Androgen Insensitivity Syndrome

A number sign (#) is used with this entry because androgen insensitivity syndrome (AIS) is caused by mutation in the androgen receptor gene (AR; 313700) on chromosome Xq12.

Description

The androgen insensitivity syndrome is an X-linked recessive disorder in which affected males have female external genitalia, female breast development, blind vagina, absent uterus and female adnexa, and abdominal or inguinal testes, despite a normal male 46,XY karyotype. Partial androgen insensitivity (PAIS; 312300), also called Reifenstein syndrome, results in hypospadias and micropenis with gynecomastia.

Nomenclature

The androgen insensitivity syndrome was referred to earlier, in both the human and the mouse, as testicular feminization (TFM). The designation androgen insensitivity more accurately reflects the basic pathophysiology and is clearly more satisfactory to patients and their families.

Clinical Features

Patients with androgen insensitivity syndrome often come to medical attention because of a presumed inguinal hernia. Many have absent pubic and axillary hair ('hairless pseudofemale'). The hair of the head is luxuriant, without temporal balding. The phenotype is often very feminine (Netter et al., 1958; Polaillon, 1891). Marshall and Harder (1958) reported affected monozygotic twins.

In a patient studied by Wilkins (1957), the hair follicles of the axillary and pubic areas, although anatomically normal, were unresponsive to local or parenteral administration of androgens and the beard, voice, and clitoris were similarly unresponsive. This was the first demonstration that the basic defect in cases of the hairless pseudofemale type is end-organ unresponsiveness to androgen, a situation comparable to nephrogenic diabetes insipidus and pseudohypoparathyroidism. (These conditions are analogous to the situation in the Sebright Bantam cock which has a female comb structure despite obvious demonstrations of virility. Albright et al. (1942) misspelled 'Sebright' in their classic article.) It is likely that more than one distinct entity is included in the testicular feminization syndrome. Wilkins stated: 'in about one-third of the cases of male pseudohermaphroditism 'of feminine type' sexual hair has been entirely lacking.'

Morris (1962) called attention to the following case of Gayral et al. (1960): a woman, who was sister, mother, and grandmother of affected males, showed asymmetry in the development of the breasts, body hair, and vulva. The right breast was smaller than the left and there was no pubic hair to the right of the midline. She had always had menstrual irregularity but had 3 children, an affected male, a carrier daughter, and a daughter who was the mother of 3 unaffected sons. The findings may be best explained by an X-linked recessive (or incompletely recessive) gene whose effects are to render tissues resistant to male hormone, the patchy changes in the heterozygous female representing the Lyon phenomenon.

Wilson (1981) studied 35 families with 1 of the 4 forms of androgen insensitivity classified as testicular feminization, incomplete testicular feminization, Reifenstein syndrome, or infertile male syndrome. In 31 of the families, he found an abnormality of the androgen receptor: abnormal binding, qualitatively abnormal receptor or decreased amount of receptor. In the other 4, no abnormality of receptor could be demonstrated.

Bals-Pratsch et al. (1990) found qualitative and quantitative abnormalities of the androgen receptor in 3 brothers with prepenile scrotum (congenital transposition of the penis), bifid scrotum, scrotal hypospadias, and bilateral undescended testes.

Kaufman et al. (1984) studied an XY patient, with ambiguous genitalia at birth and breast development at puberty, whose cultured fibroblasts showed normal initial formation of low-affinity androgen-receptor complexes but defective transformation of these complexes to a higher affinity state. They presumed that the defect was in the X-linked structural gene for androgen receptor. A qualitative defect of the androgen receptor was demonstrated (Kovacs et al., 1984); although its binding properties were normal, it was unstable on sucrose density gradient centrifugation.

Hughes and Evans (1986) described 2 sibs with classic complete androgen insensitivity syndrome (CAIS) but increased androgen receptor concentrations in genital skin fibroblasts. The steroid-receptor complex appeared to be translocated normally into the nucleus. They concluded that 'the gene coding for the androgen receptor is intact and does not account for the androgen insensitivity.' But is it not possible that the mutation is in the part of the receptor that is concerned with its effects on DNA? Pinsky et al. (1987) described a family in which the proposita and her aunt had partial androgen resistance of a type different from those previously described. Although there was normal maximum binding capacity, there was an increased apparent equilibrium dissociation constant with dihydrotestosterone and 2 synthetic androgens.

Grino et al. (1988) described a family in which gynecomastia and undervirilization occurred in 5 men, 4 of whom had fathered children, in a pedigree pattern consistent with X-linked recessive inheritance. In fibroblasts cultured from genital skin from 2 of the men, the levels of androgen receptor and the affinity of receptor for dihydrotestosterone were normal. However, androgen binding in fibroblast monolayers was thermolabile, upregulation of receptor levels did not occur after prolonged incubation with dihydrotestosterone or methyltrienolone, and dissociation rates at 37 degrees centigrade were increased with the synthetic androgen mibolerone. In addition, in cytosol preparations the androgen receptor protein was unstable. Grino et al. (1988) suggested that this disorder represents the most subtle functional abnormality of androgen receptor characterized to date, since it was compatible with normal male phenotypic development and in some affected men with fertility.

Davies et al. (1997) described 2 patients with complete androgen insensitivity syndrome and mental retardation associated with submicroscopic deletion of the AR gene. They pointed to the report of another patient with associated CAIS and MR. They postulated that the deletion involves, in addition to the AR gene, 1 or more neighboring genes that are implicated in nonspecific MR.

Holterhus et al. (2000) reported a family with 4 affected individuals, 3 brothers (B1-3) and their uncle, displaying strikingly different external genitalia: B1, ambiguous; B2, severe micropenis; B3, slight micropenis; and uncle, micropenis and penoscrotal hypospadias. All had been assigned a male gender. They detected the same mutation in the AR gene (313700.0050) in each subject. Holterhus et al. (2000) demonstrated that the mutant AR could switch its function from subnormal to normal within the physiologic concentration range of testosterone. This was reflected by an excellent response to testosterone therapy in B1, B2, and the uncle. The authors concluded that, taking into account the well documented individual and time-dependent variation in testosterone concentration in early fetal development, their observations illustrated the potential impact of varying ligand concentrations for distinct cases of phenotypic variability in AIS.

Biochemical Features

Amrhein et al. (1976) presented evidence for 2 types of testicular feminization: in one, the receptor for dihydrotestosterone (DHT) was deficient; in the other, the receptor (androgen receptor, AR; 313700) was apparently present but the receptor-DHT complex was for some reason ineffective. The second type, 'receptor-positive' cases, included the 3 sibs pictured by McKusick (1964). They displayed some pubic hair. The first type included a patient with the 'hairless female' phenotype, also pictured by McKusick (1964). All were longtime patients of Dr. Lawson Wilkins, and it was in the last patient that he demonstrated unresponsiveness to locally administered androgens.

Griffin (1979) found a qualitative abnormality of androgen receptor, manifested by thermolability, in some cases of testicular feminization. Binding overlapped the normal range at 26 degrees C. It was half-normal at 37 degrees and less than 20% of normal at 42 degrees. Gerli et al. (1979) described a case of complete testicular feminization syndrome in a person with the 47,XXY karyotype. Obviously, nondisjunction occurred in the carrier mother, who was 40 years old. Two sibs and a daughter of each of 2 sisters of the patient also had testicular feminization. Unlike the usual cases, the patient had low plasma testosterone and high gonadotropins. German and Vesell (1966) reported this situation in monozygotic twins. Kaufman et al. (1979) reported 2 'receptor-positive' cases of complete androgen insensitivity. One of these had maternally related affected relatives in 3 successive generations.

Kaufman et al. (1981) suggested that whereas one class of mutation that affects the structural domain of the androgen receptor confers increased dissociability and defective upregulation (a term they coined), a second impairs upregulation only.

AIS results from the incapacity for testosterone and dihydrotestosterone to virilize male embryos and is mainly attributable to molecular defects of the AR gene. In normal males, testosterone and LH rise during the first few months of life, and this physiological surge is commonly used to evaluate the gonadotropic axis at this age. This neonatal surge had not yet been evaluated in detail in newborns with AIS. Bouvattier et al. (2002) sequentially measured plasma testosterone, LH, and FSH during the first 3 months of life in 15 neonates with AIS and AR mutations. A GNRH (152760) and a human CG (see 118860) stimulation test were also performed. Patients were divided into 2 groups with complete or partial AIS (CAIS or PAIS). In patients with PAIS, testosterone levels were in the high-to-normal range at day 30 (18.4 +/- 6.9 nM) and day 60 (12.8 +/- 3.8 nM). In contrast, plasma testosterone values were below the normal range in 9 of 10 patients with CAIS at day 30 (1 +/- 0.3 nM) and day 60 (1.4 +/- 0.7 nM, both P less than 0.004 vs PAIS). Plasma LH values were low in CAIS at day 30 and increased normally in PAIS. Bouvattier et al. (2002) concluded that the postnatal testosterone and LH surge occurs expectedly in neonates with PAIS but is absent in those with CAIS, and that the postnatal testosterone rise requires the receptivity of the hypothalamopituitary axis to testosterone.

To investigate the interaction of androgens with the IGF system, Elmlinger et al. (2001) compared the expression of IGFs and IGFBPs in cultured genital skin fibroblasts from 9 patients with the syndrome of complete androgen insensitivity with that in genital skin fibroblasts from 10 normally virilized males. Complete AIS genital skin fibroblast strains produced significantly lower IGF2 (147470) and IGF2 mRNA than control genital skin fibroblast strains. The production of IGFBP2 (146731) was also decreased in complete AIS genital skin fibroblasts, whereas that of IGFBP3 (146732) did not differ. The authors concluded that in addition to the endocrine actions of IGF1 (147440), IGF2, and IGFBP2, they are also involved as local growth factors in the mediation of androgen action and growth of genital tissues.

Sobel et al. (2006) studied subjects with complete androgen insensitivity and 5-alpha-reductase-2 deficiency (see 264600) to determine the direct effect of androgens on bone mineral density (BMD). In CAI subjects, BMD was significantly decreased in the spine and hip, whereas subjects with 5-alpha-reductase-2 deficiency had normal BMD values. Sobel et al. (2006) concluded that androgens are of direct importance in the development and/or maintenance of BMD and that testosterone and/or low levels of dihydrotestosterone appear to be sufficient for BMD development and/or maintenance.

Inheritance

The means for establishing X-linked inheritance include demonstration of linkage with an X chromosome marker, demonstration of lyonization in heterozygous females, and demonstration that the proportion of new mutation cases is one-third rather than one-half (expected of an autosomal dominant). Meyer et al. (1975) found 2 clones of fibroblasts in heterozygous females, one with androgen-binding and one without, thus clinching the X-linkage of this disorder.

Cytogenetics

Muller et al. (1990) described an almost 12-year-old black female with testicular feminization and 47,XXY Klinefelter syndrome. Using DNA markers, they demonstrated that the supernumerary X chromosome resulted from maternal nondisjunction during meiosis II. The error at this stage provided the basis for homozygosity of the mutation at the androgen receptor locus.

Xu et al. (2003) described a 3-month-old girl with CAIS in whom the diagnosis was made during elective repair of inguinal hernia, which had been noted shortly after birth. She had a 46,XY karyotype with inversion of the X chromosome with one break disrupting the AR gene. The phenotypically normal 46,XX mother also carried the inversion in one X chromosome; a maternal aunt had CAIS and a 46,inv(X),Y karyotype. At the age of 5 years this aunt had undergone repair of inguinal hernias, at which time testes were identified. She underwent gonadectomy 1 year later because of concerns of potential malignancy. At age 16 years she had primary amenorrhea and a height of 180 cm.

Mapping

In the most extensively affected kindred known with complete androgen insensitivity, one living in the Dominican Republic, Imperato-McGinley et al. (1990) found linkage to DXS1 and PGK1, localizing the AR gene to an area between Xq11 and Xq13. Linkage between DXS1 and AR showed a peak lod score of 3.2 at theta = 0.06. No recombination was found between PGK1 and AR; peak lod score was 2.9 at theta = 0.0. Although both AR and PGK1 are distal to DXS1, it was not possible to determine the sequence of the 2. Using 3 cDNA probes spanning various parts of the AR gene, they could demonstrate no abnormality in restriction fragment patterns, suggesting that the gene defect is not a deletion but rather a point mutation or a small insertion/deletion.

Also see the mapping section under androgen receptor (AR; 313700).

Molecular Genetics

See androgen receptor (AR; 313700).

Heterogeneity

Genetic Heterogeneity

Adachi et al. (2000) described a 19-year-old woman with primary amenorrhea in whom the complete androgen insensitivity syndrome was diagnosed on the basis of phenotypic and endocrinologic findings, but who had no mutations in the androgen receptor gene (AR; 313700). The patient had normal breast development and normal female external genitalia, but she had no pubic or axillary hair, and the vagina was short and ended in a blind pouch. Abdominal exploration revealed no uterus, but testes were present, which were resected. Histologic examination of the testes revealed small numbers of immature Sertoli cells and germ cells and a moderate number of Leydig cells. The karyotype was 46,XY. Detailed studies revealed that transmission of the activation signal from the AF-1 region of the androgen receptor was disrupted, suggesting that a coactivator interacting with the AF-1 region of the androgen receptor was lacking in this patient. Adachi et al. (2000) stated that this would represent a newly identified form of steroid hormone insensitivity, a coactivator disease.

Genotype/Phenotype Correlations

Boehmer et al. (2001) analyzed the genotype-phenotype relationship in AIS and the occurrence of possible causes of phenotypic variation in families with multiple affected cases. Of 49 index cases with possible AIS identified, 59% had affected relatives. A total of 17 families were studied, 7 families with CAIS (18 patients), 9 families with PAIS (24 patients), and 1 family with female prepubertal phenotypes (2 patients). No phenotypic variation was observed in families with CAIS. However, phenotypic variation was observed in 1 of 3 families with PAIS resulting in different sex of rearing and differences in requirement of reconstructive surgery. Intrafamilial phenotypic variation was observed for mutations R846H (313700.0040) and M771I (313700.0039). Patients with a functional complete defective AR had some pubic hair, Tanner stage P2, and vestigial wolffian duct derivatives despite absence of AR expression. Vaginal length was functional in most but not all CAIS patients. Boehmer et al. (2001) concluded that while phenotypic variation was absent in families with CAIS, distinct phenotypic variation was observed relatively frequent in families with partial AIS.

Pathogenesis

French et al. (1966) found that testosterone failed to affect the urinary excretion of nitrogen, phosphorus and citric acid when given in a dosage much greater than that which in controls decreased excretion of all three. Plasma estrogen levels were the same as those observed in the normal female. Leydig cell stimulation to estrogen production occurs probably because of failure of the feedback repression of the pituitary which shares the unresponsiveness to testosterone. Southern and Saito (1961) showed normal testosterone levels in this disorder.

Clinical Management

Ong et al. (1999) identified a met807-to-thr mutation (313700.0044) in the AR gene in a 46,XY infant with female-appearing genitalia. An AR construct bearing the met807-to-thr mutation, when expressed in COS-7 and HeLa cells, did not bind to its natural ligand testosterone at its upper physiologic range of 3 nM/L but bound dihydrotestosterone to near normal levels at the same concentration. The transactivation function of the receptor was reduced to 15% of normal at physiologic doses of the hormone. However, DHT was able to induce receptor transactivity 10-fold more efficiently than testosterone, in parallel with androgen-binding assays. Intramuscular injections of increasingly higher doses of depo-testosterone, though elevating serum testosterone level, did not result in significant development of the infant's male external genitalia. Subsequently, he was treated with a DHT gel, applied topically to the periscrotal region 3 times a day, for 5 weeks. Serum DHT rose from 1.8 nmol/L to 8.8 nmol/L and resulted in improved male genital development. The authors concluded that in vitro functional assays can help identify the subset of patients with ambiguous genitalia who could respond well to androgen therapy, providing them an option to be reared in accordance with their chromosomal sex.

Wisniewski et al. (2000) assessed by questionnaire and medical examination the physical and psychosexual status of 14 women with documented complete androgen insensitivity syndrome (CAIS). They determined participant knowledge of CAIS as well as opinion of medical and surgical treatment. As a whole, secondary sexual development of these women was satisfactory, as judged by both participants and physicians. In general, most women were satisfied with their psychosexual development and sexual function. All of the women who participated were satisfied with having been raised as females, and none desired a gender reassignment. Although not perfect, the medical, surgical, and psychosexual outcomes for women with CAIS were satisfactory; however, specific ways for improving long-term treatment of this population were identified.

Population Genetics

Mainly using data on the frequency of inguinal hernia in females, Jagiello and Atwell (1962) estimated the frequency of testicular feminization as being about 1 in 65,000 males.

Edwards et al. (1992) demonstrated that the distribution of the number of CAG repeats in exon 1 of the AR gene was lowest in African Americans, intermediate in non-Hispanic whites, and highest in Asians. The distribution of allele size was bimodal in African Americans, and only in African Americans was there a deviation from Hardy-Weinberg equilibrium. Irvine et al. (1995) studied the distribution of the CAG and GC repeats (microsatellites) in exon 1 of the AR gene in African Americans, non-Hispanic whites, and Asians (Japanese and Chinese) and confirmed the findings of Edwards et al. (1992). The frequency of prostate cancer (176807) in the 3 racial groups is inversely proportional to the length of the repeats. One of the critical functions of the product of the AR gene is to activate the expression of target genes. This transactivation activity resides in the N-terminal domain of the protein which is encoded in exon 1 which contains the polymorphic repeats. The smaller size of the CAG repeat is associated with a higher level of receptor transactivation function, thereby possibly resulting in a higher risk of prostate cancer. Irvine et al. (1995) noted that Schoenberg et al. (1994) had observed a somatic mutation resulting in a contraction of the CAG repeat from 24 to 18 in an adenocarcinoma prostate and the effects of the shorter allele were implicated in the development of the tumor.

Based on patients with molecular proof of the diagnosis in a nationwide study in the Netherlands and previous estimates from the Danish patient registry, Boehmer et al. (2001) estimated that the minimal incidence of AIS is 1:99,000.

Animal Model

Lyon and Hawkes (1970) described a homologous phenotype in the mouse and showed that it is genetic, the Tfm locus being situated in the middle of the X chromosome. Ohno and Lyon (1970) showed that in these mice certain enzymes of the mouse kidney, e.g., alcohol dehydrogenase, are not inducible by testosterone as is usually possible. They postulated that the Tfm locus is a repressive regulatory locus controlling many testosterone inducible enzymes. In affected hemizygotes all these enzymes become noninducible. According to their suggestion, this is a regulator mutation like the noninducible mutation in the lac-repressor locus of E. coli as elucidated by Jacob and Monod (1963). Bardin et al. (1970) described studies of the pseudohermaphroditic rat which seems to have a disorder analogous to testicular feminization. Androgen-dependent differentiation is absent. Defective formation of dihydrotestosterone was apparently not the explanation. Goldstein and Wilson (1972) studied the Tfm mouse and showed, by giving dihydrotestosterone to pregnant mothers, that there is resistance to androgen-mediated sexual differentiation in embryos. Low serum testosterone and low production of testosterone in adult Tfm testis of the mouse were features different from those in man, but were considered by them as secondary to the defect in differentiation. They showed deficient binding of testosterone in the nuclei of the submaxillary gland of these adult Tfm animals, but again this may be the result of incomplete differentiation of an androgen-sensitive cell line.

Bullock and Bardin (1972) concluded that androgen-binding proteins are absent from the cytosol of preputial gland of Tfm rats and from the kidney of Tfm mice. Testicular feminization rats, despite female external sexual development, show masculine sexual behavior and little feminine sexual behavior. In the Tfm mouse, Charest et al. (1991) demonstrated a single base deletion in the N-terminal domain of the androgen receptor, resulting in a frameshift mutation. Gaspar et al. (1991) independently demonstrated the same abnormality. They found no structural aberration in the coding region of the messenger by a series of RNase-protection assays. However, cell-free translation of RNAs transcribed in vitro from enzymatically amplified overlapping segments of exon 1 demonstrated a truncated receptor protein. Sequence analysis showed deletion of a single nucleotide in the hexacytidine stretch at position 1107-1112 altering the reading frame of the messenger and introducing 41 missense amino acids before the premature termination codon at position 1235-1237.

In female mice heterozygous for the Tfm gene, Takeda et al. (1987) demonstrated mosaicism in 2 androgen target tissues by steroid autoradiographic techniques, thus documenting X-linked inheritance. See also Takeda et al. (1987).

History

The variety of sex anomaly described in this entry has been of relatively long interest to geneticists, largely through the publication of Pettersson and Bonnier (1937), who concluded that the affected persons are genetic males. Dieffenbach (1912), an American geneticist, had pointed out the hereditary pattern. Morris (1953), in a classic paper, first used the term testicular feminization.