Rodent Habenulo–Interpeduncular Pathway Expresses a Large Variety of Uncommon nAChR Subtypes, But Only the α3β4 ^{and α3β3β4 Subtypes Mediate Acetylcholine Release}

Abstract

Recent studies suggest that the neuronal nicotinic receptors (nAChRs) present in the habenulo–interpeduncular (Hb–IPn) system can modulate the reinforcing effect of addictive drugs and the anxiolytic effect of nicotine. Hb and IPn neurons express mRNAs for most nAChR subunits, thus making it difficult to establish the subunit composition of functional receptors. We used immunoprecipitation and immunopurification studies performed in rat and wild-type (+/+) and β2 knock-out (−/−) mice to establish that the Hb and IPn contain significant β2* and β4* populations of nAChR receptors (each of which is heterogeneous). The β4* nAChR are more highly expressed in the IPn. We also identified novel native subtypes (α2β2*, α4β3β2*, α3β3β4*, α6β3β4*). Our studies on IPn synaptosomes obtained from +/+ and α2, α4, α5, α6, α7, β2, β3, and β4^−/− mice show that only the α3β4 and α3β3β4 subtypes facilitate acetylcholine (ACh) release. Ligand binding, immunoprecipitation, and Western blotting studies in β3^−/− mice showed that, in the IPn of these mice, there is a concomitant reduction of ACh release and α3β4* receptors, whereas the receptor number remains the same in the Hb. We suggest that, in habenular cholinergic neurons, the β3 subunit may be important for transporting the α3β4* subtype from the medial habenula to the IPn. Overall, these studies highlight the presence of a wealth of uncommon nAChR subtypes in the Hb–IPn system and identify α3β4 and α3β3β4, transported from the Hb and highly enriched in the IPn, as the subtypes modulating ACh release in the IPn.

Introduction

Evidence suggests the habenular complex (Hb) participates in various behaviors, including nociception, sleep–wake cycle, locomotion, anxiety-related responses, mood regulation, learning, and memory (De Biasi and Salas, 2008). Its role in reward phenomena has recently attracted much attention. The Hb and its inputs and outputs support self-stimulation (Morissette and Boye, 2008). Indeed, the Hb is activated by negative reward or absence of positive reward in humans (Shepard, 2006) and monkey (Matsumoto and Hikosaka, 2007). Stimulation of the lateral Hb (LHb) suppresses dopamine neuron activity when an expected reward is withheld in rat (Ji and Shepard, 2007) and monkey (Matsumoto and Hikosaka, 2007). The medial Hb (MHb) projects almost exclusively to the interpeduncular nucleus (IPn) through the fasciculus retroflexus (fr) (Herkenham and Nauta, 1979; Klemm, 2004), whereas the LHb projects to a number of midbrain regions (Klemm, 2004; Lecourtier and Kelly, 2007; Geisler and Trimble, 2008).

The Hb and IPn are both highly enriched in cholinergic markers. The Hb receives cholinergic input from septal and basal forebrain nuclei through the stria terminalis (Contestabile and Fonnum 1983). Cholinergic cells, densely packed in the ventral portion of MHb, innervate rostral, central, and intermediate subnuclei of the IPn (Kimura et al., 1981; Houser et al., 1983; Contestabile et al., 1987; Eckenrode et al., 1987). The IPn also receives direct cholinergic input from forebrain nuclei through the fr bypassing the Hb (Contestabile and Fonnum, 1983; Fonnum and Contestabile, 1984; Albanese et al., 1985; Klemm, 2004):

The wealth of cholinergic neurons in the habenulo–interpeduncular pathway is paralleled by a high density of nicotinic acetylcholine receptors (nAChRs). The MHb and IPn express the highest density of nicotinic binding sites in mammalian brain and have the largest variety of nicotinic subunits (Clarke et al., 1985; Wada et al., 1989; Marks et al., 1992; Le Novère et al., 1996; Han et al., 2000, 2003; Whiteaker et al., 2000; Yeh et al., 2001; Cui et al., 2003). Electrophysiological and pharmacological experiments in rats or wild-type (+/+) and nicotinic subunit knock-out (−/−) mice have shown that nAChRs in MHb and IPn are heterogeneous, including both β2* and β4* receptors (Mulle et al., 1991; Connolly et al., 1995; Zoli et al., 1998; Quick et al., 1999; Whiteaker et al., 2002; Adams et al., 2004; Fonck et al., 2009). Although many nAChRs are presynaptic, some are postsynaptic in the IPn (Clarke et al., 1986; Mulle et al., 1991).

The habenulo–interpeduncular system may mediate some nicotine-induced behaviors. Effects of acute nicotine administration in rodent MHb and IPn include increases in Fos immunoreactivity (Seppa et al., 2001; Choi et al., 2006; Fonck et al., 2009) and cerebral blood volume (Choi et al., 2006). nAChRs in MHb and IPn may partially mediate locomotor effects of nicotine (Hentall and Gollapudi, 1995) and self-administration of addictive drugs (Glick et al., 2006; Taraschenko et al., 2007). However, functional roles of nAChRs in MHb and IPn remain essentially unknown. Aims of this study were to determine native nAChR subtypes in this system and define selected subtype function.

Materials and Methods

Animals.

Adult male pathogen-free Sprague Dawley rats were obtained from Harlan-Nossan. Rat animal care and experimental procedures were approved by the University of Modena and Reggio Emilia and are in accordance with the European Communities Council Directive of November 24, 1986 (86/609/EEC) Mice with nAChR subunit null mutations were bred and maintained at The Institute for Behavioral Genetics (Boulder, CO) in accordance with the guidelines and approval of the Animal Care and Utilization Committee of the University of Colorado, Boulder. All mice were generated from heterozygous matings, and genotype was determined by PCR analysis of tail DNA (Salminen et al., 2004). Original sources of the mutated mice were as follows: Dr. Jim Boulter (University of California, Los Angeles, Los Angeles, CA) (α2), Dr. John Drago (University of Melbourne, Victoria, Australia) (α4), Dr. Arthur Beaudet (Baylor College, Houston, TX) (α5, α7, β4), Dr. Uwe Maskos (Pasteur Institute, Paris, France) (α6), Dr. Marina Picciotto (Yale University, New Haven, CT) (β2), and Dr. Stephen Heinemann (Salk Institute, San Diego, CA) (β3).

Materials.

(+/−)[³H]Epibatidine ([³H]Epi) (specific activity, 70.6 Ci/mmol), [¹²⁵I]αbungarotoxin ([¹²⁵I]αBgtx) (specific activity, 200 Ci/mmol), [³H]choline (methyl-³H) (60–90 Ci/mmol), and ⁸⁶Rb⁺ (initial specific activity, ∼12 Ci/mg) were purchased from PerkinElmer Life and Analytical Sciences. The nonradioactive αBgtx, epibatidine, acetylcholine iodide, diisopropylfluorophosphate, phenylmethylsulfonylfloride (PMSF), EDTA, EGTA, bovine serum albumin (BSA), atropine sulfate, acetylcholine iodide, tetrodotoxin, Tris, Triton X-100, Tween 20, glucose, nicotine sulfate, NaCl, KCl, MgSO₄, CaCl₂, and polyethylenimine were purchased from Sigma.

Antibody production and characterization.

The subunit-specific polyclonal antibodies (Abs) used were produced in rabbit against peptides derived from the C-terminal (COOH) and/or intracytoplasmic loop (CYT) regions of rat, human, or mouse subunit sequences and affinity purified as described previously (Zoli et al., 2002) Most of the Abs have been described previously (Zoli et al., 2002; Champtiaux et al., 2003; Gotti et al., 2005a,b, 2008) (supplemental Table 1A, available at www.jneurosci.org as supplemental material).

Antibody specificity was checked by means of quantitative immunoprecipitation or immunopurification experiments using nAChRs from different areas of the CNS of wild-type (+/+) and null mutant (−/−) mice, which allowed selection of Abs specific for the subunit of interest and established the immunoprecipitation capacity of each Ab (Zoli et al., 2002; Gotti et al., 2005a,b, 2007) (supplemental Table 1B, available at www.jneurosci.org as supplemental material).

Ab specificity was also tested by Western blotting (Gotti et al., 2008), which indicated that some of the Abs were less specific for Western blotting than for immunoprecipitation.

Preparation of membranes and 2% Triton X-100 extracts from Hb and IPn.

The tissues obtained from rats or mice were dissected, immediately frozen on dry ice or in liquid nitrogen, and stored at −80°C for later use. In every experiment, the tissues from Hb (0.05–0.20 g) or IPn (0.05–0.10 g) were homogenized in 5 ml of 50 mm Na phosphate, pH 7,4, 1 m NaCl, 2 mm EDTA, 2 mm EGTA, and 2 mm PMSF with a Potter homogenizer. The homogenates were then diluted and centrifuged for 1.5 h at 60,000 × g.

The procedures of homogenization, dilution, and centrifugation of the total membranes were performed twice, after which the pellets were collected, rapidly rinsed with 50 mm Tris HCl, pH 7, 120 mm NaCl, 5 mm KCl, 1 mm MgCl₂, 2.5 mm CaCl₂, and 2 mm PMSF, and then resuspended in the same buffer containing a mixture of 20 μg/ml of each of the following protease inhibitors: leupeptin, bestatin, pepstatin A, and aprotinin. Triton X-100 at a final concentration of 2% was added to the washed membranes, which were extracted for 2 h at 4°C. The extracts were then centrifuged for 1.5 h at 60,000 × g, supernatants were recovered, and an aliquot of these resultant supernatants was collected for protein measurement using the BCA protein assay (Pierce) with bovine serum albumin as the standard.

Selective [³H]epibatidine binding.

To ensure that the α7 nAChRs did not contribute to [³H]Epi binding, in both membrane or solubilized extracts, binding was performed in the presence of 2 μm αBgtx, which specifically binds to α7* nAChR (and thus prevents [³H]Epi binding to these sites).

Binding to the homogenates obtained from Hb or IPn membranes was performed overnight by incubating aliquots of the membrane with 2 nm [³H]Epi at 4°C. Nonspecific binding (averaging 5–10% of total binding) was determined in parallel samples containing 100 nm unlabeled epibatidine. At the end of the incubation, the samples were filtered on a glass microfiber filter GF/C (Whatman) filter soaked in 0.5% polyethylenimine, washed with 15 ml of buffer (Na phosphate, 10 mm, pH 7.4, and 50 mm NaCl), and counted in a β counter.

The Triton X-100 extracts were labeled with 2 nm [³H]epibatidine. Tissue extract binding was performed using DE52 ion-exchange resin (Whatman) as described previously (Champtiaux et al., 2003).

[¹²⁵I]αBungarotoxin binding.

Binding experiments were performed by incubating Hb or IPn membranes overnight with 2–4 nm [¹²⁵I]αBgtx at 20°C in the presence of 2 mg/ml BSA. Specific radioligand binding was defined as total binding minus nonspecific binding determined in the presence of 1 μm cold αBgtx.

Immunoprecipitation by anti-subunit-specific antibodies of [³H]Epi-labeled Hb or IPn receptors.

Hb or IPn dissected from 30–40 rats or mice differing in β3 or β2 genotypes were immediately frozen at − 70°C. The membranes (10–30 ml) were then extracted by addition of 2% Triton X-100 as described above and centrifuged. The extracts (100 μl) were labeled with 2 nm [³H]Epi and incubated overnight with a saturating concentration of anti-subunit affinity-purified IgG (anti-α2, -α3, -α4, -α5, -α6, -β2, -β3, or -β4). Each immunoprecipitate was recovered by incubating the samples with beads containing bound anti-rabbit goat IgG (Technogenetics). The level of immunoprecipitation with each Ab was expressed as the percentage of [³H]Epi-labeled receptors immunoprecipitated by the Abs (taking the amount present in the Triton X-100 extract solution before immunoprecipitation as 100%) or as femtomoles of immunoprecipitated receptors per milligram of protein.

To immunopurify the β2* nAChRs, extracts (30 ml) were incubated three times with 5 ml of Sepharose-4B with bound anti-β2 CYT Abs. The β2-depleted flow-through fraction was collected for additional processing. The bound β2* population was eluted by incubation with 100 μm β2 CYT peptide. The purified β2 recovered populations (or the flow through devoid of β2-containing receptors) were analyzed by immunoprecipitation using subunit-specific Abs.

Immunoblotting and densitometric quantification of Western blot bands.

The membranes obtained from mouse Hb and IPn were diluted 1:1 (v/v) with Laemli's buffer and then underwent SDS-PAGE using 9% acrylamide by loading 5 μg of membrane protein for both the Hb and IPn (except for the analysis of the β4 subunit in which only 2.5 μg of IPn membrane protein was loaded). After SDS-PAGE, the proteins were electrophoretically transferred to nitrocellulose membranes with 0.45 μm pores (Schleicher and Schuell). The blots were blocked overnight in 5% nonfat milk in Tris-buffered saline (TBS), washed in a buffer containing 5% nonfat milk and 0.3% Tween 20 in TBS, incubated for 2 h with the primary antibody (1–2.5 μg/ml), and then incubated with the appropriate peroxidase conjugated secondary Abs. After another series of washes, peroxidase was detected using a chemiluminescent substrate (Pierce).

The signal intensity of the Western blot bands was acquired using an Epson 4500 gel scanner. The developed films were scanned as a Tiff scale in eight-bit grayscale format at setting of 300 dpi. All of the films obtained from the separate experiments of different genotypes were acquired in the same way and scanned in parallel with the calibrated optical density step tablet from Stouffer. The images were analyzed using NIH ImageJ software (National Technical Information Service). The pixel values of the images were transformed to optical density values by the program using the calibration curve obtained by acquiring the calibrated tablet with the same parameters as those used for the images.

The immunoreactive bands were quantified in four separate experiments for both Hb and IPn. The optical density ratio was calculated by taking the optical density of the wild-type β3 as 1. The values are the mean ± SEM of four separate experiments for each tissue for each genotype

Release experiments.

For both [³H]ACh release and ⁸⁶Rb⁺ efflux, crude synaptosomes were prepared from freshly dissected IPn and Hb. For [³H]ACh release experiments, uptake of [³H]choline into crude IPn synaptosomes as well as perfusion and release procedures of Grady et al. (2001) were used with the following modifications. Perfusion speed was 0.7 ml/min, with a 10 min wash period before stimulation with ACh or high potassium for 20 s. Fractions (10 s) were collected into 96-well plates using a Gilson FC204 fraction collector with multicolumn adapter, and radioactivity was determined after addition of 150 μl of Optiphase Supermix mixture using a 1450 Microbeta scintillation counter (both from PerkinElmer Life and Analytical Sciences). Instrument efficiency was 20%. Baseline counts per minute were subtracted from peak samples, and then peak samples were summed and normalized to baseline to give units of release.

For ⁸⁶Rb⁺ efflux experiments, the methods of Marks et al. (2002) were followed for uptake, superfusion, and detection of efflux. For these experiments, crude synaptosomes were prepared from IPn and Hb of β2 and β4 subunit null mutant genotypes (+/+, +/−, −/−).

Results

Because rats have been consistently used for behavioral and lesioning studies of the Hb–IPn system and mice for the availability of genetically modified animals, we analyzed in detail the nAChR subtype content and composition of Hb and IPn in both species.

Subunit composition of nAChRs in rat Hb

We first determined the proportions of αBgtx-sensitive ([¹²⁵I]αBgtx binding) and αBgtx-insensitive ([³H]Epi binding in the presence of 2 μm αBgtx) nAChRs in homogenates of rat Hb. [¹²⁵I]αΒgtx binding (33.0 ± 13.2 fmol/mg protein) was much less than [³H]Epi binding (461.0 ± 26.2 fmol/mg protein).

To determine the subunit composition of the αBgtx-insensitive nAChRs in the Hb, we immunoprecipitated [³H]Epi-labeled receptors with subunit-specific Abs. Figure 1A and supplemental Table 2 (available at www.jneurosci.org as supplemental material) show the mean immunoprecipitation values obtained in four to five separate experiments for each subunit. By quantifying the number of receptors immunoprecipitated by the specific Ab as the percentage of the total number of [³H]Epi receptors, we found that adult rat Hb contains detectable levels of almost all of the αBgtx-insensitive nAChR subunits: high levels of α3, α4, β2, β3, and β4, moderate levels of α5, and no statistically significant levels of α2 and α6 subunits. The sum of the immunoprecipitated β2* and β4* (femtomoles per milligram of protein) [³H]Epi-labeled receptors (414) was very close to the total number of [³H]Epi labeled receptors (385) in the total extract, thus suggesting that β2* and β4* nAChRs are two primarily independent receptor populations in the Hb.

Figure 1.

Subunit composition of rat Hb and IPn nAChRs before and after immunodepletion of β2* receptors and mouse Hb and IPn nAChR from wild-type and β2 subunit null mutant mice. Two percent Triton X-100 extracts were prepared as described in Materials and Methods. Solubilized HB and IPn membrane extracts from rat before (white) and after (gray) immunodepletion of β2* receptors were preincubated with 2 nm [³H]Epi and immunoprecipitated with the indicated subunit-specific Abs as described in Materials and Methods using saturating concentrations (20 μg) of anti-subunit Abs. The amount immunoprecipitated by each antibody was subtracted from the value obtained in control samples containing an identical concentration of normal rabbit IgG, and the results are expressed as femtomoles of immunoprecipitated labeled [³H]Epi nAChR per milligram of protein. Results are the mean ± SEM values of four to five experiments performed in triplicate per antibody. Rat extracts are shown in A and B. Extracts from solubilized mouse +/+ (white) and −/− (black) Hb and IPn are shown in C and D. Statistical analysis was performed by t test. Each Ab in control animals was compared with the same Ab tested in the β2*-depleted extract by means of paired Student's t test, * p < 0.05; **p < 0.01; ***p < 0.001

To examine the composition of the β2* and β4* receptor populations, we immunodepleted the habenular extract of β2* receptors using an affinity column bearing an anti-β2 antibody. Selective β2* nAChR immunodepletion was confirmed by the fact that the number of immunoprecipitated β2-containing [³H]Epi-labeled receptors decreased from >60% in the total habenular extract to <1% in the flow through of the anti-β2 antibody column (Fig. 1A) (supplemental Table 2, available at www.jneurosci.org as supplemental material). To identify the subunit composition of the β4* receptors, we immunoprecipitated [³H]Epi-labeled receptors from the flow through of the anti-β2 column by means of subunit-specific antisera. The flow through showed a considerable decrease in the number of the α4* and α5* receptors, a significant but more modest reduction in β3* receptors, but no significant reduction in the number of α3* and β4* receptors; the relative proportion of the sparsely expressed α2* and α6* receptors was only slightly modified (Fig. 1A) (supplemental Table 2, available at www.jneurosci.org as supplemental material) and did not differ significantly from zero.

Combining the results obtained from the total and β2-depleted rat Hb extracts, it can be concluded that the β2* and β4* nAChR populations are quantitatively similarly represented in Hb with (1) β2* subunits contribute to 51% of the total habenular nAChRs, and 78% of these contain the α4 subunit (40% of the total), and 14% of the β2* nAChRs were associated with the α3 subunits (7% of the total); and (2) β4* subunits contribute to 49% of total habenular nAChRs, 88% of these contain the α3 subunit (43% of the total), and 24% of β4* nAChRs were associated with the α4 subunits (12% of the total) (supplemental Table 2, available at www.jneurosci.org as supplemental material).

The β2* and β4* receptor populations of rat Hb both contain the accessory α5 and β3 subunits but to different extents: higher levels of the α5 subunit are associated with the β2* receptor (41% of β2* vs 10% of β4* nAChRs) and higher levels of the β3 subunit with the β4* receptors (45% of β4* vs 26% of β2* nAChRs).

Subunit composition of nAChRs in rat IPn

Binding studies revealed that the level of [³H]Epi (974.1 ± 65.8 fmol/mg protein) was much higher than that of [¹²⁵I]αBgtx binding (37.4 ± 5.2 fmol/mg protein) and also higher than the levels of [³H]Epi binding in the habenular extract (see above).

Immunoprecipitation of [³H]Epi-labeled receptors by means of subunit-specific antisera showed that adult rat IPn contains all of the nAChR subunits except α6: namely, high levels of α3, α4, β2, β3, and β4, and moderate levels of α2 and α5 (Fig. 1B) (supplemental Table 2, available at www.jneurosci.org as supplemental material). As in the Hb, the sum of the immunoprecipitated β2* and β4* [³H]Epi-labeled receptors (742.8) was very close to the number of [³H]Epi-labeled receptors in the total IPn extract (746.0), thus suggesting that β2* and β4* nAChRs are two primarily independent receptor populations in the IPn. However, unlike the Hb, the IPn has more β4* than β2* receptors (supplemental Table 2, available at www.jneurosci.org as supplemental material).

After immunodepleting the β2* receptors, the immunoprecipitation of the flow through showed an almost total depletion of the α2 subunit, marked depletion of the α4 and α5 subunits, and a slight decrease in the number of receptors containing the α3, β3, and β4 subunits There are, therefore, two main nAChR populations in the IPn: (1) β2 subunits contribute to 38% of IPn nAChRs, and 35% of these are associated with the α2 subunit (13% of the total), 33% with the α3 subunit (13% of total), and 54% with the α4 subunit (21% of the total); and (2) β4 subunits contribute to 62% of IPn nAChRs, and of these 99% are associated with the α3 subunit (61% of the total) and 32% with the α4 subunit (20% of the total).

In addition, in the rat IPn, the β2* and β4* receptors are associated with α5 and β3 accessory subunits: 46% of β2* and 16% of β4* nAChRs contain an α5 subunit, whereas 55% of the β4* and 18% of β2* nAChRs contain a β3 subunit (supplemental Table 2, available at www.jneurosci.org as supplemental material) (see Fig. 6A).

Figure 6.

Diagrams reporting the content of nAChR subtypes in percentage of the total number of nAChRs in the rat and mouse Hb (A) and IPn (B). Note that, in the instances of β2* nAChRs of rat and mouse interpeduncular nucleus and β4* nAChRs of mouse interpeduncular nucleus, the sum of the α subunits is larger than the total number of receptors. In these cases, the existence of subtypes with different ligand-binding interfaces must be postulated, although present analysis does not allow their identification.

Subunit composition of immunopurified rat β2* nAChRs from Hb and IPn

To further characterize the possible β2* receptor subtypes present in the Hb and IPn, we identified the subunits assembled with the β2 subunit by immunoprecipitating the β2* receptors bound by anti-β2 antibodies. The captured β2* nAChRs were eluted from the affinity column using an excess of the β2 CYT peptide, labeled with [³H]Epi, and then immunoprecipitated with subunit specific antisera. As shown in Figure 2, the subunit-specific Abs revealed that the mean ± SEM percentages (n = 3) of the purified [³H]Epi-labeled β2* nAChRs from rat Hb containing the different subunits were 1.6 ± 1.1% (α2), 26.6 ± 11% (α3), 85.4 ± 3.2% (α4), 26.6 ± 14.7% (α5), 4.0 ± 4.0% (α6), 100.0% (β2), 23.5 ± 5.8% (β3), and 13.5 ± 2.4% (β4) (Fig. 2A).

Figure 2.

Immunoprecipitation analysis of the subunit content of purified β2* nAChRs from Hb and IPn of rat. Extracts prepared from control rat Hb or IPn were incubated on an affinity column with bound anti-β2 CYT Abs (see Materials and Methods) to bind the β2* population, which was eluted from the column by means of incubation with the β2 CYT peptide. The recovered β2* nAChRs were labeled with 2 nm [³H]Epi and then immunoprecipitated by the indicated subunit-specific Abs. The results are expressed as the percentage of the immunoprecipitation obtained with the β2 Abs and are the mean ± SEM values of two to three immunopurification experiments from each tissue.

This immunopurification experiment qualitatively and quantitatively confirms that, in Hb, the β2 subunit is principally associated with the α4 and to a much lower extent with the α3 subunit. Only a small fraction of the β2* receptors have more than one principal ligand-binding subunit (α2, α3, α4, or α6). Levels of α2 and α6 subunits do not differ significantly from zero.

The same immunopurification and immunoprecipitation experiments on the rat IPn showed that the mean ± SEM percentage (n = 3) of the purified [³H]Epi-labeled β2* nAChRs containing the different subunits were as follows: 26.9 ± 3.8% (α2), 38.7 ± 16.0% (α3), 60.7 ± 7.4% (α4), 35.9 ± 5.6% (α5), 0 ± 0.0% (α6), 100.0% (β2), 20.9 ± 6.3% (β3), and 36.8 ± 8% (β4) (Fig. 2B). This immunopurification experiment established that the recovery of the α2, α3, and α4 ligand-binding subunits was higher than that of the β2 subunit alone, thus indicating the presence of receptors with a complex subunit composition such as α2α4β2*, α2α3β2*, and/or α3α4β2*. However, the complexity of these receptors is even greater because some of them also contain the accessory subunits α5 and β3, and, in addition, a fraction appears to contain the β4 subunit.

Subunit composition of nAChRs in mouse Hb and IPn

When the number of receptors immunoprecipitated by the specific Abs was expressed as femtomoles per milligram of protein, we found that, similar to rat, Hb of adult wild-type mouse contains high levels of α3, α4, β2, β3, and β4, and low levels of α2 and α6. A difference in expression was noted for α5, which was markedly less concentrated in mouse than in rat Hb (9% in mouse and 28% in rat) (Fig. 1A,C).

To distinguish the β2* and β4* receptors, we compared immunoprecipitation experiments using subunit-specific antibodies in both β2^−/− and β2^+/+ mice (Fig. 1C) (supplemental Table 3, available at www.jneurosci.org as supplemental material). As in the case of β2-immunodepleted extracts of rat Hb, mouse β2^−/− Hb extracts were devoid of β2 as expected, and α4 (−90%) and β3 (−30%) subunits were significantly decreased. Similarly to rat, levels of α2, α3, α6, and β4 were not significantly changed. Unlike the rat β2-immunodepleted extracts, in which significant decrease in the α5 subunits was observed, no statistically significant decrease in α5 was observed in β2^−/− mice.

We repeated the same approach using β2^+/+ and β2^−/− IPn extracts, and we found that the β2* and β4* population were present at similar levels in the +/+, whereas, as expected, there was no β2* population remaining in the β2^−/− extract. In addition, the β2^−/− extracts showed a selective loss of receptors containing the α2 (−100%), α3 (−47%), and α4 (−56%) subunits. This pattern differs from the rat in that α5 subunits are unchanged, whereas α3 subunits are decreased.

nAChR subtypes in rat and mouse Hb and IPn

Assuming the current stoichiometric rules on nAChR composition (Gotti et al., 2007) and the near 100% immunoprecipitating efficacy of our subunit-specific antibodies (for caveats, see Materials and Methods), some additional quantitative deductions can be made (supplemental Tables 2, 3, available at www.jneurosci.org as supplemental material) (see Fig. 6A,B); note that, when two percentages are in brackets, the first refers to rat and the second to mouse.

Habenula

In both rat and mouse Hb, β2* nAChRs consist of a major α4β2* (78 and 79%), a minor α3β2 (14 and 13%), and two very small α2β2* and α6β2* (<5%) populations. More than half of the β2* receptors in rat Hb contain either α5 (40%) or β3 (26%) accessory subunits, in mouse Hb β3 subunit is present in 12% of β2* receptors, whereas only a few β2* receptors contain the accessory α5 subunit (7%)

In both rat and mouse Hb, β4* nAChRs consist of a major α3β4* (89 and 100%) and a smaller α4β4* (24 and 17%) population. Because the sum of α3β4 and α4β4 clearly exceeds 100%, we hypothesize the presence of a mixed α3α4β4 receptor population in both rat (13%) and mouse (17%) Hb. This implies, indeed, that (non-α3)α4β4* receptors may not exist in mouse. Many of these receptors also contain β3 (45 and 56%) or α5 (11 and 15%) accessory subunits.

Interpeduncular nucleus

In both rat and mouse IPn, there are significant populations of β2* nAChRs with α2 (35 and 33%), α3 (33 and 47%), and α4 (54 and 44%) subunits. Very few of the β2* nAChRs contain α6 subunits. Both rat and mouse IPn have large populations of β4* nAChRs combined with α3 (99 and 67%) or α4 (32 and 44%) subunits. In addition, the mouse shows a minor α6β4* population (12%). None of the β4* nAChRs in the IPn contain the α2 subunit. Given that the sum of the ligand-binding α subunits (α2, α3, α4, α6) is higher than the immunoprecipitation value of the β2 or β4 subunits, we have to assume that some receptors contain two different ligand-binding interfaces and form complex subtypes. In addition, many β2* receptors in rat IPn also contain α5 (47%), whereas none of the mouse β2* nAChRs do. β3 is at moderate to low level (18 and 8%) in β2* nAChRs of both species. Many β4* nAChRs of IPn also contain β3 (55 and 83%) or α5 (16 and 35%) accessory subunits.

Effects of the genetic deletion of the β3 subunit on nAChR subunit composition in the Hb and IPn

In situ hybridization studies have shown that the MHb expresses high levels of β3 subunit mRNA (Deneris et al., 1989; Le Novère et al., 1996). The β3 subunit is considered important not only for its role in determining the biophysical characteristics of the receptors containing this subunit (Broadbent et al., 2006; Kuryatov et al., 2008) but also for its role in determining assembly and/or transport of receptors (Gotti et al., 2005a; Drenan et al., 2008). Our immunoprecipitation and immunopurification studies showed that the Hb and IPn are enriched in β3* nAChRs, particularly an α3β3β4* subtype that accounts for more than half of the β4* nAChRs in the Hb and IPn of both rat and mouse. Therefore, in a series of experiments, we investigated the contribution of the β3 subunit to the composition of nAChRs in the habenulo–interpeduncular system by means of immunoprecipitation and Western blotting studies of the Hb and IPn of β3^+/+ and β3^−/− mice using subunit-specific antibodies.

Immunoprecipitation

The proportion of [³H]Epi-labeled β3* nAChR in the habenular extract was decreased from 26% of the total in the β3^+/+ mice to only 1% in the β3^−/− mice; however, there was no significant difference in either the total [³H]Epi-labeled extract or the relative amount of non-β3 subunit-specific immunoprecipitates between β3^+/+ and β3^−/− mice. However, although the proportion of immunoprecipitated [³H]Epi-labeled β3* nAChR in the IPn was 32% of the total in the β3^+/+ mice and <1% in the β3−/− mice, this decrease was paralleled by a decrease in [³H]Epi labeling in total extract and in the relative content of several subunits: α3 (−39%), α4 (−26%), β2 (−34%), and β4 (−44%) subunits (Fig. 3A).

Figure 3.

Characterization of the nAChR subunit content of the Hb and IPn from β3 ^+/+ and β3 ^−/− mice. A, Subunit composition of mouse habenula and nucleus interpeduncularis of β3^+/+ and β3^−/− mice determined by immunoprecipitation. Results are the mean ± SEM values of three to four experiments performed in triplicate per antibody. Statistical analysis was performed by t test. Each Ab in β3^+/+ mouse was compared with the same Ab tested in the β3^−/− extract by means of paired Student's t test, *p < 0.05; **p < 0.01; ***p < 0.001. B, The subunit content of habenula and nucleus interpeduncularis membranes from β3^+/+ and β3^−/− mice was determined by Western blot analysis. The proteins (5 μg) for all except the β4 subunit (2.5 μg) were separated on 9% acrylamide SDS gels, electrotransferred to nitrocellulose, probed with 1–2.5 μg/ml of the indicated primary Abs, and then incubated with the secondary Ab (anti-rabbit conjugated to peroxidase; dilution, 1:40,000). The bound Abs were revealed by a chemiluminescent substrate (Pierce). C, Comparison of the quantitated results obtained in the habenula and nucleus interpeduncularis of β3^+/+ and β3^−/− mice by immunoprecipitation (gray) and Western blotting (black). The ratios (mean values ± SEM) between β3^+/+ (white) and β3^−/− measured by immunoprecipitation (IP; gray) and β3^−/− or Western blotting (WB; black) were calculated by taking the immunoprecipitation or optical density values of the wild type as 1 and are the results of three to four separate experiments for each tissue and genotype. Statistical analysis was performed by paired t test. Each Ab in β3^+/+ was compared with the same Ab tested in the β3^−/− mice by means of paired Student's t test, *p < 0.05; **p < 0.01. KO, Knock-out.

Western blotting

Samples obtained from the Hb and IPn of β3^+/+ and β3^−/− mice were analyzed by loading the same amount of membrane protein (except for the β4 sample in which the IPn protein loaded was half of Hb). Figure 3B shows the typical anti-α3, anti-α4, anti-β2, anti-β3, and anti-β4 Ab labeling of samples.

To further define the difference in subunit composition between β3^+/+ and β3^−/− mice and to compare the results obtained in immunoprecipitation and Western blotting, the α3, α4, β2, β3, and β4 subunit signal intensity in the Hb and IPn of the two genotypes were expressed as the ratio of Ab labeling (taking the amount present in the samples from wild-type mice as 1.0) (Fig. 3C). The results are from four independent experiments using three separate membrane preparations. The immunoprecipitation and Western blotting data showed that, in the Hb, there were no significant changes in the levels of any of the subunits besides β3: the mean ± SEM optical density ratios of the subunits were 1.23 ± 0.11 (α3 subunit), 1.11 ± 0.08 (α4 subunit), 1.12 ± 0.11 (β2 subunit), and 1.18 ± 0.09 (β4 subunit). However, in the case of the IPn, the ratios ± SEM were as follows: 0.68 ± 0.09 (α3 subunit, p = 0.006), 0.82 ± 0.05 (α4 subunit, p = ns), 0.71 ± 0.05 (β2 subunit, p = 0.03), and 0.42 ± 0.07 (β4 subunit, p = 0.0013), thus indicating a significant decrease in the amount of receptors containing these subunits. Immunoprecipitation and Western blotting approaches are compared in Figure 3C and found to give highly concordant results.

In conclusion, our data showed that the lack of the β3 subunit differentially affects the levels of the nAChR subtypes in the Hb and IPn: there is no change in the level of heteromeric receptors in the Hb, but a significant decrease is seen in the IPn, of mainly α3β4* receptors with a lesser decrease in α4β2*.

Effects of the genetic deletion of α2, α4, α5, α6, α7, β2, β3, or β4 subunits on acetylcholine release from IPn synaptosomes

Previously published data have shown that the nAChR mediating release of ACh from IPn synaptosomes is not affected by the β2 null mutation (Grady et al., 2001) or by the α3/α6β2*-selective antagonist α-conotoxin MII, but it is decreased by treatment with the α3β4*-selective antagonist α-conotoxin Au1B (Luo et al., 1998; Grady et al., 2001). We extended this analysis to include the subunit null mutations for α2, α4, α5, α6, α7, β3, and β4 and repeated that of β2 for purpose of comparison. Figure 4A shows the results for the β2 and β4 subunits, as well as for the accessory subunits α5 and β3. As expected, the β2 null mutation did not affect the release of [³H]ACh from the IPn. Furthermore, the almost complete loss of release in the β4 null mutants supported previous indications that the relevant receptor was likely of the α3β4* subtype. Surprisingly, the α5 null mutation had no effect, whereas the β3 homozygous as well as heterozygous null mutants at all concentrations showed significantly decreased [³H]ACh release. The α2, α4, α6, and α7 subunit null mutations showed no differences among genotypes (Fig. 4B). These results indicate that ACh release in the IPn is mediated by the α3β3β4 and α3β4 subtypes.

Figure 4.

Effect of nAChR subunit null mutations on [³H]acetylcholine release from IPn synaptosomes. [³H]ACh release from IPn synaptosomes of β3^+/+ (white bars), β3^+/− (gray bars), and β3^−/− (black bars) mice was stimulated by ACh. Release stimulated by 50 mm potassium did not differ by strain or genotype (mean ± SEM for K⁺-stimulated release, 31.4 ± 0.8). In addition, no significant differences were seen among +/+ mice for the different concentrations of ACh by strain. Data shown in A from n = 6 β2^+/+, n = 6 β2^+/−, n = 6 β2^−/−, n = 5 β4^+/+, n = 7 β4^+/−, n = 6 β4^−/−, n = 11 β3^+/+, n = 11 β3^+/−, and n = 11 β3^−/− mice, and in B from n = 5 α2^+/+, n = 5 α2^+/−, n = 5 α2^−/−, n = 6 α4^+/+, n = 7 α4^+/−, n = 7 α4^−/−, n = 7 α6^+/+, n = 7 α6^+/−, n = 8 α6^−/−, n = 5 α7^+/+, n = 5 α7^+/−, and n = 5 α7^−/− mice. Significant differences by one-way ANOVA with Tukey's post hoc test are indicated by * for different from both +/+ and +/− of same strain or + for different from +/+. No significant differences were found for the β2, α5, α2, α4, α6, or α7 genotypes. For the β4 null mutation, the +/− mice did not differ from the +/+; however, the −/− mice differed significantly from both the +/+ and +/− mice (F_(2,15) = 13.30, p < 0.05; F_(2,15) = 12.15, p < 0.05; F_(2,15) = 16.00, p < 0.05 for 10, 30, and 100 μm ACh, respectively). Significant differences were also seen for the β3 genotypes. The −/− mice differed from both the +/+ and +/− mice at the two lower ACh concentrations and from the +/+ at the highest concentration, whereas in addition, the +/− mice differed from the +/+ at all concentrations (F_(2,30) = 18.10, p < 0.05; F_(2,30) = 18.10, p < 0.05; F_(2,30) = 6.98, p < 0.05 for 10, 30, and 100 μm ACh, respectively).

Effects of the genetic deletion of β2 or β4 subunits on ⁸⁶Rb⁺ efflux from Hb and IPn synaptosomes

To assess whether the two major subclasses β2* and β4* nAChRs are functional at presynaptic sites in Hb and IPn at terminals other than the cholinergic terminals described above, we made use of the ⁸⁶Rb⁺ efflux assay from synaptosomes prepared from these regions of mice with the β2 or β4 subunit null mutations. Although this assay does not determine which neurotransmitter is released, it does indicate which subtypes of nAChR are functional at terminals in the region assayed. Results (Fig. 5) indicate that, in the Hb, functional β2* nAChRs are found, whereas there are few or no functional β4* nAChRs. Evidence for function of both major subtypes is seen in the IPn synaptosomal preparation, in which significant decreases in ⁸⁶Rb⁺ efflux were found for both null mutations. A portion of the functional β4* nAChRs are likely those on cholinergic terminals measured with the [³H]ACh release assay, although receptors of this subclass may be present on other terminals as well.

Figure 5.

Effect of nAChR subunit null mutations on ⁸⁶Rb⁺ efflux from Hb and IPn synaptosomes. ⁸⁶Rb⁺ efflux from IPn synaptosomes of the β2 and β4 nAChR subunit +/+ (white bars), +/− (gray bars), and −/− (black bars) mice was stimulated by ACh. Data shown are from n = 5 β2^+/+, n = 5 β2^+/−, n = 4 β2^−/−, n = 5 β4^+/+, n = 5–6 β4^+/−, and n = 4–5 β4^−/− mice. Statistical analysis according to one-way ANOVA followed by Tukey's post hoc test; *p < 0.05 vs both +/+ and +/− mice of the same strain; ⁺p < 0.05 vs +/+ mice. No significant differences in ACh-stimulated ⁸⁶Rb⁺ efflux were seen between β2^+/+ and β4^+/+ mice. For the β2 null mutation with Hb synaptosomes, the β2^+/− mice did not differ from the β2^+/+ mice; however, the β2^−/− mice differed significantly from both the β2^+/+ and β2^+/− mice (F_(2,11) = 36.04, p < 0.001; F_(2,11) = 60.77, p < 0.001; F_(2,11) = 36.98, p < 0.001 for 3, 30, and 300 μm ACh, respectively). Significant differences were also seen for the β2 genotypes in IPn. The β2^−/− mice differed from the β2^+/+ mice at the two lower ACh concentrations (F_(2,11) = 8.19, p < 0.01; F_(2,11) = 5.81, p < 0.05; F_(2,11) = 1.28, NS, for 3, 30, and 300 μm ACh, respectively). No significant differences were found for the β4 genotypes in Hb; however, β4 gene deletion reduced ⁸⁶Rb⁺ efflux in IPn at the two higher ACh concentrations (F_(2,13) = 0.79, NS; F_(2,13) = 5.70, p < 0.05; F_(2,13) = 6.27, p < 0.05 for 3, 30, and 300 μm ACh, respectively).

Discussion

Native subtypes in the habenulo–interpeduncular system of rats and mice

In this study, Hb and IPn were found to express two major and distinct nAChR populations, β2* and β4*, with minor populations of α7. The diagrams in Figure 6 summarize the nAChR composition in Hb (A) and IPn (B) determined by quantitative immunoprecipitation.

The major β2* nAChR expressed in Hb of both rats and mice is α4β2* with minor amounts of α3β2* nAChR (Fig. 6A). A significant fraction of β2* receptors in rat contain accessory subunits, α5 or β3. Because only one accessory position exists per receptor (Kuryatov et al., 2008), these two accessory subunits are mutually exclusive, and their presence indicates even greater nAChR heterogeneity. In a distinction between species, the mouse β2* nAChRs in Hb include fewer accessory subunits.

In IPn of both species, β2* nAChRs exist as three approximately equal populations of α2β2*, α3β2*, and α4β2* (Fig. 6B), with a small fraction containing β4 subunits. Restriction of α2β2* to IPn agrees with previous in situ hybridization and reverse transcription-PCR studies showing localization of α2 mRNA in IPn and its absence in Hb (Wada et al., 1989; Marks et al., 1992; Sheffield et al., 2000). Detection of α3β2* receptors in both regions agrees with decreases in α-conotoxin MII binding seen there in α3^−/− mice (Whiteaker et al., 2002). Significant α5 was found in β2* nAChRs in rat IPn, whereas α5 is absent in this population in mouse. For both species, low levels of β3 were associated with β2* nAChR in IPn.

The β4* nAChR population, in both regions and species, is mainly associated with α3, in agreement with previously reported autoradiography studies in null mutant mice (Zoli et al., 1998; Marks et al., 2002; Whiteaker et al., 2002). Many of the β4* nAChRs contain accessory subunits; more have β3 than α5.

α6 subunit was detectable in mouse IPn but absent in rat. α6* nAChRs were not significantly decreased in β2^−/− mice but disappeared in IPn of β3^−/− mice, thus indicating a small population of α6β3β4* nAChRs exists in mouse IPn.

Although most β2 and β4 subunits form separate receptor populations, some β2β4* nAChRs exist in both regions (Fig. 2). Electrophysiological studies in rat Hb also presented evidence for β2β4* nAChRs (Quick et al., 1999). Functional measures at individual cells may detect smaller populations of a subtype than methods applied here.

Location and function of subtypes in IPn

The numerous cholinergic terminals in the IPn are mainly projections from Hb through the fr (Clarke et al., 1986). Our [³H]ACh release studies on these terminals clearly indicate that presynaptic β4* nAChRs, but not β2* nAChRs, modulate ACh release. Although deletion of α2, α4, α5, α6, or α7 subunits had no effect on nAChR-mediated ACh release, deletion of β3 subunits reduced this response ∼50%. Deletion of β3 also decreased expression of α3 and β4 subunits by a similar amount (Fig. 3), establishing the presence of presynaptic α3β3β4 and α3β4 controlling ACh release in the IPn.

Our immunochemical and functional experiments demonstrate that the IPn contains sizeable populations α2β2*, α3β2*, and α4β2* nAChRs. α2β2* nAChRs are expressed by intrinsic IPn neurons because, among the relevant nuclei, α2 mRNA is expressed only by the IPn (Wada et al., 1989). These are possibly the postsynaptic β2* nAChRs identified previously using electrophysiological methods (Brown et al., 1983; Clarke et al., 1986; Mulle et al., 1991; Dineley-Miller and Patrick, 1992).

α3β2* nAChRs must be located presynaptically in IPn because α3 mRNA is not expressed there. The afferents likely originate from MHb, highly enriched in α3 mRNA (Wada et al., 1989; Le Novère et al., 1996).

α4β2* nAChRs could be expressed on intrinsic IPn neurons and/or on afferents to the IPn, and, therefore, could have both presynaptic and postsynaptic locations. Besides the massive input from the MHb, afferents project to the IPn from several other nuclei (Herkenham and Nauta, 1979; Groenwhegen et al., 1986; Shibata et al., 1986; Araki et al., 1988). These nuclei all express α4 and β2 mRNAs (Wada et al., 1989) and may, in principle, be the origin of α4β2* nAChR. Nevertheless, MHb is likely the primary source of presynaptic α4β2* nAChRs in IPn. Indeed, we measured decreases in levels of α4 and β2 subunits in the IPn of β3^−/− mice. These α4β2β3 nAChRs are likely synthesized in the MHb and transported to the IPn, because β3 mRNA is not expressed in the IPn but is highly enriched in the MHb (Le Novère et al., 1996; Cui et al., 2003). This conclusion is consistent with the observation that bilateral habenular lesions decrease [³H]nicotine binding sites (now known to be α4β2*) in the rat IPn by 35% (Clarke et al., 1986)

Our Rb⁺ efflux results provide evidence that presynaptic β2* nAChRs in IPn are functional, but β2* nAChRs do not mediate ACh release. β2* nAChRs may modulate release of other neurotransmitters because the IPn receives important noncholinergic inputs (Klemm, 2004; Lecourtier and Kelly, 2007). Although we have not yet identified which neurotransmitter(s) are released by β2* nAChR, electrophysiological studies on IPn neurons suggest glutamate as a likely candidate (Girod et al., 2000).

In Hb synaptosomes, Rb⁺ efflux indicates that β2* nAChRs may be the only presynaptic functional subtype and may mediate release of several neurotransmitters here because dopaminergic (Phillipson and Pycock, 1982), GABA, glutamatergic (Meshul et al., 1998), and noradrenergic (Li et al., 1998) terminals have been identified in Hb.

Accessory subunits in α3β4* nAChR populations

A significant fraction of the α3β4* nAChR in the Hb and IPn of both rat and mouse contains the β3 subunit. This is the first biochemical and functional characterization of this unique subtype from brain. This result was unexpected because peripheral α3β4* nAChR, a well characterized subtype, contains α5 as accessory subunit (Conroy and Berg, 1995). β3 subunits have been shown previously to be associated with the α6β2nAChR by immunopurification of mesostriatal dopaminergic and visual pathways (Zoli et al., 2002; Gotti et al., 2005b). The presence of β3 not associated with α6 subunits in the Hb–IPn pathway agrees with in situ hybridization showing high levels of β3 but not α6 subunit mRNA in Hb (Le Novère et al., 1996; Champtiaux et al., 2003; Cui et al., 2003). β3 subunit mRNA is absent in IPn (Deneris et al., 1989; Cui et al., 2003); therefore, nAChRs containing β3 are not synthesized in the IPn. The β3 subunit may have a targeting role as suggested previously (Gotti et al., 2007; Kuryatov et al., 2008).

α3β3β4 nAChR has been expressed in oocytes (Groot-Kormelink et al., 1998; Boorman et al., 2000) in which functional studies indicate that incorporation of β3 decreases mean channel open time and burst length with only minor changes in calcium permeability and pharmacological properties (Boorman et al., 2003). Expression of human β3 with α3 and β4 subunits in oocytes resulted in only small changes in surface expression but suppressed function by ∼65% (Broadbent et al., 2006). In our studies on α3β4 and α3β3β4 nAChRs mediating [³H]ACh release in mouse IPn, deletion of β3 subunits decreased function by ∼50%, with no obvious change in EC₅₀. Absence of β3 subunits also decreased the amount of α3 and β4 subunits in IPn by ∼50%. It appears that absence of accessory subunit β3 has not changed function as expected from oocyte studies using human sequences (Broadbent et al., 2006). Perhaps distribution of receptors normally localized at cholinergic terminals differs. Alternatively, the unchanged calcium permeability and pharmacology of these subtypes may be more relevant for this assay, or mouse may differ significantly from human.

Pathophysiology of α3β4* nAChR subtypes

Nicotine can have both anxiolytic and anxiogenic effects. Behavioral studies indicate that subunit deletion of either β3 or β4 decreases anxiety-like behavior (Salas et al., 2003; Booker et al., 2007), suggesting that β3* and β4* nAChRs are involved in neuronal processes associated with anxiety. Our studies show that β4 and β3 subunits are not only colocalized but also coassembled in Hb and IPn and that this α3β3β4 nAChR modulates ACh release in the IPn. The IPn modulates impulses descending from the limbic forebrain to the source of serotoninergic mesencephalic projections, a pathway implicated in anxiety and depression. Our data may link the α3β3β4nAChR with the control of anxiety exerted by the IPn through the serotoninergic pathways (Klemm, 2004).

Components of the Hb–IPn system are thought to play a major role in the physiology and pathophysiology of reward phenomena (see Introduction). In particular, a strong case has recently been made that α3β4* nAChRs in the Hb–IPn system play an important role in mediating self-administration of morphine and other addictive drugs (Glick et al., 2006; Taraschenko et al., 2007).

α3β4* nAChRs have a restricted expression pattern in brain but are highly expressed in the Hb and IPn. Other uncommon nAChR subtypes (α2β2*, α4β3β2*, α3β3β4*, or α6β3β4*) appear unique to this system. Although the role of α3β3β4 receptors in the Hb–IPn is partially defined, the anatomy and functions of the wide variety of subtypes described in this study remain almost unknown. Studies with cellular and subcellular anatomical approaches, as well as additional functional studies, will be necessary to clarify this complex issue. The characterization of nAChR subtypes present in the Hb–IPn system may be an important step in the development of new therapies for neuropsychiatric diseases, such as drug dependence, depression, and anxiety disorders.