Mechanisms of modulation of voltage-dependent calcium channels by G proteins (original) (raw)

J Physiol. 1998 Jan 1; 506(Pt 1): 3–11.

Department of Pharmacology, University College London and Royal Free Hospital School of Medicine, Gower Street, London WC1 6BT, UK

Received 1997 Sep 15; Accepted 1997 Oct 6.

Voltage-dependent calcium channels (VDCCs) are key components in the complex functioning of excitable cells. Because of this, their regulation is of paramount importance to the control of cellular activity. VDCCs consist of a pore-forming α1 subunit, which has four domains each containing six putative transmembrane segments. From purification studies, these are associated with an intracellular β subunit. They also co-purify with an α2 subunit, which is entirely extracellular, linked into the membrane by S-S bonding to a transmembrane δ subunit (Witcher et al. 1993; Liu, De Waard, Scott, Gurnett, Lennon & Campbell, 1996) (Fig. 1). A number of different α1 subunits have been cloned; α1C, D and S all form 1,4-dihydropyridine (DHP)-sensitive L-type calcium channels, whereas α1A, B and E form P/Q-, N- and possibly R- or T-type channels, respectively. There are several means by which these channels may be modulated, but for neuronal channels, particularly N and P/Q, a major mechanism involves inhibitory modulation via the activation of heterotrimeric G proteins by seven transmembrane (7TM) receptors (for review see Dolphin, 1995). The key features of this inhibition are that it is always partial and is typified by a slowing of the current activation kinetics, which is thought to be due to a time-dependent recovery from voltage-dependent inhibition (Bean, 1989). The voltage dependence is manifested by a shift to more depolarized potentials of the current activation-voltage relationship, and the loss of inhibition at large depolarizations (Bean, 1989). There may also be additional mechanisms that are not voltage dependent, manifested by an incomplete ability of a depolarizing prepulse to reverse the inhibition (Diversé-Pierluissi & Dunlap, 1993), and the continuing presence of inhibition at large depolarizations measured from the tail current amplitude.

An external file that holds a picture, illustration, etc. Object name is tjp0506-0003-f1.jpg

The VDCC oligomeric complex

Binding sites on the α1 I-II loop for VDCC β and Gβγ are shown. All VDCC α1 subunits have a binding site for β (α interaction domain, AID) on this loop, but Gβγ binding has only been shown for α1A, α1B and α1E. Regions of the C-terminal tail of human α1E which may be involved in VDCC β and Gβγ binding are also indicated. BID, β interaction domain.

Calcium channels are present in many tissues, where they fulfil a number of different specialized roles. In neurons the distribution of the channels is non-uniform, with α1B and α1A being particularly concentrated at synaptic terminals (Westenbroek, Hell, Warner, Dubel, Snutch & Catterall, 1992; Westenbroek et al. 1995). G protein-mediated modulation of these channels has been shown to occur at presynaptic terminals (Toth, Bindokas, Bleakman, Colmers & Miller, 1993). Evidence suggests that this mechanism may be responsible for at least some of the presynaptic inhibition of synaptic transmission mediated by a wide variety of 7TM receptors in many areas of the nervous system (Man-Son-Hing, Zoran, Lukowiak & Haydon, 1989; Hille, 1992; Toth et al. 1993; Dolphin, 1995). Activation of such receptors will reduce calcium entry into presynaptic terminals via VDCCs, but the effect should be frequency dependent. Inhibition will be reduced during a high frequency train as a result of the voltage dependence of the inhibitory modulation, providing a gain-setting mechanism. Relief of inhibition of calcium currents, evoked by action potential-like voltage waveforms, has been reported during high frequency trains (Williams, Serafin, Mühlethaler & Bernheim, 1997), and might contribute to the modulation of presynaptic inhibition depending on input frequency. The search for the molecular mechanism of this modulation is hotting up. Nevertheless, it should not be forgotten that indirect mechanisms such as presynaptic hyperpolarization, or direct mechanisms involving inhibition of exocytosis, are also likely to play a role in the modulation of presynaptic release of transmitter (Man-Son-Hing et al. 1989).

Direct coupling between G proteins and VDCCs

Several lines of evidence have suggested that the link between G protein activation and N- or P/Q-type calcium channel inhibition is a direct one, rather than acting through a downstream soluble intracellular messenger (reviewed in Hille, 1992). Of particular importance was the finding that inhibition of calcium channels recorded in the cell-attached patch mode only occurs when the receptor agonist is present in the patch pipette, and not when it bathes the remainder of the cell membrane (Forscher, Oxford & Schulz, 1986). Several groups began to study whether there was any direct link between activated G proteins and one of the subunits of the calcium channel complex. My group was interested in how the activated G protein species produced its inhibitory effect on VDCCs, and made an educated guess that the intracellular VDCC β subunit might be involved. In order to investigate this, we developed an antisense strategy to deplete dorsal root ganglion neurons of their VDCC β subunits by microinjection of an antisense oligonucleotide complementary to a region common to all β subunits (Berrow, Campbell, Fitzgerald, Brickley & Dolphin, 1995). We found that loss of the VDCC β subunit resulted in smaller calcium channel currents with altered kinetics, as would be expected for depletion of this accessory subunit (Berrow et al. 1995). More interestingly, it also resulted in an enhancement of the effect of the GABAB receptor agonist (-)-baclofen in inhibiting the residual currents (Campbell, Berrow, Fitzgerald, Brickley & Dolphin, 1995_b_). We hypothesized from these results that there was competition between the activated G protein and the VDCC β subunit for binding to a site on the channel (Campbell et al. 1995_b_; Dolphin, 1995). The VDCC β subunit has been shown clearly to bind to a site (called the α interaction domain or AID) on the intracellular loop between transmembrane domains I and II of all VDCC α1 subunits (De Waard, Pragnell & Campbell, 1994). This led to the inescapable possibility that the activated G protein moiety might also bind to this region. The hypothesis that there was competition for an overlapping binding site was supported by evidence that overexpression of VDCC β in oocytes blocked agonist-mediated inhibition (Roche, Anantharam & Treistman, 1995), whereas expression of VDCCs in the absence of the VDCC β subunit led to an enhancement of receptor-mediated modulation (Bourinet, Soong, Stea & Snutch, 1996).

Which G protein subunit is responsible for coupling?

The modulation of neuronal VDCCs in most systems is mediated by receptors coupled to pertussis toxin-sensitive G proteins, and many groups performed experiments with blocking antibodies and antisense oligonucleotides complementary to G protein α subunits which showed that Gαo was primarily responsible for the response (McFadzean, Mullaney, Brown & Milligan, 1989; Baertschi, Audigier, Lledo, Israel, Bockaert & Vincent, 1992; Campbell, Berrow & Dolphin, 1993). However, in several systems, both Gαi and Gαo were involved (Ewald, Pang, Sternweis & Miller, 1989; Toselli, Lang, Costa & Lux, 1989), and to complicate matters further, in a few systems, Gq- or Gs-coupled receptors produced similar modulation (Shapiro & Hille, 1993; Golard, Role & Siegelbaum, 1994; Zhu & Ikeda, 1994). This led two groups to test the hypothesis that the species involved was the moiety common to all these G proteins, Gβγ, rather than any particular Gα (Ikeda, 1996; Herlitze, Garcia, Mackie, Hille, Scheuer & Catterall, 1996). There was a clear precedent for this in the G protein-activated potassium channels (GIRKs). Although for many years a controversy reigned concerning which G protein subunit was responsible for modulation of the native GIRKs (e.g. Yatani, Codina, Brown & Birnbaumer, 1987), they were eventually shown conclusively to be activated by Gβγs (Logothetis, Kurachi, Galper, Neer & Clapham, 1987; Clapham & Neer, 1993; Krapivinsky, Krapivinsky, Wickman & Clapham, 1995). Furthermore all Gβγ species tested, except transducin Gβ1γ1, are similarly effective (Wickman & Clapham, 1995; Wickman et al. 1997; Yamada et al. 1997). From the work of both Ikeda (1996) and Herlitze et al. (1996), it became clear that transfection either of primary neurons or of cell lines with Gβ1γ2 or Gβ2γ3 led to the tonic inhibition of the calcium current, which could be reversed by a depolarizing prepulse, applied just before the test pulse, a hallmark of voltage-dependent inhibition of these channels (see example in Fig. 2). Gβγ overexpression also occluded modulation by agonist.

An external file that holds a picture, illustration, etc. Object name is tjp0506-0003-f2.jpg

Prepulse potentiation of α1B modulated by co-expressed Gβγ

The VDCC α1 subunit α1B was co-expressed with the VDCC accessory subunits α2-δ and β2a and the G protein subunits Gβ1 and Gγ2 in COS-7 cells, as described previously (Page et al. 1997). The cell was held at -100 mV, and the voltage was stepped according to the protocol shown, with two 50 ms steps P1 and P2 to varying voltages (-40, -30 and -20 mV) separated by an 80 ms step (prepulse) to a large depolarizing voltage (+120 mV). Leak subtraction was performed according to a P/8 protocol. The amplitude of the currents in P2 after the depolarizing prepulse are clearly larger than in P1 before the prepulse and their rate of activation is faster at all potentials (from data supplied by Dr G. J. Stephens).

The possible reason that many workers (Ewald et al. 1989; McFadzean et al. 1989; Degtiar, Wittig, Schultz & Kalkbrenner, 1996), including my own group (Campbell et al. 1993; Menon-Johansson, Berrow & Dolphin, 1993), had observed the importance of Gαo was the examination of receptor-mediated inhibition in neurons, where Go is the major G protein. In the absence of the Gαo subunit, coupling will be markedly attenuated, as it depends on the G protein heterotrimer. However, several groups directly investigating the involvement of Gβγ in calcium channel modulation previously, had not found any such effect during its infusion (Hescheler, Rosenthal, Trautwein & Schultz, 1987). The reason for this is unclear, but this result highlights one difficulty of handling these hydrophobic subunits, which, except for Gβ1γ1, are generally only soluble in detergent. There are extensive results from the work of Schultz and colleagues, concerning the specificity of different Gβγ subunits for signalling pathways between different receptors and calcium current modulation (Kleuss, Hescheler, Ewel, Rosenthal, Schultz & Wittig, 1991; Degtiar et al. 1996; Kalkbrenner, Dippel, Wittig & Schultz, 1996). These may also be reconciled with the finding that several Gβγ subunits are able to transduce the signal to calcium channels (Ikeda, 1996; Herlitze et al. 1996), by the interpretation that a specific G protein heterotrimer combination may selectively couple to a particular receptor in intact cells and the selectivity is therefore largely at the receptor-G protein interaction step, rather than at the G protein-calcium channel level. It will be possible to test this hypothesis by the synthesis and use of receptor-Gα fusion proteins (Milligan, 1997).

Is the α1 I-II loop involved in G protein modulation?

The combination of two findings: (1) that Gβγ subunits are the important mediators of inhibitory modulation, and (2) that this modulation may involve the I-II loop, since the VDCC β subunit binds there and may functionally compete with Gβγ subunits, led a number of groups to examine the intracellular I-II linker in detail. Gβγ subunits have been found previously to bind to sites on type 2 adenylyl cyclase and phospholipase Cβ2 (Chen et al. 1995) which have a characteristic central motif consisting of QXXER. Whilst this motif is not necessarily indicative of a functional Gβγ binding site, it was also found to occur in the I-II loop of α1A, B and E, intriguingly within the binding site described for the VDCC β subunit (Table 1). However, in reconstituted systems consisting of a VDCC α1/β combination and either an endogenous or an expressed G protein-coupled receptor, classical modulation could be demonstrated for α1B (the N-type channel) and to a lesser extent for α1A, but not to any clear extent for α1E (Toth, Shekter, Ma, Philipson & Miller, 1996). Therefore my group, and others, made chimeric channels between α1B, which shows the greatest G protein modulation, and those α1 subunits that showed no or less modulation, in an attempt to define the regions involved in this process (Zhang, Ellinor, Aldrich & Tsien, 1996; Page, Stephens, Berrow & Dolphin, 1997; Zamponi, Bourinet, Nelson, Nargeot & Snutch, 1997; De Waard, Liu, Walker, Scott, Gurnett & Campbell, 1997). Unfortunately, although the consenus of opinion is that the I-II intracellular linker may be important in bestowing at least some of the ability to be G protein modulated onto a particular channel, the same conclusion was not reached by all groups. However, a careful analysis of the data may indicate why this divergence of opinion has arisen.

Table 1

Calcium channel I-II loop mutations

Ca2+ channel α1 subunit QXXER motif in I-II loop Mutations
α1B (N type) QQIER R→E no effect on modulation (Zhang et al. 1996)
α1A (P/Q type) QQIER Effect on modulation: I→L, R→E reduced (Herlitze et al. 1997); R→E enhanced (Herlitze et al. 1997) or reduced (De Waard et al. 1997)
Effect on Gβγ binding: either Q→A or R→E reduced binding (De Waard et al. 1997)
α1E (R or T type) QQIER
α1C (cardiac L-type) QQLEE L→I and final E→R, did not produce modulation (Zhang et al. 1996)
α1D (neurosecretory L-type) QQLEE No published studies

Snutch and colleagues have constructed a chimera between α1B and α1A, making the construct consisting of α1A with the I-II loop of α1B (chimera AbAAA, using capitals for the four transmembrane domains and small letters for the intracellular linkers or terminals). They found that this showed greater modulation by somatostatin, although still not as high as α1B itself (Zamponi et al. 1997). They also showed that inclusion of Gβγ subunits in the patch pipette mimicked the effect of somatostatin, although it is surprising that no clear slowing of the activation phase of the current was observed in the presence of Gβγ. Furthermore two peptides, one containing the region of the I-II loop including the QQIER sequence of α1B and one with a sequence located more towards the C-terminal on the same loop, blocked the effect of Gβγ, indicating that two, possibly interacting, sites on this loop are involved in G protein modulation. The suggestion is that these peptides act by competing for Gβγ binding, or by sterically hindering its effect; however, these experiments do not prove that the effect of Gβγ is on the I-II loop of the α1 subunit. These authors also showed that phosphorylation of one of the peptides prevented its effect, presumably by reducing Gβγ binding, as might be expected for an increase in charge in the binding domain of a hydrophobic protein complex. This may provide an explanation as to why activation of protein kinase C (PKC) antagonizes receptor-mediated calcium current inhibition (Stea, Soong & Snutch, 1995). However, it must be said that not all groups support such an antagonistic role for PKC, and there are several reports that it may mediate aspects of calcium current inhibition (e.g. Diversé-Pierluissi, Goldsmith & Dunlap, 1995).

Two groups have now shown that the I-II loops of VDCCs bind Gβγ (Zamponi et al. 1997; De Waard et al. 1997), and the residues critical for that binding have been mapped (De Waard et al. 1997). In agreement with electrophysiological data, the AID part of the loop (the region involved in binding VDCC β) is one important domain, and this contains the QQIER sequence, some residues of which have been shown to be essential for Gβγ binding (De Waard et al. 1997) (Table 1). Moreover, Campbell's group also showed that the part of the I-II loop C-terminal to AID could also bind Gβγ on its own (De Waard et al. 1997), although the residues involved have not yet been mapped. This result correlates well with the peptide blocking data (Zamponi et al. 1997).

We have found, by making chimeras between α1A or α1B and rat α1E, that we can confer aspects of G protein-mediated inhibition onto α1E, which is not significantly G protein-modulated in our system, or in several others that have been tested (Bourinet et al. 1996; Toth et al. 1996). The two parameters that we have measured are kinetic slowing and inhibition of the calcium channel current amplitude. These may both be aspects of the same mechanism of G protein modulation. Kinetic slowing is due to the time- and voltage-dependent partial dissociation of Gβγ that occurs during the test pulse to activate inward calcium channel currents, whereas the inhibition of current amplitude, which is reversible by a very large depolarizing prepulse, may be due to inhibitory modulation by the Gβγ which remains bound during the small depolarization represented by the test pulse, but which is driven off by depolarization to a large positive potential. We examined modulation by GTPγS, following transient transfection of the relevant VDCC α1 subunits, together with β1b and α2-δ, in COS-7 cells. In this study, we substituted the I-II loop and part of the Is6 transmembrane segment from α1B into the α1E backbone, making an α1EbEEE chimera. This replacement conferred on α1E the ability to exhibit some kinetic slowing in the presence of GTPγS, whereas α1E itself shows very little modulation. However, for currents resulting from α1B, GTPγS also produces a large inhibition of the maximum amplitude of the current, and this can be reversed by a depolarizing prepulse (Fig. 2). This property is not conferred on the chimera. It is possible that the Gβγ‘unbinding’ from the α1EbEEE chimera shows less voltage dependence than for α1B, so that all (rather than only some) of the Gβγ is removed slowly by the test pulse. Nevertheless it also remains a possibility that modulation of current amplitude represents a different mechanism. Our findings therefore support a role for the Is6 and/or the I-II linker region in G protein modulation, but suggest that this is not the major region of the α1B channel that is important in conferring this property, because no amplitude modulation was observed for α1 EbEEE.

We have now repeated these experiments using the β subunit β2a, which dramatically slows entry into the inactivated state, in order to remove possible confounding effects of voltage-dependent inactivation on accurate measurement of the current amplitude. We have observed similar results under these conditions with co-expression of Gβγ, again indicating that the α1B Is6, I-II loop region mediates some Gβγ-induced kinetic slowing of calcium currents, but no amplitude modulation. We have now further shown that the complete domain I of α1B is the most important region for conferring all the G protein modulatory properties of α1B on α1E (G. J. Stephens, C. Canti & A. C. Dolphin, in preparation).

Catterall and colleagues provide additional data that the I-II loop is important in G protein modulation, firstly by using peptides (Herlitze, Hockerman, Scheuer & Catterall, 1997), in a similar approach to Zamponi et al. (1997). Peptides alone do not prove that the α1 I-II loop is the site of modulation, but rather indicate whether the peptides bind to Gβγ and can therefore effectively compete for this mediator. However, in a second approach this group has mutated the QQIER sequence in α1A to that in α1C, which is QQLEE (Table 1). This produced a reduction of modulation by GTPγS, using a HEK 293 cell expression system (Herlitze et al. 1997). They have also shown that mutation only of one amino acid in the QQIER sequence to QQIEE increased, rather than decreased, modulation by GTPγS, indicating that isoleucine 381 rather than the arginine at position 383 is an important amino acid for transmitting the inhibitory modulation by Gβγ. One inconsistency, compared with the results of De Waard et al. (1997), is that the latter group found that the R→E mutation in the QQIER motif reduced α1A modulation by GTPγS and in parallel, reduced Gβγ binding in vitro (Table 1).

In contrast, the results of Zhang et al. (1996), who have made an extensive range of chimeras between α1A, α1C and α1B, indicate that the I-II linker has no role in conferring G protein modulatory properties to the channel. Neither the I-II loop from α1A nor that from α1C reduced modulation in an α1B backbone. The expression system used was oocytes, and the VDCC α1 subunits were co-expressed with α2-δ and β1b, as well as the somatostatin I receptor and G protein βγ subunits. The overexpression of G protein βγ subunits in these experiments may be of relevance, since in other expression systems Gβγ produces constitutive modulation, and occludes the effect of agonists (Ikeda, 1996; Herlitze et al. 1996). Whatever the explanation for the divergence of results concerning the I-II loop, Zhang et al. (1996) have also made an important advance in this study, by identifying a number of other regions of importance for G protein modulation: including domain I and the C-terminal tail of α1B. Nevertheless, it is also of interest that the chimera consisting of the first domain only of A in the B backbone (ABBBb) showed greater modulation even than α1B. Furthermore, a chimera with the first domain and tail of α1C inserted into B (CBBBc) still showed some modulation and still contained the I-II loop of α1B (Zhang et al. 1996), whereas α1C itself showed no modulation, so the situation is clearly complex.

Using the well-established prepulse protocol, Zhang et al. (1996) have also made some observations on the kinetics of G protein modulation of the channels. It is commonly accepted, although by no means proven, that the relief of G protein-mediated inhibition is a result of rapid dissociation of activated G protein (presumably Gβγ) from the channel at depolarized potentials. This process is thought to be strongly voltage and time dependent, and, as suggested above, is also believed to cause the slow relaxation observed in response to a test pulse (Jones & Elmslie, 1997). Furthermore, re-establishment of inhibition resulting from a period at the holding potential is likely to result from rebinding of Gβγ. Whether these processes actually result in physical dissociation and reassociation between the G protein subunits and channel remains to be established. However, consistent with this view is the finding that reblock is dependent on the concentration of activated G protein (Elmslie & Jones, 1994; G. J. Stephens, N. L. Brice, N. S. Berrow & A. C. Dolphin, in preparation). Along these lines, another observation made by Zhang et al. (1996) was that the rate of reblock, as revealed by a depolarizing prepulse protocol with a variable duration at the holding potential between the prepulse and test pulse (Fig. 2), was more rapid for α1A than for α1B. Furthermore, the rate of relief of block, as revealed by varying the duration of the depolarizing prepulse, was also more rapid for α1A than α1B. It is suggested by this group that the main difference between the two VDCCs is that the dissociation rate of activated G protein is greater from α1A than from α1B. This could also account for the smaller degree of steady-state inhibition of α1A, rather than there being an intrinsic difference between the efficacy of the inhibition resulting from interaction between the activated G protein and the two channel subtypes. This hypothesis remains to be tested quantitatively in direct binding studies, and it has to be stated that a very wide variety of rates for reblock and relief of block have been observed in different cell types (Dolphin, 1996). However, the hypothesis has recently been examined on native, pharmacologically dissected, current components in bovine chromaffin cells, and the conclusion drawn was that the rates of dissociation and re-association of activated G protein, measured by the prepulse protocol, were not different between N-type (α1B) and P/Q-type (putative α1A) currents, using either ATP as an agonist or GTPγS to activate the G protein pool directly. These results would therefore lead to the opposite conclusion from Zhang et al. (1996), that some intrinsic property of the channels must account for the greater steady-state inhibition of N-, compared with P/Q-type calcium channels (Currie & Fox, 1997).

Does α1E have the capacity to be G protein modulated?

At this point it should be emphasized that although α1E is either not (or only slightly) G protein modulated in most studies (Toth et al. 1996; Page et al. 1997), its I-II loop contains a QQIER sequence (Table 1) and the I-II loop isolated as a fusion protein does bind to Gβγ_in vitro_ (De Waard et al. 1997). This has been taken by some as evidence that the I-II loop is not important for G protein modulation (Zhang et al. 1996). However, there are now two reports that, following expression in Xenopus oocytes without exogenous VDCC β subunits, human α1E does show a small degree of G protein modulation, which is lost on co-expression of a β2a subunit and reduced by other β subunits (Mehrke, Pereverzev, Grabsch, Hescheler & Schneider, 1997; Qin, Platano, Olcese, Stefani & Birnbaumer, 1997). There are two C-terminal isoforms of human α1E, one of which has a 129 bp deletion; the published mouse, rat and rabbit sequences are all similar to the C-terminal deleted isoform. This has led to some confusion in the literature, and might have been the reason why some groups found G protein modulation (Mehrke et al. 1997; Qin et al. 1997), whereas others did not (Toth et al. 1996; Page et al. 1997). However, Schneider and colleagues (Mehrke et al. 1997) observed somatostatin-induced inhibition of both splice variants of human α1E in HEK 293 cells, although only in the absence of a β subunit. This supports the notion that the VDCC β subunit inhibits receptor-mediated modulation (Campbell et al. 1995_b_), possibly by competing with Gβγ for an overlapping binding site (see below). It is possible that the reason that inhibition of α1E is either not seen at all or is much reduced in the presence of a β subunit is because binding of VDCC β to α1E may be particularly strong, although there is, as yet, no direct evidence for this. One confounding result which must temper this conclusion is that Xenopus oocytes contain an endogenous β subunit, which is thought to be present in sufficient amount to allow surface expression of the α1E subunit alone in this system, but insufficient to produce all its characteristic biophysical effects (Tareilus et al. 1997). It remains unclear whether the reason for the limited extent of modulation of α1E in the absence of exogenous β (Qin et al. 1997) was because of the presence of the endogenous β, or whether the ability of the α1E subunit to translate Gβγ binding into an effect on channel gating is intrinsically smaller than for α1B.

Birnbaumer and colleagues (Qin et al. 1997) have further challenged the hypothesis that the I-II loop of human α1E is involved in G protein modulation, by making a chimera of human α1E containing the I-II loop of α1C. This remained G protein modulated (when expressed in the absence of β subunits), whereas a chimera containing part of the C-terminal tail of α1C was not modulated (Qin et al. 1997). This group therefore concluded that the C-terminal rather than the I-II linker is involved in G protein modulation in α1E, and further showed that there is a binding site for Gβγ on this tail (distal to the deletion described above), coinciding with the presence of an additional binding site for the VDCC β subunit that the group has previously reported in the last 277 amino acids of α1E (Tareilus et al. 1997). Interestingly, the human, but not the rat clone contains a QXXER motif in the C-terminal tail, but this is situated proximal to the Gβγ binding site identified here (Qin et al. 1997).

Molecular mechanism of G protein-mediated inhibition

There are few clues as to how Gβγ binding to the VDCC α1 (on the I-II loop or elsewhere) results in kinetic slowing and steady-state inhibition of the current. Modelling of the whole-cell current modulation by agonists had suggested either that the G protein-bound channels opened with different gating properties (Kasai & Aosaki, 1989; Delcour & Tsien, 1993), or that there was in addition a dissociation of bound G protein from the channel (Bean, 1989; Elmslie, Zhou & Jones, 1990). The simplest model would involve opening only of the free and not the G protein-bound channel (Dolphin, 1991). Models also suggested that more than one activated G protein may be bound per channel in a co-operative manner (Boland & Bean, 1993). Early work identified a change in modal gating of native single calcium channels in sympathetic neurons in the presence of noradrenaline (Delcour & Tsien, 1993; Delcour, Lipscombe & Tsien, 1993). The channels involved were identified as N-type. However, recent evidence (Elmslie, 1997) now suggests that the channel examined in those studies was possibly not N-type, but might have been E-type. This conclusion stems from the fact that its activation was observed at voltages that were too negative in the 100 mM Ba2+ used to study single channels, to correspond to the N-type whole-cell current component in 2 mM Ba2+, taking into account the shift in voltage dependence resulting from additional surface charge screening in the high Ba2+ (Elmslie, 1997). Patil, De Leon, Reed, Dubel, Snutch & Yue (1996) have now shown that the only clear effect at the single channel level of muscarinic modulation of cloned α1B channels expressed in HEK 293 cells together with muscarinic m2 receptors, is a prolonged latency to first opening in the presence of a muscarinic agonist. Once a channel has opened, they observed no effect on subsequent open probability or gating pattern. This result suggests that the delay to first opening is due to dissociation of the activated G protein species from the channel, allowing it to open, and that the G protein-bound channel does not open, even with large depolarizations. This indicates either that the Gβγ binding is itself strongly voltage dependent, or that Gβγ binds to a site on the channel that produces a voltage-dependent block.

A further piece of evidence concerning mechanism comes from the same group (Jones, Patil, Snutch & Yue, 1997). The effect of GTPγS has been examined on gating currents generated by α1B channels. GTPγS both reduced gating current amplitude, and introduced a slow component of ‘on’ gating charge. It also reduced the tightness of coupling on the voltage axis between charge movement and channel opening. Therefore, activated G protein (presumably Gβγ, although this has not been demonstrated directly) not only slows the movement of charge in the voltage sensor, in response to a given depolarization, but also reduces the ability of the charge once moved to result in channel opening.

Termination of G protein-mediated inhibition

The rates of onset and offset of agonist-mediated inhibition of calcium currents have now been compared accurately with the dissociation and reassociation rates of the activated G protein-channel complex, obtained by the prepulse protocol (Zhou, Shapiro & Hille, 1997). As expected, the onset of agonist-mediated inhibition is much slower than reinhibition following a prepulse (time constant, τ= 0.7 s compared with 0.2 s, using 10 μM noradrenaline). This difference is accounted for by the time taken for agonist binding to the receptor and for G protein activation. The offset of the agonist-mediated response (τ=∼6 s for 10 μM noradrenaline) is also slower than dissociation measured during the depolarizing prepulse. This difference may be accounted for by the slower dissociation of Gβγ from the calcium channel at polarized rather than depolarized potentials (the basis for the voltage dependence of inhibition), and the slow decay of the free Gβγ concentration. This will depend on lateral diffusion in the membrane and reassociation of Gβγ and Gα-GDP, which will in turn be dependent on the rate of hydrolysis of activated Gα-GTP to Gα-GDP. The intrinsic hydrolysis rates of most heterotrimeric G proteins are much slower than the measured off rate of this response. For example, the rate of catalysis of GTP for purified Gαo is 0.03 s−1 (Higashijima, Ferguson, Smigel & Gilman, 1987). It is therefore possible that either the calcium channel or an associated protein acts as a GTPase activator for the Gα-GTP, as has been suggested for other effectors (Arshavsky & Bownds, 1992; Berstein, Blank, Jhon, Exton, Rhee & Ross, 1992). The ubiquitous RGS proteins that stimulate the GTPase activity of heterotrimeric G proteins (Watson, Linder, Druey, Kehrl & Blumer, 1996; Dohlman & Thorner, 1997) may fulfil this role, but might require targeting to the signal transducing complex. Of interest in this regard, we have shown that the GTPase activity of Gαo in neuronal membranes is blocked by an antibody to the VDCC β subunit (Campbell, Berrow, Brickley, Page, Wade & Dolphin, 1995_a_). It remains to be determined whether a physical interaction between Gαo and VDCC α1 or other subunits can be demonstrated.

Conclusion

A number of experiments indicate that the calcium channel α1 I-II loop is involved in the modulation of α1A and α1B calcium channels by G protein βγ subunits. However, several pieces of evidence suggest that this is not the main site involved, and is certainly not sufficient for full modulation (Zhang et al. 1996; Page et al. 1997). There are now several pointers to the role of the C-terminal tail, particularly in the small degree of modulation shown by α1E (Mehrke et al. 1997; Qin et al. 1997). Goals for the future include determination of other sites of G protein interaction, and elucidation of the molecular mechanism of modulation by Gβγ, since there is still little understanding of the way in which G protein binding is converted into an effect on latency of channel opening (Patil et al. 1996). It will also be of interest to evaluate whether the G protein α subunit plays a role in terminating the signal transduction process, as may be the case for GIRKs (Schreibmayer et al. 1996), and to examine whether indeed there is competition between VDCC β and G protein βγ subunits for functional binding to calcium channels.

Acknowledgments

I would like to thank Drs K. Page and N. Berrow for their critical comments on the manuscript and for their role in introducing molecular techniques into my laboratory. I would also like to acknowledge the invaluable contribution of all members of my group, past and present to our own work that is described here. The work was supported by The Wellcome Trust.

References

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society