ZAP70-Independent Signaling Pathway

ORIGINAL ARTICLE

NKG2D Triggers Cytotoxicity in Murine Epidermal cd T Cells via PI3K-Dependent, Syk/ZAP70-Independent Signaling Pathway Atsuko Ibusuki1, Kazuhiro Kawai1,2, Shigeru Yoshida3,4, Youhei Uchida1, Ayano Nitahara-Takeuchi5, Kimiko Kuroki6, Mizuho Kajikawa7, Toyoyuki Ose6, Katsumi Maenaka6, Masanori Kasahara4 and Takuro Kanekura1 Murine epidermal gd T cells, known as dendritic epidermal T cells (DETCs), survey tissue stress through the invariant T-cell receptor (TCR) and non-clonotypic receptors such as NKG2D. NKG2D signaling via the DAP10phosphatidylinositol 3-kinase (PI3K) pathway directly stimulates cytotoxicity in natural killer (NK) cells and costimulates CD8 þ T cells to augment TCR signals. In activated murine NK cells, NKG2D signals also via the DAP12-Syk/ZAP70 pathway that triggers both cytotoxicity and cytokine production. It remains controversial whether NKG2D on DETCs is a primary activating receptor or functions only as a costimulatory receptor, and signaling pathways initiated by NKG2D ligation in DETCs have not been analyzed. We show that stimulation of short-term DETC lines with recombinant NKG2D ligands triggers degranulation (exocytosis of cytotoxic granules) via the PI3K-dependent signaling pathway, but does not induce cytokine production or Syk/ ZAP70 activation. Coengagement of TCR or Syk/ZAP70 signaling was not crucial for DETC-mediated killing of NKG2D ligand-expressing target cells. Thus, NKG2D can function as a coactivating stress receptor that directly triggers cytotoxicity in DETCs, at least after priming, via the PI3K-dependent, Syk/ZAP70-independent signaling pathway. Journal of Investigative Dermatology (2014) 134, 396–404; doi:10.1038/jid.2013.353; published online 19 September 2013

INTRODUCTION Most epithelial tissues contain resident T cells that express tissue-specific invariant or restricted gd T-cell receptors (TCRs) (Witherden and Havran, 2011). These epithelial gd T cells mediate stress surveillance in the given tissue and exert various effector and regulatory functions to maintain tissue integrity (Hayday, 2009; Bonneville et al., 2010; Witherden and Havran, 1

Department of Dermatology, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan; 2Department of Dermatology, Kido Hospital, Niigata, Japan; 3Department of Medical Laboratory Science, Faculty of Health Sciences, Hokkaido University, Sapporo, Japan; 4Department of Pathology, Hokkaido University Graduate School of Medicine, Sapporo, Japan; 5Division of Dermatology, Niigata University Graduate School of Medical and Dental Sciences, Niigata, Japan; 6Laboratory of Biomolecular Science, Faculty of Pharmaceutical Sciences, Hokkaido University, Sapporo, Japan and 7Laboratory for Infectious Immunity, Research Center for Allergy and Immunology (RCAI), RIKEN Yokohama Institute, Yokohama, Japan Correspondence: Kazuhiro Kawai, Department of Dermatology, Kagoshima University Graduate School of Medical and Dental Sciences, 8-35-1 Sakuragaoka, Kagoshima 890-8544, Japan. E-mail: [email protected] Abbreviations: DETC, dendritic epidermal T cell; MFI, mean fluorescence intensity; NK, natural killer; NKG2DL, NKG2D ligand; PI3K, phosphatidylinositol 3-kinase; r, recombinant; TCR, T-cell receptor Received 18 May 2013; revised 17 July 2013; accepted 29 July 2013; accepted article preview online 20 August 2013; published online 19 September 2013

396

Journal of Investigative Dermatology (2014), Volume 134

2011). Dendritic epidermal T cells (DETCs) are prototypic epithelial gd T cells that reside in the murine skin and express a canonical Vg3Vd1 (nomenclature according to Garman et al. (1986)) TCR (Witherden and Havran, 2011). DETCs secrete growth factors, cytokines, and chemokines, and release cytotoxic granules in response to tissue stress, and contribute to wound healing (Jameson et al., 2002), epidermal homeostasis (Sharp et al., 2005; Girardi et al., 2006), hair cycle regulation (Kloepper et al., 2013), immunoregulation (Shiohara et al., 1996; Girardi et al., 2002; Strid et al., 2011), and tumor surveillance (Girardi et al., 2001; Strid et al., 2008). DETCs recognize an as-yet-undetermined self ligand induced on stressed, damaged, or transformed keratinocytes through the invariant TCR (Havran et al., 1991; Tigelaar and Lewis, 1994; Jameson et al., 2004). In situ immunofluorescence staining with soluble DETC TCR tetramers demonstrated that the DETC TCR ligand is undetectable in the normal epidermis, but upregulated on periwound keratinocytes rapidly and transiently after wounding (Komori et al., 2012). Accumulating evidence suggests, however, that low levels of the DETC TCR ligand, which might not be detectable by immunofluorescence staining, are expressed on healthy keratinocytes in vivo (Mallick-Wood et al., 1998; Aono et al., 2000; Hara et al., 2000; Minagawa et al., 2001; Jameson et al., 2004; Leandersson et al., 2006). Constitutive expression of the DETC TCR ligand in the normal epidermis & 2014 The Society for Investigative Dermatology

A Ibusuki et al. NKG2D-Mediated Signal in Epidermal gd T Cells

is supported by the recent observation using intravital dynamics–immunosignal correlative microscopy that TCRs on DETCs are clustered and triggered at steady state in immunological synapse-like structures on the apical dendrites located near the keratinocyte tight junctions (Chodaczek et al., 2012). In response to tissue stress, these TCR-triggered proximal signals are relocated from the apical dendrites to the newly formed synapses, but whole-cell amount of TCR signals does not increase (Chodaczek et al., 2012). Therefore, stress-induced DETC activation would require relocalization of the TCR–ligand interactions and/or additional signals through non-TCR stress receptors such as junctional adhesion molecule-like protein (JAML) (Witherden et al., 2010), CD100 (Witherden et al., 2012), and NKG2D (Jamieson et al., 2002). NKG2D is a homodimeric C-type lectin–like receptor expressed on natural killer (NK) cells, CD8 þ T cells (only after activation in mice), and subsets of CD4 þ T cells (in humans), NKT cells, and gd T cells (Raulet, 2003). NKG2D recognizes a diverse family of major histocompatibility complex class I–related proteins that are poorly expressed on healthy cells, but upregulated on stressed, infected, or transformed cells (Raulet, 2003; Champsaur and Lanier, 2010; Raulet et al., 2013). NKG2D ligands (NKG2DLs) comprise three members of the histocompatibility 60 family (H60a–c), five members of the retinoic acid early inducible-1 family (Rae-1a-e), and murine UL16-binding protein–like transcript 1 (MULT1) in mice, and major histocompatibility complex class I chain–related proteins A (MICA) and B (MICB), and six UL16-binding proteins (ULBP1–6) in humans (Champsaur and Lanier, 2010; Raulet et al., 2013). Functional outcome of NKG2D ligation varies depending on the cell type and activation state. This is, at least in part, explained by the differential usage of two transmembrane adaptor proteins, DAP10 and DAP12, that associate with NKG2D and transduce signals (Raulet, 2003; Lopez-Larrea et al., 2008). In mice, alternative splicing generates two functionally distinct NKG2D isoforms, NKG2D-L (long) and NKG2D-S (short), that differ in the cytoplasmic domain and associated signaling adaptors (Diefenbach et al., 2002). NKG2D-L associates with DAP10, whereas NKG2D-S, which is induced on murine NK cells after activation, can associate with both DAP10 and DAP12 (Diefenbach et al., 2002). Human NKG2D is an equivalent of murine NKG2D-L and associates only with DAP10 (Wu et al., 1999, 2000; Billadeau et al., 2003; Andre et al., 2004; Rosen et al., 2004). DAP10 contains a YXNM motif (Wu et al., 1999) that recruits the p85a subunit of phosphatidylinositol 3-kinase (PI3K) (Wu et al., 1999; Billadeau et al., 2003; Upshaw et al., 2006) as well as Grb2 and Vav1 (Billadeau et al., 2003; Cella et al., 2004; Graham et al., 2006; Upshaw et al., 2006). DAP12 contains an immunoreceptor tyrosine-based activation motif (ITAM) that recruits Syk family protein tyrosine kinases, Syk and ZAP70 (Lanier et al., 1998). NKG2D-DAP10mediated PI3K and Grb2-Vav1 signaling triggers cytotoxicity in both murine and human NK cells, but is insufficient for inducing cytokine production (Wu et al., 2000; Diefenbach et al., 2002; Gilfillan et al., 2002; Billadeau et al., 2003;

Zompi et al., 2003; Cella et al., 2004; Graham et al., 2006; Upshaw et al., 2006). By contrast, NKG2D-S-DAP12mediated Syk/ZAP70 signaling triggers both cytotoxicity and cytokine production in activated murine NK cells (Diefenbach et al., 2002; Gilfillan et al., 2002; Zompi et al., 2003). As most T cells do not express DAP12, NKG2D signals solely through DAP10 in murine as well as human CD8 þ T cells (Lanier et al., 1998; Diefenbach et al., 2002; Gilfillan et al., 2002). Unlike NK cells, NKG2D engagement alone is generally insufficient to activate CD8 þ T cells, and provides only costimulatory signals that augment TCR signals (Groh et al., 2001; Roberts et al., 2001; Diefenbach et al., 2002; Jamieson et al., 2002; Maasho et al., 2005) under restricted conditions (Ehrlich et al., 2005; Rajasekaran et al., 2010). DETCs can be activated simply by acute, keratino cyte-specific upregulation of NKG2DLs (presumably without upregulation of the DETC TCR ligand) in vivo (Strid et al., 2008, 2011). Without NKG2DL recognition, promotion of wound healing by DETCs is impaired (Jung et al., 2012; Yoshida et al., 2012). NKG2D has also been implicated in DETC-mediated tumor surveillance (Girardi et al., 2001; Oppenheim et al., 2005; Strid et al., 2008). It has been controversial, however, whether NKG2D directly stimulates DETC activation or serves only as a costimulatory receptor in DETCs. We reported previously that unlike most other T cells, DETCs constitutively express both NKG2D-L and NKG2D-S isoforms as well as DAP10 and DAP12 adaptors, and NKG2D ligation alone is sufficient to trigger both cytotoxicity and cytokine production in DETCs in the absence of TCR coengagement (Nitahara et al., 2006). Another group, conversely, reported that NKG2D functions only as a costimulatory receptor in DETCs (Whang et al., 2009). In this study, we reanalyzed the functional outcome of NKG2D engagement in DETCs using recombinant NKG2DL proteins, because different mAbs were used for NKG2D crosslinking in the previous studies (Nitahara et al., 2006; Whang et al., 2009). We show that NKG2D engagement with NKG2DLs triggers degranulation via the PI3K-dependent signaling pathway, but does not induce cytokine production or Syk/ZAP70 activation in DETCs. Consistent with these data, coengagement of TCR or Syk/ZAP70 signaling was not crucial for DETC-mediated lysis of NKG2DL-expressing target cells. RESULTS Recombinant NKG2DL proteins stimulate degranulation but not cytokine production in DETCs

Consequence of NKG2D ligation in DETCs was controversial in the previous studies (Nitahara et al., 2006; Whang et al., 2009). We confirmed that this was caused by the difference in the mAbs used for NKG2D crosslinking (Supplementary Text 1 and Supplementary Figure S1 online). To analyze the functional outcome of NKG2D ligation in DETCs in a more physiological setting, we stimulated DETCs with recombinant NKG2DL proteins instead of anti-NKG2D mAbs. Stimulation of DETCs with plate-bound H60a-Fc, Rae-1d-Fc, and Rae-1e-Fc fusion proteins induced degranulation (Figure 1a) but virtually no IFN-g, tumor necrosis factor a, or IL-13 production (Figure 1b and data not shown). The degranulation www.jidonline.org

397

A Ibusuki et al. NKG2D-Mediated Signal in Epidermal gd T Cells

H60a-Fc

H60a-Fc

***

50 40

50 40

**

30

30

**

20

20 10

10

0 0.1

1

CD107a+ cells (%)

40

0.1

**

20

* 0.1

1

10

30 20 10

100

0.1

100

NKG2DL triggers degranulation in DETCs via the PI3Kdependent, Syk/ZAP70-independent signaling pathway

20

**

10

*

0 0.1

1

10

100

0.1

1

10

100

Concentrations of immobilized NKG2DL-Fc (μg ml–1)

rH60a

50

rH60a

50

rNKG2DL Control

40

40

30

30

***

20

20

***

10

10 0.1

1

10 100

rH60c C150S

40

IFN-γ+ cells (%)

CD107a+ cells (%)

10

0 0.1 50

1

10 100

rH60c C150S

40 30

30

**

***

20 10 0

0 0.1 1 10 100 Concentrations of immobilized rNKG2DL (μg ml–1)

0.1 1 10 100 Concentrations of immobilized rNKG2DL (μg ml–1)

Figure 1. Recombinant NKG2D ligand (NKG2DL) proteins stimulate degranulation but not cytokine production in dendritic epidermal T cells (DETCs). (a, b) DETCs were stimulated with indicated NKG2DL-Fc or control-Fc, and (a) degranulation (cell surface CD107a expression) and (b) intracellular IFN-g production were analyzed by flow cytometry. (c, d) DETCs were stimulated with indicated recombinant extracellular domain proteins of NKG2DL (rNKG2DL) or control protein (BSA), and (c) degranulation and (d) intracellular IFN-g production were analyzed by flow cytometry. Data are expressed as mean positive cells (%) and SD. Representative data from three independent experiments are shown. Significant differences as compared with control are denoted with asterisks (*Po0.05, **Po0.01, ***Po0.001).

398

responses were completely blocked with soluble anti-NKG2D mAb (Supplementary Figure S2 online). H60c is an NKG2DL that is predominantly expressed in the skin (Takada et al., 2008; Whang et al., 2009; Yoshida et al., 2012) and upregulated on stressed or damaged keratinocytes (Whang et al., 2009; Yoshida et al., 2012). As spontaneously formed recombinant H60c (rH60c) dimers bind to NKG2D with very low affinity (Yoshida et al., 2012), the monomeric form of the recombinant extracellular domain protein, rH60c C150S, was used for DETC stimulation. Similar to the NKG2DL-Fc fusion proteins, plate-bound rH60c C150S induced degranulation (Figure 1c) but not IFN-g production in DETCs (Figure 1d). The degranulation responses to rH60c C150S were lower than those to NKG2DL-Fc fusion proteins, but this would be due to the inefficient immobilization or inappropriate dimensions of the recombinant extracellular domain proteins of NKG2DL, because rH60a also induced lower responses than H60a-Fc (Figure 1a and c).

30

Concentrations of immobilized NKG2DL-Fc (μg ml–1)

10

1

40

**

20

20

*

50

***

30

50

** Rae-1ε-Fc

***

40

0

100

0

50

0

10

40

Rae-1ε-Fc

10

1

Rae-1δ-Fc

50

*** ***

30

0

100

Rae-1δ-Fc

50

10

10

IFN-γ+ cells (%)

0

NKG2DL-Fc Control-Fc

Journal of Investigative Dermatology (2014), Volume 134

In murine NK cells, either NKG2D-DAP10-mediated PI3K/ Grb2-Vav1 signaling or NKG2D-S-DAP12-mediated Syk/ ZAP70 signaling is sufficient to trigger cytotoxicity, but the latter signaling pathway is crucial for cytokine production (Diefenbach et al., 2002; Gilfillan et al., 2002; Zompi et al., 2003). Induction of degranulation but not cytokine production by NKG2D engagement with NKG2DLs (Figure 1) suggests that NKG2DLs cannot trigger NKG2D-S-DAP12 signaling in DETCs. As expected, stimulation of DETCs with H60a-Fc rapidly induced PI3K-dependent phosphorylation of Akt, which was confirmed by blocking with a PI3K-specific inhibitor LY294002 (Figure 2a), but did not trigger phosphorylation of Syk/ZAP70 (Figure 2b). The degranulation responses of DETCs to H60a-Fc were also blocked with the PI3K inhibitor in a dose-dependent manner, whereas Syk/ZAP70-specific inhibitor piceatannol showed modest blocking only at a high concentration (50 mM) owing to the off-target effects (Colucci et al., 2002) (Figure 3). In contrast, IFN-g production in DETCs on stimulation with the potent anti-NKG2D mAb (clone 191004) was dose-dependently attenuated with the Syk/ZAP70 inhibitor (Figure 3). TCR signaling is not necessary for DETC-mediated killing of NKG2DL-expressing target cells

We next determined whether DETCs can kill NKG2DLexpressing target cells in the absence of TCR coengagement. For this purpose, we used H60a- and H60c-transduced Ba/F3 cells and primary cultured keratinocytes from TCRd  /  mice on a C57BL/6 background as targets cells for the cytotoxicity assay. Mock-transduced Ba/F3 cells that expressed only low levels of MULT1 but not other NKG2DLs or the DETC TCR ligand (Supplementary Figure S3 online) were not killed by DETCs (Figure 4a). Expression of either H60a or H60c alone, without induction of other NKG2DLs or the DETC TCR ligand (Supplementary Figure S3 online), was sufficient to convert

A Ibusuki et al. NKG2D-Mediated Signal in Epidermal gd T Cells

H60a-Fc

H60a-Fc

***

100

*** ***

Fold change in MFI of phosphorylated Akt

3

***

80 60

2

40 1

20

0

0 0

1

5

15

100

30

101

102

103

H60a-Fc LY294002-treated

4

100

3

80 60

2

40 1

Fold change in MFI of phosphorylated Syk/ZAP70

4

4

100

3

80 60

2

40 1

20

0

0 0

1

5

15

***

1 5 15 30 Stimulation time (minutes)

0

103

80

3

*** 2

60

** 40 20

0 100 101 102 103 Phosphorylated Akt

102

100

20 0

101

H2O2 4

1

0

100

30

0 0

5 15 30 1 Stimulation time (minutes)

100 101 102 103 Phosphorylated Syk/ZAP70

Figure 2. NKG2D ligand (NKG2DL) triggers phosphatidylinosital 3-kinase (PI3K) but not Syk/ZAP70 signaling in dendritic epidermal T cells (DETCs). DETCs were preincubated with DMSO or 10 mM LY294002, and stimulated with H60a-Fc or hydrogen peroxide (H2O2) (positive control for Syk/ZAP70 phosphorylation; Schieven et al., 1994; Haas et al., 2008). Intracellular expression levels of (a) phosphorylated Akt and (b) Syk/ZAP70 were analyzed by flow cytometry. Fold change in mean fluorescence intensity (MFI) was determined as geometric MFI at indicated time point/geometric MFI before stimulation. Data are expressed as mean fold change and SD. Significant increases in the phosphorylation on stimulation are denoted with asterisks (**Po0.01, ***Po0.001). Representative profiles at 5 minutes (open histograms) from three independent experiments are also shown. Shaded histograms indicate cells before stimulation.

Ba/F3 cells to susceptible target cells for DETC-mediated killing, and the lysis was completely blocked with antiNKG2D mAb (Figure 4a). DETC-mediated killing of the H60c-Ba/F3 cells was also blocked with anti-H60c mAb but not with anti-TCR mAb (Figure 4b). Consistent with the previous report (Whang et al., 2009), H60c but not other NKG2DLs was induced on cultured keratinocytes (H60a protein cannot be expressed in the C57BL/6 background and H60b is not expressed on keratinocytes; Takada et al., 2008; Whang et al., 2009; Yoshida et al., 2012) (Supplementary Figure S3 online). We also confirmed expression of the DETC TCR ligand on cultured keratinocytes using the soluble DETC TCR (Supplementary Figure S3 online). DETCs could efficiently kill cultured keratinocytes (Figure 5), but in contrast to the absolute dependence on NKG2D in the lysis of NKG2DL-transduced Ba/F3 cells (Figure 4), the lysis of keratinocytes by DETCs was blocked only partially with the anti-NKG2D mAb (Figure 5). Complete inhibition of DETC-mediated killing of keratinocytes with another anti-NKG2D mAb (clone MI6) was reported previously (Whang et al., 2009), but we found that the mAb MI6 inhibited the lysis less efficiently than the blocking mAb 191004 used in this study (data not shown). The DETCmediated lysis of keratinocytes was also partially blocked with anti-CD3 and anti-TCR mAbs (Figure 5), and additive blockade was observed in the presence of both anti-NKG2D

and anti-TCR/CD3 mAbs (Figure 5b). Thus, DETC-mediated killing of stressed keratinocytes can occur in the absence of either NKG2D or TCR ligation, indicating that both receptors act as independent coactivating receptors in DETCs. As expected, the lysis of keratinocytes by DETCs was completely abolished with the PI3K inhibitor (Figure 6). In contrast, inhibition of Syk/ZAP70 only partially blocked the lysis (Figure 6). DISCUSSION We showed that NKG2D engagement with NKG2DLs alone is sufficient to trigger degranulation but not cytokine production in DETCs. Contrary to our earlier conclusion (Nitahara et al., 2006), NKG2DL-triggered DETC activation is mediated via the PI3K-dependent signaling pathway, but not via the NKG2D-SDAP12-Syk/ZAP70 pathway. Although DETCs constitutively express NKG2D-S and DAP12 transcripts (Nitahara et al., 2006) and can secrete IFN-g when stimulated with some potent anti-NKG2D mAbs (Nitahara et al. (2006) and Supplementary Figure S1b online), NKG2DLs may not be able to engage enough NKG2D-S-DAP12 receptor complexes, which might be fewer than NKG2D-DAP10 on DETCs, to trigger Syk/ ZAP70 signaling. We also showed that DETCs can kill NKG2DL-transduced Ba/F3 target cells that do not express the TCR ligand. Our results that DETCs cannot kill mock-transduced Ba/F3 cells www.jidonline.org

399

A Ibusuki et al.

Inhibition of degranulation (%)

NKG2D-Mediated Signal in Epidermal gd T Cells

100

100

H60a-Fc

80

80

60

***

40

***

40

***

20

20

***

0

0

Inhibition of degranulation (%)

0

3.1 6.3 12.5 25 50 LY294002 (μM)

100

H60a-Fc

100 101 102 103 CD107a

100

80

80

60

60

40

40

***

20 0

20 0

0

Inhibition of IFN-γ production (%)

60

3.1 6.3 12.5 25 50 Piceatannol (μM)

Anti-NKG2D (191004)

100 80

100 101 102 103 CD107a

100 80

***

60

***

40

***

20

**

0

60 40 20 0

0

3.1 6.3 12.5 25 50 Piceatannol (μM)

100 101 102 103 IFN-γ

Figure 3. NKG2D ligand (NKG2DL) triggers degranulation in dendritic epidermal T cells (DETCs) via the phosphatidylinositol 3-kinase (PI3K)– dependent, Syk/ZAP70-independent signaling pathway. DETCs were preincubated with indicated concentrations of LY294002 or piceatannol, and stimulated with 10 mg ml  1 of immobilized H60a-Fc or anti-NKG2D mAb (clone 191004). Degranulation and intracellular IFN-g production were analyzed by flow cytometry. Inhibition is expressed as a percentage of CD107a þ and IFN-g þ cells observed in the absence of the inhibitors (mean and SD). Significant inhibition of the degranulation and IFN-g production as compared with those in the absence of the inhibitors is denoted with asterisks (**Po0.01, ***Po0.001). Representative profiles in the presence of 25 mM LY294002 or piceatannol (open histograms) from three independent experiments are also shown. Shaded histograms indicate cells in the absence of the inhibitors.

and that DETC-mediated lysis of NKG2DL-transduced Ba/F3 cells is completely blocked with anti-NKG2D and anti-H60 mAbs, but not with anti-TCR mAb, indicate that NKG2D can act as a primary activating receptor in DETC-mediated killing of NKG2DL-expressing target cells. In contrast to our results, a previous study demonstrated that DETCs cannot kill the Fc receptor–bearing human B cell lymphoma Daudi cells or murine mastocytoma P815 cells coated with anti-NKG2D mAb (Whang et al., 2009), but this could be expected because the ‘‘non-stimulating’’ anti-NKG2D mAb (clone MI6) was used for the redirected cytotoxicity assay. It was also reported that DETCs cannot kill H60c-transduced Daudi cells (Whang et al., 2009). The basis of this discrepancy is unknown, but the levels of H60c expression on H60ctransduced Daudi cells might be low to trigger NKG2D 400

Journal of Investigative Dermatology (2014), Volume 134

signaling in DETCs. Alternatively, murine Ba/F3 cells, but not human Daudi cells, might express ligands for additional receptors on DETCs that augment NKG2D signals to induce cytotoxicity. In this case, NKG2D functions as a synergistic coactivating receptor rather than the primary activating receptor. Nevertheless, TCR-independent stimulation of DETC cytotoxicity through NKG2D is distinct from the classical costimulation in conventional CD8 þ T cells, for which TCR signaling is crucial. Partial blocking of the DETC-mediated lysis of keratinocytes with both anti-NKG2D and anti-TCR mAbs is consistent with the previous study using the PDV-transformed keratinocyte cell line as target cells (Girardi et al., 2001) and with our prior observation that DETC-mediated lysis of the Pam 212– transformed keratinocyte cell line, which expresses the DETC TCR ligand but only low levels of H60, is completely inhibited with anti-TCR mAb alone (Nitahara et al. (2006), Ito et al. (2008), and unpublished data). Contrary to these results, a previous study demonstrated that DETC-mediated killing of keratinocytes was absolutely dependent on NKG2D (Whang et al., 2009). In that study, however, the lysis of keratinocytes by DETCs was generally low as compared with that observed in our study (Whang et al., 2009), suggesting that some additional coactivating/costimulatory receptors other than NKG2D might be lost during preparation of DETCs. Involvement of additional receptor–ligand interactions in the DETC-mediated lysis of keratinocytes is also supported by the residual cytotoxicity in the presence of both antiNKG2D and anti-TCR mAbs (Figure 5b). DETCs may recognize keratinocytes through a set of receptors including NKG2D and TCR that bind to stress-induced as well as constitutively expressed ligands, and these receptors may act as synergistic coactivating/costimulatory receptors to induce cytotoxicity. Syk/ZAP70 act upstream of PI3K (Jiang et al., 2002; LopezLarrea et al., 2008) and are crucial not only for NKG2D-SDAP12 signaling but also for TCR proximal signaling (Raulet, 2003; Lopez-Larrea et al., 2008). As DETC-mediated lysis of keratinocytes is completely abolished with the PI3K inhibitor, but only partially blocked with the Syk/ZAP70 inhibitor, NKG2D-DAP10 signaling via the PI3K/Grb2-Vav1 pathway alone is sufficient to trigger cytotoxicity against keratinocytes in DETCs in the absence of both NKG2D-S-DAP12 and TCR signals. As we used short-term DETC lines that had been stimulated through TCR and expanded in the presence of IL-2, the TCRindependent promiscuous cytotoxicity through NKG2D might be induced during in vitro culture (Supplementary Text 2 online). In fact, NKG2D cannot trigger cytotoxicity in freshly isolated, purified DETCs ex vivo or after in vitro culture for 24 hours in the presence of IL-15, which upregulates NKG2D expression (Roberts et al., 2001) and primes NKG2D-DAP10 signaling to induce cytotoxicity (Horng et al., 2007) (unpublished data). In general, however, freshly isolated DETCs do not exhibit cytotoxicity ex vivo, and their acquisition of cytotoxicity requires 10–14 days of in vitro culture in the presence of IL-2 (Nixon-Fulton et al., 1988; Kaminski et al., 1993). Therefore, DETCs may have to be

A Ibusuki et al. NKG2D-Mediated Signal in Epidermal gd T Cells

H60c-Ba/F3 80

60

60

60

40

40

40

20

20

20

0

0 1

3

10 30

** * ** ***

0

** *** *** *** 1

1 3 10 30 Effector/target ratio

3

***

NS

80 60 40 20 0

10 30

None

80

100

Anti-TCR

80

Control Ig Anti-NKG2D

Control Ig

100

Anti-H60c

H60a-Ba/F3 100

Specific lysis (%)

Specific lysis (%)

Mock-Ba/F3 100

Figure 4. TCR signaling is not necessary for dendritic epidermal T cell (DETC)–mediated killing of NKG2D ligand (NKG2DL)–transduced target cells. (a) DETCs and indicated target cells were coincubated in the presence of isotype control or blocking anti-NKG2D mAb (clone 191004). (b) DETCs and H60c-Ba/F3 cells were coincubated at the effector/target ratio of 10 in the presence of isotype control or anti-H60c mAb (clone 5G6), or in the absence or presence of Fab fragments of anti-TCR mAb. Data are expressed as mean specific lysis (%) and SD. Representative data from three independent experiments are shown. Significant inhibition of the cytotoxicity as compared with control is denoted with asterisks (*Po0.05, **Po0.01, ***Po0.001). NS, not significant.

None

Anti-NKG2D

Anti-CD3

Specific lysis (%)

Inhibition of cytotoxicity (%)

100

100 80

80

***

**

60

60 40

***

40

**

20 0

100

* 20

*

0

3 10 1 30 Effector/target ratio

*

1 3 10 30 Effector/target ratio

Specific lysis (%)

100

*

80

***

80 60

***

*** ***

***

40 20

100 Inhibition of cytotoxicity (%)

Control Ig

80

***

60 40 20 0

0 0

3.1 6.3 12.5 25 50 LY294002 (μM)

0

3.1 6.3 12.5 25 50 Piceatannol (μM)

Figure 6. Syk/ZAP70 signaling is not crucial for dendritic epidermal T cell (DETC)–mediated killing of keratinocytes. DETCs were preincubated with indicated concentrations of LY294002 or piceatannol, and coincubated with keratinocytes at the effector/target ratio of 10. Inhibition of the cytotoxicity is expressed as a percentage of specific lysis observed in the absence of the inhibitors (mean and SD). Representative data from three independent experiments are shown. Significant inhibition of the cytotoxicity as compared with that in the absence of the inhibitors is denoted with asterisks (***Po0.001).

** *

60

*

40 20 0 –

+

–

+

–

+

Anti-NKG2D

–

–

+

+

–

–

Anti-TCR

–

–

–

–

+

+

Anti-CD3

Figure 5. TCR signaling is not crucial for dendritic epidermal T cell (DETC)–mediated killing of keratinocytes. (a) DETCs and keratinocytes were coincubated in the presence of isotype control or anti-NKG2D mAb (clone 191004), or in the absence or presence of Fab fragments of anti-CD3 mAb. (b) DETCs and keratinocytes were coincubated at the effector/target ratio of 10 in the absence or presence of anti-NKG2D mAb (191004) and/or Fab fragments of anti-TCR or antiCD3 mAb. Data are expressed as mean specific lysis (%) and SD. Representative data from three independent experiments are shown. Significant differences in the cytotoxicity are denoted with asterisks (*Po0.05, **Po0.01, ***Po0.001).

preactivated and/or primed with IL-2/IL-15 to exert cytotoxicity through activating receptors including but not limited to NKG2D. Once primed, however, DETCs can mount

cytotoxicity only by recognizing stress ligands through NKG2D and/or other non-TCR receptors in the absence of TCR engagement. Priming of DETCs may occur under some conditions in vivo, as suggested by our observation in vivo (Supplementary Text 3 and Supplementary Figure S4 online). In summary, we have demonstrated that NKG2D-mediated recognition of stress ligands is sufficient to trigger DETC cytotoxicity via the PI3K-dependent, Syk/ZAP70-independent signaling pathway, at least under restricted conditions. Unlike the classical two-signal model in conventional T cells, DETC activation may not always require simultaneous coengagement of TCR and rather is regulated by the balance between positive and negative signals provided through various coactivating, costimulatory, and coinhibitory receptors. Expression of multiple independent stress receptors such as NKG2D would be beneficial for DETCs not only to mount rapid, innate-like responses to tissue stress but also to avoid inappropriate activation by accidentally expressed stress ligands (Supplementary Text 4 online). www.jidonline.org

401

A Ibusuki et al. NKG2D-Mediated Signal in Epidermal gd T Cells

MATERIALS AND METHODS Cells Short-term DETC lines from C57BL/6J mice were generated in each experiment as described previously (Uchida et al., 2011). Ba/F3 cells expressing H60a (H60a-Ba/F3) and H60c (H60c-Ba/F3) were generated as described (Takada et al., 2008). Epidermal keratinocytes from TCRd  /  mice on a C57BL/6 background were cultured for 10 days as described (Yano and Okochi, 2005).

Stimulation assay Fusion proteins of the extracellular domain of NKG2DL (H60a, Rae1d, or Rae-1e) and the Fc region of human IgG1 (NKG2DL-Fc) and control human IgG1-Fc protein (control-Fc) were purchased from R&D Systems (Minneapolis, MN). Recombinant extracellular domain proteins of NKG2DL were produced in Escherichia coli as described previously (Takada et al., 2008; Yoshida et al., 2012). In the rH60c protein, the cysteine residue at position 150 was substituted with serine (rH60c C150S) to prevent dimer formation (Yoshida et al., 2012). For stimulation assays, 96-well plates were coated overnight at 4 1C with indicated concentrations of NKG2DL-Fc or control-Fc, or recombinant extracellular domain proteins of NKG2DL or BSA (Invitrogen, Carlsbad, CA) as control protein. DETCs (1  105) were stimulated for 4 hours on the plates, and degranulation and intracellular cytokine production were analyzed by flow cytometry as described (Uchida et al., 2011). In some experiments, DETCs were preincubated for 15 minutes at 4 1C with the predetermined saturating concentration (50 mg ml  1) of blocking anti-NKG2D (clone 191004; R&D Systems) or isotype control mAb (eBioscience, San Diego, CA), or for 30 minutes at 37 1C with indicated concentrations of LY294002 (Cell Signaling Technology, Danvers, MA), piceatannol (SigmaAldrich, St Louis, MO), or DMSO as vehicle control before stimulation.

Intracellular phospho-flow cytometry DETCs were preincubated for 30 minutes at 37 1C with 10 mM LY294002 or DMSO before stimulation with H60a-Fc or 10 mM H2O2 (Schieven et al., 1994; Haas et al., 2008). To crosslink H60aFc, DETCs incubated for 20 minutes on ice with 10 mg ml  1 H60a-Fc were further incubated for 20 minutes with 10 mg ml  1 F(ab’)2 goat anti-human Fcg antibody (Jackson ImmunoResearch, West Grove, PA) before being transferred to a 37 1C water bath. At indicated time points, stimulated cells were fixed for 10 minutes at 37 1C with 2% paraformaldehyde (BD Cytofix Fixation Buffer; BD Biosciences, San Jose, CA) and permeabilized for 30 minutes on ice with 90% methanol (BD Phosflow Perm Buffer III; BD Biosciences). After preincubation with anti-CD16/CD32 mAb (2.4G2; BD Biosciences), cells were stained with phycoerythrin-conjugated anti-phosphorylated Akt (Ser473) mAb (M89-61; BD Biosciences) or unlabeled rabbit anti-phosphorylated Syk (Tyr352)/ZAP70 (Tyr319) antibody (Cell Signaling Technology), followed by phycoerythrin-conjugated F(ab’)2 donkey anti-rabbit IgG antibody (eBioscience). Expression levels of the intracellular phosphoproteins were determined by flow cytometry.

Cytotoxicity assay DETCs and target cells were coincubated for 4 hours, and specific lysis was determined as described (Uchida et al., 2011). Where stated, effector cells were preincubated for 15 minutes at 4 1C with 402

Journal of Investigative Dermatology (2014), Volume 134

50 mg ml  1 of blocking anti-NKG2D (clone 191004) or isotype control mAb and/or the predetermined saturating concentration (1 mg ml  1) of Fab fragments of anti-TCR Cd (UC7-13D5; a gift from G Matsuzaki, Molecular Microbiology Group, Department of Tropical Infectious Diseases, Tropical Biosphere Research Center, University of the Ryukyus, Okinawa, Japan) or anti-CD3 (KT3; prepared in our laboratory) mAb, or for 30 minutes at 37 1C with indicated concentrations of LY294002, piceatannol, or DMSO before addition to the target cells. Blocking with anti-H60c mAb (clone 5G6) was performed as described (Yoshida et al., 2012). Further information about other materials and methods used in this study is provided in the Supplementary Material online. CONFLICT OF INTEREST The authors state no conflict of interest.

ACKNOWLEDGMENTS We thank Nobue Uto and Sawako Komure for technical assistance, Rania Hassan Mohamed for mAb purification, Hideo Yagita for the HMG2D mAb, John Kappler and Philippa Marrack for the pBACp10pH plasmid, and Goro Matsuzaki for the Fab fragments of the UC7-13D5 mAb. SUPPLEMENTARY MATERIAL Supplementary material is linked to the online version of the paper at http:// www.nature.com/jid

REFERENCES Andre P, Castriconi R, Espeli M et al. (2004) Comparative analysis of human NK cell activation induced by NKG2D and natural cytotoxicity receptors. Eur J Immunol 34:961–71 Aono A, Enomoto H, Yoshida N et al. (2000) Forced expression of terminal deoxynucleotidyl transferase in fetal thymus resulted in a decrease in gammadelta T cells and random dissemination of Vgamma3Vdelta1 T cells in skin of newborn but not adult mice. Immunology 99:489–97 Billadeau DD, Upshaw JL, Schoon RA et al. (2003) NKG2D-DAP10 triggers human NK cell-mediated killing via a Syk-independent regulatory pathway. Nat Immunol 4:557–64 Bonneville M, O’Brien RL, Born WK (2010) Gammadelta T cell effector functions: a blend of innate programming and acquired plasticity. Nat Rev Immunol 10:467–78 Cella M, Fujikawa K, Tassi I et al. (2004) Differential requirements for Vav proteins in DAP10- and ITAM-mediated NK cell cytotoxicity. J Exp Med 200:817–23 Champsaur M, Lanier LL (2010) Effect of NKG2D ligand expression on host immune responses. Immunol Rev 235:267–85 Chodaczek G, Papanna V, Zal MA et al. (2012) Body–barrier surveillance by epidermal gammadelta TCRs. Nat Immunol 13:272–82 Colucci F, Schweighoffer E, Tomasello E et al. (2002) Natural cytotoxicity uncoupled from the Syk and ZAP-70 intracellular kinases. Nat Immunol 3:288–94 Diefenbach A, Tomasello E, Lucas M et al. (2002) Selective associations with signaling proteins determine stimulatory versus costimulatory activity of NKG2D. Nat Immunol 3:1142–9 Ehrlich LI, Ogasawara K, Hamerman JA et al. (2005) Engagement of NKG2D by cognate ligand or antibody alone is insufficient to mediate costimulation of human and mouse CD8 þ T cells. J Immunol 174: 1922–31 Garman RD, Doherty PJ, Raulet DH (1986) Diversity, rearrangement, and expression of murine T cell gamma genes. Cell 45:733–42 Gilfillan S, Ho EL, Cella M et al. (2002) NKG2D recruits two distinct adapters to trigger NK cell activation and costimulation. Nat Immunol 3:1150–5 Girardi M, Oppenheim DE, Steele CR et al. (2001) Regulation of cutaneous malignancy by gammadelta T cells. Science 294:605–9

A Ibusuki et al. NKG2D-Mediated Signal in Epidermal gd T Cells

Girardi M, Lewis J, Glusac E et al. (2002) Resident skin-specific gammadelta T cells provide local, nonredundant regulation of cutaneous inflammation. J Exp Med 195:855–67

Minagawa M, Ito A, Shimura H et al. (2001) Homogeneous epithelial gamma delta T cell repertoire of the skin is shaped through peripheral selection. J Dermatol Sci 25:150–5

Girardi M, Lewis JM, Filler RB et al. (2006) Environmentally responsive and reversible regulation of epidermal barrier function by gammadelta T cells. J Invest Dermatol 126:808–14

Nitahara A, Shimura H, Ito A et al. (2006) NKG2D ligation without T cell receptor engagement triggers both cytotoxicity and cytokine production in dendritic epidermal T cells. J Invest Dermatol 126: 1052–8

Graham DB, Cella M, Giurisato E et al. (2006) Vav1 controls DAP10-mediated natural cytotoxicity by regulating actin and microtubule dynamics. J Immunol 177:2349–55 Groh V, Rhinehart R, Randolph-Habecker J et al. (2001) Costimulation of CD8alphabeta T cells by NKG2D via engagement by MIC induced on virus-infected cells. Nat Immunol 2:255–60 Haas A, Weckbecker G, Welzenbach K (2008) Intracellular Phospho-Flow cytometry reveals novel insights into TCR proximal signaling events: a comparison with Western blot. Cytometry A 73:799–807 Hara H, Kishihara K, Matsuzaki G et al. (2000) Development of dendritic epidermal T cells with a skewed diversity of gamma delta TCRs in V delta 1-deficient mice. J Immunol 165:3695–705 Havran WL, Chien YH, Allison JP (1991) Recognition of self antigens by skinderived T cells with invariant gamma delta antigen receptors. Science 252:1430–2 Hayday AC (2009) Gammadelta T cells and the lymphoid stress-surveillance response. Immunity 31:184–96 Horng T, Bezbradica JS, Medzhitov R (2007) NKG2D signaling is coupled to the interleukin 15 receptor signaling pathway. Nat Immunol 8:1345–52 Ito A, Shimura H, Nitahara A et al. (2008) NK cells contribute to the skin graft rejection promoted by CD4 þ T cells activated through the indirect allorecognition pathway. Int Immunol 20:1343–9 Jameson J, Ugarte K, Chen N et al. (2002) A role for skin gammadelta T cells in wound repair. Science 296:747–9 Jameson JM, Cauvi G, Witherden DA et al. (2004) A keratinocyte-responsive gamma delta TCR is necessary for dendritic epidermal T cell activation by damaged keratinocytes and maintenance in the epidermis. J Immunol 172:3573–9 Jamieson AM, Diefenbach A, McMahon CW et al. (2002) The role of the NKG2D immunoreceptor in immune cell activation and natural killing. Immunity 17:19–29 Jiang K, Zhong B, Gilvary DL et al. (2002) Syk regulation of phosphoinositide 3-kinase-dependent NK cell function. J Immunol 168:3155–64 Jung H, Hsiung B, Pestal K et al. (2012) RAE-1 ligands for the NKG2D receptor are regulated by E2F transcription factors, which control cell cycle entry. J Exp Med 209:2409–22 Kaminski MJ, Cruz PD Jr, Bergstresser PR et al. (1993) Killing of skinderived tumor cells by mouse dendritic epidermal T-cells. Cancer Res 53:4014–9 Kloepper JE, Kawai K, Bertolini M et al. (2013) Loss of gammadelta T cells results in hair cycling defects. J Invest Dermatol 133:1666–9

Nixon-Fulton JL, Hackett J Jr., Bergstresser PR et al. (1988) Phenotypic heterogeneity and cytotoxic activity of Con A and IL-2-stimulated cultures of mouse Thy-1 þ epidermal cells. J Invest Dermatol 91: 62–8 Oppenheim DE, Roberts SJ, Clarke SL et al. (2005) Sustained localized expression of ligand for the activating NKG2D receptor impairs natural cytotoxicity in vivo and reduces tumor immunosurveillance. Nat Immunol 6:928–37 Rajasekaran K, Xiong V, Fong L et al. (2010) Functional dichotomy between NKG2D and CD28-mediated co-stimulation in human CD8 þ T cells. PLoS One 5:e12635 Raulet DH (2003) Roles of the NKG2D immunoreceptor and its ligands. Nat Rev Immunol 3:781–90 Raulet DH, Gasser S, Gowen BG et al. (2013) Regulation of ligands for the NKG2D activating receptor. Annu Rev Immunol 31:413–41 Roberts AI, Lee L, Schwarz E et al. (2001) NKG2D receptors induced by IL-15 costimulate CD28-negative effector CTL in the tissue microenvironment. J Immunol 167:5527–30 Rosen DB, Araki M, Hamerman JA et al. (2004) A structural basis for the association of DAP12 with mouse, but not human, NKG2D. J Immunol 173:2470–8 Schieven GL, Mittler RS, Nadler SG et al. (1994) ZAP-70 tyrosine kinase, CD45, and T cell receptor involvement in UV- and H2O2-induced T cell signal transduction. J Biol Chem 269:20718–26 Sharp LL, Jameson JM, Cauvi G et al. (2005) Dendritic epidermal T cells regulate skin homeostasis through local production of insulin-like growth factor 1. Nat Immunol 6:73–9 Shiohara T, Moriya N, Hayakawa J et al. (1996) Resistance to cutaneous graftvs.-host disease is not induced in T cell receptor delta gene-mutant mice. J Exp Med 183:1483–9 Strid J, Roberts SJ, Filler RB et al. (2008) Acute upregulation of an NKG2D ligand promotes rapid reorganization of a local immune compartment with pleiotropic effects on carcinogenesis. Nat Immunol 9:146–54 Strid J, Sobolev O, Zafirova B et al. (2011) The intraepithelial T cell response to NKG2D-ligands links lymphoid stress surveillance to atopy. Science 334:1293–7 Takada A, Yoshida S, Kajikawa M et al. (2008) Two novel NKG2D ligands of the mouse H60 family with differential expression patterns and binding affinities to NKG2D. J Immunol 180:1678–85

Komori HK, Witherden DA, Kelly R et al. (2012) Cutting edge: dendritic epidermal gammadelta T cell ligands are rapidly and locally expressed by keratinocytes following cutaneous wounding. J Immunol 188: 2972–6

Tigelaar RE, Lewis JM (1994) Factors involved in the localization and activation of murine gammadelta positive dendritic epidermal T cells. In: Lambert WC, Giannotti B, van Vloten WA (eds). Basic Mechanisms of Physiologic and Aberrant Lymphoproliferation in the Skin. Plenum Press: New York, NY, USA, pp 39–55

Lanier LL, Corliss BC, Wu J et al. (1998) Immunoreceptor DAP12 bearing a tyrosine-based activation motif is involved in activating NK cells. Nature 391:703–7

Uchida Y, Kawai K, Ibusuki A et al. (2011) Role for E-cadherin as an inhibitory receptor on epidermal gammadelta T cells. J Immunol 186:6945–54

Leandersson K, Jaensson E, Ivars F (2006) T cells developing in fetal thymus of T-cell receptor alpha-chain transgenic mice colonize gammadelta T-cellspecific epithelial niches but lack long-term reconstituting potential. Immunology 119:134–42

Upshaw JL, Arneson LN, Schoon RA et al. (2006) NKG2D-mediated signaling requires a DAP10-bound Grb2-Vav1 intermediate and phosphatidylinositol-3-kinase in human natural killer cells. Nat Immunol 7:524–32

Lopez-Larrea C, Suarez-Alvarez B, Lopez-Soto A et al. (2008) The NKG2D receptor: sensing stressed cells. Trends Mol Med 14:179–89

Whang MI, Guerra N, Raulet DH (2009) Costimulation of dendritic epidermal gammadelta T cells by a new NKG2D ligand expressed specifically in the skin. J Immunol 182:4557–64

Maasho K, Opoku-Anane J, Marusina AI et al. (2005) NKG2D is a costimulatory receptor for human naive CD8 þ T cells. J Immunol 174: 4480–4 Mallick-Wood CA, Lewis JM, Richie LI et al. (1998) Conservation of T cell receptor conformation in epidermal gammadelta cells with disrupted primary Vgamma gene usage. Science 279:1729–33

Witherden DA, Verdino P, Rieder SE et al. (2010) The junctional adhesion molecule JAML is a costimulatory receptor for epithelial gammadelta T cell activation. Science 329:1205–10 Witherden DA, Havran WL (2011) Molecular aspects of epithelial gammadelta T cell regulation. Trends Immunol 32:265–71

www.jidonline.org

403

A Ibusuki et al. NKG2D-Mediated Signal in Epidermal gd T Cells

Witherden DA, Watanabe M, Garijo O et al. (2012) The CD100 receptor interacts with its plexin B2 ligand to regulate epidermal gammadelta T cell function. Immunity 37:314–25

404

Yano S, Okochi H (2005) Long-term culture of adult murine epidermal keratinocytes. Br J Dermatol 153:1101–4

Wu J, Song Y, Bakker AB et al. (1999) An activating immunoreceptor complex formed by NKG2D and DAP10. Science 285:730–2

Yoshida S, Mohamed RH, Kajikawa M et al. (2012) Involvement of an NKG2D ligand H60c in epidermal dendritic T cell-mediated wound repair. J Immunol 188:3972–9

Wu J, Cherwinski H, Spies T et al. (2000) DAP10 and DAP12 form distinct, but functionally cooperative, receptor complexes in natural killer cells. J Exp Med 192:1059–68

Zompi S, Hamerman JA, Ogasawara K et al. (2003) NKG2D triggers cytotoxicity in mouse NK cells lacking DAP12 or Syk family kinases. Nat Immunol 4:565–72

Journal of Investigative Dermatology (2014), Volume 134