In recent years, the microbiome has gained increased attention for its effects on the immune system, and it is well known to influence the CD4⁺ T cell compartment; however, little is known of its effect on CD8⁺ T cells. In a study recently published in Nature, Tanoue and Morita et al. show that the microbiome plays a role in local and systemic accumulation of IFNγ⁺CD8⁺ T cells. They identified a consortium of 11 strains of bacteria from the human microbiome that maximize this effect and explored the potential to utilize this microbiome-driven influence on CD8⁺ T cells to enhance immunity.

Kicking off their research, Tanoue and Morita et al. found that in specific pathogen-free (SPF) mice, IFNγ⁺CD8⁺ T cells were constitutively present at high frequencies in the intestinal lamina propria and that many exhibited an activated or a memory phenotype. Comparatively, IFNγ⁺CD8⁺ T cells were much less frequent in germ-free mice or in SPF mice treated with an antibiotic cocktail, indicating a relationship between the presence and/or composition of the microbiome and IFNγ⁺CD8⁺ T cell accumulation.

To identify specific bacterial strains that influence IFNγ⁺CD8⁺ T cell accumulation in the gut, the researchers took fecal samples from six healthy human volunteers and transplanted them into germ-free mice. IFNγ⁺CD8⁺ T cell induction varied by sample, and researchers selected the mouse with the strongest induction for further analysis. The team isolated and analyzed 206 bacterial colonies, identifying a total of 26 strains, which together constituted most of the strains present. Of these, 11 strains were found to be positively associated with IFNγ⁺CD8⁺ T cell frequency. Inoculating germ-free mice with a mixture of these 11 strains (11-mix) robustly induced accumulation of IFNγ⁺CD8⁺ T cells without inducing any obvious signs of inflammation or toxicity. Of the 11 strains, 7 strains were bacteroidales and 4 were not; when administered as separate groups, the mix of 4 non-bacteroidales effectively induced IFNγ⁺CD8⁺ T cells (though not as strongly as the full 11-mix), while the mix of 7 bacteroidales failed to induce IFNγ⁺CD8⁺ T cells. This indicated that the non-bacteroidales play an active effector role, while the bacteroidales play a supporting role, together forming an effective consortium.

Exploring the underlying mechanism by which the consortium of 11 bacterial strains induce IFNγ⁺CD8⁺ T cell accumulation, Tanoue and Morita et al. found that the strains enter the colonic mucus layer and induce nearby colonic epithelial cells to express chemokines and IFNγ-inducible genes, likely mediating the recruitment and accumulation of IFNγ⁺CD8⁺ T cells. IFNγ⁺CD8⁺ T cells were also found to be actively proliferating, and many recognized bacterial antigens derived from the 11 strains, suggesting accumulation may also be enhanced by antigen-mediated differentiation and expansion. Furthermore, the accumulation of T cells was found to be dependent on CD103⁺ dendritic cells, and on enhanced expression of MHC-I on dendritic cells in the context of 11-mix colonization. While local accumulation of IFNγ⁺CD8⁺ T cells was most notable, accumulation was also observed in several other organs (but interestingly, not in peripheral blood), indicating a systemic effect. Examination of caecal contents from transplanted mice revealed differences in overall metabolomic profiles, and overlap between metabolites in the caecal contents and sera, suggesting that the systemic effect may be related to circulating metabolites. Molecules including mevalonate and dimethylglycine were found in both caecal contents and sera suggesting that these may play a role in the systemic effect of the consortium.

Given that the consortium of gut bacteria augments the accumulation of IFNγ⁺CD8⁺ T cells, the researchers wondered whether this effect influenced effective immune function. SPF or germ-free mice treated with the 11-mix showed better protective immunity against both wild-type and invasive mutant Listeria monocytogenes, clearing infection faster, reducing colon damage, and protecting against dissemination compared to untreated SPF or germ-free mice. Treatment was shown to induce antigen-specific IFNγ⁺CD8⁺ T cells, and the anti-microbial immunity was dependent on CD8⁺ T cells.

To determine the impact on antitumor immunity, the researchers treated germ-free or SPF mice with 11-mix, then injected tumor cells. The consortium of 11 bacterial strains enhanced both spontaneous host antitumor responses and the efficacy of checkpoint blockade (anti-PD-1 or anti-CTLA-4). This effect was dependent on CD8⁺ T cells that were phenotypically distinct from those produced in the colon. In the context of checkpoint blockade, much like in previous models, the researchers observed an increased frequency of IFNγ⁺CD8⁺ TILs (including tumor antigen-specific IFNγ⁺CD8⁺ T cells), increased co-expression of granzyme B, and an increase in tumor-infiltrating dendritic cells expressing high levels of MHC-I.

Using metagenomic human data sets, Tanoue and Morita et al. found that the 11 strains identified in their consortium are typically rare, low-abundance components of the human microbiome. Because of this, the researchers hope that the consortium may serve as a broadly applicable biotherapeutic in humans to enhance antimicrobial or antitumor immunity.

by Lauren Hitchings