Early Growth Response Gene 1 Modulates Androgen Receptor Signaling in Prostate Carcinoma Cells*

The transcription factor early growth response gene 1 (EGR1) has been implicated in diverse roles in the regulation of cell growth, apoptosis, and differentiation. Previous studies suggest that the effects of EGR1 on tumorigenesis are critically dependent on the cellular context. In a majority of prostate cancers, EGR1 is overexpressed and promotes prostate tumor progression. In contrast, in other tumor types such as breast cancers and glioblastomas, EGR1 is expressed at low levels and when overexpressed can inhibit tumor growth. To explore the role of EGR1 in prostate tumorigenesis, we examined the impact of EGR1 expression on the androgen receptor (AR) signaling pathway. We show here that EGR1 binds to the AR in prostate carcinoma cells, and an EGR1-AR complex can be detected by chromatin immunoprecipitation at the enhancer of an endogenous AR target gene. Overexpression of EGR1 enhanced AR-mediated transactivation, whereas EGR1 knockdown by small interfering RNA inhibited AR signaling pathway activity. Furthermore, Western blot and immunocytochemical analyses showed that constitutive overexpression of EGR1 promotes the translocation of AR from the cytoplasm to the nucleus. These results indicate that EGR1 may promote prostate cancer development by modulating the androgen receptor signaling pathway.

The early growth response gene 1 (EGR1, 1 also known as NGFI-A, Zif 268, Krox 24, and Tis 8) is a zinc finger transcription factor that belongs to a multigene family that includes EGR2, EGR3, EGR4, and WT1 (1)(2)(3)(4). EGR1 is rapidly induced by multiple growth factors and cytokines. The induction of EGR1 by external stimuli is generally transient but is sustained in a large fraction of prostate tumors (5)(6)(7)(8), suggesting that EGR1 plays a critical role in the initiation and/or progression of prostate cancer. Analysis of two different prostate cancer-prone transgenic mice (TRAMP and CR2TAg) indicates that deletion of Egr1 significantly delays prostate tumor pro-gression (9). EGR1 overexpression in prostate cancer cells can up-regulate some growth factors, such as insulin-like growth factor-II, transforming growth factor-␤1, and platelet-derived growth factor-A, which have previously been implicated in enhancing tumor progression (7). These results were confirmed by two recent studies. In one study, it was shown that Egr1 promotes cell growth and cell cycle progression and increases cell resistance to apoptotic signals by up-regulating cyclin D2 and inhibiting p19 ink4d and FAS expression in mouse prostate cancer cells (10). Another study shows that inhibition of Egr1 expression could reverse the transformation of prostate carcinoma cells in vitro and in vivo (11). However, in contrast to the pro-tumorigenic effects of EGR1 observed in prostate cancer, in other tumor types such as breast cancers and glioblastomas/ astrocytomas, EGR1 expression is often absent or reduced and, when re-expressed, results in growth suppression (12)(13)(14). In sum, the available evidence strongly suggests that the effects of EGR1 on tumor progression are critically dependent on the cellular context.
We tested the hypothesis that interaction between EGR1 and the androgen receptor (AR) signaling pathway plays a role in the context-dependent pro-tumorigenic role of EGR1 in prostate tumorigenesis. Androgen receptor function plays a pivotal role in normal prostate development and physiology as well as prostate tumorigenesis. The transcriptional activity of the AR is modulated by a large and growing number of interacting proteins that function as coactivators and corepressors (15). We found that EGR1 functionally interacts with the androgen receptor in human prostate carcinoma cells. Overexpression and siRNA knockdown experiments indicate that EGR1 expression enhances AR-mediated activation of AR target genes. In addition, EGR1 enhanced the translocation of AR from the cytoplasm to the nucleus. Thus EGR1 overexpression may promote prostate tumorigenesis at least in part by modulating androgen receptor function.
Cell Culture and Stable Transfectants-The androgen receptor-positive human prostate cancer cell line LNCaP and the androgen-independent human prostate cancer cell line DU145 were obtained from ATCC and maintained in RPMI 1640 supplemented with 5% FBS and Dulbecco's modified Eagle's medium with 5% FBS, respectively. For the generation of LNCaP cell line stably expressing EGR1, LNCaP cells were plated overnight in 60-mm dishes. The cells were transfected with pcDNA3.1-EGR1 or pcDNA3.1 control vector using FuGENE 6 transfection reagent (Roche Applied Science) according to the manufacturer's protocol. The cells were subjected to selection using 800 g/ml G418 (Mediatech) 48 h after transfection. The resistant clones were pooled 10 -14 days later, and the cells were maintained in RPMI 1640 containing 600 g/ml G418.
Western Blot Analysis-Whole cells lysates were prepared using extraction buffer (50 mM Tris-HCl-buffered saline, pH 7.4, 1% Triton X-100, 1% Nonidet P-40, 5 mM CaCl 2 , 2 mM phenylmethylsulfonyl fluoride, and 3 mM hydrogen peroxide). Nuclear protein was extracted by the method of Dignam et al. (19). Protein concentrations of supernatants were measured using the Bio-Rad D C protein assay reagent. Extracts containing 10 -20 g of protein were electrophoresed on a 12% SDS-polyacrylamide gel and blotted onto an Immobilon TM membrane (Millipore). The blotted membrane was treated with 5% fat-free dry milk at 4°C overnight and incubated for 2 h at room temperature with the antibodies described below. The membrane was then incubated for 1 h at room temperature with a peroxidase-labeled goat anti-mouse or goat anti-rabbit antibody (Bio-Rad). The membrane was rinsed, treated with ECL reagent (PerkinElmer Life Sciences) for 1 min and exposed to x-ray film at room temperature for 1-3 min. The following antibodies from Santa Cruz Biotechnology were used: rabbit anti-human EGR1 antibody (1:500), mouse anti-human AR N-terminal antibody (1:500), and goat anti-actin antibody (1:1000).
Immunoprecipitation-For immunoprecipitation experiments, DU145 cells were grown in medium with 5% charcoal/dextran-stripped FBS (HyClone) for 24 h and then transfected with pCMV-hAR vector with FuGENE 6 transfection reagent. The cells were treated with 10 nM 5␣-dihydrotestosterone (DHT; Sigma) 24 h after transfection. For LN-CaP cells, the cells were grown in RPMI 1640 containing 5% charcoal/ dextran-stripped FBS for 3 days. The cells were treated with 10 nM DHT or vehicle for 1 h before harvest. The cells were lysed in 0.5 ml of extraction buffer. The lysates were incubated on ice for 30 min and subsequently cleared by centrifugation at 12,000 rpm for 15 min at 4°C. The protein concentration of the lysates was determined by Bio-Rad DC protein assay reagent. 10 l of protein A/G Plus-agarose beads (Santa Cruz) were added to 1.5-ml microtubes containing 250 l of cellular lysate (1 mg protein/ml) and rotated at 4°C for 1 h. The samples were centrifuged for 1 min at 2000 rpm. Primary antibodies (anti-EGR1, anti-cyclin E (Santa Cruz), or rabbit IgG) and 10 l of protein A/G Plus-agarose beads were added to the supernatant, and the mixture was rotated overnight at 4°C. The beads were pelleted by gentle centrifugation and washed three times with 1 ml of ice-cold extraction buffer. After the final wash, the precipitated protein complexes were resuspended in SDS sample loading buffer and boiled for 5 min. The samples were centrifuged after vortexing, and the supernatants were analyzed by Western blotting.
Production of GST-AR Fusions-BL21 cells (Amersham Biosciences) transformed with pGEXGST-AR1-562, pGEXGST-AR544 -634, and pGEXGST-AR624 -919 plasmids (17) were inoculated on a LB-agarose plate containing ampicillin and incubated at 37°C overnight. Colonies of BL21 cells with the above plasmids were cultured in 12 ml of LB medium containing ampicillin to an A 600 value of 0.6 -0.8. GST fusion protein expression was induced by adding 1.0 mM isopropyl-1-thio-␤-Dgalactopyranoside (Amersham Biosciences). Bacterial pellets were lysed by sonication in 2 ml of ice-cold phosphate-buffered saline (PBS) containing protease inhibitor mixture. The lysates were mixed 2 h at 4°C with glutathione-Sepharose 4B (Amersham Biosciences). After beads with fusion protein were washed three times with PBS, fusion protein was collected by glutathione elution buffer (Amersham Biosciences) and detected by Western blot using an anti-GST antibody (Santa Cruz).
GST Pull-down Assay-Beads with the above GST-AR fusion protein were incubated with nuclear lysates from LNCaP cells. GST-AR-EGR1 complexes were washed with nuclear protein extraction buffer three times, eluted with SDS sample loading buffer, and then analyzed by Western blot.
Dual Luciferase Gene Reporter Experiment-DU145 cells (2.5 ϫ 10 5 ) were plated overnight in 6-well plates in phenol red-free Dulbecco's modified Eagle's medium with 5% charcoal/dextran-stripped FBS. 24 h after transfection with FuGENE 6 Transfection Reagent, the medium was replaced with fresh medium with or without 10 nM DHT. The cells were harvested 24 h later, and luciferase was measured using a Promega (Madison, WI) dual luciferase assay kit.
Quantitative RT-PCR Analysis-LNCaP-EGR1 and LNCaP-Neo control cells were cultured in phenol red-free RPMI 1640 with 5% charcoal/ dextran-stripped FBS for 3 days. The cells were treated with 10 nM DHT 90 min before harvesting. RNA was extracted with Trizol reagent (Invitrogen), and 1 g of total RNA was reverse transcribed. PCR was performed by SYBR® green PCR Master Mix (Applied Biosystems) as described using the relative standard curve method (7). The increase in fluorescence of the SYBR green dye was monitored using a GeneAmp 5700 sequence detection system (Applied Biosystems). All of the PCR reactions were performed in triplicate. The values were normalized to the relative amounts of 18 S mRNA. The sequences of primers used for PCR analyses are as follows: PSA, 5Ј-CAACCCTGGACCTCACAC-CTA-3Ј (sense) and 5Ј-GGAAATGACCAGGCCAAGAC-3Ј (antisense); 18 S, 5Ј-CGCCGCTAGAGGTGAAATTCT-3Ј (sense) and 5Ј-CGAAC-CTCCGACTTTCGTTCT-3Ј (antisense).
Chromatin Immunoprecipitation (ChIP) Assay-LNCaP cells were grown in RPMI 1640 with 5% charcoal/dextran-stripped FBS for 3 days. The cells were then treated with 10 nM DHT for 1 h and cross-linked with 1% formaldehyde at 37°C for 10 min. The cells were then rinsed three times with ice-cold PBS, collected into PBS with protease inhibitor mixture, and centrifuged for 5 min. The pellets were resuspended in 0.3 ml of lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl, pH 8.1) with protease inhibitor mixture, incubated for 10 min on ice, and sonicated six times at 10 s each at 50% input using a Branson Ultrasonics Sonicator (Danbury, CT), followed by centrifugation for 10 min. The supernatants were diluted in dilution buffer (0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl, 167 mM NaCl, pH 8.1), followed by immunoclearing with salmon sperm DNA/protein A agarose (Upstate Biotechnology, Inc.). Immunoprecipitation was performed for 2 h at 4°C with rabbit anti-human AR polyclonal antibody (Upstate Biotechnology, Inc.), rabbit anti-human EGR1 polyclonal antibody, or rabbit IgG. After immunoprecipitation, protein A-agarose was added and incubated overnight. Agarose beads were washed sequentially for 10 min each in low salt wash buffer, high salt wash buffer, and LiCl immune complex wash buffer. The beads were then washed two times FIG. 1. EGR1 physically interacts with the AR in prostate cancer cells. A, DU145 cells were transfected with AR expression plasmid, and 24 h after transfection, the medium was replaced with fresh medium with 5% charcoal/dextran-stripped FBS with (ϩ) or without (Ϫ) DHT. The cells were harvested 24 h later for immunoprecipitation (IP) using anti-EGR1 antibody or control rabbit IgG. The immunoprecipitates were subjected to immunoblotting using anti-AR antibody. Input (15%) is loaded as control. B, immunoprecipitation of AR by anti-EGR1 antibody but not control rabbit IgG in LNCaP cells cultured in medium containing regular 5% FBS. Input (15%) is loaded as control. Lanes 2 and 3 are duplicates. C, LNCaP cells were cultured in medium with 5% charcoal/dextran-stripped FBS. 1 h after treatment with 10 nM DHT, the cells were harvested for immunoprecipitation using anti-cyclin E (negative control) or anti-EGR1 antibodies. Input (5%) is loaded as control. In A-C, IgG refers to IgG heavy chain. D, LNCaP cells cultured in charcoal/dextran-stripped FBS were treated with 0, 1, 2, or 4 nM of the synthetic androgen R1881 for 24 h and then subjected to immunoprecipitation using anti-EGR1 antibody followed by AR immunoblotting. Note dose-dependent increase in complex formation. Input (15%) is loaded as control.
with TE buffer and extracted two times with 1% SDS, 0.1 M NaHCO 3 . Histone-DNA cross-links were reversed by heating at 65°C for 4 h. DNA was recovered by a Geneclean® III kit (Midwest Scientific). For PCR, 5 l of 50 l of DNA was used in 30 cycles of amplification. The AREIII primers used are 5Ј-GAGGTTCATGTTCACATTAGTACAC-3Ј (sense), and 5Ј-ATTCTGGGTTTGGCAGTGGAGTGC-3Ј (antisense).
Immunofluorescence-LNCaP-EGR1 and LNCaP-Neo cells were grown in RPMI 1640 containing 5% charcoal/dextran-stripped FBS on coverslips fixed in cold acetone for 10 min and air-dried. The cells were blocked in 4% normal serum in 0.1% Tween 20/PBS for 30 min. Primary rabbit polyclonal anti-human AR antibody was used at 5 mg/ml and incubated for 2 h at room temperature followed by three 10-min washes. Secondary antibodies conjugated to fluorophores (Molecular Probes) were used at 5 mg/ml and incubated for 1 h at room temperature followed by three 10-min washes. The coverslips were mounted with mounting medium containing 4Ј,6-diamidino-2-phenylindole (DAPI) (Vector Laboratories). The images were captured with a SPOT camera (Diagnostic Instruments) mounted on a Zeiss Axioskop 40 microscope.

EGR1 Binds to the AR in Prostate Carcinoma
Cells-In previous studies using recombinant proteins produced in bacteria, EGR1 was shown to interact specifically with the androgen receptor in vitro (20). To determine whether EGR1 interacts with AR in human prostate carcinoma cells, we performed coimmunoprecipitation experiments. The human prostate carcinoma cell line DU145, which does not express detectable levels of AR, was transfected with the pCMV-AR plasmid encoding the full-length human androgen receptor cDNA (17). As shown in (Fig. 1A), a 110-kDa band corresponding to the AR was detected in immunoprecipitates using anti-EGR1 antibody but not in those from control rabbit IgG. The interaction between EGR1 and AR was strongest in cells treated with DHT. This is probably a reflection of the fact that EGR1 is a nuclear protein, and the AR translocation to the nucleus is dependent on DHT. To investigate whether EGR1 binds to endogenous AR in prostate cancer cells, we performed coimmunoprecipitation experiments using LNCaP cells, which express both AR and EGR1. Anti-EGR1 antibody, but not control IgG, coimmunoprecipitated AR in LNCaP cells grown in medium containing normal FBS (Fig. 1B). When grown in the absence of androgens, a weaker but readily detectable interaction between EGR1 and AR was observed in LNCaP cells, and this interac- tion is further increased by treatment with DHT or the testosterone analogue R1881 (Fig. 1, C and D). An additional control coimmunoprecipitation with anti-cyclin E antibody further demonstrates the specificity of the EGR1-AR interaction (Fig.  1C).
EGR1 Binds to N-terminal Domain of AR-To determine which portion of AR interacts with EGR1, we used GST pulldown assays. We expressed and purified GST fusion proteins encoding different domains of the androgen receptor corresponding to the N-terminal domain, the DNA-binding domain, and the C-terminal domain (Fig. 2, A and B). GST pull-down experiments indicate that EGR1 interacts exclusively with the N-terminal activation domain of AR and that this interaction is independent of DHT (Fig. 2C).
EGR1 Binds to Androgen Response Elements (AREs) in the Upstream Domain of the PSA Gene-After translocation to the nucleus, AR forms a complex on the androgen response elements (AREs) present in the regulatory regions of its target genes. To investigate whether EGR1 forms a complex with AR on the AREs of a native promoter in vivo, we used ChIP assays. We examined the interaction of AR and EGR1 on the AREIII element of the well known AR target gene, PSA (Fig. 3A). LNCaP cells were grown in medium containing charcoal/dex-tran-stripped serum for at least 3 days followed by treatment with 10 nM DHT. Specific antibodies against EGR1 or AR were used to immunoprecipitate bound genomic DNA fragments. As expected, AR antibody precipitated AREIII in a DHT-dependent manner (Fig. 3B). Antibody against EGR1 also precipitated the AREIII element, indicating that EGR1 complexed with AR could interact with a native AR-target gene promoter/enhancer in vivo. Importantly, this interaction is dependent on DHT (Fig. 3B), supporting the conclusion that EGR1 does not directly bind to the AREIII but is recruited through complex formation with AR. Additional control ChIP experiments with normal rabbit IgG further demonstrate the specificity of the interactions.
EGR1 Augments AR-mediated Transcription-To assess the functional significance of the EGR1-AR interaction, we assessed the effect of EGR1 expression on AR-mediated activation of an ARE luciferase promoter-reporter construct. In DU145 cells, transfection of EGR1 and AR in the absence of DHT had a modest stimulatory effect on ARE luciferase reporter activity (Fig. 4A). As expected, DHT treatment stimulated AR activity, and this ligand-dependent transcriptional activity of AR was further enhanced significantly by EGR1 in a dose-dependent manner (Fig. 4A). Expression of EGR1 alone in the absence of AR had no effect on ARE luciferase reporter activity (Fig. 4B). These data indicate that EGR1 augments AR-mediated transcription, and this effect is mediated through the interaction between the two proteins.
EGR1 Expression Enhances Androgen-induced PSA mRNA Expression-To extend the above results to an endogenous AR target gene, we monitored the effect of EGR1 overexpression on the expression of prostate-specific antigen (PSA) mRNA in androgen-dependent LNCaP cells. We first established LNCaP cell lines stably overexpressing EGR1 (Fig. 5A). As shown in Fig. 5B, relative to untreated LNCaP-Neo cells, DHT treatment induced PSA gene expression about 9-fold in LNCaP-Neo cells and ϳ18-fold in LNCaP-EGR1 cells. Interestingly, LN-CaP-EGR1 cells consistently showed elevated expression of PSA even in the absence of DHT (Fig. 5B).
Effect of EGR1 on Nuclear Translocation of AR in LNCaP Cells-The observation that LNCaP-EGR1 cells expressed elevated levels of PSA even in the absence of androgens prompted us to examine whether EGR1 expression can promote the translocation of AR from cytoplasm to nucleus. In the absence of androgens, the AR is localized to the cytoplasm in an inactive complex that includes heat shock proteins. Upon binding to its cognate ligand, the AR undergoes a conformational change that results in a more compact and stable form of the AR. The activated AR dissociates from heat shock protein and translocates to the nucleus where it binds to consensus DNA sequences as a homodimer to influence transcription of down- stream genes (21). We observed changes in intracellular distribution of AR in LNCaP cells stably expressing EGR1 by Western blot and immunocytochemical analyses. Western blot for nuclear protein showed that most of AR protein translocated from cytoplasm to nucleus in LNCaP-Neo and LNCaP-EGR1 1 h after exposure to 10 nM DHT (Fig. 6A). In the absence of DHT, there was a modestly higher level of nuclear AR protein in LNCaP-EGR1 cells compared with LNCaP-Neo cells. Immunocytochemical analysis confirmed these results. Without DHT, AR protein in LNCaP-Neo control cells was mainly present in the cytoplasm and translocates to the nucleus with DHT treatment (Fig. 6B). In LNCaP-EGR1 cells on the other hand, AR expression could be detected in the nucleus in addition to the cytoplasm in the absence of DHT. Nuclear AR reactivity was further increased upon DHT treatment. These results suggest that EGR1 overexpression can promote translocation of AR from cytoplasm to nucleus.
RNA Interference with EGR1-siRNA Inhibits PSA Expression in LNCaP Cells-To determine whether down-regulation of endogenously expressed EGR1 in prostate carcinoma cells can modulate PSA gene expression, we employed siRNA targeting EGR1. LNCaP cells growing in medium containing 10% regular FBS were transfected with siEGR1 RNA duplexes or control siRNA specific for GFP. It should be noted that regular FBS contains testosterone and DHT, which drive androgen receptor-dependent PSA gene expression. Western blot analysis performed on extracts prepared 48 h after transfection indicates that the siEGR1 leads to a significant reduction in EGR1 protein levels without an effect on AR levels (Fig. 7A). PSA levels were measured by RT-PCR. The results show that EGR1 knockdown by siEGR1 resulted in down-regulation of PSA gene expression 3-fold relative to the expression in siGFP transfected cells (Fig. 7B). These results further confirm that EGR1 expression promotes AR transcriptional activity in prostate carcinoma cells. DISCUSSION The androgen receptor signaling pathway plays a crucial role in normal prostate physiology and prostate tumorigenesis (21). Human prostate carcinomas are generally androgen-sensitive and react to hormonal therapy by temporary remission, which is invariably followed by relapse to an androgen-independent state. The molecular mechanisms of transition from androgen dependence to androgen independence remain poorly understood, and a number of possibilities have been explored to explain this phenomenon. Earlier theories postulated that androgen-independent cells might not express the AR, or the AR gene could be mutated from androgen-independent tumors to a molecule unable to be activated by androgen. However, several studies show that there is a higher average of AR in androgendependent tumors, and the mutation in the AR is present in a low percentage of local recurrence of androgen-independent tumor (22-24). Furthermore, the AR signaling pathway remains active in androgen-independent prostate cancers (21). How are androgen-dependent genes activated in a low androgen medium such as that of patients under maximum androgen blockage? AR coactivators may contribute to the AR activity in androgen-independent prostate cancer. Their overexpression in prostate cancer could make AR signaling pathway be in a superactive state so that AR transcription is activated in a very low androgen concentration (25,26). There is increasing evidence that signaling pathways of some cytokines and growth factors may activate unliganded AR. They modify AR protein through interactions with AR or AR coactivators, activating AR signaling (27)(28)(29)(30).
Our results have identified functional interaction between EGR1 and the androgen receptor. Several studies have impli-cated EGR1 in prostate tumorigenesis (5,9,10). However, in other tumor types such as breast cancer and glioblastoma, EGR1 behaves as a growth suppressor (12,14). Our findings suggest that the effects of EGR1 in promoting prostate tumorigenesis are mediated, at least in part, by functional interaction with the AR signaling pathway in prostate cells. EGR1 physically interacts with AR and is part of a complex that forms on the regulatory regions of endogenous AR target genes. Furthermore, this interaction translates into increased gene expression for the target gene we have examined, PSA. Our in vitro GST pull-down experiments mapped EGR1 interaction to the N-terminal domain of the androgen receptor. Quantitation of PSA expression in LNCaP-EGR1 cells showed that EGR1 modestly enhanced AR-mediated gene activation even in the absence of androgens. This effect is likely due to the fact that significant EGR1 overexpression promotes the translocation of AR from the cytoplasm into the nucleus. It is possible that the outcome of EGR1-AR interaction could differ for other AR target genes. Also, our study has not examined the impact of EGR1-AR interaction on the regulation of EGR1 target genes. This is an interesting area for future investigation.
In summary, we have identified interaction between EGR1 and AR in prostate carcinoma cells. The fact that EGR1 is frequently overexpressed in human prostate cancer makes this interaction potentially significant. We propose that the interaction between EGR1 and the AR signaling pathway may explain part of the prostate-specific pro-tumorigenic effect of EGR1. Further investigation of this interaction will increase our knowledge of context-dependent regulation of tumorigenesis and may help us to identify new steps that can be targeted for prostate cancer therapy.