Abstract 5169: Omics approach to identify driver genes for peritoneal dissemination of gastric cancer cells

Abstract

Abstract Background Little is known about the molecular mechanisms of peritoneal dissemination in gastric cancer cells. In this study, to clear the molecular mechanisms of accelerating the peritoneal dissemination, we performed to identify significant genes which give rise to peritoneal dissemination and poor prognosis of gastric cancer. We conducted expression array analysis in mice model of peritoneal dissemination and in 200 cases of gastric cancer cases, and compared with each other to identify driver genes. Methods As described previously, three parental scirrhous gastric carcinoma-derived cell lines (HSC-39, HSC-44PE and HSC-58) were established from human scirrhous gastric carcinoma. Moreover, inoculating each parental cell lines into the gastric wall of nude mice, four matched highly peritoneal-metastatic cell lines (HSC-39As8, HSC-44As3, HSC-58As1 and HSC-58As9) were established after 12 cycles of stepwise selection. To identify gene expression signatures for peritoneal dissemination, we performed expression analysis for matched parental and each resultant cell line. Next, to examine the relationship between gene expression signatures and clinical behaviors, we performed synthetic analysis of gene expression signatures from in vivo peritoneal dissemination mouse model data and expression data set containing 200 gastric tumors from the Singapore cohort. Results (1) Our synthetic analysis, EEM analysis, which tested the functionality of the input genes sets on the basis of their coherence, showed that one gene set from the 58As9 cell line harbored a coherently expressed subset of significant size. (2) We found 57 genes in the coherent subset which had clinical correlation. (3) We found that peritoneal metastases patients tended to show higher expression of such genes, and patients prognosis with higher expression of such genes tended to be poor. (4) Among such genes, we focused on chemokine(C-X-C motif) receptor 7(CXCR7). CXCR7 expression was significantly higher in 58As9 than in HSC58(P<0.005). We also found that overall survival of patients with CXCR7 high expression gastric cancer tended to be poorer than that with CXCR7 low expression gastric cancer in both the Singapore and Japanese cohorts. Conclusions In conclusion, synthetic analysis from two data sets indicated several candidate driver genes for strongly accelerating gastric cancer peritoneal dissemination, the current approach to compare gene signatures in vivo and clinical samples with accurate clinical annotations enabled us to identify brand-new driver candidate genes. Among them, CXCR7 might play an important role in peritoneal dissemination and CXCR7 might be a therapeutic target for suppression of gastric cancer peritoneal dissemination in the near future. Note: This abstract was not presented at the meeting. Citation Format: Sho Nambara, Junji Kurashige, Tomoko Saito, Hisateru Komatsu, Masami Ueda, Shotaro Sakimura, Hidenari Hirata, Ryutaro Uchi, Yuki Takano, Yoshiaki Shinden, Tomohiro Iguchi, Hidetoshi Eguchi, Keishi Sugimachi, Yoshihiko Maehara, Koshi Mimori. Omics approach to identify driver genes for peritoneal dissemination of gastric cancer cells. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 5169. doi:10.1158/1538-7445.AM2015-5169