Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Extracellular nucleotides, such as diphosphates and triphosphates of adenosine and uridine, regulate diverse physiological functions through the activation of membrane receptors that include both ion channel (P2X) and G protein-coupled (P2Y) receptor subtypes. A number of P2 receptor subtypes have been identified through molecular cloning studies: seven subtypes of P2X receptors (P2X₁-P2X₇) and five distinct subtypes of P2Y receptors (P2Y₁, P2Y₂, P2Y₄, P2Y₆, and P2Y₁₁; Harden et al., 1998; Kunapuli and Daniel, 1998; Ralevic and Burnstock, 1998; Turner et al., 1998). However, these structurally distinct P2 receptor subtypes have not been clearly assigned to physiological functions, in part because of the paucity of subtype-selective antagonists.

One compound that has been proposed as an antagonist to help distinguish among P2 receptor subtypes is pyridoxal-phosphate-6-azophenyl-2',4'-disulfonate (PPADS). Lambrecht et al. (1992) were the first to report that PPADS blocked the ATP-induced contractions of rabbit vas deferens and proposed it as a novel and selective P2 receptor antagonist. This compound has subsequently been reported to antagonize P2 receptors of various cell types. PPADS appears to be relatively selective for P2X receptors. However, in endothelial cells that lack P2X receptors, PPADS can also antagonize P2Y receptors (Windscheif et al., 1994; Ralevic and Burnstock, 1996; Ralevic et al., 1997). Although PPADS interacts with all the known P2X receptor subtypes (Ralevic and Burnstock, 1998), its effect on P2Y receptor subtypes show selectivity, such that PPADS blocks P2Y₁- but not P2Y₂-mediated accumulation of inositol phosphates and contractile responses of isolated tissues (Ralevic and Burnstock, 1996; Hansmann et al., 1997; Westfall et al., 1997; McLaren et al., 1998).

Agonist selectivity has also been used to characterize P2Y receptor subtypes. For instance, 2-methylthioadenosine triphosphate (2MeSATP) and UTP are believed to stimulate P2Y₁ and P2Y₂ receptors, respectively (Harden et al., 1998). Previous studies from our laboratory have identified both P2Y₁ and P2Y₂ receptors in the Madin-Darby canine kidney (MDCK-D₁) cells and demonstrated a functional difference, such that the cyclooxygenase inhibitor indomethacin blocks the cAMP accumulation stimulated by UTP, whereas the same response produced by 2MeSATP is insensitive to indomethacin (Firestein et al., 1996; Insel et al., 1996; Post et al., 1998). However, the characterization of receptors on the grounds of agonist-evoked responses may be misleading. For instance, 2MeSATP and UTP have been shown to also activate P2Y₁₁ and P2Y₄ receptors, respectively (Communi et al., 1997; Harden et al., 1998). Because the expression of P2Y₁, P2Y₂, and P2Y₁₁ in MDCK-D1 cells has been demonstrated by reverse transcription-polymerase chain reaction analysis (Post et al., 1998), the response to 2MeSATP may not be exclusive for P2Y₁ receptors. The present study was initiated to use PPADS, the putative P2Y₁-selective antagonist, to define the function of P2Y₁ receptors in MDCK-D1 cells. We found unexpectedly, that at least in these cells, this compound primarily acts at a site distal to P2Y receptors.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. Dulbecco's modified Eagle's medium (DMEM) was obtained from Life Technologies BLR (Grand Island, NY). PPADS and 2MeSATP were purchased from Research Biochemicals Inc. (Natick, MA) Anti-cAMP antibody was obtained from Calbiochem (San Diego, CA). ¹²⁵I-cAMP was obtained from NEN Life Science Products (Boston, MA). Melittin was purchased from Peninsula Laboratories (Belmont, CA). All other reagents were purchased from Sigma Chemical Co.

Cell Culture. The MDCK-D₁ cells were grown and maintained in DMEM supplemented with 10% heat-inactivated serum (7.5% horse serum and 2.5% fetal calf serum).

Measurement of cAMP Accumulation. The cells, grown to 60 to 70% confluency, were equilibrated with freshly prepared HEPES-buffered DMEM for 30 min at 37°C. Subsequent to a 10-min incubation with PPADS, the stimulation of cAMP was initiated by the addition of 2MeSATP and 200 µM isobutylmethyl xanthine (a cyclic nucleotide phosphodiesterase inhibitor), and the incubation at 37°C was allowed to continue for 5 min. The reaction was terminated by the addition of 7.5% trichloroacetic acid. The cellular cAMP was measured in trichloroacetic acid-soluble fraction by radioimmunoassay after acetylation according to the manufacturer's instructions (Calbiochem). The generation of cAMP was normalized to the protein determined according to the Bio-Rad (Hercules, CA) protein assay.

Measurement of [³H]Arachidonic Acid and Metabolite (AA) Release. As described earlier (Xing et al., 1997), the cells were trypsinized, plated onto 24-well plates and allowed to grow to about 60% confluency (2 days), followed by an overnight incubation with 0.5 µCi/ml [³H]AA (from NEN; specific activity, 3396 GBq/mmol). The cells were washed twice with bicarbonate- and serum-free DMEM, pH 7.4, and then equilibrated with the same medium for 45 min at 37°C. After a 10-min incubation with PPADS, the reaction was initiated by the addition of the stimulating agents, and the incubation was continued for another 20 min. The reaction was stopped by the addition of ice-cold EDTA and EGTA to a final concentration of 5 mM each. The medium was aspirated into the scintillation vials and mixed with the scintillant, and [³H] was determined. The cells were treated with 0.2% of Triton X-100, scraped from the plates, mixed with the scintillant, and quantified through determination of unreleased ³H. The AA release was normalized based on the amount of radioactivity incorporated into the cells. Previous studies from our laboratory document that [³H]AA release includes both free arachidonic acid and arachidonic acid metabolites, in particular, prostaglandin E₂ (Slivka and Insel, 1988; Post et al., 1998).

Measurement of Phospholipase A₂ (PLA₂) Activity. The MDCK-D₁ cells grown to about 70% confluency were harvested by centrifugation; resuspended in 100 mM Tris · HCl buffer, pH 8, containing 1 mM CaCl₂; and homogenized by sonication. The activity was measured in the homogenate according to the method described by Conde-Frieboes et al. (1996). Briefly, reactions were carried out in a buffer containing 80 mM HEPES (pH 7.4), 150 mM NaCl, 2 mM CaCl₂, 1 mM 1-palmitoyl-2-arachidonyl-sn-glcero-3-phosphocholine (with 100,000 cpm [¹⁴C]1-palmitoyl-2-arachidonyl-sn-glcero-3-phosphocholine), 2 mM Triton X-100, and 30 µg of MDCK cell homogenate. After 30 min, reactions were quenched by the addition of 2.5 ml of Dole reagent [2-propanol/heptane/0.5 M H₂SO₄ (400:100:20 v/v/v)] and vortexed. The amount of hydrolysis was determined using a procedure described previously (Ulevitch et al., 1988).

Statistical Analysis. Data represent mean ± S.E. values from three to six independent experiments. The significance of the inhibitory effect of PPADS was confirmed with Student's t test by unpaired and one-tailed comparisons with values of P < .05 considered statistically significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of PPADS on 2MeSATP-, UTP-, ADP-, and Adenosine-5'-O-(2-thio)diphosphate (ADPS)-Mediated cAMP Accumulation. PPADS, at a concentration of 100 µM, did not substantially affect the concentration-dependent increase in cAMP accumulation caused by 2MeSATP in MDCK-D₁ cells but blocked the accumulation of cAMP evoked by UTP (Fig. 1). As described earlier, these observations are opposite to what would be expected for P2Y₁ receptor effects because 2MeSATP is presumed to activate P2Y₁ receptors, and UTP would selectively activate P2Y₂ receptors. Moreover, PPADS failed to block cAMP accumulation in response to the P2Y₁-selective agonists ADP and ADPS (Harden et al., 1998; Fig. 1b). Thus, based on effects on agonist-promoted cAMP accumulation, PPADS appears not to selectively block P2Y₁ receptors in MDCK-D₁ cells.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Effect of PPADS on 2MeSATP- and UTP-induced (panel a) and ADP- and ADPS-induced (panel b) accumulation of cAMP. Before the addition of 2MeSATP, the cells were incubated with PPADS (100 µM) for 10 min. The reaction was started by the addition of the nucleotides at varying concentrations (a) and at 100 µM (b) and continued at 37°C for 5 min. The samples were then assayed for cAMP. Each value is mean ± S.E. of three to six independent experiments. **P < .01, ***P < .001, inhibition compared with the respective basal value was significant as determined by Student's t test.

PPADS Inhibited AA Release Mediated by 2MeSATP, UTP, and Other Nucleotides. Because the generation of cAMP in response to UTP, but not 2MeSATP, is sensitive to inhibition by indomethacin (Post et al., 1996, 1998), we tested whether the release of AA stimulated by these nucleotides is also differentially blocked by PPADS. We found that the increase in AA release promoted by 2MeSATP was blocked by PPADS and that the effect was concentration dependent (Fig. 2, left). In addition, PPADS blocked the release of AA in response to other nucleotides: release promoted by ADP and ADPS, as well as that by UTP, was blocked by PPADS (Fig. 2).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Effect of PPADS on AA release mediated by 2MeSATP, ATP, and UTP (left) and ADP and ADPS (right). Cells were incubated with PPADS for 10 min and then with nucleotides for 20 min at 37°C. AA release was measured as described in Materials and Methods. Both 2MeSATP and UTP caused 4- to 5-fold increases over basal levels, and ADP and ADPS caused 2-fold increases over basal levels. The values for unstimulated (basal) responses have been subtracted from the stimulated responses, which were then calculated as the percentage of control (in the absence of PPADS), which has been considered 100%. Each value represents the average ± S.E. of at least six independent experiments. *P < .05, **P < .01, ***P < .001, inhibition compared with the respective control value was significant as determined by Student's t test.

Uncompetitive Inhibition of 2MeSATP-Induced AA Release by PPADS. To help define the mode of antagonism of P2Y receptors by PPADS, we measured AA release in response to various concentrations of 2MeSATP. Because PPADS did not shift the agonist-response curves rightward in a parallel fashion without affecting the maximal response (Fig. 3), the mode of inhibition appears to be uncompetitive, thus suggesting a nonreceptor interaction of PPADS. Lower concentrations (10 and 30 µM) of PPADS were ineffective, whereas larger concentration (100 and 300 µM)-mediated blockade were not reversed by increasing the concentrations of the agonist. In addition, PPADS blocked the basal release of AA in a concentration-dependent manner (Fig. 3, inset).

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Concentration-response curves of 2MeSATP-promoted AA release in the absence and presence of different concentrations of PPADS. Inset, effect of PPADS on basal release of AA. Cells were incubated with 2MeSATP and PPADS as indicated in the legend to Fig. 2. Each point represents mean ± S.E. of three to six independent experiments.

Inhibition of Non-nucleotide-Mediated AA Release by PPADS. In MDCK-D₁ cells, bradykinin, acting via B₂ receptors, and phenylephrine, acting via _1b-adrenergic receptors, also generate AA, albeit through different signaling pathways (Slivka and Insel, 1987, 1988; Xing and Insel, 1996; Xing et al., 1997). We found that PPADS inhibited the release of AA mediated by both of these agonists (Fig. 4, left).

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Effect of PPADS on AA release mediated by bradykinin and phenylephrine (left); PMA, ionomycin, and thapsigargin (middle); and melittin (right). Cells were incubated with varying concentrations of PPADS for 10 min and then with agents that stimulate AA release for 20 min. The data have been calculated as the percentage of control response (absence of PPADS) and corrected for response in the absence of stimulants. Control response for the stimulants was approximately 2-, 2-, 5-, 6-, 3-, and 12-fold over basal levels by bradykinin, phenylephrine, PMA, ionomycin, thapsigargin, and melittin, respectively. Each value represents mean ± S.E. of at least six independent experiments. *P < .05, **P < .01, ***P < .001, inhibition compared with the respective control value was significant as determined by Student's t test.

Effect of PPADS on PLA₂ Activity in MDCK-D1 Cell Homogenate and Recombinant Enzyme. To assess directly whether PPADS could block PLA₂ activity in MDCK-D₁ cells, we assayed PLA₂ activity in MDCK-D₁ cell homogenate. We found that this activity was substantially dependent on calcium (as shown by the sensitivity to EDTA) but was not inhibited by 100 µM PPADS (Fig. 5). Sensitivity to EDTA may represent blockade of calcium-dependent activation of the 85-kDa cytosolic PLA₂ (cPLA₂) in these cells (Xing and Insel, 1996). This implies that cPLA₂ is not sensitive to inhibition by PPADS. Moreover, PPADS does not inhibit recombinant human cPLA₂ activity up to a concentration of 1 mM (M.A.B. and E. A. Dennis, data not shown). Thus, PPADS is likely to block AA release via a mechanism other than inhibition of cPLA₂.

View larger version (50K): [in this window] [in a new window]   Fig. 5.   Effect of PPADS on PLA2 activity in the MDCK-D1 cell homogenate. PLA2 activity was measured as described in Materials and Methods with 30 µg of the cell homogenate protein per assay. Each point represents mean ± S.E. of at least three independent experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The current study was initially undertaken with the expectation that PPADS might be useful to distinguish the role of P2Y receptor subtypes in MDCK-D₁ cells. In contradiction to several published reports (see introduction), in which PPADS preferentially blocked P2Y₁ receptors, our data are more consistent with an action of PPADS at a postreceptor site. Evidence in support of this conclusion is that PPADS blocked AA release in response to several types of agonists but appeared not to block 2MeSATP-stimulated cAMP generation or to show a pattern of competitive inhibition in antagonism of 2MeSATP-promoted AA release. The inability of PPADS to block cAMP response to 2MeSATP may relate to an activation of P2Y₁₁, rather than P2Y₁, receptors by 2 MeSATP (Communi et al., 1997; Torres et al., 1999). We are aware of no available data regarding blockade of P2Y₁₁ receptors by PPADS, but our results suggest that PPADS may not block the P2Y₁₁ subtype.

In contrast to the inability of PPADS to inhibit cAMP formation by 2MeSATP, we found that cAMP generated in response to UTP was suppressed by PPADS. Moreover, PPADS blocked AA release stimulated by both 2MeSATP and UTP and by a variety of agents: nucleotides, phenylephrine, bradykinin, PMA, ionomycin, thapsigargin, and melittin. Taken together, these findings indicate that PPADS can inhibit the release of AA and its metabolites via an action distal to P2Y receptors. A postreceptor action of PPADS was also suggested by Vigne et al. (1996, 1998), who demonstrated that PPADS could inhibit endothelin-stimulated Ca²⁺ mobilization by blocking inositol trisphosphate-stimulated Ca²⁺ channels in endothelial cells. Other recent data question whether PPADS is a P2Y₁ antagonist in vascular tissue (Guibert et al., 1998).

PPADS has also been shown to inhibit ectonucleotidase activity in primary cultures of guinea pig vas deferens smooth muscle cells, endothelial cells, C6 glioma cells, and RAW 264.7 macrophages (Ziganshin et al., 1995; Chen et al., 1996; Lambrecht, 1996). However, inhibition of the ectonucleotidase would not account for the blockade of nucleotide-mediated responses, and thus, the two effects appear to be independent of each another.

The inhibition of AA release mediated by melittin, a direct activator of PLA₂, suggested that PLA₂ might be the site of PPADS action, but the lack of inhibition of Ca²⁺-dependent PLA₂ activity appears to rule out cPLA₂ as the target of PPADS action. However, it is possible that the PLA₂ activity assayed in the MDCK-D₁ cell homogenate fails to measure other AA-generating phospholipases (Balsinde et al., 1999) and that PPADS blocks the activity of one or more of these enzymes. It is also possible that PPADS inhibits AA-metabolizing enzymes or transport of AA across the plasma membrane. Further studies will be needed to assess each of these possible sites of action; however, the current data clearly show that the use of PPADS to discriminate the functions of various P2 receptor subtypes can be misleading, especially in settings in which AA contributes to cellular responses.