This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Chang, B.-l. Articles by Xu, J. Search for Related Content PubMed PubMed Citation Articles by Chang, B.-l. Articles by Xu, J.

Integration of Somatic Deletion Analysis of Prostate Cancers and Germline Linkage Analysis of Prostate Cancer Families Reveals Two Small Consensus Regions for Prostate Cancer Genes at 8p

¹ Center for Human Genomics, Wake Forest University School of Medicine, Winston-Salem, North Carolina and ² Johns Hopkins Medical Institutions, Baltimore, Maryland

Requests for reprints: William B. Isaacs, Johns Hopkins Hospital, Marburg 115, 600 North Wolfe Street, Baltimore, MD 21287. Phone: 410-955-2518; Fax: 410-955-0833; E-mail: wisaacs{at}jhmi.edu or Jianfeng Xu, Center for Human Genomics, Medical Center Boulevard, Winston-Salem, NC 27157. Phone: 336-713-7500; Fax: 336-713-7566; E-mail: jxu{at}wfubmc.edu.

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
The evidence for tumor suppressor genes at 8p is well supported by many somatic deletion studies and genetic linkage studies. However, it remains a challenge to pinpoint the tumor suppressor genes at 8p primarily because the implicated regions are broad. In this study, we attempted to narrow down the implicated regions by incorporating evidence from both somatic and germline studies. Using high-resolution Affymetrix arrays, we identified two small common deleted regions among 55 prostate tumors at 8p23.1 (9.8–11.5 Mb) and 8p21.3 (20.6–23.7 Mb). Interestingly, our fine mapping linkage analysis at 8p among 206 hereditary prostate cancer families also provided evidence for linkage at these two regions at 8p23.1 (5.8–11.2 Mb) and at 8p21.3 (19.6–23.9 Mb). More importantly, by combining the results from the somatic deletion analysis and genetic linkage analysis, we were able to further narrow the regions to 1.4 Mb at 8p23.1 and 3.1 Mb at 8p21.3. These smaller consensus regions may facilitate a more effective search for prostate cancer genes at 8p. [Cancer Res 2007;67(9):4098–103]

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Deletion of sequences from chromosome 8p is the most common deletion event in the genome of prostate tumors (1). In a recent study that estimated the frequency of DNA copy number alterations in the prostate cancer genome based upon all published comparative genomic hybridization studies of prostate cancers, we found that one third of 891 prostate cancers had a deletion at 8p21.3, considerably higher than the second most commonly deleted region at 6q15 (22.4%; ref. 2). Despite the overwhelming evidence for 8p deletions, few specific genes have been consistently implicated as prostate tumor suppressor genes in this region. One of the major obstacles in the identification of tumor suppressor genes at 8p is the size of the deleted regions, which is affected by the resolution of methods used to detect deletions. For example, in our study cited above (2), the deleted region at 8p21.3 spans a 27.1-Mb interval extending into 8p23.3 and 8p21.1 and contains many genes. Higher-resolution detection methods that can detect small deletions and complex deletion patterns are needed to identify 8p tumor suppressor genes (3).

Furthermore, results from genetic linkage studies have provided evidence for prostate cancer linkages at 8p (4, 5). In the largest genome-wide linkage analysis done to date, Xu et al. (5) found suggestive evidence for linkage at 8p21, one of the five most significant in the genome, among 1,233 prostate cancer families of the International Consortium for Prostate Cancer Genetics (ICPCG). Similar to the results of somatic deletion studies, few genes in this 8p region have been consistently implicated as major prostate cancer susceptibility genes accounting for the 8p linkage. One of the major difficulties is the low resolution of genetic linkage studies, which are typically in the range of 10 to 20 cM, due to limited meiosis events in families. For example, the 1-LOD drop interval of 8p21 linkage identified in the ICPCG study was 13 cM (39–52 cM) or 10 Mb (22–32 Mb).

Cancers are thought to arise as a result of alterations in expression of tumor suppressor genes and oncogenes in prostate epithelial cells. Altered gene expression may result from inherited genetic changes and acquired somatic genetic changes, including deletions, as hypothesized by the "two-hit" model (6). Therefore, assuming that at least some fraction of prostate cancers arise from a combination of inherited and acquired genomic events affecting the same gene or combination of genes, studies that simultaneously examine inherited genetic changes and somatic genetic alterations of chromosomal regions or genes may improve the likelihood of identifying genes involved in cancer development.

In this study, we have taken three steps to identify genomic regions that contain prostate cancer genes. First, we used high-resolution Affymetrix single nucleotide polymorphism (SNP) arrays to examine detailed deletion patterns at 8p among 55 prostate cancers. Second, we did a fine mapping linkage analysis in 206 hereditary prostate cancer (HPC) families. Finally and more importantly, we integrated results from somatic deletion analysis and germline linkage analysis to identify a consensus region.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Detection of 8p deletions in somatic DNA from prostate cancers. All subjects were prostate cancer patients undergoing radical prostatectomy for treatment of clinically localized disease at Johns Hopkins Hospital. For the somatic DNA deletion analysis of prostate cancers, we selected 55 subjects from whom genomic DNA of sufficient quantity (>5 µg) and purity (>70% cancer cells for cancer specimens, no detectable cancer cells for normal samples) could be obtained by macrodissection of matched nonmalignant (hereafter referred to as "normal") and cancer containing areas of prostate tissue as determined by histologic evaluation of H&E-stained frozen sections of snap-frozen radical prostatectomy specimens. Genomic DNA was isolated from trimmed frozen tissues as previously described (7).

We used Affymetrix SNP array panels to detect DNA copy number alterations, and for this study, we focused entirely on 8p deletions. We used the 100K SNP array for the first 22 subjects (3) and then used the 500K SNP array for the final 41 subjects (eight samples were analyzed using both 100K and 500K arrays). For all subjects, we analyzed both tumor DNA and normal DNA from the same subject using Affymetrix SNP arrays (either 100K or 500K array). The normal DNAs were extracted from histologically normal prostate tissue from the same prostate or from the seminal vesicle of the same patient. DNA copy number was calculated based on allele intensity using two different software packages: Copy Number Analyzer for Affymetrix GeneChip (CNAG2.0; ref. 8) and dChip analyzer (9). Allele-specific analysis was also done to estimate DNA copy number for each chromosome using CNAG2.0. The physical positions of detected deletions were based on the Human hg17 Assembly (NCBI Build 35). The criteria used in dChip analysis for this study are similar to those used for the 100K SNP array analyses using CNAT, which has been described in our previous publication (3). Briefly, to reduce random noise in allele intensity at individual SNPs, we first estimated DNA copy number based on flanking SNPs in the region, using the 10-SNP smoothing setting of dChip software to obtain a genome smooth average copy number (GSACN) for each SNP. We then defined deletions using the following working criteria: a minimum of four consecutive SNPs with at least three of them having the following characteristics: the GSACN ratios of tumor/matched normal <0.75, the GSACN of the tumor DNA <1.9, and the minimum physical length of the putative deletion 2 kb. To define deletions using CNAG2.0 software (for both intensity-based and allele-specific analyses), we also used a 10-SNP smoothing setting to minimize random variations at individual SNPs. Each deletion is defined by whether the log 2 ratio of probe intensity is below (no overlap) with the baseline log 2 ratio defined by the matched normal DNA. The baseline log 2 ratio has a theoretical value of zero, with small variations due to random noise (as shown in Fig. 1 ). The results from all three analyses are in agreement with each other, except that the boundaries of the deletions defined by different analyses varied from one to five SNPs.

View larger version (39K): [in this window] [in a new window]   Figure 1. Small and complex partial deletions of 8p. A, a visual summary for five of the smallest and most complex tumors that were detected at 8p. Top, distances in Mb beginning at the pter (Chr8). Allele-specific CNAG output for each of these five tumors, with red or green for each respective chromosome. Solid vertical lines highlight a region at 8p21.3-8p21.2, spanning from 20.6 to 23.7 Mb, that was deleted in all 30 tumors. Dotted vertical lines indicate a deleted region spanning from 9.8 to 11.5 Mb at 8p23.1 that is shared by 29 of the 30 tumors. B, results of deletion confirmation of the homozygous deleted regions by quantittaive real-time PCR analyses. Left, results from DBC2/RHOBTB1 primers located within the putative homozygous deletion. Right, results from LOXL2 primers located outside the homozygous deleted interval. Ct values of the control (X-axis) and test (Y-axis) amplicons for the three dilutions of each DNA sample were plotted against each other. Offsets between best-fit lines for the samples along the test amplicon axis at 25 Ct of the control amplicon axis are used to infer DNA copy number. Tumor DNA is defined as having a hemizygous deletion when Ct < 0.68 and as having homozygous deletion when Ct > 0.68, assuming 25% normal DNA contamination.

Linkage analysis in prostate cancer families and construction of a recombinant map. All 206 HPC families were collected and studied at the Brady Urology Institute at Johns Hopkins Hospital (Baltimore, MD) as described previously (4). Prostate cancer diagnosis was verified by medical records for each affected male studied. Age at diagnosis of prostate cancer was confirmed either through medical records or from two other independent sources. The mean age at diagnosis was 64.3 years for the cases in these families. Eighty-four percent of the families are non-Jewish Caucasians, 6.9% are Ashkenazi Jewish, and 8.8% are African Americans.

Thirty fine mapping microsatellite markers spanning about 35 Mb at 8p were genotyped in these 206 HPC families. Following multiplex PCR using fluorescently labeled primers, the resulting PCR fragments were separated using capillary electrophoresis using an ABI 3700 sequencer. Marker allele frequencies were estimated from the 214 independent individuals in the data set. The marker order and distances were primarily based on information available from the MAP-O-MAT web site (10). Four markers were not available in the MAP-O-MAT web site; their order and distances were interpolated from the University of California Santa Cruz (UCSC) Genome Browser.³ Multipoint linkage analyses were done using both parametric and nonparametric methods implemented by the computer program GENEHUNTER-PLUS (11, 12). For the parametric analysis, the same autosomal-dominant model that was used by Smith et al. (13) was assumed. For the nonparametric analysis, the estimated marker identical by descent (IBD) sharing of alleles for the various affected relative pairs was compared with its expected values under the null hypothesis of no linkage (NPL). A statistical "Z-all" in the program was used (14). Allele sharing LOD scores were then calculated based on the statistical "Z-all" and assigning equal weight to all families using the computer program ASM (12).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Detection of somatic DNA deletions. Detectable deletions at 8p were observed in 29 of the 55 prostate cancers (52.73%) examined in this study (Table 1 ). Although many of these deletions involved almost the entire short arm of chromosome 8, we detected partial 8p deletions in 10 of these tumors. Among these partial deletions of 8p, five were smaller or more complex (Fig. 1A). This includes a tumor with a small deletion at 8p21 (G7-022), three tumors containing multiple interstitial deletions (G7-013, G9-010, and G7-028), and two tumors with several small homozygous deletions (G7-028 and G6-002). No copy number polymorphisms were detected in these samples, and these homozygous deletions are due to somatic DNA loss.

The pattern of deletions observed among the partial 8p deletions suggests the presence of two smaller deletion regions. In particular, a region at 8p21.3-8p21.2, spanning from 20.6 to 23.7 Mb, was deleted in all 29 tumors (Fig. 1A, solid vertical lines). We detected another deleted region spanning from 9.8 to 11.5Mb at 8p23.1 that is shared by 28 of the 29 tumors (Fig. 1, dotted vertical lines). The primary reason these regions seem to be separated is due to the three tumors with interstitial deletions.

Prostate cancer linkage region. Linkage analysis of 206 prostate cancer families provided evidence for a susceptibility gene at 8p from both parametric (using a dominant model) and nonparametric linkage analyses (Fig. 2 ). Interestingly, two separate linkage peaks were observed. One peak was found at the marker D8S258 of 8p21.3 (20,411,446), with a LOD score of 2.51 (P = 0.0007) and an NPL score of 3.14. The 1-LOD drop interval spanned 4 Mb, between 19.6 and 23.9 Mb. The other peak was found at the marker D8S503 of 8p23.1 (9,270,543), with a LOD score of 1.50 (P = 0.009) and an NPL score of 2.72. The 1-LOD drop interval spanned 5.4 Mb, between 5.8 and 11.2 Mb. There were 49 families with LOD scores > 0.588 (P_nominal = 0.05) within the 8pter-8p12 region, with all but two of these families being linked to at least one of the two regions described above. Among these families, 18 families had positive LOD scores across these two regions, 16 families had positive LOD scores only at 8p21.3, and 13 families had positive LOD scores only at 8p23.1.

View larger version (19K): [in this window] [in a new window]   Figure 2. Linkage analysis results at 8p among 206 prostate cancer families. Two primary linkage peaks that we observed. Y-axis, LOD (light purple lines) and NPL (dark blue lines) scores; X-axis, physical position along the 8p arm. For the peaks observed at 8p21.3 and 8p23.1, vertical dashed lines have been used to indicate the span of 1-LOD drop intervals.

View larger version (40K): [in this window] [in a new window]   Figure 3. Combined results of somatic deletion study and germline linkage study. Overlapping results provided by our analyses of germline linkage and somatic deletion. Two consensus regions at 8p21.3 and 8p23.1 (solid vertical lines). Bottom, known and predicted genes located within these two consensus regions, based on release hg18 of the UCSC database.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
Chromosome 8p has received a great deal of attention from cancer researchers in the past decade because it is commonly deleted in prostate cancer (1, 16) as well as in many other cancers, including colon, breast, ovarian, liver, lung, bladder, and head and neck cancer. Furthermore, results from multiple genetic linkage studies provide evidence that 8p may harbor major prostate cancer susceptibility genes. Considerable efforts have been devoted to the identification of specific prostate cancer genes at 8p that account for the observations from deletion and linkage studies. Although several candidate genes at 8p have been reported to be involved in prostate cancer development, including NKX3.1, N33, MSR1, and DLC1, few are consistently implicated among different studies. One of the major difficulties is the broad genomic regions implicated in these deletion and linkage studies; for example, most of the observed deletions involve the entire 8p arm. In this study, we used two complementary methodologies in an attempt to effectively narrow the genomic region(s) harboring prostate cancer genes. We used high-resolution Affymetrix SNP arrays to define detailed deletion patterns at 8p among 55 prostate cancers. This analysis led to the identification of two small deleted regions at 8p21.3 and 8p23.1. We did a fine mapping linkage analysis at 8p among 206 HPC families and obtained evidence for linkage at these two regions. Most importantly, we combined the results from the somatic deletion analysis and genetic linkage analysis to further narrow the regions to 3.1 Mb at 8p21.3 and 1.4 Mb at 8p23.1. These much smaller consensus regions will likely facilitate more effective searches for prostate cancer genes at 8p.

The high-resolution SNP arrays provide a better tool to identify small DNA copy number alterations and to examine detailed patterns of deletions. With a denser resolution of SNPs covering the 8p region, combined with allele specific analysis, we were able to detect small deletions and better define the boundaries of deletions. In addition, the high density of SNPs revealed interstitial 8p deletions in three cancers. These findings allowed us to identify two small overlapping deleted regions at 8p21.3 and 8p23.1. It is interesting to note that these two separate deleted regions are within the single 27.1-Mb deleted region at 8p identified from a combined analysis of 891 prostate cancers as the most deleted region in the genome (2). Our current study provides evidence that the previously known commonly deleted region may consist of two separate deleted regions.

The fine mapping panel of 29 markers at 8p in our linkage study, with an average of 1-cM resolution, provides a better tool to dissect detailed linkage patterns among the 206 HPC families. In this study, we were able to confirm prostate cancer linkage at 8p among a large number of prostate cancer families. More importantly, we were able to obtain statistical evidence for two separate linkage regions. One of the linkage regions (19.6–23.9 Mb at 8p21.3) overlapped with the 1-LOD drop interval at 8p21 (22–32 Mb) reported from 1,233 ICPCG prostate cancer families (5).

As hypothesized in the "two-hit" model (10, 11), inherited genetic defects, combined with acquired somatic changes, ultimately alter the expression and/or function of tumor suppressor genes and lead to cancer. Therefore, approaches that combine information from germline and somatic studies may provide better power to identify cancer genes. This combined approach has been successfully used to identify the APC gene for familial adenomatous polyposis (FAP). Results from genetic linkage studies in FAP families, somatic loss of heterogeneity analysis, and an interstitial germline deletion all converged to a small region at 5q21 and led to the identification of the APC gene (17). Although there are large differences between the rare syndrome of FAP and prostate cancer, the principle of the "two-hit" model may still apply, and our combined approach represents a critical step toward the identification of prostate cancer genes at 8p.

It is interesting that both somatic deletion analysis of prostate tumors and germline linkage analysis of prostate cancer families identified the same genomic regions. Although by no means conclusive, this overlap is consistent with the hypothesis that the same gene or genes is affected both at the germline and somatic levels. Unfortunately, because tumor tissue is not available from the families linked to this region, we can not determine whether the non-linked allele is more likely to undergo somatic deletion, as is observed in multiple inherited cancer syndromes. The observation that most tumors with 8p deletions have deleted both of the implicated regions, and that at least some prostate cancer families are linked to both regions, suggests that multiple genes in these intervals may need to be affected before prostate carcinogenesis can proceed effectively. In any event, the results of this integrated analysis improve the confidence that these two regions most likely contain prostate cancer genes as well as provide more detailed positional information regarding the genomic regions harboring these genes.

In summary, we have combined genetic linkage information with somatic deletion mapping in an attempt to refine the localization of prostate cancer genes on the short arm of chromosome 8. The genomic intervals narrowed by this combined approach provide novel positional information useful for the eventual identification of specific genes important in prostate carcinogenesis.

	Acknowledgments

 
Grant support: National Cancer Institute grants CA106523 and CA95052 (J. Xu) and Department of Defense grant PC051264 (J. Xu).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank all the study subjects who participated in this study.

	Footnotes

 
Note: B-L. Chang and W. Liu contributed equally to this study.

³ http://genome.ucsc.edu/

Received 12/12/05. Revised 1/24/07. Accepted 2/ 9/07.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Dong JT. Chromosomal deletions and tumor suppressor genes in prostate cancer [review]. Cancer Metastasis Rev 2001;20:173–93.[CrossRef][Medline]
2. Sun J, Liu W, Adams TS, et al. DNA copy number alterations in prostate cancers: a combibed analysis of published CGH studies. Prostate 2007 Mar 6; [Epub ahead of print].
3. Liu W, Chang B, Sauvageot J, et al. Comprehensive assessment of DNA copy number alterations in human prostate cancers using Affymetrix 100K SNP mapping array. Genes Chromosomes Cancer 2006;45:1018–32.[CrossRef][Medline]
4. Xu J, Zheng SL, Hawkins GA, et al. Linkage and association studies of prostate cancer susceptibility: evidence for linkage at 8p22-23. Am J Hum Genet 2001;69:341–50.[CrossRef][Medline]
5. Xu J, Dimitrov L, Chang BL, et al. A combined genomewide linkage scan of 1,233 families for prostate cancer-susceptibility genes conducted by the international consortium for prostate cancer genetics. Am J Hum Genet 2005;77:219–29.[CrossRef][Medline]
6. Knudson AG, Jr. Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci U S A 1971;68:820–3.[Abstract/Free Full Text]
7. Bova GS, MacGrogan D, Levy A, Pin SS, Bookstein R, Isaacs WB. Physical mapping of chromosome 8p22 markers and their homozygous deletion in a metastatic prostate cancer. Genomics 1996;35:46–54.[CrossRef][Medline]
8. Nannya Y, Sanada M, Nakazaki K, et al. A robust algorithm for copy number detection using high-density oligonucleotide single nucleotide polymorphism genotyping arrays. Cancer Res 2005;65:6071–9.[Abstract/Free Full Text]
9. Lin M, Wei LJ, Sellers WR, Lieberfarb M, Wong WH, Li C. dChipSNP: significance curve and clustering of SNP-array-based loss-of-heterozygosity data. Bioinformatics 2004;20:1233–40.[Abstract/Free Full Text]
10. Kong X, Matise TC. MAP-O-MAT: Internet-based linkage mapping. Bioinformatics 2005;21:557–9.[Abstract/Free Full Text]
11. Kruglyak L, Daly MJ, Reeve-Daly MP, Lander ES. Parametric and nonparametric linkage analysis: a unified multipoint approach. Am J Hum Genet 1996;58:1347–63.[Medline]
12. Kong A, Cox NJ. Allele-sharing models. LOD scores and accurate linkage tests. Am J Hum Genet 1997;61:1179–88.[CrossRef][Medline]
13. Smith JR, Freiji D, Carpten JD, et al. A genome wide search reveals a major susceptibility locus for prostate cancer on chromosome 1. Science 1996;274:1371–4.[Abstract/Free Full Text]
14. Whittemore A, Halpern J. A class of tests for linkage using affected pedigree members. Biometrics 1994;50:118–27.[CrossRef][Medline]
15. Zheng SL, Ju JH, Chang BL, et al. Germ-line mutation of NKX3.1 cosegregates with hereditary prostate cancer and alters the homeodomain structure and function. Cancer Res 2006;66:69–77.[Abstract/Free Full Text]
16. Macoska JA, Trybus TM, Benson PD, et al. Evidence for three tumor suppressor gene loci on chromosome 8p in human prostate cancer. Cancer Res 1995;15:5390–5.
17. Kinzler KW, Nilbert MC, Vogelstein B, et al. Identification of a gene located at chromosome 5q21 that is mutated in colorectal cancers. Science 1991;251:1366–70.[Abstract/Free Full Text]

This article has been cited by other articles:

				A. Ramalingam, J. B. Duhadaway, E. Sutanto-Ward, Y. Wang, J. Dinchuk, M. Huang, P. S. Donover, J. Boulden, L. M. McNally, A. P. Soler, et al. Bin3 Deletion Causes Cataracts and Increased Susceptibility to Lymphoma during Aging Cancer Res., March 15, 2008; 68(6): 1683 - 1690. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Chang, B.-l. Articles by Xu, J. Search for Related Content PubMed PubMed Citation Articles by Chang, B.-l. Articles by Xu, J.