Plasmacytoid dendritic cell: vive le dilettante!

Comment on Schlecht et al, page 1808

Whether plasmacytoid dendritic cells (pDCs) are effective in presenting antigens has been a matter of debate. Schlecht and colleagues provide evidence that pDCs can indeed prime CD8 T cells specific for viral antigens.

Plasmacytoid dendritic cells (pDCs) are a rare population morphologically similar to plasma cells that express major histocompatibility complex (MHC) class II and costimulatory molecules, thereby triggering T-cell proliferation and differentiation. Moreover, they secrete large amounts of type I interferon (IFN; IFN-α and IFN-β) when exposed to DNA and RNA viruses. The ability of pDCs to stimulate T cells and secrete type I IFN has raised 3 fundamental questions. (1) Do pDCs present antigens as effectively as classical DCs? While it is generally agreed that pDCs are less efficient than DCs, whether pDCs can prime naive T cells, as DCs do, or just expand memory T cells, as B cells do, remains a matter of debate. (2) Does activation of pDCs affect the quality of antigen-specific T-cell responses? While activated pDCs up-regulate T-cell stimulatory molecules, it has been proposed that immature pDCs may in fact induce regulatory T cells that actively suppress antigen-specific T-cell responses. (3) Does pDC secretion of type I IFN contribute directly or indirectly to antigen presentation?

In this issue of Blood, Schlecht and colleagues address these issues using an original and clever approach that involves activating pDCs with different stimuli in vivo, recovering activated pDCs, loading them with antigen in vitro, and adoptively transferring them into mice. Remarkably, pDCs injected intravenously home efficiently to the spleen, making it possible to study their ability to stimulate or inhibit naive and memory T-cell responses. The results of these experiments demonstrate that the antigen-presenting capacity of pDCs is critically dependent on their state of activation and, in particular, the stimuli used to activate them in vivo. Two types of stimuli were evaluated: a cytosine-guanine (CpG) oligonucleotide, which activates pDCs through toll-like receptor (TLR) 9, and heat-inactivated influenza virus, which stimulates pDCs through TLR7. The pDCs activated with CpG can stimulate memory CD8 T cells but not naive CD8 T cells. In contrast, pDCs activated with influenza virus can prime naive CD8 T cells, although less efficiently than DCs. Moreover, T-cell responses primed by pDCs can be recalled by a second antigenic challenge as efficiently as those primed by classical DCs. Finally, immature pDCs pulsed with a peptide antigen induce neither antigen-specific CD8 T-cell responses nor tolerance to subsequent challenge with the same peptide.

That antigen presentation is induced by influenza virus but not CpG is surprising, as both stimuli activate the same TLR/myeloid differentiation factor 88 (MyD88) signaling pathway. How can this be? One clue to this difference can be found in one of the earliest experiments performed by Schlecht and colleagues, showing that influenza virus triggers IFN-α secretion while CpG does not. Previous studies have shown that IFN-α is unique in its ability to stabilize peptide-MHC class I complexes, possibly by inducing tapasin, which edits peptide loading onto class I by selecting high-affinity versus low affinity peptides. Thus, secretion of IFN-α by pDCs may not only induce an antiviral state in neighboring cells, protecting them from the cytopathic effect of the virus, but also enable pDCs to selectively present viral antigens by increasing the half-life of MHC class I-viral peptide complexes. In conclusion, despite their original image as dilettantes of antigen presentation, pDCs may be unexpectedly skilled in triggering T-cell responses specific for viral antigens.