The interaction of diadenosine polyphosphates with P2X-receptors in the guinea-pig isolated vas deferens

Authors


Department of Physiology and Pharmacology, University of Strathclyde, Royal College, 204, George Street, Glasgow G1 1XW

Abstract

  • The site(s) at which diadenosine 5′,5′′′-P1, P4-tetraphosphate (AP4A) and diadenosine 5′, 5′′′-P1, P5-pentaphosphate (AP5A) act to evoke contraction of the guinea-pig isolated vas deferens was studied by use of a series of P2-receptor antagonists and the ecto-ATPase inhibitor 6-N, N-diethyl-D-β,γ-dibromomethyleneATP (ARL 67156).

  • Pyridoxalphosphate-6-azophenyl-2′,4′-disulphonic acid (PPADS) (300 nM–30 μM), suramin (3–100 μM) and pyridoxal-5′-phosphate (P-5-P) (3–1000 μM) inhibited contractions evoked by equi-effective concentrations of AP5A (3 μM), AP4A (30 μM) and α,β-methyleneATP (α,β-meATP) (1 μM), in a concentration-dependent manner and abolished them at the highest concentrations used.

  • PPADS was more potent than suramin, which in turn was more potent than P-5-P. PPADS inhibited AP5A, AP4A and α,β-meATP with similar IC50 values. No significant difference was found between IC50 values for suramin against α,β-meATP and AP5A or α,β-meATP and AP4A, but suramin was more than 2.5 times more potent against AP4A than AP5A. P-5-P showed the same pattern of antagonism.

  • Desensitization of the P2X1-receptor by α,β-meATP abolished contractions evoked by AP5A (3 μM) and AP4A (30 μM), but had no effect on those elicited by noradrenaline (100 μM).

  • ARL 67156 (100 μM) reversibly potentiated contractions evoked by AP4A (30 μM) by 61%, but caused a small, significant decrease in the mean response to AP5A (3 μM).

  • It is concluded that AP4A and AP5A act at the P2X1-receptor, or a site similar to the P2X1-receptor, to evoke contraction of the guinea-pig isolated vas deferens. Furthermore, the potency of AP4A, but not AP5A, appears to be inhibited by an ecto-enzyme which is sensitive to ARL 67156.

British Journal of Pharmacology (1997) 121, 57–62; doi:10.1038/sj.bjp.0701099

Introduction

Diadenosine polyphosphates are naturally occurring molecules which are involved in numerous intracellular biochemical pathways. However, they may also be important extracellular signalling agents as the release of micromolar concentrations of compounds such as diadenosine 5′,5″-P1, P4-tetraphosphate (AP4A) and diadenosine 5′,5″-P1, P5-pentaphosphate (AP5A) into the extracellular space can be measured from platelets (Flodgaard & Klenow, 1982; Schlüter et al., 1994), chromaffin cells (Pintor et al., 1991; Castillo et al., 1992) and neurones (Pintor et al., 1992). Furthermore, at these concentrations AP4A and AP5A have widespread extracellular actions, causing contraction of visceral and vascular smooth muscle, excitation of neurones, release of catecholamines from chromaffin cells and inhibition of platelet aggregation (see Ogilvie, 1992 for review).

AP4A and AP5A appear to produce their effects through a number of different receptor types and may act as agonists at P2X1- (Evans et al., 1995; Ralevic et al., 1995), P2Y- (Castro et al., 1992) and P2U-receptors (Lazarowski et al., 1995) and antagonists at P2T-receptors (see Cusack, 1993). The existence of a separate site which recognises di-, but not mono-adenosine nucleotides such as adenosine 5′-triphosphate (ATP), has also been proposed (Hilderman et al., 1991; Walker et al., 1993; Pintor et al., 1993; Pintor & Miras-Portugal, 1995) and the name P2D-purinoceptor suggested (Pintor et al., 1993), (see Burnstock & King, 1996 for discussion of nomenclature). A further complicating factor is that some actions of AP4A and AP5A are inhibited by selective P1-receptor antagonists (Klishin et al., 1994; Ziganshin et al., 1995; Hoyle et al., 1996; Rubino & Burnstock, 1996). It is not clear if this is due to a direct action of AP4A and AP5A on P1-receptors or if it depends upon their breakdown to adenosine. Thus, great care must be taken when trying to classify the receptors at which AP4A and AP5A act.

AP4A and AP5A cause contraction of the guinea-pig vas deferens (Stone, 1981; MacKenzie et al., 1988; Bailey & Hourani, 1995; Hoyle et al., 1995b). It has been assumed that they act through the P2X1-receptor as the contractions were similar to those elicited by α,β-methyleneATP (α,β-meATP), which is thought to act in this tissue via the recently cloned P2X1-receptor (Evans et al., 1995). Also, the P2-receptor antagonists suramin (Bailey & Hourani, 1995) and pyridoxalphosphate-6-azophenyl-2′,4′-disulphonic acid (PPADS) (Hoyle et al., 1995b) inhibited AP4A and AP5A, but only single or high antagonist concentrations were used. Thus, there is little quantitative data with which to characterize their site of action.

Another approach often used is agonist cross-desensitization. In mechanical studies AP5A, but not AP4A, mimicked the ability of α,β-meATP to desensitize the P2X1-receptor and so inhibit the purinergic component of neurogenic contractions (MacKenzie et al., 1988). However, desensitization produced by AP5A and α,β-meATP had different time-courses. Thus, it is not clear if AP4A, AP5A and α,β-meATP act at the same or separate receptors to elicit contraction in this tissue. In contrast, in an electrophysiological study, desensitization to AP4A or AP5A attenuated depolarization evoked by ATP (Hoyle et al., 1995a). The reason for the discrepancy between the two types of study is not known.

It is clear from the above that previous studies on the site(s) through which AP4A and AP5A act in the guinea-pig isolated vas deferens have been largely qualitative and have not reached the same conclusion. Therefore, the aim of this study was to characterize these site(s) quantitatively, by comparing the ability of a full range of concentrations of the P2-receptor antagonists suramin, PPADS and pyridoxal-5′-phosphate (P-5-P) to inhibit their contractions and those to α,β-meATP. The influence of breakdown of AP4A and AP5A on their potency was also investigated with the ecto-ATPase inhibitor 6-N, N-diethyl-d-β,γ-dibromomethyleneATP (ARL 67156) (Crack et al., 1995; Westfall et al., 1996a). A preliminary account of these results has been published (Westfall et al., 1996b).

Methods

Albino male guinea-pigs (250–400 g) were killed by asphyxiation with CO2 and subsequent cervical dislocation. The vasa deferentia were removed, cleaned of connective tissue and mounted in 2 ml horizontal baths. The tissues were allowed to equilibrate under a resting tension of 1 g at 35°C for 1 h in a physiological salt solution of the following composition (mm): NaCl 118.4, NaHCO3 25, KCl 4.7, MgSO4 1.2, KH2PO4 1.2, CaCl2 2.5 and glucose 11; bubbled with 95% O2, 5% CO2. Tension was recorded by Scaime transducers connected via Universal amplifiers (Gould) to a pen recorder (Gould 2200S).

Concentration-response curves to AP4A, AP5A and ATP were obtained in individual tissues by addition of the drugs to the bath in a non-cumulative manner. Shortly after the contraction had reached a peak the agonist was washed out by several changes of the bathing solution. Care was taken to avoid desensitization by leaving 30 min before the next concentration of agonist was added.

The effects of suramin, PPADS and P-5-P on responses to exogenous agonists were examined in individual tissues. Three reproducible control responses to equi-effective concentrations of AP5A (3 μm), AP4A (30 μm) or α,β-meATP (1 μm) were obtained at 30 min intervals. The lowest concentration of suramin, PPADS or P-5-P used was then applied to the tissue for 30 min and the agonist readded. This procedure was repeated until steady state inhibition was reached. Thereafter, progressively higher concentrations of antagonist were administered in the same manner.

The effect of desensitization of the P2X-receptor on contractions to AP4A or AP5A was determined as follows. Control responses to AP5A (3 μm), AP4A (30 μm) or noradrenaline (100 μm) were first obtained, then α,β-meATP (50 μm) was applied, evoking a large transient contraction. Once tension had returned to the baseline level a further 50 μm α,β-meATP was added. This procedure was repeated until the tissue had been exposed to a cumulative concentration of 200 μm α,β-meATP. The drug was washed out and when α,β-meATP (50 μm) was added two min later no contraction was seen, confirming that the P2X-receptors had been desensitized. α,β-meATP was again washed out and two min later the test agonist administered.

Similarly, when the effects of ARL 67156 were studied, three reproducible control responses to AP5A (3 μm) or AP4A (30 μm) were obtained at 30 min intervals and ARL 67156 (100 μm) was then added 10 min before the fourth application of agonist. We have previously shown ARL 67156 to equilibrate within 10 min (Westfall et al., 1996a).

Statistics

Values in the text refer to mean ± s.e.mean or geometric mean with 95% confidence limits for IC50 values (after Fleming et al., 1972). IC50 values were calculated for each individual concentration-inhibition curve. Data were compared by Student's paired t test or one way analysis of variance and Tukey's comparison as appropriate. Differences were considered significant when P<0.05. Concentration-inhibitory response curves for the antagonists were fitted to the data by logistic (Hill equation), non-linear regression analysis (FigP, Biosoft, Cambridge, U.K.).

Drugs

α,β-meATP (lithium salt), AP4A (ammonium salt), AP5A (sodium salt), ATP (disodium salt), P-5-P (all Sigma), ARL 67156 (provided by Astra Charnwood) and suramin (Bayer) were dissolved in distilled water and stored as 100 mm stock solutions. PPADS, a gift from Dr G. Lambrecht, University of Frankfurt, was dissolved in distilled water as a 10 mm stock solution and stored frozen in darkness. (–)-Noradrenaline bitartrate (Sigma) was dissolved in acid saline and frozen as a 100 mm stock solution. Potassium chloride (Sigma) was kept as a 2 m stock. The firefly luciferin-luciferase assay (Sigma) was used to analyse the ATP content of the AP4A and AP5A solutions.

Results

Concentration-response relationships of agonists and antagonists

AP5A (30 nm–100 μm) (n=7), AP4A (300 nm–100 μm) (n=7) and ATP (1 μm–1 mm) (n=6–10) evoked concentration-dependent, rapid, transient contractions, which reached a peak in about 5 s and then subsided rapidly, even in the continued presence of the agonist (see Figure 3). We have previously shown that α,β-meATP evokes similar contractions (McLaren et al., 1994) and comparing the data showed that the rank order of agonist potency was α,β-meATp>AP5A>AP4A>ATP. However, none of the curves had a clear maximum and so it was not possible to calculate EC50 values.

Figure 3.

The effect of ARL 67156 on contractions evoked by AP5A and AP4A. (a) Typical responses to AP4A (30 μm) and AP5A (3 μm) before and 10 min after addition of ARL 67156 (100 μm). (b) The mean responses to AP4A (30 μm, n = 8) and AP5A (3 μm, n = 6) after addition of ARL 67156 (100 μm, solid columns) expressed as a percentage of control responses (open columns). Statistical significance was determined by analysis of the raw data in g (*P<0.05, ***P<0.001).

Suramin, PPADS and P-5-P (all 100 μm) shifted the concentration-response curves to AP4A and AP5A to the right such that the response to most concentrations of agonists was abolished (not shown). This flattening of the curves limited the range of antagonist concentrations which could be meaningfully studied. Therefore, in order to characterize the actions of the antagonists over as wide a concentration range as possible, we studied their effects against equi-effective concentrations of α,β-meATP (1 μm), AP5A (3 μm) and AP4A (30 μm). These each evoked contractions of approximately 50% amplitude of that obtained to 1 mm ATP. α,β-meATP was used rather than ATP as ATP may act at more than one site to evoke contraction in this tissue (Bailey & Hourani, 1995) and the potency of ATP, but not α,β-meATP, is decreased by breakdown by ecto-ATPase (Westfall et al., 1996a).

PPADS (300 nm–30 μm), suramin (3–100 μm) and P-5-P (3–1000 μm) inhibited the contractions evoked by α,β-meATP (1 μm), AP5A (3 μm) and AP4A (300 μm) in a concentration-dependent manner and abolished them at the highest concentrations used (Figure 1). In each case, PPADS was significantly more potent than suramin, which in turn was significantly more potent than P-5-P (P<0.05, Table 1). PPADS had a similar antagonist potency against α,β-meATP, AP5A and AP4A. Likewise, no significant difference was found between the IC50 values for suramin against α,β-meATP and AP5A or α,β-meATP and AP4A. However, suramin was significantly more potent against AP4A than AP5A (P<0.05). The same pattern of relative potency was also seen with P-5-P.

Figure 1.

Inhibition of contractions evoked by (a) α,β-meATP (1 μm), (b) AP5A (3 μm) and (c) AP4A (30 μm) by PPADS (▴), suramin (•) and P-5-P (▪). The graphs show mean data (n = 6);

Table 1. Potency of P2-receptor antagonists in the guinea-pig isolated vas deferens
 α,β-meATP (1 μm)AP5A (3 μm)AP4A (30 μm)
  1. Values shown are IC50 and 95% confidence limits (μm) for the antagonists; n = 6 for each agonist.

PPADS3.6 (1.6–8.1)3.4 (1.8–6.4)3.1 (2.6–3.6)
Suramin10.6 (6.7–16.8)14.2 (8.2–24.6)6.0 (4.7–7.5)
P-5-P95.9 (76.3–120.6)166.0 (115.3–238.9)63.1 (49.1–81.0)

Effects of αβ-meATP-induced desensitization

To further characterize the site(s) of action of AP4A and AP5A we studied the effect of desensitization of the P2X1-receptor (see Methods). This procedure abolished contractions evoked by AP5A (3 μm) (control = 2.27 ± 0.12 g, n = 5, Figure 2a) and AP4A (10 μm) (control = 2.48 ± 0.19 g, n = 5, Figure 2b). Higher concentrations of AP5A (10 μm) and AP4A (100 μm) were now also ineffective (not shown). In contrast, contractions evoked by noradrenaline (100 μm) were unaffected (control = 4.11 ± 1.03 g, test = 3.71 ± 0.52 g, n = 6, Figure 2c). This suggests that α,β-meATP selectively desensitized the P2X1-receptor and that both AP4A and AP5A act through this receptor to evoke contraction.

Figure 2.

The effect of desensitization of the P2X-receptor by α,β-meATP. Control responses were obtained to (a) AP4A (30 μm), (b) AP5A (3 μm) or (c) noradrenaline (NA, 100 μm) (first panel); α,β-meATP (50 μm) was then applied (second panel) and once tension had returned to baseline level further 50 μm aliquots were added until the tissue had been exposed to a cumulative concentration of 200 μm α,β-meATP. The drug was washed out and when α,β-meATP (50 μm) was added 2 min later no contraction was seen (third panel), confirming that the P2X-receptor had been desensitized. α,β-meATP was again washed out and two min later AP4A and AP5A no longer evoked contraction, but responses to noradrenaline were unchanged (fourth panel). Note that the vertical scale represents 2 g for contractions evoked by α,β-meATP and noradrenaline, but 1 g for those to AP4A and AP5A.

Effects of ARL 67156 on contractions

We have previously shown that 100 μm ARL 67156 potentiates the peak amplitude of contractions of the guinea-pig vas deferens evoked by ATP (100 μm) by approximately 60%, but has no effect on those to α,β-meATP (Westfall et al., 1996a), consistent with ARL 67156 inhibiting ecto-ATPase. ARL 67156 (100 μm) potentiated the peak response to AP4A (30 μm) by 61% (control = 3.55 ± 0.27 g, test = 5.71 ± 0.39 g, n = 8), but caused a small decrease in the peak amplitude of contractions evoked by AP5A (3 μm) (control = 3.22 ± 0.14 g, test = 2.54 ± 0.24 g, n = 6) (Figure 3). In either case the effect of ARL 67156 reversed rapidly on washout of the drug.

AP4A and AP5A as provided by Sigma are only 95% pure, but analysis with the firefly luciferin-luciferase assay for ATP showed that it accounted for only 1% and 0.5% of the AP5A and AP4A solutions respectively. As both AP4A and AP5A are more potent than ATP in this tissue, it is unlikely that the effects of ARL 67156 on AP4A and AP5A are due to the presence of this small amount of ATP.

Discussion

The results of this study show that α,β-meATP, AP4A and AP5A all evoked contractions of the guinea-pig isolated vas deferens which were inhibited by the P2X-receptor antagonists suramin, PPADS and P-5-P. Of the cloned P2X-receptors, α,β-meATP is a potent agonist at the P2X1- and P2X3-subtypes only. Whilst the P2X1-receptor is present in many tissues, including visceral smooth muscle, the P2X3-receptor has a highly restricted distribution, being selectively expressed at high levels in nociceptive sensory neurones (Collo et al., 1996). Thus, α,β-meATP is likely to have been acting through the P2X1-receptor to evoke contraction in this study.

These experiments showed that for each antagonist, the potency against AP4A than AP5A was not significantly different from that against α,β-meATP, consistent with AP4A and AP5A acting at the same site as α,β-meATP, i.e. the P2X1-receptor. This is supported by the finding that AP5A is an agonist at the cloned P2X1-receptor (Evans et al., 1995). However, both suramin and P-5-P were approximately 2.5 times more potent as antagonists against AP4A than against AP5A. This could suggest that AP4A and AP5A are not acting at the same receptor. This conclusion is difficult to reconcile with the data discussed above, which suggests that both AP4A and P5A act at the same site as α,β-meATP. The apparent differences in antagonist potency are small and the potencies of suramin, PPADS and P-5-P seen in this study are similar to those previously found for antagonism at the P2X1-receptor cloned from human urinary bladder (Evans et al., 1995) and at the P2X1-receptor in smooth muscle preparations (Hoyle et al., 1990; Leff et al., 1990; Lambrecht et al., 1992; Ziganshin et al., 1993; McLaren et al., 1994; Trezise et al., 1994; Bailey & Hourani, 1995; Ralevic et al., 1995; Usune et al., 1996). Thus, the small differences seen may simply be due to experimental variability.

Perhaps the strongest evidence for a common site of action is that contractions evoked by AP4A and AP5A were abolished by desensitization of the P2X1-receptor by α,β-meATP. Contractions elicited by noradrenaline were unaffected, suggesting that desensitization was selective for the P2X1-receptor. Cross-desensitization between α,β-meATP, AP4A and AP5A has also been demonstrated in the urinary bladder of the guinea-pig (Usune et al., 1996) and rat (Hashimoto & Kokubun, 1995) and in the rat mesenteric bed (Ralevic et al., 1995), again consistent with these agonists acting at the same receptor. AP4A and AP5A also show cross-desensitization with ATP in the rat vas deferens (Stone & Paton, 1989). This contrasts with the results of MacKenzie et al. (1988) where AP4A did not mimic the ability of AP5A and α,β-meATP to inhibit the purinergic (P2X1) component of neurogenic contractions in the guinea-pig vas deferens. However in an electrophysiological study, desensitization to AP4A or AP5A attenuated depolarization of the guinea-pig vas deferens evoked by ATP (Hoyle et al., 1995a).

An alternative possibility is that AP4A and AP5A act at separate receptors, which have similar sensitivity to antagonists as the P2X1-receptor, but the results do not fit the properties of any currently known P2-receptor. AP4A, but not AP5A, is a potent agonist at the P2Y2- (P2U-) receptor (Lazarowski et al., 1995), but PPADS has little effect at this site (Charlton et al., 1996). Neither AP4A nor AP5A are agonists at the P2Z-receptor (Steinberg et al., 1987; Tatham et al., 1988) and both are antagonists at the P2T-receptor (see Cusack, 1993). In contrast to our results, suramin is not an antagonist at the proposed separate site for diadenosine nucleotides (Pintor & Miras-Portugal, 1995). Finally, there is no evidence for a functional P2Y-receptor in the guinea-pig vas deferens (Bailey & Hourani, 1995).

In this study the peak amplitude of contractions evoked by AP4A was potentiated by 61% by ARL 67156. We previously showed that ARL 67156 enhanced contractions to ATP to a similar extent, but had no effect on those to α,β-meATP (Westfall et al., 1996a). This suggests that the action of AP4A in the guinea-pig vas deferens is limited by its breakdown by ecto-enzymes, similar to the coronary bed of the rabbit (Pohl et al., 1991), bovine cultured adrenal chromaffin cells (Ramos et al., 1995), guinea-pig left atrium (Hoyle et al., 1996) and several preparations where the action of AP4A depends upon its breakdown to adenosine (Klishin et al., 1994; Ziganshin et al., 1995; Rubino & Burnstock, 1996). In contrast, AP4A is not broken down in the coronary bed of the guinea-pig (Nees, 1989) and rabbit perfused aorta and mesenteric artery (Busse et al., 1988). Thus, the stability of AP4A is tissue-dependent.

The metabolism of AP4A by intracellular and plasma enzymes is much better characterized than that by ecto-enzymes (see Ogilvie, 1992). Outside the cell it has been proposed that AP4A is broken down by ecto-enzymes distinct from ecto-ATPase (Ogilvie, 1992; Ramos et al., 1995), but it is not clear if ARL 67156 potentiated AP4A in the present study because AP4A was metabolized by ecto-ATPase, or if AP4A was broken down by a separate ecto-AP4Aase which is also sensitive to ARL 67156. To date, biochemical studies on the actions of ARL 67156 have only examined the activity of ecto-ATPase and further studies are required against other nucleotidases. Interestingly, of the three agonists used here, the IC50 for suramin was lowest against AP4A. Suramin, as well as acting as a P2-antagonist, also inhibits ecto-ATPase (Bailey & Hourani, 1994) and this can decrease its antagonist potency. This would suggest that AP4A was not metabolized here by a suramin-sensitive enzyme.

In contrast to the potentiation of AP4A, contractions evoked by AP5A were inhibited by ARL 67156. Thus, AP5A appears to be metabolically stable in the guinea-pig vas deferens, unlike in bovine cultured adrenal chromaffin cells where a single ecto-enzyme is thought to break down both AP4A and AP5A (Ramos et al., 1995). How ARL 67156 inhibited AP5A is not clear. ARL 67156 has a pA2 of 3.3 as an antagonist at the P2X-receptor (Crack et al., 1995), but the concentration of ARL 67156 used here (100 μm), would not be expected to produce significant antagonism. Furthermore, 100 μm ARL 67156 did not inhibit α,β-meATP when acting through the P2X1-receptor in this tissue (Westfall et al., 1996a). An alternative possibility is that part of the action of AP5A is dependent upon its breakdown by an ecto-enzyme to other active substances such as ATP. However, if this was the case then it would not appear to be an important pathway as ARL 67156 only inhibited the contractions by about 20%.

Acknowledgments

This work was supported by grants from Astra Charnwood and the Wellcome Trust. C.A.M. was supported by the Physiological Society. PPADS was a kind gift from Dr G. Lambrecht, University of Frankfurt.

Ancillary