HCC is one of the most lethal human cancers. With the development of clinical management of liver cancer, some prognostic factors including tumor volume, grade and stage and the number of lesions are characterized 7, 8. But, an effective molecular biomarker for HCC prognosis monitoring is still urgently needed. There is an increasingly emerging evidence suggest that autophagy is closely implicated in tumorigenesis and progression 9. The exploration of mechanism about autophagy opens up a new prospect approaching for liver cancer. In present, high-throughput biological technologies have been widely applied in the early diagnosis of cancer. Therefore, using large-scale databases will help to explore the expression pattern of ARGs and reveal the prognosis of HCC patients.
In the current study, based on the high-throughput expression data from two datasets (TCGC and ICGC), we aimed to screen key ARGs having intimate connection with the prognosis of patients with HCC. Firstly, 22 ARGs were found having differentially expression between liver tumor and normal tissues, including 3 up-regulated and 19 down-regulated genes. The results of GO terms and KEGG pathways analysis showed the major enrichment in tumor biological process and molecular function. It was observed that the KEGG pathways enriched by the differentially expressed ARGs was involved in several types of cancer. The results suggested that specific autophagy pattern may played a role in occurrence and progression of liver cancer. Then we observed 14 risk ARG related to OS in the TCGA database by dimensionality reduction in univariate survival analysis. And further multivariate survival analysis decided five key prognostic ARGs (ATIC, BAX, BIRC5, CAPNS1, FKBP1A) to construct prognostic risk models, which could provide accurate prognosis for patients with liver malignant tumor. Multivariate Cox regression analysis of prognostic risk model and clinical parameters showed that the five-gene risk score is conducive to independently predict the prognosis of patients with liver cancer. Meanwhile the results of ROC curve analysis suggested that the prognostic risk model could distinguished healthy people from liver cancer patients accurately.
ATIC, a bi-functional protein enzyme, catalyzes the final two steps of the de novo purine biosynthetic pathway. Recent studies have shown that ATIC is expressed at high levels in lung cancer and related to poor patient prognosis 10. There are few reports on its function in liver cancer. It has reported that up-regulated of ATIC in HCC was correlated with poor survival and it supported propagation of HCC cells by regulating AMPK-mTOR-S6 K1 signature 11.
Bax has been proved to be one of the most widely-characterized participants in autophagy. It is a member of the BCL2 protein family and functions as an apoptotic activator by forming a heterodimer with BCL2. Bax participates in the mitochondria-initiated intrinsic apoptotic signaling and the extrinsic apoptotic pathway that is triggered by transmembrane death receptors 12, 13. Also, Bax has the effect in the crosstalk between endoplasmic reticulum (ER) signaling pathway and mitochondrial pathways 14. Considering Bax be important role in apoptosis, it is not unexpected that Bax participates in multiple anticancer drugs that induce apoptosis of cancer cells. SKA3, a component of the spindle and kinetochore-related complexes, influences cell apoptosis by regulation of BAX/Bcl-2 expression in HCC cells 15. The modulation of MT1G (a low-molecular weight protein with high affinity for zinc ions) on p53 resulted in upregulation of Bax, which leads HCC cells apoptosis 16.
BIRC5 belongs to the inhibitor of apoptosis (IAP) family, which can prevent apoptotic cell death. It has been reported that BIRC5 exerted its influence on HCC cells in promoting proliferation 17. Recent studies have shown that BIRC5 could be regarded as potential biomarkers for molecular diagnosis as well as therapeutic intervention of HCC and with significant influence on prognosis of HCC patients 18, 19.
CAPNS1, a member of the calpain small subunit family, operate as heterodimers and essential for the calpain activity and function. Several studies have shown that CAPNS1 is associated with biological function to tumorigenesis. Indeed, the expression of CAPNS1 has been detected in various cancers, including hepatocellular carcinoma 20, ovarian carcinoma 21, colorectal cancer 22, nasopharyngeal carcinoma 23, and other tumors. In vitro experiments found that CAPNS1 enhances the growth and metastasis of HCC by activating FAK-Src signaling pathway and MMP2 24.
FKBP1A is a cis-trans prolyl isomerase that binds to FK506 and rapamycin, and inhibits calcineurin and the activity of mTOR 25. FKBP1A was validated to have overexpressed in HCC and predicted the poor prognosis 26. However, the specific mechanism of FKBP1A in HCC is still unclear, which deserve to be further investigated.
In summary, there is a close relationship between autophagy and the prognosis of liver cancer patients. In addition, this study suggests that risk ARGs be potential candidates for prognostic biomarkers in HCC with the value to guide making decisions for the choice of clinical treatment. Further researches are needed to provide more specific information about these results. At the same time, based on the discovery of autophagy-related patterns which could impact the prognosis of patients with tumor, our work conducted a risk model to benefit distinguish high-risk liver cancer patients. However, the main limitation of our study is that we used already available data from two public databases, and above results are required to be further investigated in prospective studies.