Functional modulation of GABA_B receptors by protein kinases and receptor trafficking

Abstract

GABA_B receptors (GABA_BR) are heterodimeric G-protein coupled receptors (GPCRs) that mediate slow and prolonged inhibitory signals in the central nervous system. The signaling of GPCRs is under stringent control and is subject to regulation by multiple posttranslational mechanisms. The β-adrenergic receptor is a prototypic GPCR. Like most GPCRs prolonged exposure of this receptor to agonist induces phosphorylation of multiple intracellular residues that is largely dependent upon the activity of G protein-coupled receptor kinases (GRKs). Phosphorylation terminates receptor effector coupling and promotes both interaction with β-arrestins and removal from the plasma membrane via clathrin-dependent endocytosis. Emerging evidence for GABA_BRs suggests that these GPCRs do not conform to this mode of regulation. Studies using both native and recombinant receptor preparations have demonstrated that GABA_BRs do not undergo agonist-induced internalization and are not GRK substrates. Moreover whilst GABA_BRs undergo clathrin-dependent constitutive endocytosis, it is generally accepted that their rates of internalization are not modified by prolonged agonist exposure. Biochemical studies have revealed that GABA_BRs are phosphorylated on multiple residues within the cytoplasmic domains of both the R1 and R2 subunits by cAMP-dependent protein kinase and 5′AMP-dependent protein kinase (AMPK). Here we discuss the role that this phosphorylation plays in determining GABA_BR effector coupling and their trafficking within the endocytic pathway and go on to evaluate the significance of GABA_BR phosphorylation in controlling neuronal excitability under normal and pathological conditions.

1. Introduction

GABA_BRs are GPCRs that mediate the slow and prolonged inhibitory action of GABA, the principal inhibitory neurotransmitter in the brain, via activation of Gαi- and Gαo-type heteromeric type G proteins. GABA_BRs mediate their inhibitory action via multiple effectors; post-synaptically they activate inwardly rectifying K⁺ channels (GIRKs), leading to hyperpolarization, whilst pre-synaptically they inhibit voltage-gated N/P/Q type Ca²⁺ channels, leading to reduced neurotransmitter release (Couve et al., 2000, 2004; Bettler et al., 2004, 2006). Through their activation of Gαi they also inhibit the activity of adenylate cyclase to reduce PKA signaling pathways.

Given the central roles of GABA_BRs in mediating neuronal inhibition there is considerable interest in understanding the cellular mechanisms neurons use to modulate their activity. This is in determined in part by the efficacy of the effector coupling of GABA_BRs and their residence time on the plasma membrane. Classically the signaling of monomeric GPCRs, as exemplified by the prototypic β-adrenergic receptor, is tightly controlled by agonist desensitization. Prolonged exposure to agonist leads to decreased effector coupling due to enhanced phosphorylation by G protein-coupled receptor kinases (GRKs). This decreased effector coupling ultimately leads to endocytosis followed by recycling or degradation (Ferguson, 2001; von Zastrow, 2003; Gainetdinov et al., 2003; Marchese et al., 2008). In this review we discuss the molecular mechanisms that neurons use to regulate the functional activity of GABA_BRs and the roles that these processes play as determinants of the efficacy of neuronal inhibition.

2. Phosphorylation of GABA_BR and its functional modulation

It is well documented that the activation and inactivation of GPCRs is modulated by the phosphorylation within the carboxyl (C)-terminus and the third intracellular loop by various protein kinases (Ferguson 2001; von Zastrow 2003; Marchese et al., 2008). In general, phosphorylation of GPCRs by intracellular kinases induces the desensitization of the receptor followed by interaction with cytosolic cofactor proteins called arrestins, which uncouple the receptor from G proteins. GPCRs are internalized from the plasma membrane principally via clathrin-dependent endocytosis. They are then recycled back to the plasma membrane for re-insertion or targeted for lysozomal degradation.

To evaluate the significance of this mode of regulation for GABA_BRs measurements of phosphorylation have been performed in both expression systems and in neurons using labeling with ³²P-orthophosphate followed by immunoprecipitation with or without agonist exposure. Whilst GABA_BRs exhibit significant levels of basal phosphorylation this is not subject to agonist-induced modulation. Consistent with this neither the R1 or R2 subunits are substrates of GRKs 1–4 when co-expressed in HEK-293 cells. The major intracellular domains of these proteins are not phosphorylated by purified GRKs in vitro (Perroy et al., 2003; Fairfax et al., 2004). In contrast to other GPCRs, GABA_BRs are not GRK substrates and do not appear to undergo agonist-induced internalization. We discuss which kinases mediate GABA_BR phosphorylation and consider the significance of this covalent modification in determining the efficacy of receptor signaling and trafficking itineraries. We then go on to assess the physiological significance of phosphorylation and its impact on GABA_BR signaling under normal and pathological conditions.

I. PKA

The phosphorylation of GPCRs by PKA has been shown to regulate receptor activity (Bouvier et al., 1988; Moffet et al., 1996; Bunemann and Hosey 1999). For instance, for the β-adrenergic receptor and rhodopsin receptors, PKA phosphorylation enhances receptor desensitization by inducing internalization. Consistent with other GPCRs a role for PKA in mediating GABA_BR receptor desensitization had been postulated prior to their structural characterization (Yoshimura et al., 1995). In agreement with this study a strong consensus site for PKA phosphorylation is found at serine 892 (S892) within the R2 subunit (Q-R-R-L-S-L). Studies using in vitro phosphorylation with purified PKA coupled with metabolic labeling of recombinant receptors expressed in fibroblast has confirmed that S892 is the sole site of phosphorylation for cAMP-mediated phosphorylation within GABA_BRs. Using phospho-specific antibodies significant levels of basal phosphorylation of S892 are evident within the brain, a phenomenon that is dependent principally upon the activity of PKA (Couve et al., 2001).

The role of PKA-mediated phosphorylation in mediating GABA_BR effector coupling has been examined by measuring their ability to activate GIRKs. Upon brief exposure to agonist phosphorylation of S892 via PKA decreases the time-dependent decrease (rundown) in the efficacy of GABA_BR- dependent activation of GIRKs in both HEK-293 cells and hippocampal neurons (Couve et al., 2001). This effect likely results from stabilization of GABA_BRs on the neuronal plasma membrane. In contrast to the majority of GPCRs, PKA-mediated phosphorylation of S892 in the R2 subunit thus appears to promote GABA_BR signaling. PKA regulation of GABA_BR might not occur as a result of agonist stimulation but as a result of cAMP accumulation by neurotransmitters that activate Gαs-coupled receptors. It remains to be determined which signaling pathways mediate S892 phosphorylation and thus GABA_BR effector coupling; however, it is well accepted that prolonged activation of GABA_BRs via the activation of Gαi leads to decreased PKA activity. Consistent with this prolonged activation of GABA_BRs leads to dephosphorylation of S892 and receptor degradation (Fairfax et al., 2004). Thus the phosphorylation status of S892, which is in part dependent upon agonist activation, is likely to play critical roles in maintaining GABA_BR cell surface stability and the strength of synaptic inhibition.

II. PKC

PKC activity had long been known to attenuate GABA_BR mediated inhibition of neurotransmitter release and receptor effector coupling (Dutar and Nicoll, 1988; Taniyama et al., 1991, 1992; Gahwiler et al., 1992). More recently the phosphorylation of R1 subunit by PKC has been reported by Bouvier and colleagues (Pontier et al., 2006). The activation of PKC induces the phosphorylation of R1 subunit followed by N-ethylmaleimide-sensitive fusion (NSF) protein dissociation from GABA_BR and promotes desensitization. Like PKA, PKC phosphorylation does not induce GABA_BR internalization. NSF is a molecular chaperone that is known to be crucial element in membrane fusion events. It has also been identified as a regulator of the α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPAR) (Nishimune et al., 1998; Osten et al., 1998; Noel et al., 1999), the γ-aminobutyric acid (GABA) type A receptor (GABA_AR) (Goto et al., 2005) and β2 adrenergic receptor cell surface expression (Cong et al., 2001). In Chinese hamster ovary (CHO) cells GABA_BR activity promotes PKC recruitment to the plasma membrane, inducing R1 subunit phosphorylation. Phosphorylation of R1 subunit is disrupted by preventing the binding of NSF to GABA_BRs using selective peptides; the association of NSF to GABA_BRs is clearly crucial for GABA_BR phosphorylation. As measured in CHO cells recruitment of PKC dependent upon NSF blocked baclofen-induced G protein activation by GABA_BRs but did not modify basal activity as measured using ³⁵S-GTP-γS binding. It remains to be determined which residues within GABA_BRs are PKC substrates and to ascertain the relevance of this novel regulatory mechanism for neuronal GABA_BRs.

III. AMPK

It is evident that, in addition to PKA, GABA_BRs are substrates for AMPK (Kuramoto et al, 2008). AMPK is a serine/threonine protein kinase that exists as a heterotrimer consisting of the catalytic α-subunit and regulatory β- and γ-subunits (Spasic et al., 2009). This kinase acts as an energy sensor that is rapidly activated when the cellular levels of AMP increase due to high metabolic activity or the pathological states of anoxia and ischemia (Kahn et al., 2005; Carling, 2005). To date the majority of AMPK substrates are metabolic enzymes; however, AMPK also binds with high affinity to the C-terminus of R1 but not R2 subunit. Interestingly, AMPK phosphorylates the C-tails of both R1 and R2 subunits in vitro. Using mass spectrometry and Edman degradation analysis two phosphorylation sites in R1 subunit (S917/923), and one site in R2 subunit (S783) have been identified. The physiological relevance of this phosphorylation has been examined by measuring the effects of AMPK activity on the activation of GIRKs in both expression systems and neurons. This revealed that the phosphorylation of R2 subunit on S783 decreases the desensitization of GABA_BRs by stabilizing the receptor complex at the plasma membrane. These results suggest that, similar to PKA, the AMPK-mediated phosphorylation of GABA_BRs either inhibits receptor endocytosis or increases receptor exocytosis to maintain synaptic inhibition.

3. Phosphorylation-independent desensitization of GABA_BRs by protein kinases

Desensitization to prevent overstimulation is a common feature of GPCRs. For most GPCRs, a time-dependent desensitization of the receptor appears to be mediated by direct phosphorylation of the GPCRs by GRK followed by arrestin and dynamin-dependent receptor internalization via clathrin-coated vesicle and recycling/degradation signaling pathways (von Zastrow, 2003; Marchise et al., 2006). Whilst GABA_BRs do not appear to be directly phosphorylated by GRKs, GRK4 and GRK5 have been reported to play a central role in agonist-induced desensitization (Perroy et al., 2003; Kanaide et al., 2007). For instance, the suppression of GRK4 expression in cerebellar granule cells strongly inhibits GABA_BR desensitization (Perroy et al 2003). Similarly, in Xenopus oocytes and baby hamster kidney (BHK) cells, expression of GABA_BR and GIRKs does not result in desensitization unless co-expressed with GRK4 or GRK5 (Kanaide et al., 2007). Interestingly, the association of GRKs to R2 subunits is also observed but the GABA_BR phosphorylation is independent of GRK association (Perroy et al., 2003; Fairfax et al., 2004; Kanaide et al., 2007). Unlike most GPCRs, GRKs may thus function as anchoring proteins to regulate GABA_BR activity but not phosphorylation.

4. GABA_BR endocytic sorting and the control of receptor cell surface stability

It is generally agreed that GABA_BRs do not show agonist-induced endocytosis; however, they do exhibit significant rates of constitutive endocytosis (Couve et al., 2001; Perroy et al., 2003; Fairfax et al., 2004; Vargas et al., 2008). Under basal conditions GABA_BRs endocytose as dimers via clathrin- and dynamin-dependent mechanisms and localize to Rab11-positive recycling endosomes. After constitutive endocytosis, large numbers of GABA_BRs recycle back to the plasma membrane to maintain steady-state cell surface numbers, presumably reflecting the long cell surface half-lives for these proteins as measured via biotinylation (Fairfax et al., 2004; Vargas et al., 2008). Interestingly endocytosis is detected only in dendrites, not in axons. The mechanisms underlying this compartmentalization of GABA_BR endocytosis remain obscure; however, R1 subunit has many splice variants, which likely results in multiple distinct modes of phosphorylation. Consistent with steady-state measurement GABA_BR endocytosis is agonist-independent; however, cell surface accumulation is dramatically reduced by exposure to glutamate via a mechanism dependent upon the activity of the proteasome (Vargas et al., 2008). In this study glutamate treatment did not result in enhanced intracellular accumulation of GABA_BRs, suggesting that this excitatory neurotransmitter may regulate lysozomal targeting of GABA_BRs. Further studies are needed to determine how glutamate treatment of neurons leads to degradation of GABA_BRs.

5. GABA_BR phosphorylation and diseases

I. Addiction

Recently GABA_BR agonist and allosteric modulators have been successfully used to treat the symptoms associated with withdrawal from cocaine, opiates, nicotine and ethanol (Filip and Frankowska, 2008; Frankowska et al, 2009; Addolorato et al, 2009). This may be due to the phosphorylation of GABA_BRs. Kalivas and colleagues reported that repeated cocaine administration leads to an increase in basal extracellular GABA in the nucleus accumbens and to dephosphorylation of R2 subunits (Xi et al., 2003). Because prolonged activation of GABA_BRs decreases cAMP production and PKA activation through Giα, elevated extracellular GABA by repeated cocaine injection may regulate S892 phosphorylation in R2 subunit. In this study, the authors performed an immunoprecipitation assay followed by immunoblotting using anti-phospho-serine antibody. Additional experiments using phospho-specific antibody against S892 and S783 is clearly essential. Repeated cocaine administration can regulate the release of various neurotransmitters, and thus would also be able to regulate many of the second messenger kinases that phosphorylate GABA_BRs.

II. Ischemia

Our group recently demonstrated that the phosphorylation of GABA_BRs by AMPK indicates a possible neuroprotective role for GABA_BRs (Kuramoto et al., 2007). AMPK is expressed in neurons throughout the brain. Rapid AMPK activation has been observed following ischemia, hypoxia, and glucose deprivation (Culmsee et al., 2001; Gadalla et al., 2004; Kahn et al., 2005; Carling, 2005; Li et al., 2007; Spasic et al., 2009). Phosphorylation of S783 but not S892 in R2 subunit has been documented in the CA3 and dentate gyrus of hippocampus after ischemic injury induced by MCAO (middle cerebral artery occlusion). The overexpression of R2 subunits in cultured hippocampal neurons also shows phosphorylation of S783 and enhanced neuronal survival in experiments using a model of anoxia (oxygen-glucose deprivation). Since mutation of S783 to alanine significantly decreases neuronal survival compared to wild-type in this model, phosphorylation of S783 in R2 subunit is likely to be a critical mechanism for neuronal survival.

6. Conclusion

GABA_BRs are unique GPCRs that function as heterodimers composed of R1 and R2 subunits. They mediate slow and prolonged inhibitory signals. Deficits in GABA_BR function have been reported in many neurological and psychiatric disorders. We have discussed here current findings on the involvement of protein kinases in GABA_BR modulation; however, the existence of phosphorylation-dependent GABA_BR endocytosis is still unclear. Identifying the molecular machinery that regulates GABA_BR trafficking is thus key in developing novel therapeutics to treat neurological and psychiatric disorders.

Abbreviations

AMPA

α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid

AMPK

AMP-activated protein kinase

BHK

baby hamster kidney

cAMP

cyclic adenosine monophosphate

CHO

Chinese hamster ovary

GABA

γ-aminobutyric acid

GABA_BR

GABA_B receptor

GIRKs

G protein-activated inwardly rectifying K⁺ channels

GPCRs

G protein-coupled receptors

GRKs

G protein-coupled receptor kinases

HEK

human embryonic kidney

MCAO

middle cerebral artery occlusion

NSF

N-ethylmaleimide-sensitive fusion

PKA

cAMP-dependent protein kinase

PKC

protein kinase C