Exploration and validation of a novel signature of seven necroptosis-related genes to improve the clinical outcome of hepatocellular carcinoma

Abstract

Necroptosis has been reported to be involved in cancer progression and associated with cancer prognosis. However, the prognostic values of necroptosis-related genes (NRGs) in hepatocellular carcinoma (HCC) remain largely unknown. This study aimed to build a signature on the basis of NRGs to evaluate the prognosis of HCC patients. In this study, using bioinformatic analyses of transcriptome sequencing data of HCC (n = 370) from The Cancer Genome Atlas (TCGA) database, 63 differentially expressed NRGs between HCC and adjacent normal tissues were determined. 24 differentially expressed NRGs were found to be related with overall survival (OS). Seven optimum NRGs, determined using Lasso regression and multivariate Cox regression analysis, were used to construct a new prognostic risk signature for predicting the prognosis of HCC patients. Then survival status scatter plots and survival curves demonstrated that the prognosis of patients with high-Riskscore was worse. The prognostic value of this 7-NRG signature was validated by the International Cancer Genome Consortium (ICGC) cohort and a local cohort (Wenzhou, China). Notably, Riskscore was defined as an independent risk factor for HCC prognosis using multivariate cox regression analysis. Immune infiltration analysis suggested that higher macrophage infiltration was found in patients in the high-risk group. Finally, enhanced 7 NRGs were found in HCC tissues by immunohistochemistry. In conclusion, a novel 7-NRG prognostic risk signature is generated, which contributes to the prediction in the prognosis of HCC patients for the clinicians.

Keywords: Immune infiltration; Macrophages; Necroptosis; Prognostic signature; USP21..

Fig. 1

The general workflow of this study

Fig. 2

Screening prognostic NRGs and seven NRGs signature construction: (A) Heatmap of DENRGs between HCC tissues samples and adjacent normal samples. (B) Topology soft-thresholding analysis of network. (C) Clustering dendrogram of DENRGs. (D) Module-trait relationships. (E) Univariate cox regression analysis identified 24 prognostic DENRGs. (F) LASSO analysis for 24DENRGs. (G) Forest plot for multivariate cox regression analysis

Fig. 3

Accuracy validation of seven NRGs signature (A) Riskscores in TCGA cohort. (B) Riskscores in ICGC cohort. (C) Survival status scatterplot of patients in TCGA cohort. (D) Survival status scatterplot of patients in ICGC cohort. (E) Analysis of PCA for TCGA cohort. (F) Analysis of PCA for ICGC cohort. (G) Survival curves of patients in TCGA cohort. (H) Survival curves of patients in ICGC cohort. (I) ROC curve analysis for the value of prognosis in the signature for TCGA cohort. (J) ROC curve analysis for the value of prognosis in the signature for ICGC cohort

Fig. 4

Independent prognostic analysis, nomogram construction and calibration curves (A) Univariate cox regression analyses. (B) Multivariate cox regression analyses. (C) Nomogram based on prognostic factors. (D-F) Calibration curve of the nomogram for the 1st, 2nd and 3rd year

Fig. 5

Correlation analyses between Riskscore and clinical features (A) Association between Riskscore and grade. (B) Association between Riskscore and stage. (C) Kaplan-Meier survival curves of G1-2 patients. (D) Kaplan-Meier survival curves of G3-4 patients. (E) Kaplan-Meier survival curves of patients with stage I-II. (F) Kaplan-Meier survival curves of patients with III-IV. (G) Correlation between grade and FAF1 expression. (H) Correlation between stage and FAF1 expression. (I) Correlation between grade and HSP90AB1expression. (J) Correlation between stage and HSP90AB1expression. (K) Correlation between grade and PPIA expression. (L) Association between stage and PPIA expression. (M) Correlation between grade and HSP90AA1 expression. (N) Correlation between grade and PGAM5 expression

Fig. 6

Immune infiltration analysis (A) GSEA enrichment analyses of high-Riskscore group. (B) GSEA enrichment analyses of low-Riskscore group. (C) The barplot of signature gene GO enrichment analysis. (D) Differential analyses of immune-related functions. Adjusted P values were demonstrated as: ns, namely, not significant; *P < 0.05; **P < 0.01; ***P < 0.001. (E) Survival curves of patients with different macrophages scores. (F) Relative infiltration of M0, M1 and M2 macrophages between high- and low-risk subgroups in TCGA cohort. (G) Relative levels of PD-1, PD-L1 and CTLA4 between high- and low-risk subgroups in TCGA cohort. (H) Tumor mutation burden between high- and low-risk subgroups in TCGA cohort

Fig. 7

Immunohistochemistry of 7 NRGs in HCC patients 7 NRGs were enhanced in HCC tissues (n = 20) compared with adjacent tumor tissue samples (n = 20)

Fig. 8

TF network construction (A) Heatmap of differentially expressed TFs. (B) Volcano plot of differentially expressed TFs. (C) TF network. Circular nodes represent NRGs, triangular nodes represent TFs. The line between nodes indicates a regulatory relationship between two nodes. (D) Co-expression relationship between NRF1 and USP21. (E) NRF1 expression. (F) USP21 expression. (G) Survival curves of patients with high- and low-NRF1. (H) Survival curves of patients with high- and low-USP21

Fig. 9

NRF1 enhances Huh7 proliferation and migration through upregulating USP21 expression (A) NRF1 mRNA expression in Huh7 cells after si-NRF1 transfection. (B) USP21 mRNA expression in Huh7 cells after si-NRF1 transfection. (C) USP21 protein level. (D) Effect of NRF1 knockdown on cell proliferation. (E) USP21 mRNA expression in Huh7 cells after si-USP21 transfection. (F) Effect of USP21 knockdown on cell proliferation. (G) Transwell assay, magnification: ×200. (H) Levels of MMP2, MMP9, CDH1, VIM and CCL2 in Huh7 cells after si-USP21 transfection. (I) Drug sensitivity test. ***p < 0.001, ****p < 0.0001, ns. no signification