Abstract
Neuronal nicotinic receptors are comprised of combinations of α2–9 and β2–4 subunits arranged to form a pentameric receptor. Currently, the principal central nervous system (CNS) subtypes are believed to be α4β2 and a homomeric α7 receptor, although other combinations almost certainly exist. The identity of the nicotinic receptor subtype(s) involved in the rewarding effects of nicotine are unknown. In the present study, using some recently described subtype selective nicotinic agonists and antagonists, we investigated the role of the α7 nicotinic receptor in the mediation of nicotine-induced hyperactivity and self-administration in rats. The α7 receptor agonists AR-R 17779 and DMAC failed to stimulate locomotor activity in both nicotine-nontolerant and -sensitized rats. In contrast, nicotine and the putative α4β2 subtype selective agonist SIB1765F increased activity in both experimental conditions. In nicotine-sensitized rats, the high affinity (including the α4β2 subtype) nicotinic antagonist dihydro-β-erythroidine (DHβE), but not the selective α7 antagonist methyllycaconitine (MLA), antagonized a nicotine-induced hyperactivity. Similarly, DHβE, but not MLA, pretreatment reduced nicotine self-administration. Electrophysiology experiments using Xenopus oocytes expressing the human α7 receptor confirmed AR-R 17779 and DMAC to be potent agonists at this site, and further studies demonstrated the ability of systemically administered AR-R 17779 to penetrate into the CNS. Taken together, these results indicate a negligible role of α7receptors in nicotine-induced hyperlocomotion and reward in the rat, and support the view for an involvement of a member from the high-affinity nicotinic receptor subclass, possibly α4β2. Issues such as drug potency, CNS penetration, and desensitization of the α7 receptor are discussed.
The neuronal nicotinic acetylcholine (ACh) receptor (nAChR) gene family is comprised of eight α-subunits (α2–α9) and three β-subunits (β2–β4). Although expression studies in Xenopus laevis oocytes have demonstrated that any one of the α2, α3, or α4 subunits in combination with β2 or β4 subunits can form functional pentameric receptors, in the rodent central nervous system (CNS), the most extensively characterized forms are the α4β2 receptor and a homomeric pentamer comprised solely of α7subunits (Lena and Changeux, 1997; Lukas et al., 1999). Each shows a distinct pharmacology, with nicotine having nanomolar affinity at the α4β2 receptor, which also desensitizes relatively slowly to nicotine (Buisson et al., 1996;Chavez-Noriega et al., 1997). In contrast, nicotine has micromolar affinity for the α7 receptor, which rapidly desensitizes after its activation (Séguéla et al., 1993;Briggs et al., 1995; Chavez-Noriega et al., 1997; Briggs and McKenna, 1998). Further distinctions can be made on the relative potencies of antagonists at these receptors: dihydro-β-erythroidine (DHβE) is a more potent antagonist at the α4β2 compared with the α7 receptor, whereas α-bungarotoxin (α-BGT) and methyllycaconitine (MLA) show selectivity for the α7 receptor (Chavez-Noriega et al., 1997;Holladay et al., 1997; Davies et al., 1999).
There has been considerable interest in trying to establish the functional roles of these receptor subtypes, in large part due to the wide ranging effects of nicotine on behavior, some of which suggest potential therapeutic benefit (Decker et al., 1995; Holladay et al., 1997; Lena and Changeux, 1997). Nicotine also has dependence liability, and like other drugs of abuse, it can maintain self-administration behavior in rats (Corrigall and Coen, 1989; Shoaib et al., 1997;Watkins et al., 1999) and stimulate locomotor activity (Clarke and Kumar, 1983a,b; Reavill and Stolerman, 1990; Louis and Clarke, 1998). Lesion and discrete microinjection studies have demonstrated a critical role for the mesolimbic dopamine (DA) pathway in nicotine self-administration and behavioral activation, with a direct effect of nicotine at the ventral tegmental area (VTA) at least contributing to each behavior (Clarke et al., 1988; Damsma et al., 1989; Reavill and Stolerman, 1990; Corrigall et al., 1992, 1994; Louis and Clarke, 1998). Detailed pharmacological studies are lacking, although the noncompetitive nicotinic antagonist mecamylamine robustly blocks nicotine-induced behavioral activation and reduces self-administration behavior (Clarke and Kumar, 1983a,b; Corrigall and Coen, 1989; Reavill and Stolerman, 1990). More recent studies have shown similar effects after systemically administered DHβE (Stolerman et al., 1997; Watkins et al., 1999). These latter findings are consistent with the observation that direct VTA injection of DHβE also reduces nicotine self-administration (Corrigall et al., 1994).
While these data support an involvement of the nicotine α4β2 receptor in the rewarding effects of nicotine, as yet there is little information about the role of the α7 receptor in this process. In a recent study, Schilstrom et al. (1998) reported that direct VTA injections of MLA attenuated the increased nucleus accumbens DA release elicited by systemically administered nicotine. However, the doses of MLA infused directly into the VTA via the microdialysis probe (0.1–0.3 mM) were high, being approximately 5 orders of magnitude above theKi value of this compound at the α7 receptor (Drasdo et al., 1992; Davies et al., 1999). Even accounting for the variety of factors which should result in the actual local tissue levels of MLA being much less (seeSchilstrom et al., 1998), it is feasible that at these concentrations, MLA may also block nicotine-induced currents at other sites, including the α4β2 receptor (Drasdo et al., 1992; Buisson et al., 1996). It was therefore the purpose of this study to investigate the role of α7 receptors in mediating nicotine-induced behavioral activation and self-administration further in rats. For this study, we used some of the agonist ligands recently described as being selective for the α7 receptor [e.g., AR-R 17779 (Gordon et al., 1998; Gurley et al., 1998) and DMAC (De Fiebre et al., 1995); see Fig. 1]. For comparison, the α4β2receptor agonist SIB 1765F (Sacaan et al., 1997) was also included. Specifically, we studied the effects of these drugs on locomotor activity in both nicotine-nontolerant and nicotine-sensitized rats. Furthermore, the effect of DHβE and MLA on nicotine-induced activation and self-administration was also studied. Because of the lack of published data on AR-R 17779, we conducted detailed electrophysiological studies on this compound using oocytes expressing the human α7 receptor to confirm its reported agonist properties at this site (Gurley et al., 1998). Finally, we determined the CNS levels of AR-R 17779 at the dose ranges used in the present experiments.
Chemical structures of the nicotinic agonists used.
Materials and Methods
All studies were conducted at F. Hoffmann-La Roche Ltd. (Basel, Switzerland), with the exception of the nicotine self-administration experiments, which were conducted at the Center for Addiction and Mental Health, ARF Division (Toronto, Canada). All studies complied with the appropriate local and national guidelines relating to animal experimentation.
Electrophysiology Studies
Expression of α7 Receptors in X. laevis Oocytes.
To isolate the human nAChR α7 cDNA, poly(A)+ RNA from the human neuroblastoma cell line SK-N-MC (HTB10; American Type Culture Collection, Manassas, VA) was reverse-transcribed to cDNA with random hexamers as primers using a GeneAmp RNA-PCR kit (Perkin-Elmer Cetus, Norwalk, CT) and then amplified by means of polymerase chain reaction (PCR) using the Expand High Fidelity enzyme (Boehringer-Mannheim Biochemicals, Indianapolis, IN). For the PCR procedure, the primer pair (nucleotides 54–78 as sense and 1693–1717 as antisense) derived from the human nAChR α7published sequence was used (Elliot et al., 1996). The PCR-amplified fragment was subcloned into the pCR2.1 vector of the TA cloning kit (Invitrogen, San Diego, CA). The nucleotide sequence of the insert was determined by automated cycle sequencing (Applied Biosystems, Foster City, CA). The sequence of α7 cDNA clone was identical with GenBank/EMBL accession number U62436 (Elliot et al., 1996). The human α7 cDNA was subcloned into the eukaryotic expression vector pCMVneo. The plasmid was diluted (1.5 μg/ml) in a buffer containing 88 mM NaCl, 1 mM KCl, and 15 mM HEPES (pH 7.0) for intranuclear oocyte injection.
Ovarian lobes were excised from South African frogs (X. laevis) anesthetized with tricaine methanesulfonate (2 g/l). The tissue was agitated (15–20 min, 37°C) in modified Barth's medium [88 mM NaCl, 2.4 mM NaHCO3, 1 mM KCl, 0.33 mM CaCl2, 0.33 mM Ca(NO3)2, 0.82 mM MgSO4, 10 I.U./ml penicillin, 10 μg/ml streptomycin] containing collagenase (400 U/ml, type 1A; Sigma Chemical Co., St. Louis, MO) and trypsin inhibitor (0.2 mg/ml, type 1S; Sigma Chemical Co.) to separate individual oocytes. After washing in medium without enzymes, oocytes at maturation stage V to VI were shortly (1 min) exposed to a 2-fold concentrated solution, and the follicle cell layer was removed by agitation in a pipette. The nuclei of the oocytes were shifted toward the surface of the animal pole by centrifugation (relative centrifugal force 800, 10 min). For microinjection, pointed glass capillaries of approximately 60-μm diameter at the opening were fabricated on a microforge made in-house. Using a Nanoliter Injector (WPI, Sarasota, FL), oocytes were injected intranuclearly with 9 nl of buffer containing the expression plasmid. Injected oocytes were kept at 20°C for the experiments on the following 2 days.
Electrophysiology.
Single oocytes were put onto a grid in a small chamber (volume 60 μl) close to the outlet for solution application. The basal saline superfusing the oocytes contained 90 mM NaCl, 1 mM KCl, 2.5 mM BaCl2, 1 mM MgCl2, and 1 μM atropine sulfate. ACh chloride and (−)-nicotine tartrate were dissolved directly in the basal saline. For the other compounds, 1000-fold concentrated stock solutions were prepared in dimethyl sulfoxide, and equal amounts of this or pure dimethyl sulfoxide were added to the test and control solutions, respectively. Exchange between solutions containing different additions was achieved by a motor-driven, 4-port MVP valve (Hamilton Co., Reno, NV) that switched between two different feeding tubes close to the outlet. The flow of solutions toward this valve was controlled by an array of electromagnetic valves (General Valve, Fairfield, NJ) farther upstream. All connections were made with Teflon tubing.
Oocytes were impaled by two glass microelectrodes filled with 2 M KCl, and the membrane potential was set to −80 mV by a voltage-clamp amplifier (NPI Electronic, Tamm, Germany). The signal from the current monitoring output of the amplifier was low-pass filtered (30 Hz), digitized (ITC-16; Instrutech), and stored in a personal computer. WinTida software (HEKA Elektronik, Lambrecht, Germany) was used to control the valves, for signal recording, and for primary analysis. Current amplitude values measured in WinTida were transferred to Excel (Microsoft) for statistical analysis. XLFit (IDBS, Guildford, Surrey, UK) was used to fit the data in concentration-response relations by the function
In Vivo Studies
Locomotor Activity Studies.
Male Sprague-Dawley rats (RCC Ltd., Fullinsdorf, Switzerland) were used throughout the study. The animals were housed four per cage in a light- and temperature-controlled environment (lights on 6:00 AM to 6:00 PM) with food available ad libitum. All testing was conducted during the animals' light phase. To study the effect of the various nicotinic agonists on locomotor activity, two approaches were taken: a study in nicotine-nontolerant rats and a second study in nicotine-sensitized rats. A repeated measures design was used for both studies, with the rats habituated to the test apparatus (36 × 24 × 19 cm; Benwick Electronics, Cambridge, UK) for three times 2-h daily sessions before formal activity testing commenced. In the nicotine-nontolerant studies, rats received (−)-nicotine or test drug as a single injection before testing, which was of a 90-min duration. A 30-min acclimatization period to the test apparatus preceded drug treatment, and a washout period of 2 to 3 days intervened between each treatment cycle. This methodology was adapted from that reported byClarke and Kumar (1983) in which the animals were described as “nicotine-nontolerant”. An identical protocol was used for the nicotine-sensitized rats, except they received 10 daily injections of nicotine (0.4 mg/kg s.c.) before drug testing began. The animals continued to receive nicotine injections on the days between treatment cycles. In the sensitized animals, except for test days, the daily nicotine injections were given noncontingently with exposure to the test apparatus. A dose of 0.4 mg/kg s.c. nicotine was included in each study as a positive control. All compounds were studied in separate groups of rats, with eight rats per group.
The antagonist studies were conducted only in nicotine-sensitized rats, which were prepared as previously described. For each experiment, a 2 × 2 factorial design was used; thus, for each study, only a single dose of antagonist was used against a standard nicotine dose of 0.4 mg/kg.
Nicotine Self-Administration Studies.
Male Long-Evans rats (Charles River, Lachine, Quebec, Canada) were singly housed in a reversed light-dark holding room (lights on 7:00 PM to 7:00 AM). Before the start of experimental procedures, the animals had ad libitum access to food.
The procedures used for training and surgery were similar to those described in previous reports (Corrigall and Coen, 1989). First, rats were food deprived for 24 h and trained to lever press for food (45-mg Noyes pellet) under a continuous reinforcement schedule. Once trained, each animal was surgically implanted with a chronic i.v. catheter placed in the jugular vein and exiting between the scapulae. All surgery was performed under xylazine (10 mg/kg i.p.) and ketamine HCl (90 mg/kg i.p.) anesthesia. A single injection of buprenorphine (0.01 mg/kg s.c.) was administered for postoperative analgesia and penicillin (30,000 I.U. i.m.) was also given at this time to minimize wound infection. Animals were allowed to recover for 1 week before drug self-administration training commenced.
Drug self-administration was initiated on a continuous reinforcement schedule with a 1-min timeout (TO; lights off) period after each infusion. During the TO, responses were recorded but not reinforced. Over a training period of approximately 3 weeks, the response requirements were increased to a final value of fixed-ratio 5, with the TO remaining at 1 min. The unit dose of nicotine used for the self-administration sessions was 0.03 mg/kg/infusion. Self-administration sessions were conducted in two lever operant chambers, with only one of the levers active. Responses on each lever were recorded. Test sessions were of 60-min duration and were run each weekday. Responding was considered stable, and therefore amenable for drug testing, when the mean number of weekly infusions did not vary by more than 15%.
Two experiments were conducted. In the first experiment, MLA (5–10 mg/kg i.p.) or vehicle was administered 30 min before the test session according to a randomized design. At least 2 days separated successive treatments with the animals run in self-administration sessions as normal. After a 1- week washout period, during which the animals were run in daily self-administration sessions but received no drug pretreatment, a second experiment studied the effect of DHβE (3–10 mg/kg s.c.) or vehicle pretreatment on nicotine self-administration. Again, at least 2 days separated successive treatments that were given in a randomized sequence. A total of 12 rats were used in both of these experiments.
Studies to Investigate CNS Penetration of AR-R 17779.
Male RORO rats (RCC Ltd., Fullinsdorf, Switzerland) were dosed with AR-R 17779 (30 mg/kg i.p.). Either 30 min or 1 h later, the rats were culled, and brain, cerebrospinal fluid (CSF), and plasma samples were taken. CSF was sampled according to the method described by Huang et al. (1996); 100 μl of CSF and plasma, and brain homogenate prepared in 2 volumes of saline, were each mixed with 100 μl of 0.2 M K2CO3, and AR-R 17779 was extracted using 1 ml of ethyl acetate. The organic phase was collected and taken to dryness under nitrogen. Residual was dissolved using 50 μl of mobile phase (formic acid/methanol 1:99). All biological samples were subsequently analyzed by liquid chromatography and tandem mass spectrometry.
Compounds and Injections.
The following compounds were used (source in parentheses): (−)-nicotine hydrogen tartrate (Sigma Chemical Co.), MLA, and DHβE (Research Biochemicals International, Natick, MA). SIB 1765F, DMAC, and AR-R 17779 were all synthesized within the Roche CNS Chemistry Department. All compounds were dissolved in 0.9% NaCl solution (saline), and the pH of nicotine and SIB 1765F was adjusted to 7.0 by the addition of sodium hydroxide. Doses are expressed as that of the base, and all compounds were injected in a dose volume of 1 ml/kg. The route of administration was s.c., except for MLA, DMAC, and AR-R 17779, which were given by the i.p. route. Pretreatment times were as follows: nicotine and SIB 1765F, 5 min; DHβE, 10 min; MLA and DMAC, 15 min; and AR-R 17779, 30 min. All doses and pretreatment times were based on published work (Meyer et al., 1994; Menzaghi et al., 1997; Stolerman et al., 1997; Kaiser et al., 1998; Turek et al., 1998).
For the self-administration studies (−)-nicotine bitartrate (Sigma Chemical Co.) was used. The nicotine solution was prepared in saline, and the pH was adjusted to 7.0 with sodium hydroxide; this was then filtered (0.22-μm pore size) for sterilization purposes before use.
Statistical Analysis
Data consisted of the number of nicotine infusions or the number of locomotor activity counts. Although activity counts were taken at 10-min time bins; for the sake of brevity, these data are collapsed into the total number of counts recorded over the 90-min test session. These data were analyzed by single-factor repeated measures ANOVA, and where appropriate followed by simple main effects tests.
Results
Electrophysiology Studies
Activation and Desensitization.
The potency of compounds for activation of the α7 nAChR was determined by voltage-clamp experiments on X. laevis oocytes 1 to 2 days after nuclear injection of cDNA encoding the human variant of the receptor. When oocytes were clamped to −80 mV rapid application of ACh, AR-R 17779, or the other agonists evoked a transient inward current as exemplified in Fig. 2A. Agonist applications (6-s duration) were usually repeated every 3 min, but longer rest intervals (5–10 min) were chosen when testing the higher concentrations of nicotine or DMAC (>10 μM) because of the long-lasting desensitization induced by these compounds. ACh (500 μM) was repeatedly applied between the other compounds to obtain reference responses for normalization. Peak current amplitudes were expressed as a fraction of the closest reference response and plotted versus concentration to determine dose-response relations (Fig. 2, B–D, circles). The data were fitted by eq. 1 to estimate the values of the EC50, the Hill coefficient, and the maximal effect. The mean values of the fitted parameters are given in Table1. The comparison of the results shows that AR-R 17779 as an agonist is about 3 times more potent than nicotine and produces a 40% larger maximal effect. DMAC is even more potent, however, reaching only 48% of the maximal effect of nicotine.
Concentration-dependent activation and desensitization of nicotinic α7 receptor expressed inX. laevis oocytes. A, representative current traces recorded during stimulation by ACh or AR-R 17779 (AR). Agonist applications are indicated by the bars above the traces and lasted 6 s, longer than the duration of the displayed recordings. B through D, concentration-response relations for the activating (●) and inhibitory or desensitizing (▪) effects of AR-R 17779 (B), DMAC (C), and nicotine (D). Peak current amplitudes were expressed as a fraction of reference responses evoked by 500 μM acetylcholine in the absence of any other compound. Symbols and error bars indicate arithmetic mean and S.E. values from 3 to 10 oocytes. S.E. was smaller than the symbol size when no error bars are seen. Several, but not all, concentrations of a given compound were applied to each oocyte. Curves were fitted to the populations of all data points by using eq. 1.
Nicotinic α7 receptors expressed in Xenopusoocytes: parameter values of fitted dose-response curves for selected drugs used in this study
Briggs and McKenna (1998) recently showed that ACh, nicotine, and some other agonists of the α7 nicotinic receptor are more potent as inhibitors than as activators due to receptor desensitization. We therefore looked for the desensitizing properties of AR-R 17779 and DMAC and compared them with nicotine. After recording a series of control responses evoked by short (6-s) applications of ACh (500 μM), oocytes were continuously (5–10 min) superfused by low concentrations of the agonists while continuing with the 500 μM ACh stimuli (rate 1/3 min−1). As expected, prolonged exposure to any agonist decreased the amplitudes of the ACh responses in a concentration-dependent fashion (Fig. 2, B and C, squares). The IC50 values and Hill coefficients estimated by fitting eq. 1 to the data are summarized in Table 1. All tested compounds were roughly 100-fold more potent as inhibitors than as activators.
Following the same experimental protocol, we also studied the inhibitory effects of the competitive antagonists MLA and DHβE for comparison. The IC50 values recorded from these experiments are presented in Table 1.
Nondesensitizing Current at Low Agonist Concentrations.
The superimposed activation and desensitization curves of Fig. 2, B to D, suggest that a small (0.2% of the maximum peak current) steady-state activity of the α7 nicotinic receptor will occur in a narrow range around the concentration where the curves cross each other (i.e., 1 μM AR-R 17779, 0.7 μM DMAC, and 5 μM nicotine). In a second series of studies, we investigated whether low concentrations of AR-R 17779 could induce a larger nondesensitized current than nicotine or ACh, by comparing the effects of AR-R 17779 (0.5–4 μM) and either nicotine (2.5–10 μM) or ACh (5–40 μM) applied to the same oocyte. AR-R 17779 (1 μM), nicotine (5 μM), and ACh (10 μM) induced similarly large currents hardly desensitizing during 90 to 100 s of application (Fig.3, A and C). The responses to 2-fold lower concentrations of each agonist were smaller or even undetectable (data not shown). Two-fold higher concentrations evoked a current peak decaying to a plateau approximately equal to the current induced by the lower concentrations (Fig. 3, B and D). Four-fold higher concentrations stimulated larger peaks, but the plateau current was either the same or smaller than that induced by the lower concentrations (not shown). Similar effects were seen in three other oocytes for each pair of compounds. On average, the plateau current induced by either 5 or 10 μM nicotine (whichever was larger) was 94 ± 15%, and that induced by 10 or 20 μM ACh was 130 ± 14%, of the plateau current induced by 1 or 2 μM AR-R 17779.
Comparison of responses induced by prolonged (90–100 s) application of low agonist concentrations. A and B, comparison of nicotine (dashed lines) and AR-R 17779 (solid lines). C and D, comparison of ACh (dashed lines) and AR-R 17779 (solid lines) responses. The following concentrations are superimposed in the four panels: A, 5 μM nicotine and 1 μM AR-R 17779; B, 10 μM nicotine and 2 μM AR-R 17779; C, 10 μM ACh and 1 μM AR-R 17779; D, 20 μM ACh and 2 μM AR-R 17779.
In Vivo Studies
Locomotor Activity Experiments.
In nontolerant rats habituated to the test apparatus, nicotine (0.1–0.4 mg/kg s.c.) produced a significant hyperactivity (F3,21 = 3.6, P < .05) over the 90-min test session. In common with previous studies, this effect of nicotine was of considerably greater magnitude and overall significance in sensitized rats (F3,21 = 61.4, P < .001; Fig. 4). In the nicotine-nontolerant experiment, there was no evidence of sensitization with successive treatment cycles (i.e., cycle 1 mean activity counts: 0.4 mg/kg nicotine = 668; cycle 4 mean activity counts: 0.4 mg/kg nicotine = 526; n = 2 rats per cycle).
Effects of nicotine, DMAC, AR-R 17779, and SIB1765F on locomotor activity in nicotine-nontolerant (a) and nicotine-sensitized (b) rats. Each compound was tested in separate groups of rats (n = 8 per group) according to a randomized design. Data presented as the mean ± S.E. of locomotor activity counts recorded during the 90-min test session. Note the different y-axis scale for the SIB1765F study in sensitized rats. N, 0.4 mg/kg nicotine s.c. *P < .05, **P < .01 versus vehicle pretreatment (simple main effects test after significant ANOVA).
Neither AR-R 17779 (10–30 mg/kg i.p.) nor DMAC (1–10 mg/kg i.p.) produced any stimulant affect on locomotor activity in either nicotine-nontolerant or -sensitized rats, despite nicotine (0.4 mg/kg s.c.) eliciting a significant hyperactivity in each experiment (Fig.4). In fact, in nontolerant rats, DMAC tended to reduce activity, although this was of borderline significance and was not replicated in nicotine-sensitized rats. At a dose of 30 mg/kg, DMAC further depressed activity and affected body posture (data not shown). SIB 1765F (1–20 mg/kg s.c.) produced a robust hyperactivity in both nontolerant and sensitized rats (nontolerant rats:F4,28 = 8.3, P < .001; sensitized rats: F4,44 = 60.4,P < .001). In each experiment, the magnitude of peak effect produced by SIB 1765F over the 90-min test period was equivalent to that produced by nicotine (0.4 mg/kg s.c.; Fig. 4). However, in the nontolerant experiment, there were trends to suggest sensitization developed with successive treatment cycles (i.e., cycle 1/2 mean activity counts: 0.4 mg/kg nicotine = 1771 ± 443, 20 mg/kg SIB = 2095 ± 281; cycle 4/5 mean activity counts: 0.4 mg/kg nicotine = 2793 ± 789, 20 mg/kg SIB = 3694 ± 1353; n = 3–4 rats).
In the antagonist studies conducted in sensitized rats, the hyperactivity produced by nicotine (0.4 mg/kg s.c.) was significantly reduced by DHβE (3 mg/kg s.c.) but not MLA (5 mg/kg i.p.; Fig.5). At these doses, each antagonist alone had no significant effect on baseline activity. At a higher dose of 10 mg/kg, MLA did produce a partial (∼30%) attenuation of the nicotine response of borderline significance (vehicle/vehicle, 396 ± 47; MLA/vehicle, 477 ± 158; vehicle/nicotine, 3793 ± 452; MLA/nic, 2701 ± 494). This higher dose of MLA did not block a cocaine (15 mg/kg i.p.)-induced hyperactivity (vehicle/vehicle, 271 ± 96; MLA/vehicle, 288 ± 94; vehicle/cocaine, 2847 ± 708; MLA/cocaine, 2561 ± 543).
Effects of DHβE (3 mg/kg s.c.) and MLA (5 mg/kg i.p.) pretreatment against a nicotine (0.4 mg/kg s.c.)-induced hyperactivity in nicotine-sensitized rats (n = 8) (a) and the i.v. self-administration of nicotine in Long-Evans rats (n = 12) (b). Data presented as the mean ± S.E. of locomotor activity counts recorded during the 90-min test session or the mean ± S.E. number of nicotine infusions taken during a 60-min self-administration session. V, vehicle/vehicle group; A, antagonist/vehicle group; N, 0.4 mg/kg nicotine/vehicle group, A + N, antagonist/nicotine group. *P < .05, **P < .01 versus vehicle pretreatment.#P < .01 versus vehicle/nicotine (simple main effects test after significant ANOVA).
Nicotine Self-Administration Experiments.
In accordance with previous studies, robust self-administration of nicotine (0.03 mg/kg/infusion) was acquired over a training period of 3 to 4 weeks with the animals (n = 12) taking 14 ± 2 infusions per 60-min session at asymptote. In the first experiment, MLA pretreatment failed to affect the number of nicotine infusions recorded over the test session (F2,22 = 1.9, N.S.). In a second study using the same rats after a 1-week washout period, during which the animals received daily self-administration sessions, DHβE produced a significant reduction of nicotine self-administration (F2,22 = 36.1,P < .001), with significant reductions noted at both the 3 and 10 mg/kg doses (Fig. 5).
Studies to Investigate CNS Penetration of AR-R 17779.
AR-R 17779 was detected in the CSF at 30 min and 1 h after an i.p. injection of a 30 mg/kg dose. The CSF concentration was in the range of 4 to 7 μM, with corresponding plasma concentrations in the range of 45 to 75 μM, thus yielding a CSF/plasma ratio of 9.0% at 30 min and 9.1% at 1 h. The brain levels of AR-R 17779 were lower than CSF, and correspondingly the brain/plasma ratios were 5.3% (30 min) and 5.5% (1 h). These data are summarized in Table2.
Plasma, brain, and CSF levels of AR-R 17779 at 30 and 60 min after the administration of a 30 mg/kg i.p. dose. Values are mean ± S.E. The number of animals per group are in parentheses.
Discussion
The nicotine α7 receptor is characterized by a rapid desensitization and low affinity to nicotine. Because subtype-selective ligands for this receptor have only recently become available, a consideration of the value of the ligands used to evaluate α7 receptor function in vivo is necessary before drawing conclusions about the role of this site in nicotine reward.
Utility of Drugs Used to Probe α7 Receptor Function In Vivo.
We used the recently described nicotine α7 receptor agonists AR-R 17779 and DMAC (De Fiebre et al., 1995; Gordon et al., 1998; Gurley et al., 1998). The anabaseine derivative DMAC has a Kivalue of 34 nM to rat brain 125I-BGT binding sites (presumably α7 receptors) compared with 348 nM at high-affinity nicotinic receptors (non-α7 receptors) measured by [3H]cytisine binding assay (De Fiebre et al., 1995). Such selectivity is also supported by functional data using oocytes expressing either rat α7 or α4β2 receptors; DMAC has an EC50 value of 4 μM at the former site and only very weak partial agonist properties at other nicotinic receptors, including the α4β2 site (De Fiebre et al., 1995). Using an oocyte system expressing human α7 receptors, we also found DMAC to be a relatively potent agonist, although unlike the study of De Fiebre et al. (1995), the maximal response was significantly less than that attainable by ACh. This may be reflective of a species difference between the rat and human α7 receptors (Séguéla et al., 1993; Chavez-Noriega et al., 1997). Regarding AR-R 17779, we could also confirm the preliminary report ofGurley et al. (1998) that this drug is a relatively potent α7 receptor agonist, with a maximal response equivalent to that of ACh. Our EC50 value of 26 μM is almost identical to that of Gurley et al. (1998), who report an EC50 value of 25 ± 4 μM at the rat α7 receptor expressed in oocytes. Binding data using 125I-BGT and [3H]nicotine assays support the selectivity of this ligand for α7 relative to the various high-affinity nicotine receptors, because AR-R 17779 has aKi value approximately 180-fold lower at the former site (Ki = 0.09 μM versus 16 μM; Gordon et al., 1998). In addition, the i.p. injection of AR-R 17779 resulted in CSF and brain levels in the micromolar range, confirming the ability of this compound to penetrate the CNS after systemic administration.
A further issue is the rapid desensitization that is characteristic of agonist-mediated responses at the α7 receptor (Séguéla et al., 1993;Briggs et al., 1995; Chavez-Noriega et al., 1997; Briggs and McKenna, 1998). Given that any exogenously administered drug will be present in the extracellular space of the receptor for a time period of minutes to hours, the α7 receptor may be in a predominantly desensitized state. Papke and Thinschmidt (1998)suggested that steady-state effects may be more important than transient responses when the α7 receptor is used as a target for drug therapy. We therefore looked more closely at the nondesensitizing responses evoked in the oocytes by low agonist concentrations. The experiments indicated that the CSF levels of AR-R 17779 reached in the experiments on rats should have some steady-state effect at α7 nicotinic receptors, but we did not see a significant difference between the maximum steady-state currents induced by nicotine and AR-R 17779.
Taken together, these data suggest that a failure of AR-R 17779 or DMAC to mimic any responses produced by nicotine is unlikely to be due to explanations such as low potency, stronger receptor desensitization, or poor CNS penetration (see Meyer et al., 1994 with respect to CNS penetration of DMAC). Both DMAC and AR-R 17779 have clear demonstrable agonist effects at the α7 receptor, with potencies greater than that of nicotine. Furthermore, studies by Turek et al. (1998) have demonstrated that MLA will penetrate the CNS after i.p. injection. More specifically, their data suggest that at a dose of 6.2 μmol/kg i.p. (corresponding to 4.2 mg/kg), brain levels of approximately 50 nM are attained 30 to 60 min after injection. This concentration is approximately 50-fold greater than theKi of MLA at the α7 receptor (Davies et al., 1999), thus presumably sufficient to block, or at least attenuate, any nicotine-mediated responses at this site.
Role of Nicotine α7 Receptor on Nicotine Hyperactivity and Self-Administration.
The acute systemic administration of nicotine (0.1–0.4 mg/kg) produced a modest hyperactivity in nicotine-nontolerant rats habituated to the test apparatus and most notably in rats sensitized to nicotine by 10 daily injections. Under each experimental condition, neither AR-R 17779 nor DMAC increased activity. Indeed, DMAC at the lowest dose tested produced a modest hypoactivity in the nicotine-nontolerant rats. The lack of effect of AR-R 17779 in nontolerant rats is in agreement with the preliminary report of Kaiser et al. (1998). This study extends these observations to nicotine-sensitized rats. We are unaware of equivalent studies having been conducted with DMAC, although Meyer et al. (1994) reported no overt behavioral changes at doses of 0.6 to 6 μmol/kg i.p. (corresponding to 0.25–2.5 mg/kg), which improved passive avoidance performance.
At the present time, some of the most compelling evidence to support a role for nicotinic α7 receptors in nicotine reward and hyperactivity involves MLA. Nicotine, like other locomotor stimulants and drugs of abuse, increases the release of DA in the nucleus accumbens (Clarke et al., 1988; Damsma et al., 1989).Schilstrom et al. (1998) reported that direct infusion of MLA into the VTA by reverse microdialysis blocked this nicotine-mediated effect. However, notwithstanding issues of in vivo recovery using this approach, the doses of MLA used in this study were high, with only the 0.3 mM concentration having a significant effect. Because theKi value of MLA for the α7 receptor is in the nanomolar range (Chavez-Noriega et al., 1997; Davies et al., 1999; this study), one might have predicted blockade of α7 receptors at lower doses. Indeed, at micromolar concentrations, MLA has affinity for a number of nicotinic sites, including α4β2 (Drasdo et al., 1992; Buisson et al., 1996). We report here that MLA at 5 to 10 mg/kg failed to influence nicotine self-administration and produced only a slight attenuation of nicotine hyperactivity at the highest (10 mg/kg) dose. Although this dose of MLA did not block a cocaine-induced hyperactivity, implying some behavioral specificity, this could reflect actions at nicotinic receptors other than α7, due to available data on CNS levels of MLA after systemic administration (Turek et al., 1998).
These data seem to support a critical role for a high-affinity nicotinic receptor subtype, possibly α4β2, as underlying the stimulant and rewarding effects of nicotine. First, SIB 1765F, a compound described as having some selectivity for the α4β2 subtype (at least relative to α4β4 and α7; Sacaan et al., 1997), produced a robust hyperactivity similar in magnitude to nicotine. These data essentially confirm the observations of Menzaghi et al. (1997), although these workers reported that SIB 1765F produced locomotor activity increases of considerably greater magnitude than nicotine. This discrepancy may be attributable to the development of sensitization to nicotine during the course of this experiment, as evidenced by the fact that the nicotine response in the SIB 1765F study was greater than that observed in the other agonist studies in nontolerant rats. Thus, cross-sensitization may occur between SIB 1765F and nicotine (seeGrottick et al., 2000). Second, DHβE, an antagonist with approximately 100-fold selectivity for human α4β2 versus human α7 receptors (Chavez-Noriega et al., 1997), potently blocked a nicotine hyperactivity and reduced nicotine self-administration, a finding consistent with other reports (Stolerman et al., 1997; Watkins et al., 1999). Third, studies using mice deficient in particular nicotinic subunits suggest that a β2 knockout line shows reduced propensity to self-administer nicotine and has attenuated accumbens DA release after acute nicotine challenge (Picciotto et al., 1998; Epping-Jordan et al., 1999). Although equivalent studies have not been reported with the other nicotinic subunit knockouts described to date [e.g., α7 (Orr-Urtreger, 1997), α4 (Marubio et al., 1999), α3 (Xu et al., 1999a), or β4 (Xu et al., 1999b)], these studies support a role for a nicotinic receptor containing the β2 subunit. An assessment of nicotine-induced hyperactivity in these various knockout lines would be of further interest. Our data imply that a nicotine hyperactivity should be similar in terms of magnitude and effective dose range between the α7 knockout and wild-type controls. Of final note, recent pharmacological investigations using rat striatal synaptosome preparations have suggested a role for α4β2 and α3β2 receptors in the presynaptic control of dopamine release by nicotine (Kulak et al., 1997; Sharples et al., 2000). Thus, multiple nAChRs belonging to the high-affinity subclass may be involved in the hyperlocomotor and rewarding effects of nicotine.
In conclusion, using a pharmacological approach, we have been unable to identify any involvement of the nicotinic α7 receptor in maintaining nicotine self-administration and hyperactivity in rats. Further pharmacological studies, perhaps applied to various mouse lines deficient in particular nicotine receptor subunits, will be of value in extending these observations, especially in the delineation of the various subunit combinations believed to comprise the high-affinity CNS nicotinic receptor family.
Footnotes
-
Send reprint requests to: Dr. G. A. Higgins, PRBN-B, Bau 72/150, F. Hoffmann-La Roche Ltd., Basel, Switzerland. E-mail: guy_a.higgins{at}roche.com
-
↵1 Funding for this work was provided by F. Hoffman-La Roche AG. We would like to thank Kathy Coen and Laurie Adamson for their contribution to the nicotine self-administration experiments and Eva Pflugfelder for assistance with the oocyte experiments.
- Abbreviations:
- ACh
- acetylcholine
- nAChR
- nicotinic acetylcholine receptor
- CNS
- central nervous system
- DA
- dopamine
- TO
- timeout
- PCR
- polymerase chain reaction
- VTA
- ventral tegmental area
- DMAC
- [4-[(1E,3E)-3-(5,6-dihydro-4H-[2,3′]bipyridinyl-3-ylidene)-propenyl]phenyl]dimethyl-amine
- DHβE
- dihydro-β-erythroidine
- MLA
- methyllycaconitine
- α-BGT
- α-bungarotoxin
- AR-R 17779
- (−)-spiro[1-azabicyclo[2.2.2]octane-3,5′-oxazolidin]-2′-one
- SIB-1765F
- 3-ethynyl-5-(1-methyl-2-pyrrolidinyl)pyridine
- Received February 18, 2000.
- Accepted May 8, 2000.
- The American Society for Pharmacology and Experimental Therapeutics