Immune infiltration landscape on prognosis and therapeutic response and relevant epigenetic and transcriptomic mechanisms in lung adenocarcinoma

OBJECTIVE: Lung adenocarcinoma (LUAD) is the most prevalent lung cancer subtype, but its immune infiltration features are not comprehensively understood. To address the issue, the present study was initiated to describe the immune infiltrations across LUAD from cellular compositional, functional, and mechanism perspectives. METHODS: We adopted five LUAD datasets (GSE32863, GSE43458, GSE75037, TCGA-LUAD, and GSE72094). Differentially expressed genes between LUAD and controls were selected for co-expression network analysis. Risky immune cell types were determined for classifying LUAD patients as diverse subtypes, followed by a comparison of antitumor immunity and therapeutic response between subtypes. Then, LUAD- and subtype-related key module genes affected by DNA methylation were determined for quantifying a scoring scheme. EXO1 was chosen for functional analysis via in vitro assays. RESULTS: Two immune cell infiltration-based subtypes (C1 and C2) were established across LUAD, with poorer prognostic outcomes and lower infiltration of immune cell types in C1. Additionally, C1 presented higher responses to immune checkpoint blockade and targeted agents (JNK inhibitor VIII, BI-D1870, RO-3306, etc.). The scoring system (comprising GAPDH, EXO1, FYN, CFTR, and KLF4) possessed higher accuracy in estimating patients’ prognostic outcomes. EXO1 upregulation contributed to the growth, migration, and invasion of LUAD cells. In addition, EXO1 facilitated PD-L1 and sPD-L1 expression in LUAD cells. CONCLUSION: Altogether, our findings offer a comprehensive understanding of the immune infiltration landscape on prognosis and therapeutic response of LUAD as well as unveil potential epigenetic and transcriptomic mechanisms, which might assist personalized treatment.