The signatures and crosstalk of gut mycobiome, microbiome, and metabolites in drug-induced liver injury

This study is led by Prof. Huikuan Chu (Department of Gastroenterology, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology) and Prof. Ling Yang (Department of Gastroenterology, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology). Dr. Weiyan Huang, Master's student Yirui Hu, and Master's student Zexuan Li completed the sequencing analysis and machine learning modeling work on fecal samples from 27 patients with traditional Chinese medicine (TCM)-associated drug-induced liver injury (DILI) and 30 healthy controls. They found that the gut fungi of patients with TCM-associated DILI exhibited significant dysbiosis, characterized by an expansion of potentially pathogenic flora (such as Candida and Aspergillus) and a reduction in beneficial flora (such as Saccharomyces and Cutaneotrichosporon). "This is the first systematic analysis of gut fungal characteristics in patients with TCM-associated DILI," said Prof. Huikuan Chu.

To investigate specific fungal biomarkers with diagnostic potential, the research team used LEfSe analysis to identify differentially abundant fungal species. Combined with the random forest model, the potential of A. jensenii and Cu. curvatum as specific diagnostic biomarker for DILI was determined. These fungal biomarkers were positively correlated with liver injury markers (e.g., TB, AST, ALT, ALP, GGT), suggesting that the gut fungal composition is closely associated with the severity of liver damage. These biomarkers provide a novel tool for non-invasive early diagnosis and intervention.

The research team further investigated the role of fungal-bacterial interactions in DILI. The results showed that the diagnostic model established by combining the optimal fungal biomarkers (A. jensenii and Cu. curvatum) with bacteria (Lactobacillus gasseri, Faecalibacterium sp., Mediterraneibacter gnavus, and Escherichia coli) performed significantly better than the fungal-only model. In the stratification of disease severity, the fungal-bacterial combined diagnosis also exhibited excellent performance. Dr. Weiyan Huang stated: “Fungi and bacteria do not change in isolation. Their interactions collectively shape the microecological characteristics of DILI, and combined analysis can significantly improve diagnostic performance.”

The study also examined the potential interaction between intestinal fungi and serum metabolites. KEGG enrichment analysis revealed that serum metabolites in DILI patients were significantly altered in liver metabolism and inflammation-related pathways, such as primary bile acid biosynthesis, arachidonic acid metabolism, cholesterol metabolism, and the PPAR signaling pathway. The combined diagnostic model based on fungi and metabolites achieved near-perfect diagnostic accuracy.

Prof. Huikuan Chu concluded: "This study confirms that gut fungal features can serve as specific non-invasive biomarkers for DILI. Integrating fungal with bacterial and metabolomic data can further enhance diagnostic accuracy, opening new avenues for early intervention and personalized treatment of DILI."

Currently, DILI still lacks specific non-invasive diagnostic methods. The study by Professor Chuhuikuan's team first revealed that patients with DILI have significant intestinal fungal dysbiosis, and the diagnostic accuracy of key fungal species exceeds 90%. By integrating fungal, bacterial, and metabolite biomarkers, the combined model can effectively diagnose DILI and identify the severity of DILI, providing a new strategy for non-invasive and accurate diagnosis and early intervention of DILI.

See the article:

The signatures and crosstalk of gut mycobiome, microbiome, and metabolites in drug-induced liver injury