Androgen receptor coregulators in prostate cancer: mechanisms and clinical implications.

Abstract

The transcriptional activity of androgen receptor (AR) is modulated by coregulators that have a significant influence on a number of functional properties of AR, including the ligand specificity as well as the DNA binding capacity. Because androgens and AR have pivotal roles in the development and growth of normal prostate as well as prostate cancer progression, these coregulators may significantly modulate the events of cancer development and progression. An aberrant regulation or expression of AR coregulators may contribute to androgen and AR-related diseases including prostate cancer. A precise understanding of the roles and mechanisms of individual coregulators may provide valuable information in designing rational therapy against prostate cancer. In this review, we analyzed the up-dated information regarding AR coregulators and their potential roles in prostate cancer progression.Androgens play essential roles in male sexual differentiation, sexual maturation, and the maintenance of spermatogenesis (1, 2). In addition to such vital roles in reproductive organ development, androgens also mediate physiological responses in a wide range of other tissues and organs including the skin, skeletal muscle, cardiac muscle, liver, kidney, central nervous system, and the hematopoietic system (3).The prostate gland is one of the main target organs for androgens, which mediate their biological responses through the AR, a Mr 110,000 ligand-inducible transcription factor that regulates tissue-specific expression of target genes. The AR belongs to the nuclear receptor (NR) superfamily, which also includes estrogen receptor, glucocorticoid receptor, progesterone receptor, vitamin D receptor, thyroid receptor, and retinoic acid receptor (4). It is ubiquitously expressed in human organs with the exception of the spleen and bone marrow, which show complete absence of AR expression. The AR gene is composed of eight exons encoding four functional domains similar to other members of the NR superfamily: a conserved DNA-binding domain, a hinge region, a COOH-terminal ligand-binding domain (LBD), and the less conserved NH2-terminal domain [A/B domain (5)]. Mutational and deletion analyses revealed two transcriptional activation functions including an NH2-terminal activation function (AF-1) involved in ligand-independent activation of AR and a ligand-dependent AF-2 function in the LBD. The AF-2 seems to be involved in more general transcriptional activity because mutation or deletion of the AF-2 domain dramatically reduces the transcriptional activation of AR (6, 7).Since cloning of its cDNA structure in 1988 (8), the AR has been extensively studied to elucidate its biological role in the development and progression of prostate cancer. The growth, differentiation, and maintenance of prostate tissue are dependent on AR and androgens, particularly testosterone and dihydrotestosterone (DHT). Overactivation of the AR signaling pathway may link to the progression of prostate cancer, which remains the most frequently diagnosed malignancy in aging men. Androgens and the AR play key roles in this malignancy, and each year about 30,000 men lose their lives to this malignancy in the United States (9). In 1941, Huggins and Hodges (10) first reported that the normal prostate and early stages of prostate cancer were highly dependent on androgens. Since then, androgen ablation, either by surgical castration or by pharmaco-logical/chemical castration, such as treatment with luteinizing hormone-releasing hormone analogs, has been the cornerstone of treatment in patients with locally advanced or metastatic prostate cancer. Androgen deprivation or the blockade of androgens at the level of the AR usually results in a favorable clinical response and a dramatic tumor regression (11, 12). Although complete disappearance of symptoms is achieved initially, in most cases the response does not imply cure. Upon continued treatment, the therapy fails, and most of the patients develop androgen-independent spread of the disease despite androgen ablation. At present there is no effective therapy for such an androgen-independent prostate cancer. One reason for failure of this androgen ablation therapy may be because of the presence of residual adrenal androgens in the circulation. Castration only abolishes testicular androgens, but the circulating adrenal androgens are readily available and can be converted to testosterone or DHT, the more active form of testosterone, in the prostate. To overcome this effect, antiandrogens are given in combination with castration (the so-called maximal androgen blockade therapy) to block the effects of androgens derived from the adrenals. However, support for additional benefit from this combined therapy is controversial, and a recent 5-year survival study failed to offer much hope for prostate cancer patients (13). Even after complete blockade of androgen action, the response to therapy is lost, tumors continue to grow, and the disease progresses into an androgen-independent state (14, 15, 16). The exact mechanisms of such resistance against ablation therapy remain largely unknown.One of the hallmarks of androgen-independent prostate cancer is that tumor growth may still depend mostly on a functional AR signaling pathway. Prostate cancers, even when becoming androgen-independent, may still contain many AR-positive cells. Fig. 1 depicts some of the possible mechanisms as to how AR signaling can operate in both androgen-dependent and androgen-independent prostate cancers. One possible mechanism through which prostate cancer cells circumvent ablation therapy is by hypersensitive AR signaling during androgen ablation therapy, part of which can be due to overexpression of 5α-reductase and AR coactivators in prostate tissues. The 5α-reductase may help tumor cells rapidly convert available testicular androgens into more active DHT, and the coactivators may render AR more sensitive to usually weak adrenal androgens. Prostate cancer cells are then able to survive and grow at very low levels of androgens. The AR-mediated growth despite low androgen levels could also be achieved via an increased expression of AR itself, mostly by AR gene amplification, leading to an enhanced ligand occupied AR. Support for this hypothesis comes from the observation that about a third of tumors that become hormone refractory after ablation therapy have an amplified AR gene, compared with none in the primary tumors from the same patients (17, 18, 19, 20, 21).Another possible mechanism by which prostate cancer cells circumvent the androgen requirement is through the development of mutations in the AR gene. The AR LBD, which is responsible for androgen binding and dimerization, may undergo somatic mutations during prostate cancer progression into the hormone-refractory state (14, 15). Mutations of AR allow the cells to circumvent the normal growth regulation by androgens. The AR ligand specificity becomes modulated and broadened so that the receptor now can be activated by nonandrogenic molecules including androgen antagonists, corticosteroids, and 17β-estradiol (E2). For example, the AR mutant T877A, isolated from patients with advanced-stage prostate cancer, is activated not only by androgens, the normal ligands, but also by antiandrogens, such as hydroxyflutamide [HF (22, 23)]. Another AR mutant, L701H, has been reported to significantly decrease androgen response but significantly potentiate the responses by glucocorticoids (24). The broadening of AR specificity thus seems to offer prostate cancer cells a selective growth advantage to evade androgen ablation therapy.Other growth signaling pathways may cross-talk with the AR signaling pathway and activate AR-mediated growth of prostate cancer cells, bypassing the requirement for androgens. For example, a number of growth factors, including insulin-like growth factor-I, keratinocyte growth factor, epidermal growth factor, and so forth, have been reported to activate AR, inducing AR target genes in the absence of androgens (25). A recent report indicated that interleukin 6 could induce activation of AR via signal transducers and activators of transcription-3. Signal transducers and activators of transcription-3 can physically interact with AR and enhance its transcriptional activity, whereas a dominant-negative signal transducers and activators of transcription-3 mutant blocks interleukin 6-induced AR transactivation (26).HER2/neu, a transmembrane glycoprotein that belongs to the epidermal growth factor receptor family of tyrosine kinases, is found to be overexpressed in androgen-independent cells isolated from implanted xenografts in castrated mice. Craft et al. (27) have shown that HER2/neu can modulate the AR response in the absence of androgens or in the presence of extremely low levels of androgens. This was further supported by the findings that HER2/neu can activate AR signaling through the mitogen-activated protein kinase (MAPK) pathway (28). Sequence analysis revealed that the AR gene possesses putative consensus binding motifs (29) for DNA-dependent protein kinase, protein kinase A, protein kinase C, MAPK, and casein kinase II, suggesting a possibility of AR regulation by these kinases. Such possibilities have been supported by the reports that protein kinase A and protein kinase C can activate AR signaling in the absence of androgens (30, 31). Activation of MAPK signaling was found to be associated with prostate cancer progression (32). When constitutively activated, the Ras/MAPK pathway sensitizes hormone-responsive LNCaP cells to subphysiological levels of androgens (33), indicating a role of MAPK pathway in the acquisition of hormone-refractory growth of cancer cells. Taken together, these studies suggested that aberrant regulation of AR signaling pathways might play a role in developing hormone-independent progression of prostate cancer. The existence of so-called AR coregulators, which can modulate AR functional activity in response to agonists and antagonists, may contribute to all of the possible mechanisms described above, and therefore AR coregulators play significant roles in AR-mediated growth and progression of prostate cancers.Substantial amounts of evidence indicate that NRs function as tripartite receptor systems, involving the receptor, its ligand, and its coregulators (34). Coregulators are biological macromolecules that interact with NR to either enhance receptor activity (coactivators) or reduce transcriptional activity (corepressors), without affecting the basal transcriptional level (35). Although coregulators interact with other DNA-binding proteins such as AP-1 and Smad3, they generally do not bind to DNA. Recently, a number of NR coactivators including SRC/pl60 family members such as SRC-1 (36), GRIP1/TIF2 (37), TIF1 (38), RIP140 (39), PGC-1 (40), SNURF (41), and so forth have been identified as being able to modulate receptor transactivation. The AR is a ligand-dependent transcriptional factor that has also been found to be modulated by a large number of coregulators (42). Table 1 summarizes those AR coregulators that can influence AR transactivation in prostate cancer cells and play some roles in prostate cancer progression.The detailed mechanisms of how AR coregulators regulate AR transactivation are unclear and remain to be further investigated. Some AR coregulators, including SRC-1, CBP/p300, and p300/CBP-associated factor, possess histone acetyltransferase activity (43, 44) and may modulate AR activity via their ability to remodel chromatin structure. In addition, posttranslational modifications, such as AR acetylation, may also contribute to modulation of AR activity by these coregulators because it has been shown that AR can be acetylated by CBP in vitro. Mutations on AR acetylation sites dramatically reduce AR activity (45). On the other hand, corepressors, such as histone deacetylase-1, contain histone deacetylase activity and reduce AR activity either directly by affecting AR acetylation status or indirectly by regulating the status of histone acetylation (46, 47). Other AR coregulators lacking such enzymatic activity, such as ARA70, ARA54, ARA55, ARA24/Ran, ARA267, ARA160, supervillin (SV), gelsolin, ARIP3, AES, and TIF2, modulate AR activity via as yet undefined mechanisms (48, 49, 50). Modulation of AR activity by these coregulators can be achieved through multiple mechanisms, such as control of ligand/AR binding affinity, AR expression, AR stability, AR nuclear translocation, and transcription machinery recruitment. The filamin and SNURF1 coregulators can bind to the hinge region of AR and enhance AR activity via facilitation of AR nuclear translocation (51).Although the exact biological significance of AR coregulators interaction has yet to be defined, by and large all AR coregulators can either enhance (coactivate) or diminish (corepress) AR functional activity, and some of the AR coregulators seem to play significant roles in prostate cancer progression. It is possible that selective expression of AR coregulators may allow tumor cells to grow and survive in androgen-depleted conditions.ARA70 was the first AR coregulator identified by yeast two-hybrid screening of the human brain cDNA library (48). It is highly expressed in the brain, thymus, adrenal gland, prostate, and testis. Although initially described as relatively specific to AR, recent studies have indicated that ARA70 could also interact with peroxisome proliferator-activated receptor (PPAR)α or PPARγ (52, 53, 54). It interacts with AR in a ligand-enhanced manner that can induce AR transcriptional activity up to 5–10-fold in some selective cells. Although ARA70 contains a classical LXXLL motif in its NH2-terminal domain, it is not essential for interaction with AR because a mutated LXXLL motif has little influence on AR-ARA70 interaction. Instead, ARA70 may use a newly identified signature motif, FXXLF, as described by Hsu et al. (55). Although the LXXLL motif does not play any role in AR interaction, it may be necessary for interaction with other NRs such as PPARα because mutations in the LXXLL motif strongly interrupt the ARA70 interaction with PPARα, suggesting that ARA70 may use distinct domains to interact with different NRs to modulate their transactivating ability. The NH2-terminal domain (ARA70N) containing amino acids 1–401 shows better coregulator activity than full-length ARA70 and can translocate into the nucleus along with AR. ARA70 can be found both in the cytosol and nuclei of Leydig cells of testes, however, ectopically expressed ARA70 is located predominantly in the cytosol in some cell lines, such as COS-1 cells. Coexpression of ARA70 with AR significantly enhances AR nuclear staining by promotion of AR nuclear translocation and/or stabilization or increase of newly synthesized AR (55).Apart from the general enhancement of AR transcriptional activity, ARA70 has been extensively characterized as having the capacity to enhance AR transcriptional activity in response not only to testicular or normally weak adrenal androgens (56) but also to the antiandrogen HF and E2 in prostate cancer cells (57, 58). Using CV-1 cells, Zhou et al. (59) also showed that ARA70 could induce E2-mediated AR transactivation. In contrast, Gao et al. (53) reported that E2 had little induction of AR transactivation in the absence or presence of ARA70 in CV-1 cells. The reasons for these discrepancies using the same CV-1 cells remain unclear. However, using HeLa, PC-3, and TSU-Pr1 cells, other laboratories also demonstrated that ARA70 could induce E2-mediated AR activity (60, 61, 62). In the CWR22 prostate cancer xenograft model, in which the CWR22 progresses from an androgen-dependent to an androgen-independent state after castration, ARA70 mRNA has been found to be elevated in the recurrent androgen-independent tumors (63). Moreover, adding hypophysiological concentrations of androgens into cells cultured under charcoal-stripped serum conditions or in cancer cells derived from castrated nude mice xenograft results in an increased ARA70 expression. These observations suggest that long-term androgen ablation therapy may result in increased ARA70 expression in tumor cells, which in turn trigger androgen- or antiandrogen-mediated AR activity. Expression of a dominant-negative ARA70 mutant (dARA70) and/or RNA interference-mediated silencing of ARA70 gene reduces the agonist activity of E2 or HF in LNCaP prostate cancer cells (64). A number of studies have reported the important roles of AR mutations in modulation of AR specificity in prostate cancer cells (65, 66, 67). The LNCaP cells used in these studies express a mutant AR (T877A), which is also found in prostate cancer patients, that can change the AR specificity. The inhibition of the agonist activity of antiandrogens in this cell line by dARA70 or ARA70 RNA interference clearly indicates a probable dominant role of ARA70 over AR mutations in the process. Surprisingly, dARA70 showed a very similar effect on DHT, HF, and E2. These blocking effects suggested that the presence of ARA70 may be able to broaden the specificity of the mutant AR, without reducing the affinity for DHT, so that optimal AR transactivation can be maintained by any of these molecules in the presence of ARA70. Apart from mutant AR, the expression of ARA70 may provide prostate cancer cells with a selective growth advantage during antiandrogen therapy. Further in vivo evidence using the LNCaP xenograft nude mice model may be needed for conclusive verification of such phenomena.ARA55 is another AR coregulator that has been characterized as having the capacity to modulate AR specificity in response to agonists and antagonists (50). It was identified from a human prostate cDNA library by yeast two-hybrid screening and has >90% homology with mouse hic-5, a hydrogen peroxide-inducible gene (68). The four LIM domains of ARA55 are similar in sequence and organization to the LIM cytoskeleton-binding protein paxillin and zyxin, consistent with the localization of ARA55 to the nuclear matrix and focal adhesions (69, 70). Previous reports described LIM proteins functioning as bridging molecules between transcription factors, and a number of coregulators have been described to interact with LIM proteins to modulate their transcriptional activity (71, 72). It is also possible that ARA55 may recruit LIM coregulators or other transcription factors to regulate AR transactivation in response to extracellular signals.ARA55 has been reported to enhance AR transcriptional activity in response to normally weak adrenal androgens, an antiandrogen HF, and other nonandrogenic steroids including E2 (56, 57). A frequently occurring AR mutation (T877A) in prostate cancer patients promotes hypersensitivity to ARA55 in the up-regulation of androgen-regulated genes. The commonly used antiandrogens including HF, cyproterone acetate, bicalutamide, and E2 promote (instead of block) AR transcriptional activity in the presence of ARA55 in prostate cancer cells (56). ARA55 may also be involved in the “superactivation” of AR signaling by HER2/neu in the presence of very low levels of androgens in prostate cancer cells (28). In addition, expression levels of ARA55 were found to be higher in tissue samples from hormone-refractory prostate cancers than in those from androgen-dependent cancers (73). Inactivation of endogenous ARA55 by a dominant-negative ARA55 mutant [COOH-terminal fragment with a point mutation (Ala→Thr) at amino acid 413 (413T)] leads to an inhibition of AR transactivation and a reduction of the agonist activity of antiandrogens in prostate cancer cells (74). The COOH-terminal fragment of ARA55 (413A) also weakly suppressed AR transactivation in prostate cancer cells expressing full-length ARA55. Interestingly, ARA55 (443T), but not ARA55 (413A), was detected in prostate cancer tissues. However, the full-length of ARA55 (both 413A and 413T) can equally enhance AR transactivation (75). These findings clearly indicated a role of ARA55 in the acquired agonist activity of antiandrogens in prostate cells. Similar results obtained by using RNA interference-mediated silencing of the ARA55 gene further strengthen these possible roles of ARA55 in prostate cancer progression.ARA55, in addition to AR, has also been shown to interact with the cytoplasmic tyrosine kinase CsK, focal adhesion kinase FAK (76), and FAK-related PYK2 kinase (77). The proline-rich tyrosine kinase PYK2, originally identified as a kinase linked to MAPK and c-Jun-NH2-terminal kinase signaling pathways, has recently been reported as being an ARA55-interacting protein that can repress AR transactivation via inactivation of the ARA55 coregulator (78). PYK2 phosphorylates ARA55 at tyrosine 43 and impairs its coregulatory function. The disruption of the interaction between AR and ARA55 by PYK2 may also contribute to this inhibition of AR transactivation. PYK2 thus seems to have multiple ways to mediate ARA55-enhanced AR transcriptional activity. PYK2 is expressed in neurons, bone marrow, smooth muscles, and prostate (79, 80). Tissue staining showed that PYK2 was decreased with increased malignancy of prostate cancers (80), indicating a regulatory role of PYK2 on ARA55-enhanced AR-mediated growth of cancer cells. Taken together, ARA55 seems to have an important role in prostate cancer progression.ARA54, a RING finger protein, was initially identified as a coregulator of AR and progesterone receptor (49). In homomeric form, ARA54 interacts with AR and enhances its transcriptional activity in a ligand-enhanced manner. In the presence of 10 nm E2 or 1 μm HF, ARA54 can further enhance the transcriptional activity of a mutant AR (T877A) in LNCaP cells, but not wild-type AR or another mutant AR (E708K), suggesting that in addition to a particular AR mutation, ARA54 coregulator may contribute to the partial agonist activity of antiandrogens or other nonandrogenic molecules. ARA54 is functional as a dimer, and disruption of ARA54 dimerization by a dominant-negative mutant (dARA54) inhibits ARA54-enhanced AR transactivation, as well as AR-mediated growth of prostate cancer cells (81). The dARA54, although lacking AR interaction, still retains the ability to form a heteromer with ARA54 and thereby prevents ARA54 interaction with AR. Ectopic expression of this dARA54 mutant suppresses both androgen-mediated and antiandrogen-mediated AR activity and prostate-specific antigen (PSA) expression in prostate cancer cells, further strengthening the role of ARA54 in the development of partial agonist activity of antiandrogens in prostate cancer cells.Recently, ARA54 has been reported to possess ubiquitin ligase activity and ubiquitinate ARA54 itself, which may contribute to its proteosomal degradation (82). ARA54 can interact with E2-conjugating enzymes through its RING finger domain, and possibly can act as an ubiquitin ligase in the ubiquitination of nuclear proteins. The ubiquitination of cellular proteins plays important roles in multiple biological functions, including cell cycle regulation, apoptosis, transcription, and so forth (83). The major role of ubiquitination is to target substrate proteins for proteosomal degradation. Hence, it is somewhat surprising that ARA54, being a AR coactivator, still possesses ubiquitin ligase activity and may be involved in the degradation of substrate proteins. However, it is quite possible that ARA54 may be involved in the degradation of corepressor(s) involved in the regulation of AR transcriptional activity. Alternatively, targeted degradation of the preinitiation complex by the ubiquitin ligase activity of ARA54 after the initiation of transcription by AR may facilitate the reinitiation of transcription. However, additional studies need to be done to define its role as a ubiquitin ligase as well as AR coregulator with respect to AR transcriptional activity and AR-mediated growth of tumor cells.ARA160, an AR NH2-terminal-interacting protein (84), can function as a coregulator through interaction with the AR AF-1 domain and shares an identical sequence homology with TATA element modulatory factor (85). Both the far Western blotting and the immunoprecipitation assays demonstrate that ARA160 interacts directly with AR in a ligand-enhanced manner and enhances its transcriptional activity in prostate cancer cells. ARA160 can also interact (although to a significantly lesser extent) through the AR AF-2 domain to increase AR activity, suggesting that ARA160 may physically and functionally serve as a bridging factor between AR NH2- and COOH-terminal activation domain. Cotransfection with other AR COOH-terminal coregulators, such as ARA70 and ARA160, results in a cooperative enhancement of AR transactivation. A somatic mutation in the AR gene (E231G) from primary prostate tumor of TRAMP mice showed an increased AR activity in response to androgen but not E2 in the presence of ARA160, suggesting a role of this coactivator in determining the ligand specificity of mutant AR in prostate cancer (62).ARA267 is a unique AR-interacting protein that contains the SET domain (86), derived from three originally identified proteins named Su(var)3–9, Enhancer-of-zeste, and Trithorax (87, 88). Apart from this evolutionarily conserved motif, ARA267 also contains two LXXLL motifs, three nuclear translocation signal sequences, and four plant homeodomain fingers. In Drosophila, these proteins have been suggested to play important roles in homeotic gene expression (89). Human homologs of these genes, such as ALR, huASH, or ALL-1, have also been reported to play important roles in the regulation of transcriptional activity via direct modulation of the chromatin structure, which in turn may control cell growth or disease progression (90, 91).ARA267 shares 80% homology with NSD1, another SET domain protein (92), which may function as either a coactivator or a corepressor due to two distinct NR-interacting domains. ARA267 can interact with AR LBD and enhances AR transcriptional activity in a ligand-dependent manner. In LNCaP prostate cancer cells, transfection of ARA267 significantly increases expression of PSA, suggesting a role of ARA267 in prostate cancer progression (86). It also can very weakly enhance progesterone receptor activity in response to progesterone. When transfected in the presence of p300/CBP-associated factor, a histone acetyltransferase capable of enhancing AR activity, ARA267 showed an additive increase of AR transcriptional activity (86). These results indicate that ARA267 may function as a general coregulator for NR and modulate the recep-tor activity via cooperation with other chromatin remodeling factors.ARA24/Ran, another AR-interacting protein, physically interacts with the polyglutamine region of AR and enhances AR-dependent transcription of target genes (93). ARA24 is an AR NH2-terminal-associated coactivator that can interact differentially with varying lengths of polyglutamine within AR. It has an identical amino acid sequence as Ran, a Ras-like small nuclear GTPase (94). ARA24/Ran belongs to the superfamily of GTP-binding proteins that function as molecular switches for cycling between the GDP- and GTP-bound states. Recent studies indicated an 81% increase of ARA24/Ran in cancer tissue compared with benign tissue isolated from prostate cancer patients (95). Higher levels of ARA24/Ran in prostate tumor tissues clearly indicated a role for ARA24/Ran in the proliferation of prostate cancer cells. Similar levels of overexpression were also found in high-grade prostate intraepithelial neoplasia, indicating that aberrant expression of ARA24/Ran may be involved in prostate cancer initiation (95).Gelsolin is a multifunctional actin-binding protein well characterized as having implications in cell signaling, cell motility, apoptosis, and carcinogenesis (96). It regulates actin polymerization and depolymerization by sequestering actin monomers, and it can sever and cap actin filaments (97). Gelsolin has also been identified as being a substrate for caspase-3 and possesses a dual role in promoting apoptosis and protecting cells from apoptosis (98, 99). Several reports have indicated a differential expression of gelsolin in various cancers including prostate cancer (100, 101). Gelsolin expression is down-regulated in a number of malignancies including prostate, breast, lung, and bladder cancer (102, 103, 104), suggesting it may function as a tumor suppressor. Also, it has been proposed to be a sensitive marker for renal cystadenomas and carcinoma.Recently, in a yeast two-hybrid screen, gelsolin was identified as an AR-interacting protein that can interact with AR and enhance its transactivation in DU145 prostate cancer cells (105). A novel function of gelsolin as an antiandrogen potentiated AR coactivator has been documented. HF-potentiated AR coactivation and an increased expression of gelsolin after androgen ablation therapy (castration plus antiandrogen HF) suggested a possible role of gelsolin in the development of “flutamide withdrawal syndrome” (105). It is possible that increased expression of gelsolin could enhance AR activity under HF treatment to maintain AR-mediated growth and survival of tumor cells. The interruption of HF-induced interaction between AR and gelsolin could thus offer a potential target for therapeutic design in the treatment of prostate cancer.SV was initially identified from blood cells as an actin-binding protein and found to be expressed in muscle-enriched tissues such as skeletal muscles and in several cancer cell lines (106). SV has structural homology to gelsolin and villin and has also been identified as an AR-interacting protein (107, 108). The prostate cancer cells DU145 and PC-3 express a low level of endogenous SV, and cotransfection of SV with AR into these cells results in 2–3-fold enhancement of AR transactiva