Resolution of non-activated and activated androgen receptors based on differences in their hydrodynamic properties

This study shows that cytosolic androgen receptor of rat ventral prostate sediments at 10-11 S on conventional low salt sucrose density gradients (SDG), and at 4.6 S on high salt SDG, whether it is activated or not; inclusion of 10 mM Na2MoO4 in all buffers does not alter these sedimentation coefficients. In the presence of 50 mM Na2MoO4 non-activated and activated androgen receptors sediment in high salt SDG at 7-8 S and 4.6 S, respectively. Thus the presence of high concentrations of molybdate during centrifugation inhibits the KCl induced disaggregation of receptor into subunits. Similar effects are observed on Sephacryl-S200 gel filtration; in 50 mM MoO2-4 and 0.4 M KCl non-activated receptor has an estimated Stokes radius of 67 A; this value decreases to 52 A upon activation in the presence of proteolysis inhibitors; omission of molybdate during chromatography yielded 52 A and 27 A entities. Estimated mol. wts are 198,000 Daltons for the non-activated 67 A form and 98,000 Daltons for the activated 52 A receptor. Sodium molybdate (50 mM) prevents temperature (18 degrees C) and high ionic strength (0.4 M KCl) induced receptor activation. This inhibition was overcome by removing molybdate by centrifugal gel filtration, or by increasing the KCl concentration to 0.8 M. The inhibitory effects of molybdate on salt induced receptor disaggregation into activated subunits are no longer observed at pH greater than 7.4 or after chemical modification of sulfhydryl groups. Once androgen receptor has been disaggregated into its activated subunits the activated state is maintained even upon reassociation to 10-11 S aggregates in low salt. The relative concentrations of KCl and molybdate are critical; thus, 10 mM Na2MoO4/0.4 M KCl and 50 mM Na2MoO4/0.8-1.2 M KCl did not differentiate activated from non-activated androgen receptor based on their hydrodynamic properties. In the presence of 0.4 M KCl and 50 mM molybdate, however, the hydrodynamic properties of androgen receptor can be correlated with receptor activation.     

Studies on the non-activated and activated forms of the estrogen receptor.

The activation of the steroid receptor is a necessary process for the biological role of the receptor. Many factors are involved in this mechanism; in addition to time, temperature, and salt concentration, RNA and RNAase can also affect the transformation of the non-activated to the activated form of the receptor. Using as a model the estrogen receptor of fetal uterus of guinea-pig, the studies of the interaction with three different monoclonal antibodies (D547, H222 and H226) reveal structural transformation during the process of the receptor activation. These conformational transformations suggest that a change in the exposure of the functional domains of the estrogen receptor occurs during activation.     

Inactivation of rat liver glucocorticoid receptor by molybdate. Effects on both non-activated and activated receptor complexes.

When freshly prepared glucocorticoid-receptor complex from rat liver cytosol was incubated at 23 degrees C in the presence of sodium molybdate, its subsequent binding to isolated nuclei, DNA-cellulose and ATP-Sepharose was blocked. In addition, binding to these acceptors by cytosol receptor complex fractionated with (NH4)2SO4 was also blocked by incubation of the complexes with 50 mM-sodium molybdate. However, molybdate had no effect on the binding of activated receptor complexes to ATP-Sepharose. Molybdate was also effective in extracting the nuclear- and DNA-cellulose-bound glucocorticoid-receptor complexes in a dose-dependent manner. Molybdate appears to exert its effects directly on the receptor by interacting with both non-activated and activated receptor forms.     

Analysis of steroid- and DNA-binding domains of the calf uterine androgen receptor by limited proteolysis

The DNA-binding form of the calf uterine androgen receptor (AR) was subjected to limited protease digestion using chymotrypsin, trypsin and a rat prostate cytosol protease. The properties of the generated polypeptide fragments were identified and compared with those of the intact AR. Physicochemical characterization was achieved through sedimentation analysis, gel filtration chromatography and DEAE anion exchange chromatography. Intactness of functional binding domains was evaluated by measuring the retention of steroid- and DNA-binding capacity. Under non-denaturing conditions the intact AR is a highly asymmetrical molecule with a Stokes radius (RS) of 45A, a sedimentation coefficient of 4.3S and a relative molecular mass of 80,000 daltons. This form of AR has an intrinsic binding affinity for DNA and was eluted from DNA-cellulose with 9 mM MgCl2. Chymotrypsin produced a more globular polypeptide (RS: 31A; 3.1S; 41,000 daltons) with a decreased net negative charge. This fragment also displayed DNA-binding affinity but required a higher concentration of MgCl2 (14 mM) for DNA-cellulose elution, indicating an increased affinity for DNA. The observed reduction in molecular size upon chymotrypsin treatment was confirmed when analysed by SDS-polyacrylamide gel electrophoresis after covalently labelling of the AR with [3H]R1881. Rat prostate cytosol contains a protease which is very active in generating an AR polypeptide with an increased affinity for DNA, without changing the AR net negative charge (RS: 33A; 3.7S; 51,000 daltons). The specificity of this protease remained unknown since none of a large number of inhibitors was able to inactivate this enzyme. The fragment generated is different from that obtained with chymotrypsin since significant differences in size as well as in charge were measured. Trypsin treatment generated a much smaller polypeptide (RS: 25A; 2.9S; 30,000 daltons) which had lost its DNA-binding capacity, but not its steroid binding site. This form probably represents the so-called meroreceptor. When intact AR was treated sequentially with prostate cytosol and trypsin, a polypeptide fragment with identical properties was obtained, indicating the spatial separation of two of the proteolytic cleavage sites. These studies provide evidence for the distinct nature of the molecular domains for androgen and DNA interaction on the calf uterine AR.     

Characteristics of the calf uterine androgen receptor

The highest molecular weight form of the calf uterine androgen receptor separates as an 11S form in glycerol gradients. This "cytosolic" receptor, prepared in the presence of molybdate, polyethyleneimide and low ionic strength, dissociates into 9S and 7.2S forms with increasing KCl concentration. A 4.5S androgen binding component appears as the predominant form of the receptor in the absence of polyethyleneimide and this unit quantitatively converts to a stable 3.5S form in the absence of molybdate. Renaturation of partially purified protein, separated by SDS-PAGE electrophoresis, demonstrates the presence of an androgen binding component in the 110 kDa region of the gel. This renatured protein separates as a 4.5S component in glycerol gradients and has a Stokes radius of 6 nm. Photoaffinity labelling of partially purified receptor preparations, followed by SDS-PAGE electrophoresis, reveals the presence of an androgen binding component having a molecular weight of 115 kDa. The binding characteristics and specificity of the receptor binding to R1881 have been studied and a DHT-affinity chromatography resin used to purify the receptor.     

Physicochemical analysis of reversible molybdate effects on different molecular forms of glucocorticoid receptor

Several distinct molecular forms of glucocorticoid receptor have been identified in a melanoma model system. We have used velocity sedimentation to monitor molybdate dependent alterations in receptor size and heterogeneity. In the absence of molybdate, native glucocorticoid receptor from dexamethasone-sensitive tumors sediments at 7–8 S and 12–13 S. Under identical conditions, receptor isolated from dexamethasone-resistant tumors sediments at 7–8 S only. However, when molybdate is introduced, either during homogenization or immediately prior to centrifugation, glucocorticoid receptors from both dexamethasone-sensitive and -resistant tumors sediment sharply at 9–10 S. These molybdate induced phenomena are reversible. The activated forms of glucocorticoid receptor isolated from both dexamethasone-sensitive and -resistant tumors by DEAE-cellulose chromatography have similar sedimentation coefficients (4–5 S) which are unaffected by molybdate.     

Factors modifying equilibrium between activated and non-activated forms of steroid-receptor complexes.

Steroid-receptor complexes formed in concentrated cytosol at low temperature, low ionic strength and neutral pH are unable to bind to nuclei. Various procedures are known to promote their 'activation'. In the present work it is shown that an increase in temperature only enhances the rate of the reaction whereas no change in the equilibrium between activated and non-activated complexes is observed. On the contrary an increase in ionic strength or pH, as well as a removal of a low-molecular-weight inhibitor, not only accelerate the reaction but also increase the concentration of activated complexes at equilibrium. Using two steroids differing 3-fold in their affinity for the receptor, no difference was seen in the effect of the bound steroid on receptor activation. When combining various activation procedures it was observed that they acted independently of each other and additively. In all cases they retained their property of either modifying only the rate of the reaction or both its rate and equilibrium. Using changes in pH, it was also possible to induce shifts in the equilibrium between activated and non-activated complexes. After activation at pH 6.5, a first equilibrium was attained. When the pH was increased to 8 the equilibrium was displaced towards higher concentrations of activated complexes. A lowering of the pH resulted in a reversal of steroid-receptor complexes from the activated to the non-activated state. To clearly establish that this was not due to irreversible damage of the receptor, which would render it unable to bind to nuclei, it was shown that the complexes which had reverted to the non-activated state were still susceptible to activation. Regulatory events may thus exist which, for a given level of hormone and receptor, modulate the concentration of activated steroid-receptor complexes.     

DNA-binding of androgen receptor overexpressed in mammalian cells

In order to investigate the DNA-binding properties of the rat androgen receptor (rAR) in mammalian cells after addition of androgens and antiandrogens, we established a gel-shift assay with extract from COS-1 cells (CV-1 cells transformed with the DNA-tumour virus SV40) overexpressing the rAR. First, the rAR was overexpressed in COS-1 cells. Therefore the full-length AR cDNA was inserted immediately downstream from the SV40 early promoter of pECE to generate pECE-AR. Expression of the rAR driven by the SV40 early promoter yields constant and high levels of rAR protein. In addition, the vector contains the SV40 origin of replication for obtaining high copy vector numbers in COS-1 cells. The rAR-containing expression vector was transiently transfected into COS-1 cells using Transfectam Reagent, in order to achieve high transfection efficiency. Expression of biologically active receptor was tested by analyzing the effect of the synthetic androgen R1881 on induction of transiently transfected pMMTV-CAT. Steroid binding assays were carried out to confirm overexpression of biologically active AR and to determine the binding of different hormones and antihormones to AR in COS-1 cells transiently transfected with pECE-AR. Gel-shift experiments performed with whole cell extract of those cells, containing 700 fmol AR/mg protein, and labeled AR-binding GRE (glucocorticoid responsive element) showed that R1881 induced the formation of a protein-GRE complex. Furthermore, the R1881-induced formation of the protein-GRE complex could be competed by addition of unlabeled excess of GRE but not of unspecific oligonucleotides, confirming sequence-specific binding of the R1881-induced protein-GRE complex.     

Purification of the intact monomeric 110 kDa form of the androgen receptor from calf uterus to near homogeneity

In the present study, culf uterine tissue has been used for isolation of androgen receptors. This tissue appeared to be a favourable source for large-scale purification of androgen receptors, because of the relatively high level of androgen receptors and the low concentration of proteolytic enzymes. The purification involved differential phosphocellulose and DNA affinity chromatography as first steps. The non-transformed receptor was passed through these matrices in order to remove contaminating DNA-binding proteins. After a transformation step to the DNA-binding state, the receptor was bound to DNA cellulose and subsequently eluted with MgCl₂. A 0.5% pure androgen receptor preparation was obtained. Photoaffinity labelling with [³H]R1881 (methyltrienolone) was used to determine the size of the receptor at this stage of purification and during the following steps. Subsequently, isoelectric focussing of the partially purified androgen receptor preparation in an aqueous glycerol gradient was performed. In this step, the progesterone receptor, which is copurified with the androgen receptor protein during the first part of the purification procedure, focussed at pH 5.5, while the androgen receptor could be isolated at pH 5.8. The isoelectric focussing procedure could be applied in a preparative way for further purification of androgen receptors. After this step an approx. 8% pure preparation was obtained. Polyacrylamide gel electrophoresis of S-carboxymethylated androgen receptor was used as the final purification step. The [³H]methyltrienolone labelled androgen receptor from calf uterus was purified to homogeneity and consisted of one polypeptide with a molecular mass of 110 kDa.     

Purification of the intact monomeric 110 kDa form of the androgen receptor from calf uterus to near homogeneity

In the present study, calf uterine tissue has been used for isolation of androgen receptors. This tissue appeared to be a favourable source for large-scale purification of androgen receptors, because of the relatively high level of androgen receptors and the low concentration of proteolytic enzymes. The purification involved differential phosphocellulose and DNA affinity chromatography as first steps. The non-transformed receptor was passed through these matrices in order to remove contaminating DNA-binding proteins. After a transformation step to the DNA-binding state, the receptor was bound to DNA cellulose and subsequently eluted with MgCl2. A 0.5% pure androgen receptor preparation was obtained. Photoaffinity labelling with [3H]R1881 (methyltrienolone) was used to determine the size of the receptor at this stage of purification and during the following steps. Subsequently, isoelectric focussing of the partially purified androgen receptor preparation in an aqueous glycerol gradient was performed. In this step, the progesterone receptor, which is copurified with the androgen receptor protein during the first part of the purification procedure, focussed at pH 5.5 while the androgen receptor could be isolated at pH 5.8. The isoelectric focussing procedure could be applied in a preparative way for further purification of androgen receptors. After this step an approx. 8% pure preparation was obtained. Polyacrylamide gel electrophoresis of S-carboxymethylated androgen receptor was used as the final purification step. The [3H]methyltrienolone labelled androgen receptor from calf uterus was purified to homogeneity and consisted of one polypeptide with a molecular mass of 110 kDa.     

DNA-binding and non-DNA-binding forms of the transformed glucocorticoid receptor

In this work we have probed the mechanism responsible for two non-DNA-binding states of the mouse glucocorticoid receptor. In the first case, transformed receptors were treated with hydrogen peroxide. It is known that oxidizing agents promote the formation of disulfide bonds in the glucocorticoid receptor, but it has not been determined what domains are involved in any disulfide bond formation that leads to inactivation of DNA-binding activity. We show here that hydrogen peroxide inhibits DNA-binding by the 15-kDa tryptic fragment containing the DNA-binding fingers with the same concentration dependency as it inhibits DNA-binding by the uncleaved receptor. This suggests that all of the effect of peroxide is on sulfhydryl groups within the zinc fingers. After dissociation (transformation) of cytosolic heteromeric glucocorticoid receptor complexes, only a portion (40–60%) of the dissociated receptors can bind to DNA-cellulose. We show that the 15-kDA tryptic fragment derived from the portion of transformed receptors that do not bind to DNA is itself competent at DNA-binding.     

Evidence for DNA-binding domain--ligand-binding domain communications in the androgen receptor

DNA binding as well as ligand binding by nuclear receptors has been studied extensively. Both binding functions are attributed to isolated domains of which the structure is known. The crystal structure of a complete receptor in complex with its ligand and DNA-response element, however, has been solved only for the peroxisome proliferator-activated receptor γ (PPARγ)-retinoid X receptor α (RXRα) heterodimer. This structure provided the first indication of direct interactions between the DNA-binding domain (DBD) and ligand-binding domain (LBD). In this study, we investigated whether there is a similar interface between the DNA- and ligand-binding domains for the androgen receptor (AR). Despite the structural differences between the AR- and PPARγ-LBD, a combination of in silico modeling and docking pointed out a putative interface between AR-DBD and AR-LBD. The surfaces were subjected to a point mutation analysis, which was inspired by known AR mutations described in androgen insensitivity syndromes and prostate cancer. Surprisingly, AR-LBD mutations D695N, R710A, F754S, and P766A induced a decrease in DNA binding but left ligand binding unaffected, while the DBD-residing mutations K590A, K592A, and E621A lowered the ligand-binding but not the DNA-binding affinity. We therefore propose that these residues are involved in allosteric communications between the AR-DBD and AR-LBD.     

Interaction of estrogen receptor of calf uterus with a monoclonal antibody: probing of various molecular forms

A monoclonal antibody to estrogen receptor (JS34/32) is able to recognize, in the calf uterine cytosol, a protein (approximately 65 000 daltons) giving a single band on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Two molecules of this antibody are able to simultaneously interact with the native 8S form of the receptor present in the calf uterine cytosol ("twin antibody" assay). This indicates the presence of two antigenic determinants on the "low-salt" 8S form of the receptor. This form of the receptor shows an increase in Mr from 345 000 to 665 000 after interaction with the soluble antibody. Dissociating agents that induce the dissociation of the 8S form to smaller forms also induce the dissociation of the two antigenic determinants. The 4S "high-salt" form of the estrogen receptor has one determinant per molecule, appearing to be the smallest form of the receptor not containing repetitive structures associated with the steroid binding site. The nuclear receptor also shows the presence of more than one antigenic determinant on its molecule.     

Physicochemical properties of non-activated and activated renin from human amniotic fluid.

The main physicochemical and enzymic properties of non-activated and activated human amniotic renin (EC 3.4.99.19) were studied in order to clarify the relationships between the two enzymes. Human amniotic renin was activated by dialysis against acidic buffer (pH 3.3), direct acidification or trypsin treatment. All procedures produced similar activation. The physicochemical characteristics of non-activated and activated renin were compared to those of human renal renin. Non-activated renin had a molecular weight of 45,500. A similar molecular weight was obtained by gel eluate activation and by acid treatment of renin prior to gel filtration. Similar isoelectric points were also found for non-activated and activated renin. One major renin peak focused at pH 6.6, whereas no similar renin peak was detected in extracts from normal human kidney. In addition, non-activated and activated renin forms were found to have the same optimal pH, the same Km and the same inhibiting pepstatin concentrations.     

Effect of molybdate on the molecular organization of the estrogen receptor system of porcine uterus

Molybdate was shown to have complex effects in modulating the molecular organization of the constituents of the estrogen receptor (ER) system of porcine uterus. We showed previously the presence of one basic ER molecule (vero-ER) (sedimentation coefficient, 4.5S; Stokes radius, 44 A) and ER-binding factors (ERBFs) ["8S" ER-forming factor ("8S" ER-FF), (component A) X (component B)6; "6S" ER-FF, (component B)6; "5S" ER-FF, component A] in the porcine uterus [Fukai, F. & Murayama, A. (1981) J. Biochem. 95, 1697-1704]. Molybdate regulates the specific interaction of vero-ER with ERBFs in a complex way. The apparent Kd value (6.7 X 10(-10) M) of vero-ER with "8S" ER-FF in the presence of molybdate (30 mM) was decreased remarkably as compared with that (2.7 X 10(-9) M) in the absence of molybdate. In contrast, the apparent Kd value (3.7 X 10(-9) M) of vero-ER with "5S" ER-FF observed in the presence of molybdate (30 mM) was increased over ten-fold as compared with that in the absence of molybdate. Meanwhile, the affinity (Kd, 5 X 10(-9) M) of vero-ER for "6S" ER-FF was scarcely influenced by molybdate. These results reveal the mechanism by which molybdate selectively stabilizes "8S" ER. Molybdate further affected the molecular constitution of ERBFs. The dissociation of "8S" ER-FF into component A and component B, which takes place under hypertonic (0.4 M KCl) conditions at higher temperature (25 degrees C), was suppressed almost completely by molybdate (30 mM).(ABSTRACT TRUNCATED AT 250 WORDS)     

Differential effects of androgens and glucocorticoids on regulation of androgen receptor concentrations and cell growth

This report describes androgen induced up-regulation of intracellular androgen receptor concentrations in the DDT₁MF-2 cell line derived from the hamster ductus deferens and the R3327H-G8-A1 line derived from the Dunning prostate adenocarcinoma. Within 6 h of exposure to androgens the receptor concentration is increased approximately 2-fold. Incorporation of dense aminoacids and subsequent analysis by sucrose density gradient centrifugation indicates that the observed increase in receptors is possibly due to de novo androgen receptor synthesis. In both cell lines this increase is inhibited by 30–50% within 6 h and completely during a subsequent 18 h period by the potent glucocorticoid triamcinolone acetonide (TA). TA also inhibits the growth of both cell lines and antagonizes the stimulatory effect of androgens. Flow cell cytometry studies indicate that TA blocks the cells in the G1 phase of the cell cycle. This event may be associated with regulation of androgen receptor concentrations.     

Kinetic and thermodynamic evidence of a molybdate interaction with glucocorticoid receptor in calf thymus

The effect of molybdate on the kinetic and thermodynamic properties of the dexamethasone-receptor interaction was studied in calf thymus cytosol. In the presence of molybdate both the equilibrium binding studies and the association and dissociation experiments reveal a significantly lower affinity of the receptor for [3]dexamethasone. At 0 degrees C the equilibrium dissociation constant increases from 0.8 nM to 1.8 nM, the association rate constant shifts from 1.5 X 10(8) M-1 h-1 to 0.2 X 10(8) M-1 h-1, whereas the rate of dissociation of the untransformed receptor increases from 0.04 h-1 to 1.1 h-1 in the molybdate-containing buffer. All these effects appear dependent on the concentration of molybdate but the dissociation of the transformed receptor (0.01 h-1) is unaffected. The enthalpy for the association, delta H not equal to, increases at least twofold whereas the entropy, both for the association (delta S not equal to = -25 to +104 J K-1 mol-1) and for the equilibrium (delta S degrees = -100 to +38 J K-1 mol-1), is markedly influenced by the presence of molybdate. Taken all together these data suggest that molybdate interacts with the receptor molecule turning it into a form that displays low affinity for steroid, in addition to the well-documented incapacity to transform itself. This fact leads us to think that both the binding and the transformation are the expression of conformational modifications involving molybdate-sensitive groups.     

Thermal inactivation and molecular forms of the estrogen receptor: Effects of molybdate

Exposure at 40°C of low salt calf uterine cytosol leads to the “transformation” of the estradiol-receptor complexes in 5 min, immediately followed by the formation of thermostable > 12 S “aggregated” receptor forms. Molybdate prevents this phenomenon but does not reverse it. Molybdate has a protective effect against thermal inactivation of the 9 S non-aggregated form of the receptor in the absence of hormone. In the presence of molybdate, the inactivation rate of this 9 S receptor is the same with and without hormone, and follows a first order reaction ($\text{t}{}_{\mathrm{<mtext>1</mtext><mtext>2</mtext>}}\text{= 7–8 min}$). The biphasic kinetics of thermo-inactivation of estradiol-receptor complexes is ascribable to the relative amounts of non-aggregated and aggregated forms.     

Interaction of androgen response elements with the DNA-binding domain of the rat androgen receptor expressed in Escherichia coli

A fragment of the rat androgen receptor (amino acids 533-637) containing the DNA-binding domain was produced in Escherichia coli as a fusion product with protein A of Staphylococcus aureus. The fusion protein was purified on IgG-Sepharose, a method that does not involve the use of denaturing agents. Approximately 4 mg of fusion protein was obtained from 500 ml of bacterial culture. In gel shift assays, the recombinant DNA-binding domain displays an affinity for a fragment of the long terminal repeat of mouse mammary tumor virus and for an intronic fragment of the gene coding for the C3 component of the androgen-regulated rat prostatic binding protein. In a DNase I footprinting assay, the fusion protein protects a sequence in the C3 fragment that has previously been shown to act as a functional androgen response element. Interestingly, a single base pair mutation in the response element, which abolishes androgen inducibility, also destroys the ability to interact with the recombinant androgen receptor DNA-binding domain.