Splicing of a novel androgen receptor exon generates a constitutively active androgen receptor that mediates prostate cancer therapy resistance

Abstract

The standard systemic treatment for prostate cancer (PCa) is androgen ablation, which causes tumor regression by inhibiting activity of the androgen receptor (AR). Invariably, PCa recurs with a fatal androgen-refractory phenotype. Importantly, the growth of androgen-refractory PCa remains dependent on the AR through various mechanisms of aberrant AR activation. Here, we studied the 22Rv1 PCa cell line, which was derived from a CWR22 xenograft that relapsed during androgen ablation. Three AR isoforms are expressed in 22Rv1 cells: a full-length version with duplicated exon 3 and two truncated versions lacking the COOH terminal domain (CTD). We found that CTD-truncated AR isoforms are encoded by mRNAs that have a novel exon 2b at their 3' end. Functionally, these AR isoforms are constitutively active and promote the expression of endogenous AR-dependent genes, as well as the proliferation of 22Rv1 cells in a ligand-independent manner. AR mRNAs containing exon 2b and their protein products are expressed in commonly studied PCa cell lines. Moreover, exon 2b-derived species are enriched in xenograft-based models of therapy-resistant PCa. Together, our data describe a simple and effective mechanism by which PCa cells can synthesize a constitutively active AR and thus circumvent androgen ablation.

Figure 1

Short AR isoforms specifically mediate ligand-independent transcriptional activation of AR target genes. A, RNA was isolated from VCaP and 22Rv1 cells grown in vehicle or 1nM mibolerone (Mib) and subjected to quantitative real-time RT-PCR using primers specific for PSA, TMPRSS2, and maspin. Expression levels are relative to vehicle treatment, which was arbitrarily set to 1. B, 22Rv1 cells were transfected with MMTV-Luc, non-targeted control (CTRL) siRNA, or siRNAs targeted to AR Exon 1 or 7 (schematic on right; DBD* denotes a three zinc-finger DBD). Cells were grown 24h in serum-free medium and treated with 1nM Mib or EtOH (vehicle control) for 24h. Luciferase activity was determined. Data represent the mean +/- S.E. from at least three independent experiments, each performed in duplicate. MMTV promoter activity without androgens and siRNAs was arbitrarily set to 1. Lysates from vehicle-treated cells were subjected to Western blot using a polyclonal antibody targeted to the AR NTD. ERK-2 levels are shown as a control. C, 22Rv1 cells were transfected with non-targeted control (CTRL) siRNA, or siRNAs targeted to AR Exon 1 (siAR-2) or 7 (siAR-1). RNA was isolated from transfected cells grown for 72h without androgens. RT-PCR was performed using primers specific for hK2, PSA, TMPRSS2, NKX3.1, SCAP, and maspin. Expression levels are relative to vehicle treatment, which was arbitrarily set to 1.

Figure 2

Short AR isoforms specifically mediate ligand-independent growth of androgen-refractory 22Rv1 cells. 22Rv1 cells were transfected with non-targeted control (CTRL) siRNA, or siRNAs targeted to AR Exon 1 (siAR-2) or 7 (siAR-1). A, Transfected cells were cultured 72h post-transfection, lysed, and subjected to Western blot using a polyclonal antibody targeted to the AR NTD. ERK-2 levels are shown as a control. B, Transfected cells were seeded and cultured for 24h in medium supplemented with 5% charcoal-stripped serum (CSS). Cells were treated with 1nM DHT or ethanol (EtOH, vehicle control). The relative number of viable cells 0h and 96h following exposure to these compounds was determined by MTS assay. Values shown are relative to siCTRL-treated cells without androgens, which was arbitrarily set to 100%. Data represent the mean +/- S.E.M. from two experiments performed in quadruplicate. C, Transfected cells were seeded and treated as described in B. The relative number of proliferating cells 24h after treatment was determined by BrdU incorporation. Values are relative to siCTRL-treated cells grown without androgens, which was arbitrarily set to 100%. Data represent the mean +/- S.E.M. from two experiments performed in quadruplicate.

Figure 3

Short AR isoforms are encoded by novel mRNAs containing a novel AR Exon 2b. A, 22Rv1 cells were transfected with non-targeted control (siCTRL) siRNA, or siRNAs targeted to AR Exon 1 (siAR-2) or 7 (siAR-1) as indicated. RNA was isolated and subjected to 2-step 3′-RACE using Exon 1-anchored forward primers. PCR products were cloned and sequenced. A schematic of this approach and resultant sequence data are shown on the right. B, Relative location of Exon 2b in the AR locus and summary of 3′-RACE and RT-PCR experiments. Details are discussed in the text. C, A second 3′RACE reaction was performed on RNA as described in A using an Exon 2b-anchored forward primer indicated in B. PCR products were cloned and sequenced, which revealed 3′termini 91-113bp downstream from the start of Exon 2b. These putative sites of polyadenylation are indicated in B. D, RT-PCR was performed on RNA used in A and C with the Exon 1-anchored forward primer used for 3′RACE and a reverse primer indicated in B. PCR products were cloned and sequenced and confirmed to consist of spliced Exons 1/2/2b or 1/2/3/2b. E, 22Rv1 cells were transfected with non-targeted control (CTRL) siRNA, or siRNAs targeted to Exon 1 (siAR-2) or 2b. Transfected cells were cultured 48h post-transfection, lysed, and subjected to Western blot using a polyclonal antibody targeted to the AR NTD.

Figure 4

AR^1/2/2b and AR^1/2/3/2b are constitutively active. A-C, 22Rv1 cells were transfected with non-targeted control (CTRL) siRNA, or siRNA targeted to Exon 2b. Transfected cells were cultured 24h post-transfection, and subjected to A, Western blot, B cell proliferation assay (*P < 0.05, **P < 0.002, compared to vehicle-treated cells), and C, quantitiative RT-PCR exactly as described in the legends to Figs. 1 and 2. D, DU-145 cells were transfected with expression vectors encoding full-length wild-type AR (AR^1-8), AR^Ex1/2/2b, or AR^Ex1/2/3/2b in conjunction with MMTV-LUC or 4XARE-E4-LUC. Cells were treated under serum-free conditions with 1nM mibolerone (Mib) or ethanol (EtOH, vehicle control) for 24h. Luciferase activity was determined. Data represent the mean +/- S.E. from at least three independent experiments, each performed in duplicate. The activity of reporters in the absence of transactivator or androgens was arbitrarily set to 1. Lysates were analyzed by Western blot using an antibody specific for the AR NTD.

Figure 5

AR^1/2/2b mRNA and protein is expressed in PCa cell lines. A, Schematic of AR mRNA species and PCR primer sets used for their detection. B, RT-PCR analysis of PCa cell lines. cDNAs were amplified by 2-step nested PCR using forward and reverse primer sets as indicated. Major bands were cloned and sequenced. The identity of sequenced products is indicated on the right. C, The absolute abundance of full-length AR mRNA and AR Exon 1/2/2b mRNA were determined using specific PCR primers in quantitative RT-PCR reactions. Threshold amplification cycle (Ct) values were converted to copy number by extrapolating against standard curves generated from serial dilutions of plasmids harboring Exon 1/2/2b or Exon 1-8 cDNAs. Data represent the mean +/- S.E.M. from four independent experiments. D, VCaP cells were transfected with non-targeted control (CTRL) siRNA, or siRNAs targeted to Exon 7 (siAR-1), Exon 1 (siAR-2), or Exon 2b. Transfected cells were cultured 48h post-transfection, lysed, and subjected to Western blot using a polyclonal antibody targeted to the AR NTD.

Figure 6

AR^1/2/2b mRNA is enriched in xenograft-based models of androgen-refractory prostate cancer. A, cDNAs were amplified by 2-step PCR using the same forward and reverse primer sets depicted in the legend for Fig. 5. Major bands were cloned and sequenced. The identity of sequenced products is indicated on the right. B, The absolute abundance of full-length AR mRNA and AR Exon 1/2/2b mRNA were determined using specific PCR primers in quantitative RT-PCR reactions. Ct values were converted to copy number by extrapolating against standard curves generated from serial dilutions of plasmids harboring Exon 1/2/2b or Exon 1-8 cDNAs. Data represent the mean +/- S.E.M. from four independent experiments (*P < 0.05, **P < 0.001). C, Lysates from androgen dependent (AD) and androgen independent (AI) versions of LuCaP 23.1 and 35 xenografts were subjected to Western blot using polyclonal antibodies recognizing the AR NTD or CTD, or a monoclonal antibody recognizing the AR NTD. The putative AR^1/2/2b isoform is indicated on the right. A schematic of the antibody recognition sites is shown at the bottom of the figure.