Regulation of phospholipase C-β1 GTPase-activating protein (GAP) function and relationship to Gq efficacy
Department of Molecular and Cellular Pharmacology, University of Miami Miller School of Medicine, Miami, FL, USA
Address correspondence to: Irene Litosch, Department of Molecular and Cellular Pharmacology, University of Miami Miller School of Medicine, Miami, Florida 33101-6189, USA. Tel.: 3052435862. Fax: +13052434555. E-mail: email@example.com
The physiological role of PLC-β1 as a GTPase activating protein (GAP) remains unclear and controversial [1, 2]. GAPs are generally perceived to limit signal kinetics and signal amplitude for the Gq/11 (Gq) and Gi/o subfamily of G proteins. GAPs accelerate the Gα GTPase cycle to shorten the life-span of Gαq-GTP. Overexpression of GAPs has been shown to deactivate Gq signaling.
Unlike most GAPs, PLC-β1 GAP is specific to Gαq . PLC-β1 is also an effector for Gq. An alternative function for PLC-β1 GAP to increase signaling efficiency has been proposed [1, 2]. PLC-β1 GAP was found to kinetically couple with ligand-activated GPCR to stabilize the three protein complex consisting of m1 muscarinic cholinergic receptor, Gq and PLC-β1, to sustain signal amplitude [4, 5].
Which paradigm correctly describes the physiological function of PLC-β1 GAP? The answer to this challenging question is important to basic research and drug discovery. Gq is a key regulator of most human behavior but its potential as a high-value therapeutic target remains to be fully realized . Gq receives and processes information from the ligand-activated GPCR to implement changes in signaling networks as is necessary to achieve appropriate response. PLC-β1 GAP is positioned to determine Gq efficacy and GPCR pharmacology.
This article proposes a new hypothesis that could unify both paradigms for PLC-β1 GAP function. It is hypothesized that PA stimulates PLC-β1 GAP activity. The physiological function of PLC-β1 GAP is regulated by PA and is dependent on signaling via the unique PLC-β1 PA binding domain.
The Life Cycle of Gαq-GTP and PLC-β1 GAP
GPCRs function as guanine nucleotide exchange factors (GEF) to promote the exchange of cytosolic GTP for the tightly bound GDP on Gαq-GDP of the heterotrimer . This rate limiting step in the G protein cycle generates Gαq-GTP. Signaling is terminated by the intrinsic Gαq GTPase activity, which converts Gαq-GTP back to Gαq-GDP. The intrinsic Gαq GTPase activity is accelerated by GAPs. Overexpression of GAPs inhibits Gq signaling [1, 2].
PLC-β1 GAP however can also increase signaling efficiency [4, 5]. PLC-β1 GAP was found to co-ordinate with the ligand-activated m1 muscarinic cholinergic receptor to regulate the amplitude of the PLC-β1 signal. Essentially, PLC-β1 GAP and GPCR GEF worked as a team to maintain a fraction of Gαq in the active GTP-bound form to stabilize the complex of GPCR, Gq, and PLC-β1. PLC-β1 GAP was subsequently shown to potentiate GPCR GEF activity to increase the rate of dissociation of bound-GDP, the rate limiting step in G protein activation .
Residues required for high affinity Gαq stimulation, GAP activity, maximum activity and association with membrane have been localized to the PLC-β C-terminal extension, which contains a coiled-coil domain [9-14]. This domain is unique to the PLC-β1–4 family. PLC-β also have a conserved core, common to all inositide-specific PLCs, that contains the pleckstrin homology domain, four tandem EF hand domains, the triose phosphate isomerase barrel-like catalytic domain and a C2 domain [15, 16].
The crystal structure of a truncated PLC-β3 in complex with activated Gαq revealed that the Gαq Ras-like GTPase domain makes multipoint contacts with the enzyme . PLC-β3 GAP was found to depend on residues within a loop off the EF hand. The Gαq construct was truncated at the amino terminus and therefore lacked the sites for attachment of palmitate. The truncated PLC-β3 lacked the coiled-coil domain, a region of ∼300 residues that is unique to the PLC-β family. The affinity of the truncated PLC-β3 for binding to Gαq was found to be similar to that of the full length enzyme. It was therefore concluded that the role of the coiled-coil domain was to mediate membrane association, that is, indirect regulation. This conclusion however is clouded by two inconvenient truths. First, dependence on the coiled-coil domain is specific to Gαq. This domain is not required for stimulation of lipase activity by Gβγ subunits, Rac or Ca2+. Second, isolated coiled-coil domain fragments exhibit GTPase activity for Gαq . Therefore, a second interaction site with Gαq must exist within the coiled-coil domain.
Indeed, the recent structure of full-length PLC-β3 in complex with activated Gαq reinstated the central role of the coiled-coil domain in regulation, independent of binding . The coiled-coil domain, specifically a hydrophobic patch of the Dα5 helix, was shown to directly interact with the N-terminal α-helical domain of Gαq during catalysis and was required for full efficacy. Perturbation of the hydrophobic patch in Dα5 resulted in a marked reduction in Gαq-stimulated lipase activity despite an intact membrane-binding surface on the coiled-coil domain. Gαq palmitoylation was found to be necessary for full efficacy . The coiled-coil domain also interacted with the core structure of PLC-β3.
Interaction between the coiled-coil domain and the N-terminal α-helical domain of Gαq strategically positions PLC-β GAP to mediate kinetic coupling. All Gα subunits share a similar tertiary structure composed of two domains, a Ras-like GTPase domain and α-helical domain . The Ras-like domain is conserved among all members of the G protein superfamily. This domain contains the site of GTP hydrolysis and binding sites for Gβγ, receptor and effectors. Truncated PLC-β3 interacted with this domain .
The Gα helical domain however is unique to the heterotrimeric Gα subunit and appears crucial for G protein/activated receptor interaction . The Gα helical domain covers the guanine nucleotide binding site of the Ras-like domain. Structural data of Gs in complex with ligand-bound β2 adrenergic receptor showed that the α-helical domain undergoes a marked displacement from the Ras-like domain exposing the nucleotide binding pocket [20, 21]. PLC-β1 GAP was shown to increase the rate of dissociation of bound-GDP, the rate limiting step in G protein activation .
Context-Dependent Regulation of PLC-β1 GAP Function
We demonstrated that the novel phospholipid mediator, phosphatidic acid (PA) regulated Gq efficacy [22, 23]. PA could increase Gq efficacy by stimulating PLC-β1 GAP. Regulation by PA depended on signaling via the unique PLC-β1 PA binding domain that we localized to the Gαq binding region of the coiled-coil domain [24, 25]. Residues required for stimulation were unique to PA and did not overlap with residues necessary for stimulation by Gαq or Gβγ, or for membrane-association [22, 26]. The PLC-β1 PA binding region is unique to the PLC-β1 isoform. In short, PLC-β1 appears programmed for regulation by PA.
Building on all this knowledge, a new hypothesis is proposed that could unify the two conflicting paradigms for PLC-β1 GAP function, as it pertains to physiological signaling and native levels of protein (Fig. 1). In the absence of ligand, PLC-β1 GAP activity could indeed function to deactivate Gq signaling. PLC-β1 GAP would determine the threshold for stimulation, that is, the ligand concentration at which a response is detected. What would be the physiological advantage to this function? PLC-β1 GAP would limit spurious or leaky Gq signaling to sharpen signaling kinetics. Many GPCRs have been found to exhibit basal or ligand independent activity that allows them to activate G proteins .
At threshold, the ligand-mediated increase in GPCR GEF activity increases the fraction of Gαq-GTP. PA would stimulate PLC-β1 GAP activity to increase signaling efficiency as dependent on the PLC-β1 PA binding domain. PA-stimulated PLC-β1 GAP would kinetically couple (coordinate) with GPCR GEF to increase signal amplitude and signal strength, that is, Gq efficiency.
In vitro assays using purified proteins reconstituted at controlled concentrations into phospholipid vesicles have shown that PLC-β1 is a very good GAP [4, 5]. PLC-β1 GAP accelerates hydrolysis of Gα-GTP by about 1000-fold . Furthermore, kinetic coupling was observed in the absence of PA. However PLC-β1 GAP activity in cells is not known and cannot be extrapolated from the in vitro data. Protein concentrations of GPCR, Gq, and PLC-β1 at the membrane and regulated association of these proteins is not known [8, 15].
Intriguingly, context-dependent regulation of PLC-β1 GAP function could contribute to Gq efficacy through Gβγ (Fig. 2). Gβγ signals independently of Gα but how Gβγ achieves specificity is unclear [28, 29]. It is thought that the Gβγ released from activation of Gi/o is exclusively responsible for all response. However, Gq could regulate localized signaling through its pool of Gβγ as dependent on PA-regulation and kinetic coupling.
The two major pathways that lead to an increase in levels of PA are the mammalian phospholipase D (PLD1, PLD2)  and the large family of diacylglycerol kinases (DGK; ). We identified RhoA-regulated PLD1 and DGKζ in the regulation of Gq efficacy in transfected COS-7 cells [23, 32]. DGKζ phosphorylates PLC-generated DAG to PA and therefore could regulate response through a localized increase in PA levels or inhibition of negative feedback regulation by protein kinase C.
PLC-β1 GAP and the Connection to Co-signaling with G12/13
GPCRs are classically depicted as monomers that signal with a single G protein subfamily. G protein subfamilies are grouped by their Gα subunit based on sequence identity and effectors . GPCRs, however, can co-signal with various G proteins [34, 35]. Muscarinic receptors co-signal with Gq and G12/13 (G12). The protease-activated receptor 1 (PAR1) co-signals with Gq, G12, Gs, and Gi. Many intriguing questions surround the mechanistic basis for G protein co-signaling. Does a GPCR signal independently with different G proteins? Could co-signaling depend on GPCR heterodimers? Does each G protein regulate a unique aspect of the overall response or do they coordinate to shape GPCR pharmacology?
A GPCR could co-signal by engaging different G proteins through unique binding sites. Mutagenesis showed that PAR1 uses different residues within the second intracellular loop to couple to G proteins . Replacement of five amino acids individually within the PAR1 i2 loop inhibited Gq but not G12 or Gi signaling.
Co-signaling could depend on GPCR heterodimers. GPCRs can form transient heterodimers [37, 38]. Dimerization of the rhodopsin class A GPCRs does not appear to be important for receptor function . Heterodimerization however may shape receptor pharmacology by altering ligand-binding properties, efficiency of G protein activation and selective engagement of downstream signaling pathways.
The herein proposed context-dependent regulation of PLC-β1 GAP function by PA suggests a mechanism for regulation of Gq efficacy by co-signaling as mediated by GPCR monomers or transient heterodimers (Fig. 3). In both scenarios, PA is supplied from ligand-activated GPCRs, coupled to G12-RhoA-PLD1 signaling, to stimulate PLC-β1 GAP to kinetically couple with GPCR GEF to increase Gq efficiency.
Lipidomics has revealed that cells generate many different species of lipids . Regulation of PLC-β1 GAP function could depend on the species of PA and their spatiotemporal dynamics. We do not know whether PA could be a common mechanism to regulate PLC-β GAP activity and kinetic coupling. PA was shown to stimulate PLC-β3 lipase activity .