HCC is a global cancer health concern. Its complexity and bad prognosis preclude effective access to treatment. Finding a helpful biomarker and building a reliable model to predict patient’s prognosis are of general interest for HCC therapy.
Assorted biomarkers for HCC diagnosis and prognosis have been screened in the past few decades, and several prognostic prediction models were constructed. A risk model was established based on 7 autophagy-related genes (SPHK1, HSPB8, ITGA3, CDKN2A, BIRC5, IKBKE and TMEM74) in HCC (Jiangtao Wang et. al15). Potential core genes associated with HCC progression and prognosis were identified by bioinformatics analysis: CCNB1, CCNA2, CCNB2, NCAPG, PBK, NUSAP1, AURKA, ZWINT, PRC1 and KIF4A (Xiudao Song et. al16). This paper identified 6 gene signatures (EIF2S1, BIRC5, SQSTM1, ATG7, HDAC1, FKBP1A) by differential expression analysis, univariate analysis and multivariate analysis in TCGA-LIHC. These autophagy-related DEGs in HCC were used to build a prognostic prediction risk model.
Regulation of these 6 genes were proved in assorted cancers except HCC. Sequestosome 1 (SQSTM1) encodes a multifunctional protein that binds to ubiquitin and activates the nuclear factor kappa B (NF-kB) signaling pathway. P62 was confirmed to inhibit autophagy flux and promote epithelial-mesenchymal transformation in metastatic prostate cancer by maintaining the level of HDAC617. Sequestosome 1 is an effective prognostic factor associated with cell proliferation in human colorectal cancer18. In addition, p62 is upregulated in the prophase of HCC and induces cancer by maintaining the survival of stress-induced HCC initiating cells19. HDAC1 plays a key role in regulating eukaryotic gene expression. It has been reported to promote glycolysis in gastric cancer, and its high expression is an independent adverse prognostic factor for OS and disease-free survival20. Silencing of HDAC1 enhances the sensitivity of ovarian cancer to chemotherapy21. HDAC1 restrains Snail2-mediated epithelial-mesenchymal transition (EMT) in the process of metastasis of HCC (Yue Hu et al.22). Viewed in toto, these genes participate in progression of HCC, which is consistent with this paper.
Besides, 4 feature genes were not yet well defined in HCC. EIF2S1 catalyzes the first regulatory step in the initiation of protein synthesis to promote the binding of the initial tRNA to the 40S ribosome subunit. Phosphorylated eIF2α has been found to predict disease-free survival in triple-negative breast cancer patients23. Estrogen-induced apoptosis of breast cancer cells takes place by blocking dephosphorylation of eIF2α protein24. In addition, Li Suzhen et al.25 demonstrated that TIPRL enhances survival of lung cancer by inducing autophagy through the eIF2α-ATF4 pathway. BIRC5 is a member of the IAP gene family, which encodes negative regulatory proteins that prevent the death of apoptotic cells. It has been shown that the long non-coding RNA (lncRNA) nR2F1-AS1 promotes the malignant phenotype of osteosarcoma cells through the miR-485-5p/miR-218-5p/BIRC5 axis26. Inhibition of BIRC5 improves cervical cancer cells sensitivity to radiotherapy27. Nine genes, including BIRC5, may be biomarkers for HCC28. FKBP1a encodes proteins that are members of the immunomodulatory family, and it plays a role in immunomodulatory and fundamental cellular processes involved in protein folding and transport. Ten genes, including FKBP1A, were identified as biomarkers for breast cancer29. LncRNA SNHG15 promotes prostate cancer cell proliferation, migration and EMT by regulating miR-338-3p/FKBP1A axis30. Luo Yue et. al31 reported FKBP1A overexpression in HCC. ATG7encodes an E1-like activator that is critical for autophagy and cytoplasmic transport to vacuoles. Studies have shown that ATG7 adjusts three negative breast cancer tumor progression32. MiR-154 exerts suppressor role by directly targeting ATG7 in bladder cancer33. In the context, the investigation of mechanism of SQSTM1 and HDAC1 in the 6 feature genes was in its infancy. EIF2S1, BIRC5 and FKBP1A were only identified as biomarkers of HCC, and ATG7 was not reported to be associated with HCC. Hence, this paper may provide theoretic basis for studying these genes in HCC.
On the whole, 6 autophagy-associated genes were identified via bioinformatics analysis (EIF2S1, BIRC5, SQSTM1, ATG7, HDAC1, FKBP1A), and corresponding prognostic risk model was constructed. Our finding will yield valuable insight into early diagnosis, prognosis, and development of new therapies. However, application of these 6 feature genes requires validation by incremental clinical experiments and animal experiments.