Organ-specific Microenvironments Drive Divergent T Cell Evolution in Acute Graft-Versus-Host Disease

Tissue-specific T cell immune responses play a critical role in maintaining organ health, but can also drive immune pathology during both auto- and alloimmunity. The mechanisms controlling intra-tissue T cell programming remain unclear. Here, we leverage a non-human primate model of acute graft-versus-host disease (aGVHD) after allogeneic hematopoietic stem cell transplantation (HCT) to probe the biological underpinnings of tissue-specific alloimmune disease, using a comprehensive systems immunology approach including multiparameter flow cytometry, population-based transcriptional profiling, and multiplexed single-cell RNA- and TCR- sequencing. Transcriptional profiling revealed substantial biological differences between T cells, infiltrating the lung and liver during aGVHD. These included enrichment for transcriptional pathways controlling extracellular matrix remodeling and chemotaxis in the lung, and enrichment for transcriptional pathways linked to nucleic acid metabolism and proliferation in the liver. Single cell RNA- and TCR-sequencing substantiated divergent organ-specific transcriptional programing of tissue-infiltrating T cells, which was linked to clonal expansion, with expanded clones progressively enriched for CX3CR1-expressing CD8 effector T cells in lung, and EOMES-expressing CD8 effector-memory T cells in liver. This divergent evolution of T cells was maintained even for T cells sharing the same TCRs, indicating its independence from antigen-specificity. Together, these results provide new insights into the principal role that tissue microenvironment-derived signals play in local T cell transcriptional programming during alloimmune-mediated clonal expansion, and which suggest potential opportunities to develop 64 tissue-specific therapeutics to curtail pathogenic immunity after transplant. Cryopreserved NHP cells were flow cytometrically sorted for Live, CD45+ CD3+, and CD14/CD20-cells. The cells were then immediately processed for single cell sequencing using the Next GEM 5’ v2 Gel Beads and Single Cell 5’ v2 Sample Index Plate TT. Gene expression alignment and TCR reconstruction was performed using Cellranger “multi” and “aggregate” pipelines (10X Genomics) to integrate both gene expression (GEX) and VDJ libraries. For the reference transcriptome, we first filtered Ensembl’s gtf file for v107 of the Macaca mulatta gene annotation using Cellranger’s “mkgtf” command. We provide the raw fastq files following sequencing in this repository. We also provide the individual sample outputs from Cellranger “multi” as well as the .hdf5 files used in our analysis pipeline as processed data.