Active GSK-3 and IL-12 are necessary for Treg induction by LPS-stimulated DCs. (A) Murine (B6) CTR- and RAPA-DCs were generated, purified, and treated with LiCl where indicated and then exposed to LPS as before. These DCs were then used to stimulate normal BALB/c CD4⁺ T cells for 5 days and flow cytometry used to quantify the percentage CD4⁺Foxp3⁺ cells under each condition. These flow cytometric data were normalized by calculating the percentage CD4⁺ Foxp3⁺ cells for each condition compared with that for CTR-DCs alone (mean ± SEM); n = 3. *P < .05. Unstimulated RAPA-DCs promoted an increase in the incidence of Foxp3⁺ cells relative to CTR-DCs. However, exposure of RAPA-DCs to LPS, or treatment of CTR-DCs with LiCl before LPS stimulation markedly reduced the capacity of these DCs to maintain Foxp3⁺ cells. (B) WT or IL-12p40^−/− mDCs were used to stimulate highly purified CD4⁺CD25⁻ BALB/c T cells. (B) Both IL-12p40^−/− mDCs and LiCl-treated DCs displayed a reduced capacity to support the induction of Foxp3 in CD25⁻ T cells. Data indicate percentage CD4⁺CD25^hi gated cells and are representative of 3 experiments performed. (C) The addition of exogenous IL-12p70 (5 ng/mL) at the start of cocultures of LiCL-treated, LPS-stimulated CTR-DCs and CD4⁺CD25⁻ T cells reestablished the population of CFSE^hiFoxp3^hi cells. Data indicate percentage of CD4⁺CD25^hi gated cells and are representative of 3 experiments performed.