Identification of Prognostic Genes in the Tumor Microenvironment of Hepatocellular Carcinoma

Abstract

Background: Hepatocellular carcinoma (HCC) is one of the most common malignant tumors in the world. The efficacy of immunotherapy usually depends on the interaction of immunomodulation in the tumor microenvironment (TME). This study aimed to explore the potential stromal-immune score-based prognostic genes related to immunotherapy in HCC through bioinformatics analysis.Methods: ESTIMATE algorithm was applied to calculate the immune/stromal/Estimate scores and tumor purity of HCC using the Cancer Genome Atlas (TCGA) transcriptome data. Functional enrichment analysis of differentially expressed genes (DEGs) was analyzed by the Database for Annotation, Visualization, and Integrated Discovery database (DAVID). Univariate and multivariate Cox regression analysis and least absolute shrinkage and selection operator (LASSO) regression analysis were performed for prognostic gene screening. The expression and prognostic value of these genes were further verified by KM-plotter database and the Human Protein Atlas (HPA) database. The correlation of the selected genes and the immune cell infiltration were analyzed by single sample gene set enrichment analysis (ssGSEA) algorithm and Tumor Immune Estimation Resource (TIMER).Results: Data analysis revealed that higher immune/stromal/Estimate scores were significantly associated with better survival benefits in HCC within 7 years, while the tumor purity showed a reverse trend. DEGs based on both immune and stromal scores primarily affected the cytokine–cytokine receptor interaction signaling pathway. Among the DEGs, three genes (CASKIN1, EMR3, and GBP5) were found most significantly associated with survival. Moreover, the expression levels of CASKIN1, EMR3, and GBP5 genes were significantly correlated with immune/stromal/Estimate scores or tumor purity and multiple immune cell infiltration. Among them, GBP5 genes were highly related to immune infiltration.Conclusion: This study identified three key genes which were related to the TME and had prognostic significance in HCC, which may be promising markers for predicting immunotherapy outcomes.

Identification of Novel Tumor Microenvironment-Related Long Noncoding RNAs to Determine the Prognosis and Response to Immunotherapy of Hepatocellular Carcinoma Patients

“…HCC is an inflammation-associated malignancy that comprise numerous immune cell subtypes (i.e., tumor-associated macrophages, tumor-associated neutrophils, and myeloidderived suppressor cells), forming a complex immune tolerance microenvironment and contributing to the development of HCC (Nishida and Kudo, 2017;Kurebayashi et al, 2018). In recent years, with the wide application of bioinformatics, increasing numbers of research works have applied statistical algorithms to investigate the characteristics of the TME and explore new targets for immunotherapy (Xiang et al, 2021). The study by Li et al reported that the TME of HCC could be classified into different molecular subtypes based on the immune gene expression profiles, which exhibited different immunogenetic features and survival outcomes (Li et al, 2019a).…”

Guanylate-Binding Protein 1 as a Potential Predictor of Immunotherapy: A Pan-Cancer Analysis

“…GBP4 and other eight differentially expressed genes constituted an immune-relevant gene signature for predicting the prognosis of patients with muscle-invasive bladder cancer (MIBC) ( Jiang et al, 2020 ). GBP5 was identified as a prognostic gene in the TME of hepatocellular carcinoma and gastrointestinal stromal tumors ( Blakely et al, 2018 ; Xiang et al, 2021 ). Low GBP6 expression was correlated with poor cell differentiation and lymph node metastasis in tongue squamous cell carcinoma (TSCC), and low GBP7 expression was linked with short OS in HNSC patients ( Liu et al, 2020 ; Wu et al, 2020 ).…”

Identifying a Novel Endoplasmic Reticulum-Related Prognostic Model for Hepatocellular Carcinomas

“…Various genes may play roles in the pathogenesis of human diseases, including cancers, by affecting the normal physiologi-cal functions of specific organelles and maintaining cellular homeostasis [4,5]. By utilizing different machine learning approaches, a series of survival prediction models targeting specific biological events were developed [6][7][8][9][10][11]. For instance, Lasso (least absolute shrinkage and selection operator) regression was applied to build several prognostic models for HCC patients targeting ferroptosis [8], reactive oxygen species [9], amino acid metabolism [11], or the tumour microenvironment [7].…”

“…By utilizing different machine learning approaches, a series of survival prediction models targeting specific biological events were developed [6][7][8][9][10][11]. For instance, Lasso (least absolute shrinkage and selection operator) regression was applied to build several prognostic models for HCC patients targeting ferroptosis [8], reactive oxygen species [9], amino acid metabolism [11], or the tumour microenvironment [7]. In this study, we first constructed an HCC prognostic model from the perspective of ER using an integrated modelling strategy (differentially expressed genes, univariate/multivariate Cox regression, Lasso regression, and nomogram prediction).…”