Development and validation of hierarchical signature for precision individualized therapy based on the landscape associated with necroptosis in clear cell renal cell carcinoma

Abstract

Background: Increasing evidence is showing that necroptosis has unique clinical significance in the occurrence and development of multiple diseases. Here, we systematically evaluate the role of necroptosis in clear cell renal cell carcinoma (ccRCC) and analyze its regulatory patterns.

Methods: First, we evaluated the expression and enrichment of necroptotic factors in ccRCC using gene set enrichment analysis (GSEA) and survival analysis in the expression profile from The Cancer Genome Atlas (TCGA) to demonstrate the overall mutation of necroptotic pathway genes. Then, we used unsupervised clustering to divide the samples into two subtypes related to necroptosis with significant differences in overall survival (OS) and subsequently detected the differentially expressed genes (DEGs) between them. Based on this, we constructed the necroptosis scoring system (NSS), which also performed outstandingly in hierarchical data. Finally, we analyzed the association between NSS and clinical parameters, immune infiltration, and the efficacy of immunotherapy containing immune checkpoint inhibitors (ICIs), and we suggested potential therapeutic strategies.

Results: We screened 97 necroptosis-related genes and demonstrated that they were dysregulated in ccRCC. Using Cox analysis and least absolute shrinkage and selection operator (LASSO) regression, a prognostic prediction signature of seven genes was built. Receiver operating characteristic (ROC) curves and Kaplan-Meier (KM) analyses both showed that the model was accurate, and univariate/multivariate Cox analysis showed that as an independent prognostic factor, the higher the risk score, the poorer the survival outcome. Furthermore, the predicted scores based on the signature were observably associated with immune cell infiltration and the mutation of specific genes. In addition, the risk score could potentially predict patients' responsiveness to different chemotherapy regimens. Specifically, Nivolumab is more effective for patients with higher scores.

Conclusion: The necroptosis-related signature we constructed can accurately predict the prognosis of ccRCC patients and further provide clues for targeted, individualized therapy.

Keywords: clear cell renal cell carcinoma; necroptosis; necroptosis scoring system; precise treatment; survival analysis.

FIGURE 1

Necroptosis is involved in the development of clear cell renal cell carcinoma (ccRCC). (A) Flow chart of this study. (B) Gene set enrichment analysis (GSEA) of necroptotic factors between tumor and normal samples. (C) GSEA of necroptotic factors in N- and M-stages. (D) Kaplan–Meier curves of the most significant six genes in necroptotic factors. (E) Interactions between necroptotic factors.

FIGURE 2

Gene transcription and gene alteration of necroptosis in clear cell renal cell carcinoma (ccRCC). (A) Abundance of 24 differentially expressed genes (DEGs) between ccRCC and normal samples. (B, C) Differences in expression levels of factors between stage (B) and grade (C). (D) Mutations of necroptotic genes in The Cancer Genome Atlas (TCGA) cohort. (E) Copy number variation (CNV) of the top 30 necroptotic genes. (NS, nonsignificant; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).

FIGURE 3

Different regulation models of necroptosis in clear cell renal cell carcinoma (ccRCC). (A–C) K-elbow curve (A), sample matrix of consistent clustering analysis (B), and PCA (C) of two clusters. (D–I) Kaplan–Meier curves (D), necroptotic factor expression (E), gene set variation analysis (GSVA) (F), immune infiltration (G), immune checkpoint expression (H), and estimate scores (I) between subtypes. (J) Gene set enrichment analysis (GSEA) between differentially expressed genes (DEGs). (NS, nonsignificant; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).

FIGURE 4

Construction of necroptosis scoring system (NSS). (A, B) least absolute shrinkage and selection operator (LASSO) regression (A) and weight coefficient of genes (B) of model. (C, D) Distribution of NSS scores (C) and survival time and status of samples (D) in the training group. (E, F) Expression of model genes (E) and Kaplan–Meier curves (F) between high- and low-NSS score groups. (G) Receiver operating characteristic (ROC) curves of model. (H) Univariate and multivariate Cox analyses in training group. (I) Nomogram to predict OS for 1/3/5 years.

FIGURE 5

Necroptosis scoring system (NSS) differences between subtypes and clinicopathological features. (A) Sankey diagram containing subtypes, NSS group, and survival outcomes. (B) NSS differences between the necroptosis subtypes and differentially expressed gene (DEG) subtypes. (C) NSS differences between clinical characteristics groups. (NS, nonsignificant; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).

FIGURE 6

Molecular mechanism analysis involved in necroptosis scoring system (NSS). (A) Correlation between NSS scores and enrichment scores of hallmark pathways. (B) Gene set enrichment analysis (GSEA) of NSS groups in Gene Otology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG). (C–E) Differences of immune infiltration (C), immune checkpoint expression (D), and immune score (E) between groups. (F, G) Mutations (F), amplification and deletion (G) in low and high NSS groups. (NS, nonsignificant; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).

FIGURE 7

Necroptosis scoring system (NSS) reflects potential therapeutic strategies. (A) Five drugs were significantly associated with NSS scores. (B–E) Differences of drug resistance (B), overall survival, progression-free survival (C), immunotherapy efficacy (D, E) between high- and low-NSS groups. (F) Receiver operating characteristic (ROC) curves of NSS in the immunotherapy patient cohort. (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).

FIGURE 8

Seven genes affecting the viability of clear cell renal cell carcinoma (ccRCC) cells. (A) Immunoblotting analysis of CDH2 protein expression in four pairs of fresh tumors and normal adjacent tissues. (B) Comparison of CDH2 protein expression in ccRCC tissues and adjacent tissues via immunohistochemistry (IHC) staining. (C) Knockdown efficiencies of seven genes. (D) Cell viability after treatment with TNF-α, Smac mimetic, and z-VAD (TSZ). (E) Cell proliferation capacity of 786-O transfected with siRNAs or control vectors. (F) Proliferative abilities of 786-O cells measured by colony formation after the knockdown and overexpression of CDH2. (G) Wound healing assay assessing the migration potential of 786-O cells transfected with CDH2 siRNAs or si-NC. (NS, nonsignificant; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).