The GARP KO mice that were constructed for these studies had two features. First, the gene expressing M-cherry was knocked into the second exon of the GARP gene so that all cells that express GARP could be identified in flow-cytometric studies by the expression of a fluorochrome, M-cherry. Second, the 2nd exon of the GARP gene was floxed so that breeding of the GARP KO mice with mice expressing cell-specific Cre would lack GARP in the Cre-expressing cells . The studies to be described involve mice that express M-cherry in their GARP gene but have not yet been bred with Cre-expressing mice; thus these mice retain GARP expression. Initial studies of the mice described above confirmed that GARP is expressed and stored intracellularly in freshly isolated Foxp3-expressing Tregs from the spleen and that GARP is translocated to the cell surface where it binds LAP upon activation. Nevertheless, Garp expression is to some extent independent of Foxp3 expression. Thus, while it is present on all natural Foxp3+ cells in the thymus, it is not present on TGF-β -induced Foxp3 cells unless the inducing culture contains IL-2. In addition, the IL-2 inducing effect is reversed by inflammatory cytokines such as IL-6 and IL-27 that inhibit GARP expression to a somewhat greater degree than they inhibit Foxp3 expression. These regulatory effects of cytokines correlate with in silico analysis of the genomic regions flanking the GARP gene that are found to contain several STAT3 and STAT5 binding sites. Functional studies of GARP+ and GARP- Tregs revealed that both Treg populations have an equal capacity to inhibit proliferation of CD4+ T cell proliferation in vitro and that such inhibition is not blocked by anti-GARP, ant-LAP or an inhibitor of TGF-β-receptor function. On the other hand, GARP+ but not GARP- cells induce CD4+ T cells to produce IL-17 when cultured with the latter in the presence of IL-6. Such induction was equal to that induced by TGF-β and IL-6 in the absence of Tregs. Thus, the inability to inhibit the suppressor capability of GARP+ cells with an agent that blocks TGF-β or TGF-β function was not due to the lack of ability of GARP+ cells to express or secrete TGF-β. It was more likely due to the fact that GARP+ (or GARP-) Tregs can suppress T cell proliferation in vitro by a mechanism that only partially depends on TGF-β. Whether this is true of Treg cell suppressor function in vivo which has been closely tide to TGF-β-mediated affects awaits further studies. In any case these functional studies highlight the fact that GARP+ Tregs can participate in T cell differentiation function. Interestingly, induction of robust colonic inflammation by administration of Dextran Sulfate induced a 2-5-fold increase in the percentage of cells expression Foxp3 but no significant increase in the percentage of cells expressing GARP (mCherry). This again indicates that Foxp3 and GARP expression are independent of one another and may depend on the cytokine milieu in an inflamed tissue. In addition, the resullts of this study suggest that certain Tregs may become functionally inactive due to lack of GARP expression. Finally, we have taken advantage of the fact that we can isolate GARP-/Foxp3+ that do not express intracellular GARP (i.e., are mCherry negative) to study Treg gene expression in the absence of GARP with microarray studies. We have found that GARP expression is indeed associated with the increased expression of proteins such as 4-1BB (CD137)a cell surface co-receptor that enhances proliferation of CD4+ T cells upon interaction with its ligand. It therefore appears that GARP has an important intra-cellular function that relates to the expression of surface receptors that affect regulatory cell proliferation.
Showing the most recent 10 out of 42 publications
