Introduction

Tobacco use is a global epidemic, and urgent actions are needed to prevent the continuing detrimental effects of tobacco smoking on health and the millions of deaths that result from this harmful addiction (World Health Organization 2009). The main psychoactive ingredient in tobacco is nicotine, which is thought to contribute substantially to the maintenance of the tobacco smoking habit (Stolerman and Jarvis 1995). Tobacco smoking rapidly delivers nicotine to the brain, where nicotine binds to nicotinic acetylcholine receptors (nAChRs) that are physiologically activated by the endogenous excitatory neurotransmitter acetylcholine.

nAChRs are ligand-gated cation channels that consist of five subunits. Twelve types of subunits have been identified for neuronal nAChR, α2-α10 and β2-β4, that together form the pentameric structure of the nAChR (Picciotto et al. 2001). The present study was designed to study the role of α7- and β4-containing nAChRs in the anhedonic and somatic aspects of nicotine withdrawal using mutant mice that lack these nAChR subunits.

α7 nAChR subunits predominantly form homomeric nAChR structures that consist of five α7 subunits. These receptors are widely and abundantly expressed throughout the brain in regions that include the hippocampus, cortex, thalamus (Quik et al. 2000), nucleus accumbens (Fu et al. 2000), and ventral tegmental area (VTA; Clarke et al. 1985; Jones and Wonnacott 2004). Homomeric α7 nAChR subunits typically exhibit relatively low affinity for nicotine and rapidly desensitize upon nicotine binding (Fenster et al. 1997; Gerzanich et al. 1994). Presynaptic, postsynaptic, and perisynaptic localization of α7 receptors has been demonstrated (Cordero-Erausquin et al. 2000). An important role has been implicated for presynaptic α7 nAChRs in brain reward processes by positively regulating several neurotransmitters, including glutamate (McGehee et al. 1995), dopamine (Schilstrom et al. 1998), and norepinephrine (Barik and Wonnacott 2006). Chronic nicotine exposure can upregulate α7 nAChRs (Marks et al. 1983), including those localized in the VTA (Pakkanen et al. 2005). The upregulation of α7 nAChRs, however, requires high concentrations of nicotine that exceed the nicotine levels achieved by smokers (Kawai and Berg 2001; Molinari et al. 1998).

Unlike α7 nAChR subunits that predominantly form homomeric receptors, β4 subunits combine with α subunits to form heteromeric nAChR structures. The β4 subunit is frequently found to combine with the α3 subunit but can also form functional nAChRs with other subunits (Gotti et al. 2009; e.g., α4 and α2 subunits). Markedly high expression of β4-containing nAChRs has been seen in the habenula-interpeduncular pathway and epiphysis, and moderate to high expression of β4-containing nAChRs has been found in the cerebellum, dentate gyrus, substantia nigra, and thalamus (Gotti et al. 2009; Grady et al. 2009; Hernandez et al. 2004; Quik et al. 2000). Specifically, co-distribution of the β4 nAChR subunit with the α3 nAChR subunit was shown for many of the brain regions where the β4 receptor has been shown to be highly expressed, including the epiphysis (Hernandez et al. 2004), medial habenula, interpeduncular nucleus, hippocampus, frontal cortex, and VTA (Winzer-Serhan and Leslie 1997). Compared with β2-containing nAChRs, β4-containing nAChRs desensitized at a relatively slow rate (Cachelin and Jaggi 1991; Papke et al. 1989) and exhibited markedly less affinity for nicotine (Parker et al. 1998; Wei et al. 2005). β4-containing nAChRs also showed high resistance to nicotine-induced upregulation compared with β2-containing nAChRs. Although upregulation of β4-containing nAChRs has been previously demonstrated, this upregulation was quickly reversed (Meyer et al. 2001) and not consistently demonstrated (Nguyen et al. 2003; Wang et al. 1998).

Mice that carry null mutations for specific nAChR subunits provide a unique opportunity for studying the role of receptors that contain these subunits in nicotine dependence. The use of such knockout mice has significantly advanced our understanding of the involvement of nAChRs that contain different nAChR subunits in aspects of nicotine dependence and other nicotine-induced effects. Specifically, studies have demonstrated a key role for β2-containing nAChRs in the reinforcing effects of nicotine. β2 knockout mice do not self-administer nicotine (Picciotto et al. 1998). In contrast, mice null for the α5 nAChR subunit exhibited increased nicotine self-administration behavior, suggesting that nAChRs that contain this subunit, particularly in the habenula-interpeduncular pathway, are involved in limiting nicotine intake (Fowler et al. 2011). Furthermore, somatic signs of nicotine withdrawal were absent in α5 and α2 knockout mice (Salas et al. 2009). Finally, α3 heterozygous and β4 knockout mice exhibit resistance to nicotine-induced seizures (Kedmi et al. 2004; Salas et al. 2003a, 2004a).

Deficits in brain reward function, such as depressed mood or anhedonia (i.e., the inability to experience pleasure from rewarding stimuli), reflect the main symptoms of nicotine withdrawal in humans (American Psychiatric Association 2000; Hughes 2007). The involvement of nAChRs in the reward deficits associated with nicotine withdrawal has not been studied extensively. The intracranial self-stimulation (ICSS) procedure is a unique procedure that allows the quantitative assessment of brain reward function in rodents by measuring elevations in current-intensity thresholds (Stoker and Markou 2011). Chronic nicotine administration lowers brain reward thresholds in C57BL/6J mice in the ICSS procedure, reflecting reward enhancement (Stoker et al. 2008), whereas nicotine withdrawal is associated with elevations in brain reward thresholds, reflecting anhedonia in mice (Johnson et al. 2008; Stoker et al. 2008). Although other affective signs of nicotine withdrawal have been assessed in mutant mice that lack various nAChR subunits (see “Discussion” section), the anhedonic aspects of nicotine withdrawal have not been extensively investigated in nAChR subunit knockout mice.

In addition to the anhedonic aspects of nicotine withdrawal, the nicotine withdrawal syndrome in humans is also characterized by somatic signs that include gastrointestinal disturbances, weight gain, decreased heart rate (American Psychiatric Association 2000), sweating, dizziness (Hughes and Hatsukami 1986), fatigue, nausea, constipation, and diarrhea (Shiffman 1979). In mice, somatic signs observed during nicotine withdrawal include body shakes, head shakes, paw tremor, scratching, and grooming (Semenova et al. 2003; Balerio et al. 2004; Stoker et al. 2008; Salas et al. 2004b; Damaj et al. 2003; Isola et al. 1999). Somatic signs of nicotine withdrawal have been assessed previously for several nAChR subunit knockout lines. Somatic signs were attenuated during nicotine withdrawal precipitated by the broad-spectrum nAChR antagonist mecamylamine in mice that carried null mutations of the β4 (Salas et al. 2004b), α5 (Jackson et al. 2008), and α7 (Salas et al. 2007), but not β2 (Salas et al. 2004b), nAChR subunits.

The present study explored the somatic and anhedonic aspects of nicotine withdrawal in α7 knockout (α7−/−), α7 wildtype (α7+/+), β4 knockout (β4−/−), and β4 wildtype (β4+/+) mice using the ICSS procedure and observations of somatic signs. Nicotine was administered through subcutaneously implanted osmotic minipumps. Nicotine withdrawal was induced either spontaneously by removing the minipumps or by administering the nonselective nAChR antagonist mecamylamine.

Materials and methods

Subjects

The α7 (Orr-Urtreger et al. 1997) and β4 (Xu et al. 1999) nAChR subunit null mouse lines were generated as described previously. The null lines were backcrossed to the C57BL/6J strain for 10 generations. Breeders were generously provided to our laboratory by Drs. Allan Collins and Michael Marks from the University of Colorado Boulder. Knockout and wildtype animals were bred in our laboratory from heterozygous breeding pairs. Wildtype mice were littermates of the knockout mice. Mice assessed in the present experiments were of the 5–8th generation since their transfer from the University of Colorado Boulder to the University of California San Diego, and on the 9–12th generation since the last backcross to the C57BL/6J background strain. Genotypes were determined by three-way polymerase chain reaction with the following primer sequences: α7 forward, 5′-CCTGGTCCTGCTGTGTTAAACTGCTTC-3′; α7 wildtype reverse, 5′-CTGCTGGGAAATCCTAGGCACACTTGAG-3′; α7 knockout reverse, 5′-GACAAGACCGGCTTCCATCC-3′; β4 forward, 5′-TGTAGAGCGAGCATCCGAACA-3′, β4 wildtype reverse, 5′-TCTCTACTTAGGCTGCCTGTCT-3′; β4 knockout reverse, 5′-AGTACCTTCTGAGGCGGAAAGA-3′. Upon weaning, the mice were maintained in groups of two to four until the beginning of the experiment, after which time the mice were housed individually. The subjects were males, 8–12 weeks of age at the beginning of the experiments. The animals were housed in a humidity- and temperature-controlled animal facility on a 12/12 h (lights off at 7:00 A.M.) reverse light/dark cycle with ad libitum access to food and water except during testing. Behavioral testing was conducted during the dark phase of the light/dark cycle, with the exception of the ICSS assessment conducted at the 12 h post-pump removal time-point that was conducted during the light phase of the cycle. During all ICSS testing and somatic signs observations, the testing room was illuminated by red light. All of the experiments were performed in accordance with the guidelines of the American Association for the Accreditation of Laboratory Animal Care and the National Research Council’s Guide for the Care and Use of Laboratory Animals and were approved by the University of California San Diego Institutional Animal Care and Use Committee.

Drugs

Nicotine hydrogen tartrate (Sigma-Aldrich, St. Louis, MO, USA) was dissolved in sterile 0.9% saline solution and infused through subcutaneous osmotic minipumps for 28 days (model 2004, Alzet, Palo Alto, CA, USA). Mecamylamine hydrochloride (Sigma-Aldrich, St. Louis, MO, USA) was dissolved in saline and injected subcutaneously in a volume of 0.1 ml/10 g. Nicotine doses are reported as the base, and mecamylamine doses are reported as the salt based on conventions in the literature.

Minipump implantation and removal

The mice were anesthetized with an isoflurane (1–3%)/oxygen vapor mixture, and osmotic minipumps (model 2004, Alzet, Palo Alto, CA, USA) were inserted using aseptic surgery techniques. Nicotine minipumps were placed subcutaneously (SC) parallel to the spine at the shoulder level with the flow moderator directed away from the wound. The wound was closed with 7 mm stainless steel wound clips (Reflex, Cellpoint Scientific, Gaithersburg, MD, USA). On day 29, nicotine minipumps were surgically removed under isoflurane anesthesia using aseptic surgery techniques.

Assessment of somatic signs of nicotine withdrawal

The mice were habituated to plastic cylinders (22 cm diameter, 25 cm height) used for the observation of somatic signs prior to testing for two 20-min periods. The observation of somatic signs occurred at 24 h of nicotine withdrawal, immediately after the light–dark box test, and was 20 min in duration. The scored somatic signs included forelimb tremors, body shakes, headshakes, and abdominal constrictions. Between observations, the cylinders were cleaned by changing the bedding. Somatic signs were analyzed as the total number of somatic signs during each observation period.

Intracranial self-stimulation: surgery, apparatus, and procedure

ICSS surgery

For ICSS surgery, the mice were anesthetized by inhalation of 1–3% isoflurane in oxygen and positioned in a stereotaxic frame (Kopf Instruments, Tujunga, CA, USA). Stainless steel bipolar electrodes (0.20 mm diameter; 6 mm length; Plastics One, Roanoke, VA, USA) were implanted into the medial forebrain bundle at the level of the lateral hypothalamus (coordinates: anterior/posterior, 1.58 mm; medial/lateral, 1.0 mm; dorsal/ventral, 5.3 mm from flat skull; Paxinos and Franklin 2001) as described previously (Gill et al. 2004; Stoker et al. 2008). Four stainless steel screws (3.2 mm length, Plastics One, Roanoke, VA, USA) were fixed to the skull to keep the electrode in place together with the application of a resin ionomer (Den-Mat, Santa Maria, CA, USA) and dental acrylic (Ortho-Jet, Lang Dental, Wheeling, IL, USA). The mice were allowed to recover for 7 days after surgery.

ICSS apparatus

ICSS training and testing were conducted in six Plexiglas operant chambers (30.5 × 24 × 27 cm; Med Associates, St. Albans, VT, USA). Each operant chamber was enclosed within a light- and sound-attenuated chamber (40 × 60 × 63.5 cm). Intracranial stimulation was delivered by constant-current stimulators (Stimtek model 1200c, San Diego Instruments, San Diego, CA, USA). The mice were connected to the stimulation circuit through flexible bipolar leads (Plastics One, Roanoke, VA, USA) attached to gold-contact swivel commutators (model SL2C, Plastics One, Roanoke, VA, USA) mounted above the operant chamber. The operant response required by the mice was a simple one-quarter turn of a wheel manipulandum (5.5 cm diameter, 4 cm width) that extended 1.5 cm from one wall of the operant chamber. The stimulation parameters, data collection, and all test session functions were controlled by a computer.

ICSS procedure

The ICSS procedure was described previously for mice (Gill et al. 2004; Stoker et al. 2008) and adapted from a discrete-trial current-intensity threshold procedure originally used in rats (Kornetsky and Esposito 1979; Markou and Koob 1992). Initially, the mice were trained to turn the wheel manipulandum on a fixed-ratio one schedule of reinforcement. The electrical stimulus consisted of 50 pulses with a pulse duration of 100 and a 10,000 μs interpulse interval. After successful acquisition of this schedule (i.e., two sessions in which the mouse received 200 reinforcers in less than 10 min), the mice were tested in the discrete-trial current-intensity threshold procedure. Each trial began with the delivery of a noncontingent electrical stimulus followed by a 7.5 s response window within which the subject could make a response to receive a second contingent stimulus identical in all parameters to the initial noncontingent stimulus. A response during this time window was labeled a positive response, and the lack of a response was labeled a negative response. During a 2 s period immediately after a positive response, additional responses had no consequences. The intertrial interval that followed either a positive response or the end of the response window (in the case of a negative response) had an average duration of 10 s (ranging from 7.5 to 12.5 s). Responses that occurred during the intertrial interval were labeled timeout responses and resulted in a further 12.5 s delay of the onset of the next trial. During training on the discrete-trial procedure, the duration of the intertrial interval and delay periods induced by timeout responses were gradually increased until both reached a duration of 10 s (ranging from 1 to 10 s during training). The animals were subsequently tested in the current-intensity threshold procedure in which stimulation current intensities were varied according to the classical psychophysical method of limits. A test session consisted of four alternating series of descending and ascending current intensities, starting with a descending series. Blocks of three trials were presented to the subject at a given stimulation intensity, and the intensity changed by 5 μA steps between blocks of trials. The initial stimulus intensity was set at ~30–40 μA above the baseline current-threshold for each animal. The mice were originally presented with a stimulus that consisted of 50 pulses, each 100 μs in duration, with an interpulse interval of 10,000 μs that resulted in 99 Hz stimulation. If mice exhibited thresholds below 70 μA during the first seven test sessions, then electrical stimuli that were half of the original frequency (25 pulses of 100 μs in duration at an interpulse interval of 20,000 μs) were made available to them. This procedure was implemented to ensure that baseline thresholds would not be exceedingly low and thus avoid potential floor effects. Each test session typically lasted 30–40 min and provided two dependent variables for behavioral assessment: threshold and response latency. The threshold value of each series was defined as the midpoint in microamperes between the current intensity level at which the animal made two or more positive responses out of the three stimulus presentations and the level at which the animal made less than two positive responses. The animal’s estimated current-intensity threshold for each test session was the mean of the four series’ thresholds. The response latency was defined as the average time in seconds that elapsed between the delivery of the electrical stimulus and the turning of the wheel manipulandum for all of the trials that led to a positive response.

Experimental design

Naive α7−/−, α7+/+, β4−/−, and β4+/+ mice were prepared with stimulating electrodes in the lateral hypothalamus and trained in the ICSS procedure. After the establishment of stable ICSS performance (defined as ≤10% variation in brain reward thresholds over three consecutive daily test sessions), the mice were prepared with 28-day subcutaneous osmotic minipumps that delivered saline or 40 mg/kg/day nicotine (base) and tested once daily. ICSS thresholds and response latencies during nicotine administration and withdrawal are expressed as a percentage of the baseline value (i.e., 5-day mean before nicotine/saline administration). The broad-spectrum nAChR antagonist mecamylamine hydrochloride was injected subcutaneously on days 9, 11, 13, and 15 of exposure to nicotine/saline using a within-subjects Latin square design for the factor Dose (0, 1.5, 3, and 6 mg/kg salt). The mice were tested in the ICSS procedure immediately after mecamylamine/saline injection. ICSS thresholds and response latencies after mecamylamine/saline administration are expressed as the percentage of the previous day’s baseline values. The mice were habituated to the cylinders for the observation of somatic signs on days 25 and 27 of nicotine/saline exposure. On day 29, the pumps were removed, and the mice were tested in the ICSS procedure at 3, 6, 8, 12, 24, 28, 48, 52, 72, 76, 96, 100 h after pump removal. At the 24 h time-point post-pump removal, immediately after completion of the ICSS test, the mice were subjected to a 5-min light–dark box session followed immediately by a 20 min observation of somatic signs.

Because no pronounced or long-lasting withdrawal effects were observed in β4−/− and β4+/+ mice after the termination of 40 mg/kg/day nicotine administration (see “Results” section), a separate naive cohort of β4−/− and β4+/+ mice were treated with 80 mg/kg/day nicotine or saline for 28 days via subcutaneous osmotic minipumps. On day 29, the pumps were removed, and somatic signs were evaluated 24 h post-pump removal. Upon the successful demonstration of somatic signs of 80 mg/kg/day nicotine withdrawal in β4+/+ mice, this nicotine dose was administered to an additional set of naive β4−/− and β4+/+ mice that were assessed in the ICSS procedure using identical experimental procedures as those described above for mice treated with 40 mg/kg/day nicotine.

Statistical analyses

All analyses were performed using the PASW 18 Statistical Package (SPSS, Chicago, IL, USA). The data were analyzed using appropriate two- or three-way analyses of variance (ANOVAs) with Genotype and Nicotine/Saline Exposure as the between-subjects factors and Administration Day/Withdrawal Day and Mecamylamine Dose as the within-subjects factors. Newman-Keuls post hoc analyses followed statistically significant interaction effects in the ANOVAs. The level of significance was set at 0.05. To evaluate the effects of mecamylamine on ICSS performance, thresholds and response latencies are expressed as a percentage of the baseline values obtained during the last ICSS session before the mecamylamine/saline injection. To evaluate the effects of nicotine administration, thresholds and response latencies are expressed as a percentage of the baseline values obtained during the last five daily ICSS sessions before pump implantation. To evaluate ICSS performance during spontaneous nicotine/saline withdrawal, thresholds and response latencies are expressed as a percentage of the baseline values obtained during the last five ICSS daily sessions before pump removal. For ICSS withdrawal, data time-points were combined for analysis (3 + 6 h, 8 + 12 h, 24 + 28 h, 48 + 52 h, 72 + 76 h, 96 + 100 h) to provide a robust and reliable estimate of the effects of nicotine/saline withdrawal, similar to our previous work (Stoker et al. 2008). Mice that demonstrated instability in ICSS thresholds upon minipump withdrawal (three or more unstable time points [i.e., more than 10% deviation from the 3-day running average] during the 3–100 h withdrawal time window) were excluded from the analysis of the ICSS data during withdrawal. The following numbers of mice were excluded from these analyses: saline-treated α7+/+, n = 2; nicotine-treated α7+/+, n = 1; saline-treated α7−/−, n = 2; saline-treated β4+/+, n = 4; nicotine-treated β4+/+, n = 2; saline-treated β4−/−, n = 4; nicotine-treated β4−/−, n = 2.

Results

Nicotine administration

The baseline values used to calculate all of the percentages of baseline thresholds are shown in Table 1. Chronic nicotine administration did not affect ICSS thresholds (Fig. 1) or latencies (data not shown) in α7+/+, α7−/− (Fig. 1a), β4+/+, or β4−/− (Fig. 1b) mice, reflected by a lack of significant interaction or main effects in the ANOVAs.

Table 1 Mean baseline thresholds (μA) of the experimental groups used to express the data as percent of baseline
Full size table
Fig. 1
figure 1

Effects of chronic administration of nicotine (40 mg/kg/day base, 28 days) or saline on ICSS thresholds in α7 (a) and β4 (b) knockout and wildtype mice. The data (mean ± SEM) are expressed as a percentage of baseline thresholds (see text for details). No statistically significant effects were observed. The effects of nicotine administration on ICSS thresholds are not shown for days 9–15 in this figure because during these days the effects of acute administration of various mecamylamine doses on ICSS thresholds were assessed using a Latin square design (see text for details)

Full size image

Mecamylamine-precipitated nicotine withdrawal

An ANOVA of the effects of mecamylamine on ICSS thresholds in α7−/− and α7+/+ mice (Fig. 2a) revealed a significant Mecamylamine Dose × Nicotine/Saline Exposure × Genotype interaction (F 3,144 = 3.777, P < 0.05) and significant main effects of Mecamylamine (F 3,144 = 7.320, P < 0.001) and Nicotine/Saline Exposure (F 1,48 = 34.197, P < 0.001) but not Genotype. Post hoc tests revealed that 1.5 mg/kg mecamylamine significantly elevated ICSS thresholds in nicotine-treated α7−/− mice compared with their saline-treated counterparts (P < 0.05), whereas thresholds in α7+/+ mice were not significantly altered. Administration of 3 mg/kg mecamylamine significantly elevated threshold in nicotine-treated α7+/+ (P < 0.05), but not α7−/−, mice compared with their saline-treated counterparts. Administration of 6 mg/kg mecamylamine resulted in threshold elevations in nicotine-treated α7+/+ mice compared with saline-treated α7+/+ mice (P < 0.01) and nicotine-treated α7−/− mice (P < 0.01). Confirming the lack of effect of 6 mg/kg mecamylamine on thresholds in nicotine-treated α7−/− mice, thresholds in nicotine- and saline-treated α7−/− mice did not differ significantly.

Fig. 2
figure 2

Effects of mecamylamine-precipitated nicotine (40 mg/kg/day base, 28 days) or saline withdrawal on ICSS thresholds in α7 (a) and β4 (b) knockout and wildtype mice. The data (mean ± SEM) are expressed as a percentage of baseline thresholds (see text for details). Asterisks indicate significant differences between nicotine- and saline-treated mice within the same genotype (*P < 0.05, ***P < 0.001). Pound signs indicate significant differences between nicotine-treated α7−/− and α7+/+ mice (## P < 0.001). Delta signs (∆∆∆ P < 0.001) indicate a significant main effect of Mecamylamine in an ANOVA

Full size image

Mecamylamine had similar effects on ICSS thresholds in β4+/+ and β4−/− mice at all doses tested (Fig. 2b), reflected by the absence of a significant Mecamylamine × Nicotine/Saline Exposure × Genotype interaction and no main effect of Genotype. Significant main effects were found for Nicotine/Saline Exposure (F 1,59 = 13.984, P < 0.001) and Mecamylamine (F 3,177 = 7.215, P < 0.001), reflecting the fact that mecamylamine elevated thresholds in nicotine-treated mice, but this effect was not different in the mutant and wildtype mice.

The latency to turn the wheel manipulandum remained unaffected after mecamylamine administration in both α7 and β4 knockout lines treated with either nicotine or saline (data not shown).

Spontaneous nicotine withdrawal

During spontaneous withdrawal from 40 mg/kg/day nicotine, α7−/− mice exhibited a delay in the onset of ICSS threshold elevations, reflected by a statistically significant difference between nicotine-treated α7+/+ and α7−/− mice at 3–6 h of withdrawal (Fig. 3a) in a t test. Although an ANOVA of ICSS thresholds in α7 mice during spontaneous withdrawal did not show a significant Nicotine/Saline Exposure × Genotype × Withdrawal Day interaction or a significant main effect of Genotype, significant main effects were found for Withdrawal Day (F 5,215 = 3.579, P < 0.01) and Nicotine/Saline Exposure (F 1,43 = 6.422, P < 0.05). Pre-planned comparisons did not reveal any differences between groups at any time-points, but a t test revealed a significant difference between nicotine- and saline-treated α7+/+ mice at 3–6 h of withdrawal (t 24 = 2.201, P < 0.05). The latency to turn the wheel manipulandum was not affected by spontaneous nicotine/saline withdrawal in either α7−/− or α7+/+ mice (data not shown).

Fig. 3
figure 3

Effects of spontaneous nicotine (40 mg/kg/day base, 28 days) or saline withdrawal on ICSS thresholds in α7 (a) and β4 (b) knockout and wildtype mice. The data (mean ± SEM) are expressed as a percentage of baseline thresholds (see text for details). The pound sign in (a) indicates a significant main effect of Nicotine/Saline Exposure (# P < 0.05) in a three-way repeated-measures ANOVA. The asterisk indicates a significant difference between nicotine- and saline-treated wildtype mice in a t test (*P < 0.05). The pound sign in (b) indicates a significant difference between nicotine-withdrawing β4−/− and β4+/+ mice (# P < 0.05) in a t test

Full size image

Nicotine-withdrawing β4−/− mice did not show threshold elevations at any time-point, whereas threshold elevations in nicotine-withdrawing β4+/+ mice were short-lasting and returned to baseline levels 8–12 h after the termination of chronic nicotine administration (Fig. 3b). An ANOVA of the thresholds did not yield a significant Nicotine/Saline Exposure × Genotype × Withdrawal Day interaction or significant main effects of Nicotine/Saline Exposure, Genotype, or Withdrawal Day. Pre-planned comparisons between experimental groups did not indicate any differences between groups at any time-point, but a t test revealed a significant difference between nicotine- and saline-treated β4+/+ mice at 3–6 h of withdrawal (t 21 = 2.082, P < 0.05). Thresholds in nicotine-withdrawing β4−/− mice were significantly lower than in β4+/+ mice (t 24 = 1.141, P < 0.05). The elevation of ICSS thresholds of saline-treated mice after the pump removal was likely caused by the repeated daily assessments that took place because the mice were only tested once daily in the ICSS procedure before pump removal. The latency to turn the wheel manipulandum was not affected by spontaneous nicotine/saline withdrawal in β4−/− or β4+/+ mice (data not shown).

The increase in somatic signs observed in α7+/+ mice during withdrawal from 40 mg/kg/day nicotine remained intact in nicotine-withdrawing α7−/− mice (Fig. 4a), reflected by a significant main effect of Nicotine/Saline Exposure (F 1,43 = 17.584, P < 0.001), no Nicotine/Saline Exposure × Genotype interaction, and no main effect of Genotype. The increase in somatic signs in mice withdrawn from nicotine, independent of genotype, was also shown by pre-planned comparisons. Nicotine-withdrawing α7+/+ (P < 0.05) and α7−/− (P < 0.01) mice exhibited significantly more somatic signs of withdrawal than their saline-treated counterparts. Light–dark box data are not reported due to the absence of reliable detection of anxiety-like behavior in wildtype mice of either mouse line. Because of the non-invasive nature of the light–dark box test, it is highly unlikely that exposure to the light–dark box affected the observations of somatic signs of nicotine withdrawal that were conducted in some cases shortly after the light–dark box.

Fig. 4
figure 4

Effects of spontaneous withdrawal from nicotine (40 mg/kg/day base, 28 days) or saline on somatic signs in α7 (a) and β4 (b) knockout and wildtype mice. The data are expressed as the mean number of somatic signs ± SEM. Asterisks indicate significant differences between nicotine- and saline-treated mice within the same genotype (*P < 0.05, **P < 0.01). The pound sign indicates a significant main effect of Nicotine/Saline Exposure (# P < 0.05)

Full size image

After the termination of 40 mg/kg/day nicotine administration, β4+/+ mice exhibited a tendency toward an increased number of somatic signs compared with saline-treated β4+/+ mice, whereas nicotine-withdrawing β4−/− mice did not show increases in somatic signs compared with saline-treated β4−/− mice (Fig. 4b). This observation was supported by a trend toward a Nicotine/Saline Exposure × Genotype interaction (F 1,46 = 2.903, P < 0.1) in the ANOVA. However, pre-planned comparisons did not reveal any significant differences between experimental groups.

Somatic signs were significantly attenuated in β4−/− mice compared with β4+/+ mice during withdrawal from 80 mg/kg/day nicotine (Fig. 5), reflected by a significant Nicotine/Saline Exposure × Genotype interaction (F 1,22 = 13.167, P < 0.01) and significant main effects of Nicotine/Saline Exposure (F 1,22 = 54.450, P < 0.001) and Genotype (F 1,22 = 7.601, P < 0.05). Post hoc comparisons revealed significant differences between nicotine- and saline-treated β4+/+ mice (P < 0.01), between nicotine- and saline-treated β4−/− mice (P < 0.05), and between nicotine-treated β4+/+ and β4−/− mice (P < 0.01). Although withdrawal from 80 mg/kg/day nicotine induced an increase in the somatic signs of withdrawal in β4+/+ mice (Fig. 5), it did not induce long-lasting elevations in ICSS thresholds in these mice (data not shown) beyond those seen after the termination of 40 mg/kg/day nicotine administration (Fig. 3b).

Fig. 5
figure 5

Effects of spontaneous nicotine withdrawal (80 mg/kg/day base, 28 days) or saline on somatic signs in β4−/− and β4+/+ mice. The data are expressed as the mean number of somatic signs ± SEM. Asterisks indicate significant differences between groups linked by the horizontal lines (*P < 0.05, **P < 0.01)

Full size image

Discussion

Spontaneous nicotine withdrawal elevated ICSS thresholds in α7+/+, but not α7−/−, mice 3–6 h post-pump removal. After this initial delay in the onset of spontaneous nicotine withdrawal in α7−/− mice, thresholds were gradually elevated in α7−/− mice until they reached levels similar to those in nicotine-withdrawing α7+/+ mice at the 24 h and later time-points of withdrawal. Consistent with these threshold findings, α7−/− and α7+/+ mice showed similar increases in the somatic signs of spontaneous nicotine withdrawal at 24 h of withdrawal. Interestingly, administration of the broad-spectrum nAChR antagonist mecamylamine in nicotine- and saline-treated mice induced effects in opposite directions at the low (1.5 mg/kg) and high (6 mg/kg) mecamylamine doses tested in nicotine-treated α7−/− mice. Specifically, the low dose of 1.5 mg/kg mecamylamine significantly elevated ICSS thresholds in α7−/−, but not α7+/+, mice. The intermediate and highest doses of 3 and 6 mg/kg/day mecamylamine significantly elevated thresholds only in α7+/+, and not α7−/−, mice.

Similar to the data in α7 mice, ICSS thresholds in β4+/+ mice were elevated only at the 3–6 h time-point of withdrawal from 40 mg/kg/day nicotine, whereas thresholds in β4−/− mice remained at baseline levels. Starting at 8–12 h of withdrawal, however, ICSS thresholds in β4+/+ mice returned to baseline levels. The lack of robust spontaneous nicotine withdrawal effects, reflected by the absence of threshold elevations in β4+/+ mice, was also reflected in the lack of significant increases in somatic signs at 24 h of withdrawal from 40 mg/kg/day nicotine in these mice, although a clear trend was observed. Nevertheless, a clear increase in the somatic signs of withdrawal was observed in β4+/+ mice after the termination of 80 mg/kg/day nicotine administration, which was absent in β4 knockout animals.

The present study was one of the first attempts to characterize the nicotine withdrawal syndrome in α7 and β4 nAChR knockout mice using the ICSS procedure. Previous studies have assessed α7 and β4 nAChR knockout mice in various other measures of nicotine withdrawal, including somatic signs and anxiety-related behavioral tests. Under baseline conditions, the behavior of α7−/− mice resembles α7+/+ mice in various behavioral procedures, including locomotor activity, anxiety-like behavior, fear conditioning, and spatial learning (Paylor et al. 1998; Salas et al. 2007; Wehner et al. 2004). Subtle differences between α7−/− and α7+/+ mice include impairments in the sympathetic responses to vasodilatation, indicating autonomic dysfunction in these mutant mice (Franceschini et al. 2000). Additionally, α7−/− mice demonstrated a slight increase in the time spent in the center of the open field, although this anxiogenic-like effect was not seen in the light–dark box test, another procedure that assesses anxiety-like behavior (Paylor et al. 1998). The same study demonstrated a subtle improvement in spatial learning in α7−/− mice compared with their wildtype counterparts. α7−/− mice located the hidden platform in the Morris water task faster than their wildtype counterparts, but this genotype difference was not supported by the distance traveled (Paylor et al. 1998). By contrast, mild cognitive impairment was observed in α7−/− mice in terms of an attentional deficit under specific conditions in the five-choice serial reaction time task (Hoyle et al. 2006; Young et al. 2004), signaled nose-poke task (Keller et al. 2005), and radial arm maze (Levin et al. 2009). In the present experiments, no differences were observed in the acquisition of the discrete-trial current-intensity ICSS procedure between α7−/− and α7+/+ mice.

Nicotine administration had similar effects in α7−/− and α7+/+ mice, including seizure sensitivity, lever-pressing behavior, locomotor activity, hyperthermia, cued and contextual fear conditioning, nicotine discrimination, and nicotine-induced conditioned place preference (Franceschini et al. 2002; Naylor et al. 2005; Salas et al. 2007; Stolerman et al. 2004; Tritto et al. 2004; Walters et al. 2006; Wehner et al. 2004). A recent study suggested that nicotine intake was similar in α7−/− and α7+/+ mice throughout the first 21 days of administration; however, after this 21-day period, α7−/− mice self-administered significantly less nicotine in a two-bottle procedure (Levin et al. 2009). The present study did not find differences in ICSS thresholds between nicotine-treated α7−/− and α7+/+ mice during the 28-day period of chronic nicotine administration via subcutaneous osmotic minipumps.

During spontaneous withdrawal from 28 days of 40 mg/kg/day nicotine, α7−/− mice exhibited a delay in the onset of ICSS threshold elevations. After the initial delay in the onset of withdrawal symptoms seen at the 3–6 h post-nicotine time-point, thresholds in α7−/− mice progressively elevated until reaching levels similar to those in α7+/+ mice at the 48–52 h post-nicotine time-point. Furthermore, ICSS thresholds in α7−/− mice were differentially affected by mecamylamine-precipitated nicotine withdrawal, depending on the dose of mecamylamine. The lower dose of 1.5 mg/kg mecamylamine significantly elevated thresholds in nicotine-treated α7−/− mice, whereas thresholds in nicotine-treated α7−/− mice were not significantly different from those in saline-injected mice of the same genotype at 3 and 6 mg/kg mecamylamine. In contrast, α7+/+ mice treated with nicotine exhibited a trend toward a linear dose-dependent elevation in thresholds with increasing doses of mecamylamine. These interesting effects seen in α7−/− mice may reflect compensatory mechanisms in other nAChRs that result from constitutive null mutation of the α7 nAChR gene. However, no significant differences were detected in the mRNA expression levels of other nicotinic receptor subunits between α7−/− and α7+/+ mice (Franceschini et al. 2002). Thus, these differential effects of low and high mecamylamine doses in α7−/− mice may be explained instead by adaptations in other central nervous system receptors, besides nAChRs, in these mice. Mecamylamine is considered a broad-spectrum noncompetitive antagonist of neuronal nAChRs, although it differentially affects different nAChR subtypes. The effects of mecamylamine on β subunit-containing nAChRs are prolonged compared with the effects of mecamylamine on α7 receptors, on which the inhibition by mecamylamine is rapidly reversed (Frazier et al. 1998; Papke et al. 2001). The short-lasting effects that mecamylamine exerts on α7 nAChR subunits (Frazier et al. 1998; Papke et al. 2001) suggests that null mutation of the α7 receptor does not likely affect the overall effects of mecamylamine. Mecamylamine also acts as an N-methyl-d-aspartate (NMDA) receptor antagonist (Chavez-Noriega et al. 1997; Frazier et al. 1998) but only at high concentrations (Papke et al. 2001). Previous work showed that chronic nicotine administration induced changes in NMDA receptor subunit expression (Kenny et al. 2009). Moreover, the NMDA receptor antagonist dizocilpine (MK-801) showed a strong tendency to lower ICSS thresholds in nicotine-treated rats compared with saline-treated rats (Kenny et al. 2003). Considered in the context of this previous literature, the present findings in α7 knockout mice may be explained by antagonism of NMDA receptors at high mecamylamine doses, leading to threshold lowering that counterbalances the threshold elevations induced by mecamylamine administration in nicotine-treated α7−/− mice.

A recent study indicated that somatic signs were similar between nicotine-treated α7−/− and α7+/+ mice 10 min after mecamylamine administration (Jackson et al. 2008). In an earlier study, however, Salas and colleagues observed significantly fewer somatic signs of mecamylamine-precipitated nicotine withdrawal in α7−/− mice compared with α7+/+ mice when tested immediately after mecamylamine administration (Salas et al. 2007), which may indicate a delayed onset of the nicotine withdrawal symptoms in α7−/− mice. The present study found similar increases in somatic signs in α7+/+ and α7−/− mice at 24 h of spontaneous nicotine withdrawal, consistent with results from a previous study that found similar increases in somatic signs at 48 h post-nicotine between wildtype and mutant mice (Grabus et al. 2005). Overall, this pattern of results supports the present findings of a delay in the onset of the nicotine withdrawal syndrome in α7−/− mice in the ICSS procedure. In the present experiments, ICSS thresholds were assessed at several time-points, whereas somatic signs were assessed only at 24 h of spontaneous nicotine withdrawal; therefore, the present data could not inform about potential alterations in the time-course of the somatic signs of nicotine withdrawal. Importantly, somatic signs and brain reward thresholds may be differentially regulated. While brain reward threshold elevations associated with nicotine withdrawal are regulated by the central nervous system (Olds and Fobes 1981; Watkins et al. 2000), somatic signs were demonstrated to be primarily mediated by the peripheral nervous systems (Watkins et al. 2000). It should be noted, however, that results from Malin and colleagues (Malin et al. 1997) indicated only central, and not peripheral, mediation of the somatic signs of nicotine withdrawal. In either case, the delay in the onset of brain reward threshold elevations and the onset of somatic signs in α7−/− mice may be indicative of a delay in the overall effects of the nicotine withdrawal syndrome in these mice.

Although a linkage between allelic variation in chromosome region 15q25, which holds the α5-α3-β4 nAChR subunit gene cluster, and an increased risk of tobacco addiction was recently demonstrated (Liu et al. 2010; Saccone et al. 2007), the behavioral phenotype of β4 knockout mice has not been investigated as extensively as α7 knockout mice. Studies have assessed, however, the behavioral profile of these mice under various conditions, including under baseline conditions, during nicotine administration and withdrawal. Under baseline conditions, β4 knockout mice generally resemble wildtype mice. For example, β4 knockout mice and their wildtype counterparts exhibited similar behavior in the contextual and cued fear conditioning procedures (Wehner et al. 2004), indicating that β4-containing nAChRs may not be involved in these learning processes. An anxiolytic phenotype has been indicated for β4 knockout mice in the elevated plus maze and staircase maze (Salas et al. 2003b). Interestingly, this decrease in the anxiety-like response in β4−/− mice was not detected in other tests of anxiety, including the light–dark box, open field, and mirrored chamber (Salas et al. 2003b). Under baseline conditions, core body temperature is lower in β4−/− than in β4+/+ mice (Sack et al. 2005), consistent with the observation that the β4 nAChR subunit is an important component of nAChRs in the autonomic nervous system (Xu et al. 1999). Furthermore, social isolation of β4−/− mice induced an augmented increase in heart rate above that seen in β4+/+ mice during social isolation (Salas et al. 2003b). Nevertheless, social isolation unlikely influenced the results of the present studies. Social isolation per se has not been known to alter ICSS thresholds under baseline conditions. The administration of nicotine had similar effects in β4−/− and β4+/+ mice during contextual and cued fear conditioning (Wehner et al. 2004). β4−/− mice, however, were less sensitive to the effects of nicotine than β4+/+ mice in the open field test. Specifically, the administration of low nicotine doses in β4−/− mice did not decrease the open field exploration behavior that was observed in β4+/+ mice (Salas et al. 2004a). Moreover, seizures induced by high nicotine doses in β4+/+ mice were absent in β4−/− mice (Kedmi et al. 2004; Salas et al. 2004a).

Mecamylamine administration in nicotine-treated β4−/− mice resulted in an attenuated increase in somatic signs compared with β4+/+ mice (Salas et al. 2004b). Furthermore, null mutation of the β4 receptor reduced hyperalgesia induced by spontaneous nicotine withdrawal (Salas et al. 2004b). In the present study, β4−/− mice withdrawn from 40 mg/kg/day chronic nicotine administration did not show ICSS threshold elevations throughout the 5-day assessment period or an increase in somatic signs at 24 h of withdrawal. β4+/+ mice showed a clear trend toward an increase in somatic signs induced by the cessation of 40 mg/kg/day nicotine administration and displayed a short-lasting elevation in ICSS thresholds at 3–6 h post-pump removal, in contrast to the longer-lasting affective and somatic symptoms of spontaneous nicotine withdrawal exhibited by their C57BL/6J background strain (Stoker et al. 2008) and α7+/+ mice in the present study. The cessation of 28 days of 80 mg/kg/day nicotine administration significantly increased the somatic signs of withdrawal in β4+/+ mice and significantly attenuated these increases in β4−/− mice. However, the same high dose of nicotine did not prolong the elevations of ICSS thresholds in β4+/+ mice in a separate pilot experiment (data not shown). The lack of pronounced and long-lasting threshold elevations in β4+/+ mice leads to somewhat inconclusive findings about the role of β4-containing nAChRs in the anhedonic aspects of nicotine withdrawal, although a delay or rather lack of onset of the anhedonic aspects of nicotine withdrawal may occur in these mice.

Summary of results and conclusions

α7−/− mice showed a delay in the onset of ICSS threshold elevations during spontaneous nicotine withdrawal. The administration of a low mecamylamine dose induced significant elevations in ICSS thresholds in nicotine-treated α7−/− mice that were absent in α7+/+ mice, whereas a high mecamylamine dose had no effect in nicotine-treated α7−/− mice and elevated thresholds of nicotine-treated α7+/+ mice. The somatic signs of spontaneous nicotine withdrawal were similar in α7−/− and α7+/+ mice. These results indicate that the α7 nAChR is involved in the initiation of the anhedonic aspects of nicotine withdrawal, but these nAChRs are not involved in the expression of the somatic signs of nicotine withdrawal. Furthermore, the effects of mecamylamine in nicotine-treated α7 knockout and wildtype mice suggests that compensatory adaptations may occur in constitutive α7−/− mice or that mecamylamine may interact with other receptors besides nAChRs in these mice.

β4−/− mice withdrawn from nicotine did not exhibit threshold elevations at 3–100 h of withdrawal when testing occurred, whereas β4+/+ mice showed short-lasting threshold elevations that were observed only at the 3–6 h time-point of withdrawal. Therefore, the results of the present study can only lead to the conclusion of an apparent delay or lack of onset of the anhedonic aspects of the nicotine withdrawal syndrome in β4 knockout mice. Mecamylamine-induced threshold elevations were similar in β4−/− and β4+/+ mice. Increases in the somatic signs of spontaneous nicotine withdrawal were only seen in β4+/+ mice after the cessation of 80 mg/kg/day, but not 40 mg/kg/day, nicotine administration for 28 days. Thus, these findings indicate that β4-containing receptors are involved in both the anhedonic and somatic signs of nicotine withdrawal.

The aversive aspects of the early withdrawal syndrome have been implicated as powerful motivators in the continuation of the tobacco smoking habit (Baker et al. 2004a; 2004b; Piasecki et al. 2002). In addition, depressive symptoms associated with tobacco withdrawal were found to be especially important in predicting relapse (Hughes 2007). Pharmacological treatment of the anhedonic symptoms of tobacco withdrawal may thus help prevent relapse (Leventhal et al. 2008). The present studies in α7 and β4 knockout mice may suggest that blockade of α7- or β4-containing nAChRs may alleviate the aversive anhedonic signs of the early nicotine withdrawal syndrome, thereby potentially decreasing the risk for relapse and thus increasing the chance of successful discontinuation of the tobacco smoking habit.

In summary, the results of the present study confirm previous findings of attenuated somatic signs of nicotine withdrawal in β4, but not α7, knockout mice. A delay in the onset of the anhedonic aspects of spontaneous nicotine withdrawal, measured with the ICSS procedure, was demonstrated in mice null for the α7 and β4 nAChR subunits.