G-Protein αq (GNAQ)
Hormone receptors elicit cellular responses most often not directly, but through diffusible second messengers. In the early 1970s it was found that second messenger synthesis requires guanosine triphosphate (GTP). Yet, GTP binds neither to hormone receptors nor to enzyme effectors. Instead, GTP binds to GTP-sensitive transducers (see Gilman 1987). These G-proteins are heterotrimers of α, β, and γ subunits. The first Gα subunits to be identified were Gαs, and transducin. Gαs triggers the formation of cAMP. Transducin mediates the effects of rhodopsin. Signaling by Gαs was found sensitive to pertussis toxin. Signaling by Gαi, which inhibits cAMP formation, was found sensitive to cholera toxin. Formation of the second messengers inositol trisphosphate (IP3) and diacylglycerol (DAG) by phospholipase C (PLC) was neither affected by pertussis toxin nor by cholera toxin. Therefore, a third class of Gα subunits was proposed. In 1990 Strathmann and Simon cloned murine Gα subunits that lack the sites modified by pertussis and cholera toxin. They termed them Gαq and suggested that they mediate signaling to PLC (Strathmann and Simon 1990).
Gene and Protein
The official gene symbol for Gαq in Homo sapiens and other mammals is GNAQ (guanine nucleotide-binding protein alpha q). GNAQ is the prototypical member of a G-protein family formed by GNAQ, GNA11, GNA14, and GNA15. As for the related Gαi family, the genomic organization is such that two Gαq genes are located in sequence, likely the result of tandem duplication (Wilkie et al. 1992). GNAQ is located on chromosome 9q21 together with GNA14. A very similar pseudogene is at 2q14.3–q21 (Dong et al. 1995).
The tertiary structure of Gαq has been solved (e.g., Nishimura et al. 2010). It is (reversibly) palmitoylated at N-terminal cysteins (C9, C10, Tsutsumi et al. 2009). The GTPase domain of Gαq is similar to the small, monomeric GTPase Ras. For isoforms of Gβ and Gγ not all combinations are possible, but any of these Gβγ dimmers can form a heterotrimer with Gαq. The heterotrimeric G-protein αqβγ is termed Gq after the α subunit.
Plasma membrane receptors signaling through Gq include the following G-protein coupled (GPCR) or seven-transmembrane domain (7TM) receptors: α1 adrenoreceptor, M1, M3, M5 muscarinic receptors, several P2Y purinergic (ATP, UTP) receptors, 5-HT2 serotonin, B2 bradykinin, H1 histamine, GPR55 cannabinoid, PAR2 trypsin, and group I metabotropic glutamate receptors. A close homologue of Gq mediates light perception in drosophila photoreceptors. Downstream responses include a rise in cytosolic calcium, which induces the secretion of hormones and neurotransmitters and the contraction of smooth muscles in vasculature and eye. Many of the GPCR activated by Gq are important drug targets. For instance, α1 receptors mediate the actions of blood pressure-lowering drugs and drugs treating benign prostatic hyperplasia, M1 receptors mediate the effects of cholinergic drugs treating Alzheimer’s disease, and H1 receptors are targeted by drugs treating allergies.
Mutations in GNAQ (and GNA11) have been associated with cancers. Eighty-three percent of uveal melanoma harbor mutations in GNAQ or GNA11 (Van Raamsdonk et al. 2009). Most mutations prevent GTP hydrolysis and make Gq constitutively active, but the exact mechanism by which Gq mutants cause malignancy are not known. GNAQ mutations are also frequent in Sturge-Weber syndrome and cutaneous “port-wine stains,” disorders involving vascular malformations (Shirley et al. 2013). These recent findings have made cells and animals expressing mutant Gq important cancer models for the development of new treatment strategies.
Overview of the G-Protein Cycle
In the classical view, Gαq has GDP bound at rest and forms a heterotrimer with Gβγ subunits. Activation of a Gq-coupled receptor (by ligand binding or other means) will lead to binding of Gq to the receptor and promote nucleotide exchange at Gαq, i.e., the unbinding of GDP and the binding of GTP. Gαq-GTP then dissociates from the receptor and from Gβγ. This free Gαq-GTP is the active form of Gαq. It binds to effectors, most importantly isoforms of phospholipase C β (PLCβ). PLCβ hydrolyzes the membrane phospholipid phosphatidylinositol (4,5)-bisphosphate (PIP 2), forming the cytosolic second messenger inositol trisphosphate (IP3) and the lipid diacylglycerol (DAG). In some cases, PLC activation depletes PIP2. IP3 releases calcium from intracellular stores; DAG activates protein kinase C (PKC); PIP2 is an activator for many plasma membrane proteins. The endogenous GTPase activity of Gαq hydrolyzes GTP to GDP. This ends the activation of Gαq. Gαq-GDP then binds again to Gβγ. If the receptor is still active, it can bind Gq again, catalyze nucleotide exchange, and induce another round of Gαq activity.
A More Detailed Description of the G-Protein Cycle
This section considers the individual steps of the G-protein cycle in more detail. Receptor activation evokes a conformational change in the receptor. This conformational changes occurs relatively fast (<100 ms in M1R, Jensen et al. 2009). The affinity of the activated receptor for Gq is higher than the affinity of inactive receptor, leading to net binding of G-proteins to the receptor. For M1R, G-protein binding takes about 200 ms. Because of the law of mass action, the speed of binding may depend on the local density of G-proteins in the vicinity of the receptor (Falkenburger et al. 2010).
Some investigators suggest that heterotrimeric Gq is preassembled with Gq-coupled receptors. Whether – and to which extent – this is the case has remained controversial (discussed, e.g., in Hein and Bünemann 2009 and Falkenburger et al. 2010). Likely there is some finite affinity, and the fraction of receptors that have Gq bound at rest depends on the local density of Gq molecules in the vicinity of the receptor. A related question is whether an inactive receptor can catalyze nucleotide exchange, i.e., whether the receptor is “precoupled” to Gq. Some effects of the M1R antagonist atropine can be interpreted in this way, but the extent of precoupling is certainly smaller than in adrenoreceptors. Finally, there is some evidence that different ligands may stabilize the receptor in different conformations, favoring binding of certain downstream partners over others and thus biasing signaling through one of several alternative pathways.
The active receptor acts as a GEF (guanine exchange factor) for Gαq, speeding up the exchange of GDP to GTP. The rate-limiting step of nucleotide exchange is the unbinding of GDP from Gαq. This step is accelerated when Gq is bound to activated receptor (Ross 2008). Without a GAP (GTPase activating protein), the endogenous GTPase activity of Gαq is very slow (time constant of 30 s, Falkenburger et al. 2010), leaving Gαq (and Gβγ) active for a fairly long time. Some effectors, including PLCβ, act as GAPs and accelerate the GTPase activity of Gαq, speeding up recovery. Other GAPs for Gαq include RGS proteins (regulators of G-protein signaling) and the Gαq effector p115RhoGEFe (Ross and Wilkie 2000).
If PLCβ would only accelerate GTPase activity, it would turn off the molecule that activates it, and PLCβ activation should be transient. However, this is not what was observed experimentally. PLCβ can be continually active for minutes. One possibility is that PLCβ also acts as GEF and accelerates nucleotide exchange (see Falkenburger et al. 2010). A different explanation is “kinetic scaffolding” (see Ross 2008), meaning that the GTPase activity of Gαq bound to PLCβ is so fast that GTP is hydrolyzed even before Gαq-GTP has had a chance to dissociate from the receptor. The receptor (a GEF) then catalyzes again nucleotide exchange and Gαq is kept at the receptor until the latter is no longer active. In both cases, fast GTPase activity is balanced by fast nucleotide exchange, and Gαq bound to PLCβ is kept active by a rapid cycle of nucleotide exchange and GTP hydrolysis. The purpose of such seemingly futile GTP consumption is to allow rapid signaling. Without the GAP/GEF effect, responses would persist and blur into one another.
The classical view holds that upon nucleotide exchange, Gq dissociates into Gαq and Gβγ. And indeed, some G-proteins do physically dissociate (Lambert 2008). Some Gβγ subunits even translocate to intracellular membranes upon activation while Gαq stays at the plasma membrane (Saini et al. 2009). However, for some G-proteins, activation results in a decreased distance between the Gαand Gβ subunits, which is inconsistent with subunit dissociation. Here, nucleotide exchange may merely result in a conformational rearrangement without physical dissociation (see Lohse et al. 2008). A related question is whether Gαq can stay bound to the receptor while binding and activating PLC. For Gαq, an increase in the average distance to Gβ was observed upon activation along with a decrease in the distance to PLCβ (Jensen et al. 2009). This is consistent with the classical view. Nonetheless, effectors can be activated without dissociation of Gαq and Gβγ (Yuan et al. 2007). In addition to the mentioned activation-induced translocation of G-protein subunits, recent evidence suggests a continuous shuttling of G-protein heterotrimers between the plasma membrane and intracellular membrane compartments at rest (Saini et al. 2009).
It has become clear that Gq can signal through other pathways than PLC. The Gq mutants associated with cancer lead to translocation of the transcription factor YAP (yes-associated protein 1) into the nucleus. YAP is known to stimulate proliferation and is inhibited by the Hippo signaling pathway when organs reach a given size. This effect of Gq mutants on YAP is independent of PLC but dependent on the GEF Trio and the Rho family GTPases RhoA and Rac (Vaque et al. 2013; Feng et al. 2014; Yu et al. 2014). Gq can activate RhoA through p115RhoGEFe by favoring nucleotide exchange (Rojas et al. 2007). Proliferative effects of Gq mutants also require the small GTPase ARF6, likely for trafficking of Gq to the plasma membrane (Yoo et al. 2016). Next to the activation of p115RhoGEFe-RhoA and Trio-YAP, Gq can bind to and activate Bruton’s tyrosine kinase (Ma and Huang 1998) and possibly further kinases.
Gαq mediates signaling from ubiquitous Gq-coupled plasma membrane receptors to PLCβ. Since its identification, considerable progress has been made in understanding kinetics and functioning of the GDP/GTP cycle. In recent years, insight into Gq-coupled signaling has been boosted by the discovery of frequent GNAQ mutations in uveal melanoma and Sturge Weber syndrome. Further work is expected to advance our understanding particularly of the PLC-independent aspects of Gq signaling.