Direct androgen regulation of PDE5 gene or the lack thereof

Transcription

1 International Journal of Impotence Research (2013) 25, & 2013 Macmillan Publishers Limited All rights reserved /13 REVIEW Direct androgen regulation of PDE5 gene or the lack thereof C-S Lin 1, Z Xin 2, M Namiki 3, M Albersen 4, D Muller 5 and TF Lue 1 Inhibition of phosphodiesterase-5 (PDE5) is a well-known mechanism for the effective treatment of erectile dysfunction (ED). Androgen supplementation has also been prescribed for treating ED. However, it has been widely accepted that androgen can upregulate PDE5 expression, and thus creating a paradox in which a positive regulator of erectile function (androgen) could possibly increase the level of a negative regulator (PDE5). To solve this paradox, we conducted a systematic search of the PubMed and a non-systematic search of the Internet using PDE5, erectile, penis, testosterone and androgen as keywords. The retrieved papers were analyzed for data concerning the expression and regulation of PDE5 by androgens. Human and rat PDE5A gene sequences were retrieved from GenBank and computer-analyzed. The results showed that a putative androgen-response element (ARE) was reported in a study of human PDE5A gene promoter, and this prompted a separate study on whether androgen regulates PDE5 expression. The positive outcome in the latter study has since been cited in 17 review and editorial articles as the underlying mechanism for androgen s therapeutic effects on ED. In addition, five other research studies also reached the same conclusion. On the other hand, two independent studies on the genome-wide searches for androgen-regulated genes did not find PDE5A as a candidate. Sequence analysis conducted in this study also failed to find ARE in rat PDE5A gene. Two independent studies on Leydig cells also failed to find positive regulation of PDE5 expression by androgen. Two other studies found concomitant reduction of cavernous smooth muscle and PDE5 expression in castrated rats. One of these studies also found no effect of androgen on PDE5 expression in cultured cavernous smooth muscle cells. Thus, it appears that reduced PDE5 expression in castrated animals is due to reduced smooth muscle content and that PDE5A gene is not directly regulated by androgens. International Journal of Impotence Research (2013) 25, 81 85; doi: /ijir ; published online 14 March 2013 Keywords: androgen; erectile dysfunction; PDE5; penis INTRODUCTION Erectile dysfunction (ED) is a prevailing health problem that seriously impacts the quality of life of men and their spouses/ partners. 1 Among many treatment options, a class of medications known as PDE5 inhibitors have been the most effective. 2 PDE5 is an enzyme (phosphodiesterase, PDE) that catalyzes the breakdown of cyclic guanosine monophosphate (cgmp). 3 In the penis, this enzymatic activity leads to penile flaccidity, and therefore, the inhibition of PDE5 enables penile erection that otherwise cannot be achieved. 4,5 Thus, as far as erectile function is concerned, PDE5 is a negative regulator. 6 Androgen is critical for male sexual development, but whether it is important for erectile function has been a subject of much heated debate. In recent years, there has been a resurgence of interest in using androgens to treat ED, and many leading experts in this field have explicitly expressed their favorable opinion about this practice. However, a substantial number (X17) of these experts review/editorial articles also cite positive androgen regulation of PDE5 gene as an underlying mechanism for androgen s therapeutic effects Thus, on the one hand, androgen supplementation is advocated as a treatment option for ED; on the other, androgen supplementation is believed to increase the level of PDE5, which obviously would lead to increased severity of ED or increased difficulty to manage ED. Thus, the purpose of this review article is to summarize all available data concerning this issue of androgen regulation of PDE5 expression, and to offer an explanation for why and how the paradox arose. METHODS Retrieval and analysis of relevant articles To investigate how the androgen PDE5 paradox came about, we conducted systematic search of the PubMed and non-systematic search of the Internet using PDE5, penis, erectile, testosterone and androgen as keywords. Because PubMed permits search in the title and abstract fields only, articles that contain these keywords only in the main text cannot be located. On the other hand, Internet permits search in the main text, but the results invariably contain too many irrelevant hits. Thus, as a compromise, best effort was applied to identify the relevant hits from the Internet, which were then followed by retrieving the pertinent articles from PubMed or from their publishers. All retrieved articles were then searched for contents that specifically address the issue of androgen regulation of PDE5. The positively identified articles were then analyzed and used for discussion in this study. Retrieval and analysis of PDE5A gene sequences Human and rat PDE5 gene sequences were retrieved from GenBank. They were then analyzed for sequence similarities with each other and for the presence of the androgen-response element (ARE, 5 0 -AGAACAnnnTGTTCT- 3 0 lowercase letters indicate nucleotides not critical for androgen receptor binding) 24 using the MacVector software (MacVector, Cary, NC, USA). 1 Knuppe Molecular Urology Laboratory, Department of Urology, School of Medicine, University of California, San Francisco, San Francisco, CA, USA; 2 Andrology Center, Peking University First Hospital, Beijing, China; 3 Department of Integrative Cancer Therapy and Urology, Kanazawa University Graduate School of Medical Science, Kanazawa, Ishikawa, Japan; 4 Department of Urology, University Hospitals Leuven, Leuven, Belgium and 5 Institute of Anatomy and Cell Biology, Justus-Liebig University, Giessen, Germany. Correspondence: Dr C-S Lin, Knuppe Molecular Urology Laboratory, Department of Urology, School of Medicine, University of California, San Francisco, San Francisco, CA , USA. clin@urology.ucsf.edu Received 16 October 2012; revised 17 January 2013; accepted 9 February 2013; published online 14 March 2013

2 82 BLAST search The web-based Basic Local Alignment Search Tool (BLAST) was used to search for similarity between sequences. Specifically, it was used to identify the actual gene whose nucleotide sequence was used in a cited study for the design of reverse transcription-polymerase chain reaction (PCR) primers. RESULTS AND DISCUSSION Identification of a putative ARE in human PDE5 gene PDE5 is encoded from a single gene, PDE5A, in humans, rats and presumably all other mammals. 25 Human PDE5A gene encodes three PDE5 isoforms (A1, A2 and A3) from two alternative promoters, PDE5A1 and PDE5A2 25 (Figure 1). These isoforms differ only in the amino terminal, and except for PDE5A3 being smooth muscle-specific, no known functional differences exist. 26 The human PDE5A1 promoter consists of a 139-base pair (bp) core sequence that confers full basal promoter activity. 27 Approximately 1600 bp upstream, a putative ARE was identified by computer-assisted sequence alignment 27 (Figure 1). This sequence, 5 0 -AGAAAAaacTGTACT-3 0, differs from the consensus ARE, 5 0 -AGAACAnnnTGTTCT-3 0, 24 by one nucleotide on each of the 6-bp palindromic half-sites (Figure 1). To test whether androgen can upregulate PDE5 gene, human vascular and cavernous smooth muscle cells were treated with dihydrotestosterone (DHT) and then analyzed by reverse transcription-pcr. The results were negative for all smooth muscle cells tested; therefore, the data were not published, as stated in two review articles by the researchers who conducted these experiments. 4,26 This negative finding was not unexpected because the putative ARE was imperfect, and it has been shown that multiple copies of imperfect ARE need to coexist to have functionality As only one imperfect ARE was identified in PDE5 gene promoter, its existence is likely accidental, not functional. Genome-wide searches for androgen-responsive genes There have been two studies that conducted genome-wide searches for androgen-responsive genes in human cells. One of these studies employed two complimentary techniques, namely expression profiling and chromatin immunoprecipitation, on an immortalized human prostate epithelial cell line (HPr-1AR), which expresses androgen receptor (AR) and is androgen-responsive. The expression profiling technique identified 205 androgenresponsive genes, whereas the chromatin immunoprecipitation technique identified 524. Among the identified genes, PDE5A is not one of them. 31 The second study also employed the chromatin immunoprecipitation technique, but on a better-known cell line, namely the human prostate cancer cell line LNCaP, which also expresses AR and is androgen-responsive. This resulted in the identification of 1532 potential AR-binding sites; PDE5A is not one of them. 32 Figure 1. Regulatory regions of human and rat PDE5A genes. Nucleotide sequences of human and rat PDE5A genes were retrieved from GenBank and analyzed by the MacVector software. A1, A2 and A3 boxes represent coding regions for PDE5A1, A2 and A3 isoforms. A3? box represents a region having homology with human PDE5A3 sequence but not found to be transcribed. ARE? indicates possible androgen-responsive elements, whose sequences are shown on top of the consensus ARE sequence, with * indicating mismatches. Animal studies on whether PDE5 gene is regulated by androgen Between 2004 and 2007 four papers were published from the same laboratory that concluded that PDE5 gene was positively regulated by androgen (Table 1) In the first paper, the putative ARE in the human PDE5A promoter reported by Lin et al. 27 was cited as incentive for investigating whether PDE5 gene was regulated by androgen in a rabbit model of hypogonadotropic hypogonadism (induced by treatment with GnRH analog triptorelin). 33 The results showed that PDE5 expression in the corpus cavernosum was downregulated in triptorelin-treated rabbits, and was restored by testosterone supplementation. However, it should be noted that in the western blot analysis, no control (such as actin or glyceraldehyde 3-phosphate dehydrogenase) was included, and in the reverse transcription-pcr analysis a questionable control was used. Specifically, the control was said to be g-non-muscle actin, citing GenBank accession number X60733 as the source for the primer pair s nucleotide sequences. However, X60733 is linked to a paper that reports the nucleotide sequence of rabbit b-actin. 37 Furthermore, to resolve this discrepancy, we conducted BLAST sequence searches for the primer pair, and the results showed that indeed they are b-, not g-, actin sequences. In any event, b-actin is expressed in both smooth muscle and non-muscle cells, and therefore, not an appropriate control for testing gene expression changes in smooth muscle cells. Specifically, it has been shown that castration led to reduced smooth muscle content and accumulation of adipocytes in rabbit corpus cavernosum. 38 Likewise, it has also been shown that castration resulted in the dedifferentiation of smooth muscle cells and reduced expression of smooth muscle-specific myosin and smoothelin in the prostate. 39,40 Thus, in the corpus cavernosum or prostate of hypogonadal animals, smooth muscle cells are replaced by non-smooth muscle cells. These non-smooth muscle cells, however, still express b-actin, and thus when b-actin is used as a control, it is incapable of normalization for hypogonadismassociated gene changes that occur specifically in smooth muscle cells. The above-mentioned PDE5 study also evaluated the effect of androgens on cgmp-hydrolyzing activity in the homogenates prepared from the corpus cavernosum of the same animal model. The results were that triptorelin treatment significantly reduced the activity, while testosterone supplementation not only restored but also increased the activity. However, it should be noted that in this experiment a questionable negative control was used. Specifically, based on the study s own western blot analysis that showed little or no PDE5 expression in human uterus, rabbit uterine homogenates were used as negative control in the cgmphydrolyzing activity assay. However, the negative finding about uterine PDE5 expression contradicts all other studies (total of 11) that have reported uterine PDE5 expression In these studies, uterus from various species (human, rat and mouse) was found to express PDE5 and/or respond to PDE5 inhibitor sildenafil. Particularly, one of such studies was conducted by a dedicated PDE research team, and it showed that in mouse uterus phospho- PDE5 was easily detectable and PDE5 is phosphorylated in intact tissues (mouse uterus) upon activation of PKG. 44 While it can be argued that rabbit was not among the species tested positive for uterine PDE5 expression, it is noted that in these authors second androgen-pde5 paper, rat uterus was also found to express little PDE5. 34 In the second androgen-pde5 paper, PDE5 expression in rat corpus cavernosum was found downregulated by castration and restored by testosterone supplementation. 34 However, like their first paper, no control was used in the western blot analysis, and b- 2-microglobulin was used for normalization in the real-time PCR analysis. Note that similar findings have been reported by a separate group of researchers, but their western blot analysis was also conducted without control. 52 In the third androgen PDE5 International Journal of Impotence Research (2013), & 2013 Macmillan Publishers Limited

3 83 Table 1. Studies examining androgen regulation of PDE5 expression Publication year/ first author Animal or cell Treatment Target tissue Assay method Assay control PDE5 expression Additional notes 2004/Morelli 33 Rabbit Triptorelin CC qpcr None Decrease Decrease 2005/Zhang 34 Rat Castration CC qpcr b-2-mg Decrease None Decrease 2006/Zhang 35 Rat STZ CC qpcr b-2-mg Decrease Rabbit Alloxan None Decrease 2006/Armagan 52 Rat Castration CC None Decrease 2007/Filippi 36 Rat Castration Bladder qpcr b-2-mg Decrease 2009/Yang 64 Rat Castration CC IF, PCR, qpcr and SMA No change Rat DHT western SMA No change CSMC 2010/Andric 67 Rat Rat LC T Triptorelin T LC LC qpcr and western qpcr and western Increase No change No change 2011/Muller 68 Rat EDS Prostate None Increase 2012/Zhang 53 Rat Castration Prostate qpcr L19 Decrease Decrease mistaken for g-actin PDE5 promoter assay also showed no change Abbreviations: b-2mg, b-2-microglobulin; CC, corpus cavernosum; CSMC, cavernous smooth muscle cells; DHT, dihydrotestosterone; EDS, ethane dimethane sulfonate; IF, immunofluorescence; LC, Leydig cells; PCR, reverse transcription-polymerase chain reaction; qpcr, quantitative PCR; SMA, smooth muscle actin; STZ, streptozotocin; T, testosterone. paper, PDE5 expression in the corpus cavernosum of streptozotocin-induced diabetic rats and alloxan-induced diabetic rabbits was downregulated when compared with non-diabetic counterparts, and this could be normalized by testosterone supplementation. 35 Again, in this study no control was used in the western blot analysis and b-2-microglobulin was used for normalization in the real-time PCR analysis. In the fourth paper, based on realtime PCR normalized by b-2-microglobulin, castration was found to reduce, and testosterone supplementation restore, PDE5 expression in rat bladder. 36 In a more recent study, 53 castration was found to reduce, and testosterone supplementation restore, PDE5 expression in rat prostate. In this study, realtime PCR was normalized by ribosomal protein L19 and western blot controlled by b-actin. Note that the evaluation of PDE5 expression in the prostate has been highly controversial, with levels ranging from extremely low to highly abundant, and this discrepancy also occurs between western blot and immunostaining analyses within this article. 54 One common denominator shared by all of the abovementioned androgen-for-pde5-regulation papers is the lack of smooth muscle control. It is well known that castration causes reduction of cavernous and prostatic smooth muscles. 38,55 59 It has also been shown that androgen supplementation increased bladder smooth muscle content in castrated rats. 60 Furthermore, streptozotocin-induced diabetic ED has been shown to be associated with reduced cavernous smooth muscle content. 61,62 Thus, when assessing gene expression changes in the penis, bladder and prostate of castrated and/or diabetic animals, it is critically important to use a smooth muscle marker (for example, a-smooth muscle actin) in western blotting and histological analyses to determine the smooth muscle content. Indeed, in a 2005 study, which focused on the effects of icariin on gene expression changes in the corpus cavernosum of castrated rats, PDE5 mrna and protein expression was found to be reduced, but so was the smooth muscle content as determined by trichrome staining. 63 In a more recent study, which focused on the effects of castration on PDE5 expression in the penis, the simultaneous reduction of smooth muscle content and PDE5 protein was again demonstrated, this time with far clearer histological images and quantitative data, as afforded by immunofluorescence staining with anti-a-smooth muscle actin antibody. 64 Moreover, in this study PDE5 expression was found to be unchanged in DHT-treated cavernous smooth muscle cells, and PDE5A promoter was unresponsive to DHT treatment. Thus, while PDE5 expression in castrated and diabetic rats could be restored by testosterone supplementation, the effect was likely due to restored smooth muscle content, and not by direct androgen regulation. Search for ARE in rat PDE5A gene All of the above-mentioned studies that concern androgen regulation of PDE5A gene were conducted with rats or rabbits. However, Homo sapiens is the only species in which PDE5A gene promoters have been functionally tested and their nucleotide sequences positively identified. 25,27,65 Thus, while a putative ARE was identified in human PDE5A gene, whether it also exists in the PDE5A gene of other species is unknown. In particular, as all but one of the above-mentioned studies investigated androgen regulation of PDE5 in the rat, it is important that we find out whether rat PDE5A gene has the same putative ARE as human PDE5A gene does. Although not functionally tested, the rat PDE5A gene promoters have been tentatively identified by nucleotide sequence analysis. 66 Besides, cdnas encoding rat PDE5A1 and PDE5A2 mrnas have been isolated and their sequences matched to the rat PDE5A gene sequence 45 (Figure 1). In addition, although repeated efforts to isolate the PDE5A3 cdna were not successful, the PDE5A3- equivalent sequence has been found in rat PDE5A gene 45 (Figure 1). Thus, rat and human PDE5A genes appear to be homologous in both nucleotide sequence and promoter arrangement. To test this hypothesis, we performed sequence alignment between human and rat PDE5A gene sequences and conducted a search for the putative ARE in rat PDE5A gene. First, we conducted sequence alignment between human and rat PDE5A genes in the regions (approximately 2770 nucleotides) upstream of their respective first exons (encoding the PDE5A1 isoform) (Figure 1). The results showed a high degree of homology (38%), considering that these are non-transcribed sequences. We then searched for the consensus ARE (5 0 -AGAACAnnnTGTTCT-3 0 ) 24 within the 2768-nucleotide sequence upstream of rat PDE5A1 ATG codon. As mentioned earlier, the putative ARE in human PDE5A gene has two mismatches out of the 12-bp palindrome of the consensus ARE (Figure 1). However, at this search stringency, no putative ARE could be found in rat PDE5A gene, and only when & 2013 Macmillan Publishers Limited International Journal of Impotence Research (2013), 81 85

4 84 the stringency was lowered to allowing five mismatches, a sequence was found. This sequence, 5 0 -ACAACAcccCTCTCC-3 0, located 2665 bp upstream from the ATG codon of the PDE5A1 isoform, differs from the consensus sequence by one and four nucleotides on the left and right 6-bp palindromic half-sites, respectively (Figure 1). As both half-site sequences of the ARE are important for AR binding (as homodimers), 24 a 4-out-6 mismatch on the right half-site points to the unlikelihood that this sequence could be a true ARE. Other relevant studies A 2010 study found that, in contrast to the study by Morelli et al., 33 treatment with triptorelin had no effect on PDE5 mrna and protein expression in Leydig cells of the chemically induced hypogonadal rats (Table 1). 67 However, treatment with testosterone did result in increases of PDE5 mrna and protein in Leydig cells isolated from normal or triptorelin-treated rats. 67 Nevertheless, cultured Leydig cells expressed similar levels of PDE5 when treated with increasing concentrations of testosterone (0, 1, 10 and 20 mm). 67 Thus, the increase of PDE5 expression in testosterone-treated rats is not due to direct androgen regulation of PDE5A gene, but probably depends on increased nitric oxide/ cgmp production. 67 In a 2011 study, several cgmp pathway proteins, including PDE5, were investigated in an androgen ablation rat model, in which Leydig cells were transiently destroyed between 2 and 14 days after a single intraperitoneal injection of ethane dimethane sulfonate. 68 In the prostate of this rat model 43.5-fold increases of PDE5 protein expression were noted at 7 and 14 days after ethane dimethane sulfonate injection. Thus, while it remains unknown why androgen ablation resulted in elevated PDE5 expression, this study demonstrated that PDE5A gene was not positively regulated by testosterone. In a 2011 study, androgen replacement was found to improve lower urinary tract symptoms in hypogonadal patients with benign prostate hyperplasia. 69 This finding also contradicts the notion that androgen is a positive regulator for PDE5 expression because PDE5 inhibitors have been rather consistently shown to improve lower urinary tract symptoms. 54 Thus, direct positive androgen regulation of PDE5 expression is inconsistent with clinical findings whether in ED or lower urinary tract symptoms. CONCLUDING REMARKS Although direct androgen regulation of PDE5 expression has been widely accepted, 7 23 its demonstration was based on animal studies, in which many factors can indirectly influence the expression of a particular gene. Specifically, a decrease of gene expression following castration does not prove direct positive androgen regulation, neither does restoration of gene expression by androgen supplementation. On the other hand, cell culture studies, which permit better control of the experimental condition, usually provide cleaner data. For example, in the case of PDE5 expression in Leydig cells, testosterone treatment was found to upregulate PDE5 expression in animals but not in cultured cells; and the upregulation in animals is possibly mediated by cgmp, not a consequence of direct action by androgens. 67 In the case of PDE5 expression in cavernous smooth muscle, a decreased level in castrated rats was found to be associated with decreased smooth muscle content, 63,64 and no change was noted in DHT-treated vascular or cavernous smooth muscle cells. 4,26,64 At the molecular level, the existence of a single and imperfect ARE in human PDE5A gene promoter indicates that it is non-functional, 29 and this is further supported by genome-wide studies that failed to identify PDE5A as an androgen-regulated gene in human prostate epithelial cells. 31,32 Furthermore, searches conducted in this study failed to identify ARE in rat PDE5A gene. Thus, androgen regulation of PDE5A gene is likely a conclusion reached by interpretation of experimental data that lack the critical smooth muscle control. And, with this realization, this study hopes to have solved the androgen-pde5 paradox. CONFLICT OF INTEREST The authors declare no conflict of interest. ACKNOWLEDGEMENTS This work was supported by grants from the National Institutes of Health (DK045370, DK64538 and DK069655). REFERENCES 1 Litwin MS, Nied RJ, Dhanani N. Health-related quality of life in men with erectile dysfunction. J Gen Intern Med 1998; 13: Dorsey P, Keel C, Klavens M, Hellstrom WJ. Phosphodiesterase type 5 (PDE5) inhibitors for the treatment of erectile dysfunction. Expert Opin Pharmacother 2010; 11: Thomas MK, Francis SH, Corbin JD. Characterization of a purified bovine lung cgmp-binding cgmp phosphodiesterase. J Biol Chem 1990; 265: Lin CS. Phosphodiesterase type 5 regulation in the penile corpora cavernosa. J Sex Med 2009; 6(Suppl 3): Lin CS, Lin G, Lue TF. Cyclic nucleotide signaling in cavernous smooth muscle. J Sex Med 2005; 2: Lin CS, Xin ZC, Wang Z, Lin G, Lue TF. Molecular Yin and Yang of erectile function and dysfunction. Asian J Androl 2008; 10: Shabsigh R. Testosterone therapy in erectile dysfunction and hypogonadism. J Sex Med 2005; 2: Vignozzi L, Corona G, Petrone L, Filippi S, Morelli AM, Forti G et al. Testosterone and sexual activity. J Endocrinol Invest 2005; 28: Morelli A, Filippi S, Zhang XH, Luconi M, Vignozzi L, Mancina R. Peripheral regulatory mechanisms in erection. Int J Androl 2005; 28(Suppl 2): Seftel AD. Testosterone regulates PDE5 expression and in vivo responsiveness to tadalafil in rat corpus cavernosum. J Urol 2005; 174: Morelli A, Vignozzi L, Filippi S, Mancina R, Maggi M. Erectile dysfunction: molecular biology, pathophysiology and pharmacological treatment. Miner Urol Nefrol 2005; 57: Gooren LJ, Saad F. Recent insights into androgen action on the anatomical and physiological substrate of penile erection. Asian J Androl 2006; 8: Buvat J, Bou Jaoude G. Significance of hypogonadism in erectile dysfunction. World J Urol 2006; 24: Shabsigh R, Rajfer J, Aversa A, Traish AM, Yassin A, Kalinchenko SY et al. The evolving role of testosterone in the treatment of erectile dysfunction. Int J Clin Pract 2006; 60: Greco EA, Spera G, Aversa A. Combining testosterone and PDE5 inhibitors in erectile dysfunction: basic rationale and clinical evidences. Eur J Urol 2006; 50: Burnett AL. Molecular pharmacotherapeutic targeting of PDE5 for preservation of penile health. J Androl 2008; 29: Gurbuz N, Mammadov E, Usta MF. Hypogonadism and erectile dysfunction: an overview. Asian J Androl 2008; 10: Hwang TI, Lin YC. The relationship between hypogonadism and erectile dysfunction. Int J Impot Res 2008; 20: Traish AM. Androgens play a pivotal role in maintaining penile tissue architecture and erection: a review. J Androl 2009; 30: Yassin AA, Saad F. Testosterone and erectile dysfunction. J Androl 2008; 29: Yassin AA, Akhras F, El-Sakka AI, Saad F. Cardiovascular diseases and erectile dysfunction: the two faces of the coin of androgen deficiency. Andrologia 2011; 43: Zhang XH, Melman A, Disanto ME. Update on corpus cavernosum smooth muscle contractile pathways in erectile function: a role for testosterone? J Sex Med 2011; 8: Mazzola CR, Mulhall JP. Impact of androgen deprivation therapy on sexual function. Asian J Androl 2012; 14: Faisst S, Meyer S. Compilation of vertebrate-encoded transcription factors. Nucleic Acids Res 1992; 20: Lin CS, Chow S, Lau A, Tu R, Lue TF. Human PDE5A gene encodes three PDE5 isoforms from two alternate promoters. Int J Impot Res 2002; 14: Lin CS, Lin G, Xin ZC, Lue TF. Expression, distribution and regulation of phosphodiesterase 5. Curr Pharm Des 2006; 12: International Journal of Impotence Research (2013), & 2013 Macmillan Publishers Limited

5 27 Lin CS, Chow S, Lau A, Tu R, Lue TF. Identification and regulation of human PDE5A gene promoter. Biochem Biophys Res Commun 2001; 280: Kasper S, Rennie PS, Bruchovsky N, Sheppard PC, Cheng H, Lin L et al. Cooperative binding of androgen receptors to two DNA sequences is required for androgen induction of the probasin gene. J Biol Chem 1994; 269: Scheller A, Hughes E, Golden KL, Robins DM. Multiple receptor domains interact to permit, or restrict, androgen-specific gene activation. J Biol Chem 1998; 273: Lin CS, Lau A, Yeh CC, Chang CH, Lue TF. Upregulation of L-plastin gene by testosterone in breast and prostate cancer cells: identification of three cooperative androgen receptor-binding sequences. DNA Cell Biol 2000; 19: Bolton EC, So AY, Chaivorapol C, Haqq CM, Li H, Yamamoto KR. Cell- and genespecific regulation of primary target genes by the androgen receptor. Genes Dev 2007; 21: Massie CE, Adryan B, Barbosa-Morais NL, Lynch AG, Tran MG, Neal DE et al. New androgen receptor genomic targets show an interaction with the ETS1 transcription factor. EMBO Rep 2007; 8: Morelli A, Filippi S, Mancina R, Luconi M, Vignozzi L, Marini M et al. Androgens regulate phosphodiesterase type 5 expression and functional activity in corpora cavernosa. Endocrinology 2004; 145: Zhang XH, Morelli A, Luconi M, Vignozzi L, Filippi S, Marini M et al. Testosterone regulates PDE5 expression and in vivo responsiveness to Tadalafil in rat corpus cavernosum. Eur Urol 2005; 47: Zhang XH, Filippi S, Morelli A, Vignozzi L, Luconi M, Donati S et al. Testosterone restores diabetes-induced erectile dysfunction and sildenafil responsiveness in two distinct animal models of chemical diabetes. J Sex Med 2006; 3: Filippi S, Morelli A, Sandner P, Fibbi B, Mancina R, Marini M et al. Characterization and functional role of androgen-dependent PDE5 activity in the bladder. Endocrinology 2007; 148: Harris DE, Warshaw DM, Periasamy M. Nucleotide sequences of the rabbit alphasmooth-muscle and beta-non-muscle actin mrnas. Gene 1992; 112: Traish AM, Toselli P, Jeong SJ, Kim NN. Adipocyte accumulation in penile corpus cavernosum of the orchiectomized rabbit: a potential mechanism for venoocclusive dysfunction in androgen deficiency. J Androl 2005; 26: Hayward SW, Baskin LS, Haughney PC, Foster BA, Cunha AR, Dahiya R et al. Stromal development in the ventral prostate, anterior prostate and seminal vesicle of the rat. Acta Anat (Basel) 1996; 155: Niu YJ, Ma TX, Zhang J, Xu Y, Han RF, Sun G. Androgen and prostatic stroma. Asian J Androl 2003; 5: Stacey P, Rulten S, Dapling A, Phillips SC. Molecular cloning and expression of human cgmp-binding cgmp-specific phosphodiesterase (PDE5). Biochem Biophys Res Commun 1998; 247: Lin CS, Lau A, Tu R, Lue TF. Expression of three isoforms of cgmp-binding cgmpspecific phosphodiesterase (PDE5) in human penile cavernosum. Biochem Biophys Res Commun 2000; 268: Agha AM, Taha RA. Sildenafil inhibits agonist-evoked rat uterine contractility: influence of guanylyl cyclase inhibition. Eur J Pharmacol 2001; 428: Rybalkin SD, Rybalkina IG, Feil R, Hofmann F, Beavo JA. Regulation of cgmpspecific phosphodiesterase (PDE5) phosphorylation in smooth muscle cells. J Biol Chem 2002; 277: Lin CS, Lin G, Lue TF. Isolation of two isoforms of phosphodiesterase 5 from rat penis. Int J Impot Res 2003; 15: Buhimschi CS, Garfield RE, Weiner CP, Buhimschi IA. The presence and function of phosphodiesterase type 5 in the rat myometrium. Am J Obstet Gynecol 2004; 190: Khan RN, Hamoud H, Warren A, Wong LF, Arulkumaran S. Relaxant action of sildenafil citrate (Viagra) on human myometrium of pregnancy. Am J Obstet Gynecol 2004; 191: Belmonte A, Ticconi C, Dolci S, Giorgi M, Zicari A, Lenzi A et al. Regulation of phosphodiesterase 5 expression and activity in human pregnant and non-pregnant myometrial cells by human chorionic gonadotropin. J Soc Gynecol Invest 2005; 12: Dolci S, Belmonte A, Santone R, Giorgi M, Pellegrini M, Carosa E et al. Subcellular localization and regulation of type-1c and type-5 phosphodiesterases. Biochem Biophys Res Commun 2006; 341: Ticconi C, Zicari A, Belmonte A, Realacci M, Rao ChV, Piccione E. Pregnancypromoting actions of HCG in human myometrium and fetal membranes. Placenta 2007; 28(Suppl A): S137 S Chiossi G, Costantine MM, Betancourt A, Hankins GD, Longo M, Saade GR et al. Does sildenafil citrate affect myometrial contractile response to nifedipine in vitro? Am J Obstet Gynecol 2010; 203: 252. e1 e5. 52 Armagan A, Kim NN, Goldstein I, Traish AM. Dose response relationship between testosterone and erectile function: evidence for the existence of a critical threshold. J Androl 2006; 27: Zhang X, Zang N, Wei Y, Yin J, Teng R, Seftel A et al. Testosterone regulates smooth muscle contractile pathways in the rat prostate: emphasis on PDE5 signaling. Am J Physiol Endocrinol Metab 2012; 302: E243 E Lin CS et al. PDE5 expression and function in the lower urinary tract: a critical review. Urology 2013; doi: /j.urology (in press). 55 Traish AM, Park K, Dhir V, Kim NN, Moreland RB, Goldstein I et al. Effects of castration and androgen replacement on erectile function in a rabbit model. Endocrinology 1999; 140: Palese MA, Crone JK, Burnett AL. A castrated mouse model of erectile dysfunction. J Androl 2003; 24: Rogers RS, Graziottin TM, Lin CS, Kan YW, Lue TF. Intracavernosal vascular endothelial growth factor (VEGF) injection and adeno-associated virus-mediated VEGF gene therapy prevent and reverse venogenic erectile dysfunction in rats. Int J Impot Res 2003; 15: Antonioli E, Cardoso AB, Carvalho HF. Effects of long-term castration on the smooth muscle cell phenotype of the rat ventral prostate. J Androl 2007; 28: Buttyan R, Ghafar MA, Shabsigh A. The effects of androgen deprivation on the prostate gland: cell death mediated by vascular regression. Curr Opin Urol 2000; 10: Tek M, Balli E, Cimen B, Efesoy O, Oğuz I, Cayan S. The effect of testosterone replacement therapy on bladder functions and histology in orchiectomized mature male rats. Urology 2010; 75: Albersen M, Lin G, Fandel TM, Zhang H, Qiu X, Lin CS et al. Functional, metabolic, and morphologic characteristics of a novel rat model of type 2 diabetes-associated erectile dysfunction. Urology 2011; 78: e471 e Zhou F, Xin H, Liu T, Li GY, Gao ZZ, Liu J et al. Effects of icariside II on improving erectile function in rats with streptozotocin-induced diabetes. J Androl 2012; 33: Liu WJ, Xin ZC, Xin H, Yuan YM, Tian L, Guo YL. Effects of icariin on erectile function and expression of nitric oxide synthase isoforms in castrated rats. Asian J Androl 2005; 7: Yang R, Huang YC, Lin G, Wang G, Hung S, Dai YT et al. Lack of direct androgen regulation of PDE5 expression. Biochem Biophys Res Commun 2009; 380: Lin CS, Chow S, Lau A, Tu R, Lue TF. Regulation of human PDE5A2 intronic promoter by camp and cgmp: identification of a critical Sp1-binding site. Biochem Biophys Res Commun 2001; 280: Kotera J, Fujishige K, Imai Y, Kawai E, Michibata H, Akatsuka H et al. Genomic origin and transcriptional regulation of two variants of cgmp-binding cgmpspecific phosphodiesterases. Eur J Biochem 1999; 262: Andric SA, Janjic MM, Stojkov NJ, Kostic TS. Testosterone-induced modulation of nitric oxide-cgmp signaling pathway and androgenesis in the rat Leydig cells. Biol Reprod 2010; 83: Muller D, Mukhopadhyay AK, Davidoff MS, Middendorff R. Cyclic GMP signaling in rat urinary bladder, prostate, and epididymis: tissue-specific changes with aging and in response to Leydig cell depletion. Reproduction 2011; 142: Shigehara K, Sugimoto K, Konaka H, Iijima M, Fukushima M, Maeda Y et al. Androgen replacement therapy contributes to improving lower urinary tract symptoms in patients with hypogonadism and benign prostate hypertrophy: a randomised controlled study. Aging Male 2011; 14: & 2013 Macmillan Publishers Limited International Journal of Impotence Research (2013), 81 85