sign in sign up
<<
Home , Acetylcholine, Calcium channel blocker, Channel blocker, Drug withdrawal, GABAA receptor, Neurochemical

Mechanisms Underlying Withdrawal

John Littleton, MD, Ph.D.

More than 50 years ago, C.K. Himmelsbach first suggested that physiological mechanisms responsible for maintaining a stable state of equilibrium (i.e., ) in the patient’s body and are responsible for tolerance and the drug withdrawal syndrome. In the latter case, he suggested that the absence of the drug leaves these same homeostatic mechanisms exposed, leading to the withdrawal syndrome. This theory provides the framework for a majority of neurochemical investigations of the adaptations that occur in and how these adaptations may precipitate withdrawal. This article examines the Himmelsbach theory and its application to alcohol withdrawal; reviews the animal models being used to study withdrawal; and looks at the postulated neuroadaptations in three systems—the gamma-aminobutyric acid (GABA) system, the glutamate neurotransmitter system, and the channel system that regulates various processes inside . The role of these neuroadaptations in withdrawal and the clinical implications of this research also are considered. KEY WORDS: AOD withdrawal syndrome; ; biochemical mechanism; AOD tolerance; brain; homeostasis; biological AOD dependence; biological AOD use; disorder theory; biological adaptation; animal model; GABA receptors; glutamate receptors; ; ; detoxification; ; disease severity; AODD (alcohol and other drug dependence) ; literature review

uring the past 25 years research- science models used to study with- of the reasons why advances in basic ers have made rapid progress drawal neurochemistry as well as a research have not yet been translated Din understanding the chemi- reluctance on the part of clinicians to into therapeutic gains and suggests cal activities that occur in the nervous consider new treatments. No reason areas for future improvement. system (i.e., the neurochemistry) during exists to criticize either scientists or alcohol withdrawal (AW), primarily clinicians for this situation; from the from experiments using animal models basic science viewpoint, it was necessary Neurochemical Concepts of alcohol dependence (Samson and to understand the neurochemical of Physiological Dependence Harris 1992). At the same time, advances processes in relatively simple models have been made in the clinical treat- first, and from the clinicians’ viewpoint, It has been said that Darwin’s theory ment of AW, making detoxification it made no to change withdrawal of natural selection is the “single best a routine procedure that no longer is treatments (e.g., the use of benzodi- idea” in biology. A candidate for a life-threatening (Romach and Sellers azepines) that had proven successful. similar tribute in drug-dependence 1991). For the most part, however, Now may be the right time to research is the suggestion made by these treatment advances have evolved rethink our goals for both research independently from the developments and treatment of AW, however. This in neurochemical understanding. The article reviews current concepts and JOHN LITTLETON, MD, PH.D., reasons for this independence are com- developments in basic research on the is a professor in the Department of plex, and they include the limited neurochemical mechanisms underlying , University of Kentucky, clinical relevance of some of the basic AW. The article also considers some Lexington, Kentucky.

Vol. 22, No. 1, 1998 13 C.K. Himmelsbach more than 50 years seesaw (see figure 1). The brain is in same: an adaptive neurochemical ago regarding the relation of toler- a balanced state before alcohol expo- response designed to oppose alcohol’s ance, dependence, and withdrawal sure. When a person consumes acute neurochemical effects. It should (Himmelsbach 1941). Based on his alcohol, its acute effect (which often be emphasized, however, that some observations of -dependent is the effect the person is seeking) is to other types of tolerance mechanisms patients, Himmelsbach suggested that unbalance the neurochemical equilib- can develop without necessarily causing physiological mechanisms which rium of the brain, thereby tilting the a withdrawal syndrome and, as dis- maintain a stable state of equilibrium seesaw. If the alcohol exposure contin- cussed later, this simple scheme does (i.e., homeostasis) in the patient’s ues, the brain institutes an opposing not take into account what may hap- body and brain are responsible for adaptation (i.e., neuroadaptation) to pen after repeated episodes of alcohol , but absence of the balance the effect of the alcohol and administration and withdrawal. drug exposes these same homeostatic restore neurochemical equilibrium. In Himmelsbach’s idea has provided a mechanisms and to the with- other words, the brain develops toler- much-needed framework for investiga- drawal syndrome. Clinically, this ance for the drug. However, the brain tions of the neurochemical adaptations theory provides the framework for the is also in a state of physiological depend- that occur in alcohol dependence and modern category of “physiological ence at this point: When alcohol expo- the way in which these adaptations dependence” and its current diagnos- sure ceases, the new neuroadaptation may precipitate withdrawal. Although criteria—the presence of tolerance now unbalances the brain’s neuro- the adoption of this framework has and/or a specific withdrawal syndrome . And, like tilting the seesaw been largely beneficial, it also has (American Psychological Association to the opposite side, the unbalanced been restrictive in many ways, not 1994). Himmelsbach’s concept has adaptation produces a functional least because it has been interpreted dominated almost all neurochemical effect on the brain (i.e., the with- to imply that all the elements involved explanations for withdrawal from drawal syndrome) that is opposite in tolerance, dependence, and with- alcohol and other since it was from the effects originally produced drawal should be found within a first proposed. by alcohol. Thus, in Himmelsbach’s single neurochemical system. For At the neurochemical level, Him- scheme, the cause of both tolerance example, if alcohol is found to inhibit melsbach’s idea can be expressed as a and the withdrawal syndrome is the (i.e., reduce the activity of) a particu-

Administration of alcohol Alcohol-free Acute state alcohol Alcohol effect

1 CNS Opposing equilibrium neuroadaptation

Adaptation Alcohol Adaptation Recovery from Alcohol neuroadaptation tolerance

Withdrawal Removal syndrome of alcohol

Figure 1 Pictorial representation of the Himmelsbach hypothesis as it applies to alcohol use. The balanced seesaw on the upper left side of the cycle represents brain neurochemistry in an alcohol-free state (i.e., before the brain has been exposed to alcohol). Consuming alcohol initially unbalances brain chemistry to produce the acute effects associated with alcohol use (e.g., and incoordination). The brain then responds to this disruption by inducing an opposing chemical adaptation that tends to restore the neurochemical balance. At this stage, the effects of a given dose of alcohol are diminished (i.e., tolerance exists). If alcohol is removed, the adaptation is exposed, unbalancing the brain’s neurochemistry in the opposite direction. The result is a withdrawal syndrome that includes (e.g., agitation and ) that are opposite to alcohol’s initial effects. These disturbances will continue until the adaptation can be removed from the brain (or until alcohol is consumed again), restoring equilibrium.

1 CNS = central nevous system.

14 Alcohol Health & Research World Neurochemical Mechanisms Underlying Alcohol Withdrawal

lar chemical messenger (i.e., neuro- several ways. Short-term chemical Animal Models of Alcohol transmitter1) system in the brain, then modifications of proteins on Withdrawal the subsequent adaptation is predicted the target neurons2 can increase (or to increase activity in the same neuro- decrease) the neurotransmitter’s Meaningful research on neuroadapta- transmitter system. This increase in effects. To use an example from another tion underlying AW became possible neurochemical activity should coun- drug of dependence, the immediate only in the early 1970’s, when investiga- terbalance alcohol’s effects, causing effects of on nerves are limited tors developed models of physiological tolerance when alcohol is still present because “nicotinic” receptors for the dependence and withdrawal in labora- and a withdrawal syndrome when it transmitter become tory animals (for example, see Goldstein is removed. An alteration of this type desensitized (both to nicotine and and Pal 1971). Before this development, (i.e., within the same system) is acetylcholine). Desensitization of most preclinical research focused on referred to as a “homologous” adapta- those receptor proteins is caused by the alcohol “preference” of laboratory tion. One of the complexities of the short-lived modifications in their rodents that consumed alcohol volun- , however, is that an structure. Alternatively, for longer tarily. Because of the limited intake effect on one transmitter system may term changes, the system may increase of alcohol by most rodents and the be counterbalanced by an adaptation or decrease the number of receptor rapid elimination of alcohol from in a totally different system (i.e., a proteins on the surface of the target their bloodstreams, such models rarely heterologous adaptation) (Littleton neurons. Thus, when nicotine is pre- produced a recognizable physical and Little 1994). Such heterologous sent for a long period in the brain, the withdrawal syndrome. Consequently, adaptations are much more difficult number of nicotinic receptors for any neurochemical findings obtained to discern. However, it may be these acetylcholine is altered. This type of could not be related to withdrawal adaptations—rather than the simpler change in receptor numbers takes severity or duration. The key to suc- homologous adaptations—that are much longer to reverse when the drug cess in the development of research responsible for causing tolerance, is removed than do the short-term models for withdrawal proved to be dependence, and withdrawal. modifications. Similar changes occur the use of “forced” alcohol administra- To complicate matters further, in receptor numbers in response to tion. For example, some common homologous and heterologous adapta- alcohol, as discussed below. As a models allow animals access only to liquid diets containing alcohol, tions are not mutually exclusive, and result, researchers studying these whereas other models maintain con- both may occur in the same neurons. neurochemical phenomena at various tinuous intoxication by keeping Alternatively, each type of adaptation time points might obtain entirely may occur in different parts of the animals in an atmosphere containing different, although equally correct, alcohol vapor. brain in response to the same drug. explanations for the adaptations that In addition, neurochemical adapta- Those animal models clearly do accompany tolerance, whether to not reflect all aspects of human alcohol tions appear to build up at different alcohol or to other drugs. rates in a hierarchical process. For dependence. An obvious deficiency As might be expected, the com- is that the neurochemical effects of example, when alcohol inhibits a plexities involved in this process specific neurotransmitter system, the self-administration of alcohol, such have given rise to disagreements about as changes in forebrain re- system may first adapt by increasing which neurochemical changes are the amount of the chemical messen- lease (Weiss et al. 1996), are relatively most important in , ignored. However, the forced- ger released. If the presence of alcohol dependence, and withdrawal. Some persists, the system’s next adaptation administration models have provided consensus has been reached, however. valuable data on the neurochemical might be to amplify the effect of the This article examines postulated neurotransmitter on its targets (i.e., changes that accompany the induc- neuroadaptations in three systems: tion of physiological dependence on receptors) in the brain. The adaptation (1) the gamma-aminobutyric acid alcohol. In turn, many of the alter- of target cells can be accomplished in (GABA) neurotransmitter system; ations in brain chemistry provide (2) the glutamate neurotransmitter 1 fairly convincing explanations for the Nerve cells (or neurons) are the key players in the system; and (3) the calcium channel activity of the . In general, they physical changes that occur during signal to each other by releasing . system, which regulates various withdrawal (Samson and Harris 1992). For more information on the principles of nerve processes inside neurons. The focus cell communication and research into the neuro- on these three mechanisms is not chemistry of alcohol , please consult the intended to discount others that recent issue of Alcohol Health & Neuroadaptation in the Research World (Vol. 21, No. 2, 1997). may prove equally or more important GABA Neurotransmitter for AW. Even the limited focus on System 2Neurotransmitters produce their effects on the neurons receiving their message by combining these few systems allows several with specific receptor proteins found on the important points about clinical rele- When neurons in the brain release surfaces of these cells. vance to be made. GABA, this neurotransmitter delivers

Vol. 22, No. 1, 1998 15 its chemical message by combining with Criswell and colleagues (1995) sug- may become less responsive to the specific receptor proteins for GABA gested that GABAA receptor proteins effects of alcohol, a change that could on the surfaces of other neurons. GABA’s which contain the α1 subunit are the represent a potential mechanism for effect is inhibitory; that is, it reduces most susceptible to the acute effects adaptation within the GABA system. the electrical activity of the target neu- of alcohol. By reducing the number of Although most researchers consider α rons. When alcohol is introduced into receptors with an 1 subunit, neurons this alteration in GABAA receptors to this system, its immediate (i.e., acute) effect is to add to (i.e., potentiate) the inhibition caused by GABA, primarily on a particular receptor designated “GABAA” (see figure 2). Major inhibitory Consequently, neurons receiving mes- transmitter system sages through GABA are even more inhibited by this transmitter than usual when alcohol is present in the brain. GABA Because GABA neurons are widespread, alcohol probably potenti- ates inhibition of neuronal activity in Inhibits electrical signal several areas of the brain. The effect is VOCC Nerve cell body GABAA similar to that of the - Ca receptor Terminal tranquilizing drugs, which also poten- Axon VOCC tiate the action of GABA at GABAA Ca Neurotransmitter receptors, and helps explain why both VOCC Ca Nucleus Electrical signal alcohol and are effec- release Ca tive antianxiety agents (see figure 3). Because alcohol increases the VOCC effects of GABAA receptor activation, NMDA receptor the Himmelsbach concept of adapta- tion predicts that in the long-term Excites electrical signal presence of alcohol, neurons will respond by reducing the number of Glutamate GABAA receptor proteins on their surfaces. This reduction should “bal- ance” the acute effects of alcohol and Major excitatory produce the consequences (i.e., toler- transmitter system ance, dependence, and withdrawal) mentioned earlier. However, GABAA receptors in the brain appear to be altered in a much more subtle way in Figure 2 Schematic representation of some of the major neurochemical systems response to alcohol. Several smaller affected by alcohol. Nerve cells (i.e., neurons) convert chemical messages proteins make up the GABAA receptor received at the cell body (at left in this simplified ) into an electrical protein, and by shuffling the way in signal that is conducted along the axon to the terminal (at right). At the which these subunits are arranged, terminal, the electrical signal is converted back into a chemical message neurons may be able to produce new (i.e., a neurotransmitter) that is released from the terminal and carries the information to the next neuron in the circuit. Alcohol increases (i.e., GABAA receptors that are affected very little by alcohol (Samson and Harris potentiates) the effects of the major inhibitory neurotransmitter in the 1992; Littleton and Little 1994). A brain, gamma-aminobutyric acid (GABA), at the GABAA receptor. GABA’s effects tend to inhibit electrical signaling through the neuron. Alcohol common research finding is that one further decreases electrical activity by inhibiting the major excitatory of the small proteins, designated the α ), neurotransmitter, glutamate, particularly at a glutamate-receptor protein alpha 1 subunit (or 1 decreases in called the N-methyl-D-aspartate (NMDA) receptor. By inhibiting glutamate number (i.e., is downregulated) after at the NMDA receptor, alcohol slows the flow of calcium (Ca) into cells. chronic exposure to alcohol, although Regulation of the cell’s calcium balance is essential for normal cell even this reduction does not occur in function. In addition to its effects at the NMDA receptor, alcohol can alter all areas of the brain (Matthews et al. the flow of calcium through voltage-operated calcium channels (VOCC’s) 1998). The downregulation of α1 at the cell body as well as at the terminal, where calcium is necessary for subunits may be in keeping with the neurotransmitter release. Himmelsbach concept of adaptation.

16 Alcohol Health & Research World Neurochemical Mechanisms Underlying Alcohol Withdrawal

be involved in tolerance, it is not clear Because acute alcohol administra- ter system validate the Himmelsbach whether the change plays an important tion inhibits NMDA receptors, the scheme as it applies to the neuro- role in AW. To have a role in with- Himmelsbach concept predicts that chemistry of alcohol tolerance and drawal, the adaptive mechanism (i.e., neurons will compensate by producing withdrawal: Initially, alcohol inhibits the reorganization in GABAA receptor more NMDA receptors. Several studies the actions of glutamate on the subunits) must cause a change in have indeed reported an increase (i.e., NMDA receptors; adaptation involves GABAA receptor function when alco- upregulation) in the number of NMDA a straightforward homologous upreg- hol is no longer present. This implies receptors in the brain in response to ulation in the number and/or a reduction in the effects of GABA alcohol (Samson and Harris 1992). function of receptors; and withdrawal itself on the function of the receptor. Researchers theorize that neurons of alcohol results in hyperactivity of Investigators disagree over whether accomplish receptor upregulation by the glutamate system, at least partially this is the case. For example, using rats, somehow “sensing” that their NMDA from NMDA receptor upregulation. Whittington and colleagues (1995) receptors are no longer being activated It remains to be seen whether the found no evidence for underactivity to the normal extent when alcohol is picture will stay so clear when more is of GABAA receptors in an area of the continuously present in the brain. learned about other adaptations that brain associated with AW (i.e., the hip- This message is sent to the nucleus of occur in the glutamate neurotransmit- pocampus). Similar experiments using the neuron, where it causes the ter system. a different regime of alcohol adminis- specifically responsible for producing tration, however, produced clear evi- the NMDA receptor proteins to dence in favor of reduced sensitivity of increase their activity (i.e., their Neuroadaptation in GABAA in the same brain area (Kang et expression). Enhanced expression Calcium Channel Proteins al. 1996). Regardless of these disparate results in the production of more findings, the alteration in GABAA proteins and gradually leads to an Each time a neuron is electrically receptors is a potential cause of several increase in the number of NMDA activated, calcium flows into the cell aspects of the AW syndrome, includ- receptors on the surface of the affected through proteins called calcium chan- ing anxiety and seizures (see figure 3). neurons. As with many other aspects nels, which are located on the outer Despite scientists’ uncertainty over of neurochemistry in AW, however, surface of the nerve. Because these the role of GABAA receptors in AW these events probably are not quite as channels are opened by electrical syndrome, clinicians commonly use simple as described here. Like the activity, they are known as voltage- drugs that act on GABAA receptors GABAA receptors, the NMDA receptors operated calcium channels (VOCC’s). (e.g., the benzodiazepine ) in the brain are made up of smaller Other channels (i.e., receptor- to suppress the signs of AW (Romach protein subunits. It recently has been operated calcium channels, or ROCC’s) and Sellers 1991). This “detoxification” suggested that some of these subunits admit calcium in response to neuro- treatment usually succeeds in allowing increase in number because of an transmitters attaching to receptors. someone who is physiologically depen- increase in gene expression, whereas The NMDA receptor is an example dent on alcohol to avoid seizures and other subunits may increase for a of this mechanism. When glutamate other AW symptoms. Evidence suggests different reason (Hu et al. 1996). combines with the NMDA receptor, that less obvious harmful effects of The upregulation process (producing a channel through the protein opens withdrawal might still occur, however, new receptors and transporting them and allows calcium to enter the neu- and alternative approaches to treatment to the cell surface) takes several days, ron. When this entry of calcium might reduce this possibility. a time course that correlates well through VOCC’s and ROCC’s occurs with both the development of alcohol in the “body” of the neuron (see fig- tolerance and the emergence of a sus- ure 2), the calcium entering the nerve Neuroadaptation in the ceptibility to withdrawal upon alcohol regulates many important processes, Glutamate Transmitter cessation. The alcohol-induced probably including the expression of System increase in NMDA receptors also the genes that produce the NMDA persists long enough after alcohol is receptor, GABAA receptor proteins, As the most common excitatory neuro- removed to account for the duration and the proteins that form VOCC’s. transmitter in the human brain, glu- of withdrawal. This increase in the At the other end of the nerve cell (i.e., tamate increases the electrical activity of actions of the excitatory transmitter the terminal), the entry of calcium neurons. This neuronal activation occurs glutamate resulting from the increas- causes the neuron to release through several different types of receptor ed number of NMDA receptors is a stored neurotransmitters. Because proteins. Although alcohol inhibits many potential cause of several manifestations is the main func- of these receptors, one specific type, the N- of the AW syndrome, such as halluci- tion of all neurons (i.e., it is the way methyl-D-aspartate (i.e., NMDA) recep- nations and seizures. in which they communicate), any- tor,3 appears to be most sensitive to alco- In many ways the alcohol-induced thing that interferes with calcium hol’s effects (Samson and Harris 1992). alterations in the glutamate transmit- channel function has important and

Vol. 22, No. 1, 1998 17 far-reaching effects. Alcohol reduces resulting from an increase in VOCC’s explain why neurochemical researchers the flow of calcium into nerves (Littleton and Little 1994). Other have concentrated on the slower types through NMDA receptors and, prob- withdrawal syndrome changes may of adaptations (e.g., alterations in ably, other ROCC’s as well (Lovinger be explained specifically by neural receptor proteins and calcium channels) 1997). It also reduces calcium entry pathways in which excitability is con- in the brain: This type of change is through VOCC’s, but in this case the trolled by either NMDA receptors or the easiest to measure. mechanism is less clear. GABAA receptors (Samson and Harris Most neurochemical experiments At initial administration, alcohol’s 1992) (see figure 3). in this area involve postmortem mea- direct effects on VOCC’s are probably surements of changes in the of small at the alcohol concentrations animals at various times during alcohol typically achieved in the human brain Limitations of Studies administration and withdrawal. Semi- (Samson and Harris 1992). However, of Withdrawal permanent changes in protein numbers alcohol also has indirect effects on these (e.g., an increase in glutamate recep- calcium channels through its ability The changes discussed previously that tors) are much easier to measure under to potentiate GABA neurotransmission occur in the brain during long-term these conditions than are more ephem- and inhibit glutamate neurotransmis- alcohol administration are fairly con- eral changes (e.g., an increase in glu- sion, thereby reducing the electrical vincing neurochemical explanations tamate transmitter release). The latter activity of nerves. The continued pres- for the mechanisms involved in AW. changes are not necessarily any less ence of alcohol eventually leads to an However, there are certain limitations important as a mechanism for with- increase in the numbers of VOCC’s, to these studies that should not be drawal, however. For example, during presumably representing an adaptation ignored. In particular, both the methods AW, dopamine release may be designed to compensate for alcohol’s used to evaluate the neurochemical reduced in the area of the brain (i.e., inhibitory effects. This upregulation changes themselves and the models in the ) in which perhaps results from increased gene used to mimic clinical alcohol depen- this transmitter is increased by alcohol expression (Littleton and Little 1994). dence may affect the types of results acutely (Weiss et al. 1996). Although Likewise, the increase in the number obtained. These limitations do not short-lived changes in transmitter release of VOCC’s probably follows a time invalidate the conclusions, but they do such as this are difficult to measure, course similar to the alcohol-induced suggest that other mechanisms of with- the , anxiety, and emotional upregulation in NMDA receptors, both drawal should be considered as well. discomfort (i.e., ) that may during the development of tolerance result from the reduction in dopamine and during withdrawal. Types of Neuroadaptation are an important psychological aspect of The increase in calcium channels, the withdrawal syndrome (see figure 3). whether through upregulation of There are several similarities in the ROCC’s (like the NMDA receptor) or types of changes in GABAA receptors, Types of Models VOCC’s, follows the pattern predicted NMDA receptors, and calcium channels by the Himmelsbach scheme. Although that were suggested above to underlie Another practical limitation of the current understanding is probably an AW. All of these changes represent results discussed applies to the type oversimplification of these changes (e.g., modifications in the synthesis of nerve of model of dependence most com- several different types of VOCC’s exist cell proteins, which in turn results in monly used. Most researchers strive that may be regulated in quite different a slowly developing alteration in either to induce physiological dependence in ways), adaptation in calcium channels, the composition or the number of these laboratory rodents as rapidly as possi- like the changes in receptor proteins, proteins in the brain. Slow adaptations ble, using a single continuous period is one of the established mechanisms like these also are reversed relatively of alcohol administration (24 hours for alcohol tolerance and withdrawal slowly, a phenomenon that accords per day for several days) followed by (Samson and Harris 1992; Littleton with the Himmelsbach explanation its removal (e.g., Goldstein and Pal and Little 1994). of withdrawal (Littleton and Little 1971). This single cycle of dependence Thus, during the induction of alco- 1994). The Himmelsbach explanation is the quickest way to produce a with- hol tolerance, increased numbers of predicts that the functional imbalance drawal syndrome. The drawback of VOCC’s and NMDA receptors develop in the brain which causes the with- this simple model is that the continu- on nerves in the brain, and changes drawal syndrome should last only as ous administration/single-cycle model occur in the GABAA receptors. These long as it takes for the adaptive mech- does not reflect the usual pattern of adaptive changes persist during alco- anism to be removed (see figure 1). alcohol self-administration in humans, hol withdrawal and are believed to Adaptations that are reversed quickly which is much more episodic, with contribute to the withdrawal syndrome. cannot explain withdrawal syndromes periods of voluntary abstinence (Ball- Some aspects of the withdrawal syn- of long duration, such as that from enger and Post 1978). drome probably are caused by a gen- alcohol. Another, more pragmatic, This is important because different eralized increase in neuron excitability explanation may exist, however, to patterns of alcohol administration may

18 Alcohol Health & Research World Neurochemical Mechanisms Underlying Alcohol Withdrawal

A

Relaxation Intoxication Modulation of effects of Intoxication glutamate and GABA Anesthesia

Potentiation of GABA at Inhibition of glutamate at Inhibition of calcium flow Increase in DA release GABAA receptors NMDA receptors through VOCC’s

Major inhibitory Major excitatory Major regulatory Major “reward” pathway neurotransmitter neurotransmitter mechanism

Administration of alcohol Acute Alcohol-free alcohol state Alcohol effect

CNS 1 Opposing equilibrium neuroadaptation

Adaptation Alcohol Adaptation Recovery from Alcohol neuroadaptation tolerance Withdrawal Removal syndrome of alcohol

Major inhibitory Major excitatory Major regulatory Major “reward” neurotransmittter neurotransmitter mechanism pathway

B

Altered GABA receptor Increased numbers of Increased numbers Reduced DA release function NMDA receptors of VOCC’s

Potentiation of effects of Tolerance to Dysphoria glutamate and GABA benzodiazepines Confusion transmitters Dysphoria Anxiety Autonomic effects Seizures Seizures Seizures

Figure 3 Effects of alcohol and possible relevance to dependence and withdrawal. (A) The major acute actions of alcohol and their possible relation to the behavioral consequences of drinking alcohol are shown in relation to the Himmelsbach

model for dependence. The potentiation of gamma-aminobutyric acid (GABA) at GABAA receptors, inhibition of glutamate at N-methyl-D-aspartate (NMDA) receptors, and inhibition of voltage-operated calcium channels (VOCC’s) may underlie the relaxation, intoxication, anesthesia, and amnesia caused by alcohol. Alcohol also increases release of the neurotransmitter dopamine (DA) in a specific area of the brain, the nucleus accumbens. This action of alcohol is not very well understood but may play an important role in the rewarding effects of drinking, such as euphoria. (B) Examples of the adaptive changes thought to oppose the acute effects of alcohol. The bottom panels show possible consequences of these adaptations during withdrawal. For example, the adaptive changes in GABA receptor proteins caused by alcohol may make benzodiazepine tranquilizers, which also act on GABA receptors, less effective (i.e., may produce tolerance). A reduction in DA release in the nucleus accumbens may accompany alcohol withdrawal and may contribute to depression, anxiety, and emotional discomfort (i.e., dysphoria), perhaps leading an alcoholic to resume his or her drinking.

1 CNS = central nevous system.

Vol. 22, No. 1, 1998 19 produce different types of adaptation scientists have not uncovered consistent Nevertheless, benzodiazepines and, consequently, different mechanisms alterations in the human brain. This continue to be used effectively and for withdrawal. For example, the vari- finding is not entirely surprising given safely in detoxification, the main ations in the effects on the GABAA the variability in human and the goal of which is to prevent the life- receptors in the , as previ- confounding incidence of brain damage threatening onset of major with- ously described, came from models in in many alcohol-dependent subjects. drawal signs. It is possible, however, which alcohol administration was either These limitations aside, the neuro- that other harmful sequelae of with- constant (Whittington et al. 1995) or chemical explanations for AW, as drawal exist that are not controlled by intermittent (Kang et al. 1996). This described previously, probably include current treatments. For example, ben- does not mean that withdrawal syn- some fundamental truths; for example, zodiazepines are known to suppress drome itself will appear to be different, hyperexcitability of nerves in the brain the convulsions that occur during AW because different mechanisms may occurs during AW, and alterations in in animals. It is unclear, however, if produce identical signs and symptoms. the GABA and glutamate transmitter this class of drugs suppresses all of the For example, if alcohol leads to an systems probably contribute to this increased electrical activity that occurs adaptive increase in glutamate recep- excitability (Samson and Harris 1992). in the brain at this time. This sup- tors in a nerve pathway in the brain, Similarly, any increase in the electrical pression is important because any transmission along this pathway will activity of nerves that occurs during remaining nerve excitation could increase in AW, in this case because severe withdrawal is inevitably associ- produce at least two harmful conse- the glutamate released as a transmitter ated with increased calcium entry into quences: (1) nerve damage and will be more effective (i.e., because of nerves through calcium channel proteins (2) changes in the brain that make an increased number of receptors for (whether these channels are increased subsequent episodes of withdrawal this transmitter). Alternatively, suppose or not). Even this “minimalist” view more severe. the number of glutamate receptors of the research findings has some clini- stays the same, but the nerves in the cal implications. Nerve Damage Associated With pathway become more excitable because Alcohol Withdrawal they have increased the number of calcium channels on their surface. Once Treatment Implications Alcohol use has been shown to cause again, transmission through the pathway brain damage, and at least some of will increase, although the mechanism A key reason for understanding the this damage may occur during periods involved is different. The outward neurochemical mechanisms underlying of AW rather than during periods of signs and symptoms (i.e., the behavior) the withdrawal syndrome is to improve alcohol administration (Lovinger 1993). associated with withdrawal would be methods of treatment. The following One mechanism for this damage is identical for both mechanisms. However, section examines areas in which that excessive amounts of calcium if potential treatments are to be tar- advances in neurochemistry may be ions enter the excited nerves, switching geted to a particular mechanism, it is particularly useful. on processes which cause the cells to necessary to first determine the true self-destruct. Several connections exist mechanisms underlying withdrawal. Detoxification between this type of damage and the Although the changes in receptor neurochemical changes associated and calcium channel proteins described Based on the neurochemical evidence with AW. During withdrawal, hyper- previously have been found in several described earlier, the signs and symp- excitation of nerves occurs because of animal models by different investigators, toms of AW clearly are not simply increased activity in the glutamate this is not always the case. In fact, the result of a reduced stimulation transmitter system. This hyperexcitation some models that indubitably produce of GABAA receptors. Despite this fact, is further exacerbated by an increase physiological dependence and a recog- the most common pharmacological in the number of NMDA receptors. nizable withdrawal syndrome never treatment given during withdrawal These receptors, which are receptor- show the expected alterations in recep- continues to be a benzodiazepine or operated calcium channel proteins, tors and channel proteins. An example some other that acts to enhance allow calcium ions to readily enter the would be the difference in the relative activation of GABAA receptors (Romach nerve cell. Even more calcium ions importance of GABAA receptors for and Sellers 1991). In addition, the enter the excited nerves through the withdrawal-induced excitation in the changes in GABAA receptor proteins increased numbers of voltage-operated hippocampus (Whittington et al. 1995; believed to confer alcohol tolerance calcium channels on the nerve surface. Kang et. al. 1996). The acid test for may, in fact, reduce the effects of Most likely, all of these mechanisms all of these mechanisms will be whether benzodiazepines on these receptors work together to make nerves vulnerable they are found to occur in humans. (see figure 3), thus explaining why to the kind of damage resulting from In general, the human results have not relatively high doses of these drugs overexcitation (i.e., ) that been convincing. Despite repeated often are needed to suppress the with- may occur during AW (Lovinger 1993). postmortem studies in alcoholics, drawal syndrome. More extensive research is necessary,

20 Alcohol Health & Research World Neurochemical Mechanisms Underlying Alcohol Withdrawal

however, to determine the exact link The Severity of Repeated Alcohol other symptoms. This similarity does between neuronal excitation during Withdrawal not mean that the mechanisms for the withdrawal and the possible develop- two phenomena are identical, however. ment of nerve cell damage. The continuous administration/single- Although several potential neuro- Although alterations in GABAA cycle models of alcohol dependence chemical mechanisms exist both for receptors may play a part in the may miss an important aspect of alcohol kindling and for the progressive increase increased excitation of nerves in with- dependence in humans: that is, repeated in seizures in withdrawal, none has yet drawal, it seems unlikely that drugs cycles of dependence and withdrawal been accepted universally. For example, acting on the GABA system, such as may have progressively worse conse- some of the mechanisms proposed to benzodiazepines, will be able to pre- quences. This outcome is not always explain kindling in withdrawal involve vent all the excitotoxic damage taking the case experimentally; it depends on changes in the GABA transmitter place in nerves during withdrawal. variables, such as the duration of alcohol system (Kang et al. 1996). However, This area is one in which findings administration and the length of time once again, it is unlikely that drugs from animal research could be imme- between withdrawal episodes. Perhaps which affect this system selectively diately applied to therapeutic inter- the best evidence that repeated with- will be able to completely prevent the vention. If experiments could show drawal becomes progressively worse is progressive increase in withdrawal that the addition of another type of that the incidence and severity of seizures severity. Although this idea has not yet drug (such as an NMDA receptor antag- increases during repeated withdrawal, been proven, evidence does suggest that onist or a calcium ) both in animal experiments and in current methods of detoxification have increased the of current little effect on the progression of with- detoxification treatments against drawal severity (Becker and Littleton withdrawal-induced brain damage, Some brain 1996). the findings would be highly relevant Clearly, important questions remain to clinical practice. As an example, damage occurs as to the best AW treatment, and the recent experiments by Corso and use of animal models of repeated with- colleagues (1998) have shown that during periods drawal could be of immediate clinical when rats are subjected to intermittent of alcohol relevance to address these questions. exposure to alcohol (with frequent Again, if it could be shown that ben- minor episodes of withdrawal), damage withdrawal. zodiazepines do not fully suppress the occurs in nerves within the hippocam- mechanisms responsible for kindling pus. The researchers administered of withdrawal severity but that an calcium channel blockers to the ani- additional treatment during detoxifi- mals to stop the influx of calcium humans (Becker and Littleton 1996). cation interrupts this progression, the into nerve cells and, in theory, the It also has been suggested that other finding would have an immediate resulting cell damage. Surprisingly, aspects of withdrawal, such as anxiety, impact on clinical practice. Excitotoxicity although the drugs were effective also increase progressively. This is often (the putative mechanism for nerve against damage in some parts of the called kindling of withdrawal severity damage in withdrawal) also may be hippocampus, they were ineffective (Ballenger and Post 1978). The term subject to kindling by repeated episodes and even harmful in other parts. The originated from research on , of withdrawal (Becker and Littleton contrasting results emphasize the impor- in which it was observed that small 1996). Other mechanisms of nerve tance of mimicking clinical situations electrical stimulations (i.e., excitation) cell damage that may occur during as closely as possible. Conclusions in some areas of the brain have a repeated withdrawal (Corso et al. 1998) drawn from simple cellular models progressively greater effect when they also are areas for investigation of ther- might suggest that withdrawal excitation are repeated. Thus, a stimulation that apeutic interventions. is always directly equated with excito- at first produces no observable effect As evidenced by the amount of space toxic nerve damage in withdrawal; may produce sufficient activation of in this review dedicated to repeated such conclusions, however, would be the brain to cause convulsions when withdrawal, a return to drinking after an oversimplification. In addition, it has been repeated, for example, 20 detoxification (i.e., a relapse) is a com- accumulating evidence shows that times at daily intervals. The small mon event. Relapse itself is linked to intermittent alcohol administration stimulations “kindled” an epilepsylike some aspect of withdrawal (see sidebar, and repeated withdrawal can produce . pp. 22–23), but once relapse has occurred different neurochemical effects, as Thus, at least as far as the occurrence in an abstinent patient, the possibility well as more intense signs of with- of seizures is concerned, a similarity of “reinstatement” of dependence must drawal, from those seen in the single- appears to exist between kindling and be considered. It is worth considering cycle models most often used in the repetitive brain excitation that whether the neurochemical mechanisms research. This approach is discussed occurs with repeated episodes of AW behind reinstatement of tolerance, in the following section. and which may to seizures and dependence, and withdrawal are the

Vol. 22, No. 1, 1998 21 same mechanisms that operate during these mechanisms is discussed here, again, they can be recycled back to the the first episode of alcohol exposure. and it illustrates the deficiencies in nerve cell surface. This recycling may current understanding. be what takes place when a patient Reinstatement The proteins (e.g., NMDA receptors to drinking soon after detoxi- and calcium channels) that increase in fication. The adaptations necessary to Once relapse occurs, the general clini- number during the induction of oppose alcohol’s effects already are pre- cal consensus is that physiological dependence rapidly decrease during sent in nerve cells in the brain; the pro- dependence is reinstated rapidly withdrawal. The speed with which teins need only be recycled (or reassem- (Ballenger and Post 1978). In other these potentially damaging proteins bled) to reinstate alcohol tolerance words, only a relatively short period disappear (Samson and Harris 1992) and dependence. Recycling may of drinking is necessary before toler- implies that they must be actively explain why the onset of physiological ance develops and the effects of with- removed from the nerve cell surface dependence is so much more rapid the drawal from alcohol again become during withdrawal. In similar conditions, second time around. It also may mean severe. There are several potential when proteins are removed they rarely that the mechanism for reinstatement neurochemical explanations as to why are immediately destroyed, however. of dependence differs subtly from the reinstatement should be so rapid, but Instead they are removed from the nerve mechanisms that have to date no real evidence for or against cell surface and sequestered inside the uncovered in the common single-cycle these explanations exists. Only one of cell. Then, if the proteins are required models of dependence and withdrawal.

Relapse Prevention

A key therapeutic challenge in treating alcohol dependence to those that occur during the early stages of alcohol with- lies in finding ways of preventing relapse. Yet because relapse drawal. These symptoms may be caused by the cue- is a highly complicated behavior to model, much of the induced initiation of neuroadaptation intended to oppose emphasis in neurochemical research has been on the alcohol. Based on past experience, the person knows that physical withdrawal syndrome. However, some of the these very unpleasant sensations can be relieved only by findings from withdrawal research may help address ques- drinking alcohol. If he or she does accept the , further tions about relapse. For example, the depression, anxiety, neuroadaptation will occur in the brain. To offset these and emotional discomfort (i.e., dysphoria) that occurs during alterations, the person may drink even more alcohol; withdrawal can occur during extended periods of absti- consequently, relapse will be well under way. nence and often is considered a precipitant of relapse Although similarities exist among the consequences (Connors et al. 1996). of cue-induced relapse and alcohol withdrawal, the adapta- These feelings appear to be triggered by exposure to tions responsible for each are unlikely to be identical. stimuli (i.e., cues) that previously were associated with Those adaptations thought to underlie alcohol withdrawal drinking alcohol (e.g., the smell of alcohol, a bar environ- involve chemical alterations, such as increases in receptor ment, or being offered an ). The cues protein production, which are slow to develop, usually trigger adaptive changes in the brain designed to oppose requiring several days. To precipitate relapse, however, the effects of alcohol. If the patient does not immediately conditioned adaptation must be rapid, beginning within drink alcohol, then these adaptations “overbalance” the minutes of cue exposure. Alterations in electrical activity neurochemistry of the brain in much the same way as in specific circuits in the brain are a more likely explanation withdrawal from alcohol affects the dependent brain for such rapid changes. (for a review, see Connors et al. 1996). As a result, the If similarities do exist between withdrawal and cue- patient experiences pseudo-withdrawal symptoms with- induced craving, they probably occur in the neuro- out ever taking a drink. Faced with these symptoms, the transmitters involved in the two processes (e.g., the patient then may relapse to alcohol consumption. glutamate and gamma-aminobutyric acid [GABA] As an example, suppose an alcohol-dependent person systems). Such similarities might be especially useful for who is trying to maintain abstinence is offered an alcoholic developing more effective treatments. For example, the drink. In the past, this cue always would have preceded new anticraving drug called —which has acceptance of the drink and the arrival of alcohol in the recently undergone several successful clinical trials for brain; consequently, the cue is strongly associated with the relapse prevention in Europe—inhibits the N-methyl- neuroadaptation to alcohol (see figure). In response, the D-aspartate (NMDA) receptor and has been found to person becomes highly dysphoric and may even start reduce some nonphysical aspects of alcohol withdrawal sweating and shaking, all of which are symptoms similar in animal models (Spanagel and Ziegelgansberger 1997).

22 Alcohol Health & Research World Neurochemical Mechanisms Underlying Alcohol Withdrawal

In other words, in the single-cycle statement is not intended as a criticism, from this field of neuroscience represent models, the neurochemical mecha- however; without a thorough under- one of the success stories in alcohol nisms are developed “from scratch,” standing of the simple, single-cycle research. The disappointing aspect of implying that the changes that occur models, none of this speculation on this research is that advances in under- in the production of receptor and mechanisms for reinstatement would standing have had relatively little impact channel proteins take place slowly. In be possible. Nevertheless, now may be on clinical treatment of alcohol depend- reinstatement the same end results are the time to reevaluate the models of ence or withdrawal. As stated earlier, achieved, but the mechanisms already dependence and withdrawal in relation to however, the situation is now chang- exist and thus come into play much the most important clinical objectives. ing. As basic science begins to address more rapidly. Treatments aimed at slow questions more directly relevant to the changes (in gene expression, for exam- major clinical problems, findings in ple), which would be effective against Conclusion neurochemistry and neuropharmacol- the primary induction of dependence, ogy will be translated into advances might be useless against reinstatement Since the first introduction of appro- in treatment. The therapeutic problems of dependence. The mechanisms behind priate animal models for AW, our of minimizing long-term consequences the reinstatement of dependence are understanding of the neurochemical of detoxification, the prevention of important therapeutic targets but they mechanisms involved in this process relapse, and the delay of reinstatement have not been extensively studied. This has increased dramatically. Findings are all examples that will benefit from

Acamprosate’s actions also may prove useful for inhibit- References ing the cue-induced withdrawallike symptoms that lead CONNORS, G.J.; MAISTO, S.A.; AND DONOVAN, D.M. Conceptualizations to relapse. Further research is needed to determine the of relapse: A summary of psychological and psychobiological models. exact mechanism underlying acamprosate’s effects. Still, Addiction 91(Suppl.):S5–S13, 1996. this example represents yet another area in which basic SPANAGEL, R., AND ZIEGELGANSBERGER, W. Anti-craving compounds for research may prove highly useful for solving the major : New pharmacological tools to study addictive processes. Trends clinical problem of relapse to drinking.—John Littleton in Pharmacological Science 18:54–59, 1997.

Setting and behavior are Alcohol-free state Alcohol neutral stimuli.

Alcohol Initial use of alcohol Setting and behavior are Acute effect neutral stimuli.

Adaptation Cues contribute to Alcohol Setting and behavior become repeated alcohol use conditioned and contribute

Adaptation Setting and behavior initiate Same setting and/or behavior unopposed adaptation, without consuming alcohol possibly resulting in craving.

The seesaw analogy for neuroadaptation applied to conditioning and craving. The top row shows brain neurochemistry in an alcohol-free state. At this time the environment in which alcohol is consumed (i.e., the setting) and the behavior involved in drinking are neutral stimuli; that is, they have no association with drinking and no special impact on the brain. The same is true of the initial use of alcohol (second row), in which the brain’s neurochemistry is unbalanced by alcohol’s actions. If the drug is taken repeatedly in the same setting and in the same way, however, these stimuli become capable of eliciting neuroadaptations similar to those elicited by alcohol itself (third row). Exposure to such conditioned stimuli in the absence of alcohol may thus lead to unopposed neuroadaptation, which potentially may be expressed as a craving for alcohol.

Vol. 22, No. 1, 1998 23 neurochemical study. The experiments on GABA-induced inhibition predicts LOVINGER, D.M. Alcohol and neurotransmitter required are not as straightforward as the interaction of ethanol with GABA on individ- gated ion channels: Past, present and future. ual neurons in several rat brain regions. Journal of Naunyn Schmiedeberg’s Archives of Pharmacology those that have taken place over the Pharmacology and Experimental Therapeutics 356:267–282, 1997. past 25 years, but eventually they should 273:526–536, 1995. have a greater clinical impact. ■ MATTHEWS, D.B.; DEVAUD, L.L.; FRITSCHY, GOLDSTEIN, D.M., AND PAL, N. Alcohol depen- J.M.; SIEGHART, W.; AND MORROW, A.L. dence produced in mice by inhalation of ethanol: Differential regulation of GABAA receptor gene Grading the withdrawal reaction. Science 177: 288– expression by ethanol in the rat hippocampus 290, 1971. References versus . Journal of Neurochemistry HIMMELSBACH, C.K. The morphine abstinence 70:1160–1166, 1998. American Psychiatric Association (APA). Diagnostic syndrome, its nature and treatment. Annals of and Statistical Manual of Mental Disorders, Fourth Internal Medicine 15:829–843, 1941. ROMACH, M.K., AND SELLERS, E.M. Management Edition. Washington, DC: APA, 1994. of the alcohol withdrawal syndrome. Annual HU, X.J.; FOLLESA, P.; AND TICKU, M.K. Chronic Review of Medicine 42:323–340, 1991. BALLENGER, J.C., AND POST, R.M. Kindling as a ethanol treatment produces a selective upregula- model for alcohol withdrawal syndromes. British tion of the NMDA receptor subunit gene expres- SAMSON, H.H., AND HARRIS, R.A. Neurobiology Journal of 133:1–14, 1978. sion in mammalian cultured cortical neurons. Brain of . Trends in Pharmacological Science 13:206–211, 1992. BECKER, H.C., AND LITTLETON, J.M. The alcohol Research and Molecular Brain Research 36:211– 218, 1996. withdrawal “kindling” phenomenon: Clinical and WEISS, F.; PARSONS, L.H.; SCHULTEIS, G.; experimental findings. : Clinical and KANG, M.; SPIGELMAN, I.; SAPP, D.W.; AND HYYTIA, P.; LORANG, M.T.; BLOOM, F.E.; AND Experimental Research 20:121A–124A, 1996. OLSEN, R.W. Persistent reduction of GABAA KOOB, G.F. Ethanol self-administration restores receptor-mediated inhibition in rat hippocampus CORSO, T.D.; MOSTAFA, H.M.; COLLINS, M.A.; withdrawal-associated deficiencies in accumbal after chronic intermittent ethanol treatment. AND NEAFSEY, E.J. Brain neuronal degeneration dopamine and 5-hydroxytryptamine release in Brain Research 709:221–228, 1996. caused by episodic in rats: dependent rats. Journal of Neuroscience Effects of , 6-dinitro-quinoxaline–2,3- LITTLETON, J.M., AND LITTLE, H.J. Current 16:3474–3485, 1996. dione and MK 801. Alcoholism: Clinical and concepts of the neurobiology of ethanol depen- WHITTINGTON, M.A.; LAMBERT, J.D.; AND Experimental Research 22:217–224, 1998. dence. Addiction 89:1397–1412, 1994. LITTLE, H.J. Increased NMDA receptor and CRISWELL, H.E.; SIMSON, P.E.; KNAPP, D.J.; LOVINGER, D.M. Excitotoxicity and alcohol- calcium channel activity underlying ethanol DEVAUD, L.L.; MCCOWN, T.J.; DUNCAN, G.E.; related brain damage. Alcoholism: Clinical and withdrawal hyperexcitability. Alcohol and MORROW, A.L; AND BREESE, G.R. Effect of Experimental Research 17:19–27, 1993. Alcoholism 30:105–114, 1995.

Does a “nightcap” promote restful or daytime drowsiness? Can research change the way clinicians treat ? he answers to these and other questions can be found in Alcohol Alert, the quarterly bulletin published by the National Institute on Alcohol Abuse and Alcoholism. Alcohol Alert provides timely information T on alcohol research and treatment. Each issue addresses a specific topic in alcohol research and summarizes critical findings in a brief, four-page, easy-to-read format. Our most recent issue — Alcohol and Sleep (No. 41) — describes how sleep patterns may be influenced by alcohol consumption, the potential health consequences that may result from disturbed sleep, and how abnormal sleep patterns may place people with alcoholism at greater risk for relapsing to drinking. Forthcoming issues include the following:

¥ Alcohol and the Liver (No. 42), which focuses on recent research on alcohol-induced liver disease, including prevention and treatment implications.

For a free subscription to Alcohol Alert, write to: National Institute on Alcohol Abuse and Alcoholism, Attn.: Alcohol Alert, Publications Distribution Center, P.O. Box 10686, Rockville, MD 20849Ð0686. Fax: (202) 842Ð0418. 24 Alcohol Alcohol Alert is available on the World Wide Web at http://www.niaaa.nih.gov