Abstract
Functional ionotropic nucleotidic receptors responding to diadenosine pentaphospate and nicotinic receptors responding to epibatidine coexpress in 19% of the total rat midbrain cholinergic terminals, as determined by the combination of immunological and microfluorimetric techniques. Activation of each independent receptor induces the intrasynaptosomal [Ca2+]i and acetylcholine (ACh) release in a dose-dependent way. The responses are inhibited by antagonists of the dinucleotide receptor and nicotinic receptors, thus confirming the involvement of specific receptors in both functions. Stimulation of single cholinergic terminal with both agonists altogether results in a significant decrease of the [Ca2+]i signaling compared with responses of each independent agonist. Inhibitory interaction between both receptors is reverted when one of them is blocked by specific antagonists, both in [Ca2+]i, and subsequent ACh release. The receptor′s inhibitory cross talk confirm the involvement of calcium/calmodulin-dependent protein kinase II, CaMKII, as the inhibitory effects are reverted in the presence of the specific inhibitors KN-62 (2-[N-(4′-methoxybenzenesulfonyl)]-amino-N-(4′-chlorophenyl)-2-propenyl-N-methylbenzylamine phosphate) and KN-93 (N-(2-[N-[4-chlorocinnamyl]-N-methylaminomethyl]phenyl)-N-(2-hydroxyethyl)-4-methoxybenzenesulphonamide). These results demonstrate the existence of an efficient interaction between these two channel populations, opening a new understanding of the functioning of the cholinergic synaptic terminals or terminals containing other neurotransmitters but exhibiting these receptor types or ones that are similar.
Neuronal nicotinic acetylcholine receptors are widely expressed in the vertebrate central nervous system (CNS) and can exhibit a pre- or postsynaptic location (Wonnacott, 1997; Dani, 2001). These acetylcholine-gated cation channels are pentameric structures containing α and β subunits (McGehee and Role, 1995). At the CNS, six vertebrate α subunits (α2–α8) and three β subunits (β2–β4) have so far been cloned. In heterologous expression systems, α7, α8, and α9 subunits can form functional homomeric receptors, which are blocked by α-bungarotoxin. On the other hand, α2, α3, α4, and α6 subunits always need to be coexpressed with β2–β4 subunits to generate functional receptors insensitive to α-bungarotoxin and activated by epibatidine (Albuquerque et al., 1995; Lindstrom et al., 1995). In addition, other authors described the possible agonistic effect of epibatidine on α7 homomeric nicotinic receptors (Delbono et al., 1997). Nicotinic receptors containing the α7 subunit have been described as the most abundant at the presynaptic level (Wonnacott, 1997; Dani, 2001); studies in synaptosomes from hippocampus, striatum, and locus coeruleus have demonstrated the existence of nicotinic receptors with the participation of α4 subunits (Egan and North, 1986; Nayak et al., 2001). Finally, pharmacological similarities between native brain and heterologously expressed α4β2 nicotinic receptors have also been described (Shafaee et al., 1999).
Diadenosine polyphosphates, also termed adenine dinucleotides, are a family of nucleotidic compounds formed by two adenosines linked by a variable number of phosphates (Baxi and Vishwanatha, 1995; Pintor et al., 2000). These compounds, together with ATP, are stored in cholinergic synaptic vesicles and released to synaptic clef after stimulation (Potter and White, 1980; Pintor et al., 2000). In the CNS, diadenosine polyphosphates are able to activate specific presynaptic ionotropic receptors (Pintor and Miras-Portugal, 1995), which exhibit permeability to Ca2+ ions (Pivorun and Nordone, 1996; Pintor et al., 2000). Studies on isolated single terminals have confirmed the existence of independent responses for diadenosine pentaphospate (Ap5A) and ATP on the same or different synaptic terminals from rat midbrain (Díaz-Hernández et al., 2001).
In the last years, the existence of an inhibitory interaction between cholinergic and purinergic systems has been described, and the existence of mutual occlusion between currents induced by ATP and acetylcholine (ACh) activating their specific ionotropic receptors has been shown. Although some theoretical models have been proposed, the mechanism of this inhibitory interaction is not well understood (Nakazawa, 1994; Khakh et al., 2000). The interaction between the cholinergic and purinergic systems became more relevant when the existence of functional-specific ATP or dinucleotide receptors together with nicotinic receptors, both sensitive or insensitive to epibatidine, were found on the same individual synaptic terminal (Díaz-Hernández et al., 2002).
The aim of the present work is to study the possible interaction between dinucleotide and nicotinic receptors and the way it occurs. This objective is mainly addressed studying ACh secretion and Ca2+ signaling. The Fura-2 technique for Ca2+ determination will be carried out in synaptosomal population and by video microscopy in isolated single cholinergic terminals. This study reports a significant degree of colocalization of specific dinucleotide ionotropic and epibatidine-responding nicotinic receptors in cholinergic terminals leading to reciprocal inhibition in ACh secretion and [Ca2+]; signaling, with the calcium/calmodulin-dependent protein kinase II (CaMKII) being involved in these effects.
Materials and Methods
Synaptosomal Preparation. Synaptosomes were prepared from rat midbrain (diencephalon/brainstem) of cervically dislocated and decapitated male Wistar rats (6 and 7 weeks old). All experiments carried out at the Universidad Complutense de Madrid followed the guidelines of the International Council for Laboratory Animal Science (ICLAS). The isolation procedure was different depending on the aim of the synaptosomal preparations. Synaptosomes used for ACh release and pharmacological studies were obtained according to Pintor and Miras-Portugal (1995). Synaptosomes used for intracellular calcium measurements in isolate single synaptic terminals and for immunochemical analyses were obtained by means of a Percoll gradient following the procedure described by Dunkley et al. (1986).
Acetylcholine Release Determination. Synaptosomal pellets containing 1 mg of protein were resuspended in 1 ml of incubation medium (composition: 122 mM NaCl, 3.1 mM KCl, 0.4 mM KH2PO4, 5 mM NaHCO3, 1.2 mM MgSO4, 10 mM glucose, 1.33 mM CaCl2, and 20 mM TES buffer, pH 7.4) (buffer A) and equilibrated at 37°C for 5 min. The synaptosomes were stimulated for 30 s with epibatidine and/or Ap5A at indicated concentrations and pelleted at 13,000 rpm for 1 min at 4°C. When the effect of nicotinic receptor antagonists, mecamylamine and hexamethonium as well as purinergic antagonist, suramin, a wide P2 receptor antagonist, or Ip5I were assayed, synaptosomes were incubated in the presence of the antagonist for 2 min prior to the addition of the agonist. In the case of α-bungarotoxin (antagonist of α7 nicotinic receptors) synaptosomes were preincubated for 10 min prior to the addition of the agonist. In some cases, the effect of the epibatidine and Ap5A were assayed on synaptosomes in a virtually Ca2+-free medium ([Ca2+] ∼200 nM). This medium was obtained by substituting 1.33 mM CaCl2 with an EGTA (50 μM)/CaCl2 (38 μM) mixture in buffer A.
The ACh released from synaptosomal preparations was measured by using the choline oxidase chemiluminescent method as previously described by Israel and Lesbats (1981). Briefly, aliquots of the samples (200 μl) were diluted in 500 μl of fresh physiological medium containing: 122 mM NaCl, 3.1 mM KCl, 0.4 mM KH2PO4, 5 mM NaHCO3, 1.2 mM MgSO4, 10 mM glucose, and 20 mM Tris buffer, pH 8.6) (buffer B). ACh measurements were performed by applying choline oxidase (EC 1.1.3.17) 2.5 U after a 2-min incubation with physiological medium containing the other reagents for the chemiluminescent assay [luminol 40 μM, peroxidase (EC 1.11.1.7) 5.2 U, and AChE (EC 3.1.1.7) 1 U]. All reagents were from Sigma-Aldrich (St. Louis, MO). The light emitted by the chemiluminescent reaction was detected by a BioOrbit 1250 luminometer connected to a recorder. The calibration curve was performed by additions of known amounts of ACh standard as previously described by Israel and Lesbats (1981). In all cases, the quantity of ACh present in control situations (in the absence of any stimulation) was subtracted to calculate the amount of ACh released. These data are presented as mean ± S.E.M. of at least three determinations in triplicate from different synaptosomal preparations. Purinergic and cholinergic agonists and antagonists used in pharmacological studies were added directly to the medium, and at the concentrations used, they showed no effect on the ACh assay.
Ca2+Response in Synaptosomal Populations. The cytosolic-free calcium concentration was determined using Fura-2 as described by Grynkiewicz et al. (1985). Five minutes after synaptosomal resuspension, CaCl2 (1.33 mM) and Fura-2/acetoxymethyl ester (5 μM) were added. Following incubation for 35 min, the synaptosomes were pelleted at 800 rpm for 1 min, washed twice, and resuspended into fresh medium containing 1.33 mM CaCl2. Fluorescence was measured in a Perkin Elmer Spectrofluorimeter LS-50 and monitored at 340 and 510 nm. Data were collected at 0.5 s intervals.
Epibatidine was assayed at concentrations ranging from 10–11 to 10–8 M to obtain the dose-response curve. In some cases, specific protein kinase C (PKC) inhibitors such as staurosporine (100 nM), 2-{1-[3-(amidinothio)propyl]-1H-indol-3-yl}-3-(1-methylindol-3-yl)-maleimide (Ro 31-8220, 100 nM), 2-[1-(3-dimethylaminopropyl)-5-metoxyindol-3-yl]-3-(1H-indol-3-yl) maleimide (Gö 6983, 10 nM), and 5,6,7,13-tetrahydro-13-methyl-5-oxo-12H-indolo[2,3-a]pyrrolo[3,4-c]carbazole-12-propanenitrile (Gö 6976, 10 nM) as well as specific CaMKII inhibitors N-(2-[N-[4-chlorocinnamyl]-N-methylaminomethyl]phenyl)-N-(2-hydroxyethyl)-4-methoxybenzenesulphonamide (KN-93, 9 μM) and 2-[N-(4′-methoxybenzenesulfonyl)]amino-N-(4′-chlorophenyl)-2-propenyl-N-methylbenzylamine phosphate (KN-62, 9 μM) were preincubated for 30 min together with the synaptosomes before the application of the agonist.
Image Acquisition and Analysis of Ca2+Response in Isolated Synaptic Terminal. Synaptosomal pellets containing 0.5 mg of protein were resuspended in 1 ml of incubation medium (composition: 122 mM NaCl, 3.1 mM KCl, 0.4 mM KH2PO4, 5 mM NaHCO3, 1.2 mM MgSO4, 10 mM glucose, and 20 mM TES buffer, pH 7.4) (buffer C) and were loaded with Fura-2/AM (5 μM) for 1 h at 37°C. After pelleting and washing, synaptosomes were deposited onto coverslips pretreated with poly-l-lysine and maintained for 45 min to allow setting and sticking to the substrate. This period also allowed for intrasynaptosomal hydrolysis of the Fura-2/AM. The coverslips were washed with fresh medium and mounted in a small superfusion chamber in the stage of a Nikon TE-200 microscope. The synaptosomes were then superfused with 50 μM Ap5A and/or 10 nM epibatidine in the presence or absence of other compounds. A pulse of 30 mM KCl was applied at the end of each experiment to confirm the functionality and viability of the synaptosomes under study.
Synaptosomes were imaged through a Nikon x100 (oil, 1.3 NA). Emitted light was isolated with a dichroic mirror (430 nm) and a 510-nm bandpass filter (Omega Optical Inc., Brattleboro, VT). The wavelength of the incoming light was selected with the aid of a monochromator (12-nm bandwidth; PerkinElmer Life and Analytical Sciences, Boston, MA) to 340 and 380 nm. These wavelengths corresponded to the peaks of fluorescence of Ca2+-saturated and Ca2+-free Fura-2 solutions determined empirically in our system. We checked that the isosbestic point for Fura-2 was between these two wavelengths (around 390 nm) as described by Díaz-Hernández et al. (2001). Twelve-bit images were acquired by Ultrapix 2000 Mono CCD camera controlled by Ultraview PC software (PerkinElmer Life and Analytical Sciences). The exposure time was 822 ms for each wavelength and the changing time <5 ms. The images were acquired continuously and buffered in a fast Small Computer System Interface disk. Time course data represent the average light intensity in a small elliptical region inside each terminal.
The background and autofluorescence components were subtracted at each wavelength and the ratio 340 over 380 nm calculated using 32-bit float arithmetics (real numbers). The ratioed images were calibrated into [Ca2+] values using the equation of Grynkiewicz et al. (1985). The Rmax, Rmin, and β parameters were calculated from the spectra of small droplets of Fura-2 in Ca2+-saturated and Ca2+-free solutions (composition: 100 mM KCl, 10 mM NaCl, 1 mM MgCl2, 10 mM Tris, 10 mM MOPS and 100 μM Fura-2; Ca2+-free: plus 2 mM EGTA; Ca2+-saturated: plus 2 mM CaCl2) (Díaz-Hernández et al., 2001). The in vitro Rmax and Rmin were corrected for differences with the cytosolic environment using the procedure described by Poenie (1990). All image calculations were performed with Lucida 3.0 (Kinetic Imaging, Research Triangle Park, NC). For the measurement of the average [Ca2+]i of each object, a binary mask was created for each field image containing the outlines of the synaptosomes of each field. The background zones outside synaptosomal profiles were set to zero in the rational images. Binary masks were created with free image processing software, (ScionImage; Scion Corporation, Frederick, MD and ImageTool; University of Texas Health Science Center, San Antonio, TX). Measurements were performed in Lucida 3.0 averaging nonzero pixels within ellipsoidal regions fitted to each object of interest in the image.
Immunocytochemical Studies. Synaptosomes were taken after the experiments and treated with 4% paraformaldehyde (PFA) for 15 min at room temperature and washed twice with PBS medium. The following step was to incubate the synaptosomes with PBS medium containing 3% BSA, 0.1% Triton X-100, and 5% normal rat serum for 1 h. Next, synaptosomes were washed twice with PBS in the presence of 3% BSA and afterward were incubated for 1 h at 37°C with primary antibodies diluted in PBS in the presence of 3% BSA. The primary antibodies used were monoclonal mouse and rabbit antisynaptophysin used at a concentration of 2 μg/ml, goat anti-postsynaptic density protein (PSD)-95 (1:500), rabbit anti-glial fibrillary acidic protein (GFAP) (1:100), monoclonal rat anti-α4 nicotinic acetylcholine receptor, and polyclonal goat anti-vesicular acetylcholine transporter (VAT) (1:200). Subsequently, synaptosomes were washed three times with PBS in the presence of 3% BSA and incubated for 1 h at 37°C with the secondary antibody, which were diluted in PBS in the presence of 3% BSA. Secondary antibodies used were goat anti-rabbit IgG marked with rhodamine (40 μg/ml), goat anti-mouse IgG marked with fluorescein (40 μg/ml), donkey anti-goat IgG marked with rhodamine or fluorescein (1:500), donkey anti-rabbit IgG marked with fluorescein (1:500), and donkey anti-rat IgG marked with fluorescein (1:500). Then, synaptosomes were washed three times with PBS and mounted following standard procedures. Controls were performed by following the same procedure but substituting the primary antibodies by PBS in the presence of 3% BSA.
The terminals displaying colocalization of two markers were quantified by taking successive fluorescein isothiocyanate (FITC), tetramethylrhodamine isothiocyanate, and fluorescent images using a Nikon TE-200 microscope and a CCD camera (Ultrapix 2000 Mono). Images were analyzed using Lucida 3.0 software (Kinetic Imaging). Synaptosomes were considered positive for the different antibodies if the mean intensity value was >130 on a 0 to 255 scale with 0 = white and 255 = black. The cut off value of 130 was determined from visual analysis of immunolabeling and by comparison with control (maximal level obtained with preabsorbed antibodies). Data are represented as mean ± S.E.M. of 10 to 15 fields from at least three different synaptosomal preparations.
Statistical Analysis. Each experiment reported here was carried out with at least three independent synaptosomal preparations and two and three fields for each one studied. Usually, a number of about 60 to 70 synaptosomes per field were individually analyzed in their ability to respond to the different tested compounds in each protocol and in each experiment. Data are presented as mean ± S.E.M. of three or more experiments in triplicate. Significant differences were determined by ANOVA test. Bonferroni post-test analysis was only applied when a significant (p < 0.05) main effect was indicated by the ANOVA. When appropriate, single experimental traces are represented in the figures; these represent at least three determinations performed in triplicate with equivalent results.
Materials. Ap5A, epibatidine, AChE (EC 3.1.1.7), choline oxidase (EC 1.1.3.17), peroxidase (EC 1.11.1.7), ACh, luminol, hexamethonium, α-bungarotoxin, mecamylamine, goat anti-PSD-95, rabbit anti-GFAP, monoclonal rat anti-α4 nicotinic receptor subunit, goat anti-rabbit IgG marked with rhodamine, and goat anti-mouse IgG marked with fluorescein were obtained from Sigma-Aldrich. KN-62, KN-93, Ro 31-8220, Gö 6976, Gö 6983, staurosporine, and goat anti-VAT were obtained from Calbiochem (Schwalbach, Germany). Suramin, Fura-2/AM, donkey anti-goat IgG marked with rhodamine or fluorescein, and donkey anti-rat IgG marked with fluorescein were obtained from Molecular Probes (Eugene, OR). Monoclonal mouse and rabbit antisynaptophysin was obtained from Roche Diagnostics (Indianapolis, IN). Donkey anti-rabbit IgG marked with fluorescein was obtained from Santa Cruz Biochemicals (Santa Cruz, CA).
Results
Functional Dinucleotide-Sensitive Receptor Induces Acetylcholine Release from Rat Midbrain Cholinergic Terminals. Cholinergic synaptic terminals were identified by immunolabeling with an anti-VAT antibody as previously described by Díaz-Hernández et al. (2002). In rat midbrain, the cholinergic population VAT positive represents 22 ± 4% of the total of the synaptic terminals identified with antisynaptophysin antibody (Fig. 1A; Table 1). Immunochemical studies carried out with the postsynaptic density PSD-95 marker and the GFAP marker showed no contamination by these structures (results not shown) (Díaz-Hernández et al., 2002).
To characterize the presence of dinucleotide-sensitive receptor in cholinergic terminals, it was necessary to combine immunological and microfluorimetric techniques. To perform these studies, rat midbrain synaptic terminals were loaded with the Ca2+ dye Fura-2, adhered to glass coverslips, and placed in a microperfusion chamber. Then they were challenged with ATP, Ap5A, and their Ca2+ responses followed by fluorescence microscopy coupled to video imaging (Fig. 1B). After the functional calcium studies were carried out, the synaptic terminals were fixed with PFA and the cholinergic terminals identified immunologically, as can be seen in Fig. 1, A and B. The analysis of roughly 3000 cholinergic terminals showed that more than one-third of them presented a dinucleotide-sensitive receptor (Table 1). This percentage of response was not modified when synaptosomes were first challenged with ATP demonstrating the lack of desensitization mediated by the ATP receptor (results not shown). The large variability in the response intensity elicited by Ap5A between cholinergic terminals is noteworthy. Calibration of fluorescence ratio (see Materials and Methods) to calculate intrasynaptosomal [Ca2+]i concentration showed a basal value of 115 ± 30.5 nM. It is noticeable that the calcium increase induced by 50 μM Ap5A in total synaptosomal populations was 201 ± 50.3 nM and in single cholinergic terminals showed values ranging from 40 to 120 nM (Fig. 1B).
Once the presence of dinucleotide-sensitive receptors on cholinergic terminals was demonstrated, studies on acetylcholine secretion were carried out according to the Israel and Lesbats (1981) protocol. When rat midbrain synaptic terminals were stimulated with Ap5A in a concentration range from 10–7 to 10–4 M, a clear increase in the extrasynaptosomal ACh concentration was observed. The ACh release induced by Ap5A was dose-dependent and required the presence of extracellular Ca2+ (Fig. 1, C and D; Table 2). The EC50 value for ACh release was 1.5 ± 0.1 μM, and the maximal secretory response was 46.8 ± 7.3 pmol/mg of synaptosomal protein (Fig. 1C). In the same preparation, the massive synaptosomal depolarization induced by high levels of potassium (60 mM) or the lysis of the synaptic terminals by hypo-osmotic shock produced 88.33 ± 4.9 and 141.6 ± 18.6 pmol/mg of synaptosomal protein, respectively (Fig. 1D).
To characterize the pharmacological profile of the dinucleotide-sensitive receptor from cholinergic terminals, specific purinergic receptor antagonists were tested. In this way, preincubation of the synaptic terminals with suramin (100 μM), a P2 receptor antagonist, 2 min before the Ap5A was applied, did not significantly reduce the ACh release induced by this compound (Table 2); however, the addition to these preparations of the specific dinucleotide receptor antagonist Ip5I (100 nM) (Pintor et al., 1997) almost completely abolished ACh release induced by Ap5A (Fig. 1D; Table 2). These data suggest that Ap5A on rat midbrain cholinergic terminals is activating specific dinucleotide receptors rather than P2 receptors (Pintor and Miras-Portugal, 1995; Díaz-Hernández et al., 2002).
Epibatidine-Responding Nicotinic Receptors on Cholinergic Synaptic Terminals Induce the ACh Release. After the functional calcium microfluorimetric studies were carried out, the synaptic terminals were fixed with PFA and the cholinergic ones immunologically identified with anti-VAT antibody. In these experiments, the synaptic terminals were also labeled with antibodies against one of the most abundant nicotinic receptor subunit in the CNS, the α4 subtype (Fig. 2A; Table 1). The analysis of these experiments showed that more than 80% of the cholinergic terminals responded to nicotine, and a subpopulation reaching 44% of them responded not only to nicotine but also to epibatidine (Table 1). In the same way that it occurs with Ap5A, a large variability in the intensity of the responses elicited by epibatidine and nicotine among cholinergic terminals was observed. Calibration of fluorescence ratio to get intrasynaptosomal [Ca2+]i concentration showed a basal value of 115 ± 20.3 nM. It is also noteworthy that in total synaptosomal populations the calcium increase corresponding to 10 nM epibatidine and 100 μM nicotine were 210.2 ± 51.5 and 215 ± 49.5 nM, respectively, whereas single cholinergic terminals showed values ranging from 80 to 140 nM for both agonists (Fig. 2B). Finally, it is noteworthy that as much as two-thirds of the cholinergic terminals responding to epibatidine also presented positive labeling to anti-α4 nicotinic receptor subunit antibody (Table 1), as is shown in Fig. 2, A and B (synaptic terminal labeled as 2).
When the presence of epibatidine-sensitive nicotinic receptors in the cholinergic terminals was demonstrated, studies on acetylcholine secretion were carried out. Once stimulated with epibatidine, synaptosomal preparations released acetylcholine in a dose-dependent way (Fig. 2C; Table 2). The dose-response curve for epibatidine exhibited an EC50 value of 1.3 ± 0.15 nM and a maximal secretory response value of 39.9 ± 7.5 pmol/mg of synaptosomal protein. In Fig. 2C and for comparative purposes, the epibatidine dose-response curve in intrasynaptosomal [Ca2+]i is also shown. The EC50 and the maximal intrasynaptosomal [Ca2+]i increase values obtained were 0.3 ± 0.02 and 52.1 ± 1.31 nM, respectively. The secretory effect evoked by epibatidine was absolutely dependent on the presence of extrasynaptosomal calcium, as shown in Fig. 2D. When the extrasynaptosomal Ca2+ concentration was reduced to nanomolar values, as described under Materials and Methods, epibatidine-evoked ACh release was severely reduced.
Figure 2D shows the effect of specific nicotinic receptor inhibitors on acetylcholine secretion. In this way, mecamylamine and hexamethonium induced a pronounced inhibition of epibatidine-induced ACh release (Table 2). Concerning α-bungarotoxin, which is a specific inhibitor for homomeric α7 nicotinic receptors, it did not significantly modify the ACh release induced by epibatidine when all the experiments are analyzed together for statistical purposes, although a slight decrease was always observed in each individual experiment. These results confirmed that epibatidine acting on presynaptic nicotinic receptors is able to induce the exocytotic release of acetylcholine.
Colocalization of Ap5A and Epibatidine Responses in Cholinergic Synaptic Terminals. Rat midbrain single synaptic terminals were stimulated with Ap5A and epibatidine and subsequently characterized by immunostaining with VAT and α4 antibodies. Cholinergic terminals (red color), α4 nicotinic receptor subunit (green color), and the colocalization of both proteins (yellow color) is shown in Fig. 3A (left hand panel) and the scheme of this distribution is shown in Fig. 7. In the right-hand panel of Fig. 3A, the fluorescence records of Ca2+ response in individual synaptic terminals are shown. Cholinergic terminals 1 and 2 both respond to Ap5A and epibatidine and represent a 19 ± 3% of the cholinergic terminals (Table 1). However, the presence of α4 subunits is only detectable in type 1; this result indicates the involvement of other subtypes of α nicotinic subunits on epibatidine responses (Delbono et al., 1997). As summarized in Table 1, it is noteworthy to indicate that 25 ± 4% of the cholinergic terminals responded only to epibatidine and 18 ± 2% responded only to Ap5A.
Nonadditive Intrasynaptosomal Ca2+Increase Induced by Epibatidine and Ap5A in the Same Single Cholinergic Terminal. Once the coexpression of epibatidine-sensitive nicotinic and dinucleotide receptors on the same cholinergic terminal was demonstrated (Fig. 3A), the next step was to determine whether the [Ca2+]i increase induced by Ap5A or epibatidine could be modified when their specific receptors are simultaneously stimulated.
To perform this study, the cholinergic synaptic terminals were stimulated first with 50 μMAp5A, followed by a second pulse of 10 nM epibatidine, and afterward by both agonists together in a third pulse. Before each agonist pulse, the preparation was always washed with the perfusion media for at least one minute. In terminals only responding to Ap5A (Fig. 3B, synaptic terminal labeled as 1), the intrasynaptosomal Ca2+ increase induced by both agonists was not significantly different from that induced by Ap5A alone, as is shown in Fig. 3B, right panel. A similar result was obtained in the terminals only responding to epibatidine (Fig. 3B, synaptic terminal labeled as 2). However, a different situation occurs in those terminals sensitive to both agonists (Fig. 3B, synaptic terminal labeled as 3). In 95% of these terminals, coapplication of both agonists produced responses that represent only 46.2 ± 5.6% of the addition of [Ca2+]i increase induced by each independent agonist (n = 55) (Fig. 3B, synaptic terminal labeled as 3, right side).
Activation of Epibatidine-Responding Nicotinic Receptors Decreases the Ap5A-Induced Intrasynaptosomal Calcium Increase and ACh Release. Because [Ca2+]i increase mediates the release of neurotransmitters at the synaptic level, studies on ACh secretion in the same experimental conditions, as reported in previous paragraph, were carried out. Similar nonadditive effects of 50 μM Ap5A and 10 nM epibatidine on the ACh secretion was observed. In this case, the ACh release induced by the coapplication of both agonists was only 44.8 ± 7.5% (n = 5) of the addition of the ACh release induced by each independent agonist. To analyze the nonadditive effect on ACh secretion, the costimulation experiments were carried out in the presence of inhibitors for nicotinic receptors such as mecamylamine and hexamethonium or dinucleotide receptor inhibitors as the Ip5I, as shown in Table 2.
Based on these results, the next studies were designed to better understand how [Ca2+]i increase or the ACh release induced by Ap5A could be modified by the prior activation of epibatidine-sensitive nicotinic receptors. To carry out this study, cholinergic terminals were preincubated with 10 nM epibatidine 2 min before 50 μMAp5A was assayed. As shown in Fig. 4A, in the cholinergic terminals analyzed presenting both functional receptors, the prior activation of epibatidine receptors induced a decrease of 46.5 ± 6.8% on [Ca2+]i increase elicited by Ap5A (n = 32).
Similar inhibitory effects were observed in ACh release induced by Ap5A at nonsaturating concentrations. Prestimulation of synaptic terminals with 1 nM epibatidine decreased to 48.78 ± 8.23% the ACh release induced by 10 μM Ap5A (Fig. 4B). This inhibitory effect was dose-dependent, presenting IC50 values of 0.12 ± 0.03 nM (Fig. 4C). It is noteworthy that the inhibitory effect of epibatidine on the ACh secretion induced by Ap5A can be observed at epibatidine concentrations almost ineffective to induce its own ACh release but enough to increase intrasynaptosomal calcium concentration (Fig. 2C). However, at higher epibatidine concentrations (5–10 nM), the observed increase in the ACh release is due to the epibatidine action on its specific terminals where the Ap5A receptors are not present (see scheme Fig. 7).
The inhibitory interaction between epibatidine and dinucleotide receptors was confirmed by means of antagonists of nicotinic receptors. In the presence of mecamylamine or hexamethonium in the extrasynaptosomal media, the inhibitory effect induced by epibatidine on ACh release was reverted (Fig. 4D). However, the preincubation of synaptosomes with α-bungarotoxin, antagonists of the α7 nicotinic receptors, did not significantly modify the inhibitory effect, although showing a slight decrease tendency similar to the one described in Fig. 2D (Fig. 4D).
Activation of Dinucleotide-Sensitive Receptors Decreases the Epibatidine-Induced Intrasynaptosomal Calcium Increase and ACh Release. Once it was demonstrated that the inhibitory effect of epibatidine on the dinucleotide receptor response was mediated through its specific nicotinic receptors, the effects of dinucleotide receptor activation on functional responses of epibatidine receptors were studied. Two parameters of functionality were analyzed, the [Ca2+]i increase in isolated cholinergic terminals and the ACh release. Studies in single terminals were carried out by preincubation of synaptic terminals adhered to coverslips preincubated with 5 μM Ap5A for 2 min before 10 nM epibatidine was assayed. The analysis of cholinergic terminals presenting both functional receptors showed that the [Ca2+]i induced by epibatidine in the presence of Ap5A was 47.8 ± 5.7% smaller than that obtained on the same cholinergic terminal in the absence of this compound (n = 29). Identification of cholinergic terminals followed the functional studies (Fig. 5A).
The ACh release studies showed that low Ap5A concentration had no effect on the secretion induced by 5 nM epibatidine. Besides, at higher Ap5A concentrations, a dose-response curve was clearly observed, which is superimposed to the epibatidine response (Fig. 5B). Nevertheless, the maximal response is clearly lower (40.2 ± 7.6%) than that expected from the independent addition of both secretory responses and in good agreement with the existence of terminals coexpressing both receptors (scheme Fig. 7). Concerning the EC50 value (2.86 ± 0.73 μM), it is not significantly different from that obtained with Ap5A exclusively (Fig. 1C and Fig. 5C).
The specific interaction between dinucleotide and epibatidine receptor has also been proved by the action of specific antagonists on ACh secretion, as shown in Fig. 5D. The ACh secretion increase mediated by preincubation with Ap5A can be selectively blocked by the dinucleotide receptor inhibitor Ip5I, as shown in Fig. 5D. On the other hand, the broad spectrum P2 antagonist, suramin, had no effect on the Ap5A or epibatidine responses.
Effect of Protein Kinase Inhibitors on the Epibatidine-Responding Nicotinic Receptor and Dinucleotide Receptor Interaction: Role of CaM Kinase II. Studies were carried out to understand the signaling pathway involved in the epibatidine inhibition of dinucleotide responses at the presynaptic level. As the epibatidine-sensitive nicotinic receptor located presynaptic level is able to induce a significant increase in the [Ca2+]i, the effect of some specific and broad spectrum PKC inhibitors were studied to try to revert the inhibition. As shown in Fig. 6A, the preincubation of synaptosomal populations with epibatidine results in a significant decrease of the Ap5A response. Besides, all the protein kinase inhibitors tested are able to increase in a variable extent the response of Ap5A alone, as it has been previously reported by our group, and the same occurs when epibatidine is present in the media (Díaz-Hernández et al., 2000; Gómez-Villafuertes et al., 2003). The PKC-specific inhibitors for the calcium-dependent (Gö 6976) or -independent (Gö 6983) isoforms, or the broad spectrum inhibitor staurosporine, were not able to revert the epibatidine inhibitory effect (Gschwendt et al., 1998). However, another inhibitor the Ro 31-8220, reported for being not only a good inhibitor of PKC but also for CaMKII, was able to revert the inhibitory action of epibatidine. These results pointed to a role for the CaMKII in synaptic terminals containing epibatidine and dinucleotide receptors altogether.
Effect of CaM Kinase II Inhibitors on the Nicotinic Receptor and Dinucleotide Receptor Interaction. To test the role of CaMKII in the intrasynaptosomal inhibitory effects, two specific enzyme inhibitors were used, the KN-62 and the KN-93. Both compounds had a similar behavior than Ro 31-8220 when studied in synaptosomal populations (data not shown), and for a more specific and accurate analysis, studies of their effects were undertaken in single isolated cholinergic terminals. Figure 6B shows a single synaptic terminal exhibiting epibatidine response and followed by an immediate administration of Ap5A. In 63% of the cases, there is no Ca2+ response to the subsequent administration of Ap5A or it is reduced by more than 90% (Fig. 6B). Another population, close two one-third of the total, did not exhibit a significant reduction on the Ca2+ signaling. It is noteworthy that all the terminals where the Ap5A response was abolished after epibatidine stimulation were able to respond to the dinucleotide after 60 s under washing by the perfusion media. In the presence of the CaMKII inhibitors (Fig. 6, C and D), epibatidine does not inhibit the immediate response to Ap5A in synaptosomes where both receptors are present. These results confirm the relevant role played by CaMKII on the regulation of Ca2+ responses mediated by presynaptic ionotropic dinucleotide receptors and agree with the events at the acetylcholine release under similar experimental situations.
Discussion
The present study confirms the existence of a fine-tuning interaction between dinucleotide receptor and the epibatidine-responding nicotinic receptor in rat midbrain cholinergic synaptic terminals. The relevance of these results relies on the fact that as much as one-third of all cholinergic terminals exhibit Ca2+ responses to Ap5A and to epibatidine altogether. Coapplication of both agonists results in a significant reduction on their specific Ca2+ responses and ACh release.
The abundance of cholinergic interneurons in mammalian midbrain, mainly in striatum, and the innervation of thalamic and subthalamic areas from the cholinergic nuclei present in the basal forebrain and diencephalon suggested this brain area as a good choice to study the events in depth at the cholinergic synaptic terminals (Perry et al., 1999). The costorage of acetylcholine together with nucleotides and dinucleotides has been demonstrated in cholinergic vesicles, and their simultaneous release to the synaptic cleft raises the question of their receptor's distribution and interaction at the synaptic level (Zimmermann, 1994). The existence of presynaptic nicotinic receptors, able to modulate the transmitter release, is firmly established (McGehee and Role, 1995; Wonnacott, 1997). Indeed, the presence of presynaptic ionotropic receptors for nucleotides and dinucleotides has been confirmed in synaptosomal populations and isolated synaptic terminals (Pintor and Miras-Portugal, 1995; Díaz-Hernández et al., 2001; Gómez-Villafuertes et al., 2001). The simultaneous presence of ionotropic nucleotidic and nicotinic receptors in individual terminals point out that some relationship can exist (Díaz-Hernández et al., 2002). Thus, this study mainly focuses on the response interaction between the dinucleotide receptor, responding to Ap5A, and the subset of nicotinic receptors responding to epibatidine, which contain mainly the α3 and α4 subunits that are very abundant in midbrain nuclei (Nayak et al., 2001; Whiteaker et al., 2002). From our studies and due to the scarce effect of α-bungarotoxin, the involvement of α7 subunit, although present, is very limited.
The [Ca2+]i responses elicited by dinucleotides or epibatidine are very similar in intensity and show no significant differences whether they coexist or not on the same cholinergic terminal. Moreover, the number of terminals responding only to Ap5A, or only to epibatidine, or to both compounds is quite similar and represents approximately one-third each from the total cholinergic population responding to these agonists. These data give an idea about the wide and extensive interaction between both neurotransmitter systems and the consequences for the physiological and pathological functioning in CNS. On the other hand, heterogeneity exists concerning the α4 subunit in cholinergic terminals, which is present in about two-thirds of the individual cholinergic terminals responding to epibatidine. The absence of α4 subunits in some of the epibatidine responsive cholinergic terminals agrees with the data reported by the Nichols group showing a clear segregation on α3 and α4 subunits in different striatal synaptic populations (Nayak et al., 2001). However, no changes in the Ca2+ response intensity have been observed with respect to the presence or absence of the α4 subunit. Mutation of the gene coding for this protein (CHRNA4) provided the first demonstration of human epilepsy caused by a genetically determined channel dysfunction (Avanzini and Franceschetti, 2003).
The fact that the responses to epibatidine can be abolished by mecamylamine or hexamethonium, but not modified by the dinucleotide receptor antagonist, confirms the presence of specific nicotinic receptors sensitive to this agonist (Shafaee et al., 1999). Similar behavior is obtained for the dinucleotide responses, which are not abolished by nicotinic antagonists or by the broad spectrum P2 antagonist, suramin, but exclusively by Ip5I, which at the concentration used, only blocks the dinucleotide receptor response in the synaptic terminals (Pintor et al., 1997; Díaz-Hernández et al., 2001).
The response to epibatidine and dinucleotide on the same individual cholinergic terminal makes the study of their receptor interaction possible. In all cases, when both receptors coexist, the coapplication of both agonists results in a significant decrease on the expected calcium signal. Moreover, when the terminals are sequentially challenged with epibatidine and dinucleotides, a minimum of a 1-min interval between both applications is required to restore their respective responses. The mutual negative control of these two ionotropic presynaptic receptors is perhaps the physiological way to turn off the synaptic excitability and prevent the excessive release of neurotransmitters. The reduction in the calcium entrance has been reported for other members of the nucleotidic and nicotinic receptor families when they are coexpressed together (Nakazawa, 1994; Searl et al., 1998; Zhou and Galligan, 1998). Some possible explanations for the mutually negative control have been proposed, such as the existence of a “channels overlap” (Nakazawa, 1994), the two channels could form a particular functional unit and would never be open at the same time (Barajas-López et al., 1998), or a dependent cross-inhibition state between transmittergated cation channel (Khakh et al., 2000). The last concerns a nicotinic receptor containing α3β4 subunits, which is also sensitive to epibatidine, and the P2X2 subunits of the nucleotidic receptors when both are expressed in Xenopus oocytes.
From the analysis of the data of acetylcholine secretion induced by Ap5A and epibatidine in synaptosomal populations, the negative interaction between receptors is confirmed. Moreover, from the inhibition studies, they appear to be independent receptors keeping their specificity for the pharmacological inhibitors. It is noteworthy that epibatidine is able to reduce by half the acetylcholine released by Ap5A and at concentrations related with its dose-response curve in Ca2+ response. This reduction means that the terminals where both receptors coexist are inhibited to a great extent by the epibatidine presence, and the secretory component induced by epibatidine only appears at much higher epibatidine concentrations mimicking its secretory dose-response curve but much more reduced in maximal effect, which is close to one-half of the total as expected.
As nicotinic and dinucleotide receptors, once present at the same terminal, establish a reciprocal modulation, the main question is to understand the biochemical components involved in their interaction. From the inhibitory studies on Ca2+ signaling and acetylcholine secretion, it becomes clear that they keep their functional and pharmacological identity so a nonpermanent modification mediated by some of the abundant presynaptic protein kinases was most likely. Previous work from the group has shown that dinucleotide presynaptic receptors are regulated by the action of protein kinases, mainly PKA, upon metabotropic receptors stimulation (Díaz-Hernández et al., 2000; Gómez-Villafuertes et al., 2003). Here, we report the cross talk between two ionotropic receptors as also the relevant role played by CaMKII. The reversion of the epibatidine-inhibitory effects on dinucleotide receptor response, by inhibiting the CaMKII, confirmed the role of this protein kinase on the reported effect.
The presence of CaMKII has been described at pre- and postsynaptic levels, although most of the relevant roles described so far are related with its postsynaptic actions. This is the case for the effect on synaptic plasticity and remodeling required in long-term potentiation and memory where they appear to play a role on the clustering of N-methyl-d-aspartate and other ionotropic receptors and their interaction with other proteins necessary to stabilize the postsynaptic structure (Fukunaga and Miyamoto, 2000; Soderling, 2000). At the presynaptic level, the CaMKII is associated with synaptic vesicles, where it phosphorylates synapsin I, allowing the movement of synaptic vesicles and further secretory events (Chapman et al., 1995; Bayer and Schulman, 2001). In Angelman′s syndrome, a disorder of human cognition characterized by severe mental retardation and epilepsy, a reduction in CaMKII activity has been reported (Weeber et al., 2003). From the present work it becomes evident that CaMKII plays a pivotal role allowing the biochemical communication between ionotropic presynaptic receptors, such as the epibatidine-sensitive nicotinic receptor and the dinucleotide receptor. The activation of this enzyme results in a significant decrease of the calcium entrance (Fig. 7) and tries to emphasize the powerful control of presynaptic responses mediated through CaMKII. It is to emphasize that in the present work three elements relevant in epileptic behavior, such as the nicotinic receptor, nucleotidic receptors, and CaMKII enzyme, appear for the first time working jointly on the same terminal (Avanzini and Franceschetti, 2003; Weeber et al., 2003).
Multiple temporarily regulated mechanisms are used to modulate the efficiency of synaptic transmission; from the present findings, there is a new one at the presynaptic level involving two ionotropic receptors, which when once costimulated, result in an inhibitory effect of both secretion and [Ca2+]i increase. The mechanisms involved for this control are due to the activation of CaMKII. It can be suggested that this type of interaction is able to moderate the excessive exocytotic release and the exocytotoxic effects of large amounts of intracellular Ca2+. In addition, it is relevant to emphasize that these facts have been found in cholinergic terminals, which number and functionality has been compromised in neurodegenerative diseases, as is the case for Alzheimer′s disease. In this pathology, there is not only a reduction in the cholinergic terminals, but also a significant decrease in nicotinic receptor containing the α4 subunits also at the presynaptic level (Maelicke and Alburquerque, 2000). What happens with the nucleotide and dinucleotide receptors mediated signaling at the presynaptic level in this situation needs to be accurately evaluated, but there is no doubt from the present study that cholinergic synapses offer much more possibilities of regulation and control than previously envisaged.
Footnotes
-
Financial support: BFI 2002-03626, Ministerio de Ciencia y Tecnología; CAM-08.5/0004/2003.
-
doi:10.1124/jpet.104.072249.
-
ABBREVIATIONS: CNS, central nervous system; ACh, acetylcholine; Ap5A, diadenosine pentaphosphate; CaMKII, calcium/calmodulin-dependent protein kinase II; TES, 2-{[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]amino}ethanesulfonic acid; AChE, acetyl cholinesterase; PKC, protein kinase C; Ro 31-8220, 2-{1-[3-(amidinothio)propyl]-1H-indol-3-yl}-3-(1-methylindol-3-yl)-maleimide; Gö 6983, 2-[1-(3-dimethylaminopropyl)-5-metoxyindol-3-yl]-3-(1H-indol-3-yl) maleimide; Gö 6976, 5,6,7,13-tetrahydro-13-methyl-5-oxo-12H-indolo[2,3-a]pyrrolo[3,4-c]carbazole-12-propanenitrile; KN-93, N-(2-[N-[4-chlorocinnamyl]-N-methylaminomethyl]phenyl)-N-(2-hydroxyethyl)-4-methoxybenzenesulphonamide; KN-62, 2-[N-(4′-methoxybenzenesulfonyl)]amino-N-(4′-chlorophenyl)-2-propenyl-N-methylbenzylamine phosphate; AM, acetoxymethyl ester; MOPS, 4-morpholinepropanesulfonic acid; PFA, paraformaldehyde; PBS, phosphate-buffered saline; BSA, bovine serum albumin; PSD, postsynaptic density; GFAP, glial fibrillary acidic protein; VAT, vesicular acetylcholine transporter; ANOVA, analysis of variance; FITC, fluorescein isothiocyanate; Ip5I, diinosine pentaphosphate.
- Received June 3, 2004.
- Accepted July 9, 2004.
- The American Society for Pharmacology and Experimental Therapeutics