Molecular mechanism for the multiple sclerosis risk variant rs17594362

Abstract

Multiple sclerosis (MS) is known as an autoimmune demyelinating disease of the central nervous system. However, its cause remains elusive. Given previous studies suggesting that dysfunctional oligodendrocytes (OLs) may trigger MS, we tested whether single nucleotide polymorphisms (SNPs) associated with MS affect OL enhancers, potentially increasing MS risk by dysregulating gene expression of OL lineage cells. We found that two closely spaced OL enhancers, which are 3 Kb apart on chromosome 13, overlap two MS SNPs in linkage disequilibrium-rs17594362 and rs12429256. Our data revealed that the two MS SNPs significantly up-regulate the associated OL enhancers, which we have named as Rgcc-E1 and Rgcc-E2. Analysis of Hi-C data and epigenome editing experiments shows that Rgcc is the primary target of Rgcc-E1 and Rgcc-E2. Collectively, these data indicate that the molecular mechanism of rs17594362 and rs12429256 is to induce Rgcc overexpression by potentiating the enhancer activity of Rgcc-E1 and Rgcc-E2. Importantly, the dosage of the rs17594362/rs12429256 risk allele is positively correlated with the expression level of Rgcc in the human population, confirming our molecular mechanism. Our study also suggests that Rgcc overexpression in OL lineage cells may be a key cellular mechanism of rs17594362 and rs12429256 for MS.

Figure 1

The NIH Roadmap Epigenomics Project H3K27ac ChIP-seq data for rs17594362 and rs12429256. These H3K27ac ChIP-seq data were visualized by the WASHU Epigenome Browser.

Figure 2

The enhancer activity of Rgcc-E1 and Rgcc-E2 in OL lineage cells and the effect of rs17594362 and rs12429256. (A) The enhancer activity of Rgcc-E1 and Rgcc-E2 cloned in pGL3-promoter was compared with that of pGL3-promoter (the empty vector) in primary mouse OPCs cultured in the proliferating and differentiating conditions. Shown are the means and standard errors (n = 8). ^*P value < 2.0 × 10⁻³ by one sample two-sided Student’s t test corrected by the Bonferroni procedure. (B) The rat OL ChIP-seq data underlying Rgcc-E1 and Rgcc-E2. (C) The effect of rs17594362 and rs12429256 on the enhancer activity of Rgcc-E1 and Rgcc-E2 in primary mouse OPCs cultured in proliferating and differentiating conditions. Shown are the means and standard errors (n = 8). ^*P < 3.5 × 10⁻² by unpaired two-sided Student’s t test corrected by the Bonferroni procedure.

Figure 3

Chromatin interaction data suggest that Rgcc is the primary target of Rgcc-E1 and Rgcc-E2. (A) The publicly available 5 Kb-resolution Hi-C data for 7 diverse cell types from human and mouse were analyzed to delineate the TAD for Rgcc-E1 and Rgcc-E2 (. The interaction frequency between two loci is indicated by color: white means no interaction, and red the strongest possible interaction. The position of Rgcc-E1 and Rgcc-E2 is marked by thin black crossing lines. The position of Rgcc is marked by thin blue crossing lines. The Rgcc-E1/E2 TAD is demarcated in green for IMR90 and CH12-LX. IMR90: lung fibroblast. K562 and KBM7: chronic myelogenous leukemia cells. HeLa: cervical cancer cell. HUVEC: human umbilical vein endothelial cell. NHEK: normal human epidermal keratinocyte. CH12-LX: murine CH12 B-cell lymphoma cell. (B) The TAD for Rgcc-E1 and Rgcc-E2 contains one protein-coding gene, Rgcc. Naa16 and Vwa8 are its neighbors, but their promoters are found out of the TAD.

Figure 4

Epigenome editing shows that Rgcc is the primary target of Rgcc-E1 and Rgcc-E2. (A) gRNAs used for epigenome editing experiments. Scr is a non-targeting negative control gRNA. For each of Rgcc-E1 and Rgcc-E2, four gRNAs were designed. (B) Rgcc-E1 and Rgcc-E2 were silenced by dCas9-KRAB in Oli-neu cells, and its effect on the expression of Rgcc, Vwa8 and Naa16 determined by RT-qPCR. Shown are the means and standard errors (n = 3). For Rgcc and Rgcc-E1, ^*P < 4.7 × 10⁻⁴. For Rgcc and Rgcc-E2, ^*P < 1.2 × 10⁻³. For Vwa8 and Rgcc-E1, ^*P < 3.3 × 10⁻³. For Vwa8 and Rgcc-E2, ^*P < 7.0 × 10⁻⁴. For Naa16 and Rgcc-E2, ^*P < 5.1 × 10⁻³. All P values by unpaired two-sided Student’s t test corrected by the Bonferroni procedure.