International Journal of Impotence Research (2003) 15, 391–396 & 2003 Nature Publishing Group All rights reserved 0955-9930/03 $25.00 www.nature.com/ijir Downregulation of , and and is involved in aging-related erectile dysfunction

M Shirai1, M Yamanaka1, H Shiina1, M Igawa2, M Fujime3, TF Lue4 and R Dahiya1*

1VA Medical Center and UCSF, Urology, San Francisco, California USA; 2Shimane Medical University, Izumo, Shimane, Japan; 3Juntendo University, Bunkyo-ku, Tokyo, Japan; and 4UCSF/Mt. Zion Medical Center, Department of Urology, San Francisco, California, USA

We hypothesize that downregulation of sex receptors (androgen, estrogen and progesterone receptors) is involved in aging-related erectile dysfunction. To test this hypothesis, we investigated the expression of receptors in penile crura of aging rats. A total of 40 rats were divided into four groups based on age (6, 12, 18 and 24 months), and the erectile function was analyzed by the measurement of intracavernous pressure. and protein expressions of sex hormone receptors were analyzed by RT-PCR and immunostaining, respectively. The mean intracavernous pressures of 6-, 12-, 18- and 24-month-old rats were 110.1, 89.6, 73.5 and 42.7 cmH2O, respectively. Gene and protein expressions for , - beta and were present in similar levels in 6-, 12- and 18-month-old rat crura, but significantly lower or absent in 24-month-old crura. This is the first study to demonstrate that downregulation of sex hormone receptors in aging rat crura is associated with erectile dysfunction. International Journal of Impotence Research (2003) 15, 391–396. doi:10.1038/sj.ijir.3901050

Keywords: genes; erectile dysfunction; aging

Introduction and progesterone are present in the circulation in men, and it may be possible that the altered balance between sex and their receptors exerts an Erectile dysfunction (ED) is common in elderly influence on male sexual function. In the central 1,2 men. The Massachusetts Male Aging Study nervous system, progesterone could mediate male reported that the combined prevalence of minimal, sexual behavior by interacting with progesterone moderate and complete ED reached up to 52% in receptor (PR) located in the olfactory bulb.4 aging men. The prevalence of complete ED tripled Although male organs produce and pro- 1 from 5 to 15% in subjects aged 40–70 y old. Older gesterone, their effects on penis are not clear. In a men are more frequently affected by systemic previous study, we have demonstrated that the diseases and often take a lot of medication. estrogen receptor-beta (ER-b) was Although it might be possible that these factors are significantly reduced in diabetic rat penis as inter-related with the higher incidence of ED, the compared to controls.5 However, there have been aging process itself undoubtedly plays a significant few reports on the relationship of sex hormone role in the development of ED. receptors (SHR) with aging-related ED. We hypothe- Circulating androgen in serum is considered to be size that SHR (androgen receptor (AR), ER and PR) one of the principal factors that affect male sexual play a significant role in maintaining the male function. It has been shown in rats that an adequate sexual function and altered levels of SHR expression level is essential to maintain the are involved in the etiology of aging-related ED. To availability of nitric oxide in the cavernous test this hypothesis, we analyzed gene and protein compartment through stimulation of expression expressions of SHR in aging rat crura and correlated and activity of penile endothelial and neuronal them with erectile function. nitric oxide synthase (NOS) isoforms.3 Estrogens Materials and methods

*Correspondence: R Dahiya, VA Medical Center and UCSF, Urology, 112F, 4150 Clement Street, San Francisco, Experimental animals CA 94121, USA. E-mail: [email protected] Received 26 October 2002; revised 7 March 2003; accepted A total of 40 male Fisher rats, divided into four 11 March 2003 groups (n ¼ 10) based on age (6, 12, 18 and 24 Downregulation of genes M Shirai et al 392 months) were used in this study. These rats Louis, MO, USA) and PCR reaction buffer provided were purchased from the National Institute of by the manufacturer. The primers used for rat AR, Aging, Washington, D.C. and/or Simonson, Gilroy, ER-a, ER-b, PR and glyceraldehyde 3-phosphate California. dehydrogenase (G3PDH) as a reference gene in RT-PCR were as follows: AR forward, 50-TGTTATC TAGTCTCAACGAGC-30; AR reverse, 50-CATCATTT CAGGAAAGTCC-30; ER-a forward, 50-GTCCAATTC Electrostimulation TGACAATCG-30; ER-a reverse, 50-CTTCAACATTC TCCCTCC-30; ER-b forward, 50-TGTACCATAGACA AGAACC-3; ER-b reverse, 50-GGAGTATCTCTGTGT After anesthetizing by intraperitoneal injection of 0 0 (40 mg/kg), a lower abdominal inci- GAAG-3 ; PR forward, 5 -CTCCTGGATGAGCCTGA TG-30; PR reverse, 50-CCCGAATATAGCTTGACCTC- sion was made and the cavernous nerve was 0 0 0 exposed. Electrostimulation was performed using a 3 , G3PDH forward, 5 -TCCCATCACCATCTTCCA-3 and G3PDH reverse, 50-CATCACGCCACAGTTTCC- delicate stainless-steel bipolar hook electrode. The 0 diameter of each pole was 0.2 mm and the two poles 3 . PCR reactions were performed in a PTC-200 were separated by 1-mm distance. Monophasic thermal cycler (MJ Research, Watertown, MA, USA) rectangular voltage pulses were generated by a at 941C for 3 min; 32 cycles at 941C for 1 min, 521C computer and converted to current pulses by a for 1 min, 721C for 1 min; followed by final extension computer-assisted stimulator. The stimulus para- at 721C for 5 min. For semiquantitative analysis, meters were 1.5 mA, 20 Hz frequency, pulse width three other dilution sets of each sample (original, 1/ 0.2 ms and duration 60 s. Each cavernous nerve was 2, 1/4 and 1/8 dilution) underwent PCR. Annealing stimulated. For monitoring intracavernous pressure temperature was 521C for all RT-PCR reactions. The (ICP), the skin overlying the penis was incised and PCR products were electrophoresed on 2.0% agarose the penile crura were exposed. A 25-gauge needle gel and the expression level of these genes was filled with heparin solution (1000 U/ml) was inserted evaluated by an ImageJ software (http://rsb.info. into the right crus and connected to a pressure nih.gov/ij), and the areas under the curves were monitor with polyethylene tubing.6 ICP was mea- calculated and analyzed. PCR cycles were adjusted sured and recorded using a computer with Lab View until these four preparations in each sample were 5.01 software (National Instruments, Austin, TX, within linear range. Between 28 and 32 cycles all USA). The ICP was defined as the maximum SHR PCR reactions were within linear range. The pressure obtained by the stimulation minus the expression of each gene was quantified relative to basal pressure before the stimulation. G3PDH expression and expressed as arbitrary units (AU). The expected sizes of PCR products were 587 (bp), 346, 633, 293 and 380 bp in AR, ER-a, ER-b, PR and G3PDH, respectively. Tissue preparation

Penile crura of each rat were collected immediately Immunohistochemistry after completing electrostimulation. Half of the sample was fixed in 10% buffered formalin (pH Tissue sections were cut (4 mm) and mounted on 7.0) and embedded in paraffin wax, and used for a slide. Tissue block sections were dewaxed, immunostaining. The remaining half of each sample rehydrated, incubated with 0.3%(v/v) hydrogen 1 was immediately frozen and stored at –80 C until peroxide for 20 min, and autoclaved at 1211C for analyzed. 15 min in 10 mM citrate buffer (pH 6.0). The samples were incubated with serum block (Santa Cruz Bio- technology, Inc., Santa Cruz, CA, USA) for 20 min at Reverse -polymerase chain reaction room temperature followed by overnight at 41C with (RT-PCR) the following antibodies: 1:150 dilution of a mono- clonal antibody against mouse ER-a (Ab-10) (Lab Vision Corp., Fremont, CA, USA), 1:400 dilution of a Total RNA was extracted from the samples of penile polyclonal antibody against rabbit AR (C-19), 1:400 crura using Tri-Reagent (Molecular Research Center, dilution of a polyclonal antibody against rabbit ER-b Cincinnati, OH, USA). Complementary DNA (cDNA) (H-150) and 1:100 dilution of a polyclonal antibody was constructed by reverse transcription (Promega against rabbit PR (H-190) (Santa Cruz Biotechnology, Corp., Madison, WI, USA) using the total RNA as a Inc.). After the slides were washed, they were template. The samples then underwent PCR ampli- incubated with a biotinylated goat anti-mouse anti- fication, cDNA samples were diluted in 20 mlof body and streptavidin-peroxidase (Lab Vision Corp.) solution containing 200 mM of dNTP, 500 nM of each for ER-a; or biotinylated goat anti-rabbit antibody primer, 0.5 U of Red Taq DNA polymerase (Sigma, St and HRP-streptavidin (Santa Cruz Biotechnology,

International Journal of Impotence Research Downregulation of sex hormone receptor genes M Shirai et al 393 Inc.) for AR, ER-b and PR at room temperature for Gene expression of AR, ER-a, ER-b and PR in aging 30 min. Following incubation with the avidin–biotin rat crura (Figures 1 and 2). complex, the antigen–antibody complexes were visualized with diaminobenzidine (Lab Vision Corp.). The sections were counterstained using Typical results and summarized data of each RT- hematoxylin. As a negative control, normal goat PCR are shown in Figures 1 and 2, respectively. IgG was substituted for the specific antibody at There was no significant difference in AR mRNA the same dilution. Paraffin sections of a normal expression between 6-, 12- and 18-month-old rats. rat and prostate were used as a positive In 24-month-old rats, expression of AR mRNA control. transcript was significantly reduced (Po0.01, A Leitz LaboLux S-microscope and a Nikon digital respectively). camera were used to determine each SHR immuno- For ER-a mRNA, there were two splicing isoforms reactivity within the cavernous tissue. The immuno- within the primer spanning region, and a strong staining pictures were analyzed on a PC with Image band corresponded to 346 bp (Figure 1). There was Pro software. The nucleus in the cavernous tissue no significant difference in the expression of ER-a was identified at  400. The pixels in the color mRNA among rats of different ages. As shown in range that corresponded to the stain of each SHR Figure 1, expression level of the splicing variant of and negative nucleus were counted and expressed ER-a mRNA was decreased with advancing age. as the percentage of the total number of pixels in the field. The number of pixels that corresponded to all nucleus (positive and negative nuclei) was counted, and defined it as 100% in each area. Three different sections in each animal were used for each SHR immunostaining, and 10 different fields were ana- lyzed in each specimen.

Statistical analysis

Parametric data were expressed as mean7s.e. Data were analyzed by one-way analysis of variance (ANOVA) with Fisher’s protected least-significance t post hoc test using StatView 5.01 software (SAS Institute Inc., Cary, NC, USA). The P-value of less than 0.05 was considered to be statistically significant.

Results

Functional study of aging rat penis

The ICP of rats after electrostimulation was reduced with advancing age. In 6-month-old rat, the pattern of ICP curve was sharp and the peak pressure was the highest. On the other hand, in 24-month-old rat, the peak pressure was the lowest. The mean ICP was 110.178.3, 89.6712.1, 73.5713.9 and 42.77 13.3 cm H2O in 6-, 12-, 18- and 24-month-old rats, respectively, indicating that ICP was decreased with advancing age. Significant differences in ICP were observed between 6- and 12-month-old rats, or between 18- and 24-month-old rats (P 0.05, respec- Figure 1 Typical RT-PCR results of SHR in aging rats are shown. o Note that in 24-month-old rat, expression level of AR and ER-b tively). The difference in ICP between 12- and 18- messenger RNA transcript is weak but preserved, while that of PR month-old rats was close to statistical significance is lost (* ¼ 100 bp DNA ladder). Relative mRNA value is shown on (P ¼ 0.05). each band.

International Journal of Impotence Research Downregulation of sex hormone receptor genes M Shirai et al 394

Figure 3 Typical immunostainings of AR (a and c), ER-a (b and d). Note that protein expression of AR was decreased in 24-month- old rat, whereas that ER-a was not significantly different: (a and b) 6-month ( Â 400); (c and d): 24-month ( Â 400).

Figure 2 Alterations of SHR mRNA expression with aging are shown. Expression of AR, ER-b and PR mRNA was decreased along with aging, whereas that of ER-a was not significantly different among all aging groups. Values are expressed as means7s.e. (AU ¼ arbitrary units).

ER-b mRNA expression among 6-, 12- and 18- month-old rats did not differ, but was significantly reduced in 24-month-old rats (Po0.01, respectively). The expression of PR mRNA transcript was the highest in 6-month and was lost in 24-month-old rats. The difference in PR mRNA expression between 6- and 24-month-old rats reached statistical significance (P 0.01), but no significant difference o Figure 4 Typical immunostainings of ER-b (a and c) and PR (b was observed between 12- and 18-month-old rats. and d). Note that protein expression of ER-b and PR was decreased and lost in 24-month-old rat, respectively: (a and b) 6-month ( Â 400); (c and d) 24-month ( Â 400). Protein expression of AR, ER-a, ER-b and PR in aging rat crura (Figures 3–5). ER-a-positive cells were distributed in the nuclei of cells in 6-month-old rat crura Typical result of each immunostaining is shown in (Figure 3b). The number of the positive cells was Figures 3 and 4, and summarized data are shown in similar among 6- and 24-month-old groups (Figures Figure 5. 3b and d) and no significant difference in ER-a AR protein expression was the highest in 6- immunoreactivity was observed among rats of month-old rat. The immunoreactivity was observed different ages. in the nuclei of smooth muscle cells of small artery ER-b immunoreactivity was reduced with advan- as well as smooth muscle fibers within the crura cing age of rats (Figure 4a and c). The difference in (Figure 3a). In 24-month-old rat crura, the number of the immunoreactivity reached statistical signifi- AR-positive cells was the lowest as compared to 6- cance between 6- and 12-month as well as between month-old rats (Figure 3c). A significant difference 18- and 24-month-old rats (both Po0.05). was observed between 12- and 18-month-old rats or PR immunoreactivity was significantly reduced 18- and 24-month-old rats (Po0.05); however, the with advancing age (Figure 4b and d). Interestingly difference between 6- and 12-month-old rats did not in 24-month-old rat crura, the PR immunoreactivity reach statistical significance. was completely lost (Figure 4d).

International Journal of Impotence Research Downregulation of sex hormone receptor genes M Shirai et al 395 cause relative androgen deprivation status in aging rat crura. These alterations are probably associated with the enhancement of degenerative processes of smooth muscle cells and/or reduction in eNOS and nNOS.3,14 Estrogen functions as regulator of growth and differentiation in various target tissues. Its receptors, ER-a and ER-b mediate the functional role of estrogen in target organs, but the distribution of ERs is quite different among rat tissues. The pituitary, kidney, epididymis and adrenal are pre- dominant for ER-a, whereas prostate, ovary, lung, bladder, and testis are higher in ER-b.15 In rat crura, gene and protein expression of ER-b was reduced with advancing age and markedly decreased in 24-month-old rats, whereas ER-a gene and protein expressions were not significantly changed among 6-, 12-, 18- and 24-month-old rats. In addition, the age-related ICP alteration correlated with the change of ER-b expression, but not with that of ER-a expression. These findings indicate that Figure 5 Alterations of SHR immunoreactivity with aging. the functionally predominant form of ER in rat crura Reduction of immunoreactivity of AR, ER-b and PR was is ER-b and age-related alteration of ER-b is probably associated with aging, while ER-a immunoreactivity was not related to the pathogenesis of ED in older rats. The significantly changed among all age groups. In 24-month-old rats, mechanism of ER-b involved in aging-related ED AR and ER-b protein expressions were significantly lower than 6-, still remains to be elucidated. Estrogen exerts direct 12- and 18-month-old rats (Po0.05). PR protein expression was lost in 24-month-old rats. Values are expressed as means7s.e. vascular protection effect on endothelial and/or smooth muscle cells, and this effect is exclusively mediated by ER-b, but not by ER-a.16,17 Estrogen also exerts an inhibitory effect on in human Discussion endothelial cells with upregulation of antiapoptotic Bcl-2.18,19 Loss of cytoprotective role of ER-b with aging might be one of the mechanisms of develop- Previous studies in our laboratory and in others ment and/or progression of aging-related ED. Alter- 7,8 have shown that aging is associated with ED. The natively, potential antiapoptotic effect by estrogen molecular mechanisms of aging-related ED are not might be impaired by reduced expression of ER-b fully understood. We hypothesized that alterations with advancing age. In fact, a recent publication in SHR are involved in the pathogenesis of ED in clearly demonstrated that loss of antiapoptotic genes aging rats. To test this hypothesis, we analyzed gene in rat crura is associated with the pathognesis of and protein expression of AR, ER and PR in aging rat aging-related ED.20 crura. First, we measured ICP to verify the presence There are no reports of the relationship between of ED in the aging rat. ICP was lowered with development and/or progression of ED and PR advanced age, which is consistent with the finding expression in the aging rat crura. Progesterone has 9 observed in aging ED. been shown to be involved in the modulation of Several lines of studies have demonstrated that NOS activity21 and may be due to PR. The most androgen deprivation is related to apoptosis of interesting finding was that PR gene and protein smooth muscle cells in addition to an increase in expression were almost lost in 24-month-old rat 10–12 connective tissue contents. Penile eNOS and crura. In addition, the mean ICP drop between 18- nNOS are two major contributors to maintain and 24-month-old rats (about 31 cm H2O) is greater erectile function, of which activity is restored after than those observed between 6- and 12-month androgen stimulation in castrated rats. In the (about 20 cm H2O) or between 12- and 18-month present study, both levels of protein and mRNA (about 16 cm H2O). Taken together, it might be transcript of AR were significantly decreased with plausible that the impaired functional role of PR as advancing age in rat crura. In addition, this aging- modulator of NOS activity reflects the lowest level related alteration of AR expression was well corre- of ICP in 24-month-old rats. lated with reduction of ICP. Circulating level of In summary, ICP alteration in relation to aging serum testosterone has been reported to be reduced crura is well correlated with AR, ER-b and PR with advancing age in rat.13 Even if sufficient expression. Even between 18- and 24-month of amount of androgen is present in the older rats, age, the abrupt reduction in AR, ER-b and PR significant reduction of AR expression itself might still displayed significant correlation with ICP

International Journal of Impotence Research Downregulation of sex hormone receptor genes M Shirai et al 396 reduction. These findings strongly suggest that AR, 6 Bakircioglu ME et al. Decreased trabecular smooth muscle and ER-b and PR are involved in the pathogenesis of ED caveolin-1 expression in the penile tissue of aged rats. J Urol with advancing age. 2001; 166: 734 – 738. 7 Dahiya R et al. Differential gene expression of growth factors in young and old rat penile tissues is associated with erectile dysfunction. Int J Impot Res 1999; 11: 201 – 206. Conclusions 8 Ferrini M et al. Aging-related expression of inducible nitric oxide synthase and markers of tissue damage in the rat penis. Biol Reprod 2001; 64: 974 – 982. 9 Garban H et al. Effect of aging on nitric oxide-mediated penile To our knowledge, this is the first study demonstrat- erection in rats. Am J Physiol 1995; 268: H467 – H475. ing that downregulation of gene and protein expres- 10 Baskin LS et al. The effect of testosterone on androgen sion of ER-b and PR is associated with ED of aging receptors and human penile growth. J Urol 1997; 158: 1113 – rats. These results may be important in under- 1118. 11 Shabsigh R. The effect of testosterone on the cavernous tissue standing the pathogenesis of aging-related ED and and erection. World J Urol 1997; 15: 21 – 26. also in providing better strategies for the manage- 12 Traish AM et al. Effects of castration and androgen replace- ment of aging-related ED. ment on erectile function in a rabbit model. 1999; 140: 1861 – 1868. 13 Banerjee PP et al. Age-dependent and lobe-specific sponta- neous hyperplasia in the brown Norway rat prostate. Biol Acknowledgements Reprod 1998; 59: 1163 – 1170. 14 Aversa A et al. Androgen and penile erection: evidence for a direct relationship between free testosterone and cavernous This research was supported by the National in men with erectile dysfunction. Clin Endo- crinol 2000; 53: 517 – 522. Institutes of Health Grants RO1DK055040 and 15 Kuiper GGJM et al. Comparison of the binding RO1AG016870. specificity and transcript tissue distribution of estrogen receptors alpha and beta. Endocrinology 1997; 138: 867 – 870. 16 Zhou Y et al. Hepatic stellate cells contain the functional but not the in References male and female rats. Biochem Biophys Res Commun 2001; 286: 1059 – 1065. 17 Lindner V et al. Increased expression of estrogen receptor-beta 1 Feldmann HA et al. Impotence and its medical and psycho- mRNA in male blood vessels after vascular injury. Circ Res social correlates: results of the Massachusetts Male Aging 1998; 83: 224 – 229. Study. J Urol 1994; 151: 54 – 61. 18 Spyridopoulos I et al. Estrogen-receptor-mediated inhibition 2 Jonler M et al. The effect of age, ethnicity and geographical of human endothelial apoptosis. as a survival location on impotence and quality of life. Br J Urol 1995; 75: factor. Circulation 1997; 95: 1505 – 1514. 651 – 655. 19 Choi KC et al. Estradiol upregulates antiapoptotic Bcl-2 3 Penson DF et al. Androgen and pituitary control of penile messenger ribonucleic acid and protein in tumorigenic nitric oxide synthase and erectile function in the rat. Biol ovarian surface cells. Endocrinology 2001; 142: Reprod 1996; 55: 567 – 574. 2351 – 2360. 4 Guerra-Araiza C et al. Progesterone receptor isoforms expres- 20 Yamanaka M et al. Loss of anti-apoptotic genes in aging rat sion in the prepuberal and adult male rat brain. Brain Res Bull crura. J Urol 2002; 168: 2296 – 2300. 2001; 54: 13 – 17. 21 Labombarda F et al. Modulation of NADPH-diaphorase and 5 El-Sakka A et al. Effects of on nitric oxide synthase glial fibrillary acidic protein by progesterone in astrocytes and genes and protein expression in an animal from normal and injured rat spinal cord. J Steroid Biochem model. Int J Impot Res 1999; 11: 123 – 132. Mol Biol 2000; 73: 159 – 169.

International Journal of Impotence Research