A Multitransmitter Approach to Dementia Treatment

Ensaculin (KA-672 × HCl): A Multitransmitter Approach to Dementia Treatment

Robert Hoerr1 and Michael Noeldner2

1Clinical Research Department and 2Department of Pharmacology, Dr. Willmar Schwabe GmbH & Co., Karlsruhe, Germany

Key Words: Ensaculin—Benzopyranone—Dementia—Neurotransmitters—Me- mory—Neuroprotection—5-HT1A—5HT7 —NMDA antagonists.

ABSTRACT

Ensaculin, a novel benzopyranone substituted with a piperazine moiety, showed memory-enhancing effects in paradigms of passive and conditioned avoidance in both normal and artificially amnesic rodents. It exhibited neuroprotective activities in an NMDA toxicity model and neurotrophic effects in primary cultured rat brain cells. The compound could be characterized as a weak NMDA receptor–operated channel blocker. In receptor-binding studies, ensaculin was found to have high affinities to serotonergic

5-HT1A and 5-HT7 receptors, adrenergic á1, and dopaminergic D2 and D3 receptors. Due to its unique pharmacodynamic profile, ensaculin may have potential as an antidementia agent acting on various transmitter systems.

INTRODUCTION

Alzheimer’s disease (AD) is a neurodegenerative disorder that predominantly affects aged people. It is the most frequent cause of dementia, thus gaining increasing importance as the population continues to age. Even though extensive research has been directed towards the understanding of this devastating illness, the formal and causal interrelation- ships of its pathogenetic determinants are still an enigma. The most distinctive morphological features of AD are amyloid deposition resulting in neuritic plaques, the formation of neurofibrillary tangles consisting of hyperphosphory-

Address correspondence and reprint requests to: Dr. Robert Hoerr, Clinical Research Department, Dr. Willmar Schwabe GmbH & Co., Willmar-Schwabe-Strasse 4, 76227 Karlsruhe, Germany. Tel: +49 (721) 4005-492; E-mail: [email protected]

143 144 R. HOERR AND M. NOELDNER lated tau protein, neuronal degeneration with a subsequent loss of synapses, and general cortical atrophy. At a biochemical level, AD is characterized by impairments in neurotransmitter systems. The cholinergic system is probably the neurotransmitter system most affected. The activity of choline acetyltransferase (ChAT), the key enzyme in acetylcholine syn- thesis, is reduced by 50–85% in various cortical regions and the hippocampus of AD patients. Moreover, muscarinic receptor binding is decreased in the frontal cortex. The biochemical changes in this transmitter system, such as reductions in cortical ChAT activity and cerebrospinal fluid (CSF) cholinesterase activity, correlate with both the tangle load in the frontal cortex and the severity of dementia symptoms prior to death (8,26). The concentration of serotonin (5-HT) and its metabolite 5-hydroxyindoleacetic acid (5-HIAA) has been shown to be lowered in cortical areas and the basal ganglia by up to 37 and 54%, respectively (26,8). Norepinephrine and dopamine concentrations are reduced in cortical areas of AD brains (26,8), and the density of striatal D2 dopaminergic receptors is decreased (25). The quantitatively most important neurotransmitter system in the cerebral cortex, the glutamatergic system, is also affected by AD, with a decrease in glutamatergic N-methyl- D-aspartate (NMDA) receptors, predominantly in the frontal cortex (4,26,28). On the one hand, the physiological stimulation of the NMDA receptors has a major role in memory formation. On the other hand, over-stimulation may cause serious damage to neurons by increasing the influx of calcium ions (16). Cellular energy deficiency, which is likely to prevail in AD due to a disruption of the oxidative glucose metabolism, may render neurons particularly vulnerable to (relative) glutamatergic over-stimulation by compro- mising the channel-blocking action of magnesium ions. In regard to neuropeptides, somatostatin levels have been found to be decreased (1,26) in the cortex of AD patients. While nerve growth factor (NGF) concentrations appear to remain unchanged, NGF receptor defects probably arise (9). Possible treatment strategies for AD include inhibition of ß-amyloid peptide secretion and/or plaque formation, pre- vention of tau hyperphosphorylation and fibrillary tangling, activation of neurotransmitter systems, inhibition of NMDA-mediated neurotoxicity, oxygen free radical scavenging, and restoration of the oxidative glucose metabolism. The most advanced treatment options are based on cholinergic stimulation and are justified by the marked deficits in this neurotransmitter system and its close link to cognitive functioning. While cholinesterase inhibitors have proven to be modestly effective (2), attempts on direct muscarinic M1 receptor stimulations by agonists have been disappointing. Low-affinity blockers of the NMDA receptor-mediated ion channel and substances with antioxidant properties have yielded promising results in clinical trials. Given the limited treatment success achieved with drugs selectively directed towards one neurotransmitter system, the idea of concurrently addressing a range of the neurotransmitter systems affected by AD is intriguing. Ensaculin (KA-672 × HCl) was identified as a compound with a unique profile of pharmacodynamic effects on the central nervous system (CNS) when a series of syntheti- cally modified compounds derived from the central structure of natural coumarins were screened for their potential therapeutic value.

CNS Drug Reviews, Vol. 8, No. 2, 2002 ENSACULIN 145

Fig. 1. Chemical structure of ensaculin (KA-672.HCl).

CHEMISTRY

The chemical structure of ensaculin (KA-672, 7-methoxy-6-{3-[4-(2-methoxyphenyl)- 1-piperazinyl]propoxy}-3,4-dimethyl-2H-1-benzopyran-2-one) corresponds to a benzopyranone with a piperazine moiety (Fig. 1). It is crystallized as hydrochloride salt, which has the most favorable physicochemical properties. Ensaculin hydrochloride is a white, crystalline powder with a molecular weight of 489.02 and a melting point of about 240°C (decomposition). It usually contains 3–4% water. Its solubility in water is 3,000 ppm at room temparature. The structure of ensaculin is evident from the starting materials entering the synthetic pathway, and it has been verified by elemental analysis, infrared and mass spectrometry, 1H and 13C NMR spectroscopy, and x-ray structure analysis.

PHARMACOLOGY

Modulation of Memory Functions The antidementia activity of ensaculin was initially detected in a rat conditioned avoidance (CAR) and a mouse passive avoidance (PAR) model (15) and later confirmed in different variations of these animal models.

1. Reversal of scopolamine-induced amnesia in a passive avoidance paradigm Reversal of scopolamine-induced amnesia for passive avoidance task is widely used to screen putative cognition enhancers. The method described by De Wied (7) was used with minor modifications. In short, mice were placed in a light compartment connected by a hole to a large dark compartment. After the mice enter the dark compartment they get a foot shock through the steel grid floor. Immediately after termination of the foot shock the animals were removed from the dark compartment. The same manipulation was repeated after 24 h, 7 days, and 14 days without foot shock (retention test). The latency time

CNS Drug Reviews, Vol. 8, No. 2, 2002 146 R. HOERR AND M. NOELDNER

150

*P < 0.05 125 Student'st -test

100

(sec) 75

Iatency 50

0 0.0 0.1 0.3 1.0 mg/kg ensaculin

Fig. 2. Effects of ensaculin on passive avoidance 24 h latency in scopolamine-induced amnesia (acquisition). Values are means ± S.E.M. of at least eight animals per group.

(before entry in the dark compartment) was registered. Intact animals remember the foot shock and do not change compartments during the retention tests. After induction of amnesia (scopolamine or electroshock) animals do not remember the foot shock and enter the dark compartment within a few seconds. The tests were performed in three modifications: (1) acquisition, ensaculin was administered 1 h before the first learning session; (2) memory consolidation, ensaculin was administered immediately after the first session; (3) memory retrieval, ensaculin was administered on day 2, 1 h before the retention test. In the case of scopolamine amnesia, scopolamine hydrochloride, 1 mg/kg, was applied subcutaneously (s.c.) 30 min before the training session. In the case of electroshock amnesia, a shock was applied by cornea electrodes (50 mA, 0.2 sec) immediately after the training session. The activity profile of ensaculin in these models is summarized in Figs. 2 and 3. It antagonized scopolamine- and electroshock-induced amnesia in mice after oral administration in a dose range between 0.1 and 1 mg/kg. A U-shaped dose-response relationship was found with significantly (P < 0.05) longer latencies for the 0.3 mg/kg dose compared with drug-free controls. In summary, ensaculin caused antiamnesic effects by increasing the latency on passive avoidance response in scopolamine-induced and to a lesser degree in electroshock-induced amnesia. Some of these effects were maintained for 14 days after single administration of the drug (15).

2. Effects of ensaculin on learning ability in intact animals

In a second set of experiments, the effects of ensaculin on learning ability in non- amnesic animals were investigated. Mice were tested in the PAR test described above (without amnesia) and rats were tested in a CAR paradigm described by De Wied (6) with minor modifications. The conditioned avoidance box (Ugo Basile, Comerio, Italy) was di-

CNS Drug Reviews, Vol. 8, No. 2, 2002 ENSACULIN 147

70 *P < 0.05 Student'st -test 60

(sec) 40

Iatency 20

0 0.0 0.1 0.3 1.0 mg/kg ensaculin

Fig. 3. Effects of ensaculin on passive avoidance 24 h latency in electroshock-induced amnesia (acquisition). Values are means ± S.E.M. of at least eight animals per group. vided in two connected compartments containing steel grids for delivering the uncondi- tioned stimulus (foot shock). The conditioned stimulus (sound of a buzzer) was given 3 sec prior to the uncondi- tioned stimulus. The foot shock was terminated as soon as the animals changed compartments (escape behavior). If the rats changed during the conditioned stimulus (avoidance behavior), no shock was applied. Forty trials were carried out every day for 7 days. Ap- proximately 80% or more avoidances during 3–5 consecutive sessions served as criterion of learning in the training period. After an extinction time of 10 days, the animals were re- examined without shock application. The animals were treated with ensaculin or vehicle on all training days, 1 h prior to the test procedure or in other experiments on 5 consecutive days before or after the training period. In the CAR test, ensaculin shortened avoidance latency and increased the number of correct responses of intact rats (Fig. 4). These effects were seen even when the drug was administered before, during, or after the training period. In the PAR test, ensaculin facili- tated learning acquisition, memory consolidation, and memory retrieval in intact mice. As in amnesia experiments, some of these effects were maintained until 14 days after single administration of the agent. In summary, ensaculin increases the learning ability of rats and mice in different versions of CAR and PAR test in intact animals (15).

Neuroprotective activities

In another set of experiments, the possible neuroprotective effects of ensaculin were investigated. Because it was known that excitatory amino acids (EAAs) are involved in various neurodegenerative diseases, we were interested to know whether ensaculin in- teracts with different EAAs and, if so, in what way. In initial experiments, ensaculin seemed to be a very potent inhibitor of neurotoxicity induced by N-methyl-D-aspartate

CNS Drug Reviews, Vol. 8, No. 2, 2002 148 R. HOERR AND M. NOELDNER

Fig. 4. Effects of ensaculin on the number of correct responses by intact rats in the conditioned avoidance test. Values are means ± S.E.M. of at least eight animals per group. (* Student’s t-test P £ 0.05).

(NMDA) and DL-homocysteic acid. It inhibited the convulsions and mortality induced by intravenous (i.v.) NMDA, 25 mg/kg, with an ED50 of about 0.1–0.2 mg/kg and the convulsions and mortality induced by i.v. administration of DL-homocysteic acid, 200 mg/kg, with an ED50 of about 0.3 mg/kg (22,24). Of interest, even relatively high doses of ensaculin (up to 10 mg/kg) did not influence the neurotoxicity induced by the conventional chemoconvulsants, pentylenetetrazol and nicotine, or electroshock-induced seizures. A series of reference compounds, including NMDA-antagonists, calcium channel blockers, adrenergic antagonists, neuroleptics, conventional anticonvulsants, etc., were tested in the mouse NMDA toxicity model. Only the adrenergic á1-adrenoreceptor antagonists prazosin and urapidil and the á2-adrenoreceptor antagonist clonidine protected mice in a dose range comparable with that of ensaculin. In contrast, compounds with NMDA antagonistic properties like MK-801, dextrometho- rphan, or ketamine were less active (21). Such findings indicate that these NMDA-pro- tective effects of ensaculin are not specific NMDA receptor-mediated effects. For a more precise characterization of the presumable EAA-antagonistic properties of ensaculin, a rotation-mediated aggregating brain cell culture, first described by Honegger et al. (12), was used. It has been published that this in vitro model can detect neurotoxic effects of a series of anticonvulsant drugs (29) and EAA-induced pathological processes (5). Incubation of aggregating brain cell cultures with L-glutamate, kainate, or NMDA re-

CNS Drug Reviews, Vol. 8, No. 2, 2002 ENSACULIN 149 sults in a clear loss in the activity of the two neuronal marker enzymes, glutamate decar- boxylase (GAD) and choline acetyltransferase (ChAT), and a strong decrease in lactate dehydrogenase (LDH) content of the cells (5,20). These neurotoxic effects of the EEAs could be prevented by co-incubation with the specific noncompetitive NMDA-antagonist, (+)MK-801, whereas co-incubation with ensaculin did not influence these EAA-induced toxicities (22). The direct effects of ensaculin on the NMDA-receptor complex were investigated in patch clamp experiments in acutely isolated hippocampal neurons. In these experiments, ensaculin antagonizes NMDA responses in a voltage-dependent manner. At a holding potential of –90 mV, the IC50 value was 20 ± 7 ìM. So it seems that ensaculin is a weak NMDA receptor-operated channel blocker (17). To investigate possible radical scavenging activities of ensaculin, the agent was first incubated in a cell-free in vitro Fenton system to assess direct scavenging activity. Secondly, the effects of ensaculin on the reduction of L-glutamate-induced hydroxyl free radical formation was tested in vivo in comparison with the NMDA antagonist MK-801 (32). In the in vitro test system, ensaculin inhibited in a concentration-dependent manner the generation of hydroxyl radicals in the micromolar range. In the microdialysate study, the administration of L-glutamate produced a dose-dependent increase in radical formation measured as dihydrobenzoic acid (DHBA) concentration. Intraperitoneal (i.p.) injection of ensaculin (0.1, 1, or 10 mg/kg) reduced the L-glutamate induced elevated DHBA-levels significantly to values of about 55–71% of vehicle-treated animals. At 1 mg/kg i.p. MK-801 reduced the L-glutamate-induced high DHBA levels to values of about 62–70% of controls. In these experiments, ensaculin was about 10-fold more potent at attenuating DHBA levels than MK-801 (32). Ensaculin shows putative neurotrophic effects in primary cultured rat brain septal cells. The cells were prepared according to methods described by Matsumoto et al. (19), with minor modifications. The cells were incubated with ensaculin (100 pmol/mL) or nerve growth factor (NGF) as a reference or with both agents and the neuronal outgrowth was measured after 5 days. Compared with a control culture, the ensaculin-treated culture had a consid- erably higher density of cells with longer neurites with a higher network. The effect of ensaculin on neurite network in this model is comparable to that of NGF (14).

Interaction with Neurotransmitters

Treatment of 2-year-old rats with ensaculin (1 mg/kg/d) for a period of 3 or 6 weeks causes a significant change in 5-HT concentration in the hippocampus and striatum. The level of 5-HT and the metabolite 5-HIAA in rat hippocampus increased to 176 and 181%, respectively, compared with vehicle-treated animals. In these experiments, the brain concentration of norepinephrine, dopamine, or the metabolites DOPAC and HVA were not affected (13). Teissmann and Ferger (33) investigated the effects of ensaculin on 5-HT and dopamine levels in the cortex and striatum of rat brains. After single oral administration of ensaculin (0.1 or 1.0 mg/kg), marked decreases in the 5-HT level of the striatum and in the dopamine content of the cortex were observed. In these animals, dopamine turnover was increased in both areas, whereas the 5-HT turnover was only increased in striatal tissue. In other experiments performed in mice, single oral administration of ensaculin 0.3 mg/kg caused a significant elevation of norepinephrine and 5-HT levels in whole

CNS Drug Reviews, Vol. 8, No. 2, 2002 150 R. HOERR AND M. NOELDNER brain homogenates (15). In these experiments, ATP and glucose levels were also measured and found to remain uninfluenced by ensaculin treatment. Two hours after oral treatment of mice with ensaculin, 1 mg/kg, coronal brain sections were prepared for quantitative autoradiography studies. Ensaculin affected both 5-HT1A and 5-HT2A receptor binding in a number of brain regions. However, the slight drug-induced increase in 5-HT1A receptor binding was restricted to the cingulate, frontal, and pa- rietal cortex, as well as the hind/forelimb area and medial septum. In contrast, [3H]ketanserin binding to the 5-HT2A receptor was markedly increased by about 30–40% in nearly all cortical regions studied (including hippocampal CA1 through CA4 subfields, and the dentate gyrus), in the striatum, globus pallidus, medial septum, and horizontal diagonal band (27).

Ensaculin inhibited rat brain acetylcholinesterase (AChE) activity in vitro with an IC50 of 0.36 ìM without affecting the activity of butyrylcholinesterase or arylacylamidase. However, in microdialysis experiments in rats, no effect of ensaculin on the acetylcholine release was measured, indicating that it does not inhibit AChE under in vivo conditions. It was concluded that ensaculin is a potent AChE inhibitor, but does not reach sufficient concentrations in the extracellular fluid to exhibit this activity, apparently due to cellular sequestration (10). This conclusion was supported by in vivo experiments in which ensaculin (up to 10 mg/kg) caused no measurable cholinergic side effects like tremor or salivation. In electrophysiological experiments, the influence of ensaculin on stimulus responses in hippocampal area CA1 and long-term potentiation induced by repetitive stimulation of Schaffer collaterals was investigated. In an additional test, the possible effects of ensaculin on epileptiform activities induced by lowering Mg2+ levels were investigated. The experiments were performed in combined entorhinal cortex/hippocampal slices. Ensaculin caused a concentration-dependent increase in the amplitude of population spikes measured in the stratum pyramidale of area CA1 that does not recover upon wash-out for at least 2 h. Although ensaculin increased the amplitude of the population spikes, LTP could still be induced in the presence of this agent. However, the efficacy of high frequency stratum radiatum stimulation to induce LTP was not enhanced when stimulus intensity was adjusted to correct for substance-dependent increases in stimulus response. The low magnesium-induced epileptic discharges were not affected by ensaculin (Hoffmann, in preparation).

Receptor Binding Profile

The receptor-binding profile of ensaculin was performed by NovaScreen, USA, and Cerep laboratories, France. Of the 96 different ligands tested, ensaculin specifically dis- placed (more than 50%) 32 ligands at a concentration of 10–5 M, six ligands at a concentration of 10–6 M, and five ligands at a concentration of 10–9 M. Such data indicate that ensaculin (> 1 ìM) could induce various pharmacological and/or toxicological activities via its effects on several receptor or binding sites. The binding sites most affected by ensaculin are summarized in Table 1. It seems that ensaculin causes negative cooperative binding to D2 and á1b receptors. Perhaps the binding sites are clustered so that there are some binding sites on one molecule.

CNS Drug Reviews, Vol. 8, No. 2, 2002 ENSACULIN 151 [nM] (nH) 50 compound IC Affinity of reference Reference compound [nM] (nH) 50 IC Affinity of ensaculin -propylpiperidine. N TABLE 1. Receptor binding profile of ensaculin Guinea pig brain membranes 848 haloperidol 43.9 H]spiperoneH]spiperoneH]spiperoneH]prazosinH]prazosin Human recombinant (A9L cells)H]RX 781094 Human recombinant (CHO cells)H]OH-DPAT Human recombinant (CHO cells)H]ketanserin 16.6 (2.05) Rat submaxillary 27.7H]LSD gland (1.10) Rat liver 277H](+)-pentazocine Rat (1.04) cortical membraneH]DTG Mouse (+)butaclamol + brain 100 nM (+)butaclamol Rat cortical membrane Guinea pig brain membranes clozapine 4.7 15.6 (1.77) (2.03) Rat recombinant 9.4 (HEK (0.89) 293 cells) > 100 126 868 (0.96) 2.4 (1.61) WB 4101 > 100 9.3 (1.34) phentolamine 8 1.5 (0.97) (+)-3-PPP 5-HT methysergide spiperone 77.1 1.8 2.8 (1.38) (1.96) 3 3 3 3 3 3 3 3 3 3 3 [ [ [ [ [ [ [ [ [ [ [ (+)pentazocine Radioligand Tissue 1A 1B 2 2 3 4.4 á á á 1 2 Binding site 1A 2 7 DTg-1,3-di-O-tolylguanidine; 3-PPP-3-(3-hydroxyphenyl)- Dopamine D Dopamine D Adrenergic Adrenergic Adrenergic 5-HT 5-HT 5-HT Sigma Sigma Dopamine D

CNS Drug Reviews, Vol. 8, No. 2, 2002 152 R. HOERR AND M. NOELDNER

Functional Tests

Receptor-binding data, which showed affinity of ensaculin to serotonergic 5-HT1A, 5-HT7, adrenergic á1, and dopaminergic D2 and D3 binding sites in nanomolar range, en- couraged us to study the possible functional relevance of these receptor interactions.

First, the effects of ensaculin on 5-HT1A receptors were investigated by Winter et al. (34). They used the Drug Discrimination Test, which is a sensitive technique for studying the interoceptive states produced in animals by drugs acting at the CNS. They found that ensaculin is an effective discriminative stimulus and its stimulant effects persist for about 2 hs. Because of the possible memory-enhancing activity of ensaculin, its high affinity for

5-HT1A receptors, and the possible role of those receptors in memory, they investigated the interactions of ensaculin and the known 5-HT1A antagonistic agent WAY-100635 and the known 5-HT1A agonistic agent 8-OH-DPAT. Administration of 8-OH-DPAT, 0.2 mg/kg, completely substituted the stimulus effects induced by ensaculin (0.3 or 1.0 mg/kg). Fur- thermore, both ensaculin-appropriate responding and rate suppression induced by this dose of 8-OH-DPAT were completely antagonized by WAY-100635, whereas the stimulant effects induced by ensaculin alone were only partially antagonized by WAY-100635. This indicates that part of ensaculin-induced stimulus response is brought about by

5-HT1A receptor-mediated activity, but also the presence of stimulus elements mediated by other receptors (34). Dopaminergic functions were tested in two animal models using the dopamine-receptor agonist, apomorphine, and the dopamine receptor antagonist, haloperidol. Administration of apomorphine stimulates dopamine receptors and causes climbing behavior in mice. Ensaculin inhibited this climbing behavior in a dose-dependent manner. Although ensaculin shows a very high affinity to dopamine receptors, the doses needed to antagonize apomorphine-induced climbing effects were relatively high compared with the doses in other animal models (22). In the second experimental paradigm we investigated the possible cataleptic or anticataleptogenic activity of ensaculin (22). Catalepsy is a phenomenon generally defined as the long-term maintenance of the animal in an abnormal posture (11). This behavior response has been related to the blockade of nigrostriatal dopaminergic neurons in the brain and could be induced by neuroleptic drugs like haloperidol. In addition, there is sub- stantial evidence supporting a role for 5-HT modulation of this neuroleptic-induced catalepsy (18). Adrenergic functions were tested in different in vitro and in vivo models. The functional affinities of ensaculin were determined at subtype A in rat vas deferens, at subtype

B in mouse spleen, and at subtype D in rat aorta. Ensaculin exhibited high affinity at á1A adrenergic receptors in rat vas deferens (pA2 = 9.17) and a slightly lower affinity at rat aortic á1D adrenergic receptors (pA = 8.45), whereas its affinity at mouse spleenic á1B adrenergic receptors was least (pA = 7.96). Thus, in these experiments ensaculin appears to be selective (16-fold) for the A over the B subtype and 5-fold selective for the A over the D subtype of á1 adrenergic receptors (affinity rank orderA>D>B). The functional relevance of á1-adrenergic receptor interaction was tested in mice. By intravenous administration the á-adrenergic stimulant phenylephrine causes death of mice within a few minutes. Oral coadministration of ensaculin inhibited this phenylephrine-induced mortality with an ED50 of about 0.3 mg/kg, which indicates that ensaculin is an á-adrenergic antagonist.

CNS Drug Reviews, Vol. 8, No. 2, 2002 ENSACULIN 153

In conclusion, the pharmacological data demonstrate that ensaculin is a potential antidementia agent with a novel broad-spectrum profile including neuroprotective and memory-improving activities that could be useful in the management of AD or other dementia diseases.

TOXICOLOGY

A series of subchronic and chronic toxicity studies were performed in rats and dogs. Ensaculin fumarate was well tolerated, in spite of minor symptoms, by rats at oral daily doses of 5, 10, and 20 mg/kg administered over a 28-day period. There was no effect on food consumption, and necropsy did not reveal any findings suggestive of target organ toxicity. Reproductive tract changes found in females, above all at the higher doses, were suggestive of a synchronization of the estrus cycle towards the stage of diestrus. This may be interpreted as an exaggerated pharmacodynamic action rather than a real toxic effect and may be due to central dopamine antagonism resulting in high prolactin plasma levels. In the dog, 10 mg/kg of ensaculin proved to be the maximal tolerated dose, and some clinical signs were visible at all doses tested in a 28-day oral subchronic toxicity study (1, 3, and 10 mg/kg/day). A “hammer-like” pulse and an increase in heart rate seen in the

ECG were presumably due to the pharmacodynamic activity of the compound (á1-adrenergic antagonism). Body weight gain and food consumption were decreased at interme- diate and high dose levels. In a chronic toxicity study in rats, ensaculin was well tolerated at daily oral doses up to 13 mg/kg. An increase in body weight among female rats of the high-dose group corre- sponded to endocrine-dependent histological changes seen during the 26-week dosing period. The no-observable-effect level in this study was 0.3 mg/kg/day. In a 26-week toxicity study in dogs, the no-observable-effect level was found to be 0.3 mg/kg/day. Clinical signs and organ weight changes observed at higher doses were at least partly reversible within a 4-week, treatment-free period. No changes were seen in the reproductive tract of dogs, suggesting that the effects observed in rats are specific to high- fertility species. Ensaculin did not induce mutation in five standard strains of Salmonella typhimurium, neither in the absence nor in the presence of metabolic activation when tested up to the lower limit of toxicity. Moreover, ensaculin was not clastogenic when tested in the “in vivo” mouse micronucleus assay up to an oral dose of 40 mg/kg on 2 consecutive days, a dose equivalent to approximately 50% of the LD50 (when administered twice).

PHARMACOKINETICS

The pharmacokinetics of ensaculin have been studied in mice, rats, and humans (Table 2). All pharmacokinetic studies in humans were performed in healthy elderly subjects under nonfasting conditions (drug intake after standard breakfast).

CNS Drug Reviews, Vol. 8, No. 2, 2002 154 R. HOERR AND M. NOELDNER

TABLE 2. Pharmacokinetics of ensaculin after oral administration

AUC Levels in Cmax 0–¥ (ng/mL) (ng × h/mL) AUC T1/2 Dose plasma Tmax 0–ô Species (sd, md) brain (ng/g) (h) (ng × h/g) (n × h/mL) (h) Rat 0.5 mg/kg sd Plasma 2.33 4 17.76 3.84 Brain 28.3 8 1466 43.6 1.5 mg/kg sd Plasma 16 2 100.9 3.2 Brain 53.8 2 2107 48.7 4.5 mg/kg sd Plasma 54.7 1 291.4 4.14 Brain 101.8 1 2374 63.8 Mouse 0.5 mg/kg sd Plasma 6.2 0.25 20.7 3.88 Brain 23.1 4 688 28 1.5 mg/kg sd Plasma 26.3 0.25 64.6 3.83 Brain 47 0.5 1052 24.9 4.5 mg/kg sd Plasma 69.8 0.25 195.9 3.81 Brain 125.7 0.25 1476 25.6 Humans 5.4 mg sd Plasma 3.5 2.9 nd nd 10 mg sd Plasma 7.6 2.9 61.8 17.5 20 mg sd Plasma 35.6 2.3 202.9 13.7 (m = 19.2) (m = 149.3) 60 mg sd Plasma 25.4 5.2 255 16.1 10 mg md Plasma 9.6 2.1 43.8 8.3 20 mg md Plasma 16.8 3.0 99.5 11.1 Mean values; sd, single dose; md, multiple dose; m, median; nd, not determined.

Upon oral administration, ensaculin was readily absorbed by all three species. Plasma levels increased rapidly in mice, but more slowly in rats and humans. Pharmacokinetics was found to be linear in both animal species across the dose range tested and at least up to single and repeated doses of 20 mg in humans, while maximum plasma concentrations

(Cmax) and the areas under the concentration-time curve (AUC) appear to be less than expected after oral intake of 60 mg. The AUC of ensaculin was markedly higher in brain compared with blood in both mice and rats. After oral administration of ensaculin at 0.5 mg/kg the blood to brain ratios of drug levels were 33 and 80, respectively. The ratios declined with increasing doses, thus indicating saturation. The plasma clearance was higher in female than in male mice (1640 mL/min × kg vs. 400 mL/min × kg), whereas brain concentration seemed to be the same in both sexes. In rats, no relevant sex-related differences regarding pharmacokinetic parameters were found. More than 95% of ensaculin was bound to protein in mouse, rat, and human plasma. In human subjects after single ensaculin doses below 5.4 mg, plasma levels of the drug did not exceed the lower level of quantification (LLQ). With between-subject coefficients of variation of 58.3 and 68.4% after doses of 10 and 20 mg, respectively, pharmacokinetics of ensaculin exhibited high variability. The geometric means for dose-normalized AUC0–24 were 38.7 and 42.9 ng × h/mL following repeated doses of 10 and 20 mg, respectively

CNS Drug Reviews, Vol. 8, No. 2, 2002 ENSACULIN 155

Fig. 5. Steady state pharmacokinetics of ensaculin: mean plasma concentrations over time.

(Fig. 5), which support the assumption of dose proportionality in this range (31). Ensa- culin was almost quantitatively metabolized, with an amount of less than 1% of unchanged substance excreted in urine. Its cerebral bioavailability in humans has been de- monstrated by electroencephalography (EEG).

CLINICAL STUDIES

Three phase I clinical trials were performed to investigate the safety and tolerability as well as the pharmacokinetics of ensaculin (31). As AD is nearly exclusively associated with late life, only healthy elderly subjects aged ³50 years were enrolled in these trials. To match the real life situation as closely as possible, the drug was administered in the morning at nonfasting conditions, i.e., after a standard breakfast. At first, low doses of ensaculin were tested in a single rising dose study. Twenty-four subjects were randomly assigned to two treatment groups receiving, in a two-fold cross- over design and under double-blind conditions, either 0.2 and 0.6 mg or 1.8 and 5.4 mg ensaculin in an ascending sequence. As an internal control, placebo was given to each pa- tient on 1 of the 3 test days in a random and double-blind manner. None of the adverse events observed in this study showed dose-dependence or were for any other plausible reason considered likely to be an adverse reaction to the investigational drug. Therefore, as a second step, a maximum tolerated dose (MTD) study was conducted. Starting at 10 mg, the dose was increased stepwise over 20 and 40 mg up to 60 mg. In a random and double-blind fashion, three subjects were assigned to receive either ensaculin (n = 2) or placebo (n = 1) at each dose step. Only at 60 mg, one of the two subjects re-

CNS Drug Reviews, Vol. 8, No. 2, 2002 156 R. HOERR AND M. NOELDNER ceiving the active substance suffered moderate postural hypotension with dizziness. This was expected as a consequence of the known á1-adrenergic antagonism. To verify the maximum tolerated dose, a further nine subjects were randomly assigned to receive either 60 mg ensaculin (n = 6) or placebo (n = 3) in a double-blind manner. Moderate postural hypotension with dizziness was seen in another two subjects and dizziness was reported by one subject with no significant decrease in blood pressure. In a subsequent repeated dose study, 24 subjects were randomly assigned to receive ensaculin at once-daily doses of 10 or 20 mg (n = 9, each dose group) or placebo (n =6)in a double-blind fashion, on 14 consecutive days. Prior to the repeated-dose treatment, each subject took one single dose for the determination of single-dose kinetics. A slight decrease in mean blood pressure was observed in some subjects after a 15-day intake of 20 mg/day; however, in no case did this cause any symptoms of hypotension or ortho- static dysregulation. No effect on blood pressure was seen in the 10 mg/day dose group. There was no other kind of adverse event that showed a dose-frequency relationship. On the contrary, adverse events generally occurred more frequently in the group receiving ensaculin, 10 mg/day, suggesting that there was no causal relationship with the investigational drug.

DISCUSSION

Ensaculin is another new chemical entity that has been derived from a class of natural substances. By linking a piperazine moiety to the benzopyranone molecule, which is the central structure of naturally occurring coumarins, a novel compound was created that exhibited a unique profile of pharmacodynamic activities when subjected to the pharmacological screening process. The improvement of learning and memory by ensaculin in various animal models, the unique combination of effects of the substance on neurotransmitter systems, as well as its potential neurotrophic and neuroprotective properties, make this compound an interesting candidate for further development as a treatment of AD. In toxicology studies, ensaculin was found to have a low target organ toxicity, and it was well tolerated in daily doses up to 20 mg in clinical studies. Its pharmacokinetic cha- racteristics were in favor of once-daily administration, which is an invaluable advantage in the treatment of AD. Given the limited success of treatments directed towards the enhancement of one single neurotransmitter system, a new approach with a sort of “polyvalent” drug is a logical step forward. On the basis of the properties of the substance as described above, further clinical testing of ensaculin is certainly warranted.

REFERENCES

1. Beal MF, Uhl G, Mazurek MF, Kowall N, Martin GB. Somatostatin: Alterations in the central nervous system in neurological disorders. In: Martin GB, Barchas JD, eds. Neuropeptides in Neurological and Psychiatric Disease. New York: Raven Press, 1986;215–257.

CNS Drug Reviews, Vol. 8, No. 2, 2002 ENSACULIN 157

2. Benzi G, Moretti A. Is there a rationale for the use of acetylcholinesterase inhibitors in the therapy of Alz- heimer’s disease? Eur J Pharmacol 1998;346:1–13. 3. Anonymous. Two more Alzheimer’s drugs fail. Scrip 1999;2436:21. 4. Chalmers DT, Dewar D, Graham DI, Brooks DN, McCulloch J. Differential alterations of cortical glutamatergic binding sites in senile dementia of the Alzheimer type. Proc Natl Acad Sci USA 1990;87: 1352–1356. 5. Chatterjee S, Nöldner M. An aggregate brain cell culture model for studying neuronal degeneration and re- generation. J Neural Transm 1994;44(Suppl 1):47–60. 6. De Wied D. Inhibitory effects of ACTH and related peptides on extinction of conditioned avoidance behavior in rats. Proc Soc Exp Biol Med 1966;122:28–32. 7. De Wied D, Bohus B, van Ree I, Urban I. Behavioral and electrophysiological effects of peptides related to lipotropin (ß-LPH). J Pharmacol Exp Ther 1978;3:570–580. 8. Gsell W, Moll G, Sofic E, Riederer P. Cholinergic and monoaminergic neurotransmitter systems in patients with Alzheimer’s disease and senile dementia of Alzheimer type: A critical evaluation. In: Maurer K, ed. Dementias: Neurochemistry, Neuropathology, Neuroimaging, Neuropsychology and Genetics. Braunschweig: Vieweg, 1993;25–51. 9. Higgins GA, Mufson EJ. NGF receptor gene expression is decreased in the nucleus basalis in Alzheimer’s disease. Exp Neurol 1989;106:222–236. 10. Hilgert M, Noeldner M, Chatterjee S, Klein J. KA-672 inhibits brain acetylcholinesterase in vitro but not in vivo. Neurosci Lett 1999;263:193–196. 11. Hoffmann D, Donovan H. Catalepsy as a rodent model for detecting antipsychotic drugs with extrapyra- midal side effect liability. Psychopharmacology 1995;120:128–133. 12. Honegger P, Lenoir D, Favrod P. Growth and differentiation of aggregating foetal brain cells in a serum-free defined medium. Nature 1979;282:305–308. 13. Hoyer S. Internal report. Karlsruhe: Dr. Willmar Schwabe, 1993. 14. Klimaviciusa L, Chatterjee SS, Nöldner M, Svirskis S, Klusa V. Neurotrophic effects of KA-672 × HCl on cultured embryonic cells. Pharmacol Toxicol 1997;80(Suppl 1):28. 15. Klusa V, Germane S, Chatterjee S, Noeldner M. KA-672. HCl: A new orally active antidemntia agent. Eur J Pharm Sci 1994;2:123. 16. Kornhuber J, Weller M. Psychotogenicity and N-methyl-D-aspartate receptor antagonism: Implications for neuroprotective pharmacotherapy. Biol Psychiatry 1997;41:135–144. 17. Lishko P, Maximyuk O, Chatterjee S, Noeldner M, Krishtal O. The putative cognitive enhancer KA-672 × HCl is an uncompetitive voltage-dependent NMDA receptor antagonist. Neuro Report 1998;9:4193–4197.

18. Lucas G, Bonhomme N, De Deurwaerdère P, Le Moal M, Spampinato U. 8-OH-DPAT, a 5HT1A agonist, and ritanserin, a 5HT2A/C antagonist, reverse haloperidol-induced catalepsy in rats independently of striatal dopamine release. Psychopharmacology 1997;131:57–63. 19. Matsumoto T, Tsuda S, Nakamura S. The neurotrophic effects of ebiratide, an analog of ACTH 4-9, on cultured septal cells and aged rats. J Neural Transm Gen Sect 1995;100:1–15. 20. Nöldner M, Chatterjee SS. Excitatory amino acid induced toxicity in the aggregate cell cultures from fetal rat brain. Naunyn Schmiedeberg’s Arch Pharmacol 1993;347(Suppl):529. 21. Nöldner M, Chatterjee SS. Antagonism of N-methyl-D-aspartic acid (NMDA) induced convulsions by adrenergic, serotonergic and dopaminergic agents. Naunyn Schmiedeberg’s Arch Pharmacol 1995; 351(Suppl 1):642. 22. Nöldner M, Hauer H, Chatterjee SS. Ensaculin hydrochloride. Drugs Future 1996;21:779–781. 23. Nöldner M, Chatterjee SS. Modulation of neurotransmitter by ensaculin. Pharmacol Toxicol 1997;80(Suppl 1):38. 24. Nöldner M, Chatterjee SS. Identification of KA-672 × HCl as a new functional antagonist of N-methyl-D-aspartic acid (NMDA) induced convulsions and mortality in mice. Naunyn Schmiedebergs Arch Pharmacol 1999;359(Suppl):368. 25. Pizzolato G, Chierichetti F, Fabbri M, et al. Reduced striatal dopamine receptors in Alzheimer’s disease. 123 Single photon emission tomography study with the D2 tracer [ I]-IBZM. Neurology 1996;47:1065–1068. 26. Reinikainen KJ, Soininen H, Riekkinen PJ. Neurotransmitter changes in Alzheimer’s disease: Implications to diagnostics and therapy. J Neurosci Res 1990;27:576–586. 27. Reisner E, Noeldner M, Rossner S, Chatterjee S, Bigl V, Schliebs R. Acute effects of KA-672, a putative cognitive enhancer on neurotransmitter receptor binding in mouse brain. Nerosci Lett 1999;274:187–190. 28. Scheuer K, Maras A, Gattaz WF, Cairns N, Förstl H, Müller WE. Cortical NMDA receptor properties and membrane fluidity are altered in Alzheimer’s disease. Dementia 1996;7:210–214. 29. Schilter B, Noeldner M, Chatterjee S, Honegger P. Anticonvulsant drug toxicity in rat brain cell aggregate cultures. Toxic In Vitro 1995;47:381–386.

CNS Drug Reviews, Vol. 8, No. 2, 2002 158 R. HOERR AND M. NOELDNER

30. Skujinš A, Svirskiš Š, Noeldner M, Chatterjee S, Klusa V. Influence of KA-672 × HCl on the NMDA-induced neurotoxicity in mice: The contents of the brain monoamines. Pharmacol Toxicol 1997;80(Suppl1):48. 31. Sourgens H, Hoerr R, Biber A, Steinbrede H, Derendorf H. KA 672 × HCl, a neuronal activator against dementia: Tolerability, safety, and preliminary pharmacokinetics after single and multiple oral doses in healthy male and female volunteers. J Clin Pharmacol 1998;38:373–381. 32. Teismann P, Ferger B. Comparison of the novel drug ensaculin with MK-801 on the reduction of hydroxyl radical production in rat striatum after local application of glutamate. Brain Res 2000;857:165–171. 33. Teismann P, Ferger B. In vivo effects of the putative cognitive enhancer KA-672 × HCl in comparison with 8-hydroxy-2-(Di-N-propylamine)tetralin and haloperidol on dopamine, 3,4,-dihydroxy phenylacetic acid, serotonin and 5-hydroxyindolacetic acid levels in striatal and cortical brain regions. Prog Neuropsychophar- macol Biol Psychiatry 2000;24:337–348. 34. Winter JC, Helsley SE, Rabin RA. The discriminative stimulus effects of KA-672, a putative cognitive

enhancer: Evidence for a 5HT1A component. Pharmacol Biochem Behav 1998;60:703–707.

CNS Drug Reviews, Vol. 8, No. 2, 2002