Date Published: June 6, 2019
Publisher: Public Library of Science
Author(s): Naoki Tanimine, Sharon K. Germana, Martin Fan, Keli Hippen, Bruce R. Blazar, James F. Markmann, Laurence A. Turka, Bhavana Priyadharshini, Jian Zhang.
Regulatory T cells (Tregs) are required for the maintenance of immune tolerance and adoptive Treg infusion therapy has become a promising approach to suppress immune responses in diseases such as autoimmunity and transplant rejection. However, one critical challenge of Treg therapy is the requirement of in vitro expansion of functionally stable Tregs while preventing either the contamination of T effector and/or emergence of unstable pathogenic Tregs. Recent studies showing distinct metabolic requirements of T effectors and Tregs suggest that manipulation of cell metabolism may be an attractive strategy to achieve this goal. Here we show that human thymically derived Tregs (tTregs) and in vitro induced Tregs (iTregs) from naive T cells engage glycolysis equivalently upon activation. However, inhibiting glucose metabolism via 2-deoxy-D-glucose (2DG) has distinct effects on each of these subsets. While 2DG treatment at the onset of activation significantly reduced the proliferation and expression of suppressive molecules such as ICOS and CTLA-4 in tTregs, its effect on FOXP3 expression was small. In contrast, 2DG treatment during iTreg induction modestly decreased their proliferation but strongly reduced both ICOS and FOXP3 expression. Importantly, both Treg subsets became insensitive to 2DG after day 3 post activation with little effect on either proliferation or FOXP3 expression while T conventional Th0 cells showed reduced proliferation under the same conditions. Moreover, 2DG treatment at day 3 did not impair the suppressive capabilities of Treg subsets. Collectively, these findings suggest that there is a distinct temporal requirement of glycolysis in each of the activated human Treg subsets and T conventional cells. Furthermore, 2DG treatment at the onset as a strategy to impair contaminating T effector cell proliferation is unfavorable for optimal Treg generation as well.
Regulatory T cells (Tregs) are critical for the prevention of autoimmunity  and for transplantation tolerance . Given their therapeutic potential, clinical trials of Treg immunotherapy are already underway in patients with autoimmune diseases and recipients of stem cell and solid organ transplants. Despite these advances, it remains challenging to manufacture sufficient numbers of functionally stable Tregs and optimize in vivo conditioning treatments, which highlight the issue of feasibility and safety with current approaches [4, 5].
To determine the metabolic differences between human tTregs and iTregs, we first isolated ex vivo Tregs (tTregs) and naïve CD4+ T cells (S1 Fig). tTregs were stimulated with CD3/CD28 antibody beads in the presence of IL-2 , while iTregs were induced by culturing naïve CD4+ T cells with CD3/CD28 antibody beads in the presence of TGF-β, ATRA and IL-2  (Fig 1A). We observed that iTregs divided more than tTregs by day 3 post activation as assessed by their higher proliferation index (PI) (mean PI, tTreg 1.4±0.4 vs iTreg 3.0±0.5, P<0.001, (Fig 1B). As a result, the mean population doubling was 3 fold higher in iTregs at day 5 compared to tTregs (Fig 1C). Concomitantly, we found that tTregs substantially increased FOXP3 expression upon activation. In contrast, the kinetics of FOXP3 expression in iTreg was transient. FOXP3 was induced by day 3 although not to the same degree as tTregs; reached its maximum around day 5, and decreased by day 7 (Fig 1D). The successful implementation of Treg therapy (either polyclonal or antigen specific) relies upon efficient Treg proliferation and maintenance of functional stability both in vitro and in vivo. Data from murine studies with metabolic inhibitors demonstrated that differentiation of iTregs require FAO, whereas that of functional T effectors require glycolysis and blocking glycolysis promotes iTreg differentiation [10, 11]. Genetically modified murine models also showed that enhanced glycolysis in Tregs can in fact diminish FOXP3 expression and disrupt their lineage stability [14, 16, 25]. However, it is becoming increasingly evident that this is not a set paradigm. Others and we show that activated ex vivo Tregs (both mice and human) can engage in glycolytic metabolism [18, 29]. In our observations, we found that both human tTreg and iTreg subsets equally engage in glycolysis post activation, but iTregs concurrently show greater utilization of oxidative metabolism potentially to support their higher proliferative capacity. TGF-β containing iTreg skewing conditioning did not affect glycolytic metabolism and proliferation of tTregs. This is in contrast to murine tTregs, where exposure to TGF-β inhibits their glycolytic metabolism and consequently their proliferation  indicating a species specific difference of Treg responses to TGF-β. Source: http://doi.org/10.1371/journal.pone.0217761