Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

A large body of evidence suggests that nicotinic cholinergic neuronal systems play a role in cognitive deficits related to Alzheimer's disease (Newhouse et al., 1997). Post-mortem studies have revealed a substantial loss of nicotinic acetylcholine receptors (nAChRs) in cortical and basal forebrain regions of patients with Alzheimer's disease (Shimohama et al., 1986; Whitehouse et al., 1986; Schroder et al., 1991; Aubert et al., 1992), which may be partially responsible for some of the cognitive and behavioral deficits seen with this disorder. Activation of nAChRs by the prototypic cholinergic ion channel agonist nicotine has been reported to consistently improve cognitive functions in intact and dysfunctional animals, healthy volunteers, and patients with Alzheimer's disease (Levin, 1992; Rusted et al., 1995; Newhouse et al., 1997). Cognitive-enhancing properties of nicotine in patients with Alzheimer's disease include improved vigilance, attention, information processing, and potentially memory function (Newhouse et al., 1988; Jones et al., 1992). The overall beneficial effects of nicotine might be explained by its general ability to stimulate the release of several neurotransmitters, neuropeptides, amino acids, and growth factors in various brain regions (Decker and Brioni, 1997).

Nicotine is of limited application in humans, however, because of dose-limiting side effects (gastrointestinal, neuromuscular, cardiovascular), which preclude its use as a therapeutic agent (Benowitz, 1986). On the other hand, the diversity of nAChR subtypes makes it possible to develop novel synthetic ligands selective for specific neuronal nAChR subtypes, which may potentially have better safety and efficacy profiles compared with the nonselective nAChR agonist nicotine (Lindstrom, 1997).

(±)-4-{[2-(1-Methyl-2-pyrrolidinyl)ethyl]thio}phenol hydrochloride (SIB-1553A) is a newly developed nAChR ligand that shows agonist selectivity for 4 subunit-containing human neuronal nAChRs and, therefore, exhibits a different receptor profile compared with previously described nAChR ligands (Decker et al., 1994, 1997; Lippiello et al., 1996; Bencherif et al., 1998; Kaiser et al., 1998). In vivo, SIB-1553A stimulates the release of acetylcholine from both the prefrontal cortex and hippocampus of young adult rats (Menzaghi et al., 1998). SIB-1553A also enhances cognitive performance in rodents with cholinergic dysfunction (Menzaghi et al., 1997) coupled with an improved safety profile relative to nicotine (Menzaghi et al., 1998). This pharmacological profile led us to hypothesize a potential usefulness of SIB-1553A for the symptomatic treatment of cognitive disorders, especially those of the Alzheimer's type where cognitive deficits have been predominantly attributed to cholinergic dysfunction of the basal forebrain (Palmer, 1996). Yet, Alzheimer's disease is not solely characterized by cholinergic dysfunction, with recent evidence demonstrating specific depletion of neurotransmitters such as serotonin, norepinephrine, and dopamine (Palmer, 1996). SIB-1553A has also been shown to stimulate the release of dopamine and norepinephrine from the hippocampus and frontal cortex in rats (Menzaghi et al., 1998). This suggests that SIB-1553A may affect multiple neuronal systems, which may be beneficial for the symptomatic treatment of Alzheimer's disease (Mohr et al., 1994).

The purpose of the present study was to characterize the potential cognitive-enhancing properties of SIB-1553A in aged rodents and nonhuman primates and to determine the effects of SIB-1553A on the release of hippocampal acetylcholine after acute and repeated administration. Cognitive deficits in aged animals are related to multiple neurotransmitter dysfunctions. Although aged animals cannot be considered as a model of Alzheimer's disease, they can provide a physiological model with which to mimic a subset of the memory deficits observed in patients with Alzheimer's disease, especially those associated with the early and middle stages of the disease. Indeed, Alzheimer's disease is characterized by a progressive dysfunction of several forms of memory (Zec, 1993; Kesner and Ragozzino, 1997). Three stages of Alzheimer's disease can be grossly defined: 1) an early stage characterized by an inability to remember novel information where the initial encoding of that information is affected but not the retrieval of previously learned information; 2) a middle stage where deficits in semantic memory, working memory, spatial orientation, and attention are especially observed; and 3) a late stage (or dementia) characterized by a global intellectual breakdown associated with changes in personality and behavior. In the present experiments, the effects of SIB-1553A on different forms of memory were explored in aged animals. Spatial and nonspatial working and reference memory were measured in various behavioral paradigms in mice, rats, and rhesus monkeys (Macaca mulatta) following acute and repeated administration of SIB-1553A.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals

Young (2-2.5-month-old) and aged (24-26-month-old) male C57BL/6 mice were obtained from the National Institute of Aging (Bethesda, MD). They were housed individually and maintained in a humidity (50-55%)- and temperature (21-23°C)-controlled room on a 12-h light/dark cycle (lights on at 6:30 AM). Mice were given a minimum of 1-week acclimation period before testing and their food was restricted to maintain body weight at 85% of their ad libitum weight throughout the duration of the training process.

Young adult (3-4-month-old) and aged (22-23-month-old) male Fisher 344/NHsd rats were purchased from the National Institute of Aging. Animals were maintained four per cage in an animal room under conditions identical to those described above. Food and water were available ad libitum.

Aged female rhesus monkeys (20-23 years old) were individually housed at the Animal Behavior Center of the Medical College of Georgia in stainless steel cages (127- × 71- × 66-cm units) and kept on a 12-h light/dark cycle. Each subject had been used in previous experiments assessing the effects of potential memory-enhancing agents. Because most of these compounds are proprietary, they cannot be listed. However, all of them have been shown to produce reversible pharmacological effects and none of them produced untoward effects in these animals. Furthermore, the monkeys were administered with ascending doses of these compounds for a short period of time only (<3 weeks). Compounds were not administered more than twice a week. No repeated dose studies were performed. Overall, these monkeys were tested on a weekly basis (baseline or drug test) for a 2-year period. A 1-month washout period and reestablishment of baseline performance were allowed between each drug test and prior to the onset of the present study. Each animal performed the delayed matching to sample (DMTS) task at least 5 days per week for 1 year before the initiation of the first drug study. Monkeys were allowed ad libitum access to water and were maintained on a feeding schedule that allowed approximately 15% of their normal daily food intake to be derived from 300-mg banana-flavored food pellets (P. J. Noyes Company, Inc., Lancaster, NH), which served as rewards during testing sessions. Standard laboratory monkey chow, fresh fruits, and vegetables comprised the remainder of their daily food intake after completion of a test session. Between testing sessions, monkeys were given toys and were exposed to television programs to provide environmental stimulation.

Apparatus and Behavioral Procedures

Behavioral testing was conducted during the light portion of the day (between 9:00 AM and 5 :00 PM) according to protocols approved by the institutional Animal Care and Use Committee and National Institutes of Health Guidelines.

Working Memory Tasks. Delayed nonmatching to place procedure in mice. Aged mice were trained in a semiautomated elevated (88 cm above the floor) eight-arm radial maze based on that described by Olton et al. (1979). The maze was made of black metal and consisted of an octagonal-shaped central platform (36 cm in diameter) from which eight arms (70 cm in length × 10 cm in width) radiated in a symmetrical manner (Lafayette Instrument, Indiana, IN). A circular food pellet cup was located at the end of each arm. Individually controlled, clear Plexiglas vertical doors were at the entrance of each arm (WCB-ARAM 1000; West Coast Biotech, La Mesa, CA). The maze was located in the center of a dimmed room (about 100 lux) with various pictures and objects placed around the room to serve as spatial cues. A remote control box and a closed circuit video system allowed the experimenter to activate the doors of the maze and to observe the behavior of each animal from an adjacent room.

Delayed matching to sample procedure in monkeys. On testing day, computer-automated test panels were attached to the monkey home cages. Stimuli on the test panels were pushable 2.54-cm-diameter clear plastic disks that could be illuminated from behind via red, yellow, or green light-emitting diodes. A trial began with the illumination of the "sample disk" with one of the three colors. A press on the sample disk extinguished the light and initiated one of four preprogrammed delay intervals, during which no disks were illuminated. Following the delay interval, two choice disks located below the sample disk were illuminated. One of the choice disks matched the hue of the sample disk, whereas the nonmatching choice was one of the other two colors. Both choice disks remained illuminated until the monkey pressed one of the two. A 300-mg banana-flavored pellet rewarded disk presses that matched the hue of the sample stimulus. Nonmatching choices were neither rewarded nor punished. Color patterns were fully counterbalanced for side, delay, and matching to sample. A new trial was initiated 5 s after a disk was pressed on the choice trial. Monkeys completed 96 trials on each daily session of testing. The average session length for the group was 25 min. Four delay intervals between a monkey's response to the sample disk and the presentation of the two choice-disks were used, including a 0-s, short, medium, and long delay. These delay intervals were individually adjusted for each animal to produce stable performance approximating the following levels of correct response: 75 to 84% (short), 65 to 74% (medium), and 55 to 65% (long). Performance for 0-s delay trials averaged 85% correct or greater. For the subjects involved in this study, the delay intervals ranged from 0 to 15 s. The rationale for this procedure was to normalize performance based on the widely varying capabilities of the monkeys.

Reference Memory Tasks. Acquisition and retention of a two-arm discrimination task in mice. The left-right arm discrimination task in mice is a rapid, positively motivated memory task that requires a mouse to discriminate between a baited and a nonbaited arm in a T-maze. Because the baited arm for a given animal is identical from trial to trial, this task involves reference memory. The T-maze was constructed of gray Plexiglas, with a main stem (70 × 10 × 20 cm) and two arms (30 × 10 × 20 cm). A food tray was placed at the end of each arm. Horizontal sliding doors separated the start box and arm entrance from the main stem. A halogen lamp positioned above the apparatus provided dimmed illumination (about 25 lux). Cues (e.g., posters) were placed above the arms to serve as spatial reference points.

Acquisition of escape to hidden platform procedure in rats. The apparatus consisted of a circular black tank (pool) (152 cm in diameter and 75 cm in height) filled with clear water (26 ± 1°C) to a depth of 40 cm. The pool was located in a dimly lit room surrounded by various extra maze cues, including experimenter, posters, shelves, and suspended toys. Rats were tested for their ability to escape from the water by climbing onto a submerged circular platform (11 cm in diameter) hidden 1 cm below water level. Performance of each animal (swim distance) was automatically monitored by a video tracking system (San Diego Instruments, San Diego, CA) consisting of a charge-coupled device video camera placed above the center of the pool and connected to a TV monitor and image analyzer coupled to a computer. The pool surface was divided into four quadrants of equal area (NW, NE, SW, and SE). Acquisition of escape onto the hidden platform was measured during two consecutive trials per day for 10 days. The platform was submerged in the center of the NW quadrant of the water maze and remained in the same location for all training trials. Each trial was initiated by placing the rat in the pool facing the wall at one of three start positions (center of the NE, SW, or SE quadrant). The trial ended when the rat climbed onto the escape platform. Start positions over days were determined randomly. If the rat did not find the platform within 90 s, it was gently guided toward it by the experimenter. The rat stayed on the platform for 15 s before starting the second trial. Following the second trial, the rat was towel-dried and placed under an infrared heat lamp for 15 min before being returned to its home cage. SIB-1553A or saline was administered s.c. 15 min prior to each daily session.

In Vivo Microdialysis

Some of the rats tested in the water maze continued to receive daily injections of SIB-1553A (4 mg/kg/day s.c.) to evaluate the level of hippocampal acetylcholine release induced by repeated administration of SIB-1553A using in vivo microdialysis. Five weeks after testing in the water maze, these rats along with rats repeatedly treated with saline were anesthetized with isoflurane (5%) and mounted in a Kopf stereotaxic apparatus with the incisor bar set at 3.3 mm below the interaural line. A hole was drilled in the skull and a 20-gauge stainless steel guide cannula (10.0 mm in length; Plastics One, Roanoke, VA) aimed at the dorsal hippocampus was implanted according to the following coordinates: 3.5 mm posterior to bregma, ±2.0 mm lateral, and 2 mm below the surface of the skull (Paxinos and Watson, 1986). The cannula was secured to the skull using three stainless steel crews and dental cement and was closed with a 24-gauge dummy cannula.

After a 7-day postsurgical recovery period during which animals continued to receive daily injections, the rats were briefly anesthetized with isoflurane and the dummy cannulae were removed. A microdialysis probe (ESA, Inc., Chelmsfold, MA; loop type with a rigid shaft and molecular weight cutoff of 6000; length 12 mm) was inserted into the guide cannula. Under these conditions, the microdialysis probe extended 2 mm beneath the guide cannula. The animal was placed in a plastic bowl with a harness around its neck (CMA 120; CMA Microdialysis, Acton, MA). The microdialysis probe was connected to a Hamilton syringe and perfused with artificial cerebrospinal fluid (145 mM NaCl, 2.7 mM KCl, 1.0 mM MgCl₂, and 1.2 mM CaCl₂, pH 7.4) containing 100 nM neostigmine at a rate of 1.0 µl/min. Twenty-minute fractions were collected and automatically injected into a column via a sample loop and an autoinjector. The on-line microdialysis was comprised of a CMA/100 microsyringe pump connected to a CMA/111 syringe selector.

The mobile phase [100 mM disodium hydrogen phosphate, 2.0 mM 1-octane sulfonic acid sodium salt, 0.005% reagent MB (ESA, Inc., pH 8.0 with phosphoric acid)] was pumped using a model 580 ESA pump through a polymeric reverse phase column (ACH-3; ESA, Inc.). The effluent from the column was passed through an enzyme reactor containing immobilized acetylcholinesterase and choline oxidase (ACH-SPR; ESA, Inc.). The high-performance liquid chromatography column and the enzyme reactor were placed in a housing with a constant temperature of 35°C. Acetylcholine and choline in microdialysis samples were converted into hydrogen peroxide, which was detected by amperometric oxidation in a ESA model 5041 analytical cell containing a glassy carbon target electrode and a palladium reference electrode. The oxidation potential was 250 mV and the signal was detected by an ESA model 5200 A Coulochem detector. The retention times for choline and acetylcholine under these conditions were 4 and 6 min, respectively. The limit of detection for acetylcholine was less than 20 fmol.

On the day of the experiment, 10 to 12 fractions were collected to establish the baseline level of acetylcholine release. Rats were then injected s.c. with 4 mg/kg SIB-1553A or saline and samples were collected until the levels of acetylcholine in the dialysate samples returned to baseline (2-3 h). Rats were tested 6 weeks after daily administration of saline or SIB-1553A.

Compound

SIB-1553A was synthesized at SIBIA Neurosciences, Inc. (now Merck Research Laboratories) as per methods previously described (Vernier et al., 1999). The compound was dissolved in 0.9% sterile saline. Rats and mice were injected s.c. with the test compound in volumes of 1 ml/kg and 10 ml/kg of body weight, respectively. Monkeys received drug treatment either i.m. (0.035 ml/kg of body weight in the gastrocnemius muscle) or p.o. (drug dissolved and pipetted onto a sugar cube). The salt form of the drug was used and doses were prepared immediately prior to injection.

Statistical Analysis

Data were analyzed by Student's t test or analysis of variance (ANOVA) using one- or two-factor analysis with repeated measures when applicable, followed by post hoc comparisons using Dunnett's or Newman-Keuls tests (SigmaStat Software; Jandel, San Rafael, CA). Percentage data followed a normal distribution and therefore were not normalized prior to statistical analysis. Values of p < 0.05 were considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effects on Working Memory

Delayed Nonmatching to Place Procedure in Mice. Aged mice required a minimum of 16 sessions to reach the criterion of at least 70% correct responses at 0-s delay, as opposed to young mice, which only required 10 sessions to acquire the same DNMTP rule (data not shown). As illustrated in Fig. 1, choice accuracy of young animals was delay-dependent (0 versus 180 s) and their performance at both delays was superior to the performance of aged animals [young versus aged at 0-s delay: t(20) = 2.98, p = 0.007; 180-s delay: t(20) = 2.61, p = 0.02]. At the 180-s delay, the performance of aged saline-treated mice was at chance level (53.4 ± 1.7%) and was significantly lower than their performance at the 0-s delay [72.6 ± 2.9%, t(12) = 7.40, p < 0.001]. A one-way ANOVA with repeated measures revealed a significant dose-dependent effect of SIB-1553A at the 180-s delay [F(6,51) = 3.73, p < 0.01], with the doses of 0.5 and 2.5 mg/kg (equivalent to 1.825 and 9.125 µmol/kg) significantly increasing the percentage of correct choices in aged mice. A maximal level of performance was observed at a dose of 2.5 mg/kg. SIB-1553A did not affect performance at the 0-s delay [F(6,51) = 0.8; N.S.].

View larger version (30K): [in this window] [in a new window]   Fig. 1.   Effects of SIB-1553A on DNMTP performance in aged mice as measured in the eight-arm radial maze. Mice were tested 20 min after s.c. administration of SIB-1553A or saline. Ascending dose responses were established for each animal (i.e., the same animal was repeatedly treated with saline and ascending doses of SIB-1553A). Drug testing was conducted for 4 days followed by 2 days of washout and then a baseline day. On baseline days, animals were injected with saline. Data are presented as percentage of correct responses (mean ± S.E.M). **p < 0.01, *p < 0.05 young versus aged saline-treated group, Student's t test; #p < 0.001 versus respective performance at 0-s delay, paired t test; *p < 0.05 versus aged saline (0)-treated mice at 180-s delay, one-way ANOVA with repeated measures followed by Dunnett's test (n = 9-13/group).

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Repeated best dose effect of SIB-1553A on DNMTP in naive aged mice as measured in the eight-arm radial maze. A, effect of SIB-1553A as a function of delay interval. Mice were injected for three consecutive days with saline or the selected best dose (i.e., 2.5 mg/kg). The best dose was selected based on the dose eliciting the highest level of performance in the ascending dose-response study. The animals were tested 20 min after s.c. administration. Data are presented as average of percentage of correct response over 3 days of testing (mean ± S.E.M, n = 7/group). A two-way repeated ANOVA revealed that SIB-1553A produced an overall main effect on performances. B, effect of SIB-1553A on DNMTP performance at 0-s delay interval. All the mice received a saline administration on baseline day. On subsequent days (day 1-3), the mice were allocated to separate group and were either administered with SIB-1553A or saline, 20 min prior testing. Data are presented as percentage of correct response at 0-s delay interval (mean ± S.E.M). *p < 0.05 versus saline-baseline, paired t test; #p < 0.05 versus saline-day 1, Student's t test.

Delayed Matching to Sample in Aged Monkeys. The performance of the aged monkeys after administration of saline (vehicle) averaged 85 ± 2, 75 ± 2.2, 68 ± 2.2, and 59 ± 1.2% correct responses, respectively, for zero, short, medium, and long delay intervals. Saline administration had no significant effect on DMTS performance (data not shown). As shown in Fig. 3A, SIB-1553A (i.m.) produced an inverted U-shaped dose-response effect. SIB-1553A, at a dose of 20 µg/kg, improved performance overall, independent of delay interval [F(8,278) = 4.78, p < 0.001]. In contrast, the highest dose (100 µg/kg) significantly decreased performance across delays (p < 0.02).

View larger version (34K): [in this window] [in a new window]   Fig. 3.   Effect of SIB-1553A as a function of dose and delay interval on DMTS performance in aged rhesus monkeys. The top and bottom panels represent the responses after i.m. (10 min before testing) and p.o. (30 min before testing) administration, respectively. Ascending dose responses were established for each animal. A drug washout period of 2 days was maintained between administration of drug doses (i.e., testing days with no injection). Data are presented as percentage of baseline (vehicle) levels of performance (horizontal line at 0%) (mean ± S.E.M) (n = 5/group).

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Effect of SIB-1553A as a function of individualized best dose on DMTS performance in aged rhesus monkeys. The best dose was selected based on the dose eliciting the highest level of performance in the ascending dose response study for each monkey. Ten minutes and 30 min refer to the time interval between drug administration and initiating testing; 24 h refers to those sessions conducted the day after drug administration. Data are presented as percentage of baseline (vehicle) levels of performance (horizontal line at 0%) (mean ± S.E.M) (n = 5/group).

Effects on Reference Memory

Acquisition and Retention of Two-Arm Discrimination Task in Mice. At the doses tested, SIB-1553A had no significant effects on reference memory performance of aged mice in the left-right arm discrimination task during both the acquisition [F(5,38) = 0.93, N.S.; Fig. 5] and retention [F(5,38) = 0.62, N.S.; Fig. 5] phases of the task.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Effects of SIB-1553A on acquisition and retention of a two-arm discrimination task as measured in a T-maze in aged mice. Performance was assessed by the number of trials needed to reach a criterion of five consecutive correct choices (e.g., baited arm). The retention tests was performed 24 h after the acquisition test. SIB-1553A and saline were administered s.c. 20 min prior to the acquisition and retention sessions. Data are presented as number of trials to criterion (mean ± S.E.M, n = 7-8/group). No significant differences were observed.

Acquisition and Retention of Escape to Hidden Platform in Rats. On the first 2 days of acquisition, young and aged rats did not differ significantly in their swim distance (Fig. 6A). Over the 10 days of acquisition training (10 sessions = 20 trials), both groups improved performance as indicated by a significant decrease in swim distance to the submerged platform [main session effects: F(9,153) = 5.94, p < 0.001]. The aged rats, however, appeared to have more difficulty in locating the escape platform, as shown by a main age effect and interaction with training sessions [main age effect: F(1,17) = 47.43, p < 0.001; interaction age × session F(9,153) = 3.28, p = 0.001]. The retention test (probe trial) indicated that aged rats swam significantly less in the target quadrant than young rats [t(17) = 6.86, p < 0.001; Fig. 6B]. Furthermore, aged rats swam significantly further from the target than did the young rats (data not shown), indicating that although they found the platform during acquisition, they did not learn the accurate spatial location of the platform.

View larger version (32K): [in this window] [in a new window]   Fig. 6.   Effects of repeated administration of SIB-1553A on acquisition (A) and retention (B) of escape to hidden platform procedure. Rats received daily s.c. injection of SIB-1553A 15 min prior to exposure to a Morris water maze. A, swim distance or length of the path that the animals swam to find the hidden platform in the water maze during each acquisition session. Data are presented as swim distance to platform averaged over two trials per session (one session per day) (mean ± S.E.M., n = 8-10/group). S.E.M. values and statistical significance were not shown for the clarity of the graphs. B, performance in the retention test. Retention test was performed 24 h after the 10th day of acquisition. The platform was removed from the pool and retention of the location of former platform position was measured by the percentage of swim distance in each quadrant of the pool (e.g., target, adjacent left or right, and opposite). The target quadrant represents the quadrant where the platform used to be located. Data are presented as mean ± S.E.M. (n = 8-10/group). *p < 0.05 versus respective target quadrant; #p < 0.05 versus aged (0)-target quadrant, two-way ANOVA with repeated measures followed by Newman-Keuls test. ***p < 0.001 versus young-target quadrant, Student's t test.

Microdialysis Studies

The location of microdialysis probe tips in dorsal hippocampus was carefully verified and no animals were discarded from the study (Fig. 7A). Under the conditions used, the average basal level of acetylcholine in the hippocampal dialysates was 104.96 ± 6.96 fmol (average of four samples collected every 20 min) in rats repeatedly injected with saline. Repeated administration of SIB-1553A for 6 weeks did not increase or decrease the baseline level of hippocampal acetylcholine (baseline average of 103.00 ± 3.72 fmol). On the day of the microdialysis test, subcutaneous administration of SIB-1553A (4 mg/kg) in rats chronically treated with saline evoked a statistically significant increase in the levels of acetylcholine in hippocampal dialysates compared with saline control [main group effect: F(1,9) = 6.67, p = 0.03]. The peak increase was observed at 40 min postinjection and the increase in acetylcholine levels persisted for 80 min (Fig. 7B). Similarly, subcutaneous administration of SIB-1553A (4 mg/kg) in rats chronically treated with SIB-1553A induced a statistically significant increase in the levels of acetylcholine in hippocampal dialysates compared with saline control [main group effect: F(1,13) = 4.94, p = 0.04]. The peak increase was observed at 40 min postinjection and the increase in acetylcholine levels persisted for at least 80 min (Fig. 7B).

View larger version (30K): [in this window] [in a new window]   Fig. 7.   Effects of SIB-1553A on acetylcholine release in the hippocampus as measured by microdialysis. A, graphic representation of the microdialysis probe placement (solid lines) in the rat dorsal hippocampus. Probe placements are plotted on the same side of the brain (coronal drawings adapted from the atlas of Paxinos and Watson, 1986). B, effects of repeated administration of SIB-1553A on acetylcholine levels in hippocampal dialysates. SIB-1553A (4 mg/kg) was administered s.c. in rats that received daily injection of saline (n = 6) or SIB-1553A (4 mg/kg/day) (n = 10) for 6 weeks prior to testing. Data are compared with rats that received saline only for the same period of time (n = 5). Data are presented as mean ± S.E.M. *p < 0.05 versus chronic saline + saline-treated group, two-way ANOVA with repeated measures followed by Newman-Keuls test.

No significant differences were observed between the level of acetylcholine measured in rats chronically treated with SIB-1553A and the level of acetylcholine measured in rats chronically treated with saline and subsequently challenged with an acute dose of SIB-1553A [F(1,14] = 0.30, N.S.].

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study examined the putative cognitive-enhancing properties of SIB-1553A, a novel subtype-selective nAChR ligand, in aged animals. SIB-1553A significantly improved cognitive performances in both aged rodents and monkeys tested in a variety of paradigms. The effects of parenteral and oral administration of SIB-1553A were observed after both acute and daily repeated administration. SIB-1553A markedly improved performances in spatial and nonspatial working memory tests with a less pronounced effect on spatial reference memory tests across species. Overall, it appears that SIB-1553A may predominantly affect attentional processes rather than producing true memory-enhancing effects.

SIB-1553A improved performances of aged mice on spatial working memory test as measured by the DNMTP task. In the ascending dose-response study, performances were improved after the long delay interval of 180 s but not after the 0-s delay, suggesting that SIB-1553A may facilitate the temporary storage of a novel information and its retrieval. However, the lack of effect of SIB-1553A at the 0 delay may be attributable to a ceiling effect inherent to the DNMTP procedure. Aged mice were initially trained to 70% correct responses at 0-s delay, thus making it difficult to observe further improvements. The repeated best dose study confirmed the facilitative effects of SIB-1553A. SIB-1553A enhanced performances in a delay-independent manner over 3 days of repeated administration. The DNMTP procedure requires the animal to first learn a rule (i.e., not to return to a previously visited arm) and then to express the memory of a visited arm according to this rule. Consequently, SIB-1553A may have improved not only the working memory capability (i.e., temporary storage of information) per se but also the ability of processing the reference memory components of the DNMTP task (i.e., utilization of preestablished spatial representations), which are required for successful execution of the task. Interestingly, in the repeated best dose protocol, performance of aged saline-treated mice at 0-s delay was reduced on the first day of testing compared with baseline levels, whereas the performance of SIB-1553A-treated mice remained unchanged. In this protocol, the first day of testing corresponded to the first day of the exposure of animals to various delay intervals. Thus, while mice had mastered the DNMTP rule, they had no previous experience with delays greater than 0 s. This novel situation may have "disturbed" the aged saline-treated mice, without affecting the SIB-1553A-treated aged mice. This result suggests that SIB-1553A may have facilitated the ability of aged animals to use flexible representations of previously acquired information in novel situations, an ability that is critically dependent on the hippocampal formation (Eichenbaum, 2000). Alternatively, the introduction of novel delays in the testing procedure may have distracted the animals, thus suggesting that SIB-1553A could also enhance attentional processes.

In agreement with the results of the rodent studies, SIB-1553A improved performances of aged rhesus monkeys in a nonspatial working memory task, the DMTS task. Performance improvement after i.m. administration of SIB-1553A was distributed across all delays. This effect is somewhat different from that of nicotine, which exhibits clear delay dependence in its response (Buccafusco et al., 1999). In fact, nicotine was not found to be particularly effective in improving zero or short delay interval-related performance. The ability of SIB-1553A to improve performance on short delay trials might also indicate a potential for improving attention (Prendergast et al., 1998). The results from the p.o. study were somewhat more variable than were those from the i.m. study. This result could possibly be due to variability in absorption and the subsequent pharmacodynamics of the drug when given via the oral route. Nevertheless, the overall improving effect achieved via either route was comparable. SIB-1553A was associated with average improvements in performance that ranged from 15 to 30% of baseline levels across delay intervals. Although there was some trend toward improved levels of performance on the day after i.m. administration, this was clearly not the case for oral administration and none of the day-after effects were statistically significant. Taken together, these data are in agreement with the rodent data, suggesting that the effect of SIB-1553A lasted less than 24 h in these animal models of working memory, an effect that can be explained by SIB-1553A's biological half-life. This effect contrasts, however, with that of nicotine showing the maintenance of an improved level of DMTS performance in monkeys 24 h after nicotine administration (Buccafusco and Jackson, 1991), in spite of comparable half-life and bioavailability. Although there are no clear explanations for these differences, this could be related to differences in receptor subtype selectivity profile. The lack of effect of SIB-1553A on response latencies is consistent with that of other nicotinic compounds and indicates that the drug is not simply acting as a general stimulant.

Overall, the results of the present study indicate that SIB-1553A was less effective in improving performances in spatial reference memory tests compared with performances in spatial working memory tests in rodents. In the T-maze discrimination task, no significant effects of SIB-1553A on acquisition or retention were observed in aged mice. In the water maze, SIB-1553A did not affect the acquisition of escape to a hidden platform, whereas a moderate, but significant, improvement was detected in the probe trial, suggesting a potential effect on retrieval of previously stored information. Alternatively, SIB-1553A-treated animals may have used a different behavioral strategy (not necessarily mnemonic) to locate the hidden platform during the initial acquisition of the task that would subsequently facilitate, at the time of the probe trial, a more focused search in the direction of the former location of the platform. The preferential effect of SIB-1553A on working memory tasks relative to reference memory tasks is in agreement with previous studies of nAChR ligands (Hodges et al., 1991; Jones et al., 1992; Levin, 1992), which suggest a predominant effect of this class of compounds on attentional processes. Indeed, working memory paradigms require the identification and use of novel information on each trial and therefore rely more heavily on attentional processes than do reference memory paradigms, which require the use of the same information across trials. This suggests that compounds like SIB-1553A may have predominant potential therapeutic benefit for a specific class of cognitive symptoms.

In the present study, SIB-1553A was able to induce the release of acetylcholine in the hippocampus of aged rats even after 6 weeks of daily administration. SIB-1553A was equally efficacious at releasing acetylcholine when administered acutely or repeatedly. These data indicate that rats chronically treated with SIB-1553A exhibited neither tolerance nor sensitization to the effect of the compound on acetylcholine release. Previous studies have demonstrated an age-dependent reduction in the expression of nAChRs such as the 4 subtype in the rodent brain (Rogers et al., 1998). Interestingly, SIB-1553A-induced acetylcholine release in the hippocampus was comparable in aged and young adult rats (Menzaghi et al., 1998), suggesting that the activation of the remaining and/or activation of intact populations of nAChRs by SIB-1553A in aged rats was sufficient for the induction of acetylcholine release in the hippocampus.

Cholinergic dysfunction is known to be responsible, at least in part, for some of the cognitive impairments observed in aged experimental animals (Gallagher and Colombo, 1995). Thus, although not directly proven, it is possible that SIB-1553A-induced cognitive improvement is partially, but not exclusively, related to an increase in hippocampal cholinergic function. However, it should not be excluded that SIB-1553A may produce other effects at other brain sites, which could also account for its behavioral actions. For instance, the effect of SIB-1553A on attentional processes could be mediated via a modulation of cholinergic neurons from the basal forebrain (Baxter and Chiba, 1999; Sarter and Bruno, 2000). In addition, the pharmacological profile of SIB-1553A demonstrates that the compound is not only a potent releaser of acetylcholine but also that it possesses the ability to stimulate the release of other neurotransmitters in various brain regions, including dopamine in the frontal cortex (Menzaghi et al., 1998), a brain region where the activation of dopamine receptors is critical for the cognitive process of working memory (Sawaguchi and Goldman-Rakic, 1991). The cortical release of norepinephrine is also likely to play an important role in the SIB-1553A-mediated effects on both attentional processes and working memory function (Clark et al., 1987). While the noradrenergic neurotransmitter system has been shown to be involved in processes of selective attention (Robbins, 1997) and attentional orientation (Coull, 1998), the dopaminergic system has been proposed to mediate more executive aspects of attention such as attentional set-shifting or working memory (Coull, 1998).

In conclusion, SIB-1553A exhibits significant cognitive-enhancing properties in aged rodents and nonhuman primates. It is hypothesized that SIB-1553A exerts greater effects on attentional processes than memory functions per se. The specificity of the cognitive-enhancement of novel agents like SIB-1553A is important to characterize because this may determine the potential therapeutic application. Altogether, these data provide additional support for the use of subtype-selective nAChR ligands as a potential therapy for the symptomatic treatment of certain cognitive deficits associated with aging and neurological diseases.