Nicotine Activation of α4* Receptors: Sufficient for Reward, Tolerance, and Sensitization

See allHide authors and affiliations

Science  05 Nov 2004:
Vol. 306, Issue 5698, pp. 1029-1032
DOI: 10.1126/science.1099420


The identity of nicotinic receptor subtypes sufficient to elicit both the acute and chronic effects of nicotine dependence is unknown. We engineered mutant mice with α4 nicotinic subunits containing a single point mutation, Leu9′ → Ala9′ in the pore-forming M2 domain, rendering α4* receptors hypersensitive to nicotine. Selective activation of α4* nicotinic acetylcholine receptors with low doses of agonist recapitulates nicotine effects thought to be important in dependence, including reinforcement in response to acute nicotine administration, as well as tolerance and sensitization elicited by chronic nicotine administration. These data indicate that activation of α4* receptors is sufficient for nicotine-induced reward, tolerance, and sensitization.

Nicotine addiction, the largest cause of preventable mortality in the world, leads to >4 million smoking-related deaths annually. Nicotine dependence begins with nicotine binding to nicotinic acetylcholine receptors (nAChRs) in the central nervous system. Nicotinic receptors are pentameric cation-permeable ligand-gated ion channels that are activated by the endogenous neurotransmitter, acetylcholine (ACh), as well as by the naturally occurring alkaloid found in tobacco, nicotine. Currently, 12 neuronal nAChR subunits have been identified (α2 to α10 and β2 to β4), and the subunit composition of each channel determines its electrophysiological properties and agonist-binding affinities (1, 2). Thus, many nAChR subtypes exist because most subunits can form heteromeric channels, whereas a subset, α7 to α10, may form homomeric channels.

Nicotine, like most drugs of abuse, interacts with the mesocorticolimbic pathways. Rats readily self-administer nicotine through intravenous infusion (3), and this behavior is attenuated when nicotinic antagonists are directly infused into the ventral tegmental area (VTA), indicating that the rewarding effects of nicotine are mediated through this brain region (4). Chronic nicotine exposure, such as that which occurs in smokers, leads to nAChR desensitization and various long-term physiological alterations, including nAChR upregulation (57), modulation of gene expression (8, 9), and induction of long-term potentiation and depression at glutamatergic synapses (10, 11). In addition, chronic nicotine also induces behavioral tolerance, sensitization, dependence, and withdrawal.

Multiple nicotinic receptor subtypes are expressed in the mesocorticolimbic system and may be linked to both acute and chronic effects of nicotine that underlie initiation and onset of dependence. Identification of these nAChR subtypes will provide insights into the pathophysiology of addiction.

Previous work has focused on knocking out specific nAChR subunit genes and assessing their role in nicotine dependence. β2 knock-out (KO) mice self-administer cocaine but fail to maintain self-administration when the cocaine is switched to nicotine, suggesting that β2* receptors are necessary for the maintenance of nicotine self-administration (12). Compared to wild-type mice, α4 and β2 KO mice exhibit striatal dopamine (DA) levels that do not increase in response to nicotine, supporting the idea that α4β2* nAChRs are necessary for dopamine release (12, 13), a nicotinic response thought to be involved in dependence.

We have taken a complementary approach to the KO strategy by replacing an endogenous exon with one containing a single point mutation, Leu9′ → Ala9′ (Leu9′Ala), within the putative pore-forming M2 domain of the α4 subunit, rendering α4* receptors hypersensitive to nicotine. In the absence of specific agonists, this “knock-in” Leu9′Ala mouse line allows for the selective activation of α4* receptors with small doses of nicotine that do not activate other nAChR subtypes. Although KO animals provide answers to the question of necessity, the hypersensitive knock-in approach allows us to address the question of sufficiency.

The Leu9′Ala mutation was introduced into an α4 knock-in construct, and Leu9′Ala mice were engineered by using exon replacement via homologous recombination as previously described (14). In contrast to the previous Leu9′ → Ser9′ line, both homozygous and heterozygous Leu9′Ala animals are viable and fertile. Mutant animals exhibit no gross developmental abnormalities and are born at close to the expected Mendelian ratios.

To assess mutant receptor expression levels, we quantified cytisine inhibition of [125I]-epibatidine binding in the superior and inferior colliculus, striatum, olfactory tubercles, hypothalamus, and midbrain (Fig. 1A). Heterozygous Leu9′Ala mice express between 71.3 and 89.6% of wild-type high affinity nAChRs, whereas homozygous mice express 40.6 to 77% of that of wild type. In homozygous mice, three of these brain regions [superior colliculus (P < 0.05, n = 3), inferior colliculus (P < 0.05, n = 3), and midbrain (P < 0.01, n = 3)] displayed significantly lower levels of cytisine-sensitive epibatidine binding compared with those of the wild type, possibly because of down-regulation in drugnaïve animals or residual interference with expression by the single remaining loxP site (15). Cytisine-insensitive binding, a measure of other non-α4β2 nAChR expression levels, was not significantly different between Leu9′Ala mutant and wild-type mice (Fig. 1B).

Fig. 1.

(A) Cytisine-sensitive and (B) -insensitive [125I]-epibatidine binding in Leu9′Ala homozygous, heterozygous, and wild-type (WT) mice. Whole particulate fractions prepared from brain regions dissected from WT and heterozygote and homozygous Leu9′Ala mice were incubated with 200 pM125I-epibatidine and varying concentrations of cytisine (0 nM and 10 concentrations from 0.1 nM to 3000 nM). After subtraction of nonspecific binding determined in the presence of 1 mM nicotine, results from each individual experiment were subjected to a nonlinear least squares fit to a two-site logistic binding model to estimate the amount of 125I-epibatidine binding sensitive to [median inhibitory concentration (IC50) about 2 nM] and less sensitive to (IC50 about 200 nM) inhibition by cytisine, as well as the IC50 values for these two components. Results represent mean ± SEM of three to four individual determinations for each brain region, and those values significantly different from WT mice are indicated by *P < 0.05 or **P < 0.01, one-way analysis of variance (ANOVA).

We recorded optical measurements of nicotine-induced calcium influx in cultured ventral midbrain neurons. Figure 2 illustrates Ca2+ influx in response to a 2-s application of nicotine (10 nM to 100 μM) in both 4-week-old wild-type and Leu9′Ala heterozygous cultures (Fig. 2, A and B, respectively). In mutant cultures, nicotine-elicited responses were observed at nicotine concentrations as low as 10 nM, a concentration that had no effect on wild-type cultures. The nicotine concentration response relation for mutant Ca2+ influx was shifted about 40-fold to the left compared with wild-type measurements (Fig. 2C). The median effective concentration (EC50) for mutant cultures was 0.12 ± 0.02 μM compared with 4.9 ± 10.4 μM for wild-type cultures. Mutant cultures exhibited a 4.6-fold increase in maximal response to ACh. The Hill coefficient was slightly higher in mutants at 0.64 ± 0.06 compared with 0.4 ± 0.15 for wild type. The low Hill coefficients most likely indicate at least two α4* receptor populations that differ in affinity to agonist.

Fig. 2.

Hypersensitive nicotine responses and functional up-regulation assayed via optically measured calcium influx in ventral midbrain neuronal cultures from Leu9′Ala heterozygous and WT mice. Average responses from 4-week-old WT (A) and Leu9′Ala (B) ventral midbrain cultures in response to 10 nM to 100 μM nicotine. (C) Concentration-response relations for Leu9′Ala and WT neurons measured in (A) and (B). Data were normalized to the peak response elicited by 100 μM nicotine in Leu9′Ala cultures and were fitted to the Hill equation. EC50 equalled 0.12 ± 0.02 μM for Leu9′Ala het cultures, nh = 0.64 ± 0.06. EC50 equalled 4.9 ± 10.4 μM for WT cultures, nh = 0.4 ± 0.15. Functional up-regulation: Four-week-old ventral midbrain neuronal cultures from (D) WT or (E) Leu9′Ala heterozygotes were incubated in 10 nM nicotine for 3 to 4 days. Average Ca2+ influx before (open circles) and after (solid squares) nicotine incubations are shown in response to 2-s applications of saturating doses of ACh: (D) 1 mM or (E) 100 μM. Recordings were taken in the presence of MLA. (Insets) (D) WT and (E) Leu9′Ala averaged Ca2+ response before and after nicotine incubation. Data represented as mean ± SEM. Significance was determined by one-way ANOVA. ***P < 0.001.

Chronic nicotine administration causes nAChR functional up-regulation (16). To determine whether low doses of nicotine could produce up-regulation in Leu9′Ala mice, we cultured mutant and wild-type ventral midbrain neurons and compared Ca2+ influx between cultures that had been incubated in only 10 nM nicotine, a concentration ∼50-fold less than found in smokers' blood (17). After 3 days of incubation in 10 nM nicotine, responses to 1 mM ACh were not changed in wild-type cultures compared to control non-incubated cultures [Fig. 2D, n values from 43 to 35, not significant (NS)]. However, Leu9′Ala cultures exhibited a robust increase in functional expression (Fig. 2E, n values from 33 to 48) in response to 100 μM ACh. The average influx of Ca2+ after nicotine incubation was 253 ± 12% of the control response (Fig. 2E, inset; n = 40; P < 0.001).

Nicotine activates VTA dopaminergic neurons, increasing action potential (AP) firing frequency and causing release of DA in the striatum (18). We recorded whole-cell currents to measure nicotinic responses in midbrain slices. VTA dopaminergic neurons were identified by anatomy and the presence of the hyperpolarizing activated potassium current, Ih (fig. S2A). Dopaminergic neurons from Leu9′Ala homozygous, heterozygous, and wild-type mice did not differ significantly in the magnitude of Ih, the resting membrane potential, or baseline action potential frequency (19) (n values from 5 to 8). Robust whole-cell nicotinic currents were recorded in response to 250-ms puffs of 1 μM nicotine in heterozygous (–210 ± 43 pA, P < 0.01, n = 8) and homozygous (–600 ± 125 pA, P < 0.001, n = 5) Leu9′Ala neurons, but only small responses were recorded in the wild type (–9 ± 2 pA, n = 6) (fig. S2C). In addition, these responses were completely blocked by 10 μM mecamylamine, a noncompetitive nicotinic-specific antagonist (fig. S2B, 95.9 ± 1.59% blockade, P < 0.01, n = 4). In current-clamp mode, 1 μM nicotine significantly increased the frequency (f) of AP firing in Leu9′Ala heterozygous (Δf = 2.8 ± 1.0 Hz, P < 0.05, n = 6) and homozygous (Δf = 11.2 ± 3.1 Hz, P < 0.01, n = 4) dopaminergic neurons at concentrations that had little effect on wild-type neurons (fig. S2D, Δf = 0.1 ± 0.2 Hz, n = 5).

We hypothesized that challenging mutant mice with small acute or chronic doses of nicotine that have little effect in wild-type animals would produce behavioral responses specifically corresponding to α4* nAChR-elicited events involved in nicotine addiction. On the basis of the slice electrophysiology, homozygous animals are significantly more sensitive to nicotine than heterozygous animals. Thus, we focused on Leu9′Ala homozygous mice for behavioral studies.

Acute nicotine administration can produce reinforcing effects in rodents depending on the concentration delivered (20, 21). This effect most likely represents the first stage of nicotine dependence. By using the conditioned place preference paradigm, we measured nicotine-induced reinforcement in wild-type and Leu9′Ala mice. In agreement with previous reports (22), wild-type mice exhibited a significant place preference for 0.5 mg/kg nicotine compared to saline (Fig. 3B, F = 12.8, P < 0.01, n = 5). We predicted that, in Leu9′Ala mice, if activation of α4* nAChRs is sufficient for a nicotine-induced reinforcement response, then a 50-fold lower nicotine dose would produce conditioned place preference. Confirming the hypothesis, a significant behavioral reinforcement response was measured in homozygous Leu9′Ala mice but not in wild-type animals at 10 μg/kg nicotine (Fig. 3D).

Fig. 3.

Nicotine-elicited reward behavior in Leu9′Ala homozygous and WT mice: Nicotine reinforcement was measured with the use of the conditioned place preference assay. (A) Bar graph representation of total time spent in black (solid) or white (open) compartments before (stage 1) or after (stage 3) nicotine training with 0.5 mg/kg nicotine in WT mice. Because WT mice had a slight preference for the black compartment before training, nicotine was paired with the less-preferred white chamber. (B) Difference score for WT mice injected with 0.5 mg/kg nicotine or saline. Bar graphs represent the time difference spent in each chamber before and after nicotine or saline pairings. (C) Graphs show total time spent in the black or white chamber during stage 1 in WT and Leu9′Ala homozygous mice. Mutant mice did not exhibit a preference for either chamber. (D) Difference score for WT and Leu9′Ala homozygous mice injected with 10 μg/kg nicotine and saline. All data are expressed as mean ± SEM. One-way ANOVA indicated a significant place preference for the nicotine-paired chamber in mutant mice (F = 18.252, P < 0.001, n = 10) but not in WT (F = 0.00189, NS, n = 7). In addition, two-way ANOVA between groups indicated a significant effect of treatment (F = 11.9, P < 0.01) but not genotype (F = 0.263, NS) and a significant interaction between the two (F = 12.1, P < 0.01). **P < 0.01, ***P < 0.001.

Multiple nAChR subtypes may be involved in reward behavior. VTA microinfusion of methyllcaconitine, an antagonist of (presumably homopentameric) α7 receptors, switches the valence of nicotine from reward to aversion, suggesting α7 receptors can modulate reward (23). Microinfusion of the high affinity nAChR antagonist, dihydro-β-erythroidine, on the other hand, blocks both the rewarding and aversive stimulus of nicotine (23). At a nicotine dose 50-fold less than that required to elicit reinforcing and reward behavior in wild-type mice, Leu9′Ala mice developed a significant place preference to nicotine. Therefore, selective activation of α4* nAChRs is sufficient for the reinforcing effects of nicotine.

Tolerance to nicotine is thought to play a critical role in the development and maintenance of dependence. We analyzed whether activation of α4* nAChR was sufficient for behavioral tolerance induction. Tolerance to nicotine-induced hypothermia in mice occurs with chronic nicotine infusion (7, 24, 25) or acutely with preinjections of large nicotine doses (4 mg/kg) 2 to 4 hours before a nicotine test dose (2 mg/kg) (26). Homozygous Leu9′Ala mice displayed hypothermia in response to small doses of nicotine (7.5 to 30 μg/kg). A 15 μg/kg intraperitoneal nicotine injection produced a 3.7° ± 0.3° decrease in temperature compared with saline injection (Fig. 4A, F = 58.42, P < 0.0001, n = 14). This dose of nicotine had no effect on wild-type mice (n = 11, NS). Preinjection of the nicotinic receptor antagonist, mecamylamine, blocked the effect (Fig. 4B, F = 8.35, P < 0.05, n = 6). The dose response relation for Leu9′Ala homozygous mice was shifted about 50-fold to the left at concentrations of 10 to 30 μg/kg for mutant compared with 500 to 1500 μg/kg for wild type (Fig. 4C).

Fig. 4.

Nicotine-induced hypothermia and tolerance in Leu9′Ala mice. (A) Representative homozygous Leu9′Ala temperature recording and response to a single intraperitoneal injection of 15 μg/kg nicotine. The nicotine injection is indicated by the arrow. (B) Temperature response elicited by intraperitoneal injection of 30 μg/kg nicotine 15 min after either a saline (left) or a mecamylamine (right, 1 mg/kg, n = 6) injection. (C) Dose-response relationship for WT (white squares) and Leu9′Ala homozygous (black squares) mice tested with nicotine. Each data point represents peak temperature drops from five to nine animals. (D) Nicotine-induced tolerance in Leu9′Ala homozygous mice. WT (n = 7) or Leu9′Ala (n values from 10 to 14) mice were given a 15 μg/kg nicotine injection intraperitoneally daily for 9 days. “Pre” represents the average response to saline injection 1 day before nicotine injections. (E) Tolerance to 30 μg/kg (triangles), 15 μg/kg (squares), 7.5 μg/kg (circles), and 3.25 μg/kg (diamonds) nicotine (n ≥ 5 for each treatment). Nicotine was administered as in (D). (F) Tolerance to 1500 μg/kg, 750 μg/kg, 500 μg/kg, or 15 μg/kg nicotine in WT mice (n ≥ 5 for each treatment). Nicotine was administered as in (D). In all experiments, WT and mutant mice were injected with saline once daily for ≥ 7 days before nicotine injections. Data are expressed as mean ± SEM in (B) to (F). Significance was measured via one-way ANOVA. *P < 0.05.

To assay tolerance, we monitored nicotine-induced hypothermia in response to single daily intraperitoneal injections of 15 μg/kg nicotine in homozygous Leu9′Ala and wild-type animals (Fig. 4D). Initial injections in mutant animals produced large decreases in temperature and had no effect in wild-type mice. However, by the eighth day of injections, the nicotine-induced temperature effect was significantly attenuated by 81.4 ± 9.0% (F = 29.04, P < 0.0001, day 1 compared to day 8, n values from 10 to 14), indicating the development of tolerance. In addition, significant tolerance occurred in mutant mice at all nicotine doses tested (30, 15, and 7.5 μg/kg, Fig. 4E). Wild-type mice tested with intraperitoneal injections of nicotine that yielded temperature decrements (Fig. 4F) only once per day did not develop tolerance, suggesting that the mutant animals were either hypersensitive to tolerance onset or that activation and/or desensitization of other nicotinic receptor subtypes distorted tolerance.

Sensitization or “reverse tolerance” of nicotine-induced motor activity occurs in response to repeated systemic injections in rats (27, 28). We assayed home cage locomotor activity in response to single daily injections of 15 μg/kg nicotine in Leu9′Ala homozygous and wild-type mice. Initial daily injections produced steadily increasing locomotor activity in mutant but not wild-type mice (figs. S3B, n = 12, and S3A, n = 7). By the seventh day of nicotine, there was a significant increase in nicotine-induced locomotor activity compared with baseline (F = 15.11, P < 0.001, n = 12), continuing through the ninth and final day of the experiment. The locomotor-stimulating effect could be blocked by preinjection with mecamylamine (fig. S3E, F = 9.90, P < 0.05, n = 7). In addition, significant sensitization was also seen in response to daily injections of the partial nicotinic agonist, cytisine (fig. S3D, F = 14.26, P < 0.01, n = 5). In wild-type rats and mice, sensitization is thought to be caused by nicotine-induced increases of dopamine in the nucleus accumbens. This is also likely in Leu9′Ala mice, because nicotine-induced locomotor responses in sensitized mice are attenuated by pretreatment with 1 mg/kg SCH23390, a D1 dopamine receptor antagonist (fig. S3F, F = 35.03, P < 0.001, n = 5), but not by the D2 receptor antagonist, sulpiride (19).

Fura-2 and slice electrophysiology data from the Leu9′Ala line indicate that (i) hypersensitive Leu9′Ala α4* nAChRs are functionally expressed in VTA DA midbrain neurons in mutant animals, (ii) one can selectively and potently activate these receptors by applying low doses of agonist that do not activate other subtypes, and (iii) chronic exposure to sub-threshold levels of nicotine cause functional up-regulation. Dependence-related behaviors, including reward, tolerance, and sensitization, occur strongly and at remarkably low nicotine doses in the Leu9′Ala mice, rendering this strain an excellent candidate for studies on molecular, behavioral, and pharmacological aspects of nicotine addiction. Data from the gain-of-function α4* nAChR knock-in mice complement earlier KO experiments by providing insights into nAChR subtypes sufficient for nicotine-elicited events important for dependence. Our data indicate that α4* nAChR activation is sufficient for nicotine-induced reward, tolerance, and sensitization.

Supporting Online Material

Materials and Methods

Figs. S1 to S3


References and Notes

Stay Connected to Science

Navigate This Article