GWAS Evidence for a Neuronal Contribution to Multiple Sclerosis Susceptibility
In this preprint (“GWAS highlights the neuronal contribution to multiple sclerosis susceptibility,” posted January 6, 2025), De Jager and colleagues revisit a long-standing assumption in multiple sclerosis (MS) genetics: that most inherited risk operates principally through peripheral immune cells and microglia. While acknowledging strong immune contributions, the authors explicitly design their analysis to test whether a non-trivial subset of MS risk variants exerts functional effects in central nervous system (CNS) parenchymal cells, including neurons and glia, thereby potentially influencing the earliest events that predispose the CNS to immune-mediated injury.
Study Architecture and Statistical Yield From an Expanded GWAS
The authors assemble an expanded European-ancestry meta-analysis by harmonizing three large biobank/EHR-linked cohorts (UK Biobank, eMERGE-III, All of Us) and then integrating these results with prior MS GWAS efforts, excluding the extended MHC region to focus on non-MHC effects. In the resulting European-ancestry analysis (19,865 cases and 623,043 controls), they report 236 independent non-MHC susceptibility variants, including 38 additional variants relative to their earlier catalog and four loci flagged as not previously associated with MS; Table 1 lists lead variants at these newly highlighted loci (rs6975311, rs10758669, rs77267834, rs3935549; marked with asterisks in the table spanning pages 5–7). This expanded signal set provides the statistical substrate for downstream functional mapping, particularly in cell types that are typically underrepresented in QTL reference resources.
Multi-ancestry Signals and the Context of Autoimmune Pleiotropy
To address transferability across ancestries, the study also evaluates African-American (AFR) and Admixed American (AMR) participants drawn from the same biobanks, recognizing limited power relative to European-ancestry GWAS. Notwithstanding this constraint, the authors report one genome-wide significant locus in AFR (rs76911648) and two in AMR (rs59061674 and rs113284638), while emphasizing that replication is not currently feasible due to data scarcity and that these findings should be interpreted cautiously. In parallel, the authors situate MS susceptibility within a broader autoimmune landscape by comparing MS loci to GWAS of 12 autoimmune diseases and additional neurodegenerative, psychiatric, and metabolic traits; consistent with prior observations, many MS loci show sharing with other autoimmune conditions, but with a meaningful fraction of directionally discordant effects, underscoring that “autoimmunity” is not a single genetic axis but a phenotype-specific configuration of immune pathway perturbations.
A Genome-wide Polygenic Score With Cross-ancestry Portability
A major translational deliverable is a genome-wide polygenic score (GPS) built using PRS-CSx, explicitly incorporating AFR, EUR, and AMR summary statistics and accounting for ancestry-specific linkage disequilibrium. In the independent eMERGE-III testing cohort, the GPS is strongly associated with MS in European-ancestry participants (odds ratio 1.70 per standard deviation), and individuals in the top 1% of the score distribution show more than a seven-fold increased risk relative to the remaining 99% (Table 2, page 11). The GPS also shows statistically significant—though attenuated—associations in AFR (OR 1.26 per SD) and AMR (OR 1.46 per SD), illustrating partial portability while simultaneously highlighting the practical consequence of underpowered non-European discovery data: performance disparities that are likely to persist until substantially larger diverse cohorts become available.
Cell-type-resolved Colocalization Identifies Inhibitory Neurons as a CNS Hotspot
The most conceptually disruptive result emerges from integrating the MS GWAS with single-cell/single-nucleus eQTL resources from peripheral blood mononuclear cells (PBMCs) and from dorsolateral prefrontal cortex (DLPFC) tissue, using COLOC to prioritize loci where the same variant plausibly drives both disease association and gene expression effects in a specific cell type. As expected, blood-derived immune cells—particularly CD4⁺ naïve T cells—show prominent colocalization, consistent with prior MS genetics. However, among CNS cell types, inhibitory neurons exhibit the largest number of colocalized MS eQTL signals (reported as 15), exceeding microglia (reported as 6) despite the neuroimmunology-centric framing that often privileges microglial mechanisms; moreover, several of these neuronal colocalizations appear unique to inhibitory neurons relative to both other CNS cell types and PBMC populations. This pattern supports a model in which inherited susceptibility includes cell-intrinsic neuronal expression programs that may shape vulnerability, immune recognition, or response-to-inflammation in ways that precede—or amplify—classical immune effector phases.
The IL7–STAT3 Axis as a Neuron-specific Bridge Between Immunity and CNS Targeting
Two loci provide a particularly instructive mechanistic example: IL7 and STAT3, canonical immune signaling components that the authors find to have susceptibility-expression colocalization specifically in inhibitory neurons, not in the surveyed blood cell types. The paper highlights these loci as a plausible bridge between peripheral autoimmune propensity and CNS targeting: IL7-driven signaling is known to intersect with STAT3 biology, and a neuron-restricted regulatory effect at these genes could, in principle, reconfigure how inhibitory neurons respond to inflammatory cues or interact with immune cells. The authors further strengthen the STAT3 finding through replication in additional single-nucleus datasets and provide protein-level support by immunofluorescence in postmortem MS DLPFC tissue, reporting STAT3 expression in GAD1⁺/GAD2⁺ inhibitory neurons and noting a subset with elevated STAT3 signal (while not observing overt morphological differences in the assessed features). Together, these layers of evidence elevate the neuronal signal beyond a statistical curiosity and toward a testable neuroimmune pathway hypothesis.
Implications, Limitations, and the Next Experimental Agenda
If inhibitory neurons and select glial populations are genuine loci of causal regulatory perturbation in MS susceptibility, then MS etiology should be conceptualized as a coupled system: immune dysregulation that enables autoreactivity, plus CNS-intrinsic variation that governs how neural cells present, resist, or propagate inflammatory injury and neurodegeneration. This reframing has practical consequences for prevention and early intervention, because it encourages investigation of CNS-targeted protective strategies alongside immune modulation, and it motivates mechanistic experiments that interrogate neuron–T cell and neuron–microglia interactions under genetically defined perturbations. At the same time, key constraints remain: non-European GWAS signals are underpowered and require future replication; COLOC’s single-causal-variant assumption can mislead at complex loci; and the DLPFC reference, while well-powered, represents a specific brain region and sampling context that may not capture disease-relevant states such as active demyelinating lesions or spinal cord pathology. The central contribution of this work is therefore not simply an expanded variant catalog, but a prioritized, cell-type-specific hypothesis set—especially in inhibitory neurons—that can now be pursued with targeted functional genomics, perturbation experiments, and longitudinal human tissue profiling.
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:
De Jager, P., Zeng, L., Khan, A., Lama, T., Chitnis, T., Weiner, H., ... & Kiryluk, K. (2025). GWAS highlights the neuronal contribution to multiple sclerosis susceptibility. Research Square, rs-3.
