INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

ACTIVATION OF PURINERGIC RECEPTORS on the vascular endothelium triggers a wide spectrum of physiological actions including an increase in cytosolic calcium that may be followed by secretion of nitric oxide, ATP, prostacyclin, endothelin, or other substances. Furthermore, lengthy exposure to ATP is reported to produce apoptotic or necrotic cell death (6, 20, 30). Many of the actions of ATP have been attributed to members of the P2Y family of metabotropic purinoreceptors, which have been well documented in the endothelium (5, 14, 36).

Only recently has the presence of members of the P2X family of ionotropic purinergic receptors been reported in the endothelium (10, 11, 13, 21, 41, 42), and there is still little functional characterization of the receptors in these cells. P2X receptors are calcium-permeable, cation-selective channels with diverse pharmacological and desensitization phenotypes that are expected to affect their role in cell signaling (1, 2, 16, 31, 32, 38). In addition, several members of the P2X family, most notably P2X7 but also P2X2-4, have the interesting capacity to initiate a second state of high conductance (pores) when activated by agonist (15, 25, 26, 37). In the case of P2X7, this may lead to sustained calcium influx and cell death.

In previous studies, we characterized the ion channels of bovine aortic endothelial cells (BAECs) that open in response to ATP-induced depletion of calcium stores via G protein-coupled P2Y receptor activation (35). In the course of those studies, it became apparent that ATP also activated a second calcium-permeable ion channel with different characteristics than the store-operated channel (SOC). With the P2X receptor/channels as likely candidates, the present study had the following two objectives: to obtain electrophysiological characterization of the structurally identified receptors in BAECs and to establish whether, in addition to activation of small cation channels, pore formation occurred in these cells in response to agonist.

We show that BAECs contain mRNA for P2X4, P2X5, and P2X7 and not for the other P2X receptors. Channels with the pharmacological and electrophysiological properties of P2X7 and P2X4 are both functionally present in the surface membrane. Channels with properties particular to P2X5 were not seen, presenting the possibility that P2X5 exists in combination with P2X4 as a heteromultimer that expresses the properties of P2X4. Pore formation was not present under a variety of conditions that elicit permeability to large molecules such as N-methyl-D-guanine (NMDG) and YO-PRO {quinolinium, 4-[3-(3-methyl-2(3H)-benzoxazolylidene)- 1-propenyl]-1-[3-(trimethylammonio)propyl]-diiodide} in other cell types in response to ATP.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Tissue. BAECs were obtained from Cell Systems (Kirkland, WA). They were grown in Dulbecco's modified Eagle's medium (GIBCO-BRL) supplemented with 10% fetal calf serum, 100 U/ml penicillin, and 100 µg/ml streptomycin. For mRNA preparation, adult rats placed under halothane anesthesia were decapitated, and the nodose ganglia and brains were removed in compliance with Case Western Reserve University Animal Research Committee guidelines.

mRNA extraction. mRNA from BAECs and the rat brain and nodose ganglia was isolated using the MicroPoly (A) Pure kit (Ambion) following the manufacturer's instructions. Poly(A⁺) mRNA was quantitated by spectrophotometric absorbence at 260 nm and stored in aliquots at 80°C. Human brain Poly(A⁺) RNA was purchased from Clontech Laboratories.

PCR amplification of ATP receptor cDNA fragments: primer design. The set of specific primers was designed to amplify unique DNA fragments corresponding to regions highly conserved in human and rat P2X purinoreceptors by RT-PCR. Primer sequences and locations from published cDNA sequences in GenBank of the National Center for Biotechnology Information were as follows: P2X1 (Accession Nos. AF020498 and RNP2XMR), 5'-AGC ATC AGC TTT CCA CGC TTC AAG GTC-3' (sense primer, nucleotides 850-876 and 783-809) and 5'-CTT GTA GTA GTG CCT CTT AGG CAG GAT G-3' (antisense primer, nucleotides 1,368-1,341 and 1,301-1,274); P2X2 (Accession Nos. NM_012226.1 and RNU14414), 5'-TTC GTG TGG TAC GTA TTC ATC GTG CAG-3' (sense primer, nucleotides 130-156 and 166-192) and 5'-TTG GGG CCA TCG TAC CCA GAA ATT GG-3' (antisense primer, nucleotides 544-519); P2X3 (Accession Nos. AB016608 and RNRNAP2X3), 5'-ACC AAG TCG GTG GTT GTG AAG AGC TG-3' (sense primer, nucleotides 37-62 and 208-233) and 5'-ACC CAG CCG ATC TTA ATA CCC AGA AC-3' (antisense primer, nucleotides 737-712 and 908-883); P2X4 (Accession Nos. HSU83993 and RNRNAP2X4), 5'-TAT CCA GAT CAA GTG GGA CTG CAA C-3' (sense primer, nucleotides 1,072-1,095 and 777-801) and 5'-TCT TCA TGC AGT AGA GGA CTA TGA C-3' (antisense primer, nucleotides 1,396-1,373 and 1,102-1,078); P2X5 (Accession Nos. RNRNAP2X5 and AF005156), 5'-GAA TGG GAC TGT GAC CTT GAT AAA GC-3' (sense primer, nucleotides 962-987 and 272-297) and 5'-GAG GTA GAT AAG TAC CAG GTC GCA G-3' (antisense primer, nucleotides 1,267-1,242); and P2X6 (Accession Nos. XM_009854.1 and RNRNAP2X6), 5'-ACC CAC AGG ACC TGT GAG ATC TGG AG-3' (sense primer, nucleotides 514-539 and 488-513) and 5'-TCC TCC AGT AGA AAC CGG CCT CTC TAT C-3' (antisense primer, nucleotides 1,129-1,102 and 1,103-1,076). For P2X7, because of the difference between rat and human P2X7 receptors, a human/Boss taurus-like pair of oligos was designed to amplify this protein (Accession Nos. NM_002562.1 and AF083073), 5'-AAG AGC CTG TCA TCA GTT CTG TGC AC-3' (sense primer, nucleotides 187-212 and 161-186) and 5'-AGA TCT CAA TGC CCA TGA TTC CTC CC-3' (antisense primer, nucleotides 795-774). For RT-PCR, 100 ng poly-A mRNA was heat denatured at 70°C for 5 min and then reverse transcribed into first-strand cDNA using a mixture of random hexameric, unlabeled deoxynucleotides and MuLV Reverse Transcriptase (First-Strand cDNA Synthesis kit, Perkin-Elmer) at 42°C for 1 h. The first-strand cDNA products were used directly as templates for PCR amplification. PCR was performed with the Advantage PCR System (Clontech) as follows. Each PCR vessel contained 5 pmol each of the sense and antisense primers, 0.2 mM deoxynucleotides, and 1 unit AdvanTaq in a final volume of 25 µl. The amplifications were performed using the following cycling program: one cycle (2 min at 95°C), 35 cycles (15 s at 94°C, 15 s at 55°C, and 1 min at 68°C), and one cycle (10 min at 72°C). PCR products of P2X1-7 purinoreceptors from endothelial cells, the nodose ganglia (P2X1-6), and the brain (P2X7) were resolved by electrophoresis on 1.2% agarose gels and transferred to BrightStar-Plus nylon membranes (Ambion) as recommended by the manufacturer. After the transfer, the membranes were baked at 80°C in a vacuum oven. Subunit- specific PCR products were identified by hybridization to radiolabeled internal oligonucleotides specific for each one of the P2X receptors. The internal oligonucleotides were as follows: P2X1, 5'-TCA CCT CTT CAA GGT GTT TGG GAT TC-3'; P2X2, 5'-AGA GCT CCA TCA TCA CCA AGG TCA AGG GGA TCA C-3'; P2X3, 5'-ACT ACA GCT CTG TTC TCC GGA CCT GTG-3'; P2X4, 5'-GGC ATC CGC TTT GAC ATC ATC GTG TTT G-3'; P2X5, 5'-CTG ATG AAA GCC TAC GGG ATC CGC TTT G-3'; P2X6, 5'-GTG TTC CGC ATT GGG GAC CTC GTG G-3'; and P2X7 5'-TAA AAA GGG ATG GTT GGG CCC GCG GAG CAA AG-3'. Radiolabeling was accomplished with T4 polynucleotide kinase (Promega) in the presence of [-³²P]ATP (Amersham). After the blots were prehybridized in Ultrasensitive Hybridization solution (ULTRAhyb, Ambion) at 42°C for 1 h, ³²P-labeled probe (at 10⁶ cpm/ml) was added and hybridized at 42°C for up to 16 h. After hybridization, the blots were washed first with 2× saline-sodium citrate (SSC)-0.1% SDS at room temperature followed by a more stringent wash of 0.1× SSC-0.1% SDS at 44°C. Blots were exposed to BioMaxMS film (Kodak) at 80°C with an intensifying screen.

Western blot. Samples of BAECs and the rat brain and nodose ganglia were homogenized in 10 volumes of lysis buffer (1% Triton X-100, 150 mM NaCl, 50 mM Tris, and 1 mM EDTA at pH 7.5) containing freshly added protease inhibitor cocktail (Complete, Roche Molecular Biologicals) with 50 mM sodium fluoride and 0.2 mM sodium vanadate. The homogenate was incubated on ice for 1 h. Debris and insoluble material were pelleted by centrifugation at 3,000 g for 10 min at room temperature. The concentration of protein in the supernatant was quantified using the bicinchoninic acid method (Pierce). For the Western blot experiments with the P2X7 receptor, a gradient fractionation from BAECs was made, and a sample from the membrane protein fraction (7.5 µg) was used for immunoblots. To prepare membranes from cultured cells, monolayers in 100-mm tissue culture dishes were washed twice with cold PBS, collected in 1 ml cold PBS by scraping, and pelleted in microfuge tubes at 1,000 g for 5 min. Cells were lysed by resuspension in a hypotonic lysis buffer (250 ml/cells from one 100-mm dish) consisting of (in mM) 10 HEPES (pH 7.9), 10 KCl, 1.5 MgCl₂ plus a protease inhibitor cocktail (Complete, Roche Molecular Biologicals), 50 NaF, and 1 Na₃VO₄. After a 15-min incubation period on ice, cells were passed 10 times through a 26-gauge needle. Nuclei and unbroken cells were removed by spinning at 1,000 g for 10 min at 4°C. The supernatant was spun at 10,000 g (10 min at 4°C) to remove mitochondria, after which membranes were pelleted by spinning at 50,000 g for 1 h. The supernatant was saved as the cytosolic fraction. The membrane pellet was washed once by resuspension in lysis buffer and recentrifuged to remove cytosolic contaminants. The final membrane pellet was solubilized in an isotonic lysis buffer containing 1% Triton X-100. After a 30-min incubation period on ice, the insoluble debris was removed by centrifugation at 20,000 g for 10 min. Lysate from the rat brain (1 µg) was used as a positive control. For P2X4, a sample from rat nodose ganglia lysate was used as a positive control and run in parallel with total protein from BAECs (10 µg/line). In both cases, the samples of protein were separated on 7.5% polyacrilamide SDS gels and transferred to polyvinylidene difluoride (PVDF) membranes. The PVDF membranes were blocked with 5% nonfat dry milk in PBS-0.1% Tween 20 (PBS-T) overnight at 4°C and then incubated with primary antibodies, polyclonal anti-P2X7 and anti-P2X4. We used commercial antibodies, anti-P2X4 (Alomone Labs) and anti-P2X7 (Lots 01-03, Alomone Labs). Anti-P2X4 is a polyclonal antibody raised in the rabbit against the highly purified peptide KKYKYVEDYEQGLSGEMNQ, corresponding to residues 370-388 of the rat P2X4 receptor with an additional NH₂-terminal cysteine. The rabbit polyclonal anti-P2Z/P2X7 antibody was raised against the synthetic peptide corresponding to the last 20 amino acids of the P2X7 (576-595) protein, KIRKEFPKTQGQYSGFKYPY. After the wash with PBS-T (30 min), blots were incubated 1 h at room temperature with anti-rabbit horseradish peroxidase-linked secondary antibody (Amersham Pharmacia) in blocking buffer. After the wash, the blots were developed using the ECL-PLUS kit (Amersham), and the images were captured on Hyperfilm-ECL.

Immunohistochemistry. Endothelial cells were seeded on glass slides, grown to confluency, washed with PBS, and fixed with paraformaldehyde (3% in PBS) with 0.1% Triton X-100. After 20 min, the cells were washed with physiological saline and incubated in 10% normal goat serum with 0.1% Triton X-100 (PBS-3% albumin) for 30 min. The cells were then incubated for 1 h with the anti-P2X4 or anti-P2X7 polyclonal antibody (1:100 dilution in PBS-albumin). Cells were washed in physiological saline and incubated with the secondary antibody (fluorescein or rhodamine isothiocyanate-conjugated goat anti-rabbit IgG at 1:100, Jackson Labs) for 1 h. Cells were mounted with Vectashield (Vector Labs), coverslipped, and examined with an Olympus BH-2 microscope. Images were obtained with a Spot 2.1 digital camera (Diagnostic Instruments) or examined using confocal microscopy (Leica).

Electrophysiology. Whole cell and outside-out patch configurations were used. The current-voltage relationship was obtained using either a voltage ramp (from 120 to +60 mV) or a series of step depolarizations (from 100 to +40 mV) from a holding potential of 60 mV. A BAEC monolayer was mechanically dissociated to single cells and replated. Single cells were studied after allowing 10 min for reattachment. Whole cell and outside-out recordings were made using an Axopatch-1C amplifier (Axon instruments). Patch pipettes (3-7 M) were made from borosilicate glass 7052 (Garner Glass). Data were digitized at 50-100 µs/point, and the results were analyzed using the pCLAMP analysis package (Axon Instruments). Four solutions were used in the electrophysiological experiments. Normal physiological solution contained (in mM) 147 NaCl, 2 KCl, 2 CaCl₂, 1 MgCl₂, 10 HEPES, and 12 glucose (pH 7.3). The low divalent solution contained (in mM) 147 NaCl, 2 KCl, 0.3 CaCl₂, 0 MgCl₂, 10 HEPES, and 12 glucose (pH 7.3), and the sodium-free low divalent solution contained (in mM) 147 NMDG chloride, 2 KCl, 0.3 CaCl₂, 0 MgCl₂, 10 HEPES, and 12 glucose (pH 7.3). CaCl₂ (2.0 mM) and MgCl₂ (1 mM) were added to the latter solution for a sodium-free normal divalent solution. The intracellular pipette solution contained (in mM) 154 CsCl or 150 CsMeSO₃, 10 EGTA, and 5 HEPES (pH 7.2 adjusted with CsOH). 2',3'-O-(4-benzoylbenzoyl)ATP (BzATP) and ATP-K were obtained from Sigma. ,-Methylene-ATP, suramin, 2-methyl-thio-ATP (2-MeSATP), and reactive blue were purchased from RBI. Agonists and antagonists were prepared fresh unless previously prepared as stock solutions and frozen in aliquots. They were defrosted and diluted in the buffer just before use. The agonists and antagonists were applied from a large multibore pipette placed close to the cell just before application.

YO-PRO uptake. Confluent monolayers of BAECs plated on glass chips were incubated in the presence of 3 µM YO-PRO-3 iodide (Molecular Probes, molecular weight 655) in either low or normal divalent buffer for 10 min at 22 or 37°C. At the end of this time, BzATP (100 µM) and YO-PRO (3 µM) in the same buffer replaced the incubation solution. Monolayers remained in this solution for 1-60 min. At the end of the incubation times, the monolayer was washed with buffer, and the field of cells was examined for YO-PRO fluorescence (excitation 612 nm; emission 631 nm).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Detection of P2X receptors subunits in BAECs. Specific pairs of oligonucleotides designed for unique regions of each subunit identified mRNA for P2X4, P2X5, and P2X7 in BAECs. mRNA for P2X1, P2X2, P2X3, or P2X6 was not detected. In controls, mRNA for P2X1-P2X6 was present in the nodose ganglia (4) and for P2X7 in the brain (33). PCR products of the expected size (325, 305, and 608 bp) were detected in blots hybridized with internal oligonucleotide probes of P2X4, P2X5, and P2X7 with reverse transcriptase present but not when it was absent (Fig. 1). Protein from BAECs and the rat nodose ganglia or brain was examined by Western blot for the presence of P2X4 and P2X7 (Fig. 2). An antibody against P2X5 was not available. Protein bands of ~60 kDa were recognized in both samples for P2X4 (19, 40). For Western blot analysis of P2X7 expression, a membrane-enriched fraction from BAECs was run in parallel with a sample of rat brain lysate. In BAECs, as in other tissues (3), a distinct protein band of ~70 kDa is recognized by the antibody. In the brain, two strong bands are recognized by the antibody: one in the expected range of 70 kDa and a smaller one around 60 kDa. In the cerebellum and hippocampus, two bands (67 and 79 kDa) have also been reported (17). The smaller one was attributed to a nonglycosylated form.

View larger version (62K): [in this window] [in a new window]   Fig. 1.   Expression of P2X receptor mRNA in bovine aortic endothelial cells (BAECs). Oligonucleotide primers designed to specifically amplify P2X1-7 were used in RT-PCR reactions from the rat nodose ganglia (P2X1-6) or human brain (P2X7) and BAECs Poly(A+) RNA. As a control, first-strand cDNA reactions were performed either with (+) or without () reverse transcriptase (RT). In all reactions, the RT lanes had no signal. The sizes of the DNA fragments were as follows: 518, 414, 676, 325, 305, 615, and 608 bp for P2X1, P2X2, P2X3, P2X4, P2X5, P2X6, and P2X7, respectively. mRNA for P2X4 and P2X7 and a weaker signal for P2X5 was detected.

View larger version (50K): [in this window] [in a new window]   Fig. 2.   Western blot analysis. A: Western blot analysis of P2X7 expression in membrane-enriched fractions from BAECs. A sample of membrane protein from BAECs (7.5 µg) was run in parallel with a sample of rat brain lysate (1 µg). In BAECs, the expected protein band of 70 kDa was recognized. Two much weaker larger bands were also evident. In the brain, two strong bands were recognized, one at ~70 kDa and a second at ~60 kDa. B: samples of total protein (10 µg/lane) from rat nodose ganglia and BAECs were examined by Western blot with polyclonal anti-P2X4. Protein bands of ~60 kDa were recognized in both samples. Molecular mass markers (in kDa) are indicated to the left.

Electrophysiological characterization of the response to ATP. To correlate the phenotype of the functional response with the three P2X receptors we identified by RT-PCR, we measured whole cell currents and outside-out patch single channel currents elicited in response to MeSATP (10-20 µM), an agonist for both P2X4 and P2X5, whereas BzATP (10-100 µM) was used as an agonist for the P2X7 receptor (24, 25). Although BzATP is not uniquely specific for P2X7, the two agonists activated distinctly different currents in the BAECs as shown below. With the exception of two experiments, all cells had both responses. In those two experiments (n = 6 cells) in late passage cells (passage 16), the response to MeSATP was normal, but a response to BzATP was absent. Functional studies of purinergic responses in BAECs can be complicated by the presence of P2Y1, P2Y2, and P2Y11 metabotropic receptors, which also respond to MeSATP. Thus electrophysiological studies were confined to single cells by using open patch recording and outside-out patches to eliminate currents from G protein-linked P2Y receptors, which deplete calcium stores and open the SOC. This channel is not elicited by agonist in isolated patches nor in the open patch whole cell mode, where, presumably, second messengers such as D-myo-inositol 1,4,5-trisphosphate [Ins(1,4,5)P₃] that lead to store depletion and activation of the channel have been diluted or lost (35). And, as will be shown below, the channels formed by the P2X receptors have different properties than the SOC.

P2X4 and P2X5. Isolated individual endothelial cells were held at a membrane potential of 60 mV and voltage ramps from 120 to +60 mV were applied at 5-s intervals to obtain the current-voltage relationship in control bathing solution. The cells were then activated with MeSATP (10-20 µM), and voltage ramps were repeated. In a simultaneous set of experiments, we used ATP in a concentration range (10-100 µM) that would be expected to activate P2X4 and P2X5 but not P2X7. These concentrations elicited current with pharmacology and kinetics equivalent to that of MeSATP. In the presence of either MeSATP or ATP, an inward current developed (Fig. 3, A and B). The current-voltage relationship for the agonist-sensitive current was obtained by subtraction of the control current from that obtained near the peak of the response. The profile of the agonist-activated response was an inwardly rectifying current that is characteristic of P2X4 as expressed in heterologous expression systems (2, 8, 32). The response was transient, decaying to control values within 30 s in the continued presence of either agonist. Application of ATP immediately after decay of the MeSATP-activated current did not elicit a further response (Fig. 3A, arrow). The current could be fully recovered only after intervals of 3-4 min (Fig. 4A). The reversal potential was close to 0 mV (ATP: 2.6 ± 1.8, n = 5; MeSATP: 4.0 ± 2.8, n = 5). The current amplitude varied among cells with 271 ± 75 pA at 100 mV (range 40-605, n = 8) for 20 µM MeSATP and 139 ± 40 pA (range 25-340, n = 7) for 10 µM ATP. In six of these cells, CsCl in the pipette was substituted by CsMeSO₃ to eliminate the possibility that chloride carried the current. Importantly, suramin (20-300 µM) in the presence of 20 µM MeSATP or 10-100 µM ATP did not block the agonist-activated current, consistent with the reported insensitivity of P2X4 to this antagonist (2, 32). ,-MeATP (100 µM) elicited either no response (n = 2) or <5% of the MeSATP response (n = 2), as seen in Fig. 4B. Reactive blue at 60 µM was also ineffective in blocking the agonist current (Fig. 4C).

View larger version (22K): [in this window] [in a new window]   Fig. 3.   ATP- and 2-methyl-thio-ATP (MeSATP)-activated currents. The agonist-induced current response recorded from an isolated cell held a membrane potential of 60 mV and was transient, decaying to near control values within 30 s in the continued presence of 20 µM MeSATP (A, left) or 10 µM ATP (B, left). Right, current-voltage relationship of the agonist-induced current obtained in response to ramp stimuli delivered at 5-s intervals. The agonist-sensitive component was isolated by subtracting the current response to a ramp voltage before the application of agonist (1) from that near the peak of the response (2). The reversal potential in both cases was near 0 mV. The arrow in A, left, indicates the subsequent application of 100 µM ATP after MeSATP. There was no response to the ATP.

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Characterization of the MeSATP and ATP responses. A: a 215-ms recovery interval between repeated applications of 10 µM ATP elicited a full response to the second application of ATP. B: application of 100 µM ,-methylene-ATP (,-MetATP) elicited no current response in the same cell in which ATP (100 µM) elicited inward current. The responses to the ramp stimuli are truncated. C: current elicited by 100 µM ATP was unaffected in the presence of 60 µM reactive blue (RB), as shown by the overlap of the current-voltage relationships in the presence and absence of RB. The currents were obtained by subtraction of control current from that in the presence of ATP or ATP plus RB. The cells were held at 60 mV between ramp stimuli delivered at 5-s intervals.

View larger version (21K): [in this window] [in a new window]   Fig. 5.   MeSATP and ATP activate channels with similar conductances in outside-out patches. A: examples of currents elicited in response to a 380-ms ramp stimulus from 120 to +60 mV in the presence of MeSATP (20 µM; data filtered at 500 Hz) and ATP (10 µM, data filtered at 1 kHz). B: single-channel amplitudes (n = 6 patches) were plotted against voltage, and a linear least-squares fit gave a conductance of 36 ± 3 pS. C: examples of channel openings in response to MeSATP (20 µM) at the holding potential of 60 mV during the intervals between ramps in a patch with only one apparent active channel.

P2X7. In contrast to the MeSATP-activated current described above, BzATP-activated currents in BAECs decayed toward a sustained value in the continued presence of the agonist (Fig. 6A). The reversal potential remained near 0 mV whether the pipette contained Cl or MeSO (n = 3), thus eliminating chloride as the current-carrying species. Furthermore, this current was not carried by the large cations that are characteristic of the pore formation described for P2X7 in other cell types. In the continued presence of BzATP with the membrane potential held at 60 mV, NMDG⁺, substituted for Na⁺ in the bath, eliminated the BzATP-activated inward current (Fig. 6). During a second application of the NMDG bathing solution, the holding potential was briefly switched from 60 mV through 0 to +60 mV. The extrapolated zero current potential was shifted toward negative values, as shown in Fig. 6B. These currents were obtained in low divalent bath solution (0.3 mM Ca²⁺ and no added Mg²⁺) because the NMDG-permeable pore formation is favored by the low divalent solution (26). Similar results were obtained in two additional cells. Further studies at 35°C, another condition that favors pore formation, also did not produce a NMDG-permeable channel (26). Finally, because long-duration ATP application in cells expressing P2X4 or P2X7 has been shown to elicit pore formation, we continuously (>5 min) perfused the cells with the ATP (1 mM) or BzATP (100 µM) low divalent bathing solution. We were unable to elicit a current that was carried by NMDG. As a control, BzATP-activated a NMDG-permeable current in a P2X7-expressing cell line under these conditions (data not shown). To further characterize the P2X7 response, we examined the effects of the two antagonists suramin and reactive blue. High concentrations of suramin (300 µM) produced a reversible 50% block of the current at the most negative potentials, whereas 100 µM had only a small (7 and 10%, n = 2) effect (Fig. 6C). Reactive blue (60 µM) completely blocked the BzATP-activated response (n = 2; Fig. 6D).

View larger version (24K): [in this window] [in a new window]   Fig. 6.   2',3'-O-(4-benzoylbenzoyl)ATP (BzATP)-activated whole cell currents. A: BzATP (50 µM) activated a current in single cells that decayed very slowly toward a sustained level in the continued presence of the agonist. While the membrane potential was held at 60 mV, N-methyl-D-guanine ion (NMDG+) was substituted for Na+ in the bath. The BzATP-activated inward current was eliminated, and the reversal potential was shifted toward more negative values, as shown in B, where the current at each of the voltages indicated by the letters in A is plotted. At the gaps between 60 and +60 mV, the potential was at 0 mV. C: current response to a ramp of voltage from 120 to +60mV in the control bath solution, in the presence of BzATP (50 µM), and when suramin (300 µM) was added to the BzATP perfusate. D: current elicited in a cell clamped at 60 mV in response to 20 µM Bz-ATP was completely blocked by 60 µM RB.

View larger version (21K): [in this window] [in a new window]   Fig. 7.   BzATP-activated channels in outside-out patches. A: in an outside-out patch, BzATP (100 µM) elicited multiple-channel activity, as shown at membrane potentials of 80 and 100 mV. B: example of the BzATP-activated current from another patch with at least two channels is shown at a series of holding potentials. C: linear fit to the current-voltage relationship for the cell in B gave a conductance of 9.8 pS. Pipette contained CsMeSO4. The bath contained the normal physiological solution.

YO-PRO uptake. We examined the cells for permeability to YO-PRO in the presence of BzATP, ATP, or MeSATP. A typical YO-PRO uptake experiment is shown in Fig. 8. A monolayer of BAECs was incubated in the extracellular solution containing 3 µM YO-PRO. After 5 min of incubation, the solution was changed to one containing 3 µM YO-PRO plus 100 µM BzATP and maintained for 30 min. YO-PRO did not appear as intranuclear fluorescence, although there was a weak increase in cytosolic fluorescence after the 30-min time period (Fig. 8B). This may suggest the presence of a few large channels passing the YO-PRO that binds to cytoplasmic nucleic acids. As a control, permeablization with Triton X-100 produced strong nuclear labeling. Similar experiments (where n  3 for each condition) were performed in normal and low divalents, at both room temperature (19-22°C) and at 37°C, but YO-PRO labeling of the nuclei was not observed.

View larger version (109K): [in this window] [in a new window]   Fig. 8.   Lack of YO-PRO uptake in the presence of BzATP. A monolayer of BAECs was incubated with standard solution in presence of YO-PRO for 5 min (top left). BzATP (200 µM) was added to the bath in presence of YO-PRO and incubated for 30 min at 37°C in low divalent solution (top middle). The small amount of fluorescence in the cytoplasm (likely bound to RNA) indicates a very low penetration of YO-PRO. This result suggests that the size of the channels induced by BzATP is not large enough to allow entrance of the YO-PRO. Top right: as a control, at the end of the experiment, 10 µl Triton X-100 (TX-100) were added to the bath to permit the entrance of YO-PRO, as confirmed by the fluorescent nuclei. Bottom left, middle, and right: corresponding differential interference contrast (DIC) images of the cell in top left, middle, and right, respectively.

Immunocytochemical localization provides further support for functional presence of P2X4 and P2X7 receptors. We next investigated the distribution of P2X4 and P2X7 immunoreactivity at the cellular level with two commercially available antibodies (Fig. 9). Staining with anti-P2X4 showed a localization of this subunit in perinuclear regions as well as diffuse distribution in the cytoplasm up to the cell borders. Anti-P2X7 staining also localized to the perinuclear area, the cytoplasm and, in addition, was also very clearly delineated at the cell borders. A Z section confocal line scan shows P2X4 and P2X7 bordering the cell surface in what appear to be patches of immunoreactive material.

View larger version (175K): [in this window] [in a new window]   Fig. 9.   Distribution of P2X4 and P2X7 receptors in BAECs. Monolayers of BAECs were incubated with anti-P2X4 or anti-P2x7 for 1 h. The distribution of the receptor was visualized using a secondary antibody conjugated with FITC or RITC. A: P2X4 was localized throughout the cytoplasm up to the cell borders and more strongly in perinuclear areas. B: DIC image of the same field as in A. C: confocal line scan from top to bottom through cells in a monolayer with anti-P2X4 antibody labeling. D: staining with anti-P2X7 antibody shows strong labeling near the cell borders as well as in the cytoplasm and in the perinuclear area. Consistent with a previous report (11), P2X7 nuclear staining is present. E: DIC image of cells shown in D. F: confocal line scan through a monolayer with anti-P2X7 antibody labeling.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This is the first electrophysiological evaluation of P2X receptors in endothelial cells. One aspect of the present study supports a recent report that demonstrated that the P2X4 message is abundant in the endothelium and that the P2X4 subunit contributes to ATP-activated calcium influx, as reported using indo 1 (41). These authors also report the presence of mRNA for P2X7, although the protein was not functionally characterized. In that study, human aortic endothelial cells produced only a very weak signal from P2X5, whereas ours was stronger in BAECs, although weaker than the signals for P2X4 and P2X7. The complement of P2X receptor/channels in endothelial cells is likely to depend on the anatomic location and species because P2X2 receptors have been localized to endothelial cells of several major arteries (13) and to cerebral vessels (21) and P2X3 has been observed in endothelial cells in the thymus gland (11). Glass and Burnstock (10) also localized P2X3 as well as P2X4 and P2X7 to endothelial cells in the rat thyroid gland.

Because at least five different P2 receptors (P2X4, P2X5, P2X7, P2Y1, and P2Y2) have been reported in the BAECs, all of which are able to increase calcium influx, it is logical to ask whether they play distinct roles in regulating intracellular calcium and cellular function. This is important because calcium has been implicated not only in modulating secretion but also in regulating gene expression. The interesting problem is whether distinct receptors are tied to specific responses through their surface localization. Calcium influx through a particular receptor might activate calcium-sensitive pathways that are colocalized at those sites. This could lead to secretion of a specific substance or expression of particular genes. Knowledge of receptor distribution, kinetics, and pharmacology of specific receptors may contribute to unraveling the contributions of specific subtypes.

P2X4 and P2X5 channels. The characteristics of the MeSATP-activated channels in BAECs are consistent with those of the expressed P2X4 receptor but not those of the P2X5 receptor. The response was transient in the continued presence of the agonist, was not blocked by the antagonist suramin, and was an inwardly rectifying current (2, 8, 32, 38). The slow desensitization of P2X4 has been reported by several groups. Garcia-Guzman et al. (8) applied agonist for 5 s and showed that the current decayed to <20% of peak in that time. Collo et al. (4) projected a tau decay (_decay) = 17 s in response to 5-s application of agonist. Soto et al. (32) showed ~40% decay for MeSATP in 2 s. North and Barnard (24) gave a value of >2 s, and Koshimizu et al. (18) also showed currents expressed in GT1 cells for P2X4 and P2X7 that are quite similar to those of the present study. On the other hand, P2X5 expressed alone has been shown to be blocked by suramin and shows little desensitization (9). In an extensive study designed to examine coassembly of the P2X family members, Torres et al. (34) used coprecipitation studies to demonstrate that P2X4 and P2X5, in heterologous expression in HEK cells, can form heteromultimeric complexes, as can P2X4 and P2X6 but, interestingly, not the other members of the P2X family. In the same study, P2X7 did not coassemble with any other members of the family. On the basis of the electrophysiology and pharmacology in the present study, we hypothesize that P2X4 subunits are expressed at the surface membrane in BAECs as P2X4 homomeric channels or, alternatively, exist as heteromultimers, P2X4/P2X5, while adopting the properties of P2X4. The latter hypothesis could be tested through coexpression studies. Yamamoto et al. (42) introduced an antisense oligonucleotide for P2X4 into endothelial cells and blocked a component of the calcium influx in response to ATP to demonstrate that the P2X4 subunit participates in activating the calcium influx. This did not rule out the possibility of heteromultimer formation with P2X5, whose message they also observed, although in much smaller quantities than P2X4. They also related shear stress-activated calcium influx with the P2X4 receptor (41).

P2X channels differ from the P2Y receptor-activated SOCs. The P2Y receptors in endothelial cells (P2Y1, P2Y2, and P2Y11) are G protein linked and activate the phosphoinositol system, leading to store-operated calcium influx after Ins(1,4,5)P₃-induced release of calcium from intracellular stores and, in the case of P2Y11, are also linked to the cAMP/protein kinase A pathway. The SOC differs from that observed in the present studies. It exhibits anomalous mole fraction behavior and a reversal potential that reflects the high calcium permeability. Its single channel conductance is ~2 pS in 2 mM extracellular calcium (35). Furthermore, the P2Y receptors are reported to be blocked by suramin, in contrast to the MeSATP-activated channel in the present studies.

P2X7 channels. The P2X7 receptor is, uniquely among the P2X4, P2X5, and P2X7 receptors, insensitive to low concentrations of ATP (100 µM) but activated by low concentrations of BzATP (10 µM). The response of the endothelial cells to BzATP (10-100 µM) is clearly distinct from that of MeSATP or ATP (10-100 µM). The BzATP-activated inward current response shows only a slow decay over many minutes in the continued presence of agonist. The single channel amplitude in the present study was 9 pS, consistent with measurements of 10 pS in B-lymphocytes, where P2X7 is a prevalent channel (22). The BzATP-activated response is 50% blocked by 300 µM suramin, which did not block the MeSATP response, consistent with the reports of the P2X7 receptor. Finally, the BzATP-activated current was completely blocked by 60 µM reactive blue, whereas the MeSATP response was not blocked. These data support the contention that the BzATP-activated response is via the P2X7 receptor.