Development and validation of M2 macrophage-related genes in a prognostic model of lung adenocarcinoma based on bulk RNA and ScRNA datasets.

Abstract

Objective: This study aimed to investigate the correlation between M2 macrophages activity with the prognosis of lung adenocarcinoma (LUAD). We sought to identify key genes associated with M2 macrophage activity and examine their relationship with clinicopathological features to elucidate the underlying mechanism.

Methods: Published datases were analyzed for differentially expressed genes. After quality control, batch effect removal, and annotation, the scRNA dataset identified M2 macrophage-associated differentially expressed genes in the LUAD group, which were cross-analyzed and referred to as M2 macrophage-linked genes. A risk model was generated using machine learing for these genes. Thereafter, two bulk RNA-seq datasets were used to evaluate the model. We computed risk scores for all samples and grouped them into low and high risk, aiding in the comparison of clinical characteristics, immune and stromal infiltration, and drug sensitivity. Finally, key genes were validated through immunohistochemistry in IPA samples.

Results: We identified four key M2 macrophage-linked genes: TIMP1, CAV2, MIF, and SELENBP1. Survival durations in the high-riskscore cluster were lower across the TCGA-LUAD (P = 1.2 × 10^-4), GSE14814 (P = 0.02), and GSE37745 (P = 0.01) data sets. The stromal score, fibroblast infiltration, and cytokinesis activation were increased in the high-risk subgroup. Neutrophil and endothelial cell infiltration and activation of the linolenic acid pathway occurred in the low-risk group. IHC confirmed that CAV2 and SELENBP1 expression was significantly reduced, while TIMP1 and MIF were significantly increased in LUAD, which was consistent with the bioinformatics findings.

Conclusion: The role of M2 macrophages in tumor progression could anticipate the prognosis of LUAD and develop novel immunotherapy strategies.