Androgens Regulate the Immune/Inflammatory Response and Cell Survival Pathways in Rat Ventral Prostate Epithelial Cells
http://www.100md.com
《内分泌学杂志》
School of Molecular Biosciences (A.J.A.), Center for Reproductive Biology (M.S., J.C.)
Bioinformatics Core Laboratory (M.S., B.G.), Washington State University, Pullman, Washington 99164
Abstract
A major hurdle in understanding the role of androgens is the heterogeneity of androgen receptor (AR) expression in the prostate. Because the majority of prostate cancer arises from the AR-positive secretory luminal epithelial cells, identifying the androgen-mediated pathways in the prostate epithelium is of great significance to understanding their role in prostate pathogenesis. To meet this objective, the current study was designed to identify immediate-early genes expressed in response to the synthetic androgen R1881 in cultured rat ventral prostate epithelial cells. Rat ventral prostate epithelial cells, purified from 20-d-old rats, were cultured, and the presence of AR and the response to androgen were established. The cells were then treated with R1881 for 2 and 12 h to capture immediate-early genes in an Affymetrix-based gene chip platform. A total of 66 nonredundant genes were identified that were responsive to R1881. The functional androgen response elements were identified in the proximal promoter to determine possible molecular mechanism. Cluster analysis identified five distinct signatures of R1881-induced genes. Pathway analysis suggested that R1881 primarily influences cell proliferation/differentiation and inflammatory/immune response pathways. Androgens appear to regulate cell renewal by regulating differentiation, cell proliferation, and apoptosis. Two mutually exclusive inflammatory response pathways were observed. The interferon pathway was up-regulated, and the ILs were down-regulated. The data identified novel androgen-regulated genes (e.g. Id1, Id3, IL-6, IGF-binding protein-2 and -3, and JunB). The loss of androgen regulation of these genes can have important consequences for cellular transformation and transition to androgen-independent growth and survival.
Introduction
DESPITE THE WELL-documented role of androgens in prostate growth and function (1), the molecular pathways elicited by androgens that lead to prostate morphogenesis and pathogenesis are still largely unknown (2, 3, 4). One of the major hurdles in understanding the role of androgens is the heterogeneity of androgen receptor (AR) expression in the prostate. Most of the cell types in the prostate, such as luminal epithelial, basal epithelial, stromal, and smooth muscle cells, express ARs at some point of development and hence are capable of mediating androgen’s actions (5). In response to androgens, these cell types interact in an autocrine-paracrine manner to influence various aspects of prostate growth in normal and diseased states (1). This diverse expression profile has led to believe that androgens act as global endocrine mediators of prostate function.
The classical mechanism by which androgens exert their biological effect is through binding and subsequent activation of the AR. The androgen-bound AR then translocates to the nucleus and interacts with sequence-specific androgen response elements (ARE) in the promoters/enhancers/introns of target genes. The interaction between AR and ARE facilitates the assembly of general transcription machinery, leading to gene transcription (reviewed in Ref.6). Alternatively, AR activation may initiate cellular events leading to activation of latent transcription factors that can mediate or mimic the androgen response. This transcriptional or genomic mode of androgen action is independent of second messenger signal transduction cascades (7, 8).
The most dramatic effect of androgens on the prostate can be observed in castrated rats. After castration, the prostate undergoes involution accompanied by the loss of epithelial cells (9, 10, 11). Upon androgen administration, the prostate of the castrated rat undergoes rapid proliferation and cellular differentiation until it reaches its normal size and function (12). Large scale gene expression profiling has identified a number of genes and pathways, such as stress (13) and inflammatory/immune response, that are potentially involved in androgen-dependent prostate regeneration/morphogenesis (14, 15, 16, 17, 18, 19). Surprisingly, there has been little to no overlap between the genes regulated by androgens in these studies, possibly due to the age of rats used in each study and the timing and dose of androgen replacement after castration, but more importantly, the complex heterogeneity of AR expression itself. These models have significantly advanced our understanding of androgen-dependent prostate regeneration, but provide little evidence of androgen-regulated gene expression in specific cell types of the prostate. Of interest to us is the androgen-regulated gene expression specifically in prostate epithelial cells. A strong body of evidence suggests that the majority of prostate cancer arises from AR-positive secretory luminal epithelial cells (20, 21). For example, overexpression of AR specifically in the prostate epithelial cells leads to the development of prostatic intraepithelial neoplasia, the earliest recognizable stage of prostate cancer (22), and loss of AR reduces the risk of prostate carcinogenesis (23). Hence, AR expression and response to androgens are key determinants of prostate cancer initiation.
The current understanding of androgen-regulated gene expression in prostate epithelial cells is derived from the well-characterized androgen-responsive human prostate carcinoma cell line LNCaP and prostate cancer xenografts (24, 25, 26, 27, 28, 29, 30). It is apparent that these in vitro cancer models do not reflect the true molecular pathways elicited by androgens in normal prostate epithelial cells. Hence, the primary aim of this study was to identify the androgen-regulated gene expression profile and pathways in normal prostate epithelial cells. The epithelial cells were isolated from the rat ventral prostate cells (rVPEC), a representative model for hormone-dependent prostate hyperplasia (31) and cancer (32). The epithelial cells from these glands are hormone dependent, and other cell populations, such as basal and stromal cells, are essentially hormone independent (5). In culture and during initial passages, rVPEC also demonstrate functional androgen-dependent gene expression (33, 34, 35). The experimental design of the study allowed us to capture immediate early androgen-dependent events, specifically in prostate epithelial cells. The results suggest that the androgenic events in prostate epithelial cells are unique and are not necessarily represented in the whole prostate model. These androgen-mediated events, specifically in prostate epithelial cells, are of paramount importance for understanding the normal developmental events associated with the androgen response, and, more importantly, the significance of AR-regulated gene expression in prostate cancer.
Materials and Methods
Animals
Male Sprague Dawley rats, bred and raised in the Washington State University vivarium, were used. The university animal care committee approved all animal procedures performed in this study.
Culture of ventral prostate (VP) cells
The 20-d-old rats were killed by cervical dislocation, and VPs removed. The prostate epithelial cells were purified according to the published procedure, with some modifications (36). Briefly, the pooled tissue from 20 rats was incubated with Hanks’ balanced salt solution (HBSS; Invitrogen Life Technologies, Inc., Gaithersburg, MD) containing 7000 U collagenase type II (Sigma-Aldrich Corp., St. Louis, MO) and 3 mg deoxyribonuclease I/ ml (Sigma-Aldrich Corp.) at 37 C for 4 h with occasional shaking. After incubation, the tissue was allowed to gravity settle for 10 min. The supernatant, which consisted mainly of stromal cells, was removed. The pellet was resuspended in HBSS, and the mixture was spun at 30 x g for 4 min to pellet the epithelial cells. The pellet was resuspended in 1 ml HBSS and plated in Ham’s F-12 containing 10% fetal bovine serum with appropriate antibiotics (penicillin/streptomycin) and plated in 150-mm culture plates (Nunclon, Roskilde, Denmark). At passage 2 the cells were plated in 150-mm culture plates and grown to 80% confluence. The cells were then washed in serum-free medium and starved for 48 h. R1881 (1–100 nM) was added to fresh serum-free medium and added to the plates. RNA was collected after 0 (no treatment), 2, 6, 12, and 24 h of R1881 treatment and analyzed for fibroblast growth factor 8 (FGF8) expression. The optimum FGF8 expression was observed at 10 nM R1881. After this observation, the cells were treated with 10 nM R1881, and RNA was collected for microarray hybridizations. For quantitative PCR analysis of gene expression, the cells were also treated with the antiandrogen hydroxyflutamide (1 μm) in the presence of 10 nm R1881 for 2 and 12 h before harvesting for RNA isolation.
RNA preparation
Total RNA was obtained using the TRIzol (Invitrogen Life Technologies, Inc.) method described previously (36). The final RNA pellet was washed with 75% ethanol, dried, resuspended in diethylpyrocarbonate-treated H2O at a concentration of 1 mg/ml, and stored at –80 C until analysis.
RT-PCR
Total RNA (2 μg) was reverse transcribed in a final volume of 25 μl containing 20 U RNasin (Promega Corp., Madison, WI), 1.25 mM each of deoxy-NTPs, 250 ng oligo(deoxythymidine) (Pharmacia Biotech, Peapack, NJ), 10 mM dithiothreitol, and 200 U Moloney murine leukemia virus (MMLV) reverse transcriptase (Invitrogen Life Technologies, Inc.) in the MMLV first-strand synthesis buffer according to the manufacturer’s instructions. Each PCR contained 250 pg reverse transcribed DNA, 1 μM of each 5' and 3' oligonucleotide primer, 0.5 U Taq polymerase, and 0.1 mM of each deoxy-NTP. The primer pair sequence used was obtained from published sequences and is mentioned in the figure legends. The possible contamination of RNA with DNA was distinguished by performing the RT reaction without MMLV reverse transcriptase. The absence of any product in the amplification reaction using such a reverse transcribed preparation indicated the absence of any contaminating DNA in our RNA samples. Each RT reaction was performed using three different samples. The PCR-based amplification reactions were carried out in duplicate on each reverse transcribed RNA sample. Simultaneous PCRs were also carried out using primers designed to rat cyclophilin or S2 to monitor the efficiency of the RT-PCRs. Cyclophilin and S2 were faithfully amplified in all PCRs, indicating consistency in the quality of RT and PCRs. The identity of the corresponding PCR products was sequence confirmed by the Center for Reproductive Biology Molecular Biology Laboratory. The data presented are representative of at least three different RT-PCRs and DNA preparations.
Immunocytochemistry (ICC)
The rVPEC, LNCaP cells (positive control for AR), and DU145 cells (positive control for AR) were cultured in four-well Nunc cell culture slides to 80% confluence. The cells were treated with R1881 (10 nM) for 4 h before the ICC was performed. The previously established protocol was followed (37). Briefly, the cells were washed with 1x PBS twice and fixed in 10% neutral buffered formalin (Sigma-Aldrich Corp., HT509-1-2) for 10 min. After fixation, the cells were washed again with 1x PBS three times and then blocked in 10% goat serum for 1 h at room temperature. The cells were then incubated with primary AR (Upstate Biotechnology, Inc., Lake Placid, NY) or c-Met (Santa Cruz Biotechnology, Santa Cruz, CA) antibody for 1 h at 1:25, 1:50, and 1:100 dilutions. This was followed by washes in blocking solution. The cells were then incubated in secondary antibody [biotinylated goat anti rabbit IgG (H+L)] solution (1:300) for 1 h and subsequently washed three times with blocking solution. The well chambers were removed, and the horseradish peroxidase-conjugated streptavidin (Pierce Chemical Co., Rockford, IL) at a dilution of 1:3125 was added to the cells and incubated for 10 min. The 3,3'-diaminobenzidine staining protocol was then followed (peroxidase substrate kit, Vector Laboratories, Inc., Burlingame, CA). The color reaction was allowed to run for 5 min. After this, the nuclei were stained with hematoxylin for 30 sec. The cells were subsequently washed in water and increasing gradients of ethanol, then placed in xylene before mounting in a xylene-based mounting solution and observed under a microscope.
RNA quality for microarray
RNA quality was determined by both electrophoretic methods using a denaturing agarose gel and absorption readings at 260 and 280 nm. If excessive degradation or protein contamination of the RNA was evident, the sample was not used for microarray hybridization. A minimum OD (260/280) ratio of 1.8 was a requirement for microarray hybridization.
Microarray hybridizations
The Affymetrix gene chip platform, described previously (38), was used to determine transcriptional changes in the epithelial cells in response to androgens. Ten micrograms of total RNA from each of the samples was used to create the target for the microarray. The biotinylated CTP- and UTP-labeled cRNA were fragmented, hybridized to RAE230A arrays (Affymetrix, Santa Clara, CA), and stained in accordance with the manufacturer’s standard protocol. The arrays were stained and washed using the Affymetrix GeneChip Fluidics Station 400 and were scanned using a GeneArray Scanner 2500A (Agilent, Palo Alto, CA). The resulting data were viewed, and preliminary assessment was made using GCOS software (Affymetrix). All reactions and microarray hybridization procedures were performed at the Laboratory for Biotechnology and Bioanalysis I (Washington State University).
Absolute and statistical analysis for microarrays
Microarray output was examined visually for excessive background noise and physical anomalies. The default GCOS statistical values were used for all analyses. All probe sets on each array were scaled to a mean target signal intensity of 125, with the signal correlating to the amount of transcript in the sample. An absolute analysis using GCOS was performed to assess the relative abundance of the 45,000 represented transcripts based on signal and detection (present, absent, or marginal). Although 15,923 probe sets are represented on these arrays, there is a certain level of redundancy within and between each array, which results in a lower number of unique transcripts represented on the RAE230A chipset; however, the level of redundancy has not been firmly established.
The absolute analysis from GCOS was imported into GeneSpring 7.0 software (Silicon Genetics, Redwood City, CA). The R1881 treatment time-course data were normalized within GeneSpring using the default/recommended normalization methods. These included setting of signal values less than 0.01 to 0.01, total chip normalization to the 50th percentile, and normalization of each gene to the median. These normalizations allowed for the visualization of data based on relative abundance at any given time point rather than compared with a specific control value.
Data restrictions and analytical tools in GeneSpring were applied to isolate noteworthy and possibly important patterns of gene expression during the course of R1881 treatment on epithelial cells. Transcripts expressed differentially at a statistically significant level were determined using a P value cutoff of 0.05 and a Benjamini and Hochberg false discovery rate multiple testing correction. This was applied to all three time points and considered all transcripts represented on the arrays. Subsequently, expression restrictions were applied to the transcripts expressed in a significant manner. These restrictions were designed so that the remaining transcripts met the following requirements in addition to being expressed in a significant manner: 1) each transcript must have a signal value of at least 100 in a minimum of one of three points in the time course; and 2) the range of the replicates must not exceed 1 (in the normalized scale) in the time points comprising the time course. The resulting transcripts were screened using Excel (Microsoft, Redmond, WA) for redundant UniGene entries. Transcripts that passed these restrictions (top list) were considered for additional analysis, which included clustering and pathway analysis.
Quantitative PCR analysis
The relative gene expression levels of selected genes were determined by real-time quantitative PCR based on TaqMan chemistry. The primer probes [complement component 3 (C3), IL-6, TNF receptor superfamily/osteoprotegerin (Tnfrs 11b/OPG) 11b, WNT1 inducible signaling pathway protein 2 (WISP2), JunB, and -actin] were obtained from Applied Biosystems (TaqMan Probes, Applied Biosystems, Foster City, CA). The culture and treatment of cells were essentially similar to those described above. Total RNA was extracted from these cells using the TRIzol reagent and was reverse transcribed (2 μg/μl) using the oligo(deoxythymidine) protocol. Real-time quantitative PCR was used to measure the mRNA expression levels of the respective genes under various treatment conditions relative to -actin mRNA expression. All PCRs were performed in a final volume of 50 μl on an ABI PRISM 7700 detection system (Applied Biosystems) according to the manufacturer’s instructions. The following temperature profile was used: 10 min at 95 C, followed by 40 cycles of 15 sec at 95 C (denaturation step) and 1 min at 60 C (annealing/extension step). The cycle threshold (Ct) was used to calculate relative amounts of target RNA. Samples that did not reach the threshold line were considered not to be above background. All experiments were performed in triplicate and repeated three times.
The Ct method (Applied Biosystems User Bulletin 2, ABI PRISM 7700 detection system) was used for relative quantification of gene expression. The Ct values of the target genes from triplicate PCRs were normalized to the levels of -actin (endogenous control) from the same cDNA preparations. The average Ct for each gene is calculated by subtracting the Ct of the sample RNA from that of the control RNA for the same time measurement. This value is called the Ct, reflects the relative expression of the treated sample compared with the control, and becomes the exponent in the calculation for amplification 2Ct control – Ct sample, the equivalent to the fold change in expression. Statistically significant difference between treatments was determined by Student’s t test.
The semiquantitative PCR analysis was used to estimate changes in the levels of the established sequence tag (EST) AI071994 [dickkopf homolog 2 (DKK2)], induced by androgens.
Pathway analysis
The genes in the top list were classified according to their roles in regulating major cellular processes using the Affymetrix Netaffx database. A more detailed analysis of cellular function influenced by the genes in the top list was performed using the PathwayAssist (version 3.00) software (Stratagene, La Jolla, CA). The list was imported into pathway assist according to the manufacturer’s instructions. The software updated the list by excluding the redundant and nonannotated genes. This final list was then used to develop the common targets pathway for the identification of cell processes influenced by the genes in the list. Each identified cellular process was confirmed through the PubMed/Medline hyperlink embedded in each node.
Identification of potential ARE
The rat genome 9999 program built in GeneSpring 7.0 was used to find functional AREs. The rat genome 9999 software identified the promoter sequences of the genes used to build the homology table. Literature searches identified a consensus functional ARE sequence that was used to identify the AREs in the genes of interest that were regulated by R1881. The primary criteria for the identification of functional ARE was the presence of conserved C and G at positions 5 and 11, respectively. The functional ARE with associated redundancy is shown in Table 3. The rat genome searched 9999 upstream of the start site upstream in the promoter region for the consensus sequence.
Statistical analysis
The data were obtained from a minimum of three different experiments unless stated otherwise. The time-matched controls were used to rule out any time-dependent changes in control/treated sets in microarray analysis. Standard statistical tests for significance (t test) were used for semiquantitative and real-time PCR datasets.
Results
Characterization of rVPEC
The androgen target genes were identified in the epithelial cells isolated from VPs of 20-d-old rats. At this age (20 d postpartum), the cytosolic AR, particle-bound dihydrotestosterone (DHT), and uptake of [3H]DHT into the 100,000 x g sediment are maximum in prostate epithelial cells (39).
There have been many previous attempts to establish a reliable primary prostate culture system to study androgenic responsiveness. Some primary epithelial cultures yielded only a limited number of passages, and others yielded a down-regulation of AR expression upon culture (40, 41, 42). Based on these observations, we used the early passage (passage 2) prostate epithelial cells that maintained androgen responsiveness as described below. At passage 2, a homogeneous population of cells was obtained that consisted primarily of epithelial cells. The epithelial cells were characterized by the presence of AR and c-Met receptor (hepatocyte growth factor receptor) by PCR as previously described and ICC. As shown in Fig. 1, the purified cells expressed both receptors. The presence of AR and c-Met immunoreactivity in the cultured cells essentially confirmed the PCR data presented previously (36). ICC also confirmed that the majority of the cell population in culture at passage 2 was of epithelial origin. Additional characterization of these cells, i.e. epithelial vs. stromal or basal vs. luminal epithelial, was performed by identifying cell-specific markers in the gene expression data discussed below.
Response to androgens
The synthetic AR agonist R1881 (methyltrienolone) was used to evaluate the response of prostate epithelial cells to androgens in culture. Dose-response experiments were performed to optimize the effective concentration of R1881 (0.1, 10, and 100 nM) required to stimulate prostate epithelial cells. The expression of FGF8, a known androgen response gene in prostate epithelial cells (43), by semiquantitative PCR was used to monitor the response of epithelial cells to R1881. Maximum FGF8 gene expression was observed within 2 h of stimulation with R1881 at a concentration of 10 nM (Fig. 2). After 2 h, the expression of FGF8 was significantly reduced, but was still higher than that at 0 h. The gene expression changes in prostate epithelial cells were therefore captured within 2 and 12 h of treatment with R1881 (10 nM). This short time frame was used identify events that were immediately downstream of the activated AR and not due to secondary or tertiary responses that could have occurred at 24 h.
Gene expression profiling
The chip hybridizations (RAE230A, Affymetrix) were performed using RNA from at least two different sets of experiments, resulting in a minimum of two replicates for each time point to minimize random errors. Microarray data were processed with GCOS (Affymetrix) and exported into GeneSpring for additional analysis that consisted essentially of filtering the gene list over the specified restriction as outlined in Materials and Methods. A statistically significant change in gene expression does not necessarily dictate a large quantitative change. The data analysis was thus restricted to only those genes that had raw expression levels greater than 100 and changed by 2-fold or more in any of the conditions tested. Based on these restrictions, a total of 66 genes were found to be regulated by R1881 in rat prostate epithelial cells (Fig. 3).
Clustering of the genes based on Spearman’s correlation was performed to identify major expression profile signatures induced by R1881. As shown in Fig. 3, at least five major R1881-induced expression profiles were observed. The K-mean clustering using Spearman’s correlation of the regulated genes showed a similar profile.
Prostate epithelial cell-specific marker(s) expression
Because AR and c-Met expression determined by RT-PCR and ICC and the functional androgen response determined by semiquantitative RT-PCR of FGF8 expression confirmed that the cells used in this study were primarily epithelial, additional analysis based on gene expression profile was considered necessary. The gene expression profile in Table 1 shows the summary of this analysis. The genes normally expressed in the stromal compartment of the prostate, such as keratinocyte growth factor (44) and hepatocyte growth factor (45) were absent in the array data. This profile suggested that stromal cell contamination was negligible or below detection in the epithelial cell preparation. The gene expression was also analyzed to determine whether the epithelial cell population consisted of basal and/or luminal epithelial cells. The luminal epithelial-specific genes, cytokeratins 8 and 18, CD10, and CD24, and the basal epithelial-specific genes, BTG2 and 17-hydroxysteroid dehydrogenase, were highly expressed, as determined by their raw signal (>350) (reviewed in Ref.46). The predominance of luminal epithelial-expressed genes suggested that the epithelial cells were primarily luminal. Raw data analysis also showed the presence of AR mRNA (mean raw intensity, 95), which did not change with the treatment. Previous reports have shown a down-regulation of AR mRNA expression in LNCaP cell lines treated with androgens for more than 48 h (47). This result should be carefully compared with our data that show a lack of effect of androgens on AR mRNA possibly due to 1) relatively short treatment (12 h), and 2) the use of normal prostate epithelial cells rather than cancer cell lines. This gene expression profile was a significant observation that allowed interpretation of data in the context of androgen-regulated genes in the epithelial compartment of the VP.
Androgen-regulated genes in prostate epithelial cells
The immediate-early events elicited by R1881 were faithfully captured by the experimental design of this study. Twenty-seven (12 known and 14 ESTs) genes were up-regulated within 2 h of R1881 stimulation and returned to basal levels within 12 h (<2-fold change at 12 h; Fig. 4). Surprisingly, only eight genes (seven known and one EST) were up-regulated by R1881 after 12 h of treatment. These eight genes were unique compared with the 2 h up-regulated gene list. Twenty-eight genes (21 known and seven ESTs) were down-regulated by R1881 within 2 h of R1881 treatment (Fig. 4). Of the 10 genes (six known and four ESTs) that were down-regulated after 12 h of R1881 treatment, five genes were also down-regulated by R1881 after 2 h of treatment [IL-6, Id1, growth differentiation factor 8 (GDF8), and ESTs AB032395 and AI071994]. Collectively, this trend suggested a strong time-dependent transient expression mechanism.
Validation of androgen response and reliability of microarray analysis
The reliability of the microarray data shown in Fig. 3 is attested by the observations that 1) in instances where there were more than one oligonucleotide set for a particular gene (e.g. CCAAT/enhancer binding protein ), the clustering algorithm ordered these two probe sets adjacent to each other, or both the probe sets showed a similar profile (e.g. gap junction membrane channel protein 5); and 2) among the known transcripts that were influenced by androgens, there are several that have been previously reported to be rapidly induced or repressed in response to testosterone, DHT, or R1881. Consistent with these previous observations, the expression of C3 (48), IGF-binding protein-2 (IGFBP-2) (49), osteoprotegerin (OPG) (50), and tissue inhibitor of metalloprotease 3 (18) expression increased, whereas the expression of p16 (51), IGFBP-3 (14, 49), and CEBPd (14) decreased in response to R1881 in prostate epithelial cells. The exception to this list was the down-regulation of adrenomedullin in our experiment, which was previously shown to be up-regulated by androgens in the whole VP (15, 17, 52). The change in the expression of the above-mentioned, known androgen response markers in the prostate confirmed that the top list (Table 2) is the true representative of the immediate-early androgen-responsive genes in the rVPEC.
Additional confirmation of the microarray data was obtained using quantitative PCR on a subset of five genes from Table 2. The selected genes were representative of the complete dataset that was either up- or down-regulated after 2 or 12 h of treatment. The two other goals of the quantitative PCR were 1) to include additional replicate samples, and 2) confirm the androgen response by including the antiandrogen hydroxyflutamide in the culture medium. Overall, the expression profile of the selected genes, as determined by real-time PCR analysis (Fig. 5), was consistent with the microarray data shown in Fig. 3 and Table 2. The magnitude of fold changes observed in real-time PCR data were usually higher compared with microarray data, possibly due to differences in the techniques used for quantification in these methods. The addition of hydroxyflutamide also reversed the effect of R1881 on the expression of C3 and JunB (Fig. 5), demonstrating the specificity of the androgen response. Semiquantitative RT-PCR analysis was also performed for the EST AI071994, which showed a significant homology to the known human/mouse gene DKK2. This approach allowed us to visualize the gene expression that strengthened the microarray analysis of unknown transcripts. It is evident from Fig. 6 that treatment of prostate epithelial cells with androgen down-regulated the expression of this EST. Densitometric scanning and quantitation of RT-PCR analysis by normalizing with the constitutively expressed gene S2 also supported the RT-PCR data. Collectively, the postmicroarray validation of androgen response on selected genes reassured the authenticity of the androgen-regulated gene expression profile shown in Fig. 3 and Table 2.
Major genes/pathways influenced by androgens in prostate epithelial cells
Functional annotations of the known genes suggested that R1881 treatment influenced a variety of pathways, as listed in Table 2. For gene classification, the overall emphasis was on the major cellular process in which a particular gene was involved. Based on this classification, the primary cellular process influenced by R1881 was the inflammatory/immune response pathway. A similar observation that regulation of the inflammatory response pathway by androgens in intact prostate and prostate cancer cell lines was also reported previously (14, 28). However, the specific gene set reported in these studies was not identical with that reported in Table 2. This difference can be due to multiple reasons, for example, use of the whole prostate and/or cancer cell lines, but the significance of the overall regulation of the inflammatory/immune response pathway in prostate epithelial cells cannot be discounted and strengthens these observations. The use of cultured and relatively pure epithelial cell populations from the VP also suggests that this is the major pathway influenced by androgens.
The cellular processes influenced by the R1881-regulated genes shown in Table 2 were reevaluated by building the pathways in PathwayAssist software. This was considered necessary, because any single gene could influence multiple processes; hence, the information shown in Table 2 could be biased. An objective assessment of the major pathways was determined by the number of arrows pointing in or out of each cellular process (connectivity). Based on this analysis, the proliferation pathway (total number of connecting nodes or connectivity = 18) appeared to be the primary pathway influenced by androgens (Fig. 7). The genes (connecting nodes) involved in promoting proliferation, such as Id1, Id3, JunB, and IGFBP-3, decreased, and those inhibiting proliferation, for example, IGFBP-2 and 2',5'-oligoadenylate synthetase I (OAS1), increased, suggesting that, in general, the rate of proliferation decreased or was not affected after R1881 treatment. These data also supported our observation that R1881 has no effect of the proliferation of cultured VP epithelial cells during the course of treatment as determined by [3H]thymidine incorporation (data not shown). Concomitant with this was the observation that the differentiation process (connectivity = 13) increased. The decrease in the genes down-regulated during differentiation, such as Id1 and Id3, and the increase in IGFBP-2 and matrix Gla protein (MGP), which are known to promote differentiation, also support this analysis. Therefore, androgens appear to determine the intricate balance between proliferation and differentiation. The other pathways that were influenced by R1881 (in the order of decreasing connectivity) were apoptosis, immune response (immune response plus immunity), secretion, maturation, regulation of signal transduction, synthesis, pathogenesis, complement activation, metabolism, and DNA damage recognition.
Inflammatory/immune response pathway
The expression of 10 known genes in the inflammatory response pathway changed in response to R1881 treatment. Of these, six that were increased within 2 h of R1881 treatment were the CXC chemokine ligand-10 (CXCL-10), proteosome subunit- type 9 (PSMB9); interferon (IFN)-inducible, double-stranded RNA-dependent, protein kinase; RT1.S3, a nonclassical type 1 major histocompatibility complex (MHC) gene; OAS1; and myxovirus resistance protein-2 (Mx2). The expression of all these genes returned to pretreatment levels within 12 h, suggesting an immediate-early and transient expression in response to androgens.
The more than 6-fold increase in the expression of Mx2, a member of both the dynamin family and the family of large guanosine triphosphatases in response to R1881 was the largest. This protein is up-regulated by IFN-, but does not contain the antiviral activity of a similar Mx1. The expression of guanylate-binding protein-2 (Gbp2) and CXCL-10, which are also induced by IFNs, increased more than 3- and 2-fold, respectively, in response to androgens. In addition to Mx2, GBP-2, and CXCL-10, all other R1881 up-regulated genes within this pathway are also known to be regulated by IFNs. These results prompted us to investigate the expression of IFNs and other genes within the IFN pathway in the raw data. The expression of IFN- and - was observed (>100 raw intensity), but did not change significantly in response to R1881 treatment. A similar expression profile was observed for IFN pathway-associated genes, such as IFN-inducible protein-16, IFN regulatory factor-3, IFN--inducible protein-27-like, and IFN- receptor IFN regulatory factor-7. These results suggest that R1881 may mimic or modulate the IFN response in prostate epithelial cells. Because our data were also normalized against time-matched control values at 2 and 12 h, the likelihood of these effects being mediated through IFN expression and release during the course of the investigation in the absence of R1881 is negligible.
The expression of cytokines IL-6 and IL-12, the mediators of the inflammatory response pathway, was repressed by androgens. Functional pathway analysis suggested that the IL-6 node, with highest connectivity, could be the key mediator of the actions of R1881 in prostate epithelial cells.
Transcription factors
AR activation led to a decrease in the expression of transcription factors Id1, Id3, CEBPd, and early growth response-1 (Egr-1) and an increase (2.5-fold) in the expression of testis-specific histone-2a. The decrease in the expression of Id1 and Id3, the dominant-negative helix-loop-helix transcription factors, may support the role of androgens in initiating or maintaining differentiated functions (36). CEBPd activity plays a role in proliferation, differentiation, acute inflammatory response, and apoptosis depending on the cell type and specific physiological response. In rat VP, androgens also down-regulate CEBPd levels (53). The down-regulation of Egr-1 transcription factor is a significant observation, because one of the mechanisms that may lead to the development of aggressive prostate cancer is loss of androgen-mediated transcriptional regulation of Egr-1 (54).
Signal transduction
Significant changes in the expression of signal transduction genes were not observed after R1881 treatment of prostate epithelial cells. This observation is not unexpected because 1) the primary events elicited by androgens, at least in the current model system, appear to be mediated by cytosolic receptor and not through the proposed membrane-bound receptor (55); and 2) the signal transduction genes are expected to be phosphorylated, which is not manifested as a change in expression. However, the raw data were analyzed for the expression of common signal transduction genes. Among these, a raw signal intensity of 100 or greater for phosphatidylinositol 4-kinase, serum-inducible kinase, MAPK6, diacylglycerol kinase (), serine-threonine kinase 2, ERK1b, MAPK4, MAP2k5, Rho-associated, coiled-coil containing protein kinase 2 (ROCK2), mothers against DPP homolog 4 (Drosophila) (SMAD4), and cAMP response element-binding protein-1 was observed.
Regulation of growth
The expression of IGFBP-3 decreased in the presence of androgen within 2 h, but returned to basal levels by 12 h. The expression of IGFBP-2 increased in response to androgens after 12 h of treatment. In the normal prostate epithelial cells, IGFBP-2 may act as a growth inhibitor (56). Hence, androgen-mediated up-regulation of IGFBP-2 may limit the growth of prostate epithelial cells. In the castrated rat model, androgens were shown to down-regulate IGFBP-3 expression. Our data suggest that down-regulation of IGFBP-3 by androgens occurs at the level of prostate epithelial cells immediately downstream of AR activation and may protect cells from undergoing apoptosis (57). Analysis of the raw data indicated high and constitutive levels of TGF-3 (raw signal, 3600). The constitutive expression of TGF-1 (raw signal, 97), TGF- (raw signal, 70), IGF-I (raw signal, 302), and IGF-II receptor (raw signal, 654) was also observed. Collectively, these observations, for example, high levels of TGF-3, may also explain the low rate of proliferation of prostate epithelial cells.
The expression of MGP, an extracellular matrix protein that acts as an inhibitor of pathological calcification, was down-regulated by androgens. In androgen-dependent Shionogi medullary carcinoma 115 cells, the expression of MGP is also negatively regulated by androgens (13), thus confirming the androgen-mediated response observed in our studies. However, the relevance of MGP is not known, but detailed analysis of its significance during the transformation of prostate epithelial cells may provide evidence of the process of microcalcification observed in prostate cancer.
WISP-2 is a growth arrest-specific gene in vascular smooth muscle cells and is a component of the WNT1 signaling pathway (58). Thus, up-regulation of WISP-2 expression does support the role of androgens as growth inhibitors. The expression of WISP-2 may also have broader implications on prostate biology: for example, as a component of the WNT signaling pathway that is aberrantly activated in prostate cancer (59) and as a secreted factor possibly involved in paracrine/autocrine interactions.
Transport activity
A decrease in the expression of monocarboxylate transporter, fatty acid-binding protein 3, retinol-binding protein 1 (cellular), and lipocalin 2 and an increase in the expression of peptidyl isomerase (Ppicap) were observed in response to androgens.
Apoptosis
Tnfrs11b (OPG), a decoy receptor for TNF-related apoptosis-inducing ligand that inhibits TNF-related apoptosis-inducing ligand-induced apoptosis, was the only apoptosis-related gene identified in our dataset. The role of androgens in the regulation of OPG in prostate epithelial cells or cancer cell lines is not known, but androgen-independent cell lines secrete significantly more OPG than androgen-responsive LNCaP cell lines (60). In the osteoblast cells, the androgens specifically inhibit OPG mRNA levels and protein secretion (50). In contrast to these studies, the observed decrease in OPG levels produced by androgens may suggest a cell-specific, androgen-mediated regulatory mechanism.
Cell-cell signaling
The expression of late gestation lung protein-1 (Lgl1) in prostate epithelial cells increased after 12 h of treatment with R1881. Lgl1 is a novel, glucocorticoid-inducible gene that regulates mesenchymal-epithelial interactions. Interestingly, the expression of cysteine knot superfamily, homolog (Xenopus laevis) Cktsf1b1 (gremlin), an extracellular bone morphogenic protein (BMP) antagonist (61) also involved in mesenchymal-epithelial interactions, decreased in the presence of R1881. It is speculated that the opposing role of androgens on germlin (decrease) and BMP (increase) (62) may regulate the bioactivity of BMP, thus influencing prostate growth and development. Together, Lgl1 and Cktsf1b1 may play a role in prostate branching morphogenesis (63).
Adrenomedullin (AM) is hypotensive peptide highly expressed in the rat prostate ventral epithelium and is a direct androgen-responsive gene (15). In situ hybridization of normal rat prostate tissue showed that AM expression is localized in the epithelial cells (52). Our results are contradictory to these reports and suggest that AM is down-regulated in prostate epithelial cells after a brief exposure to androgens.
Cell cycle
A decrease in the expression of the cyclin-dependent kinase inhibitor (CDKN) 2A (p16INK4) tumor suppressor gene in response to androgens was observed. CDKN2A functions as an inhibitor of cyclin-dependent kinase (CDK) 4 and CDK6, which initiate phosphorylation of the retinoblastoma tumor suppressor protein. Thus, a decrease in CDKN2A can promote the cell cycle. Its probable physiological role is in the implementation of irreversible growth arrest or cellular senescence that can be induced by stresses such as DNA damage and aberrant mitogenic signaling in human primary cells. The observed decrease in CDKN2A may support the role of androgens in maintaining proliferation and delaying replicative senescence, at least in the early-passage cells used in this study (64). Previously, androgens have also been shown to down-regulate p16 (51),
Identification of ARE in androgen-regulated genes
The experimental design of this study allowed us to identify immediate downstream targets of androgens in rVPEC. The change in the expression of genes (Table 2) is anticipated to be under direct androgen control, although a change in the expression after 12 h of treatment can be indirect due to subsequent secondary or even tertiary events elicited by AR activation. One of the approaches that may support the direct regulation of gene transcription shown in Table 2 by the activated AR is to determine the presence/absence of potential ARE in the promoter of androgen-regulated genes as outlined in Materials and Methods.
The GeneSpring (rat genome 9999) software allowed us to identify ARE within 9999 bp of the predicted transcriptional start site. However, we focused on the ARE within the proximal 1000 bp because these proximal AREs are more likely to have a regulatory role. The rat genome 9999 software could identify the promoter sequence in 40 of the 66 genes listed in Table 2. Of the 40 genes for which promoter sequences were available, the ARE consensus sequence was present in 23 genes (Table 3). The presence of the ARE in the majority of the androgen-regulated genes identified in this study is therefore a significant observation and warrants additional experimental confirmation. The functional ARE in C3, a known androgen-responsive gene, is in the first intron (48) and, hence, was not included in the list (Table 3). Thus, including introns in the search may provide evidence for the presence of functional ARE in other androgen-regulated genes listed in Table 2.
Fifteen of the 23 genes with putative ARE (Table 3) were down-regulated by androgens within 2 h of treatment. Interestingly, of these eight genes, the expression of four genes is up-regulated only after 12 h of treatment. The significance of this observation is not known, but it can be speculated that androgens appear to actively down-regulate the majority of the genes, possibly through the ARE directly, whereas up-regulation may involve secondary events, such as the expression/phosphorylation of novel accessory proteins.
Discussion
The role of androgens in regulating normal prostatic growth, development, and function is complex and involves interactions between various cell types that is mediated by paracrine and autocrine interactions involving growth factors and cytokines. In the pubertal prostate, the androgens primarily regulate the differentiated functions, but also maintain epithelial cell numbers by regulating basal rates of proliferation and apoptosis (65). The complex heterogeneity of the prostate makes it difficult to understand the specific mechanism by which androgens regulate epithelial cell function. Understanding these mechanisms is of significant interest because 99% of prostate cancer arises from these AR-positive epithelial cells. Our dataset on the androgen-regulated genes in prostate epithelial cells therefore assumes greater significance in understanding the cellular pathways influenced by androgen in the cell type that eventually becomes transformed.
Androgens regulate a significantly small fraction of genes in prostate epithelial cells compared with the whole prostate gland. This is obvious, because diverse cell types within the prostate express AR and respond to androgens. More importantly, the lack of paracrine interactions and brief exposure to androgens may have also resulted in the small number of androgen-regulated genes identified in this study. The presence of known androgen-regulated genes in our dataset is consistent with the published literature. Our results therefore confirm many investigations of the androgen-regulated genes despite the fact that there was a minimal overlap of the results independently compared with data obtained through various large-scale profiling techniques, such as subtractive hybridization, SAGE (serial analysis of gene expression), and microarray, reported in the literature (13, 14, 15, 16, 17, 18, 19, 24, 25, 26, 27, 28, 29, 30).
A significant observation was the lack genes involved in major signal transduction pathway, possibly due to the absence of secondary paracrine events and the predominant genomic actions of androgens. If androgens were to act through nongenomic pathways, then the 2 and 12 h points were long enough to capture membrane receptor/nongenomic events (66).
Based on published results and data obtained in this study, it seems that the androgens influence at least two major pathways in prostate epithelial cells. These include the inflammatory/immune response and the cell survival pathway (broadly involving apoptosis, proliferation, and differentiation).
The down-regulation of JunB, Id1, and Id3, the positive regulators of the cell cycle, also supports the absence of an overall effect of androgens on proliferation. Apart from these classical cell cycle markers, an increase in WISP2, a negative regulator of the cell cycle (58), and the overall up-regulation of the IFN pathway, previously shown as an androgen-dependent proliferation inhibitory pathway in transformed tumorigenic benign prostatic hyperplasia (BPH1; CAFTD) compared with parental nontumorigenic BPH1 cells (28), also qualify our results. The decrease in CDKN2A, the inhibitor of CDK, and increases in CXC chemokine ligand 10 may sustain basal proliferation of prostate epithelial cells in response to androgens (64, 67). Thus, it appears that androgens regulate the self-renewal in which neither regression nor overgrowth of prostate epithelium occurs, thus recapitulating the in vivo effects of androgens on prostate epithelial cells (4). It is speculated that the down-regulation of CDKN2A may also lead to the delaying of cellular senescence in vitro (64). The gene expression profiling suggests that androgens may support the differentiation of prostate epithelial cells in culture, as is evident from the decreased expression of Id1 and Id3 and the increased expression of C3 and Ppap2b (endoplasmic reticulum transmembrane protein Dri42) (68).
The inflammatory response pathway may be elicited by the components of the IFN pathway, such as guanosine triphosphatases (Mx2), Gbp2, protein catabolism [proteosome (prosome, macropain) subunit, type 9], and chemokines (CXC chemokine ligand 10). A previous report that C3, a component of the complement activation system of the immune response pathway, is a direct androgen target independently confirms these observations (69).
In general, androgens are known to suppress the immune response, although the exact mechanism is not known (70). In the VP, androgen withdrawal induces the immune/inflammatory genes IL-15 and IL-18, possibly due to influx of T cells, macrophages, and mast cells (14). In the current in vitro model, androgens also appear to modulate the inflammatory response, but specifically in epithelial cells. At least two androgen-dependent immune/inflammatory gene signatures are observed in prostate epithelial cells. The first is the general up-regulation of the IFN pathway such as CXCL10; IFN-inducible, double-stranded RNA-dependent, protein kinase; Mx2; proteosome subunit- type 9; OAS1; and RT1.S3. The second is the down-regulation of IL-6 and IL-12.
The up-regulation of the genes within the IFN pathway by androgens can have important consequences for the immunosurveillance. Loss of the immunosurveillance or onset of immune senescence can lead to BPH and cancer (20). For example, the IFN pathway that is generally suppressed in cancer is also suppressed by androgens in transformed tumorigenic BPH (CAFTD) cells (28). Hence, our data provide direct evidence that in normal cells, androgens positively regulate the immune surveillance pathway.
The down-regulation of IL-6 by R1881 observed in this study is particularly interesting. IL-6 is a multifunctional cytokine that acts as a survival factor in carcinoma of the prostate. IL-6 can stimulate the proliferation of prostate epithelial cells in culture (71, 72), but in LNCaP cells, growth stimulatory and inhibitory effects have been reported (73, 74, 75, 76). Of significance, however, is the crosstalk between IL-6 and AR. IL-6 can induce ligand-independent and synergistic activation of AR, possibly mediated through the MAPK pathway (77). In the absence of androgen, IL-6 leads to increased prostate-specific antigen mRNA levels and activates several androgen-responsive promoters, but not the nonandrogen-responsive promoters, in LNCaP cells (78). These data indicate the IL-6 induces an androgen response in prostate cancer cells through the AR even in the absence of androgens (77, 78). Hence, IL-6 can promote prostate cancer progression through induction of an androgen response.
The decrease in IL-6 can also be due to down-regulation of CEBPd expression through a positive feedback loop. CEBPd is a known androgen-repressed gene and IL-6 positively regulates CEBPd expression (72). Hence, a decrease in CEBPd after androgen treatment can also lead to decreased levels of IL-6. Alternatively, a direct androgen-mediated repression of IL-6 may also lead to the decreased CEBPd expression. Thus, our studies for the first time have established a link between IL-6, CEBPd, and androgens in normal prostate epithelial cells. The down-regulation of IL-6 by androgens directly or indirectly through CEBPd can serve at least three functions: 1) blocking proliferation, 2) suppressing the immune response directly or through CEBPd, and 3) suppression of androgen-independent expression of AR-responsive genes, thus allowing these genes to be expressed only in the presence of physiological levels of androgens. During transition to hormone-independent cancer, the inhibitory effect of androgens on IL-6 may be lost, thus allowing IL-6 to promote proliferation and induce the expression of AR-responsive genes and up-regulation of CEBPd. Deregulation of CEBPd and IL-6 also occurs when prostate cancer progresses to the androgen-independent state (79).
Consistent with the effects of androgens on other transcription factors, such as Id1, Id3, and CEBPd, a down-regulation in the expression of Egr-1 was also observed. Egr-1 is a crucial regulator of cell growth, differentiation, and survival (80). Egr-1 expression is known to be increased in prostate cancer and is constitutively expressed in hormone-independent cancers. Egr-1 plays a functional role in the transformed phenotype and may represent a valid target for prostate cancer therapy (81). Our results suggest that androgens negatively regulate the expression of Egr-1 in normal cells. This is a novel observation and suggests that androgen-mediated down-regulation of Egr-1 in normal prostate epithelium may be involved in growth regulation and differentiation.
IGFBPs, particularly IGFBP-3, also have IGF-independent, antiproliferative, and proapoptotic functions (57). After castration, the levels of IGFBP-3 increase, but they decrease after androgen administration in the whole prostate, suggesting that IGFBP-3 expression is androgen dependent, negatively regulates growth, and promotes apoptosis. The androgen-dependent decrease in IGFBP-3 in our in vitro analysis thus supports the role of androgens in blocking apoptosis, but promoting the growth that is required for cell renewal. IGFBP-2 is the major androgen-regulated IGFBP in prostatic fluid and has a modest suppressive effect on the growth of normal prostate epithelial cells (56). An increase in IGFBP-2 may also explain the low growth of epithelial cells in culture, but, more importantly, supports the significance of the androgen-responsive genes obtained from an in vitro system.
A number of other genes that were regulated by androgens, listed in Table 2, can be used to develop a broader understanding of how androgens influence prostate epithelial cell functions. These include enzymes involved in lipid metabolism and transport, cell-cell signaling, and growth factors that may unfold new pathways in cell-cell interactions within the prostate, as outlined in Results.
The identification of functional AREs in most of the genes for which the promoter sequences were available strengthens the results of this study. The presence of conserved cytosine at 5 bp and guanine at 11 bp and compliance with the known heterogeneity with the consensus ARE (17) strongly support the possibility that the identified genes are direct androgen targets. The presence of known androgen-regulated genes, such as MGP (13) and Adm (17), validates this analysis. A number of additional genes regulated by androgens, possibly through ARE, were also identified. These include Id3, Egr-1, growth differentiation factor 8 (GDF8), IL-6, and Mx2 among others. Functional analysis of the significance of these ARE sites on the expression of the individual genes listed in Table 2 will help in understanding the mechanism of action of androgens in normal and pathogenic prostates.
The major pathways influenced by androgen are expected to form the basis for future investigations to gain a comprehensive understanding of prostate epithelial cell function. The development of similar pathways using more defined, androgen-dependent and independent in vitro and in vivo models is expected to unravel the molecular events associated with the transformation of prostate epithelial cells.
In conclusion, our results have for the first time identified a set of novel immediate-early androgen-regulated genes in prostate epithelial cells. The genes that are normally regulated by androgens were part of this dataset, thus independently validating our results. Based on the combined data analysis presented here and published elsewhere, it seems that androgens regulate the cell survival and immune response pathway with a significant crosstalk. We propose that at the core of this process are genes such as Egr-1, CEBPd, IL-6, IGFBPs, and Id. The regulated expression of these genes is required for the maintenance on normal prostate epithelial cell homeostasis. Loss of androgenic control may lead to altered expression profiles of these genes that may contribute to initiation, maintenance, or progression toward hormone-independent prostate cancer. Although the significance of other androgen-regulated genes in Table 2 cannot be discounted, detailed analysis will be required to gain a better understanding of androgen-mediated events in the prostate by these genes.
Footnotes
This work was supported by the Center for Reproductive Biology, Washington State University.
First Published Online September 29, 2005
Abbreviations: AM, Adrenomedullin; AR, androgen receptor; ARE, androgen response element; BMP, bone morphogenic protein; BPH, benign prostatic hyperplasia; C3, complement component 3; CDK, cyclin-dependent kinase; CDKN, CDK inhibitor; Ct, cycle threshold; CXCL-10, CXC chemokine ligand-10; DHT, dihydrotestosterone; DKK2, dickkopf homolog 2; Egr-1, early growth response-1; EST, established sequence tag; FGF8, fibroblast growth factor 8; GBP2, guanylate-binding protein-2; HBSS, Hanks’ balanced salt solution; ICC, immunocytochemistry; IFN, interferon; IGFBP, IGF-binding protein; Lgl1, late gestation lung protein-1; MGP, matrix Gla protein; MMLV, Moloney murine leukemia virus; Mx, myxovirus resistance protein; OAS1, 2',5'-oligoadenylate synthetase-I; OPG, osteoprotegerin; rVPEC, rat ventral prostate epithelial cell; VP, ventral prostate; WISP2, WNT1 inducible signaling pathway protein 2.
Accepted for publication September 19, 2005.
References
Cunha GR, Donjacour AA, Cooke PS, Mee S, Bigsby RM, Higgins SJ, Sugimura Y 1987 The endocrinology and developmental biology of the prostate. Endocr Rev 8:338–362
Marker PC, Donjacour AA, Dahiya R, Cunha GR 2003 Hormonal, cellular, and molecular control of prostatic development. Dev Biol 253:165–174
Abate-Shen C, Shen MM 2000 Molecular genetics of prostate cancer. Genes Dev 14:2410–2434
Heinlein CA, Chang C 2004 Androgen receptor in prostate cancer. Endocr Rev 25:276–308
Prins GS, Birch L, Greene GL 1991 Androgen receptor localization in different cell types of the adult rat prostate. Endocrinology 129:3187–3199
Lee DK, Chang C 2003 Molecular communication between androgen receptor and general transcription machinery. J Steroid Biochem Mol Biol 84:41–49
Heinlein CA, Chang C 2002 Androgen receptor (AR) coregulators: an overview. Endocr Rev 23:175–200
Heinlein CA, Chang C 2002 The roles of androgen receptors and androgen-binding proteins in nongenomic androgen actions. Mol Endocrinol 16:2181–2187
Kyprianou N, Isaacs JT 1988 Activation of programmed cell death in the rat ventral prostate after castration. Endocrinology 122:552–562
Kwong J, Choi HL, Huang Y, Chan FL 1999 Ultrastructural and biochemical observations on the early changes in apoptotic epithelial cells of the rat prostate induced by castration. Cell Tissue Res 298:123–136
Banerjee S, Banerjee PP, Brown TR 2000 Castration-induced apoptotic cell death in the Brown Norway rat prostate decreases as a function of age. Endocrinology 141:821–832
Evans GS, Chandler JA 1987 Cell proliferation studies in the rat prostate. II. The effects of castration and androgen-induced regeneration upon basal and secretory cell proliferation. Prostate 11:339–351
Pang ST, Dillner K, Wu X, Pousette A, Norstedt G, Flores-Morales A 2002 Gene expression profiling of androgen deficiency predicts a pathway of prostate apoptosis that involves genes related to oxidative stress. Endocrinology 143:4897–4906
Desai KV, Michalowska AM, Kondaiah P, Ward JM, Shih JH, Green JE 2004 Gene expression profiling identifies a unique androgen-mediated inflammatory/immune signature and a PTEN (phosphatase and tensin homolog deleted on chromosome 10)-mediated apoptotic response specific to the rat ventral prostate. Mol Endocrinol 18:2895–2907
Wang Z, Tufts R, Haleem R, Cai X 1997 Genes regulated by androgen in the rat ventral prostate. Proc Natl Acad Sci USA 94:12999–13004
Furuya Y, Isaacs JT 1993 Differential gene regulation during programmed death (apoptosis) versus proliferation of prostatic glandular cells induced by androgen manipulation. Endocrinology 133:2660–2666
Nantermet PV, Xu J, Yu Y, Hodor P, Holder D, Adamski S, Gentile MA, Kimmel DB, Harada S, Gerhold D, Freedman LP, Ray WJ 2004 Identification of genetic pathways activated by the androgen receptor during the induction of proliferation in the ventral prostate gland. J Biol Chem 279:1310–1322
Jiang F, Wang Z 2003 Identification of androgen-responsive genes in the rat ventral prostate by complementary deoxyribonucleic acid subtraction and microarray. Endocrinology 144:1257–1265
Porkka KP, Visakorpi T 2001 Detection of differentially expressed genes in prostate cancer by combining suppression subtractive hybridization and cDNA library array. J Pathol 193:73–79
De Marzo AM, Meeker AK, Zha S, Luo J, Nakayama M, Platz EA, Isaacs WB, Nelson WG 2003 Human prostate cancer precursors and pathobiology. Urology 62:55–62
Okada H, Tsubura A, Okamura A, Senzaki H, Naka Y, Komatz Y, Morii S 1992 Keratin profiles in normal/hyperplastic prostates and prostate carcinoma. Virchows Arch A Pathol Anat Histopathol 421:157–161
Stanbrough M, Leav I, Kwan PW, Bubley GJ, Balk SP 2001 Prostatic intraepithelial neoplasia in mice expressing an androgen receptor transgene in prostate epithelium. Proc Natl Acad Sci USA 98:10823–10828
Eng MH, Charles LG, Ross BD, Chrisp CE, Pienta KJ, Greenberg MN, Hsu CX, Sanda MG 1999 Early castration reduces prostatic carcinogenesis in transgenic mice. Urology 54:1112–1119
Nelson PS, Clegg N, Arnold H, Ferguson C, Bonham M, White J, Hood L, Lin B 2002 The program of androgen-responsive genes in neoplastic prostate epithelium. Proc Natl Acad Sci USA 99:11890–11895
Velasco AM, Gillis KA, Li Y, Brown EL, Sadler TM, Achilleos M, Greenberger LM, Frost P, Bai W, Zhang Y 2004 Identification and validation of novel androgen-regulated genes in prostate cancer. Endocrinology 145:3913–3924
Oosterhoff JK, Grootegoed JA, Blok LJ 2005 Expression profiling of androgen-dependent and -independent LNCaP cells: EGF versus androgen signalling. Endocr Relat Cancer 12:135–148
DePrimo SE, Diehn M, Nelson JB, Reiter RE, Matese J, Fero M, Tibshirani R, Brown PO, Brooks JD 2002 Transcriptional programs activated by exposure of human prostate cancer cells to androgen. Genome Biol 3:RESEARCH0032
Shou J, Soriano R, Hayward SW, Cunha GR, Williams PM, Gao WQ 2002 Expression profiling of a human cell line model of prostatic cancer reveals a direct involvement of interferon signaling in prostate tumor progression. Proc Natl Acad Sci USA 99:2830–2835
Amler LC, Agus DB, LeDuc C, Sapinoso ML, Fox WD, Kern S, Lee D, Wang V, Leysens M, Higgins B, Martin J, Gerald W, Dracopoli N, Cordon-Cardo C, Scher HI, Hampton GM 2000 Dysregulated expression of androgen-responsive and nonresponsive genes in the androgen-independent prostate cancer xenograft model CWR22–R1. Cancer Res 60:6134–6141
Xu LL, Su YP, Labiche R, Segawa T, Shanmugam MN, McLeod DG, Moul JW, Srivastava S 2001 Quantitative expression profile of androgen-regulated genes in prostate cancer cells and identification of prostate-specific genes. Int J Cancer 92:322–328
Fujimoto N, Suzuki T, Honda H, Kitamura S 2004 Estrogen enhancement of androgen-responsive gene expression in hormone-induced hyperplasia in the ventral prostate of F344 rats. Cancer Sci 95:711–725
Homma Y, Kondo Y, Kaneko M, Kitamura T, Nyou WT, Yanagisawa M, Yamamoto Y, Kakizoe T 2004 Promotion of carcinogenesis and oxidative stress by dietary cholesterol in rat prostate. Carcinogenesis 25:1011–1014
Rhim JS 2003 Research into molecular and genetic mechanisms underlying prostate carcinogenesis would be greatly advanced by in vitro prostate cell models. Drugs Today 39:837–847
Cussenot O, Berthon P, Cochand-Priollet B, Maitland NJ, Le Duc A 1994 Immunocytochemical comparison of cultured normal epithelial prostatic cells with prostatic tissue sections. Exp Cell Res 214:83–92
Ilio KY, Sensibar JA, Lee C 1995 Effect of TGF-1, TGF-, and EGF on cell proliferation and cell death in rat ventral prostatic epithelial cells in culture. J Androl 16:482–490
Chaudhary J, Schmidt M, Sadler-Riggleman I 2005 Negative acting HLH proteins Id1, Id2, Id3, and Id4 are expressed in prostate epithelial cells. Prostate 64:253–264
Hill CM, Anway MD, Zirkin BR, Brown TR 2004 Intratesticular androgen levels, androgen receptor localization, and androgen receptor expression in adult rat Sertoli cells. Biol Reprod 71:1348–1358
Li Y, Friel PJ, Robinson MO, McLean DJ, Griswold MD 2002 Identification and characterization of testis- and epididymis-specific genes: cystatin SC and cystatin TE-1. Biol Reprod 67:1872–1880
Moeller H, Blank B, Lander K, Mates G 1987 Ontogeny of the androgen receptor in rat ventral prostate during sexual development. Res Exp Med (Berl) 187:287–294
Chopra DP, Grignon DJ, Joiakim A, Mathieu PA, Mohamed M, Sakr WA, Powell IJ, Sarkar FH 1996 Differential growth factor responses of epithelial cell cultures derived from normal human prostate, benign prostatic hyperplasia, and primary prostate carcinoma. J Cell Physiol 169:269–280
Chaproniere DM, McKeehan WL 1986 Serial culture of single adult human prostatic epithelial cells in serum-free medium containing low calcium and a new growth factor from bovine brain. Cancer Res 46:819–824
Chowdhury A, Harber GJ, Chopra DP 1989 Characterization and serial propagation of mouse prostate epithelial cells in serum-free medium. Biol Cell 67:281–287
Gnanapragasam VJ, Robson CN, Neal DE, Leung HY 2002 Regulation of FGF8 expression by the androgen receptor in human prostate cancer. Oncogene 21:5069–5080
Yan G, Fukabori Y, Nikolaropoulos S, Wang F, McKeehan WL 1992 Heparin-binding keratinocyte growth factor is a candidate stromal-to-epithelial-cell andromedin. Mol Endocrinol 6:2123–2128
Kasai S, Sugimura K, Matsumoto K, Nishi N, Kishimoto T, Nakamura T 1996 Hepatocyte growth factor is a paracrine regulator of rat prostate epithelial growth. Biochem Biophys Res Commun 228:646–652
Peehl DM 2005 Primary cell cultures as models of prostate cancer development. Endocr Relat Cancer 12:19–47
Krongrad A, Wilson CM, Wilson JD, Allman DR, McPhaul MJ 1991 Androgen increases androgen receptor protein while decreasing receptor mRNA in LNCaP cells. Mol Cell Endocrinol 76:79–88
Claessens F, Celis L, Peeters B, Heyns W, Verhoeven G, Rombauts W 1989 Functional characterization of an androgen response element in the first intron of the C3(1) gene of prostatic binding protein. Biochem Biophys Res Commun 164:833–840
Arnold JT, Le H, McFann KK, Blackman MR 2005 Comparative effects of DHEA vs. testosterone, dihydrotestosterone, and estradiol on proliferation and gene expression in human LNCaP prostate cancer cells. Am J Physiol 288:E573–E584
Chen Q, Kaji H, Kanatani M, Sugimoto T, Chihara K 2004 Testosterone increases osteoprotegerin mRNA expression in mouse osteoblast cells. Horm Metab Res 36:674–678
Lu S, Tsai SY, Tsai MJ 1997 Regulation of androgen-dependent prostatic cancer cell growth: androgen regulation of CDK2, CDK4, and CKI p16 genes. Cancer Res 57:4511–4516
Pewitt EB, Haleem R, Wang Z 1999 Adrenomedullin gene is abundantly expressed and directly regulated by androgen in the rat ventral prostate. Endocrinology 140:2382–2386
Yang G, Gregory CW, Shang Q, O’Brien DA, Zhang YL 2001 Differential expression of CCAAT/enhancer-binding protein- (c/EBP) in rat androgen-dependent tissues and human prostate cancer. J Androl 22:471–480
Yang SZ, Abdulkadir SA 2003 Early growth response gene 1 modulates androgen receptor signaling in prostate carcinoma cells. J Biol Chem 278:39906–39911
Castoria G, Lombardi M, Barone MV, Bilancio A, Di Domenico M, Bottero D, Vitale F, Migliaccio A, Auricchio F 2003 Androgen-stimulated DNA synthesis and cytoskeletal changes in fibroblasts by a nontranscriptional receptor action. J Cell Biol 161:547–556
Moore MG, Wetterau LA, Francis MJ, Peehl DM, Cohen P 2003 Novel stimulatory role for insulin-like growth factor binding protein-2 in prostate cancer cells. Int J Cancer 105:14–19
Rajah R, Valentinis B, Cohen P 1997 Insulin-like growth factor (IGF)-binding protein-3 induces apoptosis and mediates the effects of transforming growth factor-1 on programmed cell death through a p53- and IGF-independent mechanism. J Biol Chem 272:12181–12188
Brigstock DR 2003 The CCN family: a new stimulus package. J Endocrinol 178:169–175
Yardy GW, Brewster SF 2005 Wnt signalling and prostate cancer. Prostate Cancer Prostatic Dis 8:119–126
Holen I, Croucher PI, Hamdy FC, Eaton CL 2002 Osteoprotegerin (OPG) is a survival factor for human prostate cancer cells. Cancer Res 62:1619–1623
Pearce JJ, Penny G, Rossant J 1999 A mouse cerberus/Dan-related gene family. Dev Biol 209:98–110
Thomas R, Anderson WA, Raman V, Reddi AH 1998 Androgen-dependent gene expression of bone morphogenetic protein 7 in mouse prostate. Prostate 37:236–245
Levedakou EN, Strohmeyer TG, Effert PJ, Liu ET 1992 Expression of the matrix Gla protein in urogenital malignancies. Int J Cancer 52:534–537
Sandhu C, Peehl DM, Slingerland J 2000 p16INK4A mediates cyclin dependent kinase 4 and 6 inhibition in senescent prostatic epithelial cells. Cancer Res 60:2616–2622
Isaacs JT, Furuya Y, Berges R 1994 The role of androgen in the regulation of programmed cell death/apoptosis in normal and malignant prostatic tissue. Semin Cancer Biol 5:391–400
Hoffman GE, Smith MS, Verbalis JG 1993 c-Fos and related immediate early gene products as markers of activity in neuroendocrine systems. Front Neuroendocrinol 14:173–213
Nagpal ML, Chen Y, Lin T 2004 Effects of overexpression of CXCL10 (cytokine-responsive gene-2) on MA-10 mouse Leydig tumor cell steroidogenesis and proliferation. J Endocrinol 183:585–594
Barila D, Plateroti M, Nobili F, Muda AO, Xie Y, Morimoto T, Perozzi G 1996 The Dri 42 gene, whose expression is up-regulated during epithelial differentiation, encodes a novel endoplasmic reticulum resident transmembrane protein. J Biol Chem 271:29928–29936
Sahu A, Lambris JD 2001 Structure and biology of complement protein C3, a connecting link between innate and acquired immunity. Immunol Rev 180:35–48
Cutolo M, Sulli A, Capellino S, Villaggio B, Montagna P, Seriolo B, Straub RH 2004 Sex hormones influence on the immune system: basic and clinical aspects in autoimmunity. Lupus 13:635–638
Culig Z, Klocker H, Bartsch G, Hobisch A 2002 Androgen receptors in prostate cancer. Endocr Relat Cancer 9:155–170
Sanford DC, DeWille JW 2005 C/EBP is a downstream mediator of IL-6 induced growth inhibition of prostate cancer cells. Prostate 63:143–154
Spiotto MT, Chung TD 2000 STAT3 mediates IL-6-induced growth inhibition in the human prostate cancer cell line LNCaP. Prostate 42:88–98
Lou W, Ni Z, Dyer K, Tweardy DJ, Gao AC 2000 Interleukin-6 induces prostate cancer cell growth accompanied by activation of stat3 signaling pathway. Prostate 42:239–242
Okamoto M, Webber MM, Quader S, Oyasu R 1998 Interleukin-6 and epidermal growth factor promote anchorage-independent growth of immortalized human prostatic epithelial cells treated with N-methyl-N-nitrosourea. Prostate 35:255–262
Mori S, Murakami-Mori K, Bonavida B 1999 Interleukin-6 induces G1 arrest through induction of p27Kip1, a cyclin-dependent kinase inhibitor, and neuron-like morphology in LNCaP prostate tumor cells. Biochem Biophys Res Commun 257:609–614
Lee SO, Lou W, Hou M, de Miguel F, Gerber L, Gao AC 2003 Interleukin-6 promotes androgen-independent growth in LNCaP human prostate cancer cells. Clin Cancer Res 9:370–376
Hobisch A, Eder IE, Putz T, Hominger W, Bartsch G, Klocker H, Culig Z 1998 Interleukin-6 regulates prostate-specific protein expression in prostate carcinoma cells by activation of the androgen receptor. Cancer Res 58:4640–4645
Zerbini LF, Wang Y, Cho JY, Libermann TA 2003 Constitutive activation of nuclear factor B p50/p65 and Fra-1 and JunD is essential for deregulated interleukin 6 expression in prostate cancer. Cancer Res 63:2206–2215
Thiel G, Cibelli G 2002 Regulation of life and death by the zinc finger transcription factor Egr-1. J Cell Physiol 193:287–292
Virolle T, Krones-Herzig A, Baron V, De Gregorio G, Adamson ED, Mercola D 2003 Egr1 promotes growth and survival of prostate cancer cells. Identification of novel Egr1 target genes. J Biol Chem 278:11802–11810(A. J. Asirvatham, M. Schmidt, B. Gao and)
Bioinformatics Core Laboratory (M.S., B.G.), Washington State University, Pullman, Washington 99164
Abstract
A major hurdle in understanding the role of androgens is the heterogeneity of androgen receptor (AR) expression in the prostate. Because the majority of prostate cancer arises from the AR-positive secretory luminal epithelial cells, identifying the androgen-mediated pathways in the prostate epithelium is of great significance to understanding their role in prostate pathogenesis. To meet this objective, the current study was designed to identify immediate-early genes expressed in response to the synthetic androgen R1881 in cultured rat ventral prostate epithelial cells. Rat ventral prostate epithelial cells, purified from 20-d-old rats, were cultured, and the presence of AR and the response to androgen were established. The cells were then treated with R1881 for 2 and 12 h to capture immediate-early genes in an Affymetrix-based gene chip platform. A total of 66 nonredundant genes were identified that were responsive to R1881. The functional androgen response elements were identified in the proximal promoter to determine possible molecular mechanism. Cluster analysis identified five distinct signatures of R1881-induced genes. Pathway analysis suggested that R1881 primarily influences cell proliferation/differentiation and inflammatory/immune response pathways. Androgens appear to regulate cell renewal by regulating differentiation, cell proliferation, and apoptosis. Two mutually exclusive inflammatory response pathways were observed. The interferon pathway was up-regulated, and the ILs were down-regulated. The data identified novel androgen-regulated genes (e.g. Id1, Id3, IL-6, IGF-binding protein-2 and -3, and JunB). The loss of androgen regulation of these genes can have important consequences for cellular transformation and transition to androgen-independent growth and survival.
Introduction
DESPITE THE WELL-documented role of androgens in prostate growth and function (1), the molecular pathways elicited by androgens that lead to prostate morphogenesis and pathogenesis are still largely unknown (2, 3, 4). One of the major hurdles in understanding the role of androgens is the heterogeneity of androgen receptor (AR) expression in the prostate. Most of the cell types in the prostate, such as luminal epithelial, basal epithelial, stromal, and smooth muscle cells, express ARs at some point of development and hence are capable of mediating androgen’s actions (5). In response to androgens, these cell types interact in an autocrine-paracrine manner to influence various aspects of prostate growth in normal and diseased states (1). This diverse expression profile has led to believe that androgens act as global endocrine mediators of prostate function.
The classical mechanism by which androgens exert their biological effect is through binding and subsequent activation of the AR. The androgen-bound AR then translocates to the nucleus and interacts with sequence-specific androgen response elements (ARE) in the promoters/enhancers/introns of target genes. The interaction between AR and ARE facilitates the assembly of general transcription machinery, leading to gene transcription (reviewed in Ref.6). Alternatively, AR activation may initiate cellular events leading to activation of latent transcription factors that can mediate or mimic the androgen response. This transcriptional or genomic mode of androgen action is independent of second messenger signal transduction cascades (7, 8).
The most dramatic effect of androgens on the prostate can be observed in castrated rats. After castration, the prostate undergoes involution accompanied by the loss of epithelial cells (9, 10, 11). Upon androgen administration, the prostate of the castrated rat undergoes rapid proliferation and cellular differentiation until it reaches its normal size and function (12). Large scale gene expression profiling has identified a number of genes and pathways, such as stress (13) and inflammatory/immune response, that are potentially involved in androgen-dependent prostate regeneration/morphogenesis (14, 15, 16, 17, 18, 19). Surprisingly, there has been little to no overlap between the genes regulated by androgens in these studies, possibly due to the age of rats used in each study and the timing and dose of androgen replacement after castration, but more importantly, the complex heterogeneity of AR expression itself. These models have significantly advanced our understanding of androgen-dependent prostate regeneration, but provide little evidence of androgen-regulated gene expression in specific cell types of the prostate. Of interest to us is the androgen-regulated gene expression specifically in prostate epithelial cells. A strong body of evidence suggests that the majority of prostate cancer arises from AR-positive secretory luminal epithelial cells (20, 21). For example, overexpression of AR specifically in the prostate epithelial cells leads to the development of prostatic intraepithelial neoplasia, the earliest recognizable stage of prostate cancer (22), and loss of AR reduces the risk of prostate carcinogenesis (23). Hence, AR expression and response to androgens are key determinants of prostate cancer initiation.
The current understanding of androgen-regulated gene expression in prostate epithelial cells is derived from the well-characterized androgen-responsive human prostate carcinoma cell line LNCaP and prostate cancer xenografts (24, 25, 26, 27, 28, 29, 30). It is apparent that these in vitro cancer models do not reflect the true molecular pathways elicited by androgens in normal prostate epithelial cells. Hence, the primary aim of this study was to identify the androgen-regulated gene expression profile and pathways in normal prostate epithelial cells. The epithelial cells were isolated from the rat ventral prostate cells (rVPEC), a representative model for hormone-dependent prostate hyperplasia (31) and cancer (32). The epithelial cells from these glands are hormone dependent, and other cell populations, such as basal and stromal cells, are essentially hormone independent (5). In culture and during initial passages, rVPEC also demonstrate functional androgen-dependent gene expression (33, 34, 35). The experimental design of the study allowed us to capture immediate early androgen-dependent events, specifically in prostate epithelial cells. The results suggest that the androgenic events in prostate epithelial cells are unique and are not necessarily represented in the whole prostate model. These androgen-mediated events, specifically in prostate epithelial cells, are of paramount importance for understanding the normal developmental events associated with the androgen response, and, more importantly, the significance of AR-regulated gene expression in prostate cancer.
Materials and Methods
Animals
Male Sprague Dawley rats, bred and raised in the Washington State University vivarium, were used. The university animal care committee approved all animal procedures performed in this study.
Culture of ventral prostate (VP) cells
The 20-d-old rats were killed by cervical dislocation, and VPs removed. The prostate epithelial cells were purified according to the published procedure, with some modifications (36). Briefly, the pooled tissue from 20 rats was incubated with Hanks’ balanced salt solution (HBSS; Invitrogen Life Technologies, Inc., Gaithersburg, MD) containing 7000 U collagenase type II (Sigma-Aldrich Corp., St. Louis, MO) and 3 mg deoxyribonuclease I/ ml (Sigma-Aldrich Corp.) at 37 C for 4 h with occasional shaking. After incubation, the tissue was allowed to gravity settle for 10 min. The supernatant, which consisted mainly of stromal cells, was removed. The pellet was resuspended in HBSS, and the mixture was spun at 30 x g for 4 min to pellet the epithelial cells. The pellet was resuspended in 1 ml HBSS and plated in Ham’s F-12 containing 10% fetal bovine serum with appropriate antibiotics (penicillin/streptomycin) and plated in 150-mm culture plates (Nunclon, Roskilde, Denmark). At passage 2 the cells were plated in 150-mm culture plates and grown to 80% confluence. The cells were then washed in serum-free medium and starved for 48 h. R1881 (1–100 nM) was added to fresh serum-free medium and added to the plates. RNA was collected after 0 (no treatment), 2, 6, 12, and 24 h of R1881 treatment and analyzed for fibroblast growth factor 8 (FGF8) expression. The optimum FGF8 expression was observed at 10 nM R1881. After this observation, the cells were treated with 10 nM R1881, and RNA was collected for microarray hybridizations. For quantitative PCR analysis of gene expression, the cells were also treated with the antiandrogen hydroxyflutamide (1 μm) in the presence of 10 nm R1881 for 2 and 12 h before harvesting for RNA isolation.
RNA preparation
Total RNA was obtained using the TRIzol (Invitrogen Life Technologies, Inc.) method described previously (36). The final RNA pellet was washed with 75% ethanol, dried, resuspended in diethylpyrocarbonate-treated H2O at a concentration of 1 mg/ml, and stored at –80 C until analysis.
RT-PCR
Total RNA (2 μg) was reverse transcribed in a final volume of 25 μl containing 20 U RNasin (Promega Corp., Madison, WI), 1.25 mM each of deoxy-NTPs, 250 ng oligo(deoxythymidine) (Pharmacia Biotech, Peapack, NJ), 10 mM dithiothreitol, and 200 U Moloney murine leukemia virus (MMLV) reverse transcriptase (Invitrogen Life Technologies, Inc.) in the MMLV first-strand synthesis buffer according to the manufacturer’s instructions. Each PCR contained 250 pg reverse transcribed DNA, 1 μM of each 5' and 3' oligonucleotide primer, 0.5 U Taq polymerase, and 0.1 mM of each deoxy-NTP. The primer pair sequence used was obtained from published sequences and is mentioned in the figure legends. The possible contamination of RNA with DNA was distinguished by performing the RT reaction without MMLV reverse transcriptase. The absence of any product in the amplification reaction using such a reverse transcribed preparation indicated the absence of any contaminating DNA in our RNA samples. Each RT reaction was performed using three different samples. The PCR-based amplification reactions were carried out in duplicate on each reverse transcribed RNA sample. Simultaneous PCRs were also carried out using primers designed to rat cyclophilin or S2 to monitor the efficiency of the RT-PCRs. Cyclophilin and S2 were faithfully amplified in all PCRs, indicating consistency in the quality of RT and PCRs. The identity of the corresponding PCR products was sequence confirmed by the Center for Reproductive Biology Molecular Biology Laboratory. The data presented are representative of at least three different RT-PCRs and DNA preparations.
Immunocytochemistry (ICC)
The rVPEC, LNCaP cells (positive control for AR), and DU145 cells (positive control for AR) were cultured in four-well Nunc cell culture slides to 80% confluence. The cells were treated with R1881 (10 nM) for 4 h before the ICC was performed. The previously established protocol was followed (37). Briefly, the cells were washed with 1x PBS twice and fixed in 10% neutral buffered formalin (Sigma-Aldrich Corp., HT509-1-2) for 10 min. After fixation, the cells were washed again with 1x PBS three times and then blocked in 10% goat serum for 1 h at room temperature. The cells were then incubated with primary AR (Upstate Biotechnology, Inc., Lake Placid, NY) or c-Met (Santa Cruz Biotechnology, Santa Cruz, CA) antibody for 1 h at 1:25, 1:50, and 1:100 dilutions. This was followed by washes in blocking solution. The cells were then incubated in secondary antibody [biotinylated goat anti rabbit IgG (H+L)] solution (1:300) for 1 h and subsequently washed three times with blocking solution. The well chambers were removed, and the horseradish peroxidase-conjugated streptavidin (Pierce Chemical Co., Rockford, IL) at a dilution of 1:3125 was added to the cells and incubated for 10 min. The 3,3'-diaminobenzidine staining protocol was then followed (peroxidase substrate kit, Vector Laboratories, Inc., Burlingame, CA). The color reaction was allowed to run for 5 min. After this, the nuclei were stained with hematoxylin for 30 sec. The cells were subsequently washed in water and increasing gradients of ethanol, then placed in xylene before mounting in a xylene-based mounting solution and observed under a microscope.
RNA quality for microarray
RNA quality was determined by both electrophoretic methods using a denaturing agarose gel and absorption readings at 260 and 280 nm. If excessive degradation or protein contamination of the RNA was evident, the sample was not used for microarray hybridization. A minimum OD (260/280) ratio of 1.8 was a requirement for microarray hybridization.
Microarray hybridizations
The Affymetrix gene chip platform, described previously (38), was used to determine transcriptional changes in the epithelial cells in response to androgens. Ten micrograms of total RNA from each of the samples was used to create the target for the microarray. The biotinylated CTP- and UTP-labeled cRNA were fragmented, hybridized to RAE230A arrays (Affymetrix, Santa Clara, CA), and stained in accordance with the manufacturer’s standard protocol. The arrays were stained and washed using the Affymetrix GeneChip Fluidics Station 400 and were scanned using a GeneArray Scanner 2500A (Agilent, Palo Alto, CA). The resulting data were viewed, and preliminary assessment was made using GCOS software (Affymetrix). All reactions and microarray hybridization procedures were performed at the Laboratory for Biotechnology and Bioanalysis I (Washington State University).
Absolute and statistical analysis for microarrays
Microarray output was examined visually for excessive background noise and physical anomalies. The default GCOS statistical values were used for all analyses. All probe sets on each array were scaled to a mean target signal intensity of 125, with the signal correlating to the amount of transcript in the sample. An absolute analysis using GCOS was performed to assess the relative abundance of the 45,000 represented transcripts based on signal and detection (present, absent, or marginal). Although 15,923 probe sets are represented on these arrays, there is a certain level of redundancy within and between each array, which results in a lower number of unique transcripts represented on the RAE230A chipset; however, the level of redundancy has not been firmly established.
The absolute analysis from GCOS was imported into GeneSpring 7.0 software (Silicon Genetics, Redwood City, CA). The R1881 treatment time-course data were normalized within GeneSpring using the default/recommended normalization methods. These included setting of signal values less than 0.01 to 0.01, total chip normalization to the 50th percentile, and normalization of each gene to the median. These normalizations allowed for the visualization of data based on relative abundance at any given time point rather than compared with a specific control value.
Data restrictions and analytical tools in GeneSpring were applied to isolate noteworthy and possibly important patterns of gene expression during the course of R1881 treatment on epithelial cells. Transcripts expressed differentially at a statistically significant level were determined using a P value cutoff of 0.05 and a Benjamini and Hochberg false discovery rate multiple testing correction. This was applied to all three time points and considered all transcripts represented on the arrays. Subsequently, expression restrictions were applied to the transcripts expressed in a significant manner. These restrictions were designed so that the remaining transcripts met the following requirements in addition to being expressed in a significant manner: 1) each transcript must have a signal value of at least 100 in a minimum of one of three points in the time course; and 2) the range of the replicates must not exceed 1 (in the normalized scale) in the time points comprising the time course. The resulting transcripts were screened using Excel (Microsoft, Redmond, WA) for redundant UniGene entries. Transcripts that passed these restrictions (top list) were considered for additional analysis, which included clustering and pathway analysis.
Quantitative PCR analysis
The relative gene expression levels of selected genes were determined by real-time quantitative PCR based on TaqMan chemistry. The primer probes [complement component 3 (C3), IL-6, TNF receptor superfamily/osteoprotegerin (Tnfrs 11b/OPG) 11b, WNT1 inducible signaling pathway protein 2 (WISP2), JunB, and -actin] were obtained from Applied Biosystems (TaqMan Probes, Applied Biosystems, Foster City, CA). The culture and treatment of cells were essentially similar to those described above. Total RNA was extracted from these cells using the TRIzol reagent and was reverse transcribed (2 μg/μl) using the oligo(deoxythymidine) protocol. Real-time quantitative PCR was used to measure the mRNA expression levels of the respective genes under various treatment conditions relative to -actin mRNA expression. All PCRs were performed in a final volume of 50 μl on an ABI PRISM 7700 detection system (Applied Biosystems) according to the manufacturer’s instructions. The following temperature profile was used: 10 min at 95 C, followed by 40 cycles of 15 sec at 95 C (denaturation step) and 1 min at 60 C (annealing/extension step). The cycle threshold (Ct) was used to calculate relative amounts of target RNA. Samples that did not reach the threshold line were considered not to be above background. All experiments were performed in triplicate and repeated three times.
The Ct method (Applied Biosystems User Bulletin 2, ABI PRISM 7700 detection system) was used for relative quantification of gene expression. The Ct values of the target genes from triplicate PCRs were normalized to the levels of -actin (endogenous control) from the same cDNA preparations. The average Ct for each gene is calculated by subtracting the Ct of the sample RNA from that of the control RNA for the same time measurement. This value is called the Ct, reflects the relative expression of the treated sample compared with the control, and becomes the exponent in the calculation for amplification 2Ct control – Ct sample, the equivalent to the fold change in expression. Statistically significant difference between treatments was determined by Student’s t test.
The semiquantitative PCR analysis was used to estimate changes in the levels of the established sequence tag (EST) AI071994 [dickkopf homolog 2 (DKK2)], induced by androgens.
Pathway analysis
The genes in the top list were classified according to their roles in regulating major cellular processes using the Affymetrix Netaffx database. A more detailed analysis of cellular function influenced by the genes in the top list was performed using the PathwayAssist (version 3.00) software (Stratagene, La Jolla, CA). The list was imported into pathway assist according to the manufacturer’s instructions. The software updated the list by excluding the redundant and nonannotated genes. This final list was then used to develop the common targets pathway for the identification of cell processes influenced by the genes in the list. Each identified cellular process was confirmed through the PubMed/Medline hyperlink embedded in each node.
Identification of potential ARE
The rat genome 9999 program built in GeneSpring 7.0 was used to find functional AREs. The rat genome 9999 software identified the promoter sequences of the genes used to build the homology table. Literature searches identified a consensus functional ARE sequence that was used to identify the AREs in the genes of interest that were regulated by R1881. The primary criteria for the identification of functional ARE was the presence of conserved C and G at positions 5 and 11, respectively. The functional ARE with associated redundancy is shown in Table 3. The rat genome searched 9999 upstream of the start site upstream in the promoter region for the consensus sequence.
Statistical analysis
The data were obtained from a minimum of three different experiments unless stated otherwise. The time-matched controls were used to rule out any time-dependent changes in control/treated sets in microarray analysis. Standard statistical tests for significance (t test) were used for semiquantitative and real-time PCR datasets.
Results
Characterization of rVPEC
The androgen target genes were identified in the epithelial cells isolated from VPs of 20-d-old rats. At this age (20 d postpartum), the cytosolic AR, particle-bound dihydrotestosterone (DHT), and uptake of [3H]DHT into the 100,000 x g sediment are maximum in prostate epithelial cells (39).
There have been many previous attempts to establish a reliable primary prostate culture system to study androgenic responsiveness. Some primary epithelial cultures yielded only a limited number of passages, and others yielded a down-regulation of AR expression upon culture (40, 41, 42). Based on these observations, we used the early passage (passage 2) prostate epithelial cells that maintained androgen responsiveness as described below. At passage 2, a homogeneous population of cells was obtained that consisted primarily of epithelial cells. The epithelial cells were characterized by the presence of AR and c-Met receptor (hepatocyte growth factor receptor) by PCR as previously described and ICC. As shown in Fig. 1, the purified cells expressed both receptors. The presence of AR and c-Met immunoreactivity in the cultured cells essentially confirmed the PCR data presented previously (36). ICC also confirmed that the majority of the cell population in culture at passage 2 was of epithelial origin. Additional characterization of these cells, i.e. epithelial vs. stromal or basal vs. luminal epithelial, was performed by identifying cell-specific markers in the gene expression data discussed below.
Response to androgens
The synthetic AR agonist R1881 (methyltrienolone) was used to evaluate the response of prostate epithelial cells to androgens in culture. Dose-response experiments were performed to optimize the effective concentration of R1881 (0.1, 10, and 100 nM) required to stimulate prostate epithelial cells. The expression of FGF8, a known androgen response gene in prostate epithelial cells (43), by semiquantitative PCR was used to monitor the response of epithelial cells to R1881. Maximum FGF8 gene expression was observed within 2 h of stimulation with R1881 at a concentration of 10 nM (Fig. 2). After 2 h, the expression of FGF8 was significantly reduced, but was still higher than that at 0 h. The gene expression changes in prostate epithelial cells were therefore captured within 2 and 12 h of treatment with R1881 (10 nM). This short time frame was used identify events that were immediately downstream of the activated AR and not due to secondary or tertiary responses that could have occurred at 24 h.
Gene expression profiling
The chip hybridizations (RAE230A, Affymetrix) were performed using RNA from at least two different sets of experiments, resulting in a minimum of two replicates for each time point to minimize random errors. Microarray data were processed with GCOS (Affymetrix) and exported into GeneSpring for additional analysis that consisted essentially of filtering the gene list over the specified restriction as outlined in Materials and Methods. A statistically significant change in gene expression does not necessarily dictate a large quantitative change. The data analysis was thus restricted to only those genes that had raw expression levels greater than 100 and changed by 2-fold or more in any of the conditions tested. Based on these restrictions, a total of 66 genes were found to be regulated by R1881 in rat prostate epithelial cells (Fig. 3).
Clustering of the genes based on Spearman’s correlation was performed to identify major expression profile signatures induced by R1881. As shown in Fig. 3, at least five major R1881-induced expression profiles were observed. The K-mean clustering using Spearman’s correlation of the regulated genes showed a similar profile.
Prostate epithelial cell-specific marker(s) expression
Because AR and c-Met expression determined by RT-PCR and ICC and the functional androgen response determined by semiquantitative RT-PCR of FGF8 expression confirmed that the cells used in this study were primarily epithelial, additional analysis based on gene expression profile was considered necessary. The gene expression profile in Table 1 shows the summary of this analysis. The genes normally expressed in the stromal compartment of the prostate, such as keratinocyte growth factor (44) and hepatocyte growth factor (45) were absent in the array data. This profile suggested that stromal cell contamination was negligible or below detection in the epithelial cell preparation. The gene expression was also analyzed to determine whether the epithelial cell population consisted of basal and/or luminal epithelial cells. The luminal epithelial-specific genes, cytokeratins 8 and 18, CD10, and CD24, and the basal epithelial-specific genes, BTG2 and 17-hydroxysteroid dehydrogenase, were highly expressed, as determined by their raw signal (>350) (reviewed in Ref.46). The predominance of luminal epithelial-expressed genes suggested that the epithelial cells were primarily luminal. Raw data analysis also showed the presence of AR mRNA (mean raw intensity, 95), which did not change with the treatment. Previous reports have shown a down-regulation of AR mRNA expression in LNCaP cell lines treated with androgens for more than 48 h (47). This result should be carefully compared with our data that show a lack of effect of androgens on AR mRNA possibly due to 1) relatively short treatment (12 h), and 2) the use of normal prostate epithelial cells rather than cancer cell lines. This gene expression profile was a significant observation that allowed interpretation of data in the context of androgen-regulated genes in the epithelial compartment of the VP.
Androgen-regulated genes in prostate epithelial cells
The immediate-early events elicited by R1881 were faithfully captured by the experimental design of this study. Twenty-seven (12 known and 14 ESTs) genes were up-regulated within 2 h of R1881 stimulation and returned to basal levels within 12 h (<2-fold change at 12 h; Fig. 4). Surprisingly, only eight genes (seven known and one EST) were up-regulated by R1881 after 12 h of treatment. These eight genes were unique compared with the 2 h up-regulated gene list. Twenty-eight genes (21 known and seven ESTs) were down-regulated by R1881 within 2 h of R1881 treatment (Fig. 4). Of the 10 genes (six known and four ESTs) that were down-regulated after 12 h of R1881 treatment, five genes were also down-regulated by R1881 after 2 h of treatment [IL-6, Id1, growth differentiation factor 8 (GDF8), and ESTs AB032395 and AI071994]. Collectively, this trend suggested a strong time-dependent transient expression mechanism.
Validation of androgen response and reliability of microarray analysis
The reliability of the microarray data shown in Fig. 3 is attested by the observations that 1) in instances where there were more than one oligonucleotide set for a particular gene (e.g. CCAAT/enhancer binding protein ), the clustering algorithm ordered these two probe sets adjacent to each other, or both the probe sets showed a similar profile (e.g. gap junction membrane channel protein 5); and 2) among the known transcripts that were influenced by androgens, there are several that have been previously reported to be rapidly induced or repressed in response to testosterone, DHT, or R1881. Consistent with these previous observations, the expression of C3 (48), IGF-binding protein-2 (IGFBP-2) (49), osteoprotegerin (OPG) (50), and tissue inhibitor of metalloprotease 3 (18) expression increased, whereas the expression of p16 (51), IGFBP-3 (14, 49), and CEBPd (14) decreased in response to R1881 in prostate epithelial cells. The exception to this list was the down-regulation of adrenomedullin in our experiment, which was previously shown to be up-regulated by androgens in the whole VP (15, 17, 52). The change in the expression of the above-mentioned, known androgen response markers in the prostate confirmed that the top list (Table 2) is the true representative of the immediate-early androgen-responsive genes in the rVPEC.
Additional confirmation of the microarray data was obtained using quantitative PCR on a subset of five genes from Table 2. The selected genes were representative of the complete dataset that was either up- or down-regulated after 2 or 12 h of treatment. The two other goals of the quantitative PCR were 1) to include additional replicate samples, and 2) confirm the androgen response by including the antiandrogen hydroxyflutamide in the culture medium. Overall, the expression profile of the selected genes, as determined by real-time PCR analysis (Fig. 5), was consistent with the microarray data shown in Fig. 3 and Table 2. The magnitude of fold changes observed in real-time PCR data were usually higher compared with microarray data, possibly due to differences in the techniques used for quantification in these methods. The addition of hydroxyflutamide also reversed the effect of R1881 on the expression of C3 and JunB (Fig. 5), demonstrating the specificity of the androgen response. Semiquantitative RT-PCR analysis was also performed for the EST AI071994, which showed a significant homology to the known human/mouse gene DKK2. This approach allowed us to visualize the gene expression that strengthened the microarray analysis of unknown transcripts. It is evident from Fig. 6 that treatment of prostate epithelial cells with androgen down-regulated the expression of this EST. Densitometric scanning and quantitation of RT-PCR analysis by normalizing with the constitutively expressed gene S2 also supported the RT-PCR data. Collectively, the postmicroarray validation of androgen response on selected genes reassured the authenticity of the androgen-regulated gene expression profile shown in Fig. 3 and Table 2.
Major genes/pathways influenced by androgens in prostate epithelial cells
Functional annotations of the known genes suggested that R1881 treatment influenced a variety of pathways, as listed in Table 2. For gene classification, the overall emphasis was on the major cellular process in which a particular gene was involved. Based on this classification, the primary cellular process influenced by R1881 was the inflammatory/immune response pathway. A similar observation that regulation of the inflammatory response pathway by androgens in intact prostate and prostate cancer cell lines was also reported previously (14, 28). However, the specific gene set reported in these studies was not identical with that reported in Table 2. This difference can be due to multiple reasons, for example, use of the whole prostate and/or cancer cell lines, but the significance of the overall regulation of the inflammatory/immune response pathway in prostate epithelial cells cannot be discounted and strengthens these observations. The use of cultured and relatively pure epithelial cell populations from the VP also suggests that this is the major pathway influenced by androgens.
The cellular processes influenced by the R1881-regulated genes shown in Table 2 were reevaluated by building the pathways in PathwayAssist software. This was considered necessary, because any single gene could influence multiple processes; hence, the information shown in Table 2 could be biased. An objective assessment of the major pathways was determined by the number of arrows pointing in or out of each cellular process (connectivity). Based on this analysis, the proliferation pathway (total number of connecting nodes or connectivity = 18) appeared to be the primary pathway influenced by androgens (Fig. 7). The genes (connecting nodes) involved in promoting proliferation, such as Id1, Id3, JunB, and IGFBP-3, decreased, and those inhibiting proliferation, for example, IGFBP-2 and 2',5'-oligoadenylate synthetase I (OAS1), increased, suggesting that, in general, the rate of proliferation decreased or was not affected after R1881 treatment. These data also supported our observation that R1881 has no effect of the proliferation of cultured VP epithelial cells during the course of treatment as determined by [3H]thymidine incorporation (data not shown). Concomitant with this was the observation that the differentiation process (connectivity = 13) increased. The decrease in the genes down-regulated during differentiation, such as Id1 and Id3, and the increase in IGFBP-2 and matrix Gla protein (MGP), which are known to promote differentiation, also support this analysis. Therefore, androgens appear to determine the intricate balance between proliferation and differentiation. The other pathways that were influenced by R1881 (in the order of decreasing connectivity) were apoptosis, immune response (immune response plus immunity), secretion, maturation, regulation of signal transduction, synthesis, pathogenesis, complement activation, metabolism, and DNA damage recognition.
Inflammatory/immune response pathway
The expression of 10 known genes in the inflammatory response pathway changed in response to R1881 treatment. Of these, six that were increased within 2 h of R1881 treatment were the CXC chemokine ligand-10 (CXCL-10), proteosome subunit- type 9 (PSMB9); interferon (IFN)-inducible, double-stranded RNA-dependent, protein kinase; RT1.S3, a nonclassical type 1 major histocompatibility complex (MHC) gene; OAS1; and myxovirus resistance protein-2 (Mx2). The expression of all these genes returned to pretreatment levels within 12 h, suggesting an immediate-early and transient expression in response to androgens.
The more than 6-fold increase in the expression of Mx2, a member of both the dynamin family and the family of large guanosine triphosphatases in response to R1881 was the largest. This protein is up-regulated by IFN-, but does not contain the antiviral activity of a similar Mx1. The expression of guanylate-binding protein-2 (Gbp2) and CXCL-10, which are also induced by IFNs, increased more than 3- and 2-fold, respectively, in response to androgens. In addition to Mx2, GBP-2, and CXCL-10, all other R1881 up-regulated genes within this pathway are also known to be regulated by IFNs. These results prompted us to investigate the expression of IFNs and other genes within the IFN pathway in the raw data. The expression of IFN- and - was observed (>100 raw intensity), but did not change significantly in response to R1881 treatment. A similar expression profile was observed for IFN pathway-associated genes, such as IFN-inducible protein-16, IFN regulatory factor-3, IFN--inducible protein-27-like, and IFN- receptor IFN regulatory factor-7. These results suggest that R1881 may mimic or modulate the IFN response in prostate epithelial cells. Because our data were also normalized against time-matched control values at 2 and 12 h, the likelihood of these effects being mediated through IFN expression and release during the course of the investigation in the absence of R1881 is negligible.
The expression of cytokines IL-6 and IL-12, the mediators of the inflammatory response pathway, was repressed by androgens. Functional pathway analysis suggested that the IL-6 node, with highest connectivity, could be the key mediator of the actions of R1881 in prostate epithelial cells.
Transcription factors
AR activation led to a decrease in the expression of transcription factors Id1, Id3, CEBPd, and early growth response-1 (Egr-1) and an increase (2.5-fold) in the expression of testis-specific histone-2a. The decrease in the expression of Id1 and Id3, the dominant-negative helix-loop-helix transcription factors, may support the role of androgens in initiating or maintaining differentiated functions (36). CEBPd activity plays a role in proliferation, differentiation, acute inflammatory response, and apoptosis depending on the cell type and specific physiological response. In rat VP, androgens also down-regulate CEBPd levels (53). The down-regulation of Egr-1 transcription factor is a significant observation, because one of the mechanisms that may lead to the development of aggressive prostate cancer is loss of androgen-mediated transcriptional regulation of Egr-1 (54).
Signal transduction
Significant changes in the expression of signal transduction genes were not observed after R1881 treatment of prostate epithelial cells. This observation is not unexpected because 1) the primary events elicited by androgens, at least in the current model system, appear to be mediated by cytosolic receptor and not through the proposed membrane-bound receptor (55); and 2) the signal transduction genes are expected to be phosphorylated, which is not manifested as a change in expression. However, the raw data were analyzed for the expression of common signal transduction genes. Among these, a raw signal intensity of 100 or greater for phosphatidylinositol 4-kinase, serum-inducible kinase, MAPK6, diacylglycerol kinase (), serine-threonine kinase 2, ERK1b, MAPK4, MAP2k5, Rho-associated, coiled-coil containing protein kinase 2 (ROCK2), mothers against DPP homolog 4 (Drosophila) (SMAD4), and cAMP response element-binding protein-1 was observed.
Regulation of growth
The expression of IGFBP-3 decreased in the presence of androgen within 2 h, but returned to basal levels by 12 h. The expression of IGFBP-2 increased in response to androgens after 12 h of treatment. In the normal prostate epithelial cells, IGFBP-2 may act as a growth inhibitor (56). Hence, androgen-mediated up-regulation of IGFBP-2 may limit the growth of prostate epithelial cells. In the castrated rat model, androgens were shown to down-regulate IGFBP-3 expression. Our data suggest that down-regulation of IGFBP-3 by androgens occurs at the level of prostate epithelial cells immediately downstream of AR activation and may protect cells from undergoing apoptosis (57). Analysis of the raw data indicated high and constitutive levels of TGF-3 (raw signal, 3600). The constitutive expression of TGF-1 (raw signal, 97), TGF- (raw signal, 70), IGF-I (raw signal, 302), and IGF-II receptor (raw signal, 654) was also observed. Collectively, these observations, for example, high levels of TGF-3, may also explain the low rate of proliferation of prostate epithelial cells.
The expression of MGP, an extracellular matrix protein that acts as an inhibitor of pathological calcification, was down-regulated by androgens. In androgen-dependent Shionogi medullary carcinoma 115 cells, the expression of MGP is also negatively regulated by androgens (13), thus confirming the androgen-mediated response observed in our studies. However, the relevance of MGP is not known, but detailed analysis of its significance during the transformation of prostate epithelial cells may provide evidence of the process of microcalcification observed in prostate cancer.
WISP-2 is a growth arrest-specific gene in vascular smooth muscle cells and is a component of the WNT1 signaling pathway (58). Thus, up-regulation of WISP-2 expression does support the role of androgens as growth inhibitors. The expression of WISP-2 may also have broader implications on prostate biology: for example, as a component of the WNT signaling pathway that is aberrantly activated in prostate cancer (59) and as a secreted factor possibly involved in paracrine/autocrine interactions.
Transport activity
A decrease in the expression of monocarboxylate transporter, fatty acid-binding protein 3, retinol-binding protein 1 (cellular), and lipocalin 2 and an increase in the expression of peptidyl isomerase (Ppicap) were observed in response to androgens.
Apoptosis
Tnfrs11b (OPG), a decoy receptor for TNF-related apoptosis-inducing ligand that inhibits TNF-related apoptosis-inducing ligand-induced apoptosis, was the only apoptosis-related gene identified in our dataset. The role of androgens in the regulation of OPG in prostate epithelial cells or cancer cell lines is not known, but androgen-independent cell lines secrete significantly more OPG than androgen-responsive LNCaP cell lines (60). In the osteoblast cells, the androgens specifically inhibit OPG mRNA levels and protein secretion (50). In contrast to these studies, the observed decrease in OPG levels produced by androgens may suggest a cell-specific, androgen-mediated regulatory mechanism.
Cell-cell signaling
The expression of late gestation lung protein-1 (Lgl1) in prostate epithelial cells increased after 12 h of treatment with R1881. Lgl1 is a novel, glucocorticoid-inducible gene that regulates mesenchymal-epithelial interactions. Interestingly, the expression of cysteine knot superfamily, homolog (Xenopus laevis) Cktsf1b1 (gremlin), an extracellular bone morphogenic protein (BMP) antagonist (61) also involved in mesenchymal-epithelial interactions, decreased in the presence of R1881. It is speculated that the opposing role of androgens on germlin (decrease) and BMP (increase) (62) may regulate the bioactivity of BMP, thus influencing prostate growth and development. Together, Lgl1 and Cktsf1b1 may play a role in prostate branching morphogenesis (63).
Adrenomedullin (AM) is hypotensive peptide highly expressed in the rat prostate ventral epithelium and is a direct androgen-responsive gene (15). In situ hybridization of normal rat prostate tissue showed that AM expression is localized in the epithelial cells (52). Our results are contradictory to these reports and suggest that AM is down-regulated in prostate epithelial cells after a brief exposure to androgens.
Cell cycle
A decrease in the expression of the cyclin-dependent kinase inhibitor (CDKN) 2A (p16INK4) tumor suppressor gene in response to androgens was observed. CDKN2A functions as an inhibitor of cyclin-dependent kinase (CDK) 4 and CDK6, which initiate phosphorylation of the retinoblastoma tumor suppressor protein. Thus, a decrease in CDKN2A can promote the cell cycle. Its probable physiological role is in the implementation of irreversible growth arrest or cellular senescence that can be induced by stresses such as DNA damage and aberrant mitogenic signaling in human primary cells. The observed decrease in CDKN2A may support the role of androgens in maintaining proliferation and delaying replicative senescence, at least in the early-passage cells used in this study (64). Previously, androgens have also been shown to down-regulate p16 (51),
Identification of ARE in androgen-regulated genes
The experimental design of this study allowed us to identify immediate downstream targets of androgens in rVPEC. The change in the expression of genes (Table 2) is anticipated to be under direct androgen control, although a change in the expression after 12 h of treatment can be indirect due to subsequent secondary or even tertiary events elicited by AR activation. One of the approaches that may support the direct regulation of gene transcription shown in Table 2 by the activated AR is to determine the presence/absence of potential ARE in the promoter of androgen-regulated genes as outlined in Materials and Methods.
The GeneSpring (rat genome 9999) software allowed us to identify ARE within 9999 bp of the predicted transcriptional start site. However, we focused on the ARE within the proximal 1000 bp because these proximal AREs are more likely to have a regulatory role. The rat genome 9999 software could identify the promoter sequence in 40 of the 66 genes listed in Table 2. Of the 40 genes for which promoter sequences were available, the ARE consensus sequence was present in 23 genes (Table 3). The presence of the ARE in the majority of the androgen-regulated genes identified in this study is therefore a significant observation and warrants additional experimental confirmation. The functional ARE in C3, a known androgen-responsive gene, is in the first intron (48) and, hence, was not included in the list (Table 3). Thus, including introns in the search may provide evidence for the presence of functional ARE in other androgen-regulated genes listed in Table 2.
Fifteen of the 23 genes with putative ARE (Table 3) were down-regulated by androgens within 2 h of treatment. Interestingly, of these eight genes, the expression of four genes is up-regulated only after 12 h of treatment. The significance of this observation is not known, but it can be speculated that androgens appear to actively down-regulate the majority of the genes, possibly through the ARE directly, whereas up-regulation may involve secondary events, such as the expression/phosphorylation of novel accessory proteins.
Discussion
The role of androgens in regulating normal prostatic growth, development, and function is complex and involves interactions between various cell types that is mediated by paracrine and autocrine interactions involving growth factors and cytokines. In the pubertal prostate, the androgens primarily regulate the differentiated functions, but also maintain epithelial cell numbers by regulating basal rates of proliferation and apoptosis (65). The complex heterogeneity of the prostate makes it difficult to understand the specific mechanism by which androgens regulate epithelial cell function. Understanding these mechanisms is of significant interest because 99% of prostate cancer arises from these AR-positive epithelial cells. Our dataset on the androgen-regulated genes in prostate epithelial cells therefore assumes greater significance in understanding the cellular pathways influenced by androgen in the cell type that eventually becomes transformed.
Androgens regulate a significantly small fraction of genes in prostate epithelial cells compared with the whole prostate gland. This is obvious, because diverse cell types within the prostate express AR and respond to androgens. More importantly, the lack of paracrine interactions and brief exposure to androgens may have also resulted in the small number of androgen-regulated genes identified in this study. The presence of known androgen-regulated genes in our dataset is consistent with the published literature. Our results therefore confirm many investigations of the androgen-regulated genes despite the fact that there was a minimal overlap of the results independently compared with data obtained through various large-scale profiling techniques, such as subtractive hybridization, SAGE (serial analysis of gene expression), and microarray, reported in the literature (13, 14, 15, 16, 17, 18, 19, 24, 25, 26, 27, 28, 29, 30).
A significant observation was the lack genes involved in major signal transduction pathway, possibly due to the absence of secondary paracrine events and the predominant genomic actions of androgens. If androgens were to act through nongenomic pathways, then the 2 and 12 h points were long enough to capture membrane receptor/nongenomic events (66).
Based on published results and data obtained in this study, it seems that the androgens influence at least two major pathways in prostate epithelial cells. These include the inflammatory/immune response and the cell survival pathway (broadly involving apoptosis, proliferation, and differentiation).
The down-regulation of JunB, Id1, and Id3, the positive regulators of the cell cycle, also supports the absence of an overall effect of androgens on proliferation. Apart from these classical cell cycle markers, an increase in WISP2, a negative regulator of the cell cycle (58), and the overall up-regulation of the IFN pathway, previously shown as an androgen-dependent proliferation inhibitory pathway in transformed tumorigenic benign prostatic hyperplasia (BPH1; CAFTD) compared with parental nontumorigenic BPH1 cells (28), also qualify our results. The decrease in CDKN2A, the inhibitor of CDK, and increases in CXC chemokine ligand 10 may sustain basal proliferation of prostate epithelial cells in response to androgens (64, 67). Thus, it appears that androgens regulate the self-renewal in which neither regression nor overgrowth of prostate epithelium occurs, thus recapitulating the in vivo effects of androgens on prostate epithelial cells (4). It is speculated that the down-regulation of CDKN2A may also lead to the delaying of cellular senescence in vitro (64). The gene expression profiling suggests that androgens may support the differentiation of prostate epithelial cells in culture, as is evident from the decreased expression of Id1 and Id3 and the increased expression of C3 and Ppap2b (endoplasmic reticulum transmembrane protein Dri42) (68).
The inflammatory response pathway may be elicited by the components of the IFN pathway, such as guanosine triphosphatases (Mx2), Gbp2, protein catabolism [proteosome (prosome, macropain) subunit, type 9], and chemokines (CXC chemokine ligand 10). A previous report that C3, a component of the complement activation system of the immune response pathway, is a direct androgen target independently confirms these observations (69).
In general, androgens are known to suppress the immune response, although the exact mechanism is not known (70). In the VP, androgen withdrawal induces the immune/inflammatory genes IL-15 and IL-18, possibly due to influx of T cells, macrophages, and mast cells (14). In the current in vitro model, androgens also appear to modulate the inflammatory response, but specifically in epithelial cells. At least two androgen-dependent immune/inflammatory gene signatures are observed in prostate epithelial cells. The first is the general up-regulation of the IFN pathway such as CXCL10; IFN-inducible, double-stranded RNA-dependent, protein kinase; Mx2; proteosome subunit- type 9; OAS1; and RT1.S3. The second is the down-regulation of IL-6 and IL-12.
The up-regulation of the genes within the IFN pathway by androgens can have important consequences for the immunosurveillance. Loss of the immunosurveillance or onset of immune senescence can lead to BPH and cancer (20). For example, the IFN pathway that is generally suppressed in cancer is also suppressed by androgens in transformed tumorigenic BPH (CAFTD) cells (28). Hence, our data provide direct evidence that in normal cells, androgens positively regulate the immune surveillance pathway.
The down-regulation of IL-6 by R1881 observed in this study is particularly interesting. IL-6 is a multifunctional cytokine that acts as a survival factor in carcinoma of the prostate. IL-6 can stimulate the proliferation of prostate epithelial cells in culture (71, 72), but in LNCaP cells, growth stimulatory and inhibitory effects have been reported (73, 74, 75, 76). Of significance, however, is the crosstalk between IL-6 and AR. IL-6 can induce ligand-independent and synergistic activation of AR, possibly mediated through the MAPK pathway (77). In the absence of androgen, IL-6 leads to increased prostate-specific antigen mRNA levels and activates several androgen-responsive promoters, but not the nonandrogen-responsive promoters, in LNCaP cells (78). These data indicate the IL-6 induces an androgen response in prostate cancer cells through the AR even in the absence of androgens (77, 78). Hence, IL-6 can promote prostate cancer progression through induction of an androgen response.
The decrease in IL-6 can also be due to down-regulation of CEBPd expression through a positive feedback loop. CEBPd is a known androgen-repressed gene and IL-6 positively regulates CEBPd expression (72). Hence, a decrease in CEBPd after androgen treatment can also lead to decreased levels of IL-6. Alternatively, a direct androgen-mediated repression of IL-6 may also lead to the decreased CEBPd expression. Thus, our studies for the first time have established a link between IL-6, CEBPd, and androgens in normal prostate epithelial cells. The down-regulation of IL-6 by androgens directly or indirectly through CEBPd can serve at least three functions: 1) blocking proliferation, 2) suppressing the immune response directly or through CEBPd, and 3) suppression of androgen-independent expression of AR-responsive genes, thus allowing these genes to be expressed only in the presence of physiological levels of androgens. During transition to hormone-independent cancer, the inhibitory effect of androgens on IL-6 may be lost, thus allowing IL-6 to promote proliferation and induce the expression of AR-responsive genes and up-regulation of CEBPd. Deregulation of CEBPd and IL-6 also occurs when prostate cancer progresses to the androgen-independent state (79).
Consistent with the effects of androgens on other transcription factors, such as Id1, Id3, and CEBPd, a down-regulation in the expression of Egr-1 was also observed. Egr-1 is a crucial regulator of cell growth, differentiation, and survival (80). Egr-1 expression is known to be increased in prostate cancer and is constitutively expressed in hormone-independent cancers. Egr-1 plays a functional role in the transformed phenotype and may represent a valid target for prostate cancer therapy (81). Our results suggest that androgens negatively regulate the expression of Egr-1 in normal cells. This is a novel observation and suggests that androgen-mediated down-regulation of Egr-1 in normal prostate epithelium may be involved in growth regulation and differentiation.
IGFBPs, particularly IGFBP-3, also have IGF-independent, antiproliferative, and proapoptotic functions (57). After castration, the levels of IGFBP-3 increase, but they decrease after androgen administration in the whole prostate, suggesting that IGFBP-3 expression is androgen dependent, negatively regulates growth, and promotes apoptosis. The androgen-dependent decrease in IGFBP-3 in our in vitro analysis thus supports the role of androgens in blocking apoptosis, but promoting the growth that is required for cell renewal. IGFBP-2 is the major androgen-regulated IGFBP in prostatic fluid and has a modest suppressive effect on the growth of normal prostate epithelial cells (56). An increase in IGFBP-2 may also explain the low growth of epithelial cells in culture, but, more importantly, supports the significance of the androgen-responsive genes obtained from an in vitro system.
A number of other genes that were regulated by androgens, listed in Table 2, can be used to develop a broader understanding of how androgens influence prostate epithelial cell functions. These include enzymes involved in lipid metabolism and transport, cell-cell signaling, and growth factors that may unfold new pathways in cell-cell interactions within the prostate, as outlined in Results.
The identification of functional AREs in most of the genes for which the promoter sequences were available strengthens the results of this study. The presence of conserved cytosine at 5 bp and guanine at 11 bp and compliance with the known heterogeneity with the consensus ARE (17) strongly support the possibility that the identified genes are direct androgen targets. The presence of known androgen-regulated genes, such as MGP (13) and Adm (17), validates this analysis. A number of additional genes regulated by androgens, possibly through ARE, were also identified. These include Id3, Egr-1, growth differentiation factor 8 (GDF8), IL-6, and Mx2 among others. Functional analysis of the significance of these ARE sites on the expression of the individual genes listed in Table 2 will help in understanding the mechanism of action of androgens in normal and pathogenic prostates.
The major pathways influenced by androgen are expected to form the basis for future investigations to gain a comprehensive understanding of prostate epithelial cell function. The development of similar pathways using more defined, androgen-dependent and independent in vitro and in vivo models is expected to unravel the molecular events associated with the transformation of prostate epithelial cells.
In conclusion, our results have for the first time identified a set of novel immediate-early androgen-regulated genes in prostate epithelial cells. The genes that are normally regulated by androgens were part of this dataset, thus independently validating our results. Based on the combined data analysis presented here and published elsewhere, it seems that androgens regulate the cell survival and immune response pathway with a significant crosstalk. We propose that at the core of this process are genes such as Egr-1, CEBPd, IL-6, IGFBPs, and Id. The regulated expression of these genes is required for the maintenance on normal prostate epithelial cell homeostasis. Loss of androgenic control may lead to altered expression profiles of these genes that may contribute to initiation, maintenance, or progression toward hormone-independent prostate cancer. Although the significance of other androgen-regulated genes in Table 2 cannot be discounted, detailed analysis will be required to gain a better understanding of androgen-mediated events in the prostate by these genes.
Footnotes
This work was supported by the Center for Reproductive Biology, Washington State University.
First Published Online September 29, 2005
Abbreviations: AM, Adrenomedullin; AR, androgen receptor; ARE, androgen response element; BMP, bone morphogenic protein; BPH, benign prostatic hyperplasia; C3, complement component 3; CDK, cyclin-dependent kinase; CDKN, CDK inhibitor; Ct, cycle threshold; CXCL-10, CXC chemokine ligand-10; DHT, dihydrotestosterone; DKK2, dickkopf homolog 2; Egr-1, early growth response-1; EST, established sequence tag; FGF8, fibroblast growth factor 8; GBP2, guanylate-binding protein-2; HBSS, Hanks’ balanced salt solution; ICC, immunocytochemistry; IFN, interferon; IGFBP, IGF-binding protein; Lgl1, late gestation lung protein-1; MGP, matrix Gla protein; MMLV, Moloney murine leukemia virus; Mx, myxovirus resistance protein; OAS1, 2',5'-oligoadenylate synthetase-I; OPG, osteoprotegerin; rVPEC, rat ventral prostate epithelial cell; VP, ventral prostate; WISP2, WNT1 inducible signaling pathway protein 2.
Accepted for publication September 19, 2005.
References
Cunha GR, Donjacour AA, Cooke PS, Mee S, Bigsby RM, Higgins SJ, Sugimura Y 1987 The endocrinology and developmental biology of the prostate. Endocr Rev 8:338–362
Marker PC, Donjacour AA, Dahiya R, Cunha GR 2003 Hormonal, cellular, and molecular control of prostatic development. Dev Biol 253:165–174
Abate-Shen C, Shen MM 2000 Molecular genetics of prostate cancer. Genes Dev 14:2410–2434
Heinlein CA, Chang C 2004 Androgen receptor in prostate cancer. Endocr Rev 25:276–308
Prins GS, Birch L, Greene GL 1991 Androgen receptor localization in different cell types of the adult rat prostate. Endocrinology 129:3187–3199
Lee DK, Chang C 2003 Molecular communication between androgen receptor and general transcription machinery. J Steroid Biochem Mol Biol 84:41–49
Heinlein CA, Chang C 2002 Androgen receptor (AR) coregulators: an overview. Endocr Rev 23:175–200
Heinlein CA, Chang C 2002 The roles of androgen receptors and androgen-binding proteins in nongenomic androgen actions. Mol Endocrinol 16:2181–2187
Kyprianou N, Isaacs JT 1988 Activation of programmed cell death in the rat ventral prostate after castration. Endocrinology 122:552–562
Kwong J, Choi HL, Huang Y, Chan FL 1999 Ultrastructural and biochemical observations on the early changes in apoptotic epithelial cells of the rat prostate induced by castration. Cell Tissue Res 298:123–136
Banerjee S, Banerjee PP, Brown TR 2000 Castration-induced apoptotic cell death in the Brown Norway rat prostate decreases as a function of age. Endocrinology 141:821–832
Evans GS, Chandler JA 1987 Cell proliferation studies in the rat prostate. II. The effects of castration and androgen-induced regeneration upon basal and secretory cell proliferation. Prostate 11:339–351
Pang ST, Dillner K, Wu X, Pousette A, Norstedt G, Flores-Morales A 2002 Gene expression profiling of androgen deficiency predicts a pathway of prostate apoptosis that involves genes related to oxidative stress. Endocrinology 143:4897–4906
Desai KV, Michalowska AM, Kondaiah P, Ward JM, Shih JH, Green JE 2004 Gene expression profiling identifies a unique androgen-mediated inflammatory/immune signature and a PTEN (phosphatase and tensin homolog deleted on chromosome 10)-mediated apoptotic response specific to the rat ventral prostate. Mol Endocrinol 18:2895–2907
Wang Z, Tufts R, Haleem R, Cai X 1997 Genes regulated by androgen in the rat ventral prostate. Proc Natl Acad Sci USA 94:12999–13004
Furuya Y, Isaacs JT 1993 Differential gene regulation during programmed death (apoptosis) versus proliferation of prostatic glandular cells induced by androgen manipulation. Endocrinology 133:2660–2666
Nantermet PV, Xu J, Yu Y, Hodor P, Holder D, Adamski S, Gentile MA, Kimmel DB, Harada S, Gerhold D, Freedman LP, Ray WJ 2004 Identification of genetic pathways activated by the androgen receptor during the induction of proliferation in the ventral prostate gland. J Biol Chem 279:1310–1322
Jiang F, Wang Z 2003 Identification of androgen-responsive genes in the rat ventral prostate by complementary deoxyribonucleic acid subtraction and microarray. Endocrinology 144:1257–1265
Porkka KP, Visakorpi T 2001 Detection of differentially expressed genes in prostate cancer by combining suppression subtractive hybridization and cDNA library array. J Pathol 193:73–79
De Marzo AM, Meeker AK, Zha S, Luo J, Nakayama M, Platz EA, Isaacs WB, Nelson WG 2003 Human prostate cancer precursors and pathobiology. Urology 62:55–62
Okada H, Tsubura A, Okamura A, Senzaki H, Naka Y, Komatz Y, Morii S 1992 Keratin profiles in normal/hyperplastic prostates and prostate carcinoma. Virchows Arch A Pathol Anat Histopathol 421:157–161
Stanbrough M, Leav I, Kwan PW, Bubley GJ, Balk SP 2001 Prostatic intraepithelial neoplasia in mice expressing an androgen receptor transgene in prostate epithelium. Proc Natl Acad Sci USA 98:10823–10828
Eng MH, Charles LG, Ross BD, Chrisp CE, Pienta KJ, Greenberg MN, Hsu CX, Sanda MG 1999 Early castration reduces prostatic carcinogenesis in transgenic mice. Urology 54:1112–1119
Nelson PS, Clegg N, Arnold H, Ferguson C, Bonham M, White J, Hood L, Lin B 2002 The program of androgen-responsive genes in neoplastic prostate epithelium. Proc Natl Acad Sci USA 99:11890–11895
Velasco AM, Gillis KA, Li Y, Brown EL, Sadler TM, Achilleos M, Greenberger LM, Frost P, Bai W, Zhang Y 2004 Identification and validation of novel androgen-regulated genes in prostate cancer. Endocrinology 145:3913–3924
Oosterhoff JK, Grootegoed JA, Blok LJ 2005 Expression profiling of androgen-dependent and -independent LNCaP cells: EGF versus androgen signalling. Endocr Relat Cancer 12:135–148
DePrimo SE, Diehn M, Nelson JB, Reiter RE, Matese J, Fero M, Tibshirani R, Brown PO, Brooks JD 2002 Transcriptional programs activated by exposure of human prostate cancer cells to androgen. Genome Biol 3:RESEARCH0032
Shou J, Soriano R, Hayward SW, Cunha GR, Williams PM, Gao WQ 2002 Expression profiling of a human cell line model of prostatic cancer reveals a direct involvement of interferon signaling in prostate tumor progression. Proc Natl Acad Sci USA 99:2830–2835
Amler LC, Agus DB, LeDuc C, Sapinoso ML, Fox WD, Kern S, Lee D, Wang V, Leysens M, Higgins B, Martin J, Gerald W, Dracopoli N, Cordon-Cardo C, Scher HI, Hampton GM 2000 Dysregulated expression of androgen-responsive and nonresponsive genes in the androgen-independent prostate cancer xenograft model CWR22–R1. Cancer Res 60:6134–6141
Xu LL, Su YP, Labiche R, Segawa T, Shanmugam MN, McLeod DG, Moul JW, Srivastava S 2001 Quantitative expression profile of androgen-regulated genes in prostate cancer cells and identification of prostate-specific genes. Int J Cancer 92:322–328
Fujimoto N, Suzuki T, Honda H, Kitamura S 2004 Estrogen enhancement of androgen-responsive gene expression in hormone-induced hyperplasia in the ventral prostate of F344 rats. Cancer Sci 95:711–725
Homma Y, Kondo Y, Kaneko M, Kitamura T, Nyou WT, Yanagisawa M, Yamamoto Y, Kakizoe T 2004 Promotion of carcinogenesis and oxidative stress by dietary cholesterol in rat prostate. Carcinogenesis 25:1011–1014
Rhim JS 2003 Research into molecular and genetic mechanisms underlying prostate carcinogenesis would be greatly advanced by in vitro prostate cell models. Drugs Today 39:837–847
Cussenot O, Berthon P, Cochand-Priollet B, Maitland NJ, Le Duc A 1994 Immunocytochemical comparison of cultured normal epithelial prostatic cells with prostatic tissue sections. Exp Cell Res 214:83–92
Ilio KY, Sensibar JA, Lee C 1995 Effect of TGF-1, TGF-, and EGF on cell proliferation and cell death in rat ventral prostatic epithelial cells in culture. J Androl 16:482–490
Chaudhary J, Schmidt M, Sadler-Riggleman I 2005 Negative acting HLH proteins Id1, Id2, Id3, and Id4 are expressed in prostate epithelial cells. Prostate 64:253–264
Hill CM, Anway MD, Zirkin BR, Brown TR 2004 Intratesticular androgen levels, androgen receptor localization, and androgen receptor expression in adult rat Sertoli cells. Biol Reprod 71:1348–1358
Li Y, Friel PJ, Robinson MO, McLean DJ, Griswold MD 2002 Identification and characterization of testis- and epididymis-specific genes: cystatin SC and cystatin TE-1. Biol Reprod 67:1872–1880
Moeller H, Blank B, Lander K, Mates G 1987 Ontogeny of the androgen receptor in rat ventral prostate during sexual development. Res Exp Med (Berl) 187:287–294
Chopra DP, Grignon DJ, Joiakim A, Mathieu PA, Mohamed M, Sakr WA, Powell IJ, Sarkar FH 1996 Differential growth factor responses of epithelial cell cultures derived from normal human prostate, benign prostatic hyperplasia, and primary prostate carcinoma. J Cell Physiol 169:269–280
Chaproniere DM, McKeehan WL 1986 Serial culture of single adult human prostatic epithelial cells in serum-free medium containing low calcium and a new growth factor from bovine brain. Cancer Res 46:819–824
Chowdhury A, Harber GJ, Chopra DP 1989 Characterization and serial propagation of mouse prostate epithelial cells in serum-free medium. Biol Cell 67:281–287
Gnanapragasam VJ, Robson CN, Neal DE, Leung HY 2002 Regulation of FGF8 expression by the androgen receptor in human prostate cancer. Oncogene 21:5069–5080
Yan G, Fukabori Y, Nikolaropoulos S, Wang F, McKeehan WL 1992 Heparin-binding keratinocyte growth factor is a candidate stromal-to-epithelial-cell andromedin. Mol Endocrinol 6:2123–2128
Kasai S, Sugimura K, Matsumoto K, Nishi N, Kishimoto T, Nakamura T 1996 Hepatocyte growth factor is a paracrine regulator of rat prostate epithelial growth. Biochem Biophys Res Commun 228:646–652
Peehl DM 2005 Primary cell cultures as models of prostate cancer development. Endocr Relat Cancer 12:19–47
Krongrad A, Wilson CM, Wilson JD, Allman DR, McPhaul MJ 1991 Androgen increases androgen receptor protein while decreasing receptor mRNA in LNCaP cells. Mol Cell Endocrinol 76:79–88
Claessens F, Celis L, Peeters B, Heyns W, Verhoeven G, Rombauts W 1989 Functional characterization of an androgen response element in the first intron of the C3(1) gene of prostatic binding protein. Biochem Biophys Res Commun 164:833–840
Arnold JT, Le H, McFann KK, Blackman MR 2005 Comparative effects of DHEA vs. testosterone, dihydrotestosterone, and estradiol on proliferation and gene expression in human LNCaP prostate cancer cells. Am J Physiol 288:E573–E584
Chen Q, Kaji H, Kanatani M, Sugimoto T, Chihara K 2004 Testosterone increases osteoprotegerin mRNA expression in mouse osteoblast cells. Horm Metab Res 36:674–678
Lu S, Tsai SY, Tsai MJ 1997 Regulation of androgen-dependent prostatic cancer cell growth: androgen regulation of CDK2, CDK4, and CKI p16 genes. Cancer Res 57:4511–4516
Pewitt EB, Haleem R, Wang Z 1999 Adrenomedullin gene is abundantly expressed and directly regulated by androgen in the rat ventral prostate. Endocrinology 140:2382–2386
Yang G, Gregory CW, Shang Q, O’Brien DA, Zhang YL 2001 Differential expression of CCAAT/enhancer-binding protein- (c/EBP) in rat androgen-dependent tissues and human prostate cancer. J Androl 22:471–480
Yang SZ, Abdulkadir SA 2003 Early growth response gene 1 modulates androgen receptor signaling in prostate carcinoma cells. J Biol Chem 278:39906–39911
Castoria G, Lombardi M, Barone MV, Bilancio A, Di Domenico M, Bottero D, Vitale F, Migliaccio A, Auricchio F 2003 Androgen-stimulated DNA synthesis and cytoskeletal changes in fibroblasts by a nontranscriptional receptor action. J Cell Biol 161:547–556
Moore MG, Wetterau LA, Francis MJ, Peehl DM, Cohen P 2003 Novel stimulatory role for insulin-like growth factor binding protein-2 in prostate cancer cells. Int J Cancer 105:14–19
Rajah R, Valentinis B, Cohen P 1997 Insulin-like growth factor (IGF)-binding protein-3 induces apoptosis and mediates the effects of transforming growth factor-1 on programmed cell death through a p53- and IGF-independent mechanism. J Biol Chem 272:12181–12188
Brigstock DR 2003 The CCN family: a new stimulus package. J Endocrinol 178:169–175
Yardy GW, Brewster SF 2005 Wnt signalling and prostate cancer. Prostate Cancer Prostatic Dis 8:119–126
Holen I, Croucher PI, Hamdy FC, Eaton CL 2002 Osteoprotegerin (OPG) is a survival factor for human prostate cancer cells. Cancer Res 62:1619–1623
Pearce JJ, Penny G, Rossant J 1999 A mouse cerberus/Dan-related gene family. Dev Biol 209:98–110
Thomas R, Anderson WA, Raman V, Reddi AH 1998 Androgen-dependent gene expression of bone morphogenetic protein 7 in mouse prostate. Prostate 37:236–245
Levedakou EN, Strohmeyer TG, Effert PJ, Liu ET 1992 Expression of the matrix Gla protein in urogenital malignancies. Int J Cancer 52:534–537
Sandhu C, Peehl DM, Slingerland J 2000 p16INK4A mediates cyclin dependent kinase 4 and 6 inhibition in senescent prostatic epithelial cells. Cancer Res 60:2616–2622
Isaacs JT, Furuya Y, Berges R 1994 The role of androgen in the regulation of programmed cell death/apoptosis in normal and malignant prostatic tissue. Semin Cancer Biol 5:391–400
Hoffman GE, Smith MS, Verbalis JG 1993 c-Fos and related immediate early gene products as markers of activity in neuroendocrine systems. Front Neuroendocrinol 14:173–213
Nagpal ML, Chen Y, Lin T 2004 Effects of overexpression of CXCL10 (cytokine-responsive gene-2) on MA-10 mouse Leydig tumor cell steroidogenesis and proliferation. J Endocrinol 183:585–594
Barila D, Plateroti M, Nobili F, Muda AO, Xie Y, Morimoto T, Perozzi G 1996 The Dri 42 gene, whose expression is up-regulated during epithelial differentiation, encodes a novel endoplasmic reticulum resident transmembrane protein. J Biol Chem 271:29928–29936
Sahu A, Lambris JD 2001 Structure and biology of complement protein C3, a connecting link between innate and acquired immunity. Immunol Rev 180:35–48
Cutolo M, Sulli A, Capellino S, Villaggio B, Montagna P, Seriolo B, Straub RH 2004 Sex hormones influence on the immune system: basic and clinical aspects in autoimmunity. Lupus 13:635–638
Culig Z, Klocker H, Bartsch G, Hobisch A 2002 Androgen receptors in prostate cancer. Endocr Relat Cancer 9:155–170
Sanford DC, DeWille JW 2005 C/EBP is a downstream mediator of IL-6 induced growth inhibition of prostate cancer cells. Prostate 63:143–154
Spiotto MT, Chung TD 2000 STAT3 mediates IL-6-induced growth inhibition in the human prostate cancer cell line LNCaP. Prostate 42:88–98
Lou W, Ni Z, Dyer K, Tweardy DJ, Gao AC 2000 Interleukin-6 induces prostate cancer cell growth accompanied by activation of stat3 signaling pathway. Prostate 42:239–242
Okamoto M, Webber MM, Quader S, Oyasu R 1998 Interleukin-6 and epidermal growth factor promote anchorage-independent growth of immortalized human prostatic epithelial cells treated with N-methyl-N-nitrosourea. Prostate 35:255–262
Mori S, Murakami-Mori K, Bonavida B 1999 Interleukin-6 induces G1 arrest through induction of p27Kip1, a cyclin-dependent kinase inhibitor, and neuron-like morphology in LNCaP prostate tumor cells. Biochem Biophys Res Commun 257:609–614
Lee SO, Lou W, Hou M, de Miguel F, Gerber L, Gao AC 2003 Interleukin-6 promotes androgen-independent growth in LNCaP human prostate cancer cells. Clin Cancer Res 9:370–376
Hobisch A, Eder IE, Putz T, Hominger W, Bartsch G, Klocker H, Culig Z 1998 Interleukin-6 regulates prostate-specific protein expression in prostate carcinoma cells by activation of the androgen receptor. Cancer Res 58:4640–4645
Zerbini LF, Wang Y, Cho JY, Libermann TA 2003 Constitutive activation of nuclear factor B p50/p65 and Fra-1 and JunD is essential for deregulated interleukin 6 expression in prostate cancer. Cancer Res 63:2206–2215
Thiel G, Cibelli G 2002 Regulation of life and death by the zinc finger transcription factor Egr-1. J Cell Physiol 193:287–292
Virolle T, Krones-Herzig A, Baron V, De Gregorio G, Adamson ED, Mercola D 2003 Egr1 promotes growth and survival of prostate cancer cells. Identification of novel Egr1 target genes. J Biol Chem 278:11802–11810(A. J. Asirvatham, M. Schmidt, B. Gao and)