Gene x environment interaction analysis confirms genetic modifier effects on steroid efficacy via TGF-β pathway in...

This paper continues our development of methods for discovery of genetic modifiers of the Duchenne muscular dystrophy (DMD) phenotype. DMD is an X-linked recessive disorder involving progressive muscle tissue loss with replacement by fat and fibrotic tissue, leading in most cases to loss of ambulation (LOA) by early to mid-adolescence. The standard pharmacologic treatment is corticosteroid administration, which increases average LOA by 2-3 years. There is variation in LOA due to specific DMD mutations, some of which permit the production of residual or partial dystrophin protein and lead to milder phenotypes. But there is also believed to be variation due to genetic modifiers acting even in patients whose DMD mutations preclude dystrophin production altogether, based in part on animal models, and several genes have been implicated as potential modifiers of LOA in DMD patients. Here we consider whether the mechanism of action of any of these genes might be to influence LOA by modifying the effects of corticosteroid exposure. We develop and evaluate a novel statistic, the PPI_GxE; we consider the issue of potential "phenocopies," or individuals whose late LOA might be due to residual dystrophin production; and we apply our approach to 12 candidate SNPs using our DMD dataset. We find evidence of genotype x steroid interaction effects for 4 out of the 12 SNPs we tested, which can be linked to the TGF-β pathway. These results corroborate the hypothesis that modifiers in the TGF-β pathway affect LOA by modulating the efficacy of corticosteroid administration.