Sanne Hoeks

PD-1 signalling and PKB in augmented CB Treg induction 27 2 INTRODUCTION In newborns the induction of productive immune responses is generally blunted in comparison with adults, resulting in tolerogenic immune reactivity. 1 This immune status results in an increased potency to engraft neonatal animals, ineffective vaccination responses in newborns, and reduced occurrence of graft-versus-host disease when cord blood (CB) derived allografts are used. 2, 3 We hypothesized that forkhead box protein 3 (FOXP3)–positive regulatory T (Treg) cells are pivotal in this phenomenon because these cells are key players in immune homeostasis. 4 When comparing the number of FOXP3 + cells in human CB and adult peripheral blood (APB), we found less FOXP3 + cells in CB than APB (Fig 1, A , ex vivo, uncultured cells). However, when naive (CD25 - CD45RO - ) T cells were activated by plate-bound anti- CD3, significantly more FOXP3 + T cells were induced from CB precursors. Remarkably, we only observed this difference between CB and APB when viable antigen-presenting cells (APCs; T cell–depleted CB or APB mononuclear cell fraction) were included in the culture; replacement with irradiated APCs abrogated the effect, whereas co-stimulation by soluble anti-CD28 showed a less pronounced difference (Fig 1, A ). Induced CB FOXP3 + T cells (sorted as CD4 + CD25 + CD127 low cells) were able to suppress dose dependently the proliferation of both CD4 and CD8 effector T cells when cultured together in different ratios (not shown), confirming the Treg nature of these cells. Thus on the first activation, CB T cells have a tendency to become functional FOXP3 + Treg cells. To substantiate the role of CB APCs in the induction of high percentages of Treg cells, we activated naive T cells in an alloreaction with APCs. Again we observed a higher number of Treg cells when CB T cells were cultured. CB APCs were able to induce higher numbers of Treg cells than APB APCs from both CB and APB precursor T cells (Fig 1, B ). Nevertheless, CB T cells always contained higher percentages of Treg cells, indicating a T-cell intrinsic mechanism as well. We investigated the percentages of different APC populations because the APCs used in our previous experiments consisted of a mixture of different cell types. In both APB and CB, we found comparable numbers of conventional dendritic cells (HLA-DR + CD11c + ) and plasmacytoid dendritic cells (HLA-DR + CD123 + ), with slightly more CD14 + monocytes in APB (not shown). When we sorted different subsets as APCs and cultured them with anti-CD3–activated T cells, we observed that all APC subsets induced more Treg cells from CB than from APB precursor cells (Fig 1, C ). The augmented Treg cell induction was not correlated with a reduced proliferation of CB T cells, which we showed to be identical in CB and APB by using carboxyfluorescein succinimidyl ester dilution assays (data not shown). Moreover, we found no difference in the kinetics of FOXP3 upregulation when we measured the percentage of FOXP3 + cells daily; both APB and