Lab.pharm.uic.edu

Neuropsychopharmacology (2007) 32, 505–513& 2007 Nature Publishing Group Cognitive Deficits in Rats after Forebrain CholinergicDepletion are Reversed by a Novel NO Mimetic Nitrate Ester  Brian M Bennett*,1, James N Reynolds1, Glen T Prusky2, Robert M Douglas3, Robert J Sutherland2 and Department of Pharmacology & Toxicology, Queen’s University, Kingston, ON, Canada; 2Canadian Centre for Behavioural Neuroscience, The University of Lethbridge, Lethbridge, AB, Canada; 3Department of Ophthalmology, University of British Columbia, Vancouver, BC, Canada; Department of Medicinal Chemistry & Pharmacognosy, College of Pharmacy, University of Illinois at Chicago, Chicago, IL, USA Many conditions adversely affecting learning, memory, and cognition are associated with reductions in forebrain acetylcholine (ACh), most notably aging and Alzheimer’s disease. In the current study, we demonstrate that bilateral depletion of neocortical and hippocampal ACh in rats produces deficits in a spatial learning task and in a recently described, delayed visual matching-to-sample task. Oral administration of the novel nitrate, GT1061 (4-methyl-5-(2-nitroxyethyl) thiazole HCl), and the acetylcholinesterase inhibitor, donepezil, reversed the cognitive deficits in both memory tasks in a dose-dependent manner. GT1061 was superior in the delayed matching-to- sample task. GT1061 was absorbed rapidly after oral administration, crossed the blood brain barrier, and achieved brain concentrations that were slightly higher than those found in plasma. The activity of GT1061 was NO mimetic: soluble guanylyl cyclase (sGC) was activated, but selectivity was observed for sGC in the hippocampus relative to the vasculature; and hippocampal levels of phosphorylated ERK1/2, which is a postulated intermediary in the formation of long-term memory, were increased. The beneficial effect on visual and spatial memory task performance supports the concept that stimulating the NO/sGC/cGMP signal transduction system can provide new, effective treatments for cognitive disorders. This approach may be superior to that of current drugs that attempt only to salvage the residual function of damaged cholinergic neurons.
Neuropsychopharmacology (2007) 32, 505–513. doi:10.1038/sj.npp.1301054; published online 8 March 2006 Keywords: nitric oxide; Alzheimer’s disease; ERK; cGMP; cognition; memory 2004), and memantine is not approved for mild to moderateforms of AD (Areosa et al, 2005), thus the imperative for Alzheimer’s disease (AD) is characterized by disruption of new therapeutic agents to treat AD and other dementias both excitatory amino acid and cholinergic neurotransmis- sion. Loss of cholinergic neurons, and subsequent deficits in Signal transduction by the NO/soluble guanylyl cyclase cholinergic neurotransmission in the hippocampus and (sGC)/cGMP system modulates synaptic transmission and cerebral cortex, is strongly correlated with clinical signs of plasticity in the CNS, and in vivo studies suggest an cognitive impairment and dementia in AD patients important role for NO/cGMP in the acquisition of new (Mesulam, 2004). The use of acetylcholinesterase inhibitor learning and memory (Bernabeu et al, 1997; Prickaerts et al, drugs, such as donepezil (Aricept), is based upon the 2002; Prickaerts et al, 1997). This is supported by numerous cholinergic hypothesis of memory loss and cognitive deficit in vitro studies indicating a role in long-term potentiation (Bartus et al, 1982). These drugs have represented the sole (LTP), clustering of presynaptic and postsynaptic proteins therapeutic response to AD until the recent introduction of (suggesting a role in synaptogenesis), and enhancement of the N-methyl-D-aspartate (NMDA) receptor low affinity presynaptic neurotransmitter release. Aberrant NO signal- antagonist, memantine. The clinical efficacy of acetylcholi- ing in the brains of AD patients has also been reported nesterase inhibitors has been questioned (Courtney et al, (Bonkale et al, 1995). Thus, sGC activation and cGMPformation in the brain may be an effective strategy for *Correspondence: Dr BM Bennett, Department of Pharmacology & mitigating the cognitive dysfunction that occurs as a Toxicology, Queen’s University, Kingston, ON, Canada K7L 3N6, consequence of cholinergic deficits in the CNS. In contrast Tel: + 613 533 6473, Fax: + 613 533 6412, to the acetylcholinesterase inhibitors that attempt to E-mails: brian.bennett@queensu.ca; thatcher@uic.edu augment the residual function of a degenerating cholinergic Received 27 December 2005; revised 17 January 2006; accepted 18 system, NO mimetics are postulated to bypass this system January 2006Online publication: 30 January 2006 at http://www.acnp.org/citations/ and modulate the function of signaling pathways down- stream from cholinergic receptor activation.
Most recently, two studies have directly implicated a role MWT testing. In the moving hidden platform version of the for dysfunction of NO/cGMP signaling in AD. In hippo- MWT, the platform was moved to a new location within the campal slices, disruption of LTP by amyloid-b peptide (Ab) pool every other day according to a random sequence of was shown to result from a malfunction of NO/cGMP locations. Thus, for two consecutive days the platform signaling (Puzzo et al, 2005). Further, Snyder et al (2005) remained in the same location. On the third day, the reported that Ab produced a depression of NMDA-evoked platform was moved to a new location for another pair of currents via endocytosis of NMDA receptors, mediated via days, and this was continued until six locations had been dissociation of the synaptic protein, PSD95, from the completed. Rats received eight trials each day. Data were NMDA receptor. As both sGC and neuronal nitric oxide collected using a video camera and microcomputer auto- synthase (NOS) are known to be associated with PSD95 in mated system (HVS 2020 Plus system). On every trial, time the brain (Russwurm et al, 2001), this work suggests that Ab and distance to find the platform was automatically specifically disrupts hippocampal NO/cGMP signaling. It recorded and analyzed. Following recovery from surgery, was concluded that intervention with NO donors or drugs the rats were divided into 10 groups; sham operated vehicle enhancing NO/cGMP signaling might provide a novel control (n ¼ 10), lesion vehicle control (n ¼ 10), four approach to treatment of AD (Puzzo et al, 2005).
groups, each receiving one dose of donepezil (0.05, 0.5, Classical nitrates, of which glyceryl trinitrate (GTN) is the 0.1, and 1 mg/kg) (n ¼ 10), and four groups, each receiving archetype, are NO mimetics, in that they manifest many one dose of GT1061 (0.5, 1, 5, and 10 mg/kg) (n ¼ 10). Each aspects of the biological activity of NO in vivo (Bennett et al, rat received the appropriate drug by gavage 60 min before 1994; Thatcher et al, 2004). Although the hypotensive effect testing each day. Each day the rats were tested in the MWT.
of GTN itself precludes its use in CNS-directed therapy, we Following the last day of testing, the animals were perfused have previously reported a novel nitrate that demonstrates: with 4% paraformaldehyde and their brains removed, (a) attenuated hypotensive activity relative to GTN; (b) cryoprotected, and sectioned for acetylcholinesterase stain- neuroprotection in the middle cerebral artery occlusion model of ischemic stroke; and (c), reversal of scopolamine-induced cognition deficits in rats tested in the Morris water Delayed matching-to-sample task. The testing apparatus task (MWT) (Reynolds et al, 2002; Smith et al, 2000).
and DMTS training and testing procedures are detailed in a Prusky et al (2004) recently demonstrated long-delay previous publication (Prusky et al, 2004). Briefly, adult trial-unique visual matching-to-sample (DMTS) perfor- female Long–Evans rats (n ¼ 10) were first trained on a mance in rats, implicating essential roles for both the simple visual discrimination task, and then in four stages, hippocampus and perirhinal cortex. Thus, the hippocampus were gradually shifted to the trial-unique, one-trial learning is important not only for spatial memory, but also for visual task. Throughout the testing, rats were rewarded for recognition memory, a cognitive process that is disrupted in identifying and swimming toward a correct picture where AD (Lindeboom et al, 2002). In this paper, we use this they could escape from water to a submerged platform. The DMTS task to examine the reversal of a cholinergic deficit in black-and-white pictures were displayed on computer recognition memory by a novel nitrate, GT1061 (4-methyl- monitors to eliminate nonvisual cues. The pictures were 5-(2-nitroxyethyl) thiazole HCl), and compare both to drawn from a set of more than 100 and on any given day no effects on spatial working memory and to the effects of picture appeared on more than one trial, such that, relative donepezil in the same tasks. A potential mechanism of familiarity of pictures is irrelevant. Similarly, the probability action of this novel compound in the hippocampus also is of any specific picture being used as a sample or a foil was equal. In the final stage, during each trial, rats viewed apicture in a sample phase, and then in a choice phase,discriminated the sample picture from a novel picture with high accuracy in a different pool located in the same room.
All procedures for animal experimentation were undertaken The left/right position of the correct picture varied in accordance with the principles and guidelines of randomly to make the task nonspatial. The delay depen- dence of memory for the sample picture was measuredsystematically by varying the delay between the sample andchoice phases. All animals received approximately 6 weeks of presurgical training. At 1-week after surgery, DMTS Moving hidden platform Morris water task: surgery. In performance was remeasured. On the basis of the pre- these studies, 90 of 100 young adult female Long–Evans rats surgical memory performance we selected a 4-min interval received bilateral injections of 192 IgG-saporin into the for postsurgical testing of drug effects. Presurgical perfor- nucleus basalis magnocellularis and medial septum to mance at this delay interval between sample and choice was deplete the cholinergic innervation of the cerebral cortex well-above chance for each rat, but was less than maximum and hippocampus, respectively (Pizzo et al, 1999). Rats were accuracy (Figure 1). Thus, it afforded us the opportunity to anaesthetized with isoflurane and under stereotaxic gui- measure deficits, drug effects, and dose response, at a dance (coordinates: medial septum; AP + 0.6, L 1.2, D À7.5, moderate level of memory difficulty (thereby avoiding floor À6.8 from bregma; nucleus basalis; AP À1, L73, D À7.5), or ceiling effects). Performance in the DMTS task before 192 IgG-saporin (0.1 mg in 1.0 ml of phosphate-buffered and after surgery was compared at the 4-min interval. Drug saline, pH 7.4, over 15 s) was infused into each site. The or drug vehicle was administered by gavage 30-min before remaining 10 animals received sham surgery. All rats were testing. Each rat was tested with vehicle solution and three doses each of GT1061 and donepezil, thus, the experimental Nitrates and cognition enhancementBM Bennett et al parameters. To assess the distribution of GT1061 intothe CNS, rats were fasted overnight and then administered25 mmol/kg (5.6 mg/kg) GT1061 by gavage. Five minuteslater the animals were killed and a blood sample wasobtained. The brain was removed and one hippocampusdissected out and frozen in liquid nitrogen, procedureslasting approximately 5 min to complete. The hippocampuswas weighed, pulverized, and then both plasma andhippocampus samples assayed for GT1061 by gas chromato-graphy. To assess potential sex and strain differences in thepharmacokinetic properties of GT1061, female Long–Evansrats were instrumented as described above, and bloodlevels of GT1061 determined after oral administration of50 mmol/kg GT1061. The area under the plasma concentration– Mean preoperative visual matching-to-sample performance at time curve from 0 to 120 min was 82.1736.2 mmol-min/l each training delay interval. Choice accuracy was assessed in each animal (SEM, n ¼ 5) for female Long–Evans rats, and 125728.1 during 24 trials at each interval, and declined systematically with increasing mmol-min/l (SEM, n ¼ 6) for male Sprague–Dawley rats.
delays. Data are expressed as the mean7SEM (n ¼ 7).
These values were not significantly different (p ¼ 0.354,Student’s t-test for unpaired data).
design was fully ‘within subjects’. The doses of GT1061 were Hippocampal and aortic sGC activity. Cholinergic inner- 1, 5, and 10 mg/kg and for donepezil were 0.1, 0.5, and vation of the hippocampus and cerebral cortex of female 1.0 mg/kg. There was at least a 48-h time interval between Long–Evans rats was depleted by intraventricular micro- successive tests of drugs or vehicle. The order of testing was injection of 192-IgG saporin, as described above, and 2 always vehicle, GT1061 5, 10, and 1 mg/kg and then weeks later the aortae and hippocampi from sham-operated donepezil 0.5, 1.0, and 0.1 mg/kg. Between GT1061 and and lesioned animals were removed. The preparation of the donepezil testing we retested vehicle administration, and 100 000 g supernatant fractions of aorta and hippocampus, found that performance was not different from that and enzyme assay conditions for assessment of sGC activity observed with the first vehicle test (55.072.2% correct vs were as described (Bennett et al, 1992). Samples (0.2 ml) 57.371.9% correct, p ¼ 0.35, Student’s t-test for paired contained 2 mM L-cysteine and GT1061 (10 mM–10 mM).
data). On each day of testing, each rat received 12 memory Reactions were initiated by the addition of 44–79 mg trials and the percentage correct choices was calculated. In a hippocampal supernatant protein or 20–31 mg aortic super- separate experiment, we assessed whether GT1061 enhanced natant protein and terminated by the addition of 0.8 ml of cognitive performance in nonlesioned animals. Rats (n ¼ 4) 50 mM sodium acetate (pH 4.0) followed by heating at 901C were trained as described above and DMTS performance for 3 min. The cGMP was quantitated by radioimmunoassay evaluated after oral administration of 10 mg/kg GT1061.
and the protein content determined using bovine serum This was compared to DMTS performance assessed on the day before and the day subsequent to drug administration.
Phosphorylation of ERK1/2. To test for in vivo activation of Surgery involving DMTS rats. After initial training, the ERK1/2, separate groups of nonlesioned animals were cholinergic innervation of the hippocampus and cerebral administered GT1061 (1 mg/kg i.p. (4.5 mmol/kg)) or cortex was depleted by intraventricular microinjection of phosphate-buffered saline. At 60 min after drug adminis- 192-IgG saporin (1.8 mg in 5 ml of phosphate-buffered saline, tration, the brain of each animal was rapidly removed, pH 7.4, over 15 s) into the lateral ventricle of each cerebral dissected into regions (hippocampus, frontal cortex, and hemisphere (coordinates: AP À1.1, L 1.5, D À4.2 from cerebellum), and homogenized in lysis buffer containing 50 mM TRIS, pH 7.4, 50 mM sodium pyrophosphate, 1 mMEDTA, 1% Triton X-100, 1 mM dithiothreitol, 1 mM PMSF, Pharmacokinetic studies. Male Sprague–Dawley rats (250– 1 mM NaF, 1 mM NaVO4 and protease inhibitors (Roche 275 g) were fasted overnight, anaesthetized for surgery with Applied Science). Phosphorylated ERK1/2 (pERK) was a combination of Hypnorm/Versed and then catheters (30- determined by ELISA using a commercially available kit gauge Teflon fused to 0.02-inch inside diameter Tygons) (Assay Designs Inc., Ann Arbor, MI). In some experiments, were inserted into the abdominal aorta and abdominal vena rats were administered the NMDA receptor antagonist, MK- cava for blood sampling and drug administration, respec- 801 (1 mg/kg, i.p.), 20 min before GT1061. For in vitro brain tively. Animals were administered 50 mmol/kg (11.2 mg/kg) slice experiments, rats were anesthetized with halothane, GT1061 (prepared in saline) by the i.v., i.p., or oral routes decapitated, and the brain rapidly excised into ice-cold and blood samples (250 ml) obtained from the aortic sucrose-substituted Krebs’ buffer (342 mM sucrose, 4.8 mM catheter at various times. Blood was immediately centri- KCl, 1.2 mM KH2PO4, 1.3 mM CaCl2, 1.2 mM MgSO4, 25 mM fuged and 100 ml aliquots of plasma were assayed for NaHCO3, and 10 mM glucose, pH 7.4, saturated with 95% GT1061 by gas chromatography with electron capture O2/5% CO2). The hippocampus was isolated, 400 mm slices detection. The data were fitted to a 2-compartment open were cut with a McIlwain tissue chopper, and a small scalpel model using PKSolutions software (Summit Research blade was used to isolate the CA1 region of each slice.
Hippocampal mini-slices containing the CA1 region were equilibrated for 3 h in Krebs’ buffer (118 mM NaCl, 4.8 mMKCl, 1.2 mM KH2PO4, 1.3 mM CaCl2, 1.2 mM MgSO4, 25 mMNaHCO3, and 10 mM glucose, saturated with 95% O2/5%CO2) at room temperature, and then exposed to drug orvehicle for 30 min. At the end of the incubation, the sliceswere homogenized in lysis buffer, and the content of pERKdetermined by ELISA.
Synthesis of GT1061. GT1061, prepared under GMPconditions, was obtained from GB Therapeutics Ltd(Mississauga, ON, Canada). GT1061 was synthesized from4-methyl-5-(2-hydroxyethyl)-1,3-thiazole by adaptation ofstandard acidic nitration procedures, followed by saltformation with HCl and crystallization. Purity as assessedby chromatography and spectroscopy was greater than 98%.
GT1061 is a crystalline white solid: mp 107–1081C (d).
Acetylcholinesterase staining in rat brain after either bilateral Statistical analyses. In the MWT and the DMTS task, data injections of 192-IgG saporin into the nucleus basalis magnocellularis and were evaluated with a repeated measures ANOVA procedure medial septum (middle panels), or intraventricular injections of 192 IgG-saporin (bottom panels). Stained sections from vehicle control (top panels), (SPSS 11.0) as previously recommended (Kirk, 1982; Ringo, or 192 IgG-saporin-treated animals were analyzed using NIH image 1991). Post hoc tests were conducted using Fisher’s LSD densitometry. Animals treated with 192 IgG-saporin exhibited significantly method. Data from other experiments were analyzed by the less staining in both the cerebral cortex (t ¼ À12.9, po0.0001) and appropriate statistical test, as indicated. Differences were hippocampus (t ¼ À7.38, po0.0001) compared to sections from the considered to be statistically significant if po0.05.
Examination of the acetylcholinesterase stained sectionsusing NIH Image densitometry showed that after bilateralinjections of 192 IgG-saporin into the nucleus basalismagnocellularis and medial septum, animals exhibitedsignificantly less acetylcholinesterase staining in both theneocortex and hippocampus as compared to sectionsobtained from control animals (t ¼ À12.9, po0.0001, andt ¼ À7.38, po0.0001, respectively, Figure 2). Compared tothe focal injections of 192 IgG-saporin, we found a similardegree of depletion of acetylcholinesterase in the neocortexand hippocampus after intraventricular administration of192 IgG-saporin (both p’s40.6).
We conducted separate ANOVAs for donepezil-treated andGT1061-treated groups on latency to find the platform ondays when it was in novel locations. There were two withinsubject factors, day and trial. In the case of donepezil, theANOVA shows that the groups did differ significantlyfrom one another (F(5,46) ¼ 2.55, po0.04). The animalsimproved significantly over days (F(5,230) ¼ 6.07, po0.0001)and trials (F(7,322) ¼ 184.67, po0.0001). In the case ofGT1061, the ANOVA shows that the groups did differsignificantly from one another (F(5,48) ¼ 3.33, po0.01).
Effect of oral administration of GT1061 (0.5–10 mg/kg) and The animals improved significantly over days (F(5,240) ¼ donepezil (0.05–1 mg/kg) in the moving hidden platform Morris water task.
7.52, po0.0001) and trials (F(7,336) ¼ 90.3, po0.0001).
(a) The data represents the average time to find the hidden platform on In both cases, post hoc analysis shows that the sham control trial 2 on days when the platform was in novel locations. Data are group exhibited shorter latencies compared to all other expressed as the mean7SEM (n ¼ 10). The groups did not differsignificantly from one another (F(9,78) ¼ 1.66, p ¼ 0.11, repeated measures ANOVA). (b) The data represents the average time to find the hidden For further illustration of the main trend, Figure 3a shows platform on trial 2 on days when the platform was in the same location.
the average latency to find the hidden platform on trial 2 on Data are expressed as the mean7SEM (n ¼ 10). *po0.05 vs lesion, days when the platform was in novel locations. Previous repeated measures ANOVA (F(9,78) ¼ 2.09, p ¼ 0.04).
Nitrates and cognition enhancementBM Bennett et al work in our laboratory has shown that this trial provides agood index of new learning. Trial 1 performance was notdifferent between any of the groups and all of the groupsimproved by approximately 16 s between trial 1 and 2. AnANOVA on the trial 2 data revealed that the treatmentgroups were not different from one another (F(9,78) ¼ 1.66,p ¼ 0.11), indicating that in lesioned animals, neitherdonepezil nor GT1061 altered the animal’s capacity fornew learning.
In similar fashion, we analyzed the latency to find the platform on days when the hidden platform remainedin the same location as the previous day for groupsreceiving donepezil. The ANOVA shows that the groups didnot differ significantly from one another (F(5,46) ¼ 0.42, Effect of oral administration of GT1061 (1–10 mg/kg) and p ¼ 0.83). The groups did improve significantly over days donepezil (0.1–1 mg/kg) in the DMTS task, which measures recognition (F(5,230) ¼ 4.4, po0.0007), and trials (F(7,322) ¼ 59.4, memory performance. The decreased accuracy in the task after cholinergic po0.0001). For the groups receiving GT1061 there was a depletion was reversed at all doses of GT1061 and at higher doses ofdonepezil. Data are expressed as the mean7SEM (n ¼ 10). *po0.01 vs significant difference between the groups (F(5,48) ¼ 2.39, postsurgery, repeated measures ANOVA (F(7,63) ¼ 25.0, po0.001).
p ¼ 0.05). The groups did improve significantly over days(F(5,25) ¼ 3.62, po0.004). Again, for simplicity of illustra-tion, Figure 3b shows the group data for performance on doses of donepezil (p’so0.01). It should be noted that the trial 2 for all days when the platform was in the same oral bioavailability of donepezil approaches 100%, whereas location as the previous day. Earlier work has shown that that of GT1061 is about 20% (see below). In the study that this is a sensitive index of memory performance, which assessed whether GT1061 improved performance in nonle- should reveal effects from the previous day’s learning sioned animals, rats made 7772% (SEM) correct choices on and the first trial reminder. The trial 1 performance the trial day before GT1061, and 83.574.9% correct choices showed a similar trend to that of trial 2, but we selected on the day after. Oral administration of 10 mg/kg GT1061 trial 2 because of its lower within group variance and hence improved performance to 87.874.3% correct choices, but greater sensitivity. A repeated measures ANOVA examining the increase was not statistically significant (F(2,6) ¼ 3.4, the group differences on trial 2 only when the platform was in the same locations as the previous day showsthat the groups differed significantly from one another(F(9,78) ¼ 2.09, p ¼ 0.04). With donepezil, doses of 0.1 mg/ kg or higher produced significant benefit in cholinergic-depleted rats (p’so0.01). GT1061, at an oral dose of 10 mg/ GT1061 was rapidly absorbed after i.p. or oral administra- kg, produced the same beneficial effect on memory tion with peak plasma concentrations occurring 3 min after performance as did donepezil (p ¼ 0.54) and produced dosing (Figure 5). Calculation of the AUCs indicated an oral performance that was not different from sham lesion rats bioavailability of 20.5711.7%, (range of 9–41.3%), and an (p ¼ 0.60) but superior to depleted control rats (po0.002).
i.p. bioavailability of 52.8710.4% (range of 36.6 to 64.7%).
There were no significant differences in swim speeds After i.v. administration, values for Vd (l/kg), Cl (l/min/kg) associated with cholinergic depletions or either drug (all and elimination half-life (min) (derived from Vd and Cl) were 2.5, 0.079, and 21.6, respectively. For oral administra-tion the values were: 3.4, 0.072, and 32.6, and for i.p.
administration the values were: 2.2, 0.08, and 19.0.
Calculations for oral and i.p. Vd and Cl values used the The overall repeated measures ANOVA on percent correct derived bioavailability values for the fractional dose showed statistically significant differences between condi- absorbed. In the brain bioavailability study, the mean tions (F (7,63) ¼ 25.0, po0.001). Before surgery, rats made plasma level of GT1061 5 min after oral dosing was 75.0% correct choices at a 4-min delay between sample and 0.5270.25 nmol/ml (SD, n ¼ 6), whereas the mean tissue choice phases and this was significantly decreased to concentration of GT1061 in hippocampus (expressed as approximately 55.0% in the same animals after cholinergic nmol/g wet weight, roughly equivalent to nmol/ml assuming depletion (po0.0006). This deficit in the visual memory that 1 g of tissue has a volume of 1 ml) was 0.6770.42 nmol/ task was large and sustained throughout the experiment. All g wet weight (SD, n ¼ 6) (Figure 6). Plasma and hippocam- doses of GT1061 significantly improved performance.
pal levels of GT1061 were not significantly different from Indeed, the highest dose of GT1061 restored performance each other (Student’s t-test for paired data), and in every in the DMTS task to a level that was not significantly animal the brain levels of GT1061 were higher than the different from the presurgery level (Figure 4). The two plasma levels (Figure 6). The ratio of brain to blood GT1061 higher doses of donepezil improved performance, but the was 1.370.28, a value that may underestimate the true brain lowest dose (0.1 mg/kg) had no effect. In this assay, GT1061 penetration of GT1061, since during the time between was equipotent (on a per mg basis) with donepezil at killing and completion of the dissection of the hippocampus the 1 mg/kg dose, and at 10 mg/kg, the improvement in (ca. 5 min) further metabolic biotransformation of GT1061 performance was significantly greater than that of any of the in the brain tissue may have occurred.
Time course of plasma concentrations of GT1061 after i.v., i.p., and oral administration of 50 mmol/kg GT1061. Data are expressedas the mean7SEM (n ¼ 5–6). The oral bioavailability of GT1061 was Comparison of activation of sGC by GT1061 in the hippocampus and aorta after depletion of cholinergic innervation of thehippocampus and cerebral cortex by intraventricular microinjection of 192IgG-saporin. The basal levels of sGC activity in hippocampal homogenatesfrom control and lesioned animals were 7.771.8 and 8.472.2 pmol/min/mg protein, respectively, and in aortic homogenates from control andlesioned animals were 86.179.4 and 82.379.1 pmol/min/mg protein,respectively. Data are expressed as the fold increase in activity over basal(mean7SEM, n ¼ 7). Enzyme activation was not altered in homogenatesfrom lesioned animals. Enzyme activation was significantly greater (po0.05)in homogenates from hippocampus compared to aorta at concentrationsof GT1061 of 0.3 mM and above (one-way ANOVA ((F(3,27) ¼ 6.02–9.59,po0.05) and Newman–Keuls post hoc test).
lesioned animals was virtually identical, indicating thatcyclic GMP signaling remained intact in cholinergic-deficient animals. Enzyme activity in hippocampal homo-genates increased 16- to 20-fold at the highest concentration Plasma and hippocampal concentrations of GT1061 (25 mmol/ of GT1061, whereas in the aorta, the maximal increase in kg) 5 min after oral administration. The mean plasma concentration of activity was only 2- to 3-fold. In broken cell preparations, GT1061was 0.5270.25 nmol/ml (SD, n ¼ 6) and the mean tissue with very rare exceptions, nitrates do not activate sGC concentration of GT1061 in hippocampus was 0.6770.42 nmol/g wetweight (SD, n ¼ 6). These values were not significantly different to each above basal levels in the absence of certain added thiols other (Student’s t-test for paired data). The ratio of brain to plasma (Artz et al, 2001; Zavorin et al, 2001); GT1061 is not an GT1061 exhibited pharmacokinetic properties similar to classic nitrates such as GTN and isosorbide dinitrate.
Absorption was rapid, and bioavailability after oral In the in vivo assessment of phosphorylated ERK1/2 administration was somewhat variable (9–41%). The large (pERK), GT1061 increased the levels of pERK in the volume of distribution (2.2–3.4 l/kg, depending on route of cerebral cortex, but had no effect in the cerebellum administration) indicated that the drug was distributed (Figure 8a). In the hippocampus, pERK levels were extensively in tissues. This was clearly evident in the brain, increased by GT1061, but the increase did not reach where drug concentrations exceeded those in plasma. Extra- statistical significance. However, when GT1061 was admi- polation of the plasma levels of drug obtained in the nistered to animals that had been pretreated with the pharmacokinetic study to the doses of drug that demon- NMDA receptor antagonist, MK-801, the increase in strated efficacy in the behavioral tests suggest that hippocampal pERK caused by GT1061 was significant therapeutic concentrations of GT1061 are in the submi- (Figure 8a, inset), suggesting that in the hippocampus, the effects of GT1061 are more evident in circumstances whereinput from upstream effectors has been compromised.
Qualitatively, this is consistent with the lack of a significanteffect of GT1061 on cognition in unlesioned animals. We GT1061 caused a concentration-dependent increase in sGC also compared the effect of GT1061 to that of the cholinergic activity in both aortic and hippocampal homogenates muscarinic receptor agonist, carbachol, in hippocampal (Figure 7), although the degree of activation in the mini-slices containing the CA1 region (Figure 8b). Both hippocampus was markedly greater. Enzyme activation by drugs significantly increased the amount of pERK in this in GT1061 in hippocampal homogenates from control and Nitrates and cognition enhancementBM Bennett et al cognitive deficits for the purpose of identifying potentialnew therapeutic compounds for the treatment of AD.
It is well established that dysfunction in the hippocampus causes spatial learning deficits in tasks such as the MWT(Redish and Touretzky, 1998; Silva et al, 1998a) However,the role of the hippocampus in visual memory andrecognition tasks has been controversial. It was recentlyconfirmed that the hippocampus is essential for nonspatialvisual recognition memory in the rat (Prusky et al, 2004).
The observation of similarities between cognitive proces-sing in the rat and human is important, but the develop-ment of a reliable rat model of recognition memory deficitalso provides a powerful tool for development of drugs forAD and neurodegenerative diseases. It is noteworthy thatvisual association tests are reported to detect, with highspecificity, a substantial proportion of AD patients up to ayear before diagnosis (Lindeboom et al, 2002). This paperdemonstrates that novel nitrates are a new class of cognitionenhancing agents, and also validates the use of the ratDMTS model by assay of the AD drug donepezil. In theoriginal DMTS work (Prusky et al, 2004), the perirhinalcortex was lesioned by aspiration and the hippocampuslesioned by direct injection of excitotoxic doses of NMDA.
We show herein that intraventricular infusion of 192 IgG-saporin induces a reliable cholinergic lesion resulting invisual recognition memory impairment, which is reversed Effect of GT1061 on ERK phosphorylation. (a) Levels of by treatment with GT1061 and donepezil. We note here that pERK1/2 in different brain regions 60 min after a single i.p. injection of with the present results we cannot determine the extent to GT1061 (1 mg/kg). Data are expressed as a percentage of the levels of which damage to noncholinergic cells may contribute to the pERK measured in saline-injected animals (mean7SEM, n ¼ 6). *po0.05 vs control, Student’s t-test for unpaired data. Inset: Levels of pERK in the Our observation that a novel nitrate, NO mimetic hippocampus 60 min after a single i.p. injection of GT1061 (1 mg/kg) with compound exerts beneficial effects in multiple learning or without prior treatment of animals with the NMDA receptor antagonist,MK-801 (1 mg/kg, i.p.). Data are expressed as a percentage of the levels of paradigms, supports the idea that activation of the sGC/ pERK measured in saline-injected animals (mean7SEM, n ¼ 6). *po0.05 vs cGMP signal transduction cascade enhances synaptic control, Student’s t-test for unpaired data. (b) Levels of pERK1/2 in the efficacy in brain regions responsible for the acquisition, CA1 region of hippocampal mini-slices in vitro after a 30 min incubation with storage, and retention of new learning. Activation of NO/ GT1061 (50 mM) or the cholinergic muscarinic receptor agonist, carbachol cGMP signaling represents a new drug target for AD, a (50 mM). Both compounds induced an equivalent increase in the content of disease characterized by generalized, progressive decline in pERK. Data are expressed as the mean7SEM (n ¼ 5). *po0.05 vs control, cognitive function due to diffuse neuronal loss and synaptic repeated measures ANOVA (F(2,8) ¼ 9.62, p ¼ 0.0074. Pair wise compar-isons were performed using Newman–Keuls post hoc test.
failure in the hippocampus and cerebral cortex (Scheff andPrice, 2003; Selkoe, 2002).
Acetylcholine (ACh), a major modulatory neurotransmit- ter in the brain, is critically involved in learning andmemory. In the cortex and hippocampus, the procognitive Development of pharmacologically effective treatments that actions of ACh are mediated via muscarinic receptors whose alleviate cognitive disorders is a central objective in activity is primarily excitatory, and involves multiple neuroscience. The high incidence of dementia, in particular different ionic conductances. In the hippocampus, choli- AD, in the aging population provides strong impetus to nergic muscarinic receptor activation leads to increased search for new compounds with the potential to reverse the tissue levels of cGMP (Tonnaer et al, 1991). Importantly, the deficits in short-term memory. The neuropathological inhibition of the slow afterhyperpolarizing current (a changes underlying age-related cognitive decline and AD calcium-activated potassium conductance that underlies are complex and poorly understood. However, the loss of spike-frequency adaptation) induced by muscarinic recep- forebrain cholinergic neurons that project to the cerebral tor activation in hippocampal CA1 pyramidal neurons can cortex and hippocampus is a hallmark of this disease. In be blocked by inhibitors of sGC and PKG (Krause and addition, the deposition of Ab in both the brain paren- Pedarzani, 2000). These data suggest that the neuromodu- chyma and cerebral blood vessels, and the formation of latory effects of ACh in the brain involve, at least in part, the intracellular neurofibrillary tangles containing tau-protein NO/sGC/cGMP signal transduction cascade.
are also observed. Studies in Alzheimer’s patients have Behavioral studies have demonstrated that NMDA revealed that the extent of cholinergic depletion correlates receptors also play an important role in both spatial with the severity of memory deficits (Mesulam, 2004). Thus, working memory and long-term memory processes. Block- animal models of cortical and hippocampal cholinergic ade of NMDA receptors, or inhibition of NOS activity, depletion have been widely used to induce relevant impairs performance in the Y-maze test, a model of spatial working memory (Yamada et al, 1996). The impairment in CREB phosphorylation (Chien et al, 2003). In in vitro induced by NMDA receptor blockade could be reversed brain slice experiments, we found that GT1061 increased the by intracerebroventricular administration of L-arginine, a levels of pERK in the CA1 region of the hippocampus. In nitrosothiol NO donor, or a cGMP analog. Taken together, addition, the finding that in vivo administration of GT1061, these studies suggest that both ACh and glutamate activate at doses associated with cognitive improvement, increased receptor systems coupled to the NO/sGC/cGMP signal ERK1/2 phosphorylation in the hippocampus and cerebral transduction cascade and that this biochemical cascade cortex, provides evidence for a mechanism of GT1061 is important for synaptic plasticity and the formation of cognition enhancement via sGC/cGMP signaling mediated at least in part by the ERK cascade.
In previous studies, we found that the novel nitrate, Arancio and co-workers reported recently that activation GT715, was less potent than GTN for relaxation of vascular of sGC or addition of a cGMP analog were able to protect smooth muscle and for decreasing mean arterial pressure, against the Ab-induced impairment of hippocampal CA1 and more effective than GTN for activating sGC in the region LTP (Puzzo et al, 2005). In mouse hippocampal slices hippocampus. In simile with these data, the activation of treated with agents in the presence of Ab1–42 before sGC in broken cell preparations of aorta and hippocampus induction of LTP, Ab inhibited LTP and blocked elevation revealed that GT1061 selectively increased enzyme activity of pCREB in response to tetanus; the activation of sGC had in the hippocampus (16- to 20-fold in hippocampus vs 2- to no effect on basal LTP and pCREB levels in the absence of 3-fold in aorta), confirming that the structure of the nitrate the tetanus stimulation. A further study provided a pathway NO mimetic influences tissue selectivity for sGC activa- for Ab-induced disruption of NO/cGMP signaling initiated tion. Additionally, sGC activity was not altered in tissue by binding of Ab to the a7 nicotinic ACh receptor (Snyder homogenates from lesioned animals, indicating that cGMP et al, 2005). Treatment of neurons with Ab was reported to signaling remained intact after depletion of cholinergic produce a depression of NMDA-evoked currents and innervation of the hippocampus. Soluble GC is a hetero- reduced CREB phosphorylation via a mechanism involving dimer of a and b subunits, and recent studies of the dissociation of the synaptic protein, PSD95, from the distribution of a1 and a2 subunits of sGC have shown that NMDA receptor. It is reasonable to assume that this event the a2 subunit is expressed in relatively high abundance in will seriously disrupt NO/cGMP signaling, since both sGC brain, especially in the hippocampus, where the a2b1 rather and nNOS are known to be bound to NMDA receptor- than the a1b1 heterodimer of sGC is the major isoform associated PSD95 (Brenman and Bredt, 1997; Mergia et al, present (Mergia et al, 2003). Furthermore, the a2 subunit 2003; Russwurm et al, 2001). Thus, evidence is growing interacts with a PDZ domain of PSD95 (Russwurm et al, justifying intervention with NO mimetic drugs in treatment 2001). As this scaffolding protein couples the NMDA receptor with neuronal NOS, this suggests an important In summary, we have shown for the first time that the role for the a2b1 heterodimer isoform in synaptic transmis- visual recognition memory deficit induced by cholinergic sion. The enhanced activation of hippocampal sGC by depletion in a rat model can be fully reversed by drug GT1061 suggests that there is selectivity for targeting the administration. All current clinical agents approved for use in mild to moderate AD attempt to salvage the residual A number of kinase signaling cascades may be involved in functionality of damaged and degenerating cholinergic signal transduction by NO/cGMP resulting in enhanced neurons, whereas we demonstrate herein an alternative cognition and memory; chief among these being the ERK approach whereby the function of the signaling apparatus is cascade (Lu et al, 1999). Activation of ERK1/2 is believed to modulated to compensate for damage to the cholinergic play a critical role in cognition and memory mediated circuitry. The drug that supports this postulate is a novel through phosphorylation and activation of CREB, and may nitrate NO mimetic, GT1061. Importantly, this compound serve as a point of convergence for a number of upstream displays significant oral bioavailability and brain penetra- signaling molecules, including glutamate acting via NMDA tion, and efficacy that, in the visual memory recognition receptors and ACh acting via cholinergic muscarinic and DMTS task, was superior to the clinical AD drug, donepezil.
nicotinic receptors (Sweatt, 2004). Activation of transcrip- Moreover, GT1061 reverses the cognitive deficits induced by tion regulating factors, such as CREB, leading to protein depletion of forebrain ACh in two different learning synthesis is a requirement for the consolidation of new paradigms, reinforcing both the previous observations with learning into long-term memory (Silva et al, 1998b). ERK1/2 the novel nitrate, GT715, and the potential for development activation can be mediated via the protein kinases, PKA or of nitrate NO mimetics as new therapeutic agents. GT1061 PKC, whereas CREB can be activated and phosphorylated induces a selective activation of sGC in the hippocampus via CaMK IV, PKA, and RSK2, the last mediating the compared to the vasculature, predicting that the beneficial activation of CREB by the ERK cascade (Xing et al, 1996).
CNS actions of the drug will occur at lower plasma con- Studies suggest that PKG provides a parallel pathway to centrations than will any potential undesirable hypotensive PKA-signaling in both phases of LTP, with PKG and PKA actions. GT1061 was observed to activate a pathway in the pathways performing complementary roles (Lu et al, 1999).
hippocampus, the ERK cascade, (a) known to be modulated NO/cGMP and PKG contribute to CREB phosphorylation, by sGC/cGMP signaling, and (b) which represents a mediated by the ERK cascade, and also possibly in part via necessary pathway in the formation of long-term memory.
the CaMK pathway (Lu et al, 1999). Interestingly, experi- Novel nitrates, as represented by GT1061, are a class of ments with YC-1, an agent that augments sGC activation, compounds that can be exploited in the discovery and produced an enhancement of LTP in rat hippocampus and development of new therapeutic agents for the treatment of amygdala via an NO/cGMP/PKG/ERK pathway culminating neurodegenerative diseases such as Alzheimer’s disease.
Nitrates and cognition enhancementBM Bennett et al selective and discrete lesioning of cholinergic basal forebrainnuclei. J Neurosci Methods 91: 9–19.
We thank Dr Khem Jhamandas for helpful discussions and Prickaerts J, de Vente J, Honig W, Steinbusch HW, Blokland A critical reading of the manuscript. Mrs Diane Anderson, Ms (2002). cGMP, but not cAMP, in rat hippocampus is involved in Margo Poklewska-Koziel, and Neuroinvestigations Inc., early stages of object memory consolidation. Eur J Pharmacol Lethbridge, Alberta are thanked for technical assistance.
Financial support for this project was provided by GB Prickaerts J, Steinbusch HW, Smits JF, de Vente J (1997). Possible Therapeutics Ltd, Mississauga Ontario, The Institute for the role of nitric oxide-cyclic GMP pathway in object recognition Study of Aging (to GRJT), and The Heart and Stroke memory: effects of 7-nitroindazole and zaprinast. Eur J Foundation of Ontario (Grant T-5162 to BMB). RJS is an Prusky GT, Douglas RM, Nelson L, Shabanpoor A, Sutherland RJ (2004). Visual memory task for rats reveals an essential role forhippocampus and perirhinal cortex. Proc Natl Acad Sci USA 101: Puzzo D, Vitolo O, Trinchese F, Jacob JP, Palmeri A, Arancio O Areosa SA, Sherriff F, McShane R (2005). Memantine for dementia.
(2005). Amyloid-beta peptide inhibits activation of the nitric Cochrane Database Syst Rev 3: CD003154.
oxide/cGMP/cAMP-responsive element-binding protein pathway Artz JD, Toader V, Zavorin SI, Bennett BM, Thatcher GR (2001). In vitro activation of soluble guanylyl cyclase and nitric oxide release: a comparison of NO donors and NO mimetics.
Redish AD, Touretzky DS (1998). The role of the hippocampus in solving the Morris water maze. Neural Comput 10: 73–111.
Bartus RT, Dean III RL, Beer B, Lippa AS (1982). The cholinergic Reynolds JN, Bennett BM, Boegman RJ, Jhamandas K, Ratz JD, hypothesis of geriatric memory dysfunction. Science 217: 408–414.
Zavorin SI et al (2002). Neuroprotection against ischemic brain Bennett BM, McDonald BJ, Nigam R, Simon WC (1994).
injury conferred by a novel nitrate ester. Bioorg Med Chem Lett Biotransformation of organic nitrates and vascular smooth muscle cell function. Trends Pharmacol Sci 15: 245–249.
Ringo JL (1991). Memory decays at the same rate in macaques with Bennett BM, McDonald BJ, St James MJ (1992). Hepatic and without brain lesions when expressed in d’ or arcsine terms.
cytochrome P-450-mediated activation of rat aortic guanylyl cyclase by glyceryl trinitrate. J Pharmacol Exp Ther 261: 716–723.
Russwurm M, Wittau N, Koesling D (2001). Guanylyl cyclase/PSD- Bernabeu R, Schroder N, Quevedo J, Cammarota M, Izquierdo I, 95 interaction: targeting of the nitric oxide-sensitive alpha2beta1 Medina JH (1997). Further evidence for the involvement of a guanylyl cyclase to synaptic membranes. J Biol Chem 276: hippocampal cGMP/cGMP-dependent protein kinase cascade in memory consolidation. Neuroreport 8: 2221–2224.
Scheff SW, Price DA (2003). Synaptic pathology in Alzheimer’s Bonkale WL, Winblad B, Ravid R, Cowburn RF (1995). Reduced disease: a review of ultrastructural studies. Neurobiol Aging 24: nitric oxide responsive soluble guanylyl cyclase activity in the superior temporal cortex of patients with Alzheimer’s disease.
Selkoe DJ (2002). Alzheimer’s disease is a synaptic failure. Science Brenman JE, Bredt DS (1997). Synaptic signaling by nitric oxide.
Silva AJ, Giese KP, Fedorov NB, Frankland PW, Kogan JH (1998a).
Molecular, cellular, and neuroanatomical substrates of place Chien WL, Liang KC, Teng CM, Kuo SC, Lee FY, Fu WM (2003).
learning. Neurobiol Learn Mem 70: 44–61.
Enhancement of long-term potentiation by a potent nitric oxide- Silva AJ, Kogan JH, Frankland PW, Kida S (1998b). CREB and guanylyl cyclase activator, 3-(5-hydroxymethyl-2-furyl)-1-ben- memory. Annu Rev Neurosci 21: 127–148.
zyl-indazole. Mol Pharmacol 63: 1322–1328.
Smith S, Dringenberg HC, Bennett BM, Thatcher GR, Reynolds JN Courtney C, Farrell D, Gray R, Hills R, Lynch L, Sellwood E et al (2000). A novel nitrate ester reverses the cognitive impairment (2004). Long-term donepezil treatment in 565 patients with caused by scopolamine in the Morris water maze. Neuroreport Alzheimer’s disease (AD2000): randomised double-blind trial.
Snyder EM, Nong Y, Almeida CG, Paul S, Moran T, Choi EY et al Karnovsky MJ, Roots L (1964). A ‘direct-coloring’ thiocholine (2005). Regulation of NMDA receptor trafficking by amyloid- method for cholinesterases. J Histochem Cytochem 12: 219–221.
Kirk RE (1982). Experimental Design, 2nd edn. Wadsworth: Sweatt JD (2004). Mitogen-activated protein kinases in synaptic plasticity and memory. Curr Opin Neurobiol 14: 311–317.
Krause M, Pedarzani P (2000). A protein phosphatase is involved Thatcher GR, Nicolescu AC, Bennett BM, Toader V (2004). Nitrates in the cholinergic suppression of the Ca(2+)-activated K(+) and NO release: contemporary aspects in biological and current sI(AHP) in hippocampal pyramidal neurons. Neuro- medicinal chemistry. Free Radic Biol Med 37: 1122–1143.
Tonnaer JA, Cheung CL, De Boer T (1991). cGMP formation and Lindeboom J, Schmand B, Tulner L, Walstra G, Jonker C (2002).
phosphoinositide turnover in rat brain slices are mediated by Visual association test to detect early dementia of the Alzheimer pharmacologically distinct muscarinic acetylcholine receptors.
type. J Neurol Neurosurg Psychiatry 73: 126–133.
Lu YF, Kandel ER, Hawkins RD (1999). Nitric oxide signaling Xing J, Ginty DD, Greenberg ME (1996). Coupling of the RAS- contributes to late-phase LTP and CREB phosphorylation in the MAPK pathway to gene activation by RSK2, a growth factor- hippocampus. J Neurosci 19: 10250–10261.
regulated CREB kinase. Science 273: 959–963.
Mergia E, Russwurm M, Zoidl G, Koesling D (2003). Major Yamada K, Noda Y, Hasegawa T, Komori Y, Nikai T, Sugihara H occurrence of the new alpha2beta1 isoform of NO-sensitive et al (1996). The role of nitric oxide in dizocilpine-induced guanylyl cyclase in brain. Cell Signal 15: 189–195.
impairment of spontaneous alternation behavior in mice.
Mesulam M (2004). The cholinergic lesion of Alzheimer’s disease: J Pharmacol Exp Ther 276: 460–466.
pivotal factor or side show? Learn Mem 11: 43–49.
Zavorin SI, Artz JD, Dumitrascu A, Nicolescu A, Scutaru D, Smith Pizzo DP, Waite JJ, Thal LJ, Winkler J (1999). Intraparenchymal SV et al (2001). Nitrate esters as nitric oxide donors: SS-nitrates.
infusions of 192 IgG-saporin: development of a method for

Source: http://lab.pharm.uic.edu/thatcher/pub/p_1.pdf