Whole Exome Sequencing reveals new candidate genes in host genomic susceptibility to Respiratory Syncytial Virus Disease

Abstract

Respiratory syncytial virus (RSV) is an important cause of serious lower respiratory tract disease in infants. Several studies have shown evidence pointing to the genome of the host as an important factor determining susceptibility to respiratory disease caused by RSV. We sequenced the complete exomes of 54 patients infected by RSV that needed hospitalization due to development of severe bronchiolitis. The Iberian sample (IBS) from The 1000 Genomes Project (1000G) was used as control group; all the association results were pseudo-replicated using other 1000G-European controls and Spanish controls. The study points to SNP rs199665292 in the olfactory receptor (OR) gene OR13C5 as the best candidate variant (P-value = 1.16 × 10^-12; OR = 5.56). Genetic variants at HLA genes (HLA-DQA1, HLA-DPB1), and in the mucin 4 gene (MUC4) also emerge as susceptibility candidates. By collapsing rare variants in genes and weighing by pathogenicity, we obtained confirmatory signals of association in the OR gene OR8U1/OR8U8, the taste receptor TAS2R19, and another mucin gene (MUC6). Overall, we identified new predisposition variants and genes related to RSV infection. Of special interest is the association of RSV to olfactory and taste receptors; this finding is in line with recent evidence pointing to their role in viral infectious diseases.

Figure 1

(A) MDS analysis carried out on a matrix of pair-wise individual identity-by-state values between RSV patients vs. different 1000G population reference sets (see legend inset). (B) Enlarged detail of the European cluster observed in (A). (C) Admixture analysis of the samples analyzed in (A).

Figure 2

(A) Manhattan plot of common variants observed in RSV patients. (B) QQ-plot of P-values for common variation observed in IBS controls. P-values obtained under a permutation approach (1,000 permutations) are shown in blue.

Figure 3

(A) P-values of SNP association between RSV patients and controls in different control groups and in the merged control group “ALL” (IBS + CEU + GBR + TSI). (B) P-values of gene burden association test between RSV patients and controls in different control groups and in the merged control group “ALL” considering all variants in genes. (C) P-values of gene burden association test between RSV patients and controls in different control groups and in the merged control group “ALL” considering only rare variants in genes. The vertical grey shadow in the figures indicates the threshold for the adjusted Bonferroni P-value according to the number of independent tests considering CEU, GBR, TSI, and ALL as additional control groups of the best SNPs (n = 12) and gene candidates (one gene when collapsing common variants and four genes when collapsing rare variants). The red vertical line indicates the genome Bonferroni threshold considering genes and all the common SNPs that are shared between cohorts.