MS Code: How Mendelian Randomization Reveals Promising Drug Targets

Multiple sclerosis (MS) remains one of the most complex autoimmune diseases, affecting millions worldwide. Although over nine classes of disease-modifying therapies are currently available, they mostly target the inflammatory stage of MS and are particularly effective in relapsing–remitting MS. However, patients living with progressive forms of MS—where neurodegeneration behind an intact blood–brain barrier becomes the main driver—continue to face limited therapeutic options. As the authors emphasize in the introduction (page 2), the growing need for new, biologically grounded drug targets is urgent, and genetics-based approaches can provide a powerful path forward.

Using Mendelian Randomization to Find Causal Proteins
To pinpoint biologically meaningful drug targets, the researchers relied on Mendelian randomization (MR)—a method that uses genetic variants as “natural experiments” to infer causal relationships between proteins and disease. The team analyzed 734 plasma proteins and 154 cerebrospinal fluid (CSF) proteins, using large genetic datasets from the International Multiple Sclerosis Genetics Consortium, UK Biobank, and FinnGen.

As illustrated in the study’s design diagram on page 2, this multistep approach combined MR, bidirectional causality checks, Bayesian co-localization, and phenotype scanning. The goal was to find proteins whose genetically predicted levels truly influence MS risk—not just correlate with it.

Six Proteins Emerge as Causal Players in MS
From their proteome-wide screen, the authors identified six proteins significantly associated with MS risk at the stringent Bonferroni threshold.

Protective in plasma:

FCRL3 (OR = 0.83)

TYMP (OR = 0.59)

AHSG (OR = 0.88)

Causal in CSF:

MMEL1 (OR = 5.03; increases risk)

SLAMF7 (OR = 0.42; protective)

CD5L (OR = 0.30; protective)

These findings are visualized in Table 1 on page 4 and further supported by the forest plots shown on page 7, which summarize replication results. Interestingly, MMEL1 stood out as the only protein validated across both UK Biobank and FinnGen cohorts.

Strengthening the Evidence: Co-localization and Reverse Causality Tests
To ensure the six proteins truly act upstream of MS, the authors conducted extensive sensitivity analyses. As shown in Table 2 (page 6), none of the proteins showed evidence of reverse causality, meaning MS itself does not appear to influence their levels.

Furthermore, five of the six proteins (FCRL3, TYMP, AHSG, MMEL1, SLAMF7) displayed strong Bayesian co-localization (PPH4 > 0.8), indicating that the same genetic variants drive both the protein levels and MS susceptibility. This helps rule out chance associations or confounding due to nearby genetic signals. Phenotype scanning also revealed known links between some protein-associated SNPs and autoimmune or metabolic traits, adding biological plausibility.

Linking New Targets to Existing MS Treatments
One of the article’s most interesting insights comes from the protein–protein interaction (PPI) network shown on page 6 (Figure 3). The analysis revealed meaningful connections between the newly identified proteins and targets of approved MS therapies.

For example:

FCRL3 interacts with CD20 (MS4A1), the target of ocrelizumab and ofatumumab—two cornerstone B-cell–depleting treatments.

TYMP interacts with enzymes involved in DNA metabolism that are targets of cladribine, a therapy used for relapsing MS.

SLAMF7 interacts with CD52, the target of alemtuzumab.

Additionally, the study notes that existing drugs such as elotuzumab (targeting SLAMF7) and floxuridine (influencing TYMP) could potentially be explored for repurposing in MS.

What This Means for Future MS Drug Development
By combining large-scale proteomics with rigorous genetic analyses, this study provides a shortlist of five high-confidence drug targets: FCRL3, TYMP, AHSG, MMEL1, and SLAMF7. Among them, FCRL3 and SLAMF7 appear particularly promising, given their strong genetic support and existing druggable pathways. The findings also highlight the importance of studying CSF proteins—critical in progressive MS—alongside plasma proteins, as reflected in the contrasting signals across compartments.

While the authors acknowledge limitations such as measurement variation across protein studies and the predominantly European ancestry of participants (page 8), the results lay a foundation for future experimental and clinical research.

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:
Lin, J., Zhou, J., & Xu, Y. (2023). Potential drug targets for multiple sclerosis identified through Mendelian randomization analysis. Brain, 146(8), 3364-3372.