Introduction

Approximately 1.1 billion people use tobacco worldwide and it is the leading preventable cause of death. Each year, the global tobacco epidemic kills nearly 6 million people (http://www.who.int/tobacco). Nicotine is the major addictive component of tobacco and exerts its effects by binding to nicotinic acetylcholine receptors (nAChRs) (Benowitz 2010). Neuronal nAChRs are pentameric ligand-gated ion channels that mediate the addictive properties of nicotine (Greenbaum and Lerer 2009) and contribute to various neuropsychiatric disorders (Gotti and Clementi 2004; Greenbaum and Lerer 2009). To date, eight alpha (α2–α7 and α9–α10) and three beta (β2–β4) subunits have been localized in the mammalian nervous system. Neuronal nAChRs can be homopentamers or heteropentamers and their subunit composition and stoichiometry confer their pharmacological and functional properties (Albuquerque et al. 2009; Gotti et al. 2009). The high-affinity α4β2-nAChR is the most abundant receptor in the brain and the principal mediator of nicotine dependence (Maskos et al. 2005; Mineur and Picciotto 2008; Tapper et al. 2004). Until recently, most research on the role of nAChRs in nicotine dependence has focused on these subunits. This has recently changed and new studies (mainly human genetics) have shown the importance of the CHRNA5/A3/B4 cluster in nicotine dependence phenotypes (Baker et al. 2009; Berrettini et al. 2008; Bierut et al. 2008; Doyle et al. 2011; Li et al. 2010a, b; Saccone et al. 2009a, b; Schlaepfer et al. 2008; Wang et al. 2009). Evidence for association of genetic variation in the cluster with nicotine dependence-related phenotypes is more robust and convincing than for any other nAChR gene. Although the most plausible evidence for a risk variant is for the non-synonymous SNP (rs16969968) in the CHRNA5 gene, multiple SNPs in the CHRNA5/A3/B4 gene cluster are reported to influence the transition from smoking to nicotine dependence (Greenbaum and Lerer 2009; Saccone et al. 2009b). Some of these SNPs are located in non-coding sequences positioned in the β4 3′-untranslated region (McDonough and Deneris 1997; Schlaepfer et al. 2008) or in CHRNA5 and CHRNA3 promoter regions (Doyle et al. 2011), suggesting that rather than influence the functionality of a specific subunit these SNPs may alter the regulation of the entire cluster. Indeed, some genetic variants described in humans give rise to overexpression of the genomically clustered genes, found in postmortem brain and lung cancer, suggesting their coordinate regulation (Falvella et al. 2009; Wang et al. 2009). Previous studies in mouse have been carried out by knocking out or overexpressing individual nAChR subunits. These studies have demonstrated the role of α5 and β4 in fear-related memories (Zhu et al. 2005), anxiety (Kedmi and Orr-Urtreger 2007; Salas et al. 2003b), nicotine withdrawal (Salas et al. 2004), and nicotine intake (Fowler et al. 2011; Frahm et al. 2011), but currently there are no data regarding the role of the entire cluster in the reinforcing properties of nicotine.

To this end we have generated a BAC transgenic mouse model (TgCHRNA5/A3/B4) that overexpresses the human CHRNA5/A3/B4 cluster. These transgenic mice have increased epibatidine-binding sites and increased sensitivity to the pharmacological effects of nicotine. Moreover, overexpression of the CHRNA5/A3/B4 cluster creates a behavioral profile prone to nicotine consumption.

Methods

Ethics statement

All animal work has been conducted according to relevant national and international guidelines. Experimental procedures were approved by the Ethical Committee of Animal Experimentation of the PRBB (Comitè Ètic d’Experimentació Animal del Parc de Recerca Biomèdica de Barcelona, CEEA-PRBB, Catalan law 5/1995 and EU directives 86/609/2001-486) and met the Standards for Use of Laboratory Animals no A5388-01 (NIH). The local approval numbers are RML-09-1175-MDS MDS-09-1165.

Generation and genotyping of transgenic mice

We used a human BAC-clone (NCBI) containing the nicotinic receptor cluster. The presence of all promoter regions was assessed by PCR on maxiprep-extracted DNA from the RP11-335K5 (AC067863) clone, and rearrangements within the BAC were checked [EagI (BshTI) restriction pattern]. DNA plugs were digested using EagI and electrophoresed by Pulsed-Field Gel Electrophoresis. Transgenic mice were generated by standard pronucleus microinjection of the 111-kb fragment containing the three nicotinic receptor subunits (α5, α3, and β4) on a hybrid B6/SJL-F1 J genetic background (Nucleis, France). Genotyping was performed routinely by PCR analysis using two primer pairs: CHRNB4-F 5′-GAGCCAAGATCCCACCACTC-3′ and CHRNB4-R 5′-CCAGGCATCCGGATTTGTAT-3′ and CHRNA5-F 5′-GAAAGACTTGAGTGGGCAGC-3′ and CHRNA5-R 5′-CAACCCTGTCTGTCTCTAGC-3′. The presence of the transgene was tested on tail DNA by slot-blot analysis using a 342-bp 32P-labeled fragment against intronic sequence CHRNA3-CHRNB4 inserted in the RP11-335K05 BAC (using the primer pairs F 5′-GGGGTGGGTGGAAGTTAGTT-3′ and R 5′-TACACCGCTTCCTTTCAACC-3′) as a probe [Slot Blot Manifold, Hoefer (PR 648)]. DNA was quantified by Hoechst 33258 dye (Sigma-Aldrich Chemie, Germany) using a calf thymus DNA standard curve (Sigma-Aldrich, USA) in a multi-label microplate luminescence reader (TECAN GmbH Infinite TM M200, Austria). Hybrid founders were crossed using B6/SJL-F1J females (F1–F5) to obtain transgenic mice. Non-transgenic (WT) littermates served as controls. Mice were housed 3–5 per cage in standard macrolon cages (40 × 25 × 20 cm) under a 12-h light/dark schedule.

mRNA expression analysis

For analysis of the human CHRNA5, CHRNA3, and CHRNB4 gene expression, total RNA from cortex, hippocampus and cerebellum of WT and TgCHRNA5/A3/B4 mice was extracted using TRIzol (Invitrogen, UK) and analyzed by RT-PCR. cDNA was synthesized from 1 μg of total RNA using Omniscript® RT Kit (Qiagen, Germany). The cDNA solution was subjected to 35 cycles of PCR amplification using specific primers against the transgene: CHRNA3-F 5′-TGAGCACCGTCTATTTGAGC-3′ and CHRNA3-R 5′-ATGAACTCTGCCCCACCA-3′; CHRNB4-F 5′-GCTCTACCCCGGTGGCTAT-3′ and CHRNB4-R 5′-GTAGGGCCCCTCAGAAGC-3′; CHRNA5-F 5′-AGAAAGAGGAAACTGAGAGTGGT-3′ and CHRNA5-R 5′-AAAGCCCAAGAGATCCAACA-3′. Quantitative-PCR was performed in duplicates with cDNA (TgCHRNA5/A3/B4 n = 12; WT n = 12) using the LightCycler® 480 SYBR Green I Master kit (Roche Diagnostics GmbH, Germany) in a LightCycler® 480 Real-Time PCR (Roche Diagnostics, Germany). 45 cycles of PCR amplification were run using specific primers for endogenous nAChR subunits: Chrna3-F 5′-AGCACCGCCTGTTCCAGT-3′ and Chrna3-R 5′-CATGAACTCCACCCCTTGG-3′; Chrnb4-F 5′-GCAGTCAGCTCCCACACG-3′ and Chrnb4-R 5′-TAGCCTAGAGGCCCTTGGAG-3′; Chrna5-F 5′-AGAAGAAGCCGAGAAAGACG-3′ and Chrna5-R 5′-AGCCCTAGCGTCCCAATG-3′. Pgk1 was used as reference gene. Duplicates with SD ≥0.38 were excluded and re-amplified. ΔΔCt method [Relative Expression Software Tool (REST)] was used to quantify the relative amount of mRNA in comparison with control samples and 2−ΔΔCt was transformed into logarithmic scale (log2 ratio).

Protein expression analysis

Cortex and hippocampus extracts from adult (3 month) TgCHRNA5/A3/B4 (n = 6) and WT (n = 6) mice were prepared in ice-cold RIPA buffer (1% Nonidet P-40, 0.5% sodium deoxycholate, and 0.1% SDS in PBS 0.1 M, pH 7.4). Samples were separated on a 10% SDS-polyacrylamide gel and transferred to a PVDF membrane (Bio-Rad, USA), blocked and incubated with anti-nAChRα3 (1:200, Santa Cruz, USA) anti-nAChRα5 (1:200, Santa Cruz, USA) and anti-nAChRβ4 (1:200, Abcam, UK) antibodies at 4ºC overnight. Specific secondary IgG/HRP (1:2000, Dako Denmark) was followed by ECL assay (Pierce, USA). Tubulin was used as internal standard (1:2000, Upstate, USA).

Immunohistochemistry

Mice were habituated to the injection protocol 3 days prior to the experiment. For each experiment two groups per genotype (n = 8) were injected subcutaneously (0.1 mL/10 g weight) with saline or nicotine tartrate (Sigma-Aldrich, USA). 120 min after injection mice were euthanized with isofluorane and perfused with 50 mL of PBS 0.1 M and pH 7.4, followed by 150 mL of chilled 4% paraformaldehyde (PFA) (Sigma-Aldrich, USA). To check the activation of VTA dopaminergic neurons, we used a non-aversive dose of 0.5 mg/kg of nicotine, previously described in the literature to induce conditioned place preference in mice and increase the firing rate and burst firing of dopaminergic neurons. Serial coronal frozen sections (30 μm; VTA Bregma −2.92 and −4.04 mm) were incubated with Tyrosine Hydroxylase (TH, 1:4000; Sigma, USA) and c-Fos (1:500, Santa Cruz, USA) antibodies in 1% FBS/PBS overnight at 4ºC, followed by incubation with secondary fluorescent-labeled antibodies (1:300, goat anti-rabbit Alexa® Fluor 488 and goat anti-mouse Alexa® Fluor 594; Molecular Probes, USA) at room temperature 1 h in dark The same dose of 0.5 mg/kg of nicotine was used to check the activation of the MHb in serial coronal frozen sections (30 μm; MHb Bregma −0.82 and −2.30 mm) incubated with c-Fos primary antibody. The color reaction was developed with a biotinylated secondary antibody, addition of avidin/biotinylated complex followed by incubation with 3,3-diamio-benzide-tetra-hydrochloride (DAB) solution, and counterstained with hematoxylin to delimitate the area of the MHb. Since the dose of 0.5 mg/kg of nicotine was not able to significantly activate c-Fos expression in the MHb of any genotype, we further examined if the MHb/IPN circuit was altered in our transgenic mice using a dose of nicotine of 6 mg/kg. Serial coronal frozen sections (30 μm; MHb Bregma −0.82 and −2.30 mm) were incubated with c-Fos primary antibody following the above-mentioned protocol and using a secondary fluorescent-labeled antibody (1:300, goat anti-rabbit Alexa® Fluor 488; Molecular Probes, USA). Hoechst staining was used to delimitate the area of the MHb. The number of positive neurons was counted under fluorescence/bright microscope at 40× and the area of the MHb was delimitated using the ImageJ Software.

Receptor-binding autoradiography

Mice (n = 3 per genotype) were killed by cervical dislocation; the brain was rapidly (<1 min) removed from the skull and quickly frozen by immersion in isopentane (−35°C) for 10 s, and stored at −70°C until sectioning. Coronal sections (14 μm thick) were obtained using a Leica CM 1850 cryostat/microtome (Leica, Germany) and thaw-mounted on Fisher Suprafrost/Plus microscope slides. A series of ten sets of sections was prepared from each brain to allow comparison of results for several different experiments on adjacent or near-adjacent sections. Slides containing the brain sections were desiccated and stored, at −70°C until use.

[125I]Epibatidine autoradiography

Slides were warmed to room temperature and subsequently transferred to Bel-Art slide racks and rehydrated by incubation at 22°C for 15 min in isotonic buffer (NaCl, 144 mM; KCl, 2.2 mM, CaCl2, 2.0 mM, MgSO4, 1.0 mM; HEPES, 25 mM; pH = 7.5). The racks containing the rehydrated slides were subsequently transferred to the isotonic buffer containing 200 pM [125I]-epibatidine. [125I]-Epibatidine (Perkin-Elmer NEN, Boston, MA, original specific activity 2200 Ci/mmol) was mixed with unlabeled I-epibatidine to yield a final specific activity of 110 Ci/mmol (20-fold dilution). Samples were incubated for 2 h at 22°. A second set of incubations with [125I]-epibatidine also included 10 nM 3-((2S-azetidinylmethoxy)-5-iodo)-pyridine (5I-A-85380, Tocris Bioscience, Ellisville, MO) to selectively inhibit binding to β2*-nAChRs (Mukhin et al. 2000) such that the residual signal represented binding of β4*-nAChRs. Following the incubation the slides were redistributed to slide racks containing 25 slides and washed as follows (all solutions at 4°C): Twice for 30 s in isotonic buffer, twice for 5 s in hypotonic buffer (0.1×) and twice for 5 s in 10 mM HEPES, pH = 7.5. The samples were then dried and stored desiccated at room temperature in vacuum overnight before exposure to Packard Super Resolution Cyclone Storage Phosphor Screens to yield images for quantitation and subsequently to Kodak MR autoradiography film to yield higher resolution images for photography. Each Phosphor Screen was also simultaneously exposed to a series of tissue paste standards containing measured amounts of 125I to allow quantitation of the image intensity. Images were captured using a Cyclone Storage Phosphor Screen (Perkin-Elmer).

Quantitation

Tissue paste samples prepared from whole brain homogenates were used to construct standard curves. The Phosphor screens yield a linear relationship between signal intensity and tissue radioactivity content over several orders of magnitude. The regression line calculated for the standard curve was used to convert the measure value of pixels/mm2 to the cpm/mg wet weight from which signal intensity in fmol/mg wet weight was estimated from the specific activity of each ligand. Brain regions were identified using the mouse brain atlas (Franklin and Paxinos 1997). Several measurements were made in each brain region of each mouse and the average of these measurements defined the signal intensity for each region.

Behavioral characterization

Male TgCHRNA5/A3/B4 and WT littermates (2–6 months of age; see number of animals per each experiment below) from 6–8 different litters were used. 30 mice (15 per genotype) were tested at 2 months of age on behavioral measurements of gait, balance, reflexes, and startle response using SHIRPA protocol (Nolan et al. 2000). Behaviors were scored to provide a semi-quantitative assessment. In the open-field apparatus (70 × 70 × 60 cm) (Altafaj et al. 2001; Escorihuela et al. 1995), 14–18 mice per genotype were individually placed in the center of the field (500 Lux; 5 min) and total distance travelled, speed, time in each zone and number of rearing, grooming and fecal boli (Maccarrone et al. 2002) were recorded (SMART, Panlab, Spain). Locomotor activity was evaluated in activity cages (LE 881 Panlab, Spain; 45 × 45 cm) for 10 min (Actitrack software, Panlab, Spain). Randomly assigned groups of TgCHRNA5/A3/B4 and WT (n = 16 per treatment and genotype) were injected with saline or nicotine (0.5 or 1 mg/kg s.c.). Five minutes after injection mice were placed in the activity cages (20–25 Lux). For nicotine-induced seizures mice received either nicotine tartrate (Sigma-Aldrich, USA dissolved in 0.9% NaCl; n = 24) or saline (s.c. 0.1 mL/10 g of body weight, n = 22). We used different nicotine doses per genotype (WT: 12, 25 and 50 mg/kg vs. TgCHRNA3/B4/A5: 4, 6 and 12 mg/kg). After injection, behavior was recorded for 5 min. The effects of nicotine were scored from 0 to 7: 0, no effects; 1, locomotor alterations; 2, tachypnea and tremors; 3, back arching; 4 myoclonic seizures; 5, clonic seizures; 6, tonic seizures; and 7, death. Concentration–response curves were fitted with a logistic curve to determine the ED50.

Nicotine self-administration

Adult mice, WT (n = 14) and TgCHRNA5/A3/B4 (n = 16) were implanted with a cannula into the right jugular vein. Nicotine self-administration sessions (Soria et al. 2005) started 3 days after surgery. Responding was maintained by nicotine (0.015 mg/kg per infusion) delivered in 23.5 μl over 2 s. Nicotine was infused via a syringe mounted on a microinfusion pump (PHM-100A, Med-Associates, USA) connected to a single-channel liquid swivel (375/25, Instech Laboratories, USA). Two hours daily self-administration sessions were conducted 6 days per week. The house light was “on” at the beginning of the session for 3 s and then remained “off”. Each daily session started with a priming injection of the drug. Mice were trained under a fixed ratio 1 (FR1) schedule of reinforcement. A 10-s time-out period was established after reinforcement. Responses on inactive hole and all responses during the 10-s time-out period were recorded. Session terminated after 50 reinforcers were delivered or after 2 h. The criteria for acquisition was achieved when mice maintained a stable response with less than 20% deviation from the mean of the total number of reinforcers earned in three consecutive sessions (80% of stability), with at least 65% responding on the active hole, and a minimum of four reinforcers per session. Once achieved acquisition, mice were moved from FR1 to progressive ratio (PR) with response requirement to earn an injection escalating according to the following series: 1-2-3-5-12-18-27-40-60-90-135-200-300-450-675-1000. The PR session lasted for 3 h or until mice did not complete the ratio for delivery of one reinforcer within 1 h.

Data analysis

Parametric data are reported as mean ± standard error of mean (SEM). Whenever differences were observed per transgenic line, data were analyzed separately. For the open field we used Student’s t test for comparison between genotypes. In the SHIRPA protocol, we used the Fisher non-parametric test since data did not fit parametric assumptions. Data analysis for nicotine effect was performed using non-linear regression analysis and two-way ANOVA to analyze genotype X treatment interaction. qRT-PCR results were analyzed using Pair Wise Reallocation Randomisation Test©. For the binding assays and western blot, values were compared by one-way ANOVA, using Tukey HDS post hoc test. Nicotine self-administration was analyzed using repeated measures analysis. Student’s t test was used to analyze differences in number of nose-pokes between the active and inactive hole, and differences between genotypes for the progressive ratio and day of criteria acquisition. The statistical analysis was performed using the SPSS 12.0 software, except for the qRT-PCR analysis that was performed using Relative Expression Software Tool—384 (REST-384© v.2).

Results

Generation of TgCHRNA5/A3/B4 mice

For the generation of TgCHRNA5/A3/B4 mice we used a BAC inserted with the human cluster (Fig. 1a). Two lines were generated carrying between 16–18 (Tg22) and 4–5 copies (Tg30) of the transgene, respectively, as detected by Slot-blot (Fig. 1b). Similar mRNA expression of the α3, β4, and α5 subunits was confirmed in both lines by RT-PCR (Fig. 1c). Western blot analysis (Fig. 1d, e) confirmed increased expression of the β4 subunit in transgenic hippocampus (F (2,10) = 4.90, p < 0.05, one-way ANOVA), while expression of the α3 and α5 subunits was similar to WT. Conversely, in the cerebral cortex a significant increase of the α3 subunit was observed in both transgenic lines compared with WT (p < 0.05, one-way ANOVA).

Fig. 1
figure 1

Generation and general characterization of TgCHRNA5/A3/B4 mice. a Vector pBACe3.6 information carrying the Homo sapiens chromosome 15; clone RP11-335K5. b Slot-blot analysis of the transgene. Control of copy number is shown from 1 to 80 copies of the transgene. Line 22 and line 30 show 16–18 and 4–5 copies of the transgene, respectively. c Reverse transcriptase PCR of CHRNA5, CHRNA3, and CHRNB4. Note that cDNA control of all samples was performed hybridizing cDNA and DNA samples with Gdx primers. d, e Representative Western blot showing CHRNA5, CHRNA3, and CHRNB4 immunolabeling of WT (open bars) and TgCHRNA5/A3/B4 (filled bars) mice hippocampus and cortex. Tubulin was used as loading control. Relative immunoreactivity was determined densitometrically. Ratios are expressed as mean percentage of control (WT) ± SEM. *p < 0.05, **p < 0.01 comparison between genotypes, one-way ANOVA

Full size image

We investigated the regulation of the endogenous Chrna5, Chrna3, and Chrnb4 genes. qRT-PCR analysis (Table 1) with probes designed to detect mouse receptor subtypes did not show significant differences between genotypes in endogenous nAChR subunits expression in the cerebral cortex. However, TgCHRNA5/A3/B4 mice showed a reduction in the Chrna3 (−42.8%, p < 0.01) expression in the hippocampus, without differences in the Chrnb4 and Chrna5 mRNA levels with respect to WT.

Table 1 Expression of the endogenous nAChR subunits by quantitative RT-PCR analysis in TgCHRNA5/A3/B4 versus WT mice
Full size table

[125I]-epibatidine binding sites

Images for [125I]-epibatidine binding to sections from WT and transgenic mice at five anatomical levels are shown in Fig. 2. This figure includes images from near adjacent section illustrating total [125I]-epibatidine binding and binding in the presence of 5I-A-85380, in which labeling of β2*-nAChR sites have been selectively inhibited. Both total binding and that in the presence of 5I-A-85380 indicated that there is no ectopic expression in the transgenic mice. Significant increases in total [125I]-epibatidine binding were detected in olfactory bulb, CA1 region of hippocampus, superficial gray area of the superior colliculus and pyriform cortex (Table 2). Subsequent analysis of the [125I]-epibatidine binding in the presence of 5I-A-85380, where labeling is restricted to β4*-nAChR binding sites, allowed the detection of additional sites. Significant increases in 5I-A-85380-resistant [125I]-epibatidine binding sites were noted in olfactory bulb, medial habenula, CA1 region of hippocampus, superior colliculus (both superficial gray and deeper layers), dorsal tegmental nucleus, and pyriform cortex. Although there was a tendency for increased binding in fasciculus retroflexus and interpeduncular nucleus, these changes were not statistically significant. In addition, lower levels of 5I-A-85380-resistant [125I]-epibatidine binding were found in the thalamus, cerebral cortex, hippocampus, and striatum, being slightly increased in transgenic mice but reaching statistically significance in the hippocampus and striatum.

Fig. 2
figure 2

Increased [125I]-epibatidine binding sites in TgCHRNA5/A3/B4 mice. [125I]-epibatidine binding in sections from WT and TgCHRNA5/A3/B4 mice at five anatomical levels. Columns on the left side show α3β4* and α4β2* binding sites. Columns on the right side show α3β4* specific binding sites. Mi mitral cells, Gi granule layer, Pyr pyriform cortex, mHb medial habenulla, CA1 CA1 region of the hippocampus, SC superior colliculus, IPN interpeduncular nucleus, DTN dorsal tegmental nucleus

Full size image
Table 2 Brain regions of increased [125I]-epibatidine binding sites
Full size table

General phenotypic characterization

In the behavioral/neurological experiments we first ensured that the cluster overexpression had no consequences on the somatometric, neurologic or gross sensorimotor aspects (Table 3). TgCHRNA5/A3/B4 mice showed significantly reduced sensitivity in the toe-pinch test (p = 0.01, Fisher Test) and performed better the balance test (p < 0.05, Fisher Test). In the open field, transgenic mice spent less time in the center (t (1,29) = 2.52, p < 0.05, Student’s t test; Fig. 3a) and remained in the periphery (t (1,29) = −2.52, p < 0.05; Fig. 3a). Besides, TgCHRNA5/A3/B4 mice showed increased speed when crossing the center of the apparatus compared with WT mice (t (1,29) = −2.69, p = 0.01; Fig. 3b). Finally, the latency to enter the periphery was reduced in TgCHRNA5/A3/B4 mice (t (1,29) = 2.72, p = 0.01; Fig. 3c). In pharmacological studies, increasing doses of nicotine progressively decreased locomotion, with genotype-dependent differences (Two-way ANOVA genotype × treatment F (4,153) = 6.34, p = 0.0001; Fig. 3d). At 1 mg/kg, the effect of nicotine was significantly more marked in TgCHRNA5/A3/B4 mice (t (1,32) = 3.26, p < 0.01, Student’s t test). For nicotine-induced seizure studies nicotine administration induced dose-dependent seizures in both genotypes, but more marked in TgCHRNA3/B4/A5 (two-way ANOVA genotype × treatment F (2,62) = 9.59, p = 0.00). Dose–response curve was shifted to the left (Fig. 4a) and ED50 was reduced (26.33 ± 1.15 in WT vs. 6.5 ± 1.23 in TgCHRNA3/B4/A5 F (1,4) = 12.02, p < 0.05, one-way ANOVA, Fig. 4b).

Table 3 General phenotypic characterization (SHIRPA)
Full size table
Fig. 3
figure 3

Behavioral characterization of TgCHRNA5/A3/B4 mice. a–c Open field test. Graphs show time spent (a) and speed (b) in the aversive (centre) and non-aversive (periphery) zones, and latency to periphery (c). Nicotine effects on locomotor activity (d). Distance traveled was registered during 10 min after injection of saline, 0.5 and 1 mg/kg of nicotine. Data are expressed as mean ± SEM. *p < 0.05, **p < 0.01 comparison between genotypes, Student’s t test

Full size image
Fig. 4
figure 4

Nicotine-induced seizures. a Dose–response curve expressed as the seizures score in TgCHRNA3/B4/A5 (gray square) and WT (black dot) mice administered with increasing doses of nicotine (one dose per animal). Dose–response curve was shifted to the left in TgCHRNA3/B4/A5 as compared with WT mice denoting hypersensitivity to nicotine-induced seizures. b Median effective dose (ED50) was reduced in TgCHRNA3/B4/A5 (filled bars) being three times more sensitive to nicotine than WT (open bars) mice. Data are expressed as mean ± SEM. *p < 0.05 comparison between genotypes and treatment, one-way ANOVA

Full size image

Nicotine self-administration

Repeated measures analysis showed main effects of active versus inactive nose-pokes in WT (F (1,22) = 4.55, p < 0.05; Fig. 5a) and TgCHRNA5/A3/B4 mice (F (1,30) = 9.018, p < 0.01; Fig. 5b) along days. WT mice discriminated the active and inactive holes on the seventh day of acquisition (t (1,27) = 2.08, p = 0.05, Student’s t test), whereas TgCHRNA5/A3/B4 mice started to discriminate the two holes since the first day (t (1,30) = 2.13, p < 0.05, Student’s t test). Four WT and five TgCHRNA5/A3/B4 mice (28.57 and 31.25%, respectively) achieved the acquisition criteria and were moved to the progressive ratio test (Fig. 5c). TgCHRNA5/A3/B4 mice reached the FR1 stability criteria before WT littermates (6.60 ± 1.29 vs. 9.50 ± 0.87, respectively), but not significantly. Progressive ratio analysis revealed that TgCHRNA5/A3/B4 mice reached a higher breaking point than the WT (t (1,7) = −3.11, p < 0.05, Student’s t test), indicating enhanced motivation of transgenic mice to seek for nicotine.

Fig. 5
figure 5

Nicotine self-administration in WT and TgCHRNA5/A3/B4 mice. Number of nose-pokes in the active (black dots) and inactive hole (gray squares) for WT (a) and TgCHRNA5/A3/B4 (b) mice. Mice were trained daily in 2-h sessions to obtain nicotine (0.015 mg/kg per infusion) for 10 days under a fixed ratio 1 schedule of reinforcement. Data are expressed as mean ± SEM. *p < 0.05, **p < 0.01 comparison between holes, Student’s t test. c Progressive ratio of WT (open bar) and TgCHRNA5/A3/B4 (filled bar) mice that reached the criteria of acquisition. Data are expressed as mean ± SEM. *p < 0.05 comparison between genotypes, Student’s t test

Full size image

Nicotine-induced c-Fos expression

In the VTA the number of c-Fos/TH double-positive cells was quantified in mice receiving an acute injection of either saline or nicotine (0.5 mg/kg s.c.; Fig. 6a, b). Both WT and TgCHRNA5/A3/B4 mice showed increased c-Fos/TH-positive cells after nicotine injection as compared with saline (WT; t (1,5) = 0.32, p < 0.001, and TgCHRNA5/A3/B4; t (1,6) = 1.55, p = 0.01, Student’s t test), but the effects of nicotine were significantly different between genotypes (genotype × treatment F (3,15) = 23.50, p < 0.001, two-way ANOVA), hence nicotine-induced VTA c-Fos expression was significantly reduced in TgCHRNA5/A3/B4 mice as compared with WT littermates (t (1,5) = 0.58, p < 0.001, Student’s t test). The same dose of nicotine of 0.5 mg/kg tended to increase c-Fos expression in the MHb of both genotypes (Fig. 7a) although it was not statistically significant [WT; (t (1,5) = −1.91, NS, Student’s t test), TgCHRNA5/A3/B4; (t (1,8) = −1.94, NS, Student’s t test)]. Nicotine produced the same effect in both genotypes (genotype x treatment F (1,17) = 0.28, NS, two-way ANOVA); however, there was a tendency to be higher c-Fos expression in transgenic mice. To further examine the activation of the MHb after an acute administration of nicotine we used a dose of 6 mg/kg. c-Fos expression was increased in mice receiving nicotine compared with the saline group both in WT (t (1,4) = −12.11, p < 0.001, Student’s t test; Fig. 7b, c) and TgCHRNA5/A3/B4 mice (t (1,4) = −7.72, p < 0.001, Student’s t test). The effects of nicotine were significantly different between genotypes (genotype x treatment F (1,12) = 5.12, p = 0.05, two-way ANOVA) although the comparison among nicotine-treated groups did not reach statistical significance (t (1,4) = −2.27, NS, Student’s t test).

Fig. 6
figure 6

c-Fos immunoreactivity in VTA DAergic neurons. a Photomicrograh illustrating c-Fos immunoreactive DAergic cells in VTA of WT and TgCHRNA5/A3/B4 mice treated with an acute injection of saline or nicotine (0.5 mg/kg). Slices are fluorescently double-labeled with anti-tyrosine hydroxylase (red) and anti c-Fos (green). Scale bar = 20 μm. b Graph shows the number of TH/c-Fos positive cells per slice in VTA. Data are expressed as mean ± SEM. **p < 0.01, ***p < 0.001 comparison between treatment, and ooo p < 0.001 comparison between genotypes, Student’s t test

Full size image
Fig. 7
figure 7

c-Fos immunoreactivity in the MHb. a, b Number of c-Fos positive nuclei per area of the MHb of WT and TgCHRNA5/A3/B4 mice treated with an acute injection of saline or nicotine [0.5 (a) or 6 (b) mg/kg, s.c.]. c Photomicrograh illustrating c-Fos immunoreactive neurons in the MHb of WT and TgCHRNA5/A3/B4 mice treated with an acute injection of saline or nicotine (6 mg/kg, s.c.). Slices are fluorescently labeled with anti c-Fos (green) and Hoechst (blue). Scale bar = 20 μm. Data are expressed as mean ± SEM. ***p < 0.001 comparison between treatment, Student’s t test

Full size image

Discussion

Human genetic studies have demonstrated that polymorphisms in the CHRNA5/A3/B4 gene cluster located on chromosome 15 may influence smoking-related behaviors. Furthermore, some of these variants give rise to increased mRNA levels of the subunits (Doyle et al. 2011; Falvella et al. 2009; Schlaepfer et al. 2008; Wang et al. 2009; Xu et al. 2006). Even though overexpression of these genes has been associated with cigarette smoking, previous studies have only explored the role of single receptor subunits by knocking out or overexpressing individual mouse nAChR subunits. Thus, we explored the effects of overexpression of the whole human CHRNA5/A3/B4 cluster containing all three subunits using a bacterial artificial chromosome (BAC) transgenic mouse model. Our study demonstrates that overexpression of the human CHRNA5/A3/B4 genomic region in mice increases the reinforcing properties of nicotine (0.015 mg/kg per infusion) as revealed by an increased break point for intravenous self-administration. In addition, we observed reduced activation of VTA dopaminergic neurons upon an acute nicotine injection, which may reflect alterations in the balance between the rewarding and the aversive effects of nicotine. This change in balance between the rewarding and aversive properties may prompt the animals to initially consume the drug. We also found increased sensitivity to the pharmacological effects of nicotine, including higher activation of the MHb after acute nicotine administration together with an enhanced anxiety-like phenotype.

Multiple lines of evidence suggest coordinate regulation of the CHRNA5, CHRNA3, and CHRNB4 genes. In our study western blot analysis showed increased expression of the α3 and β4 subunits in our Tg mice. Consistently, TgCHRNA5/A3/B4 mice showed increased 5I-A-85380—resistant [125I]-epibatidine binding sites in brain regions where α3β4*-nAChRs are known to be normally expressed (Soria et al. 2005), including regions implicated in the control of nicotine addiction, such as the MHb and the interpeduncular nucleus (IPN) (Fowler et al. 2011; Frahm et al. 2011). Interestingly, our Tg mice showed a substantial increase in the expression of β4*-nAChRs in the CA1 region of hippocampus, also described previously in Tabac mice (Frahm et al. 2011), thus implicating the hippocampus as a putative region mediating the reinforcing effects of nicotine. The observed increase in binding sites in Tg mice was accompanied by increased sensitivity to nicotine-induced hypolocomotion and seizures. The present data are in accordance with previous studies reported in which deletion of the α3, β4 or α5 subunits individually, or in combination, result in a phenotype resistant to the acute effects of nicotine, specifically hypolocomotion and seizures (Salas et al. 2003a, b).

The main finding of our study was that Tg mice showed increased nicotine-reinforcing effects, as suggested by improved acquisition of nicotine self-administration and better performance in the progressive ratio schedule. TgCHRNA5/A3/B4 mice showed a preference for the active hole over the inactive hole on day one, whereas WT mice started to discriminate after the seventh day. Under the progressive-ratio TgCHRNA5/A3/B4 reached a higher break point to obtain a nicotine infusion than WT animals. This reveals an enhanced motivation of TgCHRNA5/A3/B4 mice to obtain the drug and suggests higher sensitivity to the reinforcing effects at this dose (0.015 mg/kg per infusion).

The habenula-interpeduncular system has been proposed as a pathway putatively involved in nicotine-induced locomotion, seizures, and consumption. Thus, we checked if the higher nicotine sensitivity observed in our TgCHRNA5/A3/B4 mice was correlated with higher MHb activation. We used a high dose of nicotine (6 mg/kg) that induced more pronounced seizures in our Tg mice compared with WT animals. As anticipated, the analysis of c-Fos expression in the MHb after an acute injection of nicotine revealed increased activation in Tg mice. Interestingly, 0.5 mg/kg of nicotine induced less activation of DAergic VTA neurons in TgCHRNA5/A3/B4 mice compared with their WT littermates. This dose activated the MHb in both genotypes, being higher (though not reaching statistical significance) in Tg mice. Since recent work demonstrates a critical role for the MHb in the circuitry controlling nicotine consumption (Fowler et al. 2011; Frahm et al. 2011), these data suggest that overexpression of the CHRNA5/A3/B4 cluster could alter the rewarding and aversive properties of nicotine. According to the study of Frahm et al. (2011) overexpression of the β4-nAChR subunit regulates the aversive effects of nicotine through increasing MHb activity. We hypothesize that the decreased activation of the DAergic neurons found in our Tg mice could be mediated by activation of the habenula-interpeduncular pathway through increased activity of α3β4*-nAChRs. Although this effect is translated to decreased nicotine consumption (Frahm et al. 2011), contrary to what we observed in our Tg mice, this could be due to the lower dose of nicotine that we used in the self-administration paradigm. Indeed, similar results were found between α5 null mutant and WT mice in the study of Fowler et al. (2011) when low doses of nicotine were used. The study by Fowler et al. (2011) demonstrated that nicotine activates the habenula-interpeduncular pathway through α5-containing nAChRs, triggering an inhibitory motivational signal that limits nicotine intake. Moreover, the same study showed that α5 subunit knockdown in the MHb does not alter the rewarding effects of nicotine, suggesting that α5-containing nAChRs located in the VTA may have a role in the rewarding effects of nicotine. Given that our Tg mice overexpress the α5 subunit we hypothesize that augmented α5-containing nAChRs in VTA may enhance the sensitivity to low doses of nicotine and thus increase the rewarding effects of the drug. The high-affinity α4β2-nAChR is the most abundant receptor in the brain and the principal mediator of nicotine dependence (Maskos et al. 2005; Mineur and Picciotto 2008; Tapper et al. 2004). Both α4β2 and α4α5β2 subtypes are present on the soma of dopaminergic neurons (Drenan et al. 2008; Klink et al. 2001; Marubio et al. 2003). Since the introduction of α5 subunit in 293 cells expressing α4 and β2 strongly favors assembly of α4α5β2 receptors and increases constitutive binding density (Gahring and Rogers 2010), overexpressing the α5 subunit in our Tg mice could lead to increased levels of α4α5β2-nAChRs on the cell surface of dopaminergic neurons, thus enhancing the rewarding effects of nicotine when low doses of the drug are used.

In summary, overexpression of the CHRNA5/A3/B4 cluster increases nicotine preference and sensitivity to its pharmacological effects due to augmented nicotine-binding sites that may enhance liability to consume nicotine through alterations in its reinforcing pathways. The present results suggest that the CHRNA5/A3/B4 genomic cluster is directly via the VTA or indirectly via the habenula-interpeduncular pathway an important determinant of tobacco consumption. Our study provides the first in vivo evidence that increased expression of these nicotinic receptor subunits enhances the motivational effects of low doses of nicotine and leads to nicotine addiction-related behaviors. However, further behavioral studies will be needed to clarify the aversive effects of high doses of nicotine in this Tg mouse model.