¹Division of Pathology and Laboratory Medicine, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, TX 77030-4095, USA

²Department of Urology, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, TX 77030-4095, USA

³Department of Epidemiology, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, TX 77030-4095, USA

Correspondence to: Jacob Kagan, Section of Experimental Laboratory Medicine, Division of Pathology and Laboratory Medicine, Box 054, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, Texas TX 77030-4095, USA

We studied loss of heterozygosity (LOH) on human chromosome 13q in prostate cancer specimens to determine the location of a putative tumor suppressor gene (TSG) and to correlate these losses with the clinicopathological stage of the disease. Overall 13 (21%) of 61 specimens analysed had an allele loss on the long arm of chromosome 13. The most frequent (37%) LOH among the informative cases with allele losses was detected at the D13S284 locus on chromosome 13q14.3. A portion of the DNA segment that spans this locus and is flanked by the microsatellite loci D13S153 and D13S163 was lost in 85% of the specimens with allele losses and was designated as a LOH cluster region (LCR). The LCR spans more than 6 Mbp of DNA. The results suggest that a TSG relevant for the development of prostate cancer is located telomeric to the RB locus. There was a significant correlation (P=0.0024) between chromosome 13q LOH and advanced metastatic disease, suggesting that loss of 13q14.3 region is associated with prostate cancer progression. However, further research must be conducted to establish the identity and function of this putative TSG.

chromosome 13; prostate cancer; loss of heterozygosity; tumor suppressor gene

Introduction

Prostate cancer (PC) is the most common cancer in males, with an estimated 179 000 new cases to be diagnosed in 1999 in the US. The death rate due to PC remains second only to that of lung cancer with an estimated 37 000 PC deaths for 1999 (Landis et al., 1999). Despite the magnitude of the problem and efforts to improve the outcome of the disease, progress in developing new therapies to combat PC has been hampered by a lack of knowledge about the affected genes, genes that are targets for inactivation or deregulation in PC, and about their biological functions.

Known genetic alterations in PC include allele losses and increases in gene copy numbers (Visakorpi et al., 1995). While an increase in the gene copy number in a specific chromosomal region is highly associated with activation/deregulation of an oncogene (e.g., ErbB-2/Her-2/Neu and Int-2 are amplified in ~20% of advanced breast cancers and N-Myb is frequently amplified in Stage III and IV neuroblastoma) (Weinstein, 1996), the frequent loss of alleles at a specific chromosomal region implicates that region as the site of a tumor suppressor gene (TSG) that become inactivated during tumor development and or progression (Weinberg, 1991).

In PC, the most frequent allele losses have been detected on chromosomes 7q, 8p, 10q, 13q, 16q and 18q (Visakorpi et al., 1995; Carter et al., 1990; Kunimi et al., 1991; Bergerheim et al., 1991; Bova et al., 1993; Zenklusen et al., 1995; MacGrogan et al., 1994; Trapman et al., 1994; Kagan et al., 1995; Suzuki et al., 1995; Macoska et al., 1995; Cooney et al., 1996; Latil et al., 1995; Cunningham et al., 1996; Li et al., 1998). However, except for the chromosome 10q23 gene, MMACI/PTEN, a candidate TSG in the very advanced stage of PC, no candidate TSGs has been cloned (Steck et al., 1997; Li et al., 1997; Teng et al., 1997; Cairns et al., 1997; Suzuki et al., 1998). Genetic alterations in the p53 TSG were detected in a small percentage of primary prostate tumors and cell lines (Mottaz et al., 1997; Salem et al., 1997; Brooks et al., 1996). p53 gene mutation is a late event in the progression of prostate cancer and is associated with advanced (metastatic) stage, loss of differentiation, and the transition from androgen-dependent to androgen-independent growth (Navone et al., 1993). Alterations at the DCC locus were associated with progression and metastasis (Ueda et al., 1997; Brewster et al., 1994).

The involvement of chromosome 13q allele losses in PC was reported as a frequent event in a limited number of studies. A comparative genomic hybridization (CGH) study reported allele losses in 32% of untreated primary tumors and in 56% of locally progressive primary tumors after hormonal therapy (Visakorpi et al., 1995). A wide range (9 - 48%) of loss of heterozygosity (LOH) in the chromosomal region that includes the retinoblastoma (RB) TSG locus, on chromosome 13q14, were reported in several studies (Carter et al., 1990; Kunimi et al., 1991; Cooney et al., 1996; Cunningham et al., 1996; Li et al., 1998). However, conflicting results were reported regarding the involvement of the RB TSG in prostate cancer. Bookstein et al. (1990) reported loss of expression of RB in a single specimen out of seven tumors, including primary and metastatic specimens, due to the loss of one allele and inactivation of the other copy by a small deletion in the promoter region. Phillips et al. (1994) and Brooks et al. (1995), using a RB intragenic microsatellite probe reported LOH in 67 and 27% of their respective specimens. However, Sarkar et al. (1992) were unable to identify any specific alterations in the RB gene in any of the primary PC specimens examined. Furthermore, more recent studies (Cooney et al., 1996; Li et al., 1998) found no correlation between specimens with LOH at the RB intragenic microsatellite locus and the expression of the RB1 protein in PC specimens, suggesting that another gene, in close proximity to the RB locus, might be the target for gene inactivation/deregulation. The involvement of another chromosome 13q TSG, BRCA2, located at 13q11 - 12, was not confirmed in these studies.

In the present study, we attempted to further refine the location of a chromosome 13q PC TSG and to correlate allele losses with the clinicopathological stage of the disease. We analysed 61 PC specimens in 18 microsatellite loci spanning the entire long arm of chromosome 13.

Results

Sixty-one tumor samples derived from 55 patients were subjected to chromosome 13q LOH analysis in 18 microsatellite loci (Figure 1). LOH occurred in at least one locus of chromosome 13q in 13 (21%) of the 61 tumors including 12 primary tumors (12/57, 21%) and one metastasis (1/4, 25%). Some of the allele losses in these specimens are shown in Figure 2. The most frequent LOH was detected at D13S284, where six (37%) of 16 informative specimens had an allele loss (Figure 1). Furthermore, the portion of the DNA segment flanked by the microsatellite markers D13S153 and D13S163 was lost in 11 (85%) of the 13 specimens with chromosome 13q allele losses. This region was designated as the LOH cluster region (LCR) (Figure 3). The probability that three or more allelic losses occurred within the LCR by independent events is very small, so such a series of LOH in contiguous markers was more likely due to deletion of the entire segment. In the majority of the PC specimens with LOH, these deletions were interstitial. Further analysis of a histogram of the LOH data from our tumor samples (Figure 1) by the Kolmogorov-Smirnov test (Zar, 1991) indicated that the data was normally distributed, as would be expected from a stochastic process such as the inactivation of a TSG.

Most of the allele losses clustered between markers D13S153 and D13S163 (Figures 1 and 3) separated by an ~1-cM region, according to the UK Linkage Database, (LDB) and Collins et al. (1996), or by an ~3-cM region, according to the Genethon linkage map (Dib et al., 1996). The most frequent allele losses occurred at D13S284, which is located in close proximity but telomeric to D13S153, the RB intragenic microsatellite marker (Phillips et al., 1994; Brooks et al., 1995). To further limit this region to the smallest common region of deletion (SCD), we used additional three microsatellite probes located between D13S153 and D13S163. However, the additional data did not further restrict this region (Figure 3). The physical map of the LCR, based on the Whitehead Institute yeast artificial chromosome (YAC) contig, is provided in Figure 4. The LCR spans more than 6 Mbp of DNA (Figure 4).

A low frequency of LOH was detected in close proximity to or within chromosome 13q TSGs loci BRCA2 and RB. Only 7.4% of the specimens had an allele loss at the D13S260 locus, which is located in close proximity to the BRCA2 TSG (Pearce et al., 1996; Collins et al., 1996). A similar frequency of LOH (6.25%) was observed in the RB intragenic microsatellite locus, D13S153 (Figure 1). Other loci also demonstrated low frequency (0 - 6.4%) of LOH (Figure 1). Allele losses in these chromosomal regions seem to be unrelated to the inactivation of the PC TSG.

Correlation of LOH with clinicopathological data

LOH on chromosome 13 was present in 27% (12/57) of the primary tumors and in 25% (1/4) of the metastasis (Table 1). LOH was detected in 24.5% (12/49) of the high grade (Gleason score 7) and in none (0/8) of the low grade (Gleason score 6) primary tumors (Table 1). There was no statistically significant difference in chromosome 13q LOH between PC specimens with Gleason score <7 and PC specimens with Gleason score 7 (² test, P=0.081) (Table 2). However, the P value was close to statistically significant value. The incidence of LOH among organ confined tumors (12.5%), tumors that demonstrated extraprostatic extension (23%), or tumors with seminal vesicle involvement (20%) was not significantly different. However, there was a statistically significant difference in the incidence of LOH between the tumors without (9/49, 18.3%) and with metastasis (5/7, 71.4%), P=0.0024 (Table 2). LOH was present in one of the three pelvic lymph node metastasis (33%) and was not detected in the single bone metastasis analysed. These results suggest that loss of chromosome 13q14.3 region is associated with the progression of PC.

Discussion

Frequent loss of alleles at a specific chromosomal region is highly associated with the inactivation of a TSG (Weinberg, 1991). Frequent LOH on the long arm of chromosome 13 in a variety of cancers was previously reported to cluster within two chromosomal bands, 13q12 (Pearce et al., 1996; Gudmundsson et al., 1995; Kerangueven et al., 1995) and 13q14 (Visakorpi et al., 1995; Cooney et al., 1996; Pei et al., 1995; Yoo et al., 1994, Melamed et al., 1997). These chromosomal regions harbor the TSGs BRCA2 and RB, respectively (Visakorpi et al., 1995; Tavtigian et al., 1996).

Although there is little evidence to implicate BRCA1 and BRCA2 in PC, a recent report suggests that there is high risk for PC in Icelandic families with a founder mutation in the BRCA2 TSG (Sigurdsson et al., 1997). To determine whether BRCA2 is inactivated by allelic loss in PC we examined LOH at the D13S260 locus, which is in close proximity to the BRCA2 gene (Pearce et al., 1996). The low frequency of LOH (7.4%) at the D13S260 locus (Figure 1) suggests a limited role for BRCA2 TSG in PC pathogenesis. These results are in agreement with those from previous studies (Kunimi et al., 1991; Cooney et al., 1996; Li et al., 1998).

The involvement of the RB TSG was assessed by detection of LOH at the D13S153 microsatellite, an intragenic microsatellite in the RB locus. Although the frequency of informative cases at this locus was low (40%), the very low (6.2%) frequency of LOH detected, suggests that RB is not the primary target for inactivation by allelic loss in primary PC (Figures 1 and 3). This data further confirms recent observations (Cooney et al., 1996; Li et al., 1998). However, in contrast to the recent publication by Li et al. (1998), we were able to map with high precision only a single LCR, suggesting that chromosome 13q harbors a single PC TSG. The second region which was located centromeric to the RB locus (Li et al., 1998) was not confirmed in our study. The LCR was located between D13S153 and D13S163, separated by more than 6 Mbp of DNA (Figure 4). The physical distance between D13S284 and the RB locus was previously estimated to be ~200 Kbp (UK LDB; Collins et al., 1996). However, our physical map of the region, based on the Whitehead Institute YACs, indicates that D13S284 is located more than 3 Mbp telomeric to D13S153, the RB intragenic microsatellite (Figure 4). The closest microsatellite to the RB locus in our study, D13S165, is located ~1.5 Mbp telomeric to D13S153. D13S284 is the microsatellite locus with the highest frequency (37.5%) of allele losses among the specimens with chromosome 13q allele losses (Figure 1). The results suggest that a putative PC TSG is located telomeric to the RB locus. Interestingly, Melamed et al. (1997), reported high frequency of LOH at the RB and D13S272 loci, especially in clinically advanced (stage C and D) PC. While the RB locus is close to but outside the LCR, D13S272 is within (Figures 3 and 4). In Melamed's study loss of allele at D13S153 was always accompanied by a loss of allele at D13S272 and conversely (Melamed et al., 1997). Similar association was not observed in our study. This should not be a surprise since the previously estimated physical distance (UK LDB; Collins et al., 1996) between these markers is incorrect. According to the YAC contig map, D13S153 and D13S272 are separated by more than 2 Mbp of DNA (Figure 4). The region adjacent to D13S272 was previously named DBM (deleted in B cell malignancies) and was suspected to harbor a B cell chronic lymphocytic leukemia (CLL) TSG (Brown et al., 1993). Two genes, Leu-1 and Leu-2, were located within 130 Kbp segment of DNA centromeric to D13S272, and were proposed as candidate TSGs in CLL (Liu et al., 1997). Although currently we cannot rule out that these genes are not candidate TSGs in PC, the relative low frequency of LOH at D13S272 (14.3%) as compared to D13S284 (37.5%) may suggest that the PC TSG is located closer to D13S284. It is obvious that further efforts will be needed to pinpoint the location of the PC candidate TSG within the LCR.

There was a significant correlation (P=0.0024) between chromosome 13q LOH and advanced metastatic PC (Table 2). Although there was no significant association between higher Gleason score (7) and LOH in the present study (P=0.081), it should be noted that 29% of the higher grade tumors had allele losses, while none of the lower grade tumors had any allelic loss (Tables 1 and 2). This observation is further supported by a previous study that found positive correlation between chromosome 13q LOH and higher Gleason grade tumors (Cooney et al., 1996; Melamed et al., 1997). Overall, the data suggests that loss of chromosome 13q14.3 region is highly associated with progressive disease.

Interestingly, in several studies of different cancers including prostate, pituitary, CLL, breast and small cell carcinoma of the head and neck (Visakorpi et al., 1995; Cooney et al., 1996; Pearce et al., 1996; Kerangueven et al., 1995; Pei et al., 1995; Yoo et al., 1994; Mastero et al., 1996; Melamed et al., 1997), some of the minimal regions with LOH overlapped our LCR (Figure 5). It is possible that inactivation of the PC candidate TSG is common to some or all of these cancers.

Materials and methods

Prostate cancer specimens

Sixty-one tumor specimens derived from 55 patients were included in the study. The specimens included 57 primary tumors (single tumor focus from 49 patients; two separate, well-separated, tumor foci from six patients) and metastatic tumors (three pelvic lymph nodes and one bone metastasis). The primary tumors were derived from radical prostatectomy (53 cases) and transurethral resection specimens (four cases). Metastatic tumors were obtained from retroperitoneal lymph node dissection (three cases) and by true-cut biopsy of the iliac crest. Twenty-five specimens were fresh frozen tumors and 36 were formalin-fixed paraffin embedded tumors. The fresh tumors were obtained from radical prostatectomy specimens following a detailed procedure as described previously (Kagan et al., 1995). A mirror image section of each tumor specimen was analysed histologically, and the content of tumor cells was estimated. Only specimens composed of more than 70% of cancer in the control section were selected for the study. As a normal control we used peripheral blood or the tip of the seminal vesicle contralateral to the tumor, if it was histologically confirmed to be tumor free. For formallin-fixed paraffin embedded specimens, a hematoxylin and eosin stained section of the selected block was used to map the area of tumor and normal tissue. The corresponding areas were scraped from unstained-sections using a scalpel (eight sections of 8 m). For most of the archival specimens selected for the study a control DNA was also extracted form blood leukocytes of the same patient. There was no difference in heterozygosity between DNA extracted form normal prostatic tissue from archival specimen and the blood leukocyte DNA from the same patient. Tumor specimens were graded according to the Gleason system (Gleason, 1977). The primary tumors were organ confined (16 cases, pT₂N₀M₀), locally advanced (13 cases with extraprostatic extension, pT₃aN₀M₀) and with seminal vesicle involvement (20 cases, pT3bN0M0) and metastatic to regional (six cases; two pT_xN₁M₀, two pT₃aN₁M₀ and two pT₃bN₁M₀) or distant (one case, pT_xN_xM₁) lymph nodes (Classification according to Fleming et al., 1997). The pathological stage for one primary tumor was not available. The clinical or pathologic stage was not available for one case.

DNA extraction

Extraction of DNA from fresh and frozen tissue was described in detail elsewhere (Kagan et al., 1995). DNA extraction from paraffin-embedded specimens was performed as follows: A paraffin-blocked specimen was serially sectioned into eight 8 m sections, which were applied to glass slides. Every first slide was stained with hematoxylin and eosin to visualize areas of stroma and low- and high-grade carcinomas. Cancerous areas were circumscribed with ink. From adjacent sections, the portions of tissue corresponding to the outlined cancerous areas were scraped with a sterile scalpel blade into a plastic Eppendorf tube. Deparaffinization was performed by twice adding 400 l of xylene and incubating the tube for 1 h at 55°C. The specimens were washed three times with ethanol (100, 95 and 100%) and air dried. Next, 500 l of a lysis buffer (100 mM NaCl; 10 mM Tris-HCl, pH 7.5; 100 mM EDTA; 0.5% SDS; 1 mg/ml proteinase K) was added and the specimens were incubated overnight at 50°C. The DNA was phenol/chloroform extracted and ethanol precipitated with 20 g glycogen as a carrier. The DNA pellet was washed with 70% ethanol dried and resuspended in 100 l of water.

Polymerase chain reaction (PCR) amplification of microsatellite sequences

Primers for microsatellite sequences were obtained from Research Genetics (Huntsville, AL, USA) and are listed in Figure 1. PCR amplification was carried out as described in detail (Kagan et al., 1995).

Analysis of loss of heterozygosity (LOH)

Normal DNA samples (from archival material, tip of seminal vesicle or leukocytes), which were heterozygous at a given locus (having two alleles of a microsatellite locus of different sizes) were considered `informative', where homozygotes were `noninformative'. The signal intensity of ³²P-labeled PCR amplified microsatellite alleles were determined from X-ray film autoradiography. X-Omat (Kodak, Rochester, New York, USA), by densitometry (Pharmacia Biotech, Uppsala, Sweden), and or by visual examination (three independent observers). LOH was considered to occur if an entire allele was absent or if the intensity of one of the two alleles in the tumor DNA was less than 50% of that in corresponding normal-tissue DNA for a given amount of DNA. Although PCR amplification cannot be considered to be quantitative, we optimized the PCR conditions so that equal amount of template should have produced equal amounts of amplified product. We used 27 amplification cycles, which we demonstrated to be in the linear part of the amplification process, i.e., before product saturation (data not shown). We also conducted a series of titrations with different proportion of homozygous and heterozygous templates to assess the influence of normal tissue `contamination' in our PCR amplified tumor samples. We determined that we could detect as little as 30% `contamination' by heterozygous template in homozygous DNA (data not shown). Thus, our limit of 50% intensity for LOH is very conservative. Results of LOH were repeated at least three times.

Microsatellite sequences and genetic maps

Microsatellite sequences and the heterozygosity of each marker were obtained from the genome database at http://www.gdb.org/. Chromosomal maps and distances were obtained from the LDB at http://cedar.genetics.soton.ac.uk/pub/chrom13/map (updated on September 9, 1997) (Collins et al., 1996) and Genethon (Dib et al., 1996). YAC contig data was obtained from http://www.genome.wi.edu/cgi-bin/contig/lookup_contig

Statistical analysis

The normality of the percentage of LOH distribution was tested using the Kolmogorov-Smirnov continuous cumulative distribution test (Zar, 1991). The significance of allelic losses was correlated with clinicopathological data, stage, Gleason score, PSA level, familial history and ethnicity, using the ² test essentially as described previously (Kagan et al., 1995).

Acknowledgements

This study was supported by NCI grant CA68578-01 (MR Spitz and J Kagan) and by the University of Texas MD Anderson Cancer Center Physician Referral Service Award (J Kagan).

Bergerheim US, Kunimi K, Collins VP and Ekman P. (1991). Genes, Chromosomes Cancer 3, 215-220. MEDLINE

Bookstein R, Rio P, Madreperla SA, Hong F, Allred C, Grizzle EW and Lee WH. (1990). Proc. Natl. Acad. Sci. USA 87, 7762-7766. MEDLINE

Bova GS, Carter BS, Bussemakers MJ, Emi M, Fujiwara Y, Kyprianou N, Jacobs SC, Robinson JC, Epstein JI, Walsh PC and Issacs WB. (1993). Cancer Res. 53, 3869-3873. MEDLINE

Brewster SF, Browne S and Brown KWZ. (1994). J. Urol. 151, 1073-1077. MEDLINE

Brooks JD, Bova GS and Issacs WB. (1995). Prostate 26, 35-39. MEDLINE

Brooks JD, Bova SG, Ewing CM, Piantadosi S, Carter BS, Robinson JC, Epstein JI and Issacs WB. (1996). Cancer Res. 56, 3814-3822. MEDLINE

Brown AG, Ross FM, Dunne EM, Steel CM and Weir-Thompson EM. (1993). Nat. Genet. 3, 67-72. MEDLINE

Cairns P, Okami K, Halachmi S, Halachmi N, Esteller M, Herman JG, Jen J, Issacs PW, Bova SG and Sidransky D. (1997). Cancer Res. 57, 4997-5000. MEDLINE

Carter BS, Ewing CM, Ward WS, Treioger BF, Aalders TW, Schalken JA, Epstein JI and Isaacs WB. (1990). Proc. Natl. Acad. Sci. USA 87, 8751-8755. MEDLINE

Collins A, Frezal J, Teague J and Morton NE. (1996). Proc. Natl. Acad. Sci. USA 93, 14771-14775. Article MEDLINE

Cooney KA, Wetzel JC, Merajver SD and Macoska JA. (1996). Cancer Res. 56, 1142-1145. MEDLINE

Cunningham JM, Shan A, Wick MJ, McDonnell SK, Schaid DJ, Tester DJ, Qian J, Takahashi S, Jenkins RB, Bostwick DG and Thibodeau SN. (1996). Cancer Res. 56, 4475-4482. MEDLINE

Dib C, Faure S, Fizames C, Samson D, Drouot N, Vignal A, Millasseau P, Marc S, Hazan J, Sebon E, Lathrop M, Gyapay G, Morissette J and Weissenbach J. (1996). Nature 380, 152-154. MEDLINE

Fleming ID, Cooper JS, Henson DE, Huntter RVP, Kennedy BJ, Murphy GP, O'Sullivan B, Sobin LH and Yarbro JW (eds). (1997). AJCC Cancer Staging Manual, 5th edition. Lippincott-Raven, Philadelphia: New York.,

Gleason D. (1977). Urologic Pathology: The prostate. Tannenbaum M (ed.). Lea & Febiger: Philadelphia, pp. 171-197.

Gudmundsson J, Johannesdottir ZG, Bergthorsson JT, Aaronson A, Ingvarsson S, Egilsson V and Barkardottir RB. (1995). Cancer Res. 55, 4830-4832. MEDLINE

Kagan J, Stein J, Babaian RJ, Joe Y, Pisters LL, Glassman AB, von Eschenbach AC and Troncoso P. (1995). Oncogene 11, 2121-2126. MEDLINE

Kerangueven F, Allione F, Noguchi T, Adelaide J, Sobol H, Jacquemier J and Birinbaum D. (1995). Genes, Chromosomes Cancer 13, 291-294. MEDLINE

Kunimi K, Bergerheim USR, Larsson IL, Ekman P and Collins VP. (1991). Genomics 11, 530-536. MEDLINE

Landis SH, Murray T, Bolden S and Wingo PA. (1999). CA: Cancer J. Clin. 4, 8-31.

Latil A, Baron JC, Cussento O, Fournier G, Soussi T, Boccon-Gbod L, Le Duc A, Rouse J and Lidereau R. (1995). Bulletin de Cancer 82, 589-597.

Li C, Larsson C, Futreal A, Lancaster J, Phelan C, Aspenblad U, Sundelin B, Liu Y, Ekman P, Auer G and Bergerheim USR. (1998). Oncogene 16, 481-487. MEDLINE

Li J, Yen C, Liaw D, Podsypanina K, Bose S, Wang SI, Puc J, Miliaresis C, Rodgers L, McCombie R, Bigner SH, Giovanella BC, Itterman M, Tycko B, Hibshoosh H, Wigler MH and Parsons R. (1997). Science 275, 194-197.

Liu Y, Corcoran M, Rasool O, Ivanova G, Ibbotson R, Grander D, Iyengar A, Baranova A, Kashuba V, Merup M, Wu X, Gardiner A, Mullenbach R, Poltaraus A, Hultstrom AL, Juliusson G, Chapman R, Tiller M, Cotter F, Gahrton G, Yankovsky N, Zabarovsky E, Einhorn S and Oscier D. (1997). Oncogene 15, 2463-2473. MEDLINE

MacGrogan D, Levy A, Bostwick D, Wagner M, Wells D and Bookstein R. (1994). Genes, Chromosomes Cancer 10, 151-159. MEDLINE

Macoska JA, Trybus TM, Benson PD, Sakr WA, Grignon DJ, Wojno KD, Pietruk T and Powell IJ. (1995). Cancer Res. 55, 5390-5395. MEDLINE

Mastero R, Piccinin S, Doglioni C, Gasparotto D, Vukosavljevic T, Sulfaro S, Barzan L and Biocchi M. (1996). Cancer Res. 56, 1146-1150. MEDLINE

Melamed J, Einhorn JM and Ittmann MM. (1997). Clin. Cancer Res. 3, 1867-1872. MEDLINE

Mottaz AE, Markwalder R, Fey MF, Klima I, Merz VW, Thalmann GN, Ball RK and Studer UE. (1997). Prostate 31, 209-215. MEDLINE

Navone NM, Troncoso P, Pisters LL, Goodrow TL, Palmer JL, Nichols WW, von Eschenbach AC and Conti CJ. (1993). Natl. Cancer Inst. 85, 1657-1669.

Pearce SH, Trump D, Woodding C, Sheppard MN, Clayton RN and Thakker RV. (1996). Clin. Endocrinol. 45, 195-200.

Pei L, Melmed S, Scheithauer B, Kovacs K, Benedict WF and Prager D. (1995). Cancer Res. 55, 1613-1616. MEDLINE

Phillips SM, Morton DG, Lee SJ, Wallace DM and Neoptolemos JP. (1994). Br. J. Urol. 73, 390-395. MEDLINE

Salem CE, Tonasic NA, Elmajian DA, Esring D, Nichols PW, Taylor CR, Skinner DG, Roy-Burman P, Lieskovsky G and Cote RJ. (1997). J. Urol. 158, 510-514. MEDLINE

Sarkar FH, Sakr W, Macoska J, Ball DE and Crissman JD. (1992). Prostate 21, 145-152. MEDLINE

Sigurdsson S, Thorlaciius S, Tomasson J, Tryggvadottir L, Benediktsdottir K, Eyfjord JE and Jonsson E. (1997). J. Mol. Med. 75, 758-761. MEDLINE

Steck PA, Pershouse MA, Jasser SA, Yung WAK, Lin H, Ligon AH, Langford LA, Baumgard ML, Hattier T, Davis T, Frye C, Hu R, Swedlund B, Teng DHF and Tavtigian SV. (1997). Nat. Genet. 15, 356-362. MEDLINE

Suzuki H, Emi M, Komiya A, Fujiwara Y, Yatani R, Nakamura Y and Shimazaki J. (1995). Genes, Chromosomes Cancer 13, 168-174. MEDLINE

Suzuki H, Freije D, Nusskern DR, Okami K, Cairns P, Sidransky D, Issacs WB and Bova SG. (1998). Cancer Res. 58, 204-209. MEDLINE

Tavtigian SV, Simard J, Rommens J, Couch F, Shattuck-Eidens D, Neuhausen S, Merajver S, Thorlacius S, Offit K, Stoppa-Lyonnet D, Belanger C, Bell R, Berry S, Bogden R, Chen Q, Davis T, Dumont M, Frye C, Hattier T, Jammulapati S, Janecki T, Jiang P, Kehrer R, Leblanc JF and Goldgar DE. (1996). Nat. Genet. 12, 333-337. MEDLINE

Teng DH-F, Hu R, Lin H, Davis T, Iliev D, Frye C, Swedlund B, Hansen KL, Vinson VL, Gumpper KL, Ellis L, El-Naggar A, Frazier M, Jasser S, Langford LA, Lee J, Mills GB, Pershouse MA, Pollack RE, Tornos C, Troncoso P, Yung WKA, Fujii G, Berson A, Bookstein R, Bolen JB, Tavtigian SV and Steck P. (1997). Cancer Res. 57, 5221-5225. MEDLINE

Trapman J, Sleddens HFBM, van der Weiden MM, Dinjens WMM, Koning JJ, Schroder FH, Faber PW and Bosman FT. (1994). Cancer Res. 54, 6061-6064. MEDLINE

Ueda T, Komiya A, Emi M, Suzuki H, Shiraishi T, Yatani R, Masai M, Yasuda K and Ito H. (1997). Genes, Chromosomes Cancer 20, 140-147. MEDLINE

Visakorpi T, Kallioniemi AH, Syvanen AC, Hytinen ER, Karhu R, Tammela T, Isola JJ and Kallioniemi OP. (1995). Cancer Res. 55, 342-347. MEDLINE

Weinberg RA. (1991). Science 254, 1138-1146. MEDLINE

Weinstein IB. (1996). J. Cell. Biochem. 25, 23-28. [Suppl].

Yoo GH, Xu HJ, Breennan JA, Westra W, Hruban RH, Koch W, Benedict WF and Sidransky D. (1994). Cancer Res. 54, 4603-4606. MEDLINE

Zar JH. (1991). Biostatistical Analysis. Prentice Hall: Englewood Cliffs, NJ.,

Zenklusen JC, Thompson JC, Troncoso P, Kagan J and Conti CJ. (1995). Cancer Res. 54, 6370-6373.

Figure 1 Frequency of chromosome 13q allele losses for each of the microsatellites loci in the informative PC specimens. Microsatellites are listed from the long arm centromer (left) to the telomer (right)

Figure 2 Representative PCR amplifications of microsatellite DNA from PC specimens. Only specimens with allele losses are presented. The case numbers are shown at the top of the lane. T and N, matched DNA samples isolated from tumor tissue and peripheral leukocytes, respectively. The microsatellite locus analysed is indicated at the bottom of each matched specimen. Arrows indicate the lost allele in the tumor specimen

Figure 3 Summary of allelic losses by specimen. Only specimens with allele losses are presented. Specimen numbers are indicated above each column. Microsatellite markers are listed in order from the most centromeric (top), to the most telomeric (bottom), and their chromosomal location is indicated. Vertical boxes in each specimen represent the extent of allele loss around regions with LOH. Arrow indicates the LOH cluster region, LCR. Filled circle, LOH; open circle, informative specimen, retention of both alleles; crossed circle, uninformative specimen for the marker tested

Figure 4 Overlapping YACs that span the LCR. YAC addresses were obtained from the Whitehead Institute Web site (see Materials and methods) and are indicated above each YAC. The size and chimerism of each YAC are listed. Markers, which were located within the YACs are indicated by short vertical bars on the YAC map. LCR, loss of heterozygosity cluster region

Figure 5 Chromosome 13q allele losses in different cancers. Distances on the map are according to UK LDB and Collins et al. (1996). The LCR in the present study is indicated by shaded region. Vertical lines represent the SCD in the nine referenced studies. The data suggest that a chromosome 13q TSG might be common to some or all of these cancers. LCR, loss of heterozygosity cluster region. HNSCC; head and neck squamous cell carcinoma

Table 1 Allele losses and pathological data

Table 2 Statistical correlation of allele losses on chromosome 13q with clinicopathological data