Bacteroides Thetaiotaomicron Induce Tregs In The Colon In A Capsule Independent Mechanism

It has become clear that there is a range in colonic Treg induction by different commensal bacteria through mechanisms that are still unclear. Bacteroides fragilis (B.frag), a relatively poor inducer of Tregs, utilizes Polysaccharide A (PSA) on the capsule for this function. Conversely, Bacteroides thetaiotaomicron (B.theta) is an excellent Treg inducer. We hypothesized that the mechanism used by strong Treg inducers (B.theta) would not be a polysaccharide antigen and would likely be a more conventional protein antigen. To test this, we generated a strain of B.theta lacking capsule as well as a transgenic mouse expressing CD4+ T cells that specifically responds to B.theta antigens in outer membrane vesicles (OMVersus). Using this system, we aim to define the B.theta-specific antigens that are important for Treg induction in the colon. We adoptively transferred CFSE-labeled naïve B.theta-specific T cells into B6. Rag1−/− mice that were gavaged with B.theta 3 days before cell transfer. One week later, the colon, colon draining mesenteric lymph nodes (cdLNs) and spleen were harvested for analysis by flow cytometry. To test whether T cell responses in the colon were dependent on the capsule of B.theta, recipients were gavaged with acapsular mutant of B.theta lacking capsule. Mice gavaged with PBS or a stain of B. theta that lacks sulfatase activity and is unable to access host APCs were used as controls. B.theta-specific naive CD4+ T cells transferred into recipients gavaged with B.theta significantly proliferated (PBS = 33 cells, N = 3 mice; B.theta = 1.6e105 cells, N = 2 mice) and upregulated FoxP3+ in the colon (N = 2 mice). As a negative control, gavage with sulfatase-deficient B.theta led to no detectable donor T cells in the colon (N = 2 mice). Because OMVersus from sulfastase-deficient B.theta do not localize in host macrophages, T cell recruitment and conversion to Tregs is likely an antigen-dependent process. We next tested the polysaccharide antigens of B.theta. We found that B.theta-specific CD4+ T cells converted to Tregs in mice exposed to acapsular B.theta (N = 2 mice) showing diversity in the mechanisms used by different Bacteroides to induce Treg responses. Our data using B.theta-specific T cells for the first time show that Treg induction is an antigen-driven process because B.theta lacking sulfatases is unable to access host APCs and induce Tregs. We also found that the antigen driving this response is not a carbohydrate found on the B.theta's capsule, because the acapsular B.theta mutant can drive Treg conversion. This data together with previous studies point out the diversity in the mechanisms utilized by commensals to drive T cell responses in the colon.