The Role of Positively Charged Amino Acids in ATP Recognition by Human P2X1 Receptors*

P2X receptors for ATP are a family of ligand-gated cation channels. There are 11 conserved positive charges in the extracellular loop of P2X receptors. We have generated point mutants of these conserved residues (either Lys → Arg, Lys → Ala, Arg → Lys, or Arg → Ala) in the human P2X1receptor to determine their contribution to the binding of negatively charged ATP. ATP evoked concentration-dependent (EC50 ∼ 0.8 μm) desensitizing responses at wild-type (WT) P2X1 receptors expressed inXenopus oocytes. Suramin produced a parallel rightward shift in the concentration response curve with an estimated pKB of 6.7. Substitution of amino acids at positions Lys-53, Lys-190, Lys-215, Lys-325, Arg-202, Arg-305, and Arg-314 either had no effect or only a small change in ATP potency, time course, and/or suramin sensitivity. Modest changes in ATP potency were observed for mutants at K70R and R292K/A (20- and 100-fold decrease, respectively). Mutations at residues K68A and K309A reduced the potency of ATP by >1400-fold and prolonged the time course of the P2X1 receptor current but had no effect on suramin antagonism. Lys-68, Lys-70, Arg-292, and Lys-309 are close to the predicted transmembrane domains of the receptor and suggest that the ATP binding pocket may form close to the channel vestibule.

P2X receptors for ATP are ligand-gated cation channels present on many different cell types including neurons, blood cells, and smooth muscle (1). The P2X 1 receptor was originally cloned from the rat vas deferens, and its properties, in particular the rapid desensitization and sensitivity to ␣,␤-meATP, correspond closely to those of the native smooth muscle phenotype (2). This has been confirmed in recent studies on P2X 1 receptor-deficient mice that showed the P2X 1 receptor is essential for the expression of functional P2X receptors in smooth muscle (3).
Seven P2X receptors (P2X 1-7 ) have been identified at the molecular level (4), and they constitute a novel family of ion channels with two transmembrane domains, intracellular amino and carboxyl termini and a large extracellular loop (5). The receptors form as either homo-or heteromultimers (6,7) from the association of at least three P2X receptor subunits (8). The second transmembrane domain lines the ion conducting pore (9), and residues on the amino and carboxyl termini are involved in determining the time course of the response of P2X 2 receptors (10 -13). The extracellular loop is thought to be the site of ATP binding, and residues that affect antagonist action have been described (14,15). A Walker ATP binding motif (16) is not present in P2X receptors, and to date no residues associated with agonist binding have been identified.
In many ATP-binding proteins, positively charged amino acids have been shown to be important in co-ordinating ATP binding. One of the key components of the Walker motif is the lysine residue, which is thought to interact directly with one of the phosphate groups of the ATP molecule (17) (for a review, see Ref. 18). Lysine residues are also important in determining the actions of ATP at proteins that do not have the Walker motif, e.g. the P2Y 2 receptor (19) and ATP-sensitive potassium channels (20). Arginine residues have also been suggested to play a role in ATP binding at P2Y receptors (19,21). In the extracellular loop of the P2X receptors, there are seven conserved lysine residues, three conserved arginine residues, and one position where there is either an arginine or lysine (Fig. 1). It is thus possible that conserved positively charged residues may contribute the binding of the negatively charged phosphate group of ATP at P2X receptors.
P2X receptors may provide novel targets for the development of drugs for the treatment of a variety of conditions including male fertility (P2X 1 receptors) (3) and pain management (P2X 3 receptors) (6,22). An understanding of the molecular basis of agonist recognition by P2X receptors may therefore aid in rational drug design. We have used site-directed mutagenesis and the electrophysiological characterization of recombinant receptors in Xenopus oocytes to determine the role of conserved positively charged amino acids in the extracellular loop of the human P2X 1 receptor in mediating the actions of ATP.

EXPERIMENTAL PROCEDURES
Cloning and Mutagenesis of the Human P2X 1 Receptor-Oligonucleotide primers designed to amplify positions 43-1631 of the published human P2X 1 cDNA sequence (23) (EMBL A47363) were utilized for reverse transcription-PCR 1 on a surgical sample of human bladder. Total RNA was isolated using an RNeasy kit (Qiagen), and 5 g was used in a first strand cDNA reaction using oligo(dT) 17 primer and Superscript II reverse transcriptase according to the manufacturers instructions (Life Technologies, Inc.). The resulting first strand cDNA (1 l) was then used directly as template in a PCR reaction containing 200 M each dNTP, 1.5 mM MgCl 2 , 25 pmol of forward primer (5Ј-CACCGCCCTGCTCTTCCTAA-3Ј), 25 pmol of reverse primer (5Ј-GC-CCAACCCCCAGCCATTCC-3Ј), 1ϫ Opti-Buffer (Bioline), and 2.5 units of Bio-X-Act Taq DNA polymerase (Bioline). Thermal cycling consisted of 30 repetitions of 94°C for 1 min, 60.9°C for 1 min, and 72°C for 1.5 min. This was followed by incubation at 72°C for 7 min. PCR product was separated on a 1% agarose gel, recovered, and cloned into a pcDNA3.1 (Invitrogen) plasmid, which had been modified to contain a poly(A) tail that would be adjacent to the 3Ј-untranslated region of the cloned P2X 1 cDNA. The cloned insert was sequenced on both strands using vector and insert-specific primers (Automated ABI Sequencing Service, Leicester University, Leicester, United Kingdom).
Point mutations in the human P2X 1 plasmid construct were introduced using the QuikChange™ mutagenesis kit (Stratagene) according to the manufacturer's instructions. Lysines at positions 53, 68, 70, 190, 215, 309, and 325 were mutated to arginine and alanine. Arginines at positions 202, 292, 305, and 314 were mutated to lysine and alanine. Introduction of the correct mutation and the absence of spontaneous mutations was confirmed by DNA sequencing (Automated ABI Sequencing Service, Leicester University).
Xenopus Oocyte Expression System-Mutant and wild type plasmids were digested with MluI to release a 2531-bp fragment containing the vector T7 RNA polymerase site, the P2X 1 cDNA, and the vector poly(A) tail. Sense strand cRNA was generated from this fragment with a T7 mMessage mMachine™ kit (Ambion), quantified by spectrophotometry, and dissolved in nuclease-free water at a concentration of 1.0 g/l. Manually defolliculated stage V Xenopus oocytes were injected with 50 nl (50 ng) of cRNA using an InjectϩMatic microinjector (J. Alejandro Gaby, Genéva, Switzerland) and stored at 18°C in ND96 buffer (96 mM NaCl, 2 mM KCl, 1.8 mM CaCl 2 , 1 mM MgCl 2 , 5 mM sodium pyruvate, 5 mM HEPES, pH 7.6) prior to recording 3-7 days later.
Electrophysiological Recordings-Two-electrode voltage clamp recordings were made from oocytes using a GeneClamp 500B amplifier (Axon Instruments). Microelectrodes were filled with 3 M KCl and the external solution consisted of ND96 with 1.8 mM BaCl 2 replacing the 1.8 mM CaCl 2 in order to prevent activation of endogenous calcium-activated chloride channels (2). Membrane currents were recorded at a holding potential of Ϫ60 mV and were acquired using a Digidata 1200 analog to digital converter with pClamp 7 acquisition software (Axon Instruments). ATP (magnesium salt, Sigma) was applied from a nearby U-tube perfusion system (24), whereas the antagonist suramin (Bayer) was bath-perfused and also present at the appropriate concentration in the U-tube application of ATP. Repeated exposures of agonist were separated by 5 min in order to allow recovery from receptor desensitization. Non-injected and water-injected oocytes tested from at least seven separate batches of oocytes (Ͼ50 oocytes) gave no detectable currents in response to ATP application (range: 100 M to 1 mM). Data are presented as mean Ϯ S.E. Differences between means were tested using Student's paired t test. Concentration response data were fitted with the equation Y ϭ ((X) H ⅐M)/((X) H ϩ (EC 50 ) H ) where Y ϭ response, X ϭ agonist concentration, H is the Hill coefficient, M is maximum response, and EC 50 is the concentration of agonist evoking 50% of the maximum response. pEC 50 is the Ϫlog 10 of the EC 50 value. The concentration ratio for suramin (3 M) was determined (EC 50 value for ATP in suramin/EC 50 value for ATP with no suramin) and used to estimate antagonist potency; pK B estimate ϭ log 10 (concentration ratio Ϫ 1) Ϫ log 10 [suramin concentration]. For the mutants K68A and K309A, concentration-response curves did not reach a plateau at the highest ATP concentration tested (10 mM).
Protein Expression Analysis-Expression levels of mutant receptors, which were either nonfunctional or gave very low peak currents, were estimated by Western blot analysis and biotinylation of cell surface proteins. For Western blot analysis, oocytes injected previously with 50 ng of cRNA were homogenized in buffer H (100 mM NaCl, 20 mM Tris⅐Cl, pH 7.4, 1% Triton X-100, 10 l/ml protease inhibitor mixture (Sigma P8340)) at a dilution of 20 l/oocyte. After centrifugation at 16,000 ϫ g (4°C) for 2 min, a 7.5-l aliquot of the supernatant was mixed 50:50 with gel sample buffer (50 mM Tris⅐Cl, pH 6.8, 100 mM dithiothreitol, 2% SDS, 10% glycerol), heated for 2 min at 90°C, and separated on a 10% SDS-PAGE gel. The gel was transferred to nitrocellulose and blocked in TBST (20 mM Tris⅐Cl, pH 7.6, 145 mM NaCl, 0.05% Tween 20) plus 5% milk powder overnight. The membrane was incubated with anti-P2X 1 antibody (1:500 dilution) (Alamone, Israel) in TBST ϩ 5% milk powder for 1.5 h, washed three times for 5 min in TBST, and incubated with anti-rabbit horseradish peroxidase secondary antibody (1:1000 dilution) (Sigma) for 40 min. After three 5-min washes in TBST, visualization of the protein bands was achieved with an ECL (Plus) kit (Amersham Pharmacia Biotech) according to the manufacturer's instructions. In order to confirm that synthesized mutant receptors had been processed correctly and transported to the cell membrane of the oocyte, biotinylation of all surface proteins with Sulfo-NHS-LC-Biotin (Pierce) was used. This compound is impermeable to the cell membrane and, in intact cells, will only biotinylate proteins present on the cell surface. Oocytes were washed to remove contaminating proteins and incubated in 0.5 mg/ml Sulfo-NHS-LC-Biotin in ND96 buffer for 30 min. After washing five times in ND96, oocytes were homogenized in buffer H (20 l/oocyte) and centrifuged at 16,000 ϫ g (4°C) for 2 min. Anti-P2X 1 antibody (20 g/ml final concentration) (Alamone) was added to an aliquot of the supernatant, which was then incubated on ice for 1 h. The antibody-antigen complex was then precipitated by the addition of Protein A-Sepharose (Amersham Pharmacia Biotech) according to the manufacturer's instructions. Gel sample buffer was added and the samples heated to 90°C for 2 min before separation on a 10% SDS-PAGE gel, which was subsequently transferred to nitrocellulose. The membrane was blocked overnight in TBSTϩ 5% milk powder and incubated with streptavidin-horseradish peroxidase (Pierce) (0.2 g/ml) in TBST for 30 min. After washing extensively in TBST, biotinylated protein bands on the membrane were visualized with ECL (Plus) (Amersham Pharmacia Biotech). 1 Receptors-ATP evoked concentration-dependent P2X 1 receptor-mediated currents from WT receptors with an EC 50 value of ϳ0.8 M (as reported previously) (25) (Fig. 2, Table I). The FIG. 1. Alignment of amino acid residues in the extracellular region of P2X receptors. The amino acid sequence of the human P2X 1 extracellular loop is shown aligned with equivalent regions of the human P2X 3-7 and the rat P2X 2 sequences. The rat sequence was used, as no amino acid sequence for the human P2X 2 receptor has been published. Positive charges (lysine or arginine) that are conserved throughout the whole family of P2X isoforms are boxed and correspond to the positions mutated in this study. Numbering shown above the sequence corresponds to the human P2X 1 sequence. These positively charged residues are conserved in all 16 known P2X sequences from rat, human, guinea pig, and mouse. Only one example of each isoform is shown here for clarity. SwissProt accession numbers for sequences in the alignment are P51575, P49653, P56373, Q99571, Q93086, O15547, and Q99572 for P2X 1-7 , respectively. ability of ATP to evoke functional responses, its potency, is the result of two steps: 1) the binding of ATP to the receptor (affinity) and 2) the ability of bound ATP to open the ion channel (efficacy). Therefore, in the present study, we refer to the potency of ATP and can only make inferences about the effects of the mutations on the affinity of ATP for the receptor.

Effects of Mutations on ATP Concentration Response of P2X
To investigate the role of conserved lysine residues in the activation of P2X 1 receptors by ATP, lysine to arginine (to maintain the positive charge; Fig. 2) or lysine to alanine (to neutralize the positive charge; Fig. 3) point mutants were generated and tested (Table I). Point mutants at conserved lysines 53, 215, and 325 either had no effect (K215R, K325R (Fig. 2, A and B), K325A) or only a small (Ͻ2-fold) decrease in ATP potency (K53A/R and K215A). K190R resulted in a small (ϳ2fold) decrease in ATP potency. However, when the positive charge was removed at this position (K190A), the potency was decreased by ϳ5-fold (Fig. 3). Mutations at K70 resulted in either an ϳ5or ϳ18-fold decrease in ATP potency for K70A and K70R (EC 50 values for ATP were ϳ4 and 15 M, respectively). The conservative substitution K309R (Fig. 2, A and B) resulted in a modest ϳ25-fold decrease in ATP potency (EC 50 ϳ 20 M); however, when the positive charge was removed, there was a more substantial ϳ1400-fold decrease in ATP potency (EC 50 Ͼ1.2 mM, Fig. 3, A and B). This suggests that both the charge and chemical properties of the lysine residue are important at this position. The point mutation K68A resulted in a substantial Ͼ1800-fold decrease in ATP potency (EC 50 Ͼ1.5 mM, Fig. 3, A and B). The K68R P2X 1 receptor did not form functional channels, although the protein was expressed at the membrane (see "P2X Receptor Expression Analysis").
There are three conserved arginine residues in the extracellular loop of P2X receptors, and at the position equivalent to amino acid 202 there is either an arginine or a lysine residue. These residues have been individually mutated to either lysine or alanine residues (Fig. 4, Table I). Mutations at residues Arg-202, Arg-305, and Arg-314 resulted in either no effect on the EC 50 for ATP (R202K, R314K) or only a small ϳ2-3-fold decrease in ATP potency (R202A, R305K, R305A, R314A). Mutation of Arg-292 to either lysine or alanine resulted in an ϳ90 -120-fold decrease in ATP potency at the P2X 1 receptor.
Effects of Mutations on the Time Course of P2X 1 Receptor Responses-Responses of WT P2X 1 receptors to ATP were transient and decayed during continued agonist application; this was particularly noticeable at high agonist concentrations where the response to 10 M ATP (a concentration that evoked 90% of the maximum response; EC 90 ) decayed to 50% of the peak response in 0.85 s (Fig. 5, Table I). When constructing the concentration-response curves for mutant P2X 1 receptors, it was apparent that the time course of the currents evoked by ATP at some of the mutant receptors was affected (Figs. 2-4). To quantify this we determined the time from peak current to 50% decay after an EC 90 concentration of ATP was applied (adjusted for each mutant receptor) ( Table I, Fig. 5). The decay of mutant K53R, R305A, and R314A P2X 1 receptor mediated responses was approximately twice as fast as for WT P2X 1 receptors. The mutations K53A, K70R, K215R, K215A, K325R, K325A, R202A, R202K, R305K, and R314K had no effect on the time course of the P2X 1 receptor current. A small Ͻ3-fold decrease in the rate of decay of P2X 1 receptor currents was seen for the point mutants K70A, K190R, and K190A. The mutants K309R, K309A, and K292 showed a modest 8 -12-fold increase FIG. 2. Effects of lysine to arginine mutations on ATP potency. Two-electrode voltage clamp recordings were made from oocytes expressing WT or lysine to arginine mutant P2X 1 receptors. A, P2X 1 receptor currents recorded in response to ATP (micromolar, unless otherwise stated, indicated by bar) from WT, K325R, and K309R mutants. B, concentration-response curves for P2X 1 WT and mutant receptors. Mean currents were normalized to the maximal response (n ϭ 6 -8 oocytes). Holding membrane potential was Ϫ60 mV.

TABLE I
Summary of data for wild-type and mutant P2X 1 receptors pEC 50 is the Ϫlog 10 of the ATP EC 50 value. 50% time corresponds to the time taken from peak current to 50% decay of the current after an application of ATP at EC 90 . Peak I corresponds to the peak current after an application of ATP that gave the maximal response for that particular mutant. Suramin pK B ϭ log 10 (concentration ratio Ϫ 1) Ϫ log 10 [suramin concentration]. Value are shown as means Ϯ standard error of the mean (n ϭ 6 -8 oocytes, except for suramin pK B values where n ϭ 4 oocytes). Significant differences from the wild type are indicated as ** (p Ͻ 0.01) and * (p Ͻ 0.05). The mutant K68R was nonfunctional. Antagonism by suramin at the R305K mutation was not competitive; therefore, no estimate of pK B was made. NA, not applicable. a Full concentration-response relationships could not be constructed in the presence of suramin; pK B was therefore estimated from concentration ratio of EC 20 .
b Full concentration-response relationships could not be constructed in the presence of suramin; pK B was therefore estimated from concentration ratio of EC 30 . in the time course of the P2X 1 current. The most marked changed in time course was seen for the K68A P2X 1 receptor mutant where there was an ϳ90-fold increase in the time to 50% decay of the response to ϳ75 s (Fig. 5).

Effects of Mutations on Antagonism by Suramin-The P2
receptor antagonist suramin (3 M) resulted in a parallel shift (16-fold increase in the EC 50 for ATP in the presence of suramin) in the concentration response to ATP at recombinant WT P2X 1 receptors similar to that reported previously (25) (Fig.  6A). The parallel shift in the concentration response indicates that the suramin antagonism is competitive (see Ref. 26) and yields a pK B estimate of ϳ6.7 for suramin. We have estimated the pK B value for suramin (at 3 M) on mutant P2X 1 receptors to test whether the mutations in the P2X 1 receptor have affected the ability of suramin to antagonize the response to ATP. The mutants K68A, K70A, K190R, K190A, R202K, R292K, R305A, K309A, R314K, R314A, K325R, and K325A had no effect on the pK B for suramin. For R305K, the antagonism by suramin was not competitive and maximum responses in the presence of suramin were only ϳ60% of the response in the absence of suramin. Suramin was a less effective antagonist at K53A, K53R, K202A, K215A, and K292A mutant P2X 1 receptors resulting in shifts in the EC 50 for ATP of between 4-and 8-fold. In contrast suramin was a more effective antagonist at K70R (Fig. 6B) and K215R receptors leading to a 40 -120-fold shift in the EC 50 for ATP. For K68A and K309A in the presence of suramin, responses to the maximal concentration of ATP tested (10 mM) were 25.8 Ϯ 4.0 and 35.8 Ϯ 1.9%, respectively, of those in the absence of suramin. We have therefore estimated the pK B from the concentration ratio at the EC 20 and EC 30 values, respectively. These results are of particular interest as in general mutants that caused the largest increases in EC 50 for ATP either had no effect (K68A, K70A, R292K, K309A) or increased (K309R, K7OR) the affinity of suramin,  4. Effects of arginine to lysine and arginine to alanine mutations on ATP potency. Two-electrode voltage clamp recordings from oocytes expressing arginine to lysine and arginine to alanine mutants of the P2X 1 receptor. A, representative currents recorded in response to the application of different concentrations (micromolar unless otherwise stated) of ATP (indicated by bar) for R314A, R292A, and R292K mutants. B, concentration-response curves. Mean currents were normalized to the maximal response (n ϭ 6 -8 oocytes). Holding membrane potential was Ϫ60 mV.

FIG. 5. Effects of mutations on the time course of P2X 1 receptor responses.
A, P2X 1 receptor currents recorded from oocytes injected with wild type, K309R, and K68A cRNA. ATP was applied at EC 90 concentration (indicated by bar). Wild type currents decayed rapidly (ϳ0.9 s from peak current to 50% decay), and K309A mutants showed an intermediate time of decay (ϳ7 s), whereas K68A mutants showed a greatly extended (ϳ75 s) decay time. B, 50% decay times for mutants that have a significant difference (**, p Ͻ 0.01; *, p Ͻ 0.05) from the wild type (n ϭ 6 -8 oocytes). Holding membrane potentials were Ϫ60 mV.
indicating that these mutations have not had a major effect on the conformation of the P2X 1 receptor.
P2X Receptor Expression Analysis-Inward currents in response to ATP application were not recorded for the K68R mutant P2X 1 receptor, and only small responses were recorded from the mutants K68A, R305A, and R305K (Table I). We have used a P2X 1 receptor-selective antibody to investigate the level of receptor expression. In Western analysis bands corresponding to a protein of ϳ60 kDa were detected with an antibody directed to the carboxyl terminus of the P2X 1 receptor in extracts from oocytes injected with P2X 1 receptor cRNAs but not water-injected controls (Fig. 7A). These correspond to the glycosylated form of the P2X 1 receptor (23). Sulfo-NHS-LC-Biotin treatment was used to estimate the expression of P2X 1 receptors on the cell surface (Fig. 7B). The surface expression of Lys-68 mutant P2X 1 receptors demonstrates that the lack of functional response (K68R) or small functional response (K68A) to ATP application does not result from low levels of receptor expression and may result in part from changes in the kinetics of channel opening. In contrast, for the mutants at Arg-305, there was reduced (barely detectable for R305A) expression of biotinylated P2X 1 receptors compared with WT controls, and this appears to correlate with the amplitude of the functional response for these mutants (Table I). DISCUSSION Point mutations of conserved positively charged amino acids in the extracellular loop of the P2X 1 receptor have indicated that residues Lys-68, Lys-70, Lys-309, and Arg-292 contribute to the binding of ATP. Mutation of these residues also prolonged the time course of the P2X 1 receptor-mediated current, indicating that they may have some effect on the kinetics of channel opening. The comparative lack of effect on antagonism by suramin indicates that these mutations did not substantially modify the structure of the P2X 1 receptor.
The majority of mutants of the conserved positively charged residues (Lys-53, Lys-190, Lys-215, Lys-325, Arg-202, Arg-305, and Arg-314) either had no effect or only a small change in ATP potency, time course, and/or suramin sensitivity. As ATP is a negatively charged molecule, positive charge on the surface of the protein may act to attract the negatively charged phosphate group of ATP toward the binding site of the receptor. The small changes in potency could be accounted for by electrostatic effects resulting from changes in the distribution of surface charge on the receptor or from small changes in the conformation of the P2X 1 receptor. These results suggest that these lysine and arginine residues do not play an essential role in the formation of the ATP binding pocket of P2X receptors. Modest effects on ATP potency were recorded for mutations at Lys-70. The potency of ATP was reduced more for K70R with maintenance of positive charge than for K70A, where the charge was neutralized. These results suggest that in terms of determining ATP potency no charge at position 70 is better than the wrong positively charged amino acid. In addition, the mutation K70R resulted in an ϳ8-fold increase in the effectiveness of suramin as an antagonist (for K70A, there was no effect). The increase in suramin pK B estimate indicates that this region of the receptor also contributes to the binding of suramin. This is consistent with a previous study on P2X 4 receptors that demonstrated that residue 78 can contribute markedly to suramin activity (15).
Mutation of Arg-292 resulted in an ϳ100-fold decrease in ATP potency following substitution with either the conservative lysine mutation to maintain the positive charge or neutralizing the charge at this residue. This suggests that it is not just the charge but also the arginine residue that is important at this position. Previously arginine residues have been shown to contribute to ATP binding, e.g. at metabotropic P2Y 1 receptors (21) and the sarcoplasmic reticulum ATPase (28).
The most substantial decreases on ATP potency were recorded for mutations at residues Lys-68 and Lys-309. Previously lysine residues have been shown to be important in a variety of ATP-binding proteins including kinases (18), the cystic fibrosis transmembrane conductance regulator (29), the multidrug resistance protein (17), K ATP potassium channels (20), and P2Y receptors (21). It has been suggested that the lysine may interact directly with one of the phosphates of ATP (17,30). In the present study, the lack of effect on suramin indicates that there has not been a substantial change in the conformation of the mutant receptors, and these lysine residues are important for ATP binding but not suramin antagonism. At WT P2X 1 receptors, a fully saturated maximal response (presumably 100% receptor occupancy) is achieved at ϳ100 M ATP. For K68A and K309A mutant receptors, Ͻ10% of the maximal responses is recorded at this concentration of ATP. The shifts in the concentration-response relationships with these mutants are therefore most likely to correspond to a decrease in the affinity of ATP for the receptor and a concomitant decrease in ATP receptor occupancy at a given agonist concentration.
The kinetics of channel opening of P2X 1 receptors following ATP binding are also affected by some of the mutations in positive charge in the extracellular loop of the receptor. In particular for residues Lys-68, Lys-309, and Arg-292, which resulted in a substantial decrease in ATP potency, there is a prolongation of the current consistent with a slowing of the desensitization of the response. It is likely that P2X 1 receptor desensitization corresponds to the channel going into a ligandbound closed state similar to that described for nicotinic acetylcholine receptors (31), where agonist binding leads to change in the structure of the acetylcholine binding site (32). Mutations K68A, K309A/R, and R292K/A may thus result in a stabilized/prolonged open state of the P2X 1 receptor channel. The K68R mutant of the P2X 1 receptor is expressed at the membrane, although ionic currents were not recorded and it is possible that this may result from an inhibition of channel opening. The residues equivalent to Lys-68, Arg-292, and Lys-309 of the P2X 1 receptor are conserved throughout the P2X receptor family, including relatively non-desensitizing forms of the receptor. This suggests that desensitization results from their interaction with some other variant amino acid(s) to affect the time course of the P2X receptor currents. Previous studies have identified other regions of the P2X receptor that contribute to the time course of the response, including the second transmembrane domain (33), the intracellular carboxyl terminus (10 -12), and a protein kinase C motif in the intracellular amino terminus (13). Thus, the control of the time course of P2X receptor currents is a complicated multifactorial process.
The residues that we have identified to be involved in ATP binding (Lys-68, Lys-70, Arg-292, Lys-309) are in two clusters close to the predicted transmembrane domains (Fig. 8) and suggest that the ATP binding pocket may form as an extension of the ion channel pore vestibule. P2X receptors form from the multimeric assembly of at least three P2X receptor subunits (8). It is therefore possible that the ATP binding site could be formed from the interaction of residues within a P2X receptor subunit (Fig. 8A) and/or between adjacent subunits (Fig. 8B), as has been demonstrated for other ligand-gated ion channel families, e.g. ␥-aminobutyric acid A (34) and nicotinic acetylcholine receptors (27).
The present study has identified positively charged amino acid residues that are involved in ATP binding and activation of the P2X 1 receptor. The conservation of these residues throughout the family of P2X receptors suggests that they contribute to a common binding pocket for the phosphate group(s) of ATP. Given the range of pharmacological properties of P2X receptors, in addition, other regions of the receptor are also likely to contribute to ligand recognition. whereas Lys-292 and Lys-309 are in close proximity to the second transmembrane domain, suggesting that the ATP binding site might bridge these two locations and be situated at the entrance to the pore. If this is the case, then the interaction with ATP could either be intrasubunit as depicted in A or intersubunit as depicted in B.