Pharmacological properties of AC-3933, a novel benzodiazepine receptor partial inverse agonist

Neuroscience 256 (2014) 352–359

PHARMACOLOGICAL PROPERTIES OF AC-3933, A NOVEL BENZODIAZEPINE RECEPTOR PARTIAL INVERSE AGONIST T. HASHIMOTO, a* T. KIYOSHI, a H. KOHAYAKAWA, b Y. IWAMURA b AND N. YOSHIDA b

INTRODUCTION

a

Drug Development Research Laboratories, Dainippon Sumitomo Pharma Co., Ltd., 33-94 Enoki-cho, Suita, Osaka 564-0053, Japan

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder of the central nervous system, and is clinically characterized by progressive loss of cognitive function represented by memory impairment (Katzman, 1986; Perry and Hodges, 1999; Selkoe, 2001; Marin et al., 2002). Neuropathologically, AD is characterized by deposit of b-amyloid protein in senile plaques and aberrant aggregation of highly phosphorylated tau proteins in the form of neurofibrillary tangles (Kosik, 1991; Carter and Lippa, 2001). Additional pathological hallmarks of AD include loss of synapses and decrease in neuron density in specific regions of the brain. Although there is currently no cure for AD, various genetic, biochemical, and molecular approaches have been taken to help with the symptoms of this disease (Farlow et al., 2001; Tariot, 2001; Bullock and Dengiz, 2005; O’Brien et al., 2011; Piau et al., 2011; Singh et al., 2012; Wiwanitkit, 2013). Since acetylcholine (ACh) is known to play an important role in learning and memory, and remarkable dysfunction of cholinergic neurons is found in the brain of AD patients, a number of agents that augment cholinergic function by inhibiting acetylcholinesterase, a serine protease that hydrolyzes ACh, including, tacrine, donepezil, rivastigmine, and galantamine have been developed (Farlow et al., 1992, 2001; Knapp et al., 1994; Tariot, 2001; Bullock and Dengiz, 2005). While acetylcholinesterase inhibitors (AChEIs) improve cognitive impairment in mild to moderate AD patients, they have limited efficacy with undesirable side effects, such as nausea, vomiting, diarrhea, and anorexia. Therefore, new therapeutic agents with a better safety profile have long been needed. GABA is a well-known inhibitory neurotransmitter that binds to the GABAA receptor/channel complex and regulates neural excitability throughout the nervous system (Schwartz, 1988; Rabow et al., 1995; Michels and Moss, 2007). The central benzodiazepine receptor (BzR) has been reported to modulate the GABAA receptor/channel complex by allosteric mechanism (Haefely et al., 1985; Polc, 1988; Christian et al., 2013). As BzR ligands are known to positively or negatively modulate GABAA receptor function, they are classified into agonists, inverse agonists and antagonists (Haefely et al., 1993; Teuber et al., 1999). BzR agonists, such as diazepam, enhance GABA effects and can therefore induce anxiolytic and antiepileptic effects as well as

b Innovative Drug Discovery Laboratories, Dainippon Sumitomo Pharma Co., Ltd., 33-94 Enoki-cho, Suita, Osaka 564-0053, Japan

Abstract—We investigated in this study the pharmacological properties of AC-3933 (5-(3-methoxyphenyl)-3-(5-methyl1,2,4-oxadiazol-3-yl)-1,6-naphthyridin-2(1H)-one), a novel benzodiazepine receptor (BzR) partial inverse agonist. AC3933 potently inhibited [3H]-flumazenil binding to rat whole brain membrane with a Ki value of 5.15 ± 0.39 nM and a GABA ratio of 0.84 ± 0.03. AC-3933 exhibited almost no affinity for the other receptors, transporters and ion channels used in this study. In addition, AC-3933, in the presence of GABA (1 lM), gradually but significantly increased [35S] tert-butylbicyclophosphorothionate binding to rat cortical membrane to 117.1% of the control (maximum increase ratio) at 3000 nM. However, this increase reached a plateau at 30 nM with hardly any change at a concentration range of 100–3000 nM (from 115.2% to 117.1%). AC-3933 (0.1–10 lM) significantly enhanced KCl-evoked acetylcholine (ACh) release from rat hippocampal slices in a concentrationdependent manner. Moreover, in vivo brain microdialysis showed that intragastric administration of AC-3933 at the dose of 10 mg/kg significantly increased extracellular ACh levels in the hippocampus of freely moving rats (area under the curve (AUC0–2 h) of ACh level; 288.3% of baseline). These results indicate that AC-3933, a potent and selective BzR inverse agonist with low intrinsic activity, might be useful in the treatment of cognitive disorders associated with degeneration of the cholinergic system. Ó 2013 IBRO. Published by Elsevier Ltd. All rights reserved.

Key words: benzodiazepine receptor, inverse agonist, GABA ratio, TBPS binding, acetylcholine, hippocampus.

*Corresponding author. Address: Ikeda Laboratory, Drug Development Research Laboratories, Dainippon Sumitomo Pharma Co., Ltd., 33-94 Enoki-cho, Suita-city, Osaka 564-0053, Japan. Tel: +81-66337-5918; fax: +81-6-6337-5109. E-mail address: [email protected] (T. Hashimoto). Abbreviations: AC-3933, 5-(3-methoxyphenyl)-3-(5-methyl-1,2,4oxadiazol-3-yl)-1,6-naphthyridin-2(1H)-one; ACh, acetylcholine; AChEI, acetylcholinesterase inhibitor; AD, Alzheimer’s disease; AUC, area under the curve; b-CCM, b-carboline-3-carboxylate; BzR, benzodiazepine receptor; DMCM, methyl-6,7-dimethoxy-4-ethylbeta-carboline-3-carboxylate; DMSO, dimethylsulfoxide; GABA, caminobutyric acid; HPLC, high-performance liquid chromatography; SEM, standard error of the mean; TBPS, tert-butylbicyclophosphorothionate; Tris, tris(hydroxymethyl)aminomethane.

0306-4522/13 $36.00 Ó 2013 IBRO. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.neuroscience.2013.10.031 352

T. Hashimoto et al. / Neuroscience 256 (2014) 352–359

sedation and amnesia as side effects (Podhorna et al., 2002; Whiting, 2006; McMullan et al., 2010). Conversely, BzR inverse agonists, such as b-carboline, reduce GABA effects and hence increase the overall excitability of neurons. Since BzR inverse agonists with high intrinsic activity, (i.e. full inverse agonists) strongly attenuate GABAergic inhibition, they induce undesirable effects, including anxiety and convulsion (Gardner, 1992; Conto´ et al., 2005). On the other hand, BzR inverse agonists with low intrinsic activity (i.e. partial inverse agonist) are expected to exert modest excitability of neurons and hence produce minimal side effects. As cholinergic neurotransmission in the hippocampus and cortex is known to be regulated by GABAergic inhibition, BzR partial inverse agonists can enhance cholinergic transmission and therefore provide a new therapeutic strategy for the treatment of AD (Appleyard et al., 1990; Imperato et al., 1993; Moore et al., 1993; Abe et al., 1998; Chorvat et al., 1998; Maubach, 2003). In our search for a new class of non-benzodiazepines that bind to BzR, we found AC-3933 (5-(3methoxyphenyl)-3-(5-methyl-1,2,4-oxadiazol-3-yl)-1,6naphthyridin-2(1H)-one) as a novel BzR partial inverse agonist (Fig. 1). In the present study, we assessed some of the in vitro and in vivo pharmacological properties of AC-3933.

EXPERIMENTAL PROCEDURES Animals Male Wistar rats were obtained from Japan SLC (Shizuoka, Japan) and housed in groups of two to three in plastic cages kept in a temperature (23 ± 3 °C)- and relative humidity (55 ± 15%)-controlled animal room under a 12/12-h light/dark cycle (light on at 07:00). The animals had free access to food and water, and were acclimatized for at least 7 days before all experiments. All experimental procedures used were approved by the Institutional Animal Care and Use Committee of Dainippon Sumitomo Pharma Co., Ltd. BzR binding assay Receptor preparation. Tissue preparation and binding to BzR in rat brain membrane were performed as reported in the literature (Wood et al., 1984; Bertz et al., 1995).

353

Briefly, male Wistar rats weighting 210–230 g were decapitated, and the whole brain was dissected and homogenized in 10 volumes of ice-cold 50 mM tris(hydroxymethyl)aminomethane (Tris)–HCl buffer (pH 7.7) with a Potter–Elvehjem tissue grinder. The homogenate was centrifuged three times at 50,000g at 4 °C for 15 min, and the pellet was suspended in 10 volumes of the original wet tissue weight of ice-cold 50 mM Tris–HCl buffer (pH 7.7). The resultant pellet was stored overnight at –80 °C. The frozen pellet was next defrosted and resuspended in 10 volumes of the original wet tissue weight of ice-cold 50 mM Tris–HCl buffer (pH 7.7), and the suspension was centrifuged twice at 50,000g at 4 °C for 15 min. The final pellet was resuspended in ice-cold 15 mM Krebs-Tris buffer [(in mM) NaCl 118, KCl 4.8, CaCl2 1.28, MgSO4 1.2, Tris– HCl 15, pH 7.4] at a concentration of 20 mL/g of original wet tissue weight, and the suspension was frozen at 80 °C until use for the binding assay. Protein concentration was determined using the BCA protein assay kit (PIERCE, Rockford, IL, USA). Binding assay. Competitive binding was performed by incubating the crude membrane preparation (200 lL) with both [3H]-flumazenil (1 nM) and various concentrations of test-compounds. The binding assay was performed in the presence or absence (i.e. use of 10 lM bicuculline) of GABA (100 lM) in parallel experiments. Assay mixtures (total volume of 1 mL) were incubated at 25 °C for 15 min, and the reaction was terminated by rapid filtration through a Whatman GF/B glass fiber filter, which was immediately rinsed three times with 5 mL of ice-cold 50 mM Tris–HCl buffer (pH 7.7) using a cell harvester (MB-48, Brandel, MD, USA). The filter-bound radioactivity was quantified after addition of 10 mL scintillator (ACS II, Amersham, Bucks, UK) using a liquid scintillation analyzer (Tri-Carb 3100TR, Packard, CA, USA). Nonspecific binding was determined in the presence of 1 lM clonazepam. All assays were carried out in duplicate, except for the total binding and nonspecific binding, which were in quadruplicate. The IC50 value of each test compound was determined by non-linear least squares curve-fitting analysis using the SASÒ system (SAS Institute Inc., Cary, NC, USA). Ki values with and without GABA were calculated using KD and IC50 values with and without GABA, respectively, according to the following equation: Ki = IC50/(1 + [L]/KD, where [L] is the concentration of [3H]-flumazenil and KD is the dissociation constant of flumazenil. Flumazenil KD value was determined from Scatchard equation using linear regression analysis in SASÒ system. The KD values of [3H]-flumazenil binding with and without GABA were 3.40 and 4.05 nM, respectively. GABA ratios were calculated from the ratio of Ki values with and without GABA using the following equation: GABA ratio = Ki without GABA/Ki with GABA. Non-specific binding

Fig. 1. Chemical structure of AC-3933.

To determine AC-3933 non-specific binding to off-target proteins, binding assays for various receptors, transporters and ion channels were carried out on our

354

T. Hashimoto et al. / Neuroscience 256 (2014) 352–359

behalf by Sekisui Medical Co., Ltd. (Tokyo, Japan), a contract research organization. The receptors, transporters and ion channels tested were as follows: receptors of adenosine (A1, A2 and A3), adrenergic (a1, a2, b1, b2 and b3), bradykinin BK2, dopamine (D1, D2, D3, D4 and D5), endothelin (ETA and ETB), GABA (GABAA and GABAB), galanin (GAL1 and GAL2), glutamate (AMPA, kainate, NMDA and glycine), histamine (H1, H2 and H3), muscarinic (M1, M2, M3, M4 and M5), neurokinin (NK1 and NK2), neuropeptide (Y1 and Y2), nicotine, opioid (l, d and j), serotonin (5-HT1, 5-HT2, 5-HT3, 5-HT4, 5-HT5, 5-HT6 and 5-HT7), sigma (r1 and r2), vasopressin V1b, and hormones (estrogen, glucocorticoid, progesterone and testosterone); transporters of adenosine, dopamine, GABA and monoamine (VMAT2); ion channels of calcium (L-type, N-type), potassium (voltage-dependent) and sodium (voltage-dependent). [35S]-tert-butylbicyclophosphorothionate (TBPS) binding assay Receptor preparation. Tissue preparation and [35S]TBPS binding assay were performed essentially as described by Serra et al. (1994). Briefly, male Wistar rats weighting 230–250 g were decapitated, and the cerebral cortex was dissected and homogenized in 50 volumes of ice-cold 50 mM Tris–citrate buffer (pH 7.4) containing 100 mM NaCl with a Potter–Elvehjem tissue grinder. The homogenate was centrifuged twice at 20,000g at 4 °C for 20 min, and the pellet was stored overnight at 80 °C. The frozen pellet was next defrosted, suspended in ice-cold 50 mM Tris–citrate buffer containing 100 mM NaCl, and then centrifuged twice at 20,000g at 4 °C for 20 min. The final pellet was resuspended in ice-cold 50 mM Tris–citrate buffer (pH 7.4) at a concentration of 20 mg wet tissue/mL and used for the binding assay. Binding assay. The resuspended crude membrane preparation (200 lL) was mixed with [35S]-TBPS (2 nM), NaCl (200 mM), GABA (1 lM) and various concentrations of test-compounds. Assay mixtures (total volume of 1 mL) were incubated at 25 °C for 90 min, and the reaction was terminated by rapid filtration through a 0.1% polyethylenimine-pretreated Whatman GF/B glass fiber filter, which was immediately rinsed three times with 5 mL of ice-cold 50 mM Tris–HCl buffer (pH 7.4), using a cell harvester (MB-48, Brandel). The filter-bound radioactivity was quantified after the addition of 10 mL scintillator (ACS II, Amersham) using a liquid scintillation analyzer (Tri-Carb 3100TR, Packard). Nonspecific binding was determined in the presence of 100 lM picrotoxin. [35S]-TBPS specific binding was calculated by subtracting nonspecific binding from total binding, and the count of specific binding was accepted as 100% control. [35S]-TBPS binding in the presence of testcompounds was expressed as percentage of specific binding; i.e., % of control.

In vitro ACh release experiment ACh release. In vitro ACh release experiment was performed as previously described (Kar et al., 1998; Yamamoto et al., 2000) with some modifications. Briefly, male Wistar rats (300–350 g) were decapitated, and the brain was removed and placed in oxygenated ice-cold superfusion buffer [Krebs buffer: (in mM) NaCl 124, KCl 3, KH2PO4 1.25, MgSO4 1.0, CaCl2 2.0, NaHCO3 26, glucose 10, (pH 7.4) containing 0.1 mM physostigmine and 2 lM choline chloride]. The hippocampus was cut into 400-lm-thick slices using a microslicer (DTK-1000; Dosaka EM, Kyoto, Japan), and the slices were preincubated with oxygenated superfusion buffer at 37 °C for 60 min before being transferred to a Brandel Superfusion apparatus (SF-12, Superfusion 1000, BRANDEL, Gaithersburg, MD, USA). The flow rate of the oxygenated superfusion buffer was 1.0 mL/min. Following a 60-min stabilization period, the slices were stimulated with oxygenated high K+-Krebs buffer (25 mM KCl) at 37 °C for 50 min, and the perfusate of the last 20 min was collected as no-drug phase (S1). The slices were next stabilized again in fresh oxygenated superfusion buffer for 50 min and then stimulated with oxygenated high K+-Krebs buffer containing test compounds/vehicle at 37 °C for 70 min. The perfusates of the 30th to the 40th min and that of the 60th to the 70th min were collected as drugtreatment phase (S2). The perfusates S1 and S2 were stored at 4 °C until use for ACh measurement. ACh measurement. ACh in the perfusates S1 and S2 was measured by high-performance liquid chromatography (HPLC) using an electrochemical detector (ECD-300, Eicom Co., Kyoto, Japan), an analytical reverse phase column (AC-GEL, Eicom Co., Kyoto), and a cholinesterase/choline oxidase column (Eicom Co., Kyoto) containing immobilized cholinesterase and choline peroxide for conversion of ACh to hydrogen peroxide. The hydrogen peroxide was detected by a platinum electrode (WE-PT, Eicom Co., Kyoto) set at 450 mV. The mobile phase was composed of 50 mM pyrophosphoric acid buffer (pH 8.2) containing sodium 1-decansulfonate (375 mg/L) and di-sodium dihydrogen ethylenediaminetetraacetate dihydrate (5 mg/L), and the system flow rate was 0.15 mL/min. Each perfusate (10 lL) containing isopropylhemicholine (1 pmol) as an internal standard was injected into the HPLC system, and the release of ACh was expressed as % of the value of S2/S1 in the control group. In vivo microdialysis Surgery. Male Wistar rats weighting 260–360 g were anesthetized with sodium pentobarbital (50 mg/kg, i.p.) and stereotaxically implanted with a guide cannula as dialysis probe into the hippocampus (5.8 mm posterior to bregma, 5 mm lateral from midline and 2.5 mm depth from dura) according to the brain atlas of Paxinos and Watson (1982). The implanted guide cannula was fixed

355

T. Hashimoto et al. / Neuroscience 256 (2014) 352–359

to the skull with dental cement. On completion of the surgery, a polyethylene catheter (1 mm in o.d.) was inserted 10 mm into the gastric body of each animal, tunneled under the skin, and exteriorized in the back of the neck. The free end of the catheter was protruded 2 cm and covered with a polyethylene cap. Microdialysis. Three to eight days after surgery, the extracellular ACh level in the hippocampus was measured by microdialysis in freely moving rats kept in plastic boxes (30  30  30 cm). A microdialysis probe (A-I-8-05, Eicom Co., Kyoto) was inserted through the guide cannula and perfused with Ringer’s solution [(in mM) NaCl 147, KCl 4, and CaCl2 2.3 in distilled water] at a flow rate of 1 lL/min. Dialysate aliquots were collected every 20 min and directly injected into the HPLC system for ACh measurement. Each dialysate was supplemented with isopropylhemicholine (0.2 pmol) as an internal standard. After a stabilization period of at least 60 min, three samples were collected as baseline. The rats were then intragastrically treated with AC-3933 or vehicle, and samples were collected for the following 3 h. Statistical analysis Data are expressed as mean and standard error of the mean (SEM). Statistical analysis was performed using the SASÒ system. In the [35S]-TBPS binding assay, statistical significance of the difference between the control group and each test compound-treated group was assessed using Dunnett’s multiple comparison test. In the microdialysis experiment, the ACh level in each dialysate was calculated as fmol/sample and expressed as percentage of the baseline value (average of each rat’s three pre-administration samples). The area under the curve for 2 h [AUC0–2 h (%  h)] was also calculated. Dose–response data from in vitro ACh release and in vivo microdialysis were statistically analyzed using Dunnett’s multiple comparison test. Drugs AC-3933 was synthesized in our Chemical Research Laboratories. Diazepam and clonazepam were purchased from Wako Pure Chemical Industries (Osaka, Japan). Methyl b-carboline-3-carboxylate (b-CCM), flumazenil, GABA, (+)-bicuculline, and picrotoxin were purchased from Sigma–Aldrich (St Louis, MO, USA). [3H]-flumazenil and [35S]-TBPS were purchased from PerkinElmer Life Science Inc. (Boston, MA, USA). In the in vitro experiments, test compounds were dissolved in dimethylsulfoxide (DMSO), and then diluted to test concentrations with deionized water. The concentration of DMSO was less than 0.1% in all experiments. Clonazepam and (+)-bicuculline were initially dissolved in ethanol and 1 N HCl, respectively, and diluted with 15 mM Krebs-Tris buffer (pH 7.4). GABA was dissolved in 15 mM Krebs-Tris buffer (pH 7.4). In the in vivo experiments, test compounds were suspended in 0.5% tragacanth solution.

RESULTS Receptor binding assays BzR binding assay. Competitive binding assay showed that AC-3933 has high affinity for BzR in crude membrane prepared from rat whole brain with a Ki value of 5.15 ± 0.39 nM (Table 1). Flumazenil and b-CCM also showed potent affinity for BzR with Ki values of 5.10 ± 0.11 and 5.97 ± 0.48, respectively. Diazepam has somewhat lower affinity for BzR than AC-3933, b-CCM, and flumazenil. Other receptor binding assays. To assess AC-3933 selectivity for the BzR, we examined the binding affinity of AC-3933 for other receptors, transporters or ion channels (a total of 90 assays) and its inhibitory activity for two enzymes, i.e. acetylcholinesterase and choline acetyltransferase. AC-3933 showed no notable binding affinity for any of the tested receptors, transporters and ion channels tested at a concentration of 1 lM (inhibition < 30%), and no inhibitory activity for the tested enzymes at a concentration of 30 lM. Intrinsic efficacy. GABA shift assay. AC-3933 GABA ratio, which is defined as the ratio of AC-3933 Ki values with and without GABA was 0.84 ± 0.03 (Table 1). GABA ratios of flumazenil, b-CCM and diazepam were 1.19 ± 0.05, 0.52 ± 0.04 and 3.59 ± 0.19, respectively. [35S]-TBPS binding assay. In the presence of 1 lM GABA, AC-3933 at concentrations over 3 nM significantly increased [35S]-TBPS binding to the rat cortical membrane, although the slope of the concentration–response curve was gentle with hardly any change in [35S]-TBPS binding at a AC-3933 concentration range of 100–3000 nM (from 115.2 ± 0.4% to 117.4 ± 1.2%, Fig. 2). The effect of b-CCM on [35S]-TBPS binding was biphasic, i.e. it reached a maximum at 300 nM (130.4 ± 0.8%) and decreased at the higher concentrations of 1000 nM (123.3 ± 1.1%), 3000 nM (119.4 ± 1.4%) and 10,000 nM (105.2 ± 0.9%). Flumazenil at a concentration of 1 nM had no effect on [35S]-TBPS binding, although a slight decrease was observed at 10–1000 nM (95.5 ± 1.2% to 93.3 ± 1.6%). Diazepam Table 1. Biochemical properties of AC-3933 and other BzR ligands Compound

AC-3933 b-CCM Flumazenil Diazepam

Ki (nM)

GABA ratio

GABA( )

GABA(+)

GABA( )/GABA(+)

5.15 ± 0.39 5.97 ± 0.48 5.10 ± 0.11 94.5 ± 6.90

6.11 ± 0.26 11.5 ± 1.10 4.30 ± 0.08 26.3 ± 0.50

0.84 ± 0.03 0.52 ± 0.04 1.19 ± 0.05 3.59 ± 0.19

Binding of [3H]flumazenil (1 nM) was performed by incubating crude membrane preparation of rat whole brain at 25 °C for 15 min. Nonspecific binding was determined in the presence of clonazepam (1 lM). Data are expressed as the mean ± SEM of three independent experiments.

356

T. Hashimoto et al. / Neuroscience 256 (2014) 352–359

Fig. 3. Effects of AC-3933 on K+-evoked ACh release from rat hippocampal slices. Tissue slices were depolarized with 25 mM KCl. The perfusates were collected for 20 min as pre-stimulation phase (S1; in the absence of AC-3933) and as post-stimulation phase (S2; in the presence of AC-3933 at each concentration). Results are expressed as % of S2/S1 ratio in the control group (mean ± SEM) from three independent experiments. n = 9 slices per group. # P < 0.05, ##P < 0.01, significantly different from the control group (Dunnett’s multiple comparison test).

K+-evoked release of ACh, as assessed by comparing the S2/S1 ratio. The increasing effect of AC-3933 was significant at 1 lM (20.8% increase) and 10 lM (32.9% increase).

Fig. 2. Effects of AC-3933 and other BzR ligands on [35S]-TBPS binding to rat cortical membrane. The effects of test-compounds (A; AC-3933, B; b-CCM, C; flumazenil and diazepam) were expressed as percentage of [35S]-TBPS binding. The reaction was performed with [35S]-TBPS (2 nM) in the presence of GABA (1 lM) at 25 °C for 90 min. Data represent mean ± SEM from three independent experiments. #P < 0.05, ###P < 0.001 significantly different from the control (Dunnett’s multiple comparison test).

concentration-dependently decreased [35S]-TBPS binding to 46.2 ± 2.0% of the control at 10,000 nM. ACh release In vitro experiment. In control preparations, the S2/S1 ratios for ACh release from rat hippocampal slices under 25 mM KCl were 0.94 ± 0.04 (S1; 808 ± 72.4 fmol/mL, S2; 755 ± 65.3 fmol/mL) (Fig. 3). AC-3933 (0.1–10 lM) produced a concentration-dependent increase in high

In vivo experiment. To evaluate the effect of AC-3933 on endogenous ACh outflow in the rat hippocampus under normal conditions, we conducted an in vivo microdialysis where the hippocampal dialysis membrane was perfused with Ringer’s solution without physostigmine. Basal ACh outflow from the hippocampus of free-moving rats was 46.4 ± 1.77 fmol/ fraction (n = 24) and remained stable throughout the experiment in vehicle-treated rats. Fig. 4A shows the time course of ACh outflow from the hippocampus of free-moving rats. Intragastric administration of AC-3933 at 3 and 10 mg/kg increased rats’ hippocampal ACh outflow 134 ± 15.4% (n = 8) and 175 ± 16.7% (n = 8) over the baseline value, respectively. Fig. 4B shows AUC0–2 h of the ACh level in the hippocampus of freemoving rats after intragastric administration of AC-3933 at 3 and 10 mg/kg. AUC0–2 h of the ACh level in the vehicle-treated group was 187.7 ± 7.4% of baseline. AC-3933 at the doses of 3 and 10 mg/kg increased the ACh level to 236.0 ± 11.9% and 288.3 ± 31.2% of baseline, respectively. The effect of AC-3933 at 10 mg/kg was statistically significant (p < 0.01), when compared to the vehicle.

DISCUSSION In the present study, we evaluated some of the pharmacological properties of AC-3933, a novel BzR partial inverse agonist developed for the treatment of AD. Radioligand binding experiments confirmed that AC-3933 is a potent and selective BzR partial inverse agonist. In addition, in vitro and in vivo ACh release experiments showed that AC-3933 significantly augments ACh release from rat hippocampus.

T. Hashimoto et al. / Neuroscience 256 (2014) 352–359

357

Fig. 4. Effects of AC-3933 on extracellular ACh level in the hippocampus of freely moving rats. (A) The results are expressed as percentage change (mean ± SEM) from the mean of three collection periods (basal value) before AC-3933 administration, shown by the arrow. Animals were intragastrically given AC-3933 at 3 (closed circles) or 10 (closed squares) mg/kg, or the vehicle (open circles). n = 8 per group. (B) Each column represents mean ± SEM of AUC for each group (n = 8) 2 h after administration of AC-3933 or the vehicle. ##P < 0.01, significantly different from the vehicle-treated group (Dunnett’s multiple comparison test).

In BzR binding assay, AC-3933 showed high binding affinity for rat brain BzR with a Ki value almost equivalent to those of b-CCM and flumazenil. In addition, AC-3933 showed no noticeable binding affinity for any of the other receptors, transporters, and ion channels tested in this study, and no inhibitory activity for acetylcholinesterase and choline acetyltransferase, suggesting that AC-3933 is a selective BzR ligand. On the other hand, the BzR agonist diazepam had high Ki values in the absence of GABA. Several reports have shown that BzR agonists binding to the BzR is strongly affected by GABA due to a conformational change in BzR caused by GABA (Mo¨hler and Richards, 1981; Chiu and Rosenberg, 1983). In our experiment, BzR binding assay was carried out using frozen membrane preparations in the presence of bicuculline to completely inhibit GABA endogenous effect. Under these conditions, it is believed that the binding affinity of diazepam for BzR is lower than that under normal conditions. The BzR is located in the site of GABAA receptor/ channel complex where BzR ligands can allosterically modulate the activity of GABA-coupled chloride channel. Hence, BzR ligands are classified into three groups (agonists, antagonists and inverse agonists) depending on their intrinsic activity (Braestrup et al., 1983; Haefely et al., 1993). It is believed that BzR inverse agonists with a low GABA ratio have high intrinsic activity (Evans and Lowry, 2007). Since AC-3933 had in this study a GABA ratio (0.84) higher than that of the full BzR inverse agonist b-CCM, it is suggested that AC-3933 intrinsic activity is lower than that of b-CCM. The [35S]-TBPS binding assay is known as a reliable method to evaluate the in vitro intrinsic activity of BzR ligands (Im and Blakeman, 1991; Serra et al., 1994). [35S]-TBPS is a specific ligand for the picrotoxin binding site that is presumably located near the chloride ionophore of the GABAA receptor/channel complex.

Change in TBPS binding is considered to be related to a conformation change of the GABA-coupled chloride channel allosterically induced by the binding of a BzR ligand to BzR (Squires et al., 1983; Othman et al., 2012). It has been reported that, in the presence of micromolar concentrations of GABA, BzR inverse agonists enhance [35S]-TBPS binding, while BzR agonists decrease this binding (Squires et al., 1983; Gee et al., 1986). Our results show that AC-3933 significantly increases [35S]-TBPS binding in the presence of 1 lM GABA, although the slope of the concentration–response curve of AC-3933 was gentle making the increase in [35S]-TBPS binding reach a plateau at over 100 nM. The maximum increase in [35S]TBPS binding induced by AC-3933 was smaller than that induced by b-CCM. As AC-3933 affinity for BzR was almost equivalent to that of b-CCM, and AC-3933 GABA ratio was higher than that of b-CCM, it is believed that AC-3933 is a BzR partial inverse agonist with low intrinsic activity. In the present study, the concentration–response curve of AC-3933 in [35S]-TBPS binding assay was monophasic, while that of the b-carboline inverse agonist b-CCM was biphasic. Im et al. (1995) have reported that the b-carboline inverse agonist methyl-6,7dimethoxy-4-ethyl-beta-carboline-3-carboxylate (DMCM) produces a biphasic effect on GABA-induced chloride ion currents in the GABAA receptor. In addition, Ma¨kela¨ et al. (1999) have demonstrated that DMCM modulates [35S]-TBPS binding biphasically. These results suggest that b-carbolines interact not only with the BzR binding site, but also with a novel site distinct from BzR on the GABAA receptor. Although a detailed investigation is needed to clarify AC-3933 binding site on the GABAA receptor, the results of our [35S]-TBPS binding experiment suggest that the binding site of AC-3933 on the GABAA receptor is different from that of b-carbolines.

358

T. Hashimoto et al. / Neuroscience 256 (2014) 352–359

Many studies have shown that GABA, an endogenous inhibitory neurotransmitter, affects cholinergic neurotransmission in the brain, including the hippocampus (Moore et al., 1993; Moor et al., 1998; Roland and Savage, 2009; Yousefi et al., 2012), and that ligands of the GABAA receptor modulate the activity of cholinergic neurons in the brain. For example, muscimol, a GABAA receptor agonist, decreases ACh release in the hippocampus (Moor et al., 1998), whereas bicuculline, a GABAA receptor antagonist, increases ACh release in the hippocampus (Moor et al., 1998; Roland and Savage, 2009). Our preliminary experiments also demonstrated that bicuculline significantly increases high K+-evoked endogenous ACh release from rat hippocampus slices (data not shown). Since the GABAA receptor/channel complex is allosterically modulated by BzR, BzR ligands are known to modulate cholinergic neuronal activity. It has been reported that BzR agonists produce a marked depression of the rhythmically bursting activity of septohippocampal neurons (Bassant et al., 1988), and that the BzR inverse agonist increases ACh in the frontparietal cortex (Abe et al., 1998). In the present study, AC-3933 significantly increased ACh release from the hippocampus both in vitro and in vivo, suggesting that this compound is a negative modulator for GABAA receptor/channel complex and enhances cholinergic neuronal activity in the hippocampus. The hippocampus plays an important role in learning and memory and its cholinergic neuronal system is a major modulator of memory formation (Carre and Harley, 2000; Kirby and Rawlins, 2003). As cholinergic neuronal dysfunction and cholinergic neuronal loss have been observed in patients with senile dementia, including AD, AChEIs that can increase the ACh level in the brain by inhibiting the degradation of this neurotransmitter, such as donepezil, are prescribed for amelioration of memory impairment. Our two ACh release experiments showed that AC-3933 increases endogenous ACh release from the hippocampus. In addition, AC-3933 showed no inhibition of cholinesterase. These findings indicate that AC-3933 has great potential in the treatment of AD both when used alone or in combination with existing therapies. Further studies will be conducted to confirm the usefulness of AC-3933 as a therapeutic drug for AD.

CONCLUSION In the present study, we showed that AC-3933, a potent and highly selective ligand for the BzR with low intrinsic activity, increases hippocampal ACh release both in vitro and in vivo. These findings indicate that AC-3933 has great potential in the treatment of cognitive disorders associated with degeneration of the cholinergic system. Acknowledgment—The authors thank Dr. Osamu Odai, Kaoru Masumoto and Dr. Kazunori Ono for the chemical synthesis of AC-3933. We also thank Yoshiaki Ochi for his contribution to the research of this paper.

REFERENCES Abe K, Takeyama C, Yoshimura K (1998) Effects of S-8510, a novel benzodiazepine receptor partial inverse agonist, on basal forebrain lesioning-induced dysfunction in rats. Eur J Pharmacol 347:145–152. Appleyard ME, Taylor SC, Little HJ (1990) Acetylcholinesterase activity in regions of mouse brain following acute and chronic treatment with a benzodiazepine inverse agonist. Br J Pharmacol 101:500–604. Bassant MH, Jobert A, Dutar P, Lamour Y (1988) Effect of psychotropic drugs on identified septohippocampal neurons. Neuroscience 27:911–920. Bertz RJ, Reynolds IJ, Kroboth PD (1995) Effect of neuroactive steroids on [3H]flumazenil binding to the GABAA receptor complex in vitro. Neuropharmacology 34:1169–1175. Braestrup C, Nielsen M, Honore´ T, Jensen LH, Petersen EN (1983) Benzodiazepine receptor ligands with positive and negative efficacy. Neuropharmacology 22:1451–1457. Bullock R, Dengiz A (2005) Cognitive performance in patients with Alzheimer’s disease receiving cholinesterase inhibitors for up to 5 years. Int J Clin Pract 59:817–822. Carre GP, Harley CW (2000) Glutamatergic activation of the medial septum complex: an enhancement of the dentate gyrus population spike and accompanying EEG and unit changes. Brain Res 861:16–25. Carter J, Lippa CF (2001) Beta-amyloid, neuronal death and Alzheimer’s disease. Curr Mol Med 1:733–737. Chiu TH, Rosenberg HC (1983) Conformational changes in benzodiazepine receptors induced by the antagonist Ro 151788. Mol Pharmacol 23:289–294. Chorvat RJ, Zaczek R, Brown BS (1998) Ion channel modulators that enhance acetylcholine release: potential therapies for Alzheimer’s disease. Expert Opin Investig Drugs 7:499–518. Christian CA, Herbert AG, Holt RL, Peng K, Sherwood KD, PangratzFuehrer S, Rudolph U, Huguenard JR (2013) Endogenous positive allosteric modulation of GABAA receptors by diazepam binding inhibitor. Neuron 78:1063–1074. Conto´ MB, de Carvalho JG, Benedito MA (2005) Behavioral differences between subgroups of rats with high and low threshold to clonic convulsions induced by DMCM, a benzodiazepine inverse agonist. Pharmacol Biochem Behav 82:417–426. Evans AK, Lowry CA (2007) Pharmacology of the beta-carboline FG-7,142, a partial inverse agonist at the benzodiazepine allosteric site of the GABA A receptor: neurochemical, neurophysiological, and behavioral effects. CNS Drug Rev 13:475–501. Farlow M, Gracon SI, Hershey LA, Lewis KW, Sadowsky CH, DolanUreno JA (1992) Controlled trial of tacrine in Alzheimer’s disease. The Tacrine study group. J Am Med Assoc 268:2523–2529. Farlow MR, Hake A, Messina J, Hartman R, Veach J, Anand R (2001) Response of patients with Alzheimer’s disease to rivastigmine treatment is predicted by the rate of disease progression. Arch Neurol 58:417–422. Gardner CR (1992) A review of recently-developed ligands for neuronal benzodiazepine receptors and their pharmacological activities. Prog Neuropsychopharmacol Biol Psychiatry 16:755–781. Gee KW, Lawrence LJ, Yamamura HI (1986) Modulation of the chloride ionophore by benzodiazepine receptor ligands: influence of gamma-aminobutyric acid and ligand efficacy. Mol Pharmacol 30:218–225. Haefely W, Kyburz E, Gereche M, Mo¨hler H (1985) Recent advances in the molecular pharmacology of benzodiazepine receptors and in the structure–activity relationships of their agonists and antagonists. In: Testa B, editor. Advances in drug research, Vol. 14. London: Academic Press. p. 165–322. Haefely WE, Martin JR, Richards JG, Schoch P (1993) The multiplicity of actions of benzodiazepine receptor ligands. Can J Psychiatry 38:S102–S108.

T. Hashimoto et al. / Neuroscience 256 (2014) 352–359 Im WB, Blakeman DP (1991) Correlation between gammaaminobutyric acidA receptor ligand-induced changes in t-butylbicyclophosphoro[35S]thionate binding and 36Cl- uptake in rat cerebrocortical membranes. Mol Pharmacol 39:394–398. Im HK, Im WB, Carter DB, McKinley DD (1995) Interaction of betacarboline inverse agonists for the benzodiazepine site with another site on GABAA receptors. Br J Pharmacol 114:1040–1044. Imperato A, Dazzi L, Obinu MC, Gessa GL, Biggio G (1993) Inhibition of hippocampal acetylcholine release by benzodiazepines: antagonism by flumazenil. Eur J Pharmacol 238:135–137. Kar S, Issa AM, Seto D, Auld DS, Collier B, Quirion R (1998) Amyloid beta-peptide inhibits high-affinity choline uptake and acetylcholine release in rat hippocampal slices. J Neurochem 70:2179–2187. Katzman R (1986) Alzheimer’s disease. N Engl J Med 314:964–973. Kirby BP, Rawlins JN (2003) The role of the septo-hippocampal cholinergic projection in T-maze rewarded alternation. Behav Brain Res 143:41–48. Knapp MJ, Knopman DS, Solomon PR, Pendlebury WW, Davis CS, Gracon SI (1994) A 30-week randomized controlled trial of highdose tacrine in patients with Alzheimer’s disease. The Tacrine study group. J Am Med Assoc 271:985–991. Kosik KS (1991) Alzheimer plaques and tangles: advances on both fronts. Trends Neurosci 14:218–219. Marin DB, Swell MC, Schlechter A (2002) Alzheimer’s disease. Accurate and early diagnosis in the primary care setting. Geriatrics 57:36–40 (Quiz 43). Maubach K (2003) GABA(A) receptor subtype selective cognition enhancers. Curr Drug Targets CNS Neurol Disord 2:233–239. McMullan J, Sasson C, Pancioli A, Silbergleit R (2010) Midazolam versus diazepam for the treatment of status epilepticus in children and young adults: a meta-analysis. Acad Emerg Med 17:575–582. Michels G, Moss SJ (2007) GABAA receptors: properties and trafficking. Crit Rev Biochem Mol Biol 42:3–14. Mo¨hler H, Richards JG (1981) Agonist and antagonist benzodiazepine receptor interaction in vitro. Nature 294:763–765. Moor E, DeBoer P, Westerink BH (1998) GABA receptors and benzodiazepine binding sites modulate hippocampal acetylcholine release in vivo. Eur J Pharmacol 359:119–126. Moore H, Sarter M, Bruno JP (1993) Bidirectional modulation of stimulated cortical acetylcholine release by benzodiazepine receptor ligands. Brain Res 627:267–274. O’Brien JT, Burns A, BAP Dementia Consensus Group (2011) Clinical practice with anti-dementia drugs: a revised (second) consensus statement from the British Association for Psychopharmacology. J Psychopharmacol 25:997–1019. Othman NA, Gallacher M, Deeb TZ, Baptista-Hon DT, Perry DC, Hales TG (2012) Influences on blockade by t-butylbicyclophosphoro-thionate of GABA(A) receptor spontaneous gating, agonist activation and desensitization. J Physiol 590:163–178. Paxinos G, Watson C (1982) The rat brain in stereotaxic coordinates. San Diego: Academic Press. Perry RJ, Hodges JR (1999) Attention and executive deficits in Alzheimer’s disease. A critical review. Brain 122(Part 3):383–404.

359

Piau A, Nourhashe´mi F, Hein C, Caillaud C, Vellas B (2011) Progress in the development of new drugs in Alzheimer’s disease. J Nutr Health Aging 15:45–57. Podhorna J, McCabe S, Brown RE (2002) Male and female C57BL/6 mice respond differently to diazepam challenge in avoidance learning tasks. Pharmacol Biochem Behav 72:13–21. Polc P (1988) Electrophysiology of benzodiazepine receptor ligands: multiple mechanisms and sites of action. Prog Neurobiol 31:349–423. Rabow LE, Russek SJ, Farb DH (1995) From ion currents to genomic analysis: recent advances in GABAA receptor research. Synapse 21:189–274. Roland JJ, Savage LM (2009) Blocking GABA-A receptors in the medial septum enhances hippocampal acetylcholine release and behavior in a rat model of diencephalic amnesia. Pharmacol Biochem Behav 92:480–487. Schwartz RD (1988) The GABAA receptor-gated ion channel: biochemical and pharmacological studies of structure and function. Biochem Pharmacol 37:3369–3375. Selkoe DJ (2001) Alzheimer’s disease: genes, proteins, and therapy. Physiol Rev 81:741–766. Serra M, Ghiani CA, Motzo C, Cuccheddu T, Floris S, Giusti P, Biggio G (1994) Imidazenil, a new partial agonist of benzodiazepine receptors, reverses the inhibitory action of isoniazid and stress on gamma-aminobutyric acidA receptor function. J Pharmacol Exp Ther 269:32–38. Singh S, Kushwah AS, Singh R, Farswan M, Kaur R (2012) Current therapeutic strategy in Alzheimer’s disease. Eur Rev Med Pharmacol Sci 16:1651–1664. Squires RF, Casida JE, Richardson M, Saederup E (1983) [35S]tbutylbicyclophosphorothionate binds with high affinity to brainspecific sites coupled to gamma-aminobutyric acid-A and ion recognition sites. Mol Pharmacol 23:326–336. Tariot PN (2001) Maintaining cognitive function in Alzheimer disease: how effective are current treatments. Alzheimer Dis Assoc Disord 1:S26–S33. Teuber L, Wa¨tjens F, Jensen LH (1999) Ligands for the benzodiazepine binding site – a survey. Curr Pharm Des 5:317–343. Whiting PJ (2006) GABA-A receptors: a viable target for novel anxiolytics? Curr Opin Pharmacol 6:24–29. Wiwanitkit V (2013) Physiological genomics analysis for Alzheimer’s disease. Ann Indian Acad Neurol 16:72–74. Wood PL, Loo P, Braunwalder A, Yokoyama N, Cheney DL (1984) In vitro characterization of benzodiazepine receptor agonists, antagonists, inverse agonists and agonist/antagonists. J Pharmacol Exp Ther 231:572–576. Yamamoto M, Koshimura K, Kawaguchi M, Sohmiya M, Murakami Y, Kato Y (2000) Stimulating effect of erythropoietin on the release of dopamine and acetylcholine from the rat brain slice. Neurosci Lett 292:131–133. Yousefi B, Nasehi M, Khakpai F, Zarrindast MR (2012) Possible interaction of cholinergic and GABAergic systems between MS and CA1 upon memory acquisition in rats. Behav Brain Res 235:231–243.

(Accepted 12 October 2013) (Available online 25 October 2013)