Reactivation of androgen receptor-regulated TMPRSS2:ERG gene expression in castration-resistant prostate cancer

Abstract

It seems clear that androgen receptor (AR)-regulated expression of the TMPRSS2:ERG fusion gene plays an early role in prostate cancer (PC) development or progression, but the extent to which TMPRSS2:ERG is down-regulated in response to androgen deprivation therapy (ADT) and whether AR reactivates TMPRSS2:ERG expression in castration-resistant PC (CRPC) have not been determined. We show that ERG message levels in TMPRSS2:ERG fusion-positive CRPC are comparable with the levels in fusion gene-positive primary PC, consistent with the conclusion that the TMPRSS2:ERG expression is reactivated by AR in CRPC. To further assess whether TMPRSS2:ERG expression is initially down-regulated in response to ADT, we examined VCaP cells, which express the TMPRSS2:ERG fusion gene, and xenografts. ERG message and protein rapidly declined in response to removal of androgen in vitro and castration in vivo. Moreover, as observed in the clinical samples, ERG expression was fully restored in the VCaP xenografts that relapsed after castration, coincident with AR reactivation. AR reactivation in the relapsed xenografts was also associated with marked increases in mRNA encoding AR and androgen synthetic enzymes. These results show that expression of TMPRSS2:ERG, similarly to other AR-regulated genes, is restored in CRPC and may contribute to tumor progression.

Figure 1

ERG expression in TMPRSS2:ERG positive primary PC and CRPC. A and B, expression in fusion negative versus fusion positive primary PC (androgen dependent, AD) and CRPC (androgen independent, AI). C, ERG and TMPRSS2 expression by RT-PCR in these 29 CRPC tumors (AI), versus another group of 10 untreated primary PC (AD) and 6 normal prostates (N). Brackets indicate 95% confidence intervals for fusion positive tumors. P-values for differences between fusion negative and positive AD or AI tumors are shown (*p<.05, **p<.01, ***p<.001).

Figure 2

Androgen regulated TMPRSS2:ERG expression in VCaP. A, cells in CSS medium were treated with DHT and immunoblotted. B, RT-PCR for ERG (exon 9/10), TMPRSS2 (exon 5/6), and PSA mRNA after DHT stimulation. C, cells in CSS medium treated with DHT and bicalutamide for 24 hours. D, cells in 5%CSS for 3 days were transfected with 10 nM control or ERG siRNA (Dharmacon). DHT was added 8 hours after transfection and cells were assayed for ERG protein and cell recovery (by MTT assay) after 3 days (with comparable results at 5 days).

Figure 3

TMPRSS2:ERG expression in VCaP xenografts. A, average normalized xenograft size (+/− SD) at 1–6 weeks after castration (N=6). B, ERG, TMPRSS2, PSA, AR and MMP-3 mRNA in xenografts from 4 mice before castration (AD, blank bar), 4 days post-castration (C, gray bar), or at relapse (AI, black bar). C, AR protein levels in xenografts before castration (D), 4 days post-castration (C), and at relapse (I). D, IHC for ERG and AR in representative xenograft.

Figure 4

Androgen synthetic enzymes in VCaP xenografts. A, enzymes in androgen metabolism. B, relative expression of indicated enzymes by RT-PCR in VCaP xenografts (before castration, blank bar; 4 days post-castration, gray bar; relapse, black bar). C, VCaP cells in CSS medium were treated with 0, 0.1, 1 or 10 nM DHT for 24 hours and assessed by RT-PCR for the indicated transcripts. D, VCaP cells grown in CSS medium were treated with (left panel) 0, 0.1, 1, or 10 nM DHT minus or plus bicalutamide (10 µM) for 24 hours, (middle panel) actinomycin D (10 µM) minus or plus DHT (10 nM) for 0–24 hours, or (right panel) cycloheximide (10 ng/ml) with minus or plus DHT (10 nM) for 0–8 hours. Actinomycin D and cycloheximide were added 1 hour before DHT. Results were normalized to 18S RNA.