Address for Correspondence Marcel Jiménez VMD, PhD, Department of Cell Biology, Physiology and Immunology, Edifici V, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain. Tel: +34 93 581 1566; fax: +34 93 581 2006; e-mail: email@example.com
Background In the present study, we further characterize the purinergic receptors mediating the inhibitory junction potential (IJP) and smooth muscle relaxation in the human colon using a new, potent and selective agonist (MRS2365), and antagonists (MR2279 and MRS2500) of the P2Y1 receptor. The P2Y12 antagonist AR-C66096 was tested as well. Using this pharmacological approach, we tested whether β-nicotinamide adenine dinucleotide (β-NAD) fulfilled the criteria to be considered an inhibitory neurotransmitter in the human colon.
Methods We carried out muscle bath and microelectrode experiments on circular strips from the human colon and calcium imaging recordings on HEK293 cells, which constitutively express the human P2Y1 receptor.
Key Results Both the fast component of IJP and non-nitrergic relaxation was concentration-dependently inhibited by MRS2279 and MRS2500. This antagonism was confirmed in HEK293 cells. However, AR-C66096 did not modify either inhibitory response. Adenosine 5′-Ο-2-thiodiphosphate and MRS2365 caused a smooth muscle hyperpolarization and transient inhibition of spontaneous motility that was antagonized by MRS2279 and MRS2500. β-Nicotinamide adenine dinucleotide inhibited the spontaneous motility (IC50 = 3.3 mmol L−1). Nevertheless, this effect was not antagonized by high concentrations of P2Y1 antagonists.
Conclusions & Inferences Inhibitory purinergic neuromuscular transmission in the human colon was pharmacologically assessed by the use of new P2Y1 receptor antagonists MRS2179, MRS2279, and MRS2500. The rank order of potency of the P2Y1 antagonists is MRS2500 > MRS2279 > MRS2179. We found that β-NAD partially fulfills the criteria to be considered an inhibitory neurotransmitter in the human colon, but the relative contribution of each purine (ATP/ADP vsβ-NAD) requires further studies.
Purinergic and nitrergic neurotransmission are the main inhibitory mechanisms causing smooth muscle relaxation in the gastrointestinal (GI) tract. However, the lack of specific agonists and antagonists has made it difficult to establish the identity of the receptors involved in purinergic relaxation. In 1998, Camaioni et al. described the compound MRS2179 (the N6-methyl modification of 2′-deoxyadenosine 3′,5′-bisphosphate) as a potent P2Y1 receptor antagonist.1 This purinergic antagonist is currently considered competitive and specific for the P2Y1 receptor.2 Using MRS2179 it has been shown, in several species and regions of the GI tract that P2Y1 receptors mediate purinergic neurotransmission.3–7 We have recently demonstrated that P2Y1 receptors mediate the fast component of the inhibitory junction potential (IJP) and therefore phasic smooth muscle relaxation both in pig small intestine and in rat and human colon.3,7,8 However, it has also been postulated that other P2Y receptors might also be involved in neuromuscular transmission.9,10 Recent studies suggest that both P2Y1 and P2Y11 receptors mediate fast and slow relaxations in the guinea-pig tenia coli.11 P2Y4 receptors expressed in interstitial cells of Cajal might also participate in inhibitory neurotransmission.12–14
Several pharmacological modifications of MRS2179 have lead to the development of a new agonists and new antagonists for the P2Y1 receptors.15,16 MRS2279 and MRS2500 are selective P2Y1 antagonists that inhibit purinergic neurotransmission in the rat colon and internal anal sphincter.8,17,18 MRS2365 is a preferential P2Y1 agonist, whereas ADPβS is a stable purinergic agonist acting on P2Y1, P2Y11 and P2Y12 receptors. P2Y12 receptors are pharmacological targets to treat thrombosis in humans19 and so, it is important to know whether these receptors participate in neuromuscular transmission in the human GI tract. AR-C66096 is a preferential antagonist of P2Y12 receptors.
The neurotransmitter responsible for purinergic neurotransmission is still under debate. Classically, the terminology ‘ATP or a related purine’ is employed in the majority of articles because ATP is quickly degraded by ectonucleases to ADP, AMP, and adenosine.20 It has been recently proposed that β-nicotinamide adenine dinucleotide (β-NAD) is potentially an inhibitory purinergic neurotransmitter in the murine GI tract, possibly acting on P2Y1 receptors.14β-Nicotinamide adenine dinucleotide might also activate P2Y11 receptors in human granulocytes.21 Thus, two putative families of compounds: β-NAD and ATP and their respective products of degradation are possible candidates to be identified as inhibitory neurotransmitters in the GI tract.14 While, we were preparing the manuscript it has been proposed that β-NAD as inhibitory neurotransmitter in the human colon.22
According to these results, it is important to translate data from animals to humans and the main aim of the present work was to study the purinergic inhibitory neurotransmission in the human colon using these new available pharmacological tools. We also wanted to test whether β-NAD fulfills the criteria to be considered an inhibitory neurotransmitter in the human colon. Pharmacological characterization of receptors and neurotransmitters involved in human colonic smooth muscle relaxation mediated by purines might help to develop pharmacological strategies to treat colonic motility disorders with impaired purinergic neurotransmission as has been described in animal models.23
Materials and methods
Specimens of distal and sigmoid colon (n = 39) were obtained from patients (aged 46–91 years) during colon resections for neoplasm. Colon segments from macroscopically normal regions were collected and transported to the laboratory in cold saline buffer. The tissue was placed in Krebs solution on a dissection dish, and the mucosal layer was gently removed. Circular muscle strips, 10 × 4 mm, were cut. The experimental procedure was approved by the Ethics Committee of the Hospital of Mataró (Barcelona, Spain).
Circular muscle strips were studied in a 10 mL organ bath filled with Krebs solution containing phentolamine, atropine, and propanol (all 1 μmol L−1) at 37 ± 1 °C. An isometric force transducer (Harvard VF-1; Harvard Apparatus Inc., Holliston, MA, USA) connected to a computer through an amplifier was used to record the mechanical activity. Data were digitalized (25 Hz) using Data 2001 software (Panlab, Barcelona, Spain) coupled to an ISC-16 A/D card installed in the computer. A tension of 4 g was applied and the tissue was allowed to equilibrate for 1 h. After this period, strips displayed spontaneous phasic activity. Electrical field stimulation (EFS) was applied for 2 min (pulse duration 0.4 ms, frequency 2 Hz, and amplitude 50 V). The area under the curve (AUC) of contractions from the baseline was used to measure the spontaneous motility.
Muscle strips were dissected parallel to the circular muscle and placed in a Sylgard-coated chamber continuously perfused with Krebs solution containing phentolamine, atropine, propranolol (all 1 μmol L−1), and l-NNA (1 mmol L−1) at 37 ± 1 °C. Strips were meticulously pinned in a cross-sectioned slab allowing microelectrode recordings from circular muscle. This procedure was previously reported in the canine ileum24 and human colon.3 Preparations were allowed to equilibrate for approximately 1 h before experiments started. Circular muscle cells were impaled with glass microelectrodes (40–60 MΩ) filled with 3 mol L−1 KCl. Membrane potential was measured using standard electrometer Duo773 (WPI Inc., Sarasota, FL, USA). Tracings were displayed on an oscilloscope 4026 (Racal-Dana Ltd., Windsor, England) and simultaneously digitalized (100 Hz) using PowerLab 4/30 system and Chart 5 software for Windows (all from AD Instruments, Castle Hill, NSW, Australia). Electrical field stimulation was applied using two silver chloride plates placed perpendicular to the longitudinal axis of the preparation and 1.5 cm apart. Train stimulation had the following parameters: total duration, 100 ms; frequency, 30 Hz; pulse duration, 0.3 ms, and increasing amplitude strengths of 5, 10, 12, 15, 17, 20, 25, 30, and 50 V. Resting membrane potential was measured before and after drug addition. The amplitude of IJPs was measured under control conditions and after infusion of each drug. To obtain stable smooth muscle cell impalements, nifedipine (1 μmol L−1) was used to abolish its mechanical activity.
Calcium imaging technique
HEK293 cells, which constitutively express the human P2Y1 receptor (Fischer 2005) were studied by using the calcium imaging technique. HEK 293 cells were grown in Dulbecco’s Modification of Eagles Medium (DMEM) (37 °C, pH 7.4) and seeded in a culture dish with a coverglass 48 h before the experimental procedure. Cells were loaded with Fluo-4 AM (5 μmol L−1, room temperature, 45 min) in extracellular medium solution. The coverglass containing the Fluo-4 AM loaded HEK 293 cells was placed over a metal ring, immobilized by a rubber o-ring and transferred to a chamber. Fluo-4 AM was used to monitor changes in cytosol calcium level. After washing out the remaining dye, cells were incubated in the recording medium. Cells were imaged with IX-FLA Camera connected to an Olympus IX70 microscope and scanned using CellR software (Olympus Biosystems, Heidelberg, Germany). In preliminary experiments, optimal UV light intensity was set to obtain sufficient image quality, whereas minimizing phototoxicity and bleaching. Cells were perfused in a constant flow of extracellular medium solution (1 mL min−1) at RT and a manual valve ALA VM-8 channel bath perfusion system (npi Electronic Instruments, Tamm, Germany) allowed switching between normal and drug-containing solutions. Changes in Fluo-4 fluorescence were recorded at 2.5Hz with a spatial resolution of 512 × 480 pixels. At the end of the experiments, images were analyzed over time using regions of interest (ROIs). Fluorescence intensity of ROIs (ΔF/F) analysis was performed and plotted over time.
Solutions and drugs
The composition of the Krebs solution was (in mmol L−1): 10.10 glucose, 115.48 NaCl, 21.90 NaHCO3, 4.61 KCl, 1.14 NaH 2 PO4, 2.50 CaCl 2, and 1.16 MgSO4 (pH 7.3–7.4). It was maintained at 37 ± 1 °C and bubbled with a mixture of 5% CO2 and 95% O2. In all the experiments, phentolamine, atropine and propranolol (1 μmol L−1) were added to the Krebs solution to block α- and β-adrenoceptors and muscarinic receptors. The composition of the extracellular medium solution was (in mmol L−1): 140 NaCl, 4.8 KCl, 1 MgCl2 6H2O, 1.8 CaCl2 2H2O, 10 glucose.
The following drugs were added to the Krebs solution: Nifedipine, Nε-nitro-l-arginine (l-NNA), adenosine 5′-Ο-2-thiodiphosphate (ADPβS), β-nicotinamide adenine dinucleotide sodium salt (β-NAD) and phentolamine (Sigma Chemicals, St Louis, MO, USA); tetrodotoxin (TTX) (Latoxan, Valence, France); atropine sulphate (Merck, Darmstadt, Germany); propranolol, 2′-deoxy-N6-methyladenosine 3′,5′-bisphosphate tetrasodium salt (MRS2179), (1R,2S,4S,5S)-4-[2-Chloro-6-(methylamino)-9H-purin-9-yl]-2-(phosphonooxy)bicyclo[3.1.0]hexane-1-methanol dihydrogen phosphate ester diammonium salt (MRS2279), (1R,2S,4S,5S)-4-[2-Iodo-6-(methylamino)-9H-purin-9-yl]-2-(phosphonooxy)bicyclo[3.1.0]hexane-1-methanol dihydrogen phosphate ester diammonium salt (MRS2500), [[(1R,2R,3S,4R,5S)-4-[6-Amino-2-(methylthio)-9H-purin-9-yl]-2,3 dihydroxybicyclo[3.1.0]hex-1-yl]methyl] diphosphoric acid mono ester trisodium salt (MRS2365), 2-(Propylthio)adenosine-5′-O-(β,γ-difluoromethylene)triphosphate tetrasodium salt (AR-C66096) (Tocris, Bristol, UK); and Fluo-4 AM (Teflabs Inc. Texas, USA). Stock solutions were made by dissolving drugs in distilled water except for nifedipine, which was dissolved in ethanol (96%), l-NNA was dissolved in Krebs solution by sonication and Fluo-4 AM was dissolved in DMSO (0.002% final concentration).
Data analysis and statistics
One-way analysis of variance (anova) was used to evaluate the effect of MRS2279 or MRS2500 on: (i) the inhibition of spontaneous motility induced by ADPβS and MRS2365 and 2- EFS-induced relaxation. Two-way anova was used to analyze the effect of MRS2279 or MRS2500 on: (i) the inhibition of spontaneous mechanical activity caused by β-NAD, and (ii) the amplitude of the IJPs (drug and voltage). Paired t-test was used to evaluate the effect of MRS2279 or MRS2500 on the hyperpolarization induced by ADPβS.
To normalize data, the inhibitory effect of ADPβS, MRS2365, β-NAD, and EFS on spontaneous rhythmic colonic contractions was calculated as percentage of inhibition using the following formula: 1 − [area under the curve (AUC) during drug addition or EFS/AUC previous drug addition or EFS] × 100. When motility was completely abolished, inhibition was considered 100%, whereas inhibition was 0% when no changes in AUC were observed. To calculate IC50 and Hill slopes, data were fitted to a sigmoid concentration-response curve of variable slope: Y = 100/(1 + 10((LogEC50-X)*Hill slope)), where X is the logarithm of concentration and Y is the response.
All image analysis was performed with the CellR sofware (Olympus Biosystems). Regions of interest were drawn over each cell, fluorescence intensity was normalized to the basal fluorescence at the onset of the recording for each ROI, and peaks were analyzed. Paired student’s t-test or anova test was used before and after drug addition.
A P < 0.05 was considered statistically significant. ‘n’ values indicate the number of samples from different patients or the number of cells analyzed (using at least three to four different plates). Statistical analysis was performed with GraphPad Prism version 4.00, (GraphPad Software, San Diego, CA, USA).
Effect of P2Y1 antagonists on the inhibition of spontaneous motility induced by EFS
Two minutes EFS caused complete cessation of spontaneous motility. Incubation with MRS2279 (1) or MRS2500 (1) did not modify the inhibitory response. To study the purinergic component of the relaxation, different concentrations of the P2Y1 receptors antagonists (MRS2279 and MRS2500 were tested in strips previously incubated with l-NNA (1 m, 20 min). In the presence of l-NNA, the purinergic antagonists caused a concentration-dependent inhibition of the non-nitrergic relaxation induced by EFS, the IC50 was 0.26 μmol L−1 (n = 7) for MRS2279 and 0.088 μmol L−1 (n = 5) for MRS2500 (Fig. 1). Table 1 summarizes these results. A comparison with the effect of MRS2179 (Gallego et al., 2006)3 is shown.
Table 1. Table summarizing the IC50 (μmol L−1) of the purinergic antagonists obtained vs the endogenous inhibitory transmitter, released by EFS and assessed on mechanical activity (relaxation) and membrane potential (IJP), and vs the exogenous agonist ADPβS, assessed in HEK cells
Relaxation induced by EFS
IJP induced by EFS
Calcium transients induced by ADPβS 1 μmol L−1
IJP, inhibitory junction potential; EFS, electrical field stimulation.
AR-C66096 did not affect the IJP and relaxation induced by EFS, data are expressed as mean (95% confidence interval).
Effect of P2Y1 antagonists on the inhibitory junction potential
Increasing the voltage of stimulation caused a progressive increase in the amplitude of the IJP (Fig. 2). MRS2279 and MRS2500 concentration-dependently reduced the IJP amplitude (anovaP < 0.0001; n = 5 each drug) (Fig. 2). To calculate the IC50 of MRS2279 and MRS2500, a protocol using supramaximal IJPs was performed (using 30 V stimuli). The IC50 was 0.28 μmol L−1 (n = 5) for MRS2279 and 0.071 μmol L−1 (n = 5) for MRS2500 (Fig. 2). Table 1 summarizes these results. A comparison with the effect of MRS2179 (Gallego et al., 2006)3 is shown.
Effect of other P2Y antagonists on the IJP and inhibition of spontaneous motility induced by EFS
The following antagonists were tested in the presence of l-NNA (1 mmol L−1) to isolate the purinergic component of the inhibitory response. AR-C66096, a P2Y12 receptor antagonist, was tested at 0.1, 1 and 5 μmol L−1, but neither the IJP (n = 4 each concentration) nor the inhibition of spontaneous motility induced by EFS (n = 4 each concentration) was affected.
Effect of exogenous addition of ADPβS and MRS2365 on spontaneous motility
Human colonic strips spontaneously developed rhythmic phasic contractions. In the presence of TTX (1 μmol L−1), the purinergic agonist ADPβS (10 μmol L−1) inhibited the spontaneous contractions displayed by circular muscle strips by 71.3 ± 8.2% (n = 10). After 10-min preincubation with MRS2279 (1 μmol L−1) or MRS2500 (1 μmol L−1), the inhibition caused by ADPβS (10 μmol L−1) was reduced to 29.7 ± 9.0% (n = 5; P < 0.01) and 13.2 ± 4.3% (n = 5, P < 0.001), respectively (Fig. 3A). MRS2365, a selective P2Y1 agonist, caused a concentration-dependent inhibition of spontaneous mechanical activity (0.1 μmol L−1: 1.8 ± 8.8%, n = 3; 1 μmol L−1: 16.9 ± 6.2%, n = 8; and 10 μmol L−1: 51.5 ± 5.5%, n = 11). Previous incubation with MRS2500 1 μmol L−1 blocked the inhibitory effect of MRS2365 10 μmol L−1 (9.4 ± 4.5%, n = 4, P < 0.001). MRS2279 (10 μmol L−1) completely blocked the inhibitory effect induced by MRS2365 10 μmol L−1 (7.1 ± 2.4%, n = 3, P < 0.001) (Fig. 3B).
Effect of exogenous addition of ADPβS and MRS2365 on smooth muscle membrane potential
Transient superfusion of the tissue with ADPβS (1 μmol L−1) hyperpolarized the smooth muscle by −10.7 ± 1.7 mV (n = 8). This hyperpolarization was reduced by MRS2279 (1 μmol L−1) to −2.2 ± 2.2 mV (P < 0.05, n = 3) and MRS2500 (0.5 μmol L−1) to −2.4 ± 1.3 mV (P < 0.05, n = 3). The selective P2Y1 agonist MRS2365 caused a concentration-dependent hyperpolarization (0.1 μmol L−1: −4.2 ± 1.9 mV, n = 5; 1 μmol L−1: −5.4 ± 1.4 mV, n = 6 and 10 μmol L−1: −9.5 ± 2.1 mV, n = 8) (Fig. 4). The hyperpolarization induced by MRS2365 at 0.1 and 1 μmol L−1 was completely antagonized by MRS2500 1 μmol L−1 and at 10 μmol L−1, the hyperpolarization induced by MRS2365 was reduced to −5.7 ± 4.7 mV by MRS 2500 1 μmol L−1 (n = 3). AR-C66096 5 μmol L−1 (n = 2) did not modify the hyperpolarization induced by 10 μmol L−1 ADPβS.
Effect of ADPβS on calcium transients in HEK293 cells
ADPβS (1 μmol L−1) elicited a calcium increase (ratio 4.09, n = 179) in HEK 293 cells. This calcium increase was concentration-dependently inhibited by MRS2179 IC50 = 0.13 μmol L−1 (n = 30), MRS2279 IC50 = 4.6 nmol L−1 (n = 30) and MRS2500 IC50 = 2.89 nmol L−1 (n = 30). Table 1 summarizes these data.
Effect of β-NAD on spontaneous motility, resting membrane potential, and calcium transient induced in HEK 293 cells
At high concentrations, β-NAD induced a concentration-dependent inhibition of spontaneous motility (IC50 = 3.3 mmol L−1; 95% confidence interval 2.4–4.4 mmol L−1; log IC50 = −2.49 ± 0.06; n = 8), which was not antagonized by MRS2279 or MRS2500 (1 μmol L−1) (Fig. 3C). Higher concentrations of MRS2279 and MRS2500 (5 μmol L−1) did not revert the inhibition caused by β-NAD (data not shown). β-Nicotinamide adenine dinucleotide caused a slight hyperpolarization of −3.2 ± 0.6 mV at 1 mmol L−1 (n = 6) and −4.1 ± 1.1 mV at 10 mmol L−1 (n = 2) (Fig. 4). The hyperpolarization caused by β-NAD at 10 mmol L−1 was abolished by MRS2500 1 μmol L−1 (n = 2). In HEK 293 cells, β-NAD did not induce a significant calcium increase at concentrations up to 5 mmol L−1. At 10 mmol L−1, β-NAD induced a calcium increase in about 50% of the ADPβS responding cells (n = 58) and the amplitude of the response was a third of those induced by ADPβS (ratio 1.37; n = 28 responding cells; P < 0.0001).
The aim of the present study has been to characterize the purinergic receptor(s) that participate in the inhibitory effect at the neuromuscular junction in the human colon and to find out the nature of the neurotransmitter(s) involved in the purinergic inhibition. Despite the evidence that purinergic receptors play a key role in purinergic inhibitory neuromuscular transmission in the GI tract,25 there are very few articles characterizing this neurotransmission in human tissues. In the human colon, NO and a purine are two of the main neuromuscular inhibitory transmitters.26 Both in the human small intestine and colon repetitive electrical field stimulation causes an IJP with two phases, a fast component followed by an l-NNA-sensitive sustained component.27,28 The fast component is partially inhibited by apamin, suramine and it is desensitized by ADPβS agonist suggesting an involvement of P2 receptors.29 Moreover, PPADS and suramine partially inhibited the non-nitrergic relaxation induced by field stimulation both in the human small intestine and colon.30,31 All these results suggest a putative role of P2 receptors in purinergic inhibitory neuromuscular transmission.
Recently, we and other research groups have shown that MRS2179, a selective P2Y1 antagonist, inhibits the inhibitory neurotransmission in the human GI tract. In the colon: (i) Both the IJP and the non-nitrergic relaxation are concentration-dependently inhibited by MRS2179 with an IC50 of about 1 μmol L−1 3, (ii) when long pulses of 5Hz are applied, the fast component is sensitive to MRS2179, whereas the slow component is sensitive to l-NNA32 and, (iii) non-nitrergic neural-mediated relaxations induced by stimulation of nicotinic receptors are blocked by MRS2179.33 In the human small intestine, non-nitrergic relaxation is sensitive to MRS2179 at concentration between 3 and 10 μmol L−1.34 All these data suggest an involvement of P2Y1 receptors in different areas of the human GI tract. However, data from these articles show that a concentration of 10 μmol L−1 of MRS2179 is needed to inhibit purinergic responses and it is possible that at this concentration the antagonist might partially loose selectivity. Data from the present article show that: (i) the fast component of IJP and the non-nitrergic relaxation are inhibited in a concentration-dependent manner by MRS2279 and MRS2500, (ii) 2 the rank of potency of these drugs is MRS2500 > MRS2279 > MRS2179 (see Table 1), (iii) data from the antagonists obtained with organ bath and microelectrode are similar showing that the underlying mechanism responsible of the non-nitrergic relaxation is the purinergic IJP mediated by P2Y1 receptors (iv) concentrations of 1 μmol L−1 MRS2279 and 0.5 μmol L−1 MRS2500 are able to completely block the inhibitory response, and (v) P2Y1 antagonists alone without NOS inhibition are not able to inhibit the relaxation induced by EFS. The nitrergic slow component of the IJP is probably able to cause a sustained hyperpolarization underlying the mechanical relaxation.32 It is important to note that AR-C66096, a P2Y12 receptor antagonist, did not modify the neural-mediated inhibitory responses. P2Y12 receptor antagonists are in development as antithrombotic agents.19 In the present article, we show for the first time that they do not inhibit inhibitory neuromuscular transmission in the human colon.
We strongly believe that it is mandatory to translate data obtained in animals to humans as important differences between species have been reported. In rodents, MRS2179 is comparatively less potent inhibiting the IJP8,9,35 than in humans,3,32 pigs 7 or guinea-pigs.6 However, when MRS2279 and MRS2500 were tested, the fast component of the IJP in the rat colon was completely inhibited by very low concentrations of these antagonists.8 The rank of potency of the antagonists is similar in both rats and humans (MRS2500 > MRS2279 > MRS2179). MRS2179 is comparatively more potent in humans (IC50 1.2 μmol L−1)3 than in rats (IC50 13.1 μmol L−1),8 but both MRS2279 and MRS2500 are more potent in rats (17.8 nmol L−1 and 14 nmol L−1 respectively)8 than in humans (280 nmol L−1 and 71 nmol L−1, respectively) (present study). Differences in the structure of the receptor between species might be responsible for these findings. An important difference between these antagonists is that MRS2179 is reversed on washout (after 30 min washing with Krebs solution) but MRS2279 and MRS2500 are not washable. This might be due to varying sensitivity to ectonucleotidases.36
To test the efficacy of MRS antagonists on P2Y1 receptors we used HEK-293 cells that endogenously express P2Y1, P2Y2 and P2Y4 receptors.37 It has been previously reported that calcium increase, using fura-2 microfluometry, induced by ADPβS was blocked by 30 μmol L−1 MRS2179. This effect is probably due to P2Y1 receptor activation. In contrast, UTP, a preferential P2Y2,4, agonist, induced a calcium increase that was not reduced by high concentrations (30 μmol L−1) of MRS2179.38 MRS2179 was a competitive antagonist in HEK-293 cells transfected with the P2Y1 guinea-pig receptor.39 In line with these results, our data show that calcium increase induced by ADPβS in HEK-293 cells was concentration-dependently decreased by P2Y1 antagonists. The IC50 was 2.9 nmol L−1 (MRS2500), 4.6 nmol L−1 (MRS2279) and 0.12 μmol L−1 (MRS2179). Due to the high selectivity and potency of these antagonists,40 it seems clear that P2Y1 receptors are implicated in the response. It is important to notice that the range of concentration varies greatly when the antagonist is used in experiments with dispersed cells (calcium transients elicited with ADPβS 1 μmol L−1) and experiments, where tissue is used and the response induced by the endogenous neurotransmitter is measured. In this case, data from microelectrode and muscle bath are quite similar.
As we previously reported, ADPβS causes an inhibition of spontaneous motility and a transient hyperpolarization.3 Both effects were inhibited by MRS21793 and in the present study; we have demonstrated that they are also inhibited with MRS2279 and MRS2500. P2Y1 antagonists tested in the present study are selective for P2Y1 receptors.2 However, the orphan receptor GPR17 is also blocked by low concentrations of MRS2179.41 GPR17 is coupled with Gi proteins leading to both adenylyl cyclase inhibition and calcium increase. However, GPR17 is not activated by either ATP or ADP. Moreover, GPR17 agonists such as UDP-glucose, which is also an agonist of P2Y14 receptors, cause gastric smooth muscle contractions, which are not recorded in P2Y14 KO mice.42 Altogether these results suggest that GPR17 is probably not involved in smooth muscle hyperpolarization and relaxation although further experiments are needed to characterize its effect on smooth muscle. In the present study, we demonstrate that MRS2365, a preferential P2Y1 agonist, hyperpolarises and transiently inhibits spontaneous motility, mimicking endogenous IJP and the corresponding mechanical inhibition. This effect is antagonised by P2Y1 antagonists confirming the involvement of this receptor in the inhibitory pathways in the human colon.
The nature of the neurotransmitter involved in purinergic response is still unknown and most articles use the term ‘ATP or a related purine’, which encompasses several putative purinergic neurotransmitters. It has been recently suggested that β-NAD released by enteric nerves and acting postjunctionally on P2Y1 receptors fits the criteria to be considered an inhibitory neurotransmitter.14,22 Although in the present study we did not measured endogenous release of purines our data with exogenous addition of β-NAD are similar to those recently reported: (i) Exogenous addition of β-NAD causes inhibition of spontaneous motility with an IC50 in the mM range, and (ii) β-NAD causes a slight hyperpolarization of human colonic smooth muscle cells (50 mmol L−1: about 6–7 mV22vs 10 mmol L−1 : 3 mV in the present study), and (iii) β-NAD -induced hyperpolarization is sensitive to the P2Y1 antagonist MRS2500. These data suggest that β-NAD might be an endogenous inhibitory mediator in the human colon acting on P2Y1 receptors. However, (i) β-NAD-induced inhibition of spontaneous motility was not antagonized by MRS2279 or MRS2500. This suggests that β-NAD-induced relaxation independently of the membrane potential. Moreover, β-NAD (10 mmol L−1) induced a very low response both in number of responding cells and in amplitude of the response in HEK-293 cells that constitutively express the P2Y1 receptor. This might explain why the hyperpolarization induced by high concentrations of β-NAD is comparatively smaller than those obtained with MRS2365 (1 μmol L−1). Alternatively, activation of non-selective cation channels possibly not located at the neuromuscular junction might counterbalance the hyperpolarization caused by exogenous administration of the compound.22 However, a contractile response has never been observed in our muscle bath recordings. Further studies comparing endogenous release of purines vs activation of P2Y1 receptors are needed to evaluate the relative contribution of ATP/ADP and β-NAD as inhibitory neurotransmitters in the GI tract.
In conclusion, our data support that P2Y1 receptors mediate the fast component of the IJP and the purinergic relaxation. P2Y1 antagonists show a rank of potency: MRS2500 > MRS2279 > MRS2179 in all three techniques tested. Preferential P2Y1 agonists such as ADPβS and MRS2365 mimicked the effect of the endogenous mediator. These antagonists are valuable pharmacological tools to study purinergic neurotransmission in the GI tract and to help in the development of future putative treatments, where purinergic neurotransmission is impaired.
The authors thank Claudia Arenas, Antonio Acosta, and Emma Martinez for the technical assistance. Diana Gallego is supported by the Instituto de Salud Carlos III, Centro de Investigación Biomédica en red de enfermedades hepáticas y digestivas (CIBERehd). Victor Gil is supported by the Ministerio de Ciencia e Innovación (AP2007-01583). We thank Drs Joan Martí-Ragué. Alex Sáenz, M Marti-Gallostra and J. Cases (Clínica Sagrada Familia) and Dr. Xavier Suñol, Dr Oscar Estrada, Dr. Fran Espin, Dr. Adolfo Heredia, Dr. Eva García, and Dr. Luís Antonio Hidalgo (Hospital de Mataró) for providing human colonic tissue. This work has been funded by the following grant: BFU2009-11118.