the experimental alzheimer drug phenserine: preclinical pharmacokinetics and pharmacodynamics

11
The experimental Alzheimer drug phenserine: preclinical pharmacokinetics and pharmacodynamics Greig NH, De Micheli E, Holloway HW, Yu Q-S, Utsuki T, Perry TA, Brossi A, Ingram DK, Deutsch J, Lahiri DK, Soncrant TT. The experimental Alzheimer drug phenserine: preclinical pharmacokinetics and pharmacodynamics. Acta Neurol Scand 2000: Supplement 176: 74–84. # Munksgaard 2000. Phenserine, a phenylcarbamate of physostigmine, is a new potent and highly selective acetylcholinesterase (AChE) inhibitor, with a >50-fold activity versus butyrylcholinesterase (BChE), in clinical trials for the treatment of Alzheimer’s disease (AD). Compared to physostigmine and tacrine, it is less toxic and robustly enhances cognition in animal models. To determine the time-dependent effects of phenserine on cholinergic function, AChE activity, brain and plasma drug levels and brain extracellular acetylcholine (ACh) concentrations were measured in rats before and after phenserine administration. Additionally, its maximum tolerated dose, compared to physostigmine and tacrine, was determined. Following i.v. dosing, brain drug levels were 10-fold higher than those achieved in plasma, peaked within 5 min and rapidly declined with half- lives of 8.5 and 12.6 min, respectively. In contrast, a high (>70%) and long-lasting inhibition of AChE was achieved (half-life >8.25 h). A comparison between the time-dependent plasma AChE inhibition achieved after similar oral and i.v. doses provided an estimate of oral bioavailability of 100%. Striatal, in vivo microdialysis in conscious, freely-moving phenserine-treated rats demonstrated >3-fold rise in brain ACh levels. Phenserine thus is rapidly absorbed and cleared from the body, but produces a long-lasting stimulation of brain cholinergic function at well tolerated doses and hence has superior properties as a drug candidate for AD. It selectively inhibits AChE, minimizing potential BChE side effects. Its long duration of action, coupled with its short pharmacokinetic half-life, reduces dosing frequency, decreases body drug exposure and minimizes the dependence of drug action on the individual variations of drug metabolism commonly found in the elderly. N. H. Greig 1 , E. De Micheli 1 , H. W. Holloway 1 , Q.-S. Yu 1 , T. Utsuki 1 , T. A. Perry 1 , A. Brossi 1 , D. K. Ingram 1 , J. Deutsch 2 , D. K. Lahiri 3 , T. T. Soncrant 4 1 Laboratory of Neurosciences, Intramural Research Program, National Institute on Aging, Gerontology Research Center*, 5600 Nathan Shock Dr., Baltimore, MD 21224; 2 Department of Pharmaceutical Chemistry, Hebrew University, Jerusalem, Israel; 3 Department of Psychiatry, Institute of Psychiatric Research, Indiana University School of Medicine, Indianapolis, Indiana 46202; 4 Cure Inc., Silver Spring, MD20906 Key words: Alzheimer’s disease; anticholinesterases; acetylcholine; cholinomimetics; acetylcholinesterase; butyrylcholinesterase; in vivo microdialysis Nigel H. Greig, Drug Design & Development, LNS, Gerontology Research Center, 5600 Nathan Shock Dr., Baltimore, MD 21224–6825, USA Tel.: (410) 558 8278 Fax: (410) 558 8323 e-mail: [email protected] There is considerable present interest in the devel- opment of therapeutics to augment memory in patients with Alzheimer’s disease, AD, and slow the disease course (for review see 1, 2). AD is the fourth leading cause of death in Western societies, affecting over 4 million Americans and 7 million Europeans. It afflicts all races world wide, and has a mean survival time from diagnosis of approximately 8 years. AD is characterized by a progressive cognitive impairment and eventually leads to a total loss of memory, language, mobility and consciousness. From a neuropathological viewpoint, it is charac- terized by i) the presence of senile plaques, extra- cellular deposits primarily composed of a 39–43 amino acid peptide, termed b-amyloid peptide, Ab (3, 4); ii) the presence of neurofibrillary tangles * Fully accredited by the American Association for the Accreditation of Laboratory Animal Care. Acta Neurol Scand 2000: Supplement 176: 74–84 Printed in UK. All rights reserved Copyright # Munksgaard 2000 ACTA NEUROLOGICA SCANDINAVICA ISSN 0065-1427 74

Upload: n-h-greig

Post on 06-Jul-2016

213 views

Category:

Documents


0 download

TRANSCRIPT

The experimental Alzheimer drugphenserine: preclinical pharmacokinetics andpharmacodynamics

Greig NH, De Micheli E, Holloway HW, Yu Q-S, Utsuki T, Perry TA,Brossi A, Ingram DK, Deutsch J, Lahiri DK, Soncrant TT. Theexperimental Alzheimer drug phenserine: preclinical pharmacokineticsand pharmacodynamics.Acta Neurol Scand 2000: Supplement 176: 74±84. # Munksgaard 2000.

Phenserine, a phenylcarbamate of physostigmine, is a new potent andhighly selective acetylcholinesterase (AChE) inhibitor, with a >50-foldactivity versus butyrylcholinesterase (BChE), in clinical trials for thetreatment of Alzheimer's disease (AD). Compared to physostigmine andtacrine, it is less toxic and robustly enhances cognition in animal models.To determine the time-dependent effects of phenserine on cholinergicfunction, AChE activity, brain and plasma drug levels and brainextracellular acetylcholine (ACh) concentrations were measured in ratsbefore and after phenserine administration. Additionally, its maximumtolerated dose, compared to physostigmine and tacrine, was determined.Following i.v. dosing, brain drug levels were 10-fold higher than thoseachieved in plasma, peaked within 5 min and rapidly declined with half-lives of 8.5 and 12.6 min, respectively. In contrast, a high (>70%) andlong-lasting inhibition of AChE was achieved (half-life >8.25 h). Acomparison between the time-dependent plasma AChE inhibitionachieved after similar oral and i.v. doses provided an estimate of oralbioavailability of 100%. Striatal, in vivo microdialysis in conscious,freely-moving phenserine-treated rats demonstrated >3-fold rise inbrain ACh levels. Phenserine thus is rapidly absorbed and cleared fromthe body, but produces a long-lasting stimulation of brain cholinergicfunction at well tolerated doses and hence has superior properties as adrug candidate for AD. It selectively inhibits AChE, minimizingpotential BChE side effects. Its long duration of action, coupled with itsshort pharmacokinetic half-life, reduces dosing frequency, decreasesbody drug exposure and minimizes the dependence of drug action on theindividual variations of drug metabolism commonly found in the elderly.

N. H. Greig1, E. De Micheli1,H. W. Holloway1, Q.-S. Yu1,T. Utsuki1, T. A. Perry1, A. Brossi1,D. K. Ingram1, J. Deutsch2,D. K. Lahiri3, T. T. Soncrant4

1Laboratory of Neurosciences, Intramural Research

Program, National Institute on Aging, Gerontology

Research Center*, 5600 Nathan Shock Dr., Baltimore,

MD 21224; 2Department of Pharmaceutical Chemistry,

Hebrew University, Jerusalem, Israel; 3Department of

Psychiatry, Institute of Psychiatric Research, Indiana

University School of Medicine, Indianapolis, Indiana

46202; 4Cure Inc., Silver Spring, MD20906

Key words: Alzheimer's disease; anticholinesterases;

acetylcholine; cholinomimetics; acetylcholinesterase;

butyrylcholinesterase; in vivo microdialysis

Nigel H. Greig, Drug Design & Development, LNS,

Gerontology Research Center, 5600 Nathan Shock Dr.,

Baltimore, MD 21224±6825, USA

Tel.: (410) 558 8278

Fax: (410) 558 8323

e-mail: [email protected]

There is considerable present interest in the devel-opment of therapeutics to augment memory inpatients with Alzheimer's disease, AD, and slow thedisease course (for review see 1, 2). AD is the fourthleading cause of death in Western societies, affectingover 4 million Americans and 7 million Europeans.

It af¯icts all races world wide, and has a meansurvival time from diagnosis of approximately 8years. AD is characterized by a progressive cognitiveimpairment and eventually leads to a total loss ofmemory, language, mobility and consciousness.From a neuropathological viewpoint, it is charac-terized by i) the presence of senile plaques, extra-cellular deposits primarily composed of a 39±43amino acid peptide, termed b-amyloid peptide, Ab(3, 4); ii) the presence of neuro®brillary tangles

* Fully accredited by the American Association for theAccreditation of Laboratory Animal Care.

Acta Neurol Scand 2000: Supplement 176: 74±84Printed in UK. All rights reserved

Copyright # Munksgaard 2000

ACTA NEUROLOGICASCANDINAVICA

ISSN 0065-1427

74

(NFTs), composed of phosphorylated tau protein,and iii) a neuronal loss, and particularly a synapticloss, with brain atrophy (5±7). These pathologicalchanges likely occur over a one or two decadeduration prior to the loss of intellectual capacity (8).

It is the synaptic loss, rather than the presence orquantity of plaques and NFTs, that best correlatesto the psychometric measures of declining cognitiveperformance in AD (7). Associated with thissynaptic loss is a loss of cholinergic neuronalmarkers, particularly of the enzymes cholineacetyltransferase (ChAT) and acetylcholinesterase(AChE), in the forebrain and its projection areas inthe cortex and hippocampus, and this represents theearliest determined neurochemical event that leadsto AD (5, 9). These enzymes are markers of thepre- and post-synaptic element of cholinergicneurons which is the most affected neurotransmittersystem involved in AD and, additionally, is aprimary system involved in memory processing (10,11). Whereas multiple neurotransmitter systemde®ciencies simultaneously occur in AD, these doso always in addition to, and not instead of, thedramatic cholinergic system loss which affectselements of both the muscarinic and nicotinicsystem (for review see 12). In turn, normalfunctioning of the muscarinic and nicotinic systemregulates the processing of b-amyloid precursorprotein, b-APP, to produce either potentiallyneurotoxic Ab peptide or nonamyloidogenic com-ponents (13, 14).

These ®ndings led to the hypothesis thatcholinergic augmentation might improve cognitiveperformance in AD (5, 15). The most widelyexplored and ef®cient approach to achieve thishas involved the use of anticholinesterase agentsto prolong the life of endogenously releasedacetylcholine, ACh (1, 2). The approach hasproved preferable to direct receptor agonisttherapy, as it ampli®es the natural spatial andtemporal pattern of ACh release, rather thantonically or globally stimulating either nicotinic ormuscarinic receptors. Clinical studies that haveutilized relatively unsophisticated ®rst and secondgeneration cholinesterase inhibitors, ChEIs, havedemonstrated encouraging improvements in mea-sures of cognitive performance in patients withdementia of the AD type (2, 16±18). In general,patients have better responses with the use ofhigher doses of ChEIs, likely concomitant withhigher levels of central nervous system acetylcho-linesterase, AChE, inhibition and ACh concentra-tion. However, the appearance of ChEI-inducedside effects, primarily peripherally mediated cho-linergic overdrive, has precluded the use of themajority of ChEIs in adequate doses to drama-

tically improve cognition (1, 2, 19). We hypothe-sized that it was possible to dissociate theclinically bene®cial actions of anticholinesterasetherapy in brain from its adverse central andperipheral ones by designing a new class of brain-directed and enzyme subtype selective ChEIs.Phenserine represents the ®rst of this class, intowhich we additionally incorporated features toprovide it optimal characteristics, from a phar-macokinetic and pharmacodynamic standpoint,for use in the elderly to maximize the action ofand test the true utility of cholinergic augmenta-tion in the therapy of AD.

We describe, herein, the time-dependent effects ofphenserine, an AChE-selective ChEI, on AChEactivity, brain extracellular ACh concentrations,brain and plasma drugs levels, and its maximumtolerated dose in rodents. This new physostigmineanalogue has demonstrated dramatic action inimproving cognitive performance in rats (20, 21)and is now entering Phase II clinical studies.

Materials and methods

Phenserine synthesis

Phenserine, (x)-phenylcarbamoyleseroline, wasprepared from (x)-eseroline and phenylisocyanate,as described previously (22), as its highly water-soluble (L)-tartrate salt by reaction with (L)-tartaricacid.

Anticholinesterase activity

An in vitro enzyme inhibition study was undertakento quantitate, side-by-side, the activity of phenser-ine, in salt and free base form, and to compare theseto several agents already in clinical assessment,physostigmine (Synapton, Forest Labs.), tacrine(Cognex, Warner Lambert, MI), (x)-heptylcarba-moyleseroline (Eptastigmine, Mediolanum, Italy),galanthamine (Johnson and Johnson, PA), andaricept (Donepezil, Eisai/P®zer, Teaneck, NJ)against human erythrocyte AChE and plasmabutyrylcholinesterase, BChE, as described pre-viously (23), using the spectrophotometric methodof Ellman et al. (24). Anticholinesterase activity wasexpressed as an IC50, which is de®ned as theconcentration (nM) required to inhibit 50% of theenzyme activity of AChE and BChE. The com-pounds were analyzed in duplicate on four differentoccasions.

Octanol-water partition coef®cient

Five milliliters of 0.2 mM octanol solutions ofphenserine and physostigmine were prepared andtheir UV absorbencies, A1, were determined by

Phenserine: preclinical pharmacokinetics & pharmacodynamics

75

spectrophotometer at 254 nm wavelength. Theoctanol solutions then were vigorously mixed withan equal volume of water for 15 min. Followingseparation, the absorbency of the octanol was againmeasured, A2. An octanol-water partition coef®-cient, P, was calculated from the formula: P=A2/A1xA2.

Duration of cholinesterase inhibition

Three-month-old male Fischer-344 rats wereanesthetized with halothane (Ayerst, New York,NY) and PE 50 catheters, ®lled with heparinizedisotonic saline were tied into their right femoralartery and vein. Animals then were restrained in aplaster cast that allowed them to move their headand forequarters only, and they were placed in atemperature-controlled enclosure to allow them torecover from anesthesia. A plasma sample then waswithdrawn for determination of resting AChEactivity. For animals to be administered eitherphysostigmine or tacrine, at 90 min followingsurgery, the quaternary nicotinic antagonist hex-amethonium bromide (5 mg/kg) (Sigma ChemicalCo., St Louis, MO) was administered i.p., followedby s.c. administration of the quaternary muscarinicantagonist atropine methyl bromide (4 mg/kg)(Sigma Chemical Co.) 10 min later. These antago-nists do not cross the blood±brain barrier andreduce peripheral cholinergic overdrive, caused byChE inhibition, which otherwise would be deleter-ious to the rats. At 2 h after surgery, eitherphenserine (1.0 mg/kg), physostigmine (2 mg/kg),eptastigmine (2 mg/kg) or tacrine (10 mg/kg) wasadministered i.v. (1 ml/kg in isotonic saline). Thesedoses were chosen to achieve approximately 50%inhibition of AChE. Blood samples were removedprior to administration of the ChEIs and from 2 to480 min thereafter, plasma samples were immedi-ately collected by centrifugation (10,000 g, 2 min)and frozen to x70uC.

In an additional series of eight anesthetized rats,PE 50 catheters were tied into the femoral artery asdescribed above, and phenserine (0.7 mg/kg) wasadministered either as an i.v. bolus into thesaphenous vein (1 mg/0.5 ml) or p.o. (1 mg/ml) byguavage into the stomach by intragastric intuba-tion. Blood samples were removed prior to admin-istration of phenserine and from 5 min to 24 hthereafter, plasma samples were immediately col-lected by centrifugation (10,000 g, 2 min) andfrozen to x70uC. In a further six animals, eitherphenserine (1 mg/kg, i.v.) or saline was adminis-tered and the animals were killed at 4 h.Cerebrospinal ¯uid was withdrawn from thecisterna magna, blood was taken by cardiac

puncture, and plasma and CSF samples werefrozen to x70uC.

ChEI assay

Samples were assayed for cholinesterase activityspectrophotometrically (23). In rat plasma, unlikethat in humans, both AChE and BChE are present.Therefore a speci®c inhibitor of BChE, Iso-OMPA(1r10x4 M), was used during the quantitativedetermination of AChE inhibition.

Pharmacokinetic studies

Three-month-old male Fischer-344 rats wereanesthetized with halothane and their saphenousvein exposed. Phenserine (1.0 mg/kg) then wasadministered i.v. (1 ml/kg in isotonic saline) andthe skin closed. At times from 5 to 60 min, aminimum of three animals were decapitated underhalothane anesthesia. Trunk blood was collectedand the brain removed, washed and the two cerebralhemispheres separated and immediately frozen tox70uC. Blood was immediately centrifuged, theplasma removed and frozen to x70uC. Drugconcentrations were later quanti®ed by high per-formance liquid chromatography, HPLC.

Determination of phenserine concentrations in biological

samples

Quantitation of phenserine was undertaken byHPLC. Chromatographic separation was per-formed using a normal phase narrow bore column(200r2 mm internal diameter (i.d.), AppliedBiosystems, Brownlee Laboratories, San Jose,CA). The mobile phase was 52% 0.01 M formicacid, 27% acetonitrile and 21% 0.05 M tris buffer,with a 0.15 ml/min ¯ow rate. Phenserine wasdetected by ¯uorescence spectroscopy (250 nmexcitation and 345 nm emission wavelengths).Quantitation was achieved by using the N-methylanalogue of physostigmine (Research BiochemicalsInc., Natick, MA) as an internal standard duringsolid phase extraction of phenserine from biologicalsamples, and chromatography. Solid phase extrac-tion was undertaken on an Adsorbex-CN extractioncolumn (model 19859±1, EM Science, Gibbstown,NJ). Calibration curves of phenserine in bothplasma and brain proved highly linear(r2=>0.99) throughout the range of concentra-tions determined in the biological samples, with alower limit of detection of 1 nM and a relative errorof less than 10%.

Greig et al.

76

Microdialysis and determination of brain ACh concentrations

Studies were performed on male Sprague±Dawleyrats (Charles River), 250±350 g weight, exposed to a12 h light±dark cycle and provided free access tofood and water. On the day prior to experimenta-tion, rats were anesthetized (equithesin 3 ml/kgi.p.), placed in a stereotaxic frame (David KopfInstruments, Tujunga, CA) and kept on a heatingpad maintained at 37uC. The scalp was incised andthe skin in®ltrated with 1% lidocaine. A custom-made microdialysis concentric probe was stereo-taxically placed in the right corpus striatum atcoordinates 0.4 mm anterior, 2.7 mm lateral frombregma, and 6.5 mm below the dura surface (25).The probe was af®xed to the skull with dentalcement, the inlet and outlet tubes of the probe weresealed with epoxy glue, and animals then werereturned to their cage for recovery. Twenty-four hlater, microdialysis studies were undertaken on thefreely moving, awake animals. The right striatumwas perfused at 2 ml/min using a microperfusionpump (Model 100, Carniegie Medicin AB,Stockholm, Sweden) with an oxygenated modi®edRinger's solution of the following composition(mM): NaCl 121, KCl 3.5, CaCl2 1.2, MgCl2 1.2,NaH2PO4 1.0 and NaHC3 25. The Ringer's solutionadditionally contained neostigmine bromide(1r10x5 M) (Sigma Chemical Co.). After a mini-mum of 2 h equilibration, at least three consecutive10 min dialysate samples were collected for thedetermination of mean basal ACh levels in brainextracellular ¯uid. In the event that basal AChlevels remained stable, phenserine was injected i.p.at either of two doses (2 and 4 mg/kg, 3 rats pergroup), and 9 consecutive 10 min dialysate sampleswere collected to determine the compound's time-dependent effect on ACh levels.

Prior to emplacement in the striatum, eachmicrodialysis probe was tested for ACh recoverywith a 20 ng/ml solution. Average probe recoverywas 12.4t0.5% (S.E). ACh concentrations weredetermined using an ESA HPLC system (ESA Inc.,Bedford, MA), equipped with an electrochemicaldetector. ACh was separated on a polymericreverse-phase column (150r3 mm i.d.) (ESA Inc.)using a mobile phase consisting of 100 mMNa2HPO4, 0.5 mM tetramethylammonium chlor-ide, 0.005% Kathon G, and 2.0 mM octane sulfonicacid, at a ®nal pH of 8.0. The ¯ow rate was 0.35 ml/min. ACh then was enzymatically converted tohydrogen peroxide by a post-column solid phasereactor (ESA Inc.) (containing immobilized AChEand choline oxidase) and the peroxide, in turn, wasmeasured electrostatically (5200 Coulochem detec-tor, ESA Inc.) using a platinum electrode set at+300 mv (versus a pallidium reference electrode).

Improved enzyme ef®ciency was obtained bymaintaining the column and reactor at 35uC.Data were not corrected for probe recoveries.

Maximum tolerated dose

Phenserine and physostigmine were administeredi.p. to male Fischer-344 rats in increasing doses,three animals per dose, and acute adverse effectswere noted. Animals in obvious distress wereeuthanized with excess halothane anesthetic.

Calculations and statistical analysis

Phenserine time-dependent concentration or AChEinhibition data were ®tted by nonlinear regressionanalysis (26) to a single exponential equation:

C=Aexat

where ``C'' equals either the concentration ofphenserine (nM) or percent inhibition of AChE attime ``t'' (min), ``A'' is the theoretical zero-timeconcentration or percent inhibition in a centralcompartment, and ``a'' is the apparent ®rst-orderelimination rate constant (minx1). Plasma andbrain pharmacokinetic half-lives, t1/2, and plasmaAChE inhibition pharmacodynamic t1/2 then werecalculated from the parameters by the generalformula: t1/2=0.693/a.

The volume of distribution and plasma clearancewere calculated from the formulae dose/A and dosea/A, respectively. Areas under the concentration-time pro®les (AUC) were calculated by thetrapezoidal rule (27) and were utilized to calculatethe brain/plasma time-dependent concentrationratio of phenserine after its i.v. administration,and its oral bioavailability (AUC oral/AUC i.v.).

A two-tailed Student's t-test was carried out tocompare two means in both pharmacokinetic andtime-dependent AChE inhibition studies. Whenmore than two means were compared, one-wayanalysis of variance and the Bonferroni multiple-

Table 1. Comparison of the IC50 values of phenserine against human erythrocyte

AChE and plasma BChE with those of several ChEIs of clinical interest

IC50 Value (nM)

Compound AChE BChE Selectivity

Physostigmine 28t2.4 16t2.9 2-fold BChE

Phenserine (L)-tartrate 22t1.4 1560t270 70-fold AChE

Phenserine free base 24t6.0 1300t85 54-fold AChE

Eptastigmine 22t2.0 5t0.1 4-fold BChE

Tacrine 190t40 47t10 4-fold BChE

Aricept 22t8.1 4150t1700 188-fold AChE

Huperzine A 47t22 >30000 >638-fold AChE

Galanthamine 800+60 7300+830 9-fold AChE

IC50 values were determined in duplicate on a minimum of 4 different occasions.

Phenserine: preclinical pharmacokinetics & pharmacodynamics

77

test were used (28). In microdialysis studies, timedsamples were compared to their correspondingbaseline values with a one-tailed paired t-test (28).Statistical signi®cance for all tests was taken asP<0.05. Unless otherwise stated, means t stand-ard errors are given routinely.

Results

ChE inhibitory action

As shown in Table 1, phenserine, both as the L-tartrate salt and free base, is a potent inhibitor ofAChE. It possesses an IC50 similar to that ofphysostigmine and aricept, and is some 8- and35-fold more potent against AChE than are tacrineand galanthamine. In contrast, like aricept, phen-serine possesses minimal BChE inhibitory activity.Indeed, unlike physostigmine and tacrine that arerelatively unselective in their cholinesterase inhibi-tion with a preference for BChE, signi®cantly largeramounts of phenserine, >50-fold, were required toproduce 50% inhibition of BChE compared toAChE.

Pharmacodynamics

As illustrated in Fig. 1, phenserine (1.0 mg/kg i.v.)caused a long duration of AChE inhibition anddemonstrated high in vivo activity. It rapidlyachieved and maintained a steady-state inhibitionof over 70% between 5 and 120 min, remaining atover 50% for 6 h. Compared to untreated animals,

phenserine induced greater than 90% inhibition ofAChE in CSF, compared to 60% plasma inhibitionat 240 min. Levels of BChE inhibition remainedbelow 30% (data not shown). In contrast, anapproximately similar dose of physostigmine pro-duce a peak plasma AChE inhibition of 45.9%2 min after administration. Thereafter, inhibitiondeclined rapidly and was not detected at 60 min.Levels of BChE inhibition mirrored those of AChE(data not shown). The inhibition half-lives ofphenserine and physostigmine were 8.25 h and19 min, respectively. A 10-fold higher dose oftacrine was required to produce a peak AChEinhibition of 33% at 2 min. Thereafter, inhibitiondeclined to less than 20%. The percent AChEinhibition induced by both physostigmine andtacrine was signi®cantly lower than that achievedby phenserine throughout the study.

Bioavailability

The time-dependent AChE activity associated withphenserine 0.7 mg/kg i.v. and p.o. administration isshown in Fig. 2. Following i.v. drug administration,a peak plasma AChE inhibition of 46.7t3.3% wasachieved at 30 min and this declined to 7.9t6.2%inhibition at 24 h. In contrast, a peak AChEinhibition of 37.1t9.2% was likewise achieved at30 min post p.o. drug administration, whichdeclined to 21.2t4.6% at 24 h. There was nosigni®cant difference (P>0.05) between the percentAChE inhibition induced by either i.v. or p.o. at anytime during the study. Calculated between zero and

Fig. 1. Time-dependent inhibition of plasma AChE in rats(minimum 3 per group) administered phenserine,physostigmine and tacrine. The level of inhibitionachieved by phenserine was signi®cantly higher than thoseachieved by physostigmine and tacrine throughout the study(P<0.05).

Fig. 2. Time-dependent inhibition of plasma AChE in rat (4per group) administered phenserine 0.7 mg/kg by theintravenous (i.v.) and oral (p.o.) routes. The levels ofinhibition, i.v. versus p.o., were not signi®cantly different atany time (P>0.05).

Greig et al.

78

24 h, the p.o./i.v. integrals of the time-dependentmean AChE inhibition were similar (p.o. integral:32,064.5; i.v. integral: 19,942.5), with a p.o./iv. ratioof 1.6.

Pharmacokinetics

In contrast to its long duration of plasma AChEinhibition, phenserine (1.0 mg/kg i.v.) was rapidlycleared from plasma (Fig. 3). It reached a peakconcentration of 117t19 ng/ml at 5 min anddisappeared with a t1/2 of 12.6 min. Ten-foldhigher levels (P<0.05) were achieved in brain,with a peak phenserine concentration of1246t101 ng/g at 5 min. Likewise, phenserinedisappeared rapidly from brain with a t1/2 of8.5 min. Its volume of distribution and clearancewere 5516 ml.kgx1 and 303 ml.minx1.kgx1,respectively. As shown in Table 2, the octanol-water partition coef®cients of phenserine and

physostigmine were 7.2 and 1.2 (log 0.86 and0.08), respectively.

In vivo microdialysis of ACh

The mean baseline concentration of ACh inextracellular ¯uid of striatum of untreated ratswas 224t46.8 fmol/10 min dialysate (n=6).Phenserine elevated ACh levels immediately follow-ing its administration to 193t38% and 344t111%of pretreatment values in 10 min dialysate samplesafter phenserine 2 and 4 mg/kg i.p., respectively. Asillustrated in Fig. 4, ACh levels remained raisedthroughout the duration of the study (P<0.05 at30 min and after), remaining at 187t23% and208t15% of their pretreatment levels, respectively,at 90 min.

Toxicity

Phenserine was dramatically less acutely toxic thanwas physostigmine in vivo. To undertake cholines-terase inhibition studies, described above, periph-eral cholinergic antagonists, speci®callyhexamethonium bromide and atropine methylbromide, were administered to reduce cholinergicover-drive which otherwise would have been fatalfollowing the administration of physostigmine at adose of 2 mg/kg. The maximum tolerated dose ofphysostigmine alone was 0.5 mg/kg by i.p. admin-istration. Doses as high as 15 mg/kg of phenserinein the form of its (L)-tartrate salt, and of 25 mg/kgof phenserine free base were administered withoutprior administration of peripheral cholinergicantagonists before distressing cholinergic over-

Fig. 3. Time-dependent brain and plasma levels ofphenserine in rat (minimum 3 per time point) followingthe administration of 1 mg/kg i.v. (brain concentrationswere signi®cantly higher than those achieved in plasmathroughout the study, P<0.05).

Table 2. Octanol/water partition coef®cients (P*) and structures of physostigmine,

phenserine and eptastigmine

Compound R A1 A2 P*

Physostigmine CH3 2.518 1.357 1.17

Phenserine Ph 2.083 1.803 7.23

Eptastigmine (CH2)6CH3 2.312 2.310 1150

Fig. 4. Effect of phenserine on brain extracellular ¯uidlevels of ACh in striatum, as measured by in vivomicrodialysis in conscious rats (3 per dose group). Asigni®cant elevation, compared to pre-treatment baselinevalue, was found at 30 min and thereafter (P<0.05).

Phenserine: preclinical pharmacokinetics & pharmacodynamics

79

drive was evident. The maximum tolerated dose oftacrine was 15 mg/kg.

Discussion

Recent clinical trials have demonstrated that oraland i.v. physostigmine can produce short-termmemory improvements in patients with AD (2,17, 19). Unfortunately, as in studies of tacrine,cognition improvement is limited by the appearanceof disabling, dose-limiting adverse effects which areprimarily systemic cholinergic-mediated parasym-pathetic actions (i.e., hypersalivation, sweating andnausea (2, 16, 17). The appearance of thesecoincides with doses that produce only modestcholinesterase inhibition and cognitive improve-ment. Additionally, as described, the short durationof action of the agents, their unselective inhibitionof all cholinesterase forms throughout the body,and their toxic effects, hepatic in particular fortacrine, limit their potential clinical value (2, 16, 17).Such limited clinical utility, however, should not besurprising as physostigmine and tacrine are classicalagents that were not speci®cally designed for use intreating the cognitive decline associated with ADand aging.

We thus synthesized novel carbamates of (x)-eseroline, (x)-noreseroline, (x)-physovenol and(x)-thiaphysovenol to assess their potential asexperimental therapeutics for AD (for review see22). Unlike alkylcarbamates of these, such asphysostigmine and its long-acting analogue eptas-tigmine (29), arylcarbamates possess a uniqueselectivity for AChE versus BChE inhibition (22).Of the arylcarbamates, the unsubstituted phenyl-carbamate, phenserine, additionally appears topossess optimal properties as a drug candidate.Phenserine, similar to physostigmine, is a potentinhibitor of AChE in the 20 nM range, beingsigni®cantly more active (8-fold) than tacrine.However, unlike physostigmine and tacrine whichnon-selectively co-inhibit BChE with a slightpreference of 2- and 4-fold, respectively, dramati-cally larger doses of phenserine (up to 70-fold) arerequired to co-inhibit BChE. This potency andselectivity compare favorably with that of eptas-tigmine, which is 4-fold BChE selective, andgalanthamine (30), which is 9-fold AChE selectivebut 40-fold less potent. Galanthamine has beenreported to reverse memory de®cits in animals andhas now been approved in some European countriesfor treating patients with AD although its potencyagainst brain-derived AChE is some 10-fold lowerthan against peripheral AChE (30), as studiedherein.

The role of BChE remains to be fully elucidatedboth systemically as well as in brain (31, 32). BChEis found in particularly high amounts in liver,suggesting a potential role in lipid metabolism, inheart, as well as in serum, where it is involved in thedegradation of a variety of drugs, includingsuccinylcholine, heroin, cocaine and physostigmine.In brain, it is found in association with neuronalcells, potentially possessing a modulatory role forthe dopaminergic system, as well as in associationwith glial cells, the meninges and in particular withthe blood±brain barrier. Interestingly, certaincholinergic neurons contain only AChE andothers only BChE (33). Additionally, there isstrong support of the involvement of BChE andAChE in brain development, speci®cally in cellularproliferation and differentiation, respectively, andin particular in the formation of adhesion synapses.Clearly, the AChE selectivity of phenserine is ofpotential pharmacologic bene®t to minimize BChE-induced effects of anticholinesterase therapy. In thisregard, the co-administration of the BChE inhibi-tor, iso-OMPA, with a selective AChE inhibitor hasbeen reported to dramatically increase the periph-eral versus central side-effects of the combinationversus either agent singly, and provide an unfavor-able side effect pro®le similar to an unselectiveChEI, such as tacrine (34). This suggests that AChEor BChE selective inhibition may be preferable tothe coinhibition of both enzymes.

Likewise, as AChE is intimately involved atcholinergic synapses within both the central andperipheral nervous systems, its preferential inhibi-tion in brain in the treatment of cognitive disorderssuch as AD should minimize potentially deleteriousperipheral effects. Whereas physostigmine is atertiary amine with a pKa of 7.9, an unionized toionized ratio of 3:1 at physiological pH (35) and anoctanol/water partition coef®cient of close to unity(1.2); phenserine, by virtue of its phenyl group, ismore lipophilic with an octanol/water partitioncoef®cient of 7.2. In contrast, eptastigmine pos-sesses an octanol/water partition coef®cient of 1150(log 3.06). As a consequence of its balancedlipophilicity, phenserine readily and preferentiallyenters the brain following its systemic administra-tion and achieves levels that are 10-fold higher thanconcomitant ones in plasma. Calculated from thearea under the time-dependent concentration curves(21,730.7 ng.min/g and 2624.2 ng.min/ml in brainand plasma respectively), phenserine possesses abrain/plasma ratio of 8, as well as a volume ofdistribution that is in excess of body water andconsistent with its balanced lipophilicity. In con-trast, brain concentrations of physostigmine closelymirror those achieved in plasma after its systemic

Greig et al.

80

administration. Physostigmine has a brain/plasmaratio of 1.2 (calculated from the AUC between 2and 120 min from data derived (36)). We havepreviously reported the importance of maintaininga balanced lipophilicity in designing compoundswhose target is within the extracellular compart-ment of the brain (37), as is the case for synapticAChE. Agents that possess a very high lipophilicity(octanol/water partition coef®cient >log 2.0) pre-ferentially enter the intracellular compartment ofthe brain, maintaining only relatively low levels inthe aqueous extracellular ¯uid, bind heavily toplasma proteins and sequester in lipid tissue.Conversely, agents that are water-soluble (octa-nol/water partition coef®cient <log 0) are restrictedfrom entering the brain.

The disappearance of phenserine from both brainand plasma was rapid with half-lives of 8.5 and12.6 min, respectively. This is in contrast to its longduration of selective AChE inhibition, half-life8.25 h. A rapid and close to steady-state inhibitionof plasma AChE was induced within 5 min (73%),maintained at >50% for 6 h, falling to 43% by 8 h.The IC50 value and selectivity of phenserine againsthuman brain-derived AChE and BChE (36t3 and2500t1100, respectively (23)) are similar to thosereported herein. As levels of drug in brain were10-fold greater than in plasma, phenserine likelyachieves even greater AChE inhibition in brain,compared to plasma. Indeed, greater than 90%inhibition of AChE in CSF was achieved at240 min. Physostigmine also disappears rapidlyfrom plasma and brain after i.v. administration(data not presented) with elimination half-lives of15 and 11 min, respectively (35). A two-fold higherdose caused a maximum inhibition of only 45.9%,and this declined to an undetectable level by 60 min.The relatively short duration of cholinesteraseinhibition achieved by physostigmine, half-life19 min, is in accord with reported studies (35, 36)and is determined by the stability of the methyl-carbamylated enzyme and its rate of decarbamyla-tion (0.011 minx1 in plasma). Phenserine, likeeptastigmine (38), forms a more stable carbamy-lated enzyme intermediate. This provides continuedinhibition long after drug elimination; the dissocia-tion between the pharmacokinetics (i.e., drugconcentration) and pharmacodynamics (i.e., dura-tion of drug action) is often termed anticlockwisehysteresis. Recent studies have shown that phenser-ine functions as a long-acting noncompetitiveAChE inhibitor (39).

Similar to phenserine, eptastigmine is potent atlow doses and possesses a long duration ofcholinesterase inhibition in rat (40). Its elimination,however, is dramatically slower with a t1/2 of

approximately 5 h (17) which likely is due to itssequestration in lipid. In contrast and as predictedby its higher IC50 value compared to phenserine, alarge dose of tacrine (10 mg/kg) was required toinduce a maximum AChE inhibition of 33%, andthis rapidly declined to <15% within 2 h.

In accord with its potency for AChE and highbrain uptake, phenserine increased ACh levels instriatum in a dose-dependent manner, as deter-mined by in vivo microdialysis with HPLC analysis.Doses of 2 and 4 mg/kg elevated basal levels to193% and 344% of pretreatment levels, respectively.Extracellular concentrations of ACh are determinedby not only the level of neuronal ACh release, butalso by the activity of AChE. Due to the presence ofendogenous AChE activity in vivo, extracellularlevels of ACh during basal conditions are oftenbelow the detection limit of available sampling anddetection techniques. Consequently, it is customaryto perform microdialysis studies utilizing a choli-nesterase inhibitor, such as neostigmine, in theperfusion ¯uid administered through the dialysisprobe to recover detectable quantities of ACh(41±43). This results in an over-estimation of thebasal ACh level and complicates interpretation ofthe actions of a systemically administered com-pound (44). Our basal value for ACh,224t46.8 fmol/10 min dialysate (20 ml totalvolume), is signi®cantly higher than recentlyreported values, 70t8.0 fmol/20 ml, in the striatumof conscious male Wistar rats (45) where noanticholinesterase was added to the perfusion¯uid. It is therefore likely that the percent elevationsin ACh levels in striatum that were achieved bysystemic administration of phenserine in our studyare dramatic under-estimations, as a consequence ofthe arti®cially elevated basal ACh levels.Nevertheless, these increases compare favorably tothose achieved following the systemic administra-tion of eptastigmine, physostigmine, tacrine (42, 44,46), SM-1088, (45) and E2020 (Donepezil/Aricept)(47) to rats. Together with our own, these resultssupport the rationale for using cholinesteraseinhibitors as cholinomimetics for the treatment ofAD, based on their ability to protect ACh fromenzymatic hydrolysis and thereby increase its half-life and concentration in regions of survivingcholinergic neurons. Indeed, aricept has provedquite effective in alleviating cognitive de®cits inpatients with early to middle stage AD (18).

Phenserine, similar to eptastigmine, is dramati-cally less acutely toxic than is physostigmine.Physostigmine could not be administered in doseshigher than 0.5 mg/kg i.p. without prior adminis-tration of quaternary, peripheral-acting muscarinicand nicotinic antagonists before life-threatening

Phenserine: preclinical pharmacokinetics & pharmacodynamics

81

cholinergic overdrive became evident. This is inaccord with its reported LD50 of 0.6 mg/kg i.p. inrodents (48). In contrast, 30- and 50-fold largerdoses of phenserine were administered in the formof its (L)-tartrate salt and free base, respectively (15and 25 mg/kg, i.p.). The maximum tolerated dose oftacrine was 15 mg/kg, i.p., which is in accord withits reported LD50 of 20 mg/kg after i.m. adminis-tration (49). The reported LD50 of eptastigmine inrodents is 35 mg/kg by i.p. administration (48, 49).

Phenserine therefore possesses selectivity of invivo activity on two levels, i) by its selective andpotent inhibition of AChE, and ii) by its preferentialbrain uptake. This may account for its unusuallywide therapeutic window in behavioral tests as acognition enhancer in rats (50), compared to otheranticholinesterases (40). Phenserine has beenreported to attenuate both scopolamine-inducedand glutamatergic-induced memory impairments inrats over a broad dose range (21), whereasphysostigmine, eptastigmine and tacrine possess anarrow effective dose range and a wide individualvariability in best dose (21, 51, 52). In addition,phenserine has been shown to improve memoryacquisition in elderly rats (20), which show an age-related decline in working memory, and is active indoses as low as 0.25 mg/kg i.p. (21); some 60-foldlower than its daily maximally tolerated dose.Furthermore, it possesses the unusual characteristicof being able to favorably modify the processing ofb-amyloid precursor protein in vivo (53) and in vitro(54).

In summary, phenserine is a potent, reversibleand highly selective inhibitor of AChE, whichpreferentially enters brain. It possesses an optimallylong duration of enzyme inhibition, t1/2 8.25 h,

which reduces dosing frequency. In contrast, itrapidly disappears from the body, which minimizestotal body exposure. Indeed, the independence ofphenserine's pharmacodynamic activity (durationof AChE inhibition) from its pharmacokinetics(rapid disappearance), shown in Fig. 5, minimizesthe effects of the high variability in drug metabolismand clearance, which is associated with the elderly(55, 56), from drug action. Drugs with such anindependence likely are safer in the elderly assigni®cant changes in clearance and distribution willonly minimally impact on drug response. Finally,phenserine produces a robust increase in brainextracellular ACh levels at doses that are well belowthose that are associated with acute toxicity andthat have previously improved cognitive perfor-mance in rats (20, 21). Indeed, recent studies havedemonstrated that phenserine (in the form of its L-tartrate salt) can be administered daily in doses ofup to 15 mg/kg p.o. in rats and 5 mg/kg p.o. in dogsfor 13 weeks without toxicity. Taken together, theseresults support the clinical development of phen-serine, and suggest that it may produce cognitiveimprovements in patients with AD without beingovershadowed by the dose-limiting adverse choli-nergic actions which have limited the utility ofprevious anticholinesterases. Phenserine was welltolerated in healthy elderly volunteers in phase Iclinical trials. Hence, results from early ef®cacytrials in AD are awaited with interest.

References

1. GIACOBINI E. Cholinesterase inhibitors do more than inhibitcholinesterase. In: BECKER R, GIACOBINI E, eds. Alzheimer'sDisease: from Molecular Biology to Therapy. Boston:Birkhauser, 1997: pp. 187±204.

2. BECKER RE, MORIEARTY P, UNNI L, VICARI S. Cholinesteraseinhibitors as therapy in Alzheimer's disease: bene®t to riskconsiderations in clinical application. In: BECKER R,GIACOBINI E, eds. Alzheimer's Disease: from MolecularBiology to Therapy. Boston: Birkhauser, 1997: pp. 257±66.

3. SELKOE DJ. Alzheimer's disease: a central role of b-amyloid.J Neuropathol Exp Neurol 1994;53:438±47.

4. CHECLER F. Processing of b-amyloid precursor protein andits regulation in Alzheimer's disease. J Neurochem1995;65:1431±44.

5. WHITEHOUSE D, PRICE DL, STRUBLE R. Alzheimer's diseaseand senile dementia: loss of neurons in the basal forebrain.Science 1982;215:1237±9.

6. DOUCETTE R, FISHMAN M, HACHINSKI V, MERSKY H. Cell lossfrom the nucleus basalis of Meynert in Alzheimer's disease.Can J Neurol Sci 1986;13:435±40.

7. TERRY RD, MASLIAH E, SALMON DP. Physical basis ofcognitive alterations in Alzheimer's disease: synapse loss isthe major correlate of cognitive impairment. Ann Neurol1991;30:572±80.

8. WISNIEWSKI HM, WEIGEL J. Neuropathological basis ofAlzheimer's disease, implications for treatment. In: BECKER

RE, GIACOBINI E, eds. Alzheimer's Disease: TherapeuticStrategies. Boston: Birkhauser, 1994: pp. 17±22.

Fig. 5. Time-dependent plasma AChE inhibition andphenserine concentration in rats (minimum 3 per timepoint) following the administration of phenserine 1 mg/kgi.v.

Greig et al.

82

9. PERRY EK, TOMLINSON BE, BLESSED G. Correlation ofcholinergic abnormalities with senile plaques and mentaltest scores in senile dementia. B M J 1978;2:1457±9.

10. DRACHMAN DA, LEAVITT J. Human memory and thecholinergic system: a relationship to aging? Arch Neurol1974;30:113±21.

11. DRACHMAN DA. Aging and dementia: insights from thestudy of anticholinergic drugs. In: KATZMAN R, eds.Biological Aspects of Alzheimer's Disease. BanburyReport, 1982;15:363±9.

12. GEULA C, MESULAM M. Cholinergic systems and relatedneuropathological predilection patterns in AlzheimerDisease. In: TERRY RD, KATZMAN K, BICK KL, eds.Alzheimer Disease. New York, NY: Raven Press, 1994:pp. 263±91.

13. BUXBAUM JD, RUEFLI AA, PARKER CA, CYPRESS AM,GREENGARD P. Calcium regulates processing of theAlzheimer's amyloid protein precursor in a protein kinaseC-independent manner. Proc Natl Acad Sci USA1994;91:4489±93.

14. NITSCH RM, GROWDON JH, FARBER SA, DENG M, WURTMAN

RJ. Regulation of APP processing by ®rst messengers. In:BECKER R, GIACOBINI E, eds. Alzheimer Disease: TherapeuticStrategies. Boston: Birkhauser, 1994:54±61.

15. BARTUS RT, DEAN RL III, BEER B, LIPPA AS. The cholinergichypothesis of geriatric memory dysfunction. Science1982;217:408±17.

16. BECKER R, GIACOBINI E. Mechanism of cholinesteraseinhibition in senile dementia of the Alzheimer-type: clinical,pharmacological and therapeutic aspects. Drug Dev Res1988;12:163±95.

17. BECKER R, MORIEARTY P, UNNI L. The second generation ofcholinesterase inhibitors: clinical and pharmacologicaleffects. In: BECKER R, GIACOBINI E, eds. Cholinergic Basisof Alzheimer's Disease. Boston: Birkhauser, 1991: pp.263±96.

18. BARNER EL, GRAY SL. Donepezil use in Alzheimer disease.Ann Pharmacother 1988;32:70±7.

19. ASTHANA S, RAFFAELE K, BERARDI A, HAXBY J, GREIG NH,SONCRANT TT. Treatment of Alzheimer's disease bycontinuous intravenous infusion of physostigmine. AlzDisease Assoc Disorders 1995;9:223±32.

20. IKARI H, SPANGLER E, GREIG NH et al. Performance of agedrats in a 14-unit T-maze is improved following chronictreatment with phenserine, a novel long-acting anticholi-nesterase. NeuroReport 1995;6:481±4.

21. PATEL N, SPANGLER EL, GREIG NH et al. Phenserine, a novelacetylcholinesterase inhibitor, attenuates impaired learningof rats in a 14-unit T-maze induced by blockade of the N-methyl-D-aspartate receptor. NeuroReport 1998;9:171±6.

22. BROSSI A, PEI XF, GREIG NH. Phenserine, a novel anti-cholinesterase related to physostigmine: total synthesis, andbiological properties. Austr J Chem 1996;49:171±90.

23. ATACK JR, YU QS, SONCRANT TT, BROSSI A, RAPOPORT S.Comparative inhibitory effects of various physostigmineanalogs against acetyl- and butyrylcholinesterase. JPharmacol Exp Ther 1989;249:194±202.

24. ELLMAN GL, COURTNEY KD, ANDRES V JR, FEATHERSTONE

RM. A new and rapid colorimetric determination ofacetylcholinesterase activity. Biochem Pharmacol 1961;7:88±95.

25. PAXINOS G, WATSON C. The Rat Brain In StereotaxicCoordinates, New York, NY: Academic Press, 1982.

26. MARQUARDT D. An alogrithm for least squares estimation ofnonlinear parameters. J Soc Ind Appl Math 1963;11:431±41.

27. WAGNER JG. Drug Intelligence, Hamilton, Illinois 1975:64.

28. MILLER, R. Simultaneous Statistical Inferences, New York,NY: McGraw-Hill, 1966: pp. 76±81.

29. BRUFANI M, CASTELLANO C, MARTA M et al. A long-lastingcholinesterase inhibitor affecting neural and behavioralprocesses. Pharmacol Biochem Behav 1987;26:625±33.

30. THOMSEN T, KEWITZ H, BICKEL V, STRASCHILL M, HOLL G.Preclinical and clinical studies with galanthamine. In:BECKER R, GIACOBINI E, eds. Cholinergic Basis ofAlzheimer's Disease, Boston: Birkhauser, 1991: pp. 328±36.

31. CHATANET A, LOCKRIDGE O. Comparison of butyrylcholi-nesterase and acetylcholinesterase. Biochem J 1989;260:625±34.

32. SOREQ H, ZAKUT H. Human Cholinesterases andAnticholinesterases. New York, NY: Academic Press,1993.

33. MESULAM MM, VOLICER L, MARQUIS JK, MUFSON EJ, GREEN

RC. Systematic regional differences in the cholinergicinervation of the primate cerebral cortex; distribution ofenzyme activities and some behavioral implications. AnnNeurol 1986;19:144±51.

34. LISTON D, NIELSEN JA, VILLALOBOS A et al. CP-118,954: apotent and selective AChE inhibitor. The importance ofselective AChE inhibition in vivo. Proc Soc Neurosci1994;20:608.

35. SOMANI SM, KHALIQUE A. Pharmacokinetics and pharma-codynamics of physostigmine in the rat after intravenousadministration. Drug Metab Dis 1987;15:627±33.

36. SOMANI SM, KAHLIQUE A. Distribution and pharmacoki-netics of physostigmine in rat after intramuscular admin-istration. Fund Appl Toxicol 1986;6:327±34.

37. GREIG NH. Drug entry to the brain and its pharmacologicmanipulation. In: BRADBURY MWB, ed. Physiology andPharmacology of the Blood±Brain Barrier. Handb ExpPharm 1992;103:485±523.

38. FREEDMAN SB, IVERSON LL, RUGARLI PL, HARLEY EA.Heptylphysostigmine: a potent inhibitor of acetylcholines-terase with long duration of action. Proc SecondInternational Spring®eld Symposium on Advances inAlzheimer's Therapy 1991:19.

39. AL-JAFARI AA, KAMAL MA, GREIG NH, ALHOMIDA AS,PERRY ER. Kinetics of human erythrocyte acetylcholines-terase inhibition by a novel derivative of physostigmine:phenserine. Biochem Biophys Res Commun 1998;248:180±5.

40. IIJIMA S, GREIG NH, GAROFALO P et al. The long-actingcholinesterase inhibitor heptyl-physostigmine attenuatesthe scopolamine-induced learning impairments of rats ina 14-unit T-maze. Neurosci Lett 1992;144:79±83.

41. DURKIN TP, MESSIER C, DE BOER P, WESTERINK BHC. Raisedglucose levels enhance scopolamine-induced acetylcholineover¯ow of the hippocampus: an in vivo microdialysis studyin the rat. Behav Brain Res 1992;49:181±8.

42. MESSAMORE E, OGANE N, GIACOBINI E. Cholinesteraseinhibitor effects on extracellular acetylcholine in the ratstriatum. Neuropharmacol 1993;32:291±6.

43. TAGUCHI K, HAGIWARA Y, SUZUKI Y, KUBO T. Effects ofmorphine on release of acetylcholine in the rat striatum: anin vivo microdialysis study. Naunyn-Schmiedeberg's ArchPharmacol 1993;347:9±13.

44. MESSAMORE E, WARPMAN U, OGANE N, GIACOBINI E.Cholinesterase inhibitor effects on extracellular acetylcho-line in rat cortex. Neuropharmacol 1993;32:745±50.

45. XU M, NAKAMURA Y, YAMAMOTO T et al. Determination ofbasal acetylcholine release in vivo by rat brain dialysis with aU-shaped cannula: effect of SM-10,888, a putative ther-apeutic drug for Alzheimer's disease. Neurosci Lett1991;123:179±82.

46. MESSAMORE E, WARPMAN U, WILLIAMS E, GIACOBINI E.

Phenserine: preclinical pharmacokinetics & pharmacodynamics

83

Muscarinic receptors mediate attenuation of extracellularacetylcholine levels in rat cerebral cortex after cholinester-ase inhibition. Neurosci Lett 1993;158:205±8.

47. KOSSA T, YAMANISHI Y, OGURA H, YAMATSU K. Effect ofE2020 on the extracellular level of acetylcholine in the ratcerebral cortex measured by microdialysis without additionof cholinesterase inhibitor. Eur J Pharmacol 1990;183:1936±41.

48. MARTA M, POMPONI M. Inhibition of acetylcholinesterase bynew physostigmine derivatives. Biomed Biochem Acta1988;47:285±8.

49. GIACOBINI E. Cholinomimetic replacement of cholinergicfunction in Alzheimer's disease. In: MEYER EM, SIMPKINS

JW, YAMAMOTO J, CREW F, eds. Treatment of Dementias, aNew Generation of Progress. (Advances in Behavioral Bio-logy, vol. 40). New York, NY: Plenum Press, 1992: pp.19±34.

50. IIJIMA S, GREIG NH, GAROFALO P et al. Phenserine: aphysostigmine derivative that is a long-acting inhibitor ofacetylcholinesterase and demonstrates a wide dose-rangefor attenuating a scopolamine-induced learning impairment

of rats in a 14-unit T-maze. Psychopharmacol 1993;112:415±20.

51. MANDEL RJ, THAL LJ. Physostigmine improves water mazeperformance following nucleus basalis magnocellularis lesionin rats. Psychopharmacol 1988;96:421±5.

52. MURRAY TK, CROSS AJ, GREEN AR. Reversal by tetrahy-droaminoacridine of scopolamine-induced memory andperformance de®cits in rats. Psychopharmacol 1991;105:134±6.

53. HAROUTUNIAN V, GREIG NH, UTSUKI T, DAVIS KL, WALLACE

WC. Pharmacological modulation of Alzheimer's b-amyloidprecrusor protein levels in the CSF of rats with forebraincholinergic system lesions. Mol Brain Res 1997;46:161±8.

54. LAHIRI DK, FARLOW MR, HINTZ N, UTSUKI T, GREIG NH.Cholinesterase inhibitors, b-amyloid precursor protein andamyloid b-peptides in Alzheimer's disease. Acta NeurolScand 2000;176:60±7.

55. GREENBLATT DJ, SELLER EM, SHADER I. Drug therapy: drugdisposition in old age. N Eng J Med 1982;306:1081±8.

56. SCHMACKER DK. Aging and drug disposition: an update.Pharmacol Rev 1985;37:133±48.

Greig et al.

84