To investigate the role of E proteins in regulatory T cell development and Foxp3 expression, we utilized inducible E protein conditional KO mice in which both E2A and HEB (E2Af/fHEBf/f ER-Cre) can be deleted by treatment with tamoxifen for 6-8 weeks. To determine thymic development of Treg cells in E protein-deficient mice, we subjected thymocytes from mice with single KO (E2A or HEB deletion only), double KO (both E2A and HEB deletion) or mice with deletion of three alleles of the four E2A/HEB alleles to intracellular staining for Foxp3+ cells. In all of these KO mice, the Foxp3+ population constituted approximately 15-30% of the total CD4+SP thymocyte population compared to 3-5% in the WT mice. The finding that E protein levels in the thymus is inversely related to thymic Treg cell differentiation prompted us to investigate if E protein expression affects the generation of Treg cells in peripheral lymphoid organs. Indeed, we found a small but significant increase of the percentage of CD4+Foxp3+cells in peripheral lymphoid tissues of E protein KO mice as compared to WT mice In further studies we determined the functional (suppressor) capacity of FACS-sorted splenic Foxp3+CD4+ cells of E protein KO mice. To examine the function of these cells in vitro we employed standard T suppressor assay and found that splenic E protein KO CD4+ T cells had a similar capacity as WT CD4+ T cells to inhibit proliferation of conventional CD4+ T cells stimulated with anti-CD3/CD28 plus IL-2. To examine the function of splenic Treg cells from E protein KO mice in vivo, we determined the capacity of the cells to inhibit cell transfer colitis. Here, we transferred wild-type nave CD45RBhi CD4+ T cells alone or in combination with WT or E protein KO Foxp3+ Treg cells into recipient mice deficient in recombination-activating gene 2(Rag2-/-) and then monitored body weight and the development of colitis in recipient mice. Mice administered naive CD4+ T cells alone developed colitis within 12 weeks;in addition, while mice administered either naive CD4+ T cells plus E protein KO Tregs or WT Tregs inhibited naive CD4+ T cell-triggered body-weight loss and colitis, the inhibitory effect of E protein KO Tregs was more pronounced. These finding demonstrate that E protein regulates both the number of Foxp3+ Treg cells as well as its function. We next turned our attention to the mechanism by which E protein deletion could lead to increased induction of Foxp3+ Tregs. It is well established that IL-2 signaling dependent on IL-2R signaling plays an indispensable role in nTreg development;we therefore focused initially on the possibility that deletion of E protein promotes increased Foxp3 expression because such deletion results in increased IL-2 receptor expression and/or sensitivity. In a first set of studies along these lines we determined the expression of IL-2Ra, IL-2Rb and IL-2Rcommon gamma chain on CD4+ SP and CD8+ SP thymocytes by flow cytometry. Both IL-2Ra (CD25) and IL-2Rb (CD122) were up-regulated in the thymocytes of E protein KO mice as compared to WT mice, whereas IL-2Rcommon gamma chain was not upregulated. We next explored the capacity of IL-2 to induce Foxp3 expression in thymocytes in the absence of TCR stimulation. To this end, we obtained purified CD4+CD25+Foxp3- thymocyte populations by flow cytometric sorting from E protein KO and WT mice and then cultured the sorted cells with IL-2 or IL-7. We found that the extent of conversion of thymocytes to Foxp3+ cells is much higher in thymocytes exposed to IL-2 from E protein KO mice than WT mice. Thus, down-regulation of E protein not is only associated with increased IL-2R expression, but also with a lower threshold of response to IL-2 signaling. In a related set of studies along these lines we determined if the increased IL-2 and IL-7 receptor expression and sensitivity in thymocytes of E protein deleted mice could be associatted with increased activation of STAT5, a transcription factor activated by these receptors that has been found to be essential for Foxp3 expression and Treg cell development. Indeed, we found that thymocytes from E protein KO mice exhibited greatly increased spontaneous STAT5 phosphorylation in vivo by both flow cytometry and immunoblot. Such increased STAT5 phosphorylation was noted in CD4+ SP and CD8+ SP cells, but not DP cells, consistent with the fact that CD25 was not up-regulated in DP cells. In a final set of studies exploring the effect of effect of E proteins on IL-2 signaling we conducted extensive molecular studies to determine if E proteins affected IL-2Ra expression at the molecular level. In these studies we performed luciferase reporter assays with cells transfected with constructs containing a luciferase reporter driven by the IL-2Ra promoter and a conserved enhancer region and co-transfected with an E protein expression plasmid. We found that E-protein expression inhibits the luciferase signal in cells with the enhancer-containing contruct but not in cells without the enhancer or a mutated enhancer. It was thus apparent that E protein exerts a negative effect on the transcription of IL-2Ra via its capacity to bind to an IL-2Ra enhancer element. This negative effect thus provides one important reason why E-protein deletion is associated with increased expression of Foxp3+ cells. Another way in which E protein could be regulating Foxp3 expression is by affecting the expression of NF-kB, a transcription complex previously shown to be critical to Foxp3 transcription. Indeed, we found that thymocytes for E protein deleted mice expressed reduced levels of various NF-kB components including Ikkbeta, p50 and c-Rel. In accompanying molecular studies exploring the mechanism of this effect, we showed in luciferase reporter gene studies that luciferase signal in cells expressing an NF-kB reporter gene was inhibited by co-expression of an E protein expressing construct. Since the NF-kB reporter gene is responsive to various activated NF-kB components, this study indicated that E protein has a negative effect on the transcription of an as yet unidentified upstream NF-kB component. In micro-array studies supporting this conclusion we showed that Foxp3-negative cells from E protein deleted mice expressed increased levels of mRNA of a variety of NF-kB components including Ikkbeta as compared to Foxp3-negative cells from WT mice;this indicates that in the absence of E protein NF-kB is activated even before cells are subjected to TCR stimulation that ordinarily induces Foxp3 expression. In summary, these studies indicate that E protein exerts a profound negative effect effect on the transcription of various genes whose activation is necessary for the expression of Foxp3. Thus, the down-regulation of E protein during the course of thymocyte development sets the stage from the emergence of Foxp3 regulatory cells in the thymus.
Showing the most recent 10 out of 23 publications
