Physical Mapping of a Commonly Deleted Region, the Site of a Candidate Tumor Suppressor Gene, at 12q22 in Human Male Germ Cell Tumors

GENOMICS

35, 562–570 (1996) 0398

ARTICLE NO.

Physical Mapping of a Commonly Deleted Region, the Site of a Candidate Tumor Suppressor Gene, at 12q22 in Human Male Germ Cell Tumors V. V. V. S. MURTY,* BEATRICE RENAULT,† CATHERINE T. FALK,‡ GEORGE J. BOSL,* RAJU KUCHERLAPATI,† AND R. S. K. CHAGANTI*,1 *Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, New York 10021; †Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461; and ‡New York Blood Center, 310 East 67th Street, New York, New York 10021 Received March 5, 1996; accepted May 29, 1996

A candidate tumor suppressor gene (TSG) site at 12q22 characterized by a high frequency of loss of heterozygosity (LOH) and a homozygous deletion has previously been reported in human male germ cell tumors (GCTs). In a detailed deletion mapping analysis of 67 normal-tumor DNAs utilizing 20 polymorphic markers mapped to 12q22–q24, we identified the limits of the minimal region of deletion at 12q22 between D12S377 (proximal) and D12S296 (distal). We have constructed a YAC contig map of a 3-cM region of this band between the proximal marker D12S101 and the distal marker D12S346, which contained the minimal region of deletion in GCTs. The map is composed of 53 overlapping YACs and 3 cosmids onto which 25 polymorphic and nonpolymorphic sequence-tagged sites (STSs) were placed in a unique order. The size of the minimal region of deletion was approximately 2 Mb from overlapping, nonchimeric YACs that spanned the region. We also developed a radiation hybrid (RH) map of the region between D12S101 and D12S346 containing 17 loci. The consensus order developed by RH mapping is in good agreement with the YAC STS-content map order. The RH map estimated the distance between D12S101 and D12S346 to be 246 cR8000 and the minimal region of deletion to be 141 cR8000 . In addition, four genes that were previously mapped to 12q22 have been excluded as candidate genes. The leads gained from the deletion mapping and physical maps should expedite the isolation and characterization of the TSG at 12q22. q 1996 Academic Press, Inc.

chromosomal arm (Mitelman, 1994) or loss of heterozygosity (LOH) of markers located on it have been reported in various tumor types (Fey et al., 1989; Sano et al., 1991; Seymour et al., 1994; Reifenberger et al., 1995; Schneider et al., 1995; Hahn et al., 1995). Male germ cell tumors (GCTs) have been shown to carry chromosomal deletions affecting 12q or monosomy of chromosome 12 (Murty et al., 1990; Samaniego et al., 1990; Rodriguez et al., 1992). Restriction fragment length polymorphism (RFLP) studies of male GCTs have identified LOH at two sites, 12q13 and 12q22, suggesting the presence of at least two candidate TSGs on 12q (Murty et al., 1992). The 12q22 deletions were characterized by LOH affecting the markers D12S7 and D12S12. In addition, a homozygous deletion at 12q22 in one case of GCT has been identified that gave further support to the hypothesis of a candidate TSG in this chromosomal region. As a prelude to isolating this candidate TSG, we further characterized the region of deletion, constructed a physical map, and excluded several candidate genes (MGF, BTG1, TMPO, and NEDD1) mapped to the region. A recently published YAC contig map of chromosome 12 (Krauter et al., 1995) contained several ordered polymorphic markers in the 12q22 region. In the present study, we have refined this map and added additional markers. We have utilized 20 polymorphic markers from this new map to assess for LOH on 67 pairs of normal–tumor DNAs. Based on the pattern of LOH in these tumors, we now define a commonly deleted region of 12q22 whose size is approximately 2 Mb.

INTRODUCTION

MATERIALS AND METHODS

Chromosome 12q appears to harbor one or more tumor suppressor genes (TSGs) since deletion of this 1 To whom correspondence should be addressed at Department of Human Genetics, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021. Telephone: (212) 639-8121. Fax: (212) 794-5830.

0888-7543/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

AID

Genom 4225

/

6r1b$$$261

Tumor specimens, normal cells, and cell lines. Tumor tissues and the corresponding peripheral blood mononuclear cells were obtained, after informed consent, from patients with GCT evaluated at the Memorial Sloan-Kettering Cancer Center as described previously (Rodriguez et al., 1992; Murty et al., 1994a,b). A total of 63 tumor specimens derived from 57 patients were included in the present study. Of these 63 specimens, 8 were also examined as cell lines

562

07-01-96 14:31:48

gnma

AP: Genomics

563

PHYSICAL MAP OF 12q22 DELETED REGION IN GERM CELL TUMORS

TABLE 1 STSs on 12q22 Utilized in PCR Analysis No.

Locus

Probe

Source

Polymorphic 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.

D12S81 D12S379 D12S101 D12S362 D12S309 D12S2087 D12S1671 D12S2085 D12S1716 D12S377 D12S2084 D12S1051 D12S1657 D12S1300 D12S393 D12S296 D12S346 D12S58 D12S234 D12S367 D12S392

AFM102xg9 GATA4B09 AFM234tg11 AFM336yf9 AFM199wb10 c10A8/CA AFMb328xc5 c12G9/CA AFMa065yd5 GATA4A06 c2C11/CA GATA24A01 AFMb293ye5 GATA85A04 GATA15A03 UT5022 AFM298xe5 MFD73 c5D8/CA AFMa123xe1 GATA13D05

Gyapay et al. (1994) GDB Gyapay et al. (1994) Gyapay et al. (1994) Weissenbach et al. (1992) LeBlanc-Straceski et al. (1994) GDB LeBlanc-Straceski et al. (1994) GDB GDB LeBlanc-Straceski et al. (1994) GDB GDB GDB GDB GDB Gyapay et al. (1994) GDB LeBlanc-Straceski et al. (1994) Gyapay et al (1994) GDB

Nonpolymorphic 22. 23. 24. 25. 26. 27. 28. 29. 30.

D12S2086 D12S1280 D12S1100 D12S1442 D12S1430 D12S1279 D12S2083 D12S1098 D12S1082

c10A8/T7 952b1-R 850b8-R WI-4038 WI-3193 781b3-L c2C11/T7 681g7-L WI-1945

LeBlanc-Straceski et al. (1994) AECOM AECOM GDB GDB AECOM LeBlanc-Straceski et al. (1994) GDB GDB

Gene-based 31. 32.

HAL MGF

RK835/RK836 —

AECOM Present study

Note. NA, not available; GDB, Human Genome Database; AECOM, Albert Einstein College of Medicine. developed in our laboratory. Histologically, the tumors comprised 6 seminomas, 56 nonseminomas (22 teratomas, 16 embryonal carcinomas, 4 yolk sac tumors, 1 choriocarcinoma, and 13 mixed tumors), and 1 combined tumor. They represented all sites of presentation and grades at primary and metastatic sites. In addition to tumor specimens ascertained by us, 4 GCT cell lines (833K-E, 577M-F, 577M-Lu, 2061H) (2 teratomas, and 2 embryonal carcinomas) with the corresponding lymphoblastoid cell lines (provided by Dr. D. L. Bronson, University of Minnesota, Minneapolis) were also analyzed. Primers and probes. Details of the polymorphic and nonpolymorphic sequence-tagged sites (STSs) utilized in the LOH analysis and their sources are shown in Table 1. A primer pair RK835 (5*-AGACCCACCAGGTATTTTCAGA-3*) and RK836 (5*-AATTAGCCAAAGAAAGTCAACTGG-3*) was designed to amplify the HAL gene. Based on the published sequence of the BTG1 gene (Rouault et al., 1992), primers were generated for single-strand conformational polymorphism analysis (SSCP) to screen for mutations, and an antisense sequence primer was constructed for evaluation of mRNA expression by Northern blot analysis. For the SSCP analysis, three sets of primers covering the entire coding sequence of BTG1 were synthesized to generate PCR products of less than 240 bp: Exon 1: 1F 5*-TGAGGAAGCCCGGGGTGGCT-3*; 1R 5*-CCCTGCTCACCTGCCAGCA-3*; Exon 2: 2.1F 5*TCTCTGTTCTTCTTTTCTTCTAT-3*; 2.1R 5*-TTCTGTAGGACACTTCATAGG-3*; 2.2F 5*-GAACTCACACTCTGGGTTGA-3*; 2.2R 5*GATCCATCCACAGACTATATC-3*. A 42-bp antisense oligonucleotide

AID

Genom 4225

/

6r1b$$$262

07-01-96 14:31:48

(positions 475–516) in the second exon was used as a hybridization probe in Northern blot analysis. In addition, the following primers were synthesized from the cDNA sequence of the MGF gene (Martin et al., 1990) to analyze a radiation hybrid panel by PCR: forward 5*CCAAGTCATTGTTGGATAAG-3* (positions 375–395) and reverse 5*CTTAGATGAGTTTTCTTTCAC-3* (positions 525–546). The polymorphic probes D12S52 (a gift from J. P. Magaud, France; Rouault et al., 1991) and D12S7 (ATCC) were used for LOH analysis by Southern blotting as described (Murty et al., 1992). Analysis of LOH by PCR. The following 18 microsatellite polymorphic markers were utilized in the analysis of LOH: D12S81, D12S379, D12S101, D12S309, D12S2087, D12S2085, D12S1716, D12S377, D12S1051, AFMb293ye5, D12S1300, D12S393, D12S296, D12S346, D12S58, D12S234, D12S367, and D12S392. PCR was carried out in a 25-ml reaction volume containing 40–100 ng of genomic DNA under standard conditions, with 20 pmol of primers, in which one-fifth of the forward primer was end-labeled with [g-32P]ATP and amplified for 25–30 cycles. The amplified PCR products were denatured in sequencing stop solution and electrophoresed on 6% denaturing polyacrylamide gels containing 10% formamide. The dried gels were autoradiographed for 5–48 h, and allelic deletions were scored by visual examination. Reduction of signal intensity by more than 50% of one allele over the other allele in tumor DNA compared with the intensity of constitutional alleles was considered LOH (Murty et al., 1994c).

gnma

AP: Genomics

564

MURTY ET AL.

Analysis of radiation hybrids. A panel of 83 radiation hybrid clones (Cox et al., 1990) was obtained from Research Genetics (Huntsville, AL). The radiation hybrid DNAs, along with positive (monochromosome 12 hybrid, GM10868) and negative (murine parental cell line, RH.G3/A3) control DNAs, were analyzed for the presence or absence of 19 loci (13 polymorphic and 6 nonpolymorphic markers) by PCR with the relevant primers. PCRs were performed at least twice on each hybrid for each set of primers with 25 ng of radiation hybrid DNA along with controls in a final reaction volume of 15 ml using standard conditions in a Perkin–Elmer Cetus 9600 thermal cycler. PCR products were visualized on 2% agarose gels following ethidium bromide staining. If a particular hybrid could not be scored definitively as positive or negative for a given marker, it was coded as unknown. Preliminary ordering of the loci was carried out using the computer programs MDMAP (Falk, 1991), MINBR (Falk and Falk, 1995), and RHMINBR from the RHMAP package (Boehnke et al., 1991). These ordering methods provided ranked lists of candidate locus orders based on either maximizing correlations between adjacent loci (MDMAP) or minimizing the number of obligate breaks between adjacent loci (MINBR and RHMINBR). Based on the preliminary results, the data were further analyzed using the more rigorous statistical program, RHMAXLIK from the RHMAP package (Boehnke et al., 1991). Distances were estimated in rays, with one cR corresponding to 1% breakage. Southern and Northern blot analysis. High-molecular-weight DNA and RNA was isolated, and Southern and Northern blot analyses were performed by standard methods. SSCP analysis. PCR was performed as described above except that 1 mCi of [32P]dCTP was added to the reaction. The PCR products were diluted in 0.1% SDS/10 mM EDTA, denatured in sequencing stop solution, and electrophoresed overnight on 6% nondenaturing polyacrylamide gels containing 10% glycerol at RT. Dried gels were autoradiographed and examined for conformational changes. Identification of YACs and generation of STS content map. To generate a complete physical map of chromosome 12, we previously identified chromosome 12 specific YACs through examination of the CEPH quickmap database. The DNA from subsets of these YACs was pooled based on their location along the length of chromosome 12. YACs corresponding to each marker were identified by a hierarchial screening of the pools by a strategy referred to as quickmapbased pooling strategy (QPS; Krauter et al., 1995). This strategy has permitted the identification of YACs that corresponded to 12q22 markers. The location of individual STSs on YACs was confirmed by PCRs (Krauter et al., 1995). STSs used in the PCR analysis and their sources are shown in Table 1. In addition, cosmids previously mapped to the 12q22 region (Montgomery et al., 1993) were also utilized in constructing the contig map. Pulsed-field gel electrophoresis (PFGE). The YAC clones were grown and agarose plugs were prepared in low-melting agarose by standard methods. Proteinase K-digested blocks were run on pulsedfield gels by standard methods. Southern-transferred YAC DNA was hybridized with radioactively labeled total genomic DNA. Size estimations were made by standard methods. Fluorescence in situ hybridization (FISH). FISH was performed by standard methods as described previously (Mathew et al., 1992).

RESULTS

We have previously shown a high frequency of LOH of the polymorphic RFLP markers D12S7 and D12S12 mapped to 12q22 in a panel of male GCTs (Murty et al., 1992). In addition, one tumor in this panel showed a homozygous deletion of regions containing the markers D12S7 and MGF; the latter also mapped to 12q22 (Murty et al., 1992), which suggested the site of a candidate TSG in their vicinity. In the present study, we have further characterized the 12q22 deletions and constructed a physical map of the region.

AID

Genom 4225

/

6r1b$$$262

07-01-96 14:31:48

TABLE 2 Frequency of LOH at 12q22–q23 in Male GCT Locus

No. studied

Informative

LOH

%

D12S81 D12S379 D12S7a D12S52a D12S101 D12S309 D12S2087 D12S2085 D12S1716 D12S377 D12S1051 AFMb293ye5 D12S1300 D12S393 D12S296 D12S346 D12S58 D12S234 D12S367 D12S392

67 67 54 50 66 67 66 67 67 66 66 67 65 67 66 65 66 59 66 66

36 50 26 26 30 49 31 41 32 50 55 37 37 38 46 47 37 36 50 45

8 15 8 6 8 13 8 18 17 21 23 16 14 13 19 16 12 8 12 11

22 30 31 23 27 27 26 44 53 42 42 43 38 34 41 34 32 22 24 24

a

RFLP markers analyzed by Southern blotting.

Characterization of 12q22 deletions. In the present study, a panel of 67 normal–tumor DNA pairs were assayed for LOH by PCR utilizing 18 polymorphic STSs and by Southern blotting using the RFLP probes D12S52 and D12S7 mapped to the 12q22–q24 regions. The relative order of these loci and their map positions are shown in Table 2 and Fig. 1. The frequency of heterozygosity among these loci in this set of normal DNAs varied from 0.46 (D12S101) to 0.83 (D12S1051) (Table 2). Examples of data obtained from the LOH analysis are shown in Fig. 2. For example, typing of 3 of the normal DNAs from T-218A, T-318A, T-318B, and T246A revealed all the individuals to be heterozygous for the markers shown. Examination of the corresponding tumor DNAs revealed deletions of one of the alleles in T-218A, while the tumors T-318A, T-318B, and T-246A showed loss at D12S1051 and D12S393 and retained heterozygosity of both proximal (D12S309) and distal (D12S346) markers. Analysis of this type revealed deletions in at least one locus in 34 of the 67 (51%) tumors examined. The frequency of deletions of individual loci varied between 22 and 53% with the highest frequency (53%) at D12S1716 (Table 2). Among the 34 tumors that showed LOH, 10 exhibited alterations at all informative loci, suggesting monosomy or large deletions affecting 12q22–qter. Of the 10 tumors that showed large deletions, 3 also exhibited microsatellite instability at certain loci (data not shown). The remaining 24 tumors exhibited LOH at 1 or more loci while retaining heterozygosity at the other loci, which suggested regional deletions (Figs. 1 and 2). The patterns of LOH from these 24 tumors, which will be useful in deducing the common region of deletion, are shown in Fig. 1. In this subgroup of 24 tumors with

gnma

AP: Genomics

PHYSICAL MAP OF 12q22 DELETED REGION IN GERM CELL TUMORS

565

FIG. 1. Ideogram showing patterns of LOH at 12q22 in male GCTs. Twenty-three tumors showing partial deletions only are represented. (A) G-banded ideogram of chromosome 12. (B) Corresponding regions of deletions at 12q13.2–q13.3 and 12q22 demonstrated by a previous study. (C) Patterns of deletions against each loci are shown. Tumor numbers are shown on top and the status of deletion against each locus are indicated in square boxes. Filled boxes, LOH; hatched boxes, retention of heterozygosity; empty boxes, homozygous and uninformative; ND, not done; T, tumor; CL, cell line. (D) Black bar indicates the minimal region of deletion. Small filled or empty squares within large square boxes indicate replication error-type microsatellite instability.

regional deletions, 19 showed LOH common to all in a region mapped to 12q22 and an additional 2 tumors (T-195A, T-146A) exhibited simultaneous LOH and microsatellite instability within this region of deletion. The remaining 3 showed either proximal or distal deletions. The pattern of LOH among the 19 tumors that showed a common region of deletion at 12q22 identified a minimal region flanked by the marker D12S1716 proximally (T-167A) and D12S346 distally (T-318A, T318B, and T-246A) (Figs. 1 and 2). Tumor T-154A showed LOH at all informative loci except one interstitial RFLP marker, D12S7, which was studied by Southern blot analysis. Of the 3 tumors that retained heterozygosity of markers spanning the common region of deletion and that exhibited LOH outside the region, tumors T-222A and CL-2061H showed deletions proximally (D12S81 and D12S379), and T-155B exhibited deletion at a distal marker (D12S234), implying that these tumors are exceptions. The pattern of LOH in the present study suggested that the loci D12S377, D12S1051, AFMb293ye5, D12S1300, D12S393, and D12S296 are located within the minimal region of deletion flanked by the marker D12S1716 proximally and the marker D12S346 distally. Six of the 19 tumors that defined the common region

AID

Genom 4225

/

6r1b$$$263

07-01-96 14:31:48

of deletion at 12q22 and also exhibited LOH at the distal markers D12S367 and D12S392 mapped to 12q24.33 while retaining heterozygosity of interstitial markers (Fig. 1). This pattern of LOH suggested that the deletions at 12q24.33 are distinct from those at the 12q22 common region of deletion and may identify an unrelated distal deletion. Construction of a physical map of the 12q22 deleted region. To construct a YAC contig map of 12q22, we identified 53 YACs contained within the interval defined by the proximal marker D12S101 and the distal marker D12S346. These YAC clones were screened for STS content of each of the 25 markers in the region. These included 15 polymorphic markers, 9 nonpolymorphic markers, and 1 gene-based STS. If two YACs share 1 or more markers they were considered to have sequence overlap. Such information permitted the organization of all of the 53 YACs into one overlapping set of contiguous YACs (contig). Three cosmids (c10A8, c12G9, c2C11) from which the markers were derived were also placed in this map. The resulting map is shown in Fig. 3. The map contains 25 markers including the Ge´ne´thon marker D12S101 mapped to 109 cM on the genetic map and D12S346 localized to 112 cM on this map. Therefore, the map covers more than 3

gnma

AP: Genomics

566

MURTY ET AL.

FIG. 2. Illustration of common region of deletion at 12q22 by LOH analysis of polymorphic loci. (A) LOH at all loci suggesting a large deletion in this case. (B and C) Two cases showing LOH at D12S1051 and D12S393 while retaining heterozygosity at the proximal (D12S309) and the distal (D12S346) markers. T-318A and B are two different specimens resected at different times. Tumor numbers are shown on top of each panel and markers on left. N, normal; T, tumor; cen, centromere; tel, telomere; Asterisks indicate LOH.

cM of genetic distance. Twenty-one of the 25 markers were unambiguously ordered on this map. To verify the exact map location of the YAC contig on chromosome 12, we performed FISH on 10 YACs (766a11, 779d2, 781b3, 784b3, 803g12, 814e12, 910e8, 922d7, 947h10, 952b1), which constituted the tiling path of the contig. All the YACs mapped to 12q22, indicating that the contig was within the 12q22 region (data not shown). The FISH analysis also identified that YACs 781b3, 784b3, 803g12, 814e12, 910e8, 922d7, 947h10, and 952b1 were nonchimeric based upon the yield of exclusive signals at 12q22. The YACs 766a11 and 779d2 were found to be chimeric based on signals at 12q22 and elsewhere in the genome. On chromosome 12, 1 cM of genetic interval corresponds to 700 kb of DNA (Krauter et al., 1995). Therefore, the estimated size covered by the physical map is more than 2100 kb. To estimate the size of the contig we analyzed the sizes of 2 YACs, 952b1 and 922d7, which contained all the markers in the contig except D12S346, by PFGE. The distal YAC 952b1 lacked D12S296 but retained both proximal and distal markers, suggesting an internal deletion in this YAC. The sizes of these two YACs were estimated to be 1900 kb and 2.4 Mb, respectively. To assess the size of the minimal region of deletion we attempted to establish sets of minimally overlapping YACs that contained the boundary markers between D12S1716 and D12S346. The overlapping YACs spanning the minimal deleted region were estimated for their sizes by PFGE analysis. The minimal deleted region was spanned by various combinations of overlapping, nonchimeric YACs within the region. With one set of two YACs, 947h10 (1.8 Mb)

AID

Genom 4225

/

6r1b$$$263

07-01-96 14:31:48

and 781b3 (805 kb), with an overlap of one marker and a lack of the distal marker (D12S296), the size was estimated to be 2.6 Mb DNA. With a second set of YACs, 910e8 (1350 kb) and 814e12 (580 kb), which lacked the proximal (D12S377) and distal (D12S296) markers, the size was estimated to be 1.9 Mb, and with a third set of YACs, 781b3 (805 kb) and 803g12 (930 kb), with an overlap of one marker and a lack of distal marker (D12S296), the size was estimated to be 1.7 Mb. Thus, the size of the minimal region of deletion between D12S377 and D12S296 can be estimated at a minimum of 1.7 Mb with a possible upper limit of 2.6 Mb of DNA. The upper limit cannot be more precisely estimated due to the lack of a marker distal to D12S393 in any YAC that covered the deletion. Generation of a radiation hybrid (RH) map. To assess independently the order of the markers we typed a panel of 83 radiation hybrids by PCR utilizing 19 markers in the physical map. These included 14 polymorphic (D12S101, D12S309, D12S2087, D12S2085, D12S1716, D12S377, D12S2084, D12S1051, AFM-b293ye5, D12S1300, AFMb328xc5, D12S393, D12S296, and D12S346) and 5 nonpolymorphic (MGF, D12S1279, D12S1100, D12S1442, and D12S1430) loci. One or more of these markers were retained in 28 hybrids. The retention frequencies for individual loci ranged from 7.2 (D12S346) to 24.4% (D12S393) with an overall average of 14.4%. This is close to the expected average retention frequency. The RH analysis included: (a) Two-point analysis using RH2PT from the RHMAP package to identify identical retention patterns and reliable linkage groups. This reduced the number of informative loci to 14. (b) Preliminary ordering using the programs MDMAP, MINBR, and RHMINBR. All three programs identified the same highly ranked orders (see Fig. 4). (c) Multipoint analysis using RHMAXLIK, which calculates likelihood ratios among highly ranked orders and estimates distances between adjacent loci. Although it was not possible to order the 14 loci with a high level of statistical confidence, the orders with the highest likelihood ratios agreed with those identified in preliminary ordering, as well as with the physical map order. The resulting radiation hybrid map is shown in Fig. 4. Distances are given in centirays and converted into kilobases based on an estimate of 30 kb/cR (Cox, 1995). The total map length between the markers D12S101 and D12S346 is 246.2 cR or approximately 7.39 Mb. The minimal deleted region between D12S377 and D12S296 is approximately 4.2 Mb. The distance from D12S377 to D12S293 is approximately 3.2 Mb, which compared reasonably well with the corresponding physical distances estimated to be 2.6 Mb. The difference may be due to the amount of information available for these markers in this RH panel, reflected in only 28 informative hybrids out of 83 (34%) and the lack of statistical power to separate the top ranked orders. Increased information in this region would likely increase the statistical power and reduce some distance estimates.

gnma

AP: Genomics

PHYSICAL MAP OF 12q22 DELETED REGION IN GERM CELL TUMORS

567

FIG. 3. Contig map of overlapping YACs in the 12q22 deleted region in male GCTs. The map is composed of 53 YACs and 3 cosmids that cover the interval between D12S101 and D12S346. Each bar below the solid line represents a YAC or cosmid (denoted with prefix ‘‘c’’). YAC and cosmid clones shown in thick lines indicate mapping by FISH. The markers that have been ordered on the contig are shown above the solid bar by vertical lines (distances in the map are not to scale). Small brackets above the solid bar indicate that the relative order of the markers could not be unambiguously determined. The numbers below the solid line indicate the genetic map positions on chromosome 12. Filled circle, polymorphic marker; filled square, nonpolymorphic marker; triangle, gene; A bracket within the YAC indicates the lack of marker(s). Empty small square or circle indicates that the marker was not tested. Filled small rectangle indicates the YAC-end marker. Large bracket on the bottom indicates the minimal region of deletion.

Analysis of candidate genes. Four known genes mapped to 12q22 were considered as candidate genes for the TSG because of their role in the development of stem cell lineages or cell growth. These genes included MGF (mast cell growth factor) (Martin et al., 1990; Mathew et al., 1992), the ligand for c-kit protooncogene, which plays an important role in the normal development of germ cell lineages; BTG1 (B-cell translocation gene 1) (Rouault et al., 1992, Kucherlapati et al., 1994), which belongs to a family of antiproliferative genes; TMPO (human thymopoietin) (Harris et al., 1995), human homologue of the rat lamina-associated polypeptide 2, which plays a role in regulation of nuclear architecture during mitosis; and NEDD1 (Takai et al., 1995),

AID

Genom 4225

/

6r1b$$$263

07-01-96 14:31:48

which has been suggested to play a role in development, differentiation, or growth arrest (Kumar et al., 1994). The MGF gene, which was homozygously deleted in a single seminomatous tumor (tumor 143A), was not placed in the contig since no YAC could be identified using STSs specific to this gene. RH map analysis also suggested that the gene is outside this linkage group. The BTG1 gene was mapped immediately distal to the RFLP locus D12S7, the marker shown to be homozygously deleted in male GCT. We have performed LOH analysis on BglII-digested DNA using the RFLP marker D12S52 (R7) in the promoter region of BTG1 and found that 23% (6 of 26) of informative tumors showed LOH. Tumor 143A was homozygously deleted

gnma

AP: Genomics

568

MURTY ET AL.

FIG. 4. Comparison of radiation hybrid map with the YAC contig map generated by STS-content analysis between D12S101 and D12S346 at 12q22. The loci mapped on each map are shown on either side of the map (distances in the map are not to scale). Chromosomal region is shown in thick vertical line. Small bracket at left on the physical map indicates that the relative map order of these markers is not determined. Brackets at right on the RH map indicate sets of loci that exhibited identical retention patterns in hybrids and were therefore considered to be in the same linkage group. cen, centromere; tel, telomere; cR, centiray.

for BTG1 gene by PCR analysis of tumor DNA isolated from microdissected paraffin-tissue sections (data not shown). Northern blot analysis of total RNA isolated from 7 cell lines derived from GCTs, however, showed higher levels of expression than in normal testis (data not shown). The coding region of BTG1 gene was screened for mutations by PCR-SSCP in tumor DNA from all cases that showed deletions at 12q22, and this analysis did not reveal any evidence suggestive of mutations (data not shown). The other 2 genes, TMPO and NEDD1, were also ruled out as candidate genes as they showed consistently higher mRNA expression in 7 GCT cell lines compared to normal testis and did not exhibit any genomic alterations in 45 tumor DNAs by Southern blot analysis (data not shown). DISCUSSION

Although the genetic mechanisms involved in transformation of germ cells are unknown, the cytogenetic data support the conclusion that genes on 12p and 12q play a role in the development of male GCT (Chaganti et al., 1996). The possibility of a potential tumor sup-

AID

Genom 4225

/

6r1b$$$263

07-01-96 14:31:48

pressor gene on 12q in male GCT was initially suggested based on high frequency of cytogenetic deletions (Murty et al., 1990; Rodriguez et al., 1992) and from the data that showed that more than 40% of tumors with LOH identified sites of candidate TSGs at 12q13.2–q13.3 and 12q22. The fact that the 12q22 region exhibited a homozygous deletion justified further search of this region for a candidate TSG (Murty et al., 1992). In the present study we have characterized the 12q22 deletion in detail and showed that 51% of the tumors exhibited LOH at this site. We observed LOH at 12q22 in all histologic subtypes of male GCTs, although with higher frequencies in teratomas (62.5%) and mixed tumors (69%) compared to embryonal carcinomas (28%), consistent with our previous observations (Murty et al., 1992; 1994b). The presence of 12q22 deletions in all histologic groups suggested an important role for a TSG at this site in the development of male GCTs. The pattern of LOH on 12q22 defined the minimal region between D12S377 and D12S296, identifying boundaries for the TSG in male GCTs. We have described the construction of a physical map of a 3-cM region encompassing the 12q22-deleted region between the markers D12S101 proximally and D12S346 distally. The map is composed of 53 overlapping YACs onto which we have placed 25 STS markers in a unique order. Several overlapping YACs span the common region of deletion between D12S377 and D12S296. Although the sizes of these YACs are known, the exact size of DNA shared by them is not known. The sizes of YACs in the minimal deleted region in several combinations estimated the region of deletion to be a minimum of 1.7 Mb and a maximum of at least 2.6 Mb based on these estimates. Several of the YACs mapped in the region of deletion were determined to be nonchimeric by FISH analysis; they should prove to be useful reagents in further analysis of the region. To define whether the homozygous deletion in T143A also spans the common region of deletion in the present study, we performed a multiplex PCR analysis using markers on 12q22 combined with markers mapped to different chromosomal regions (D12S123, D5S432) on DNA isolated from normal, frozen tumor tissue, and tumor cells microdissected from paraffin sections of this tumor. The combined data of LOH and multiplex PCR analyses suggested that the deletion in T-143A maps between D12S81 proximally and D12S346 distally. The region between the markers D12S7 proximally and D12S2085 (Krauter et al., 1995) distally showed suggestive evidence for homozygous deletion, which is proximal to the common region of deletion identified in the present study (data not shown). To confirm independently the physical map order derived by YAC STS-content mapping, we have constructed an RH map of the 12q22-deleted region ordering 14 markers to unique positions. The RH map included all the markers used for LOH except the

gnma

AP: Genomics

PHYSICAL MAP OF 12q22 DELETED REGION IN GERM CELL TUMORS

proximal D12S81, D12S379, D12S7, and D12S52 loci and the distal D12S58 and D12S234 loci. Comparison of YAC STS-content map order with the consensus RH map order revealed no discrepancies (Fig. 4). In addition, one pair of markers, D12S1300/D12S393, that could not unequivocally be ordered by STS-content mapping were resolved by the RH map (Fig. 4). The distance for the radiation hybrid map between D12S101 and D12S346 was estimated to be 246.2 cR8000 or 7.4 Mb of DNA. The RH map of the region between D12S377 and D12S296 is 141 cR8000 with an estimated physical distance of 4.23 Mb. The estimated size of the RH map between D12S377 and D12S393 compares reasonably well with the YAC size estimate (3.2 Mb vs 2.6 Mb). Although the reason for the larger estimate of distance by RH mapping is not clear, it could be due to a relatively low retention of the 12q22 region in the hybrids or scoring false-negatives due to low amplification with certain primer pairs. The RH mapping unit may vary in size of DNA per centiray from chromosome to chromosome. Low retention frequency of marker loci located in distal regions of chromosomes has also been reported (Walter et al., 1994; Shaw et al., 1995). The identification of the smallest common region of deletion in male GCTs and construction of extensive YAC contig and RH maps in the consensus region of deletions on 12q22 presented here provide a first detailed physical map of the region. The construction of such detailed maps will allow the subsequent placement of additional RFLPs, STSs, ESTs, and genes in this physical map. Construction of small genomic clone contigs such as P1s and cosmids will be valuable in generating additional polymorphic markers in the region. Analysis of such polymorphic markers will permit narrowing the region of deletion to a size that is amenable to positional cloning strategies. ACKNOWLEDGMENTS This work was supported in part by the Cancer center grants from the NIH to MSKCC (CA05826) and AECOM (CA13330), NIH Grants HG00965 (R.K.) and GM29177 (C.T.F.), and the Byrne Fund (V.V.V.S.M., R.S.K.C.). We acknowledge the technical help of R-G. Li. We thank Dr. Sharad Kumar (Adelaide) for providing a cDNA probe of NEDD1 gene and Dr. Crafford Harris for the TMPO gene probe. We also thank Dr. P. H. Rao for help in FISH mapping of YACs and Dr. Janine M. LeBlanc-Straceski for help in the initial stages of the work.

REFERENCES Boehnke, M., Lange, K., and Cox, D. R. (1991). Statistical methods for multipoint radiation hybrid mapping. Am. J. Hum. Genet. 49: 1174–1188. Chaganti, R. S. K., Murty, V. V. V. S., and Bosl, G. J. (1996). Molecular genetics of male germ cell tumors. In ‘‘Comprehensive Textbook of Genitourinary Oncology’’ (N. J. Vogelzang, P. T. Scardino, W. V. Shipley, and D. S. Coffey, Eds.), pp. 932–940, Williams & Wilkins, Baltimore, MD. Cox, D. R. (1995). Mapping with radiation hybrids. Genome Digest 2: 14–15. Cox, D. R., Burmeister, M., Price, E. R., Kim, S., and Myers, R. M.

AID

Genom 4225

/

6r1b$$$263

07-01-96 14:31:48

569

(1990). Radiation hybrid mapping: A somatic cell genetic method for constructing high-resolution maps of mammalian chromosomes. Science 250: 245–250. Falk, C. T. (1991). A simple method for ordering loci using data from radiation hybrids. Genomics 9: 120–123. Falk, C. T., and Falk, H. (1995). Chromosome mapping: Radiation hybrid data and stochastic spin models. Phys. Rev. E 52: 895–900. Fey, M. F., Hesketh, C., Wainscoat, J. S., Gendler, S., and Thein, S. L. (1989). Clonal allele loss in gastrointestinal cancers. Br. J. Cancer 59: 750–754. Gyapay, G., Morissette, J., Vignal, A., Dib, C., Fizames, C., Millasseau, P., Marc, S., Bernardi, G., Lathrop, M., and Weissenbach, J. (1994). The 1993–1994 Genethon human genetic linkage map. Nature Genet. 7: 246–339. Hahn, S. A., Seymour, A. B., Hoque, A. T. M. S., Schutte, M., da Costa, L. T., Redston, M. S., Caldas, C., Weinstein, C. L., Fisher, A., Yeo, C. J., Hruban, R. H., and Kern, S. E. (1995). Allelotype of pancreatic adenocarcinoma using xenograft enrichment. Cancer Res. 55: 4670–4675. Harris, C. A., Andryuk, P. J., Cline, S. W., Mathew, S., Siekierka, J. J., and Goldstein, G. (1995). Structure and mapping of the human thymopoietin (TMPO) gene and relationship of human TMPO b to rat lamin-associated polypeptide 2. Genomics 28: 198–205. Krauter, K., Montgomery, K., Yoon, S-J., LeBlanc-Straceski, J., Renault, B., Marondel, I., Herdman, V., Cupelli, L., Banks, A., Lieman, J., Menninger, J., Bray-Ward, P., Nadkarni, P., Weissenbach, J., Le Paslier, D., Figault, P., Chumakov, I., Cohen, D., Miller, P., Ward, D., and Kucherlapati, R. (1995). A second-generation YAC contig map of human chromosome 12. Nature 377(Suppl.): 321–333. Kucherlapati, R., Craig, I., and Marynen, P. (1994). Report of the second international workshop on human chromosome 12 mapping 1994. Cytogenet. Cell Genet. 67: 245–276. Kumar, S., Matsuzaki, T., Yoshida, Y., and Noda, M. (1994). Molecular cloning and biological activity of a novel developmentally regulated gene encoding a protein with b-transducin-like structure. J. Biol. Chem. 269: 11318–11326. LeBlanc-Straceski, J. M., Montgomery, K. T., Kissel, H., Murtaugh, L., Tsai, P., Ward, D. C., Krauter, K. S., and Kucherlapati, R. (1994). Twenty-one polymorphic markers from human chromosome 12 for integration of genetic and physical maps. Genomics 19: 341–349. Martin, F. H., Suggs, S. V., Langley, K. E., Lu, H. S., Ting, J., Okino, K. H., Morris, C. F., McNiece, I. K., Jacobsen, F. W., Mendiaz, E. A., Birkett, N. C., Smith, K. A., Johnson, M. J., Parker, V. P., Flores, J. C., Patel, A. C., Fisher, E. F., Erjavec, H. O., Herrera, C. J., Wypych, J., Sachdev, R. K., Pope, J. A., Leslie, I., Wen, D., Lin, C-H., Cupples, R. L., and Zsebo, K. M. (1990). Primary structure and functional expression of rat and human stem cell factor DNAs. Cell 63: 203–211. Mathew, S., Murty, V. V. V. S., Hunziker, W., and Chaganti, R. S. K. (1992). Subregional mapping of 13 single-copy genes on the long arm of chromosome 12 by fluorescence in situ hybridization. Genomics 14: 775–779. Mitelman, F. (1994). Catalog of chromosome aberrations in cancer, 5th ed., Wiley–Liss, New York. Montgomery, K. T., LeBlanc, J. M., Tsai, P., McNinch, J. S., Ward, D. C., DeJong, P. J., Kucherlapati, R., and Krauter, K. S. (1993). Characterization of two chromosome 12 cosmid libraries and development of STSs from cosmids mapped by FISH. Genomics 17: 682– 693. Murty, V. V. V. S., Dmitrovsky, E., Bosl, G. J., and Chaganti, R. S. K. (1990). Nonrandom chromosome abnormalities in testicular and ovarian germ cell tumor cell lines. Cancer Genet. Cytogenet. 50: 67–73. Murty, V. V. V. S., Houldsworth, J., Baldwin, S., Reuter, V., Hunziker, W., Besmer, P., Bosl, G., and Chaganti, R. S. K. (1992). Allelic deletions in the long arm of chromosome 12 identify sites

gnma

AP: Genomics

570

MURTY ET AL.

of candidate tumor suppressor genes in male germ cell tumors. Proc. Natl. Acad. Sci. USA 89: 11006–11010. Murty, V. V. V. S., Bosl, G. J., Houldsworth, J., Meyers, M., Mukherjee, A. B., Reuter, V., and Chaganti, R. S. K. (1994a). Allelic loss and somatic differentiation in human male germ cell tumors. Oncogene 9: 2245–2251. Murty, V. V. V. S., Li, R-G., Houldsworth, J., Bronson, D. L., Reuter, V. E., Bosl, G. J., and Chaganti, R. S. K. (1994b). Frequent allelic deletions and loss of expression characterize the DCC gene in male germ cell tumors. Oncogene 9: 3227–3231. Murty, V. V. V. S., Li, R-G., Mathew, S., Reuter, V. E., Bronson, D. L., Bosl, G. J., and Chaganti, R. S. K. (1994c). Replication errortype genetic instability at 1q42–43 in human male germ cell tumors. Cancer Res. 54: 3983–3985. Reifenberger, G., Reifenberger, J., Ichimura, K., and Collins, V. P. (1995). Amplification at 12q13–14 in human malignant gliomas is frequently accompanied by loss of heterozygosity at loci proximal and distal to the amplification site. Cancer Res. 55: 731–734. Rodriguez, E., Mathew, S., Reuter, V., Ilson, D. H., Bosl, G. J., and Chaganti, R. S. K. (1992). Cytogenetic analysis of 124 prospectively ascertained male germ cell tumors. Cancer Res. 52: 2285–2291. Rouault, J. P., Rimokh, R., Santalucia, B., Gadoux, M., Dorleac, E., Tessa, C., Germain, D., Samarut, J., and Magaud, J. P. (1991). Isolation and mapping of a polymorphic DNA sequence (R7) on chromosome 12 (D12S52). Nucleic Acids Res. 19: 193. Rouault, J-P., Rimokh, R., Tessa, C., Paranhos, G., Ffrench, M., Duret, L., Garoccio, M., Germain, D., Samarut, J., and Magaud, J-P. (1992). BTG1, a member of a new family of antiproliferative genes. EMBO J. 11: 1663–1670. Samaniego, F., Rodriguez, E., Houldsworth, J., Murty, V. V. V. S.,

AID

Genom 4225

/

6r1b$$$264

07-01-96 14:31:48

Ladanyi, M., Lele, K. P., Chen, Q., Dmitrovsky, E., Geller, N. L., Reuter, V., Jhanwar, S. C., Bosl, G. J., and Chaganti, R. S. K. (1990). Cytogenetic and molecular analysis of human male germ cell tumors: Chromosome 12 abnormalities and gene amplification. Genes Chromosomes Cancer 1: 289–300. Sano, T., Tsujino, T., Yoshida, K., Nakayama, H., Haruma, K., Ito, H., Nakamura, Y., Kajiyama, G., and Tahara, E. (1991). Frequent loss of heterozygosity on chromosomes 1q, 5q, and 17p in human gastric carcinomas. Cancer Res. 51: 2926–2931. Schneider, B. G., Pulitzer, D. R., Brown, R. D., Prihoda, T. J., Bostwick, D. G., Saldivar, V., Rodriguez-Martinez, H. A., GutierrezDiaz, M. E. C., and O’Connel, P. (1995). Allelic imbalance in gastric cancer: An affected site on chromosome arm 3p. Genes Chromosomes Cancer 13: 263–271. Seymour, A. B., Hruban, R. H., Redston, M., Caldas, C., Powell, S. M., Kinzler, K. W., Yeo, C. J., and Kern, S. E. (1994). Allelotype of pancreatic adenocarcinoma. Cancer Res. 54: 2761–2764. Shaw, S. H., Farr, J. E. W., Thiel, B. A., Matise, T. C., Weissenbach, J., Chakaravarti, A., and Richard, C. W., III (1995). A radiation hybrid map of 95 STSs spanning human chromosome 13q. Genomics 27: 502–510. Takai, S., Yoshida, Y., Noda, M., Yamada, K., and Kumar, S. (1995). Assignment of the developmentally regulated gene NEED1 to human chromosome 12q22 by fluorescence in situ hybridization. Hum. Genet. 95: 96–98. Walter, M. A., Spillett, D. J., Thomas, P., Weissenbach, J., and Goodfellow, P. N. (1994). A method for constructing radiation hybrid maps of whole genomes. Nature Genet. 7: 22–28. Weissenbach, J., Gyapay, G., Dib, C., Vignal, A., Morissette, J., Millasseau, P., Vaysseix, G., and Lathrop, M. (1992). A second-generation linkage map of the human genome. Nature 359: 794–801.

gnma

AP: Genomics