Abstract
Androgen antagonists or androgen deprivation is a primary therapeutic modality for the treatment of prostate cancer. Invariably, however, the disease becomes progressive and unresponsive to androgen ablation therapy (hormone refractory). The molecular mechanisms by which the androgen antagonists inhibit prostate cancer proliferation are not fully defined. In this report, we demonstrate that sirtuin 1 (SIRT1), a nicotinamide adenosine dinucleotide-dependent histone deacetylase (HDAC) linked to the regulation of longevity, is required for androgen antagonist-mediated transcriptional repression and growth suppression. Androgen antagonist-bound androgen receptor (AR) recruits SIRT1 and nuclear receptor corepressor to AR-responsive promoters and deacetylates histone H3 locally at the prostate-specific antigen promoter. Furthermore, SIRT1 down-regulation by small interfering RNA or by pharmacological means increased the sensitivity of androgen-responsive genes to androgen stimulation, enhanced the sensitivity of prostate cancer cell proliferative responses to androgens, and decreased the sensitivity of prostate cancer cells to androgen antagonists. In this study, we demonstrate the ligand-dependent recruitment of a class III HDAC into a corepressor transcriptional complex and a necessary functional role for a class III HDAC as a transcriptional corepressor in AR antagonist-induced transcriptional repression. Collectively, these findings identify SIRT1 as a corepressor of AR and elucidate a new molecular pathway relevant to prostate cancer growth and approaches to therapy.
PROSTATE CANCER IS the second leading cause of cancer death in men (1–3). The androgen receptor (AR), a hormone-dependent transcription factor, plays a major role in promoting the development and progression of prostate cancer (4–7). Androgen ablation and blockade of androgen actions through the AR remain the mainstay of treatment for advanced prostate cancer (8, 9). Whereas initial responses to androgen deprivation are the norm, most tumors eventually recur in what is termed an androgen-independent (refractory) state (10–12).
The transcriptional activity of the AR is modulated by nuclear coregulatory proteins, known as coactivators and corepressors (13–15). During activation by ligands, such as dihydrotestosterone (DHT), the AR translocates to the nucleus, whereupon it binds to androgen-response elements (AREs) on target genes and regulates their transcription (16–18). The balance of corepressors and coactivators in the AR complex determines AR transcriptional activity (19, 20). Binding of androgens induces recruitment of coactivators such as SRC-1 (steroid receptor coactivator-1), transcriptional intermediary factor 2 (Tif2)/glutamate receptor-interacting protein 1-1 (GRIP1-1), and ACTR (activator of thyroid and retinoic acid receptor)/AIB1 (amplified in breast cancer 1)/RAC3 (receptor-associated coactivator 3)/pCIP (p300/CBP interacting protein), as well as p300, CBP [CREB (cAMP response element binding protein) binding protein], and pCAF (p300/CBP-associated factor), which contain intrinsic histone acetylase activity (21–30). p160 appears to mediate the binding of the AR to the histone acetylase complex (16, 17). In contrast, binding of AR antagonists induces AR to form a complex with corepressors, such as nuclear receptor corepressor (NCoR), silencing mediator of retinoic acid and thyroid hormone receptor (SMRT), and histone deacetylase 1 (HDAC-1), HDAC2, and HDAC3 (24, 25, 31–34). Although many coactivators of the AR have been identified and well studied, the currently known AR corepressors are fewer and less well characterized, and, in some cases, their necessity in transcriptional repression has not yet been established (13, 15). Identification of new corepressors and developing an understanding of the precise mechanisms underlying the regulation of AR function is of critical importance for the design and development of novel therapies and pharmaceutical targets for treating prostate cancer.
Sirtuin 1 (SIRT1) is a mammalian nicotinamide adenosine dinucleotide (NAD)-dependent deacetylase belonging to the class III HDAC family (35). Recent studies have demonstrated that SIRT1 plays a role in a wide variety of processes, including stress responses (36), metabolism (37), apoptosis (38), embryogenesis (39), and calorie restriction and aging (40, 41). SIRT1 binds to, and regulates the activity of, several transcription factors, including p53 (42–44), FOXO1 (forkhead box O1), FOXO3a, and FOXO4 (45–47), PPARγ (peroxisome proliferator-activated receptor γ) (48), HES-1 (hairy and enhancer of split 1) and HEY-2 (hairy/enhancer-of-split related with YRPW motif 2) (49), MyoD (myogenic differentiation) (50), CTIP2 [chicken ovalbumin upstream promoter transcription factor (COUP-TF)-interacting protein] (51), NF-κB (nuclear factor κB) (52), and PGC1α (peroxisome proliferator-activated receptor γ coactivator 1) (53).
In the present study, we establish SIRT1 as a specific corepressor of the AR. We find that androgen antagonists induce recruitment of SIRT1 to AR-responsive promoters and that AR-dependent transcriptional suppression by androgen antagonists requires SIRT1. We demonstrate that SIRT1 suppresses AR-dependent gene transcription through its deacetylase activity and alters local histone H3 acetylation. Furthermore, we find that SIRT1 is required for androgen antagonist-mediated growth suppression and demonstrate that down-regulation or suppression of SIRT1 activity increases the sensitivity of prostate cancer cells to the transcriptional and proliferative activities of androgens.
Results
SIRT1 Suppresses AR-Dependent Gene Transcription
To determine whether the SIRT1 could regulate androgen-dependent and -independent transcription, we assessed the effect of altering SIRT1 activity on the regulation of AR-mediated transcription of two androgen-responsive promoter-reporter vectors, a human prostate-specific antigen (PSA) promoter-driven reporter (PSA-LUC) and a mouse mammary tumor virus (MMTV) long terminal repeat-driven reporter (MMTV-LUC), in a human prostate cancer (LNCaP) cell line. First, cell-based HDAC activity assays were performed to confirm that the SIRT1 agonist and SIRT1 inhibitor were affecting HDAC enzymatic activity in vivo, at the concentrations used. These studies demonstrated that exposure to resveratrol increased NAD-dependent HDAC activity greater than 2.5-fold (P < 0.05), whereas nicotinamide (NAM) exposure diminished enzymatic activity by 3.3-fold (P < 0.05) (Fig. 1A). Second, we analyzed the effect of these SIRT1 modulators on AR-dependent and -independent transcription. DHT-induced luciferase (LUC) activity of PSA-LUC was suppressed 16-fold (P < 0.01) by treatment of the transfected cells with resveratrol, an agonist of SIRT1 (Fig. 1B). Conversely, PSA-LUC activity was induced 3.5-fold (P < 0.01) by exposure to the SIRT1 inhibitor NAM and 3-fold by Sirtinol, a structurally unrelated SIRT1 inhibitor (data not shown). Similar effects were seen using pMMTV-LUC as the reporter (data not shown). In the absence of DHT, NAM or resveratrol treatment did not result in significant change in the PSA-LUC activity. Immunoblot assays demonstrated that neither resveratrol nor NAM treatment changed AR protein levels (Fig. 1B). Collectively, these results demonstrated that pharmacological agents capable of regulating SIRT1 deacetylase activity modulated AR-mediated PSA and pMMTV promoter-driven gene activity.
SIRT1 Represses AR-Dependent Gene Transcription A, Cell-based NAD-dependent HDAC activity assay of cells treated with 50 μm resveratrol (RES) or 10 mm NAM for 24 h. B, Top, The effects of resveratrol and NAM on PSA-LUC transcription. LNCaP cells cultured in media containing charcoal-stripped serum were transfected with the PSA-LUC reporter vector plasmid. Transfected cells were then exposed to resveratrol (RES) at 50 μm or NAM at 10 mm for 24 h and treated with 10 nm DHT or vehicle treated (Veh) for 24 h before assay of LUC activity, expressed here in arbitrary units. Inset shows the results from vehicle treatment, with an expanded y-axis. Bottom, Immunoblot analysis of AR protein levels in cells treated with NAM or resveratrol. C, Top, Effects of SIRT1 overexpression or SIRT1 knockdown on PSA-LUC transcription. LNCaP cells were cotransfected with PSA-LUC and an empty control vector (pcDNA3.1), a wt-SIRT1 expression vector (pcDNA3.1-SIRT1), a vector expressing SIRT1 siRNA (pSUPER-SIRT1), or the empty siRNA expression vector (pSUPER) and then treated with 10 nm DHT or vehicle treated for 24 h, and cells were harvested for LUC assay. Inset shows the results from vehicle treatment, with an expanded y-axis. Bottom, Immunoblot analysis of SIRT1 protein levels in cells transfected with wt-SIRT (pCDNA3.1-SIRT1), dominant-negative-SIRT1 (pCDNA3.1-H363Y), empty vector (pCDNA3.1), empty siRNA vector (pSUPER), or SIRT1 siRNA expression vector (pSUPER-SIRT1). Cell extracts were normalized for protein content, separated by PAGE, transferred, probed with an anti-SIRT1 antibody, anti-AR antibody, or an anti-β-actin antibody, and developed with a chemiluminescence kit. D, The effect of resveratrol or NAM on endogenous AR-dependent and -independent PSA gene transcription. LNCaP cells cultured in media containing charcoal-stripped serum were exposed to 10 mm NAM, 50 μm resveratrol (RES), or vehicle (control) for 2 h and treated with 10 nm DHT or vehicle treated for 48 h. Transcript levels of PSA were measured by quantitative RT-PCR analysis of RNA extracted from the cells. Transcript levels are expressed relative to β-actin transcripts. E, The effect of resveratrol or NAM on endogenous AR-dependent and -independent KLK2 gene transcription. LNCaP cells cultured in media containing charcoal-stripped serum were exposed to 10 mm NAM, 50 μm resveratrol (RES), or vehicle (control) for 2 h and treated with 10 nm DHT or vehicle treated for 48 h. Transcript levels of KLK2 were measured by quantitative RT-PCR analysis of RNA extracted from the cells. Transcript levels are expressed relative to β-actin transcripts. F, Top, SIRT1 knockdown increases endogenous AR-dependent PSA genes transcription. Transcript levels of PSA were measured by quantitative RT-PCR analysis of RNA extracted from empty-vector-transfected LNCaP cells (Ctrl) and LNCaP cell lines in which SIRT1 expression levels had been knocked down by stable expression of siRNA (RNAi). The cells were cultured under androgen-deprivation conditions for 3 d, followed by treatment with DHT or vehicle for 48 h. Bottom, Immunoblot analysis of SIRT1 and AR protein levels in an empty-vector-transfected LNCaP cell line (Ctrl) and LNCaP cell lines in which SIRT1 expression levels had been knocked down by stable expression of siRNA (RNAi). G, SIRT1 knockdown increases endogenous AR-dependent KLK2 gene transcription. Transcript levels of KLK2 were measured by quantitative RT-PCR analysis of RNA extracted from empty-vector-transfected LNCaP cells (Ctrl) and an LNCaP cell line in which SIRT1 expression levels had been knocked down by stable expression of siRNA (RNAi). The cells were cultured under androgen-deprivation conditions for 3 d, followed by treatment with DHT or vehicle for 48 h. H, SIRT1 deacetylase activity is required for SIRT1 effects on AR-dependent gene transcription. LNCaP cells cultured in media containing charcoal-stripped serum were cotransfected with the PSA-LUC vector plus an empty vector (Vector), a SIRT1 expression vector (SIRT1), a dominant-negative-SIRT1 vector (H363Y), or SIRT1 expression vector plus dominant-negative-SIRT1 vectors (SIRT1+H363Y), then treated with 10 nm DHT or vehicle, and harvested for assay of LUC activity. Inset shows the results from vehicle treatment, with an expanded y-axis. In all relevant figures, relative LUC activities were normalized to β-gal activity to control for transfection efficiency. The error bars represent the sem. *, P < 0.05 and **, P < 0.01, significant differences between two groups.
We next tested the role of SIRT1 in AR-dependent and -independent gene transcription, using ectopic overexpression of SIRT1 or SIRT1 knockdown by RNA interference (RNAi), in LNCaP cells. Cotransfection of PSA-LUC with a SIRT1 expression vector reduced DHT-stimulated PSA-LUC transcription by 3-fold (P < 0.01) (Fig. 1C). Conversely, SIRT1 knockdown by expression of SIRT1 RNAi resulted in a 2.5-fold (P < 0.01) increase in DHT-stimulated PSA-LUC transcription. In parallel studies, ectopic expression of SIRT1 reduced pMMTV-LUC transcription by 7.5-fold, whereas SIRT1 knockdown by RNAi induced a 3-fold (P < 0.01) increase in pMMTV-LUC transcription (data not shown). The levels of SIRT1 protein in cells transfected with the SIRT1 expression vector or with SIRT1 RNAi were measured and confirmed increased or decreased levels of SIRT1 protein, respectively. SIRT1 overexpression or down-regulation did not affect AR protein expression levels (Fig. 1C). To assess the specificity of SIRT1 in affecting promoter activity, the effect of altering SIRT1 activity on expression of an simian virus 40 promoter-driven LUC reporter was studied. The activity of this vector after transient transfection was not affected by overexpression of SIRT1 (data not shown).
The effect of SIRT1 activity on endogenous AR-responsive genes was next examined. The PSA and kallikrein 2 (KLK2) genes were chosen for study because they are well-recognized targets of AR-dependent transcriptional regulation in vivo (54). Inhibition of SIRT1 activity by NAM treatment induced PSA transcript levels by 3-fold (Fig. 1D) and KLK2 transcripts by 4-fold (Fig. 1E) in the presence of DHT (P < 0.05). Conversely, activation of SIRT1 by exposure to resveratrol suppressed both PSA and KLK2 transcripts below basal levels by 31-fold and 5-fold (P < 0.05), respectively (Fig. 1, D and E). The ability of SIRT1 to suppress transcription of endogenous AR-responsive genes was confirmed independently by analysis of cell lines in which SIRT1 had been stably knocked down by transfection with a vector expressing a SIRT1 hairpin (for SIRT1 and AR protein levels in these cells, see Fig. 1F). Compared with control-transfected cells, knockdown of SIRT1 resulted in induction of PSA transcripts by 3.6-fold (Fig. 1F) and KLK2 transcripts by 4.7-fold (P < 0.05) (Fig. 1G), in the presence of DHT. In the absence of DHT, a minimal but nonsignificant induction of the very low basal levels of transcription was detected after treatment with NAM and in the SIRT1 knockdown cell lines, although no additional suppression was detected after exposure to resveratrol (Fig. 1, D–G). Collectively, these experiments demonstrate that SIRT1 specifically suppresses AR-dependent gene transcription. SIRT1 may also exert a modest repressive effect on AR target genes in the absence of DHT, but the levels of basal expression are so low in the absence of DHT that the significance of any such effect is difficult to discern.
The Deacetylase Activity of SIRT1 Is Required for Suppression of AR-Dependent Transcription
SIRT1 is an NAD-dependent HDAC, the enzymatic activity of which is induced by resveratrol (55) and reduced by NAM (56). Our finding that AR-dependent transcription is regulated by these reciprocal modulators of SIRT1 deacetylase activity is consistent with the enzymatic activity of SIRT1 being responsible for this effect. To independently demonstrate that SIRT1 regulates AR-dependent transcription through its deacetylase activity, we used a dominant-negative mutant that abolishes the deacetylase activity of SIRT1 (SIRT1H363Y-mutation of His363 to Tyr). Coexpression of SIRT1 H363Y protein in LNCaP cells failed to suppress PSA-LUC transcription (Fig. 1H) or pMMTV-LUC transcription (data not shown), whereas cotransfection of wild-type (wt)-SIRT1 in parallel inhibited transcription, as demonstrated previously. When both wt-SIRT1 and SIRT1H363Y expression vectors were cotransfected, the repressive effect of wt-SIRT1 expression was blocked by the dominant-negative mutant. These results establish that the suppression of AR-mediated transcription by SIRT1 requires its deacetylase activity.
SIRT1 Associates with the PSA Promoter and Acts as a Corepressor of AR
To investigate the mechanism underlying SIRT1 regulation of AR-dependent gene transcription, we first determined whether SIRT1 associates with known AR-binding sites in the promoter regions of the PSA gene. Three sets of PCR primer pairs were generated to amplify genomic fragments (∼150 bp in size) encompassing the promoter, the enhancer, and a control region distal to the PSA gene (7 kb upstream of the start site). LNCaP cells were cultured under androgen deprivation for 3 d, followed by treatment with DHT or bicalutamide [Casodex (CDX)], an AR antagonist. Chromatin immunoprecipitation (ChIP) assays were performed using antibodies against SIRT1 and AR. SIRT1 and AR proteins were both detected at the promoter and enhancer regions of the PSA gene (Fig. 2A–D). Exposure to the androgen antagonist bicalutamide increased the recruitment of SIRT1 to both promoter and enhancer regions (Fig. 2, A and B). In contrast, treatment with DHT did not induce SIRT1 occupancy (Fig. 2, A and B). AR recruitment to both regions was induced by exposure to either DHT or bicalutamide (Fig. 2, C and D). Binding of SIRT1 or AR to a control DNA region 7 kb upstream of the PSA gene was not observed (data not shown). These results indicate that SIRT1 and AR bind to the same general promoter/enhancer regions of the PSA gene and that SIRT1 recruitment to the promoter and enhancer is stimulated by androgen antagonists.
ChIP Analysis of the Endogenous PSA Gene Promoter Region LNCaP cells were cultured in charcoal-stripped serum (S) for 3 d, and certain cultures were treated for 4 h with DHT (10 nm) (D), bicalutamide (15 μm) (CDX), or vehicle control (S). ChIP assays were performed using primers sets that amplified the PSA promoter region, the enhancer region, or a distal region upstream of known PSA regulatory elements. Immunoprecipitations were performed using antibodies directed against SIRT1, AR, NCoR, Pol II, and an irrelevant protein (RAG-1). The bound and input DNA were analyzed by Applied Biosystems 7500 Real-Time PCR system by the ΔΔCt cycle threshold method. The results are presented as the relative level of the protein associated with the PSA promoter or enhancer, normalized to irrelevant control antibody and input DNA. A, SIRT1 on the PSA promoter. B, SIRT1 on the PSA enhancer. C, AR on the PSA promoter. D, AR on the PSA enhancer. E, Pol II on the PSA promoter. F, Pol II on the PSA enhancer. G, NCoR on the PSA promoter. H, NCoR on the PSA enhancer. I, Immunoblot analysis of AR protein levels in DU145 cells transfected with wt-AR or empty vector. J, Bicalutamide-induced recruitment of SIRT1 to the endogenous PSA promoter requires the AR. DU145 cells were cultured in DMEM plus 10% charcoal-treated FBS. The cells were transfected with AR or mock transfected and treated with bicalutamide (CDX) or DHT (D). ChIP assays were performed using antibodies directed against SIRT1.
Parallel ChIP assays examining polymerase II (Pol II) recruitment to the promoter and enhancer as a marker for active transcription demonstrated increases in Pol II occupancy after exposure to DHT but not when bicalutamide was present (Fig. 2, E and F). Pol II recruitment was inversely related to occupancy by SIRT1, and bicalutamide uncoupled AR occupancy at the promoter from Pol II recruitment.
NCoR is an established repressor of AR-driven transcription and is known to physically associate with the AR (32, 57, 58). SIRT1 has been shown to interact with NCoR and suppress the transcriptional activity of certain genes, such as the PPAR receptor (48). To determine whether NCoR occupancy parallels SIRT1 occupancy at the PSA enhancer and promoter, ChIP assays were conducted with an anti-NCoR antibody. We observed increased occupancy by NCoR at both the endogenous PSA promoter and enhancer sites in the presence of bicalutamide (Fig. 2, G and H).
Finally, we determined whether the recruitment of SIRT1 to the PSA promoter by androgen antagonists required the AR. ChIP assays were performed using anti-SIRT1 antibody in DU145 cells, an AR-negative prostate cancer cell line, with or without transfection of an AR expression vector, in the presence of DHT or CDX (for AR protein expression levels in DU145 cells after transfection, see Fig. 2I). Exposure to DHT did not induce SIRT1 occupancy at the PSA promoter, whether or not the AR was expressed (Fig. 2J) Bicalutamide treatment did not induce SIRT1 occupancy at the PSA promoter in control (empty-vector transfected) AR-negative DU145 cells (Fig. 2J). When AR was introduced into the cells by transfection, however, exposure to bicalutamide stimulated recruitment of SIRT1 to the PSA promoter approximately 4-fold more efficiently than in the absence of AR. These results suggest that antagonist-bound AR mediates SIRT1 occupancy at the PSA promoter.
These results are consistent with a model in which SIRT1 and NCoR complex with the AR at the PSA promoter/enhancer under situations of transcriptional silencing by androgen antagonists, thus facilitating corepression, in which NCoR acts as a corepressor adaptor for the AR and may bridge SIRT1 with the AR at the PSA promoter/enhancer.
Histone H3 in the PSA Promoter Region Is a Potential Target of SIRT1
Several lines of evidence support the importance of local histone acetylation in transcriptional activation by nuclear receptors (16, 17, 59). The acetylation level of histone H3 at the PSA promoter has been reported to increase after exposure to DHT (58). We examined the acetylation state of histone H3 at the PSA promoter and enhancer using an antiacetylated histone H3 antibody in a ChIP assay. Exposure to DHT produced a marked increase in local histone H3 acetylation. Enhancement of SIRT1 activity by resveratrol treatment reversed the H3 acetylation induced by DHT (Fig. 3, A and B). To determine whether SIRT1 itself was required for these ligand-dependent changes in local histone acetylation, we examined local H3 acetylation in cell lines in which SIRT1 levels had been knocked down by SIRT1 hairpin RNA expression (Fig. 3C). SIRT1 knockdown elevated the levels of acetylated histone H3 at the enhancer and promoter compared with levels in control vector-transfected LNCaP cells, resulting in approximately 4-fold more acetylated H3 in the SIRT1 knockdown cells at the PSA enhancer and 2-fold more at the PSA promoter (P < 0.05) in response to DHT (Fig. 3C). These results suggest that histone H3 in the PSA promoter/enhancer regions may be one direct target of SIRT1 activity (or alternatively, and less likely, that SIRT1 is influencing the activity of other coregulatory histone acetylases or deacetylases at AR-responsive promoters).
Androgen and Antagonists Alter Histone H3 Acetylation at the PSA Promoter A, Histone H3 acetylation at the PSA promoter. Cross-linked chromatin was extracted from cells cultured under androgen-deprivation conditions and then treated with 10 nm DHT (D), DHT plus 25 μm resveratrol (D+RES), or vehicle (S) for 4 h. Anti-histone H3 antibody (AcH3) or an irrelevant antibody (anti-RAG) were used for immunoprecipitation. The ethidium-stained PCR products of the ChIP assay are shown. B, Quantitative PCR results from the same ChIP assays, analyzing both the PSA promoter and enhancer. C, Quantitative PCR analysis of ChIP assays for acetylated histone H3 at the PSA promoter or enhancer in the presence of DHT in control-transfected LNCaP cells (Ctrl) and in LNCaP cell lines in which SIRT1 expression levels had been knocked down by stable expression of siRNA (RNAi). *, P < 0.05, significant differences between two groups.
SIRT1 Is Required for Bicalutamide-Mediated Transcriptional Suppression of the PSA Promoter and Growth Suppression by Bicalutamide
The finding that bicalutamide exposure induces recruitment of SIRT1 to the PSA promoter, coincident with changes in local promoter histone acetylation and transcriptional repression, raised the possibility of a functional role for SIRT1 in the transcriptional repression induced by androgen antagonists. To test for a necessary role for SIRT1 in bicalutamide-mediated transcription repression, the activity of the androgen-responsive PSA-LUC vector transfected into SIRT1-knockdown LNCaP cells or control (empty-vector)-transfected LNCaP cells was assessed. Treatment with bicalutamide suppressed DHT-induced PSA-LUC transcriptional activity in cells expressing SIRT1 by 80% but only suppressed transcription by 28% in SIRT1 knockdown cells (P < 0.05) (Fig. 4A; for SIRT1 and AR protein levels in these cells, see B). To verify that the observed defect of bicalutamide-mediated transcription suppression in the SIRT1-knockdown cells was indeed attributable to the suppression of SIRT1, we preformed a rescue experiment by reintroducing wt-SIRT1 vector back into the stable SIRT1-knockdown cell lines and tested for restoration of the activity of bicalutamide. Whereas treatment with bicalutamide suppressed DHT-induced PSA-LUC transcriptional activity by a nonsignificant 25% in SIRT1-deficient cells cotransfected with an empty vector and PSA-LUC, bicalutamide suppressed DHT-induced PSA-LUC transcriptional activity by 55% in the cells cotransfected with wt-SIRT1 and PSA-LUC (P < 0.05) (Fig. 4C). However, when cotransfected with a dominant-negative mutant that abolishes the deacetylase activity of SIRT1 (SIRT1H363Y) and PSA-LUC, bicalutamide did not significantly suppress DHT-induced PSA-LUC transcriptional activity (for levels of wt-SIRT1, SIRT1H363Y, and AR protein expression in these cells, see Fig. 4D). These studies demonstrate that the defect in bicalutamide-mediated transcriptional suppression in the knockdown cells is attributable to deficiency of SIRT1 deacetylate activity rather than to an off-target effect of small interfering RNA (siRNA), consistent with the results independently obtained using chemical inhibitors of SIRT1.
Reversal of Bicalutamide-Mediated Transcriptional and Growth Suppression of Cells by SIRT1 Depletion A, SIRT1 knockdown impairs bicalutamide-mediated PSA transcription repression. A SIRT1 knockdown cell line (RNAi) and control-transfected LNCaP cells (Ctrl) were cultured in charcoal-stripped serum, transfected with the PSA-LUC reporter vector, and then treated with DHT (1 nm) (D), DHT plus bicalutamide (10 μm) (D+CDX), or vehicle (S). Cells were harvested after 24 h, and lysates were assayed for LUC activity. The data are presented as a percentage of the activity obtained in DHT alone (assigned the value of 100). B, Immunoblot analysis of SIRT1, AR, and β-actin levels in control-transfected LNCaP cells (Ctrl) and a LNCaP line in which SIRT1 had been knocked down by stable expression of siRNA (RNAi). C, SIRT1 overexpression can partially rescue the defect of bicalutamide-mediated PSA transcription suppression induced by SIRT1 depletion. SIRT1 knockdown cells were cultured in charcoal-stripped serum, cotransfected with PSA-LUC and empty-vector (RNAi), with PSA-LUC and SIRT1 expression vector (RNAi+SIRT1wt), or with PSA-LUC and SIRT1 catalytic inactive mutant (RNAi+SIRT1H363Y) and then treated with DHT (1 nm) (D), DHT plus bicalutamide (10 μm) (D+CDX), or vehicle (S). Cells were harvested after 24 h, and lysates were assayed for LUC activity. The data are presented as a percentage of the activity obtained with exposure to DHT alone (assigned the value of 100). D, Immunoblot analysis of SIRT1 and AR protein levels in SIRT1 knockdown cell lines transfected with empty vector (RNAi), wt-SIRT (SIRT1wt), or catalytic inactive mutant (SIRT1H363Y). E, SIRT1 is required for bicalutamide-mediated cell growth suppression. Parental empty-vector-transfected LNCaP cells (Ctrl) or LNCaP cells in which SIRT1 had been stably knocked down by siRNA expression (RNAi) were cultured in charcoal-stripped serum for 3 d, exposed to DHT (1 nm) for another 3 d, and then treated with addition of bicalutamide (2.5 μm) (D+CDX) or vehicle (D) for 48 h. Viable cells were quantitated by MTS assay, and the results are expressed relative to the values obtained from wells cultured without added bicalutamide (assigned as value of 100). *, P < 0.05, significant differences between two groups.
Together with the data on Pol II recruitment shown in Fig. 2, E and F, these results suggest that SIRT1 is required for bicalutamide-mediated AR-dependent transcriptional suppression. Parallel studies using treatment with NAM to suppress SIRT1 activity confirmed the requirement for SIRT in the action of androgen antagonists (data not shown).
Bicalutamide suppresses both AR-dependent transcription and prostate cancer cell proliferation. We next assessed whether SIRT1 plays a role in bicalutamide-mediated prostate cancer cell growth suppression, using the SIRT1 partial-knockdown LNCaP cells and control (empty-vector transfected) cells. Exposure to bicalutamide suppressed AR-dependent growth by 60% in the control (empty-vector)-transfected cells but induced only 30% growth suppression in the stable SIRT1 partial knockdown cells (P < 0.05) (Fig. 4E). Together with the studies presented above, these findings suggest that SIRT1 is required both for bicalutamide-mediated transcriptional suppression of AR-responsive genes and for bicalutamide-mediated prostate cancer cell growth suppression.
SIRT1 Down-Regulation Enhances LNCaP Cell Sensitivity to DHT with Respect to AR-Dependent Gene Transcription and Cell Growth
It has been proposed that the inevitable androgen “independence” and antagonist insensitivity that occurs during the progression of prostate cancer may in some cases be the result of an acquired hypersensitivity of the AR to androgen, resulting in the ability of the tumor cells to respond to very low levels of androgen (3, 12, 60). The studies described above demonstrate that SIRT1 can regulate AR-responsive genes in a ligand-dependent manner. We asked, therefore, whether alterations in SIRT1 levels or activity could increase the sensitivity of hormone-responsive cells to androgen. We determined the relative androgen sensitivity of AR-responsive genes using the androgen-responsive PSA promoter-LUC vector in control-transfected LNCaP cells and in stable SIRT1-knockdown LNCaP cell lines. Cells were androgen deprived for 72 h, transfected with the PSA-LUC reporter, and treated with various concentrations of DHT in the absence or presence of the SIRT1 inhibitor NAM. At very low concentrations of DHT, ranging from 0.01 to 0.1 nm, there was minimal induction of PSA-LUC activity above basal levels in the control (empty-vector)-transfected LNCaP cells (Fig. 5A). In contrast, both the SIRT1 knockdown cells and NAM-treated control-transfected LNCaP cells demonstrated significant increases in PSA-LUC transcription at these low doses of DHT. For example, at 0.1 nm DHT, PSA-LUC transcription in SIRT1-knockdown or NAM-treated cells was induced by more than 10-fold (P < 0.05) compared with parental LNCaP cells. Thus, inhibition of SIRT1 activity by two independent methods enhanced the transcriptional sensitivity of the androgen-responsive PSA promoter to DHT.
SIRT1 Depletion Increases the Sensitivity of AR-Dependent Gene Transcription and Cell Proliferation to DHT A, SIRT1 knockdown increases the sensitivity of AR-dependent gene transcription to DHT. Parental empty-vector-transfected LNCaP cells (Ctrl) or LNCaP cells in which SIRT1 had been stably knocked down by siRNA expression (RNAi) were cultured in charcoal-stripped serum, transfected with the PSA-LUC reporter vector, treated with NAM or vehicle (Ctrl), and exposed to the indicated concentrations of DHT (0–1 nm). Cells were harvested after 48 h and assayed for LUC activity. B, Immunoblot analysis of SIRT1, AR, and β-actin levels in control-transfected LNCaP cells (Ctrl) and three LNCaP lines in which SIRT1 had been knocked down by stable expression of siRNA (RNAi-1, RNAi-2, and RNAi-3). C, SIRT1 knockdown increases the sensitivity of the proliferative response of LNCaP cells to DHT. Control, empty-vector-transfected LNCaP cells (Ctrl) or LNCaP cell lines in which SIRT1 had been stably knocked down by siRNA expression (lines RNAi-1, RNAi-2, and RNAi-3) were made quiescent by culture in charcoal-stripped serum and exposed to the indicated concentrations of DHT (0–10 nm). Viable cells were quantitated at 72 h by MTT assay, and the results are expressed relative to values obtained from plates cultured without added DHT (assigned an arbitrary value of 1). In all relevant figures, relative LUC activities were normalized to β-gal activity to control for transfection efficiency. The error bars represent the sem. *, P < 0.05, significant differences between two groups.
To determine whether SIRT1 knockdown increased the sensitivity of hormone-responsive cells to the mitogenic effects of androgen, the androgen responsiveness of control (empty-vector)-transfected LNCaP cells and three SIRT1-knockdown LNCaP cell lines (Fig. 5B) was compared. After androgen deprivation, cell lines were treated with varying concentrations of DHT for 6 d, and their proliferation was assessed. Each of the three SIRT1-knockdown cell lines exhibited significantly enhanced sensitivity to the mitogenic effects of DHT at concentrations less than 1 nm (Fig. 5C). For example, at concentrations of 0.05 nm DHT, the SIRT1-knockdown cell line RNAi-2 proliferated 300% compared with only approximately 50% for control cells (P < 0.05). Maximal proliferation was observed at 0.5 nm DHT in the RNAi-2 cell line compared with 1 nm DHT in the control cells. Similar degrees of sensitization to DHT were observed in the RNAi-1 line and other SIRT1-knockdown cell lines (Fig. 5C). These data demonstrate that suppression of SIRT1 increases the sensitivity of hormone-responsive prostate cancer cells to the mitogenic effects of DHT.
Discussion
In this study, we demonstrate a role for SIRT1 in the regulation of AR-dependent gene transcription and androgen-dependent prostate cancer cell growth. We identify SIRT1 as a novel corepressor of AR suppressing AR-dependent gene transcription. In addition, we demonstrate that SIRT1 is recruited to AREs by AR antagonists and is required for bicalutamide-mediated transcriptional repression and prostate cancer cell growth suppression. We further demonstrate that down-regulation of SIRT1 increases the sensitivity of prostate cancer cells to the proliferative and transcriptional actions of androgens. Finally, we provide evidence to suggest that the mechanism of SIRT1-mediated transcriptional inhibition on AR-responsive genes may be attributable to local deacetylation of histone H3 in AR-dependent gene promoters.
AR corepressor complexes play a critical role in regulating AR activity with precision and efficiency (15). To date, several corepressors of AR have been characterized, including NCoR, SMRT, and HDAC1–HDAC3 (24, 25, 31–34). HDAC1–HDAC3 are class I deacetylases and form a holo-corepressor complex with AR and NCoR at the PSA promoter, suppressing AR-dependent transcription (24, 25). SIRT1 is a member of the class III HDAC family, the deacetylase activity of which is NAD dependent. We report herein that SIRT1 is required for antagonist (bicalutamide)-mediated transcription suppression of AR-dependent genes. To our knowledge, this is the first functional demonstration of ligand-induced recruitment of a class III HDAC to transcriptionally repress any promoter and the first demonstration of a necessary role for a class III HDAC as a transcriptional corepressor in steroid hormone-responsive gene regulation. Together with the report of Zhu et al. (24), our data indicate that both class I HDACs (1, 2, and 3) and class III HDACs (SIRT1) are required for androgen antagonist-mediated transcriptional repression. We also found that SIRT1 inhibits AR-dependent transcription tonically during transcriptional activation in response to androgens, in that modulation of SIRT1 activity influences the amplitude of the transcriptional response. Several biochemical and molecular genetic studies have shown that a chimeric AR/coactivator/corepressor complex exists at the promoter at the onset of the androgenic response. Corepressors (HDAC1 and SMRT) and coactivators [TIP60 (TAT-interacting protein 60)] have simultaneously been identified in this complex (61). The corepressor elements may attenuate agonist-induced transactivation, acting transiently as part of a cycle of cofactors recruited to target promoters by ligand-bound receptors. We suggest that SIRT1 may be another element in the incremental and constant regulation of AR activity, playing a role to overcome or reduce coactivator-mediated effects on the receptor, thereby preventing excessive gene expression and finely tuning the transcriptional response. The dramatic recruitment of SIRT1 to ARE-containing promoter and enhancer elements after exposure to androgen antagonists and the finding that SIRT1 is required for transcriptional repression of exogenous and endogenous AR-responsive promoters clearly demonstrate its functional necessity for the action of androgen antagonists.
Resveratrol is recognized as an agonist of SIRT1 (55) and is a natural chemopreventive agent in prostate cancer models (63, 64). Both SIRT1 overexpression and exposure to resveratrol suppress AR-dependent transcription, although their efficiency in suppressing AR-dependent transcription differs. Resveratrol suppresses transcription by 16-fold, whereas SIRT1 overexpression represses it by approximately 2.5-fold. There are several possible mechanisms accounting for this difference. First, although resveratrol is a agonist of SIRT1, it also targets a number of other proteins and signaling pathways (65). It is therefore possible that resveratrol may be acting on additional proteins contributing to transcriptional repression. In addition, as prostate cancer cell lines take up DNA with low efficiency, the level of SIRT1 expression may partially limit the effects of SIRT1 overexpression on the PSA gene transcription. Another possible explanation could be that resveratrol might be acting not directly on SIRT1 but rather is blocking androgen binding to the AR, particularly in prostate cancer cell lines in which the AR is mutated and exhibits relaxed ligand specificity, such as LNCaP cells. Blocking androgen binding to the AR would then impair AR translocation, prevent recruitment of AR to AREs, and suppress AR-mediated transcriptional activity. As shown in supplemental Fig. S1 (published as supplemental data on The Endocrine Society’s Journals Online web site at http://mend.endojournals.org), we ruled out this possible mechanism by demonstrating that exposure to resveratrol does not affect AR binding to the PSA promoter in response to androgens. To exclude the possibility that resveratrol acts differentially on the mutated AR in LNCaP cells compared with wt-AR, we also performed a PSA-reporter assay using a 293 cell line, transfected and expressing wt-AR, in the presence or absence of resveratrol. Exposure to resveratrol produced similar levels of suppression on the transcriptional activity of this wt-AR as it did on the activity of the mutant AR in LNCaP cells (supplemental Fig. S2, published as supplemental data on The Endocrine Society’s Journals Online web site at http://mend.endojournals.org).
Unlike the thyroid hormone and retinoid receptors, the AR does not bind to and repress target gene transcription in the absence of ligand (66). Instead, the switch to transcriptional repression is induced by binding of an androgen antagonist. Antagonist-bound AR is then rendered unable to bind coactivator proteins and is also newly able to interact with, and recruit, transcriptional repressors, including an NCoR holo-repressor complex containing TBL (transducing β-like protein)/TBLR1 (transducing β-like protein receptor 1), HDAC1–HDAC3, Brg1 (Brm-related gene 1), and Sin3 (24). NCoR and HDAC1–HDAC3 appear to be required for transcriptional repression by androgen antagonists (24). The state of gene repression is correlated with histone deacetylation by corepressors and their associated HDACs (24, 34, 66, 67). We observed significant recruitment of SIRT1 to the PSA promoter in the presence of bicalutamide and found that SIRT1 is required for bicalutamide-mediated transcription suppression and growth suppression. This recruitment was dependent on the presence of the AR. We found that local deacetylation of histone H3 and transcriptional suppression of AR-responsive genes was dependent on SIRT1 recruitment to the promoter. There is no evidence that SIRT1 can bind to DNA directly, and targeted deacetylation of histones at promoters is thought to occur through interaction and recruitment of HDACs by specific DNA-binding proteins. Our results suggest that NCoR, a known adaptor protein for SIRT1, works as a corepressor adaptor for AR, may bridge SIRT1 with AR at the PSA promoter, thus making SIRT1 available to deacetylate H3 histones at the promoter, and suppress AR-dependent transcription. In unpublished studies, we found that the AR associates with SIRT1 in a ligand-dependent manner in cell lysates and that NCoR can bind to SIRT1, also in an antagonist-dependent manner, raising the possibility that the AR directly recruits the NCoR/SIRT1 complex when bound to androgen antagonists. Our observation that new recruitment of SIRT1 to the PSA promoter requires the AR is consistent with such a model. A rigorous testing of this model is in progress. It is also possible that SIRT1 may act on additional protein targets to achieve transcriptional repression. Because p300, an AR coactivator, is a deacetylation target of SIRT1 (68), it is tempting to speculate that SIRT1 may act in part through deacetylation of p300 to control AR-dependent transcription. It is also possible that SIRT1 inhibits AR-dependent transcription in part through deacetylation of the AR itself. In unpublished studies, we find that SIRT1 can associate with AR, and inhibition of SIRT1 increases the level of acetylation of the AR. Interestingly, an independent report published (69) during the submission of this paper indicates that SIRT1 can partially regulate AR-dependent transcription through deacetylation of the AR.
The formation of an active coactivator complex on the PSA promoter in response to androgens involves the recruitment of AR to two discrete regions of the gene, an enhancer (containing ARE III) and a promoter region (containing ARE I and ARE II) (25, 70). AR recruitment to the PSA promoter and enhancer regions in response to androgen antagonists, however, remains controversial. One previous study suggested that formation of an AR repressor complex involves only the promoter region (25), because no recruitment of NCoR, SMRT, HDAC1, or HDAC2 was observed in the enhancer region after exposure to an antagonist, although another study found NCoR to be recruited to both the promoter and the enhancer (58). We also observed recruitment of SIRT1 and NCoR to both the enhancer and the promoter in response to antagonist stimulation in our studies. The reason for these disparate findings is not clear, although each study used different primer sets in the ChIP, and neither of the previous studies attempted to determine whether binding of a corepressor complex to the enhancer region was functionally important.
Androgen depletion, or blockade of androgen signaling, represents the major therapeutic strategy for the treatment of advanced prostate cancer. Invariably, however, the disease becomes progressive and unresponsive to androgen ablation therapy (hormone-refractory) within an average of 18 months (71). This evolution occurs through one of several discrete and incompletely understood molecular mechanisms, some of which permit AR signaling in the absence of, or at very low concentrations of, androgens (6, 60, 73). We report herein that down-regulation of SIRT1 increases the sensitivity of LNCaP cells to DHT, as assessed by at least two functional outcomes. Down-regulation of SIRT1 activity by pharmacological or genetic means increased the sensitivity of cells to transcriptional activation of AR-responsive target genes by androgens. In parallel, the proliferation of androgen-responsive cells (which is not necessarily directly linked to the transcriptional activation of AR-regulated genes such as PSA and KLK) was also significantly enhanced at low concentrations of DHT by SIRT1 inhibition. Conversely, we also show that SIRT1 is required for bicalutamide-mediated growth suppression, suggesting that loss of SIRT1 might contribute to the development of antagonist resistance in prostate cancer. Thus, loss of SIRT1 increases the sensitivity of prostate cancer cells to growth in response to androgens but simultaneously decreases their sensitivity to the antiproliferative effects of androgen antagonists.
These findings suggest the possibility of a role for SIRT1 loss in the evolution to hormone-refractory prostate cancer. It is perhaps noteworthy that global histone modification in prostate tumor tissues, including acetylation of H3, predict risk of prostate cancer recurrence (74), and we found that SIRT1 down-regulation increases the acetylation level of H3 at AR-dependent gene promoters. We also note that the SIRT1 expression level is higher in the AR-dependent cell line LNCaP than in the AR-independent cell lines DU145 and PC3 (data not shown). In other studies not reported here, we found that higher concentrations of bicalutamide can decrease SIRT1 protein levels in LNCaP cells. Chronic bicalutamide treatment in patients may thus result in depletion of SIRT1 protein levels in prostate cancer cells, thereby increasing their sensitivity to circulating androgens and their progression to androgen independence.
SIRT1 has been shown to play roles in aging (40) and in diseases and pathways related to aging, such as diabetes (75), fat mobilization (48), and insulin signaling (76, 77). Prostate cancer is considered a disease of aging because its incidence increases with age more rapidly than do other types of cancer (62). A link between cellular aging and SIRT1 protein expression in humans has been proposed, because the expression of endogenous SIRT1 protein progressively decreases during replication/aging of normal human fibroblasts in culture (72). Similarly, fetal tissues show higher expression of SIRT1 than adult tissues (79). It is therefore possible that the age-related decline in SIRT1 expression in humans may result in abnormal AR activity or function, promoting prostate cancer development during aging, and this hypothesis is currently being investigated.
Materials and Methods
Cells, Plasmids, and Antibodies
LNCaP and DU145 cells were obtained from the American Type Culture Collection (Manassas, VA). LNCaP cells were maintained in RPMI 1640 medium with 10% fetal bovine serum (FBS) or in 10% charcoal-treated FBS (HyClone, Logan, UT). DU145 cells were maintained in DMEM with 10% FBS. The human SIRT1 expression vector pcDNA3.1-SIRT1 was generated by subcloning SIRT1 cDNA into a pcDNA3.1(+)-based V5His vector, generating a C-terminal V5His-tagged fusion protein, using standard PCR-based strategies. The deacetylase-defective SIRT1H363Y mutant expression plasmid was generously provided by Dr. Melanie Ott (University of California, San Francisco, San Francisco, CA). The SIRT1 RNAi vectors (pSUPER.retro.puro-SIRT1 and pSUPER.retro.neo-SIRT1) were generously provided by Dr. F. Picard (Laval University, Quebec, Canada) and Dr. Melanie Ott, respectively. The pSUPER.retro.puro-SIRT1 vector contains SIRT1 sequence 5′-GATGAAGTTGACCTCCTCA-3′ and the puromycin-resistance gene, and the pSUPER.retro.neo-SIRT1 vector contains 5′-CTTGTACGACGAAGACGA-3′ and the neomycin-resistance gene. The AR expression vector was provided by Dr. Marco Marcelli (Baylor College of Medicine, Houston, TX). The reporter plasmid PSA-LUC, containing the LUC gene under the control of a fragment of the human PSA gene promoter, was provided by Dr. A. O. Brinkmann (Erasmus Medical Center, Rotterdam, The Netherlands). The pMMTV-LUC reporter plasmid, containing the LUC gene driven by the MMTV long terminal repeat, was provided by Dr. R. Spanjaard (Boston University, Boston, MA).
Antibodies to SIRT1 (catalog no. 05-707), AR (PG-21; catalog no. 06-680), and acetyl-histone H3 (catalog no. 06-599) were purchased from Upstate Biotechnology (Lake Placid, NY). The rabbit polyclonal SIRT1 antibody was generously provided by Dr. Roy A. Frye (Veterans Affairs Medical Center, Pittsburgh, PA). Antibodies to NCoR (H303; sc-8994), Pol II (A-10; sc-17798), AR (441; sc-7305), and RAG-1 (recombination activating gene 1) (sc-363) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Antibody to β-actin (A-2066) was purchased from Sigma (St. Louis, MO).
DHT (A-8380), NAM (N3376), resveratrol (R5010), and NAD (N1636) were purchased from Sigma, and bicalutamide (CDX; s210183) was provided by AstraZeneca (London, UK).
Retroviral Infection and Establishment of Stable SIRT1-Knockdown Cell Lines
The Phoenix packaging cell line was transfected with the pSUPER.retro.puro-SIRT1, pSUPER.retro.neo-SIRT1, or the pSUPER.retro vectors separately, using Lipofectamine plus (Invitrogen, Carlsbad, CA). After 48 h, the medium containing retrovirus was collected, filtered, treated with polybrene, and transferred to LNCaP cell cultures. Infected cells were selected with G418 or puromycin plus G418 for isolation of stably infected colonies. SIRT1-knockdown stable cell line RNAi-2 was isolated from pSUPER.retro.neo-SIRT1-infected cells and is G418 resistant. SIRT1-knockdown stable cell lines RNAi-1 and RNAi-3 were selected from cells coinfected with pSUPER.retro.neo-SIRT1 and pSUPER.retro.puro-SIRT1 and are puromycin and G418 double resistant.
LUC Assay
LNCaP cells were cultured in six-well plates in RPMI 1640 with 10% charcoal-treated FBS (HyClone) (androgen-deprivation conditions) for 3 d and then cotransfected with 1 μg PSA-LUC or pMMTV-LUC reporter vectors together with SIRT1 expression vectors, using Lipofectamine plus (Invitrogen). Fresh medium was added after overnight transfection. Transfected cells were exposed to resveratrol or NAM or vehicle for 24 h and then treated with DHT or vehicle control for 24 h before assay for LUC activity (Luciferase Assay System; catalog no. E1500; Promega, Madison, WI). The relative LUC activities were normalized to the activity of a cotransfected β-galactosidase (β-gal) expression vector, as an internal control for transfection efficiency. Each experiment was performed in triplicate and repeated a minimum of three times. These values were used to determine sds, with error bars indicating ±sem.
Cell-Based HDAC Assay
Cell-based HDAC assays were performed as described previously (55). LNCaP cells were cultured under androgen-deprivation conditions for 3 d and then treated with resveratrol or NAM. Cells were washed with PBS and lysed with 1× HDAC lysis buffer. Equal amounts of lysates were analyzed for enzyme activity using the HDAC Fluos de Lys Fluorescent Assay System (catalog no. AK-500; Biomol, Plymouth Meeting, PA).
Real-Time RT-PCR
Total cellular RNA was isolated using Trizol reagent (Invitrogen) according to the instructions of the manufacturer. A two-step RT-PCR method was used to synthesize single-stranded cDNA (SuperScript TM III First Strand kit; catalog no. 18080-051; Invitrogen). Target genes were analyzed by real-time PCR using Applied Biosystems (Foster City, CA) 7500 Fast Real-Time PCR System with SYBR Green I dye (catalog no. 4309155). The primers used were as follows: PSA forward, TGCCCACTGCATCAGGAACA; PSA reverse, GTCCAGCGTCCAGCACACAG; KLK2 forward, CCTGGCAGGTGGCTGTGTAC; KLK2 reverse, TGTGCCGACCCAGCCA; β-actin forward, GAGAAAATCTGGCACCACACC; and β-actin reverse, ATACCCCTCGTAGATGGGCAC. PCR reactions were performed in triplicate. β-Actin mRNA abundance was analyzed in each sample. The PSA and KLK2 mRNA levels were normalized to the β-actin mRNA level. The specificities of the RT-PCR products were monitored by melting curve analysis and also verified by agarose gel electrophoresis.
ChIP
LNCaP cells (107 cells) were grown in RPMI 1640 (Invitrogen) supplemented with 10% charcoal-dextran-stripped FBS. After 3 d of cultivation, cells were treated with DHT or bicalutamide for 4 h, trypsinized, and washed twice with PBS. DU145 (2 × 107 cells) were cultured in DMEM plus 10% FCS and transfected with 20 μg AR expression vector or empty vector using Lipofectamine plus (Invitrogen). Fresh medium was added after overnight transfection. Transfected cells were cultured for another 48 h then treated with bicalutamide or vehicle for 4 h, trypsinized, and washed twice with PBS. The harvested cells (LNCaP or DU145) were resuspended in 10 ml culture medium, cross-linked with 1% formaldehyde at room temperature for 10 min, and washed three times with ice-cold PBS. The pellets were then resuspended in 0.4 ml lysis buffer [1% SDS, 10 mm EDTA, 50 mm Tris-HCl (pH 8.0), 1 mm phenylmethylsulfonylfluoride (PMSF), and protease inhibitor cocktail (Roche Molecular Biochemicals, Indianapolis, IN)], incubated 20 min at room temperature, and sonicated four times for 10 sec each, followed by a 1 min pulse, at 20% power (Sonic Dismembrator, model 550; Fisher, Pittsburgh, PA). The samples were centrifuged for 10 min, and the supernatants were collected and diluted 5-fold in dilution buffer [1% Triton X-100, 2 mm EDTA, 150 mm NaCl, 20 mm Tris-HCl (pH 8.1), and protein inhibitor cocktail], followed by immunoclearing with protein A/G-Sepharose (50 μl of 50% slurry) for 1 h at 4 C. One hundred microliters of the supernatant was reserved as input. Immunoprecipitation was performed for 1 h at room temperature with specific antibodies (anti-SIRT1, anti-AR, antiacetylated histone H3, anti-NCoR, anti-Pol II, or irrelevant control antibody anti-RAG-1). Seventy microliters of protein A/G-Sepharose and 1 μg salmon sperm DNA were added, and the incubation was continued overnight at 4 C. Sepharose beads were washed sequentially for 5 min each in wash I [0.1% SDS, 1% Triton X-100, 2 mm EDTA, 20 mm Tris-HCl (pH 8.0), 150 mm NaCl, and 1 mm PMSF], wash II [0.1% SDS, 1% Triton X-100, 2 mm EDTA, 20 mm Tris-HCl (pH 8.1), 500 mm NaCl, and 1 mm PMSF], wash III [1% Nonidet P-40, 0.25 m LiCl, 1 mm EDTA, 10 mm Tris-HCl (pH 8.0), and 1 mm PMSF], and Tris-EDTA buffer. Beads were then extracted twice with 1% SDS and 0.1 m NaHCO3 for 15 min at room temperature, with rotation. Eluates were pooled and heated to 65 C in 0.2 m NaCl for 6 h to reverse the formaldehyde cross-linking. The samples were treated with Protein K for 1 h at 37 C, the DNA fragments were purified by phenol/chloroform extraction, followed by ethanol precipitation at − 20 C for 10 min, and the precipitates were washed with 70% ethanol. For PCR, a 2 μl aliquot of the total 50 μl of extracted DNA was amplified in 21–25 PCR cycles for gel analysis or, in real-time PCR, a 2 μl aliquot of the total 50 μl of immunoprecipitation, and input DNA were analyzed using Applied Biosystems 7500 Fast Real-Time PCR System with SYBR Green I dye (catalog no. 4309155). Triplicate PCR reactions for each sample were preformed, and each ChIP assay was performed on at least three independent experiments. The primer sequences were as follows: PSA promoter forward, ACAATCTCCTGAGTGCTGGTGT; PSA promoter reverse, GCAGAGGAGACATGCCCA G; PSA enhancer forward, GAGAATTGCCTCC CAACACTG; PSA enhancer reverse, TGCCAGA CACAGTCGATCG; distal control primer forward, TTCACCGTGTTGGCCAGG; and distal control primer reverse, ATGGTGGCTCACGCC TG.
Immunoblots
Cells (5 × 106) treated with different reagents were lysed in 200 μl 1% Nonidet P-40 immunoblotting lysis buffer. The samples were separated on 8% PAGE gels and further analyzed after transfer by immunoblot analysis with anti-AR (441), anti-SIRT1 (05-707), or anti-β-actin antibodies.
Cell Viability Assay
MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] (M2128; Sigma) or MTS [3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium] (G3580; Promega) assays were used to quantitate cell viability. Cells were plated at a density of 4 × 104 cells per well on 24-well plates (for MTT assay) or 104 cells per well on 96-well plates (for MTS assay) and cultured under androgen-deprivation conditions for 3 d, followed by treatment with different concentrations of DHT for another 3 d. MTT (50 mg/ml) was added to the medium for 3 h, then the supernatant was removed, and the formazone crystals were dissolved using dimethylsulfoxide. The absorbance was read at 690 nm on an ELISA plate reader. For the MTS assay, 20 μl of CellTiter96Aqueous one solution reagent was added to the medium for 1 h, and the absorbance was read at 490 nm.
Acknowledgments
We sincerely thank Dr. Margie Oettinger (Genetic Department of Harvard Medical School, Boston, MA) for her continuous support. We thank Dr. A. Brinkmann (Erasmus Medical Center, Rotterdam, The Netherlands), Dr. M. Marcelli (Baylor College of Medicine, Waco, TX), Dr. M. Ott (University of California, San Francisco, La Jolla, CA), Dr. F. Picard (Laval University, Quebec, Canada), Dr. S. Qin (Brigham and Women’s Hospital, Harvard Medical School, Boston, MA), Dr. R. A. Frye (Veterans Affairs Medical Center, Pittsburgh, PA), Dr. D. Sinclair (Department of Pathology, Harvard Medical School, Boston, MA), and Drs. R. Spanjaard and X. Zhao (Boston University School of Medicine, Boston, MA) for reagents and helpful discussions.
This work was supported American Cancer Society Grant IRG-72-001-27-IRG (to Y.D.), a Boston University School of Medicine Department of Medicine Pilot Project Grant award (to Y.D.), National Cancer Institute Grant CA101992-03, and the Karin Grunebaum Cancer Research Foundation (to D.V.F.).
Disclosure Statement: The authors have nothing to disclose.
Abbreviations
- AR
Androgen receptor;
- ARE
androgen-response element;
- CDX
Casodex;
- ChIP
chromatin immunoprecipitation;
- DHT
dihydrotestosterone;
- FBS
fetal bovine serum;
- β-gal
β-galactosidase;
- HDAC
histone deacetylase;
- LUC
luciferase;
- MMTV
mouse mammary tumor virus;
- NAD
nicotinamide adenosine dinucleotide;
- NAM
nicotinamide;
- NCoR
nuclear receptor corepressor;
- PMSF
phenylmethylsulfonylfluoride;
- Pol II
polymerase II;
- PSA
prostate-specific antigen;
- RNAi
RNA interference;
- siRNA
small interfering RNA;
- SIRT1
sirtuin 1;
- SMRT
silencing mediator of retinoic acid and thyroid hormone receptor;
- wt
wild type.
