Zaprinast stimulates extracellular adenosine accumulation in rat pontine slices

Neuroscience Letters 371 (2004) 12–17

Zaprinast stimulates extracellular adenosine accumulation in rat pontine slices Minou Lea,1 , Yin Lua,1 , Ya Lia , Robert W. Greeneb , Paul M. Epsteinc , Paul A. Rosenberga,∗ a

Department of Neurology and Program in Neuroscience, Enders Research Building, Room 349, Children’s Hospital and Harvard Medical School, Longwood Avenue, Boston, MA 02115, USA b Department of Psychiatry and Dallas VAMC, UTSW, Dallas, TX 75216, USA c Department of Pharmacology, University of Connecticut Health Center, Farmington, CT 06030, USA Received 11 June 2004; received in revised form 28 July 2004; accepted 28 July 2004

Abstract Adenosine appears to be an endogenous somnogen. The lateral dorsal tegmental/pedunculopontine nucleus (LDT/PPT) located in the mesopontine tegmentum is important in the regulation of arousal. Neurons in this nucleus are strongly hyperpolarized by adenosine and express neuronal nitric oxide synthase. Zaprinast is a cyclic nucleotide phosphodiesterase inhibitor, and has been shown in the hippocampal slice to inhibit the field excitatory postsynaptic potential. This action could be blocked by an adenosine receptor antagonist, and therefore is presumably due to adenosine release stimulated by zaprinast. In the present study we tested the effect of zaprinast on extracellular adenosine accumulation in pontine slices containing the LDT. Zaprinast at 10 ␮M evoked an increase in extracellular adenosine concentration. This effect was blocked by impermeant inhibitors of 5 -nucleotidase, indicating that the extracellular adenosine was derived from extracellular AMP. However, inhibitors of cAMP degradation had little or no effect on zaprinast-evoked adenosine accumulation, suggesting that extracellular cAMP was not the source. Removal of extracellular calcium inhibited the effect of zaprinast. These results demonstrate that a pathway exists by which zaprinast stimulates extracellular adenosine accumulation, and the presence of this pathway in the pontine slice suggests the possibility that it may be relevant for the regulation of behavioral state. © 2004 Elsevier Ireland Ltd. All rights reserved. Keywords: Adenosine; Sleep; Zaprinast; cAMP; cGMP; Cyclic nucleotide phosphodiesterase

The LDT and PPT nuclei are neighboring anatomically and functionally similar pontine nuclei that contain most of the brainstem cholinergic neurons. Compelling evidence associates these nuclei with arousal mechanisms [21,22,33,37,38]. Evidence suggesting an effect of adenosine on behavioral state derives in part from the widely appreciated stimulant effect of caffeine and other methylxanthines that block adenosine receptors [36]. It has been observed using in vivo microdialysis techniques that extracellular adenosine concentrations in the basal forebrain (BF) increased during wakefulness, and decreased during sleep. Furthermore, an ∗ 1

Corresponding author. Tel.: +1 617 355 6962; fax: +1 617 730 0243. E-mail address: [email protected] (P.A. Rosenberg). These two authors contributed equally.

0304-3940/$ – see front matter © 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2004.07.090

adenosine transport inhibitor introduced unilaterally into the BF produced both significant increases in extracellular adenosine and slow wave sleep [25]. Another site of adenosine action that could contribute to changes in behavioral state, and that would be expected to have high efficacy in doing so would be the cholinergic neurons of the LDT/PPT that controls thalamocortical activation. In fact, whole-cell and extracellular recordings in an LDT/PPT slice showed that mesopontine cholinergic neurons are under tonic inhibitory control by endogenous adenosine, mediated postsynaptically by the inwardly rectifying potassium conductance and by inhibition of IH [26]. Therefore, adenosine is a strong candidate for an endogenous sleep regulatory substance with important actions in the BF and the LDT/PPT. In our previous experiments using rat forebrain neuronal cultures, we obtained evidence for a mechanism of extracel-

M. Le et al. / Neuroscience Letters 371 (2004) 12–17

lular adenosine accumulation coupled to cyclic nucleotide metabolism [27–30,32]. This pathway seems to be involved in the regulation of extracellular adenosine levels by catecholamines and neuropeptides, and has been shown to be operative in more intact preparations [3,6,11,20,41]. Others also have provided indirect evidence that zaprinast, an inhibitor of cyclic nucleotide phosphodiesterases (PDEs), stimulates extracellular adenosine accumulation in the hippocampal slice [4,5]. We have shown recently [18] that zaprinast stimulates extracellular adenosine accumulation in cultured rat forebrain neurons. These observations prompted us to test the effect of zaprinast on extracellular adenosine in a more intact preparation, such as LDT slices. We have obtained direct evidence, reported here, that zaprinast evokes extracellular adenosine accumulation. These results have been reported in preliminary form [12,17]. The methods for preparation and handling of the pontine slice have been previously described (see below) [16,26], and were conducted according to the ethical standards of the Children’s Hospital Animal Care and Use Committee. All efforts were made to minimize animal suffering. The LDT slices were made from the upper pons and midbrain region of one Sprague–Dawley rat (P19–P30). Tissue was sliced using a vibratome, and slices (300 ␮m) were blocked to include the central gray containing the LDT. This was done by taking slices caudal to the superior cerebellar peduncle, and cutting away all tissue dorsal to the dorsal raphe (a transverse cut was made at the level of the aqueduct). Slices were further blocked by cutting away tissue lateral and ventral to the central gray. The resulting slices contained almost all of the LDT, but also some medial structures, in particular the pericentral dorsal tegmental nucleus. Four to six slices from one animal were used per experiment. Slices were placed in artificial cerebrospinal fluid (ACSF) containing (in mM): NaCl, 124; KCl, 2.5; CaCl, 2; MgSO4 , 1.3; NaH2 PO4 , 1; glucose, 10; and NaHCO3 , 26; equilibrated to pH 7.4 with a constant supply of 95% O2 , 5% CO2 [11], and were allowed to recover for 90 min at room temperature. Slices were then transferred to a superfusion chamber with small volume (0.5 ml), and were trapped by a nylon net. Slices were perfused with ACSF containing 10 ␮M nitrobenzylthoinosine (NBTI) to block reuptake of adenosine [40], and 10 ␮M deoxycoformycin to block adenosine deaminase catalyzing degradation of adenosine [35]. These two drugs were used to allow the measurement of adenosine production uncomplicated by adenosine clearance mechanisms. We found that while the zaprinast effect on adenosine occurred in the absence of isobutylmethylxanthine (IBMX), a non-specific PDE inhibitor [2], it was more robust in the presence of IBMX. Therefore, in most experiments, 100 ␮M IBMX was also added. Slices were perfused at a rate of 1 ml/min. Inclusion of any additional drug depended on the nature of the experiment. After a 10-min equilibration, collection of perfusion fluid was begun, in 1 ml aliquots. After 15 min to establish a baseline, 10 ␮M zaprinast plus 100 ␮M IBMX were added to the superfusion fluid. Application of drug was continued for 5 min, washed out for 15 min, and then

13

100 ␮M veratridine was added for an additional 5 min to verify slice viability. Slices were perfused for a total of 50 min. Inclusion of the LDT in slices used in experiments was verified by fixing and staining for nitric oxide synthase [8,13,14]. After the conclusion of the experiment, slices were placed in 4% paraformaldehyde at 4 ◦ C, incubated overnight, washed three times with PBS, and placed in PBS/Triton X-100 (1%) for 30 min. Slices were then transferred to a dish with PBS/Triton X-100 containing 1 mg/ml NADPH, 0.1 mg/ml nitro blue tetrazolium, and incubated at 37 ◦ C. After the LDT turned dark purple, which occurred after approximately 1–2 h, slices were washed three times with PBS, and mounted on slides. For biochemical assays, perfusion fluid in 1-min (1 ml) samples was collected on ice upon exiting from the chamber. Each collection tube contained EDTA to produce a final concentration of 10 mM to block any calcium-dependent PDEs that might be present. Medium was assayed by HPLC for adenine containing compounds exactly as described previously [30]. The method described by Thompson et al. [39] was followed closely to measure cAMP- and cGMP-PDE activity. In brief, we dissected the central gray containing LDT from Sprague–Dawley rats (P26) as noted above except that tissue was not sliced. Tissue was added to a homogenization buffer containing: Tris–HCl, 40 mM (pH 7.5); EDTA, 1 mM; DTT, 1 mM; benzamidine, 15 mM; PMSF, 35 ␮g/ml; leupeptin, 1 ␮g/ml; antipain, 1 ␮g/ml; pepstatin, 1 ␮g/ml; and aprotinin, 1 ␮g/ml, homogenized (3–5 s) and then centrifuged (7 min, 1000 rpm). Supernatant from this centrifugation was used as the enzyme source. Enzyme reactions (0.4 ml total volume) were carried out in glass tubes and contained: Tris–HCl, 40 mM (pH 8.0); MgCl2 , 10 mM; ␤mercaptoethanol, 2.5 mM; and [3 H] cAMP or [3 H]cGMP, 3 × 105 or 1 × 106 cpm; cAMP or cGMP, 2 ␮M; bovine serum albumin, 50 ␮g/ml; and diluted supernatant, 100 ␮l. The protein concentration of the supernatant was 1.04 mg/ml. For the cAMP-PDE assay the supernatant was diluted 1:100. For the cGMP-PDE assay, the supernatant was diluted 1:30. Various combinations of PDE inhibitors were also added to the mixture. The tubes were immediately placed in a water bath (30 ◦ C, 10 min) to activate the reaction, transferred to a boiling bath (100 ◦ C, 1 min) to stop the reaction, and then placed on ice. Snake venom [100 ␮l of a 1-mg/ml solution (Sigma V-0376)], as a source of 5 -nucleotidase to convert AMP to adenosine, was added. Tubes were heated in a bath (30 ◦ C, 10 min), placed on ice, and 1 ml of methanol was then added. To separate reaction products from substrate, the entire contents were passed through a Dowex 1-X2 column positioned over a scintillation vial, aided by additional methanol (1 ml) and vacuum. Scintillation fluid (8 ml) was added and vials were counted for 10 min. NBTI, IBMX, veratridine, snake venom, ␣␤methylene-ADP (AMPCP), and ethylene glycol-bis(␤-aminoethylether)-N,N,N ,N -tetraacetic acid (EGTA) were obtained from Sigma (St. Louis, MO). Deoxyco-

14

M. Le et al. / Neuroscience Letters 371 (2004) 12–17

formycin was purchased from CalBiochem (La Jolla, CA). Zaprinast, RO20-1724, 8-methoxymethyl-IBMX, and dipyridamole were purchased from Biomol (Plymouth Meeting, PA). UK114,542 was a gift from Pfizer Inc. MK801, CNQX, and CGS19755 were obtained from Research Biochemicals International (Natick, MA). GMP was purchased from Roche Diagnostics Corporation (Indianapolis, IN). All other chemicals were purchased from Sigma.

Statistical comparisons were performed by two-tailed Student’s t-test using the Instat 2 program from GraphPad Inc. (San Diego, CA). To investigate the effect of zaprinast in a more intact preparation relevant to the regulation of behavioral states, we tested whether zaprinast would elicit an increase in extracellular adenosine in the pontine slices. All experiments, except as noted, were performed with 10 ␮M zaprinast and 100 ␮M IBMX in the perfusion fluid. As shown in Fig. 1A, zaprinast-evoked extracellular adenosine accumulation, peaking at 5–10 min following the initial application of the drug, and gradually returned to baseline. In order to quantify the increase we used the percent increase of the zaprinast effect over the baseline, where the zaprinast value represents the average of the peak and the four values surrounding the peak, and the baseline value represents the average of minutes 8–13 prior to application of zaprinast. In the four experiments that were performed, the mean baseline adenosine concentration was 6.1 ± 1.5 nM. Zaprinast increased the extracellular adenosine 98–142%, with a mean percent increase of 120 ± 9.6% (13.5 ± 3.6 nM, p < 0.001). In all slice experiments, cAMP and AMP concentrations were below the detection of HPLC. For comparison, we tested the effect of zaprinast in the absence of IBMX. In three experiments, we found a basal concentration of adenosine of 3 ± 0.6 nM. In each of these three experiments, zaprinast significantly (p < 0.05) increased the concentration of adenosine to a mean maximal value of 5.2 ± 0.9 nM (73% increase). There are several routes by which extracellular adenosine may be increased, including degradation of extracellular cAMP transported out of cells, degradation of extracellular ATP, and transport of adenosine itself across the plasma membrane. In the cases of degradation of extracellular ATP and cAMP, the last step leading to the production of extra-

Fig. 1. Zaprinast increased extracellular adenosine accumulation in pontine slices by a mechanism that involves extracellular AMP but not cAMP. In all experiments, slices were perfused with ACSF at a rate of 1 ml/min. After 15 min, 10 ␮M zaprinast plus 100 ␮M IBMX was added to the perfusion fluid (Z—first down arrow) for 5 min (first up arrow), washed out for 15 min, and then veratridine (100 ␮M) was added for 5 min (V—second down arrow). In experiments involving the inclusion of an inhibitor, these drugs were added to the perfusion fluid during minutes 0–35 and were absent during the application of veratridine. (A) Zaprinast evoked an increase in extracellular adenosine. In this experiment there was a 118% increase in extracellular adenosine over baseline levels (baseline: 11 ± 2 nM, peak: 24 ± 2.1 nM, p < 0.0001). Veratridine evoked a large increase in extracellular adenosine. This experiment is representative of four that were performed. (B) AMPCP (5 ␮M, present throughout the duration of the experiment) completely blocked the effect of zaprinast. Adenosine was not detectable before the application of veratridine. A similar result was observed in another experiment. (C) RO20-1724 (200 ␮M, present throughout the duration of the experiment) did not block the effect of zaprinast. The first downward and upward arrows indicate the addition and removal of zaprinast while the second downward and upward arrows indicate the addition and removal of forskolin (100 ␮M in ACSF solution containing 10 ␮M NBTI, 10 ␮M deoxycoformycin, and 100 ␮M IBMX), which was added to stimulate the adenylate cyclase. In this experiment, there was a 215% increase in extracellular adenosine in the presence of RO20-1724.

M. Le et al. / Neuroscience Letters 371 (2004) 12–17

cellular adenosine is the hydrolysis of AMP to adenosine by 5 -ectonucleotidase. This enzyme is inhibited by AMPCP [19,24]; therefore, AMPCP should prevent an increase in extracellular adenosine if it is derived from extracellular ATP or cAMP. Indeed, AMPCP (5 ␮M) completely blocked the zaprinast-evoked accumulation of extracellular adenosine (Fig. 1B). As shown in Fig. 1B, in the presence of AMPCP, there was no detectable extracellular adenosine (<0.2 nM) in the presence or absence of zaprinast. In three additional experiments, we tested 5 mM GMP, another 5 -nucleotidase inhibitor [19,24,31,32], and observed no detectable extracellular adenosine in the presence or absence of zaprinast (data not shown). From these data we conclude that extracellular adenosine accumulation evoked by zaprinast is not due to the transport of adenosine itself across the plasma membrane, but rather derives from the hydrolysis of extracellular AMP. To further investigate whether the adenosine that accumulated in response to zaprinast was derived from extracellular cAMP, PDE inhibitors were tested for their ability to block the conversion of cAMP to AMP. If formation of adenosine were dependent upon hydrolysis of cAMP by PDE, then inhibition of this enzyme would prevent the accumulation of adenosine. RO20-1724 has been shown to block the formation of adenosine from cAMP in vitro [30] and in vivo [3], but did not block the zaprinast-evoked increase of adenosine when present throughout the duration of the experiments in this study. Fig. 1C represents one of four experiments that were performed. In this experiment, the baseline in the presence of RO20-1724 was 3.2 ± 0.2 nM and the peak after zaprinast addition was 10 ± 1.1 nM, a 213% increase (p < 0.0001). In four experiments, we obtained similar results, with a mean of 240 ± 42% increase in adenosine in the presence of RO20-1724 (200 ␮M). We also directly measured extracellular cAMP in these experiments by enzyme-linked immunoassay. No consistent changes in cAMP levels were detected (data not shown). To test the possibility that ATP released from vesicles is the source for extracellular adenosine, the calcium chelator, EGTA was tested. EGTA (1 mM) significantly blocked the effect of zaprinast (Fig. 2A). In three experiments, basal adenosine was 4.7 ± 1.9 nM, and peak adenosine was 5.8 ± 1.8 nM, representing an increase of 23 ± 23%, which was significantly reduced when compared to zaprinast alone in the absence of EGTA (23 ± 23% versus 120 ± 9.6%, p < 0.05). Using calcium-free perfusate with 5 mM EGTA, the zaprinast effect was completely blocked (data not shown). We, therefore, conclude that zaprinast-evoked accumulation of adenosine is a calcium-dependent event. In our neuronal cultures, we found that zaprinast-evoked extracellular adenosine accumulation was blocked by NMDA receptor antagonists, MK801 and APV, and was reduced significantly by a non-NMDA receptor antagonist, NBQX [18]. To test the effect of glutamate receptor antagonists on zaprinast-mediated adenosine accumulation in pontine slices, we performed experiments using a combination of glutamate receptor blockers: 10 ␮M MK801 [42] (n = 2); 10 ␮M

15

Fig. 2. The effect of zaprinast on extracellular adenosine was dependent upon extracellular calcium. (A) The effect of zaprinast was diminished in Ca-free ACSF. The increase for this experiment was 25%, which was significantly lower than the average increase for zaprinast experiments performed in calcium containing medium (p < 0.05). This experiment is representative of three that were performed. (B) The effect of zaprinast was not dependent upon NMDA receptor activation. Here, in the presence of 10 ␮M MK801 (throughout the duration of the experiment), zaprinast evoked a 134% increase in extracellular adenosine. This experiment was carried out in the presence of calcium.

MK801 plus 50 ␮M CNQX [43] (n = 2); or 10 ␮M MK801 plus 50 ␮M CNQX plus 50 ␮M CGS19755 [23] (n = 1). Slices were exposed to inhibitors throughout the duration of the experiment. As shown in Fig. 2B, zaprinast evoked an increase in extracellular adenosine in the presence of MK801 (134% increase; p < 0.0001). Similar results were observed in all five experiments. Thus, zaprinast stimulated 152 ± 25% increase of adenosine in the presence of glutamate receptor antagonists, which was not significantly different from zaprinast alone (120 ± 9.6%).

16

M. Le et al. / Neuroscience Letters 371 (2004) 12–17

To test whether the effect of zaprinast in pontine slices was due to an action on PDEs, we attempted to block all PDE activity in the slices using combinations of PDE inhibitors. We tested the effect of various PDE inhibitors on cAMP and on cGMP hydrolysis in homogenates of the same region from which the slices were obtained. We found that the combination of 8-methoxymethyl-IBMX (50 ␮M), UK114,542 (60 ␮M) and dipyridamole (100 ␮M) with IBMX (100 ␮M) suppressed 99% and 97% of the hydrolysis of cAMP and cGMP, respectively. We then tested this combination of inhibitors on the intact pontine slices. All inhibitors were present throughout the duration of the experiment. In the two experiments that were performed, the baseline was 17.7 ± 5.3 nM. Thus the inhibition of PDE activity significantly increased baseline adenosine concentration (190 ± 86% increase, p < 0.05). In these experiments, zaprinast evoked an increase in extracellular adenosine to 24.8 ± 8.1 nM (40 ± 4% increase, p < 0.05). We report here that it was possible to detect adenosine in the superfusion fluid bathing pontine slices. More importantly, we found that zaprinast stimulated adenosine accumulation in these slices. The zaprinast effect was potentiated in the presence of IBMX, which might be due to its ability either to block adenosine receptors or to inhibit PDEs [7]. The PDE target for zaprinast in pontine slices is not clear, but possible candidate(s) are: PDE1 [34], PDE5 [15], or PDE9 [1,10] since these PDEs are present in the brain and are inhibited by zaprinast. In experiments using a concentration of PDE inhibitors that blocked 97–99% of cGMP and cAMP hydrolyzing activity in homogenates derived from the pons, we found a significant elevation of extracellular adenosine, suggesting that PDE activity in this region could be involved in the regulation of extracellular adenosine. However, addition of zaprinast still evoked an increase in adenosine, albeit diminished, suggesting that some of the effect of zaprinast might be due to actions other than inhibition of PDE. The immediate source of adenosine that accumulated in response to zaprinast appears to be extracellular AMP, since AMPCP and GMP, inhibitors of 5 -nucleotidase, blocked the zaprinast effect. Furthermore, this AMP is most likely derived from ATP not cAMP, because RO20-1724 failed to block zaprinast-evoked adenosine accumulation. ATP is a possible source of the adenosine that accumulates in response to zaprinast stimulation. ATP is co-released with neurotransmitters by a calcium-dependent mechanism, and therefore removal of extracellular calcium would be expected to block the accumulation of adenosine if the adenosine derived from extracellular ATP. We found that the effect of zaprinast was indeed diminished in calcium-free perfusate suggesting that a calcium-dependent process is involved in the effect of zaprinast on extracellular adenosine. In neuronal cultures, we found that zaprinast-evoked adenosine accumulation was mediated by increase of glutamate release and activation of NMDA receptors [18]. However, in pontine slices, zaprinast stimulated adenosine accumulation was not blocked by the NMDA receptor antagonist

MK801, either alone or in combination with non-NMDA receptor antagonists suggesting that zaprinast-evoked adenosine accumulation is mediated by different mechanisms in neuronal cultures and in pontine slices. The basal adenosine concentration measured in the perfusate of LDT slices was 3–6 nM. Since the perfusate of slices greatly expands the volume of the extracellular space, measurement of adenosine in the perfusate is not a reliable measure of extracellular adenosine. Considering the volume expansion involved, the adenosine concentration measured in the perfusate in the present study is consistent with previous estimates of adenosine concentration in the extracellular space (140–200 nM) [9]. In summary, we have found that zaprinast, a PDE inhibitor, consistently evoked extracellular adenosine accumulation derived from extracellular AMP but not cAMP by a process that appears to be calcium-dependent. The mechanism of this effect is as yet unknown, but may be important in the physiological regulation of extracellular adenosine concentration. Acknowledgments This work was supported by the National Heart, Lung, and Blood Institute (HL 59595 and HL 60292), the National Institute of Child Health and Human Development (HD 18655), and from the Department of Veteran Affairs. References [1] S.G. Andreeva, P. Dikkes, P.M. Epstein, P.A. Rosenberg, Expression of cGMP-specific phosphodiesterase 9A mRNA in the rat brain, J. Neurosci. 21 (2001) 9068–9076. [2] J.A. Beavo, D.H. Reifsnyder, Primary sequence of cyclic nucleotide phosphodiesterase isozymes and the design of selective inhibitors, Trends Pharmacol. Sci. 11 (1990) 150–155. [3] A. Bonci, J.T. Williams, A common mechanism mediates long-term changes in synaptic transmission after chronic cocaine and morphine, Neuron 16 (1996) 631–639. [4] C.L. Boulton, A.J. Irving, E. Southam, B. Potier, J. Garthwaite, G.L. Collingridge, The nitric oxide–cyclic GMP pathway and synaptic depression in rat hippocampal slices, Eur. J. Neurosci. 6 (1994) 1528–1535. [5] M.R. Broome, G.L. Collingridge, A.J. Irving, Activation of the NOcGMP signalling pathway depresses hippocampal synaptic transmission through an adenosine receptor-dependent mechanism, Neuropharmacology 33 (1994) 1511–1513. [6] J.M. Brundege, L.H. Diao, W.R. Proctor, T.V. Dunwiddie, The role of cyclic AMP as a precursor of extracellular adenosine in the rat hippocampus, Neuropharmacology 36 (1997) 1201–1210. [7] O.H. Choi, M.T. Shamim, W.L. Padgett, J.W. Daly, Caffeine and theophylline analogs: correlation of behavioral effects with activity as adenosine receptor antagonists and as phosphodiesterase inhibitors, Life Sci. 43 (1988) 387–398. [8] T.M. Dawson, D.S. Bredt, M. Fotuhi, P.M. Hwang, S.H. Snyder, Nitric oxide synthase and neuronal NADPH diaphorase are identical in brain and peripheral tissues, Proc. Natl. Acad. Sci. U.S.A. 88 (1991)7797–7801. [9] T.V. Dunwiddie, L. Diao, Extracellular adenosine concentrations in hippocampal brain slices and the tonic inhibitory modulation of

M. Le et al. / Neuroscience Letters 371 (2004) 12–17

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22] [23]

[24]

[25]

[26]

evoked excitatory responses, J. Pharmacol. Exp. Ther. 268 (1994) 537–545. D.A. Fisher, J.F. Smith, J.S. Pillar, S.H. St. Denis, J.B. Cheng, Isolation and characterization of PDE9A, a novel human cGMP-specific phosphodiesterase, J. Biol. Chem. 273 (1998) 15559–15564. R.W. Gereau, P.J. Conn, Potentiation of cAMP responses by metabotropic glutamate receptors depresses excitatory synaptic transmission by a kinase-independent mechanism, Neuron 12 (1994) 1121–1129. E.F. Hearn, H.C.R. Grunze, D.G. Rainnie, P.A. Rosenberg, R.W. McCarley, R.W. Greene, Control of extracellular adenosine levels in the laterodorsal/pedunculopontine tegementum nuclei (LDT7PPT), Soc. Neurosci. Abstr. 21 (1995) 366. B.T. Hope, G.J. Michael, K.M. Knigge, S.R. Vincent, Neuronal NADPH diaphorase is a nitric oxide synthase, Proc. Natl. Acad. Sci. U.S.A. 88 (1991) 2811–2814. J.-Y. Koh, D.W. Choi, Vulnerability of cultured cortical neurons to damage by excitotoxins: differential susceptibility of neurons containing NADPH-diaphorase, J. Neurosci. 8 (1988) 2153–2163. J. Kotera, N. Yanaka, K. Fujishige, Y. Imai, H. Akatsuka, T. Ishizuka, K. Kawashima, K. Omori, Expression of rat cGMP-binding cGMPspecific phosphodiesterase mRNA in Purkinje cell layers during postnatal neuronal development, Eur. J. Biochem. 249 (1997) 434–442. C.S. Leonard, R. Llinas, Serotonergic and cholinergic inhibition of mesopontine cholinergic neurons controlling REM sleep: an in vitro electrophysiological study, Neuroscience 59 (1994) 309–330. Y. Li, H.J. Chung, X. Gan, R.W. Greene, P.A. Rosenberg, Zaprinast evokes extracellular adenosine accumulation in slices of the lateral dorsal tegmental nucleus, Soc. Neurosci. Abstr. 23 (1997) 1782. Y. Lu, Y. Li, G.A. Herin, E. Aizenman, P.M. Epstein, P.A. Rosenberg, Elevation of intracellular cAMP evokes activity-dependent release of adenosine in cultured rat forebrain neurons, Eur. J. Neurosci. 19 (2004) 2669–2681. W.F. MacDonald, T.D. White, Nature of extrasynaptosomal accumulation of endogenous adenosine evoked by K+ and veratridine, J. Neurochem. 45 (1985) 791–797. O. Manzoni, D. Pujalte, J. Williams, J. Bockaert, Decreased presynaptic sensitivity to adenosine after cocaine withdrawal, J. Neurosci. 18 (1998) 7996–8002. D.A. McCormick, Neurotransmitter actions in the thalamus and cerebral cortex and their role in neuromodulation of thalamocortical activity, Prog. Neurobiol. 39 (1992) 337–388. G. Moruzzi, The sleep-waking cycle, Ergebn. Physiol. 64 (1972) 1–165. D.E. Murphy, A.J. Hutchison, S.D. Hurt, M. Williams, M.A. Sills, Characterization of the binding of [3H]-CGS19755: a novel Nmethyl-d-aspartate antagonist with nanomolar affinity in rat brain, Br. J. Pharmacol. 95 (1988) 932–938. J.D. Pearson, Ectonucleotidases. Measurement of activities and use of inhibitors, in: D.M. Paton (Ed.), Methods in Pharmacology, vol. 6, Plenum Press, New York, 1987, pp. 83–107. T. Porkka-Heiskanen, R.E. Strecker, M. Thakkar, A.A. Bjorkum, R.W. Greene, R.W. McCarley, Adenosine: a mediator of the sleepinducing effects of prolonged wakefulness, Science 276 (1997) 1265–1268. D.G. Rainnie, H.C.R. Grunze, R.W. McCarley, R.W. Greene, Adenosine inhibition of mesopontine cholinergic neurons: implications for EEG arousal, Science 263 (1994) 689–692.

17

[27] P.A. Rosenberg, Functional significance of cyclic AMP secretion in cerebral cortex, Brain Res. Bull. 29 (1992) 315–318. [28] P.A. Rosenberg, M.A. Dichter, Extracellular cAMP accumulation and degradation in rat cerebral cortex in dissociated cell culture, J. Neurosci. 9 (1989) 2654–2663. [29] P.A. Rosenberg, Y. Li, Adenylyl cyclase activation underlies intracellular cyclic AMP accumulation, cyclic AMP transport, and extracellular adenosine accumulation evoked by ␤3-adrenergic receptor stimulation in mixed cultures of neurons and astrocytes derived from cerebral cortex, Brain Res. 692 (1995) 227–232. [30] P.A. Rosenberg, Y. Li, Beta-adrenergic receptor-mediated regulation of extracellular adenosine in cerebral cortex in culture, J. Neurosci. 14 (1994) 2953–2965. [31] P.A. Rosenberg, Y. Li, Forskolin evokes extracellular adenosine accumulation in rat cortical cultures, Neurosci. Lett. 211 (1996) 49–52. [32] P.A. Rosenberg, Y. Li, Vasoactive intestinal peptide regulates extracellular adenosine levels in rat cortical cultures, Neurosci. Lett. 200 (1995) 93–96. [33] K. Semba, H.C. Fibiger, Organization of central cholinergic systems, Prog. Brain Res. 79 (1989) 37–63. [34] R.K. Sharma, A.M. Adachi, K. Adachi, J.H. Wang, Demonstration of bovine brain calmodulin-dependent cyclic nucleotide phosphodiesterase isozymes by monoclonal antibodies, J. Biol. Chem. 259 (1984) 9248–9254. [35] P. Skolnick, Y. Nimitkitpaisan, L. Stalvey, J.W. Daly, Inhibition of brain adenosine deaminase by 2 -deoxycoformycin and erythro9-(2-hydroxy-3-nonyl) adenine, J. Neurochem. 30 (1978) 1579– 1582. [36] S.H. Snyder, Adenosine as a neuromodulator, in: W.M. Cowan, E.M. Shooter, C.F. Stevens, R.F. Thompson (Eds.), Annual Review of Neuroscience, vol. 8, Annual Reviews, Inc., Palo Alto, CA, 1985, pp. 103–124. [37] M. Steriade, S. Datta, D. Pare, G. Oakson, R. Curro Dossi, Neuronal activities in brain-stem cholinergic nuclei related to tonic activation processes in thalamocortical systems, J. Neurosci. 10 (1990) 2541–2559. [38] M. Steriade, D. Pare, S. Datta, G. Oakson, R. Curro Dossi, Different cellular types in mesopontine cholinergic nuclei related to pontogeniculo-occipital waves, J. Neurosci. 10 (1990) 2560–2579. [39] W.J. Thompson, W.L. Terasaki, P.M. Epstein, S.J. Strada, Assay of cyclic nucleotide phosphodiesterase and resolution of multiple molecular forms of the enzyme, Adv. Cyclic Nucleotide Res. 10 (1979) 69–92. [40] J.A. Thorn, S.M. Jarvis, Adenosine transporters, Gen. Pharmacol. 27 (1996) 613–620. [41] D.G. Winder, P.S. Ritch, R.W. Gereau, P.J. Conn, Novel glialneuronal signalling by coactivation of metabotropic glutamate and P-adrenergic receptors in rat hippocampus, J. Physiol. (Lond.) 494 (1996) 743–755. [42] E.H.G. Wong, J.A. Kemp, T. Priestley, A.R. Knight, G.N. Woodruff, L.L. Iversen, The anticonvulsant MK-801 is a potent N-methyld-aspartate antagonist, Proc. Natl. Acad. Sci. U.S.A. 83 (1986) 7104–7108. [43] K.A. Yamada, J.M. Dubinsky, S.M. Rothman, Quantitative physiological characterization of a quinoxalinedione non-NMDA antagonist, J. Neurosci. 9 (1989) 3230–3236.