Modulating the immune response to the live oral vaccine vector Streptococcus gordonii with the use of exogenous TNF.

摘要

The commensal bacterium Streptococcus gordonii has recently been investigated as a mucosal vaccine delivery vector. Although proficient at colonizing murine oral mucosa, S. gordonii strains often fail to elicit significant antibody titers against their vaccine antigen payloads. We hypothesized that this poor response may be due to an inability of S. gordonii to elicit cytokines needed to suppress mucosal tolerance, and that exogenously supplied cytokines such as TNF could overcome this effect. To test this, murine bone marrow-derived dendritic cells (BM-DCs) were stimulated with UV-killed S. gordonii PM14, a vaccine strain which surface expresses a fragment of the immunodominant S1 subunit of pertussis toxin. Peptidoglycan (PGN), lipoteichoic acid (LTA), lipoprotein (LP), and DNA were also purified from the bacteria, and used to stimulate BM-DCs. Twenty four-hour stimulations were carried out with the agents alone, or in combination with exogenous TNF. Following stimulation, DC expression of surface markers was analyzed by flow cytometry, and supernatant cytokines quantified. Stimulation with TNF, PM14, PGN, LTA, or LP all resulted in increased surface expression of MHCII, CD80, and CD86, compared to unstimulated DCs. Stimulation with PM14 elicited TNF, IL-6, IL-10, and IL-12p70 production from the DCs, while stimulation with the bacterial components induced some or all of the four cytokines. When DCs were simultaneously stimulated with PM14 and TNF, an additive increase in surface marker upregulation was observed, and the two stimuli synergized to elicit substantially greater quantities of IL-6, IL-10, and IL-12p70. Synergy between TNF and the purified bacterial components was also observed. The priming effect of TNF was abolished when DCs were obtained from mice deficient for either TNFR1 or TNFR2, and cytokine induction by PM14 was entirely dependent on functional MyD88. Synergistic IL-10 induction by S. gordonii and TNF was not observed in TLR-2-/- BM-DCs, and TNF was found to cause TLR-2 upregulation, providing at least a partial mechanism for the observed synergy. When PM14 and TNF were used to immunize mice, splenocyte antigen-specific recall without a corresponding increase in antibody production was observed. Finally, an S. gordonii strain capable of secreting rTNF was successfully engineered. These findings demonstrate that TNF is able to prime BM-DCs to better respond to S. gordonii, through a mechanism at least partially involving TLR-2 upregulation. TNF did not conclusively improve the in vivo response to the bacteria, but future work can investigate this further using a TNF secreting strain. Ideally, the immonumodulatory effects of TNF could be exploited in future developments of S. gordonii mucosal vaccines.