Transcript specific regulation of expression influences susceptibility to multiple sclerosis

Abstract

Genome-wide association studies (GWAS) have identified over 100 loci containing single nucleotide variants (SNVs) that influence the risk of developing multiple sclerosis (MS). Most of these loci lie in non-coding regulatory regions of the genome that are active in immune cells and are therefore thought to modify risk by altering the expression of key immune genes. To explore this hypothesis we screened genes flanking MS-associated variants for evidence of allele specific expression (ASE) by quantifying the transcription of coding variants in linkage disequilibrium with MS-associated SNVs. In total, we were able to identify and successfully analyse 200 such coding variants (from 112 genes) in both CD4+ and CD8+ T cells from 106 MS patients and 105 controls. Fifty-six of these coding variants (from 43 genes) showed statistically significant evidence of ASE in one or both cell types. In the Lck interacting transmembrane adaptor 1 gene (LIME1), for example, we were able to show that in both cell types, the MS-associated variant rs2256814 increased the expression of some transcripts while simultaneously reducing the expression of other transcripts. In CD4+ cells from an additional independent set of 96 cases and 93 controls we were able to replicate the effect of this SNV on the balance of alternate LIME1 transcripts using qPCR (p = 5 × 10^-24). Our data thus indicate that some of the MS-associated SNVs identified by GWAS likely exert their effects on risk by distorting the balance of alternate transcripts rather than by changing the overall level of gene expression.

Fig. 1. ASE in genes surrounding the MS-associated SNV rs2256814.

Box-whisker plots represent median, quartiles and 1.5× interquartile range. Statistically significant ASE is shown in grey, with the transcripts captured by the rs914559 SNV showing increased expression of the risk allele and those transcripts captured by rs2236510 showing decreased expression of risk allele.

Fig. 2. Replication analysis of transcript specific LIME1 regulation.

Box-whisker plots represent median, quartiles and 1.5x interquartile range. a Quantitative PCR relative gene expression capturing ENST00000487026.5 and ENST00000465591.1 LIME1 transcripts with increased expression in rs914559_C risk allele carriers. b Quantitative PCR relative gene expression capturing ENST00000493265.2 and ENST00000621325.1 LIME1 transcripts with decreased expression in rs2236510_T risk allele carriers. c Ratio of the ENST00000487026.5/ENST00000465591.1: ENST00000493265.2/ENST00000621325.1 correlated with rs914559 genotype.