Unveiling potential drug targets for lung squamous cell carcinoma through the integration of druggable genome and genome-wide association data

Abstract

Background: Lung squamous cell carcinoma (LSCC) is a major subtype of lung cancer with poor prognosis and low survival rate. Compared with lung adenocarcinoma, yet no FDA-approved targeted-therapy has been found for lung squamous cell carcinoma. Methods: To identify potential drug targets for LSCC, Summary-data-based Mendelian randomization (SMR) analysis was used to examine the potential association between 4,543 druggable genes and LSCC, followed by colocalization analysis and HEIDI tests to confirm the robustness of the result. Phenome-wide association study (PheWAS) explored potential side effects of candidate drug targets. Enrichment analysis and protein-protein interaction networks revealed the function and significance of therapeutic targets. Single-cell expression analysis was used to examine cell types with enrichment expression of druggable genes in LSCC tissue. Drug prediction included screening potential drug candidates and evaluating their interactions with targets through molecular docking. Results: This research has identified ten significant drug targets for LSCC through a comprehensive SMR analysis. These targets included (COPA, PKD2L1, CCR1, C2, CYP21A2, and NCSTN as risk factors, and CCNA2, C4A, APOM, and LPAR2 as protective factors). PheWAS demonstrated that C2, CCNA2, LPAR2, and NCSTN exhibited associations with other phenotypes at the genetic level. Then, we found four potentially effective drugs with the Dsigdb database. Subsequently, molecular docking indicated that favorable binding interactions between drug candidates and potential target molecules. In the druggability evaluation, five out of ten drug target genes have been used in drug development (APOM, C4A, CCNA2, COPA, and PKD2L1). Six out of ten druggable genes showed significant expression in LSCC tissues (COPA, PKD2L1, CCR1, C2, NCSTN, LPAR2). Besides, Single-cell expression analysis revealed that C2 and CCNA2 were primarily enriched in macrophages, while COPA and NCSTN were enriched in both macrophages and epithelial cells. Conclusion: Our research revealed ten potential druggable genes for LSCC treatment, which might help to advance the precise and efficient therapeutic approaches of LSCC.

Keywords: drug targets; genome-wide association study; genomics; lung squamous cell carcinoma; summary-data-based mendelian randomization.

FIGURE 1

Outline of the study design. GWAS, genome-wide association studies; SMR, summary-data-based Mendelian Randomization; HEIDI, heterogeneity in dependent instruments; PPH4, the posterior probability of hypothesis 4; PheWAS, Phenome-Wide Association Study.

FIGURE 2

Volcano plot displaying the SMR results from the discovery phase for 23 significant genes. Dashed line on the horizontal axis represents FDR 0.05.

FIGURE 3

Forest plots displaying the SMR results from the discovery phase for 23 significant genes.

FIGURE 4

Forest plots displaying the SMR results from the replication phase for 10 significant genes.

FIGURE 5

Enrichment results of 10 drug target genes. (A) Go enrichment results. (B) KEGG enrichment results.

FIGURE 6

PPI network.

FIGURE 7

Single-cell type expression in lung squamous cell carcinoma tissue for the drug target genes identified by SMR. (A) 22 cell clusters and 14 cell types were identified. (B) and (C) show the expression of drug target genes in each cluster.

FIGURE 8

Docking results of available proteins with small molecule ligands. (A) CCNA2 docking Irinotecan hydrochloride, (B) PKD2L1 docking Irinotecan hydrochloride, (C) CYP21A2 docking simvastatin, (D) NCSTN docking isotretinoin, (E) APOM docking Irinotecan hydrochloride, (F) PKD2L1 docking isotretinoin, (G) CCR1 docking simvastatin, (H) APOM docking ciglitazone.