
Clinical and Translational Medicine, Journal Year: 2023, Volume and Issue: 13(2)
Published: Feb. 1, 2023
Growing evidence has demonstrated that pseudogenes are intricately associated with tumour progression in hepatocellular carcinoma (HCC).1 Nonetheless, related to immune infiltration and their value improving clinical outcomes remain largely unexplored. Pseudogenes, newly discovered non-coding homologs of protein-coding genes,1 have been regarded as non-functional evolutional relics. Nevertheless, increasing play vital roles tumourigenesis regulators coding genes. For example, HMGA1P6, a pseudogene transcriptionally activated by oncogene MYC, could contribute oncogenesis ovarian cancer.2 Besides, also profound impacts on anti-tumour responses via involvement regulating tumour-immune interactions. BRCA1 Pseudogene 1 (BRCA1P1) reported weaken response suppressing innate defense mechanisms.3 Through glioma-derived exosomes, TMEM198B promoted macrophage lipid accumulation increased fatty acid oxidation, further inducing macrophages M2 polarization.4 However, the significance immune-related remains unexplored HCC. Here, we enrolled four public datasets abundant expression profiles complete prognostic information, including: The Cancer Genome Atlas Liver Hepatocellular Carcinoma (TCGA-LIHC) (n = 340), GSE116174 64), GSE144269 67) GSE14520 242) (Table S1). To systematically evaluate candidate intrinsic modulators cells, introduced novel framework identify housekeeping (Supplementary Materials). In TCGA-LIHC, relative abundance 28 cells was first deciphered using ssGSEA algorithm.5 Considering purity may obscure links microenvironment,6 thus calculated first-order partial correlation coefficient (PCC) between removing effect purity.6, 7 Pseudogenes top 5% PCC were extracted for each cell. A hypothesis is if specific strong correlations all cell types, it execute role HCC microenvironment,8 which defined (HIRIP) this study. tissue specificity index (TSI)9 generally correlated different types S2). As previously reported,9 lower TSI score strongly suggesting biological functions immunity. According criteria described previous studies,8, 9 settled threshold <.2 identified 23 HIRIP essential regulation (Figure S1A). Subsequently, filtered develop an integrative signature (HIRIPS). Initially, univariate Cox regression analysis determined potential, including HNRNPA3P5, HNRNPA3P6, PTMAP5 EIF2S2P4 S1B). Kaplan–Meier high these suggested unfavorable prognosis 1A). Afterward, optimal machine-learning algorithm assessing prognosis, developed 22 survival models Materials) based TCGA-LIHC. mentioned above, final HIRIPS fitted Akritas highest average mean C-index (.667) integrated areas under curve (iAUC) (.825) TCGA training cohort three validation cohorts (GSE116174, GSE14520) considered one 1B). validate robust performance HIRIPS, Kaplan-Meier analyses applied. Notably, patients possessed adverse TCGA-LIHC (HR 1.713; log-rank test: P < .001), similar tracks observed datasets, 1.136; 1.120; 1.118; .001) 2A). time-dependent receiver operating characteristic (ROC) curves OS (1-, 2-, 3-year) exhibited excellent discrimination 2B) (AUC values: .728, .690, .699), .714, .751, .700), .703, .704, .714), .691, .694, .702). Multivariate performed remained statistical after adjusting confounding factors across 2C). Overall, presented clear superior predicting prognosis. explore mechanisms underlying proposed pipeline perform functional enrichment, maximized information retention comprehensively gene ordering over-representing revealed predominant enrichment pathways malignant progressions, such cycle p53 signaling 3A Figure Whereas low particularly evident pathways, cytokine activity T receptor another bioinformatics algorithm, GSEA, confirmed significantly linked 3B–E). remarkable differences regarding function account discrepancy Patients featured cellular proliferation status, line Low predominantly distinguished infiltration, indicated harbored more reserves immunization resources immunotherapy. conclusion, investigated HCC-infiltrating (termed HIRIPS) from algorithms. This displayed stable might serve latent biomarker immunotherapy response. our study provided promising platform optimizing precise treatment authors declare they no competing interests. Please note: publisher not responsible content or functionality any supporting supplied authors. Any queries (other than missing content) should be directed corresponding author article.
Language: Английский