at SciVerse ScienceDirect

Neuropharmacology 63 (2012) 958e965

Contents lists available

Neuropharmacology

journal homepage: www.elsevier .com/locate/neuropharm

Are biological actions of neurokinin A in the adult brain mediated by a cross-talkbetween the NK1 and NK2 receptors?

Ulrike Tauer a,1, Yi Zhao b,1, Steven P. Hunt c, Juraj Culman a,*

a Institute of Experimental and Clinical Pharmacology, University Hospital of Schleswig-Holstein, Campus Kiel, Hospitalstrasse 4, 24105 Kiel, GermanybDepartment of Nuclear Medicine, Molecular Image, Diagnostics and Therapy, University Hospital of Schleswig-Holstein, Campus Kiel, 24105 Kiel, GermanycDepartment of Cell and Developmental Biology, University College London, Gower Street, London WC1E 6BT, United Kingdom

a r t i c l e i n f o

Article history:Received 4 April 2012Received in revised form29 May 2012Accepted 19 June 2012

Keywords:Neurokinin ANK1/NK2 receptor interactionBrainMouse

Abbreviations: DMSO, dimethyl sulfoxide; FW, faHG, hind limb grooming/biting; ICV, intracerebrovenNK1R�/�mice, NK1 receptor knock-out mice; NKA, nesubstance P; V, vehicle; WDS, wet-dog-shakes.* Corresponding author. Tel.: þ49 431 597 3519; fa

E-mail addresses: utauer@gmx.de (U. Tauer),ucgaunt@ucl.ac.uk (S.P. Hunt), juraj.culman@pharmak

1 They contributed equally to this work.

0028-3908/$ e see front matter � 2012 Elsevier Ltd.http://dx.doi.org/10.1016/j.neuropharm.2012.06.041

a b s t r a c t

Mice lacking the NK1 receptor (NK1R�/� mice) and selective, high-affinity, non-peptide, NK1, NK2 andNK3 receptor antagonists were used to identify the tachykinin receptor subtype(s) mediating the centralresponses induced by neurokinin A (NKA). The peptides, substance P (SP), NKA and senktide and theantagonists were injected intracerebroventricularly (ICV) through an implanted cannula. NKA (50 pmol)was as potent as SP (50 pmol) in inducing grooming behaviour (face washing and hind limb grooming) inwild-type mice, but both peptides failed to induce behavioural responses in NK1R�/� mice. In wild-typemice, the NK1 receptor antagonist, RP 67580 (2 nmol), effectively inhibited grooming behaviour elicitedby SP, but was inactive against grooming induced by NKA, which in turn was abolished after pre-treatment with the selective NK2 receptor agonist, SR 48968 (2 nmol). Unlike NKA, the selective NK2

receptor agonists, (b Ala8) NKA 4e10 and (NLeu10) NKA 4e10, injected ICV at doses of 50 or 100 pmol didnot elicit any behavioural response in wild-type mice. The NK3 receptor antagonist, SR 142801, inhibitedbehaviours induced by the NK3 receptor agonist, senktide, but did not alter behavioural responses toeither SP or NKA in wild-type mice. The present findings demonstrate that central biological actions of SPand senktide are mediated by activation of NK1 and NK3 receptors, respectively. Our results also indicatethat NK1 receptors are essential for generating central actions induced by NKA, which are most probablymediated by a cross-talk between the NK1 and NK2 receptors.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

The most prominent functions of tachykinins in the braincomprise cardiovascular and neuroendocrine regulation andcontrol of behaviour. Neurokinin A (NKA), an endogenous NK2receptor agonist, along with substance P (SP) are the most abun-dant tachykinin peptides in the brain. NK1 and NK3 receptorspreferentially interacting with SP and neurokinin B (NKB), respec-tively, are widely distributed in the adult brain. In contrast, NK2receptors have either not been detected in the adult brain or theirdensity is extremely low (Regoli et al., 1994; Höckfelt et al., 2004).

ce washing/head scratching;tricular(ly); NK, neurokinin;urokinin A; SEN, senktide; SP,

x: þ49 431 597 3522.yi.zhao@uk-sh.de (Y. Zhao),ologie.uni-kiel.de (J. Culman).

All rights reserved.

SP and NKA injected intracerebroventricularly (ICV) inducedidentical cardiovascular and behavioural responses in rats (Culmanet al., 1993; Picard et al., 1994), calling into question which tachy-kinin receptor subtype mediates the central effects of NKA. Severallines of evidence indicate that NKA exerts its effects in the brain viaactivation of the NK1 receptor. Homologous and heterologousbinding analyses in mutant NK1 receptors convincingly demon-strate that NKA is a high-affinity ligand for the NK1 receptor and theNKA binding domain is distinct from the SP binding site (Hastrupand Schwartz, 1996; Maggi and Schwartz, 1997; Wijkhuisen et al.,1999). However, the findings obtained in experiments employinghigh-affinity, non-peptide NK1 and NK2 receptor antagonists haveled to a different conclusion. The NK1 receptor antagonists, CP96,345 or RP 67580 abolished the cardiovascular and behaviouraleffects to SP, yet these antagonists failed to modify the centraleffects of NKA which in turn were inhibited by the NK2 receptorantagonist, SR 48968 (Tschöpe et al., 1992; Picard et al., 1994). As SR48968 does not bind to the NK1 receptor but both, NKA and SR48968, bind with high affinity to the NK2 receptor (Gether et al.,1993; Jensen et al., 1994), these findings suggest that central

U. Tauer et al. / Neuropharmacology 63 (2012) 958e965 959

actions of NKA are mediated by the NK2 receptor (Picard et al.,1994). Although early studies failed to clearly identify NK2 recep-tors in the adult brain, recent findings have demonstrated thepresence of NK2 binding sites in few brain regions, including thehippocampus, the septum, the thalamus and the prefrontal cortex(Saffroy et al., 2001, 2003). The rapid onset of the behaviouralresponse to ICV injected SP or NKA indicates that the neuronalcircuits generating this response are most probably localised in thecircumventricular organs or in adjacent periventricular regions. Asthe presence of NK2 receptors in these areas has not yet beendemonstrated, the tachykinin receptor subtype mediating thephysiological actions of NKA in the brain has not been identified.We, therefore, evaluated the behavioural response induced by SP,NKA and senktide, a selective NK3 receptor agonist, in: i) micelacking the NK1 receptor (NK1R�/� mice) and ii) in wild-type micetreated with high-affinity, non-peptide NK1, NK2 and NK3 receptorantagonists, RP 67580, SR 48968 and SR 142801, respectively. Theeffects of the active enantiomers of the NK1 and NK2 receptorantagonists, RP 67580 and SR 48968, respectively, were comparedwith those of their inactive enantiomers, RP 68651 and SR 48965 todemonstrate the enantioselective effects.

2. Materials and methods

2.1. Animals

Male wild-type and NK1R�/� mice were used in the present experiments.Receptor autoradiography with [125I]Bolton-Hunter SP indicated that SP bindingsites were absent in the brain and spinal cord of NK1R�/� mice (De Felipe et al.,1998). The expression of NK2 receptor in gut tissue was not affected (unpublishedautoradiography with [125I]Bolton-Hunter NKA). Mice of 6e10 week of ageweighing25e35 g were obtained from Professor S. Hunt, University College London, London,United Kingdom. Mice were allowed free access to food and water and maintainedon 12 h light/dark cycle at room temperature 21 �C.

2.2. Implantation of ICV cannulae and ICV injections of agonists and antagonists

Mice were anaesthetised with an intraperitoneal injection of chloralhydrate(400 mg/kg, bw). An ICV polyethylene cannula (PP10) was implanted into the rightcerebral ventricle using a stereotaxic apparatus (David Kopf Instruments, Tujunga,CA, USA) and fixed to the skull with one screw and dental cement. The coordinateswere: 0.3 mm posterior to bregma,1.1 mm lateral to the midline and 2.4 mmverticalfrom the skull surface. Mice were then placed in individual plastic cages and 5 dayswere allowed for the recovery. Three days after the surgery, angiotensin II (50 pmol)was injected ICV to verify the correct position of the ICV cannula. Only the micewhich responded with an immediate drinking were included in the study. Theanimals were transferred into the test room at least 18 h before the beginning of theexperiment. All experiments were carried out in conscious, freely moving mice. Theexperiments were started when the animals were resting. Peptides in a volume of0.5 ml and non-peptide antagonists in a volume of 1 ml, flushed with 2 ml of sterilephysiological saline were slowly injected ICV and the behavioural response wasrecorded. The total volume injected ICV did not exceed 3 ml.

2.3. Determination of behavioural parameters

The experiments were carried out between 9:00 and 13:00 in a quiet testingroom. Mice remained in their resident cages with a grid cage top removed. The ICVcannula was connected to a Hamilton syringe using a polyethylene catheter (PP10)filled with physiological saline. Vehicle, peptides and antagonists were injected ICV(J.C.) when the mice were resting. Behavioural responses were independentlyrecorded by two observers (U.T. and J.C.) sitting next to the cage over a 20min periodstarting immediately after the ICV injection. The frequency of the followingbehavioural manifestations was recorded: 1) face washing/head scratching (FW)(movement of forepaws over the nose, eyes and ears or scratching of the head withthe forepaws), 2) hind limb grooming/biting (HG) (licking or biting the caudal part ofthe body including the hind limbs), 3) rearing (standing on the hind limbs withstretched back), 4) digging (manipulating saw dust) and 5) wet-dog-shakes (WDS)(whole body shaking) using a 15 s sampling procedure. During every consecutiveperiod of 15 s, a score 1 or 0was given depending onwhether themouse showed thespecific type of behaviour or not, regardless of the intensity or the duration of theresponse. Summation of scores for 20 min gave the total behavioural score for allbehavioural manifestations except of WDS. The WDS behaviour was evaluatedaccording to the number of episodes during the 20 min period. The recordings werecompared to check for concordance of the obtained results. In some experiments,

the performance was recorded and analysed post hoc under blind conditions. Onlyminimal differences between these and original recordings were detected.

2.4. Experimental protocols

In the first series of experiments, the behavioural responses induced by ICVinjections of SP, NKA and senktide were investigated in wild-type mice. The firstgroup of mice (n ¼ 7) received a single ICV injection of physiological saline (vehicle1) and the behavioural response was recorded. Twenty minutes later, SP (50 pmol)was injected ICV in 4 mice and NKA (50 pmol) in 3 mice and the behaviouralresponse was evaluated. On the second day, the treatment with the peptides wasreversed. The second group of mice (n ¼ 6) was treated ICV with vehicle 2 and thebehavioural response was recorded. Twenty minutes later, senktide (50 pmol) wasinjected ICV and the behavioural response was recorded again.

The second series of experiments was designated to compare the behaviouralresponses induced by SP, NKA and senktide in wild-type and NK1R�/� mice. Eachmouse received only one injection of the peptide. Twenty min after ICV injection ofvehicle, SP, NKA or senktide (50 pmol each) were injected ICV to wild-type- andNK1R�/� mice (SP: wild-type mice: n ¼ 8; NK1R�/� mice: n ¼ 8; NKA: wild-typemice: n ¼ 8; NK1R�/� mice: n ¼ 7; senktide: wild-type mice: n ¼ 9; NK1R�/�mice: n ¼ 6) and the behaviours were recorded.

In the third series of experiments, we investigated the effects of high-affinity,non-peptide NK1, NK2 and NK3 receptor antagonists, RP 67580, SR 48968 and SR142801, respectively, on behavioural responses induced by SP and NKA (Garret et al.,1991; Emonds-Alt et al., 1992, 1995). Importantly, RP 67580 and SR 48968 may exertoff-target actions which are not enantioselective and unrelated to the inhibition oftachykinin receptors. Therefore, the effects of the active enantiomers of the NK1 andNK2 receptor antagonists, RP 67580 and SR 48968, respectively, were comparedwiththose of their inactive enantiomers, RP 68651 and SR 48965, to demonstrateenantioselective effects. The dose of 2 nmol RP 67580 was previously demonstratedto effectively and selectively inhibit the central cardiovascular and behaviouralresponses of SP in rodents (Picard et al., 1994; Culman et al., 1995). Although thehigh-affinity NK2 receptor antagonist, SR 48968 does not bind to the NK1 receptor(Ki > 10,000 nM; Gether et al., 1993), its interaction with the NK1 receptor could notbe excluded, when high doses of the antagonist are used. To establish the dose of theantagonist, which completely antagonises the behavioural response induced byNKA, mice were pre-treated ICV with vehicle (n ¼ 8) or different doses of SR 48968(0.2 nmol: n¼ 10; 0.5 nmol, n¼ 8 and 2 nmol, n¼ 8). Twentymin later, NKA 50 pmolwas administered ICV and the behavioural responses were recorded. Each mousereceived only one dose of SR 48968 to exclude the accumulation of the antagonist inthe brain. In the second set of experiments, mice were pretreated ICV with vehicle(n ¼ 7) or with 2 nmol of the high-affinity NK1-, NK2- and NK3 receptor antagonists,RP 67580 (n ¼ 9), its inactive enantiomer, RP 68651 (n ¼ 7), SR 48968 (n ¼ 7) and SR142801 (n ¼ 6), respectively. SP (50 pmol) was injected 20 min thereafter and thebehavioural response was recorded. The same protocol was used with NKA. Vehicle(n ¼ 8), RP 67580 (n ¼ 7), SR 48968 (n ¼ 11), its inactive enantiomer, SR 48965(n ¼ 8), or SR 142801 (n ¼ 8) were injected ICV 20 min prior to NKA (50 pmol) andthe behavioural response to the peptide was evaluated.

In the fourth series of experiments, the selective NK2 receptor agonists, (b Ala8)NKA 4e10 and (NLeu10) NKA 4e10 were injected ICV. (NLeu10) NKA 4e10 is 2.5orders of magnitude less active than NKA on the NK1 receptor and (b Ala8) NKA 4e10shows almost 100-fold higher potency for NK2 than NK1 receptors (Regoli et al.,1994). Three groups of mice were used. Twenty min after ICV injections of theappropriate vehicle, the first group (n ¼ 7) received 50 pmol NKA, other two groupsof mice were treated with an equimolar dose of (b Ala8) NKA 4e10 (n ¼ 9) and(NLeu10) NKA 4e10 (n ¼ 7) and the behavioural responses were evaluated. As bothlatter peptides failed to induce any behavioural response, 100 pmol (b Ala8) NKA4e10 and (NLeu10) NKA 4e10 was administered ICV in the second and third groupsof mice 24 h later to avoid tachyphylaxis and the behavioural responses wererecorded.

In the last, fifth series of experiments, we studied the role of NK3 receptors ingenerating ofWDS, which is themost prominent behavioural manifestation inducedby senktide. Vehicle or the selective NK3 receptor antagonist, SR 140801 wasinjected ICV 20 min prior to senktide (50 pmol) in wild-type mice (vehicle: n ¼ 7;senktide: n ¼ 7; SR 140801 þ senktide: n ¼ 6) and NK1R�/� mice (vehicle: n ¼ 8;senktide: n ¼ 6; SR 140801 þ senktide: n ¼ 6) and the number of WDS episodes wasrecorded.

All experimental protocols were approved by the Governmental Committee forthe Ethical Use of Experimental Animals in the German Federal State of Schleswig-Holstein. All efforts have been made to minimise animal suffering and to reduce thenumber of animals used in our experiments.

2.5. Peptides and non-peptides

SP (Bachem, Weil am Rhein, Germany) and NKA (SigmaeAldrich, Taufkirchen,Germany) were dissolved in sterile physiological saline (vehicle 1). Senktide(Bachem, Weil am Rhein, Germany) was dissolved in a small volume of dimethylsufoxide (DMSO) and physiological saline was added to obtain the desiredconcentration of the peptide. The corresponding vehicle (vehicle 2) contained 10%

U. Tauer et al. / Neuropharmacology 63 (2012) 958e965960

DMSO in physiological saline. Stock solutions of the peptides (0.5 nmol/1 ml) weredivided into 20 ml aliquots and stored at �20 �C until used. (b Ala8) NKA 4e10 and(NLeu10) NKA 4e10 (Bachem, Weil am Rhein, Germany) were dissolved in 0.1 MNH4OH and stock solutions (2 nmol/1 ml) were divided into 10 ml aliquots and storedat �20 �C until used. Control mice were treated ICV with 0.5 ml of vehicle 3 (0.005NH4OH M in physiological saline). Dilutions of the peptides were made in physio-logical saline before each experiment. The NK1 receptor antagonist, RP 67580, itsinactive enantiomer RP 68651 (Rhône-Poulenc Rorer Aventis Pharma, Vitry surSeine, France) and the NK2 receptor antagonist, SR 48968 and its inactive enan-tiomer, SR 48965 (Sanofi Recherche, Montpellier, France) were dissolved in a smallvolume of DMSO and 0.1 M HCl (RP 67580 and RP 68651) or physiological saline (SR48968 and SR 48965) (1:1) and saline was added to obtain the concentration(2 nmol/ml). Solutions were divided in aliquots and stored at �20 �C until used. Onthe day of experiment, 1 ml was injected ICV together with 2 ml of phosphate bufferedsaline, pH 7.4. The final solution of SR 48968 or SR 48965 contained less than 10% ofDMSO. One ml of the vehicle 4 containing small amount of 0.1 M HCl and DMSO (lessthan 10%) in physiological saline was injected together with 2 ml of phosphate-buffered saline, pH 7.4. The NK3 receptor antagonist, SR 140801 (Rhône-PoulencRorer Aventis Pharma, Vitry sur Seine, France) was dissolved in ethanol (1/4 of thefinal volume) and physiological saline was added to obtain the final volume. Thesolution (4 nmol/1 ml) was divided in aliquots and stored at�20 �C until used. On theday of experiment, the solution was diluted with physiological saline (1:1) and 1 mlwas injected ICV. Controls received 1 ml of 12.5% ethanol in physiological saline(vehicle 5).

2.6. Statistical analysis

All values are expressed as the mean � SEM. Data were subjected to non-parametric KruskaleWallis test followed by Dunn’s test. Non-parametric Wilcox-oneManneWhitney test was used to compare the behavioural response induced byvehicle and senktide (first series of experiments). A significance level of P< 0.05 wasaccepted.

3. Results

SP and NKA injected ICV in wild-type mice induced behaviouralresponse of the same duration and were equipotent in elicitinggrooming behaviour; FW was the most prominent behaviour.Relative to SP, NKAwasmore effective in inducingWDS. In contrast,senktide elicited almost continuousWDS for the whole observationperiod (Table 1).

Compared to wild-type mice, SP and NKA did not induce anybehavioural effects in NK1R�/� mice (FW and HG: P < 0.001). Inthis set of experiments, the frequency of WDS triggered by NKA inwild-type, but not in NK1�/� mice was only slightly higher thanthat observed in controls treated with vehicle (P ¼ 0.08) (Fig. 1).Again, WDS was the most prominent behavioural manifestationproduced by senktide and no differences were observed betweenwild-type and NK1R�/� mice (Fig. 1).

In the third series of experiments, high-affinity, non-peptideNK1 and NK2 receptor antagonists, RP 67580 and SR 48968 wereused to further explore the role of NK1 and NK2 receptors in theinitiation of the behavioural response induced by SP and NKA in thebrain. The results obtained in a separate set of experiments showed,that 2 nmol of SR 48968 effectively inhibited all behaviours inducedby NKA (FW: P < 0.001; HG: P < 0.01; WDS: P < 0.01) and reducedthe duration of the response (Table 2). Although the 10-times lower

Table 1Behavioural response to substance P, neurokinin A and senktide (50 pmol each) injected

Treatment ICV (n) Face-washing Hind limb grooming

Vehicle 1 (7) 0.7 � 0.6 0.3 � 0.2Substance P (7) 13 � 1.8** 4.2 � 1.4*Neurokinin A (7) 11 � 0.8** 4.5 � 0.7**Vehicle 2 (6) 2.0 � 0.6 0.5 � 0.3Senktide (6) 2.3 � 0.5 0.3 � 0.3

Values represent the frequency of the duration of the response (min) and the frequency oand digging) and wet-dog-shake episodes for 20 min and are indicated by the mean � SEMtreated group (see the Materials and methods section), calculated with the KruskaleWoneManneWhitney test (senktide).

dose of the antagonist (0.2 nmol) reduced FW elicited by NKA, theduration of the response did not significantly differ from thatrecorded in mice treated with NKA (Table 2).

The selective NK1 receptor antagonist, RP 67580 (2 nmol),completely abolished FW and HG to ICV SP (P < 0.001) (Fig. 2). RP67580, however, did not alter FW and HG evoked by NKA, which inturn was effectively inhibited by the selective NK2 receptor antag-onist, SR 48968 (2 nmol) (FW and HG: P < 0.001). The NK2 receptorantagonist was inactive against the behavioural response elicitedby SP. ICV pre-treatment with the inactive enantiomers of the NK1and NK2 receptor antagonists, RP 68651 and SR 48965, respectively,did not affect the behavioural response either to SP or NKA (Fig. 2).Correspondingly, the active enantiomers of NK1 and NK2 receptorsantagonists reduced the duration of the behavioural responses toSP and NKA, respectively, to values indistinguishable from thosedetected in mice treated with vehicle; the inactive enantiomers ofthe tachykinin receptor antagonists were without effect (experi-ments employing SP: vehicle: 1.0 � 0.4; SP (50 pmol) afterpretreatment with vehicle: 11.2 � 0.9a, with RP 67580: 1.1 � 0.5b

and with RP 68651: 9.2 � 1.3a; experiments employing NKA:vehicle: 1.2 � 0.4; NKA (50 pmol) after pretreatment with vehicle:8.9� 0.9a, with SR 48968: 1.5 � 0.4c and with SR 48965: 8.4� 0.9a;values are expressed in min, aP < 0.01, statistical comparison withthe appropriate vehicle-treated group; bP < 0.001 and cP < 0.001,statistical comparison with mice pretreated with vehicle andtreated with SP and NKA, respectively). NKA was more efficaciousin eliciting WDS than SP. The selective NK3 receptor antagonist, SR140801, tended to inhibit this response, but failed to alter FW andHG induced by SP and NKA, clearly indicating that brain NK3receptors do not mediate grooming behaviour elicited by thesepeptides (Fig. 2).

Behavioural effects of NKA and NK2 receptor agonists, (b-Ala8)NKA 4e10 and (NLeu10) NKA 4e10 were compared in the fourthseries of experiments. Again, FW and HG were the two mostprominent behavioural manifestations after ICV injection of NKA(FW: P < 0.001; HG: P < 0.01, WDS: P < 0.01). In contrast, theselective NK2 receptor agonists injected ICV even at twice as highdose as NKA failed to induce behavioural effects and the duration ofthe responses was not different from that recorded in vehicle-treated mice (Table 3).

WDS was the most prominent behavioural manifestationinduced by ICV injection of senktide. The behavioural patternproduced by senktide in wild-type and NK1R�/� mice was specificand did not comprise FW and HG. The selective NK3 receptorantagonist, SR 140801 reduced WDS behaviour and the duration ofthe response in wild-type mice (vehicle: 1.3 � 0.4; senktide(50 pmol) after pretreatment with vehicle: 15.2 � 0.9a and with SR140801: 5.2� 1.2b; values are expressed inmin, aP< 0.01, statisticalcomparison with the vehicle-treated group; bP < 0.05, statisticalcomparison with mice pretreated with vehicle and treated withsenktide). SR 140801 was equally potent in inhibiting the WDSbehaviour in NK1R�/� mice (wild-type mice: P < 0.05; NK1R�/�

intracerebroventricularly (ICV) in wild-type mice.

Rearing Digging Wet-dog-shakes Duration (min)

3.8 � 1.2 1.4 � 0.7 0.4 � 0.3 1.2 � 0.43.4 � 1.4 1.6 � 0.7 2.4 � 0.7 9.9 � 1.2**3.0 � 0.8 3.1 � 1.2 4.5 � 1.4* 8.6 � 0.6*3.0 � 0.7 1.7 � 0.9 2.3 � 0.8 1.8 � 0.52.0 � 1.0 1.5 � 1.0 24 � 2.5** 13.3 � 1.7**

f individual behavioural manifestations (face washing, hind limb grooming, rearingof (n) mice. *P< 0.05, **P< 0.01, statistical comparison to the appropriate vehicle-allis test followed by Dunn’s test (substance P and neurokinin A) or by Wilcox-

Table 2Dose-dependent inhibition of the behavioural response induced by neurokinin A (NKA,antagonist, SR 48968, in wild-type mice.

Pretreatment ICV Treatment ICV (n) Face washing Hind limb g

Vehicle 4 Vehicle 1 (8) 1.6 � 0.5 0.5 � 0.3Vehicle 4 NKA (8) 12.6 � 1.7*** 4.2 � 0.8**SR 48968 0.2 nmol NKA (10) 7.1 � 0.5*y 2.5 � 0.6SR 48968 0.5 nmol NKA (8) 6.2 � 0.7y 2.4 � 1.1SR 48968 2 nmol NKA (8) 2.9 � 1.1yyy 0.2 � 0.2yyy

Values represent the duration of the response (min) and the frequency of individual behawet-dog-shake episodes for 20 min and are indicated by the mean � SEM of (n) mice. *Pvehicle 1 and vehicle 4 (see the Materials andmethods section); yP< 0.05, yyP< 0.01, yyyP<KruskaleWallis test followed by Dunn’s test.

Fig. 1. Behavioural effects induced by substance P (SP), neurokinin A (NKA) andsenktide (SEN) in wild-type mice (empty columns) and NK1R knock-out mice (solidcolumns). Values represent the frequency of face washing and hind-limb grooming andwet-dog-shake episodes for 20 min and are indicated by the mean � SEM. **P < 0.01,***P < 0.001, statistical comparison to the appropriate control group treated withvehicle 1 (V1) or vehicle 2 (V2) (see the Materials and methods section); yP < 0.05,yyyP < 0.001, statistical comparison to wild-type mice treated with SP or NKA, calcu-lated with KruskaleWallis test followed by Dunn’s test.

U. Tauer et al. / Neuropharmacology 63 (2012) 958e965 961

mice: P < 0.05). The equal efficacy of senktide to induce WDS inwild-type and NK1R�/� mice also indicate that knocking-out theNK1 receptor did not modify the expression of NK3 receptor in thebrain (Fig. 3).

The frequency of behavioural manifestations induced by ICVinjection of physiological saline (vehicle 1) was low and notspecific. Spontaneous, unspecific behavioural activity recordedafter ICV injection of vehicles 2, 3, 4 and 5 did not differ from thebehavioural pattern produced by vehicle 1 (data not shown).

4. Discussion

The NKA, an agonist at NK1 and NK2 receptors, is expressed inbrain tissue and is equipotent to SP in inducing cardiovascular andbehavioural effects (Culman et al., 1993). The present study con-ducted in wild-type and NK1R�/� mice provides the first directevidence that activation of central NK1 receptors is essential for thegeneration of behavioural and most probably other physiologicaleffects of NKA in the brain. Together with the data obtained afterpharmacological blockade of NK1 and NK2 receptors, our resultsimply a cross-talk between the NK1 and NK2 as a possible mecha-nism mediating the central effects of NKA.

The most prominent behavioural manifestation, WDS inducedby senktide was selectively blocked by the high-affinity NK3receptor antagonist SR 140801 in both wild-type and NK1R�/�mice, indicating that NK3 receptors mediate this response.

The pattern of the behavioural responses induced by SP or NKAinjected ICV in rats and mice differ. In rats, the peptides elicita typical syntactic grooming chain starting with rhythmic pawlicking followed by large unilateral and bilateral strokes whichextent behind the ears and most of the head (head scratching). Thisphase is usually terminated by a short lasting period of very rapidpaw licking immediately followed by twisting of the torso andlicking of the ventrolateral body surface and skin biting (HG)(Berridge et al., 1987; Cromwell and Berridge, 1996). This pattern ofgrooming sequence possibly repeats 2 or 3 times and the totalscores and duration of paw licking/unilateral and bilateral strokesand hind limb grooming are almost identical. In mice, however, thegrooming behaviour to ICV SP and NKA predominantly comprisespaw licking and forepaw strokes which are interrupted by a shortlasting periods of licking of the ventrolatelal body surface andrarely by skin biting.

NKA in the brain can act through activation of NK1 receptors.Data from experiments employing multiple chimeric NK1/NK2receptors and point mutations of the NK1 receptor point to thepresence of two NK1 receptor conformers, which may not beinterconvertible: an SP e preferring conformer possessing highaffinity for SP, and a general e tachykinin conformer binding withhigh-affinity SP, NKA, septide and other tachykinin peptides (sep-tide site) (Yokota et al., 1992; Fong et al., 1992; Huang et al., 1994;Pradier et al., 1994; Hastrup and Schwartz, 1996; Maggi and

50 pmol) injected intracerebroventricularly (ICV) by the high-affinity NK2 receptor

rooming Rearing Digging Wet-dog-shakes Duration (min)

1.6 � 0.8 2.1 � 0.9 1.1 � 0.4 1.1 � 0.32.3 � 1.0 1.5 � 0.6 5.7 � 0.4** 9.8 � 0.7***2.4 � 0.7 2.8 � 0.9 2.0 � 0.7 4.0 � 0.91.0 � 0.9 1.3 � 0.7 1.5 � 0.6y 2.2 � 0.5yy

0.6 � 0.5 1.1 � 0.9 1.5 � 0.8y 1.3 � 0.5yyy

vioural manifestations (face washing, hind limb grooming, rearing and digging) and< 0.05, **P < 0.01 and ***P < 0.001, statistical comparison to controls treated with0.001, statistical comparison to the group treated ICVwith NKA, calculated with the

Fig. 2. Behavioural responses induced by substance P (SP) and neurokinin A (NKA) injected intracerebroventricularly in wild-type mice 20 min after pre-treatment with vehicle orhigh-affinity, non-peptide NK1, NK2 and NK3 tachykinin receptor antagonists or their inactive enantiomers. NK1ac: active enantiomer of the NK1 receptor antagonist, RP 67580;NK1in: its inactive enantiomer, RP 68651; NK2ac: active enantiomer of the NK2 receptor antagonist, SR 48968; NK2in: its inactive enantiomer, SR 48965; NK3ac: active enantiomerof the NK3 receptor antagonist, SR 140801. Values represent the frequency of face washing and hind-limb grooming and wet-dog-shake episodes for 20 min and are indicated by themean � SEM. *P < 0.05, **P < 0.01, ***P < 0.001, statistical comparison to the vehicle-treated group; yyP < 0.01, yyyP < 0.001, statistical comparison to the group treated ICV with SP orNKA, calculated with KruskaleWallis test followed by Dunn’s test.

Table 3Behavioural response to neurokinin A, (b-Ala8) NKA (4e10), Nleu10 NKA (4e10) injected intracerebroventricularly (ICV) in wild-type mice.

Treatment ICV Dose (pmol) (n) Face-washing Hind limb grooming Rearing Digging Wet-dog-shakes Duration (min)

Vehicle 1 e (7) 1.3 � 0.4 0.0 � 0.0 1.6 � 1.0 1.7 � 1.0 0.9 � 0.4 1.0 � 0.4Neurokinin A 50 (7) 11.7 � 0.8** 4.0 � 0.9*** 2.3 � 1.1 1.6 � 0.9 4.7 � 1.1* 11.2 � 0.8**Vehicle 3 e (7) 1.3 � 0.8 0.7 � 0.5 0.6 � 0.6 0.9 � 0.9 1.3 � 0.7 0.9 � 0.5(b-Ala8) NKA (4e10) 50 (9) 2.3 � 0.8yy 0.0 � 0.0yyy 0.9 � 0.8 0.2 � 0.2 0.4 � 0.3yy 0.8 � 0.3yyy

(b-Ala8) NKA (4e10) 100 (9) 3.0 � 0.6y 0.1 � 0.1yyy 0.3 � 0.2 0.7 � 0.4 4.3 � 1.4 1.8 � 0.4y

Nleu10 NKA (4e10) 50 (7) 1.2 � 0.7yyy 0.0 � 0.0yyy 0.3 � 0.4 0.5 � 0.3 1.7 � 0.8 1.2 � 0.4yy

Nleu10 NKA (4e10) 100 (7) 2.9 � 1.1y 0.0 � 0.0yyy 1.3 � 0.7 0.9 � 0.5 1.6 � 0.9 1.9 � 0.5y

Values represent the duration of the response (min) and the frequency of individual behavioural manifestations (face washing, hind limb grooming, rearing and digging) andwet-dog-shake episodes for 20 min and are indicated by the mean� SEM of (n) mice. *P< 0.05, **P< 0.01, ***P< 0.001, statistical comparison to controls treated with vehicle1; yP < 0.05, yyP < 0.01, yyyP < 0.001, statistical comparison to mice treated ICV with NKA (50 pmol), calculated the KruskaleWallis test followed by Dunn’s test.

U. Tauer et al. / Neuropharmacology 63 (2012) 958e965962

Fig. 3. Wet-dog-shakes induced by vehicle (empty columns) and senktide (SEN) afterpre-treatment with vehicle (solid columns) or the high-affinity, non-peptide NK3

receptor antagonist, SR 140801 (NK3ac) (hatched columns) in wild-type mice (leftpanels) and NK1R�/� mice (right panels). Values represent the number of wet-dog-shake-episodes for 20 min and are indicated by the mean � SEM. ***P < 0.001,statistical comparison to the vehicle-treated group; yP < 0.05, statistical comparison tothe group treated ICV with senktide, calculated with KruskaleWallis test followed byDunn’s test.

U. Tauer et al. / Neuropharmacology 63 (2012) 958e965 963

Schwartz, 1997; Wijkhuisen et al., 1999; Alves et al., 2006). In linewith this concept, NKA failed to induce any behavioural effects inNK1R�/� mice demonstrating that the NK1 receptor mediatescentral actions of the peptide. The lack of the selective NK2 receptoragonists, (NLeu10) NKA 4e10 and (b Ala8) NKA 4e10 to producebehavioural responses also supports this assumption.

The high-affinity NK1 receptor antagonist, RP 67580, completelyabolished the SP-induced grooming and scratching behaviour inwild-type mice, clearly implying brain NK1 receptors in thegeneration of this response. Although, RP 67580 is a highly selec-tive, competitive antagonist of SP in vivo and in vitro (Garret et al.,1991), SP and RP 67580 do not bind to the same amino acids resi-dues in the NK1 receptor and only a few (His-197 and His-265)interact with both the agonist and the antagonist. Nevertheless,the binding pockets of the agonist and antagonist may spatiallyoverlap, thereby producing competitive antagonism (Fong et al.,1992, 1994, 1995). Obviously, the conformational changes of theNK1 receptor induced by RP 67580 prevent the binding of SP but donot affect the binding of NKA, and, consequently, the generation ofbiological responses induced by the peptide. As already mentionedabove, the concept of the common high-affinity binding domain forNKA and septide, which is distinct from the SP binding site, hasbeen firmly established (Hastrup and Schwartz, 1996; Wijkhuisenet al., 1999). Although the apparent affinity of RP 67580 esti-mated in NK1 receptor bioassay was higher against NKA or senktidethan against SP (Sagan et al., 1996; Maggi and Schwartz, 1997), theantagonist did not alter the behavioural response to NKA, indi-cating that this behavioural response is most probably not medi-ated by activating of the septide site on the NK1 receptor.

The antagonist for the NK2 receptor SR 48968 (Maggi et al.,1993; Bhogal et al., 1994) did inhibit cardiovascular and behav-ioural effects of NKA injected ICV in rats, but the cardiovascularresponse to the peptide was completely abolished by a co-treatment with the NK1 and NK2 antagonists, RP 67580 and SR48968, respectively (Picard et al., 1994). These findings clearlydemonstrate that NKA in the brain activates NK2 and to a lesserextent also NK1 receptors (Couture et al., 1995). However, in linewith previous findings obtained in rats (Picard et al., 1994), presentstudy demonstrates that SR 48968 completely abolished groomingbehaviour induced by NKA indicating that brain NK2 receptorsprincipally mediate behavioural effects of NKA. However, the lackof the selective NK2 receptor agonists, (b Ala8) NKA 4e10 and(NLeu10) NKA 4e10, to trigger behavioural effects is not consistentwith this assumption and rules out the hypothesis that central NKA

responses are principally mediated by activation of NK2 receptors(Ravard et al., 1994).

There are two plausible explanations for the present findings.Firstly, SP and NKA interact with distinct high-affinity bindingdomains of the NK1 receptor. The selective NK1 receptor antagonist,RP 67580 effectively inhibits SP effects but not actions mediated byNKA. The NK2 receptor antagonist, SR 48968 antagonises NKAeffects through interfering with the NKA binding to the NK1receptor (Fig. 4A). However, the interaction of the NK2 receptorantagonist with any conformer or subtype of the NK1 receptor hasnot yet been demonstrated. The possibility, that NKA exertsbehavioural effects through binding to a binding site distinct fromthe septide-sensitive binding domain can also be excluded becausethe “NK1-sensitive NKA binding site” postulated by Beaujouan et al.(2000) was insensitive to SR 48968.

An alternative possibility is the cross-linking of the NK1 and theNK2 receptor. Recent evidence suggests that G-protein-coupledreceptors form higher-order oligomers or heterodimers to attainproper surface expression and novel pharmacological properties orto achieve functional activity (Prinster et al., 2005). We hypothesisethat the activation of the second messenger pathway(s) by the NK2

receptor in brain tissue depends on the formation of heterodimerswith the NK1 receptor. The conformational changes of the NK1 andNK2 receptor resulting from the binding of NKA to both receptorsubtypes are the prerequisite for their dimerization (Fig. 4B). Thehigh-affinity NK2 receptor antagonist, SR 48968, prevents theinteraction of NKAwith the NK2 receptor and thereby the formationof the heterodimer with the NK1 receptor (Fig. 4C). The selectiveNK2 receptor agonists, (NLeu10) NKA 4e10 and (b Ala8) NKA 4e10,bind to the NK2, but not to the NK1 receptor and cannot, therefore,initiate the formation of NK1/NK2 heterodimers and, consequently,trigger behavioural responses (Fig. 4D). NK2 receptors have beendetected in discrete brain regions including the hippocampus, theseptum the thalamus and the cortex (Steinberg et al., 1998; Saffroyet al., 2001, 2003). Nevertheless, the proposed concept of NK1/NK2receptor dimerization conforms to all principle findings reported inthe present study: i) the essential role of the NK1 receptor in themediating central NKA responses, ii) the high potency of SR 48968to inhibit these responses and, iii) the inability of selective peptideNK2 receptor agonists to trigger biological effects in the brain. Thishypothesis also helps to elucidate our previous findings ondesensitisation and cross-desensitisation of cardiovascular andbehavioural effects induced by ICV injected SP or NKA. When SPwas used as the first stimulus, the responses induced by the secondinjection were abolished regardless of which peptide, SP or NKA,was used. However, when NKA was used as the first stimulus, theresponses induced by the second peptide, either SP or NKA, werenot altered (Culman et al., 1993). The homologous desensitisationinduced by the ligand binding occurs by G-protein receptor kinasephosphorylation of the receptor. SP causes a rapid internalisation ofNK1 receptors, which are no more available for activation by eitherpeptide. On the other hand, the formation of NK1/NK2 heterodimersinitiating by NKA may prevent the internalisation of NK1 receptors,which can then interact either with SP or NKA.

The selective NK1 receptor antagonist, aprepitant, whichproduces approximately 90% NK1 receptor occupancy in the brain,was ineffective in the treatment of the major depressive disorderdespite the fact, that a number of preclinical behavioural studieshad suggested antidepressant-like and/or anxiolytic effects ofpharmacological or genetic inactivation of NK1 receptor (Krameret al., 1998; Keller et al., 2006). The recent clinical study employ-ing the novel NK1 receptor antagonist, Casopitant (GW679769),providing nearly complete receptor occupancy for at least 24 h,demonstrates that NK1 receptor antagonists may be effective in thetreatment of depression. This study supports the hypothesis that

Fig. 4. Possible mechanisms of the activation of signal transduction pathways by SP and NKA in the brain. A: SP and NKA bind to the NK1 receptor and activate the correspondingsignal transduction pathways, indicated by the large black arrow. Changes in NK1 receptor conformation (not shown in the figure) induced by the high-affinity NK1 receptorantagonists (NK1 RA), like RP 67580, prevent the binding of SP but not the binding of NKA to the NK1 receptor. The high-affinity NK2 receptor antagonist (NK2 RA), SR 48968,interacts with the NK1 receptor, prevents the binding of NKA and, consequently, the activation of the corresponding signal transduction pathway. However, the interaction of SR48968 or any high-affinity NK2 receptor antagonists with the NK1 receptor could not yet be demonstrated. B: High-affinity binding of NKA to the NK1 e (at a distinct site than SP)and to the NK2 receptor induces conformation changes of both receptors (not shown in the figure), which allow the formation of NK1/NK2 receptor heterodimers resulting in anactivation of signal transduction pathways. The simultaneous binding of NKA to the NK1 and NK2 receptors is the prerequisite for their dimerization. C: The NK2 receptor antagonist,SR 48968, does not alter the binding of NKA to the NK1 receptor but prevents the interaction of NKA with the NK2 receptor and, consequently, the formation of NK1/NK2 receptorheterodimers. D: The selective NK2 receptor agonists, (NLeu10) NKA 4e10 and (b Ala8) NKA 4e10, possess high-affinities only for the NK2 receptor. They do not interact with the NK1

receptor and, therefore, they cannot initiate the formation of NK1/NK2 receptor heterodimers.

U. Tauer et al. / Neuropharmacology 63 (2012) 958e965964

nearly complete NK1 receptor blockade is required to produceclinical antidepressant effects (Ratti et al., 2011). We demonstratehere that the selective NK1 receptor antagonist inhibits only theeffects of SP mediated by the NK1 receptor, the actions triggered byNKA remain unimpaired. However, antidepressant-like effects andreduced corticotropin-releasing factor function were also reportedafter selective blockade of the NK2 receptor (Louis et al., 2008;Steinberg et al., 2001). Our data indicate that a simultaneoustreatment with selective NK1- and NK2 receptor antagonists mayeffectively antagonise SP and NKA activity at NK1 and NK2receptors.

In summary, the behavioural effects induced by SP are mediatedby NK1 receptors. The present findings do not support the view thatbiological responses induced by NKA in the brain are solely medi-ated by NK2 receptors and rather imply the cross-talk between theNK1 and NK2 receptors in triggering these responses.

Acknowledgements

The authors wish to thank Dr Nadia Rupniak for helpful reviewof this manuscript and Ms. Britta Schwarten for her excellenttechnical assistance.

U. Tauer et al. / Neuropharmacology 63 (2012) 958e965 965

References

Alves, I.D., Delaroche, D., Mouillac, B., Salamon, Z., Tollin, G., Hruby, V.J., Lavielle, S.,Sagan, S., 2006. The two NK-1 binding sites correspond to distinct, independent,and non-interconvertible receptor conformational states as confirmed byplasmon-waveguide resonance spectroscopy. Biochemistry 45, 5309e5318.

Beaujouan, J.C., Saffroy, M., Torrens, Y., Glowinski, J., 2000. Different subtypes oftachykinin NK(1) receptor binding sites are present in the rat brain.J. Neurochem. 75, 1015e1026.

Berridge, K.C., Fentress, J.C., Parr, H., 1987. Natural syntax rules control actionsequence of rats. Behav. Brain Res. 23, 59e68.

Bhogal, N., Donnelly, D., Findlay, J.B., 1994. The ligand binding site of the neurokinin2 receptor. Site-directed mutagenesis and identification of neurokinin Abinding residues in the human neurokinin 2 receptor. J. Biol. Chem. 269,27269e27274.

Couture, R., Picard, P., Poulat, P., Pratt, A., 1995. Characterization of the tachykininreceptors involved in spinal and supraspinal cardiovascular regulation. Can. J.Physiol. Pharmacol. 73, 892e902.

Cromwell, H.C., Berridge, K.C., 1996. Implementation of action sequences by a neo-striatal site: a lesion mapping study of grooming syntax. J. Neurosci. 16,3444e3458.

Culman, J., Tschöpe, C., Jost, N., Itoi, K., Unger, T., 1993. Substance P and neurokinin Ainduced desensitization to cardiovascular and behavioral effects: evidence forthe involvement of different tachykinin receptors. Brain Res. 625, 75e83.

Culman, J., Wiegand, B., Spitznagel, H., Klee, S., Unger, T., 1995. Effects of thetachykinin NK1 receptor antagonist, RP 67580, on central cardiovascular andbehavioural effects of substance P, neurokinin A and neurokinin B. Br. J. Phar-macol. 114, 1310e1316.

De Felipe, C., Herrero, J.F., O’Brien, J.A., Palmer, J.A., Doyle, C.A., Smith, A.J., Laird, J.M.,Belmonte, C., Cervero, F., Hunt, S.P., 1998. Altered nociception, analgesia andaggression in mice lacking the receptor for substance P. Nature 392, 394e397.

Emonds-Alt, X., Vilain, P., Goulaouic, P., Proietto, V., Van Broeck, D., Advenier, C.,Naline, E., Neliat, G., Le Fur, G., Brelière, J.C., 1992. A potent and selective non-peptide antagonist of the neurokinin A (NK2) receptor. Life Sci. 50, PL101ePL106.

Emonds-Alt, X., Bichon, D., Ducoux, J.P., Heaulme, M., Miloux, B., Poncelet, M.,Proietto, V., Van Broeck, D., Vilain, P., Neliat, G., et al., 1995. SR 142801, the firstpotent non-peptide antagonist of the tachykinin NK3 receptor. Life Sci. 56,PL27ePL32.

Fong, T.M., Yu, H., Huang, R.R., Strader, C.D., 1992. The extracellular domain of theneurokinin-1 receptor is required for high-affinity binding of peptides.Biochemistry 31, 11806e11811.

Fong, T.M., Yu, H., Cascieri, M.A., Underwood, D., Swain, C.J., Strader, C.D., 1994.Interaction of glutamine 165 in the fourth transmembrane segment of thehuman neurokinin-1 receptor with quinuclidine antagonists. J. Biol. Chem. 269,14957e14961.

Fong, T.M., Huang, R.C., Yu, H., Swain, C.J., Underwood, D., Cascieri, M.A.,Strader, C.D., 1995. Mutational analysis of neurokinin receptor function. Can. J.Physiol. Pharmacol. 73, 860e865.

Garret, C., Carruette, A., Fardin, V., Moussaoui, S., Peyronel, J.F., Blanchard, J.C.,Laduron, P.M., 1991. Pharmacological properties of a potent and selective non-peptide substance P antagonist. Proc. Natl. Acad. Sci. U. S. A. 88, 10208e10212.

Gether, U., Yokota, Y., Emonds-Alt, X., Brelière, J.C., Lowe 3rd, J.A., Snider, R.M.,Nakanishi, S., Schwartz, T.W., 1993. Two nonpeptide tachykinin antagonists actthrough epitopes on corresponding segments of the NK1 and NK2 receptors.Proc. Natl. Acad. Sci. U. S. A. 90, 6194e6198.

Hastrup, H., Schwartz, T.W., 1996. Septide and neurokinin A are high-affinity ligandson the NK-1 receptor: evidence from homologous versus heterologous bindinganalysis. FEBS Lett. 399, 264e266.

Höckfelt, T., Kuteeva, T., Stanic, D., Ljungdahl, Å., 2004. The histochemistry oftachykinin system in the brain. In: Holzer, P. (Ed.), Tachykinins. Handbook ofExperimental Pharmacology. Springer-Verlag, Berlin, Heidelberg, pp. 63e120.

Huang, R.R., Yu, H., Strader, C.D., Fong, T.M., 1994. Interaction of substance P withthe second and seventh transmembrane domains of the neurokinin-1 receptor.Biochemistry 33, 3007e3013.

Jensen, C.J., Gerard, N.P., Schwartz, T.W., Gether, U., 1994. The species selectivity ofchemically distinct tachykinin nonpeptide antagonists is dependent oncommon divergent residues of the rat and human neurokinin-1 receptors. Mol.Pharmacol. 45, 294e299.

Keller, M., Montgomery, S., Ball, W., Morrison, M., Snavely, D., Liu, G., Hargreaves, R.,Hietala, J., Lines, C., Beebe, K., Reines, S., 2006. Lack of efficacy of the substance p(neurokinin1 receptor) antagonist aprepitant in the treatment of majordepressive disorder. Biol. Psychiatry 59, 216e223.

Kramer, M.S., Cutler, N., Feighner, J., Shrivastava, R., Carman, J., Sramek, J.J.,Reines, S.A., Liu, G., Snavely, D., Wyatt-Knowles, E., Hale, J.J., Mills, S.G.,MacCoss, M., Swain, C.J., Harrison, T., Hill, R.G., Hefti, F., Scolnick, E.M.,Cascieri, M.A., Chicchi, G.G., Sadowski, S., Williams, A.R., Hewson, L., Smith, D.,Carlson, E.J., Hargreaves, R.J., Rupniak, N.M., 1998. Distinct mechanism forantidepressant activity by blockade of central substance P receptors. Science281, 1640e1645.

Louis, C., Stemmelin, J., Boulay, D., Bergis, O., Cohen, C., Griebel, G., 2008. Additionalevidence for anxiolytic- and antidepressant-like activities of saredutant(SR48968), an antagonist at the neurokinin-2 receptor in various rodent-models. Pharmacol. Biochem. Behav. 89, 36e45.

Maggi, C.A., Schwartz, T.W., 1997. The dual nature of the tachykinin NK1 receptor.Trends Pharmacol. Sci. 18, 351e355.

Maggi, C.A., Patacchini, R., Giuliani, S., Giachetti, A., 1993. In vivo and in vitropharmacology of SR 48,968, a non-peptide tachykinin NK2 receptor antagonist.Eur. J. Pharmacol. 234, 83e90.

Picard, P., Regoli, D., Couture, R., 1994. Cardiovascular and behavioural effects ofcentrally administered tachykinins in the rat: characterization of receptors withselective antagonists. Br. J. Pharmacol. 112, 240e249.

Pradier, L., Ménager, J., Le Guern, J., Bock, M.D., Heuillet, E., Fardin, V., Garret, C.,Doble, A., Mayaux, J.F., 1994. Septide: an agonist for the NK1 receptor acting ata site distinct from substance P. Mol. Pharmacol. 45, 287e293.

Prinster, S.C., Hague, C., Hall, R.A., 2005. Heterodimerization of g protein-coupledreceptors: specificity and functional significance. Pharmacol. Rev. 57,289e298.

Ratti, E., Bellew, K., Bettica, P., Bryson, H., Zamuner, S., Archer, G., Squassante, L.,Bye, A., Trist, D., Krishnan, K.R., Fernandes, S., 2011. Results from 2 randomized,double-blind, placebo-controlled studies of the novel NK1 receptor antagonistcasopitant in patients with major depressive disorder. J. Clin. Psychopharmacol.31, 727e733.

Ravard, S., Betschart, J., Fardin, V., Flamand, O., Blanchard, J.C., 1994. Differentialability of tachykinin NK-1 and NK-2 agonists to produce scratching andgrooming behaviours in mice. Brain Res. 651, 199e208.

Regoli, D., Boudon, A., Fauchére, J.L., 1994. Receptors and antagonists for substance Pand related peptides. Pharmacol. Rev. 46, 551e599.

Saffroy, M., Torrens, Y., Glowinski, J., Beaujouan, J.C., 2001. Presence of NK2 bindingsites in the rat brain. J. Neurochem. 79, 985e996.

Saffroy, M., Torrens, Y., Glowinski, J., Beaujouan, J.C., 2003. Autoradiographicdistribution of tachykinin NK2 binding sites in the rat brain: comparison withNK1 and NK3 binding sites. Neuroscience 116, 761e773.

Sagan, S., Chassaing, G., Pradier, L., Lavielle, S., 1996. Tachykinin peptides affectdifferently the second messenger pathways after binding to CHO-expressedhuman NK-1 receptors. J. Pharmacol. Exp. Ther. 276, 1039e1048.

Steinberg, R., Marco, N., Voutsinos, B., Bensaid, M., Rodier, D., Souilhac, J., Alonso, R.,Oury-Donat, F., Le Fur, G., Soubrie, P., 1998. Expression and presence of septalneurokinin-2 receptors controlling hippocampal acetylcholine release duringsensory stimulation in rat. Eur. J. Neurosci. 10, 2337e2345.

Steinberg, R., Alonso, R., Griebel, G., Bert, L., Jung, M., Oury-Donat, F., Poncelet, M.,Gueudet, C., Desvignes, C., Le Fur, G., Soubrié, P., 2001. Selective blockade ofneurokinin-2 receptors produces antidepressant-like effects associated withreduced corticotropin-releasing factor function. J. Pharmacol. Exp. Ther. 299,449e458.

Tschöpe, C., Picard, P., Culman, J., Prat, A., Itoi, K., Regoli, D., Unger, T., Couture, R.,1992. Use of selective antagonists to dissociate the central cardiovascular andbehavioural effects of tachykinins on NK1 and NK2 receptors in the rat. Br. J.Pharmacol. 107, 750e755.

Wijkhuisen, A., Sagot, M.A., Frobert, Y., Créminon, C., Grassi, J., Boquet, D.,Couraud, J.Y., 1999. Identification in the NK1 tachykinin receptor of a domaininvolved in recognition of neurokinin A and septide but not of substance P. FEBSLett. 447, 155e159.

Yokota, Y., Akazawa, C., Ohkubo, H., Nakanishi, S., 1992. Delineation of structuraldomains involved in the subtype specificity of tachykinin receptors throughchimeric formation of substance P/substance K receptors. EMBO J. 11,3585e3591.