). Calin GA, Ferracin M, Cimmino A, Di Leva G, Shimizu M, et al: A microRNA signature associated with prognosis and progression in chronic lymphocytic leukemia. N Engl J Med 353:1793– 1801 (2005a). Calin GA, Trapasso F, Shimizu M, Dumitru CD, Yendamuri S, et al: Familial cancer associated with a polymorphism in ARLTS1. N Engl J Med 352:1667–1676 (2005b). Catovsky D: The search for genetic clues in chronic lymphocytic leukemia. Hematol Cell Ther 39: S5–S11 (1997). Chen CZ: MicroRNAs as oncogenes and tumor suppressors. N Engl J Med 353: 1768–1771 (2005). Chen CZ, Li L, Lodish HF, Bartel DP: MicroRNAs modulate hematopoietic lineage differentiation. Science 303: 83–86 (2004). Cimmino A, Calin GA, Fabbri M, Iorio MV, Ferracin M, et al: miR-15 and miR-16 induce apoptosis by targeting BCL2. Proc Natl Acad Sci USA 102:13944–13949 (2005). Comings DE: A general theory of carcinogenesis. Proc Natl Acad Sci USA 70: 3324–3328 (1973). Corcoran MM, Rasool O, Liu Y, Iyengar A, Grander D, et al: Detailed molecular delineation of 13q14.3 loss in B-cell chronic lymphocytic leukemia. Blood 91:1382–1390 (1998). Cordone I, Masi S, Mauro FR, Soddu S, Morsilli O, et al: p53 expression in B-cell chronic lymphocytic leukemia: a marker of disease progression and poor prognosis. Blood 91: 4342–4349 (1998). Costinean S, Zanesi N, Pekarsky Y, Tili E, Volinia S, et al: Pre-B cell proliferation and lymphoblastic leukemia/high-grade lymphoma in E(mu)miR155 transgenic mice. Proc Natl Acad Sci USA 103: 7024–7029 (2006). Crespo M, Bosch F, Villamor N, Bellosillo B, Colomer D, et al: ZAP-70 expression as a surrogate for immunoglobulin-variable-region mutations in chronic lymphocytic leukemia. N Engl J Med 348:1764–1775 (2003). Criel A, Verhoef G, Vlietinck R, Mecucci C, Billiet J, et al: Further characterization of morphologically defined typical and atypical CLL: a clinical, immunophenotypic, cytogenetic and prognostic study on 390 cases. Br J Haematol 97: 383–391 (1997). Cuneo A, Rigolin GM, Bigoni R, De Angeli C, Veronese A, et al: Chronic lymphocytic leukemia with 6q– shows distinct hematological features and intermediate prognosis. Leukemia 18:476– 483 (2004). Damle RN, Wasil T, Fais F, Ghiotto F, Valetto A, et al: Ig V gene mutation status and CD38 expression as novel prognostic indicators in chronic lymphocytic leukemia. Blood 94: 1840–1847 (1999). Daugherty SE, Pfeiffer RM, Mellemkjaer L, Hemminki K, Goldin LR: No evidence for anticipation in lymphoproliferative tumors in population-based samples. Cancer Epidemiol Biomarkers Prev 14:1245–1250 (2005). De Lord C, Powles R, Mehta J, Wilson K, Treleaven J, et al: Familial acute myeloid leukaemia: four male members of a single family over three consecutive generations exhibiting anticipation. Br J Haematol 100: 557–560 (1998).
Deshpande HA, Hu XP, Marino P, Jan NA, Wiernik PH: Anticipation in familial plasma cell dyscrasias. Br J Haematol 103: 696–703 (1998). Dickinson JD, Smith LM, Sanger WG, Zhou G, Townley P, et al: Unique gene expression and clinical characteristics are associated with the 11q23 deletion in chronic lymphocytic leukaemia. Br J Haematol 128: 460–471 (2005). Dickinson JD, Gilmore J, Iqbal J, Sanger W, Lynch JC, et al: 11q22.3 deletion in B-chronic lymphocytic leukemia is specifically associated with bulky lymphadenopathy and ZAP-70 expression but not reduced expression of adhesion/ cell surface receptor molecules. Leuk Lymphoma 47:231–244 (2006a). Dickinson JD, Joshi, Iqbal J, Sanger W, Bierman PJ, Joshi SS: Genomic abnormalities in chronic lymphocytic leukemia influence gene expression by a gene dosage effect. Int J Mol Med 17: 769–778 (2006b). Dierlamm J, Wlodarska I, Michaux L, Vermeesch JR, Meeus P, et al: FISH identifies different types of duplications with 12q13–q15 as the commonly involved segment in B-cell lymphoproliferative malignancies characterized by partial trisomy 12. Genes Chromosomes Cancer 20:155–166 (1997). Dohner H, Fischer K, Bentz M, Hansen K, Benner A, et al: p53 gene deletion predicts for poor survival and non-response to therapy with purine analogs in chronic B-cell leukemias. Blood 85: 1580–1589 (1995). Dohner H, Stilgenbauer S, James MR, Benner A, Weilguni T, et al: 11q deletions identify a new subset of B-cell chronic lymphocytic leukemia characterized by extensive nodal involvement and inferior prognosis. Blood 89: 2516–2522 (1997). Dohner H, Stilgenbauer S, Benner A, Leupolt E, Krober A, et al: Genomic aberrations and survival in chronic lymphocytic leukemia. N Engl J Med 343: 1910–1916 (2000). Dreger P, Corradini P, Kimby E, Michallet M, Milligan D, et al: Indications for allogeneic stem cell transplantation in chronic lymphocytic leukemia: the EBMT transplant consensus. Leukemia 21:12–17 (2007). Eis PS, Tam W, Sun L, Chadburn A, Li Z, et al: Accumulation of miR-155 and BIC RNA in human B cell lymphomas. Proc Natl Acad Sci USA 102: 3627–3632 (2005). el Rouby S, Thomas A, Costin D, Rosenberg CR, Potmesil M, et al: p53 gene mutation in B-cell chronic lymphocytic leukemia is associated with drug resistance and is independent of MDR1/MDR3 gene expression. Blood 82:3452– 3459 (1993). Fegan C, Robinson H, Thompson P, Whittaker JA, White D: Karyotypic evolution in CLL: identification of a new sub-group of patients with deletions of 11q and advanced or progressive disease. Leukemia 9:2003–2008 (1995). Fero ML, Randel E, Gurley KE, Roberts JM, Kemp CJ: The murine gene p27Kip1 is haplo-insufficient for tumour suppression. Nature 396: 177– 180 (1998). Fraser FC: Trinucleotide repeats not the only cause of anticipation. Lancet 350: 459–460 (1997). Geisler CH, Philip P, Christensen BE, Hou-Jensen K, Pedersen NT, et al: In B-cell chronic lymphocytic leukaemia chromosome 17 abnormalities and not trisomy 12 are the single most important cytogenetic abnormalities for the prognosis: a cytogenetic and immunophenotypic study of 480 unselected newly diagnosed patients. Leuk Res 21:1011–1023 (1997).
Cytogenet Genome Res 118:310–319 (2007)
317
Geoffroy-Perez B, Janin N, Ossian K, Lauge A, Croquette MF, et al: Cancer risk in heterozygotes for ataxia-telangiectasia. Int J Cancer 93: 288– 293 (2001). Goldin LR, Sgambati M, Marti GE, Fontaine L, Ishibe N, Caporaso N: Anticipation in familial chronic lymphocytic leukemia. Am J Hum Genet 65: 265–269 (1999). Goldin LR, Ishibe N, Sgambati M, Marti GE, Fontaine L, et al: A genome scan of 18 families with chronic lymphocytic leukaemia. Br J Haematol 121:866–873 (2003). Guarini A, Gaidano G, Mauro FR, Capello D, Mancini F, et al: Chronic lymphocytic leukemia patients with highly stable and indolent disease show distinctive phenotypic and genotypic features. Blood 102:1035–1041 (2003). Haidar MA, El-Hajj H, Bueso-Ramos CE, Manshouri T, Glassman A, et al: Expression profile of MDM-2 proteins in chronic lymphocytic leukemia and their clinical relevance. Am J Hematol 54:189–195 (1997). Hamblin TJ, Davis Z, Gardiner A, Oscier DG, Stevenson FK: Unmutated Ig V(H) genes are associated with a more aggressive form of chronic lymphocytic leukemia. Blood 94: 1848–1854 (1999). Hamblin TJ, Orchard JA, Gardiner A, Oscier DG, Davis Z, Stevenson FK: Immunoglobulin V genes and CD38 expression in CLL. Blood 95: 2455–2457 (2000). Haslinger C, Schweifer N, Stilgenbauer S, Dohner H, Lichter P, et al: Microarray gene expression profiling of B-cell chronic lymphocytic leukemia subgroups defined by genomic aberrations and VH mutation status. J Clin Oncol 22:3937– 3949 (2004). He L, Thomson JM, Hemann MT, HernandoMonge E, Mu D, et al: A microRNA polycistron as a potential human oncogene. Nature 435: 828–833 (2005). Herling M, Patel KA. Khalili J, Schlette E, Kobayashi R, et al: TCL1 shows a regulated expression pattern in chronic lymphocytic leukemia that correlates with molecular subtypes and proliferative state. Leukemia 20: 280–285 (2006). Herman JG, Merlo A, Mao L, Lapidus RG, Issa, JP, et al: Inactivation of the CDKN2/p16/MTS1 gene is frequently associated with aberrant DNA methylation in all common human cancers. Cancer Res 55: 4525–4530 (1995). Horwitz M: The genetics of familial leukemia. Leukemia 11:1347–1359 (1997). Horwitz M, Goode EL, Jarvik GP: Anticipation in familial leukemia. Am J Hum Genet 59: 990– 998 (1996). Jones GG, Reaper PM, Pettitt AR, Sherrington PD: The ATR-p53 pathway is suppressed in noncycling normal and malignant lymphocytes. Oncogene 23: 1911–1921 (2004). Joshi AD, Dickinson JD, Hegde GV, Sanger WG, Armitage JO, et al: Bulky lymphadenopathy with poor clinical outcome is associated with ATM downregulation in B-cell chronic lymphocytic leukemia patients irrespective of 11q23 deletion. Cancer Genet Cytogenet 172: 120–126 (2007). Kalachikov S, Migliazza A, Cayanis E, Fracchiolla NS, Bonaldo MF, et al: Cloning and gene mapping of the chromosome 13q14 region deleted in chronic lymphocytic leukemia. Genomics 42:369–377 (1997). Kalla C, Nentwich H, Schlotter M, Mertens D, Wildenberger K, et al: Translocation t(X;11) (q13;q23) in B-cell chronic lymphocytic leukemia disrupts two novel genes. Genes Chromosomes Cancer 42: 128–143 (2005).
318
Kienle DL, Korz C, Hosch B, Benner A, Mertens D, et al: Evidence for distinct pathomechanisms in genetic subgroups of chronic lymphocytic leukemia revealed by quantitative expression analysis of cell cycle, activation, and apoptosis-associated genes. J Clin Oncol 23: 3780–3792 (2005). Kim SZ, Chow KU, Kukoc-Zivojnov N, Boehrer S, et al: Expression of ZAP-70 protein correlates with disease stage in chronic lymphocytic leukemia and is associated with, but not generally restricted to, non-mutated Ig VH status. Leuk Lymphoma 45:2037–2045 (2004). Kitamura E, Su G, Sossey-Alaoui K, Malaj E, Lewis J, et al: A transcription map of the minimally deleted region from 13q14 in B-cell chronic lymphocytic leukemia as defined by large scale sequencing of the 650 kb critical region. Oncogene 19: 5772–5780 (2000). Knauf WU, Knuutila S, Zeigmeister B, Thiel E: Trisomy 12 in B-cell chronic lymphocytic leukemia: correlation with advanced disease, atypical morphology, high levels of sCD25, and with refractoriness to treatment. Leuk Lymphoma 19:289–294 (1995). Knudson AG Jr: Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci USA 68: 820–823 (1971). Krober A, Seiler T, Benner A, Bullinger L, Bruckle E, et al: V(H) mutation status, CD38 expression level, genomic aberrations, and survival in chronic lymphocytic leukemia. Blood 100: 1410–1416 (2002). Liu Y, Corcoran M, Rasool O, Ivanova G, Ibbotson R, et al: Cloning of two candidate tumor suppressor genes within a 10 kb region on chromosome 13q14, frequently deleted in chronic lymphocytic leukemia. Oncogene 15: 2463–2473 (1997). Lozanski G, Heerema NA, Flinn IW, Smith L, Harbison J, et al: Alemtuzumab is an effective therapy for chronic lymphocytic leukemia with p53 mutations and deletions. Blood 103: 3278–3281 (2004). Mabuchi H, Fujii H, Calin G, Alder H, Negrini M, et al: Cloning and characterization of CLLD6, CLLD7, and CLLD8, novel candidate genes for leukemogenesis at chromosome 13q14, a region commonly deleted in B-cell chronic lymphocytic leukemia. Cancer Res 61: 2870–2877 (2001). Matrai Z, Lin K, Dennis M, Sherrington P, Zuzel M, et al: CD38 expression and Ig VH gene mutation in B-cell chronic lymphocytic leukemia. Blood 97:1902–1903 (2001). Matutes E, Oscier D, Garcia-Marco J, Ellis J, Copplestone A, et al: Trisomy 12 defines a group of CLL with atypical morphology: correlation between cytogenetic, clinical and laboratory features in 544 patients. Br J Haematol 92:382–388 (1996). Merlo A, Herman JG, Mao L, Lee DJ, Gabrielson E, et al: 5ⴕ CpG island methylation is associated with transcriptional silencing of the tumour suppressor p16/CDKN2/MTS1 in human cancers. Nat Med 1: 686–692 (1995). Mertens D, Wolf S, Schroeter P, Schaffner C, Dohner H, et al: Down-regulation of candidate tumor suppressor genes within chromosome band 13q14.3 is independent of the DNA methylation pattern in B-cell chronic lymphocytic leukemia. Blood 99: 4116–4121 (2002). Mertens D, Wolf S, Tschuch C, Mund C, Kienle D, et al: Allelic silencing at the tumor-suppressor locus 13q14.3 suggests an epigenetic tumorsuppressor mechanism. Proc Natl Acad Sci USA 103: 7741–7746 (2006).
Cytogenet Genome Res 118:310–319 (2007)
Merup M, Juliusson G, Wu X, Jansson M, Stellan B, et al: Amplification of multiple regions of chromosome 12, including 12q13–q15, in chronic lymphocytic leukaemia. Eur J Haematol 58: 174–180 (1997). Merup M, Moreno TC, Heyman M, Ronnberg K, Grander D, et al: 6q deletions in acute lymphoblastic leukemia and non-Hodgkin’s lymphomas. Blood 91:3397–3400 (1998). Migliazza A, Bosch F, Komatsu H, Cayanis E, Martinotti S, et al: Nucleotide sequence, transcription map, and mutation analysis of the 13q14 chromosomal region deleted in B-cell chronic lymphocytic leukemia. Blood 97: 2098–2104 (2001). Miska EA: How microRNAs control cell division, differentiation and death. Curr Opin Genet Dev 15:563–568 (2005). Morrell D, Chase CL, Swift M: Cancers in 44 families with ataxia-telangiectasia. Cancer Genet Cytogenet 50: 119–123 (1990). Neilson JR, Auer R, White D, Bienz N, Waters JJ, et al: Deletions at 11q identify a subset of patients with typical CLL who show consistent disease progression and reduced survival. Leukemia 11:1929–1932 (1997). Ng D, Marti GE, Fontaine L, Toro JR, Caporaso N, Goldin LR: High-density mapping and followup studies on chromosomal regions 1, 3, 6, 12, 13 and 17 in 28 families with chronic lymphocytic leukaemia. Br J Haematol 133: 59–61 (2006). Ng D, Toure O, Wei MH, Arthur DC, Abbasi F, et al: Identification of a novel chromosome region, 13q21.33–q22.2, for susceptibility genes in familial chronic lymphocytic leukemia. Blood 109:916–925 (2007). Nuckel H, Frey UH, Durig J, Duhrsen U, Siffert W: 1513A/C polymorphism in the P2X7 receptor gene in chronic lymphocytic leukemia: absence of correlation with clinical outcome. Eur J Haematol 72:259–263 (2004). O’Donnell KA, Wentzel EA, Zeller KI, Dang CV, Mendell JT: c-Myc-regulated microRNAs modulate E2F1 expression. Nature 435: 839–843 (2005). Olsen JH, Hahnemann JM, Borresen-Dale AL, Brondum-Nielsen K, Hammarstrom L, et al: Cancer in patients with ataxia-telangiectasia and in their relatives in the Nordic countries. J Natl Cancer Inst 93:121–127 (2001). Oscier DG, Gardiner AC, Mould SJ, Glide S, Davis ZA, et al: Multivariate analysis of prognostic factors in CLL: clinical stage, IGVH gene mutational status, and loss or mutation of the p53 gene are independent prognostic factors. Blood 100:1177–1184 (2002). Payne SR, Kemp CJ: Tumor suppressor genetics. Carcinogenesis 26:2031–2045 (2005). Pekarsky Y, Santanam U, Cimmino A, Palamarchuk A, Efanov A, et al: Tcl1 expression in chronic lymphocytic leukemia is regulated by miR-29 and miR-181. Cancer Res 66: 11590–11593 (2006). Petronis A, Kennedy JL, Paterson AD: Genetic anticipation: fact or artifact, genetics or epigenetics? Lancet 350: 1403–1404 (1997). Pettitt AR, Sherrington PD, Stewart G, Cawley JC, Taylor AM, Stankovic T: p53 dysfunction in Bcell chronic lymphocytic leukemia: inactivation of ATM as an alternative to TP53 mutation. Blood 98:814–822 (2001). Pritsch O, Troussard X, Magnac C, Mauro FR, Davi F, et al: VH gene usage by family members affected with chronic lymphocytic leukaemia. Br J Haematol 107:616–624 (1999).
Rassenti LZ, Huynh L, Toy TL, Chen L, Keating MJ, et al: ZAP-70 compared with immunoglobulin heavy-chain gene mutation status as a predictor of disease progression in chronic lymphocytic leukemia. N Engl J Med 351:893–901 (2004). Romanov VV, James CH, Sherrington PD, Pettitt AR: Basic fibroblast growth factor suppresses p53 activation in the neoplastic cells of a proportion of patients with chronic lymphocytic leukaemia. Oncogene 24: 6855–6860 (2005). Rudd MF, Sellick GS, Webb EL, Catovsky D, Houlston RS: Variants in the ATM-BRCA2CHEK2 axis predispose to chronic lymphocytic leukemia. Blood 108:638–644 (2006). Sakai A, Marti GE, Caporaso N, Pittaluga S, Touchman JW, et al: Analysis of expressed immunoglobulin heavy chain genes in familial B-CLL. Blood 95: 1413–1419 (2000). Schaffner C, Stilgenbauer S, Rappold GA, Dohner H, Lichter P: Somatic ATM mutations indicate a pathogenic role of ATM in B-cell chronic lymphocytic leukemia. Blood 94: 748–753 (1999). Secchiero P, Barbarotto E, Tiribelli M, Zerbinati C, di Iasio MG, et al: Functional integrity of the p53-mediated apoptotic pathway induced by the nongenotoxic agent nutlin-3 in B-cell chronic lymphocytic leukemia (B-CLL). Blood 107:4122–4129 (2006) Sellick GS, Rudd M, Eve P, Allinson R, Matutes E, et al: The P2X7 receptor gene A1513C polymorphism does not contribute to risk of familial or sporadic chronic lymphocytic leukemia. Cancer Epidemiol Biomarkers Prev 13: 1065–1067 (2004). Sellick GS, Webb EL, Allinson R, Matutes E, Dyer MJ, et al: A high-density SNP genomewide linkage scan for chronic lymphocytic leukemia-susceptibility loci. Am J Hum Genet 77: 420–429 (2005). Sellick GS, Catovsky D, Houlston RS: Relationship between ARLTS1 polymorphisms and risk of chronic lymphocytic leukemia. Leuk Res 30: 1573–1576 (2006). Stanbridge EJ, Der CJ, Doersen CJ, Nishimi RY, Peehl DM, et al: Human cell hybrids: analysis of transformation and tumorigenicity. Science 215:252–259 (1982). Stankovic T, Kidd AM, Sutcliffe A, McGuire GM, Robinson P, et al: ATM mutations and phenotypes in ataxia-telangiectasia families in the British Isles: expression of mutant ATM and the risk of leukemia, lymphoma, and breast cancer. Am J Hum Genet 62:334–345 (1998). Stankovic T, Weber P, Stewart G, Bedenham T, Murray J, et al: Inactivation of ataxia telangiectasia mutated gene in B-cell chronic lymphocytic leukaemia. Lancet 353: 26–29 (1999). Stankovic T, Stewart GS, Fegan C, Biggs P, Last J, et al: Ataxia telangiectasia mutated-deficient Bcell chronic lymphocytic leukemia occurs in pregerminal center cells and results in defective damage response and unrepaired chromosome damage. Blood 99:300–309 (2002). Starczynski J, Pepper C, Pratt G, Hooper L, Thomas A, et al: The P2X7 receptor gene polymorphism 1513 A]C has no effect on clinical prognostic markers, in vitro sensitivity to fludarabine, Bcl2 family protein expression or survival in B-cell chronic lymphocytic leukaemia. Br J Haematol 123: 66–71 (2003).
Starostik P, Manshouri T, O’Brien S, Freireich E, Kantarjian H, et al: Deficiency of the ATM protein expression defines an aggressive subgroup of B-cell chronic lymphocytic leukemia. Cancer Res 58:4552–4557 (1998). Stilgenbauer S, Liebisch P, James MR, Schroder M, Schlegelberger B, et al: Molecular cytogenetic delineation of a novel critical genomic region in chromosome bands 11q22.3–q23.1 in lymphoproliferative disorders. Proc Natl Acad Sci USA 93: 11837–11841 (1996). Stilgenbauer S, Nickolenko J, Wilhelm J, Wolf S, Weitz S, et al: Expressed sequences as candidates for a novel tumor suppressor gene at band 13q14 in B-cell chronic lymphocytic leukemia and mantle cell lymphoma. Oncogene 16:1891– 1897 (1998). Stilgenbauer S, Bullinger L, Benner A, Wildenberger K, Bentz M, et al: Incidence and clinical significance of 6q deletions in B cell chronic lymphocytic leukemia. Leukemia 13: 1331–1334 (1999). Su’ut L, O’Connor SJ, Richards SJ, Jones RA, Roberts BE, et al: Trisomy 12 is seen within a specific subtype of B-cell chronic lymphoproliferative disease affecting the peripheral blood/ bone marrow and co-segregates with elevated expression of CD11a. Br J Haematol 101: 165– 170 (1998). Summersgill B, Thornton P, Atkinson S, Matutes E, Shipley J, et al: Chromosomal imbalances in familial chronic lymphocytic leukaemia: a comparative genomic hybridisation analysis. Leukemia 16:1229–1232 (2002). Sutherland GR, Richards RI: Simple tandem DNA repeats and human genetic disease. Proc Natl Acad Sci USA 92:3636–3641 (1995). Takeuchi S, Koike M, Seriu T, Bartram CR, Schrappe M, et al: Frequent loss of heterozygosity on the long arm of chromosome 6: identification of two distinct regions of deletion in childhood acute lymphoblastic leukemia. Cancer Res 58: 2618–2623 (1998). Taylor AM: Ataxia telangiectasia genes and predisposition to leukaemia, lymphoma and breast cancer. Br J Cancer 66: 5–9 (1992). Taylor AM, Metcalfe JA, Thick J, Mak YF: Leukemia and lymphoma in ataxia telangiectasia. Blood 87:423–438 (1996). Thornton PD, Gruszka-Westwood AM, Hamoudi RA, Atkinson S, Kaczmarek P, et al: Characterisation of TP53 abnormalities in chronic lymphocytic leukaemia. Hematol J 5: 47–54 (2004). Thunberg U, Johnson A, Roos G, Thorn I, Tobin G, et al: CD38 expression is a poor predictor for VH gene mutational status and prognosis in chronic lymphocytic leukemia. Blood 97:1892– 1894 (2001). Thunberg U, Tobin G, Johnson A, Soderberg O, Padyukov L, et al: Polymorphism in the P2X7 receptor gene and survival in chronic lymphocytic leukaemia. Lancet 360: 1935–1939 (2002). Valganon M, Giraldo P, Agirre X, Larrayoz MJ, Rubio-Martinez A, et al: p53 Aberrations do not predict individual response to fludarabine in patients with B-cell chronic lymphocytic leukaemia in advanced stages Rai III/IV. Br J Haematol 129: 53–59 (2005). Virgilio L, Narducci MG, Isobe M, Billips LG, Cooper MD, et al: Identification of the TCL1 gene involved in T-cell malignancies. Proc Natl Acad Sci USA 91:12530–12534 (1994).
Vorechovsky I, Rasio D, Luo L, Monaco C, Hammarstrom L, et al: The ATM gene and susceptibility to breast cancer: analysis of 38 breast tumors reveals no evidence for mutation. Cancer Res 56:2726–2732 (1996). Vorechovsky I, Luo L, Ortmann E, Steinmann D, Dork T: Missense mutations at ATM gene and cancer risk. Lancet 353: 1276 (1999). Watanabe T, Hotta T, Ichikawa A, Kinoshita T, Nagai H, et al: The MDM2 oncogene overexpression in chronic lymphocytic leukemia and lowgrade lymphoma of B-cell origin. Blood 84: 3158–3165 (1994). Wiernik PH, Wang SQ, Hu XP, Marino P, Paietta E: Age of onset evidence for anticipation in familial non-Hodgkin’s lymphoma. Br J Haematol 108:72–79 (2000). Wiernik PH, Ashwin M, Hu XP, Paietta E, Brown K: Anticipation in familial chronic lymphocytic leukaemia. Br J Haematol 113: 407–414 (2001). Wiestner A, Rosenwald A, Barry TS, Wright G, Davis RE, et al: ZAP-70 expression identifies a chronic lymphocytic leukemia subtype with unmutated immunoglobulin genes, inferior clinical outcome, and distinct gene expression profile. Blood 101:4944–4951 (2003). Wiley JS, Dao-Ung LP, Gu BJ, Sluyter R, Shemon AN, et al: A loss-of-function polymorphic mutation in the cytolytic P2X7 receptor gene and chronic lymphocytic leukaemia: a molecular study. Lancet 359:1114–1119 (2002). Winkler D, Schneider C, Krober A, Pasqualucci L, Lichter P, et al: Protein expression analysis of chromosome 12 candidate genes in chronic lymphocytic leukemia (CLL). Leukemia 19: 1211–1215 (2005). Wolf S, Mertens D, Schaffner C, Korz C, Dohner H, et al: B-cell neoplasia associated gene with multiple splicing (BCMS): the candidate B-CLL gene on 13q14 comprises more than 560 kb covering all critical regions. Hum Mol Genet 10: 1275–1285 (2001). Yuille MR, Houlston RS, Catovsky D: Anticipation in familial chronic lymphocytic leukaemia. Leukemia 12:1696–1698 (1998). Yuille MR, Matutes E, Marossy A, Hilditch B, Catovsky D, Houlston RS: Familial chronic lymphocytic leukaemia: a survey and review of published studies. Br J Haematol 109: 794–799 (2000). Yuille MR, Condie A, Hudson CD, Bradshaw PS, Stone EM, et al: ATM mutations are rare in familial chronic lymphocytic leukemia. Blood 100:603–609 (2002). Zhang LY, Ibbotson RE, Orchard JA, Gardiner AC, Seear RV, et al: P2X7 polymorphism and chronic lymphocytic leukaemia: lack of correlation with incidence, survival and abnormalities of chromosome 12. Leukemia 17: 2097–2100 (2003). Zhang Y, Matthiesen P, Harder S, Siebert R, Castoldi G, et al: A 3-cM commonly deleted region in 6q21 in leukemias and lymphomas delineated by fluorescence in situ hybridization. Genes Chromosomes Cancer 27: 52–58 (2000). Zhu Y, Monni O, El-Rifai W, Siitonen SM, Vilpo L, et al: Discontinuous deletions at 11q23 in B cell chronic lymphocytic leukemia. Leukemia 13: 708–712 (1999).
Cytogenet Genome Res 118:310–319 (2007)
319
some 1q rearrangement is a particularly frequent event in a broad spectrum of human cancers. It is observed in Burkitt lymphoma (BL), Diffuse Large B Cell Lymphoma (DLBCL), Follicular Lymphoma (FL) and B cell acute leukaemias (BALL) (Le Baccon et al., 2001). Importantly, 1q rearrangement is among the most frequent events observed in multiple myeloma (MM) (Avet-Loiseau et al., 1997; Cigudosa et al., 1998; Sawyer et al., 1998, 2005; Le Baccon et al., 2001; Liebisch et al., 2003; Gutierrez et al., 2004) where it is associated with deregulated 1q gene expression, disease progression and poor prognosis (Carrasco et al., 2006; Hanamura et al., 2006; Shaughnessy Jr et al., 2006; Zhan et al., 2006a). Significantly the major 1q breakpoint observed in B cell non-Hodgkin lymphoma and MM localises to the 1q12 heterochromatin region (Le Baccon et al., 2001). Rearrangement of 1q has also been described in T cell lymphoma (Oshiro et al., 2006), in the myeloproliferative disorders (Andrieux et al., 2003; Sambani et al., 2005) and in numerous solid tumours (Hattinger et al., 2002; Lu et al., 2002; Zudaire et al., 2002; Cheng et al., 2004; Tomlinson et al., 2005; Mendrzyk et al., 2006; Stange et al., 2006; van Dekken et al., 2006). In view of the unique structural features of 1q rearrangements in lymphoma and MM, and the recurrent involvement of 1q12 constitutive heterochromatin in these events, they are likely to drive complex oncogenic processes that target multiple genes through both genetic and epigenetic mechanisms. Based on i) recent integrated gene expression and array comparative genomic hybridisation (aCGH) profiling in multiple myeloma and ii) increasingly abundant literature pointing to an important role for heterochromatin in the control of normal and pathological gene silencing, this review discusses the genetics and epigenetics of 1q rearrangements in haematological malignancies. Genetic impact of 1q rearrangements in haematological malignancies – insights from aCGH and gene expression profiling in multiple myeloma
Gene expression profiling has amply contributed to improved molecular characterisation of the haematological malignancies (Staudt, 2003). More recently, the integration of genomic- (conventional cytogenetics/FISH, arrayCGH…) with transcriptome-profiling has allowed assessment of how gene expression profiles are linked to genetic abnormalities in tumour cells. In this setting, recent landmark studies in multiple myeloma (MM) have shed considerable light on the pathological impact of 1q rearrangements in this disorder (Carrasco et al., 2006; Shaughnessy Jr et al., 2006; Zhan et al., 2006a, b). MM which is caused by malignant transformation of plasma cells that subsequently home to and accumulate in the bone marrow is characterised by numerous structural and numerical chromosome aberrations (Avet-Loiseau et al., 2007). It has been hypothesised that MM pathogenesis proceeds via two major oncogenic pathways. Hyperdiploid MM involves multiple trisomies of chromosomes 3, 5, 7, 9,
11, 15, 19 and 21 whereas non-hyperdiploid MM are associated with recurrent IGH gene translocations and are generally shown to be of poorer prognosis (Fonseca et al., 2004). In order to identify genetic events underlying the genesis and progression of MM, Carrasco et al. (2006) have recently performed a high-resolution analysis of recurrent copy number alterations (CNAs) and gene expression profiles in a collection of MM cell lines (47) and a panel of newly diagnosed MM patients (67) for which detailed clinical annotation was available. All patients were treated with high-dose chemotherapy followed by tandem transplants and had a median follow-up of 43 months (range 5–65 months) (Barlogie et al., 2006). In keeping with the molecular heterogeneity of MM, unsupervised clustering and non-negative matrix factorization of high resolution aCGH data uncovered distinct genomic subtypes of prognostic significance (Carrasco et al., 2006). A major finding was that hyperdiploid MM could be further subdivided into two groups, one exhibiting trisomies of the odd chromosomes (3, 5, 9, 11, 15, 19 and 21) and a second exhibiting, in addition, gains of chromosomes 1q and 7, deletion of chromosome 13 and absence of trisomy 11. Significantly, poor survival was associated with gain of 1q and/or loss of 13q. Further integration with expression data generated a refined list of MM gene candidates residing within these minimal common regions (MCRs). In particular this analysis revealed a marked enrichment in overexpressed genes residing at 1q21]q23 (spans approximately 143–158 kb). Significantly overexpressed genes (of a list of 44) included previously identified target genes for 1q rearrangements in B cell malignancies (BCL9, MLLT11, JTB and FCRL5), genes known to play a role in MM pathobiology (IL6R, MCL1...) and other novel, functionally diverse genes not previously shown to be involved in MM. It is also interesting to note the presence of one microRNA gene within this 1q region (hsa-mir-9–1) (Carrasco et al., 2006). Overall, this integrated aCGH/transcriptome analysis identified 87 discrete MCRs of amplification/deletion of strong biological and clinical significance in MM. Interestingly, 20% of these 87 high priority MCRs contain microRNAs (Carrasco et al., 2006). It will be important to investigate the expression status and pathological relevance of these genes in this disorder. Further evidence for the importance of 1q gene deregulation in the pathogenesis or progression of MM has been obtained from unsupervised hierarchic clustering of transcriptome data from 414 newly diagnosed MM patients (Zhan et al., 2006a). This study allowed identification of seven disease subtypes that were strongly influenced by known genetic anomalies such as recurrent IGH gene translocations that target cMAF- and MAFB, CCND1, CCND3 and WHSC1 or hyperdiploidy. MM cases falling into poorrisk groups characterised by a proliferation signature or MMSET deregulation, showed significant overexpression of genes mapping to 1q. This 1q gene overexpression was not evident in the other five groups identified and thus seemed to be linked to the presence of a proliferation signature.
Cytogenet Genome Res 118:320–327 (2007)
321
Whether 1q gene deregulation is a marker of or causal in the acquisition of the proliferation signature is not known at the present time (Zhan et al., 2006a). The importance of 1q gain and/or overexpression of 1q genes in MM progression and prognosis is also evident from transcriptome profiling in a large series of MM and the related disorders, benign monoclonal gammopathy of undetermined significance (MGUS) and smoldering MM (SMM) (Zhan et al., 2006b). In this study, Significance Analysis of Microarrays identified a set of 52 genes, that were involved in important pathways in cancer and were differentially expressed between plasma cells from healthy individuals and patients with MGUS/SMM and symptomatic MM (Zhan et al., 2006b). In addition, the presence of an MGUS signature in MM (termed MGUS-like MM) was linked to good prognosis. Interestingly, interphase FISH analysis showed 1q gain/amplification to be less frequent in MGUS-like MM relative to other MM and the negative prognostic impact of amp1q21 was only seen in the non-MGUS-like MM (Zhan et al., 2006b). In keeping with this, previous FISH or aCGHdefined gains/amplifications of 1q in MM tumor cells have been linked to inferior survival (Hanamura et al., 2006) or progression of SMM to overt disease (Rosinol et al., 2005). Detailed gene expression profiling and bioinformatics analyses (log rank tests of expression quartiles) in the tumor cells of a large series of newly diagnosed MM patients (532, treated on two separate protocols) identified 70 genes, that were linked to early disease-related death. 30% of these genes mapped to chromosome 1 (Shaughnessy Jr et al., 2006). Importantly, the majority of up-regulated genes mapped to chromosome 1q and of down-regulated genes mapped to chromosome 1p. Thus this data provides a strong indication that transcriptional deregulation of genes mapping to chromosome 1 may contribute to disease progression in MM and that expression profiling can be used to identify high-risk disease (Shaughnessy Jr et al., 2006). Finally, a recent transcriptional profiling study, in a series of 77 MM patients, identified 61 genes that distinguished MM cases with 1q gain (1q/gain) from those without (1q/normal). 41 of the 43 upregulated genes mapped to 1q12]q44 whereas a significant proportion of the downregulated genes was localized on chromosomes 13q and 11. While the 1q gene list did not include the IL6R or CKS1B genes, previously identified MM 1q target genes, such as PSMD4, UBAP2L and UBE2Q1, were present (Fabris et al., 2007). Interestingly, these 1q genes are implicated in the ubiquitin-proteasome pathway; PSMD4 encodes the 26S proteasome non-ATPase regulatory subunit 4 and UBE2Q1 encodes a putative member of the E2 ubiquitin-conjugating enzyme family. An issue that remains to be resolved is why a recurring MM-specific 1q gene expression signature has not been identified across all transcriptome analyses. This may reflect differences in sample populations and data mining approaches. Indeed, alternative data mining strategies in other MM series have revealed other candidate genes of potential interest in this disorder (Moreaux et al., 2006; Condomines et al., 2007). Of note is the identification of
322
Cytogenet Genome Res 118:320–327 (2007)
CD200 – a membrane glycoprotein involved in the suppression of T cell-mediated immune responses – as a novel prognostic indicator in MM (Moreaux et al., 2006). Alternatively, it is possible that the manner and ‘differentiation window’ in which 1q gain is achieved in tumour cells may favour particular types of 1q rearrangement (1q breakpoint heterogeneity?) and/or patterns of 1q gene deregulation. In support of this, FISH-defined amplification of 1q21 has been shown to be absent in MGUS: its presence in a subset of patients with SMM was associated with a higher risk of conversion to MM (Rosinol et al., 2005; Hanamura et al., 2006). Likewise, we have identified 1q12 juxtacentromeric heterochromatin as the single most frequent 1q breakpoint site in DLBCL, FL, BL and MM patients but this breakpoint is rarely observed in marginal zone or small lymphocytic lymphoma (Le Baccon et al., 2001). Also, 1q duplication (dup1q) is particularly frequent in Burkitt lymphoma but is relatively infrequent elsewhere. In this setting, it will be of considerable interest to assess the incidence of 1q constitutive heterochromatin rearrangements, for example, in the MM patient series that have been studied by gene expression profiling. It should be pointed out that CGH arrays generally do not contain probes to repeat-rich regions (1q12 is enriched in satellite II repeats). Addressing these issues will help to define whether 1q gain/ 1q gene deregulation is causal or simply a surrogate marker for high risk disease in MM and should help to shed light on the role of heterochromatin abnormalities in this disorder. Finally, the non-random, secondary nature of 1q anomalies in B cell non-Hodgkin lymphoma is strongly suggestive of a role in tumour progression. Indeed, early cytogenetic studies identified a negative prognostic impact for 1q rearrangements in DLBCL (Offit et al., 1991). Recent integrated aCGH/FISH and transcriptome profiling in DLBCL and BL, although confirming the frequency of 1q alterations, have not as yet delivered specific information on the biological significance of 1q alterations/gene deregulation in these disorders (Bea et al., 2005; Dave et al., 2006; Hummel et al., 2006). Array CGH has, however, identified frequent 1q gain in the germinal centre B cell-like DLBCL molecular subgroup compared to the activated B-cell-like subgroup (Tagawa et al., 2005). Constitutive heterochromatin, gene silencing and epigenetic cancer mechanisms: implications for 1q-linked oncogenic mechanisms in haematological malignancies?
While 1q rearrangements are clearly associated with 1q gene deregulation, at least in MM, they are also likely to drive oncogenesis through other mechanisms that are linked to the unusual structural features of the rearrangements. In particular, as stated earlier, we and others have found that the 1q12 satellite II-rich constitutive heterochromatin region is a remarkably recurring breakpoint in a majority of DLBCL, FL and MM cases with 1q rearrangements
(Le Baccon et al., 2001). Indeed, we have observed MM cases with multiple copies of 1q12 heterochromatin in a single cell (Le Baccon et al., 2001). Taken together, this is strong evidence of heterochromatin-dependent oncogenic mechanisms in these disorders. It should be noted that 1q12 constitutive heterochromatin rearrangements, unless specifically searched for with appropriate probe sets, are largely missed by conventional FISH or aCGH mapping approaches (repeat-sequence probes are selectively excluded from CGH arrays). Constitutive heterochromatin is now recognised as performing a key role in the epigenetic regulation of gene expression, through a process termed gene silencing. We hypothesize that heterochromatin-dependent oncogenic mechanisms may contribute significantly to the pathogenesis of DLBCL, FL and MM with 1q rearrangements. In this setting, the second part of this review provides a brief overview of constitutive heterochromatin and gene silencing, followed by examples of its perturbation in cancer cells. The epigenome, an overview The epigenome refers to the totality of the reversible and inheritable information that is not directly coded by the genome. Epigenetic information is encoded by chemical modifications to DNA or histones which, when assembled into nucleosomes, form the basic structural and functional building blocks of chromatin. A major discovery in recent years has been the realisation that these modifications (‘marks’) constitute the basis of an epigenetic code that is capable of controlling genome structure and function (Strahl and Allis, 2000). This code operates through modulation of chromatin accessibility to factors that control gene transcription, DNA replication and repair. The ‘writing’ and functional read-out of this epigenetic code depends on specific and tightly regulated signalling pathways. For example, histone acetylation which facilitates the formation of an ‘open’ chromatin structure (euchromatin) is mediated by histone acetyl transferases (HATs). Removal of ‘acetylation marks’ and subsequent formation of ‘closed’ chromatin structures (heterochromatin) is mediated by histone deacetylases (HDACs). In recent years it has become increasingly evident that histone modifications constitute a combinatorial, indexing system for the dynamic and functional compartmentalization of chromatin. For example, trimethylation of histone H3-K9 is an epigenetic mark for transcriptionally repressed heterochromatin while acetylation at H3-K4 is associated with transcriptional activation (Fischle et al., 2003). This chromatin indexing is now thought to be rendered possible in large part through the action of specialised chromatin binding proteins that specifically read or recognise and subsequently bind these modifications. For example, the binding of acetylated histones is mediated by bromodomain-bearing proteins while methylated histones are recognised and bound by chromodomain-bearing proteins such as HP1 (binez trimethylated H3-K9 (Fischle et al., 2003; Nightingale et al., 2006)) or plant homeodomainbearing proteins (PHD) in the case of trimethylated H3-K4
which epigenetically ‘marks’ transcriptionally active chromatin (Mellor, 2006). During development and differentiation, cellular phenotypes are stably propagated through numerous cell divisions. This epigenetic ‘cell memory’ helps to maintain stable patterns of gene expression. Particularly exciting is the increasing probability that DNA methylation, post-translational histone modification and differential incorporation of histone variants into individual or groups of nucleosomes play a vital role in chromatin compartmentalization and the inheritance and maintenance of cellular memory (Turner, 2002; Govin et al., 2005; Hake and Allis, 2006). In view of the importance of this epigenetic code in controlling DNA function it is not surprising to find that epigenetic deregulation has emerged as a key event in the development of diseases such as cancer (Lund and van Lohuizen, 2004). Examples of disturbed chromatin structure in cancer include global reductions in monoacetylated H4K16 and trimethylated H4-K20 – hallmarks of euchromatin and heterochromatin, respectively (Fraga et al., 2005). Indeed, an ‘epigenetic progenitor’ model for the origin of cancer has recently been proposed (Feinberg et al., 2006). This new model proposes that cancer has a fundamentally common basis that is rooted in a polyclonal epigenetic disruption of stem/progenitor cells, mediated by ‘tumour-progenitor genes’. Thus, tumour cell heterogeneity is in part due to epigenetic variation in progenitor cells and epigenetic and genetic instability in tumour progression (Feinberg et al., 2006). This new concept has far-reaching implications for cancer prevention and therapy. Constitutive heterochromatin and gene silencing Constitutive heterochromatin is a specialised compartment within the chromatin and is found primarily at centromeres, telomeres and the pericentromeric regions of certain chromosomes (Grewal and Jia, 2007). Constitutive heterochromatin is distinguished from facultative heterochromatin on the basis that it remains constitutively condensed throughout the cell cycle. It is composed of genepoor, repetitive DNA sequences (alpha satellite, satellite II and III) that show distinct epigenetic features i.e. late-replication, high levels of DNA methylation and ‘repressive’ histone codes such as histone H3-K9 and H3-K20 trimethylation (Grewal and Jia, 2007). While the precise mechanisms for the assembly and maintenance of constitutive heterochromatin in mammalian cells are not clearly established, histone modifications and HP1 binding appear to have a crucial role. Evidence also exists for a role of an RNA component in heterochromatin formation (Bernstein and Allis, 2005; Grewal and Jia, 2007). Apart from its role in the control of genomic stability, constitutive heterochromatin participates in the epigenetic regulation of gene expression in numerous species through a highly conserved mechanism known as gene silencing. This phenomenon results from heterochromatin spreading into adjacent euchromatin and induces gene repression (Fisher and Merkenschlager, 2002; Talbert and Henikoff, 2006; Grewal and Jia, 2007). Recent data indicates that this
Cytogenet Genome Res 118:320–327 (2007)
323
mechanism also operates in mammalian cells. For example, during lymphoid development, stable transcriptional repression of certain lymphoid-specific genes is tightly correlated to their association with constitutive heterochromatin and formation of repressive histone codes (Brown et al., 1997, 1999; Grogan et al., 2001; Su et al., 2004). Evidence exists that perinuclear localisation of heterochromatin, at least in yeast, facilitates heterochromatin-induced transcriptional silencing (Andrulis et al., 1998). It is also clear that genes are non-randomly positioned within the nucleus and that proximity to heterochromatin influences transcriptional activity (Cremer and Cremer, 2001; Chubb and Bickmore, 2003; Spector, 2003). For instance, it has been shown that the gene-rich human chromosome 19 preferentially localises to the nuclear interior whereas the relatively gene-poor chromosome 18 is localised to the nuclear periphery (Croft et al., 1999) although this organisation appears to be partially lost in tumour cells (Cremer et al., 2003). Furthermore, the spatial organisation of chromosome territories is largely conserved in higher primates (Tanabe et al., 2002), and depends on both cell type (Alcobia et al., 2000, 2003; Cremer et al., 2003) and cell cycle stage (Vourc’h et al., 1993; Walter et al., 2003). Differential nuclear positioning of genes down- or upregulated during B and T lymphocyte development has been described: these include the RAG1 and RAG2 genes during B and T lymphocyte development and cytokine genes during TH1 or TH2 polarisation in T lymphocyte differentiation (Brown et al., 1999; Grogan et al., 2001). Specific positioning of immunoglobulin gene loci, IGH and IGK loci has also been shown during murine B cell development (Kosak et al., 2002). A further example concerns the VpreB1/ 5 loci which have been shown to undergo dynamic changes with respect to nuclear positioning and heterochromatin association in response to pre-B-cell receptor signalling, thereby providing a possible mechanism for their transcriptional silencing (Parker et al., 2005). Indeed, it has been proposed, in a rather elegant model, that reversible gene repression most likely involves local heterochromatinisation at target promoters whereas stable gene repression or silencing requires further repositioning of the repressed gene to constitutive heterochromatin foci (Fisher and Merkenschlager, 2002). In this setting, it is interesting to note that 1q12 heterochromatin can recruit transcriptional regulators such as Ikaros and members of the Polycomb family (Saurin et al., 1998; Sewalt et al., 2002). In the case of Ikaros, heterochromatin interaction has been shown to require the DNA binding domain while Polycomb recruitment requires functional SUV39H1 histone methyltransferase activity (Saurin et al., 1998; Sewalt et al., 2002). KAP1/TIF1beta which is proposed to be a universal co-repressor protein for the KRAB zinc finger protein superfamily of transcriptional repressors has also been shown to be recruited to 1q12 heterochromatin (Ryan et al., 1999). The functional significance of these associations is not known but they may allow the formation of specialised transcriptional silencing compartments (1q12 heterochromatin-associated?) in the interphase nucleus.
324
Cytogenet Genome Res 118:320–327 (2007)
Constitutive heterochromatin, cell stress, senescence and tumour suppression Physiological processes such as the response to cell stress or the onset of cellular senescence also appear to be mediated through specialised constitutive heterochromatin compartments. At the cellular level, a very striking effect of stress is the rapid and reversible redistribution of HSF1 into nuclear structures termed nuclear stress granules which form primarily on the 9q12 locus in humans. Within these structures, HSF1 binds to satellite III repeated elements concomitant with satellite III repeat expression (Metz et al., 2004). Likewise, the formation of specialised senescence activated heterochromatin foci (SAHF) that exhibit features of constitutive heterochromatin such as enrichment in heterochromatin protein 1 (HP1), and histone H3-K9 trimethylation, appear to be linked to stable gene silencing during cellular senescence (Narita et al., 2003). These SAHF may provide a chromatin buffer preventing activation of proliferation genes in cells entering a senescent state thus providing a barrier to malignant transformation. In support of this, loss of senescence, resulting from decreased activity of the histone methyl transferase, SUV39H1, has been linked to rapid onset of lymphoma in mouse models of p53 or oncogenic RAS-driven oncogenesis (Braig et al., 2005). Finally, oncogenic signalling may globally perturb heterochromatin regulated gene silencing. Indeed, constitutive JAK signaling globally counteracts heterochromatic gene silencing in Drosophila (Shi et al., 2006). This remarkable finding would predict similar perturbations in human myeloproliferative disorders that show constitutive JAK signalling as a consequence of the recently identified JAK2 V617F mutation (Campbell and Green, 2006). Taken together these data indicate a key role for constitutive heterochromatin in the control of normal and pathological gene silencing. An important challenge will be to determine whether chromosome anomalies that disturb constitutive heterochromatin compartments impact on these important epigenetic, regulatory mechanisms. Epigenetic features of 1q12 constitutive heterochromatin rearrangements in B lymphoma cells
We have hypothesized that 1q12 constitutive heterochromatin anomalies in haematological malignancies might represent the molecular equivalent of heterochromatin rearrangements that provoke aberrant gene silencing by ‘position effects’ in other species such as Drosophila and yeast. As a first step towards testing this hypothesis, we have compared 1q12 heterochromatin organisation in normal and tumour B lymphocytes with 1q rearrangements. In keeping with our hypothesis, dramatic spatial disorganisation of 1q12 constitutive heterochromatin and partner chromosomes is observed in tumour B lymphocytes with 1q rearrangements (Barki-Celli et al., 2005). For example, in Burkitt lymphoma cells 1q12 heterochromatin duplication was associated with the formation of particularly large 1q12 het-
erochromatin foci that were displaced toward the nuclear interior (Barki-Celli et al., 2005). We also found that 1q12 heterochromatin dynamically organises to the nuclear periphery in a non-random, cell cycle-dependent manner in normal B cells. Indeed, B cell activation, through antigen receptor engagement, was found to induce dramatic relocalisation of 1q12 domains to the extreme nuclear periphery. In addition, events which favoured increased local or global chromatin acetylation (enhancer activity or inhibition of HDAC inhibition) were associated with abnormal spatial positioning of 1q12 heterochromatin. Taken together, these findings suggest that aberrations affecting constitutive heterochromatin in lymphoma cells can induce long range epigenetic alterations that have the potential to induce gene silencing. Such ‘position effects’ may underpin, at least partially, the pathological role of particularly frequent 1q12 heterochromatin rearrangements in human tumours. It is also possible that the presence of excess constitutive heterochromatin in the same tumour cells might lead to inappropriate derepression of genes elsewhere in the genome. This might occur by ‘titration’ of key transcriptional regulators/heterochromatin ‘components’ away from their target genes and into constitutive heterochromatin compartments (Le Baccon et al., 2001). In this respect, it is worth noting that 1q/gain MM patients show overexpression of various cancer testis anti-
gen (CTA) genes (Chng et al., 2007). CTA are expressed in testis and malignant tumours but rarely in nongametogenic tissues. In this respect, they represent attractive targets for cancer vaccination approaches. Efforts to systematically analyse the expression pattern of CTA are now underway (Condomines et al., 2007). Integrated genomic, transcriptomic and epigenomic profiling will be required to gain a clear picture of how longrange epigenetic alterations associated with 1q12 constitutive heterochromatin abnormalities impact on gene expression and function in tumour B cells. Particular attention will have to be paid to the identification of copy number independent alterations in gene expression since these are likely to yield important information concerning epigenetically deregulated loci (Frigola et al., 2006; Stransky et al., 2006). Ultimately, these studies will have direct clinical relevance since epigenetic therapy is now a viable option for correction of epigenetic perturbations in cancer cells (Fandy et al., 2007; Taddei et al., 2005). Acknowledgements We are grateful to Dr. Saadi Khochbin and Dr. Claire Vourc’h (INSERM U823) for stimulating discussions.
References Alcobia I, Dilao R, Parreira L: Spatial associations of centromeres in the nuclei of hematopoietic cells: evidence for cell-type-specific organizational patterns. Blood 95: 1608–1615 (2000). Alcobia I, Quina AS, Neves H, Clode N, Parreira L: The spatial organization of centromeric heterochromatin during normal human lymphopoiesis: evidence for ontogenically determined spatial patterns. Exp Cell Res 290: 358–369 (2003). Andrieux J, Demory JL, Caulier MT, Agape P, Wetterwald M, et al: Karyotypic abnormalities in myelofibrosis following polycythemia vera. Cancer Genet Cytogenet 140: 118–123. (2003). Andrulis ED, Neiman AM, Zappulla DC, Sternglanz R: Perinuclear localization of chromatin facilitates transcriptional silencing. Nature 394: 592–595 (1998). Erratum, Nature 395: 525 (1998). Avet-Loiseau H, Vigier M, Moreau A, Mellerin MP, Gaillard F, et al: Comparative genomic hybridization detects genomic abnormalities in 80% of follicular lymphomas. Br J Haematol 97:119– 122 (1997). Avet-Loiseau H, Attal M, Moreau P, Charbonnel C, Garban F, et al: Genetic abnormalities and survival in multiple myeloma: the experience of the Intergroupe Francophone du Myelome. Blood 109:3489–3495 (2007). Barki-Celli L, Lefebvre C, Le Baccon P, Nadeau G, Bonnefoix T, et al: Differences in nuclear positioning of 1q12 pericentric heterochromatin in normal and tumor B lymphocytes with 1q rearrangements. Genes Chromosomes Cancer 43: 339–349 (2005).
Barlogie B, Tricot G, Anaissie E, Shaughnessy J, Rasmussen E, et al: Thalidomide and hematopoietic-cell transplantation for multiple myeloma. N Engl J Med 354:1021–1030 (2006). Bea S, Zettl A, Wright G, Salaverria I, Jehn P, et al: Diffuse large B-cell lymphoma subgroups have distinct genetic profiles that influence tumor biology and improve gene-expression-based survival prediction. Blood 106: 3183–3190 (2005). Bernstein E, Allis CD: RNA meets chromatin. Genes Dev 19: 1635–1655 (2005). Braig M, Lee S, Loddenkemper C, Rudolph C, Peters AH, et al: Oncogene-induced senescence as an initial barrier in lymphoma development. Nature 436: 660–665 (2005). Brown KE, Guest SS, Smale ST, Hahm K, Merkenschlager M, et al: Association of transcriptionally silent genes with Ikaros complexes at centromeric heterochromatin. Cell 91: 845–854 (1997). Brown KE, Baxter J, Graf D, Merkenschlager M, Fisher AG: Dynamic repositioning of genes in the nucleus of lymphocytes preparing for cell division. Mol Cell 3: 207–217 (1999). Busson-Le Coniat M, Salomon-Nguyen F, Dastugue N, Maarek O, Lafage-Pochitaloff M, et al: Fluorescence in situ hybridization analysis of chromosome 1 abnormalities in hematopoietic disorders: rearrangements of DNA satellite II and new recurrent translocations. Leukemia 13:1975–1981 (1999). Campbell PJ, Green AR: The myeloproliferative disorders. N Engl J Med 355: 2452–2466 (2006).
Carrasco DR, Tonon G, Huang Y, Zhang Y, Sinha R, et al: High-resolution genomic profiles define distinct clinico-pathogenetic subgroups of multiple myeloma patients. Cancer Cell 9: 313– 325 (2006). Cheng KW, Lahad JP, Kuo WL, Lapuk A, Yamada K, et al: The RAB25 small GTPase determines aggressiveness of ovarian and breast cancers. Nat Med 10: 1251–1256 (2004). Chng WJ, Kumar S, Vanwier S, Ahmann G, PriceTroska T, et al: Molecular dissection of hyperdiploid multiple myeloma by gene expression profiling. Cancer Res 67: 2982–2989 (2007). Chubb JR, Bickmore WA: Considering nuclear compartmentalization in the light of nuclear dynamics. Cell 112:403–406 (2003). Cigudosa JC, Rao PH, Calasanz MJ, Odero MD, Michaeli J, et al: Characterization of nonrandom chromosomal gains and losses in multiple myeloma by comparative genomic hybridization. Blood 91:3007–3010 (1998). Condomines M, Hose D, Raynaud P, Hundemer M, De Vos J, et al: Cancer/testis genes in multiple myeloma: expression patterns and prognosis value determined by microarray analysis. J Immunol 178: 3307–3315 (2007). Cremer M, Kupper K, Wagler B, Wizelman L, von Hase J, et al: Inheritance of gene density-related higher order chromatin arrangements in normal and tumor cell nuclei. J Cell Biol 162: 809– 820 (2003). Cremer T, Cremer C: Chromosome territories, nuclear architecture and gene regulation in mammalian cells. Nat Rev Genet 2: 292–301 (2001).
Cytogenet Genome Res 118:320–327 (2007)
325
Croft JA, Bridger JM, Boyle S, Perry P, Teague P, et al: Differences in the localization and morphology of chromosomes in the human nucleus. J Cell Biol 145: 1119–1131 (1999). Dave SS, Fu K, Wright GW, Lam LT, Kluin P, et al: Molecular diagnosis of Burkitt’s lymphoma. N Engl J Med 354:2431–2442 (2006). Fabris S, Ronchetti D, Agnelli L, Baldini L, Morabito F, et al: Transcriptional features of multiple myeloma patients with chromosome 1q gain. Leukemia 21:1113–1116 (2007). Fandy TE, Carraway H, Gore SD: DNA demethylating agents and histone deacetylase inhibitors in hematologic malignancies. Cancer J 13: 40–48 (2007). Feinberg AP, Ohlsson R, Henikoff S: The epigenetic progenitor origin of human cancer. Nat Rev Genet 7:21–33 (2006). Fischle W, Wang Y, Allis CD: Histone and chromatin cross-talk. Curr Opin Cell Biol 15: 172–183 (2003). Fisher AG, Merkenschlager M: Gene silencing, cell fate and nuclear organisation. Curr Opin Genet Dev 12:193–197 (2002). Fonseca R, Barlogie B, Bataille R, Bastard C, Bergsagel PL, et al: Genetics and cytogenetics of multiple myeloma: a workshop report. Cancer Res 64:1546–1558 (2004). Fraga MF, Ballestar E, Villar-Garea A, Boix-Chornet M, Espada J, et al: Loss of acetylation at Lys16 and trimethylation at Lys20 of histone H4 is a common hallmark of human cancer. Nat Genet 37: 391–400 (2005). Frigola J, Song J, Stirzaker C, Hinshelwood RA, Peinado MA, et al: Epigenetic remodeling in colorectal cancer results in coordinate gene suppression across an entire chromosome band. Nat Genet 38: 540–549 (2006). Govin J, Caron C, Rousseaux S, Khochbin S: Testisspecific histone H3 expression in somatic cells. Trends Biochem Sci 30: 357–359 (2005). Grewal SI, Jia S: Heterochromatin revisited. Nat Rev Genet 8:35–46 (2007). Grogan JL, Mohrs M, Harmon B, Lacy DA, Sedat JW, et al: Early transcription and silencing of cytokine genes underlie polarization of T helper cell subsets. Immunity 14:205–215 (2001). Gutierrez NC, Garcia JL, Hernandez JM, Lumbreras E, Castellanos M, et al: Prognostic and biologic significance of chromosomal imbalances assessed by comparative genomic hybridization in multiple myeloma. Blood 104: 2661– 2666 (2004). Hake SB, Allis CD: Histone H3 variants and their potential role in indexing mammalian genomes: the ‘H3 barcode hypothesis’. Proc Natl Acad Sci USA 103: 6428–6435 (2006). Hanamura I, Stewart JP, Huang Y, Zhan F, Santra M, et al: Frequent gain of chromosome band 1q21 in plasma-cell dyscrasias detected by fluorescence in situ hybridization: incidence increases from MGUS to relapsed myeloma and is related to prognosis and disease progression following tandem stem-cell transplantation. Blood 108:1724–1732 (2006). Hattinger CM, Potschger U, Tarkkanen M, Squire J, Zielenska M, et al: Prognostic impact of chromosomal aberrations in Ewing tumours. Br J Cancer 86: 1763–1769 (2002). Hummel M, Bentink S, Berger H, Klapper W, Wessendorf S, et al: A biologic definition of Burkitt’s lymphoma from transcriptional and genomic profiling. N Engl J Med 354:2419–2430 (2006). Itoyama T, Nanjungud G, Chen W, Dyomin VG, Teruya-Feldstein J, et al: Molecular cytogenetic analysis of genomic instability at the 1q12–22 chromosomal site in B-cell non-Hodgkin lymphoma. Genes Chromosomes Cancer 35: 318– 328 (2002).
326
Kosak ST, Skok JA, Medina KL, Riblet R, Le Beau MM, et al: Subnuclear compartmentalization of immunoglobulin loci during lymphocyte development. Science 296: 158–162 (2002). Kuppers R, Klein U, Hansmann ML, Rajewsky K: Cellular origin of human B-cell lymphomas. N Engl J Med 341:1520–1529 (1999). Le Baccon P, Leroux D, Dascalescu C, Duley S, Marais D, et al: Novel evidence of a role for chromosome 1 pericentric heterochromatin in the pathogenesis of B-cell lymphoma and multiple myeloma. Genes Chromosomes Cancer 32:250–264 (2001). Liebisch P, Wendl C, Wellmann A, Krober A, Schilling G, et al: High incidence of trisomies 1q, 9q, and 11q in multiple myeloma: results from a comprehensive molecular cytogenetic analysis. Leukemia 17:2535–2537 (2003). Lu YJ, Hing S, Williams R, Pinkerton R, Shipley J, et al: Chromosome 1q expression profiling and relapse in Wilms’ tumour. Lancet 360:385–386 (2002). Lund AH, van Lohuizen M: Epigenetics and cancer. Genes Dev 18: 2315–2335 (2004). Mellor J: It takes a PHD to read the histone code. Cell 126:22–24 (2006). Mendrzyk F, Korshunov A, Benner A, Toedt G, Pfister S, et al: Identification of gains on 1q and epidermal growth factor receptor overexpression as independent prognostic markers in intracranial ependymoma. Clin Cancer Res 12: 2070–2079 (2006). Metz A, Soret J, Vourc’h C, Tazi J, Jolly C: A key role for stress-induced satellite III transcripts in the relocalization of splicing factors into nuclear stress granules. J Cell Sci 117: 4551–4558 (2004). Moreaux J, Hose D, Reme T, Jourdan E, Hundemer M, et al: CD200 is a new prognostic factor in multiple myeloma. Blood 108: 4194–4197 (2006). Narita M, Nunez S, Heard E, Lin AW, Hearn SA, et al: Rb-mediated heterochromatin formation and silencing of E2F target genes during cellular senescence. Cell 113: 703–716 (2003). Nightingale KP, O’Neill LP, Turner BM: Histone modifications: signalling receptors and potential elements of a heritable epigenetic code. Curr Opin Genet Dev 16: 125–136 (2006). Offit K, Jhanwar SC, Ladanyi M, Filippa DA, Chaganti RS: Cytogenetic analysis of 434 consecutively ascertained specimens of non-Hodgkin’s lymphoma: correlations between recurrent aberrations, histology, and exposure to cytotoxic treatment. Genes Chromosomes Cancer 3:189– 201 (1991). Oshiro A, Tagawa H, Ohshima K, Karube K, Uike N, et al: Identification of subtype-specific genomic alterations in aggressive adult T-cell leukemia/lymphoma. Blood 107: 4500–4507 (2006). Parker MJ, Licence S, Erlandsson L, Galler GR, Chakalova L, et al: The pre-B-cell receptor induces silencing of VpreB and lambda5 transcription. EMBO J 24: 3895–3905 (2005). Rosinol L, Carrio A, Blade J, Queralt R, Aymerich M, et al: Comparative genomic hybridisation identifies two variants of smoldering multiple myeloma. Br J Haematol 130: 729–732 (2005). Ryan RF, Schultz DC, Ayyanathan K, Singh PB, Friedman JR, et al: KAP-1 corepressor protein interacts and colocalizes with heterochromatic and euchromatic HP1 proteins: a potential role for Kruppel-associated box-zinc finger proteins in heterochromatin-mediated gene silencing. Mol Cell Biol 19: 4366–4378 (1999).
Cytogenet Genome Res 118:320–327 (2007)
Sambani C, La Starza R, Pierini V, Vandenberghe P, Gonzales-Aguilera JJ, et al: Leukemic recombinations involving heterochromatin in myeloproliferative disorders with t(1;9). Cancer Genet Cytogenet 162: 45–49 (2005). Saurin AJ, Shiels C, Williamson J, Satijn DP, Otte AP, et al: The human polycomb group complex associates with pericentromeric heterochromatin to form a novel nuclear domain. J Cell Biol 142:887–898 (1998). Sawyer JR, Tricot G, Mattox S, Jagannath S, Barlogie B: Jumping translocations of chromosome 1q in multiple myeloma: Evidence for a mechanism involving decondensation of pericentromeric heterochromatin. Blood 91: 1732–1741 (1998). Sawyer JR, Tricot G, Lukacs JL, Binz RL, Tian E, et al: Genomic instability in multiple myeloma: evidence for jumping segmental duplications of chromosome arm 1q. Genes Chromosomes Cancer 42: 95–106 (2005). Sewalt RG, Lachner M, Vargas M, Hamer KM, den Blaauwen JL, et al: Selective interactions between vertebrate polycomb homologs and the SUV39H1 histone lysine methyltransferase suggest that histone H3-K9 methylation contributes to chromosomal targeting of Polycomb group proteins. Mol Cell Biol 22: 5539–5553 (2002). Shaughnessy Jr JD, Zhan F, Burington BE, Huang Y, Colla S, et al: A validated gene expression model of high-risk multiple myeloma is defined by deregulated expression of genes mapping to chromosome 1. Blood 109: 2276–2284 (2006). Shi S, Calhoun HC, Xia F, Li J, Le L, et al: JAK signaling globally counteracts heterochromatic gene silencing. Nat Genet 38: 1071–1076 (2006). Spector DL: The dynamics of chromosome organization and gene regulation. Annu Rev Biochem 72:573–608 (2003). Stange DE, Radlwimmer B, Schubert F, Traub F, Pich A, et al: High-resolution genomic profiling reveals association of chromosomal aberrations on 1q and 16p with histologic and genetic subgroups of invasive breast cancer. Clin Cancer Res 12:345–352 (2006). Staudt LM: Molecular diagnosis of the hematologic cancers. N Engl J Med 348:1777–1785 (2003). Strahl BD, Allis CD: The language of covalent histone modifications. Nature 403: 41–45 (2000). Stransky N, Vallot C, Reyal F, Bernard-Pierrot I, de Medina SG, et al: Regional copy number-independent deregulation of transcription in cancer. Nat Genet 38: 1386–1396 (2006). Su RC, Brown KE, Saaber S, Fisher AG, Merkenschlager M, et al: Dynamic assembly of silent chromatin during thymocyte maturation. Nat Genet 36:502–506 (2004). Taddei A, Roche D, Bickmore WA, Almouzni G: The effects of histone deacetylase inhibitors on heterochromatin: implications for anticancer therapy? EMBO Rep 6:520–524 (2005). Tagawa H, Suguro M, Tsuzuki S, Matsuo K, Karnan S, et al: Comparison of genome profiles for identification of distinct subgroups of diffuse large B-cell lymphoma. Blood 106: 1770–1777 (2005). Talbert PB, Henikoff S: Spreading of silent chromatin: inaction at a distance. Nat Rev Genet 7:793– 803 (2006). Tanabe H, Muller S, Neusser M, von Hase J, Calcagno E, et al: Evolutionary conservation of chromosome territory arrangements in cell nuclei from higher primates. Proc Natl Acad Sci USA 99: 4424–4429 (2002).
Tomlinson GE, Douglass EC, Pollock BH, Finegold MJ, Schneider NR: Cytogenetic evaluation of a large series of hepatoblastomas: numerical abnormalities with recurring aberrations involving 1q12-q21. Genes Chromosomes Cancer 44: 177–184 (2005). Turner BM: Cellular memory and the histone code. Cell 111:285–291 (2002). van Dekken H, Wink JC, Vissers KJ, van Marion R, Koppert LB, et al: Genomic analysis of early adenocarcinoma of the esophagus or gastroesophageal junction: tumor progression is associated with alteration of 1q and 8p sequences. Genes Chromosomes Cancer 45: 516–525 (2006).
Vourc’h C, Taruscio D, Boyle AL, Ward DC: Cell cycle-dependent distribution of telomeres, centromeres, and chromosome-specific subsatellite domains in the interphase nucleus of mouse lymphocytes. Exp Cell Res 205:142–151 (1993). Walter J, Schermelleh L, Cremer M, Tashiro S, Cremer T: Chromosome order in HeLa cells changes during mitosis and early G1, but is stably maintained during subsequent interphase stages. J Cell Biol 160: 685–697 (2003).
Zhan F, Huang Y, Colla S, Stewart JP, Hanamura I, et al: The molecular classification of multiple myeloma. Blood 108:2020–2028 (2006a). Zhan F, Barlogie B, Arzoumanian V, Huang Y, Williams DR, et al: A gene expression signature of benign monoclonal gammopathy evident in multiple myeloma is linked to good prognosis. Blood 109:1692–1700 (2006b). Zudaire I, Odero MD, Caballero C, Valenti C, Martinez-Penuela JM, et al: Genomic imbalances detected by comparative genomic hybridization are prognostic markers in invasive ductal breast carcinomas. Histopathology 40: 547–555 (2002).
Cytogenet Genome Res 118:320–327 (2007)
327
This finding, in a morphological sense, mirrors one of the key biological features of follicular lymphoma: the circumvention of programmed cell death (apoptosis). In reactive follicles lymphoid cells with unsuccessful gene rearrangements generated by the somatic hypermutation process are readily eliminated by programmed cell death (apoptosis). This intriguing feature (the morphological equivalent of which is the presence of ‘tingible body’ macrophages), next to other processes, involves downregulation of the expression of the anti-apoptotic protein BCL-2 in reactive follicles, in which it is, in contrast to the perifollicular mantle, marginal zone and T cell areas, virtually not expressed. The formation of a cell with a t(14; 18)(q32;q21)/ IGH-BCL2 rearrangement, the prototypic chromosome translocation in roughly 85% of FL, during VDJ rearrangement of the immunoglobulin (B cell) receptor in early B cell stages, leads to the formation of cell clones with constitutive deregulation and overexpression of BCL2 (Rowley, 1988; Horsman et al., 1995). Upon entry of these cells into the follicles of peripheral lymphoid organs, the expression of BCL2 cannot be downregulated, and the respective cells are prone to survival, expansion and, ultimately, to colonization of other follicles (Sanchez-Beato et al., 2003). The occurrence of ‘ongoing’ mutations within tumour subclones is a unique feature of germinal center cells due to the somatic hypermutation process inferred by the follicular microenvironment. Laser microdissection (LMD) and micromanipulation strategies allow for molecular genetic analysis of single cell mutation patterns and hence, the analysis of tumour cell dissemination in FL (Oeschger et al., 2002). In this context, Matolcsy et al. (1999) investigated the intraclonal diversification occurring in the transformation from FL to diffuse large B cell lymphoma (DLBCL). Intriguingly, they found a cessation of hypermutation resulting in a unique sequence in all tumour cells of the DLBCL. Against this background, it is worth mentioning that compared to other low-grade B-cell lymphomas, FL is characterized by a rather high histomorphological variability, which is also reflected in a multifaceted spectrum of changes in morphology during frequently occurring relapses with changes in morphology of the tumour cells and grade of the lesion (Muller-Hermelink et al., 2001). In this study, we were interested in the analysis of changes in intraclonal diversity during FL progression occurring in a setting without transformation to a high-grade lymphoma. Furthermore, preserved antigenic selection was detected by analyzing the distribution of silent and replacement mutations in the complementarity determining regions (CDR) and framework regions (FR) of the B-cell receptor.
Material and methods Patients and tissue specimens Four patients with follicular lymphoma were included in the study (Table 1). Patient 1 was a 62-year-old male, patient 2 a 72-year-old male, patient 3 a 38-year-old female and patient 4 a 56-year-old female. In two patients (patient 1 and 2) a biopsy at one point of time was available (one case FL grade 1 and the other FL grade 2). From patient 3, two biopsies
Table 1. Clinical data of the four patients including the times until the relapses were diagnosed
Patient
Age Sex Localisation years
1 2 3 4
62 72 38 56
m m f f
2nd Relapse 1st Relapse (months after (months after diagnosis) diagnosis)
axilla (stage I) axilla (stage I) cervical (stage I) axilla (34) inguinal (stage I) cervical (21)
preauricular (32)
with an interval of 34 months (both FL grade 2) could be analysed, and from patient 4 the initial specimen (FL grade 2) and two further biopsies, after 21 months (FL grade 3A) and 32 months (FL grade 3A), respectively, after the initial diagnosis were included in the study.
Histology, immunohistochemistry, and microdissection of follicles For lymphoma diagnosis 3-m sections of paraffin-embedded tissues of the four patients were stained with H&E, periodic acid-Schiff’s reagent (PAS) and Giemsa. Immunostaining was performed with antibodies directed against CD20, BCL-2, Ki67, Ig light chains kappa/ lambda (all DakoCytomation, Hamburg, Germany), CD3, CD5, CD10, and CD23 (all Novocastra Laboratories, Newcastle upon Tyne, UK). Tumour classification and grading was done according to the World Health Organisation (WHO) classification of tumours of hematopoietic and lymphoid tissues (Jaffe et al., 2001). DNA extraction and detection of VH families used by the tumours For each case, ten sections of cryopreserved tumour tissues, each 20 m thick, were incubated overnight with 50 l DNA extraction buffer and 2.5 l proteinase K solution at 55 ° C. Proteinase K was then inactivated at 95 ° C for 10 min and insoluble components were removed by centrifugation (14,000 rpm, 5 min). Extracted DNA was dissolved in H2O at a concentration of 100 g/ml. 5 l of this solution were used in the subsequent PCR reactions. 12 separate PCR assays were performed: The primers specific for the six different VH families described by Campbell et al. (1992) and those according to Küppers and associates (1993) were applied. The former were combined with JHa(19)-primers as 3´-consensus primers, the latter with JHa(24)-primers, respectively. Amplificates were visualised after gel electrophoresis. Microdissection For microdissection of follicles, silanised glass slides were covered with a PEN-foil. Subsequently 6–8-m sections from cryopreserved tissues were spread on the foil. Reactive T-cells were stained with mouse anti-CD3 antibody (LEU 4, Beckman Coulter, Krefeld, Germany). After anti-CD3 staining, a biotinylated goat anti-mouse secondary antibody was applied (DakoCytomation, Hamburg, Germany), followed by an incubation with Streptavidin-bound alkaline phosphatase (StreptAB complex/AP, DakoCytomation, Hamburg, Germany). Bound enzyme was visualized with DAB. CD3 negative follicles were dissected using a UV-laser beam (Palm Robot-Micro Beam, Wolfratshausen, Germany). Of every case up to ten follicles were microdissected (Fig. 1). Using a pipette tip each dissected follicle was transferred to a separate PCR reaction tube and incubated with DNA extraction buffer and proteinase K (both Roche, Mannheim, Germany). After protein digestion, DNA was extracted and dissolved in H2O resulting in a concentration of 100 g/ml. 5 l of this solution were used in the subsequent family specific amplification reaction. VH family specific PCR For amplification of rearranged immunoglobulin heavy chain genes of the tumour cells, a two-step nested PCR strategy was employed. In the first round a set of six, non-degenerate oligonucleotide
Cytogenet Genome Res 118:328–336 (2007)
329
Table 2. Overall mutation frequency in the tumour cell populations
PaPatient 1 tient 2
Grade 1 2 Average mutation frequency 10.7% 10.8%
Patient 3
Patient 4
1st biopsy
2nd biopsy
1st 2nd 3rd biopsy biopsy biopsy
2
2
2
9.8%
4.2%
4.6% 7.1% 9.8%
3a
3a
a
VH family specific primers, (Campbell et al., 1992) combined with JHa(19)-primers as 3ⴕ-consensus primers was used, and in the second round, the primer set described by Küppers and associates (1993) was combined with JHa(24)-3ⴕ-primers.
Cloning of PCR products Nested PCR products were gel purified and extracted using a QIAEX II쏐 Gel Extraction Kit (Qiagen, Hilden, Germany). Using the TOPO TA-Cloning Kit for Sequencing쏐 (Invitrogen, Groningen, Netherlands) PCR products were cloned in a pCR4-TOPO쏐 vector according to the manufacturer’s instructions. Subsequently, OneShot쏐 TOP10 chemically competent Escherichia coli were transfected with the ligation products. Cultures were plated on ampicillin agar plates and incubated overnight at 37 ° C. The following day, nine colonies per follicle were transferred into ampicillin-containing LB-media for an overnight culture. Plasmid preparation Short-time bacteria cultures were centrifuged, resuspended in GET(Glucose/EDTA/Tris/HCl) solution, incubated with lysis buffer and sodium-acetate and again centrifuged. The supernatant was transferred into a new tube and plasmid DNA was precipitated with ethanol and air dried. Resulting pellets were resuspended in TE containing 20 g/ml RNAse. Successful cloning was subsequently checked by restriction digestion of 4 l plasmid DNA solution with 2.5 U EcoRI each. Resulting fragments were visualized by gel electrophoresis. Mutation analysis and calculation of phylogenetic trees For sequencing the VH-containing plasmids the ABI PRISMTM Dye Cycle Sequencing Ready Reaction Kit (Perkin Elmer, Weiterstadt, Germany) was used according to the manufacturer’s protocol. After PCR amplification and purification via a Sephadex G50 column, the products were loaded on a sequencing gel and base sequences were analysed in an ABI PRISMTM 377 DNA Sequencer (Perkin Elmer, Weiterstadt, Germany). Using NCBI BLAST (http://www.ncbi.nlm.nih. gov/BLAST/) and DNAPlot (http://www.dnaplot.de/input/human_v. html) programs, obtained VH-gene sequences were compared with all germline sequences to identify the VH segment with the highest homology, referred to as the tumour ‘stem cell’. For calculation of the genealogical trees the gene sequences of the individual tumour cell were compared to that of the ‘stem cell’ and similarity of gene sequence was depicted as relative distance in the figure, i.e. cells with few different base pairs were localised closer to each other than cells with higher sequence divergence.
Results
Clonal evolution patterns in tumour progression In all four cases the mutation rate by far exceeded the empiric error rate of the Taq polymerase used (lower than 2 ! 10 –4). Of patient 3, two separate biopsies were evaluat-
330
Cytogenet Genome Res 118:328–336 (2007)
b Fig. 1. After anti-CD3 staining, negative follicle centers were dissected using a UV-laser beam ((a) before, (b) after microdissection).
ed: a cervical primary tumour (FL grade 2) and an axillary relapse (FL grade 2) diagnosed 34 months later. The Ig VH gene sequences of the two tumours were 89.9%–90.2% and 94.9% –97.5%, respectively, identical to that of the VH 4–34.1 germ-line gene. ‘Ongoing’ mutations were detectable both in the primary tumour and also in the relapse. In sequence comparison, the two tumours shared only two base pair mutations (Fig. 2) which means that the clonal mutation pattern of the primary tumour (tumour 3.1) was not found in the relapse (tumour 3.2). The putative progenitor cell of the primary FL harbored 28 somatic point mutations as compared to the germ line sequence, whereas the progenitor cell of the relapsed FL revealed 11 somatic point mutations. Thus, the overall mutation frequency in the tumour cell population of the relapse (4.2%) was distinctly lower than that of the primary tumour (9.8%) (Table 2). The second longitudinal analysis (patient 4) consisted, next to the primary tumour, of a first lymph node relapse after 21 months and of a second relapse after 32 months. The Ig VH gene sequences of the tumours were 93.2%–99.1% (tumour 4.1), 90.2%–95.7% (tumour 4.2) and 90.2% (tumour 4.3) respectively, identical to that of the VH 4–34.1 germ-line gene. In the neoplastic population of the primary tumour (tumour 4.1), ‘ongoing’ mutations were detectable in contrast to both relapses. The relapsed tumours (tumors 4.2 and 4.3) also presented with a different morphology: both were
Fig. 2. Sequence comparison of patient 3: Two separate biopsies were evaluated: the primary tumour and a relapse diagnosed 34 months later. ‘Ongoing’ mutations were detectable both in the primary tumour and also in the relapse. In sequence comparison, the two tumours shared only two base pair mutations which means that the clonal mutation pattern of the primary tumour (red arrow) was not found in the relapse (green arrows).
FR.II VH4-34.1
Migration pattern of tumour cells in FL As a second approach, possible differences in the migration pattern between tumours of various histomorphological grades and their alterations during tumour progression were evaluated. In the first case, an FL grade 1 (tumour 1), a pronounced clustering of tumour cells was seen in the follicles they were dissected from (Fig. 4: Genealogical tree tumour 1). However, a moderate exchange of tumour cells between follicles was found. In the second tumour, FL grade 2 (Tumour 2), a distinctly higher migration of tumour cells between follicles became obvious (Fig. 5: Genealogical tree tumour 2). Nonetheless, an at least rudimental clustering in distinct follicles was preserved. Both lymphomas showed a high mutation rate (both up to 11%), indicating a persisting influence of the hypermutation machinery. In keeping with these data, both the primary and relapsed tumours of case 3 (tumours 3.1 and 3.2), both FL grade 2, had preserved a rudimental clustering of their tumour cells (not shown). The distribution of tumour cells of the primary tumour in case 4 (tumour 4.1) showed a significant similarity to the other cases of FL grade 2 (tumours 2, 3.1 and 3.2): A certain clustering in particular follicles was preserved, but there was an extensive interfollicular migration. The two relapses (tumours 4.2 and 4.3), in contrast, presented with a merely diffuse growth pattern of the tumour cells, so an assignment to distinct follicular compartments was not possible.
CDR.I
classified as FL grade 3A and showed in large areas a merely diffuse growth pattern. Molecular genetic analysis of the first relapse (tumour 4.2) revealed, in a subgroup of four tumour cells (VIII-1, VIII-5, VIII-6 and VIII-7), a nearly identical mutation pattern as compared to the primary tumour. Three of these cells showed an identical additional base pair mutation, strongly pointing at a clonal evolution in these tumour cells. Another subgroup of five tumour cells of the same VH family as the primary tumour showed an entirely different mutation pattern (Fig. 3). A possible ‘hybrid’ cell with features of both tumour cell populations, and hence indicating a clonal relation, was not detectable. Mutational analysis of the framework regions of this second cell population resulted in the detection of a significant negative selection (P = 0.012), whereas in the first cell population, no significant selection process could be demonstrated (P = 0.258).
Cytogenet Genome Res 118:328–336 (2007)
331
CDR.II
FR .III
VH4-34.1
Fig. 3. The second longitudinal analysis (patient 4) consisted of the primary tumour, a first lymph node relapse after 21 months and a second relapse after 32 months. In the neoplastic population of the primary tumour (yellow arrows), ‘ongoing’ mutations were detectable in contrast to both relapses. Molecular genetic analysis of the first relapse (green arrows) revealed, in a subgroup of four tumour cells (VIII-1,
Fig. 4. Genealogical tree of tumour 1. Similarity of gene sequence corresponds to relative distance in the figure, i.e. cells with few different base pairs are depicted closer to each other than cells with higher sequence divergence. The first number in the cell designations gives the follicle they were dissected from, the second number identifies the individual cell. In this case a pronounced clustering of tumour cells became obvious, according to the follicles they were dissected from. However, a moderate exchange of tumour cells among the follicles was detectable.
332
Cytogenet Genome Res 118:328–336 (2007)
VIII-5, VIII-6 and VIII-7), a nearly identical mutation pattern as compared to the primary tumour. Three of these cells showed an identical additional base pair mutation, strongly pointing to clonal evolution in the tumour cells. Another subgroup of five tumour cells of the same VH family as the primary tumour showed an entirely different mutation, paralleling that of the second relapse (red arrows).
generated code’ of DNA, occurrence of an S-mutation does not result in change of the amino acid encoded, so that the structural integrity of the not-antigen binding scaffold structures of the receptor is maintained. In the present study, a mutation pattern characteristic of a significant antigen dependent effect on the mutations was only found in the case of tumour 1, an FL grade 1. In all subsequent cases of higher histologic grades (2 and 3A), R/Sratios characteristic of a haphazard mutation distribution were encountered (Table 3).
Antigenic selection Antigen-dependent hypermutation is recognizable by a characteristic distribution of the mutations, following a typical pattern (Davi et al., 1996): Within the complementarity determining regions (CDR), replacement (R)-mutations prevail, i.e. a base mutation results in an exchange of the particular amino acid for a different one. The altered steric properties of the antigen binding site therefore lead to a modified specificity and binding strength of the B-cell receptor. In contrast, significantly more silent (S)-mutations occur within the framework regions (FR). Due to the ‘de-
Discussion
Follicular lymphoma accounts for roughly 40% of NonHodgkin-Lymphoma (NHL). In most cases, the normal architecture of the lymph node parenchyma is effaced and occupied by neoplastic follicles recognisable by well-elaborated morphological features and, in about 90% of cases, by atypical expression of the BCL-2 protein (Harris et al., 1999). It is still a matter of debate, how the neoplastic follicles in FL actually arise: by colonization (infiltration) of pre-existing reactive follicles, or by the formation of new – neoplastic – follicular structures. A study of Su and colleagues (2001) on FL found a reactive lymphoid population in most germinal centers pointing at a pre-existing reactive germinal center. In addition, Cong and associates (2002) recently described the condition of ‘in situ’ follicular lymphoma implying that at least initially, neoplastic follicles in FL are not generated de novo, but that neoplastic follicular cells home to and colonise reactive germinal centers. In this study, we analysed the gene sequences of rearranged immunoglobulin heavy chain genes in up to 46 tumour cells per lymph node in samples of FL including consecutive biopsies of the same patient. In all but one case of FL grades 1 and 2 ‘ongoing’ mutations were detected in primary samples as well as in relapses, whereas in the two re-
Fig. 5. Genealogical tree of tumour 2. Compared to tumour 1 (Fig. 4) a markedly increased exchange of tumour cells became obvious. Nonetheless an at least rudimental clustering in distinct follicles was preserved.
Table 3. Antigen-dependent hypermutation is characterised by prevailing of replacement (R)-mutations within the complementarity determining regions (CDR), whereas within the framework regions (FR) significantly more silent (S)-mutations occur (see text). In the present study, a mutation pattern characteristic of a significant antigen dependent effect on the mutations was only found in the case of tumour 1, an FL grade 1. In all subsequent cases of higher histologic grades (2 and 3A), R/S-ratios characteristic of a haphazard mutation distribution were encountered.
Patient 1
Patient 2
Patient 3
Patient 4
1st biopsy
2nd biopsy
1st biopsy
2nd biopsy
3rd biopsy
Grade
1
2
2
2
2
3a
3a
R/S-Ratio (CDR) expected R/S-Ratio (CDR) found P
3.4 4.9 0.04
3.7 0.4 0.71
3.0 4.0 0.23
2.0 2.0 0.78
1.5 2.0 0.35
2.0 2.5 0.26
2.0 2.4 0.37
R/S-Ratio (FR) expected R/S-Ratio (FR) found P
3.3 1.5 0.03
3.3 2.4 0.19
3.00 2.00 0.15
3.2 1.4 0.24
3.0 1.5 0.29
2.9 1.3 0.13
3.0 1.3 0.21
Cytogenet Genome Res 118:328–336 (2007)
333
lapses of patient 4 (tumours 4.2 and 4.3, both FL grade 3A) a monoclonal cell population without intraclonal diversification was found. These findings suggest that the hypermutation process effective in the germinal center microenvironment may become less important within relapses or with increasing grade, even outside the setting of transformation to a high-grade lymphoma. Matolcsy et al. (1999) described this phenomenon of loss of hypermutations in a diffuse large B-cell lymphoma (DLBCL) evolving by high-grade transformation from a pre-existing FL. In contrast, an FL grade 2 from our series (tumour 3.1), which also relapsed as grade 2 FL, revealed a striking degree of intraclonal divergence, indicating a conserved hypermutation effect. In this context, it is interesting to note that the overall mutational frequency in the tumour cell population of the relapsed lymphoma (4–5%) was much lower than that of the primary tumour (10%). These data correspond well with that of other studies describing a reduction of the mutation rate in relapsed tumours when compared to the corresponding primary lymphoma (Bahler and Levy, 1992; Ottensmeier et al., 1998), a finding that may suggest a reduced mutational capability of the tumour cells during progression. This view is supported by the findings of Aarts and colleagues (2000), who described a lost ability for hypermutation or class switch of immunoglobulin heavy chain genes during progression of FL. In our FL grade 1 and 2 samples with sustained ongoing mutations, the genetic relationship of the different tumour cells was estimated by comparing their mutation patterns. The data obtained allowed for the construction of genealogical trees from which the way of migration of the tumour cells between follicles becomes apparent. Under physiological conditions, the migration rate of lymphocytes between germinal centers in lymph nodes is low (Jacob et al., 1991). In contrast, an extensive exchange of tumour cells between different follicles became apparent in the neoplastic germinal center cells analysed in this study. In an FL grade 1, however, the tumour cells in their majority clustered according to the follicles they were dissected from (Fig. 4: Genealogical tree tumour 1), with only moderate exchange of tumour cells between follicles perceptible. In contrast, all FL grade 2 tumours displayed a marked increase in interfollicular migration resulting in a decreased follicular clustering of tumour cells. This finding may be indicative of a direct correlation between the histomorphological grade with the rate of interfollicular tumour cell exchange on a molecular level. The two relapsed lymphomas of patient 4 (tumours 4.2 and 4.3) both showed a different histological growth pattern: both were classified as FL grade 3A with an extensive diffuse growth pattern. Such changes in morphology are often observed in the progression of FL. In roughly one third of FL, a transformation to a diffuse large B-cell lymphoma (DLBCL) occurs (Muller-Hermelink et al., 2001), which was not the case within the patient cohort presented here. The clinical impact of a predominantly diffuse growth pattern in FL is still under discussion (Bastion et al., 1991; Bartlett et al., 1994; Miller et al., 1997).
334
Cytogenet Genome Res 118:328–336 (2007)
Taking into account the fact that the tumour cells of the second relapse all harbored the same mutation pattern as a tumour cell subgroup of the first relapse, a survival advantage of this population can be assumed. Possibly, the conserved negative selection mechanism contributed to an increased assertiveness of these cells. The question as to which extent a conservation of structure and functionality of the B-cell receptor is needed for clonal expansion of the tumour, is not yet finally answered. Whilst an intact B-cell receptor formerly used to be regarded as a prerequisite for tumour growth, this assumption has been doubted by others (Ottensmeier et al., 1998; Noppe et al., 1999; Aarts et al., 2000). In addition, the second identified subclone of tumour cells was significantly more mutated than the first, indicating a conserved effectiveness of the hypermutation machinery. It is now a known fact that hypermutations do not only take place within the VH regions of the immunoglobulin genes, but also at different other chromosomal sites, possibly involving protooncogenes and tumour suppressor genes. In several studies frequent mutations in protooncogenes in FL were found (Yano et al., 1992; Sander et al., 1993; Matolcsy et al., 1996). The present concept of tumourigenesis of follicular lymphoma postulates that the tumour cells retain and express characteristics of their physiologic counterpart (germinal center cell). Against this background, the finding of continuously emerging base mutations (the so-called ‘ongoing mutations’) in the B-cell receptor genes of the tumour cells indicates a preserved influence of the hypermutation machinery. This mutation process physiologically guarantees the broad variety of the antibody repertoire in the organism. Several studies identified ongoing mutations in FL as highly characteristic features of the germinal center derivation of the tumour cells (Cleary et al., 1986; Bahler and Levy, 1992). During tumour progression, narrowing down of the mutation variation width occurs (Aarts et al., 2000), and ongoing mutations suspend. Biologically the tumour cells deprive themselves from the germinal center signature. In keeping with these findings, ongoing mutations were detectable in all primary tumours in the present study. At first relapse of patient 3, 34 months after initial diagnosis, no ongoing mutations were found. In patient 4 as well, ongoing mutations failed to be detectable in both relapsed tumours, 21 and 32 months after initial diagnosis. Such an intraclonal evolution from heterogeneity to homogeneity is a typical finding in the relapsed tumours of FL: a single tumour cell population which seems to have lost the susceptibility for ongoing mutations prevails (Zhu et al., 1994). Similar observations have been made during the transformation of FL to DLBCL (Zelenetz et al., 1991; Matolcsy et al., 1999). A possible influence of therapy mostly consisting of monoagent or multiagent chemotherapy is under discussion. The physiologic counterpart of the FL tumour cell, the germinal center cell, is able to circumvent apoptosis only by antigen-dependent selection, a mechanism which is used for the optimization of antibody affinity to the correspond-
ing antigen. Ziegner et al. (1994) demonstrated different reactions in germinal centers following antigenic challenge. In addition to the establishment of germinal centers harbouring hypermutated B-cells with highly affine B-cell receptors, germinal centers with no or only very low mutated B-cells were found. This heterogeneity of antigen dependent and obviously independent mutated cells might also be the case in FL. Moreover, B-cell receptor functionality does not seem to be an essential prerequisite for the expansion of a tumour clone. Examples for this are the description of surface Ignegative FL, or of those with a stop codon at the VH-gene locus not only in the primary tumour, but also at relapse (Ottensmeier et al., 1998; Noppe et al., 1999). In a study of 55 FL, Noppe and associates (1999) found evidence of antigenic selection in only 30% of their cases. In multiple myeloma as well, a prototypic postgerminal lymphoma whose origin from a mature B-cell would strongly suggest antigen-selected hypermutation, only 20% of cases harbored a significantly higher number of R-mutations in the CDRs than predicted by chance, indicating antigen influence on the distribution of the mutations (Bakkus et al., 1992; Vescio et al., 1995). In the present study, a statistically significant (P = 0.03) antigenic selection could be demonstrated only in tumour 1. R/S-ratios are so low, however, that also in the other tumours a preserved functionality of the receptor can be assumed. Fewer R-mutations than predicted by chance were found in the framework regions (FR) essential for the steric
structure of the B-cell receptor in all the cases of this study. These data are well in line with a study of Matolcsy et al. (1999) where an antigenic selection in FL and related DLBCL was found. Although in general the possibility that some of the mutations had happened before clonal expansion of the tumour must be considered, it is reasonable to assume that the preserved B-cell receptor functionality points at an ongoing selection. In conclusion, in the present study, we found evidence of a distinct influence of the germinal center microenvironment on clonal evolution and tumour cell distribution/migration pattern in FL. Most interestingly, we obtained data on the reduction of clonal diversification and selection of subclones in relation to higher tumour grades, also outside the setting of transformation to high-grade lymphoma. Antigen-dependent hypermutation was only seen in an FL grade 1, while in progressed FL random mutation patterns and a decrease in clonal diversity were found. Our data, therefore, suggest that alterations in the mechanisms of somatic hypermutation, that have been described during high-grade transformation in FL, can already be detected during tumour progression. Acknowledgments The authors wish to thank Mrs. Eva Werder and Mr. Erwin Schmitt for their excellent technical assistance.
References Aarts WM, Bende RJ, Steenbergen EJ, Kluin PM, Ooms EC, et al: Variable heavy chain gene analysis of follicular lymphomas: correlation between heavy chain isotype expression and somatic mutation load. Blood 95: 2922–2929 (2000). Adam P, Katzenberger T, Eifert M, Ott MM, Rosenwald A, et al: Presence of preserved reactive germinal centers in follicular lymphoma is a strong histopathologic indicator of limited disease stage. Am J Surg Pathol 29: 1661–1664 (2005). Bahler DW, Levy R: Clonal evolution of a follicular lymphoma: evidence for antigen selection. Proc Natl Acad Sci USA 89: 6770–6774 (1992). Bakkus MH, Heirman C, Van Riet I, Van Camp B, Thielemans K: Evidence that multiple myeloma Ig heavy chain VDJ genes contain somatic mutations but show no intraclonal variation. Blood 80: 2326–2335 (1992). Bartlett NL, Rizeq M, Dorfman RF, Halpern J, Horning SJ: Follicular large-cell lymphoma: intermediate or low grade? J Clin Oncol 12:1349– 1357 (1994). Bastion Y, Berger F, Bryon PA, Felman P, French M, Coiffier B: Follicular lymphomas: assessment of prognostic factors in 127 patients followed for 10 years. Ann Oncol 2 Suppl 2: 123–129 (1991). Campbell MJ, Zelenetz AD, Levy S, Levy R: Use of family specific leader region primers for PCR amplification of the human heavy chain variable region gene repertoire. Mol Immunol 29: 193–203 (1992).
Cleary ML, Meeker TC, Levy S, Lee E, Trela M, Sklar J, Levy R: Clustering of extensive somatic mutations in the variable region of an immunoglobulin heavy chain gene from a human B cell lymphoma. Cell 44: 97–106 (1986). Cong P, Raffeld M, Teruya-Feldstein J, Sorbara L, Pittaluga S, Jaffe ES: In situ localization of follicular lymphoma: description and analysis by laser capture microdissection. Blood 99: 3376– 3382 (2002). Davi F, Maloum K, Michel A, Pritsch O, Magnac C, et al: High frequency of somatic mutations in the VH genes expressed in prolymphocytic leukemia. Blood 88:3953–3961 (1996). Harris NL, Jaffe ES, Diebold J, Flandrin G, MullerHermelink HK, et al: World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissues: report of the Clinical Advisory Committee meeting – Airlie House Virginia November 1997. J Clin Oncol 17:3835–3849 (1999). Horsman DE, Gascoyne R, Coupland DRW, Coldman AJ, Adomat SA: Comparison of cytogenetic analysis, Southern analysis and polymerase chain reaction for the detection of t(14; 18) in follicular lymphoma. Am J Clin Pathol 103: 472–478 (1995). Jacob J, Kelsoe G, Rajewsky K, Weiss U: Intraclonal generation of antibody mutants in germinal centers. Nature 354: 389–392 (1991).
Jaffe ES, Harris NL, Stein H, Vardiman JW (eds): World Health Organization Classification of Tumours. Pathology and Genetics of Tumours of Haematopoietic and Lymphoid Tissues (IARC Press, Lyon 2001). Kuppers R, Zhao M, Hansmann ML, Rajewsky K: Tracing B cell development in human germinal centers by molecular analysis of single cells picked from histological sections. EMBO J 12: 4955–4967 (1993). Matolcsy A, Casali P, Warnke RA, Knowles DM: Morphologic transformation of follicular lymphoma is associated with somatic mutation of the translocated Bcl-2 gene. Blood 88: 3937– 3944 (1996). Matolcsy A, Schattner EJ, Knowles DM, Casali P: Clonal evolution of B cells in transformation from low- to high-grade lymphoma. Eur J Immunol 29:1253–1264 (1999). Miller TP, LeBlanc M, Grogan TM, Fisher RI: Follicular lymphomas: do histologic subtypes predict outcome? Hematol Oncol Clin North Am 11:893–900 (1997). Muller-Hermelink HK, Zettl A, Pfeifer W, Ott G: Pathology of lymphoma progression. Histopathology 38: 285–306 (2001). Noppe SM, Heirman C, Bakkus MH, Brissinck J, Schots R, Thielemans K: The genetic variability of the VH genes in follicular lymphoma: the impact of the hypermutation mechanism. Br J Haematol 107: 625–640 (1999).
Cytogenet Genome Res 118:328–336 (2007)
335
Oeschger S, Brauninger A, Kuppers R, Hansmann ML: Tumor cell dissemination in follicular lymphoma. Blood 99: 2192–2198 (2002). Ottensmeier CH, Thompsett AR, Zhu D, Wilkins BS, Sweetenham JW, Stevenson FK: Analysis of VH genes in follicular and diffuse lymphoma shows ongoing somatic mutation and multiple isotype transcripts in early disease with changes during disease progression. Blood 91: 4292– 4299 (1998). Rowley JD: Chromosome studies in the non-Hodgkin’s lymphomas: the role of the 14; 18 translocation. J Clin Oncol 6:919–925 (1988). Sanchez-Beato M, Sanchez-Aguilera A, Piris MA: Cell cycle deregulation in B-cell lymphomas. Blood 101:1220–1235 (2003).
336
Sander CA, Yano T, Clark HM, Harris C, Longo DL, et al: p53 mutation is associated with progression in follicular lymphomas. Blood 82: 1994– 2004 (1993). Su W, Spencer J, Wotherspoon AC: Relative distribution of tumour cells and reactive cells in follicular lymphoma. J Pathol 193: 498–504 (2001). Vescio RA, Cao J, Hong CH, Lee JC, Wu CH, et al: Myeloma Ig heavy chain V region sequences reveal prior antigenic selection and marked somatic mutation but no intraclonal diversity. J Immunol 155: 2487–2497 (1995), Yahalom J: Radiation therapy in the treatment of lymphoma. Curr Opin Oncol 11: 370–374 (1999).
Cytogenet Genome Res 118:328–336 (2007)
Yano T, Jaffe ES, Longo DL, Raffeld M: MYC rearrangements in histologically progressed follicular lymphomas. Blood 80: 758–767 (1992). Zelenetz AD, Chen TT, Levy R: Histologic transformation of follicular lymphoma to diffuse lymphoma represents tumor progression by a single malignant B cell. J Exp Med 173: 197–207 (1991). Zhu D, Hawkins RE, Hamblin TJ, Stevenson FK: Clonal history of a human follicular lymphoma as revealed in the immunoglobulin variable region genes. Br J Haematol 86: 505–512 (1994). Ziegner M, Steinhauser G, Berek C: Development of antibody diversity in single germinal centers: selective expansion of high-affinity variants. Eur J Immunol 24: 2393–2400 (1994).
sion of BCL2 alone is not sufficient for tumorigenesis and additional genetic events are required for the initiation of the tumor and its further progression (McDonnell and Korsmeyer, 1991). In keeping with this, virtually every FL exhibits a number of additional recurring abnormalities (Yunis et al., 1987; Armitage et al., 1988; Tilly et al., 1994; Whang-Peng et al., 1995; Juneja et al., 1997; Knutsen, 1997; Horsman et al., 2001, 2003). However, in more than 90% of the cases 10–60% of the chromosomal abnormalities remain undefined due to karyotypic complexity. The introduction of spectral karyotyping (SKY) has made possible the identification of each of the human chromosomes by a different color, facilitating precise identification of all rearrangements in a tumor karyotype (Schrock et al., 1996). SKY analysis in other B-cell lymphomas has revealed additional cytogenetic information in more than 90% of the cases including new recurring breakpoints, translocations and regions of gains and losses (Rao et al., 1998; Sawyer et al., 1998; Nanjangud et al., 2002). In an attempt to define the full spectrum of chromosomal abnormalities in FL, we analyzed a panel of 61 FLs with an abnormal G-banded karyotype by SKY. The SKY analysis was also augmented by fluorescence in situ hybridization (FISH) using probes for specific DNA sequences to further clarify or confirm the rearrangements.
Materials and methods Tumor ascertainment The ongoing ascertainment of consecutive NHL cases for cytogenetic analysis initiated at the MSKCC in 1984 has been described (Offit et al., 1991). A subset of 61 cases with histologic diagnosis of follicular lymphoma for which archived methanol-acetic acid fixed metaphase preparations were available, were selected for SKY analysis. All patients were histologically reclassified according to the World Health Organization classification and staged by the Ann Arbor staging system (Nathwani et al., 2001). Cases that were t(14;18)(q32;q21) negative were also stained by paraffin immunohistochemistry as previously described for BCL6, BCL2, CD10, CD43, and CCND1 to histologically confirm a follicle center origin (Hedvat et al., 2002). G-banding and SKY analysis G-banding analysis was performed on metaphase spreads obtained from lymph node biopsy (47), soft tissue (11), or spleen (3), as previously described (Offit et al., 1991). Clonal chromosomal abnormalities were described according to International System of Human Cytogenetic Nomenclature (ISCN, 1995). SKY analysis was performed as previously described (Rao et al., 1998). SKY images were acquired with an SD300 Spectracube (Applied Spectral Imaging, Migdal Ha-Emck, Israel) mounted on a Nikon Eclipse E800 microscope using a custom designed optical filter (SKY-1) (Chroma Technology, Brattleboro, VT, USA). For each case, 10 to 30 metaphases were analyzed. Breakpoints on G-banded chromosomes were scored according to previously established criteria (Cigudosa et al., 1999). Breakpoints on the SKY painted chromosomes were determined by comparison with corresponding inverted DAPI image and G-banded karyotype of the same tumor as previously described (Rao et al., 1998). Fluorescence in situ hybridization (FISH) analysis FISH with whole chromosome painting and alpha satellite probes (Vysis, Downers Grove, IL, USA) was performed according to the manufacturer-supplied protocols when necessary. To further confirm the presence of t(14; 18)(q32;q21), the non-chimeric PAC clones, 408b18
338
Cytogenet Genome Res 118:337–344 (2007)
(containing BCL2), and 120H10 (region immediately telomeric to IGHV ), were labeled with digoxigenin-11-dUTP and biotin-16-dUTP respectively by nick translation and hybridized as previously described (Rao et al., 1993). For each case, 10–15 metaphases were analyzed using the Nikon Eclipse E800 microscope equipped with the Quips PathVysion system (Applied Imaging, Santa Clara, CA, USA). A case showing unambiguous fusion signal in at least 2 metaphases and 15% interphase nuclei was considered positive for the translocation.
Statistical analysis of data Differences in percentage of variables were tested for significance using 2-tailed Fischer exact test. A probability of !0.05 was considered statistically significant. Cytogenetic variables included recurring break sites (16), gains (12) and losses (4) observed in 110%. Clinical correlations could not be performed due to the small number of cases with a recurring abnormality in each clinical subgroup.
Results
Clinical and histologic features The clinical and histologic features of the 61 follicular lymphomas are summarized in Table 1. The most common histologic subtype was grade 2 (36%) followed by grade 1 (33%) and grade 3a/3b (31%). Overall, a diffuse component was present in 48% of cases. The cohort comprised 30 males and 31 females. The median age at diagnosis was 59 years (range 30–83). Thirty-five (57%) patients presented with stage IV disease, 9 (15%) with stage I and III disease each and 6 (10%) with stage II disease. Two patients could not be staged. Of the 61 patients, 38 were at low-risk, 21 at high-risk and the risk could not be determined in the remaining two. Complete clinical data was not available in two patients and five patients were lost to follow up. Chromosomal instability Of the 61 patients analyzed by SKY, the G-banded karyotype of 17 (28%) remained unchanged while several new breakpoints and aberrations were identified in the remaining 44 patients (72%). Chromosomal instability as measured by the mean number of breaks and aberrations (numerical and structural) per case was greater in SKY analysis (8.1 and 6.3 respectively) than in G-banding analysis (5.1 and 5.7 respectively). All chromosomes were also noted to be affected more frequently in the SKY analysis than G-banding (Fig. 1). Involvement of chromosome 12 was significantly more frequent in SKY analysis (P = 0.04). Analysis of breakpoints Figure 2 shows the distribution of breakpoints by Gbanding and SKY analysis in the 61 cases. A total of 469 breakpoints were identified by SKY analysis compared to 325 by G-banding analysis. Of the 325 breakpoints identified by G-banding, SKY analysis confirmed 296 breakpoints and revised the remaining 29 breakpoints. In addition, SKY analysis revealed 144 new breakpoints. The 469 breakpoints identified by SKY were located at 130 bands and of these, 85 were recurring (62 cases). In the case of Gbanding, the 325 breakpoints were located at 106 bands and of these, 56 were recurring. Figure 3A compares the inci-
Percentage 0
20
40
60
6
7
80
6 1
100
2 4
1 2 3 4 5 6 7
1
2
3
4
1
3
1
1
5
6
7
8
9
10
11
12
2
3
1
16
17
18
1
8 9 10 11 12
P = 0.04
3
2
13
14
13 14 15 16
15
1
17 18
2
1
1
19 20
19
20
21
22
X
Y
21
Fig. 2. Idiogram showing distribution of breakpoints identified by G-banding (left of chromosome) and SKY (right of chromosome). The number of breakpoints in each chromosome that were identified by SKY, but could not be precisely assigned to a band are noted on top of the chromosomes. Each square represents a single breakpoint, star represents 10 breakpoints, triangle a single breakpoint misidentified by G-banding, and the plus sign represents a single breakpoint misidentified by G-banding but within the same chromosome.
22 X Y
Fig. 1. Involvement of chromosomes detected by SKY (black bars) versus G-banding (white bars). The difference was significant for chromosome 12.
Table 1. Clinical and histologic features of the 61 follicular lymphomas
Features
Number %
Features
Total Gender Male Female Age (Years) Median Range <60 >60 Histologic subtype Grade 1 Grade 2 Grade 3a/3b Architecture Follicular alone Follicular and diffuse Diffuse alone Performance status Not known 0 1 2 3 4 Lactate dehydrogenase Normal Elevated
61
Stage (Ann Arbor) Not known I II III IV B-Symptoms Not known Absent Present Extranodal sites <2 >2 Bone marrow Clinical risk-group Not known Low risk High risk Response Not known Complete remission Partial remission No response Status Lost to follow up Alive with disease No evidence of disease Died of disease
30 31
49.2 50.8
59 30–83 32 29
52.5 47.5
20 22 19
32.9 36.1 31.0
32 23 6
52.5 37.7 9.8
2 2 24 12 15 6
3.3 3.3 39.3 19.7 24.6 9.8
41 20
67.2 32.8
Cytogenet Genome Res 118:337–344 (2007)
Number % 2 9 6 9 35
3.3 14.8 9.8 14.8 57.4
2 47 12
3.3 77.0 19.7
46 15 31
75.4 24.6 50.8
2 38 21
3.3 62.3 34.4
5 38 10 8
8.2 62.3 16.4 13.1
5 30 13 13
8.2 49.2 21.3 21.3
339
A
B
Percentage 0
20
40
60
80
Percentage 0
100
1p11–13
+1/1q
1p36
+2p/q
1q11–21
+3p/q
20
40
60
80
100
+6p
2p11–13
+7/7q
3q27
+8/8q
6q11–15
+12/12q 6q21 +16 6q25–27 +17/17q 8q24
+18/18q
9p13
+21
10q22–24
+X
12q11–13
Fig. 3. Incidence of (A) recurring (110%) breakpoints and (B) chromosomal gains and losses, by SKY (black bars) versus G-banding (white bars).
del(1p)
14q32
del(6q)
17q11–21
del(10q)
18q21
del(17p)
Table 2. Frequency of IG gene sites and 3q27 involvement by SKY and G-banding
Table 3. Translocations affecting 3q27 other than those involving 14q32 and 22q11
Translocation
Case
Histology
Translocation/derivative
t(14;18)
2163 2241 2618 1125 2528 2093 2158 2034 2409
Grade 3 Grade 2 Grade 2 Grade 1 Grade 3 Grade 3 Grade 1 Grade 2 Grade 2
t(3;7)(q27;q32) t(3;3)(p23;q27) t(3;4)(q27;p12) t(3;8)(q27;q24) t(3;10)(q27;q11) der(3)t(3;19)(q27;?) der(3)t(3;10)(q27;p?)t(10;14)(p?;?) der(3)t(3;10)(p25;q22)t(3;5)(q27;p13) t(3;18)(q27;q21)
– + + + – – – + –
14q32 t(14;18)(q32;q21) t(3;14)(q27;q32) t(1;14)(q21;q32) other 14q32 overall 14q32 22q11 t(3;22)(q27;q11) t(8;22)(q24;q11) other 22q11 overall 22q11 2p11 t(2;18)(p11;q21) t(2;3)(p11;q27) other 2p11 overall 2p11 Overall IG gene siteb 3q27 other 3q27 overall 3q27
Number of cases (%) G-banding
SKY
41 (67.21) 4 (6.55) 3 (4.92) 2 (3.28) 47a (77.05)
42 (68.85) 4 (6.55) 3 (4.92) 2 (3.28) 48a (78.69)
1 (1.64) 1 (1.64) 1 (1.64) 3 (4.92)
1 (1.64) 1 (1.64) 1 (1.64) 3 (4.92)
1 (1.64) – 1 (1.64) 2 (3.28) 49 (80.33)
1 (1.64) 1 (1.64) 2 (3.28) 4 (6.55) 51 (83.61)
7 (11.48) 12 (19.67)
9 (14.75) 15 (24.59)
a
Both the homologs of chromosome 14 were involved in three cases. Two IG gene sites were affected in four cases and three cases in SKY and G-banding, respectively.
b
dence of recurring breakpoints (110%) by SKY and G-banding analysis. The six most frequent breakpoints in both SKY and G-banding analysis were 14q32, 18q21, 3q27, 1q11–q21, 6q11–q15 and 1p36. Nine additional sites were noted in 110% of cases in SKY analysis alone and in decreasing order of incidence and these were, 17q11–q21, 1p11–p13, 2p11–p13, 6q21, 8q24, 6q21, 9p13, 12q11–q13, 10q22–q24. Among these, breaks at 1p11–p13, and 2p11–p13 were highly underrepresented in the G-banding analysis. Rearrangements affecting these two sites were primarily unbalanced and while translocations affecting 2p11–p13 lead to partial gains of 2p, those affecting 1p11–p13 lead to partial loss of 1p. Analysis of structural abnormalities Structural abnormalities were noted in all the cases by both G-banding and SKY analysis and included transloca-
340
Cytogenet Genome Res 118:337–344 (2007)
Table 4. Other recurring translocations in this series of 61 follicular lymphomas
Case
1948 2383 2019 2337 2150 1411 1949 1389
Histology
Grade 2 Grade 3 Grade 1 Grade 3 Grade 2 Grade 2 Grade 2 Grade 3
tions, derived chromosomes, deletions, duplications, isochromosomes, inversions, insertions, additions and markers. G-banding analysis detected a total of 223 structural aberrations, of which 57 remained undefined (42 additions and 15 markers). All the undefined aberrations were resolved by SKY. In addition to identifying 55 new aberrations, SKY analysis revised 87/223 aberrations detected by G-banding. Of the marker and addition chromosomes detected by G-banding, 90% represented translocations. Among the additional structural abnormalities translocations were the most frequent (69%). Translocations affecting one of the IG gene sites were the most common ones and were detected at a slightly higher rate by SKY (84%) than by G-banding (80%) (Table 2). By SKY, 42 (71%) cases had t(14; 18)(q32;q21) or its variant t(2; 18)(p11;q21) (one case), four cases had t(3; 14) (q27;q32), three cases had t(1; 14)(q21;q32), and one case each had t(2;3) (p11;q27), t(3;22)(q27;q11), and t(8;22)(q24; q11) (Table 2). Five cases had other IG gene associated translocations. Translocations affecting 3q27 (25%) were the next most frequent after those affecting the IG gene sites (Table 2). Six cases exhibited t(3;14)(q27;q32) and it’s variants and the remaining 9 involved other partner chromosomes (Table 3). Only 12 of these were detected by Gbanding. Excluding the already established recurring translocations discussed above, a total of 125 translocations were identified of which 98% were observed in FLs with t(14;18)(q32;q21) or translocations of 3q27. Sites most frequently affected in these non-IG gene translocations included 1p11–p13, 1p36, 1q11–q21, 8q24, 9p13 and 17q11– q21 (10–16%). Among the 125 additional translocations, four were recurrent within this series of which two were identified by SKY alone (Table 4). Analysis of chromosomal gains and losses Whole chromosome or partial gains and losses were noted in 97% of the cases. Resolution of all the derivatives, additions and markers by SKY not only resulted in increased detection of gains and losses but also in their correct identification. Thus, 37% of the deletions identified by G-banding represented unbalanced translocations. Figure 3B compares the incidence of recurring (110%) gains and losses by G-banding and SKY and Fig. 4 shows the regions of gains and losses delineated by SKY analysis. Overall, gains (93%)
Translocation
t(14;18) t(3q27)
SKY
G-banding
der(1)t(1;1)(p36;q21) der(1)t(1;1)(p36;q21) der(1)t(1;2)(p11;p11) der(1)t(1;2)(p11;p11) der(1)t(1;2)(p36:q31) der(1)t(1;2)(p36;q31) t(7;8)(q22;q11.2) der(7)t(7;8)(q22;q13)
der(1)t(1;1)(p36;q21) der(1)t(1;1)(p36;q21) del(1)(p33) del(1)(p34) add(1)(p36) add(1)(p36) t(7;8)(q22;q11.2) der(7)t(7;8)(q22;q13)
3
2
2
+ + + – + – + +
– – + – – + – –
1 1
2
1
3
4
5
6 1
3
7
8 1
13
9
10
1
14
11
12
17
18
1
15
16
1
19
20
21
22
X
Y
Fig. 4. Idiogram showing distribution of gain and loss of genetic material by SKY analysis. The bars on the right of the chromosomes indicate gains and those on the left of the chromosomes indicate losses.
were more common than losses (71%) and all chromosomes in the complement were affected. Chromosomes most frequently affected by gains were 18 (38%), 1 and 12 (36% each), X (34%), 2 and 8 (26% each), 21 (22%), 7 (21%), 17 (16%), 3 and 16 (12% each) and 6p (10%). Eleven regions of common cytogenetic gain were identified: 1q11–q21, 1q23–q25, 1q32, 2p13–p21, 2q11–q21, 2q31–q37, 12q12–q15, 12q21–q24, 18q21, 21q11–q13, and 17q21–q25. Gains affecting chromosomes 7, 8, 16, and X tended to be whole chromosome gains. Among deletions, 6q deletions were the most frequent (30%) and four regions of common cytogenetic deletions (RCDs), in decreasing order of incidence, were observed at 6q21, 6q23, 6q13–q15 and 6q25–q27. Other chromosomes/regions involved in deletions were 1p32–p36 (13%), 10q22– q24 (12%), and 17p11–p13 (10%).
Cytogenet Genome Res 118:337–344 (2007)
341
A 0
20
Percentage 40
B 60
80
0
20
Percentage 40
60
80 p = 0.036
+1/1q
1p11–13
+2p/q
1p36
+3p/q 1q11–21
+6/6p
2p11–13
+7
3q27
+8/8q
p = 0.027
+12/12q
6q11–15
16 6q21 +17/17q 6q25–27
+18/18q
8q24
+21 +X
9p13
Fig. 5. Incidence of (A) recurring (110%) breakpoints and (B) chromosomal gains and losses in the t(14; 18)(q32;q21) positive (black bars) and negative subsets (dotted bars) of FL.
10q22–24
del(6q) 12q11–13
del(10q)
17q11–21
Table 5. Distribution summary of the recurring break sites, gains and losses observed in >10% in the histologic grades
del(17p)
Chromosome
Breakpoints
Gains
Losses
342
del(1p)
1p11-p13 1p36 1q11-q21 2p11-p13 3q27 6q11-q15 6q21 6q25-q27 8q24 9p13 10q22 10q24 12q11-q13 17q11-q21 1/1q 2p/q 3p/q 6/6p 7/7q 8/8q 12/12q 16 17/17q 18/18q 21 X 1p32-p36 6q11-q27 10q22-q24 17p13
Cytogenet Genome Res 118:337–344 (2007)
Number (%) Grade 1 (N = 20)
Grade 2 (N = 22)
Grade 3a/3b (N = 19)
3 (15) 3 (15) 3 (15) 6 (30) 6 (30) – – 2 (10) 2 (10) 2 (10) 1 (5) – 3 (15) 1 (5) 5 (30) 6 (30) 3 (15) 4 (20) 2 (10) 1 (05) 9 (50) 3 (15) 3 (15) 6 (30) 5 (25) 8 (40) 1 (5) 3 (15) 1 (5) 2 (10)
2 (9) 3 (14) 5 (23) 2 (9) 6 (27) 4 (18) 5 (23) 3 (14) – 2 (10) 4 (18) 3 (14) 2 (10) 5 (23) 9 (41) 5 (23) 2 (9) – 1 (5) 5 (23) 4 (18) – 3 (14) 9 (41) 4 (18) 7 (32) 4 (18) 6 (27) 4 (18) 1 (5)
3 (16) 4 (21) 3 (16) 1 (5) 3 (16) 7 (37) 3 (16) 2 (11) 6 (32) 3 (16) 1 (5) 3 (16) 2 (11) 3 (16) 8 (42) 5 (26) 2 (11) 2 (11) 8 (42) 9 (47) 9 (47) 4 (21) 4 (21) 7 (37) 5 (26) 6 (32) 3 (16) 9 (47) 2 (11) 3 (16)
P value
0.041 0.008 0.006
0.006 0.009
Karyotypic differences between t(14;18)(q32;q21) positive and negative FLs Of the 61 cases, 16 (26%) were negative for the t(14;18) (q32;q21) translocation. All cases negative for t(14;18)(q32; q21) were reconfirmed for their absence by FISH. Karyotypic complexity as measured by the mean number of breaks, gains and losses was slightly higher in the t(14;18) (q32;q21) negative FLs. Clear differences were noted in the incidence of some of the recurring breakpoints and additional chromosomal abnormalities between the two groups (Fig. 5). Translocations affecting 3q27 and gain of chromosome 1/1q were significantly more frequent in the t(14; 18)(q32;q21) negative FLs (P ! 0.05). With the exception of imbalances involving chromosomes 6, 7 and 8, several of the others gains and losses were also more common in the t(14; 18)(q32;q21) negative FLs, particularly those without translocations of 3q27. Correlation of recurring clonal abnormalities with histologic grade SKY analysis enabled a more precise correlation of the additional chromosomal changes with histologic grade. As seen in Table 5, karyotypic complexity increased with grade and several additional changes were noted at a higher frequency in grade 3a/3b FL than grade 1–2 FL. Breaks at 6q11– q15, 8q24, and gain of 7/7q, and 8/8q were significantly more frequent in grade 3a/3b FL compared to grade 1–2 FL and breaks at 2p11–p13 were significantly more frequent in grade 1 FL (P ! 0.05). With the exception of 7/7q gain, these associations were not detected in the G-banding analysis. Though not significant (P = 0.09), the incidence of deletions of 6q gradually increased with higher histologic grade, 15% in grade 1, 27% in grade 2 and 47% in grade 3a/3b. A similar trend was observed in the G-banding analysis. Discussion
This study was undertaken to resolve the full spectrum of chromosomal abnormalities in FL. In more than 70% of the cases SKY analysis revealed additional cytogenetic information and in this series involvement of chromosome 12 was found to be significantly more frequent by SKY compared to G-banding. Resolution of all the derived chromosomes, additions and marker chromosomes revealed several new recurring breakpoints, translocations, gains and losses. A detailed analysis of the breakpoints revealed a total of 15 recurring breakpoints, of which six (1p36, 1q11–q21, 3q27, 6q11–q15, 14q32 and 18q21) were detected by both Gbanding and SKY at 110% and nine by SKY alone (1p11–p13, 2p11–p13, 6q21, 6q25–q27, 8q24, 9p13, 10q22–q24, 12q11– q13 and 17q11–q21). Of note were breaks at 1p11–p13 and 2p11–p13 which were highly underrepresented in the Gbanding analysis. Both regions were primarily involved in complex unbalanced translocations and as such underrepresented in the present and previous G-banding studies of FL (Yunis et al., 1987; Armitage et al., 1988; Tilly et al., 1994; Whang-Peng et al., 1995; Juneja et al., 1997; Knutsen, 1997;
Horsman et al., 2001, 2003). While rearrangements affecting 2p11–p13 lead to partial gains of 2p, those affecting 1p11–p13 resulted in partial loss of 1p32–p36. The overall incidence of the previously recognized recurring translocations, t(14; 18)(q32;q21), t(3; 14)(q32;q21), t(8;14)(q24;q32) and their variants was in the reported range. Apart from detecting these known translocations at a slightly higher frequency, SKY enabled their correct identification. SKY also unambiguously resolved all the additions and marker chromosomes detected by G-banding and 90% of these were shown to be translocations. In total, 125 nonIG gene translocations were identified, of which two were recurrent within this series: der(1)t(1;2)(p36;q31) and der(1) t(1; 2)(p11;p11). Sites frequently involved in these non-IG gene translocations included 1p11–p13, 1p36, 1q11–q21, 8q24, 9p13 and 17q11–q21 (10–16%). Interestingly, 98% of these non-IG gene translocations were observed in higher histologic grade FLs with t(14;18)(q32;q21) and/or t(3q27). This is in contrast to other lymphoma subtypes where additional changes primarily constitute chromosomal gains and losses. Genes involved in some of these recurring sites have been identified (Dave et al., 1999; Callanan et al., 2000; Dyomin et al., 2000; Chen et al., 2001) and molecular genetic analysis of the remaining translocations identified in this study can be expected to reveal additional functionally important genes in the pathogenesis of FL. Central to the pathogenesis of FL is the t(14;18)(q32;q21) translocation which results in overexpression of the antiapoptotic protein BCL2 (McDonnell and Korsmeyer, 1991). A subset of FLs lack the t(14;18)(q32;q21) translocation and this subset has not been comprehensively characterized, particularly at the chromosomal level. In the only other molecular cytogenetic study of FL, only cases with t(14; 18) (q32;q21) were analyzed (Lestou et al., 2003). In this series, the t(14; 18)(q32;q21) negative subset exhibited a greater degree of karyotypic complexity and instability than the t(14; 18)(q32;q21) positive subset. Virtually every t(14; 18) (q32;q21) negative FL exhibited gains and/or losses. Gains of 1q and X were predominant being observed in more than half of the cases. More than one third of the cases exhibited gain of chromosome 18q21, 2p/q, 12 and 17q. Several other breakpoints (6q25–q27, 9p13, 12q11–q13 and 17q11–q21) and gains (X, 3p/q, 2p/q, 16 and 17q) were also 2–3 fold higher in the t(14;18)(q32;q21) negative FL. The only abnormalities that strongly associated the t(14;18)(q32;q21) positive FL were rearrangements of 1p36, 8q24 and gain of 7. These findings suggest that genomic instability resulting from increased/decreased gene dosage may be more important in the pathogenesis of t(14;18)(q32;q21) negative FLs, particularly those without translocations of 3q27. The pathways of histologic progression and/or transformation in FL are multiple and result from diverse alterations such as chromosomal translocation, gene deletion and somatic point mutation that lead to activation or inactivation of proto-oncogenes or tumor suppressor genes (Knutsen, 1997; Lossos and Levy, 2003; Viardot et al., 2003). Resolutions of all the unbalanced translocations, additions and marker chromosomes by SKY allowed better delinea-
Cytogenet Genome Res 118:337–344 (2007)
343
tion of the chromosomal imbalances and their association with histologic progression. Thus, 37% of the deletions identified by G-banding were shown to be unbalanced translocations leading to gains by SKY. In addition to identifying all the known frequent gains and losses in FL (+7, +X, +12 or dup(12q), +18, del(6q), del(10q22–q24) and del(17p)) at a greater frequency, SKY identified recurring gains at 110% in six other chromosomes/regions (8/8q, 2p/ q, 3p/q, 6p, 16 and 17q). SKY analysis also enabled the narrowing down of several regions of gain, notably gain of 2p13–p21, 2q11–q21, 2q31–q37, 12q12–q15, 17q21–q25 and 18q21. As observed by others, karyotypic complexity and instability increased with histologic grade. Among the several additional recurring changes identified in this series, breaks at 6q11–q15 and 8q24 and gain of chromosomes 7
and 8 associated significantly with histologic progression. These results are consistent with the recent molecular cytogenetic and array-CGH studies in FL (Bentz et al., 1996; Avet-Loiseau et al., 1997; Lestou et al., 2003; Martinez-Climent et al., 2003). Our results also indicate that deregulation of genes via chromosome translocations is common in the progression of FL. In conclusion, this comprehensive molecular cytogenetic characterization of follicular lymphoma provides a detailed description of chromosomal instability associated with the t(14;18)(q32;q21) positive and negative subsets and histologic grades. Many additional recurring breakpoints, translocations and chromosomal imbalances were identified thus providing a framework for molecular characterization of the genetic events that underlie the biology of FL.
References Anderson JR, Armitage JO, Weisenburger DD: Epidemiology of the non-Hodgkin’s lymphomas: distributions of the major subtypes differ by geographic locations: Non-Hodgkin’s Lymphoma Classification Project. Ann Oncol 9: 717– 720 (1998). Armitage JO, Sanger WG, Weisenburger DD, Harrington DS, Linder J, et al: Correlation of secondary cytogenetic abnormalities with histologic appearance in non-Hodgkin’s lymphomas bearing t(14;18)(q32;q21): J Natl Cancer Inst 80: 576–580 (1988). Avet-Loiseau H, Vigier M, Moreau A, Mellerin MP, Gaillard F, et al: Comparative genomic hybridization detects genomic abnormalities in 80% of follicular lymphomas. Br J Haematol 97:119– 122 (1997). Bentz M, Werner CA, Dohner H, Joos S, Barth TF, et al: High incidence of chromosomal imbalances and gene amplifications in the classical follicular variant of follicle center lymphoma. Blood 88:1437–1444 (1996). Callanan MB, Le Baccon P, Mossuz P, Duley S, Bastard C, et al: The IgG Fc receptor, FcgammaRIIB, is a target for deregulation by chromosomal translocation in malignant lymphoma. Proc Natl Acad Sci USA 97: 309–314 (2000). Chen W, Palanisamy N, Schmidt H, Teruya-Feldstein J, Jhanwar SC, et al: Deregulation of FCGR2B expression by 1q21 rearrangements in follicular lymphomas. Oncogene 20:7686–7693 (2001). Cigudosa JC, Parsa NZ, Louie DC, Filippa DA, Jhanwar SC, et al: Cytogenetic analysis of 363 consecutively ascertained diffuse large B-cell lymphomas. Genes Chromosomes Cancer 25: 123–133 (1999). Dave BJ, Pickering DL, Hess MM, Weisenburger DD, Armitage JO, Sanger WG: Deletion of cell division cycle 2-like 1 gene locus on 1p36 in non- Hodgkin lymphoma. Cancer Genet Cytogenet 108:120–126 (1999). Dyomin VG, Palanisamy N, Lloyd KO, Dyomina K, Jhanwar SC, et al: MUC1 is activated in a B-cell lymphoma by the t(1;14)(q21;q32) translocation and is rearranged and amplified in B-cell lymphoma subsets. Blood 95: 2666–2671 (2000). Hedvat CV, Hegde A, Chaganti RS, Chen B, Qin J, et al: Application of tissue microarray technology to the study of non-Hodgkin’s and Hodgkin’s lymphoma. Hum Pathol 33: 968–974 (2002).
344
Horsman DE, Connors JM, Pantzar T, Gascoyne RD : Analysis of secondary chromosomal alterations in 165 cases of follicular lymphoma with t(14; 18). Genes Chromosomes Cancer 30: 375– 382 (2001). Horsman DE, Okamoto I, Ludkovski O, Le N, Harder L, et al: Follicular lymphoma lacking the t(14; 18)(q32;q21): identification of two disease subtypes. Br J Haematol 120: 424–433 (2003). ISCN 1995: An International System for Human Cytogenetic Nomenclature, Mitelman F (ed) (S. Karger, Basel 1995). Juneja S, Matthews J, Lukeis R, Laidlaw C, Cooper I, et al: Prognostic value of cytogenetic abnormalities in previously untreated patients with non-Hodgkin’s lymphoma. Leuk Lymphoma 25:493–501 (1997). Knutsen T: Cytogenetic mechanisms in the pathogenesis and progression of follicular lymphoma. Cancer Surv 30: 163–192 (1997). Lestou VS, Gascoyne RD, Sehn L, Ludkovski O, Chhanabhai M, et al: Multicolour fluorescence in situ hybridization analysis of t(14; 18)-positive follicular lymphoma and correlation with gene expression data and clinical outcome. Br J Haematol 122: 745–759 (2003). Lossos IS, Levy R: Higher grade transformation of follicular lymphoma: phenotypic tumor progression associated with diverse genetic lesions. Semin Cancer Biol 13: 191–202 (2003). Martinez-Climent JA, Alizadeh AA, Segraves R, Blesa D, Rubio-Moscardo F, et al: Transformation of follicular lymphoma to diffuse large cell lymphoma is associated with a heterogeneous set of DNA copy number and gene expression alterations. Blood 101:3109–3117 (2003). McDonnell TJ, Korsmeyer SJ: Progression from lymphoid hyperplasia to high-grade malignant lymphoma in mice transgenic for the t(14; 18). Nature 349: 254–256 (1991). Nanjangud G, Rao PH, Hegde A, Teruya-Feldstein J, Donnelly G, et al: Spectral karyotyping identifies new rearrangements, translocations, and clinical associations in diffuse large B-cell lymphoma. Blood 99: 2554–2561 (2002). Nathwani BHN, Weisenberger D: Follicular lymphoma, in: Jaffe EHN, Stien H, Vardiman J (eds): Pathology and Genetics of Tumours of Haematopoetic and Lymphoid tissues, pp 162– 167 (IARC press, Lyon, 2001).
Cytogenet Genome Res 118:337–344 (2007)
Offit K, Jhanwar SC, Ladanyi M, Filippa DA, Chaganti RS: Cytogenetic analysis of 434 consecutively ascertained specimens of non-Hodgkin’s lymphoma: correlations between recurrent aberrations, histology, and exposure to cytotoxic treatment. Genes Chromosomes Cancer 3:189– 201 (1991). Rao PH, Murty VV, Gaidano G, Hauptschein R, Dalla-Favera R, Chaganti RS: Subregional localization of 20 single-copy loci to chromosome 6 by fluorescence in situ hybridization. Genomics 16: 426–30 (1993). Rao PH, Cigudosa JC, Ning Y, Calasanz MJ, Iida S, et al: Multicolor spectral karyotyping identifies new recurring breakpoints and translocations in multiple myeloma. Blood 92: 1743–1748 (1998). Sawyer JR, Lukacs JL, Munshi N, Desikan KR, Singhal S, et al: Identification of new nonrandom translocations in multiple myeloma with multicolor spectral karyotyping. Blood 92: 4269– 4278 (1998). Schrock E, du Manoir S, Veldman T, Schoell B, Wienberg J, et al: Multicolor spectral karyotyping of human chromosomes. Science 273: 494– 497 (1996). Tilly H, Rossi A, Stamatoullas A, Lenormand B, Bigorgne C, et al: Prognostic value of chromosomal abnormalities in follicular lymphoma. Blood 84:1043–1049 (1994). Tsujimoto Y, Ikegaki N, Croce CM: Characterization of the protein product of bcl-2, the gene involved in human follicular lymphoma. Oncogene 2: 3–7 (1987). Viardot A, Barth TF, Moller P, Dohner H, Bentz M: Cytogenetic evolution of follicular lymphoma. Semin Cancer Biol 13: 183–90 (2003). Whang-Peng J, Knutsen T, Jaffe ES, Steinberg SM, Raffeld M, et al: Sequential analysis of 43 patients with non-Hodgkin’s lymphoma: clinical correlations with cytogenetic, histologic, immunophenotyping, and molecular studies. Blood 85: 203–216 (1995). Yunis JJ, Frizzera G, Oken MM, McKenna J, Theologides A, Arnesen M: Multiple recurrent genomic defects in follicular lymphoma. A possible model for cancer. N Engl J Med 316: 79–84 (1987).
B-cell NHLs are characterized by recurrent translocations involving the immunoglobulin heavy chain (IGH) gene in approximately 50% of cases (Willis and Dyer, 2000; Kuppers and Dalla-Favera, 2001; Bernicot et al., 2005). The IGH locus is located on chromosome 14, at band 14q32.3, and spans 1,250 kb. IGH involvement in these translocations is the result of its variation during B-cell development. Indeed, the translocations are the consequence of an illegitimate VDJ recombination at the early stage, at the hypermutation or at the switch stage of B-cell development (Kuppers, 2005). IGH translocations correlate with clinical, morphological and immunophenotypic features. Their detection is helpful in establishing the diagnosis, selecting treatment and giving prognostic information. Some are associated with specific subtypes of NHL such as t(11;14)(q13;q32) in about in 95% of mantle cell lymphomas (MCLs) (Vaandrager et al., 1996; Wlodarska et al., 1999), t(14;18)(q32;q21) in 80% of follicular lymphomas (FLs) (Yunis et al., 1982; Vaandrager et al., 2000), t(3;14)(q27;q32) in diffuse large B-cell lymphomas (DLBCL) (Bastard et al., 1992; Cigudosa et al., 1999; Dave et al., 2002; Nanjungud et al., 2002) and t(8; 14)(q24;q32) in Burkitt lymphomas (Douglass et al., 1980; Dalla-Favera et al., 1982). At the molecular level, these translocations, or more rarely insertions, activate an oncogene located near the breakpoint of the chromosomal partner by juxtaposition to IGH regulatory sequences. They led to the identification of several oncogenes involved in NHL pathogenesis (Dalla-Favera et al., 1982; Tsujimoto et al., 1984a, b; Baron et al., 1993; Bastard et al., 1994). As a consequence, translocations involving IGH have been recognized as a hallmark of NHL. However, the IGH locus is located in the terminal band of 14q, a band that is R-positive (dark) with conventional banding techniques, as are many terminal bands of the other chromosomes. Furthermore, karyotypes obtained by conventional cytogenetics are usually complex, with numerous structural abnormalities. Therefore, banding techniques may fail to identify cryptic translocations or insertions and those partners located in the terminal bands of the other chromosome (Gozzetti et al., 2002; Bernicot et al., 2005). Thus, the incidence of IGH translocations could be underestimated. Fluorescent in situ hybridization (FISH) using IGH probes is a further step in approaching the diversity of chromosomal partners and the true incidence of IGH translocations. However, most of the studies are conducted on interphase cells and use probes, such as IGH-BCL2, IGH-MYC, IGH-CCND1, specific of a few recurrent translocations (Frater et al., 2001; Tamura et al., 2001). Increasingly, molecular techniques such as Southern blot hybridization and PCR have become important tools in the diagnosis and follow-up of patients with NHL. However, these techniques have met with varying degrees of success due notably to the relatively large breakpoint regions on 11q13 and 18q21, making interphase FISH a more reliable method for identifying these translocations (Ashton-Key et al., 1995; Horsman et al., 1995; Luthra et al., 1995; Segal and Maiese, 1996; Siebert et al., 1998; Frater et al., 2001; Belaud-Rotureau et al., 2002; Kodet et al., 2003). However, although interphase
346
Cytogenet Genome Res 118:345–352 (2007)
FISH is well-suited to determine the incidence of these specific translocations, it fails to identify other IGH rearrangements. In the present study, we used different FISH techniques to detect IGH rearrangements in 97 NHL patients and identify the chromosomal partners. We compared the distribution observed in our series to that available in the literature. Materials and methods Patients Since 2000, lymph nodes suspected of lymphoma and removed at the CHU Brest have been subjected to a cytogenetic analysis, once the pathological examination had confirmed the diagnosis of NHL. A few lymph nodes from the CH Yves Le Foll in Saint-Brieuc have also been analyzed. Conventional cytogenetics Lymph node samples were received at the cytogenetic laboratory in RPMI 1640. They were manually split and crushed with scalpels. The cell suspension was incubated for 1 h at 37 ° C with 5% CO2. Then, cells were synchronized for 17 h with FrdU (10 –7M) before being released by thymidine (10 –5M) for 6 h. Colcemid exposure (15 min) and standard harvesting were performed. The chromosomes were R-banded and the karyotypes described according to ISCN (1995). At least twenty metaphases were studied for each patient. One spleen sample was also processed according to the same protocol. Molecular cytogenetics LSI IGH/CCND1 dual color probe. The LSI IGH/CCND1 dual color probe is a mixture of the LSI IGH probe labeled with SpectrumGreen and the LSI CCND1 probe labeled with SpectrumOrange (Abbott, Rungis, France). The LSI IGH probe contains two parts hybridizing the sequences on either side of the IGH J region breakpoint (450 kb). The LSI CCND1 probe extends from FGF4 point telomeric to beyond the MTC region (350 kb). LSI IGH/BCL2 dual color probe. The LSI IGH/BCL2 dual color probe is a mixture of the LSI IGH probe labeled with SpectrumGreen and the LSI BCL2 probe labeled with SpectrumOrange (Abbott). The LSI IGH probe spans 1.5 Mb and contains sequences of the entire IGH locus. The LSI BCL2 probe covers a 750 kb region, including the entire BCL2 gene. LSI IGH dual color probe. This probe is a mixture of two probes: the 900 kb LSI IGHV probe, labeled in spectrum green, covering the entire IGH variable region, and the 250 kb 3ⴕ flanking probe, labeled in red spectrum, lying completely 3ⴕ to the IGH locus (Abbott). Hybridization was performed according to the manufacturer’s recommendations. Briefly, before hybridization, slides were immersed in a jar of 2! SSC, 0.4% NP40 solution for 30 min at 37 ° C and then immediately passed through an ethanol series of growing concentrations (70%, 90%, 100%). Denaturation was performed simultaneously on the slides and probes for 1 min at 75 ° C. The slides were incubated overnight in a dark humidified chamber at 37 ° C. They were washed for 45 s in 0.4! SSC, 0.3% NP40 at 72 ° C and 20 s in 2! SSC, 0.1% NP40 at room temperature and finally counterstained with 4,6-diamidino2-phenyl-indole (DAPI). The slides were analyzed using a Zeiss Axioplan Microscope (Zeiss, Le Pecq, France). Subsequent image acquisition was performed using a CCD camera with Isis (significant in situ imaging system) (MetaSystems, Altlussheim, Germany). Whole chromosome paint and telomeric probe. Slides hybridized with the IGH probes were dehybridized and rehybridized with wholechromosome painting (WCP) probes (Qbiogene, Illkirch, France) or telomeric probes (Abbott), as previously described (Morel et al., 2003). The chromosome paints and telomeric probes were chosen according to the chromosomal rearrangement observed in conventional cytogenetics.
100%
IGH
90%
t(14;18)
80%
t(11;14)
70%
t(3;14)
60%
t(14;16)
50%
t(8;14)
40%
t(9;14)
30%
t(1;14)
20% Fig. 1. Distribution of IGH rearrangements with their chromosomal partners by subgroup of B-cell NHL.
t(X;14)
10%
ins(14;19)
0% FL
24-color FISH. In several cases, complex rearrangements could not be elucidated by banding techniques. 24-color FISH using MetaSystems 24Xcyte kit probe (MetaSystems) was applied. The kit contains chromosome-painting probes specific for the 24 different chromosomes. Each paint is labeled with four fluorochromes (FITC, spectrum orange, TexasRed, DEAC) and biotin, respectively, or a unique combination of them. Detection of the biotin-labeled fraction is performed with streptavidin-cy5 (B-tect). Briefly, after treatment with proteinase K (200 g/ml) at 37 ° C for 10 min, the slides were washed in 1! PBS two times at room temperature for 5 min and immediately placed in a 1! PBS + 50 mM MgCl2 solution at room temperature for 5 min. They were immersed in a post fixative solution (1% formaldehyde in 1! PBS + 50 mM MgCl2) for 10 min at room temperature, washed in 1! PBS for 5 min and passed through an ethanol series (70%, 90%, 100%) for 3 min. Target DNA was denatured in 70% formamide, 2! SSC at 75 ° C for 3 min and then passed through a cold ethanol series and allowed to air dry. Probe mix was denatured at 75 ° C for 5 min, put on ice for 2 min and incubated at 37 ° C for 30 min; subsequently 6 l per hybridization were applied to each slide. After four days of hybridization, the slides were washed in 50% formamide, 2! SSC twice for 7 min 30 s each at 41 ° C and then in 2! SSC two times for 7 min 30 s at 41 ° C and one time in 4! SSC containing 0.01% Tween 20 for 3 min at room temperature. Biotin-labeled probes were detected with streptavidin-Cy5. Finally, the slides were counterstained with DAPI. Image acquisition was performed with Isis/mFISH (MetaSystems) (Herry et al., 2004; Douet-Guilbert et al., 2007).
Results
Since 2000, 326 patients with a suspicion of lymphoma have been referred to the cytogenetic laboratory. Among them, 166 patients had NHL. No metaphases and sole normal metaphases were obtained in 29 and 18 cases, respectively. Among the remaining 119 patients with an abnormal karyotype following R-banding technique, 97 were diagnosed as having NHL of B-cell origin (immunohistochemically and immunophenotypically). These 97 B-cell NHL cases were distributed in 42 follicular lymphomas, 14 mantle cell lymphomas, 40 diffuse large B-cell lymphomas and one splenic lymphoma with villous lymphocytes (SLVL). In this latter case, R-banding showed a 47,XYY,t(4; 14)(p13;q32)[4]/47,XYY[50] karyotype. FISH study using LSI IGH showed one IGH rearrangement and
MCL
DLBCL
24-color FISH identified the IGH chromosomal partner as chromosome 4p13, without other abnormalities. FISH analysis using bacterial artificial chromosome (BAC) RP11395I6 containing the RHOH gene revealed three signals, one on the normal chromosome 4, one on the der(4) and one on the der(14), indicating RHOH rearrangement (Bernicot et al., 2006). An IGH rearrangement was observed in 74% of the remaining 96 cases (Fig. 1). A total of 38 cases (90.5%) among the 42 FL had an IGH rearrangement, including 37 cases of t(14;18)(q32;q21) (88.1%) and one t(3;14)(q27;q32) (2.4%). Of these, one case showed a complex rearrangement identified by 24-color FISH as der(14)(14pter]14q32::18q21:: 3q2?5] 3qter) and der(18)(18pter]18q21::14q32). Twelve of 14 MCL (85.7%) had an IGH rearrangement (Fig. 1). Ten patients had a classic t(11; 14)(q13;q32), one of them arising on a constitutional 13;14 Robertsonian translocation (der(13; 14)(13qter]13q10::14q10]14q32::11q13] 11qter)). The other two cases showed complex (11;14) rearrangements identified by 24-color FISH: der(14)(11::14cen ]14q32::11q13]11qter) in one patient and der(5)(5pter] 5q11::14q?]14q32::11q13]11qter), der(13)(5::11q2?1]11q 13::14q32::13p11]13qter),der(22)(5::11q2?1]11q13::14q32:: 22p11]22qter) in the second patient. Twenty-one (52.5%) of the 40 patients with DCLBL showed a rearrangement of the IGH gene, with several chromosomal partners involved (Fig. 1). The most frequent partners were band 18q21 involved in seven patients (17.5%), 3q27 in four (10%) and 8q24 in three (5%). One patient had a complex IGH rearrangement in which chromosomes 3, 14 and 8 were shown to be involved by 24-color FISH: 46,XY, t(3;14)(q27;q11)[14]/46,XY,t(3;14)(q27;q11),t(8;14) (q24;q32)[4]/ 46,XY,der(3)(3pter]3q27::14q11] 14q32::8q24]8qter), der(8)(8pter]8q24 ::14q32]14qter) [2]. Other chromosome band partners were identified by 24-color FISH: 16q2?2 in two cases and 9p13, 1p11, 1q21, Xp11 and 19q13 in one case each. Figure 2 shows the strategy used to identify a cryptic translocation, t(14; 16)(q32;q2?2). Figure 2A shows the Rbanded karyotype of a DLBCL patient. No abnormality is
Cytogenet Genome Res 118:345–352 (2007)
347
1
2
3
4
6
7
8
9
10
11
12
X
13
14
15
16
17
18
19
20
21
22
5
Y
Fig. 2. Identification of a t(14;16)(q32;q2?2) in a DLBCL patient using several FISH techniques.
348
Cytogenet Genome Res 118:345–352 (2007)
observed on chromosomes 14 and 16. Metaphase FISH analysis using the LSI IGH dual color probe shows disruption of all three IGH probes. Three red signals localize to chromosomes 14 and two green to two other chromosomes, signing not only two translocations involving IGH, but also a partial deletion of the third IGH probe (Fig. 2B). 24-color FISH on the same metaphase identifies both chromosomes on which the green signals were visualized (Fig. 2C); they are shown to be chromosomes 16 after 24-color FISH karyotyping (Fig. 2D). A FISH analysis using a TEL16q probe (labeled in red) to identify the chromosomal arm shows three signals (Fig. 2E), two of them being located on chromosomes 14, after dehybridization and rehybridization of the same metaphases with WCP14 (green) and WCP16 (red) probes (Fig. 2F). Discussion
The incidence of translocations involving band 14q32 in B-cell NHL is about 50% in North America and Europe. Variants of the recurrent translocations involving IGH partners have also been described. They involve the immunoglobulin light chains located at bands 2p12]p11 (IGK) and 22q11 (IGL) (Emanuel et al., 1984; Konishi et al., 1990; Larsen et al., 1990; Baron et al., 1993; Wlodarska et al., 2004). Geographic differences in the incidence of IGH translocations have been reported (Segel et al., 1998), with a lower frequency in Japan (Maseki et al., 1987; Konishi et al., 1990; Takechi et al., 1991; Hashimoto et al., 1995). Using interphase FISH with an IGH probe, Ueda et al. (1996), Tsujimoto et al. (1984b) and Taniwaki et al. (1995) found IGH rearrangements in 41% of the Japanese B-cell NHL patients studied. The majority of FISH studies was performed on interphase cells obtained from lymph nodes using dual color probes recognizing the t(11;14)(q13;q32) or t(14;18)(q32;q21). The t(14;18)(q32;q21) was present in 70.4% of the FL, ranging from 25 to 81% and showing geographic disparities (Tsujimoto et al., 1984b; Frater et al., 2001; Bosga-Bouwer et al., 2003; Guo et al., 2005). The incidence of t(11;14)(q13;q32) in MCL was 91.9%, ranging from 66.7 to 100% (Bigoni et al., 1996; Siebert et al., 1998; Li et al., 1999; Frater et al., 2001; Belaud-Rotureau et al., 2002; Kodet et al., 2003; Jarosova et al., 2004). In DLBCL, the 14;18 translocation was found in 16% of the cases (range: 13.1 to 20%) (Huang et al., 2002; Barrans et al., 2003; Iqbal et al., 2004; Cerretini et al., 2006). Gozzetti et al. (2002) used dual color FISH with an IGH probe to determine the rate of cryptic translocations and identify the chromosomal partners among 51 selected Bcell NHL patients. No IGH rearrangement was found among the three patients with a normal karyotype. Nine (69.2%) of the 13 patients with add(14)(q32) had a translocation involving IGH, the remaining four showing no IGH rearrangement. Among the 35 patients having an abnormal karyotype without visible anomalies of chromosome 14 in conventional cytogenetics, seven (20%) had a masked or cryptic IGH translocation. Two studies used SKY or 24-color FISH to identify IGH rearrangements in 66 DLBCL cases (Dave et al., 2002; Nan-
jungud et al., 2002). A t(14;18) was identified in ten cases (15.2%), a t(3; 14)(q27;q32) in eight cases (12.1%), a t(9; 14) (p13;q32) in four cases (6.1%) while other 14q32 rearrangements were found in 12 cases (18.2%). Fan and Rizkalla (2003) used SKY and FISH techniques using various probes such as IGH, MYC, IGH-BCL2 and IGH-CCND1 to determine the rate of IGH rearrangements in 46 B-cell NHL lymph node samples (Fan and Rizkalla, 2003). All five MCL had a t(11;14). Twenty-two (91.7%) of the 24 FL had an IGH rearrangement, 20 (83.3%) being a t(14;18). Fifteen (71.4%) of the 21 DLBCL had an IGH rearrangement, 12 (57.1%) being t(14;18) and two (9.5%) were t(3;14). Our results on follicular and mantle cell lymphoma are in line with those reported in the literature. Some differences have been noted in diffuse large B-cell lymphoma. Indeed, our approach allows the identification of complex translocations and insertions resulting in masked rearrangements. Furthermore, DLBCL includes several subtypes characterized by different genetic, morphologic and clinical features, which may explain the great variety of chromosomal partners of IGH translocations. This heterogeneity could be due to the developmental stage at which the B-cell is transformed (germinal center or post-germinal center) or the transformation of a lower histologic grade (Alizadeh et al., 2000). Gene expression studies have identified several subtypes of DLBCL, GCBL (germinal center B-cell like) and ABL (activated B-cell like) (Alizadeh et al., 2000; Iqbal et al., 2004). Furthermore, although DLBCL is a malignancy of late adulthood, cases have been reported in children and young adults, which appear to have a pattern of chromosomal abnormalities distinct from older adults, including IGH rearrangements which seem to be rarer (Dave et al., 2004). CD5+ diffuse large B-cell lymphoma is now considered as a distinct subgroup of DLBCL. Few cytogenetic studies have been performed in this particular subset (Katzenberger et al., 2003; Yoshioka et al., 2005). No t(14;18)(q32;q21), nor t(11;14)(q13;q32) have been identified among the 36 cases reported whereas t(3;14)(q27;q32) was found in four cases (11.1%) and t(9;14)(p13;q32) in a sole case. Complex translocations involving three or four chromosomes and insertions leading to IGH rearrangement have also been identified. Several complex t(14;18)(q32;q21) were described in FL (Tilly et al., 1994; Horsman et al., 2001; Le Baccon et al., 2001; Fan and Rizkalla, 2003; Cook et al., 2004; Bentley et al., 2005; Okano et al., 2005). In some cases, the third chromosomal band is known to participate in other recurrent IGH translocations (9p13, 8q24, 3q27) while, in some others, it contains the IGK or IGL gene (2p11, 22q11). Complex t(11;14)(q13;q32) and insertions were described in MCL (Koduru et al., 1989; Leroux et al., 1991; Wong and Chan, 1999; Tamura et al., 2001; Au et al., 2002; Mohamed et al., 2002; Aventin et al., 2003; Maravelaki et al., 2004; Michaux et al., 2004); no involvement pattern of the third chromosome is found, to the contrary of what is observed in FL. Several recurrent IGH translocations such as t(14;18), t(3;14) and t(8;14) were also found in complex rearrangements in DLBCL (Ladanyi et al., 1991; Goyns et al., 1993; WhangPeng et al., 1995; Nowotny et al., 1996; Cigudosa et al., 1999;
Cytogenet Genome Res 118:345–352 (2007)
349
IGH in DLBCL subtype 14q32 in DLBCL subtype IGH in FL subtype 1
2
3
4
5
14q32 in FL subtype IGH in the MCL subtype 14q32 in MCL subtype
6
7
8
13
14
15
19
20
9
21
Table 1. Genes involved in IGH translocations in B-cell NHL
10
11
12
16
17
18
22
Chromosomal band 1q21
3p13 3q27 8q24 9p13 11q13 18q21 19q13
Dave et al., 1999; Macpherson et al., 1999; Roumier et al., 2000; Gozzetti et al., 2002; Nanjungud et al., 2002; Nomura et al., 2005); in most of the cases, the third chromosomal band was also shown to be involved in two-way reciprocal translocation with band 14q32. Although a few IGH translocations such as t(14;18), t(11; 14) and t(3;14) are very frequently observed in B-cell NHL, many other chromosomal bands have been shown to be partners in translocations involving 14q32 or the IGH gene by banding and FISH techniques, respectively (Fig. 3). Some rearrangements have been identified in two patients or more (recurrent breakpoints), others in single patients. Besides band 11q13 involved in t(11;14), hallmark of MCL, only two bands, 8q24 and 18q21, have been thus far identified in
350
Cytogenet Genome Res 118:345–352 (2007)
X
Y
Fig. 3. Distribution of the partners of IGH gene or band 14q32. On the left, breakpoints identified in a sole case and, on the right, recurrent breakpoints (diagrams based on G-banding).
Gene name
Function
FCGR2B IRTA1/FCRL3 IRTA2/FCRL5 BCL9 FOXP1 BCL6
Contain immune receptor tyrosine-based inhibition motifs signal transduction?
MYC PAX5 CCND1 BCL2 BCL3
Forkhead box transcription factor Transcriptional repressor required for GC formation Transcriptional factor Transcriptional factor Cell cycle regulator Negative regulator of apoptosis Transcriptional co-activator
IGH rearrangements. More surprisingly, although t(14;18) is present in some 70% of FL, many other single or recurrent breakpoints have been identified, as shown in Fig. 3. It remains that only a few genes deregulated by these IGH translocations have been identified. As for fusion genes observed notably in leukemia (De Braekeleer et al., 2005), known genes deregulated in B-cell NHL are involved in transcription or apoptosis (Table 1). In conclusion, a global approach associating banding and FISH techniques should help in determining the true incidence of IGH rearrangements and reveal new translocations in B-cell NHL. The identification of new genes is of utmost importance for a better understanding of the molecular mechanisms involved in the genesis of lymphoma.
References Alizadeh AA, Eisen MB, Davis RE, Ma C, Lossos IS, et al: Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature 403:503–511 (2000). Ashton-Key M, Diss TC, Isaacson PG, Smith ME: A comparative study of the value of immunohistochemistry and the polymerase chain reaction in the diagnosis of follicular lymphoma. Histopathology 27: 501–508 (1995). Au WY, Gascoyne RD, Viswanatha DS, Connors JM, Klasa RJ, Horsman DE: Cytogenetic analysis in mantle cell lymphoma: a review of 214 cases. Leukemia Lymphoma 43:783–791 (2002). Aventin A, Nomdedeu J, Briones J, Espinosa I, Bordes R, Sierra J: Insertion of CCND1 gene into the IgH locus in a case of leukaemic small cell mantle lymphoma with normal chromosomes 11 and 14. J Clin Pathol 56: 798–800 (2003). Baron BW, Nucifora G, McCabe N, Espinosa R, III, Le Beau MM, McKeithan TW: Identification of the gene associated with the recurring chromosomal translocation t(3; 14)(q27;q32) and t(3; 22)(q27;q11) in B-cell lymphomas. Proc Natl Acad Sci USA 90:5262–5266 (1993). Barrans SL, Evans PA, O’Connor SJ, Kendall SJ, Owen RG, et al: The t(14; 18) is associated with germinal center-derived diffuse large B-cell lymphoma and is a strong predictor of outcome. Clin Cancer Res 9: 2133–2139 (2003). Bastard C, Tilly H, Lenormand B, Bigorgne C, Boulet D, et al: Translocations involving 3q27 and Ig gene regions in non-Hodgkin’s lymphoma. Blood 79: 2527–2531 (1992). Bastard C, Deweindt C, Kerckaert JP, Lenormand B, Rossi A, et al: LAZ3 rearrangements in nonHodgkin’s lymphoma: correlation with histology, immunophenotype, karyotype, and clinical outcome in 217 patients. Blood 83: 2423– 2427 (1994). Belaud-Rotureau MA, Parrens M, Dubus P, Garroste JC, de Mascarel A, Merlio JP: A comparative analysis of FISH, RT-PCR, PCR, and immunohistochemistry for the diagnosis of mantle cell lymphomas. Mod Pathol 15: 517– 525 (2002). Bentley G, Palutke M, Mohamed AN: Variant t(14; 18) in malignant lymphoma: a report of seven cases. Cancer Genet Cytogenet 157: 12–17 (2005). Bernicot I, Douet-Guilbert N, Le Bris MJ, Morice P, Abgrall JF, et al: Characterization of IGH rearrangements in non-Hodgkin’s B-cell lymphomas by fluorescence in situ hybridization. Anticancer Res 25: 3179–3182 (2005). Bernicot I, Morel F, Le Bris MJ, Bail JP, Douet-Guilbert N, De Braekeleer M: Rearrangement of the RHOH gene in a case of splenic lymphoma with villous lymphocytes. Cancer Genet Cytogenet 170:78–79 (2006). Bigoni R, Negrini M, Veronese ML, Cuneo A, Castoldi GL, Croce CM: Characterization of t(11; 14) translocation in mantle cell lymphoma by fluorescent in situ hybridization. Oncogene 13: 797–802 (1996). Bosga-Bouwer AG, van Imhoff GW, Boonstra R, van der Veen A, Haralambieva E, et al: Follicular lymphoma grade 3B includes 3 cytogenetically defined subgroups with primary t(14; 18), 3q27, or other translocations: t(14;18) and 3q27 are mutually exclusive. Blood 101: 1149–1154 (2003). Cerretini R, Noriega MF, Narbaitz M, Slavutsky I: New chromosome abnormalities and lack of BCL-6 gene rearrangements in Argentinean diffuse large B-cell lymphomas. Eur J Haematol 76:284–293 (2006).
Cigudosa JC, Parsa NZ, Louie DC, Filippa DA, Jhanwar SC, et al: Cytogenetic analysis of 363 consecutively ascertained diffuse large B-cell lymphomas. Genes Chromosomes Cancer 25: 123–133 (1999). Cook JR, Shekhter-Levin S, Swerdlow SH: Utility of routine classical cytogenetic studies in the evaluation of suspected lymphomas: results of 279 consecutive lymph node/extranodal tissue biopsies. Am J Clin Pathol 121: 826–835 (2004). Dalla-Favera R, Bregni M, Erikson J, Patterson D, Gallo RC, Croce CM: Human c-myc onc gene is located on the region of chromosome 8 that is translocated in Burkitt lymphoma cells. Proc Natl Acad Sci USA 79: 7824–7827 (1982). Dave BJ, Hess MM, Pickering DL, Zaleski DH, Pfeifer AL, et al: Rearrangements of chromosome band 1p36 in non-Hodgkin’s lymphoma. Clin Cancer Res 5: 1401–1409 (1999). Dave BJ, Nelson M, Pickering DL, Chan WC, Greiner TC, Weisenburger DD, et al: Cytogenetic characterization of diffuse large cell lymphoma using multi-color fluorescence in situ hybridization. Cancer Genet Cytogenet 132:125–132 (2002). Dave BJ, Weisenburger DD, Higgins CM, Pickering DL, Hess MM, et al: Cytogenetics and fluorescence in situ hybridization studies of diffuse large B-cell lymphoma in children and young adults. Cancer Genet Cytogenet 153: 115–121 (2004). De Braekeleer M, Morel F, Le Bris MJ, Herry A, Douet-Guilbert N: The MLL gene and translocations involving chromosomal band 11q23 in acute leukemia. Anticancer Res 25: 1931–1944 (2005). Douet-Guilbert N, Marical H, Pinson L, Herry A, Le Bris MJ, et al: Characterisation of supernumerary chromosomal markers: a study of 13 cases. Cytogenet Genome Res 116: 18–23 (2007). Douglass EC, Magrath IT, Lee EC, Whang-Peng J: Cytogenetic studies in non-African Burkitt lymphoma. Blood 55: 148–155 (1980). Emanuel BS, Selden JR, Chaganti RS, Jhanwar S, Nowell PC, Croce CM: The 2p breakpoint of a 2; 8 translocation in Burkitt lymphoma interrupts the V kappa locus. Proc Natl Acad Sci USA 81: 2444–2446 (1984). Fan YS, Rizkalla K: Comprehensive cytogenetic analysis including multicolor spectral karyotyping and interphase fluorescence in situ hybridization in lymphoma diagnosis: a summary of 154 cases. Cancer Genet Cytogenet 143: 73– 79 (2003). Frater JL, Tsiftsakis EK, Hsi ED, Pettay JMT, Tubbs RR: Use of novel t(11; 14) and t(14; 18) dual-fusion fluorescence in situ hybridization probes in the differential diagnosis of lymphomas of small lymphocytes. Diagn Mol Pathol 10: 214– 222 (2001). Goyns MH, Hammond DW, Harrison CJ, Menasce LP, Ross FM, Hancock BW: Structural abnormalities of the X chromosome in non-Hodgkin’s lymphoma. Leukemia 7: 848–852 (1993). Gozzetti A, Davis EM, Espinosa R, III, Fernald AA, Anastasi J, Le Beau MM: Identification of novel cryptic translocations involving IGH in B-cell non-Hodgkin’s lymphomas. Cancer Res 62: 5523–5527 (2002). Guo Y, Karube K, Kawano R, Yamaguchi T, Suzumiya J, et al: Low-grade follicular lymphoma with t(14; 18) presents a homogeneous disease entity otherwise the rest comprises minor groups of heterogeneous disease entities with Bcl2 amplification, Bcl6 translocation or other gene aberrances. Leukemia 19:1058–1063 (2005).
Harris NL, Jaffe ES, Diebold J, Flandrin G, MullerHermelink HK, et al: World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissues: report of the Clinical Advisory Committee meeting-Airlie House, Virginia, November 1997. J Clin Oncol 17:3835–3849 (1999). Harris NL, Stein H, Coupland SE, Hummel M, Dalla-Favera R, et al: New approaches to lymphoma diagnosis. Hematology 1: 194–220 (2001). Hashimoto K, Miura I, Chyubachi A, Saito M, Miura AB: Correlations of chromosome abnormalities with histologic and immunologic characteristics in 49 patients from Akita, Japan with non-Hodgkin lymphoma. Cancer Genet Cytogenet 81: 56–65 (1995). Herry A, Morel F, Le Bris MJ, Bellec V, Lallaoui H, et al: Molecular cytogenetic characterization of two small chromosome 8 derived supernumerary mosaic markers. Am J Med Genet 128A:33– 38 (2004). Horsman DE, Gascoyne RD, Coupland RW, Coldman AJ, Adomat SA: Comparison of cytogenetic analysis, southern analysis, and polymerase chain reaction for the detection of t(14;18) in follicular lymphoma. Am J Clin Pathol 103: 472–478 (1995). Horsman DE, Connors JM, Pantzar T, Gascoyne RD: Analysis of secondary chromosomal alterations in 165 cases of follicular lymphoma with t(14; 18). Genes Chromosomes Cancer 30: 375– 382 (2001). Huang JZ, Sanger WG, Greiner TC, Staudt LM, Weisenburger DD, et al: The t(14; 18) defines a unique subset of diffuse large B-cell lymphoma with a germinal center B-cell gene expression profile. Blood 99: 2285–2290 (2002). Iqbal J, Sanger WG, Horsman DE, Rosenwald A, Pickering DL, et al: BCL2 translocation defines a unique tumor subset within the germinal center B-cell-like diffuse large B-cell lymphoma. Am J Pathol 165: 159–166 (2004). ISCN (1995): An International System for Human Cytogenetic Nomenclature. Mitelman F (ed) (S Karger, Basel 1995). Jarosova M, Papajik T, Holzerova M, Dusek L, Pikalova Z, et al: High incidence of unbalanced chromosomal changes in mantle cell lymphoma detected by comparative genomic hybridization. Leuk Lymphoma 45:1835–1846 (2004). Katzenberger T, Lohr A, Schwarz S, Dreyling M, Schoof J, et al: Genetic analysis of de novo CD5+ diffuse large B-cell lymphomas suggests an origin from a somatically mutated CD5+ progenitor B cell. Blood 101: 699–702 (2003). Kodet R, Mrhalova M, Krskova L, Soukup J, Campr V, et al: Mantle cell lymphoma: improved diagnostics using a combined approach of immunohistochemistry and identification of t(11; 14) (q13;q32) by polymerase chain reaction and fluorescence in situ hybridization. Virchows Arch 442:538–547 (2003). Koduru PR, Offit K, Filippa DA: Molecular analysis of breaks in BCL-1 proto-oncogene in B-cell lymphomas with abnormalities of 11q13. Oncogene 4: 929–934 (1989). Konishi H, Sakurai M, Nakao H, Maseki N, Kaneko Y, et al: Chromosome abnormalities in malignant lymphoma in patients from Kurashiki: histological and immunophenotypic correlations. Cancer Res 50: 2698–2703 (1990). Kuppers R: Mechanisms of B-cell lymphoma pathogenesis. Nat Rev Cancer 5: 251–262 (2005). Kuppers R, Dalla-Favera R: Mechanisms of chromosomal translocations in B cell lymphomas. Oncogene 20:5580–5594 (2001).
Cytogenet Genome Res 118:345–352 (2007)
351
Ladanyi M, Offit K, Jhanwar SC, Filippa DA, Chaganti RS: MYC rearrangement and translocations involving band 8q24 in diffuse large cell lymphomas. Blood 77: 1057–1063 (1991). Larsen CJ, Mecucci C, Leroux D: t(2;18) and t(18;22) variant chromosomal translocations and bcl-2 gene rearrangements in human malignant lymphomas. Nouv Rev Fr Hématol 32: 401–403 (1990). Le Baccon P, Leroux D, Dascalescu C, Duley S, Marais D, et al: Novel evidence of a role for chromosome 1 pericentric heterochromatin in the pathogenesis of B-cell lymphoma and multiple myeloma. Genes Chromosomes Cancer 32:250–264 (2001). Leroux D, Le Marc’Hadour F, Gressin R, Jacob MC, Keddari E, et al: Non-Hodgkin’s lymphomas with t(11;14)(q13;q32): a subset of mantle zone/ intermediate lymphocytic lymphoma? Br J Haematol 77: 346–353 (1991). Li JY, Gaillard F, Moreau A, Harousseau JL, Laboisse C, et al: Detection of translocation t(11; 14)(q13;q32) in mantle cell lymphoma by fluorescence in situ hybridization. Am J Pathol 154: 1449–1452 (1999). Luthra R, Hai S, Pugh WC: Polymerase chain reaction detection of the t(11; 14) translocation involving the bcl-1 major translocation cluster in mantle cell lymphoma. Diagn Mol Pathol 4:4–7 (1995). Macpherson N, Lesack D, Klasa R, Horsman D, Connors JM, et al: Small noncleaved, nonBurkitt’s (Burkitt-like) lymphoma: cytogenetics predict outcome and reflect clinical presentation. J Clin Oncol 17:1558–1567 (1999). Maravelaki S, Burford A, Wotherspoon A, Joshi R, Matutes E, et al: Molecular cytogenetic study of a mantle cell lymphoma with a complex translocation involving the CCND1 (11q13) region. Cancer Genet Cytogenet 154: 67–71 (2004). Maseki N, Kaneko Y, Sakurai M, Kurihara M, Sampi K, et al: Chromosome abnormalities in malignant lymphoma in patients from Saitama. Cancer Res 47: 6767–6775 (1987). Michaux L, Wlodarska I, Theate I, Stul M, Scheiff JM, et al: Coexistence of BCL1/CCND1 and CMYC aberrations in blastoid mantle cell lymphoma: a rare finding associated with very poor outcome. Ann Hematol 83: 578–583 (2004). Mohamed AN, Ali W, Kopptich F, Al Katib A: Banded chromosomes versus fluorescence in situ hybridization in the diagnosis of mantle cell lymphoma: a lesson from three cases. Cancer Genet Cytogenet 136:108–112 (2002). Morel F, Le Bris MJ, Douet-Guilbert N, Duchemin J, Herry A, et al: Insertion of chromosome 11 in chromosome 4 resulting in a 5ⴕMLL-3ⴕAF4 fusion gene in a case of adult acute lymphoblastic leukemia. Cancer Genet Cytogenet 145: 74–77 (2003).
352
Nanjungud G, Rao PH, Hegde A, Teruya-Feldstein J, Donnelly G, et al: Spectral karyotyping identifies new rearrangements, translocations, and clinical associations in diffuse large B-cell lymphoma. Blood 99: 2554–2561 (2002). Nomura K, Kanda-Akano Y, Shimizu D, Okuda T, Yoshida N, et al: An additional segment at 1p36 derived from der(18)t(14; 18) in patients with diffuse large B-cell lymphomas transformed from follicular lymphoma. Ann Hematol 84: 474–476 (2005). Nowotny H, Karlic H, Gruner H, Hirsch J, Vesely M, et al: Cytogenetic findings in 175 patients indicate that items of the Kiel classification should not be disregarded in the REAL classification of lymphoid neoplasms. Ann Hematol 72:291–301 (1996). Okano A, Nakano S, Namura K, Yamada N, Uchida R, et al: Unique three-way translocation, t(3;14; 18)(q27;q32;q21), in follicular lymphoma. Cancer Genet Cytogenet 159:174–176 (2005). Roumier C, Galiegue-Zouitina S, Bastard C, Soenen V, Lai JL, et al: FISH analysis with a YAC probe improves detection of LAZ3/BCL6 rearrangement in non-Hodgkin’s lymphoma. Hematol J 1:117–125 (2000). Segal GH, Maiese RL: Mantle cell lymphoma. Rapid polymerase chain reaction-based genotyping of a morphologically heterogeneous entity. Arch Pathol Lab Med 120: 835–841 (1996). Segel MJ, Paltiel O, Zimran A, Gottschalk-Sabag S, Schibi G, et al: Geographic variance in the frequency of the t(14;18) translocation in follicular lymphoma: An Israeli series compared to the world literature. Blood Cells Mol Dis 24: 62–72 (1998). Siebert R, Matthiesen P, Harder S, Zhang Y, Borowski A, et al: Application of interphase cytogenetics for the detection of t(11; 14)(q13;q32) in mantle cell lymphomas. Ann Oncol 9:519–526 (1998). Takechi M, Tanaka K, Hashimoto T, Asaoku H, Dohy H, et al: Cytogenetic, molecular biological and clinical study of B-cell lymphomas with 14;18 translocation in Japanese patients. Leukemia 5:1069–1075 (1991). Tamura A, Miura I, Iida S, Yokota S, Horiike S, et al: Interphase detection of immunoglobulin heavy chain gene translocations with specific oncogene loci in 173 patients with B-cell lymphoma. Cancer Genet Cytogenet 129:1–9 (2001). Taniwaki M, Nishida K, Ueda Y, Misawa S, Nagai M, et al: Interphase and metaphase detection of the breakpoint of 14q32 translocations in B-cell malignancies by double-color fluorescence in situ hybridization. Blood 85: 3223–3228 (1995). Tilly H, Rossi A, Stamatoullas A, Lenormand B, Bigorgne C, et al: Prognostic value of chromosomal abnormalities in follicular lymphoma. Blood 84:1043–1049 (1994).
Cytogenet Genome Res 118:345–352 (2007)
Tsujimoto Y, Finger LR, Yunis J, Nowell PC, Croce CM: Cloning of the chromosome breakpoint of neoplastic B cells with the t(14;18) chromosome translocation. Science 226: 1097–1099 (1984a). Tsujimoto Y, Yunis J, Onorato-Showe L, Erikson J, Nowell PC, Croce CM: Molecular cloning of the chromosomal breakpoint of B-cell lymphomas and leukemias with the t(11; 14) chromosome translocation. Science 224: 1403–1406 (1984b). Ueda Y, Matsuda F, Misawa S, Taniwaki M: Tumorspecific rearrangements of the immunoglobulin heavy-chain gene in B-cell non-Hodgkin’s lymphoma detected by in situ hybridization. Blood 87: 292-298 (1996). Vaandrager JW, Schuuring E, Zwikstra E, de Boer CJ, Kleiverda KK, et al: Direct visualization of dispersed 11q13 chromosomal translocations in mantle cell lymphoma by multicolor DNA fiber fluorescence in situ hybridization. Blood 88:1177–1182 (1996). Vaandrager JW, Schuuring E, Raap T, Philippo K, Kleiverda K, Kluin P: Interphase FISH detection of BCL2 rearrangement in follicular lymphoma using breakpoint-flanking probes. Genes Chromosomes Cancer 27: 85–94 (2000). Whang-Peng J, Knutsen T, Jaffe ES, Steinberg SM, Raffeld M, et al: Sequential analysis of 43 patients with non-Hodgkin’s lymphoma: clinical correlations with cytogenetic, histologic, immunophenotyping, and molecular studies. Blood 85: 203–216 (1995). Willis TG, Dyer MJS: The role of immunoglobulin translocations in the pathogenesis of B-cell malignancies. Blood 96:808–822 (2000). Wlodarska I, Pittaluga S, Hagemeijer A, De WolfPeeters C, Van den Berghe H: Secondary chromosome changes in mantle cell lymphoma. Haematologica 84: 594–599 (1999). Wlodarska I, Meeus P, Stul M, Thienpont L, Wouters E, et al: Variant t(2; 11)(p11;q13) associated with the IgK-CCND1 rearrangement is a recurrent translocation in leukemic small-cell Bnon-Hodgkin lymphoma. Leukemia 18: 1705– 1710 (2004). Wong KF, Chan JKC: Cytogenetic abnormalities in chronic B-cell lymphoproliferative disorders in Chinese patients. Cancer Genet Cytogenet 111: 55–60 (1999). Yoshioka T, Miura I, Kume M, Takahashi N, Okamoto M, et al: Cytogenetic features of de novo CD5-positive diffuse large B-cell lymphoma: chromosome aberrations affecting 8p21 and 11q13 constitute major subgroups with different overall survival. Genes Chromosomes Cancer 42:149–157 (2005). Yunis JJ, Oken MM, Kaplan ME, Ensrud KM, Howe RR, Theologides A: Distinctive chromosomal abnormalities in histologic subtypes of nonHodgkin’s lymphoma. N Engl J Med 307: 1231– 1236 (1982).
nual extranodal non-Hodgkin lymphomas (Cancer in Finland, Finnish Cancer Registry, 2003; www.cancerregistry. fi). The most common forms of CTCL are mycosis fungoides (MF) and its variants, and CD30-positive forms (primary skin CD30+ lymphoproliferative disease (ca. 30% of CTCL): skin primary anaplastic large cell lymphoma, lymphomatoid papulosis and borderline cases, and the leukaemic Sézary syndrome (SS; Willemze et al., 2005), but altogether eight different subtypes are currently distinguished based on clinicopathologic findings (Willemze et al., 2005). Clinically, MF is often preceded by eczema-like large plaques, i.e. parapsoriasis en plaques, considered by some experts as already CTCL. In our own clinical material, 10% of small plaque parapsoriasis (SPP) and 35% of large plaque parapsoriasis (LPP) cases developed to MF during a mean follow-up time of nine years (Väkevä et al., 2005). In advanced CTCL, malignant cells are found in lymph nodes, viscera and blood. MF is classifed to various stages according to the TNM classification and groupings of TNM classes (IA to IVB; reviewed by Girardi et al., 2004). Sézary syndrome, showing erythroderma, lymph node affision and malignant cells in blood, may evolve directly, or from a preceding MF. The 5-year survival in MF depends on the stage of the disease so that the median survival is 87%, but in the earliest stage, stage IA it is nearly normal, but for the late stage IVB 15% or less, and in SS about 11% (Willemze et al., 1997; Whittaker et al., 2003, review). Histopathology of MF and SS In MF, small or medium-sized lymphocytes with cerebriform nuclei infiltrate the skin epidermotropically. Papillary dermis, and epidermis may show clusters of malignant cells (Pautrier’s microabscesses). The infiltrate shows an admixture of reactive, inflammatory cells. Thus, the diagnosis of early MF is difficult, since only a few malignant cells may be present. The definition of histopathologic features to differentiate early MF from benign inflammatory diseases is a difficult and debated issue in CTCL diagnostics. Usually, multiple biopsies are required to reach a definite diagnosis. As MF progresses to tumour stage, the proportion and size of tumour cells increases, and epidermotropism decreases (Willemze et al., 1997). Sézary syndrome resembles MF, but the lymphocyte infiltrate may be more monotonous and epidermotropism may be absent. The same type of infiltrate of SS cells is seen in lymph nodes. The usual criterion for Sézary syndrome diagnosis, at least 1,000 Sézary cells per mm3 of blood, can be reached in benign erythrodermas as well. Therefore, the European Organization for Research and Treatment of cancer (EORTC) has proposed an additional criterion, the ratio CD4/CD8 110 (Willemze et al., 1997). The malignant cells in MF or SS are immunophenotypically usually CD3+, CD4+, CD45RO+, CD8–, and CD30–, in rare cases CD3+, CD4–, and CD8+. They coexpress skin homing molecules, e.g. CLA and CCR4 (Ferenczi et al., 2002). In MF, the malignant cells have been considered mature Th1 type cells, while in SS, the cells are functionally Th2 type (Saed et al., 1994; Willemze et al., 1997). In SS, they
354
Cytogenet Genome Res 118:353–361 (2007)
often express T-plastin (Kari et al., 2003; Su et al., 2003), and may show loss of any, or all of T-cell antigens (CD2, CD3, CD4, CD5; Willemze et al., 2005). According to one hypothesis, in advanced stages of CTCL, T regulatory cells (Treg; CD4+, CD25+, FOXP3) are incapable of suppressing the proliferation of CD4+, CD25– cells, which are suppressable by Tregs of healthy controls ex vivo (Tiemessen et al., 2006). Reportedly, part of the malignant cells express CD25, but have not become immunosuppressive themselves (Berger et al., 2005; Tiemessen et al., 2006). Advanced CTCL may transform cytologically, developing the appearance of a large-cell lymphoma with more than 25% of pleomorphic large cells with prominent nucleoli, CD30 positivity and an increase in the clinical aggressiveness of the disease (Cerroni et al., 1990, 1992; Wolfe et al., 1995). A cytotoxic T-cell response against the tumour cells appears to control the malignancy to some extent (Bagot et al., 1998), but Bcl-2 mediated apoptosis is weak (Dummer et al., 1995; Nevala et al., 2001). A dynamic communication between Langerhans’ cells and CTCL cells stimulating the latter to proliferate against their own tumour antigens has been suggested (Berger et al., 2002). Chromosome and gene studies concerning the etiology, diagnosis and follow-up of CTCL
Etiology, general remarks The etiology of CTCL is poorly known. A consistent viral, environmental, or occupational causative factor or hereditary mutation has not been found (Ranki et al., 1990; Väkevä et al., 2000, 2005; Burg et al., 2001; Girardi, 2004 review; Morozov et al., 2005; Pawlaczyk et al., 2005). Early conventional cytogenetic studies showed a large repertoire of chromosomal abnormalities in CTCL, both clonal and non-clonal, but no specific abnormality (Whang-Peng et al., 1982). Gene-level or protein expression abnormalities or promoter hypermethyation of c-myc, p15 and p16, have been observed (Peris et al., 1991; Navas et al., 2000; Scarisbrick et al., 2002). Mutations and overexpression of p53 are associated with advanced or transformed disease (Garatti et al., 1995; Li et al., 1998; Marrogi et al., 1999). The timing and compartment of malignization are not known. When the blood, skin and lymph node of two SSpatients were studied with a combination of immunofluorescence and individual identification of malignant cells by centromere-specific in situ hybridization, it was observed that individual cytogenetic clonal cells are mainly CD3+, CD4+, CD45RO+ (Schlegelberger et al., 1994), but can at the same time show CD45RA typical to naive cells or cells reverted toward naiveness, and CDw150+, even if they are IL2+, INF␥– (Karenko et al., 2001). T-cell receptor (TCR) rearrangement clones and cytogenetic clones T-cell receptor (TCR) rearrangement studies, with diverse very sensitive PCR-based methods, show dominant TCR gamma clones in the majority of MF-cases, accompa-
nied by a TCR beta or gamma-delta rearrangement, but some cases are oligoclonal or non-clonal (Vega et al., 2002; Muche et al., 2004). In addition, a minority of patients show clonal heterogeneity between skin samples or skin and lymph node or blood samples (Vega et al., 2002; Muche et al., 2004). It has been concluded that the T-cell proliferation begins as polyclonal, but dominant clones develop later. Alternatively, clonal evolution of the dominant clone produces subclones with new rearrangements or deletions at TCRlocus (Vega et al., 2002). The TCR receptor repertoire of CTCL patients is restricted compared to normal persons (Yawalkar et al., 2003). The restricted use of V beta segments, that would suggest a common superantigen, has not been confirmed (Gorochov et al., 1995), however. Preferential use of Jbeta1 genes has been observed in advanced cases (Morgan et al., 2006). Diverse TCR gamma clones but only one TCR beta clone in cells microdissected from an early stage CTCL patient’s skin lesion could indicate a superantigen-driven cell proliferation in that particular case (Linnemann et al., 2004). Dominant TCR clones can be found also in early stage CTCL (Wood et al., 1994; Muche et al., 1997; Fraser-Andrews et al., 2000) and in large and small plaque parapsoriasis (Kikuchi et al., 1993, Southern blot; Haeffner et al., 1995; Muche et al., 1999). They may have diagnostic or prognostic significance in CTCL (Delfau-Larue et al., 2000; Fraser-Andrews et al., 2000; Muche et al., 2000), but in blood they may also be age-related when they differ from the clones in the skin (Delfau-Larue et al., 2000). Dominant TCR clones can, however, also be detected in non-malignant diseases, such as lichen sclerosus et atrophicus (Lukowsky et al., 2000) and sarcoidosis (Sawabe et al., 2000). Cytogenetics and molecular cytogenetics can be used for diagnostics, follow up and the study of CTCL etiology, and the results can be combined with TCR analysis. In most cases, samples with cytogenetic clones show dominant TCR gamma clones, but discrepant cases explainable e.g. by sensitivity differences have been detected (Muche et al., 2004). The theories of genotraumatic cells in CTCL and epigenetic progenitor origin of cancer Based on studies of CTCL-lesion-derived cells grown in vitro with emerging chromosomal abnormalities, a hypothesis of polyclonal, ‘genotraumatic’, karyotypically normal but genetically unstable non-malignant cells, developing into tumour cells with chromosomal aberrations (CIN) has been proposed (Kaltoft et al., 1992, 1994; Thestrup-Pedersen et al., 1994). More generally in cancer, such a mutator phenotype, could be caused e.g. by mutations in genes, that control the fidelity of DNA replication, efficacy of DNA repair, or chromosome segregation (Loeb et al., 2003) and could lead to increased mutation rate and chromosomal aberrations. Alternatively, normal mutation rate combined to clonal selection during many successive cell divisions could lead to accumulation of mutations, chromosomal aberrations and cancer (Loeb, 1991; Loeb et al., 2003). Recently chromatin-modifying epigenetic changes like CpG island methylation have been proposed to precede genetic muta-
tions. Global DNA hypomethylation in early stage and gene-specific promoter hypermethylation with increasing frequency with tumor progression have been observed in epithelial malignancies (reviewed by Baylin and Ohm, 2006; and Feinberg et al., 2006). Evidence of epigenetic silencing of several individual genes by methylation, as p15 (CDKN2B) and p16INK4a (CDKN2A), associated especially with progressed disease (Navas et al., 2002; Scarisbrick et al., 2002; Gallardo et al., 2004), MLH1 (Scarisbrick et al., 2003), thrombospondin 4 (THBS4), BCL7A and PTPRG (van Doorn et al., 2005) have been observed in CTCL. According to the epigenetic progenitor model, epigenetic changes in polyclonally dividing stem cells/progenitor cells (Feinberg et al., 2006), e.g. during chronic inflammation or cell injury, precede genetic aberrations that are clonally selected (reviewed by Baylin and Ohm, 2006; Feinberg et al., 2006). Global hypomethylation increases transciptional activity, and may cause chromosomal instability, loss of imprinting and oncogene activation (reviewed by Feinberg et al., 2006). According to the model (Feinberg et al., 2006), solid tumors are formed by a heterogeneous cell population representing both the polyclonal, epigenetically changed cell population and diverse cell lines developed from it with different genetic and epigenetic changes. Late recurrences of local tumors after treatment and metastases may be descendants of the progenitor cell line with only the first epigenetic changes in common with the primary tumour (Feinberg et al., 2006). Such genetic heterogeneity has been shown by CGH in breast cancer (Kuukasjärvi et al., 1997; Schmidt-Kittler et al., 2003). In CTCL, chromosomal sublones and even unrelated chromosomal clones have been observed (Fukuhara et al., 1978; Schlegelberger et al., 1994 with review to several previous studies; Karenko et al., 2003) in the same patient and even in the same tissue. Patients with large cell transformation show chromosomal clonal evolution with changes in ploidy level, aneuploidy and new structural aberrations (D’Alessandro et al., 1992; Prochazkova et al., 2005). The multilineal progression of CTCL is also suggested by microsatellite instability analyses of microdissected tumours (Rübben et al., 2004). Cytogenetic evidence of CTCL is in concordance with early dissemination of the disease. Even if early-stage CTCL usually shows less aberrations and less clone formation than late stage (Whang-Peng et al., 1982), patients with large plaque parapsoriasis may already show chromosomal abnormalities (Whang-Peng et al., 1982; Karenko et al., 1997), even chromosomal clones (Karenko et al., 2003). Chromosomal aberrations, observed by conventional cytogenetics or interphase in situ hybridization, have indicated malignant affision before histopathologically observable malignancy e.g in lymph nodes (Whang-Peng et al., 1982; Karenko et al., 2001) and cytogenetic chromosomal clones in blood during remission precede a relapse (Karenko et al., 1997). Concordantly, TCR clones have been observed in morphologically uninvolved extracutaneous tissues in an early stage CTCL patient (Veelken et al., 1995).
Cytogenet Genome Res 118:353–361 (2007)
355
Chromosomal instability and telomeres Chromosomal instability can be caused by dysfunction of telomeres, which protect chromosome ends against shortening during cell division. Telomeres consist of looping tandem TTAGGG repeats and binding proteins with roles in chromatin regulation, DNA repair, V(D)J and recombination (e.g. RB1, WRN, ATM, Ku; reviewed by Jung and Alt, 2004; Blasco, 2005; Crabbe and Karlseder, 2005; Bailey and Murnane, 2006; Stewart and Weinberg, 2006). In somatic cells telomeres themselves shorten during each cell division leading to permanent growth arrest, which can be bypassed by inactivating Rb and p53 pathways. A crisis with chromosomal instability follows, as telomeres become critically short and fuse producing dicentric chromosomes that break in subsequent cell cycles, followed by apoptosis (reviewed by Greenberg, 2005; Opitz, 2005; Stewart and Weinberg, 2006). In contrast, in germline and embryonic stem cells, telomere length is stablilized by telomerase enzyme. Cancer cells commonly show telomerase activity, short telomeres, and chromosome abnormalities, interpreted as a sign of a crisis in their past, which they have survived by activating telomerase and consequent regeneration of their short telomeres (Artandi et al., 2000; O’Hagan et al., 2002; and following reviews: Zimmermann and Martens, 2005; Chung et al., 2005; Stewart and Weinberg, 2006). Broken chromosomal ends can be repaired with diverse mechanisms, like copying the end of another chromosome with a homologous region, translocation or creation of a new telomere (Stellwagen et al., 2003; and following reviews: Feldser et al., 2003; Bailey and Murnane, 2006; Stewart and Weinberg, 2006). However, some chromosomal instability may persist (Gisselson et al., 2001). In addition to excessive shortening of the telomere sequence, telomere dysfunction can also be caused by mutant telomere proteins. Dysfuctional telomeres may fuse with each other or with DNA double strand breaks, caused by e.g. ionizing irradiation, leaving one chromatid end free and ongoing instability (Bailey et al., 2004; Bailey and Murnane, 2006). Instead of telomerase, some cancer cells conserve their telomeres by an alternative pathway, alternate lengthening of telomeres (reviewed by Muntoni and Reddel, 2005; Stewart and Weinberg, 2006). In T-cell lines derived from CTCL or parapsoriasis skin biopsies and in peripheral blood CD4+ lymphocytes of the patients, a high level of telomerase activity and short telomeres have been found (Wu et al., 1999; Wu and Hansen, 2001). In addition, CD8+ blood lymphocytes of CTCL patients and parapsoriasis patients show shortened telomeres (Wu and Hansen, 2001). Telomerase activity of blood CD8+ cells in CTCL was elevated compared to healthy controls CD4+ cells or B-cells of the CTCL patients, but lower than in CD4+ cells of parapsoriasis patients. This may indicate the effect of increased proliferation activity of reactive, inflammatory cells or the order of events in cancer progression with telomere shortening followed by a rise in telomerase activity (Wu and Hansen, 2001).
356
Cytogenet Genome Res 118:353–361 (2007)
Relative frequencies of aberrations of specific chromosomes in CTCL by conventional cytogenetics Typical to CTCL is a large spectrum of chromosomal abnormalities. Any chromosome can be aberrated, numerically or structurally. Numerical aberrations, especially missing chromosomes are the most common abnormality (Whang-Peng et al., 1982; Mao et al., 2003b). In the 41 cases studied by Whang-Peng et al. (1982), 15 of 41 patients (37%) had numerical aberrations of chromosomes 11, 21 and 22; other common numerical aberrations affected chromosomes 8, 9, 15, 16, 17, 6, 18, 20 observed from 14/41 (34%, chromosome 8) to 10/41 (24%, chromosome 20) of cases. The chromosomes most often affected structurally were 1 (10/41, 24% of cases), and 6, 7, 4, 9, 10, 12, 14, 15 and 17, from 21% to 17% of cases. Mao et al. (2003b) reviewed 166 abnormal SS-karyotypes published in the literature. The most common chromosomes lost were 10 (28% of cases), 9, 2, 17, 1, 13 and 16 (chromosome 16, 14% of cases). The most common structural abnormalities were the deletion of 6q (18% of cases) and i(17q) (17% of cases). Other structurally often aberrant chromosomes (not in order of frequency) were 1, 2, 3, 7, 8, 9, 10, 11, 12, 14, 17 and 19, with aberrations of p or q arm each in 10% to 6% of the cases. Association of numerical aberrations of specific chromosomes with the disease course Clonal or non-clonal chromosomal aberrations in CTCL are statistically non-random (Karenko et al., 1997, 2003). Simultaneous two-colour centromere in situ hybridization, detected by light microscopy, has shown that compared to healthy controls, patients with large plaque parapsoriasis have more aberrations especially in chromosomes 11 and 13, and CTCL patients showed elevated levels of aberrations of 1, 6, 8, 9, 11, 13, 15 and 17 (Karenko et al., 2003). Followup of blood samples has shown that some numerical chromosomal aberrations are observable during remission statistically more often than in healthy control individuals (aberrations of chromosomes 1, 6 and 11) indicating existing disease, but they, as many other chromosome aberrations, especially aberrations of 8 and 17, become more abundant during active or progressive disease (Karenko et al., 2003). The agreement between aberration rate and change of clinical condition was statistically significant for chromosomes 1, 8, 9, 15 and 17 (in situ hybridization or Gbanding; Karenko et al., 2003). Interestingly, CTCL patients have recently been reported to show abnormalities by FISH, and immunohistochemistry in cell cycle control genes CCND1/BCL1 (11q13), and its regulator RB1 (13q14; Mao et al., 2006). Several other genes, possibly involved in cell cycle control, are located in e.g. chromosomes 1, 6 and 12p (Mao et al., 2006). The immunohistochemical expression patterns were complex, and could suggest different stages of tumour cell de-differentiation, tumour- and cell cycle progression, or selection pressures against tumour subclones of different expression patterns (Mao et al., 2006). Recently, combination of molecular cytogenetics (CGH) and microarray gene expression analysis have revealed
chromosomal arms, where both DNA and RNA copy number are changed to the same direction, namely: 1q, 3p, 3q, 4q, 12q, 16p, and 16q. These regions are likely to harbour further gene-level aberrations in CTCL (Hahtola et al., 2006). Defining lost or amplified chromosomal regions by CGH As CTCL cells are difficult to cultivate, comparative genomic hybridization, based on DNA, has been useful. In this method, patient DNA is labeled, in most cases green, and hybridized to normal metaphases with a competing normal DNA sample labeled red. Lost chromosomal regions are shown as red and gained (amplified) regions as green on the metaphases allowing their localization in reference to normal karyotype. The analyses with calculations are done with a computer program. The sensitivity of the method varies 10 Mb for deletions (Kallioniemi et al., 1994) and can be improved to 3 Mb by using standard reference intervals, based on a series of normal samples (Kirchhoff et al., 1999). The length of amplified sequences that can be detected depends on the copy number so that the copy number times the amplicon size is at least 2 Mb (Kallioniemi et al., 1994). Polyploidy weakens the sensitivity, and chromosomes 1pter]p32, 16p, 19, and 22 may show false deletions. Centromeric regions are excluded from the analyses (Kallioniemi et al., 1994). In CTCL, CGH has shown a wide variety of changes, the most common being losses with minimal common regions in 10q25]q26, 13q21]q22 and 17p13]p11 (Karenko et al., 1999) and amplifications in 8q24]q24.3 and 17q21]q25 (Karenko et al., 1999). These findings and some others with a lower frequency (dim 6, enh 7) have been confirmed in various ethnic patient populations (Mao et al., 2002; Fischer et al., 2004), and frequent losses in 1p36]p31 and 19 and gain in 18 reported (Mao et al., 2002). In a German study, the most common pattern was enh(7,8)dim(6q,13). Patients with more than five aberrations or loss in either 6q, 10q, 13q or gain in 7 or 8q showed shortened 5-year survival in comparison to patients without such aberrations (Fischer et al., 2004). In contrast, the most common aberration, loss in 17p, did not shorten survival time. Subsequently, microsatellite markers in 10q have shown loss of heterozygosity (LOH) of the haploinsufficient tumour suppressor gene PTEN and microsatellite instability, associated especially with advanced MF (Scarisbrick et al., 2000). There may be other important genes, too, in this region, since by CGH and LOH studies, more distal parts of 10q are also involved in losses (Karenko et al., 1999), which form two, separate regions (Wain et al., 2005). Recently, additional copies of HER-2/neu gene (ERBB2) (17q11.2]q12), in relation to centromere copies have been detected by FISH in SS (Utikal et al., 2006). Since the sensitivity of conventional CGH is limited, microarray techniques have been developed to achieve gene-level information about DNA copy number aberrations. With array-CGH, areas of chromosomal aberrations detected by CGH can be significantly narrowed to allow further identification of target genes (Pollack et al., 1999).
Structural chromosome aberrations In G-banding studies, two different chromosome 14 translocations involving 14q32 have been reported (Nowell et al., 1986) in two patients. In the other patient, it was in one of three apparently unrelated chromosomally clonal cells (TCR clones were not reported). For i(17q), see above. Recently, multicolour FISH in which every chromosome pair is detected by a chromosome-specific colour combination has been used to study blood lymphocyte metaphases. There are two widely used commercial systems. In MFISH, (multifluor or multiplex FISH) every colour is imaged separately and combined with the computer software, whereas in SKY (spectral karyotyping) the image is analysed spectroscopically (Fourier analysis). Two translocations, der(1)t(1;10)(p2;q2) and der(14)t(14;15)(q;q?) were observed in two patients each with SKY (Mao et al., 2003b). A translocation occurring between centromeres of chromosomes 8 and 17, der(8)t(8;17)(p11;q11), was observed by Thangavelu et al. (1997) with G-banding and by Batista et al. (2006) by SKY. The latter group considered the chromosome 17 centromere to be active, defining the translocation as psu dic (17;8)(p11.2;p11.2). However, it was not detected in the other nine patients studied by centromere in situ hybridizations (Batista et al., 2006). In addition, we have reported (Karenko et al., 2005, supplementary data at www.cancerres.aacrjournals.org) der(8)t(8;17)(?p1;q1) by SKY in one patient. In the study of Batista et al. (2006), recurrent breakpoints in two or three patients each were observed in 1p36]p32, 6q22]q25, 17p13]p11.2, 10q23]q26 and 19p13.3, regions often shown to be lost in previous CGH studies (Karenko et al., 1999; Mao et al., 2002; Fischer et al., 2004). Vermeer et al. (2006) found with a chromosome arm-specific application of the non-commercial multicolour FISH method COBRA (COmbined Binary Ratio labeling FISH; Tanke et al., 1999; Wiegant et al., 2000) no recurrent translocations, but deletion of 10q24 and breakpoints in 17p11 in 3/7 SS patients and isochromosome 17 in one patient. A common deletion or translocation in the NAV3 gene in 12q21 In a 24-colour FISH study of the blood of seven consecutive SS patients, the most common clonal aberration was a translocation or deletion in 12q21]q22 observed in five patients, while one other patient had a loss of the whole chromosome 12 (Karenko et al., 2005). Five of six MF patients studied showed nonclonal deletions in 12q. The breakpoint of one balanced clonal translocation, t(12;18)(q21;q21), and the extension of two long clonal deletions, one from 12q21 to 12q24 and another from 12q12 to 12q21, were studied with locus-specific FISH (YAC and BAC probes). The translocation was proven to have occurred in the 40-exon long gene NAV3 (POMFIL1) dividing the YAC 855F7 and the BAC RP11-494K17 between chromosomes 12q and 18q. The gene was situated in the 7-YAC-long minimal common region of the two long deletions. Locus-specific FISH with two other BACs RP11-136F16 and RP11-36P3, involving the distal part of the gene (exons 10 to 40), indicated the loss of one copy of the corresponding region in relation to chromo-
Cytogenet Genome Res 118:353–361 (2007)
357
some 12 centromere in 4/8 early stage and 11/13 advanced stage patients (Karenko et al., 2005). Homozygous deletions or whole chromosome losses were not studied. NAV3 is one of three human homologues of UNC53, involved in axon guidance in C. elegans. Like NAV2, by structure prediction, NAV3 has calponin-like domains and SH3binding sites suggestive of a role in cell signaling (Coy et al., 2002; Maes et al., 2002). UNC53 interacts with SEM-5, the nematode homologue of GRB2 (Stringham et al., 2002), involved e.g. in proliferative cell signalling in T-lymphocytes (Frauwirth and Thompson, 2002). By RT-PCR, we have found NAV3 expression only in PHA-stimulated but not in resting human lymphocytes. In preliminary analyses, NAV3-siRNA silenced Jurkat cells and peripheral blood lymphocytes showed increased IL-2 production (Karenko et al., 2005). The IL-2-induced T-cell proliferation is mediated by two IL-2R-coupled pathways, one involving activation of Stat5 (Moriggl et al., 1999). IL-2 inducible STAT5 expression has been shown to be shifted from full-length to a truncated form in SS patients (Mitchell et al., 2003), and constitutive expression of STAT3 has been observed in CTCL-derived cells in vitro (Eriksen et al., 2001). Whether NAV3 loss contributes to the STAT-imbalance, has not been confirmed yet. As NAV2, NAV3 is structurally a helicase and exonuclease, resembling Werner and Bloom syndrome proteins with the role in maintaining the stability of chromosomes (WRN, BLM; Coy et al., 2002; Ishiguro et al., 2002; Maes et al.; 2002; Nakayama, 2002, review). Subcellularly, NAV3 has been reported to locate in nuclear pore complexes (Coy et al., 2002) and might have a role in nuclear transport, kinetochore formation and cell cycle control (Fahrenkrog and Aebi, 2003, a review). Four of ten primary neuroblastomas studied by Coy et al. (2002) showed reduced or absent expression of NAV3, and three of them had homozygous deletions of both alleles. In CTCL, a deletion or a translocation was associated with a point mutation in the remaining allele only in one of seven patients studied (Karenko et al., 2005). Thus, NAV3 could be a non-classical haploinsufficient tumour suppressor (Sherr, 2004). Possible epigenetic silencing of NAV3 remains to be studied. In comparison, the mouse NAV2 homologue shows gene dosage effects for behaviour (Peeters et al., 2004). Microarray studies in CTCL
Recently, microarray studies have been performed to reveal gene expression patterns of CTCL. Several novel genes possibly having a role in CTCL pathogenesis have been discovered. Tracey and coworkers identified an expression profile suggesting upregulation of genes involved in the TNF signaling pathway (e.g. TRAF1, BIRC3, TNFSF5 (CD40LG)) among 29 MF skin samples when compared to inflammatory dermatoses with the CNIO OncoChip array (Tracey et al., 2003). Kari and coworkers (2003) found overexpression of many Th2-specific transcription factors (like
358
Cytogenet Genome Res 118:353–361 (2007)
Gata-3 and JunB), as well as RhoB, integrin 1, and proteoglycan 2, while underexpressed genes included CD26 (DPP4), STAT4, and IL-1 receptors among 48 frozen PBMC samples analyzed with a cDNA array containing 4,500 genes. Altogether, a panel of eight genes was identified that could distinguish SS from normal controls, and ten genes were able to classify patients into short term and long term survivors (Kari et al., 2003). In the blood samples of ten Dutch SS patients, analyzed with Affymetrix U95Av2 array, decreased expression of some tumor suppressor genes such as TGF- receptor II was shown, while EPHA4 and TWIST were overexpressed. The latter two were highly expressed also in some lesional skin samples of MF (van Doorn et al., 2004). Identification of single, aberrantly expressed genes not only allows the better understanding of the pathomechanisms of the disease, but may also provide diagnostic aid. Recently, a 5-gene QPCR assay has been developed to aid in the diagnosis of SS including STAT4, GATA3, T-plastin (PLS3), CD1D and TRAIL (TNFSF10) (Nebozhyn et al., 2006). Assays including only a single gene would be easier to implement in the SS diagnostics, but the variation in expression patterns among patients has been too great to allow single gene diagnostics. For instance the expression of Tplastin alone was informative only for 50% of the patients in the study of Nebozhyn and coworkers (2006), and was even lower (30%) in previous array studies (Kari et al., 2003). In our own experiment of altogether 18 SS and MF patients studied e.g. with Affymetrix HGU133A array, downregulation of a set of Th1 specific genes, e.g. TBX21 (T-bet), NKG7, CCL5 (SCYA5, RANTES) was observed in SS and also in a proportion of MF samples thus explaining the previously observed Th2-cytokine profile at the molecular level, and also suggesting that the Th2 polarization of CTCL cells already occurs in the early CTCL stages. In lesional skin, IL7R and CD52 were upregulated suggesting their differential expression to be a hallmark of the skin-infiltrating lymphocytes. This finding explains the proliferative response of CTCL cells to locally produced IL7, and may encourage the broader use of CD52 antibody in the therapy of CTCL (Hahtola et al., 2006). As indicated above, the gene expression signatures obtained have not been uniform. This is most likely due to the different experimental designs, practically varying probe and sample sets as well as the variability in the analysis methods. Interestingly, Kari and coworkers discovered that panels of fewer than ten genes can be used to classify CTCL cases and separate them from controls even when the Sézary cell count is as few as 5%. This might indicate that even a small proportion of morphologically malignant cells is able to produce factors that induce a CTCL-specific expression pattern (Wood, 2005) or that the actual malignant transformation of the cell occurs before the cell turns morphologically malignant (Karenko et al., 2001). Apart from gene expression analysis, genomic microarrays have also been used to study the gene-level aberrations leading to CTCL. Mao and coworkers identified several oncogene copy number gains with AmpliOnc I DNA Array
containing 57 oncogenes, the most significant of which was the amplification of JunB, detected in five of seven cases with MF or SS studied. JUNB was also overexpressed in a larger series of CTCL patients (immunohistochemistry or RT-PCR; Mao et al., 2003a). The CGH of peripheral blood lymphocyte DNA of 21 SS patients on an array of approximately 3,500 BAC sequences (sensitivity 1 Mb for deletions) performed by Vermeer et al. (2006, and personal communication) revealed frequent amplifications and deletions in chromosomal regions mainly reported in earlier studies with conventional CGH (Karenko et al., 1999; Mao et al., 2002; Fischer et al., 2004), e.g. gain of 8q24 in 71% and loss of 17p12 in 67% of cases studied. Amplification of MYC (8q24) and deletion of TP53 (17p13) were shown by quantitative PCR (Vermeer et al., 2006). The main problem in the microarray approach has been the low concordance of results obtained with the same samples on different devices (Marshall et al., 2004). Also, since the malignant cells are relatively sparse and surrounded by reactive T-cells and other normal cells of the skin, methods selectively picking up the morphologically and/or immuno-
histologically/karyotypically malignant cells should be used. In the future, combination of DNA and RNA copy number changes detected by microarray technique would provide further knowledge of the poorly known etiology of CTCL, as well as provide new diagnostic and therapeutic tools. Conclusion
It is obvious that new technologies will allow better classification of CTCL and apparently new subgroups within CTCL will be discovered. The new molecular cytogenetic studies with associated gene-level and expression studies are soon ready to be validated in large patient materials and thereafter used as part of clinical diagnostics and follow-up of patients. These methods will enable the early diagosis of CTCL and will be helpful in follow-up of residual disease during therapy. Ultimately, the results of the molecular cytogenetic studies will pave the way to the development of targeted therapies for CTCL.
References Artandi SE, Chang S, Lee S-L, Alson S, Gottlieb GJ, et al: Telomere dysfunction promotes non-reciprocal translocations and epithelial cancers in mice. Nature 406: 641–645 (2000). Bagot M, Echchakir H, Mami-Chouaib F, DelfauLarue M-H, Charue D, et al: Isolation of tumorspecific cytotoxic CD4+ and CD4+CD8dim+ T-cell clones infiltrating a cutaneous T-cell lymphoma. Blood 91: 4331–4341 (1998). Bailey SM, Murnane JP: Telomeres, chromosome instability and cancer. Nucleic Acids Res 34: 2408–2417 (2006). Bailey SM, Cornforth MN, Ullrich RL, Goodwin EH: Dysfunctional mammalian telomeres join with DNA double-strand breaks. DNA Repair 3: 349–357 (2004). Batista D, Vonderheid EC, Hawkins A, Morsberger L, Long P, et al: Multicolor fluorescence in situ hybridization (SKY) in mycosis fungoides and Sézary syndrome: Search for recurrent chromosome abnormalities. Genes Chromosomes Cancer 45: 383–391 (2006). Baylin SB, Ohm JE: Epigenetic gene silencing in cancer – a mechanism for early oncogenic pathway addiction? Nat Rev Cancer 6: 107–116 (2006). Berger CL, Hanlon D, Kanada D, Dhodapkar M, Lombillo V, et al: The growth of cutaneous Tcell lymphoma is stimulated by immature dendritic cells. Blood 99: 2929–2939 (2002). Berger CL, Tigelaar R, Cohen J, Mariwalla K, Trinh J, et al: Cutaneous T cell lymphoma, malignant proliferation of T-regulatory cells. Blood 105: 1640–1647 (2005). Blasco M: Telomeres and human disease: Ageing, cancer and beyond. Nat Rev Genet 6: 611–622 (2005). Burg G, Dummer R, Haeffner A, Kempf W, Kadin M: From inflamation to neoplasia. Mycosis fungoides evolves from reactive inflammatory conditions (lymphoid infiltrates) transforming into neoplastic plaques and tumors. Arch Dermatol 137:949–952 (2001).
Cerroni L, Hödl S, Rieger E, Soyer HP, Smolle J, Kerl H: Transformation of mycosis fungoides to large cell lymphoma. J Cutan Pathol 17: 290 (1990). Cerroni L, Rieger E, Hödl S, Kerl H: Clinicopathologic and immunologic features associated with transformation of mycosis fungoides to large cell lymphoma. Am J Surg Pathol 16: 543–552 (1992). Chevret E, Prochazkova M, Beylot-Barry M, Merlio J-P: A suggested protocol for obtaining highquality skin metaphases from primary cutaneous T-cell lymphoma. Cancer Genet Cytogenet 167:89–91 (2006). Chung HK, Cheong C, Song J, Lee H-W: Extratelomeric functions of telomerase. Curr Mol Med 5: 233–241 (2005). Clark RA, Chong BF, Mirchandani N, Yamanaka K-I, Murphy GF, et al: A novel method for the isolation of skin resident T cells from normal and diseased human skin. J Invest Dermatol 126:1059–1070 (2006). Coy JF, Wiemann S, Bechmann I, Bächner D, Nitsch R, et al: Pore membrane and/or filament interacting like protein 1 (POMFIL1) is predominantly expressed in the nervous system and encodes different protein isoforms. Gene 290: 73–94 (2002). Crabbe L, Karlseder J: In the end, it’s all structure. Curr Mol Med 5: 135–143 (2005). D’Alessandro E, De Pasquale A, Ligas C, Lo Re ML, Di Cola M, et al: Cytogenetic findings in terminal large cell transformation in a case of Sézary syndrome. Cancer Genet Cytogenet 58: 100– 104 (1992). Delfau-Larue M-H, Laroche L, Wechsler J, Lepage E, Lahet C, et al: Diagnostic value of dominant T-cell clones in peripheral blood in 363 patients presenting consecutively with a clinical suspicion of cutaneous lymphoma. Blood 96: 2987– 2992 (2000).
Dummer R, Michie SA, Kell D, Gould JW, Haeffner AC, et al: Expression of bcl-2 protein and Ki-67 nuclear proliferation antigen in benign and malignant cutaneous T-cell infiltrates. J Cutan Pathol 22: 11–17 (1995). Eriksen KW, Kaltoft K, Mikkelsen G, Nielsen M, Zhang Q, et al: Constitutive STAT3-activation in Sezary syndrome: tyrphostin AG490 inhibits STAT3-activation, interleukin-2 receptor expression and growth of leukemic Sezary cells. Leukemia 15:787–793 (2001). Fahrenkrog B, Aebi U: The nuclear pore complex: nucleocytoplasmic transport and beyond. Nat Rev Mol Cell Biol 4: 757–766 (2003). Feinberg AP, Ohlsson R, Henikoff S: The epigenetic progenitor origin of human cancer. Nat Rev Genet 7:21–33 (2006). Feldser DM, Hackett JA, Greider CW: Telomere dysfunction and the initiation of genome instability. Nat Rev Cancer 3:1–5 (2003). Ferenczi K, Fuhlbrigge RC, Pinkus JL, Pinkus GS, Kupper TS: Increased CCR4 expression in cutaneous T-cell lymphoma. J Invest Dermatol 119:1405–1410 (2002). Fischer TC, Gellrich S, Muche MJ, Sherev T, Audring H, et al: genomic aberrations and survival in cutaneous T cell lymphomas. J Invest Dermatol 122:579–586 (2004). Fraser-Andrews EA, Woolford AJ, Russell-Jones R, Seed PT, Whittaker SJ: Detection of a peripheral blood T cell clone is an independent prognostic marker in mycosis fungoides. J Invest Dermatol 114:117–121 (2000). Frauwirth KA, Thompson GB: Activation and inhibition of lymphocytes by costimulation. J Clin Invest 109:295–299 (2002). Fukuhara S, Rowley JD, Variakojis D: Banding studies of chromosomes in a patient with mycosis fungoides. Cancer 42: 2262–2268 (1978). Gallardo F, Esteller M, Pujol RM, Costa C, Estrach T, Servitje O: Methylation status of the p15, p16 and MGMT promoter genes in primary cutaneous T-cell lymphomas. Haematologica 89: 1401–1403 (2004).
Cytogenet Genome Res 118:353–361 (2007)
359
Garatti SA, Roscetti E, Trecca, Fracciolla NS, Neri A, Berti E: bcl-1, bcl-2, p53,c-myc, and lyt-10 analysis in cutaneous T-cell lymphomas. Recent Result Cancer Res 139: 249–261 (1995). Girardi M, Heald PW, Wilson LD: The pathogenesis of mycosis fungoides. N Engl J Med 350: 1978–1988 ( 2004). Gisselsson D, Jonson T, Petersén Å, Strömbeck B, Dal Cin P, et al: Telomere dysfunction triggers extensive DNA fragmentation and evolution of complex chromosome abnormalities in human malignant tumors. Proc Natl Acad Sci USA 98: 12683–12688 (2001). Gollin SM: Chromosomal instability. Curr Opin Oncol 16:25–31 (2004). Gorochov G, Bachelez H, Cayuela JM, Legac E, Laroche L, et al: Expression of V· gene segments by Sézary cells. J Invest Dermatol 105: 56–61 (1995). Greenberg RA: Telomeres, crisis and cancer. Curr Mol Med 5: 213–218 (2005). Haeffner AC, Smoller BR, Zepter K, Wood GS: Differentiation and clonality of lesional lymphocytes in small plaque parapsoriasis. Arch Dermatol 131:321–328 (1995). Hahtola S, Tuomela S, Elo L, Häkkinen T, Karenko L, et al: Th1 response and cytotoxicity genes are downregulated in cutaneous T-cell lymphoma. Clin Cancer Res 12: 4812–4821 (2006). Ishiguro H, Shimokawa T, Tsunoda, T, Tanaka T, Fujii Y, et al: Isolation of HELAD1, a novel human helicase gene up-regulated in colorectal carcinomas. Oncogene 21:6387–6394 (2002). Jung D, Alt FW: Unraveling V(D)J recombination: Insights into gene regulation. Cell 116:299–311 (2004). Kallioniemi O-P, Kallioniemi A, Piper J, Isola J, Waldman FM, et al: Optimizing comparative genomic hybridization for analysis of DNA sequence copy number changes in solid tumors. Genes Chromosomes Cancer 10: 231–243 (1994). Kaltoft K, Bisballe S, Dyrberg T, Boel E, Rasmussen PB, Thestrup-Pedersen K: Establishment of two continuous T-cell strains from a single plaque of a patient with mycosis fungoides. In Vitro Cell Dev Biol 28A:161–167 (1992). Kaltoft K, Hansen BH, Thestrup-Pedersen K: Cytogenetic findings in cell lines from cutaneous Tcell lymphoma. Dermatologic Clinics 12: 295– 304 (1994). Karenko L, Hyytinen E, Sarna S, Ranki A: Chromosomal abnormalities in cutaneous T-cell lymphoma and in its premalignant conditions as detected by G-banding and interphase cytogenetic methods. J Invest Dermatol 108: 22–29 (1997). Karenko L, Kähkönen M, Hyytinen E-R, Lindlöf M, Ranki A: Notable losses at specific regions of chromosomes 10q and 13q in the Sézary syndrome detected by comparative genomic hybridization. J Invest Dermatol 112: 392–395 (1999). Karenko L, Nevala H, Raatikainen M, Franssila K, Ranki A: Chromosomally clonal T-cells in the skin, blood or lymph nodes of two Sézary syndrome patients express CD45RA, CD45RO, CDw150, and interleukin-4, but no interleukin2 or interferon-gamma. J Invest Dermatol 116: 188–193 (2001). Karenko L, Sarna S, Kähkönen M, Ranki A: Chromosomal abnormalities in relation to clinical disease in patients with cutaneous T-cell lymphoma: a 5-year follow-up study. Br J Dermatol 148:55–64 (2003).
360
Karenko L, Hahtola S, Päivinen S, Karhu R, Syrjä S, et al: Primary cutaneous T-cell lymphomas (CTCL) show a deletion or translocation affecting NAV3, the human UNC-53 homologue. Cancer Res 65: 8101–8110 (2005). Kari L, Loboda A, Nebozhyn M, Rook AH, Vonderheid EC, et al: Classification and prediction of survival in patients with the leukemic phase of cutaneous T cell lymphoma. J Exp Med 197: 1477–1488 (2003). Kikuchi A, Naka W, Harada T, Sakuraoka K, Harada R, Nishikawa T: Parapsoriasis en plaques: Its potential for progression to malignant lymphoma. J Am Acad Dermatol 29: 419–422 (1993). Kirchhoff M, Gerdes T, Maahr J, Rose H, Bentz M, et al: Deletions below 10 megabasepairs are detected in comparative genomic hybridization by standard reference intervals. Genes Chromosomes Cancer 25: 410–413 (1999). Kuukasjärvi T, Karhu R, Tanner M, Kähkönen M, Schäffer A, et al: Genetic heterogeneity and clonal evolution underlying development of asynchronous metastasis in human breast cancer. Cancer Res 57: 1597–1604 (1997). Li G, Chooback L, Wolfe JT, Rook AH, Felix CA, et al: Overexpression of p53 protein in cutaneous T cell lymphoma: Relationship to large cell transformation and disease progression. J Invest Dermatol 110: 767–770 (1998). Linnemann T, Gellrich S, Lukowsky A, Mielke A, Audring H, et al: Polyclonal expansion of T cells with the TCR V· type of the tumour cell in lesions of cutaneous T-cell lymphoma: evidence for possible superantigen involvement. Br J Dermatol 150:1013–1017 (2004). Loeb LA: Mutator phenotype may be required for multistage carcinogenesis. Cancer Res 51: 3075–3079 (1991). Loeb LA, Loeb KR, Anderson JP: Multiple mutations and cancer. Proc Natl Acad Sci USA 100: 776–781 (2003). Lukowsky A, Muche JM, Sterry W, Audring H: Detection of expanded T cell clones in skin biopsy samples of patients with lichen sclerosus et atrophicus by T cell receptor-␥ polymerase chain reaction assays. J Invest Dermatol 115: 254–259 (2000). Maes T, Barceló A, Buesa C: Neuron navigator: A human gene family with homology to unc-53, a cell guidance gene from Caenorhabditis elegans. Genomics 80: 21–30 (2002). Mao X, Lillington D, Scarisbrick JJ, Mitchell T, Czepulkowski B, et al: Molecular cytogenetic analysis of cutaneous T-cell lymphomas: identification of common genetic alterations in Sézary syndrome and mycosis fungoides. Br J Dermatol 147:464–475 (2002). Mao X, Orchard G, Lillington DM, Russell-Jones R, Young BD, Whittaker SJ: Amplification and overexpression of JUNB is associated with primary cutaneous T-cell lymphomas. Blood 101: 1513–1519 (2003a). Mao X, Lillington DM, Czepulkowski B, RussellJones R, Young BD, Whittaker S: Molecular cytogenetic characterization of Sézary syndrome. Genes Chromosomes Cancer 36: 250–260 (2003b). Mao X, Orchard G, Vonderheid EC, Nowell PC, Bagot M, et al: Heterogeneous abnormalities of CCND1 and RB1 in primary cutaneous T-cell lymphomas suggesting impaired cell cycle control in disease pathogenesis. J Invest Dermatol 126:1388–1395 (2006). Marrogi AJ, Khan MA, Vonderheid EC, Wood GS, McBurney E: p53 tumor suppressor gene mutations in transformed cutaneous T-cell lymphoma: A study of 12 cases. J Cutan Pathol 26:369– 378 (1999).
Cytogenet Genome Res 118:353–361 (2007)
Marshall E: Getting the noise out of gene arrays. Science 306:630–631 (2004). Mitchell TJ, Whittaker SJ, John S: Dysregulated expression of COOH-terminally truncated Stat5 and loss of IL2-inducible Stat5-dependent gene expression in Sézary syndrome. Cancer Res 63: 9048–9054 (2003). Morgan SM, Hodges E, Mitchell TJ, Harris S, Whittaker SJ, Smith JL: Molecular analysis of T-cell receptor  genes in cutaneous T-cell lymphoma reveals J1 bias. J Invest Dermatol 126: 1893– 1899 Moriggl R, Topham DJ, Teglund S, Sexl V, McKay C, et al: Stat 5 is required for IL-2-induced cell cycle progression of peripheral T cells. Immunity 10:249–259 (1999). Morozov VA, Syrtsev AV, Ellerbrok H, Nikolaeva EV, Bavykin AS, Pauli G: Mycosis fungoides in European Russia: No antibodies to human T cell leukemia virus type 1 structural proteins, but virus-like sequences in blood and saliva. Intervirology 48: 362–371 (2005). Muche JM, Lukowsky A, Asadullah K, Gellrich S, Sterry W: Demonstration of frequent occurrence of clonal T cells in the peripheral blood of patients with primary cutaneous T-cell lymphoma. Blood 90: 1636–1642 (1997). Muche M, Lukowsky A, Heim J, Friedrich M, Audring H, Sterry W: Demonstration of frequent occurrence of clonal T cells in the peripheral blood but not in the skin of patients with small plaque parapsoriasis. Blood 94:1409–1417 (1999). Muche JM, Lukowsky A, Ahnhudt C, Gellrich S, Sterry W: Peripheral blood T cell clonality in mycosis fungoides – an independent prognostic marker? J Invest Dermatol 115: 504–505 (2000). Muche JM, Karenko L, Gellrich S, Karhu R, Kytölä S, et al: Cellular coincidence of clonal T-cell receptor rearrangements and complex clonal chromosomal aberrations – a hallmark of malignancy in cutaneous T-cell lymphoma. J Invest Dermatology 122: 574–578 (2004). Muntoni A, Reddel RR: The first molecular details of ALT in human tumor cells. Hum Mol Genet 14: R191-R196 (2005). Nakayama H: RecQ family helicases: roles as tumor suppressor proteins. Oncogene 21: 9008–9021 (2002). Navas IC, Ortiz-Romero PL, Villuendas R, Martinez P, Garcia C, et al: p16INK4a gene alterations are frequent in lesions of mycosis fungoides. Am J Pathol 156: 1565–1572 (2000). Navas IC, Algara P, Mateo M, Martinez P, García C, et al: P16INK4a is selectively silenced in the tumoral progression of mycosis fungoides. Lab Invest 82: 123–132 (2002). Nebozhyn M, Loboda A, Kari L, Rook AH, Vonderheid EC, et al: Quantitative PCR on 5 genes reliably identifies CTCL patients with 5–99% circulating tumor cells with 90% accuracy. Blood 107:3189–3196 (2006). Nevala H, Karenko L, Ranki A: Proapoptotic and antiapoptotic markers in cutaneous T-cell lymphoma skin infiltrates and in comparison with lymphomatoid papulosis. Brit J Dermatol 145: 928–937 (2001). Nowell PC, Vonderheid EC, Besa I, Hoxie JA, Moreau L, Finan JB: The most common chromosome change in 86 chronic B cell or T cell tumors: a 14q32 translocation. Cancer Genet Cytogenet 19: 219–227 (1986). O’Hagan R, Chang S, Maser RS, Mohan R, Artandi SE, et al: Telomere dysfunction provokes regional amplification and deletion in cancer genomes. Cancer Cell 2: 149–155 (2002).
Opitz OG: Telomeres, telomerase and malignant transformation. Curr Mol Med 5: 219–226 (2005). Pawlaczyk M, Filas V, Sobieska M, Gozdzicka-Jozefiak A, Wiktorowicz K, Breborowicz J: No evidence of HTLV-1 infection in patients with mycosis fungoides and Sézary syndrome. Neoplasma 52: 52–55 (2005). Peeters PJ, Baker A, Goris I, Daneels G, Verhasselt P, et al: Sensory deficits in mice hypomorphic for a mammalian homologue of unc-53. Dev Brain Res 150:89–101 (2004). Peris K, Cerroni L, Kerl H, Radaszkiewicz T, Chimenti S, Höfler H: Expression of c-myc in cutaneous lymphomas and pseudolymphomas. Dermatologica 183: 1–6 (1991). Pollack JR, Perou CM, Alizadeh AA, Eisen MB, Pergamenschikov A, et al: Genome-wide analysis of DNA copy-number changes using cDNA microarrays. Nat Genet 23: 41–46 (1999). Prochazkova M, Chevret E, Beylot-Barry M, Vergier B, Sobotka J, Merlio J-P: Large-cell transformation of mycosis fungoides: tetraploidization within skin tumor large cells. Cancer Genet Cytogenet 163: 1–6 (2005). Ranki A, Niemi K-M, Nieminen P, Krohn K: Antibodies against retroviral core proteins in relation to disease outcome in patients with mycosis fungoides. Arch Derm Res 282: 532–538 (1990). Rübben A, Kempf W, Kadin ME, Zimmermann DR, Burg G: Multilineage progression of genetically unstable tumor subclones in cutaneous T-cell lymphoma. Exp Dermatol 13: 472–483 (2004). Saed G, Fivenson DP, Naidu Y, Nickoloff BJ: Mycosis fungoides exhibits a Th1-type cell-mediated cytokine profile whereas Sézary syndrome expresses a Th2-type profile. J Invest Dermatol 103:29–33 (1994). Sawabe T, Shiokawa S, Sugisaki K, Tsuda T, Yamamoto K: Accumulation of common clonal T cells in multiple lesions of sarcoidosis. Mol Med 9:793–802 (2000). Scarisbrick JJ, Woolford AJ, Russell-Jones R, Whittaker SJ: Loss of heterozygosity on 10q and microsatellite instability in advanced stages of primary cutaneous T-cell lymphoma and possible association with homozygous deletion of PTEN. Blood 95: 2937–2942 (2000). Scarisbrick, JJ, Woolford AJ, Calonje E, Photiou A, Ferreira S, et al: Frequent abnormalities of the p15 and p16 genes in mycosis fungoides and Sezary syndrome. J Invest Dermatol 118: 493– 499 (2002). Scarisbrick JJ, Mitchell TJ, Calonje E, Orchard G, Russell-Jones R, Whittaker SJ: Microsatellite instability is associated with hypermethylation of the hMLH1 gene and reduced gene expression in mycosis fungoides. J Invest Dermatol 121:894–901 (2003). Schlegelberger B, Weber-Matthiesen K, Sterry W, Bartels H, Sonnen R, et al: Combined immunophenotyping and karyotyping in peripheral T cell lymphomas demonstrating different clonal and nonclonal chromosome aberrations in T helper cells. Leuk Lymphoma 15: 113–125 (1994). Schmidt-Kittler O, Ragg T, Daskalakis A, Granzow M, Ahr A, et al: From latent disseminated cells to overt metastasis: Genetic analysis of systemic breast cancer progression. Proc Natl Acad Sci USA 100:7737–7742 (2003).
Sherr CJ: Principles of tumor suppression. Cell 116: 235–246 (2004). Stellwagen AE, Haimberger ZW, Veatch JR, Gottschling DE: Ku interacts with telomerase RNA to promote telomere addition at native and broken chromosome ends. Gen Dev 17: 2384–2395 (2003). Stewart SA, Weinberg RA: Telomeres: Cancer to human aging. Annu Rev Cell Dev Biol 22: 531– 557 (2006). Stringham E, Pujol N, Vandekerchove J, Bogaert T: unc-53 controls longitudinal migration in C. elegans. Development 129: 3367–3379 (2002). Su M-W, Dorocicz, I, Dragowska WH, Ho V, Li G, et al: Aberrant expression of T-plastin in Sézary cells. Cancer Res 63: 7122–7127 (2003). Tanke HJ, Wiegant J, van Gijlswijk RPM, Bezrookove V, Pattenier H, et al: New strategy for multicolour fluorescence in situ hybridisation: COBRA: COmbined Binary RAtio labelling. Eur J Hum Genet 7:2–11 (1999). Thangavelu M, Finn WG, Yelawarthi KK, Roenigk HH Jr, Samuelson E, et al: Recurring structural chromosome abnormalities in peripheral blood lymphocytes of patients with mycosis fungoides/Sézary syndrome. Blood 89: 3371–3377 (1997). Thestrup-Pedersen K, Kaltoft K: Genotraumatic T cells and cutaneous T-cell lymphoma. A causal relationship? Arch Dermatol Res 287: 97–101 (1994). Tiemessen MM, Mitchell TJ, Hendry L, Whittaker SJ, Taams LS, John S: Lack of suppressive CD4+, CD25+, FOXP3+ T cells in advanced stages of primary cutaneous T-cell lymphoma. J Invest Dermatol 126:2217–2223 (2006) Tracey L, Villuendas R, Dotor AM, Spiteri I, Ortiz P, et al: Mycosis fungoides shows concurrent deregulation of multiple genes involved in the TNF signaling pathway: An expression profile study. Blood 102:1042–1050 (2003). Utikal J, Poenitz N, Gratchev A, Klemke K-D, Nashan D, et al: Additional Her/neu gene copies in patients with Sézary syndrome. Leukemia Res 30:755–760 (2006). Väkevä L, Pukkala E, Ranki A: Increased risk of secondary cancers in patients with primary cutaneous T cell lymphoma. J Invest Dermatol 115:62–65 (2000). Väkevä L, Sarna S, Vaalasti A, Pukkala E, Kariniemi A-L, Ranki A: A retrospective study of the probablility of the evolution of parapsoriasis en plaques into mycosis fungoides. Acta Derm Venereol 85: 318–323 (2005). van Doorn R, Dijkman R, Vermeer MH, Out-Luiting JJ, van der Raaij-Helmer EMH, et al: Aberrant expression of the tyrosine kinase receptor EphA4 and the transcription factor twist in Sézary syndrome identified by gene expression analysis. Cancer Res 64: 5578–5586 (2004). Van Doorn R, Zoutman WH, Dijkman R, de Menezes RX, Commandeur S, et al: Epigenetic profiling of cutaneous T-cell lymphoma: Promoter hypermethylation of multiple tumor suppressor genes including BCL7a, PTPRG, and p73. J Clin Oncol 23:3886–3896 (2005). Veelken H, Wood GS, Sklar J: Molecular staging of cutaneous T-cell lymphoma: Evidence for systemic involvement in early disease. J Invest Dermatol 104:889–894 (1995). Vega F, Luthra R, Medeiros J, Dunmire V, Lee S-J, et al: Clonal heterogeneity in mycosis fungoides and its relationship to clinical course. Blood 100:3369–3373 (2002).
Vermeer MH, Dijkman R, Mao X, Doorn R, Whittaker S, et al: Array CGH and COBRA-FISH reveal recurrent genetic alterations in Sézary syndrome. 36th annual European Society for Dermatolocigal Research meeting, Paris, France, 7–9 September, 2006. Abstract J Invest Dermatology 126 Suppl 3:s46 (2006). Wain EM, Mitchell TJ, Russell-Jones R, Whittaker SJ: Fine mapping of chromosome 10q deletions in Mycosis Fungoides and Sézary syndrome: identification of two discrete regions of deletion at 10q23.33–q24.1 and 10q24.33–q25.1. Genes Chromosomes Cancer 42: 184–192 (2005). Whang-Peng, J, Bunn PA, Knutsen T, Matthews MJ, Schechter G, Minna JD: Clinical implications of cytogenetic studies in cutaneous T-cell lymphoma (CTCL). Cancer 50: 1539–1553 (1982). Whittaker SJ, Marsden JR, Spittle M, Russell Jones R: Joint British association of dermatologist and UK cutaneous lymphoma group guidelines for the management of primary cutaneous Tcell lymphomas. Br J Dermatol 149: 1095–1107 (2003). Wiegant J, Bezrookove V, Rosenberg C, Tanke HJ, Raap AK, et al: Differentially painting human chromosome arms with combined binary ratio-labeling fluorescence in situ hybridization. Genome Res 10: 861–865 (2000). Willemze R, Kerl H, Sterry W, Berti E, Cerroni L, et al: EORTC classification for primary cutaneous lymphomas: A proposal from the cutaneous lymphoma study group of the European Organization for Research and Treatment of Cancer. Blood 90:354–371 (1997). Willemze R, Jaffe ES, Burg G, Cerroni L, Berti E, et al: WHO-EORTC classification for cutaneous lymphomas. Blood 105: 3768–3785 (2005). Wolfe JT, Chooback L, Finn DT, Jaworsky C, Rook AH, Lessin SR: Large-cell transformation following detection of minimal residual disease in cutaneous T-cell lymphoma: Molecular and in situ analysis of a single neoplastic T-cell clone expressing the identical T-cell receptor. J Clin Oncol 13:1751–1757 (1995). Wood GS, Tung RM, Haeffner AC, Crooks CF, Liao S, et al: Detection of clonal T-cell receptor ␥ gene rearrangements in early Mycosis Fungoides/Sézary syndrome by polymerase chain reaction and denaturing gradient gel electrophoresis (PCR/DGGE). J Invest Dermatol 103: 34–41 (1994). Wood GS: cDNA microarrays and cutaneous oncology. Arch Dermatol 141: 632 (2005). Wu K, Hansen ER: Shortened telomere length is demonstrated in T-cell subsets together with a pronounced increased telomerase activity in CD4 postivie T cells from blood of patients with mycosis fungoides and parapsoriasis. Exp Dermatol 10:329–336 (2001). Wu K, Lund M, Bang K, Thestrup-Pedersen K: Telomerase activity and telomere length in lymphocytes from patients with cutaneous Tcell lymphoma. Cancer 86: 1056–1063 (1999). Yawalkar N, Ferenczi K, Jones DA, Yamanaka K, Suh K-Y, et al: Profound loss of T-cell receptor repretoire complexity in cutaneous T-cell lymphoma. Blood 102: 4059–4066 (2003). Zimmermann S, Martens UM: Telomere dynamics in hematopoietic stem cells. Curr Mol Med 5: 179–185 (2005).
Cytogenet Genome Res 118:353–361 (2007)
361
Author Index Vol. 118, No. 2–4, 2007
Adam, P. 328 Al-Taie, O. 214 Alvarez, S. 304 Anai, S. 204 Auer, R.L. 310
Hoehn, H. 166 Hogue-Angeletti, R. 92 Hop, W.C.J. 130 Huebner, K. 196 Hurst, C.D. 166
Barnoski, B.L. 196 Belbin, T.J. 92 Bergman, A. 92 Bernicot, I. 345 Boehlein, S.K. 204 Brandwein-Gensler, M. 92 Bremer, S.W. 237
Iczkowski, K.A. 204 Iliopoulos, D. 196
Calcagnile, O. 270 Callanan, M.B. 320 Cannizzaro, L.A. 91 Chaganti, R.S.K. 337 Chen, Q. 92 Childs, G. 92 Cigudosa, J.C. 304 Cotter, F.E. 310 Croce, C.M. 196, 252 De Braekeleer, M. 345 Dinjens, W.N.M. 130 Donnelly, G. 337 Douet-Guilbert, N. 345 Druck, T. 196 Ducharme-Smith, A.L. 260 Florin, A. 320 Florl, A.R. 166 Fong, L.Y.Y. 196 Fournier, A. 320 Friedl, R. 166 Ganapathiraju, S.C. 260 Garg, M. 92 Geurts van Kessel, A. 157 Gisselsson, D. 270 Goodison, S. 116, 204 Griffin, C.A. 148 Haddadin, M. 148 Hader, C. 166 Hagan, J.P. 252 Hahtola, S. 353 Haigentz, M. 92 Hameed, M. 138 Hanenberg, H. 166 Hartmann, E. 214 Hartmann, F.H. 166 Hartmann, M. 328 Hawkins, A.L. 148 Head and Neck Cancer Group 92 Hecht, F. 222 Heim, S. 190 Heino, S. 277 Heng, H.H.Q. 237 Herry, A. 345 Herterich, S. 166
Fax +41 61 306 12 34 E-Mail
[email protected] www.karger.com
© 2007 S. Karger AG, Basel
Jaffee, E. 148 James, C.D. 260 Jhanwar, S.C. 337 Kalb, R. 166 Karenko, L. 353 Kimura, A. 297 Knowles, M.A. 166 Knuutila, S. 277 Kosari, F. 260 Koss, L.G. 247 Kudlich, T. 214 Kuiper, R.P. 157 Le Bris, M.-J. 345 Lefebvre, C. 320 Leroux, D. 320 Lindholm, P.M. 277 Liu, G. 237 Lührs, H. 214 Maisch, S. 214 Mancini, R. 196 McAvoy, S. 260 McCorkell, K.A. 196 Medendorp, K. 157 Mehra, S. 337 Melcher, R. 214 Menzel, T. 214 Micci, F. 190 Mihara, K. 297 Moadel, R. 92 Morel, F. 345 Morsberger, L. 148 Müller-Hermelink, H.K. 328 Myllykangas, S. 277 Nakamura, K. 204 Nanda, I. 166 Nanjangud, G. 337 Negassa, A. 92 Neveling, K. 166
Qin, J. 337 Ramesh, K.H. 91 Ranki, A. 353 Rao, P.H. 337 Ried, T. 148 Rosenwald, A. 214, 328 Rosser, C.J. 204 Sandberg, A.A. 182 Savola, S. 277 Schepens, M. 157 Scheppach, W. 214 Schiff, B. 92 Schindler, D. 166 Schlecht, N.F. 92 Schmid, M. 166, 214 Schoenmakers, E.F.P.M. 157 Schoof, J. 328 Schrock, E. 148 Schulz, W.A. 166 Schwarz, S. 328 Shifteh, K. 92 Siersema, P.D. 130 Siprashvili, Z. 196 Siracusa, L.D. 196 Smith, D.I. 260 Smith, R.V. 92 Solly, F. 320 Sreekantaiah, C. 284 Stallings, R.L. 110 Steinlein, C. 166, 214 Suela, J. 304 Takihara, Y. 297 Teruya-Feldstein, J. 337 Thijssen, J. 157 Tibiletti, M.G. 229 Tilanus, H.W. 130 Tönnies, H. 166 Urquidi, V. 116 van Dekken, H. 130 van den Hurk, W.H. 157 van Duin, M. 130 van Groningen, J.J.M. 157 van Marion, R. 130 Vauhkonen, H. 277 Vissers, K.J. 130 Vreede, L. 157
Ott, G. 328 Ott, M. 328 Owen, R. 92
Wolff, D.J. 177
Patel, A. 148 Perez, D.S. 260 Perlman, E.J. 148 Pritchett, J.R. 260 Prystowsky, M.B. 92 Puppe, B. 328
Zanesi, N. 196 Zelenetz, A.D. 337 Zheng, X. 92 Zhu, Y. 260
Accessible online at: www.karger.com/cgr
Ye, C.J. 237
Author Index Vol. 118, 2007
Adam, P. 328 Aktas, D. 31 Al-Taie, O. 214 Alvarez, S. 304 Anai, S. 204 Annerén, G. 1 Auer, R.L. 310
Ferreira, I.A. 78 Florin, A. 320 Florl, A.R. 166 Fong, L.Y.Y. 196 Fournier, A. 320 Friedl, R. 166
Barnoski, B.L. 196 Barreiro, C. 84 Basaran, S. 38 Baumer, A. 38 Belbin, T.J. 92 Bennewitz, J. 67 Bergman, A. 92 Bernicot, I. 345 Bertollo, L.A.C. 78 Bocian, E. 31 Boduroglu, K. 31 Boehlein, S.K. 204 Boettcher, D. 67 Bondeson, M.-L. 1 Borg, Å. 13 Brandwein-Gensler, M. 92 Bremer, S.W. 237 Brenig, B. 55 Breuning, M.H. 19 Broad, T.E. 55 Brunner, H.G. 19 Calcagnile, O. 270 Callanan, M.B. 320 Cannizzaro, L.A. 91 Cefle, K. 38 Chaganti, R.S.K. 337 Chen, Q. 92 Childs, G. 92 Chmurzynska, A. 63 Cigudosa, J.C. 304 Coccé, M.C. 84 Cotter, F.E. 310 Croce, C.M. 196, 252 de Menezes, R.X. 19 De Braekeleer, M. 345 den Dunnen, J.T. 19 Dinjens, W.N.M. 130 Donev, R. 42 Donnelly, G. 337 Douet-Guilbert, N. 345 Druck, T. 196 Ducharme-Smith, A.L. 260 Dumanski, J.P. 1 Edeby, C. 1 Eick, G.N. 72 Ellis, P. 1 Enukashvily, N.I. 42
Fax +41 61 306 12 34 E-Mail
[email protected] www.karger.com
© 2007 S. Karger AG, Basel
Gallego, M.S. 84 Ganapathiraju, S.C. 260 Garg, M. 92 Geurts van Kessel, A. 19, 157 Giordano, P.C. 19 Gisselsson, D. 270 Goodison, S. 116, 204 Griffin, C.A. 148 Haddadin, M. 148 Hader, C. 166 Hagan, J.P. 252 Hahtola, S. 353 Haigentz, M. 92 Hall, A.J. 55 Hameed, M. 138 Hanenberg, H. 166 Hansen, M. 55 Hartmann, E. 214 Hartmann, F.H. 166 Hartmann, M. 328 Hawkins, A.L. 148 Head and Neck Cancer Group 92 Hecht, F. 222 Heidenblad, M. 13 Heim, S. 13, 190 Heino, S. 277 Heng, H.H.Q. 237 Herry, A. 345 Herterich, S. 166 Hoehn, H. 166 Hogue-Angeletti, R. 92 Hop, W.C.J. 130 Hovland, R. 13 Huebner, K. 196 Hurst, C.D. 166 Iczkowski, K.A. 204 Iliopoulos, D. 196 Jacobs, D.S. 72 Jaffee, E. 148 James, C.D. 260 Jhanwar, S.C. 337 Johansson, B. 13 Kalay, E. 19 Kalb, R. 166 Karenko, L. 353 Kimura, A. 297 Knorr, C. 55
Accessible online at: www.karger.com/cgr
Knowles, M.A. 166 Knuutila, S. 8, 277 Kosari, F. 260 Koss, L.G. 247 Kosyakova, N. 31 Kudlich, T. 214 Kuechler, A. 31 Kuiper, R.P. 157 Langford, C. 1 Le Bris, M.-J. 345 Lefebvre, C. 320 Lehesjoki, A.E. 19 Leroux, D. 320 Liehr, T. 31 Lindholm, P.M. 277 Liu, G. 237 Lührs, H. 214 Lundin, C. 13 Maak, S. 67 Maisch, S. 214 Mancini, R. 196 Martins, C. 78 McAvoy, S. 260 McCorkell, K.A. 196 Medendorp, K. 157 Mehra, S. 337 Melcher, R. 214 Menzel, T. 214 Micci, F. 190 Mihara, K. 297 Moadel, R. 92 Morel, F. 345 Morsberger, L. 148 Mrasek, K. 31 Müller-Hermelink, H.K. 328 Myllykangas, S. 277 Nakamura, K. 204 Nanda, I. 166 Nanjangud, G. 337 Negassa, A. 92 Neveling, K. 166 Obregon, M.G. 84 Ott, G. 328 Ott, M. 328 Owen, R. 92 Ozturk, S. 38 Palanduz, S. 38 Patel, A. 148 Paul, S. 67 Perez, D.S. 260 Perlman, E.J. 148 Pietracz, J. 31 Podgornaya, O.I. 42
363
Pritchett, J.R. 260 Prystowsky, M.B. 92 Puppe, B. 328 Qin, J. 337 Ramesh, K.H. 91 Ranki, A. 353 Rao, P.H. 337 Rauch, A. 31 Ried, T. 148 Rosenwald, A. 214, 328 Rosser, C.J. 204 Salido, M. 84 Sandberg, A.A. 182 Savola, S. 277 Schepens, M. 157 Scheppach, W. 214 Schiff, B. 92 Schindler, D. 166 Schinzel, A. 38 Schlecht, N.F. 92 Schmid, M. 166, 214 Schoenmakers, E.F.P.M. 157 Schoof, J. 328 Schrock, E. 148 Schulz, W.A. 166 Schwarz, S. 328 Shifteh, K. 92
364
Cytogenet Genome Res Vol. 118, 2007
Siersema, P.D. 130 Siprashvili, Z. 196 Siracusa, L.D. 196 Smith, D.I. 260 Smith, R.V. 92 Solé, F. 84 Solly, F. 320 Sreekantaiah, C. 284 Stallings, R.L. 110 Steinlein, C. 166, 214 Strömbeck, B. 13 Suela, J. 304 Swalve, H.H. 67 Switonski, M. 63 Szczerbal, I. 63 Takihara, Y. 297 Taralczak, M. 38 Teruya-Feldstein, J. 337 Thaller, G. 67 Thijssen, J. 157 Thuresson, A.-C. 1 Tibiletti, M.G. 229 Tilanus, H.W. 130 Tönnies, H. 166 Trautmann, U. 31 Tyybäkinoja, A. 8 Urquidi, V. 116 Utine, G.E. 31
van de Vosse, E. 19 van Dekken, H. 130 van den Hurk, W.H. 157 van Duin, M. 130 van Groningen, J.J.M. 157 van Marion, R. 130 Vauhkonen, H. 277 Veltman, J.A. 19 Villa, O. 84 Vilpo, J. 8 Vissers, K.J. 130 Vissers, L.E.L.M. 19 Volleth, M. 72 Vreede, L. 157 Waisertreiger, I.S.-R. 42 Weise, A. 31 White, S.J. 19 Wolff, D.J. 177 Yang, F. 72 Ye, C.J. 237 Zanesi, N. 196 Zelenetz, A.D. 337 Zheng, X. 92 Zhu, Y. 260
Contents Vol. 118, 2007
72 Karyotypic differences in two sibling species of Scotophilus
No. 1
from South Africa (Vespertilionidae, Chiroptera, Mammalia)
Original Articles 1 Whole-genome array-CGH for detection of submicroscopic
chromosomal imbalances in children with mental retardation Thuresson, A.-C.; Bondeson, M.-L.; Edeby, C. (Uppsala); Ellis, P.; Langford, C. (Cambridge); Dumanski, J.P. (Uppsala/Birmingham, Ala.); Annerén, G. (Uppsala) 8 High-resolution oligonucleotide array-CGH pinpoints genes
involved in cryptic losses in chronic lymphocytic leukemia Tyybäkinoja, A. (Helsinki); Vilpo, J. (Tampere); Knuutila, S. (Helsinki) 13 Tiling resolution array CGH of dic(7;9)(p11⬃13;p11⬃13)
in B-cell precursor acute lymphoblastic leukemia reveals clustered breakpoints at 7p11.2⬃12.1 and 9p13.1 Lundin, C.; Heidenblad, M.; Strömbeck, B.; Borg, Å. (Lund); Hovland, R. (Helse-Bergen); Heim, S. (Oslo); Johansson, B. (Lund)
19 Variation of CNV distribution in five different ethnic
Eick, G.N. (Stellenbosch); Jacobs, D.S. (Rondebosch); Yang, F. (Cambridge); Volleth, M. (Magdeburg) 78 Comparative chromosome mapping of 5S rDNA and 5SHindIII
repetitive sequences in Erythrinidae fishes (Characiformes) with emphasis on the Hoplias malabaricus ‘species complex’ Ferreira, I.A. (Botucatu); Bertollo, L.A.C. (São Carlos); Martins, C. (Botucatu)
Human Cytogenetics Case Report 84 Duplication dup(1)(q41q44) defined by fluorescence in
situ hybridization: delineation of the ‘trisomy 1q42]qter syndrome’ Coccé, M.C. (Buenos Aires); Villa, O. (Barcelona); Obregon, M.G. (Buenos Aires); Salido, M. (Barcelona); Barreiro, C. (Buenos Aires); Solé, F. (Barcelona); Gallego, M.S. (Buenos Aires)
populations White, S.J. (Leiden); Vissers, L.E.L.M.; Geurts van Kessel, A. (Nijmegen); de Menezes, R.X. (Leiden/Rotterdam); Kalay, E. (Nijmegen/Trabzon); Lehesjoki, A.E. (Helsinki); Giordano, P.C.; van de Vosse, E.; Breuning, M.H. (Leiden); Brunner, H.G. (Nijmegen); den Dunnen, J.T. (Leiden); Veltman, J.A. (Nijmegen)
No. 2-4 91 Preface Cannizzaro, L.A.; Ramesh, K.H. (Bronx, NY)
31 Neocentric small supernumerary marker chromosomes
(sSMC) – three more cases and review of the literature Liehr, T. (Jena); Utine, G.E. (Ankara); Trautmann, U.; Rauch, A. (Erlangen); Kuechler, A. (Jena); Pietracz, J.; Bocian, E. (Warsaw); Kosyakova, N. (Jena/Moscow); Mrasek, K. (Jena); Boduroglu, K. (Ankara); Weise, A. (Jena); Aktas, D. (Ankara) 38 Initial maternal meiotic I error leading to the formation of a
maternal i(2q) and a paternal i(2p) in a healthy male Baumer, A. (Zurich); Basaran, S. (Istanbul); Taralczak, M. (Zurich); Cefle, K.; Ozturk, S.; Palanduz, S. (Istanbul); Schinzel, A. (Zurich) 42 Human chromosome 1 satellite 3 DNA is decondensed,
demethylated and transcribed in senescent cells and in A431 epithelial carcinoma cells Enukashvily, N.I. (St. Petersburg); Donev, R. (Cardiff); Waisertreiger, I.S.-R.; Podgornaya, O.I. (St. Petersburg) 55 Sequence analysis of the equine SLC26A2 gene locus on
chromosome 14q15]q21 Hansen, M.; Knorr, C. (Göttingen); Hall, A.J. (Armidale); Broad, T.E. (Brisbane); Brenig, B. (Göttingen) 63 Cytogenetic mapping of eight genes encoding fatty acid
binding proteins (FABPs) in the pig genome Szczerbal, I.; Chmurzynska, A.; Switonski, M. (Poznan) 67 Exclusion of NFYB as candidate gene for congenital splay leg
in piglets and radiation hybrid mapping of further five homologous porcine genes from human chromosome 12 (HSA12) Boettcher, D. (Halle); Paul, S.; Bennewitz, J. (Kiel); Swalve, H.H. (Halle); Thaller, G. (Kiel); Maak, S. (Halle)
Fax +41 61 306 12 34 E-Mail
[email protected] www.karger.com
© 2007 S. Karger AG, Basel
92 Head and neck cancer: reduce and integrate for optimal
outcome Belbin, T.J.; Bergman, A.; Brandwein-Gensler, M.; Chen, Q.; Childs, G.; Garg, M.; Haigentz, M.; Hogue-Angeletti, R.; Moadel, R.; Negassa, A.; Owen, R.; Prystowsky, M.B.; Schiff, B.; Schlecht, N.F.; Shifteh, K.; Smith, R.V.; Zheng, X. (Bronx, NY) 110 Origin and functional significance of large-scale chromosomal
imbalances in neuroblastoma Stallings, R.L. (San Antonio, TX) 116 Genomic signatures of breast cancer metastasis Urquidi, V.; Goodison, S. (Jacksonville, FL) 130 High-resolution array comparative genomic hybridization
of chromosome 8q: evaluation of putative progression markers for gastroesophageal junction adenocarcinomas van Duin, M.; van Marion, R.; Vissers, K.J.; Hop, W.C.J.; Dinjens, W.N.M.; Tilanus, H.W.; Siersema, P.D.; van Dekken, H. (Rotterdam) 138 Pathology and genetics of adipocytic tumors Hameed, M. (Newark, NJ) 148 Molecular cytogenetic characterization of pancreas cancer
cell lines reveals high complexity chromosomal alterations Griffin, C.A.; Morsberger, L.; Hawkins, A.L.; Haddadin, M.; Patel, A. (Baltimore, MD); Ried, T.; Schrock, E. (Bethesda, MD); Perlman, E.J.; Jaffee, E. (Baltimore, MD) 157 Molecular mechanisms underlying the MiT translocation
subgroup of renal cell carcinomas Medendorp, K.; van Groningen, J.J.M.; Schepens, M.; Vreede, L.; Thijssen, J.; Schoenmakers, E.F.P.M.; van den Hurk, W.H.; Geurts van Kessel, A.; Kuiper, R.P. (Nijmegen)
Access to full text and tables of contents, including tentative ones for forthcoming issues: www.karger.com/cgr_issues
166 Disruption of the FA/BRCA pathway in bladder cancer Neveling, K.; Kalb, R. (Würzburg); Florl, A.R. (Düsseldorf); Herterich, S.; Friedl, R.; Hoehn, H. (Würzburg); Hader, C.; Hartmann, F.H. (Düsseldorf); Nanda, I.; Steinlein, C.; Schmid, M. (Würzburg); Tönnies, H. (Berlin); Hurst, C.D.; Knowles, M.A. (Leeds); Hanenberg, H. (Düsseldorf); Schulz, W.A.; Schindler, D. (Würzburg) 177 The genetics of bladder cancer: a cytogeneticist’s perspective Wolff, D.J. (Charleston, SC) 182 The cytogenetics and molecular biology of endometrial
stromal sarcoma Sandberg, A.A. (Phoenix, AZ) 190 Pathogenetic mechanisms in endometrial stromal sarcoma Micci, F.; Heim, S. (Oslo) 196 Influence of a nonfragile FHIT transgene on murine tumor
susceptibility
a pathological paradox? Calcagnile, O.; Gisselsson, D. (Lund) 277 Etiology of specific molecular alterations in human
malignancies Vauhkonen, H.; Heino, S.; Myllykangas, S.; Lindholm, P.M.; Savola, S.; Knuutila, S. (Helsinki) 284 FISH panels for hematologic malignancies Sreekantaiah, C. (Stratford, CT) 297 Genetic and epigenetic alterations in myelodysplastic
syndrome Mihara, K.; Takihara, Y.; Kimura, A. (Hiroshima) 304 DNA profiling by arrayCGH in acute myeloid leukemia and
myelodysplastic syndromes Suela, J.; Alvarez, S.; Cigudosa, J.C. (Madrid)
McCorkell, K.A. (Philadelphia, PA); Mancini, R. (Rome); Siprashvili, Z. (Stanford, CA); Barnoski, B.L. (Philadelphia, PA); Iliopoulos, D. (Columbus, OH); Siracusa, L.D. (Philadelphia, PA); Zanesi, N.; Croce, C.M.; Fong, L.Y.Y.; Druck, T.; Huebner, K. (Columbus, OH) 204 Exogenous mycoplasmal p37 protein alters gene expression,
growth and morphology of prostate cancer cells Goodison, S. (Jacksonville, FL); Nakamura, K.; Iczkowski, K.A.; Anai, S.; Boehlein, S.K.; Rosser, C.J. (Gainesville, FL) 214 SNP-Array genotyping and spectral karyotyping reveal
uniparental disomy as early mutational event in MSS- and MSI-colorectal cancer cell lines Melcher, R. (Würzburg); Al-Taie, O. (Aschaffenburg); Kudlich, T.; Hartmann, E.; Maisch, S.; Steinlein, C.; Schmid, M.; Rosenwald, A.; Menzel, T.; Scheppach, W.; Lührs, H. (Würzburg) 222 Familial cancer syndromes: catalog with comments Hecht, F. (Scottsdale, AZ) 229 Interphase FISH as a new tool in tumor pathology Tibiletti, M.G. (Varese) 237 The dynamics of cancer chromosomes and genomes Ye, C.J.; Liu, G.; Bremer, S.W.; Heng, H.H.Q. (Detroit, MI) 247 The mystery of chromosomal translocations in cancer Koss, L.G. (Bronx, NY) 252 MicroRNAs in carcinogenesis Hagan, J.P.; Croce, C.M. (Columbus, OH) 260 Non-random inactivation of large common fragile site genes
in different cancers McAvoy, S.; Ganapathiraju, S.C.; Ducharme-Smith, A.L.; Pritchett, J.R.; Kosari, F.; Perez, D.S.; Zhu, Y. (Rochester, MN); James, C.D. (San Francisco, CA); Smith, D.I. (Rochester, MN)
IV
270 Telomere dysfunction and telomerase activation in cancer –
Cytogenet Genome Res Vol. 118, 2007
310 Genetic alteration associated with chronic lymphocytic
leukemia Cotter, F.E.; Auer, R.L. (London) 320 Genetics and epigenetics of 1q rearrangements in
hematological malignancies Fournier, A.; Florin, A.; Lefebvre, C.; Solly, F.; Leroux, D.; Callanan, M.B. (Grenoble) 328 Cell migration patterns and ongoing somatic mutations in
the progression of follicular lymphoma Adam, P.; Schoof, J.; Hartmann, M. (Würzburg); Schwarz, S. (Regensburg); Puppe, B.; Ott, M.; Rosenwald, A.; Ott, G.; Müller-Hermelink, H.K. (Würzburg) 337 Molecular cytogenetic analysis of follicular lymphoma (FL)
provides detailed characterization of chromosomal instability associated with the t(14;18)(q32;q21) positive and negative subsets and histologic progression Nanjangud, G. (New York, NY); Rao, P.H. (Houston, TX); Teruya-Feldstein, J.; Donnelly, G.; Qin, J.; Mehra, S.; Jhanwar, S.C.; Zelenetz, A.D.; Chaganti, R.S.K. (New York, NY) 345 Molecular cytogenetics of IGH rearrangements in
non-Hodgkin B-cell lymphoma Bernicot, I.; Douet-Guilbert, N.; Le Bris, M.-J.; Herry, A.; Morel, F.; De Braekeleer, M. (Brest) 353 Molecular cytogenetics in the study of cutaneous T-cell
lymphomas (CTCL) Karenko, L.; Hahtola, S.; Ranki, A. (Helsinki) 362 Author Index Vol. 118, No. 2–4, 2007 363 Author Index Vol. 118, 2007
Contents
