Received 22 September 2004; received in revised form 1 December 2004; accepted 13 December 2004

Pain 114 (2005) 212 220 www.elsevier.com/locate/pain Inhibitors of serine/threonine protein phosphatases antagonize the antinociception induced by agonists of a 2 adrenoceptors and GABA B but not k-opioid receptors in the tail flick test in mice Ana Moncada, Cruz Miguel Cendán, José M. Baeyens, Esperanza Del Pozo* Department of Pharmacology and Neurosciences Institute, School of Medicine, University of Granada, Avenida de Madrid 12, E-18012 Granada, Spain Received 22 September 2004; received in revised form 1 December 2004; accepted 13 December 2004 Abstract We previously reported that serine/threonine protein phosphatases (PPs) play a role in the antinociception induced by the m-opioid receptor agonist morphine. In this study we evaluated the possible involvement of PPs on the antinociception induced by agonists of others G proteincoupled receptors in the tail flick test in mice. The subcutaneous administration of clonidine (0.25 4 mg/kg), baclofen (2 32 mg/kg) or U50,488H (2 16 mg/kg) (agonists of a 2 adrenoceptors, GABA B and k-opioid receptors, respectively) produced dose-dependent antinociception. The antinociceptive effects of clonidine and baclofen were antagonized in a dose-dependent way by the protein phosphatase inhibitors okadaic acid (0.001 10 pg/mouse, i.c.v.) and cantharidin (0.001 10 ng/mouse, i.c.v.), and okadaic acid was 1000 times more potent than cantharidin in producing this effect. The effects of these drugs appear to be specifically due to the blockade of PPs, since L-norokadaone (an analogue of okadaic acid that has no effect on PPs) did not modify clonidine- or baclofen-induced antinociception over the wide range of doses used (0.001 1000 pg/mouse, i.c.v.). On the other hand, the antinociception induced by activation of k-opioid receptors with U50,488H was not modified by okadaic acid or cantharidin. In conclusion, our data support the idea that serine/threonine PPs are differentially involved in the antinociceptive effects of several agonists of G protein-coupled receptors in mice. q 2004 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. Keywords: Antinociception; G protein-coupled receptors; Serine/threonine protein phosphatases; Okadaic acid; Cantharidin; Tail flick test 1. Introduction Protein phosphorylation and dephosphorylation reactions (catalysed by kinases and phosphatases, respectively) regulate the activity of many proteins that participate in synaptic transmission processes in the central nervous system, such as ion channels (Herzig and Neumann, 2000), receptors (Chan and Sucher, 2001; Hemmings et al., 1989) and transporters for neurotransmitters (Foster et al., 2003). The most abundant protein phosphatases in mammalian systems are the serine/threonine protein phosphatases (PPs), which catalyse the dephosphorylation of serine and threonine residues in proteins. They have been * Corresponding author. Tel.: C34 958 243539; fax: C34 958 243537. E-mail address: edpozo@ugr.es (E. Del Pozo). traditionally classified into type 1 (PP1) and type 2 (PP2), depending on their substrate specificity and sensitivity to inhibitors. Type 2 protein phosphatases are subdivided into three major groups: ion-independent PP2A, calciumdependent calcineurin PP2B and magnesium-dependent PP2C (Herzig and Neumann, 2000; Price and Mumby, 1999). Other types of serine/threonine PPs designed PP4, PP5, PP6 and PP7 have been described, but their role, substrate affinity and pharmacology are not well known (Herzig and Neumann, 2000; Huang and Honkanen, 1998; Price and Mumby, 1999). Previous findings by us showed that okadaic acid and cantharidin, both inhibitors of PPs, antagonize the antinociception induced by the m-opioid receptor agonist morphine, in the tail flick test in mice (Moncada et al., 2003). The m-opioid receptor is a member of the family of G protein-coupled receptors (GPCRs), and agonists of several 0304-3959/$20.00 q 2004 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.pain.2004.12.017

A. Moncada et al. / Pain 114 (2005) 212 220 213 GPCRs are known to produce antinociception by modulating the activity of similar G proteins and cellular effectors, including Ca 2C (Del Pozo et al., 1990; Dierssen et al., 1990; Dogrul et al., 2001) and K C channels (for a review see Ocaña et al., 2004). It has been reported that protein phosphatases 1 and 2A influence the activity of some G proteins (Bushfield et al., 1991), as well as that of several types of neuronal Ca 2C channels (Chik et al., 1999; Herzing and Neumann, 2000; Surmeier et al., 1995) and K C channels (Herzing and Neumann, 2000; Mullner et al., 2003). Therefore, we hypothesized that inhibition of the activity of PPs may modulate the antinociceptive effects of drugs acting through different GPCRs. Because this hypothesis has not been previously tested, we evaluated, in the tail flick test in mice, the effects of the inhibitors of PPs okadaic acid and cantharidin on the antinociception induced by clonidine, baclofen or U50,488H. These drugs are prototypic agonists of a 2 adrenoceptors, GABA B and k-opioid receptors, respectively, that have been previously shown to produce antinociception in the tail flick test in mice (Ocaña and Baeyens, 1993; Ocaña et al., 1996). We also tested the effects on the antinociception induced by the agonists of GPCRs of L-norokadaone, an analogue of okadaic acid that does not inhibit PPs (Honkanen et al., 1994), which was used as a negative control in these assays. 2. Methods 2.1. Experimental animals Female CD-1 mice (Charles River, Barcelona, Spain) weighing 25 30 g were used for all the experiments. The animals were housed in a temperature-controlled room (21G1 8C) with air exchange every 20 min and an automatic 12-h light/dark cycle (08.00 20.00 h). They were fed a standard laboratory diet and tap water ad libitum until the beginning of the experiments. The animals were placed in the experimental room 2 h before the antinociceptive test, for adaptation. The experiments were performed during the light phase (09.00 15.00 h). Naive animals were used throughout. Mice were always handled in accordance with ethical principles for the evaluation of pain in conscious animals (Zimmerman, 1983) and with the European Communities Council Directive of 24 November 1986 (86/609/ECC). The experimental work has been revised by the Ethical Research Committee of the University of Granada, Spain. 2.2. Drugs and drug administration The agonists of G protein-coupled receptors were clonidine HCl, baclofen and trans-(g)-3,4-dichloro-n-methyl-n-(2-[1-pirrolidynyl]cyclohexyl) benzeneacetamide methanesulfonate salt (U50,488H). All were provided by Sigma Química, Spain. The inhibitors of PPs used were okadaic acid and cantharidin (Sigma Química); as a negative control we used L-norokadaone (ICN Hubber, Spain). Clonidine, baclofen and U50,488H were dissolved in ultrapure water and injected subcutaneously (s.c.). The inhibitors of PPs (okadaic acid and cantharidin) and the analogue L-norokadaone were dissolved in 1% Tween 80 in ultrapure water, and were injected intracerebroventricularly (i.c.v.). Control animals received the same volume of vehicle. The s.c. injections were done in the interscapular region, in a volume of 5 ml/kg. The i.c.v. injections were done in the right lateral cerebral ventricle in a volume of 5 ml/mouse, according to the method which we previously described in detail (Ocaña et al., 1995). Briefly, the injection site was identified according to the method reported by Haley and McCormick (1957). The drug solution was injected with a 10-ml Hamilton syringe with a sleeve around the needle to prevent the latter from penetrating more than 3 mm into the skull. After the experiments were done, the position of the injection was evaluated in each brain, and the results from animals in which the tip of the needle did not reach the lateral ventricle were discarded. The accuracy of the injection technique was evaluated and the percentage of correct injections was 99%. 2.3. Antinociception experiments The tail flick test was used to evaluate the antinociceptive effects of the drugs. The test was performed as previously described (Moncada et al., 2003; Robles et al., 1996). Briefly, the animals were restrained in a Plexiglas tube and placed on the tail flick apparatus (LI 7100, Letica, S.A., Barcelona, Spain). A noxious beam of light was focussed on the tail about 4 cm from the tip, and the tail flick latency was recorded automatically to the nearest 0.1 s. The intensity of the radiant heat source was adjusted to yield baseline latencies between 3 and 5 s; this intensity was never changed and any animal whose baseline latency was outside the pre-established limits was excluded from the experiments. In order to minimise injury in the animals, a cut-off time of 10 s was used. Baseline tail flick latencies were recorded 10 min before and immediately before all injections. Once baseline latencies were obtained the animals received an s.c. injection of a GCPR agonist or its solvent, and immediately thereafter an i.c.v. injection of an inhibitor of PPs (okadaic acid or cantharidin), L-norokadaone, or their solvent. The end of the last injection was considered time 0; from this time response latency was measured again at 15, 30, 45, 60, 90 and 120 min. The degree of antinociception was calculated from the area under the curve (AUC) of tail flick latency against time, which allowed us to obtain a global value of the antinociception induced during the 2-h experimental period, according to the formula: % antinociceptionz[(auc d KAUC v )/(AUC max KAUC v )]!100, where AUC d and AUC v are the areas under the curve for drugtreated and vehicle-treated animals respectively, and AUC max is the area under the curve of maximum possible antinociception (10 s in each determination). 2.4. Data analysis The values in the control group were compared against those obtained in the treated groups using one-way or two-way analysis of variance (ANOVA) followed by Newman Keuls test. The differences between means were considered significant when the value of P was below 0.05. The ED 50 (dose of agonist drug that produced half of the maximal antinociception) values were calculated from the doseresponse curves using nonlinear regression analysis with the Sigma

214 A. Moncada et al. / Pain 114 (2005) 212 220 Plot 2000 version 6.00 program (SPSS Inc., IL, USA). The area under the curve of tail flick latency against time was calculated with the GraphPad Prism version 3.00 program (GraphPad Sofware Inc., CA, USA). 3. Results 3.1. Effect of inhibitors of PPs on clonidine-induced antinociception We studied the effects on the pain threshold produced by the i.c.v. administration of inhibitors of PPs alone. Neither inhibitor at the highest doses administered (okadaic acid 10 pg/mouse, cantharidin 10 ng/mouse) modified tail flick latency values when given alone, and neither had any hyperalgesic nor analgesic effects (data not shown). L-norokadaone (1 ng/mouse, i.c.v.) was also unable to modify tail flick latency in control animals. The s.c. administration of clonidine (0.25 4 mg/kg) induced a dose-dependent antinociceptive effect in the tail flick test in mice (Fig. 1). Okadaic acid (0.1 and 10 pg/mouse, i.c.v.) produced a dose-dependent inhibition of the clonidine effect, displacing its dose-response curve to the right (Fig. 1A) and increasing the ED 50 of clonidine from 0.41G0.02 (clonidine C vehicle) to 0.57G0.04 and 1.51G 0.33 mg/kg, when clonidine was associated to okadaic acid 0.1 pg/mouse and 10 pg/mouse. Cantharidin (i.c.v.) also produced a dose-dependent inhibition of the clonidine effect and displaced the dose-response curve of clonidine (s.c.) to the right, although the doses of cantharidin required to produce these effects were a thousand times higher than those of okadaic acid. The ED 50 of clonidine was increased from 0.41G0.02 (clonidinecvehicle) to 0.76G0.05 and 1.37G0.18 mg/kg (clonidineccantharidin 0.1 ng/mouse and 10 ng/mouse, respectively) (Fig. 1B). The association of a wide range of doses of okadaic acid (0.001 10 pg/mouse, i.c.v.) or cantharidin (0.01 10 ng/mouse, i.c.v.) together with clonidine (1 mg/kg, s.c.) produced a dose-dependent antagonism of clonidine-induced antinociception, and the dose-response lines of inhibition were parallel (Fig. 2A). Maximal reduction in the antinociceptive effect of clonidine was seen with okadaic acid 10 pg/mouse and cantharidin 10 ng/mouse, which decreased the effect of clonidine 1 mg/kg from 76G 3% (clonidine plus vehicle) to 13G3% (clonidine plus okadaic acid 10 pg/mouse) and 21G4% (clonidine plus cantharidin 10 ng/mouse). On the other hand, L-norokadaone over the large range of doses tested (1 fg/mouse to 1 ng/mouse, i.c.v.) did not significantly modify the antinociception induced by clonidine (1 mg/kg, s.c.) (Fig. 2A). When the time courses of the effects of the different treatment groups were plotted, both okadaic acid (10 pg/mouse, i.c.v.) and cantharidin (10 ng/mouse, i.c.v.) significantly antagonized the antinociception induced by clonidine (1 mg/kg, s.c.) from 45 min until 120 min after Fig. 1. Effects of the i.c.v. administration of okadaic acid and cantharidin, or their vehicle, on the antinociception induced by clonidine (s.c.) in a tail flick test in mice. (A) Effects of clonidine plus: vehicle (C), okadaic acid 0.1 pg/mouse (6) or okadaic acid 10 pg/mouse (:). (B) Effects of clonidine plus: vehicle (C), cantharidin 0.1 ng/mouse (^) or cantharidin 10 ng/mouse (%). The percentage of antinociception was calculated from the area under the curve of antinociception (see Section 2). Each point represents the meangsem (nr8). Statistically significant differences in comparison to clonidine plus vehicle: *P!0.05; **P!0.01 (two-way ANOVA followed by Newman Keuls test). the injection (Fig. 2B). In contrast, L-norokadaone (1 ng/mouse, i.c.v.) did not significantly modify the time course of the antinociceptive effect of clonidine (Fig. 2B). 3.2. Effect of inhibitors of PPs on baclofen-induced antinociception Baclofen (2 32 mg/kg, s.c.) induced a dose-dependent antinociception in the tail flick test in mice (Fig. 3). The i.c.v. injection of okadaic acid (0.01 and 1 pg/mouse) associated to baclofen (s.c.) antagonized the antinociceptive

A. Moncada et al. / Pain 114 (2005) 212 220 215 Fig. 2. (A) Effects of i.c.v. administration of okadaic acid (1 fg/mouse to 10 pg/mouse) (:), cantharidin (10 pg/mouse to 10 ng/mouse, i.c.v.) (%) or L-norokadaone (1 fg/mouse to 1 ng/mouse) (,) on the antinociception induced by clonidine (1 mg/kg, s.c.). The shaded area represents the effect of clonidinecvehicle. The percentage of antinociception was calculated from the area under the curve of antinociception (see Section 2). (B) Time course of the antinociceptive effects obtained with the tail flick test in mice treated with clonidine 1 mg/kg, s.c. plus: vehicle i.c.v. (C), okadaic acid 10 pg/mouse, i.c.v. (:); cantharidin 1 ng/mouse, i.c.v. (%), or L-norokadaone 1 ng/mouse i.c.v. (,). Results are represented as the meangsem (nr8). Statistically significant differences in comparison to clonidine plus vehicle: *P!0.05; **P!0.01 (one-way ANOVA [A] and two-way ANOVA [B] followed by Newman Keuls test). activity of this agent, displacing its dose-response curve to the right and progressively increasing its ED 50 from 4.74G 0.77 (baclofen plus vehicle) to 7.33G1.55 mg/kg (baclofen plus okadaic acid 0.01 pg/mouse) and 11.85G0.07 mg/kg (baclofen plus okadaic acid 1 pg/mouse) (Fig. 3A). Cantharidin behaved in a similar way, although it was necessary to use higher doses than for okadaic acid. Cantharidin (0.01 and 1 ng/mouse, i.c.v.) displaced Fig. 3. Effects of the i.c.v. administration of okadaic acid and cantharidin, or their vehicle, on the antinociception induced by baclofen (s.c.) in a tail flick test in mice. (A) Effects of baclofen plus: vehicle (C), okadaic acid 0.01 pg/mouse (6) or okadaic acid 1 pg/mouse (:). (B) Effects of baclofen plus: vehicle (C), cantharidin 0.01 ng/mouse (^) or cantharidin 1 ng/mouse (%). The percentage of antinociception was calculated from the area under the curve of antinociception (see Section 2). Each point represents the meangsem (nr8). Statistically significant differences in comparison to baclofen plus vehicle: *P!0.05; **P!0.01 (two-way ANOVA followed by Newman Keuls test). the dose-response curve of baclofen to the right (Fig. 3B), and increased its ED 50 from 4.74G0.77 (baclofen plus vehicle) to 8.19G0.50 mg/kg (baclofen plus cantharidin 0.01 ng/mouse) and 10.15G0.47 mg/kg (baclofen plus cantharidin 1 ng/mouse). When we compared the effects of okadaic acid (1 fg/mouse to 1 pg/mouse, i.c.v.) and cantharidin (1 pg/mouse to 1 ng/mouse, i.c.v.) on the antinociception induced by baclofen (8 mg/kg, s.c.), both inhibitors of PPs dose-dependently reversed the antinociceptive effect of

216 A. Moncada et al. / Pain 114 (2005) 212 220 Fig. 4B shows that both okadaic acid (1 pg/mouse, i.c.v.) and cantharidin (1 ng/mouse, i.c.v.) antagonized the antinociceptive effect of baclofen (8 mg/kg, s.c.) from 30 to 120 min after administration. However, L-norokadaone (1 ng/mouse, i.c.v.) did not modify the time course of the antinociceptive effect of baclofen (Fig. 4B). 3.3. Effects of inhibitors of PPs on U50,488H-induced antinociception The agonist of k-opioid receptors U50,488H (2 16 mg/kg, s.c.) induced dose-dependent antinociception (Fig. 5A). The i.c.v. administration of okadaic acid (0.01 and 1 pg/mouse) did not shift the dose-response curve of U50,488H (Fig. 5A), nor did it significantly modify the ED 50 of this drug, which reached 4.33G0.32 mg/kg in the control group (U50,488HCvehicle) and similar values (4.32G0.40 mg/kg and 4.30G0.01 mg/kg) in the groups treated with U50,488 HCokadaic acid at 0.01 and 1 pg/mouse. When a wide range of doses of okadaic acid (0.001 1 pg/mouse, i.c.v.), cantharidin (0.001 1 ng/mouse, i.c.v.) and L-norokadaone (0.001 pg/mouse to 1 ng/mouse, i.c.v.) were associated with U50,588H (8 mg/kg, s.c.), none of them significantly modified the antinociception induced by the k-opioid receptor agonist (Fig. 5B). Moreover, okadaic acid (1 pg/mouse, i.c.v.), cantharidin (1 ng/mouse, i.c.v.) and L-norokadaone (1 ng/mouse, i.c.v.) did not modify the time course of the antinociceptive effect of U50,488H (8 mg/kg, s.c.) (data not shown). 4. Discussion Fig. 4. (A) Effects of i.c.v. administration of okadaic acid (1 fg/mouse to 1 pg/mouse) (:), cantharidin (1 pg/mouse to 1 ng/mouse, i.c.v.) (%) orlnorokadaone (1 fg/mouse to 1 ng/mouse, i.c.v.) (,) on the antinociception induced by baclofen (8 mg/kg, s.c.). The shaded area represents the effect of baclofen (8 mg/kg, s.c.)cvehicle. The percentage of antinociception was calculated from the area under the curve of antinociception (see Section 2). (B) Time course of the antinociceptive effects obtained with the tail flick test in mice treated with baclofen 8 mg/kg, s.c. plus: vehicle i.c.v. (C), okadaic acid 1 pg/mouse, i.c.v. (:), cantharidin 1 ng/mouse, i.c.v. (%), or L-norokadaone 1 ng/mouse i.c.v. (,). The data shown represent tail flick latency at each time point. Results are represented as the meangsem (nr8). Statistically significant differences in comparison to baclofen plus vehicle: *P!0.05; **P!0.01 (one-way ANOVA [A] and two-way ANOVA [B] followed by Newman Keuls test). baclofen (Fig. 4A). The dose-response lines for the two drugs were parallel, cantharidin being approximately 1000 times less potent than okadaic acid in antagonizing the effect of baclofen (Fig. 4A). In contrast, L-norokadaone (1 fg/mouse to 1 ng/mouse, i.c.v.) did not modify baclofen-induced antinociception (Fig. 4A). The present research is unique in that we tested the differential modulation of the antinociceptive effects of different agonists of GPCRs by inhibitors of serine/threonine PPs. We show that the PP inhibitors okadaic acid and cantharidin dose-dependently decreased the antinociceptive effects of the a 2 adrenoceptor agonist clonidine and GABA B receptor agonist baclofen, but did not modify the antinociception of the k-opioid receptor agonist U50,488H. Okadaic acid and cantharidin are cell-permeable agents that bind the catalytic subunit of serine/threonine PPs and inhibit them with the same efficacy but different potency (okadaic acid [ cantharidin) (Honkanen and Golden, 2002; Li and Casida, 1992). In our experiments they also antagonized clonidine- and baclofen-induced antinociception with similar efficacy, although okadaic acid was a thousand times more potent than cantharidin. The PP inhibitors decreased not only the potency of clonidine and baclofen, but also their maximum antinociceptive effect. This probably reflects that the antagonism of the interaction is noncompetitive, as both types of drug exert their action at different sites. These findings are probably also related with the long-lasting effects of the inhibitors of PPs tested here. Interestingly, the okadaic acid analogue L-norokadaone,

A. Moncada et al. / Pain 114 (2005) 212 220 217 Fig. 5. (A) Effects of the s.c. administration of different doses of U50,488H associated with: vehicle i.c.v. (C), okadaic acid 10 fg/mouse, i.c.v. (6), or okadaic acid 1 pg/mouse, i.c.v. (:). (B) Effects of the i.c.v. administration of different doses of okadaic acid (0.001 1 pg/mouse) (hatched columns), cantharidin (1 pg/mouse to 1 ng/mouse) (cross-hatched columns), or L- norokadaone (1 fg/mouse to 1 ng/mouse) (white columns) or their vehicle (black column) on the antinociception induced by U50,488H (8 mg/kg, s.c.). The percentage of antinociception was calculated from the area under the curve of antinociception (see Section 2). Each point and each column represents the meangsem (nr8). Nonstatistically significant differences in comparison to U50,488H plus vehicle were found (two-way ANOVA [A] and one-way ANOVA [B]). which does not inhibit PPs (Honkanen et al., 1994; Nishiwaki et al., 1990), did not antagonize the antinociceptive effect of clonidine or baclofen. These results suggest that inhibition of PP activity may underlie the ability of okadaic acid and cantharidin to antagonize the antinociceptive effects of the a 2 -adrenoceptor and the GABA B receptor agonists we tested. Recently, we demonstrated that okadaic acid and cantharidin, but not L-norokadaone, also decreased morphine-induced antinociception in the tail flick test in mice (Moncada et al., 2003), which is in agreement with the present results and suggests that inhibitors of PPs can modulate the antinociception induced by agonists of different G protein-coupled receptors. On the other hand, okadaic acid and cantharidin had no effect on tail flick latency in control animals. This agrees with previous results showing that both PP inhibitors were unable to modify tail flick latency in control animals (Bersntein and Welch, 1998; Moncada et al., 2003). It has also been reported that PP inhibitors (okadaic acid and fostriecin administered i.t.) did not modify the response to thermal or mechanical nociceptive stimuli in the normal rat paw, although they prolonged and enhanced the intensity of mechanical and thermal hyperalgesia and mechanical allodynia in the capsaicin-treated rat paw (Zhang et al., 2003). Therefore, PPs seem not to play an important role in the perception of acute pain in a control situation, although they modulated the antinociceptive effect induced by several drugs (our study and Moncada et al., 2003) and were found to be involved in the mechanisms that maintain neuronal sensitisation to nociceptive stimuli (Zhang et al., 2003). Regarding the subtype of PP involved in the antagonism by PP inhibitors of the antinociceptive effects of clonidine and baclofen, it should be considered that okadaic acid and cantharidin block several types of PPs with different potency. PP2A and PP4 are the most sensitive, but at around 10- to 100-fold higher concentrations they also block PP1 and PP5 (Bialojan and Takai, 1988; Chen et al., 1994; Hastie and Cohen, 1998; Honkanen, 1993; Li et al., 1993). Other PPs such as PP2B, PP2C and PP7 are much less sensitive or insensitive (Bialojan and Takai, 1988; Honkanen, 1993; Huang and Honkanen, 1998; Li et al., 1993). We used very low doses of okadaic acid and cantharidin. If it is assumed that the volume of cerebrospinal in mice is about 100 ml (Bernstein and Welch, 1998; Moncada et al., 2003), the final concentration of inhibitors of PPs in the cerebrospinal fluid after injecting the doses that produce around 50% inhibition of clonidine- and baclofen-induced antinociception (0.1 1 pg/mouse of okadaic acid and 0.1 1 ng/mouse of cantharidin, see Figs. 2A and 4A) would be 0.012 0.124 nm for okadaic acid and 5 51 nm for cantharidin. These concentrations are in consonance with the IC 50 values for the inhibition of PP2A and PP4 by okadaic acid (0.1 0.3 nm) and cantharidin (40 194 nm) (Cohen et al., 1989, 1990; Girault, 1994; Hastie and Cohen, 1998; Honkanen, 1993; Honkanen and Golden, 2002; Li et al., 1993). Thus the inhibitory effect on PP2A or PP4 may play a relevant role in the antagonism by okadaic acid and cantharidin of clonidineand baclofen-induced antinociception. However, new studies with other inhibitors of PPs such as tautomycetin, a preferential blocker of PP1 over PP2A (Mitsuhashi et al., 2001), or fostriecin, a relatively selective blocker of PP2 and PP4 over PP1 (Hastie and Cohen, 1998; Walsh et al., 1997), may provide a deeper understanding of the PPs involved in the antinociceptive effects of morphine, clonidine and baclofen.

218 A. Moncada et al. / Pain 114 (2005) 212 220 Obviously, our experimental approach does not allow us to identify the exact mechanism of the interaction between inhibitors of PPs and the GPCR agonists clonidine and baclofen, nor was this the goal of our study. However, some possible explanations for the effects we observed can be offered. (1) Enhanced phosphorylation of the receptor induced by inhibitors of PPs altered the receptor-g protein interaction, uncoupling both proteins and thus reducing the receptor-mediated effects (Alcántara-Hernández et al., 2000; Ansonoff and Etgen, 2001; Vázquez-Prado et al., 2003). (2) Nucleotide-binding protein Ga i2 undergoes dynamic phosphorylation reactions catalysed by PKC and phosphatases 1 and 2A (Bushfield et al., 1991; Strassheim and Malbon, 1994), and phosphorylated Ga i2 is less able to regulate downstream cellular effectors such as adenylyl cyclase (Strassheim and Malbon, 1994). Because protein Ga i2 is a main transducer of a 2 adrenoceptors and GABA B receptor-mediated signalling in neurons, and plays a prominent role in their antinociceptive effects (Garzón et al., 1999; Odagaki and Koyama, 2001), the greater phosphorylation of this protein induced by PP inhibitors may reduce the antinociception of agonists of these receptors. (3) The inhibitors of protein phosphatases modulate the formation and activity of second messengers such as camp (Strassheim and Malbon, 1994; Wang et al., 1996) and intracellular inositol-1,4,5-triphosphate (Assari et al., 2003; Strassheim et al., 1998), which mediate the effects of the agonists of a 2 adrenoceptors and GABA B receptors, including their antinociceptive effects (Bowery, 1993; Holliday et al., 1997; Parsley et al., 1999; Wei and Roerig, 1998). (4) Finally, the inhibitors of PPs modulate the activity of potassium and calcium channels (Davare et al., 2000; Herzig and Neumann, 2000; Mullner et al., 2003). It has been reported that these channels underlie the actions of clonidine (Li and Bayliss, 1998; Ocaña and Baeyens, 1993; Ocaña et al., 1996) and baclofen (Lambert and Wilson, 1996; Ocaña and Baeyens, 1993; Ocaña et al., 1996); therefore, a change in the phosphorylation level of ion channels may alter the effects of these GPCR agonists. Neither okadaic acid or cantharidin, at any dose tested, significantly modified the antinociceptive effects of the k-opioid receptor agonist U50,488-H. These results were not due to any methodological pitfall because the experimental procedure was the same as that used with clonidine and baclofen. Our finding of a lack of activity of PP inhibitors on U50,488-H antinociception rules out the hypothesis that an unspecific effect of okadaic acid or cantharidin on pain transmission underlies the interaction between these inhibitors of PPs and agonists of GPCRs. The antagonism by PP inhibitors of the antinociception induced by agonists of m-opioid receptor, a 2 -adrenoceptors and GABA B receptors appears to be specific and not a consequence of the indiscriminate antagonism of any antinociceptive drug. One possible hypothesis to explain how okadaic acid and cantharidin modified the antinociception induced by several GPCR agonists (morphine, clonidine and baclofen) but not the antinociceptive effect of U50,488-H is that GPCR agonists exerted their effects through different effector systems that can be modulated in different way by PPs. This is especially interesting regarding K C channels, since it has been well established that the opening of some K C channels is involved in the antinociception induced by morphine, clonidine (K ATP channels) and baclofen (calcium- and voltage-dependent potassium channels) (see Ocaña et al., 2004 for a review), but not in that induced by U50,448H (Ocaña and Baeyens, 1993; Ocaña et al., 1993, 1996; Picolo et al., 2003; Welch and Dunlow, 1993). Because PP1 and PP2A control the activity of these types of K C channel (Firth et al., 2000; Herzing and Neumann, 2000; Light et al., 1995), an effect of PP inhibitors on K C channel activity may modulate the effects of GPCR agonists in different ways depending on their coupling to potassium channels. Obviously, this tempting hypothesis needs further testing before it can be accepted or refuted. In conclusion, the present data demonstrate that okadaic acid and cantharidin, at doses that probably inhibit PP2A and PP4 predominantly, antagonized in a dose-dependent and specific way the antinociception induced by the a 2 adrenoceptor agonist clonidine and the GABA B receptor agonist baclofen, but did not affect the antinociception induced by the k-opioid receptor agonist U50,488-H. These results suggest that serine/threonine protein phosphatases play a role in the antinociceptive effects of agonists of some, but not all, G protein-coupled receptors. Acknowledgements This study was supported in part by grants from the Comisión Interministerial de Ciencia y Tecnología (SAF 97/0173) and the Junta de Andalucía (CTS 109). C.M. Cendán is the recipient of a grant from the University of Granada. We thank K. Shashok for revising the English style of the manuscript. References Alcántara-Hernández R, Vázquez-Prado J, García-Sáinz JA. Protein phosphatase protein kinase interplay modulates a 1b -adrenoceptor phosphorylation: effects of okadaic acid. Br J Pharmacol 2000;129: 727 30. Ansonoff MA, Etgen AM. Receptor phosphorylation mediates estradiol reduction of alpha2-adrenoceptor coupling to G protein in the hypothalamus of female rats. Endocrine 2001;14:165 74. Assari T, Cox S, Munday MR, Pearce B. Regulation of a 1 -adrenoceptorlinked phosphoinositide metabolism in cultured glia: involvement of protein phosphatases and kinases. Cell Signal 2003;15:403 12. Bernstein MA, Welch SP. Inhibition of protein phosphatases alters the expression of morphine tolerance in mice. Eur J Pharmacol 1998;341: 173 7.

A. Moncada et al. / Pain 114 (2005) 212 220 219 Bialojan C, Takai A. Inhibitory effect of a marine-sponge toxin, okadaic acid, on protein phosphatases. Specificity and kinetics. Biochem J 1988; 256:283 90. Bowery NG. GABA B receptor pharmacology. Annu Rev Pharmacol Toxicol 1993;33:109 47. Bushfield M, Lavan BE, Houslay M. Okadaic acid identifies a phosphorylation/dephosphorylation cycle controlling the inhibitory guanine-nucleotide-binding regulatory protein G i2. Biochem J 1991; 274:317 21. Chan SF, Sucher NJ. An NMDA receptor signalling complex with protein phosphatase 2A. J Neurosci 2001;21:7985 92. Chen MX, McPartlin AE, Brown L, Chen YH, Barker HM, Cohen PT. A novel human protein serine/threonine phosphatase, which possesses four tetratricopeptide repeat motifs and localizes to the nucleus. Eur Mol Biol Org J 1994;13:4278 80. Chik CL, Li B, Karpinski E, Ho AK. Regulation of the L-type Ca 2C channel current in rat pinealocytes: role of basal phosphorylation. J Neurochem 1999;72:73 80. Cohen PTW, Klumpp S, Schelling DL. An improved procedure for identifying and quantitating protein phosphatases in mammalian tissues. Fed Eur Biochem Soc Lett 1989;250:596 600. Cohen P, Holmes CF, Tsukitani Y. Okadaic acid: a new probe for the study of cellular regulation. Trends Biochem Sci 1990;15:98 102. Davare MA, Horne MC, Hell JW. Protein phosphatase 2A is associated with class C L-type calcium channels (Ca v 1.2) and antagonizes channel phosphorylation by camp-dependent protein kinase. J Biol Chem 2000;275:39710 7. Del Pozo E, Ruiz-Garcia C, Baeyens JM. Analgesic effects of diltiazem and verapamil after central and peripheral administration in the hot-plate test. Gen Pharmacol 1990;21:681 5. Dierssen M, Florez J, Hurle MA. Calcium channel modulation by dihydropyridines modifies sufentanil-induced antinociception in acute and tolerant conditions. Naunyn Schmiedeberg s Arch Pharmacol 1990; 342:559 65. Dogrul A, Yesilyurt O, Isimer A, Guzeldemir ME. L-type and T-type calcium channel blockade potentiate the analgesic effects of morphine and selective m opioid agonist, but not to selective d and k agonist at the level of the spinal cord in mice. Pain 2001;93:61 8. Firth TA, Mawe GM, Nelson MT. Pharmacology and modulation of K(ATP) channels by protein kinase C and phosphatases in gallbladder smooth muscle. Am J Physiol Cell Physiol 2000;278:C1031 C7. Foster JD, Pananusorn B, Cervinski MA, Holden HE, Vaughan RA. Dopamine transporters are dephosphorylated in striatal homogenates and in vitro by protein phosphatase 1. Brain Res Mol Brain Res 2003; 110:100 8. Garzón J, Rodríguez-Díaz M, DeAntonio I, DeFelipe J, Rodríguez J-R, Sánchez-Blázquez P. Myr C -G i2 a and G o a subunits restore the efficacy of opioids, clonidine and neurotensin giving rise to antinociception in G-protein knock-down mice. Neuropharmacology 1999;38:1861 73. Girault JA. Protein kinases and phosphatases. Neurotransmissions 1994;10: 1 6. Haley TJ, Mccormick WG. Pharmacological effects produced by intracerebral injection of drugs in the conscious mouse. Br J Pharmacol 1957;12:12 15. Hastie CJ, Cohen PT. Purification of protein phosphatase 4 catalytic subunit: inhibition by the antitumour drug fostriecin and other tumour suppressors and promoters. Fed Eur Biochem Soc Lett 1998;431: 357 61. Hemmings HC, Nairn AC, McGuinness TL, Huganir RL, Greengard P. Role of protein phosphorylation in neuronal signal transduction. Fed Am Soc Exp Biol J 1989;3:1583 92. Herzig S, Neumann J. Effects of serine/threonine protein phosphatases on ion channels in excitable membranes. Pharmacol Rev 2000;80:173 96. Holliday ND, Tough IR, Cox HM. Inhibition of cyclic AMP-dependent chloride secretion by PP receptors and alpha 2-adrenoceptors in a human colonic epithelial cell line. Naunyn-Schmiedeberg s Arch Pharmacol 1997;355:183 9. Honkanen RE. Cantharidin, another natural toxin that inhibits the activity of serine/threonine protein phosphatases types 1 and 2A. Fed Eur Biochem Soc Lett 1993;330:283 6. Honkanen RE, Golden T. Regulators of serine/threonine protein phosphatases at the dawn of a clinical era? Curr Med Chem 2002; 9:2055 75. Honkanen RE, Codispoti BA, Tse K, Boynton AL. Characterization of natural toxins with inhibitory activity against serine/threonine protein phosphatases. Toxicon 1994;32:339 50. Huang X, Honkanen RE. Molecular cloning, expression, and characterization of a novel human serine/threonine protein phosphatase, PP7, that is homologous to Drosophila retinal degeneration C gene product (rdgc). J Biol Chem 1998;273:1462 8. Lambert NA, Wilson WA. High-threshold Ca 2C currents in rat hippocampal interneurones and their selective inhibition by activation of GABA B receptors. J Physiol 1996;492:115 27. Li YW, Bayliss DA. Activation of a 2 -adrenoceptors causes inhibition of calcium channels but does not modulate inwardly-rectifying K C channels in caudal raphe neurons. Neuroscience 1998;82:753 65. Li YM, Casida JE. Cantharidin-binding protein: identification as protein phosphatase 2A. Proc Natl Acad Sci 1992;89:11867 70. Li YM, MacKintosh C, Casida JE. Protein phosphatase 2A and its [ 3 H]cantharidin/[ 3 H]endothall thioanhydride binding site. Inhibitor specificity of cantharidin and ATP analogues. Biochem Pharmacol 1993;46:1435 43. Light PE, Allen BG, Walsh MP, French RJ. Regulation of adenosine triphosphate-sensitive potassium channels from rabbit ventricular myocytes by protein kinase C and type 2A protein phosphatase. Biochemistry 1995;34:7252 7. Mitsuhashi S, Matsuura N, Ubukata M, Oikawa H, Shima H, Kikuchi K. Tautomycetin is a novel and specific inhibitor of serine/threonine protein phosphatase type 1, PP1. Biochem Biophys Res Commun 2001; 287:328 31. Moncada A, Cendan CM, Baeyens JM, Del Pozo E. Effects of serine/threonine protein phosphatase inhibitors on morphine-induced antinociception in the tail flick test in mice. Eur J Pharmacol 2003;465: 53 60. Mullner C, Yakubovich D, Dessauer CW, Platzer D, Schreibmayer W. Single channel analysis of the regulation of GIRK1/GIRK4 channels by protein phosphorylation. Biophys J 2003;84:1399 409. Nishiwaki S, Fujiti H, Suganuma M, Furuya-Suguri H, Matsushima R, Iida Y, Ojika M, Yamada K, Uemura D, Yasumoto T. Structure activity relationship within a series of okadaic acid derivatives. Carcinogenesis 1990;11:1837 41. Ocaña M, Baeyens JM. Differential effects of KC channel blockers on antinociception induced by a2-adrenoceptor, GABA B and k-opioid receptor agonists. Br J Pharmacol 1993;110:1049 54. Ocaña M, Del Pozo E, Baeyens JM. ATP-dependent K C channel blockers antagonize morphine- but not U-504,88H-induced antinociception. Eur J Pharmacol 1993;230:203 7. Ocaña M, Del Pozo E, Barrios M, Baeyens JM. Subgroups among m-opioid receptor agonists distinguished by ATP-sensitive K C channel-acting drugs. Br J Pharmacol 1995;114:1296 302. Ocaña M, Barrios M, Baeyens JM. Cromakalim differentially enhances antinociception induced by agonists of Alpha 2 Adrenoceptors, g- Aminobutyric Acid B, Mu and Kappa opioid receptors. J Pharmacol Exp Ther 1996;276:1136 42. Ocaña M, Cendán CM, Cobos EJ, Entrena JM, Baeyens JM. Potassium channels and pain: present realities and future opportunities. Eur J Pharmacol 2004;500:203 19. Odagaki Y, Koyama T. Identification of G a subtype(s) involved in g- aminobutyric acid B receptor-mediated high-affinity guanosine triphosphatase activity in rat cerebral cortical membranes. Neurosci Lett 2001;297:137 41.

220 A. Moncada et al. / Pain 114 (2005) 212 220 Parsley S, Gazi L, Bobirnac I, Loetscher E, Schoeffter P. Functional a 2C - adrenoceptors in human neuroblastoma SH-SY5Y cells. Eur J Pharmacol 1999;372:109 11. Picolo G, Cassola AC, Cury Y. Activation of peripheral ATP-sensitive K C channels mediates the antinociceptive effect of Crotalus durissus terrificus snake venom. Eur J Pharmacol 2003;469:57 64. Price NE, Mumby MC. Brain protein serine/threonine phosphatases. Curr Opin Neurobiol 1999;9:336 42. Robles LI, Barrios M, Del Pozo E, Dordal A, Baeyens JM. Effects of K C channel blockers and openers on antinociception induced by agonists of 5-HT 1A receptors. Eur J Pharmacol 1996;295:181 8. Strassheim D, Malbon CC. Phosphorylation of G ia2 attenuates inhibitory adenylyl cyclase in neuroblastoma/glioma hybrid (NG-10815) cells. J Biol Chem 1994;269:14307 13. Strassheim D, Law PY, Loh HH. Contribution of phospholipase C-beta 3 phosphorylation to the rapid attenuation opioid-activated phosphoinositide response. Mol Pharmacol 1998;53:1047 53. Surmeier DJ, Bargas J, Hemmings Jr HC, Nairn AC, Greengard P. Modulation of calcium currents by a D 1 dopaminergic protein kinase/phosphatase cascade in rat neostriatal neurons. Neuron 1995; 14:385 97. Vázquez-Prado J, Casas-González P, García-Sáinz A. G protein-coupled receptor cross-talk: pivotal roles of protein phosphorylation and protein protein interactions. Cell Signal 2003;15:549 57. Walsh AH, Cheng A, Honkanen RE. Fostriecin, an antitumor antibiotic with inhibitory activity against serine/threonine protein phosphatases types 1 (PP1) and 2A (PP2A), is highly selective for PP2A. Fed Eur Biochem Soc Lett 1997;416:230 4. Wang L, Medina VM, Rivera M, Gintzler AR. Relevance of phosphorylation state to opioid responsiveness in opiate naive and tolerant/dependent tissue. Brain Res 1996;723:61 9. Wei ZY, Roerig SC. Spinal morphine/clonidine antinociceptive synergism is regulated by protein kinase C, but not protein kinase A activity. J Pharmacol Exp Ther 1998;287:937 43. Welch SP, Dunlow LD. Antinociceptive activity of intrathecally administered potassium channel openers and opioid agonists: a common mechanism of action? J Pharmacol Exp Ther 1993;267:399 9. Zhang X, Wu J, Fang L, Willis WD. The effects of protein phosphatase inhibitors on nociceptive behavioural responses of rats following intradermal injection of capsaicin. Pain 2003;106:443 51. Zimmermann M. Ethical guidelines for investigations of experimental pain in conscious animals. Pain 1983;16:109 10.