Complement-induced regulatory T cells suppress T-cell responses but allow for dendritic-cell maturation

Abstract

Concurrent activation of the T-cell receptor (TCR) and complement regulator CD46 on human CD4+ T lymphocytes induces Tr1-like regulatory T cells that suppress through IL-10 secretion bystander T-cell proliferation. Here we show that, despite their IL-10 production, CD46-induced T-regulatory T cells (Tregs) do not suppress the activation/maturation of dendritic cells (DCs). DC maturation by complement/CD46-induced Tregs is mediated through simultaneous secretion of GM-CSF and soluble CD40L, factors favoring DC differentiation and reversing inhibitory effects of IL-10. Thus, CD46-induced Tregs produce a distinct cytokine profile that inhibits T-cell responses but leaves DC activation unimpaired. Such "DC-sparing" Tregs could be desirable at host/environment interfaces such as the gastrointestinal tract where their specific cytokine profile provides a mechanism that ensures unresponsiveness to commensal bacteria while maintaining reactivity to invading pathogens.

Figure 1.

CD46-induced Tr1-like cell supernatants do not suppress dendritic-cell (DC) maturation despite high IL-10 content. (A) General experimental approach. Purified human blood monocytes were cultured in GM-CSF/IL-4–containing media for 72 hours. The generated DC precursors were incubated for 24 hours with supernatants derived from either CD3-, CD3/CD28-, or CD3/CD46-activated T cells (Tr1) or maintained in GM-CSF/IL-4–containing control media. Maturation of these DC populations was induced by LPS addition for 24 hours. The maturation stage of the DCs was determined by their maturation marker expression profile and their potential to induce allogeneic T-cell proliferation in an MLR. (B) DCs exposed to supernatants from CD3/CD46-activated Tr1 cells up-regulate maturation markers. iDCs were generated and the media replaced with supernatants derived from CD3/CD46-induced Tr1 cells, CD3- and CD3/CD28-activated CD4⁺ T cells, or fresh media with 500 pg/mL recombinant human IL-10 (rec. IL-10). DC maturation was induced by LPS addition and surface expression of MHCII, and CD86 was analyzed by FACS after 24 hours. (C-D) DCs matured in control media (C) or in supernatants from CD3/CD46-activated Tr1 cells (D) demonstrate strong MLR potential. DCs were generated and treated as described in panel B, and their potential to induce allogeneic T-cell proliferation was measured in an MLR: DCs were seeded in serial 2-fold dilutions beginning with 50 × 10⁴ cells/well and irradiated. Purified allogeneic PBMCs were then added at 50 × 10⁴ cells/well, the cocultures were incubated for 4 days, and proliferation of allogeneic T cells was measured via [³H] thymidine incorporation. Shown is 1 representative of 8 independently performed experiments (with each activation condition in triplicate). (E) Summary and statistical evaluation of all proliferation experiments is shown as relative data ± standard error. n indicates the number of separate proliferation conditions (ie, 4 dilutions/experiment in duplicates or triplicates) considered for this evaluation. Mo indicates monocyte; Tr1 supern., supernatant derived from CD3/CD46-induced Tr1-like cells; and MLR, mixed lymphocyte reaction. F statistics and P values are shown for the comparison.

Figure 2.

CD46-activated IL-10–secreting CD4⁺ T cells express high amounts of CD40L and GM-CSF. Purified peripheral-blood T cells (sorted CD4⁺CD45RA⁺ cell population) were activated with the indicated immobilized mAbs for 18 hours, and Monensin was added to the last 8 hours of culture. Cells were permeabilized and fixed, and intracellular cytokine staining was then performed for (A) IL-10 and CD40L or (B) IL-10 and GM-CSF. Shown is 1 representative FACS analysis of 3 independently performed experiments (with each activation condition in triplicate). The observed level of significance for the differences in the amount of IL-10/CD40L or IL-10/GM-CSF production between CD3/CD28- and CD3/CD46-activated cells was P < .001 by the paired Student t test in all cases.

Figure 3.

Soluble CD40L mediates the DC activation in supernatants derived from CD46-induced Tr1-like cells. (A-D) iDCs were incubated with control media containing GM-CSF/IL-4 (A) or supernatants derived from T cells activated with CD3 (B), CD3/CD28 (C), or CD3/CD46 (D). Each condition was performed with or without the addition of the indicated function neutralizing mAbs. A nonspecific isotype mAb was used as a control. This mAb had no effect on any of the tested activation conditions (data not shown). DC maturation was induced with LPS and the capacity of the DCs to induce allogeneic T-cell activation was analyzed in an MLR. Shown is 1 representative of 5 independently performed experiments (with each activation condition in triplicate). (E) Summary and statistical evaluation of all proliferation data obtained in this set of experiments, displayed as described in Figure 1E. F statistics and P values are shown for the comparison between untreated (gray bars) and anti-CD40L ligand mAb–treated conditions (black bars). 10, 40L, and G indicate the addition of function-neutralizing mAbs to IL-10, CD40L, and GM-CSF, respectively.

Figure 4.

MLR potential of DCs incubated with CD46-activated T-cell supernatant derived from a CD40L-deficient patient. (A) iDCs were incubated with supernatants derived from CD4⁺ T cells from a CD40L-deficient patient activated with CD3/CD46 or CD3/CD28, with or without the addition of a function neutralizing mAb to IL-10. The capacity of DCs to induce allogeneic T-cell activation after LPS addition was analyzed in an MLR. Shown is 1 representative of 3 independently performed experiments (with each activation condition in triplicate). (B) Summary and statistical evaluation of all proliferation data obtained in this set of experiments, displayed as described in Figure 1E.

Figure 5.

DCs matured in the presence of Tr1-like cell supernatants induce the proliferation of conventional effector T cells. Peripheral-blood monocytes were cultured for 3 days in media containing GM-CSF and IL-4. These immature DCs were then incubated for 24 hours in control media containing GM-CSF/IL-4 (A), and supernatants from T cells activated with CD3 (B), CD3/CD28 (C), or CD3/CD46 (D), and maturation was induced by LPS addition for 24 hours. The matured DCs were washed and used in an MLR with allogeneic PBMCs. The cytokine profile of the proliferating PBMCs was analyzed after 5 days using the Th1/Th2 cytokine bead array (BD Biosciences). Data shown represent the mean cytokine production ± SD of 3 independent experiments performed in triplicate. The observed level of significance for the differences in the amount of cytokines produced by DCs treated with either control media or different supernatants was P > .15 by the paired Student t test in all cases.