Androgen receptor splice variant-7 (AR-V7) is a constitutively active AR variant implicated in castration-resistant prostate cancers. Here, we show that the RNA splicing factor SF3B2, identified by in silico and CRISPR/Cas9 analyses, is a critical determinant of AR-V7 expression and is correlated with aggressive cancer phenotypes. Transcriptome and PAR-CLIP analyses revealed that SF3B2 controls the splicing of target genes, including AR, to drive aggressive phenotypes. SF3B2-mediated aggressive phenotypes in vivo were reversed by AR-V7 knockout. Pladienolide B, an inhibitor of a splicing modulator of the SF3b complex, suppressed the growth of tumors addicted to high SF3B2 expression. These findings support the idea that alteration of the splicing pattern by high SF3B2 expression is one mechanism underlying prostate cancer progression and therapeutic resistance. This study also provides evidence supporting SF3B2 as a candidate therapeutic target for treating patients with cancer. Significance: RNA splicing factor SF3B2 is essential for the generation of an androgen receptor (AR) variant that renders prostate cancer cells resistant to AR-targeting therapy.

Background At the beginning of tumorigenesis, newly born cancer cells must successfully avoid attack by the immune system. Although most abnormal cells are efficiently identified and destroyed by the immune system, particularly by NK cells, the molecular mechanisms by which newly born cancer cells evade NK cell surveillance are not fully understood. Methods NK cell resistance of highly tumorigenic population of human prostate cancer (PCa) cells were confirmed by xenograft in SCID mice with or without NK cell neutralization. The mechanisms by which the tumorigenic PCa cells evaded NK cell attack were investigated by RNAseq, ChIPseq, generation of several transformants and xenograft in SCID mice. Results Here, we show that PCa cells have a strengthened ability to escape NK cell attack due to NANOG, a pluripotent-related transcription factor, mediating the repression of ICAM1, a cell adhesion molecule, during tumorigenesis. Mechanistically, NANOG directly binds to the region upstream of ICAM1. As the binding between NANOG and the upstream ICAM1 region increases, p300 binding to this region is diminished, resulting in decreased ICAM1 expression. High NANOG expression confers PCa cells the ability to resist NK cell attack via the repression of ICAM1. Consistent with these results, low ICAM1 expression is significantly correlated with a high recurrence rate in patients with PCa. Conclusions Our findings indicate that repression of ICAM1 is a critical mechanism by which cancer cells evade attack from NK cells during tumorigenesis. These results suggest a pivotal role of NANOG in establishing a gene expression profile for escaping the immune system.

We designed and synthesized a novel 1,2-deoxy-pyranose and terminal epoxide methyl substituted derivatives of spliceostatin A using Julia–Kocienski olefination as a key step. With respect to the biological activity, the 1,2-deoxy-pyranose analogue of spliceostatin A suppressed AR-V7 expression at the nano level (IC50 = 3.3 nM). In addition, the in vivo toxicity test showed that the 1,2-deoxy-pyranose analogue was able to avoid severe toxicity compared to spliceostatin A.

No abstract

CRISPR-Cas9 system can be used to generate knock-out cancer cell lines. An insertion or deletion induced by a single guide RNA (gRNA) is often used to generate knock-out cells, however, some cells express the target gene by skipping the disrupted exon, or by using a splicing variant, thus losing the target exon. To overcome this unexpected expression of the target gene, almost the entire gene can be swapped with a selection marker. However, it is time-consuming to create a targeting vector which contains 5′ and 3′ homology arms flanked by a selection marker. Here, we developed a simple and easy method called SUCCESS (Single-strand oligodeoxynucleotides, Universal Cassette, and CRISPR/Cas9 produce Easy Simple knock-out System), to knock-out a target gene without constructing a targeting vector. Our method removed the targeted large genomic region by using two pX330 plasmids encoding Cas9 and gRNA, two 80mer single strand oligodeoxynucleotides (ssODN), and a blunt-ended universal selection maker sequence in B16F10 murine cancer cell and ID8 murine ovarian cancer cell. SUCCESS generated knock-out clones in two murine cancer cell lines by homozygous deletion of the target genomic region, and without constructing targeting vectors. Thus, our method can be widely applied to generate homozygous knock-out cell lines, as well as knock-in cell lines.