Neuropharmacology 89 (2015) 64e76

Contents lists available at ScienceDirect

Neuropharmacology journal homepage: www.elsevier.com/locate/neuropharm

Amplification of neuromuscular transmission by methylprednisolone involves activation of presynaptic facilitatory adenosine A2A receptors and redistribution of synaptic vesicles L. Oliveira a, b, A.C. Costa a, b, J.B. Noronha-Matos a, b, I. Silva a, b, W.L.G. Cavalcante c,  teo a, b, A.P. Corrado d, C.A. Dal Belo e, C.R. Ambiel f, W. Alves-do-Prado g, M.A. Timo  a, b, * P. Correia-de-Sa rio de Farmacologia e Neurobiologia/UMIB, Universidade do Porto, Portugal Laborato Center for Drug Discovery and Innovative Medicines (MedInUP), Universidade do Porto, Portugal c ~o Paulo (UNESP), Botucatu, Sa ~o Paulo, Brazil Instituto de Bioci^ encias, Universidade Estadual de Sa d ~o Preto, Universidade de Sa ~o Paulo, Sa ~o Gabriel, Rio Grande do Sul, Brazil Departamento de Farmacologia, Faculdade de Medicina de Ribeira e ~o Gabriel, Rio Grande do Sul, Brazil Universidade Federal do Pampa, Sa f , Parana , Brazil gicas, Universidade Estadual de Maringa Departamento de Ci^ encias Fisiolo g , Parana , Brazil Departamento de Farmacologia e Terap^ eutica, Universidade Estadual de Maringa a

b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 18 June 2014 Received in revised form 19 August 2014 Accepted 2 September 2014 Available online 16 September 2014

The mechanisms underlying improvement of neuromuscular transmission deficits by glucocorticoids are still a matter of debate despite these compounds have been used for decades in the treatment of autoimmune myasthenic syndromes. Besides their immunosuppressive action, corticosteroids may directly facilitate transmitter release during high-frequency motor nerve activity. This effect coincides with the predominant adenosine A2A receptor tonus, which coordinates the interplay with other receptors (e.g. muscarinic) on motor nerve endings to sustain acetylcholine (ACh) release that is required to overcome tetanic neuromuscular depression in myasthenics. Using myographic recordings, measurements of evoked [3H]ACh release and real-time video microscopy with the FM4-64 fluorescent dye, results show that tonic activation of facilitatory A2A receptors by endogenous adenosine accumulated during 50 Hz bursts delivered to the rat phrenic nerve is essential for methylprednisolone (0.3 mM)induced transmitter release facilitation, because its effect was prevented by the A2A receptor antagonist, ZM 241385 (10 nM). Concurrent activation of the positive feedback loop operated by pirenzepinesensitive muscarinic M1 autoreceptors may also play a role, whereas the corticosteroid action is restrained by the activation of co-expressed inhibitory M2 and A1 receptors blocked by methoctramine (0.1 mM) and DPCPX (2.5 nM), respectively. Inhibition of FM4-64 loading (endocytosis) by methylprednisolone following a brief tetanic stimulus (50 Hz for 5 s) suggests that it may negatively modulate synaptic vesicle turnover, thus increasing the release probability of newly recycled vesicles. Interestingly, bulk endocytosis was rehabilitated when methylprednisolone was co-applied with ZM241385. Data suggest that amplification of neuromuscular transmission by methylprednisolone may involve activation of presynaptic facilitatory adenosine A2A receptors by endogenous adenosine leading to synaptic vesicle redistribution.

Keywords: Neuromuscular junction Synaptic vesicle recycling Acetylcholine release Glucocorticoids Adenosine receptors Muscarinic receptors

© 2014 Elsevier Ltd. All rights reserved.

Abbreviations: ACh, acetylcholine; ADA, adenosine deaminase; DPCPX, 1,3-dipropyl-8-cyclopentyl xanthine; Dyngo-4a, 3-Hydroxy-N'-[(2,4,5-trihydroxyphenyl)methylidene] naphthalene-2-carbohydrazide; FM4-64, N-(3-triethylammoniumpropyl)-4-(6-(4-(diethylamino) phenyl)hexatrienyl) pyridinium dibromide; HC, hemicholinium-3; b,g-imidoATP, adenosine 50 -(b,g-imido)triphosphate; H-89, N-[2-(p-bromocinnamylamino)ethyl]-5-isoquinolinesulfonamide; MP, methylprednisolone; PZ, pirenzepine; D-Tc, D-tubocurarine; ZM241385, 4-(-2-[7-amino-2-{2-furyl}{1,2,4}triazolo{2,3-a}{1,3,5}triazin-5-yl-amino]ethyl)phenol. rio de Farmacologia e Neurobiologia/UMIB and Center for Drug Discovery and Innovative Medicines (MedInUP), Instituto de Cie ^ncias * Corresponding author. Laborato dicas de Abel Salazar (ICBAS), Universidade do Porto (UP), R. Jorge Viterbo Ferreira, 228, 4050-313 Porto, Portugal. Tel.: þ351 220428212; fax: þ351 220428090. Biome ). E-mail addresses: [email protected], [email protected] (P. Correia-de-Sa http://dx.doi.org/10.1016/j.neuropharm.2014.09.004 0028-3908/© 2014 Elsevier Ltd. All rights reserved.

L. Oliveira et al. / Neuropharmacology 89 (2015) 64e76

1. Introduction Corticosteroids are first line medications in the treatment of autoimmune neuromuscular transmission disorders, such as Myasthenia gravis, LamberteEaton myasthenic syndrome and neuromyotonia (reviewed in Skeie et al., 2006; Jani-Acsadi and Lisak, 2007; Vincent, 2010; Kumar and Kaminski, 2011). Besides its empirical use as immunosuppressive agents, corticosteroids may improve neuromuscular transmission deficits by acting directly on skeletal motor endplates. The mechanism(s) of action of corticosteroids at the neuromuscular junction is still a matter of debate. Anti-inflammatory and immunosuppressive effects of glucocorticoids may be mediated by cytosolic receptors (genomic mechanism) and/or by fast pathways involving non-specific interactions with cellular membranes (non-genomic mechanism) (Stahn and €sel and Wehling, 2003). While glucocorticoids Buttgereit, 2008; Lo produce no changes on skeletal muscle twitches triggered by lowfrequency nerve stimulation, these drugs significantly enhance post-tetanic potentiation and attenuate neuromuscular block produced by anti-nicotinic muscle relaxants in both slow and fast sz skeletal muscles (Baker et al., 1977; Dal Belo et al., 2002; Solte et al., 2008). Evidences for a non-genomic presynaptic action of corticosteroids at motor endplates (Makara and Haller, 2001) were corroborated by steroid-induced increases in the frequency and amplitude of miniature endplate potentials (MEPPs), without any effect on muscle resting membrane potential (Dalkara and Onur, 1987). The enhancing effect of corticosteroids on neuromuscular transmission seems to be unrelated to the blockade of phospholipase A2 (Sen et al., 1976) and to inhibition of pro-inflammatory prostanoids (Smyth et al., 2006). Prednisolone and dexamethasone antagonize the inhibitory action of hemicholinium-3 both on the rate of choline uptake and the incorporation of choline into acetylcholine (ACh) in the rat diaphragm (Veldsema-Currie et al., 1976; van Marle et al., 1986). All these properties may contribute to the usefulness of corticosteroids in autoimmune neuromuscular transmission disorders. Therefore clarification of the underlying mechanism(s) responsible for corticosteroid-induced facilitation of transmitter release may be clinically relevant. The neuromuscular transmission is controlled by a sophisticated interaction between the activation of presynaptic inhibitory and facilitatory receptors, which is dependent on the neuronal firing pattern, including the number, frequency and duration of each stimulus (Wessler, 1989). Adenosine, released as such or buildup from ATP catabolism during neuronal firing, plays a key role in adjusting the modulatory pattern of neuromuscular transmission  et al., 1996). Fineto the stimulation conditions (Correia-de-Sa tuning synaptic control by adenosine emerges via subtle modifi~o and cations at the presynaptic interreceptor dynamics (Sebastia Ribeiro, 2000) involving intracellular second messengers, such as cyclic AMP (Correia-de-S a and Ribeiro, 1994) and calcium (Correia et al., 2000). Our group provided evidence suggesting that de-Sa adenosine generated from released ATP activates preferentially facilitatory A2A receptors, both at rat hippocampal and neuromus et al., 1996; Cunha et al., 1996). Tonic cular synapses (Correia-de-Sa activation of adenosine A2A receptors operates a coordinated shift in Ca2þ channel dynamics leading to facilitation of ACh release, from the “prevalent” Cav2.1 (P-type) to the “facilitatory” Cav1 (L et al., 2000), which may contribute to type) channel (Correia-de-Sa overcome tetanic depression during high-frequency neuronal firing (Oliveira et al., 2004). This adaptive mechanism seems to be severely affected in myasthenic motor endplates, but we demonstrated that it may be rehabilitated by A2A receptor activation (Noronha-Matos et al., 2011). In addition to its stimulatory action on skeletal muscle fibres, ACh may also acts presynaptically to regulate its own release at

65

mammalia neuromuscular junctions. Neuronal nicotinic receptors expressing a3b2 subunits (Faria et al., 2003) mediate a short term positive feedback mechanism, which is terminated by rapid autodesensitization (Wessler et al., 1986; Colquhon et al., 1989). Interestingly, nicotinic-induced transmitter overflow is regulated by  teo et al., synchronous activation of adenosine A2A receptors (Timo 2003). Transmitter release is also controlled by presynaptic muscarinic receptors via the activation of facilitatory M1 and inhibitory M2 receptors (see e.g. Wessler, 1989; Oliveira et al., 2002). The synaptic levels of adenosine accumulated during neuronal firing fine-tuning modulate the muscarinic tonus (Oliveira et al., 2002). Muscarinic M1 autofacilitation predominates during low stimulation frequencies (5 Hz) when small amounts of extracellular adenosine activate preferentially inhibitory A1 receptors (Correia et al., 1996). During high frequency (50 Hz) stimulation de-Sa bursts, there is a shift from muscarinic M1 facilitation towards M2mediated inhibition, which results mainly from the activation of A2A receptors by high adenosine amounts formed from the catabolism of released adenine nucleotides. Thus, impairment of adenosine accumulation in myasthenic motor endplates due to reduced skeletal muscle activity and/or to low amounts of released adenine nucleotides (Noronha-Matos et al., 2011), might contribute to tetanic failure of neuromuscular transmission reflecting a loss of the nicotinic and muscarinic autoreceptor control in myasthenics. Considering that dominant neuromodulatory actions of corticosteroids are observed during high-frequency neuronal activity and that this situation coincides with the predominant facilitatory tonus of adenosine (via A2A receptors), we tested whether these receptors are involved in the amplification effect of methylprednisolone on transmitter exocytosis triggered by tetanic stimulation of the phrenic motor nerve using myographic recordings, radioisotope neurochemistry and video microscopy with the FM4-64 fluorescent dye (see e.g. Noronha-Matos et al., 2011; Correia-de et al., 2013). The presynaptic involvement of muscarinic (M1 Sa and M2) and adenosine A1 receptors on corticosteroid-induced facilitation of transmitter exocytosis was also investigated. 2. Material and methods 2.1. Animals Rats (Wistar, 150e250 g) of either sex (Charles River, Barcelona, Spain) were kept at a constant temperature (21  C) and a regular light (06.30e19.30 h)edark (19.30e06.30 h) cycle for at least ten days prior to the experiments, with food and water ad libitum. The animals were killed after stunning followed by exsanguination. Animal handling and experiments were in accordance with the guidelines prepared by the Committee on Care and Use of Laboratory Animal Resources (National Research Council, USA) and followed the European Communities Council Directive (86/609/EEC). All animal studies comply with the ARRIVE guidelines. 2.2. Experiments to measure the release of [3H]ACh and ATP from phrenic nervehemidiaphragm preparations The experiments were performed using either the left or the right phrenic nerve-hemidiaphragm preparations (4e6 mm width). The procedures used for labelling the preparations and measuring evoked [3H]ACh release were as described  and Ribeiro, 1996; previously (Correia-de-S a et al., 1991, 2013; Correia-de-Sa Noronha-Matos et al., 2011) with minor modifications. Briefly, the preparations were mounted in 3-ml capacity Perspex chambers and superfused with gassed (95% O2 and 5% CO2) Tyrode's solution (pH 7.4) containing (mM): NaCl 137, KCl 2.7, CaCl2 1.8, MgCl2 1, NaH2PO4 0.4, NaHCO3 11.9, glucose 11.2 and choline 0.001, at 37  C. Nerve terminals were labelled for 40 min with 1 mM [3H]choline (specific activity 2.5 mCi/nmol) under electrical stimulation at a frequency of 1 Hz (0.04 ms duration, 8 mA). The phrenic nerve was stimulated with a glass-platinum suction electrode placed near the first division branch of the nerve trunk, to avoid direct contact with muscle fibres. Washout of the preparations was performed for 60 min, by superfusion (15 ml/min) with Tyrode's solution supplemented with the choline uptake inhibitor, hemicholinium-3 (10 mM). Tritium outflow was measured in a Perkin Elmer TriCarb 2900TR scintillation spectrometer (% tritium efficiency: 58 ± 2%), after appropriate background subtraction, using aliquots of 2-ml bath samples collected automatically every 3 min. In some experiments, the ATP content of the samples was evaluated with the luciferineluciferase ATP bioluminescence assay kit HS II (Roche

66

L. Oliveira et al. / Neuropharmacology 89 (2015) 64e76

Applied Science, Indianapolis, Indiana). Luminescence was determined using a multi detection microplate reader (Synergy HT, BioTek Instruments). [3H]ACh release was evoked by two periods of electrical stimulation of the phrenic nerve, starting at 12th (S1) and 39th (S2) minutes after the end of washout (zero time). Supramaximal-intensity rectangular pulses (0.04 ms duration, 8 mA) were delivered at 5 or 50 Hz frequency. A series of five bursts of 150 pulses applied with a 20-s interburst interval were used when stimulation frequency was 50 Hz (tetanic stimulation) (Correia-de-S a et al., 1996; Oliveira et al., 2004; Noronha-Matos et al., 2011). Pulses were generated by a Grass S48 (Quincy, MA, USA) stimulator coupled to a stimulus isolation unit (Grass SIU5, USA) operating in current constant mode. Electrical stimulation of the phrenic nerve increased the release of [3H]ACh in  et al., 2000), while the a Ca2þ- and tetradotoxin-sensitive manner (Correia-de-Sa output of [3H]choline remained unchanged (Wessler and Kilbinger, 1986). Therefore, evoked [3H]ACh release was calculated by subtracting the basal tritium outflow from  et al., 1991). the total tritium outflow during the stimulation period (cf. Correia-de-Sa Likewise, stimulation-evoked release of ATP was calculated by subtracting the basal release, measured in three samples collected before stimulation, from the total release of the nucleotide determined after stimulus application. Test drugs were added 15 min before S2 and were present up to the end of the experiments. The change in the ratio between the evoked [3H]ACh release during the two stimulation periods (S2/S1) relative to that observed in control situations (in the absence of test drugs) was taken as a measure of drug effects. 2.3. Real-time video-microscopy using the FM4-64 fluorescent dye as a measure of transmitter exocytosis To follow real-time transmitter exocytosis from stimulated phrenic motor nerve terminals, the preparations were mounted on the stage of an upright epifluorescence microscope (Zeiss Axiophot, Oberkochen, Germany) and thereafter incubated as for the release of [3H]ACh, except no [3H]choline was added to the Tyrode's solution. After a 30-min equilibration period, the preparations were incubated with a-bungarotoxin (4 mM during 15e20 min, an irreversible blocker of muscle-type nicotinic receptors) to prevent nerve-evoked muscle fibre contractions, which would otherwise complicate the analysis of fluorescence signals. The procedures used to load synaptic vesicles with the membrane-selective FM4-64 fluorescent dye were as previously described (Perissinotti et al., 2008), and used with minor mod et al., 2013). After a 10ifications (see e.g. Noronha-Matos et al., 2011; Correia-de-Sa min incubation period with FM4-64 (5 mM) made up in Tyrode solution, loading of synaptic vesicles was achieved by stimulating the phrenic nerve trunk with 250 pulses of supramaximal intensity (0.04 ms duration, 8 mA) applied at 50 Hz frequency. Following a 10-min period with the dye at rest, the FM4-64 fluorescence was washed vigorously during 30 min (Perissinotti et al., 2008). Upon vigorous washout of the dye (not taken up by nerve terminals) from the incubation fluid, dissipation of FM4-64 fluorescence hotspots during electrical stimulation (e.g. 50 Hz bursts, see above) can be taken as a measure of synaptic vesicle exocytosis (cf. Betz et al., 1992; Perissinotti et al., 2008; Noronha-Matos et al., 2011). Test drugs were added to the superfusion fluid via an automatic perfusion system (ValveLink8.2; Digitimer, Welwyn Garden City, UK) connected to a fast solution heating device (TC-344B; Harvard Apparatus, March-Hugstetten, Germany).

Fluorescence images (excitation filter: BP 546/12 nm; emission filter: LP 590 nm) were acquired using a 63x/0.90 n.a. water-immersion objective lens (Achroplan, Zeiss, Germany). Images were acquired in the real-time mode with a high-resolution cooled CCD camera (CoolSnap HQ, Roper Scientific Photometrics, Tucson, AZ, USA). Absolute fluorescence measurements were converted to a percentage of the maximum fluorescence detected after staining, by the following equation: %F(t) ¼ 100x [F(t)  FNV]/[FMAX  FNV], where F(t) is the absolute fluorescence at time t, FMAX is the absolute fluorescence after maximum loading, and FNV is the non-vesicular fluorescence background (i.e. fluorescence remaining at the end of the stimulation) (Betz et al., 1992; Noronha-Matos et al., 2011; Perissinotti et al., 2008). 2.4. Myographic recordings Phrenic nerve-diaphragm muscle preparations were isolated and mounted according to Bülbring (1946). Each preparation was immersed in a 20 ml chamber containing gassed (95% O2 and 5% CO2) Krebs's solution containing (mM): NaCl 110, KCl 4.7, CaCl2 3, MgCl2 1.3, NaHCO3 25, KH2PO4 1, and glucose 11.1, at 37  C. Hemidiaphragms were connected to an isometric force transducer (Grass FT 03, Grass Instruments Division, West Warwick, RI, USA). Muscle contraction responses were continuously recorded at a resting tension of 50 mN with a PowerLab data acquisition system (Chart Software; AD Instruments, Castle Hill, NSW, Australia). The phrenic nerve was stimulated through a bipolar platinum electrode (supramaximal rectangular pulses, 0.05 ms). The preparations were indirectly stimulated at 0.2 Hz for 15 min (equilibration protocol). From this time onwards, the phrenic nerve was stimulated four times with 50 Hz tetanic trains of 10 s duration, which were applied at 15 min intervals. This interval was selected to avoid the influence of the previous tetanic stimulation on tetanic facilitation or fade. The tension at the beginning (A) and at the end (B) of each tetanus was recorded, and the ratio (R) B/A was calculated (Fig. 1A). Values of A, B and R were obtained for the second, third, and fourth tetanic train delivered 15, 30, and 45 min after the first tetanic stimulus (control), respectively. Test drugs were added after the first tetanic stimulus and were present up to the end of the experiments. 2.5. Choline uptake by RBE4 and CACO-2 cells in culture Rat brain endothelial (RBE4) and human epithelial colorectal adenocarcinoma (CACO-2) cells were routinely cultured in 75 cm2 flasks using Dulbecco's Modified Eagle's Medium (DMEM) with 4.5 g/L glucose and GlutMAX™, nonessential amino acids (NEAA), fetal bovine serum (FBS), 0.25% trypsin/1 mM EDTA, antibiotic (10,000 U/mL penicillin, 10,000 mg/mL streptomycin), fungizone (250 mg/mL amphotericin B) and human transferrin (4 mg/mL) purchased from Gibco Laboratories (Lenexa, KS). Cells were maintained in a 5% CO2 þ 95% air atmosphere at 37  C and the medium was changed every 2 days. Cultures passage was done weekly by trypsinization (0.25% trypsin/1 mM EDTA). Cells used for the experiments were taken between the 58th and 64th passages. RBE4 and CACO-2 cells were seeded onto 6-well plates at a density of 60,000 cells/cm2. After reaching confluence, the cells were washed twice with uptake buffer (137 mmol/L NaCl, 5.4 mmol/L KCl, 2.8 mmol/L CaCl2, 1.2 mmol/L MgCl2, 10 mmol/L HEPES, pH 7.4) and incubated, at 37  C for 20 min, with

Fig. 1. Methylprednisolone increases tetanic facilitation of neuromuscular transmission. Panel A, shows typical recording traces of nerve-evoked hemidiaphragm contractions obtained during brief tetanic trains (50 Hz for 5 s), in the absence (Control) and in the presence of methylprednisolone (0.3 mM). The small horizontal lines indicate the duration of tetanic stimulation (5 s); vertical calibration: 50 mN. Panel B, tetanic facilitation was calculated as increases in the ratio (R) between the tensions recorded at the end (B) and at the beginning (A) of the tetanic response (R ¼ B/A). Methylprednisolone (0.3 and 0.6 mM) was applied in a cumulative manner and contacted the preparations at least 15 min before recordings. Ordinates represent the percentage increase of the ratio (R) as compared to the control situation (in the absence of the corticosteroid). The vertical bars represent SEM of 4e5 experiments. *P < 0.05 (one-way ANOVA followed by Dunnett's modified t-test) when compared with zero percent. Panel C, shows the percentage reduction in the amplitude of muscular contraction (AMC) caused by D-tubocurarine (D-Tc, 1.5 mM) in the absence and in the presence of methylprednisolone (0.3 mM). Rat phrenic nerve-hemidiaphragm preparations were indirectly stimulated at 0.2 Hz (equilibration protocol). The vertical bars represent SEM of 6 experiments. *P < 0.05 (one-way ANOVA followed by Dunnett's modified t-test) when compared with D-Tc alone.

L. Oliveira et al. / Neuropharmacology 89 (2015) 64e76 0.5 mmol/L choline (0.005 mmol/L [3H]choline þ 0.495 mmol/L unlabelled choline dissolved in uptake buffer) in the presence and absence of methylprednisolone (0.3 mM). The effect of hemicholinium-3 (HC-3), a known inhibitor of choline uptake, was also tested by incubating the cells at 37  C for 20 min, with 0.5 mmol/L choline (0.005 mmol/L [3H]choline þ 0.495 mmol/L unlabelled choline, dissolved in uptake buffer) in the presence of 100 mM hemicholinium-3 and in the presence and absence of methylprednisolone. Choline uptake was stopped by aspiration of the incubation buffer and washing the cells three times with ice-cold stop solution (137 mmol/L NaCl, 14 mmol/L Tris, pH 7.4). The cells were then lysed on ice for 20 min by adding 500 mL of 0.5% Triton-X-100 to each well, followed by centrifugation at 300 g, 4  C, for 15 min. Radioactivity was determined by adding 3.5 ml of scintillation cocktail to 300 mL of the supernatant and tritium outflow was evaluated by liquid scintillation spectrometry (TriCarb 2900TR Perkin Elmer, Boston, USA). The cell pellet was dissolved in NaOH 0.3 M and used for protein quantification (Bio-Rad DC protein assay kit) using bovine serum albumin as protein standard. The effects of methylprednisolone and hemicholinium-3 in the choline uptake were expressed as percentage over control, after normalization to the protein content.

2.6. Drugs and solutions Methylprednisolone sodium succinate (Solumedrol™, Pfizer); Adenosine deaminase (ADA, type VI, 1803 U/mL, EC 3.3.3.4), adenosine 50 -(b,g-imido)triphosphate lithium salt hydrate (b,g-imidoATP), adenosine 50 -triphosphate disodium salt hydrate (ATP), atropine sulphate, N-[2-(p-bromocinnamylamino)ethyl]-5-isoquinolinesulfonamide dihydrochloride (H-89), a-bungarotoxin, choline chloride, hemicholinium-3, methoctramine hydrate, pirenzepine dihydrochloride, D-tubocurarine hydrochloride pentahydrate (Sigma, St Louis, MO, USA); 8-cyclopentyl-1,3dipropylxanthine (DPCPX) (Res. Biochem. Inc., Natick, MA, USA); 4-(-2-[7-amino2-{2-furyl}{1,2,4}triazolo{2,3-a}{1,3,5}triazin-5-yl-amino]ethyl) phenol (ZM 241385) (Tocris Bioscience, Bristol, UK); 3-Hydroxy-N'-[(2,4,5-trihydroxyphenyl) methylidene] naphthalene-2-carbohydrazide (Dyngo-4a) (Abcam Biochemicals, Cambridge, UK); N-(3-triethylammoniumpropyl)-4-(6-(4-(diethylamino)phenyl) hexatrienyl)pyridinium dibromide (FM4-64), tetramethylrhodamine-conjugated abungarotoxin (Invitrogen, Barcelona, Spain; Molecular Probes, Eugene, OR, USA); ATP bioluminescence assay kit HS II (Roche Applied Science, Indianapolis, Indiana). Radiolabeled [methyl-3H] choline chloride (ethanol solution, 80.6 Ci mmol1) and the scintillation cocktail (Insta-gel Plus) were obtained from Perkin Elmer (Boston, USA). Bio-Rad DC protein assay kit was purchased from Bio-Rad (Hercules, CA). FM4-64, tetramethylrhodamine-conjugated a-bungarotoxin, DPCPX and ZM 241385 were made up in dimethylsulphoxide (DMSO). All stock solutions were stored as frozen aliquots at 20  C. Dilutions of these stock solutions were made daily and appropriate solvent controls were done. No statistically significant differences between control experiments, made in the absence or in the presence of the solvents at the maximal concentrations used, were observed.

2.7. Statistics The data are expressed as mean ± S.E.M. from an n number of experiments. Statistical analysis of data was carried out using paired or unpaired Student's t-test or one-way analysis of variance (ANOVA) followed by Dunnett's modified t-test. Values of P < 0.05 were considered to represent significant differences.

67

3. Results 3.1. Methylprednisolone facilitates transmitter exocytosis from motor nerve terminals stimulated with high-frequency trains Methylprednisolone (0.3 and 0.6 mM) concentrationdependently enhanced tetanic facilitation (increase the R-value) of diaphragm muscle contractions when the phrenic nerve was stimulated with 50 Hz frequency trains, without affecting the maximal phasic tension (A-value) developed at the beginning of each tetanus (Fig. 1A and B). Methylprednisolone (0.3 and 0.6 mM) did not change (P > 0.05) the amplitude of muscle contraction when the preparations were indirectly stimulated at 0.2 Hz frequency. It also did not affect the shape of tetanic (50 Hz) muscle tension produced by direct stimulation of diaphragms paralysed with D-tubocurarine (1.5 mM) (data not shown). Conversely, the neuromuscular block caused by D-tubocurarine (1.5 mM, 75.4 ± 7.9%, n ¼ 6) was partially overcome by methylprednisolone (0.3 mM, 54.4 ± 5.9%, n ¼ 6, P < 0.05), when the corticosteroid was applied before the neuromuscular relaxing agent (Fig. 1C). Interestingly, prednisolone (0.3 mM) prevented with equal potency the neuromuscular blockade by D-tubocurarine without affecting the twitchtension of skeletal muscles with different fibre types and operating with distinct patterns of motor commands, namely the phrenic nerve-hemidiaphragm and the external popliteal/sciatic nervetibialis anterior muscle of the rat (Dal Belo et al., 2002). The same occurred in the cat tibialis-anterior (fast) and soleus (slow) muscles (Durant et al., 1984). Short-term effects of the corticosteroids apparently differ from chronic usage regarding, for instance, the degenerative changes which were more pronounced in fast-twitch than in slow-twitch muscles (Fahim, 1995). Given that the enhancing effect of the corticosteroid on tetanic facilitation suggests a dominant presynaptic action, we thought it was interesting to investigate the role of methylprednisolone directly on nerve-evoked transmitter release. Fig. 2 shows that methylprednisolone (0.3 mM) enhanced by 40 ± 11% (n ¼ 5, P < 0.05) nerve-evoked [3H]ACh release triggered by 50 Hz bursts, but the corticosteroid was devoid of effect when the phrenic nerve trunk was stimulated with 5 Hz trains (2 ± 8%, n ¼ 9, P < 0.05) keeping the number of pulses (750) and the duration of each pulse (0.04 s) constants. Note that methylprednisolone (0.3 mM) was devoid of effect on spontaneous tritium outflow under the present experimental conditions (Fig. 2).

Fig. 2. Methylprednisolone (0.3 mM) facilitates [3H]ACh release evoked by 50 Hz stimulation bursts, but not when the phrenic nerve was stimulated with 5 Hz trains. Tritium outflow (ordinates) is expressed as a percentage of the total radioactivity present in the tissue at the beginning of the collection period (Fractional release, %). Abscissa indicates the times at which samples were collected. [3H]ACh release was elicited by stimulating the phrenic nerve trunk with 750 electrical pulses delivered with frequencies of 5 Hz (A) and 50 Hz (B); a series of 5 bursts (3 s, 150 pulses, 20-s interburst interval) were delivered when the 50 Hz stimulation was applied. Each period of stimulation was applied twice, starting at 12th (S1) and 39th (S2) minutes after the end of washout (zero time). Methylprednisolone (0.3 mM, closed circles) was added to the incubation media 15 min before S2 (horizontal bar). Note that the spontaneous tritium outflow was not changed in the presence of the corticosteroid.

68

L. Oliveira et al. / Neuropharmacology 89 (2015) 64e76

Fig. 3. Facilitatory effect of methylprednisolone (MP) on nerve-evoked transmitter exocytosis measured by real-time video microscopy using the FM4-64 fluorescent dye. Transmitter exocytosis was elicited by stimulating the phrenic nerve trunk with 50 Hz bursts (five trains of 150 pulses applied with a 20-s interburst interval). MP (0.3 mM) was applied 15 min before test stimulus. Panel A, shows the FM4-64 fluorescence intensity changes in two typical motor endplates during nerve stimulation. Images were taken at the indicated times just before (0 s) and during phrenic nerve stimulation using a 63x/0.90 n.a. water-immersion objective lens (Achroplan, Zeiss, Germany). Note that fluorescence FM4-64 hotspots (white arrows) dimmed almost to the initial level (achieved before application of the loading stimulus) as a consequence of phrenic nerve stimulation. Panel B, shows the time-course of FM4-64 fluorescence intensity decay during electrical stimulation of the phrenic nerve. Fluorescence decay is expressed as a percentage of maximal loading considering that 100% is the fluorescence intensity at zero time. The vertical dashed lines represent starting and ending of the stimulus. Each value represents pooled data from 4 to 8 experiments. The vertical bars represent ± SEM. *P < 0.05 (one-way ANOVA followed by Dunnett's modified t-test) when compared with the control situation obtained in the absence of the corticosteroid.

A similar pattern was observed using video microscopy with the FM4-64 fluorescence dye to measure real-time transmitter exocytosis in preparations paralysed with the irreversible muscle-type nicotinic receptor blocker, a-bungarotoxin (4 mM), which lacks any effect on ACh release (Faria et al., 2003). Methylprednisolone (0.3 mM) significantly (P < 0.05) increased the rate of FM4-64 fluorescence intensity decay along with phrenic nerve stimulation with 50 Hz bursts (Fig. 3). Methylprednisolone-induced facilitation of transmitter exocytosis (FM4-64 unloading of hotspots) increased

progressively with the time of stimulation (Fig. 3B), which is in agreement with the predominant effect of the corticosteroid in the amplification phase of tetanic muscle tension (see Fig. 1A). 3.2. Methylprednisolone inhibits choline uptake, but this effect does not account for the neuromuscular transmission facilitation Corticosteroids interference with choline uptake and its, subsequent, incorporation in ACh has been demonstrated in the rat

Fig. 4. Effect of methylprednisolone (MP) and hemicholinium-3 (HC) on choline uptake in RBE4 and CACO-2 cells in culture compared to the action of both drugs on tetanic facilitation of neuromuscular transmission on rat phrenic nerve-hemidiaphragm preparations (NMJ). Panels A and B, the uptake of [3H]choline (0.5 mM) was measured at 20 min in presence of MP (0.3 mM), HC (100 mM) or both compounds (MP þ HC). Results are presented as mean ± SEM from four independent experiments (triplicates were performed in each experiment). Panel C, ordinates represent tetanic facilitation as percentage increases of the ratio (R) as compared to the control situation (in the absence of any added drugs). MP (0.3 mM), HC (1 mM) or both compounds (MP þ HC) were applied at least 15 min before recordings. The vertical bars represent SEM of 4e6 experiments. *P < 0.05 (one-way ANOVA followed by Dunnett's modified t-test) when compared with zero percent.

L. Oliveira et al. / Neuropharmacology 89 (2015) 64e76

69

diaphragm (Veldsema-Currie et al., 1976). Here, we used rat brain endothelial (RBE4) and human epithelial colorectal adenocarcinoma (CACO-2) cells in culture to assess the influence of methylprednisolone (0.3 mM) on [3H]choline transport. The choline uptake inhibitor, hemicholinium-3, was instrumental to discriminate the high-affinity choline transport system from low-affinity transporters. Methylprednisolone (0.3 mM) reduced by about 40% (n ¼ 4) high-affinity [3H]choline uptake by RBE4 cells considering maximal inhibition produced by hemicholinium-3 (100 mM) (Fig. 4A); yet, the same was not observed in CACO-2 cells (Fig. 4B). Pretreatment of RBE4 cells with hemicholinium-3 (100 mM) prevented further inhibition of [3H]choline uptake by methylprednisolone (0.3 mM) effect (Fig. 4A and B). Besides the conflicting effects of the corticosteroid in the two cell lineages, co-application of methylprednisolone (0.3 mM) and hemicholinium-3 (1 mM) did not increase further (P > 0.05) the facilitatory action of the corticosteroid on nerve-evoked tetanic muscle tension (increase in R-value) (Fig. 4C). These findings indicate that methylprednisolone interference with choline uptake plays a minor (if at all) role on neuromuscular transmission facilitation under the present experimental conditions. This situation totally agrees with our neurochemical findings, because methylprednisolone facilitation of nerve-evoked [3H]ACh release was obtained despite all release experiments have been conducted in the presence of hemicholinium-3 (10 mM) (see Materials and Methods). 3.3. Transmitter release facilitation by methylprednisolone depends on tonic activation of muscarinic and adenosine receptors located on motor nerve terminals Taking into account that the predominant facilitatory effect of methylprednisolone on transmitter exocytosis was observed during high-frequency (50 Hz) nerve stimulation and we showed in previous studies that ACh release is modulated by presynaptic muscarinic (M1 and M2) and adenosine (A1 and A2A) receptors depending on the stimulation conditions (see Introduction), we tested whether these receptors were involved in neuromuscular transmission amplification caused by the corticosteroid. The facilitatory effect of methylprednisolone (0.3 mM, 40 ± 11%, n ¼ 5) on nerve-evoked [3H]ACh release triggered by 50 Hz bursts was enhanced upon the blockade of muscarinic M2 and adenosine A1 inhibitory receptors with methoctramine (0.1 mM, 65 ± 13%, n ¼ 3) and DPCPX (2.5 nM, 61 ± 14%, n ¼ 5), respectively (Fig. 5). Conversely, selective blockage of facilitatory muscarinic M1 receptors with pirenzepine (10 nM) and adenosine A2A receptors (with ZM241385 10 nM) significantly (P < 0.05) attenuated methylprednisolone (0.3 mM)-induced facilitation of transmitter release to 13 ± 8% (n ¼ 6) and 8 ± 11% (n ¼ 5), respectively (Fig. 5). Inactivation of endogenous extracellular adenosine with adenosine deaminase (ADA, 0.5 U/ml) mimicked the preventive effect of the A2A receptor antagonist, ZM241385 (10 nM), on methylprednisolone facilitation of [3H]ACh release (2 ± 7%, n ¼ 4). These findings indicate that methylprednisolone-induced tetanic enhancement of neuromuscular transmission depends on tonic activation of presynaptic facilitatory muscarinic M1 and adenosine A2A receptors, but it may be partially counteracted by the activation of inhibitory M1 and A2A receptors. Moreover, Fig. 6 shows that selective blockade of both muscarinic M1 receptors and adenosine A2A receptors, respectively with pirenzepine (10 nM) and ZM241385 (10 nM), prevented (P < 0.05) the FM4-64 dye unloading caused by methylprednisolone (0.3 mM) in due course of high-frequency stimulation bursts delivered to the phrenic nerve. However unlike pirenzepine (10 nM), which attenuated the FM4-64 dye unloading immediately after the onset of the stimulus even before any measurable effect of methylprednisolone (0.3 mM) (Fig. 6A), the A2A receptor antagonist ZM 241385 (10 nM)

Fig. 5. Facilitation of nerve-evoked [3H]ACh release by methylprednisolone depends on tonic activation of muscarinic and adenosine receptors on motor nerve terminals. Transmitter release was elicited by stimulating the phrenic nerve trunk with 50 Hz bursts (five trains of 150 pulses applied with a 20-s interburst interval). Methylprednisolone (0.3 mM) was applied 15 min before test stimulus (S2). Pirenzepine (10 nM, M1 receptor antagonist), methoctramine (0.1 mM, M2 receptor antagonist), DPCPX (2.5 nM, A1 receptor antagonist) and ZM 241385 (10 nM, A2A receptor antagonist) were applied during the whole assay, including in S1 and S2. The ordinates are percentage increases in the S2/S1 ratio caused by methylprednisolone either alone (grey column) or in the presence of drugs indicated below each black column. The S2/S1 ratio obtained under these conditions, that is, with pirenzepine, methoctramine, DPCPX or ZM 241385 present in both S1 and S2, were not significantly different from the ratio obtained in control conditions. Each column represents pooled data from an n number of individual experiments (shown at the bottom of each column). The vertical bars represent ± SEM. *P < 0.05 (one-way ANOVA followed by Dunnett's modified t-test) when compared with the facilitatory effect of methylprednisolone alone.

reverted specifically the amplification phase of transmitter exocytosis caused by the corticosteroid without any effect at the initial stimulation phase (Fig. 6B). Experiments using the non-selective muscarinic receptor antagonist, atropine (0.2 mM, Fig. 6C), and ADA (0.5 U/ml, Fig. 6D) instead of pirenzepine (10 nM) and ZM241385 (10 nM) produced similar results, thus confirming that activation of muscarinic M1 and adenosine A2A receptors play predominant roles on methylprednisolone-induced facilitation of transmitter exocytosis compared to their inhibitory companion receptors, M2 and A1. 3.4. Synaptic vesicles recycling during high-frequency nerve stimulation is affected by methylprednisolone via the activation of adenosine A2A and muscarinic M1 receptors It has been demonstrated that bulk endocytosis contributes to the maintenance of neuromuscular transmission by regenerating synaptic vesicles to the readily releasable pool during highfrequency stimulation trains. Therefore, we thought it was worth to investigate the effect of methylprednisolone (0.3 mM) on bulk

70

L. Oliveira et al. / Neuropharmacology 89 (2015) 64e76

Fig. 6. The facilitatory effect of methylprednisolone (MP) on nerve-evoked transmitter exocytosis measured by video-microscopy using the FM4-64 fluorescent dye is dependent on tonic activation of presynaptic muscarinic M1 and adenosine A2A receptors. Transmitter exocytosis was elicited by stimulating the phrenic nerve trunk with 50 Hz bursts (five trains of 150 pulses applied with a 20-s interburst interval). MP (0.3 mM) in the absence and in the presence of (A) pirenzepine (PZ, 10 nM, M1 receptor antagonist) and (B) ZM 241385 (10 nM, A2A receptor antagonist) was applied 15 min before test stimulus. For comparison purposes, methylprednisolone facilitation of transmitter exocytosis was also tested in the presence of (C) atropine (0.2 mM, a non-selective muscarinic receptor antagonist) and (D) adenosine deaminase (0.5 U/ml, the enzyme that inactivates endogenous adenosine). Shown is the time-course of FM4-64 fluorescence intensity decay during electrical stimulation of the phrenic nerve. Fluorescence decay is expressed as a percentage of maximal loading considering that 100% is the fluorescence intensity at zero time. The vertical dashed lines represent starting and ending of the stimulus. Each value represents pooled data from 4 to 5 experiments. The vertical bars represent ± SEM. *P < 0.05 (one-way ANOVA followed by Dunnett's modified t-test) when compared with the facilitatory effect of the corticosteroid (see also Fig. 3B).

endocytosis during FM4-64 loading into stimulated motor nerve terminals. Fig. 7 shows that methylprednisolone (0.3 mM) attenuates FM4-64 fluorescence loading into synaptic vesicles of motor nerve terminals stimulated following a brief (5 s) high-frequency (50 Hz) tetanus. The inhibitory effect of the corticosteroid was much more evident on the late maturation phase of FM4-64 vesicles loading starting roughly 200 s after the onset of tetanic stimulus than on the initiation (0e100 s) fast fluorescence rise after beginning of the stimulus (Fig. 7B). Methylprednisolone-induced

inhibition of FM4-64 loading mimicked, although with a much lesser potency, that obtained after pretreatment of the preparations with Dyngo-4a (30 mM), a potent inhibitor of helical dynamin I (IC50 ~ 27 mM) which is thought to be responsible for facilitating endocytosis (McCluskey et al., 2013) (Fig. 7C). Inhibition of FM4-64 loading by methylprednisolone indicates for the first time that the corticosteroid may negatively modulate synaptic vesicle turnover increasing the release probability of immature recycled vesicles. Interestingly, bulk endocytosis was

L. Oliveira et al. / Neuropharmacology 89 (2015) 64e76

71

Fig. 7. FM4-64 fluorescent dye loading into stimulated motor nerve terminals in control conditions and in the presence of methylprednisolone (MP), applied either alone (open squares) or in the presence of pirenzepine (PZ, 10 nM), atropine (0.2 mM), ZM 241385 (10 nM), and adenosine deaminase (ADA, 0.5 U/ml). Stimulation (50 Hz for 5 s) of the phrenic nerve trunk was applied at zero time. Panel A, shows typical fluorescence intensity changes during FM4-64 loading into stimulated phrenic nerve terminals. Images were acquired at given time points using a 63x/0.90 n.a. water-immersion objective lens (Achroplan, Zeiss, Germany). Nerve stimulation in the presence of FM4-64 increases the fluorescence intensity in the whole terminal and elicits the appearance of FM4-64 hotspots (white arrows), which intensity is maintained fairly constant with time after washing out the dye. Panels BeG, show the average of FM4-64 fluorescence (arbitrary units) at given time points during loading of the dye into motor nerve terminals. Each value represents pooled data from an n number of experiments shown in the graphs. The vertical bars represent ± SEM. *P < 0.05 (one-way ANOVA followed by Dunnett's modified t-test) when compared with control conditions (B and C) or with the inhibitory effect of methylprednisolone alone (DeG). The effect of the selective inhibitor of endocytosis, Dyngo-4a (30 mM), is shown in Panel C for comparison purpose.

rehabilitated when methylprednisolone (0.3 mM) was applied together with pirenzepine (10 nM, Fig. 7D) or ZM241385 (10 nM, Fig. 7F), which selectively antagonize facilitatory muscarinic M1 and adenosine A2A receptors, respectively. Atropine (0.2 mM, Fig. 7E) and ADA (0.5 U/ml, Fig. 7G) also partially restored bulk endocytosis in the presence of methylprednisolone (0.3 mM), but their effects were less evident than those obtained by the selectively blockage of muscarinic M1 and adenosine A2A receptors, respectively with pirenzepine (10 nM) and ZM241385 (10 nM). Although data confirm the predominant role of facilitatory muscarinic M1 and adenosine A2A receptors on methylprednisolone-induced transmitter exocytosis, their effects may be partially counteracted by inhibitory M2 and A1 receptors that are also downregulated by atropine (0.2 mM) and ADA (0.5 U/ml), respectively.

3.5. Protein kinase A activation is required for methylprednisoloneinduced changes in synaptic vesicles recycling during highfrequency nerve stimulation The interplay between facilitatory muscarinic M1 and adenosine A2A receptors to facilitate [3H]ACh release has been demonstrated and it may occur by signal convergence to a common pathway involving protein kinase A activation and Ca2þ influx through Cav1 , 2005). In agreement (L-type) channels (Oliveira and Correia-de-Sa with this hypothesis, Fig. 8 shows that the selective and potent inhibitor of cAMP-dependent protein kinase (PKA), H-89 (10 mM), restored bulk endocytosis in the presence of methylprednisolone (0.3 mM) (Fig. 8A) in a similar manner to that observed when the corticosteroid was applied together with pirenzepine (10 nM,

72

L. Oliveira et al. / Neuropharmacology 89 (2015) 64e76

Fig. 8. Effect of the selective inhibitor of cAMP-dependent protein kinase (PKA), H-89, on FM4-64 fluorescent dye loading and unloading from stimulated motor nerve terminals in the presence of methylprednisolone (MP). Panel A, shows the average of FM4-64 fluorescence (arbitrary units) at given time points during loading of the dye into motor nerve terminals. Stimulation (50 Hz for 5 s) of the phrenic nerve trunk was applied at zero time. Panel B shows the time-course of FM4-64 fluorescence intensity decay during electrical stimulation of the phrenic nerve with 50 Hz bursts (five trains of 150 pulses applied with a 20-s interburst interval). Fluorescence decay is expressed as a percentage of maximal loading considering that 100% is the fluorescence intensity at zero time. MP (0.3 mM) in the absence and in the presence of H-89 (10 mM) was applied 15 min before stimulation. The vertical dashed lines represent starting and ending of the stimulus. Each value represents pooled data from 3 to 4 experiments. The vertical bars represent ± SEM. *P < 0.05 (oneway ANOVA followed by Dunnett's modified t-test) when compared with the effect of methylprednisolone alone.

Fig. 7B) or ZM241385 (10 nM, Fig. 7C). Like that occurring with ADA (see Fig. 6D) and the A2A receptor antagonist (see Fig. 6B), H-89 (10 mM) reverted mainly the amplification phase of transmitter exocytosis caused by the methylprednisolone (0.3 mM) without much affecting the initial stimulation phase (Fig. 8B). 3.6. The mechanism underlying facilitation of transmitter exocytosis by methylprednisolone involve an increase in the resting ATP outflow Taking into consideration that adenosine generated from released ATP activates preferentially facilitatory A2A receptors (Correia-de-S a et al., 1996; Cunha et al., 1996), we measured the ATP content in some of the samples used for quantification of [3H] ACh release from phrenic nerve-hemidiaphragm preparations. Methylprednisolone (0.3 mM) increased significantly (P < 0.05) the amount of ATP in the incubation fluid collected under resting conditions from a basal level of 19.5 ± 4.4 pM/mg to 39.6 ± 9.5 pM/ mg (n ¼ 12). Despite the ATP content of the samples increased from baseline following stimulation of the phrenic nerve with 50 Hz bursts, the nucleotide reached similar levels (P > 0.05) in the absence (83.8 ± 3.7 pM/mg) and in the presence (85.2 ± 3.9 pM/mg) of methylprednisolone (0.3 mM). These results suggest that the corticosteroid favours the outflow of ATP from resting hemidiaphragm preparations, which may anticipate adenosine accumulation at the neuromuscular synapse and favour the facilitatory A2A receptor tonus during 50 Hz bursts. A direct presynaptic action of ATP mediating methylprednisolone-induced facilitation of transmitter exocytosis is highly unlikely given that the nucleotide cause on its own an overall depression of ACh release, both quantal and non-quantal, via the activation of P2Y receptors coupled to Gq/11 and phospholipase C/protein kinase C pathways at mammalian neuromuscular junctions (Galkin et al., 2001; Malomouzh et al., 2011). In our hands, ATP (30 mM) increased by 19 ± 5% (n ¼ 5, P < 0.05) and 32 ± 4%

(n ¼ 5, P < 0.05) nerve-evoked [3H]ACh release triggered by 5 Hz trains and 50 Hz bursts, respectively. Both ATP (30 mM) effects were completely reversed when the preparations were pretreated with ADA (0.5 U/ml), indicating that the facilitatory effects of ATP are indeed mediated by adenosine. That is, in the presence of ADA (0.5 U/ml), ATP (30 mM) consistently reduced [3H]ACh release when the phrenic nerve was stimulated with 5 Hz trains (12 ± 8%, n ¼ 4, P > 0.05) and 50 Hz bursts (24 ± 8%, n ¼ 4, P < 0.05). The effects of ATP (30 mM) in the presence of ADA (0.5 U/ml) reproduced the transmitter release inhibition caused by the non-hydrolysable ATP analogue, b,g-imidoATP (30 mM); this compound reduced [3H]ACh release by 5 ± 4% (n ¼ 3, P > 0.05) and 15 ± 5% (n ¼ 3, P < 0.05) when the phrenic nerve was stimulated with 5 Hz trains and 50 Hz bursts, respectively (data not shown). These results fully agree with data from our group using the adenosine precursor, AMP, to assess the facilitatory role of adenosine originating from the extracellular catabolism of adenine nucleotides at both hippocampal and neuromuscular synapses of the rat (Correia-de-S a et al., 1996; Cunha et al., 1996). 4. Discussion In this study, we present evidences demonstrating that methylprednisolone facilitates neuromuscular transmission by increasing the amount of ACh release during high-frequency nerve activity. Methylprednisolone-induced facilitation of nerve-evoked transmitter exocytosis was confirmed using “indirect” methods, such as myographic recordings (tetanic enhancement of skeletal muscle tension), in addition to “direct” measurements of transmitter release, like a radiochemical technique to detect evoked [3H] ACh outflow and video microscopy with the membrane-selective FM4-64 fluorescent dye to assess real-time vesicle exocytosis. We show here for the first time that beneficial effects of the corticosteroid on neuromuscular transmission depend on the activation of presynaptic muscarinic and adenosine receptors. Data suggest that

L. Oliveira et al. / Neuropharmacology 89 (2015) 64e76

tonic activation of facilitatory A2A receptors by endogenous adenosine generated from ATP released under resting conditions is vital for methylprednisolone-induced facilitation of transmitter release during high-frequency bursts. Concurrent activation of the positive feedback loop operated by muscarinic M1 autoreceptors by released ACh may also play a role. Although the molecular mechanism(s) underlying the link between the glucocorticoid action and stimulation of presynaptic muscarinic and adenosine receptors deserves further investigations, our results strongly indicate that it may involve the negative modulation of synaptic vesicle maturation favouring the release of newly recycled vesicles. These novel findings prompt for further elucidation in the context of their clinical relevance for the management of autoimmune neuromuscular transmission deficits, namely in that concerning favourable changes in neuromuscular junction morphology (Sieck et al., 1999), synthesis and stabilization of postjunctional nicotinic receptors (Braun et al., 1993) and decreased acetylcholinesterase activity (Brank et al., 1998) after chronic glucocorticoid treatment (reviewed in Askanas et al., 1992). Our group demonstrated previously that endogenous adenosine generated in myasthenic motor endplates during high-frequency nerve stimulation may be insufficient to maintain transmitter release demand through tonic activation of presynaptic facilitatory A2A receptors (Noronha-Matos et al., 2011). Reduced levels of adenosine (less than 10%) were found during paralysis of the diaphragm with m-conotoxin GIIIB, a toxin that abolishes muscle action potentials by blocking voltage-gated Naþ channels of skeletal muscle without affecting ACh release (see e.g. Faria et al., 2003). Interestingly, deficient endogenous adenosine accumulation and neuromuscular transmission deficits in myasthenics could be rehabilitated by A2A receptor activation using the nucleoside precursor AMP. It is, therefore, tempting to speculate that glucocorticoid might favour adenosine accumulation at the neuromuscular synapse to levels high enough to activate presynaptic facilitatory A2A receptors leading to increases in ACh release and to potentiation of tetanic muscle contractions. In fact, we show here that methylprednisolone favours the outflow of ATP from resting phrenic nerve-hemidiaphragm preparations, which upon extracellular hydrolysis may contribute to increase the synaptic adenosine levels and, thus, the preferential activation of facilitatory A2A  et al., 1996; Cunha et al., 1996). Such a receptors (Correia-de-Sa mechanism would be clinically relevant to explain corticosteroid benefits (avoidance of neuromuscular tetanic failure) in patients with Myasthenia gravis. Data from the present study corroborate this hypothesis, given that inactivation of endogenous adenosine with ADA or selective blockade of adenosine A2A receptors with ZM 241385 significantly attenuated methylprednisolone-induced facilitation of ACh release measured using both a radioisotope method and real-time fluorescence video microscopy. This effect was only evident during the amplification phase of transmitter exocytosis when the phrenic nerve was stimulated with highfrequency (50 Hz) bursts, but not when we used a frequency (5 Hz) that mimics the mean physiological respiratory rhythm in non-anesthetized rats (cf. Monteiro and Ribeiro, 1987). Interestingly, neuromuscular fatigue in Myasthenia gravis patients is accentuated when the nerve stimulation frequency increases from 5 to 50 Hz (Conti-Fine et al., 2006; Hirsch, 2007). These results gain pathophysiological interest because it has been proven that during daily voluntary movements and resistive inspiration drive to thoracic muscles the pattern of motoneuron firing frequencies is variable (ranging from 100 to

Amplification of neuromuscular transmission by methylprednisolone involves activation of presynaptic facilitatory adenosine A2A receptors and redistribution of synaptic vesicles.

The mechanisms underlying improvement of neuromuscular transmission deficits by glucocorticoids are still a matter of debate despite these compounds h...
2MB Sizes 0 Downloads 10 Views