A comprehensive scientific review has revealed that certain gut bacteria are not merely associated with colorectal cancer (CRC) but actively contribute to its development by damaging DNA, altering gene regulation, and suppressing the body's immune defenses. The findings provide new insights into how the gut microbiome influences cancer progression and could pave the way for precision microbiome-based therapies.
The review, conducted by researchers from the Institute of Digestive Disease at The Chinese University of Hong Kong, highlights how advances in multi-omics technologies—including genomics, transcriptomics, epigenomics, and metabolomics—are transforming scientists' understanding of the complex interactions between gut microbes and human cells.
One of the strongest examples involves Escherichia coli strains carrying the pks genomic island, which produce colibactin, a toxin capable of causing DNA damage. Researchers note that the characteristic DNA mutation pattern produced by colibactin has been identified in more than 12% of colorectal cancer cases, providing direct molecular evidence that bacterial toxins can initiate cancer-causing mutations.
The review also identifies Fusobacterium nucleatum as another important cancer-promoting bacterium. Its virulence protein, FadA, binds to the E-cadherin receptor on intestinal cells, activating the Wnt/β-catenin signaling pathway, a key driver of uncontrolled cell growth and tumor development.
Beyond bacterial toxins, microbial metabolites also play a significant role. Secondary bile acids, particularly deoxycholic acid (DCA), produced by specific gut bacteria, were found to suppress cancer-fighting CD8⁺ T cells, allowing tumors to escape immune surveillance and continue growing.
Despite rapid growth in microbiome research, the authors emphasize that many previous studies have focused only on identifying bacterial species present in patients rather than understanding their biological functions. Microbiome datasets are highly complex due to their compositional nature, data sparsity, and thousands of microbial variables, often resulting in misleading correlations.
To address these challenges, researchers highlight the growing use of artificial intelligence (AI) and advanced machine-learning techniques such as Random Forest algorithms and deep-learning models like MetaNN. These computational approaches help distinguish genuine biological interactions from statistical noise.
Emerging technologies are also expanding the field. Long-read sequencing platforms, including PacBio Single-Molecule Real-Time (SMRT) sequencing and Oxford Nanopore Technologies, together with bacterial single-cell spatial transcriptomics, now allow scientists to precisely identify where individual bacteria reside within tumors and how they communicate with neighboring human cells.
According to the authors, combining multi-omics technologies with AI is shifting microbiome research from simple association studies toward establishing direct cause-and-effect relationships. Experimental models such as organ-on-chip systems and gnotobiotic mice are expected to further validate these microbial mechanisms.
The findings have significant clinical implications. Microbiome signatures may eventually serve as non-invasive biomarkers for the early detection of colorectal cancer or help predict which patients are most likely to benefit from immunotherapy. Researchers also envision targeted elimination of harmful bacteria using bacteriophages, alongside engineering beneficial microbes to deliver anti-cancer therapies or restore intestinal barrier function.
Looking ahead, the authors propose the development of patient-specific "digital twins" that integrate multi-omics data to predict how diet, probiotics, prebiotics, or live biotherapeutics could reshape an individual's gut microbiome, bringing personalized colorectal cancer prevention and treatment closer to reality.
Reference : Lu Y, Yu J, et al. Cancer Biology & Medicine. 2025. DOI: 10.20892/j.issn.2095-3941.2025.0762
