Genetic variants in glutamine metabolic pathway genes predict cutaneous melanoma‐specific survival

Abstract

Glutamine dependence is a unique metabolic defect seen in cutaneous melanoma (CM), directly influencing the treatment and prognosis. Here, we investigated the associations between 6025 common single‐nucleotide polymorphisms (SNPs) in 77 glutamine metabolic pathway genes with CM‐specific survival (CMSS) using genotyping datasets from two published genome‐wide association studies (GWASs). In the single‐locus analysis, 76 SNPs were found to be significantly associated with CMSS (P < .050, false‐positive report probability < 0.2 and Bayesian false discovery probability < 0.8) in the discovery dataset, of which seven SNPs were replicated in the validation dataset and three SNPs (HAL rs17676826T > C, LGSN rs12663017T > A, and NOXRED1 rs8012548A > G) independently predicted CMSS, with an effect‐allele attributed adjusted hazards ratio of 1.52 (95% confidence interval = 1.19‐1.93) and P < .001, 0.68 (0.54‐0.87) and P = .002 and 0.62 (0.46‐0.83) and P = .002, respectively. The model including the number of unfavorable genotypes (NUGs) of these three SNPs and covariates improved the five‐year CMSS prediction (P = .012) than the one with other covariates only. Further expression quantitative trait loci (eQTL) analysis found that the LGSN rs12663017 A allele was significantly associated with increased messenger RNA (mRNA) expression levels (P = 8.89 × 10 −11) in lymphoblastoid cell lines of the 1000 Genomes Project database. In the analysis of the genotype tissue expression (GTEx) project datasets, HAL rs17676826 C and NOXRED1 rs8012548 G alleles were significantly associated with their mRNA expression levels in sun‐exposed skin of the lower leg (P = 6.62 × 10−6 and 1.37 × 10−7, respectively) and in sun‐not‐exposed suprapubic skin (P < .001 and 1.43 × 10−8, respectively). Taken together, these genetic variants of glutamine‐metabolic pathway genes may be promising predictors of survival in patients with CM.