A. Yatani

Roles of G proteins in coupling of receptors to ionic channels and other effector systems

Guanine nucleotide binding (G) proteins are heterotrimers that couple a wide range of receptors to ionic channels. The coupling may be indirect, via cytoplasmic agents, or direct, as has been shown for two K+ channels and two Ca2+ channels. One example of direct G protein gating is the atrial muscarinic K+ channel K+[ACh], an inwardly rectifying K+ channel with a slope conductance of 40 pS in symmetrical isotonic K+ solutions and a mean open lifetime of 1.4 ms at potentials between -40 and -100 mV. Another is the clonal GH3 muscarinic or somatostatin K+ channel, also inwardly rectifying but with a slope conductance of 55 pS. A G protein, Gk, purified from human red blood cells (hRBC) activates K+ [ACh] channels at subpicomolar concentrations; its alpha subunit is equipotent. Except for being irreversible, their effects on gating precisely mimic physiological gating produced by muscarinic agonists. The alpha k effects are general and are similar in atria from adult guinea pig, neonatal rat, and chick embryo. The hydrophilic beta gamma from transducin has no effect while hydrophobic beta gamma from brain, hRBCs, or retina has effects at nanomolar concentrations which in our hands cannot be dissociated from detergent effects. An anti-alpha k monoclonal antibody blocks muscarinic activation, supporting the concept that the physiological mediator is the alpha subunit not the beta gamma dimer. The techniques of molecular biology are now being used to specify G protein gating. A bacterial alpha i-3 expressed in Escherichia coli using a pT7 expression system mimics the gating produced by hRBC alpha k.

Isolation and physiological characterization of taicatoxin, a complex toxin with specific effects on calcium channels

Taicatoxin is a new complex oligomeric toxin that was isolated from the venom of the Australian taipan snake Oxyuranus scutellatus scutellatus. It is composed of three different molecular entities: an alpha-neurotoxin-like peptide of mol. wt 8000, a neurotoxic phospholipase of mol. wt of 16,000 and a serine protease inhibitor of mol. wt 7000, linked by non-covalent bonds, at an approximate stoichiometry of 1:1:4. The most active form of the complex was isolated by ion exchange chromatography through DE-Cellulose followed by two steps of CM-Cellulose chromatography at pH 4.7 and pH 6.0, respectively. At this stage the complex migrates as a single component in beta-alanine-acetate-urea gel electrophoresis and is very toxic to mice (1 or 2 micrograms of the complex protein kills a mouse of 20 g within 2 hr). It blocks the high threshold calcium channel current of excitable membranes in heart and does not affect the low threshold calcium channel current. The block occurs at a site that is accessible extracellularly but not intracellularly. The block is selective for calcium channels, reversible, does not affect single channel conductance but only changes channel gating, and is voltage dependent with higher affinity for inactivated channels. The phospholipase activity of the complex toxin can be separated by affinity-chromatography using a phospholipid analog (PC-Sepharose). The resulting complex contains only alpha-neurotoxin and protease inhibitor and is still capable of blocking calcium channels, although with less potency than the native oligomeric form. Sephadex G-50 gel filtration chromatography in the presence of high salt (1M NaCl) at alkaline pH (8.2), separates the alpha-neurotoxin-like peptide from the protease inhibitor, but at this stage the resulting peptides lose physiological activity towards the calcium channels. The amino acid sequence of the protease inhibitor was determined by automatic Edman degradation. The alpha-neurotoxin-like peptide and two isosubunits displaying phospholipase activity were sequenced at the N-terminal part of the molecule.

Hormonal regulation of pituitary GH3 cell K+ channels by Gk is mediated by its α-subunit

The resolved α‐GTPγS (α*) and βγ‐subunits of human erythrocyte Gk, the stimulatory regulatory component of hormone‐responsive K+ channels, were tested for their potential stimulatory activities on the K+ channel of the endocrine GH3 cell. Concentrations as low as 0.5 pM αk* consistently activated K+ channels in isolated membrane patches, and saturating effects were obtained with 50 pM αk*. In contrast 2000–4000 pM βγ was without effect. We conclude that Gk acts on K+ channels through its α‐subunit in a manner akin to that of Gs acting on adenylyl cyclase and transducin acting on cGMP‐specific phosphodiesterase of photo‐receptor cells.

Disruption of Rho signaling results in progressive atrioventricular conduction defects while ventricular function remains preserved

Recent studies suggest that RhoA and Rac1 mediate hypertrophic signals in cardiac myocyte hypertrophy. However, effects on cardiac function caused by inhibition of their activity in the heart have yet to be evaluated. Cardiac‐specific inhibition of Rho family protein activities was achieved by expressing Rho GDIα, an endogenous specific GDP dissociation inhibitor for Rho family proteins, using the α‐myosin heavy‐chain promoter. Increased expression of Rho GDIα led to atrial arrhythmias and mild ventricular hypertrophy in adult mice (4–7 months). However, left ventricular systolic and diastolic function was largely preserved before and after the development of cardiac hypertrophy, indicating that Rho GTPases are not required to maintain ventricular contractile function under basal physiological condition. Electrocardiography and intracardiac electrophysiological studies revealed first‐degree atrioventricular (AV) block in the transgenic heart at 1 week of age, which further progressed into second‐degree AV block at 4 weeks of age before the development of cardiac hypertrophy. Expression of connexin 40 dramatically decreased from 1 week to 4 weeks of age in the transgenic heart, which may contribute in part to the conduction defects in the transgenic mice. This study provides novel evidence for an important role of Rho GTPases in regulating AV conduction.

Direct coupling of the somatostatin receptor to potassium channels by a G protein

G proteins couple receptors to ionic channels indirectly by acting on membrane enzymes which modulate channel activity through second or third messengers such as cytoplasmic kinases, IP3 or Ca++. Recently, it has been shown that G proteins can act on ionic channels in a membrane-delimited or direct manner; from our experience this phenomenon seems to be widespread. A G protein purified from human red blood cells (hRBC) Gk when preactivated with GTP gamma S acts directly on muscarinic acetylcholine receptor-regulated K+ channels (K+[ACh]) in atrial cells and the stimulatory regulator of adenylyl cyclase, Gs from hRBCs acts directly on two distinct voltage-gated Ca++ channels, one in cardiac muscle and the other in skeletal muscle T-tubules. In many cells, including clonal GH3 pituitary cells, somatostatin (SST) inhibits secretion by a complex mechanism that involves a pertussis toxin (PTX)-sensitive step. This is not due to lowering cAMP since secretion induced by cAMP analogs and K+ depolarization are also inhibited. SST also causes membrane hyperpolarization, which is similar to the effect of ACh on cardiac pacemaking cells and may lead to decreases in intracellular Ca++ needed for secretion. ACh acting through a muscarinic recpetor in GH3 cells has the same effects as SST.(ABSTRACT TRUNCATED AT 250 WORDS)

Amino acid sequence and physiological characterization of toxins from the venom of the scorpion Centruroides limpidus tecomanus Hoffmann

The complete amino acid sequence of the major toxic component (II.20.3.4), named toxin 1, from the venom of the Mexican scorpion C. l. tecomanus is reported. The sequence (66 amino acids) was obtained by direct Edman degradation of reduced and alkylated toxin, followed by sequence determination of selected peptides separated after enzymatic cleavage with S. aureus V8 protease. In cultured chick dorsal root ganglion cells, 0.5 microM toxin 1 slowed down specifically the time course of Na+ current inactivation, while Ca2+ currents from the same preparation were little affected. In neonatal rat ventricular heart cells, toxin 1, at concentrations between 0.1 and 0.5 microM, reduced Na+ currents without changing the kinetics and Ca2+ currents were unaffected. Comparative analysis of the primary structure of this toxin with other scorpion toxins shows a high degree of similarity with the north American scorpion toxins. This analysis suggests that the fine tuning of the molecular mechanism of action of these toxins is related to variations in the primary structure as well as to the type of membrane under study (tissue specificity).

Direct G‐Protein Regulation of Ca2+ Channels

When we first identified specific guanine nucleotide binding, or G proteins, that could directly gate muscarinic cholinergic atrial K+ channels,*z we wondered how widespread the phenomenon was. K+ channels in other tissues were prime candidates if they, like muscarinic atrial K+ channels, could also be activated by neurotransmitters or hormones in a cyclic adenosine 3,5-cyclic monophosphate (CAMP)-independent manner. Our expectations were confirmed when we found that the same exogenous G protein, G,, that was effective in heart mimicked the effects of somatostatin and acetylcholine on specific K+ channels in GH3 cells, a clonal anterior pituitary cell line., The GH3 K+ channels had a larger conductance than the atrial K+ channels indicating that direct gating by G proteins was not restricted to one type of K+ channel and this encouraged us to broaden the scope of our enquiry by asking whether a completely different category of channels might be involved. We turned to Caz+ channels for the following reasons: ( 1) guanosine triphosphate (GTP)-altered dihydropyridine (DHP) binding in cardiac sarcolemmal and skeletal muscle t-tubule (2) G proteins were implicated in the coupling of a variety of neurotransmitters to Caz+ channels7.*; and (3) G proteins were shown to couple opioid receptors to Caz+ channel^.^ The situation for DHP-sensitive Ca2+ channels is more complicated because they are also modulated by second messengers (FIG. l) , and this mechanism has to be excluded when the presence of direct pathways is being determined. Heart and skeletal muscle were satisfactory tissues because at least three types of pertussis toxin (PTX) s~bs t r a t e l~ and two types of cholera (CTX) substrate have been reported for cardiac sarcolemma, and two types of PTX and CTX substrates have been found in skeletal muscle t-tubules. We found that G proteins could directly gate DHP-sensitive Ca2+ channels in heart. and skeletal muscle16 and we proved, using recombinant DNA techniques, that a single G protein, G,, so named because it is the stimulatory regulator of adenylyl cyclase, could also gate Caz+ channel~. ~~

Signal transduction by G proteins.

Publisher Summary This chapter deals with the mechanism of activation/deactivation, and the actions of the heterotrimeric G proteins responsible for transducing the binding of a signaling molecule to its seven-transmembrane receptor into cellular responses. The process by which the extracellular ligand–receptor interaction leads to changes inside the cell is commonly referred to as “signal transduction” and a ligand is the extracellular signal whose message is transduced into an intracellular signal of a different chemical nature. The signals generated in this way are second messengers, such as ions entering through ligand-gated ion channels; signaling molecules derived from the receptor itself; products of the activation of the receptors intrinsic enzymatic activities; tyrosine phosphorylation, first of self and then of proteins recruited to the phosphorylated receptor, as happens when EGF, PDGF, and insulin bind to their respective homodimeric receptors; and sequential serine/threonine phosphorylation, also first of self and then of cytosolic transducing proteins such as R-SMADs by phosphorylated receptor I, as it happens in response to interaction of the TGF β and bone morphogenic protein (BMP) family of signaling molecules with their heteromeric receptors. The use of heterotrimeric G proteins as signal transducers constitutes yet a different type of signal transduction process.

Control of K+ channels by G proteins

Heterotrimeic G proteins are thought to couple receptors to ionic channels via cytoplasmic mediators such as cGMP in the case of retinal rods, cAMP in the case of olfactory cells, and the cAMP cascade in the case of cardiac myocytes. G protein-mediated second messenger effects on K+ channels are dealt with elsewhere in this series. Recently, membrane-delimited pathways have been uncovered and an hypothesis proposed in which the α subunits of G proteins directly couple receptors to ionic channels, particularly K+ channels. While direct coupling has not been proven, the membrane-delimited nature has been established for specific G proteins and their specific K+ channel effectors.

Multiple Roles of G Proteins in Coupling of Receptors to Ionic Channels and Other Effectors

The central role of G proteins in coupling receptors to effector systems can be best illustrated by the contents of Table 1, which lists receptors that exert their actions by interacting with a G protein, according to the general schemes shown in Fig. 1. Like receptors, which are increasing in number rapidly, effectors affected by the activated forms of G proteins are also increasing, most notably through the discovery in 1986–1987 that ionic channels form part of the family of molecules regulated by G proteins. Using cell-free systems such as provided by excision of membrane patches from cells and incorporation of plasma membrane vesicles into lipid bilayers, it was shown that ionic channels are indeed regulated by activated G proteins. Some of these channels had long before been predicted to be under the control of G-protein-coupled receptors by means other than soluble second messengers. The mechanism by which G proteins regulate some of these channels is at the very least “membrane delimited” and independent of any phosphorylation event or of changes in cytoplasmic levels of second messengers such as cAMP, Ca2+, or IP3, and is very likely due to direct interaction of the G protein α-subunit and the channel proper (for review, see Brown and Birnbaumer 1988). Our group was prominent in providing some of the initial as well as subsequent supporting data for these conclusions (Brown and Birnbaumer 1988; Yatani et al. 1987a,b,c, 1988a,b,c; Codina et al. 1987a,b, 1988; Imoto et al. 1988; Kirsch et al. 1988). Another research group has claimed that ionic channels, specifically the heart muscarinic K+ channel, now referred to as Gi-gated K+ channel (see below), may be regulated by receptors, not via a specific G protein α-subunit, but via the βγ-subunits of G proteins (Logothetis et al. 1987).