Extended Data Fig. 4: In vitro and in vivo suppression, apoptotic rate and exT_reg cell formation of CARMA1-deficient T_reg cells.

a, CD4⁺CD45RB^highYFP^– T_conv cells and CD4⁺CD45RB^lowYFP^bright T_reg cells were double-sorted to more than 98% purity from LNs and spleens of F^cre × C1^+/+ × ROSA26-stop^f/f-YFP mice, which allow for clear differentiation of Cre-expressing T_reg cells based on high expression of soluble enhanced yellow fluorescent protein (eYFP) in addition to the YFP–Cre fusion protein. b, YFP^bright T_reg from F^cre × C1^+/+ or F^cre/+ × C1^f/+ or C1 ^f/f mice and CellTrace Violet-labelled T_conv cells from F^cre × C1^+/+ mice were co-cultured at indicated ratios for three days in the presence of anti-CD3 antibodies and T-cell-depleted splenocytes, and suppression measured as reduction of T_conv cell proliferation. c, T_reg cells of various genotypes and T_conv cells were co-adoptively transferred into Rag-deficient hosts and their respective frequency in peripheral blood was determined eight weeks later. d, CD4⁺YFP⁺ T_reg cells of indicated genotypes were cultured without exogenous IL-2 on anti-CD3/CD28-coated or uncoated plates for 6 or 18 h and examined for reactivity with annexin V and the viability dye ZombieRed. e, CD4⁺YFP^bright cells were sorted from LNs of one-year-old and F^cre/+ × C1^+/+ (or C1 ^f/+, C1 ^f/f) × ROSA26-stop^f/f-YFP mice and subsequently stained for expression of FOXP3 protein to determine the frequency of FOXP3^– exT_reg cells. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (two-way ANOVA with Sidak post hoc test in b, c; two-tailed Student’s t-test in d).