The Prostate 33:9–12 (1997) Genetic Variation of 3b-Hydroxysteroid Dehydrogenase Type II in Three Racial/Ethnic Groups: Implications for Prostate Cancer Risk Sunita A. Devgan,1 Brian E. Henderson,2 Mimi C. Yu,2 Chen-Yang Shi,3 Malcolm C. Pike,2 Ronald K. Ross,2 and Juergen K.V. Reichardt1,4* 1 Department of Biochemistry and Molecular Biology, University of Southern California School of Medicine, USC/Norris Comprehensive Cancer Center, Los Angeles, California 2 Department of Preventive Medicine, University of Southern California School of Medicine, USC/Norris Comprehensive Cancer Center, Los Angeles, California 3 Department of Community, Occupational and Family Medicine, National University of Singapore, Singapore 4 Institute for Genetic Medicine, University of Southern California School of Medicine, USC/Norris Comprehensive Cancer Center, Los Angeles, California BACKGROUND. Elevated prostatic dihydrotestosterone (DHT) has been suggested to increase the risk of prostate cancer. The HSD3B2 gene encodes the type II 3b-hydroxysteroid dehydrogenase: one of two enzymes that initiate the inactivation of DHT. Thus, the HSD3B2 gene is a candidate gene for predisposition to prostate cancer. METHODS. We have determined the distribution of a complex dinucleotide repeat in the HSD3B2 gene in high-risk African-Americans, intermediate-risk Euro-Americans, and lowrisk Asians. Genomic DNA from 312 individuals was amplified by polymerase chain reaction (PCR) and analyzed by electrophoresis on denaturing polyacrylamide gels. RESULTS. We have found that certain alleles are either unique to or much more common in either African-Americans, Asians, or Euro-Americans. Our data also substantially expand the number of alleles reported for the complex dinucleotide repeat polymorphism in the HSD3B2 gene. CONCLUSIONS. Our report demonstrates substantial genetic variation in the HSD3B2 gene. We hypothesize that allelic variants of the HSD3B2 gene may play a role in predisposition to prostate cancer, and in explaining the substantial racial/ethnic variation in risk. Prostate 33:9–12, 1997. © 1997 Wiley-Liss, Inc. KEY WORDS: androgen metabolism; PCR; complex dinucleotide repeat polymorphism; racial/ethnic variation of prostate cancer risk INTRODUCTION Prostate cancer in the United States will be diagnosed in some 317,000 men in 1996 and over 41,000 will die of this disease . This cancer is characterized by marked variations in incidence among U.S. population groups: African-American men have a 70% increase in risk when compared to Euro-American (i.e., non-Hispanic White) males, while Asian-American men (of either Chinese or Japanese ancestry) have an approximately 60% decrease in risk when compared © 1997 Wiley-Liss, Inc. to males of European ancestry . We have previously suggested that elevated testosterone (T) levels and, in particular, intraprostatic dihydrotestosterone (DHT) Contract Grant sponsor: National Institutes of Health; Contract Grant numbers: CA17054, CA53890, CA54281, and CA68581. *Correspondence to: Juergen K.V. Reichardt, Institute for Genetic Medicine, USC/Norris Comprehensive Cancer Center, HMR413, 2011 Zonal Ave., Los Angeles, CA 90033. E-mail: email@example.com Received 29 April 1996; Accepted 17 June 1996 10 Devgan et al. levels may be responsible for some of this variation in risk . Molecular studies thus far have revealed few mutations responsible for predisposition to prostate cancer and early tumor progression [4,5]. T is converted to the more active intracellular metabolite, DHT, by steroid 5a-reductase . DHT then binds to the androgen receptor (AR) and the DHT-AR complex transactivates a number of genes with ARresponsive elements . This results in a variety of physiologic effects including cell division. DHT is inactivated by either 3a- or 3b-hydroxysteroid dehydrogenase, further modified and ultimately excreted . Thus, the 3-hydroxysteroid dehydrogenases are involved in the regulation of DHT levels which modulate prostatic cell division. Two isozyme forms of 3b-hydroxysteroid dehydrogenase have been reported in humans. The type I enzyme is encoded by the HSD3B1 gene and is expressed mostly in the breast, placenta, and skin . The type II enzyme is encoded by the HSD3B2 gene and is primarily expressed in adrenals, testis, and ovary . The human HSD3B1 and HSD3B2 genes have been shown to be closely linked and mapped to a single band on the short arm of chromosome 1 [1p13; 7,8]. Both genes have been cloned and shown to be very similar spanning about 8 kb of genomic DNA in four exons each . A number of mutations in the HSD3B2 gene have been found to cause a rare human disorder, congenital adrenal hyperplasia . These (and other) mutations, however, do not appear to be involved in prostatic diseases in adults. Verreault et al.  reported a complex (TG)n (TA)n (CA) n dinucleotide repeat polymorphism in the HSD3B2 gene which can be detected by polymerase chain reaction (PCR). The variable region, located within the third intron of the gene and reported to consist of eight alleles, was typed in 52 Caucasian individuals . We have initiated systematic molecular investigations into the role of androgen-metabolic genes in the racial/ethnic variation of prostate cancer risk. We recently reported genetic variation of the SRD5A2 gene encoding steroid 5a-reductase which synthesizes DHT from T [6,11]. The high-risk African-American sample contained alleles that were absent from intermediate-risk Euro-Americans and low-risk Asians . Here, we examined the hypothesis that HSD3B2 is also a candidate gene for prostate cancer predisposition by genotyping males from three population groups at very different risks for prostate cancer: highrisk African-American, intermediate-risk EuroAmerican, and low-risk Asian men. We report significant polymorphism in the HSD3B2 gene with some alleles restricted to high-risk African-Americans. Several other alleles are characterized by significant dif- ferences in their racial distribution. These data are consistent with our hypothesis that androgenmetabolic genes such as HSD3B2 and SRD5A2 play a role in predisposition to prostate cancer and may explain, at least in part, the substantial racial/ethnic variation in risk. MATERIALS AND METHODS HSD3B2 Genotyping Our study subjects consisted of multiethnic, male residents of Los Angeles County (of African, European, Chinese, and Japanese ancestry) and Chinese male residents of Singapore. The Los Angeles component has been described previously [11,12] and Chinese subjects in Singapore are part of an ongoing cohort study similar in both design and goals to the multiethnic cohort study in Los Angeles [11,12]. Genomic DNA of study subjects of AfricanAmerican, Asian, and Euro-American men  was PCR amplified using the primer pair of Verreault et al. . Syntheses of primers were performed in 30 nM scales on a Beckman Oligo1000 (Beckman Instruments, Fullerton, CA). One primer was radiolabeled by kinasing with g[32P]ATP (6,000 Ci/mmol; NEN, Boston, MA) and T4 polynucleotide kinase (Promega, Madison, WI; 13). PCR products were obtained in a TwinBlock thermal cycler (Ericomp, San Diego, CA) by repeating the following cycle 30 times: denaturation at 92°C for 2 min, annealing for 1 min at 62°C, and extension at 72°C for 2 min. PCR reactions were denatured and then fractionated on 4.5% denaturing polyacrylamide gels (‘‘sequencing gels’’; 13) in parallel with DNA sequencing reactions obtained with Sequenase 2.0 (USB, Cleveland, OH). Gels were dried and exposed overnight to Kodak BIOMAX autoradiography film (Rochester, NY). Statistical Analysis The ‘‘exact’’ method of Mehta and Patel  was used to compare the overall prevalence of alleles among the three racial/ethnic groups and between each pair of groups. All tests were statistically highly significant (two sided, P < 0.000001). Since the overall tests were highly significant, we then tested individual allele comparison using Fisher’s exact test (comparing each allele against all others combined) between each set of two racial/ethnic groups. Significant differences in these results are quoted as one-sided P values in the Results section . HSD3B2 Polymorphism and Prostate Cancer Risk TABLE I. Distribution of HSD3B2 Alleles in Three Populations* Allele (bp) 271–273 275 281 283–287 289 291 293–295 302–334 338 340 342 344–372 AfricanAmerican (n = 256) Euro-American (n = 248) Asian (n = 120) 0 0.055 0.016 0.004 0.336 0.254 0.051 0.066 0.039 0.137 0.023 0.020 0.008 0 0.020 0.020 0.516 0.105 0.012 0.024 0.032 0.222 0.028 0.016 0.033 0 0.175 0.008 0.367 0.150 0.033 0.125 0.008 0.075 0.017 0.008 *Allele frequency figures were rounded as necessary and, therefore, may not add up to 1.0 (or 100% in the text). RESULTS We have further investigated the published HSD3B2 genotyping system which involves a complex (TG)n (TA)n (CA)n dinucleotide repeat and was reported to consist of eight alleles . Thus far, we have identified a total of 25 alleles. Table I presents allele frequencies for the HSD3B2 gene by race in 312 control subjects of African-American, Asian, and European heritage (624 alleles). Some alleles were combined into ‘‘families’’ to condense the data for easier analysis (cf. Table I). All alleles are identified by their size in base pairs (bp). The 289 bp allele is the most common allele in all three population groups examined (Table I) in agreement with the previously published data on Caucasians alone . This marker is significantly more common among Euro-Americans (allele frequency = 51.6%) than among African-Americans (allele frequency = 33.6%; P = 0.00003) or among Asians (allele frequency = 36.7%; P = 0.005). Several other allelic markers show statistically significant differences in allele distribution in the three populations examined. The second most common allele in all populations is 291 bp in size but its frequency differs dramatically between African- (allele frequency = 25.4%) and EuroAmericans (allele frequency = 10.5%; P = 0.000009; Table I). The same allele is intermediate in frequency among Asians (15.0%; P = 0.015 vs. AfricanAmericans). The 275 bp allele has thus far been found only in African-American men with a frequency of 5.5% (P = 0.00064 vs. Euro-Americans; P = 0.0041 vs. Asians; Table I). The 293–295 bp family of alleles appears to be more common in African-Americans (allele 11 frequency = 5.1%; Table I) than it is among EuroAmericans (allele frequency = 1.2%; P = 0.011; Table I). Furthermore, alleles ranging from 302 to 334 bp in size are most common among people of Asian ancestry (allele frequency = 12.5%). This allele frequency is much lower among Euro-Americans (allele frequency = 2.4%; P = 0.0002; Table I). These alleles are, however, intermediate among African-American men when compared to Asians (allele frequency = 6.6%; P = 0.047). The 281 bp allele is much more common among Asians (allele frequency = 17.5%) than among African(allele frequency = 1.6%) or Euro-Americans (allele frequency = 2%; Table I). Finally, the 340 bp allele is more prominent among people of European ancestry (allele frequency = 22.2%; Table I) than among those of African (allele frequency = 13.7%; P = 0.009) or Asian descent (allele frequency = 7.5%; P = 0.0002). DISCUSSION We report here extensive genotypic data on three populations at markedly different risk for prostate cancer: high-risk African-Americans, intermediaterisk Euro-Americans, and low-risk Asians. We PCR typed the previously reported complex (TG)n (TA)n (CA)n dinucleotide repeat polymorphism in the human HSD3B2 gene . Our data indicate 1) that this gene is more polymorphic than previously reported, and 2) that certain polymorphisms are distributed differently among the three racial groups examined (Table I). Thus far, we have identified 25 different alleles which expand substantially the original report of 8 alleles . For ease of analysis, we have grouped these 25 alleles into 12 families (Table I). In agreement with the data on Caucasians reported earlier , the 289 bp allele is the most prominent allele in all three populations examined (Table I). We also found one allele that is unique to the African-American men sampled (275 bp; Table I). Furthermore, we identified two families of alleles—291 and 293–295 bp—that are significantly more common among AfricanAmericans than they are among either EuroAmericans or Asians (Table I). The 340 bp allele is significantly more common among people of European ancestry than among the two other groups we genotyped in this study. Finally, we have identified two families of alleles, 281 and 302–334 bp, that are more common in Asian men (Table I). We have previously proposed that the difference in the racial/ethnic incidence of prostate cancer  may be determined in part by different levels of T and more specifically its intraprostatic metabolite DHT . The present study provides additional molecular genetic support for our hypothesis. The statistically significant finding of HSD3B2 repeat alleles that 12 Devgan et al. are present only or more prominently so in high-risk African-Americans or low-risk Asians (Table I) supports our initial study (3). We hypothesize here based on our polymorphic data (Table I) that certain 3bhydroxysteroid dehydrogenase enzyme variants encoded by mutant HSD3B2 genes may result in altered enzyme activity. Such mutations could result in slower than normal degradation of DHT through decreased enzyme activity. Intraprostatic DHT levels would rise accordingly, resulting in an increased risk for developing prostate cancer. Conversely, other mutations that increase HSD3B2 activity might reduce the risk for developing prostate cancer by increasing the rate of DHT degradation. These hypothetical alleles that may influence prostate cancer risk could be marked by particular repeat alleles. The complex dinucleotide repeat polymorphism we typed in this study is intronic  and, therefore, unlikely to be of functional significance in this context. However, this polymorphic repeat is a useful marker for molecular investigations into prostate cancer risk. The present report complements our previous work on another androgen metabolic gene, SRD5A2, encoding steroid 5a-reductase type II . In that communication we characterized substantial racial/ethnic variation at this locus that correlated with risk. These two investigations into genetic variability of two candidate genes support a role for genes encoding androgen-metabolic enzymes in explaining the racial/ethnic differences in prostate cancer risk. ACKNOWLEDGMENTS This work was supported by NIH grants CA17054, CA53890, CA54281, and CA68581. We thank Kristine Monroe and Peggy Wan for help with data recording and statistical computation. REFERENCES 1. American Cancer Society: ‘‘Cancer Facts and Figures.’’ New York: American Cancer Society, 1994. 2. Bernstein L, Ross RK: ‘‘Cancer in Los Angeles County: A Portrait of Incidence and Mortality.’’ Los Angeles: University of Southern California, 1991, pp 56–57. 3. Ross RK, Bernstein L, Lobo RA, Shimizu H, Stanczyck FZ, Pike MC, Henderson BE: 5-Alpha-reductase activity and risk of prostate cancer among Japanese and US white and black males. Lancet 339:887–889, 1992. 4. Nature Genetics Editorial: The prognosis for prostate cancer. Nature Genet 6:215–216, 1994. 5. Cannon-Albright L, Eeles R: Progress in prostate cancer. Nature Genet 9:336–338, 1995. 6. Coffey DS: The molecular biology of the prostate. In Lepor H, Lawson RK (eds): ‘‘Prostate Diseases.’’ Philadelphia: W.B. Saunders, pp 28–56. 7. Labrie F, Simard J, Luu-The V, Belanger A, Pelletier G: Structure, function and tissue-specific gene expression of 3bhydroxysteroid dehydrogenase/5-ene-4-ene isomerase enzymes in classical and peripheral intracrine and steroidogenic tissue. J Steroid Biochem Mol Biol 43:805–826, 1992. 8. Berube D, Luu-The V, Lachance Y, Gagne R, Labrie F: Assignment of the human 3b-hydroxysteroid dehydrogenase gene to the p13 band of chromosome 1. Cytogenet Cell Genet 52:199– 200, 1989. 9. Rheaume E, Simard J, Morel Y, Mebarki F, Zachman M, Forest MG, New MI, Labrie F: Congenital adrenal hyperplasia due to point mutations in the type II of 3b-hydroxysteroid dehydrogenase gene. Nature Genet 1:239–245, 1992. 10. Verreault H, Dufort I, Simard J, Labrie F, Luu-The V: Dinucleotide repeat polymorphisms in the HSD3B2 gene. Hum Mol Genet 3:384, 1994. 11. Reichardt JKV, Makridakis N, Henderson BE, Yu MC, Pike MC, Ross RK: Genetic variability of the human SRD5A2 gene: Implications for prostate cancer risk. Cancer Res 55:3973–3975, 1995. 12. Monroe KR, Yu MC, Kolonel LN, Coetzee GA, Wilkens LR, Ross RK, Henderson BE: Evidence of an X-linked or recessive genetic component to prostate cancer risk. Nature Med 1:827– 829, 1995. 13. Sambrook J, Fritsch EF, Maniatis T: ‘‘Molecular Cloning,’’ 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory. 14. Mehta CR, Patel NR: A network algorithm for performing Fischer’s exact test in RxC contingency tables. J Am Stat Assoc 78: 427–434, 1983. 15. EPILOG, Epicenter Software, Pasadena, CA.