Meta-analysis of microarray data to determine gene indicators involved in cisplatin resistance in non-small cell lung cancer

Abstract

Background: Lung cancer is a major cause of cancer-related mortality worldwide, with a 5-year survival rate of approximately 22%. Cisplatin is one of the standard first-line chemotherapeutic agents for non-small cell lung cancer (NSCLC), but its efficacy is often limited by the development of resistance. Despite extensive research on the molecular mechanisms of chemoresistance, the underlying causes remain elusive and complex.

Aims: We analyzed three microarray datasets to find the gene signature and key pathways related to cisplatin resistance in NSCLC.

Methods and results: We compared the gene expression of sensitive and resistant NSCLC cell lines treated with cisplatin. We found 274 DEGs, including 111 upregulated and 163 downregulated genes, in the resistant group. Gene set enrichment analysis showed the potential roles of several DEGs, such as TUBB2B, MAPK7, TUBAL3, MAP2K5, SMUG1, NTHL1, PARP3, NTRK1, G6PD, PDK1, HEY1, YTHDF2, CD274, and MAGEA1, in cisplatin resistance. Functional analysis revealed the involvement of pathways, such as gap junction, base excision repair, central carbon metabolism, and Notch signaling in the resistant cell lines.

Conclusion: We identified several molecular factors that contribute to cisplatin resistance in NSCLC cell lines, involving genes and pathways that regulate gap junction communication, DNA damage repair, ROS balance, EMT induction, and stemness maintenance. These genes and pathways could be targets for future studies to overcome cisplatin resistance in NSCLC.

Keywords: BER; EMT; PARP3; cisplatin resistance; microarray; non-small cell lung cancer.

FIGURE 1

The principal component analysis (PCA) and volcano plot. (A) The PCA plot was conducted based on the normalized integrated expression data. Clustering the samples into two distinct groups using this unsupervised approach demonstrates the removal of unwanted variations among datasets. (B) The volcano plot for differentially expressed genes (DEGs) in cisplatin‐resistant NSCLC cell lines versus sensitive ones shows the fold‐change (x‐axis) versus the significance (−log10 (p.value) on the y‐axis) of the identified DEGs. The significant DEGs were identified based on p.value <.01 and |logFC| > 2. Two vertical lines show the 2‐fold change boundaries and the horizontal line shows the cutoff of statistical significance (p.value $\le$ .01). Pink and blue dots display upregulated and downregulated genes, respectively.

FIGURE 2

Heatmap and KEGG pathway analysis of differentially expressed genes. (A) The heatmap provides a visual representation of the varying levels of gene expression in cisplatin‐sensitive and ‐resistant NSCLC cell lines. High expression is denoted by the color red, while low expression is indicated by blue (S represents sensitive; R represents resistance). (B) The bar plot presents the 20 most representative altered canonical KEGG pathways that are affected by the differentially expressed genes (DEGs) associated with cisplatin resistance in NSCLC. Each term's statistical significance was evaluated using a p.value <.05, with the most significantly enriched pathways depicted in a progressively redder color.

FIGURE 3

Gene Ontology (GO) of DEGs. The gene ontology of differentially expressed genes (DEGs) reveals the top 20 terms in three categories: biological process (BP), molecular function (MF), and cellular component (CC). The color of the bubbles ranges from red to blue, indicating a progression from smaller to larger p.value. The size of the bubbles corresponds to the number of genes, with larger bubbles representing a greater number of genes.

FIGURE 4

Protein–protein interaction network of DEGs. This figure illustrates the network of differentially expressed genes (DEGs) with potential roles in cisplatin resistance in NSCLC cells, constructed using the STRING database (version 12.0). The nodes represent proteins, and the edges denote interactions. The red nodes represent proteins encoded by seven DEGs (TUBB2B, MAPK7, TUBA3D, TUBB4A, TUBAL3, ADCY1, and MAP2K5) selected for their roles in the gap junction pathway. This pathway was the only one significantly over‐represented among the proteins in the network. The module reveals the connection between these DEGs and connexin43 (GJA1 gene), a key effector in the gap junction pathway. Notably, GJA1 was not identified among the DEGs contributing to cisplatin‐resistant NSCLC.

FIGURE 5

Receiver operating characteristic (ROC) analysis for DEGs. ROC plots illustrate the predictive accuracy of 14 DEGs in relation to the occurrence of cisplatin resistance in NSCLC cell lines. AUC refers to the area under the ROC curve. (A–N) are ROC plots of G6PD, MAGEA1, MAPK7, PDK1, MAP2K5, NTRK1, PARP3, NTHL1, HEY1, SMUG1, TUBB2B, TUBB4A, TUBAL3 and ADCY1 genes, respectively.