Khorana, R. | Search bioRxiv

bioRxiv2026

Bacteroides intestinalis-Driven Arabinoxylan Fermentation Mitigates Inflammatory and Metabolic Dysfunction

Insufficient dietary fiber intake is closely linked to gut microbiome dysfunction and increased risk of noncommunicable diseases. Synergistic synbiotics, pairing defined microbes with their dietary substrates, offer a precise strategy to restore microbiome function. Here, we show that pairing the human colonic commensal Bacteroides intestinalis with insoluble wheat arabinoxylan (inWAX) yields pronounced metabolic and anti-inflammatory benefits. Using mono-associated gnotobiotic mice and high-fat diet-induced obese mice, we demonstrate that this synbiotic enhances resistance to intestinal inflammation and improves glucose homeostasis. Mechanistically, this synbiotic increases production of 6-hydroxylated bile acids with anti-diabetic and anti-steatotic effects. It also promotes microbial transformation of phenolic compounds and redirects protein metabolism toward neuroactive metabolites. These metabolic shifts are accompanied by transcriptional remodeling in the colon, spleen and liver, involving induction of genes essential for circadian rhythm, lipid metabolism, immune defense, and bile acid 6-hydroxylation. Notably, Y chromosome-linked genes associated with epigenetic regulation and protein turnover are also induced, suggesting a potential sex-specific response to synbiotics. Together, our findings establish a mechanistic framework for targeted synbiotic interventions to combat inflammatory and metabolic disorders.

microbiology↗

bioRxiv2025

Protein Language Models Capture Structural and Functional Epistasis in a Zero-Shot Setting

Protein language models (PLMs) learn from large collections of natural sequences and achieve striking success across prediction tasks, yet it remains unclear what biological principles underlie their representations. We use epistasis, the dependence of a mutations effect on its sequence context, as a lens to probe what PLMs capture about proteins. Comparing PLM-derived scores with deep mutational scanning data, we find that epistasis emerges naturally from pretrained models, without supervision on experimental fitness. Raw model scores align with residue-residue contacts, indicating that PLMs internalize structural proximity. Applying a nonlinear transformation to bring model outputs onto the experimental scale, however, shifts the signal toward functional couplings between distant sites. These findings show that PLMs capture both structural and functional dependencies from sequence data alone, and that epistasis provides a powerful window into the biological principles embedded in their representations.

bioinformatics↗