Residual Activity of Mutant Androgen Receptors ... - Oxford Journals

0021-972X/04/$15.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 89(11):5815–5822 Copyright © 2004 by The Endocrine Society doi: 10.1210/jc.2004-0709

Residual Activity of Mutant Androgen Receptors Explains Wolffian Duct Development in the Complete Androgen Insensitivity Syndrome SABINE E. HANNEMA, IAN S. SCOTT, JOHN HODAPP, HOWARD MARTIN, NICK COLEMAN, JOHN W. SCHWABE, AND IEUAN A. HUGHES Department of Pediatrics, University of Cambridge, Addenbrooke’s Hospital (S.E.H., J.H., H.M., I.A.H.), Cambridge, United Kingdom CB2 2QQ; Medical Research Council Cancer Cell Unit, Hutchison/Medical Research Council Research Center (I.S.S., N.C.), Cambridge, United Kingdom CB2 2XZ; and Medical Research Council Laboratory of Molecular Biology (J.W.S.), Cambridge, United Kingdom CB2 2QH Development of the Wolffian ducts (WD) into epididymides and vasa deferentia is dependent on testosterone. Patients with the complete androgen insensitivity syndrome (CAIS) are therefore not expected to develop these structures. However, WD derivatives have been described in cases of CAIS. It is thought that these may be remnants. This study assesses the degree of WD development in 33 patients with CAIS and investigates whether this development was androgen dependent. Epididymides and vasa deferentia were identified in 70% of patients with substitution mutations in the androgen receptor ligand-binding domain. They were more developed than epididymides and vasa deferentia from 16- to 20-wk-old male fetuses, suggesting that the WD had been stimulated to

T

HE ANDROGEN INSENSITIVITY syndrome (AIS) is a disorder caused by mutations in the androgen receptor (AR) gene, located at Xq11.2-q12. This leads to deficient masculinization in 46,XY individuals despite normal or increased androgen production by the testes. There is a wide variation in the degree of undermasculinization, ranging from infertility in an otherwise normal male (mild AIS), to ambiguous genitalia (partial AIS), to a completely female external phenotype [complete AIS (CAIS)]. Although most attention tends to focus on masculinization of the external genitalia, the internal genitalia are expected to be affected in a similar way, because stabilization of the Wolffian ducts (WD) is dependent on testosterone (1). In patients with CAIS, the lack of external masculinization suggests that the AR is completely unresponsive to androgens. The WD are therefore not expected to develop into the epididymides and vasa deferentia. Morris (2), in his classic review of 82 cases of CAIS, noted that rudimentary anlagen of the internal genitalia may be present, including spermatic ducts. Several reports have since confirmed, but not explained, this observation (3–5). Most observers believe that such Wolffian structures have Abbreviations: AIS, Androgen insensitivity syndrome; AR, androgen receptor; CAIS, complete androgen insensitivity syndrome; DBD, DNAbinding domain; DHT, 5␣-dihydrotestosterone; LBD, ligand-binding domain; WD, Wolffian duct. JCEM is published monthly by The Endocrine Society (http://www. endo-society.org), the foremost professional society serving the endocrine community.

grow, rather than failed to regress. Receptors with substitutions in the ligand-binding domain were normally expressed and showed residual response to androgens in transactivation assays. Patients with premature stop codons or frameshift mutations, which prevented androgen receptor expression, or DNA-binding domain mutations that abolished transcriptional activity did not have epididymides or vasa deferentia. We hypothesize that mutant receptors with residual activity in vitro respond to high local testosterone concentrations in vivo, thereby stimulating WD development. The classification of androgen insensitivity in such patients should be considered severe rather than complete. (J Clin Endocrinol Metab 89: 5815–5822, 2004)

failed to completely regress, as indicated by the use of terms such as remnants. Because none of the reports provides objective information on the degree of development, it remains unclear whether all WD derivatives identified are indeed vestigial. It is important to distinguish between remnants and well-developed WD in relation to whether androgen responsiveness is maintained to some degree in CAIS. True development of the WD despite the presence of an AR mutation severe enough to prevent masculinization of the external genitalia would clearly be unexpected based on the dogma derived from the studies by Jost (1). This study was designed to ascertain WD development in a large group of CAIS patients. A detailed histological analysis of gonadal material removed from 33 patients in whom an AR mutation had been identified was undertaken, and the findings were compared with internal genital development in normal male fetuses. We report the presence of well developed epididymides and vasa deferentia in a group of CAIS patients with substitution mutations in the AR ligandbinding domain (LBD). It is proposed that this shows evidence for residual activity of mutant ARs in vivo, and we suggest a further subclassification of androgen insensitivity. Subjects and Methods Subjects Thirty-three patients with CAIS were selected from the Cambridge Intersex Database. All had a 46,XY karyotype and normal female external genitalia, as assessed by experienced pediatric endocrinologists or gynecologists. In each case a mutation in the AR gene had been identified

5815

5816

J Clin Endocrinol Metab, November 2004, 89(11):5815–5822

Hannema et al. • WD Development in CAIS

by direct sequencing. Mutations were described using the numbering convention of Lubahn et al. (6). Duplication of exon B in patients 120, 229, and 231 was detected by RNA extraction from genital skin fibroblasts using TRIzol reagent (Invitrogen Life Technologies, Paisley, UK), followed by RT using SuperScript II reverse transcriptase (Invitrogen Life Technologies) and amplification of the AR cDNA using two primer pairs: AR1 forward, GATAGCTACTCCGGACCTTA; AR1 reverse, CATCTGGTCGTCCACGTGTA; AR2 forward, ACAGCTTGTACACGTGGTCA; and AR2 reverse, TAACAGGCAGAAGACATCTGA. PCR products of abnormal size were sequenced. Gonadectomy had been performed in all patients.

protein/lane) were resolved by SDS-PAGE and electrotransferred to Hybond ECL nitrocellulose membranes (Amersham Biosciences, Chalfont St. Giles, UK). AR protein was detected with primary antibody F 39.4.1 (Biogenex, San Ramon, CA) or AR(N-20) (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). Horseradish peroxidase-conjugated secondary antibodies (DakoCytomation, Ely, UK) and ECL Plus Blotting Detection Reagents (Amersham Biosciences) were used according to the manufacturers’ instructions. Telomerase-immortalized human genital skin fibroblasts, BJ1 (BD Clontech, Oxford, UK), served as a positive control.

Site-directed mutagenesis

Histological analysis Hematoxylin- and eosin-stained slides of the gonads from all patients were obtained for review. Archival slides from nine male fetuses aged between 16 and 20 wk gestation, retained from postmortem studies, were used as controls of normal WD development. The peritesticular area was examined for the presence of an epididymis or vas deferens. If tubes of Wolffian origin were found that were too underdeveloped or too few to fit the description of an epididymis or vas deferens (7), these were classified as Wolffian remnants. These tubules were similar to the vestigial WD that can be found in females (8). The observers were blinded to the genotype of the patients. Any case with contentious histology was examined by three individuals, one of whom is a consultant pathologist. Luminal diameter, epithelial cell height, and size of the muscular layer of the epididymis and vas deferens were measured using an Olympus BX41 microscope linked to image analysis software (AnalySIS, Olympus UK, London, UK).

Western blots Fibroblast lines established from genital skin biopsies, when available, were grown as previously described (9). Cell lysates (50 ␮g total

Mutations Y223X, deletion F582, G724D, and D864N were introduced into the wild-type AR cDNA vector pSVAR (10) by site-directed mutagenesis using a two-stage PCR as described previously (11). Mutations L700M and D732N were introduced into wild-type pSVAR using a QuikChange XL kit (Stratagene, La Jolla, CA) according to the manufacturer’s instructions. Sequencing confirmed the presence of the desired mutation and the absence of PCR-induced errors. A765T had previously been recreated (12). Nucleotide changes for each case are described in Table 1.

Transactivation assays COS-1 cells (American Type Culture Collection, Manassas, VA) were seeded in 12-well plates at a density of 105 cells/well in DMEM containing 10% fetal bovine serum, charcoal stripped as described previously (13), 100 U/ml penicillin, 0.1 mg/ml streptomycin, 2 mm lglutamine (all from Sigma-Aldrich Corp., St. Louis, MO), and 1⫻ MEM nonessential amino acids (Invitrogen Life Technologies). After 24 h cells were transfected with 15 ng pSVAR (wild-type or mutant), 500 ng pGRE-luciferase (an androgen- and glucocorticoid-responsive firefly

TABLE 1. Overview of subjects ID

Base change

Protein change

Epididymis

Vas deferens

WR

Age (yr)

AR

750 112 740 120 229 231 181 1286 1287 402 1365 902 453 429 432 167 537 1189 65 854 1273 374 801 861 376 395 396 502 604 101 423 150 178

C1031G Del. C1745 C1802A Dupl. exon B Dupl. exon B Dupl. exon B Del. TTC2106-8 G2206A G2206A G2424T A2431T G2445A T2460A G2491C G2491C C2529T GC2533-4AT G2556A G2655A Del. T2660 Del. T2660 G2702A G2702A C2853T G2926A C2930G and T2955C C2930G and T2955C G2952A A2953G G2958A G2958A G3027A C3081T

Y223X Frameshift Y480X Frameshift Frameshift Frameshift Del. F582 R615H R615H G688X D690V D695N L700M R710T R710T P723S G724D D732N A765T Frameshift Frameshift M780I M780I R831X R855H F856L and S865P F856L and S865P D864N D864G V866M V866M V889M L907F

⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫺ ⫺ ⫺ ⫹ ⫹ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫹ ⫺ ⫹

⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫹ ⫹ ⫹ ⫺ ⫹ ⫹ ⫺ ⫺ ⫺ ⫹ ⫹ ⫺ ⫹ ⫺ ⫺ ⫹ ⫺ ⫹ ⫺ ⫺ ⫹

⫹ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹ ⫹ ⫹ ⫺ ⫺ ⫹ ⫺ ⫺ ⫺ ⫺ ⫹ ⫺ ⫺ ⫹ ⫺

2 2 18 18 12 14 11 4 0.17 0.17 13 8 2 11 8 0.25 0.5 1 0.58 4 17 0.08 15 1 1 10 10 14 15 17 12 2 4

⫺ np np ⫾ ⫾ np ⫹ np np ⫺ np np ⫹ np ⫹ np np ⫹ ⫹ np np np np np np np np ⫹ np np ⫹ np np

Comments

Sister 231 Sister 229 Sister 1287 Sister 1286

Sister 432 Sister 429

Cousin 801 Cousin 374 Twin of 396 Twin of 395

WR, Wolffian remnants; Age, age at gonadectomy; AR, AR expression in genital skin fibroblasts; Del., deletion; Dupl., duplication; ⫹, present; ⫺, absent; ⫾ indicates a faint band detected in some experiments; np, no Western blot performed.



5817

luciferase reporter vector) (14), 25 ng pRLTK, a Renilla luciferase vector used as an internal control for transfection efficiency (Promega Corp., Madison, WI), and 1.22 ␮g salmon sperm DNA/well using calcium precipitation. Sixteen hours after transfection, fresh medium with 0, 1, or 10 nm mibolerone (17-hydroxy-7,17-dimethylestr-4-en-3-one; Steraloids, Inc., Wilton, NH) was added. After an additional 24 h, cells were harvested in lysis buffer [25 mm glycylglycine (pH 7.8), 15 mm MgSO4, 4 mm EGTA, 1% Triton X, and 1 mm dithiothreitol]. Luciferase assays were performed with reagents from NanoLight Technology (Pinetop, AZ), and the ratio of firefly to Renilla luciferase was measured using a Turner TD-20/20 luminometer (Turner Designs, Sunnyvale, CA). Three independent experiments were performed, each in triplicate. In each experiment, values were normalized to the average ratio obtained for wild-type AR with 10 nm mibolerone.

Statistical analysis Comparisons of luminal diameter, epithelial cell height, and size of the muscular layer of epididymis and vas deferens in CAIS patients and normal male fetuses were assessed by the Mann-Whitney U test. Differences were considered significant at P ⱕ 0.05.

Ethics Local research ethics committee approval and informed consent were obtained for the use of patient samples as part of a sexual development disorders research program. Separate local research ethics committee approval was obtained for the use of slides of fetal gonads retained from postmortem examinations with parental consent.

Results Subjects

All 33 patients included in the study had normal female external genitalia, with no sign of clitoromegaly. Many patients were prepubertal at the time of the study, but of those who were postpubertal, information about pubic hair development was available in three cases. Patient 604 (with mutation D864G) showed no pubic hair, whereas patients 101 (V866M) and 502 (D864N) had pubic hair development (Tanner stages P5 and P3– 4, respectively). Testes were inguinal in 20 cases, abdominal in five cases, unilaterally inguinal and unilaterally abdominal in five cases, and labial in one case, and the position was unknown in two patients. All had undergone gonadectomy at ages ranging from 1 month to 19 yr (median, 8 yr; Table 1). Twenty-six different AR mutations had been identified in these 33 patients (Table 1). Family history was positive in 76% of individuals, negative in 15%, and unknown in 9%. Histological analysis

Screening of the paratesticular area revealed the presence of well developed epididymides and/or vasa deferentia in 14 cases (42%; Table 1). The epididymides showed the typical characteristics of a highly coiled duct lined by tall columnar cells, whereas the vasa deferentia had a typical slightly folded mucosa and thick muscular coat consisting of three layers (Fig. 1, A–D). These characteristics appeared to have developed to a greater extent in the CAIS patients than in the normal male 16- to 20-wk-old fetuses used for comparison (Fig. 1, E and F, note the difference in magnification). In nine cases (27%) no WD derivatives were identified. The remaining 10 cases (30%) had Wolffian remnants. These structures resembled fetal WD, but they were fewer in number (Fig. 1, G and H).

FIG. 1. WD derivatives in CAIS patients and in normal male 16- to 20-wk-old fetuses. A, Epididymis in patient 178 (with mutation L907F). B, Vas deferens in patient 1365 (D690V). C, Epididymis in patient 1189 (D732N). D, Vas deferens in patient 429 (R710T). E, Epididymis in 20-wk-old fetus. F, Vas deferens in 20-wk-old fetus. G, Wolffian remnants in patient 150 (V889M). H, Wolffian remnants in patient 65 (A765T). Note the highly coiled ductus epididymis (A and C) and the well developed muscular layer of the vas deferens (B and D) in CAIS patients. Scale bars, 200 ␮m.

To quantify the degree of development, lumen diameter, epithelial cell height, and size of the muscular layer of both the epididymis and vas deferens were measured. All parameters were significantly greater in CAIS patients than in normal male fetuses (Fig. 2). Genotype-phenotype correlation

There was a clear difference between the type of AR mutation found in CAIS patients with well developed WD and those without well developed WD (Fig. 3). Frameshift mutations, premature stop codons, and mutations in the DNA-binding domain (DBD) were associated with the absence of well-developed WD, whereas all patients with a well developed epididymis and/or vas deferens had a

5818



FIG. 2. Morphometric comparison of the epididymis and vas deferens in CAIS patients and normal male 16- to 20-wk-old fetuses. A–C, Epididymis in CAIS patients (n ⫽ 12) and normal male 16- to 20-wk-old fetuses (n ⫽ 8). A, Lumen diameter (P ⬍ 0.001). B, Epithelial cell height (P ⬍ 0.001). C, Muscular layer (P ⬍ 0.001). D–F, Vas deferens in CAIS patients (n ⫽ 12) and normal male 16- to 20-wk-old fetuses (n ⫽ 3). D, Lumen diameter (P ⫽ 0.014). E, Epithelial cell height (P ⫽ 0.009). F, Muscular layer (P ⫽ 0.009). Boxes represent the median with the interquartile ranges. F, Outliers. NMF, Normal male fetuses.

or deletions (Table 1). No AR could be detected in fibroblasts from patients with premature stop codons. Two patients had a duplication of the first DBD exon (exon B), thereby creating a frameshift. In fibroblasts from these patients, a very faint band was detected in some experiments, suggesting the presence of a minimal amount of normally spliced AR. Functional analysis of mutant ARs

FIG. 3. Comparison between AR mutations found in CAIS patients with and without WD development. Diagrammatic representation of the AR, divided into the N terminus (N-term), DBD, and LBD. The location of AR mutations is shown in patients without WD development (WD⫺) and in patients with WD development (WD⫹). Note that more severe mutations are associated with the absence of WD development, whereas only milder substitution mutations in the LBD are found in the group with WD development. ⴱ, Stop codon; #, frameshift; ‚, deletion; 2, substitution (dashed arrows indicate two substitution mutations found in the same patient).

single amino acid substitution in the AR LBD. However, some single amino acid substitution mutations in the LBD (D695N, A765T, D864G, and V889M) were associated with the absence of an epididymis or vas deferens, as was the double mutation F856L and S865P. The type of mutation and the presence of WD were not related to the position of the gonads. AR expression in genital skin fibroblasts

Western blots performed on genital skin fibroblasts from 11 patients with several types of mutations showed normal AR expression in those with single amino acid substitutions

The results of transactivation assays showed that the activity of all mutant ARs was impaired in the presence of 1 nm mibolerone, confirming that these mutations cause androgen insensitivity (Fig. 4A). In the presence of 10 nm mibolerone, four of the five ARs with LBD substitution mutations (L700M, G724D, D732N, and D864N) had transcriptional activity similar to that of the wild-type AR. However, a premature stop codon (Y223X), a single amino acid deletion in the DBD (deletion F582), and one of the five substitution mutations in the LBD (A765T) completely abolished transcriptional activity. This was not due to the absence of AR protein, as a Western blot of the transfected cells showed that all mutant ARs were expressed (Fig. 4B). The level of activity of each mutant AR varied between experiments and appeared to correlate with variation in expression levels of the transfected ARs. However, the four mutant ARs that showed residual activity did so consistently in all experiments. Discussion

The results of this study show that the epididymides and/or vasa deferentia are well developed in 42% of this series of patients with CAIS. This is surprising in view of the


FIG. 4. Androgen-dependent transcriptional activity of seven mutant ARs. A, COS-1 cells were transiently transfected with plasmids encoding wild-type or mutant ARs (15 ng) and the reporter plasmid GRE-luciferase (500 ng). After a 24-h incubation with 0, 1, or 10 nM mibolerone, cells were harvested and assayed for luciferase activity. Shown are results from a representative experiment. The activity of each mutant is expressed as a percentage of wild-type receptor activity; error bars indicate 1 SD. B, Western blot of lysates from transfected cells shows the expression of all mutant ARs. The blot was probed with antibody AR(N-20).

androgen dependence of the development of these structures (1). WD derivatives have previously been found in CAIS patients, and it has been suggested that they represent remnants that have failed to completely regress (4). If that were the case, then WD development should be similar to that found in the normal male fetus before the onset of testosterone production at 9 wk gestation (15, 16). We have compared the epididymides and vasa deferentia in CAIS patients with those found in male fetuses of 16 –20 wk gestation, when WD have differentiated (16). The epididymis and vas deferens in CAIS patients had developed well beyond the stage observed in the control fetal tissues. Compared with data for normal children (17), lumen diameter and epithelial height of the epididymis were within the 2.5 sem range in 83% of cases. Lumen diameter and epithelial height in the normal male fetuses were within the 2.5 sem range reported for the prenatal epididymis (17) in all cases, indicating that our control measurements are comparable to those reported by others. The fact that the WD in this subgroup of CAIS patients are more developed than control fetal WD and in some respects similar to those in normal subjects suggests that not only have they failed to regress, but they have been stimulated to grow. As epididymides and vasa deferentia were only found in patients with single amino acid substitutions in the LBD, residual function of these receptors may be responsible for stimulating WD development. We confirmed by Western blotting that ARs with single amino acid substitutions were expressed in genital skin fibroblasts. In contrast, more severe mutations, such as premature stop codons, associated with the absence of an epididymis or vas deferens, lead to undetectable AR levels. Others have found a similar relationship between the type of mutation and AR expression (4, 18). The amount of normally spliced AR in genital skin fibroblasts


5819

from patients with a duplication of exon B was close to the detection limit of the assay; it was only visible in some experiments. This minimal amount of AR is presumably not sufficient to cause masculinization of the internal or external genitalia. Functional studies showed that a premature stop codon (Y223X) and an amino acid deletion from the DBD (deletion F582) abolished transcriptional activity despite expression of the mutant ARs in COS-1 cells. One of the five LBD mutants tested (A765T) also failed to respond to mibolerone. All three mutations were associated with the absence of epididymides and vasa deferentia. However, the other four ARs with substitution mutations in the LBD (L700M, G724D, D732N, and D864N), found in patients with an epididymis and vas deferens, showed activity similar to that of the wild-type AR in the presence of 10 nm mibolerone. With low concentrations of androgen, the activity of the mutants was reduced compared with that of the wild-type AR, suggesting that although the mutations do affect AR function, the defect can be overcome in the presence of high concentrations of androgen. The synthetic androgen mibolerone was used for these experiments rather than the natural ligand testosterone, because testosterone is metabolized by COS-1 cells (19). Consequently, high concentrations of testosterone are necessary to induce AR activity in COS-1 cells. However, 1 ␮M testosterone as well as 100 nm 5␣-dihydrotestosterone (DHT) can activate the ARs with substitution mutations in the LBD to the same extent as 10 nm mibolerone (Hannema, S. E., unpublished observation), indicating that this is not a ligandspecific effect. In this study seven mutant ARs were recreated for functional analysis. Other mutations identified in the 33 patients included in this study have been recreated and studied by other groups. Mutant R615H, affecting the DBD, is inactive in the presence of 10 nm DHT (20) and is associated with the absence of epididymides and vasa deferentia. Similarly, two LBD substitution mutations, D864G and S865P, which abolish AR activity in response to androgens (11, 21) also lead to absent WD development. The following LBD mutations, most of which were associated with well developed WD, were found to have residual activity: D695N (28% of wildtype activity with 10 nm R1881, a synthetic androgen) (13), M780I (100% activity with 3 nm R1881) (14), R855H (79% activity with 2 nm mibolerone) (22), V866M (79% activity with 4 nm mibolerone) (23), V889M (89% activity with 50 nm mibolerone) (24), and L907F (38% activity with 10 nm mibolerone) (12). A comparison of the mutant AR activity levels is not possible because of nonuniform experimental conditions between the various studies, but it is clear that the mutant ARs are androgen responsive. No functional studies have been performed on D690V, R710T, and P723S. Taken together these observations strongly suggest that most mutant ARs with LBD substitution mutations may respond to androgens in vivo to stimulate WD development in patients who are assumed to be completely insensitive to androgens. Whether this residual AR activity is also responsible for pubic hair development is not clear. Quigley et al. (25) and Sinnecker et al. (26) suggested that the presence of pubic hair should be regarded as evidence of some degree of androgen

5820


responsiveness and proposed a separate category, AIS grade 6 and type 5a, respectively, for patients with female external genitalia and pubic hair. However, others do not take pubic hair development into account when determining the degree of masculinization (27). Boehmer et al. (4) reported a patient with a frameshift mutation with pubic hair Tanner stage P3. It is difficult to imagine residual AR activity with this type of mutation; therefore, it seems that pubic hair may develop independently from AR activity (4). In the three cases from whom we have information on pubic hair development, there is a correlation between the presence of well developed WD and the presence of pubic hair. However, because of the small number of cases from whom this information is available, it is not possible to say whether this is true in general. It is interesting that only three LBD substitution mutations completely abolished AR activity in vivo and in vitro (A765T, D864G, and S865P). This suggests that these mutations have a particularly disruptive effect on the protein structure or on critical functional regions within the LBD. Examination of the crystal structures (28, 29) and model (30) of the AR LBD in part helps to rationalize the severity of these defects. However, it is inherently difficult to clearly differentiate between mutants that show a low level of residual activity and those that are totally inactive. Significantly, none of the three amino acids causing total loss of function is in direct contact with the ligand. They are, however, all in highly ordered parts of the structure. Alanine 765 is completely buried and is in contact with residues in the loop between helixes 1 and 3 and in helix 3 (Fig. 5A). The change from alanine to the larger and more polar threonine is stereochemically unfavorable and may perturb the structure of the ␤-sheet on one side of the ligand-binding cavity. This might, in turn, affect ligand binding affinity or on/off rates, or conceivably perturb the folding of the LBD. Mutation of aspartic acid 864 to glycine (patient 604), but not asparagine (patient 502), also results in a total loss of function. This residue makes hydrogen bonds to the backbone amides of phenylalanine 916 and histidine 917, thus stabilizing the very C terminus of the LBD (Fig. 5B). Mutation to glycine, but not to asparagine, would result in complete loss of these interactions and destabilization of the C terminus of the LBD. This suggests that this region of the protein has a critical functional role that remains to be fully understood. Mutation of the adjacent residue serine 865 to proline is also severely deleterious to protein function. The role of this residue is not as clear; however, we have noted previously that the proline substitution may perturb helix 10/11, and this might explain the severity of the phenotype (11). Mutations associated with residual function seem to be less severe for two main reasons. Some are relatively conservative amino acid changes. Others are located in loops on the surface of the AR. They are therefore expected to be less disruptive to the structure and function of the AR. It is noteworthy that WD have developed in some CAIS patients, whereas the external genitalia are not masculinized. This may be due to differences in local androgen concentrations to which the tissues have been exposed. The WD are attached to the testis, and several lines of evidence suggest that testosterone is secreted directly into the WD. In unilaterally castrated rabbit fetuses, the WD is only maintained on


FIG. 5. Locations of mutations A765T, D864G, and S865P in the AR LBD. A, Ribbon diagram showing alanine 765, located in a tightly packed area of the AR LBD (PDB code, 1I37) (29) as part of a hydrophobic cluster. It is not in direct contact with the ligand (DHT; shown in purple), but interacts with residues in the loop between helixes 1 and 3 and in helix 3. Mutation of the alanine to the larger and more polar threonine may cause substantial perturbation. Side chains of amino acids 686 – 687, 707, and 763–769 are indicated on the structure. B, Ribbon diagram showing the location of aspartic acid 864 and serine 865. Aspartic acid 864 forms hydrogen bonds with the amides of residues 916 and 917 (shown as dashed lines). These interactions would be lost if the aspartic acid was mutated to glycine, thereby destabilizing the very C terminus of the LBD. Serine 865 is situated in helix 10/11, which may be perturbed by mutation of this residue to a proline. Side chains of amino acids 861, 864, 865, 867 and 913–917 are indicated on the structure. The figure was prepared using Pymol (www.pymol.org).

the unoperated side (1). Furthermore, when testicular insufficiency is induced in fetal rabbits by decapitation before sexual differentiation, the degree of underdevelopment increases along a gradient extending from the testis along the genital ducts (1). Additional evidence for diffusion of testosterone down the WD is provided by studies in which fluorescent androgen was injected into fetal mouse testis, showing rapid localization of fluorescence in the proximal and subsequently distal WD (31). Testicular testosterone concentrations are very high in the fetus during the phase of WD stabilization. Between 12 and 15 wk gestation the testis contains between 500 and 5200 pg testosterone/mg wet tissue (32). If tissue density is assumed to be approximately 1 g/ml, then this equals about 1.7–18


␮M. In infancy, the epididymal testosterone content is 30% that of testicular testosterone (33). As this percentage declines with age (34), the epididymal concentration may be even higher in the fetus. Thus, the local concentration of testosterone to which the WD are exposed is likely to be in the micromolar range. The developing external genitalia, in contrast, are exposed to systemic levels of androgens. Serum testosterone reaches a peak in the fetus between 16 and 18 wk gestation when levels are 200 – 600 ng/dl (35), equaling approximately 6.9 –21 nm. Although testosterone is converted to DHT, which is 10 times more potent (19), in the external genitalia, AR stimulation is likely to be less intense than in the WD, where testosterone concentrations are probably more than 100-fold higher. However, other factors may play a role in the difference between development of the internal and external genitalia in CAIS patients, such as AR coactivators that may be specifically expressed in the WD. Two patients in this series have residual mutant AR activity, but the WD do not seem to have developed (patient 150 with mutation V889M and patient 902 with mutation D695N). It is possible that epididymides and vasa deferentia were present, but were not included in the tissue samples available for review. Sometimes an epididymis, but not a vas deferens, or a vas deferens, but not an epididymis, was found (Table 1). Because these structures are contiguous in development, it suggests that sampling may be incomplete in some cases. However, the absence of epididymides and vasa deferentia may be due to other factors. Testicular testosterone concentrations at the time of WD stabilization show a wide variation between individuals (500 –5200 pg/mg tissue) (32). Epididymal testosterone concentrations are closely correlated to testicular concentrations (33); therefore, a similar wide range may be expected in testosterone concentrations in the WD. If an AR mutation is present, concentrations toward the lower end of the range may not be sufficient to stimulate growth of the WD. Alternatively, other unknown factors that vary between individuals may influence WD development. These may be the same factors that underlie variations in the degree of undermasculinization of the external genitalia between individuals, and even siblings, who have the same AR mutation (36). The finding of epididymides and vasa deferentia in patients diagnosed with CAIS poses a problem with the classification of AIS. Because there is evidence of AR activity in vivo, it is incorrect to say that these patients are completely insensitive to androgens. However, the term partial androgen insensitivity is historically associated with partial masculinization of the external genitalia. We therefore suggest using the term severe androgen insensitivity syndrome to describe patients with normal female external genitalia but male internal genitalia. Acknowledgments We are grateful to Dr. A. O. Brinkmann (Erasmus University, Rotterdam, The Netherlands) for generously providing the wild-type AR vector, to Dr. A. Allera (Friedrich Wilhelms University, Bonn, Germany) for providing the GRE-luciferase vector, to Drs. M. Williams and P. G. Ransley for providing patient slides, to T. Bunch for technical assistance, and to Drs. J. Murphy and J. Jaä¨skelaïnen for advice.


5821

Received April 23, 2004. Accepted August 12, 2004. Address all correspondence and requests for reprints to: Prof. I. A. Hughes, Department of Pediatrics, University of Cambridge, Addenbrooke’s Hospital, Hills Road, Cambridge, United Kingdom CB2 2QQ. E-mail: [email protected]. This work was supported by the Birth Defects Foundation. S.E.H. is a recipient of a 2001 Nuffic Talents Award (Dutch Ministry of Education) and a Gates Cambridge Scholarship.

References 1. Jost A 1953 Problems or fetal endocrinology: the gonadal and hypophyseal hormones. Recent Prog Horm Res 8:379 – 418 2. Morris JM 1953 The syndrome of testicular feminization in male pseudohermaphrodites. Am J Obstet Gynecol 65:1192–1211 3. Hodapp J 1999 The presence of Wolffian duct structures in patients with complete androgen insensitivity. Horm Res 51(Suppl 2):76 (Abstract) 4. Boehmer AL, Bruggenwirth H, van Assendelft C, Otten BJ, VerleunMooijman MC, Niermeijer MF, Brunner HG, Rouwe CW, Waelkens JJ, Oostdijk W, Kleijer WJ, van der Kwast TH, de Vroede MA, Drop SL 2001 Genotype versus phenotype in families with androgen insensitivity syndrome. J Clin Endocrinol Metab 86:4151– 4160 5. Bale PM, Howard NJ, Wright JE 1992 Male pseudohermaphroditism in XY children with female phenotype. Pediatr Pathol 12:29 – 49 6. Lubahn DB, Joseph DR, Sar M, Tan J, Higgs HN, Larson RE, French FS, Wilson EM 1988 The human androgen receptor: complementary deoxyribonucleic acid cloning, sequence analysis and gene expression in prostate. Mol Endocrinol 2:1265–1275 7. Trainer TD 1997 Testis and excretory duct system. In: Sternberg SS, ed. Histology for pathologists. 2nd ed. Philadelphia: Lippincott-Raven; 1019 –1037 8. Robboy SJ, Bentley RC 1997 Vagina. In: Sternberg SS, ed. Histology for pathologists. 2nd ed. Philadelphia: Lippincott-Raven; 867– 877 9. Batch JA, Williams DM, Davies HR, Brown BD, Evans BA, Hughes IA, Patterson MN 1992 Androgen receptor gene mutations identified by SSCP in fourteen subjects with androgen insensitivity syndrome. Hum Mol Genet 1:497–503 10. Brinkmann AO, Faber PW, van Rooij HC, Kuiper GG, Ris C, Klaassen P, van der Korput JA, Voorhorst MM, van Laar JH, Mulder E, Trapman J 1989 The human androgen receptor: domain structure, genomic organization and regulation of expression. J Steroid Biochem 34:307–310 11. Mongan NP, Jaä¨skelaïnen J, Green K, Schwabe JW, Shimura N, Dattani M, Hughes IA 2002 Two de novo mutations in the AR gene cause the complete androgen insensitivity syndrome in a pair of monozygotic twins. J Clin Endocrinol Metab 87:1057–1061 12. Bevan CL, Hughes IA, Patterson MN 1997 Wide variation in androgen receptor dysfunction in complete androgen insensitivity syndrome. J Steroid Biochem Mol Biol 61:19 –26 13. Ris-Stalpers C, Trifiro MA, Kuiper GG, Jenster G, Romalo G, Sai T, van Rooij HC, Kaufman M, Rosenfield RL, Liao S, Schweikert H-U, Trapman J, Pinsky L, Brinkmann AO 1991 Substitution of aspartic acid-686 by histidine or asparagine in the human androgen receptor leads to a functionally inactive protein with altered hormone-binding characteristics. Mol Endocrinol 5:1562– 1569 14. Knoke I, Allera A, Wieacker P 1999 Significance of the CAG repeat length in the androgen receptor gene (AR) for the transactivation function of an M780I mutant AR. Hum Genet 104:257–261 15. Hughes IA 2001 Minireview: sex differentiation. Endocrinology 142:3281–3287 16. Sultan C, Paris F, Terouanne B, Balaguer P, Georget V, Poujol N, Jeandel C, Lumbroso S, Nicolas JC 2001 Disorders linked to insufficient androgen action in male children. Hum Reprod Update 7:314 –322 17. De Miguel MP, Marino JM, Martinez-Garcia F, Nistal M, Paniagua R, Regadera J 1998 Pre- and post-natal growth of the human ductus epididymidis. A morphometric study. Reprod Fertil Dev 10:271–277 18. Avila DM, Wilson CM, Nandi N, Griffin JE, McPhaul MJ 2002 Immunoreactive AR and genetic alterations in subjects with androgen resistance and undetectable AR levels in genital skin fibroblast ligand-binding assays. J Clin Endocrinol Metab 87:182–188 19. Deslypere JP, Young M, Wilson JD, McPhaul MJ 1992 Testosterone and 5␣-dihydrotestosterone interact differently with the androgen receptor to enhance transcription of the MMTV-CAT reporter gene. Mol Cell Endocrinol 88:15–22 20. Mowszowicz I, Lee HJ, Chen HT, Mestayer C, Portois MC, Cabrol S, Mauvais-Jarvis P, Chang C 1993 A point mutation in the second zinc finger of the DNA-binding domain of the androgen receptor gene causes complete androgen insensitivity in two siblings with receptor-positive androgen resistance. Mol Endocrinol 7:861– 869 21. De Bellis A, Quigley CA, Cariello NF, El-Awady MK, Sar M, Lane MV, Wilson EM, French FS 1992 Single base mutations in the human androgen receptor gene causing complete androgen insensitivity: rapid detection by a modified denaturing gradient gel electrophoresis technique. Mol Endocrinol 6:1909 –1920

5822


22. Marcelli M, Zoppi S, Wilson CM, Griffin JE, McPhaul MJ 1994 Amino acid substitutions in the hormone-binding domain of the human androgen receptor alter the stability of the hormone receptor complex. J Clin Invest 94:1642–1650 23. Kazemi-Esfarjani P, Beitel LK, Trifiro M, Kaufman M, Rennie P, Sheppard P, Matusik R, Pinsky L 1993 Substitution of valine-865 by methionine or leucine in the human androgen receptor causes complete or partial androgen insensitivity, respectively with distinct androgen receptor phenotypes. Mol Endocrinol 7:37– 46 24. De Bellis A, Quigley CA, Marschke KB, El-Awady MK, Lane MV, Smith EP, Sar M, Wilson EM, French FS 1994 Characterization of mutant androgen receptors causing partial androgen insensitivity syndrome. J Clin Endocrinol Metab 78:513–522 25. Quigley CA, De Bellis A, Marschke KB, El-Awady MK, Wilson EM, French FS 1995 Androgen receptor defects: historical, clinical, and molecular perspectives. Endocr Rev 16:271–321 26. Sinnecker GHG, Hiort O, Nitsche EM, Holterhus P-M, Kruse K 1997 Functional assessment and clinical classification of androgen sensitivity in patients with mutations of the androgen receptor gene. Eur J Pediatr 156:7–14 27. Ahmed SF, Cheng A, Dovey L, Hawkins JR, Martin H, Rowland J, Shimura N, Tait AD, Hughes IA 2000 Phenotypic features, androgen receptor binding, and mutational analysis in 278 clinical cases reported as androgen insensitivity syndrome. J Clin Endocrinol Metab 85:658 – 665 28. Matias PM, Donner P, Coelho R, Thomaz M, Peixoto C, Macedo S, Otto N, Joschko S, Scholz P, Wegg A, Ba¨sler S, Scha¨fer M, Egner U, Carrondo MA 2000 Structural evidence for ligand specificity in the binding domain of the human androgen receptor. J Biol Chem 275:26164 –26171


29. Sack JS, Kish KF, Wang C, Attar RM, Kiefer SE, An Y, Wu GY, Scheffler JE, Salvati ME, Krystek SR, Weinmann R, Einspahr HM 2001 Crystallographic structures of the ligand-binding domains of the androgen receptor and its T877A mutant complexed with the natural agonist dihydrotestosterone. Proc Natl Acad Sci USA 98:4904 – 4909 30. Poujol N, Wurtz J-M, Tahiri B, Lumbroso S, Nicolas J-C, Moras D, Sultan C 2000 Specific recognition of androgens by their nuclear receptor. A structurefunction study. J Biol Chem 275:24022–24031 31. Tong SY, Hutson JM, Watts LM 1996 Does testosterone diffuse down the Wolffian duct during sexual differentiation? J Urol 155:2057–2059 32. Reyes FI, Winter JS, Faiman C 1973 Studies on human sexual development. I. Fetal gonadal and adrenal sex steroids. J Clin Endocrinol Metab 37:74 –78 33. Bidlingmaier F, Dorr HG, Eisenmenger W, Kuhnle U, Knorr D 1983 Testosterone and androstenedione concentrations in human testis and epididymis during the first two years of life. J Clin Endocrinol Metab 57:311–315 34. Leinonen P, Hammond GL, Vihko R 1980 Testosterone and some of its precursors and metabolites in the human epididymis. J Clin Endocrinol Metab 51:423– 428 35. Reyes FI, Boroditsky RS, Winter JS, Faiman C 1974 Studies on human sexual development. II. Fetal and maternal serum gonadotropin and sex steroid concentrations. J Clin Endocrinol Metab 38:612– 617 36. Rodien P, Mebarki F, Mowszowicz I, Chaussain JL, Young J, Morel Y, Schaison G 1996 Different phenotypes in a family with androgen insensitivity caused by the same M780I point mutation in the androgen receptor gene. J Clin Endocrinol Metab 81:2994 –2998

JCEM is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.