for alcohol, both of which are influenced by genetic effects. Heritability of drinking frequency was estimated at 0.66 for women and 0.42 to 0.75 for men; heritability of drinking quantity was estimated to be 0.57 for women and between 0.24 and 0.61 for men.

Because the prevalence, age at onset, clinical features, course, and outcome of alcoholism appear to differ in men and women, it may be assumed that risk factors for alcoholism also differ for the two sexes. Kendler et al. (1992) interviewed 1,030 same-sex female twin pairs to examine whether, as in men, genetic factors have a role in the etiology of alcoholism in women. Within the sample population, 185 women met the criteria established by the revised edition of DSM-III (DSM-III-R) (American Psychiatric Association 1987) for alcohol dependence and 357 met the broad criteria for alcoholism. The investigators found that concordance for alcoholism was consistently higher in identical than in fraternal twin pairs. Based on these data, the researchers estimated that genetic influences account for 50 to 61 percent of the risk for alcoholism in women.

Various factors, such as assortative mating and the effects of alcohol on examined individuals, can complicate efforts to interpret data from studies that examine familial transmission of alcoholism (Hesselbrock et al. 1991; Newlin and Thomson 1990). Assortative mating occurs when individuals with similar traits tend to mate with one another more often than chance would dictate. Such mating can confound efforts to identify the mode of inheritance of alcoholism. Further, among drinkers, spousal consumption of alcohol is significantly intercorrelated (Gleiberman et al. 1992; Tambs and Vaglum 1990). In addition, offspring of alcoholics and heavy drinkers tend to marry alcoholics (Boye-Beaman et al. 1991). These factors complicate efforts to separate the effects of alcoholism in one parent from the effects of drinking or the family drinking history in the other parent.

Maternal drinking during pregnancy is another confounding factor from which a spectrum of congenital injuries in offspring can arise (see Chapter 9, Effects of Alcohol on Fetal and Postnatal Development). As well, some evidence suggests that paternal drinking may adversely affect offspring. Little and Sing (1986) noted that infants fathered by regular drinkers weighed less than those fathered by nondrinkers, even after they had controlled for the effects of maternal alcohol and nicotine use. At present, however, it is

not clear whether paternal drinking directly affects birth weight in human offspring. An association between paternal drinking and injury in offspring has been demonstrated in animals: Offspring of male rats exposed to large amounts of alcohol prior to and at the point of conception have tended toward lower birth weight (Abel and Lee 1988), increased susceptibility to infection and stress (Abel et al. 1990), poorer maze learning (Abel and Lee 1988), improved passive avoidance (Abel and Lee 1988), and reduced active avoidance (Abel and Tan 1988). In light of these findings, it is possible that some of the effects that appear to be genetically influenced in alcoholic families, in part, may be consequences of paternal and maternal drinking before and after conception.

variables that may determine drinking behavior. Animals also can be used to test various hypotheses about the underlying neurochemical, neurophysiological, and neuroanatomical bases for responsiveness to alcohol.

Although most animals are innately averse to drinking alcohol, some will voluntarily consume large quantities. By selectively breeding animals with similar preferences for alcohol (i.e., breeding preferring animals with preferring animals and nonpreferring animals with nonpreferring animals), researchers can establish animal strains that either avoid or prefer alcohol. Studies of selectively bred animals have demonstrated that genetic factors affect many alcohol-related traits. Using these animals, scientists have determined that genetic factors may independently moderate many alcohol-related behaviors and responses that are expressed in humans-among them, the tendency to self-administer alcohol; sensitivity to locomotor activation by alcohol (which may model the euphoric effects of alcohol in humans); anxiolytic, sedative, and analgesic effects induced by alcohol; development of tolerance for alcohol's effects; rate and process of alcohol metabolism; alcohol-induced motor discoordination and hypothermia; and sensitivity to withdrawal from alcohol. The ability to breed lines of animals with high or low measures of these traits indicates that the traits are influenced by genetic factors (table 2).

Genetic factors help to determine variation in central nervous system neurotransmitter systems. These genetically determined differences may influence both the voluntary intake of alcohol and the development of complications that can follow alcohol consumption. Analysis of the behav

ior, neurophysiology, psychophysiology, and neurochemistry of various animal strains may help determine the features that differentiate a person who is at heightened risk for an alcoholrelated problem from one who does not have this vulnerability.

Alcohol-Seeking Behavior

Animal studies have proved to be useful in research addressing the complex issue of why people drink. Alcohol has reinforcing properties that may contribute to initial drinking, continued consumption, and development of alcohol dependence. Researchers have used animals to examine the behavioral and biological processes of reinforcement and the relationship of these processes to alcohol-seeking behavior.

There are two basic types of reinforcement, positive and negative (see Chapter 5, Neurobehavioral Aspects of Alcohol Consumption). Positive reinforcement involves a rewarding reinforcer and is demonstrated by an animal or a person performing an act to obtain the effect of a reward. The positively reinforcing effects of alcohol induce a spectrum of subjective states ranging from pleasant sensations to euphoria. Negative reinforcement, on the other hand, involves the relief of something unpleasant and is characterized by an animal or a person performing an act to remove an aversive effect, such as anxiety or pain. The anxiety-reducing effects of alcohol are considered to be negatively reinforcing. It appears that distinctive neural circuits provide the physiological basis for euphoria and anxiety. Evidence suggests that alcohol may differentially affect these systems and that genetic

factors may influence the differential effect (Pihl and Peterson 1992).

Researchers have begun to investigate the underlying mechanisms responsible for alcohol's rewarding effects. Alcohol appears to act on brain systems that are affected by the neurotransmitter dopamine (Wise 1988). The dopaminergic psychomotor system in the brain seemingly provides the neurophysiological underpinnings for behaviors that occur with various motivated states, such as those that follow the presentation of a positive reinforcer. When activated, this brain system causes an animal or a person to interact with or explore relevant environmental stimuli (Wise 1988). Such activation appears to be intrinsically rewarding. In animal studies, this behavior is used to assess the positively reinforcing effects of alcohol. Activity following alcohol administration can indicate that an animal is sensitive to alcohol's rewarding effects. Activation of this brain system in humans is accompanied by sensations of euphoria, expansiveness, and enhanced power and energy (Pihl and Peterson

1992).

Several rodent lines have been bred selectively for traits that may be associated with the positively reinforcing properties of alcohol. Among these are the alcohol-preferring (P) line of rats, with a high preference for alcohol, and the alcohol-nonpreferring (NP) line of rats, with a low alcohol preference (Li et al. 1981). The successful breeding of P and NP rats provided an experimental model for the controlled study of possible genetic and biological determinants of alcohol-seeking behavior. In addition, various characteristics of the P line make it a useful animal model for alcoholism. For example, these animals voluntarily consume large amounts of alcohol to achieve high blood alcohol concentrations; they will work to obtain alcohol even when food and water are readily available; they develop a physical dependence on alcohol and a tolerance to its depressive effects, and they become susceptible to the withdrawal syndrome.

Schwarz-Stevens et al. (1991) noted that P rats tend to maintain bar-pressing that is reinforced by alcohol administration in a variety of circumstances, even when high concentrations of alcohol are administered. Other investigators have noted that these rats have an increased resistance to alcohol-induced sedation (Schechter 1992) and an increased sensitivity to alcohol's psychomotor stimulant properties (Murphy et al. 1990). The psychomotor stimulant and sedative effects of alcohol appear to be under independent ge

netic control (Dudek et al. 1991; Phillips and Dudek 1991).

Overall, studies using P and NP rats suggest a relation between the rewarding properties of alcohol and the subsequent chronic self-administration of the drug to the point of maladaptive intoxication.

Two selectively bred strains of mice, labeled FAST and SLOW, have also been used to study traits that may be linked to the reinforcing properties of alcohol. FAST and SLOW mice display heightened and diminished responses, respectively, to the stimulatory effect of a low dose of alcohol on their activity in an open field (Crabbe et al. 1987). A recent study by Phillips et al. (1991) showed that FAST mice given daily doses of alcohol for 9 days responded increasingly to alcohol's stimulatory effect. With the same dosing regimen, SLOW mice initially exhibited diminished locomotor activity. After 9 days of dosing, however, these mice developed tolerance to alcohol's depressant effects. Findings from a subsequent study demonstrated that SLOW mice eventually developed a stimulant response to alcohol after 31 days of dosing. These results demonstrate that tolerance develops to the depressant but not the activating effects of alcohol on locomotor activity. Such information may provide insight into the role of reinforcement in the development of alcohol addiction. If humans, like FAST mice, do not develop tolerance to the rewarding effects of alcohol, alcohol use may be reinforced (Crabbe and Phillips 1990).

Molecular biology has provided valuable tools for exploring the genetic mechanisms responsible for alcohol-seeking behavior.

Molecular biology has provided valuable tools for exploring the genetic mechanisms responsible for alcohol-seeking behavior. For example, Goldman et al. (1987) used molecular biology techniques in research on inbred mouse lines that differed in alcohol preference or nonpreference. The investigators noted that variation in the gene responsible for the synthesis of the LTW-4 protein is associated with increased alcohol consumption in both mouse strains. This protein is abundant in the brain, liver, and kidneys of these animals. A correlation was found

between the possession of a variant of this gene and preference for alcohol. The correlation may indicate that (a) alcohol preference arises directly from the properties of this gene and its product or (b) that alcohol preference arises from the properties of a nearby gene that is closely linked to the LTW-4 gene.

Sedation

Various animal lines have been selectively bred for their responsiveness to the sedative, or anesthetic, effects of alcohol. Researchers have measured this responsiveness by an animal's sensitivity or resistance to alcohol-induced loss of righting reflex, referred to as sleep time, which is a genetically correlated trait. For example, the much-studied long-sleep (LS) and shortsleep (SS) mouse lines were bred for the characteristic differences they exhibit in the duration of their loss of righting reflex after alcohol administration (McClearn and Kakihana 1981). LS mice require a longer period of time to regain the righting reflex (i.e., sleep time) than do SS mice. Further, LS mice lose the righting reflex at approximately one-half the concentration of alcohol in the brain that causes SS mice to demonstrate this behavior (Smolen and Smolen 1989). Sensitivity to alcohol-induced loss of righting reflex in these mice appears to be determined by additive gene effects (Dudek and Abbott 1984).

Studies in these mouse lines have demonstrated that Purkinje cells in the cerebellum are a possible site for the sedative actions of alcohol.2 Sorenson et al. (1980) demonstrated that local administration of alcohol directly to Purkinje cells inhibited spontaneous activity of these cells at doses 30-fold lower in LS mice than in SS mice.

More recently, studies of LS and SS mice have focused on the neurochemical mechanisms that may underlie the differential sensitivity of these mice to the sedative effects of alcohol. Alcohol, like benzodiazepines and barbiturates, acts at the receptor for the inhibitory neurotransmitter y-aminobutyric acid (GABA). The actions of alcohol at the GABA receptor augment the flux of chloride ions through channels regulated by GABA, thus reducing the ability of neurons to fire (Warneke 1991) (see Chapter 4, Actions of Alcohol on the Brain). Several investigators have observed that alcohol and barbiturates may en

hance chloride movement independent of GABA (Mehta and Ticku 1988; Zorumski and Isenberg 1991). Evidence suggests that a genetically determined relationship exists between sensitivity to alcohol and the function of chloride channels (Harris and Allan 1989).

Allan et al. (1988) examined the effects of alcohol on chloride channels regulated by GABA in brain membranes isolated from LS and SS mice. These investigators found that membranes from LS mice were more sensitive to alcohol's effect on GABA-regulated chloride channels than membranes prepared from SS mice.

Wafford et al. (1990) used a newly developed and innovative technique in molecular biology to transfer messenger ribonucleic acid (mRNA) from the brains of LS and SS mice to oocytes of the Xenopus frog. mRNA, which is transcribed from deoxyribonucleic acid (DNA), serves as a template to direct the synthesis of proteins. Although frog oocytes do not naturally possess GABA receptors, the transfected oocytes expressed receptors that showed pharmacological responses typical for the A subtype of the GABA receptor (GABAA) from LS and SS mice. In oocytes injected with mRNA from LS mice, alcohol augmented the action of GABA; this response was similar to the response cited above for LS brain preparations. In contrast, alcohol inhibited GABA effects in oocytes injected with mRNA from SS mice. The investigators proposed that genetic differences in GABA receptor function may be related to the differential sensitivity to alcohol of LS and SS mice. A subsequent study by Wafford et al. (1991) demonstrated that a specific subunit of the GABA receptor is required for alcohol's enhancing effect on GABA actions. Whether this subunit contributes to genetic differences in sensitivity to alcohol has yet to be determined.

Another selectively bred strain of rats, known as high-alcohol-sensitive (HAS) and low-alcoholsensitive (LAS) rats, also have been characterized for their sensitivity to the sedative effects of alcohol (reviewed in Crabbe 1989, Crabbe and Phillips 1990, and Deitrich et al. 1989). Using these selectively bred rats, Allan et al. (1991) demonstrated that HAS rats were more sensitive than LAS rats to the potentiating effect of alcohol on GABA, thus suggesting that a genetic correlation exists between initial sensitivity to alcohol and the GABA-receptor chloride complex.

Serotonergic Interactions

A number of lines of evidence suggest that serotonin, a neurotransmitter involved in mood, sleep, and consummatory behavior, may modify the reactions of humans and animals to alcohol. Various studies have linked central serotonergic dysfunction in humans to depressed (Meltzer 1990), anxious (Charney et al. 1990; Taylor 1990), and aggressive or antisocial behavior (reviewed in Pihl et al. 1990a and Tabakoff and Hoffman 1991). These are interesting connections with regards to alcoholism, especially given the strength of the connection between antisocial behavior and alcoholism (see Chapter 2, Psychiatric Comorbidity With Alcohol Use Disorders).

Animal studies have demonstrated that in several regions of the brain P rats (Murphy et al. 1982, 1987) and high-alcohol-drinking rats (HAD) (McBride et al. 1990) have lower levels of serotonin and its primary metabolite 5-hydroxyindoleacetic acid (5-HIAA) than NP and lowalcohol-drinking rats. McBride et al. (1990) suggested that diminished serotonin levels may be a consequence of decreased activation of serotonin systems. The acute administration of high doses of alcohol appears to increase central serotonin activity (McBride et al. 1991). In particular, this increase in P and HAS rats appears to be due to enhanced activity of serotonin systems in dorsal raphe nuclei in the brain (McBride et al. 1990) (see chapter 4). These findings suggest that certain serotonin pathways are involved in controlling drinking behavior.

Withdrawal

Withdrawal-seizure-prone (WSP) mice and withdrawal-seizure-resistant (WSR) mice are characterized by their display of severe or mild convulsions, respectively, after chronic alcohol intake (Crabbe et al. 1985). The symptoms of withdrawal that result from discontinuation of drug intake are the hallmark of physical dependence. These symptoms are 10 times more severe in WSP mice than in WSR mice, despite the fact that the two lines achieve equal blood alcohol levels during chronic alcohol treatment (Kosobud and Crabbe 1986; Phillips et al. 1989). The ability to select for intensity of withdrawal in these mice demonstrates that the development of physical dependence has a genetic component. These strains, however, do not differ in their sensitivity to other alcohol effects, including the development of tolerance (Crabbe and

Kosobud 1986; Phillips et al. 1989). Thus, it appears that genes that convey vulnerability to physical dependence on alcohol may differ from those that influence the initial sensitivity and development of tolerance to alcohol. Further, the increased sensitivity that WSP mice also display to other drugs, such as selected barbiturates, other alcohols, benzodiazepines, and nitrous oxide, suggests that individuals who have inherited facilitative genes for alcoholism may also be vulnerable to dependence on other drugs (Belknap et al. 1988, 1989; Crabbe et al. 1991a).

A number of lines of evidence suggest that serotonin, a neurotransmitter involved in mood, sleep, and consummatory behavior, may modify the reactions of humans and animals to alcohol.

Researchers have begun to investigate the various neurochemical pathways that may be involved in alcohol withdrawal as manifested by WSP mice. Glutamate, the major excitatory neurotransmitter in the brain, has been central to such studies (see chapter 4). Alcohol primarily affects the glutamate receptor subtype named for the chemical N-methyl-D-aspartate (NMDA). Binding of glutamate or NMDA to this receptor causes an influx of calcium ions into the cell, thereby augmenting the neuron's potential to fire. Low concentrations of alcohol inhibit the NMDA-stimulated calcium flux (Hoffman et al. 1989), whereas chronic alcohol administration produces an adaptive increase in the number of NMDA receptors in the hippocampus, a region of the brain associated with seizures (Grant et al. 1990). This adaptive process could be a major factor in physical dependence.

WSP mice have more hippocampal NMDA receptors than WSR mice (Valverius et al. 1990), and NMDA administration exacerbates the severity of the withdrawal exhibited by WSP mice (Crabbe et al. 1991b). In contrast, NMDA has no effect on alcohol-naive WSP mice or alcoholnaive or alcohol-withdrawing WSR mice.

Tolerance

The phenomenon of tolerance-the diminished effect of a drug on repeated exposurehas also been studied in various genetic animal models. Such studies are advancing our