Carcinogenesis vol.29 no.9 pp.1725–1733, 2008
doi:10.1093/carcin/bgn117
Advance Access publication May 16, 2008
Prohibitin and the SWI/SNF ATPase subunit BRG1 are required for effective androgen
antagonist-mediated transcriptional repression of androgen receptor-regulated genes
Yan Dai, Duyen Ngo, Johanna Jacob, Lora W.Forman
and Douglas V.Faller
Department of Medicine, Cancer Research Center, Boston University School
of Medicine, Boston, MA 02118, USA
To whom correspondence should be addressed. Tel: þ1 617 638 5650;
Fax: þ1 617 638 5609;
Email: yandai@bu.edu
Correspondence may also be addressed to Douglas V.Faller.
Tel: þ1 617 638 4173; Fax: þ1 617 638 4176;
Email: dfaller@bu.edu
Introduction
Prostate cancer is the second leading cause of cancer death in men
(1,2). Androgen ablation and blockade of androgen actions through
the androgen receptor (AR) remain the mainstay of treatment for
advanced prostate cancer (3,4). While initial responses to androgen
deprivation are the norm, most tumors eventually recur in what is
termed an androgen-independent (refractory) state (5–7).
The AR, a hormone-dependent transcription factor, plays a major
role in promoting the development and progression of prostate cancer
(8–11). The transcriptional activity of the AR is modulated by nuclear
coregulatory proteins, designated coactivators and corepressors (12–
Abbreviations: AR, androgen receptor; ChIP, chromatin immunoprecipitation; DHT, dihydrotestosterone; ER, estrogen receptor; FBS, fetal bovine serum; HDAC, histone deacetylase; mRNA, messenger RNA; NCoR, nuclear
corepressor; PCR, polymerase chain reaction; PSA, prostate-specific antigen;
PSA-LUC, prostate-specific antigen promoter-driven luciferase reporter; SiRNA,
short-inhibitory RNA.
Published by Oxford University Press 2008.
Materials and methods
Cells, plasmids and antibodies
LNCaP and DU145 cells were obtained from the American Type Culture
Collection (Manassas, VA) and were maintained in RPMI 1640 medium with
10% fetal bovine serum (FBS) or in 10% charcoal-treated FBS (HyClone, Logan
CO). The prohibitin, BRG1 and BRM expression vectors (pcDNA3.1-prohibitin,
BRG1, BRM) and RNAi vectors (pRNAT-U6.1/Neo-Prohibitin, BRG1, BRM)
were generously provided by Dr S.Wang (Boston University School of
Medicine). The reporter plasmid prostate-specific antigen promoter-driven
1725
Downloaded from http://carcin.oxfordjournals.org/ at Pennsylvania State University on March 6, 2014
Androgen antagonists or androgen deprivation are the primary
therapeutic modalities 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 androgen antagonists inhibit
prostate cancer proliferation are not fully defined. In this study,
we identify two molecules which are required for effective prostate cancer cell responsiveness to androgen antagonists. We establish that androgen receptor (AR)-dependent transcriptional
suppression by androgen antagonists requires the tumor suppressor prohibitin. This requirement for prohibitin was demonstrated
using structurally-distinct androgen antagonists, stable and transient knockdown of prohibitin and transfected and endogenous
AR-responsive genes. The SWI–SNF complex core ATPase BRG1,
but not its closely-related counterpart ATPase BRM, is required
for this repressive action of prohibitin on AR-responsive promoters. Androgen antagonists induce recruitment of prohibitin
and BRG1 to endogenous AR-responsive promoters and induce
a physical association between AR and prohibitin and BRG1. The
recruitment of prohibitin to endogenous AR-responsive promoters is dependent upon antagonist-bound AR. Prohibitin binding in the prostate-specific antigen (PSA) promoter results in
the recruitment of BRG1 and the dissociation of p300 from the
PSA promoter. These findings suggest that prohibitin may
function through BRG1-mediated local chromatin remodeling activity and the removal of p300-mediated acetylation to produce
androgen antagonist-mediated transcriptional repression. Furthermore, in addition to its necessary role in AR-mediated transcriptional repression, we demonstrate that prohibitin is required
for full and efficient androgen antagonist-mediated growth suppression of prostate cancer cells.
14). Upon activation by ligands such as dihydrotestosterone (DHT),
the AR translocates to the nucleus, whereupon it binds to androgen
response elements on target genes and regulates their transcription.
Binding of androgens induces recruitment of coactivators such as SRC-1,
p300, CBP and pCAF, which contain intrinsic histone acetylase activity
(15–18). In contrast, binding of AR antagonists induces the AR to form
a complex with corepressors, such as nuclear corepressor (NcoR), SMRT
and histone deacetylase (HDAC)-1, -2 and -3 (19–22). A complete understanding of the regulation of AR antagonist-mediated transcriptional
repression is of critical importance for the design and development of
novel endocrine-based therapies and identification of pharmaceutical
targets for treating refractory prostate cancer.
Prohibitin (PHB1) is a highly-conserved protein with homologues
in all animal species, yeasts and plants. Prohibitin is recognized as
a potential tumor suppressor protein and inhibits cellular proliferation
(23,24). Prohibitin interacts with transcriptional repressor proteins,
such as NCoR, HDAC1 and retinoblastoma to repress activation of
genes regulated by the E2F family of transcription factors (25). Prohibitin protein levels have been reported to be downregulated in the
LNCaP prostate cancer cell line following growth stimulation with the
androgen DHT (26), and overexpression of prohibitin represses
androgen-dependent prostate-specific antigen (PSA) gene transcription and growth in prostate cancer cells (27). In some circumstances,
prohibitin acts in concert with the SWI–SNF complex to repress
transcriptional activity (28–30). The human SWI–SNF complex is
a ubiquitous multimeric complex that regulates gene expression by remodeling nucleosomal structure in an ATP-dependent manner (31–33).
The SWI–SNF complex contains one of two core ATPases, BRG1
or BRM. These complexes can interact with sequence-specific transcription factors to either promote or repress target gene activation,
depending upon promoter context and the protein subunits comprising the complex (34,35). SWI/SNF activity has been shown to be
essential for glucocorticoid receptor (36,37) and estrogen receptor
(ER) transcriptional activation (38,39), and BRM, but not BRG1,
appears to required for AR-dependent transcriptional activation
(40). Our previous studies have demonstrated that prohibitin can coordinate with BRG1 and BRM to regulate transcription factor activity
and that recruitment of prohibitin, and then both BRG1 and BRM,
inhibits E2F activation and causes growth arrest in breast cancer cells
treated with estrogen antagonists (28,29).
In the present study, we establish that efficient AR-dependent transcriptional suppression by androgen antagonists required prohibitin
and BRG1, but not BRM (in contrast to AR-dependent transcriptional
activation). We find that androgen antagonists induce recruitment of
prohibitin and BRG1 to AR-responsive promoters and induce a physical interaction among AR and prohibitin and BRG1. The recruitment
of BRG1 to, and the dissociation of p300 from, the AR complex at
responsive promoters is prohibitin-dependent, suggesting that prohibitin may act through BRG1-mediated chromatin remodeling activity
and inhibition of p300-mediated acetylation to function in androgen
antagonist-mediated transcriptional suppression. Furthermore, we
find that prohibitin is required for full and efficient androgen antagonist-mediated growth suppression of prostate cancer cells.
Y.Dai et al.
luciferase reporter (PSA-LUC), containing the luciferase gene under the control of a fragment of the human PSA gene promoter, was provided by Dr
A.O.Brinkmann (Erasmus University Medical Center, The Netherlands).
Antibody to prohibitin was obtained from NeoMarkers, Fremont, CA (Cat#
MS-261-p1). Antibodies to the androgen receptor (PG-21, Cat# 06-680), BRG1
(H-88) and p300 (N-15) were purchased from Santa Cruz Biotechnology (Santa
Cruz, CA). Antibody to BRM (Cat# 610389) was from (BD Biosciences, San Jose,
CA). DHT (A-8380) was purchased from (Sigma–Aldrich, St Louis, MO). Bicalutamide (Casodex) (Cat# s210183) was provided by AstraZeneca, London, UK.
Real-time reverse transcription–polymerase chain reaction
Total cellular RNA was isolated using Trizol reagent (Invitrogen Life Technologies, Carlsbad, CA) according to the manufacturer’s instructions. A two-step
reverse transcription–polymerase chain reaction (PCR) method was employed
to synthesize single-stranded cDNA (SuperScriptTM III First Strand kit, Invitrogen, 18080-051). Target genes were analyzed by real-time PCR using an
Applied Biosystems 7500 Fast Real-Time PCR System with SYBR Green I
dye (Applied Biosystems, Foster City, CA, 4309155). The primers used were
as follows: PSA forward, TGCCCACTGCATCAGGAACA; PSA reverse,
GTCCAGCGTCCAGCACACAG; KLK2 forward, CCTGGCAGGTGGCTGTGTAC; KLK2 reverse, TGTGCCGACCCAGCCA; b-actin forward, GAGAAAATCTGGCACCACACC; b-actin reverse, ATACCCCTCGTAGATGGGCAC. PCRs were performed in triplicate. b-Actin messenger RNA (mRNA)
abundance was analyzed in each sample. The PSA mRNA levels were normalized to the b-actin mRNA level. The specificities of the reverse transcription–
PCR products were monitored by melting curve analysis and also verified by
agarose gel electrophoresis.
Chromatin immunoprecipitation
LNCaP cells (107 cells) were grown in RPMI 1640 (Gibco BRL, Gaithersburg,
MD) supplemented with 10% charcoal–dextran-stripped FBS. After 3 days of
cultivation, cells were treated with DHT or bicalutamide for 4 or 24 h, then
trypsinized and washed with PBS. The harvested cells were cross-linked with
1% formaldehyde. The cell pellets were then resuspended in lysis buffer and
sonicated four times for 10 s each time at 20% power (Fisher Sonic Dismembrator, Model 550). Immunoprecipitations were performed for 1 h at room
temperature with specific antibodies (anti-prohibitin, anti-AR, anti-PolII, antiBRG1, anti-BRM, anti-p300 or the irrelevant control antibody anti-RAG-1).
Protein A/G sepharose was added, and the incubation was continued overnight
at 4°C. The sepharose beads were washed sequentially in Wash I, Wash II, Wash
III and TE buffer, and then extracted with 1% SDS, 0.1 M NaHCO3 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 and the DNA fragments were purified.
Triplicate PCRs for each sample were performed and each chromatin immunoprecipitation (ChIP) assay was performed on samples from at least three independent experiments. The primer sequences were as follows: PSA promoter
forward, ACAATCTCCTGAGTGCTGGTGT; PSA promoter reverse, GCAGAGGAGACATGCCCAG; distal control primer forward, TTCACCGTGTTGGCCAGG; distal control primer reverse, ATGGTGGCTCACGCCTG.
Immunoblot assays
Cells (5 106) treated under different conditions as described were lysed in 200
ll of a 1% NP-40 immunoblotting lysis buffer. The samples were separated on 8%
polyacrylamide gel electrophoresis and further analyzed after transfer by immunoblot analysis with anti-prohibitin, -BRG1, -BRM, -p300, -AR or -actin antibodies.
Results
Prohibitin is required for efficient androgen antagonist-mediated AR
transcriptional gene repression
To determine if prohibitin plays a role in androgen antagonist-mediated
AR-dependent transcriptional repression, we assessed the effect of
1726
Prohibitin is recruited to the PSA gene promoter in response to
exposure to androgen antagonists
To investigate the mechanism underlying the regulation of AR
antagonist-mediated AR-dependent gene transcriptional repression
by prohibitin, we first determined if prohibitin associates with known
AR-binding regions in the promoter of the endogenous PSA gene in
a ligand-dependent fashion. Two sets of PCR primer pairs were generated to amplify genomic fragments (150 bp in size) encompassing
the endogenous PSA promoter region or a control region distal to the
PSA gene (7 kb upstream of the start site). LNCaP cells were cultured
under androgen deprivation for 3 days, followed by treatment with
DHT or bicalutamide, an AR antagonist. ChIP assays were performed
using antibodies directed against prohibitin or the AR. Exposure to the
androgen antagonist bicalutamide significantly increased the recruitment of prohibitin to the endogenous PSA promoter region by .6fold compared with androgen-deprivation conditions (Figure 2A). In
contrast, exposure to DHT produced no statistically significant change
in prohibitin occupancy at the PSA promoter. AR recruitment to the
PSA promoter was induced by exposure to either DHT or bicalutamide as expected (Figure 2B). Binding of prohibitin or AR to a ‘control’ distal DNA region, located 7 kb upstream of the PSA gene, was
not observed (Figure 2A and data not shown). These results indicate
that prohibitin and the AR bind to the same general promoter region of
the PSA gene and that prohibitin recruitment to the promoter region is
stimulated by exposure to androgen antagonists.
Parallel ChIP assays examining PolII recruitment to the promoter as
a marker for active transcription demonstrated increases in PolII occupancy after exposure to DHT, but not after exposure to bicalutamide
(Figure 2C). PolII occupancy at the endogenous PSA promoter was inversely correlated with occupancy by prohibitin and exposure to bicalutamide uncoupled AR occupancy at the promoter from PolII recruitment.
Downloaded from http://carcin.oxfordjournals.org/ at Pennsylvania State University on March 6, 2014
Luciferase reporter assay
LNCaP cells were cultured in six-well plates in RPMI 1640 with 10% charcoaltreated FBS (HyClone Laboratories, Logan, UT) (androgen-deprivation conditions) for 3 days, then cotransfected with PSA-LUC reporter vectors (1 lg),
together with prohibitin or BRG1 or BRM expression vectors, using Lipofectamine plus (Invitrogen, Carlsbad, CA). Fresh medium was added after an
overnight transfection. Transfected cells were treated with DHT or DHT plus
bicalutamide or vehicle control for 24 h before lysis and assay for luciferase
activity (Promega Luciferase Assay System, #E1500). The relative luciferase
activities were normalized to the activity of a cotransfected b-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 standard deviations, with error bars indicating ±SEM.
altering cellular levels of prohibitin, using ectopic overexpression
of prohibitin or prohibitin knockdown by RNAi, on the regulation
of an androgen-responsive promoter–reporter vector (human PSALUC) in a human prostate cancer cell line (LNCaP). Cotransfection
of PSA-LUC with a prohibitin expression vector produced a modest
and not statistically significant suppression of DHT-stimulated PSApromoter-driven luciferase activity, but enhanced bicalutamidemediated PSA-LUC transcriptional repression by 4-fold (Figure 1A).
Conversely, prohibitin knockdown by expression of prohibitin
RNAi reversed bicalutamide-mediated PSA-LUC transcriptional
suppression (i.e., there was no statistically-significant repression
of transcription by bicalutamide when prohibitin levels were
knocked down), but resulted in no significant enhancement of DHTstimulated expression of the PSA-LUC vector (Figure 1B). In
parallel studies, using the structurally-distinct androgen antagonist
cyproterone acetate, ectopic expression of prohibitin similarly increased cyproterone acetate-mediated transcriptional repression of
PSA-LUC by 4-fold (data not shown). Conversely, prohibitin knockdown by expression of prohibitin RNAi resulted in a 2.5-fold abrogation of cyproterone acetate-mediated PSA-LUC transcriptional
repression (data not shown). Collectively, these data suggest that prohibitin plays a significant role in androgen antagonist-mediated PSA
transcriptional suppression, while having only a marginal role in regulating agonist-mediated transcription.
The effect of altering prohibitin protein levels on the transcriptional
repression of endogenous AR-responsive genes by AR antagonists was
next examined. In prohibitin partial-knockdown cell lines, bicalutamide
treatment suppressed AR-dependent PSA and KLK2 gene transcriptional induction by only 55 and 20%, respectively, whereas suppressing
PSA and KLK2 transcript levels by 80 and 90% in the control cell lines
(Figure 1C). This result suggests that inhibition of prohibitin activity
can partially release antagonist-mediated transcriptional suppression
of endogenous AR-responsive genes. Measurement of the levels of
prohibitin protein in those cells transfected with prohibitin RNAi confirmed decreased levels of prohibitin protein (Figure 1C). Collectively,
these experiments demonstrate that prohibitin is required for efficient
AR antagonist-mediated transcriptional suppression.
Prohibitin regulates AR-dependent transcription by androgen antagonists
We next determined whether the recruitment of prohibitin to the
PSA promoter by androgen antagonists required the AR. ChIP assays
were performed using anti-prohibitin antibody in DU145 cells, an
AR-negative prostate cancer cell line with lower levels of prohibitin
expression than LNCaP cells, by cotransfection of AR and prohibitin,
or transfection of prohibitin itself, in the presence of DHT or bicalutamide. Exposure to DHT did not induce prohibitin occupancy at the
PSA promoter, whether or not the AR was expressed (Figure 2D).
Bicalutamide treatment did not induce prohibitin occupancy at the
PSA promoter in AR-negative DU145 cells. When the AR was introduced into the cells by transfection, however, exposure to bicalutamide efficiently stimulated recruitment of prohibitin to the PSA
promoter. These results suggest that antagonist-bound AR is required
for, and likely mediates, prohibitin antagonist-induced prohibitin recruitment to the PSA promoter
The SWI/SNF core ATPase BRG1 is required for androgen antagonistmediated AR transcriptional gene repression
BRM and BRG1, ATPase members of the chromatin remodeling SWI–
SNF complex, have been shown to be involved in certain types of
steroid-dependent gene transcription, and BRM is required for efficient induction of the PSA promoter by androgens (40). In contrast, no
role for BRG1 in the regulation of AR-dependent transcription has been
discovered to date. We used short-inhibitory RNA (siRNA) knockdown
to determine whether BRG1 or BRM play a necessary role in androgen
antagonist-mediated transcriptional repression. Knock down of BRG1
by a specific siRNA significantly reversed bicalutamide-induced transcriptional repression of a PSA promoter-driven reporter (Figure 3A).
In contrast, cotransfection of a control (empty) siRNA vector, or of
a BRM-specific siRNA, had no effect on the transcriptional repression
induced by bicalutamide. We next examined whether BRG1 or BRM is
recruited to the PSA promoter in response to exposure to an androgen
antagonist. As shown in Figure 3B, amplification of DNA coprecipitated with chromatin complexes by anti-BRG1 antibodies demonstrated
an androgen antagonist-dependent recruitment of BRG1 to the endogenous PSA promoter within the same time frame during which transcription is repressed (as determined by prohibitin association and PolII
dissociation). In contrast, exposure to androgen antagonists did not
induce recruitment of BRM. Rather, BRM recruitment was selectively
induced only by exposure to androgen (DHT).
1727
Downloaded from http://carcin.oxfordjournals.org/ at Pennsylvania State University on March 6, 2014
Fig. 1. Prohibitin is required for androgen antagonist-mediated AR-dependent gene transcriptional repression. (A) Effects of prohibitin overexpression on
bicalutamide-mediated PSA promoter-luciferase reporter (PSA-LUC) transcriptional repression. LNCaP cells were cotransfected with PSA-LUC and an empty
control vector, or a wild-type-prohibitin expression vector, then treated with vehicle alone (starvation), DHT (10 nM) or DHT plus bicalutamide (DHT þ CDX) for
24 h, and cell lysates were harvested for luciferase assay. (B) Effects of prohibitin knockdown on bicalutamide-mediated PSA-LUC transcriptional repression.
LNCaP cells were cotransfected with PSA-LUC and a vector expressing prohibitin-specific siRNA (siProhibitin) or the empty siRNA expression vector, then
treated as in panel A, and cell lysates were harvested for luciferase assay. (C) Transcriptional repression of the endogenous AR-dependent PSA (left panel) and
KLK2 (right panel) genes by bicalutamide require prohibitin. LNCaP cells were cotransfected with a vector expressing prohibitin siRNA (siProhibitin) or the
empty siRNA expression vector (control) and pcDNA3.1-zeosin. The cells were selected with zeosin (50 lg/ml) for 1 week, then cultured in media containing
charcoal-stripped serum for 3 days, and treated with vehicle (starvation), DHT (1 nM) (DHT) or DHT plus bicalutamide (DHT þ CDX) for 48 h. Levels of PSA
and KLK2 transcripts were measured by quantitative reverse transcription–PCR analysis. Transcript levels are expressed relative to b-actin transcripts and the data
are presented as a percent of the activity obtained during exposure to DHT alone (assigned the value of 100). Asterisks indicate significant differences between two
groups ( P , 0.05). Inset photograph shows the immunoblot analysis of prohibitin protein levels.
Y.Dai et al.
Downloaded from http://carcin.oxfordjournals.org/ at Pennsylvania State University on March 6, 2014
Fig. 2. Ligand-dependent recruitment of prohibitin, AR and RNA polymerase II to the PSA gene promoter region. LNCaP cells were cultured under androgendeprivation conditions (charcoal-stripped serum) for 3 days and certain cultures were then treated for 4 h with DHT (10 nM) (DHT) or bicalutamide (10 lM)
(CDX) or dimethyl sulfoxide as a vehicle control (S). ChIP assays were performed using primers which amplify the PSA promoter region (promoter) or a distal
region upstream of known PSA regulatory elements (Distal region). Immunoprecipitations were carried out using antibodies directed against prohibitin, AR, RNA
polymerase II (PolII), or an irrelevant protein (Rag1). The bound and input DNAs were analyzed by the ABI 7500 Real-Time PCR system (Applied Biosystems,
Foster City, CA) by the DDCt method. The results are presented as the relative level of the protein associated with the PSA promoter, normalized to irrelevant
control antibody and input DNA. (A) Prohibitin occupancy at the PSA promoter. Inset photograph shows ethidium-stained, PCR-amplified fragments after
separation on an agarose gel and immunoblot analysis of prohibitin protein levels. (B) AR occupancy at the PSA promoter. (C) PolII occupancy at the PSA
promoter. (D) Bicalutamide-induced recruitment of prohibitin to the endogenous PSA promoter requires the AR. DU145 cells were cultured in Dulbecco’s
modified Eagle’s medium þ 10% charcoal-treated FBS. The cells were transfected with prohibitin- and AR expression vectors or prohibitin alone and treated with
bicalutamide (CDX) or DHT. ChIP assays were performed using antibodies directed against prohibitin. The immunoblot show the AR and prohibitin expression
level in DU145 cells. Asterisks indicate significant differences between two groups ( P , 0.05).
1728
Prohibitin regulates AR-dependent transcription by androgen antagonists
Downloaded from http://carcin.oxfordjournals.org/ at Pennsylvania State University on March 6, 2014
Fig. 3. Role of BRG1 in repression of AR-dependent transcription by androgen antagonists. (A) BRG1 is required for bicalutamide-mediated transcriptional
repression. LNCaP cells were cotransfected with PSA-LUC reporter vector and a vector expressing BRG1 siRNA (siBRG1), BRM siRNA (siBRM) or the empty
siRNA expression vector (siVector:). At 24 h after transfection, the cells were treated with vehicle alone, DHT (10 nM) or DHT plus bicalutamide (CDX) for 24 h,
and cells lysates were harvested for luciferase assay. Inset photograph shows the immunoblot analysis of BRG1 and BRM protein levels. (B) LNCaP cells were
cultured in charcoal-stripped serum for 3 days and certain cultures were treated for 4 h with DHT (10 nM) (DHT) or bicalutamide (10 lM) (CDX) or dimethyl
sulfoxide as a vehicle control (S). ChIP assays were performed using antibodies directed against BRG1 or BRM. The results are presented as the relative level of
the protein associated with the PSA promoter, normalized to irrelevant control antibody and input DNA. (C) Prohibitin is required for BRG1 recruitment to the
endogenous PSA promoter. LNCaP cells were cotransfected with a vector expressing prohibitin siRNA (siProhibitin) or the empty siRNA expression vector
(siVector) and pcDNA3.1-zeosin. The cells were selected with zeosin (50 lg/ml) for 1 week, then cultured in media containing charcoal-stripped serum for 3 days
and treated with DHT (1 nM) (DHT) or bicalutamide (CDX) for 4 h. ChIP assays were performed using antibodies directed against BRG1. The results are
presented as the relative level of BRG1 protein associated with the PSA promoter, normalized to irrelevant control antibody and input DNA. Asterisks indicate
significant differences between two groups ( P , 0.05).
1729
Y.Dai et al.
We have demonstrated previously that prohibitin can serve as a transcriptional repressor at certain promoters through direct recruitment
of SWI–SNF complexes (29,30), raising the possibility that the occupancy of BRG1 and prohibitin at AR-responsive promoters after exposure to androgen antagonists might be linked. We therefore
determined if the recruitment of BRG1 to AR-responsive promoters
is prohibitin dependent. As shown in Figure 3C, prohibitin knockdown reduced the recruitment of BRG1 to the PSA promoter by
bicalutamide. These data suggest that prohibitin may act through
BRG1-mediated chromatin remodeling to regulate bicalutamidemediated gene transcriptional suppression.
Prohibitin is required for efficient androgen antagonist-mediated
growth suppression of prostate cancer cell lines
Androgen antagonists, such as bicalutamide and cyproterone acetate,
suppress both AR-dependent gene transcription and prostate cancer
cell proliferation. While these activities are probably linked, a direct
causal relationship has not been established. Therefore, the necessary
role of prohibitin in each of these functional outcomes was studied
independently. A colony formation assay was carried out to assess
whether prohibitin plays a role in bicalutamide-mediated prostate
cancer cell growth suppression, using the prohibitin partial knockdown LNCaP cells and control (empty vector-transfected) LNCaP
cells. Knock down of prohibitin significantly ameliorated the sensitivity of LNCaP cells to androgen antagonists, with exposure to
bicalutamide resulting in the formation of 3-fold and 6-fold more
colonies in the prohibitin partial-knockdown cells than in the control
(empty vector) transfected cells (P , 0.05), at two concentrations of
bicalutamide (Figure 5). Interestingly, exposure to DHT resulted in
a modest but consistent increase in colony formation in the prohibitin
partial-knockdown cells compared with the control (empty vector)transfected cells, suggesting that endogenous levels of prohibitin may
exert some tonic antiproliferative effect on the cells even in the
Fig. 4. AR corepressor and coactivator complexes as a function of antagonist or agonist binding. (A) p300 occupies the PSA promoter in the presence of DHT.
LNCaP cells were cultured in charcoal-stripped serum for 3 days and certain cultures were treated for 4 h with DHT (10 nM) (DHT) or bicalutamide (10 lM)
(CDX) or dimethyl sulfoxide as a vehicle control (S). ChIP assays were performed using antibodies directed against p300. (B) Prohibitin facilitates p300
dissociation from the PSA promoter. LNCaP cells were cotransfected with a vector expressing prohibitin siRNA (siProhibitin) or the empty siRNA expression
vector (siVector) and pcDNA3.1-zeosin. The cells were selected with zeosin (50 lg/ml) for 1 week, then cultured in media containing charcoal-stripped serum for
3 days and treated with DHT (1 nM) (DHT) or bicalutamide (CDX) for 4 h. ChIP assays were performed using antibodies directed against p300. (C) Liganddependent association of antagonist-bound AR with prohibitin and BRG1 and association of DHT-bound AR with BRM and p300. Cell lysates were made from
LNCaP cells cultured in DHT and treated or not treated with bicalutamide (CDX) for 4 h. Protein complexes were immunoprecipitated with anti-AR antibody and
immunoblotted with anti-prohibitin, anti-BRG1, anti-BRM, anti-p300 or anti-AR antibodies. ‘Input’ represents the relative total amounts of protein in the lysates.
1730
Downloaded from http://carcin.oxfordjournals.org/ at Pennsylvania State University on March 6, 2014
Prohibitin enhances antagonist-mediated dissociation of p300 at an
AR-responsive promoter
AR corepressor and coactivator complexes play critical roles in regulating AR activity with precision and efficiency. Binding of androgens induces recruitment of coactivators such as p300 and CBP, which
contain intrinsic histone acetylase activity. In contrast, binding of AR
antagonists induces the AR to form a complex with corepressors, such
as HDAC-1, -2 and -3. To further understand the possible mechanism
of prohibitin-mediated transcriptional repression, we determined
whether prohibitin has a role in coactivator dissociation or on the
balance between coactivators and corepressors in the AR complex.
Analysis of p300 occupancy at the endogenous PSA promoter by
ChIP assay demonstrated that p300 occupies the PSA promoter in
the presence of DHT, but dissociates from the promoter in the presence of bicalutamide (Figure 4A). When prohibitin levels were
knocked down by siRNA, this antagonist-mediated dissociation of
p300 occupancy was partially reversed (Figure 4B). Thus, prohibitin
occupancy at the PSA promoter correlates inversely with p300 occupancy. These results suggest that prohibitin recruitment may result in
p300 dissociation from the PSA promoter, and this release of an
essential coactivator may be one mechanism contributing to prohibitinmediated transcriptional repression.
In parallel studies of AR complexes in whole-cell lysates, prohibitin
association with the AR complex in vitro was also inversely associated
with p300 binding in the presence of bicalutamide (Figure 4). Furthermore, these studies demonstrated an androgen antagonist-inducible
association of the AR and BRG1, but not BRM, with prohibitin
(Figure 4C). In contrast, in the presence of DHT, AR associated with
BRM and p300, but not BRG1 or prohibitin. This finding is consistent
with a previous report that the AR associates and colocalizes with
BRM in the presence of DHT (41).
Collectively, these results demonstrate ligand (androgen antagonist)dependent recruitment of prohibitin and BRG1 to AR-responsive promoters and are consistent with a mechanism whereby androgen
antagonist-bound AR recruits prohibitin to AR-responsive promoters
and prohibitin then facilitates recruitment of BRG1 and dissociation
of p300 from the AR complex at the promoter, thus facilitating
corepression and suppressing AR-dependent transcription.
Prohibitin regulates AR-dependent transcription by androgen antagonists
absence of antagonists, as has been observed by other investigators in
different cell types.
Together with the studies presented above, these findings demonstrate that prohibitin is required both for androgen antagonistmediated AR-dependent gene transcriptional repression and for
efficient antagonist-mediated prostate cancer cell growth suppression.
Discussion
A number of studies have elucidated some of the transcription factors
and cofactors involved in, or modifying, transcriptional activation by
the AR. These have included the SWI/SNF ATPase core subunit
BRM, but not BRG1. The factors involved in transcriptional repression have not been as well characterized, despite their clinical relevance to endocrine modalities of prostate cancer treatment. In this
study, we demonstrate that prohibitin is required for AR-responsive
gene transcriptional repression by androgen antagonists. This requirement for prohibitin was demonstrated using multiple, structurallydistinct androgen antagonists, stable and transient knockdown of
prohibitin, and transfected and endogenous AR-responsive genes. In
addition, prohibitin is required for efficient AR antagonist-mediated
prostate cancer cell growth suppression. Prohibitin is recruited to androgen response elements in endogenous gene promoters by androgen
antagonist-bound AR. Finally, we provide evidence to suggest that the
mechanism of prohibitin-mediated transcriptional inhibition on ARresponsive genes may be due to chromatin remodeling by BRG1, but
not BRM, and the dissociation of acetylase p300 from the promoter.
AR corepressor complexes play a critical role in regulating AR
activity with precision and efficiency (12). To date, several corepressors of AR have been characterized, including NCoR, SMRT, HDACs
1–3 and SIRT1 (19–21,42,43). Transcriptional repression mediated by
androgen antagonists is through repressor complex formation at the
gene promoter. Prohibitin has been shown to be involved in transcriptional suppression through formation of repressor complexes incorporating several distinct transcriptional corepressors. We and others
have shown that prohibitin interacts with proteins such as NCoR,
HDAC1 and retinoblastoma to repress activation of genes regulated
by the E2F family of transcription factors (25,44–46). More recently,
we have demonstrated that the recruitment of prohibitin and both
BRG1 and BRM inhibits E2F activation and causes growth arrest in
breast cancer cells treated with the antiestrogen tamoxifen (29). In this
study, we observed significant recruitment of prohibitin to an ARresponsive promoter during exposure to androgen antagonists and
found that prohibitin is required for androgen antagonist-mediated
transcriptional repression and growth suppression. In the absence of
androgen antagonists, however, prohibitin may also play a minor role
in the regulation of AR-dependent transcription. Gamble et al. (27)
have shown that overexpression of prohibitin can suppress androgeninduced transcription. Constitutive expression of AR and prohibitin in
the naturally androgen-unresponsive cells lines, Cos-7, MCF-7 and
PC3, significantly suppressed androgen-induced transcription of cotransfected reporter genes. However, in the androgen-responsive cell
line LNCaP, knock down of endogenous prohibitin resulted only in
a marginal increase in endogenous androgen-dependent gene transcription, consistent with our finding’s results.
We found that the BRG1 associates with prohibitin and AR in an
androgen antagonist-dependent fashion in cell lysates, and that
BRG1 recruitment to the PSA promoter is dependent on the prohibitin, raising the possibility that the prohibitin suppresses AR-dependent transcription through BRG1-mediated chromatin remodeling in
response to androgen antagonists. SWI–SNF complex-mediated remodeling of chromatin is a well-documented mechanism of transcriptional regulation and appears to play a role in transcriptional
activation of AR-responsive and ER-responsive genes. Using the
AR-deficient and BRM/BRG1-deficient SW13 cell line, Marshall
et al. (40) have shown that BRM, but not BRG1, is required for
DHT-mediated transcriptional activation of a transfected PSALUC reporter. In contrast, we and others have shown that BRG1
(38) and BRM (28,40) are both required for ligand-dependent activation of ER-responsive genes. Interestingly, the SWI–SNF complex can also participate in transcriptional repression. We have
recently demonstrated that BRG1 and BRM are each required for
both activation and suppression of ER-dependent gene transcription
in response to agonists and to antagonists (30). Estrogen and its
antagonists both induce the recruitment of BRG1/BRM to ERresponsive promoters, but recruitment of BRG1/BRM induced by
antagonists is dependent upon prohibitin, which is not the case for
BRG1/BRM recruitment in response to estrogen (an agonist) (30).
1731
Downloaded from http://carcin.oxfordjournals.org/ at Pennsylvania State University on March 6, 2014
Fig. 5. Prohibitin is required for efficient antagonist-mediated prostate cell growth suppression. (A) wild-type-LNCaP cells or prohibitin partial-knockdown
LNCaP cells were cultured in charcoal-stripped serum for 3 days, and the cells were split 1:10 and cultured with DHT (1 nM; control) or DHT (1 nM) plus
bicalutamide (CDX) at the indicated concentrations for 2 weeks. Cells were fixed and stained. Colonies consisting of 20 or more cells were scored. Each
experiment was performed in triplicate. (B) Immunoblotting studies to demonstrate prohibitin protein levels. (C) Representative plates of siVector- or siProhibitintransfected cells treated with 10 or 25 lM CDX.
Y.Dai et al.
1732
genes required for entry into S phase, we speculate that the upregulation
of prohibitin induced by androgen antagonists over time may also
suppress cell proliferation by downregulation of E2F activity, thus
reenforcing its more immediate repressive effects on AR-dependent
gene induction.
Funding
American Cancer Society (IRG-72-001-27-IRG to Y.D.); BUSM Department of Medicine Pilot Project Grant Award (to Y.D.); National
Cancer Institute (CA101992 to D.V.F.); Karin Grunebaum Cancer
Research Foundation (to D.V.F.)
Acknowledgements
We thank Dr A.O.Brinkmann (Erasmus University Medical Center, The Netherlands) and Drs Sheng Wang and Christina N.Mork (Boston University
School of Medicine) for reagents.
Conflict of Interest Statement: None declared.
References
1. Chen,C.D. et al. (2004) Molecular determinants of resistance to antiandrogen therapy. Nat. Med., 10, 33–39.
2. Boring,C.C. et al. (1994) Cancer statistics, 1994. CA Cancer J. Clin., 44, 7–26.
3. Sharifi,N. et al. (2006) Androgen receptor as a therapeutic target for androgen independent prostate cancer. Am. J. Ther., 13, 166–170.
4. Roscigno,M. et al. (2005) Medical therapy of prostate cancer. Minerva
Urol. Nefrol., 57, 71–84.
5. Hobisch,A. et al. (2006) The androgen receptor pathway is by-passed in
prostate cancer cells generated after prolonged treatment with bicalutamide. Prostate, 66, 413–420.
6. Nishiyama,T. et al. (2006) Stepping-stones to the further advancement of
androgen-deprivation therapy for prostate cancer. Expert Rev. Anticancer
Ther., 6, 259–268.
7. Isaacs,J.T. (1999) The biology of hormone refractory prostate cancer. Why
does it develop? Urol. Clin. North Am., 26, 263–273.
8. Brinkmann,A.O. et al. (1999) Mechanisms of androgen receptor activation
and function. J. Steroid Biochem. Mol. Biol., 69, 307–313.
9. Heinlein,C.A. et al. (2004) Androgen receptor in prostate cancer. Endocr.
Rev., 25, 276–308.
10. Hirawat,S. et al. (2003) The androgen receptor: structure, mutations, and
antiandrogens. Cancer Invest., 21, 400–417.
11. Chang,C.S. et al. (1988) Molecular cloning of human and rat complementary DNA encoding androgen receptors. Science, 240, 324–326.
12. Wang,L. et al. (2005) Androgen receptor corepressors: an overview.
Prostate, 63, 117–130.
13. Culig,Z. et al. (2004) Expression and function of androgen receptor coactivators in prostate cancer. J. Steroid Biochem. Mol. Biol., 92, 265–271.
14. Heinlein,C.A. et al. (2002) Androgen receptor (AR) coregulators: an overview. Endocr. Rev., 23, 175–200.
15. Chakravarti,D. et al. (1996) Role of CBP/P300 in nuclear receptor signalling. Nature, 383, 99–103.
16. Chen,H. et al. (1997) Nuclear receptor coactivator ACTR is a novel histone
acetyltransferase and forms a multimeric activation complex with P/CAF
and CBP/p300. Cell, 90, 569–580.
17. Li,J. et al. (2000) p300 requires its histone acetyltransferase activity and
SRC-1 interaction domain to facilitate thyroid hormone receptor activation
in chromatin. Mol. Cell. Biol., 20, 2031–2042.
18. Spencer,T.E. et al. (1997) Steroid receptor coactivator-1 is a histone acetyltransferase. Nature, 389, 194–198.
19. Nagy,L. et al. (1997) Nuclear receptor repression mediated by a complex
containing SMRT, mSin3A, and histone deacetylase. Cell, 89, 373–380.
20. Liao,G. et al. (2003) Regulation of androgen receptor activity by the
nuclear receptor corepressor SMRT. J. Biol. Chem., 278, 5052–5061.
21. Hodgson,M.C. et al. (2005) The androgen receptor recruits nuclear receptor
CoRepressor (N-CoR) in the presence of mifepristone via its N and
C termini revealing a novel molecular mechanism for androgen receptor
antagonists. J. Biol. Chem., 280, 6511–6519.
22. Zhu,P. et al. (2006) Macrophage/cancer cell interactions mediate hormone
resistance by a nuclear receptor derepression pathway. Cell, 124, 615–629.
Downloaded from http://carcin.oxfordjournals.org/ at Pennsylvania State University on March 6, 2014
Consistent with a previous report which used androgen non-responsive
cells rendered responsive by ectopic expression of an AR and BRM
(40), we found, using mRNA knockdown in androgen-responsive
cells, that depletion of BRM, but not BRG1, strongly reduces DHTinduced transcriptional activation of androgen-responsive reporter
genes and of endogenous cellular genes (Y.Dai. and D.V.Faller, unpublished data). DHT-bound AR can physically associate with BRM
and p300 in the presence of DHT. More interestingly, our data presented herein conversely demonstrate that BRG1, but not BRM, is
required for androgen antagonist-mediated transcriptional suppression of AR-responsive genes. BRG1 is recruited to the endogenous
PSA gene promoter by androgen antagonists, whereas BRM is recruited to the PSA gene promoter by agonists, such as DHT. These
results suggest that BRG1 and BRM function differentially in agonistdependent transcriptional activation, versus antagonist-mediated
transcriptional suppression, of AR-responsive genes, in distinct contrast to their joint requirements for ER-dependent gene regulation.
As discussed above, prohibitin requires association with one or
more corepressor proteins, such as HDAC1, to suppress local gene
transcription and appears to serve as multivalent recruiter of corepressors rather than possessing any intrinsic corepressor activity. In spite
of a demonstrable physical interaction of prohibitin with HDAC1 in
a breast cancer cell line, Gamble et al. (27) utilized a general
HDAC inhibitor to suggest that HDAC activity is not required for
AR-responsive gene repression initiated by overexpression of prohibitin. In contrast, we show here that BRG1, a known binding partner
of prohibitin, is recruited to the androgen response element complex
in an antagonist-dependent fashion and is required for transcriptional
repression. Thus, BRG1 may play the major corepressor role for the
transcriptional repression of AR-responsive genes by prohibitin. It is
also noteworthy that exposure to androgen antagonists reduces p300
occupancy at AR-responsive promoters and that the absence of prohibitin reduces this effect of antagonists, suggesting that, in addition
to facilitating recruitment of BRG1, prohibitin may also act through the
removal of p300 acetylase activity to suppress androgen antagonistmediated transcriptional suppression. We show that suppression of
prohibitin or BRG1 protein levels reverses, but does not completely
abrogate, the repressive effects of androgen antagonists on ARdependent transcription. One potential explanation for this is that,
although very effective, siRNA techniques do not completely deplete
all prohibitin or BRG1 protein from the cell, and the small amount of
protein remaining may allow for some repression by antagonists.
Alternatively, there may be recruitment of other corepressors induced
by exposure to androgen antagonists. Indeed, we have recently reported that the class III HDAC SIRT1 is recruited to the promoter of
AR-responsive genes after exposure to androgen antagonists and contributes independently to transcriptional repression (42).
Prohibitin and BRG1 are not DNA-binding proteins, but can be
recruited to promoter regions by other transcription factors. We have
shown a ligand (androgen antagonist)-dependent physical association
of AR and prohibitin and prohibitin and BRG1, and that BRG1 recruitment to endogenous AR-responsive promoters is reduced when
prohibitin is knocked down, suggesting that recruitment of prohibitin
to an AR-dependent promoter could lead to prohibitin-dependent recruitment of BRG1. Our analysis of AR-responsive promoters in
DU145 cells, expressing or not expressing an AR protein, demonstrates that recruitment of prohibitin to the PSA promoter is androgen
antagonist-bound AR-dependent.
Collectively, our data demonstrate a critical role for prohibitin in
the transcriptional suppression of AR-mediated genes, and in the
growth suppression of prostate cancer cell lines, mediated by androgen antagonists. Repression of AR-mediated transcriptional activity is
therefore likely to be one mechanism by which prohibitin inhibits
androgen-dependent growth of prostate cells. It is interesting to note,
however, that we observed significant induction of cellular prohibitin
levels following longer exposures (72 h) to bicalutamide (Y.Dai. and
D.V.Faller, unpublished data). As prohibitin is well-established to
interact with the E2F family of transcription factors and to suppress
cell proliferation through repression of E2F-mediated transcription of
Prohibitin regulates AR-dependent transcription by androgen antagonists
36. Wallberg,A.E. et al. (2000) Recruitment of the SWI-SNF chromatin remodeling complex as a mechanism of gene activation by the glucocorticoid
receptor tau1 activation domain. Mol. Cell. Biol., 20, 2004–2013.
37. Bilodeau,S. et al. (2006) Role of Brg1 and HDAC2 in GR trans-repression
of the pituitary POMC gene and misexpression in Cushing disease. Genes
Dev., 20, 2871–2886.
38. DiRenzo,J. et al. (2000) BRG-1 is recruited to estrogen-responsive promoters and cooperates with factors involved in histone acetylation. Mol.
Cell. Biol., 20, 7541–7549.
39. Belandia,B. et al. (2002) Targeting of SWI/SNF chromatin remodelling
complexes to estrogen-responsive genes. EMBO J., 21, 4094–4103.
40. Marshall,T.W. et al. (2003) Differential requirement of SWI/SNF for androgen receptor activity. J. Biol. Chem., 278, 30605–30613.
41. Klokk,T.I. et al. (2007) Ligand-specific dynamics of the androgen receptor
at its response element in living cells. Mol. Cell. Biol., 27, 1823–1843.
42. Dai,Y. et al. (2007) Sirt1 is required for antagonist-induced transcriptional
repression of androgen-responsive genes by the androgen receptor. Mol.
Endocrinol, 26, 1807–1821.
43. Fu,M. et al. (2006) Hormonal control of androgen receptor function
through SIRT1. Mol. Cell. Biol., 26, 8122–8135.
44. Wang,S. et al. (1999) Prohibitin, a potential tumor suppressor, interacts
with RB and regulates E2F function. Oncogene, 18, 3501–3510.
45. Fusaro,G. et al. (2002) Differential regulation of Rb family proteins and
prohibitin during camptothecin-induced apoptosis. Oncogene, 21, 4539–
4548.
46. Wang,S. et al. (1999) Rb and prohibitin target distinct regions of E2F1 for
repression and respond to different upstream signals. Mol. Cell. Biol., 19,
7447–7460.
Received August 25, 2007; revised March 23, 2008; accepted May 6, 2008
1733
Downloaded from http://carcin.oxfordjournals.org/ at Pennsylvania State University on March 6, 2014
23. Rastogi,S. et al. (2006) Prohibitin facilitates cellular senescence by recruiting specific corepressors to inhibit E2F target genes. Mol. Cell. Biol., 26,
4161–4171.
24. Joshi,B. et al. (2003) A putative coiled-coil domain of prohibitin is sufficient to repress E2F1-mediated transcription and induce apoptosis. Biochem. Biophys. Res. Commun., 312, 459–466.
25. Wang,S. et al. (2002) Prohibitin co-localizes with Rb in the nucleus and recruits
N-CoR and HDAC1 for transcriptional repression. Oncogene, 21, 8388–8396.
26. Gamble,S.C. et al. (2004) Androgens target prohibitin to regulate proliferation of prostate cancer cells. Oncogene, 23, 2996–3004.
27. Gamble,S.C. et al. (2007) Prohibitin, a protein downregulated by androgens, represses androgen receptor activity. Oncogene, 26, 1757–1768.
28. Wang,S. et al. (2004) BRG1/BRM and prohibitin are required for growth
suppression by estrogen antagonists. EMBO J., 23, 2293–2303.
29. Wang,S. et al. (2002) Prohibitin requires Brg-1 and Brm for the repression
of E2F and cell growth. EMBO J., 21, 3019–3028.
30. Zhang,B. et al. (2007) Reprogramming of the SWI/SNF complex for coactivation or co-repression in prohibitin-mediated estrogen receptor regulation. Oncogene, 26, 7153–7157.
31. Aoyagi,S. et al. (2002) Nucleosome remodeling by the human SWI/SNF
complex requires transient global disruption of histone-DNA interactions.
Mol. Cell. Biol., 22, 3653–3662.
32. Havas,K. et al. (2000) Generation of superhelical torsion by ATPdependent chromatin remodeling activities. Cell, 103, 1133–1142.
33. Schnitzler,G. et al. (1998) Human SWI/SNF interconverts a nucleosome
between its base state and a stable remodeled state. Cell, 94, 17–27.
34. Sif,S. et al. (2001) Purification and characterization of mSin3A-containing
Brg1 and hBrm chromatin remodeling complexes. Genes Dev., 15, 603–618.
35. Kadam,S. et al. (2000) Functional selectivity of recombinant mammalian
SWI/SNF subunits. Genes Dev., 14, 2441–2451.