The inhibitory effect of photodepletion cells requires cell-cell contact and CTLA-4. (A) Photodepletion GVHD cells were physically separated from GVHD targets using Transwell plates to prevent cell-cell contact, but not cytokine interaction. After 5-day cultures, proliferation of target cells exposed to untreated (clear) versus Photodepletion-treated (filled bars) cells was measured in a ³H-thymidine incorporation assay, and (B) IL-10 levels were measured in the supernatants using ELISA. Bars represent mean ± SEM of 3 independent experiments. (C) PBMCs from GVHD patients were cultured with untreated (clear bars) or photodepletion-exposed PBMCs from the same patient for 5 days in the absence (filled bars) or presence of mAbs directed against CTLA-4, GITR, or an irrelevant antigen. Proliferation in each of the above conditions was determined by measuring ³H-thymidine incorporation (mean ± SEM). (D) FOXP3 expression in CD4⁺ cells was evaluated by flow cytometry after GVHD patient cells were cultured with autologous untreated cells, photodepletion-treated cells, or photodepletion-treated cells with anti–CTLA-4 mAbs. Results are expressed according to (D) the proportion of FOXP3⁺ cells within CD4⁺ cells and (E) the absolute number of CD4⁺FOXP3⁺ cells (results are shown as mean ± SD). *P < .05 and ***P < .001.