A2-CAR CD8⁺ Treg therapy efficiently controls HLA-A*02 mismatched skin graft allogeneic rejection. (A) Schematic depicting the model of human skin graft in NSG mice. Skin from HLA-A*02⁺ human healthy volunteer (HV#1) was grafted on immuno-incompetent NSG mice, and 1 month later (= day 0 of graft survival challenge), mice were IV injected with 5 × 10⁶ PBMCs from a HLA-A*02 mismatched healthy volunteer (HV#2, HLA-A*02⁻) to induce an allogeneic rejection of the skin graft. A2-CAR or Her2-CAR CD8⁺ Tregs were generated from PBMCs of HV#2 donor and co-injected with PBMCs in grafted mice at a PBMCs:Tregs ratio of 3:1 or 1:1. (B) Survival of skin graft over time based on macroscopic features of rejection in mice treated with A2-CAR CD8⁺ Tregs at a ratio of PBMCs:Tregs of 3:1 (orange line, n = 6) or 1:1 (red line, n = 4) or Her2-CAR at a ratio of PBMCs:Tregs of 3:1 (green line, n = 3) or 1:1 (blue line, n = 4) or not treated (purple line, n = 9). Log rank (Mantel Cox) test, *P < .05; **P < .01. (C) Percentage human PBMCs in mouse blood over time after cell transfer. One line represents 1 mouse. (D) Representative photos of HPS coloration of human skin tissue grafted 100 days after cell transfer. Arrows show immune cell infiltrate, stars delimit human skin graft interface with mouse skin, scales are indicated on photos. (Upper) Representative photo of a skin graft of mice injected with PBMCs, macroscopic score, 3. (Lower) Representative photo of skin graft of mice injected with PBMCs and treated with A2-CAR CD8⁺ Tregs, macroscopic score, 2. (E) Human skin grafts were assessed 100 days after cell transfer of PBMCs ± A2-CAR CD8⁺ Tregs by immunofluorescence stains of 4′,6-diamidino-2-phenylindole (DAPI) and involucrin staining. Images are representative stains; scale is indicated on the bottom left.