Abstract
Transcripts for P2X2 and P2X6 subunits are present in rat CNS and frequently colocalize in the same brainstem nuclei. When rat P2X2 (rP2X2) and rat P2X6 (rP2X6) receptors were expressed individually in Xenopus oocytes and studied under voltage-clamp conditions, only homomeric rP2X2 receptors were fully functional and gave rise to large inward currents (2–3 μA) to extracellular ATP. Coexpression of rP2X2 and rP2X6 subunits in Xenopus oocytes resulted in a heteromeric rP2X2/6 receptor, which showed a significantly different phenotype from the wild-type rP2X2receptor. Differences included reduction in agonist potencies and, in some cases (e.g., Ap4A), significant loss of agonist activity. ATP-evoked inward currents were biphasic at the heteromeric rP2X2/6 receptor, particularly when Zn2+ions were present or extracellular pH was lowered. The pH range was narrower for H+ enhancement of ATP responses at the heteromeric rP2X2/6 receptor. Also, H+ions inhibited ATP responses at low pH levels (<pH 6.3). The pH-dependent blocking activity of suramin was changed at this heteromeric receptor, although the potentiating effect of Zn2+ on ATP responses was unchanged. Thus, the rP2X2/6 receptor is a functionally modified P2X2-like receptor with a distinct pattern of pH modulation of ATP activation and suramin blockade. Although homomeric P2X6 receptors function poorly, the P2X6subunit can contribute to functional heteromeric P2X channels and may influence the phenotype of native P2X receptors in those cells in which it is expressed.
P2X receptors are ligand-gated cation channels that when activated by extracellular ATP mediate fast excitation in various cells, including central and peripheral neurons (Burnstock, 1997). Neuronal P2X receptors show considerable differences in their sensitivity to naturally occurring agonists, P2 receptor antagonists, and allosteric modulators and, furthermore, show differences in kinetics of receptor activation and inactivation (Khakh et al., 1995; King, 1998). Such diversity in the operational profiles of ATP-gated ion channels may be attributable to the subunit composition of native P2X receptors, because other classes of ionotropic receptors show differing phenotypes that depend on subunit composition (Barnard et al., 1998). Seven P2X receptor subunits (P2X1–7) have been cloned, each of which is believed to form functional homomeric assemblies (Buell et al., 1996). They can also coassemble with other P2X subunits to form heteromeric P2X receptors of three, or possibly four, protein subunits per ATP-gated ion channel (Kim et al., 1997; Nicke et al., 1998; Torres et al., 1999). Three functional heteromeric P2X receptors have been reported: P2X2/3 (Lewis et al., 1995; Radford et al., 1997), P2X4/6 (Lê et al., 1998), and P2X1/5 (Torres et al., 1998; Haines et al., 1999;Lê et al., 1999). Heteromeric channels composed of splice variants of the same P2X subunit (e.g., mP2X4 and mP2X4a) can also generate a different phenotypic form of the wild-type P2X receptor (Townsend-Nicholson et al., 1999).
The potential for heteropolymerization among P2X1–7 receptor subunits was recently investigated using coimmunoprecipitation procedures (Torres et al., 1999). For P2X subunits concentrated in the CNS (namely P2X2, P2X4, and P2X6) (Collo et al., 1996), epitope-tagged P2X2 and P2X6 subunits or P2X4 and P2X6 subunits (but not P2X2 and P2X4 subunits) were shown to form immunopositive heteromeric assemblies. The functional properties of heteromeric P2X4/6receptors have been established (Lê et al., 1998), but not yet the phenotype of heteromeric P2X2/6 receptors. The result of P2X2 and P2X6subunit coexpression is of considerable interest because of (1) the distinct pH modulation of ATP responses at the homomeric P2X2 receptor (King, 1998), (2) a growing belief that the P2X6 subunit might only contribute to functional channels when other P2X subunits are present (Torres et al., 1999), and (3) the recent identification of a pH-modulated ATP receptor in those nuclei of rat brainstem where P2X2 and P2X6 transcripts have been detected (Thomas et al., 1999; Thomas and Spyer, 2000).
Thus, it was of interest to examine the contribution of the P2X6 subunit, when coexpressed with the pH-modulated P2X2 subunit, to the operational profile of the resultant heteromeric P2X2/6receptor expressed in defolliculated Xenopus oocytes. Differences in the ways heteromeric P2X2/6 and homomeric P2X2 receptors respond to nucleotidic agonists, suramin, pH, and Zn2+ ions were investigated in the oocyte expression system. The results establish the P2X2/6 receptor as the fourth example of a heteropolymeric ATP-gated ion channel that, in this case, possesses a pattern of pH modulation of ATP responses distinct from other known homomeric and heteromeric P2X receptors.
MATERIALS AND METHODS
Oocyte preparation. Xenopus laevis frogs were killed by immersion in a lethal dose in Tricaine (0.4% w/v, in tap water) and then decapitated, and ovarian lobes were removed by blunt dissection. Xenopus oocytes (stages V and VI) were defolliculated by a two-step process involving (1) collagenase treatment (Type IA, 2 mg/ml in Ca2+-free Ringer's solution, for 2–3 hr) and (2) stripping away the follicle cell layer with fine forceps. Defolliculated oocytes do not possess native P1 and P2 receptors (King et al., 1996a,b) and are largely devoid of ecto-ATPases (Ziganshin et al., 1995). Oocytes were stored in Barth's solution (pH 7.5, at 4°C) containing (in mm): NaCl 110, KCl 1, NaHCO3 2.4, Tris HCl 7.5, Ca(NO3)2 0.33, CaCl2 0.41, MgSO4 0.82, supplemented with gentamycin sulfate, 50 μg/l. Cells were injected cytosolically with cRNA for rP2X2 (40 nl, 0.002 μg/μl) or rP2X6 (40 nl, 1 μg/μl) or both rP2X2 and rP2X6 (40 nl of each) and incubated for 48 hr at 18°C in Barth's solution. Thereafter, injected oocytes were kept at 4°C for up to 12 d until they were used in electrophysiological experiments.
Electrophysiology. Membrane currents were recorded from cRNA-injected oocytes using a twin-electrode voltage-clamp amplifier (Axoclamp 2A). The holding potential (Vh) was −50 mV, unless stated otherwise. The voltage-recording and current-recording microelectrodes (1–5 MΩ tip resistance) were filled with 3.0 mKCl. Oocytes were placed in an electrophysiological chamber (volume, 0.5 ml) and superfused with Ringer's solution (5 ml/min, at 18°C) containing (in mm): NaCl 110, KCl 2.5, HEPES 5, CaCl2 1.8, adjusted to pH 7.5. Extracellular pH (pHe) was adjusted with HCl (1.0N) or NaOH (1.0N) to reach the desired level. Electrophysiological data were stored on magnetic tape using a DAT recorder (Sony 1000ES) and displayed using a pen recorder (Gould 2200S).
Drug solutions. ATP and other nucleotides were prepared in Ringer's solution, and the pH of stock solutions was readjusted to the desired level. Agonists were superfused, at the concentrations given in the text, by a gravity-feed continuous flow system allowing the rapid addition and washout of drugs. ATP was added for 120 sec or until the current reached a peak, then washed off with Ringer's solution for a period of 5 min. Where used, antagonists were applied for 5 min before and during the application of agonists.
Agonist responses were normalized to the maximum inward current (Imax) evoked by ATP at pH 7.5, including agonist responses recorded at lower pH levels. At pH 7.5, maximum responses were evoked by 300–1000 μm ATP. The agonist concentration required to evoke 50% of the maximum response (EC50) was taken from Hill plots, using the transform log (I/Imax− I), where I is the peak current evoked by each concentration of ATP.
The potentiating effects of extracellular Zn2+ ions on agonist activity were investigated in two ways. Zn2+ ions were either applied simultaneously with ATP or added to the Ringer's solution for 5 min before ATP was applied (with Zn2+ present).
Statistics and graphs. Data are presented as mean ± SEM of four to seven sets of data from different oocyte batches. Concentration–response curves and inhibition curves were fitted by nonlinear regression analysis using Prism v2.0 (GraphPad). Significant differences were determined by unpaired Student's t test or one-way ANOVA followed by Dunnett's post hoc test, again using Prism v2.0 (GraphPad).
Drugs and reagents. All common salts and reagents were AnalaR grade (Aldrich Chemicals, Poole, UK). ATP and ATPαS were purchased from Boehringer (Mannheim, Germany). 2-Methylthio ATP (2-MeSATP) was obtained from RBI (Natick, MA), and other nucleotides [ATPγS, ADP, AMP, adenosine, UTP, UDP, UMP, uridine, CTP, GTP, ITP, diadenosine polyphosphates (ApnA;n = 2–6), α,β-meATP, β,γ-meATP, and 2′- and 3′-O-(4-benzoyl-benzoyl)ATP (BzATP)] came from Sigma (Poole, UK). Suramin was a gift from Bayer (Newbury, UK).
RESULTS
ATP responses of P2X receptors
In initial experiments, the functionality of homomeric rP2X2 and rP2X6 receptors expressed in Xenopus oocytes was tested against a near-saturating concentration of ATP (100 μm), according to available pharmacological data on homomeric P2X receptors (King, 1998). At a holding potential of −50 mV, ATP-activated rP2X2 receptors produced fast-activating and slowly inactivating inward currents (1993 ± 147 nA;n = 6) (Fig.1A,C). rP2X6 receptors failed to respond to ATP at a holding potential −50 mV, but where increased to −90 mV, the agonist did evoke low-amplitude slowly activating inward currents (4.57 ± 1.31 nA, n = 7) (Fig.1A,C). Control (water-injected) oocytes failed to respond to ATP, at either −50 or −90 mV.
Expression of homomeric and heteromeric P2X receptors. A, Whole-cell inward currents by homomeric rP2X2 and rP2X6 receptors activated by a near-saturating ATP concentration (100 μm, for 60 sec), at the given holding potentials (Vh).B, Whole-cell inward currents by ATP-activated heteromeric rP2X2/6 receptors. ATP responses were often biphasic, showing a transient component (filled arrow) followed by a sustained current. The deactivation of inward current occasionally showed two phases of current decay (open arrow). C, Averaged whole-cell inward currents by homomeric rP2X2, rP2X6, and heteromeric rP2X2/6 receptors activated by ATP (100 μm). The y-axis of the histogram has been truncated to help reveal the small responses by rP2X6 receptors. Data are expressed as mean ± SEM for six to seven cells per determination.
In further experiments, coexpression of rP2X2 and rP2X6 subunits resulted in fast-activating and slowly inactivating inward currents (1516 ± 286 nA,n = 6) (Fig. 1B,C), which were broadly similar in their time course to the ATP responses produced by homomeric rP2X2 receptors. However, ATP-activated heteromeric rP2X2/6 receptors uniquely showed biphasic (transient and sustained) components to the evoked inward currents (Fig. 1B, closed arrow). Such biphasic responses were seen in all cRNA-injected oocytes tested (n = 175), although the amplitude of each component of biphasic currents was variable from response to response. Furthermore, evoked responses would change in an unpredictable manner from biphasic to monophasic currents (and back again) over several successive ATP applications. However, the reproducibility of biphasic inward currents by rP2X2/6 receptors was enhanced when pHe was lowered or Zn2+ ions were present in the bathing solution (see Fig. 5). Deactivation of rP2X2/6receptors frequently comprised two phases of current decay (Fig.1B, open arrow). Neither biphasic inward currents nor biphasic current decays were seen at rP2X2 receptors.
The concentration–response (C–R) relationship was studied for ATP responses at rP2X2 and rP2X2/6 receptors, at pH 7.5 (Fig.2A). ATP was more potent (approximately twofold) at rP2X2 receptors (EC50, 18 ± 2.1 μm; nH = 2.0 ± 0.2) than rP2X2/6 receptors (EC50, 32 ± 1.6 μm; nH = 1.7 ± 0.2) (p < 0.05, unpaired ttest). Since the potency of agonists at rP2X2receptors is strongly affected by pHe, the above differences in ATP activity could potentially be attributed to incorrect pHe measurements. However, rP2X2 and rP2X2/6 receptors responded in different ways to changes in pHe(Fig. 2B). The amplitude of ATP responses at rP2X2 receptors increased over the range of pH 8.0 to 6.3 and was maintained at lower pHe levels (up to pH 5.0). ATP responses at rP2X2/6receptors initially increased in size over the range of pH 8.0 to 6.3, then decreased in amplitude as pHe levels were lowered further. The pKa value for the potentiating phase of the H+ effect was 7.04 ± 0.05 (n = 4) at P2X2/6 receptors, a value not significantly different from that of P2X2 receptors (7.05 ± 0.05; n = 4). However, the slopes of the curves describing the potentiating H+ effect were significantly different (P2X2/6, 1.83 ± 0.31; P2X2, 3.04 ± 0.22; p< 0.05).
ATP activity at homomeric and heteromeric P2X receptors. A, C–R relationship for ATP-activated inward currents at rP2X2 and rP2X2/6 receptors, at pH 7.5. B, The relationship between the amplitude of ATP responses (rP2X2, 3 μm; rP2X2/6, 10 μm; each producing 5% of the maximum response) and the extracellular pH level (range, pH 8.3–5.0) at homomeric and heteromeric P2X receptors. C, The C–R curves for ATP activation of rP2X2/6 receptors at the pHe levels indicated. ATP efficacy was markedly reduced at pH 5.5. D, The C–R curves for ATP activation of rP2X2 receptors. ATP efficacy was not altered at pH 5.5. Curves were fitted by the Hill equation inA–D (solid lines) and by a single exponential function in B (dashed line). Data given as mean ± SEM for four to six cells per curve.
The C–R relationship for ATP was reexamined at different pHe levels for rP2X2/6receptors. ATP potency was increased fourfold at pH 6.5 and 15-fold at pH 5.5 (Table 1; see EC50 values). The maximum response to ATP was unchanged at pH 6.5, but agonist efficacy was significantly reduced (by 76 ± 3%) at pH 5.5 (Fig. 2C). At rP2X2 receptors, acidification of the bathing solution shifted the ATP C–R curve to the left without a reduction in the maximum (Fig. 2D). ATP potency was increased 12-fold at pH 6.5 and 30-fold at pH 5.5 at rP2X2receptors (Table 1; see EC50 values). The effects of lowering pHe were reversed on restoration to pH 7.5 for both rP2X2 and rP2X2/6 receptors.
Effect of extracellular pH on ATP potency
Agonist activity at P2X receptors
ATP, ATPαS, ATPγS, and 2-meSATP are known to be full agonists at rP2X2 receptors (King et al., 1997), and consequently their ability to activate rP2X2/6receptors was investigated. Each nucleotide (30 μm) elicited large, slowly inactivating inward currents at rP2X2/6 receptors, with an apparent potency order of (estimated EC50 value) ATP (29.9 μm) = ATPγS (30.8 μm) > 2-MeSATP (34.8 μm) > ATPαS (40.6 μm) (Fig.3A,C). BzATP was a weak agonist at rP2X2/6 receptors (EC50, 399 μm) (Fig.3A,C). P2X2/6receptors did not respond to ADP, AMP, adenosine, UTP, UDP, UMP, uridine, CTP, GTP, ITP, α,β-meATP, and β,γ-meATP (each tested at 30 and 100 μm) (data not shown). Of the diadenosine polyphosphates tested (ApnA,n = 2–6), Ap4A alone showed activity but proved to be a weak agonist (EC50, >1 mm) (Fig.3B,C). This weak activity contrasted with results from rP2X2 receptors, at which Ap4A is a full agonist (EC50, 15.2 μm) (Pintor et al., 1996).
Nucleotide activation of rP2X2/6receptors. In A, whole-cell inward currents at the heteromeric rP2X2/6 receptor were evoked by ATP, ATPγS, 2-MeSATP, ATPαS, and BzATP (30 μm), each of which is a known agonist of rP2X2 receptors (King et al., 1997). InB, the rP2X2/6 receptor was activated weakly by Ap4A (30 and 300 μm), and the kinetics of activation and deactivation were considerably slower than ATP responses. C, C–R relationship for agonist activation of rP2X2/6 receptors at pH 7.5. Estimates of EC50 values (micromolar concentration) were made using the “2 + 2 assay” method of Arunlakshana and Schild (1959): ATP, 29.9 ± 1.9; ATPγS, 30.8 ± 2.9; ATPαS, 40.6 ± 8.0; 2-MeSATP, 34.8 ± 5.1; BzATP, 399 ± 66; Ap4A, >1000 (n = 4–6). The dashed lineshows the position of the full C–R curve for ATP (redrawn from Fig.2A). Open and filled arrows (in A and B) draw attention to biphasic components of receptor activation and deactivation. Data are given as mean ± SEM for four to six cells per determination.
Suramin blockade at P2X receptors
Suramin is an effective antagonist at rP2X2receptors, at which its potency is enhanced when pHe levels are lowered (King et al., 1997). Similar results were obtained for P2X2/6receptors, with suramin reducing ATP responses in a concentration-dependent manner and its potency enhanced with acidification of the bathing solution (Fig.4A,B). At pH 7.5, suramin was equipotent at rP2X2 and rP2X2/6 receptors (Table2; see IC50values). Differences in blocking activity were only observed at lower pHe levels where, at pH 6.5, the inhibition curve for suramin was biphasic for rP2X2/6 receptors and monophasic for rP2X2 receptors (Fig.4C). Comparison of IC50 values at pH 6.5 revealed that activity indices for each of the two phases of P2X2/6 receptor blockade was significantly different (p < 0.05, unpaired ttest) compared with the IC50 value for rP2X2 receptors (Table 2). The blocking activity of suramin at rP2X2 and rP2X2/6 receptors was reversed on washout, at all pHe levels studied.
Suramin antagonism of rP2X2/6receptors. Shown is antagonism of ATP responses (Vh = −50 mV) at heteromeric rP2X2/6 receptors by suramin at pH 7.5 (A) and pH 6.5 (B). Suramin was effective at micromolar concentrations at pH 7.5, but the concentration range for suramin blockade was extended at pH 6.5.C, Inhibition curves for suramin blockade of ATP responses at rP2X2 and rP2X2/6 receptors at the given pH levels. At pH 6.5, the inhibition curve for rP2X2/6 was fitted best by a biphasic curve. IC50 values are given in Table 2. Open arrows draw attention to biphasic current decays. Data are expressed as mean ± SEM for four to eight cells per curve. The biphasic curve for rP2X2/6 was constructed from eight sets of data, using the results from the first two log10 units of concentration (suramin, 0.001–0.01 μm) to represent the first component of the inhibition curve.
Blockade by suramin of P2X receptors
Actions of Zn2+ ions at P2X receptors
Extracellular Zn2+ is known to potentiate ATP responses at P2X2 receptors (Wildman et al., 1998), although the degree of potentiation depends on whether Zn2+ is applied before, or simultaneously with, the agonist. When applied 5 min before ATP, Zn2+ ions (1–30 μm) progressively increased ATP responses at rP2X2/6receptors, by 6- to 14-fold (averaging 9.82 ± 2.29,n = 6), whereas higher concentrations (30–300 μm) progressively decreased and abolished ATP responses in a concentration-dependent manner (Fig.5A). The potentiating and inhibitory effects were reversed on washout. Zn2+ preincubation also affected ATP responses by clearly increasing the incidence of biphasic inward currents, a phenomenon also seen when pHe levels were lowered (Fig. 5A,C). Where applied simultaneously with ATP, Zn2+ ions (1–300 μm) only caused a concentration-dependent increase (8- to 25-fold; averaging 16.35 ± 4.28, n = 5) in the amplitude of ATP responses at P2X2/6 receptors (Fig. 5B). Without Zn2+ preincubation, the above inhibitory Zn2+ effect was not seen, and the incidence of biphasic ATP responses was inconsistent and infrequent for each oocyte tested. EC50 values for the potentiating effects of Zn2+ ions at rP2X2 and rP2X2/6 receptors were similar (Fig. 5, see legend).
Modulation of ATP responses by Zn2+ and H+ at rP2X2/6 receptors. A, Concentration-dependent potentiation and inhibition of agonist-evoked inward currents by extracellular Zn2+ (1–100 μm) given 5 min before and during ATP application at rP2X2/6 receptors, at pH 7.5. EC50 values (micromolar concentration) for Zn2+ potentiation of ATP responses was rP2X2/6, 6.8 ± 1.0 versus rP2X2, 6.9 ± 1.1 (n = 4).B, Concentration-dependent potentiation of ATP-evoked inward currents by extracellular Zn2+ (1–100 μm) applied simultaneously with the agonist. Under these circumstances, biphasic currents were rarely seen, and the inhibitory action of Zn2+ was lost. EC50 values (micromolar concentration) for Zn2+ potentiation of ATP responses were rP2X2/6, 8.2 ± 0.5 versus rP2X2, 11.7 ± 2.8 (n = 6).C, Concentration-dependent potentiation and inhibition of ATP-evoked inward currents by extracellular H+ions (pH 7.0–5.5) at rP2X2/6 receptors. pKa values (−log10[H+] causing 50% potentiation) were rP2X2/6, 7.04 ± 0.05 versus rP2X2, 7.05 ± 0.05 (n = 4). D, Paired biphasic inward currents evoked by ATP (100 μm, at pH 7.5) at rP2X2/6 receptors with either Mg2+ or Ca2+ (1.8 mm) present in the bathing solution. Substitution of Ca2+ with Mg2+ resulted in a reduction of ATP potency [as shown for rP2X2 receptors (King et al., 1997)] without significantly altering the appearance of biphasic currents. Filled arrows draw attention to transient component of ATP-evoked inward currents (A, C). Data are expressed as mean ± SEM for four to six cells per determination.
DISCUSSION
In the present study, expression of homomeric rP2X6 receptors in defolliculatedXenopus oocytes resulted in functional P2X receptors that, even under heightened conditions for channel activation, only managed to produce low-amplitude responses. Such weak ATP responses were attributed to the activation of just a small number of functional rP2X6 receptors, because defolliculated oocytes do not possess native P1 or P2 receptors to complicate the analysis of agonist actions (King et al., 1996a,b). Homomeric rP2X6 receptors have thus far been reported to function well in human embryonic kidney (HEK) 293 cells (Buell et al., 1996; Collo et al., 1996), to be silent in HEK 293 cells (Torres et al., 1999), or not to function at all in Xenopus oocytes (Soto et al., 1996; Lê et al., 1998). Our initial experiments thus confirmed that there are difficulties associated with rP2X6 receptor expression in Xenopusoocytes and, in all probability, in other cell systems. It is possible that Xenopus oocytes and occasionally HEK 293 cells fail to produce an essential protein necessary to insert P2X6 subunits into the cell membrane. One plausible candidate for this protein is another P2X subunit, perhaps the P2X4 subunit, on the grounds that a P2X4-like cDNA (AF012903) has been isolated from HEK 293 cells (direct submission GenBank by Chang and Chang in 1996) and the P2X4 protein is present at low levels in these cells (Worthington et al., 1999). The heteromeric rP2X4/6 receptor is similar in its functional properties to the operational profile of homomeric rP2X6 receptors (Lê et al., 1998). Because <5% of HEK 293 cells transfected with rP2X6cDNA go on to assemble a functional P2X6 (or possibly P2X4/6-like) receptor (Collo et al., 1996), the P2X4 subunit may not be present in all HEK 293 cells.
Where coexpression of P2X2 and P2X6 subunits in Xenopus oocytes was concerned, our experiments were based on a comparison of the operational profiles of wild-type rP2X2 receptors and heteromeric rP2X2/6 receptors. The rP2X2 receptor has already been characterized in our laboratory in an extensive survey of agonists, antagonists, and modulators at this ATP-gated ion channel (King et al., 1996c, 1997;Pintor et al., 1996; Wildman et al., 1997, 1998, 1999a,b,c). Torres and colleagues (1999) have demonstrated that epitope-tagged rP2X2 and rP2X6 subunits will coprecipitate when expressed in a heterologous expression system. Thus, our present results confirm that functional heteromeric P2X2/6 receptors are indeed formed and inserted into the membrane of Xenopus oocytes. Several key observations were made on this new heteromeric P2X receptor, particularly (1) the nature of the evoked inward currents, (2) the potency of agonists, and (3) the effect of pH on ATP responses and suramin blockade.
ATP-evoked inward currents at heteromeric rP2X2/6receptors were sometimes biphasic in nature, involving transient and sustained components that varied in amplitude from response to response in the same cell. However, the incidence and reproducibility of biphasic responses in each oocyte varied in an unpredictable manner. The incidence of biphasic currents was greater and reproducibility more consistent when extracellular Zn2+ was present (Fig. 5A) or extracellular pH was lowered (Fig.5C). Biphasic currents have already been reported at homomeric rP2X4 receptors, at which time-dependent changes in channel permeability were observed and shifts in the reversal potential for ATP-evoked currents noted (Khakh et al., 1999; Virginio et al., 1999). The binary permeability properties of rP2X4 receptors were seen only when extracellular Ca2+ levels were lowered or zero Ca2+o conditions imposed (Khakh et al., 1999). Therefore, we explored this possibility and found that biphasic responses at heteromeric rP2X2/6 receptors were not enhanced when Ca2+ was replaced with equimolar Mg2+ (Fig. 5D). Ca2+-independent binary permeability properties have been reported for homomeric rP2X2receptors, although the time- and concentration-dependent changes in permeability do not result in biphasic currents to ATP (Khakh et al., 1999; Virginio et al., 1999). Others have reported, however, that rP2X2 receptor ion channels do not show significant changes in unitary conductance or reversal potential of whole-cell currents (Ding and Sachs, 1999b). This inconsistency with the P2X2 receptor is reminiscent of the variability of agonist responses (monophasic and biphasic) at the heteromeric P2X2/6 receptor. Currently, there is no satisfactory explanation for biphasic ATP responses at heteromeric P2X2/6 receptors.
The potency of ATP was lower at heteromeric rP2X2/6 receptors than homomeric rP2X2 receptors, regardless of the pH level studied (Table 1). Although ATP potency was decreased overall, the rank potency order for mononucleotidic agonists at the heteromeric receptor remained the same as at the rP2X2 receptor, namely ATP = ATPγS > 2-MeSATP > ATPαS > BzATP. One significant difference in agonist activity involved the dinucleotide diadenosine tetraphosphate (Ap4A), which is a full and potent agonist at rP2X2receptors (Pintor et al., 1996; Wildman et al., 1999a) and only a weak agonist at rP2X2/6 receptors. This difference in Ap4A activity is potentially important, because this dinucleotide occurs naturally and is released in a Ca2+-dependent manner from central synaptosomes in rat brain (Pintor et al., 1992). Therefore, Ap4A may subserve a transmitter role at homomeric rP2X2 receptors but not at heteromeric rP2X2/6 receptors.
Extracellular pH is known to exert a profound effect on ATP potency at homomeric rP2X2 receptors (King et al., 1996c,1997; Stoop et al., 1997; Wildman et al., 1997, 1998, 1999b,c; Stoop and Quayle, 1998; Ding and Sachs, 1999a). A secondary inhibitory effect is observed at very low pH levels (e.g., pH 4.2), at which ATP responses rapidly desensitize, yet recover quickly, if pHe is reversed to levels above pH 5.0, a phenomenon called “fade and rebound” (Stoop and Quayle, 1998). The heteromeric rP2X2/6 receptor showed both the potentiating and inhibitory effects of extracellular H+, and the pH ranges for these two separate effects are compressed when compared with rP2X2 receptors. The inhibitory effect was caused by a reduction in agonist efficacy alone and not a decrease in agonist potency, as evidenced by the lower maximum for the ATP C–R curve at pH 5.5 (Fig. 2C). Because some homomeric P2X receptors (rP2X1, rP2X3, rP2X4, and rP2X7) also show a reduction in ATP activity when pHe is lowered (Virginio et al., 1997; Wildman et al., 1999c), it is conceivable that the observed H+inhibitory effect at rP2X2/6 receptors is caused as much by an action of H+ ions at the rP2X6 subunit as at the rP2X2 subunit.
The potency of suramin is progressively enhanced at rP2X2 receptors as pH is lowered, with blockade occurring at nanomolar concentrations at pH 5.5 (King et al., 1997). The present results now show that this is an attribute shared by heteromeric rP2X2/6 receptors, although subtle differences were observed at pH 6.5 for suramin blockade of homomeric rP2X2 and heteromeric rP2X2/6 receptors. There appeared to be high-affinity (I1) and low-affinity (I2) sites for suramin at rP2X2/6 receptors, and activity indices for each component failed to match the corresponding IC50value at rP2X2 receptors. The precise cause of this unusual effect is as yet unresolved. However, one possibility may involve differences in the subunit composition of heteromeric P2X2/6 receptors, if subpopulations of oligomeric assemblies containing different numbers of rP2X6subunits were generated. Where shown to be functional, the homomeric rP2X6 receptor (or even the heteromeric rP2X4/6 receptor) has been reported to be relatively insensitive to suramin blockade (Collo et al., 1996;Lê et al., 1998). The suramin insensitivity of the P2X6 subunit might help contribute to biphasic inhibition curves seen at pH 6.5 with the heteromeric P2X2/6 receptor.
The potentiating effect of extracellular Zn2+ was not significantly different at rP2X2 and rP2X2/6receptors. However, one subtle difference was noted when using high concentrations (≥ 100 μm) of this transition metal, which appeared to directly activate the heteromeric rP2X2/6 receptors without the need for exogenous ATP (data not shown). It is known that Xenopus oocytes continuously extrude small amounts of intracellular ATP via a mechanogated transport pathway (Nakamura and Strittmatter, 1996), and consequently the potency of locally released ATP may be sufficiently elevated by Zn2+ ions to explain the apparent Zn2+-activated inward currents. The subsequent inhibition of ATP responses by high concentrations of Zn2+ ions may be caused by a gradual desensitization of the receptor pool by locally released ATP.
In conclusion, the heteromeric P2X2/6 receptor possesses a significantly different operational profile from the wild-type P2X2 receptor. It is of interest to us that rP2X2 and rP2X6transcripts are found in rat brainstem (Collo et al., 1996; Comer et al., 1997) in nuclei with demonstrable pH-dependent chemoreceptive inputs (Thomas et al., 1999). The pH modulation of the homomeric P2X2 and heteromeric rP2X2/6 receptor forms an interesting basis for examining the recently discovered involvement of ATP receptors in the CO2-evoked (and pH-dependent) changes in central respiratory drive in rat (Thomas et al., 1999; Thomas and Spyer, 2000). At this point in time, however, the present results establish the P2X2/6 receptor as the fourth example of a heteropolymeric ATP-gated ion channel, which in this case possesses a pattern of pH modulation of ATP responses distinct from other known homomeric and heteromeric P2X receptors.
Footnotes
This work was supported by grants from British Heart Foundation, Biotechnology and Biological Sciences Research Council (BBSRC, UK), and Roche Bioscience (Palo Alto, CA). We acknowledge Dr. David Julius (University of California San Francisco) and Dr. Gary Buell (Ares Sereno, Geneva) for the gifts of cDNAs encoding rP2X2 and rP2X6 subunits.
Correspondence should be addressed to Dr. Brian F. King, Autonomic Neuroscience Institute, Royal Free and University College Medical School, Royal Free Campus, Rowland Hill Street, Hampstead, London NW3 2PF, UK. E-mail: b.king{at}ucl.ac.uk.
Dr. Thomas's current address: Department of Physiology, University of Birmingham, Birmingham B15 2TT, UK.
Dr. Townsend-Nicholson's current address: Department of Physiology, University College London, Royal Free Campus, London NW3 2PF, UK.
Dr. Wildman's current address: Imperial College, Biophysics Section, London SW7 2BZ, UK.
