Androgen-regulated genes differentially modulated by the androgen receptor coactivator L-dopa decarboxylase in human prostate cancer cells

The identification of AR-regulated genes that are affected by DDC overexpression may provide important clues regarding the biology of this catecholamine synthesis enzyme and its influence on AR function. Toward this end, the change in gene expression of androgen-regulated genes caused by sustained DDC overexpression were analyzed in vitro by the generation of LNCaP cells stably expressing Dox-inducible DDC. We used human androgen-dependent prostate cancer cells (LNCaP), since our primary goal was to assess the co-activation function of DDC, known to enhance AR activity through an androgen-dependent mechanism [18]. To verify the effects of Dox treatment on expression of the DDC gene, LNCaP cells, stably expressing Dox-inducible DDC (LNCaP-DDC) or the vector control (LNCaP-pDEST) were treated for 48 hours under mock-induced (-Dox) and Dox-induced (+Dox) conditions. With Dox treatment, we observed an optimal 6-fold increase in DDC protein levels when compared with mock-induced (-Dox) LNCaP-DDC cells. No detectable DDC protein was found under the Dox-induced (+Dox) or mock-induced (-Dox) condition in the LNCaP-pDEST cell line, confirming the fact that endogenous DDC protein expression levels in this cell line are below detection when directly compared to ectopic overexpression (Figure 1A) [18]. To investigate the extent of DDC overexpression at the RNA level, we compared by RT-PCR the mRNA levels in DDC overexpressing cells, with the vector control cells after Dox stimulation. In the LNCaP-DDC cells, higher levels of DDC mRNA were observed than in the vector control LNCaP-pDEST cells, where only a faint band was detected (Figure 1B).

The availability of the LNCaP-DDC cell line afforded us an opportunity to study the influence of DDC overexpression on global gene expression using the Human Operon 21 K oligonucleotide array. Our goal in this study was to identify which genes among the AR-regulated genes, are affected by DDC overexpression. We prepared total RNA from DDC overexpressing cells (LNCaP-DDC) and from the LNCaP control cell lines (LNCaP-DDC-Dox, LNCaP-pDEST- Dox, and LNCaP-pDEST+Dox) after 48 h treatment with or without Dox, 24 h of which was in the presence or absence of R1881 synthetic hormone (Figure 2). Independent probe synthesis from different batches of RNA and hybridizations were performed in duplicate. To account for dye bias a dye swap was also performed. To identify the androgen-regulated genes in both LNCaP-DDC and LNCaP control cells, total RNA isolated from hormone treated (+R1881) samples was combined with total RNA isolated from hormone-untreated (-R1881) samples and the relative abundance of each gene was calculated as a ratio between hormone-treated and hormone-untreated samples. The data set was normalized, and a filter was applied to select only those genes whose expression level was significant (p ≤ 0.05) in at least 1 out of the 2 conditions. A total of 3,127 genes were identified, and presented as a scatter correlation plot in Figure 3.

Among the androgen-regulated genes we identified, using Venn diagram analysis, 130 genes that were androgen-regulated in LNCaP-DDC and LNCaP control cells (Figure 4A). The hormone induction response of these 130 genes was substantiated by the altered expression of the classically androgen-regulated genes, such as PSA, FKBP5, NKX3A, TMEPAI, KLK2, ODC1, and TMPRSS2 (data not shown). Comparison of LNCaP-DDC and LNCaP control cells revealed a number of changes in the expression profile of these androgen-regulated genes. Out of the 130 genes, 35 were differentially regulated as shown in Table 1. Of these, 25 genes were up-regulated at least two fold and 10 were down-regulated by at least two fold with DDC overexpression. In particular, among the set of 35 genes, we could identify four different responses to hormone treatment and DDC overexpression: i) 2 genes were hormone and DDC up-regulated ii) 4 genes were hormone and DDC down-regulated iii) 23 genes were hormone down-regulated in control cells and hormone up-regulated in DDC overexpressing cells iv) 6 genes were hormone up-regulated in control cells and hormone down-regulated in DDC overexpressing cells (Figure 4B).

Gene ontology classification of these 35 differentially expressed genes was performed by using GeneBank accession number, Operon ID, and annotation tools database available on-line [28, 29]. The gene sets were classified according to their putative main molecular function. While the list of 35 genes contained 15 unclassified transcripts, of the remaining 20, ontological molecular function analyses revealed signal transducer, binding, and catalytic activity as the predominant divergences between DDC overexpressing cells and controls (Table 1). Gene ontology classifications with overlapping gene lists were combined. The signal transducer classification defines those transcripts that mediate the transfer of a signal from the outside to the inside of a cell by a means other than the introduction of a signal molecule itself into the cell. The binding activity category defines those transcripts encoding for proteins capable of selective, often stoichiometric, interaction with one or more specific sites on another protein. The catalytic activity group refers to those proteins able to perform the catalysis of a biochemical reaction. Further annotation analysis of the transcripts involved in signal transduction, subclassified four genes into the G-protein-coupled receptors (GPCRs) signaling pathway (DRD3, PTGDR, GPR15, and GABRE). Interestingly, we found a 4.3-fold down-regulation of the DRD3 gene in DDC overexpressing cells when compared with controls. The DRD3 gene encodes the D3 subtype dopamine receptor which binds dopamine, one of the enzymatic products of DDC [30]. The D3 receptor, along with the other GPCRs, are known to modulate a plethora of signal transduction pathways that can cross-talk with known AR activation pathways [4]. Among the genes up-regulated by DDC overexpression, there was a well established androgen regulated TMEPAI gene [31, 32] (Table 1). Another gene up-regulated by DDC overexpression was TGM2, which belongs to the transglutaminase (hTGP) enzyme family (EC 2.3.2.13) [33]. Interestingly, the expression of one of these family members, prostate transglutaminase enzyme (TGM4), has been shown to be androgen-regulated [34]. SYTL2 was among the androgen down-regulated genes, and has been previously shown to be differentially regulated by AR reduction in prostate cancer cells [35]. Overall, our results indicated that 35 genes were differentially expressed after induced overexpression of DDC in LNCaP cells, including genes known to be up- or down-regulated by AR.

To verify the results of our microarray analyses, we selected seven representative genes from the list in Table 1 and determined their expression profiles by quantitative real-time RT-PCR. TMEPAI, SYTL2, TGM2, DRD3, and PTGDR were chosen because of the above consideration, whereas UNC5C and YME1L1 genes were selected based on the large magnitude of change (Table 1) and GAPDH was used as an endogenous control. The same RNA samples used for the microarray hybridization were examined. For simplicity we did not consider the mock-induced (-Dox) samples, but compared the RNA extracted from LNCaP-DDC Dox-induced cells (+Dox) versus the RNA extracted from the empty vector LNCaP-pDEST Dox-induced cells (+Dox). Although the magnitudes of up-or down-regulation of the TMEPAI, SYTL2, TGM2, DRD3, YME1L1 and PTGDR genes in the array analysis were different than that observed by real time RT-PCR, there was a similar trend in the changes with respect to both R1881 treatment and DDC overexpression (Table 2). A constant level of GAPDH was observed. Unfortunately no voluble template was obtained for the UNC5C gene. Overall, the above results verify the data observed in our microarray studies.

LNCaP cells stably expressing tetracycline-inducible DDC (LNCaP-DDC) were generated using ViraPower T-REx Lentiviral Expression System and Gateway Technology vectors, according to the manufacturer's protocol (Invitrogen) (Wafa LA, 2007 manuscript in preparation). Briefly, 3 μg of each lentiviral vector, pLenti4/TO/V5-DEST carrying the DDC gene or the empty vector pLenti4/TO/V5-DEST and the pLenti6/TR containing the TetR gene, together with 9 μg of the ViraPower packaging mix, were transfected into 293T cells, using Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA). The DNA- Lipofectamine 2000 complexes diluted in Opti-MEM I Medium (Gibco-BRL) were allowed to form for 20 min at room temperature before addition to 293T cells. Cells were maintained for 24 hours at 37°C and 5% CO₂ before removing the media containing the DNA-Lipofectamine 2000 complexes and replacing with DMEM media (10% FBS, 2 mM L-glutamine, 0.1 mM MEM Non-Essential Amino Acids, 1% penicillin/streptomycin, and 1 mM MEM Sodium Pyruvate).

LNCaP-DDC, and LNCaP-pDEST cell lines were plated in 10-cm plates (2 × 10⁶ per plate) in RPMI 1640 containing 5% charcoal-stripped serum (CSS; HyClone, VWR, West Chester, PA). When the cells reached 60% confluence, they were seeded in presence of 1 μg/ml of Dox (Dox-induced) or in the absence of Dox (mock-induced). After 24 hours the cells were stimulated with or without 0.1 nM of synthetic androgen (± R1881) for an additional 24 hours. Total RNA was extracted using Trizol reagent by following the manufacturer's instructions (Invitrogen). Total RNA from each hormone-treated sample (+R1881) was compared with that of hormone-untreated sample (-R1881) on the same microarray slide. Two independent biological replicates were assayed for each sample and a dye swap was performed to account for dye bias. Microarrays of 21,521 (70-mer) human oligos representing 21,521 genes (Operon Technologies, Inc., Alameda, CA) printed on aminosilane coated microarray slides (Matrix Technologies; Hudson, NH) were supplied by the Array Facility of The Prostate Centre at Vancouver General Hospital. Microarrays were hybridized with 10 μg of total RNA from duplicate samples of LNCaP-DDC or LNCaP-pDEST cells treated with or without (±) Dox, and with or without (±) R1881, using the 3DNA Array 350™ Expression Array Detection Kit and according to the manufacturer's instructions (Genisphere, Hatfield, PA) (Figure 2). The arrays were immediately scanned on a Scan Array Express Microarray Scanner (Perkin Elmer). Signal quality and quantity were assessed using ImaGene 7.0 software (BioDiscovery, San Diego, CA).

The DDC antibody was purchased from Chemicon (Chemicon Inc., Temecula, CA) and the polyclonal antibody to actin was obtained from Sigma (Sigma Chemical Co., St. Louis, MO). LNCaP-DDC and LNCaP-pDEST cells were plated in six-well plates (3 × 10⁵ per well), treated with Dox as described above and immunoblotted as previously reported [18]. Total cell protein (50 μg), measured by BCA™ Protein Assay (PIRCE, Rockford, IL), was used for immunoblotting. Band intensity was quantified using a Bio-Rad Gel Doc 2000 software (Bio-Rad Laboratories, USA).

Raw signal data files obtained with ImaGene 7.0 software were subsequently analyzed on GeneSpring 7.2 software (Silicon Genetics, Redwood City, CA) for profiling significant changes in gene expression. The fluorescent signal ratios (+R1881/-R1881) were subjected to Lowess normalization with background correction. Experimental error was based on replicated dye pair values. Comparison analysis of the gene expression data from LNCaP-DDC+Dox cells, and LNCaP-DDC-Dox, LNCaP-pDEST-Dox, and LNCaP-pDEST+Dox cells, treated with or without R1881, was conducted to first identify androgen-regulated genes in all experimental conditions, and then the differentially regulated genes due to DDC overexpression.

Only genes with a p value of ≤ 0.05 in at least one out of two conditions were analyzed further. Data was transformed to log ratio (Log₁₀) for display and analysis. All genes showing a normalized expression value ≥ 2 or ≤ 0.5 were classified as androgen up- or down-regulated, respectively. Genes showing a normalized expression value between 0.5- and 2- were classified as unchanged and not considered further. Lists of androgen-regulated genes (both up-regulated and down-regulated) were created for each of the cell lines and were compared by Venn diagram analysis. Genes were considered differentially expressed as a result of DDC overexpression if normalized values from induced LNCaP-DDC cells were at least 2-fold greater or 2-fold less than those from the control cells (LNCaP-DDC-Dox, LNCaP-pDEST- Dox, and LNCaP-pDEST+Dox). Functional classifications were based on gene ontology (GO) annotation obtained through the GeneTools database [28, 29].

Complementary DNA (cDNA) for real time PCR and semiquantitative RT-PCR was made using 2 μg of total RNA treated with RNase-free DNase according to the manufacturer's instructions (Promega). First-strand cDNA was synthesized using random hexamers (Perkin-Elmer Applied Biosystems, Branchburg, NJ) with 20U of Moloneymurine leukemia virus reverse transcriptase, M-MLV (Invitrogen). The ABI PRISM 7700 Sequence Detection System (Perkin-Elmer Applied Biosystems, Foster City, CA) was used for real time monitoring of PCR amplification of cDNA. The forward and reverse primers used for the real time RT-PCR are listed in Table 3. All primers were selected by PRIMER EXPRESS v.3 software and were commercially synthesized by Integrated DNA Technologies laboratories (IDT, Coralville IA). A SYBR green PCR kit was used following the manufacturer's instructions and the analyses were performed in triplicate (Invitrogen). In brief, RT-PCR amplification mixtures (25 μl) containing 25 ng template cDNA, 2x SYBR Green I Master Mix buffer (12.5 μl), and 300 nM forward and reverse primer was prepared. Target mRNA values were normalized using GAPDH mRNA as an internal control. A comparative threshold cycle (Ct) was used to determine gene expression relative to a calibrator. For each sample, the Ct values were calculated using the formula ΔC_T = C_{t sample}- C_{t GAPDH}. To determine relative expression levels, the following formula was used ΔΔC_T = ΔC_{T sample}- ΔC_{T calibrator} and the value used to indicate relative gene expression was calculated using the formula 2^-ΔΔCT. Primers for the semiquantitative PCR were synthesized by Integrated DNA Technologies laboratories (IDT, Coralville IA); DDC, sense 5'- ACA CCA TGA ACG CAA GTG AA-3' and antisense 5'- CAC CCC AGG CAT GAT TAT CT-3'. The PCR products (209 bp) were resolved by electrophoresis using 1.5% agarose gels containing ethidium bromide.