Cuproptosis predicts the risk and clinical outcomes of lung adenocarcinoma

Abstract

Copper is an essential microelement for the body and a necessary coregulator for enzymatic reactions, yet an unbalanced copper level promotes reactive oxidation and cytotoxicity, which ultimately induces cell death. Several small molecules targeting copper-induced cell death have been investigated, yet few showed promising therapeutic effects in clinical trials. In March 2022, Science first introduced the concept and mechanisms of cuproptosis, suggesting that copper-induced cell death targets the tricarboxylic acid (TCA) cycle via protein lipoylation. Does this novel form of cell death take part in tumorigenesis or tumor progression? Is cuproptosis related to clinical outcomes of diseases? Is there a cuproptosis-related panel for clinical practice in cancer treatment? Herein, based on 942 samples of lung adenocarcinoma (LUAD), we analyzed on gene set level the existence and predictive value of cuproptosis in disease diagnosis and treatment. We screened out and identified the "cupLA" panel which indicates the risk of LUAD occurrence, clinicopathological features of LUAD patients, and could guide clinicians to refine LUAD subtypes and make treatment choices.

Keywords: Copper-induced cell death; chemotherapy sensitivity; clinical outcomes; cuproptosis; immunotherapy sensitivity; lung cancer; tumor microenvironment.

Figure 1

Genomics and transcriptomics-based bioinformatic analysis on cuproptosis-related genes in lung adenocarcinoma. (A) CNV frequencies showed the loss and gain of ten cuproptosis biomarkers. (B) TMB of cuproptosis biomarkers and the proportion of base-pair alterations in LUAD samples. (For exmale, C>T: cytidine being replaced by adenosine, and so on in a similar fashion.) (C) ROC curves showed overall survival of single cuproptosis-related gene expression in LUAD patients. (D) Association between cuproptosis biomarkers expression and LUAD risks. (For example, DLD was up-regulated in LUAD samples; DLD expression was a risk factor for LUAD patients; and DLD expression was positively correlated with LIPT1, LIAS, FDX1, PDHB, PDHA1, and DLAT).

Figure 2

Preliminary screening of cuproptosis-related genes indicative of LUAD survival. (A) Kaplan-Meier curves for OS of two LUAD subgroups automatically clustered by transcriptomic profiles. (B) Enrichment analysis showing characteristics of two LUAD subgroups. Cluster B with poorer five year-survival are enriched in pathways including DNA repair, MYC targets, MTORC1 signaling, E2F targets, and G2M checkpoints. (C) Expression of six cuproptosis biomarkers related to poor overall survival and their relationships with clinicopathological features. (D) GO enrichment analysis of the six-cuproptosis marker genes.

Figure 3

The establishment of cupLA panel and the risk assessment potency. (A) Consensus matrix of two LUAD clusters based on cuproptosis-related gene set demonstrated in Figure 2 . Based on the sis-gene model, two new LUAD clusters (C1 and C2) were identified. (B) Kaplan-Meier curves for OS of two C1 and C2. (C) Expression of cuproptosis biomarkers in two LUAD clusters. (D) Heatmap showing clinicopathological features of patients from the high and low-risk groups and the correspondence with the cuproptosis-based gene set. (E) Expression of cuproptosis biomarkers in the high and low-risk group (*, p<0.05; **, p<0.01' ***, p<0.001).

Figure 4

Cuproptosis-related gene model predicts the clinical outcomes of LUAD patients. (A) Kaplan-Meier curves of LUAD subgroups divided by cuproptosis-related gene model. (B, C) Cuproptosis-related gene model predictive of poor overall survival among patients older than 65 years and smoking history. (D) Nomogram showing the efficacies of cuproptosis-related gene model for predicting the RFS of 1, 3, and 5 years in LUAD patients. (E) Calibration curves of the nomogram for predicting 1-, 3-, and 5-year OS in LUAD samples. (F, G) A risk prediction model based on the cuproptosis-related gene set.

Figure 5

Inferring the TME of LUAD samples with different cuproptosis statuses. (A) Correlation between cuproptosis gene model and immune cell infiltration inferred by different bioinformatic tools. (XCELL, TIMER, QUANTISEQ, MCPCOUNTER, EPIC, CIBERSORT-ABS, and CIBERSORT.) (B) Correlation between the immune cell infiltration and the differentially expressed cuproptosis-related gene model. (For example, one of the cuproptosis biomarker CDKN2A has a positive correlation with CD8+T cell infiltration, CD4+T cell memory and M1 subtype macrophages.) (C) Correlation between cuproptosis-based risk assessment and TME components. (D) Estimation of TME components and enriched pathways in the high and low-risk groups determined by cuproptosis-related gene model (*, p<0.05; **, p<0.01' ***, p<0.001.

Figure 6

The association between cuproptosis status and sensitivities to immunotherapy and chemotherapy. (A, B) Mutations of tumor-associated genes in the high- (A) and low-risk (B) groups. (C) TMB of the high- and low-risk LUAD group. (D) Kaplan-Meier curves show the overall survival of patients with different TMB statuses and cuproptosis-based LUAD risks. (E) Expression of immune checkpoints in the high and low-risk groups. (F) Sensitivities to cisplatin, gemcitabine, and paclitaxel of the high- and low-risk group identified by cuproptosis statuses (*, p<0.05; **, p<0.01' ***, p<0.001.