Perhaps the best-studied example of an ion channel that undergoes direct modulation by G proteins is the G protein-gated inwardly rectifying potassium channel (GIRK) expressed in brain, heart, and pancreatic tissues. GIRK channels comprise a subfamily of inwardly rectifying K+ channels that currently contains four distinct members (GIRK 1-4) along with several alternatively-spliced isoforms (Doupnik et al. 1995). In cardiac atrial myocytes and atrial pacemaker cells, the prototype member of this potassium channel family, termed IK(ACh), mediates the parasympathetic slowing of heart rate. The binding of acetylcholine (ACh) released by vagal nerve terminals to muscarinic type 2 (M2) receptors (Chapter 18) on the surfaces of pacemaker cells activates a PTX-sensitive G protein to mediate a rapid membrane hyperpolarization which decreases excitability (Hartzell 1988). It is now widely recognized that the opening of this inward rectifier K+ channel is triggered by direct contacts between activated G protein subunits and distinct regions of the channel protein.
A series of classical observations were important in revealing the direct nature of this interaction. Early electrophysiological studies of cardiac atrial fibers described an increase in potassium flux across the outer membrane following parasympathetic nerve stimulation (Del Castillo and Katz 1955; Hutter and Trautwein 1956, 1957). The channel responsible for this increase in K+ permeability was shown to have unique biophysical properties that differed from the background potassium conductance (Ik1 ) present in the same cells (Sakmann etal. 1983). Evidence for the involvement of G proteins in the activation of IK(ACh) was provided by several findings. Pfaffinger et al. (1985) showed that receptor-induced activation of the current in atrial myocytes was abolished by pretreatment of cells with pertussis toxin, a bacterial toxin which ADP-ribosylates and inactivates G proteins of the Gi subfamily. Moreover, intracellular perfusion with a non-hydrolyzable GTP-analogue (GTP-y-S) resulted in the irreversible and receptor-independent activation of the same potassium current (Breitwieser and Szabo 1985). Further studies by Kurachi et al. (1986) determined that IK(ACh) channel activity recorded in cell-attached patches of atrial myocytes declined upon excision of the patch into GTP-free solution, but was recovered when GTP was subsequently added back to the bath.
Although the above findings allowed only for speculation regarding the steps that link G proteins to the activation of GIRK channels, a provocative discovery by Soejima and Noma (1984) shed further light on the nature of the regulatory cascade. Their studies had found that when channels were measured in cell-attached patches on atrial cells, IK(ACh)
activity increased only when acetylcholine was present in the recording pipette, but not when applied to the bath solution surrounding the cells. These results argued against the involvement of diffusible second messengers such as cAMP or cGMP in the activation of GIRK currents, and left unanswered the question of how channel openings were coupled to receptor activation.
The first straightforward evidence implicating a direct role for G proteins in the activation of IK(ACh) was provided by Logothetis etal. (1987), who showed that cardiac GIRK channels were strongly activated after the addition of purified GPy subunits to the cytoplasmic surface of a membrane patch containing the channel. By measuring channel activity in the inside-out patch configuration (Hamill et al. 1981), this study convincingly demonstrated that IK(ACh) open-probability was increased more than 500-fold in response to GPy application, and that this rise in activity could be sustained even in the absence of native cytosolic constituents. This was surprising, given that the large body of literature had implicated Ga subunits as the functional arm in the G protein complex. A number of independent studies have since confirmed the notion that GPy is the primary activator of GIRK channels expressed within the heart and throughout the brain (Logothetis etal. 1988; Ito etal. 1992;Reuveny etal. 1994; Wickman et al. 1994). The cloning of various GIRK genes has provided a means of expressing these channels in cellular environments different from those of their native tissues in order to examine their modes of activation (illustrated in Fig. 6.1). The expression of recombinant GIRK1/GIRK4 channel complexes in various cell lines gives rise to potassium currents with single-channel properties that are indistinguishable from native cardiac IK(ACh) and which are activated by purified GPy dimers (Wickman et al. 1994; Krapivinsky et al. 1995a). GIRK channels found in the brain also show specific activation by GPy subunits when studied in a variety of heterologous expression systems (Kofuji etal. 1995; Velimirovic etal. 1996; Jelacic etal. 1999, 2000).
A number of residues on the surface of the GPy molecule appear to be important for this interaction. Thus far, site-directed mutagenesis strategies have identified distinct regions on the Gp subunit that are thought to directly interact with GIRK channels. Ford et al. (1998) found that mutations of specific residues in the distal amino terminus of Gp (blade 1;
+ 80mV j-Excised patch r50nMGp7 +80_mV r-100 nM Ga-GDP
p^THm rpPWTPipnri™» ■■! ■!»¥ ' wwyw^p^ipi
Fig. 6.1 Gpy subunits directly activate GIRK channels. In cells transiently expressing recombinant GIRK1 and GIRK4 channel subunits, inwardly-rectifying potassium currents can be readily measured in the cell-attached configuration. Patch excision into GTP-free bath solution causes a decline in channel activity, which can be strongly recovered upon addition of 50 nM purified Gpy to the inside surface of the patch. The subsequent addition of 100 nM Ga-GDP reverses the effect by forming inactive Ga^y heterotrimers. The membrane potential was held at —80 mV with brief steps to +80 mV to verify inward rectification. Reproduced from Krapivinsky etal. (1995) Nature, 374, 135-41.
Chapter 4) caused a measurable decline in GIRK current activation when the subunits were transiently expressed in Xenopus oocytes. The majority of these sites are masked by the amino-terminal helix of Ga under conditions where the two G protein subunits do not dissociate (Lambright et al. 1996). Further mutations of residues on the top surface of Gß that face the switch regions of Ga-GDP have also caused these subunits to become less effective at stimulating channel activity. Albsoul-Younes et al. (2001) purified a series of mutant ßy subunits and tested their abilities to activate neuronal GIRK channels in excised patches from rat brainstem neurons. Their results revealed additional residues exposed on the outer surfaces of the blades forming the Gß-propeller (Chapter 4) that maybe important for enhancing channel activity. Though less attention has been directed towards identifying similar regions on the Gy protein, it remains possible that this subunit serves other functions than to simply anchor the ßy dimer to the membrane (see Kawano et al. 1999). The sites on Gßy that are involved in effector regulation have been more carefully defined for adenylyl cyclase and phospholipase C ß, and it seems that there is at least a partial overlap with those regions found to be important for GIRK activation (Li etal. 1998).
The previous results strongly support the idea that the stimulation of M2 muscarinic receptors in the heart activates Ik(acIi) by causing the release of free Gßy dimers which then go on to gate the channel. Proof of a direct interaction between these two participants came from studies showing that purified Gßy binds specifically to immunoprecipitated GIRK1 and GIRK4 channel subunits (Krapivinsky et al. 1995 b). The interaction was found with both native and recombinant channels, while no detectable binding was revealed for either inactive Ga-GDP or activated Ga-GTPyS. Further protein-binding studies using affinity-purified recombinant GIRK1 and Gßy subunits have produced similar results.
Considerable effort has been directed towards identifying regions on GIRK proteins that interact with Gßy subunits. Numerous sites located within the intracellular amino- and carboxyterminal tails of both GIRK1 and GIRK4 subunits have been implicated in Gßy binding based on mutagenesis and peptide-competition results (Inanobe et al. 1995; Kunkel and Peralta 1995; Huang et al. 1997). Native cardiac IK(ACh) exists as a heterotetrameric complex formed by two GIRK1 and two GIRK4 channel subunits (Corey et al. 1998). Purified Gßy interacts specifically with fusion proteins containing the hydrophilic amino- and carboxyterminal domains of GIRK1 (Huang etal. 1995). Studies by Krapivinsky etal. (1998) showed that a C-terminal region of the GIRK4 protein close to the second transmembrane domain is critically involved in Gßy-dependent activation of the GIRK1/GIRK4 channel complex. It seems likely that the intracellular stretches of both channel subunits combine with one another to form a high-affinity binding site for Gßy inside the cell. The precise mechanisms by which the binding of G proteins to the cytoplasmic tail regions translates into the opening of the channel pore remain to be revealed.
Was this article helpful?