PD-1 Signaling Has a Critical Role in Maintaining Regulatory T Cell Homeostasis; Implication for Treg Depletion Therapy By PD-1 Blockade

CD4⁺ Foxp3⁺ regulatory T cells (Treg) play a central role in the maintenance of immune tolerance. In the setting of allogeneic hematopoietic stem cell transplantation (HSCT), Treg recovery after transplant has been reported as a critical factor to suppress the onset of graft-versus-host disease (GVHD) and stabilize the immune condition. We previously demonstrated that low-dose IL-2 administration preferentially increased Treg in patients with active chronic GVHD and resulted in clinical improvement of the symptoms (NEJM 2011). In the clinical trial, IL-2 induced selective and rapid proliferation of Treg in the first week of therapy but proliferation subsequently returned to baseline levels despite continued daily administration of IL-2 (Sci Trans Med 2013). By detailed analysis of lymphocytes from patients receiving IL-2 therapy, we found that inhibitory molecule Programmed death-1 (PD-1) expression rapidly increased selectively on Treg 2 weeks after starting IL-2, suggesting PD-1 may work to suppress the excessive proliferation of Treg and stabilize homeostasis of this subset during IL-2 therapy (ASH 2014). However, the cellular mechanisms of inhibition of Treg proliferation to maintain Treg homeostasis has not been characterized. To tackle the issue, we here studied the impact and role of PD-1 expression on Treg by using murine IL-2 therapy model. B6 mice were administrated 5x10³ IU IL-2 once a day for 14 days and CD4+CD25+Foxp3+ Treg were analyzed comparing with CD4+CD25-Foxp3- conventional CD4 T cells (Tcon) and CD8+ T cells. These subsets were further divided into subpopulations by the expression of CD44 and CD62L. The expressions of pSTAT5, Ki-67, Bcl-2, Fas, PD-1, CTLA-4, LAG3 and TIM-3 in each subset were also examined. As in human, low-dose IL-2 rapidly enhanced PD-1 on Treg, but not other inhibitory molecules including CTLA-4, LAG-3 and TIM-3. Low-dose IL-2 did not affect the expression level of any inhibitory molecules on Tcon and CD8T cells. To clarify the role of PD-1 expression on Treg, we compared PD-1^-/- B6 with wild-type B6. PD-1^-/- mice were administrated 5x10³ IU IL-2 once a day for 28 days. In the first 7 days, IL-2 induced significantly higher levels of pSTAT5 expression in Treg in PD-1^-/- cohort than WT cohort (MFI of pSTAT5; mean 4.8 vs 4.0, p<0.01). This resulted in the stronger proliferation and expansion of Treg in PD-1^-/- cohort in the first week of IL-2. However, after 7 days, IL-2 expanded PD-1^-/- Treg rapidly decreased and returned to baseline level, while IL-2 treated wild type Treg maintained to increase during the 4 weeks. We confirmed this phenomenon by another experimental system using the combination of low-dose IL-2 and anti-PD-1 antibody, and again found that the initial strong proliferation by IL-2 was just temporally and followed by the collapse of Treg homeostasis. To reveal the mechanism of Treg decrease in the absence of PD-1 signal, we assessed Annexin-V+ cells to detect apoptotic cells at day15. IL-2 treated PD-1^-/- Treg showed significantly higher level of Annexin-V+ cells than WT Treg (mean 10.5% vs 4.6%, p<0.01). IL-2 treated WT Treg were predominantly CD44^high CD62L^high central memory phenotype and this phenotype showed increase of PD-1 expression (%PD-1⁺ cells; mean 44.6% vs 31.7%, vs control, p<0.0001), in contrast, IL-2 treated PD-1^-/- Treg showed the accumulation of CD44^high CD62L^low effector memory phenotype, which showed high positivity of Annexin-V (mean 15.4% vs 6.4%, p<0.01). These data suggested that PD-1 modulates subset-balance of Treg and inhibits activation-induced-cell death (AICD). We also examined the expression of surface death receptor Fas and intracellular anti-apoptic protein Bcl-2 at day15 to investigate the pathway of apoptosis by lacking of PD-1 signal. IL-2 treated PD-1^-/- Treg showed higher increase of Fas expression (mean MFI, 1.5 vs 1.1, p<0.05) and decrease of Bcl-2 expression (mean MFI, 2.6 vs 3.4, p<0.05) than IL-2 treated WT Treg after IL-2 therapy, suggesting that Fas and Bcl-2 might participate in the regulation of apoptosis. In conclusion, our findings indicate that the PD-1 pathway can limit over activation and proliferation of Treg and prevent apoptosis via Fas and Bcl-2 activity during IL-2 therapy. Our finding may implicate the development of Treg depletion therapy by blocking PD-1 signaling.