Tobacco Documents Online

I I I TH E N EU ROCH EM I CAL MECHANISMS UNDERLYING NICOTINE TOLERANCE AND DEPENDENCE I ,bepart David J. K. Balfour ment of Pharmacology and Clinical Pharmacology, University Me School, Ninewells Hospital, Dundee DD1 9SY, UK dical I Table of Contents 5.1 Introduction 121 I 2 5 Behavioural and Endocrinological Consequences of . Chronic Nicotine Administration 122 5.2.1 Locomotor activity 122 I 5.2.2 Models of nicotine dependence 123 5.3 Neurochemical Mechanisms 127 5.3.1 Studies on the role of central nicotinic receptors 127 I 5.3.2 Studies on the role of specific neurotransmitter systems in the responses to chronic nicotine 131 5.3.2.1 Noradrenaline 132 5.3.2.2 Dopamine turnover and release 134 I 5.3.2.3 The role of dopamine in behavioural responses to nicotine 136 5.3.2.4 5-Hydroxytryptamine 139 5.4 Concluding Comments 146 I 5.1 Introduction 1 I I I I I I It is now generally accepted that a majority of people who smoke tobacco do so in order to experience the psychopharmacological properties of the nicotine present in the smoke and that a significant proportion of habitual smokers become dependent upon nicotine (Balfour, 1984b). This conclusion is supported by the fact that nicotine, given in the form of Nicorette chewing gum or as a skin patch, is reported to alleviate the sy%pptoms associated with THE BIOLOGICAL BASES OF DRUG TOLERANCE Copyright © 1991 Academ c Press Limited AND DEPENDENCE Alinghts of reproduction in any form reserved ISBN 0-12-564250-4 CHAPTER 5

I I I I I I I I 1 I I I I I D.J.K. Balfour smoking withdrawal ( Fagerstrom, 1988; Abelin el al., 1989). The behavioural responses to nicotine are complex in nature and appear to be influenced bv a number of factors, including the dose, the test environment used for the measurements and the duration of the experimental trial (Clarke, 1987). Nevertheless, the predominant effect of chronic nicotine in most animals is psychostimulation and it is clear that there are many similarities between the responses to chronic nicotine and those seen in animals given other psychostimulants such as amphetamine. As a result it has been suggested that the mechanisms which underlie the development of nicotine dependence may be the same as those that mediate amphetamine dependence (Wise and Bozarth, 1987) although it is important to remember that the psychopharmacological properties of nicotine are by no means identical to those of amphetamine (for review see Balfour, 1984b ). In addition it is clear that many people become regular smokers without apparently developing a strong dependence upon nicotine and are able to quit the habit without much difficulty (Balfour, 1984b; Gilbert, 1979; Chapter 2). Thus, any hypothesis proposed to explain the role of nicotine in the tobacco-smoking habit must take account of this fact. This review will focus on the evidence which supports the hypothesis that nicotine exerts effects on pathways in the limbic system which control emotional responses to aversive environmental stimuli and that exposure to such stimuli enhances the development of nicotine dependence. 5.2 Behavioural and endocrinological consequences of chronic nicotine administration 5.2.1 Locomotor activity In experimental animals the acute administration of moderate or high degrees of nicotine often results in suppression of spontaneous locomotor activity (Morrison and Stephenson, 1972; Stolerman et al., 1973) although if a long trial is used a stimulation of activity may be observed towards the end of the trial (Clarke and Kumar, 1983). In contrast, if animals are pretreated with a number of injections of nicotine prior to the test day, they seem to becomf tolerant to the depressant effects of the drug and locomotor stimulatior becomes the predominant response observed (Morrison and Stephenson, 1972: Stolerman et al., 1973). In nicotine-tolerant rats locomotor activity is increased in a dose-dependent way up to a subcutaneous dose of 0.4 mg/ kg. If a higher dose (0.8 mg/kg s.c.) is used, the animals do not appear to become fully tolerant and the activity observed during the first 60 min after the injection. although often higher than that observed for saline-treated controls, is lower 122 I

I I I I I t I I I I I I I I I I I Mechanisms underlying nicotine tolerance and dependence than that recorded for rats treated chronically with 0.2 or 0.4 mg j kg (Clarke and Kumar, 1983). In some reviews (Gilbert, 1979) the suppressant effect of nicotine on locomotor activity has been equated to a`sedative' action of the drug. While this conclusion cannot be totally excluded, it seems more likely that the decrease in activity observed in rats treated acutely with moderate doses of nicotine or chronically with higher doses is associated with the mechanisms which evoke ataxia and prostration in these animals rather than a sedative action similar to that seen in animals given a`conventional' sedative drug. There is no evidence that experimental animals become tolerant to the stimulant action of nicotine (Clarke, 1987). 5.2.2 Models of nicotine dependence Addictive drugs are thought to exert two effects which contribute to the maintenance of drug-taking behaviour. Firstly they are said to be 'rewarding' to take and many addicted individuals appear to develop a`craving' to re-experience this potent reinforcing property of the drug. Wise and Bozarth (1987) have proposed that for most drugs of dependence, including particularly am-)hetamine and cocaine, this effect is so potent that it alone accounts for the desire to maintain the drug-taking behaviour. Secondly, however, there is evidence that the withdrawal of many, if not all, drugs of dependence can be associated with an unpleasant abstinence syndrome which addicts seek to avoid by maintaining their intake of drug. In the case of drugs, such as the opiates or the barbiturates, which cause physical dependence, the abstinence syndrome can be severe and even life-threatening and the use of drugs to ameliorate the symptoms has a clear place in the treatment of withdrawal. There is no evidence, however, that nicotine evokes a withdrawal effect of this type, and nicotine dependence appears to be predominantly psychological in nature (Balfour, 1984b, 1990a). Nevertheless, many habitual smokers experience significant and unpleasant withdrawal effects when they first stop smoking which can be ameliorated by giving nicotine in another form (Fagerstrom, 1988; West, 1988). Experimental studies designed to investigate the mechanisms which mediate the development of nicotine dependence have, therefore, examined both the ability of nicotine to act as a rewarding stimulus and the effects of nicotine withdrawal. The putative rewarding properties of nicotine have been examined, )rimarily, using two paradigms-self-administration and conditioned place )reference. When compared with many other drugs of dependence, nicotine s found to be a relatively weak substrate in drug self-administration schedules Lnd its effects in this type of experiment appear to be influenced to a significant legree by the conditions of the experiment (Balfour, 1984b; Clarke, 1987). 1 ,or example, Lang et al. (1977) reported that nicotine self-administration 123 I

I I I 1 i I I I I I I I I I I I D.J.K. Balfour could be demonstrated in rats providing the animals were maintained at a reduced body weight. They also found that nicotine self-administration was enhanced if the injections were paired with food delivery. A subsequent study from the same group found that it was not necessary to pair food delivery with the nicotine injections in order to demonstrate nicotine self-administration although reduced body weight did seem to be important (Smith and Lang, 1980). A more recent study by Cox et al. (1984) suggested that neither reduced body weight nor schedule reinforcement was essential for the demonstration of nicotine self-administration in rats although pressing rates for nicotine were always fairly slow when compared with those reported for rats trained to self-administer amphetamine or cocaine and the increased rates observed for the rats pressing for nicotine could be explained, in part, by the stimulant effects of the nicotine per se. Nicotine self-administration has also been demonstrated in squirrel monkeys (Goldberg et al., 1981) using a schedule in which a visual stimulus was occasionally paired with a nicotine injection. In another study, Hutchinson and Emley (1985 ) showed that, whereas non-stressed squirrel monkeys preferentially drank nicotine-free water, this preference was reversed if they were exposed to a stressful stimulus (tail shocks ). In a conditioned place preference study, drug administration is paired with a specific and distinguishable environment. If the drug is rewarding, when the animals are subsequently given a choice, they will opt to spend most time in the environment which was paired with the drug. Fudala el al. (1985) reported that the rewarding properties of nicotine could be demonstrated using this technique providing the nicotine injection was paired with the less-preferred compartment of the apparatus. The study reported by Clarke and Fibiger ( 1987), using a more standard procedure in which the initial preference was not used as a factor in the experimental design, failed to confirm the ability of nicotine to act as a reward in a place preference paradigm. More recently Iwamoto (1990), using a conditioned place preference paradigm in which the nicotine was given intracerebroventricularly, has reported that a place preference for nicotine can be demonstrated which is independent of the animal's initial preference for one of the compartments of the apparatus. It may be, therefore, that place preference studies with systemic nicotine may be compromised by aversive properties of the drug which are mediated peripherally. It is consistently reported that tobacco smoke can exert a`tranquillizing` effect and that the craving to smoke can be enhanced by exposure to aversive environments (Gilbert, 1979; Pomerleau and Pomerleau, 1987). A number of investigators, therefore, have sought to examine the putative anxiolytic properties of nicotine. Early studies failed to find any convincing evidence that nicotine did exert anxiolytic activity (Morrison, 1969; Morrison and Stephenson, 1970). More recently, however, Costall et al. ( 1989) have shown 124

I ~ ~ I I I I I I I 1 I a I Mechanisms underly ng nicot ne tolerance and dependence that nicotine does exert anxiolytic-like activity in mice tested in a two- compartment apparatus in which one compartment is made more aversive than the other. In contrast, Balfour el at. ( 1986b) failed to demonstrate anxiolytic-like activity in the elevated X-maze test for anxiety, a paradigm which is clearly able to detect the anxiolytic activity of the benzodiazepine anxiolytics (Pellow et al., 1985). A subsequent study examined the possibility that the anxiolytic properties of nicotine were only revealed following a longer period of treatment with the drug. The results of these experiments (Table 1) did appear to suggest that chronic nicotine was anxiolytic. Analysis of covariance, however, showed that the increase in open runway entries was not independent of the increase in enclosed runway entries and, therefore, that the increased proportion of total activity occurring in the open runways OE:TA ratio could not be taken as unequivocal evidence for an anxiolytic response to chronic nicotine (Pellow et al., 1983). Thus, if nicotine does, indeed, exert anxiolytic activity it seems reasonable to suggest that the neural systems which underlie the effect differ from those which mediate the responses to the benzodiazepines. There appear to be no reliable reports that the withdrawal of nicotine fr-)m animals treated chronically with the drug has any effects on spontaneous activity per se. Benwell and Balfour (1979), however, showed that the withdrawal of nicotine from unstressed rats resulted in a modest but significant Table 1 The effects of chronic nicotine and its withdrawal on the behavioural and plasma corticosterone responses to the elevated X-maze test for anxiety. Drug treatment Chronic -Test day Total activity Responses (OE:TA ratio) PC (µg/100 ml) Saline Saline - 18±2 0.23±0.08 18±2 Nicotine Nicotine 38 ± 5' 0.38 ± 0.03 24 ± 4 Nicotine Saline - 18 ± 3 0.23 ± 0.08 24 ± 4 The responses were.measured using male Sprague-Dawley rats pretreated with daily subcutaneous.injections of saline or nicotine (0.4 mg/kg) for 19 days prior to the test day (chronic treatment). The treatment on the test day was given 5 min before the animals were placed in the maze for 20 min. Plasma corticosterone (PC) was measured using blood samples collected at the end of the trial. The total activity was calculated by summing the total entries into all four arms of the maze; the OE:TA ratio was calculated by dividing the number of entries made into the open runways by the total activity. Significantly different when compared with rats treated on the test day with saline: ' P < 0.01. 125

I I I I I i 1 I I I I I I I I o.J K Balfour increase in the plasma corticosterone concentration. These data are consistent with the hypothesis that nicotine withdrawal is mildly anxiogenic in rats although other explanations for the effect were also proposed. Nevertheless, the drug discrimination study of Harris et al. ( 1986) also suggested that nicotine withdrawal was anxiogenic to the extent that it generalized to an anxiogenic stimulus ( pentylenetetrazol ) and that it could be attenuated by the prior administration of diazepam. The generalization, however, was not impressive and, as a result, the authors concluded that nicotine withdrawal was only weakly anxiogenic. Costall et al. (1989 ), using their two-compartment model of anxiety, were able to demonstrate that the withdrawal of nicotine from mice following a period of chronic treatment evoked changes in behaviour which were characteristic of an anxiogenic drug. In contrast, studies in our laboratory (Balfour, 1991 a) have failed to demonstrate an anxiogenic response to nicotine withdrawal in the rat using the elevated X-maze test for anxiety (Table 1). At this time it is not possible to provide a clear explanation for this discrepancy although it is, perhaps, worth noting that the tests described by some groups are not specific for nicotine withdrawal (Costall et al., 1990; Emmett-Oglesby et al., 1990). Similar behavioural responses were observed following the withdrawal of addictive drugs with totally different mechanisms of action and it is possible, therefore, that the tests may detect the novelty of the no-drug state rather than being a specific measure of nicotine withdrawal. If this is the case then it is, perhaps, also likely that the changes in plasma corticosterone reported by Benwell and Balfour ( 1979) may also be a measure of drug withdrawal which is not a true indicator of dependence. Operant conditioning schedules have also revealed behavioural changes which occur as a consequence of nicotine withdrawal. For example, Carroll el al. (1989 ) have reported that if nicotine is withdrawn from rats trained to lick for a sweetened glucose solution, the response is markedly suppressed. It was reinstated immediately the nicotine injections were restarted. In another study Corrigal et al. ( 1989) showed that rats can become tolerant to the suppressant effects of high nicotine doses ( 2 mg/kg per day s.c. ) on lever-press responding for food. When, after a 50-day treatment schedule, the nicotine was withdrawn, responding for the food reward was initially depressed, although after 3 days it had returned to control levels. In both cases the authors suggested that the behavioural deficits could be a measure of nicotine abstinence although in neither study was there evidence that the effects were sustained for many days or that the effects were not the result of state- dependent learning upon nicotine. In a much earlier study, Morrison ( 1974) reported that the withdrawal of nicotine from rats trained on a shock avoidance schedule under the influence of the drug caused disruption of avoidance performance to the extent that it became significantly worse than that observed for saline-treated controls. In addition she reported that the 126

I I I I I I I I I I I I I I Mechanisms underlying nicotine tolerance and dependence disruption was significantly less marked if the schedule was made less stressful by the introduction of a warning signal prior to the delivery of each shock or a feedback signal after each successful lever-pressing response. She concluded that, since the acute administration of nicotine to rats trained with saline had no effect on avoidance performance, the effects of nicotine withdrawal could not be attributed to state-dependent learning and that the behavioural deficit, therefore, reflected the development of nicotine dependence. In addition she suggested that the dependence appeared to develop more readily in more stressful test environments. These conclusions, however, must remain speculative until confirmed by other studies although recent experiments in this laboratory have confirmed the development of the behavioural deficit following nicotine withdrawal (Balfour, 1990b) and shown that it persists for at least 5 days (Balfour, unpublished observations). 5.3 Neurochemical mechanisms 5.3.1 Studies on the role of central nicotinic receptors In vitro radioligand binding studies have shown that [3H ] ( - )-nicotine binds reversibly, stereoselectively and with high affinity to membranes prepared from mammalian brain (Reavill et al., 1988; Romano and Goldstein, 1980). The nicotine binding site also has high affinity for acetylcholine ( ACh ) and for other componds which act as agonists at the nicotinic cholinoreceptor located on autonomic ganglia but little or no affinity for agonists at non-cholinergic receptors ( Martino-Barrows and Kellar, 1987; Clarke et al., 1985b; Romano and Goldstein, 1980; Benwell and Balfour, 1985). Thus, it is widely assumed that the binding site represents the ACh recognition site on a central nicotinic receptor (Clarke et al., 1985b; Wonnacott, 1987). This receptor appears to mediate a number of the behavioural responses to nicotine, including, particularly, its ability to act as a locomotor stimulant (Clarke and Kumar,, 1983) and a cue in drug discrimination paradigms (Reavill et al., 1988; Stolerman, 1988), to the extent that the potency of many, although not all, nicotonic agonists in these behavioural tests appears to correlate well with their affinity for the receptor (Reavill and Stolerman, 1990; Reavill et al., 1990). In addition, both the locomotor stimulant response to nicotine and its ability to act as a discriminative cue can be attenuated by the systemic administration of pempidine or mecamylamine, non-quaternary compounds which block the nicotinic receptors found on autonomic ganglia but which also penetrate into the central nervous system (Clarke and Kumar, 1983; Stolerman; 1988; Romano ee al., 1981). In contrast the systemic 127 I ~ U1 00 GD O ~ ~

I I I I I I I I I I I I I I 1 I D.J.K. Balfour administration of quaternary antagonists which do not readily cross the blood-brain barrier has little or no effect on the responses to nicotine. Although these observations provide clear support for the hypothesis that the responses are mediated by a central nicotinic receptor with characteristics similar to those of the receptor on autonomic ganglia, they do not correlate with the ability of the antagonists to displace [ 3H ](-)-nicotine in in vitro binding assays using membranes prepared from brain tissue. Indeed, when assayed in vitro, the central nervous system ( CNS ) nicotinic receptor is found to have little or no affinity for almost all nicotinic antagonists (Benwell and Balfour, 1985; Romano and Goldstein, 1980), the only exception being dihydro-p- erythroidine, whose affinity for the binding site is still approximately 50 times lower than that for (-)-nicotine (Reavill et al., 1988; Wonnacott, 1987). The reason for the marked discrepancy between the effects of nicotinic antagonists such as mecamylamine on the behavioural responses to nicotine thought to be mediated by the central nicotinic receptor with high affinity for nicotine and their apparent lack of affinity for the receptors in vitro remains to be established with certainty. One relatively simple explanation, proposed by Wonnacott (1987 ) and Reavill et al. ( 1988), is that these antagonists act noncompetitively at another site on the receptor complex. There is evidence that the nicotinic receptor in the CNS is an allosteric protein which appears to undergo a fairly rapid desensitization if it is exposed to agonist for a period of time (Wonnacott, 1987). Recently, Takayama et al. (1989 ) have suggested that radioligand binding studies using tritiated agonist as the ligand essentially measure binding to the desensitized form of the receptor, a conformation of the receptor complex which has a particularly high affinity for agonist. In their studies, in which they used [ 3H ] methylcarbamylcholine as the ligand, they found that mecamylamine did reduce agonist binding to the receptor, but that, in agreement with the hypothesis proposed by Wonnacott (1987 ) and by Reavill et al. (1988 ), the antagonist appeared to exert its effects allosterically rather than by competing for the agonist recognition site. Studies with experimental animals have shown that the chronic administra- tion of nicotine can result in an increase in the density of nicotinic receptors in the brain (Marks et al., 1983; Nordberg et al., 1985; Schwartz and Kellar, 1983). Studies with human tissue taken at postmortem also suggest that tobacco smoking is associated with an increase in the density of the receptors (Table 2). The mechanism which mediates the upregulation of the receptors remains to be clarified although the data currently available appear most consistent with the conclusion that the effect probably reflects an adaptation to the prolonged or repeated desensitization of the receptor complex by nicotine (Wonnacott, 1987). In the animal studies, upregulation of the receptors has been observed following both the chronic infusion of nicotine (Marks et al., 1983) and the repetitive subcutaneous administration of the drug, often at 128 I

I I Mechanisms underlying nicotine tolerance and dependence I I ~ I I I I Table 2 Effects of tobacco smoking on the density of nicotinic receptors in human brain. Receptor density (fmol/mg protein) Brain region Non-smokers Smokers Hippocampal formation 71 ± 8 144 ± 25" Hippocampal neocortex 144 + 13 310 ± 39" Gyrus rectus 68 ± 6 98 ± 8' Cerebellar cortex 84 + 7 140 + 17" Median raphe nuclei 75 ± 10 112 ± 10" Medulla oblongata 48 ± 6 50 ± 5 The density of nicotinic receptors.was measured in membrane fractions prepared from postmortem brain tissue taken from 18 non-smokers (7 male and 11 female) and 12 smokers (6 male and 6 female) using [3H](-)-nicotine as the radioligand. The results are expressed as means ± SEM. Significantly different when compared with the data for non-smokers: ' P < 0.05; " P < 0.01. The data are derived from the study reported by Benwell et al. (1988). relatively high doses (Marks et al., 1986a; Schwartz and Kellar, 1983, 1985). In contrast, the chronic administration of a centrally acting inhibitor of acetylcholinesterase activity has been shown to cause a decrease in the receptor density (Costa and Murphy, 1983; Schwartz and Kellar, 1985), which implies that prolonged exposure to the putative natural agonist for the receptor evokes down- rather than upregulation of the receptor. However, a recent study by De Sarno and Giacobini (1989) has shown that the repeated intra- cerebroventricular administration of the short-acting cholinesterase inhibitor physostigmine, over a period of 21 days, evokes an increase in the receptor density similar to that found in animals given chronic nicotine. These authors concluded that, for ACh at least, desensitization of the receptor complex was a response to repetitive rather than prolonged exposure to agonist and they suggested that desensitization occurred when 'activated' agonist-receptor complex was exposed to a further dose of agonist. Interestingly, although the plasma nicotine level of habitual smokers tends to increase steadily during the day, substantial peaks of plasma nicotine are superimposed on the basal levels following each cigarette and, possibly, following the inhalation of each puff of cigarette smoke (Russell, 1988). Thus, it seems likely that the brains of habitual smokers are exposed to nicotine in a way which, according to De Sarno and Giacobini (1989), is most likely to cause receptor desensitization and, thus, upregulation. It is, perhaps, not surprising, therefore, that the 129

I I I I I I I I I I I I I I I I I I D.J.K. Balfour increases in receptor dens'ity evoked by smoking tobacco appear to be amongst the highest reported (Benwell el al., 1988). The molecular mechanism underlying the change in receptor density evoked by chronic nicotine administration remains to be established with certainty. Since it is likely that the receptor is an allosteric protein and that the radioligand studies used to measure the receptor density defect, primarily, a conformation of the receptor with high affinity for agonist, it is possible that the apparent increase in receptor density reflects an increased prevalence of this conformation of the receptor. Indeed, Romanelli el al. (1988) have suggested that this is the case. This view, however, has not been accepted by others who believe that the increased binding of [3H] (-)-nicotine or [ 3H ] ACh to brain membranes prepared from chronically treated rats reflects an increase in the synthesis of receptor protein (Wonnacott, 1987). It is clearly tempting to suggest that the effects of chronic nicotine on the density of nicotinic receptors in the brain may be related to the development of tolerance to and ( or dependence upon the drug. Both Marks et al. ( 1983) and Ksir et al. (1985 ) have drawn attention to the fact that the development of tolerance to the depressant effects of nicotine on locomotor behaviour correlates reasonably closely with the increase in receptor density. The data reported in a more recent paper by Ksir el al. ( 1987) suggest that locomotor simulation to nicotine is only observed if the doses given are high enough to evoke an increase in receptor density, results which, the authors claimed, imply a possible causative relationship. Other studies (Benwell and Balfour, 1985; Marks ct al., 1986a; Collins el al., 1988), however, suggest that the development of tolerance to the depressant effects of nicotine on locomotor activity and the enhanced locomotor stimulant responses observed in animals treated chronically with the drug are unlikely to be associated with changes in the density of the receptors. Indeed, it would be surprising if this was the case since there is evidence that tolerance to the depressant effects of the drug persists for many weeks (Stolerman el al., 1973) whereas the receptor density returns to control levels within 21 days of cessation of treatment (Schwartz and Kellar, 1985). Another attractive possibility is that the increased density of receptors mediates the effects of nicotine withdrawal although, to date, this hypothesis has been the subject of very little direct investigation. If it is the case, then it is important to establish if the withdrawal effects are associated with increased responses to ACh mediated by an increased density of 'active' nicotinic receptors or diminished responses to the transmitter caused by a proliferation of inactive or desensitized receptors. One of the few studies to address this problem is that described by Lapchak el al. (1989). It showed that the release of ACh from slices of frontal cortex or hippocampus evoked by the addition of the nicotinic agonist methylcarbamylcholine was abolished 130