Bioinformatics Research Laboratory, IBI Biosolutions Pvt. Ltd. India

Signal Transduction Pathways for Catecholamine Receptors

INTRODUCTION

Until relatively recently, synaptic transmission was conceptualized as a set of processes by which neurotransmitters, acting through their receptors, caused changes in the conductances of specific ion channels to produce excitatory or inhibitory postsynaptic potentials. According to this view, the human brain could be viewed as a very complex digital computer with its complexity derived largely from its wiring diagram. Over the past twenty years, however, it has become evident that neurotransmitters elicit diverse and complicated effects in target neurons. This has led to a much more complete view of synaptic transmission (20). Thus, in addition to the rapid elicitation of postsynaptic potentials, neurotransmitter-receptor interactions influence virtually every aspect of a target neuron's functioning through a complex network of intracellular messenger systems (Figure 1). The purpose of this Chapter is to present a brief overview of these intracellular messenger systems in the brain and to describe how these systems mediate the many effects of catecholamines and their numerous receptors on target neuron functioning.

OVERVIEW OF INTRACELLULAR MESSENGER SYSTEMS IN BRAIN

Activation of all catecholamine receptors and of most other types of receptors is transmitted to intracellular sites via G-proteins (Figure 1). G-proteins then "couple" receptors to various effector proteins, which include ion channels and numerous intracellular second messenger pathways. Generation of second messengers leads to diverse physiological effects via cascades of intracellular messengers. In most cases, these intracellular cascades involve ultimately changes in protein phosphorylation--the addition (via protein kinases) or removal (via protein phosphatases) of phosphate groups from target phosphoproteins. Altered phosphorylation of phosphoproteins, which can be considered "third messengers," alters their physiological activity. As with all neurotransmitters, catecholamine regulation of second messenger and protein phosphorylation pathways influences virtually all aspects of neuronal function through the phosphorylation of diverse types of neural proteins (Figure 1). Such intracellular processes produce some of the rapid responses to the neurotransmitter, such as regulation of ion channels and neuronal firing rate. In addition, these processes produce short-term modulatory effects on neuronal function, such as regulation of the responsiveness of the neuron to the same or different neurotransmitters (e.g., via altered receptor sensitivity). Finally, these processes produce more long-term modulatory effects on neuronal function, including changes that are achieved through the regulation of gene expression. Such changes can include altered synthesis of receptors, ion channels, and other cellular proteins, and ultimately forms of learning and memory. Individual steps in these intracellular cascades are given below in greater detail.

G-Proteins in Brain Signal Transduction

With the exception of synaptic transmission mediated via receptors that contain intrinsic ion channels or enzyme activity, the family of G-proteins may be involved in all other transmembrane signaling in the nervous system (Figure 2; see 29,40). G-proteins are so-named because of their ability to bind the guanine nucleotides, guanosine triphosphate (GTP) and guanosine diphosphate (GDP). Four major types of G-proteins are involved in transduction of signals produced by neurotransmitter binding, Gs, Gi/o, Gq, and G12 and multiple subtypes exist for each. Rhodopsin, the light sensitive molecule of photoreceptor cells in the retina, can also be viewed as a G-protein-linked receptor: light activates rhodopsin which then, through a fifth type of G-protein called transducin (Gt), regulates the electrical properties of photoreceptor cells. Each type of G-protein is a heterotrimer composed of single a, b, and g subunits. As for the a subunits, there are also multiple b and g subtypes. Distinct a subunits, as well as distinct bg subunits, confer specific functional activity on the different types of G-proteins.

G-protein regulation of ion channels. G-proteins have been shown to couple neurotransmitter receptors to multiple types of intracellular effector proteins. In some cases, G-proteins couple neurotransmitter receptors directly to certain types of ion channels (e.g., Figure 3) (33). In this case, it appears that the bg subunits released from the G-protein-receptor interaction directly gate (i.e., opens or closes) the channel. The best established example of this type of mechanism in brain is the coupling of many types of receptors, via subtypes of Go and Gi in many types of neurons, to the activation of an inward rectifying K+ channel and to the inhibition of a voltage-dependent Ca2+ channel, actions that hyperpolarize cells. There are also reports that a subunits may regulate ion channel function in some cases.

G-protein regulation of intracellular second messengers. In addition to direct regulation of ion channels, G-proteins transduce the activation of neurotransmitter receptors into alterations in intracellular levels of second messengers in target neurons. Prominent second messengers in brain include cyclic AMP (cAMP), cyclic GMP (cGMP), calcium, the major metabolites of phosphatidylinositol (PI) [inositol triphosphate (IP3) and diacylglycerol] and of arachidonic acid, and nitric oxide (NO). As discussed above, altered levels of second messengers mediate the actions of neurotransmitter-receptor activation on some types of ion channels, as well as on numerous other physiological responses as outlined in Figure 1.

Second Messengers in Brain Signal Transduction

cAMP. The molecular mechanism by which neurotransmitters regulate cAMP levels is well-established (Figure 3). Gs couples certain receptors to adenylyl cyclase, the enzyme responsible for the synthesis of cAMP, such that the enzyme is stimulated by receptor activation. In contrast, Gi (and possibly Go) couples other receptors to adenylyl cyclase such that the enzyme is inhibited by receptor activation (40,42). A variant form of Gi, termed Gz, may also mediate receptor inhibition of adenylyl cyclase in some cell types. Nine forms of adenylyl cyclase have been cloned to date (8,10,12,13,). These enzymes show different regional distributions in brain, as well as distinct regulatory properties. The enzymes differ in their ability to be activated or inhibited upon binding free bg complexes or Ca2+/calmodulin complexes (Figure 3) (8).

cGMP and nitric oxide. Neurotransmitters regulate cellular cGMP levels via two general mechanisms (Figure 3) (8,44). In some cases, nitric oxide appears to act as an intracellular second messenger in mediating the ability of certain receptors to activate guanylyl cyclase, the enzyme that catalyzes the synthesis of cGMP. It is thought that these receptors elicit an increase in intracellular Ca2+ levels (as described below), which activates nitric oxide synthase, the enzyme responsible for the synthesis of nitric oxide (4). Nitric oxide then directly activates cytoplasmic forms of guanylyl cyclase. In other cases, as with the atrial natriuretic peptide receptor and related systems, the enzyme guanylyl cyclase resides within the receptor protein.

In addition to activating guanylyl cyclase, nitric oxide has been shown to regulate ADP-ribosylation, a process whereby ADP-ribose groups are transferred from NAD (nicotinamide adenine dinucleotide) to specific substrate proteins. The addition of ADP-ribose then alters the physiological activity of the protein. Brain contains high levels of ADP-ribosyltransferases which catalyze this reaction, and some forms of the enzyme are activated by nitric oxide (8,43). Although most of the physiological substrates for these nitric oxide-sensitive and -insensitive enzymes remain unknown, recent studies demonstrate that growth-associated protein of 43 kD (GAP-43) and Gas are substrates in vitro, although more work is needed to define how the function of these proteins is altered by ADP-ribosylation. [Cholera and pertussis toxins can also be considered ADP-ribosyltransferases in that they catalyze the ADP-ribosylation of specific G-protein a subunits, but their activity is not affected by nitric oxide.] Further research is needed to determine which of the second messenger roles of nitric oxide in the brain are mediated via regulation of ADP-ribosylation. In any case, it is notable that recent studies to identify a retrograde messenger involved in long-term potentiation have implicated nitric oxide and ADP-ribosylation: inhibition of nitric oxide synthesis or ADP-ribosylation has been reported to block the formation of long-term potentiation. However, this remains controversial, since not all laboratories have been able to replicate these findings (29).

Phosphodiesterases. Cyclic nucleotide levels in neurons are highly regulated by the metabolism, as well as the synthesis, of these second messengers. This is accomplished by phosphodiesterases (PDEs), a large family of enzymes which catalyze the conversion of cAMP and cGMP into 5'-AMP and 5'-GMP, respectively (1,8). There are at least eight forms of PDE, which display different affinities for cyclic nucleotides and are differentially regulated and distributed in brain. PDE1 isozymes account for more than 90 percent of enzyme activity in brain and can hydrolyze either cAMP or cGMP. These PDEs are stimulated by Ca2+/calmodulin and are thereby regulated by extracellular stimuli that regulate Ca2+ levels. PDE2 enzymes are regulated by binding cGMP and also hydrolyze both cAMP and cGMP. PDE4 is specific for cAMP, and there are reports that the enzyme is regulated by guanine nucleotides, suggesting a possible role for G proteins. PDE3 and 5 are expressed in peripheral tissues and rod outer segments but to date have not been found in brain. Less is known about the relative distribution of the remaining PDE isoforms and their relevance to the regulation of cyclic nucleotides in the brain.

Calcium and the phosphatidylinositol system. The ways in which neurotransmitters alter intracellular Ca2+ levels are more complex compared to those for cyclic nucleotides and involve two types of mechanisms that operate to different extents in different cell types (Figure 3). Neurotransmitter receptor activation can alter the flux of extracellular Ca2+ into neurons or can regulate release of Ca2+ from intracellular stores. Once released, Ca2+ can exert multiple actions on neuronal function via intracellular regulatory proteins (Figure 3). Receptors can directly regulate the conductance of specific voltage-gated Ca2+ channels via coupling with G-proteins, as mentioned above. In addition, activation of other second messenger systems can alter Ca2+ channel conductance; for example, cAMP, and neurotransmitters that act through cAMP, can increase the conductance of some voltage-gated Ca2+ channels (see below). Depolarization of a neuron by any means will activate voltage-gated Ca2+ channels, which will lead to the flux of Ca2+ into the cells. Finally, extracellular Ca2+ can pass through some ligand-gated channels, such as the nicotinic cholinergic and NMDA-glutamate receptors.

Receptor activation can increase intracellular levels of free Ca2+ through regulation of the phosphatidylinositol system and subsequent actions on intracellular Ca2+ stores (Figure 3) (2). Many types of neurotransmitter receptors are coupled through G-proteins to an enzyme termed phospholipase C (PLC) (also referred to as phosphoinositidase C). This effect is mediated predominantly by Gq, although Gi and Go may be involved in some cell types (40). Multiple forms of phospholipase C have been identified in brain, which show different anatomical and regulatory properties: PLC-b1 is stimulated by G-protein a subunits, PLC-b2 is stimulated by bg subunits, and PLC-g is activated upon phosphorylation by protein tyrosine kinases (see below). Phospholipase C catalyzes the breakdown of phosphatidylinositol resulting in the generation of inositol triphosphate (IP3), which, through binding to a specific inositol triphosphate receptor located on intracellular organelles (e.g., endoplasmic reticulum), releases Ca2+ from intracellular stores. The inositol triphosphate receptor, like the related ryanodine receptor, forms a Ca2+ channel that responds to inositol triphosphate by releasing Ca2+ stores. In addition, Ca2+ itself can exert a stimulatory effect on inositol triphosphate and ryanodine receptors, which may underlie "oscillations and waves" in Ca2+ levels in some neurons and other cell types (2): this effect of Ca2+ represents a type of positive feedback which promotes the spread of the Ca2+ signal throughout the cell.

Arachidonic acid metabolites. The prostaglandins and leukotrienes represent another family of intracellular messengers (see Piomelli, this volume, for detailed discussion). Briefly, this family of messengers is generated by activation of an enzyme termed phospholipase A2, which cleaves membrane phospholipids to yield free arachidonic acid. The activity of phospholipase A2 may be regulated by certain neurotransmitter-receptor interactions via G-proteins, although this remains speculative. Next, arachidonic acid is cleaved by cyclooxygenase (an enzyme inhibited by aspirin and other non-steroidal anti-inflammatory drugs) to yield, after numerous additional enzymatic steps, several types of prostaglandins and other cyclic endoperoxides (e.g., prostacyclins and thromboxanes) or by lipoxygenase to yield the leukotrienes. These endoperoxides and leukotrienes exert many effects on cell function by influencing directly the activity of adenylyl cyclase, guanylyl cyclase, ion channels, protein kinases, and other cellular proteins (37; also see Piomelli, this volume, for detailed discussion).

Protein Phosphorylation as a Final Common Pathway in the Regulation of Neuronal Function

Despite the large number of second messengers that can be activated within neurons, there is a relatively uniform way in which these signaling pathways work. While second messenger molecules may occasionally have direct actions as effectors (e.g., cAMP can bind to and directly gate certain ion channels, and Ca2+ can bind to and directly regulate the activity of several enzymes), most of the known effects of intracellular second messengers are mediated, as stated earlier, by protein phosphorylation: by stimulating the addition or removal of phosphate groups from specific amino acid residues in target proteins. Phosphate groups alter the conformation and charge of proteins and thereby alter their function (30,31).

The regulation of protein function by phosphorylation plays a paramount role in signal transduction within the brain, a view originally proposed by Greengard and co-workers 30 years ago. In most cases, neurotransmitters regulate protein phosphorylation through second messenger-mediated activation of enzymes called protein kinases. Protein kinases transfer phosphate groups from ATP to serine, threonine, or tyrosine residues in specific substrate proteins. Neurotransmitters can also regulate protein phosphorylation through second messenger-mediated regulation of protein phosphatases, enzymes that remove phosphate groups from proteins through hydrolysis. Each protein kinase and protein phosphatase acts on a specific array of substrate proteins.

Protein kinases. The best studied protein kinases in brain are those activated by the second messengers cAMP, cGMP, Ca2+, and diacylglycerol (21,30,31). These protein kinases are named for the second messengers that activate them. Brain contains one major class of cAMP-dependent protein kinase and one major class of cGMP-dependent protein kinase (Figure 3), although isoforms of these enzymes are now known. In contrast, two major classes of Ca2+-dependent protein kinases have been described (Figure 3). One is activated by Ca2+ in conjunction with the Ca2+-binding protein calmodulin and are referred to as Ca2+/calmodulin-dependent protein kinases. The other is activated by Ca2+ in conjunction with diacylglycerol and other lipids and is referred to as protein kinase C (34). The brain contains several subtypes of each of these Ca2+-dependent enzymes, which exhibit different regulatory properties and are expressed differentially in neuronal cell types throughout the nervous system.

The brain contains numerous additional types of protein serine/threonine kinases, which are not directly activated by second messengers, and numerous types of protein tyrosine kinases, which phosphorylate substrate proteins specifically on tyrosine residues (see Figure 1). It is likely that many of these protein kinases also play critically important roles in brain signal transduction, although the mechanisms involved are not as clearly established as for the second messenger-dependent enzymes.

An example of a second messenger-independent, protein serine/threonine kinase are the MAP-kinases (also referred to as ERKs, Extracellular signal-Regulated Kinases), first identified on the basis of their association with, and phosphorylation of, microtubule-associated proteins (MAPs) (3,6). MAP-kinases have since been shown to phosphorylate a number of other proteins in brain and elsewhere, including other protein tyrosine kinases, tyrosine hydroxylase and numerous DNA-binding proteins. The activity of MAP-kinases are regulated by many extracellular signals, apparently through cAMP-dependent and Ca2+-dependent protein kinases and protein tyrosine kinases. Thus, those neurotransmitters, including catecholamines (see below), that influence the cAMP and Ca2+ pathways initially, would be expected ultimately to regulate (and produce certain physiological effects via) the MAP-kinase system.

The activity of MAP-kinases themselves is controlled through complex cascades involving protein phosphorylation. MAP-kinases are activated via their phosphorylation on threonine and tyrosine residues by another protein kinase, termed MEK (Mitogen and Extracellular-regulated Kinase), which in turn can be phosphorylated and activated by several MEK kinases (also referred to as ERK-kinase kinases), such as Raf-kinase. The mechanisms by which these MAP-kinases are influenced by second messenger-dependent protein kinases and by protein tyrosine kinases is becoming increasingly well known and are summarized in Figure 2. This system highlights the complex inter-relationships among intracellular messenger pathways and their regulation of cell function.

The best studied protein tyrosine kinases are those that are associated with plasma membrane receptors for many types of growth factors (see 21). Receptors for most growth factors, including insulin, epidermal growth factor, nerve growth factor (NGF) and related proteins (i.e., brain derived neurotrophic factor and neurotrophin-3 and -4), possess protein tyrosine kinase enzyme activity within the receptor complex. Many forms of these receptor-kinases are known. The neurotrophins activate a class of receptor-kinases, termed Trk proteins (Tropomyosin receptor kinase). Recent studies (see 3,6) have revealed some of the mechanisms by which activation of these receptors lead to biological responses (see Figure 2). Binding of growth factor to its receptor leads to dimerization and activation of the receptor-associated protein tyrosine kinase, and then autophosphorylation of multiple tyrosine residues on the receptor itself, which creates SH (Src Homology) adapter domains. This leads to the coupling of Trk with other proteins with SH (Src Homology) domains. Some of these SH-containing proteins (i.e., SHC, Grb, and SOS) are involved in activation of the MAP kinase pathway, while others (e.g., phospholipase C-g and phosphatidylinositol-3 kinase) are effector proteins that lead to biological responses (Figure 2).

The other major class of protein tyrosine kinase (e.g., Src) lacks a receptor domain. The mechanism underlying their regulation has remained elusive, although early evidence indicates that some of these enzymes might transiently become associated with specific receptors or other membrane proteins via SH domains and thereby transduce extracellular signals into changes in intraneuronal function.

Protein phosphatases. Protein phosphatases can be divided into two major classes based on the types of amino acids they dephosphorylate: serine/threonine phosphatases and tyrosine phosphatases (21,39). There are two known mechanisms by which neurotransmitters can influence protein phosphorylation through the regulation of protein serine/threonine phosphatases. One phosphatase, referred to as calcineurin or phosphatase 2B, can be activated directly by binding Ca2+/calmodulin. Presumably, neurotransmitters that alter cellular Ca2+ levels influence the phosphorylation of cellular proteins through alterations in calcineurin activity. The other mechanism is indirect and involves a class of protein referred to as protein phosphatase inhibitors. The best known protein phosphatase inhibitors are phosphatase inhibitors-1 and -2 and DARPP-32, the latter an inhibitor protein expressed in specific neuronal cell types in the brain (see below). These proteins are potent inhibitors of protein phosphatase1, and their phosphorylation by cAMP-dependent or other protein kinases alters their inhibitory activity. Presumably, in neurons that contain these phosphatase inhibitors, neurotransmitters that alter cellular cAMP levels influence the phosphorylation of cellular proteins through alterations in protein phosphatase 1 activity. Less is known about the physiological regulation of protein tyrosine phosphatases in the brain.

Regulation of proteins by phosphorylation. Following the regulation of protein kinase or protein phosphatase activity, the next step in intracellular signal transduction involves regulation of the phosphorylation state of specific neuronal phosphoproteins. These phosphoproteins are referred to as third messengers. Virtually every type of neural protein is now known to be regulated by phosphorylation, indicating the widespread role of protein phosphorylation in the regulation of diverse aspects of neuronal function. This includes the regulation of ion channel conductance, activity of various transporters, neurotransmitter receptor sensitivity, neurotransmitter synthesis and release, axoplasmic transport, elaboration of dendritic and axonal processes, development and maintenance of differentiated characteristics of neurons, and gene expression. (See Nestler and Duman, this volume, for further discussion of gene expression and neuronal plasticity.)

The above discussion of signal transduction pathways portrays protein phosphorylation as the major molecular currency with which protein function is regulated in response to extracellular stimuli, a view supported by over a generation of research. Thus, although proteins are known to be covalently modified in many other ways, e.g., by ADP-ribosylation, carboxymethylation, acetylation, myristoylation, palmitoylation, tyrosine sulfation, isoprenylation, and glycosylation, none of these mechanisms is as widespread and readily subject to regulation by synaptic stimuli as phosphorylation.

Heterogeneity in Brain Signal Transduction Pathways

As with receptors and ion channels, molecular biological studies have demonstrated extraordinary heterogeneity in intracellular messenger pathways, a degree of heterogeneity not suspected by classical biochemical, pharmacological, or physiological studies. For example, whereas biochemical and pharmacological studies indicated the existence of five major types of G-proteins (i.e., Gs, Gi, Go, Gq, and Gt), two types of cAMP-dependent protein kinase, and just one type of protein kinase C, molecular cloning studies now indicate the existence of over 20 distinct G-protein subunits, 7 distinct subunits of cAMP-dependent protein kinase, and over 7 subtypes of protein kinase C. Such heterogeneity is due to a combination of the existence of numerous distinct genes for each of the proteins plus alternative splicing of some common genes. Comparison of the individual subtypes of these proteins has indicated that they possess different regulatory properties and exhibit varying levels of expression in different neuronal cell types. This high degree of heterogeneity indicates still greater potential for functional specificity within and between neuronal cell types in the brain. Such heterogeneity also raises the possibility of developing drugs targeted for specific subtypes of intracellular messengers, drugs that would represent novel approaches in the treatment of neuropsychiatric disorders.

INTRACELLULAR SYSTEMS COUPLED TO CATECHOLAMINE RECEPTORS: PROXIMAL EFFECTS

All known receptors for catecholamine neurotransmitters belong to the family of G-protein-coupled receptors, which possess seven transmembrane domains and produce all of their physiological effects via interactions with G-proteins. These receptors, and the intracellular messenger pathways through which they produce their physiological effects, can be categorized based on the species of G-protein with which they interact initially. In this section, we focus on the mechanisms by which catecholamine receptors, via their interactions with G-proteins and intracellular messengers, produce rapid electrophysiological effects. In the next section, we illustrate how the same recruitment of intracellular messengers also results in the regulation of many additional neural processes.

Before discussing signal transduction pathways for specific catecholamine receptors, it is important to emphasize the technical difficulties in delineating these pathways experimentally. For example, to support a second messenger role of cAMP in the electrophysiological actions of a catecholamine receptor, it is necessary to show that agents that directly activate the cAMP pathway mimic receptor activation. This is not straightforward, because cAMP analogues applied extracellularly cannot elevate intracellular cAMP levels as rapidly and to the same extent as receptor activation. As a result, numerous reports of the inability of cAMP or other second messengers to mimic receptor activation must be viewed with extreme caution. Similar technical problems exist for demonstrating that agents that directly inhibit cAMP, or other second messenger pathways, block the consequences of receptor activation. Ultimately, the signal transduction cascades for individual receptors must be studied by patch clamp and related techniques, which permit direct access to the intracellular milieu.

Receptors Coupled to Gs

b1- and b2-adrenergic and D1- and D5-dopamine receptors are believed to produce their physiological actions via interactions with Gs and the subsequent stimulation of adenylyl cyclase and cAMP-dependent protein kinase (see 41). This cascade mediates the electrophysiological actions of these receptors through the phosphorylation of ion channels and pumps. Most, and possibly all, types of channels and pumps are acutely regulated via their phosphorylation by many types of protein kinases. The electrophysiological actions of b-adrenergic and D1/5-dopamine receptors therefore depend on the types of channels and pumps expressed in a particular type of target cell that can be phosphorylated by cAMP-dependent protein kinase. For example, b-adrenergic receptor stimulation depolarizes cardiac myocytes via the phosphorylation and activation of voltage-dependent Ca2+ channels. b-adrenergic receptor stimulation also promotes depolarization of hippocampal pyramidal and many other neurons, although in this case the effect is mediated via phosphorylation and inhibition of Ca2+-activated K+ channels (33), as well as the phosphorylation and facilitation of glutamate receptor function (14). It should be noted that this latter mechanism may mediate the ability of b-adrenergic receptors to influence long-term potentiation in the hippocampus. In contrast, b-adrenergic receptor stimulation promotes GABA-induced hyperpolarization of cerebellar Purkinje cells (38), possibly via the phosphorylation and facilitation of GABAA receptor-chloride channel function (25). Most known effects of D1/5 receptor stimulation are hyperpolarizing, although the specific ion channels involved have not yet been identified in most cases.

Other signal transduction pathways for these receptors have been reported in the literature (41), although the extent to which they mediate physiological actions of the receptors in brain remains uncertain. In the heart, free Gas generated via activation of b-adrenergic receptors may activate voltage-dependent Ca2+ channels in two ways: via stimulation of adenylyl cyclase leading eventually to channel phosphorylation (as outlined above), as well as by directly binding to and activating the channels. In certain tissues, activation of D1-dopamine receptors is claimed to activate phosphatidylinositol hydrolysis, leading to the speculation that some of the effects of D1 receptors are mediated via the inositol triphosphate, Ca2+, and protein kinase C cascade. However, these studies have not ruled out the alternative explanation that D1-induced activation of phosphatidylinositol hydrolysis may be mediated via the cAMP pathway.

Receptors Coupled to Gi/Go

Catecholamine receptors coupled to the Gi/Go family of G-proteins, and include several subtypes of the a2-adrenergic receptor and the D2-, D3-, and D4-dopamine receptors (41). In addition, there are reports that the a1A-adrenergic receptor subtype also utilizes these G proteins. The role of Gi and Go in mediating the actions of these receptors is based on the ability of pertussis toxin, which ADP-ribosylates and inactivates these G-proteins, to block various physiological actions of receptor activation. However, it remains unanswered in most cases as to which subtype of Gi and/or Go mediates the various effects of a certain receptor in a given cell type.

The a2-adrenergic and D2-4-dopamine receptors, and probably all other types of receptors that are coupled to Gi and Go, produce their rapid physiological actions via two major mechanisms, which can occur in the same target neurons (33,41). In one mechanism, receptor stimulation leads to the activation of an inward rectifying K+ channel and the inhibition of a voltage-dependent Ca2+ channel. In smooth muscle, a1A-adrenergic receptors are reported to stimulate voltage-dependent Ca2+ channels; whether a similar mechanism operates in brain remains to be determined. These various actions are thought to be mediated via direct G-protein coupling: free ba subunits bind to and gate the channel. In the other mechanism, receptor stimulation leads to inhibition of adenylyl cyclase. This action is thought to be mediated primarily via receptor-G-protein interaction and the generation of free Gai/o subunit, which then binds to and inhibits adenylyl cyclase. Inhibition of adenylyl cyclase would then lead to some of the electrophysiological effects of receptor stimulation via inhibition of cAMP-dependent phosphorylation of various types of channels and pumps, depending on the neuronal cell type involved, as described above.

Free bg subunits generated by receptor-G-protein interactions may also contribute to adenylyl cyclase inhibition, at least for some forms of the enzyme (e.g., the calmodulin-sensitive type I, III, and VIII enzymes) expressed in some cell types (8). Other forms of adenylyl cyclase (e.g., the calmodulin-insensitive type II and IV enzymes) have been shown to be stimulated by free bg subunits in vitro. This leads to the theoretical possibility that receptors coupled to Gi and Go might even stimulate adenylyl cyclase in some cell types. Although this has not been directly demonstrated for the a2-adrenergic or D2-4-dopamine receptors, such a mechanism could explain reports of a2-adrenergic receptor enhancement of cAMP responses in brain.

There are isolated reports that some of these same catecholamine receptors may activate the phosphatidylinositol system in vitro (41). These actions are reported to occur via coupling with Gi/o through mechanisms analogous to those described below for Gq.

Receptors Coupled to Gq

Of the large number of neurotransmitter receptors known to activate phosphatidylinositol hydrolysis, only one class of catecholamine receptor, the a1-adrenergic receptor, is known to produce its physiological actions primarily via this second messenger pathway (41). In most cell types, neurotransmitter receptor-induced activation of the phosphatidylinositol pathway is mediated via pertussis toxin-insensitive G-proteins (Gq), although in some cases toxin-sensitive G-proteins (Gi and/or Go) may be involved (40). Despite the fact that activation of several subtypes of the a1-adrenergic receptor has been shown to lead to phosphatidylinositol hydrolysis in numerous tissues, the specific subtype of Gq involved in various neuronal cell types remains unknown in most cases.

Activation of the phosphatidylinositol pathway could theoretically lead to regulation of ion channels via several mechanisms, although the specific mechanisms pertinent to the a1-receptor are not known. Activation of the phosphatidylinositol pathway would lead, via the generation of diacylglycerol and inositol triphosphate-mediated release of Ca2+ from internal stores, to the activation of protein kinase C and to the phosphorylation and regulation of many types of channels and pumps, depending on the cell type. Release of Ca2+ from internal stores would also lead to activation of Ca2+/calmodulin-dependent protein kinases and the subsequent phosphorylation and regulation of other channels and pumps. In addition, release of Ca2+ from internal stores would influence directly Ca2+-activated K+ channels.

a1-adrenergic receptors also indirectly activate the cAMP system in brain (7). While activation of a1-adrenergic receptors alone has little or no effect, a1-adrenergic receptor activation enhances the cAMP response to receptors that couple to Gs, including the b-adrenergic and vasoactive intestinal peptide receptors. This may occur via formation of inositol triphosphate, elevated Ca2+ levels, and activation of protein kinase C. Alternatively, it could involve activation of adenylyl cyclase by free G-protein bg subunits released upon receptor-G-protein coupling to Gq.

Activation of a1-adrenergic receptors is also known to increase cellular levels of cGMP in nervous tissue (41). The most likely mechanism is via Ca2+-induced activation of nitric oxide synthase and the subsequent activation of guanylyl cyclase by nitric oxide. The physiological consequences of a1-adrenergic receptor stimulated increases in the cGMP and nitric oxide pathways remain to be determined. However, it should be mentioned that certain ion channels are known to be phosphorylated and regulated by cGMP-dependent protein kinase or to be gated directly by cGMP (30,31).

INTRACELLULAR SYSTEMS COUPLED TO CATECHOLAMINE RECEPTORS: DISTAL EFFECTS

This section discusses the diverse effects that catecholamine receptors exert on target cells through the regulation of intracellular messenger pathways. As discussed in the preceding section and in other Chapters of this volume (see, Grace and Bunney, Foote and Aston-Jones, and Valentino and Aston-Jones), catecholamine regulation of the electrical properties of target neurons are important for mediating the most rapid effects of catecholamines in the brain. However, in addition to these rapid effects, catecholamines exert many other actions on their target neurons that produce short- and long-term modulatory effects on neuronal function (see Figure 1). These modulatory effects, which are mediated predominantly if not solely via intracellular messenger pathways, include regulation of the activity of receptors, ion channels, second messenger effector enzymes, and neurotransmitter synthetic enzymes, as well as regulation of the synthesis and degradation of these and other neuronal proteins. In fact, these modulatory effects can be viewed as catecholamine-mediated neural plasticity, some examples of which are discussed below. Given that the formation of mental symptoms, and their reversal in response to psychotropic drug treatment and other therapies, are gradual, it is possible that these modulatory effects of catecholamines are more relevant for psychiatric phenomena than the regulation of ion channels or pumps per se. The physiological relevance of these modulatory processes to specific psychiatric phenomena is the subject of a later Chapter (Nestler and Duman, this volume).

Short-Term Modulatory Processes

Activation of catecholamine receptors exerts short-term modulatory effects on several cellular systems. In most cases, these actions are mediated through the regulation of specific protein kinases and protein phosphatases. A few examples of such short-term modulatory effects are discussed below. The reader is referred elsewhere for a more detailed discussion (22).

Regulation of receptor function. Activation of catecholamine receptors can trigger numerous processes that influence the functional state of that receptor system. Agonist treatment has been shown to alter receptor affinity for its ligand, receptor coupling to G-proteins, receptor accessibility to the extracellular space (e.g., sequestration and internalization), receptor degradation, and receptor synthesis. Receptor phosphorylation by multiple protein kinases appears to mediate many of these phenomena. These processes are best established for the b-adrenergic receptor and are presented in greater detail below.

Regulation of neurotransmitter metabolism. Activation of catecholamine receptors can influence the ability of target neurons to synthesize their own neurotransmitter. One mechanism by which this is achieved is through the phosphorylation of neurotransmitter synthetic enzymes (Figure 4). This is best established for tyrosine hydroxylase, the rate-limiting enzyme in the synthesis of the catecholamines, which is phosphorylated and activated by cAMP-dependent protein kinase, Ca2+/calmodulin-dependent protein kinase II, and protein kinase C, as well as by MAP-kinase and probably other second messenger-independent protein kinases (30,46). In most cases, phosphorylation increases the Vmax of tyrosine hydroxylase (i.e., increases the maximal catalytic activity of a single enzyme molecule) or the affinity of the enzyme for its pterin co-factor (which would make the enzyme more active at subsaturating concentrations of co-factor). Dephosphorylation of the enzyme (achieved via inhibition of the cAMP or Ca2+ pathways or via activation of a phosphatase as discussed above) could mediate the ability of a2-adrenergic or D2-dopamine autoreceptors to reduce catecholamine synthesis. Phosphorylation of tyrosine hydroxylase also mediates the ability of many other types of neurotransmitter receptors, that act through the cAMP or Ca2+ systems, to rapidly regulate tyrosine hydroxylase activity and, as a result, the capacity of catecholaminergic neurons to synthesize their neurotransmitter. This provides a critical homeostatic control mechanism that enables catecholaminergic neurons to alter their functional activity in response to a variety of synaptic inputs.

The other major mechanism by which activation of catecholamine receptors can influence the synthesis of other neurotransmitters is by regulating the expression of peptide neurotransmitters in target neurons, as discussed below.

Regulation of neurotransmitter release. Activation of catecholamine receptors located on presynaptic nerve terminals can regulate the release of neurotransmitter from those terminals (Figure 4). One mechanism probably involves regulation of the phosphorylation of nerve terminal ion channels or pumps and the subsequent regulation of Ca2+ entry into the terminals and the release of neurotransmitter. For example, by hyperpolarizing nerve terminals, a2-adrenergic and D2-4-dopamine receptors would be expected to reduce neurotransmitter release via activation of K+ channels or inhibition of Ca2+ channels.

Another critical mechanism appears to involve the phosphorylation of a family of synaptic vesicle-associated proteins, the best-studied of which are the synapsins (Figure 4) (15). The synapsins comprise a family of phosphoproteins, present in virtually all nerve terminals in brain, that are phosphorylated by cAMP-dependent and Ca2+/calmodulin-dependent protein kinases. Synapsin phosphorylation increases the amount of neurotransmitter released from nerve terminals in response to physiological stimuli. Phosphorylation of synapsins appears to augment neurotransmitter release by altering their binding affinity for synaptic vesicles and other cytoskeletal proteins. Such changes in synapsin binding affinities are thought to regulate synaptic vesicle traffic within nerve terminals and, possibly, the process of exocytosis. Phosphorylation of the synapsins is regulated by a number of neurotransmitters, which influence cAMP or Ca2+ levels in nerve terminals, and appears to mediate the ability of these neurotransmitters to produce relatively long-lasting changes in the functional activity of those terminals. For example, stimulation of b-adrenergic and D1-dopamine receptors, via activation of the cAMP pathway, has been shown to stimulate synapsin phosphorylation in a variety of neural preparations, and this probably contributes to the ability of these receptors to increase neurotransmitter release. In contrast, a2-adrenergic and D2-4-dopamine receptor inhibition of the cAMP and Ca2+ pathways and of synapsin phosphorylation would be expected to contribute to receptor inhibition of neurotransmitter release.

Long-Term Modulatory Processes

Activation of catecholamine receptors can also result in more long-term regulation of neuronal function. A brief overview of these processes is given below. More detailed information of these mechanisms, and their role in mediating the chronic actions of psychotropic drug treatments on the brain, is presented in a later Chapter (Nestler and Duman, this volume).

Regulation of protein levels. A prominent mechanism by which catecholamines exert long-term effects on the brain involves regulation of the types and amounts of proteins present in target neurons. Thus, in addition to regulation by covalent modifications, the functional activity of a given protein in a neuron can be influenced by the amount of protein that is expressed. Catecholamine-induced alterations in protein levels can thereby exert profound and persistent changes in target neurons. As some examples, altered levels of a neurotransmitter receptor would produce long-lasting changes in the neuron's responsiveness to that neurotransmitter, altered levels of an ion channel would produce long-lasting changes in the neuron's electrical excitability, and altered levels of a neurotransmitter synthetic enzyme would produce long-lasting changes in the neuron's capacity to transmit its signals to subsequent neurons.

Catecholamine regulation of protein levels appears to be achieved by the regulation of every conceivable step involved, including alterations in gene transcription, the processing of mRNA and its transport into the cytoplasm, the stability and translatability of mRNA, the post-translational processing of proteins and their localization to specific subcellular compartments, as well as their enzymatic degradation. Once again, protein phosphorylation appears to be the most important mechanism by which each of these processes is influenced by extracellular signals. This occurs through the phosphorylation, and consequent regulation of the physiological activity, of specific regulatory proteins involved in transcription, translation, and post-translational processing.

Through these various mechanisms it is becoming increasingly apparent that catecholamines influence the expression of diverse types of neuronal proteins. One prominent example is regulation of the expression of peptide neurotransmitters and growth factors. For example, based on the actions of haloperidol (as a D2 receptor antagonist) and stimulants (as indirect dopamine agonists), it is clear that dopamine can regulate the expression of proenkephalin and other neuropeptides in specific brain regions in vivo (see 22).

One aspect of gene expression which has received a great deal of attention is regulation of nuclear transcription factors. Transcription factors are proteins that bind to specific sequences of DNA present in certain genes and thereby increase or decrease the rate of transcription of those genes (22,28). One class of transcription factor (e.g., Fos and Jun), encoded by immediate early genes, would appear to have an important role in long-term modulatory processes, since they are induced rapidly in brain in response to a variety of extracellular stimuli, including activation of adrenergic and dopaminergic receptors (28). Another class of transcription factor, an example of which is the Cyclic AMP Response Element Binding protein or CREB, is also rapidly regulated in brain, but in this case activation occurs by phosphorylation of the transcription factor by cAMP, as well as Ca2+, -dependent protein kinases. Catecholamine receptors that result in alterations in the cAMP or Ca2+ pathways would be expected to result in altered CREB phosphorylation and altered transcriptional activity (22). Regulation of these transcription factors, and other pathways for regulation of gene expression in the brain, are discussed in further detail elsewhere in this volume (see Nestler and Duman).

Regulation of neuronal growth and differentiation. It is likely that catecholamines influence cell growth, differentiation, and movement (including axoplasmic transport and sprouting of dendrites and axons) in their target cells, although the details of the mechanisms involved remain obscure. This view is based on the ability of catecholamines to regulate, as discussed above, the critical intracellular messenger pathways known to control these cellular processes. Moreover, recent studies have demonstrated that activation of adrenergic and dopaminergic receptors increases the expression of neurotrophins in primary neuronal cultures and in brain (27,35,45). These findings, while preliminary, support trophic-like consequences of catecholamine receptor activation and suggest that regulation of neurotropins may be one of the mechanisms involved.

Regulation of learning and memory. Every instance in which a protein is phosphorylated, or the amount of a protein changes, can be viewed as molecular memory. This is because a change in a protein's phosphorylation or amount leads to a change in that protein's, and hence its neuron's, function--a molecular record of that neuron's prior experience. These individual examples of molecular memory then accumulate to lead successively to changes in the physiological properties of individual neurons, to changes in the physiological properties of larger neural networks, and ultimately to changes in the behavior of the organism. It is well-known that catecholamines can influence processes of learning and memory at the behavioral level (see Valentino and Aston-Jones, Robbins and Everitt, this volume). Identification of the myriad molecular steps underlying such phenomena is a major challenge for the future (see Nestler and Duman, this volume).

Regulation of Intracellular Messenger Pathways

The ability of catecholamine receptors to initiate specific intracellular cascades means that these receptors also influence, albeit less directly, numerous other intracellular messenger pathways in their target cells. This is based on the now extensive evidence that most of the protein components of intracellular messenger systems are themselves regulated by phosphorylation. This permits extraordinarily complex cross-talk between signaling pathways, which permits cells to coordinate their responses to environmental stimuli (22,30,31).

Several types of G-proteins have been reported to undergo phosphorylation by a variety of protein kinases. Proteins that control the synthesis of the cyclic nucleotide second messengers (adenylyl cyclase and guanylyl cyclase), as well as the degradation of cyclic nucleotides (phosphodiesterases), are regulated by phosphorylation. Similarly, proteins that control intracellular Ca2+ levels or the phosphatidylinositol system (e.g., phospholipase C, Ca2+ channels, the Ca2+/Mg2+-ATPase pump, the inositol triphosphate receptor) are regulated by phosphorylation. Moreover, phospholipase A2, which generates arachidonic acid metabolites (e.g., prostaglandins) that modulate cyclic nucleotide and Ca2+ levels, is also subject to phosphorylation. Many protein kinases are themselves phosphorylated and regulated by other protein kinases, and protein phosphatase type 1 is regulated by protein phosphatase inhibitor proteins, which are regulated by phosphorylation. In addition, most, and possibly all, protein kinases undergo autophosphorylation, whereby they phosphorylate themselves.

It is clear from the above discussion that each second messenger system in the brain influences all the others. This means that although the systems are drawn as distinct pathways in Figure 3, they do not operate as distinct pathways, but operate instead as a complex web of interacting pathways (see Figure 1). Thus, any time a catecholamine or other neurotransmitter produces its primary effect on one second messenger system, many other systems will also be influenced eventually, with such interactions mediated for the most part through protein phosphorylation. For example, b-adrenergic and D1/5-dopamine receptors, which produce their primary effects through the activation of the cAMP pathway, could potentially influence the Ca2+ and phosphatidylinositol systems via cAMP-dependent phosphorylation of: G-proteins, phospholipases, Ca2+ and K+ channels, electrogenic pumps, Ca2+-dependent protein kinases, and the inositol triphosphate receptor, as well as the many proteins that can be phosphorylated by both cAMP-dependent and Ca2+-dependent protein kinases.

In addition, there is also potential for interactions between these catecholamine receptor-activated pathways and the second messenger-independent protein kinase pathways, including those regulated by the receptor-associated protein tyrosine kinases (see Figure 1 and Figure 2). Thus, second messenger kinases can phosphorylate and activate Raf-kinase and possibly other MAP-kinase kinase kinases. These enzymes are the first step in the MAP-kinase pathway, and function, as stated earlier in this Chapter, by phosphorylating and activating MAP-kinase kinases, which, in turn, function by phosphorylating and activating MAP-kinases. This provides a mechanism whereby catecholamine receptor-activated second messenger pathways may interact with and regulate, in either a stimulatory or inhibitory manner, the same pathways regulated by neurotrophins and other growth factor receptors.

Examples of Distal Actions of Catecholamine Receptor Activation

b-adrenergic receptor. Activation of the b-adrenergic receptor (bAR) results, via Gs, in activation of the cAMP second messenger system and thereby initiates a cascade of intracellular events regulated by this pathway. Some of these events, which have been studied in detail, include receptor desensitization and down-regulation, activation and translocation of cAMP-dependent protein kinase, regulation of transcription factors, and expression of specific target genes (see Figure 5) (23,24,26).

Desensitization of the bAR has been studied extensively and involves alterations of practically every point of the bAR-coupled cAMP system, including regulation of receptor and effector protein expression (Figure 4). Agonist binding to the bAR leads to formation of cAMP and activation of cAMP-dependent protein kinase which in turn phosphorylates the receptor and functionally uncouples it from Gs. This involves phosphorylation of specific serine residues in the third cytoplasmic and carboxy-terminus domains of the receptor and results in a reduced sensitivity to agonist, as measured by a rightward shift in dose response. Continued exposure to high concentrations of agonist results in phosphorylation of the receptor by a second protein kinase, termed bAR kinase (bARK), which only phosphorylates the agonist- activated form of the receptor at serine residues in the carboxy-terminus domain. Once these sites are phosphorylated, another protein, b-arrestin, binds to this domain of the receptor and competes with Gs; this reduces the maximal response to agonist stimulation. More recent studies indicate that various forms of bAR kinase and b-arrestin are not specific for the bAR and probably mediate agonist-induced desensitization of many other G protein-coupled receptors, including otehr catecholamine receptors.

Loss of bAR binding sites from the cell membrane involves at least two mechanisms, receptor sequestration and degradation (down- regulation). In the presence of high concentrations of agonist, receptors are internalized and are then sequestered into intracellular vesicles; these receptors are accessible to hydrophobic ligands, which penetrate the vesicle membrane, but not to hydrophilic ligands. This pool of receptors is available for recycling back to the plasma membrane, apparently upon receptor dephosphorylation. Alternatively, internalized receptors may be transported to lysosomes where they are degraded rapidly.

Down-regulation of the bAR in response to agonist exposure also occurs via regulation of receptor expression, an action apparently mediated through the cAMP-dependent protein phosphorylation pathway. The amount of receptor protein expressed appears to be regulated by changes in bAR mRNA stability as well as in the rate of bAR gene transcription, depending on the cell type or tissue being examined. Agonist treatment is reported to regulate both b2AR mRNA stability and gene expression in a smooth muscle cell line (5,16,17), whereas agonist treatment regulates b1AR gene expression, with no change in mRNA stability, in a glioma cell line (18,19). In either case, decreased expression of bAR mRNA and protein would contribute to the down-regulation of receptor in response to agonist treatment. These mechanisms of receptor regulation may have particular relevance to the actions of psychotropic drugs, such as antidepressants, which require several weeks of treatment (see Nestler and Duman, this volume).

In some systems, prolonged activation of the bAR has been shown to result in altered levels of proteins, in addition to the receptor itself, in the bAR signal transduction pathway. This would contribute further to agonist-induced down-regulation of bAR function. Such bAR-induced down-regulation has been observed for G-proteins and cAMP-dependent protein kinase (see 17,30).

The physiological actions of bAR activation are presumably mediated by many target proteins. bAR-stimulation of the cAMP pathway would be expected to lead to the phosphorylation and regulation of numerous cellular proteins and consequently to the regulation of numerous cellular processes, as discussed in previous sections of this Chapter. This would include changes in gene expression via regulation of transcription factors (e.g., Fos, CREB). One mechanism by which bAR activation may lead to transcription factor regulation is by inducing the nuclear translocation of cAMP-dependent protein kinase (see 32). While regulation of these transcription factors can be used as a marker for studying the brain regions influenced by bAR activation, at this time there is little known about the specific target genes for these transcription factors in specific neuronal cell types in the brain. Potential target genes of the bAR, as mentioned above, are those for neuropeptides and neurotrophins, which possess DNA response elements sensitive to these transcription factors and show regulation by bAR ligands in cultured cells and brain.

The D1 -dopamine receptor and DARPP-32. Activation of the D1-dopamine receptor would be expected to result in many of the same types of effects in target neurons as outlined above for the bAR, due to the fact that the actions of both receptors are mediated via the cAMP pathway. However, an additional type of signal transduction mechanism has been elaborated for the D1 receptor. This mechanism involves DARPP-32 (dopamine and cAMP regulated phosphoprotein of 32 kD) and highlights the complex interactions, mediated via intracellular messenger pathways, that occur among neurotransmitter actions in the brain (11, 36).

DARPP-32 was discovered during a study of the regional distribution of neuronal phosphoproteins in rat brain (36). It is one of several substrates for cAMP-dependent protein kinase that are highly concentrated in the basal ganglia. DARPP-32 is phosphorylated in vitro on a single threonine residue by cAMP-dependent or by cGMP-dependent protein kinase. Phospho-DARPP-32, but not the dephospho form of the protein, is a highly potent and specific inhibitor of protein phosphatase 1. DARPP-32 is also phosphorylated on serine residues by casein kinases I and II; casein kinases are second messenger-independent serine/threonine protein kinases. Such phosphorylation influences the ability of the threonine residue to be phosphorylated by cAMP-dependent protein kinase.

DARPP-32 is highly enriched in neurons in the brain that possess D1-dopamine receptors, and it appears to be present in all such neurons. It is also present in renal tubular epithelial cells, parathyroid hormone-producing cells in the parathyroid gland, and tanocytes, all of which are known to express the D1 receptor. However, DARPP-32 is also found in several cell types that do not possess D1-dopamine receptors, where the protein is regulated by other neurotransmitters.

The state of phosphorylation of DARPP-32 can be regulated in many cell types by various hormones and neurotransmitters that activate the cAMP or cGMP pathway; one notable example is stimulation of DARPP-32 phosphorylation in striatal neurons via activation of D1-dopamine receptors. Changes in the phosphorylation state and phosphatase inhibitory activity of DARPP-32 indirectly influence the phosphorylation state of other proteins, and thereby mediate some of the effects of dopamine and other first messengers on cell function. The full spectrum of proteins regulated by DARPP-32 phosphorylation in this way have not yet been identified, although the Na+/K+-ATPase represents one target protein. Regulation of this protein by DARPP-32 provides one mechanism by which alterations in DARPP-32 phosphorylation can lead to changes in the electrical excitability of neurons and in ion transport properties of non-excitable peripheral tissues.

Several types of physiological actions for DARPP-32 can be envisioned. First, DARPP-32 phosphorylated and activated in response to dopamine (or another first messenger) and cAMP (or cGMP), can enhance the signal for these messengers by reducing the dephosphorylation of other substrates for the same protein kinase. Support for this scheme comes from recent analysis of DARPP-32 knock out mice (11). Second, DARPP-32 can reduce the dephosphorylation of substrate proteins for other protein kinases and, in so doing, can mediate the effects of first- and second-messenger systems on one another. Third, DARPP-32 through its phosphorylation by cAMP (or cGMP)-dependent protein kinase and its dephosphorylation by Ca2+/calmodulin-dependent protein phosphatase (calcineurin) can integrate certain physiological effects of first messengers that influence the cAMP and Ca2+ systems.

An example of this latter mechanism is illustrated in Figure 6. In this scheme, extracellular signals that activate the cAMP pathway would phosphorylate and activate DARPP-32, whereas extracellular signals that activate the Ca2+ pathway would dephosphorylate and inactivate DARPP-32. Changes in DARPP-32 activity would then lead to altered activity of protein phosphatase 1 and, as a result, to altered dephosphorylation of Na+/K+-ATPase, a prominent substrate for this enzyme. Changes in the phosphorylation state of the Na+/K+-ATPase would result in altered sodium transport across the cell membrane and, in excitable cells, to altered membrane potential. Considerable evidence has been obtained to support this scheme in several cell types. Moreover, the scheme can account for some of the antagonist actions of dopamine (acting through cAMP) and glutamate (acting through Ca2+) on neuronal excitability in striatal neurons (36).

TOP

Figure 1.

Schematic illustration of the role played by intracellular messenger systems in synaptic transmission in the brain. Recent studies in neuroscience have provided a complex view of synaptic transmission. These studies have focused on the involvement of intracellular messenger systems, involving coupling factors (termed G- proteins), second messengers [e.g., cAMP, cGMP, Ca2+, nitric oxide (NO), and the metabolites of phosphatidylinositol (PI) and arachidonic acid (AA)], and protein phosphorylation (involving the phosphorylation of phosphoproteins by protein kinases and their dephosphorylation by protein phosphatases), in mediating multiple actions of neurotransmitters on their target neurons.

Prominent in brain are numerous protein serine/threonine kinases that are activated directly by various second messengers, and are referred to as second messenger-dependent protein kinases. Brain also contains numerous protein serine/threonine kinases that are not regulated directly by second messengers (see Table 1). In addition, the brain contains numerous types of protein tyrosine kinases, some of which reside in the receptors for most growth factors, others that are not associated with growth factor receptors (see Table 1). These various protein kinases are all highly regulated by extracellular stimuli. The second messenger-dependent protein kinases are regulated by receptor-second messenger pathways as shown in Figure 3. The receptor-associated protein tyrosine kinases are activated upon growth factor binding to the receptor. The second messenger-independent protein serine/threonine kinases and the protein tyrosine kinases that are not receptor-associated seem to be regulated indirectly by second messengers and second messenger-dependent protein kinases and by protein tyrosine kinases, although the precise mechanisms are not yet known in most cases. The brain also contains numerous types of protein serine/threonine and protein tyrosine phosphatases, which are also subject to regulation by extracellular stimuli.

The figure illustrates three major roles subserved by these intracellular messenger pathways. In some cases, intracellular messengers mediate the actions of some neurotransmitters in opening or inhibiting particular ion channels. However, intracellular messengers mediate most of the many other actions of neurotransmitters on their target neurons. Some are relatively short-lived and involve modulation of the general metabolic state of the neurons, their ability to synthesize or release neurotransmitter, and the functional sensitivity of their various receptors and ion channels to various synaptic inputs. Others are relatively long-lived and are achieved through the regulation of gene expression in the target neurons. Thus, neurotransmitters, through the regulation of intracellular messenger pathways and alterations in gene transcription and protein synthesis, alter the numbers and types of receptors and ion channels in target neurons, the functional activity of the intracellular messenger systems in those neurons, and even the shape and numbers of synapses the neurons form. The figure is drawn to illustrate the amplification that intracellular messenger systems can give to neurotransmitter action. Thus, a single event of a neurotransmitter binding to its receptor (the 1st messenger level) can act through the 2nd, 3rd, 4th, etc. messenger levels to produce an increasingly wider array of physiological effects.

TOP

Figure 2.

Schematic illustration of the mechanisms by which G-protein coupled receptors and receptor-associated protein tyrosine kinases function in the brain. G-protein coupled receptors and receptor-associated protein tyrosine kinases employ distinct mechanisms for activation of intracellular signal transduction pathways. Activation of G-protein coupled receptors (R) leads to association of the receptor with a heterotrimeric G-protein comprised of single a, b, and g subunits. This leads to the binding of GTP to the a subunit and the displacement of GDP from the subunit. GTP binding induces the generation of free a subunit by causing the dissociation of the a subunit from its bg subunits and the receptor. Free a subunit bound to GTP, and free ba subunits, are functionally active and directly regulate a number of effector proteins, which, depending on the type of a and ba subunits and cell involved, can include ion channels, adenylyl cyclase, phospholipase C, and phosphodiesterases. GTPase activity intrinsic to the a subunit hydrolyzes GTP to GDP. This leads to the reassociation of the a and bg subunits, which, along with the dissociation of ligand from the receptor, leads to restoration of the basal state.

Nerve growth factor receptors (i.e., Trks), as well as other growth factor receptors, stimulate protein tyrosine kinase activity intrinsic to the receptor. Binding of two molecules and dimerization of Trk receptors is necessary for receptor activation. This leads to phosphorylation of the receptor itself and adaptor proteins with Src homology (SH) domains, such as Shc. The SH proteins, or phosphorylated receptor in some cases, then associates with Grb2. Grb2 then complexes with Sos (and related Ras effector proteins) via SH domains to promote GTP binding to Ras. The function of Ras is also negatively regulated by GAP (GTPase activating protein), which stimulates Ras GTPase activity and the hydrolysis of GTP to GDP. Ras, bound to GTP, is functionally active and directly stimulates Raf by an unknown mechanism. Raf, a protein serine/threonine kinase, phosphorylates and activates MEK (Mitogen and Extracellular-regulated Kinase), which then activates MAP kinase (Microtubule Associated Protein kinase; also referred to as ERK, Extracellular signal-Regulated Kinase) by phosphorylating it on threonine and tyrosine residues. Raf and possibly other MEK kinases can also be activated by second messenger-dependent protein kinases. This provides a pathway by which G-protein coupled receptors can influence this major growth factor-Ras cascade.

TOP

Figure 3.

Schematic illustration of major second messenger pathways in the brain. Gs and Gi/o, respectively, mediate the ability of neurotransmitter receptors (R) to activate or inhibit adenylyl cyclase, the enzyme that catalyzes the synthesis of cAMP. Also shown in the figure is the ability of G-protein bg subunits, released potentially by any type of G-protein, or Ca2+/calmodulin to stimulate or inhibit different forms of adenylyl cyclase. Gq and perhaps Gi/o mediate the ability of neurotransmitter receptors to regulate phospholipase C (PLC), which metabolizes phosphatidylinositol (PI) into the second messengers inositol triphosphate (IP3) and diacylglycerol (DAG). IP3 then acts on specific IP3 receptors (IP3R) to increase intracellular levels of free Ca2+ (also a second messenger in brain) by releasing Ca2+ from internal stores. Increased levels of intracellular Ca2+ also result from the flux of Ca2+ across the plasma membrane through Ca2+ and other ion channels, a flux stimulated by nerve impulses and certain neurotransmitters. As discussed in the text, G-proteins mediate many of the actions of neurotransmitters on such channels. Increased levels of Ca2+ activate nitric oxide synthase (NOS), leading to increased levels of nitric oxide and the activation of cytoplasmic guanylyl cyclase (the enzyme that catalyzes the synthesis of cGMP) and of ADP-ribosyltransferases (not shown). Other forms of guanylyl cyclase reside in specific plasma membrane receptors.
These second messengers, in turn, activate specific types of protein kinases. Brain contains one major type of cAMP-dependent protein kinase and of cGMP-dependent protein kinase, although subtypes of these enzymes are differentially expressed throughout the brain. These enzymes phosphorylate a specific array of substrate proteins, which can be considered third messengers. cAMP-dependent protein kinase has a broad substrate specificity, that is, it phosphorylates many substrate proteins and mediates most of the numerous second messenger actions of cAMP in the nervous system. The substrate specificity of cGMP-dependent protein kinase appears to be less broad, although by analogy with the cAMP system it is likely that it mediates many of the second messenger functions of cGMP. Brain contains two major classes of Ca2+-dependent protein kinase. One is activated by Ca2+ and calmodulin and is referred to as Ca2+/calmodulin-dependent protein kinase. Brain contains at least six distinct types of this enzyme: 1-4) Ca2+/calmodulin-dependent protein kinases I, II (several subtypes of this enzyme are known), III, and IV; 5) phosphorylase kinase; and 6) myosin light chain kinase. The other major class is activated by Ca2+ in conjunction with DAG and various phospholipids and is referred to as Ca2+/DAG-dependent protein kinase or protein kinase C; there are at least 7 closely-related variants of this enzyme present in the brain. Protein kinase C and Ca2+/calmodulin-dependent protein kinases II and perhaps I and IV have broad substrate specificities (as indicated by the multiple arrows in the figure) and each probably mediates many of the numerous second messenger actions of Ca2+ in the nervous system. [The figure also illustrates that some of the second messenger actions of Ca2+ in brain are mediated through proteins other than protein kinases.] Not shown in the Figure is the fact that some protein phosphatases are also subject to regulation by second messengers. For example, protein phosphatase 2B, or calcineurin, is activated upon binding Ca2+/calmodulin. Phosphorylation of substrate proteins by these various second messenger-dependent protein kinases alters their physiological activity in such a way as to lead to the biological responses of the extracellular messengers either directly or indirectly through intervening fourth, fifth, sixth, etc. messengers. Modified from Hyman and Nestler, 1993.

TOP

Figure 4.

Schematic illustration of the mechanisms contributing to receptor regulation of neurotransmitter synthesis and release. The mechanisms which underlie the ability of catecholamine autoreceptors to inhibit neurotransmitter synthesis probably involve a combination of second messenger pathways and phosphoproteins. Illustrated in this figure are the effects of dopamine (DA) (D2) or norepinephrine (NE) (a2) autoreceptors on the cAMP and Ca2+ second messenger systems and ion channels. The activity tyrosine hydroxylase (TH), the rate limiting enzyme in the synthesis of both dopamine and norepinephrine, is increased by phosphorylation, which can occur in response to activation of a number of different kinases, including cAMP and Ca2+-dependent kinases (CaMK and PKC). Activation of D2 or a2 autoreceptors inhibits adenylyl cyclase and would be expected to reduce cAMP-dependent protein kinase activity and phosphorylation of TH. These autoreceptors may also inhibit Ca2+ channels and decrease the activity of Ca2+-dependent protein kinases and the level of phospho-TH.

One mechanism by which D2 and a2 autoreceptors may influence the release of neurotransmitters is by inhibition of Ca2+ channels. D2 or a2 autoreceptors may also hyperpolarize the nerve terminal by activation of K+ channels. Another mechanism by which these receptors may influence neurotransmitter release is by regulation of synaptic-vessicle associated proteins, such as synapsins. The synapsins are a family of proteins found in nerve terminals that are phosphorylated by cAMP and Ca2+/calmodulin-dependent protein kinases. Phosphorylation of synapsin results in a greater amount of neurotransmitter being released in response to a physiological stimulus. Decreased activity of cAMP or Ca2+/calmodulin-dependent kinases by activation of autoreceptors would be expected to decrease the phosphorylation state of synapsin and decrease the release of neurotransmitter.

TOP

Figure 5.

Schematic illustration of multiple mechanisms underlying regulation of the bAR. One of the best characterized actions of bAR activation is regulation of the receptor itself, which involves several points controlling receptor function and expression. 1) bAR stimulation of the cAMP system results in phosphorylation of the receptor by cAMP-dependent protein kinase, which leads to uncoupling of the receptor from Gs. The activated protein kinase would also phosphorylate many other proteins, not shown in the Figure, which would then mediate many of the actions of bAR activation. 2) Prolonged activation of the receptor leads to phosphorylation by another kinase, bAR kinase (bARK), which only phosphorylates the agonist-activated form of the receptor. This results in binding of the receptor to b-arrestin, which competes with Gs and thereby inhibits bAR-stimulation of adenylyl cyclase (AC). 3) Loss of bARs from the membrane occurs when receptors are internalized and sequestered into intracellular vesicles. This pool of receptors is then available for either recycling back to the membrane or degradation. Such sequestration, internalization, degradation, and membrane re-insertion may be mediated via receptor phosphorylation and dephosphorylation involving the cAMP and/or bARK pathways. Another mechanism by which receptor activation leads to down regulation of the bAR is via regulation of receptor mRNA levels, which may occur by two primary mechanisms. 4) The level of receptor mRNA is regulated by the stability or half-life of the mRNA. Although the mechanisms responsible for regulation of mRNA stability have not been identified, they may also involve cAMP-dependent protein kinase. 5) The level of receptor mRNA is also regulated via changes in bAR gene transcription. This effect is mediated by the cAMP pathway and appears to involve the translocation of cAMP-dependent protein kinase catalytic subunit into the nucleus and the phosphorylation of constitutively expressed transcription factors (e.g., CREB). It might also depend upon the subsequent induction of other transcription factors [e.g., immediate early gene (IEG) products such as c-Fos]. In addition to regulation of receptor gene transcription, such regulation of transcription factors would mediate the effects of bAR activation on the expression of many other genes, for example, those for G-proteins, cAMP-dependent protein kinase, neurotrophins, and neuropeptides. This, in turn, would mediate many of the more long-term effects of bAR activation on brain function.

TOP

Figure 6.

Scheme illustrating hypothetical role of DARPP-32, in mediating effects of first messengers with opposing physiological actions, in regulation of Na+/K+-ATPase activity. This scheme, involving bidirectional control of Na+/K+-ATPase activity, may be applicable to various tissues, including brain and kidney. cAMP-PK, cAMP-dependent protein kinase; PKC, protein kinase C. (From Pessin et al., 1993).

TOP

REFERENCES

1. Beavo JA(1988) Multiple isozymes of cyclic nucleotide phosphodiesterase. Adv Sec Mess Phosphoprotein Res 22: 1-37.

2. Berridge MJ (1993) Inositol triphosphate and calcium signaling. Nature 361: 315-325.

3. Blenis, J (1993) Signal transduction via the MAP kinases: Proceed at your own RSK. Proc Natl Acad Sci USA 90: 5889-5892.

4. Bredt DS, Snyder SH (1992) Nitric oxide, a novel neuronal messenger. Neuron 8:3-11.

5. Collins S, Bouvier M, Bolanowski MA, Caron MG, and Lefkowitz RJ (1989) cAMP stimulates transcription of the b2-adrenergic receptor gene in response to short-term agonist exposure. Proc Natl Acad Sci USA 86: 4853-4857.

6. Davis RJ (1993) The mitogen-activated protein kinase signal transduction pathway. J Biol Chem 268:14553-14556.

7. Duman RS, Karbon EW, Harrington C, and Enna SJ (1986) An examination of the involvement of phospholipases A2 and C in the a-adrenergic and g-aminobutyric acid receptor modulation of cyclic AMP accumulation in rat brain slices. J Neurochem 47: 800-810.

8. Duman RS, Nestler EJ (1999) Cyclic Nucleotides. In: Basic Neurochemistsry: Molecular, Cellular and Medical Aspects, 6th Ed., ed. by G. Siegel, B. Agranoff, R. Albers, S. Fisher and M. Uhler, Lippincott-Raven Press, New York, 20:433-451.

9. Duman RS, Terwilliger RZ, Nestler EJ (1991) Endogenous ADP-ribosylation in brain: initial characterization of substrate proteins. J Neurochem 57:2124-2132.

10. Feinstein P, Schrader K, Bakalyar H, et al. (1991) Molecular cloning and characterization of a Ca2+/calmodulin-insensitive adenylyl cyclase from rat brain. Proc Natl Acad Sci USA 88:10173-10177.

11. Fienberg AA, Hiroi N, Mermelstein P, Song W-J, Snyder GL, Nishi A, Cheramy A, O'Callaghan JP, Miller D, Cole DG, Corbett R, Haile C, Cooper DC, Onn S, Grace A, Ouimet CC, White FJ, Hyman SE, Surmeier DJ, Girault JA, Nestler EJ, Greengard P (1998) DARPP-32, regulator of the efficacy of dopaminergic neurotransmission. Science 281:838-842.

12. Gao B, Gilman AG (1991) Cloning and expression of a widely distributed (type IV) adenylyl cyclase. Proc Natl Acad Sci USA 88:10178-10182.

13. Glatt, CE, Snyder, SH (1993) Cloning and expression of an adenylyl cyclase localized to the corpus striatum. Nature 361: 536-538.

14. Greengard P, Jen J, Nairn AC, Stevens CF (1991) Enhancement of the glutamate response by cAMP-dependent protein kinase in hippocampal neurons. Science 253:1132-1335.

15. Greengard P, Valtorta F, Czernik AJ, Benfenati F (1993) Synaptic vesicle phosphoproteins and regulation of synaptic function. Science 259:780-785.

16. Hadcock JR, Wang H-Y, and Malbon CC (1989) Agonist-induced destabilization of b-adrenergic receptor mRNA. J Biol Chem 264: 19928-19933.

17. Hadcock JR and Malbon CC (1993) Agonist regulation of gene expression of adrenergic receptors and G proteins. J Neurochem 60: 1-9.

18. Hosoda, K., Feussner, G., Fishman, P.H., and Duman, R.S. (1992) b1-adrenergic receptor down-regulation in C6 glioma cells: evidence for inhibition of b1AR gene transcription by a transient repressor. Soc Neurosci Abstr 18, 816.

19. Hough C, Chuang D-M (1990) Differential down-regulation of b1- and b2-adrenergic receptor mRNA in C6 glioma cells. Biochem Biophys Res Commun 170, 46-52.

20. Huganir RL, Greengard P (1990) Regulation of neurotransmitter receptor desensitization by protein phosphorylation. Neuron 5:555-567.

21. Hunter T, Sefton BM, Eds. (1991) Protein phosphorylation. Parts A and B. Methods in Enzymology, vols. 200 and 201, Academic Press, NY.

22. Hyman SE, Nestler EJ. (1993) The Molecular Foundations of Psychiatry, Washington DC: American Psychiatric Press.

23. Kobilka B (1992) Adrenergic receptors as models for G protein-coupled receptors. Annu Rev Neurosci 15:87-114.

24. Lefkowitz RJ (1993) G protein coupled receptor kinases. Cell 74:409-412.

25. Leidenheimer NJ, Browning MD, Harris RA (1991) GABAA receptor phosphorylation: multiple sites, actions, and artifacts. Trends Pharmacol Sci 12:84-87.

26. Liggett SB, Ostrowski J, Chesnut LC, Kurose H, Raymond JR, Caron MG, Lefkowitz RJ (1992) Sites in the third intracellular loop of the a2A-adrenergic receptor confer short term agonist-promoted desensitization. J Biol Chem 267:4740-4746.

27. Mocchetti, I., De Bernardi, M.A., Szekely, A.M., Alho, H., Brooker, G., and Costa, E. (1989) Regulation of nerve growth factor biosynthesis by b-adrenergic receptor activation in astrocytoma cells: a potential role of c-Fos protein. Proc Natl Acad Sci USA 86: 3891-3895.

28. Morgan JI, Curran T (1991) Stimulus-transcription coupling in the nervous system. Annu Rev Neurosci 14:421-452.

29. Nestler EJ, Duman RS (1999) G Proteins. In: Basic Neurochemistry: Molecular, Cellular and Medical Aspects, 6th Ed., ed. by G. Siegel, B. Agranoff, R. Albers, S. Fisher, and M. Uhler, Lippincott-Raven Press, New York, 20:401-414.

30. Nestler EJ, Greengard P (1984) Protein Phosphorylation in the Nervous System, J. Wiley and Sons, NY.

31. Nestler EJ, Greengard P (1999) Serine and Threonine phosphorylation and the regulation of neuronal function. In: Basic Neurochemistry, 6th Ed., ed. by G. Siegel, B. Agranoff, R. W. Albers, S. Fisher and M. Uhler, Lippencott-Raven Press, New York, 471-495.

32. Nestler EJ, Terwilliger RZ, Duman RS (1989) Chronic antidepressant administration alters the subcellular distribution of cyclic AMP-dependent protein kinase in rat frontal cortex. J Neurochem 53:1644-1647.

33. Nicoll RA (1988) The coupling of neurotransmitter receptors to ion channels in the brain. Science 241:545-551.

34. Nishizuka Y (1992) Intracellular signaling by hydrolysis of phospholipids and activation of protein kinase C. Science 258: 607-614.

35. Okazawa H, Murata M, Watanabe M, Kamei M, Kanazawa I (1992) Dopaminergic stimulation up-regulates the in vivo expression of brain-derived neurotrophic factor (BDNF) in the striatum. FEBS 313: 138-142.

36. Pessin MS, Snyder GL, Halpain S, Girault J-A, Aperia A, Greengard P (1992) DARPP-32/Protein phosphatase-1/Na+/K+ATPase System: A mechanism for bidirectional control of cell function. Proceedings of the 1992 Wenner Gren International Symposium, Stockholm, in press.

37. Piomelli D, Greengard P (1990) Lipoxygenase metabolites of arachidonic acid in neuronal transmembrane signalling. Trends Pharmacol. Sci. 11:367-373.

38. Sessker FN, Mouradian RD, Cheng J-T, Yeh HH, Liu W, Waterhouse BD (1989) Noradrenergic potentiation of cerebellar Purkinje cell responses to GABA: evidence for mediation through the b-adrenoceptor-coupled cyclic AMP system. Brain Res 499:27-38.

39. Shenolikar S, Nairn AC (1991) Protein phosphatases: Recent progress. Adv Second Messenger Phosphoprotein Res 23:1-121.

40. Simon MI, Strathman MP, Gautam N (1991) Diversity of G proteins in signal transduction. Science 252:802-808.

41. Summers RJ, McMartin LR. (1993) Adrenoceptors and their second messenger systems. J Neurochem 60:10-23.

42. Taussig R, Iniguez-Lluhi JA, Gilman AG (1993) Inhibition of adenylyl cyclase by Gia. Science 261:218-221.

43. Williams MB, Li X, Gu X, Jope RS 1992) Modulation of endogenous ADP-ribosylation in rat brain. Brain Res 592:49-56.

44. Yuen PST, Garbers DL (1992) Guanylyl cyclase-linked receptors. Annu Rev Neurosci 15:193-225.

45. Zafra F, Lindholm D, Castren E, Hartikka J, and Thoenen H (1992) Regulation of brain-derived neurotrophic factor and nerve growth factor mRNA in primary cultures of hippocampal neurons and astrocytes. J. Neurosci. 12: 4793-4799.

46. Zigmond RE, Schwarzschild MA, Rittenhouse AR (1989) Acute regulation of tyrosine hydroxylase by nerve activity and by neurotransmitters via phosphorylation. Annu Rev Neurosci 12:415-461.

TOP