Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Alzheimer's disease (AD) is a chronic neurodegenerative disorder characterized by progressive cognitive decline, debilitating behavioral impairment, and ultimately death. Neurofibrillary tangles and amyloid plaques are the pathological hallmarks of this disease, which is also characterized by a dramatic loss of cholinergic neurons in brain regions involved in higher cognitive functions. It is widely accepted that the loss of cholinergic markers is one of the major neurochemical deficits in AD and that this deficit correlates well with loss of cognitive abilities (Bierer et al., 1995; Palmer, 1996). Despite this, cholinergic therapies, such as inhibitors of acetylcholinesterase, have had only partial success (e.g., Volger, 1991), possibly due to the multifactorial nature of the deficits and poorly understood etiology of AD. Thus, there is a major need for novel therapeutic approaches that can alleviate the symptoms of the disease.

In addition to the cholinergic deficit, loss of several other neurotransmitter systems has been reported in AD, including norepinephrine, dopamine (see Gottfries, 1990 for a review), and 5-hydroxytryptamine (5-HT) (Palmer, 1996; Meltzer et al., 1998). This therefore provides other opportunities for therapeutic intervention, either alone or in combination with compounds acting on the cholinergic system (Cassel and Jeltsch, 1995; Steckler and Sahgal, 1995; Richter-Levin and Segal, 1996). The numerous subtypes of 5-HT receptor that have been described (Hoyer et al., 1994) allow a variety of ways in which serotonergic compounds might affect cognition.

Although there is evidence for the benefic activity of several 5-HT receptor subtypes in cognitive processes (e.g., Barnes et al., 1990; Carli et al., 1997), the role of the 5-HT_1A receptor in cognition is controversial since abundant literature has stressed the inhibitory role of 5-HT_1A receptors on memory formation and retrieval in hippocampus and other brain structures, but other studies have shown that 5-HT_1A agonists reverse an acquisition deficit induced by pharmacological agents, such as scopolamine or 7-chloro-kynurenic acid, in the rat (see for example Carli et al., 1998, 2001 or Meneses and Hong, 1999). An increasing body of evidence supports a major role for the 5-HT₄ receptor subtype in learning and memory. This receptor is a member of the G-protein-coupled 7-transmembrane domain receptor superfamily and is positively coupled to adenylyl cyclase. 5-HT₄ receptors are heterogeneously located in the central nervous system, with high concentrations in numerous structures considered important for mnemonic processes, such as the hippocampus and regions of the cortex (Jakeman et al., 1994; Compan et al., 1996). In the hippocampus, they have been localized to pyramidal neurons (Roychowdhury et al., 1994), where there is in vitro evidence that their activation results in the inhibition of voltage-dependent K⁺ channels via stimulation of a cAMP-dependent protein kinase (Ansanay et al., 1995; Torres et al., 1995). Experiments have shown that only a brief stimulation of the 5-HT₄ receptor is necessary for a slowly reversible blockade of these K⁺ channels, which results in a long-lasting increase in neuronal excitability (Andrade and Nicoll, 1987; Ansanay et al., 1995). Interestingly, there is evidence for a loss of 5-HT₄ receptors in Alzheimer's disease in the cortex (23%) and particularly in the hippocampus (66%) (Reynolds et al., 1995), consistent with their distribution and localization on pyramidal cells and with a potential role in cognitive processing.

In a recent study (Robert et al., 2001), activation of the human 5-HT₄ receptor was shown to stimulate the secretion of the nonamyloidogenic soluble form of the amyloid precursor protein, sAPP. Given the neuroprotective and memory-enhancing effects of sAPP (Mattson et al., 1993; Schubert and Behl, 1993; Behl et al., 1994), 5-HT₄ receptor agonists may have a therapeutic potential for the treatment of Alzheimer's disease. Further support for a therapeutic potential of 5-HT₄ receptor agonists in learning and memory disorders comes from a range of in vivo studies showing that they can improve performance in animal models of learning and memory. Most of these studies have used the relatively nonselective compounds BIMU-1 and BIMU-8 (Eglen et al., 1995), which have shown activity in an olfactory discrimination task (Marchetti-Gauthier et al., 1997), a social memory task (Letty et al., 1997), acquisition of an operant autoshaping task when administered preconditioning (Meneses and Hong, 1997), and in reversing scopolamine-induced amnesia in a mouse passive avoidance task (Galeotti et al., 1998). A specific role for 5-HT₄ receptors in some of these studies is supported by the antagonism of these effects by selective 5-HT₄ antagonists (Letty et al., 1997; Marchetti-Gauthier et al., 1997; Galeotti et al., 1998). More selective compounds, such as RS 17017 and RS 67333, have also been shown to enhance cognition in a delayed-match-to-sample task in aged primates (Terry et al., 1998) and reversal of an atropine-induced deficit in the Morris water maze in rats (Fontana et al., 1997).

A potential drawback of 5-HT₄ agonists as therapeutic agents is their effects mediated in the periphery by this receptor subtype. These include increased gastric motility, cardiac inotropy and tachycardia, and release of adrenal hormones (Hegde and Eglen, 1996), all of which would be undesirable in a clinical setting. We anticipated that a compound with less efficacy at the 5-HT₄ receptor would retain the promnesic activity demonstrated by agonist compounds while avoiding the peripheral effects associated with this receptor. This report describes the properties of a selective 5-HT₄ partial agonist compound, SL65.0155 [5-(8-amino-7-chloro-2,3-dihydro-1,4-benzodioxin-5-yl)-3-[1-(2-phenylethyl)-4-piperidinyl]-1,3,4-oxadiazol-2(3H)-one monohydrochloride] (Fig. 1).

View larger version (8K): [in this window] [in a new window]   Fig. 1.   The structure of SL65.0155.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Radioligand Binding Studies

All radioligand binding studies were carried out either by CEREP (Celle L'Evescault, France) or MDS Pharma Services (Taiwan, Republic of China). Unless otherwise indicated in the results, all studies were initially carried out at 10⁵ M. If there was greater than 75% displacement at this concentration, specific K_i values were determined. Over 100 binding sites were examined, including all other 5-HT receptors and all major classes of neurotransmitter receptor, uptake systems, ion channels, and enzymes. In the case of the 5-HT receptor subtypes, all experiments were carried out using recombinant human receptors cloned into various cell types as indicated in Table 1; also shown in Table 1 are the ligands used and their concentration.

Affinity of SL65.0155 for Different Splice Variants of the 5-HT₄ Receptor. Membranes were prepared from either CHO cells stably transfected with human isoforms of the 5-HT₄ receptor (isoforms c, d, and e) or from transiently transfected COS-7 cells (isoform b), and the ability of SL65.0155 to displace [³H]GR 113808 (Amersham Biosciences, Piscataway, NJ) was examined as previously described (Blondel et al., 1998). Briefly, radioligand binding studies were performed in 500 µl of HEPES buffer (50 mM, pH 7.4) to which was added 50 µl of membrane preparation (100-200 µg), 20 µl of buffer solution containing the competing agent (eight concentrations of SL65.0155 were used or 10 µM of GR 113808 to determine nonspecific binding), and 20 µl of buffer containing [³H]GR 113808 to give a final concentration of 50% of the K_d value. The reaction was stopped by rapid vacuum filtration through Whatman GF/B filter paper (Clifton, NJ) [presoaked in a solution of polyethylenimine (0.1%) to reduce binding to filters] using a Brandel 48R cell harvester (Brandel Inc., Gaithersburg, MD). Filters were subsequently washed with ice-cold buffer (50 µM Tris-HCl, pH 7.4) and placed overnight in 4 ml of Ready-Safe scintillation cocktail (Beckman Coulter, Inc., Fullerton, CA). Radioactivity was measured using a Beckman model LS 6500C liquid scintillation counter. Each experiment was carried out two or three times in triplicate. Binding data were analyzed by computer-assisted nonlinear regression analysis (GraphPad Prism; GraphPad Software, San Diego, CA).

Functional Effects at the 5-HT₄ Receptor

Stimulation of cAMP Production in CHO Cells Expressing Splice Variants of the Human 5-HT₄ Receptor. The 5-HT_4(a) variant of the human 5-HT₄ receptor was transiently expressed in COS-7 cells, and changes in cAMP formation in response to compounds were evaluated as described previously using [³H]adenine to label the ATP pool (Claeysen et al., 1997). The 5-HT_4(b) and 5-HT_4(e) variants of the human 5-HT₄ receptor were expressed in CHO cells (transiently and stably, respectively), and changes in cAMP formation in response to compounds were evaluated as described previously using a radioimmunological kit (ERIA model 79830; Pasteur Diagnostics, Marnes-La-Coquette, France) (Blondel et al., 1998). At least six different concentrations were used, and each concentration was evaluated in two or three separate experiments. Nonlinear curve fitting was used to determine the maximal effect and EC₅₀ values.

Rat Esophagus. The 5-HT₄ agonist activity of SL65.0155 was assessed by measuring the relaxation of the muscularis mucosae of rat distal esophagus precontracted with carbachol compared with that induced by 5-HT. The thoracic esophagus was isolated from male Sprague-Dawley rats (350-450 g) and placed in Krebs' solution (118 mM NaCl; 4.7 mM KCl; 1.64 mM MgSO₄; 1.18 mM KH₂PO₄; 11.5 mM glucose; 24.88 mM NaHCO₃; and 2.52 mM CaCl₂). The outer striated muscle coat was cut longitudinally and peeled away to expose the inner muscularis mucosae. Tissues were mounted in 20-ml organ baths containing Krebs' solution maintained at 32°C and gassed with 95% O₂/5% CO₂, pH 7.4. Contractions were recorded isometrically with an initial resting tension of 0.5 g applied to the preparation. After 30 min, carbachol (0.5 µM) was added to the bath, and the response to 5-HT (1 µM) was determined. After washing out the 5-HT and a 20-min re-equilibration period, cumulative concentration-effect curves were constructed for SL65.0155 (0.1-1 µM) in carbachol-precontracted tissue to determine its agonist effect. Responses to the addition of SL65.0155 were expressed as a percentage of the effect of 5-HT at 1 µM. At the highest concentration of SL65.0155, a cumulative concentration-effect curve to 5-HT was constructed and compared with a pre-established 5-HT control curve to determine the pK_b as a measure of the antagonist properties of SL65.0155.

Guinea Pig Ileum. Following sacrifice, the distal segment of the ileum was rapidly removed from male Dunkin-Hartley guinea pigs weighing approximately 350 g. The ileum segment was cleaned and then cut into longitudinal segments approximately 1 cm in length. These segments were then placed in an organ bath containing oxygenated (95% O₂/5% CO₂) physiological solution (118 nM NaCl, 4.7 mM KCl, 1.2 mM MgSO₄, 2.5 mM CaCl₂, 1.2 mM KH₂PO₄, 25 mM NaHCO₃, and 11 mM glucose) at pH 7.4 and a temperature of 37°C and connected to an isotonic strain gauge (model 7006; Ugo Basile, Comerio, Italy) to measure smooth muscle contraction. Experiments were carried out in the presence of pyrilamine (1 µM), methysergide (1 µM), and ondansetron (10 µM). After installation in the organ bath, the tissue was placed under 1 g of tension and allowed to stabilize for 60 min, after which the tension was readjusted to 1 g. Agonist dose-response curves were obtained by administering different concentrations of SL65.0155 or 5-HT to the bath at 40-min intervals. Each concentration was left in contact with the tissue until the maximal effect had been obtained or for a maximum of 5 min. The tissue was washed between drug administrations. Antagonist effects of SL65.0155 were evaluated by exposing the tissue to one of three concentrations (10, 30, or 100 nM) 60 min prior to the determination of a concentration-response curve to 5-HT (six determinations per concentration of SL65.0155).

Evaluation of Cognition-Enhancing Activity

Visual Recognition Memory in Young Rats. The effects of SL65.0155 on visual recognition memory were assessed using an object recognition task similar to that described by Ennaceur and Delacour (1988). Male Sprague-Dawley rats (180-230 g at testing) were housed in pairs with free access to food and water in a temperature-controlled animal room (21 ± 1°C) on a 12-h light/dark cycle (dark period from 7:00 AM-7:00 PM). The apparatus consisted of a uniformly lit hardboard enclosure (65 × 45 × 45-cm) observed from an adjacent room via a video monitoring system. Each experiment consisted of three sessions. During the first session, the animals were allowed 2 min to become familiar with the experimental environment. Time spent active (animal moving around with or without sniffing and exploration) was measured. The animals were again placed in the enclosure 24 h later in the presence of two identical objects for the amount of time necessary to spend 20 s exploring these two objects to a limit of 3 min (exploration was defined as the animal having its head within 2 cm of the object while looking at, sniffing, or touching it). Any animal not exploring the objects for 20 s within the 3-min period was eliminated. After an interval of either 1 min or 24 h, the rats were again placed in the enclosure with a previously presented familiar object and a novel object for a period of 3 min. Time spent exploring the familiar and novel objects was recorded. Animals exploring the objects for less than 10 s were eliminated from the study. The objects used were a metal triangle (7 × 3 × 8-cm) and a plastic pyramid (9 × 3 × 7-cm). Two different sets of objects were used to enable them to be wiped between one rat and the next, to prevent the possibility of olfactory recognition. SL65.0155 was studied at doses of 0.00001 to 0.1 mg/kg p.o. (doses expressed as salt) with tacrine (0.5, 1, and 2 mg/kg p.o.) as a reference. In the study examining the interaction of SL65.0155 with the 5-HT₄ receptor antagonist SDZ 205,557 [4-amino-5-chloro-2-methoxy-benzoic acid-(diethylamino) ethyl ester hydrochloride], doses of 0.1 mg/kg p.o. and 1 mg/kg i.p. were used, respectively. In a study to determine the interaction between SL65.0155 and the cholinesterase inhibitor rivastigmine, doses of 0.0001 mg/kg p.o. and 0.03 mg/kg i.p. were used, respectively. This dose of rivastigmine was chosen because preliminary studies showed that it did not affect behavior in this task and was not associated with cholinergic side effects. Higher doses (0.1 and 0.3 mg/kg, i.p.) slightly but significantly improved performance with the appearance of side effects at the higher dose. Drugs were administered either 30 (i.p.) or 60 min (p.o.) before each of the three experimental sessions. Data were analyzed by analysis of variance for comparisons between treatments and between familiar and novel objects for each treatment.

Acquisition of a Linear Maze Task in Aged Rats. The effects of SL65.0155 on acquisition of a linear maze task were investigated in aged rats (21-23 months old at the start of experiments). These experiments used a total of 63 male Wistar rats housed five animals per cage (42 × 42 × 18-cm) in a temperature-controlled room (21 ± 1°C) with a 12h/12h dark/light cycle (lights on at 7:00 AM). Water was freely available, but animals were kept on a food deprivation schedule of 30 min access to food per day. The linear maze consisted of a start box, six-choice units (50 × 40 × 35-cm) with 10-cm wide openings between each of them, and a goal box (25 × 25 × 35-cm). All units and boxes were made of beige polyvinyl chloride, and start and goal boxes were separated from other units by sliding doors. The task required rats to learn a sequence of six successive left-right choices to reach the goal box and obtain food reinforcement. Vertical barriers made of clear Plexiglas could be changed from one side to the other in each unit to form blind alleys. Illumination was provided by six 60-W lamps 2.5 m above each choice point to avoid shadows in the apparatus. Six different maze configurations were used and randomized between animals: LRRLLR, LRLLRR, RRLLRL, RLRRLL, RLLRRL, and LLRRLR (where R = right and L = left). The reinforcement (Kellogg's Chocopops) was placed in a white porcelain receptacle in the goal box. Prior to training, there was a habituation period (4 or 5 days), which allowed the animals to become familiar with the environmental context and to associate reinforcement and goal box. No barriers were present during this stage. The learning period consisted of two trials per day (one in the morning and one in the afternoon) for 10 days. For each trial, the rat was placed in the start box, the door was immediately opened, and the rat was allowed to reach the goal box and obtain the reinforcement (maximum 15 min). The number of errors (defined as an entry into a blind alley) made by rats was recorded and summed for each pair of trials per day. After an error, rats could return toward the start box, thus allowing them to re-enter a blind alley more than once during the same trial. Statistical analyses were performed according to a multistage procedure. Data were analyzed using one- or two-factor ANOVA for repeated measures, as appropriate, followed by additional ANOVA or t test comparisons if significant (p < 0.05). SL65.0155 (0.1 and 1.0 mg/kg i.p.) or rivastigmine (0.1 and 0.25 mg/kg i.p.) were dissolved in saline (0.9% NaCl w/v) and injected in a volume of 2 ml/kg 30 min before each acquisition trial. Doses are expressed as salt.

Reversal of Scopolamine-Induced Deficits in the Water Maze in Mice. The water maze task evaluates spatial memory acquisition by testing the ability of animals to find an invisible platform located in a circular arena (1.5-m diameter filled with water at 23°C) using distal environmental cues. The platform (10-cm diameter transparent polymethyl methacrylate) is made invisible by placing it 0.5 cm below the surface of the water. Male B6D2F₁ mice were placed in the water maze at one of four start points located around the perimeter and were required to swim in the maze until they found the platform. If they did not find the platform in 60 s, they were placed on the platform for 10 s. Mice were subjected to four trials per day (each trial separated by 5 min, during which the mouse was replaced in the home cage) for 4 days. The fourth trial of the final day was a retention test, in which the platform was removed and mice allowed to swim in the maze for 30 s. Possible confounding effects of motivational or visual deficits were evaluated with a training session using a visible platform (an opaque platform 0.5 cm above the surface of the water) carried out 4 days after the retention test. Four treatment groups were studied, each consisting of nine mice: controls, scopolamine HBr (0.3 mg/kg i.p. in 0.9% NaCl), and two groups receiving scopolamine and either 0.1 or 0.3 mg/kg i.p. of SL65.0155 dissolved in a solution of 1% Tween 80 and 0.01% dimethicone. All mice received two injections of drug or vehicle 20 min prior to the first trial on each day of testing. The trajectory taken by each animal was monitored by an image analysis system (Videotrack; Viewpoint, Lyon, France), which measured latency and distance taken to the platform for all acquisition and visible platform trials. During the retention trial, the additional parameters of time spent in each quadrant and annulus crossing (i.e., number of occasions the mouse swam over the former platform location and corresponding positions in other quadrants) were recorded. Statistical analysis consisted of an initial ANOVA comparison of control and scopolamine-treated groups (treatment and session as factors) followed by a similar comparison of the scopolamine-treated group with those receiving scopolamine and SL65.0155.

Improvement of Social Olfactory Recognition in Rats. The effects of SL65.0155 on social olfactory recognition were assessed using the procedure previously described (Pério et al., 1989). Juvenile Wistar rats (3 weeks old, 50-100 g, housed seven or eight per cage) and 10 adult Wistar rats [400-500 g, housed individually (home cage dimensions, 30 × 40 × 18 cm)] were maintained under conditions of constant temperature (22 ± 2°C) on an inverted light/dark cycle (lights on from 7:00 PM-7:00 AM). Food and water were freely available. Experiments were conducted in the animal room, which was under red illumination during the dark phase. The juveniles were isolated for 30 min before the start of each experiment. In the spontaneous forgetting procedure, the juveniles were exposed to the tested conspecific, in the home cage of the conspecific, on two successive presentations of 5-min duration separated by a 120-min interval during which the juveniles were returned to their isolation box. The experiments were monitored via a camera by an observer in an adjacent laboratory. Duration of investigatory behavior (nosing, sniffing, grooming, and close following of the juvenile) was recorded by striking precoded keys on a computer system.

Side Effect Liability

Effect on General Behavior in Rats (Irwin Observation Test). Potential behavioral and neurological effects of SL65.0155 and its effects on body temperature and weight were studied in male Sprague-Dawley rats (150-200 g at testing; n = 5/group). SL65.0155 (0.1-10 mg/kg expressed as the salt) or vehicle (1% Tween 80 and 0.01% dimethicone) were administered orally in a volume of 5 ml/kg. Each animal was observed continuously for autonomic and neurological changes over a 30-min period postdose and then at times 1, 2, 4, 6, and 24 h after treatment. Observation of animals treated with SL65.0155 was carried out in parallel with that of control rats (receiving vehicle). Body temperature was measured using a rectal probe, and animals were weighed before treatment and after 24 h.

Arterial Blood Pressure and Heart Rate in Conscious Normotensive Rats. Male Sprague-Dawley rats (288-324 g at testing; n = 5/group) were anesthetized using an intraperitoneal injection of ketamine (116 mg/kg). A catheter filled with anticoagulant (povidone) solution was inserted into the femoral artery, passed under the skin, and exteriorized between the scapula. The day after the implantation, the catheter was connected to a blood pressure transducer (Statham P23XL model; PTC Electronics, Wyckoff, NJ). After a stable baseline was obtained, SL65.0155 was administered orally, dissolved in water at the doses of 0.1, 1, and 10 mg/kg expressed as the salt in a volume of 5 ml/kg. The effect of SL65.0155 on arterial blood pressure and heart rate was monitored continuously for 4 h. Hemodynamic parameters (systolic and diastolic arterial blood pressure and heart rate) were recorded every 15 min during the first 2 h, then every 30 min. The results are expressed as the difference from baseline values.

Hemodynamic Profile in Anesthetized Dogs. Male beagle dogs, weighing 10 to 16 kg at testing (n = 5 per treatment group) were fasted for 24 h before use. On the day of testing, anesthesia was induced with pentobarbital (30 mg/kg, i.v.) and then maintained with an i.v. infusion of pentobarbital (0.1 mg/kg/min) into the cephalic vein. Animals were placed under artificial ventilation (25 cycles/min and 200 ml/kg/min). Arterial blood pressure and pulmonary artery pressure were measured from catheters inserted into the right femoral aorta and a jugular vein, respectively. Left ventricular pressure was measured by means of a cannula inserted via the left carotid artery. Doppler probes (Transonic Systems T206; Ithaca, NY) were positioned around the right carotid artery, left circumflex coronary artery, aorta, and left renal artery to monitor arterial flow rates. Three electrodes were placed to measure the electrocardiogram. The hemodynamic parameters were measured before drug administration and at various times between 1 and 120 min after the intravenous administration of SL65.0155 at doses of 0.01, 0.1, and 0.5 mg/kg, expressed as the salt, or its vehicle (saline). In addition, arterial blood samples were taken at 15, 30, 60, and 120 min post-treatment for analysis of blood gases and electrolytes.

Intestinal Transit in Rats. The effects of SL65.0155 were studied in male Sprague-Dawley rats weighing 150 to 190 g (n = 15 per treatment group). Rats were fasted for 24 h prior to oral administration of either vehicle (1% w/v Tween 80 in sterile water) or SL65.0155 (0.3, 3, or 30 mg/kg p.o., expressed as base) on the test day. Thirty minutes later, the rats received a charcoal meal (12.5 ml/kg of 10% w/v activated charcoal suspended in 0.25% w/v methyl cellulose and 0.5% w/v Tween 80). A further 30 min later, the animals were sacrificed by an intracardiac injection of a lethal dose of pentobarbital. The small intestine was removed, and the length of the intestine from the pylorus to the ileocaecal junction was measured as well as the length of the part colored by the charcoal. Intestinal transit was expressed as the percentage of intestine colored by charcoal.

Compounds. SL65.0155 HCl, rivastigmine HCl, tacrine HCl, and SDZ 205,557 were all synthesized by the Sanofi-Synthélabo Medicinal Chemistry Group. Scopolamine HBr, carbachol, noradrenaline, and serotonin were purchased from Sigma-Aldrich (St Quentin Fallavier, France). The vehicle and volume of injection used is indicated for each study. All doses refer to salt except where stated.

Animals. Unless otherwise stated, all animals were obtained from Charles River Laboratories (two sources were used: Iffa Credo, L'Arbresele, France and Charles River, St. Aubin les Elbeuf, France). Rats arrived at our animal quarters at least 1 week before use and were housed in groups of two to five per cage (see individual experimental methods for details). The beagle dogs used for the cardiovascular studies were obtained from Marshall Farms (North Rose, NY) and acclimatized to our animal quarters, where they were individually housed for at least 10 days prior to use. All animals were kept under standard conditions of humidity and temperature with lights on between 7:00 AM and 7:00 PM. Food and water were freely available except in the case of the aged Wistar rats, which were placed on a restricted diet of 30-min access to food per day.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Radioligand Binding Studies

SL65.0155 demonstrated high affinity for the human 5-HT₄ receptor (Table 1). In addition, it was found to be extremely selective for this subtype, with more than a 1000-fold difference in affinities for the 5-HT₄ receptor compared with all other 5-HT receptors (Table 1). These studies used the 5-HT_4(e) splice variant of the human 5-HT₄ receptor.

In a separate series of studies, the affinity of SL65.0155 was evaluated against several of the splice variants of the human 5-HT₄ receptor. SL65.0155 displaced binding to the b, c, d, and e isoforms of the 5-HT₄ receptor with affinities of 4.7 ± 0.6, 6.1 ± 2.5, 7.0 ± 1.0, and 16.0 ± 3.0, respectively (all values are K_i in nM). Although these values are slightly higher than those of the initial study due to the different experimental conditions used, this experiment demonstrates that SL65.0155 does not discriminate between isoforms of the 5-HT₄ receptor.

In contrast to its high affinity for 5-HT₄ receptors, SL65.0155 was largely devoid of activity (K_i >1 µM) at most of the other 100 binding sites studied (CEREP and MDS Pharma Services binding profiles; data not shown). The only receptors/binding sites for which SL65.0155 had significant affinity were the rat  receptor (K_i = 94 nM), human noradrenergic _1D receptor (K_i = 111 nM), dopamine D_4.4 receptor (K_i = 129 nM), and histamine H₁ receptor (K_i = 171 nM). Taken overall and comparing studies carried out under similar conditions, these data demonstrate that SL65.0155 can be considered as a selective ligand for the 5-HT₄ receptor, although the selectivity of the compound over , _1D, H₁, and D_4.4 receptors is potentially narrower.

Functional Studies at the 5-HT₄ Receptor

Stimulation of cAMP Production in CHO Cells Expressing Splice Variants of the Human 5-HT₄ Receptor. In cells expressing the 5-HT_4(a) variant of the 5-HT₄ receptor, none of the concentrations of SL65.0155 increased cAMP production and actually reduced basal conversion of ATP to cAMP with a maximal reduction of 32% at 100 nM. This inverse agonist effect was confirmed by the ability of SL65.0155 to antagonize the effect of 5-HT (100 nM) with a K_d of 2.6 nM. In contrast to its antagonist activity against the 5-HT_4(a) variant, SL65.0155 behaved as a partial agonist in cell lines expressing either the 5-HT_4(b) or 5-HT_4(e) isoform of the human 5-HT₄ receptor, stimulating cAMP production to a lesser extent than 5-HT (Table 2).

Rat Esophagus. SL65.0155 was without agonist activity in this preparation but antagonized the effect of 5-HT with a pK_b of 8.81.

Guinea Pig Ileum. SL65.0155 was without contractile effects on the guinea pig ileum preparation. However, pre-exposure to SL65.0155 resulted in a rightward shift of the concentration-response curve to 5-HT with no depression of the maximum response, consistent with competitive antagonism, allowing the calculation of a pA₂ value of 8.29.

Cognitive Effects

Object Recognition Task. This test exploits the tendency of rats to preferentially explore novel elements of their environment. Thus, when a rat is presented with both a novel and a recently presented familiar object, it will spend significantly more time exploring the novel object. If this discrimination task is carried out 24 h after the presentation of the familiar object, however, the rat will no longer differentially explore the two objects, as seen in the 24-h control group (Fig. 2A). Treatment with either tacrine or SL65.0155 prior to each of the three sessions reinstated recognition of the previously exposed object, as demonstrated by significantly more exploration of the novel one (Fig. 2). Although the magnitude of the effect was similar for both compounds, SL65.0155 was considerably more potent (minimal effective dose = 0.001 mg/kg p.o. or i.p.) and had a much wider range of active doses than did tacrine (doses of tacrine higher than 2 mg/kg could not be tested due to the appearance of cholinergic side effects). SL65.0155 showed similar potency and activity via both the i.p. and p.o. routes of administration.

View larger version (44K): [in this window] [in a new window]   Fig. 2.   Effect of SL65.0155 on object recognition performance in the young rat. Introducing a 24-h delay between test and recall sessions induced a marked decrease in discrimination between familiar and novel objects (A). This deficit was reversed by either i.p. (B) or p.o. (C) administration of SL65.0155. Oral tacrine was also effective in improving object recognition (D). The black and gray columns show exploration time for the familiar and novel object, respectively. *, p < 0.05; **, p < 0.01 paired t test comparison between familiar and novel object exploration following a significant ANOVA.

View larger version (42K): [in this window] [in a new window]   Fig. 3.   Reversal of the promnesic effect of SL65.0155 (0.1 mg/kg p.o.) in the object recognition task by the 5-HT4 antagonist SDZ 205,557 (1 mg/kg i.p.). **, p < 0.01 paired t test comparison between familiar and novel object exploration following a significant treatment by object interaction by ANOVA.

View larger version (43K): [in this window] [in a new window]   Fig. 4.   Synergism between SL65.0155 (0.0001 mg/kg p.o.) and rivastigmine (0.03 mg/kg i.p.) in the object recognition test. **, p < 0.01 paired t test comparison between familiar and novel object exploration following a significant treatment by object interaction by ANOVA.

Linear Maze Task. Aged rats show a marked deficit in the acquisition of this task compared with young rats (2-3 months), as shown in the inset to Fig. 5. Not only do they show much slower learning of the task, requiring many more trials to achieve asymptotic levels of performance, but they also continue to make more errors per trial after a stable performance has been reached. Pretreatment of aged rats prior to each trial with either SL65.0155 (0.01 and 0.1 mg/kg i.p.) or rivastigmine (0.1 and 0.25 mg/kg i.p.) improved task performance as measured by both these parameters: both the initial rate of learning and the asymptotic level of performance were significantly better than aged animals receiving vehicle (Fig. 5).

View larger version (32K): [in this window] [in a new window]   Fig. 5.   Effect of SL65.0155 and rivastigmine on linear maze performance in the aged rat. The insert in the right-hand panel shows the deficit in aged rats compared with young animals. Statistical analysis of the SL65.0155 data (following a significant ANOVA result for factor treatment) revealed that both doses significantly improved performance: 0.01 mg/kg, p < 0.05 and 0.1 mg/kg, p < 0.01. A similar analysis of the rivastigmine data revealed that the group receiving 0.1 mg/kg differed from the controls (p < 0.05) but that the higher dose, 0.25 mg/kg, just failed to reach significance (p = 0.058).

Reversal of Scopolamine-Induced Deficits in the Water Maze in Mice. Scopolamine pretreatment (0.3 mg/kg i.p.) induced a significant acquisition deficit in the water maze task as demonstrated by the increased latency to find the platform compared with vehicle-treated mice [control versus scopolamine-treated groups: F(1,16) = 54.6, p < 0.0001; Fig. 6]. This effect was particularly marked on the second and third day of training, and overall latencies averaged across trials were also significantly increased. Similar differences were observed for distance [F(1,16) = 66.4, p < 0.0001; data not shown], and scopolamine-treated mice swam slightly but significantly faster during the third and fourth sessions.

View larger version (30K): [in this window] [in a new window]   Fig. 6.   Reversal of the scopolamine-induced deficit of water maze performance in mice by SL65.0155. The value for each session is the mean of four trials per animal. The insert shows the data collapsed over all four acquisition trials. ##, p < 0.01 compared with vehicle-treated controls; *, p < 0.05, **, p < 0.01 compared with scopolamine alone, Dunnett's test following significant ANOVA.

Improvement of Social Olfactory Recognition in Rats. The promnesic effects of SL65.0155 are presented in Fig. 7. Increasing doses of SL65.0155 reduced the amount of time spent investigating the juvenile during the second presentation as indicated by the significantly decreased RID values [Kruskal-Wallis test for repeated measures: ² value = 17.26 (4 degrees of freedom), p < 0.01, n = 10). Post hoc Wilcoxon comparisons between groups indicated that RIDs were significantly decreased compared with control values at 0.01 and 0.03 mg/kg.

View larger version (12K): [in this window] [in a new window]   Fig. 7.   Promnesic effect of SL65.0155 in the social olfactory recognition test in rats. SL65.0155 was administered immediately after the first presentation of the juvenile rat. The results are expressed as the ratio of the time spent investigating the juvenile during the second presentation compared with the first (RID) and are the means of 10 rats, each tested under each of treatment conditions. The time spent investigating the juvenile during the first period did not differ significantly between groups (the values for the five treatment conditions were 98.3, 103.1, 86.1, 103.7, and 110.5 s, Kruskall-Wallace comparison 2 = 3.94, nonsignificant). *, p < 0.05, **, p < 0.01 compared with control group (black histogram), Wilcoxon test.

Side Effect Liability

Irwin Test. No deleterious effects of SL65.0155 were observed up to the highest dose tested (10 mg/kg p.o.) on any of the parameters measured. This represents a dose more than 1,000-fold greater than that found to be active in the object recognition task.

Cardiovascular Effects in Conscious Normotensive Rats. Neither heart rate nor blood pressure was modified by SL65.0155 at any of the tested doses (0.1-10 mg/kg p.o.) compared with rats receiving vehicle.

Hemodynamic Profile in Anesthetized Dogs. No changes in arterial blood pressure, pulmonary artery pressure, left ventricular pressure (and derived parameters), blood flows, heart rate, or PR, QRS, and QT durations were observed at any of the tested doses of SL65.0155 (0.01-0.5 mg/kg i.v.) (data not shown).

Effect of SL65.0155 on Gastric Motility. SL65.0155 had no significant effects on gastrointestinal motility in rats. Control animals had 85 ± 2% of their intestine colored by charcoal compared with 88 ± 2, 87 ± 2, and 90 ± 1% for the 0.3, 3, and 30 mg/kg-treated animals, respectively.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present results characterize SL65.0155 as a potent and selective ligand for the 5-HT₄ receptor that, in common with several other 5-HT₄ compounds, shows robust cognition-enhancing activity across a range of tasks in rodents. This activity was comparable in magnitude to that seen with acetylcholinesterase inhibitors, currently the only class of compound available for the symptomatic treatment of Alzheimer's disease, but was maintained over a much broader dose range. In functional tests, SL65.0155 behaved as a competitive antagonist when evaluated in the rat esophagus and guinea pig ileum preparations but demonstrated a broader range of functional interactions with the 5-HT₄ receptor when studied against individual 5-HT₄ receptor splice variants. This activity ranged from inverse agonism at the 5-HT_4(a) isoform to partial agonist activity (43-48% of the maximal activity of 5-HT) at the 5-HT_4(b) and 5-HT_4(e) isoforms (see below for further discussion). Thus, unlike other 5-HT₄ compounds that have demonstrated cognition-enhancing activity, SL65.0155 demonstrated much less intrinsic activity. Compounds such as BIMU-1, BIMU-8, and RS 67333 all show moderate to marked efficacy in comparable in vitro preparations, ranging from 50 to 90% of that obtained with 5-HT (Eglen et al., 1995; Mialet et al., 2000a), increasing the likelihood that they will show side effects. Preclinical studies with 5-HT₄ compounds and clinical experience with compounds possessing 5-HT₄ agonist activity (such as cisapride) suggest a propensity to induce tachycardia, increased gastrointestinal activity, and urinary incontinence. SL65.0155 clearly lacks cardiovascular and gastrointestinal effects, consistent with its antagonist activity in isolated peripheral tissues. The data presented here therefore not only demonstrate that SL65.0155 is an effective cognition enhancer with a novel mechanism of action but also suggest that it will have a better clinical side effect profile than currently available treatments for memory deficits.

Given that the literature supports a role for agonist effects at the 5-HT₄ receptor in mediating procognitive effects (see Introduction) and that the effect of SL65.0155 in the object-recognition task was antagonized by the selective 5-HT₄ receptor antagonist SDZ 205,557, it seems reasonable to conclude that SL65.0155 is also acting as an agonist to improve learning and memory performance. The absence of promnesic effects of SDZ 205,557 also supports this suggestion. The partial agonist effect demonstrated in some of the in vitro functional tests is therefore sufficient to produce significant promnesic activity. Given the clear absence of any of the peripheral effects normally associated with 5-HT₄ agonists, it is tempting to speculate further that SL65.0155 might have greater efficacy at central 5-HT₄ receptors (i.e., acting as an agonist or partial agonist) than at peripheral 5-HT₄ receptors, where it appears to act as an antagonist or at least lacks functional agonist activity.

Data obtained in the studies evaluating SL65.0155 at individual splice variants of the 5-HT₄ receptor provide evidence that such a functional dissociation between differently located 5-HT₄ receptors is a possibility. In humans, at least eight splice variants of the 5-HT₄ receptor have been identified to date (labeled a-h), all differing in their C-terminal region (except the h isoform) and resulting in receptors with different degrees of intrinsic coupling to their second messenger systems (Blondel et al., 1998; Claeysen et al., 1999). Corresponding isoforms have been reported in rats and mice, where the isoforms originally referred to as 5-HT_4S and 5-HT_4L correspond to human 5-HT_4(a) and to 5-HT_4(b), respectively (Claeysen et al., 1996, 1997). Although the data currently available suggest a heterogeneous distribution of mRNA for the different isoforms in brain regions and peripheral tissues, none are purely central or peripheral, apart from the 5-HT_4(d), which is limited to the intestine, and no brain region expresses only a single isoform (Gerald et al., 1995; Ullmer et al., 1996; Blondel et al., 1998; Claeysen et al., 1999; Bender et al., 2000; Medhurst et al., 2001).

Although there is increasing evidence for pharmacological differences between isoforms, it is interesting that this is manifested not so much in terms of affinity, as few compounds appear to discriminate between the isoforms in binding, but in terms of compound efficacy in stimulating coupling to adenylate cyclase studies (Blondel et al., 1998; Bender et al., 2000; Mialet et al., 2000a,b; Pindon et al., 2002). The data presented here support this: SL65.0155 did not differ markedly in its affinity for different isoforms of the human 5-HT₄ receptor (when measured under similar conditions) but showed marked variation in intrinsic activity. This ranged from inverse agonism at the 5-HT_4(a) variant to partial agonist activity at the 5-HT_4(b) and 5-HT_4(e) variant. Although these experiments evaluating the functional interaction of SL65.0155 with splice variants of the 5-HT₄ receptor were not carried out under identical conditions, they nevertheless demonstrate that under certain circumstances, SL65.0155 is capable of interacting with 5-HT₄ receptors as either a partial agonist or an antagonist.

It thus seems likely that SL65.0155 could have sufficient agonist activity at one of the 5-HT₄ receptor splice variants to elicit an improvement in cognition but insufficient agonist activity at any isoform to elicit unwanted side effects, particularly in peripheral tissues. Whether this functional selectivity reflects a difference in the way SL65.0155 interacts with the various 5-HT₄ receptor isoforms or a difference in the local environment around central and peripheral 5-HT₄ receptors that affects efficiency of coupling remains to be determined. Whatever the mechanism, the data presented in this paper suggest that SL65.0155 can be described as a functionally selective partial agonist for those 5-HT₄ receptor isoforms mediating improved cognition.

The improvements in learning and memory induced by SL65.0155 were observed in several tests assessing different areas of cognitive performance and against a variety of deficits. The object recognition task assesses recognition memory in young, nominally unimpaired rats, but even under these conditions, SL65.0155 improved performance at 24 h when control animals no longer discriminated between objects, i.e., they could no longer recognize one as familiar and one as novel. Reversal of a recognition deficit was also observed in the social recognition test, where SL65.0155 improved the ability of an adult rat to remember a juvenile rat presented 2 h earlier. In the linear maze, aged animals demonstrate a marked deficit in their ability to remember a sequence of left-right turns. Such a deficit might reflect an inability to remember sequences or lists but might also include an element of disorientation, features of cognitive decline in both age and Alzheimer's disease. It is also important to note that in this experiment, rats received 20 administrations of SL65.0155 over 10 days and that the cognitive improvement was maintained over this period. Finally, SL65.0155 was also effective against the cognitive deficit induced by the muscarinic acetylcholine receptor antagonist scopolamine in the Morris water maze task for spatial memory. Scopolamine induces a cognitive deficit in humans and has been used as a means of mimicking the hypocholinergic state of AD. These studies therefore suggest that SL65.0155 has a broad potential for treating a variety of cognitive deficits across a number of disease states.

One area of potential difficulty for the therapeutic application of 5-HT₄ compounds as cognition-enhancing agents is that of side effect liability (Eglen et al., 1995). In particular, compounds with agonist activity are likely to have unwanted gastric and cardiovascular effects. The predictive value of the assays we have used is attested by the fact that compounds like cisapride, which increase heart rate and gastrointestinal motility in humans, also induce these effects in these animal models. It is now well documented that cisapride prolongs the QT interval in humans and leads to Torsades de Pointes and abnormal ECG and modifications of the heart rate (see Tonini et al., 1999 for a review), and this has led to its withdrawal from the market in several countries. Side-by-side comparison with cisapride was not performed in our studies, since abundant literature data have documented the effects of cisapride in these animal models (see for instance Sugiyama and Hashimoto, 1998 or Ueki et al., 1999). In keeping with its antagonist profile as measured in the rat esophagus and at the 5-HT_4(a) isoform, SL65.0155 was devoid of 5-HT₄ agonist-like actions in two studies evaluating a wide range of cardiovascular (including ECG) parameters in rats and in dogs. Similarly, and in contrast with cisapride, which is known to have prokinetic activity in the rat (Yamano et al., 1997), SL65.0155 had no gastrointestinal prokinetic activity as measured by intestinal transit times in the rat. This absence of agonist-like activity at doses over 1000-fold higher than those active in cognitive tests demonstrates a wide safety margin and also reinforces the suggestion that SL65.0155 has some degree of functional selectivity for central nervous system 5-HT₄ receptors. It is also important to note that despite the moderately high affinity of SL65.0155 for ₁ adrenoceptors, there were none of the unwanted effects (particularly cardiovascular) normally associated with antagonists at this receptor, and in the rat vas deferens it only weakly antagonized (IC₅₀>1 µM) the effect of noradrenaline.

Despite the recent advances in our understanding of Alzheimer's disease, there is still a clear need for treatments that act via different mechanisms than those currently available. Such compounds not only offer the prospect of improved efficacy and reduced side effect liability but may also be useful as add-on therapies for patients already receiving cholinesterase inhibitors, to enhance the modest improvements in cognitive function reported with this class of compounds. This is particularly supported by our demonstration that SL65.0155 can act synergistically with rivastigmine. A similar finding was recently reported with the 5-HT₄ agonist RS 67333 and a galanthamine analog in a place and object recognition task (Lamirault and Simon, 2001).

In conclusion, the data presented here demonstrate that SL65.0155 has potent and robust cognitive-enhancing activity, which is mediated by an agonist activity at 5-HT₄ receptors. This is supported by the partial agonist activity of SL65.0155 at the b and e isoforms of 5-HT₄ receptors in vitro and by the ability of a 5-HT₄ antagonist to reverse the cognition-enhancing activity of SL65.0155. In contrast, SL65.0155 acts as an antagonist at 5-HT₄ receptors in peripheral tissues in vitro and at the cloned a isoform of 5-HT₄ receptors. This functional selectivity is almost certainly responsible for the absence of gastrointestinal and cardiovascular effects and suggests that SL65.0155 will have a low propensity for side effects in the clinic. We propose that SL65.0155 represents a promising new agent for the treatment of patients suffering from memory deficits or dementia.