1521-0103/351/1/134–145$25.00 http://dx.doi.org/10.1124/jpet.114.217372THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS J Pharmacol Exp Ther 351:134–145, October 2014Copyright ª by The American Society for Pharmacology and Experimental Therapeutics

Cyclooxygenase 2 Inhibitor Celecoxib Inhibits GlutamateRelease by Attenuating the PGE2/EP2 Pathway in RatCerebral Cortex Endings

Tzu-Yu Lin, Cheng-Wei Lu, Chia-Chuan Wang, Shu Kuei Huang, and Su-Jane WangDepartment of Anesthesiology, Far-Eastern Memorial Hospital, New Taipei, Taiwan ( T.-Y.L., C.-W.L., S.K.H.); Department ofMechanical Engineering, Yuan Ze University, Taoyuan, Taiwan ( T.-Y.L., C.-W.L.); and Graduate Institute of Basic Medicine(S.-J.W.) and School of Medicine (C.-C.W., S.-J.W.), Fu Jen Catholic University, New Taipei, Taiwan

Received June 11, 2014; accepted July 18, 2014

ABSTRACTThe excitotoxicity caused by excessive glutamate is a criticalelement in the neuropathology of acute and chronic braindisorders. Therefore, inhibition of glutamate release is a poten-tially valuable therapeutic strategy for treating these diseases. Inthis study, we investigated the effect of celecoxib, a selectivecyclooxygenase-2 (COX-2) inhibitor that reduces the level ofprostaglandin E2 (PGE2), on endogenous glutamate release inrat cerebral cortex nerve terminals (synaptosomes). Celecoxibsubstantially inhibited the release of glutamate induced by theK1 channel blocker 4-aminopyridine (4-AP), and this phenom-enon was prevented by chelating the extracellular Ca21 ions andby the vesicular transporter inhibitor bafilomycin A1. Celecoxibinhibited a 4-AP–induced increase in cytosolic-free Ca21 con-centration, and the celecoxib-mediated inhibition of glutamaterelease was prevented by the Cav2.2 (N-type) and Cav2.1 (P/Q-type) channel blocker v-conotoxin MVIIC. However, celecoxib

did not alter 4-AP–mediated depolarization and Na1 influx. Inaddition, this glutamate release–inhibiting effect of celecoxibwas mediated through the PGE2 subtype 2 receptor (EP2) be-cause it was not observed in the presence of butaprost (an EP2agonist) or PF04418948 [1-(4-fluorobenzoyl)-3-[[6-methoxy-2-naphthalenyl)methyl]-3-azetidinecarboxylic acid; an EP2 an-tagonist]. The celecoxib effect on 4-AP–induced glutamaterelease was prevented by the inhibition or activation of proteinkinase A (PKA), and celecoxib decreased the 4-AP–inducedphosphorylation of PKA. We also determined that COX-2 andthe EP2 receptor are present in presynaptic terminals becausethey are colocalized with synaptophysin, a presynaptic marker.These results collectively indicate that celecoxib inhibits glu-tamate release from nerve terminals by reducing voltage-dependent Ca21 entry through a signaling cascade involvingEP2 and PKA.

IntroductionCyclooxygenase-2 (COX-2) is an essential enzyme that

converts arachidonic acid into prostaglandin. In the centralnervous system (CNS), COX-2 is constitutively expressed,mainly in hippocampal and cortical pyramidal neurons whereit plays a role in synaptic transmission and plasticity (Murrayand O’Connor, 2003; Cowley et al., 2008). In addition, COX-2is a crucial mediator of neuroinflammation, and its overex-pression in neurons is implicated in numerous brain injuries(Nogawa et al., 1997; Minghetti, 2004). COX-2 inhibition hasalso been reported to confer neuroprotective effects in various

animal models with neurologic disorders, including cerebralischemia, epilepsy, Alzheimer’s disease, and Parkinson’sdisease (Nakayama et al., 1998; Hunter et al., 2007; Trepanierand Milgram, 2010; Serrano et al., 2011; Akram et al., 2013).Consequently, COX-2 inhibitors have been suggested to bea potential therapeutic application for treating these CNSdisorders. Celecoxib is a selective COX-2 inhibitor, and itsneuroprotective action has been observed in several in vitroand in vivo experimental studies. For example, celecoxib atten-uates oxygen and glucose deprivation–induced neuronal celldeath (López-Villodres et al., 2012), protects against ischemia-or lipopolysaccharide-induced brain damage (Abd El-Aalet al., 2013; Fan et al., 2013; Kaizaki et al., 2013), improveskainic acid–induced cognitive impairment (Gobbo and O’Mara,2004), and delays the progress of brain degeneration (Drachmanet al., 2002; Small et al., 2008). Celecoxib is known to possess

This work was supported by the National Science Council of Taiwan (NSC101-2314-B-418-001).

dx.doi.org/10.1124/jpet.114.217372.

ABBREVIATIONS: 4-AP, 4-aminopyridine; BAPTA, 1,2-bis(2-aminophenoxy) ethane-N,N,N9,N9-tetraacetic acid-acetoxymethyl ester; [Ca21]C,cytosolic free Ca21 concentration; CGP37157, 7-chloro-5-(2-chloropheny)-1,5-dihydro-4,1-benzothiazepin-2(3H)-one; v-CgTX MVIIC, v-conotoxinMVIIC; CNS, central nervous system; COX-2, cyclooxygenase-2; DiSC3(5), 39,39,39-dipropylthiadicarbocyanine iodide; DL-TBOA, DL-threo-b-benzyl-oxyaspartate; DON, 6-diazo-5-oxo-L-norleucine; EP2, PGE2 subtype 2 receptor; ER, endoplasmic reticulum; GF109203X, bisindolylmaleimide I; H89, N-[2-(p-bromocinnamylamino)-ethyl]-5-isoquinolinesulfonamide dihydrochloride; HBM, HEPES buffer medium; KT5720, hexahydro-10-hydroxy-9-methyl-1-oxo-9,12-epoxy-1H-diindolo-benzodiazocine-10-carboxylic acid; PB, phosphate buffer; PD98059, 2-(2-amino-3-methoxyphenyl)-4H-1-benzopyran-4-one;PF04418948, 1-(4-fluorobenzoyl)-3-[[6-methoxy-2-naphthalenyl)methyl]-3-azetidinecarboxylic acid; PGE2, prostaglandin E2; PKA, protein kinase A; SBFI-AM, sodium-binding benzofuran isophthalate-AM; TCS2510, (5R)-5[(3S)-3-hydroxy-4-phenyl-1-buten-1-yl]-1-[6-(2H-tetrazol-5-yl)hexyl]-2-pyrrolidinone;TTX, tetrodotoxin; VDCC, voltage-dependent Ca21 channel.

134

at ASPE

T Journals on A

ugust 19, 2020jpet.aspetjournals.org

Dow

nloaded from

anti-inflammatory activity, and its neuroprotective action isgenerally assumed to be associated with this property (Hunteret al., 2007; López-Villodres et al., 2012; Kaizaki et al., 2013).However, the pharmacological mechanisms underlying theneuroprotective effects of celecoxib have not been fully clarified.In the brain, glutamate is a major excitatory neurotransmit-

ter that plays a vital role in numerous brain functions, such assynaptic plasticity, learning, and memory (Fonnum, 1984;Greenamyre and Porter, 1994). However, in addition to thephysiologic role of glutamate, excessive glutamate release andactivation of the glutamate receptors induces an increase inintracellular Ca21 levels, subsequently triggering a cascade ofcellular responses, including enhanced oxygen free-radicalproduction, disturbed mitochondrial function, and proteaseactivation, which ultimately kills the neurons (Coyle andPuttfarcken, 1993; Schinder et al., 1996; Greenwood andConnolly, 2007). This process has been implicated as a patho-physiological factor inmultiple neurologic disorders, both acute(such as stroke and head trauma) and chronic (such as neu-rodegenerative disorders) (Meldrum and Garthwaite, 1990;Meldrum, 2000). Thus, inhibiting glutamate release mightprovide a potential target for neuroprotective action. Consistentwith this hypothesis, several clinical neuroprotective drugs (e.g.,riluzole, memantine, and minocycline) have been revealed todecrease glutamate release in rat brain tissues (Wang et al.,2004; Gonzalez et al., 2007; Lu et al., 2010).Because celecoxib has a neuroprotective profile, and the

excessive release of glutamate is a critical element in theneuropathology of acute and chronic brain disorders, assessingthe effects of celecoxib on glutamate release is warranted.Based on a review of the literature, no study has thus faraddressed whether celecoxib directly affects glutamate releaseat the presynaptic level. Hence, we used isolated nerveterminals (synaptosomes) purified from rat cerebral cortex toinvestigate the effect of celecoxib on glutamate release and tocharacterize its underlying molecular mechanisms. The syn-aptosome was adopted because it can accumulate, store, andrelease neurotransmitters, and is devoid of functional glial- andnerve-cell body elements that might obscure interpretationbecause of modulatory loci at non-neuronal, postsynaptic, ornetwork levels (Nicholls and Sihra, 1986). Therefore, thesynaptosome provides a useful model for directly studyingthe specific presynaptic regulation of neurotransmitter release.Using this model, we investigated the effects of celecoxib on thelevels of released glutamate, the synaptosomal plasma mem-brane potential, and the activation of the voltage-dependentCa21 channels (VDCCs). Because COX-2 inhibition suppressesglutamatergic neurotransmission and long-term potentiationin the hippocampus through prostaglandin E2 (PGE2), thepredominant reaction product of COX-2 (Chen et al., 2002;Murray and O’Connor, 2003), we also determined whethera relationship exists between the regulation of glutamaterelease by celecoxib and the PGE2 signaling cascade.

Materials and MethodsChemicals. v-Conotoxin MVIIC (v-CgTX MVIIC), dantrolene,

CGP37157 [7-chloro-5-(2-chlorophenyl)-1,5-dihydro-4,1-benzothiazepin-2(3H)-one], PF04418948 [1-(4-fluorobenzoyl)-3-[[6-methoxy-2-naphthalenyl)methyl]-3-azetidinecarboxylic acid], TCS2510 [(5R)-5[(3S)-3-hydroxy-4-phenyl-1-buten-1-yl]-1-[6-(2H-tetrazol-5-yl)hexyl]-2-pyrrolidinone],H89 (N-[2-(p-bromocinnamylamino)-ethyl]-5-isoquinolinesulfonamide

dihydrochloride), PD98059 [2-(2-amino-3-methoxyphenyl)-4H-1-benzopyran-4-one], KT5720 (hexahydro-10-hydroxy-9-methyl-1-oxo-9,12-epoxy-1H-diindolo-benzodiazocine-10-carboxylic acid), GF109203X(bisindolylmaleimide I), and DON (6-diazo-5-oxo-L-norleucine) wereobtained from Tocris Cookson (Bristol, UK). Fura-2-AM, sodium-binding benzofuran isophthalate-AM (SBFI-AM), and DiSC3(5) (39,39,39-dipropylthiadicarbocyanine iodide) were purchased from Invitrogen(Carlsbad, CA). PGE2 and sulprostone were purchased from CaymanChemical (Ann Arbor, MI). Celecoxib, butaprost, EGTA, tetrodotoxin(TTX), and all other reagents were purchased from Sigma-Aldrich(St. Louis, MO).

Animals. Adult male Sprague-Dawley rats (1502200 g) (BioLASCOTaiwan Co., Ltd, Taipei, Taiwan) were used in this study. All animalexperiments were conducted in accordance with theNational Institutesof Health Guide for the Care and Use of Laboratory Animals, and wereapproved by the Institutional Animal Care and Use Committee at theFar-Eastern Memorial Hospital (100-1-19-A). All efforts were made tominimize the number of animals tested and their suffering.

Synaptosomal Preparation. Percoll-purified synaptosomeswere prepared using the cerebral cortex of the rats, as describedpreviously (Nicholls and Sihra, 1986). The adult rats were sacrificedby decapitation, and the brains were removed at 4°C. The cerebralcortex was rapidly dissected and homogenized in a medium containing320 mM sucrose at pH 7.4. The homogenate was centrifuged at 3000g(5000 rpm in a JA 25.5 rotor; Beckman Coulter, Inc., Brea, CA) for10 minutes at 4°C, and the supernatant was centrifuged again at14,500g (11,000 rpm in a JA 25.5 rotor) for 12minutes at 4°C. The pelletwas gently resuspended in 8 ml of 320 mM sucrose at pH 7.4. Twomilliliters of this synaptosomal suspension was placed in 3ml of Percolldiscontinuous gradients containing (in mM) 320 sucrose, 1 EDTA, 0.25DL-dithiothreitol, and 3, 10, and 23% Percoll at pH 7.4. The gradientswere centrifuged at 32,500g (16,500 rpm in a JA 20.5 rotor) for7 minutes at 4°C. Synaptosomes sedimenting between the 10 and 23%Percoll bands were collected and diluted to a final volume of 30 ml ofHEPES buffer medium (HBM) consisting of (in mM) 140 NaCl, 5 KCl,5 NaHCO3, 1 MgCl2×6H2O, 1.2 Na2HPO4, 10 glucose, and 10 HEPES atpH 7.4. Protein concentration was determined using a Bradford assay.Synaptosomes were centrifuged in a final wash to obtain synaptosomalpellets containing 0.5 mg of protein. The synaptosomal pellets werestored on ice and used within 4–6 hours.

Glutamate Release Assay. The glutamate release was assayedusing on-line fluorimetry, as described previously (Nicholls and Sihra,1986; Wang and Sihra, 2004). Synaptosomal pellets (0.5 mg of protein)were resuspended in HBM containing 16 mM bovine serum albumin andincubated in a stirred and thermostated cuvette at 37°C in a PerkinElmerLS-55 spectrofluorimeter (PerkinElmer Life and Analytical Sciences,Waltham,MA). NADP1 (2mM), glutamate dehydrogenase (50U/ml), andCaCl2 (1.2 mM) were added after 3 minutes. After an additional10 minutes of incubation, 4-aminopyridine (4-AP; 1 mM) or high externalKCl (15 mM) was added to stimulate glutamate release. Glutamaterelease was monitored by measuring the increase in fluorescence (ex-citation and emission wavelengths of 340 and 460 nm, respectively)caused by NADPH being produced by the oxidative deamination ofreleased glutamate by glutamate dehydrogenase. Data were accumulatedat 2-second intervals. A standard of exogenous glutamate (5 nmol) wasadded at the end of each experiment, and the fluorescence response usedto calculate released glutamate was expressed as nanomoles of glutamateper milligram of synaptosomal protein. Values quoted in the text anddepicted in bar graphs represent the levels of glutamate cumulativelyreleased after 5 minutes of depolarization, and are expressed as nmol/mgper 5 minutes. Estimation of the IC50 was based on a one-site model

inhibition5�inhibitionMAX � ½celecoxib����IC50 1 ½celecoxib��;

and was calculated using the nonlinear curve-fitting function pro-vided in MicroCal Origin (OriginLab Corporation, Northampton,MA). Cumulative data were analyzed using Lotus 1-2-3 (IBM Cor-poration, Armonk, NY).

Inhibition of Glutamate Release by Celecoxib 135

at ASPE

T Journals on A

ugust 19, 2020jpet.aspetjournals.org

Dow

nloaded from

Synaptosomal Plasma Membrane Potential. The plasmamembrane potential was determined using a membrane potential–sensitive dye, DiSC3(5). DiSC3(5) is a positively charged carbocyaninethat accumulates in polarized synaptosomes that are negativelycharged on the inside. At high concentrations, the dye moleculesaccumulate and the fluorescence is quenched. On depolarization, thedye moves out, and hence, the fluorescence increases (Akerman et al.,1987). Synaptosomes were resuspended in HBM and incubated ina stirred and thermostated cuvette at 37°C in a PerkinElmer LS-55spectrofluorimeter. After 3 minutes of incubation, 5 mMDiSC3(5) wasadded and allowed to equilibrate before the addition of 1 mM CaCl2after 4 minutes of incubation. 4-AP was then added to depolarize thesynaptosomes for 10minutes, andDiSC3(5) fluorescencewasmonitoredat excitation and emissionwavelengths of 646 and 674 nm, respectively.Cumulative data were analyzed using Lotus 1-2-3 and expressed influorescence units.

Cytosolic Free Ca21 Concentration. Cytosolic free Ca21 con-centration ([Ca21]C) was measured using the Ca21 indicator fura-2.Synaptosomes (0.5 mg/ml) were preincubated in HBM with 16 mMbovine serum albumin in the presence of 5 mM fura-2-AM and 0.1 mMCaCl2 for 30 minutes at 37°C in a stirred test tube. After fura-2-AMloading, synaptosomes were centrifuged in a microcentrifuge for30 seconds at 3000g (5000 rpm). The synaptosomal pellets wereresuspended inHBMwith bovine serumalbumin, and the synaptosomalsuspension was stirred in a thermostated cuvette in a PerkinElmerLS-50B spectrofluorimeter. CaCl2 (1 mM) was added after 3 minutes,and further additions were made after an additional 10 minutes.Fluorescence data were accumulated at excitation wavelengths of 340and 380 nm (emission wavelength 505 nm) at 7.5-second intervals.[Ca21]C (in nanomoles) was calculated using calibration procedures(Sihra et al., 1992) and equations described previously (Grynkiewiczet al., 1985). Cumulative data were analyzed using Lotus 1-2-3.

Cytosolic Free Na1 Concentration. Na1 measurements wereperformed essentially as described for [Ca21]C determinations, exceptthat synaptosomes were preincubated with 5 mM SBFI-AM instead offura-2 (Minta and Tsien, 1989). SBFI fluorescence was monitored asin fura-2 experiments. Results are expressed as ratios of fluorescence(emission wavelength 505 nm) at excitation wavelengths of 340 and380 nm.

Immunocytochemistry. The synaptosomes were allowed toattach to coverslips (diameter 20 mm) precoated with poly-L-lysinefor 40 minutes at 4°C before being fixed with 4% paraformaldehyde in0.1 M phosphate buffer (PB; pH 7.4) for 30 minutes. After rinsing withPB three times, the synaptosomes were incubated in blocking buffercontaining 3% normal goat serum and 0.2% Triton X-100 for 60minutes. They were then incubated with a mixture of primary mousemonoclonal antibodies against synaptophysin (1:200; Abcam, Cam-bridge, UK) and rabbit monoclonal antibodies against COX-2 (1:50;Cell Signaling Technology, Beverly, MA) or PGE2 subtype 2 receptor(EP2; 1:100; Abcam) for 90 minutes at room temperature. Afterrinsing with blocking buffer, the synaptosomes were incubated witha mixture of goat anti-mouse DyLight 549– and goat anti-rabbitfluorescein isothiocyanate–conjugated secondary antibodies (1:200;Jackson ImmunoResearch Inc., West Grove, PA) for 1 hour at roomtemperature. The synaptosomes were then washed three times withPB and 0.1 M carbonate buffer (pH 9.2) and coverslipped withfluorescencemountingmedium (DAKONorthAmerica, Inc., Carpinteria,CA). Double immunofluorescence images were observed at a magnifica-tion of 400� using upright fluorescence microscopy (Leica DM2000 LED;Leica Microsystems, Wetzlar, Germany), and images were capturedusing a charge-coupled device camera (SPOT RT3; Diagnostic Instru-ments, Sterling Heights, MI).

Western Blotting. Synaptosomes (0.5mg of protein/ml) from controland drug-treated groups were lysed in a pH-7.5 ice-cold Tris-HCl buffersolution containing (in mM) 20 Tris-HCl, 1% Triton X-100, 1 EDTA, 1EGTA, 150 NaCl, 2.5 sodium pyrophosphate, 1 b-glycerophosphate, 1phenylmethanesulfonyl fluoride, 1 sodium orthovanadate, and 1 mg/mlleupeptin. The lysates were sonicated for 10 seconds and then

centrifuged at 13,000g at 4°C for 10 minutes. Equal amounts ofsamples were separated through electrophoresis on 7.5% SDS-PAGE,and then transferred to nitrocellulose membranes. The membraneswere blockedwith Tris-buffered saline that contained 5% low-fat milk andincubated with appropriate primary antibodies [anti-phospho–proteinkinase A (PKA), 1:1000; Cell Signaling Technology]. After incubationwith appropriate horseradish peroxidase–conjugated donkey anti-rabbit IgG secondary antibodies (1:1000; Cell Signaling Technology),protein bands were detected using the enhanced chemiluminescencesystem (Amersham, Buckinghamshire, UK). An aliquot of samples wasloaded and probed with anti–b-actin antibodies for detecting b-actin asa loading control. Films were scanned using a scanner, and the level ofexpression or phosphorylation was assessed using band density, whichwas quantified through densitometry.

Statistical Analysis. All data are expressed as the mean 6 S.E.M. Student’s t tests were used for statistical analysis between twogroups, whereas an analysis of variance with least significantdifference comparisons and post-hoc tests were used to analyze morethan two groups. Analysis was completed using SPSS software(version 17.0; SPSS Inc., Chicago, IL). P , 0.05 was considered torepresent a significant difference.

ResultsCelecoxib Inhibits 4-AP–Induced Glutamate Release

in Rat Cerebrocortical Synaptosomes by ReducingClassic External Ca21-Dependent Exocytosis. To inves-tigate the effect of celecoxib on glutamate release, isolatednerve terminals were depolarized with the potassium channelblocker 4-AP, which has been shown to open VDCCs andinduce the release of glutamate (Nicholls, 1998). In synapto-somes incubated in the presence of 1.2 mM CaCl2, 4-AP(1mM) induced a glutamate release of 7.36 0.1 nmol/mg per 5minutes, and this release was blocked by the Na1 channelblocker TTX (2 mM) (P 5 0.000; Fig. 1A). This is consistentwith the observation that 4-AP is able to initiate TTX-sensitive depolarization and Ca21-dependent glutamaterelease in synaptosomes (Nicholls, 1998). Application ofcelecoxib (30mM) reduced the amount of 4-AP–induced glutamaterelease to 3.9 6 0.2 nmol/mg per 5 minutes (P 5 0.000)without altering the basal release of glutamate (Fig. 1A).The inhibitory effect of celecoxib on 4-AP–induced gluta-mate release was concentration-dependent, and producedan IC50 value of approximately 7 mM, which was derivedfrom a dose-response curve (Fig. 1B).We next investigated whether the inhibition of 4-AP–induced

glutamate release by celecoxib was mediated by an effect onexocytotic vesicular release, or on Ca21-independent releaseattributable to cytosolic efflux through the reversal ofthe glutamate transporter (Nicholls et al., 1987). The Ca21

-independent glutamate efflux was measured by depolarizingthe synaptosomes with 1 mM 4-AP in extracellular Ca21-freesolution that contained 300 mM EGTA (Fig. 1C). Under theseconditions, the release of glutamate induced by 4-AP was 2.3 60.3 nmol/mg per 5 minutes [F(2,12) 5 164.837; P 5 0.000;Fig. 1C]. The Ca21-independent release induced by 4-AP, how-ever, was not affected by 30 mMcelecoxib (2.46 0.1 nmol/mg per5 minutes; P 5 0.882; Fig. 1C). Similar results were also ob-tained using BAPTA [1,2-bis(2-aminophenoxy) ethane-N,N,N9,N9-tetraacetic acid-acetoxymethyl ester], a cell-permeableCa21 chelator. BAPTA (50 mM) alone reduced the 4-AP–inducedglutamate release [F(2,12) 5 163.562; P 5 0.000; Fig. 1C].However, the release measured in the presence of bothBAPTA and celecoxib was similar to that obtained in the

136 Lin et al.

at ASPE

T Journals on A

ugust 19, 2020jpet.aspetjournals.org

Dow

nloaded from

presence of BAPTA alone (P5 0.963; Fig. 1C). In addition, weexamined the action of celecoxib in the presence of DL-threo-b-benzyl-oxy aspartate (DL-TBOA), a nonselective inhibitor ofall excitatory amino acid transporter subtypes, or bafilomycinA1, which depletes vesicle content by inhibiting the synapticvesicle H1-ATPase that drives glutamate uptake. In thepresence of DL-TBOA (10 mM), although the 4-AP–inducedglutamate release was increased by the inhibitor (because ofinhibition of reuptake of the released glutamate) (P , 0.001),celecoxib (30 mM) continued to reduce the 4-AP–inducedrelease of glutamate significantly [F(2,12)5 71.69; P5 0.000;Fig. 1C]. In contrast to DL-TBOA, bafilomycin A1 (0.1 mM)

reduced the control 4-AP (1 mM)–induced glutamate release[F(2,12) 5 120.558; P 5 0.000; Fig. 1C], and completely pre-vented the inhibitory effect of celecoxib (30 mM) on the4-AP–induced glutamate release. Therefore, in the presence ofbafilomycin A1, celecoxib (30 mM) induced a statistically non-significant inhibition (P5 0.972; Fig. 1C). We also examined theeffect of DON, an inhibitor of glutaminase, which catalyzes theconversion of glutamine to glutamate. Figure 1C shows thatDON (100mM) had no effect on 4-AP–induced glutamate release(P 5 0.332). In the presence of DON, celecoxib (30 mM) stilleffectively inhibited the 4-AP–induced glutamate release[F(2,9) 5 38.17; P 5 0.000; Fig. 1C).

Fig. 1. Celecoxib inhibits the 4-AP–induced release of glutamate in rat cerebrocortical synaptosomes. (A) Glutamate release was induced by theaddition of 1 mM 4-AP in the absence (control) or presence of 30 mΜ celecoxib or 2 mM TTX, added 10 minutes before depolarization. (B) The log dose-response curve for celecoxib inhibition of 4-AP–induced glutamate release, fitted using a logistic function. (C) 4-AP–induced glutamate release undercontrol conditions or in the presence 30 mΜ celecoxib, 300 mΜ EGTA (without CaCl2), 300 mΜ EGTA (without CaCl2) and 30 mΜ celecoxib, 50 mΜ BAPTA,50 mΜ BAPTA and 30 mΜ celecoxib, 10 mΜ DL-TBOA, 10 mΜ DL-TBOA and 30 mΜ celecoxib, 0.1 mΜ bafilomycin A1, 0.1 mΜ bafilomycin A1 and 30 mΜcelecoxib, 100 mMDON, or 100 mMDON and 30 mΜ celecoxib. Celecoxib was added 10 minutes before depolarization, and other drugs 10 minutes beforethis. Results are presented as the mean6 S.E.M. of independent experiments, using synaptosomal preparations from five animals. ***P, 0.001 versusthe control group; #P , 0.05 versus the DL-TBOA–treated group.

Inhibition of Glutamate Release by Celecoxib 137

at ASPE

T Journals on A

ugust 19, 2020jpet.aspetjournals.org

Dow

nloaded from

Celecoxib Reduces the 4-AP–Induced Increase in[Ca21]C and Glutamate Release by Inhibiting Cav2.2(N-Type) and Cav2.1 (P/Q-Type) Channels. Transmitterrelease can bemodulated by regulating the plasmamembranelevel, and consequently altering the calcium influx. To in-vestigate the potential mechanisms underlying the celecoxib-mediated inhibition of glutamate release, the effect of celecoxibon intrasynaptosomal Ca21 levels was determined using theCa21 indicator fura-2. As indicated in Fig. 2A, 1 mM 4-APcaused an increase in [Ca21]C from 157.46 2.6 nM to a plateau

level of 212.8 6 5.6 nM (P 5 0.001). Applying 30 mM celecoxibdid not affect basal Ca21 levels (155.66 2.9 nM), but caused anapproximate 13% decrease in the 4-AP–induced increase in[Ca21]c (184.6 6 5.2 nM; P 5 0.01; Fig. 2A).In the adult rat cerebrocortical nerve terminals, elevation of

the [Ca21]C and glutamate release induced by depolarizationwas supported by the Ca21 influx through Cav2.2 (N-type) andCav2.1 (P/Q-type) channels (Vazquez and Sanchez-Prieto,1997; Millan and Sanchez-Prieto, 2002). To investigatewhether the decrease in Ca21 channel activity was involved in

Fig. 2. Celecoxib reduces the 4-AP–induced increase in cytosolic Ca2+ levels, and this effect is prevented by blocking the Cav2.2 (N-type) and Cav2.1(P/Q-type) channels. (A) Cytosolic-free Ca2+ concentration (in nanomoles) was measured in the absence (control) and presence of 30 mΜ celecoxib, added10 minutes before depolarization with 1 mM 4-AP. (B and C) The elevation of cytosolic Ca2+ levels and glutamate release were induced by 1 mM 4-AP, inthe absence (control) or presence of 30 mΜ celecoxib, 2 mM v-CgTX MVIIC, 2 mM v-CgTX MVIIC and 30 mΜ celecoxib, 50 mM dantrolene, 50 mMdantrolene and 30 mΜ celecoxib, 100 mMCGP37157, or 100 mMCGP37157 and 30 mΜ celecoxib. Celecoxib was added 10 minutes before depolarization,whereas the other drugs were added 30 minutes before depolarization. Results are presented as the mean 6 S.E.M. of independent experiments, usingsynaptosomal preparations from five animals. ***P , 0.001 versus the control group; **P , 0.01 versus the control group; #P , 0.05 versus thedantrolene- or CGP37157-treated group.

138 Lin et al.

at ASPE

T Journals on A

ugust 19, 2020jpet.aspetjournals.org

Dow

nloaded from

the celecoxib-mediated inhibition of 4-AP–induced increase in[Ca21]C and glutamate release, we examined the effect ofcelecoxib in the presence of v-CgTX MVIIC, a wide spectrumblocker of Cav2.2 (N-type) and Cav2.1 (P/Q-type) channels. InFig. 2, B and C, 2 mM v-CgTX MVIIC substantially reducedthe 1mM4-AP–induced increase in [Ca21]C [F(2,12)5 65.968;P 5 0.000] and glutamate release [F(2,12) 5 151.722; P 50.000]. Although the 4-AP–induced increase in [Ca21]C andglutamate release was considerably reduced in the presence ofcelecoxib (30 mM) (P 5 0.000), this effect was prevented by thepresence of v-CgTXMVIIC. The [Ca21]C and release measuredin the presence of both v-CgTX MVIIC and celecoxib wassimilar to that obtained in the presence of v-CgTX MVIICalone (P. 0.05) (Fig. 2, B and C). In addition, to test the role ofCa21 release from intracellular stores such as the endoplasmicreticulum (ER) and mitochondria (Berridge, 1998), we exam-ined the effect of dantrolene, an inhibitor of intracellular Ca21

release from the ER, and CGP37157, a membrane-permeantblocker of mitochondrial Na1/Ca21 exchange. As indicated inFig. 2, B and C, 100 mM dantrolene reduced the 1 mM4-AP–induced increase in [Ca21]C [F(2,12)5 55.693; P5 0.000;Fig. 1C] and glutamate release [F(2,12) 5 138.982; P 5 0.000;Fig. 1C], indicating that Ca21 release from ER ryanodinereceptors contributes substantially to the 4-AP–induced in-crease in [Ca21]C and glutamate release. In the presence ofdantrolene, however, celecoxib (30 mΜ) still effectively in-hibited the 4-AP–induced increase in [Ca21]C and glutamaterelease (P , 0.05; Fig. 2, B and C). Similar results were alsoobserved using 100 mΜ CGP37157 (Fig. 2, B and C).Celecoxib Does Not Alter the Synaptosomal Mem-

brane Potential and Depolarizaton-Induced Na1 In-flux. Inhibition of [Ca21]C elevation by celecoxib might beattributable either to a direct reduction in the amount of Ca21

entering through VDCCs, or to secondary effects caused by,for example, the modulation of potassium channels and theconsequently altered plasma membrane potential. To discernbetween these two possibilities, the effect of celecoxib on thesynaptosomal plasma membrane potential under restingconditions and on depolarization was examined using mem-brane potential–sensitive dye DiSC3(5). Figure 3A shows that1 mM 4-AP caused an increase in DiSC3(5) fluorescence (11.960.5 fluorescence units/5 minutes). Preincubation of the synap-tosomes with 30 mM celecoxib for 10 minutes before adding4-AP did not alter the resting membrane potential, andproduced no substantial change in the 4-AP–mediated increasein DiSC3(5) fluorescence (12.16 0.6 units/5 minutes; P5 0.937;Fig. 3A). This suggested that the observed inhibition of4-AP–induced glutamate release by celecoxib is unlikely to bedue to a hyperpolarizing effect of the drug on the synaptosomalplasmamembrane potential or an attenuation of depolarizationproduced by 4-AP. Confirmation that the celecoxib effect did notaffect synaptosomal excitability was obtained through experi-ments using high external [K1]-mediated depolarization.Elevated extracellular KCl depolarized the plasma membraneby shifting the K1 equilibrium potential above the threshold foractivation of voltage-dependent ion channels. Although Na1

channels are inactivated under these conditions, VDCCs areactivated nonetheless to mediate Ca21 entry, which supportsneurotransmitter release (Barrie et al., 1991). As indicated bythe inset in Fig. 3A, 15 mM KCl induced a glutamate release of6.9 6 0.4 nmol/mg per 5 minutes, which was reduced to 4.8 60.4 nmol/mg per 5 minutes in the presence of 30 mΜ celecoxib

(P5 0.006). In addition, the Na1-sensitive probe SBFIwas usedto measure cytosolic Na1 levels. Figure 3B indicates that 4-AP(1 mM) caused a clear rise in Na1 influx, but celecoxib (30 mM)failed to affect this increase (P5 0.996). The failure of celecoxibto produce an effect on this increase was not caused by aninsufficient level of sensitivity of the SBFI probe to alterationsin Na1 channel activity because in parallel experiments, theNa1 channel blocker TTX (2 mM) caused an 86% inhibition of4-AP–induced Na1 influx (P 5 0.000; Fig. 3B).PGE2/EP2 Is Involved in the Celecoxib-Mediated

Inhibition of Glutamate Release. PGE2, the predominantreaction product of COX-2, has been reported to increaseglutamate release (Nishihara et al., 1995). If the effect ofcelecoxib on glutamate release results from the blocking ofPGE2 synthesis, exogenous application of PGE2 shouldocclude the celecoxib-mediated inhibition of glutamate re-lease. To test this hypothesis, the effect of celecoxib on4-AP–induced glutamate release in the absence or presence ofPGE2 was compared. As shown in Fig. 4A, 1 mM 4-AP induceda glutamate release of 6.76 0.3 nmol/mg per 5minutes, whichwas facilitated by 45% (9.7 6 0.5 nmol/mg per 5 minutes) inthe presence of PGE2 (5 mM) [F(2,13) 5 7.86; P 5 0.006].Celecoxib (30 mM) alone reduced the 4-AP–induced glutamaterelease to 3.9 6 0.4 nmol/mg per 5 minutes (P 5 0.000), butthis inhibition was abolished by pretreatment with PGE2, andno statistical difference was observed between the releaseafter PGE2 alone and after the PGE2 and celecoxib treatment(9.56 0.6 nmol/mg per 5 minutes; P5 0.986; Fig. 4, A and B).Similarly, 5 mM butaprost, a EP2 agonist, enhanced the4-AP–induced glutamate release [F(2,13)5 10.539; P5 0.003]and completely prevented the inhibition of glutamate releaseby 30 mM celecoxib (P5 0.492; Fig. 4B). Furthermore, 100 mMPF04418948 (an EP2 antagonist) reduced 4-AP–inducedglutamate release [F(2,13) 5 45.782; P 5 0.000] andcompletely prevented the effect of celecoxib (P 5 0.954;Fig. 4B). By contrast, 5 mM sulprostone, a PGE2 receptor 3(EP3) agonist, exerted no effect on 4-AP–induced glutamaterelease (P 5 1.000). In the presence of sulprostone, 30 mMcelecoxib still effectively inhibited 4-AP–induced glutamaterelease [F(2,12)5 27.516; P5 0.000; Fig. 4B]. A similar resultwas also obtained using TCS2510, a PGE2 receptor 4 (EP4)agonist. As reported in Fig. 4B, TCS2510 (100 mM) exertedno effect on either control 4-AP–induced glutamate release(P 5 0.603) or inhibition of glutamate release by celecoxib[F(2,12) 5 83.813; P 5 0.000; Fig. 4B]. These results sug-gested that the observed effect of celecoxib on glutamaterelease is mediated by PGE2/EP2 receptors.COX-2 and EP2 Are Expressed in Cerebrocortical

Nerve Terminals. To determine whether COX-2 and EP2are expressed in purified cerebrocortical synaptosomes, weperformed a double immunostaining of COX-2 and EP2 usingsynaptophysin, a presynaptic marker. As shown in Fig. 5, thenerve terminals that contained synaptophysin also containedCOX-2 or EP2, confirming that COX-2 and EP2 are present inpresynaptic terminals, which is consistent with previousstudies (Sang et al., 2005; Zhu et al., 2005).The PKA Pathway Is Involved in the Celecoxib-

Mediated Inhibition of Glutamate Release. BecauseEP2 is linked to the Gs-cAMP/PKA pathway (Sang et al.,2005), we used PKA inhibitors H89 and KT5720 to examinewhether the celecoxib-mediated inhibition of glutamate re-lease can be prevented. Figure 6A shows that 100 mM H89

Inhibition of Glutamate Release by Celecoxib 139

at ASPE

T Journals on A

ugust 19, 2020jpet.aspetjournals.org

Dow

nloaded from

reduced 1 mM 4-AP–induced glutamate release [F(2,13) 57.115; P 5 0.01], which indicated basal PKA activity. Al-though celecoxib (30 mM) reduced the 4-AP–induced gluta-mate release (P , 0.001), this effect was prevented by H89;the release measured in the presence of H89 and celecoxibwas similar to that obtained in the presence of H89 alone (P50.837; Fig. 6, A and B). Similar results were also observed inthe presence of KT5720 (1 mM) (P 5 1.0; Fig. 6B). We alsodetermined whether direct activation of PKA is capable ofpreventing the inhibitory effect of celecoxib on glutamaterelease. The application of PKA activator forskolin (100 mM)facilitated 4-AP–evoked glutamate release (P , 0.001; Fig.6B). Interestingly, however, when celecoxib (30 mM) wasapplied together with forskolin, the inhibitory effect ofcelecoxib on the release of glutamate evoked by 4-AP wasprevented (P 5 0.462; Fig. 6B). In addition, the mitogen-activated protein kinase inhibitor PD98059 (50 mM) reduced 1

mM 4-AP–induced glutamate release (P , 0.001). In thepresence of PD98059, however, 30 mΜ celecoxib still effec-tively inhibited 4-AP–induced glutamate release [F(2,12) 576.364; P 5 0.000; Fig. 6B]. Similar to PD98059, the proteinkinase C inhibitor GF109203X (10mM) decreased 4-AP–inducedglutamate release [F(2,12) 5 68.696; P 5 0.000], but exertedno effect on the celecoxib-mediated inhibition of 4-AP–inducedglutamate release (Fig. 6B).To further confirm that the PKA signaling pathway was

suppressed by celecoxib during its inhibition of 4-AP–inducedglutamate release, we determined the effect of celecoxib onthe phosphorylation of PKA in cerebrocortical synaptosomes.As indicated in Fig. 6C, depolarization of synaptosomes with1 mM 4-AP in the presence of 1.2 mM CaCl2 increased thephosphorylation of PKA (126.1 6 2.1%; P , 0.001). Whensynaptosomes were pretreated with 30 mM celecoxib for 10minutes before depolarization with 4-AP, a significant decrease

Fig. 3. Celecoxib does not alter the synaptosomal mem-brane potential and Na+ influx. Synaptosomal membranepotential (A) or Na+ influx (B) was measured in the absence(control) and in the presence of 30 mM celecoxib or 2 mMTTX, added 10 minutes before depolarization with 1 mM4-AP. (Inset) Glutamate release was induced by 15 mM KClin the absence (control) or presence of 30 mM celecoxib,added 10 minutes before depolarization. Results are pre-sented as the mean 6 S.E.M. of independent experiments,using synaptosomal preparations from six animals. ***P ,0.001 versus the control group; **P , 0.01 versus thecontrol group.

140 Lin et al.

at ASPE

T Journals on A

ugust 19, 2020jpet.aspetjournals.org

Dow

nloaded from

in the 4-AP–induced PKA phosphorylation was observed[106.6 6 2.5%; F(2,6) 5 52.215; P 5 0.001; Fig. 6C].

DiscussionCelecoxib, a selective COX-2 inhibitor, has recently received

considerable attention because of its beneficial effects onnumerous CNS diseases, including cerebral ischemia andneurodegenerative disorders (Drachman et al., 2002; Small

et al., 2008; Abd El-Aal et al., 2013). Excessive glutamate releaseis involved in the pathophysiology of these diseases, and re-ducing glutamate release might provide a potential strategyfor treating these diseases (Meldrum and Garthwaite, 1990;Meldrum, 2000). Therefore, we elucidated the potential roleof celecoxib in glutamate release in rat cerebral cortex nerveterminals (synaptosomes). We demonstrated for the first timethat celecoxib inhibits depolarization-induced glutamate release.The possible underlying mechanisms for the celecoxib-mediated

Fig. 4. Inhibition of glutamate release by celecoxib is mediated by a PGE2 EP2 receptor. (A) Glutamate release was induced by 1 mM 4-AP in theabsence (control) or presence of 30 mΜ celecoxib, 5 mM PGE2, and 5 mM PGE2 and 30 mΜ celecoxib. (B) Glutamate release was induced by 1 mM 4-AP inthe absence (control) or presence of 30 mΜ celecoxib, 5 mMPGE2, 5 mMPGE2 and 30 mΜ celecoxib, 5 mM butaprost, 5 mM butaprost and 30 mΜ celecoxib,100 mMPF04418948, 100 mMPF04418948 and 30 mΜ celecoxib, 5 mM sulprostone, 5 mM sulprostone and 30 mΜ celecoxib, 100 mMTCS2510, or 100 mMTCS2510 and 30 mΜ celecoxib. Celecoxib was added 10 minutes before depolarization, and other drugs 10 minutes before this. Results are presented asthe mean6 S.E.M. of independent experiments, using synaptosomal preparations from five to six animals. ***P, 0.001 versus the control group; **P,0.01 versus the control group; #P , 0.05 versus the sulprostone-treated group.

Fig. 5. Double-labeled immunofluorescence of synaptophysin, a major presynaptic vesicle-associated protein, COX-2, and EP2 in synaptosomes.Example pictures of immunofluorescence staining demonstrate synaptophysin (red) (A and E), COX-2 (green) (B), EP2 (green) (F), and merged images(orange) (C and G) of two proteins. (D and H) Higher magnifications of colocalization (orange) of the rectangles in C and G, respectively. Red arrowsindicate synaptophysin-positive staining only. Green arrows indicate synaptosomes labeled with COX-2 and EP2. Orange arrows indicate doublelabeling. The scale bar for A–C and E–G is 50 mm, and that for D is 12 mm.

Inhibition of Glutamate Release by Celecoxib 141

at ASPE

T Journals on A

ugust 19, 2020jpet.aspetjournals.org

Dow

nloaded from

inhibition of glutamate release were investigated in the presentstudy.The process through which neurotransmitter release oc-

curs is complex, involving Na1, K1, and Ca21 channels (Wuand Saggau, 1997; Nicholls, 1998). The inhibition of Na1 chan-nels and the activation of K1 channels stabilizes membrane

excitability and, consequently, causes a reduction in the levelsof Ca21 entry and neurotransmitter release (Rehm andTempel, 1991; Li et al., 1993). According to this concept, theobserved inhibitory effect of celecoxib on induced glutamaterelease might occur through a reduction of nerve terminalexcitability, but this is unlikely because of the following

Fig. 6. PKA pathway is involved in the celecoxib-mediated inhibition of 4-AP–induced glutamate release. (A) Glutamate release was induced by 1 mM4-AP in the absence (control) or presence of 30 mΜ celecoxib, 100 mMH89, and 100 mMH89 and 30 mΜ celecoxib. (B) Glutamate release was induced by 1mM 4-AP, in the absence (control) or presence of 30 mΜ celecoxib, 100 mMH89, 100 mMH89 and 30 mΜ celecoxib, 1 mMKT5720, 1 mMKT5720 and 30 mΜcelecoxib, 100 mM forskolin, 100 mM forskolin and 30 mΜ celecoxib, 50 mMPD98059, 50 mMPD98059 and 30 mΜ celecoxib, 10 mMGF109203X, or 10 mMGF109203X and 30 mΜ celecoxib. H89, KT5720, PD98059, or GF109203X was added 40 minutes before depolarization, whereas celecoxib was added10 minutes before depolarization. (C) Phosphorylation of PKA was detected in the absence (control) or presence of 1 mM 4-AP, or 1 mM 4-AP and 30 mMcelecoxib. Data are expressed as a percentage of the control phosphorylation obtained in the absence of 4-AP stimulation. Results are presented as themean 6 S.E.M. of independent experiments, using synaptosomal preparations from four to five animals. ***P , 0.001 versus the control group; **P ,0.01 versus the control group; #P , 0.05 versus the PD98059- or GF109203X-treated group.

142 Lin et al.

at ASPE

T Journals on A

ugust 19, 2020jpet.aspetjournals.org

Dow

nloaded from

observations. First, celecoxib inhibited the release of glutamateinduced by 4-AP and KCl. Because 4-AP–induced glutamaterelease involves the activation of Na1 and Ca21 channels, 15mM external KCl–induced glutamate release involves onlyCa21 channels (Barrie et al., 1991; Nicholls, 1998); thus, Na1

channels are not involved in the effect of celecoxib on glutamaterelease. This was supported by our observation that celecoxibdid not affect 4-AP–induced Na1 influx. Second, celecoxib wasnot observed to substantially affect the synaptosomal plasmamembrane potential, indicating that celecoxib does not affectK1 conductance. Third, celecoxib did not significantly inhibit4-AP–induced Ca21-independent glutamate release, a compo-nent of glutamate release that is exclusively dependent onmembrane potential (Nicholls et al., 1987; Attwell et al., 1993).Furthermore, the celecoxib-mediated inhibition of 4-AP–inducedglutamate release was prevented by the vesicular transporterinhibitor bafilomycin A1, but was insensitive to the glutamatetransporter inhibitor DL-TBOA. This indicated that celecoxibdoes not affect glutamate release by reversing the direction ofthe plasma membrane glutamate transporter. These datasuggest that the celecoxib-mediated inhibition of 4-AP–inducedglutamate release is mediated by a decrease in the exocytoticpool of release. Moreover, this phenomenon is not the resultof a reduction in synaptosomal excitability caused by ionchannel (e.g., Na1 or K1 channels) modulation. This findingis inconsistent with that of previous electrophysiologicalstudies, which have shown that celecoxib modulates Na1

and K1 currents in rat retinal and ganglion neurons (Parket al., 2007; Du et al., 2011; Mi et al., 2013). The reasons forthe difference in results are unclear, but might be attribut-able to the distinct experimental models used; previousstudies have used a cell culture model, whereas we useda nerve terminal (synaptosomal) model.Therefore, if the effect is not caused by the modulation of

synaptosomal excitability, celecoxib might inhibit glutamaterelease by decreasing the levels of Ca21 entry through theCav2.2 (N-type) and Cav2.1 (P/Q-type) Ca21 channels that arecoupled to glutamate exocytosis in the nerve terminals (Vazquezand Sanchez-Prieto, 1997; Millan and Sanchez-Prieto, 2002).

This hypothesis is plausible because the present studydemonstrated that celecoxib reduced the 4-AP–induced in-crease in [Ca21]C and glutamate release, and that this effectwas prevented by blocking the Cav2.2 (N-type) and Cav2.1(P/Q-type) channels. Conversely, the reduced release of storedCa21 from the ER ryanodine receptors and mitochondriaduring the celecoxib-mediated inhibition of glutamate releasecan be excluded. This is because the inhibitory effect ofcelecoxib on the 4-AP–induced increase in [Ca21]C and glu-tamate release was insensitive to the ER ryanodine receptorinhibitor dantrolene and the mitochondrial Na1/Ca21 exchangeinhibitor CGP37157. Although no direct evidence has indicatedthat celecoxib acts on presynaptic Ca21 channels, these resultsimplied that the suppression of Ca21 influx through Cav2.2 andCav2.1 channels is involved in the inhibition of glutamaterelease induced by celecoxib.Determining how celecoxib inhibits the Cav2.2 and Cav2.1-

channels and glutamate release is critical. Celecoxib isa selective COX-2 inhibitor that inhibits the production ofPGE2. PGE2 is lipophilic, and diffuses rapidly and activatesits specific membrane receptors (EP1–4). These receptors areG-protein–coupled receptors, and their intracellular signal-ing is distinct. The EP1 receptors couple with the Gq-phospholipase C–inositol triphosphate pathway. The EP2 andEP4 receptors are linked to the Gs-cAMP/PKA pathway andincrease cAMP levels. Conversely, activation of the EP3receptors reduces cAMP production through a pertussistoxin–sensitive Gi-coupled pathway (Breyer et al., 2001;Sugimoto and Narumiya, 2007). PGE2 has been shown toincrease glutamate release through presynaptic EP2 signal-ing (Nishihara et al., 1995; Sang et al., 2005). Thus,speculating that the inhibitory effect of celecoxib on Ca21

influx and glutamate release is a consequence of reducedPGE2 signaling is reasonable. In elucidating this hypothesis,we observed that 1) a PGE2 or EP2 agonist increasedglutamate release and abrogated the inhibitory effect ofcelecoxib on 4-AP–induced glutamate release; 2) an EP3 orEP4 agonist did not elicit effects on either 4-AP–inducedglutamate release or the celecoxib-mediated inhibition of

Fig. 7. Potential mechanisms by which celecoxibinhibits glutamate release. In rat cerebrocorticalnerve terminals, the inhibition of COX-2 activity bycelecoxib induces the reduction of PGE2 levels,causing a decrease in EP2 activation and PKAactivity. This, in turn, reduces the Ca2+ influxthrough N- and P/Q-type Ca2+ channels to causea decrease in glutamate release.

Inhibition of Glutamate Release by Celecoxib 143

at ASPE

T Journals on A

ugust 19, 2020jpet.aspetjournals.org

Dow

nloaded from

glutamate release; 3) the inhibitory effect of celecoxib on4-AP–induced glutamate release was blocked by an EP2antagonist; 4) the cerebrocortical nerve terminals expressedCOX-2 and EP2 receptors, and these two proteins were co-expressed with synaptophysin, a presynaptic marker, withinthe same nerve terminals; 5) PKA inhibitors preventedthe inhibitory effect of celecoxib on glutamate release, butneither the mitogen-activated protein kinase inhibitor nor theprotein kinase C inhibitor exerted any effects; 6) activatingPKA prevented the celecoxib-mediated inhibition of glu-tamate release; and 7) celecoxib significantly reduced the4-AP–induced phosphorylation of PKA. These results sug-gested that presynaptic EP2, not EP3 or EP4, is involved inthe inhibition of 4-AP–evoked glutamate release produced bycelecoxib. Furthermore, these data confirmed the involve-ment of the PKA pathway in the action of celecoxib. In nerveterminals, PKA has been shown to phosphorylate VDCCs andseveral synaptic proteins, subsequently increasing glutamaterelease (Herrero and Sanchez-Prieto, 1996; Chheda et al.,2001). Thus, a reduction of presynaptic Cav2.2 and Cav2.1channel phosphorylation should be considered when deter-mining the possible mechanism of celecoxib-mediated pre-synaptic inhibition. However, this possibility cannot beexamined in the current study, because antibodies for thephosphorylation of the N- and P/Q-type calcium channels arenot available. Based on our data, Fig. 7 presents a model toexplain the mechanism through which celecoxib inhibitsglutamate release from cortical nerve terminals. In brief,the inhibition of COX-2 activity by celecoxib induces thereduction of PGE2 levels, causing a decrease in EP2 activationand PKA activity. This, in turn, reduces Ca21 influx andglutamate release. COX-2–derived prostaglandins other thanPGE2, such as prostaglandin D2 and prostaglandin F2a, areamong the most abundant prostaglandins in the brain andregulate numerous functions, including pain and sleep(Hertting and Seregi, 1989; Urade and Hayaishi, 1999). Thesefunctions are thought to be associated with the glutamatesystem (Gilmour et al., 2013; Suto et al., 2014). Therefore, therole of prostaglandin D2 or prostaglandin F2a in the celecoxib-mediated inhibition of glutamate release cannot be ruled out.In addition, the possible involvement of other presynapticpathways should be considered. For example, GABAA recep-tors are expressed in the brain and localize both pre- andpostsynaptically. At the presynaptic level, GABAA receptorshave been shown to inhibit Ca21 influx and glutamate release(Long et al., 2009). Future studies are needed to determinewhether GABAA receptors play any role in the celecoxib-mediated inhibition of glutamate release.Celecoxib at 10–30 mg/kg exhibits numerous pharmacolog-

ical effects, such as the inhibition of inflammatory processes,stabilization of mitochondrial function, and attenuation ofoxidative stress. These effects are likely to be associated withthe neuroprotective activity of celecoxib (Hunter et al., 2007;López-Villodres et al., 2012; Kaizaki et al., 2013). We observedthat celecoxib reduced glutamate release from nerve termi-nals, and this effect was dose-dependent and peaked at 30 mMwith an IC50 of 7 mM. This suggested that the decrease inreleased glutamate presents an additional explanation for theneuroprotective effect of celecoxib. This hypothesis was basedon the excessive glutamate release and activation of gluta-mate receptors resulting in neurotoxic cell damage that hasbeen implicated in the neuropathology of acute and chronic

brain disorders (Meldrum and Garthwaite, 1990; Meldrum,2000). The blockade of glutamate neurotransmission, such asglutamate receptor antagonists, has conferred neuroprotec-tion in in vitro and in vivo studies (Schauwecker, 2010;Yeganeh et al., 2013). The reduction of glutamate release is aneven earlier phenomenon, and can prevent the postsynaptictoxicity of glutamate at all glutamate receptors.In conclusion, the current study provides evidence that the

COX-2 inhibitor celecoxib exerts an inhibitory effect on glu-tamate release. This effect might be exerted mainlythrough the suppression of the presynaptic EP2/PKApathway. COX-2 expression is significantly increased inpatients with Alzheimer’s disease and in experimental strokeor epilepsy, suggesting that COX-2 and its reaction productsare involved in several neurologic disorders (Nogawa et al.,1997; Minghetti, 2004). Thus, our finding is valuable becauseit provides new insight into the mechanism by which COX-2inhibitors act in the brain.

Authorship Contributions

Participated in research design: Lin, S.-J. Wang.Conducted experiments: Lin.Performed data analysis: Lu, Lin, C.-C. Wang, Huang, S.-J. Wang.Wrote or contributed to the writing of the manuscript: S.-J. Wang.

References

Abd El-Aal SA, El-Sawalhi MM, Seif-El-Nasr M, and Kenawy SA (2013) Effect ofcelecoxib and L-NAME on global ischemia-reperfusion injury in the rat hippo-campus. Drug Chem Toxicol 36:385–395.

Akerman KE, Scott IG, Heikkilä JE, and Heinonen E (1987) Ionic dependence ofmembrane potential and glutamate receptor-linked responses in synaptoneur-osomes as measured with a cyanine dye, DiS-C2-(5). J Neurochem 48:552–559.

Akram A, Gibson CL, and Grubb BD (2013) Neuroprotection mediated by the EP₄receptor avoids the detrimental side effects of COX-2 inhibitors following ischaemicinjury. Neuropharmacology 65:165–172.

Attwell D, Barbour B, and Szatkowski M (1993) Nonvesicular release of neuro-transmitter. Neuron 11:401–407.

Barrie AP, Nicholls DG, Sanchez-Prieto J, and Sihra TS (1991) An ion channel locusfor the protein kinase C potentiation of transmitter glutamate release from guineapig cerebrocortical synaptosomes. J Neurochem 57:1398–1404.

Berridge MJ (1998) Neuronal calcium signaling. Neuron 21:13–26.Breyer RM, Bagdassarian CK, Myers SA, and Breyer MD (2001) Prostanoid recep-tors: subtypes and signaling. Annu Rev Pharmacol Toxicol 41:661–690.

Chen C, Magee JC, and Bazan NG (2002) Cyclooxygenase-2 regulates prostaglandinE2 signaling in hippocampal long-term synaptic plasticity. J Neurophysiol 87:2851–2857.

Chheda MG, Ashery U, Thakur P, Rettig J, and Sheng ZH (2001) Phosphorylation ofSnapin by PKA modulates its interaction with the SNARE complex. Nat Cell Biol3:331–338.

Cowley TR, Fahey B, and O’Mara SM (2008) COX-2, but not COX-1, activity isnecessary for the induction of perforant path long-term potentiation and spatiallearning in vivo. Eur J Neurosci 27:2999–3008.

Coyle JT and Puttfarcken P (1993) Oxidative stress, glutamate, and neurodegener-ative disorders. Science 262:689–695.

Drachman DB, Frank K, Dykes-Hoberg M, Teismann P, Almer G, Przedborski S,and Rothstein JD (2002) Cyclooxygenase 2 inhibition protects motor neuronsand prolongs survival in a transgenic mouse model of ALS. Ann Neurol 52:771–778.

Du XN, Zhang X, Qi JL, An HL, Li JW, Wan YM, Fu Y, Gao HX, Gao ZB, and Zhan Y,et al. (2011) Characteristics and molecular basis of celecoxib modulation on K(v)7potassium channels. Br J Pharmacol 164:1722–1737.

Fan LW, Kaizaki A, Tien LT, Pang Y, Tanaka S, Numazawa S, Bhatt AJ, and Cai Z(2013) Celecoxib attenuates systemic lipopolysaccharide-induced brain in-flammation and white matter injury in the neonatal rats. Neuroscience 240:27–38.

Fonnum F (1984) Glutamate: a neurotransmitter in mammalian brain. J Neurochem42:1–11.

Gilmour G, Broad LM, Wafford KA, Britton T, Colvin EM, Fivush A, Gastambide F,Getman B, Heinz BA, and McCarthy AP, et al. (2013) In vitro characterisation ofthe novel positive allosteric modulators of the mGlu₅ receptor, LSN2463359 andLSN2814617, and their effects on sleep architecture and operant responding in therat. Neuropharmacology 64:224–239.

Gobbo OL and O’Mara SM (2004) Post-treatment, but not pre-treatment, with theselective cyclooxygenase-2 inhibitor celecoxib markedly enhances functional re-covery from kainic acid-induced neurodegeneration. Neuroscience 125:317–327.

González JC, Egea J, Del Carmen Godino M, Fernandez-Gomez FJ, Sánchez-Prieto J,Gandía L, García AG, Jordán J, and Hernández-Guijo JM (2007) Neuroprotectantminocycline depresses glutamatergic neurotransmission and Ca(21) signalling inhippocampal neurons. Eur J Neurosci 26:2481–2495.

Greenamyre JT and Porter RH (1994) Anatomy and physiology of glutamate in theCNS. Neurology 44(11, Suppl 8)S7–S13.

144 Lin et al.

at ASPE

T Journals on A

ugust 19, 2020jpet.aspetjournals.org

Dow

nloaded from

Greenwood SM and Connolly CN (2007) Dendritic and mitochondrial changes duringglutamate excitotoxicity. Neuropharmacology 53:891–898.

Grynkiewicz G, Poenie M, and Tsien RY (1985) A new generation of Ca21 indicatorswith greatly improved fluorescence properties. J Biol Chem 260:3440–3450.

Herrero I and Sánchez-Prieto J (1996) cAMP-dependent facilitation of glutamaterelease by beta-adrenergic receptors in cerebrocortical nerve terminals. J BiolChem 271:30554–30560.

Hertting G and Seregi A (1989) Formation and function of eicosanoids in the centralnervous system. Ann N Y Acad Sci 559:84–99.

Hunter RL, Dragicevic N, Seifert K, Choi DY, Liu M, Kim HC, Cass WA, Sullivan PG,and Bing G (2007) Inflammation induces mitochondrial dysfunction and dopami-nergic neurodegeneration in the nigrostriatal system. J Neurochem 100:1375–1386.

Kaizaki A, Tien LT, Pang Y, Cai Z, Tanaka S, Numazawa S, Bhatt AJ, and Fan LW(2013) Celecoxib reduces brain dopaminergic neuronaldysfunction, and improvessensorimotor behavioral performance in neonatal rats exposed to systemic lipo-polysaccharide. J Neuroinflammation 10:45.

Li M, West JW, Numann R, Murphy BJ, Scheuer T, and Catterall WA (1993) Con-vergent regulation of sodium channels by protein kinase C and cAMP-dependentprotein kinase. Science 261:1439–1442.

Long P, Mercer A, Begum R, Stephens GJ, Sihra TS, and Jovanovic JN (2009) Nerveterminal GABAA receptors activate Ca21/calmodulin-dependent signaling to inhibitvoltage-gated Ca21 influx and glutamate release. J Biol Chem 284:8726–8737.

López-Villodres JA, De La Cruz JP, Muñoz-Marin J, Guerrero A, Reyes JJ,and González-Correa JA (2012) Cytoprotective effect of nonsteroidal antiin-flammatory drugs in rat brain slices subjected to reoxygenation after oxygen-glucose deprivation. Eur J Pharm Sci 45:624–631.

Lu CW, Lin TY, and Wang SJ (2010) Memantine depresses glutamate releasethrough inhibition of voltage-dependent Ca21 entry and protein kinase C in ratcerebral cortex nerve terminals: an NMDA receptor-independent mechanism.Neurochem Int 57:168–176.

Meldrum BS (2000) Glutamate as a neurotransmitter in the brain: review of physi-ology and pathology. J Nutr 130(4S, Suppl)1007S–1015S.

Meldrum B and Garthwaite J (1990) Excitatory amino acid neurotoxicity and neu-rodegenerative disease. Trends Pharmacol Sci 11:379–387.

Mi Y, Zhang X, Zhang F, Qi J, Gao H, Huang D, Li L, Zhang H, and Du X (2013) Therole of potassium channel activation in celecoxib-induced analgesic action. PLoSONE 8:e54797.

Millán C and Sánchez-Prieto J (2002) Differential coupling of N- and P/Q-type calciumchannels to glutamate exocytosis in the rat cerebral cortex. Neurosci Lett 330:29–32.

Minghetti L (2004) Cyclooxygenase-2 (COX-2) in inflammatory and degenerativebrain diseases. J Neuropathol Exp Neurol 63:901–910.

Minta A and Tsien RY (1989) Fluorescent indicators for cytosolic sodium. J BiolChem 264:19449–19457.

Murray HJ and O’Connor JJ (2003) A role for COX-2 and p38 mitogen activatedprotein kinase in long-term depression in the rat dentate gyrus in vitro. Neuro-pharmacology 44:374–380.

Nakayama M, Uchimura K, Zhu RL, Nagayama T, Rose ME, Stetler RA, Isakson PC,Chen J, and Graham SH (1998) Cyclooxygenase-2 inhibition prevents delayeddeath of CA1 hippocampal neurons following global ischemia. Proc Natl Acad SciUSA 95:10954–10959.

Nicholls DG (1998) Presynaptic modulation of glutamate release. Prog Brain Res116:15–22.

Nicholls DG and Sihra TS (1986) Synaptosomes possess an exocytotic pool of gluta-mate. Nature 321:772–773.

Nicholls DG, Sihra TS, and Sanchez-Prieto J (1987) Calcium-dependent and -independentrelease of glutamate from synaptosomes monitored by continuous fluorometry.J Neurochem 49:50–57.

Nishihara I, Minami T, Watanabe Y, Ito S, and Hayaishi O (1995) Prostaglandin E2stimulates glutamate release from synaptosomes of rat spinal cord. Neurosci Lett196:57–60.

Nogawa S, Zhang F, Ross ME, and Iadecola C (1997) Cyclo-oxygenase-2 gene ex-pression in neurons contributes to ischemic brain damage. J Neurosci 17:2746–2755.

Park SY, Kim TH, Kim HI, Shin YK, Lee CS, Park M, and Song JH (2007) Celecoxibinhibits Na1 currents in rat dorsal root ganglion neurons. Brain Res 1148:53–61.

Rehm H and Tempel BL (1991) Voltage-gated K1 channels of the mammalian brain.FASEB J 5:164–170.

Sang N, Zhang J, Marcheselli V, Bazan NG, and Chen C (2005) Postsynapticallysynthesized prostaglandin E2 (PGE2) modulates hippocampal synaptic trans-mission via a presynaptic PGE2 EP2 receptor. J Neurosci 25:9858–9870.

Schauwecker PE (2010) Neuroprotection by glutamate receptor antagonists againstseizure-induced excitotoxic cell death in the aging brain. Exp Neurol 224:207–218.

Schinder AF, Olson EC, Spitzer NC, and Montal M (1996) Mitochondrial dysfunctionis a primary event in glutamate neurotoxicity. J Neurosci 16:6125–6133.

Serrano GE, Lelutiu N, Rojas A, Cochi S, Shaw R, Makinson CD, Wang D, FitzGeraldGA, and Dingledine R (2011) Ablation of cyclooxygenase-2 in forebrain neurons isneuroprotective and dampens brain inflammation after status epilepticus. J Neu-rosci 31:14850–14860.

Sihra TS, Bogonez E, and Nicholls DG (1992) Localized Ca21 entry preferentiallyeffects protein dephosphorylation, phosphorylation, and glutamate release. J BiolChem 267:1983–1989.

Small GW, Siddarth P, Silverman DHS, Ercoli LM, Miller KJ, Lavretsky H, BookheimerSY, Huang SC, Barrio JR, and Phelps ME (2008) Cognitive and cerebral metaboliceffects of celecoxib versus placebo in people with age-related memory loss: randomizedcontrolled study. Am J Geriatr Psychiatry 16:999–1009.

Sugimoto Y and Narumiya S (2007) Prostaglandin E receptors. J Biol Chem 282:11613–11617.

Suto T, Severino AL, Eisenach JC, and Hayashida K (2014) Gabapentin increasesextracellular glutamatergic level in the locus coeruleus via astroglial glutamatetransporter-dependent mechanisms. Neuropharmacology 81:95–100.

Trepanier CH and Milgram NW (2010) Neuroinflammation in Alzheimer’s disease:are NSAIDs and selective COX-2 inhibitors the next line of therapy? J AlzheimersDis 21:1089–1099.

Urade Y and Hayaishi O (1999) Prostaglandin D2 and sleep regulation. BiochimBiophys Acta 1436:606–615.

Vázquez E and Sánchez-Prieto J (1997) Presynaptic modulation of glutamate releasetargets different calcium channels in rat cerebrocortical nerve terminals. Eur JNeurosci 9:2009–2018.

Wang SJ and Sihra TS (2004) Noncompetitive metabotropic glutamate5 receptorantagonist (E)-2-methyl-6-styryl-pyridine (SIB1893) depresses glutamate releasethrough inhibition of voltage-dependent Ca21 entry in rat cerebrocortical nerveterminals (synaptosomes). J Pharmacol Exp Ther 309:951–958.

Wang SJ, Wang KY, and Wang WC (2004) Mechanisms underlying the riluzole in-hibition of glutamate release from rat cerebral cortex nerve terminals (synapto-somes). Neuroscience 125:191–201.

Wu LG and Saggau P (1997) Presynaptic inhibition of elicited neurotransmitter re-lease. Trends Neurosci 20:204–212.

Yeganeh F, Nikbakht F, Bahmanpour S, Rastegar K, and Namavar R (2013) Neu-roprotective effects of NMDA and group I metabotropic glutamate receptorantagonists against neurodegeneration induced by homocysteine in rat hippo-campus: in vivo study. J Mol Neurosci 50:551–557.

Zhu P, Genc A, Zhang X, Zhang J, Bazan NG, and Chen C (2005) Heterogeneousexpression and regulation of hippocampal prostaglandin E2 receptors. J NeurosciRes 81:817–826.

Address correspondence to: Su-Jane Wang, Graduate Institute of BasicMedicine, Fu Jen Catholic University, 510, Chung-Cheng Rd., Hsin-Chuang,New Taipei, Taiwan 24205. E-mail: med0003@mail.fju.edu.tw

Inhibition of Glutamate Release by Celecoxib 145

at ASPE

T Journals on A

ugust 19, 2020jpet.aspetjournals.org

Dow

nloaded from