from combinations of two or three subunit types (usually an alpha, beta, and gamma) account for much of the activity of the native receptor (Pritchett et al. 1989).

Researchers have begun defining the GABAA subunits needed for alcohol sensitivity. This work began with the observation that alcohol potentiated GABAA receptor function in brain tissue from mice bred for high alcohol sensitivity but had no effect in tissue from low-sensitivity mice (Allan and Harris 1986). It was then observed that receptors made by using mRNA from brains of the two strains of mice differed in their alcohol sensitivity (Wafford et al. 1990). Only receptors from high-sensitivity animals showed alcohol sensitivity. The actions of barbiturates, benzodiazepines, and receptor antagonists did not differ between strains, suggesting different mechanisms of action for alcohol and benzodiazepine. Next, it was found that one of the gamma subunits is necessary for alcohol sensitivity. This gamma subunit is one of a pair of subunits made by the alternative RNA splicing phenomenon described previously (Whiting

et al. 1990). The major difference between this subunit and the other gamma subunits is the presence of a site for the addition of phosphate. Thus, alcohol sensitivity of the GABAA receptor may depend on both the type of subunit made and its posttranslational modifications thereafter. Both the alcohol-sensitive and the alcoholinsensitive mice possess the GABA receptor subunit necessary for alcohol sensitivity. However, the subunits in the two strains need to be characterized with regard to their phosphorylation states. Thus, although the exact sites of molecular differences between the sensitive and insensitive mice have yet to be determined, this exciting line of research demonstrates how selective genetic breeding can be integrated with molecular biological and pharmacological techniques to provide information on the neural actions of alcohol.

The studies mentioned above did not, however, provide information about the location of alcohol-sensitive GABA receptors within the brain. In vivo studies suggest the importance of alcohol's effects on GABAA receptors in a subcor

tical area known as the medial septal nucleus (Givens and Breese 1990). Administration of alcohol to rats potentiates GABA's ability to inhibit neuronal activity in this nucleus. Blockade of GABAA receptors in this region reduces acute effects of alcohol (Givens and Breese 1990). The medial septal nucleus is thought to be involved in the "arousal state" of the brain via its interactions with structures in the limbic system, i.e., the regions of the brain that are associated with emotions and feelings. It will be interesting to determine whether actions of alcohol in this structure influence other brain regions to produce sedation.

A brain region believed to be important for alcohol's impairing effect on movement is the cerebellum. The large Purkinje neurons in the cerebellar cortex (the outer layer of the cerebellum) are involved in mediating sensory and motor activity and may be the main targets of alcohol's action. These neurons receive input from a number of smaller neurons within the cerebellar cortex and from neurons in nuclei deep within the cerebellum. The Purkinje neurons send integrated information out of the brain to neurons elsewhere in the body to refine movement. It has long been known that Purkinje neurons are especially sensitive to the actions of alcohol (Chu 1983; Rogers et al. 1979). Alcohol may act directly on Purkinje neurons or it may act on other cells that in turn influence Purkinje neuron activity.

Purkinje cells have been studied as single cells and as explants in culture. Short-term exposure of these cells or tissue containing these cells to concentrations of alcohol alters Purkinje cell electrical activity (Franklin and Gruol 1987; Palmer et al. 1988). Alcohol also reduces the ability of the neurotransmitter glutamate to stimulate Purkinje cell firing. Most important for this discussion, alcohol potentiates responses to GABA; i.e., it furthers GABA's inhibitory effect on Purkinje cell activity (Lin et al. 1991). However, this potentiation is seen only when receptors for the neurotransmitter norepinephrine are also activated. This fact suggests a complex interaction between transmitters that allows for alcohol sensitivity. In addition, RO 15-4513 reverses alcohol's effect on Purkinje cell firing (Palmer et al. 1988). The interaction between alcohol and GABA at the Purkinje cell is of interest in light of observations that RO 15-4513 can overcome alcohol's motor-impairing effects (Lister and Nutt 1988). The findings to date suggest that alcohol directly affects Purkinje cells as well as excita

tory and inhibitory neurotransmission onto these

neurons.

There are several clinical implications of the effects of alcohol on "GABAergic" (GABAstimulating) transmission. For example, the anxiolytic action of alcohol may be an underlying factor that contributes to the use of alcohol. It may be possible to design drugs that compete with alcohol for GABA receptors and reduce alcohol's anxiolytic effects. Such drugs might also reduce the anesthetic effects of alcohol and allow individuals to remain conscious after ingesting large amounts of alcohol, thus reducing the incidence of problems incurred during alcoholic loss of consciousness.

Alcohol's effects on glutamatergic
neurotransmission

Glutamate is the major excitatory transmitter in the mammalian central nervous system. The excitatory actions of glutamate are produced by activation of at least three receptor types (Mayer and Westbrook 1987; Sladeczek et al. 1985). Two of these receptor types belong to the family of ligand-gated ion channels, which also includes the GABAA receptor. These receptors can be distinguished by their differential activation by the compounds N-methyl-D-aspartate acid (NMDA) and α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA); the receptors are named for these compounds. Different receptors can also be distinguished on the basis of blockade of their function by selective antagonists (see Watkins and Olverman 1987 for a review).

A brain region believed to be important for alcohol's impairing effect on movement is the cerebellum.

The NMDA and AMPA receptors excite cells because their associated ion channels allow positively charged ions to enter the neuron. The AMPA receptor allows sodium and potassium ions to pass; the NMDA receptor allows calcium, in addition to sodium and potassium, to pass into the cell. Thus, in addition to exciting the neuron, the NMDA receptor also activates calcium-sensitive proteins within the cell. Activation of calcium-dependent enzymes, in turn, may produce changes in neuronal function that outlast the effects of depolarization. Indeed, NMDA receptor activation appears to be important for the

initiation of long-lasting changes in synaptic function (see Collingridge and Bliss 1987 for a review). NMDA receptor activation is crucial for initiation of long-term potentiation (LTP) (Collingridge et al. 1983; Harris et al. 1984), i.e., an increase in the strength of synaptic transmission brought about by repetitive synaptic activation. Because LTP can persist for weeks, it has been suggested that an LTP-like process is involved in memory formation. This idea is reinforced by the observation that NMDA receptor blockade can produce amnesia in animals (Morris et al. 1986; Staubli et al. 1989). Another outcome of excessive NMDA receptor activation is initiation of epilepsy-like neuronal activity (Dingledine et al. 1986). The AMPA-type receptors are responsible for the bulk of excitatory transmission at brain synapses (Andreasen et al. 1989; Lovinger 1991b). Thus, significant decreases in the function of these receptors may have disastrous consequences, including loss of consciousness, severe sensory and movement impairment, and respiratory failure.

A large body of evidence suggests that the inhibitory effect of alcohol on NMDA receptors contributes to

intoxication.

The actions of alcohol on these two types of glutamate receptors have been the subject of intense investigation. Alcohol inhibits glutamate receptor function; NMDA receptors are more sensitive to alcohol than AMPA receptors (DildyMayfield and Leslie 1991; Gothert and Fink 1989; Hoffman et al. 1989; Lovinger et al. 1989). Alcohol inhibits NMDA receptor function in brain slices (Lovinger et al. 1989; Martin et al. 1991); a similar effect is seen in isolated neurons in culture (Dildy-Mayfield and Leslie 1991; Hoffman et al. 1989; Lovinger et al. 1989; White et al. 1990) and even in single NMDA receptor molecules examined in small regions of neuronal membrane (Lima-Landman and Albuquerque 1989) when the patch-clamp recording technique is used. These data suggest that the effects of alcohol at this site are derived from an action on the NMDA receptor itself or the membrane surrounding the receptor and are not secondary to other changes within the cell. It has been demonstrated that the amino acid glycine, which sometimes acts as a neurotransmitter, can over

come this inhibitory action of alcohol (DildyMayfield and Leslie 1991; Rabe and Tabakoff 1990; Woodward and Gonzalez 1990). This action may result from glycine's interaction with a portion of the receptor at which glycine acts, thereby increasing channel function (Johnson and Ascher 1987).

A large body of evidence suggests that the inhibitory effect of alcohol on NMDA receptors contributes to intoxication. First, animals trained to respond in a particular fashion during alcohol intoxication will respond similarly when given drugs such as phencyclidine (PCP, or “angel dust") which also act on NMDA receptors (Grant, Knisely et al. 1991). Second, the effects of alcohol in vivo will potentiate the effects of NMDA antagonists in a manner that suggests an overlap in their sites of action (Balster and Wessinger 1983; Fidecka and Langwinski 1989; Wilson et al. 1990). Finally, alcohol can inhibit electrophysiologic responses to NMDA in the intact animal (Simson et al. 1991).

What behavioral changes during intoxication might involve NMDA receptors? The most obvious is the cognitive impairment and amnesia observed at moderate to high alcohol concentrations. Acute alcohol exposure decreases the learning of new information (Lister et al. 1987). Blockade of LTP by alcohol has been reported (Mulkeen et al. 1987; Sinclair and Lo 1986), even at very low concentrations (Blitzer et al. 1990). Research in several laboratories is directed at elucidating alcohol's effects on cognitive phenomena (Cermak 1990). Knowledge gained from such studies will allow for a better understanding of the effect of alcohol on NMDA receptors in these phenomena. Alcohol also reduces convulsant (e.g., epileptic) effects of certain drugs (Ticku 1990); such effects may arise from NMDA receptor inhibition. In addition, alcohol reduces the severity of a stimulus-induced bursting model of epilepsy in a manner similar to other NMDA receptor antagonists (Martin et al. 1991).

Clinical implications for the study of acute alcohol intoxication and NMDA receptors are numerous. Alcohol and other abused agents, such as PCP, act at the NMDA receptor. Because these drugs may be co-abused, understanding their interactions at the molecular level might help explain synergistic behavioral effects of the drugs. It may also be possible to develop treatments involving the NMDA receptor that could overcome the amnesic effects of alcohol.

It is difficult to evaluate the role of AMPA receptors in alcohol's actions because few drugs that act at this receptor are suitable for in vivo use. However, high concentrations of alcohol appear to affect AMPA receptors (Gothert and Fink 1989; Hoffman et al. 1989; Lovinger et al. 1989). Some reports have also demonstrated effects at lower concentrations in certain preparations (Dildy-Mayfield et al. 1991). In mice, high doses of alcohol have been shown to prevent the lethal effects of kainic acid on its receptor, which, like the AMPA receptor, functions as an excitatory receptor (Ticku 1990). These receptors may also participate in the general anesthesia and respiratory depression that occur with extremely high blood alcohol levels.

Recent studies have led to the elucidation of the structure of AMPA- and NMDA-type receptors (Hollman et al. 1989; Moriyoshi et al. 1991). Receptors of the AMPA type may be formed from a variety of subunits in a manner similar to that seen with the GABAA receptor (Keinanen et al. 1990). Although data on the NMDA receptor are not extensive, it is probable that multiple subunits for this receptor will be found (Monyer et al. 1992).

The discovery that subtypes of glutamate receptors exist suggests that they may exhibit subtle differences in the response to alcohol. Further, understanding the nature of alcohol's action at glutamate receptors may aid in the development of treatments for acute intoxication. For example, drugs that target only certain receptor subtypes may be useful if they can counteract effects on the most alcohol-sensitive subtypes. However, more basic research on alcohol's effects at these receptors, such as research to determine which combination of subunits exhibits maximum sensitivity to alcohol, is needed before therapeutic uses can be pursued.

Alcohol's effects on dopaminergic and
serotonergic neurotransmission

Part of the reason for long-term alcohol abuse lies in the fact that the neural effects of alcohol are in some way reinforcing or rewarding. Neuroscientific approaches have been quite helpful in determining the neural basis of alcohol reinforcement. Studies of the neural basis of action of strongly reinforcing and highly addictive drugs such as cocaine and amphetamines suggest a role for the monoamine neurotransmitters dopamine and serotonin in this process (see chapter 5).

Alcohol ingestion produces increases in dopamine levels in the nucleus accumbens and other reward centers (Carboni et al. 1989; Wozniak et al. 1990). In addition, neuronal activity is increased just after alcohol consumption in areas of the brain that are rich in dopamine, including the accumbens and the olfactory tubercle (Lewis et al. 1990). The evidence that these increases in dopamine levels contribute to reinforcement by alcohol is reviewed in chapter 5. Studies suggest the involvement of the D2 type of dopamine receptor in the reinforcing effects of alcohol (McBride et al. 1990).

The discovery that subtypes of glutamate receptors exist suggests that they may exhibit subtle differences in the response to alcohol.... Understanding the nature of alcohol's action at glutamate receptors may aid in the development of treatments for acute intoxication.

The mechanism of alcohol's action on dopaminergic transmission is unclear. The most likely mechanism underlying alcohol-induced increases in brain dopamine levels is increased activity of dopaminergic neurons in the ventral tegmental area (VTA) of the brain (Gessa et al. 1985). Increases in the activity of VTA neurons would increase release of dopamine from the presynaptic terminals of these neurons in the nucleus accumbens. The molecular basis of this increased activity is not yet understood. Changes in the electrical properties of the VTA neurons themselves appear to occur, but changes in transmission onto these neurons have yet to be determined. Alcohol's effect appears to occur within the VTA because altered firing rates have been observed mainly in this region of the brain (Brodie et al. 1990; Shefner 1990). Alcohol can also stimulate dopamine release from slices of the neostriatum (Russell et al. 1988); this action presumably takes place at dopamine-containing axon terminals. Additional evidence suggests that serotonin may be involved in alcohol's effect on dopamine levels, as discussed below.

Serotonin, or 5-hydroxytryptamine (5-HT), another monoamine neurotransmitter, has a long history of association with the effects of alcohol. For example, it has been demonstrated repeatedly that drugs that increase 5-HT levels in the brain by blocking the reuptake of 5-HT into

neurons also reduce alcohol consumption (Amit et al. 1984; Lawrin et al. 1986; Murphy et al. 1988). (These effects are discussed in depth in chapter 5.) Acute alcohol exposure also appears to affect 5-HT levels in the brain by potentiating the action of a serotonin transport protein located on presynaptic terminals (Alexi and Azmitia 1991).

it appears that both dopaminergic and serotonergic neurotransmissions are altered during acute exposure to alcohol.

Serotonin uptake inhibitors have a general agonistic effect because they increase 5-HT levels wherever the neurotransmitter is released. Research is now aimed at determining the types of 5-HT receptors activated and the brain loci at which these effects take place. At last count, seven types of 5-HT receptor were known to exist (Julius 1991). However, drugs that selectively act at some of the subtypes have not been developed, so differentiation of receptor involvement is difficult. Several studies suggest a direct interaction between alcohol and the 5-HT3 receptor, which belongs to the ligand-gated ion channel family (Derkach et al. 1989; Peters and Lambert 1989; Yakel and Jackson 1988). These studies, in turn, have led researchers to investigate the location of the 5-HT3 receptors that are affected by alcohol. For example, selective 5-HT3 antagonists attenuate alcohol-induced increases in dopamine in the nucleus accumbens in vivo (Carboni et al. 1989; Wozniak et al. 1990). This effect was observed even with local alcohol administration, thus suggesting that the receptors of interest are present in the nucleus accumbens. It has also been reported that 5-HT3 antagonists can block alcohol-induced increases in 5-HT uptake in rat forebrain slices (Alexi and Azmitia 1991). These brain regions may be the appropriate choices in the search for the receptors affected by alcohol.

The function of the 5-HT3 receptor can be directly assayed by studying cells that contain the receptor. Acute exposure to alcohol potentiates 5HT3 receptor function in isolated cells (Lovinger 1991a, Lovinger and White 1991)—an action similar to the effect of alcohol at the GABAA receptor in that not all cells exhibit potentiation. With the cloning of the receptor (Maricq et al. 1991), researchers can begin to determine

whether alcohol sensitivity is critically dependent on structural features of the receptor.

In sum, it appears that both dopaminergic and serotonergic neurotransmissions are altered during acute exposure to alcohol. The sites responsible for controlling neurotransmitter levels in the brain, as well as the receptors that respond to these transmitters, are the focus of continuing research.

Chronic Effects of Alcohol on Neural Function

Tolerance

Prolonged alcohol exposure renders the brain less sensitive to the acute actions of alcohol. For example, the doses of alcohol needed to produce sedation in an animal that has been given alcohol for several days are higher than those needed to produce sedation upon initial exposure to the drug. This tolerance to alcohol develops as a result of changes in alcohol's effects on the brain as well as an increased capacity of the body to reduce alcohol levels via metabolism. The following overview deals with the aspects of tolerance that are due to changes in alcohol's neural effects. It is thought that tolerance results from the brain's reversing or overcoming the cellular and molecular effects of acute alcohol exposure. This process involves long-term alterations in neuronal function and structure, most notably at the synapses. Recent work suggests that lasting changes in receptor-mediated effects account for some forms of tolerance.

Changes in GABAA receptors following chronic alcohol exposure. Evidence that acute alcohol exposure alters the function of certain receptors suggests that neurons may compensate for this deleterious effect by altering the number or function of these receptors. Research supporting this hypothesis has begun to accumulate. For example, drugs that potentiate GABAA receptor function often show cross-tolerance to ethanol (see Dietrich 1987 for a review). Thus, an animal's sensitivity to these drugs will decrease with chronic alcohol exposure, suggesting some change in GABAA receptors in tolerant animals.

Animals exposed to alcohol for days have decreased GABAA receptor function in neuronal tissue (Allan and Harris 1987; Morrow et al. 1988). One possible explanation for this change in receptor function is a change in the expression (i.e., the DNA and RNA-driven production) of the subunit components of the receptor. Re