We greatly appreciate the innovative work by Zheng et al. 1, which integrates Mendelian randomization (MR) and machine learning to identify pivotal plasma protein targets (e.g., GCKR, MSP, FCG2A) for inflammatory bowel disease (IBD). This methodological approach aligns with recent advancements in IBD research that leverage multi-omics integration to uncover causal relationships between plasma proteins and disease risk. Boasting a rigorous study design and robust data backing, this research offers valuable insights for advancing precision medicine in IBD, similar to other cutting-edge studies that have identified preclinical protein signatures predictive of IBD development up to 16 years before diagnosis. We would like to respectfully raise two points that we feel merit further clarification, with the goal of enhancing the translational potential and mechanistic depth of this important work. To begin with, the disease-specific pharmacological compatibility of “drug repurposing” for these identified targets has yet to be fully elucidated. The study highlights MSP (targeted by fostamatinib and crizotinib) and FCG2A (targeted by cetuximab and abciximab) as core therapeutic candidates, though these agents are currently indicated for non-IBD conditions (e.g., autoimmune disorders, malignancies, and acute myocardial infarction) 1. IBD exhibits distinct mucosal pathological features (e.g., mucosal macrophage infiltration, dysregulated cytokine networks) that differ fundamentally from those of systemic immune diseases or tumors, which is a critical consideration for drug repurposing strategies 2. For example, Fostamatinib—a SYK inhibitor that exerts anti-inflammatory effects in autoimmune diseases—raises questions about whether intestinal epithelial barrier dysfunction might compromise its mucosal penetration and therapeutic efficacy in IBD 3. A recent study by Wang et al. demonstrated that SYK inhibition by Fostamatinib modulates neutrophil immune-responses via the mTOR/RUBCNL-dependent autophagy pathway to alleviate intestinal inflammation in ulcerative colitis, though mucosal bioavailability remained a concern 4. Additionally, the intestinal pharmacokinetic profiles of these agents have not been evaluated, which could pose challenges for their direct repurposing, as emphasized in a comprehensive review on peptide-based therapeutics for IBD that highlighted delivery and bioavailability as major obstacles. We suggest that supplementary preclinical assessments of their intestinal-specific bioactivity and safety in IBD animal models would be highly valuable to strengthen the pharmacological evidence base, as recommended by recent guidelines on drug development for gastrointestinal diseases 5. Second, the temporal ordering and cell-type specificity of the epigenetic mediating mechanism call for further clarification. The study identifies 10 DNA methylation loci that indirectly modulate plasma proteins and IBD susceptibility, yet the causal timing of the “methylation-protein-IBD” axis and the critical cell populations involved remain unclear 1. Notably, the cg03555424 locus—located in the intronic region of APEH—regulates MSP expression, but it remains unclear if aberrant methylation of cg03555424 precedes perturbations in MSP levels and the subsequent onset of intestinal inflammation. Moreover, it requires clarification whether this methylation locus is enriched in cell types relevant to IBD pathogenesis (e.g., intestinal epithelial cells, lamina propria macrophages), as recent single-cell multi-omic studies have revealed that DNA methylation patterns are highly cell-type specific and can profoundly influence gene expression in disease-relevant cell populations 2. We believe that addressing these points could help consolidate the “genetic-epigenetic-protein” causal chain, which has been shown to mediate genetic risk in IBD through integrative epigenome-wide analyses, and inform the development of cell-type-specific intervention strategies. In summary, Zheng et al.'s study lays a solid foundation for understanding the molecular mechanisms underlying IBD, complementing recent proteome-wide MR and colocalization analyses that have identified additional therapeutic targets for gastrointestinal diseases 1, 5. These observations are not meant to diminish the study's significance; rather, they aim to provide constructive insights to guide future research. We anticipate that further exploration of these aspects will enhance the translational impact of this valuable work, potentially leading to novel precision medicine approaches for IBD similar to those being developed for other complex diseases using integrated MR and machine learning methodologies. This study was supported by Natural Science Foundation of Guangdong Province of China (2025A1515011990). The authors declare no conflicts of interest. The data that support the findings of this study are available from the corresponding author upon reasonable request.
Yang et al. (Fri,) studied this question.