Filipin III br SPDEF transcriptionally represses ZBTB in pro
3.2. SPDEF transcriptionally represses ZBTB46 in prostate cancer cells
Unlike luminal-type tumor cells, NE Filipin III do not express AR, and AR signaling is repressed . We next validated ZBTB46 expression in the Cancer Genome Atlas (TCGA) prostate dataset by a gene set enrichment analysis (GSEA) with the gene signatures that reflect AR signaling components [51,52]. We found that the tissues expressing low levels of ZBTB46 were more likely to have gene signatures associated with ac-tivated androgen-responsive signatures (Supplementary Fig. S2A). We identified SPDEF as a potential regulator of ZBTB46 in the core en-richment group (Supplementary Fig. S2B). SPDEF is a critical tumor suppressor in prostate cancer, and it is highly expressed in normal prostate epithelial cells . To further confirm the inverse association between SPDEF and ZBTB46, we found that knockdown of SPDEF in the AR-positive cells induced ZBTB46 mRNA expression (Fig. 2A), whereas the SPDEF-overexpressing AR-negative cells had significantly lower ZBTB46 mRNA levels (Fig. 2B). In addition, SPDEF could directly reg-ulate ZBTB46 by transcriptional function. Based on our ChIP-Seq data analyzed from a published dataset (GSE48930) , SPDEF may be physically associated with the ZBTB46 promoter (Supplementary Fig. S2C). We hypothesized that SPDEF transcriptionally represses ZBTB46 in prostate cancer cells by directly binding to the SPDEF response ele-ment (SRE) on the ZBTB46 promoter. We next looked for sequences resembling the putative SRE in the ZBTB46 promoter region and found 10 SREs located between −3 and −1 kb upstream of the transcriptional start site (Fig. 2C). The ChIP results confirmed that the most-enriched ChIP product was found at the putative SPDEF-binding sites, SRE2 and SRE3/4 (Fig. 2D). Moreover, the SRE2 and SRE3/4 sites were found to have increased SPDEF binding in response to DHT (Fig. 2E), but this binding was reduced in the cells after MDV3100 treatment (Supplementary Fig. S2D). We next performed reporter assays with serially deleted promoter constructs containing SREs (Fig. 2C) and found that E2 and E3 reporters had the lowest reporter activities compared with the E1 reporter (Fig. 2F). Wild-type (WT) E2 and E3 reporter activities were reduced in the presence of DHT (Fig. 2G), whereas E2 or E3 mutants did not change (Fig. 2H). Moreover, de-creased reporter gene activities were detected when WT E2 and E3 reporter constructs were cotransfected with SPDEF expression vector in RasB1 cells, whereas E2 and E3 mutants disrupted the repressive ability of SPDEF (Fig. 2I). These data are consistent with a mechanism whereby SPDEF represses ZBTB46 transcription by directly and physi-cally interacting with the ZBTB46 promoter.
3.3. Loss of SPDEF is associated with NE diﬀerentiation of prostate cancer cells
We next analyzed the relationship between SPDEF and ZBTB46 in a panel of prostate cancer cells and found an inverse relationship between SPDEF and ZBTB46 (Fig. 3A). These results were confirmed in the prostate cancer cells treated with DHT or MDV3100. As expected, DHT increased SPDEF abundance and ZBTB46 reduction, whereas the treatment with MDV3100 resulted in a decrease in SPDEF and an in-crease in ZBTB46 in the AR-positive cells (Fig. 3B, left and middle). These results were confirmed in the AR-negative cells, which showed nonsignificant changes in SPDEF and ZBTB46 following perturbation of AR signaling (Fig. 3B, right). We next examined whether ZBTB46 abundance was aﬀected by SPDEF, which is involved in NEPC diﬀer-entiation. We found that the NE markers (CHGA, CHGB, and ENO2) increased in the LNCaP and C4-2B cells following SPDEF-knockdown (Fig. 3C). These results were confirmed by Western blots, which showed that the SPDEF protein decreased and that the ZBTB46 and ENO2 proteins increased following SPDEF-knockdown (Fig. 3D). Moreover, induction of ZBTB46 and ENO2 proteins was observed in the C4-2B cells in the presence of SPDEF shRNA regardless of DHT treatment (Fig. 3E). We also found that SPDEF overexpression decreased the protein levels of endogenous ZBTB46 and ENO2 even in the cells
Fig. 1. ZBTB46 induces NE marker expressions of prostate cancer. (A) Immunoblotting for AR, ZBTB46, and ENO2 in various prostate cancer cell lines. (B–D) ZBTB46, CHGA, and ENO2 mRNA levels after MDV3100 treatment at 10 μM in FBS-containing medium for 2 weeks of various prostate cancer cell lines. (E) Immunoblotting of extracts from 22Rv1 and C4-2B cells following MDV3100 treatment for 2 weeks. (F) ZBTB46 and NE markers (CHGA, CHGB, and ENO2) in 22Rv1 and C4-2B cells following transient transfection with ZBTB46 or an empty vector (EV) by qRT-PCR. (G) ZBTB46, CHGA, and ENO2 proteins in 22Rv1 and C4-2B cells following stable transfection with EV and ZBTB46 expression vectors. (H) ZBTB46 and NE markers in RasB1 and PC3 cells stably expressing ZBTB46 shRNA or control vector (shLuc) by qRT-PCR. (I) Immunoblots showing ZBTB46, CHGA, and ENO2 in RasB1, PC3, and NE-1-8 cells following control (NC) or ZBTB46 siRNA (si46) transfection. (J) ZBTB46 and NE markers (CHGA, CHGB, and ENO2) in C4-2B cells transiently transfected with NC or ZBTB46 siRNA and then treated with MDV3100 in FBS-containing medium for 24 h by qRT-PCR. The quantification results of mRNA are presented as the mean ± SEM, n = 3 biological replicates. Significance was determined by Student's t-test. *p < 0.05, **p < 0.01, ***p < 0.001.