Effect Of CeRNA Network Mediated By Hypoxia-related LncRNAs On Prognosis, Treatment Responsiveness And Biological Function Analysis Of Hepatocellular Carcinoma

Objective: Long non-coding RNAs can regulate hypoxia-induced remodeling of the tumor microenvironment and tumor progression.However,the clinical significance of hypoxiarelated biomarkers in hepatocellular carcinoma(HCC)remains largely unclear.The aim of this study was to screen hypoxia-related lncRNAs by bioinformatics approach to construct a risk score model and to assess the prognosis of HCC patients and to explore the underlying mechanisms.Methods: RNA sequencing data downloaded from the cancer genome atlas(TCGA)and gene expression omnibus(GEO)databases and clinical follow-up data were used,while hypoxia-related genes were downloaded from the molecular signature database.Correlation analysis,differential analysis and univariate Cox regression analysis were applied to obtain candidate hypoxia-related lncRNAs.365 HCC patients were randomly divided into training and test datasets at the ratio of 7:3,and least absolute shrinkage and selection operator(LASSO)-Cox regression was performed to screen and develop hypoxia-related lncRNAs signature(HRLS)models in the training dataset.The median risk score in the training dataset was used as the cut-off value to classify patients into high and low risk groups,and survival curves and receiver operating characteristic(ROC)curves were plotted to assess the prognostic value of the model in clinical patients.A comprehensive assessment of its efficacy was performed using nomogram,internal calibration curve,consistency index,and decision curve analysis.Patients’ correlations with clinical parameters were analyzed using chi-square tests.Gene waterfall plots were used to compare the differences in mutation frequencies between high/low HRLS groups.The maftools algorithm was used to calculate the mutation frequency and tumor mutational burden(TMB)and compare the relationship between them and HRLS scores.Microsatellite instability(MSI)scores were inferred by the Pre MSIm algorithm.Afterwards,the single-sample gene-set enrichment analysis algorithm was used to estimate the fraction of immune cell types and to sequentially validate the relationship between HRLS and chemotherapy drug responsiveness,immunotherapy and crucial gene expression patterns.Results: The HRLS model contains eight core biomarkers,namely SNHG3,NRAV,AC073611.1,AL031985.3,AL049840.6,ZFPM2-AS1,AC074117.1,and MAFG-DT.Survival curves showed that patients in the low-risk group had a good prognosis,while patients in the high-risk group had a lower survival rate and a poor prognosis(P < 0.001).The area under the ROC curve at 1,3,and 5 years was 0.767,0.710,and 0.718,respectively.The model was validated internally and externally on the test dataset and the GSE76427 dataset,respectively,and was consistent with the results of the training dataset.Multivariate Cox regression analysis also confirmed that tumor stage(HR = 2.304,P <0.001)and HRLS(HR = 4.164,P < 0.001)remained independent risk factors for overall survival after adjusting for clinical factors.The nomogram,calibration curve,concordance index and decision curve all indicated stable and strong predictive efficacy of the model.The high HRLS group had a relatively high mutation frequency and a significantly higher TMB scores.The MSI-L/MSS group had a higher HRLS scores than the MSI-H group.Patients with high HRLS had abundant macrophage infiltration and higher expression levels of immune checkpoints(PDCD1,CTLA4 and CD8A).However,patients with high HRLS had better immune checkpoint blockade(ICB)efficacy and sorafenib response rates.The high HRLS group was involved in multiple oncogenic-related pathways,while immune-related pathways were mainly enriched in the low HRLS group.In the GSE155505 dataset,the expression of AC073611.1 and AL031985.3 was found to be significantly decreased in Hepatoma Hep3 B cell line(Hep3B)after 48 hours of hypoxic exposure.In addition,the expression of MAFG-DT was significantly upregulated in sorafenib responders in the GSE109211 dataset,and this abnormal expression may be related to hypoxic status.A ceRNA network based on 3lncRNA-8mi RNA-21 m RNA interactions was predicted to have competitive binding sites.Single-cell RNA-sequencing(sc RNA-seq)analysis revealed the distribution of TNFAIP3 and VEGFA in the tumor microenvironment of HCC.Conclusion: In this study,a novel HRLS model based on hypoxia-related lncRNAs was developed and validated to accurately predict the prognostic survival of HCC patients with stable predictive performance.The risk score model was significantly correlated with multiple clinical parameters of HCC and was also able to assess immunological features and infer drug sensitivity and immunotherapy responsiveness.Hypoxia-related lncRNAs may play a role in HCC progression through genes in the ceRNA regulatory network.Hypoxia-related biomarkers may become important targets for HCC treatment and provide new perspectives for designing personalized therapies.