When Genes Meet the Outside World: Rethinking Multiple Sclerosis in the Post-GWAS Era
Multiple sclerosis (MS) is a chronic inflammatory disease of the central nervous system, and its cause remains stubbornly complex. This review argues that the biggest payoff from genome-wide association studies (GWAS) is no longer just “finding more loci,” but using GWAS data as a scaffold to build etiologic models that explicitly incorporate nongenetic drivers—environmental exposures, lifestyle, and other nonheritable influences—because twin and population data imply that these forces materially shape MS risk and vary across settings.
What classic GWAS delivered—and why it is still not enough
Over the past decade, GWAS has identified more than 200 MS-associated loci, showing that MS genetic risk is highly polygenic: many common variants with small effects scattered across the genome, with notable weight in the HLA region. The review highlights why this creates a translation bottleneck: association signals rarely tell you which gene is causal, what cell type matters, or what pathway is being perturbed. Still, “mechanism-first” successes exist (for example, variants affecting immune receptors or splicing; and a clinically important cautionary story where a TNF receptor–region variant mirrored the risk-increasing effect of anti-TNF drugs on demyelinating disease), illustrating how careful functional follow-up can convert statistical signals into actionable biology.
Reworking GWAS: looking beyond only genome-wide significant hits
A central theme is that strict genome-wide thresholds (e.g., p < 5×10⁻⁸) are excellent for discovery, but can be overly conservative for understanding biology in a disease driven by thousands of modest effects. The paper summarizes “GWAS reworking” strategies that intentionally use looser cutoffs (often p < 0.05) and then test whether implicated genes converge in pathways or protein–protein interaction (PPI) networks. As compiled in Table 1 (page 3), approaches such as pathway enrichment plus PPI networks and network-based pathway analysis (PINBPA) helped surface coherent immune and neural subnetworks and suggested additional susceptibility loci that were later reinforced by larger datasets—illustrating how aggregation can rescue signal that single-variant testing underpowers.
Adding the missing “regulatory layer”: epigenomics and cell-type specificity
Many MS-associated variants fall in noncoding regions, implying regulatory—not protein-coding—mechanisms. The review emphasizes how ENCODE/Roadmap-style annotations enable a practical next step: ask where in the body and immune system these variants likely act by mapping them onto cell-type-specific regulatory elements. A highlighted systems-biology framework (using large case-control cohorts) builds cell-specific networks from susceptibility variants and regulatory annotations, and points to a dual story: peripheral immune cells remain central, but CNS-resident microglia emerge as key contributors to susceptibility as well—an important shift from purely “outside-in” immune narratives.
From variant to consequence: eQTLs that can depend on disease context
Because regulatory variants often act through gene expression, the review discusses expression quantitative trait loci (eQTL) analyses as a bridge from genotype to mechanism. A crucial nuance is that eQTL effects may not be identical in healthy and diseased states: one study described in the review found multiple eQTL associations in blood immune cells where a substantial fraction appeared more pronounced in MS patients than in noninflammatory neurological controls. The article also walks through a concrete example at CD58, where an MS-associated SNP within a microRNA stem-loop may alter Drosha processing, with downstream effects on CD58 and microRNA expression—an illustration of how “noncoding” can still be mechanistically specific.
A more “human” model: genes interacting with the environment via the interactome
Where the review becomes most distinctive is its interactome-based proposal: instead of treating environment as an add-on, model it through known physical interactions between environmental factors (notably viruses) and human proteins. The authors define interactomes as sets of host genes whose protein products directly interact with a given exposure, then test whether MS-associated GWAS signals are enriched near those genes (they operationally used a 20 kbp window to map variants to nearby genes). In the “candidate-interactome” analysis, viral interactomes—especially Epstein–Barr virus (EBV)—stand out, supporting EBV as a major environmental risk factor acting in concert with host genetics. The workflow is visualized in Figure 1 (page 6): start from polygenic predisposition, intersect with exposure-specific protein interaction modules, and then interrogate pathways enriched in the resulting gene set.
The next step: polygenic risk scores that finally include environment (and maybe the virus itself)
The review closes with a forward-looking but cautious argument: polygenic risk scores (PRS) could eventually help quantify an individual’s inherited burden and relate it to endophenotypes (imaging or fluid biomarkers), but PRS alone will be incomplete if MS risk is fundamentally G×E. The authors propose integrating PRS-style genetics with environmental measurements and even incorporating EBV genomic variation alongside host loci—conceptually moving toward calibrated models that include genetics (G), environment (E), and interaction (G×E). They are explicit that this is not yet clinically “within reach” due to challenges like pleiotropy and causal inference, but the direction is clear: the post-GWAS era in MS will be defined by synthesis—networks, regulation, expression, and real-world exposures—rather than ever-longer lists of loci.
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:
Mechelli, R., Umeton, R., Manfrè, G., Romano, S., Buscarinu, M. C., Rinaldi, V., ... & Ristori, G. (2020). Reworking GWAS data to understand the role of nongenetic factors in MS etiopathogenesis. Genes, 11(1), 97.
