Mol Neurobiol DOI 10.1007/s12035-014-8815-5

Adenosine A1 Receptor-Dependent Antinociception Induced by Inosine in Mice: Pharmacological, Genetic and Biochemical Aspects Francisney Pinto Nascimento & Sérgio José Macedo-Júnior & Fabrício Alano Pamplona & Murilo Luiz-Cerutti & Marina Machado Córdova & Leandra Constantino & Carla Inês Tasca & Rafael Cypriano Dutra & João B. Calixto & Allison Reid & Jana Sawynok & Adair Roberto Soares Santos

Received: 24 April 2014 / Accepted: 11 July 2014 # Springer Science+Business Media New York 2014

Abstract Inosine is an endogenous nucleoside that has antiinflammatory and antinociceptive properties. Inosine is a metabolite of adenosine, and some of its actions suggest the involvement of adenosine A1 receptors (A1Rs). The purpose of this study was to better understand mechanisms of inosineinduced antinociception by investigating the role of A1Rs and F. P. Nascimento : S. J. Macedo-Júnior : M. Luiz-Cerutti : M. M. Córdova : A. R. S. Santos Laboratório de Neurobiologia da Dor e Inflamação, Departamento de Ciências Fisiológicas, Centro de Ciências Biológicas, Universidade Federal de Santa Catarina, Campus Universitário, Trindade, Florianópolis, SC 88040-900, Brazil F. P. Nascimento : S. J. Macedo-Júnior : R. C. Dutra : A. R. S. Santos (*) Programa de Pós-Graduação em Farmacologia, Departamento de Farmacologia, Centro de Ciências Biológicas, Universidade Federal de Santa Catarina, Campus Universitário, Trindade, Florianópolis, SC 88040-900, Brazil e-mail: [email protected] A. R. S. Santos e-mail: [email protected] F. A. Pamplona : L. Constantino : C. I. Tasca Departamento de Bioquímica, Centro de Ciências Biológicas, Universidade Federal de Santa Catarina, Campus Universitário, Trindade, Florianópolis, SC 88040-900, Brazil F. A. Pamplona Instituto D’Or de Pesquisa e Ensino (IDOR), Rio de Janeiro, RJ 22230-100, Brazil J. B. Calixto Centro de Inovação e Ensaios Pré-clínicos, Cachoeira do Bom Jesus, Florianópolis, SC 88056-000, Brazil A. Reid : J. Sawynok Department of Pharmacology, Dalhousie University, Halifax, NS, Canada

purine metabolism inhibitors. Inosine antinociception was evaluated using the formalin test in mice. An A1R-selective antagonist (DPCPX), A1R knockout mice (gene deletion) and mice with A1R reduced expression (antisense oligonucleotides) were used to assess the role of A1Rs in the antinociceptive action of inosine. Binding assays were performed to compare the affinity of inosine and adenosine for A1Rs. Finally, the role of adenosine and inosine breakdown was assessed using deoxycoformycin (DCF) and forodesine (FDS) as enzymatic inhibitors of adenosine deaminase and purine nucleoside phosphorylase, respectively. Inosine induced antinociception in the formalin test when given by systemic, spinal and peripheral routes. Systemically, inosine exhibited a potency similar to adenosine, and its effects were inhibited by DPCPX. Inosine did not induce antinociception in A1R knockout mice or in mice with reduced A1R expression. In binding studies, inosine bound to A1Rs with an affinity similar to adenosine. DCF had no effect on inosine actions. FDS augmented the antinociceptive effect of a low systemic dose of inosine and, at a higher dose, induced antinociception by itself. Collectively, these data indicate that inosine is an agonist for A1Rs with antinociceptive properties and a potency similar to adenosine and can be considered another endogenous ligand for this receptor. Keywords Inosine . Adenosine . Adenosine A1 receptor . Antinociception . Deoxycoformycin . Forodesine

Introduction Inosine is an endogenous nucleoside belonging to the purinergic family and induces neuroprotective and antiinflammatory effects [1]. It is a product of ATP metabolism via an enzymatic cascade including ATPase and adenosine

Mol Neurobiol

deaminase (ADA) enzymes; adenosine is a proximal intermediary, and further metabolism via purine nucleoside phosphorylase (PNP) produces inosine (Fig. 1) [1–3]. Adenosine has been known to exhibit effects via adenosine receptors for some time [4], but the biological activity of inosine has been appreciated only more recently [1, 5]. There are four adenosine receptor subtypes, A1 A2A, A2B and A3 [4, 6, 7]; the adenosine A 1 receptor (A1R) is the most widely distributed and underlies adenosine antinociceptive effects [2, 3, 8–10]. When activated, this G protein-coupled receptor inhibits adenylyl cyclase, opens potassium channels and blocks calcium channels; additionally, A1R activation through phospholipase C can increase both intracellular calcium and inositol triphosphate [4, 6, 7]. Several studies demonstrate a variety of antiinflammatory effects of inosine and report in vivo and in vitro actions in various models [11–15]. A2A and A3Rs generally have been implicated in such antiinflammatory effects of inosine [16–19]. Some studies also implicate A1Rs in biological effects of inosine [20]. Antinociceptive effects of inosine have been reported in both acute and chronic pain models [5]. Pharmacological

evidence has suggested that antinociceptive effects of inosine involve activation of adenosine receptors (potentially A1Rs), the protein kinase C pathway [5], as well as pertussis toxin-sensitive G-proteins, K + channels and voltage-gated Ca2+ channels [21]. The goal of the present study was to further characterize the mechanism of action of inosine in producing antinociception by using pharmacological, genetic and biochemical approaches. In pharmacological studies, we investigated the antinociceptive effects of inosine and adenosine in the formalin test, a model of ongoing pain involving elements of peripheral and central sensitization where A1Rs are known to modulate nociception, and examined effects of a selective A1R antagonist and different routes of administration. We also used inhibitors of ADA (deoxycoformycin, or DCF) and PNP (forodesine, or FDS) to manipulate endogenous levels of adenosine and inosine and evaluated the consequence of these drugs on adenosine/inosine antinociception. Genetic studies used knockout (gene deletion) and knockdown (antisense oligonucleotide) approaches. Finally, biochemical studies characterized receptor-binding properties of inosine at A1Rs.

Fig. 1 Purinergic metabolism. Schematic diagram summarizing purinergic metabolism in the extracellular and intracellular space. ATP is converted to AMP by ATPase. AMP under the action of 5′ecto-nucleotidase (or 5′endo-nucleotidase inside the cell) is metabolized to adenosine. In the other direction, adenosine phosphorylated by AK becomes AMP, which can be converted to ATP by adenylate kinase. Adenosine activates four metabotropic receptors, A1, A2A, A2B and A3. Adenosine can be transported in two directions (inside/outside cell) by ENTs. ADA converts

adenosine to inosine, which is then converted to hypoxanthine by PNP. DCF is an ADA inhibitor that increases adenosine levels, while FDS is a PNP inhibitor that increases inosine levels. ADA adenosine deaminase, A1, A2A, A2B and A3 adenosine receptors, AK adenosine kinase, AMP adenosine monophosphate, ATP adenosine triphosphate, DCF deoxycoformycin, ENT equilibrative nucleoside transporter, FDS forodesine, HXT hypoxanthine, PNP purine nucleoside phosphorylase

Mol Neurobiol

Methods Animals Male/female C57BL/6 mice and male Swiss mice (Mus musculus) from Universidade Federal de Santa Catarina (Florianópolis, Brazil) were used for experiments involving mice with A1R reduced expression due to oligonucleotide treatment (immunohistochemistry analysis and behaviour). All other experiments were performed in Canada with male/female C57BL/6 mice, wild-type (A1R+/+) and knockout mice (A1R-/-) from Charles River (Quebec, Canada) or raised in-house. Mice (between 70 and 90 days old) were housed at 22±2 °C under a 12-h light/12-h dark cycle (lights on at 6.00 am) and had access to food and water ad libitum. They were acclimatized to the laboratory for at least 1 h before testing and were used only once. To perform the binding assays, Sprague-Dawley rats from Universidade Federal de Santa Catarina were used. Experiments were performed according to protocols approved by the Committee for Animal Research of the Universidade Federal de Santa Catarina (protocol number PP00484) and by the Dalhousie University Committee on Laboratory Animals (protocol number 11-021). All experiments were carried out in accordance with current guidelines for the care of laboratory animals and the ethical guidelines for investigations of experimental pain in conscious animals [22]. The number of animals and intensities of noxious stimuli used were the minimum necessary to demonstrate consistent effects. Drug Administration Drugs were administered by intraperitoneal (i.p.), intraplantar (i.pl.) or intrathecal (i.t.) routes. Drugs administered by the i.p. or i.pl. routes were given in a volume of 5 ml/kg or 20 μl, respectively. Mice were briefly anaesthetized with isoflurane just before the i.t injections of drugs or antisense oligonucleotide (5 μl using a 30-gauge needle) delivered by acute lumbar puncture between the L5 and L6 vertebrae and a characteristic tail-flick indicated successful penetration of the spinal compartment. Within 5 min of induction, mice were fully recovered from anaesthesia. The i.t. injection methodology was initially validated using injections of dye and post-mortem inspection of spread of dye. Formalin Test The formalin test, a model of ongoing and inflammatory pain, was used to assess nociceptive behaviour. Briefly, each mouse received a subcutaneous i.pl. injection of 2 % formalin (20 μl, in saline) into the plantar surface of the hind paw [23, 24]. The number of flinches, defined as elevations of the hind paw and/ or episodes of rapid shaking, was counted for 60 min. Two

mice were monitored simultaneously in alternating 2-min bins in separate adjacent plexiglass observation chambers. The biphasic nature of formalin-evoked pain behaviours was analyzed as the cumulative number of flinches during phase 1 (0–8 min) and phase 2 (12–60 min). Phase 1 flinching behaviour was generally unaffected by inosine or adenosine [5]. As such, only data from phase 2 are shown. Dose Response Curve of Systemic Inosine and Adenosine in the Formalin Test In the first block of experiments, inosine and adenosine were given by systemic (i.p.) injection at doses between 1 and 100 mg/kg. Mice in the control group received only vehicle (dimethyl sulfoxide (DMSO), 5 % in saline) at a dose of 5 ml/ kg i.p. Animals were injected with formalin 20 min later. Involvement of A1Rs in Inosine-Induced Antinociception The selective A1R antagonist DPCPX and A1R knockout animals were used to investigate the involvement of A1Rs in antinociception induced by inosine. A dose of 10 mg/kg of inosine was used, as it induced the same level of antinociception as 100 mg/kg. To assess A1R participation at systemic and spinal levels, animals were divided into three groups and treated with vehicle (5 ml/kg i.p.), DPCPX 10 nmol/5 μl i.t. or DPCPX 0.1 mg/kg i.p. Following this, they received vehicle (5 ml/kg i.p.) or inosine (10 mg/kg i.p.) and, after a further 20 min, underwent the formalin test. In other sets of experiments, wild-type (A1R+/+) and knockout mice (A1R−/−) were treated systemically (i.p.) with vehicle (5 ml/kg) or inosine (10 mg/kg), or spinally (i.t.) with vehicle (5 μl) or inosine (10 μg/5 μl). Formalin injection was given 20 or 5 min later to i.p. and i.t. groups, respectively. Genotyping of animals was performed using polymerase chain reaction analysis. Peripheral Antinociceptive Effect of Inosine To investigate the local effect of inosine, this nucleoside was co-administered with 2 % formalin under the plantar surface of the hind paw (i.pl. injection). Based on previous studies in our laboratory (data not shown), doses of 10 and 20 μg/20 μl of inosine were injected. To assess the involvement of A1Rs in the local effect of inosine, a dose of 20 μg was chosen because it induced antinociception locally. DPCPX doses were based on previous work [5] and/or pilot experiments. These animals were pre-treated with DPCPX (5/20 μl i.pl.) and 5 min later were injected with inosine+formalin (20 μg+ 2 %/20 μl i.pl.), and nociceptive behaviour was immediately assessed. Finally, we used A1R knockout animals to evaluate the peripheral participation of A1Rs in inosine antinociception. Wild-type (A1R+/+) animals received vehicle

Mol Neurobiol

(20 μl i.pl.) or inosine (20 μg i.pl.), and the knockout animals (A1R−/−) received inosine (20 μg i.pl.). Antinociceptive Effect of Inosine in Mice with Reduced A1R Expression C57Bl/6 mice received two daily injections of A1R antisense oligonucleotide over 5 days [25]. The control group received injections of antisense oligodeoxynucleotide-mismatch (AS-ODN-mismatch; sequence 5′-AAGTTGGCGGGG AAGCAGGG-3′, 1 μM/5 μl i.t., designed by pDRAW32, AcaClone Software; the test group received injections of antisense oligodeoxynucleotide to A1R (AS-ODN-A1R; 5′GTCCTTGCTCTCCCTTCCTC-3′, 1 μM/5 μl i.t., designed by pDRAW32, AcaClone Software). One hour after the final injection of AS-ODN, each group received inosine (10 mg/kg i.p.) or vehicle (5 ml/kg i.p.) and 20 min later underwent the formalin test. Following the formalin test, animals were anaesthetized (mixture of ketamine 10 mg/kg and xylazine 30 mg/kg i.m.) and perfused with 4 % paraformaldehyde. Spinal cords were harvested for the immunohistochemical detection of A1R expression. Immunohistochemistry of A1Rs in the Spinal Cord Spinal cords were removed and kept in 4 % paraformaldehyde for 24 h. Transverse sections of the lumbar spinal cord were embedded in paraffin and cut using a microtome. The staining procedure has been previously reported [26]. Protein immunodetection was performed with a primary antibody against the A1R (1:500) and a secondary biotinylated antibody anti-IgG rabbit. Immunodetection was completed with a chromogen solution of 0.03 % 3,3′-diaminobenzidine (3,3′,4,4′-tetraaminobiphenyltetrahydrochloride) and 0.3 % hydrogen peroxide. Images were obtained using a Sight DS5 M-L1 digital camera connected to an Eclipse 50i light microscope (both from Nikon, Melville, NY, USA). Settings for image acquisition were identical for control and experimental tissue. Four ocular fields per section (eight to ten mice per group) were captured, and a threshold optical density that best discriminated staining from the background was obtained using the NIH ImageJ 1.36b imaging software (NIH, Bethesda, MD, USA). The total pixel intensity was determined, and data were expressed as optical density, using a counting grid at×200 and×400 magnification [26]. Competitive Binding Assay of [3H]-DPCPX Versus Inosine and Adenosine to A1Rs in the Brain and Spinal Cord Membranes Competitive binding assays using the radiolabeled A1Rselective antagonist [3H]-DPCPX were used to detect binding of inosine and adenosine to A1Rs, according to methodology

previously published [27]. Briefly, rat whole brain and mouse spinal cord synaptosome-enriched membranes (P2 fraction) were incubated with adenosine or inosine in the presence of the A1R antagonist [3H]-DPCPX (5 nM) for 40 min at 37 °C. Non-specific binding was determined with cold DPCPX and was never higher than 20 % of the total binding. The number of samples was four per group, run in triplicate. These experiments allow the detection of competitive specific binding at one site. Results were normalized to total specific binding and expressed as a percentage of inhibition. Ki values were calculated by the non-linear fitting of the one-site competitive specific binding assay equation using GraphPad software (Statsoft), using Kd value of 1 nM for [3H]-DPCPX [28]. Influence of Deoxycoformicin (DCF) on Inosine- and Adenosine-Induced Antinociception Administration of DCF, an ADA inhibitor, was used to evaluate whether this inhibitor could alter the antinociceptive effect of adenosine or inosine. In these experiments, animals received vehicle (5 ml/kg i.p.) or DCF (50 mg/kg i.p.) followed by vehicle (5 ml/kg i.p.), inosine (10 mg/kg i.p.) or adenosine (30 mg/kg i.p.) 20 min later. After an additional 20 min, they underwent the formalin test. To assess ADA activity, a second group of animals (n=5 per group) was administered vehicle (5 ml/kg i.p.) or DCF (50 mg/kg i.p.), and the blood was collected after 30 min. The serum was processed using a spectrophotometric technique [29]. ADA activity was estimated by standard curve interpolation, and results were expressed as units per litre (UI/l). Influence of FDS on Inosine-Induced Antinociception and the Role of A1Rs To assess whether blocking the breakdown of inosine altered antinociception, we inhibited PNP, the enzyme that converts inosine to hypoxanthine, using FDS. Mice were pre-treated with vehicle (5 ml/kg i.p.) or FDS (0.1 mg/kg i.p.) and 20 min later received vehicle (5 ml/kg i.p.) or inosine (3 or 10 mg/kg i.p.). After an additional 20 min, they underwent the formalin test. Antinociception Induced by FDS and Its Dependence on A1Rs To assess how the antinociceptive effect of FDS is related to A1Rs, mice were pre-treated with vehicle (5 ml/kg i.p.) or DPCPX (0.1 mg/kg i.p.). After 20 min, those that had been pre-treated with vehicle received vehicle (5 ml/kg i.p.) or FDS (0.1 or 1 mg/kg i.p.); those pre-treated with DPCPX received FDS (1 mg/kg i.p.). Twenty minutes later, all animals underwent the formalin test.

Mol Neurobiol

Drugs Formalin and DMSO were purchased from Merck, Darmstadt, Germany; adenosine, inosine, ketamine hydrochloride, xylazine hydrochloride and hydrogen peroxide were purchased from Sigma-Aldrich, USA; DCF and DPCPX were purchased from Tocris Bioscience; FDS was purchased from BioCryst, North Caroline, USA; and diaminobenzidine was purchased from Molecular Brasil, São Paulo, Brazil. Adenosine, inosine, DCF, DPCPX and FDS were dissolved in saline with 5 % DMSO. The following agents were sourced as indicated: isoflurane (Baxter Corporation, Canada), A1R primary antibody (1:500; Abcam®, Cambridge, MA, USA), secondary biotinylated antibody anti-IgG rabbit (DakoCytomation, CA, USA), oligonucleotide A1R antisense (Prodimol, São Paulo, Brazil), and paraformaldehyde (Fisher Scientific, USA). Statistical Analysis The results are presented as mean± S.E.M. Data were analyzed using a one-way ANOVA followed by Bonferroni’s post hoc test or Student’s t test. P values less than 0.05 were considered indicative of significance. The statistical software used was Prism 4.0 (GraphPad Software Inc., San Diego, CA).

percentage inhibition of flinching was 44±10 % (P=0.007) (Fig. 2a). Adenosine also reduced flinching induced by formalin, with 44± 3 % inhibition at 10 mg/kg (P = 0.009) (Fig. 2b). Adenosine 100 mg/kg did not induce antinociception and differed from adenosine 10 mg/kg (t test, P=0.011) (Fig. 2b). The efficacy and potency of inosine in the formalin test were similar to those of adenosine. The Systemic Effect of Inosine Depends on Central A1Rs The A1R-selective antagonist DPCPX alone, given spinally or systemically, did not alter the nociceptive response to formalin but a pre-emptive i.t (10 nmol) or i.p (0.1 mg/kg) injection of DPCPX that reversed the antinociceptive effect of systemic inosine (P