Project Summary
Multiple sclerosis (MS) is an immune-mediated disease that destroys the protective myelin sheath surrounding
nerve cells, and its cause is unknown. MS prevalence is higher in regions farther from the equator, leading to
the hypothesis that sun exposure, and consequently vitamin D exposure, has a protective effect on MS risk.
Recent studies utilizing Mendelian randomization (MR) analysis have demonstrated strong evidence for a causal
role of low serum 25-hydroxyvitamin D (25(OH)D) levels in MS pathogenesis; however, the molecular
mechanisms underlying the association are unknown. It is well established that 25(OH)D signaling operates
through the nuclear vitamin D receptor (VDR), a ligand-regulated transcription factor. VDRs recognize specific
binding sites in DNA, leading to activation or suppression of a target gene. Previous research has found that
genetic variation of tagging SNPs (VDR-BVs) within these VDR binding sites, can alter binding affinity. These
VDR-BVs are enriched in genomic regions associated with MS, as well as with several other autoimmune
diseases, which suggests regulation of specific genes mediated by vitamin D could affect MS risk and severity.
No formal investigation of genetic variation of individual VDR-BVs in MS cases and controls has been reported
and VDR-BVs are strong candidates to investigate for genetic variation relevant to MS. The overall objective of
this F31 application is to estimate the effect of VDR-BVs on MS risk and severity. We hypothesize that
genetic variation in VDR-BVs is associated with MS susceptibility and severity, and further, that the effect of this
genetic variation on MS is modulated by bioavailability of vitamin D. Our approach will use a dataset of
approximately 10,000 clinically definite MS cases and 35,000 controls frequency matched on age and
race/ethnicity with whole genome genotypes, demographic, clinical, and environmental exposure data and will:
1) Estimate the association between VDR-BVs near established MS risk loci and MS risk and severity; 2)
Estimate the association between VDR-BVs across the human genome and MS risk and severity; and 3)
Replicate findings in independent datasets of MS cases and controls and combine replicated associations across
all datasets via meta-analysis. VDR-BVs were previously identified through ChIP-seq analysis in Lymphoblastoid
Cell Lines and we will use machine learning to prioritize candidate VDR-BVs for Aim 2. Analyses will incorporate
a genetic instrumental variable as measure of 25(OH)D bioavailability to test for interaction between VDR-BVs
and 25(OH)D bioavailability. Results from the proposed aims will determine whether MS GWAS findings
are partially explained by genetic variation in VDR binding affinity, identify genetic risk factors for MS
related to vitamin D, and identify new genes and regulatory pathways involved in the development of
MS. Ultimately, the goal of this work understand the mechanisms by which vitamin D affects this immune-
mediated neurological disorder.