点击显示 收起
Institut de Pharmacologie et de Biologie Structurale, Centre National de la Recherche Scientifique, Unitee Mixte Recherche 5089, Toulouse, France
Abstract
To elucidate the mechanism of the cellular antiopioid activity of neuropeptide FF (NPFF), we have transfected the SH-SY5Y neuroblastoma cell line, which expresses e?and -opioid receptors, with the human NPFF2 receptor. The selected clone, SH2-D9, expressed high-affinity NPFF2 receptors in the same range order as e? and -opioid receptors (100eC300 fmol/mg of protein). The NPFF analog [D-Tyr1, (NMe)Phe3]NPFF (1DMe) did not modify the binding parameters of the e? and -specific agonists [D-Ala2,N-Me-Phe4,Gly5-ol]-enkephalin and deltorphin-I, respectively. 1DMe dose dependently inhibited 75 to 80% of the cAMP production stimulated by forskolin. Preincubation with 1DMe halved the maximal inhibition of N-type Ca2+ channels by opioid agonists. In the presence of carbachol, acting on muscarinic receptors to release Ca2+ from the intracellular stores, deltorphin-I and 1DMe enhanced this release. Preincubation with 1DMe reduced the maximal effect of deltorphin-I by 40%, demonstrating an antiopioid effect in this experimental model for the first time. By using peptides corresponding to the carboxyl terminus of the i1,2, i3, o, and s subunits of G proteins, which specifically uncouple receptors from G proteins, we demonstrated that e?opioid and NPFF2 receptors couple to the four subunits assayed. The Ca2+ release from the intracellular stores by 1DMe resulted from the coupling of NPFF2 receptors with Go and Gi1,2, whereas the coupling with Gs reduced the antiopioid effect of 1DMe in the modulation of N-type channels. This SH2-D9 cell line now provides the opportunity to study the interaction between both receptors.
Neuropeptide FF (NPFF), FLFQPQRFamide, is representative of a family of mammalian amidated neuropeptides whose precursors pro-NPFFA and pro-NPFFB and G protein-coupled receptors NPFF1 and NPFF2 have been recently cloned (Zajac, 2001). Although NPFF does not interact with opioid receptors (Gouardeeres et al., 1998), a close relationship between neuropeptide FF and opioid systems has been clearly demonstrated in the central nervous system, especially in pain perception (Harrison et al., 1998; Roumy and Zajac, 1998; Panula et al., 1999). For instance, supraspinal injection of NPFF analogs, which has little or no effect in pain tests, decreases morphine-induced analgesia (antiopioid activity, Roumy and Zajac, 1998; Panula et al., 1999), whereas spinal administration induces a naloxone-sensitive analgesia and potentiates morphine-induced analgesia (proopioid activity, Roumy and Zajac, 1998; Panula et al., 1999).
In neurons, opioids, including nociceptin, 1) inhibit adenylyl cyclase activity and 2) stimulate inwardly rectifying K+ channels and inhibit voltage-dependent Ca2+ currents by activating four types, e? , , and Opioid Receptor-Like 1, of opioid receptors (Law et al., 2000). This leads to postsynaptic neuronal inhibition and to presynaptic inhibition of transmitter release, respectively. In contrast to opioids, no data report a direct modulation of K+ and Ca2+ conductances by NPFF or analogs but rather describe a blockade of the opioid activity on ionic conductances when cells are pretreated with NPFF. NPFF or analogs, which are inactive by themselves, reverse the opioid-induced inhibition of Ca2+ conductance in NPFF2 receptor-expressing neurons dissociated from rat dorsal root ganglion (Rebeyrolles et al., 1996), dorsal raphe (Roumy and Zajac, 1999), and NPFF1-expressing neurons from the hypothalamic periventricular nucleus (Roumy et al., 2003). This functional antagonism also has been observed by others in different models such as acetylcholine release in the myenteric plexus of the guinea pig (Takeuchi et al., 2001), electrical response in hippocampus slices (Miller and Lupica, 1997), or Met-enkephalin release in the spinal cord (Ballet et al., 1999). In this latter case, blockade of presynaptic -opioid autoreceptors after activation of NPFF receptors leads to enhanced release of Met-enkephalin that activates e?opioid receptors (Mauborgne et al., 2001). In addition, it concomitantly reduces the level of pronociceptive dynorphin (Ballet et al., 2002). Therefore, such an antiopioid activity, paradoxically, could account for the spinal opioid-dependent analgesia induced by NPFF.
Importantly, in experiments on isolated neurons, the effect of NPFF was restricted to opioid receptors and not to other G protein-coupled receptors, although involved in the same transduction pathway as, for example, 5-HT1A receptors in the rat dorsal raphe nucleus (Roumy and Zajac, 1999). This suggests that the NPFF-induced inhibition is specific to opioid receptors and is not mediated by the spread out of an intracellular messenger or by a modification of the effector such as the phosphorylation of the voltage-gated Ca2+ channel. The mechanism of this cellular antiopioid effect is unknown and merits to be studied for bringing new information on G protein-coupled receptor signaling regulation.
To elucidate this mechanism, a model reproducing the cellular antiopioid effect of NPFF has been established by transfecting the neuroblastoma cell line SH-SY5Y with the human NPFF2 (hNPFF2) receptor, which is the most extensively characterized NPFF receptor. SH-SY5Y cells were chosen because they naturally express opioid (e? , and Opioid Receptor-Like 1) receptors that inhibit voltage-dependent N-type Ca2+ channels (for review, see Vaughan et al., 1995). We present here the characterization of this cell line (SH2-D9), with the demonstration that NPFF2 receptors exert a functional antiopioid activity similar to that previously observed on isolated neurons. Furthermore, we demonstrate, for the first time, that NPFF2 receptors antagonize the opioid-induced potentiation of Ca2+ release from intracellular stores triggered by activation of Gq-coupled muscarinic receptors. Therefore, this cell line constitutes a model of choice to study the interactions between opioid and NPFF receptors.
Materials and Methods
Materials. NPFF-related peptides [1DMe, [D-Tyr1, (NMe)Phe3]NPFF; SQA-NPFF, SQAFLFQPQRFa; EYF, EYFSLAAPQRFa], deltorphin-I (Tyr-D-Ala-Phe-Asp-Val-Val-Gly-NH2), Cys-dermorphin (Tyr-D-Ala-Phe-Gly-Tyr-Pro-Lys-Cys-NH2), and the peptides analogs to the C-terminal regions of the subunits of G protein i1,2 (345eC354, [C]KNNLKDCGLF), i3 (345eC354: [C]KNNLKECGLY), o (345eC354, [C]ANNLRGCGLY), and s (385eC394, [C]RMHLRQYELL) were synthesized with an automated peptide synthesizer (model 433A; Applied Biosystems, Foster City, CA). The integrity of peptides was confirmed by mass spectrometry analysis. [D-Ala2,N-Me-Phe4,Gly5-ol]-enkephalin (DAMGO) was purchased from Bachem (Bubendorf, Switzerland), levorphanol was from F. Hoffman-La Roche (Paris, France), clonidine was from Sigma/RBIR (Natick, MA) and -conotoxin GVIA was from Alomone Labs (Jerusalem, Israel). Pertussis toxin (PTX), cholera toxin (CTX), H89, 1,2-dimethoxy-N-methyl(1,3)benzodioxolo(5,6-c)phenanthridinium chloride (chelerythrine), and Ro 31-8220 were purchased from Sigma (Lyon, France). Porcine NPY was a generous gift from Dr. R. Quirion (McGill University, Montreeal, QC, Canada). [3H]DAMGO (67 Ci/mmol) and [3H]adenine (26 Ci/mmol) were purchased from Amersham Biosciences Inc. (St. Quentin Fallavier, France). 125I-EYF and 125I-deltorphin-I were iodinated by electrophilic substitution, as described previously (Gouardeeres et al., 2001).
Cell Culture. Human neuroblastoma SH-SY5Y cells were kindly provided by F. Noble (Universitee Renee Descartes, Paris, France). Cells were grown in Dulbecco's modified Eagle's medium (4.5 g/l of glucose; GlutaMAX I) containing 10% fetal calf serum and 50 e蘥/ml gentamicin (Invitrogen, Cergy Pontoise, France), in a humidified atmosphere containing 5% CO2. SH-SY5Y cells were transfected with the human NPFF2 receptor subcloned into the bicistronic vector pEFIN3 (Kotani et al., 2001), by using FuGENE 6 according to the manufacturer's instructions (Roche Diagnostics, Indianapolis, IN). Recombinant cells were selected by addition of 400 e蘥/ml G418 (Geneticin; Invitrogen) to the culture medium. The SH2-D9 clone, obtained by limit dilution, was chosen for all experiments. Wild-type and transfected cells were used undifferentiated between passages 8 to 20.
Binding on Cell Membrane. Membrane preparation was obtained as described previously (Mollereau et al., 2002). Binding of 125I-EYF, a high-affinity NPFF2 receptor agonist (Gouardeeres et al., 2001), was performed in polypropylene tubes in a final volume of 0.5 ml containing 5 to 15 e蘥 of protein, 50 mM Tris-HCl, pH 7.4, 60 mM NaCl, 25 e bestatin (Sigma), 0.1% bovine serum albumin (BSA), and the radioligand at the desired concentration. The nonspecific binding was determined in the presence of 1 e EYF. For the binding of 125I-deltorphin-I, a specific -opioid agonist, 10 e levorphanol was used to determine the nonspecific binding, and the incubation buffer was 50 mM Tris-HCl, pH 7.4, 0.1% BSA, and the radioligand at the desired concentration. Binding of [3H]DAMGO, a specific e?opioid receptor agonist, was performed in polypropylene tubes in a final volume of 1 ml containing 100 e蘥 of protein, 50 mM Tris-HCl, pH 7.4, and the radioligand at the desired concentration. The nonspecific binding was determined in the presence of 10 e levorphanol.
After a 60-min incubation at 25°C, samples were rapidly filtered on Whatman GF/B filters preincubated in 50 mM Tris-HCl, pH 7.4, 0.1% BSA for the binding of 125I-EYF and 125I-deltorphin-I and 0.3% polyethylenimine for the binding of [3H]DAMGO. The filters were rinsed three times with 4 ml of ice-cold 10 mM Tris-HCl, pH 7.4, containing 0.1% BSA for 125I-EYF and 125I-deltorphin-I. The bound radioactivity was measured with a gamma-counter (PerkinElmer Life and Analytical Sciences, Boston, MA) for 125I-EYF and 125I-deltorphin-I and with a liquid scintillation analyzer (PerkinElmer Life and Analytical Sciences) for [3H]DAMGO.
Binding on Living Cells. Confluent cells, grown in 100-mm dishes, were rinsed and collected in ice-cold HEPES buffer (10 mM HEPES, pH 7.3, 150 mM NaCl, 2.5 mM KCl, 10 mM glucose, and 0.1% BSA). After centrifugation at 1000 rpm for 1 min at 4°C, cells were recovered in 2 ml of buffer. The cell suspension was divided and incubated for 20 min at room temperature with, or without, 1 e 1DMe. The cell suspension (0.2 ml) was then added to polypropylene tubes in a final volume of 0.5 ml of HEPES buffer, containing 1 nM 125I-deltorphin-I and 10 e levorphanol for determination of non-specific binding. After a 90-min incubation at 25°C, samples were rapidly filtered on Whatman GF/B filters preincubated in HEPES buffer. The filters were rinsed three times with 4 ml of ice-cold HEPES buffer, and the bound radioactivity was counted.
Measurement of cAMP. Cells (4eC5 x 105) were seeded into 24-well plates and incubated for 24 h. The culture medium was then replaced by 0.3 ml of fresh medium containing 0.1 e adenine and 1 e藽i of [3H]adenine. After a 60-min incubation at 37°C under a 5% CO2 in air atmosphere, the cells were rinsed twice with 0.5 ml of Krebs-Ringer-HEPES (KRH) (124 mM NaCl, 5 mM KCl, 1.25 mM MgSO4, 1.5 mM CaCl2, 1.25 mM KH2PO4, 25 mM HEPES, 8 mM glucose, and 0.5 mg/ml BSA, pH 7.4). Prewarmed KRH (0.2 ml) was added to each well, and the reaction was initiated by addition of 0.1 ml of KRH containing 15 e forskolin (Sigma), 0.3 mM 3-isobutyl-1-methylxanthine (Sigma), 0.3 mM Ro 20-1724 (Fisher, Illkirch, France) and the agonist to be tested. After 10 min at 37°C, the reaction was stopped by addition of 0.03 ml of 2.2 N HCl. The [3H]cAMP content of each well was isolated by chromatography on acid alumina (Sigma) columns (Mollereau et al., 2002).
Measurement of [Ca2+]i. SH2-D9 or SH-SY5Y cells were seeded at 3 to 5 x 105 cells in 35-mm Petri dishes and cultured for 24 or 48 h. On the day of the experiment, the culture medium was replaced with 1 ml of HEPES-buffered medium (without BSA), and the cells were incubated for 10 min. The medium was changed to 0.7 ml of HEPES buffer (without BSA) containing 3.6 e Fluo-4 AM (Molecular Probes, Leiden, The Netherlands) and 0.1% Pluronic acid F127 (Sigma), and the cells were incubated for 30 min in the dark at 37°C. The medium was then replaced with 1 ml of HEPES medium containing 0.1% BSA (Euromedex, Souffelweyersheim, France), and the cells were further incubated for 30 min at room temperature in the dark to allow for complete Fluo-4 AM deesterification. All subsequent perfusing media contained 0.1% BSA.
Cells were viewed with a 40/0.65 objective, illuminated at 488 nm (10-nm bandwidth interference filter), and imaged with a cooled charge-coupled device camera (MicroMax 782 Y; Princeton Instruments, Evry, France) driven by MetaView software (Universal Imaging Corporation, Downingtown, PA). The average pixel intensity within user-defined regions of interest was measured and saved on a computer hard drive. To achieve a rate of measurement of 1 per second (with a 500-ms exposure time set by a Uniblitz shutter), 6 x 6to9 x 9 binning of pixels were used. The fluorescence intensity was expressed as F/F0, where F and F0 are the fluorescence intensity at any time and the mean resting fluorescence intensity preceding the first stimulation, respectively. In many cells, there was a slow linear decrease of F0 as a function of time, and not as a function of previous illumination, that probably represents Fluo-4 leakage from the cells. This was corrected for by using the parameters of the linear regression of F0 against time.
The cells were gravity perfused at a rate of 1.8 ml/min with a medium containing 150 mM NaCl, 2.5 mM KCl, 2 mM CaCl2, 1 mM MgCl2, 10 mM glucose, and 10 mM HEPES. The pH was adjusted to 7.30 and 0.1% BSA was added.
Modulation of N-Type Ca2+ Channels. To study the modulation of N-type Ca2+ conductance, the L-type Ca2+ channels were blocked by adding 3 e nifedipine in all perfusing media. Two Petri dishes seeded at the same cell density and cultured for the same duration were used. In the control culture, cells were returned to HEPES-buffered medium after incubation with Fluo-4, whereas for the test culture, this medium contained 1 to 1000 nM 1DMe, which also was added to all perfusing media. The cells were depolarized during 5 or 10 s with a 140 mM KCl medium (NaCl reduced to 12.5 mM). An electrically operated valve effected a rapid change of solution that was complete within 8 s. The response to a 5- or 10-s depolarization was recorded. The response to the same depolarization was recorded 5 min later, at the end of a 30-s perfusion of DAMGO, deltorphin-I, NPY, or clonidine. Five minutes later, a third response to the same depolarization was measured to test for recovery of the calcium transient. Only the cells that recovered from the opioid (or the other modulators) were included in the present results. For each Petri dish, the responses were recorded in two distinct microscope fields. Each experiment was performed on at least two different cultures. The reduction (%) of the depolarization-induced [Ca2+]i transient by DAMGO was calculated as 100 x [(F/F0)DAMGO eC (F/F0)C]/(F/F0)C, in which (F/F0)C = ((F/F0)max eC 1)C and (F/F0)DAMGO = ((F/F0)max eC 1)DAMGO are the magnitude of the fluorescence increase during depolarization in the control test and after 30-s perfusion with DAMGO, respectively.
In experiments designed to test the effects of protein kinase A or C inhibitors on the antiopioid activity of 1DMe , the control culture was returned to HEPES medium plus BSA plus the protein kinase inhibitor for 45 min, after Fluo-4 AM incubation, before perfusion was started. For the test culture, 100 nM 1DMe were added after 15 min of incubation with protein kinase inhibitors, and the cells were further incubated for 30 min. For both the control and test cultures all perfusing media contained the kinase inhibitor. Two inhibitors of protein kinase C were used: 1 e chelerythrine and 0.1 e Ro 31-8220. H89 (1 e), a protein kinase A inhibitor, was also tested.
Interaction between Muscarinic Receptors and Opioid and/or NPFF2 Receptors. For each experiment, two 35-mm culture dishes were used: one served as control and the other as the test culture. Regions of interest were laid over the all cell bodies present in the field. Unless otherwise indicated, the cells were perfused with 1 e deltorphin-I for 40 s and/or 1 e 1DMe for 40 s in the continuing presence of 5 e carbachol. Because all cells in a field were stimulated by carbachol, the responses were represented by the mean ± S.E.M. of F/F0 for all the cells, measured every second. However, the increase in fluorescence induced by deltorphin-I or 1DMe was measured in each cell.
To study a possible antiopioid effect of 1DMe on the response to deltorphin-I in the presence of carbachol, the cells were returned, after incubation with Fluo-4 AM, to either HEPES buffer or HEPES buffer plus 100 nM 1DMe, and all the perfusing media contained 100 nM 1DMe. The response to deltorphin-I (1eC1000 nM, 40 s) in the presence of 5 e carbachol was measured in control and 1DMe-treated cells. An identical experiment was performed in which the roles of deltorphin-I and 1DMe were reversed.
Identification of G Proteins Coupled to Opioid and NPFF2 Receptors in SH2-D9 Cells. Peptides corresponding to the last 10 C-terminal residues of G protein subunits were tested for their ability to inhibit the specific binding of radiolabeled agonists and to block 1DMe and opioid responses in intracellular Ca2+ and cAMP assays. For functional assays, peptides (6 or 60 e蘥 in 5 e蘬 of water) were introduced in SH2-D9 cells (1.5eC3 x 106) by the Nucleofactor technology (Amaxa Biosystems, Cologne, Germany) according to the manufacturer's instructions. After electroporation, cells were rapidly plated in 35-mm dishes for Ca2+ imaging or in 24-well plates (5 x 105 cells/well) for cAMP and left to recover for 4 h in a humidified 37°C, 5% CO2 incubator. To assess peptide delivery efficiency, cells were electroporated in the presence of 8 e蘥 of dermorphin labeled with the fluorophore Alexa 488. Cells were left to recover for 4 h, washed four times, and observed with a fluorescence microscope.
Because electropermeabilization technique could not be applied for Ca2+ conductance measurement, 1.5 e蘥 of Gs peptide was delivered into the cells with the Chariot transfection reagent (Active Motif, Rixensart, Belgium). Complexes were allowed to form for 30 min at room temperature and were then added to 50 to 70% confluent cells in 35-mm dishes. After a 2-h incubation according to manufacturer's instruction, the effect of 1DMe on the opioid-induced Ca2+ transient inhibition was investigated.
Data Analysis. Binding experiments and dose-response relationships were analyzed with Prism software (GraphPad Software Inc., San Diego, CA) using the appropriate functions. Statistical comparisons of more than two samples were done with one-way ANOVA followed by post hoc tests. Comparisons of two samples were done with the appropriate t test. In all cases, the level of significance was chosen as 0.05.
Results
Characterization of hNPFF2-Transfected SH-SY5Y Cells (SH2-D9 Clone). In SH-SY5Y membrane preparation, no specific binding of 125I-EYF could be detected. In membranes from the SH2-D9 clone, 125I-EYF labeled one class of high-affinity sites (Fig. 1A; Table 1). Significantly, NPFF2 receptors were expressed in the same range order as e? and -opioid receptors (Table 1). Preincubation of SH2-D9 membranes, for 20 min with 1 e 1DMe, did not change significantly the binding parameters of [3H]DAMGO and 125I-deltorphin-I (Table 1). In addition, the specific binding of 1 nM 125I-deltorphin-I in living SH2-D9 cells was not affected by preincubation with 1 e 1DMe (20 min at room temperature), being 111 ± 19% (n = 3) of the control.
In SH2-D9 cells, binding experiments also were performed on membranes preincubated (1DMe) for 20 min at room temperature with 1 e 1DMe. Data are means ± S.E.M. of three to four experiments performed in duplicate.
In SH-SY5Y and in SH2-D9 cells, DAMGO inhibited 50% of the forskolin-induced cAMP production, with an EC50 of 529 ± 274 (n = 5) and 1114 ± 217 nM (n = 4), respectively (Fig. 1B). Likewise, in SH2-D9 cells, 1DMe dose dependently inhibited with higher efficacy (75eC80%) and potency (EC50 = 0.8 ± 0.1 nM, n = 6) the intracellular cAMP production (Fig. 1). It was totally inactive in SH-SY5Y cells (Fig. 1). When SH2-D9 cells were preincubated with 1DMe (100 nM, 15 min, 37°C) and washed three times, no modification of the response to 10 e DAMGO was observed (Fig. 1). The inhibition of adenylate cyclase by 1DMe and DAMGO was totally prevented by overnight PTX pretreatment (100 ng/ml), indicating that both e?opioid and NPFF receptors mediated cAMP inhibition through a Gi/o protein coupling.
Effect of NPFF2 Receptor Activation on the Modulation of N-Type Ca2+ Channels by Opioids in SH2-D9 Cells. In the presence of nifedipine to block the L-type Ca2+ channels, some SH-SY5Y and SH2-D9 cells responded to depolarization (140 mM K+, 5 or 10 s) by an increase in F/F0 (Fig. 2A). This response was totally and irreversibly suppressed (n = 14) after incubation with -conotoxin GVIA (1 e, 90 s), a highly specific antagonist of N-type calcium channels. When the depolarization was repeated at the end of a 30-s perfusion with 1 e DAMGO, the magnitude of the [Ca2+]i transient was reduced in all cells, compared with the control depolarization (Fig. 2A). The reduction of the F/F0 transients by DAMGO was reversible and could be repeated at 10-min intervals without any sign of desensitization (Fig. 2A). The magnitude of the reduction was the same with 5- or 10-s depolarizations (eC44.5 ± 1.2, n = 104 and eC45.5 ± 2.0%, n = 67, respectively; p > 0.05, unpaired t test) and was identical in SH2-D9 and SH-SY5Y cells (eC44.9 ± 1.1%, n = 171 versus eC48.0 ± 1.8%, n = 62, respectively; p > 0.05, unpaired t test). The reduction of the F/F0 transients by DAMGO was dose-dependent with an EC50 value of 1.7 nM (Fig. 2B).
In SH2-D9 cells preincubated (30 min) and perfused with 100 nM 1DMe, there was no change in either the resting [Ca2+]i or the magnitude of the response to depolarization: (F/F0) = 2.233 ± 0.132 (n = 123) versus 2.227 ± 0.167 (n = 110) in control cells (p > 0.05, unpaired t test). However, the reduction of (F/F0) by DAMGO was less after 1DMe than in control cells (Fig. 3A). The dose-response curve of DAMGO in the presence of 100 nM 1DMe (Fig. 2B) demonstrated that the EC50 value of DAMGO was unchanged (3.1 versus 1.7 nM in control cells) but that its maximal effect was approximately halved (eC53.6% of the control) (Fig. 2B). By varying the 1DMe concentration and maintaining the DAMGO concentration constant at 1 e, we found that 1DMe reduced the inhibitory effect of DAMGO on (F/F0) with an EC50 value of 4.2 nM and a maximal reduction of 40% (Fig. 3B). In SH-SY5Y cells, 100 nM 1DMe did not modify the magnitude of the inhibition of (F/F0) by DAMGO (Fig. 3B).
In SH2-D9 cells, 100 nM SQA-NPFF, the 11-amino acid peptide contained in the human pro-NPFFA precursor, reduced the DAMGO-induced decrease in (F/F0) by the same amount as 100 nM 1DMe (eC65.96% ± 6.04, n = 46 versus eC59.95% ± 3.52, n = 67, respectively; p > 0.05, unpaired t test). In SH-SY5Y cells, 100 nM SQA-NPFF had no effect on the inhibition of (F/F0) by DAMGO.
To test for the possible involvement of protein kinases A and C in the antiopioid effect of 1DMe , two inhibitors of protein kinase C, 1 e chelerythrine and 1 e Ro 31-8220, and one inhibitor of protein kinase A, 0.1 e H89, were used. Neither chelerythrine nor H89 affected the responses to 1 e DAMGO in the absence and presence of 100 nM 1DMe (not illustrated). Ro 31-8220 strongly reduced the response to 1 e DAMGO in the absence of 1DMe so that it was difficult to assess whether 1DMe retained its antiopioid effect in presence of the inhibitor (data not shown).
SH-SY5Y cells also express -opioid, NPY Y2, and 2-adrenergic receptors (for review, see Vaughan et al., 1995). In SH2-D9 cells, 1 e deltorphin-I (30 s), 300 nM porcine NPY (30 s), and 10 e clonidine (30 s) reduced the calcium transients triggered by depolarization (Fig. 4), although to a lesser extent than DAMGO (p < 0.05, one-way ANOVA followed by Bonferroni's multiple comparison test). In each case, 100 nM 1DMe significantly (p < 0.05) reduced the inhibitory effect of these agonists on N-type calcium conductance (Fig. 4).
Interactions of -Opioid and NPFF2 Receptors with Muscarinic Receptors. In SH-SY5Y cells, opioid agonists enhance the release of Ca2+ from the intracellular stores triggered by carbachol acting on muscarinic receptors (Connor and Henderson, 1996). In all SH2-D9 cells, perfusion with 5 e carbachol produced an immediate increase in F/F0 that decayed slowly toward a new steady state (Fig. 5, A and B) and that was not abolished by the removal of extracellular Ca2+ (not illustrated). Perfusions with deltorphin-I or DAMGO alone increased F/F0 in less than 5% of the cells, whereas 1DMe alone never produced any change.
In the presence of 5 e carbachol, 1 e deltorphin-I and 1 e DAMGO (40 s) increased F/F0 in all the cells. The response to deltorphin-I was larger than that to DAMGO: (F/F0) = 3.59 ± 0.15 (n = 84) versus 0.85 ± 0.03 (n = 90), respectively (p < 0.05, unpaired t test). In view of this difference in magnitude, deltorphin-I was used in subsequent experiments. The NPFF agonist 1 e 1DMe (40 s) also caused an increase in F/F0 in all the cells, although of lesser magnitude than deltorphin-I: (F/F0) = 2.15 ± 0.16 versus 3.59 ± 0.15, respectively (n = 84, p < 0.05, paired t test). When applied for longer periods, in the presence of 5 e carbachol, the time courses of the effects of deltorphin-I and 1DMe were different. During perfusion with 1 e deltorphin-I, F/F0 decreased to its predeltorphin-I level within 100 s (Fig. 5A), and upon switching back to carbachol alone, no change in F/F0 was recorded (Fig. 5A). During perfusion with 1 e 1DMe, the F/F0 decrease was slower than with deltorphin-I (Fig. 5B), and F/F0 remained above its pre-1DMe level during the 5 min of perfusion (Fig. 5B). As a consequence, upon switching back to carbachol alone a clear and fast decrease in F/F0 was recorded (Fig. 5B). Removing extracellular Ca2+ did not alter the responses to deltorphin-I or 1DMe in the presence of carbachol (not illustrated). In SH-SY5Y cells, perfusion with 1 e 1DMe (40 s) plus 5 e carbachol caused no change in F/F0, whereas 1 e deltorphin-I (40 s) produced the same increase as in SH2-D9 cells (not illustrated).
To test whether 1DMe could have an antiopioid effect on the response to deltorphin-I, we measured the dose-response relationship of 1 to 1000 nM deltorphin-I (40 s) in the continuing presence of 5 e carbachol in control SH2-D9 cells and in cells pretreated for 30 min and perfused with 100 nM 1DMe. In the presence of 100 nM 1DMe, the increase in F/F0 in response to deltorphin-I was reduced (Fig. 6A). The EC50 value for deltorphin-I was not significantly altered (5.3 versus 7.1 nM in control cells), but the maximal effect was reduced by 40% (Fig. 6B), indicating that the inhibition was not competitive. When SH2-D9 cells were preincubated with 100 nM deltorphin-I before perfusion with media containing 100 nM deltorphin-I, the response to 1 e 1DMe (40 s) in the presence of carbachol was identical to that of control cells [(F/F0) = 2.74 ± 0.20, n = 37 versus 2.23 ± 0.23, n = 19, in control cells; p > 0.05, unpaired t test]. Thus, the interaction between opioid and NPFF2 receptors is not reciprocal.
Identification of G Proteins Associated with Opioid and NPFF2 Receptors in SH2-D9 Cells. Peptides corresponding to the carboxyl terminus of the G subunits of G proteins, which represent an important site of interaction with the receptor, have been reported to specifically uncouple receptors from G proteins in several systems, such as adenosinergic, adrenergic, and serotonergic, leading either to a low- or high-affinity state of the receptors (Gilchrist et al., 1998; Chang et al., 2000; Mazzoni et al., 2000). The Gi1,2 Gi3, Go, and Gs inhibitory peptides were therefore tested for their ability to modulate the specific binding of 125I-EYF and [3H]DAMGO (Fig. 7). The Gi1,2,Gi3, and Go synthetic peptides inhibited (maximum 25eC30%) the binding of [3H]DAMGO. In contrast, Gi1,2 peptide displayed a strong dose-dependent inhibition (up to 75% at 300 e) of the specific binding of 125I-EYF. A smaller inhibition (50% at 300 e) was observed for the Go peptide, whereas a slight (25%) increase in binding was recorded for Gi3. Surprisingly, the Gs peptide was the most potent inhibitor of both NPFF and opioid receptor binding. It was as efficacious (70% inhibition at 300 e) as the Gi1,2 peptide to inhibit the 125I-EYF-specific binding. Although the effect was smaller than with 125I-EYF, it produced the greatest reduction of the [3H]DAMGO-specific binding. Thus, NPFF and opioid receptors interact with all G subunits assayed.
To characterize more precisely the functional association of G proteins with opioid and NPFF receptors, inhibitory peptides were then tested for their ability to block cellular responses. In SH2-D9 cells that were electroporated in the presence of Alexa 488 dermorphin (8 e蘥), 77% of the cells were fluorescent, whereas no fluorescence was detected in cells incubated with the labeled peptide but not electroporated (not illustrated). Electroporation was thus an efficient method to deliver the G peptides into SH2-D9 cells, although differences in loading efficiency between the different peptides could not be excluded.
SH2-D9 cells were electroporated in the absence and presence of 50 or 500 e Gi1,2, Gi3, or Go peptides, because 100 ng/ml PTX (18 h) suppressed the responses to 1 e deltorphin-I (40 s) and 1 e 1DMe (40 s) in the presence of 5 e carbachol (not illustrated). As shown in Fig. 8, the response to deltorphin-I was reduced by Gi1,2 and Go peptides at 50 and 500 e, whereas Gi3 only was efficient at 500 e. The activation of the -opioid receptor in the presence of carbachol released Ca2+ from intracellular stores through a preferential coupling with Gi1,2, Go, and to a lesser extent to Gi3. The response to 1DMe was reduced by the Go peptide at 50 and 500 e and by the Gi1,2 peptide at 500 e. The Gi3 peptide, at 500 e concentration, caused a nonsignificant reduction of the response to 1DMe (Fig. 8), indicating that the activation of the NPFF2 receptor, in the presence of carbachol, released Ca2+ from intracellular stores through a preferential coupling with Go and to a lesser extent with Gi1,2.
The inhibitory effect of DAMGO on N-type Ca2+-channel was reduced to 24% of its control value (p < 0.05, unpaired t test) after pretreatment with 100 ng/ml PTX (20eC22 h), precluding the measurement of the antiopioid effect of 1DMe. Only we could test the effect of the Gs peptide, after delivery within the cells by the Chariot peptide carrier: the Gs peptide did not change the inhibitory effect of DAMGO, but increased the antiopioid activity of 1DMe (Fig. 9). To confirm the involvement of Gs, SH2-D9 cells were incubated overnight with 500 ng/ml CTX. This resulted in a very low density of cells that were unresponsive to depolarization, so that it was not possible to measure the antiopioid activity of 1DMe.
To exclude a nonspecific effect of the Gs peptide, the inhibition of the forskolin-stimulated production of cAMP by 1 nM 1DMe was compared in SH2-D9 cells electroporated in the presence of Gi1,2 or Gs peptides (60 e蘥). The effect of 1 nM 1DMe in cells electroporated with Gi1,2, was reduced to 74.3 ± 5.8% (n = 4) of that measured in the absence of peptide (p < 0.05, unpaired t test), whereas it was not modified in cells electroporated with Gs (108.7 ± 6.4% of control cells, n = 3, p > 0.05, unpaired t test).
Discussion
We describe in the present study a cellular model (human NPFF2 receptor transfected SH-SY5Y cells: SH2-D9) that reproduces the functional antagonism exerted by NPFF on opioid activity in neurons, thus allowing the possibility to explore the molecular mechanisms responsible for this process.
SH2-D9 cells express high-affinity NPFF2 receptors with a Bmax of the same order of magnitude (100eC300 fmol/mg) as for e? and -opioid receptors (Kazmi and Mishra, 1987; Toll et al., 1997; Noble et al., 2000; this study). Activation of both e?opioid and NPFF2 receptors inhibit adenylate cyclase, in a PTX-sensitive way, as already demonstrated in SH-SY5Y cells for opioid receptors (Kazmi and Mishra, 1987) and in NPFF2-transfected Chinese hamster ovary cells (Kotani et al., 2001). We have, therefore, obtained a neuroblastoma cell line expressing high-affinity and functional opioid and NPFF2 receptors. We demonstrate that activation of NPFF2 receptors, in this cell line, exerts an antiopioid effect in two experimental paradigms: modulation of N-type voltage-activated Ca2+ channels as observed in neurons (Roumy and Zajac, 1999) and coincident signaling between Gq-coupled receptors (muscarinic) and opioid receptors.
In all SH2-D9 cells responding to depolarization in the presence of nifedipine (a blocker of the L-type Ca2+ channels), DAMGO, a highly specific e?opioid agonist reduces the magnitude of the depolarization-induced increase in [Ca2+]i. In cells preincubated with an agonist of NPFF2 receptors (1DMe), the effect of DAMGO on the Ca2+ transient is reduced noncompetitively, with no change in the EC50 value and a reduction of 40 to 50% of its maximal effect. This antiopioid effect is caused by the activation of the NPFF2 receptor, because it is absent in nontransfected SH-SY5Y cells. It is identical to that demonstrated in dorsal root ganglion (Rebeyrolles et al., 1996) and dorsal raphe neurons (Roumy and Zajac, 1999). In contrast to the specificity toward opioid receptors established in neurons (Roumy and Zajac, 1999) and in in vitro models (Takeuchi et al., 2001), the activation of NPFF2 receptors in SH2-D9 cells also reduces the inhibition of N-type Ca2+ channels induced by NPY Y2 and 2-adrenergic receptors. It should be recognized, however, that the interaction of NPFF2 receptors with other G protein-coupled receptors has not been exhaustively studied in isolated neurons and thus it cannot be entirely excluded that the activation of NPFF2 receptors could antagonize the activity of receptors other than the opioids. On the other hand, this lack of specificity may relate to the nature of the SH-SY5Y cells, which are undifferentiated neuroblast-like cells that can spontaneously transdifferentiate into an epithelial-like S cell (Ross et al., 1983) and be induced to differentiate to a mature neuronal phenotype. It is therefore possible that the compartmentation of receptors and/or G proteins within the membrane of these cells might be very different from that of mature neurons. This could be approached experimentally by differentiating SH2-D9 cells.
In SH-SY5Y cells, e? and -opioid receptor agonists are able to release Ca2+ from the intracellular stores only when applied in the presence of carbachol, acting at the muscarinic receptor, that causes Ca2+ release through activation of Gq (Connor and Henderson, 1996). In the presence of 1 e carbachol, DAMGO and [D-Pen2,D-Pen5]-enkephalin, e? and -specific agonists, increase [Ca2+]i with an EC50 value of 270 and 10 nM, respectively (Connor and Henderson, 1996). Opioid agonists act through stimulation of phospholipase C by G subunits from Gi/o proteins, although an increase in inositol trisphosphate is still debated (Smart et al., 1995; Yeo et al., 2001). In SH2-D9 but not in SH-SY5Y cells, 1DMe increases [Ca2+]i in the presence of 5 e carbachol, as do the opioid agonists in both cell lines. However, there may exist subtle differences between the mechanism of response to opioid and NPFF agonists because the time courses of the responses to 5-min applications of deltorphin-I and 1DMe are not identical. It should be pointed out that this is the first report of coincident signaling between NPFF receptors and Gq-coupled receptors.
Furthermore, pretreatment of SH2-D9 cells with 100 nM 1DMe reduces the response to deltorphin-I, whereas preincubation with deltorphin-I does not affect the response to 1DMe, demonstrating that the interaction between opioid and NPFF receptors is not reciprocal. Therefore, as for the modulation of N-type Ca2+ channels, NPFF2 receptors exert a noncompetitive antiopioid activity on the release of intracellular Ca2+.
Thus, we demonstrate, for the first time, that NPFF2 receptors have a cellular antiopioid activity in two experimental situations where the final targets, an N-type voltage-sensitive calcium channel and the inositol triphosphategated Ca2+ channel, are different. This suggests that a modification of these channels (e.g., a phosphorylation) reducing their sensitivity to the activation of opioid receptors is not responsible for the antiopioid effect of NPFF agonists but that an upstream modification of the signaling cascade should be considered. A decrease in opioid agonist affinity or receptor number does not explain the antiopioid effect of NPFF because the binding parameters of opioid agonists in membranes or living cells remained unchanged after pretreatment with 1DMe. A competition for a common pool of G protein, resulting in decreased signaling of opioid receptors, might be responsible for the antiopioid effect of NPFF, as in the case of human cannabinoid receptor that reduces noradrenaline- and somatostatin-induced inhibition of Ca2+ channels (for review, see Cordeaux and Hill, 2002). However, we would expect the interaction between opioid and NPFF2 receptors to be competitive, which is not the case, and the opioid agonists to reduce the Ca2+ release induced by 1DMe in the presence of carbachol, which is not the case either.
We have taken advantage of the inhibitory properties of peptides corresponding to the carboxyl terminus of the G subunits (Gilchrist et al., 1998) to carry a more precise characterization of the G proteins associated with NPFF2 and opioid receptors in SH-SY5Y cells. As expected from pertussis toxin experiments, the specific binding to NPFF receptors is strongly decreased by Gi1,2 and to a lesser extent by Go peptide. Accordingly, these two peptides decrease the effect of 1DMe in the presence of carbachol, suggesting that NPFF receptors are preferentially coupled to Gi1,2 and Go proteins. However, even if it does not block the functional effect of NPFF in the carbachol assay, an interaction with the Gi3 subunit is not excluded because a 25% increase in binding is observed with the Gi3 peptide. Thus, our results indicate that NPFF2 receptors are probably coupled to Gi1,2, Go, and Gi3 proteins. They are in accordance with previous experiments performed in human embryonic kidney 293 and COS7 cells cotransfected with chimeric G proteins, showing an interaction between NPFF receptors with Gq/i2 and Gq/o (Elshourbagy et al., 2000) as well as Gq/i3 and Gq/z (Bonini et al., 2000).
The Gs peptide strongly inhibits the binding to NPFF receptors and potentiates the antiopioid effect of 1DMe in Ca2+ conductance measurements. That the action of the Gs peptide is specific is suggested by the fact that although the Gi1,2 peptide reduces the inhibitory effect of 1DMe on the forskolin-induced cAMP production, the Gs peptide is inactive. Some data also suggest a coupling between NPFF receptors and Gs. Analogs of NPFF, at concentrations in the micromolar range, stimulate adenylate cyclase in membranes from the mouse olfactory bulb (Gherardi and Zajac, 1997), and a cellular response to NPFF through the activation of the chimeric Gq/s has been demonstrated (Bonini et al., 2000). Furthermore, the antiopioid activity of NPFF on the nociceptin-induced inhibition of Ca2+ conductances in isolated rat dorsal raphe neurons is prevented by CTX (Roumy and Zajac, 2001). Together, these observations firmly suggest that NPFF2 receptors couple to Gs, although their activation does not clearly increase cAMP production. Furthermore, this coupling does not mediate NPFF antiopioid activity, but it is involved in its modulation.
As described previously in SH-SY5Y (Carter and Medzihradsky, 1993; Laugwitz et al., 1993), opioid receptors are coupled to Gi1,2, Gi3, and Go in SH2-D9 cells. In addition, we find that the binding of [3H]DAMGO is strongly inhibited by the Gs peptide, but this does not result in a reduced effect of the opioid agonist on the N-type Ca2+ channels measured at a saturating concentration. The existence of Gs-coupled excitatory opioid receptors has been suggested (Crain and Shen, 2000), but such a coupling through Gs is not involved in either the inhibition of the N-type Ca2+ channels or the cross-talk with Gq-coupled receptors because both effects are suppressed by PTX.
Together, these results indicate that NPFF2 receptors exert a functional antagonism on opioid receptors that is not reciprocal. There are many peptides, including the opioid peptides themselves, able to modulate the activity of opioid receptors, but to our knowledge only NPFF and cholecystokinin (Heinricher et al., 2001) exert a cellular antiopioid activity. The SH2-D9 model provides the opportunity to characterize the molecular mechanisms involved in the interaction between NPFF and opioid receptors.
doi:10.1124/mol.104.004614.
References
Ballet S, Braz J, Mauborgne A, Bourgoin S, Zajac JM, Hamon M, and Cesselin F (2002) The neuropeptide FF analogue, 1DMe, reduces in vivo dynorphin release from the rat spinal cord. J Neurochem 81: 659eC662.
Ballet S, Mauborgne A, Gouarderes C, Bourgoin AS, Zajac JM, Hamon M, and Cesselin F (1999) The neuropeptide FF analogue, 1DMe, enhances in vivo metenkephalin release from the rat spinal cord. Neuropharmacology 38: 1317eC1324.
Bonini JA, Jones KA, Adham N, Forray C, Artymyshyn R, Durkin MM, Smith KE, Tamm JA, Boteju LW, Lakhlani PP, et al. (2000) Identification and characterization of two G protein-coupled receptors for neuropeptide FF. J Biol Chem 275: 39324eC39331.
Carter BD and Medzihradsky F (1993) Go mediates the coupling of the mu opioid receptor to adenylyl cyclase in cloned neural cells and brain. Proc Natl Acad Sci USA 90: 4062eC4066.
Chang M, Zhang L, Tam JP, and Sanders-Bush E (2000) Dissecting G protein-coupled receptor signaling pathways with membrane-permeable blocking peptides. Endogenous 5-HT(2C) receptors in choroid plexus epithelial cells. J Biol Chem 275: 7021eC7029.
Connor M and Henderson G (1996) delta- and mu-Opioid receptor mobilization of intracellular calcium in SH-SY5Y human neuroblastoma cells. Br J Pharmacol 117: 333eC340.
Cordeaux Y and Hill SJ (2002) Mechanisms of cross-talk between G-protein-coupled receptors. Neurosignals 11: 45eC57.
Crain SM and Shen KF (2000) Antagonists of excitatory opioid receptor functions enhance morphine's analgesic potency and attenuate opioid tolerance/dependence liability. Pain 84: 121eC131.
Elshourbagy NA, Ames RS, Fitzgerald LR, Foley JJ, Chambers JK, Szekeres PG, Evans NA, Schmidt DB, Buckley PT, Dytko GM, et al. (2000) Receptor for the pain modulatory neuropeptides NPFF and NPAF is an orphan G-protein-coupled receptor. J Biol Chem 275: 25965eC25971.
Gherardi N and Zajac J-M (1997) Neuropeptide FF receptors of mouse olfactory bulb: binding properties and stimulation of adenylate cyclase activity. Peptides 18: 577eC583.
Gilchrist A, Mazzoni MR, Dineen B, Dice A, Linden J, Proctor WR, Lupica CR, Dunwiddie TV, and Hamm HE (1998) Antagonists of the receptor-G protein interface block Gi-coupled signal transduction. J Biol Chem 273: 14912eC14919.
Gouardeeres C, Mollereau C, Tafani JA, Mazarguil H, and Zajac J-M (2001) [125I]EYF: a new high affinity radioligand to neuropeptide FF receptors. Peptides 22: 623eC629.
Gouardeeres C, Tafani J-AM, and Zajac J-M (1998) Affinity of neuropeptide FF analogs to opioid receptors in the rat spinal cord. Peptides 19: 727eC730.
Harrison LM, Kastin AJ, and Zadina JE (1998) Opiate tolerance and dependence: receptors, G-proteins and antiopiates. Peptides 19: 1603eC1630.
Heinricher MM, McGaraughty S, and Tortorici V (2001) Circuitry underlying antiopioid actions of cholecystokinin within the rostral ventromedial medulla. J Neurophysiol 85: 280eC286.
Kazmi SM and Mishra RK (1987) Comparative pharmacological properties and functional coupling of e?and opioid receptor sites in human neuroblastoma SH-SY5Y cells. Mol Pharmacol 32: 109eC118.
Kotani M, Mollereau C, Detheux M, Le Poul E, Brezillon S, Vakili J, Mazarguil H, Vassart G, Zajac J-M, and Parmentier M (2001) Functional characterization of a human receptor for neuropeptide FF and related peptides. Br J Pharmacol 133: 138eC144.
Laugwitz KL, Offermanns S, Spicher K, and Schultz G (1993) mu and delta Opioid receptors differentially couple to G protein subtypes in membranes of human neuroblastoma SH-SY5Y cells. Neuron 10: 233eC242.
Law PY, Kouhen OM, Solberg J, Wang W, Erickson LJ, and Loh HH (2000) Deltorphin II-induced rapid desensitization of -opioid receptor requires both phosphorylation and internalization of the receptor. J Biol Chem 275: 32057eC32065.
Mauborgne A, Bourgoin S, Polienor H, Roumy M, Simonnet G, Zajac J-M, and Cesselin F (2001) The neuropeptide FF analogue, 1DMe, acts as a functional opioid autoreceptor antagonist in the rat spinal cord. Eur J Pharmacol 430: 273eC276.
Mazzoni MR, Taddei S, Giusti L, Rovero P, Galoppini C, D'Ursi A, Albrizio S, Triolo A, Novellino E, Greco G, et al. (2000) A Gs carboxyl-terminal peptide prevents Gs activation by the A2A adenosine receptor. Mol Pharmacol 58: 226eC236.
Miller KK and Lupica CR (1997) Neuropeptide FF inhibition of morphine effects in the rat hippocampus. Brain Res 750: 81eC86.
Mollereau C, Mazarguil H, Marcus D, Quelven I, Kotani M, Lannoy V, Dumont Y, Quirion R, Detheux M, Parmentier M, et al. (2002) Pharmacological characterization of human NPFF(1) and NPFF(2) receptors expressed in CHO cells by using NPY Y(1) receptor antagonists. Eur J Pharmacol 451: 245eC256.
Noble F, Szucs M, Kieffer B, and Roques BP (2000) Overexpression of dynamin is induced by chronic stimulation of e? but not -opioid receptors: relationships with e?related morphine dependence. Mol Pharmacol 58: 159eC166.
Panula P, Kalso E, Nieminen M, Kontinen VK, Brandt A, and Pertovaara A (1999) Neuropeptide FF and modulation of pain. Brain Res 848: 191eC196.
Rebeyrolles S, Zajac JM, and Roumy M (1996) Neuropeptide FF reverses the effect of mu-opioid on Ca2+ channels in rat spinal ganglion neurones. Neuroreport 7: 2979eC2981.
Ross RA, Spengler BA, and Biedler JL (1983) Coordinate morphological and biochemical interconversion of human neuroblastoma cells. J Natl Cancer Inst 71: 741eC747.
Roumy M, Garnier M, and Zajac JM (2003) Neuropeptide FF receptors 1 and 2 exert an anti-opioid activity in acutely dissociated rat dorsal raphe and periventricular hypothalamic neurones. Neurosci Lett 348: 159eC162.
Roumy M and Zajac J-M (1998) Neuropeptide FF, pain and analgesia. Eur J Pharmacol 345: 1eC11.
Roumy M and Zajac J-M (1999) Neuropeptide FF selectively attenuates the effects of nociceptin on acutely dissociated neurons of the rat dorsal raphe nucleus. Brain Res 845: 208eC214.
Roumy M and Zajac J-M (2001) Neuropeptide FF receptors couple to a cholera toxin-sensitive G-protein in rat dorsal raphe neurones. Eur J Pharmacol 417: 45eC49.
Smart D, Smith G, and Lambert DG (1995) mu-Opioids activate phospholipase C in SH-SY5Y human neuroblastoma cells via calcium-channel opening. Biochem J 305: 577eC581.
Takeuchi T, Fujita A, Roumy M, Zajac J-M, and Hata F (2001) Effect of 1DMe, a neuropeptide FF analog, on acetylcholine release from myenteric plexus of guinea pig ileum. Jpn J Pharmacol 86: 417eC422.
Toll L, Polgar WE, and Auh JS (1997) Characterization of the delta-opioid receptor found in SH-SY5Y neuroblastoma cells. Eur J Pharmacol 323: 261eC267.
Vaughan PF, Peers C, and Walker JH (1995) The use of the human neuroblastoma SH-SY5Y to study the effect of second messengers on noradrenaline release. Gen Pharmacol 26: 1191eC1201.
Yeo A, Samways DS, Fowler CE, Gunn-Moore F, and Henderson G (2001) Coincident signalling between the Gi/Go-coupled delta-opioid receptor and the Gq-coupled m3 muscarinic receptor at the level of intracellular free calcium in SH-SY5Y cells. J Neurochem 76: 1688eC1700.
Zajac J-M (2001) Neuropeptide FF: new molecular insights. Trends Pharmacol Sci 22: 63.