Deep Resequencing Identifies Enrichment of Nonsynonymous Variants in Leprosy Susceptibility Genes

Vinicius M. Fava^1,2; Guillaume Lettre^3,4, and Erwin Schurr^1,2

1. Program in Infectious Diseases and Immunity in Global Health, The Research Institute of the McGill University Health Centre, Montreal, Quebec, Canada; 2. McGill International TB Centre, Faculty of Medicine,McGill University, Montreal, Quebec, Canada; 3. Montreal Heart Institute, Montreal, Quebec H1T 1C8, Canada. 4. Department of Medicine, Faculty of Medicine, Université de Montréal, Montreal,Quebec, H3T 1J4, Canada

Background: Among infectious diseases, leprosy has the largest number of replicated host susceptibility loci identified by genetic studies. However, leprosy risk variants described by GWAS and positional cloning (PC) are typically common, non-coding variants. Especially GWAS loci can be difficult to assign to a specific gene due to extensive linkage disequilibrium in the locus. Therefore, a direct biological translation of such genetic associations can be challenging. Here, we investigated the enrichment of non-synonymous variants in the coding regions of 36 genes located in leprosy per se candidate susceptibility loci by resequencing all exons in 551 leprosy cases and 480 healthy controls.

Methods: Deep resequencing was performed employing Illumina TrueSeq v.1.5 custom libraries. Standard Illumina and GATK protocols were used to perform alignment, variant calling and filtering. Additional QC thresholds were applied to reduce false positives calls in targeted approaches. The association analysis was carried out by grouping variants per gene and performing association testing with SKAT-O and Variable threshold (VT) methods as implemented in EPACTS.

Results: Of the 36 genes evaluated for association with leprosy, six displayed only a single variant or were non-polymorphic. Of the remaining 30 genes, two showed gene-wise evidence of association with leprosy per se. For the LACC1 gene an enrichment of non-synonymous variants was observed (p_VT= 4.2 e^-6). However, this association was driven entirely by the strong effect of a single common variant p.I254V (rs3764147; p_single = 5.9 e^-10) with known association with leprosy per se. The second gene-wise association was found for IL18R1 where six variants with allele frequencies ranging from 0.0005 to 0.01 were enriched in healthy controls (p_SKAT-O=5.1e^-4; p_VT=3.7e^-3). The initial leprosy GWAS signal in chromosome 2q12.1 encompassed four genes: IL1RL1, IL18R1, IL18RAP and SLC9A4. Here we found only the IL18R1 gene associated with leprosy per se suggesting this gene as functional candidate for leprosy susceptibility at the 2q12.1 locus.