Introduction: Antibody-mediated rejection (ABMR) is the major cause of late kidney allograft loss, with a lack of effective treatment options. Associations between Natural Killer (NK) cell gene signatures and graft infiltration with ABMR have been established, however, NK cell disease-specific phenotypes, pathogenic mechanism and factors regulating kidney NK cell function are not well defined. Further, many transplant recipients experience ABMR despite maintenance immunosuppression, suggesting inadequate suppression of pathogenic populations, including NK cells.

Methods: Using single cell and single nucleus RNA sequencing (scRNAseq/snRNAseq), we are mapping NK cell phenotypes in transplant recipients with ABMR. Using cell interaction analysis, we are identifying potential NK cell interactions with other kidney cell populations that could contribute to ABMR pathology. As effective therapies for ABMR are needed, we are assessing effects of immunosuppression on NK cells, with particular attention to memory-like NK cells and NK cell phenotypes observed in ABMR. 

Results: We identified two NK cell populations in kidneys of transplant recipients experiencing ABMR; one NKG2A+FCER1G+ population and another NKG2C+CD57+ population, in line with memory-like NK cell identity. Compared to NK cells from healthy living donor kidney, NK cells in ABMR had increased expression of genes associated with activation and degranulation and high TNF (TNF-a) expression, and reduced expression of the growth factor amphiregulin. Further, cell interaction analysis revealed a greater number of predicted interactions between NK cells and endothelial cells in ABMR, supporting NK cell pathogenicity in ABMR. To assess the impact of immunosuppression, we are systematically examining effects of mycophenolic acid, cyclosporine A, and prednisolone alone or in combination on NK cell proliferation, cytotoxicity, and cytokine production. Studies to date support that mycophenolic acid suppresses NK cell proliferation and IFN-gproduction, however it does not fully suppress TNF-a production, with differential impacts on CD56dimCD16+and CD56brightCD16- NK cells. Above 100 ng/ml, Cyclosporine A suppresses IFN-g and TNF-a production in CD56dimCD16+ NK cells but is unable to inhibit IFN-g production in CD56brightCD16- NK cells at this concentration. Below 100 ng/ml, however, Cyclosporine A is unable to inhibit either NK cell population. 

Conclusions: Collectively these studies are mapping NK cells in ABMR and identifying mechanisms by which NK cells contribute to pathology. Ongoing studies seek to identify novel approaches to limit NK cell activity in ABMR, informed by patient-based scRNAseq studies.