Advertisement

A polymorphism of the CYP17 gene related to sex steroid metabolism is associated with female-to-male but not male-to-female transsexualism

Published:September 03, 2007DOI:https://doi.org/10.1016/j.fertnstert.2007.05.056

      Objective

      To assess the association between transsexualism and allele and genotype frequencies of the common cytochrome P450 (CYP) 17 −34 T>C single nucleotide polymorphism (SNP).

      Design

      Case-control study.

      Setting

      Academic research institution.

      Patient(s)

      102 male-to-female (MtF) and 49 female-to-male (FtM) transsexuals, 756 male controls, and 915 female controls.

      Intervention(s)

      Buccal swabs and multiplex polymerase chain reaction on a microarray system.

      Main Outcome Measure(s)

      Analysis of the CYP17 −34 T>C SNP.

      Result(s)

      CYP17 −34 T>C SNP allele frequencies were statistically significantly different between FtM transsexuals and female controls (CYP17 T: 55/98 [56%] and CYP17 C: 43/98 [44%] versus CYP17 T: 1253/1826 [69%] and CYP17 C: 573/1826 [31%], respectively). In accordance, genotype distributions were also different between FtM transsexuals and female controls using a recessive genotype model (CYP17 T/T+T/C: 39/49 [80%] and C/C 10/49 [20%] vs. CYP17 T/T+T/C: 821/913 [90%] and C/C 92/913 [10%], respectively). The CYP17 −34 T>C allele and genotype distributions were not statistically significantly different between MtF transsexuals and male controls. Of note, the CYP17 −34 T>C allele distribution was gender-specific among controls (CYP17 C: males; 604 of 1512 [40%] vs. females; 573 of 1826 [31%]). The MtF transsexuals had an allele distribution equivalent to male controls, whereas FtM transsexuals did not follow the gender-specific allele distribution of female controls but rather had an allele distribution equivalent to MtF transsexuals and male controls.

      Conclusion(s)

      These data support CYP17 as a candidate gene of FtM transsexualism and indicate that loss of a female-specific CYP17 T −34C allele distribution pattern is associated with FtM transsexualism.

      Key Words

      Gender dysphoria is a mental state characterized by a conflict between a person's genetic sex and his/her gender perception. Gender dysphoria comprises several different entities, among them transvestism, cross-dressing, and transsexualism (
      • Brown G.R.
      A review of clinical approaches to gender dysphoria.
      ). The Diagnostic and Statistical Manual of Mental Disorders defines transsexualism as the desire to be of the opposite sex or the assertion that one is of the sex opposite from the one assigned at birth (
      Diagnostic and statistical manual of mental disorders.
      ). It is a rare condition, with incidence rates between 1:12,000 to 1:40,000 (
      • Landen M.
      • Walinder J.
      • Lundstrom B.
      Prevalence, incidence and sex ratio of transsexualism.
      ,
      • Bakker A.
      • van Kesteren P.J.
      • Gooren L.J.
      • Bezemer P.D.
      The prevalence of transsexualism in the Netherlands.
      ). The etiology of transsexualism is unknown, but it has been speculated that the influence of sex steroids on early brain development may play an important role (
      • van Goozen S.H.
      • Slabbekoorn D.
      • Gooren L.J.
      • Sanders G.
      • Cohen-Kettenis P.T.
      Organizing and activating effects of sex hormones in homosexual individuals.
      ,
      • Walinder J.
      • Thuwe I.
      A study of consanguinity between the parents of transsexuals.
      ). Transsexualism may also have a genetic component, based on rare reports of twin–twin concordance and families with several affected members (
      • Green R.
      Family cooccurrence of ‘gender dysphoria’: ten sibling or parent–child pairs.
      ). Chromosomal aberrations are not found at an increased rate among both male-to-female (MtF) and female-to-male (FtM) transsexuals based on a study in 30 and 31 affected individuals, respectively (
      • Hengstschlaeger M.
      • van Trotsenburg M.
      • Repa C.
      • Marton E.
      • Huber J.C.
      • Bernaschek G.
      Sex chromosome aberrations and transsexualism.
      ). In a small series of 29 Swedish MtF transsexuals, however, Henningsson et al. (
      • Henningsson S.
      • Westberg L.
      • Nilsson S.
      • Lundstrom B.
      • Ekselius L.
      Sex steroid-related genes and male-to-female transsexualism.
      ) observed an association between MtF transsexualism and a CA repeat polymorphism in the estrogen receptor (ER) beta gene.
      CYP17 A2 T>C is a functional single nucleotide polymorphism (SNP) associated with elevated serum and plasma levels of estradiol (E2), progesterone, and testosterone (
      • Feigelson S.
      • Shames L.S.
      • Pike M.C.
      • Coetzee G.A.
      • Stanczyk F.Z.
      • Henderson B.E.
      Cytochrome P450c17alpha gene (CYP17) polymorphism is associated with serum estrogen and progesterone concentrations.
      ). In a case-control study, we assessed the genotype frequencies of the CYP17 A2 T>C SNP in a series of Caucasian transsexuals and compared these with controls. We hypothesized that the CYP17 A2 T>C SNP is associated with transsexualism and that mutant alleles will therefore be overrepresented among individuals with this condition.

      Materials and methods

      Approval for this study was obtained by the institutional review board of the Department of Obstetrics and Gynecology, Medical University of Vienna, Austria. From September 1994 to December 2006 we enrolled 104 MtF and 49 FtM transsexuals treated at the Transgender Outpatient Clinic of our department. Transsexualism was diagnosed according to DSM IV criteria (
      Diagnostic and statistical manual of mental disorders.
      ). Signed written consent was obtained from all participants. The male controls were 756 men participating in a health prevention program. The female controls were 915 women seeking counseling for perimenopausal disorders in university hospitals, primary and secondary care facilities, and private offices in Germany and Austria.
      We extracted DNA from the participants'; blood or buccal swabs, and CYP17 −34 T>C genotyping was performed as previously described elsewhere (
      • Tempfer C.B.
      • Riener E.
      • Hefler L.A.
      • Huber J.C.
      • Muendlein A.
      DNA microarray-based analysis of single nucleotide polymorphisms may be useful for assessing the risks and benefits of hormone therapy.
      ,
      • Huber M.
      • Mundlein A.
      • Dornstauder E.
      • Schneeberger C.
      • Tempfer C.B.
      • Mueller M.W.
      • et al.
      Accessing single nucleotide polymorphisms in genomic DNA by direct multiplex polymerase chain reaction amplification on oligonucleotide microarrays.
      ). Two cases were excluded from analysis because of DNA degradation, which left 102 MtF and 49 FtM transsexuals for analysis.
      Serum concentrations of E2 and testosterone were measured using commercial assays.

       Statistics

      Differences in the frequencies of alleles and genotypes were analyzed by chi-square test. Continuous variables were compared using the Wilcoxon rank-sum test. For statistical analysis, we used a dominant genotype model comparing homozygous wild-type versus heterozygous mutant and homozygous mutant genotypes; a recessive genotype model comparing homozygous wild-type and heterozygous mutant versus homozygous mutant genotypes; and a gene dosage model comparing homozygous wild-type versus heterozygous mutant versus homozygous mutant genotypes. A power calculation demonstrated that, with a sample size of 1822, the study has a power of >95% to detect a 10% difference in genotype distribution at a statistical significance level of 0.05 using the Yates correction factor based on published genotype distributions of the CYP17 SNP (
      • Feigelson S.
      • Shames L.S.
      • Pike M.C.
      • Coetzee G.A.
      • Stanczyk F.Z.
      • Henderson B.E.
      Cytochrome P450c17alpha gene (CYP17) polymorphism is associated with serum estrogen and progesterone concentrations.
      ). The odds ratio (OR) was used as a measure of the strength of the association between allele and genotype frequencies and transsexualism. All P values are two-tailed, and 95% confidence intervals (CI) were calculated. P<.05 was considered statistically significant.

      Results

      In our patient population of 151 transsexuals, 102 (68%) were MtF and 49 (32%) were FtM, which is in accordance with patterns previously described elsewhere (
      • Cohen-Kettenis P.T.
      • Gooren L.J.G.
      Transsexualism: a review of etiology, diagnosis, and treatment.
      ). The mean age at presentation of FtM transsexuals was 33.2 (±7.7 SD) years. The mean age of MtF transsexuals was 41.8 (±9.8 SD) years. This is also in accordance with previous reports indicating that the coming out of FtM transsexuals occurs significantly earlier compared to MtF transsexuals (
      • Landen M.
      • Walinder J.
      • Lundstrom B.
      Prevalence, incidence and sex ratio of transsexualism.
      ,
      • Bakker A.
      • van Kesteren P.J.
      • Gooren L.J.
      • Bezemer P.D.
      The prevalence of transsexualism in the Netherlands.
      ,
      • Cohen-Kettenis P.T.
      • Gooren L.J.G.
      Transsexualism: a review of etiology, diagnosis, and treatment.
      ). The genotype distributions were in Hardy-Weinberg equilibrium among FtM transsexuals, MtF transsexuals, female controls, and male controls (P=.9, P=.9, P=.9, and P=.7, respectively), indicating that there is no loss of specific CYP17 −34 T>C genotypes due to an association with life-threatening or life-shortening diseases.
      Allele and genotype frequencies among MtF transsexuals and male controls are shown in Table 1. Allele and genotype frequencies among FtM transsexuals and female controls are shown in Table 2. As shown in Table 2, both allele frequencies and genotype distributions were statistically significantly different between FtM transsexuals and female controls The genotype distributions were different using a recessive genotype model (P=.04) and a gene dosage model (P=.03). Using a dominant genotype model, the difference of genotype distributions was of borderline statistical significance (P=.06).
      Table 1Cytochrome P450 17 (CYP17) −34T>C polymorphism: allele and genotype frequencies among male-to-female transsexuals (cases) and male controls.
      AlleleCases (n = 102; 204 alleles) n (%)Controls (n = 756; 1512 alleles) n (%)P value (OR; 95% CI)
      CYP17 T (wt)126 (62%)908 (60%)
      CYP17 C (mt)78 (38%)604 (40%)0.7 (0.93; 0.65–1.34)
      GenotypeCases (n = 102) n (%)Controls (n = 756) n (%)P value (OR; 95% CI)
      CYP17 TT40 (39%)278 (37%)
      CYP17 CT46 (45%)352 (46%)T/T+C/T vs. C/C
      Recessive genotype model.
      0.9 (0.19; 0–∞)
      CYP17 CC16 (16%)126 (17%)T/T vs. C/T+C/C
      Dominant genotype model.
      0.7 (0.90; 0.52–1.57)
      Note: wt = wild type; mt =mutant; OR = odds ratio; CI = confidence interval.
      a Recessive genotype model.
      b Dominant genotype model.
      Table 2Cytochrome P450 17 (CYP17) –34T>C polymorphism: allele and genotype frequencies among female-to-male transsexuals (cases) and female controls.
      AlleleCases (n = 49; 98 alleles) n (%)Controls (n = 913; 1826 alleles) n (%)P value (OR; 95% CI)
      CYP17 T (wt)55 (56%)1253 (69%)
      CYP17 C (mt)43 (44%)573 (31%)0.01 (1.71; 1.12–2.62)
      GenotypeCases (n = 49) n (%)Controls (n = 913) n (%)P value (OR; 95% CI)
      CYP17 TT16 (33%)432 (47%)
      CYP17 CT23 (47%)389 (43%)T/T+C/T vs. C/C
      Recessive genotype model.
      0.04 (2.89; 1.05–7.97)
      CYP17 CC10 (20%)92 (10%)T/T vs. C/T+C/C
      Dominant genotype model.
      0.06 (1.85; 0.97–3.54)
      Note: wt = wild type; mt = mutant; OR = odds ratio; CI = confidence interval.
      a Recessive genotype model.
      b Dominant genotype model.
      The CYP17 −34 T>C allele and genotype distributions were not statistically significantly different between MtF transsexuals and male controls, as shown in Table 1. Of note, the CYP17 −34 T>C allele distribution was gender-specific among male and female controls in that male controls had a statistically significant higher mutant C allele frequency compared with females (CYP17 C: 604/1512 [40%] vs. 573/1826 [31%], respectively; P=.001). The MtF transsexuals had an allele distribution equivalent to their male pairs. In contrast, FtM transsexuals did not follow the gender-specific allele distribution of their female pairs; rather, they had an allele distribution equivalent to MtF transsexuals and male controls (P=.4 and P=.5, respectively).
      The mean serum concentrations of E2 and testosterone in MtF transsexuals before the start of hormone treatment were 30.8 (SD 15.8) pg/mL and 5.1 (SD 2.1) ng/mL, respectively. The mean serum concentrations of E2 and testosterone in FtM transsexuals before the start of hormone treatment were 89.4 (SD 68.2) pg/mL, and 0.6 (SD 0.9) ng/mL, respectively. Among MtF transsexuals, the median serum concentrations of E2 and testosterone were not statistically significantly different between CYP17 −34 T>C genotypes comparing CYP17 T/T+C/T and C/C genotypes (E2: 28.5 [min 24.5; max 35.0] pg/mL vs. 26.5 [min 22.5; max 29.5] pg/mL, P=.2; and testosterone: 5.2 [min 0.6; max 13.1] ng/mL vs. 4.7 [min 2.3; max 7.6] ng/mL, P=.6, respectively). Among FtM transsexuals, the median serum concentrations of E2 and testosterone were also not statistically significantly different (E2: 76.0 [min 44.0; max 101.5] pg/mL vs. 53.5 [min 37.0; max 70.0] pg/mL, P=.3; and testosterone: 0.5 [min 0.2; max 1.3] ng/mL vs. 0.5 [min 0.4; max 0.6] ng/mL, P=.5, respectively).

      Discussion

      Our study found that carriage of the mutant CYP17 −34 T>C C allele is statistically significantly associated with FtM, but not MtF transsexualism. The CYP17 −34 T>C allele distribution was gender-specific among controls. The MtF transsexuals had an allele distribution equivalent to male controls, whereas the FtM transsexuals did not follow the gender-specific allele distribution of female controls but rather had an allele distribution equivalent to MtF transsexuals and male controls.
      To the best of our knowledge, this is the largest case-control study to date evaluating a genetic risk factor of transsexualism. The results of this study support CYP17 as a candidate gene of FtM transsexualism. These data also indicate that loss of a female-specific CYP17 T−34C SNP allele distribution pattern predisposes women to develop transsexualism.
      Dominant, recessive, and gene-dosage genetic models were performed. All three models confirmed the association between the CYP17 T−34C SNP and FtM transsexualism. This consistency supports a cause–effect relation in a gene-dosage–dependent manner. The CYP17 enzyme converts 17-hydroxypregnenolone to dehydroepiandrosterone (DHEA) and 17-hydroxyprogesterone to androstenedione (AND). The DHEA and AND are then further metabolized to testosterone, estrone, and estradiol (E2) (
      • Gruber C.J.
      • Tschugguel W.
      • Schneeberger C.
      • Huber J.C.
      Production and actions of estrogens.
      ). Because the CYP17 T−34C SNP is a gain of function mutation, carriers of the mutant C allele have higher serum and tissue concentrations of both testosterone and E2 (
      • Feigelson S.
      • Shames L.S.
      • Pike M.C.
      • Coetzee G.A.
      • Stanczyk F.Z.
      • Henderson B.E.
      Cytochrome P450c17alpha gene (CYP17) polymorphism is associated with serum estrogen and progesterone concentrations.
      ). The association between the CYP17 T−34C SNP T allele and FtM transsexualism is therefore compatible with the hypothesis that increased tissue availability of both testosterone and E2 leads to an aberration of early brain development regarding central sex differentiation (
      • van Goozen S.H.
      • Slabbekoorn D.
      • Gooren L.J.
      • Sanders G.
      • Cohen-Kettenis P.T.
      Organizing and activating effects of sex hormones in homosexual individuals.
      ,
      • Dorner G.
      • Poppe I.
      • Stahl F.
      • Kolzsch J.
      • Uebelhack R.
      Gene- and environment-dependent neuroendocrine etiogenesis of homosexuality and transsexualism.
      ,
      • Bosinski H.A.
      • Peter M.
      • Bonatz G.
      • Arndt R.
      • Heidenreich M.
      • Sippell W.G.
      • et al.
      A higher rate of hyperandrogenic disorders in female-to-male transsexuals.
      ).
      In our series, however, females without the CYP17 T−34C SNP were FtM transsexuals and females who were carriers of this SNP were not transsexuals. Thus, carriage of the mutant CYP17 T−34C SNP C allele is neither necessary nor sufficient for developing transsexualism. These data indicate that the CYP17 T−34C SNP is at best a high-frequency, low-penetrance susceptibility marker of FtM transsexualism.
      Our study has limitations. For example, selection bias has to be acknowledged when interpreting the results of this study as we only enrolled transsexual individuals visiting our outpatient clinic; thus, we cannot rule out bias by self-selection. Furthermore, we have only documented an association between carriage of a CYP17 SNP and FtM transsexualism. A cause–effect relationship cannot be established by an association study. Therefore, the interpretation of our data that loss of a female-specific CYP17 T−34C SNP allele distribution pattern predisposes women to develop transsexualism has to be interpreted with caution. In addition, the control population was not specifically investigated as to their sexual orientation and may also be biased by self-selection.
      Our study also has strengths. Given the rarity of this disorder, the number of investigated individuals is high, representing the largest association study in this patient population to date. Also, recruitment in a university department that specializes in gender disorders ensures adequate quality regarding the diagnostic criteria of transsexualism.
      We present the largest case-control study of Caucasian transsexuals to date investigating a sex steroid metabolizing gene polymorphism. In this series, the presence of the CYP17 −34 T>C SNP was statistically significantly overrepresented among FtM but not MtF transsexuals, thus supporting CYP17 as a candidate gene of FtM transsexualism.

      References

        • Brown G.R.
        A review of clinical approaches to gender dysphoria.
        J Clin Psychiatry. 1990; 51: 57-64
      1. Diagnostic and statistical manual of mental disorders.
        4th ed. American Psychiatric Association, Washington, DC1994
        • Landen M.
        • Walinder J.
        • Lundstrom B.
        Prevalence, incidence and sex ratio of transsexualism.
        Acta Psychiatr. 1996; 93: 221-223
        • Bakker A.
        • van Kesteren P.J.
        • Gooren L.J.
        • Bezemer P.D.
        The prevalence of transsexualism in the Netherlands.
        Acta Psychiatr Scand. 1993; 87: 237-238
        • van Goozen S.H.
        • Slabbekoorn D.
        • Gooren L.J.
        • Sanders G.
        • Cohen-Kettenis P.T.
        Organizing and activating effects of sex hormones in homosexual individuals.
        Behav Neurosci. 2002; 116: 982-988
        • Walinder J.
        • Thuwe I.
        A study of consanguinity between the parents of transsexuals.
        Br J Psychiatr. 1977; 131: 73-74
        • Green R.
        Family cooccurrence of ‘gender dysphoria’: ten sibling or parent–child pairs.
        Arch Sex Behav. 2000; 29: 499-507
        • Hengstschlaeger M.
        • van Trotsenburg M.
        • Repa C.
        • Marton E.
        • Huber J.C.
        • Bernaschek G.
        Sex chromosome aberrations and transsexualism.
        Fertil Steril. 2003; 79: 639-640
        • Henningsson S.
        • Westberg L.
        • Nilsson S.
        • Lundstrom B.
        • Ekselius L.
        Sex steroid-related genes and male-to-female transsexualism.
        Psychoneuroendocrinology. 2005; 30: 657-664
        • Feigelson S.
        • Shames L.S.
        • Pike M.C.
        • Coetzee G.A.
        • Stanczyk F.Z.
        • Henderson B.E.
        Cytochrome P450c17alpha gene (CYP17) polymorphism is associated with serum estrogen and progesterone concentrations.
        Cancer Res. 1998; 58: 585-587
        • Tempfer C.B.
        • Riener E.
        • Hefler L.A.
        • Huber J.C.
        • Muendlein A.
        DNA microarray-based analysis of single nucleotide polymorphisms may be useful for assessing the risks and benefits of hormone therapy.
        Fertil Steril. 2004; 82: 132-137
        • Huber M.
        • Mundlein A.
        • Dornstauder E.
        • Schneeberger C.
        • Tempfer C.B.
        • Mueller M.W.
        • et al.
        Accessing single nucleotide polymorphisms in genomic DNA by direct multiplex polymerase chain reaction amplification on oligonucleotide microarrays.
        Annal Biochem. 2002; 303: 25-33
        • Cohen-Kettenis P.T.
        • Gooren L.J.G.
        Transsexualism: a review of etiology, diagnosis, and treatment.
        J Psychosom Res. 1999; 46: 315-333
        • Gruber C.J.
        • Tschugguel W.
        • Schneeberger C.
        • Huber J.C.
        Production and actions of estrogens.
        N Engl J Med. 2002; 346: 340-352
        • Dorner G.
        • Poppe I.
        • Stahl F.
        • Kolzsch J.
        • Uebelhack R.
        Gene- and environment-dependent neuroendocrine etiogenesis of homosexuality and transsexualism.
        Exp Clin Endocrinol. 1991; 98: 141-150
        • Bosinski H.A.
        • Peter M.
        • Bonatz G.
        • Arndt R.
        • Heidenreich M.
        • Sippell W.G.
        • et al.
        A higher rate of hyperandrogenic disorders in female-to-male transsexuals.
        Psychoneuroendocrinology. 1997; 22: 361-380