Integration of QTL and comprehensive analysis in the circulating inflammatory cytokines for pan-cancer

Chemokines and cytokines are components of the tumor microenvironment and also influence tumorigenesis and its composition. However, whether they genetically proxy tumorigenesis is unclear. For causal inferences, eQTL and pQTL were used to determine the role of chemokines and cytokines in pan-cancer. The impact on the tumor immune microenvironment was also explored.

This study leveraged summary statistics from respective genome-wide association studies (GWAS) of 109 cytokines and chemokines in 18 types of solid tumors. Single nucleotide polymorphisms (SNPs) robustly associated with the cytokines and chemokines, located in or close to their coding gene (cis), were used as instrumental variables. A two-sample MR design was employed, followed by comprehensive sensitivity analyses to validate the robustness of results. The impact on immune infiltration was investigated using the TIMER and TISIDB websites. Survival analysis was conducted using the K-M plotter and TIMER 2.0 websites. The TISCH and GEO databases were used to carry out scRNA cell analysis.Analyzing relevant proteins using the STRING database and conducting enrichment pathways for GO analysis of the identified proteins.

The results of the inverse-variance weighted (IVW) method using cis-protein QTL (cis-pQTL) instruments showed the causal effects of TNF in reducing the risk of squamous cell lung cancer (LUSC) and HGF in reducing the risk of head and neck cancer (HNSC).The results were consistent with the eQTL. HGF was associated with better overall survival (OS) in HNSC, regardless of the types of cells enriched. However, high expression of the ligand MET for HGF leads to a decrease in overall survival in LUSC. TNF was related to poor OS in LUSC with no significant impact. However, in CD8 + T cell-enriched, eosinophil-enriched, macrophage-enriched, and NK cell-deficient types of LUSC, high expression of TNF leads to a poor prognosis, and there is statistical significance. The results showed a significant positive correlation between TNF and most immune cell infiltration, immunomodulator and chemokine in LUSC. HGF is positively correlated with the majority of immune cells except CD56 + cells, as well as some immune regulatory factors and chemotactic factors. According to single-cell sequencing results, HGF is mainly secreted by fibroblasts and myofibroblasts in HNSC, while in LUSC, it is primarily secreted by macrophages and CD8 + T cells secrete TNF. The GO/KEGG analysis suggests that proteins related to HGF are mainly involved in regulating peptidyl-tyrosine phosphorylation and positive regulation of the MAPK cascade. Proteins related to TNF are primarily associated with the regulation of I-kappaB kinase/NF-kappaB signaling and cytokine-mediated signaling pathway.

HGF is primarily secreted by fibroblasts in HNSC and may have a protective effect on the occurrence and prognosis of HNSC. These effects are independent of immune cell influence, and this role may not necessarily be mediated through the HGF/MET pathway. On the other hand, TNF in LUSC is mainly secreted by immune cells like CD8 + T cell, and it may have a protective effect on the occurrence of LUSC. However, it's impact on the prognosis of LUSC through the immune microenvironment may have a different effect.