When Vitamin D Meets Your DNA: Tiny Binding Changes, Big Shifts in MS Risk
Multiple sclerosis (MS) shows a striking latitude gradient—people living farther from the equator, where sunlight (and thus vitamin D synthesis) is lower, tend to have more MS. Observational data have long hinted that low vitamin D is linked to MS, but “is it causal?” has remained a hard question. A new study in PNAS (Feb 12, 2024) takes a clever Mendelian randomization (MR) approach to show that very specific DNA changes that alter vitamin D receptor (VDR) binding at the genome can nudge MS risk—pinning part of the vitamin D–MS connection on gene regulation itself.
Why this matters
Past MR studies used variants that change circulating 25-hydroxyvitamin D (25[OH]D) levels and found lower vitamin D causes higher MS risk. What’s been missing is a mechanistic bridge: does variation in VDR’s ability to bind DNA at particular loci—and thereby control transcription—actually contribute to MS susceptibility? This study argues yes, pointing to concrete VDR-binding variants (VDR-BVs) that track with risk.
What the team actually did
They started from calcitriol-stimulated lymphoblastoid cell lines (LCLs) with ChIP-exo to map places where VDR binds one allele more strongly than the other—allele-specific VDR binding. These sites harbor VDR-BVs.
For each VDR-BV present in their case–control datasets, they built a genetic instrumental variable (GIV): the allele dosage multiplied by the measured effect of that SNP on VDR binding strength at that locus. Think of this as a per-locus “proxy” for how much VDR sits on DNA there.
They also built standard GIVs for serum 25(OH)D using two large vitamin-D GWAS (Jiang et al.; Revez et al.) to capture a person’s genetically predicted vitamin D level.
Then they asked, in four large European-ancestry cohorts (KPNC, two Swedish cohorts, UK Biobank; in total 13,598 cases and 38,887 controls), whether VDR-BV instruments associate with MS—and crucially, whether those associations interact with the vitamin-D GIV (as you’d expect if the biology truly flows through VDR + vitamin D).
Key numbers at a glance
Low vitamin D → higher MS risk (replicated): OR = 1.85 (Jiang GIV), OR = 1.32 (Revez GIV), consistent direction across studies.
Two VDR-binding variants associated with MS after FDR correction:
rs2881514 (A allele): OR = 1.10 (95% CI: 1.05–1.15) → higher MS risk.
rs2531804 (A allele): OR = 0.82 (0.73–0.92) → lower MS risk.
Mechanistic hint via interaction: rs2881514’s effect was stronger when genetically predicted vitamin D was lower (meta-analysis interaction OR = 2.17, 1.10–4.29), consistent with true VDR-dependent biology rather than off-target pleiotropy.
A closer look at rs2881514 near RFTN1
rs2881514 sits ~1.5 kb downstream of the RFTN1 transcription start site on chromosome 3. In GTEx, this variant is an eQTL/sQTL for RFTN1 in several tissues. RFTN1 encodes Raftlin, a lipid-raft protein important for B-cell receptor signaling and Toll-like receptor pathways—immune circuits relevant to MS biology. Putting it together: an allele that increases VDR binding at this locus associates with higher MS risk, and the risk is accentuated when vitamin D is genetically lower. That’s a tidy mechanistic story: vitamin D → VDR binding at a specific regulatory element → changes in immune signaling gene regulation → altered MS susceptibility.
What this (does and doesn’t) say about vitamin D supplements
Yes: The MR replication again supports that lower lifetime vitamin D exposure causally elevates MS risk.
Careful: This study doesn’t test supplementation trials or dosing regimens; it shows that genetic variation in VDR-DNA interactions can modulate risk, and that effect scales with vitamin D bioavailability. Interventional implications need clinical trials that consider genotype-by-vitamin-D interactions as part of design and analysis.
Strengths worth calling out
Mechanistic MR: Rather than stopping at circulating vitamin D, the authors move down the pathway to receptor–DNA occupancy, getting closer to causal gene regulation.
Interaction-based pleiotropy check: Because VDR binding should matter most when vitamin D is low, testing the GIV(VDR-BV) × GIV(25[OH]D) interaction is a biologically grounded way to argue against horizontal pleiotropy for at least one top signal.
Large, multi-cohort meta-analysis with harmonized adjustments (sex, birth year, HLA-DRB1*15:01, PCs) increases robustness.
LD independence: Almost all VDR-BVs were not in substantial LD with known MS GWAS hits, suggesting new regulatory angles beyond established risk loci.
Important caveats
Single-SNP instruments per locus: That limits standard MR sensitivity analyses; the interaction framework helps but doesn’t fully erase pleiotropy concerns.
Cell model: VDR-BVs were mapped in EBV-transformed B-cell lines (LCLs). Key MS actors include T cells and microglia; allele-specific VDR binding in those cells remains to be charted.
Coverage: Only 112 VDR-BVs were analyzable—just a slice of all potential VDR sites across the immune epigenome.
Generalizability: All participants were of European ancestry. The findings need testing across ancestries.
Phenotyping differences: UK Biobank cases included self-report/EHR-derived diagnoses and showed a lower DRB1*15:01 frequency, which may attenuate effects.
Where this points next
Cell-type–specific maps: Profile allele-specific VDR binding in primary CD4⁺/CD8⁺ T cells, B cells, and myeloid cells under vitamin D stimulation. Overlay with ATAC-seq/ChIP-seq to pinpoint causal regulatory elements.
Fine-mapping + perturbation: CRISPRi/a at rs2881514’s regulatory element near RFTN1 to test directionality on immune signaling and cytokine outputs; assess vitamin D dose-response.
Gene–environment trial design: Incorporate VDR-BV genotypes (and 25[OH]D polygenic scores) into vitamin D supplementation RCTs for MS risk/progression to test genotype-by-supplement interactions prospectively.
Pathway integration: Connect VDR-BV targets to HLA-DRB1*15:01 status, EBV serostatus, and interferon signatures to model multi-hit MS risk architecture.
This study moves beyond “vitamin D levels” and shows that where VDR binds the genome—tuned by specific SNPs—can causally influence MS risk, with the clearest case at rs2881514 near RFTN1. The effect is stronger when genetically predicted vitamin D is lower, aligning with VDR biology. It’s the most direct population-scale evidence so far that allele-specific receptor–DNA interactions are part of the causal chain linking vitamin D to MS.
Disclaimer: This blog post is based on the provided research article and is intended for informational purposes only. It is not intended to provide medical advice. Please consult with a healthcare professional for any health concerns.
References:
Adams C. et al., “Evidence supports a causal association between allele-specific vitamin D receptor binding and multiple sclerosis among Europeans,” PNAS 121(8): e2302259121 (Published Feb 12, 2024).
