|Home | About | Journals | Submit | Contact Us | Français|
High affinity, β2 subunit-containing (β2*) nAChRs are essential for nicotine reinforcement; however, these nAChRs are found on both GABA and dopaminergic (DA) neurons in the VTA, and also on terminals of glutamatergic and cholinergic neurons projecting from the pedunculopontine tegmental area and the laterodorsal tegmental nucleus. Thus, systemic nicotine administration stimulates many different neuronal subtypes in various brain nuclei. To identify neurons in which nAChRs must be expressed to mediate effects of systemic nicotine, we investigated responses in mice with low-level, localized expression of β2* nAChRs in the midbrain/VTA. Nicotine-induced GABA and DA release were partially rescued in striatal synaptosomes from transgenic mice as compared to tissue from β2 KO mice. Nicotine-induced locomotor activation, but not place preference, was rescued in mice with low-level VTA expression, suggesting that low-level expression of β2* nAChRs in DA neurons is not sufficient to support nicotine reward. In contrast to control mice, transgenic mice with low level β2* nAChR expression in the VTA showed no increase in overall levels of CREB but did show an increase in CREB phosphorylation in response to exposure to a nicotine-paired chamber. Thus, CREB activation in the absence of regulation of total CREB levels was not sufficient to support nicotine place preference in β2tr mice. This suggests that partial activation of high affinity nAChRs in VTA might block the rewarding effects of nicotine, providing a potential mechanism for the ability of nicotinic partial agonists to aid in smoking cessation.
Activation of the ventral tegmental area (VTA) and striatum/nucleus accumbens (NAc) is likely to be the first step in drug reinforcement, and ultimately, the development of addiction. Local infusion of nicotine or nicotinic agonists into the VTA results in increased locomotor activity (Museo & Wise, 1994b; Laviolette & Van Der Kooy, 2003) and can result in conditioned place preference (Museo & Wise, 1994a). In addition, rats and mice self-administer nicotine directly into the VTA (Ikemoto et al., 2006, Maskos et al., 2005). Nicotine binds to nicotinic acetylcholine receptors (nAChRs), leading to depolarization of cell bodies and neurotransmitter release from presynaptic terminals in several brain areas (Marks et al., 1999). When administered in VTA slices, nicotine can activate (and desensitize) nAChRs located on DA and GABA nerve terminals. In addition, nicotine can stimulate glutamatergic neurons originating in the pedunculopontine tegmentum (Mansvelder et al., 2002, Mansvelder & Mcgehee, 2000) and cholinergic neurons originating in the laterodorsal tegmental nucleus (Blaha et al., 1996) both of which synapse onto DA neurons in the VTA. Thus, activation of nAChRs by systemically- administered nicotine can result in stimulation of many different neuronal subtypes resulting in release of a number of neurotransmitters, in various brain nuclei (inside and outside the VTA), leading to behavioral outcomes that may depend on distinct circuits. It is therefore of interest to identify the neuronal pathways in which nAChRs must be expressed in order to mediate the effects of systemic nicotine on behaviors related to reinforcement and addiction.
Although some of the neural circuitry involved in acute nicotine action has been defined pharmacologically, less is known regarding β2*nAChR contributions to nicotine-associated changes in cellular and molecular signaling. Previous studies have demonstrated that β2 subunit-containing (β2*) nAChRs are critical for nicotine-mediated dopamine (DA) release, locomotor activation and reward learning (Grady et al., 2001, King et al., 2004, Marubio et al., 2003, Maskos et al., 2005, Picciotto et al., 1998, Tapper et al., 2004, Walters et al., 2006). It has also been shown that chronic administration of nicotine in a regimen that induces locomotor activation (King et al., 2004) leads to adaptations in intracellular signaling pathways within the meso-cortico-limbic system (Brunzell et al., 2003), including signaling through the cAMP response element-binding protein (CREB), a transcription factor known to be critical for long-term adaptations to treatment with drugs of abuse. CREB αδ knockout mice also fail to show nicotine conditioned place preference (Walters et al., 2003). It is not known whether these adaptations are downstream of nAChRs expressed in the VTA. In the current study we investigated the effects of local, low-level rescue of β2* nAChRs in the neurons of the VTA on locomotor activation and nicotine conditioned place preference. In addition, we sought to determine whether changes in CREB activity during these behavioral paradigms are dependent on β2* nAChRs expressed in VTA neurons. These experiments aim to clarify some of the underlying neuronal and molecular events that could explain the very sharp dose-response function for nicotine reward and may provide new points of intervention to support smoking cessation.
β2 trangenic mice (β2tr) express β2* nicotinic acetylcholine receptors only in the neurons originating in the VTA region. These animals were generated as described (King et al., 2003) and backcrossed at least 12 generations onto the C57BL/6J background. Mice from the founder line carrying a β2 cDNA transgene driven by a tetracycline-operon (tet-β2) promoter were crossed with Line C NSE-tTA mice (Chen et al., 1998; Kelz et al., 1999); subsequently, the offspring were crossed with β2 KO mice on the C57BL/6J background (Picciotto et al., 1995). From this crossing, mice heterozygous for the β2KO and containing a copy of the Tet-β2 and NSE-tTA transgenes were bred together. The offspring obtained from this final cross and compared in this study were: β2KO mice with both transgenes, i.e. specific rescue in the VTA region (β2tr); β2 heterozygous mice with or without the transgenes (used as controls), and β2 KO mice with no transgenes or only a single transgene (β2KO). Mice were group housed up to five per cage. Temperature was maintained at 22°C and the light/dark cycle was 12:12 with lights on at 7:00 A.M. Food and water were available ad libitum. All animal procedures were carried out in strict accordance with the NIH Care and Use of Laboratory Animals Guidelines and were approved by the Yale Animal Care and Use Committee. Briefly, mice carrying a tetracycline operon-β2 transgene were crossed with NSE-tTA mice and subsequently with β2KO mice. “Control” mice (animals heterozygous for the β2KO allele with one of the transgenes) and β2KO mice were obtained from the same litters as β2tr mice. Upon weaning, mice were genotyped as previously described (King et al., 2003).
Naïve mice were injected with an overdose of chloral hydrate, and perfused intracardially with cold PBS (0.1 M, pH 7.3) followed by cold 4% PFA for 5-10 min each. Brains were post-fixed for 12 hr in PFA at 4°C, and then placed for 24 hr in 30% sucrose/PBS (0.1 M, pH 7.3). 40 μm sections were mounted onto microscope slides and processed for in situ hybridization with a β2 subunit riboprobe as described (King et al., 2003), exposed to a phosphor screen for 4 days and read using a phosphor-Imager (Cyclone Imager, Perkin Elmer).
Radioligand binding and immunohistochemistry was performed on brain sections from mice of Control, β2KO and β2tr genotypes to visualize assembled β2*nAChRs, β2α6*/β2α3* nAChRs, and levels of the α4 subunit. Multiple, parallel series of coronal tissue sections were cut from frozen brains, were thaw mounted onto glass slides and stored at -70°C until used. For β2*nAChR binding,15-μm fresh frozen sections were thawed to room temperature and incubated with 200 pM [125I]A-85380 (Perkin Elmer) for 45 min in cold 50 mM Tris-HCl, pH7.4, washed three times in the same buffer, rinsed with water, and exposed for 3 days (Controls) or 7 days (β2tr) on barium screens. Methods for [125I]α-CtxMII (β2α6*/β2α3*) binding and [125I]mAb299 (α4) immunohistochemistry were as previously described (Whiteaker et al., 2006, Whiteaker et al., 2000). 14 μm brain sections were incubated with either [125I]-α-CtxMII (0.5 nM) for 2h at 22 °C or with [125I]mAb299 (0.3 nM) for 48h at 4 °C in humidified chambers. In both cases PMSF, EDTA, EGTA, BSA, aprotinin, leupeptin trifluoroacetate, and pepstatin A were used to prevent degradation by endogeneous proteases. Sections were then exposed to Kodak MR film for 14 days.
Methods were as previously described for DA (Salminen et al., 2007, Salminen et al., 2004) and GABA release (King et al., 2003) with minor modifications. Briefly, mice were sacrificed by cervical dislocation, brains removed and placed on ice for immediate dissection of the entire striatum and olfactory tubercle. Mouse brain regions were independently homogenized in 0.5 ml cold isotonic sucrose (0.32 M) buffered with HEPES (5 mM) pH 7.5. Following 20 min of centrifugation at 12,000 X g, the resulting pellets were resuspended (1.6 ml/mouse brain region) in uptake buffer (128 mM NaCl, 2.4 mM KCl, 3.2 mM CaCl2, 1.2 mM KH2PO4, 1.2 mM MgSO4, 25 mM HEPES, pH 7.5, 10 mM glucose, with 1 mM ascorbic acid, and 0.01 mM pargyline added for DA). Synaptosomes were incubated at 37° C for 10 min before addition of 100nM [3H]DA or [3H]GABA and the suspension incubated for another 5-10 min. Aliquots (80μl) were placed on filter paper and perfused at room temperature with uptake buffer plus additions (DA: 0.1% BSA, 10 μM nomifensine, and 1 μM atropine; GABA: 0.1% BSA 1μM atropine, and 10 nM NC711) for 10 min at 0.7 ml/min. Release was stimulated by exposure (20s) to nicotine (10 μM in buffer for DA release and 30 μM for GABA release) or elevated potassium (20 mM substituting for 20 mM NaCl in buffer). 10s fractions were collected and radioactivity was determined by scintillation counting. Baseline measures were determined before and after stimulation and subtracted from each fraction. Units of release were summed fractions of 10% or more above baseline and expressed as fraction of baseline before normalization to Control animals.
For locomotor activation mice received 200 μg/mL nicotine bitartrate (Nicotine; 2.31 mM) in 2% saccharin or 2% saccharin and hydrogen bitartrate (Vehicle; 2.31 mM) in their drinking water (Brunzell et al., 2003, King et al., 2004).
Mice undergoing place conditioning and those for western blot studies received i.p. injections of drug (CREB studies: 0.09 mg/kg, 10 ml/kg in 0.9% saline; conditioned place preference studies: 0.03, 0.06, 0.09, 0.35 or 0.43 mg/kg, 10 ml/kg in 0.9% saline, or saline alone during the conditioning phase of the conditioned place preference). Previous pilot experiments in our lab in C57BL/6J mice helped us to determine the optimal nicotine dose (0.09 mg/kg) that induced nicotine place conditioning with our paradigm in mice (Brunzell et al., submitted).
Mice were single-housed three days before the beginning of the experiments. After this acclimation period, the water bottles were changed and each animal received Nicotine or Vehicle in their drinking water as their sole source of fluid. Subsequently, beam breaks for β2tr, β2KO, and Control mice were measured in 1 h time bins over 24 h for 7 days in the homecage.
The standard place preference apparatus (Med Associates, St. Albans, VT) was modified so mice showed no bias for one of the chambers. Two conditioning chambers (16.8 cm × 12.7 × 12.7 cm each) with retractable doors were separated by a neutral chamber (7.2 cm × 12.7 × 12.7 cm). The walls and the floors of the chambers varied but lighting was equilibrated so that there was no baseline preference for either chamber. Time spent in each chamber was recorded by beam breaks and calculated using Med-PC IV software (Med Associates, St. Albans, VT).
On the first day of the experiment (baseline; day 1), mice were placed in the neutral chamber and were allowed to explore the apparatus for 15 min; at the end of this session, mice were returned to their home cage. Over the next 3 days (conditioning; days 2-4), mice received an injection of saline (i.p.) in the morning and were then confined to one chamber for 30 min. In the afternoon, animals received an injection of nicotine (i.p.) and were placed in the opposite chamber. The choice of the conditioning chamber was pseudorandomly assigned and counterbalanced in order to test under unbiased conditions. On the day after training was completed (testing; day 4), mice were placed in the neutral compartment and allowed to visit both conditioning chambers without restrictions, for a total of 15 min. Baseline and testing sessions were carried out around noon, an intermediate time between the AM and PM conditioning sessions. Total time spent in each chamber was measured and changes from baseline preference were calculated.
Biochemical studies of CREB signaling were performed under similar conditions to place conditioning studies, but animals were conditioned to only a single chamber. Control animals were wild-type C57BL/6J mice. Separate “Saline” and “Nicotine” groups underwent place conditioning training. Mice in these groups received a saline or nicotine (0.09 mg/kg, i.p.) injection immediately before placement in a training chamber for 30 min. This treatment was repeated once daily for 3 days. Brains were harvested 15 min after the third injection. The “Paired-Chamber” group corresponded to those animals that were undergoing place conditioning testing. This group had the identical treatment to the “Nicotine” group above, however this group of mice then received a saline injection 24 h after the last conditioning session prior to placement in the nicotine-associated chamber. As with the other groups, brains were harvested 15 min after saline injection. All data were normalized to an appropriate genetic control group that received saline injections in the home-cage only.
Mouse brains were harvested following decapitation, placed in ice cold PBS, sectioned in a chilled matrix, and placed in fresh, chilled PBS. 16 gauge punches of NAc core and shell were collected from 1 mm sections and frozen in dry ice. Tissues and blots were processed as described previously (Brunzell et al., 2003). 10 μg of protein was loaded on a polyacrylamide gradient gel and transferred to nitrocellulose. Blots were incubated in polyclonal antisera (Cell Signaling) to CREB (1:1000) and pCREB (Ser 133; 1:500),were washed then incubated in 1:1000 peroxidase labeled anti-rabbit secondary (Vector) and detected using ECL and X-Ray film detection. Samples were normalized to GAPDH levels (1:10,000; Upstate Biotechnology).
For nicotine-dependent locomotor activation a 2 × 3 repeated measures ANOVA with between subject factors for drug treatment effects (Nicotine versus Vehicle) and genotype (β2KO, β2tr and Control) was used to identify differences in nicotine-dependent locomotor activity on Days 4 and 7 of oral Nicotine/Vehicle exposure. For place conditioning, significant effects of treatment and genotype were determined by testing the time spent in the conditioning chambers (time spent at baseline - time spent on the testing day) against the null-hypothesis (no difference between baseline and test) as within-subject measures, with between-subjects measures for genotype. Western blot optical densities were normalized to vehicle-treated controls to enable comparisons across blots. One-way ANOVAs were performed within genotype for experiments using a subchronic, rewarding dosing regimen of nicotine. A 2 × 2 (genotype × drug exposure) ANOVA was performed for control versus β2KO mice chronically drinking nicotine. Post-hoc two-tailed t-tests were performed for all statistics except where previous data predicted a directional finding to justify use of one-tailed t-tests. Mann Whitney U-tests were used when homogeneity of variance could not be assumed.
Expression of both the NSE-tTA and TetOp-β2 transgenes on a β2KO background (β2tr mice) resulted in expression of the β2 subunit mRNA restricted to the VTA/substantia nigra (SN) region in the midbrain (Fig. 1A) consistent with binding studies performed previously (King et al., 2003). β2tr mice also show nicotinic binding (reflecting assembled nAChRs) in the VTA/SN and striatum with preferential recovery of A85380 binding (specifically targeting β2*nAChRs) in the NAc region (Fig. 1B). Sparse binding was also observed in cortical regions, likely reflecting β2 nAChRs on the terminals of the VTA/SN projection neurons. The binding was much stronger in the VTA compared to the NAc, suggesting that α4/β2 nAChRs on the cell bodies of DA neurons in the VTA and on their terminals impinging onto the striatum were preferentially rescued in these mice. β2tr mice also showed recovery of both α4 (as measured by immunohistochemistry using mAb299; Fig. 1C) and α3/α6 subunits (as measured by binding of αCtxMII; Fig. 1D), indicating that the rescue was not selective for one subclass of β2* nAChRs. mAb299 labeling, an α4 specific antibody (Whiting & Lindstrom, 1988); Whiteaker et al, 2006), likely represents assembled α4/β2 nAChRs since no labeling is seen with this antibody in brain sections from β2 knockout mice. Similarly, αCtxMII binding in the VTA/SN and striatum likely represents assembled α6/β2 nAChRs since little α3 subunit is expressed in these neurons (Zoli et al., 1998). In addition, αCtxMII binding in β2tr mice is partially restored in the striatum and almost completely recovered in the nucleus accumbens, while, while no αCtxMII binding was seen in the β2KO further suggesting that α6/β2 nAChRs are assembled in these brain regions. Interestingly, there is also significant αCtxMII binding in regions that receive input from the retina, in which both the α3 and α6 subunits are highly expressed (Gotti et al, 2005; observed at similar levels in Control and β2tr animals) while β2KO do not show any binding. Expression did not completely recover to the level of Control subjects, however. A small amount of non-specific αCtxMII binding was evident in the hippocampal region of the β2KO slices.
There was an overall genotype difference on nicotine-induced dopamine release in the striatum (Normalized to Control animals (100%): KO = 7.89 ± 1.22; β2tr = 16.72 ± 4.16; F2,32 = 71.38, p < 0.0001) and the olfactory tubercules (Normalized to Control animals (100%): KO = 2.69 ± 0.51; β2tr = 11.75 ± 2.79; F2,32 = 92.69, p < 0.0001). Nicotine-stimulated DA release from striatal synaptosomes was significantly greater in β2tr than β2KO mice (z = -2.29, p = 0.0218), but only a partial rescue of this phenotype was observed in β2tr mice compared to Control animals (z = 3.83, p = 0.0001) (Fig. 2A). Similar findings were observed in olfactory tubercle (F2,32 = 92.69, p < 0.0001; Control vs. β2tr: z = 3.83, p = 0.0001; β2tr vs. β2KO: z = 3.42, p = 0.0006; Fig. 2A). Potassium-mediated DA release was similar across genotype in the striatum and olfactory tubercle, indicating that differences in DA release were specific to activity at β2*nAChRs (data not shown). Thus, the β2tr line of mice has partial recovery of functional β2*nAChRs in the VTA and NAc.
GABA release was also partially restored in the β2tr line of mice, with an overall genotype difference in nicotine-induced GABA release in the striatum (Normalized to Control animals (100%): KO = 9.44 ± 1.77; β2tr = 19.89 ± 4.25; F2,32 = 66.59, p < 0.0001) and the olfactory tubercule (Normalized to Control animals (100%): KO = 18.10 ± 3.83; β2tr = 26.13 ± 4.11; F2,30 = 67.40, p < 0.0001; Fig. 2B). Nicotine-stimulated GABA release from striatal synaptosomes was greater in β2tr mice compared to β2KO animals (z = -2.3, p = 0.0214; β2tr = 19.89% ± 4.26, β2KO = 9.44% ± 1.78 compared to Controls), but was significantly less than release measured in synaptosomes from Control animals (z = 3.81, p = 0.0001). Similar results were obtained in synaptosomes prepared from the olfactory tubercle (F2,30=33.70, p < 0.0001; Control vs. β2tr: z = 3.78, p = 0.0002), although the difference between the β2tr and β2KO mice did not reach significance (z = 1.18, p = 0.23). Potassium-mediated GABA release was similar across genotypes in these brain areas (not shown).
No changes in locomotor activity were observed after 4-days of nicotine exposure (Fig. 3A); whereas a significant genotype × treatment effect for nicotine-dependent locomotor activation reached significance by Day 7 (F2,251 = 12.06, p < 0.001; Fig.3B). Vehicle-exposed animals showed baseline levels of activity that did not vary with genotype and that were stable across the entire test period. As expected from previous studies in C57BL/6J mice (King et al., 2004), Control animals demonstrated nicotine-induced locomotor activation compared to vehicle-treated mice (t16 = 13.51, p < 0.001). Unlike β2KO mice that failed to show nicotine-dependent locomotor activation (t14 = 6.51, p = 0.544), β2tr mice showed a significant increase in locomotor activity in response to nicotine (t25 = 9.39, p = 0.015) that was similar to Control mice. VTA and striatal expression of β2*nAChRs was therefore sufficient to rescue locomotor sensitizing effects of nicotine. No differences in fluid intake were observed regardless of the treatment or the genotype (data not shown)
There was a genotype × treatment interaction for nicotine conditioned place preference (F2,26 = 3.72, p = 0.037) due to a significant increase in preference for the nicotine-paired chamber exhibited by Control mice (t7 = 4.05, p = 0.0049, Fig. 4A) at an optimal dose of nicotine that was not observed for β2KO or β2tr mice. Indeed, the β2tr mice showed an increase in time spent in the saline-paired compartment after nicotine administration. Control mice show a very steep dose-response curve for nicotine preference, with the optimal dose at 0.09 mg/kg (Fig. 4A) and some preference at 0.04 mg/kg, but no preference at other doses we have tested (not shown). To determine whether there might be a shift in the dose-response curve, β2tr mice were tested for place preference across a broad range of nicotine doses but no preference was seen at any dose tested (all p values > 0.10; Fig. 4B). Interestingly, β2tr mice treated with 0.09 mg/kg nicotine, the dose that is rewarding in control mice, showed a marked increase in time spent in the saline chamber following conditioning. While this may represent a nicotine-dependent place aversion, it was not possible to make this conclusion since the change came at the expense of time spent in the unpaired central compartment, rather than the nicotine-paired compartment.
Western blot analysis revealed that exposure to a novel chamber induced increases in total CREB levels in response to both saline and nicotine injections. There was a significant effect of experimental condition on levels of CREB (F4, 27 = 5.273, p = 0.004) and pCREB (F4, 27 = 3.236, p = 0.030) in the NAc shell of control mice; compared to homecage vehicle-injected animals. Exposure to a nicotine-paired chamber (Nicotine) increased both total CREB and pCREB levels in the NAc shell of control mice (p = 0.005; Fig. 5A). In contrast to control mice, and like β2KO mice, β2tr mice did not show saline- or nicotine-induced increases in total CREB in the NAc shell (F’s < 1.0; Fig. 5B and 5C). In fact, β2tr mice showed a significant reduction in total CREB in the NAc shell during nicotine conditioning when compared against vehicle-injected homecage Control animals (p = 0.01). Interestingly, these changes in CREB were seen in response to the training chamber and were not significantly different between training chamber-exposed Saline- and Nicotine-treated mice. Similar to Control animals, β2tr mice did show a significant increase in pCREB following exposure to the nicotine-paired chamber (Paired chamber), however (p = 0.003).
Our results indicate that limited expression of β2*nAChRs in the VTA region and on VTA terminals results in low levels of nicotine-elicited DA and GABA release in striatal synaptosomes from transgenic mice (β2tr). This was sufficient to rescue nicotine-dependent locomotor activation that is abolished in β2*nAChR knockout mice (β2KO), whereas nicotine-induced conditioned place preference was not restored. Biochemically, we observed that conditioning with nicotine in a training chamber increases total CREB levels in NAc shell of control, but not β2KO or β2tr, mice. β2tr mice showed diminished CREB in NAc shell when exposed to a nicotine-paired training chamber. Overall, these data suggest that there is a differential dependence on β2*nAChRs for nicotine-dependent psychostimulant effects than for nicotine’s rewarding effects.
In the β2tr mice examined here, the β2 transgene was rescued in both DA and GABA projection neurons from the VTA to the NAc. We have shown previously that nicotine-dependent locomotor activation in the paradigm used here is abolished by administration of a DA antagonist (King et al., 2004) but it is not known whether GABA signaling is also necessary for this behavior. Electrophysiological studies suggest that desensitization of β2* nAChRS on GABA neurons may contribute to the long time course of nicotine-elicited DA release, thus nAChRs on both populations of neurons may contribute to nicotine-induced locomotor activation. The existing data suggest that in β2tr mice, β2* nAChRs on DA neurons are likely to be necessary for the rescue of nicotine-dependent locomotor activation, however. The fact that this behavior develops over several days and is maintained as long as nicotine continues to be administered suggests that β2* nAChRs on DA neurons are critical for driving ongoing DA release in response to long-term nicotine exposure.
It is not clear why β2* nAChRs on both DA and GABA terminals were restored in the midbrain in β2tr mice, but this likely suggests that there is a common neuronal precursor population for both GABA and DA neurons that we have targeted with the insertion of these transgenes. DA innervation to the striatum is provided by terminals of the DA neurons emerging from the VTA and substantia nigra, but GABAergic projection neurons in the VTA and substantia nigra also innervate striatal regions (Ferreira et al., 2008, Ikemoto, 2007, Ikemoto et al., 2006). Less is known about the role of GABA projection neurons in the VTA and substantia nigra in behaviors related to drug abuse than about DA neurons, but it seems likely that these neurons are important for modulating behavioral responses to nicotine. While it is known that DA neurons are critical for various nicotine-induced behaviors including locomotor activation and sensitization, conditioned place preference and self-administration, stimulation of nAChRs on VTA GABA neurons could also have an influence on the rewarding effects of nicotine. It has been proposed that rapid desensitization of β2 nAChRs on GABA terminals disinhibits DA neurons and increases DA release in the nucleus accumbens, perhaps enhancing DA-dependent nicotine reinforcement (Mansvelder et al., 2002). Similarly, VTA GABA neurons may play a critical role in the motivational salience of drugs of abuse including nicotine since blockade of GABA neurons in the VTA can induce robust rewarding effects of a low dose of nicotine as measured by conditioned place-preference (Laviolette & Van Der Kooy, 2001), likely through a facilitation of DA release from VTA DA neurons. Thus, the limited rescue of β2* nAChRs on GABA terminals may limit the tonic inhibition of VTA/SN DA neurons that occurs when β2* nAChRs are expressed at higher levels in these neurons.
It was somewhat surprising that transgenic expression of β2*nAChRs in the VTA and on NAc DA terminals was sufficient to rescue nicotine-induced locomotion but not nicotine conditioned place preference, since previous studies have demonstrated that lentiviral-mediated expression of mesolimblic β2*nAChRs rescues intra-VTA self administration (Maskos et al., 2005). Several differences could explain this apparent discrepancy. Perhaps most likely, viral rescue leads to a more complete recovery of nicotine-stimulated DA activity than what we observed in β2tr mice. Multiple subtypes of neurons with cell bodies or terminals in the VTA can be modulated by nicotine administration, and the level of nAChR stimulation may be important for the balance between nicotine-mediated DA and GABA release, as well as for the rewarding and stimulating effects of nicotine. Incomplete recovery of the β2 nAChR subunit likely favors the low affinity α43 β22 stoichiometry (Zwart & Vijverberg, 1998), further limiting neuronal stimulation by nicotine. In fact, low levels of DA release observed in β2tr animals may simulate effects of partial agonism at β2*nAChRs, similar to the administration of partial agonists including the smoking cessation drug, varenicline (Coe et al., 2005). Interestingly, central infusion of a β2*nAChR partial agonist results in locomotor activation and this effect is abolished by NAc 6-hydroxydopamine lesions (Clarke, 1990), in line with our results. However, data obtained in rats have also demonstrated that the partial agonist cytisine could induce conditioned place preference when infused into the VTA (Museo & Wise, 1994).
It is also possible that re-expression of β2*nAChRs in β2tr mice is mainly localized on cell bodies of the VTA/DA neurons and their projections and that β2*nAChRs on neurons that project to the VTA, such those in the pedonculonpontine tegmental nucleus (PPTg), are critical for nicotine-induced reward. The importance of these projections to nicotine reward is supported by studies demonstrating that lesions of the PPTg decrease nicotine self-administration (Corrigall et al., 1994) and can even induce nicotine-induced aversion to doses that are usually sufficient for nicotine place preference (Laviolette et al., 2002).
Another consideration is that self-administration and conditioned place preference likely measure different aspects of drug response as has been seen in studies of other mouse lines such as those lacking dopamine D1 receptors (see (Caine et al., 2007) and (Karasinska et al., 2005)). In these studies, knockout of D1 receptors had little effect on cocaine place preference but greatly decreased cocaine self-administration. This further suggests that there may be a dissociation between the circuits involved in conditioned place preference and those related to self-administration. It is worth noting, however that there are likely to be differences in the neuronal circuitry involved in the effects of cocaine and nicotine as well. For instance, studies on fosB expression emphasize these potential differences (Zhu et al., 2007).
Another important difference is the route of administration in the viral and transgenic β2 nAChR rescue studies. The current experiments on place preference in transgenic mice were carried out with systemic nicotine administration, whereas Maskos and colleagues used intra-VTA self-administration in the viral vector studies. Also, the paradigms used in the two studies (conditioned place preference and self-administration) involve very different learning processes (pavlovian condtioning, contextual memory, sensory processing, etc.) which likely involve both overlapping and non-overlapping neuronal pathways. Few studies have been performed to determine whether VTA administration is more effective than systemic injection in stimulating behavioral responses to nicotine, but the localized stimulation of VTA neurons would likely limit potential aversive/inhibiting effects of nicotine thought to take place in the periphery and/or in VTA projection areas (Laviolette et al., 2002; Laviolette & Van Der Kooy, 2001).
In addition to measuring behavioral responses to nicotine, we were interested in whether downstream intracellular signaling related to neuronal plasticity was rescued in parallel with nicotine-dependent locomotor activation. The transcription factor CREB has been shown to be involved in the rewarding effects of nicotine (Walters et al., 2005). In the current study, regulation of total CREB levels was observed in control mice, but not β2KO or β2tr mice, following a nicotine dosing regimen that leads to conditioned place preference. This paralleled the behavioral findings that 0.09 mg/kg nicotine supported conditioned place preference in Control, but not β2KO or β2tr, mice, suggesting that β2*nAChR-dependent increases in total CREB levels in the NAc shell may contribute to some aspects of nicotine reward as measured by nicotine CPP, but are not essential for nicotine-dependent locomotor activity. In contrast, β2tr mice did show induction of pCREB when re-exposed to a chamber previously paired with nicotine. Phosphorylation of serine 133 (pCREB) is a marker of CREB activation. These data suggest that context-induced activation of pCREB in the NAc shell is not sufficient to induce nicotine seeking.
Reductions in NAc pCREB appear to enhance psychostimulant-induced locomotor sensitization (Dong et al., 2006). Consistent with this possibility, control mice show both locomotor activation and reductions in NAc pCREB following chronic nicotine exposure (Brunzell et al., 2003, King et al., 2004). Thus, nicotine-dependent regulation of DA release appears to be sufficient for locomotor activation, but DA release alone does not support reward learning or associated regulation of CREB in β2tr mice.
In summary, these studies demonstrate the dissociation between the psychostimulant effects of nicotine and its ability to induce conditioned place preference. A low level of β2* nAChR expression in VTA neurons is sufficient for nicotine-induced locomotor activation but does not support nicotine reward, potentially as a result of a shift in the balance of nicotine-induced DA vs. GABA stimulation in the mesolimbic circuitry. Alternatively, the lack of high affinity nAChRs in other brain areas, particularly those with afferent inputs into the VTA, could also contribute to the blunting of nicotine conditoned place preference. Taken together, this further suggests that partial activation of high affinity nAChRs might block the rewarding effects of nicotine, providing a molecular mechanism that could explain the ability of partial agonists such as cytisine and varenicline to aid in smoking cessation.
This work was supported by the State of Connecticut, Department of Mental Health and Addiction Services, National Institutes of Health grants AA15632, DA14241, DA10455, DA00436 to MRP, DA003194 to MJM and SRG, DA012242 to JMM, MJM and SRG, MH53631, GM48677 to JMM, NS11323 to JML and a Lieber NARSAD Young Investigator Award to DHB.