In This Issue Free

The process of T cell activation can be divided into Phase I, an initial period of transient contacts between the T cell and APC, and Phase II, during which these contacts become stable. It has been difficult to determine whether calcium flux, which is indicative of TCR signaling, occurs during Phase I in vivo, and this difficulty is compounded by the tendency of calcium-sensitive dyes to leach out of cells. To clarify this issue, Le Borgne et al. (p. 1471) generated inducible knock-in mice bearing a modified form of the genetically encoded FRET-based calcium biosensor Cameleon (mCameleon), which could measure calcium flux with a high degree of sensitivity. In vitro imaging of mCameleon⁺ TCR-transgenic naive T cells stimulated with peptide-loaded APCs revealed calcium signaling patterns that could be transient, sustained, or oscillating. Different TCRs favored distinct patterns, although T cells bearing the same TCR could often demonstrate all three patterns. These patterns were somewhat affected by Ag quality but were unrelated to Ag quantity or the type of APC. The authors then used two-photon imaging of lymph nodes (LNs) in C57BL/6J recipients of mCameleon⁺ OT-I TCR transgenic T cells and Ag-loaded dendritic cells to analyze in vivo calcium signaling during Phase I of T cell activation. Compared with T cells in control LNs, T cells in draining LNs had increased concentrations of intracellular calcium that did not require continuous contact between T cells and dendritic cells. In addition to presenting a useful method for in vivo assessment of calcium flux, this study addresses an unresolved question in T cell biology by revealing that the transient T cell-APC contacts formed during Phase I are productive contacts that induce TCR signaling. This suggests that the accumulation of multiple low-level TCR-APC interactions resulting in TCR signaling allows T cell activation to progress to Phase II.

In addition to using their Ig-like TCRs to bind soluble Ags, some γδ T cells have been shown to recognize processed Ags presented on the surface of APCs. Although evidence for stimulation of γδ T cells by CD1c, which displays lipid Ags, was published in 1989, a detailed analysis of γδ T cell recognition of CD1c-Ag complexes has remained lacking. Roy et al. (p. 1933) have now identified the mycobacterial cell wall lipid Ag phosphomycoketide (PM) as an Ag recognized by human CD1c-restricted γδ T cells and investigated γδ T cell interactions with CD1c in depth. Use of PM-loaded CD1c tetramers identified Vδ1⁺ γδ T cells that directly bound to CD1c to induce T cell stimulation. These cells all used the TRDV1 gene but had otherwise variable TCRs, and the authors found that the TCR δ chain played a dominant role in TCR–CD1c binding, whereas different γ-chains varied greatly in their effects on binding. Surprisingly, in addition to binding to CD1c presenting PM or a related lipid Ag, γδ T cells selected by CD1c–PM tetramers also bound to CD1c proteins bearing endogenous lipids from the expression system, which had been intended to serve as negative controls. This observation suggested that these γδ T cells could bind a broader variety of Ags than would be expected for αβ T cells, and that the binding between the TCR and CD1c was the key factor in ligand recognition. In support of this idea, endogenous permissive Ags that allowed direct interactions between CD1c and the TCR were shown to bind γδ TCRs with varying affinities, whereas nonpermissive Ags that blocked these interactions could not. This analysis suggested that, unlike the generally foreign Ag-restricted nature of αβ T cell responses, γδ T cell stimulation may occur in a more graded manner in response to a variety of CD1c-presented lipid Ags.

Optimal activation and differentiation of CD8⁺ T cells depends on three signals: TCR engagement, costimulation, and cytokine signaling. TCR activation without cytokine signaling can lead to defective expansion and memory formation, and even result in tolerance. Different infections lead to distinct cytokine milieus, and Booty et al. (p. 1822) examined the role that several key cytokines have in CD8⁺ T cell differentiation in response to Mycobacterium tuberculosis. As different cytokine receptor knockout (KO) mice have varying susceptibility to M. tuberculosis infection, 1:1 mixed bone marrow chimeras were established to compare wild-type (WT) and IL-12R^−/−, IFNAR^−/−, or IL-27R^−/− CD8⁺ T cells within the same host. Reconstitution of WT and cytokine receptor KO CD8⁺ T cells was equivalent, indicating that these cytokine receptors are dispensable for CD8⁺ T cell homeostasis. Four weeks after M. tuberculosis infection, WT CD8⁺ T cells were more prevalent in the lungs than any of the cytokine receptor KO CD8⁺ T cells, with the most dramatic difference observed with IL-12R^−/− CD8⁺ T cells, which also were less capable of producing IFNγ following Ag stimulation ex vivo. KLRG1 and CD127 phenotypic analysis of these cells revealed that, relative to WT Ag-specific CD8⁺ T cells, IL-12R^−/− Ag-specific CD8⁺ T cells had a lower proportion of KLRG1^hiCD127^lo short-lived effector cells. Although all three types of cytokine receptor KO cells expressed less granzyme B than WT CD8⁺ T cells, no defects in specific killing were apparent in other sets of chimeras in which all CD8⁺ T cells lacked individual cytokine receptors. Collectively, these results suggest that although type 1 IFN and IL-27 are also involved in CD8⁺ T cell expansion in the lung, IL-12 is the dominant cytokine in CD8⁺ T cell priming and differentiation following M. tuberculosis infection.

During T cell activation, phosphatidylinositol 4,5-bisphosphate (PIP2) serves as a precursor for several second messenger molecules and modulates cytoskeletal dynamics, bringing lipid rafts together with the TCR and costimulatory molecules at the immune synapse. Synthesis of PIP2 requires type I phosphatidylinositol 4-phosphate 5-kinase (PIP5K), which exists in α, β, and γ isoforms in CD4⁺ T cells. In this issue, Kallikourdis et al. (p. 1955) sought to elucidate the role of the PIP5Kβ isoform in T cell activation. In human primary CD4⁺ T cells, PIP5Kβ was recruited to the T:APC interface through a CD28-mediated, TCR-independent mechanism that required the C-terminal region of PIP5Kβ but not its lipid kinase activity. This recruitment following CD28 stimulation also required the presence of the guanine-nucleotide exchange factor Vav1, and PIP5Kβ activity was found to act downstream of Vav1. The C-terminal region of PIP5Kβ was also important for actin remodeling and accumulation of filamin-A and lipid rafts during T cell activation. Knockdown of PIP5Kβ using small interfering RNA directly impaired T cell activation, as measured by reductions in intracellular calcium levels and expression of CD69 and CD25. Combined with data from previous studies, these data suggest that different isoforms of PIP5K generate pools of PIP2 with distinct functions during T cell activation, and that PIP5Kα and PIP5Kβ have nonredundant activities downstream of CD28 stimulation.

Phenotypically and functionally specialized dendritic cells (DCs) reside in the tissues and lymphoid organs, and are crucial to mounting effective T cell responses. Transcription factors such as IRF4 and IRF8 are expressed in precursors to conventional DCs (pre-cDCs) and can dictate DC differentiation. In this issue, Bajaña et al. (p. 1666) used conditional knockout mice (CD11c-cre-Irf4 and CD11c-cre-Irf8 mice) to determine the impact of these transcription factors on the homeostasis of DC subsets. In the CD11c-cre-Irf4^−/− mice, Cre is driven by the CD11c promoter and the Irf4 gene is replaced with GFP in CD11c⁺ cells, including NK cells, pre-DCs, mature DCs, and some macrophages. Cells with intermediate CD11c expression consisted of mixed populations of GFP⁺ and GFP^- cells, indicating that the target gene deletion may not have been complete. Numbers of pre-cDCs in the spleen and bone marrow were unaffected by Irf4 deficiency and these cells had no compensatory increase in IRF8 expression. The CD11c-cre-Irf4^−/− mice did, however, have reduced numbers of CD11c^hiMHCII^hiCD11b⁺ DCs and CD11c^hiMHCII^hiSIRPα⁺CD24^hi DCs in the lung, also with unaltered levels of IRF8. In contrast, splenic CD11b⁺CD4⁺ DCs were reduced by Irf4 deficiency, but the cells that were present had increased expression of IRF8. Strikingly, CD103⁺ DCs were absent in the lungs of CD11c-cre-Irf8^−/− mice. Irf8 deficiency also decreased splenic CD8α⁺ DCs and pre-cDCs, and the remaining pre-cDCs had increased expression of IRF4, which might explain the increased numbers of CD11b⁺ DCs observed in the spleen and lungs of CD11c-cre-Irf8^−/− mice. These results demonstrate that modulation of IRF4 and IRF8 drastically alters DC differentiation in a tissue-specific manner, with likely downstream effects on T cell priming and adaptive immunity.