Loss of Long Noncoding RNA NXTAR in Prostate Cancer Augments Androgen Receptor Expression and Enzalutamide Resistance

Abstract

Androgen receptor (AR) signaling continues to play a dominant role in all stages of prostate cancer, including castration-resistant prostate cancers (CRPC) that have developed resistance to second generation AR antagonists such as enzalutamide. In this study, we identified a long noncoding RNA (lncRNA), NXTAR (LOC105373241) that is located convergent with the AR gene and is repressed in human prostate tumors and cell lines. NXTAR bound upstream of the AR promoter and promoted EZH2 recruitment, causing significant loss of AR (and AR-V7) expression. Paradoxically, AR bound the NXTAR promoter, and inhibition of AR by the ACK1/TNK2 small molecule inhibitor (R)-9b excluded AR from the NXTAR promoter. The histone acetyltransferase GCN5 bound and deposited H3K14 acetylation marks, enhancing NXTAR expression. Application of an oligonucleotide derived from NXTAR exon 5 (NXTAR-N5) suppressed AR/AR-V7 expression and prostate cancer cell proliferation, indicating the translational relevance of the negative regulation of AR. In addition, pharmacologic restoration of NXTAR using (R)-9b abrogated enzalutamide-resistant prostate xenograft tumor growth. Overall, this study uncovers a positive feedback loop, wherein NXTAR acts as a novel prostate tumor-suppressing lncRNA by inhibiting AR/AR-V7 expression, which in turn upregulates NXTAR levels, compromising enzalutamide-resistant prostate cancer. The restoration of NXTAR could serve as a new therapeutic modality for patients who have acquired resistance to second generation AR antagonists. SIGNIFICANCE: This study identifies NXTAR as a tumor suppressive lncRNA that can epigenetically downregulate AR/AR-V7 expression and provides a therapeutic strategy to reinstate NXTAR expression for treating recurrent CRPC.

Figure 1.. NXTAR is a novel tumor suppressor lncRNA in human prostate cancer.

A, Location of NXTAR gene with respect to AR gene on chromosome X, shown in a graphical format. The exons 1-5 in NXTAR are shown in Roman numerals (I to V), to distinguish from AR exons. B-L, Total RNA isolated from paired normal human prostates and tumor samples and subjected to qRT-PCR and relative expression of NXTAR (B and F) and KLK3 (D and H) is shown. Data representing mean of relative NXTAR (C, G and J) and KLK3 (E, I and K) expression between Normal and Tumor samples is shown. L, A relative NXTAR expression in various Gleason grades of prostate tumor samples is shown. M, Total RNA isolated from various cell lines and subjected to qRT-PCR with NXTAR and 18S primers. Data are represented as mean ± SEM as in (B-L). *** p<0.001, **p < 0.01, *p ≤0.05, two-tailed Student’s t-test. NS, not significant.

Figure 2.. AR and ACK1 negatively regulates NXTAR expression in prostate cancer cell lines.

A-D, Androgen-deprived LNCaP, C4-2B, 22Rv1 and VCaP cells were transfected with scrambled (Sc) or AR siRNA. Total RNA was isolated, followed by qRT-PCR with AR, NXTAR and 18S rRNA primers. E and F, Androgen-deprived (E) LNCaP and (F) 22Rv1 cells were treated with vehicle or (R)-9b (3.5 and 5 μM) for 48 hrs. Total RNA was isolated, followed by qRT-PCR with NXTAR and 18S rRNA primers. G, qRT-PCR was performed to assess suppression of ACK1 expression in 22Rv1 cells upon transfection with two different sets of (set I and II) ACK1 siRNA. H and I, Total RNA isolated from (H) LNCaP or (I) 22Rv1 cells in which ACK1 expression was downregulated using two different sets of siRNAs at the indicated time points and subjected to qRT-PCR. Data are represented as mean ± SEM. *** p<0.001, **p < 0.01, *p ≤0.05, two-tailed Student’s t-test. NS, not significant.

Figure 3:. (R)-9b inhibits Enzalutamide-resistant CRPC xenograft tumor growth in vivo and induces NXTAR expression.

A, Enzalutamide-resistant VCaP cells were injected subcutaneously in castrated male SCID mice. Once the tumors were palpable the mice were treated with either Vehicle (Captisol; n=7) or (R)-9b at 12mg/Kg (n=7) or 20mg/Kg (n=7) subcutaneously, five times a week. B and C, Tumor weights were recorded (B) and a photograph was taken (C). D, Graph represents weights of the vehicle and (R)-9b treated mice. E and F, Prostates (E) and brains (F) of the mice were harvested, RNA prepared, followed by qRT-PCR to determine the levels of the AR mRNA (n=4 each). G-J, Tumors were harvested, RNA prepared, followed by qRT-PCR to determine the levels of the (G) AR, (H) KLK3, (I) TMPRSS2 mRNAs and (J) NXTAR (n=4 each). For G-J, (n=4 each, 3 replicates). K, Tumor lysates (n=3 tumors in each arm) were immunoblotted by AR, pAR, pACK1, ACK1 and Actin antibodies, as shown. Data (A, B and D-J) are represented as mean ± SEM. *p ≤0.05; ** p<0.01; ***p<0.001, two-tailed Student’s t-test. NS, not significant.

Figure 4:. ACK1 inhibition induces NXTAR expression by increasing acetylation of H3K14 at NXTAR promoter.

A, Graphical representation of primers (Ppr3 and Ppr4) upstream of the NXTAR transcription start site. The numbers indicate nucleotide position. B and C, 22Rv1 (B) and LNCaP (C) cells treated with either vehicle or (R)-9b alone or in combination with a GCN5 inhibitor, CPTH2 were processed for total RNA extraction and qRT-PCR was performed with primers for NXTAR and 18S rRNA. D-E, 22Rv1 cells were treated with either vehicle or (R)-9b alone or in combination with CPTH2 and ChIP was performed using H3K14ac antibody or IgG (Supplementary Fig. S9A and B), followed by qPCR using (D) PPr4 and (E) PPr3 primers. F-G, LNCaP cells treated with (R)-9b alone or in combination with CPTH2 were subjected to ChIP using GCN5 antibody or IgG (Supplementary Fig. S9C and D), followed by qPCR using (F) Ppr4 and (G) Ppr3 primers. H-I, LNCaP cells were treated with either (R)-9b alone or in combination with CPTH2 were harvested and processed for ChIP using antibody against AR or IgG (Supplementary Fig. S9E and F), followed by qPCR using (H) PPr4 and (I) PPr3 primers. J, 22Rv1 cells were treated with either vehicle or (R)-9b, subjected to ChIP using GCN5 antibody, followed by qPCR using IRF8 binding site primers. K, VCaP cells were treated with either vehicle or (R)-9b, subjected to ChIP using NCOR1 antibody, followed by qPCR using NXTAR enhancer primers. L, VCaP cells were transfected with control siRNA and three sets of NCOR1 specific siRNAs (1, 2, 3 and pool) and the cell lysate was subjected to immunoblotting. M, RNA isolated processed from these cells was subjected to qRT-PCR using NXTAR and 18S rRNA specific primers. Data are represented as mean ± SEM. *p ≤0.05; ** p<0.01; ***p<0.001, two-tailed Student’s t-test. NS, not significant.

Figure 5:. NXTAR over-expression causes an increase in repressive methylation on AR promoter.

A, Location of primers with respect to AR gene shown in a graphical format. B, LNCaP cells were infected with either pBabe control vector or NXTAR expressing constructs and grown in androgen-free medium for 48 hrs. ChIP was performed using H3K27me3 antibody or IgG (Supplementary Fig. S9G-K and L-O), followed by qPCR with using primers spanning regions upstream and downstream of AR TSS (−0.7 kb to +3.4 kb) as shown in (A). C, LNCaP cells infected with pBabe or NXTAR expressing constructs and grown in androgen free medium for 48 hrs and ChIP was performed using EZH2 and IgG antibodies (Supplementary Fig. S9P-R), followed by qPCR. D, Biotinylated oligos complementary to NXTAR (or lacZ as control) were used to pull down NXTAR from cell lysates prepared from VCaP cells retrovirally-infected to over-express NXTAR, followed by immunoblotting for EZH2. E, VCaP cells were infected with NXTAR expressing construct and treated with EZH2 inhibitor, EPZ6438 (1μM) overnight, followed by qRT-PCR. F, NXTAR (or lacZ as control) immobilized onto streptavidin beads were incubated with purified EZH2. Pull-down were washed, followed by immunoblotting with EZH2 antibodies. Data (B, C and E) are represented as mean ± SEM. *p ≤0.05; ** p<0.01; ***p<0.001, two-tailed Student’s t-test.

Figure 6:. NXTAR restoration inhibits prostate cancer proliferation.

A and B, VCaP (A) and 22Rv1 (B) cells were infected with either pBabe vector or NXTAR expressing constructs and were seeded in androgen-deprived condition. Graph represents number of viable cells determined by trypan blue exclusion assay. C and D, VCaP (C) and 22Rv1 (D) cells infected with pBabe vector or NXTAR expressing constructs and were treated with either vehicle (10% DMSO) or enzalutamide for 9 or 5 days respectively, and viable cells were counted. E, 22Rv1 cells infected with pBabe vector or NXTAR expressing constructs, treated with either vehicle (10% DMSO) or abiraterone for 5 days and the viable cells were counted. F, 1.5 million VCaP cells that were infected either with pBabe or NXTAR vectors were implanted subcutaneously per SCID mice, and allowed to grow till they reached an average volume of ~1200mm³ for pBabe mice. G-I, The tumors were excised, photographed (G) and the tumor weights (H) and volume (I) were recorded. (J) Tumor lysates (n=3 per group) were immunoblotted. Densitometry measurement is provided below the AR blot. Data are represented as mean ± SEM. *p ≤0.05; ** p<0.01; ***p<0.001, two-tailed Student’s t-test. NS, not significant.

Figure 7:. NXTAR derived oligonucleotide NXTAR-N5 suppresses prostate cancer proliferation.

A, Location of NXTAR-N5 binding region upstream of AR gene shown in a graphical format. B, BLASTN analysis showing regions of significant homology (complementarity to other strand) in the upstream regions of AR and exon 5 of NXTAR. C, The NXTAR-N5 oligonucleotide sequence is shown. *Represents a Phosphorothioate bonds modifications to avoid degradation by exonucleases. D, VCaP and 22Rv1 cells were transfected with NXTAR-N5 oligonucleotide and cell proliferation was measured using trypan blue exclusion method. E, VCaP and 22Rv1 cells were transfected with NXTAR-N5, and RNA was prepared, followed by qRT-PCR for AR and AR-V7 mRNA. F, Biotin conjugated NXTAR-N5 and Globin oligos were incubated with lysate from fixed VCaP cells and qPCR was performed for AR Up1.1 primers (see Fig. 7A). G, VCaP cell lysate was subjected to chromatin pull down using NXTAR-N5 biotin conjugated oligos, followed by immunoblotting for EZH2. Actin was used as loading control. H, NXTAR-N5 and Globin oligos were incubated with purified EZH2. Pull-down were washed, followed by immunoblotting with EZH2 antibodies. I, VCaP cells were transfected with Globin and NXTAR-N5 oligos and subjected to ChIP with H3K27me3 antibody followed by qPCR for site upstream of AR TSS (−0.7kb). Data are represented as mean ± SEM. **p<0.01. *** p<0.001.