expression profiles of PC3CR by microarray to investigate the mecha- nisms of CBZ resistance. RESULTS: We incubated PC3 cells with gradually increasing concentrations of CBZ for 1.5 years, and established a CBZ-resistant cell line, PC3CR. PC3CR cells underwent cell division with 3 nM CBZ. We compared the cytotoxic effect of CBZ on PC3 and PC3CR cells using a cell viability assay. The half maximal inhibitory concentration (IC50) of CBZ in PC3 and PC3CR cells were 11.0 nM and 3.7 nM, respectively. Mice PC3CR xenograft tumors also had CBZ resistance (Fig. A) Functional annotation clustering (FAC) analysis using micro- array data demonstrated cell division (GO: 0051301) and mitotic nu- clear division (GO: 0007067) were the most enhanced clusters in PC3CR compared to PC3. These results suggested that enhancement of cell cycle progression signaling was related with CBZ resistance in CRPC. We perform in silico screening for compounds overcoming CBZ resistance by Connectivity Map (CMAP) analysis. Etoposide (VP16) was one of the candidate drugs which reverted gene expression pattern of CBZ-resistant cells into CBZ-sensitive cells. Topoisomerase IIa (TOP2A) is a main target of VP16, so we examined TOP2A expression in PC3CR. Quantitative PCR showed up-regulation of TOP2A in PC3CR. In cell viability assay, 1mM etoposide significantly inhibited PC3CR proliferation (80.3 Æ 3.2%). VP16 is ordinary used with plat- inum-based agents for neuroendocrine tumor. We tested cytotoxic ef- fect of CDDP for CBZ-resistant CPRC. The relative cell viabilities of PC3CR treated with 3mM CDDP was 86.4 Æ 1.6%. CDDP was also effective for CBZ-resistant CRPC cells. Next, we tested anti-tumor effect of VP16 and CDDP using PC3CR xenograft tumor model. Both single- agent treatments with VP16 and CDDP significantly inhibited PC3CR xenograft tumors (Fig. B). Moreover, VP16 and CDDP in combination use had a synergic effect for the xenograft tumors. CONCLUSIONS: Etoposide based chemotherapy may be an optimal treatment for CPRC in the post-cabazitaxel setting. Source of Funding: This work was supported by JSPS KAKENHI Grant Number 17K16813 to Hongo H, and 25861451 to Kosaka T from the Ministry of Education, Culture, Sports, Science, and Technology of Japan. MP70-13 DEPLETION OF PMEPA1 GENE CONFERS INCREASED CELL PROLIFERATION IN MOUSE PROSTATE EPITHELIUM Hua Li*, Shashwat Sharad, Talai Barbiev, Wei Huang, Yingjie Song, Denise Young, Rockville, MD; Issabell Sesterhenn, Silver Spring, MD; Inger Rosner, Bethesda, MD; Albert Dobi, Shiv Srivastava, Taduru Sreenath, Rockville, MD INTRODUCTION AND OBJECTIVES: Dysfunctions of androgen receptor (AR) and transforming growth factor-b (TGF-b) signaling play important roles in tumorigenesis of prostate cancer (CaP). PMEPA1 gene is androgen and TGF-b -responsive with abundance in human prostate gland. The PMEPA1 biological func- tions involve suppression of AR and TGF-b signaling through similar negatively regulated feedback loop. Loss of PMEPA1 gene expression was reported in primary tumors of over two thirds of CaP patients may facilitate the activation AR and TGF-b signaling and supporting androgen dependent and independent progression of CaP respectively. Here we describe the evaluation of Pmepa1 gene knock-out model with conditionally silencing in mouse prostate epithelium to further dissect its biological functions in prostate development and tumorigenesis in vivo. METHODS: Pmepa1 gene was conditionally deleted in C57BL/ 6 mouse prostate epithelium by ARR2PB-Cre-Lox system. Male mice of genotypes of wild-type, Pmepa1 flox/wild-type-ARR2PB-Cre, Pmepa1 flox/flox-ARR2PB-Cre were analyzed at age of 3, 9 and 15 months. The prostate and control tissues were sectioned and stained with Hema- toxylin & Eosin (H&E). The protein levels of Pmepa1, Ar, Nkx3.1, Pten, Cre, Akt and proliferating cell nuclear antigen (Pcna) were analyzed by immunohistochemistry (IHC). The transcript levels of Pmepa1 were evaluated by RNA in situ hybridization (RISH). RESULTS: The data of IHC and RISH showed Pmepa1 gene was suppressed in Pmepa1 flox/flox and Pmepa1 flox/wild-type ARR2PB-Cre mice. Compared to wild-type mice, protein levels of Ar and AR responsive gene Nkx3.1 were enhanced in prostate epithelium in Pmepa1 heterozygous and homozygous knockout mice at the age of 3, 9 and 15 months. The protein level of Pten was not significantly different in Pmepa1 conditional knockout mouse. However, the protein levels of Akt were increased in prostate epithelium with Pmepa1 deletion at the age of 9 and 15 months. Importantly, stronger Pcna IHC staining was observed in knockout prostate epitheliums, highly suggesting Pmepa1 silencing leading to accelerated prostate epithelium proliferation via activated Ar and Akt signaling. CONCLUSIONS: Conditional deletion of Pmepa1 gene in mouse prostate epithelium resulted in enhanced Ar protein, activated Ar/Akt signaling and accelerated cell growth. This new mouse model provides opportunities to assess Pmepa1 functions in prostate and other cancers where PMEPA1 is altered. Source of Funding: CPDR, USUHS, HU0001-10-2-0002 to I.L.R. MP70-14 REGULATION OF PROSTATE CANCER METABOLISM AND INVASIVENESS BY THE LIVER X RECEPTOR Allison May*, Suomia Abuirqeba, Zachary Hamilton, Colin Flaveny, Sameer Siddiqui, St. Louis, MO INTRODUCTION AND OBJECTIVES: Prostate cancer (PCa) progression is marked by characteristic changes in PCa cell meta- bolism. During early PCa cell transformation, activation of the oncogene, c-MYC, fosters a metabolic switch from glycolytic to oxidative metabolism and increases lipid synthesis. This switch in metabolic activity and increased expression of lipid synthesis genes is characteristic of castration resistant metastatic PCa. The liver X receptor (LXR) is a transcriptional regulator expressed in prostate epithelial cells that plays an important role in cholesterol, lipogenic and carbohydrate homeostasis. Recently, it has been demonstrated that loss of the bona-fide PCa tumor-suppressor, PTEN, results in oncogene driven dysregulation of LXR-target gene expression. Therefore, we theorized that LXRs may play a role in directing PCa metabolism and treatment resistance. METHODS: First, we assessed DNA and RNA sequencing data from prostate tumor samples to determine if lipid synthesis enzyme expression correlated with LXR expression in tumors. Second, we compared the expression of LXRs and LXR target genes in culture PCa cells that display differential dependence on androgens for growth. Third, we tested whether disruption of LXR transcriptional activity using pharmacological and genetic ap- proaches could inhibit prostate tumor metabolism and block PCa metastasis. RESULTS: Prostate tumor samples had more than twofold higher LXR target gene expression than normal prostate samples. Expression of LXR b positively correlated with the expression of lipid synthesis genes SREBP1c, FASN, and SCD1 (pearson r¼.93, p¼.03) that drive cancer cell metabolic activity. LXR expression Vol. 199, No. 4S, Supplement, Sunday, May 20, 2018 THE JOURNAL OF UROLOGY â e939