Characterization of the nontransformed and transformed androgen receptor and heat shock protein 90 with high-performance hydrophobic-interaction chromatography

DHT selectively reverses Smad3-mediated/TGF-beta-induced responses through transcriptional down-regulation of Smad3 in prostate epithelial cells

Androgens suppress TGF-β responses in the prostate through mechanisms that are not fully explored. We have recently reported that 5α-dihydrotestosterone (DHT) suppresses the ability of TGF-β to inhibit proliferation and induce apoptosis of prostatic epithelial cells and provided evidence that such suppression was fueled by transcriptional down-regulation of TGF-β receptor II (ΤβRII). We now show that androgen receptor (AR) activated by DHT suppresses the TGF-β-induced phosphorylation of Sma- and Mad-related protein (Smad)3 in LNCaP cells overexpressing TβRII under the control of a cytomegalovirus promoter, which is not regulated by DHT, suggesting that transcriptional repression of TβRII alone does not fully account for the impact of DHT on TGF-β responses. Instead, we demonstrate that such suppression occurs through loss of total Smad3, resulting from transcriptional suppression of Smad3. We provide evidence that DHT down-regulates the promoter activity of Smad3 in various prostate cancer cell lines, including NRP-154+AR, DU145+AR, LNCaP, and VCaP, at least partly through androgen-dependent inactivation of Sp1. Moreover, we show that overexpression of Smad3 reverses the ability of DHT to protect against TGF-β-induced apoptosis in NRP-154+AR, supporting our model that loss of Smad3 by DHT is involved in the protection against TGF-β-induced apoptosis. Together, these findings suggest that deregulated/enhanced expression and activation of AR in prostate carcinomas may intercept the tumor suppressor function of TGF-β through transcriptional suppression of Smad3, thereby providing new mechanistic insight into the development of castration-resistant prostate cancer.

Androgenic control of transforming growth factor-beta signaling in prostate epithelial cells through transcriptional suppression of transforming growth factor-beta receptor II

The androgen receptor cross-talks with transforming growth factor-beta (TGF-beta) through mechanisms that remain poorly understood. Here we provide strong evidence that 5alpha-dihydrotestosterone (DHT) intercepts the ability of prostate epithelial cells to undergo TGF-beta-induced apoptosis, and present a new model for this androgenic effect. We report that DHT decreases the level of TGF-beta receptor II (TbetaRII) through a transcriptional mechanism, leading to suppression of the ability of TGF-beta to down-regulate expression of Bcl-xL and cyclin Ds, activate caspase-3, and induce apoptosis. Promoter analysis, DNA pulldown, and electrophoretic mobility shift assays support that transcriptional down-regulation of TbetaRII by DHT occurs through Sp1/Sp3 response elements, with the binding of Sp1 to the TbetaRII promoter being suppressed by DHT, largely driven by loss of Sp1 protein and/or activity. These results provide fresh insight on the mechanism of growth control by androgens and the progression of prostate cancer to androgen independence. [Cancer Res 2008;68(19):8173-82].

PRMT2, a member of the protein arginine methyltransferase family, is a coactivator of the androgen receptor

The basal transcriptional activity of nuclear receptors (NRs) is regulated by interactions with additional comodulator proteins (coactivator/corepressor). Here, we describe a new androgen receptor (AR)-associated coactivator, PRMT2, which belongs to the arginine methyltransferase protein family. To search for AR-interacting proteins a fragment of the AR was used in a library screen exploiting the yeast two-hybrid technique and identifying the C-terminal region of PRMT2. We demonstrated that PRMT2 acts as a strong coactivator of the AR, had modest or none influence on transcriptional activation mediated by other NRs. Interestingly, PRMT2 interaction with the estrogen receptor (ER) was strongly dependent on the cellular background, thus, suggesting the involvement of additional, differentially expressed coregulators. We also demonstrated synergistic interaction of PRMT2 with other known nuclear receptor coactivators, such as GRIP1/TIF-2. Potentiation of AR-mediated transactivation by PRMT2 alone and in synergism with GRIP1 was prevented by a competitive inhibitor of methyltransferase activity. The PRMT2 expression profile overlaps with the distribution of AR, with strongest PRMT2 abundance in androgen target tissues. Immunofluorescence experiments showed that the intracellular localization of PRMT2 depends on the presence of the cognate receptor ligand. Under androgen-free conditions, both AR and PRMT2 are confined to the cytoplasm, whereas in the presence of androgens both proteins colocalize and translocate into the nucleus. Treatment with the AR antagonist hydroxyflutamide results in nuclear translocation of the AR, but not the coactivator PRMT2. Thus, it appears that the ligand-dependent AR conformation is essential for the recruitment and nuclear translocation of PMRT2 which acts as AR-coactivator, presumably by arginine methylation.

FoxG1, a member of the forkhead family, is a corepressor of the androgen receptor

The androgen receptor (AR) is a ligand-dependent transcriptional regulator which belongs to the nuclear receptor superfamily. The basal transcriptional activity of the androgen receptor is regulated by interaction with coactivator or corepressor proteins. The exact mechanism whereby comodulators influence target gene transcription is only partially understood, especially for corepressors. Whereas several coactivators are described for the AR, only a few corepressors are known. Here, we describe the discovery of a new androgen receptor corepressor, FoxG1, which belongs to the forkhead family. By using a fragment of the AR (aa 325-919) as bait in a yeast two hybrid screen, the C-terminal region (aa 175-489) of FoxG1 (also known as BF1), was identified as AR-interacting protein. Binding of AR to the FoxG1 fragment was verified by one- and two-hybrid assays, and pull-down experiments. In addition, we show that the full-length form of FoxG1 functions as a strong corepressor in the AR-mediated transactivation. The FoxG1 expression profile in adult individuals is restricted to brain and testis in human and decreases during aging in the rodent brain. Both AR and FoxG1 expression are developmentally regulated. Besides its reported role in neurogenesis, the strong expression of FoxG1 in AR-abundant areas of the adult brain suggests possible involvement in neuroendocrine regulation. Taken together, the data presented suggest that, in addition to repression of transcription by direct binding to DNA, FoxG1 may interact with AR in vivo, thereby targeting its repressor function specifically to sex hormone signaling.

A novel variant of the putative demethylase gene, s-JMJD1C, is a coactivator of the AR

Evidence is accumulating in support of the view that tissue-specific effects of steroid hormones depend on the recruitment of nuclear receptor comodulator proteins. The latter interact directly with the hormone receptors and modify their transcriptional effects on specific target genes. The mechanisms of comodulator influence on nuclear receptor-controlled gene transcription is only partially understood. Here, we describe the discovery of a new AR coactivator which belongs to the JmjC containing enzyme family as a novel variant of JMJD1C (jumonji domain-containing 1C). By using a fragment of the human AR (aa 325-919) as bait in a yeast two-hybrid screen, a region of the human JMJD1C gene was identified as interacting with AR. A novel splice variant s-JMJD1C was amplified by RACE, and the binding to AR was analysed by GST-pull-down and mammalian one-hybrid experiments. As a nuclear-localized protein, the s-JMJD1C gene is expressed in a variety of human tissues. In the brain, this protein is present in several, but not confined to, AR-expressing neuronal populations and its abundance varies with the hormonal status in a region-specific fashion. Interestingly, the expression of s-JMJD1C is reduced in breast cancer tumors and significantly higher in normal breast tissues indicating a putative role in tumor suppression. As s-JMJD1C has putative demethylase activity, removal of methylation seems to be important for nuclear receptor-based gene regulation.

Aggregation and proteasome: the case of elongated polyglutamine aggregation in spinal and bulbar muscular atrophy

Aggregates, a hallmark of most neurodegenerative diseases, may have different properties, and possibly different roles in neurodegeneration. We analysed ubiquitin-proteasome pathway functions during cytoplasmic aggregation in polyglutamine (polyQ) diseases, using a unique model of motor neuron disease, the SpinoBulbar Muscular Atrophy. The disease, which is linked to a polyQ tract elongation in the androgen receptor (ARpolyQ), has the interesting feature that ARpolyQ aggregation is triggered by the AR ligand, testosterone. Using immortalized motor neurons expressing ARpolyQ, we found that a proteasome reporter, YFPu, accumulated in absence of aggregates; testosterone treatment, which induced ARpolyQ aggregation, allowed the normal clearance of YFPu, suggesting that aggregation contributed to proteasome de-saturation, an effect not related to AR nuclear translocation. Using AR antagonists to modulate the kinetic of ARpolyQ aggregation, we demonstrated that aggregation, by removing the neurotoxic protein from the soluble compartment, protected the proteasome from an excess of misfolded protein to be processed.

Antimyeloma activity of heat shock protein-90 inhibition

We show that multiple myeloma (MM), the second most commonly diagnosed hematologic malignancy, is responsive to hsp90 inhibitors in vitro and in a clinically relevant orthotopic in vivo model, even though this disease does not depend on HER2/neu, bcr/abl, androgen or estrogen receptors, or other hsp90 chaperoning clients which are hallmarks of tumor types traditionally viewed as attractive clinical settings for use of hsp90 inhibitors, such as the geldanamycin analog 17-AAG. This class of agents simultaneously suppresses in MM cells the expression and/or function of multiple levels of insulin-like growth factor receptor (IGF-1R) and interleukin-6 receptor (IL-6R) signaling (eg, IKK/NF-kappaB, PI-3K/Akt, and Raf/MAPK) and downstream effectors (eg, proteasome, telomerase, and HIF-1alpha activities). These pleiotropic proapoptotic effects allow hsp90 inhibitors to abrogate bone marrow stromal cell-derived protection on MM tumor cells, and sensitize them to other anticancer agents, including cytotoxic chemotherapy and the proteasome inhibitor bortezomib. These results indicate that hsp90 can be targeted therapeutically in neoplasias that may not express or depend on molecules previously considered to be the main hsp90 client proteins. This suggests a more general role for hsp90 in chaperoning tumor- or tissue-type-specific constellations of client proteins with critical involvement in proliferative and antiapoptotic cellular responses, and paves the way for more extensive future therapeutic applications of hsp90 inhibition in diverse neoplasias, including MM.

Mechanism of p21-activated kinase 6-mediated inhibition of androgen receptor signaling

PAK6 was first identified as an androgen receptor (AR)-interacting protein able to inhibit AR-mediated transcriptional responses. PAK6 is a serine/threonine kinase belonging to the p21-activated kinase (PAK) family implicated in actin reorganization and cell motility, gene transcription, apoptosis, and cell transformation. We investigated the biochemical basis for inhibition of AR signaling by PAK6. We compared the kinase activity of PAK6 with two other well characterized members of the PAK family, PAK1 and PAK4. Like PAK4, PAK6 possesses a constitutive basal kinase activity that, unlike PAK1, is not modulated by the binding of active Rac or Cdc42 GTPases. In order to test the involvement of PAK6 kinase activity in suppression of AR-mediated transcription, we generated kinase-dead (K436A) and kinase-active (S531N) mutants of PAK6. We show that PAK6 kinase activity is required for effective PAK6-induced repression of AR signaling. Suppression does not depend upon GTPase binding to PAK6 and is not mimicked by the closely related PAK1 and PAK4 isoforms. Kinase-dependent inhibition by PAK6 extended to the enhanced AR-mediated transcription seen in the presence of coactivating molecules and to the action of AR coinhibitors. Active PAK6 inhibited nuclear translocation of the stimulated AR, suggesting a possible mechanism for inhibition of AR responsiveness. Finally, we observe that autophosphorylated, active PAK6 protein is differently expressed among prostate cancer cell lines. Modulation of PAK6 activity may be responsible for regulation of AR signaling in various forms of prostate cancer.

Ribosomal RNA transcriptional activation and processing in hamster rubrospinal motoneurons: effects of axotomy and testosterone treatment

Rubrospinal motoneurons (RSMN) represent a population of androgen receptor-expressing central motoneurons with limited regenerative potential relative to their peripheral counterparts. A key determinant of regenerative capability lies in the nucleolar reaction of injured neurons. To date, characterization of the nucleolar reaction in injured central motoneurons has not been accomplished. Furthermore, it has been documented that testosterone propionate (TP) augments peripheral motoneuron regeneration through regulation of the nucleolar reaction to injury. In this study, the effects of injury alone, or in conjunction with TP, on the nucleolar response of injured RSMN were examined using in situ hybridization (ISH) techniques. Castrated adult male hamsters were subjected to right spinal cord hemisection at the C7/T1 vertebral level. Half the animals were subcutaneously implanted with one Silastic TP capsule, with the other half sham implanted. ISH for precursor 45S and mature 28S rRNA was accomplished with a (3)H-labeled ribosomal DNA probe specific to the external transcribed spacer region or to the 28S region of the ribosomal gene, respectively. Postoperative times of 2, 6, and 24 hours were selected for examination of precursor 45S rRNA (i.e., rRNA transcriptional activation) levels and 0.25, 2, 4, and 14 days for examination of mature rRNA (i.e., ribosome) levels. Transcriptional activation of the rRNA gene was rapidly and transiently increased in injured RSMN, analogously to previously documented effects of injury on rRNA transcription in peripheral motoneurons, but, in contrast, this did not translate into an increase in mature ribosomes. TP administration failed to affect positively the nucleolar response of injured RSMN at all. From this study, a key component underlying inherent differences in the regenerative capacity of peripheral vs. central motoneurons has been identified, which can be targeted in future experiments designed to enhance the regenerative potential of selective neuronal populations.

Loss of androgen receptor transcriptional activity at the G(1)/S transition

Androgens are essential for the differentiation, growth, and maintenance of male-specific organs. The effects of androgens in cells are mediated by the androgen receptor (AR), a member of the nuclear receptor superfamily of transcription factors. Recently, transient transfection studies have shown that overexpression of cell cycle regulatory proteins affects the transcriptional activity of the AR. In this report, we characterize the transcriptional activity of endogenous AR through the cell cycle. We demonstrate that in G0, AR enhances transcription from an integrated steroid-responsive mouse mammary tumor virus promoter and also from an integrated androgen-specific probasin promoter. This activity is strongly reduced or abolished at the G(1)/S boundary. In S phase, the receptor regains activity, indicating that there is a transient regulatory event that inactivates the AR at the G(1)/S transition. This regulation is specific for the AR, since the related glucocorticoid receptor is transcriptionally active at the G(1)/S boundary. Not all of the effects of androgens are blocked, however, since androgens retain the ability to increase AR protein levels. The transcriptional inactivity of the AR at the G(1)/S junction coincides with a decrease in AR protein level, although activity can be partly rescued without an increase in receptor. Inhibition of histone deacetylases brings about this partial restoration of AR activity at the G(1)/S boundary, demonstrating the involvement of acetylation pathways in the cell cycle regulation of AR transcriptional activity. Finally, a model is proposed that explains the inactivity of the AR at the G(1)/S transition by integrating receptor levels, the action of cell cycle regulators, and the contribution of histone acetyltransferase-containing coactivators.

Interaction of the 90-kDa heat shock protein with native and in vitro translated androgen receptor and receptor fragments

Androgen receptor (AR) from rat ventral prostate and AR synthesized in vitro by translation in rabbit reticulocyte lysate of AR mRNA, transcribed from a pGEM-4Z DNA template were compared by gel permeation chromatography and by sucrose gradient ultracentrifugation. Under non-activating conditions the AR from rat prostate migrated as an 8-9 S complex of approx. 300 kDa. The addition of chicken antibodies against HSP90 shifted this complex to the void volume of the column or to the bottom of the ultracentrifugation gradient. Under activating conditions, on the other hand, the AR migrated as a 110 kDa, 5.2 S protein and was no longer displaced by HSP90 antibodies. Under all these conditions, the behaviour of in vitro synthesized AR was very similar to that of AR from rat prostate. By selective use of restriction enzymes on the template of transcription AR mutants could be prepared from which an increasing part was deleted at their carboxy terminal end. The interaction with HSP90 was conserved for AR1-758 missing the last 145 amino acids, but was lost in AR1-703. Furthermore, a large internal deletion (ARd41-469) of the major part of the amino terminal half of the AR did not result in the loss of HSP90 binding. These results indicate that a specific subregion (amino acids 704-758) of the carboxy terminal half of the AR is required for the interaction with HSP90.

Association of the 90-kDa heat shock protein does not affect the ligand-binding ability of androgen receptor

An N-terminal truncated androgen receptor with putative DNA- and ligand-binding domains (AR438) and that with a ligand-binding domain (AR612) were expressed under control of the T7 promoter in E. coli or translated in vitro with rabbit reticulocyte lysate, and their ligand-binding properties and the interaction with HSP90 were investigated. Bacterially expressed AR438 and AR612 bound a synthetic androgen, [3H]R1881, with apparent dissociation constants of 2.6 +/- 0.2 and 3.1 +/- 0.7 nM, respectively, values which are comparable to those of androgen receptor in target tissues. The recombinant androgen receptors sedimented at the 4-5 S region irrespective of the presence of 10 mM tungstate, indicating that the receptor exists free from HtpG, which is the bacterial homolog of eukaryotic HSP90. The apparent dissociation constant of truncated androgen receptors translated in vitro was 0.1 nM for AR438 and 0.2 nM for AR612. Sedimentation coefficients of in vitro translated molecules were converted from 7-8 S in the presence of tungstate to 3 S in the absence of tungstate. Both AR438 and AR612 translated in vitro were retained by anti-rat HSP90 antibody-protein A Sepharose. Exposure to 0.3 M NaCl in the presence of ligand caused dissociation of AR438 and AR612 from HSP90, and concomitantly, the DNA-cellulose binding ability of AR438 was enhanced. Thus, we conclude that the androgen receptor associates with HSP90 through the ligand-binding domain and that this association prevents the interaction of the androgen receptor with DNA. However, HSP90 seems to have little effect on the ligand-binding characteristics of the androgen receptor.

Estradiol increases phosphorylation of the 90 kDa heat shock protein not associated with estradiol receptor in MCF-7 cells in culture

MCF-7 cells in monolayer culture were incubated with [32P]orthophosphate for 18 h followed by covalent whole cell labelling of the estradiol receptor with tritiated tamoxifen aziridine [( 3H]TA). The heat shock protein (hsp-90) bound to receptor was precipitated with monoclonal antibodies H222 or JS 34/32, coupled to protein A-Sepharose and purified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) under reducing conditions. Hsp-90 not associated with receptor was similarly purified after isolation with the monoclonal antibody AC88. It was found that estradiol treatment of the cells markedly increased phosphate incorporation in the free hsp-90, without affecting heat shock protein bound to receptor. A 6-fold increase in phosphate content was observed after 10 min incubation of the cells with estradiol. A similar effect was seen after treatment of the cells with the phorbol ester 12-O-tetradecanoylphorbol 13-acetate (TPA). The calcium ionophore A23187 had no influence on hsp-90 phosphorylation, and treatment of the cells with forskolin to increase the cellular content of cAMP had a reverse effect. A 50% reduction of the phosphate content in the free hsp-90 was observed after 15 min treatment. The observation that estradiol, TPA and forskolin had effect only on hsp-90 not bound to receptor is an indication that the receptor-hsp-90 complex exists in vivo. Time course studies show that the effect of estradiol is non-genomic. Two possible explanations of the results seem to exist. Either estradiol induces an increase in the degree of phosphorylation of hsp-90, or hsp-90 is translocated to the cytosol from a different cellular compartment.

Antibodies against synthetic peptides recognize the human and rat androgen receptor

Antibodies against two synthetic peptides (aa 299-311 and aa 544-559) selected in different immunogenic domains of the human AR, were induced in rabbits. Antiserum reactivity against the native receptor was investigated by gel permeation chromatography and sucrose density gradient centrifugation using [3H]mibolerone-labeled rat prostate cytosol and [3H]5 alpha-dihydrotestosterone-labeled T-47D cytosol as a source of AR. The absence of cross-reactivity of the antisera with estrogen, progesterone and glucocorticoid receptor was confirmed by density gradient centrifugation of rat uterus cytosol labeled with [3H]E2 or [3H]ORG 2058 and rat liver cytosol labeled with [3H]dexamethasone. After partial proteolytic breakdown of rat prostate AR by endogenous proteases the steroid-labeled receptor was recognized only by the second peptide (aa 544-559) antibody. This proteolytic breakdown could be prevented to a large degree by addition of a high concentration of soybean trypsin inhibitor. The specific AR antibodies provide new tools for the functional analysis of AR, since they interact selectively with specific domains of the receptor.

Reconstitution of the 9 S estrogen receptor with heat shock protein 90

As a first step in the investigation of the reconstitution of steroid hormone receptor systems, we studied the reconstitution of 9 S estrogen receptor (ER) from purified vero ER, which is the estradiol binding subunit, and heat shock protein 90 (hsp 90). By using a phosphate buffer containing molybdate, thiocyanate, dimethylformamide, glycerol, etc., vero ER could be converted to 9 S ER with hsp 90, but not with the control protein, ovalbumin. Inactivation of ER during the reconstitution was suppressed partially by hsp 90, but not by ovalbumin. Like native 8 S ER, the reconstituted ER was sedimented at about 8.9 S and 4.6 S on glycerol gradient centrifugation in low and high salt buffers, respectively.

The transformed glucocorticoid receptor has a lower steroid-binding affinity than the nontransformed receptor

High-salt treatment of cytosolic glucocorticoid receptor (GR) preparations reduces the steroid-binding ability of the receptor and induces the conversion of the receptor from a nontransformed (non-DNA-binding) 9S form to a transformed (DNA-binding) 4S entity. Therefore, we decided to investigate the possible relationship between these two phenomena. Steroid-free GR was converted from a 9S to a 4S form by exposure to 0.4 M NaCl. The binding of [3H]triamcinolone acetonide [( 3H]TA) to the 9S form was almost saturated at a concentration of 20 nM, whereas [3H]TA was hardly bound to the 4S form at this concentration. The 4S form was efficiently labeled at 200 nM. Scatchard analysis of the GR exposed to 0.4 M NaCl in the presence of 10 mM molybdate showed the presence of two types of binding sites with apparent dissociation constants of 0.52 +/- 0.07 and 64.1 +/- 16.2 nM, respectively. In the absence of molybdate, the ratio of the lower affinity site was increased, but the total number of binding sites was not modified. The GR with the low [3H]TA-binding affinity bound to DNA-cellulose even in its unliganded state, whereas the form with the high affinity did not. Immunoblot analysis using anti-GR monoclonal antibody revealed no difference in molecular size (Mr 94000) between the high- and low-affinity entities. These results indicate that the transformed GR has a reduced [3H]TA-binding affinity as compared to the nontransformed GR.(ABSTRACT TRUNCATED AT 250 WORDS)

Separation of two molecular forms of human estrogen receptor by hydrophobic interaction chromatography. Gradient optimization and tissue comparison

High-performance hydrophobic interaction chromatography (HPHIC) was used to separate and characterize two molecular forms of estrogen receptor with a SynChropak propyl hydrophobic column (300 A pore size). The linear gradient utilized earlier with a polyether-bonded column (2 to 0 M) ammonium sulfate in 40 min, gave poor resolution with the propyl column. However, resolution was maximized with either an initial ammonium sulfate concentration of 1 M (40-min gradient) or with a two-phase gradient (2 to 0.5 M in 10 min, 0.5 to 0 M in 30 min). This indicated that the propyl column was more hydrophobic than the polyether column. Estrogen receptor separated into two isoforms, either in the presence [MI, retention time (tR) = 13-14 min; MII, tR = 20-21 min] or absence (I, tR = 21-23 min; II, tR = 31-33 min) of the estrogen receptor stabilizing reagent, sodium molybdate. Similar isoforms were observed in cytosols from human breast tumors, uterus, and MCF-7 breast cancer cells. Unlike others, MCF-7 estrogen receptor did not show MI. Since MCF-7 cells contain 90,000 dalton heat shock proteins (HSP-90), HSP-90 is probably not directly involved in MI formation. Sodium molybdate selectively interacted with isoform II and converted it to MI. All isoforms appeared to be high-molecular-weight proteins (greater than 60 A) when subsequently analyzed by high-performance size-exclusion chromatography. Interestingly, when estrogen receptor was immobilized on the stationary phase, no change was detected in either hydrophobicity or steroid-binding capacity. After 16-18 h, immobilized receptor was eluted with a slightly longer tR. During incubation on the column, component MI was converted into I and/or II. HPHIC appears to be a rapid, yet gentle procedure for isolating large receptor complexes in significant quantities with high recoveries. This allows one to discern the complicated structure-function relationships of estrogen receptor and associated non-receptor proteins and provides information about the on-column behavior of complex proteins.