A Genome-Wide Association Study Identifies SERPINB10, CRLF3, STX7, LAMP3, IFNG-AS1 and KRT80 As Risk Loci Contributing to Cutaneous Leishmaniasis In Brazil.

Castellucci, Léa C; Almeida, Lucas; Cherlin, Svetlana; Fakiola, Michaela; Francis, Richard W; Carvalho, Edgar M; Santos da Hora, Anadílton; do Lago, Tainã Souza; Figueiredo, Amanda B; Cavalcanti, Clara M; Alves, Natalia S; Morais, Katia L P; Teixeira-Carvalho, Andréa; Dutra, Walderez O; Gollob, Kenneth J; Cordell, Heather J; Blackwell, Jenefer M

BACKGROUND Our goal was to identify genetic risk factors for cutaneous leishmaniasis (CL) caused by Leishmania braziliensis. METHODS Genotyping 2066 CL cases and 2046 controls using Illumina HumanCoreExomeBeadChips provided data for 4,498,586 imputed single nucleotide variants (SNVs). Genome-wide association testing using linear mixed models took account of genetic diversity/ethnicity/admixture. Post-GWAS positional, expression quantitative trait locus (eQTL), and chromatin interaction mapping was performed in FUMA. Transcriptional data were compared between lesions and normal skin, and cytokines measured using flow cytometry and Bioplex assay. RESULTS Positional mapping identified 32 genomic loci associated with CL, none achieving genome-wide significance (P&5x10 -8). Lead SNVs at 23 loci occurred at protein coding or non-coding RNA genes, 15 with eQTLs for functionally relevant cells/tissues and/or showed differential expression in lesions. Of these, the 6 most plausible genetic risk loci were: SERPINB10 (Pimputed_1000G=2.67x10 -6), CRLF3 (Pimputed_1000G=5.12x10 -6), STX7 (Pimputed_1000G=6.06x10 -6), KRT80 (Pimputed_1000G=6.58x10 -6), LAMP3 (Pimputed_1000G=6.54x10 -6) and IFNG-AS1 (Pimputed_1000G=1.32x10 -5). LAMP3 (Padjusted=9.25x10 -12; +6-fold), STX7 (Padjusted=7.62x10 -3; +1.3-fold) and CRLF3 (Padjusted=9.19x10 -9; +1.97-fold) were expressed more highly in CL biopsies compared to normal skin; KRT80 (Padjusted=3.07x10 -8; -3-fold) was lower. Multiple cis-eQTLs across SERPINB10 mapped to chromatin interaction regions of transcriptional/enhancer activity in neutrophils, monocytes, B cells and haematopoietic stem cells. Those at IFNG-AS1 mapped to transcriptional/enhancer regions in T, natural killer, and B cells. The percent peripheral blood CD3 + T cells making antigen-specific interferon-γ differed significantly by IFNG-AS1 genotype. CONCLUSIONS This first GWAS for CL identified multiple genetic risk loci including a novel lead to understanding CL pathogenesis through regulation of interferon-γ by IFNG antisense RNA 1.