- Email: [email protected]
Nicotine-induced inhibition of cerebellar purkinje neurons: Specific actions of nicotine and selective blockade by mecamylamine
OO28-3908/89 $3.00+ 0.00
Neuropharmacology Vol. 28, No. 5, pp. 495-501, 1989 Printed in Great Britain
Pergamon Press plc
NICOTINE-INDUCED INHIBITION OF CEREBELLAR PURKINJE NEURONS: SPECIFIC ACTIONS OF NICOTINE AND SELECTIVE BLOCKADE BY MECAMYLAMINE R.
DE LA
GARZA,‘~~R. FREEDMAN’,~”and J. HOFFER~*
‘Veterans Administration Medical Center, Denver, Colorado 80262, U.S.A., *University of Colorado, Health Sciences Center, Department of Psychiatry C-268, Denver, Colorado, 80262, U.S.A. and ‘University of Colorado, Health Sciences Center, Department of Pharmacology C-236, Denver, Colorado 80262, U.S.A. (Received 17 August 1988) Summary-The specificity and pharmacological characteristics of the effects of local administration of nicotine on cerebellar Purkinje cells in the rat were examined electrophysiologically. Local application of nicotine, whether by pressure-ejection or by iontophoresis, depressed the spontaneous discharge of Purkinje neurons in a reversible and dose-dependent manner. This action could not be mimicked by local application of vehicle alone. The inhibitory effects of (-)-nicotine were several-fold more potent than that of the (+)-enantiomer. Systemic administration of the ganglion blocker mecamylamine reliably and reproducibly antagonized the nicotine-induced inhibitions of Purkinje cells whereas nicotine-induced excitation of interneurons was not altered. Local pressure-ejection of mecamylamine also antagonized the inhibitory actions of nicotine, administered by iontophoresis. Since the central effects of nicotine on behavior are stereospecific and sensitive to mecamylamine, the data in this study further support the hypothesis that the actions of nicotine on Purkinje neurons arc mediated by ganglionic-like receptors, These findings also suggest that the Purkinje cell may serve as a good cellular model for studies on central pharmacology of nicotine. Key words-nicotine,
mecamylamine, Purkinje neurons.
The reinforcing properties of nicotine are often proposed to mediate addiction to tobacco smoking (Clarke, 1987; Kumar and Lader, 1981). Indeed, several lines of evidence implicate the central actions of nicotine in tobacco smoking. For instance, the intake of nicotine in humans is directly related to the dose of nicotine in a cigarette (Gritz, 1980). Similarly, self-administration studies in animals have shown
that the administration of nicotine is reinforcing in rats (Hanson, Ivester and Morton, 1979) and monkeys (Spealman, 1983). Moreover, the nicotine ganglionic antagonist, mecamylamine blocks cigarette smoking in humans (Stolerman, Goldfarb, Fink and Jarvik, 1973) and the self-administration of nicotine in primates (Spealman, 1983). Although a recent study reported that the direct actions of nicotine on single units of the substantia nigra were blocked by mecamylamine (Clarke, Hommer, Pert and Skirboll, 1985), electrophysiological studies that directly assess the central cellular actions of nicotine-mecamylamine interactions are lacking. These studies are important because evidence from radioligand binding and behavioral studies indicate that the majority of nicotinic receptors in the CMS resemble those present in sympathetic ganglia
*To whom correspondence
should be addressed. 495
in which mecamylamine works as an antagonist of the actions of nicotine (Brown, 1979). Previous findings have shown that the actions of nicotine on cerebellar Purkinje cells were differentially sensitive to hexamethonium, a ganglionic blocker (de la Garza, Bickford-Wimer, Hoffer and Freedman, 1987a). In order to determine further whether Purkinje cells can serve as a cellular model for the ganglionic-like central actions of nicotine, in this study whether the administration of mecamylamine could also selectively block the inhibitory actions of nicotine on Purkinje cells was investigated. In addition, several other types of experiments were conducted to substantiate further the specificity of actions of the nicotine on Purkinje cells. First, local actions of nicotine on Purkinje cells, using iontophoresis, were compared with those using pressure-ejection to dissociate pressure or currentrelated artifacts from the pharmacological actions of nicotine. Second, actions of locally applied nicotine on Purkinje cells were compared with that of vehicle alone. Third, effects of local administration of the two stereoisomers of nicotine on the activity of the Purkinje cells were compared. Finally, nicotinemecamylamine interactions were examined on cerebellar interneurons, which appear to possess neuromuscular- rather than ganglionic-like putative receptors (de la Garza et al., 1987a; de la Garza, McGuire, Freedman and Hoffer, 1987b).
R. DE LA GARZA et al.
496 METHODS
Data from 32 male Sprague-Dawley rats (200-300g Sasco Omaha, New England) were suitable for analysis. The rats were anesthetized with urethane (1.5 g/kg i.p.) and prepared for the recording of neurons in lobules VI, VII, and VIII of the cerebellum, as previously described (Freedman, Hoffer, Woodward, 1975; Hoffer, Siggins and Bloom, 1971). Cells were identified by their characteristic spontaneous discharge patterns and by their evoked responses to surface stimulation of parallel fibres (de la Garza et al., 1987a). Action potentials from Purkinje cells and cerebellar inhibitory interneurons were recorded with the 3 M NaCl-filled barrel of a multi-barrelled micropipette. Spikes were separated from the background activity and converted to constant voltage pulses with a window discriminator. Pulses were also integrated over intervals of 1 set by a ratemeter and displayed on a strip chart recorder. Nicotine bitartrate (Sigma) was either dissolved in 0.1 M phosphate buffered saline (pH 7.0-7.4; 100 PM) or physiological saline (pH 6; 100 PM) and applied by pressure-ejection (Palmer, 1982). Nicotine bitartrate was also dissolved in distilled water (pH 4; 0.75 M) and aplied by microiontophoresis. The (+)-isomer of nicotine (kindly supplied by Dr Allan Collins, IBG) was also tested in some experiments (pH 7; 1OOpM) and applied by pressure-ejection. Mecamylamine HCl (Sigma) was dissolved in physiological saline and applied locally (pH 6; 10 FM) by pressure-ejection. In some experiments, mecamylamine HCl (0.12550.75 mg/kg) was also dissolved in physiological saline and administered intravenously. Drug-induced responses were quantified from the ratemeter records using a Tektronix graphics tablet and a Nova 3/12 computer. The digitized ratemeter information was transformed into percentages of drug-induced excitations or inhibitions of the firing rate of the cell by the computer. Because the resting discharge rate was sometimes altered in a nonspecific manner by application of the drug, this method for analyzing the data permitted the measuring of effects of drug independently of the resting discharge rate (Freedman et al., 1975). For each cell, the ejection pressure or iontophoretic current for nicotine, and the time of application were initially adjusted to elicit a 40 to 80% inhibition of Purkinje cell, or at least a 100% excitation of interneuron. Since the amount of drug ejected is linearly related to the product of time and ejection pressure or ionotophoretic current (Palmer, Wuerthele and Hoffer, 1980), this quantity can be utilized as an index of the dose of the drug. Drug-induced responses were considered to be blocked by mecamylamine if the inhibitory or excitatory effects of nicotine were reversibly reduced to at least 30% of the effects obtained in the absence of the antagonist. The multiple response to drug, obtained in the absence and presence of mecamylamine were averaged for each and represented the correlated observations obtained for that particular experiment.
After the intravenous injection of mecamylamine, an hour was allowed to pass in order to observe recovery of the local actions of nicotine. Several agonistic responses were obtained prior to administration of antagonist to ensure reproducibility of responses. In some experiments, dose-response curves for the isomers of nicotine were obtained and the dose required to elicit a 50% inhibition of the activity of Purkinje cell was calculated for each compound. Post hoc statistical analyses employed Student’s t-test for correlated samples to calculate the average doseresponse obtained between experiments and to establish the degree of significance of these responses to drug (Hayes, 1981). RESULTS
Studies of nicotine receptor specificity
Several control studies were conducted to investigate the pharmacological specificity of the actions of nicotine. In order to maintain pH at levels that would not disrupt cell activity, two vehicles were used in this study. Nicotine was either dissolved in physiological saline or in phosphate-buffered saline. Neither vehicle had physiological actions at doses that produced inhibition of Purkinje cell if nicotine was dissolved in the vehicle. In Figure 1 are shown two cells that were treated with nicotine or its vehicle, alternately, from separate barrels of the same micropipette assembly. In Panel A, a continuous ratemeter record shows that when nicotine was pressuredejected, an approximately 45% inhibition of activity was produced. When only the vehicle for the solution was tested, however, the same dose did not alter the activity of the Purkinje cell. Similarly, in the bottom panel of the same figure is shown a continuous ratemeter record of a second Purkinje cell. Nicotine produced an approximately 81% inhibition of the cell firing, but the phosphate buffer vehicle failed to produced any changes in rate when the same dose was used. A total of 10 cells failed to respond to either vehicle, at a dose that produced marked inhibition of the activity of Purkinje cells, using nicotine-containing solutions in an adjacent barrel of the same pipette assembly. The pharmacological specificity of the actions of nicotine was also studied by comparing the actions of isomers of nicotine. Barrels in the same pipette assembly were filled with 100pM of either (-) or (+)-nicotine and their actions were tested on the same Purkinje cell. Figure 2 top panel shows the ED,,s for inhibition of the discharge rate of Purkinje cells, the ED,,+ were obtained from individual dose-response curves for each isomer in 14 Purkinje cells. The inhibitory actions of the (-)-isomer were almost three times more potent than the (+)-isomer. The mean EDS, dose for (-)-nicotine was 9.4 f 2.5 (&SE) and the ED,, dose for (+)-nicotine was 25.1 f 5.4 (*SE). A t-test for correlated samples showed that the difference between these two doses
491
Mecamylamine blocks nicotine actions - NlCOTlNE
40 psi x set
PliYSlOLOGlCAL
l
SALlNE
- NICOTlNE
100 psi x set
.
PHOSPHATE BUFFEAEO SALME
25pulsesI set lmin Fig. I. Effects of nicotine and vehicle solutions on Purkinje neurons. Panels A and B shows ratemeter records of two Purkinje cells, treated with a fixed-dose of pressure-ejected nicotine or vehicle, respectively. In these two records, the solid bars above the traces indicate the time of application of nicotine and the solid circles indicate the time of application of vehicle. The numbers above the first bar indicates the dose, used in pounds per square inch time the duration of application of drug or vehicle (p.s.i. x set). Percentages above the ratemeter responses in this and succeeding figures represent calculated drug-induced percentage changes from baseline discharge rate. Note that the two different doses of nicotine produced an immediate reduction in discharge rate, but the same dose failed to disrupt the discharge rate of the Purkinje cell when vehicle alone was pressure-ejected.
was statistically significant [t(l3) = 2.86, P < 0.021. In the bottom two panels are representative ratemeter records of the same Purkinje cell, treated with both isomers of nicotine. Both enantiomers showed recoverable dose-dependent effects but smaller doses were required with (-)-nicotine, as seen in the middle panel. Studies of mecamylamine antagonism The pharmacological specificity of the actions of nicotine was also evaluated by investigating the actions of the ganglion blocker mecamylamine. Local pressure-ejection of mecamylamine produced variable results when interacted with pressure-ejected nicotine. Therefore, nicotine was delivered by microiontophoresis, in combination with pressure-ejected mecamylamine. Nicotine-induced inhibitions, produced by iontophoresis, were reliably and reversibly blocked by application of local mecamylamine in 5 of 6 cells. In Figure 3 is shown a representative nicotine-induced response before, during, and after pressure-ejected mecamylamine. Panels A and B show a continuous ratemeter record of a Purkinje cell. Nicotine initially produced an average 54% inhibition, but during the concurrent application of mecamylamine, the response to nicotine was markedly reduced. The average percentage nicotineinduced inhibition before mecamylamine, observed in 6 cells, was 53.8% +4.0 (X + SE) and during mecamylamine was 9.5% f 4.0 (X If: SE). A t-test for correlated samples showed that the reduction in percentage inhibition was highly significant [t(5) = 8.610, P < O.OOl].
Nicotine-induced inhibitions were also blocked by intravenously-administered mecamyamine in 10 of 14 Purkinje cells. Figure 4 shows representative data for two Purkinje cells. In panels A,-A, are shown a ratemeter record for a single Purkinje cell before, during and after administration of mecamylamine. Nicotine initially produced an average 85% inhibition, but during the application of the antagonist, the response to nicotine was reduced to about half the original response. An hour after the administration to mecamylamine, the response to the application of nicotine recovered. In panels B,-B, are shown similar responses for another Purkinje cell. In this cell, nicotine produced an initial 67% inhibition. The administration of mecamylamine completely antagonized the response to nicotine in a reversible manner, as seen in the last panel. The average percentage inhibition before the intravenous administration of mecamylamine, observed in 14 cells, was 69.1% + 3.59 (X k SE) and during mecamylamine was 17.73% + 6.84 (X + SE). A t-test for correlated samples showed that the reduction in the percentage inhibition was significant [t (13) = 9.007, P < O.OOl]. Intravenous administration of mecamylamine was also tested on cerebellar interneurons, where previous studies had shown nicotine-induced excitation to be mediated by neuromuscular-type receptors. In Figure 5, panels A,-A2 and B,-B2 show ratemeter records for two representative cerebellar interneurons, before and during the application of two doses of mecamylamine (see figure legend). The excitation, elicited by nicotine, of these two cells was not blocked by mecamylamine. In the entire interneuron sample of 9 cells, doses of mecamylamine that blocked
R. DE LA GARZA
et al.
lI
(-)
Nicotine
(T) Nicotine
17.6
f-J
30.1
psi x set
Nicotine 56.0
17.6
psi x set
25 puk.es/sec (tl
Nicotine
1 min
Fig. 2. The agonistic actions of the two stereoisomers of nicotine. The top bar graph shows the dose required to produce a 50% reduction on the discharge rate of Purkinje neurons with both isomers of nicotine. The lower two ratemeter records are from a single Purkinje neuron that was treated with increasing doses of (-)-nicotine (middle) or (+)-nicotine (bottom). See Figure I for identification of symbols. Note that (-)-nicotine was almost three times more potent than (+)-nicotine in inhibiting the activity of Purkinje cells. Also note that the dose of (-)-nicotine that produced approximately 40% inhibition in this cell failed to decrease the activity of the same cell when (+)-nicotine was pressure-ejected. Nicotine 100 nA / 10 set
Mecomylomine 2 psi
A
I
7.5 %
I 47.4 %
I 64.6 %
25 pulses/set 1 min
Fig. 3. Interactions between iontophoretically-applied nicotine and pressure-ejected mecamylamine on Purkinje neurons. Panels A and B show a continuous ratemeter record of a Purkinje neuron before, during and after the local pressure-ejection of mecamylamine. See Figure 1 for identification of symbols. The dashed line indicates the time application of mecamylamine. Note the blockade of the actions of nicotine during the application of mecamylamine and the subsequent recovery after the removal of this ganglionic antagonist.
Mecamylamine
blocks
50 spikes/ set
nicotine
499
actions
Bs
25spikes/ set
1 min
1 min
Fig. 4. Interactions between nicotine and parenterally-administered mecamylamine on Purkinje neurons. A,-A, and B,-B, show ratemeter records from two Purkinje cells before, during and after intravenous injection of mecamylamine. The solid bars above the traces indicate the time of application of nicotine and the number above the first indicates the dose of drug used in pounds per square inch times the duration of application of drug (PSI/set). Panels A, and B, show the response to nicotine 5 min after 0.25 mg,‘kg mecamylamine. Note that mecamylamine partially blocked the response to nicotine in A, and completely blocked nicotine-elicited inhibitions in B,.
inhibition of Purkinje cells did not block the excitatory actions of nicotine, but rather produced a modest enhancement in percentage excitation that was not significantly different from controls [t(8) = 1.46, P > 0.101. DISCUSSION
The present investigation provides further evidence that the actions of nicotine on Purkinje cells are pharmacologically specific because the vehicles used did not induce depressions of the activity of Purkinje cells at doses that produced marked inhibition if the solution contained nicotine. Moreover, the enantiomers of nicotine differed significantly in potency and reversible dose-dependent depressions of the
activity of Purkinje cell were elicited by local administration of nicotine, either by pressure ejection or by iontophoresis. The actions of nicotine were also shown to be pharmacologically specific because mecamylamine, whether applied locally or systemically blocked the actions of nicotine in a reproducible and reversible manner. These findings extend previous data that indicated that the actions of nicotine appeared to be mediated by two separate nicotinic sites: the inhibitory effects of nicotine through ganglioniclike (C,) receptors and the excitatory effects of nicotine through neuromuscular-like (C,,) receptors (de la Garza et al., 1987a). Since other studies have suggested that C, receptors mediate the reinforcing actions of nicotine in animals (Spealman, 1983) and humans (Stolerman et al., 1973) the present study 9 psi /set
8 psi/set 1
A,
I
I
291%
B,
I
204%
I
224 %
A2
17 spikes/ 5ec
1 min
1 min
Fig. 5. Interactions between nicotine and mecamylamine on cerebellar interneurons. A,-A, and B,-B, show ratemeter records from two interneurons before and during the intravenous injection of 0.25 mg/kg (A*) and 0.125 mg/kg (B2) of mecamylamine. Note that mecamylamine failed to block nicotine-induced excitations.
500
R. DE LA GARZA et al.
also suggests that the actions of nicotine on Purkinje neurons may be a valid cellular model for the study of the central actions of nicotine relevant to the pharmacological properties that may underline the reinforcing properties of nicotine. Because of the ability of mecamylamine to block some of the behavioral actions of nicotine with small doses and others with large doses, other studies have suggested that mecamylamine acts through at least two types of receptor (Collins, Evans, Miner and Marks, 1986). Although, it remains to be seen whether the site sensitive to large doses in their study, is also sensitive to neuromuscular blockers, like curare or a-bungarotoxin (a-BTX), the present data, together with behavioral work of Collins, seems to indicate that neuronal nicotinic receptors may not be homogeneous in their functional properties. This issue is particularly important because neuronal receptors are traditional identified as of the C, type (Brown, 1979). The present work, however, proposes that subtypes of functional neuronal receptors are more adequate hypotheses to account for the central actions of nicotine. Further studies should investigate the extent to which both the electrophysiological and behavioral actions of nicotine may be dependent on C, or C,, receptors. Local application of mecamylamine produced inconsistent results when studied with pressure-ejected nicotine. However, when nicotine was administered by iontophoresis, mecamylamine produced inconsistent results when studied with pressure-ejected nicotine. However, when nicotine was administered by iontophoresis, mecamylamine produced a marked and reversible blockade of the actions of nicotine. The difference between these findings may be related to the release characteristics of drugs, administered by pressure-ejection compared to iontophoresis. The pattern of distribution of drug is more localized when drugs are delivered by iontophoresis than when the drugs are administered by pressure-ejection. Since mecamylamine is a highly lipophihc agent, which rapidly diffuses into white matter, it is possible that it does not reach the more distal sites of the Purkinje cell membrane where pressure-ejected nicotine could act. The iontophoresis or pressure-ejection of drugs each have their own set of potential artifacts (Hoffer et al., 1971; Palmer et al., 1980). However, since in this study both the pressure ejection and iontophoresis of nicotine produced similar inhibitory effects it is suggested that the actions of nicotine were not related to drug-administered artifacts. The difference in potency between the two stereoisomers of nicotine is consistent with other data that also report (-)-nicotine to be more potent than (+)-nicotine. However, in some studies, the (-) form of nicotine is 20 to 30 times, more potent than the (+) form in inhibiting the binding of /-[-‘HInicotine and 4 times more potent inhibiting the binding of a-[‘251]-bungarotoxin to tissue from the rodent brain
(Marks, Stitzel, Romm, Wehner and Collins, 1986). Similarly, in behavioral studies (Meltzer, Rosecrans, Aceto and Harris, 1980), the (-) form of nicotine is 7-10 times more potent than the (+) form of nicotine. The extent to which one isomer is more potent than the other may depend on the type of assays used, but the data as a whole support the receptormediated nature of the central actions of nicotine. Systemic administration of drugs may affect peripheral systems and, indirectly, modify activity in the CNS. Systemically-administered mecamylamine sometimes produced increases in firing rate that may partially account for the antagonism observed in those cells. However, experiments conducted several minutes after the discharge rate had decreased to normal also showed blockade of the responses to nicotine; therefore, the transient changes in rate were not considered to be responsible for the blockade of the actions of nicotine seen here. Additionally the analysis of the drug-induced percentage of rate change from baseline, used here, permits the study of effects of drug, independently of the baseline rate (Freedman et al., 1975). In conclusion, the present findings suggest that the Purkinje neuron may be adeq,tate for the study of the central pharmacological actions of nicotine. Moreover, since mecamylamine is able to block the reinforcing properties of nicotine in humans and primates (Spealman, 1983; Stolerman et al., 1973), the Purkinje cell may be an appropriate cellular model for the study of the pharmacological processes involved in the addiction to nicotine. Acknowledgement-This work was supported by the Veterans Administration Medical Research Service Award and USPHS (DA 02429). REFERENCES
Brown D. A. (1979) Neurotoxins and the ganglionic (C6) type of nicotinic receptor. In: Advances in Cyropharmacology,
Neurotoxins
: Tools
in
Neurobiology
(Ceccarelli B. and Clementi F., Eds), Vol. 3, pp. 225-230. Raven Press, New York. Clarke P. B. S. (1987) Nicotine and smoking: a perspective from animal studies. Psychopharmacology 93: 135-143. Clarke P. B. S., Hommer, D. W., Pert A. and Skirboll L. R. (1985) Electrophysiological actions of nicotine on substantia nigra single units. Br. J. Pharmac. 85: 827-835. Collins A. C., Evans C. B., Miner L. L. and Marks M. J. (1986) Mecamylamine blockade of nicotine responses: evidence for two brain nicotinic receptors. Pharmac. Biothem. Behav. 24: 1767-1773.
Freedman R., Hoffer B. J. and Woodward D. J. (1975) A quantitative microiontophoretic analysis of the responses of central neurons to noradrenaline; interactions with cobalt, manganese, verapamil, and dicholoiosprenaline. Br. J. Pharmac. 54: 529-539.
Garza R. de la, Bickford-Wimer P. C., Hoffer B. J. and Freedman R. (1987a) Heterogeneity of nicotine actions in the rat cerebellum: an in vivo electronhvsioloaical studv. J. Pharmac. exp. Ther. 240: 689-695:
_
-
I
Garza R. de la. McGuire T. J.. Freedman R. and Hoffer B. J. (1987b)‘Selective antagonism of nicotine actions in the rat cerebellum with a-bungarotoxin. Neuroscience 23: 887-89 I.
Mecamylamine
blocks
Gritz E. R. (1980) Smoking behavior and tobacco abuse. In: Advances in Substance Abuse (Mello N. K., Ed.), pp. 91-158. JAI Press Inc., Greenwich, CT. Hanson H. M., Ivester C. A. and Morton B. R. (1979) Nicotine self-administration in rats. In: Nafional Institute on Drug Abuse Research Monograph 23 (Krasnegor N. A., Ed.) DHEW Pub No (ADM) 79-800, Washington DC, Supt of Dots, US Govt Print Off, pp. 7&90. Haves W. L. (1981) Statistics, 3rd edn. Holt, Rinehart and Winston, New York. Hoffer B. J.. Sieeins __ G. R. and Bloom F. E. (1971) Studies on norepinepbrine-containing afferents to cerebellar Purkinje cells of rat cerebellum. II. Sensitivity to Purkinje cells to norepinephrine and related substances administered by microiontophoresis. Brain Res. 25: 523-534. Kumar R. and Lader M. H. (1981) Nicotine and smoking. In: Current Developments in Psychopharmacology (Essman W. B. and Valzelli K., Eds), Vol. 6, pp. 127-164. Spectrum, New York. Marks M. J., Stitzel J. A., Romm E., Wehner J. M. and Collins A. C. (1986) Nicotinic binding sites in rat
nicotine
actions
501
and mouse brain: comparison of acetylcholine, nicotine, and a-bungarotoxin. Molec. Pharmac. 30: 427-436. Meltzer L. T., Rosecrans J. A., Aceto M. D. and Harris L. S. (1980) Discriminative stimulus properties of the optical isomers of nicotine. Psychopharmacology 68: 283-286. Palmer M. R. (1982) Micro pressure-ejection: a complementary technique to microiontophoresis for neuropharmacological studies in the mammalian central nervous system. J. Electrophysiol. Tech. 9: 1233139. Palmer M. R., Wuerthele S. M. and Hoffer B. J. (1980) Physical and physiological characteristics of micropressure ejection of drugs from multibarrelled pipettes. Neuropharmacology 19: 931-938. Spealman R. D. (1983) Maintenance of behavior by postponement of scheduled injections of nicotine in squirrel monkeys. J. Pharmac. exp. Ther. 221: 154159. Stolerman I. P., Goldfarb T., Fink R. and Jarvik M. E. (1973) Influencing cigarette smoking with nicotine antagonists. Psychopharmacology 28: 247-259.
Neuropharmacology Vol. 28, No. 5, pp. 495-501, 1989 Printed in Great Britain
Pergamon Press plc
NICOTINE-INDUCED INHIBITION OF CEREBELLAR PURKINJE NEURONS: SPECIFIC ACTIONS OF NICOTINE AND SELECTIVE BLOCKADE BY MECAMYLAMINE R.
DE LA
GARZA,‘~~R. FREEDMAN’,~”and J. HOFFER~*
‘Veterans Administration Medical Center, Denver, Colorado 80262, U.S.A., *University of Colorado, Health Sciences Center, Department of Psychiatry C-268, Denver, Colorado, 80262, U.S.A. and ‘University of Colorado, Health Sciences Center, Department of Pharmacology C-236, Denver, Colorado 80262, U.S.A. (Received 17 August 1988) Summary-The specificity and pharmacological characteristics of the effects of local administration of nicotine on cerebellar Purkinje cells in the rat were examined electrophysiologically. Local application of nicotine, whether by pressure-ejection or by iontophoresis, depressed the spontaneous discharge of Purkinje neurons in a reversible and dose-dependent manner. This action could not be mimicked by local application of vehicle alone. The inhibitory effects of (-)-nicotine were several-fold more potent than that of the (+)-enantiomer. Systemic administration of the ganglion blocker mecamylamine reliably and reproducibly antagonized the nicotine-induced inhibitions of Purkinje cells whereas nicotine-induced excitation of interneurons was not altered. Local pressure-ejection of mecamylamine also antagonized the inhibitory actions of nicotine, administered by iontophoresis. Since the central effects of nicotine on behavior are stereospecific and sensitive to mecamylamine, the data in this study further support the hypothesis that the actions of nicotine on Purkinje neurons arc mediated by ganglionic-like receptors, These findings also suggest that the Purkinje cell may serve as a good cellular model for studies on central pharmacology of nicotine. Key words-nicotine,
mecamylamine, Purkinje neurons.
The reinforcing properties of nicotine are often proposed to mediate addiction to tobacco smoking (Clarke, 1987; Kumar and Lader, 1981). Indeed, several lines of evidence implicate the central actions of nicotine in tobacco smoking. For instance, the intake of nicotine in humans is directly related to the dose of nicotine in a cigarette (Gritz, 1980). Similarly, self-administration studies in animals have shown
that the administration of nicotine is reinforcing in rats (Hanson, Ivester and Morton, 1979) and monkeys (Spealman, 1983). Moreover, the nicotine ganglionic antagonist, mecamylamine blocks cigarette smoking in humans (Stolerman, Goldfarb, Fink and Jarvik, 1973) and the self-administration of nicotine in primates (Spealman, 1983). Although a recent study reported that the direct actions of nicotine on single units of the substantia nigra were blocked by mecamylamine (Clarke, Hommer, Pert and Skirboll, 1985), electrophysiological studies that directly assess the central cellular actions of nicotine-mecamylamine interactions are lacking. These studies are important because evidence from radioligand binding and behavioral studies indicate that the majority of nicotinic receptors in the CMS resemble those present in sympathetic ganglia
*To whom correspondence
should be addressed. 495
in which mecamylamine works as an antagonist of the actions of nicotine (Brown, 1979). Previous findings have shown that the actions of nicotine on cerebellar Purkinje cells were differentially sensitive to hexamethonium, a ganglionic blocker (de la Garza, Bickford-Wimer, Hoffer and Freedman, 1987a). In order to determine further whether Purkinje cells can serve as a cellular model for the ganglionic-like central actions of nicotine, in this study whether the administration of mecamylamine could also selectively block the inhibitory actions of nicotine on Purkinje cells was investigated. In addition, several other types of experiments were conducted to substantiate further the specificity of actions of the nicotine on Purkinje cells. First, local actions of nicotine on Purkinje cells, using iontophoresis, were compared with those using pressure-ejection to dissociate pressure or currentrelated artifacts from the pharmacological actions of nicotine. Second, actions of locally applied nicotine on Purkinje cells were compared with that of vehicle alone. Third, effects of local administration of the two stereoisomers of nicotine on the activity of the Purkinje cells were compared. Finally, nicotinemecamylamine interactions were examined on cerebellar interneurons, which appear to possess neuromuscular- rather than ganglionic-like putative receptors (de la Garza et al., 1987a; de la Garza, McGuire, Freedman and Hoffer, 1987b).
R. DE LA GARZA et al.
496 METHODS
Data from 32 male Sprague-Dawley rats (200-300g Sasco Omaha, New England) were suitable for analysis. The rats were anesthetized with urethane (1.5 g/kg i.p.) and prepared for the recording of neurons in lobules VI, VII, and VIII of the cerebellum, as previously described (Freedman, Hoffer, Woodward, 1975; Hoffer, Siggins and Bloom, 1971). Cells were identified by their characteristic spontaneous discharge patterns and by their evoked responses to surface stimulation of parallel fibres (de la Garza et al., 1987a). Action potentials from Purkinje cells and cerebellar inhibitory interneurons were recorded with the 3 M NaCl-filled barrel of a multi-barrelled micropipette. Spikes were separated from the background activity and converted to constant voltage pulses with a window discriminator. Pulses were also integrated over intervals of 1 set by a ratemeter and displayed on a strip chart recorder. Nicotine bitartrate (Sigma) was either dissolved in 0.1 M phosphate buffered saline (pH 7.0-7.4; 100 PM) or physiological saline (pH 6; 100 PM) and applied by pressure-ejection (Palmer, 1982). Nicotine bitartrate was also dissolved in distilled water (pH 4; 0.75 M) and aplied by microiontophoresis. The (+)-isomer of nicotine (kindly supplied by Dr Allan Collins, IBG) was also tested in some experiments (pH 7; 1OOpM) and applied by pressure-ejection. Mecamylamine HCl (Sigma) was dissolved in physiological saline and applied locally (pH 6; 10 FM) by pressure-ejection. In some experiments, mecamylamine HCl (0.12550.75 mg/kg) was also dissolved in physiological saline and administered intravenously. Drug-induced responses were quantified from the ratemeter records using a Tektronix graphics tablet and a Nova 3/12 computer. The digitized ratemeter information was transformed into percentages of drug-induced excitations or inhibitions of the firing rate of the cell by the computer. Because the resting discharge rate was sometimes altered in a nonspecific manner by application of the drug, this method for analyzing the data permitted the measuring of effects of drug independently of the resting discharge rate (Freedman et al., 1975). For each cell, the ejection pressure or iontophoretic current for nicotine, and the time of application were initially adjusted to elicit a 40 to 80% inhibition of Purkinje cell, or at least a 100% excitation of interneuron. Since the amount of drug ejected is linearly related to the product of time and ejection pressure or ionotophoretic current (Palmer, Wuerthele and Hoffer, 1980), this quantity can be utilized as an index of the dose of the drug. Drug-induced responses were considered to be blocked by mecamylamine if the inhibitory or excitatory effects of nicotine were reversibly reduced to at least 30% of the effects obtained in the absence of the antagonist. The multiple response to drug, obtained in the absence and presence of mecamylamine were averaged for each and represented the correlated observations obtained for that particular experiment.
After the intravenous injection of mecamylamine, an hour was allowed to pass in order to observe recovery of the local actions of nicotine. Several agonistic responses were obtained prior to administration of antagonist to ensure reproducibility of responses. In some experiments, dose-response curves for the isomers of nicotine were obtained and the dose required to elicit a 50% inhibition of the activity of Purkinje cell was calculated for each compound. Post hoc statistical analyses employed Student’s t-test for correlated samples to calculate the average doseresponse obtained between experiments and to establish the degree of significance of these responses to drug (Hayes, 1981). RESULTS
Studies of nicotine receptor specificity
Several control studies were conducted to investigate the pharmacological specificity of the actions of nicotine. In order to maintain pH at levels that would not disrupt cell activity, two vehicles were used in this study. Nicotine was either dissolved in physiological saline or in phosphate-buffered saline. Neither vehicle had physiological actions at doses that produced inhibition of Purkinje cell if nicotine was dissolved in the vehicle. In Figure 1 are shown two cells that were treated with nicotine or its vehicle, alternately, from separate barrels of the same micropipette assembly. In Panel A, a continuous ratemeter record shows that when nicotine was pressuredejected, an approximately 45% inhibition of activity was produced. When only the vehicle for the solution was tested, however, the same dose did not alter the activity of the Purkinje cell. Similarly, in the bottom panel of the same figure is shown a continuous ratemeter record of a second Purkinje cell. Nicotine produced an approximately 81% inhibition of the cell firing, but the phosphate buffer vehicle failed to produced any changes in rate when the same dose was used. A total of 10 cells failed to respond to either vehicle, at a dose that produced marked inhibition of the activity of Purkinje cells, using nicotine-containing solutions in an adjacent barrel of the same pipette assembly. The pharmacological specificity of the actions of nicotine was also studied by comparing the actions of isomers of nicotine. Barrels in the same pipette assembly were filled with 100pM of either (-) or (+)-nicotine and their actions were tested on the same Purkinje cell. Figure 2 top panel shows the ED,,s for inhibition of the discharge rate of Purkinje cells, the ED,,+ were obtained from individual dose-response curves for each isomer in 14 Purkinje cells. The inhibitory actions of the (-)-isomer were almost three times more potent than the (+)-isomer. The mean EDS, dose for (-)-nicotine was 9.4 f 2.5 (&SE) and the ED,, dose for (+)-nicotine was 25.1 f 5.4 (*SE). A t-test for correlated samples showed that the difference between these two doses
491
Mecamylamine blocks nicotine actions - NlCOTlNE
40 psi x set
PliYSlOLOGlCAL
l
SALlNE
- NICOTlNE
100 psi x set
.
PHOSPHATE BUFFEAEO SALME
25pulsesI set lmin Fig. I. Effects of nicotine and vehicle solutions on Purkinje neurons. Panels A and B shows ratemeter records of two Purkinje cells, treated with a fixed-dose of pressure-ejected nicotine or vehicle, respectively. In these two records, the solid bars above the traces indicate the time of application of nicotine and the solid circles indicate the time of application of vehicle. The numbers above the first bar indicates the dose, used in pounds per square inch time the duration of application of drug or vehicle (p.s.i. x set). Percentages above the ratemeter responses in this and succeeding figures represent calculated drug-induced percentage changes from baseline discharge rate. Note that the two different doses of nicotine produced an immediate reduction in discharge rate, but the same dose failed to disrupt the discharge rate of the Purkinje cell when vehicle alone was pressure-ejected.
was statistically significant [t(l3) = 2.86, P < 0.021. In the bottom two panels are representative ratemeter records of the same Purkinje cell, treated with both isomers of nicotine. Both enantiomers showed recoverable dose-dependent effects but smaller doses were required with (-)-nicotine, as seen in the middle panel. Studies of mecamylamine antagonism The pharmacological specificity of the actions of nicotine was also evaluated by investigating the actions of the ganglion blocker mecamylamine. Local pressure-ejection of mecamylamine produced variable results when interacted with pressure-ejected nicotine. Therefore, nicotine was delivered by microiontophoresis, in combination with pressure-ejected mecamylamine. Nicotine-induced inhibitions, produced by iontophoresis, were reliably and reversibly blocked by application of local mecamylamine in 5 of 6 cells. In Figure 3 is shown a representative nicotine-induced response before, during, and after pressure-ejected mecamylamine. Panels A and B show a continuous ratemeter record of a Purkinje cell. Nicotine initially produced an average 54% inhibition, but during the concurrent application of mecamylamine, the response to nicotine was markedly reduced. The average percentage nicotineinduced inhibition before mecamylamine, observed in 6 cells, was 53.8% +4.0 (X + SE) and during mecamylamine was 9.5% f 4.0 (X If: SE). A t-test for correlated samples showed that the reduction in percentage inhibition was highly significant [t(5) = 8.610, P < O.OOl].
Nicotine-induced inhibitions were also blocked by intravenously-administered mecamyamine in 10 of 14 Purkinje cells. Figure 4 shows representative data for two Purkinje cells. In panels A,-A, are shown a ratemeter record for a single Purkinje cell before, during and after administration of mecamylamine. Nicotine initially produced an average 85% inhibition, but during the application of the antagonist, the response to nicotine was reduced to about half the original response. An hour after the administration to mecamylamine, the response to the application of nicotine recovered. In panels B,-B, are shown similar responses for another Purkinje cell. In this cell, nicotine produced an initial 67% inhibition. The administration of mecamylamine completely antagonized the response to nicotine in a reversible manner, as seen in the last panel. The average percentage inhibition before the intravenous administration of mecamylamine, observed in 14 cells, was 69.1% + 3.59 (X k SE) and during mecamylamine was 17.73% + 6.84 (X + SE). A t-test for correlated samples showed that the reduction in the percentage inhibition was significant [t (13) = 9.007, P < O.OOl]. Intravenous administration of mecamylamine was also tested on cerebellar interneurons, where previous studies had shown nicotine-induced excitation to be mediated by neuromuscular-type receptors. In Figure 5, panels A,-A2 and B,-B2 show ratemeter records for two representative cerebellar interneurons, before and during the application of two doses of mecamylamine (see figure legend). The excitation, elicited by nicotine, of these two cells was not blocked by mecamylamine. In the entire interneuron sample of 9 cells, doses of mecamylamine that blocked
R. DE LA GARZA
et al.
lI
(-)
Nicotine
(T) Nicotine
17.6
f-J
30.1
psi x set
Nicotine 56.0
17.6
psi x set
25 puk.es/sec (tl
Nicotine
1 min
Fig. 2. The agonistic actions of the two stereoisomers of nicotine. The top bar graph shows the dose required to produce a 50% reduction on the discharge rate of Purkinje neurons with both isomers of nicotine. The lower two ratemeter records are from a single Purkinje neuron that was treated with increasing doses of (-)-nicotine (middle) or (+)-nicotine (bottom). See Figure I for identification of symbols. Note that (-)-nicotine was almost three times more potent than (+)-nicotine in inhibiting the activity of Purkinje cells. Also note that the dose of (-)-nicotine that produced approximately 40% inhibition in this cell failed to decrease the activity of the same cell when (+)-nicotine was pressure-ejected. Nicotine 100 nA / 10 set
Mecomylomine 2 psi
A
I
7.5 %
I 47.4 %
I 64.6 %
25 pulses/set 1 min
Fig. 3. Interactions between iontophoretically-applied nicotine and pressure-ejected mecamylamine on Purkinje neurons. Panels A and B show a continuous ratemeter record of a Purkinje neuron before, during and after the local pressure-ejection of mecamylamine. See Figure 1 for identification of symbols. The dashed line indicates the time application of mecamylamine. Note the blockade of the actions of nicotine during the application of mecamylamine and the subsequent recovery after the removal of this ganglionic antagonist.
Mecamylamine
blocks
50 spikes/ set
nicotine
499
actions
Bs
25spikes/ set
1 min
1 min
Fig. 4. Interactions between nicotine and parenterally-administered mecamylamine on Purkinje neurons. A,-A, and B,-B, show ratemeter records from two Purkinje cells before, during and after intravenous injection of mecamylamine. The solid bars above the traces indicate the time of application of nicotine and the number above the first indicates the dose of drug used in pounds per square inch times the duration of application of drug (PSI/set). Panels A, and B, show the response to nicotine 5 min after 0.25 mg,‘kg mecamylamine. Note that mecamylamine partially blocked the response to nicotine in A, and completely blocked nicotine-elicited inhibitions in B,.
inhibition of Purkinje cells did not block the excitatory actions of nicotine, but rather produced a modest enhancement in percentage excitation that was not significantly different from controls [t(8) = 1.46, P > 0.101. DISCUSSION
The present investigation provides further evidence that the actions of nicotine on Purkinje cells are pharmacologically specific because the vehicles used did not induce depressions of the activity of Purkinje cells at doses that produced marked inhibition if the solution contained nicotine. Moreover, the enantiomers of nicotine differed significantly in potency and reversible dose-dependent depressions of the
activity of Purkinje cell were elicited by local administration of nicotine, either by pressure ejection or by iontophoresis. The actions of nicotine were also shown to be pharmacologically specific because mecamylamine, whether applied locally or systemically blocked the actions of nicotine in a reproducible and reversible manner. These findings extend previous data that indicated that the actions of nicotine appeared to be mediated by two separate nicotinic sites: the inhibitory effects of nicotine through ganglioniclike (C,) receptors and the excitatory effects of nicotine through neuromuscular-like (C,,) receptors (de la Garza et al., 1987a). Since other studies have suggested that C, receptors mediate the reinforcing actions of nicotine in animals (Spealman, 1983) and humans (Stolerman et al., 1973) the present study 9 psi /set
8 psi/set 1
A,
I
I
291%
B,
I
204%
I
224 %
A2
17 spikes/ 5ec
1 min
1 min
Fig. 5. Interactions between nicotine and mecamylamine on cerebellar interneurons. A,-A, and B,-B, show ratemeter records from two interneurons before and during the intravenous injection of 0.25 mg/kg (A*) and 0.125 mg/kg (B2) of mecamylamine. Note that mecamylamine failed to block nicotine-induced excitations.
500
R. DE LA GARZA et al.
also suggests that the actions of nicotine on Purkinje neurons may be a valid cellular model for the study of the central actions of nicotine relevant to the pharmacological properties that may underline the reinforcing properties of nicotine. Because of the ability of mecamylamine to block some of the behavioral actions of nicotine with small doses and others with large doses, other studies have suggested that mecamylamine acts through at least two types of receptor (Collins, Evans, Miner and Marks, 1986). Although, it remains to be seen whether the site sensitive to large doses in their study, is also sensitive to neuromuscular blockers, like curare or a-bungarotoxin (a-BTX), the present data, together with behavioral work of Collins, seems to indicate that neuronal nicotinic receptors may not be homogeneous in their functional properties. This issue is particularly important because neuronal receptors are traditional identified as of the C, type (Brown, 1979). The present work, however, proposes that subtypes of functional neuronal receptors are more adequate hypotheses to account for the central actions of nicotine. Further studies should investigate the extent to which both the electrophysiological and behavioral actions of nicotine may be dependent on C, or C,, receptors. Local application of mecamylamine produced inconsistent results when studied with pressure-ejected nicotine. However, when nicotine was administered by iontophoresis, mecamylamine produced inconsistent results when studied with pressure-ejected nicotine. However, when nicotine was administered by iontophoresis, mecamylamine produced a marked and reversible blockade of the actions of nicotine. The difference between these findings may be related to the release characteristics of drugs, administered by pressure-ejection compared to iontophoresis. The pattern of distribution of drug is more localized when drugs are delivered by iontophoresis than when the drugs are administered by pressure-ejection. Since mecamylamine is a highly lipophihc agent, which rapidly diffuses into white matter, it is possible that it does not reach the more distal sites of the Purkinje cell membrane where pressure-ejected nicotine could act. The iontophoresis or pressure-ejection of drugs each have their own set of potential artifacts (Hoffer et al., 1971; Palmer et al., 1980). However, since in this study both the pressure ejection and iontophoresis of nicotine produced similar inhibitory effects it is suggested that the actions of nicotine were not related to drug-administered artifacts. The difference in potency between the two stereoisomers of nicotine is consistent with other data that also report (-)-nicotine to be more potent than (+)-nicotine. However, in some studies, the (-) form of nicotine is 20 to 30 times, more potent than the (+) form in inhibiting the binding of /-[-‘HInicotine and 4 times more potent inhibiting the binding of a-[‘251]-bungarotoxin to tissue from the rodent brain
(Marks, Stitzel, Romm, Wehner and Collins, 1986). Similarly, in behavioral studies (Meltzer, Rosecrans, Aceto and Harris, 1980), the (-) form of nicotine is 7-10 times more potent than the (+) form of nicotine. The extent to which one isomer is more potent than the other may depend on the type of assays used, but the data as a whole support the receptormediated nature of the central actions of nicotine. Systemic administration of drugs may affect peripheral systems and, indirectly, modify activity in the CNS. Systemically-administered mecamylamine sometimes produced increases in firing rate that may partially account for the antagonism observed in those cells. However, experiments conducted several minutes after the discharge rate had decreased to normal also showed blockade of the responses to nicotine; therefore, the transient changes in rate were not considered to be responsible for the blockade of the actions of nicotine seen here. Additionally the analysis of the drug-induced percentage of rate change from baseline, used here, permits the study of effects of drug, independently of the baseline rate (Freedman et al., 1975). In conclusion, the present findings suggest that the Purkinje neuron may be adeq,tate for the study of the central pharmacological actions of nicotine. Moreover, since mecamylamine is able to block the reinforcing properties of nicotine in humans and primates (Spealman, 1983; Stolerman et al., 1973), the Purkinje cell may be an appropriate cellular model for the study of the pharmacological processes involved in the addiction to nicotine. Acknowledgement-This work was supported by the Veterans Administration Medical Research Service Award and USPHS (DA 02429). REFERENCES
Brown D. A. (1979) Neurotoxins and the ganglionic (C6) type of nicotinic receptor. In: Advances in Cyropharmacology,
Neurotoxins
: Tools
in
Neurobiology
(Ceccarelli B. and Clementi F., Eds), Vol. 3, pp. 225-230. Raven Press, New York. Clarke P. B. S. (1987) Nicotine and smoking: a perspective from animal studies. Psychopharmacology 93: 135-143. Clarke P. B. S., Hommer, D. W., Pert A. and Skirboll L. R. (1985) Electrophysiological actions of nicotine on substantia nigra single units. Br. J. Pharmac. 85: 827-835. Collins A. C., Evans C. B., Miner L. L. and Marks M. J. (1986) Mecamylamine blockade of nicotine responses: evidence for two brain nicotinic receptors. Pharmac. Biothem. Behav. 24: 1767-1773.
Freedman R., Hoffer B. J. and Woodward D. J. (1975) A quantitative microiontophoretic analysis of the responses of central neurons to noradrenaline; interactions with cobalt, manganese, verapamil, and dicholoiosprenaline. Br. J. Pharmac. 54: 529-539.
Garza R. de la, Bickford-Wimer P. C., Hoffer B. J. and Freedman R. (1987a) Heterogeneity of nicotine actions in the rat cerebellum: an in vivo electronhvsioloaical studv. J. Pharmac. exp. Ther. 240: 689-695:
_
-
I
Garza R. de la. McGuire T. J.. Freedman R. and Hoffer B. J. (1987b)‘Selective antagonism of nicotine actions in the rat cerebellum with a-bungarotoxin. Neuroscience 23: 887-89 I.
Mecamylamine
blocks
Gritz E. R. (1980) Smoking behavior and tobacco abuse. In: Advances in Substance Abuse (Mello N. K., Ed.), pp. 91-158. JAI Press Inc., Greenwich, CT. Hanson H. M., Ivester C. A. and Morton B. R. (1979) Nicotine self-administration in rats. In: Nafional Institute on Drug Abuse Research Monograph 23 (Krasnegor N. A., Ed.) DHEW Pub No (ADM) 79-800, Washington DC, Supt of Dots, US Govt Print Off, pp. 7&90. Haves W. L. (1981) Statistics, 3rd edn. Holt, Rinehart and Winston, New York. Hoffer B. J.. Sieeins __ G. R. and Bloom F. E. (1971) Studies on norepinepbrine-containing afferents to cerebellar Purkinje cells of rat cerebellum. II. Sensitivity to Purkinje cells to norepinephrine and related substances administered by microiontophoresis. Brain Res. 25: 523-534. Kumar R. and Lader M. H. (1981) Nicotine and smoking. In: Current Developments in Psychopharmacology (Essman W. B. and Valzelli K., Eds), Vol. 6, pp. 127-164. Spectrum, New York. Marks M. J., Stitzel J. A., Romm E., Wehner J. M. and Collins A. C. (1986) Nicotinic binding sites in rat
nicotine
actions
501
and mouse brain: comparison of acetylcholine, nicotine, and a-bungarotoxin. Molec. Pharmac. 30: 427-436. Meltzer L. T., Rosecrans J. A., Aceto M. D. and Harris L. S. (1980) Discriminative stimulus properties of the optical isomers of nicotine. Psychopharmacology 68: 283-286. Palmer M. R. (1982) Micro pressure-ejection: a complementary technique to microiontophoresis for neuropharmacological studies in the mammalian central nervous system. J. Electrophysiol. Tech. 9: 1233139. Palmer M. R., Wuerthele S. M. and Hoffer B. J. (1980) Physical and physiological characteristics of micropressure ejection of drugs from multibarrelled pipettes. Neuropharmacology 19: 931-938. Spealman R. D. (1983) Maintenance of behavior by postponement of scheduled injections of nicotine in squirrel monkeys. J. Pharmac. exp. Ther. 221: 154159. Stolerman I. P., Goldfarb T., Fink R. and Jarvik M. E. (1973) Influencing cigarette smoking with nicotine antagonists. Psychopharmacology 28: 247-259.