Identification Of A Peroxisome-related Gene Signature For Predicting Prognosis In Hepatocellular Carcinoma And Molecular Mechanisms Of HAO2-induced Cell Cycle Arrest In Human Hepatoma Cells

Background: Hepatocellular carcinoma(HCC)is the sixth most common tumor type and the fourth leading cause of cancer-related death in the world.Much progress has been made in the clinical diagnosis and treatment of HCC,but the overall prognosis of HCC patients still needs to be improved.Prognostic assessment is very important for monitoring followup and treatment strategy selection of HCC patients.In recent years,the popularity of second-generation sequencing technology has greatly enriched our understanding of cancer biology.Through bioinformatics analysis of genomics data,combined with the clinicopathological characteristics of HCC patients,the identification and development of novel and reliable prognostic prediction models have broad application prospects,and may also provide a reference for the screening of tumor therapeutic targets.Abnormal metabolism is one of the hallmark features of many malignant tumors including liver cancer.Peroxisomes are monolayer membrane organelles that contain numerous metabolic enzymes and are directly involved in multiple metabolic pathways.In HCC and many other tumor types,a variety of peroxisomal enzymes and their metabolic activities have been reported to change,and the corresponding enzyme inhibitors or alteration in gene expression levels can inhibit or promote tumor growth,indicating that peroxide Enzymes are involved in the development and progress of various tumors.However,the current research on peroxisomes in hepatocellular carcinoma mainly focuses on certain peroxisomal genes,and few studies have explored the prognostic value of peroxisomes as a whole in hepatocellular carcinoma at the subcellular level.Hydroxyacid oxidase HAO2 is an enzyme in the peroxisomes that catalyzes the oxidation of hydroxyl fatty acids to keto acids and hydrogen peroxide.In our earlier analysis of transcriptomic data from HCC and adjacent tissues using public databases,we found that HAO2 expression in human HCC tissues decreased significantly,and the low expression of HAO2 was associated with a worse prognosis of liver cancer.At present,there are very limited research about HAO2.Some studies have found that the expression level of HAO2 in drug-induced HCC animal models and human liver cancer tissue samples is significantly reduced,and overexpression of HAO2 can inhibit the growth of transplanted tumors,suggesting that HAO2 may play a tumor suppressing role in HCC,but the molecular mechanism involved has not been reported.In this study,we first explored the overall value of peroxisomal related genes in the prognosis prediction of HCC using bioinformatics methods.Second,focusing on the differences in peroxisomal related genes that have both prognostic significance and differential expression alteration.HAO2 was the survival-related gene with the most significant change in expression,and its possible mechanism on inhibiting HCC was explored.Methods:1.We searched HCC related genomics data set in the public databases,included 2 HCC data sets(n = 547),and defined the peroxisomal gene set.Using R and Bioconductor packages,we pre-processed and carried out subsequent analysis of transcriptomics data and clinical information in the 2 data sets.First,we screened out peroxisomal genes with prognostic significance and analyzed differential expression of peroxisomal genes related to prognosis.In the TCGA-LIHC data set,LASSO algorithm was used to select genes from prognosis-related genes and established a prognostic model.We evaluated the predictive power and independence of the genetic model,and validated it with external data set.We compared our model with other HCC gene prognostic models,and combined clinical features to optimize the prognostic model.2.First,expression change of HAO2 in the large sample HCC data set and the relationship with the prognosis of HCC were systematically analyzed.Then,we used clinical samples to verify the expression change of HAO2.GSEA analysis was used to explore the possible KEGG pathway related to HAO2.The stable hepatoma cell lines with overexpression of HAO2 were constructed by lentivirus infection.Western blot and qPCR verified the HAO2 overexpression.CCK8 and flow cytometry were used to detect the effect of overexpressing HAO2 on the proliferation and cell cycle of hepatoma cells.Transcriptome of mRNA from HAO2 overexpressed cells suggested the major molecules that may be affected downstream and possible mechanisms were explored.Results:1.This study included 2 HCC data sets,a total of 547 patients with liver cancer.There are 351 cases in the training set TCGA-LIHC and 196 cases in the validation set LIRI-JP.A peroxisomal gene set containing 113 genes was defined.The prognosis analysis found that 47 peroxisomal genes were related to the prognosis of HCC patients.Some genes also showed differential expression between liver cancer and adjacent tissues,most of which were in downregulated.HAO2 was the prognostic gene with maximum fold change.10 optimal prognostic genes(ACAT1,AGPS,ATAD1,LDHA,MTARC2,PECR,ASCL6,MPV17,PRDX1,TRIM37)were selected from 47 genes using LASSO algorithm.A new gene signature based on these 10 peroxidases was established.Survival analysis,ROC curve,Cox analysis and external data set validation showed that the genetic model exhibited independent and robust predictive ability.Compared with other six genetic models,the model’s predictive ability has cross-platform stability,and finally combined with TNM staging optimized the gene signature.GSEA analysis showed that the gene model was closely related to the metabolic functions of peroxisomes.2.The expression of HAO2 in two independent liver cancer data sets(a total of 860 liver cancer and adjacent tissue samples)was significantly reduced in liver cancer tissues,and the low expression of HAO2 was associated with poor prognosis of liver cancer patients.Immunohistochemical staining of liver cancer patients confirmed that HAO2 expression decreased in liver cancer tissues.Single-gene GSEA analysis suggested that the expression of HAO2 was closely related to the cell cycle and indicated that the high expression of HAO2 might be related to cell cycle inhibition.Hepatoma cell lines with stable overexpression of HAO2 were successfully constructed using lentivirus.CCK8 detection found that HAO2 overexpression inhibited the proliferation of liver cancer cells.Flow cytometry detection found that HAO2 overexpression can induce G1 phase arrest of the cell cycle,upregulate p21 expression,and inhibit CDK2,CDK6 expression.Transcriptome sequencing data analysis and verification experiments show that HAO2 can inhibit the expression of ID1 in hepatoma cells,and that overexpression of ID1 can rescue the upregulation of p21 by HAO2,suggesting that HAO2 may induce cell cycle arrest through the ID1-p21 pathway,thereby inhibiting liver cancer progression.Conclusions:1.We identified a prognostic model for HCC based on 10 peroxisomal-related genes from the public HCC data sets.2.HAO2 is significantly decreased in liver cancer tissues compared with adjacent tissues.HAO2 may induce the cycle arrest in hepatoma cells through the ID1-p21 pathway,thereby inhibiting the progression of liver cancer.