Daan Pieren

174 Chapter 6 NKG2A-HLA-E interaction counteracts suppressive responses in CD8 + Tregs Murine CD8 + Ly49 + Tregs can prevent autoantibody-induced autoimmune disease by binding Qa-1b (HLA-E in humans) present on dysregulated follicular helper T cells [54,65]. Interestingly, a subset of CD8 + Tregs recognizing HLA-E has also been shown in humans and failure of these cells to recognize HLA-E may lead to type-1 diabetes [70]. Moreover, senescent human fibroblasts have recently been shown to express HLA-E, which interacted with NKG2A on CD8 + T cells and thereby prevented fibroblast elimination [71]. Thus, HLA-E can be bound by either the T-cell receptor (TCR) or the NKG2A/CD94 receptor of CD8 + Tregs, with different outcomes: binding of the TCR to HLA-E on target T cells leads to suppression of target T cells by killing of these T cells in mice [72], whereas this suppressive mechanism can be inhibited when NKG2A binds HLA-E, as NKG2A-signalling abrogates TCR signaling [73] ( Figure 4 ). The peptide presented in HLA-E that is recognized by human CD8 + Tregs is still unknown, but heat shock protein 60 (HSP60) as well as self-antigens are thought to be presented by HLA-E and recognized by CD8 + Tregs [68,70,74]. T cells can upregulate expression of HSP60 presented in MHC class I and II after activation and can in turn be suppressed by Tregs that recognize MHC/HSP60 complexes [75]. Thus, these findings suggest that the interaction of NKG2A with HLA-E limits the TCR-mediated capacity of CD8 +  Tregs to suppress, whereas interaction of KIR with its ligand results in survival by preventing overactivation ( Figure 4 ). Do NKG2A + RA + T cells convert to KIR + RA + T cells during aging? In chapter 5 , we show that the proportion of NKG2A + RA + T cells declines at older age, whereas the proportion of KIR + RA + T cells increases. We describe NKG2A + RA + and KIR + RA + T cells as two independent cell subsets that were previously regarded to be one. However, we cannot exclude the possibility that KIR + RA + T cells present at older age are derived from NKG2A + RA + T cells that lost NKG2A and gained KIR expression during aging. An argument for this hypothesis can be derived from natural killer cells (NK cells). KIR and NKG2A are classical NK-cell markers and it has been shown that NKG2A + NK cells may lose NKG2A during differentiation and start expressing KIRs [76,77]. KIR + NK cells, but not NKG2A + NK cells, degranulate when in contact with HLA-E expressing cells, indicating killing of these HLA-E + cells by KIR + NK cells. Based on our findings on KIR + RA + T cells in the context of the studies mentioned above, we established a hypothesis ( Figure 4 ). We propose that due to a yet to be identified process that is accelerated by aging, NKG2A + CD8 + T cells may lose expression of NKG2A and gain expression of KIR. Gain of KIR expression ensures