G Proteins:
Implications K. Manji,
Husseini
There
is mounting
evidence
that
a family
their
treatments.
available
form
offunctional implicated
psychotropic
drugs
modulate
neuronal
G proteins
the basis
affect
and the complexities of the nervous tion ofneuronal function, it seems various major psychiatric illnesses. G protein targets (AmJ Psychiatry
remains 1992;
The
may
be one
an exciting 149:746-760)
prospect
for
ne of the most exciting recent advances in neuroscience has been the elucidation of the molecular mechanisms underlying neuronal communication. Although rapidly increasing numbers of potential neurotransmitters and receptors continue to be identified, it has become clear that the translation of the extracellular signals (into a form that can be interpreted by the complex intracellular enzymatic machinery) is generally achieved in relatively few ways. Generally speaking, these transmembrane signaling systems, and the receptors that use them, can be divided into two major groups: 1 Those with relatively self-contained structures and messages that take the form of transmembrane ion fluxes. .
with
multiple
components
and
messages
that
generate various intracellular “second messengers.” The first class of receptors contain in their stable molecular complex an intrinsic ion channel. Receptors of this class include those for a number of amino acidsincluding glutamate, y-aminobutyric acid (GABA) (through the GABAA receptor), and glycine-the nicotinic acetylcholine receptor, and the 5-HT3 receptor (1). Receptors of this type have also been described as “ionotropic” and are generally composed of four or five Received May 22, 1991; revision received Nov. 6, 1991; accepted Nov. 21, 1991. From the Section on Clinical Pharmacology, Experimental Therapeutics Branch, NIMH. Address reprint requests to Dr. Manji, Section on Clinical Pharmacology, Experimental Therapeutics Branch, NIMH, Bldg. 10, Rm. 2D46, 9000 Rockville Pike, Bethesda,
MD 20892. The Hsiao
746
author wishes for their helpful
proteins
(G pro-
in the CNS,
the involvement conditions and
endowing
the neuron
in the function and/or expression of G states, and a number of currently
understanding ofthe
keys
ofthe
mechanisms
to understanding
by which
the functioning
system. Given their widespread, critical roles in the regulalikely that G proteins are involved in the pathophysiology of The development of novel, site-specific drugs with primary
O
2. Those
integration
Abnormalities ofpathophysiologic
G proteins.
activity
binding
action in the CNS, and an emphasis on psychiatric
with
ofsignal
diversity. in a variety
triphosphate
ofa vast array ofextracellular, receptor-detected effectors. The author reviews the literature dealing generation, the role of G proteins in regulating
divergence,[neurotransmiuer ofclinical ctnditions,
G proteins
with a large degree proteins have been
M.D.
ofguanosine
teins) play an obligatory role in the transduction signals across cell membranes to intracellular with G protein coupling to second messenger both the convergence and of G proteins in a variety
for Psychiatry
to thank Drs. William Z. Potter and suggestions, input, and encouragement.
John
K.
the future.
subunits surrounding a central ionic pore, which opens transiently when the transmitter binds, allowing ions to flow either into the neuron (e.g., Na, Ca2, CF) or out of it (e.g., K), thereby generating synaptic potentials. Neurotransmission of this type is very fast: ion channels open and close within milliseconds (2). Most receptors, however, do not have ionic conductance channels within their structure; rather, they regulate cellular activity by generating various second messengers. In genera!, receptors of this class do not directly interact with the various second-messengergenerating enzymes but transmit information to the appropriate “effector” by the activation of interposed “coupling proteins.” It is now clear that a family of structurally related, guanosine-triphosphate-binding proteins (or G proteins) play an obligatory role in the transduction of a vast array of extracellular, receptordetected signals across cell membranes to intracellular effectors. Indeed, Birnbaumer (3) has suggested that about 80% of all known hormones, neurotransmitters, and neuromodulators (numbering around 100) elicit cellular responses through G proteins coupled to a Variety
of
cellular
effectors,
including
adenyly!
cyclase
(frequently referred to as adenylate cyclase), phospholipase C, phospholipase A2, and several ion channels (4). Receptors coupled to G proteins include those for catecholamines, serotonin, acetylcholine, various peptides (e.g., vasopressin, substance P), and even sensory signals, such as light and various odorants (3). Receptors coupled to G proteins are prototypic “metabotropic” receptors. Activation of these receptors initiates a cas-
Am
J
Psychiatry
1 49:6,
June
1992
HUSSEINI
cade of events, usually through regulation of intracellu!ar second messengers, that lead to modification of enzymatic phosphorylation of various membrane and cytosolic proteins (enzymatic activation or inhibition), including receptors, ion channels, and possibly G proteins themselves (1 ). As might be predicted, receptors coupled to G proteins generate electrical signals much more slowly, often with a latency of signal onset of at least 30 msec (2). The ability of many G proteins to interact with multiple receptors also provides an elegant mechanism for the neuron to respond to a large number of stimuli and to regulate both the convergence and divergence of neurotransmitter action (5, 6). The high degree of complexity generated by the interactions of G-protein-coupled receptors may be one mechanism by
which
neurons
acquire
the
flexibility
necessary
for
generating the wide range of responses observed in the nervous system; this has led to the suggestion that Gprotein-coupled receptors may be involved in pathways regulating such diverse vegetative functions as mood, appetite, and wakefulness (2). Thus, understanding the mechanisms by which G proteins
modulate
to understanding of the nervous
neuronal
of second
functioning Additionally,
is one
of
the
keys
the complexities there is a growing body of evidence implicating both quantitative and qualitative G protein abnormalities in a variety of clinical disorders. Similarly, there is increasing evidence that a number of psychotropic agents exert their clinically relevant effects at postreceptor sites, including G proteins. The identification of these putative G protein targets may thus lead not only to the development of more potent, site-specific treatments but also to the localization of molecular and biochemical factors predisposing individuals to various psychiatric illnesses. This overview will explore the growing knowledge of G protein coupling to chemical and ionic second messenger generation. Specifically, I will discuss the biochemistry
the system.
activity
messenger
and
synthesis,
the
role
of
ARE
G PROTEINS?
The G proteins are a family of GTP-binding that play an obligatory role in the transduction tracellular, receptor-detected signals across membrane to various intracellular effectors These proteins have been the focus of extensive since
the
cyclase
(the
demonstration enzyme
that which
stimulation
generates
both hormones and GTP. demonstrated that -adrenergic GTPase activity (discussed later) that guanine nucleotides-such as hydrolyzable analogues [Gpp(NH)p lated the affinity of this receptor quires
Am
J Psychiatry
1 49:6,
June
1992
cAMP)
proteins of exthe cell (3, 4). research of
adeny!y!
activity
antagonists. These and other indirect observations led to the proposal that the site of GTP’s action is on a protein distinct from both the hormone receptor and adenylyl cyclase. This hypothesis was confirmed by the subsequent purification of a GTP-binding protein from liver. By using the 549 cell mutant (deficient in hormoneand GTP-dependent activity), now termed “cyc,” it was possible to demonstrate that the addition of the purified “transducer protein” restores hormoneand GTP-dependent adenylyl
cyclase
activity.
Since
the
addition
of
this
pu-
rified protein stimulated adenylyl cyclase activity, it was termed “G,” (s=stimulatory) (reference 7 is an excellent historical overview). Shortly thereafter, the purification of a protein necessary for inhibition of adenylyl cyclase (“G1”) provided a mechanistic explanation for the bidirectional regulation of this enzymatic activity (8). G-like proteins were shown to activate a phospholipase C that hydrolyzes phosphatidylinositol 4,5bisphosphate to generate diacylglycerol and inositol 1,4,S-trisphosphate and to modulate the activity of specific
ion
channels
(3, 4, 9, 10).
Subsequently,
another
G
protein for which no obvious function was evident was purified from brain (11); this protein was named “G0” (o=other protein). The identification of G0 led to the realization that the G proteins are, in fact, a family of homologous proteins serving diverse roles in a wide variety of receptor-mediated extracellular signals to vanous intracellular second messenger systems (3, 4, 9, 10). It is now clear that G proteins involved in signal transduction are cx1y hetenotrimers. The greatest diversity observed thus far is with a subunits, which are generally believed to directly modulate the activities of vanous effectors (discussed later), but diversity also exists for and y subunits (see table 1).
MECHANISM
OF G PROTEIN
ACTIVATION
G
proteins in neurotransmitter-neurotransmitter interactions, and the possible involvement of G proteins in a variety of clinical conditions, with an emphasis on psychiatric conditions and their treatments.
WHAT
K. MANJI
re-
It was subsequently agonists stimulated in erythrocytes and GTP or similar nonor GTPyS]-modufor agonists but not
The molecular events underlying G protein activation/deactivation have been best characterized for Gprotein-mediated regulation of adenylyl cyclase (3, 4, 13), but the following sequence of events is assumed to occur for most (if not all) G proteins (see figure 1). The G protein a subunit cycles between an inactive GDP-bound oligomeric (4y) form and an active GTPbound monomeric form. Gilman (4, 13) has described these two forms of the a subunit as the “off” and “on” positions of a carefully timed molecular switch. Activation of a receptor by an agonist induces a conformational change in the receptor, allowing it to interact with the G protein, and forms a short-lived “high-affinity ternary complex” consisting of these components. The interaction of the receptor with the G protein in turn results in a conformational change in the a subunit of the G protein, which facilitates the displacement of GDP by GTP. The binding of GTP is a crucial step in the activation cycle and has two major consequences. 1 The G protein dissociates from the receptor. This destabilization of the high-affinity ternary complex re.
747
G PROTEINS
TABLE
1. Key Features
of G Protein Subunitsa
Subunit
Molecular Weight (x103)
Ga, (four types)
44.5-46
G, olf Ga1 (three
40.4-40.5
Toxin Targeting Subunit (cholera or pertussis)
Effector(s)’
Examples
Cholera
AC (+); L-type
Cholera
AC
(+)
Pertussis
AC
(-);
40.5
Both
Cyclic
Ga Ga0 (two types)
40.5 39.9
Both Pertussis
Cyclic GMP phosphodiesterase K channels (+); Ca2 channels PLC (+) (sensitive to pertussis
Gct
40.9
Neither
PLC
42
Neither
PLC (+) (insensitive
Direct interaction with AC (-)‘; inactivates a y required for interaction of a subunit with receptor
Ga
r
G
l
44.7
types)
,
G14
,
?four
types)
37.4
Neither
#{149}1 (three
types)
8-10
Neither
aAdapted from Birnbaumer (3), Taylor bAC=adenylyl cyclase, PLC=phospholipase CFinding uncertain.
et al. (9), Freissmuth C, +=stimulatory
sults
to the
in the
receptor’s
reversion
K channels GMP
low-affinity
con-
of
is terminated to
GDP.
the
However,
additional
adenylyl
ct-GDP
by the action
a subunit, this process amplification
of a GTPase
enzyme
which hydrolyzes GTP to is relatively slow, allowing of the signal. The forma-
is presumed
to
cause
dissociation
of
a
effector (which is now free to interact with activated a-GTP subunits); the reassociation with y is thermodynamically stable and the cycle with the formation of the inactive holo-G protein. appear to exert their inhibitory effects on cyclase
by two
distinct
mechanisms.
First,
in a
manner closely paralleling that just outlined for there appears to be a direct inhibitory effect of aj on catalytic unit of adenylyl cyclase. Gilman (4, 13) proposed that, in addition to the direct inhibition by the ‘y subunits, released by receptor activation of may by mass action attenuate the dissociation of Since the quantity of G is considerably greater than quantity of G, in most tissues and y subunits appear be largely interchangeable, this sequestration model received
much
acceptance.
Although
some
debate
G, the has a, G, G. the to has
per-
sists as to whether the a, or y subunits have the greater role in the inhibition of adenylyl cyclase, more studies suggest that they exert distinct but complementary effects
subunits
748
in
and Simon -=inhibitory
Unknown Thromboxane
A2, vasopressin
(12). effect.
stimulation
of
that
a is the primary
adenylyl
cyclase
(14).
mediator
It is
of inhibi-
active.
intrinsic
of
toxin) toxin) toxin)
tion of adenylyl cyclase when G is stimulated by a hormone or neunotransmitter but that y exerts a major role in the inhibition of basal and postreceptor-stimulated adenylyl cyclase activity.
a-GTP
from the additional of a-GDP completes GDP-bound Receptors
(-);
to pertussis
possible
The ticular, lected
tion
(+) (+)
to pertussis
dependent thus
12, D1, A2, H2, ACTH, CRH, V2, PGE1 Olfactory signals 2’ D2, A1, t, M2, S-HTIA Retinal rods (rhodopsins) Retinal cones (rhodopsins) a2, GABAB ,
(+)
(insensitive
(+)C
(+)
phosphodiesterase
2. The second major consequence of GTP binding to the G protein is to promote dissociation into a-GTP and ysubunits. It is generally accepted that, for a numben of effector systems, the a-GTP complex activates the effector enzyme by an as yet undetermined mechanism. The continued activation of adenylyl cyclase by
for
channels
et al. (10), and Strathmann effect of subunit on effector,
formation and the dissociation of the agonist from the receptor. Thus, the receptor is not permanently associated with the G-adenylyl cyclase complex, and this a!lows the receptor to recycle and function catalytically, activating several G proteins during the time that one
G remains
Ca2
of Receptors
the
may
inhibition
be more
of
potent
adenylyl
inhibitors
cyclase.
Thus,
of receptor-in-
‘y
G PROTEINS
AS TARGETS
discovery pertussis G proteins
various
subunits
TOXINS
that certain bacterial toxins (in parand cholera) can cova!ent!y modify sehas been invaluable in the delineation
structural
homologous and pertussis diphosphate
FOR BACTERIAL
and
functional
aspects
of
highly
G proteins. In brief, both cholera toxin toxin catalyze the transfer of an adenosine ribose (ADP-nibose) moiety onto the a
of various
G proteins.
These
ADP-nibosylation
reactions, and their subsequent effects on signal transduction, are believed to be responsible (at least in part) for the pathophysiology of the diarrhea and whooping cough caused by cholera toxin and pertussis toxin, respectively. A major difference between the stimu!atory
(Ga) and other
subtypes
of G proteins
is their
suscepti-
bility
to toxin-catalyzed ribosylation. The a subunits of G5 and G (transducin, the major retinal G protein) are substrates for ADP-nibosylation by cholera toxin (3, 13). G is not generally a target of cholera toxin; rather, it is preferentially ADP-ribosy!ated in an analogous manner by pertussis toxin. ADP-ribosy!ation by pertussis toxin appears to stabilize the G protein in its inactive ay undissociated conformation. Additionally, pertussis
toxin
thereby
“uncouples” attenuating
receptors
from
receptor-mediated
G1 (3,
7),
4,
inhibition
of
adenylyl cyclase. Studies shown that several G proteins sis toxin, namely, G1, G12, additional techniques have tween the various pertussis
to be used to distinguish betoxin substrates. Neverthe-
less,
of these
the
relative
selectivity
Am
using cDNA cloning have are substrates for pertus-
G3,
G,
J Psychiatry
and G0 (15).
toxins
I 49:6,
has
Thus,
resulted
June
1992
HUSSEINI
in their extensive use in identifying the presence and G, in various tissues, as well as the coupling ous receptors to these G proteins.
of G of van-
FIGURE
1. Mechanisms
K. MANJI
of G Protein Activation/Deactivationa Aonlst
High-Affinity
Ternary
Beta
Complex
Receptor
G PROTEIN
REGULATION
OF PHOSPHOLIPASE
The phosphoinositide pathway is a major signaltransducing pathway that has a ubiquitous role in the regulation of various aspects of cellular and neuronal function. The binding of hormones and neurotransmittens to a variety of cell surface receptors-e.g., a1-adrenergic, serotonin (5-HT1, 5-HT2), and muscaninic (M1, M3)-activates, through interposed G proteins, a phosphoinositide-specific phospholipase C (16). This enzyme hydrolyzes phosphatidyl inositol 4,5-bisphosphate (PIP2) to generate two second messengers: inosito! 1,4,5-trisphosphate (1P3), which mobilizes Ca2 from intracellular stores, and 1,2-diacylglycerol, which activates protein kinase C (16). Protein kinase C occupies a pivotal position in the biochemical pathways that relay information into the cell, because it is able to influence the cellular response to numerous other stimuli. Indeed, activation of protein kinase C has been linked to the
regulation
of cell
surface
receptors,
ion
elusive
G
proteins
coupled
to
stimulation
J Psychiatry
1 49:6,
June
1992
Subunit DissOciatiOn
IL__i
Subunit Reassociation
I
Affinity)
(Low
P1 GTPase
Intrinsic
-
GDP
Inactive
GTP Active
as
as
Adenyiate Cyclase
AlP
cAMP
M+
aAt
rest
an equilibrium
exists
between
the
receptor
in the
high-affinity
state (coupled to the G protein) and the low-affinity state. Activation of receptors by an agonist induces a conformational change in the receptor, allowing it to interact with the G protein, and forms a short-lived “high-affinity ternary complex” consisting of these components. The receptor-G protein interaction facilitates the replacement of GDP by GTP on the guanine nucleotide site (on the a subunit of the G protein). Binding of guanine nucleotides to the G protein causes a destabilization of the high-affinity complex and a dissociation of the G protein into a-GTP and 1y subunits. For most effector systems, the a-GTP complex activates the enzyme or ion channel. The continued activation is terminated by the action of a GTPase enzyme intrinsic to the a subunit, which hydrolyzes GTP to
GDP; the
the reassociation
formation
of the
cumulated pase
C,
that
of GDP-a inactive
similar
G proteins
with 1y completes
the cycle
with
G protein.
to the may
regulation
couple
various
of phospholireceptors
to
phosphatidylcholine by means of phospholipase (21 ). The existence of this signal transduction pathway has generated considerable interest, since it provides mechanistic explanation for the frequently reported temporal dissociation between production of diacylglyerol and 1P3 (21).
D a
of
phospholipase C in a pertussis-insensitive manner (12). Similar to the dual regulation of adenylyl cyclase, a number of receptors appear to inhibit phosphoinositide turnover, including receptors for adenosine, dopamine, S-HT, and glutamate (20). In addition, evidence has ac-
Am
GDP
Stimulatory G Protein Complex
channels,
secretion, gene expression, and neuronal plasticity (17). The details of the phosphoinositide turnover pathway have been outlined previously in an excellent review in the Journal ( 1 7) and will not be repeated here. The G protein that stimulates phosphoinositide turnover has been termed “Ge” (p=phosphoinositide), a!though its exact nature remains to be determined. What is clear, however, is that there appears to be more than one type of G protein activating phospholipase C, since pertussis toxin (which ADP-ribosylates and inactivates several G proteins [see preceding discussion]) inhibits agonist-stimulated PIP2 hydrolysis in some cell types but not others (16). Indeed, functionally distinct G proteins may even selectively couple different receptors to PIP2 hydrolysis in the same cell (18). Studies using Xenopus oocytes (large frog eggs frequently used in transfection studies) suggest that G0 rather than G may represent the pertussis-toxin-sensitive “Ge,” at least in certain cell types (16). The identification of a novel G protein termed “G7,” which is not a substrate for nibosylation by either cholera toxin or pertussis toxin, has led to the proposal that G may represent the “G” in a variety of tissues, including brain (where it is abundant) and platelets (19). More recently, however, several groups have purified and cloned members of a novel, ubiquitously distributed family of G proteins known as Gq. The a subunits of this family, including aq and , have unequivocally been shown to stimulate phospholipase C, making them leading candidates for the heretofore
(Low Affinity)
(High Affinity)
C
G PROTEIN
COUPLING
OF RECEPTORS
TO
ION
CHANNELS It has become increasingly large number of ion channels
clear in recent are regulated
years that a both directly
749
G PROTEINS
and indirectly by G proteins; at least 12 separate ion channels appear to be directly regulated (under cell free conditions) by G proteins (3, 22). In the CNS, a variety of receptors regulate neuronal K conductances by means of pertussis-toxin-sensitive G proteins. Activation of these K channels results in neuronal hyperpolanization and is an important element in the regulation of the neuronal firing rate (23). Although both a, and
a0 are able
to activate
inwardly
rectifying
K
currents
in neurons, a0 is markedly more potent. This is of considerable interest since a0 can be said to be predominantly localized in the nervous system (24), where its major role may be to regulate ion channels. Indeed, the most abundant G protein in the brain is G0, which cornpnises 0.S%-1.0% of the total membrane protein (24). Within the brain, G0 has a heterogenous distribution; it is particularly rich in the hippocampus and frontal contex but is also widely distributed in other areas of the brain (25). This distribution parallels that of protein kinase C (but not adenylyl cyclase) in the brain, leading to the suggestion that G0 may represent the pertussistoxin-sensitive “G” in the brain (25). In addition to regulating K channels, G proteins (in particular, G0) also directly modulate neuronal Ca2 channels (3, 22, 26, 27). Voltage-activated Ca2 channels are present throughout neurons and regulate a variety of cellular functions, including, of course, neurotransmitter release. A variety of neurotransmitter receptors (including a2-noradrenergic, GABAB, and adenosine Al ) inhibit neuronal voltage-sensitive Ca2 channels, and this receptor-stimulated modulation of Ca2 ion channels may be (at least in part) the molecular mechanism whereby nerve terminal autoreceptors inhibit neurotransmitter release (26, 27).
crease in Ga,, mRNA, suggesting in vivo physiologic regulation (33). It has also been demonstrated that pretreatment of mouse stniatal neurons in primary culture with 1 7 3-estradiol or testosterone increases the pertussis-toxin-catalyzed ADP-nibosylation of Ga0 and/on
Ga1 subunits
as evidenced
tion, and methasone
by
OF
G PROTEINS
BY
However,
it has
been
HORMONES
demonstrated
that
7-day
administration of corticosterone increases the transcniption and expression of Ga5 while decreasing both the mRNA and immunoreactivity of Gaj in rat cerebral cortex (33). Moreover, adrenalectomy without corticosterone replacement results in a significant 20% de-
750
ACTH
hypersecre-
of plasma cortisol after dexa(35). Similarly, at least some studies have shown the administration of thyroid hormones to be beneficial in the treatment of refractory depression and rapid-cycling bipolar disorder. Although a variety of biochemical effects may contribute to the CNS manifestations of abnormalities in thyroid or conticosteroid status, alterations of G protein function or content (with the inherent amplification of receptor responses) by these permissive hormones are an attractive mechanism. Similarly, the effects of gonadal hormones on G proteins, resulting in subsequent modification of signal transduction (34, 36), may be one mechanism by which biological maturation (puberty or menopause) triggers the expression of certain psychotic illnesses
cholic
(e.g.,
schizophrenia
depression
in late
in late
adolescence,
melan-
adulthood).
AS MEDIATORS
OF
NEUROTRANSMITfER-NEUROTRANSMITfER AND RECEPTOR-RECEPTOR INTERACTIONS
The CNS is a major target for the actions of glucocorticoids, thyroid hormones, and gonadal hormones, but the biochemical alterations ultimately responsible for producing the effects remain unclear. Malbon et a!. (28, 29) coined the term “permissive hormones” to describe agents such as glucocorticoids and thyroid hormones that modulate the actions of a variety of agents acting through cAMP. Increasing evidence suggests that these modulatory actions are exerted to a significant degree through G protein regulation. Thus, in vivo alterations of hormonal levels have been demonstrated to alter the steady-state levels of several G protein subunits and thereby regulate the overall sensitivity of transmembrane signaling in a variety of peripheral tissues (3032). Less information is available about the effects of hormonal manipulation on G protein regulation in the CNS.
hypercortisolemia,
nonsuppression administration
G PROTEINS REGULATION
(34).
Abundant evidence supports an interaction between thyroid and adrenal hormones and psychiatric illness, particularly in affective disorders. Thus, primary disordens of both the thyroid and hypothalamic-pituitaryadrenal (HPA) axis have been linked with depressive, manic, and anxiety symptoms. Additionally, there is general agreement that more than 50% of patients with major depression exhibit hyperactivity of the HPA axis,
ple
Most clinical measures
terms
studies in psychiatry, are obtained, analyze
of independent
measures,
e.g.,
even data
when multiprimarily in
Is norepinephnine
deficient? Is serotonin deficient? However, considerable preclinical evidence shows that monoamine systems interact, and a major question for neuroscience is emenging with regard to elucidating the mechanism(s) by which one neurotransmitten influences the response of a neuron to all the other converging afferent inputs. The CNS is extremely complex, both anatomically and chemically, and there is a remarkable convergence of different receptors in common cortical layers (37) and a considerable convergence of neurotransmitter action (38). A single neuron in the brain receives thousands of synaptic inputs on the cell body and dendrites, and neuronal response is also modulated by a variety of hormonal and neurohormonal substances not dependent on synaptic organization. The neuron needs to integrate all the synaptic and nonsynaptic inputs impinging on it; this integration of a multitude of signals determines the ultimate excitability, firing pattern, and response characteristics of the neuron, and the neuron’s
Am
J
Psychiatry
1 49:6,
June
1992
HUSSEINI
response is then conveyed to succeeding means of synaptic transmission. How does neuron decipher and integrate the multitude it receives and, additionally, generate unique to each
of these
signals
or combinations
of signals?
do G proteins
form work
the basis of a complex information processing netin the plasma membrane (39). Thus, the ability of
to interact
an elegant
signals,
with
mechanism
they
multiple
also
appear
receptors
to organize
the
signals
from
messenger
. NE
according
to
their
G protein
and
inte-
pathway.
Similarly, the dual (positive and negative) regulation of adenylyl cyclase and perhaps of phospholipase C by G proteins allows stimulatony and inhibitory signals for these pathways to be balanced at the G protein level, yielding an integrated, dampened output (39). Convergence
of a variety
of neurotransmitters
on
the
same
Alpha-2 Blockade
A,
and
and
same
the
hippocampus,
K-opiate pool
through creasing
receptors
of
G
is mutual
appear
proteins
voltage-sensitive norepinephnine
there
a2-adrenergic, to
modulate
Ca2 release
antagonism
Ca2
channels, (44-46).
for
these
de-
2).
Similarly,
heterologous
of desensitization
agent leads of stimuli, nisms) may protect the stimulation provide the receptors to tions linking with
signals
exposure
to
(a form a desensitizing
to diminished responsiveness to a number apparently through postreceptor mechabe a compensatory mechanism designed to neuron from the deleterious effects of overby multiple receptors. Thus, G proteins first opportunity for signals from different be integrated; this complex web of interacreceptors, G proteins, and their effectons converging to shared detectors appears to
be crucial
for the
CNS
47,
(39,
desensitization
whereby
integrative
functions
Kappa opiate
receptors
with respect to modulating both Ca2 influx and nonepinephnine release. That is, stimulating one of these receptors diminishes the effectiveness of the other receptons; this may be a mechanism for allowing neurons to “escape” from excessive inhibitory input (46) (see figure
NE
the
influx
thereby Additionally,
between
Alpha-2 Stimulation
adenosine
to compete
NE
ion
channel has been demonstrated to occur in the locus coeruleus, hippocampus, thalamus, and substantia nigra (38, 40-43). Thus, it appears that a variety of receptons which coexist on neurons and mediate similar responses may share signal transduction mechanisms (38, 40, 43, 44). This convergence may occur at the level of the G protein or the effector itself. In both the cortex
performed
by the
48).
aReceptors
that
coexist
EFFECTOR
CROSS-TALK
In addition
G proteins between
Am
also various
J Psychiatry
OF RECEPTOR-
IN THE
to their role as receptor-effector serve as targets for regulating second messenger systems.
I 49:6,
June
1992
mediate
number of potential sites of interaction ase C with the cAMP-generating system ure
3).
rylation
A large of the
body
of evidence
3-adrenergic
similar
responses
biochemical
effect of -adrenergic
by cross-talk evidence
that
phospho-
by various
kinases,
desensitizes the receptor that the most commonly of antidepressants, receptors,
through
protein
has
accumulated
also
of protein kin(49, 50; see fig-
suggests
receptor
including protein kinase C, (50). Indeed, evidence suggests
Increasing
couplers, cross-talk There are a
and
possibly by making available a greater proportion of the G protein pool. In an analogous manner, stimulation of a2 (bottom) diminishes the effectiveness ofthe other receptors a disproportionate use of the common pool of G proteins. This may be a mechanism for allowing a neuron to “escape” from excessive inhibitory input.
diated
CNS
neurons
(middle), shared receptors through
down-regulation
AS MEDIATORS
on
may share signal transduction mechanisms. The top portion of the figure depicts a neuron with a,-adrenergic, adenosine A,, and K-opiate receptors all converging onto a common pool of G-like proteins to inhibit Ca2 influx and norepinephrine (NE) release. Alpha-2-adrenergic antagonism increases the effectiveness of the other receptors
observed
G PROTEINS
by a Corn-
these
grated
second
Interactions
provides
them to a relatively Signals from a van-
a single
of Receptor-Receptor Systerna
to
multiple receptors and to transmit much smaller number of effectors. ety of receptors can be “weighted” intrinsic ability to activate a given to stimulate
FIGURE 2. Integration mon G Protein-Effector
Not
only
G proteins
amplify
targets by the single of signals responses
K. MANJI
kinase
namely, may
be me-
C (51, that
the
52). po-
tentiating effects of protein kinase C on cAMP accumulation are mediated by G. Thus, in a number of cell types, including hepatocytes and stniatal membranes (30, 53-55), activation of protein kinase C appears to
751
G PROTEINS
FIGURE 3. Mediation G Proteinsa
of Receptor-Effector
Cross-Talk
in the CNS by
duction pathways, it is not surprising that abnormalities in the function and/or expression of various G proteins have been implicated in a variety of pathophysio!ogic states: pseudohypoparathynoidism type I, heart failure, certain endocrine tumors, McCune-Albright syndrome, diabetes, alcoholism, schizophrenia, mitral valve prolapse, chronic cocaine/opiate ingestion, aging, hypo/hyperthyroidism, and adrena!ectomy/corticosteroid administration (58-81). It should be noted that G protein dysfunction appears to be the primary pathology in Albright’s hereditary osteodystrophy, McCune-Albnight syndrome, and endocrine tumors; in several of the other conditions, G protein abnormalities are probably secondary but are nonetheless implicated in the pathophysiology of the condition. I will briefly highlight some of the clinical conditions with well-characterized
G protein dysfunction. The first disease in which Calcium
normality was called because
Mobilization
ma! (or even aThe
figure
depicts
ceptors
activating
ceptors
coupled
pholipase
Ca2
the
C results protein
interaction
intracellular
the phosphoinositide
to adenylyl in the
from intracellular
tivates
complex
kinase
of protein
cyclase. generation
stores, C. There
kinase
Protein kinase an uncoupling rylate adenylyl
C phosphorylates of the receptor
on differences
among
cross-talk
turnover
Receptor
of IP3 (IP3),
are a number
of protein
kinase
subtypes
which
(DAG),
of potential
re-
and re-
of phos-
sites
ac-
for the
system.
the 3-adrenergic receptor, causing from G,. It also appears to phospho-
of G proteins
of adenylyl cycit appears to phosits inhibitory tone. probably depends
and the relative
abun-
C isozymes.
enhance adenyly! cyclase activity by attenuating the inhibitory influence of G. Since the susceptibility of G proteins to be phosphorylated (and therefore regulated) by protein kinase C depends in large part on their conformational state (56), the degree of cross-talk is negulated by the relative degree of simultaneous stimulation of various receptors. In sum, it is clear that interactions between distinct second-messenger-generating systems represent a fine-tuning cellular network that regulates the neuron’s reactions to the large number of extracellular signals it encounters. The net effect of the various potential interactions probably depends on the summation of the effects on individual components. In this context, quantitative and qualitative (e.g., conformation states) differences among subtypes of G proteins and the relative abundance of protein kinase C isozymes (57) in various cells may be major factors in determining the final integrated output.
CLINICAL
752
their widespread regulation, and
and crucial amplification
role in the inteof signal trans-
ab-
have
so nor-
than
normal)
levels
of parathyroid
demonstrate
a resistance
to parathynoid
hormone and to a number of hormones using cAMP as the second messenger. In these patients, lower than norma! expression or function (approximately 50%) of Gcx, has been demonstrated in cultured cells and in cells from freshly obtained tissue (58 and references therein). Additionally, distinct mutations have been identified in the gene encoding Ga in different kindreds with the disease (58, 59), thereby providing the first demonstrations of an inherited mutation in a human G protein gene with important implications. The existence of more than one type of mutation resulting in abnormal function of Gas, demonstrates the genetic heterogeneity of this autosoma! dominant disease. Moreover, the presence of an identical mutation in the Ga,, gene, both in individuals with multiple hormone resistance and in those without hormone resistance, clearly suggests that
G5 deficiency
is necessary
but not sufficient
for the full
phenotypic expression of the disease (59). Thus, as has been suggested for a variety of psychiatric disorders, additional factors, including modifying genes and perhaps environmental factors, appear to be involved. Studies have also demonstrated substantial abnormalities in G proteins in failing human and animal heart. A 50% lower than normal apparent concentnation and function of G in sarcolemma from a canine model of left ventricular failure has been reported. In contrast, end-stage idiopathic human congestive failure is associated with greater than normal activity of G, suggesting that various manipulations of G/G1 stoichiometry may result in similar pathophysiology (60). Perhaps more akin to psychiatric research with respect cessible
Given gration,
higher
to the use of peripheral
IMPLICATIONS
G protein
hormone but have the manifestations of the disease because of cellular resistance to the hormone. In one form of the illness (Albright’s hereditary osteodystrophy),
the patients
mobilizes
which
C with the cAMP-generating
cyclase, resulting in an enhancement lase activity. Finally, at least in certain systems, phorylate and inactivate G, thereby removing The net effect of the various potential interactions dance
pathway
activation
and diacylglycerol
between
an intrinsic
found is pseudohypoparathyroidism, individuals with this condition
80% jects
tissue
lower with
blood
(CNS,
than congestive
cells to represent
heart),
normal
study
lymphocyte
heart
Am
one
failure.
J
Psychiatry
(60)
less acshowed
G5 levels More
intriguing,
1 49:6,
June
in subsuc-
1992
HUSSEINI
cessful treatment with captopnil was associated significant two-fold increase in lymphocyte G
The following ciation
are summaries
of G proteins
of research
to psychiatric
with levels.
a
on the asso-
conditions.
The evidence for genetic factors in alcoholism is compelling (62), and, given the widespread medical, behavioral, and societal manifestations of this disorder, it is not surprising that there has been considerable research
factors
that may predispose
individuals
to
alcoholism. A variety of putative electrophysiologic, biochemical, and neuroendocrine markers have been studied, but increasing research has focused on G proteins not only as the mediators of the biochemical effects of alcohol but also as possible sites of underlying pathophysiology in individuals predisposed to the development of alcoholism (63). Ethanol can profoundly affect the G protein-adenylyl cyclase signal transduction pathway (63). Chronic ethanol ingestion or in vitro exposure results in a desensitization of cAMP production in a variety of tissues, including brain (63, 64). The heterologous nature of desensitization of the adenylyl cyclase activity following chronic in vitro exposure of neuroblastoma cells to ethanol suggests a postreceptor (e.g., G,) site. Chronic exposure of cultured cell lines to ethanol results in a more than 30% decrease in both the a protein and its mRNA (65, 66). In mice chronically fed ethanol, there is a reduction in the cholera-toxin-catalyzed [32P]ADPribosylation of G and a reduction of high-affinity [3H]forskolin binding (which is thought to reflect binding to the complex of a and adenylyl cyclase) (63). Clinical studies have also implicated G-adenylyl cyclase as a putative marker in alcoholism. One study showed that basal and adenosine-receptor-stimulated cAMP 1evels
in fresh
lymphocytes
of alcoholic
subjects
were
75%
lower than those of normal subjects or subjects with nonalcoholic liver disease (67). Cultured lymphocytes from alcoholics, grown in the absence of alcohol, also show some abnormalities in the cAMP signal transduction pathway (64). Similarly, platelet membranes of alcoholics reportedly abstinent from alcohol for 12-48 months demonstrate lower receptorand G5-stimulated adenylyl cyclase activity than do platelet membranes of age- and sex-matched comparison subjects (68). These findings suggest that low platelet adenylyl cyclase activity may be
a genetic by studies
marker by my
in alcoholics; colleagues
tussis-toxin-catalyzed
this finding and
me
of cholera-
[32P]ADP-nibosylation
G, and G, respectively,
in platelet
is supported and
per-
of platelet
membranes
from
ab-
stinent alcoholics, subjects with concomitant diagnoses of major depression and alcoholism (major depression as the primary diagnosis), and age- and sex-matched companison subjects. We observed significantly less 32P-labeling of G (but not G#{149}) in the alcoholic group than in the comparison subjects. These findings are entirely consistent
with
koff
et a!. already
Am
the
J Psychiatry
platelet
adenylyl
described
1 49:6,
cyclase
(68).
June
1992
findings
Opiate/Cocaine
To my knowledge,
Use there have been no clinical
studies
of the gestion clinical
Alcoholism
on biological
Chronic
K. MANJI
by Taba-
effects of acute or chronic opiate or cocaine inon G proteins. Nevertheless, considerable preevidence suggests that these substances exert at least some of their chronic effects (e.g., tolerance, dependence, withdrawal) through G proteins. A variety of opiate receptors are coupled to their effectors by G proteins, and pertussis toxin uncouples opiate receptors from the inhibition of adenylyl cyclase in cultured cells and in locus coeruleus neurons (3). Studies have shown that alterations in G protein levels are temporally conrelated with neurona! activity in the locus coeruleus (more firing) and with the behavioral manifestations of the morphine withdrawal syndrome in rats (69, 70). These results suggest that alterations in the levels of G proteins may underlie the enhancement of neuronal excitability (and possibly withdrawal symptoms) observed after abrupt opiate discontinuation. In vitro studies of cultured neurons from rat spinal cord dorsal root ganglia have demonstrated that exposure to K-opiate agonists is accompanied by a 60%-70% decrease in levels of aj, in the absence of alterations in levels of a,
a0, and
subunits
(71),
thereby
providing
tic explanation for opiate-induced sitization. In contrast to the studies recent study has directly examined
G proteins, ventral coeruleus
demonstrating
tegmental in rat
low levels
a mechanis-
heterologous on opiates, cocaine’s
desenonly one effects on
of a and a0 in the
area, nucleus accumbens, brain after chronic cocaine
and locus use (72).
Schizophrenia Several years ago, studies brain tissue demonstrated ylyl
cyclase
activity
using post-mortem a greaten response
to fluoride
stimulation
human of aden(presumably
acting through G) in the caudate and accumbens of schizophrenic patients than in the corresponding brain areas of comparison subjects (73). These studies suggested that a postreceptor defect may contribute to the presumed dopaminergic hyperactivity in schizophrenia, but it was several more years before they received additional, more direct support. Seeman et al. (74) used radioligand binding studies to explore the D1-D2 link in post-mortem human brain tissues. They proposed that the D1 receptor may modulate ligand binding to the D2 receptor by modifying the levels of G protein subunits
(1)
that these
receptors
share
(see figure
4), and they
reported
that the D1-D2 link was missing in more than 50% of tissue samples from patients with schizophrenia and Huntington’s disease, a finding apparently not due to prior neuroleptic use. In a study that examined G protein levels in post-mortem schizophrenic brains more directly, pertussis-toxin-catalyzed nibosylation in the left putamen was 42% lower in schizophrenic patients than in comparison subjects (75). These findings (suggesting low G1 and/or G0) would, of course, be compatible with enhanced dopamine-adenylyl cyc!ase function. In keeping with these lateralized abnormali-
753
G PROTEINS
FIGURE
4. Reciprocal
Modulation of Receptor Affinity by G Proteinsa
ribosylation
of GIG0
sent
1MB
link
D1-D2
Two
studies
jects
GDP
inactiveai
as
cussion). strated
02 AgonIst
U
in depressed
84),
abnormalities
ing
these
allows
further
interac-
released may drive the equilibrium for another G protoward the undissociated state and thereby modulate receptor affinity (in this case, the D, receptor). This may be one mechanism for the observed synergism between the D, and D2 receptors despite opposite effects on adenylyl cyclase. (e.g.,
thus G)
For
ties
in postreceptor
pathways
is thought adenylyl
to
reflect
cyclase,
in schizophrenic
binding higher
than
patients,
[3H]forskolin bindparahippocampal gyrus brains of schizophrenic [3H]forskolin binding to
the normal
complex levels
of of
a and a, and!
or adenylyl cyclase would be the simplest explanation for these results. However, low levels of G1 3y subunits would also result in a greater availability of free
a
to interact
with
adenylyl
cyclase.
Indeed,
although
this suggestion is highly speculative, a low level of y subunits would be compatible with the higher than norma! adenylyl cyclase activity (73), lower pertussis toxin
754
platelet
manic-
[3HJ-
are
to
in reference interpret
cAMP
levels)
lower
lymphocyte
accumulation
(85).
demonstrated
and
in in dewas
Similarly,
a significant
correlations
between
release and basal cyclase activity
platelets
(unpublished
urinary and postin both
observations).
It is presently unclear whether these abnormalities receptor and postreceptor sensitivity are primary the sequelae of abnormalities in circulating levels echolamines cultured
and/or cells
needed
ob-
patients with high circulat-
terminal insomnia and blunted accumulation in depressed patients (86). Additionally, my colleagues
measures of norepinephnine receptor-stimulated adenylyl lymphocytes
given
documenting abnormalities and glucocorticoids
in a subgroup of depressed agitation (and presumably
I have
rereceptor
(reviewed difficult
significantly
between cAMP observed
demon-
a2-adrenergic
patients
example,
from
to examine
confounding
greater than normal high-affinity ing has been found in the left and CA 1 region in post-mortem patients (76 ). Since high-affinity
from
isoproterenol-stimulated
literature catecholamines
catecholamine
and
tion of G proteins. When receptors coexist on neurons, a potential mechanism for modulation of receptor affinity by other receptors exists. Binding of an agonist to a receptor (e.g., D1 receptor) promotes the dissociation of the G protein into a and ly subunits. The tein
and
sensitivity
correlation lymphocyte has been
of 3y subunits
tissue
a number of studies have in lymphocyte -adrenergic
sensitivity
served only psychomotor
D2 Receptor
13’)’units
brain
isoproterenol-stimulated
(Low affinity)
interchangeability
G proteins in afreported that the
significantly
by
Although abnormalities
pression. cii
or ab-
higher than those These findings are intriguing of Schreiber et al. (83), who G proteins in manic sub-
were
determined
the extensive circulating
‘p affuiity)
minimal
examined al. (82)
et
in post-mortem subjects
ceptor
(High
be fewer
GTPyS binding in leukocytes. Although these findings need to be replicated, they are exciting since lithium is reported to attenuate G protein function (see later dis-
)GDP
GD$
relative
directly
of comparison subjects. since they parallel those reported “hyperfunctional”
/
would
heterotnimers-the
and greater [3H]forskolin binding from schizophrenic subjects (76).
Young
of Ga
depressive
aThe
have
disorders.
levels
D1 Agonist
GTP#{231})
there
Disorders
fective
Active
(74), tissue
in post-mortem Affective
as
(since
subunits available to form ay substrate for the nibosylation reaction),
D1 Receptor
inactive
(73)
1-
stress these
hormones. patient
possible
effects
of
Thus,
populations
G protein circulating
defects
in or are of cat-
studies are
of
clearly
free
of the
catecholamines
and
hormones. Similarly, a common postreceptor abnormality (e.g., at the level of G proteins) would provide a mechanism for depressed patients’ blunted growth hormone and prolactin responses to a variety of provocative
challenges
(35),
but
such
an
abnormality
remains
to be established. Other
Neuropsychiatric
Conditions
Lower than normal levels of platelet basal and postreceptor-stimulated adenylyl cyclase activity have been reported in both panic disorder (87) and posttraumatic stress
disorder
possibility dary
(88).
that
to abnormalities
these
Once
again,
platelet in plasma
Am
J
one
is faced
abnormalities
with
are
catecholamines
Psychiatry
1 49:6,
the
seconand/or
June
1992
1-LUSSEINI
stress
hormones.
symptoms
Another
illness
in common
hyperadrenergic
with
with
panic
dysautonomia
a constellation
disorder
is a form
associated
with
of
duced
[3H]GTPyS
of
manic
patients
mitral
valve prolapse (77). Elegant reconstitution studies have recently demonstrated that upon reconstitution into cyc lymphoma cells, erythrocyte G, from patients with mitral valve prolapse is hypenresponsive with respect to both adenylyl cyclase activation and receptor coupling (78). If a similar G abnormality is present centrally in panic disorder, it would provide an attractive mechanistic explanation (i.e., desensitization of 3-adrenengicreceptor-coupled adenylyl cyclase) for the therapeutic efficacy of both tnicyclic antidepressants and monoamine oxidase inhibitors in this condition. Finally, abnormalities in G proteins have also been implicated
in
disease
(80),
Huntington’s
and
G PROTEINS
disease
aging
(79),
Alzheimer’s
(81).
AS TARGETS
OF PSYCHOTROPIC
DRUGS
Lithium Although
lithium
recurrent its
is widely
affective
disorders,
mood-stabilizing
ever,
system
depletion
suggests
and
the
(since
blood-brain
barrier,
phosphatase phoinositide
could deplete turnover) (91,
sive
research
on
cyclase. In rat vivo (i.e., after has
been
adenylyl pressants,
exerts sys-
of the
does
not
inhibition
of
inosito!
the
inositol-1-
inositol and reduce 92). This hypothesis
phoshas
shared
namely, reported
Am
J
effects
attenuate
on
neuronal
adenylyl
both in vitro ofthe animal), receptor-
Gpp(NH)p,
and
and ex lithium
postrecep-
fluoride,
(93-96). In contrast has been reported
by these
two
signal
forskolin]
to most to induce
transduction
In this context, attenuate the binding
and carbachol et a!. (83)
Psychiatry
lithium
delays
antidefew, if
in response
(97). observed
1 49:6,
June
Using greater
1 992
systems,
lithium has agonist-induced to
both
been in-
isopno-
a similar method, isoproterenol-in-
from
subjects
with
lithium.
activation
untreated
In renal
of
bi-
or euthymic
epithelial
adenylyl
cyclase
by
Gpp(NH)p, which is compatible with attenuation of the formation of the active Gpp(NH)p-cx subunit (98). Similanly, the inhibitory effects of chronic lithium use on rat brain adenylyl cyclase are reversed by increasing concentrations of GTP (93, 94). Taken together, these results
suggest
that
the
physiologically
relevant
effects
of
lithium on adenylyl cyclase may be exerted at the level of G proteins (presumably at a GTP-responsive step). My colleagues and I have recently demonstrated significant increases in basal and postreceptor stimulated adenyly! cyclase activity in platelets (but not lymphocytes) obtained from normal volunteers after 14 days of lithium administration (95, 96). Since inactivation of G appears to exert significant effects on adenylyl cyclase activity in platelets but not in lymphocytes (96), we postulated that the striking tissue-specific effects could be explained by a lithium-induced attenuation of G function. To examine this more directly, we used platelet membranes from these same subjects to quantitate G proteins by using specific lyzed [32P}ADP-nibosylation.
antibodies We did
and
toxin-cata-
not observe any abnormalities in immunolabeling of platelet G but did observe a significant 37% increase in pertussis-toxincatalyzed [32P]ADP-ribosylation of platelet G. Since the undissociated, inactive afry heterotnimeric form of gest
that
for pertussis
lithium
inactivates
toxin,
G
by
our results
stabilizing
the
sugmac-
tive, undissociated conformation. Corroborating our human platelet findings, we have recently observed similar increases in pentussis-toxin-catalyzed [32P]ADPnibosylation in rat cortex after 4 weeks of lithium use but no alterations in the immunolabeling of G, G1, or
G0. Similar
Whatever
in [3H]GTP
terenol Schneiber
cells,
lithium’s
G proteins. to markedly
crease
stabilized
question, however, since alterations in have not been consistently found after (see 92). There has also been exten-
any, changes in the density of rat brain -adrenergic receptors. Lithium also attenuates receptor-stimulated phosphoinositide turnover in rat brain and in human platelets (reviewed in references 92 and 96). Since lithium affects both phosphoinositide turnover and adenylyl cyclase, recent attention has focused on mechanisms
patients
G1 is the substrate
cross
in leukocytes
in normal
polar
laboratory mechanism rently under with Mg2
[GTP, cyclase lithium
for
How-
lithium
generation
lithium
to
ton-stimulated
that
inositol
brain preparations, chronic treatment
shown
basis
unknown.
inhibits the activity of the en(89, 90) has resulted in exthe phosphatidylinositol signal
hypothesis
been called into brain PIP2 levels lithium treatment
of
second-messenger-generating that lithium, at therapeutically
relevant concentrations, zyme inositol-1-phosphatase tensive research on transduction
treatment
biochemical
remains
evidence
effects on discovery
for the
the
actions
accumulating
substantial tems. The
used
binding than
K. MANJI
thus
findings
been
the mechanism,
multiple
cyclase,
have
reported
by another
(R.S. Jope, personal communication). The by which lithium may inactivate G is curinvestigation and may involve competition on cross-talk with protein kinase C (99). signal
by affecting
transduction
phosphoinositide
G proteins
mechanisms
turnover)
and
(adenylyl
common
to
many
different neurotransmitters, lithium may be in a unique position to affect the functional balance between neurotransmitter systems. One might speculate that it is precisely this effect on the functional balance between interacting neurotransmitter systems that underlies lithium’s mood-stabilizing effects. Antidepressants Despite nisms
of
extensive action
research,
of antidepressant
the
molecular drugs
mecha-
have
not
been
clearly established. The most commonly observed biochemical effect of most clinically effective antidepressant treatments is a reduction in the norepinephninestimulated production of cAMP (desensitization), usually
(but
not
invariably)
accompanied
by
a reduc-
755
G PROTEINS
tion
in
the
number
of
3-adnenengic
receptors
(down-
CONCLUSIONS
regulation) (100). This has generally been regarded as an adaptation to the elevation of intrasynaptic norepinephnine through pnesynaptic mechanisms (e.g., reuptake blockade or monoamine oxidase inhibition). However, a direct postsynaptic effect of these drugs has
been
vitro
suggested
glioma
of
cells
(102,
to
receptor
have
attempted
sants
at the
to cause
G
by the
exposure
103)
to level
a functional state
adenylyl
to tnicyclic
cyclase,
effects of the
interferes
perhaps
f
receptor
with
(103,
105).
from
the high-afactivation
decreasing
the
Okada
affinity
of
et a!. (106)
by pertussis
sensitizing
toxin
simply
masks
desipramine’s
de-
In studies using high in vitro jtM), it has been demonstrated
concentrations that a variety
(50-300
of antidepressants interfere with G activation of adenylyl cyclase (107). Interestingly, greater inhibition was observed when the antidepressants were added before, rather than after, the addition of GTP; this finding is also compatible with an antidepressant-induced inhibition of G protein dissociation. However, given the high doses of antidepressants used, the physiologic relevance of these findings remains unclean. It was reported (in an abstract) that chronic administration of imipramine decreased
the
levels
ADP-ribosylation, These results are of some (109-111) gesting
of Ga
immunolabeling,
cholera
toxin
and Ga, mRNA in rat brain (108). difficult to reconcile with the results but not all (104, 112) studies, sug-
increased
postreceptor-stimulated
adenylyl
cyclase activity in rat cortical membranes after administration of antidepressants. Nevertheless, it remains possible that antidepressants attenuate -adrenergicmediated activation of G, while enhancing the effects of agents operating by pathways independent of the adrenergic receptor. Like most antidepressants,
(ECS) cyclase
desensitizes activity
the in rat
electroconvulsive
f3-adrenergic-mediated cortex
and
shock
adenylyl
hippocampus
(100).
Fewer studies have investigated the effects of repeated ECS on postreceptor sites. However, repeated ECS is reported to decrease both the agonist-induced [3H]GTP (113) and [3H]forskolin (114) binding in rat cortex. In a study using [35S]GTPyS binding (115), repeated ECS was
found
hippocampus to increase
756
mechanisms
to decrease
binding
in prefrontal
cortex
but to have no effect in the stniatum I35SIGTP’yS binding in the amygdala.
and
and
specific
clearly
come
the
a long
way
family.
The
neuronal
in the
past
diversity
and
has
clear
decade.
In particu-
and
family
serve the critical role extracellularly genertransmitting
these
inte-
forming the basis for a network (39). of G protein in a given the influence of physipharmacologic
implications
both
of the G protein
then
pathophysiologic,
tions,
of a neurotransmit-
progress in defining of the signal-transducing
grated signals to effectors, thus complex information processing The finding that the amount tissue is not static, but is under ologic,
a
the
surface is translated into a a physiologic) effect has
The G proteins or attenuating
signals
plays
Understanding
binding
lan, there has been dramatic the structure and the function continues to grow. of first amplifying
receptors
activation.
by which
not
only
perturba-
for
research
into
the etiology of various psychiatric conditions but also for the development of better treatments. Abnormalities in G protein function have now been identified in the etiology/pathophysiology of a variety of medical diseases. With the exception of chronic alcohol exposure,
opiate
neuroleptic
effects.
through
in neuronal
ten to a receptor at the cell biochemical (and ultimately
ated of
reported that pertussis toxin treatment of rats overcomes desipramine-induced receptor desensitization (as assessed by measurements of isoproterenol-stimulated adenylyl cyclase activity) without attenuating (in fact, promoting) desipramine-induced receptor downregulation. While these results may suggest that G and/or G0 are involved in desipramine’s effects, it is equally plausible that the removal of the inhibitory
tone
transduction
role
G protein
appears
subsequent
by
nucleotides
leads
of antidepres-
uncoupling and
C6
investigators
Desipramine
breakdown
G for guanine
the
and
in
and
antidepressants
G proteins.
(104)
chronic
(101)
Other
examine of
that
fibroblasts
down-regulation.
in rat cortex
finity
observations
human
Signal
central
withdrawal,
and
exposure,
the
dysfunction in psychiatric rect. Nevertheless, given
of G proteins seems
likely
yield
insights
in the that
schizophrenia!
for
postreceptor
disorders is currently the widespread, critical
regulation
of neuronal
future,
more
the
involvement
into
perhaps
evidence
less
diroles
function,
sophisticated
it
studies
of G proteins
will in the
pathogenesis of various major psychiatric illnesses. These putative G protein abnormalities may be subtle but sufficient to modify G protein functioning in response to neurotransmitter changes, permissive hormones, or other environmental events or stressors. Finally, given the increasing evidence that the currently available psychotropic drugs affect G proteins, the development of novel drugs with primary G protein targets
remains
an exciting
prospect
for
the
future.
Indeed,
it is intriguing that a number of the currently pharmacologic agents, although not developed appear to modify physiologic events that are
G proteins.
These
include
hepanin
available as such, linked to
derivatives
(116),
volatile anesthetics ( 1 1 7), and caffeine-like agents (118). It has also been demonstrated that a variety of cationic amphiphilic neuropeptides activate G proteins in a receptor-independent manner (119); they represent a novel class of agents (receptomimetics, rather than receptor agonists). Interestingly, tricyclic antidepressants are also cationic amphiphilic compounds (95, 120), so they
too
tion
to
may allow
adopt for
a membrane-spanning an
interaction
with
conformathe
guanine
cleotide binding site on the G protein. Recent have also identified a novel class of agents that allostenic enhancers and nist-preferring conformation
stabilize the high-affinity agoof the adenosine A1 recep-
ton (121). This may be viewed manner in which benzodiazepines
Am
nu-
reports serve as
J
as being analogous to the potentiate GABA-er-
Psychiatry
1 49:6,
June
1992
HUSSEINI
gic
neurotransmission.
Such
agents
have
a built-in
safety mechanism, since the maximal effect of the drug depends on the availability of the endogenous ligand. This may be one mechanism underlying the relative safety of benzodiazepines in overdose. Additionally, the use of such allostenic modulators may serve to effectively re-equilibrate dysregulated neuronal systems (which
are
likely
without the tachyphylaxis sensitivity.
to be involved
commonly (receptor Finally,
in psychiatric
encountered down-regulation)
given
the
fact
that
a fixed niques stimulate
phenomena and
super-
GTPase
activ-
the
of
as a “clock,” allowing the system to cycle at rate, it seems likely that molecular biologic techwill lead to the development of agents that either or inhibit
GTPase
by altering
Baraban
18.
and psychoactive drug action: focus on the phosphoinositide system and lithium. AmJ Psychiatry 1989; 146:1251-1260 Ashkenazi A, Peralta EG, Winslow JW, Ramachandran J, Capon DJ: Functionally distinct G proteins selectively couple different receptors to P1 hydrolysis in the same cell. Cell 1989;
its catalytic
rate,
thereby dampening or amplifying the hormonal (without constitutively activating the G protein). Appendix 1 provides more specific information methods for G protein research.
signal
19.
56:487-493 Casey PJ, Fong cleotide-binding
20. 21.
Exton
22. 23. 24.
25.
2131 I 1 . Birnbaumer ofeffector
1989;
13. 14.
15.
16.
Am
Psychiatry
June
1992
Signaling
Brown
AM,
Birnbaumer
by G protein
subunits.
that
inhibit
Sci 1989; phosphatidylcholine
through
1990;
340:
a guanine properties.
nu-
J
phosphoinosi10:114-120 breakdown.
265:1-4 L: Ionic
Annu
channels
Rev Physiol
and their regulation 1990; 52:197-213 and neuronal function. ofthe
Go protein.
PF, BarabanJM, Snyder SH: Beyond receptors: systems in brain. Ann Neurol 1 987;
Soc
multiple 21:217-
modulation of calcium currents in neu1990; 52:243-255 5, Tsien RW: Alpha-adrenergic inhineurotransmitter release mediated by calcium-channel gating. Nature 1989;
639-642
Ros M, Northup JK, Malbon CC: teins and beta-adrenergic receptors
29.
effects
ofthyroid
Malbon
CC,
hormones. Rapiejko
PJ,
30.
31
.
Steady-state levels of G-proin rat fat cells: permissive
J Biol Chem Watkins
regulation of hormone-sensitive macol Sci 1988; 9:33-36
1988;
DC:
effector
Houslay MD, Milligan G: G-Proteins Signalling Processes. New York, John Levine MA, Feldman AM, Robishaw
263:4362-4368
Permissive
hormone
systems.
Trends
Phar-
as Mediators of Cellular Wiley & Sons, 1990 JD, Ladenson PW, Ahn
TG, MoroneyJF, Smallwood PM: Influence ofthyroid hormone status on expression ofgenes encoding G protein subunits in the rat heart. J Biol Chem 1990; 265:3553-3560 32.
Ros
M,
Northup
adenylate
cyclase:
JK,
Malbon CC: Adipocyte of adrenalectomy.
G-proteins Biochem
effects
J
and I 989;
257:737-744
33.
34.
Saito N, Guitart X, Hayward M, Tallman JF, Duman RS, Nestier EJ: Corticosterone differentially regulates the expression of Gs alpha and Gi alpha messenger RNA and protein in rat cerebral cortex. Proc NatI Acad Sci USA 1989; 86:3906-3910 Maus M, Homburger V, Bockaert J, Glowinski J, Premont J: Pretreatment of mouse striatal neurons in primary culture with 17 beta-estradiol
ribosylation
enhances
ofG
alpha
the
pertussis
o,i protein
toxin-catalyzed
subunits.
ADP-
J Neurochem
1990;
SS:1244-12S1 35.
Nemeroff
36.
ogy. New York, Maus M, Bertrand
CB: Handbook
Glowinski
L: Transduction of receptor signal into modulation activity by G proteins: the first 20 years or so. FASEB
1 49:6,
JH:
Dolphin AC: G protein rons. Annu Rev Physiol Lipscombe D, Kongsamut bition of sympathetic modulation of N-type
28.
3:2125-
Strathmann M, Simon MI: G protein diversity: a distinct class of alpha subunits is present in vertebrates and invertebrates. Proc Natl Acad Sci USA 1990; 87:9113-9117 Gilman AG: The Albert Lasker Medical Awards: G proteins and regulation of adenylyl cyclase. JAMA 1989; 262:1819-1825 Hildebrandt JD, Kohnken RE: Hormone inhibition of adenylyl cyclase: differences in the mechanisms for inhibition by hormones and G protein beta gamma. J Biol Chem 1990; 265: 9825-9830 Jones DT, Reed RR: Molecular cloning of five GTP-binding protein cDNA species from rat olfactory neuroepithelium. J Biol Chem 1987; 262:14241-14249 Michell RH, Drummond AH, Downes CP: Inositol Lipids in Cell Signalling. London, Academic Press, 1989
J
systems
second-messenger
Dl
J 1990; 4:3178-3188 12.
messenger
229
1 . McGeer PL, Eccles JC, McGeer EG: Molecular Neurobiology of the Mammalian Brain, 2nd ed. New York, Plenum, 1987 2. North RA: Neurotransmitters and their receptors: from the clone to the clinic. Seminars in Neuroscience 1989; 1:81-90 3. Birnbaumer L: G proteins in signal transduction. Annu Rev Pharmacol Toxicol 1 990; 30:675-705 4. Gilman AG: G proteins: transducers of receptor-generated signals. Annu Rev Biochem 1987; 56:615-649 5. Nicoll RA, Malenka RC, Kauer JA: Functional comparison of neurotransmitter receptor subtypes in mammalian central nervous system. Nature 1989; 87:741-746 6. Ewald DA, Pang IH, Sternweis PC, Miller RJ: Differential G protein-mediated coupling of neurotransmitter receptors to Ca2+ channels in rat dorsal root ganglion neurons in vitro. Neuron 1989; 2:1185-1193 7. Bokoch GM, Katada T, Northup JK, Hewlett EL, Gilman AG: Identification of the predominant substrate for ADP-ribosylation by islet activating protein.J BiolChem 1983; 258:2072-2075 8. Sternweis PC, RobishawJD: Isolation oftwo proteins with high affinity for guanine nucleotides from membranes of bovine brain. J Biol Chem 1984; 259:13806-13813 9. Taylor SJ, Smith JA, Exton JH: Purification from bovine liver membranes of a guanine nucleotide-dependent activator of phosphoinositide-specific phospholipase C: immunologic identification as a novel G-protein alpha subunit. J Biol Chem 1990; 265:17150-17156 I 0. Freissmuth M, Casey PJ, Gilman AG: G proteins control diverse
FASEBJ
Second
HK, Simon MI, Gilman AG: Gz, protein with unique biochemical
J Biol Chem
Worley
27.
signaling.
SH:
Miller RJ: Multiple calcium channels Science 1987; 235:46-52 Neer EJ: Structural and functional studies Gen Physiol 5cr 1990; 45:143-15 1
on
REFERENCES
oftransmembrane
PF, Snyder
1990; 265:2383-2390 Linden J, Delahunty TM: Receptors tide breakdown. Trends Pharmacol
26.
pathways
Worley
Biol Chem
serves
ity
17.
disorders)
JM,
K. MANJI
37.
38.
39.
40.
41.
and
J, PremontJ, D2
of Clinical
Psychoneuroendocrinol-
Guilford Press, 1987 P, Drouva S, Rasolonjanahary
C, of adenylate cyclases by 17 betaneurons and anterior pituitary cells.
Enjalbert
dopamine-sensitive
estradiol in cultured striatal J Neurochem 1989; 52:410-418 Goldman-Rakic PS, Lidow MS. Galiager
paminergic, adrenergic, and serotoninergic plementarity of their subtypes in primate Neurosci 1990; 10:2125-2138 McCormick DA, Williamson A: Convergence neurotransmitter action in human cerebral Acad Sci USA 1989; 86:8098-8 102 Ross EM: Signal sorting and amplification coupled receptors. Neuron 1989; 3:141-152
Aghajanian GK, Wang YY: Common tor mechanisms in the locus coeruleus: brain North
slices. RA,
Neuropharmacology Williams JT:
On
R, Kordon
A: Differential
alpha
modulation
DW:
Overlap
receptors prefrontal
of doand comcortex. J
and divergence cortex. Proc through
of NatI
G protein-
2- and opiate
intracellular studies 1 987; 26:793-799 the potassium conductance
effecin in-
757
G PROTEINS
creased (Lond) 42.
43.
by
opioids
1985;
locus
coeruleus
J Physiol
neurones.
Andrade R, Malenka RC, Nicoll RA: A G protein couples serotonin and GABAB receptors to the same channels in hippocampus. Science 1986; 234:1261-1265 Lacey MG, Mercuri NB, North RA: On the potassium conductance
increase
activated
in rat substantia 437-453 44.
in rat
cology: The Third Generation HY. New York, Raven Press,
364:265-280
by GABAB
nigra
J
neurones.
and dopamine D2 receptors Physiol (Lond) 1988; 401:
pie to G proteins
B, SieverlingJ, on noradrenergic
and interact
with
Hertting nerve
64.
Nagy LE, Diamond I, Collier K, Lopez L, Ullman B, Gordon AS: Adenosine is required for ethanol-induced heterologous desensitization. Mochly-Rosen
the alpha
G: Presynaptic terminals cou-
2-adrenoceptors.
tion
66.
45.
46.
47. 48.
49.
50.
Si.
52.
53.
54.
Duman RS, Enna SJ: Modulation of receptor-mediated AMP production in brain. Neuropharmacology 1987;
26:98
986 Bouvier
Ann
57.
58.
59.
between
second
messengers.
C phosphorylates
the inhibitory
RM,
gene
ES, Spiegel
AM:
in Albright hereditary gradient gel electrophoresis.
Mutations
of the Gs alpha-
osteodystrophy Proc Natl
68.
.
69.
61.
Horn
Bilezkian JP: Mechanism of abnormal transmembrane signaling of the beta-adrenergic receptor in congestive heart failure. Circulation 1990; 80:271-283 Weinstein LS, Shenker A, Gejman PV, Merino MJ, Friedman E, Spiegel AM: Activating mutations of the stimulatory G protein in McCune-Albright syndrome. N EnglJ Med 1991; 325:1688-
72.
758
Shuckit
MA:
Biology
of risk
for alcoholism,
in Psychopharma-
Biochem
B, Estrin
W, Gordon
levels of cAMP
alcoholic
patients.
Nature
differentially
Biophys
Res
Corn-
Proc
A: Basal
and
are reduced
Natl
Acad
adenosine
in lymphocytes
Sci
USA
1987;
84:
Tabakoff B, Hoffman PL, Lee JM, Saito T, Willard B, Dc-Leon Jones F: Differences in platelet enzyme activity between alcoholics and nonalcoholics. N EnglJ Med 1988; 318:134-139 Nestler EJ, Erdos JJ, Terwilliger R, Durnan RS, Tallman iF: Regulation of G proteins by chronic morphine in the rat locus coeruleus. Brain Res 1989; 476:230-239
.
K, Beitner-Johnson EJ: Opiate withdrawal
DB, KrystalJH, Aghajanian and the rat locus coeruleus:
havioral, electrophysiological, and biochemical Neurosci 1990; 10:2308-2317 Attali B, Vogel Z: Long-term opiate exposure leads of the alpha i-i subunit of GTP-binding proteins.
GK, be-
J
correlates. to reduction
J Neurochem
1989; 53:1636-1639 Nestler EJ, Terwilliger RZ, Walker JR. Sevarino KA, Duman RS: Chronic cocaine treatment decreases levels of the G protein subunits
Gi alpha
J Neurochem
and
1990; M, Kleinman
Go
alpha
in discrete
55:1079-1082 JE, Hanbauer
regions
brain.
Memo
74.
recognition sites with adenylate cyclase in nuclei accumbens and caudatus of schizophrenics. Science 1983; 221:1304-1307 Seeman P, Niznik HB, Guan HC, Booth G, Ulpian C: Link be-
tween nia 75.
76.
77.
Dl and D2 dopamine and
Huntington
receptors
diseased
brain.
of dopamine
is reduced Proc
Kerwin RW, Beats BC: parahippocampal gyrus phrenic brain determined
Acad
Davies
(Gi,
Go)
brain.
AA: Mitral
hypersensitivity: with desensitization exposure. Am J Med 1987; 82:193-20 1 AO, Su CJ, Balasubramanyam A, Codina
L: Abnormal dysautonomia:
Sci
USA in the
J Neural
Increased forskolin binding in the left and CAl region in post mortem schizoby quantitative autoradiography.
Neurosci Lett 1990; 118:164-168 Davies AO, Mares A, PoolJL, Taylor
terenol
Dl
in schizophre-
Nati
1989; 86:10156-10160 Okada F, Crow TJ, Roberts GW: G-proteins basal ganglia of control and schizophrenic Transrn Gen Sect 1990; 79:227-234
with symptoms renergic receptor 78.
I: Coupling
of rat
73.
of beta-adrenergic supercoupling
guanine nucleotide regulatory evidence from reconstitution
valve prolapse beta 2-adon isopro-
J, Birnbaumer protein in MVP of Gs. J Clin En-
docrinol
79.
80.
81.
82.
1695 62.
I, Wrubel
Nestler
detected by Acad Sci USA
EM,
RNA.
M: Ethanol
cells.
I,
desensitiza-
s messenger
LA, Henteleff
Rasmussen
70.
1990; 87:8287-8290 60.
Diamond
1413-14 16
1986; 47:890-897
Gershon
subunit denaturing
alpha
in neural
M,
heterologous
1988; 155:138-143
from
guanine-nucleotide-
Sagi-Eisenberg R: GTP-binding proteins as possible targets for protein kinase C action. Trends Biochem Sci 1989; 14:355-357 Nishizuka T: The molecular heterogeneity of protein kinase C and its implications for cellular regulation. Nature I 988; 85: 661-665 Patten JL, Levine MA: Immunochemical analysis of the alphasubunit of the stimulatory G-protein of adenylyl cyclase in patients with Albright’s hereditary osteodystrophy. J Clin Endocrinol Metab 1990; 71:1208-1214 Weinstein LS, Gejman PV, Friedman E, Kadowaki T, Collins
causes
by reducing
receptor-stimulated
NY
binding regulatory component and apparently suppresses its function in hormonal inhibition of adenylate cyclase. Eur J Biochem 1985; 151:431-437 Olianas MC, Onali P: Phorbol esters increase GTP-dependent adenylate cyclase activity in rat brain striatal membranes. J Neu-
ethanol
G proteins
Diamond
1-
A2. Psychopharmacol Bull 1991; 27:247-253 Jakobs KH, Bauer S, Watanabe Y: Modulation of adenylate cyclase of human platelets by phorbol ester: impairment of the hormone-sensitive inhibitory pathway. Eur J Biochem 1985; 151:425-430 Katada T, Gilman AG, Watanabe Y, Bauer S, Jakobs KH: Pro-
rochem 56.
67.
cyclic
Acad Sci 1990; 594:120-129 Asakura M, Tsukamoto T, Kubota H, Osada K, Imafuku J, Nishizaki J, Sato A, Nakanishi J, Shimbo K, Shibata M: Involvement of protein kinase in the regulation of beta-adrenergic receptors by antidepressants. Int J Clin Pharmacol Res 1989; 9: 123-130 Manji HK, Chen G, Bitran JA, Potter WZ: Downregulation of beta adrenergic receptors in vitro involves PKC/phospholipase
tein kinase
55.
Cross-talk
of receptors
mun
71 M:
AS: Chronic
regulates
and guanine nucleotide bindFASEB J 1990; 4:2612-2622
Mol Pharmacol 1989; 36:744-748 D, Chang FH, Cheever L, Kim
1988; 333:848-850 Charness ME, Querimit
J
Neurochem 1989; 53:1629-1635 Adamson P, Mantzouridis T, Xiang JZ, Hajimohammadreza I, Brammer MJ, Campbell IC: Alpha-2-adrenergic, kappa-opiate, and P1-purinergic autoreceptors have mutually antagonistic effects: a new regulatory mechanism? J Neurochem 1989; 53: 1077-1082 Limberger N, Spath L, Starke K: Presynaptic alpha 2-adrenoceptor, opioid kappa-receptor and adenosine Al-receptor interactions on noradrenaline release in rabbit brain cortex. Naunyn Schmiedebergs Arch Pharmacol 1988; 338:53-61 Taylor CW: The role of G proteins in transmembrane signalling. BiochemJ 1990; 272:1-13 Rodbell M: Programmable messengers: a new theory of hormone action. Trends Biochem Sci 1985; 10:461-464
by Meltzer
Hoffman PL, TabakoffB: ing proteins: a selective
65.
Ethanol interaction.
Edited
63.
Gordon
Allgaier C, Daschmann kappa-opioid receptors
of Progress. 1987
83.
Metab 1991; 72:867-875 J, Dc Backer JP, Ebinger G, Vauquelin G: Coupling of Dl dopamine receptors to the guanine nucleotide binding protein Gs is deficient in Huntington’s disease. Brain Res I 989; 496:327-330 Dewar D, Horsburgh K, Graham DI, Brooks DN, McCulloch J: Selective alterations of high affinity E3Hlforskolin binding sites in Alzheimer’s disease: a quantitative autoradiographic
Dc Keyser
study. Brain Res 1990; 511:241-248 Nomura Y, Kitamura Y, Kawai M, Segawa T: Alpha 2-adrenoceptor-GTP binding regulatory protein-adenylate cyclase systern in cerebral cortical membranes of adult and senescent rats. Brain Young
Res 1986; 379:118-124 LT, Li PP, Kish Si, Siu KP, Warsh JJ: Postmortem cerebral cortex G, a-subunit levels are elevated in bipolar affective disorder. Brain Res 1991; 553:323-326 Schreiber G, Avissar S, Danon A, Belmaker RH: Hyperfunc-
Am
J
Psychiatry
1 49:6,
June
1992
HUSSEINI
84.
85.
86.
tional G proteins in mononuclear leukocytes of patients with mania. Biol Psychiatry 1991; 29:273-280 Pandey GN, Pandey SC, Davis JM: Peripheral adrenergic receptors in affective illness and schizophrenia. Pharmacol Toxicol 1990; 3:13-36 Mann JJ, Brown RP, Halper JP, Sweeney JA, Kocsis JH, Stokes PE, Bilezikian JP: Reduced sensitivity of lymphocyte beta-adrenergic receptors in patients with endogenous depression and psychomotor agitation. N Engi J Med 1985; 3 13:715-720
Ebstein
RP, Lerer
B, Shapira
B, Shemesh
Z, Moscovich
Kindler S: Cyclic AMP second-messenger signal in depression. BrJ Psychiatry 1988; 152:665-669
87.
Charney
DS, Innis
Platelet
alpha-2-receptor
in panic 107 88.
89.
90.
91.
disorder.
93.
94.
95.
96.
97.
RS, Woods
binding
and
cyclase 1989;
Berridge
MJ, Downes of lithium:
CP, Hanley a unifying
MR: Neural hypothesis.
99.
Cell
1989;
59:411-
101.
cerebral
Jpn J Psychiatry
cortex.
Okada
F, Tokumitsu
substrates
T: Effects
adenylate
Neurol
Y, Ui M: Possible
coupling
1988;
cyciase
in cerebral
involvement
cortices
in rat
42:858-860 of pertussis
(Gi, Go) in desipramine-induced of rats.
refractoriness J Neurochem
1988; 51:194-199 Yamaoka K, Nanba
T, Nomura S: Direct influence of antideGTP binding protein of adenylate cyclase in cell membranes of the cerebral cortex of rats. J Neural Transm 1988; 71:165-175 Duman RS, Terwilliger RZ, Nestler EJ: Chronic antidepressant regulation ofGsa and cyclic AMP-dependent protein kinase (abstract). Pharmacologist 1989; 31:182 Menkes DB, Rasenick MM, Wheeler MA, Bitensky MW: Guaon
nosine triphosphate activation of brain hancement by long-term antidepressant 1 10.
of desipramine
cyclase
adenylate
cyclase:
treatment.
en-
Science
1983; 219:65-67 Newman ME, Lerer B: Post-receptor-mediated increases in adenylate cyclase activity after chronic antidepressant treat-
ment: relationship
to receptor
desensitization.
Eur J Pharmacol
1989; 162:345-352 1 1 1.
Ozawa binding
H, Rasenick MM: Coupling of the protein Gs to rat synaptic membrane
is enhanced subsequent Mol Pharmacol 1989; 112.
Duman
RS,
Strada
to chronic 36:803-808 Si,
Enna
SJ:
antidepressant Effect
of
GTPcyclase
treatment.
imipramine
Goldberg H, Clayman P, Skorecki K: Mechanism of Li inhibition of vasopressin-sensitive adenylate cyclase in cultured renal epithelial cells. Am J Physiol 1988; 255(5, part 2):995-1 002 Manji HK, Bitran JA, Masana MI, Chen G, Hsiao JK, Risby
MV, Potter
WZ:
Signal
transduction
to the tricyclic
antidepressant
Manji Chronic
HK, Chen exposure
104.
Okada
F, Tokumitsu
1987;
gic
receptor-coupled
in vivo treatment 47:454-459
J
Y, Ui M: Desensitization
Psychiatry
adenylate
of rats with
cyclase
desipramine.
cortex
J Neurochem
June
1992
Gleiter
CH,
DeckertJ,
and repeated
tors, adenylate 1 16.
chem Huang
1989; RR,
Nutt
DJ, Marangos
cyclase, and the adenosine
Strader
CD: Identification
guanine
nucleotide-binding
118.
1 19.
120.
121.
122.
123.
after
1986;
PJ: Electroconvulsive
neuromodulatory ECS on the adenosine
52:641-646 Dehaven RN,
col 1990; 37:304-3 10 1 1 7. Anthony BL, Dennison
124.
1 49:6,
teins in rat cortex. EurJ Pharmacol 1990; 189:99-103 Nishida A, Kaiya H, Tohmatsu T, Wakabayashi S, Nozawa Y: Electroconvulsive treatment: effects on phospholipase C activity and GTP binding activity in rat brain. J Neural Transm Gen Sect
of single
of beta-adrener-
in cerebral
norepi-
and electroconvulsive shock attenuate betaand muscarinic cholinoceptor coupling to G pro-
shock (ECS) and the adenosine
49:282-289
GA, Bitran JA, Gusovsky F, Potter WZ: of C6 glioma cells to desipramine desensitizes beta-adrenoceptors, but increases KLJKH ratio. Eur J Pharmacol 1991; 206:159-162
brain
1990; 81:121-1 30 1 15.
desipra-
mine reduces the number of beta-adrenoceptors. Biochem Pharmacol 1986; 35:1899-1902 Fishman PH, Finberg JP: Effect of the tricyclic antidepressant desipramine on beta-adrenergic receptors in cultured rat glioma
C6 cells. J Neurochem
1 14.
modulation
by lithium: cell culture, cerebral microdialysis, and human studies. Psychopharmacol Bull 1991; 27:199-208 Sulser F: Antidepressant treatments and regulation of norepinephrine-receptor-coupled adenylate cyclase systems in brain. Adv Biochem Psychopharmacol 1984; 39:249-261 Honegger UE, Disler B, Wiesmann UN: Chronic exposure of
cells in culture
Carbamazepine adrenoceptor
on the rat
and
nephrine-coupled cyclic nucleotide generating system: alteration in alpha and beta components. J Pharmacol Exp Ther 1985; 234:409-414 Avissar S, Schreiber G, Aulakh CS, Wozniak KM, Murphy DL:
1 13.
administration
stimulatory adenylate
Chuang DM: Neurotransmitter receptors and phophoinositide turnover. Annu Rev Pharmacol Toxicol 1989; 29:71-110 Mork A, Geisler A: Effects of GTP on hormone-stimulated adenylate cyclase activity in cerebral cortex, striatum, and hippocampus from rats treated chronically with lithium. Biol Psychiatry 1989; 26:279-288 Mork A, Geisler A: The effects of lithium in vitro and cx vivo on adenylate cyclase in brain are exerted by distinct mechanisms. Neuropharmacology 1989; 28:307-311 Manji HK, Hsiao JK, Risby ED, Oliver J, Rudorfer MV, Potter WZ: The mechanisms of action of lithium, I: effects on serotoninergic and noradrenergic systems in normal subjects. Arch Gen Psychiatry 1991; 48:505-512 Risby ED, Hsiao JK, Manji HK, Bitran J, Moses F, Zhou DF, Potter WZ: The mechanisms of action of lithium, II: effects on adenylate cyclase activity and beta-adrenergic receptor binding in normal subjects. Arch Gen Psychiatry 1991; 48:513-524 Avissar S, Schreiber G, Danon A, Belmaker RH: Lithium inhibits adrenergic and cholinergic increases in GTP binding in rat
103.
Am
Y, Saito
adrenocorticotropin
human
102.
Hatta
receptor
419
ED, Rudorfer
100.
109.
H,
on
pressants
and developmen-
cortex. Nature 1988; 331:440-442 98.
107.
GR:
activity 98:102-
F, Ikeda
administration
of adenylate
108.
(Berlin)
Tsuchiya
toxin
DG,
SW, Heninger
adenylate
Psychopharmacology
106.
amplification
Lerer B, Bleich A, Bennett ER, Ebstein RP, Balkin J: Platelet adenylate cyclase and phospholipase C activity in posttraumatic stress disorder. Biol Psychiatry 1990; 27:735-740 Berridge MJ, Downes CP, Hanley MR: Lithium amplifies agonist-dependent phosphatidylinositol responses in brain and salivary glands. Biochem J 1982; 206:587-595 Hallcher LM, Sherman WR: The effects oflithium ion and other agents on the activity of myoinositol1 -phophatase from bovine brain. J Biol Chem 1980; 255:10896-10901 tal actions
92.
RB, Duman
105.
K. MANJI
Cheung
of allosteric protein
RL, Aronstam
AH,
system: effect Al and A2 recepuptake site. J Neuro-
Diehl
RE,
antagonists interactions.
Dixon
RA,
of receptorMol
RS: Influence
Pharma-
of volatile
anesthetics on muscarinic regulation of adenylate cyclase activity. Biochem Pharmacol I 990; 40:376-379 Parsons WJ, Ramkumar V, Stiles GL: Isobutylmethylxanthine stimulates adenylate cyclase by blocking the inhibitory regulatory protein, Gi. Mol Pharmacol 1988; 34:37-41 Mousli M, Bueb JL, Bronner C, Rouot B, Landry Y: G protein activation: a receptor-independent mode of action for cationic amphiphilic neuropeptides and venom peptides. Trends PharmacolSci 1990; 11:358-362 Moor M, Honegger UE, Weismann UN: Organospecific, qualitative changes in the phospholipid composition of rats after chronic administration of the antidepressant drug, desipramine. Biochem Pharmacol 1987; 37:2035-2039 Bruns RF, Fergus JH: Allosteric enhancement of adenosine Al receptor binding and function by 2-amino-3-benzoylthiophenes. Mol Pharmacol 1990; 38:939-949 Milligan G: Techniques used in the identification and analysis of function of pertussis toxin-sensitive guanine nucleotide binding proteins. Biochem J 1988; 255:1-13 Simonds WF, Goldsmith PK, CodinaJ, Unson CG, Spiegel AM: Gi2 mediates alpha 2-adrenergic inhibition of adenylyl cyclase in platelet membranes: in situ identification with G alpha C-terminal antibodies. Proc Natl Acad Sci USA 1989; 86:7809-78 13 Akita Y, Saito T, Yajima Y, Sakuma S: The stimulatory and
759
G PROTEINS
inhibitory cyclase roidism Spiegel
125.
guanine
nucleotide-binding
in erythrocytes type I. J Clin AM, Levine
ciency
of hormone
from
receptor-adenylate
cyclase resistance. GM:
127.
128.
stimulatory guanosine with pressure-overload 1988; 81:420-424 Koski G, Simonds WF,
of agonists
J Biol
Chem
coupling Prog
Biol
Res
analysis mutants.
Klee
WA:
Guanine
nucleotides
opiate
use
Studying
Several distinct and used to study G proteins
in Clinical Populations
G Proteins
complementary approaches in clinical populations. The
can range
be of
techniques developed in recent years to study G protein interactions with both receptors and effectors is described in detail in an excellent
tial
problem
cold
NAD
the method
GTP-dependent
of
(127),
though
not,
when
tional changes.
used
alone,
provide
changes occurring For this reason,
valuable
adjunctive aspects
structural
namide-adenine
information
in the pertussis
tools
for
about
investigating
of various
G proteins.
dinucleotide
(NAD)
functional By using
and
form.
in ADP-ribosylating In contrast,
G protein
G in the activated,
the inactive,
is ADP-ribosylated
mal susceptibility ribosylation has
760
only
undissociated
by pertussis
toxin.
example,
a comparison
used
ADPcon-
does
not
exist.
Al-
can be generated pertussis G1 (122).
adenylyl
G protein),
cholera
as probes
by
toxin
cyclase
and
forskolin
of For
activity
and G protein (which
in re-
analogues
acts
at G, and
G protein,
and
toxin
pertussis
of the signal
and
effector
transduction
pathway
may
be
because
of
G and to inactivate/un-
activate GTP
should,
diof
to be identoxin
binding,
agonist-induced
agonist-induced increase in GTPase in theory, serve as good indexes for G protein in the receptor-effector coupling being studied.
and
activity involveHow-
ever, a number of practical problems exist. First, the inability of GTP or its analogues to enter intact cells across the plasma membrane necessitates the use of membrane preparations. Second,
in a number
of systems,
the
basal
GTP
binding
and
GTPase activity are high, perhaps because signal-transducing G proteins represent only one group of GTP binding and
hy-
drolyzing
the
high
proteins.
nonspecific
increases.
abnor-
memand
allow for the dissection protein/effector system.
fluoride
in receptor,
Similarly,
to preclude
of the G proteins to toxin-catalyzed been shown to result from an abnormal
of the
agonists,
at the
abnormalities
dissociated of the
a G
the catalytic unit of adenylyl cyclase), and Mn2 (which rectly stimulates adenylyl cyclase) may allow identification
ment
the
of
a model
Certain pharmacologic agents the multicomponent receptor/G
form
Thus,
of
G protein but have allows for the de-
!ymphoma
such
GDP release,
[32Plnicoti-
lack a specific and effector(s)
S49
in-
transfer of its [32P]ADP-ribose moiety to membrane components, it is possible to obtain information about functional aspects of the G protein. In such investigations, [32P]ADP-ribosylation of membrane components can be used because both cholera and pertussis toxin preferentially catalyze the ADP-ribosylation of a given conformational state (dissociated/undissociated) of the G proteins. Cholera toxin is very
effective
cyc
their ability to constitutively couple G, respectively. Finally, agonist-induced
and
determining
excess
extraction of G, from a donor full reconstitution of -adrenergic
do
conforma-
absence of quantitative and cholera toxin are
a large
treating appropriate membranes with then examining the effects of exogenous
tified.
they
including
activity
cumbersome,
adenylyl cyclase inhibition (123). era recognize both the dissociated
However, since these antisa subunits and the corre-
that functional
of the
more
act
complex,
by
mixture.
adenylyl cyclase activity in this cell line. Such studies have been used to demonstrate G, abin pseudohypoparathyroidism (125, 126), heart and mitral valve prolapse (78). At present, a
(which
4y
overcome
detergent for the
to receptor
heterotrimeric
pertussis toxin ribosylcases, it may be possible
Cyc
sponse
undissociated,
in various
problems as a means of tissues. First,
possess cyclase
Thus, allows
the
choice for quantitating G protein subunits. Similarly, these peptide antibodies can be used as specific probes of G protein receptor-effector coupling. Thus, although human platelets contain G, G2, G13, G , and G, only G.2 appears to functionally couple to the a2-a#{228}renergic receptor and thereby mediate
sponding
potential
protein.
(124). brane
of
failure
yields the greatest conformational (and G proteins. Several
is clearly
be
reaction
receptor(s)
normalities
of researchers have generated specific antisera against peptides corresponding to the predicted amino acid of various G protein subunits.
antisera
are
to intracel-
variants of the S49 lymphoma cell line lack G, but -adrenergic receptors and catalytic units of adenylyl
equivalent”
specific
there
G proteins
membranes
termination
era and toxin-catalyzed ADP-ribosylation information about the quantity and therefore functional) state of specific
use of these
can
in the
using
reconstitution
The
in relay-
receptors
ADP-ribosylation
quantifying
appropriate
review by Milligan (122); only a brief overview of clinically applicable methods is presented here. The simultaneous use of immunoblotting techniques with specific antis-
groups synthetic sequence
abnormalities
Another functional probe is the so-called reconstitution assay, in which the G protein from the tissue in question is cxtracted and used to restore deficient adenylyl cyclase activity.
inhibit
receptors.
256:1536-1538
for
However,
toxin-catalyzed
and
with
extracellular
both cholera toxin and, in particular, ate more than one G protein. In some
Thus, 1. Methods
to correlate from
(1 13).
of
detecting
the APPENDIX
effectors
in the
and
signals
to separate the G protein a subunits electrophoretically. Second, [32PJNAD+ can be degraded by endogenous NAD-glycohydrolases, thereby altering substrate availability. This poten-
Ci: Decreased
to soluble
state
molecular
lular defi-
protein
Clin
Somatic genetic of unresponsive
SF, Homcy
ing
triphosphate binding protein in dogs left ventricular failure. J Clin Invest
and antagonists
1981;
formational
with
J Cell Physiol 1975; 85:611-620 Longabaugh JP, Vatner DE, Vatner
binding
of adenylate
pseudohypoparathyEndocrinol Metab 1985; 61:1012-1017 MA: Pseudohypoparathyroidism:
as a cause of hereditary hormone 1982; 97:327-340 Coffino P, Bourne HR, Tomkins ofcyclic AMP action: characterization
126.
proteins
patients
in
in the
GTPase
an This
the study
Thus,
accurate assay
activities
(122),
perhaps
membrane
J
to
be
more
toxin
greater
abundance
of their
“noise”
of receptor-mediated
to pertussis
and their greater
Am
sufficient
however,
linked
because
of instances,
generate
determination appears,
of receptors
plasma
majority
enzymatic
Psychiatry
effective
G proteins
capacity
1 49:6,
June
in the (128).
1992