(Brennan et al. 1990). These findings suggest, but do not prove, that the observed interstrain differences are directly responsible for the differences in seizure susceptibility between WSP and WSR mice. Finding the genetic loci that differ between the strains may help us to locate key factors in the development of withdrawal seizures.

It must be emphasized that the neural consequences of alcohol withdrawal are reversible with continued abstinence. However, the time it takes symptoms to disappear varies depending on the behavioral patterns established during the period of chronic alcohol intake. The severity of alcohol withdrawal seizures is usually maximal 8 to 12 hours after the beginning of withdrawal (Gulya, Grant et al. 1991). Seizure susceptibility is considerably decreased 1 day after withdrawal. In contrast, reduction of anxiety following the initial anxiogenic withdrawal period can take several days (Lal et al. 1991). The causes of these different time courses of recovery probably stem from the lifetimes of the underlying cellular and molecular events. For example, it has been observed that changes in NMDA receptor numbers parallel withdrawal seizure susceptibility (Gulya, Grant et al. 1991).

Alcoholics with a history of decades of abuse often exhibit the type of vitamin B1 (thiamine)-deficiencyinduced brain damage characteristic of WernickeKorsakoff disorder, which appears to involve excessive NMDA receptor activation.

The long-term consequences of changes in NMDA and GABA receptors may be disastrous. It is known that excessive glutamate receptor activation leads to death of the neuron (see Choi 1988 for a review). This process is known as excitotoxicity because the protracted increase in neuronal activity caused by glutamate receptor activation ultimately leads to neuronal death. Thus, this excitatory process has the potential to contribute to brain damage after prolonged alcohol abuse. Further, increases in numbers of NMDA receptors would increase the potential for excitotoxicity. Decreases in GABAergic transmission would unmask excitatory events and perhaps also contribute to excitotoxicity. Alcoholics with a history of decades of abuse often exhibit the type of vitamin B1 (thiamine)-deficiency

induced brain damage characteristic of WernickeKorsakoff disorder, which appears to involve excessive NMDA receptor activation (Berman 1990). It now appears that NMDA receptor antagonists can prevent neuronal loss in certain brain areas in an animal model of thiamine deficiency (Langlais and Mair 1990). These data point out a possible involvement of excitotoxicity in alcoholic brain damage.

Brain Imaging Techniques:
A Window Into the Effects of
Alcohol on Brain Structure and
Function

Until recently, studies of the effects of alcohol on the human brain could be conducted only after a person's death, when the brain could be physically examined. Postmortem studies permitted researchers to examine the pathological changes that can result from heavy drinking, yet they provided little information about the effects of alcohol on the biochemical pathways in the living brain.

The development of noninvasive imaging techniques has enabled researchers to explore the living brain and probe its biochemistry and physiology. Some imaging tools provide structural information about the size, shape, and physical integrity of the brain. Other techniques assess functional activity by measuring electrical activity, blood flow, oxygen and glucose use, and neurotransmitter activity. Each technique has distinctive strengths; the coupling of two or more approaches can provide multifaceted information about brain structure and function.

Computerized tomography and magnetic resonance imaging give structural information

The earliest technique developed for in vivo imaging of the human brain, computerized tomography (CT), uses x-ray bombardment at multiple angles into the head to give a threedimensional picture of structures in the brain (figure 2). Brain images are divided into sections 5 to 10 mm thick and are displayed on a computer screen. Areas of approximately 1 mm2 can be discerned. Thus, individual cells cannot be seen, but the gross structure of a small region is observable. Using this technique, researchers have seen shrinkage of brain tissue in alcoholics after long-term abuse of alcohol (Lishman 1990; Wilkinson 1987). The sites of shrinkage in these images correspond to neural tissue. Some evi

modest decreases in blood flow in patients suffering from Korsakoff's syndrome, a form of psychosis often seen in patients with a long history of severe alcohol abuse. Decreases in blood flow in the frontal cortex are commonly associated with cognitive dysfunction in these patients (Berglund et al. 1987).

Blood flow measurements can be altered by changes in cerebral blood vessel function as well as by neuronal activity. Thus, to measure neuronal activity more accurately, researchers have used 2-deoxyglucose, a sugar that is taken up into neurons but cannot be used for energy (figure 3). Because neurons use only glucose as an energy source, they require more glucose during periods of increased activity. The deoxyglucose is not metabolized immediately, like normal glucose, and thus is present in the cell for periods long enough to permit detection. Studies have demonstrated that selected brain areas show increased deoxyglucose uptake during periods of enhanced activity.

The few studies conducted of alcoholic subjects have reported that neuronal activity in “resting" alcoholics is not markedly different from that in nonalcoholics.

Studies of nonalcoholic subjects reveal that alcohol ingestion produces a general decrease in neuronal activity (de Wit et al. 1990). The few studies conducted of alcoholic subjects have reported that neuronal activity in "resting" alcoholics is not markedly different from that in nonalcoholics (Eckardt et al. 1990). It remains to be seen whether levels in particular regions differ when subjects are asked to perform cognitively demanding tasks. PET imaging can also be used to detect isotopes that bind to specific neurotransmitter receptors in the brain. Studies using PET technology to examine dopamine receptors in psychotic individuals have already been carried out (Waddington 1989), and thus it should be possible to study this receptor system in vivo in chronic alcoholics.

The imaging techniques discussed show great promise for clinical use. The use of such techniques should help to determine differences in alcohol sensitivity between individuals. These techniques may also be useful in evaluating the extent of alcohol tolerance by examining

changes in the response to acute alcohol exposure following chronic abuse. Changes in activity in specific brain regions thought to be important for "reward" may be useful for assessing alcohol dependence. Finally, structural and functional information will probably allow for better determination of the extent of brain damage following prolonged alcohol abuse.

Summary

Advances in neuroscience have provided much knowledge of alcohol's acute and chronic actions on the central and peripheral nervous systems. These advances have been facilitated by the development of such sophisticated new research tools as electrophysiological, imaging, and molecular biology techniques, which are enabling neuroscience researchers to analyze alcohol's effects on systems and regions in the brain. Through these research efforts, we are gaining insight into the chemical and physiological processes that underlie alcohol addiction.

Acute exposure to alcohol produces numerous behavioral effects. Although the molecular site(s) of alcohol's actions on neurons is not yet clear, researchers have hypothesized that alcohol may work by perturbing lipids in the cell membrane of the neuron, interacting directly with the hydrophobic region of neuronal membrane proteins, or interacting directly with a lipid-free enzyme protein in the membrane.

Whether alcohol acts on lipids or proteins in the neuronal cell membrane, it is clear that alcohol alters the function of neuron-specific proteins. For example, evidence suggests that the activity of the chloride ion channel linked to the A-type receptor of the GABA neurotransmitter increases during exposure to intoxicating amounts of alcohol. GABA is the major inhibitory neurotransmitter in the mammalian brain. The effects of alcohol on the GABAA receptor may contribute to the anxiolytic, sedative, and motor impairment actions of alcohol. Recent findings have shown that many subtypes of the GABAA receptor exist; these subtypes consist of a combination of varying forms of five subunits. Using mice selectively bred for their sensitivity to alcohol, researchers have compared the GABAA receptor in these animals and in alcohol-insensitive mouse strains to determine whether differences in this receptor may account for increased sensitivity to alcohol. These studies have shown that a specific GABAA receptor subunit (e.g., gamma) with a site for a phosphate molecule (added as a post