Takahiro Kodama1, Emilie A. Bard-Chapeau2, Justin Y. Newberg1, Michiko Kodama1,
Roberto Rangel1, Kosuke Yoshihara3, Jerrold M. Ward2, Nancy A. Jenkins1, and Neal G. Copeland1.
1 Cancer Research Program, Houston Methodist Research Institute, Houston, Texas
2 Institute of Molecular and Cell Biology,
Agency for Science, Technology and Research, Biopolis, Singapore
3 Department of Obstetrics and Gynecology, Niigata University Graduate School of Medical and Dental Sciences, Niigata, Japan
Published in Gastroenterology in August 2016.
The paper was profiled on the cover and in an editorial in the journal.
BACKGROUND & AIMS: High-throughput sequencing technologies have identified thousands of infrequently mutated genes in hepatocellular carcinomas (HCCs). However, high intratumor and intertumor heterogeneity, combined with large numbers of passenger mutations, have made it difficult to identify driver mutations that contribute to the development of HCC. We combined transposon mutagenesis with a high-throughput screen of a small-hairpin RNA (shRNA) library to identify genes and pathways that contribute to HCC development.
METHODS: Sleeping beauty transposons were mobilized in livers of transgenic mice predisposed to develop hepatocellular adenoma and HCC owing to expression of the hepatitis B virus surface antigen. This whole-genome mutagenesis technique was used to generate an unbiased catalogue of candidate cancer genes (CCGs). Pooled shRNA libraries targeting 250 selected CCGs then were introduced into immortalized mouse liver cells and the cells were monitored for their tumor-forming ability after injection into nude mice.
RESULTS: Transposon-mediated mutagenesis identified 1917 high-confident CCGs and highlighted the importance of Ras signaling in the development of HCC. Subsequent pooled shRNA library screening of 250 selected CCGs validated 27 HCC tumor-suppressor genes. Individual shRNA knockdown of 4 of these genes (Acaa2, Hbs1l, Ralgapa2, and Ubr2) increased the proliferation of multiple human HCC cell lines in culture and accelerated the formation of xenograft tumors in nude mice. The ability of Ralgapa2 to promote HCC cell proliferation and tumor formation required its inhibition of Rala and Ralb. Dual inhibition of Ras signaling via Ral and Raf, using a combination of small-molecule inhibitor RBC8 and sorafenib, reduced the proliferation of HCC cells in culture and completely inhibited their growth as xenograft tumors in nude mice.
CONCLUSIONS: In a 2-step forward genetic screen in mice, we identified members of the Ral guanosine triphosphatase–activating protein pathway and other proteins as suppressors of HCC cell proliferation and tumor growth. These proteins might serve as therapeutic targets for liver cancer.
Figure. 4. Dual inhibition of the Ral and Raf pathways synergistically suppresses HCC cell proliferation in vitro and completely
inhibits tumor growth in vivo. (A) Diagram of Ras signaling pathways and their inhibitors used in this study. (B) WST-1 assay of
SNU-398 (left), Hep3B (center), and PLC/PRF/5 (right) cells after treatment with sorafenib (2 μmol/L for SNU-398 and Hep3B
and 4 μmol/L for PLC/PRF/5) and/or 10 μmol/L of RBC8 for 72 hours (N ¼ 4 each; *P < .05 vs all). (C) Cell counts of SNU-398
(left), Hep3B (center), and PLC/PRF/5 (right) cells after treatment with sorafenib (2 μmol/L for SNU-398 and Hep3B and 4
μmol/L for PLC/PRF/5) and/or 10 μmol/L of RBC8 at the indicated time courses (N ¼ 3 each; *P < .05 vs all). (D) Clonogenic
assay of SNU-398 (top), Hep3B (middle), and PLC/PRF/5 (bottom) cells after treatment with sorafenib (2 μmol/L for SNU-398
and Hep3B and 4 μmol/L for PLC/PRF/5) and/or 10 μmol/L of RBC8 for 72 hours. (E) WST-1 assay of SNU-387 (left), SNU-449
(center), and SNU-475 (right) cells after treatment with 4 μmol/L of sorafenib and/or 10 μmol/L of RBC8 for 72 hours (N ¼ 4
each; *P < .05 vs all). (F) SNU-398 cells (2.0 x 106) were injected subcutaneously into nude mice. Once xenograft tumor
volumes reached 100 mm3 (6 days after injection), mice were assigned randomly into 4 groups: control, RBC8, sorafenib, or
RBC8þsorafenib. Mice then were treated with oral gavage of 50 mg/kg RBC8 and/or intraperitoneal injection of 50 mg/kg
sorafenib 5 days a week for 12 days. Xenograft tumor volumes then were measured at different times after the initiation of
treatment (N ¼ 4 for RBC8þsorafenib and 3 for all other cohorts; *P < .05 vs all).