Masaki Mori, Ryutaro Uchi, Atsushi Niida, Sho Nambara, Yuki Takano, Masami Ueda, Hidetoshi Eguchi, Keishi Sugimachi, Yoshiaki Shinden, Yuichiro Doki, Hidenari Hirata, Shotaro Sakimura, Hisateru Komatsu, Tomohiro Iguchi, Koshi Mimori, and Tomoko Saito
Introduction Hepatocellular carcinoma (HCC) is the most common type of liver cancer and the third leading cause of all cancer deaths worldwide. A fraction of HCC has increased activity of Wnt/β-catenin (Wnt) pathway, which reportedly enhances malignant phenotypes like epithelial-mesenchymal transition (EMT) in various kinds of cancer. Identification of Wnt pathway target and regulator genes gives us better understanding of molecular biological behavior of HCC. In this study, we performed in silico screening for a novel Wnt pathway target and regulator genes in human HCC using public databases. Method First, we obtained TCGA data sets of human HCC including mRNA expression (268 samples) and mutational status (198 samples) from the Broad GDAC Firehose web site (gdac.broadinstitute.org/). Next, for β-catenin transcriptional target genes prepared by literature curation, we performed EEM analysis (Niida et al. 2009) on the HCC expression data set to obtain Wnt pathway target genes in HCC. Assuming the sum of mRNA expression of the Wnt pathway target genes as Wnt pathway activity, we then measured association between the pathway alteration and the activity by linear multiple regression analysis ([Wnt pathway activity] = a0 + a1[CTNNB1 mutation] + a2[AXIN1 mutation] + a3[AXIN2 mutation]+a4[APC mutation]+ a5[CTNNB1 mRNA expression] : a0∼a5; partial regression coefficient). By focusing on known epithelial markers (CDH1, KRT18, MUC1, OCLN, TJP1) and mesenchymal markers (ACTA2, CDH2, FN1, VIM, VTN), we also examined association between the Wnt pathway activity and EMT-manifesting genotype. Finally, we performed multiple regression analysis to search for novel Wnt pathway regulator genes ([Wnt pathway activity] = b0 + b1[CTNNB1 mutation] + b2[APC mutation]+ b3[CTNNB1 mRNA expression] + b4[mRNA of a regulator] : b0∼b4; partial regression coefficient). Result (1)EEM identified AASS, ASPSCR1, BMP4, C6orf97, DUT, GGH, GNPAT, GREB1, HGD, IKBKAP, ITPR2, LAMA3, LGR5, NEK3, RELN, RHBG, RHOBTB1, SLC13A3, SMYD2, TBX3, TDGF1, TRIB2, VSNL1, ZBTB38 as the Wnt pathway target genes in HCC. (2)Multiple regression analysis revealed the Wnt pathway activity had strong correlations with CTNNB1 mutation (p-value < 2.00×10-16), APC mutation (p-value = 2.73×10-3), and CTNNB1 expression (p-value = 2.19×10-07) (3)The Wnt pathway activity had a strong correlation with the EMT genotype (p-value = 3.58×10-4). (4)We identified 100 statistically significant Wnt pathway regulator candidates (q-values < 1.00×10-10). Conclusion Our in silico screening identified novel Wnt pathway target genes in HCC, whose expression profiles were correlated with Wnt pathway alterations and the EMT-manifesting gene signature. We also have predicted Wnt pathway regulator genes, and are now conducting in vitro validation. Citation Format: Hisateru Komatsu, Atsushi Niida, Masami Ueda, Hidenari Hirata, Ryutaro Uchi, Sho Nambara, Tomoko Saito, Shotaro Sakimura, Yuki Takano, Yoshiaki Shinden, Tomohiro Iguchi, Hidetoshi Eguchi, Keishi Sugimachi, Yuichiro Doki, Masaki Mori, Koshi Mimori. In silico screening for novel Wnt/β-catenin pathway target and regulator genes in human hepatocellular carcinoma. [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 1948. doi:10.1158/1538-7445.AM2015-1948