Multiple sclerosis (MS) is a chronic and debilitating autoimmune condition of the central nervous system (1). A complex interplay of genetic and environmental factors contributes to disease onset and progression in MS patients (2). Vitamin D deficiency is a known susceptibility factor for MS, however the underlying mechanism of vitamin D gene-environment interactions in MS etiology is still poorly understood (3).

There are several lines of evidence supporting the role of vitamin D in MS susceptibility, including higher MS prevalence in regions with lower annual UVB exposure (4–7), and migrant studies showing changes to MS risk based on latitude changes (8–10). Increased actinic sun damage, and reporting more time outdoors during childhood/adolescence, have both been associated with reduced risk of MS (11–13). Large prospective studies measuring vitamin D status prior to MS onset showed that participants with low serum vitamin D levels were at greater risk of developing MS (14–16). Clinical studies have reported that vitamin D supplementation in MS patients lowered disease relapse risk (17–21), while lower vitamin D status at MS onset is associated with worse cognitive function and neuronal integrity (22). MS genetic risk loci, identified by GWAS, are enriched with binding sites for the vitamin D receptor (VDR) (23), and several MS risk variants have been localized to VDR binding sites (24).

Recent in silico evidence suggests that vitamin D receptor (VDR) super-enhancers (VSE) may be influenced by nearby MS risk SNPs, thereby impacting the activity of these VSEs in relation to disease (25). Vitamin D exerts its biological effects when the active hormone [1,25-dihydroxyvitamin D, 1,25(OH)₂D] binds VDR, which then acts as a transcription factor upon binding to VDR response elements in target genes (26). These VDR target genes are often associated with clusters of VDR binding sites that form super-enhancers, regulatory genomic regions consisting of dense clusters of enhancers (27). VSEs contain clusters of VDR binding sites, and loop to hundreds of promoter regions, upregulating vitamin D target genes (27).

Lu et al. (25) identified specific genes near VSE regions that overlap with MS risk variants. We investigated the expression of ten VSE-MS-linked genes in a cohort of 64 persons with MS representing the two extremes of disease activity, to assess whether expression differences were associated with mild vs. severe MS phenotype.

Clinical characteristics of the study participants are listed in Table 1. There were large differences in ARMSS scores between the mild and severe phenotype groups, as expected due to study design (Table 1). Across all study participants, the mean ARMSS scores were 1.34 and 6.90, for the mild and severe phenotypes, respectively. Other variables, including the age at onset, age at recruitment, and sex (% female) were not significantly different between mild and severe phenotypes (Table 1). The severe phenotype groups in both cohorts consisted of only female patients (Table 1). The number of participants in each group receiving at least one of the MS disease modifying therapies, as well as the types of treatments, did not differ significantly between the groups; 58.1% in the mild group, 61.9% in the severe group (data not shown).

In the discovery cohort, whole transcriptome sequencing was conducted for the ten candidate MS-linked-VSE genes selected based on prior evidence (Figure 1). Mean expression levels were analyzed for significant differences in the mild vs. severe phenotype groups (≥ 2-fold-change, p < 0.05). Four of the candidate MS-linked-VSE genes GRINA, PLEC, PARP10, and LRG1 were significantly upregulated in the severe phenotype MS subgroup (Table 2), and thus were evaluated in the validation cohort (n = 44) using custom-designed ddPCR assays. This analysis demonstrated three genes that were differentially expressed between MS phenotypes including GRINA, PLEC, and LRG1 (Figure 2).

This pilot study compared expression of MS-VSE genes in mild vs. severe MS phenotypes, in order to elucidate a potential role for vitamin D in MS etiology, as well as identifying potential future biomarkers for differentiating these phenotype subgroups. Vitamin D super-enhancers in MS is a relatively new area of research and this is the first study, to our knowledge, to directly investigate the putative role of VSE-linked genes associated with MS risk SNPs in MS phenotypic severity.

Vitamin D deficiency has long been implicated in MS susceptibility, yet the exact vitamin D gene-environment or epigenetic interactions in MS etiology remain unclear. A better understanding may at best help in prevention or modulation of disease severity, or at least create awareness of the importance of vitamin D sufficiency. Active vitamin D, 1,25(OH)D exerts its biological effects by binding to the VDR and altering gene expression. Vitamin D super-enhancers contain multiple VDR binding sites, which then loop to transcription start site of target gene(s) to stimulate transcription (29). VSEs are believed to have more pronounced effects on gene expression than traditional enhancers (and many are signal-inducible via 1,25(OH)D exposure (26, 27). Recent in silico evidence identified specific genes with VSE regions overlapping with MS risk variants as well as VSE genes overlapping with MS risk variants (25). We compared the expression of ten of these identified candidates in our cohort. It is hypothesized that MS risk SNPs within VSEs, or genes associated with VSEs may alter the VSE activity to impact disease risk and activity (30).

RNA-sequencing and ddPCR were used to assess differential expression of select candidates. Three VSE genes (PLEC, GRINA, LRG1) that were previously identified as linked to MS variants (25) showed consistent differences in expression between the mild vs. severe phenotype groups in both the discovery and confirmation cohorts. It should be noted that for neurodegenerative diseases, even small fold-changes in expression levels can relate to disease development and activity (24, 31, 32). Interestingly, a significant increase in transcription levels was associated with these VSE-genes in the severe phenotype. The findings implicate a role for vitamin D super-enhancers in modulating disease via environment-gene interactions. In addition, expression levels may have some utility as prognostic biomarkers in the future.

The molecular mechanisms involved in VSE regulation of gene expression and its impact on MS has not been studied. GRINA encodes a glutamate ionotropic receptor N-methyl-D-aspartate (NMDA)-type subunit associated protein 1, expressed on the postsynaptic membrane of neuronal synapses (33). The receptor is an excitatory glutamate-gated ion channel found throughout the body (34). Excessive glutamate at the sites of demyelination and overstimulation of glutamate receptors leading to neuronal death have been reported (35, 36). The PLEC gene encodes the cytoskeleton plectin protein, which is found in many tissues, including nervous tissue (37, 38). The degradation of cerebral tissue integrity and elasticity in MS patients, especially in glial cells (39), may be affected. LRG1 encodes the leucine-rich alpha-2-glycoprotein, which has recently been reported to promote angiogenesis in the brain (40, 41).

While it is hypothesized that VSE-related MS risk variants may lead to changes in the epigenomic landscape involved in MS pathogenesis, we cannot rule out that the converse is true—it may be that disease severity affects VDR related changes in chromatin accessibility and thus regulation of gene expression. It may also be a combination of both effects. Studying expression in individuals with first de-myelinating events, and/or longitudinally, would be needed to better support any potential cause-and-effect relationship.

Given the small sample size in this pilot study, follow up in larger phenotype cohorts or case-control designs, will be useful to validate a role for VSE genes in MS risk and severity, as well as to explore a potential role as clinically useful biomarkers. In particular, a comparison of mild cases vs. healthy controls and severe cases vs. healthy controls would be warranted in this validation. Furthermore, the severe phenotype group in the validation cohort contained exclusively females, and obtaining larger sample sizes with more males are warranted to confirm the findings. While the cohorts do not differ in the baseline demographic characteristics, we acknowledge the limitation on the results of not having more precise 1:1 propensity score matching between the individuals in the discovery and validation cohorts.

Another potential limitation of this study involves the use of whole blood, which has both advantages and disadvantages, as we have discussed previously (28). Many recent publications studying MS biomarkers have used whole blood (42). Other blood sample types (serum, plasma) require more onerous standardization to avoid variability in sample contents introduced (43). We believe that whole blood remains a reasonable choice for gene expression studies in MS because the collection is easily standardized, and there is precedent from numerous prior studies. Recent publications continue to demonstrate the important role of peripheral blood cells in the pathogenesis of MS (44, 45). Nonetheless, future studies extracting transcripts from different cell types could help determine if the results are more generalizable, given that the genomic binding pattern of VDR can be variable between tissues (29). While our findings support a link between VSE genes and MS risk variants, the mechanisms involved have not been studied. Extension to murine EAE models could explore functional effects.

The study of vitamin D super-enhancers, and their effect on human health, is a relatively new and growing area of study. There are very limited publications in this area for the field of multiple sclerosis. The candidate genes investigated may be important players in the interaction between the environment (vitamin D) and genetic risk associated with multiple sclerosis.

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author/s.

The studies involving human participants were reviewed and approved by University of Calgary Conjoint Health Research Ethics Board (REB17-1193). The patients/participants provided their written informed consent to participate in this study.

SO: conceived study and design, lab work, data analysis, data interpretation, manuscript writing, and manuscript editing. AS: lab work, data collection, data analysis, and manuscript writing. MG: data collection, data analysis, and manuscript editing. KM: technical assistance and experiments. LM: cohort collection and study design. AK: cohort collection, study design, and data analysis. GP: cohort collection, study design, data interpretation, and manuscript revision. All authors reviewed and approved the final manuscript version.