Photoaffinity labelling with [3H]R1881 of androgen receptors from human cultured genital fibroblasts of normal individuals and patients with androgen receptor abnormalities

Androgen receptors were partially purified by affinity chromatography of cytosols prepared from cultured genital fibroblasts of normal individuals and patients with androgen receptor abnormalities. Partially purified receptors were covalently labelled with [3H]R1881 (triated methyltrienolone) by ultraviolet photoactivation. Gel electrophoresis of cytosols from normal individuals showed a single radioactive peak, Mr approximately 97 K. Cytosols from patients with decreased nuclear transfer showed a similar peak, Mr approximately 97 K; cytosols from patients with partial androgen insensitivity (PAIS) or receptor positive complete androgen insensitivity (CAIS, R+) showed a peak Mr approximately 30 K-43 K.     

Photoaffinity labelling of androgen receptors from human foreskin and cultured genital fibroblasts with tritiated methyltrienolone [3H]R1881

Androgen receptors were partially purified by affinity chromatography of cytosols prepared from either normal foreskins or normal cultured fibroblasts. Partially purified receptors were covalently labelled with [3H]R1881 (tritiated methyltrienolone) by ultraviolet photoactivation. Gel electrophoresis of cytosols from both sources showed two peaks of specific binding, Mr approximately 40 K and approximately 85 K, under denaturing conditions. Photoaffinity labelling of the human androgen receptor may thus provide a useful tool in further studies on disorders of androgenization.     

Isoelectric focusing and 2D electrophoresis of the human androgen receptor.

Nuclear androgen receptors from cultured genital skin fibroblasts were analyzed by non-denaturing isoelectric focusing (IEF) in ultrathin polyacrylamide gels before and after photoaffinity labeling with [3H]methyltrienolone. Both reversibly and covalently labeled receptors focused at pH 5.28 +/- 0.20 when extracted from nuclei with high salt. Lowering of the salt concentration yielded, in both cases, a second species which focused at pH 7.16. This species became predominant when nuclei were sonicated in IEF sample buffer containing no salt, even after extensive nucleic acid digestion. Low salt cytosols from both prostate and foreskin focused as a single peak of pI: 4.93 +/- 0.31 which remained unchanged when KCl was added to the cytosol up to a concentration of 0.6 M. SDS-polyacrylamide gel electrophoresis of photoaffinity labeled receptors revealed labeled proteins with Mw 90-95 kDa. Two-dimensional electrophoresis of photoaffinity labeled nuclear receptors, extracted in low or high salt, showed that the two isoforms (pI 5.28 and 7.16) contain the same steroid-binding subunit with Mw 90-95 kDa. Nuclear receptors from 4 patients with the receptor positive form of the Complete Androgen Insensitivity Syndrome (CAIS, Rc+) were analyzed by non-denaturing IEF: a single species was observed, focusing at pH 6.0 whether in high or low salt conditions. These results indicate that the nuclear androgen receptor is an acidic protein with pI 5.28 and Mw 90-95 kDa under maximum protein dissociation conditions. When extracted under low salt conditions, it can be isolated in a neutral form (pI 7.16) suggesting its association with a nuclear protein. Receptors of (CAIS, Rc+) patients have an abnormal charge and show no pI shift upon lowering of the salt concentration suggesting that this shift could be a significant step in the mechanism of action of androgens.     

Partial purification and characterisation of the human skin fibroblast androgen receptor: detection of abnormal receptor complexes in cells from patients with androgen insensitivity syndromes

After incubation of hGSF with [3H]5 alpha-dihydrotestosterone, 17 beta-hydroxy-7 alpha, 17 alpha-dimethyl-4-estrene-3-one, or 17 beta-hydroxy-17 alpha-methyl-4,9,11-estrien-3-one, androgen-receptor complexes were extracted with 0.5 M KCl and precipitated by 35% ammonium sulphate. Receptor complexes from control hGSF sedimented at approximately 4S on linear 5-20% sucrose gradients. The 4S peak was diminished or absent in cells from androgen insensitive patients exhibiting absent, deficient or unstable binding of androgens in intact hGSF. This procedure may be a useful means of distinguishing quantitative and qualitative defects in androgen binding to receptor, since one cell line found to have normal levels of androgen receptor complexes in whole cell assays had a profile resembling that of receptor negative cells on sucrose gradients. The complexes from one patient with complete androgen insensitivity having normal androgen binding in intact hGSF were indistinguishable from control complexes after sucrose gradient analysis and ADP-Sepharose chromatography. Receptor complexes were eluted from the ADP-Sepharose between 0.5-1.0 M KCl. HPLC-gel filtration of androgen receptor complexes at 22 degrees C revealed two peaks, the larger had a Mr of 60-65K, Stokes radius of 3.16 nm and a frictional ratio between 1.21 and 1.43. The second peak, Mr of 15K, was believed to represent a fragment of the receptor containing the steroid binding domain. On gel filtration at 22 degrees C the complexes from a patient with partial androgen insensitivity, who showed a diminished 4S receptor peak on sucrose gradients, revealed only the small "meroreceptor" fragment, suggesting that the mutation in this individual might render the androgen receptor more susceptible to proteolysis in vitro.     

The 56 kDa protein of human genital skin fibroblasts is identical to that radiolabelled by [3H]dihydrotestosterone 17 beta-bromoacetate

Analysis of soluble proteins from human genital skin fibroblasts by two-dimensional polyacrylamide gel electrophoresis reveals an abundant protein doublet of mol. wt 56,000 with isoelectric points (pI) of 6.7 and 6.5. This protein is absent in non-genital skin fibroblasts as well as in genital skin fibroblasts of most patients with complete forms of androgen insensitivity. The protein specifically binds androgen. A protein of similar estimated molecular weight (58,000) from human genital skin fibroblasts has recently been found to be covalently radiolabelled by the affinity ligand dihydrotestosterone 17 beta-bromoacetate (DHT-BA). In the present study these proteins have been found to be indistinguishable on one- and two-dimensional gel electrophoresis. Antibodies raised against the 56 kDa pI 6.7/6.5 protein also recognized the protein covalently radiolabelled by DHT-BA. A third protein of estimated mol. wt 59,000 has been found to be associated with several steroid hormone receptor complexes but has no known ligand binding activity. This protein was found to be clearly separable from the 56/58 kDa protein on two-dimensional gel electrophoresis as it has a more acidic pI of approximately 5.4. Furthermore, antibodies against the 59 kDa protein do not recognize the 56 kDa species, and vice versa.     

The 56 kDa androgen binding protein is an aldehyde dehydrogenase.

We have described a 56 kDa protein from genital skin fibroblasts that specifically binds androgen and that is generally not expressed in genital skin fibroblasts from patients with androgen insensitivity due to genetic defects of the androgen receptor. We have isolated a partial cDNA clone for the 56 kDa protein from an expression library of genital skin fibroblasts. In vitro translation of message selected with this clone faithfully produces the 56 kDa protein which can be immuneprecipitated with an anti-56 kDa antiserum. Northern blots probed with this clone show a 2.2 kb message, which parallels the expression of the 56 kDa protein. The sequence of this 998bp clone is identical to human liver aldehyde dehydrogenase 1, the cytoplasmic isoenzyme. On activity gels of genital skin fibroblast cytosol covalently labelled with androgen, aldehyde dehydrogenase activity comigrates with the single band labelled specifically with androgen. Thus, the 56 kDa androgen binding protein is an aldehyde dehydrogenase, which is prominently expressed in normal genital skin fibroblasts, but not in non-genital skin fibroblasts.     

Identification of proteins involved in intracellular copper metabolism. Low levels of a approximately 48-kDa copper-binding protein in the brindled mouse model of Menkes disease

Little is known about copper metabolism at the cellular level. The brindled mouse is an animal model of Menkes disease which is an inborn error of copper metabolism. Control and brindled mice were used to identify copper-binding proteins with possible roles in normal copper metabolism that are affected by the defect in the brindled mice. When 64Cu-labeled hepatic or renal cytosols from control mice were applied to Mono Q or Superose columns, a approximately 48-kDa protein coeluted with the protein fractions which contained the radiolabeled copper. Large decreases in copper binding were detected in these fractions from the brindled mice. The same column fractions which showed decreased copper binding showed large decreases in the levels of the approximately 48-kDa protein. Decreased copper binding and approximately 48-kDa protein were not simply secondary to the abnormal hepatic and renal copper levels that are found in the brindled mice since although their liver copper levels are low, their kidney copper levels are high. Elevated levels of an approximately 80-kDa heat shock protein were also detected in the hepatic and renal cytosols from the brindled mice. Consistent with expression of the primary defect in both the liver and kidney, the levels of the approximately 48- and approximately 80-kDa proteins were affected similarly in both organs. Irrespective of how the low levels of the approximately 48-kDa protein may be related to the basic defect in the brindled mice, the data are consistent with an important role for the approximately 48-kDa protein in intracellular copper metabolism.     

Multiple forms of the glucocorticoid receptor steroid binding protein identified by affinity labeling and high-resolution two-dimensional electrophoresis

Potential charge heterogeneity within the glucocorticoid binding protein (GBP) of the glucocorticoid receptor was examined by a combination of affinity labeling, immunopurification, and high-resolution two-dimensional (2D) gel electrophoresis. One-dimensional sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) of [3H]dexamethasone 21-mesylate ([3H]DM) labeled cytosol identified a major, competable, component of Mr approximately equal to 92 000 (92K). This component was recognized by anti-human glucocorticoid receptor antibodies but not by nonimmune serum, indicating that the 92K component was the reduced denatured GBP. Examination of [3H]DM-labeled GBP by conventional 2D electrophoresis utilizing equilibrium isoelectric focusing in the first dimension failed to resolve the 92K GBP into discrete isoelectric components. This behavior was not representative of other, nonspecifically [3H]DM-labeled proteins or proteins in general. Nonequilibrium pH gradient electrophoresis (NEPHGE) was therefore employed to achieve separation in the first dimension. Immunopurified, [3H]DM-labeled GBP subjected to NEPHGE reached isoelectric equilibrium after 6 h of electrophoresis at 400 V. A single, broad peak of radioactivity was identified at pH approximately equal to 6.3. Second-dimension analysis of the NEPHGE-separated GBP by SDS-PAGE resolved this peak into two discrete, 92K, isoforms of apparent pI = 5.7 and 6.0-6.5. The GBP charge heterogeneity was confirmed by NEPHGE 2D analysis of [3H]DM-labeled GBP prepared directly from crude cytosol. Two isoforms indistinguishable from those observed in immunopurified samples were identified. An additional, more acidic, isoform (apparent pI approximately equal to 5.2) was also identified. Thus, there are at least two, and perhaps three, isoforms of the GBP. These data therefore suggest that there is significant charge heterogeneity in the GBP of the glucocorticoid receptor.     

An abundant 56 kD protein with low affinity androgen binding: another member of the steroid/thyroid receptor family?

We have found in genital skin fibroblasts an abundant 56 kD protein which appears related to the androgen receptor. The protein is detected in the soluble fraction by two-dimensional polyacrylamide gel electrophoresis as two spots with isoelectric points of 6.7 and 6.5 respectively. Photoaffinity labelling of GSF with 8 nM or 50 nM [3H]-methyltrienolone selectively labels the two protein spots but with an apparently lower affinity than is known for the androgen receptor. The two protein spots stem from one protein as judged from peptide patterns of partial proteolytic digests and the equal labelling of both spots with methyltrienolone. Cells from subjects with mutant androgen receptors generally lack the 56 kD protein and labelling with methyltrienolone fails, but the protein is not the androgen receptor itself. We propose the hypothesis that the 56 kD protein is synthesized from the same gene as the androgen receptor and that androgen may not be its natural ligand.     

A DNA-binding fraction of mouse kidney 3-ketosteroid reductase: Comparison with androgen receptors

Since approximately 1% of 3-ketosteroid reductase (which metabolizes dihydrotestosterone [17β-hydroxy-5α-androstan-3-one] to 5α-androstane-3α,17β-diol or 5α-androstane-3α,17β-diol) from mouse kidney cytosol adheres to DNA under conditions that allow virtually complete androgen receptor binding, these two DNA-binding activities were compared in cytosol extracts of mouse kidney and hypothalamus-preoptic area. This DNA-binding fraction of 3-ketosteroid reductase was distinguished from androgen receptor in several ways: (1) its pattern of elution from DNA-cellulose with steps of increasing NaC1 concentration differed from that for receptors from wild-type kidney; (2) it was influenced differently by the mutation Tfm, both in level and in DNA-cellulose elution pattern; (3) in mouse kidney cytosol it was relatively stable at moderate (25°C) temperatures which rapidly inactivated ligand-free androgen receptors in the same cytosols; (4) the DNA-binding was not proportional to androgen receptor levels between two wild-type tissues, the hypothalamus-preoptic area and kidney. By these criteria, a simple relationship of androgen receptors and a DNA-binding fraction of 3-ketosteroid reductase activity is unlikely.     

Cultured human skin fibroblasts: a model for the study of androgen action

Human skin may be considered as a target organ for androgens, as are male sex accessory organs, since all events involved in testosterone action have been observed in this tissue. As a corollary, the mechanism of androgen action can be studiedin vitro in cultured skin fibroblasts. The advantages of this system are that studies can be performed with intact human cells under carefully controlled conditions, differentiated genetic and biochemical characteristics of the cells are faithfully preserved and the biological material is renewable from a single biopsy specimen. The metabolism of androgens, in particular the 5α-reduction of testosterone to the active metabolite, dihydrotestosterone, the intracellular binding of androgen to its specific receptor protein and its subsequent translocation to the nucleus have been studied in skin fibroblasts. The intracellular androgen receptor content of genital skin fibroblasts is higher than that from nongenital skin sites. In addition, the androgen receptor has been characterized as a specific macromolecule with properties of high affinity and low capacity similar to that of other steroid hormone receptors. The pathophysiology of three genetic mutations which alter normal male sexual development and differentiation has been identified in the human skin fibroblast system. In 5α-reductase deficiency, an autosomal recessive disorder in which dihydrotestosterone formation is impaired, virilization of the Wolffian ducts is normal but the external genitalia and urogenital sinus derivatives are female in character. At least two types of X-linked disorders of the androgen receptor exist such that the actions of both testosterone and dihydrotestosterone are impaired and developmental abnormalities may involve both Wolffian derivatives and the external genitalia as well. These two forms of androgen insensitivity result from either the absence of androgen receptor binding activity (receptor(−)form) or apparently normal androgen receptor binding with absence of an appropriate biological response (receptor (+) form). In addition, studies with human skin fibroblasts may also be of value in defining the cellular mechanisms underlying the broad spectrum of partial defects in virilization. In summary, we have correlated our studies of the molecular mechanism of androgen action in human genital skin fibroblasts with those of other investigators as these studies contribute to our understanding of male sexual development and differentiation.     

Putative androgen receptors distinguished in wild-type and testicular-feminized (Tfm) mice.

The reduced level of putative androgen receptor in the mouse mutant, testicular feminization (Tfm), chromatographs on DNA-cellulose differently from the bulk of wild-type receptors. While the elution maximum for extracts of Tfm/Y kidney is in the 180–190 mM NaCl range, wild-type kidney extracts exhibit two maxima of elution at 140–150 mM NaCl and 180–190 mM NaCl, respectively. For hypothalamus-preoptic area, Tfm/Y has one elution maximum at approximately 180 mM NaCl, while the wild-type exhibits a major elution maximum at 140–150 mM NaCl, with a minor peak at approximately 180 mM NaCl. Mixing experiments between wild-type and Tfm/Y cytosols reveal that the different characteristic elution patterns are intrinsic to the binding complexes and are not conveyed simply by other soluble factors. The distinctive pattern for Tfm indicates that the mutation does not cause merely a reduced level of wild-type receptor. Rather the residual receptor of the mutant may be either an abnormal protein or a minor form of wild-type receptor, not readily seen in wild-type tissue due to the presence of more preponderant species. Differences in the elution profiles of androgen receptor species of wild-type kidney with the two bound androgens, testosterone and dihydrotestosterone, are also presented. A model of the androgen receptor system is proposed which includes several binding classes for androgen ligands and metabolites. In light of aromatization of androgens to estrogens and its probable role in some androgenic responses, we include the “estrogen receptor” in this mechanism.     

Syndromes d’insensibilite aux androgenes: Aspects moleculaires

The molecular analysis of clinical syndromes involving androgen insensitivity has been facilitated by the availability of increasingly powerful molecular biological techniques. Complementary DNA of the androgen receptor gene (wich was recently cloned and sequenced) has been used as a probe to investigate DNA restriction fragment length polymorphisms in patients with partial or complete androgen insensitivity. Such studies have demonstrated that deletions are rare. Using enzymatic amplification and sequencing of exons of the androgen receptor gene, several groups have described point mutations in patients with androgen insensitivity. PCR, accompanied by denaturing gradient gel electrophoresis or single strand conformation polymorphism assays, has revealed further mutations. These powerful tools, together with studies of mRNA from expression of the mutant gene, have also illustrated structure-function associations of the androgen receptor gene in some patients with androgen insensitivity.     

Androgen receptors and partial androgen insensitivity in male pseudohermaphroditism

Male pseudohermaphroditism can occur either as a result of inadequate fetal androgen secretion or ineffective androgen action. Partial androgen insensitivity syndrome is related to a decrease in the binding capacity of target cells androgen receptors. However, the biochemical expression of these syndromes is variable: although a low concentration of androgen receptors is observed in 75% of the cases, a normal binding capacity is sometimes reported. If the presence of androgen receptors is necessary for hormonal action, it is not sufficient since the receptor has to conserve its functional integrity. Several qualitative abnormalities of the androgen receptor have been described and could thus explain the disturbance in the androgen mechanism of action. The study of androgen-dependent proteins and other androgen-receptor "markers" should offer the clinician a better management of partial androgen insensitivity and allow, in amniotic or trophoblastic cells, a prenatal diagnosis of androgen resistance in high-risk families.