Naunyn-Schmiedeberg’s Arch Pharmacol (2000) 361 : 521–528 Digital Object Identifier (DOI) 10.1007/s002100000222

O R I G I N A L A RT I C L E

Kazuo Nakamura · Mitsue Kurasawa

Serotonergic mechanisms involved in the attentional and vigilance task performance of rats and the palliative action of aniracetam Received: 2 August 1999 / Accepted: 24 January 2000 / Published online: 8 March 2000 © Springer-Verlag 2000

Abstract Central serotonergic systems play an important role in regulating mood/emotion, cognition, sleep and wakefulness, appetite and locomotion and body temperature via multiple receptor subtypes. Among them, 5-HT1A and 5-HT2A/2C receptors have opposite effects with respect to certain functions. The aim of the present study was to compare the effects of 8-hydroxy-2-(di-n-propylamino)tetralin (8-OH-DPAT), a selective 5-HT1A receptor agonist, and 1-(2,5-dimethoxy-4-iodophenyl)-2-aminopropane (DOI), a selective 5-HT2A/2C receptor agonist, on the performance of middle-aged rats in a two-lever choice reaction task that assessed attention and vigilance functions. We also examined the effects of aniracetam, a cognition enhancer, and its major metabolites on the induced performance impairments. 8-OH-DPAT (0.3 mg/kg s.c.) reduced response speed and choice accuracy and increased response omission with a reduction of task-associated motor activity without inducing motor inability or motivational changes. These findings indicate a specific disturbance of attentional and vigilance processes. DOI caused similar impairments at the highest dose tested (3 mg/kg s.c.); at a lower dose (1 mg/kg s.c.), however, it selectively attenuated the response speed, suggesting a selective attention deficit. (–)-Alprenolol, a non-selective 5-HT1A receptor antagonist, and ritanserin, a preferential 5-HT2A receptor antagonist, blocked the 8-OH-DPAT- and DOI-induced performance impairments respectively. Aniracetam ameliorated all the performance deficits, and the metabolites N-anisoyl-GABA and 2-pyrrolidinone partially mimicked the aniracetam effect in the 8-OHDPAT-induced attentional and vigilance impairments. Nefiracetam, another cognition enhancer, improved only the 8-OH-DPAT-induced impairments. Each compound tested alone had no effect on task performance. These results in-

K. Nakamura (✉) · M. Kurasawa CNS Supporting Laboratory, Nippon Roche Research Centre, 200 Kajiwara, Kamakura 247-8530, Japan e-mail: [email protected], Fax: +81-467-472219

dicate that both serotonergic regulations, possibly via presynaptic 5-HT1A receptors and more likely via postsynaptic 5-HT2A receptors, lead similarly to attention deficits. Key words Attention and vigilance · Choice reaction task · 5-HT1A and 5-HT2A receptors · Aniracetam · Metabolites

Introduction Attention deficits and low vigilance are pivotal components of various mental dysfunctions in the elderly and in patients with cerebrovascular and neurodegenerative diseases. Such symptoms may often provide a psychopathological basis for many psychiatric syndromes, such as emotional disturbances, cognition impairments, sleep disorders and behavioural abnormalities (Lipowski 1990; Sarter 1994). Although the biochemical mechanisms underlying attentional and vigilance impairments are poorly understood, it has been hypothesized that they may be related to an abnormality in central noradrenergic and cholinergic neurotransmission (Robbins 1997). Brain serotonergic systems modulate mood and emotion, learning and memory, sleep and wakefulness, appetite, locomotion and thermoregulation via multiple receptor subtypes (Hoyer et al. 1994; Leonard 1996). The global manipulation of central serotonergic neurotransmission disrupts attentional and arousal processes. In a five-choice serial reaction time task performed by rats, serotonin (5-HT) synthesis inhibition by p-chlorophenylalanine (PCPA) induces attention deficits (Jäkälä et al. 1992). Similarly, 5-HT loss or lesion of the dorsal raphé nucleus by 5,7-dihydroxytryptamine (5,7-DHT) reduces the arousal process with different effects on attentional performance (Harrison et al. 1997a, 1997b). Moreover, 5-HT depletion induced by PCPA or dorsal raphé lesions disrupt latent inhibition (Solomon et al. 1978, 1980) that has been used to assess attentional dysfunction in various pathological conditions (Lubow 1997).

522

In contrast, specific stimulation of the 5-HT1A and 5-HT2A/2C receptors is believed to result in opposite effects on various functions, such as feeding behaviour, thermoregulation, and anxiolytic, antidepressive and antipsychotic activities (Leonard 1996). Moreover, 8-hydroxy-2-(di-n-propylamino)tetralin (8-OH-DPAT), a selective 5-HT1A receptor agonist, has no effect on latent inhibition (Cassaday et al. 1993), whereas 1-(2,5-dimethoxy4-iodophenyl)-2-aminopropane (DOI), a 5-HT2A/2C receptor agonist, disrupts it (Hitchcock et al. 1997), as does ritanserin, a preferential 5-HT2A receptor antagonist (Cassaday et al. 1993). As the results from many reports are controversial, the nature of serotonergic implications in attentional and vigilance processes remains unclear. The use of more specific ligands and appropriate evaluation models will be required to resolve these inconsistencies. Aniracetam is a clinically prescribed cognition enhancer, and its therapeutic effects on behavioural abnormalities (nocturnal delirium and wandering), sleep disorders and emotional disturbances (anxiety/agitation, depressive mood) have been demonstrated in a double-blind study with stroke patients (Otomo et al. 1991; Kumar 1999). The compound was first characterized in animal models as a learning and memory enhancer, and thought to act through mechanisms such as positive modulation of cholinergic and glutamatergic nervous systems and an increase in synaptic efficacy and energy metabolism (Himori et al. 1992; Martin and Haefely 1993; Pizzi et al. 1993). We have reported recently that both scopolamine and apomorphine cause attentional deficits and low vigilance in rats that had mastered previously a two-lever choice reaction task, and that aniracetam reverses the induced performance impairments (Nakamura et al. 1998a, 1998b). As yet, however, no studies have evaluated experimentally the interaction between aniracetam and serotonergic compounds on attention and vigilance. The present study was therefore designed to elucidate whether central 5-HT1A receptor manipulation by 8-OHDPAT disrupts the attentional and vigilance performances of middle-aged rats in the two-lever choice reaction task, and to compare these results with those obtained with DOI. We also examined the effects of aniracetam and its major metabolites on the performance impairments in the models. Nefiracetam, a structural analogue of aniracetam and under clinical development as a cognition (memory) enhancer, was evaluated similarly to assess whether there is a structure-related common action or mechanism.

Materials and methods Animals. Male Wistar rats aged 8–9 months were obtained from Charles River Japan. Animals were housed in groups of three in a room with controlled temperature (22±2 °C) and relative humidity (55±10%) illuminated from 7.00 a.m. to 7.00 p.m. The diet was restricted so that the animals’ weights were maintained at 80% of their free-feeding weight (CRF-1, Charles River Japan). The study was carefully performed in accordance with the guidelines dictated by the Animal Care and Use Committee of Nippon Roche Research Centre.

Drugs. Aniracetam (1-p-anisoyl-2-pyrrolidinone), N-anisoyl-GABA and nefiracetam were synthesized by F. Hoffman-La Roche (Basle, Switzerland). p-Anisic acid and 2-pyrrolidinone were purchased from Tokyo Kasei Kogyou (Tokyo, Japan) and Wako Pure Chemical Industries (Osaka, Japan), respectively. (–)-Alprenolol D-tartrate was obtained from Sigma (St. Louis, Mo., USA), and DOI hydrochloride, 8-OH-DPAT and ritanserin were from RBI (Natick, Mass., USA). Aniracetam, N-anisoyl-GABA, p-anisic acid, nefiracetam and ritanserin were suspended in 0.25% carboxymethylcellulose solution containing 1–2 drops of Triton X-100. 8-OH-DPAT, DOI and (–)-alprenolol were dissolved in physiological saline, and 2-pyrrolidinone was dissolved in deionized water. In the choice reaction task with 8-OH-DPAT, the inducer (0.3 mg/kg s.c.) or saline vehicle was injected 0.5 h prior to behavioural testing, while aniracetam (10, 30 or 100 mg/kg p.o.), its metabolites (10 or 30 mg/kg p.o.) and nefiracetam (3, 10 or 30 mg/kg p.o.) or vehicle were administered 0.5 h before 8-OHDPAT. (–)-Alprenolol (1 or 3 mg/kg i.p.) was given 10 min before the treatment. In the choice reaction task with DOI, rats were given DOI (1 mg/kg s.c.) or saline vehicle simultaneously with the oral test compounds or vehicle 2 h prior to behavioural testing. To examine the effect of each compound alone on the task performance, test compounds were administered with saline vehicle according to the respective treatment schedule. Compounds were administered orally in a volume of 5 ml/kg and injected s.c. or i.p. in a volume of 1 ml/kg. All compound solutions were freshly prepared for each experiment. Choice reaction task. The choice reaction apparatus consisted of 14 two-lever operant conditioning chambers (Skinner boxes) enclosed in wooden, sound-attenuating compartments. The basic equipment and interface-controller system are detailed elsewhere (Nakamura et al. 1998a). Briefly, rats were trained daily (Monday through Friday) for 1–2 months from the age of 9–10 months. They were reinforced to press correctly the appropriate lever with a continuous schedule of a fixed ratio 1, immediately after random presentation of a visual stimulus (cue lamp) above the response lever. The session onset was indicated by switching off the house-light. Animals were trained to refrain from pressing either of the two levers during the random period (2 to 5 s) of differential reinforcement of other behaviour (DRO). If they pressed the levers during the DRO period, the period was repeatedly reset within the period of 10 s for advancing to the next choice reaction. During the choice reaction period (5 to 8 s), the time between illumination of the cue lamp and the correct response was defined as the choice reaction time (CRT) and a food pellet (45 mg, Bioserv, New Jersey, USA) as a positive reinforcer was provided through the pellet dispenser. Pressing the wrong lever was counted as an incorrect response. The cue lamp remained illuminated until the rats pressed either the correct or incorrect lever during the choice reaction period. With the lever pressing or the end of time limit the house-light was switched on and an intertrial interval (ITI) period (time-out period for 30 s) began. One trial took approximately 40 s, and one session consisted of 30 trials and lasted for 20 min. Training was completed successfully when the following criteria for the behavioural measures were attained in three consecutive sessions: greater than 90% correct responses, 0.5–2.0 s CRT and fewer than 50 premature responses per session. The baseline data were obtained in this manner. The performance and behavioural measurements are summarized as follows: 1. CRT: the latency (s) between the onset of the visual stimulus and the choice/pressing of the correct lever; 2. Percentage correct (% correct): the proportion of correct responses (pressing a lever in response to the signal of the cue lamp) over the total number of responses during the choice reaction period; 3. Percentage omission (% omission): the proportion of no responses over the total number of responses during the choice reaction period.

523 To minimize data variation in the experiment for the evaluation of drug efficacy, high responders showing 30–90% correct responses and 0.7–2.5 s CRT following 8-OH-DPAT (0.3 mg/kg s.c.) or 0.7–2.5 s CRT following DOI (1 mg/kg s.c.) were selected from the trained animals and used repeatedly at 2-week intervals. The rejected animals were 11.8% for 8-OH-DPAT and 24.5% for DOI. Animals were aged 13–15 months at the beginning of the experiment. The experimental sessions with either test drugs or vehicle were always conducted on Tuesdays and Wednesdays. The experiments were repeated at least twice for confirmation of the results of the first experiment, and the data were pooled for statistical analysis. Task-associated motor activity. Motor activity related to the performance of the choice reaction task in the Skinner box was also measured during a test session (20 min) with the AB system (Neuroscience, Tokyo, Japan), whereby the infrared sensor was located on the ceiling of the box. Statistical analysis. Data were analysed by one-way ANOVA followed by a post-hoc multiple comparison using Dunnett’s t-test (2tailed) or by a Kruskal-Wallis non-parametric analysis followed by Mann-Whitney’s U-test (2-tailed). P<0.05 was considered significant. The results are represented as means±SEM.

Results Choice reaction task with 8-OH-DPAT Of the tested 8-OH-DPAT doses (0.03, 0.1 and 0.3 mg/ kg), only the highest dose caused significant prolongation of CRT [F(3, 26)=9.71, P<0.05], reduction in % correct [F(3, 26)=3.49, P<0.05] and increase in % omission [F(3, 26)=3.63, P<0.05] 0.5 h after the injection (Fig. 1). These impairments were not observed 1 h after the dosage. In contrast, the compound had no effect on premature response (the number of lever pressings in the periods of DRO and ITI). Animals ate the food pellet given as a reward after the correct response. 8-OH-DPAT dose-dependently elicited the classical 5-HT behavioural syndrome about 0.5 h after the injection (Goodwin and Green 1985). For example, hypothermia was seen in three animals at 0.03 mg/kg, in six at 0.1 mg/kg and in seven out of eight rats at 0.3 mg/kg, and flat body posture was observed in five out of eight at 0.03 mg/kg and in all animals at both 0.1 and 0.3 mg/kg. Furthermore, when task-associated motor activity was measured in the animals performing the choice reaction task in the Skinner box, 8-OH-DPAT (0.3 mg/kg) significantly reduced the motor activity at 0.5 h (Fig. 1). The change in motor activity had disappeared at 1 h [F(3, 26)=4.79, P<0.05]. We thus decided to evaluate the effects of the test substances 0.5 h after the injection of 0.3 mg/kg 8-OH-DPAT, under the conditions in which there was a behavioural change. (–)-Alprenolol, a 5-HT1A receptor antagonist, at 1 or 3 mg/kg significantly attenuated the impairments of task performance as measured by CRT [F(3, 30)=3.64, P<0.05] and % correct [F(3, 30)=3.49, P<0.05] to the same degree at both doses and also tended to reduce % omission (P<0.10) (Fig. 1). Administration of aniracetam (10–100 mg/kg) 0.5 h prior to the 8-OH-DPAT injection produced significant amelioration, which was maximal at

Fig. 1 Dose-dependent impairment of choice reaction performance of rats by 8-hydroxy-2-(di-n-propylamino)tetralin) (8-OHDPAT) and its reversal by (–)-alprenolol. Animals (n=6–8/group in the left panels; n=8–10/group in the right panels) were injected with 8-OH-DPAT or vehicle and the test session was begun 0.5 h thereafter. (–)-Alprenolol (i.p. dose in milligram/kilogram body weight indicated above the dotted line at bottom right) or vehicle was administered 0.5 h before 8-OH-DPAT treatment. Means± SEM (Choice reaction time latency between visual signal and correct performance of task, % Correct percentage of total trials in which the task was performed correctly, % Omission percentage of total trials in which no response was made, Veh vehicle) #P<0.05 vs. vehicle control; *P<0.05 and **P<0.01 vs. 8-OH-DPAT alone

30 mg/kg, in CRT [F(4, 49)=8.08, P<0.01], % correct [F(4, 49)=10.5, P<0.01] and % omission [F(4, 49)=7.79, P<0.01] compared with the performance of the group treated with 8-OH-DPAT alone (Fig. 2). The increase in % correct [F(4, 49)=10.5, P<0.05] was significant at 10 mg/ kg. Doses above 30 mg/kg did not elicit any further improvement. Of the three major metabolites of aniracetam, N-anisoyl-GABA (30 mg/kg) significantly ameliorated the 8-OH-DPAT-induced changes in % correct (P<0.05) and % omission (P<0.05) and tended to restore CRT (P<0.10) (Table 1). 2-Pyrrolidinone shortened the prolonged CRT [F(3, 45)=14.8, P<0.05] only at 10 mg/kg, whereas p-anisic acid elicited no improvements at all. The patterns of the partial improvement of performance impairments by the metabolites were consistent with that of aniracetam. Nefiracetam (3–30 mg/kg), a cognition enhancer, produced a significant but dose-independent improvement in % correct [F(4, 64)=3.49, P<0.05] and %

524

omission [F(4, 64)=4.43, P<0.05 or P<0.01] and tended to shorten the CRT (P<0.10) (Fig. 2). The administration of each compound alone at the respective active doses had no effect on the task performance (Table 2), nor did they affect either the task-associated motor activity in the absence of 8-OH-DPAT or the motor activity suppressed by 8-OHDPAT (Figs. 1 and 2 and Tables 1 and 2). The effects of the test compounds on the 8-OH-DPAT-induced 5-HT syndrome were not examined in detail but aniracetam appeared to reduce the increased urination and defecation. Choice reaction task with DOI

Fig. 2 Effects of aniracetam and nefiracetam on the 8-OH-DPATinduced performance impairment in the choice reaction task of rats. Animals (n=10–12/group in the left panels; n=12–15/group in the right panels) were injected with 8-OH-DPAT or vehicle 0.5 h after the administration of either compound (p.o. dose in milligram/kilogram body weight indicated above the dotted lines at bottom) or vehicle, and the test session was begun 0.5 h after 8-OH-DPAT treatment. Means±SEM *P<0.05 and **P<0.01 vs. 8-OH-DPAT alone

DOI dose-dependently (0.3–3 mg/kg) impaired all behavioural parameters (P<0.05 or P<0.01) measured 0.5, 1 or 2 h after the treatment. The dose-dependency was clearest at 2 h (Fig. 3). Although the highest dose caused strong and long-lasting performance impairments, the intermediate dose prolonged only CRT significantly 1 h [F(3, 37)=30 4, P<0.05] and 2 h [F(3, 37)=17.9, P<0.05] after the injection. Animals completely consumed the food pellet rewards during the ITI period. The 5-HT behavioural syndrome elicited dose-dependently by DOI differed from that induced by 8-OH-DPAT. DOI (1 mg/kg) caused hyperthermia in seven out of eight rats, head-twitch in all of six and sedation in two out of six. The highest dose produced robust sedation in all animals, and the subsequent motor depression was thought to seriously impair task performance. DOI significantly increased the task-associated motor activity (P<0.05 or P<0.01) at all times after treatment with the lowest dose, but significantly depressed it (P<0.01) at the highest dose (Fig. 3). Given that the intermediate dose had little influence on motor activity, we decided to evaluate the test compounds’ effects 2 h after the injection of 1 mg/kg DOI.

Table 1 Effects of the major metabolites of aniracetam on the 8hydroxy-2-(di-n-propylamino)tetralin) (8-OH-DPAT)-induced performance impairment in a choice reaction task performed by rats. Animals were injected with 8-OH-DPAT or vehicle 0.5 h after the administration of each metabolite or vehicle and the test session

was begun 0.5 h after 8-OH-DPAT treatment. Means±SEM (Choice reaction time latency between visual signal and correct performance of task, % Correct percentage of total trials in which the task was performed correctly, % Omission percentage of total trials in which no response was made)

Compound Vehicle 8-OH-DPAT +N-anisoyl-GABA Vehicle 8-OH-DPAT +2-Pyrrolidinone Vehicle 8-OH-DPAT +p-Anisic acid

Dose (mg/kg)

No. of animals

Choice reaction time (s)

% Correct

% Omission

Motor activity (counts/20 min)

0.3 s.c. 10 p.o. 30 p.o.

13 13 15 15

1.4±0.1** 2.2±0.1 2.1±0.2 1.9±0.1

99.2±0.6** 87.9±3.6 84.9±3.6 93.6±1.2*

0.3±0.3** 11.3±1.9 13.8±3.6 5.8±1.2*

888±55.8* 657±47.6 670±47.3 646±44.4

0.3 s.c. 10 p.o. 30 p.o.

11 11 13 14

1.2±0.1** 2.4±0.1 2.0±0.1* 2.3±0.2

97.9±0.9** 82.7±4.1 85.1±4.9 73.6±6.5

1.2±0.5** 16.4±4.0 14.4±4.8 24.1±5.9

922±49.8** 682±63.4 622±39.1 699±53.3

0.3 s.c. 10 p.o. 30 p.o.

19 17 19 19

1.3±0.1** 2.2±0.2 2.1±0.2 2.3±0.2

97.9±0.6** 74.9±5.8 66.9±6.9 69.5±5.8

1.7±0.6** 25.1±5.7 31.9±7.2 29.3±5.8

847±33.3** 563±29.7 515±40.3 527±32.0

*P<0.05 and **P<0.01 vs. 8-OH-DPAT alone

525 Table 2 Effects of test compound alone on the choice reaction performance of rats. Animals were administered with each compound or vehicle and the test session begun 1 or 2 h after treatment. Means±SEM Compound 1 h after treatment Vehicle Aniracetam (–)-Alprenolol Nefiracetam 2 h after treatment Vehicle Aniracetam Ritanserin Nefiracetam

Dose (mg/kg)

No. of animals

Choice reaction time (s)

% Correct

% Omission

Motor activity (counts/20 min)

10 p.o. 30 p.o. 3 i.p. 30 p.o.

5 6 6 6 5

1.7±0.3 1.7±0.1 1.5±0.1 1.5±0.1 2.0±0.2

98.7±1.3 93.4±2.3 95.6±1.4 97.2±1.0 90.7±1.9

0.7±0.7 2.2±1.1 1.1±1.1 1.1±0.7 6.0±2.2

912± 36.0 817±120 866± 67.4 875± 44.1 868± 45.7

10 p.o. 30 p.o. 1 p.o. 30 p.o.

5 6 6 6 5

1.5±0.2 1.6±0.1 1.5±0.1 1.5±0.1 1.9±0.2

96.7±1.1 87.8±2.7 92.8±3.3 90.6±3.3 88.7±4.8

0 ±0 6.1±2.7 1.7±1.1 1.7±0.7 4.7±3.3

950± 33.3 861±104 896± 66.4 840± 20.9 957± 56.3

Fig. 3 Dose-dependent impairment of choice reaction performance of rats by 1-(2,5-dimethoxy-4-iodophenyl)-2-aminopropane (DOI) and its reversal by ritanserin. Animals (n=10–13/group in the left panels; n=5–7/group in the right panels) were injected with DOI or vehicle and the test session was begun 2 h thereafter. Ritanserin (p.o. dose in milligram/kilogram body weight indicated above the dotted line at bottom right) was administered concomitantly with DOI. Means±SEM #P<0.05 and ##P<0.01vs. vehicle control; *P<0.05 and **P<0.01 vs. DOI alone

The preferential 5-HT2A receptor antagonist ritanserin significantly and dose-dependently (0.1–1 mg/kg) improved the CRT (Fig. 3). The improvement at 1 mg/kg

Fig. 4 Effects of aniracetam and nefiracetam on the DOI-induced prolongation of choice reaction time. Animals (n=7–12/group in the left panels; n=8/group in the right panels) were concomitantly administered with DOI and either compound (p.o. dose in milligram/kilogram body weight indicated above the dotted lines at bottom) or vehicle and the test session was begun 2 h after DOI treatment. Means±SEM *P<0.05 and **P<0.01 vs. DOI alone

was approximately 100% [F(4, 25)=3.25, P<0.05]. Aniracetam (3–30 mg/kg) significantly ameliorated the impaired CRT [F(4, 39)=4.25, P<0.05] only at 10 mg/kg (Fig. 4). The three major metabolites (10 or 30 mg/kg) were ineffective (data not shown). Nefiracetam also did not ameliorate the DOI-induced deficit at any dose (3–30 mg/kg) (Fig. 4). As described before, the test compound itself did not affect any of the behavioural measures or task-associated motor activity (Table 2). None of the com-

526

pounds changed the task-associated motor activity after the DOI treatment (Figs. 3 and 4).

Discussion It is generally believed that the stimulation of 5-HT1A receptors decreases neuronal activity, whereas 5-HT2 receptors mediate an increase in neuronal activity (Brandao et al. 1991). The present studies, however, demonstrate that the 5-HT1A agonist 8-OH-DPAT impaired attentional and vigilance performances and the 5-HT2A agonist DOI also caused attention deficits in the two-lever choice reaction performance of middle-aged rats. These results suggest that the stimulation of either 5-HT1A or 5-HT2A receptors in the brain disturbs attentional and/or vigilance processes similarly. 8-OH-DPAT impaired all measured parameters, indicating selective and sustained attention deficits and reduced vigilance and arousal. These results are in partial agreement or consistent with some previously published reports (Jäkälä et al. 1992; Evenden et al. 1995), but not others (Harrison et al. 1997a, 1997b). In the latter studies, 5,7-DHT decreased 5-HT level by 62–89% in the cerebral cortex without altering noradrenaline (NA) levels and elicited high accuracy and fast task responses, whereas in Jäkälä’s studies PCPA induced complete 5-HT depletion and partial depletion of NA (30%) and elicited low accuracy and slow task responses. With respect to this discrepancy, it can be hypothesized that in the former group of studies the intact noradrenergic systems may compensate the attention deficits normally induced by serotonergic hypofunction (Robbins 1997), whereas the second group may reflect directly the strict serotonergic dysfunction by 5-HT depletion. Thus, our results with 8-OH-DPAT seem to be consistent with those of Jäkälä et al. (1992). The 8-OH-DPAT-induced performance impairments may be explained in relation to appetite, 5-HT behavioural syndrome or anxiolytic/antidepressive effects, since each behavioural measure reflects different or combined psychological properties (attention, vigilance/arousal, food motivation) and motor abilities of the tested animals as detailed previously (Nakamura et al. 1998a). Firstly, 8OH-DPAT elicits hyperphagia (Hoyer et al. 1994; Leonard 1996). If the compound enhances feeding behaviour in the fasted rats, CRT (response speed) would probably be shortened and the % omission decreased. A prefeeding experiment, however, has shown the present task to be relatively resistant to alterations in food motivation (Nakamura et al. 1998a). Secondly, 8-OH-DPAT caused a severe 5-HT behavioural syndrome (flat body posture, reciprocal forepaw treading, salivation and defecation) and hypothermia, most of which are thought to be mediated by a direct stimulation of postsynaptic 5-HT1A receptors (Tricklebank et al. 1985), thereby reducing motor activity associated with the task performance. It is thus possible that the performance impairments result from secondary effects of such non-specific behavioural side effects. However, since all the test compounds ameliorated

the 8-OH-DPAT-induced performance impairments without reversing the decreased task-performing motor activity, the deterioration in the behavioural measures would rather seem to derive from a presynaptically inducible, specific effect of 8-OH-DPAT on attentional and vigilance processes, which is presumably independent of motor inability. Thirdly, it has been postulated that anxiety arises from hyperactivity of the stimulatory serotonergic pathways, while decreased serotonergic tone is associated with depression (Leonard 1996). If 8-OH-DPAT preferentially acts on the dorsal raphé, the stimulation of 5-HT1A somatodendritic autoreceptors would reduce serotonergic neuronal firing and 5-HT release (Sprouse and Aghajanian 1988; Wright et al. 1990). In contrast, the activation of postsynaptic 5-HT1A receptors increases wakefulness and leads to its antidepressant-like activity (Leonard 1996; Bjorvatn and Ursin 1998). We therefore tested also the effects of tandospirone, which has 5-HT1A receptor agonist properties and is used clinically as an anxiolytic, to evaluate further the role of 5-HT1A receptors in the attentional and vigilance task performance (Wieland and Lucki 1990; Shimizu et al. 1992). Tandospirone at anxiolytic and antidepressive doses (5–20 mg/kg p.o.) had no effect on the behavioural measures [CRT: 20 mg/kg 1.30±0.19 s (n=7) vs. vehicle 1.44±0.13 s (n=7) and % correct: 20 mg/kg 96.2±2.11 vs. vehicle 96.7±1.26 1 h after the treatment] or task-associated motor activity (20 mg/kg 884±31.7 counts/ 20 min vs. vehicle 795±35.8 counts/20 min) but elicited various signs of the 5-HT syndrome (K. Nakamura and M. Kurasawa, unpublished data). The findings indicate that emotional changes and the 5-HT behavioural syndrome are not causally related to either attentional and vigilance dysfunction or to the decrease in task-associated motor activity. The discrepancy in the task between 8-OH-DPAT and tandospirone may relate to differences in their preferences for presynaptic 5-HT1A receptors. 8-OHDPAT interacts with 5-HT1A receptors pre- and postsynaptically (Sprouse and Aghajanian 1988; Bjorvatn and Ursin 1998), whereas tandospirone may produce its anxiolytic (and probably antidepressant-like) efficacy postsynaptically (Shimizu et al. 1992). Aniracetam reversed the 8-OH-DPAT-induced changes in CRT, % correct and % omission. Each of the aniracetam metabolites (N-anisoyl-GABA or 2-pyrrolidinone) partially mimicked this effect. The additive or co-operative effects obtained by the combined administration of both metabolites might more completely resemble the improvement by aniracetam, but further study is needed to confirm this. The performance improvements achieved by aniracetam may be due partly to its vigilance-enhancing effects that have been shown in some animal species (Martin and Haefely 1993). Consequently, aniracetam functionally increases behavioural states (attention, vigilance and arousal), so that the subjects concentrate on the performance of the choice reaction task, thus restoring the lowered choice accuracy and response speed. Both aniracetam and nefiracetam accelerate 5-HT turnover or increase the levels of 5-HT and 5-hydroxyindolacetic acid

527

in the brain of rodents (Petkov et al. 1984; Luthman et al. 1991; Murai et al. 1991). Such serotonergic activation may counteract the reduction in neuronal activity induced by 8-OH-DPAT. There is no evidence suggesting direct interaction of aniracetam and its metabolites with 5-HT1, 5-HT2A and 5-HT3 receptor subtypes (Martin and Haefely 1993; Tanaka et al. 1998). In addition to the serotonergic mechanism, the cholinergic activation mechanism of both drugs potentially may also be involved in the improvement of attention deficits and low vigilance (Kawajiri et al. 1990; Nakamura and Shirane 1999), since reticulothalamic and cortical cholinergic projections have fundamental roles in attentional and vigilance processes (Robbins 1997; Nakamura et al. 1998a). Moreover, involvement of the serotonergic-cholinergic interaction is suggested by the following: activation of 5-HT1A receptors aggravates scopolamine-induced cognitive impairments, inhibits carbachol-stimulated phosphoinositide turnover, down-regulates muscarinic M2 receptors (Claustre et al. 1991; Riekkinen 1994) and decreases acetylcholine release (Rada et al. 1993). Thus, although our results were obtained from animal experiments, they may support the clinical efficacy of aniracetam in improving behavioural and psychiatric symptoms (i.e. sleep disturbance, delirium, wandering and abnormal rhythmicity) related to cerebrovascular and neurodegenerative diseases (Otomo et al. 1991; Senin et al. 1993; Katsunuma et al. 1998; Kumar 1999). With respect to the effects of DOI, it seems unlikely that the selective decrease in response speed was caused by motor inability and reduced food motivation, since there was no alteration in task-associated motor activity or food consumption. In addition, CRT in well-trained rats is mostly unaffected by the locomotion change elicited by scopolamine and apomorphine (Nakamura et al. 1998a, 1998b). These findings indicate that the DOI-elicited activation of 5-HT2A receptors may disturb preferentially selective attention and then impair sustained attention and vigilance/arousal functions. This suggestion is supported by recent reports that DOI-induced disruption of latent inhibition in rats is reversed by 5-HT2A receptor antagonists (Hitchcock et al. 1997) and that poor performers with low choice accuracy in an attention task had higher 5-HT utilization rate in the frontal cortex than did well-performing rats (Puumala and Sirviö 1998). In addition, Carli and Samanin (1992) have found that 5-HT2 receptor agonists (lysergic acid diethylamide and quipazine) cause a ritanserin-sensitive deterioration in choice accuracy and omission in the five-choice serial reaction time task, as seen here with the highest dose of DOI. In contrast, there is evidence suggesting presynaptic inhibition of serotonergic activity by DOI (Wright et al. 1990). Altogether, DOI may not only activate 5-HT2A receptors postsynaptically but also produce an indirect serotonergic inhibition (i.e. suppression of dorsal raphé neuronal firing and of terminal release of 5-HT in the frontal cortex) by reducing glutamatergic neuronal firing (Ashby et al. 1989).Thus, the potential glutamatergic-serotonergic interaction by DOI may mimic the presynaptic action of 8-OH-DPAT.

Aniracetam shortened the CRT prolongation induced by DOI. The vigilance-enhancing effects of aniracetam may be implicated in this action (Martin and Haefely 1993). Although 10 mg/kg aniracetam improves the CRT, it has no effect on the DOI (1 mg/kg s.c.)-induced headtwitch response in rats (Tanaka et al. 1998), whereas ritanserin (1 mg/kg p.o.) completely blocks both DOI responses. These results show that the DOI-induced attention deficits are not due to 5-HT behavioural symptoms and indicate that aniracetam can restore selectively the attention deficits, probably by modulating presynaptic sites as discussed above. The results with nefiracetam suggest that these effects were mediated by non-cholinergic and non-serotonergic mechanisms. A potential alternative mechanism might involve the positive allosteric modulation by aniracetam of ionotropic and metabotropic glutamate receptors (Martin and Haefely 1993; Pizzi et al. 1993). This effect could, at least in part, serve to reverse the DOI-induced reduction in glutamatergic synaptic transmission, such as an inhibition of both glutamate release and glutamate-induced neuronal firing (Ashby et al. 1989; Maura et al. 1991). In conclusion, we demonstrated that 8-OH-DPAT causes attention deficits and low vigilance/arousal in a choice reaction task performed by rats, as evidenced by slow response speed, low choice accuracy and high response omission. The performance impairment by DOI at a relatively low dose was evident only in response speed. These changes seem to be mainly mediated by serotonergic dysfunction via the stimulation of 5-HT1A and 5-HT2A receptors (and probably glutamatergic dysfunction for DOI). Aniracetam effectively ameliorated both 8-OHDPAT- and DOI-induced attentional perturbation, probably by facilitating either serotonergic or cholinergic neurotransmission or both, or by positively modulating glutamatergic systems. These findings strongly suggest the functional involvement of the central serotonergic systems in attention and vigilance function.

References Ashby CR Jr, Edwards E, Harkins K, Wang RY (1989) Effects of (±)-DOI on medial prefrontal cortex cells: a microiontophoretic study. Brain Res 498:393–396 Bjorvatn B, Ursin R (1998) Changes in sleep and wakefulness following 5-HT1A ligands given systemically and locally in different brain regions. Rev Neurosci 9:265–273 Brandao ML, Lopez-Garcia JA, Graeff FG, Roberts MHT (1991) Electrophysiological evidence for excitatory 5-HT2 and depressant 5-HT1A receptors on neurons of the rat midbrain tectum. Brain Res 556:259–266 Carli M, Samanin R (1992) Serotonin2 receptor agonists and serotonergic anorectic drugs affect rats’ performance differently in a five-choice serial reaction time task. Psychopharmacology 106:228–234 Cassaday HJ, Hodges H, Gray JA (1993) The effects of ritanserin, RU 24969 and 8-OH-DPAT on latent inhibition in the rat. J Psychopharmacol 7:63–71 Claustre Y, Benavides J, Scatton B (1991) Potential mechanisms involved in the negative coupling between serotonin 5-HT1A receptors and carbachol-stimulated phosphoinositide turnover in the rat hippocampus. J Neurochem 56:1276–1285

528 Evenden J, Ryan C, Palejko W (1995) The effects of repeated treatment with 8-hydroxy-2-(di-n-propylamino)tetralin (8-OHDPAT) on the lever press responding of the rat under FI and DRL schedules of food reinforcement. Psychopharmacology 120:81–92 Goodwin GM, Green AR (1985) A behavioral and biochemical study in mice and rats of putative selective agonists and antagonists for 5-HT1 and 5-HT2 receptors. Br J Pharmacol 84:743– 753 Harrison AA, Everitt BJ, Robbins TW (1997a) Central 5-HT depletion enhances impulsive responding without affecting the accuracy of attentional performance: interactions with dopaminergic mechanisms. Psychopharmacology 133:329–342 Harrison AA, Everitt BJ, Robbins TW (1997b) Doubly dissociable effects of median- and dorsal-raphé lesions on the performance of the five-choice serial reaction time test of attention in rats. Behav Brain Res 89:135–149 Himori N, Imai M, Nakamura Y (1992) Potentiating action of the cerebral insufficiency enhancer, aniracetam, on adenosine triphosphate production in rat brain mitochondria. Pharmacol Commun 2:211–216 Hitchcock JM, Lister S, Fischer TR, Wettstein JG (1997) Disruption of latent inhibition in the rat by the 5-HT2 agonist DOI: effects of MDL 100,907, clozapine, risperidone and haloperidol. Behav Brain Res 88:43–49 Hoyer D, Clarke JR, Fozard JR, Hartig PR, Martin GR, Mylecharane EJ, Saxena PR, Humphrey PPA (1994) VII. International union of pharmacology classification of receptors for 5-hydroxytryptamine (serotonin). Pharmacol Rev 46:157–203 Jäkälä P, Sirvio J, Jolkkonen J, Riekkinen P Jr, Acsady L, Riekkinen P (1992) The effects of p-chlorophenylalanine-induced serotonin synthesis inhibition and muscarinic blockade on the performance of rats in a 5-choice serial reaction time task. Behav Brain Res 51:29–40 Katsunuma H, Shimizu T, Ogawa K, Kubo H, Ishida H, Yoshihama A (1998) Treatment of insomnia by concomitant therapy with zopiclone and aniracetam in patients with cerebral infarction, cerebroatrophy, Alzheimer’s disease and Parkinson’s disease. Psychiatry Clin Neurosci 52:198–200 Kawajiri S, Taniguchi K, Sakurai T, Yamasaki T (1994) Nefiracetam enhances acetylcholine outflow from the frontal cortex: in vivo microdialysis study in the rat. J Neural Transm 98:15–22 Kumar V (1999) Post-stroke depression and treatment strategies including Aniracetam. Int J Geriat Psychopharmacol 2:40–46 Leonard BE (1996) Serotonin receptors and their function in sleep, anxiety disorders and depression. Psychother Psychosom 65: 66–75 Lipowski ZJ (1990) In: Lipowski ZJ (eds) Delirium: Acute Confusional States. Oxford University Press, New York Lubow RE (1997) Latent inhibition as a measure of learned inattention: some problems and solutions. Behav Brain Res 88: 75–83 Luthman J, Lindqvist E, Dell’anna E, Kojima H, Shiotani T, Tachizawa H, Olson L (1991) Effects of DM-9384, a pyrrolidone derivative, on ischemia-induced changes in the central monoamine systems. Pharmacol Biochem Behav 41:231–234 Martin JR, Haefely WE (1993) Pharmacology of aniracetam, a novel pyrrolidinone derivative with cognition enhancing activity. Drug Invest 5:4–49 Maura G, Carbone R, Guido M, Pestarino M, Raiteri M (1991) 5-HT2 presynaptic receptors mediate inhibition of glutamate release from cerebellar mossy fiber terminals. Eur J Pharmacol 202:185–190 Murai S, Saito H, Itoh T (1991) Acute and long-term effects of aniracetam on the levels of monoaminergic, cholinergic, and amino acidergic substrates in discrete brain regions of mice. Clin Rep 25:285–301 Nakamura K, Shirane M (1999) Activation of the reticulothalamic cholinergic pathway by the major metabolites of aniracetam. Eur J Pharmacol 380:81–89

Nakamura K, Kurasawa M, Tanaka Y (1998a) Scopolamine model of delirium in rats and reversal of the performance impairment by aniracetam. Drug Dev Res 43 (2):85–97 Nakamura K, Kurasawa M, Tanaka Y (1998b) Apomorphine-induced hypoattention in rats and reversal of the choice performance impairment by aniracetam. Eur J Pharmacol 342:127– 138 Otomo E, Hirai S, Terashi A, Hasegawa K, Tazaki Y, Araki G, Itoh E, Nishimura T, Furukawa T (1991) Clinical usefulness of aniracetam for psychiatric symptoms in patients with cerebrovascular disorders: Double-blind trial against placebo. J Clin Exp Med 156:143–187 Petkov VD, Grahovska T, Petkov VV, Konstantinova E, Stancheva S (1984) Changes in the brain biogenic monoamines of rats, induced by piracetam and aniracetam. Acta Physiol Pharmacol Bulg 10:6–15 Pizzi M, Fallacara C, Arrighi V, Memo M, Spano P (1993) Attenuation of excitatory amino acid toxicity by metabotropic glutamate receptor agonists and aniracetam in primary cultures of cerebellar granule cells. J Neurochem 61:683–689 Puumala T, Sirviö J (1998) Changes in activities of dopamine and serotonin systems in the frontal cortex underlie poor choice accuracy and impulsivity of rats in an attention task. Neuroscience 83:489–499 Rada PV, Mark GP, Hoebel BG (1993) In vivo modulation of acetylcholine in the nucleus accumbens of freely moving rats: I. inhibition by serotonin. Brain Res 619:98–104 Riekkinen P Jr (1994) 5-HT1A and muscarinic acetylcholine receptors jointly regulate passive avoidance behavior. Eur J Pharmacol 262:77–90 Robbins TW (1997) Arousal system and attentional processes. Biol Psychol 45:57–71 Sarter M (1994) Neuronal mechanisms of attentional dysfunction in senile dementia and schizophrenia: two sides of the same coin? Psychopharmacology 114:539–550 Senin U, Parnetti L, Cucinotta D, Criscuolo D, Longo A, Marini G (1993) Clinical experience with aniracetam in the treatment of senile dementia of the Alzheimer’s type and related disorders. Drug Invest 5 (Suppl. 1):96–105 Shimizu H, Tatsuno T, Tanaka H, Hirose A, Araki Y, Nakamura M (1992) Serotonergic mechanism in anxiolytic effect of tandospirone in the Vogel conflict test. Jpn J Pharmacol 59:105– 112 Solomon PR, Kiney CA, Scott DR (1978) Disruption of latent inhibition following systemic administration of para-chlorophenylalanine (PCPA). Physiol Behav 20:265–271 Solomon PR, Nichols GL, Kiernan JM, Kamor RS Kaplan LJ (1980) Differential effects of lesions in medial and dorsal raphé of the rat: latent inhibition and septohippocampal serotonin levels. J Comp Physiol Psychol 94:145–154 Sprouse JS, Aghajanian GK (1988) Responses of hippocampal pyramidal cells to putative serotonin 5-HT1a and 5-HT1b agonists: a comparative study with dorsal raphé neurons. Neuropharmacology27:707–715 Tanaka Y, Nakamura K, Kurasawa M (1998) Aniracetam attenuates the 5-HT2 receptor-mediated head-twitch response in rodents as a hallucination model. Drug Dev Res 44:131–139 Tricklebank MD, Forler C, Fozard JR (1985) The involvement of subtypes of the 5-HT1 receptors and of catecholaminergic systems in the behavioural response to 8-hydroxy 2-(di-n-propylamino)tetralin in the rat. Eur J Pharmacol 106:271–282 Wieland S, Lucki I (1990) Antidepressant-like activity of 5-HT1A agonists measured with the force swim test. Psychopharmacology 101:497–504 Wright IK, Garratt JC, Marsden CA (1990) Effects of a selective 5-HT2 agonist, DOI, on 5-HT neuronal firing in the dorsal raphé nucleus and 5-HT release and metabolism in the frontal cortex. Br J Pharmacol 99:221–222