British Journal of Anaesthesia 89 (1): 62±78 (2002)
Anaesthetic actions on other targets:
protein kinase C and guanine nucleotide-binding proteins
M. J. Rebecchi* and S. N. Pentyala
Departments of Anesthesiology and Physiology & Biophysics, School of Medicine,
State University of New York, Stony Brook, NY 11794-8480, USA
*Corresponding author
Br J Anaesth 2002; 89: 62±78
Keywords: anaesthetics, molecular targets; enzymes, protein kinase C; proteins, G-proteins
Introduction
Current research into the mechanism of anaesthesia has
been focused on a few ligand-gated ion channels that are
sensitive and appear to be relevant targets for general
anaesthetic agents.138 This is, however, no reason to rule out
the hundreds of other proteins likely to bind these drugs.35 It
is reasonable to suppose that among these, more than a
handful are subject to anaesthetic perturbation, and a few
could be relevant to the anaesthetic state. There is no reason
a priori to believe that metabotropic receptors, which
modulate synaptic transmission and bind the same ligands,
should not be as sensitive or important to anaesthetic action.
Similarly the downstream elements (guanine nucleotide
(G)-proteins) that couple these receptors to various effector
pathways (cAMP or inositol trisphosphate [InsP3]/Ca2+) or
the feedback and response elements (protein kinase C
[PKC]) may well be sensitive and relevant, yet few were
examined until recently. One of the major reasons that
attention has moved away from these second messenger
generating pathways is the perception that they and their
components are insensitive to anaesthetic agents. While this
is undoubtedly true for some, this does not eliminate others
± many ion channels are insensitive to general anaesthetics.
Measurements of second messengers and downstream
responses also do not necessarily eliminate upstream
components, since their sensitivity will depend on the
levels of receptor reserve and feedback. Here the bias
towards ion channels is re¯ected in the nature of the
experimental measures of anaesthetic action.
Sensitivity to any perturbation depends on what steps are
measured and the concurrent feedback. Ligand-gated and
other ion channels behave as self-amplifying units. Opening
a channel produces a measurable change in current under
the appropriate conditions. Although other channels may
respond rapidly, and could thereby dampen or reinforce
voltage changes, when measured under voltage clamp
conditions (the typical experiment), feedback (generally)
requires some mechanism intrinsic to the channel itself;
hence, at this level of measurement, one examines directly
the effects on the ion channel itself. Concepts such as
receptor reserve and feedback are not meaningful here.
By contrast, G-protein coupled signals rely on many
components operating in series. In tracing the cascade of
events towards the response elements, the further away one
gets from agonist binding, the more likely that some step has
already been saturated with respect to the preceding signal.
Such systems often demonstrate signi®cant receptor reserve,
which must be depleted before a perturbant, such as an
anaesthetic, will have a measurable effect. If one measures
the most proximal events, for example receptor/G-protein
coupling, and works forwards to the response elements,
performing measurements well below the reserve threshold,
it should be possible to rule out certain pathways as being
sensitive to anaesthetics. For PKC-modulated, G-protein
coupled pathways this has rarely been done. Moreover, in
systems where feedback is strong, changes in signal output
caused by anaesthetics are likely to be suppressed.
Therefore, the ability to measure responsiveness to anaesthetic depends very much on the system under study, its
receptor reserve and the strength of its feedback elements,
all of which will vary among different synapses. This
problem is compounded when examining an ensemble of
neurones, for several reasons. The most critical is related to
measurement of second messengers (other than calcium
ions), which normally entails sampling of large tissue
volumes containing thousands of neurones. Even with
receptor subtype speci®c agonists, spatial and time averaging of the signals across the many synapses present in this
Ó The Board of Management and Trustees of the British Journal of Anaesthesia 2002
Protein kinase C and guanine nucleotide-binding proteins
Fig 1 PKC isoforms.
Fig 2 Metabotropic receptor signalling: this scheme represents the regulatory mechanisms of two key signalling molecules, G-protein and protein
kinase C (PKC). Both interact with heptahelical transmembrane receptors directly. G-protein transduces the signal from the receptor to downstream
channels (Na+, Ca2+, K+) and effectors (phosholipase C [PLC], phospholipase A [PLA], phospholipase D [PLD], adenylyl cyclase [AC]) and the
feedback regulation is under the control of regulators of G-protein subunits (RGS) proteins and PKC, which negatively modulates the receptor. PKC
interacts with diacylglycerol (DAG), calcium (Ca2+) and phosphatidyl serine (PS) and also transduces the signal downstream to effectors and channels.
volume may well preclude observing any anaesthetic
effects.
In this article we review the results obtained for two
alternative anaesthetic targets ± PKC and G-proteins ± that
are integral to metabotropic receptor signalling. A signi®cant body of evidence has accumulated suggesting that
certain forms of these proteins are sensitive to general
anaesthetic agents in concentration ranges relevant to
anaesthetic action.
messengers. Their critical role in the release and response to
neurotransmitters, and their own regulation, suggest a
potential site for the action of general anaesthetic agents.
Nonetheless, the sensitivities of only a few protein kinases
have been examined. The best studied is PKC, a group of
serine/threonine kinases activated by lipid and calcium
signals generated by receptor-stimulated phospholipase C
(PLC).50
The PKC family targets a wide range of substrates,
including ligand-,129 G-protein-29 and voltage-gated ion
channels,134 and metabotropic receptors.96 Mammals express eleven different PKC subtypes, all multi-domain
proteins containing a common catalytic core, composed of
an ATP-binding C3 region and a substrate binding C4
region. The various subtypes can be broadly classi®ed on
the basis of sensitivity to calcium, and diacylglycerol
(DAG) or the xenobiotic, phorbol ester. The conventional
subtypes a, bI, bII, and g contain a C1 region consisting of
Protein kinase C
Protein kinases catalyse phosphoryl transfers from ATP to
serine, threonine, tyrosine and occasionally histidine residues of numerous regulatory proteins, many of which
control neuronal excitability. Located pre- and postsynaptically, protein kinases are allosterically controlled,
often directly driven by the levels of intracellular second
63
Rebecchi and Pentyala
two tandem C1 domains that bind DAG or phorbol esters,
and a C2 domain that binds calcium and acidic phospholipids (usually phosphatidylserine [PS]); crystallographic
structures of these domains have been published.75 The
calcium-independent (novel) forms d, eI, eII, h and q, have a
`novel' C2 domain, but are activated by DAG or phorbol
esters through the C1 region. So called atypical forms z and
l are missing the C2 region and have only a single C1
domain that does not bind lipid; thus, their activity is
independent of DAG and calcium. All PKC subtypes
contain a 17-residue pseudosubstrate sequence, located
near the amino terminus, that constitutively blocks catalytic
activity and ®gures prominently in their activation (Fig. 1).
Regulation
A general model of conventional PKC regulation involves
tethering of the otherwise soluble enzyme to the cytoplasmic aspect of the membrane surface.31 126 The initial event
involves binding of the C2 domain to calcium complexed
with acidic phospholipids. This alters domain±domain
interactions, aligning the C1 region with the membrane
surface where it binds DAG or phorbol esters (Fig. 2).
Subsequent conformational changes pull the pseudosubstrate sequence from the mouth of the catalytic core, thereby
facilitating binding and phosphorylation of a subtypespeci®c array of substrates. Control of PKC activity,
however, also involves upstream protein kinases and
phosphatases in addition to calcium and lipid activators.
Three sites, located in the catalytic core of conventional
PKCs, are constitutively phosphorylated in living cells;
autophosphorylation (by PKC itself) accounts for two, while
the third, part of the activation loop of the catalytic core, is
subject to phosphorylation by a newly discovered 3phosphoinositide-dependent protein kinase, PDK-1.130
Newly synthesized PKC is constitutively phosphorylated
by PDK-1, triggering autophosphorylation.126 In this state,
the active site of the now cytoplasmic PKC is constitutively
blocked by the pseudosubstrate sequence. Subsequent
recruitment to the membrane by DAG, PS or other acidic
lipids and calcium (conventional forms) leads to expulsion
of the pseudosubstrate peptide, and activation. Thus PKCs
are tightly controlled by coordinated, and in some cases,
opposing pathways that include PDK-1, protein phosphatases, generation of DAG from polyphosphoinositides, and
elevation of cytoplasmic calcium from both external and
internal sources. Perturbation of any one of these pathways
is predicted to alter conventional PKC activity.
Fig 3 Control of protein kinase C (PKC) activity (conventional): PKC
isozymes interact with several partners in transducing signals. Where
diacylglycerol (DAG), calcium (Ca2+), phosphatidyl serine (PS) and
phosphatase act as positive modulators of PKC, 3-phosphatidylinositoldependent (3-PI dependent) kinase exerts negative regulation. Interactions with downstream receptors for activated C kinases (RACK)
bind PKC, laterally organizing the enzymes with their substrates.
to PKC and proteins associated with these pathways,
RACKs are also capable of binding components of other
signalling pathways and could therefore serve as coincidence detectors and sites of signal integration. Other
proteins bind PKC, as well, laterally organizing kinases
with their substrates. For example, inactivation-no-afterpotential D (INAD), a scaffold protein containing multiple
postsynaptic density-95/discs large/zona occludens (PDZ)
domains, organizes eye-speci®c PKC, ¯y PLC-b isozyme
and the transient-receptor-potential calcium channel, a PKC
substrate that regulates photoreceptor sensitivity.131 How
PKC and its substrates are organized in mammalian
synapses is only now being unravelled.
Tissue and subcellular distribution
Immunoblotting (to measure protein), northern blotting and
in situ hybridization (to measure mRNA levels) and indirect
immuno¯uoresence techniques have been used to map
distribution of the various PKC subtypes in the central
nervous system (CNS). Neuronal PKC isoforms (a, bI, bII,
g, d, e, h, z, l) are differentially distributed among the major
brain and spinal cord structures89 101 and all have distinct
distributions within each area.111 Unfortunately, there are
no de®nitive rules concerning distribution among different
neurones classi®ed by transmitter, and multiple different
forms of each group are commonly expressed in the same
cell. At a subcellular level, many PKC isoforms have their
own distinct distributions, implying unique neuromodulatory roles for each subtype/class. For example, PKC-g, a
conventional form expressed primarily in the CNS, is often
located presynaptically,111 whereas the a form is often
Lateral organization in the membrane
Several PKC subtypes have protein binding partners known
as receptors for activated C kinases (RACKs) that serve to
organize the kinases, their substrates and associated regulatory proteins (Fig. 3).116 RACK1 is selective for PKCbII108 whereas RACK2 is selective for PKC-e.28 In addition
64
Protein kinase C and guanine nucleotide-binding proteins
in living cells modulate PKC isoforms associated with nonmembranous structures remains to be determined.
Activation of PKC by anaesthetics and n-alkanols has
also been reported. Comparing three different substrate/
lipid activating systems, Hemmings and Adamo52 observed
that partially puri®ed PKC from brain is stimulated by
halothane (EC50 2.2 vol%) and propofol (EC50 240 mM). In
this case the stimulatory effects were dependent on substrate
and the type of activating surface, yielding twofold stimulation with histone-H1 (substrate), and PC/PS/DAG vesicles, a more physiological activating model membrane.
Lipid dispersions that maximally activated PKC showed no
further stimulation by anaesthetic.50 52 A constitutively
active, trypsin-cleaved form of PKC was insensitive. The
stimulatory effect of halothane appeared to increase Vmax
rather than decrease Km for substrates and reduced the
amount of PS required for activation.49 Unfortunately the
use of Michaelis±Menton kinetics can lead to erroneous
conclusions when the enzyme and substrate bind to, or are
part of, a membrane surface. Nonetheless, anaesthetics
appear to activate, contradicting earlier results.122 Whether
the differences are due to assay conditions, substrate or the
source of enzyme is still unsettled.
Extending the observations of Hemmings and Adamo,
Shen and colleagues119 showed that n-alkanols stimulated
PKC-a catalysed phosphorylation of histone substrate when
PC/PS/DAG membranes were used as the activating
surface. In this case, n-alkanols like octanol increased
activity up to fourfold (EC50 ~250 mM). The EC50 was 10fold higher when DAG was omitted. Similar results were
obtained for C4 through C7, although the maximum
activation declined with decreasing chain length.
Importantly, binding to the PC/PS membrane was greatly
enhanced, suggesting that n-alkanol could substitute for
DAG. The degree of activation was dependent on the acyl
chain composition of the phospholipid. Greater activation
occurred when saturated PC and PS were combined with
unsaturated DAG. Whilst PKC activity was measured at
temperatures where the saturated lipid was ¯uid, increasing
the mol fraction of octanol or pentanol caused a lowering
and broadening of the transition temperature range.
Maximal activation by pentanol was observed under
conditions where both gel and liquid crystalline phases
coexist, implying that n-alkanols activate PKC by both
altering lipid structure and enhancing membrane binding.
Indeed, the PKC enzyme and its constitutive domains
(especially C1) are exquisitely sensitive to subtle changes in
the membrane surfaces,55 supporting a modi®ed lipid
hypothesis. One problem with this idea is the lack of
evidence for such transitions in natural membranes,
although indirect evidence for the existence of liquid
ordered phases (rafts) enriched in saturated lipids, cholesterol and sphingomyelin has been obtained.21 The possibility that these are potential sites for PKC activation, much
less modulation by anaesthetics, is an open question.
found throughout the neuronal cell body, and PKC-e is
associated with actin ®bers.54
Anaesthetics and their effects on puri®ed PKC
subtypes
The effects of n-alkanols and volatile agents on the
enzymatic activities of partly puri®ed brain PKC containing
conventional subtypes, as well as individual recombinantly
expressed isoforms, have been examined, and both inhibition and stimulation have been reported.50 Slater and
colleagues121 found that n-alkanols and volatile anaesthetics
inhibit PKC activity when protamine sulphate is used as the
activating surface and histone III-S as substrate. (Activation
by protamine sulphate appears to be different from that
induced by membrane cofactors;95 moreover, this compound is a substrate for PKC.) Under these conditions,
halothane and en¯urane were also inhibitory, albeit at high
concentrations (IC50 0.8 and 0.95 mM, respectively). The nalkanol potency followed relative hydrophobicity, which
parallels anaesthetic potency. Inhibition was also observed
using phosphatidylcholine (PC)/PS membranes containing
phorbol ester, indicating that inhibition was not due to the
absence of lipid. Although obtained using a highly arti®cial
system, these observations suggested a direct lipid-independent effect of these compounds on PKC. Allosteric
modulation of the enzyme was supported by the insensitivity of the constitutively active, trypsinized form of PKC to
n-alkanols. Comparable results were also obtained with high
anaesthetic concentrations of ethanol, halothane, thiopental,
etomidate and diethylether, with modest inhibition observed
using various substrates and PS/DAG as the activating lipid
dispersion.37
Differences between short- and long-chain n-alkanols
were observed in PC/PS membranes containing either
phorbol ester or DAG, and myelin basic protein peptide
(MBP) as substrate.123 Short-chained n-alkanols (C2±C6)
inhibit DAG-stimulated PKC-a, whereas the 12-O-tetradecanoylphorbol 13-acetate (TPA)-activated enzyme was
insensitive. By contrast, long-chain n-alkanols (C7±C10)
potentiated both DAG- and TPA-stimulated activities, and
were more effective in potentiating the latter; long-chain
alcohols also enhanced binding of ¯uorescent phorbol ester.
These results point to distinct actions of long- and shortchain n-alkanols.
F-actin binds and activates PKC, and alcohols affect this
interaction.124 125 In complex in vitro assays containing
calcium, phorbol ester, and MBP, n-alkanols from C2 to C6
inhibited stimulation by F-actin, with potencies that paralleled hydrophobicity. By contrast, n-alkanols of more than
six carbons were less potent. Ethanol (EC50 ~100 mM)
inhibited calcium/phorbol-ester-stimulated PKC-a and bI,
but had no effect on g or novel or atypical isoforms. The
effects of n-alkanols on F-actin binding and activation were
complex, and depended on the subtype and the presence and
levels of activating cofactors. Whether and how anaesthetics
65
Rebecchi and Pentyala
Halothane also enhances the effectiveness of phorbol
esters, which induce rapid translocation of conventional
PKC isoforms from the cytosol to the membrane, coincident
with activation and down-regulation through a proteolytic
pathway.92 When halothane alone is applied to synaptosomes, a small decrease in cytosolic and membrane PKC-a
immunoreactivity is observed, indicative of translocation
followed by rapid degradation.51 Phorbol 12-myristate-13acetate (PMA) alone causes a substantial translocation but
halothane enhances this effect, clearing the entire cytosolic
fraction of any remaining PKC-a. Similar results were
obtained for PKC-g. Unfortunately, the relationship between
phorbol ester concentration and anaesthetic dose was not
explored and many other questions remain. What happens
when a physiological stimulus is applied? To which
membrane compartments do the PKC isoforms localize?
Which isoform is responsible for phosphorylation of
MARCKs and other relevant substrates? It is clear that
further work is needed to explore the relationship between
anaesthetics and the events leading to PKC activation/
translocation.
Possible molecular explanations
The in vitro sensitivity of conventional PKCs suggests that
anaesthetics disturb or enhance the normal allosteric
modulation by DAG and calcium. Thus far no threedimensional structure of the whole enzyme has been
obtained, but several recent structures of the C1 and C2
domains26 suggest how anaesthetics might operate at the
protein/lipid interface. Since the C1 domain must penetrate
the bilayer to access DAG or phorbol ester,82 it would seem
that increased lateral pressures near the interface, caused by
the partitioning of anaesthetic into the interface,22 should
inhibit PKC activation by raising the energetic costs of
surface penetration, yet activation is generally reported
when arti®cial membranes are used52 49 119 and inhibition is
observed when the conventional PKCs are tested in the
absence of membranes.123±125 Alternatively, anaesthetics
may disrupt the normal allosteric transitions at domain±domain interfaces. The realignment of the C1 region,
thought to occur upon Ca:PS binding to the C2 domain,
would suggest a focus on this interface. Here, either
activation or inhibition could be proposed, depending on
which enzyme conformers are stabilized. In the former case,
little further activation should be observed at maximal
concentrations of calcium, PS or DAG. Activation of PKC
by halothane is observed at saturating levels of calcium and
DAG; however, sensitivity to halothane appears to depend
on the mol fraction of PS in the membrane surface,49
suggesting enhanced C2 domain function. Signi®cant
activation by halothane is also observed in the absence of
DAG. Similar DAG independence was noted for long-chain
n-alkanols,119 consistent with alterations in how the C1
domain interacts with the membrane surface. How anaesthetics alter the membrane and C1/C2 interface remains to
be investigated.
Oocytes
Xenopus oocytes are a convenient system to express
recombinant CNS proteins, especially metabotropic receptors coupled to PLC and ion channels. The read-out for
coupling to PLC is a calcium-activated chloride conductance, endogenous to the oocyte, that is stimulated by
calcium, mobilized from internal stores by the PLC product,
InsP3. Anaesthetic agents (iso¯urane, halothane or en¯urane) and n-alkanols reversibly suppress a calcium-stimulated chloride conductance induced by agonist activation of
M1, M3 muscarinic receptors, 5-HT2A, 5-HT1C serotoninergic and mGlu5 glutamate receptors.72 84±87 114
Importantly, the chloride conductance itself is not directly
affected by these agents. Oocytes injected with cerebrocortical mRNA express M1 muscarinic and 5HT-1C serotoninergic receptors. En¯urane (1.8 mM) applied just before a
low dose of acetylcholine or serotonin completely and
reversibly blocked the increase in the calcium-dependent
current.73 The lack of any effect on InsP3-induced calcium
release (re¯ected in calcium-activated current) pointed to
en¯urane blockade upstream of InsP3, at the level of the
receptor, G-protein or PLC. En¯urane also blocked the
GTPgS stimulated current, further suggesting that the site(s)
of inhibition involved receptor/G-protein/PLC coupling.
Similar results were obtained with halothane (0.34 mM) in
oocytes injected with M1 receptor mRNA, although the
AT1A angiotensin II receptor was resistant.34
Ethanol also reversibly inhibits M1 and 5HT-1C
receptors in oocytes,114 although the GTPgS-stimulated
current was not affected. This n-alkanol was most potent at
low agonist concentrations, increasing the agonist EC50
fourfold and decreasing the maximum response.
Suppression was blocked by a PKC inhibitor peptide or
Biological membranes and synaptosomes
Results obtained in synaptosomes (resealed membrane
vesicles pinched off from nerve terminals) are generally
consistent with activation by long-chain n-alkanols and
halothane. Halothane induced a twofold increase in PKCcatalysed phosphorylation of endogenous myristoylated
alanine-rich C kinase (MARCK), an 83 kDa PKC substrate,
as well as several other proteins (EC50 ~1.8 vol%) in whole
and lysed synaptosomes.51 A selective peptide inhibitor of
PKC blocked halothane-induced phosphorylation whereas a
protein phosphatase inhibitor failed to affect the increase,
consistent with stimulation of PKC. By contrast, basal
phosphorylation of synapsin 1, a key snare complex protein
that is under the control of multiple other kinases, was
unaffected, although halothane did enhance synapsin 1
phosphorylation induced by calcium ionophore or membrane depolarization. It should be noted that these results
were obtained in the absence of any overt physiological
stimulus.
66
Protein kinase C and guanine nucleotide-binding proteins
through G proteins, second messengers and PKC.64
Although anaesthetics can directly inhibit some high
voltage activated calcium channels, they also perturb
receptor/PKC-mediated regulatory pathways.20 M1 receptors, and phorbol esters or DAG, which activate R-type
channels through PKC, are differentially affected by volatile
agents.65 When R-type channel subunit combinations are
expressed in oocytes, pre-application of the volatile agents
halothane (0.59 mM) or iso¯urane (0.7 mM) blocks the
effect of direct PKC activators,64 consistent with inhibition
of PKC. Effects on receptor modulation, however, are more
complex and depend on agonist dose and timing. (It is also
worth noting that a relationship has been established
between PKC and the anaesthetic sensitivity of voltageregulated sodium channels from skeletal muscle and
brain.88 98 These channels are insensitive to halothane
when expressed alone in Xenopus oocytes but acquire
sensitivity when co-expressed with PKC-a.)
In summary, it is dif®cult to reconcile these results with a
simple model wherein one predominant species of PKC is
activated by anaesthetic. Whether an explanation lies in the
multiplicity of PKC isoforms in oocytes is unclear.
staurosporine (a non-speci®c kinase inhibitor), suggesting
that PKC, known to down-regulate these receptors, was
essential to the ethanol response. In addition, ethanol,
halothane, F3, an anaesthetic cyclobutane, and the related
non-immobilizer, F6, inhibit 5HT-2A receptor-induced
currents.87 n-alkanols inhibited best at low doses, with
increasing potency up to C8; however, C10±C12 n-alkanols
were comparatively ineffective. (Note that long-chain
alcohols are most effective in stimulating conventional
PKC isoforms in vitro; see above.) The PKC inhibitor
bisindolylmalemide I (BIS) increases 5HT-enhanced current by ~2.5-fold, while pretreatment with BIS prevents
nearly all the effects of halothane, F3, ethanol and octanol,
consistent with a requirement for PKC. Extending these
experiments, when types 1 and 5 mGlu receptors were
compared, it was found that ethanol (200 mM), halothane
(0.25 mM) and F3 substantially inhibited mGlu5- but not
mGlu1-induced chloride current.84 BIS prevented the
anaesthetic effects on mGlu5-induced currents, again suggesting that PKC is required for the action of these
anaesthetics. Results of two other experiments further
support this interpretation. First, pretreatment of oocytes
with calyculin A, a protein phosphatase inhibitor, prolonged
the inhibitory effect of these drugs after wash out but did not
affect control mGlu5-induced current. Secondly, S890G
substitution in mGlu5 (an important PKC phosphorylation
site) resulted in a leftwards shift in the glutamate activation
curve, and resistance to anaesthetic inhibition. These results
support the conclusion that anaesthetics somehow stimulate
PKC activity and thereby reduce mGlu5 receptor generated
signals through an enhanced negative feedback at the level
of receptor phosphorylation. However, this fails to explain
why the closely related mGlu1 receptor is anaesthetic
resistant, since it is subject to PKC down-regulation via
multiple phosphorylation sites.1
In comparable experiments, the receptor for substance P,
an important transmitter and modulator of nociception, is
similarly inhibited by halothane, iso¯urane, en¯urane, ether
and ethanol.86 These effects are also blocked by the PKC
inhibitor, BIS. Likewise, iso¯urane (1.25±2.5 vol%)
reversibly suppressed M3 receptor induced current.86
Chelerythrine (another PKC inhibitor) blocked the effects
of iso¯urane, whereas protein phosphatase inhibitors, such
as okadiac acid, prolonged iso¯urane action, similar to the
mGlu5 receptor,33 except that chelerythrine did not affect
the agonist response in the absence of anaesthetic, whereas
BIS increased responses to M1 stimulation. Interestingly,
iso¯urane is much more effective at suppressing M3 than
M1 receptor responses,33 which is curious because en¯urane,73 halothane,34 87 and F385 are quite effective at
suppressing the latter. Thus, anaesthetics do not have
equivalent effects on different receptor subtypes, implying
that stimulation of a single PKC isoform is unlikely to
account for these results.
Muscarinic and other G-protein/PLC coupled receptors
also modulate high-voltage-operated calcium channels
Experiments in whole animals
Early experiments in tadpoles suggested an important role
for protein kinases, possibly PKC, in anaesthesia.
Staurosporine, a relatively non-speci®c kinase inhibitor,
caused behavioural effects, immobility and loss of righting
re¯ex, similar to that of halothane.38 More recently the
anaesthetic sensitivity of mice lacking PKC-g or e isoforms
has been tested. (Null mice have been generated for PKC g,
b, e and t, but neurological and anaesthetic testing has been
limited to e and g isoforms.91 128) Of particular interest to
anaesthetic mechanisms, Harris and colleagues have shown
that g null mice have a reduced sensitivity to the sedative48
and anxiolytic17 effects of ethanol. The changes observed,
however, are, to a large extent, dependent on the genetic
background of the particular mouse strain,18 suggesting that
sedative-related phenotypes are produced through the
actions of multiple genes, in addition to PKC-g.
Interestingly, mice lacking PKC-g have a reduced GABAA
receptor response to ethanol,48 which normally enhances
GABAA activity. The mechanism, however, is unclear. PKC
generally down-regulates the GABAA receptor through
direct phosphorylation of the channel19 and receptor
internalization.27 36 Thus the absence of PKC-g does not
provide a simple explanation.
PKC-e null mice also have complex phenotypes. These
animals have diminished pain receptor sensitization, suggesting that PKC-e plays a role in hyperalgesic states.67
Markedly enhanced sensitivity to ethanol, benzodiazapines
and barbiturates is also observed,56 and this appears to be
related to GABAA receptor sensitivity, suggesting that
phosphorylation by this PKC subtype leads to changes in
67
Rebecchi and Pentyala
receptor changed? Could anaesthetics bound to the receptor,
PLC or G-protein alter the receptor conformation so as to
make it a better PKC substrate? These three proteins form
an activation complex wherein the agonist-occupied
receptor maintains an active heterodimer of G-protein and
PLC, both of which are subject to PKC phosphorylation. It
is possible that anaesthetics bound to this complex favour an
agonist-occupied state of the receptor, one that is more
readily phosphorylated, thereby accelerating the natural
desensitization process and requiring PKC.
Despite the wealth of anaesthetic data, some important
information is lacking, especially the phospho-states of the
receptors and identities of the PKC isoforms involved.
Identi®cation of the PKC subtypes should be possible with
isoform-speci®c peptide inhibitors117 ± these results would
help focus the attention of in vitro experiments. With
current expression technologies it should be possible to test
directly whether n-alkanols or volatile anaesthetics truly
enhance PKC-catalysed phosphorylation of the receptors in
a reconstituted system ± which has yet to be demonstrated
directly. Procedures for reconstituting muscarinic receptors
with Gq and PLC in arti®cial membranes have been worked
out.110 Using this experimental model and truncated forms
of PKC that are catalytically active but insensitive to
anaesthetics, it should be possible to test whether the
receptor or other components become better substrates for
the enzyme when anaesthetics are present. Given that
localization of PKC is key to its action, studies of whether
and how anaesthetic compounds affect subcellular distribution of speci®c PKC subtypes are also needed. This
approach has already provided some insight into how
ethanol and other short-chain n-alkanols may modulate
PKC-d and PKC-e in cultured cells.43 44
receptor function and responses to these allosteric modulators.
The contrary sensitivities of the g and e null mice suggest
that PKC-g facilitates while PKC-e diminishes the effects of
ethanol, altering neuronal responses that affect GABAergic
transmission. Whether the phenotypes of these animals
re¯ect any direct action of ethanol or other drugs on PKC is
unclear. Other complications cloud even these broad
interpretations. Unfortunately the behaviour of these mice
indicates some generalized developmental changes that
in¯uence global patterns of CNS function, consequently
affecting anaesthetic sensitivity. Similar dif®culties have
been encountered in mice lacking GABAA receptor subunits.57 To eliminate developmental differences, an inducible knock-out or knock-down (antisense RNA or interfering
protein) design would be required. Alternatively, more
subtle changes in anaesthetic-sensitive PKC subtypes could
be engineered by gene knock-ins (replacing the natural exon
with a mutated sequence) but this requires knowledge of the
anaesthetic binding site(s). Even this elegant approach may
provide equivocal results, since the actions of anaesthetics
are so closely linked to binding of the endogenous ligand.
Thus, it is unlikely that an anaesthetic-related mutation that
did not also affect binding and action of the natural activator
would be found.
Future directions
Whilst experiments in oocytes implicate PKC (or closely
related kinases) as essential for the actions of anaesthetic
compounds, the mechanistic implications are less clear. The
simple interpretation that anaesthetics activate PKC, resulting in receptor and channel phosphorylation and thereby
dampening subsequent agonist response, fails to account for
all results, especially when one considers that ethanol has
only been shown to inhibit PKC in vitro and that there are
mechanistic differences among the anaesthetics, especially
en¯urane, which also blocks GTPgS-stimulated chloride
currents (calcium mobilization).73 115 Moreover, the differential sensitivity of receptor subtypes does not correlate
with their susceptibility to down-regulation by PKC.
Whether this inconsistency is the result of differential
sensitivity and action of the various PKC subtypes or other
kinase requirements is unclear. Also uncertain is whether
some effects of anaesthetic on PKC activity involve changes
in the cellular levels of activators such as DAG and calcium
rather than a direct effect on PKC itself. (Although AT1A
and mGlu1 receptor subtypes were insensitive to anaesthetic
compounds, at least in the oocyte expression system, these
receptors contain multiple PKC phosphorylation sites,104
and they do desensitize via PKC-catalysed phosphorylation,
especially at low doses of agonist.11 39)
Do the effects of anaesthetics on metabotropic receptor
coupling re¯ect a direct or permissive role for PKC? In
altering the receptor, has the mutation simply erased a
phosphorylation site or has the allosteric behaviour of the
G-proteins
Heterotrimeric guanine nucleotide-binding proteins (Gproteins) are bound to the inner lea¯et of the plasma
membrane, where they couple metabotropic and other
receptors to an array of ion channels and second messenger
generating effector enzymes.107 Widely expressed in the
CNS, G-proteins directly and indirectly modulate neurotransmitter release from the presynaptic terminal and
excitability of the post-synaptic membrane. Thus, any
actions of anaesthetics that interfere with the function of
G-proteins should have a profound effect on neurotransmission, in addition to processes outside direct CNS control.
Recent evidence supports the view that early steps in
metabotropic signalling are sensitive to haloalkanes, ethers
and n-alkanols. Whether the molecular targets include Gproteins, receptors, or the protein kinases that modulate their
coupling to receptor and effector proteins, is presently
unclear.
G-proteins consist of a guanine nucleotide binding a
subunit and a bg heterodimer. Ga subunits are comprised of
an a-helical and a ras-like domain.127 Nucleotide occupies
68
Protein kinase C and guanine nucleotide-binding proteins
dimer diffuse in the plane of the membrane, engaging
downstream signalling components. Two modes of desensitization are normally triggered by activation: one involves
receptor/G-protein phosphorylation by various protein
kinases, including PKC, and the other involves regulators
of G-protein signalling (RGS) proteins30 that down-regulate
a cognate set of G-protein subtypes (Fig. 4). By stimulating
GTP hydrolysis, RGS proteins return the a subunit to an
inactive GDP state from which it can retrieve the released bg
heterodimer. In many cases the effector enzyme itself serves
an RGS-like function, such as PLC.106
Combinatorial regulation
There are 23 different a subunits, divided into four major
classes: as (olf), ai (o, t, g, z), aq (11, 14, 15)and a12 (13),
each linked to a speci®c set of downstream effectors. In
addition there are six different b and twelve unique g
subunits. G-protein heterotrimers are generally classi®ed
according to which a subunit they contain (Table 1).
Particular combinations of a, b and g appear to dictate the
speci®city for coupling each receptor subtype to a characteristic set of signalling pathways.137
Most a, b and g subunit combinations are possible, giving
rise to more than 1600 possible combinations. When
coupled to more than 2000 different receptor genes encoded
in mammals,78 and considering that some receptors can
interact with more than one G-protein subtype,41 the number
of potential combinations becomes truly enormous. Adding
further layers of combinatorial control and complexity,
these receptors form homo- and heterodimers,41 suggesting
mechanisms for modulating signal output, coincidence
detection and signal integration. Finally, a subunits are
subject to phosphorylation, giving rise to functionally
distinct phospho-states of each G-protein.24
Fig 4 Halothane promotes the interaction of a with bg subunits: Gbg
subunits were incorporated into arti®cial phosphatidylcholine membranes
and unbound protein was separated from the bound form by gel ®ltration.
The membranes incorporated with Gbg were transferred to assay tubes
containing non-myristoylated, coumarin-labelled Gai and the interaction
assay performed with and without halothane. The samples were
incubated for 30 min at 30°C and the complex was centrifuged at 100
000 g for 45 min. Gai, bound to bg (pelleted as membrane fraction) was
separated from free Gai (supernatant). The pellet was resuspended in the
buffer and the ¯uorescence of the sample was read at excitation and
emission wavelengths of 350 nm and 470 nm, respectively, in an ISS
spectro¯uorometer. Binding is represented as normalized ¯uorescence
values. The enhanced af®nity is independent of the guanine nucleotide
(GDP or GTP) bound to the a subunits; normally GTP-charged a
subunits have low af®nity for their bg-binding partners.
the cleft between the domains engaging ¯exible loops. The
loops, part of the so-called switch region, recognize
distinctive determinants on GDP and Mg2+-GTP, giving
rise to three conformations: GDP, Mg2+-GTP, and empty,
which are coupled to N and C terminal structures that form
the binding/activation surfaces for bg subunits, receptors,
effectors and other proteins. The N or C terminus or both are
lipid modi®ed, having subunit-speci®c combinations of
myristoyl, palmitoyl, farnesyl and geranylgeranyl groups.
Gb subunits have propeller-like projections (WD-40
motifs) that form the binding surfaces for a subunit,
activated receptor and down-stream effector proteins.86 b
and g subunits are coordinately expressed whereupon they
form an irreversible complex associated through a coiledcoil structure. Each g subunit is also lipid modi®ed, linking
the heterodimer to the membrane surface.
In the current view of G-protein activation, agonist
favours a state of the receptor that binds the G-protein
heterotrimer, catalysing exchange of GDP for GTP on the a
subunit. The GTP-charged subunit and free Gbg hetero-
Tissue distribution
G-protein tissue distribution is generally widespread, with a
few exceptions such as a15 and a16, con®ned to hematopoietic cells, agust, expressed primarily in taste organs, and
at1 and at2 (transducins), expressed in retinal rods and
cones. Speci®c enrichments of the other subtypes in
differing brain and spinal cord regions have been observed
but no clear rules governing patterns have been uncovered.9 10 61
Work on the central nervous system
Studies of the CNS support the view that volatile
anaesthetics disrupt receptor/G-protein coupling, perhaps
at the level of the G-protein itself. Halothane prevents
the shift to the low-af®nity state of the muscarinic
receptor, which is observed in the presence of the
non-hydrolysable GTP derivative, guanyl-5¢-yl imidodiphosphate (GMPPNP).5 Similarly, halothane affects
69
Rebecchi and Pentyala
oxotremorine binding in rat brainstem membranes.32
Persistence of this high-af®nity receptor state is also found
with other volatile agents like chloroform, en¯urane,
iso¯urane and diethylether.3 Volatile anaesthetics also
cause a rightwards shift in the dose of clonidine, a partial
a2-agonist, to inhibit adenylyl cyclase (AC). This same dose
of halothane antagonizes the decrease in clonidine binding
af®nity observed in the presence of GMPPNP. These results
suggest that volatile agents prevent agonist-induced dissociation of receptor from the heterotrimeric G-protein or
decrease the binding of GTP analogue or both, thereby
preserving the high-af®nity state and uncoupling metabotropic receptors from their downstream effector proteins.
Not all metabotropic receptors are affected by halothane in
membrane preparations. Serotoninergic 5HT-1A and adensoine A1 receptors were insensitive to halothane
(0.4±1.9 mM).5 81 Although a modest decrease in highaf®nity binding was observed, the change in af®nity due to
GMPPNP was unaffected. (There is a disparity between the
work of Aronstam and colleagues on the CNS and that of
Bazil and Minneman,13 who found no effect of 1.25%
halothane on binding of muscarinic agonist when a comparable experiment was performed at 37°C. They suggested
that the effects reported by Aronstam and colleagues5 were
obtained at tissue concentrations that were too high to be
relevant, because they did not correct tissue concentrations
for the lower [20°C] temperature, which was chosen
because it produced the optimal coupling between receptors
and G proteins in these isolated membranes.7 However, later
studies by Aronstam and colleagues performed at 37°C2 and
at 30°C90 con®rmed earlier observations.)
Ethanol also affects the af®nity of muscarinic receptors in
rat brainstem membranes.2 Muscarinic-stimulated GTPase
activity and GTP binding to membranes were also suppressed, consistent with an affect of this n-alkanol on
receptor/G-protein coupling.120 Ethanol at low concentration also blocks Gi/o mediated inhibition of AC activity in
rat brain cortex and in neuronal cell lines.12 23 These results
may, however, be explained by activation of Gs rather than
inhibition of the opposing Gi/o dependent pathway; indeed
most evidence supports the view that ethanol promotes
activation of AC by Gas.140 Although other n-alkanols
operate similarly, agents like chloroform and halothane do
not. In general, the volatile anaesthetics have a much less
consistent effect on Gs coupling to AC, perhaps because of
the differences in AC isoforms and their differential
distribution among and within tissues. For example,
halothane inhibits isoprenaline activation of lipolysis, but
this appears to result from suppression of hormone-sensitive
lipase downstream of cAMP production rather than an effect
on coupling of the b-adrenoceptor to AC.102
Acute ethanol treatment (50±200 mM) has been shown to
decrease the binding af®nity of agonists but not antagonists
to the d-opioid receptor expressed in neuroblastoma cells.42
The magnitude of the effect was temperature dependent: at
37°C, 200 mM ethanol caused a 20-fold decrease in
apparent af®nity and 2.5-fold increase in Bmax. This
increase in maximum binding was accounted for by ethanol
70
Protein kinase C and guanine nucleotide-binding proteins
Na+ channel conductivity in guinea-pig ventricular myocytes is substantially reduced by halothane (1.2 mM) and
iso¯urane (1 mM).136 Effects of both were reduced by
GDPbS, suggesting the involvement of one or more Gproteins, but here the effect of halothane was enhanced by
pretreatment with PTX whereas iso¯urane was resistant.
Further differences among anaesthetics were suggested by
the ®nding that the PKC inhibitor BIS enhanced the effects
of iso¯urane but not halothane. The a1-adrenergic suppression of Na+ channels was enhanced by pre-application of
iso¯urane or halothane. Co-application of GDPbS, PTX or
BIS markedly reduced the effects of agonist and anaesthetic,
although as a percentage of the agonist response the effects
of halothane were still observed, leading the authors to
conclude that the interactions between this anaesthetic and
agonist were independent of G-proteins and PKC. The
residual effects of iso¯urane on agonist inhibition were
blocked however, leading to the converse conclusion with
respect to this drug. In both cases, anaesthetic effects were
already attenuated by these broadly acting reagents; thus,
de®nitive mechanistic conclusions seem unjusti®ed. Rather,
the effects of PTX and GDPbS in modulating halothane
sensitivity suggests a complex cross-talk among G-proteincoupled pathways that somehow governs anaesthetic sensitivity. How this in¯uences a1-adrenergic regulation of the
Na+ channel is unclear.
blockade of agonist-induced receptor internalization.
Agonist-stimulated binding of GTPgS to membranes was
also suppressed (~50% by 50 mM ethanol), although the Gproteins involved were not identi®ed. By contast, n-alkanols
(C4, C6, C8) inhibit binding of agonist and antagonist to
5HT1A receptors in hippocampal membranes in the presence or absence of GTPgS. Here, ethanol (10±100 mM) had
no effect on agonist but inhibited antagonist binding. In the
presence of GTPgS, inhibition by ethanol was diminished.47
These results highlight differences among the n-alkanols
and suggest that the effects of ethanol may be linked to
receptor-speci®c G-protein interactions.
Work on myocardium
Many halogenated anaesthetics have a negative ionotropic
and chronotropic effect on the heart in isolation and in
patients.99 Paradoxically, these agents, especially halothane,
sensitize the myocardium to the effects of exogenous
catecholamines, through a direct action on the heart. For
example, halothane enhances contraction induced by
coadministration of the b-adrenoceptor agonist isoprenaline
in electrically driven human ventricular preparations.15 In
myocardial membranes, halothane increases stimulation of
AC by isoprenaline, NaF, cholera toxin and GMPPNP, but
not the stimulation produced by forskolin, which acts
directly on this enzyme.15 90 118 Treatment with pertussis
toxin (PTX) increases isoprenaline-, NaF- and GMPPNPstimulated AC activity, but halothane is unable to increase
this response further. Halothane completely suppresses
muscarinic receptor GTPase activity in rat atrial membranes, with an EC50 of 0.3 mM, further evidence implicating receptor/G-protein coupling as a target.4 These
observations imply that Gi is uncoupled from downstream
effectors by PTX and halothane, operating through analogous mechanisms. However, many of these effects are
observed in failing but not healthy myocardium,15 suggesting that the levels and activities of other signal transduction
components (receptors, kinases, RGS proteins, etc.) may be
critical for the effects of anaesthetics to rise above the
natural feedback mechanisms and relative receptor reserves.
In ventricular myocytes, muscarinic receptor activation
of Ik, a Gbg regulated ion channel, is slowed by halothane
and iso¯urane, although the magnitude of Ik conductance is
increased by the former.77 The slowing of Ik activation is
consistent with an uncoupling of receptor/G-protein from
the channel. One plausible mechanism for slowed activation
could entail the sequestration of released bg subunits by
receptor-associated Ga bound to anaesthetic. However, the
increased current, which was attributed to an increased
opening probability, is most consistent with an additional
direct effect on the channel. This work has recently been
extended in oocytes (see below).
General anaesthetics also effect Na+ channel activity ±
through a mechanism that is modulated by G-protein, albeit
at high anaesthetic concentrations. Depolarization-induced
Work in smooth muscle
Volatile agents like halothane cause relaxation of certain
smooth muscles and appear to disrupt the coupling of
muscarinic receptors to G-protein dependent pathways.63 In
permeabilized canine tracheal smooth muscle, halothane
(0.8±0.9 mM) suppresses acetylcholine-induced potentiation of regulatory myosin light chain (MLC) phosphorylation,60 which normally enhances contraction at
submaximal calcium levels. This potentiation is dependent
on the presence of GTP. AlF-induced potentiation, a mimic
of the GTPgP, is also blocked, but much more effectively
when halothane is added before AlF,63 suggesting that
halothane inhibits dissociation of the AlF-GDP charged a
monomer from the bg heterodimer. Once these subunits are
allowed to dissociate, the anaesthetic is less effective.
Potentiation is also dependent on a monomeric rho-related
GTPase downstream of the muscarinic receptor, the Gprotein heterotrimer and MLC phosphatase, which is
normally inhibited by this low molecular weight Gprotein.63 These results are consistent with halothane
causing a net increase in MLC phosphatase activity without
changing MLC kinase activity.46 It is worth noting that
earlier studies showed that pretreatment of guinea-pig ileum
myenteric plexus longitudinal muscle with PTX greatly
reduced the sensitivity to halothane, shifting the IC50 from
2% to 9% (vol).103 Clearly, at least some of the actions of
halothane are intimately linked to Gi/o activity in smooth
muscle.
71
Rebecchi and Pentyala
In addition, butanol, hexanol and octanol are able to
completely relax airway smooth muscle.113 When a
muscarinic agonist is applied, these n-alkanols suppress
the enhanced calcium sensitivity in a manner similar to
halothane, although the exact mechanism has yet to be
described. Ethanol also relaxes smooth muscle and inhibits
rMLC phosphorylation induced by muscarinic stimulation,45 but here calcium sensitivity is increased, presumably
via an independent mechanism.
Other systems
In turkey erythrocytes, low concentrations of halothane
(0.1±0.3 mM) signi®cantly inhibit stimulation of PLC by
isoprenaline and P2Y purinergic agonists, whose receptors
are coupled to PLC through Gi.109 In the case of isoprenaline, PLC activity was reduced nearly 50%; however, the
effects were biphasic, with stimulation of PLC observed at
high anaesthetic concentrations, obscuring any further
inhibition. In endothelial cells cultured from aorta, ethanol
(10±100 mM) enhances nitric oxide synthase activity in a
PTX-sensitive manner.53 Interestingly, ethanol also enhances PTX-catalysed ADP-ribosylation of Gi, suggesting
that this n-alkanol alters a subunit conformation or ADPribosyl transferase activity.
Work on oocytes
As discussed in the section on PKC, Xenopus oocytes have
been used to study the coupling between metabotropic
receptors, G-proteins and PLC. From these studies it
appears that receptors which couple strongly to PLC
through Gq are suppressed by general anaesthetic agents
via PKC-catalysed phosphorylation of the receptor.
Nonetheless, anaesthetic can suppress GTPgS-stimulated
calcium mobilization,73 suggesting that some of these
agents also block at the level of GTP binding proteins or
downstream effectors, such as PLC. Here, an action on Gi or
Go is also plausible, since stimulation of PLC in oocytes is
PTX-sensitive.105
The effects of anaesthetic agents on receptor/G-protein
regulated ion channels have also been examined. G-proteincoupled inwardly rectifying K+ channels (GIRKs), which
play an important role in neurotransmission,79 80 are clearly
affected when expressed in oocytes. Both brain (GIRK1/2)
and cardiac (GIRK1/4) types are activated by ethanol,70 71
whereas other inwardly rectifying K+ channels are not. A
homologous series of n-alkanols from C1 to C3 activate
GIRK1/2 and GIRK1/4 with decreasing ef®cacy, whereas
pentanol is inhibitory.70 71 Heterodimers containing the
GIRK2 subunit and GIRK2 homodimers are most sensitive.71 PTX had no signi®cant effect on ethanol activation of
either channel heterodimer,70 suggesting a direct action on
the channel itself. GABAB- and m-opioid-receptor coupling
to the channel was also enhanced by ethanol,70 71 with the
degree of stimulation being independent of protein kinase A
and PKC.71 By contrast with these results, clinical concentrations of halothane, iso¯urane and en¯urane and F3
inhibited basal GIRK1/2-, GIRK2- and m-opioid-receptoractivated GIRK2 channel currents, whereas GIRK1/4 was
resistant to all except F3.138 This demonstrates clear
differences between the anaesthetics and their effects on
speci®c subunit combinations. In a similar study,
0.1±0.3 mM halothane suppressed M2 muscarinic receptor
activated GIRK1/4 and GIRK1 channels.135 Halothane also
blocked activation by GTPgS, suggesting that this anaesthetic disrupted the coupling of bg subunits with the
channel. The apparent differential halothane sensitivity of
M2135 and m opioid138 receptors and stimulation of the
GIRK1/4 channel is consistent with receptor-speci®c
actions of this anaesthetic at the level of G-protein coupling.
Work in Caenorhabditis elegans
The nematode species C. elegans has been exploited to
identify genes that modulate sensitivity to volatile anaesthetics. Among the most notable are syntaxin and the Gao
subunit. Mutations in syntaxin, part of the presynaptic
complex of proteins regulating transmitter release, modulate
anaesthetic sensitivity over a 33-fold range, measured as a
loss of coordinated locomotion.132 133
Go and related G-proteins are negative modulators of
transmitter release, suppressing, in a voltage-dependent
manner, calcium channel opening and vesicle fusion in C.
elegans and man.66 Four different loss-of-function mutations in the a subunit resulted in decreased sensitivities to
iso¯urane (~2-fold), but only one, an amino acid substitution near the C-terminus (sy192), was associated with both
iso¯urane and halothane resistance. Genetic analysis of
sy192 suggested an interference with other components of
pathways under Go control.132 Crossing a hypersensitive
syntaxin strain with an iso¯urane-resistant Gao mutant
leads to the latter phenotype, showing that syntaxin-related
hypersensitivity depends on normal Go function. Consistent
with these results, over-expression of the RGS protein,
EGL-10, a negative modulator of Go, or a reduced function
mutation in eat-16, an RGS that operates on Gq, produced
similar resistance to both anaesthetics.132
Whatever the mechanism, it seems unlikely that Gao is
the sole/principal anaesthetic target of halothane in C.
elegans. Rather, another component associated with transmitter release machinery (SNARE complex) of syntaxin,
SNAP-25, VAMP and N/P type calcium channel is more
likely. The interfering effects of the RGS, egl-10, and the
sy192 mutation, which are located near the bg binding
interface of Gao, suggest a shift in the normal equilibrium
between a and bg subunits, resulting in a higher proportion
of Gao-GDP and abg heterotrimer complex and a reduction
in the amounts of free bg heterodimer. Since free bg appears
to inhibit the SNARE complex,141 these mutations should
increase neurotransmission, which appears to be the case.
Nonetheless, enhanced transmitter release opposing the
72
Protein kinase C and guanine nucleotide-binding proteins
Fig 5 Structure of Gai in the GTP and GDP bound state. This ®gure shows ribbon diagrams of the ai1 subunit charged with a non-hydrolysable GTP
analogue or GDP (based on x-ray crystallographic structures).127 The a-helical and ras-like domains, and the guanine nucleotide binding and switch
regions are indicated. In the GDP state, a micro-domain is formed from the disordered N- and C-terminal regions that do not appear in the GTP state.
Other changes in loop conformations are also induced, creating a new binding surface for receptor, Gbg subunit and effector. A large cavity, unique to
the GTP state, is shown in purple and is of suf®cient volume to bind volatile anaesthetics. This cavity is lost in the GDP state. Anaesthetic binding to
this site may affect binding of guanine nucleotide and prevent the normal coupling of the switch region and the N/C-terminal micro-domain
conformations.
Comparisons of halothane and other volatile anaesthetics,
iso¯urane, en¯urane and sevo¯urane, showed that all drugs
suppress exchange on ai2 at clinical concentrations, with
iso¯urane being the most effective. Similar results were
obtained when recombinant Gai1 lacking fatty acyl chain
modi®cation was tested in the absence of detergents and in
membranes prepared from Sf9 cells overexpressing Gai1
and bg subunits.100 By contrast to the sensitive a subunits,
the low molecular weight G-protein, ras, was resistant to
high concentrations of each drug. A recent study of
transducin showed that this Gi/Go family member is
insensitive to halothane, as measured by photocrosslinking
of radiolabelled drug.58 Thus, the anaesthetic agents display
measurable degrees of G-protein speci®city in vitro.
Since binding and exchange of the guanine nucleotide are
coupled to the conformations of the switch regions and the
receptor/bg subunit interface, we tested whether anaesthetics could affect the binding of the a subunit to the bg
heterodimer. When charged with Mg2+-GTPgS, the soluble
a subunit has a low af®nity for the bg vesicles, but in the
presence of anaesthetic, formation of abg complex is
greatly enhanced (Fig. 4). A similar, although less dramatic,
enhancement of abg complex formation is observed when
the a subunit is charged with GDP. These effects are
reversible and dose dependent (EC50 0.25 mM).
These results obtained with the puri®ed protein suggest
an uncoupling of the normal conformational changes
between the switch regions and the N/C-terminal regions
of the Ga subunits that form the bg binding interface. (An
actions of anaesthetics operating elsewhere cannot be the
sole explanation for resistance to these drugs.132 Likewise,
developmental defects cannot account for these observations, since inducing wild-type Gao expression after
development recovers anaesthetic sensitivity.
Studies in transgenic animals
Most of the genes encoding Ga subunits have been
inactivated in mice, but targeted mutations of Gb or Gg
have not been reported.93 Interestingly, Gao-de®cient mice
appear to be hyperalgesic,59 Gaz-de®cient mice exhibit
altered responses to a variety of psychoactive drugs,139 and
Gaolf-de®cient mice exhibit dramatically reduced electrophysiological responses to odorants.14 Mice lacking Gaq
develop ataxia, with clear signs of motor coordination
de®cits, although, like PKC-g null mice, this appears to be a
developmental problem.68 94 Despite the availability of Ga
transgenic mice, the sensitivity of these mice to anaesthetic
agents has yet to be tested.
Puri®ed GTP-binding proteins
To test the idea that volatile anaesthetics directly affect Ga
subunits, we examined exchange of bound GDP for GTP,100
a reaction normally stimulated by activated receptors.16 At
clinical doses halothane substantially reduced the extent of
nucleotide exchange for ai1, ai2, ai3 and as, but not ao. Of
all the subunits tested, Gai2 was most sensitive.
73
Rebecchi and Pentyala
membrane calcium ATPase,40 74 or actin.62 Other targets are
also undoubtedly important, as shown in studies in C.
elegans. Thus, the data do not support a single site theory
unless one argues that all these other effects are irrelevant. A
more realistic theory would be similar to the multi-site
agent-speci®c hypothesis proposed by MacIver,76 wherein
each drug is associated with its own array of protein targets
(some common and some agent speci®c), with distinct
subsets for each aspect of anaesthesia (sedation, amnesia
and immobility).
effect of anaesthetics on bg seems unlikely given the
stability of the heterodimer, whose WD-40 motifs form a
rigid platform for the docking of other proteins; thus, the
conformation of the bg binding interface in Mg2+-GTPgScharged a subunits must resemble that of the GDP state.) In
the GDP state, the disorder in the switch region permits the
folding of the N-terminal and C-terminal residues into a
stable micro-domain. Conversely, these regions become
highly disordered when GTP and Mg2+are bound (Fig. 5).
Thus, Ga subunits use the favorable energy of binding GTP
and Mg2+ to select a stable protein conformer, ¯icking open
a switch that triggers the release of bg subunits, thereby
modulating downstream effector proteins.69 127 In the
presence of anaesthetic, the binding interface is altered,
enhancing a subunit af®nity for bg. Hence, anaesthetics are
predicted to lock the metabotropic receptor into a persistent,
high-af®nity complex with the G-protein, which would
explain how volatile anaesthetics prevent the decrease in
receptor af®nity for agonist that is normally induced by nonhydrolysable GTP analogues, as reported by Aronstam and
others.5±8 42 112 This could have profound implications for
the operation of this molecular switch, since it would
effectively prevent GTP-charged a and bg subunits from
coupling to their downstream effectors.
Acknowledgement
The work in the authors' laboratory is supported by NIH Grant GM-60376.
References
1 Alagarsamy S, Sorensen SD, Conn PJ. Coordinate regulation of
metabotropic glutamate receptors. Curr Opin Neurobiol 2001; 11:
357±62
2 Anthony BL, Dennison RL, Jr., Narayanan TK, Aronstam RS.
Diethyl ether effects on muscarinic acetylcholine receptor
complexes in rat brainstem. Biochem Pharmacol 1988; 37: 4041±6
3 Anthony BL, Dennison RL, Aronstam RS. Disruption of
muscarinic receptor-G protein coupling is a general property
of liquid volatile anesthetics. Neurosci Lett 1989; 99: 191±6
4 Aronstam RS, Anthony BL, Dennison RL, Jr. Halothane effects on
muscarinic acetylcholine receptor complexes in rat brain.
Biochem Pharmacol 1986; 35: 667±72
5 Aronstam RS, Narayanan TK. Temperature effect on the
detection of muscarinic receptor-G protein interactions in
ligand binding assays. Biochem Pharmacol 1988; 37: 1045±9
6 Aronstam RS. G-Protein systems. In: Yaksh TL, Lynch C, Zapol
WM, Maze M, Biebuyck JF, Saidman LJ, eds. AnesthesiaÐBiological
Foundations. Philadelphia: Lippincott-Raven, 1997; 45±53
7 Aronstam RS, Dennison RL, Jr. Anesthetic effects on muscarinic
signal transduction. Int Anesthesiol Clin 1989; 27: 265±72
8 Aronstam RS, Triggle DJ, Eldefrawi ME. Structural and
stereochemical requirements for muscarinic receptor binding.
Mol Pharmacol 1979; 15: 227±34
9 Asano T, Morishita R, Kato K. Two forms of G(o) type G
proteins: identi®cation and distribution in various rat tissues and
cloned cells. J Neurochem 1992; 58: 2176±81
10 Asano T, Nagahama M, Kato K. Subcellular distribution of GTPbinding proteins, Go and Gi2, in rat cerebral cortex. J Biochem
1990; 107: 694±8
11 Balmforth AJ, Shepherd FH, Warburton P, Ball SG. Evidence of
an important and direct role for protein kinase C in agonistinduced phosphorylation leading to desensitization of the
angiotensin AT1A receptor. Br J Pharmacol 1997; 122: 1469±77
12 Bauche F, Bourdeaux-Jaubert AM, Giudicelli Y, Nordmann R.
Ethanol alters the adenosine receptor-Ni-mediated adenylate
cyclase inhibitory response in rat brain cortex in vitro. FEBS Lett
1987; 219: 296±300
13 Bazil CW, Minneman KP. Effects of clinically effective
concentrations of halothane on adrenergic and cholinergic
synapses in rat brain in vitro. J Pharmacol Exp Ther 1989; 248:
143±8
14 Belluscio L, Gold GH, Nemes A, Axel R. Mice de®cient in G(olf)
are anosmic. Neuron 1998; 20: 69±81
15 Bohm M, Schmidt U, Gierschik P, Schwinger RH, Bohm S,
Erdmann E. Sensitization of adenylate cyclase by halothane in
Future directions
Many studies that measure responses downstream of Gprotein coupling use high concentrations of anaesthetics,
bringing into question the relevance of these pathways to the
anaesthetic state. Nonetheless, as more reductionist approaches have simpli®ed these systems, it is clear that
receptor/G-protein coupling is suf®ciently sensitive to a
wide range of anaesthetics. Once the most sensitive
metabotropic receptor/G-protein pairs are identi®ed, clear
predictions can be generated that should be testable in
neuronal circuits of the brain and spinal cord.
More work is needed on a molecular level to explore the
anaesthetic sensitivities of the G-proteins themselves,
particularly Gq, which couples to many of the receptors
examined in oocyte experiments. Thus far it appears that
coupling to Gq is sensitive at the level of the receptor
(through PKC phosphorylation) whereas coupling to Gi/o
seems sensitive at the level of G-protein or receptor/Gprotein complex. De®nitive answers will only be provided
by studying puri®ed receptors and G-proteins reconstituted
into arti®cial bilayers.
Conclusion and perspective
This review has focused exclusively on PKC and G-proteins
because they are integral to metabotropic receptor signalling, which is clearly affected by general anaesthetics.
Because of space limitations we did not discuss other
important targets including calcium release channels such as
the ryanodine receptor,25 97 and pumps such as the plasma
74
Protein kinase C and guanine nucleotide-binding proteins
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
human myocardium and S49 lymphoma wild-type and cyc cells:
evidence for inactivation of the inhibitory G protein Gi alpha.
Mol Pharmacol 1994; 45: 380±9
Bourne HR. How receptors talk to trimeric G proteins. Curr
Opin Cell Biol 1997; 9: 134±42
Bowers BJ, Elliott KJ, Wehner JM. Differential sensitivity to the
anxiolytic effects of ethanol and ¯unitrazepam in PKCgamma null
mutant mice. Pharmacol Biochem Behav 2001; 69: 99±110
Bowers BJ, Owen EH, Collins AC, Abeliovich A, Tonegawa S,
Wehner JM. Decreased ethanol sensitivity and tolerance
development in gamma-protein kinase C null mutant mice is
dependent on genetic background. Alcohol Clin Exp Res 1999; 23:
387±97
Brandon NJ, Delmas P, Kittler JT, et al. GABAA receptor
phosphorylation and functional modulation in cortical neurons
by a protein kinase C-dependent pathway. J Biol Chem 2000; 275:
38856±62
Bremerich DH, Hirasaki A, Jones KA, Warner DO. Halothane
attenuation of calcium sensitivity in airway smooth muscle.
Mechanisms of action during muscarinic receptor stimulation.
Anesthesiology 1997; 87: 94±101
Brown DA, London E. Structure and origin of ordered lipid
domains in biological membranes. J Membr Biol 1998; 164:
103±14
Cantor RS. The lateral pressure pro®le in membranes: a physical
mechanism of general anesthesia. Biochemistry 1997; 36: 2339±44
Charness ME, Querimit LA, Henteleff M. Ethanol differentially
regulates G proteins in neural cells. Biochem Biophys Res Commun
1988; 155: 138±43
Chen CA, Manning DR. Regulation of G proteins by covalent
modi®cation. Oncogene 2001; 20: 1643±52
Chini EN. Selected contribution: effect of volatile anesthetics on
cADP-ribose-induced Ca(2+) release system. J Appl Physiol 2001;
91: 516±21; discussion 504±5
Cho W. Membrane targeting by C1 and C2 domains. J Biol Chem
2001; 276: 32407±10
Connolly CN, Kittler JT, Thomas P, et al. Cell surface stability of
gamma-aminobutyric acid type A receptors. Dependence on
protein kinase C activity and subunit composition. J Biol Chem
1999; 274: 36565±72
Csukai M, Chen CH, De Matteis MA, Mochly-Rosen D. The
coatomer protein beta'-COP, a selective binding protein
(RACK) for protein kinase Cepsilon. J Biol Chem 1997; 272:
29200±6
Dascal N. Ion-channel regulation by G proteins. Trends Endocrinol
Metab 2001; 12: 391±8
De Vries L, Zheng B, Fischer T, Elenko E, Farquhar MG. The
regulator of G protein signaling family. Annu Rev Pharmacol Toxicol
2000; 40: 235±71
Dempsey EC, Newton AC, Mochly-Rosen D, et al. Protein kinase
C isozymes and the regulation of diverse cell responses. Am J
Physiol 2000; 279: L429±38
Dennison RL, Jr., Anthony BL, Narayanan TK, Aronstam RS.
Effects of halothane on high af®nity agonist binding and guanine
nucleotide sensitivity of muscarinic acetylcholine receptors from
brainstem of rat. Neuropharmacology 1987; 26: 1201±5
Do SH, Kamatchi GL, Durieux ME. The effects of iso¯urane on
native and chimeric muscarinic acetylcholine receptors: the role
of protein kinase C. Anesth Analg 2001; 93: 375±81
Durieux ME. Halothane inhibits signaling through m1 muscarinic
receptors expressed in Xenopus oocytes. Anesthesiology 1995;
82: 174±82
Eckenhoff MF, Chan K, Eckenhoff RG. Multiple speci®c binding
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
75
targets for inhaled anesthetics in the mammalian brain. J
Pharmacol Exp Ther 2002; 300: 172±9
Filippova N, Sedelnikova A, Zong Y, Fortinberry H, Weiss DS.
Regulation of recombinant gamma-aminobutyric acid (GABA)(A)
and GABA(C) receptors by protein kinase C. Mol Pharmacol
2000; 57: 847±56
Firestone S, Firestone L. Protein kinase C. In: Yaksh TL, Lynch C,
Zapol WM, Maze M, Biebuyck JF, Saidman LJ, eds.
AnesthesiologyÐBiological Foundations. Philadelphia: LippincottRaven, 1997; 75±86
Firestone S, Firestone LL, Ferguson C, Blanck D. Staurosporine,
a protein kinase C inhibitor, decreases the general anesthetic
requirement in Rana pipiens tadpoles. Anesth Analg 1993; 77:
1026±30
Francesconi A, Duvoisin RM. Opposing effects of protein kinase
C and protein kinase A on metabotropic glutamate receptor
signaling: selective desensitization of the inositol trisphosphate/
Ca2+ pathway by phosphorylation of the receptor-G proteincoupling domain. Proc Natl Acad Sci USA 2000; 97: 6185±90
Franks JJ, Wamil AW, Janicki PK, et al. Anesthetic-induced
alteration of Ca2+ homeostasis in neural cells: a temperaturesensitive process that is enhanced by blockade of plasma
membrane Ca2+-ATPase isoforms. Anesthesiology 1998; 89:
149±64
Gomes I, Jordan BA, Gupta A, Rios C, Trapaidze N, Devi LA. G
protein coupled receptor dimerization: implications in
modulating receptor function. J Mol Med 2001; 79: 226±42
Gomes I, Trapaidze N, Turndorf H, Devi LA, Bansinath M. Acute
ethanol treatment modulates delta opioid receptors in N18TG2
cells. Anesthesiology 2000; 92: 1789±98
Gordon AS, Yao L, Jiang Z, Fishburn CS, Fuchs S, Diamond I.
Ethanol acts synergistically with a D2 dopamine agonist to cause
translocation of protein kinase C. Mol Pharmacol 2001; 59:
153±60
Gordon AS, Yao L, Wu ZL, Coe IR, Diamond I. Ethanol alters
the subcellular localization of delta- and epsilon protein kinase C
in NG108±15 cells. Mol Pharmacol 1997; 52: 554±9
Hanazaki M, Jones KA, Perkins WJ, Warner DO. Halothane
increases smooth muscle protein phosphatase in airway smooth
muscle. Anesthesiology 2001; 94: 129±36
Hanazaki M, Jones KA, Perkins WJ, Warner DO. The effects of
ethanol on Ca(2+) sensitivity in airway smooth muscle. Anesth
Analg 2001; 92: 767±74
Harikumar KG, Chattopadhyay A. Effect of alcohols on Gprotein coupling of serotonin(1A) receptors from bovine
hippocampus. Brain Res Bull 2000; 52: 597±601
Harris RA, McQuilkin SJ, Paylor R, Abeliovich A, Tonegawa S,
Wehner JM. Mutant mice lacking the gamma isoform of protein
kinase C show decreased behavioral actions of ethanol and
altered function of gamma-aminobutyrate type A receptors. Proc
Natl Acad Sci USA 1995; 92: 3658±62
Hemmings HC Jr. General anesthetic effects on protein kinase C.
Toxicol Lett 1998; 100±101: 89±95
Hemmings HC Jr., Adamo AI. Effects of halothane and propofol
on puri®ed brain protein kinase C activation. Anesthesiology 1994;
81: 147±55
Hemmings HC, Adamo AI, Hoffman MM. Biochemical
characterization of the stimulatory effects of halothane and
propofol on puri®ed brain protein kinase C. Anesth Analg 1995;
81: 1216±22
Hemmings HC Jr., Adamo AI. Activation of endogenous protein
kinase C by halothane in synaptosomes. Anesthesiology 1996; 84:
652±62
Hendrickson RJ, Cahill PA, Sitzmann JV, Redmond EM. Ethanol
Rebecchi and Pentyala
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
enhances basal and ¯ow-stimulated nitric oxide synthase activity
in vitro by activating an inhibitory guanine nucleotide binding
protein. J Pharmacol Exp Ther 1999; 289: 1293±300
Hernandez RM, Wescott GG, Mayhew MW, McJilton MA,
Terrian DM. Biochemical and morphogenic effects of the
interaction between protein kinase C-epsilon and actin in vitro
and in cultured NIH3T3 cells. J Cell Biochem 2001; 83: 532±46
Ho C, Slater SJ, Stagliano BA, Stubbs CD. Conformation of the
C1 phorbol-ester-binding domain participates in the activating
conformational change of protein kinase C. Biochem J 1999; 344
Pt 2: 451±60
Hodge CW, Mehmert KK, Kelley SP, et al. Supersensitivity to
allosteric GABA(A) receptor modulators and alcohol in mice
lacking PKCepsilon. Nature Neurosci 1999; 2: 997±1002
Homanics GE, DeLorey TM, Firestone LL, et al. Mice devoid of
gamma-aminobutyrate type A receptor beta3 subunit have
epilepsy, cleft palate, and hypersensitive behavior. Proc Natl
Acad Sci USA 1997; 94: 4143±8
Ishizawa Y, Sharp R, Liebman PA, Eckenhoff RG. Halothane
binding to a G protein coupled receptor in retinal membranes by
photoaf®nity labeling. Biochemistry 2000; 39: 8497±502
Jiang M, Gold MS, Boulay G, et al. Multiple neurological
abnormalities in mice de®cient in the G protein Go. Proc Natl
Acad Sci USA 1998; 95: 3269±74
Jones KA, Hirasaki A, Bremerich DH, Jankowski C, Warner DO.
Halothane inhibits agonist-induced potentiation of rMLC
phosphorylation in permeabilized airway smooth muscle. Am J
Physiol 1997; 273: L80±5
Jordan FL. Distribution and expression of G-protein in rat
cerebral cortical cells. II. Primary tissue culture. Brain Res Dev
Brain Res 1992; 67: 11±18
Kaech S, Brinkhaus H, Matus A. Volatile anesthetics block actinbased motility in dendritic spines. Proc Natl Acad Sci USA 1999;
96: 10433±7
Kai T, Jones KA, Warner DO. Halothane attenuates calcium
sensitization in airway smooth muscle by inhibiting G-proteins.
Anesthesiology 1998; 89: 1543±52
Kamatchi GL, Durieux ME, Lynch C, 3rd. Differential sensitivity
of expressed L-type calcium channels and muscarinic M(1)
receptors to volatile anesthetics in Xenopus oocytes. J Pharmacol
Exp Ther 2001; 297: 981±90
Kamatchi GL, Tiwari SN, Durieux ME, Lynch C, 3rd. Effects of
volatile anesthetics on the direct and indirect protein kinase Cmediated enhancement of alpha1E-type Ca(2+) current in
Xenopus oocytes. J Pharmacol Exp Ther 2000; 293: 360±9
Keren O, Gafni M, Sarne Y. Opioids potentiate transmitter
release from SK-N-SH human neuroblastoma cells by modulating
N-type calcium channels. Brain Res 1997; 764: 277±82
Khasar SG, Lin YH, Martin A, et al. A novel nociceptor signaling
pathway revealed in protein kinase C epsilon mutant mice.
Neuron 1999; 24: 253±60
Kim JJ, Thompson RF. Cerebellar circuits and synaptic
mechanisms involved in classical eyeblink conditioning. Trends
Neurosci 1997; 20: 177±81
Kleuss C, Raw AS, Lee E, Sprang SR, Gilman AG. Mechanism of
GTP hydrolysis by G-protein alpha subunits. Proc Natl Acad Sci
USA 1994; 91: 9828±31
Kobayashi T, Ikeda K, Kojima H, et al. Ethanol opens G-proteinactivated inwardly rectifying K+ channels. Nature Neurosci 1999;
2: 1091±7
Lewohl JM, Wilson WR, May®eld RD, Brozowski SJ, Morrisett
RA, Harris RA. G-protein-coupled inwardly rectifying potassium
channels are targets of alcohol action. Nat Neurosci 1999; 2:
1084±90
72 Lin LH, Chen LL, Harris RA. En¯urane inhibits NMDA, AMPA,
and kainate-induced currents in Xenopus oocytes expressing
mouse and human brain mRNA. FASEB J 1993; 7: 479±85
73 Lin LH, Leonard S, Harris RA. En¯urane inhibits the function of
mouse and human brain phosphatidylinositol-linked acetylcholine
and serotonin receptors expressed in Xenopus oocytes. Mol
Pharmacol 1993; 43: 941±8
74 Lopez MM, Kosk-Kosicka D. Spectroscopic analysis of halothane
binding to the plasma membrane Ca2+-ATPase. Biophys J 1998;
74: 974±80
75 Ma D. Recent advances in the discovery of protein kinase C
modulators based on the structures of natural protein kinase C
activators. Curr Med Chem 2001; 8: 191±202
76 MacIver MB. General anesthetic actions on transmission of
glutamate and GABA synapses. In: Yaksh TL, Lynch C, Zapol
WM, Maze M, Biebuyck JF, Saidman LJ, eds. AnesthesiaÐBiological
Functions. Philadelphia: Lippincott-Raven, 1997; 277±86
77 Magyar J, Szabo G. Effects of volatile anesthetics on the G
protein-regulated muscarinic potassium channel. Mol Pharmacol
1996; 50: 1520±8
78 Marinissen MJ, Gutkind JS. G-protein-coupled receptors and
signaling networks: emerging paradigms. Trends Pharmacol Sci
2001; 22: 368±76
79 Mark MD, Herlitze S. G-protein mediated gating of inwardrecti®er K+ channels. Eur J Biochem 2000; 267: 5830±6
80 Mark MD, Ruppersberg JP, Herlitze S. Regulation of GIRK
channel deactivation by Galpha(q) and Galpha(i/o) pathways.
Neuropharmacology 2000; 39: 2360±73
81 Martin DC, Dennison RL, Introna RP, Aronstam RS. In¯uence of
halothane on the interactions of serotonin1A and adenosine A1
receptors with G proteins in rat brain membranes. Biochem
Pharmacol 1991; 42: 1313±6
82 Medkova M, Cho W. Interplay of C1 and C2 domains of protein
kinase C-alpha in its membrane binding and activation. J Biol
Chem 1999; 274: 19852±61
83 Milligan G, White JH. Protein±protein interactions at G-proteincoupled receptors. Trends Pharmacol Sci 2001; 22: 513±8
84 Minami K, Gereau RWt, Minami M, Heinemann SF, Harris RA.
Effects of ethanol and anesthetics on type 1 and 5 metabotropic
glutamate receptors expressed in Xenopus laevis oocytes. Mol
Pharmacol 1998; 53: 148±56
85 Minami K, Minami M, Harris RA. Inhibition of 5hydroxytryptamine type 2A receptor-induced currents by nalcohols and anesthetics. J Pharmacol Exp Ther 1997; 281:
1136±43
86 Minami K, Shiraishi M, Uezono Y, Ueno S, Shigematsu A. The
inhibitory effects of anesthetics and ethanol on substance P
receptors expressed in Xenopus oocytes. Anesth Analg 2002; 94:
79±83
87 Minami K, Vanderah TW, Minami M, Harris RA. Inhibitory effects
of anesthetics and ethanol on muscarinic receptors expressed in
Xenopus oocytes. Eur J Pharmacol 1997; 339: 237±44
88 Mounsey JP, Patel MK, Mistry D, John JE, Moorman JR. Protein
kinase C co-expression and the effects of halothane on rat
skeletal muscle sodium channels. Br J Pharmacolol 1999; 128:
989±98
89 Naik MU, Benedikz E, Hernandez I, Libien J, Hrabe J, Valsamis M,
et al. Distribution of protein kinase Mzeta and the complete
protein kinase C isoform family in rat brain. J Comp Neurol 2000;
426: 243±58
90 Narayanan TK, Confer RA, Dennison RL, Jr., Anthony BL,
Aronstam RS. Halothane attenuation of muscarinic inhibition of
adenylate cyclase in rat heart. Biochem Pharmacol 1988; 37:
1219±23
76
Protein kinase C and guanine nucleotide-binding proteins
91 Nechushtan H, Leitges M, Cohen C, Kay G, Razin E. Inhibition of
degranulation and interleukin-6 production in mast cells derived
from mice de®cient in protein kinase Cbeta. Blood 2000; 95:
1752±7
92 Oda T, Shearman MS, Nishizuka Y. Synaptosomal protein kinase
C subspecies: B. Down-regulation promoted by phorbol ester
and its effect on evoked norepinephrine release. J Neurochem
1991; 56: 1263±9
93 Offermanns S. In vivo functions of heterotrimeric G-proteins:
studies in Galpha-de®cient mice. Oncogene 2001; 20: 1635±42
94 Offermanns S, Hashimoto K, Watanabe M, et al. Impaired motor
coordination and persistent multiple climbing ®ber innervation of
cerebellar Purkinje cells in mice lacking Galphaq. Proc Natl Acad
Sci USA 1997; 94: 14089±94
95 Orr JW, Newton AC. Intrapeptide regulation of protein kinase
C. J Biol Chem 1994; 269: 8383±7
96 Otani S, Connor JA. Requirement of rapid Ca2+ entry and
synaptic activation of metabotropic glutamate receptors for the
induction of long-term depression in adult rat hippocampus. J
Physiol 1998; 511: 761±70
97 Pabelick CM, Prakash YS, Kannan MS, Warner DO, Sieck GC.
Effects of halothane on sarcoplasmic reticulum calcium release
channels in porcine airway smooth muscle cells. Anesthesiology
2001; 95: 207±15
98 Patel MK, Mistry D, John JE, 3rd, Mounsey JP. Sodium channel
isoform-speci®c effects of halothane: protein kinase C coexpression and slow inactivation gating. Br J Pharmacolol 2000;
130: 1785±92
99 Pavlin EG, Su JY. Cardiopulmonary pharmacology. In: Miller RD,
ed. Anesthesia. New York: Churchill Livingstone, 1994; 125±56
100 Pentyala SN, Sung K, Chowdhury A, Rebecchi MJ. Volatile
anesthetics modulate the binding of guanine nucleotides to the
alpha subunits of heterotrimeric GTP binding proteins. Eur J
Pharmacol 1999; 384: 213±22
101 Polgar E, Fowler JH, McGill MM, Todd AJ. The types of neuron
which contain protein kinase C gamma in rat spinal cord. Brain
Res 1999; 833: 71±80
102 Prokocimer PG, Maze M, Vickery RG, Kraemer FB, Gandjei R,
Hoffman BB. Mechanism of halothane-induced inhibition of
isoproterenol-stimulated lipolysis in isolated rat adipocytes. Mol
Pharmacol 1988; 33: 338±43
103 Puig MM, Turndorf H, Warner W. Synergistic interaction of
morphine and halothane in the guinea pig ileum: effects of
pertussis toxin. Anesthesiology 1990; 72: 699±703
104 Qian H, Pipolo L, Thomas WG. Identi®cation of protein kinase C
phosphorylation sites in the angiotensin II (AT1A) receptor.
Biochem J 1999; 343 Pt 3: 637±44
105 Quick MW, Simon MI, Davidson N, Lester HA, Aragay AM.
Differential coupling of G protein alpha subunits to seven-helix
receptors expressed in Xenopus oocytes. J Biol Chem 1994; 269:
30164±72
106 Rebecchi MJ, Pentyala SN. Structure, function, and control of
phosphoinositide-speci®c phospholipase C. Physiol Rev 2000; 80:
1291±335
107 Rodbell M. The complex regulation of receptor-coupled Gproteins. Adv Enzyme Regul 1997; 37: 427±35
108 Ron D, Vagts AJ, Dohrman DP, et al. Uncoupling of betaIIPKC
from its targeting protein RACK1 in response to ethanol in
cultured cells and mouse brain. FASEB J 2000; 14: 2303±14
109 Rooney TA, Hager R, Stubbs CD, Thomas AP. Halothane
regulates G-protein-dependent phospholipase C activity in
turkey erythrocyte membranes. J Biol Chem 1993; 268: 15550±6
110 Ross EM, Berstein G. Regulation of the M1 muscarinic receptor-
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
77
Gq-phospholipase C-beta pathway by nucleotide exchange and
GTP hydrolysis. Life Sci 1993; 52: 413±9
Saito N, Tsujino T, Fukuda K, Tanaka C. Alpha-, beta II- and
gamma-subspecies of protein kinase C localized in the monkey
hippocampus: pre- and post-synaptic localization of gammasubspecies. Brain Res 1994; 656: 245±56
Saito S, Kadoi Y, Ohyama A, Goto F. Halothane facilitates the
translocation of GRK-2 and phosphorylation of beta2-adrenergic
receptor in rat synaptosomes. Can J Anaesth 2000; 47: 73±80
Sakihara C, Jones KA, Lorenz RR, Perkins WJ, Warner DO.
Effects of primary alcohols on airway smooth muscle.
Anesthesiology 2002; 96: 428±37
Sanna E, Dildy-May®eld JE, Harris RA. Ethanol inhibits the
function of 5-hydroxytryptamine type 1c and muscarinic M1 G
protein-linked receptors in Xenopus oocytes expressing brain
mRNA: role of protein kinase C. Mol Pharmacol 1994; 45:
1004±12
Sanuki M, Yuge O, Kawamoto M, Fujii K, Azuma T. Sevo¯urane
inhibited beta-adrenoceptor-G protein bindings in myocardial
membrane in rats. Anesth Analg 1994; 79: 466±71
Schechtman D, Mochly-Rosen D. Adaptor proteins in protein
kinase C-mediated signal transduction. Oncogene 2001; 20:
6339±47
Schechtman D, Mochly-Rosen D. Isozyme-speci®c inhibitors and
activators of protein kinase C. Methods Enzymol 2002; 345:
470±89
Schmidt U, Schwinger RH, Muller-Ehmsen, et al. In¯uence of
halothane on the effect of cAMP-dependent and cAMPindependent positive inotropic agents in human myocardium.
Br J Anaesth 1994; 73: 204±8
Shen YM, Chertihin OI, Biltonen RL, Sando JJ. Lipid-dependent
activation of protein kinase C-alpha by normal alcohols. J Biol
Chem 1999; 274: 34036±44
Singh SP, Handa RK, Depala V, Gao Y, McIlroy PJ, Ravindra R.
The effect of ethanol on muscarinic receptor-G protein coupling
in the rat cortex. Pharmacol Toxicol 1997; 81: 294±9
Slater SJ, Cox KJ, Lombardi JV, et al. Inhibition of protein kinase
C by alcohols and anaesthetics. Nature 1993; 364: 82±4
Slater SJ, Kelly MB, Larkin JD, et al. Interaction of alcohols and
anesthetics with protein kinase Calpha. J Biol Chem 1997; 272:
6167±73
Slater SJ, Milano SK, Stagliano BA, et al. Synergistic activation of
protein kinase Calpha, -betaI, and -gamma isoforms induced by
diacylglycerol and phorbol ester: roles of membrane association
and activating conformational changes. Biochemistry 1999; 38:
3804±15
Slater SJ, Stagliano BA, Seiz JL, et al. Effects of ethanol on protein
kinase C activity induced by ®lamentous actin. Biochim Biophys
Acta 2001; 1544: 207±16
Slater SJ, Ho C, Taddeo FJ, Kelly MB, Stubbs CD. Contribution
of hydrogen bonding to lipid±lipid interactions in membranes and
the role of lipid order: effects of cholesterol, increased
phospholipid unsaturation, and ethanol. Biochemistry 1993; 32:
3714±21
Sonnenburg ED, Gao T, Newton AC. The phosphoinositidedependent kinase, PDK-1, phosphorylates conventional protein
kinase C isozymes by a mechanism that is independent of
phosphoinositide 3-kinase. J Biol Chem 2001; 276: 45289±97
Sprang SR. G protein mechanisms: insights from structural
analysis. Annu Rev Biochem 1997; 66: 639±78
Sun Z, Arendt CW, Ellmeier W, et al. PKC-theta is required for
TCR-induced NF-kappaB activation in mature but not immature
T lymphocytes. Nature 2000; 404: 402±7
Swope SL, Moss SI, Raymond LA, Huganir RL. Regulation of
Rebecchi and Pentyala
130
131
132
133
134
135
ligand-gated ion channels by protein phosphorylation. Adv Second
Messenger Phosphoprotein Res 1999; 33: 49±78
Toker A. Protein kinases as mediators of phosphoinositide 3kinase signaling. Mol Pharmacol 2000; 57: 652±8
Tsunoda S, Sun Y, Suzuki E, Zuker C. Independent anchoring and
assembly mechanisms of INAD signaling complexes in
Drosophila photoreceptors. J Neurosci 2001; 21: 150±8
vanSwinderen B, Metz LB, Shebester LD, Mendel JE, Sternberg
PW, Crowder CM. Goalpha regulates volatile anesthetic action
in Caenorhabditis elegans. Genetics 2001; 158: 643±55
vanSwinderen B, Saifee O, Shebester L, Roberson R, Nonet ML,
Crowder CM. A neomorphic syntaxin mutation blocks volatileanesthetic action in Caenorhabditis elegans. Proc Natl Acad Sci USA
1999; 96: 2479±84
Wang YX, Dhulipala PD, Li L, Benovic JL, Kotlikoff MI. Coupling
of M2 muscarinic receptors to membrane ion channels via
phosphoinositide 3-kinase gamma and atypical protein kinase C. J
Biol Chem 1999; 274: 13859±64
Weigl LG, Schreibmayer W. G protein-gated inwardly rectifying
potassium channels are targets for volatile anesthetics. Mol
Pharmacol 2001; 60: 282±9
136 Weigt HU, Kwok WM, Rehmert GC, Bosnjak ZJ. Sensitization of
the cardiac Na channel to alpha1-adrenergic stimulation by
inhalation anesthetics: evidence for distinct modulatory
pathways. Anesthesiology 1998; 88: 125±33
137 Wittinghofer A. Deciphering the alphabet of G proteins: the
structure of the alpha, beta, gamma heterotrimer. Structure 1996;
4: 357±61
138 Yamakura T, Bertaccini E, Trudell JR, Harris RA. Anesthetics and
ion channels: molecular models and sites of action. Annu Rev
Pharmacol Toxicol 2001; 41: 23±51
139 Yang J, Wu J, Kowalska MA, et al. Loss of signaling through the G
protein, Gz, results in abnormal platelet activation and altered
responses to psychoactive drugs. Proc Natl Acad Sci USA 2000;
97: 9984±9
140 Yoshimura M, Tabakoff B. Ethanol's actions on cAMP-mediated
signaling in cells transfected with type VII adenylyl cyclase. Alcohol
Clin Exp Res 1999; 23: 1457±61
141 Zamponi GW. Determinants of G protein inhibition of
presynaptic calcium channels. Cell Biochem Biophys 2001; 34:
79±94
78