Reconstitution of RAG2^-/- recipients with host-type CD4⁺CD25⁺ T cells reduces penetrance and severity of clinical cGVHD. (A) Incidence of cGVHD. Combined data from 2 experiments. On day 0, recipient mice were lethally irradiated and reconstituted with 8 × 10⁶ BM cells alone (n = 18, both recipient types) or BM plus 6 × 10⁶ WT spleen cells (n = 27). RAG2^-/- recipients received BM + spl with either 10⁶ freshly purified BALB/c CD4⁺CD25^- T cells (spl + CD25^-, n = 19) or 10⁶ freshly purified BALB/c CD25⁺ T cells (spl + CD25⁺, n = 20). ^*P < .05 for RAG2^-/- spl + CD25⁺ versus RAG2^-/- BM + spl. (B) Average clinical disease score for mice affected with cGVHD (unaffected mice are excluded). ^*P < .05, †P < .01 for RAG2^-/- spl + CD25⁺ compared with RAG2^-/- BM + spl. BM control mice did not get cGVHD and are represented on the graph as scoring “0.” (C) Pathology scores for representative mice. Mean score is indicated by a horizontal bar. †P < .01 for RAG2^-/- spl + CD25⁺ compared with RAG2^-/- BM + spl, RAG2^-/- spl + CD25^-/-, or WT. Pathology samples were taken on day 26, 34, or 50.