In this study, we used data from three public platforms (GEO, TCGA, ICGC). First, we made a Wayne-diagram between three GSE datasets of NSCLC (GSE19188, GSE33532, GSE44077) and cuproptosis-associated genes selected after the literature review, and identified 16 cuproptosis-related DEGs. The heat map reflects the expression of these genes in NSCLC samples and normal samples, and there are obvious differences between them. GO and KEGG enrichment analysis showed that cuproptosis-related genes in NSCLC were mainly enriched in cellular energy metabolism-related pathways. Then, according to the survival analysis of these 16 genes, we found that the high expression of 13 genes (CCDC167, PKIG, SFXN1, CCNB2, KRT6A, SLC25A4, SLC25A5, SLC25A6, SLC25A10, SLC25A11, SLC25A20, SLC25A23, SLC25A31) predicted a poor overall survival rate in patients with NSCLC. Next, we used these 13 genes to establish a prognostic model of cuproptosis-related genes in NSCLC, through which we divided patients in the TCGA database into a high-risk group and a low-risk group, and analyzed the differences in various clinical outcomes between the two groups. The results suggested that the high-risk group had higher rates of advanced tumor grade, advanced tumor stage and vascular invasion. Survival curves showed that the high-risk group also had a worse survival rate than the low-risk group. To demonstrate the reliability of this finding, we performed a validation of the same model using the dataset from the ICGC database and obtained similar results. Finally, we verified the differential expression of several genes in the model in the tissues. By detecting intracellular Cu2+ levels before and after gene knockdown, we found that four genes may affect the progression of NSCLC by regulating cuproptosis.
Copper, as an important element in the composition of human cells, plays an important role in many areas such as cell composition, cell metabolism, and immune response. However, excess intracellular copper may also have toxic effects on cells, and evidence from yeast and mammalian cells (27, 28) suggests that copper toxicity is associated with altered mitochondrial function, but the specific role of copper in the process of cell death has not been fully elucidated. Recently, Tsvetkov (18) et al found that excess copper selectively disrupts a group of metabolic enzymes with a distinctive posttranslational modification, known as lipoylation. Lipoylation involves the attachment of the small sulfur-containing metabolite lipoic acid to a substrate protein. This flexible, chemically active appendage achieves catalysis by oscillating between different subunits within an enzyme complex, donating or accepting electrons. This process is specifically manifested by the aggregation of lipoylated mitochondrial enzymes, which leads to cell death (18). Furthermore, this form of cell death differs from the previously found cell necrosis, ferroptosis (12) and zinc-induced cell death (29).
Currently, only a few studies (18, 24, 25) have discussed the mechanism of cuproptosis. To summarize, cuproptosis-induced cell death does not involve a profound disruption of the electron transport chain (ETC). Copper may induce acute lethal cellular shock by inhibiting a variety of lipoylated enzymes while altering the concentration of many metabolites (24). Alternatively, the enzymes bound to copper may acquire a new catalytic activity, leading to the production of toxic metabolites that cannot be processed. In addition, protein aggregation may kill cells by disrupting the function of other mitochondrial enzymes, such as those involved in the synthesis of iron-sulfur clusters, which are important cofactors for other enzymes. In summary, excess copper promotes the aggregation of lipid-acylated proteins and the destabilization of Fe-S cluster proteins, ultimately leading to cell death.
In the 13 genes prognostic model, these genes can be classified into 8 mitochondrial vectors family genes (SLC25A4, SLC25A5, SLC25A6, SLC25A10, SLC25A11, SLC25A20, SLC25A23, SLC25A31) and 5 other genes (CCNB2, KRT6A, CCDC167, PKIG, SRXN1). The GO and KEGG enrichment analyses resulted in these genes were mainly enriched in the pathways related to cellular energy metabolism, which is more consistent with the mechanism related to cuproptosis. Cell cycle protein B2 (CCNB2), a member of the cell cycle protein family (30), mainly regulates the cell proliferation (31) process and is involved in the proliferation of NSCLC, and is considered to be a pivotal gene involved in the carcinogenesis and development of NSCLC (32). Meanwhile, high expression of CCNB2 is significantly and positively correlated with tumor size, lymph node metastasis, distant metastasis and clinical stage, and is an independent adverse prognostic factor for overall survival of NSCLC patients (33). CCDC167 is differentially upregulated in tumor tissues such as lung cancer and breast cancer, and is thought to be associated with poor prognosis of breast cancer patients (34). Keratin 6A (KRT6A) belongs to the keratin family (35), an important component of the cytoskeleton in mammalian cells, which functions downstream of LSD1 and upregulates G6PD through the MYC signaling pathway as a key driver of NSCLC progression. It has also been found that KET6A may be a biomarker in patients with smoking NSCLC (36). PKIG is thought to be associated with the regulation of osteoblast and adipocyte differentiation (37, 38), but there are no studies related to its relationship with NSCLC. SRXN1 is considered as a potential prognostic assessment for NSCLC, and previously constructed predictive models (36, 39) found that high expression of SRXN1 may be associated with poor prognosis in patients with lung adenocarcinoma correlation.
This study also has some limitations. First, cuproptosis is a recently proposed form of RCD, and there is a relative lack of basic and clinical studies on this area, and its associated mechanisms may lack confirmation from a large number of experiments. Second, the biological characteristics of NSCLC are heterogeneous, and the conclusions drawn from different studies are highly variable, although the present study reduces such differences by comprehensive analysis of multiple data sets, it still has some limitations. Third, the number of NSCLC samples available for bioinformatics analysis is limited due to the cost of gene chip technology and the difficulty of sample sourcing, and possible reporting bias may also affect the analysis results. In the future, more studies targeting cuproptosis-related genes in NSCLC are needed to continuously improve this field in order to find potential diagnostic and therapeutic targets for NSCLC.
In conclusion, this study aimed to analyze the role of cuproptosis-related genes in the pathogenesis of NSCLC, and through public data resources, we analyzed the differential expression of cuproptosis-related genes in NSCLC and their biological significance in the progression of NSCLC. In addition, we developed a new diagnostic model for NSCLC, which contains 13 cuproptosis-associated genes. This model can be used as a reference for clinical classification of NSCLC, and has value in predicting patients’ survival time. Also, this model is useful in assessing the tumor grade, tumor stage, and tumor vascular invasion of NSCLC patients.