Calcium signalling through nucleotide receptor P2X1 in rat portal vein myocytes

Abstract

1. ATP-mediated Ca²⁺ signalling was studied in freshly isolated rat portal vein myocytes by means of a laser confocal microscope and the patch-clamp technique.

2. In vascular myocytes held at −60 mV, ATP induced a large inward current that was supported mainly by activation of P2X1 receptors, although other P2X receptor subtypes (P2X3, P2X4 and P2X5) were revealed by reverse transcription-polymerase chain reaction.

3. Confocal Ca²⁺ measurements revealed that ATP-mediated Ca²⁺ responses started at initiation sites where spontaneous or triggered Ca²⁺ sparks were not detected, whereas membrane depolarizations triggered Ca²⁺ waves by repetitive activation of Ca²⁺ sparks from a single initiation site.

4. ATP-mediated Ca²⁺ responses depended on Ca²⁺ influx through non-selective cation channels that activated, in turn, Ca²⁺ release from the intracellular store via ryanodine receptors (RYRs). Using specific antibodies directed against the RYR subtypes, we show that ATP-mediated Ca²⁺ release requires, at least, RYR2, but not RYR3.

5. Our results suggest that, in vascular myocytes, Ca²⁺ influx through P2X1 receptors may trigger Ca²⁺-induced Ca²⁺ release at intracellular sites where RYRs are not clustered.

P2 receptors for nucleotides are located on the extracellular surface of a variety of mammalian cell types, including smooth muscle cells (Kunapuli & Daniel, 1998; Ralevic & Burnstock, 1998). They can be divided into two main classes according to their signalling mechanisms. P2X receptors are ligand-gated cation channels, whereas metabotropic P2Y receptors are G protein-coupled receptors (Fredholm et al. 1997). To date, seven types of P2X receptor and at least five types of P2Y receptor have been identified at the molecular level (North & Surprenant, 2000). Frequently, multiple purinoceptor subtypes have been found to co-exist in the same cell, but their relative importance is only beginning to be understood (Boarder & Hourani, 1998). In smooth muscle cells, the expression of P2Y receptors is markedly upregulated in culture, so that their effects may become predominant (Erlinge et al. 1998). In freshly dissociated or short-term cultured smooth muscle cells, P2X receptors are the major purinoceptors expressed and their activation leads to inward current through non-selective cation channels (Benham & Tsien, 1987; Honoréet al. 1989). From cytosolic Ca²⁺ measurements, it has been suggested that ATP may release Ca²⁺ from intracellular stores in response to activation of a Ca²⁺-induced Ca²⁺ mechanism (Luo et al. 1999). However, the Ca²⁺ signalling pathway that is activated by P2X receptors remains to be elucidated.

In smooth muscle, intracellular Ca²⁺ signals can be generated by inositol 1,4,5-trisphosphate-gated channels (InsP₃Rs) and ryanodine-sensitive channels (RYRs) and there are indications that these two Ca²⁺ release channels are located on the same intracellular store in some cell types. In rat portal vein myocytes, Ca²⁺ sparks have been shown to be produced by RYRs, as in other types of muscle cell (Arnaudeau et al. 1996; Mironneau et al. 1996). It has been shown that the spatiotemporal summation of Ca²⁺ sparks activated by L-type Ca²⁺ current gives rise to propagated Ca²⁺ waves (Cheng et al. 1996; Lipp & Niggli, 1996; Arnaudeau et al. 1997). Using an antisense strategy, it has been shown that triggered Ca²⁺ sparks and propagated Ca²⁺ waves both require RYR subtype 1 (RYR1) and 2 (RYR2), but not RYR subtype 3 (RYR3). This hierarchical Ca²⁺ signalling, from elementary Ca²⁺ sparks to propagated Ca²⁺ waves, is responsible for the angiotensin II-activated increase in [Ca²⁺]_i (Arnaudeau et al. 1996). In noradrenaline-induced propagated Ca²⁺ waves, Ca²⁺ sparks are activated locally by Ca²⁺ release through InsP₃-gated channels and contribute to an all-or-none increase in [Ca²⁺]_i (Boittin et al. 1999).

The aims of the present study were to characterize the Ca²⁺ signalling pathway activated by ATP in rat portal vein myocytes and to identify the Ca²⁺ release channels that are involved in the ATP-induced Ca²⁺ responses. We report that: (1) ATP induces Ca²⁺ responses, essentially through activation of P2X1 receptors, which (in contrast to membrane depolarizations) do not start from the initiation sites that produce spontaneous or triggered Ca²⁺ sparks; and (2) ATP-mediated Ca²⁺ release requires, at least, activation of RYR2, but not RYR3. These results suggest that, in vascular myocytes, ATP-activated Ca²⁺ influx through non-selective cation channels may trigger Ca²⁺-induced Ca²⁺ release at intracellular sites where RYRs are not clustered.

METHODS

Cell preparation

Experiments conformed with the European Community and French guiding principles for the care and use of laboratory animals (authorized by the French Ministre de l'Agriculture et de la Pêche). Rats (160–180 g) were killed by cervical dislocation. The portal vein was cut into several pieces and incubated in low Ca²⁺ (40 μm) physiological solution for 10 min. Thereafter, 0.8 mg ml⁻¹ collagenase (EC 3.4.24.3), 0.2 mg ml⁻¹ pronase E (EC 3.4.24.31) and 1 mg ml⁻¹ bovine serum albumin were added at 37 °C for 20 min. Subsequently, the solution was removed and the pieces of portal vein were incubated again in a fresh enzyme solution at 37 °C for 20 min. The tissues were placed in an enzyme-free solution and triturated using a fire-polished Pasteur pipette to dissociate cells. Cells were seeded at a density of 10³ cells mm⁻² on glass slides and used on the same day.

Reverse transcription-polymerase chain reaction

Total RNA was extracted from cells on one slide using an RNeasy mini kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. The reverse transcription (RT) reaction was performed using a Sensiscript RT kit (Qiagen). Total RNA was incubated with oligo-dT₍₁₅₎ primers (Promega, Lyon, France) at 65 °C for 5 min. RT mix was added after 15 min at room temperature and the total RT mix was incubated for 60 min at 37 °C. A control without reverse transcriptase was included in each experiment. The resulting cDNA was stored at −20 °C. Polymerase chain reaction (PCR) was performed with 2 μl of cDNA (in RT-PCR mix), 1.25 units of HotStartTaq DNA polymerase (Qiagen), 2.5 mm MgCl₂, 0.5 μm of each primer and 200 μm of each deoxynucleotide trisphosphate, in a final volume of 50 μl. The PCR conditions were 95 °C for 15 min, followed by 35 cycles at 94 °C for 1 min, 62 °C for 1 min and 72 °C for 1.5 min. After the PCR, samples were kept at 72 °C for 10 min for final extension and then stored at 4 °C. The optimal hybridization temperature was determined by using a gradient of temperature between 45 and 65 °C. RT was performed with a thermal cycler (Techne, Cambridge, UK) and PCR was performed with the Mastercycler gradient (Eppendorf, Paris, France). Amplification products were separated by electrophoresis (2 % agarose gel) and visualized with ethidium bromide. Gels were photographed with EDAS 120 and analysed with KDS1D 2.0 software (Kodak Digital Science, Paris, France). Sense and antisense primer pairs specific for P2X receptors were previously described by Shibuya et al. (1999) and were verified by comparison with cloned receptor sequences available in GenBank with DNAstar software (Lasergene, DNAstar, Madison, USA). The length of the expected PCR products for the P2X receptor fragments was 452 bp for P2X1, 357 bp for P2X2, 440 bp for P2X3, 447 bp for P2X4, 418 bp for P2X5, 520 bp for P2X6 and 354 bp for P2X7. PCR fragments were sequenced by the Qiagen sequencing service.

Patch-clamp measurements

Voltage-clamp and membrane current recordings were made using a standard patch-clamp technique with a List EPC7 patch-clamp amplifier (Darmstadt-Eberstadt, Germany). Patch-clamp pipettes of 2–5 MΩ resistance were used for whole-cell recording. Membrane potential and current records were stored and analysed with pCLAMP software (Axon Instruments, Foster City, CA, USA). Ba²⁺ currents were corrected digitally for leakage current. Cell capacitance was determined in each cell tested by imposing 10 mV hyperpolarizing steps from the holding potential. The normal physiological solution contained 130 mm NaCl, 5.6 mm KCl, 1 mm MgCl₂, 1.7 mm CaCl₂, 11 mm glucose and 10 mm Hepes (pH adjusted to 7.4 with NaOH). Ca²⁺-free solution was prepared by omitting CaCl₂ and by adding 0.5 mm EGTA. Low Na⁺ solution was prepared by replacing 100 mm NaCl with tetraethylammonium chloride. The standard pipette solution contained 120 mm CsCl, 10 mm NaCl and 10 mm Hepes (pH adjusted to 7.3 with NaOH). The low Cl⁻ pipette solution was prepared by replacing 105 mm CsCl with caesium aspartate. The external solution used to record Ba²⁺ currents was prepared by replacing CaCl₂ with 5 mm BaCl₂. In experiments where antibodies were added to the pipette solution, the infusion time after breakthrough in the whole-cell recording mode was at least 7 min, which is longer than the time expected theoretically for diffusion of substances in solution (Viard et al. 1999).

Cytosolic Ca²⁺ measurements

In all experiments, fluo 3 (60 μm) was dialysed into cells through the patch-clamp pipette. Images were acquired using the linescan mode of a confocal BioRad MRC1000 microscope (BioRad, Paris, France) connected to a Nikon Diaphot microscope. Excitation light was delivered by a 25 mW argon ion laser (Ion Laser Technology, Salt Lake City, UT, USA) through a Nikon Plan Apo ×60, 1.4 NA objective lens. Fluo 3 was excited at 488 nm and fluorescence emission was filtered and measured at 540 ± 30 nm. At the setting used to detect fluo 3 fluorescence, the resolution of the microscope was near 0.4 μm × 0.4 μm × 1.5 μm (x-, y- and z-axis, respectively). Images were acquired in the linescan mode at a rate of 6 ms per scan. Scanned lines were plotted vertically and each line was added to the right of the preceding line to form the linescan image. In these images, time increased from the left to the right, and the position along the scanned line was given by the vertical displacement. Fluorescence signals are expressed as the pixel per pixel fluorescence ratio (F/F₀), where F is the fluorescence during a response and F₀ is the rest level fluorescence of the same pixel.

Image processing and analysis were performed by using COMOS, TCSM and MPL 1000 software (BioRad). Bay K8644, ATP, αβ-MeATP, UTP and caffeine were applied by pressure ejection from a glass pipette for the period indicated on the records. All experiments were carried out at 26 ± 1 °C.

In some experiments, myocytes were incubated in the presence of 10 μm DI-8-Anepps for 10 min and rinsed twice in physiological solution before performing patch-clamp and [Ca²⁺]_i measurements.

P2X receptor labelling

Myocytes were immunostained as previously described (Macrez-Leprêtre et al. 1997), except that donkey serum was used instead of fetal calf serum. Myocytes were incubated in the presence of anti-P2X1 receptor or anti-P2X4 receptor antibody (1:200 dilution) for 20 h at 4 °C. The secondary antibody (donkey anti-rabbit IgG conjugated to fluorescein-isothiocyanate (FITC), 1:200 dilution) was added for 3 h at 20 °C. Thereafter, myocytes were mounted in Vectashield. Controls without primary antibodies or after inactivation of the antibodies with their antigen peptides were included in each experiment. Images of the stained cells were obtained with the BioRad MRC 1000 confocal microscope.

Chemicals and drugs

Collagenase was obtained from Worthington (Freehold, NJ, USA). Fluo 3 and DI-8-Anepps (pyridinium, 4-[2-[6-(dioctylamino)-2-naphtalenyl]ethyl]-1-(3-sulfopropyl) salt) were obtained from Molecular Probes (Leiden, The Netherlands). Caffeine obtained was from Merck (Nogent sur Marne, France). Bay K8644 was obtained from Bayer (Puteaux, France). Oxodipine was a gift from A. Galiano (Instituto de Investigacion y Desarrollo Quimico Biologico, Madrid, Spain). All nucleotides, noradrenaline and heparin (from porcine intestinal mucosa, MW = 6000) were from Sigma (St Louis, MO, USA). Ryanodine and cyclopiazonic acid were obtained from Calbiochem (Meudon, France). Guanosine-5′-0-(2-thiodiphosphate) was obtained from Boehringer (Mannheim, Germany). The rabbit anti-P2X1 and anti-P2X4 receptor antibodies (Alomone Labs, Jerusalem, Israel) were directed against polypeptides corresponding to residues 382–399 and 370–388 of the rat P2X1 and P2X4 receptors, respectively. The rabbit polyclonal anti-InsP₃R (407143, Calbiochem) and mouse monoclonal anti-RYR (559279, Calbiochem) antibodies were directed against polypeptides corresponding to the last 11 or 13 residues of the C-terminus of the InsP₃R or the RYR. The mouse monoclonal anti-RYR2 antibody (clone C3–33) was from RBI (Natick, MA, USA). The rabbit polyclonal anti-RYR3 antibody was prepared in house and directed against the deduced amino acid sequence between residues 4326 and 4336 (11 amino acids) of rabbit RYR3 (Jeyakumar et al. 1998). For immunological detection, FITC-conjugated affinity-purified donkey anti-rabbit IgG and donkey serum was from Jackson Immunoresearch Laboratories (West Grove, PA, USA) and Vectashield was from AbCys (Paris, France).

Data analysis

Results are expressed as means ±s.e.m. Statistical significance was assessed by means of Student's t test. P < 0.05 was considered statistically significant.

RESULTS

ATP-mediated membrane current in single rat portal vein myocytes

Application of 10 μm ATP to a venous myocyte with the membrane potential held at −60 mV caused an inward current (Fig. 1A). The ATP-induced inward current activated rapidly and then decayed before removal of ATP. The mean current amplitude was 595 ± 45 pA (n = 41). ATP may activate both a non-selective cation current and a Cl⁻ current. Accordingly, the reversal potential under the ionic composition of the solutions used in the present study (18 ± 2 mV, n = 15) did not correspond to the equilibrium potential for Cl⁻, Na⁺ or Ca²⁺ ions. After substituting 105 mm caesium aspartate for 105 mm CsCl in the pipette solution, the Cl⁻ equilibrium potential was around −45 mV. At a holding potential of −60 mV, the Ca²⁺-dependent Cl⁻ current is expected to be strongly reduced. Application of 10 μm ATP evoked a transient inward current (525 ± 55 pA, n = 12) that had a significantly shorter time to half-maximal amplitude (115 ± 30 ms, n = 12) than that recorded with the standard pipette solution (237 ± 45 ms, n = 12). The reversal potential in the low Cl⁻ pipette solution (14 ± 4 mV, n = 12) was not significantly different from that obtained in the standard pipette solution. Finally, when the Na⁺ equilibrium potential was shifted from 67 to 28 mV in 30 mm external Na⁺ concentration, the reversal potential of the ATP-induced current was shifted to −5.1 ± 1.5 mV (n = 8), suggesting that the conductance that underlies this current was, in part, a non-selective cation conductance. The fact that Ca²⁺ influx occurred through the channel opened by ATP was supported by the effects of Ca²⁺-free solution on the ATP-induced increase in [Ca²⁺]_i (see below).

Figure 1. Membrane currents activated in rat portal vein myocytes by external application of ATP.

A, effects of 10 μm ATP, 10 μmαβ-MeATP and 100 μm UTP obtained from three different cells. B, pipette solution contained 2 mm GDPβS and the cell was dialysed with the pipette solution for 5 min before application of 10 μm ATP. C, intracellular application of 10 μg ml⁻¹ anti-P2X1 antibody for 7 min before application of 10 μm ATP. In A–C, the myocytes were held at −60 mV. D, typical Ba²⁺ currents elicited by depolarization to 10 mV from a holding potential of −40 mV and current-voltage relationships obtained in control conditions (•) and after intracellular application of 10 μg ml⁻¹ anti-P2X1 antibody for 7 min (○). Currents are expressed as a fraction of the maximal current (I/I_max) and are the means ±s.e.m. for 7–9 cells.

Because P2Y receptors, but not P2X receptors, are coupled to G proteins (Fredholm et al. 1997), a convenient way to identify P2X receptors is to measure the effect of inhibition of G proteins with GDPβS on the ATP-induced current. Figure 1B shows that including 2 mm GDPβS in the pipette solution had no significant effect on the ATP-induced currents (control cells: 615 ± 50 pA, n = 12; GDPβS-infused cells: 605 ± 45 pA, n = 12). In addition, application of 10 μmαβ-MeATP, which is a potent P2X receptor agonist, induced a current that was similar to that induced by ATP (Fig. 1A), with a mean current amplitude of 562 ± 51 pA (n = 11). By contrast, 100 μm UTP, a potent P2Y receptor agonist, was unable to induce any inward current (Fig. 1A). Thus, we conclude that the pharmacological characterization of P₂ receptors suggests that P2X receptors exist in freshly dissociated rat portal vein myocytes.

A reliable approach to identify P2X receptor subtype expression in a defined cell type is PCR analysis. mRNA was purified from rat portal vein myocytes and reverse transcribed into cDNA. Subtype-specific primers designed by Shibuya et al. (1999) to amplify cDNA were used. As illustrated in Fig. 2A, amplified fragments of P2X1, P2X3, P2X4 and P2X5 receptors were detected in rat portal vein myocytes. A comparison with the published rat P2X receptor sequences indicated 98–100 % identity with the cloned receptors (not shown).

Figure 2. Expression of P2X receptors in rat portal vein myocytes.

A, amplified DNA fragments of P2X receptors (lanes 1–7) from rat brain (a) and rat portal vein myocytes (b) were separated on a 2 % agarose gel and visualized by staining with ethidium bromide. Lane 8, RNA from brain and portal vein myocytes in the absence of reverse transcriptase served as a negative control. Numbers on the left indicate molecular size standards in base pairs (bp). For RNA purification and PCR conditions, see Methods. B, immunostaining of P2X receptor subtypes in portal vein myocytes. Myocytes were stained with anti-P2X1 receptor (a) or anti-P2X4 receptor antibody (b) and vizualization was realized with a donkey anti-rabbit IgG FITC-conjugated antibody. In the absence of primary antibodies or after inactivation of the antibodies by their antigen peptides, only a faint background fluorescence was observed (not shown). Typical confocal sections were performed above the nucleus and therefore appeared spherical. Both P2X1 (a) and P2X4 (b) receptor subtypes were distributed throughout the confocal sections with a marked staining of P2X1 receptor subtype at the cell periphery.

Immunodetection of P2X receptors in 0.5 μm confocal sections from rat portal vein myocytes was performed with the commercially available anti-P2X1 and anti-P2X4 receptor antibodies and the binding sites were revealed with FITC-conjugated secondary antibody. As shown in Fig. 2B, both P2X1 and P2X4 receptors were detected in whole-cell confocal sections. P2X1 receptors formed a dense staining at the cell periphery, whereas P2X4 receptors were homogeneously distributed in the cell sections. Anti-P2X receptor antibodies were applied intracellularly via the patch-clamp pipette to identify the P2X receptors that were responsible for the ATP-induced inward current. Intracellular application of the anti-P2X1 receptor antibody for 7 min inhibited the ATP-induced inward current in a concentration-dependent manner with a maximal inhibition (89 ± 7 %, n = 12) obtained at 10 μg ml⁻¹ (Fig. 1C). The antibody-induced inhibition was specific, because inactivated antibody obtained by pretreatment with the antigen peptide had no significant effect on the ATP-induced inward current (n = 6; not shown). In addition, intracellular application of 10 μg ml⁻¹ anti-P2X1 antibody for 7 min had no effect on either the peak Ba²⁺ current or the current-voltage relationship of the Ba²⁺ current (Fig. 1D). Intracellular application of 10 μg ml⁻¹ anti-P2X4 receptor antibody resulted in a slight, but non-significant, inhibition of the ATP-induced inward current (8 ± 6 %, n = 10). These results suggest that in freshly isolated rat portal vein myocytes, the effects of ATP are dominated by the P2X1 receptor.

ATP-mediated Ca²⁺ responses

External application of ATP either induced rather uniform increases in [Ca²⁺]_i at low ATP concentrations (Fig. 3A) or propagated Ca²⁺ waves at higher ATP concentrations (Fig. 3B and C). The Ca²⁺ responses started from the edges of a cell, but a clear initiation signal, similar to Ca²⁺ sparks, was never observed (n = 82). The ATP-induced Ca²⁺ responses were not affected by inhibition of voltage-gated Ca²⁺ channels through the continuous presence of 10 μm oxodipine (a light-stable dihydropyridine). Plotting the fluorescence signal amplitude as a function of ATP concentration resulted in a concentration-response curve with a concentration corresponding to half-maximal stimulation of about 0.3 μm (Fig. 3D). At 10 μm ATP, the maximal upstroke velocity and maximal propagation rate of the Ca²⁺ wave were estimated to be 25.1 ± 6.5 units s⁻¹ (Δ(F/F₀) s⁻¹) and 68 ± 29 μm s⁻¹ (n = 12), respectively (Fig. 3C). At lower ATP concentrations (0.1 μm or less), localized and transient Ca²⁺ responses were not detected in any of the cells tested (n = 33). When the myocytes were perfused in Ca²⁺-free 0.5 mm EGTA-containing solution for 20–30 s, the ATP-induced Ca²⁺ response was almost completely abolished (n = 7). Similarly, when the myocytes were infused with 10 μg ml⁻¹ anti-P2X1 receptor antibody for 7 min, application of 10 μm ATP was ineffective (n = 5).

Figure 3. Increase in [Ca²⁺]_i evoked by increasing concentrations of ATP.

A–C, Ca²⁺ responses shown as a linescan image (a) and spatial averaged fluorescence (b; F/F₀). Data are for a 2 μm region indicated by the vertical line to the left of the corresponding linescan image. Responses are to 0.1 μm (A), 1 μm (B) or 10 μm ATP (C). D, concentration-response curve for increasing concentrations of ATP, obtained by measuring peak [Ca²⁺]_i (Δ(F/F₀)) in a 2 μm region of the linescan images. Data are means ±s.e.m with the number of cells tested indicated in parentheses. Myocytes were loaded with fluo 3 via the patch pipette and held at −60 mV.

In order to identify more precisely the initiation sites corresponding to Ca²⁺ sparks and ATP-induced Ca²⁺ responses, we stained the plasma membrane with DI-8-Anepps in functional myocytes dialysed with the fluo 3-containing pipette solution. This staining showed the localization of the plasma membrane in the periphery of the cell section, but also the presence of several infoldings inside the cell section, which appeared as dark areas (Fig. 4A). The linescan image obtained from the scanned line shown in Fig. 4Aa revealed spontaneous Ca²⁺ sparks (Fig. 4Ab) that were generated close to the membrane delimiting such infoldings (Fig. 4Ac).

Figure 4. Localization of Ca²⁺ sparks and Ca²⁺ responses induced by ATP, membrane depolarization and Bay K8644.

Aa, vascular myocyte stained with DI-8-Anepps (10 μm) and fluo 3 (60 μm) showing the plasma membrane (red) and cytosol (green). The dashed line corresponds to the scanned line. Ab, linescan image showing Ca²⁺ sparks. Ac, superimposition of Ca²⁺ sparks and DI-8-Anepps staining, illustrating the localization of the Ca²⁺ sparks close to the infolding of the plasma membrane shown in Aa. B, Ca²⁺ sparks triggered by application of 5 nm Bay K8644 and ATP-induced Ca²⁺ response in the same cell, shown as a linescan image (a) and spatial averaged fluorescence (b). Data are for the same 2 μm region indicated by the vertical line to the left of the corresponding linescan image. C, ATP-induced Ca²⁺ response and Ca²⁺ spark induced by a depolarization step from −60 to −10 mV in the same cell, shown as a linescan image (a) and spatial averaged fluorescence (b). Data are for the same 2 μm region indicated by the vertical line on the linescan image. Myocytes were loaded with fluo 3 via the patch pipette and held at −60 mV.

In a myocyte where Ca²⁺ sparks were triggered by external application of 5 nm Bay K8644 (an L-type Ca²⁺ channel activator), a further application (2 min later) of 0.1 μm ATP activated a Ca²⁺ response that started from an edge of the cell rather than at the initiation site of the Ca²⁺ sparks (Fig. 4B). In another myocyte, 0.1 μm ATP was applied first and initiated a homogeneous Ca²⁺ response (Fig. 4C). Two minutes later, membrane depolarization (from −60 to −10 mV) activated a Ca²⁺ spark that gave rise to a propagated Ca²⁺ wave (Fig. 4C). Similar results were obtained in 11 other cells, where Ca²⁺ waves triggered by membrane depolarizations started from Ca²⁺ spark initiation sites, whereas the ATP-induced Ca²⁺ responses did not.

Ca²⁺ release channels involved in ATP-mediated Ca²⁺ responses

To assess whether Ca²⁺ release is a component of the ATP-induced increase in [Ca²⁺]_i, ATP was applied immediately after the Ca²⁺ store had been depleted by 10 mm caffeine. The ATP-induced Ca²⁺ responses were decreased by 85 ± 5 % (n = 12). Similarly, pre-treatment with 10 μm cyclopiazonic acid for 15 min, which depleted the Ca²⁺ store, inhibited the Ca²⁺ response induced by 10 μm ATP in the presence of external Ca²⁺ (n = 9). To identify the Ca²⁺ release channels that were responsible for the ATP-induced Ca²⁺ release, we used anti-RYR and anti-InsP₃R antibodies, which have been shown to be useful to immunologically detect and inhibit RYRs and InsP₃Rs in these cells (Boittin et al. 1999). The effect of the anti-RYR antibody was revealed by the concentration-dependent inhibition of the membrane depolarization (−60 to 10 mV)-induced Ca²⁺ response, with maximal inhibition obtained at 10 μg ml⁻¹ (Fig. 5A). The anti-RYR antibody-induced inhibition was specific, because boiled (95 °C for 30 min) anti-RYR antibody had no significant effect on the membrane depolarization-induced Ca²⁺ response (Fig. 5A). The anti-InsP₃R antibody inhibited the noradrenaline-induced Ca²⁺ response, which is known to depend on InsP₃ generation (Leprêtre et al. 1994), in a concentration-dependent manner (Fig. 5B). The anti-InsP₃R antibody-induced inhibition was specific, as shown by the absence of any effect of boiled anti-InsP₃R antibody on the noradrenaline-induced Ca²⁺ response (Fig. 5B). Susqeuently, we tested the effects of anti-RYR and anti-InsP₃R antibodies on ATP-induced Ca²⁺ responses. As shown in Fig. 5C, intracellular application of 10 μg ml⁻¹ anti-InsP₃R antibody for 7 min had no effect on the ATP-induced Ca²⁺ response, whereas intracellular application of 10 μg ml⁻¹ anti-RYR antibody for 7 min inhibited the ATP-induced Ca²⁺ response. The antibody-induced inhibition was specific, because boiled (95 °C for 30 min) anti-RYR antibody had no significant effect on the Ca²⁺ responses evoked by 10 μm ATP (Fig. 5C). In agreement with these results, intracellular application of 1 mg ml⁻¹ heparin for 5 min had no effect on the ATP-induced Ca²⁺ response, whereas external application of 10 μm ryanodine for 15 min inhibited the ATP-induced Ca²⁺ response (n = 12). These results indicate that the ATP-induced Ca²⁺ release involves specifically the RYRs in freshly isolated rat portal vein myocytes.

Figure 5. Effects of anti-RYR and anti-InsP₃R antibodies on ATP-induced Ca²⁺ responses.

A, peak Ca²⁺ responses evoked by membrane depolarizations (−60 to 10 mV) in control conditions (C) and in the presence of increasing concentrations of anti-RYR antibody or 10 μg ml⁻¹ boiled anti-RYR antibody, each applied intracellularly for 7 min. Data are means ±s.e.m. with the number of cells tested indicated in parentheses. B, peak Ca²⁺ responses evoked by 10 μm noradrenaline in control conditions (C) and in the presence of increasing concentrations of anti-InsP₃R antibody or 10 μg ml⁻¹ boiled anti-InsP₃R antibody, each applied intracellularly for 7 min. Data are means ±s.e.m. with the number of cells tested indicated in parentheses. C, peak Ca²⁺ responses evoked by 10 μm ATP in control conditions (C) and in the presence of 10 μg ml⁻¹ anti-InsP₃R antibody, anti-RYR antibody or boiled anti-RYR antibody, each applied intracellularly for 7 min. Data are means ±s.e.m. with the number of cells tested indicated in parentheses. [Ca²⁺]_i was measured in a 2 μm region of the linescan image. Cells were obtained from three different batches. ⋆, values significantly different from controls (P < 0.05). Myocytes were loaded with fluo 3 via the patch pipette and held at −60 mV.

We have recently shown that Ca²⁺ sparks and global Ca²⁺ responses generated by membrane depolarizations in 1.7 mm Ca²⁺-containing solution require activation of the RYR subtypes 1 and 2, whereas the RYR subtype 3 is not involved (Coussin et al. 2000). These results were obtained by using antisense oligonucleotides targeting each one of the three RYR subtypes. In the present study, we used intracellular applications of specific anti-RYR antibodies directed against RYR1+2+3 (anti-RYR antibody), RYR2 (anti-RYR2 antibody; Xu et al. 1994) or RYR3 alone (anti-RYR3 antibody; Jeyakumar et al. 1998) to assess the effect on the Ca²⁺ responses evoked by ATP, Bay K8644, membrane depolarization and caffeine. First, we tested the effect of the anti-RYR2 antibody on both the number of Ca²⁺ sparks evoked by external application of 5 nm Bay K8644 and the membrane depolarization (−60 to 10 mV)-induced Ca²⁺ response. As shown in Fig. 6, intracellular application of 10 μg ml⁻¹ anti-RYR2 antibody for 7 min strongly inhibited these Ca²⁺ responses. The anti-RYR2 antibody-induced inhibition was similar to that induced by the anti-RYR antibody. By contrast, the Ca²⁺ responses evoked by 10 μm ATP or 10 mm caffeine were almost abolished by the anti-RYR antibody, but were only inhibited by 50 % by the anti-RYR2 antibody (Fig. 7). The inhibitory effects of the anti-RYR and anti-RYR2 antibodies were considered to be specific, because boiled (95 °C for 30 min) anti-RYR receptor antibodies had no significant effect on the frequency of Ca²⁺ sparks or the Ca²⁺ responses induced by membrane depolarizations, ATP and caffeine (Fig. 6 and Fig. 7).

Figure 6. Effects of anti-RYR and anti-RYR2 antibodies on Bay K8644-induced Ca²⁺ sparks and on Ca²⁺ responses induced by membrane depolarization.

A, mean number of Ca²⁺ sparks per linescan image evoked by external application of 5 nm Bay K8644 in control conditions (C) and in the presence of 10 μg ml⁻¹ anti-RYR antibody, anti-RYR2 antibody or boiled anti-RYR2 antibody, each applied intracellularly for 7 min. B, peak Ca²⁺ responses evoked by membrane depolarization (−60 to 10 mV) in control conditions (C) and in the presence of 10 μg ml⁻¹ anti-RYR antibody, anti-RYR2 antibody or boiled anti-RYR2 antibody, each applied intracellularly for 7 min. Data are means ±s.e.m. with the number of cells tested indicated in parentheses. [Ca²⁺]_i was measured in a 2 μm region of the linescan image. Cells were obtained from three different batches. ⋆, values significantly different from controls (P < 0.05). Myocytes were loaded with fluo 3 via the patch pipette and held at −60 mV.

Figure 7. Effects of anti-RYR and anti-RYR2 antibodies on Ca²⁺ responses induced by ATP and caffeine.

A, peak Ca²⁺ responses evoked by 10 μm ATP in control conditions (C) and in the presence of 10 μg ml⁻¹ anti-RYR antibody, anti RYR2 antibody or boiled anti-RYR2 antibody, each applied intracellularly for 7 min. B, peak Ca²⁺ responses evoked by 10 mm caffeine in control conditions (C) and in the presence of 10 μg ml⁻¹ anti-RYR antibody, anti-RYR2 antibody or boiled anti-RYR2 antibody, each applied intracellularly for 7 min. Data are means ±s.e.m. with the number of cells tested indicated in parentheses. [Ca²⁺]_i was measured in a 2 μm region of the linescan image. Cells were obtained from three differents batches. ⋆, values significantly different from controls (P < 0.05). Myocytes were loaded with fluo 3 via the patch pipette and held at −60 mV.

We have shown previously that RYR3 is capable of being activated in Ca²⁺-overloaded myocytes (Mironneau et al. 2001). Therefore, we used Ca²⁺-overloaded myocytes (bathed in 10 mm[Ca²⁺]_o for 1 h) in the present study to test the effect of the anti-RYR3 antibody. As illustrated in Fig. 8A, intracellular application of 10 μg ml⁻¹ anti-RYR3 antibody reduced the amplitude of the caffeine-induced Ca²⁺ response obtained in 10 mm[Ca²⁺]_o. Interestingly, in Ca²⁺-overloaded myocytes, the amplitude of the caffeine-induced Ca²⁺ response in the presence of the anti-RYR3 antibody was similar to that obtained in control cells superfused with 1.7 mm[Ca²⁺]_o (Fig. 8A). The antibody-induced inhibition was specific, because boiled (95 °C for 30 min) anti-RYR3 antibody had no effect on the caffeine-induced Ca²⁺ response in Ca²⁺-overloaded myocytes. In myocytes superfused with 1.7 mm[Ca²⁺]_o, intracellular application of 10 μg ml⁻¹ anti-RYR3 antibody had no significant effect on ATP- and caffeine-induced Ca²⁺ responses (Fig. 8B), or the number of Ca²⁺ sparks evoked by external application of 5 nm Bay K8644 (data not shown). These results indicate that in 1.7 mm Ca²⁺-containing solution, the Ca²⁺ responses evoked by Ca²⁺ influx through L-type Ca²⁺ channels or non-selective cation channels require RYR2, but not RYR3.

Figure 8. Effects of anti-RYR3 antibody on Ca²⁺ responses induced by ATP and caffeine.

A, peak Ca²⁺ responses evoked by 10 mm caffeine in 1.7 mm[Ca²⁺]_o (C1), 10 mm[Ca²⁺]_o (C2) and in the presence of 10 μg ml⁻¹ anti-RYR3 antibody or boiled anti-RYR3 antibody (applied intracellularly for 7 min). ⋆, values significantly different from controls in 1.7 mm[Ca²⁺]_o (P < 0.05). B, peak Ca²⁺ responses evoked by 10 μm ATP and 10 mm caffeine in control conditions (C) in 1.7 mm[Ca²⁺]_o and in the presence of 10 μg ml⁻¹ anti-RYR3 antibody (applied intracellularly for 7 min). Data are means ±s.e.m. with the number of cells tested indicated in parentheses. Cells are obtained from two different batches. Myocytes were loaded with fluo 3 via the patch pipette and held at −60 mV.

DISCUSSION

Electrophysiological experiments and confocal Ca²⁺ imaging of rat portal vein myocytes in the present study revealed that the ATP-induced Ca²⁺ response is mediated by activation of P2X1 receptors and involves Ca²⁺ influx through non-selective cation channels, which is further amplified by Ca²⁺ release through RYRs of the sarcoplasmic reticulum. We were unable to detect typical Ca²⁺ sparks, suggesting that P2X1 receptors are not colocalized with clusters of RYRs, in contrast to L-type Ca²⁺ channels. Using anti-RYR subtype antibodies, we showed that RYR2 is required for ATP-induced Ca²⁺ responses, but not RYR3.

The involvement of P2X receptors in mediating a large inward current at the resting membrane potential and Ca²⁺ influx has been described previously (Benham & Tsien, 1987; Lewis et al. 2000). This study shows that in freshly dissociated rat portal vein myocytes, the ATP-induced Ca²⁺ response is mediated by P2X1 receptors. Several lines of evidence support this conclusion. First, αβ-MeATP was as effective as ATP in inducing inward current, whereas UTP had no effect. The reversal potential of the ATP-induced current was modified by a shift in the Na⁺ equilibrium potential, but not in the Cl⁻ equilibrium potential, suggesting that it is a non-selective cation current. Second, infusion of 2 mm GDPβS into the myocytes did not prevent the response to ATP, supporting the idea of a receptor-gated channel. Third, RT-PCR experiments revealed the expression of several P2X receptor isoforms (P2X1, P2X3, P2X4 and P2X5). However, infusion of the anti-P2X1 receptor antibody suppressed the electrophysiological and Ca²⁺ responses to ATP, whereas the anti-P2X4 receptor antibody was ineffective. In addition, immunostaining showed a marked localization for P2X1 receptors at the cell periphery, although both P2X1 and P2X4 receptors were detected in whole-cell sections. These results strongly suggest that the P2X1 receptors dominate the ATP-mediated Ca²⁺ response in rat portal vein myocytes. This concurs with recent data obtained from the vas deferens of P2X1 receptor-deficient mice (Mulryan et al. 2000) and from other smooth muscles (Lewis & Evans, 2000; Lewis et al. 2000; Vial & Evans, 2000). Although the presence of multiple P2X receptor isoforms has been reported at the cellular level in various vessels (Burnstock, 1997; Nori et al. 1998; Lewis et al. 2000), the role of each isoform in the control of smooth muscle function remains to be elucidated.

Previous studies have indicated that activation of P2X receptor-activated cation current leads to an increase in [Ca²⁺]_i in different cell types, including macrophages (Picello et al. 1990), thymocytes (Pizzo et al. 1991), oocytes (Nuttle & Dubyak, 1994), neuroblastoma xNG108–15 cells (Brater et al. 1999) and smooth muscle cells (Benham & Tsien, 1987). In addition, Ca²⁺ influx through L-type Ca²⁺ channels can be triggered in response to ATP-induced membrane depolarization. In our experiments, vascular myocytes were held at −60 mV, a potential at which the probability of voltage-dependent Ca²⁺ channels being open is very low, and oxodipine (an inhibitor of L-type Ca²⁺ channels) was added in most experiments. The ATP-induced increase in [Ca²⁺]_i was inhibited: (1) after intracellular application of the anti-P2X1 receptor antibody; (2) in the absence of extracellular Ca²⁺; (3) in the presence of cyclopiazonic acid; or (4) after application of caffeine to empty the intracellular Ca²⁺ store. These results support the idea that the ATP-induced increase in [Ca²⁺]_i is triggered by Ca²⁺ influx through P2X1 receptors and secondarily depends on Ca²⁺ release from the intracellular store. Interestingly, confocal Ca²⁺ imaging revealed that the initiation sites of ATP-induced Ca²⁺ responses are always located close to the plasma membrane, but do not correspond to the initiation sites of Ca²⁺ sparks detected in the same scanned lines. Localized Ca²⁺ events were never detected in response to ATP, particularly at low concentrations. These observations suggest that the Ca²⁺ influx through ATP-activated cation channels is large enough to activate isolated RYRs, which, in turn, activate neighbouring RYRs and then elicit a propagated Ca²⁺ response. We observed that when the ATP-induced Ca²⁺ response passed through a Ca²⁺ release unit previously revealed by a Ca²⁺ spark, the propagation rate was only increased locally. This is in contrast to Ca²⁺ waves activated by membrane depolarization or Bay K8644, which started preferentially from initiation sites at which Ca²⁺ sparks had been detected previously in the same scanned lines. Opening of cation channels in toad stomach smooth muscle has been reported to induce a localized Ca²⁺ event with a Ca²⁺ spark-like appearance (Zou et al. 1999), but this has been obtained with high speed digital imaging. Although co-immunostaining of L-type Ca²⁺ channels and P2X1 receptors was not performed in these vascular myocytes, our Ca²⁺ imaging experiments suggest that the two types of channel could be located in different areas of the plasma membrane. Recent data have shown that caveolae contain L-type Ca²⁺ channels and that depletion of caveolae by dextrin decreases the frequency, amplitude and spatial size of Ca²⁺ sparks, which suggests that caveolae are structural elements for the generation of Ca²⁺ sparks in arterial and cardiac myocytes (Lohn et al. 2000). Caveolae have also been suggested to play a vital role as organized centres of signal transduction, where signalling molecules are highly concentrated (Okamoto et al. 1998; Parton et al. 2000). By contrast, plasma membrane receptors would appear to be preferentially located on the sarcolemma and, therefore, could be more accessible to extracellular factors, such as mediators and hormones.

We have shown in a previous study that the three RYR subtypes are expressed in rat portal vein myocytes (Coussin et al. 2000). Based on an antisense strategy, it has been proposed that both RYR1 and RYR2 participate in Ca²⁺ sparks and waves elicited by membrane depolarization or application of Bay K8644. Using anti-RYR subtype antibodies, we have confirmed here that in 1.7 mm[Ca²⁺]_o, the RYR3 subtype is required for neither activation of Ca²⁺ sparks nor ATP- and caffeine-induced Ca²⁺ responses. Application of an anti-RYR2 antibody inhibited both the ATP- and caffeine-induced Ca²⁺ responses by about 50 %, whereas application of an antibody directed against the three RYR subtypes almost completely suppressed these responses. Taken together, these data indicate that activation of RYR2 is needed to induce a full Ca²⁺ response to ATP and that RYR1 might participate in the ATP-induced Ca²⁺ response. These observations are in good agreement with our previous studies obtained using antisense oligonucleotides, which showed that both RYR1 and RYR2 are required for the caffeine-induced Ca²⁺ response (Coussin et al. 2000).

The propagation rate of Ca²⁺ waves induced by 10 μm ATP is similar to that evoked by L-type Ca²⁺ current (Arnaudeau et al. 1997). In agreement with Niggli's hypothesis (Niggli, 1999) and our proposal that Ca²⁺ wave propagation in vascular myocytes is linked to the progressive recruitment of isolated and neighbouring RYRs following the sustained activity of a sparking site, we propose that isolated RYRs exhibit a high Ca²⁺ sensitivity, but a low Ca²⁺ flux. This is in contrast to Ca²⁺ sparks triggered by the opening of L-type Ca²⁺ channels, which might be due to activation of RYRs that exhibit a low Ca²⁺ affinity, but a high Ca²⁺ flux (Niggli, 1999). However, this explanation seems unlikely, because the same RYR subtypes (namely, RYR1 and RYR2) are involved in both types of Ca²⁺ response in vascular myocytes. Therefore, it can be envisaged that clusters of RYRs would be located in those particular cell areas where L-type Ca²⁺ channels are present, whereas ATP-gated channels would present a homogeneous distribution on the plasmalemma. Another possibility is that the opening of an L-type Ca²⁺ channel acts as a point-source of Ca²⁺ in a microdomain and generates a very high Ca²⁺ concentration in comparison to Ca²⁺ influx through P2X1 receptor-gated channels. Assuming that ∼6 % of the ATP-activated cation current is carried by Ca²⁺ ions (Schneider et al. 1991), the Ca²⁺ component of the ATP-induced current is around 50 pA, which is similar to the peak L-type Ca²⁺ current (Arnaudeau et al. 1997) and, therefore, does not support the aforementioned hypothesis. More experiments are needed to define the precise location of L-type Ca²⁺ channels, membrane receptors and RYR subtypes in vascular myocytes and to elucidate their role in Ca²⁺ responses induced by various neuromediators and hormones.

In conclusion, the results of the present study showed that, in vascular myocytes, Ca²⁺ responses can be activated by Ca²⁺ influx through P2X1 receptors. ATP might trigger Ca²⁺-induced Ca²⁺ release at intracellular sites where RYRs are not clustered, by activating RYR2, but not RYR3.