Basic and Clinical Pharmacology, 13th Ed.

Introduction to the Pharmacology of CNS Drugs

John A. Gray, MD, PhD, &

Roger A. Nicoll, MD

Drugs acting in the central nervous system (CNS) were among the first to be discovered by primitive humans and are still the most widely used group of pharmacologic agents. These include medications used to treat a wide range of neurologic and psychiatric conditions as well as drugs that relieve pain, suppress nausea, and reduce fever, among other symptoms. In addition, many CNS-acting drugs are used without prescription to increase the sense of well-being.

Due to their complexity, the mechanisms by which various drugs act in the CNS have not always been clearly understood. In recent decades, however, dramatic advances have been made in the methodology of CNS pharmacology. It is now possible to study the action of a drug on individual neurons and even single receptors within synapses. The information obtained from such studies is the basis for several major developments in studies of the CNS. First, it is clear that nearly all drugs with CNS effects act on specific receptors that modulate synaptic transmission. While a few agents such as general anesthetics and alcohol may have nonspecific actions on membranes (although these exceptions are not fully accepted), even these non-receptor-mediated actions result in demonstrable alterations in synaptic transmission.

Second, drugs are among the most valuable tools for studying CNS function, from understanding the mechanism of convulsions to the laying down of long-term memory. Both agonists that mimic natural transmitters (and in many cases that are more selective than the endogenous substances) and antagonists are extremely useful in such studies. Third, unraveling the actions of drugs with known clinical efficacy has led to some of the most fruitful hypotheses regarding the mechanisms of disease. For example, information about the action of antipsychotic drugs on dopamine receptors has provided the basis for important hypotheses regarding the pathophysiology of schizophrenia. Studies of the effects of a variety of agonists and antagonists on γ-aminobutyric acid (GABA) receptors have resulted in new concepts pertaining to the pathophysiology of several diseases, including anxiety and epilepsy.

A full appreciation of the effects of a drug on the CNS requires an understanding of the multiple levels of brain organization, from genes to circuits to behavior. This chapter provides an introduction to the functional organization of the CNS and its synaptic transmitters as a basis for understanding the actions of the drugs described in the following chapters.


The CNS comprises the brain and spinal cord and is responsible for integrating sensory information and generating motor output and other behaviors needed to successfully interact with the environment and enhance species survival. The human brain contains about 100 billion interconnected neurons surrounded by various supporting glial cells. Throughout the CNS, neurons are either clustered into groups called nuclei or are present in layered structures such as the cerebellum or hippocampus. Connections among neurons both within and between these clusters form the circuitry that regulates information flow through the CNS.


Neurons are electrically excitable cells that process and transmit information via an electrochemical process. There are many types of neurons in the CNS, and they are classified in the following ways: by function, by location, and by the neurotransmitter they release. The typical neuron possesses a cell body (or soma) and specialized processes called dendrites and axons (Figure 21–1). Dendrites, which form highly branched complex dendritic “trees,” receive and integrate the input from other neurons and conduct this information to the cell body. The axon carries the output signal of a neuron from the cell body, sometimes over long distances. Neurons may have hundreds of dendrites but generally have only one axon, though axons may branch distally to contact multiple targets. The axon terminal makes contact with other neurons at specialized junctions, called synapses, where neurotransmitter chemicals are released that interact with receptors on other neurons.


FIGURE 21–1 Neurons and glia in the CNS. A typical neuron has a cell body (or soma) that receives the synaptic responses from the dendritic tree. These synaptic responses are integrated at the axon initial segment, which has a high concentration of voltage-gated sodium channels. If an action potential is initiated, it propagates down the axon to the synaptic terminals, which contact other neurons. The axon of long-range projection neurons is insulated by a myelin sheath derived from specialized membrane processes of oligodendrocytes, analogous to the Schwann cells in the peripheral nervous system. Astrocytes perform supportive roles in the CNS, and their processes are closely associated with neuronal synapses. (see Figures 21-4 and 21-7).


In addition to neurons, there are a large number of non-neuronal support cells, called neuroglia or glia, that perform a variety of essential functions in the CNS. Astrocytes are the most abundant glial cells in the brain and play homeostatic support roles, including providing metabolic nutrients to neurons and maintaining extracellular ion concentrations. In addition, astrocyte processes are closely associated with neuronal synapses where they are involved in the removal and recycling of neurotransmitters after release and play increasingly appreciated roles in regulating neurotransmission (see below).

Oligodendrocytes are cells that wrap around the axons of projection neurons in the CNS forming the myelin sheath (Figure 22–1). Similar to the Schwann cells in peripheral neurons, the myelin sheath created by the oligodendrocytes insulates the axons and increases the speed of signal propagation. Damage to oligodendrocytes occurs in multiple sclerosis and thus is a target of drug discovery efforts.

Microglia are specialized macrophages derived from the bone marrow that are found in the CNS and are the major immune defense system in the brain. These cells are actively involved in neuroinflammatory processes in many pathological states including neurodegenerative diseases.

Blood-Brain Barrier

The blood-brain barrier (BBB) is a protective functional separation of the circulating blood from the extracellular fluid of the CNS that limits the penetration of substances, including drugs. This separation is accomplished by the presence of tight junctions between the capillary endothelial cells as well as a surrounding layer of astrocyte end-feet. Therefore, to enter the CNS, drugs must either be highly hydrophobic or engage specific transport mechanisms. For example, the second-generation antihistamines cause less drowsiness because they were developed to be significantly more polar than older antihistamines, limiting their crossing of the BBB (see Chapter 16). Many nutrients, such as glucose and the essential amino acids, have specific transporters that allow them to cross the BBB. L-DOPA, a precursor of the neurotransmitter dopamine, can enter the brain using an amino acid transporter, whereas dopamine cannot cross the BBB. Thus, an orally administered drug, L-DOPA, but not dopamine, can be used to boost CNS dopamine levels in the treatment of Parkinson’s disease. Some parts of the brain, the so-called circumventricular organs, lack a normal BBB. These include regions that sample the blood, such as the area postrema vomiting center, and regions that secrete neurohormones into the circulation.


The membranes of neurons contain two types of channels defined on the basis of the mechanisms controlling their gating (opening and closing): voltage-gated and ligand-gated channels (Figure 21–2A and B). Voltage-gated channels respond to changes in the membrane potential of the cell. The voltage-gated sodium channel described in Chapter 14 for the heart is an example of this type of channel. In nerve cells, these channels are highly concentrated on the initial segment of the axon (Figure 21–1), which initiates the all-or-nothing fast action potential and along the length of the axon where they propagate the action potential to the nerve terminal. There are also many types of voltage-sensitive calcium and potassium channels on the cell body, dendrites, and initial segment, which act on a much slower time scale and modulate the rate at which the neuron discharges. For example, some types of potassium channels opened by depolarization of the cell result in slowing of further depolarization and act as a brake to limit further action potential discharge. Plant and animal toxins that target various voltage-gated ion channels have been invaluable for studying the functions of these channels (see Box: Natural Toxins: Tools for Characterizing Ion Channels; Table 21–1).

TABLE 21–1 Some toxins used to characterize ion channels.



FIGURE 21–2 Types of ion channels and neurotransmitter receptors in the CNS. A shows a voltage-gated channel in which a voltage sensor component of the protein controls the gating (broken arrow) of the channel. B shows a ligand-gated channel in which the binding of the neurotransmitter to the ionotropic channel receptor controls the gating (broken arrow) of the channel. C shows a G protein-coupled (metabotropic) receptor, which, when bound, activates a heterotrimeric G protein. D and E show two ways metabotropic receptors can regulate ion channels. The activated G protein can interact directly to modulate an ion channel (D) or the G protein can activate an enzyme that generates a diffusible second messenger (E), eg, cAMP, which can interact with the ion channel or can activate a kinase that phosphorylates and modulates a channel.

Neurotransmitters exert their effects on neurons by binding to two distinct classes of receptor. The first class is referred to as ligand-gated channels, or ionotropic receptors. These receptors consist of multiple subunits, and binding of the neurotransmitter ligand directly opens the channel, which is an integral part of the receptor complex (see Figure 22–6). These channels are insensitive or only weakly sensitive to membrane potential. Activation of these channels typically results in a brief (a few milliseconds to tens of milliseconds) opening of the channel. Ligand-gated channels are responsible for fast synaptic transmission typical of hierarchical pathways in the CNS (see following text).

Natural Toxins: Tools for Characterizing Ion Channels

Evolution is tireless in the development of natural toxins. A vast number of variations are possible with even a small number of amino acids in peptides, and peptides make up only one of a broad array of toxic compounds. For example, the predatory marine snail genus Conus includes over 3000 different species. Each species kills or paralyzes its prey with a venom that contains 50–200 different peptides or proteins. Furthermore, there is little duplication of peptides among Conus species. Other animals with useful toxins include snakes, frogs, spiders, bees, wasps, and scorpions. Plant species with toxic (or therapeutic) substances are referred to in several other chapters of this book.

Since many toxins act on ion channels, they provide a wealth of chemical tools for studying the function of these channels. In fact, much of our current understanding of the properties of ion channels comes from studies utilizing only a small percentage of the highly potent and selective toxins that are now available. The toxins typically target voltage-sensitive ion channels, but a number of very useful toxins block ionotropic neurotransmitter receptors. Table 21–1 lists some of the toxins most commonly used in research, their mode of action, and their source.

The second class of neurotransmitter receptor is referred to as metabotropic receptors (Figure 21–2C). These are seven-transmembrane G protein-coupled receptors of the type described in Chapter 2. The binding of neurotransmitter to this type of receptor does not result in the direct gating of a channel. Rather, binding to the receptor engages a G protein, which results in the production of second messengers that mediate intracellular signaling cascades such as those described in Chapter 2.

In neurons, activation of metabotropic neurotransmitter receptors often leads to the modulation of voltage-gated channels. These interactions can occur entirely within the plane of the membrane and are referred to as membrane-delimited pathways (Figure 21–2D). In this case, the G protein (often the βγ subunit) interacts directly with a voltage-gated ion channel. In general, two types of voltage-gated ion channels are the targets of this type of signaling: calcium channels and potassium channels. When G proteins interact with calcium channels, they inhibit channel function. This mechanism accounts for the inhibition of neurotransmitter release that occurs when presynaptic metabotropic receptors are activated. In contrast, when these receptors are postsynaptic, they activate (cause the opening of) potassium channels, resulting in a slow postsynaptic inhibition. Metabotropic receptors can also modulate voltage-gated channels less directly by the generation of diffusible second messengers (Figure 21–2E). A classic example of this type of action is provided by the β adrenoceptor, which generates cAMP via the activation of adenylyl cyclase (see Chapter 2). Whereas membrane-delimited actions occur within microdomains in the membrane, second messenger-mediated effects can occur over considerable distances. Finally, an important consequence of the involvement of G proteins in receptor signaling is that, in contrast to the brief effect of ionotropic receptors, the effects of metabotropic receptor activation can last tens of seconds to minutes. Metabotropic receptors predominate in the diffuse neuronal systems in the CNS (see below).


The communication between neurons in the CNS occurs through chemical synapses in the majority of cases. (A few instances of electrical coupling between neurons have been documented, and such coupling may play a role in synchronizing neuronal discharge. However, it is unlikely that these electrical synapses are an important site of drug action.) The events involved in synaptic transmission can be summarized as follows.

An action potential propagating down the axon of the presynaptic neuron enters the synaptic terminal and activates voltage-sensitive calcium channels in the membrane of the terminal (see Figure 6–3). The calcium channels responsible for the release of neurotransmitter are generally resistant to the calcium channel-blocking agents discussed in Chapter 12 (verapamil, etc) but are sensitive to blockade by certain marine toxins and metal ions (see Tables 21–1 and 12–4). As calcium flows into the terminal, the increase in intraterminal calcium concentration promotes the fusion of synaptic vesicles with the presynaptic membrane. The neurotransmitter contained in the vesicles is released into the synaptic cleft and diffuses to the receptors on the postsynaptic membrane. The neurotransmitter binds to its receptor and opens channels (either directly or indirectly as described above) causing a brief change in membrane conductance (permeability to ions) of the postsynaptic cell. The time delay from the arrival of the presynaptic action potential to the onset of the postsynaptic response is approximately 0.5 ms. Most of this delay is consumed by the release process, particularly the time required for calcium channels to open.


FIGURE 21–3 Postsynaptic potentials and action potential generation. A (top) shows the voltage recorded upon entry of a microelectrode into a postsynaptic cell and subsequent recording of a resting membrane potential of −60 mV. Stimulation of an excitatory pathway (E1, left) generates transient depolarization called an excitatory postsynaptic potential (EPSP). Simultaneous activation of multiple excitatory synapses (E1 + E2, middle) increases the size of the depolarization, so that the threshold for action potential generation is reached. Alternatively, a train of stimuli from a single input can temporally summate to reach the threshold (E1 + E1, right). B (bottom) demonstrates the interaction of excitatory and inhibitory synapses. On the left, a suprathreshold excitatory stimulus (E3) evokes an action potential. In the center, an inhibitory pathway (I) generates a small hyperpolarizing current called an inhibitory postsynaptic potential (IPSP). On the right, if the previously suprathreshold excitatory input (E3) is given shortly after the inhibitory input (I), the IPSP prevents the excitatory potential from reaching threshold.

The first systematic analysis of synaptic potentials in the CNS was in the early 1950s by Eccles and associates, who recorded intracellularly from spinal motor neurons. When a microelectrode enters a cell, there is a sudden change in the potential recorded by the electrode, which is typically about −60 mV (Figure 21-3). This is the resting membrane potential of the neuron. Two types of pathways—excitatory and inhibitory—impinge on the motor neuron.

When an excitatory pathway is stimulated, a small depolarization or excitatory postsynaptic potential (EPSP) is recorded. This potential is due to the excitatory transmitter acting on an ionotropic receptor, causing an increase in cation permeability. As additional excitatory synapses are activated, there is a graded summation of the EPSPs to increase the size of the depolarization (Figure 21–3, top, spatial summation, middle). When a sufficient number of excitatory synapses are activated, the excitatory postsynaptic potential depolarizes the postsynaptic cell to threshold, and an all-or-none action potential is generated. Alternatively, if there is a repetitive firing of an excitatory input, the temporal summation of the EPSPs may also reach the action potential threshold (Figure 21–3, top, right).

When an inhibitory pathway is stimulated, the postsynaptic membrane is hyperpolarized owing to the selective opening of chloride channels, producing an inhibitory postsynaptic potential (IPSP) (Figure 21–3, bottom, middle). However, because the equilibrium potential for chloride (see Chapter 14) is only slightly more negative than the resting potential (~ −65 mV), the hyperpolarization is small and contributes only modestly to the inhibitory action. The opening of the chloride channel during the inhibitory postsynaptic potential makes the neuron “leaky” so that changes in membrane potential are more difficult to achieve. This shunting effect decreases the change in membrane potential during the excitatory postsynaptic potential. As a result, an excitatory postsynaptic potential that evoked an action potential under resting conditions fails to evoke an action potential during the inhibitory postsynaptic potential (Figure 21–3, bottom, right). A second type of inhibition is presynaptic inhibition. It was first described for sensory fibers entering the spinal cord, where excitatory synaptic terminals receive synapses called axoaxonic synapses (described later). When activated, axoaxonic synapses reduce the amount of transmitter released from the terminals of sensory fibers. It is interesting that presynaptic inhibitory receptors are present on almost all presynaptic terminals in the brain even though axoaxonic synapses appear to be restricted to the spinal cord. In the brain, transmitter can spill out of the synapse and activate presynaptic receptors, either on the same synapse (autoreceptors) or on neighboring synapses.


Virtually all the drugs that act in the CNS produce their effects by modifying some step in chemical synaptic transmission. Figure 21–4 illustrates some of the steps that can be altered. These transmitter-dependent actions can be divided into presynaptic and postsynaptic categories.


FIGURE 21–4 Sites of drug action. Schematic drawing of steps at which drugs can alter synaptic transmission. (1) Action potential in presynaptic fiber; (2) synthesis of transmitter; (3) storage; (4) metabolism; (5) release; (6) reuptake into the nerve ending or uptake into a glial cell; (7) degradation; (8) receptor for the transmitter; (9) receptor-induced increase or decrease in ionic conductance; (10) retrograde signaling.

Drugs acting on the synthesis, storage, metabolism, and release of neurotransmitters fall into the presynaptic category. Synaptic transmission can be depressed by blockade of transmitter synthesis or storage. For example, reserpine depletes monoamine synapses of transmitters by interfering with intracellular storage. Blockade of transmitter catabolism inside the nerve terminal can increase transmitter concentrations and has been reported to increase the amount of transmitter released per impulse. Drugs can also alter the release of transmitters. The stimulant amphetamine induces the release of catecholamines from adrenergic synapses (see Chapters 69, and 32). Capsaicin causes the release of the peptide substance P from sensory neurons, and tetanus toxin blocks the release of transmitters. After a CNS transmitter has been released into the synaptic cleft, its action is terminated either by uptake or by degradation. For most neurotransmitters, there are uptake mechanisms into the synaptic terminal and also into surrounding neuroglia. Cocaine, for example, blocks the uptake of catecholamines at adrenergic synapses and thus potentiates the action of these amines. Acetylcholine, however, is inactivated by enzymatic degradation, not reuptake. Anticholinesterases block the degradation of acetylcholine and thereby prolong its action (see Chapter 7). No uptake mechanism has been found for any of the numerous CNS peptides, and it has yet to be demonstrated whether specific enzymatic degradation terminates the action of peptide transmitters.

In the postsynaptic region, the transmitter receptor provides the primary site of drug action. Drugs can act either as neurotransmitter agonists, such as the opioids, which mimic the action of enkephalin, or they can block receptor function. Receptor antagonism is a common mechanism of action for CNS drugs. An example is strychnine’s blockade of the receptor for the inhibitory transmitter glycine. This block, which underlies strychnine’s convulsant action, illustrates how the blockade of inhibitory processes results in excitation. Drugs can also act directly on the ion channel of ionotropic receptors. For example, the anesthetic ketamine blocks the NMDA subtype of glutamate ionotropic receptors by binding in the ion channel pore. In the case of metabotropic receptors, drugs can act at any of the steps downstream of the receptor. Perhaps the best example is provided by the methylxanthines, which can modify neurotransmitter responses mediated through the second-messenger cAMP. At high concentrations, the methylxanthines elevate the level of cAMP by blocking its metabolism and thereby prolong its action.

The traditional view of the synapse is that it functions like a valve, transmitting information in one direction. However, it is now clear that the synapse can also generate signals that feed back onto the presynaptic terminal to modify transmitter release. Endocannabinoids are the best documented example of such retrograde signaling. Postsynaptic activity leads to the synthesis and release of endocannabinoids, which then bind to receptors on the presynaptic terminal. Although the gas nitric oxide (NO) has long been proposed as a retrograde messenger, its physiologic role in the CNS is still not well understood.

The selectivity of CNS drug action is based on two primary factors. First, with a few exceptions, different neurotransmitters are released by different groups of neurons. These transmitters are often segregated into neuronal systems that subserve broadly different CNS functions. That this segregation occurs has provided neuroscientists with a powerful pharmacologic approach for analyzing CNS function and treating pathologic conditions. Second, there is a multiplicity of receptors for each neurotransmitter. For example, at least 14 different serotonin receptors are encoded by different genes. These receptors also often have differential cellular distributions throughout the CNS, allowing for the development of drugs that selectively target particular receptors and CNS functions.


Most of the neuronal systems in the CNS can be divided into two broad categories: hierarchical systems and nonspecific or diffuse neuronal systems.

Hierarchical Systems

Hierarchical systems include all the pathways directly involved in sensory perception and motor control. These pathways are generally clearly delineated, being composed of large myelinated fibers that can often conduct action potentials at a rate of more than 50 m/s. The information is typically phasic and occurs in bursts of action potentials. In sensory systems, the information is processed sequentially by successive integrations at each relay nucleus on its way to the cortex. A lesion at any link incapacitates the system.

Within each nucleus and in the cortex, there are two types of cells: relay or projection neurons and local circuit neurons (Figure 21–5A). The projection neurons form the interconnecting pathways that transmit signals over long distances. Their cell bodies are relatively large, and their axons can project long distances but also emit small collaterals that synapse onto local interneurons. These neurons are excitatory, and their synaptic influences, which involve ionotropic receptors, are very short-lived. The excitatory transmitter released from these cells is, in most instances, glutamate.


FIGURE 21–5 Hierarchical pathways in the CNS. A shows parts of three excitatory relay neurons (blue) and two types of local inhibitory interneuron pathways, recurrent and feed-forward. The inhibitory neurons are shown in gray. B shows the pathway responsible for axoaxonic presynaptic inhibition in which the axon of an inhibitory neuron (gray) synapses onto the presynaptic axon terminal of an excitatory fiber (blue) to inhibit its neurotransmitter release.

Local circuit neurons are typically smaller than projection neurons, and their axons arborize in the immediate vicinity of the cell body. Most of these neurons are inhibitory, and they release either GABA or glycine. They synapse primarily on the cell body of the projection neurons but can also synapse on the dendrites of projection neurons as well as with each other. Two common types of pathways for these neurons (Figure 21–5A) include recurrent feedback pathways and feed-forward pathways. A special class of local circuit neurons in the spinal cord forms axoaxonic synapses on the terminals of sensory axons (Figure 21–5B). Although there are a great variety of synaptic connections in these hierarchical systems, the fact that a limited number of transmitters are used by these neurons indicates that any major pharmacologic manipulation of this system will have a profound effect on the overall excitability of the CNS. For instance, selectively blocking GABAA receptors with a drug such as picrotoxin results in generalized convulsions. Thus, although the mechanism of action of picrotoxin is specific in blocking the effects of GABA, the overall functional effect appears to be quite nonspecific, because GABA-mediated synaptic inhibition is so widely utilized in the brain.

Nonspecific or Diffuse Neuronal Systems

Neuronal systems containing many of the other neurotransmitters, including the monoamines and acetylcholine, differ in fundamental ways from the hierarchical systems. These neurotransmitters are produced by only a limited number of neurons whose cell bodies are located in small discrete nuclei, often in the brainstem. For example, noradrenergic cell bodies are found primarily in a compact cell group, called the locus caeruleus, located in the caudal pontine central gray matter and number only approximately 1500 neurons on each side of the brain in the rat. However, from these limited nuclei, these neurons project widely and diffusely throughout the brain and spinal cord (Figure 21–6). Because the axons from these diffusely projecting neurons are fine and unmyelinated, they conduct very slowly, at about 0.5 m/s. The axons branch repeatedly and are extraordinarily divergent. Branches from the same neuron can innervate several functionally different parts of the CNS, synapsing onto and modulating neurons within the hierarchical systems. In the neocortex, these fibers have a tangential organization and therefore can influence large areas of cortex. In addition, most neurotransmitters utilized by diffuse neuronal systems, including norepinephrine, act predominantly on metabotropic receptors and therefore initiate long-lasting synaptic effects. Based on these observations, it is clear that the monoamine systems cannot be conveying topographically specific types of information; rather, vast areas of the CNS must be affected simultaneously and in a rather uniform way. It is not surprising, then, that these systems have been implicated in such global functions as sleeping and waking, attention, appetite, and emotional states.


FIGURE 21–6 Diffuse neurotransmitter pathways in the CNS. For each of the neurotransmitter pathways shown, the cell bodies are located in discrete brainstem or basal forebrain nuclei and project widely throughout the CNS. These diffuse systems largely modulate the function of the hierarchical pathways. Serotonin neurons, for example, are found in the midline raphe nuclei in the forebrain and send extraordinarily divergent projections to nearly all regions of the CNS. Other diffusely projecting neurotransmitter pathways include the histamine and orexin systems (not shown). VTA, ventral tegmental area; SN, substantia nigra; A1-A7, adrenergic brainstem nuclei; MSN, medial septal nucleus; DB, diagonal band of Broca; C5-C8, cholinergic brainstem nuclei.


Because drug selectivity is based on the fact that different pathways use different transmitters, a primary goal of neuroscientists has been to identify the neurotransmitters in CNS pathways. Establishing that a chemical substance is a transmitter has been far more difficult for central synapses than for peripheral synapses. The following criteria have been established for transmitter identification.

1.Localization: A suspected transmitter must reside in the presynaptic terminal of the pathway of interest.

2.Release: A suspected transmitter must be released from a neuron in response to neuronal activity and in a calcium-dependent manner.

3.Synaptic mimicry: Application of the candidate substance should produce a response that mimics the action of the transmitter released by nerve stimulation, and application of a selective antagonist should block the response.

Using the criteria above, a vast number of small molecules have been isolated from the brain, and studies using a variety of approaches suggest that the agents listed in Table 21–2 are neurotransmitters. A brief summary of these compounds follows.

TABLE 21–2 Summary of neurotransmitter pharmacology in the central nervous system.



Amino Acid Neurotransmitters

The amino acids of primary interest to the pharmacologist fall into two categories: the acidic amino acid glutamate and the neutral amino acids glycine and GABA. All three compounds are present in high concentrations in the CNS and are extremely potent modifiers of neuronal excitability.

A. Glutamate

Excitatory synaptic transmission is mediated by glutamate, which is present in very high concentrations in excitatory synaptic vesicles (~100 mM). Glutamate is released into the synaptic cleft by Ca2+-dependent exocytosis. The released glutamate acts on postsynaptic glutamate receptors and is cleared by glutamate transporters present on surrounding glia (Figure 21–7). In glia, glutamate is converted to glutamine by glutamine synthetase, released from the glia, taken up by the nerve terminal, and converted back to glutamate by the enzyme glutaminase. The high concentration of glutamate in synaptic vesicles is achieved by the vesicular glutamate transporter (VGLUT).


FIGURE 21–7 Schematic diagram of a glutamate synapse. Glutamine is imported into the glutamatergic neuron (A) and converted into glutamate by glutaminase. The glutamate is then concentrated in vesicles by the vesicular glutamate transporter. Upon release into the synapse, glutamate can interact with AMPA and NMDA ionotropic receptor channels (AMPAR, NMDAR) and with metabotropic receptors (mGluR) on the postsynaptic cell (B). Synaptic transmission is terminated by active transport of the glutamate into a neighboring glial cell (C) by a glutamate transporter. It is converted into glutamine by glutamine synthetase and transported back into the glutamatergic axon terminal.

Almost all neurons that have been tested are strongly excited by glutamate. This excitation is caused by the activation of both ionotropic and metabotropic receptors, which have been extensively characterized by molecular cloning. The ionotropic receptors are divided into three subtypes based on the action of selective agonists: α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA), kainic acid (KA),and N-methyl-D-aspartate (NMDA). All the ionotropic receptors are composed of four subunits. AMPA receptors, which are present on all neurons, are heterotetramers assembled from four subunits (GluA1–GluA4). The majority of AMPA receptors contain the GluA2 subunit and are permeable to Na+ and K+, but not to Ca2+. Some AMPA receptors, typically present on inhibitory interneurons, lack the GluA2 subunit and are also permeable to Ca2+.

Kainate receptors are not as uniformly distributed as AMPA receptors, being expressed at high levels in the hippocampus, cerebellum, and spinal cord. They are formed from a number of subunit combinations (GluK1–GluK5). Although GluK4 and GluK5 are unable to form channels on their own, their presence in the receptor changes the receptor’s affinity and kinetics. Similar to AMPA receptors, kainate receptors are permeable to Na+ and K+ and in some subunit combinations can also be permeable to Ca2+.

NMDA receptors are as ubiquitous as AMPA receptors, being present on essentially all neurons in the CNS. All NMDA receptors require the presence of the subunit GluN1. The channel also contains one or two GluN2 subunits (GluN2A–GluN2D). Unlike AMPA and kainate receptors, all NMDA receptors are highly permeable to Ca2+ as well as to Na+ and K+. NMDA receptor function is controlled in a number of intriguing ways. In addition to glutamate binding, the channel also requires the binding of glycine to a separate site. The physiologic role of glycine binding is unclear because the glycine site appears to be saturated at normal ambient levels of glycine. Another important feature is that while AMPA and kainate receptor activation results in channel opening at resting membrane potential, NMDA receptor activation does not. This is due to the voltage-dependent block of the NMDA pore by extracellular Mg2+. Only when the neuron is strongly depolarized, as occurs with intense activation of the synapse or by activation of neighboring synapses, is Mg2+ expelled and the channel opened. Thus, there are two requirements for NMDA receptor channel opening: Glutamate must bind the receptor and the membrane must be depolarized. The rise in intracellular Ca2+ that accompanies channel opening results in a long-lasting enhancement in synaptic strength that is referred to as long-term potentiation (LTP). The change, which is one major type of synaptic plasticity, can last for many hours or even days and is generally accepted as an important cellular mechanism underlying learning and memory.

The metabotropic glutamate receptors are G protein-coupled receptors that act indirectly on ion channels via G proteins. Metabotropic receptors (mGluR1–mGluR8) have been divided into three groups (I, II, and III). A variety of agonists and antagonists have been developed that interact selectively with the different groups. Group I receptors are typically located postsynaptically and activate phospholipase C, leading to inositol trisphosphate-mediated intracellular Ca2+ release. In contrast, group II and group III receptors are typically located on presynaptic nerve terminals and act as inhibitory autoreceptors. Activation of these receptors causes the inhibition of Ca2+ channels, resulting in inhibition of transmitter release. These receptors are activated only when the concentration of glutamate rises to high levels during repetitive stimulation of the synapse. Activation of these receptors also causes the inhibition of adenylyl cyclase and decreases cAMP generation.

B. GABA and Glycine

Both GABA and glycine are inhibitory neurotransmitters, which are typically released from local interneurons. Interneurons that release glycine are restricted to the spinal cord and brainstem, whereas interneurons releasing GABA are present throughout the CNS, including the spinal cord. It is interesting that some interneurons in the spinal cord can release both GABA and glycine. Glycine receptors are pentameric structures that are selectively permeable to Cl. Strychnine, which is a potent spinal cord convulsant and has been used in some rat poisons, selectively blocks glycine receptors.

GABA receptors are divided into two main types: GABAA and GABAB. Inhibitory postsynaptic potentials in many areas of the brain have a fast and slow component. The fast component is mediated by GABAA receptors and the slow component by GABAB receptors. The difference in kinetics stems from the differences in coupling of the receptors to ion channels. GABAA receptors are ionotropic receptors and, like glycine receptors, are pentameric structures that are selectively permeable to Cl. These receptors are selectively inhibited by picrotoxin and bicuculline, both of which cause generalized convulsions. A great many subunits for GABAA receptors have been cloned; this accounts for the large diversity in the pharmacology of GABAA receptors, making them key targets for clinically useful agents (see Chapter 22). GABAB receptors are metabotropic receptors that are selectively activated by the antispastic drug baclofen. These receptors are coupled to G proteins that, depending on their cellular location, either inhibit Ca2+ channels or activate K+ channels. The GABAB component of the inhibitory postsynaptic potential is due to a selective increase in K+ conductance. This inhibitory postsynaptic potential is long-lasting and slow because the coupling of receptor activation to K+ channel opening is indirect and delayed. GABAB receptors are localized to the perisynaptic region and thus require the spillover of GABA from the synaptic cleft. GABAB receptors are also present on the axon terminals of many excitatory and inhibitory synapses. In this case, GABA spills over onto these presynaptic GABAB receptors, inhibiting transmitter release by inhibiting Ca2+ channels. In addition to their coupling to ion channels, GABAB receptors also inhibit adenylyl cyclase and decrease cAMP generation.


Acetylcholine was the first compound to be identified pharmacologically as a transmitter in the CNS. Eccles showed in the early 1950s that excitation of spinal cord Renshaw cells by recurrent axon collaterals from spinal motor neurons was blocked by nicotinic antagonists. Furthermore, Renshaw cells were extremely sensitive to nicotinic agonists. This early success at identifying a transmitter for a central synapse was followed by disappointment because it remained the sole central synapse for which the transmitter was known until the late 1960s, when comparable data became available for GABA and glycine. The motor axon collateral synapse remains one of the best-documented examples of a cholinergic nicotinic synapse in the mammalian CNS, despite the rather widespread distribution of nicotinic receptors as defined by in situ hybridization studies.

Most CNS responses to acetylcholine are mediated by a large family of G protein-coupled muscarinic receptors. At a few sites, acetylcholine causes slow inhibition of the neuron by activating the M2 subtype of receptor, which opens potassium channels. A far more widespread muscarinic action in response to acetylcholine is a slow excitation that in some cases is mediated by M1 receptors. These muscarinic effects are much slower than either nicotinic effects on Renshaw cells or the effect of amino acids. Furthermore, this M1 muscarinic excitation is unusual in that acetylcholine produces it by decreasing the membrane permeability to potassium, ie, the opposite of conventional transmitter action.

Eight major CNS nuclei of acetylcholine neurons have been characterized with diffuse projections. These include neurons in the neostriatum, the medial septal nucleus, and the reticular formation that appear to play an important role in cognitive functions, especially memory. Presenile dementia of the Alzheimer type is reportedly associated with a profound loss of cholinergic neurons. However, the specificity of this loss has been questioned because the levels of other putative transmitters, eg, somatostatin, are also decreased.

Monoamine Neurotransmitters

Monoamines include the catecholamines (dopamine and norepinephrine) and 5-hydroxytryptamine. The diamine neurotransmitter, histamine, has several similarities to these monoamines. Although these compounds are present in very small amounts in the CNS, they can be localized using extremely sensitive histochemical methods. These pathways are the site of action of many drugs; for example, the CNS stimulants cocaine and amphetamine appear to act primarily at catecholamine synapses. Cocaine blocks the reuptake of dopamine and norepinephrine, whereas amphetamines cause presynaptic terminals to release these transmitters.

A. Dopamine

The major pathways containing dopamine are the projection linking the substantia nigra to the neostriatum and the projection linking the ventral tegmental region to limbic structures, particularly the limbic cortex. The therapeutic action of the antiparkinsonism drug levodopa is associated with the former area (see Chapter 28), whereas the therapeutic action of the antipsychotic drugs is thought to be associated with the latter (see Chapter 29). In addition, dopamine-containing neurons in the ventral hypothalamus play an important role in regulating pituitary function. Five dopamine receptors have been identified, and they fall into two categories: D1-like (D1 and D5) and D2-like (D2, D3, D4). All dopamine receptors are metabotropic. Dopamine generally exerts a slow inhibitory action on CNS neurons. This action has been best characterized on dopamine-containing substantia nigra neurons, where D2-receptor activation opens potassium channels via the Gi coupling protein.

B. Norepinephrine

Most noradrenergic neurons are located in the locus caeruleus or the lateral tegmental area of the reticular formation. Although the density of fibers innervating various sites differs considerably, most regions of the CNS receive diffuse noradrenergic input. All noradrenergic receptor subtypes are metabotropic. When applied to neurons, norepinephrine can hyperpolarize them by increasing potassium conductance. This effect is mediated by α2 receptors and has been characterized most thoroughly on locus caeruleus neurons. In many regions of the CNS, norepinephrine actually enhances excitatory inputs by both indirect and direct mechanisms. The indirect mechanism involves disinhibition; that is, inhibitory local circuit neurons are inhibited. The direct mechanism involves blockade of potassium conductances that slow neuronal discharge. Depending on the type of neuron, this effect is mediated by either α1 or β receptors. Facilitation of excitatory synaptic transmission is in accordance with many of the behavioral processes thought to involve noradrenergic pathways, eg, attention and arousal.

C. 5-Hydroxytryptamine

Most 5-hydroxytryptamine (5-HT, serotonin) pathways originate from neurons in the midline raphe nuclei of the pons and upper brainstem. 5-HT is contained in unmyelinated fibers that diffusely innervate most regions of the CNS, but the density of the innervation varies. 5-HT acts on more than a dozen receptor subtypes. Except for the 5-HT3 receptor, all of these receptors are metabotropic. The ionotropic 5-HT3 receptor exerts a rapid excitatory action at a very limited number of sites in the CNS. In most areas of the CNS, 5-HT has a strong inhibitory action. This action is mediated by 5-HT1A receptors and is associated with membrane hyperpolarization caused by an increase in potassium conductance. It has been found that 5-HT1A receptors and GABAB receptors activate the same population of potassium channels. Some cell types are slowly excited by 5-HT owing to its blockade of potassium channels via 5-HT2 or 5-HT4 receptors. Both excitatory and inhibitory actions can occur on the same neuron. 5-HT has been implicated in the regulation of virtually all brain functions, including perception, mood, anxiety, pain, sleep, appetite, temperature, neuroendocrine control, and aggression. Given the broad roles of 5-HT in CNS function and the rich molecular diversity of 5-HT receptors, it is not surprising that many therapeutic agents target the 5-HT system (see Chapters 162930, and 32).

D. Histamine

In the CNS, histamine is exclusively made by neurons in the tuberomammillary nucleus (TMN) in the posterior hypothalamus. These neurons project widely throughout the brain and spinal cord where they modulate arousal, attention, feeding behavior, and memory (see Chapter 16). There are four histamine receptors (H1 to H4), all of which are metabotropic. Centrally acting antihistamines are generally used for their sedative properties and antagonism of H1 receptors is a common side effect of many drugs including some tricyclic antidepressants and antipsychotics.


A great many CNS peptides have been discovered that produce dramatic effects both on animal behavior and on the activity of individual neurons. In many cases, peptide hormones discovered in the periphery (see Chapter 17) also act as neurotransmitters in the CNS. As most of these peptides were initially named according to their peripheral functions, the names are often unrelated to their CNS function. The pathways for many of the peptides have been mapped with immunohistochemical techniques and include opioid peptides (eg, enkephalins, endorphins), neurotensin, substance P, somatostatin, cholecystokinin, vasoactive intestinal polypeptide, neuropeptide Y, and thyrotropin-releasing hormone.

Unlike the classical neurotransmitters above, which are packaged in small synaptic vesicles, neuropeptides are generally packaged in large, dense core vesicles. As in the peripheral autonomic nervous system, peptides often coexist with a conventional nonpeptide transmitter in the same neuron, but the release of the neuropeptides and the small molecule neurotransmitters can be independently regulated. Released neuropeptides may act locally or may diffuse long distances and bind to distant receptors. Most neuropeptide receptors are metabotropic and, like monoamine receptors, primarily serve modulatory roles in the nervous system. Neuropeptides have been implicated in a wide range of CNS functions including reproduction, social behaviors, appetite, arousal, pain, reward, and learning and memory. Thus, neuropeptides and their receptors are active targets of drug discovery efforts.

A good example of the approaches used to define the role of these peptides in the CNS comes from studies on substance P and its association with sensory fibers. Substance P is contained in and released from small unmyelinated primary sensory neurons in the spinal cord and brainstem and causes a slow excitatory postsynaptic potential in target neurons. These sensory fibers are known to transmit noxious stimuli, and it is therefore surprising that—although substance P receptor antagonists can modify responses to certain types of pain—they do not block the response. Glutamate, which is released with substance P from these synapses, presumably plays an important role in transmitting pain stimuli. Substance P is certainly involved in many other functions because it is found in many areas of the CNS that are unrelated to pain pathways.


Orexins are peptide neurotransmitters produced in neurons in the lateral and posterior hypothalamus that, like the monoamine systems, project widely throughout the CNS. Orexins are also called hypocretins because of the near simultaneous discovery by two independent laboratories. Like most neuropeptides, orexin is released from large, dense core vesicles and bind to two G protein-coupled receptors, OX1 and OX2. Orexin neurons also release glutamate and are thus excitatory. The orexin system, like the monoamine systems, projects widely throughout the CNS to influence physiology and behavior. In particular, orexin neurons exhibit firing patterns associated with wakefulness and project to and activate monoamine and acetylcholine neurons involved in sleep-wake cycles (see also Chapter 22). Animals lacking orexin or its receptors have narcolepsy and disrupted sleep-wake patterns. In addition to promoting wakefulness, the orexin system is involved in energy homeostasis, feeding behaviors, autonomic function, and reward.

Other Signaling Substances

A. Endocannabinoids

The primary psychoactive ingredient in cannabis, δ9-tetrahy-drocannabinol (δ9-THC), affects the brain mainly by activating a specific cannabinoid receptor, CB1. CB1 receptors are expressed at high levels in many brain regions, and they are primarily located on presynaptic terminals. Several endogenous brain lipids, including anandamide and 2-arachidonylglycerol (2-AG), have been identified as CB1 ligands. These ligands are not stored, as are classic neurotransmitters, but instead are rapidly synthesized by neurons in response to depolarization and consequent calcium influx. Activation of metabotropic receptors (eg, by acetylcholine and glutamate) can also activate the formation of 2-AG. In further contradistinction to classic neurotransmitters, endogenous cannabinoids can function as retrograde synaptic messengers: They are released from postsynaptic neurons and travel backward across synapses, activating CB1 receptors on presynaptic neurons and suppressing transmitter release. This suppression can be transient or long lasting, depending on the pattern of activity. Cannabinoids may affect memory, cognition, and pain perception by this mechanism.

B. Nitric Oxide

The CNS contains a substantial amount of nitric oxide synthase (NOS) within certain classes of neurons. This neuronal NOS is an enzyme activated by calcium-calmodulin, and activation of NMDA receptors, which increases intracellular calcium, results in the generation of nitric oxide. Although a physiologic role for nitric oxide has been clearly established for vascular smooth muscle, its role in synaptic transmission and synaptic plasticity remains controversial. Nitric oxide diffuses freely across membranes and thus has been hypothesized to be a retrograde messenger, although this has not been demonstrated conclusively. Perhaps the strongest case for a role of nitric oxide in neuronal signaling in the CNS is for long-term depression of synaptic transmission in the cerebellum.

C. Purines

Receptors for purines, particularly adenosine, ATP, UTP, and UDP, are found throughout the body, including the CNS. High concentrations of ATP are found in and released from catecholinergic synaptic vesicles, and ATP may be converted to adenosine extracellularly by nucleotidases. Adenosine in the CNS acts on metabotropic A1 receptors. Presynaptic A1 receptors inhibit calcium channels and inhibit release of both amino acid and monoamine transmitters. ATP co-released with other neurotransmitters can bind to two classes of receptors. The P2X family of ATP receptors comprises nonselective ligand-gated cation channels, whereas the P2Y family is metabotropic. The physiological roles for co-released ATP remain elusive, but pharmacological studies suggest these receptors are involved in memory, wakefulness, and appetite, and may play roles in multiple neuropsychiatric disorders.


Basbaum AI et al: Cellular and molecular mechanisms of pain. Cell 2009;139:267.

Berger M, Gray JA, Roth BL: The expanded biology of serotonin. Annu Rev Med 2009;60:355.

Castillo PE et al: Endocannabinoid signaling and synaptic function. Neuron 2012;76:70.

Catterall WA: Voltage-gated calcium channels. Cold Spring Harb Perspect Biol 2011;3:a003947.

Catterall WA: Voltage-gated sodium channels at 60: Structure, function and pathophysiology. J Physiol 2012;590:2577.

Daneman R: The blood-brain barrier in health and disease. Ann Neurol 2012;72:648.

Gotter AL et al: International Union of Basic and Clinical Pharmacology. LXXXVI. Orexin receptor function, nomenclature and pharmacology. Pharmacol Rev 2012;64:389.

Hille B: Ionic Channels of Excitable Membranes, 3rd ed. Sinauer, 2001.

Jan LY, Jan YN: Voltage-gated potassium channels and the diversity of electrical signalling. J Physiol 2012;590:2591.

Lewis RJ et al: Conus venom peptide pharmacology. Pharmacol Rev 2012;64:259.

Khakh BS, North RA: Neuromodulation by extracellular ATP and P2X receptors in the CNS. Neuron 2012;76:51.

Mody I, Pearce RA: Diversity of inhibitory neurotransmission through GABA(A) receptors. Trends Neurosci 2004;27:569.

Nestler EJ, Hyman SE, Malenka RC: Molecular Neuropharmacology: A Foundation for Clinical Neurosceince, 2nd ed. McGraw-Hill, 2009.

Nicoll RA, Roche KW: Long-term potentiation: peeling the onion. Neuropharmacology 2013;74:18.

Südhof TC, Rizo J: Synaptic vesicle exocytosis. Cold Spring Harb Perspect Biol 2011;3:a005637.

Traynelis SF et al: Glutamate receptor ion channels: structure, regulation, and function. Pharmacol Rev 2010;62:405.