A convenient synthesis of the side chain of loteprednol etabonate--an ocular soft corticosteroid from 20-oxopregnanes using metal-mediated halogenation as a key reaction

A facile synthesis of the side chain of loteprednol etabonate, namely, chloromethyl-17α-[(ethoxycarbonyl))oxy]-11β-hydro of loteprednol etabonate, viz., chloromethyl-17α-[(ethoxycarbonyl))oxy]-11xy-3-oxoandrosta-1,4-diene-17β-carboxylate - an ocular soft corticosteroid, has been described starting from a 20-oxopregnane, namely, 3β-acetoxy-pregn-5(6),16(17)-diene-20-one (16-dehydropregnenolone acetate, i.e., 16-DPA) using our recently developed metal-mediated halogenation as a key reaction.     

Isolation and structure elucidation of the major photodegradation products of loteprednol etabonate

Photodegradation of loteprednol etabonate (5), a steroid anti-inflammatory drug, in the solid state, in aqueous suspension, and in aqueous acetonitrile solution has been investigated. Analysis by HPLC showed that the profile of photodegradation products in the solid state was qualitatively similar to that in the aqueous suspension, although the profile in the aqueous acetonitrile solution was considerably different. The major photodegradation products were isolated from the aqueous suspension and the aqueous acetonitrile solution by using preparative reversed-phase HPLC and their structures were elucidated on the basis of spectroscopic data. Photolysis in the solid state and in aqueous suspension yielded three rearrangement products, chloromethyl 17alpha-ethoxycarbonyloxy-11beta-hydroxy-5alpha-methyl-2-oxo-19-norandrosta-1(10),3-diene-17beta-carboxylate (8), chloromethyl 17alpha-ethoxycarbonyloxy-11beta-hydroxy-1-methyl-3-oxo-6(5-->10alpha)-abeo-19-norandrosta-1,4-diene-17beta-carboxylate (9), and chloromethyl 1beta,11beta-epoxy-17alpha-ethoxycarbonyloxy-2-oxo-10alpha-androsta-4-ene-17beta-carboxylate (10). In aqueous acetonitrile solution, 10 was the major product, however, 8 and 9 were not obtained. Pathways for the formation of these compounds from loteprednol etabonate (5) are proposed.     

Effect of the 90 kDa heat shock protein, HSP90, on glucocorticoid receptor binding to DNA-cellulose

Glucocorticoid receptors in the IM-9 human lymphoblastoid cell line were affinity labeled with [3H]dexamethasone 21-mesylate and activated to a DNA-binding form by filtration through a Bio-Gel A-1.5m column. The 90 kDa heat shock protein, HSP90, was identified by labeling IM-9 cells with 35S-methionine at both 37 degrees C and 42 degrees C and purified to near homogeneity by sequential chromatography through DE52 and hydroxyapatite. Addition of purified HSP90 to activated, affinity labeled glucocorticoid receptors in a molecular ratio of 16 to 1 inhibited the binding of the receptors to DNA-cellulose. HSP90 did not affect the binding of other proteins to DNA-cellulose, indicating that the inhibitory effect of HSP90 was specific for the glucocorticoid receptor. These results suggest that HSP90 may associate with the glucocorticoid receptor, masking its DNA-binding site and thereby inhibiting receptor interaction with DNA.     

Interaction of rat liver glucocorticoid receptor with a newly synthesized antisteroid ZK98299

Steroid receptor antagonists are important biochemical probes for understanding the mode of steroid hormone action. We have studied the interaction between rat liver glucocorticoid receptor and a newly synthesized antisteroid ZK98299 (13-antigestagen; [11-beta-(4-dimethylaminophenyl)-17a-hydroxy-17 beta-(3- hydroxypropyl)-13 alpha-methyl-4,9-gonadien-3-one]). Glucocorticoid receptor from freshly prepared hepatic cytosol bound [3H]ZK98299 with affinity approximately equal to that of [3H]triamcinolone acetonide. The binding of both steroids reached a maximum at 4 h at 0 degrees C. Both ligands were able to compete for the steroid binding site but progesterone, estradiol and dihydrotestosterone (DHT) failed to compete for the [3H]ZK98299 and [3H]triamcinolone acetonide binding. While [3H]ZK98299 binding to glucocorticoid receptor could occur in the presence of iodoacetamide and N-ethylmaleimide (NEM), [3H]triamcinolone acetonide binding capacity was completely abolished following such treatments. The [3H]ZK98299-receptor complexes sedimented as 9 S and 4 S molecules under control (4 degrees C) and receptor transforming (23 degrees C) conditions, and exhibited a faster rate of dissociation at 23 degrees C when compared with [3H]triamcinolone acetonide-receptor complexes. These results indicate that ZK98299 interacts with hepatic glucocorticoid receptor. The differential effects of iodoacetamide and NEM on the interaction of glucocorticoid receptor with ZK98299 and triamcinolone acetonide, and the faster rate of dissociation of [3H]ZK98299-receptor complexes suggest that treatment with these agents (NEM and iodoacetamide) results in distinct conformational changes in glucocorticoid receptor structure with respect to triamcinolone acetonide and ZK98299 binding. Alternatively, ZK98299 may be interacting with a site which is distinct from one which accepts triamcinolone acetonide.     

The Mr 90,000 heat shock protein-free androgen receptor has a high affinity for steroid, in contrast to the glucocorticoid receptor

Transformed and bacterially expressed glucocorticoid receptors free from Mr 90,000 heat shock protein (hsp90) have a 100-fold lower steroid-binding affinity than the hsp90-bound nontransformed receptor, suggesting that hsp90 is needed for high-affinity steroid binding [Nemoto, T., Ohara-Nemoto, Y., Denis, M., & Gustafsson, J.-A. (1990) Biochemistry 29, 1880-1886]. To investigate whether or not this phenomenon is common to all steroid receptors, we investigated the steroid-binding affinities of bacterially expressed and transformed androgen receptors. The C-terminal portion of the rat androgen receptor containing the putative steroid-binding domain was expressed as a fusion protein of protein A in Escherichia coli. The recombinant protein bound a synthetic androgen, [3H]R1881, with high affinity (Kd = 0.8 +/- 0.3 nM). Glycerol gradient analysis revealed that the recombinant protein sedimented at around the 3S region irrespective of the presence of molybdate, indicating that the receptor is present in monomeric form. The steroid-free transformed androgen receptor was obtained by exposure of rat submandibular gland cytosol to 0.4 M NaCl in the absence of steroid. High-performance ion-exchange liquid chromatography analysis showed that the transformed androgen receptor bound to [3H]R1881 with high affinity. Thus these observations indicate that, in contrast to the glucocorticoid receptor, hsp90 is not required for the high-affinity steroid binding of the androgen receptor. In addition, the hsp90-free androgen receptor prebound with radioinert R1881 was efficiently relabeled with [3H]R1881, while the triamcinolone acetonide-bound, transformed glucocorticoid receptor failed in ligand exchange. The inability to achieve ligand exchange probably reflects the low steroid-binding affinity of this entity.     

Depletion of [3H]methyltrienolone cytosol binding in glucocorticoid-induced muscle atrophy

The present study was undertaken to determine cytosol binding properties of [3H]methyltrienolone, a synthetic androgen, in comparison with [3H]dexamethasone, a synthetic glucocorticoid, under conditions of glucocorticoid excess in skeletal muscle. Male hypophysectomized rats received either seven daily subcutaneous injections of cortisone acetate (CA) (100 mg X kg-1 body wt) or the vehicle, 1% carboxymethyl cellulose. Following treatment, both [3H]dexamethasone and [3H]methyltrienolone-receptor concentrations were decreased from those in vehicle-treated rats by more than 90 and 80%, respectively, in CA-treated animals. Scatchard analysis of [3H]methyltrienolone binding in muscles of vehicle-treated animals became nonlinear at high concentrations of labeled ligand and were reanalyzed by a two-component binding model. The lower affinity, higher capacity component, which was attributed to binding of methyltrienolone to a "dexamethasone" component, disappeared in muscles of CA-treated rats and Scatchard plots were linear. Receptor concentrations of the higher affinity lower capacity "methyltrienolone" component were similar in muscles of vehicle-treated and CA-treated groups. From competition studies, the high relative specificities of glucocorticoids for [3H]methyltrienolone binding in muscles of vehicle-treated animals were markedly reduced by CA treatment. In addition, the binding specificity data also showed strong competition by progesterone and methyltrienolone for [3H]dexamethasone binding and estradiol-17 beta for [3H]methyltrienolone binding. These results demonstrate that most of the [3H]methyltrienolone binding is eliminated under in vivo conditions of glucocorticoid excess. Furthermore, the competitiveness of various steroids for receptor binding suggests that rat muscle may not contain classic (ligand-specific) glucocorticoid and androgen receptors.     

Assay of primate seminiferous tubule androgen receptors using [3H]mibolerone

The synthetic radiolabelled androgen mibolerone (7 alpha, 17 alpha-dimethyl-19-nortestosterone) was used to characterize androgen receptor binding in the seminiferous tubules from Cynomolgus monkey testis. Mibolerone binding was of high affinity (Kd = 0.6-5.4 nM) and limited capacity (37-50 fmol/mg protein), and was androgen specific. Sucrose density gradient centrifugation using a vertical tube rotor permitted the identification of a 9S molybdate-stabilized receptor under low salt conditions. The receptor bound to DEAE-cellulose. Methyltrienolone, but not mibolerone, also bound to a low affinity high capacity binding site in tubule cytosol, which probably represents glucocorticoid receptor binding, since it could be displaced by excess dexamethasone. However, occupancy of this low-affinity binding site by dexamethasone in an androgen receptor assay with [3H]methyltrienolone lead to a 33% underestimation of receptor binding, which appeared to relate to radioactive decomposition. Mibolerone, as well as methyltrienolone, bound to a progestin-binding protein in seminiferous tubule cytosol. These studies provide methods for the study of seminiferous tubule androgen receptors in subhuman primates and indicate that, due to its greater stability and lack of binding to glucocorticoid receptor, mibolerone is a useful new ligand in the study of androgen receptors in testis and its constituent cells.     

Direct evidence that the glucocorticoid receptor binds to hsp90 at or near the termination of receptor translation in vitro

We have translated the rat glucocorticoid receptor in both reticulocyte lysate and in wheat germ extract. Receptor synthesized in the reticulocyte lysate is immunoadsorbed by the 8D3 monoclonal antibody directed against the 90-kDa heat shock protein (hsp90) and it has a normal ability to bind glucocorticoid in a high affinity manner. Although the wheat germ extract synthesizes the full length receptor, the receptor is not immunoadsorbed by 8D3 and we cannot demonstrate high affinity steroid binding. Receptor synthesized by the reticulocyte lysate can be immunoadsorbed by antibody directed against hsp90 as soon as the translation product is full length, suggesting that the receptor becomes associated with hsp90 late during translation or immediately at the termination of translation. When newly synthesized receptor is bound with steroid and incubated at 25 degrees C, it is converted to a form that binds to DNA. This study provides direct evidence that association of hsp90 with the glucocorticoid receptor is a very early event and that the newly formed heteromeric receptor-hsp90 complex is fully competent to undergo transformation.     

Steroid derivatives for electrophilic affinity labelling of glucocorticoid binding sites: interaction with the glucocorticoid receptor and biological activity

To investigate the possible use of electrophilic affinity labelling for the characterization of glucocorticoid receptors, different chemically reactive derivatives of deoxycorticosterone (deoxycorticosterone 21-mesylate and deoxycorticosterone 21-(1-imidazole) carboxylate), dexamethasone (dexamethasone 21-mesylate, dexamethasone 21-iodoacetate and dexamethasone 21-bromoacetate) and progesterone (21-chloro progesterone) were tested for their ability to bind irreversibly to the glucocorticoid receptor from goat lactating mammary gland. Using partially purified receptor, only one of the steroids tested, dexamethasone 21-mesylate (DXM-M) was found more effective than dexamethasone (DXM) in preventing exchange of radioactive dexamethasone in the receptor binding site. The affinity of DXM-M for the glucocorticoid receptor, measured by competitive binding assay, was 1/15 that of DXM. Polyacrylamide gel electrophoresis in sodium dodecyl sulphate of the [3H]-DXM-M labeled glucocorticoid receptor revealed a specific covalently radiolabeled fraction corresponding to an apparent molecular weight of 75,000 to 80,000. The biological activity of DXM-M was studied in RPMI 3460-clone 6 Syrian hamster melanoma cells, a cell line which is sensitive to growth inhibition by glucocorticoids. Like DXM, DXM-M inhibits the growth of RPMI 3460-clone 6 cells and it acts as a slowly reversible glucocorticoid agonist at concentrations which correlate with the affinity of DXM-M for the glucocorticoid receptor in vitro.     

Characterization of wild type and mutant glucocorticoid receptors from rat hepatoma and mouse lymphoma cells

Using a combination of immunological blotting techniques and hormone affinity labeling, we have characterized the glucocorticoid receptors present in wild type and mutant rat hepatoma (HTC) and mouse thymoma (S49 and WEHI7) cells. Mutant HTC and WEHI7 cells of the receptorless phenotype, which contain greatly reduced amounts of glucocorticoid hormone binding activity, show parallel decreases in immunoreactive material using a monoclonal antibody raised against the rat liver glucocorticoid receptor. This indicates that these receptorless mutant cells harbor defects in either the production or accumulation of receptor protein. Quantitation of immunoreactivity and hormone binding activity present in wild type and mutant S49 cells indicates that these cells contain significantly more immunoreactive material than hormone binding activity. We conclude that S49 cells produce, in addition to their well characterized wild type or mutant receptors, a mutant receptor from a second allele which is of wild type size, is immunologically reactive, but is unable to bind hormone. The S49 mutant cell line nti (nuclear transfer increase) contains a glucocorticoid receptor which has a molecular weight of 40,000, while the wild type receptor has a molecular weight of 94,000. Affinity labeling of glucocorticoid receptors in nti cells with [3H]dexamethasone mesylate indicates that nti cells do not contain wild type sized precursor molecules which bind hormone, nor do they contain immunoreactive fragments of a molecular mass smaller than 94 kDa. It is proposed that the 40-kDa nti receptor is produced as a truncated protein most likely resulting from a nonsense mutation or from a truncated messenger RNA.     

Molecules acting on receptor level at weaning, durably influence liver glucocorticoid receptors

In the experiments the effect of late hormonal imprinting to the liver glucocorticoid receptors were studied. Three-week-old (weanling) female rats were treated with five molecules acting at receptor level and four weeks later receptor kinetic analysis was done on liver glucocorticoid receptors. The tricyclic antidepressant, histamine and serotonin receptor blocker mianserin positively influenced receptor density and negatively receptor affinity. Vitamin D3 and the environmental pollutant benzpyrene elevated receptor density. Mifepristone (RU 486) which is bound by progesterone- and glucorticoid-receptor without postreceptorial effects was ineffective as well, as the H1 receptor blocker chlorpheniramine. The results demonstrate that receptor-level-acting foreign molecules can durably influence the binding capacity of glucocorticoid receptors, however, this is not a general phenomenon and it is not dependent on the type of receptors (membrane or cytosol). Those molecules were effective which 1. have receptor in the same receptor family (vitamin D3) and have postreceptorial effect, or 2. have a structure similar to steroids (benzpyrene) or 3. deeply influenced steroid receptors in earlier experiments (mianserin). This effect should be considered before administering such type of medicaments.     

Adrenocorticoid action in the spinal cord: Some unique molecular properties of glucocorticoid receptors

1. Glucocorticoid hormones affect several functions of the spinal cord, such as synaptic transmission, biogenic amine content, lipid metabolism, and the activity of some enzymes (ornithine decarboxylase, glycerolphosphate dehydrogenase), indicating that this tissue is a target of adrenal hormones. 2. Corticosterone, the main glucocorticoid of the rat, is detected at all regional levels of the spinal cord, and cold stress increases this steroid, predominantly in the cervical regions. 3. Intracellular glucocorticoid receptors have been found in the spinal cord, with higher concentrations in the cervical and lumbar enlargements. Prima facie, these receptors presented biochemical, stereospecifical, and physicochemical properties similar to those of receptors found in other regions of the nervous system. The prevalent form in the spinal cord is the type II receptor, although type I is also present in small amounts. 4. The type II glucocorticoid receptor of the spinal cord shows an affinity lower (Kd 3.5 nM) than that of the hippocampal type II site (Kd 0.7 nM) when incubated with [3H]dexamethasone. This condition may impair the nuclear translocation of the spinal cord receptor. 5. Another peculiar property of spinal cord type II site is a greater affinity for DNA-cellulose binding than the hippocampal receptor during heat-induced transformation. Also, the spinal cord receptor shows resistance to the action of RNAse A, an enzyme which increases DNA-cellulose binding of the hippocampal receptor, indicating that both receptors may be structurally different. 6. Therefore, it is possible that a different subclass of type II, or "classical glucocorticoid receptor," is present in the spinal cord. This possibility makes the cord a useful system for studying diversity of glucocorticoid receptors of the nervous system, especially the relationship between receptor structure and function.     

Cytoplasmic 8 S glucocorticoid receptor binds to actin filaments through the 90-kDa heat shock protein moiety

The glucocorticoid receptor exists in the cytoplasm of hormone-untreated cells as a complex with the 90-kDa heat shock protein (HSP90). Glucocorticoids induce dissociation of the glucocorticoid binding protein from HSP90 and translocation of the receptor to the nucleus. HSP90 binds to actin filaments, and calmodulin or tropomyosin inhibits the binding. We present here evidence that the HSP90-containing glucocorticoid receptor complexes (8 S receptor) bind to filamentous actin in vitro while the HSP90-free form of the receptor does not. The binding was detectable for both the crude cytosolic fractions and the partially purified 8 S glucocorticoid receptor. Purified HSP90 or tropomyosin completely abolished the binding. Calmodulin also inhibited the binding in a Ca(2+)-dependent manner. From these results, we conclude that the glucocorticoid receptor complex is able to bind actin filaments via the HSP90 moiety. The binding may provide an anchoring mechanism for the glucocorticoid receptor in the cytoplasm.     

Somatostatin increases glucocorticoid binding and signaling in macrophages by blocking the calpain-specific cleavage of Hsp 90

Somatostatin has direct anti-inflammatory actions and participates in the anti-inflammatory actions of glucocorticoids, but the mechanisms underlying this regulation remain poorly understood. The objective of this study was to evaluate whether somatostatin increases glucocorticoid responsiveness by up-regulating glucocorticoid receptor (GR) expression and signaling. Somatostatin promoted a time- and dose-dependent increase in [(3)H]dexamethasone binding to RAW 264.7 macrophages. Cell exposure to 10 nM somatostatin for 18 h promoted a 2-fold increase in the number of GR sites per cell without significant modification of the affinity. Analysis of GR heterocomplex components demonstrated that somatostatin increased the level of heat shock protein (Hsp) 90, whereas the level of GR remained almost unchanged. The increase in Hsp 90 was associated with a decrease in the cleavage of its carboxyl-terminal domain. Evidence for the involvement of calpain inhibition in this process was obtained by the demonstration that 1) somatostatin induced a dose-dependent decrease in calpain activity and 2) calpain inhibitors, calpain inhibitor I and calpeptin, both abolished the cleavage of Hsp 90 and induced a dose-dependent increase in [(3)H]dexamethasone binding. Increases in glucocorticoid binding after somatostatin treatment were associated with similar increases in the ability of GR to transactivate a minimal promoter containing two glucocorticoid response elements (GRE) and to interfere with the activation of nuclear factor-kappaB (NF-kappaB). Thus, the present findings indicate that somatostatin increases glucocorticoid binding and signaling by limiting the calpain-specific cleavage of GR-associated Hsp 90. This mechanism may represent a novel target for intervention to increase glucocorticoid responsiveness.     

Rat liver glucocorticoid receptor isolated by affinity chromatography is not a Mg2+- or Ca2+-dependent protein kinase

The glucocorticoid hormone receptor (92 kDa), purified 9000-fold from rat liver cytosol by steroid affinity chromatography and DEAE-Sephacel chromatography, was assayed for the presence of protein kinase activity by incubations with [gamma-32P]ATP and the photoaffinity label 8-azido-[gamma-32P]ATP. Control preparations isolated by affinity chromatography in the presence of excess steroid to prevent the receptor from binding to the affinity matrix were assayed for kinase activity in parallel. The receptor was not labeled by the photoaffinity label under photoactivation conditions in the presence of Ca2+ or Mg2+. A Mg2+-dependent protein kinase (48 kDa) that could be photoaffinity labeled with 8-azido-ATP copurified with the receptor. This kinase was also present in control preparations. The kinase could phosphorylate several minor contaminants present in the receptor preparation, including a protein (or proteins) of similar molecular weight to the receptor. The phosphorylation of 90-92-kDa proteins was independent of the state of transformation or steroid-binding activity of the receptor. These experiments provide direct evidence that neither the glucocorticoid receptor nor the 90-92-kDa non-steroid-binding protein associated with the molybdate-stabilized glucocorticoid receptor possesses intrinsic Ca2+- or Mg2+-dependent protein kinase activity.     

Circadian and seasonal variations of glucocorticoid receptors in normal human lymphocytes

Circulating human lymphocytes are known to contain specific glucocorticoid receptors. Using a competitive binding whole cell assay, we have examined the binding of [3H] dexamethasone to peripheral lymphocytes of normal male subjects. Lymphocytes were found to contain 2000-10000 glucocorticoid receptor sites/cell and the Kd value was in the range of 0.5-9 X 10(-9) M. The number and affinity of glucocorticoid receptors did not change throughout a 1-year observation time. In contrast, there was a significant diurnal variation in receptor content (38% higher at 11 p.m. than at 8 a.m.), while receptor affinity did not change.     

A new affinity matrix for mineralocorticoid receptors

The behavior of mineralocorticoid and glucocorticoid receptors of rabbit kidney cytosol was investigated on two affinity gels: a new affinity matrix prepared with a 3-O-derivative of carboxymethyloxime deoxycorticosterone (deoxycorticosterone gel) and a gel linked to a 17 beta-dexamethasone derivative (dexamethasone gel). Deoxycorticosterone gel was highly specific, since it retained mineralocorticoid but not glucocorticoid receptors, and dexamethasone gel exhibited high selectivity for glucocorticoid receptors since it did not bind mineralocorticoid receptors. The use of these two matrices allowed separation of mineralocorticoid and glucocorticoid receptors and further characterization of each type of cytosolic receptors after its isolation. Cytosolic mineralocorticoid and glucocorticoid receptors stabilized by tungstate were found to have a Stokes radius of approximately 6 nm, as determined by high performance size exclusion chromatography and a sedimentation coefficient of approximately 9 S, determined on a glycerol density gradient containing tungstate, under either high or low salt conditions. The hydrodynamic parameters, binding characteristics, and specificity of mineralocorticoid receptors were the same in the untreated and dexamethasone gel-treated cytosol. Similarly glucocorticoid receptor characteristics remained unchanged after deoxycorticosterone gel treatment, indicating biochemical independence of cytosolic mineralocorticoid and glucocorticoid receptors. The [3H]aldosterone receptor complex eluted from deoxycorticosterone gel was recovered with a 30-40% yield and a purification factor of about 1000. Purified mineralocorticoid receptors had the same sedimentation coefficient as cytosolic mineralocorticoid receptors (9 S) but a different Stokes radius (4 versus 6 nm). The decrease in the Stokes radius of the purified mineralocorticoid receptors was probably due to the gel filtration method. These results indicate that the newly synthesized matrix specific for mineralocorticoid receptors constitutes a powerful tool for their extensive purification.     

Role of cysteines 640, 656, and 661 in steroid binding to rat glucocorticoid receptors.

The involvement of a vicinally spaced dithiol group in steroid binding to the glucocorticoid receptor has been deduced from experiments with the thiol-specific reagent methyl methanethiolsulfonate and the vicinal dithiol-specific reagent sodium arsenite. The vicinally spaced dithiol appears to reside in the 16-kDa trypsin fragment of the receptor, which is thought to contain 3 cysteines (Cys-640, -656, and -661 of the rat receptor) and binds hormone with an approximately 23-fold lower affinity than does the intact 98-kDa receptor. We now report that the steroid binding specificity of preparations of this 16-kDa fragment and the intact receptor are virtually identical. This finding supports our designation of the 16-kDa fragment as a steroid-binding core domain and validates our continued use of this tryptic fragment in studies of steroid binding. To identify the cysteines which comprise the vicinally spaced dithiol group, and to examine further the role of cysteines in steroid binding, a total of five point mutant receptors were prepared: cysteine-to-serine for each suspected cysteine, cysteine-to-glycine for Cys-656, and the C656,661S double mutant. Unexpectedly, each receptor with a single point mutation still bound steroid. Even the double mutant (C656,661S) bound steroid with wild type affinity. These results suggest that none of these cysteines are directly required either for steroid binding to the glucocorticoid receptor or for heat shock protein 90 association with the receptor. However, the presence of Cys-656 was obligatory for covalent labeling of the receptor by [3H]dexamethasone 21-mesylate. Studies with preparations of the 98 and 16 kDa forms of these mutant receptors revealed both that Cys-656 and -661 comprise the vicinally spaced dithiols reacting with arsenite and that any two of the three thiols could form an intramolecular disulfide after treatment with low concentrations of methyl methanethiolsulfonate. These data, in conjunction with those from experiments on the effects of steric bulk on various receptor functions, support a model for the ligand binding cavity of the receptor that involves all three thiols in a flexible cleft but where thiol-steroid interactions are not essential for binding.