Background High-fat diet (HFD) is a recognized risk factor for colorectal cancer, while its incidence isn’t directly proportional to fat intake. Healthy individuals show varying responses to the carcinogenic effects of HFD. Although gut microbiota are known to affect colorectal cancer (CRC) development, their involvement in individual susceptibility to HFD-induced CRC remains uncertain.

Methods Apc ^Min/+ mice were fed a high-fat diet to identify individuals insensitive to its carcinogenic effects. Fecal microbiota transplantation experiments demonstrated that resistance to the carcinogenic effects of a high-fat diet depends on the gut microbiota. Differentiated bacteria in insensitive mice were identified using 16S rDNA sequencing, and their ability to counteract HFD-related carcinogenesis was assessed. Differentiated metabolites were sought through fecal metabolomics, and the association between key bacteria and metabolites was determined via bacterial genome sequencing. Organoid experiments were conducted to validate the effects of key bacteria and metabolites on tumor organoids and elucidate their related mechanisms.

Results Apc ^Min/+ mice exhibited varying tumor incidences after HFD, with some displaying significantly fewer tumors, denoted as HFD-insensitive. Recipient mice transplanted with HFD-insensitive microbiota also counteracted HFD-induced carcinogenesis. 16S rDNA sequencing revealed increased Lactobacillus abundance in HFD-insensitive mice. Among several Lactobacillus strains, Lactobacillus johnsonii (L. johnsonii) showed the strongest resistance to HFD-induced carcinogenesis. But only live L. johnsonii bacteria demonstrated the anti-tumor efficacy, highlighting the importance of bacterial metabolism. Metabolomic analysis identified elevated chenodeoxycholic acid levels in HFD-insensitive mice, while L. johnsonii encoded bile salt hydrolase (BSH) capable of chenodeoxycholic acid formation. Inhibition of BSH weakened its anti-tumor effect. Engineered bacteria expressing this enzyme suppressed HFD-induced carcinogenesis. In CRC patient tumor organoids, L. johnsonii inhibited proliferation and induced apoptosis via chenodeoxycholic acid, associated with elevated ROS levels and mitochondrial dysfunction.

Conclusions The sensitivity to HFD-induced carcinogenesis correlates with gut microbiota composition. Individuals with abundant Lactobacillus are often less sensitive to HFD. Lactobacillus converts primary bile acids to chenodeoxycholic acid via bile salt hydrolase, inhibiting tumor formation. Mechanistically, chenodeoxycholic acid may induce ROS production, leading to mitochondrial dysfunction, inhibiting tumor growth, and promoting apoptosis.