Introduction

Top Abstract Introduction Experimental Procedures Results and Discussion References

Nicotinic acetylcholine receptors (nAChRs) are excitatory ligand-gated cation channels that are widely distributed in mammalian organisms, appearing in the central and peripheral nervous systems, neuromuscular junctions, and adrenal glands. The nAChR channel complex is composed of five protein subunits, which form a pore that is permeable to Na⁺, K⁺, and Ca²⁺ (Lindstrom, 1995; Holladay et al., 1997). To date, , , , , and  subunits have been isolated and cloned from mammalian and avian tissues, with nine varieties of  and four varieties of  subunits identified. The 1, 1, , , and  subunits form the neuromuscular junction receptor, the very first nAChR to be characterized. The other subunits (2-9 and 2-4) are found predominantly throughout the nervous system (Lindstrom, 1995; Holladay et al., 1997). This subunit diversity affords a large potential for a variety of nAChR subtypes, exhibiting distinct cation-conducting properties and pharmacological heterogeneity.

Based on binding properties and pharmacological sensitivity, major nAChR subtypes in mammalian brain can be categorized as -bungarotoxin-sensitive (7) and -bungarotoxin-insensitive (e.g., 42) (Lindstrom, 1995; Holladay et al., 1997). Accordingly, ¹²⁵I--bungarotoxin has been the radioligand of choice for in vitro characterization of the 7 subtype of nAChR, whereas tritiated agonists, such as nicotine, acetylcholine, N-methylcarbamylcholine, cytisine, and epibatidine, have been used to study nAChRs of the latter group in vitro (Holladay et al., 1997). Of these ligands, epibatidine has the highest known affinity for 42 nAChRs, and outstanding in vitro binding characteristics (high specific-to-nonspecific binding ratio and slow kinetics of dissociation) (Dukat et al., 1993; Houghtling et al., 1995; Flores et al., 1996; Holladay et al., 1997; Stauderman et al., 1998; Xiao et al., 1998).

Radioligands developed for noninvasive in vivo imaging of -bungarotoxin-insensitive nAChRs have exhibited shortcomings, such as poor subtype selectivity and high levels of nonspecific binding (Nybäck et al., 1994). The radioligands [¹²³I]IPH and [¹²⁵I]IPH ((±)-exo-2-(2-[^123/125I]iodo-5-pyridyl)-7-azabicyclo[2.2.1]heptane), recently developed iodinated analogs of epibatidine, do not distinguish well between the 42 subtype and nAChRs containing 3 and 4 subunits (Dávila-García et al., 1997), much like epibatidine itself (Flores et al., 1996; Xiao et al., 1998). In contrast to 42 nAChRs, nAChRs containing 3 and 4 subunits, possibly in combination with 5 subunits, are distributed mostly in the peripheral nervous system and adrenal glands (Holladay et al., 1997). Therefore, high affinity for the latter receptors could contribute to the untoward cardiovascular effects of epibatidine and its analogs (Molina et al., 1997; Horti et al., 1998) and might limit the use of epibatidine-based compounds for imaging nAChRs in human subjects.

Recently, 3-(2(S)-azetidinylmethoxy)pyridine (A-85380, Fig. 1) has been identified as a high-affinity nAChR ligand (Abreo et al., 1996). Subsequently, a chloro analog of A-85380, ABT-594 (Fig. 1), has been developed as a promising nonopioid analgesic having affinity for 42 nAChRs comparable to that of epibatidine, but lacking its toxicity (Bannon et al., 1998). In a search for improved radioligands suitable for noninvasive in vivo imaging of nAChRs, chemists in our group synthesized several halogenated analogs of A-85380 (Koren et al., 1998). Some of these compounds, particularly 5-iodo-A-85380 (Fig. 1), exhibited extremely high affinity for nAChRs in rat brain (Koren et al., 1998).

View larger version (11K): [in this window] [in a new window]   Fig. 1.   Chemical structures of A-85380 and two of its analogs.

Initial evaluation of 5-[¹²⁵I]iodo- and 5-[¹²³I]iodo-A-85380 in vivo in mice (Musachio et al., 1998; Vaupel et al., 1998) and 5-[¹²³I]iodo-A-85380 in rhesus monkey (Chefer et al., 1998) and baboon (Musachio et al., 1999) demonstrated that these radioligands readily crossed the blood-brain barrier, bound to cerebral nAChRs with high specificity, and had low toxicity. Here, we present an in vitro characterization of 5-[¹²⁵I]iodo-A-85380, indicating that this ligand possesses excellent properties as a probe for studying the 42 nAChR subtype. 5-Iodo-A-85380 features high affinity for nAChRs, low nonspecific binding, slow dissociation from the receptor, and exceptionally high selectivity for the 42 subtype among the major mammalian nAChR subtypes. These properties, together with the results of in vivo studies with this ligand (Chefer et al., 1998; Musachio et al., 1998, 1999; Vaupel et al., 1998), suggest that 5-[¹²³I]iodo-A-85380 may have exceptional potential as a radioligand for in vivo imaging of 42 nAChRs with single photon emission computed tomography.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results and Discussion References

Materials. 5-[¹²³I]iodo-A-85380 and 5-[¹²⁵I]iodo-A-85380 were prepared according to the literature procedures (Musachio et al., 1998; Horti et al., 1999). Specific activity of 5-[¹²⁵I]iodo-A-85380 was determined as described previously (Horti et al., 1999). On the day of each synthesis, the specific activities of the three batches of 5-[¹²⁵I]iodo-A-85380 used in these studies were 1550, 1980, and 2200 Ci/mmol, respectively. ¹²⁵I--Bungarotoxin (¹²⁵I--BTX, 100 Ci/mmol), [³H]cytisine (32 Ci/mmol), and (±)-[³H]epibatidine (48 Ci/mmol) were obtained from New England Nuclear Corp. (Boston, MA). 5-Iodo-A-85380 was prepared by a published method (Koren et al., 1998). (±)-exo-2-(2-Iodo-5-pyridyl)-7-azabicyclo[2.2.1]heptane (IPH), a gift from Dr. Kellar, was synthesized by Dr. J. L. Musachio at the Johns Hopkins University as described previously (Musachio et al., 1997). 3-(2(S)-Azetidinylmethoxy)pyridine dihydrochloride (A-85380), -bungarotoxin (-BTX), (-)- and (±)-epibatidine, (-)-cytisine, (+)- and ()-stereoisomers of nicotine, acetylcholine, carbachol, dihydro--erythroidine, curare, mecamylamine, atropine, naloxone, (R)-(-)-apomorphine, and haloperidol were purchased from Research Biochemicals International (Natick, MA). Physostigmine, diisopropyl fluorophosphate (DFP), and all other chemicals used were purchased from Sigma Chemical Co. (St. Louis, MO). Male Fischer-344 and Sprague-Dawley rats were obtained from Charles River Breeding Laboratories (Wilmington, MA). Rats, shipped at the age of 12 weeks, were housed in a temperature- and light-controlled vivarium for at least 2 weeks before being used for this study. Mice were generated by mating parents heterozygous for a mutation in the 2 nAChR subunit (Picciotto et al., 1995). Frozen Torpedo californica electric organ tissue was purchased from Marinus Inc. (Long Beach, CA). Frozen samples of postmortem tissue of human cerebral cortex (four subjects, 38 to 49 years of age, death from arteriosclerotic cardiovascular disease) were obtained from the Brain and Tissue Bank for Developmental Disorders (Baltimore, MD).

Membrane Preparation. After CO₂ euthanasia and decapitation of the rats, brains were removed and prepared as follows. Brain tissue used for binding studies was obtained by a single cut just behind the inferior colliculi to exclude the cerebellum and medulla. In some experiments, specific brain regions and the adrenal glands were isolated. Frozen samples of Torpedo californica electric organ and postmortem human cerebral cortical tissue were thawed at room temperature for 30 to 60 min before membrane preparation. Total membrane fractions from all tissues were isolated by homogenization of the respective tissue with a Brinkmann Polytron homogenizer in 10 to 20 volumes of a HEPES-salt solution (HSS), containing HEPES (pH 7.4, 15 mM), 120 mM NaCl, 5.4 mM KCl, 0.8 mM MgCl₂, and 1.8 mM CaCl₂, followed by centrifugation at 40,000g for 10 min. The pellets were washed twice with HSS through rehomogenization and centrifugation at the same settings. Three additional washings were performed in the case of the total membrane fraction from rat adrenal glands and Torpedo californica electric organ. Crude membrane fractions (P2) were isolated as described previously (Koren et al., 1998), and were stored in aliquots at -70°C for at least 16 h but not more than 4 weeks before use. On the day of assay, pellets were thawed, homogenized in 30 volumes of HSS, and centrifuged at 40,000g for 10 min. The resultant pellets were resuspended in a freshly prepared HSS and used for binding assays.

Binding Assays. Assays were carried out in HSS at 22°C unless otherwise specified. Incubations were performed in polystyrene tubes except for the assays with ¹²⁵I--BTX, for which borosilicate glass tubes were used. The HSS for the studies of membranes from T. californica electric organ contained 0.1% of BSA. Nonspecific binding was determined in the presence of 300 µM (-)-nicotine except for the assays with ¹²⁵I--BTX, for which 1 µM -BTX was used instead (see the figure legends for other specific conditions for particular binding assays). Incubation was terminated by filtration through Whatman GF/B glass fiber filters, presoaked in 1% polyethyleneimine, using a Brandel 48-channel cell harvester. Filters were washed three times with 3-ml aliquots of a rinse buffer (50 mM Tris · HCl, pH 7.4). In ¹²⁵I--BTX assays, the rinse buffer also contained 1% of nonfat dry milk to reduce nonspecific binding to filter material. Radioactivity was measured using a Beckman LS 3801 liquid scintillation counting system (efficiency 43%) or a LKB Wallac 1277 gamma counter (efficiency 68%).

In Vitro Autoradiography. Sagittal slices (20 µm thick) at 0.4, 0.9, 1.4, 1.9, 3.9, and 4.2 mm lateral to midline from eight frozen Fischer-344 rat brains were obtained by sectioning in a cryostat at -20°C, and were thaw-mounted onto gelatin-coated slides. Sections were refrozen and kept frozen at -80°C until the day of the assay. Slices were preincubated for 20 min in 50 mM Tris · HCl buffer (pH 7.0) containing 120 mM NaCl, 5 mM KCl, 2.5 mM CaCl₂, and 1 mM MgCl₂. Thereafter, they were incubated with 210 pM 5-[¹²⁵I]iodo-A-85380 (specific activity 435 Ci/mmol) in the same buffer for 2 h at 25°C, then rinsed twice in ice-cold buffer for 5 min each and once in distilled water for 1 min. The slides were dried overnight in a vacuum desiccator. Nonspecific binding was assessed in adjacent slices incubated in 210 pM 5-[¹²⁵I]iodo-A-85380 containing 10 µM nicotine bitartrate. Slices and appropriate ¹²⁵I-standards were apposed to ³H-Ultrafilm for 2 days at room temperature. The resulting autoradiograms were digitized using a video camera-based system. The digitized images were analyzed using a computer program (INQUIRY, Loats Associates, Inc., Westminster, MD).

Protein Assay. Protein measurements were performed using a dye reagent kit (Bio-Rad, Richmond, CA) and BSA as a standard.

Data Analysis. Saturation binding data were subjected to Scatchard and linear regression analyses. Competition binding data were analyzed using nonlinear regression methods. Values of K_i were derived from the measured IC₅₀ and K_d values for radioactive ligands using the Cheng-Prusoff equation K_i = IC₅₀/(1+F/K_d), where F is the concentration of unbound radioligand. The K_d values were obtained from three to six independent experiments performed on the same membrane preparations that were used for the competition assays. Results of the kinetic experiments were analyzed using semilogarithmic plots and linear regression analysis. The values of equilibrium constant of dissociation, K_d, obtained from the kinetic studies, were calculated by the equation K_d = k_diss/k_ass, where k_diss and k_ass are the dissociation and association rate constants, respectively.

	Results and Discussion

Top Abstract Introduction Experimental Procedures Results and Discussion References

Kinetic and Equilibrium Binding Characteristics. The specific binding of 5-[¹²⁵I]iodo-A-85380, determined in rat brain membranes at 22°C and at a ligand concentration of 10 pM, reached one half the maximal (equilibrium) binding level in 67 ± 9 min (Fig. 2a). The binding was completely reversible and was characterized by a very slow dissociation (t_1/2 = 132 ± 9 min) (Fig. 2b). The rate constants of association (k_ass) and dissociation (k_diss) were (5.6 ± 1.4) · 10⁴/pM/min and (54 ± 4) · 10⁴/min, respectively. The K_d value, calculated as the ratio of k_diss to k_ass, was 9.7 ± 1.8 pM.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Kinetics of ligand-receptor association (A) and dissociation (B) for 5-[125I]iodo-A-85380 in rat brain. In association experiments, rat brain P2 membrane fractions (4-5 µg of protein) were incubated in a total volume of 0.4 ml with 10 pM 5-[125I]iodo-A-85380 at 22°C. In dissociation experiments, ()-nicotine at final concentration of 1 mM was added to rat brain P2 membrane fractions preincubated with 10 pM 5-[125I]iodo-A-85380 for 4 h at 22°C. Each graph represents results of a single experiment performed in triplicate (S.E.M. < 10%). Similar results were obtained in three additional experiments for both association and dissociation studies. The mean ± S.E. values from the four corresponding experiments were: t1/2ass, 67 ± 9 min [kass, (5.6 ± 1.4) · 104 /min/pM], and t1/2diss, 132 ± 9 min [kdiss, (5.4 ± 0.4) · 103 /min].

View larger version (13K): [in this window] [in a new window]   Fig. 3.   Scatchard plots of 5-[125I]iodo-A-85380 binding data obtained from saturation studies in P2 membrane fraction of Fischer-344 rat brain (A), in total membrane fraction of Sprague-Dawley rat brain (B), and in total membrane fraction of postmortem tissue of human cerebral cortex (C). Rat brain membranes (18-25 µg of protein) or human cortical membranes (26-30 µg of protein) were incubated in a total volume of, respectively, 1 or 0.5 ml with 1 to 500 pM 5-[125I]iodo-A-85380 for 4 h at 22°C. Data were analyzed as described in Experimental Procedures. Each point represents the mean from three or four replicates (S.E.M. < 7%). A, pooled data from two independent experiments performed on the same membrane preparation. Four additional experiments were performed on membranes from two separate membrane preparations. The Kd and Bmax values (mean ± S.E.) obtained from these six experiments were 10.6 ± 0.3 pM and 160 ± 25 fmol/mg protein (3.8 ± 0.6 pmol/g tissue), respectively. B, results of a single experiment. Similar binding characteristics were observed in four additional experiments performed on membranes from four independent preparations. The Kd and Bmax values (mean ± S.E.) obtained from these five experiments were 10.0 ± 0.2 pM and 178 ± 6 fmol/mg protein (3.9 ± 0.2 pmol/g tissue), respectively. C, results of a single experiment. Similar results were obtained in three additional experiments performed on membranes of human postmortem cortical tissue obtained from different subjects. The Kd and Bmax values (mean ± S.E.) obtained from all four experiments were 11.6 ± 0.7 pM and 53 ± 6 fmol/mg protein (0.98 ± 0.04 pmol/g tissue).

Competition Studies. In competition assays with 5-[¹²⁵I]iodo-A-85380, affinities for nAChRs in rat brain for six well-characterized nicotinic agonists and four nicotinic antagonists (Table 2) fell into an order that was consistent with that previously observed in assays using other radioligands for -bungarotoxin-insensitive nAChRs (Pabreza et al., 1991; Decker et al., 1995; Houghtling et al., 1995). Compounds that did not effectively inhibit binding of 5-[¹²⁵I]iodo-A-85380 (K_i > 25 µM) included a noncompetitive nicotinic antagonist (mecamylamine), an antagonist at muscarinic acetylcholine receptors (atropine), cholinesterase inhibitors (physostigmine and DFP), an antagonist at opiate receptors (naloxone), an agonist at dopamine receptors (apomorphine), and an antagonist at D₂-like dopamine receptors (haloperidol).

Regional Distribution in Brain. To test our hypothesis that 5-[¹²⁵I]iodo-A-85380 labels 42 nAChRs, we investigated distribution of the radioligand binding in the rat brain using both in vitro binding assays and autoradiography. In all brain regions studied, 5-[¹²⁵I]iodo-A-85380 binding was characterized by interaction with a single population of homogenous binding sites (Scatchard plots not shown) with K_d values close to 11 pM (Fig. 4). The average of K_d values from all regions studied was 11.0 ± 0.2 pM. These constants closely agreed with the K_d and K_i values observed for the whole rat forebrain (Fig. 3, Table 2). The regional distribution of binding sites (Fig. 4) closely matched that of the 42 nAChR subtype measured previously in rat brain using (-)-[³H]cytisine and (±)-[³H]epibatidine (Pabreza et al., 1991; Houghtling et al., 1995); with the highest densities being observed in thalamic nuclei and superior colliculi; intermediate densities in the striatum, cortex, and hippocampus; and the lowest concentrations in the hypothalamus and cerebellum.

View larger version (58K): [in this window] [in a new window]   Fig. 4.   Distribution of 5-[125I]iodo-A-85380 in Fischer-344 rat brain. The individual brain structures obtained from ten animals were pooled and total membrane fractions from each region were isolated as described in Experimental Procedures. Membrane samples from each brain region (correspond to 0.1-0.5 mg of wet tissue) were incubated in a total volume of 0.25 ml with 1 to 500 pM 5-[125I]iodo-A-85380 for 4 h at 22°C. For all brain regions, Scatchard analyses produced data consistent with homogeneous populations of binding sites with similar Kd values. Gray (Bmax values) and open (Kd values) columns represent means obtained from three to four saturation assays performed on membranes from two separate preparations. For all regions studied, S.E.M. values were less than 10% except for the hypothalamus, where S.E.M. = 25%. Cb, cerebellum; Cx, frontal cortex; F-Br, forebrain; Hipp, hippocampus; Hyp, hypothalamus; SC, superior colliculus; Str, striatum; Th, thalamus.

View larger version (87K): [in this window] [in a new window]   Fig. 5.   In vitro autoradiography of rat brain with 5-[125I]iodo-A-85380. Slices (20-µm thick; taken approximately 0.4, 1.4, and 1.9 mm lateral to the midline) were incubated with 210 pM 5-[125I]iodo-A-85380 for 2 h. Nonspecific binding was assessed in the presence of 10 µM (-)-nicotine. Pseudocolor-transformed autoradiograms were expressed in femtomoles per milligram of tissue. Acb, nucleus accumbens; Cb, cerebellum; CPu, caudate putamen; DTg, dorsal tegmental area; fr, fasciculus retroflexus; Hipp, hippocampus; IP, interpeduncular nucleus; MHb, medial habenula; Pn, pontine nucleus; RS, retrosplenial cortex; S, subiculum; SuG, superior colliculus, superficial gray; Th, thalamus; VTA, ventral tegmental area.

Subtype Selectivity. To address the issue of binding selectivity among nAChR subtypes, we measured affinities of unlabeled 5-iodo-A-85380 in four different competition assays. To determine the affinity of 5-iodo-A-85380 for the 42 subtype, we used previously described competition assays with (±)-[³H]epibatidine or (-)-[³H]cytisine and P2 membrane fractions of Fischer-344 rat forebrain. Consistent with the previous observations (Koren et al., 1998), Scatchard plots of (±)-[³H]epibatidine binding to these rat brain membranes, at ligand concentrations of 1 to 800 pM, revealed a single population of binding sites (data not shown) with a K_d of 9.5 ± 0.5 pM (n = 4). Similar high-affinity binding of (±)-[³H]epibatidine was observed in the rat (Houghtling et al., 1995), mouse (Marks et al., 1998), and human brain (Houghtling et al., 1995; Sihver et al., 1998), as well as in transfected cells stably expressing the 42 nAChR subtype (Houghtling et al., 1995; Whiteaker et al., 1998). In addition, the affinity of 5-iodo-A-85380 for the 42 nAChR subtype was measured in competition assays with ()-[³H]cytisine, which is the most widely used ligand for characterization of this subtype. In our saturation experiments, (-)-[³H]cytisine revealed binding sites with a K_d value of 300 ± 50 pM (n = 3) in the rat brain membranes, consistent with published data (Pabreza et al., 1991).

View this table: [in this window] [in a new window]   TABLE 3 Binding selectivity of 5-iodo-A-85380 and other nAChR ligands at four major mammalian nAChR subtypes In assays of the 42 subtype, rat brain P2 membrane fractions (200 µg of protein) were incubated in a total volume of 0.5 ml with 0.5 nM (±)-[3H]epibatidine and 9 to 11 concentrations of competitors for 1.5 h at 22°C. In assays of the nAChRs containing 3 and 4 subunits (34x), total membrane fractions of rat adrenal glands (250 µg of protein) were incubated in a total volume of 1 ml with 0.4 nM (±)-[3H]epibatidine and five to nine concentrations of competitors for 1.5 h at 22°C. In assays of the 7 subtype, rat brain P2 membrane fractions (100 µg of protein) were incubated in a total volume of 0.1 ml with 2 nM 125I--bungarotoxin and 9 to 11 concentrations of competitors for 3 h at 22°C. In assays of the muscle subtype, total membrane fractions from T. californica electric organ (0.1 µg of protein) were incubated in a total volume of 0.1 ml with 2 nM 125I--bungarotoxin and 9 to 11 concentrations of competitors for 1.5 h at 22°C. The inhibition constants (Ki values) were calculated by the Cheng-Prusoff equation from measured IC50 values using the following Kd values: 10 and 55 pM for (±)-[3H]epibatidine binding in rat forebrain (42) and in rat adrenal glands (34x), respectively; and 1.5 and 2.3 nM for 125I--bungarotoxin binding in rat forebrain (7) and in T. californica electric organ (11), respectively. These Kd values were obtained from three to six separate saturation assays per constant. Data represent means ± S.E. obtained from three to seven competition assays per constant.

View larger version (57K): [in this window] [in a new window]   Fig. 6.   In vitro autoradiography of wild-type mouse brain and of brains from mice homozygous for a mutation in the 2 subunit of nAChRs performed with [125I]IPH and with 5-[125I]iodo-A-85380. Slices (12-µm thick) at bregma -1.5 mm (four top panels) and at bregma -3.4 mm (four bottom panels) were incubated with 200 pM [125I]IPH (2200 Ci/mmol) and with 200 pM 5-[125I]iodo-A-85380 (420 Ci/mmol) for 0.5 h. Processed slides were exposed to Hyperfilm for 2 days ([125I]IPH) or 5 days (5-[125I]iodo-A-85380). Autoradiograms from mice homozygous for a mutation in the 2 subunit showed an absence of 5-[125I]iodo-A-85380 binding. Similar results were observed in all slices analyzed. Arrow and double arrow indicate, respectively, the medial habenula and interpeduncular nucleus, where binding of [125I]IPH, but not of 5-[125I]iodo-A-85380, was seen in brains of 2-subunit knockout mice. Exposure for 5 weeks of 5-[125I]iodo-A-85380-labeled slices from mice homozygous for a mutation in the 2 subunit of nAChRs to Hyperfilm did not result in the detection of any additional specific signal. Results shown were reproduced in additional experiments with two control and two mutant animals as well as in experiments with 5-[123I]iodo-A-85380 in three control and three mutant animals.