| Abstract: | ![]()
The activating natural killer (NK) cell receptor Ly49H recognizes the mouse cytomegalovirus (MCMV) m157 glycoprotein expressed on the surface of infected cells and is required for protection against MCMV. Although Ly49H has previously been shown to signal via DAP12, we now show that Ly49H must also associate with and signal via DAP10 for optimal function. In the absence of DAP12, DAP10 enables Ly49H-mediated killing of m157-bearing target cells, proliferation in response to MCMV infection, and partial protection against MCMV. DAP10-deficient Ly49H+ NK cells, expressing only Ly49H–DAP12 receptor complexes, are partially impaired in their ability to proliferate during MCMV infection, display diminished ERK1/2 activation, produce less IFN-γ upon Ly49H engagement, and demonstrate reduced control of MCMV infection. Deletion of both DAP10 and DAP12 completely abrogates Ly49H surface expression and control of MCMV infection. Thus, optimal NK cell–mediated immunity to MCMV depends on Ly49H signaling through both DAP10 and DAP12.NK cells mediate immunity against tumors and pathogens by producing effector cytokines, such as IFN-γ, and by secreting lytic granules that kill target cells. NK cell activation and function are determined by a balance of signals transmitted by inhibitory and activating NK cell receptors (NKRs) (1). Many of the inhibitory NKRs recognize self-ligands such as MHC class Ia, which are recognized by inhibitory Ly49 receptors in mice and inhibitory killer immunoglobulin-like receptors (KIRs) in humans, and MHC class Ib molecules, which are recognized by CD94/NKG2A in both species. Activating NKRs can recognize host-encoded molecules that are induced by transformation or infection of the host cells (e.g., NKG2D ligands), and some activating NKRs recognize non–self-ligands, for example, the mouse CMV (MCMV) m157 glycoprotein, which is recognized by Ly49H (2–6). Activating NKRs typically have short intracellular domains that lack known signaling motifs and instead associate with signaling subunits, including the immunoreceptor tyrosine-based activation motif (ITAM)–bearing CD3-ζ, FcϵRIγ, and DAP12 (also referred to as KARAP) proteins and the YINM motif-containing DAP10 adaptor (7, 8). In the absence of an appropriate signaling subunit, many of the activating NKRs (e.g., NKG2D, CD16, KIR3DS1, and the CD94/NKG2C heterodimer) are not stably expressed on the cell surface (8–11). In humans, some of the activating NKRs, such as CD16, can associate interchangeably with both the ITAM-bearing FcϵRIγ and CD3-ζ subunits for stable expression and signaling (12).DAP12 associates with multiple activating NKRs, including the human and mouse CD94/NKG2C heterodimers, the mouse Ly49H and Ly49D receptors, human NKp44, and the activating human KIR (1). Conversely, NKG2D is the only activating NKR known to associate with DAP10 and initiate NK cell immune responses in vivo (13). Based on the structural similarity of DAP10 and DAP12, particularly within their transmembrane domains, which mediate receptor–adaptor association, it is reasonable to expect that receptors known to associate with DAP12 might also complex with DAP10. However, the alternatively spliced isoform of NKG2D, which is designated NKG2D-S, expressed by activated mouse but not human NK cells, is the only receptor known to pair with both DAP10 and DAP12 in vivo (14–16). Signaling downstream of these adaptors differs as ITAM-containing adaptors recruit Syk or ZAP-70, whereas DAP10 recruits the p85 subunit of PI3 kinase and Grb2 (8, 17–21). NK cells activated via ITAM-containing subunits proliferate, produce cytokines, and are cytotoxic, whereas cells activated through DAP10 are triggered to kill but do not efficiently induce the production of cytokines (14, 15, 19, 22, 23). Importantly, in human NK cells, NKG2D signaling via DAP10 augments IFN-γ and GM-CSF production induced by an activating DAP12-associated KIR; thus, signaling through both adaptors induces a more robust immune response (22).Lack of NK cells renders both humans and mice susceptible to certain infections, particularly the herpesviruses, including human CMVs and MCMVs (24–26). Therefore, experimental infection of mice with MCMV provides an instructive model for studying NK cell responses to viral infection. The Ly49H receptor, which is expressed on a subset of NK cells in C57BL/6 (B6) mice, binds to the m157 glycoprotein encoded by MCMV and is the dominant receptor responsible for NK cell–mediated resistance to MCMV in B6 mice (2, 6, 27–29). Ly49H+ NK cells control MCMV replication by both direct cytotoxic mechanisms and by secretion of IFN-γ (30, 31). Genetic ablation of Ly49H, treatment with Ly49H blocking antibody, or infection with a mutant MCMV lacking m157 (Δm157) renders normally resistant B6 mice susceptible to MCMV (32–34). Dokun et al. (35) have reported that early after infection with MCMV, both Ly49H+ and Ly49H−, NK cells become activated and proliferate, presumably as a result of the proinflammatory cytokine environment. However, by 3 d after infection Ly49H+ NK cells preferentially proliferate, and by 6 d the percentage of Ly49H+ NK cells increases from ∼50 to 80–90% of the total NK cell population (35). In B6 mice in which the ITAM of DAP12 had been altered to prevent association with Syk or ZAP-70 (designated as DAP12ki or KARAPki mice), this preferential expansion of Ly49H+ NK cells was lost (36, 37). Furthermore, these mice exhibited increased viral burden and histopathology (38).Recently, Coudert et al. (39) suggested an association between Ly49H and DAP10 based on transient transfection experiments and coimmunoprecipitation from IL-2–cultured NK cells, and prior studies have suggested an association of mouse SIRP-β1 with DAP10 or DAP12 in a rat mast cell leukemia transfectant model (40). However, it remains unknown whether Ly49H–DAP10 complexes are functionally active and whether they contribute to host protection in vivo. In this paper, we demonstrate that Ly49H associates with DAP10 in addition to DAP12 and that Ly49H–DAP10 complexes are functional and necessary for optimal control of MCMV infection. Deletion of either signaling subunit reduces surface expression of Ly49H, and loss of both DAP10 and DAP12 completely ablates Ly49H surface expression. In the absence of DAP12, Ly49H can signal via DAP10 to induce NK cell responses that partially control MCMV infection. In contrast, the deletion of both DAP10 and DAP12 ablates control of MCMV to the level observed in NK cell–depleted mice. Although either signaling subunit is sufficient for NK cell–mediated killing of m157-bearing targets, we find that DAP10 is necessary for optimal Ly49H-mediated activation of ERK1/2, NK cell proliferation, IFN-γ production, and control of MCMV infection. Thus, Ly49H must associate with both DAP10 and DAP12 to induce optimal NK cell–mediated immunity to MCMV. |
