- Email: [email protected]
Genetics of Erectile Dysfunction
Genetics of Erectile Dysfunction Natalya A. Lopushnyan and Kanchan Chitaley* From the Department of Urology, University of Washington, Seattle, Washington
Abbreviations and Acronyms ACE ⫽ angiotensin converting enzyme Ang ⫽ angiotensin CSMC ⫽ cavernous smooth muscle cell D ⫽ deletion ED ⫽ erectile dysfunction eNOS ⫽ endothelial NOS GWAS ⫽ genome wide association study Hcy ⫽ homocysteine I ⫽ insertion IGF ⫽ insulin-like growth factor IGFBP ⫽ IGF binding protein NO ⫽ nitric oxide NOS ⫽ NO synthase PDE5i ⫽ phosphodiesterase 5 inhibitor PR ⫽ progesterone receptor SHR ⫽ sex hormone receptor SMR ⫽ submandibular rat gene 1 SNP ⫽ single nucleotide polymorphism VEGF ⫽ vascular endothelial growth factor VNTR ⫽ variable number tandem repeat vsca1 ⫽ variable coding sequence protein A1 * Correspondence: 815 Mercer St, Box 358050, Seattle, Washington 98109 (telephone: 206-8975456; FAX: 206-897-5442; e-mail: kanchanc@uw. edu).
1676
www.jurology.com
Purpose: Erectile dysfunction affects 50% of men older than 40 years. Recently more attempts have been made to identify genetic predictors of this disease. We reviewed animal and human data on genes related to the development and increased risk of erectile dysfunction. Materials and Methods: A literature search was performed using the PubMed® database. Articles addressing genes involved in erectile dysfunction were evaluated. Results: The majority of studies used a candidate gene approach to investigate genetic polymorphisms of known pathways mediating erection/detumescence. Studies in human and animal models are available. Human studies often compared the frequency of a specifically predetermined genetic polymorphism in men with erectile dysfunction to that in matched controls in whom few genes were persistently replicated. Several gene expression profiling studies are available that targeted specific erectile dysfunction models. Currently, there are few human genome wide association studies of erectile dysfunction. Conclusions: Studies investigating the genetics of erectile dysfunction are mostly derived from animal models and candidate gene approaches. Candidate gene studies omit the greater portion of the genome, a problem that can be solved using a genome wide association study approach. The lack of persistently replicated results of candidate gene studies may be related to different patient ethnic backgrounds, variations in erectile dysfunction etiology and small sample sizes. Using strict inclusion/exclusion criteria for erectile dysfunction etiology and ethnicity in human studies may lead to improved understanding of the genetics of erectile dysfunction in specific populations. Key Words: penis, erectile dysfunction, genetics, genome, ethnology ACCORDING to the Massachusetts Male Aging study, more than 50% of males 40 years old or older will have ED of various severities.1 The prevalence of complete impotence triples in men between ages 40 and 70 years, affecting multiple domains of quality of life of patients and their partners. While ED is more common in the aging population, patients with metabolic syndrome, diabetes, hypertension and obesity are also at increased risk.2
Improved understanding of penile erection and detumescence has provided us with the mainstay of ED pharmaceutical therapy, PDE5is. These agents, initially introduced in the mid 1990s, enhance NO induced signaling and are generally the first line treatment for ED. Other common treatments, such as intracavernous injections of prostaglandin or ␣-adrenergic antagonists, induce CSMC relaxation. While these pharmaceuticals improve erection in some men, they do not
0022-5347/12/1885-1676/0 THE JOURNAL OF UROLOGY® © 2012 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION
http://dx.doi.org/10.1016/j.juro.2012.07.008 Vol. 188, 1676-1683, November 2012 RESEARCH, INC. Printed in U.S.A.
AND
GENETICS OF ERECTILE DYSFUNCTION
cure the underlying problem and men must often continue medication, usually increasing the dose with time. Recently, effort has been directed toward identifying the genetic origin of impotence. Investigating ED in twins suggested a heritability component of at least 35%.2 To date most studies have focused on the NOS pathway and proteins composing ion channels. The first human trial of gene therapy was performed to attempt to increase the expression of ion channels in penile smooth muscle.3 This therapy represents a first attempt at curing the disease by modifying genetic material, rather than improving symptoms. The role of genes in ED presentation and severity is not completely understood. Most groups have used a candidate gene approach, identifying genetic polymorphisms in a pathway known to be involved in penile erection. Several groups have looked at gene expression in specific ED models. Significantly fewer groups have investigated novel gene polymorphisms or compared whole genomes of patients with ED to those of matched controls. The primary goal of this review was to discuss findings relating to the genetic basis of ED and establish possible directions for future research in this area.
MATERIALS AND METHODS In July 2011 a PubMed database search was performed for articles published between 1970 and the present. The terms used were “genes,” “erectile dysfunction,” “GWAS,”
1677
“genetic polymorphism,” “microarray,” “impotence” and combinations. The search primarily focused on original articles.
CANDIDATE GENE STUDIES In 1975 one of the first studies was published to demonstrate that many proteins have various heritable isoforms, showing that common genetic variations could lead to a change in protein structure and biological function. These genetic polymorphisms can be tested for their association with a disease by determining the genotype of these variants in affected individuals vs those without the disease. If a genotype of interest is present at a higher frequency in cases, it is taken as evidence that the variation is associated with an increased risk of the disease. The majority of studies published in the field of ED genetics are candidate gene studies that take the approach from known to unknown and allow for a focus on a particular gene with a known pathological involvement in the disease (table 1). Nitric Oxide Synthase One of the better studied candidate genes in the penile smooth muscle relaxation pathway is eNOS. To date 3 forms of NOS, originally labeled after their location, have been recognized, including neuronal NOS, eNOS and inducible NOS. It was hypothesized that rapid activation of neuronal NOS initiates the erectile process, whereas phosphorylation and activation of eNOS lead to sustained erection.4 The en-
Table 1. Genetic polymorphisms previously investigated in ED development References eNOS: Erol et al5 Lee et al6 Sinici et al4 Andersen et al7 Erkan et al9 Rosas-Vargas et al12 Safarinejad et al16 Serrano et al10 Park et al15 Slo (Davies et al17) Vcsa1 (Tong et al34) ACE: Park et al15 Mazo et al23 Andersen et al25 Eisenhardt et al13 PR (Andersen et al28) Androgen receptor (Andersen et al29) Hcy (Safarinejad et al30) IGFBP-3: Pu et al45 Soh et al46 Transforming growth factor-1 (Peters et al42)
Polymorphism, Allele, Splice Variant
Rat Model*
Ethnic Group*
7G894T, 4VNTR 7G894T T-786C 7G894T 4VNTR 7G894T T786C, 7G894T, 4a/4b T786C, 7G894T, 4VNTR 4VNTR SVcyt Not applicable
Not applicable Not applicable Not applicable Not applicable Not applicable Not applicable Not applicable Not applicable Not applicable Diabetic Diabetic, aging ⫹ neurogenic
Turkish Taiwanese Turkish Brazilian Turkish Mexican Mestizo Iranian Colombian Korean Not applicable Not applicable
II, DI, DD II, DI, DD II, DI, DD II, DI, DD PROGINS CAG repeat C677T, A1298C, G1793A Not applicable
Not Not Not Not Not Not Not
Korean Russian Brazilian German Brazilian Brazilian Iranian Not applicable
Various
* Most studies were done in specific animal model or in cohort of men of specific ethnicity.
applicable applicable applicable applicable applicable applicable applicable
Aging Diabetic Not applicable
Various
1678
GENETICS OF ERECTILE DYSFUNCTION
dothelial and neuronal forms of the NOS enzyme are encoded by different genes with eNOS mapped to chromosome 7q35 and consisting of 26 exons spanning 21 kb of DNA. Several polymorphisms of eNOS have been investigated. More frequently, investigated regions of this gene include a variable number of 27 bp tandem repeats in intron 4 (4VNTR), G894T (rs1799983) polymorphism in exon 7 and a T-786C (rs2070744) polymorphism in the promoter region.4 –7 The eNOS gene has 2 common alleles containing 4 VNTRs (allele a) and 5 VNTRs (allele b), which make 3 possible genotypes (aa, ab and bb). The biological function of the 4a/4b polymorphism is controversial. In the past it was associated with decreased eNOS concentration and activity, although no decrease in plasma NO levels was reported.8 Also, eNOS aa, ab and bb genotypes have a similar distribution in patients with ED and controls.9 The eNOS G894T polymorphism, which corresponds to a substitution of glutamate by aspartate at amino acid position 298, results in the disturbance of eNOS activity and NO production.7 The T-786C region results in a significant decrease in eNOS gene promoter activity.10 Human studies correlating these polymorphisms with ED prevalence are controversial. A limitation is that most available studies were performed in small groups of patients of various ethnic backgrounds. While the 7G894T polymorphism was noted to be an independent factor for ED in Mexican, Turkish and Taiwanese populations,5,6,11,12 findings do not support a role of this polymorphism in the Brazilian or German population.7,13 A recent meta-analysis by Wang et al showed a significant association of ED and the G894T polymorphism (generalized OR 1.64, 95% CI 1.30 –2.61) but failed to demonstrate the same correlation with 4VNTR (OR 0.96).14 4VNTR (allele a) and the distribution of eNOS genotypes in ED was first investigated in the Korean population in 1999 by Park et al.15 They found no significant difference in the distribution in the ED group compared to controls. Consistently, 4VNTR has not predisposed men to ED in a Turkish or Iranian population.5,16 T-789C is an independent predisposing factor for ED in Iranian and Turkish men.4,16 However, data supporting its role in other populations are lacking. eNOS polymorphisms occur at different frequencies among various ethnic groups.10 This difference in distribution could contribute to the lack of well replicated results across patient populations of different ethnicities. In the future it could potentially be improved by comparing the distribution of eNOS polymorphisms in larger cohorts of patients with ED vs controls across various ethnic backgrounds.
Ion Channels Penile erection and detumescence are partly a function of the smooth muscle cells of the cavernosum and arterial wall, which are regulated by changes in free cytosolic calcium. Ca2⫹ channels can be regulated by currents generated through K⫹ channels.17 These channels allow K⫹ to flow down its gradient and out of the smooth muscle cell. The hyperpolarization that occurs limits intracellular Ca2⫹ entry and results in smooth muscle relaxation. The Maxi-K or BK channel (large performance, calcium sensitive K channel) is the most studied ion channel in the field of ED genetics and physiology that was previously shown to have a role in penile erection.18 However, there are limited data on intrinsic expression of the Maxi-K gene or its variants in men with ED. The pore forming subunit of this channel is encoded by the slo gene. In slo knockout mice cavernous tissue relaxation was decreased by 50% and intracavernous pressure was attenuated in response to nerve stimulation.18 The slo gene is known to undergo alternative splicing triggered by various stimuli, including hormones, leading to several different isoforms.19 In streptozotocin induced diabetic animals alternative splicing led to up-regulation of the SV0 channel forming transcript compared to controls. Also, a higher proportion of SV0 was reported in men with vs without diabetes, suggesting that alternative splicing of the slo gene may have a role in diabetic ED.17 Angiotensin Converting Enzyme I/D The presence of ACE, Ang II and Ang II receptor in the corpora cavernosa has been confirmed. The corpus cavernosum produces physiologically relevant amounts of Ang II and tissues with higher Ang II levels appear to have a decreased response to papaverine and prostaglandin E-1.20 Increased Ang II in cavernous blood was noted during penile detumescence vs rigidity and in patients with organogenic ED vs controls.20 ACE, which converts Ang I into Ang II, is encoded by the ACE gene containing a polymorphism based on the presence (I) or absence (D) in an intron of a 287 bp nonsense DNA domain (rs4646994). This results in 3 possible genotypes, including DD, DI or II.15 An ACE polymorphism accounts for 47% of the total phenotypic variance of serum ACE levels with a higher concentration in the II genotype.21 Identification of homologous sequences in gene ontology annotated databases suggested that the DD genotype lacks molecular and biological function.22 One of the first studies of ACE polymorphisms related to ED was reported by Park et al.15 They examined the relationship between gene polymorphism and organic ED. In the ED group the proportion of the DD genotype was significantly higher
GENETICS OF ERECTILE DYSFUNCTION
1679
than in controls (54% vs 24%, p ⬍0.01). Consistent findings were published in Russian men with metabolic syndrome, with a higher frequency of the D allele and the DD genotype in the ED group vs controls (81.3% vs 63.8%, p ⬍0.001, and 68.3% vs 38.2%, respectively).23 The opposite was seen in the Brazilian population with increased frequency of the I allele in individuals complaining of ED.21 In a study of German patients with ED compared with controls there was no difference in the ACE I/D genotype.13 However, that study included patients with ED attributable to various causes, including vascular, neurogenic, psychogenic, hormonal and mixed etiologies, as well as 11% of men for whom no specific ED etiology could be identified. A lack of association was also noted between ACE genotypes in the Korean population.24 It is possible that the lack of consistent replication in the reports outlined is due to the study of diverse ethnic backgrounds, ED etiologies and comorbidities. Potentially, a larger sample size of patients with a single ED etiology would shed more light on the genetic correlation of polymorphisms and the predisposition to ED.
ED complaints. While sex hormones certainly have an important role in sexual function, currently there is a lack of evidence that genetic polymorphisms of SHR predispose to ED development.
Hormone Receptors Sex hormones have various effects on sexual function, including NOS activity, erectile frequency and the sexual behavior of male rodents.25,26 However, data are limited on possible genetic variations in sex hormone and SHR genes. Genetic expression of androgen, progesterone and estrogen receptor  decreases dramatically in the penile crura of aging rats, suggesting that SHR may have a role in age related ED.26 However, no difference in SHR polymorphisms was found in 2 population based reports. In the first report a PROGINS polymorphism of the PR gene was investigated.27 It consists of a 306 bp Alu insertion in intron G, a silent point mutation in exon 5 and a change at position 660 from valine to leucine (V660L) (rs1042838). In the past the PROGINS variant has been associated with a decreased response to progesterone due to reduced protein activity. Decreased PR activity could potentially alter erectile function. However, in the report by Andersen et al there was no difference in the frequency of PROGINS in an ED vs a nonED group.28 The limitation of the report was that only 1.2% of subjects were homozygous for Alu insertion, possibly making the sample size inadequate to detect a difference. Another study of the CAG repeat sequence of the androgen receptor gene was reported by Andersen et al.29 Longer CAG sequences were associated with decreased transcriptional activity, presumably correlating with relatively reduced androgen feedback. However, there was no association with increased
GENE EXPRESSION PROFILING
Hcy Metabolism Genes involved in the Hcy pathway have been investigated in relation to ED. The methylenetetrahydrofolate reductase gene, located on chromosome 1, is a key enzyme in folic acid metabolism. Three polymorphisms of this gene are known, including C677T (rs1801133), A1298C (rs1801131) and G1793A (rs2274976), which result in decreased enzyme activity and increased Hcy levels.30 An increase in Hcy is linked to impaired endothelial dysfunction and vascular disease. Recently it was reported that patients with early onset vasculogenic ED had higher levels of serum Hcy and a higher prevalence of C677T polymorphism compared to controls (15.8% vs 22.4%, p ⫽ 0.01).30 Additionally, patients homozygous for the C677T allele have a decreased response to PDE5i alone but improve with combination treatment of PDE5i, folate and vitamin B6.31
While the candidate gene approach provides the benefit of knowing the function of the investigated gene, it is unlikely to detect variation in gene expression. For that purpose gene expression profiling becomes useful. It is an effective tool for characterizing gene expression among various ED models and in ED vs healthy controls. Although they do not necessarily directly relate to genetic heritability, methods of gene expression profiling allow us to acquire quantitative data on the level of activity of multiple genes. The majority of all reports available are derived from rodents. One of the first groups reported altered expression of 126 genes in the neurogenic impotence model. One of the down-regulated genes was SMR1, a member of the opiorphin family, which is expressed at a significantly higher level in male than in female rats. Later, it was also suggested as an ED marker in candidate studies.32 In humans other genes from the same family, ProL1, hSMR3A and hSMR3B, are significantly down-regulated in men with ED with and without diabetes.33 The rat homologue of opiorphin, sialorphin, which is encoded by the vsca1 gene, was suggested as an ED marker in diabetic, neurogenic and aging animal models.34 Its product is a neutral endopeptidase inhibitor, which is a membrane bound peptidase that decreases peptide agonist action by removing them from membrane receptors. Substantial vsca1 up-regulation was noted following intracorporeal injection of the Slo gene containing vector, a pore forming
1680
GENETICS OF ERECTILE DYSFUNCTION
subunit of the BK channel.32 It was hypothesized that Vcsa1 may improve CSMC relaxation by neutral endopeptidase inhibition, thus, prolonging peptide agonist action at the receptors.35 Using microarray technology in diabetic rats, 529 genes were identified with differential expression.36 That group took the study 1 step further and identified genes that were previously reported to mediate vascular dysfunction. One of these genes, ceruloplasmin, and its 2 splice variants were up-regulated in diabetic rat CSMCs and endothelium. Previous reports demonstrated that ceruloplasmin impairs aortic relaxation, possibly altering NO bioavailability in the penis.37 Aberrant regulation of growth factors is often implicated in vasculogenic ED. Several penile growth factors have been identified. Originally IGF-1 and VEGF were found to promote CSMC proliferation in rats.38 One of the first groups to report these growth factors in human CSMCs was Rajasekaran et al.39 Using reverse transcriptase-polymerase chain reaction they identified 4 splice variants of VEGF mRNA, including VEGFs 121, 145, 165 and 189, with VEGFs 121 and 165 most abundant. In that study a twofold proliferation of CSMCs was seen with exposure to VEGF. In hypercholesterolemic rats with ED significantly lower VEGF gene expression and lower levels of phosphorylated eNOS were shown.40 Improved intracavernous pressure in diabetic animals transfected with the VEGF gene was also observed.41 In humans specific polymorphisms of VEGF and transforming growth factor-1 are related to ED development following radiation therapy for prostate cancer.42 IGF and its receptors have varied expression in ED. IGF-1 is a polypeptide structurally homologous to proinsulin that facilitates the regeneration of NOS containing nerve fibers.43 After pudendal artery ligation in rats, up-regulation of IGFBP-1, 3 and 5, and down-regulation of IGFBP-6 was seen.44 In the same report the most highly induced gene after pudendal artery ligation was apolipoprotein D. IGFBP-3 mRNA expression levels were increased in aging and diabetic rat models.45,46 It is possible that increased expression levels of IGFBP-3 decrease the availability of IGF-1 in penile tissue, contributing to ED in patients with diabetes. This is consistent with the report that IGF-1 injection improved erectile function in aging animals.45 While to our knowledge it is unknown whether there is a difference in VEGF or IGF-1 gene expression in nondiabetic, nonvasculogenic ED, they could possibly provide the basis for future gene therapy in specific patient populations. Gene expression profiling studies, including microarray analysis, permits analysis of the gene network interactions to identify the most highly affected components. However, frequently they are used to confirm involvement of a gene of interest or
as a starting point for candidate gene studies. This method of analysis also does not account for alternative gene splicing, a major mechanism for generating protein diversity, or for the protein expression level, which may not correlate with gene expression in a direct manner. An adjunct solution may be proteomic analysis to look at the expression of the protein rather than the gene. While those limitations exist, these studies can still provide additional information to the field of ED genetics.
GENOME WIDE ASSOCIATION STUDIES/SNP Candidate gene studies target pathways known to contribute to ED. However, they ignore most of the genome and are likely to miss many causal regions of genes or show many false-positive associations. Results of candidate gene studies are also not often easily and consistently reproduced. Gene expression studies investigate variations in gene transcription for a given condition but do not identify specific polymorphisms. GWASs improve on those shortcomings, allowing comparison of common genetic variants over the entire genome in a large number of cases to those in unaffected controls. However, there are still only a few reports available (table 2). Kerns et al identified SNPs associated with ED development in black American men.47 In the 27 men who qualified for analysis Kerns et al identified SNP rs2268363 as significantly associated with ED after multiple comparison correction (unadjusted p ⫽ 5.46 ⫻ 10⫺8, Bonferroni p ⫽ 0.028) and an additional 4 SNPs that trended toward a significant association (unadjusted p ⬍10⫺6). SNP rs2268363 is located on chromosome 2 in the follicle stimulating hormone receptor, which is involved in testis development and function. Alterations in the follicle stimulating hormone signaling pathway lead to improper gonad development in mice, and to small testis size and infertility in humans. Since to our knowledge this is the only report implicating SNP rs2268363 as possibly involved in ED, a candidate gene study may be warranted. Two more abstracts were recently presented by groups that investigated ED genetics using a GWAS Table 2. SNPs identified in GWASs in patients with ED References Kerns et al47 Hotaling et al48 Hotaling et al49
SNP
Population
Unadjusted p Value*
Black American men 5.46 ⫻ 10⫺8 Diabetes Cohort ⫹ 7 ⫻ 10⫺7, Complications Trial 9 ⫻ 10⫺7 SNP on chromosome 14 Adult Changes in 1.1 ⫻ 10⫺6 Thought rs2268363 rs9810233, rs1920201
* Standard GWAS significance criteria considered at p ⬍5 ⫻ 10⫺8.
GENETICS OF ERECTILE DYSFUNCTION
approach. The first study specifically investigated type 1 diabetes related ED using the Diabetes Control and Complications Trial cohort.48 Two SNPs located on chromosome 3 near the activated leukocyte cell adhesion molecule gene were identified, approaching the standard GWAS significance criterion of p ⬍5 ⫻ 10⫺8. Expression of activated leukocyte cell adhesion molecule, a member of the Ig superfamily, was noted in various cultured human endothelial cells and is also thought to have a role in the inflammatory response. Another series included male patients with and without ED from the Group Health Adult Changes in Thought study.49 After adjustment for smoking, diabetes and hypertension, an SNP on chromosome 14 near the ADCK1 gene was identified to have the strongest association with ED (p ⫽ 1.1E-06). ADCK1 is a highly conserved gene implicated in coenzyme Q biosynthesis, a part of the electron transport chain of the inner mitochondrial membrane. To our knowledge none of the mentioned findings in human studies have been replicated to date. Although GWASs allow analysis of the entire genome, the identified SNPs mostly do not have an obvious association with ED, making the causal variant identification challenging.50 The question then arises of whether the SNP identified represents an association as a proxy of a causal variant with a similar frequency, or an association of the causal variant via linkage disequilibrium. Linkage disequilibrium is present when some combination of alleles or genetic markers is present at a higher or lower frequency in the population than would be expected based on their random combination frequency. Detecting such a synthetic association would require extensive resequencing.50 While GWASs allow simultaneous testing of thousands of SNPs, they raise a statistical challenge of accounting for multiple comparisons. Failure to do so results in decreased power, falsepositive results and incorrect rejection of the null hypothesis, also influencing the publication bias that favors the dissemination of positive results. In addition to those statistical challenges, which are beyond the scope of this report, GWASs require sample sizes in the thousands to detect a modest association. Achievement of this power is particularly challenging while evaluating a multifactorial disease such as ED in a population with a specific phenotype.
CONCLUSIONS Multiple genetic variants potentially associated with ED have been found. Most of them are a product of candidate gene studies stemming from the
1681
pathophysiological basis of ED. Few results have been consistently replicated across various ED models. Currently, the most commonly used ED models in the literature are vasculogenic, neurogenic and aging. What may work as a genetic marker in one model may not be applicable in another model. Gene association studies have identified various gene polymorphisms thought to be etiological contributors or genetic markers of ED. However, few of them have been successfully replicated across different populations, possibly due to variations in ethnic background, sample size, different study inclusion criteria, confounding by the population substructure, potential false-positive associations and underpowered, nonsignificant studies of real associations. ED is a complex disease and, like most diseases with complex traits, it likely results from the interaction of multiple genetic and environmental factors. While the central goal of genetic ED research is to identify genetic factors and their effect of ED development, it would be most desirable to simultaneously consider gene-gene and gene-environment interaction. Gene expression profiling identifies variations in set ED models. However, it does not detect novel polymorphisms or account for post-transcriptional modification. While it is currently a useful adjunct, it could potentially be supplemented further using proteomic analysis. The possible genetic associations that are overlooked by candidate gene studies could be investigated using a GWAS. However, currently only a few reports are available that analyzed a limited number of patients. To determine whether the result of an association study represents a true association would require large sample sizes, an appropriate statistical approach with correction for multiple hypothesis testing and replication of the results in a separate but phenotypically close population. The ultimate goal of the search for the genetic origin or association of ED is the creation of better treatment and possibly cure. The search for genetic associations of ED will likely lead to the discovery of 2 groups of genes, including those that contribute to the development of or protection from impotence and those that are markers for it. Genes belonging to the former group could lead to advances in treatment and specifically tailored therapies based on the genetic makeup of the patient, such as local introduction of a gene encoding for a specific enzyme or ion channel. With the latter the prevention and identification of ED before it is clinically evident and the development of preventive strategies will likely be possible.
1682
GENETICS OF ERECTILE DYSFUNCTION
REFERENCES 1. Avasthi A, Grover S, Bhansali A et al: Erectile dysfunction in diabetes mellitus contributes to poor quality of life. Int Rev Psychiatry 2011; 23: 93. 2. Fischer ME, Vitek ME, Hedeker D et al: A twin study of erectile dysfunction. Arch Intern Med 2004; 164: 165. 3. Melman A, Biggs G, Davies K et al: Gene transfer with a vector expressing Maxi-K from a smooth muscle-specific promoter restores erectile function in the aging rat. Gene Ther 2008; 15: 364. 4. Sinici I, Güven EO, Serefog˘lu E et al: T-786C polymorphism in promoter of eNOS gene as genetic risk factor in patients with erectile dysfunction in Turkish population. Urology 2010; 75: 955. 5. Erol B, Bozdogan G, Akduman B et al: eNOS gene intron 4 VNTR and exon 7-G894T polymorphisms in Turkish men with erectile dysfunction: a case control study. J Sex Med 2009; 6: 1423. 6. Lee YC, Wu WJ, Liu CC et al: The associations among eNOS G894T gene polymorphism, erectile dysfunction, and benign prostate hyperplasia-related lower urinary tract symptoms. J Sex Med 2009; 6: 3158.
dysfunction: a meta-analysis. J Sex Med 2010; 7: 3889. 15. Park JK, Kim W, Kim SW et al: Gene-polymorphisms of angiotensin converting enzyme and endothelial nitric oxide synthase in patients with erectile dysfunction. Int J Impot Res 1999; 11: 273. 16. Safarinejad MR, Khoshdel A, Shekarchi B et al: Association of the T-786C, G894T and 4a/4b polymorphisms of the endothelial nitric oxide synthase gene with vasculogenic erectile dysfunction in Iranian subjects. BJU Int 2011; 107: 1994. 17. Davies KP, Zhao W, Tar M et al: Diabetes-induced changes in the alternative splicing of the slo gene in corporal tissue. Eur Urol 2007; 52: 1229. 18. Werner ME, Zvara P, Meredith AL et al: Erectile dysfunction in mice lacking the large-conductance calcium-activated potassium (BK) channel. J Physiol 2005; 567: 545. 19. Lai GJ and McCobb DP: Opposing actions of adrenal androgens and glucocorticoids on alternative splicing of Slo potassium channels in bovine chromaffin cells. Proc Natl Acad Sci U S A 2002; 99: 7722.
7. Andersen ML, Guindalini C, Santos-Silva R et al: Association analysis of endothelial nitric oxide synthase G894T gene polymorphism and erectile dysfunction complaints in a population-based survey. J Sex Med 2010; 7: 1229.
20. Becker AJ, Uckert S, Stief CG et al: Plasma levels of angiotensin II during different penile conditions in the cavernous and systemic blood of healthy men and patients with erectile dysfunction. Urology 2001; 58: 805.
8. Song J, Yoon Y, Park KU et al: Genotype-specific influence on nitric oxide synthase gene expression, protein concentrations, and enzyme activity in cultured human endothelial cells. Clin Chem 2003; 49: 847.
21. Andersen ML, Guindalini C, Santos-Silva R et al: Angiotensin-converting enzyme polymorphism and erectile dysfunction complaints in the Brazilian population. J Sex Med 2010; 7: 2791.
9. Erkan E, Muslumanoglu AY, Oktar T et al: Polymorphism of endothelial nitric oxide synthase gene in patients with erectile dysfunction. J Sex Med 2006; 3: 69. 10. Serrano NC, Díaz LA, Casas JP et al: Frequency of eNOS polymorphisms in the Colombian general population. BMC Genet 2010; 11: 54. 11. Lee YC, Huang CH, Wang CJ et al: The associations among eNOS G894T gene polymorphism, erectile dysfunction and related risk factors. BJU Int 2007; 100: 1116. 12. Rosas-Vargas H, Coral-Vazquez RM, Tapia R et al: Glu298Asp endothelial nitric oxide synthase polymorphism is a risk factor for erectile dysfunction in the Mexican Mestizo population. J Androl 2004; 25: 728. 13. Eisenhardt A, Stief C, Porst H et al: Genetic association study of the GNB3 C825T, the ACE I/D and the eNOS G894T polymorphisms and the risk to develop erectile dysfunction in a German ED population. Andrologia 2010; 42: 218. 14. Wang JL, Wang HG, Gao HQ et al: Endothelial nitric oxide synthase polymorphisms and erectile
22. Wiwanitkit V: I/D genetic polymorphism of angiotensin-converting enzyme: pathogenesis evaluation for erectile dysfunction by gene ontology. Fertil Steril 2008; 89: 1095. 23. Mazo EB, Gamidov SI, Mamedov MN et al: Association between the insertion/deletion polymorphism of the angiotensin-converting enzyme gene and erectile dysfunction in patients with metabolic syndrome. Int J Impot Res 2008; 20: 68. 24. Kim DS, Choi SI, Lee HS et al: Determination of human angiotensin converting enzyme (ACE) gene polymorphisms in erectile dysfunction: frequency differences of ACE gene polymorphisms according to the method of analysis. Clin Chem Lab Med 2001; 39: 11. 25. Andersen ML, Martins RC, Alvarenga TA et al: Progesterone reduces erectile dysfunction in sleep-deprived spontaneously hypertensive rats. Reprod Biol Endocrinol 2007; 5: 7. 26. Shirai M, Yamanaka M, Shiina H et al: Downregulation of androgen, estrogen and progesterone receptor genes and protein is involved in aging-related erectile dysfunction. Int J Impot Res 2003; 15: 391.
27. Romano A, Delvoux B, Fischer DC et al: The PROGINS polymorphism of the human progesterone receptor diminishes the response to progesterone. J Mol Endocrinol 2007; 38: 331. 28. Andersen ML, Alvarenga TA, Mazzotti DR et al: Hormonal profile, the PROGINS polymorphism, and erectile dysfunction complaints: data from a population-based survey. Fertil Steril 2011; 95: 621. 29. Andersen ML, Guindalini C, Santos-Silva R et al: Androgen receptor CAG repeat polymorphism is not associated with erectile dysfunction complaints, gonadal steroids and sleep parameters: data from a population-based survey. J Androl 2010; 32: 524. 30. Safarinejad MR, Safarinejad S and Shafiei N: Role of methylenetetrahydrofolate reductase gene polymorphisms (C677T, A1298C, and G1793A) in the development of early onset vasculogenic erectile dysfunction. Arch Med Res 2010; 41: 410. 31. Lombardo F, Tsamatropoulos P, Piroli E et al: Treatment of erectile dysfunction due to C677T mutation of the MTHFR gene with vitamin B6 and folic acid in patients non responders to PDE5i. J Sex Med 2010; 7: 216. 32. Calenda G, Tong Y, Tar M et al: Vcsa1 acts as a marker of erectile function recovery after gene therapeutic and pharmacological interventions. J Urol 2009; 181: 2806. 33. Tong Y, Tar M, Melman A et al: The opiorphin gene (ProL1) and its homologues function in erectile physiology. BJU Int 2008; 102: 736. 34. Tong Y, Tar M, Davelman F et al: Variable coding sequence protein A1 as a marker for erectile dysfunction. BJU Int 2006; 98: 396. 35. Davies KP: The role of opiorphins (endogenous neutral endopeptidase inhibitors) in urogenital smooth muscle biology. J Sex Med, suppl., 2009; 6: 286. 36. Sullivan CJ, Teal TH, Luttrell IP et al: Microarray analysis reveals novel gene expression changes associated with erectile dysfunction in diabetic rats. Physiol Genomics 2005; 23: 192. 37. Cappelli-Bigazzi M, Ambrosio G, Musci G et al: Ceruloplasmin impairs endothelium-dependent relaxation of rabbit aorta. Am J Physiol 1997; 273: H2843. 38. Liu X, Lin CS, Graziottin T et al: Vascular endothelial growth factor promotes proliferation and migration of cavernous smooth muscle cells. J Urol 2001; 166: 354. 39. Rajasekaran M, Kasyan A, Allilain W et al: Ex vivo expression of angiogenic growth factors and their receptors in human penile cavernosal cells. J Androl 2003; 24: 85. 40. Ryu JK, Shin HY, Song SU et al: Downregulation of angiogenic factors and their downstream target molecules affects the deterioration of erectile
GENETICS OF ERECTILE DYSFUNCTION
function in a rat model of hypercholesterolemia. Urology 2006; 67: 1329. 41. Gou X, He WY, Xiao MZ et al: Transplantation of endothelial progenitor cells transfected with VEGF165 to restore erectile function in diabetic rats. Asian J Androl 2011; 13: 332. 42. Peters CA, Stock RG, Cesaretti JA et al: TGFB1 single nucleotide polymorphisms are associated with adverse quality of life in prostate cancer patients treated with radiotherapy. Int J Radiat Oncol Biol Phys 2008; 70: 752. 43. Jung GW, Kwak JY, Yoon S et al: IGF-I and TGF-beta2 have a key role on regeneration of nitric oxide synthase (NOS)-containing nerves after cavernous neurotomy in rats. Int J Impot Res 1999; 11: 247.
44. Lin CS, Ho HC, Gholami S et al: Gene expression profiling of an arteriogenic impotence model. Biochem Biophys Res Commun 2001; 285: 565.
1683
can men after radiotherapy for prostate cancer. Int J Radiat Oncol Biol Phys 2010; 78: 1292.
45. Pu XY, Zheng XG, Zhang Y et al: Higher expression of mRNA and protein of insulin-like growth factor binding protein-3 in old rat penile tissues: implications for erectile dysfunction. J Sex Med 2011; 8: 2181.
48. Hotaling J, Waggott D, Patterson A et al: Genome wide association study identifies novel gene for erectile dysfunction risk using the Epidemiology of Diabetes Interventions and Complications (EDIC) database. J Urol, suppl., 2010; 183: e274, abstract 703.
46. Soh J, Katsuyama M, Ushijima S et al: Localization of increased insulin-like growth factor binding protein-3 in diabetic rat penis: implications for erectile dysfunction. Urology 2007; 70: 1019.
49. Hotaling J, Wessells H, Burt A et al: Novel single nucleotide polymorphisms may predict ED: data from the Group Health Adult Changes in Thought Study. J Urol, suppl., 2011; 185: e368, abstract 920.
47. Kerns SL, Ostrer H, Stock R et al: Genome-wide association study to identify single nucleotide polymorphisms (SNPs) associated with the development of erectile dysfunction in African-Ameri-
50. Takeuchi F, Kobayashi S, Ogihara T et al: Detection of common single nucleotide polymorphisms synthesizing quantitative trait association of rarer causal variants. Genome Res 2011; 21: 1122.
Abbreviations and Acronyms ACE ⫽ angiotensin converting enzyme Ang ⫽ angiotensin CSMC ⫽ cavernous smooth muscle cell D ⫽ deletion ED ⫽ erectile dysfunction eNOS ⫽ endothelial NOS GWAS ⫽ genome wide association study Hcy ⫽ homocysteine I ⫽ insertion IGF ⫽ insulin-like growth factor IGFBP ⫽ IGF binding protein NO ⫽ nitric oxide NOS ⫽ NO synthase PDE5i ⫽ phosphodiesterase 5 inhibitor PR ⫽ progesterone receptor SHR ⫽ sex hormone receptor SMR ⫽ submandibular rat gene 1 SNP ⫽ single nucleotide polymorphism VEGF ⫽ vascular endothelial growth factor VNTR ⫽ variable number tandem repeat vsca1 ⫽ variable coding sequence protein A1 * Correspondence: 815 Mercer St, Box 358050, Seattle, Washington 98109 (telephone: 206-8975456; FAX: 206-897-5442; e-mail: kanchanc@uw. edu).
1676
www.jurology.com
Purpose: Erectile dysfunction affects 50% of men older than 40 years. Recently more attempts have been made to identify genetic predictors of this disease. We reviewed animal and human data on genes related to the development and increased risk of erectile dysfunction. Materials and Methods: A literature search was performed using the PubMed® database. Articles addressing genes involved in erectile dysfunction were evaluated. Results: The majority of studies used a candidate gene approach to investigate genetic polymorphisms of known pathways mediating erection/detumescence. Studies in human and animal models are available. Human studies often compared the frequency of a specifically predetermined genetic polymorphism in men with erectile dysfunction to that in matched controls in whom few genes were persistently replicated. Several gene expression profiling studies are available that targeted specific erectile dysfunction models. Currently, there are few human genome wide association studies of erectile dysfunction. Conclusions: Studies investigating the genetics of erectile dysfunction are mostly derived from animal models and candidate gene approaches. Candidate gene studies omit the greater portion of the genome, a problem that can be solved using a genome wide association study approach. The lack of persistently replicated results of candidate gene studies may be related to different patient ethnic backgrounds, variations in erectile dysfunction etiology and small sample sizes. Using strict inclusion/exclusion criteria for erectile dysfunction etiology and ethnicity in human studies may lead to improved understanding of the genetics of erectile dysfunction in specific populations. Key Words: penis, erectile dysfunction, genetics, genome, ethnology ACCORDING to the Massachusetts Male Aging study, more than 50% of males 40 years old or older will have ED of various severities.1 The prevalence of complete impotence triples in men between ages 40 and 70 years, affecting multiple domains of quality of life of patients and their partners. While ED is more common in the aging population, patients with metabolic syndrome, diabetes, hypertension and obesity are also at increased risk.2
Improved understanding of penile erection and detumescence has provided us with the mainstay of ED pharmaceutical therapy, PDE5is. These agents, initially introduced in the mid 1990s, enhance NO induced signaling and are generally the first line treatment for ED. Other common treatments, such as intracavernous injections of prostaglandin or ␣-adrenergic antagonists, induce CSMC relaxation. While these pharmaceuticals improve erection in some men, they do not
0022-5347/12/1885-1676/0 THE JOURNAL OF UROLOGY® © 2012 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION
http://dx.doi.org/10.1016/j.juro.2012.07.008 Vol. 188, 1676-1683, November 2012 RESEARCH, INC. Printed in U.S.A.
AND
GENETICS OF ERECTILE DYSFUNCTION
cure the underlying problem and men must often continue medication, usually increasing the dose with time. Recently, effort has been directed toward identifying the genetic origin of impotence. Investigating ED in twins suggested a heritability component of at least 35%.2 To date most studies have focused on the NOS pathway and proteins composing ion channels. The first human trial of gene therapy was performed to attempt to increase the expression of ion channels in penile smooth muscle.3 This therapy represents a first attempt at curing the disease by modifying genetic material, rather than improving symptoms. The role of genes in ED presentation and severity is not completely understood. Most groups have used a candidate gene approach, identifying genetic polymorphisms in a pathway known to be involved in penile erection. Several groups have looked at gene expression in specific ED models. Significantly fewer groups have investigated novel gene polymorphisms or compared whole genomes of patients with ED to those of matched controls. The primary goal of this review was to discuss findings relating to the genetic basis of ED and establish possible directions for future research in this area.
MATERIALS AND METHODS In July 2011 a PubMed database search was performed for articles published between 1970 and the present. The terms used were “genes,” “erectile dysfunction,” “GWAS,”
1677
“genetic polymorphism,” “microarray,” “impotence” and combinations. The search primarily focused on original articles.
CANDIDATE GENE STUDIES In 1975 one of the first studies was published to demonstrate that many proteins have various heritable isoforms, showing that common genetic variations could lead to a change in protein structure and biological function. These genetic polymorphisms can be tested for their association with a disease by determining the genotype of these variants in affected individuals vs those without the disease. If a genotype of interest is present at a higher frequency in cases, it is taken as evidence that the variation is associated with an increased risk of the disease. The majority of studies published in the field of ED genetics are candidate gene studies that take the approach from known to unknown and allow for a focus on a particular gene with a known pathological involvement in the disease (table 1). Nitric Oxide Synthase One of the better studied candidate genes in the penile smooth muscle relaxation pathway is eNOS. To date 3 forms of NOS, originally labeled after their location, have been recognized, including neuronal NOS, eNOS and inducible NOS. It was hypothesized that rapid activation of neuronal NOS initiates the erectile process, whereas phosphorylation and activation of eNOS lead to sustained erection.4 The en-
Table 1. Genetic polymorphisms previously investigated in ED development References eNOS: Erol et al5 Lee et al6 Sinici et al4 Andersen et al7 Erkan et al9 Rosas-Vargas et al12 Safarinejad et al16 Serrano et al10 Park et al15 Slo (Davies et al17) Vcsa1 (Tong et al34) ACE: Park et al15 Mazo et al23 Andersen et al25 Eisenhardt et al13 PR (Andersen et al28) Androgen receptor (Andersen et al29) Hcy (Safarinejad et al30) IGFBP-3: Pu et al45 Soh et al46 Transforming growth factor-1 (Peters et al42)
Polymorphism, Allele, Splice Variant
Rat Model*
Ethnic Group*
7G894T, 4VNTR 7G894T T-786C 7G894T 4VNTR 7G894T T786C, 7G894T, 4a/4b T786C, 7G894T, 4VNTR 4VNTR SVcyt Not applicable
Not applicable Not applicable Not applicable Not applicable Not applicable Not applicable Not applicable Not applicable Not applicable Diabetic Diabetic, aging ⫹ neurogenic
Turkish Taiwanese Turkish Brazilian Turkish Mexican Mestizo Iranian Colombian Korean Not applicable Not applicable
II, DI, DD II, DI, DD II, DI, DD II, DI, DD PROGINS CAG repeat C677T, A1298C, G1793A Not applicable
Not Not Not Not Not Not Not
Korean Russian Brazilian German Brazilian Brazilian Iranian Not applicable
Various
* Most studies were done in specific animal model or in cohort of men of specific ethnicity.
applicable applicable applicable applicable applicable applicable applicable
Aging Diabetic Not applicable
Various
1678
GENETICS OF ERECTILE DYSFUNCTION
dothelial and neuronal forms of the NOS enzyme are encoded by different genes with eNOS mapped to chromosome 7q35 and consisting of 26 exons spanning 21 kb of DNA. Several polymorphisms of eNOS have been investigated. More frequently, investigated regions of this gene include a variable number of 27 bp tandem repeats in intron 4 (4VNTR), G894T (rs1799983) polymorphism in exon 7 and a T-786C (rs2070744) polymorphism in the promoter region.4 –7 The eNOS gene has 2 common alleles containing 4 VNTRs (allele a) and 5 VNTRs (allele b), which make 3 possible genotypes (aa, ab and bb). The biological function of the 4a/4b polymorphism is controversial. In the past it was associated with decreased eNOS concentration and activity, although no decrease in plasma NO levels was reported.8 Also, eNOS aa, ab and bb genotypes have a similar distribution in patients with ED and controls.9 The eNOS G894T polymorphism, which corresponds to a substitution of glutamate by aspartate at amino acid position 298, results in the disturbance of eNOS activity and NO production.7 The T-786C region results in a significant decrease in eNOS gene promoter activity.10 Human studies correlating these polymorphisms with ED prevalence are controversial. A limitation is that most available studies were performed in small groups of patients of various ethnic backgrounds. While the 7G894T polymorphism was noted to be an independent factor for ED in Mexican, Turkish and Taiwanese populations,5,6,11,12 findings do not support a role of this polymorphism in the Brazilian or German population.7,13 A recent meta-analysis by Wang et al showed a significant association of ED and the G894T polymorphism (generalized OR 1.64, 95% CI 1.30 –2.61) but failed to demonstrate the same correlation with 4VNTR (OR 0.96).14 4VNTR (allele a) and the distribution of eNOS genotypes in ED was first investigated in the Korean population in 1999 by Park et al.15 They found no significant difference in the distribution in the ED group compared to controls. Consistently, 4VNTR has not predisposed men to ED in a Turkish or Iranian population.5,16 T-789C is an independent predisposing factor for ED in Iranian and Turkish men.4,16 However, data supporting its role in other populations are lacking. eNOS polymorphisms occur at different frequencies among various ethnic groups.10 This difference in distribution could contribute to the lack of well replicated results across patient populations of different ethnicities. In the future it could potentially be improved by comparing the distribution of eNOS polymorphisms in larger cohorts of patients with ED vs controls across various ethnic backgrounds.
Ion Channels Penile erection and detumescence are partly a function of the smooth muscle cells of the cavernosum and arterial wall, which are regulated by changes in free cytosolic calcium. Ca2⫹ channels can be regulated by currents generated through K⫹ channels.17 These channels allow K⫹ to flow down its gradient and out of the smooth muscle cell. The hyperpolarization that occurs limits intracellular Ca2⫹ entry and results in smooth muscle relaxation. The Maxi-K or BK channel (large performance, calcium sensitive K channel) is the most studied ion channel in the field of ED genetics and physiology that was previously shown to have a role in penile erection.18 However, there are limited data on intrinsic expression of the Maxi-K gene or its variants in men with ED. The pore forming subunit of this channel is encoded by the slo gene. In slo knockout mice cavernous tissue relaxation was decreased by 50% and intracavernous pressure was attenuated in response to nerve stimulation.18 The slo gene is known to undergo alternative splicing triggered by various stimuli, including hormones, leading to several different isoforms.19 In streptozotocin induced diabetic animals alternative splicing led to up-regulation of the SV0 channel forming transcript compared to controls. Also, a higher proportion of SV0 was reported in men with vs without diabetes, suggesting that alternative splicing of the slo gene may have a role in diabetic ED.17 Angiotensin Converting Enzyme I/D The presence of ACE, Ang II and Ang II receptor in the corpora cavernosa has been confirmed. The corpus cavernosum produces physiologically relevant amounts of Ang II and tissues with higher Ang II levels appear to have a decreased response to papaverine and prostaglandin E-1.20 Increased Ang II in cavernous blood was noted during penile detumescence vs rigidity and in patients with organogenic ED vs controls.20 ACE, which converts Ang I into Ang II, is encoded by the ACE gene containing a polymorphism based on the presence (I) or absence (D) in an intron of a 287 bp nonsense DNA domain (rs4646994). This results in 3 possible genotypes, including DD, DI or II.15 An ACE polymorphism accounts for 47% of the total phenotypic variance of serum ACE levels with a higher concentration in the II genotype.21 Identification of homologous sequences in gene ontology annotated databases suggested that the DD genotype lacks molecular and biological function.22 One of the first studies of ACE polymorphisms related to ED was reported by Park et al.15 They examined the relationship between gene polymorphism and organic ED. In the ED group the proportion of the DD genotype was significantly higher
GENETICS OF ERECTILE DYSFUNCTION
1679
than in controls (54% vs 24%, p ⬍0.01). Consistent findings were published in Russian men with metabolic syndrome, with a higher frequency of the D allele and the DD genotype in the ED group vs controls (81.3% vs 63.8%, p ⬍0.001, and 68.3% vs 38.2%, respectively).23 The opposite was seen in the Brazilian population with increased frequency of the I allele in individuals complaining of ED.21 In a study of German patients with ED compared with controls there was no difference in the ACE I/D genotype.13 However, that study included patients with ED attributable to various causes, including vascular, neurogenic, psychogenic, hormonal and mixed etiologies, as well as 11% of men for whom no specific ED etiology could be identified. A lack of association was also noted between ACE genotypes in the Korean population.24 It is possible that the lack of consistent replication in the reports outlined is due to the study of diverse ethnic backgrounds, ED etiologies and comorbidities. Potentially, a larger sample size of patients with a single ED etiology would shed more light on the genetic correlation of polymorphisms and the predisposition to ED.
ED complaints. While sex hormones certainly have an important role in sexual function, currently there is a lack of evidence that genetic polymorphisms of SHR predispose to ED development.
Hormone Receptors Sex hormones have various effects on sexual function, including NOS activity, erectile frequency and the sexual behavior of male rodents.25,26 However, data are limited on possible genetic variations in sex hormone and SHR genes. Genetic expression of androgen, progesterone and estrogen receptor  decreases dramatically in the penile crura of aging rats, suggesting that SHR may have a role in age related ED.26 However, no difference in SHR polymorphisms was found in 2 population based reports. In the first report a PROGINS polymorphism of the PR gene was investigated.27 It consists of a 306 bp Alu insertion in intron G, a silent point mutation in exon 5 and a change at position 660 from valine to leucine (V660L) (rs1042838). In the past the PROGINS variant has been associated with a decreased response to progesterone due to reduced protein activity. Decreased PR activity could potentially alter erectile function. However, in the report by Andersen et al there was no difference in the frequency of PROGINS in an ED vs a nonED group.28 The limitation of the report was that only 1.2% of subjects were homozygous for Alu insertion, possibly making the sample size inadequate to detect a difference. Another study of the CAG repeat sequence of the androgen receptor gene was reported by Andersen et al.29 Longer CAG sequences were associated with decreased transcriptional activity, presumably correlating with relatively reduced androgen feedback. However, there was no association with increased
GENE EXPRESSION PROFILING
Hcy Metabolism Genes involved in the Hcy pathway have been investigated in relation to ED. The methylenetetrahydrofolate reductase gene, located on chromosome 1, is a key enzyme in folic acid metabolism. Three polymorphisms of this gene are known, including C677T (rs1801133), A1298C (rs1801131) and G1793A (rs2274976), which result in decreased enzyme activity and increased Hcy levels.30 An increase in Hcy is linked to impaired endothelial dysfunction and vascular disease. Recently it was reported that patients with early onset vasculogenic ED had higher levels of serum Hcy and a higher prevalence of C677T polymorphism compared to controls (15.8% vs 22.4%, p ⫽ 0.01).30 Additionally, patients homozygous for the C677T allele have a decreased response to PDE5i alone but improve with combination treatment of PDE5i, folate and vitamin B6.31
While the candidate gene approach provides the benefit of knowing the function of the investigated gene, it is unlikely to detect variation in gene expression. For that purpose gene expression profiling becomes useful. It is an effective tool for characterizing gene expression among various ED models and in ED vs healthy controls. Although they do not necessarily directly relate to genetic heritability, methods of gene expression profiling allow us to acquire quantitative data on the level of activity of multiple genes. The majority of all reports available are derived from rodents. One of the first groups reported altered expression of 126 genes in the neurogenic impotence model. One of the down-regulated genes was SMR1, a member of the opiorphin family, which is expressed at a significantly higher level in male than in female rats. Later, it was also suggested as an ED marker in candidate studies.32 In humans other genes from the same family, ProL1, hSMR3A and hSMR3B, are significantly down-regulated in men with ED with and without diabetes.33 The rat homologue of opiorphin, sialorphin, which is encoded by the vsca1 gene, was suggested as an ED marker in diabetic, neurogenic and aging animal models.34 Its product is a neutral endopeptidase inhibitor, which is a membrane bound peptidase that decreases peptide agonist action by removing them from membrane receptors. Substantial vsca1 up-regulation was noted following intracorporeal injection of the Slo gene containing vector, a pore forming
1680
GENETICS OF ERECTILE DYSFUNCTION
subunit of the BK channel.32 It was hypothesized that Vcsa1 may improve CSMC relaxation by neutral endopeptidase inhibition, thus, prolonging peptide agonist action at the receptors.35 Using microarray technology in diabetic rats, 529 genes were identified with differential expression.36 That group took the study 1 step further and identified genes that were previously reported to mediate vascular dysfunction. One of these genes, ceruloplasmin, and its 2 splice variants were up-regulated in diabetic rat CSMCs and endothelium. Previous reports demonstrated that ceruloplasmin impairs aortic relaxation, possibly altering NO bioavailability in the penis.37 Aberrant regulation of growth factors is often implicated in vasculogenic ED. Several penile growth factors have been identified. Originally IGF-1 and VEGF were found to promote CSMC proliferation in rats.38 One of the first groups to report these growth factors in human CSMCs was Rajasekaran et al.39 Using reverse transcriptase-polymerase chain reaction they identified 4 splice variants of VEGF mRNA, including VEGFs 121, 145, 165 and 189, with VEGFs 121 and 165 most abundant. In that study a twofold proliferation of CSMCs was seen with exposure to VEGF. In hypercholesterolemic rats with ED significantly lower VEGF gene expression and lower levels of phosphorylated eNOS were shown.40 Improved intracavernous pressure in diabetic animals transfected with the VEGF gene was also observed.41 In humans specific polymorphisms of VEGF and transforming growth factor-1 are related to ED development following radiation therapy for prostate cancer.42 IGF and its receptors have varied expression in ED. IGF-1 is a polypeptide structurally homologous to proinsulin that facilitates the regeneration of NOS containing nerve fibers.43 After pudendal artery ligation in rats, up-regulation of IGFBP-1, 3 and 5, and down-regulation of IGFBP-6 was seen.44 In the same report the most highly induced gene after pudendal artery ligation was apolipoprotein D. IGFBP-3 mRNA expression levels were increased in aging and diabetic rat models.45,46 It is possible that increased expression levels of IGFBP-3 decrease the availability of IGF-1 in penile tissue, contributing to ED in patients with diabetes. This is consistent with the report that IGF-1 injection improved erectile function in aging animals.45 While to our knowledge it is unknown whether there is a difference in VEGF or IGF-1 gene expression in nondiabetic, nonvasculogenic ED, they could possibly provide the basis for future gene therapy in specific patient populations. Gene expression profiling studies, including microarray analysis, permits analysis of the gene network interactions to identify the most highly affected components. However, frequently they are used to confirm involvement of a gene of interest or
as a starting point for candidate gene studies. This method of analysis also does not account for alternative gene splicing, a major mechanism for generating protein diversity, or for the protein expression level, which may not correlate with gene expression in a direct manner. An adjunct solution may be proteomic analysis to look at the expression of the protein rather than the gene. While those limitations exist, these studies can still provide additional information to the field of ED genetics.
GENOME WIDE ASSOCIATION STUDIES/SNP Candidate gene studies target pathways known to contribute to ED. However, they ignore most of the genome and are likely to miss many causal regions of genes or show many false-positive associations. Results of candidate gene studies are also not often easily and consistently reproduced. Gene expression studies investigate variations in gene transcription for a given condition but do not identify specific polymorphisms. GWASs improve on those shortcomings, allowing comparison of common genetic variants over the entire genome in a large number of cases to those in unaffected controls. However, there are still only a few reports available (table 2). Kerns et al identified SNPs associated with ED development in black American men.47 In the 27 men who qualified for analysis Kerns et al identified SNP rs2268363 as significantly associated with ED after multiple comparison correction (unadjusted p ⫽ 5.46 ⫻ 10⫺8, Bonferroni p ⫽ 0.028) and an additional 4 SNPs that trended toward a significant association (unadjusted p ⬍10⫺6). SNP rs2268363 is located on chromosome 2 in the follicle stimulating hormone receptor, which is involved in testis development and function. Alterations in the follicle stimulating hormone signaling pathway lead to improper gonad development in mice, and to small testis size and infertility in humans. Since to our knowledge this is the only report implicating SNP rs2268363 as possibly involved in ED, a candidate gene study may be warranted. Two more abstracts were recently presented by groups that investigated ED genetics using a GWAS Table 2. SNPs identified in GWASs in patients with ED References Kerns et al47 Hotaling et al48 Hotaling et al49
SNP
Population
Unadjusted p Value*
Black American men 5.46 ⫻ 10⫺8 Diabetes Cohort ⫹ 7 ⫻ 10⫺7, Complications Trial 9 ⫻ 10⫺7 SNP on chromosome 14 Adult Changes in 1.1 ⫻ 10⫺6 Thought rs2268363 rs9810233, rs1920201
* Standard GWAS significance criteria considered at p ⬍5 ⫻ 10⫺8.
GENETICS OF ERECTILE DYSFUNCTION
approach. The first study specifically investigated type 1 diabetes related ED using the Diabetes Control and Complications Trial cohort.48 Two SNPs located on chromosome 3 near the activated leukocyte cell adhesion molecule gene were identified, approaching the standard GWAS significance criterion of p ⬍5 ⫻ 10⫺8. Expression of activated leukocyte cell adhesion molecule, a member of the Ig superfamily, was noted in various cultured human endothelial cells and is also thought to have a role in the inflammatory response. Another series included male patients with and without ED from the Group Health Adult Changes in Thought study.49 After adjustment for smoking, diabetes and hypertension, an SNP on chromosome 14 near the ADCK1 gene was identified to have the strongest association with ED (p ⫽ 1.1E-06). ADCK1 is a highly conserved gene implicated in coenzyme Q biosynthesis, a part of the electron transport chain of the inner mitochondrial membrane. To our knowledge none of the mentioned findings in human studies have been replicated to date. Although GWASs allow analysis of the entire genome, the identified SNPs mostly do not have an obvious association with ED, making the causal variant identification challenging.50 The question then arises of whether the SNP identified represents an association as a proxy of a causal variant with a similar frequency, or an association of the causal variant via linkage disequilibrium. Linkage disequilibrium is present when some combination of alleles or genetic markers is present at a higher or lower frequency in the population than would be expected based on their random combination frequency. Detecting such a synthetic association would require extensive resequencing.50 While GWASs allow simultaneous testing of thousands of SNPs, they raise a statistical challenge of accounting for multiple comparisons. Failure to do so results in decreased power, falsepositive results and incorrect rejection of the null hypothesis, also influencing the publication bias that favors the dissemination of positive results. In addition to those statistical challenges, which are beyond the scope of this report, GWASs require sample sizes in the thousands to detect a modest association. Achievement of this power is particularly challenging while evaluating a multifactorial disease such as ED in a population with a specific phenotype.
CONCLUSIONS Multiple genetic variants potentially associated with ED have been found. Most of them are a product of candidate gene studies stemming from the
1681
pathophysiological basis of ED. Few results have been consistently replicated across various ED models. Currently, the most commonly used ED models in the literature are vasculogenic, neurogenic and aging. What may work as a genetic marker in one model may not be applicable in another model. Gene association studies have identified various gene polymorphisms thought to be etiological contributors or genetic markers of ED. However, few of them have been successfully replicated across different populations, possibly due to variations in ethnic background, sample size, different study inclusion criteria, confounding by the population substructure, potential false-positive associations and underpowered, nonsignificant studies of real associations. ED is a complex disease and, like most diseases with complex traits, it likely results from the interaction of multiple genetic and environmental factors. While the central goal of genetic ED research is to identify genetic factors and their effect of ED development, it would be most desirable to simultaneously consider gene-gene and gene-environment interaction. Gene expression profiling identifies variations in set ED models. However, it does not detect novel polymorphisms or account for post-transcriptional modification. While it is currently a useful adjunct, it could potentially be supplemented further using proteomic analysis. The possible genetic associations that are overlooked by candidate gene studies could be investigated using a GWAS. However, currently only a few reports are available that analyzed a limited number of patients. To determine whether the result of an association study represents a true association would require large sample sizes, an appropriate statistical approach with correction for multiple hypothesis testing and replication of the results in a separate but phenotypically close population. The ultimate goal of the search for the genetic origin or association of ED is the creation of better treatment and possibly cure. The search for genetic associations of ED will likely lead to the discovery of 2 groups of genes, including those that contribute to the development of or protection from impotence and those that are markers for it. Genes belonging to the former group could lead to advances in treatment and specifically tailored therapies based on the genetic makeup of the patient, such as local introduction of a gene encoding for a specific enzyme or ion channel. With the latter the prevention and identification of ED before it is clinically evident and the development of preventive strategies will likely be possible.
1682
GENETICS OF ERECTILE DYSFUNCTION
REFERENCES 1. Avasthi A, Grover S, Bhansali A et al: Erectile dysfunction in diabetes mellitus contributes to poor quality of life. Int Rev Psychiatry 2011; 23: 93. 2. Fischer ME, Vitek ME, Hedeker D et al: A twin study of erectile dysfunction. Arch Intern Med 2004; 164: 165. 3. Melman A, Biggs G, Davies K et al: Gene transfer with a vector expressing Maxi-K from a smooth muscle-specific promoter restores erectile function in the aging rat. Gene Ther 2008; 15: 364. 4. Sinici I, Güven EO, Serefog˘lu E et al: T-786C polymorphism in promoter of eNOS gene as genetic risk factor in patients with erectile dysfunction in Turkish population. Urology 2010; 75: 955. 5. Erol B, Bozdogan G, Akduman B et al: eNOS gene intron 4 VNTR and exon 7-G894T polymorphisms in Turkish men with erectile dysfunction: a case control study. J Sex Med 2009; 6: 1423. 6. Lee YC, Wu WJ, Liu CC et al: The associations among eNOS G894T gene polymorphism, erectile dysfunction, and benign prostate hyperplasia-related lower urinary tract symptoms. J Sex Med 2009; 6: 3158.
dysfunction: a meta-analysis. J Sex Med 2010; 7: 3889. 15. Park JK, Kim W, Kim SW et al: Gene-polymorphisms of angiotensin converting enzyme and endothelial nitric oxide synthase in patients with erectile dysfunction. Int J Impot Res 1999; 11: 273. 16. Safarinejad MR, Khoshdel A, Shekarchi B et al: Association of the T-786C, G894T and 4a/4b polymorphisms of the endothelial nitric oxide synthase gene with vasculogenic erectile dysfunction in Iranian subjects. BJU Int 2011; 107: 1994. 17. Davies KP, Zhao W, Tar M et al: Diabetes-induced changes in the alternative splicing of the slo gene in corporal tissue. Eur Urol 2007; 52: 1229. 18. Werner ME, Zvara P, Meredith AL et al: Erectile dysfunction in mice lacking the large-conductance calcium-activated potassium (BK) channel. J Physiol 2005; 567: 545. 19. Lai GJ and McCobb DP: Opposing actions of adrenal androgens and glucocorticoids on alternative splicing of Slo potassium channels in bovine chromaffin cells. Proc Natl Acad Sci U S A 2002; 99: 7722.
7. Andersen ML, Guindalini C, Santos-Silva R et al: Association analysis of endothelial nitric oxide synthase G894T gene polymorphism and erectile dysfunction complaints in a population-based survey. J Sex Med 2010; 7: 1229.
20. Becker AJ, Uckert S, Stief CG et al: Plasma levels of angiotensin II during different penile conditions in the cavernous and systemic blood of healthy men and patients with erectile dysfunction. Urology 2001; 58: 805.
8. Song J, Yoon Y, Park KU et al: Genotype-specific influence on nitric oxide synthase gene expression, protein concentrations, and enzyme activity in cultured human endothelial cells. Clin Chem 2003; 49: 847.
21. Andersen ML, Guindalini C, Santos-Silva R et al: Angiotensin-converting enzyme polymorphism and erectile dysfunction complaints in the Brazilian population. J Sex Med 2010; 7: 2791.
9. Erkan E, Muslumanoglu AY, Oktar T et al: Polymorphism of endothelial nitric oxide synthase gene in patients with erectile dysfunction. J Sex Med 2006; 3: 69. 10. Serrano NC, Díaz LA, Casas JP et al: Frequency of eNOS polymorphisms in the Colombian general population. BMC Genet 2010; 11: 54. 11. Lee YC, Huang CH, Wang CJ et al: The associations among eNOS G894T gene polymorphism, erectile dysfunction and related risk factors. BJU Int 2007; 100: 1116. 12. Rosas-Vargas H, Coral-Vazquez RM, Tapia R et al: Glu298Asp endothelial nitric oxide synthase polymorphism is a risk factor for erectile dysfunction in the Mexican Mestizo population. J Androl 2004; 25: 728. 13. Eisenhardt A, Stief C, Porst H et al: Genetic association study of the GNB3 C825T, the ACE I/D and the eNOS G894T polymorphisms and the risk to develop erectile dysfunction in a German ED population. Andrologia 2010; 42: 218. 14. Wang JL, Wang HG, Gao HQ et al: Endothelial nitric oxide synthase polymorphisms and erectile
22. Wiwanitkit V: I/D genetic polymorphism of angiotensin-converting enzyme: pathogenesis evaluation for erectile dysfunction by gene ontology. Fertil Steril 2008; 89: 1095. 23. Mazo EB, Gamidov SI, Mamedov MN et al: Association between the insertion/deletion polymorphism of the angiotensin-converting enzyme gene and erectile dysfunction in patients with metabolic syndrome. Int J Impot Res 2008; 20: 68. 24. Kim DS, Choi SI, Lee HS et al: Determination of human angiotensin converting enzyme (ACE) gene polymorphisms in erectile dysfunction: frequency differences of ACE gene polymorphisms according to the method of analysis. Clin Chem Lab Med 2001; 39: 11. 25. Andersen ML, Martins RC, Alvarenga TA et al: Progesterone reduces erectile dysfunction in sleep-deprived spontaneously hypertensive rats. Reprod Biol Endocrinol 2007; 5: 7. 26. Shirai M, Yamanaka M, Shiina H et al: Downregulation of androgen, estrogen and progesterone receptor genes and protein is involved in aging-related erectile dysfunction. Int J Impot Res 2003; 15: 391.
27. Romano A, Delvoux B, Fischer DC et al: The PROGINS polymorphism of the human progesterone receptor diminishes the response to progesterone. J Mol Endocrinol 2007; 38: 331. 28. Andersen ML, Alvarenga TA, Mazzotti DR et al: Hormonal profile, the PROGINS polymorphism, and erectile dysfunction complaints: data from a population-based survey. Fertil Steril 2011; 95: 621. 29. Andersen ML, Guindalini C, Santos-Silva R et al: Androgen receptor CAG repeat polymorphism is not associated with erectile dysfunction complaints, gonadal steroids and sleep parameters: data from a population-based survey. J Androl 2010; 32: 524. 30. Safarinejad MR, Safarinejad S and Shafiei N: Role of methylenetetrahydrofolate reductase gene polymorphisms (C677T, A1298C, and G1793A) in the development of early onset vasculogenic erectile dysfunction. Arch Med Res 2010; 41: 410. 31. Lombardo F, Tsamatropoulos P, Piroli E et al: Treatment of erectile dysfunction due to C677T mutation of the MTHFR gene with vitamin B6 and folic acid in patients non responders to PDE5i. J Sex Med 2010; 7: 216. 32. Calenda G, Tong Y, Tar M et al: Vcsa1 acts as a marker of erectile function recovery after gene therapeutic and pharmacological interventions. J Urol 2009; 181: 2806. 33. Tong Y, Tar M, Melman A et al: The opiorphin gene (ProL1) and its homologues function in erectile physiology. BJU Int 2008; 102: 736. 34. Tong Y, Tar M, Davelman F et al: Variable coding sequence protein A1 as a marker for erectile dysfunction. BJU Int 2006; 98: 396. 35. Davies KP: The role of opiorphins (endogenous neutral endopeptidase inhibitors) in urogenital smooth muscle biology. J Sex Med, suppl., 2009; 6: 286. 36. Sullivan CJ, Teal TH, Luttrell IP et al: Microarray analysis reveals novel gene expression changes associated with erectile dysfunction in diabetic rats. Physiol Genomics 2005; 23: 192. 37. Cappelli-Bigazzi M, Ambrosio G, Musci G et al: Ceruloplasmin impairs endothelium-dependent relaxation of rabbit aorta. Am J Physiol 1997; 273: H2843. 38. Liu X, Lin CS, Graziottin T et al: Vascular endothelial growth factor promotes proliferation and migration of cavernous smooth muscle cells. J Urol 2001; 166: 354. 39. Rajasekaran M, Kasyan A, Allilain W et al: Ex vivo expression of angiogenic growth factors and their receptors in human penile cavernosal cells. J Androl 2003; 24: 85. 40. Ryu JK, Shin HY, Song SU et al: Downregulation of angiogenic factors and their downstream target molecules affects the deterioration of erectile
GENETICS OF ERECTILE DYSFUNCTION
function in a rat model of hypercholesterolemia. Urology 2006; 67: 1329. 41. Gou X, He WY, Xiao MZ et al: Transplantation of endothelial progenitor cells transfected with VEGF165 to restore erectile function in diabetic rats. Asian J Androl 2011; 13: 332. 42. Peters CA, Stock RG, Cesaretti JA et al: TGFB1 single nucleotide polymorphisms are associated with adverse quality of life in prostate cancer patients treated with radiotherapy. Int J Radiat Oncol Biol Phys 2008; 70: 752. 43. Jung GW, Kwak JY, Yoon S et al: IGF-I and TGF-beta2 have a key role on regeneration of nitric oxide synthase (NOS)-containing nerves after cavernous neurotomy in rats. Int J Impot Res 1999; 11: 247.
44. Lin CS, Ho HC, Gholami S et al: Gene expression profiling of an arteriogenic impotence model. Biochem Biophys Res Commun 2001; 285: 565.
1683
can men after radiotherapy for prostate cancer. Int J Radiat Oncol Biol Phys 2010; 78: 1292.
45. Pu XY, Zheng XG, Zhang Y et al: Higher expression of mRNA and protein of insulin-like growth factor binding protein-3 in old rat penile tissues: implications for erectile dysfunction. J Sex Med 2011; 8: 2181.
48. Hotaling J, Waggott D, Patterson A et al: Genome wide association study identifies novel gene for erectile dysfunction risk using the Epidemiology of Diabetes Interventions and Complications (EDIC) database. J Urol, suppl., 2010; 183: e274, abstract 703.
46. Soh J, Katsuyama M, Ushijima S et al: Localization of increased insulin-like growth factor binding protein-3 in diabetic rat penis: implications for erectile dysfunction. Urology 2007; 70: 1019.
49. Hotaling J, Wessells H, Burt A et al: Novel single nucleotide polymorphisms may predict ED: data from the Group Health Adult Changes in Thought Study. J Urol, suppl., 2011; 185: e368, abstract 920.
47. Kerns SL, Ostrer H, Stock R et al: Genome-wide association study to identify single nucleotide polymorphisms (SNPs) associated with the development of erectile dysfunction in African-Ameri-
50. Takeuchi F, Kobayashi S, Ogihara T et al: Detection of common single nucleotide polymorphisms synthesizing quantitative trait association of rarer causal variants. Genome Res 2011; 21: 1122.