Characteristic Pattern of Chromosomal Gains and Losses in Merkel Cell Carcinoma Detected by Comparative Genomic Hybridization1
Total Page:16
File Type:pdf, Size:1020Kb
[CANCER RESEARCH 58. 1503-1508. April 1. 1998) Characteristic Pattern of Chromosomal Gains and Losses in Merkel Cell Carcinoma Detected by Comparative Genomic Hybridization1 Mireille Van Gele, Frank Speleman, Jo Vandesompele, Nadine Van Roy, and J. Helen Leonard2 Department of Medical Genetics, University Hospital Ghent. B-9000 Client. Belgium ¡M.V. G.. F. S.. J. V.. N. V. RJ, and Queensland Radium Institute Laboratory. Queens/and Institute of Medical Research, Brisbane, 4029, Queensland. Australia ¡J.H. L] ABSTRACT Ip36 were recently shown to be implicated in MCC.4 To date, only three large LOH studies on MCC have been reported, investigating the Merkel cell carcinoma or small cell carcinoma of the skin is a rare skin status of Ip, 3p. and 13q loci (21, 22).5 cancer seen in increasing numbers in Queensland, Australia. In its clinical In view of the limited available genetic data for MCC, we decided course and histopathology, it resembles small cell lung carcinoma (SCLC). to perform CGH on a series of tumors and tumor-derived cell lines Little is known of the genetic basis of this disease except for a number of cytogenetic studies and three loss of heterozygosity studies. Therefore, from 24 patients. CGH allows screening of the entire genome for comparative genomic hybridization was performed to determine the char chromosomal gains, losses, and gene amplification. Therefore, it acteristic DNA gains and losses that occur in this tumor. Comparative provides an opportunity to identify regions of the genome that un genomic hybridization analysis of 34 specimens from 24 patients revealed dergo these types of genetic alterations in MCC and to increase our a pattern of gains and losses that closely resembles that seen in SCLC. knowledge of this aggressive neoplasm. Overall frequent loss was seen for chromosomes 3p (46%), 5q (21%), 8p (21%), 10 (33%), llq (17%), 13q (33%), and 17p (25%). Significant gains were seen for chromosomes 1 (63%), 3q (33%), 5p (38%), 8q (38%), 19 MATERIALS AND METHODS (63%), and X (41%), with smaller numbers having gains for chromosomes 6, 7, 20, and 21. In contrast to SCLC, amplification in Merkel cell MCC Tumors and Cell Lines. In total, 34 DNA samples from 24 patients carcinoma is a rare event. were analyzed. These include 26 tumors and eight MCC cell lines (Table 1). Clinical and LOH data for these tumors are summari/.ed in Table 1. Seven cell INTRODUCTION lines were derived from tumors included in this study. Cell line UISO was described previously (18). Cytogenetic data of tumors and cell lines have been MCC3 is a neuroendocrine skin tumor with particularly aggressive published (12, 23, 24) or will be reported elsewhere6 (tumors UHG-VM and UHG-RM and cell lines MCC 13, MCC 14/2, MCC 15. and MCC26). Within clinical behavior, which appears to be increasing in incidence in "Results" where patient loss or gain is referred to, only one tumor sample was Queensland, Australia. In 1992, it accounted for 4% of skin malig counted for patients in which there were two samples examined and no cell line nancies causing death in Queenslanders (1). Clinical behavior is similar to that of SCLC, with 25-30% of patients dying of their data were included. CGH Analysis. Metaphase spreads were prepared from phytohemaggluti- disease within 3 years (2, 3), attributable mainly to the propensity of nin-stimulated lymphocytes from healthy individuals according to standard MCC to metastasise early (50-80% of cases having regional lymph procedures. Each batch of chromosome preparations was tested by hybridiza node involvement at presentation; Ref. 3). Histochemical markers are tion (reverse painting) with DNA from the neuroblastoma cell line IMR32 with almost identical, making differential diagnosis of these diseases dif known DNA gains and losses (25) to assess the quality of the slides for CGH. ficult where there is no obvious primary lesion. The majority of Batches with poor hybridization (e.g., strong staining of contours of sister primary MCCs occur on the head and neck region or arms (3), chromatids and uneven or granular staining) were discarded. Slides were stored implicating UV exposure in its etiology. in plastic boxes with silica gel at —¿20°Cbefore use. DNA was extracted from Cytogenetic investigations of MCC have not been common, with MCC cell lines and peripheral blood from a healthy male individual as only 25 tumors and five cell lines reported in the literature (4-20).4 So described (26). Primary tumor DNA was extracted from 50-nm cryosections from biopsies frozen at -80°C. Labeling of DNAs. in situ hybridization, far, no recurrent chromosomal rearrangements have been identified fluorescence microscopy, digital image acquisition, and processing were es that could be helpful in the identification of genes involved in MCC sentially done according to the methods of du Manoir et al. (27, 28) with minor development: however, the short arm of chromosome 1 was fre modifications (25). 4',6-Diamino-2-phenylindole images of metaphases were quently affected by inversions, translocations, and deletions. The recorded before hybridization using a Leitz DM microscope, a high-sensitivity observation of Ip36 abnormalities in two MCCs as the only karyo- integrated monochrome charge-coupled device camera (Sony IMAC-CCD typic change suggested that genes located in this region play a role in S30), and dedicated software (ISIS: MetaSystems GmbH, Altlussheim. Ger MCC tumorigenesis (14).4 Moreover, two distinct regions within many). Further processing of these images for CGH analysis was done using the ISIS CGH software (MetaSystems). For each case, 10-20 metaphases were Received 9/23/97; accepted 1/23/98. analyzed. For evaluation of CGH data, average ratio profiles with fixed limits The costs of publication of this article were defrayed in part by the payment of page at 1.25 and 0.75 and SD limits (the width of the confidence interval being three charges. This article must therefore be hereby marked advertisement in accordance with times the SD) as well as individual ratio profiles were analyzed. A chromo 18 U.S.C. Section 1734 solely to indicate this fact. ' This work was supported by grants from the Vereniging voor Kankerbestrijding somal region was considered to be over-represented (gain) or under-repre ( 1995-1998). FAVOGrants G.0328.95 and G.0085.96. GOA Grant 12051397. a grant from sented (loss) when the ratio profile crossed the SD limit. In tumors from female the Flemish Institute for the Promotion of Scientific Technological Research in Industry patients, a normal average ratio profile lies at or closely near the 1.25 limit, (IWT), and the Queensland Cancer Fund and the Queensland Radium Institute. N. V. R. whereas a clear shift to the right of the average ratio profile versus the 1.25 is a postdoctoral researcher of the Fund for Scientific Research. Flanders. 2 To whom requests for reprints should be addressed, at Queensland Radium Institute limit was considered a gain. The reliability of the CGH procedure and software Laboratory. Queensland Institute of Medical Research, Post Office, Royal Brisbane analysis was tested previously on well-characterized neuroblastoma cell lines. Hospital, Herston Road. Brisbane. 4029. Queensland. Australia. Phone: 61-7-33620309; Particular attention was given in that study to the accuracy for detection of Fax: 61-7-33620107. 3 The abbreviations used are: MCC, Merkel cell carcinoma; SCLC, small cell lung *"H. J. Leonard, D. Nancarrow, N. Hayward, M. Van Gele, N. Van Roy. and F". carcinoma: LOH, loss of heterozygosity; CGH, comparative genomic hybridization. 4 M. Van Gele. N. Van Roy. S. G. Roñan,L. Messiaen. J. Vandesompele, M. L. Geerts, Speleman. Deletion mapping on the short arm of chromosome l in Merkel cell carcinoma, J. M. Naeyaert, E. Blennow. I. Bar-Am, T. K. Das Gupta, P. van der Drift. J. H. Leonard, submitted for publication. and F. Speleman. Molecular analysis of Ip36 breakpoints in two Merkel cell carcinomas. 6M. Van Gelé.F. Speleman. N. Van Roy, S. Van Belle, V. Cocquyt, and J. H. Genes Chromosomes Cancer, in press. Leonard, unpublished data. 1503 Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 1998 American Association for Cancer Research. CGH ANALYSIS OF MERKEL CELL CARCINOMA Table 1 Clinical und LOH dala of tumors from 24 patienls wiÃhMCC on LOH on Nodes at first Survival'' Tumor"MCCIMCC2MCC3MCC4MCC5MCC7TIMCC7T2MCC8MCC9MCCSpecimen*IpNode on 3p(months)+ 13 presentation +NodeNode 14+ + 13+ + +NodeRee +*DA+ +23 40+ +Ree 26+ + +Met 42+ + +Ree +42+ +ReeRee 16+ + 9I/C+ + 11MCC +ReeReeNodeNode I/C++ 12 12MCC 9+ 13MCC14T1MCC14T2MCC *DA+ -1- 24 5ND + NDRee 51117+ ND 15MCC +Ree 16MCC +Ree 17MCC +Ree 12+ + 18MCC20MCC21MCC22MCC24MCC26TIMCC26T2UHG-VMUHG-RMUISOSexFMMM1FlMMMMFMMMMMM1MMMP1PFFLOH'+Node *DP+ + 42 NDNodeNode ND + 49*DA33 'DP+ +Ree «DP+ +24 +Ree 17+ +Node 9+ +Met 9ND +MetPrimLOH 4ND ND + 22ND ND + ND ? ? "Cell line MCC26TC is derived from MCC26T!; cell lines MCC14/ITC and MCC14/2TC are derived from MCCI4T1. ' Ree, recurrence al site; Node, specimen laken from regional lymph node; Mel. melastasis; Prim, primary tumor. ' +, LOH detected within region of sludy; —¿.noLOH delected; ND. not done. Survival is given in months, where * atier number the patient is still alive with disease (DP) or without disease (DA). In two cases (he patient died from other causes (I/C). small distal Ip deletions and MYCN amplification (25). As a control, normal- DNA gains for the remaining part of the chromosome. This is illus to-normal hybridi/ation was performed. trated by the chromosome 13 profiles for MCC5T and MCC5TC (Fig. 1). MCC5T shows under-representation of the proximal part of 13q, RESULTS whereas in MCC5TC, over-representation of the distal part of 13q is observed. The MCC5TC profile could be explained by the presence of The results of CGH analysis of 34 samples of tumor and tumor cell extra copies of the derivative chromosome 13 or translocated distal line DNA from a total of 24 patients are summarized in Fig.