T Cells Motility in the Colonic Gvhd Is Influenced By Both Cognate Interaction and Microenvironment

Graft-versus-host disease (GVHD) accounts for substantial morbidity and mortality after allogeneic bone marrow transplantation. Allograft T cells are stimulated by both host and donor-derived antigen presenting cells (APCs). After differentiating into T effector cells, donor T cells migrate to GVHD target organs where they mediate damage directly and indirectly. However, very little is known about the dynamics of T cells in GVHD target tissues and how their behaviors are affected by the local environments, including by tissue-resident APCs. We utilized two-photon intravital microscopy (2PIM) to analyze T cell locations, motility and interactions with APCs. 129S1/SVLmJ (129) (H-2^b) hosts were irradiated and reconstituted with C57Bl/6 (B6) (H-2^b) CD11-YFP transgenic mice bone marrow together with B6 dsRed (RFP) CD8 cells and cyan fluorescent protein (CFP) CD4 cells. Mice were imaged from days 25-42 post-transplant. We focused on the colon which is strongly targeted in this model and in humans and which previously has been difficult to image by 2PIM. In syngeneic B6 into B6 recipients, donor CD11c⁺, CD8⁺ and CD4⁺ cells populate the colonic mucosa. However, a greater number of cells accumulated in the allogenic as compared to syngeneic recipients. Based on the analysis of 7-8 mice per group, T cells in allogenic mice were less motile. The mean CD8 speeds in syngeneic and allogeneic mice were 3.38 ± 0.07 µm/min and 2.21 ± 0.03 µm/min, respectively. Mean CD4 speeds in syngeneic and allogeneic mice were 3.11 ± 0.08 µm/min and 2.38 ± 0.04 µm/min, respectively. The majority of T cells were stationary, with few entering or leaving the imaged volume. Perfusion was confirmed by i.v. rhodamine-dextran; therefore, stationary behaviors was not due to tissue viability. Using optical clearing we imaged whole-mount colonic tissue from mucosa to serosa. T cells were aggregated in the sub-mucosa and infiltrated crypts and surprisingly also the muscular layer. Ki67⁺ T cells were found throughout, especially in the submucosa. Given the 2PIM data which show few new T cells entering the imaged volumes, these cells were likely stimulated to divide locally. Preliminary data with EdU pulsing suggest this to be the case. To determine whether T cell stability is microenvironment- or antigen-driven, we injected OT-1 TCR transgenic effectors which do not recognize any antigen in the recipient and imaged the next day. Most colonic OT-1 cells showed the same stationary behavior as nearby donor-derived CD4 and CD8 cells, suggesting that factors in the GVHD colon microenvironment drive T cell stability. Nonetheless there was also an antigen-driven component as injection of an anti-MHCII antibody (but not isotype control Ab) increased CD4⁺ T cell motility presumably by disrupting TCR:MHCII interaction. These results show the motility of T cells in the GVHD colon is influence by both TCR:MHC cognate interactions and by the microenvironment. That T cells are dividing and may be activated in situ, suggesting GVHD may be maintained locally. Current studies are focusing on what antigen-independent factors affect T cell motility and on defining the roles played by tissue APCs.