In the present study, we provide evidence that the non-receptor tyrosine kinase Pyk2 interacts with and regulates mGluR1a signaling and desensitization. Pyk2 associates with the second intracellular loop domain and the carboxyl-terminal tail domains of mGluR1a, and similar to what has been described for GRK2, optineurin and CAIN, Pyk2 functions to disrupt Gαq/11 interactions with the receptor resulting in the attenuation of IP formation [9, 16, 17]. However, Pyk2 also facilitates mGluR1a-mediated activation of the ERK1/2 pathway, an effect that also requires the activity of Src, calmodulin and PKC. Thus, we describe a new mechanism by which Pyk2 attenuates mGluR1a by disrupting mGluR1a/Gαq/11 interactions, while simultaneously coupling the receptor to the activation of the mitogen-activated protein kinase (MAPK) signaling pathway.
Pyk2 is a non-receptor tyrosine kinase that is known to be activated by stimuli that either increase intracellular Ca2+ concentrations or activate PKC in response to stress signals, such as UV light, hyperosmotic shock and tumor necrosis factor-α . Pyk2 has also been shown to be involved in G protein-coupled-receptor signaling. For example, Felsch et al.  reported that Pyk2 phosphorylates the potassium channel Kv1.2 in response to the activation of m1 muscarinic acetylcholine receptor. In vascular smooth muscle cells, angiotensin II stimulates the association of Janus kinase 2 with the angiotensin II type1 receptor via a mechanism that involves increases in intracellular Ca2+ concentrations and the activation of both PKCδ and Pyk2 . Similarly, Heidinger et al.  showed that Pyk2 is involved in the mGluR1-mediated phopshorylation of NR2A/B subunits of the NMDA receptor in cortical neurons. Moreover, in HEK293 cells both the α1B- or α2A-adrenergic receptors stimulate ERK1/2 phosphorylation via a Ca2+-, calmodulin-, Src- and Pyk2-dependent pathway . In the present study, we provide evidence that Pyk2 interacts with mGluR1a and plays a role in of the mGluR1a-dependent activation of ERK1/2 phosphorylation via a mechanism that also requires calmodulin, Src, and PKC activity. Moreover, Pyk2 is associated with mGluR1a in the absence of agonist treatment and the mGluR1a/Pyk2 protein complex dissociates in response to quisqualate treatment for 20 min. The physiological consequence of this dissociation is unclear, but it is possible that the loss of Pyk2 from the receptor may occur as the result GRK2 binding to the receptor, which like mGluR5 might be required for agonist-stimulated mGluR1a endocytosis . Consistent with this hypothesis and the observation that Pyk2 dissociates from the receptor in response to agonist activation, we do not observe Pyk2 internalization with mGluR1a.
Pyk2 has previously been demonstrated to be autophosphorylated on Tyr402 in response to the activation of the G protein-coupled lysophosphatidic acid receptor resulting in the creation of a docking site for the Src SH2 domain . The recruitment of Src to Pyk2 subsequently leads to Pyk2 phosphorylation at both Tyr579 and Tyr580 and results in enhanced Pyk2 kinase activity [28–30]. The overexpression of Pyk2 is also reported to result in increased basal ERK1/2 phosphorylation [28, 29, 31]. However, we find that in HEK293 cells basal ERK1/2 phosphorylation is increased only when mGluR1a is overexpressed with Pyk2. Thus, despite the fact that Pyk2 reduces basal mGluR1a G protein coupling, Pyk2 overexpression results in an increase in basal ERK1/2 phosphorylation. In contrast, mGluR1a-mediated ERK1/2 phosphorylation is reduced following the expression of dominant-negative Pyk2-Y402F and catalytically inactive Pyk2-K457A mutants. Thus, we conclude that Pyk2 is essential for ERK1/2 phosphorylation in response to mGluR1a activation.
Both mGluR1 and mGluR5 are expressed in cortical tissue and primary cortical neurons can be positively stained for the expression of both mGluR1a and mGluR5 proteins . However, in primary mouse cortical neurons ERK1/2 phosphorylation is selectively activated by mGluR1, as ERK1/2 phosphorylation in response to DHPG is not inhibited by a mGluR5-specific antagonist. This is similar to what was observed by Heidinger and colleagues , who demonstrated that the phosphorylation of NMDA receptors by Pyk2 is selectively mediated by the activation of mGluR1 and not mGluR5 in cortical neurons. It is likely that mGluR5 will also activate ERK1/2 phosphorylation in other neuronal cell types for several reasons. First, Pyk2 can be co-immunoprecipitated from rat brain with mGluR5. Second, mGluR5-mediated nociception in the spinal cord involves the activation of ERK1/2 and we have recently demonstrated that mGluR5-dependent ERK1/2 phosphorylation is increased in mutant huntingtin protein knockin mice [32, 33]. Finally, MacDonald and coworkers have demonstrated that mGluR5-dependent activation of Pyk2 stimulates NMDA and AMPA receptor currents in hippocampal neurons .
It has previously been reported that Group I mGluRs increase the extent of Pyk2 phosphorylation in mouse neuronal cortical cultures via a mechanism that is both calmodulin- and Src-dependent, but that is independent of PKC activity . We find that Pyk2-mediated activation of ERK1/2 phosphorylation is both calmodulin- and Src-dependent, but it also involves a mechanism that requires PKC activation. Calmodulin binds to both mGluR1 and mGluR5 in a Ca2+-dependent manner and can be regulated by PKC-mediated phosphorylation of the receptor within the calmodulin binding domain [35–37]. The identification of Pyk2 and CaM as Group I mGluR interacting proteins suggests the possibility that Pyk2, Src and calmodulin may exist as a preformed complex that is scaffolded on the intracellular face of mGluR1. Thus, in response to PKC-mediated phosphorylation of mGluR1a, the complex may be released from the receptor to activate ERK1/2 phosphorylation.
We also find that siRNA treatment of cortical neurons results in a significant reduction in Pyk2 protein expression, but unexpectedly this also results in increased basal ERK1/2 phosphorylation. The increase in ERK1/2 phosphorylation is associated with increased phosphorylation of the remaining fraction of Pyk2 protein expressed in the neuronal cultures. The depletion of Pyk2 expression in COS7 cells also leads to increased basal Pyk2 phosphorylation, but mGluR1a expression is required for this Pyk2-mediated increase in ERK/1/2 phosphorylation. The mechanism by which a loss of Pyk2 expression leads to alterations in the phosphorylation of the residual pool of Pyk2 and increased ERK1/2 phosphorylation is unclear. However, it appears to require mGluR1 expression suggesting that constitutive mGluR1 activity underlies the phenomenon and that Pyk2 plays a role in regulating basal mGluR1 activity.
In summary, we show here that Pyk2 interacts with Group I mGluR1a via the intracellular loop2 and the carboxyl-terminal tail domains of mGluR1a and functions to uncouple the receptor from Gαq/11 protein, while facilitating ERK1/2 phosphorylation. The direct association of Pyk2 with mGluR1a appears to coordinate the formation of a protein complex that includes Src bound to Pyk2 and potentially calmodulin bound to the carboxyl-terminal tail of the receptor. The formation of this complex may be important for the spatial temporal regulation of Pyk2 activity at post-synaptic densities that is expected to be required for the efficient regulation of NMDA and AMPA receptor function as well as the regulation of other ERK1/2-dependent activities in synaptic transmission. The observation that several signaling proteins, such as Pyk2, optineurin CAIN and GRK2 regulate mGluR signaling via their interaction with the interface provided by the second intracellular loop suggests that peptides which mimic this interface may be useful for the modulation of Group I mGluR signaling.