In the initiation and progression of hepatocellular carcinoma, genetic factors usually play an important role. Meanwhile, mRNA gene signature based on a certain characteristic like glycolysis [11] and immune [12] have been developed for prognosis of cancers. In this research, we explored specific function to identify genes by comparing survival group and death group in GSEA. According to our results, six signal pathways was found to be highly related to survival and we established the gene signature with mTORC1 signal pathway. As we all known, mTOR pathway is a serine/threonine protein kinase belonging to the PI3K-related kinase family [13], which comprised of two distinct complexes (mTORC1 and mTORC2). With Raptor as its unique and key protein component, mTORC1 plays an important role in cell survival, autophagy, and metabolism [14]. Concerning for the mTOR signal pathway in HCC, it has been found that aberrant mTOR signaling was present in half of the HCC cases [15]. Meanwhile, an intact mTORC1 axis [16] and mTORC2-Akt1 cascade [17] were required for c-Myc-driven hepatocarcinogenesis. Moreover, some researches [15, 18] provided the theoretical basis of mTOR signaling pathway-oriented targeting treatment for HCC in clinic. Overall, mTOR signal pathway plays an important role in the development and progression of HCC.
In this study, we identified six genes in signature by performing the differentially expressed analysis, univariate Cox regression analysis and lasso regression analysis. Among our included genes, five genes (ETF1, GSR, HSPD1, CACYBP and PNP) have been found to be related to HCC from previous studies. Singh et al found that ETF1, CNOT6 and XRN1 gene in HepG2 cell led to significant alteration in stability of specific mRNAs and this mechanism may hold novel cancer therapeutic targets [19]. In another research, McLoughlin concluded that GSR, TRXR1, NRF2 and oxidative stress determined hepatocellular carcinoma malignancy [20]. Lee’s study [21] found that HSPD1 was down-regulated during early apoptosis of the hepatoma cell mediated by Paeoniae Radix. In terms of CACYBP, it have been verified that CACYBP can promote hepatocellular carcinoma progression in the absence of RNF41 mediated degradation [22]. Moreover, a study [23] found that PNP/fludarabine suicide gene system induced HCC cell apoptosis and inhibited the growth of HCC cells. Although we found no evidence supporting the correlation between SKAP2 and HCC, it has been verified that SKAP2 promotes podosome formation to facilitates tumor-associated macrophage infiltration and metastatic progression [24].
Being different from the previous prognostic studies in HCC, our predictive model firstly concentrated on mTORC1 signal pathway. More importantly, mTORC1 signal pathway was identified by GSEA, which indicated the underlying mechanism between survival of HCC and mTORC1 pathway. Moreover, the validation from three different datasets and a rigorous screening process enabled the identification of a reliable signature. However, our study has some limitations. First, prognostic signature showed a relatively low diagnostic performance, which may be attributed to the inadequate genes (199 related genes) in the initiation of screening process. More genes associated with mTORC1 signal pathways need to be confirmed in the future. Second, using a single characteristic (mTORC1 signal pathway) to establish the predictive model is an intrinsic weakness. Indeed, many other mechanisms, such as metabolism [25] and immune [8], have an effect on the progression of HCC. Furthermore, it’s necessary to perform more independent trials and functional experiments to shed light on the mechanism linking mTORC1 signal pathway and HCC.