Costimulation of dendritic epidermal gammadelta T cells by a new NKG2D ligand expressed specifically in the skin

Abstract

Dendritic epidermal T cells (DETCs) are a highly specialized population of gammadelta T cells that resides in the murine skin and participates in wound healing and tumor surveillance. Despite the expression of other stimulatory receptors on these cells, mechanisms involving activation have focused primarily on the invariant Vgamma3-Vdelta1 TCR expressed by DETCs. All DETCs also express the activating NKG2D receptor, but the role of NKG2D in DETC activation remains unclear, as does the identity of NKG2D ligands that are functionally expressed in the skin. In this study, we document the cloning of an NKG2D ligand H60c that is expressed specifically in the skin and in cultured keratinocytes and demonstrate its role in the activation of DETCs and NK cells. The ligand is unique among NKG2D ligands in being up-regulated in cultured keratinocytes, and its interaction with NKG2D is essential for DETC activation. Importantly, it is shown that engagement of NKG2D is not sufficient to activate DETCs, but instead provides a costimulatory signal that is nevertheless essential for activating DETCs in response to stimulation with keratinocytes.

FIGURE 1

Activation of NK cells by novel NKG2D ligand. A, The T cell lymphoma RMA was transduced with retroviruses encoding the indicated proteins or with empty vector (RMA-control). Cells were stained with NKG2D tetramers (open histogram) or streptavidin (shaded histogram). B, IL-2-stimulated NK cells were used as effector cells against the indicated target cells in a standard 4 hour ⁵¹Cr-release assay. NKG2D antibody (MI-6) or isotype control antibody (each at 50 μg/ml final concentration) were added as indicated. C, IL-2-activated NK cells were stimulated for 6 hours with the indicated RMA transductants. Accumulation of IFNγ in NK cells was determined by intracellular cytokine staining and electronic gating on NK1.1⁺CD3⁻ cells. The results are representative of three or more experiments.

FIGURE 2

Expression of H60c mRNA in tissues and in tumor cell lines. A and B, Total RNA (30 μg) from the indicated tissues (A) or cell lines (B) was electrophoresed on agarose gels and immobilized on a charged nylon membrane. The membrane was hybridized with a ³²P-labeled H60c probe (top panel). After the final exposure, the membranes were stripped and rehybridized with a GAPDH probe to control for loading of all lanes (middle panel). Ethidium bromide staining of ribosomal RNA in the gel lanes prior to transfer is shown at the bottom of panel A. As determined in other studies ((25), unpublished data), most of the cell lines examined in panel B expressed one or more of the other known NKG2D ligands (Rae1 (R), Mult1 (M) and/or H60a (A)), including Fibroblast (R, M), Nobo-1 (R, M), Yac-1 (R, M, A), SM-1 (R, M), RMA (no ligands), Tramp-C1 (R), DC2.1 (R, M), and B16-BL6 (no ligands). The results are representative of three experiments.

FIGURE 3

H60c functionally activates epidermal γδ T cells. A, Staining of dissociated skin cells shows that Vg3⁺ epidermal γδ T cells express NKG2D. Contour plots show NKG2D and Vγ3 TCR expression on ex vivo cells (18-hours post dissociation) (left panel) or on in vitro expanded (for 6 weeks) wild-type DETC (middle panel) or DETC from Klrk1^−/− (Klrk1 is the mouse NKG2D gene) (right panel). B, The indicated keratinocyte tumor cell lines and their H60c transductants were stained with NKG2D tetramers (open histograms) or streptavidin (shaded histograms). C and D, In vitro expanded wild-type and Klrk1^−/− DETC were used as effector cells against the indicated target cells in a standard 5 hour ⁵¹Cr-release assay. NKG2D monoclonal antibody (MI-6) or isotype control antibody was used to block interactions. The results are representative of three or more experiments.

FIGURE 4

Cultured primary keratinocytes express H60c and activate DETC. A, Dissociated C57BL/6 skin cells were cultured for the indicated period before analysis by flow cytometry. Cells were stained with antibodies that recognize Rae1, MULT1, or H60c (open histogram) or control antibodies (shaded histogram). B, Two color contour plot shows H60c and β4-integrin expression on ex vivo keratinocytes (18-hours post dissociation). C, Increased H60c transcripts in keratinocytes after in vitro culture for 18 hours, determined by quantitative RT-PCR. The data were normalized to GAPDH transcript amounts. D, Target cells used in panel E were stained with antibodies that recognize Rae1, MULT1, or H60c or with NKG2D tetramers (open histograms) or control antibodies or streptavidin (shaded histograms). E, In vitro expanded DETCs from wild-type or Klrk1^−/− mice were used as effector cells for cytotoxicity assays against keratinocytes harvested from primary cultures. NKG2D antibody (MI-6) or isotype control antibody was used to block interactions. Keratinocytes were in culture for 30 days. The results are representative of three or more experiments.

FIGURE 5

NKG2D stimulation costimulates DETC-mediated killing. A, FcγR⁺ Daudi cells and Daudi cells transduced with H60c (Daudi-H60c) were stained with NKG2D tetramers (open histogram) or streptavidin (shaded histogram). B, In vitro expanded DETC from wildtype or NKG2D-deficient mice pre-coated with NKG2D antibody (MI-6) or CD3ε antibody (145-2C11) were used as effector cells against Daudi target cells. C, IL-2 activated NK cells pre-coated with NKG2D antibody were used as effector cells against Daudi target cells. D, DETC pre-coated with a limiting dose of anti-CD3ε antibody (4 ng/ml) were used as effector cells against Daudi and Daudi-H60c target cells. The results are representative of three experiments.

FIGURE 6

NKG2D stimulation costimulates DETC degranlulation and IL-2 production. A and B, In vitro expanded DETC were stimulated with the indicated concentrations of either platebound CD3ε antibody (145-2C11) or NKG2D antibody (MI-6) for 4 hours. Degranulation (left of panel A, top of panel B) or intracellular IL-2 accumulation (right of panel A, bottom of panel B) were determined by flow cytometry. C, DETC were stimulated with a suboptimal concentration of platebound CD3ε antibody (1 ug/ml) plus increasing concentrations of either NKG2D antibody or control antibody. The results are representative of three or more experiments.

FIGURE 7

H60c mRNA is upregulated in wounded skin. C57BL/6 mice received full thickness wounds in their back skin. Relative levels of H60c mRNA were analyzed by quantitative RT-PCR in non-wounded and wounded skin. There was a dramatic increase in overall cellularity in the wounded tissue immediately after injury as a result of infiltrating lymphohematopoietic cells, which prevented the use of conventional housekeeping transcripts for normalization when comparing unwounded skin and wounded skin. Instead, given our finding that only CD104⁺ cells expressed H60c, we normalized the samples based on CD104 (integrin β4) transcript amounts. The results are representative of three experiments.