D-antagonist in the rat prefrontal cortex

Neuroscience Letters 402 (2006) 253–258

Yohimbine acts as a putative in vivo ␣2A/D-antagonist in the rat prefrontal cortex P´eter Kov´acs, Istv´an Hern´adi ∗ Department of Experimental Zoology and Neurobiology, University of P´ecs, 6. Ifj´us´ag str., H-7624 P´ecs, Hungary Received 21 November 2005; received in revised form 27 March 2006; accepted 10 April 2006

Abstract Yohimbine has been widely used as ␣2 -adrenergic receptor antagonist in neurophysiological research and in clinical therapy. In this study, we provide in vivo electrophysiological evidence, that microiontophoretic application of yohimbine (YOH) inhibits spontaneous activity of prefrontal neurons of the rat. By microiontophoretic application of the ␣2A -receptor antagonist BRL44408 (BRL), the effects of YOH could be mimicked, indicating that the action of YOH is manifested through ␣2A/D -receptor mechanisms. Furthermore, the inhibiting effects of YOH or BRL were blocked by ␣2B -receptor antagonist imiloxan. In concert with previous microiontophoretic data, the present findings suggest that ␣2 -receptor antagonist YOH predominantly acts on the ␣2A/D -receptor subtype in vivo. Furthermore, we hypothesize that this action is manifested via deactivation of autoreceptors causing increased norepinephrine release, finally inhibiting postsynaptic neurons through the activation of ␣2B -receptors. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Alpha-2 adrenergic receptors; Microiontophoresis; Norepinephrine; Prefrontal cortex; Yohimbine

Yohimbine (YOH) is a widely applied substance previously described as a non-selective ␣2 -adrenoceptor antagonist in the nervous system. It has significant antinociceptive action [8], and it has been shown to be effective in the treatment of obesity [19] and psychiatric disorders, such as depression or narcolepsy [28]. Yohimbine has been used in neuropharmacological research of stress [21], anxiety [7] and memory processes [3]. In rodent experimental models, these functions have been suggested to be highly dependent on norepinephrinergic (NEergic) mechanisms intrinsic to the medial prefrontal cortex (mPFC) [12,23,27]. Intracerebral ␣2 -adrenoceptors have been divided into four subtypes: ␣2A , ␣2B , ␣2C and ␣2D [4]. The human ␣2A -subtype is homologous with the rodent ␣2D -subtype [16] often referred as ␣2A/D -subtype [29]. Although ␣2A/D -subtype also acts presynaptically as autoreceptor [29], the majority of ␣2A/D -receptors in the brain are postsynaptic [1]. The ␣2A/D -autoreceptors inhibit release of norepinephrine (NE) into the synaptic cleft, as their stimulation decreases the extracellular levels of NE [5,13,31]. In addition to ␣2A/D -receptors, on the postsynaptic side, ␤1 -, ␣1 -, ␣2B - and ␣2C -receptors are also expressed [6,16]. While ␤1 - and

∗

Corresponding author. Tel.: +36 72 503600x4816; fax: +36 72 501517. E-mail address: [email protected] (I. Hern´adi).

0304-3940/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2006.04.011

␣1 -receptors depolarise both subcortical [2,6] and cortical neurons [17,20], through the activation of intracellular cAMP- and inositol 1,4,5-trisphosphate (IP3 ) second messenger systems, only ␣2A/D , ␣2B - and ␣2C -receptors are known to hyperpolarize neurons by the inhibition of cAMP synthesis [2,6]. There are three major conflicting interpretations of the action mechanisms of YOH in the brain. Firstly, in most cases, YOH has been shown as a general ␣2 -adrenoceptor antagonist [3,10]. Secondly, YOH reportedly acts as a subtype-specific antagonist on ␣2B /␣2C [16,30] or ␣2A/D receptors [11,32] in vitro. Thirdly, YOH has been used as a presynaptic ␣2 -receptor antagonist without considering a confounding effect through postsynaptic ␣2 receptor mechanisms [9,24]. If YOH mainly antagonizes the presynaptic autoreceptors, it should increase the release of NE as it was formerly described [5,24,28,31,32]. Previous data from our laboratory [15] and other independent studies [26] demonstrated that YOH decreases maintained firing rate in the rat and the monkey PFC, respectively. Yohimbine inhibition was successfully mimicked with ␣2 -agonist clonidine (CLON), but not by ␣1 -antagonist prazosin, suggesting a primary action of YOH on ␣2 receptors. Although, in vitro data is available suggesting the ␣2A/D receptor-specific action of YOH [11,32], there is no in vivo electrophysiological evidence for similar action of YOH. Because of the widespread pharmacological [8,19,28] and behavioural

254

P. Kov´acs, I. Hern´adi / Neuroscience Letters 402 (2006) 253–258

[3,7,21] application of YOH, it is important to define its in vivo action accurately. In this study, we aimed to determine the in vivo pharmacological action of YOH on the maintained firing activity of medial PFC (mPFC) neurons in the rat. The action of YOH was confirmed by means of combined extracellular single neuron recording and microiontophoresis of YOH and various other ␣2 -adrenergic receptor agents. Experiments were approved by the Animal Care Committee at our Institution (University of P´ecs, Hungary), in compliance of international standards (NIH Guidelines) for the care and use of laboratory animals. Twenty male Long Evans rats (Charles River Laboratories, G¨od¨oll˝o, Hungary) were used during experiments. Anaesthesia was induced with a single injection of ketamine (100 mg/kg, SBH, Hungary) and maintained with 20% of the initial dose administered in approximately every 45 min thereafter for a maximum of three times. Stereotaxic coordinates for the targeted regions of the mPFC according to Paxinos and Watson [22] were: AP (from bregma) 2.7–3.7 mm; L 0.2–1.0 mm; V (from the dura) 1.0–5.0 mm. Seven-barrelled micropipettes were used for recording and microiontophoresis (Carbostar-7, Kation Scientific, USA). The impedance of the central recording channel was 0.2–0.6 M
(at 50 Hz), and the impedance for each drug channel was 10–50 M
. One of the drug-channels (filled with 0.5 M NaCl or pontamine sky blue or methylene blue solution) was used for the application of a continuous balancing current, while four other capillaries were filled with the following substances: ␣2 antagonist YOH (ICN, 10 mM, dissolved in 30% DMSO/70% distilled water), subtype selective ␣2A/D -antagonist BRL44408 (BRL) (Tocris 10 mM, dissolved in distilled water), subtype selective ␣2B -antagonist imiloxan (IMI) (Tocris, 10–80 mM dissolved in distilled water) and subtype specific ␣2A/D -agonist guanfacine (GUA) (Tocris, 10–80 mM, dissolved in distilled water). The remaining electrode channel was filled with GABA (Sigma, 50 mM, dissolved in distilled water) or kainic acid (Sigma, 50 mM, dissolved in distilled water) for inhibitory or excitatory control purposes, respectively. All compounds, except kainic acid, were ejected as cations by individual constant current circuits (Neurophore BH2, Medical Systems Corp., USA). To determine the overall electric charge (nC) passed through the electrode tip during drug applications, we multiplied the applied electric current (nA) by the duration of pump opening (s). Extracellular single unit activity was recorded and passed through an analogue-digital converter interface (Power 1401, CED, Cambridge, UK) to an IBM-compatible microcomputer. Spike sorting and data analysis was performed by Spike2 software (CED, Cambridge, UK) to ensure that data were always recorded from single neurons. Frequency histograms of neuronal discharge activity were computed and displayed in cycles per seconds (cps, Hz). Baseline activity was recorded for 60 s before starting a drug application trial. A neuron was considered responsive to a treatment when its firing rate changed ±20% respective to its baseline level. Responses were assessed during microiontophoresis and, typically, for 60 s after the termination of drug ejection. In the case of long lasting drug application trials (>100 s), neuronal activity was processed for a minimum of 60 s, or until the activity of the cell first reached its pre-trial fir-

ing rate. Neuronal activity was converted to normalized values (Table 1, Fig. 2B) representing the percentage of activity changing between the neuron’s baseline firing rate and its response to treatment (±standard deviation, S.D.). Statistical comparison between pre- and post treatment firing activity was performed using Student’s paired t-tests. Neuronal responses were then analysed by one-way ANOVA between three response categories (excitation, inhibition, no response). The threshold for significance for all statistical comparisons was set to p < 0.01. Recorded areas were microiontophoretically labelled with pontamine sky blue or methylene blue through continuous application from the current balancing channel of the electrode during the recording sessions, as described elsewhere [14]. At the end of each recording session, animals were perfused transcardially with saline solution followed by 4% paraformaldehyde. Brains were postfixed and rinsed in PBS. Native 40 ␮m slices were made and sections were studied under a light microscope. We recorded extracellular single unit activity of 119 mPFC neurons in conjunction with 889 different drug application trials. The mean baseline firing activity of the neuronal pool was 12.37 ± 3.24 Hz. Localization of each electrode track was successful, but histological analysis for the localization of individual neurons was possible only in 90% (107/119) of recorded positions. If a neuron fell outside the target area, or its position could not be located within the electrode track, it was excluded from further analyses. The distribution of localized recorded neurons among the target subregions of the mPFC were the following: anterior cingulate: 27, prelimbic: 49, infralimbic: 31. As neuronal responsiveness did not reveal subregional differences, the target area will be referred as mPFC throughout the text. Yohimbine: The effect of YOH microiontophoresis was inhibitory in 55 of 63 tested mPFC neurons (87%) to 0.51 ± 0.18 of the pre-treatment spontaneous activity (one-way ANOVA, F(1,191) = 48.85, p < 0.0001) (Fig. 1A, B and D). Firing rate suppression typically lasted 100–1500 s and was proportional to the applied electric charge. BRL 44408: The subtype selective ␣2A/D -adrenoceptor antagonist BRL induced similar electrophysiological effects to YOH (Fig. 1A–D). Similarly to YOH, BRL also decreased the spontaneous firing rate (0.49 ± 0.17) in 68 of 82 neurons (83%) (oneway ANOVA, F(2,320) = 282.54, p < 0.0001). BRL-induced suppression was recorded with all applied currents (10–70 nA, 100–500 s), while excitations were only recorded in 4% of tested neurons with lower ejection current (10–20 nA) (Table 1). Overall, the application of BRL induced shorter inhibitory responses of maintained firing activity than that of YOH (mean duration of inhibition (±S.E.M.) YOH: 504.48 ± 67.46, BRL: 104.58 ± 14.64, Student’s t-test, p < 0.0001). Fig. 2A depicts the firing rate suppression as a function of the duration of microiontophoretic YOH or BRL application. Imiloxan: The majority of tested neurons (47/73, 64%) did not change their activity during IMI ejection, even when IMI was applied with several different protocols from short to long duration (Fig. 1B, C, D and F). However, 28% of IMI applications resulted in excitation of the recorded neurons (Fig. 1E), and only a small portion of neurons (6/73, 8%) were suppressed by IMI (one-way ANOVA, F(2,161) = 64.82, p < 0.0001) (Table 1).

P. Kov´acs, I. Hern´adi / Neuroscience Letters 402 (2006) 253–258

255

Fig. 1. Typical effects of microiontophoretically applied BRL, GUA, IMI and YOH on single unit activity in the rat mPFC. (A) YOH and BRL both decrease neuronal activity. (B) YOH and BRL both decrease neural activity, while IMI antagonizes YOH-inhibition. (C) IMI antagonizes BRL-inhibition, BRL excites under IMI-application. (D) YOH facilitates neuronal firing after application of IMI. (E) IMI elicits excitation. YOH inhibitions become less extent after several IMI ejections. (F) GUA elicits firing excitation. YOH causes firing excitation after IMI-ejection. Frequency histograms are displayed in cycles per second (1/s, Hz). Ejection currents in nanoamperes (nA) are shown above the histograms. Waveform averages of corresponding single neuron action potentials are shown on the right side of each histogram.

256

P. Kov´acs, I. Hern´adi / Neuroscience Letters 402 (2006) 253–258

Table 1 Effects of NEergic agonists and antagonists on single cell unit activity of mPFC neuronsa Compound

a Effect

nneuron

%neuron

ntrials

↓ ∅ ↑

55/63 8/63 0/63

87 13 0

↓ ∅ ↑

68/82 11/82 3/82

↓ ∅ ↑ ↓ ∅ ↑

%trials

Normalized effect ± S.D.

Ejection current (nA)

188/196 8/196 0/196

96 4 0

0.51 ± 0.18 0.97 ± 0.12 –

5–20 10–50 –

48.85

p < 0.0001 – –

83 13 4

278/322 38/322 6/322

86 12 2

0.49 ± 0.17 0.98 ± 0.06 2.29 ± 1.09

10–70 20–50 10–20

282.54

p < 0.0001 – p < 0.01

6/73 47/73 20/73

8 64 28

15/165 92/165 58/165

9 56 35

0.53 ± 0.18 0.99 ± 0.03 4.04 ± 2.84

40–70 5–50 10–40

64.82

p < 0.0001 – p < 0.0001

6/23 4/23 13/23

26 17 57

21/69 18/69 30/69

30 26 44

0.55 ± 0.25 1.07 ± 0.26 2.19 ± 0.75

20–40 15–30 5–30

61.77

p < 0.0001 – p < 0.0001

Combined actions IMI antagonizes YOH-inhibition

27/29

93

48/54

89

1.69 ± 0.69

5–50

IMI antagonizes BRL-inhibition

29/33

88

61/83

73

1.49 ± 0.43

5–50

Simple effects YOH

BRL

IMI

GUA

n a

119

One-way ANOVA F value

137.27 324.24

Statistical comparison

p < 0.0001 p < 0.0001

889

Symbols: ↑ excitation, ↓ inhibition, ∅ no effect.

Fig. 2. (A) Scatterplot of firing rate suppression as a function of the duration of microiontophoretic YOH and BRL application in the mPFC. Continuous lines represent the best-fit linear regression curves over each data series. (B) Summary diagram indicating the antagonizing effects of combined microiontophoretic application of IMI + YOH (n = 27), or IMI + BRL (n = 29) on YOH (n = 55) or BRL (n = 68) induced inhibition of the spontaneous activity of mPFC neurons. Data is expressed as ‘normalized effects’ relative to pre-treatment baseline firing activity (=1) ± standard deviation (S.D.). * p < 0.01.

Guanfacine: The ␣2A/D -agonist GUA caused excitation in 57% of the tested neurons (13/23, 57%, Fig. 1F) while only 26% (6/23) of them were inhibited (one-way ANOVA, F(2,74) = 61.77, p < 0.0001) (Table 1). Combined actions: The ␣2B -antagonist IMI successfully blocked the inhibition elicited by YOH (27/29 neurons, 93%, Fig. 1B), or BRL (29/33 neurons, 88%, Fig. 1C) (oneway ANOVA: YOH/YOH + IMI F(1,92) = 137.27, p < 0.0001; BRL/BRL + IMI F(1,120) = 324.24, p < 0.0001) (Table 1). Interestingly, YOH- or BRL-induced suppression turned to excitation in 75 of 137 drug application trials (55%) during (Fig. 1C and F) or after IMI application (Fig. 1D). The combined (antagonizing) effects of IMI on YOH- or BRL-induced firing suppression on the population level are summarized in Fig. 2B. Although studies using in vitro cortical slices can provide basic information on possible neurotransmitter actions, it is of utmost importance to determine these actions under more physiological conditions such as in intact, anesthetized animals. In the present study, we have shown that microiontophoresis of YOH induced suppression of maintained firing activity of mPFC neurons in vivo. This effect of YOH was mimicked by a known ␣2A/D -antagonist BRL. In addition, firing rate suppression, induced either by YOH or BRL, was successfully blocked by a known ␣2B -antagonist IMI. Reverse effects (excitation) were induced by the application of the specific ␣2A/D -agonist GUA on neurons, where YOH or BRL exerted firing inhibition. In line with previous, in vitro studies [5,24,28,31,32], we conclude, that the most likely result of YOH action can be the increased release of NE into the synaptic cleft by the deactivation of ␣2A/D -autoreceptors. Therefore, higher concentration of

P. Kov´acs, I. Hern´adi / Neuroscience Letters 402 (2006) 253–258

NE may bind to postsynaptic ␣2B -receptors, while postsynaptic ␣2A/D -receptors are also blocked by YOH. Finally, neurons become hyperpolarized and firing will be inhibited, similarly to results of in vitro experiments [6]. In the present experimental model, the excitatory effects of GUA may act via activation of ␣2A/D -autoreceptors, resulting in decreased NE levels and reduced activity of postsynaptic ␣2B -receptors. This mechanism is also suggested by behavioural studies demonstrating that GUA reverses adverse effects of increased NE levels on memory performance [18]. In contrast, GUA-induced inhibition of firing is attributed to the activation of postsynaptic ␣2A/D -receptors. As this action was observed less frequently in the present experiment (6/23, 26%), the primary effects of GUA were exerted on the presynaptic ␣2A/D -receptor terminals of mPFC neurons. Provided that YOH is generally active, as it was previously suggested [3,10], it should antagonize all types of ␣2 receptors. Therefore, NE could only act on postsynaptic ␣1 - or ␤1 -receptors. This would lead to a YOH-induced excitation. In this case, it would not be possible to reverse the effect of YOH by selective ␣2B -receptor antagonist IMI or mimic it by selective ␣2A/D -receptor antagonist BRL. However, in the present experiment, selective blockage of ␣2B -receptors by IMI alone, or combined administration of IMI with YOH or BRL, both resulted in increased firing activity in the majority of tested neurons. This excitatory effect can be accounted for by the action of NE on postsynaptic ␣1 -receptors, as elevated extracellular NE levels could not activate ␣2 receptors due to the possible selective blockage of ␣2A (YOH or BRL) and ␣2B (IMI) receptors. A co-existence of opposite effects on membrane excitability mediated by these two types of adrenoceptors (␣2 and ␣1 ) was also observed previously on the same neuron in vitro [6]. Direct ␣1 -receptor mediated excitatory response of cortical neurons was also reported in several in vitro studies elsewhere [6,17,20]. Taken together, we hypothesize that normal levels of NE act on postsynaptic ␣2 -receptors, while increased levels of NE may lead to the activation of postsynaptic ␣1 -receptors. This occurs especially in stress-induced neurophysiological conditions [3,5]. In the present study, we described dose dependent, long lasting inhibitory effects of YOH. Similar inhibitory action of BRL was also observed, though it was less effective and shorter lasting on the neurons than that of YOH inhibitions, maybe due to its lower binding affinity on ␣2A/D -receptors (Ki,BRL = 7.17 nM, Ki,YOH = 0.42 nM) [25,32]. Further available data suggests that YOH and its dimers have 4.8× or higher affinity on ␣2A receptors than on ␣2B - or ␣2C -receptors [11,32]. According to numerous in vitro studies, biochemical features of different agents displacing [3 H]yohimbine from ␣2 -adrenoceptors in the neocortex are similar to those of BRL or other ␣2A/D -receptor selective agents (oxymethazoline) and, they are different from selective ␣2B -receptor agents such as ARC239 [13]. Although this type of BRL-like action of YOH was suggested in several in vitro experiments [5,31], the present study is the first to demonstrate direct in vivo electrophysiological evidence supporting putative ␣2A/D -receptor selectivity of NEergic receptor antagonist YOH. We presented here primary in vivo electrophysiological evidence for the action of ␣2 -receptor antagonist YOH in the mPFC

257

of the rat. Contrary to previous findings suggesting that YOH may act as a general [3,10] or specific ␣2B -receptor antagonist [16,30], in the present paper, based on simultaneous application of known selective ␣2 -receptor agents together with YOH, we demonstrated the primary involvement of ␣2A/D -receptor mechanism in the action of YOH. Although the current results do not resolve the exact localization of putatively involved ␣2A/D receptors, available evidence suggests that the action of YOH as a general ␣2 -antagonist in vivo should be revisited. The possible subtype-selectivity, as well as pre- and postsynaptic action of YOH should be brought to attention in further in vivo pharmacological studies concerning the NEergic mechanisms of the mammalian PFC. Acknowledgements The authors thank Dr. Philippe Tobler for critically reviewing the manuscript, and Dr. E. Poll´ak and T. Atlasz for valuable discussion. Financial support was given by the MTA-PTE Adaptational Biology Research Group (P.K.), OTKA 029818 (I.H.) and the Gedeon Richter Centenary Fund (I.H.). References [1] C. Aoki, C. Venkatesan, C.G. Go, R. Forman, H. Kurose, Cellular and subcellular sites for noradrenergic action in the monkey dorsolateral prefrontal cortex as revealed by the immunocytochemical localization of noradrenergic receptors and axons, Cereb. Cortex 8 (1998) 269–277. [2] D. Arcos, A. Sierra, A. Nu˜nez, G. Flores, J. Aceves, J. Arias-Monta˜no, Noradrenaline increases the firing rate of a subpopulation of rat subthalamic neurones through the activation of ␣1 -adrenoceptors, Neuropharmacology 45 (2003) 1070–1079. [3] A.F.T. Arnsten, R. Mathew, R. Ubriani, J.R. Taylor, B.M. Li, ␣-1 noradrenergic receptor stimulation impairs prefrontal cortical cognitive function, Biol. Psychiatry 45 (1999) 26–31. [4] D.B. Bylund, Subtypes of ␣1 - and ␣2 -adrenergic receptors, FASEB J. 6 (1992) 832–839. [5] L.F. Callado, J.A. Stamford, ␣2A - but not ␣2B/C -adrenoceptors modulate noradrenaline release in rat locus coeruleus: voltametric data, Eur. J. Pharmacol. 366 (1999) 35–39. [6] B. Carette, Noradrenergic responses of neurones in the mediolateral part of the lateral septum: ␣1 -adrenergic depolarization and rhythmic bursting activities, and ␣2 -adrenergic hyperpolarization from guinea pig brain slices, Brain Res. Bull. 48 (1999) 263–276. [7] D.S. Charney, G.R. Heninger, D.E. Redmond Jr., Yohimbine induced anxiety and increased noradrenergic function in humans: effects of diazepam and clonidine, Life Sci. 33 (1983) 19–29. [8] J. Dessaint, W. Yu, J.E. Krause, L. Yue, Yohimbine inhibits firing activities of rat dorsal root ganglion neurons by blocking Na+ channels and vanilloid VR1 receptors, Eur. J. Pharmacol. 485 (2004) 11–20. [9] M.V. Donoso, A. Carvajal, A. Paredes, A. Tomic, C.S. Koenig, J.P. Huidobro-Toro, alpha2-Adrenoceptors control the release of noradrenaline but not neuropeptide Y from perivascular nerve terminals, Peptides 23 (2002) 1663–1671. [10] K.S. Elmslie, D.H. Cohen, Iontophoresis of norepinephrine onto the pigeon’s lateral geniculate nucleus: characterization of an inhibitory response, Brain Res. 517 (1990) 134–142. [11] H.O. Kalkman, E. Loetscher, ␣2C -Adrenoceptor blockade by clozapine and other antipsychotic drugs, Eur. J. Pharmacol. 462 (2003) 33–40. [12] H. Kawahara, Y. Kawahara, B.H. Westerink, The role of afferents to the locus coeruleus in the handling stress-induced increase in the release of noradrenaline in the medial prefrontal cortex: a dual-probe microdialysis study in the rat brain, Eur. J. Pharmacol. 387 (2000) 279–286.

258

P. Kov´acs, I. Hern´adi / Neuroscience Letters 402 (2006) 253–258

[13] J.P. Kiss, G. Zsilla, A. Mike, T. Zelles, E. Toth, A. Lajtha, E.S. Vizi, Subtype-specificity of the presynaptic ␣2-adrenoceptors modulating hippocampal norepinephrine release in rat, Brain Res. 674 (1995) 238–244. [14] P. Kov´acs, V. D´enes, L. Kell´enyi, I. Hern´adi, Microiontophoresis electrode location by neurohistological marking: comparison of four native dyes applied from current balancing electrode channels, J. Pharmacol. Toxicol. Methods 51 (2005) 147–151. [15] P. Kov´acs, I. Hern´adi, Alpha2 antagonist yohimbine suppresses maintained firing of rat prefrontal neurons in vivo, Neuroreport 14 (2003) 833–836. [16] E. MacDonald, B.K. Kobilka, M. Scheinin, Gene targeting—homing in on ␣2 -adrenoceptor-subtype function, Trends Pharmacol. Sci. 18 (1997) 211–219. [17] G.J. Marek, G.K. Aghajanian, Alpha 1B-adrenoceptor-mediated excitation of piriform cortical interneurons, Eur. J. Pharmacol. 305 (1996) 95–100. [18] W. Marrs, J. Kuperman, T. Avedian, R.H. Roth, J.D. Jentsch, Alpha-2 adrenoceptor activation inhibits phencyclidine-induced deficits of spatial working memory in rats, Neuropsychopharmacology 30 (2005) 1500–1510. [19] M.F. McCarty, Pre-exercise administration of yohimbine may enhance the efficacy of exercise training as a fat loss strategy by boosting lipolysis, Med. Hypotheses 58 (2002) 491–495. [20] R.D. Mouradian, F.M. Sessler, B.D. Waterhouse, Noradrenergic potentiation of excitatory transmitter action in cerebrocortical slices: evidence for mediation by an alpha 1 receptor-linked second messenger pathway, Brain Res. 546 (1991) 83–95. [21] C.R. Park, A.M. Campbell, D.M. Diamond, Chronic psychosocial stress impairs learning and memory and increases sensitivity to yohimbine in adult rats, Biol. Psychiatry 50 (2001) 994–1004. [22] G. Paxinos, C. Watson, The Rat Brain in Stereotaxic Coordinates, Compact third ed., Academic Press, London, 1997. [23] A.M. Penna, S.Y. Lee, L. Scheidegger da Silva, R.W. Oliveira, C. de Freitas Gomes, E.M. Nakamura-Palacios, Behavioral and cognitive effects

[24]

[25]

[26]

[27]

[28] [29]

[30]

[31]

[32]

produced by electrical stimulation in the medial prefrontal cortex: an experimental model for high cortical activation, Neuropsychobiology 38 (1998) 241–250. C. Plewnia, M. Bartels, L. Cohen, C. Gerloff, Noradrenergic modulation of human cortex excitability by the presynaptic ␣2 -antagonist yohimbine, Neurosci. Lett. 307 (2001) 41–44. A. Renouard, P.S. Widdowson, M.J. Millan, Multiple alpha2 adrenergic receptor subtypes. I. Comparison of H RX821002-labelled rat RAlpha-2A adrenergic receptors in cerebral cortex to human HAlpha-2A adrenergic receptor and other populations of alpha-2 adrenergic subtypes, J. Pharmacol. Exp. Ther. 270 (1994) 946–957. T. Sawaguchi, Attenuation of delay-period activity of monkey prefrontal neurons by an alpha2-adrenergic antagonist during an oculomotor delayed-response task, J. Neurophysiol. 80 (1998) 2200–2205. J.R. Simpson Jr., W.C. Drevets, A.Z. Snyder, D.A. Gusnard, M.E. Raichle, Emotion-induced changes in human medial prefrontal cortex: II. During anticipatory anxiety, Proc. Natl. Acad. Sci. U.S.A. 98 (2001) 688–693. S.W. Tam, M. Worcel, M. Wyllie, Yohimbine: a clinical review, Pharmacol. Ther. 91 (2001) 215–243. A.U. Trendelenburg, W. Klebroff, L. Hein, K. Starke, A study of presynaptic alpha2-autoreceptors in alpha2A/D-, alpha2B- and alpha2Cadrenoceptor-deficient mice, Naunyn Schmiedebergs Arch. Pharmacol. 364 (2001) 117–130. S. Uhlen, R. Muceniece, N. Rangel, G. Tiger, J.E. Wikberg, Comparison of the binding activities of some drugs on alpha 2A, alpha 2B and alpha 2C-adrenoceptors and non-adrenergic imidazoline sites in the guinea pig, Pharmacol. Toxicol. 6 (1995) 353–364. E. Umeda, T. Satoh, H. Nagashima, P.E. Potter, G. Tarkovacs, E.S. Vizi, ␣2A subtype of presynaptic ␣2 -adrenoceptors modulates the release of [3H]-noradrenaline from rat spinal cord, Brain Res. Bull. 42 (1997) 129–132. W. Zheng, L. Lei, S. Lalchandani, G. Sun, R.D. Feller, D. Miller, Yohimbine dimers exhibiting selectivities for human ␣2a -versus ␣2b -adrenergic receptors, Bioorg. Med. Chem. Lett. 10 (2000) 627–630.