AI-identified peptide reduced ulcerative colitis symptoms in mice

Researchers used machine learning to identify a new antimicrobial peptide that reduced inflammation, restored gut barrier function and reshaped the microbiome in ulcerative colitis. In a study from First Hospital of Jilin University, a peptide named LR showed the most favourable balance between antibacterial activity and low cytotoxicity and produced stronger effects than 5-aminosalicylic acid and ciprofloxacin in a mouse model of colitis.

The study developed a machine-learning pipeline combining peptide prediction models with genetic algorithms to analyse structural and physicochemical properties of more than 6,000 potential candidates. The study highlighted 22 sequences that looked promising, and five peptides were synthesised for laboratory testing.

In vitro experiments showed that LR had strong bactericidal effects against pathogenic bacteria including Escherichia coli and Staphylococcus aureus, while maintaining good biocompatibility, with minimal toxicity and low haemolytic activity compared with other candidates.

To test therapeutic potential, LR was administered to mice with dextran sulphate sodium-induced colitis. Treatment led to reduced body weight loss, improved disease activity index and less colon shortening, while histological analysis showed reduced mucosal damage and decreased infiltration of inflammatory cells.

Further analyses found that LR suppressed inflammatory responses, with markedly reduced levels of tumour necrosis factor-α and interleukin-6 after treatment. The peptide also increased expression of the tight junction proteins ZO-1, claudin-1 and occludin, suggesting improved epithelial barrier function.

Sequencing of faecal microbiota showed that treatment reshaped microbial composition in mice with colitis. The abundance of Akkermansia muciniphila increased significantly, and supplementation with A. muciniphila alone partially alleviated colitis symptoms, suggesting that microbiota modulation contributes to LR’s therapeutic effect.

The findings show that integrating computational screening with experimental validation can identify stable and selective antimicrobial peptides with anti-inflammatory activity in ulcerative colitis. The study said more research is needed to assess long-term safety and applicability to humans.