Using 2 powerful high-throughput in vivo screening tools, researchers identified and validated 27 genes as suppressors of hepatocellular carcinoma (HCC), reported in the August issue of Gastroenterology. Their findings support the role of Ras signaling in development of HCC and provide new therapeutic targets.
Sorafenib is the only treatment for patients with advanced-stage HCC, but this drug prolongs patient survival by only a few months. To discover effective treatments for HCC we need to learn more about the mechanisms of HCC development.
Genetic analyses have identified only a few genes (TERT, CTNNB1, TP53, and ARID2) that are mutated in more than 5% of HCCs. In contrast, infrequent mutations (<5% of patients) have been identified in more than 10,000 genes. HCC is therefore a highly heterogenous tumor.
In contrast to these reverse genetic analyses, which start with a biologic feature and look for gene variants that cause it, in forward genetic analyses, researchers disrupt random genetic sequences and see what phenotypes develop. This approach provides an unbiased selection for genes that control a phenotype of interest, such as HCC development.
A forward genetic screen using sleeping beauty transposon mutagenesis of transgenic mice predisposed to develop hepatocellular adenoma and HCC (as a result of the expression of hepatitis B surface antigen, HBsAg) had previously identified 21 genes that regulated liver tumorigenesis and tumor progression. In this system, cells that acquire a survival advantage, through transposon-mediated activation or inactivation, give rise to tumors. However, this original system could not cover the entire genome, due to technical limitations.
Takahiro Kodama et al used a system (T2Onc3) that overcame these obstacles and increased the number of tumor-associated genes identified by sleeping beauty-based mutagenesis. They then screened a high-throughput short hairpin RNA (shRNA) library for the effects of knockdown of these genes in immortalized mouse liver cells.
Mice that carried the HBsAg transgene and were undergoing active T2Onc3 transposition in their liver started to develop tumors at approximately 1 year of age, developed more advanced tumors, and had significantly shorter survival times than HBsAg mice without active sleeping beauty transposition. Even some mice with the transposon without HBsAg developed advanced liver tumors.
Kodama et al identified 1917 genes at which transposon insertion promoted liver tumor development. Analyses of their products revealed the importance of Ras signaling in the development of HCC.
A shRNA library screen of 250 selected genes validated 27 suppressors of HCC development. Individual shRNA knockdown of 4 of these tumor suppressor genes (Acaa2, Hbs1l, Ralgapa2, and Ubr2) increased proliferation of multiple human HCC cell lines and accelerated formation of xenograft tumors in nude mice. The authors concluded that these genes are bona fide tumor suppressors in mouse and human liver.
When the authors compared the outcomes of patients whose HCCs carried a mutation in these tumor suppressors, using a publically available microarray data set, they found that expression levels of 9 of these genes had a negative correlation with patient survival.
Ralgapa2 encodes a member of the Ral guanosine triphosphatase (GTPase)-activating protein (RalGAP) family, which suppresses the oncogenic Ral pathway. The authors found that 51% of sleeping beauty-induced tumors had a transposon insertion in 1 or more of these family members.
The ability of Ralgapa2 knockdown to promote HCC cell proliferation and tumor formation required expression 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.
Ralgapa2, and the closely related Ralgapa1 and Ralgapb family members, negatively regulate the RalA and RalB Ras-like small GTPases through the hydrolysis of GTP. Although oncogenic roles for the Ral pathway have been reported in pancreas, skin, lung, colon, prostate, and bladder cancer, little is known about the mechanisms of RalGAPs in cancer development. Down-regulation of RALGAPA2 promotes invasion and metastasis of bladder cancer. Kodama et al showed that that RalGAPs suppress cell proliferation and growth of xenograft tumors from HCC cells via inactivation of the Ral pathway.
Kodama et al noted that the oncoding RNA Rian was mutated in the highest number of tumors and showed a dense pattern of insertions in the antisense strand. Sleeping beauty insertions at Rian promoted liver tumor development through activation of the promoter of an adjacent oncogene, Rtl, rather than through insertional inactivation of Rian.
The second and third most highly ranked trunk drivers (genes mutated or deregulated at an early stage of tumor development) identified were Hras and Kras. Although Ras is not frequently mutated in HCCs, this pathway is activated in 50%–100% of HCCs, through down-regulation of Ras inhibitors such as GTPase-activating proteins and Rassf and Spred family members. So the Ras pathway appears to be a good target for development of anti-HCC agents. Sorafenib targets RAF/MEK/ERK members of the Ras pathway.
In an editorial that accompanies the article, Andreas Teufel and Jean François Dufour write that the findings nicely demonstrate the usefulness of the sleeping beauty transposon system, in combination with simultaneous gene expression profile analyses, to identify genes that regulate tumor development. The findings could provide new therapeutic targets for HCC and support the investigation of transposon systems for cancer gene therapy.