The Prostate 41:253–257 (1999) Androgen Receptor CAG Repeat Length Polymorphism in Benign Prostatic Hyperplasia (BPH): Correlation With Adenoma Growth Kenji Mitsumori, Akito Terai,* Hiroya Oka, Takehiko Segawa, Keiji Ogura, Osamu Yoshida, and Osamu Ogawa Department of Urology, Faculty of Medicine, Kyoto University, Kyoto, Japan BACKGROUND. The androgen receptor (AR) gene has a polymorphic CAG microsatellite encoding variable-length glutamine repeats in the AR protein. The purpose of this study was to evaluate the association between the growth of benign prostatic hyperplasia (BPH) and the AR gene CAG repeat length. METHODS. We determined CAG repeat lengths in 176 BPH patients who underwent simple prostatectomy and in 41 control subjects without benign prostatic enlargement (non-BPE group). RESULTS. A statistically significant (P < 0.02) trend for large adenoma size with short CAG repeat length was found among the adenoma quartiles. CAG repeat length in the fourth quartile (large adenoma, 21.5 ± 2.7) was significantly shorter than in the first quartile (small adenoma, 23.3 ± 2.1, P < 0.02). It tended to be shorter than in the non-BPE group (23.1 ± 2.4), but CAG repeat lengths in the entire BPH (22.4 ± 2.5) and non-BPE groups did not significantly differ. The relative risk of large BPH (the fourth quartile) was 2.75 (95% confidence interval, 1.05–7.24; P < 0.05) on comparing CAG repeats of ⱕ22–ⱖ23. CONCLUSIONS. Shorter CAG alleles may be a genetic factor that promotes the growth of BPH. Prostate 41:253–257, 1999. © 1999 Wiley-Liss, Inc. KEY WORDS: CAG microsatellite; androgen receptor; benign prostatic hyperplasia; adenoma size; relative risk INTRODUCTION Benign prostatic hyperplasia (BPH) is a very common pathology in 50% of 50-year-old men and in 90% of 90-year-old men . Although the pathogenesis and natural course of BPH are still unclear, androgen is widely accepted as necessary for differentiation and growth of the prostatic epithelium and as having a critical role in pathological conditions of the prostate, including prostate cancer and BPH . Responses to androgens are mediated through the androgen receptor (AR), a member of the steroid receptor superfamily of ligand-dependent transcription factors. The androgen receptor (AR) gene has a polymorphic CAG microsatellite in exon one that codes for variable-length glutamine repeats in the N-terminal domain of the AR protein . Previously reported in vitro studies showed that the length of the polygluta© 1999 Wiley-Liss, Inc. mine chain is correlated inversely with the transcriptional activity of the AR [4–6]. Consistent with these observations, expansion of AR gene CAG repeat length has been found in patients with partial androgen insensitivity, such as in Kennedy’s disease . CAG repeat lengths have been intensively studied in patients with prostate cancer [8–14]. The length of CAG repeats is known to be associated with susceptibility to or age of onset of prostate cancer. Despite the central role of androgens in the development of The authors certify that all authors of this manuscript have no financial arrangement with a company whose product figures prominently in the submitted manuscript or with a company making a competing product. *Correspondence to: A. Terai, M.D., Department of Urology, Faculty of Medicine, Kyoto University, Sakyo-ku, Kyoto 606-8507, Japan. E-mail: firstname.lastname@example.org Received 25 January 1999; Accepted 11 June 1999 254 Mitsumori et al. BPH, only a few studies have analyzed CAG repeat lengths in BPH patients. Recently, Giovannucci et al.  found an inverse correlation between the AR gene CAG repeat length and the prevalence of BPH in men enrolled from the Health Professionals Follow-Up Study. Men were defined as having BPH if they had responded in the affirmative to a questionnaire as “ever [having been] diagnosed with an enlarged prostate by digital rectal examination (DRE)” and/or “ever [having] had surgery for BPH.” Additional parameters, such as prostate volume and histopathology, had not been analyzed in their study. We hypothesized that variation in the transcriptional activity of the AR gene that is mediated by different CAG repeat lengths may be a genetic factor that affects the growth of adenomas. We have examined AR gene CAG repeat lengths in BPH patients who underwent open prostatectomy and for whom adenoma weight had been measured. MATERIALS AND METHODS Study Population Between 1980–1996, 530 Japanese patients with BPH underwent surgery at Kyoto University (204 open prostatectomies, 326 transurethral resections of the prostate). Weights of the adenomas had been measured for 176 of the 204 patients who underwent open prostatectomy, and the histopathological diagnosis of benign prostatic hyperplasia was confirmed using step sections of whole specimens. The median weight of the resected adenomas was 40.0 g (25th and 75th percentiles; 25.0 and 59.3 g, respectively). All 176 specimens were subjected to CAG repeat length analysis. Men without benign prostatic enlargement (BPE) were matched against the groups of prostatectomized patients as the control group. This non-BPE group consisted of 41 men who met all the following criteria: age 65 years or older, normal serum PSA level, benign DRE finding, and normal appearance with <20 ml prostate volume on transrectal ultrasonography (TRUS). Eighteen women selected at random comprised an additional control group. DNA Extraction Because somatic mutations in the lengths of CAG repeats in the AR gene are rare, we used paraffinembedded tissues of resected adenomas to extract the target DNA with the standard protocols . Briefly, 15 × 15 mm tissue sections sliced to 10-m thickness were treated twice with xylene to remove paraffin, and then washed twice with 99.5% ethanol. After being dried in a vacuum, the samples were incubated at 50°C for 48 hr in 100 l of tissue lysis buffer (50 mM Tris, pH 8.5, 1 mM EDTA, 0.5% Tween-20) containing 200 mg/ml of Proteinase K. To inactivate proteinase, the samples were boiled for 10 min and then centrifuged. Five microliters of each supernatant were used as the template in the polymerase chain reaction (PCR). Genomic DNAs of the non-BPE subjects and women were extracted from peripheral blood lymphocytes. Analysis of CAG Repeat Length in the AR The CAG region of the AR gene was amplified by PCR. The primer pairs used, AR-F (5⬘-TGC GCG AAG TGA TCC AGA AC-3⬘) and AR-R (5⬘-CTT GGG GAG AAC CAT CCT CA-3⬘), produced fragments of 195 bp for 17 repeats and 234 bp for 29 repeats. Each 30 l of reaction mixture contained 50 pmol of each primer, 200 nM dNTPs, 2.0 mM MgCl2, 1 × PCR Buffer II (Perkin Elmer, Norwalk, CT), 0.75 U of AmpliTaq Gold (Perkin Elmer), and 5 l of the target DNA. Amplification conditions were the initial denaturation and activation of the enzyme at 95°C for 8 min followed by 38 cycles of 94°C for 1 min, 60°C for 1 min, 72°C for 1 min, and final extension at 72°C for 10 min. After amplification, the PCR products were separated by electrophoresis at 900 v for 5 hr on 8% polyacrylamide gels on nonfluorescence glass plates. The gels were stained with SYBR Green I (Molecular Probes, Eugene, OR), and PCR bands were made visible with FluoroImager SI (Amersham Pharmacia Biotech, Buckinghamshire, UK). Repeat sizes were determined by a comparison with a series of previously determined (by direct sequencing of PCR products) CAG size standards. All samples were examined at least twice. At the second electrophoresis, we applied the samples in ranked order based on size and confirmed the original determinations. Statistical Methods Simple regression analysis was used to evaluate the relationship between CAG repeat length and adenoma weight. Differences in CAG repeat length distribution were examined by the Mann-Whitney U test (between two groups) or the Kruskal-Wallis and Scheffé F tests (among quartiles and five groups). The relative risk and Miettinen 95% confidence interval were analyzed using 2 × 2 tables and the 2 test. All statistical analyses were done with StatView-J 4.5 software (Abacus Concepts, Berkeley, CA). Androgen Receptor CAG Microsatellite in BPH 255 Fig. 1. A: Association between number of AR gene CAG repeats and adenoma weight. B: Association between patient age and adenoma weight. RESULTS Excluding 18 specimens from which no amplified products were obtained after repeated PCR, CAG repeat length could be determined for 158 alleles of the BPH patients, 41 alleles of the non-BPE group, and 36 alleles of the female control group. The associations of adenoma weight with patient age and CAG repeat length in the BPH patients were assessed by regression analysis (Fig. 1). Adenoma weight was inversely related to CAG repeat length, but the coefficient of determination was low (R2 = 0.075). In contrast, adenoma weight did not have a statistically significant association with patient age. All BPH specimens were divided into quartiles by adenoma weight (first quartile, ⱕ29.0 g; second, >29.0 g and ⱕ40.0 g; third, >40.0 g and ⱕ59.3 g; and fourth, >59.3 g) to further analyze the association between the weight of adenomas and CAG repeat length. The trend for larger adenoma size with shorter CAG repeat length was statistically significant (P < 0.02), and CAG repeat length in the fourth quartile (21.5 ± 2.7) was significantly shorter than that in the first quartile (23.3 ± 2.1, P < 0.02). When the non-BPE group was considered, CAG repeat length did not differ significantly between the BPH (22.4 ± 2.5) and non-BPE groups (Table I). CAG repeat length in the fourth quartile, however, was still significantly shorter than that in the first quartile (P < 0.05) and tended to be shorter than in the non-BPE group (23.1 ± 2.4, P = 0.0584). Furthermore, CAG repeat length in the 199 men analyzed (22.6 ± 2.5) agreed with that in the randomly selected female control group (23.3 ± 2.6). The relative risk of BPH was calculated, using the median repeat length in the non-BPE group as the arbitrary cutoff point (ⱕ22 and ⱖ23 CAG repeats, Table II). There was a 1.97-fold nonsignificant excess risk for BPH associated with ⱕ22 CAG repeats. These repeats, however, showed the highest relative risk for large BPH (the fourth quartile), which reached statistical significance (relative risk, 2.75; 95% confidence interval, 1.05–7.24; P < 0.05). DISCUSSION BPH seems to be a pathobiological process that, at least in part, depends on androgen. Although androgen does not cause BPH, its development requires the presence of testicular androgens during prostate development, puberty, and aging. AR levels in the prostate remain high throughout aging, unlike in other androgen-dependent organs . The prostate, therefore, retains the capacity for further growth, despite the age-related decreasing levels of androgen in peripheral circulation. Because AR gene CAG repeat length is inversely correlated with the transcriptional activity of the receptor, it also may affect the development of BPH. Recently, Giovannucci et al.  found an inverse correlation between AR gene CAG repeat length and the prevalence of BPH in men enrolled 256 Mitsumori et al. TABLE I. Distribution of CAG Repeat Length in Adenoma Quartiles of BPH and in Control Groups* BPH No. of cases analyzed Median adenoma weight (g) (minimum– maximum) Mean age No. of CAG repeats 15–16 17–18 19–20 21–22 23–24 25–26 27–28 29–30 Noninformative Minimum–maximum Mean All BPH First quartile Second quartile Third quartile Fourth quartile Non-BPE (control) Womena 176 44 44 44 44 41 18 40.0 (4.2–155.0) 17.0 (4.2–25.0) 33.5 (26.0–40.0) 49.5 (40.3–59.0) 75.0 (60.0–155.0) N.A. N.A. 71.3 ± 6.8 (55–87) 1 69.9 ± 6.5 (57–85) 0 71.2 ± 7.2 (55–87) 0 71.5 ± 6.2 (59–83) 0 72.1 ± 7.1 (58–85) 1 77.1 ± 7.8 (65–93) 0 N.A. 6 28 49 43 20 9 2 18 15–29 22.4 ± 2.5 0 4 12 14 9 2 1 2 20–29 23.3 ± 2.1 0 6 16 8 3 4 1 6 19–29 22.6 ± 2.6 2 4 13 9 5 1 0 10 17–27 22.4 ± 2.3 4 14 8 12 3 2 0 0 15–28 21.5 ± 2.7 0 5 10 16 7 2 1 0 19–29 23.1 ± 2.4 2 2 10 11 6 5 0 0 17–28 23.3 ± 2.6 0 *N.A., not applied. Results of statistical analysis: 1) Among the 4 quartiles of BPH: P < 0.02 by the Kruskal-Wallis test. First vs. fourth quartiles: P < 0.02 by the Scheffé F test. 2) All BPH vs. non-BPE: not significant by Mann-Whitney U test. 3) Among the 5 groups (non-BPE and the 4 quartiles of BPH): P < 0.02 by Kruskal-Wallis test. First vs. fourth quartiles: P < 0.05 by Scheffé F test. Fourth quartile vs. non-BPE: P = 0.0584 by the Scheffé F test. a Thirty-six alleles were analyzed from 18 randomly selected women. TABLE II. Relative Risk of BPH Based on CAG Repeat Length† Case Control No. of CAG repeats No. of cases No. of controls RR 95% CI All BPH Non-BPE Non-BPE Second quartile Non-BPE Third quartile Non-BPE Fourth quartile Non-BPE 84 74 16 26 22 16 19 15 27 17 15 26 15 26 15 26 15 26 15 26 1.97 1.00 1.07 1.00 2.38 1.00 2.20 1.00 2.75* 1.00 0.91–4.27 First quartile ⱕ22 ⱖ23 ⱕ22 ⱖ23 ⱕ22 ⱖ23 ⱕ22 ⱖ23 ⱕ22 ⱖ23 0.24–4.79 0.85–6.64 0.75–6.46 1.05–7.24 † RR, relative risk; CI, confidence interval. *P < 0.05. from the Health Professionals Follow-Up Study. Men with ⱕ19 CAG repeat lengths had odds ratio for BPH of 1.92 relative to those with ⱖ25 repeats. Because the definition of BPH was based on questionnaire responses of a history of diagnosis of an enlarged prostate by DRE and/or a history of surgery for BPH, the additional parameters of prostate volume and histopathology were not analyzed in their study. In the study reported here, we found an inverse correlation between adenoma weight and AR gene CAG repeat length (Fig. 1), although the coefficient of determination in the regression line was low (R2 = 0.075). We also found a statistically significant trend among the adenoma quartiles for larger adenoma sizes with shorter CAG repeat lengths (Table I). These findings support our hypothesis that short CAG repeats may promote the growth of prostatic adenoma, possibly through modification of the transcriptional activity of the AR. Furthermore, this is the first study to demonstrate unambiguously that there is an inverse correlation between prostatic adenoma size and AR gene CAG repeat length. Androgen Receptor CAG Microsatellite in BPH With regard to BPH, defining the control group may be difficult because it is unusual to find a histologically normal prostate in men over 70 years of age. Berry et al.  reported that the average weight of an adult human prostate with normal histology is 20 ± 6 g and remains constant with age. In a healthcare examination, Kitagawa et al.  found that prostatic size as a function of age is separable into two groups; increasing and no-change (20.0 ± 5.0 g). On the basis of these reports, we defined the control subjects in our study as men ⱖ65 years old, of normal echogenicity with <20 ml prostate volume on TRUS, benign DRE findings, and normal serum PSA levels. Although we did not examine the pathology of the prostate in the control group, these criteria would exclude nearly all prostatic diseases of clinical significance in aged men. In this study, CAG repeat length did not differ significantly between the BPH and non-BPE groups, but the fourth quartile (large adenoma) group tended to have shorter CAG repeats than the non-BPE group (Table I). On the basis of these findings, we conclude that short CAG repeats may promote the growth of prostatic adenoma through modification of the transcriptional activity of the AR. Their impact, however, may become overt only in patients with large BPH, probably because of great interindividual variability in other risk factors such as androgens. A familial form of BPH was reported in several recent studies, indicative of the presence of genes contributing to the pathogenesis of the disease [17,18]. Segregation analysis showed that a Mendelian dominant model of transmission is consistent with the observed familial clustering of BPH . The possibility of maternal transmission also agrees with the finding  that brothers carried a greater age-specific relative risk of BPH requiring surgery (6.1) than fathers (3.5). Our findings support the hypothesis that the human AR gene, which has been mapped to chromosome Xq11–q12, may be one of the candidate genes responsible for this disorder. Further study is needed to establish molecular markers by which to identify men at high risk of BPH because, in combination with traditional epidemiologic studies, the stratification of genetic predisposition may make it possible to clarify related environmental risk factors. CONCLUSIONS We investigated the association between prostatic adenoma weight and AR gene CAG repeat length in patients who underwent open prostatectomy. A statistically significant trend for large adenomas with short CAG repeats was found among the adenoma quartiles. Our findings suggest that short CAG alleles may constitute a genetic factor which has a promoting influence on the growth of benign prostatic hyperplasia. 257 REFERENCES 1. Berry SJ, Coffey DS, Walsh PC, Ewing LL. The development of human benign prostatic hyperplasia with age. J Urol 1984;132: 474–479. 2. McConnell JD. Prostatic growth: new insights into hormonal regulation. Br J Urol 1995;76(Suppl 1):5–10. 3. Edwards A, Hammond HA, Jin L, Caskey CT, Chakraborty R. Genetic variation at five trimeric and tetrameric tandem repeat loci in four human population groups. Genomics 1992;12:241– 253. 4. Chamberlain NL, Driver ED, Miesfeld RL. The length and location of CAG trinucleotide repeats in the androgen receptor Nterminal domain affect transactivation function. Nucleic Acids Res 1994;22:3181–3186. 5. Kazemi-Esfarjani P, Trifiro MA, Pinsky L. Evidence for a repressive function of the long polyglutamine tract in the human androgen receptor: possible pathogenetic relevance for the (CAG)n-expanded neuronopathies. Hum Mol Genet 1995;4:523– 527. 6. Choong CS, Kemppainen JA, Zhou Z-X, Wilson EM. Reduced androgen receptor gene expression with first exon CAG repeat expansion. Mol Endocrinol 1996;10:1527–1535. 7. Biancalana V, Serville F, Pommier J, Julien J, Hanauer A, Mandel JL. Moderate instability of the trinucleotide repeat in spino bulbar muscular atrophy. Hum Mol Genet 1992;1:255–258. 8. Coetzee GA, Ross RK. Re: prostate cancer and the androgen receptor. J Natl Cancer Inst 1994;86:872–873. 9. Giovannucci E, Stampfer MJ, Krithivas K, Brown M, Dahl D, Brufsky A, Talcott J, Hennekens CH, Kantoff PW. The CAG repeat within the androgen receptor gene and its relationship to prostate cancer. Proc Natl Acad Sci USA 1997;94:3320– 3323. 10. Hardy DO, Scher HI, Bogenreider T, Sabbatini P, Zhang ZF, Nanus DM, Catterall JF. Androgen receptor CAG repeat lengths in prostate cancer: correlation with age of onset. J Clin Endocrinol Metab 1996;81:4400–4405. 11. Irvine RA, Yu MC, Ross RK, Coetzee GA. The CAG and GGC microsatellites of the androgen receptor gene are in linkage disequilibrium in men with prostate cancer. Cancer Res 1995;55: 1937–1940. 12. Stanford JL, Just JJ, Gibbs M, Wicklund KG, Neal CL, Blumenstein BA, Ostrander EA. Polymorphic repeats in the androgen receptor gene: molecular markers of prostate cancer risk. Cancer Res 1997;57:1194–1198. 13. Ingles SA, Ross RK, Yu MC, Irvine RA, La Pera G, Haile RW, Coetzee GA. Association of prostate cancer risk with genetic polymorphisms in vitamin D receptor and androgen receptor. J Natl Cancer Inst 1997;89:166–170. 14. Hakimi JM, Schoenberg MP, Rondinelli RH, Piantadosi S, Barrack ER. Androgen receptor variants with short glutamine or glycine repeats may identify unique subpopulations of men with prostate cancer. Clin Cancer Res 1997;3:1599–1608. 15. Giovannucci E, Platz EA, Stampfer MJ, Chan A, Krithivas K, Kawachi I, Willett WC, Kantoff PW. The CAG repeat within the androgen receptor gene and benign prostatic hyperplasia. Urology 1999;53:121–125. 16. Kitagawa N, Ichikawa T, Akimoto S, Shimazaki J. Natural course of human benign prostatic hyperplasia with relation to urinary disturbance. Prostate 1994;24:279–284. 17. Partin AW, Page WF, Lee BR, Sanda MG, Miller RN, Walsh PC. Concordance rates for benign prostatic disease among twins suggest hereditary influence. Urology 1994;44:646–650. 18. Sanda MG, Beaty TH, Stutzman RE, Childs B, Walsh PC. Genetic susceptibility of benign prostatic hyperplasia. J Urol 1994; 152:115–119.