| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Molecular and Cellular Biology, Baylor College of Medicine, Houston, Texas 77030
Address all correspondence and requests for reprints to: Dr. Jiemin Wong, Department of Molecular and Cellular Biology, Baylor College of Medicine, One Baylor Plaza, Houston, Texas 77030. E-mail: jwong{at}bcm.tmc.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Early studies indicate that in the absence of ligands, AR resides primarily in cytoplasm and is believed to associate with heat shock proteins in an inactive state (6, 7). Binding of ligand to AR is believed to trigger a series of events, including conformational change, translocation from the cytoplasm to the nucleus, and subsequent binding to specific promoter response elements, which eventually leads to activation or repression of its target genes (1, 4, 5). Like other NRs, research in the last several years has revealed an increasingly complexity of the mechanism of transcriptional regulation by AR (8). The actions of AR are subject to modulation, either positively or negatively, by an increasing number of coregulatory proteins, termed coactivators or corepressors (9). Coactivators are believed to function either as bridging factors between receptors and basal transcription machinery to enhance recruitment of the basal transcription machinery and/or as factors that have capacity to actively remodel repressive chromatin (8). While some coactivators such as ARA70 (10) or FHL2 (11) may be specific for AR, many of the coactivators identified so far, including steroid receptor coactivator (SRC) family coactivators, CREB-binding protein, p300, CBP/p300-associated factor, and TR-associated proteins/VDR-interacting proteins/activator-recruited cofactor complexes are generic to NRs (for review, see Refs. 8 and 9). Importantly, many coactivators possess intrinsic histone acetyltransferase activity (12). In contrast, corepressors such as silencing mediator of retinoid and thyroid hormone receptors (SMRT) and nuclear receptor corepressor (N-CoR) are found to associate with histone deacetylases in large protein complexes (13, 14, 15, 16). These findings provide a strong molecular connection between the modification of chromatin structure and transcriptional regulation by NRs. Indeed, a conceptual advance in our understanding of transcription control over the last several years is the recognition of chromatin structure as an integral component of transcriptional regulation in eukaryotic cells (17). In comparison to other NRs such as TR, GR, PR, and ER, little is known about how AR regulates transcription in the context of chromatin.
Uniquely among steroid hormone receptors, the hormone-dependent AF2 activity of AR is elusive. Deletion of the ligand binding domain generates an AR molecule with constitutive activity that in many transcription assays is equivalent to the activity of the full-length AR in the presence of ligands, whereas deletion of the N-terminal AF1 domain usually results in an AR molecule with low or no detectable activity even in the presence of ligands (6, 18). These observations suggest that AF1 contributes most, if not all, the activity of AR. Consistent with this idea, several studies indicate that the AF1 domain mediates primarily the interaction between the SRC family coactivators and liganded AR (19, 20, 21). Recent studies also indicate that a ligand-dependent intramolecular interaction between AF1 and AF2 domains is essential for AR transcriptional activity (22, 23). In addition, AR can be activated in the absence of androgens in different cell lines by growth factors such as IGF-I and epidermal growth factor or chemicals that directly activate the PKA signaling pathway (24, 25). The mechanism of such ligand-independent activation is not clear yet, but likely to involve phosphorylation of AR and/or its associated proteins.
Recent studies in prostate cancer provide evidence for the existence of a ligand-independent activity for AR. Androgens are known to play a crucial role in the occurrence and progression of prostate cancer. Patients with advanced prostate cancer are usually subjected to hormonal therapy by either androgen deprivation and/or blockade of AR with antiandrogens. These treatments are beneficial in the early stages of cancer but eventually lead to relapse of androgen-insensitive cancers (26). Paradoxically, many hormone-insensitive prostate cancers are found to be positive for both AR as well as the gene products that are regulated by AR (27, 28, 29), suggesting that AR may still remain functionally active and thus contribute to the progression of androgen-independent prostate cancer. While mutations in AR may lead to activation of AR in the absence of ligands or a change in its hormone specificity, recent studies indicate that mutations in AR are rare events in hormone-insensitive cancers. Instead, the amplification and consequent overexpression of the wild-type AR gene appears to be the most common event found in hormone-refractory prostate cancer (29). These observations have led to the hypothesis that overexpression of AR and its subsequent activation by growth factor-mediated cross-talk pathways could lead to the ligand-independent activation of AR in hormone-insensitive prostate cancer. However, it is not known whether overexpression of AR alone is able to activate transcription in the absence of cross-talk pathways.
An important question related to the issue of the hormone-independent activity is whether AR can bind to an androgen response element (ARE) in the absence of ligand. Although ligand is usually required for androgen-dependent transcription activation because AR is located primarily in cytoplasm in the absence of ligand, the fact that AR can be activated by other signaling pathways in the absence of ligand argues that AR has the capacity to bind DNA in a ligand-independent manner. So far, in vitro gel shift assays have yielded conflicting results on this subject. In some cases, in vitro translated AR or AR produced in insect cells is capable of binding to AREs in vitro in the absence of ligand (30, 31, 32), whereas in other cases pretreatment with ligand is required for DNA binding in vitro (33). The discrepancy over whether AR can bind DNA in the absence of ligands in vitro is at least partly due to the technical difficulty in producing sufficient amounts of recombinant unliganded AR proteins and further complicated by the fact that AR appears to have an intrinsic weak DNA binding activity.
Our previous work and that of others have established Xenopus oocytes as an excellent model system for studies of transcriptional regulation by NRs in the context of chromatin (34, 35). Xenopus oocytes contain a large storage of factors required for transcription and both histones and nonhistone proteins required for chromatin assembly. Xenopus oocytes are well suited for introduction of DNA, mRNA, or proteins through microinjection. Introduction of DNA into the nucleus of Xenopus oocytes through microinjection allows the assembly of injected DNA into chromatin through two different pathways depending upon the type of DNA injected. While microinjection of DNA templates either as single-stranded (ss) or double-stranded (ds) DNA into Xenopus oocyte nucleus leads to the assembly of both DNA templates into chromatin, the chromatin template resulted from injection of ssDNA is more refractory to basal transcription than that generated by dsDNA. This is because that the ssDNA injected into Xenopus oocyte nucleus is rapidly converted into dsDNA through the synthesis of the complementary strand. The resulting dsDNA is assembled into chromatin within 30 min after injection in a process coupled to the synthesis of the complementary strand (replication-coupled assembly pathway) (34, 36), which mimics the chromatin assembly process during S phase in cell cycle.
In this study, we have reconstituted a ligand-responsive AR transcriptional system using Xenopus oocytes in an effort to understand the molecular mechanisms of transcriptional regulation by AR in the context of chromatin. We demonstrate that, while R1881 strongly stimulated transcriptional activation by AR, a ligand-independent activity is also observed when AR is highly expressed. Expression of coactivators such as members of SRC family and p300 stimulates both ligand-independent and -dependent activation by AR. In vitro DNA binding assays indicate that ligand is not required for AR DNA binding activity. Furthermore, this hormone-independent activity is also observed in mammalian cells. Interestingly, addition of AR-antagonists such as casodex can inhibit this hormone-independent activity and this inhibitory effect appears to correlate with the recruitment of corepressor SMRT. Taken together, our results indicate that overexpression of AR can lead to activation of AR target genes in a ligand-independent manner and thus provide a possible molecular basis for the roles of AR gene amplification and consequent overexpression of AR in many hormone-refractory prostate cancers.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
, lane 2), whereas no endogenous AR proteins could be detected from the noninjected control oocyte extract (Fig. 1A, lane 1). Indeed, Western analysis using a different AR antibody (C-19; Santa Cruz Biotechnology, Inc.) also failed to detect the presence of AR protein in Xenopus oocytes (data not shown), indicating that Xenopus oocytes contain a very low level, if any, of endogenous AR proteins.
|
). Because the functional importance of chromatin structure in transcriptional regulation of the mouse mammary tumor virus-long terminal repeat (MMTV-LTR) by steroid hormone receptors has been well established (37), we also generated a MMTV-LTR-based reporter (Fig. 1B). To assemble reporter DNA into repressive chromatin through the replication-coupled chromatin assembly pathway (36), we injected both reporters in ssDNA form into the nucleus of Xenopus oocytes. After overnight incubation, the injected oocytes were collected and the chromatin structure was analyzed by micrococcal nuclease (MNase) digestion assay. As shown in Fig. 1C, limited MNase digestions revealed that injection of both reporters as ssDNA plasmids led to the assembly of the DNA into chromatin with regularly spaced nucleosomes. This result is consistent with the notion that injection of ssDNA plasmid will result in efficient assembly of chromatin through a replication-coupled assembly pathway.
We next examined whether expression of AR could activate transcription from repressive chromatin in Xenopus oocytes. Groups of Xenopus oocytes were injected with mRNA encoding AR (100 ng/µl, 18.4 nl/oocyte) and ssDNA of the MMTV reporter and treated with agonist R1881 or the antagonists casodex or flutamide at concentrations as indicated (Fig. 2). After overnight incubation, the total RNA was purified from each group of oocytes and the level of transcription from the MMTV promoter was analyzed by primer extension assay. A histone H4-specific primer, which detected the endogenous histone H4 mRNA and thus served as an internal loading control, was included in the primer extension reaction. As shown in Fig. 2, addition of R1881 at concentrations of 0.1 nM was sufficient to activate transcription from the MMTV promoter, whereas addition of casodex or flutamide in a concentration ranging from 1 nM to 100 nM failed to do so. Similar results were observed when the 4.ARE-TRßA reporter was used (data not shown). We thus conclude that AR expressed in Xenopus oocytes exhibits the expected hormone specificity and activates transcription from the MMTV LTR assembled into chromatin.
|
). Levels of transcription were assayed after overnight incubation in the presence or absence of R1881 (10 nM). Consistent with the result in Fig. 2, a R1881-dependent activation was observed from both the MMTV- and TRßA-based reporters when a low concentration of AR mRNA (100 ng/µl) was injected (Fig. 3A). However, an R1881-independent activation of transcription from both reporters (compare lanes 4 with 2 and lanes 9 with 7) was clearly detected when a high dose of AR mRNA was injected. Although addition of R1881 led to a stronger final levels of transcription (compare lanes 5 with 3 and lanes 10 with 8), the fold of R1881-dependent activation actually decreased due to the presence of R1881-independent activation. On the TRßA promoter, both R1881-independent and -dependent activation required the presence of AREs, because no activation was observed when the parental reporter without AREs was used as the reporter (compare lanes 12 and 13 with 11), indicating that both R1881-dependent and -independent activation were directly mediated by AR. Because ligand-independent activation for AR has only been reported in the cases of activation by cross-talk pathways and because the ligand-independent activity of AR has been implicated clinically in hormone-refractory prostate cancer, we focus here on the characterization of the molecular mechanism of this R1881-independent transcriptional activation by AR.
|
and incubated with or without addition of 10 nM of R1881. After overnight incubation, nuclear, cytoplasmic and total oocyte fractions were prepared. Due to the drastic difference in volume between the nucleus and cytoplasm of Xenopus oocytes (38), total proteins equivalent to three nuclei, half an oocyte of cytoplasmic and half an oocyte of the total oocyte extracts were fractionated by using a SDS-PAGE, and the distribution of AR was analyzed by Western blotting. As shown in Fig. 3B, in the absence of R1881, the majority of AR was found in the cytoplasm in both groups of oocytes injected with the low and high doses of AR mRNA (compare lanes 1 and 2). Treatment with R1881 led to a strong enrichment of AR in the nuclear fraction (compare lanes 5 and 4). This result indicates that AR expressed in Xenopus oocytes is primarily localized to the cytoplasm in the absence of R1881 and undergoes translocation to the nucleus in response to R1881 treatment. Thus, the pattern of subcellular localization of AR proteins in Xenopus oocytes is identical with that in mammalian cells.
Importantly, as shown in Fig. 3B, injection of the high dose of AR mRNA (1 µg/µl) clearly increased the level of AR protein in the nucleus in the absence of R1881 (compare lane 2 in the low and high). This result, together with the requirement of AREs for both R1881-dependent and -independent activation, suggests a model in which overexpression of AR leads to an increased level of AR protein in the nucleus, and this nuclear AR leads to subsequent activation of transcription even in the absence of R1881.
As the hormone-independent activation was only clearly observed when a high dose of AR mRNA was injected, we were concerned whether this is a phenomenon that exists only in the presence of vastly overexpression of AR. To have some sense about the levels of AR proteins, we compared the levels of AR proteins in Xenopus oocytes primed with low and high doses of AR mRNA with that that in an AR-positive prostate cancer cell line, LNCaP. When the same amount of the total proteins (5 µg) of LNCaP whole cell extract or AR mRNA primed oocyte extracts were analyzed for levels of AR by Western blotting (Fig. 3C), we found that level of AR in LNCaP cells was even higher than that in oocyte extract primed with the high dose of AR mRNA. This result indicates that a comparable level (concentration) of AR proteins can be found in prostate cancer cells such as LNCaP cells and thus suggests that the hormone-independent transcriptional activation by AR may have clinical relevance.
DNA Binding in Vitro by AR Protein Is Ligand Independent
The capacity of AR to activate transcription in the absence of hormone implies that AR can bind DNA in the absence of ligand. Because it is controversial as to whether ligand is required for DNA binding by AR, we analyzed the DNA binding activity of AR proteins expressed in Xenopus oocytes. We first carried out gel mobility shift assays using a 32P-labeled ARE-containing oligonucleotide probe and oocyte extracts prepared from oocytes injected with AR mRNA and treated with or without R1881 (10 nM). To maintain the association with R1881 of the AR derived from the R1881-treated AR-expressing oocytes, a final concentration of 10 nM of R1881 was added to all buffers used for binding assay or for making extracts derived from the R1881-treated oocytes. As shown in Fig. 4, a shifted DNA complex can be observed in lanes with both AR programmed extracts, with (lane 3) or without R1881 (lanes 8), but not in the lanes with control oocyte extract (lanes 2 and 7). In addition, this complex is ARE specific because the complex could be eliminated by addition of an excessive cold ARE competitor but not cold TRE competitor. Furthermore, in multiple experiments, we observed that the AR-DNA complex in the presence of R1881 appeared to migrate slightly slower than that in the absence of R1881 [compare lane 3 with lane 8 and use the nonspecific complex indicated by an asterisk (*) as a reference] suggest this difference in mobility may reflect the conformational changes of AR or/and the AR-DNA complex after binding of R1881.
|
, AR purified from both R1881 untreated or treated oocytes can bind to the ARE in a dose-dependent manner. No significant difference can be observed in terms of the binding (or protection) of the ARE sequence by both R1881 treated or untreated AR. Interestingly, the protection by R1881-treated AR appeared to extend more broadly than that by unliganded AR (compare lane 7 with 4). This difference may reflect the difference in conformations between liganded and unliganded AR and/or association of liganded AR with additional protein(s). Taken together, both gel mobility shift and DNase I footprinting assays demonstrate that AR binds to a consensus ARE in a ligand-independent manner, providing a crucial support for the observation of the ligand-independent activation.
|
, the peak of the unliganded AR behaved as protein complex of 600–700 kDa (peak around fractions 22–24). Clearly, the peak of the AR from R1881 treated oocyte extracts shifted toward the right and thus appeared to be smaller (500 kDa) (peaks around fractions 26–28) (Fig. 6B). This result is consistent with the idea that binding of ligand induces the conformational change and/or dissociation of AR associated heat shock proteins or other protein(s). Nevertheless, when DNA binding activity was assayed by gel mobility shift across the fractions, both R1881 untreated and treated AR bound to the ARE as already demonstrated in Figs. 5 and 6. It is noteworthy that the DNA binding activity across the fractions from both R1881-untreated and -treated oocyte extracts correlated directly with the presence of AR, but not the sizes of the AR complex. This result is important because it rules out the possibility that the DNA binding activity from R1881 untreated AR containing extract is derived from a subfraction of AR proteins that may not be integrated into the protein complex(es) and thus presumably not associate with heat shock proteins. If dissociation of heat shock proteins is required for AR DNA binding, one would expect that the smaller AR complex migrating toward the right may be free of heat shock proteins and thus exhibit a better DNA binding capacity. However, our effort to check directly the presence of heat shock protein in AR complex(es) was hindered by the lack of antibody in our hands that can recognize heat shock proteins in Xenopus oocytes.
|
where the MMTV reporter was used, coexpression of SRC-1 and RAC3 with AR led to a significant enhancement of R1881-independent activation (from 4-fold to 16- and 17-fold, respectively). Under the same conditions, SRC-1 and RAC3 only moderately stimulated the transcription in the presence of R1881 (from 23-fold to 41- and 31-fold, respectively). The stimulation of R1881-independent activity by coactivators was not restricted to the MMTV reporter, as expression of p300 also stimulated the R1881-independent activation from the 4.ARE.TRßA-based reporter from 7- to 22-fold (Fig. 7B). As a control, expression of p300 alone (Fig. 7B, compare lane 2 with 1) or SRC-1 and RAC3 alone (data not shown) in the absence of AR did not stimulate transcription, indicating that the stimulation of transcription by those coactivators is mediated through AR. Thus, much like the hormone-dependent activation, the hormone-independent activation can be enhanced by the action of coactivators such as SRC1, RAC3, and p300.
|
, although R1881-independent activation by AR was not detectable when a low level of AR plasmid (5 ng) was used, the R1881-independent activation was clearly observed when a higher dose of AR expression plasmid (50 ng) was used. In both cases, addition of R1881 further stimulated the luciferase activity 4- and 3-fold, respectively. Thus, the ligand-independent activity by AR is not unique to Xenopus oocytes but likely an inherent feature of AR.
|
). After overnight incubation, the levels of transcription were again determined by primer extension analysis. As expected, expression of AR led to a hormone-independent activation (Fig. 9A, compare lane 2 with 1). While addition of R1881 led to a further robust activation (Fig. 9A, compare lane 3 with 2), addition of either antagonists clearly inhibited the hormone-independent activation by AR (compare lanes 4, 5 and 6, 7 with 2).
|
, addition of antagonists casodex and flutamide indeed inhibited the hormone-independent activation by AR under the similar conditions as described in Fig. 8.
The Inhibition by Antagonist Correlates with the Recruitment of Corepressor SMRT
In an attempt to understand the mechanisms by which the antagonists inhibited the hormone-independent activity of AR, we analyzed whether casodex could influence the interaction of AR with coactivators and corepressors. We cotransfected a SRC-1 expression construct together with the AR expression construct into COS-1 cells. The transfected cells were then treated with or without R1881 or casodex as indicated (Fig. 9C). We then performed immunoprecipitation (IP) experiments using a FLAG-specific antibody (AR with a FLAG tag) and examined the co-IP of SRC-1. As shown in Fig. 9C, SRC-1 was co-IP with AR in the absence of hormone treatment (lane 6). Interestingly, addition of R1881 did not appear to have a significant effect the association of SRC-1 with AR (compare lane 7 with 6), whereas addition of casodex led to a slight reduction of the interaction of SRC-1 with AR (compare lane 8 with 6). Because several recent studies indicate that antagonists for estrogen receptors and progesterone receptors also have capacity to modulate interaction of corepressors SMRT and N-CoR with receptors (40, 41), we also tested whether casodex could induce interaction of AR proteins with corepressors such as SMRT and N-CoR. Western blotting using a SMRT-specific antibody revealed that addition of casodex resulted in co-IP of SMRT with AR (compare lane 6 with 4 at the right panel). Similar attempt using a N-CoR-specific antibody failed to detect N-CoR proteins in COS-1 cell extract, presumably because the level of N-CoR in COS-1 cells is low. Thus, the inhibitory effect of casodex appears to correlate with its ability to reduce the association of coactivator with AR as well as to enhance the recruitment of corepressor SMRT.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and 8); 2) AR can bind to a consensus ARE in vitro in a hormone-independent manner (Figs. 4 and 5); 3) coactivators such as SRC-1, RAC3, and p300 stimulate both ligand-independent and -dependent activation by AR (Fig. 7) and 4) antagonists such as casodex can inhibit hormone-independent activation by AR and this inhibition appears to correlate with its ability to influence the association of AR with both coactivator and corepressor (Fig. 9). While ligand-independent activation of AR by growth factors or other signaling pathways has been reported (24, 25), it is not clear whether AR itself has an intrinsic ligand-independent activity. By manipulating the levels of AR expression in oocytes through injection of different amounts of AR mRNAs, we demonstrate that high level expression of AR activated both the MMTV and TRßA-based reporters in the absence of R1881. Several lines of evidence support the conclusion that this hormone-independent activity is intrinsic to AR but not a unique feature of Xenopus oocytes. First, consistent with the observation from mammalian cells that AR proteins reside primarily in the cytoplasm, AR expressed in Xenopus oocytes also resides primarily in the cytoplasm in the absence of R1881. Second, AR expressed in Xenopus oocytes responds to agonist R1881 the same way as AR expressed in mammalian cells. These include the translocation from the cytoplasm to the nucleus and the robust trans-activation of both reporters by AR in the presence of R1881. Third, high level expression can lead to the increase of nuclear AR. This is not surprising because the subcellular localization of AR is dynamic and likely to be influenced by its concentration. We believe that this unliganded AR in the nucleus is responsible for the observed ligand-independent transcription. Fourth, the Xenopus oocytes used here were not treated with growth factors or reagents that could activate PKA pathways. In other words, the R1881-independent activity of AR that we observed is unlikely a result of the activation of AR by cross-talk pathways. Nevertheless, we also could not rule out the remote possibility that a subpopulation of AR in Xenopus oocytes could be activated by other cross-talk signaling pathways or by mysterious ligand(s) in the oocytes. Fifth, overexpression of AR in COS-1 cells also leads to a R1881-independent trans-activation, indicating that the hormone-independent activation is not unique to Xenopus oocytes. Finally, as shown in several recent publications (19, 20, 21), coactivators such as members of SRC family interact with AR primarily through the AF1 but not the AF2 domain in AR. Consistent with those observations, we show that expression of SRC1, RAC3 and p300 in Xenopus oocytes further enhanced the ligand-independent activation by AR. Taken together, we propose that this hormone-independent transcriptional activity is intrinsic to AR and may be mediated through the hormone-independent interaction of AR with coactivators such as members of SRC family and p300.
Our demonstration that AR expressed in Xenopus oocytes exhibits ligand-independent DNA binding provides strong support for the idea that AR has the capacity to activate transcription in a ligand-independent manner. By both gel mobility and DNase I footprinting assays, we demonstrated that both unliganded AR and liganded AR bind to a consensus ARE. Furthermore, gel filtration analysis revealed that unliganded AR exists in a large protein complex(es) and that R1881 treatment causes AR to migrate as a smaller complex (Fig. 6). These results are consistent with the idea that in the absence of hormone AR is associated with other proteins including heat shock proteins and that binding of hormone results in the change of conformation and/or release of heat shock proteins. Nevertheless, gel mobility shift analysis of the gel filtration fractions derived from the R1881-untreated and -treated AR extracts indicates that DNA binding activity correlates with the presence of AR, not the size of the AR complex (Fig. 6). Taken together, these results provide strong evidence that AR can bind to an ARE in a ligand-independent manner. Ligand-independent DNA binding by AR has been reported before by using either in vitro translated AR proteins (31) or AR proteins expressed in insect SF9 cells (30). However, in many other cases treatment with ligand appears to be required for preparation of AR proteins with active DNA binding activity (33). This discrepancy could, at least in part, be explained by the technical difficulty in preparation of unliganded recombinant AR. AR expressed in SF9 cells is by and large insoluble in the absence of R1881 (30). On the other hand, R1881 treatment has been shown to induce AR expression due to the presence of AREs in the AR coding region and to stabilize AR proteins (42, 43). These two factors facilitate preparation of and the DNA binding assay for the liganded AR. Thus, the hormone-independent DNA binding activity is unlikely to be unique to the AR proteins expressed in Xenopus oocytes and may be an intrinsic feature of AR.
Our results that AR exhibits hormone-independent DNA binding and transcriptional activity also have strong implications for our understanding of the possible roles of AR in hormone-refractory prostate cancer. Strong evidence suggests that AR may remain functionally active and thus contribute to the progression of androgen-independent prostate cancer (44). Many androgen-independent prostate cancers are found to express both AR and its regulated genes (27, 28, 29). However, how AR remains transcriptionally active in androgen-independent prostate cancer is largely unknown. Many hypotheses, including mutations in AR, AR gene amplification, and protein overexpression; changes in coregulators; and activation of AR by cross-talk signaling pathways have been proposed. While mutations in AR may enhance activity of AR in the absence of ligand or a change in its hormone specificity, recent studies indicate that the frequency of AR mutations is low even in hormone-insensitive cancers (45, 46). Instead, the amplification and consequent overexpression of the wild-type AR gene appears to be the most common event found in hormone-refractory prostate cancers (45). These results suggest that overexpression of AR proteins is a potential mechanism that leads to the ligand-independent activity of AR in hormone-insensitive prostate cancer. Our results that overexpression of AR in Xenopus oocytes can result in R1881-independent activation of both TRßA promoter and MMTV LTR assembled into repressive chromatin provides support for this idea. Furthermore, comparison of AR in Xenopus oocytes injected with a high dose of AR mRNA with that in LNCaP cells indicates that a level of AR protein sufficient for observation of hormone-independent activity in Xenopus oocytes could be present in prostate cancer cells (Fig. 3C). In addition, we show that expression of coactivators such as members of the SRC family and p300 can further enhance the hormone-independent trans-activation by AR (Fig. 7). This result is consistent with the previous observation that AR could interact with and thus sequester SRC-1 protein even in the absence of hormone in mammalian cells (47). This result is also consistent with the recent reports that the AF1 but not the ligand-dependent AF2 domain of AR is primarily responsible for the interaction and recruitment of the SRC-1 family coactivators by AR (19, 20, 21). Given the ability to stimulate hormone-independent activation by AR, it is tempting to speculate that changes in levels of coactivators could be a potential contributing factor for hormone-independent activation of AR in prostate cancer. One can envisage a scenario in prostate cancers in which gene amplification and overexpression of AR could result in hormone-independent activation of AR-regulated genes. The levels of this hormone-independent activation are likely to be further augmented by any increase in levels of coactivators. In addition, this hormone-independent activation could be further enhanced by cross-talk pathways mediated by growth factors (24, 25).
The finding that antagonists such as casodex can induce interaction of AR with corepressor SMRT is not surprising. It has been reported that antagonists for ERs and PRs can modulate interaction of corepressors SMRT and N-CoR with ER and PR (40, 41). Together, these findings indicate that, in addition to competing with agonists for binding of receptors, antagonists have capacity to actively repress the receptor activity by promoting their interaction with corepressor complexes.
In conclusion, the data presented here provide evidence that the interaction of AR with specific DNA sequences does not require ligand and that AR has the capacity to activate transcription in a ligand-independent manner when AR is overexpressed. This ligand-independent activity can be further enhanced by coactivators including the members of the SRC family and p300. It is of great interest to test in future whether this hormone-independent transcriptional activity is relevant to the occurrence and progression of androgen-independent prostate cancer.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
In Vitro mRNA Preparation and Microinjection of Xenopus Oocytes
To prepare AR mRNA in vitro, the pSP64poly(A)-AR was first digested with BglII. The synthesis of AR mRNA was carried out by using the linearized DNA template and a SP6 Message Machine kit (Ambion, Inc., Austin, TX) as described by the manufacturer. The in vitro synthesis of mRNAs encoding SRC-1, RAC3, and p300 was as described previously (39). A typical reaction with approximately 1 µg of linearized template in a 20 µl reaction yielded 10
15 µg of capped mRNA. All mRNAs were resuspended in ribonuclease-free water at a final concentration of 1 µg/µl. The preparation of stage VI Xenopus oocytes and microinjection were essentially as described (34). For transcriptional analysis, single-stranded reporter DNA was injected (50 ng/µl, 18.4 nl/oocyte) into the nuclei of the oocytes, whereas the indicated amount of mRNAs encoding AR or coactivators (100 ng/µl, 18.4 nl/oocyte) was injected into the cytoplasm of the oocytes. Injection of mRNAs was usually performed 2–3 h before the injection of ssDNA to allow protein synthesis. Usually a group of approximately 20 oocytes was injected for each sample to minimize variations among oocytes and injections. The injected oocytes were incubated at 18 C overnight in modified Barth’s solution (36) supplemented with antibiotics (50 U/ml penicillin/streptomycin) in the presence or absence of 10 nM R1881 or the antagonists casodex and flutamide at concentrations indicated. The oocytes were then collected for transcription analyses or other assays as described below.
MNase Assay of Chromatin Structure
The MNase assay of chromatin assembly was performed as described previously (34).
Expression and Subcellular Localization of AR in Oocytes
To examine the expression and localization of AR in the oocytes, the cytoplasm and nucleus of the injected oocytes treated with or without R1881 (10 nM) were dissected manually. The protein extracts from cytoplasm, nucleus, and the whole oocytes were then resolved by SDS-PAGE followed by immunoblotting using an antibody against the FLAG-tag (1:5000 dilution). Signals were detected with a chemiluminescence kit (Pierce Chemical Co., Rockford, IL) as described by the manufacturer.
Transcription Analysis
Transcription analysis by primer extension was performed essentially as described (34). The primer I was used for detection of transcripts from the pTRßA and p4.ARE.TRßA reporters and CAT primer was used for detection of transcripts from the MMTV construct (34). The internal control was the primer extension product of the endogenous histone H4 mRNA using a H4-specific primer as described (49). In the figures where levels of transcription were presented, the levels of transcription were quantified by using phosphorimage analysis and were the average results of at least two independent experiments.
Gel Mobility Shift Assay
To examine the DNA binding activity of AR proteins expressed in Xenopus oocytes, groups of oocytes were injected with AR mRNA (1 µg/µl) and treated with or without R1881 (10 nM) overnight. The oocytes were then collected, rinsed once and homogenized in the extraction buffer (10 µl/oocytes) [20 mM HEPES (pH 7.9), 75 mM KCl, 1 mM dithiothreitol (DTT), 0.5 mM EDTA, 0.1% NP40, 10% glycerol, 0.1 mM phenylmethylsulfonyl fluoride]. To maintain association of AR with R1881, a final concentration of 10 nM of R1881 was included in the extraction buffer for making extracts derived from R1881 treated oocytes. The clean extracts were obtained after centrifugation of crude extracts at 13,000 rpm for 20 min at 4 C to remove yolk proteins and lipids and used for gel shift assay. In brief, the oocyte extracts (1–2 µl) were preincubated with the binding buffer [HEPES (pH 7.5), 100 mM KCl, 5 mM MgCl2, 10% glycerol, 2 mM DTT, 0.1 mM EDTA, and 0.25 µg of polydeoxyinosine-deoxycytidine] and with or without 10 nM of R1881 in a final volume of 14 µl for 15 min on ice. The end-labeled oligonucleotide probe containing a consensus ARE (0.1 ng) was added to each binding reaction and the mixture was incubated for 20 min at room temperature. In competition assays, unlabeled ARE or TRE (10 ng) was added into the reaction and incubated on ice with oocyte extracts for 15 min before the addition of the probe mixture. DNA-protein complexes were resolved on 5% polyacrylamide gels (80:1 of polyacrylamide/bisacrylamide) containing 0.5x TBE and revealed by autoradiography.
Gel Filtration Analysis of AR Complexes
A Superose 6 column (Amersham Pharmacia Biotech, Piscataway, NJ) was preequilibrated with the gel filtration buffer (20 mM HEPES, pH 7.8; 150 mM KCl; 1 mM DTT; 0.2 mM phenylmethylsulfonyl fluoride) at a flow-rate of 0.3 ml/min. Clean oocyte extracts (200 µl) prepared from AR mRNA injected Xenopus oocytes treated with or without R1881 (10 nM) were fractionated at a flow-rate of 0.3 ml/min. Samples (15 µl) from every other fraction (450 µl) were analyzed either by gel mobility shift for AR-DNA binding activity or by Western blotting for the presence of AR.
DNase I Footprinting
The DNase I footprinting assay was performed essentially as described (34) with after modifications. An end-labeled DNA fragment containing a consensus ARE was prepared by PCR, purified by PAGE and used for footprinting. The liganded AR and unliganded AR proteins used for footprinting assays were first partially affinity purified using the FLAG-tag specific M2 agarose resins (Sigma, St. Louis, MO) to reduce the nonspecific binding activity from oocyte extracts.
Cell Culture, Transient Transfection, Coimmunoprecipitation, and Western Blotting
LNCaP cells were culture in Roswell Park Memorial Institute 1640 medium (Invitrogen), which was supplemented with 5% FBS and glutamine. The whole cell extract of LNCaP cells was prepared by using the lysis buffer (10 mM Tris-HCl, pH 8.0; 1 mM EDTA; 150 mM NaCl; and 0.5% NP40) followed by a centrifugation (14,000 rpm, 20 min at 4 C). COS-1 cells were cultured in DMEM with addition of 10% FBS. For luciferase assay, 12 x 104 COS-1 cells were plated in six-well plates in phenol red-free medium supplemented with 10% dextran charcoal-stripped FCS 24 h before transfection. Transient transfection was performed according to the protocol of the LipofectAMINE-plus kit (Life Technologies, Inc., Gaithersburg, MD), with addition of 100 ng of reporter MMTV-luc and the indicated amount of AR expression plasmid pCR3.1-AR for each well. After incubation for 16 h, the cells were washed and supplemented with fresh medium containing 10 nM R1881 or antagonists as indicated. After a further 24-h incubation, the cells were washed with cold PBS and lysed with the lysis buffer described above. The extracts were analyzed for luciferase activity according to a manufacturer’s instruction (Promega Corp. luciferase assay kit) and the relative luciferase activity was normalized to the protein concentration. The results were the averages from at least three independent experiments. For coimmunoprecipitation experiments, expression constructs for FLAG-tagged AR and SRC-1 were cotransfected into COS-1 cells and treated with R1881 or antagonists as described above. The whole cell extracts were prepared and used for immunoprecipitation of AR using the FLAG-tag-specific antibody (M2, Sigma). The presence of SRC-1 or SMRT was detected by Western blotting using a SRC-1-specific antibody (39) and an SMRT-specific antibody (raised against amino acid 1165–1363 of human SMRT) (16). The AR antibody (N-20) for Western shown in Fig. 1A was purchased from Santa Cruz Biotechnology, Inc.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Abbreviations: AF, Activation function; ARE, androgen response element; CAT, chloramphenicol acetyltransferase; DNase I, deoxyribonuclease I; ds, double-stranded; DTT, dithiothreitol; IP, immunoprecipitation; LTR, long terminal repeat; MMTV, mouse mammary tumor virus; MNase, micrococcal nuclease; N-CoR, nuclear receptor corepressor; NR, nuclear receptor; RAC3, receptor-associated coactivator-3; SMRT, silencing mediator of retinoid and thyroid hormone receptors; SRC-1, steroid receptor coactivator-1; ss, single-stranded; TRßA, Xenopus thyroid hormone receptor ßA; TRE, thyroid hormone response element.
Received for publication August 29, 2001. Accepted for publication January 10, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. Fu, J. Jiang, J. Li, S. Wang, G. Shi, Q. Feng, E. White, J. Qin, and J. Wong Deleted in Breast Cancer 1, a Novel Androgen Receptor (AR) Coactivator That Promotes AR DNA-binding Activity J. Biol. Chem., March 13, 2009; 284(11): 6832 - 6840. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z.-Y. Wu, K. Chen, B. Haendler, T. V. McDonald, and J.-S. Bian Stimulation of N-Terminal Truncated Isoform of Androgen Receptor Stabilizes Human Ether-a-go-go-Related Gene-Encoded Potassium Channel Protein via Activation of Extracellular Signal Regulated Kinase 1/2 Endocrinology, October 1, 2008; 149(10): 5061 - 5069. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. Dong, J. Sweet, J. R. G. Challis, T. Brown, and S. J. Lye Transcriptional Activity of Androgen Receptor Is Modulated by Two RNA Splicing Factors, PSF and p54nrb Mol. Cell. Biol., July 1, 2007; 27(13): 4863 - 4875. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Yong, Z. Yang, S. Periyasamy, H. Chen, S. Yucel, W. Li, L. Y. Lin, I. M. Wolf, M. J. Cohn, L. S. Baskin, et al. Essential Role for Co-chaperone Fkbp52 but Not Fkbp51 in Androgen Receptor-mediated Signaling and Physiology J. Biol. Chem., February 16, 2007; 282(7): 5026 - 5036. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Li, J. Fu, C. Toumazou, H.-G. Yoon, and J. Wong A Role of the Amino-Terminal (N) and Carboxyl-Terminal (C) Interaction in Binding of Androgen Receptor to Chromatin Mol. Endocrinol., April 1, 2006; 20(4): 776 - 785. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Y. Khan, G. Fu, A. Ismail, S. Srinivasan, X. Cao, Y. Tu, S. Lu, and Z. Nawaz Multifunction Steroid Receptor Coactivator, E6-Associated Protein, Is Involved in Development of the Prostate Gland Mol. Endocrinol., March 1, 2006; 20(3): 544 - 559. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. J. Burd, C. E. Petre, L. M. Morey, Y. Wang, M. P. Revelo, C. A. Haiman, S. Lu, C. M. Fenoglio-Preiser, J. Li, E. S. Knudsen, et al. Cyclin D1b variant influences prostate cancer growth through aberrant androgen receptor regulation PNAS, February 14, 2006; 103(7): 2190 - 2195. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Vlahopoulos, W. E. Zimmer, G. Jenster, N. S. Belaguli, S. P. Balk, A. O. Brinkmann, R. B. Lanz, V. C. Zoumpourlis, and R. J. Schwartz Recruitment of the Androgen Receptor via Serum Response Factor Facilitates Expression of a Myogenic Gene J. Biol. Chem., March 4, 2005; 280(9): 7786 - 7792. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. M. Jacobsen, S. A. Schittone, J. K. Richer, and K. B. Horwitz Progesterone-Independent Effects of Human Progesterone Receptors (PRs) in Estrogen Receptor-Positive Breast Cancer: PR Isoform-Specific Gene Regulation and Tumor Biology Mol. Endocrinol., March 1, 2005; 19(3): 574 - 587. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Q. Wang, J. A. Blackford Jr., L.-N. Song, Y. Huang, S. Cho, and S. S. Simons Jr. Equilibrium Interactions of Corepressors and Coactivators with Agonist and Antagonist Complexes of Glucocorticoid Receptors Mol. Endocrinol., June 1, 2004; 18(6): 1376 - 1395. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Kim, L. Jia, W. D. Tilley, and G. A. Coetzee Dynamic methylation of histone H3 at lysine 4 in transcriptional regulation by the androgen receptor Nucleic Acids Res., December 1, 2003; 31(23): 6741 - 6747. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Hara, K. Nakamura, H. Araki, M. Kusaka, and M. Yamaoka Enhanced Androgen Receptor Signaling Correlates with the Androgen-refractory Growth in a Newly Established MDA PCa 2b-hr Human Prostate Cancer Cell Subline Cancer Res., September 1, 2003; 63(17): 5622 - 5628. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. V. Litvinov, A. M. De Marzo, and J. T. Isaacs Is the Achilles' Heel for Prostate Cancer Therapy a Gain of Function in Androgen Receptor Signaling? J. Clin. Endocrinol. Metab., July 1, 2003; 88(7): 2972 - 2982. [Full Text] [PDF]
|
|
|
|
|
|
L. B. Lutz, M. Jamnongjit, W.-H. Yang, D. Jahani, A. Gill, and S. R. Hammes Selective Modulation of Genomic and Nongenomic Androgen Responses by Androgen Receptor Ligands Mol. Endocrinol., June 1, 2003; 17(6): 1106 - 1116. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. J. Lee, J. McClung, G. A. Hand, and J. A. Carson Overload-induced androgen receptor expression in the aged rat hindlimb receiving nandrolone decanoate J Appl Physiol, March 1, 2003; 94(3): 1153 - 1161. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
