Murine regulatory NK cells restrict allogeneic CD4⁺ T cell-driven gvhd without compromising cytotoxic function Free

Suppression of alloreactive responses is a major limitation to current therapies for graft-versus-host disease (GVHD), as they broadly inhibit immune function, compromising graft-versus-tumor (GVT) effects and protective immunity. CD56^brightCD16^- regulatory NK cells (NK_reg) have recently been associated with improved outcomes in human HCT recipients, but their mechanisms and therapeutic potential remain undefined. We hypothesize that murine NK cells with regulatory function exist and can be harnessed to suppress pathogenic alloreactive T cell responses and mitigate GVHD without impairing cytolytic immunity.

To identify candidate NK_reg subsets, we analyzed single-cell RNA-seq data from 3-month-old mouse spleens and found distinct NK1.1⁺ clusters. One subset expressed high Cd27 and low Cd11b, lacked cytotoxic genes (Gzmb, Prf1), and expressed stemness markers (Xcl1, Tcf7, Il7r). We focused on this subset due to its transcriptional similarities with human NK_regs. Unsupervised clustering from 30-color spectral flow cytometry across murine tissues (liver, thymus, blood, marrow, nodes, spleen; N=2; n=10) confirmed a CD27⁺CD11b⁻ NK cell population. Like human NK_regs, the cells have high expression of stemness proteins (TCF7, IL7R), low expression of cytotoxic markers, and high expression of the T-cell inhibitory ligand PD-L1.

To assess suppressive function, we co-cultured sorted CD27⁺CD11b⁻ NK cells with proliferation dye–labeled and activated CD4⁺ or CD8⁺ T cells or cytotoxic NK cells. CD27⁺CD11b⁻ NK cells suppressed CD4⁺ T cell proliferation without affecting CD8⁺ T cells or NK cells and did not induce cell death. Cytokine profiling of these assays revealed reduced CD4⁺ T cell–derived IFNγ (p<0.05) and TNFα (p<0.0005). To test whether suppression required direct contact, we used transwell plates. Suppression was observed to be contact-dependent, with reduced effect when cells were separated (p<0.005). (N=2–3; n=8–12).

To assess how allo-HCT affects this subset, we used an MHC-disparate GVHD model (C57BL/6J→BALB/cJ), profiling NK cells on days 1, 4, 7, 14, and 21 post-transplant (N=2; n=7–10/group). CD27⁺CD11b⁻ NK cells declined and failed to reconstitute in GVHD recipients (p<0.05). PD-L1 expression also decreased in the periphery and liver (p<0.005). Given the previously observed high PD-L1 expression, we tested whether PD-1–PD-L1 signaling mediated suppression. CD4⁺ T cell co-cultures with PD-1 blockade abrogated suppression (p<0.0005) (N=2; n=10).

To evaluate the role of NK cells in GVHD, we first depleted NK and ILC1 cells using the anti-NK1.1 antibody. Depletion increased GVHD scores and reduced survival (p<0.05) (N=2; n=20/group). Using anti-asialo-GM1 to selectively deplete NK cells produced similar results, suggesting the effect is NK-driven (p<0.0005) (N=2; n=20/group). Flow cytometry of recipient tissues revealed increased CD4⁺ T cell infiltration in the blood and liver, a key GVHD target and known homing site for human NK_reg cells.

To evaluate whether CD27⁺CD11b⁻ NK cells mediate these effects, we adoptively transferred sorted CD27⁺CD11b⁻ regulatory or CD27⁻CD11b⁺ cytotoxic NK cells into allo-HCT recipients with T cell–induced GVHD. First, to evaluate a prophylactic approach, cells were transferred at the time of transplant. CD27⁺CD11b⁻ NK cell transfer improved GVHD scores and survival. To evaluate a therapeutic approach, transfers were performed on day 7 post-transplant, with similarly observed improved GVHD scores and survival. In both settings, CD27⁺CD11b⁻ NK cell transfer selectively reduced CD4⁺ T cells in the periphery and liver without affecting CD8⁺ T cells.

To confirm translational relevance, we examined human CD56^brightCD16⁻ NK_reg cells from peripheral blood in functional assays paralleling the murine experiments. Human NK_reg cells similarly suppressed allogeneic CD4⁺ T cells (p<0.0005) while sparing CD8⁺ T cells and NK cells (N=5). Suppression was non-cytolytic, contact-dependent, and partially mediated by PD-1 (p<0.05). These results support functional conservation between murine and human NK_reg subsets and their selective targeting of CD4⁺ T cells.

In summary, we identify a novel murine CD27⁺CD11b⁻ NK_reg subset that suppresses CD4⁺ T cells through a contact- and PD-L1–dependent mechanism. These cells mitigate GVHD by restraining alloreactive CD4⁺ T cells in the periphery and liver, revealing a non-cytolytic, immunoregulatory NK population with translational therapeutic potential.