Computational Network Study Highlights CD44 as a Potential MS Target

A new computational study is drawing attention to CD44, a cell-surface receptor long associated with immune cell trafficking, as a potentially important therapeutic target in multiple sclerosis. By combining systems biology with molecular docking and molecular dynamics simulations, researchers suggest that CD44 occupies a highly central position within molecular interaction networks linked to neuroinflammation—and that existing drugs may structurally interact with the receptor in silico.

Multiple sclerosis is marked by striking biological and clinical heterogeneity, complicating efforts to identify durable biomarkers and treatment targets. In this study, investigators analyzed gene-expression data from human gray matter tissue, comparing samples from patients with MS to healthy controls. While more than a thousand genes were differentially expressed, network analysis revealed a smaller group of highly connected “hub” genes. CD44 emerged as the most interconnected node, indicating a prominent position within inflammatory interaction networks, though such centrality does not by itself establish a causal regulatory role in disease.

That positioning aligns with CD44’s established biology. The receptor plays a role in leukocyte adhesion, migration into the central nervous system, and hyaluronan-mediated signaling—processes implicated in immune cell infiltration and demyelinating pathology in MS. Elevated CD44 expression has been associated with inflammatory activation states in gene-expression datasets, supporting its relevance to disease-associated immune processes rather than serving as a direct marker of clinical survival outcomes.

The researchers then explored whether CD44 might be structurally tractable as a drug target. Screening known and investigational compounds identified several molecules predicted to bind the CD44 hyaluronan-binding domain in silico. Among them, obeticholic acid demonstrated the most stable and energetically favorable interaction during 100-nanosecond molecular dynamics simulations. Other candidates, including chlordiazepoxide and dextromethorphan, showed comparatively less stable binding, while hyaluronic acid—the receptor’s endogenous ligand—exhibited greater variability consistent with its known size, flexibility, and polymeric properties.

The authors emphasize that all findings are computational and require experimental validation. Rather than pointing to an immediate therapeutic shift or drug-repurposing opportunity, the study highlights how integrated network biology and structural modeling can help prioritize biologically plausible targets for further investigation as MS research moves toward more precise, mechanism-driven interventions.