The detection of fungal particles by Dectin-1 and the subsequent activation of Syk triggers various intracellular signaling pathways (Kerrigan and Brown, 2011; Mócsai et al., 2010; Osorio and Reis e Sousa, 2011
). We observed that the recognition of zymosan elicits a Src- and Syk-dependent phosphorylation of PKCδ at Tyr311, indicating that PKCδ activation occurs downstream of Syk. Moreover, Prkcd−/−
BMDCs showed impaired Card9-Bcl10 complex assembly and NF-κB control in spite of normal Syk activation, demonstrating that PKCδ acts upstream of Card9. Together, these findings indicate that PKCδ operates as a missing link between Syk signaling and Card9 complex formation for the activation of innate immunity. As shown by the fact that only Prkcd−/−
BMDCs but not cells lacking PKCα, PKCβ, or PKCθ were defective in Dectin-1-induced cytokine production, PKCδ is the specific PKC isoform for signaling in the Dectin-1 pathway.
Treatment of the cells with a small molecule PKC kinase inhibitor blocked zymosan- or curdlan-induced cytokine production. This indicates that the enzymatic serine-threonine kinase activity and not merely a scaffold function of PKCδ is responsible for its activity in Dectin-1 signaling. In vitro kinase assays revealed that Card9 is a direct PKCδ substrate and that PKCδ phosphorylates Card9 at position Thr231. PKCδ is likely to phosphorylate other Card9 residues because the Card9(T231A) mutant was still substantially phosphorylated by PKCδ in vitro. Another PKCδ target site could be Thr95 because global proteomic approaches have demonstrated Card9 phosphorylation at Thr95 in vivo (Choudhary et al., 2009
). Because we were unable to express the Card9(T95A) mutant in mammalian cells, we were not able to determine the physiological function of Thr95 phosphorylation but speculate that it might be important for protein stability. Nevertheless, our genetic experiments with Card9-deficient cells that were reconstituted with the phosphorylation-defective Card9(T231A) mutant revealed that the Thr231 residue phosphorylated by PKCδ is absolutely required for downstream signaling and cytokine production. In addition, we observed that PKCδ signaling is essential for Card9-Bcl10 complex assembly and for Card9-dependent TAK1 activation. Thus, we postulate a molecular model in which Syk-induced PKCδ activity mediates direct Card9 phosphorylation, finally resulting in Card9-Bcl10 complex assembly for TAK1 activation. TAK1 then most probably mediates Dectin-1-induced IKK activation in a fashion similar to its mode of triggering NF-κB activity in response to stimuli from other immune receptors, such as TLRs (Vallabhapurapu and Karin, 2009
). This would be consistent with the fact that pharmacological TAK1 blockage inhibited zymosan-induced IKK activation.
Intriguingly, although the Card9 signaling pathway was severely defective in Prkcd−/−
BMDCs, the phagocytosis of zymosan and the production of ROS were largely intact and the activation of Erk MAPK signaling was only slightly reduced. These findings indicate that PKCδ controls only specific subsets of the Dectin-1 responses. Because Dectin-1 ligation can activate the serine-threonine kinase Raf-1 through alternative mechanisms (Gringhuis et al., 2009
), it is possible that Raf-1 might be responsible for Dectin-1-triggered and PKCδ-independent Erk activation, consistent with the role of Raf-1 in activating MAPK pathways in numerous settings (Galabova-Kovacs et al., 2006
). Interestingly, Prkcd−/−
mice were similar to Card9−/−
mice (Gross et al., 2006
) in being highly susceptible to fungal infections. The specific PKCδ-Card9 effector response downstream of CLRs is therefore absolutely critical for host defense. The aforementioned pivotal functions of Dectin-1 and Card9 in human antifungal immunity were recently demonstrated in genetic studies (Ferwerda et al., 2009; Glocker et al., 2009
). In the future it will thus be important to investigate whether genetic defects in PKCδ might also cause human immunodeficiency syndromes.
Many of our experiments aimed at identifying the principle mechanisms of CLR signaling utilized zymosan stimulation or selective Dectin-1 triggering. Yet in the absence of PKCδ, activation of the NF-κB pathway as well as cytokine production were also severely impaired in response to whole C. albicans
cells. Recent studies have indicated that C. albicans
cells, and particularly C. albicans
hyphae, are potent activators of Dectin-2 signaling (Saijo et al., 2010
). Presumably the hyphae in addition also activate Mincle (Wells et al., 2008
). Thus, we propose that PKCδ also couples signals from those CLRs to the Card9-controlled NF-κB pathway. This hypothesis is in line with our observation that Prkcd−/−
BMDCs were defective in cytokine responses to selective agonists for Dectin-2 and Mincle, which formally establishes PKCδ as a general integrator of CLR function.
We believe that our findings have implications beyond antifungal immunity. Mincle and Dectin-1 detect ligands on mycobacteria (Ishikawa et al., 2009; Rothfuchs et al., 2007
) and Card9−/−
mice are impaired in the Mincle-induced inflammatory response to the mycobacterial cord factor (Schoenen et al., 2010; Werninghaus et al., 2009
). These animals also succumb rapidly to aerosol lung infection with Mycobacterium tuberculosis
(Dorhoi et al., 2010
). In addition, Schistosoma mansoni
activates Dectin-2 and Card9 (Ritter et al., 2010
) and viruses such as the Dengue virus
induce inflammatory responses through the ITAM-coupled CLR Clec5a (Chen et al., 2008
). Moreover, Dectin-1 also binds to unknown endogenous structures on T cells (Ariizumi et al., 2000
), the Syk-coupled CLR Clec9a recognizes ligands that are exposed upon cellular necrosis (Sancho et al., 2009
), and Mincle triggers Card9-dependent inflammatory responses upon binding to SAP130 from necrotic cells under sterile conditions (Yamasaki et al., 2008
). Together with our results, these data imply that PKCδ might also mediate innate responses to bacteria, parasites, or viruses or could be involved in immune responses to conditions of noninfectious cell injury. These hypotheses need to be investigated.
Finally, CLR-triggered Card9 signaling not only regulates the immediate innate antimicrobial responses but also couples innate recognition to the activation of adaptive immunity (Kerrigan and Brown, 2011; Osorio and Reis e Sousa, 2011
). Triggering of the Card9 pathway by Dectin-1, Dectin-2, or Mincle ligands on antigen-presenting cells controls the synthesis of distinct cytokines. This cytokine milieu instructs the development of antigen-specific Th17 cell responses and additionally induces Th1 cell immunity (LeibundGut-Landmann et al., 2007; Robinson et al., 2009; Saijo et al., 2010; Werninghaus et al., 2009
). Th17 cell responses are often associated with autoimmunity and polymorphisms in Card9 are recurrently detected in human inflammatory diseases (Pointon et al., 2010; Zhernakova et al., 2008
). It will therefore be important to test whether PKCδ signaling in innate cells couples to Th17 cell responses and whether aberrant activity in CLR-induced PKCδ-Card9 signaling contributes to human inflammatory disease. In this context, PKCδ is a drugable target and PKC inhibitors are in clinical trials. Thus, our findings raise the possibility of therapeutically manipulating CLR-mediated Card9 signaling by targeting PKCδ function.