Dopaminergic signaling is involved in a variety of biological mechanisms. Dopamine (DA) acts as a neurotransmitter in the CNS where it participates in such processes as long term memory persistence (Rossato et al. 2009) and synaptic plasticity (Reynolds and Wickens 2000); however, aberrant dopaminergic signaling contributes to schizophrenia, movement disorders, and Parkinson’s disease. DA can also act in a paracrine fashion by regulating sodium excretion within the kidney (Yu et al. 2006). Dopaminergic signaling is mediated by five G protein-coupled receptors (GPCRs) that belong to the Class A family (Gainetdinov et al. 2004). DA receptors (DARs) are further grouped into two subclasses based upon common structural, pharmacological, and physiological properties (Neve et al. 2004). The D1-like receptor subclass is composed of the D1 and D5 DARs that couple to Gs/Golf leading to an increase in cAMP accumulation upon agonist activation. The D2-like receptor subclass is composed of the D2, D3, and D4 DARs that couple to Gi/Go resulting in attenuation of cAMP accumulation and modulation of K+ and Ca2+ channel conductances.
G protein-coupled receptors comprise the largest family of genes within the mammalian genome and mediate a variety of signaling pathways (Fredriksson and Schioth 2005). Most of these receptors exhibit some degree of phosphorylation which is proposed to mediate receptor desensitization—the process of diminishing receptor response under continual agonist stimulation (Krupnick and Benovic 1998; Ferguson 2001). Homologous forms of desensitization are believed to involve G protein-coupled receptor kinase (GRK, EC 184.108.40.206) phosphorylation resulting in uncoupling of receptor from G protein and enhanced arrestin binding with a concomitant abrogation of second messenger production followed by receptor internalization (Krupnick and Benovic 1998; Ferguson 2001). Heterologous desensitization of GPCRs appears to result from receptor phosphorylation by a kinase that is activated by a separate receptor system and is generally mediated by second messenger-activated kinases such as cAMP-dependent protein kinase (PKA, EC 220.127.116.11) and/or protein kinase C (PKC, EC 18.104.22.168) (Krupnick and Benovic 1998; Ferguson 2001).
The D1DAR exhibits both basal and robust agonist-induced phosphorylation mediated by multiple classes of protein kinases. The majority of the agonist-induced phosphorylation of the D1DAR is mediated by GRK isoforms GRK2, GRK3, and GRK5 (Tiberi et al. 1996; Rankin et al. 2006), while agonist-promoted PKA phosphorylation at T268 within the third intracellular loop (ICL3) of the D1DAR regulates the onset of receptor desensitization (Jiang and Sibley 1999) and receptor trafficking (Mason et al. 2002). In contrast, little is known about PKC-mediated phosphorylation of the D1DAR, although cells treated with PKC inhibitors display a reduction in basal and phorbol 12-myristate,13-acetate (PMA)-induced receptor phosphorylation (Gardner et al. 2001; Rex et al. 2008), indicating that PKC does phosphorylate D1DAR; however, the PKC sites within the receptor and the physiological significance of this modification remain unknown.
The PKC family of kinases is comprised of ten isozymes divided into three subgroups based on sequence homology and cofactor activation requirements. These isozymes display specific subcellular localization patterns and activities (Gould and Newton 2008; Reyland 2009; Newton 2010). The conventional subgroup is comprised of PKC α, βI, βII, and γ, and requires both Ca2+ and diacylglycerol (DAG) for kinase activation. The novel PKCs (δ, ε, η, and θ) require DAG, but not Ca2+ for activation, while the atypical PKCs (ζ and λ) require neither DAG nor Ca2+ for activation, but rather rely on protein-protein interactions and phosphoinositide-dependent kinase-1 (PDK-1) phosphorylation for activation. PKCμ/PKD1 and PKCν/PKD3 were originally categorized as the fourth (PDK) subgroup within the PKC family; however, these kinases have been re-categorized as a novel subgroup within the CamK family (Manning et al. 2002b).
Our laboratory has previously reported that PKCs phosphorylate the D2DAR and that this phosphorylation results in functional desensitization and receptor internalization (Namkung and Sibley 2004). We have also shown that ethanol-dependent regulation of the D1DAR is mediated through modification of PKC activities, resulting in decreased D1DAR phosphorylation (Rex et al. 2008). We now report that D1DAR is phosphorylated by multiple PKC isozymes in the absence of agonist stimulation. This constitutive PKC phosphorylation of the D1DAR dampens DA activation of the receptor thus attenuating D1DAR-mediated signaling pathways.
- Top of page
- Materials and methods
- Supporting Information
The idea that phosphorylation of GPCRs represents a universal mechanism whereby agonist-bound, phosphorylated receptor becomes a substrate for arrestin binding, inducing receptor desensitization and internalization is no longer broadly applicable given the recent accumulation of evidence for phosphorylation-independent arrestin recruitment, desensitization, and internalization (Chen et al. 2004; Jala et al. 2005; Richardson et al. 2003; Namkung et al. 2009b,a). Additionally, evidence is mounting that GPCRs can be differentially regulated not only by a particular kinase but also by the location within the receptor and the time-frame in which the phosphorylation occurs (Busillo et al. 2010).
The D1DAR is the most widely expressed dopamine receptor in the CNS where it functions in a variety of neuronal processes, yet it is also expressed in peripheral tissue including the parathyroid gland (Sunahara et al. 1990), seminal vesicles (Hyun et al. 2002), heart, and kidney (Ozono et al. 1997). Given the diversity of tissue types in which the receptor is expressed, tissue-specific D1DAR functional regulation mechanisms must exist to refine tissue-specific function. Phosphorylation represents the most prevalent means of post-translationally modifying a protein to effect functional change (Manning et al. 2002a), so an attractive model for tissue-specific receptor regulation can be achieved through tissue-specific kinase expression, activation and/or subcellular localization (Tobin et al. 2008). To this end, we have begun identifying which kinases phosphorylate the D1DAR, where this phosphorylation occurs within the receptor, and the functional consequence(s) of this modification.
The major finding of our current study is that the D1DAR is phosphorylated by PKC either constitutively or heterologously, and that this negatively regulates receptor-G protein coupling and downstream signaling. Through the use of PKC inhibitors and mutagenesis (see below) we found that the vast majority of D1DAR phosphorylation in the basal state is mediated by PKC. The small amount of remaining basal phosphorylation is because of GRK(s) and an unidentified protein kinase (Rankin and Sibley, unpublished observations). The exact mechanism of basal phosphorylation of the D1DAR by PKC is unclear, but may be because of constitutively active PKC isozymes within the cell or through the scaffolding activities of PKC binding proteins. We have recently reported that RanBP9 and RanBP10 can bind to the D1DAR and to PKCδ or PKCγ, providing a mechanism by which these PKC isozymes can be directed to the receptor to increase D1DAR basal phosphorylation (Rex et al. 2010). We also found that PKC phosphorylation of the D1DAR could be increased through either direct PKC activation with phorbol esters or through activation of a Gq-linked muscarinic receptor leading to PKC activation. The latter mechanism may be involved in heterologous forms of GPCR desensitization. PKC activation appears not to be involved in homologous forms of agonist-induced receptor phosphorylation as we previously found that PKC inhibitors do not block this response (Gardner et al. 2001). Agonist-stimulated receptor phosphorylation appears to be predominantly mediated by GRKs (Rankin and Sibley, in preparation).
It appears as if multiple PKC isozymes are capable of phosphorylating the D1DAR. In our over-expression studies, we found that PKCs α, βI, γ, δ, and ε all increased D1DAR phosphorylation whereas PKCs η, λ, ζ, μ, and ν were ineffective. We previously found that HEK293T cells express PKCs α, βI, δ, ε, ζ, μ, and ν (Rex et al. 2008). From these comparative analyses, it would appear as if PKCs α, βI, δ, and ε may be involved in D1DAR phosphorylation, at least within the HEK293T cells. A distinct possibility, however, is that different PKC isozymes may play different roles in basal versus cell-stimulated phosphorylation and/or that distinct PKCs phosphorylate different sites on the receptor. All of these possibilities will require further investigation.
In order to specifically delineate the role of receptor phosphorylation per se, we found it necessary to identify all of the potential PKC phosphorylation sites on the receptor. This was a complicated analysis because of the fact that the D1DAR is a robustly phosphorylated receptor with 32 potential serine/threonine phospho-acceptor sites located within the intracellular domains of the receptor. We determined that the carboxyl tail and the ICL3 comprise the two domains that are phosphorylated in the D1DAR and these alone contain 28 potential sites of serine/threonine phosphorylation. A further complication in this analysis is that phosphorylation of the D1DAR appears to occur in a hierarchical fashion in that sites in the carboxyl terminus must be phosphorylated prior to phosphorylation of sites in the ICL3. We first obtained evidence for this in examining DA-stimulated (i.e., GRK-mediated) D1DAR phosphorylation (Kim et al., 2004). While this mechanism also appears to apply to PKC-mediated receptor phosphorylation, we are currently exploring this phenomenon more fully using GRK-mediated phosphorylation as a read-out (Rankin and Sibley, in preparation).
Given the above, we undertook a combination of mutational analyses, which involved ‘loss of function’ mutations, in which serine/threonine residues were mutated to alanine/valine residues and a loss of phosphorylation was examined, as well as ‘gain of function’ mutations in which all (or many) of the serine/threonine residues were first mutated to alanine/valine residues to create a phosphorylation-null or phosphorylation-reduced receptor and then specific alanine/valine residues were mutated back to serine/threonine residues and a gain of phosphorylation was detected. Examination of over 100 mutated receptor constructs (for PKC alone) resulted in the identification of five serine residues—four in the carboxyl terminus and one in the ICL3 that appeared to account for all of the receptor phosphorylation by PKC in both the basal and stimulated states. When all five of these serine residues were simultaneously mutated, basal phosphorylation of the receptor was dramatically reduced and there was no effect of PKC activators or inhibitors on receptor phosphorylation. Interestingly, the four serine residues in the carboxyl terminus are in relative close proximity to each other (Fig. 5) suggesting that this region may comprise a potential binding site for PKC(s) within the carboxyl terminus.
Importantly, PKC- and GRK-mediated phosphorylation of the D1DAR appears to occur independently from one another. This is suggested by the fact that PKC inhibitor pre-treatment did not affect DA-induced receptor phosphorylation (Gardner et al. 2001) and the PKC-null receptor construct desensitized normally in response to agonist (this study). The role of PKC phosphorylation in regulating D1DAR function proved to be rather complicated as we found that both intracellular activators and inhibitors of PKCs led to enhanced DA-stimulated cAMP accumulation. However, it has been previously shown that PKC phosphorylation of specific AC isoforms will increase hormone-stimulated cAMP accumulation (Sibley et al. 1986; Yoshimasa et al. 1987; Sunahara and Taussig 2002) whereas PKC phosphorylation of GPCRs generally leads to desensitization or reduced receptor activity (Sibley et al. 1984; Bouvier 1990). In order to differentiate between these two possibilities we created a PKC-null mutant as described above. In comparison to the WT D1DAR, the PKC-null construct was refractory to the cAMP enhancing effects of PKC inhibitors on cAMP accumulation, whereas the mutant construct responded fully (enhancement of cAMP accumulation) to PKC activators when compared to the WT receptor. These results would seem to confirm that the enhanced cAMP response subsequent to PKC activation is non-receptor mediated whereas the enhanced response to PKC inhibition is mediated by a reduction in the basal or constitutive phosphorylation of the D1DAR. These results were confirmed using [35S]GTPγS binding assays to directly measure D1DAR activation of Gs. In these experiments, the PKC-null receptor construct was significantly more effective in stimulating [35S]GTPγS binding than the WT receptor, indicating that PKC phosphorylation diminishes the ability of the receptor to activate Gs. Importantly, in this model, one must invoke the hypothesis that under basal conditions, PKC phosphorylation of the receptor, rather than non-receptor components, exhibits the predominant functional effect otherwise PKC inhibitor treatment would be expected to decrease rather than increase DA-stimulated cAMP accumulation. In contrast, when PKC is maximally activated such as when the cells are stimulated with phorbol esters, non-receptor components provide the predominant effect of enhancing cAMP accumulation.
In summary, we have shown that the D1DAR is phosphorylated either constitutively or heterologously by PKC and that this negatively regulates receptor-G protein coupling and downstream signaling. We have also previously shown that the D1DAR can be constitutively phosphorylated and desensitized by GRK4 (Rankin et al. 2006). Taken together, these results demonstrate the importance of constitutive mechanisms of GPCR regulation in general and of the D1DAR in particular, and provide a means of modulating the signaling efficacy of the D1DAR in a context-specific manner.
- Top of page
- Materials and methods
- Supporting Information
Figure S1. Quantitation of PKC phosphorylation of the D1DAR. The mean ± SEM of the normalized data collected from seven individual in situ phosphorylation assays performed as described in Fig. 1(c) are reported in the histogram (paired Student’s t-test, where *p < 0.05, **p < 0.01, and ***p < 0.001).
Figure S2. Identification of individual PKC isozymes that can mediate D1DAR phosphorylation. In situ phosphorylation assays performed on HEK293T cells expressing the D1DAR and empty vector (V) or the indicated PKC isozyme as described in Materials and Methods. (+) lanes correspond to cells treated for 30 min with the indicated PKC activator PDBu and/or the PKC inhibitors Gö6976 and Gö6983. Activation of the atypical PKCλ and PKCζ was accomplished by cotransfection of the upstream activator kinase PDK-1. Here, cells were exposed to [32P]orthophosphoric acid for 45 min prior to cell lysis. The amount of D1DAR resolved for each experiment is as follows: 0.5 pmole/lane for PKCβI and PKCδ, 1.4 pmole/lane for PKC and PKCμ, and 1.5 pmole/lane for PKCλ and PKCζ, and 1 pmole/lane for PKCγ. High molecular weight bands present in the PKC and PKCγ autoradiographs represent co-immunoprecipitated FLAG-tagged PKC expression constructs. While PKCμ is not FLAG-tagged, it efficiently co-immunoprecitated with the D1DAR. These experiments were performed at least three times with similar results.
Figure S3. Removal of PKC phosphorylation sites within the D1DAR has no effect on receptor expression. Radioligand binding assay performed on cells expressing the WT or PKC-null D1DAR. In this representative experiment, binding parameters were Bmax = 7.6 pmole/mg protein, KD = 0.20 for WT receptor and Bmax = 7.3 pmole/mg protein, KD = 0.17 for PKC-null D1DAR. This experiment was performed four times with similar results.
Figure S4. Removal of PKC phosphorylation sites within the D1DAR has no effect on DA-induced receptor desensitization. cAMP accumulation assays performed on cells expressing (a) WT or (b) PKC-null D1DAR. Cells were pre-treated for 1 h with 10 μM DA, washed, then challenged with various concentrations of DA followed by cell lysis and cAMP quantitation as described in Materials and Methods. This experiment was performed three times with similar results; the extent of DA-induced desensitization was not statistically different between the two groups (mean ± SEM %desensitization for WT = 42.5 ± 4.9 and for PKC-null = 37.0 ± 2.8), p = 0.26, paired Student’s t-test.
As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.