We tested whether the attenuation of experimental colitis by live probiotic bacteria is due to their immunostimulatory DNA, whether toll-like receptor (TLR) signaling is required, and whether nonviable probiotics are effective.

Methylated and unmethylated genomic DNA isolated from probiotics (VSL-3), DNAse-treated probiotics and Escherichia coli (DH5α) genomic DNA were administered intragastricly (i.g.) or subcutaneously (sc) to mice prior to the induction of colitis. Viable or γ-irradiated probiotics were administered i.g. to wild-type mice and mice deficient in different TLR or in the adaptor protein MyD88, 10 days prior to administration of dextran sodium sulfate (DSS) to their drinking water and for 7 days thereafter.

ntragastric and sc administration of probiotic and E. coli DNA ameliorated the severity of DSS-induced colitis, whereas methylated probiotic DNA, calf thymus DNA, and DNase-treated probiotics had no effect. The colitis severity was attenuated to the same extent by i.g. delivery of nonviable γ-irradiated or viable probiotics. Mice deficient in MyD88 did not respond to γ-irradiated probiotics. The severity of DSS-induced colitis in TLR2 and TLR4 deficient mice was significantly decreased by i.g. administration of γ-irradiated probiotics, whereas, in TLR9-deficient mice, γ-irradiated probiotics had no effect.

The protective effects of probiotics are mediated by their own DNA rather than by their metabolites or ability to colonize the colon. TLR9 signaling is essential in mediating the anti-inflammatory effect of probiotics, and live microorganisms are not required to attenuate experimental colitis because nonviable probiotics are equally effective.