Analysis of Androgen-dependent and -independent Regulation of Transcriptional Activity

Monitor activity using a sensitive and optimized magnetic ChIP kit

bySuzanne M. Smith, M.S.; Kay K. Opperman, Ph.D.; Barbara J. Kaboord, Ph.D. - 09/11/13

Many factors, including androgens, growth factors, and transcription factors, participate in the regulation of normal and cancerous prostatic development and function. The androgen receptor (AR) is a key regulator of normal prostate function and survival; however, the AR also can be converted into an inducer of uncontrolled cell growth in prostate carcinoma. This steroid receptor is sequestered in the cytoplasm by heat-shock proteins and co-chaperones in the absence of its ligand, dihydroxytestosterone (DHT). Upon androgen binding, the chaperones dissociate, AR dimerizes, and a nuclear localization sequence is revealed. This allows translocation to the nucleus where AR can bind to androgen response elements (AREs), leading to an alteration in gene expression (Figure 1).[Ref.1] The progression of signaling through DHT binding to AR is referred to as androgen-dependent.

Figure 1. Pathways involved in androgen receptor activation. Testosterone is converted to DHT, which binds to the androgen receptor (AR). AR then dimerizes, is phosphorylated and translocates to the nucleus where it can bind to different androgen response elements (AREs) to ultimately lead to increased cell survival. In LNcaP cells, AR is mutated to allow weak β-estradiol activation, which can result in some of the same responses as testosterone. EGF binds to EGFR leading to a cascade of protein activations including JAK/STAT3, PI3K/Akt, and MEK/MAPK pathways. It is possible that some of these androgen-independent responses, in turn, allow AR to be phosphorylated.

Treatment for prostate carcinomas often involves testosterone deprivation therapy to stop tumor growth. However, this effect was shown to be only temporary, leading to investigations of androgen-independent pathways for prostate cancer progression.[Ref.2] This is a highly complex pathway(s) that is not yet fully understood, potentially involving adaptation to non-androgens, prostate cell clonal selection, and AR mutations. Current evidence suggests that cross-talk between the AR/DHT axis and growth factor signaling effectors contributes to prostate tumor progression. Both epidermal growth factor (EGF) and its membrane receptor (EGFR) are frequently up-regulated in advanced stages of prostate cancer.[Ref.3] Numerous post-translational modifications have been described on the androgen receptor, including phosphorylation, acetylation, sumoylation, methylation, and ubiquitination.[Ref.1] It is possible that many of the EGFR signaling effectors are responsible for some of these modifications.

Here we examined the effects of testosterone, β-estradiol, and EGF stimulation on the androgen-sensitive prostate cancer cell line LNcaP. We used Western blotting to show activation/repression of proliferation pathways along with AR protein levels. We investigated via chromatin immunoprecipitation (ChIP) assay whether or not non-androgen stimuli (β-estradiol and EGF) could lead to AR binding to the known androgen receptor elements (AREs) of several genes: prostate specific antigen (PSA), FK506 binding protein 5 (FKBP5), insulin-like growth factor 1 (IGF-1), cyclin-dependent kinase inhibitor 1A (CDKN1A), and transmembrane protease, serine 2 (TMPRSS2). Our results indicate that proliferation pathways such as MAPK and AKT are activated upon EGF stimulation and repressed in cells treated with testosterone or β-estradiol. We also show that testosterone treatment leads to increased binding of the AR to AREs.

To assess the effects of androgen and non-androgen stimuli on LNcaP cells, we cultured and treated the cells with testosterone, β-estradiol, or EGF (Figure 2) and measured the changes in expression of various proteins and genes. To determine pathway activity, we lysed the treated cells to extract total protein and probed for the phospho-specific targets pAKT, pMAPK, pCREB, and pSTAT3 and AR protein levels by western blotting. In addition, we monitored AR mRNA levels after testosterone treatment by RT-PCR. Finally, we performed ChIP using anti-RNA polymerase II and anti-AR antibodies to determine the effect of the treatments on protein binding events at known AREs.

Figure 2. Experimental workflow to look at androgen and non-androgen effectors on AR protein expression and gene regulation. LNcaP cells were grown and treated with testosterone, β-estradiol, or EGF. Total protein and RNA were extracted to evaluate changes in protein expression at the protein or transcriptional level. Cells were also crosslinked to perform ChIP and ultimately qPCR to determine effects on gene regulation at the transcriptional level. Refer to Methods section for complete details.

Our results indicate that testosterone treatment increases pAKT activity over time, while MAPK and CREB phosphorylation levels are repressed (Figure 3A). Research has shown that activation of AKT as well as inactivation of MAPK can protect LNCaP cells from apoptosis, therefore leading to increased cell survival.[Ref.4] AR protein levels increased after two hours of testosterone treatment (Figure 3A). This result was supported by observing a five-fold increase in AR mRNA levels shortly after testosterone treatment, indicating that protein expression was controlled at the transcriptional level (Figure 3B). The increase in transcriptional activity was also verified by performing ChIP with testosterone treated cells using anti-RNA polymerase II antibody and amplifying the PSA promoter region (Figure 3C). Here, testosterone treatment led to almost a three-fold increase of RNA polymerase recruitment to the PSA promoter region. This is a fast response and returns back to normal levels after two hours of testosterone treatment. This could help explain why PSA levels are elevated in prostate cancer patients.

Figure 3. Upregulation of AR by testosterone increases transcriptional activity of AR-dependent target genes. (A) Western analysis of total cell lysate with targets and testosterone treatment indicated. (B) RT-PCR analysis of AR mRNA levels represented as fold enrichment over the control cells. (C) Transcriptional activity of the PSA promoter is indicated as fold enrichment over normal IgG. (D) AR targets are indicated and transcriptional activity is represented as fold enrichment over normal IgG.

We also performed ChIP with anti-AR antibody to look at the binding of this transcriptional regulator to different known AREs (Figure 3D). Our results show that testosterone increased AR binding to the PSA gene by 54-fold, to the CDKN1A promoter by 26-fold, to the TMPRSS2 ARE by 58-fold, and to the FKBP5 ARE by 300-fold (Figure 3D and Table 1). In most cases, this was a rapid response that occurred in the first thirty minutes of testosterone treatment and then returned to lower levels by two hours of treatment. We conjecture that this significant increase of AR binding to the FKBP5 ARE in testosterone-treated cells might cause altered FBP51 protein levels, which in turn could affect AR nuclear translocation, as FKBP1 is known be an AR co-chaperone.[Ref.5] Testosterone might increase AR binding to the TMPRSS2 promoter region because this promoter has been reported to fuse with the coding region of erythroblast transformation-specific (ETS) family of transcription factors, which appears to lead to a more aggressive prostate cancer.[Ref.6]

Table 1. Relative transcriptional regulation of AR-dependent target genes. Value for each stimulant and target time-point is the difference in fold-enrichment over the control (zero time-point).

Because LNCaP cells have a mutated AR that allows weak β-estradiol activation, we also examined the effects of β-estradiol on pathway activation and transcriptional activity (Figure 4). MAPK and CREB were inactivated and AR protein levels increased upon β-estradiol treatment, similar to what was observed with testosterone treatment; however, unlike testosterone, β-estradiol was unable to activate AKT (Figure 4A). Using ChIP, we also measured the effects of β-estradiol on AR binding to known AREs (Figure 4B). Treatment with β-estradiol only slightly increased AR binding to the PSA and CDKN1A promoters and the FKBP5 ARE. β-estradiol increased AR binding to the TMPRSS2 ARE by 75-fold and to the IGF-1 promoter by 15-fold. Nevertheless, by comparison to testosterone, β-estradiol was only a weak effector of these genes.

Figure 4. β-estradiol is unable to strongly activate transcription of AR-dependent targets. (A) Western analysis of total cell lysate with targets and β-estradiol treatment indicated. (B) Transcriptional activity of AR-dependent targets is indicated as fold enrichment over normal IgG.

Lastly, we examined the effects of the non-androgen stimulant EGF on pathway activation and transcriptional activity (Figure 5). As expected, EGF treatment quickly activated JAK/STAT3, PI3K/AKT, and MAPK pathways, while AR protein levels remained unchanged (Figure 5A). EGF treatments led to decreased AR binding to the PSA promoter by 12-fold. CDKN1A and IGF-1 promoters as well as FKBP5 ARE were mildly affected, signifying a role of non-androgen stimuli on the regulation of ARE-containing genes (Figure 5B).

In this study, steroids did not activate typical proliferation pathways, with the exception of AKT in testosterone treatment. Testosterone-treated cells exhibited an increase in AR at the mRNA transcript and protein level. AR protein levels also slightly increased with β-estradiol treatment. As expected, EGF quickly activated JAK/STAT3, PI3K/AKT, and MAPK pathways. Only testosterone treatment strongly activated AR binding to the PSA promoter (5 to 10 times more than β-estradiol and EGF; Table 1). Testosterone was also able to recruit RNA polymerase II to the PSA promoter to activate transcription. Binding of AR to the FKBP5 ARE in testosterone treated cells was 300-fold enriched over control cells. It would be interesting to evaluate FKBP5 by Western blot to determine the effect of testosterone on protein levels. Binding of AR to the TMPRSS2 ARE increased by 60- and 75-fold over the control when treated with testosterone or β-estradiol, respectively. The reason that β-estradiol can cause AR binding to the TMPRSS2 promoter so strongly is unknown; more studies need to be completed.

Prostate cancer progression relies on the interplay between androgen-dependent and -independent processes. Depending on the cell type and stimulus, the independent pathway may result in proliferation, progression, or inhibition. Here we examined the regulation of cellular signaling events, transcriptional activity, and mRNA levels by in LNCaP cells treated with testosterone, β-estradiol, and EGF. A sensitive and optimized magnetic ChIP kit allowed us to investigate transcriptional activity differences among the treatments.

Cell culture and treatments

LNCaP (ATCC, CRL-1740) cells were cultured at 37°C with 5% CO₂ in RPMI-1640 with 10% FBS until 80% confluent. Cells were then cultured in phenol red-free RPMI-1640 containing 10% charcoal-stripped FBS for 48 hours. Cells were treated with testosterone (10nM, 30 minutes and 2 hours; Sigma), β-estradiol (10nM, 4 hours; Sigma), EGF (100ng/mL, 15 minutes and 4 hours; Cell Signaling Technology), or left as no treatment control cells.

Western blot analysis

Cells were treated as described above, then rinsed twice with cold TBS. The washed cells were lysed on the plate using Pierce IP Lysis Buffer (Part No. 87788) containing 1X Halt Protease and Phosphatase Inhibitor Cocktail, EDTA-free (Part No. 78441). Total protein was quantified using Pierce 660nm Protein Assay (Part No. 22660). For Western analysis, 30µg of total protein from each sample was electrophoresed on 4-20% Tris-HCl SDS-polyacrylamide gels and transferred to PVDF membrane. Membranes were blocked with StartingBlock T20 (TBS) Blocking Buffer (Part No.37543) and incubated with primary antibody overnight at 4°C. Membranes were incubated with the appropriate horseradish peroxidase (HRP)-conjugated secondary antibody (Part No. 31430 and 31460). Bands were detected using SuperSignal West Pico Chemiluminescent Substrate (Part No. 34080). Primary antibodies used: β-actin (Ab No. MA1-744); pAKT, PKA, pMAPK, pCREB, and pAKT (Cell Signaling Technology); and PI3K ( BD Biosciences).

Total RNA extraction and reverse transcription

Cells were treated with testosterone as described above, then rinsed twice with cold TBS. Total RNA was extracted using TRIzol™ Reagent (Life Technologies), and 1µg was reverse transcribed using SuperScript™ III Reverse Transcriptase (Life Technologies). For qPCR, 1µL of the final cDNA was used.

Chromatin immunoprecipitation

Cells were treated as described above. One plate of each treatment was counted to determine total cell number. The remaining plates were cross-linked on the plate using a final concentration of 1% formaldehyde for 10 minutes at room temperature and quenched with 250mM glycine (final concentration) for 5 minutes at room temperature. ChIP reactions, 4×10^6 cells each, were performed in duplicate using Pierce Magnetic ChIP Kit (Part No. 26157) and the following antibodies: AR (4.5µg, Millipore), RNA polymerase II (10µL, from Part No. 26157), and normal rabbit IgG (1.5µg, from Part No. 26157). For qPCR, 1µL of the final DNA was used.

Quantitative PCR

Quantitative PCR was performed in triplicate using Thermo Scientific™ Luminaris™ Color HiGreen qPCR Master Mix on the Thermo Scientific™ Piko-Real™ Real-Time qPCR System. Data is represented as fold enrichment, i.e., signal (specific antibody) over background (normal IgG antibody). Analysis was done via the [delta][delta] Cq method, first normalizing to input, then using the equation: Net Cq = [Cq specific antibody] – [Cq normal IgG], where Cq is defined as quantification cycle, the cycle number where fluorescence increases above the background. To determine fold enrichment, the following equation was used: fold enrichment = 2^-Net Cq.

qPCR primer sequences

AR promoter:

• Forward 5’ AAGAGCCGCTGAAGGGAAACAG 3’
• Reverse 5’ AGCATCCTGGAGTTGACATTGG 3’

PSA promoter:

• Forward 5’ TCTGCCTTTCTCCCCTAGAT 3’
• Reverse 5’ AACCTTCATTCCCCAGGACT 3’

FKBP5 ARE:

• Forward 5’ GCATGGTTTAGGGGTTCTTGC 3’
• Reverse 5’ AACACCCTGTTCTGAATGTGGC 3’

TMPRSS2 ARE:

• Forward 5’ TGGTCCTGGATGATAAAAAAGTTT 3’
• Reverse 5’ GACATACGCCCCCACAACAGA 3’

IGF-1 promoter:

• Forward 5’GGGCACATAGTAGAGCTCACAAAATG 3’
• Reverse 5’ TGAGTCTTCTGTGTGGTTAATACATTG 3’

CDKN1A promoter primers were purchased from SA Biosciences.

1. Wang, X, et al. (2007). Targeted treatment of prostate cancer. J Cell Biochem 102: 571-579.
2. Coffey, K and Robson, CN. (2012). Regulation of the androgen receptor by post-translational modifications. J Endocrin 215:221-237
3. Zhu, ML and Kyprianou, N. (2008). Androgen receptor and growth factor signaling cross-talk in prostate cancer cells. Endocri Relat Cancer 15: 841-849.
4. Lin, HK, et al. (2003). Suppression versus induction of androgen receptor functions by the phosphatidylinositol 3-kinase/akt pathway in prostate cancer LNCaP cells with different passage numbers. J Biol Chem 278: 50902-50907.
5. Tai, PK, et al. (1986). A 59-kilodalton protein associated with progestin, estrogen, androgen, and glucocorticoid receptors. Biochem 25: 5269-5275.
6. Li, Y, et al. (2011). Inactivation of AR/TMPRSS2-ERG/Wnt signaling networks attenuates the aggressive behavior of prostate cancer cells. Cancer Prev Res 4:1495-1506.

This research was first presented as a poster:

Smith, S.M, et al. (April 10, 2013). Androgen-independent regulation of cellular signaling, transcriptional activity, and mRNA stability in a cell model of prostate cancer. American Association for Cancer Research Annual Meeting, Washington, D.C. (Abstract No. 5454).