Modulation of oxidative phosphorylation and mitochondrial biogenesis by cigarette smoke influence the response to immune therapy in NSCLC patients

Abstract

The treatment regimen of non-small cell lung cancer (NSCLC) has drastically changed owing to the superior anti-cancer effects generated by the immune-checkpoint blockade (ICB). However, only a subset of patients experience benefit after receiving ICBs. Therefore, it is of paramount importance to increase the response rate by elucidating the underlying molecular mechanisms and identifying novel therapeutic targets to enhance the efficacy of IBCs in non-responders. We analyzed the progression-free survival (PFS) and overall survival (OS) of 295 NSCLC patients who received anti-PD-1 therapy by segregating them with multiple clinical factors including sex, age, race, smoking history, BMI, tumor grade and subtype. We also identified key signaling pathways and mutations that are enriched in patients with distinct responses to ICB by gene set enrichment analysis (GSEA) and mutational analyses. We found that former and current smokers have a higher response rate to anti-PD-1 treatment than non-smokers. GSEA results revealed that oxidative phosphorylation (OXPHOS) and mitochondrial related pathways are significantly enriched in both responders and smokers, suggesting a potential role of cellular metabolism in regulating immune response to ICB. We also demonstrated that all-trans retinoic acid (ATRA) which enhances mitochondrial function significantly enhanced the efficacy of anti-PD-1 treatment in vivo. Our clinical and bioinformatics based analyses revealed a connection between smoking induced metabolic switch and the response to immunotherapy, which can be the basis for developing novel combination therapies that are beneficial to never smoked NSCLC patients.

Keywords: Cigarette smoke; Immune therapy; NSCLC; Oxidative phosphorylation.

Figure 1.. Smoking status is associated with immunotherapy response in NSCLC patients.

(A) Progression free survival (PFS) and Overall survival (OS) of patients with different smoking history treated with anti-PD-1. (B) Distribution of patients’ response in current smoker, former smoker and non-smoker. CR: complete response, PR: partial response, SD: stable disease, PD: progression disease. (C) Chi-square analysis of smoking status and responses to anti-PD-1 therapy. DCB: durable clinical benefit, NDB: no durable benefit. (D) Distribution of DCB and NDB patients in smokers and non-smokers of NSCLC patients. *p<0.05, ** p<0.001.

Figure 2.. Oxidative phosphorylation (OXPHOS) is upregulated in smokers and responders to ICB.

(A) Enrichment of hallmark pathways in smokers and non-smokers analyzed by GSEA using TCGA, GSE81089, GSE11969 and GSE13213. (B) Enrichment plots of oxidative phosphorylation signature in smokers and non-smokers. (C) Flowchart of GSEA pathways analyses using TIDE predicted responders and non-responders from TCGA and GSE81089. (D) A waterfall plot of predicted responders and non-responders in TCGA and GSE81089 based on their TIDE values. (E) KEGG pathways that are enriched in responders and non-responders of NSCLC patients. (F) Enrichment plots of OXPHOS signature between responders and non-responders.

Figure 3.. Mitochondrial biogenesis is activated in smokers and responders.

(A) Ranking of Enriched mitochondrial pathways in smokers. (B) Enrichment plots of top 2 mitochondrial pathways in smokers and non-smokers. (C) Enrichment of mitochondrial pathways in TIDE predicted responders and non-responders. (D) MitoTracker staining of lung cancer cell lines treated with or without cigarette smoke extract (CSE). Scale bar: 20 μm

Figure 4.. OPA1 is upregulated by CSE and its expression is associated with the ICB response

(A) T cell dysfunction value of mitochondrial gene in various cancer cohorts examined by TIDE. (B) Relative mRNA expression of mitochondrial biogenesis genes between smokers and non-smokers in TCGA and GSE81089. (C) OPA1 protein levels were examined in various human and mouse cell lines treated with different dose of CSE by Western Blot. (D) OPA1 Immunohistochemistry of lung cancer tissue from NSCLC patients with known smoking history and response to ICB. (E) OPA1 expression in responders and non-responders examined by ROC plotter in NSCLC patients treated with anti-PD-1Ab. (F) AUC value of OPA1 and PD-L1 calculated by ROC plotter using anti-PD-1 treated patients. (G) Survival analysis of lung cancer patients with high or low OPA1 and CTL value by TIDE. *p<0.05, ** p<0.001.

Figure 5.. EGFR mutation is associated with a decreased response to immunotherapy and OXPHOS pathway in NSCLC patients.

(A) Oncoplot of mutation landscape of top differential mutated genes between smokers and non-smokers in TCGA. (B) Somatic mutation of TP53, KRAS and EGFR between smokers and non-smokers in WFBCCC cohort. (C) Distribution of patients’ response to anti-PD-1 therapy in EGFR wild type and mutant patients. (D) Progression free survival of 45 EGFR mutant or wild type NSCLC patients. (E) Enrichment plots of OXPHOS and mitochondrial related pathways in EGFR mutant and wild type NSCLC patients from TCGA. (F) mRNA expression of mitochondrial genes in EGFR wild type and mutant patients. TCGA. *p<0.05, ** p<0.001.

Figure 6.. All-trans retinoic acid (ATRA) promotes anti-PD-1 efficacy by elevating mitochondrial activity in tumor cells.

(A) MitoSOX staining of lung cancer cells treated with or without ATRA. (B) Protein levels of OPA1 were analyzed by Western Blot in lung cancer cells treated with or without ATRA. (C) Tumor volume of LL/2 cells inoculated mice treated with either monotherapy or combination therapy. (D) Tumor volume of CMT167 inoculated mice treated with either monotherapy or combination therapy. (E) immunocytochemical analyses of CD4,CD8 and Ly6G in LL/2 tumor samples with indicated treatment. (F) CD4,CD8 and Ly6G expressions in CMT167 tumor sections derived from each group were analyzed by immunocytochemistry. Scale bar: 20 μm *p<0.05, ** p<0.001.