G Protein α i/o/z
Adenylyl cyclase inhibitory Gi alpha subunit; G alpha (o); G alpha z; G protein alpha i; G protein alpha o; G protein alpha z; G protein a i; G protein a o; G protein a z; Gi protein alpha subunit; GNAI; GNAO; GNAO1; GNAZ; Go alpha subunit; Guanine nucleotide binding protein, alpha inhibiting; Guanine nucleotide binding protein, alpha o; Guanine nucleotide binding protein, alpha z subunit; Guanine nucleotide binding regulatory protein, alpha i; Gz alpha subunit
Historical Background: Discovery of G Protein α i as the Inhibitor of Hormone-Stimulated Adenylyl Cyclase Activity
Gi heterotrimers follow a similar biosynthetic route as most G proteins, but aspects of Gαi-class subunit biosynthesis include unique processing. Gα subunits are translated on free cytosolic ribosomes and the initiator methionine is cleaved. The resultant amino terminal Gαi/o/z glycine residue becomes myristoylated. N-myristoyl transferase catalyzes the covalent attachment of a 14-carbon myristate chain to the glycine amino group during protein translation (Mumby et al. 1990). The nascent, myristoylated Gαi chain is folded by action of the cytosolic chaperone complex (CCT), as documented for Gαi family member G protein α Transducin (Farr et al. 1997). A new role for Gα non-receptor Ric-8 guanine nucleotide exchange factors (GEFs) showed that Ric-8A may aid Gαi folding with the CCT, or function after folding is mostly complete to promote the initial membrane association of nascent Gαi/o subunits. In Ric-8A−/− cells, Gαi (and also Gαq and Gα12/13) subunits exhibit membrane targeting defects and were subjected to rapid turnover (Gabay et al. 2011). Gαi binds to Gβγ on the endoplasmic reticulum membrane to form the nascent Gi heterotrimer, which is requisite for trafficking of the intact heterotrimer to the plasma membrane. Posttranslational Gαi/o/z palmitoylation at cysteine 3 is also required for G protein plasma membrane targeting (Fig. 1a). ER or Golgi-enriched DHHC palmitoyl transferases are likely responsible. The actual G protein heterotrimer trafficking mechanism has not been elucidated, but may involve a diffusive or membrane sampling type of mechanism until plasma membrane residence is achieved. Once G protein heterotrimers reach the inner leaflet of the plasma membrane, they are considered mature and sufficient to transduce signals from G protein coupled receptors (GPCRs). Plasma membrane residence is not static and Gαi/o and Gβγ subunits undergo agonist-dependent and -independent translocation to (and from) other cellular residences, including the Golgi during a process that may intersect a dynamic palmitoylation and depalmitoylation cycle (for comprehensive reviews of G protein trafficking mechanisms and accounts of primary references therein, see Chisari et al. (2007), Marrari et al. (2007), Saini et al. (2009)).
Gαi Structure and G Protein Catalytic Mechanism
Gαi/o/z subunits share primary structural features common to all heterotrimeric G protein α subunits. Each contain a core Ras small GTP-binding protein homology domain consisting of Gαi amino terminal amino acids ~1–60, interrupted by a region of ~120 amino acids, followed by the carboxyl-terminal ~175 amino acids that complete the Ras homology domain. The intervening region has high α-helical content and is commonly referred to as the Gα subunit α-helical domain. It is not found in small GTP binding proteins (Fig. 1a). Guanine nucleotide and its Mg+2 cofactor bind the Ras domain and are sandwiched between the Ras and α-helical domains. Like other Gα subunits, the Gαi carboxyl terminal residues constitute one region responsible for the specificity of G protein-receptor coupling.
G protein catalytic mechanisms have been elucidated biochemically and structurally, predominantly using Gαi1, as well as Gαs and Gαq as the prominent model G proteins (for review and primary references therein, see Elliott (2008), Gilman (1987), Sprang (1997)). The G protein catalytic cycle consists of three primary steps, GDP release (intrinsic Gαi1 rate: ~0.02–0.05 min−1), subsequent GTP binding to the open, nucleotide-free Gαi subunit (predicted to be very fast), and subsequent GTP hydrolysis (Gαi1 rate ~ 3 min−1). As is evident, GDP release is rate limiting to the steps of Gαi activation and single turnover and steady-state hydrolysis of GTP. GPCRs act as guanine nucleotide exchange factors (GEFs) for Gi heterotrimers and accelerate the GDP release rate such that it may no longer be the slowest step in the G protein catalytic cycle. In many activated-GPCR signaling contexts, the intrinsic Gαi GTP hydrolysis rate becomes limiting, underscoring the relevancy of the action of RGS GTPase activating proteins. RGS proteins bind to Gαi (and Gαq and Gα12/13-class) subunits and accelerate the GTP hydrolysis rate to keep pace with the rate of GPCR-stimulated GDP release (and apparent GTP binding) (Berman and Gilman 1998). The mechanism of Pertussis toxin inhibition of Gi is to ADP-ribosylate Gαi subunits. This renders Gi heterotrimers as non-substrates for GPCR-mediated activation.
The X-ray crystal structures of various forms of Gαi helped reveal the important features of G protein function and catalysis, and demonstrated the changes in the G protein Ras domain switch regions that occur during the transition from the GDP-bound to GTP-bound state. The conformational differences between these two states enable the G protein to interact with different sets of protein binding partners. G protein structures and mechanisms of action are reviewed in exquisite detail by Sprang (1997). Many structures of G proteins were produced using Gαi subunits as the model G protein and prominent examples include the following structures with PDB ID numbers: RCSB PDB (www.pdb.org): Gαi1-GTPγS (1GIA) (Fig. 1b) (Coleman et al. 1994), Gαi1-GDP-AlF4− (1GFI) (Coleman et al. 1994), Gαi1-GDP (1GDD) (Mixon et al. 1995), Gαi1-GDP:Gβγ (1GP2) (Wall et al. 1995), Gαi/αT chimera:Gβγ (1GOT) (Lambright et al. 1996), and Gαi1-GDP-AlF4−:RGS4 (1AGR) (Tesmer et al. 1997).
GPCR-Regulated Gi Signaling
Gαi-GTP targets – Agonist-stimulated GPCRs coupled to Gi-class heterotrimers produce Gαi-GTP and Gβγ. Gαi/z-GTP interacts directly with adenylyl cyclase (AC) isoforms I, V, and VI to inhibit catalytic production of the soluble second messenger cAMP. In comparison to Gαi, Gαo has reduced ability to inhibit adenylyl cyclase I. Gαs-GTP is the G protein stimulator of all membrane-bound adenylyl cyclase isoforms. The Gαi-GTP interaction site of AC is distinct from the Gαs-GTP binding site. Combined mutagenesis, biochemical, and kinetic modeling studies show that the opposed allosteric action of these G proteins may occur in simultaneous fashion (Chen-Goodspeed et al. 2005; Dessauer et al. 1998; Taussig et al. 1994). Gβγ liberated from Gi heterotrimers also influences AC activities in an AC isoform-dependent manner (activation or inhibition depending on isoform). The activities of Gβγ and the small molecule AC activator, forskolin are synergistic to the action of Gα-GTP (for comprehensive reviews and accounts of primary references therein, see Sadana and Dessauer (2009), Sunahara et al. (1996), Taussig and Gilman (1995)).
Free Gαi subunits bisect tyrosine kinase signaling pathways in ways that are unique from Gβγ (Gi) regulation. In vitro c-SRC tyrosine kinase activity was activated directly by purified Gαi-GTPγS (and Gαs-GTPγS), and c-SRC-dependent phosphorylation of cellular substrates appeared to be enhanced by co-expression of (activated) GTPase-deficient Gαi or Gαs (Ma et al. 2000). Reciprocally, c-SRC can phosphorylate Gαi and Gαs on tyrosine residues to alter adrenergic receptor coupling to the phosphorylated G proteins (Hausdorff et al. 1992). There are multiple modes of pathway integration and crosstalk between Gi and SRC or other tyrosine kinases (Natarajan and Berk 2006). Gαi and Gβγ subunits liberated from Gi heterotrimers have both redundant and opposed contextual effects toward tyrosine kinase signaling outputs.
Gi-dependent Gβγ targets – The majority of cellular responses to GPCR-induced Gi heterotrimer activation are arguably manifested by Gβγ. One explanation of why Gβγ dimers released from Gαi-GTP elicit signaling responses that other Gα/GPCR species do not is that Gi-class heterotrimers constitute the majority of expressed G proteins in many tissues. The “dose” of Gβγ produced from Gi heterotrimers may be above a particular response threshold for a given effector enzyme. This threshold may not be attainable by release of Gβγ from lower expressed G protein heterotrimer subtypes (Gq, Gs, G12/13). Regulated subcellular localization and scaffolding of G proteins and effectors also contributes to the specificity of Gi-mediated Gβγ signaling.
The complete list of effectors regulated by Gβγ subunits is expansive and beyond the scope of this Gαi topical essay (for comprehensive reviews of Gβγ signaling and the roles of Gi, see Clapham and Neer (1997), Smrcka (2008)). In brief, Gi-derived Gβγ activates the mitogen-activated protein kinase (MAPK) signaling cascade. Extracellular regulated kinases (ERKs) and downstream kinases are phosphorylated in response to cell treatment with agonists that stimulate Gi-coupled receptors (Gutkind 2000). Phospholipase Cβ (PLCβ) enzyme activity is co-modulated by Gαq-GTP and Gβγ subunits released from Gi heterotrimers to mediate phosphatidylinositol 3,4-bisphosphate (PIP2) hydrolysis (Exton 1994). The produced inositol trisphosphate (IP3) binds IP3 receptors to activate Ca+2 release from intracellular stores. Gβγ directly regulates the activities of many ion channels. Two prominent examples are activation of G protein inwardly rectifying potassium (GIRK) channels that regulate cellular K+ influx, and inhibition of N-type Ca+2 channels that convert neuronal action potentials to neurotransmitter release through cellular Ca+2 influx (Nathan 1997; Tedford and Zamponi 2006). Gi heterotrimer activation potentiates signaling through the phosphatidyl-inositol-3 kinase (PI3Kγ) and protein kinase B/Akt signaling pathways. The 110 kDa PI3Kγcatalytic subunit is a direct effector of Gβγ (Stephens et al. 1994).
Gi-Family Regulation of Vesicle-Mediated Protein Transport
Gi family members impart multiple modes of regulation toward intracellular trafficking and secretory processes by influencing vesicle budding, priming, and fusion events. Gαi3 overexpression or PTX treatment disrupted protein trafficking through the secretory pathway (Stow et al. 1991). The heterotrimeric G-protein activators AlF4−, mastoparan and related peptides, and compound 48/80 inhibited ER to Golgi transport (Beckers and Balch 1989; Schwaninger et al. 1992). PTX treatment also blocked vesicle budding from the trans-Golgi (Barr et al. 1991), and studies in model organisms revealed that Gαi/o proteins were important regulators of exocytosis (Ch’ng et al. 2008; Hajdu-Cronin et al. 1999; Miller et al. 1999; Lackner et al. 1999; Vashlishan et al. 2008).
One of the most well-studied trafficking pathways influenced by Gi family members is the regulation of insulin secretion by pancreatic β cells. Knowledge of PTX enhancement of glucose-stimulated insulin secretion (GSIS) predated the actual discovery of Gαi by at least a decade. PTX is known as islet activating protein (IAP) (Katada and Ui 1979, 1981a) and PTX treatment enhanced GSIS whether administered systemically or directly to primary cultures of pancreatic islets or β cell lines (Katada and Ui 1979, 1981a, b; Gulbenkian and Schobert 1968; Szentivanyi et al. 1963; Tabachnick and Gulbenkian 1969; Yajima et al. 1978). Subsequent reports showed that insulin secretion was inhibited by hormones that are Gi/o-coupled (reviewed in Sharp 1996). Consistent with these findings, selective ectopic expression of PTX in pancreatic β cells induced basal hyperinsulinemia and enhanced GSIS and glucose tolerance (Regard et al. 2007).
The specific Gαi family subunit(s) responsible for regulating insulin secretion has not been determined since the PTX effects would implicate Gαi1, Gαi2, or Gαi3 and/or Gαo. In fact, the regulation by Gαi family members appears to be redundant in part, as individual roles for Gαi, Gαo, and PTX-insensitive Gαz have been described. In the case of Gαi2, chemically-induced diabetic rodent models exhibited decreased Gαi2 expression in liver (Gαwler et al. 1987) and adipose tissue (Baculikova et al. 2008), and antisense-mediated loss of Gαi2 expression in liver and white adipose tissue lead to the development of insulin resistance (Moxham and Malbon 1996). Manipulation of the activation state of Gαi2 in mice, either by conditional expression of GTPase-deficient Gαi2-Q205L or homozygous knock-in of Gαi2-G184S RGS-insensitive alleles (suggesting a prolonged time of Gαi2 in the activated, GTP-bound state) resulted in enhanced glucose tolerance (Chen et al. 1997) or mice that were lean, resistant to high-fat diet-induced diabetes and had increased insulin sensitivity (Huang et al. 2008). These data suggest a key role for Gαi2 in insulin secretion and signal regulation.
Gαo also participates in the regulation of insulin secretion by inhibiting secretory vesicle docking. A mouse with a conditional null Gαo allele in pancreatic islet cells had improved glucose tolerance and the effect of PTX to enhance insulin secretion was blocked, which led the authors to conclude that Gαo is the primary target of the PTX effect on insulin secretion (Zhao et al. 2010a). It was subsequently demonstrated that the GαoB isoform was responsible for this effect (Wang et al. 2011).
Although most aspects of hormonal regulation of insulin secretion are PTX-sensitive, this is not always the case. Kimple and colleagues demonstrated that PGE1 inhibition of GSIS was insensitive to PTX and that this effect was blocked in a pancreatic β-cell line with reduced Gαz expression (Kimple et al. 2005). The same authors also demonstrated that Gαz-null mice are hyperinsulinemic and have increased glucose tolerance (Kimple et al. 2008). Collectively, Gαz appears to have an important role in hormonal regulation of insulin secretion apart from the roles ascribed to the PTX-sensitive Gαi and Gαo regulation.
Gi-family regulation of insulin vesicle secretion is manifested by liberated Gβγ subunits and shares a conserved mechanism with Gi regulation of neurotransmitter release. Gβγ subunits inhibit neurotransmitter release (Blackmer et al. 2005, 2001; Gerachshenko et al. 2005) by binding directly to the SNARE complex component SNAP-25 to disrupt synaptic vesicle fusion (Blackmer et al. 2005; Gerachshenko et al. 2005). SNARE complexes and intracellular calcium levels similarly regulate insulin-containing secretory vesicle fusion and therefore insulin release from pancreatic β cells (Wang and Thurmond 2009). The secretory-promoting effects of elevated Ca+2 are blocked by noradrenaline, implicating Gi-family heterotrimers. Noradrenaline-inhibited insulin secretion was also blocked by Gβ antibodies, a Gβγ-activating peptide mSIRK, or Botulinum A toxin which cleaves the Gβγ binding site from the SNAP-25 carboxyl terminus (Zhao et al. 2010b). These results show that Gβγ released from Gi heterotrimers is the responsible G protein species that regulates secretory vesicle fusion and insulin/neurotransmitter release.
Noncanonical Gαi Signaling
As G-proteins became established as signal transducers for GPCRs at the cell surface, functional evidence suggested additional roles for Gi and Gαi in subcellular regions and contexts distinct from GPCR and classic effector signaling. Cell fractionation and immunofluorescence studies demonstrated populations of Gi proteins that did not reside on the plasma membrane, and in some cases Gαi was not always associated with Gβγ (Stow et al. 1991; Denker et al. 1996; Lin et al. 1998; Maier et al. 1995; Montmayeur and Borrelli 1994; Muller et al. 1994; Ogier-Denis et al. 1995; Pimplikar and Simons 1993; Schurmann et al. 1992; Wilson et al. 1993, 1994). Noncanonical roles for Gαi proteins include regulation of Golgi structure and function, (Jamora et al. 1999; Yamaguchi et al. 2000), signaling interactions with tyrosine kinase and steroid hormone receptors (Kreuzer et al. 2004; Kumar et al. 2007), and interactions with GPR/GoLoco proteins to regulate mitotic spindle positioning and asymmetric cell division (Yu et al. 2000; Gotta and Ahringer 2001; Parmentier et al. 2000; Schaefer et al. 2000).
Gαi Regulation of Golgi Function and Structure
Gi-subunits, particularly Gαz and Gαi3, are Golgi-localized (Stow et al. 1991; Wilson et al. 1993, 1994; Stow and de Almeida 1993). In addition to regulating vesicular protein transport through the Golgi, Gi/o-family proteins regulate overall Golgi structure. Chemically induced Golgi fragmentation was blocked by exogenous application of Gα subunits, including Gαi3, to permeabilized NRK cells (Jamora et al. 1997) or by overexpression of Gαz or Gαi2 (Yamaguchi et al. 2000). Expression of a putative dominant-negative Gαz mutant in HeLa cells disrupted Golgi structure (Nagahama et al. 2002). Golgi disruption was also observed upon overexpression of the Gαz-selective GTPase-activating protein RGSz (Nagahama et al. 2002). It will be important to determine which aspects of Gi regulation of vesicular transport and Golgi structure are related to each other (or not), and to discriminate those processes regulated through receptor-independent mechanisms and/or by Gi-coupled GPCRs.
Gαi Interactions with Steroid Hormone and Tyrosine Kinase Receptors
Gαi-GDP and Gβγ directly bind a variety of steroid hormone receptors in vitro, including estrogen receptor α (ERα). The effect of estrogen on ERa-dependent eNOS signaling and monocyte adhesion to endothelial cells was blocked by disruption of the ERα-Gαi interaction (Kumar et al. 2007). The mechanistic basis of this process awaits elucidation, but may involve a nontraditional, nucleotide-independent process of Gi heterotrimer activation by ERα. Gi heterotrimers functionally interact with a variety of tyrosine kinase receptors including the insulin receptor (IR), the epidermal growth factor receptor (EGFR), and the platelet-derived growth factor receptor (PDGFR) (Patel 2004). In most part, these functional interactions were established by demonstration that natural ligand (insulin, EGF, PDGF) effects on MAP kinase pathway activity were altered by cell pretreatment with PTX. Gαi2 was actually shown to be recruited to the IR complex in a PTX- and guanine nucleotide-sensitive manner (Kreuzer et al. 2004). A role for Gαi proteins downstream of IR signaling was also supported by observation of PTX-sensitive antiautophagic responses to insulin in hepatocytes and insulin-sensitive localization of Gαi3 to autophagic endomembranes (Gohla et al. 2007).
Gαi Regulation by Accessory Proteins – GPR Motif Proteins and non-receptor GEFs
Gαi family proteins are targets of many regulatory mechanisms and interacting proteins, perhaps more than any other family of Gα subunits. The identification of accessory proteins as regulatory factors included the use of protein-protein interaction screens, purification of biochemical activities, forward genetic screens, and expression cloning methods (reviewed in Sato et al. 2006), also see Blumer & Lanier ESM Review of AGS Proteins). Among the Gαi-family accessory proteins identified are non-receptor GEFs (e.g., GAP-43, AGS1, Ric-8A, GIV), GAPs (e.g., RGS proteins), and guanine nucleotide dissociation inhibitors (GDIs) (e.g., proteins containing the GPR/GoLoco motif) (reviewed in Sato et al. 2006; Siderovski et al. 2005). Functional roles for each of these classes has perhaps been most clearly developed in model organisms, where the GPR-Gαi module appears to be involved in the integration of polarity cues with the orientation of the mitotic spindle during asymmetric cell division. Interestingly, a receptor-independent Gαi activation/deactivation cycle is implicated this process (Gonczy 2008; Knoblich 2010; Siderovski and Willard 2005). The discovery and overviews of each of these classes of accessory proteins are covered in depth elsewhere (Siderovski and Willard 2005; Blumer et al. 2011, 2007; Hollinger and Hepler 2002; McCudden et al. 2005; Ross and Wilkie 2000; Sato et al. 2006). The remainder of this subsection will serve to highlight recently reported key regulatory roles of these accessory proteins on Gαi/o function.
The Gαi-GPR module is also a substrate for Ric-8A-catalyzed nucleotide exchange in a manner that may be analogous to GPCR-mediated regulation of Gαiβγ heterotrimers (Tall and Gilman 2005; Thomas et al. 2008; Vellano et al. 2011b). A non-GPCR-mediated Gαi-GPR activation (by Ric-8) and deactivation (by C. elegans RGS7) cycle and has been proposed to be the means by which the Gαi protein switch regulates mitotic spindle positioning processes during asymmetric cell division (Afshar et al. 2004; Couwenbergs et al. 2004; Hess et al. 2004; Wilkie and Kinch 2005). The non-receptor GEF, GIV/Girdin also appears to act on GPR-Gαi complexes in the regulation of autophagy (Garcia-Marcos et al. 2011).
The Gαi-class is one of four subfamilies of G protein α subunits. The Gαi subfamily has the largest number of individual members, and in most cases constitutes the bulk of expressed G protein α subunits in a given tissue or cell type. Gαi was discovered as a key component in the hormonal inhibition of adenylyl cyclase. GPCR signals that are transduced through Gi heterotrimers are propagated directly by the activated Gαi-GTP subunit, but most Gi signaling arguably stems from the Gβγ subunits of Gi heterotrimers. New roles for Gαi subunits have emerged more recently, in which Gαi acts independently of Gβγ. Gαi interaction with GPR/Goloco domain-containing proteins provides a means to regulate distinct signaling pathways including intracellular events that do not always occur at the cell periphery.
- Gohla A, Klement K, Piekorz RP, Pexa K, von Dahl S, Spicher K, Dreval V, Haussinger D, Birnbaumer L, Nurnberg B. An obligatory requirement for the heterotrimeric G protein Gi3 in the antiautophagic action of insulin in the liver. Proc Natl Acad Sci USA. 2007;104:3003–8.PubMedCrossRefPubMedCentralGoogle Scholar
- Kimple RJ, Willard FS, Hains MD, Jones MB, Nweke GK, Siderovski DP. Guanine nucleotide dissociation inhibitor activity of the triple GoLoco motif protein G18: alanine-to-aspartate mutation restores function to an inactive second GoLoco motif. Biochem J. 2004;378:801–8.PubMedCrossRefPubMedCentralGoogle Scholar