Azacitidine Induces A Shift In Th1/Th17 Ratio and FoxP3 Expression In Anti-CD3 Stimulated CD4+ T-Cells; Implications for the Treatment of Myelodysplastic Syndromes

Abstract 4026

Aberrant DNA methylation and other epigenetic changes play a role in the development of myelodysplastic syndromes (MDS). Epigenetic drugs such as DNA methyltransferase inhibitors are therefore increasingly employed in MDS treatment regimens. Recent studies show that gene methylation processes also regulate T-cell function. Here we analyzed the in vitro effects of the DNA methyltransferase inhibitor ‘5-azacitidine (Aza) on CD4+ T-cell activation. We confirmed the previously described inhibition of proliferation and increased expression of FoxP3, the regulatory T-cell (Treg) marker, by anti-CD3 stimulated T-cells in the presence of 1mM Aza. Here we have sorted CD4+ T-cells isolated form healthy donor peripheral blood into CD25^neg resting, CD25^dim recently activated and CD25^hi Treg cells. Aza facilitated the induction of CD25hiFoxP3+ T-cells from CD25^neg (4.7% of vehicle treated cells versus 17.3% of Aza treated cells p=0.0007, n=9) and to a lesser extend from CD25^dim (1.2% versus 8.6%, p=0.0015, n=7) CD4+ T-cells, while Aza had no effect on FoxP3 expression in CD25^hi sorted cells, FoxP3 expression remained high. In addition, cytokine producing T-cells were enumerated after stimulation with phorbol-12-myristate-13-acetate (PMA) and ionomycin in the presence of Brefeldin A. Aza treatment increased the number of IFNγ producing cells in the total CD4+ population (19.1% versus 40.8%; p<0.0001, n=10) as well as among the CD25^neg (5.7% vs 41.2%; p=0.001, n=8) and CD25^dim CD4+ T-cell populations (28.4% versus 46.6%; p=0.06, n=7). TNFα producing cells were increased in the total CD4+ (36.4% versus 51.1%; p=0.011, n=9) and CD4^posCD25^dim (36.7% versus 52.3%; p=0.033, n=6) populations but not in the CD4^posCD25^neg cells (50.9% versus 51.1%; p=0.9, n=7). This increase in pro-inflammatory cytokine production indicates that Aza induces T-cell activation and that the increase in FoxP3 expression may reflect T-cell activation rather than an increase in bona fide Treg by Aza treatment. Indeed a proportion of the FoxP3+ cells was positive for TNFa or IFNg, suggesting that these are activated T-cells rather than Treg. However, the proportion of FoxP3+IFNγ- and FoxP3+TNFα- cells was significantly higher among Aza treated CD4+ cells (p=0.0037 and 0.0018 respectively, n=5), suggesting an increase in Treg as well. Functional assays to demonstrate that these FoxP3+ cells are indeed regulatory T-cells are currently being set up. Next to IFNγ and TNFα producing Th1 cells, the more recently described IL-17 committed Th17 cells have been described to play a role in low risk MDS. Furthermore, it has been shown that Treg can differentiate into IL-17 producing cells. We, therefore evaluated the effect of Aza on Th17 cells. Interestingly, in contrast to IFNγ and TNFα producing cells, the proportion of IL-17+IFNγ- Th17 cells among the total CD4+ population was reduced by Aza treatment (1.8% versus 1.1% p=0.035, n=10), leading to a significant increase in the Th1/Th17 ratio (14.0 versus 47.9, p=0.0005, n=10). In contrast, although the numbers were small, the proportion of Th17 cells was increased by Aza in the CD25^neg population (0.08% versus 0.20%, p=0.028, n=8), suggesting that Aza may have differential effects on resting and recently activated T-cells. In conclusion, our data show that Aza increases the induction of FoxP3+ Treg and Th1 cells but inhibits IL-17 production, particularly by previously activated T-cells. Aza may therefore particularly be beneficial in pathogenic immune disorders characterized by increased Th17 numbers accompanied by reduced Treg frequencies, such as low-risk MDS.