Experimental evidence has implicated genotoxic Escherichia coli (E. coli) and enterotoxigenic Bacteroides fragilis (ETBF) in the development of colorectal cancer (CRC). However, evidence from epidemiological studies is sparse. We therefore assessed the association of serological markers of E. coli and ETBF exposure with odds of developing CRC in the European Prospective Investigation into Nutrition and Cancer (EPIC) study.Serum samples of incident CRC cases and matched controls (n = 442 pairs) were analyzed for immunoglobulin (Ig) A and G antibody responses to seven E. coli proteins and two isoforms of the ETBF toxin via multiplex serology. Multivariable-adjusted conditional logistic regression analyses were used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for the association of sero-positivity to E. coli and ETBF with CRC.The IgA-positivity of any of the tested E. coli antigens was associated with higher odds of developing CRC (OR: 1.42; 95% CI: 1.05-1.91). Dual-positivity for both IgA and IgG to E. coli and ETBF was associated with >1.7-fold higher odds of developing CRC, with a significant association only for IgG (OR: 1.75; 95% CI: 1.04, 2.94). This association was more pronounced when restricted to the proximal colon cancers (OR: 2.62; 95% CI: 1.09, 6.29) compared to those of the distal colon (OR: 1.24; 95% CI: 0.51, 3.00) (pheterogeneity = 0.095). Sero-positivity to E. coli and ETBF was associated with CRC development, suggesting that co-infection of these bacterial species may contribute to colorectal carcinogenesis. These findings warrant further exploration in larger prospective studies and within different population groups.
The nematode Caenorhabditis elegans is hypersusceptible to Burkholderia pseudomallei infection. However, the virulence mechanisms underlying rapid lethality of C. elegans upon B. pseudomallei infection remain poorly defined. To probe the host-pathogen interaction, we constructed GFP-tagged B. pseudomallei and followed bacterial accumulation within the C. elegans intestinal lumen. Contrary to slow-killing by most bacterial pathogens, B. pseudomallei caused fairly limited intestinal lumen colonization throughout the period of observation. Using grinder-defective mutant worms that allow the entry of intact bacteria also did not result in full intestinal lumen colonization. In addition, we observed a significant decline in C. elegans defecation and pharyngeal pumping rates upon B. pseudomallei infection. The decline in defecation rates ruled out the contribution of defecation to the limited B. pseudomallei colonization. We also demonstrated that the limited intestinal lumen colonization was not attributed to slowed host feeding as bacterial loads did not change significantly when feeding was stimulated by exogenous serotonin. Both these observations confirm that B. pseudomallei is a poor colonizer of the C. elegans intestine. To explore the possibility of toxin-mediated killing, we examined the transcription of the C. elegans ABC transporter gene, pgp-5, upon B. pseudomallei infection of the ppgp-5::gfp reporter strain. Expression of pgp-5 was highly induced, notably in the pharynx and intestine, compared with Escherichia coli-fed worms, suggesting that the host actively thwarted the pathogenic assaults during infection. Collectively, our findings propose that B. pseudomallei specifically and continuously secretes toxins to overcome C. elegans immune responses.