European Journal of Pharmacology 590 (2008) 212–216

Contents lists available at ScienceDirect

European Journal of Pharmacology

j ourna l homepage: www.e lsev ie r.com/ locate /e jphar

Agmatine attenuates the disruptive effects of phencyclidine on prepulse inhibition

Erik Pålsson, Kim Fejgin, Caroline Wass, Daniel Klamer ⁎Department of Pharmacology, The Institute of Neuroscience and Physiology, Sahlgrenska Academy at University of Gothenburg, Gothenburg, Sweden

⁎ Corresponding author. Department of Pharmacologyand Physiology, Sahlgrenska Academy at University of GGothenburg, Sweden. Tel.: +46 31 7863403; fax: +46 31

E-mail address: daniel.klamer@pharm.gu.se (D. Klam

0014-2999/$ – see front matter © 2008 Elsevier B.V. Aldoi:10.1016/j.ejphar.2008.06.022

A B S T R A C T

A R T I C L E I N F O

Article history:

Agmatine, a decarboxylatio Received 8 February 2008Received in revised form 26 May 2008Accepted 5 June 2008Available online 11 June 2008

Keywords:AgmatinePrepulse inhibitionPhencyclidineSchizophreniaMouse

n product of arginine, is thought to be an important neuromodulator in themammalian brain. It is proposed to exert neuroprotective, anxiolytic and antidepressant effects. Thereceptor-binding profile of agmatine is complex and includes interaction with α2-adrenergic and imidazolineI1 receptors. Furthermore, agmatine is an NMDA-receptor antagonist and inhibits nitric oxide synthase.Prepulse inhibition (PPI) of the acoustic startle response is used as a measure of the pre-attentive informationprocessing. PPI is lowered in schizophrenia and this impairment can be mimicked in experimental animalsusing the psychotomimetic drug phencyclidine (PCP). The aim of the present study was to investigate theeffects of agmatine per se on the PPI response and the effects of agmatine pre-treatment on a PCP-induceddisruption of PPI. Agmatine administration (10, 20 and 40 mg/kg) did not change the PPI response or theacoustic startle response. However, pre-treatment with agmatine 20 mg/kg, but not agmatine 40 mg/kg,significantly attenuated a PCP (5 mg/kg)-induced disruption of the PPI response. These results emphasize thepotential role of agmatine as a neuromodulator and potential target for novel treatments for brain disorders.

© 2008 Elsevier B.V. All rights reserved.

1. Introduction

Agmatine, an amine and ionic cation, was thought to besynthesised only in lower life forms until the biosynthetic pathway(decarboxylation of arginine) was described in the mammalian brain(Li et al., 1994). The concentration of agmatine in the brain iscomparable to that of classical neurotransmitters and agmatine mightbe an important neuromodulator in mammals with a potential fornew drug development (Halaris and Plietz, 2007; Reis and Regu-nathan, 2000). The proposed effects of agmatine in the brain includeantineurotoxic, anxiolytic and antidepressant actions. These assump-tions originated from the observation that agmatine reduces neuronalloss produced by ischemia (Feng et al., 2002) or excitotoxins (Olmoset al., 1999) and the behavioural effects of agmatine in the elevated-plusmaze (Lavinsky et al., 2003), forced swimming and tail suspensiontests (Krass et al., 2008; Li et al., 2003; Zomkowski et al., 2002).

Agmatine acts onmany receptors in the central nervous system. Themost thoroughly studied receptor systems so far are the α2-adrenergic,imidazoline I1 and theNMDA receptors (Li et al.,1994;Olmos et al.,1999;Yang and Reis, 1999). In addition to the activity at these receptors,agmatine irreversibly inhibits neuronal nitric oxide synthase and downregulates inducible nitric oxide synthase (Demady et al., 2001; Galea

, The Institute of Neuroscienceothenburg, POB 431, SE 405 307863284.er).

l rights reserved.

et al., 1996). This interaction might have important functional conse-quences with respect to the action of agmatine in the brain (Halaris andPlietz, 2007). Interestingly, arginine decarboxylase (ADC) synthesisesagmatine from arginine, thus directly competing with nitric oxidesynthase for substrate availability.

The highest concentrations of agmatine are found in brain regionssuch as hippocampus, hypothalamus and the cortex (Iyo et al., 2006;Otake et al., 1998; Reis et al., 1998). Hence, agmatine containing neu-rons are located in areas of the brain that modulate e.g. visceral andneuroendocrine control, processing of emotions, pain perception andcognition (Halaris and Plietz, 2007; Reis and Regunathan, 2000).

Pre-attentive information processing is thought to be importantfor selective and efficient processing of sensory information and forcoherent cognitive operations. Pre-attentive information processingcan be assessed by e.g. the prepulse inhibition (PPI) of acoustic startleresponse model. PPI is the reduction in reflex response to an intensestimulus when this stimulus is immediately preceded (30–500 ms) bya weak prestimulus (Graham, 1975; Hoffman and Ison, 1980). Thepresentation of the prestimulus, although not strong enough to elicit ameasurable response by itself, evokes a short lasting sensory gatingprocess in the brain, which is manifested by the attenuated responseto the successive intense stimulus. PPI is hypothesized to reflect theability of the central nervous system to filter out distracting andirrelevant stimuli before they reach the cortex. Hence, impairments inPPI could cause deficits in inhibitory mechanisms that filters or gatesintrinsic and extrinsic stimuli prior to higher-order processing. Intheory, deficits in pre-attentive information processing can result insensory flooding and a subsequent increased distractibility and

213E. Pålsson et al. / European Journal of Pharmacology 590 (2008) 212–216

cognitive fragmentation. This, in turn, may underlie many of thesymptoms observed in e.g. schizophrenia (Braff et al., 1978; Braff,1993; McGhie and Chapman, 1961). The reduction in PPI and higher-order cognitive impairments demonstrated in patients with schizo-phrenia can be mimicked to some extent in laboratory animals byadministering the psychotomimetic drug phencyclidine (PCP). PCP-induced psychosis also includes positive and negative symptoms(Allen and Young,1978; Javitt and Zukin,1991; Jentsch and Roth,1999;Lubyet al.,1959). These observations have established the use of PCP asa drug-induced model of schizophrenia, possibly unmasking some ofthe neurobiological alterations behind the symptoms manifested inpatients with schizophrenia (Farber, 2003; Thornberg and Saklad,1996).

Based on the described receptor-binding profile of agmatine wehypothesized that it could ameliorate PCP-induced impairments inpre-attentive information processing. The first aim of this studywas toinvestigate the effects of agmatine on pre-attentive information pro-cessing, assessed by the PPI of acoustic startle model. Secondly, weexamined the effects of agmatine on PCP-induced disruption of the PPIresponse.

2. Materials and methods

2.1. Animals

Male NMRI mice (Charles River, Sulzfeld, Germany), 30–40 g,were used. The mice arrived at the animal facility one week prior tothe start of the experiments. They were housed 6–8 per cage(37×21×15 cm) in a colony room under constant temperature (20±1 °C) and humidity (55±5%). Food (B&K Feeds) and tap water wereavailable ad libitum. The daylight cycle was maintained artificially(dark 1800–0600 h). Experiments were performed during the lighthours. The Ethics Committee for Animal Experiments, Gothenburg,Sweden, approved the experimental procedures used in this study.

2.2. Drugs

Agmatine (Sigma-Aldrich, USA) and phencyclidine (1-(1-phenyl-cyclohexyl) piperidine HCl) (RBI, Natick, USA) were used. The drugswere dissolved in saline (0.9% NaCl) and administered intraperitone-ally (i.p.) in a volume of 10 ml/kg. The used dose of agmatine wasbased on published literature (Halaris and Plietz, 2007). The dose ofphencyclidine (5 mg/kg) used to disrupt prepulse inhibition in themice tested was obtained from previous findings (Bird et al., 2001;Klamer et al., 2001, 2004).

2.3. Prepulse inhibition experiments

2.3.1. ApparatusAcoustic startle was recorded by a MOPS 2b startle response record-

ing system (Metod och Produkt, Svenska AB, Göteborg, Sweden). Eachmouse was placed in a small wire-mesh cage (5.5×10×5.5 cm) made ofstainless steel, which was suspended at one point at the top to a pistonin such a way that it could freely move under the piston. A suddenmovement of the mouse inside the cage caused a displacement of thepiston, the acceleration ofwhichwas converted to an analogue signal bya moving coil transducer. The signal was sampled and digitalised with a12-bit digital resolution by a microcomputer, which also served tocontrol the delivery of acoustic stimuli. Startle amplitudewas defined asthe maximum signal amplitude (digital units) that occurred during thefirst 40 ms after the delivery of the startle-eliciting stimulus. This timeperiodwas considered to cover the response latency inmice sufficiently,which was always shorter than 15 ms. The cage was housed in dimly litand sound-attenuated enclosure (52×42×38 cm). Three cages wereused simultaneously and a mouse tested in one cage was always testedin the same cage at subsequent tests. The acoustic signal consisted of

white noise delivered to themouse by two high frequency loudspeakersbuilt into the ceiling of the enclosure. A continuous signal provided awhite background noise level of 62 dB inside the enclosure. This signalwas interrupted at stimuluspresentationbya burst ofwhite noisewith arise/decay time of less than 1 ms.

2.3.2. Testing procedureThe animals were first placed in the startle cages for a 10 min

adaptation period. After this period, they were presented with aseries of five startle pulse-alone trials followed by a series of fiveprepulse-alone trials. The pulse-alone trials served only to accom-modate the animals to the sudden change in stimulus conditions andwere omitted from the data analysis and the prepulse-alone trialswere analysed only to ensure that these stimuli did not evokeany startle responses on their own. Thereafter the animals werepresented, three times repeatedly, with a series of five prepulse-pulsetrials followed by a series of five pulse-alone trials, i.e., a total of30 trials. The time between trials was always 10 s and the timebetween any series of trials was 70s. Startle pulse intensity was set to105 dB and prepulse intensity to 70 dB. The prepulse was 60 ms induration and presented immediately before the startle pulse, whichwas 20 ms in duration. The startle pulse was set to 105 dB, since thisintensitywas found to evoke a robust startle amplitude that showed aminimum of habituation and at the same time did not cause a ceilingeffect.

Similarly, prepulse intensity was set to 70 dB (8 dB above back-ground noise) to produce a robust prepulse inhibition.

2.3.3. Drug treatmentTesting started one week after the arrival of the animals from the

breeder the first test being a pre-test with no drug treatment. Themice were then tested every third day using a semi-randomisedcrossover design, each animal receiving all treatments of an experi-ment. Each mouse was used in one experiment only.

2.3.3.1. Experiment 1— dose response trial for agmatine. Agmatine (0,2.5, 10 or 40 mg/kg) was administered 10 min prior to a secondinjection of saline (10 ml/kg), which was administered 15 min prior tothe presentation of the first startle stimulus (n=15).2.3.3.2. Experiment 2 — interaction of agmatine (20 mg/kg) on phencycli-dine (5.0mg/kg). Agmatine (0 or 20mg/kg) was administered 10minprior to a second injection of either saline (10 ml/kg) or PCP (5 mg/kg),which was administered 15 min prior to the presentation of the firststartle stimulus (n=19).2.3.3.3. Experiment 3 — interaction of agmatine (40 mg/kg) on phencycli-dine (5.0mg/kg). Agmatine (0 or 40mg/kg) was administered 10minprior to a second injection of either saline (10 ml/kg) or PCP (5 mg/kg),which was administered 15 min prior to the presentation of the firststartle stimulus (n=18).

2.3.4. Data analysisThe mean response amplitude for pulse-alone trials (P) was cal-

culated for each mouse and treatment condition. This measure wasused in the statistical analysis to assess drug-induced changes instartle reactivity. The mean response amplitude for prepulse-pulsetrials (PP) was also calculated and used to express the percent pre-pulse inhibition according to the following formula:

Prepulse inhibition kð Þ ¼ 100 − PP=Pð Þ⁎100½ �:

Using this formula, a 0% value denotes no difference betweenpulse-alone and prepulse-pulse response amplitudes and conse-quently no prepulse inhibition. Statistical analysis was performedusing a one-way (Experiment 1) or two-way (Experiments 2 and 3)repeated measures ANOVAwith pre-treatment (Experiments 2 and 3)and treatment as fixed factors followed by Bonferroni's post hoc test to

Fig. 1. The effect of agmatine (AGM, 0, 2.5, 10 and 40 mg/kg, i.p.) on prepulse inhibitionin mice. The results are represented by the mean values (%)±SEM of 15 mice testedevery third day receiving all treatments of the experiment in a semi-randomised order.

Fig. 2. The effect of agmatine (AGM, 20 mg/kg, i.p.) on a phencyclidine-induced (PCP,5 mg/kg, i.p.) disruption of prepulse inhibition in mice. The results are represented bythe mean values (%)±SEM of 19 mice tested every third day receiving all treatments ofthe experiment in a semi-randomised order. ★ Pb0.05, ★★★ Pb0.001 compared tosaline (SAL) treatment unless otherwise indicated (two-way repeated measures ANOVAfollowed by Bonferroni's test for comparisons between treatments).

214 E. Pålsson et al. / European Journal of Pharmacology 590 (2008) 212–216

assess differences between treatment groups. Two-tailed levels ofsignificance were used and Pb0.05 was considered statisticallysignificant.

3. Results

3.1. Experiment 1 — agmatine has no effect on prepulse inhibition

Fig. 1 and Table 1 illustrate that none of the agmatine doses used,2.5, 10 and 40 mg/kg, had any significant effect on PPI of acousticstartle (one-way ANOVA, effect of treatment, F (1,14=0.18, P=0.909)or the acoustic startle response (one-way ANOVA, effect of treatment,F (1,14) 0.633, P=0.608) compared to the saline treated group.

3.2. Experiment 2 — low dose agmatine (20 mg/kg) partially blocksphencyclidine-induced deficit in prepulse inhibition

As displayed in Fig. 2, statistical analysis showed a significant effectof PCP (5 mg/kg) treatment in the data set (two-way repeatedmeasures ANOVA, effect of treatment, F (1,18)=57.024, Pb0.001). Inaddition, a close to significant effect of agmatine (20 mg/kg) pre-treatment was demonstrated (two-way repeated measures ANOVA,effect of pre-treatment, F (1,18)=4.215, P=0.055). Further analysis alsorevealed a significant pre-treatment×treatment interaction effectsuggesting that agmatine at this dose blocked the effects of PCP (two-way repeated measures ANOVA, pre-treatment×treatment, F (1,18)=11.288, Pb0.01). Post hoc comparisons showed that mice treated withPCP had a decrease in their PPI response compared to both saline+saline Pb0.001 and agmatine+PCP Pb0.05, Bonferroni's test. Addi-tionally, agmatine+PCP treated mice showed a decrease in their PPIresponse compared to saline+saline treated mice (Pb0.01, Bonferro-ni's test). This suggests a partial blockade by agmatine (20 mg/kg) ofthe PCP-induced decrease in PPI. The acoustic startle response wassignificantly decreased by PCP treatment (two-way repeated mea-sures ANOVA, effect of treatment, F (1,18)=19.798, Pb0.01). Thefollowing post hoc test showed that the saline+PCP treated group

Table 1The acoustic startle response (ASR, pulse-alone response) in Experiment 1, Experiment 2and Experiment 3

Experiment 1 SAL+SAL Agmatine 2.5+SAL Agmatine 10+SAL Agmatine 40+SALASR 328±57 287±46 290±46 280±35Experiment 2 SAL+SAL Agmatine 20+SAL SAL+PCP Agmatine 20+PCPASR 263±30 254±29 141±19b 177±22Experiment 3 SAL+SAL Agmatine 40+SAL SAL+PCP Agmatine 40+PCPASR 391±76 356±65 210±40a 295±46

The results are represented by means (arbitrary digital units)±SEM. aPb0.05, bPb0.01compared to saline (SAL) treatment (statistically significant ANOVA followed byBonferroni's test for comparisons between treatments).

displayed a lowered acoustic startle response compared to the saline+saline treated group (Pb0.01, Bonferroni's test).

3.3. Experiment 3 — high dose agmatine (40 mg/kg) does not blockphencyclidine-induced deficit in prepulse inhibition

The effect of agmatine (40 mg/kg) on a PCP (5 mg/kg)-induceddeficit in PPI was similar, albeit less pronounced than the lower doseof agmatine (see Fig. 3). An effect of treatment (two-way repeatedmeasures ANOVA, effect of treatment, F (1,17)=59.841, Pb0.001) wasdemonstrated together with a significant pre-treatment+treatmentinteraction (two-way repeated measures ANOVA, pre-treatment×treatment, F (1,17)=5.249, Pb0.05). The post hoc comparisonsconfirmed the decrease in PPI following PCP treatment compared tocontrols (saline+saline Pb0.001, Bonferroni's test). However, theagmatine+PCP group also differed significantly in PPI responsecompared to the saline+saline treated group (P=0.01, Bonferroni'stest) whereas this group did not differ significantly from the saline+PCP group (P=0.395, Bonferroni's test). This could be interpreted as atrend of agmatine (40 mg/kg) towards blocking the PCP-induceddisruption of PPI. In analogy to Experiment 2, PCP lowered the acousticstartle response (two-way repeated measures ANOVA, effect oftreatment, F (1,17)=8.168, Pb0.05). In addition, this effect wascounteracted by agmatine as evidenced by a significant pre-treat-ment× treatment interaction effect (two-way repeated measuresANOVA, F (1,17)=11.648, Pb0.01). The post hoc analysis verified theeffect of saline+PCP compared to saline+saline treatment (Pb0.01,Bonferroni's test).

Fig. 3. The effect of agmatine (AGM, 40 mg/kg, i.p.) on a phencyclidine-induced (PCP,5 mg/kg, i.p.) disruption of prepulse inhibition in mice. The results are represented bythe mean values (%)±SEM of 18 mice tested every third day receiving all treatments ofthe experiment in a semi-randomised order. ★★ Pb0.01, ★★★ Pb0.001 compared tosaline (SAL) treatment unless otherwise indicated (two-way repeated measures ANOVAfollowed by Bonferroni's test for comparisons between treatments).

215E. Pålsson et al. / European Journal of Pharmacology 590 (2008) 212–216

4. Discussion

The results demonstrate that agmatine per se has no effect on pre-attentive information processing, assessed by the PPI of acoustic startlemodel. Interestingly, a low dose of agmatine (20 mg/kg) partiallyblocked the PCP-induced disruption of the PPI response, while a higherdose of agmatine (40 mg/kg) had a less pronounced effect on PCP-induced impairments on pre-attentive information processing. Thistendency for agmatine to block the effects of PCP in an inverted U-shaped dose response curve indicates thatwe likely are in the right doserange for agmatine. Moreover, U-shaped dose response curves foragmatine have been reported previously (Aricioglu and Altunbas, 2003;Feng et al., 2002). However, this interpretation of the present results iscomplicated by thedifference in thedegreeof PCP-induceddisruption ofPPI in Experiments 2 and 3. Thus, the apparent difference in effectbetween the 20 mg/kg and 40 mg/kg doses of agmatine may besecondary to a difference in the effect of PCP (see Figs. 2 and 3).

The reason for the variation in the PCP-response can only bespeculated on but may be due to a difference in sensitivity to PCPbetween the two batches of animals.

Moreover, PCP decreased startle reactivity (pulse-alone response)in this study. It should be noted that previous studies have shown nocorrelation between startle response amplitude and the level of PPI innon-treated rodents (Klamer et al., 2005a; Paylor and Crawley, 1997).These observations suggest that startle response and prepulseinhibition are mediated by independent neuronal mechanisms.Interestingly, agmatine did not have any effect by itself on startleresponse, but attenuated the effects of PCP on this measure.

The receptor-binding profile of agmatine is very complex. It hasbeen demonstrated that the effects of agmatine likely involve inter-actions with the α2-adrenergic and imidazoline I1 receptors as well asantagonism of the NMDA receptor (Li et al., 1994; Piletz et al., 1995;Reis et al., 1998; Yang and Reis, 1999). It should be noted that phar-macological studies using NMDA-receptor antagonists generally showa lowered PPI (Curzon and Decker, 1998; Klamer et al., 2005c), sug-gesting that agmatine might be expected to lower PPI. However, thepotency of agmatine as a NMDA-receptor antagonist may be too low(Yang and Reis, 1999) to result in a disrupted PPI response, at least inthe present dose interval. In addition to the receptor interactionsdiscussed above, agmatine has been shown to inhibit nitric oxidesynthase activity (Demady et al., 2001; Galea et al., 1996). As men-tioned before, both agmatine and nitric oxide are products of argininemetabolism and several studies indicate that the inhibition of nitricoxide synthase is involved in the behavioural effects of agmatine (Fenget al., 2002; Ruiz-Durantez et al., 2002). Furthermore, nitric oxidesynthase inhibitors have been demonstrated to normalize PCP-induced impairments, not only in PPI (Johansson et al., 1997; Klameret al., 2001;Wiley,1998), but also in higher cognitive functions such aslatent inhibition (Klamer et al., 2005b) and working memory (Wasset al., 2006). Themechanismwhereby agmatine exerts its action in thePCP model may depend on a combination of serotonergic, adrenergicand nitrinergic system interactions.

In summary, agmatine demonstrates an interesting profile whentested in the PCP model of schizophrenia, partially blocking a deficit inpre-attentive sensory information processing. However, this study sug-gests that agmatine does not alter the PPI response per se in mice.Further studies of the potential role of agmatine in psychiatric diseasemay answer important questions regarding pathophysiology and therole of argininemetabolism in brain function. Thus, the agmatine systemmay beof future interest both as a novel treatment target and as a part ofthe pathophysiological mechanisms underlying several brain disorders.

Acknowledgements

This research was supported by grants from Swedish MedicalResearch Council (4247), Wilhelm och Martina Lundgrens Vetenskaps-

fond, Adlerbertska Forskningsstiftelsen, the Swedish Society of Medi-cine, Åke Wibergs Stiftelse, the Swedish Society for Medical Research,the Lundbeck Foundation, Åhlen-stiftelsen, Lars Hiertas Minne andStiftelsen Tornspiran.

References

Allen, R.M., Young, S.J., 1978. Phencyclidine-induced psychosis. Am. J. Psychiatry 135,1081–1084.

Aricioglu, F., Altunbas, H., 2003. Is agmatine an endogenous anxiolytic/antidepressantagent? Ann. N. Y. Acad. Sci. 1009, 136–140.

Bird, D.C., Bujas-Bobanovic, M., Robertson, H.A., Dursun, S.M., 2001. Lack of phencyclidine-induced effects in mice with reduced neuronal nitric oxide synthase. Psychopharma-cology (Berl) 155, 299–309.

Braff, D., Stone, C., Callaway, E., Geyer, M., Glick, I., Bali, L., 1978. Prestimulus effects onhuman startle reflex in normals and schizophrenics. Psychophysiology 15, 339–343.

Braff, D.L., 1993. Information processing and attention dysfunctions in schizophrenia.Schizophr. Bull 19, 233–259.

Curzon, P., Decker, M.W., 1998. Effects of phencyclidine (PCP) and (+)MK-801 on sen-sorimotor gating in CD-1 mice. Prog. Neuropsychopharmacol. Biol. Psychiatry 22,129–146.

Demady, D.R., Jianmongkol, S., Vuletich, J.L., Bender, A.T., Osawa, Y., 2001. Agmatineenhances the NADPH oxidase activity of neuronal NO synthase and leads to oxidativeinactivation of the enzyme. Mol. Pharmacol. 59, 24–29.

Farber, N.B., 2003. The NMDA receptor hypofunction model of psychosis. Ann. N. Y.Acad. Sci. 1003, 119–130.

Feng, Y., Piletz, J.E., Leblanc, M.H., 2002. Agmatine suppresses nitric oxide productionand attenuates hypoxic-ischemic brain injury in neonatal rats. Pediatr. Res. 52,606–611.

Galea, E., Regunathan, S., Eliopoulos, V., Feinstein, D.L., Reis, D.J., 1996. Inhibition ofmammalian nitric oxide synthases by agmatine, an endogenous polyamine formedby decarboxylation of arginine. Biochem J. 316 (Pt 1), 247–249.

Graham, F.K., 1975. The more or less startling effects of weak prestimulation. Psy-chophysiology 12, 238–248.

Halaris, A., Plietz, J., 2007. Agmatine: metabolic pathway and spectrum of activity inbrain. CNS Drugs 21, 885–900.

Hoffman, H.S., Ison, J.R., 1980. Reflex modification in the domain of startle: I. Some em-pirical findings and their implications for how the nervous system processes sensoryinput. Psychol. Rev. 87, 175–189.

Iyo, A.H., Zhu, M.Y., Ordway, G.A., Regunathan, S., 2006. Expression of arginine decar-boxylase in brain regions and neuronal cells. J. Neurochem. 96, 1042–1050.

Javitt, D.C., Zukin, S.R., 1991. Recent advances in the phencyclidine model of schizo-phrenia. Am. J. Psychiatry 148, 1301–1308.

Jentsch, J.D., Roth, R.H., 1999. The neuropsychopharmacology of phencyclidine: fromNMDA receptor hypofunction to the dopamine hypothesis of schizophrenia. Neuro-psychopharmacology 20, 201–225.

Johansson, C., Jackson, D.M., Svensson, L., 1997. Nitric oxide synthase inhibition blocksphencyclidine-induced behavioural effects on prepulse inhibition and locomotoractivity in the rat. Psychopharmacology (Berl) 131, 167–173.

Klamer, D., Engel, J.A., Svensson, L., 2001. The nitric oxide synthase inhibitor, L-NAME,blocks phencyclidine-induced disruption of prepulse inhibition in mice. Psycho-pharmacology (Berl) 156, 182–186.

Klamer, D., Engel, J.A., Svensson, L., 2004. Phencyclidine-induced behaviours in miceprevented by methylen blue. Basic Clin. Pharmacol. Toxicol. 94 (2), 65–72.

Klamer, D., Pålsson, E., Fejgin, K., Zhang, J., Engel, J., Svensson, L., 2005a. Activation of anitric oxide sensitive cAMP pathway with phencyclidine: elevated hippocampalcAMP levels are temporally associated with deficits in prepulse inhibition.Psychopharmacology (Berl) 179, 479–488.

Klamer, D., Pålsson, E., Wass, C., Archer, T., Engel, J., Svensson, L., 2005b. Antagonism ofthe nitric oxide synthase inhibitor, L-NAME, of the effects of phencyclidine on latentinhibition in taste aversion conditioning. Behav. Brain Res. 161, 60–68.

Klamer, D., Zhang, J., Engel, J.A., Svensson, L., 2005c. Selective interaction of nitric oxidesynthase inhibition with phencyclidine: behavioural and NMDA receptor bindingstudies in the rat. Behav. Brain Res. 159, 95–103.

Krass, M., Wegener, G., Vasar, E., Volke, V., 2008. Antidepressant-like effect of agmatineis not mediated by serotonin. Behav. Brain Res. 188 (2), 324–328.

Lavinsky, D., Arteni, N.S., Netto, C.A., 2003. Agmatine induces anxiolysis in the elevatedplus maze task in adult rats. Behav. Brain Res. 141, 19–24.

Li, G., Regunathan, S., Barrow, C.J., Eshraghi, J., Cooper, R., Reis, D.J., 1994. Agmatine: anendogenous clonidine-displacing substance in the brain. Science 263, 966–969.

Li, Y.F., Gong, Z.H., Cao, J.B., Wang, H.L., Luo, Z.P., Li, J., 2003. Antidepressant-like effect ofagmatine and its possible mechanism. Eur. J. Pharmacol. 469, 81–88.

Luby, E.D., Cohen, B.D., Rosenbaum, G., Gottlieb, L.S., Kelly, R., 1959. Study of a new schizo-phrenomimetic drug-sernyl. Arch. Neurol. Psychiatry 81, 363–369.

McGhie, A., Chapman, J., 1961. Disorders of attention and perception in early schi-zophrenia. Br. J. Med. Psychol. 34, 103–116.

Olmos, G., DeGregorio-Rocasolano, N., Paz Regalado, M., Gasull, T., Assumpcio Boronat,M., Trullas, R., Villarroel, A., Lerma, J., Garcia-Sevilla, J.A., 1999. Protection byimidazol(ine) drugs and agmatine of glutamate-induced neurotoxicity in culturedcerebellar granule cells through blockade of NMDA receptor. Br. J. Pharmacol. 127,1317–1326.

Otake, K., Ruggiero, D.A., Regunathan, S., Wang, H., Milner, T.A., Reis, D.J., 1998. Regionallocalization of agmatine in the rat brain: an immunocytochemical study. Brain Res.787, 1–14.

216 E. Pålsson et al. / European Journal of Pharmacology 590 (2008) 212–216

Paylor, R., Crawley, J.N., 1997. Inbred strain differences in prepulse inhibition of themouse startle response. Psychopharmacology (Berl) 132, 169–180.

Piletz, J.E., Chikkala, D.N., Ernsberger, P., 1995. Comparison of the properties of agmatineand endogenous clonidine-displacing substance at imidazoline and alpha-2adrenergic receptors. J. Pharmacol. Exp. Ther. 272, 581–587.

Reis, D.J., Regunathan, S., 2000. Is agmatine a novel neurotransmitter in brain? TrendsPharmacol. Sci. 21, 187–193.

Reis, D.J., Yang, X.C., Milner, T.A., 1998. Agmatine containing axon terminals in rathippocampus form synapses on pyramidal cells. Neurosci. Lett. 250, 185–188.

Ruiz-Durantez, E., Ruiz-Ortega, J.A., Pineda, J., Ugedo, L., 2002. Effect of agmatine on locuscoeruleus neuron activity: possible involvement of nitric oxide. Br. J. Pharmacol.135,1152–1158.

Thornberg, S.A., Saklad, S.R., 1996. A review of NMDA receptors and the phencyclidinemodel of schizophrenia. Pharmacotherapy 16, 82–93.

Wass, C., Archer, T., Palsson, E., Fejgin, K., Alexandersson,A., Klamer, D., Engel, J.A., Svensson,L., 2006. Phencyclidine affects memory in a nitric oxide-dependent manner: workingand reference memory. Behav. Brain Res. 174, 49–55.

Wiley, J.L., 1998. Nitric oxide synthase inhibitors attenuate phencyclidine-induceddisruption of prepulse inhibition. Neuropsychopharmacology 19, 86–94.

Yang, X.C., Reis, D.J., 1999. Agmatine selectively blocks the N-methyl-D-aspartatesubclass of glutamate receptor channels in rat hippocampal neurons. J. Pharmacol.Exp. Ther. 288, 544–549.

Zomkowski, A.D., Hammes, L., Lin, J., Calixto, J.B., Santos, A.R., Rodrigues, A.L., 2002.Agmatine produces antidepressant-like effects in two models of depression inmice. NeuroReport 13, 387–391.