2′, 3′-O-(2,4,6,Trinitrophenyl)-ATP and A-317491 are competitive antagonists at a slowly desensitizing chimeric human P2X₃ receptor

Abstract

1. Rapid desensitization of ligand-gated ion channel receptors can alter the apparent activity of receptor modulators, as well as make detection of fast-channel activation difficult. Investigation of the antagonist pharmacology of ATP-sensitive homomeric P2X₃ receptors is limited by agonist-evoked fast-desensitization kinetics.

2. In the present studies, chimeric receptors were created using the coding sequence for the N-terminus and the first transmembrane domain of either the nondesensitizing human P2X_2a or fast-desensitizing P2X₃ receptor joined to the sequence encoding the extracellular loop, second transmembrane domain, and C-terminus of the other receptor (designated P2X_2–3 and P2X_3–2, respectively). These clones were stably transfected into 1321N1 astrocytoma cells for biophysical and pharmacological experiments using both electrophysiological and calcium-imaging methods.

3. Chimeric P2X_2–3 and P2X_3–2 receptors were inwardly rectifying and agonist responses showed desensitization properties similar to the wild-type human P2X_2a and P2X₃ receptors, respectively.

4. The P2X_2–3 chimera displayed an agonist pharmacological profile similar to the P2X₃ wild-type receptor being activated by low concentrations of both ATP and α,β-meATP. In contrast, the P2X_3–2 chimera had markedly reduced sensitivity to both agonists.

5. The P2X₃ receptor antagonists TNP-ATP and A-317491 were shown to be potent, competitive antagonists of the P2X_2–3 chimera (K_i=2.2 and 52.1 nM, respectively), supporting the hypothesis that rapid receptor desensitization can mask the competitive antagonism of wild-type homomeric P2X₃ receptors.

Introduction

The P2X₃ receptor is an ATP-sensitive ligand-gated ion channel that is highly expressed on sensory afferent neurons (Chen et al., 1995; Lewis et al., 1995; Vulchanova et al., 1997). The selective localization of P2X₃ receptors on sensory neurons has generated considerable interest in the role of this receptor in the propagation of pain signaling (Burnstock, 2000; Jarvis & Kowaluk, 2001). The P2X₃ receptor is natively expressed as a functional homomer and also as a heteromultimeric combination with the slowly desensitizing P2X₂ receptor (P2X_2/3 receptors) (Chen et al., 1995; Lewis et al., 1995; Lynch et al., 1999). These P2X₃ and P2X_2/3 receptors share highly similar pharmacological profiles (North, 2002), but differ in their acute desensitization kinetics (Collo et al., 1996; Lewis et al., 1995; Burgard et al., 1999). Consequently, activation of the heteromeric P2X_2/3 receptors can be pharmacologically distinguished from P2X₂ receptor-mediated responses, since it is sensitive to low concentrations of the P2X₃ agonist α,β-meATP, whereas the P2X₂ receptor is not (Collo et al., 1996; Lewis et al., 1995; Bianchi et al., 1999; Lynch et al., 1999).

Characterization of P2X receptor pharmacology has been hindered by the general lack of high-affinity ligands that selectively interact with specific P2X receptor subtypes (Jacobson et al., 2002). Antagonists such as suramin, PPADS, and reactive blue 2 have been traditionally used as P2X receptor antagonists (Jacobson et al., 2002). However, these compounds are relatively nonselective and have low micromolar affinity for P2X receptors (Bianchi et al., 1999). The ATP analog, 2′,3′-O-(2,4,6-trinitrophenyl)-ATP (TNP-ATP), has long been used as a probe for ATP binding sites (Mockett et al., 1994) and was later shown to be an antagonist of native P2X receptors (King et al., 1997). Subsequently, it was demonstrated that TNP-ATP had nanomolar affinity for blocking P2X₁, P2X₃, and P2X_2/3 receptors (Lewis et al., 1998; Thomas et al., 1998; Virginio et al., 1998). Recent investigation of the kinetics of receptor block by suramin, PPADS, and TNP-ATP has indicated that these compounds differ significantly in their respective rates of receptor association and dissociation, and that these differences contribute greatly to their affinity for blocking the P2X_2/3 receptor (Spelta et al., 2002). In addition to these antagonists, a structurally novel non-nucleotide P2X₃ receptor antagonist, A-317491, has been reported, which has a high degree of potency and selectivity for P2X₃-containing channels as compared to its activity at other P2X receptors (Jarvis et al., 2002).

Agonist-induced fast desensitization of homomeric P2X₃ receptors also complicates the pharmacological analysis of these receptors. For example, TNP-ATP was originally reported to be a noncompetitive blocker of homomeric P2X₃ receptors (Virginio et al., 1998). However, analysis of the blockade of nondesensitizing ATP responses on cochlear hair cells (Mockett et al., 1994) and on recombinant human P2X_2/3 receptors (Burgard et al., 2000) provided evidence that TNP-ATP functions as a competitive antagonist. This apparent discrepancy may be related to the fact that the relatively slow off-rate of TNP-ATP from P2X₃ receptors can mask competitive inhibition over time due to the fast desensitization rate of this receptor (Burgard et al., 2000). As noted above, conclusive proof of this hypothesis has been difficult to obtain using the slowly desensitizing heteromeric P2X_2/3 receptor, because the P2X₃ selective agonist, α,β-meATP, can activate P2X₂ receptors at high (>100 μM) concentrations (Spelta et al., 2002), thus obscuring direct measurement of only the heteromeric receptor.

In an attempt to further address the issue of competitive antagonism of P2X₃ receptors, we have examined the ability of TNP-ATP and A-317491 (Figure 1) to block a chimeric human P2X_2–3 receptor that exhibits the acute slow-desensitization kinetics of the human P2X_2a receptor (Lynch et al., 1999), while maintaining the pharmacological properties of the P2X₃ receptor. It has previously been shown through construction of chimeric P2X₁ and P2X₂ receptors that the two membrane-spanning domains (TMD1 and 2) are responsible for different rates of desensitization among these P2X receptor subunits (Werner et al., 1996). In addition, the first membrane-spanning domain has been shown to contribute to receptor gating and modulation of agonist affinity for the P2X₂ receptor (Haines et al., 2001). In the present studies, we have created a pair of chimeric human receptors between P2X_2a and P2X₃ in which the N-terminal domain and TMD1 of one receptor were replaced with the analogous region of the other receptor. This exchange included the initial 41 amino acids of the P2X₃ receptor and the first 59 amino acids of the P2X_2a receptor (Figure 2). To differentiate these chimeras from the heteromeric P2X_2/3 receptor, we refer to them as P2X_2–3 and P2X_3–2 (the first half being the N-terminal portion added and the second half being the original receptor host).

Figure 1.

Chemical structures of A-317419 and TNP-ATP.

Figure 2.

Schematic of the native and chimeric P2X receptors.

Methods

Human chimeric receptor construction

The hP2X_2–3 chimeric receptor cDNA was created by substituting the coding sequence for the amino terminal 41 amino acids of the human P2X₃ receptor with the analogous region of the human P2X_2a receptor (59 amino acids). Making the analogous switch into P2X₂ receptor created the hP2X_3–2 chimera. This was accomplished through the use of sequential overlap-PCR reactions. The first reaction utilized the parent P2X cDNAs as templates. For the hP2X_2–3 chimera, in one reaction, the human P2X2_2a template cDNA was combined with a sense primer directed to the region surrounding the initiation codon of P2X_2a and an antisense primer (5′AGCCTTCTCGTGCAAGAAAACGTACCACACGAAGTAGAGCAG3′). The other reaction employed the human P2X₃ cDNA as a template with an antisense primer directed to the region around the termination codon of P2X₃ and a chimeric sense primer (5′CTGCTCTACTTCGTGTGGTACGTTTTCTTGCACGAGAAGGCT3′). For the hP2X_3–2 chimera, in one reaction, the human P2X₃ template cDNA was combined with a sense primer directed to the region surrounding the initiation codon of P2X₃ and an antisense primer (5′GCTTTTCTGCACGATGAATACCCACCCTACAAAGTAGGAGAT3′). The other reaction employed the human P2X_2a cDNA as a template with an antisense primer directed to the region around the termination codon of P2X_2a and a chimeric sense primer (5′ATCTCCTACTTTGTAGGGTGGGTATTCATCGTGCAGAAAAGC3′). The amplification consisted of 2 min at 95°C, followed by 15 cycles of 94°C 20 s, 49°C 20 s, and 68°C 1 min. Products of these reactions were isolated via agarose gel electrophoresis and used as templates for a second round of amplification. The 5′ P2X_2a initiation primer and 3′ P2X₃ termination primer were used in this reaction. Amplification conditions were as follows: 2 min at 94°C, five cycles of 94°C 30 s, 62°C 20 s, 68°C 90 s, and then 15 additional cycles of 94°C 20 s, 55°C 20 s, and 68°C 90 s Platinum Pfx DNA polymerase and buffers (Life Technologies, Inc.) were used for the amplifications. The amplification product was isolated via gel electrophoresis and cloned into pCRtopo 2.1 vector (Invitrogen). The cDNA for the chimeric receptor was then transferred into the pIRESneo (Clontech) vector for generation of stable cell lines. The human P2X_2a and P2X₃ receptors were cloned as described in Lynch et al. (1999). All data were obtained from human P2X receptors and all chimeric cDNAs were confirmed by sequence analysis using previously described methods (Lynch et al., 1999).

Stable cell line construction and culturing

An expression plasmid encoding the human chimeric receptors was independently transfected into 1321N1 astrocytoma cells using Lipofectamine (Invitrogen, Carlsbad, CA, U.S.A.). Subsequently, stable clones of each receptor were selected using G418 (800 μg ml⁻¹) and grown in culture as previously described (Bianchi et al., 1999; Burgard et al., 1999; Lynch et al., 1999). Cells were then maintained at 37°C in Dulbecco's modified Eagle's medium (DMEM) containing 4.5 mg ml⁻¹ glucose and 4 mM L-glutamine, 10% fetal bovine serum, and 300 μg ml⁻¹ G418. Stable human P2X₃ and P2X_2a cell lines were constructed similarly.

Electrophysiology

For patch-clamp recordings, 1321N1 cells were maintained in an extracellular recording solution (pH 7.4,. 325 mOsm) consisting of (mM): 155 NaCl, 5 KCl, 2 CaCl₂, 1 MgCl₂, 10 HEPES, 12 glucose. Patch pipettes (3–8 MΩ), composed of borosilicate glass, were pulled using a Flaming-Brown P87 micropipette puller (Sutter Instrument Co., San Rafael, CA, U.S.A.). Pipettes were then fire polished using a microforge (Narishige, Tokyo, Japan) and filled with internal solution (pH 7.3, 295 mOsm) consisting of (mM): 140 K-aspartate, 20 NaCl, 10 EGTA, 5 HEPES. Whole-cell recording techniques (Hamill et al., 1981) were utilized to voltage-clamp 1321N1 cells expressing recombinant P2X receptors at −60 mV using an Axopatch 200B amplifier (Axon Instruments, Foster City, CA, U.S.A.). Drugs were applied for 5–15 s at 2–5 min intervals to individual cells voltage-clamped at −60 mV (unless otherwise stated) using a piezo-electric-driven glass theta tube positioned close to the cell as previously described (Burgard et al., 2000). Increasing concentrations of agonist were applied to individual cells to generate concentration–response curves. Data acquisition and analysis were performed using pCLAMP 8.02 (Axon Instruments).

Pharmacology

Pharmacological responses were measured using calcium-imaging techniques as previously described (Bianchi et al., 1999; Burgard et al., 2000). Fluo-4, a fluorescent Ca²⁺-chelating dye, was used as an indicator of the relative levels of intracellular Ca²⁺ in a 96-well format with a Fluorescence Imaging Plate Reader (FLIPR, Molecular Devices). Cells expressing recombinant human P2X receptors were grown to confluence and plated in 96-well black-walled tissue culture plates approximately 18 h prior to the experiment. At 1–2 h before the assay, cells were loaded with fluo-4 AM (2.28 μM; Molecular Probes, Eugene, OR, U.S.A.) in D-PBS and maintained in a dark environment at room temperature. Immediately before the assay, each plate was washed twice with 250 μl D-PBS per well in order to remove extracellular fluo-4 AM. Two 50 μl additions of compounds (prepared in D-PBS) were made to the cells during each experiment. The first compound addition (antagonist or vehicle control) was made and incubation continued for the remainder of the experiment. The second compound addition (test P2X receptor agonist) occurred 3 min after the first and measurement continued for 3 min after this final addition. Data shown are based on the average peak increase in relative fluorescence units, as compared to basal fluorescence for the population of cells from multiple wells treated with a single agonist or antagonist concentration on each coverslip. Concentration–response curves and Schild plots were generated by taking these average values and plotting them as a function of the concentration of the drug. Therefore, with this technique, cells were only exposed to agonist one time, thus removing potential confounding effects of progressive desensitization that may be present in the electrophysiological experiments.

Data analysis

Agonist concentration–response curves were fitted by least-squares regression to the logistic equation:

where Y, min, and max represent the measured minimum and maximum responses respectively; EC₅₀ is the ligand concentration giving half-maximal response; X is the concentration of ligand used, and n_H is the Hill coefficient. Graphs and curve fits were generated using Prism (Graphpad Software, San Diego, CA, U.S.A.).

Competitive inhibition was determined using Schild analysis (Arunlakshana & Schild, 1959). pA₂ values were determined from a least-squares linear regression fitted to the Schild plots using Prism (Graphpad Software, San Diego, CA, U.S.A.).

All reagents were obtained from Sigma Chemical Co. (St Louis, MO, U.S.A.) except A-317491 (synthesized at Abbott Laboratories, Abbott Park, IL, U.S.A.) and TNP-ATP and Fluo-4 (Molecular Probes, Eugene, OR, U.S.A.).

Results

Biophysical and pharmacological properties of chimeric receptors

Chimeric receptors stably transfected into 1321N1 astrocytoma cells were initially characterized using whole-cell patch-clamp electrophysiological techniques. Individual 1321N1 cells expressing the chimeric receptors were voltage clamped at potentials ranging from −100 to +60 mV at 20 mV increments and the average peak whole-cell currents were plotted as a function of these holding potentials. The chimeric P2X_2–3 and P2X_3–2 receptors, similar to the wild-type receptors, showed pronounced inward rectification (Figure 3a,b). For comparison, we show that P2X_2a and P2X₃ receptors expressed in 1321N1 cells also have inwardly rectifying current–voltage relations (Figure 3c,d), similar to what has previously been shown in oocytes (Lynch et al., 1999). These results indicate that this biophysical feature of the channel was not disrupted by the formation of the chimeric receptors.

Figure 3.

Current–voltage relationships of agonist-activated responses in cells expressing P2X receptors. Inwardly rectifying current–voltage curves were observed in both native and chimeric P2X receptors (a–d). Current–voltage curves are shown where the average peak currents evoked by agonist were plotted as a function of the membrane-holding potential (−80 to +60 mV) (P2X_2–3: 10 μM α,β-meATP; P2X₂, P2X_3–2: 10 μM ATP; P2X₃: 3 μM ATP). Chimeric and wild-type receptors were stably expressed in 1321N1 cells and currents were recorded using whole-cell patch-clamp techniques (a–d). Data are mean±s.e.m.

The application of ATP (30 nM–30 μM) to cells expressing the P2X_2–3 chimera produced a concentration-dependent increase in current amplitude with maximal currents reaching 1733±415 pA at 30 μM. Normalized peak currents were averaged, plotted as a function of ATP concentration, and fit with a sigmoidal logistic equation with an EC₅₀ of 327 nM (Figure 4a, Table 1). Cells expressing the P2X_3–2 chimera also had concentration-dependent currents evoked by ATP with maximal peak currents reaching 2003±279 pA. Fits of the averaged normalized data resulted in an EC₅₀=970 nM (Figure 4a). Concentration–response curves for wild-type P2X₂ (EC₅₀=7.4 μM) and P2X₃ (EC₅₀=130 nM) receptors (Figure 3a, dashed lines, Table 1) are shown for comparison.

Figure 4.

Agonist concentration–response curves for wild-type and chimeric P2X receptors. (a) ATP (30 nM–1 mM) stably expressed in 1321 cells. (b) α,β-meATP (100 nM–1 mM). Current amplitudes are expressed as a percentage of the maximum (% maximum) response for each agonist. Data are mean±s.e.m.

Table 1.

Electrophysiological characterization of chimeric and wild-type P2X receptor pharmacology

	ATP		α,β-meATP
	EC50 (μM)	nH	EC50 (μM)	nH
P2X3	0.13±0.06	1.4±0.22	0.62±0.03	1.5±0.15
P2X2–3	0.33±0.03	1.8±0.62	1.97±0.06	1.8±0.31
P2X3–2	0.97±0.01	1.3±0.01	23.4±0.02	1.8±0.07
P2X2	7.36±0.03	1.5±0.11	>300	>2

Each value represents the best fit of a logistic equation to the averaged data from concentration–response curves from three to six cells. Values are mean EC₅₀±s.e.m. (standard error of the mean); n_H=Hill slope.

A concentration-dependent activation of the chimeric receptors was also produced by α,β-meATP (100 nM–100 μM). Maximal currents of 1270±249 pA at 100 μM α,β-meATP were recorded from P2X_2–3-expressing cells. Fits of the averaged normalized α,β-meATP data resulted in an EC₅₀ of 1.97 μM (Figure 4b, Table 1). A similar concentration-dependent effect of α,β-meATP (EC₅₀=23 μM) was seen in cells expressing P2X_3–2 receptors with maximal currents of 1728±893 pA at 100 μM (Figure 4b). The concentration-dependent responses of wild-type P2X₃ are shown for comparison (EC₅₀=0.62 μM). P2X₂ wild-type receptors were insensitive to α,β-meATP up to 300 μM and at 1 mM small currents of only 536±147 pA were activated (Figure 4b). These currents were significantly smaller than that of either chimeric receptor or the wild-type P2X₃ receptor. For comparison, normalized currents were plotted as a function of α,β-meATP concentration and fit with a logistic equation. However, since the responses did not reach a plateau level, it can only be concluded that the EC₅₀ of α,β-meATP is greater than 300 μM for these channels.

In the course of these patch-clamp experiments, it was evident that the kinetics of desensitization of the P2X_2–3 chimeric receptor was more similar to the P2X_2a receptor than to the P2X₃ receptor. Likewise, the reverse chimera, P2X_3–2, showed a faster desensitizing component as compared to the P2X₂ receptor. Figure 5 illustrates the differences in desensitization kinetics in response to near-maximal concentrations of ATP applied for 16 s (3 μM for P2X₃ and P2X_2–3 chimeric receptors, 30 μM for the P2X_3–2 chimeric receptors and 100 μM for the less sensitive P2X_2a receptor).

Figure 5.

Altered desensitization kinetics of human P2X chimeric receptors. Cells were voltage-clamped at −60 mV and maximal concentrations of ATP (3, 30 or 100 μM) were applied for 16 s (denoted by bar). Representative inward currents recorded from cells expressing P2X receptors are shown, illustrating that replacing the N-terminal portion of the rapidly desensitizing P2X₃ receptor (d) with the slowly desensitizing P2X_2a receptor (a), results in a chimeric receptor with slowly desensitizing currents (b). In addition, making the analogous chimera (replacing the N-terminal domain of the P2X₂ receptor with that of P2X₃), results in a receptor with faster desensitizing currents (c). Wild-type and chimeric channel currents were recorded from stably transfected 1321N1 cells using whole-cell patch-clamp techniques.

Pharmacological profile of chimeric receptor using fluorescence-based assay

In parallel with the electrophysiological characterization of the chimeric receptor, we investigated the kinetics and pharmacology of the receptor using calcium-imaging techniques. 1321N1 cells stably expressing the chimeric P2X_2–3 receptor were loaded with Fluo-4, a fluorescent calcium dye, and challenged with 3 μM ATP. Like the electrophysiological results, the chimeric P2X_2–3 receptor did not show the agonist-induced fast-desensitization kinetics seen in the P2X₃ wild-type receptors, whereas the P2X_3–2 chimeras did show increased acute receptor desensitization (Figure 6). The extent and apparent rate of desensitization were similar to the receptor that contributed the N-terminal portion to the chimeric receptor.

Figure 6.

Kinetics of ATP-activated Ca²⁺ influx in 1321N1 cells expressing P2X receptor subtypes. Changes in Fluo-4 fluorescence were measured in response to application of 3 μM ATP (denoted by bars). These are raw FLIPR traces obtained from individual experimental runs and show that the kinetics of the P2X channel responses is similar to that recorded using patch-clamp techniques.

Concentration–response curves were generated for ATP and α,β-meATP at the wild-type P2X_2a and P2X₃ receptors and the chimeric receptors using the calcium influx assay. ATP (10 nM–100 μM) produced concentration-dependent responses at all four receptor subtypes. Both chimeric receptors had steep concentration-dependent responses in response to ATP with maximum fluorescence signals achieved at 1 μM and EC₅₀'s around 250 nM (Figure 7a, Table 2). Similar to the chimeric receptors, ATP showed high affinity for the P2X₃ receptor (EC₅₀=400 nM). In contrast, P2X_2a receptors had an EC₅₀ for ATP of 1.07 μM (Figure 7a, Table 2). Thus, calcium imaging and electrophysiology produced similar measures of agonist sensitivity at both wild-type (P2X₂ and P2X₃) and chimeric (P2X_2–3 and P2X_3–2) receptors.

Figure 7.

Effect of concentration on agonist-activated Ca²⁺ influx for wild-type and chimeric P2X receptors. The chimeric receptors have agonist pharmacology similar to the native receptor contributing the extracellular domains. Agonist concentration–response curves obtained from four different P2X receptors to ATP (a) and α,β-meATP (b). In these calcium-imaging experiments, values were normalized to peak Fluo-4 fluorescence intensities obtained in the presence of agonist. Each curve represents the mean±s.e.m. for two to six experiments under each condition.

Table 2.

Calcium-imaging experiments characterizing the pharmacology of chimeric and wild-type P2X receptors

	ATP	α,β-meATP	ATPγS	BzATP	2-meS-ATP	ADP
	EC50 (μM)	EC50 (μM)	EC50 (μM)	EC50 (μM)	EC50 (μM)	EC50 (μM)
P2X3	0.40±0.03	0.78±0.08	0.55±0.04	0.04±0.03	0.18±0.04	15.00±1.27
P2X2–3	0.27±0.02	1.21±0.11	4.35±0.48	0.01±0.004	0.04±0.004	15.27±2.54
P2X3–2	0.24±0.02	18.84±2.49	0.36±0.06	0.31±0.09	0.29±0.05	32.19±0.60
P2X2	1.07±0.13	>30	0.88±0.04	1.05±0.32	0.56±0.03	>100

Each value represents the mean EC₅₀±s.e.m. values (μM) calculated from 2 to 16 concentration–response curves run in duplicate.

Concentration-dependent responses were also evoked by application of α,β-meATP (1 nM-100 μM) in cells expressing the chimeric and native P2X₃ receptors. Maximum responses for both P2X₃ and the chimeric P2X_2–3 receptors were evoked by application of 10 μM α,β-meATP (Figure 7b). In contrast, the P2X_3–2 required much higher concentrations of α,β-meATP to reach maximal responses (∼100 μM). The calculated EC₅₀ for the chimeric P2X_2–3 receptor was similar to that of the P2X₃ receptor, but significantly less than the P2X_3–2 chimera (Table 2). At the P2X_2a receptor, α,β-meATP did not produce any significant response until 100 μM. This resulted in a steep concentration–response curve that did not plateau, similar to the electrophysiological results. A fit of the normalized data resulted in a calculated EC₅₀ of 43.4 μM with an extremely high Hill slope (9.7). However, since no upper plateau was reached, all that can be concluded from these data are that the EC₅₀ for α,β-meATP is greater than 30 μM at the P2X_2a receptor.

Concentration-dependent effects of other P2X receptors agonists (BzATP, ADP, ATPγS, 2-meS-ATP) were also tested on the two wild-type receptors and chimeric receptors, and the results are summarized in Table 2. In each case, the affinity of the agonist at the P2X_2–3 chimeric receptor was similar to that obtained at the P2X₃ receptor (EC₅₀'s were generally one- to four-fold apart), and significantly different from the P2X_2a receptor (generally greater than five-fold different). One exception to this trend was ATPγS which was relatively insensitive at the P2X_2–3 chimeric receptor compared to both wild-type receptors.

To address whether TNP-ATP is a competitive or noncompetitive antagonist, a Schild analysis was performed on cells expressing the P2X_2–3 receptor. Concentration–response curves were generated for ATP (100 nM–30 μM) using the calcium influx assay in the presence of increasing concentrations of TNP-ATP (0.6–40 nM). Peak ATP responses were plotted as a function of ATP concentration and fitted with logistic equations in the presence of increasing TNP-ATP concentrations (Figure 8a). The resulting EC₅₀'s were then plotted in relation to the concentration of TNP-ATP present. As predicted for a competitive antagonist, TNP-ATP produced a linear rightward shift in the ATP concentration–response curve, which was fitted with a linear regression resulting in a pA₂ value of 2.2 nM with a slope of 1.3 (Figure 8b).

Figure 8.

TNP-ATP is a competitive antagonist of human P2X2-3 chimeric receptor. Agonist concentration curves for ATP were generated in the presence of increasing (n=7) concentrations (0.5–40 nM) of TNP-ATP. (a) Shown are selected curves from the range tested, illustrating the rightward shift and full efficacy of the ATP concentration–response curves. Some intermediate concentrations were removed for clarity. (b) Schild plot of all the competition curve data reveals a linear relationship with slope=1.3 and pA₂=2.2 nM.

Concentration–response curves were generated for ATP (100 nM–30 μM) in the presence of increasing concentrations of the novel P2X antagonist A-317491 (12.5–500 nM) in 1321N1 cells expressing the P2X_2–3 receptor (Figure 9a). Similar to TNP-ATP, a rightward shift in the ATP concentration–response curve was observed, which resulted in a linear Schild plot with a slope of 0.9 and a pA₂ of 52.1 nM (Figure 9b). In addition, A-317491 was virtually inactive at the P2X_3–2 chimeric and showed no concentration-dependent shifts in a competition experiment (data not shown) consistent with P2X_3–2 having a pharmacological profile similar to P2X₂.

Figure 9.

The novel non-nucleotide compound, A-317491, is a competitive antagonist at the P2X_2–3 chimeric receptor. Competition curves for ATP generated in the presence of increasing (n=10) concentrations (12.5–500 nM) of A-317491. (a) Selected curves from the range tested illustrating the rightward shift and full efficacy of the ATP concentration–response curves are shown. Some intermediate concentrations were removed for clarity. (b) Schild plot of all the competition curve data reveals a linear relationship with slope=1.1 and pA₂=52.1 nM.

Discussion

The present studies were undertaken to further characterize the nature of the antagonist block of the P2X₃-containing channels. Previous work from our lab had shown that the relatively slow-off kinetics of the potent antagonist, TNP-ATP, could potentially obscure the competitive antagonism at the fast-desensitizing homomeric P2X₃ receptor (Burgard et al., 2000). Analysis of the heteromeric P2X_2/3 receptor, while overcoming the fast-desensitization properties of the homomeric channel, is complicated by the presence of homomeric P2X₃ and P2X₂ receptors that can coexist with P2X_2/3 receptors in stably transfected mammalian cells created to express P2X_2/3 heteromeric receptors. α,β-meATP shows similar potency to activate both homomeric and heteromeric P2X₃-containing channels and is a relatively poor agonist for the P2X₂ receptor. However, Spelta et al. (2002) have recently demonstrated that α,β-meATP can effectively activate P2X₂ receptors at high (>100 μM) concentrations. This complication in the pharmacological analysis of the pharmacology of the P2X_2/3 receptor may also be dependent on mammalian species homologue (rat vs human) and the method of stable receptor transfection (Jarvis & Burgard, 2002).

Previously, Werner et al. (1996) showed that substituting transmembrane and intracellular domains of P2X₂ receptors could markedly reduce agonist-induced receptor desensitization of the fast-desensitizing P2X₁ receptor. We adopted a similar strategy and created chimeras between P2X₃ and P2X₂ receptors to resolve the mode of action of receptor antagonists such as TNP-ATP and A-317491. These chimeric receptors eliminate the potential complications of both coexpression of multiple channel isoforms and of rapid receptor desensitization.

The pharmacological analysis of the P2X₃ chimeric receptors indicated that responses to receptor antagonists were very similar to the wild-type channel contributing the extracellular domain of the chimera. In addition, the agonist pharmacology of the chimeric P2X_2–3 receptor matches well with that of the wild-type P2X₃ receptor, whereas the P2X_3–2 chimera agonist profile was intermediate between what we have previously observed for the P2X₂ and P2X₃ receptors (Bianchi et al., 1999; Lynch et al., 1999).

The pharmacological and desensitization properties of the P2X_2–3 chimeras make them uniquely suitable to address the mode of action of antagonists such as TNP-ATP. While TNP-ATP had originally been characterized as a low-affinity, but competitive, antagonist of endogenous P2X receptors (Mockett et al., 1994; King et al., 1997), more recent studies had discovered that TNP-ATP was a potent (nM) antagonist of P2X₃- and P2X₁-containing channels (Lewis et al., 1998; Virginio et al., 1998). We demonstrated that TNP-ATP blocked agonist activation of recombinant heteromeric P2X_2/3 receptors in an apparent competitive fashion and provided evidence suggesting that the rapidly desensitizing P2X₃ currents masked the competitive antagonism of TNP-ATP (Burgard et al., 2000). This high-affinity antagonism is determined by the ultra-fast association rate of TNP-ATP (∼50-fold faster than other antagonists) and not its dissociation rate (Spelta et al., 2002), as we had originally concluded (Burgard et al., 2000). The present data demonstrated that TNP-ATP produced a concentration-dependent parallel rightward shift in the effect of ATP at the chimeric P2X_2–3 receptor without altering the peak response, consistent with competitive antagonism. Additionally, A-317491, a novel potent and selective non-nucleotide antagonist of P2X₃-containing channels, was also shown to competitively block chimeric P2X_2–3 responses consistent with its competitive activity at the heteromeric P2X_2/3 receptor (Jarvis et al., 2002). The potencies of both TNP-ATP and A-317491 to block the chimeric P2X_2–3 receptor were similar to what we had previously reported for the heteromeric P2X_2/3 channel (pA₂ values of 2.2 vs 2.0 nM for TNP-ATP and 52.1 vs 232 nM for A-317491) (Burgard et al., 2000; Jarvis et al., 2002). In addition, A-317491 exhibited similar on and off rate constants (unpublished observations) as compared to TNP-ATP (Burgard et al., 2000).

Formation of chimeric receptors was accomplished by swapping the N-terminal domain and first transmembrane domain (TMD) of either human P2X₂ or P2X₃ subunits and attaching them to the extracellular, TMD2, and C-terminal portion of the other channel. Expression of these chimeras resulted in functional receptors with desensitization kinetics similar to their respective native receptors that contributed the N-terminal domain. The degree of desensitization for the P2X_2–3 chimeric receptor was significantly faster than that of native P2X₂ receptors but not quite as fast as native P2X₃ receptors. Similarly, the chimeric P2X_3–2 receptor had dramatically slower desensitization kinetics as compared to the P2X₃ receptor, but not as slow as the native P2X₂ receptor. These desensitization properties were apparent in both whole-cell patch-clamp electrophysiological measurements as well as calcium-imaging experiments. The desensitization kinetics of the chimeric P2X_2–3 and P2X_3–2 receptors is in agreement with similar data reported for chimeric P2X₁ and P2X₂ receptors by Werner et al. (1996). These investigators demonstrated that switching out one TMD was sufficient to reduce the desensitization kinetics of native P2X₁ receptors, but that switching both TMDs were required to impart strong or total desensitization kinetics on P2X₂ receptors (Werner et al., 1996). Conversely, they showed that removal of either TMD1 or 2 was sufficient to remove desensitization from P2X₁ receptors. The present chimeric receptors, however, involved more extensive modifications as they also included the N-terminal domain along with TMD1. Taken together, these results indicate that like P2X₁, the N-terminal portion (along with TMD1) of P2X₃ is sufficient to alter the rate of desensitization in slowly desensitizing P2X receptors.

Based on the pharmacology of the P2X_2–3 receptor, we would have predicted that the P2X_3–2 chimera would have a pharmacological profile similar to the wild-type P2X₂ receptor. However, the P2X_3–2 chimeric receptor often displayed an agonist pharmacology that was intermediate between the two wild-type receptors. This was particularly evident when comparing the relatively potent effect of α,β-meATP at the chimeric receptor, observed in both electrophysiological and calcium-imaging assays, to that at the α,β-meATP-insensitive P2X₂ receptors. Although the P2X_3–2 chimera showed some sensitivity to α,β-meATP, which differentiated it from P2X₂ receptors, it was significantly less potent when compared to its actions at the wild-type P2X₃ receptors or the P2X_2–3 chimera receptors. Interestingly, a chimera constructed by replacing the N-terminal portion of the extracellular domain (Val⁶⁰ to Arg¹⁸⁰) of P2X₃ with the analogous region of P2X₂ was sufficient to confer α,β-meATP sensitivity to the receptor (Koshimizu et al., 2002) but with P2X₂-like desensitization. These results suggest that alteration of the P2X₂ receptor to increase the rate of desensitization of the channel alters the apparent affinity of the agonist.

Recently, two other groups have created chimeric receptors between P2X₂ and P2X₃ in order to investigate the interactions of specific receptor domains in creating the desensitizing phenotype and receptor heteromultimerization (He et al., 2002; Koshimizu et al., 2002). In each case, both intracellular domains of the P2X₃ receptor were replaced with those of the P2X₂ receptor and the resulting chimeric receptors showed altered desensitization rates and agonist sensitivities analogous to the present data for the P2X_3–2 receptor.

The alteration of agonist pharmacology of the P2X_3–2 chimeric receptor, as compared to the wild-type P2X₂ receptor, necessarily limits a clear interpretation of the mode of antagonist block of this chimera. However, these data raise interesting questions concerning the contributions of different regions of the receptor to agonist recognition/binding and the potential influence of channel kinetics on agonist pharmacology. As noted above, He et al. showed that there is a correlation between agonist affinity at the P2X₂ receptor and the degree of desensitization produced by the agonist. Furthermore, these investigators demonstrated that the construction of a chimeric receptor between P2X₂ and P2X₇ receptors reduced the receptor's sensitivity for ATP and blocked C-terminus-dependent desensitization of the chimera. The present data also show apparent shifts in agonist affinity for the P2X_3–2 chimera that correlate with the desensitization properties of the chimeric channel, thus providing evidence that the N-terminal portion of the channel may also contribute to this characteristic of P2X receptors.

In conclusion, our results demonstrate that TNP-ATP and the non-nucleotide antagonist, A-317491, competitively block agonist activation of the P2X₃-containing channels. These findings are consistent with previous reports that suggested this mechanism of action; however, the previous studies were necessarily limited by the lack of truly selective P2X₃ receptor agonists and by the ambiguity associated with the heterologous expression of multiple receptor subunits in the construction of the stably transfected P2X_2/3 receptor cell line.

Abbreviations

α,β-meATP

α,β-methylene adenosine triphosphate

TNP-ATP

2′,3′-O-(2,4,6,trinitrophenyl)-adenosine triphosphate