Splicing factor TRA2B enhances synthesis of androgen receptor variant AR-V7 in prostate cancer cells
Nicholas Brittain, Alec Paschalis, Ryan Nelson, Beth Adamson, Laura Walker, Ruaridh Duncan, Graham R. Smith, Suzanne McGill, Richard J.S. Burchmore, Denisa Bogdan, Juan M. Jiménez-Vacas, Jonathan Welti, Wei Yuan, Craig N. Robson, Pasquale Rescigno, Sara Luzzi, Adam Sharp, Johann de Bono, Luke Gaughan
Nicholas Brittain, Alec Paschalis, Ryan Nelson, Beth Adamson, Laura Walker, Ruaridh Duncan, Graham R. Smith, Suzanne McGill, Richard J.S. Burchmore, Denisa Bogdan, Juan M. Jiménez-Vacas, Jonathan Welti, Wei Yuan, Craig N. Robson, Pasquale Rescigno, Sara Luzzi, Adam Sharp, Johann de Bono, Luke Gaughan
View: Text | PDF
Research Article Clinical Research Oncology

Splicing factor TRA2B enhances synthesis of androgen receptor variant AR-V7 in prostate cancer cells

Abstract

Treatment of locally advanced and metastatic prostate cancer (PC) with androgen receptor–targeting (AR-targeting) therapies has limited durability, with disease eventually progressing to castrate-resistant PC (CRPC). Constitutively active AR splice variants (AR-Vs), such as AR-V7, play a key role in driving treatment resistance and disease progression. Importantly, the failure to attenuate AR-V function represents a major unmet clinical need, and as such, defining how AR-Vs are generated is likely to yield new therapeutic targets. Our knowledge of factors that mediate splicing of AR-V–encoding mRNAs remains limited. Here, we have employed an RNA-targeting CasRx approach to identify selective protein interactors of AR-V7 mRNA in PC. TRA2B and its ortholog, TRA2A, were identified as splicing regulators of AR transcripts that facilitate AR-V synthesis at the expense of full-length AR isoforms. TRA2B expression correlated with AR-V7 transcript in CRPC and attenuation of TRA2-mediated splicing diminished PC cell growth. Exploiting TRA2B function may therefore provide new therapeutic opportunities in advanced disease.

Authors

Nicholas Brittain, Alec Paschalis, Ryan Nelson, Beth Adamson, Laura Walker, Ruaridh Duncan, Graham R. Smith, Suzanne McGill, Richard J.S. Burchmore, Denisa Bogdan, Juan M. Jiménez-Vacas, Jonathan Welti, Wei Yuan, Craig N. Robson, Pasquale Rescigno, Sara Luzzi, Adam Sharp, Johann de Bono, Luke Gaughan

×

Figure 2

Positional proximal biotinylation at CE3 identifies numerous characterized AR-V7 splicing factors.

Options: View larger image (or click on image) Download as PowerPoint
Positional proximal biotinylation at CE3 identifies numerous characteriz...
(A) Summary of the label-free proteomics workflow used to detect splicing regulators involved in AR-V7 synthesis. CWR22Rv1-idCasRX-APEX2 were transfected with either nontargeting gRNA (gNT), AR gRNA (gAR), or no gRNA in the presence of 1 mg/μl doxycycline for 72 hours prior to biotin phenol (BP)/H2O2 treatment, streptavidin-based protein enrichment, and subsequent detection by mass spectrometry. (B) After applying the experimental setup detailed in A, a portion of extracted nuclear protein was retained for Western blot for biotinylated proteins to confirm APEX2 labeling compared with the unlabeled (-BP/H2O2). Ponceau was used to confirm equal protein loading. (C) Volcano plot of limma-analyzed protein iBAQ (derived from MaxQuant) value enrichment between gAR and gNT arms. Protein enrichment was calculated between AR g2 and NT gRNA arms using iBAQ values with the limma package. Enrichment cutoffs for visualization are limma P value (unadjusted) < 0.05 and linear fold enrichment > 1.5 (log2 0.585). (D) KEGG analysis, performed using STRING, was applied to proteins enriched by gAR at a limma FDR < 0.25. The top 5 most significant KEGG terms are displayed and ranked by –log10 STRING FDR. KEGG spliceosome is highlighted by a red asterisk. (E) KEGG spliceosome –log10 STRING FDR comparison between high- (FDR < 0.25) and low- (FDR > 0.25) confidence CE3 interactors. (F) Protein enrichment results as in C were filtered for previously published AR-V7 regulators and are shown in the filtered plot. (G) ROAST was performed using the curated list of AR-V7 splicing factors as input (Supplemental Table 4). FDR represents ROAST-calculated significance of statistical enrichment for this list. Barcode plot represents ranking of –log10 FDR values for this list among all proteins from limma analysis.