点击显示 收起
【关键词】 Sulfonic Acid Groups
We have previously identified the suramin analog 4,4',4'',4'''-(carbonylbis(imino-5,1,3-benzenetriylbis(carbonylimino)))tetrakis-benzene-1,3-disulfonic acid (NF449) as a low nanomolar potency antagonist of recombinant P2X1 receptors. Here, we characterize, by two-electrode voltage-clamp electrophysiology, three isomeric suramin analogs designated para-4,4',4'',4''''-(carbonylbis(imino-5,1,3-benzenetriylbis (carbonylimino)))tetrakis-benzenesulfonic acid (NF110), meta-(3,3',3'',3''''-(carbonylbis(imino-5,1,3-benzenetriylbis (carbonylimino)))tetra-kis-benzenesulfonic acid (NF448), and ortho-(2,2',2'',2''''-(carbonylbis(imino-5,1,3-benzenetriylbis (carbonylimino)))tetra-kis-benzenesulfonic acid (MK3) with respect to their potency in antagonizing rat P2X receptor-mediated inward currents in Xenopus laevis oocytes. Meta, para, and ortho refer to the position of the single sulfonic acid group relative to the amide bond linking the four symmetrically oriented benzenesulfonic acid moieties to the central, invariant suramin core. NF448, NF110, and MK3 were >200-fold less potent in blocking P2X1 receptors than NF449, from which they differ structurally only by having one instead of two sulfonic acid residues per benzene ring. Although the meta- and ortho-isomers retained P2X1 receptor selectivity, the para-isomer NF110 exhibited a significantly increased activity at P2X3 receptors (Ki 36 nM) and displayed the following unique selectivity profile among suramin derivatives: P2X2+3 = P2X3 > P2X1 > P2X2 >> P2X4 > P2X7. The usefulness of NF110 as a P2X3 receptor antagonist in native tissues could be demonstrated by showing that NF110 blocks -methylene-ATP-induced currents in rat dorsal root ganglia neurons with similar potency as recombinant rat P2X3 receptors. Together, these data highlight the importance of both the number and exact location of negatively charged groups for P2X subtype potency and selectivity.P2X receptors constitute an abundant class of cell surface receptors that open an intrinsic pore that becomes permeable to small cations when exposed to extracellular ATP and related nucleotides (North, 2002). The seven known subunit isoforms, designated P2X1-P2X7, assemble into homotrimeric and heterotrimeric receptor channels (Nicke et al., 1998; Aschrafi et al., 2004; Mio et al., 2005), resulting in large receptor subtype diversity.
Among the numerous sensory processes in which P2X receptors have been implicated, their involvement in nociception has attracted particular attention (North, 2004). Pain-sensing sensory neurons of trigeminal, nodose, and dorsal root ganglia express at least six of the seven P2X subunit isoforms (Collo et al., 1996). Interest has been centered on two members, P2X2 and P2X3, which seem to account for the two major types of responses of sensory neurons to ATP or -methylene-ATP (,-meATP): 1) a rapidly desensitizing inward current typical for homomeric P2X3 receptors (Rae et al., 1998), and 2) a slowly or nondesensitizing current typical for heteromeric P2X2+3 receptors assembled from P2X3 and P2X2 subunits (Lewis et al., 1995). Because P2X3 subunits are almost exclusively confined to sensory neurons, P2X3 subunit-containing receptors represent a promising target for the development of novel analgesic drugs (Chizh and Illes, 2001; Kennedy, 2005; Burnstock, 2006).
Converging evidence for an important role of the P2X3 subunit in nociception comes from gene knockout and pharmacological studies. Sensory neurons of P2X3 knockout mice lacked ATP-inducible desensitizing inward currents and showed reduced pain-related behaviors (Cockayne et al., 2000; Souslova et al., 2000). In addition, post-transcriptional silencing of P2X3 gene expression by intrathecally applied antisense oligonucleotides or short interfering RNA in rodents reduced pain intensity (Barclay et al., 2002; Honore et al., 2002; Dorn et al., 2004). Recently, the nucleotide analog 2',3'-O-(2,4,6-trinitrophenyl)-ATP (TNP-ATP), a high-affinity antagonist of P2X1, P2X3, and P2X2+3 receptors (Virginio et al., 1998; Burgard et al., 2000), and A-317491, a high-affinity non-nucleotide antagonist of P2X3 and P2X2+3 receptors, have been found to produce antinociceptive effects in various pain models in rodents (Honore et al., 2002; Jarvis et al., 2002; McGaraughty et al., 2003).
Antinociceptive responses were also observed in rodents upon administration of the nonselective P2 antagonist suramin (Driessen et al., 1994; Honore et al., 2002). We have shown the potential of suramin as a lead structure for developing highly potent P2X1 receptor-selective antagonists (Rettinger et al., 2000, 2005; Lambrecht et al., 2002; Horner et al., 2005). Here, we report about the isomeric suramin analogs NF110, NF448, and MK3, which differ from the potent P2X1 receptor-selective antagonist NF449 only in the number (one instead of two) of sulfonic acid groups present in para-, meta-, or ortho-position at each of the four terminal benzene rings (Figs. 1 and 5, A and C). Reducing the number of sulfonic acid groups from eight to four greatly reduced the potency at P2X1 receptors, yet it did not change selectivity per se, because the meta- and ortho-isomers NF448 and MK3, respectively, still behaved as a distinctly P2X1 receptor selective antagonists. In contrast, the potency of the para-isomer, NF110, at P2X3 receptors was greatly enhanced, whereas the potency at P2X1 receptors remained unaffected, resulting in an overall selectivity change in favor of P2X3 receptors. These data demonstrate that both the number and precise position of negatively charged substituents provide a structural basis for altering the selectivity and potency of antagonists at P2X receptor subtypes.
Fig. 1. Chemical structure of NF110. NF110 is designated as a tetravalent suramin derivative, because four aromatic ring systems (benzene rings in the case of NF110) carrying sulfonic acid residues are linked by amide bonds to the central, symmetrically phenyl-substituted urea moiety, which represents the invariant part of suramin. Bivalent suramin analogs such as NF023 and NF279 contain only two amide-bonded aromatic ring systems. Arrows indicate the positions where additional sulfonic acid groups are located on the otherwise structurally identical, previously characterized suramin analog NF449. NF449 blocks P2X1 receptors with subnanomolar potency, but it is a comparatively weak inhibitor of P2X3 receptors (Rettinger et al., 2005). Please note that the sulfonic acid groups of NF110 are in para-position relative to the amide bond.
Fig. 5. Antagonist potency of NF448 and MK3 at P2X1, P2X2, and P2X3 receptors. A and C, chemical structures of NF448 and MK3, which contain the sulfonic acid groups in meta- and ortho-position, respectively, in contrast to the para-isomer NF110 (Fig. 1). B, concentration-inhibition curves of NF448 at wild-type rP2X1, rP2X2, and rP2X3 receptors. , rP2X1 homomer; , rP2X2 homomer; and , rP2X3 homomer. D, concentration-inhibition curves of MK3 at wild-type rP2X1, rP2X2, and rP2X3 receptors. , rP2X1 homomer; , rP2X2 homomer; and , rP2X3 homomer.
Materials. ATP and ,-meATP were obtained from Sigma-Aldrich (Taufkirchen, Germany). The suramin derivatives NF110, NF448, and MK3 (for chemical structures, see Figs. 1 and 5, A and C) were synthesized as tetrasodium salts (Kassack et al., 2004). Enzymes were obtained from New England Biolabs (Frankfurt, Germany), Epicenter (Madison, WI), Stratagene (La Jolla, CA), or Promega (Heidelberg, Germany). Desoxyoligonucleotides were from QIAGEN GmbH (Hilden, Germany). All other chemicals were purchased from either Sigma-Aldrich or Merck (Darmstadt, Germany).
P2X Receptor Expression in Xenopus laevis Oocytes. Plasmids encoding wild-type rat P2X subunits (rP2X1, GenBank accession no. X80477; rP2X2, GenBank accession no. U14414; rP2X3, GenBank accession no. X90651; rP2X4, GenBank accession no. X87763; and rP2X7, GenBank accession no. X95882) and a rP2X2-X1 chimera comprising the first 47 N-terminal amino acids of the rP2X2 subunit joined in frame with amino acids 48Val-399Ser of the 399-amino acid-long rP2X1 subunit were available from previous studies (Nicke et al., 1998; Aschrafi et al., 2004).
An rP2X2-X3 chimera was generated by joining, in-frame, the first 47 N-terminal codons of the rP2X2 subunit (1Met-47Tyr) with codons 42Val-397His of the 397-amino acid-long rP2X3 subunit. To this end, SnaB I sites were introduced by site-directed mutagenesis, allowing for blunt-ended cleavage just before corresponding 48Val and 42Val codons of the rP2X2-pNKS2 and rP2X3-pNKS2 plasmids, respectively. The N-terminal rP2X2 sequence was excised with HindIII and SnaBI and ligated in frame between the HindIII site of the vector and the inserted SnaBI cleavage site of rP2X3-pNKS2, thus yielding rP2X2-X3-pNKS2. The construct was verified by sequencing using an ABI Prism 310 genetic analyzer (Applied Biosystems, Foster City, CA).
Capped cRNAs were synthesized from linearized plasmids and injected at 0.5 µg/µl in 23- or 50-nl aliquots into follicle cell-free X. laevis oocytes using a Nanoliter 2000 injector (WPI, Sarasota, FL). For expression of the heteromeric rP2X2+3 combination, a 1:1 (w/w) ratio of rP2X2 and rP2X3 cRNA was injected. Oocytes were maintained at 19°C in sterile oocyte Ringer's (ORi) solution (90 mM NaCl, 1 mM KCl, 1 mM CaCl2, 1 mM MgCl2, and 10 mM HEPES, pH 7.4) supplemented with 50 µg/ml gentamicin.
Two-Electrode Voltage-Clamp Electrophysiology. Agonist-evoked current responses were recorded 1 to 3 days after cRNA injection at ambient temperature (21-24°C) with a conventional two-electrode voltage-clamp (Turbo TEC-05 amplifier; npi Electronics, Tamm, Germany) at a holding potential of -60 mV as described previously (Rettinger et al., 2005). Oocytes were continuously perfused by gravity flow (5-10 ml/min) in a small flow-through chamber (volume 10 µl) with a nominally calcium-free ORi solution (designated Mg-ORi), in which CaCl2 was replaced by equimolar MgCl2 to avoid a contribution of endogenous Ca2+-dependent Cl- channels to the ATP response. rP2X7 receptor-dependent currents were recorded in a bathing solution consisting of 94 mM NaCl, 1 mM KCl, 0.1 mM flufenamic acid, and 10 mH HEPES-NaOH, pH 7.4 (Hulsmann et al., 2003).
Dilutions of agonists and antagonists in Mg-ORi were prepared daily and applied by bath perfusion. Switching between bath solutions was controlled by a set of magnetic valves, enabling computer-controlled applications of compounds in a rigorous schedule by using the CellWorks Lite 5.1 software (npi Electronics, Tamm, Germany). For analyzing the antagonist-mediated inhibition of P2X receptor currents, the receptors were activated with agonist concentrations close to the respective EC50 value (i.e., 10 µM ATP for rP2X2 and rP2X4 receptors, 1 µM ATP for rP2X1 and rP2X3 receptors) (Rettinger et al., 2000), 10 nM ATP for P2X2-X1 and P2X2-X3 chimeras, 100 µM ATP for the rP2X7 receptor, and 1 µM ,-meATP for the rP2X2+3 heteromeric receptor. How agonists and antagonists were applied was dependent on whether desensitizing or nondesensitizing P2X receptors were under study (see below).
Concentration-Inhibition Analysis of Fast Desensitizing P2X Currents. For the analysis of the activation and inhibition of the rapidly desensitizing wt-rP2X1 and wt-rP2X3 receptors, receptors were first repetitively activated in 1- or 2-min intervals with an EC50 concentration of ATP until constant current responses were obtained. After this pre-equilibration period and complete ATP washout, oocytes were perfused for 30 s with the desired concentration of antagonist alone, followed by a 5-s-long coapplication of 1 µM ATP and antagonist. Between the applications of incrementally larger concentrations of antagonist, ATP control responses were recorded to monitor rundown or run-up artifacts and to check for complete washout of the antagonist. Thus, each current amplitude (peak current) recorded in the presence of antagonist could be compared for reference with the two flanking control responses to ATP alone. Further details are given in the figure legends.
Concentration-Inhibition Analysis of Nondesensitizing P2X Currents. Oocytes expressing nondesensitizing P2X receptors (defined as a current decrease of <5% over 30 s in the presence of agonist) were superfused with agonist-containing bath solution until a steady-state current was obtained. Then, the superfusion solution was switched to one containing both antagonist and the same concentration of agonist as before. The superfusion was maintained until a steady-state level of inhibition was reached, followed by switching to bath solution alone for drug washout. The percentage of the control response was calculated as the steady-state currents in the absence and presence of antagonist.
Whole-Cell Patch-Clamp Recordings in Dorsal Root Ganglia Neurons. The isolation and culturing of thoracic and lumbar DRGs from 1-day-old male Wistar rats has been described previously (Gerevich et al., 2005). Membrane currents were recorded at a holding a potential of -70 mV by whole-cell configuration patch clamping using an Axopatch 200B amplifier (Axon Instruments). Patch pipettes (3-5 M) were filled with a solution containing 135 mM CsCl, 2 mM MgCl2, 11 mM EGTA, 1 mM CaCl2, 1.5 mM Mg-ATP, and 0.3 mM Li-GTP, and 20 mM HEPES-CsOH, pH 7.3. The external recording solution consisted of 140 mM NaCl, 5 mM KCl, 2 mM MgCl2, 2 mM CaCl2, 11 mM glucose, and 10 mM HEPES-NaOH, pH 7.4. To block the firing of cells during agonist application, 2 mM lidocaine N-ethyl bromide was included in the pipette solutions. A pressure-operated, computer-controlled, rapid drug application device (DAD-12; Adams and List, Westbury, NY) was used for drug administration. To induce P2X3 receptor-mediated currents, 10 µM ,-meATP was applied onto single DRG cells for 1 s in 5-min intervals. The responses to ,-meATP were constant (Gerevich et al., 2005). After recording of two ,-meATP-induced currents of approximately the same amplitude (control currents), NF110 was administered in increasing concentrations (0.03-30 µM) to the same cell. For each NF110 concentration, a 5-min preincubation with NF110 was followed by a 1-s test pulse of 10 µM ,-meATP in the continued presence of NF110 at the same concentration as during preincubation. To estimate the inhibitory effect of NF110, the ,-meATP-induced current amplitude after the incubation with a certain NF110 concentration was expressed as percentage ± S.E.M. of the control current.
Data Analysis. Data were plotted and fitted using Prism 4 (GraphPad Software Inc., San Diego, CA) and Origin 6.0 (OriginLab Corp., Northampton, MA). Agonist concentration-response curves were produced by measuring the maximal current induced by increasing concentrations of agonist, which were then normalized to the maximal current value obtained with a saturating ATP concentration. Concentration-response curves and half-maximal effective agonist concentrations (EC50) were obtained from curves by iteratively fitting the Hill equation (eq. 1) to the normalized data points pooled from several (n) oocytes:
(1)
where I is the current evoked by agonist concentration A, Imax the maximal current response, and nH the Hill coefficient.
Concentration-inhibition curves and IC50 values were derived from nonlinear least-squares fits of eq. 2 to the pooled data points:
(2)
where Imax is the current response in absence of antagonist (Ant), IAnt is the current response at the respective antagonist concentration, and IC50 the antagonist concentration causing 50% inhibition of the current elicited by a given agonist concentration.
Where indicated, IC50 values were transformed into inhibition constants (Ki values) by using the classic Cheng-Prusoff equation (Cheng and Prusoff, 1973):
(3)
which is warranted by the classification of NF110 as a competitive antagonist (see Results).
Results are presented as means ± S.E.M. from n experiments. Error bars were omitted in the figures when they were smaller than the symbols used.
Target Specificity of NF110. The pharmacological specificity of NF110 at a fixed concentration of 10 µM was assessed for a total of 33 cell surface receptors, ion channels, transporters, and enzymes by testing in the commercially available diversity profile at Cerep (Celle l'Evescault, France) (http://www.cerep.fr). In addition, the ability of NF110 to block recombinant human P2Y1, P2Y2, and P2Y11 receptors was evaluated by Ca2+ imaging using Oregon Green 448 as a fluorescent indicator exactly as described previously (Ullmann et al., 2005).
Inhibition of Fast-Desensitizing rP2X1 and rP2X3 Receptor-Mediated Responses by NF110. In functional experiments of native P2X receptors, NF110 was found to antagonize ,-meATP-induced contractions of the longitudinal smooth muscle of guinea pig ileum more potently than contractions of the rat vas deferens (Kassack et al., 2004). Because this is indicative of a preferential inhibition of (guinea pig) P2X3 over (rat) P2X1 receptors, we have characterized NF110 at rat P2X receptors recombinantly expressed in X. laevis oocytes.
For determining concentration-inhibition relationships of NF110 at the fast desensitizing rP2X1 and rP2X3 receptors, the peak current protocol was used (see Materials and Methods). Current traces representative for the application of this protocol to rP2X3 receptor-expressing oocytes are illustrated in Fig. 2A. To assess the time required for the development of the inhibitory effect of NF110, rP2X3 receptor-expressing oocytes were preincubated in pilot experiments for 5 to 60 s with 100 nM NF110 before a test pulse of ATP was applied. Inhibition plateaued after a 15-s preincubation period as shown in Fig. 2B. Therefore, for all subsequent concentration-inhibition measurements, oocytes expressing rP2X1 or rP2X3 receptors were preincubated for exactly 30 s at the desired NF110 concentration and then challenged with 1 µM ATP plus the same NF110 concentration used for preincubation. Fitting the Hill equation to the pooled data yielded IC50 values of 201 ± 42 and 90 ± 18 nM for the rP2X1 and rP2X3 receptor, respectively. The corresponding concentration-inhibition curves are included in Fig. 2C.
Fig. 2. Antagonist potency of NF110 at P2X1, P2X2, P2X2+3, and P2X3 receptors. Light and dark gray areas indicate the duration of application of agonist and NF110, respectively. A, original traces of rP2X3 receptor-mediated currents illustrating the assessment of the inhibitory potency of NF110 from peak current measurements. rP2X3 receptor-expressing oocytes were repetitively activated by 1 µM ATP in 1-min intervals. Between two control measurements, oocytes were pre-equilibrated with the indicated concentration of NF110 for 30 s and then challenged with 1 µM ATP in the continued presence of NF110. The extent of inhibition was judged from the reduction of the peak current amplitude compared with the two flanking control measurements. B, effect of the duration of NF110 preapplication on the inhibition of the fast desensitizing homomeric rP2X3 receptor responses. The bar graph shows the averaged inhibition of current amplitudes elicited by 1 µM ATP after the indicated preincubation times in the presence of 100 nM NF110 as referred to current amplitudes evoked in the absence of NF110 (for details, see A; n = 5). C, concentration-inhibition curves. Agonist-induced currents were recorded in the presence of the indicated concentrations of NF110 according to the procedures shown in A and D. , rP2X3 homomer; , rP2X2+3 heteromer; , rP2X1 homomer; and , rP2X2 homomer. In this and all other figures, continuous lines through symbols are fits to the Hill equation. For convenience, all Ki values of NF110 are summarized in Table 1. D, original traces of rP2X2+3 receptor-mediated currents illustrating the assessment of the inhibitory potency of NF110 from stationary current measurements. After a stationary current was reached in the presence of 1 µM ,-meATP, coapplication of NF110 was started (indicated by dark gray area). The current declined until a lower steady state was reached, indicating that inhibition by NF110 was fully developed. Then, application of ,-meATP and NF110 was terminated. The percentage of the control response was calculated as the ratio of the stationary current in the absence and presence of NF110.
Antagonist Potencies of NF110 at rP2X2, rP2X2+3, rP2X4, and rP2X7 Receptors. The above-mentioned results identified NF110 as the first suramin derivative that is more potent in inhibiting recombinant rP2X3 than rP2X1 receptors. Because the P2X3 subunit exists in vivo not only as homomeric P2X3 receptors but also as functional heteromultimeric assemblies with the P2X2 subunit (designated P2X2+3 receptors) (Lewis et al., 1995), the characterization of NF110 was extended to also include these receptor subtypes. rP2X2+3 receptors incorporate features of both P2X2 and P2X3 receptors by exhibiting a peculiar nondesensitizing, yet ,-meATP-sensitive phenotype. Therefore, the protocol for nondesensitizing receptors could be applied for rP2X2 as well as rP2X2+3 receptor subtypes. Activation of the homomeric rP2X2 receptor by 10 µM ATP in the presence of increasing concentrations of NF110 yielded an IC50 value of 6.2 ± 0.9 µM (Fig. 2C).
TABLE 1 Potencies of suramin analogs at recombinant P2X receptors The Cheng-Prusoff equation (eq. 3) was used to calculate Ki values from the IC50 values determined in this study for NF110, NF448, MK3 (Figs. 2, 4, and 5) and the published IC50 values for NF449 (Rettinger et al., 2005). Also used in the calculation were the following EC50 values obtained for the various P2X receptor subtypes in this laboratory: 0.68 µM ATP (rP2X1), 19.9 µM ATP (rP2X2), 0.67 µM ATP (rP2X3), 17.6 µM ATP (rP2X2-X3), and 1.7 µM ,-meATP (rP2X2+3). A competitive antagonism is supported by a Schild plot analysis (Fig. 4E). n indicates the number of independent experiments.
Coexpression of rP2X2 and rP2X3 subunits in oocytes results in a mixed population of heterogeneous receptors; these include both the homomeric rP2X2 and rP2X3 receptors as well as the heteromeric rP2X2+3 receptors (Liu et al., 2001). To exclude a contribution of homomeric rP2X2 receptors to the current response, 1 µM ,-meATP was used as an agonist, which activates heteromeric rP2X2+3 receptors with markedly higher potency than homomeric rP2X2 receptors. Control experiments with oocytes injected solely with cRNA for the rP2X2 subunit showed that >30 µM ,-meATP was needed to evoke significant inward currents in these cells (results not shown). Original current traces in oocytes coexpressing rP2X2 and rP2X3 subunits are shown in Fig. 2D. ,-meATP (1 µM) elicited a largely monophasic response consisting of a nondesensitizing current, which in general obscured any rapidly deactivating component attributable to homomeric rP2X3 receptors. Coapplication of NF110 decreased the current amplitude. The extent of current inhibition was quantified as the percentage of current remaining at the end of the NF110-agonist coapplication relative to the current amplitude evoked by agonist alone. Fitting eq. 2 to the pooled data yielded an NF110 IC50 value of 55 ± 8.8 nM. The corresponding concentration-inhibition curve is included in Fig. 2C.
Compared with the other rP2X subtypes examined, NF110 was significantly less potent in inhibiting rP2X4 receptors (<45% inhibition at 100 µM NF110; Fig. 3A) and rP2X7 receptors (<5% inhibition at 30 µM NF110; Fig. 3B) irrespective of whether the oocytes were preincubated with NF110 for 15-30 s or not. In regard of the limited availability of NF110, we did not attempt to obtain complete concentration-inhibition curves for these NF110-insensitive receptor subtypes.
Fig. 3. Antagonist potency of NF110 at P2X4 and P2X7 receptors. Light and dark gray areas indicate the duration of application of agonist and NF110, respectively. A, inhibition by NF110 of inward currents elicited by 10 µM ATP from rP2X4 receptor-expressing oocytes was assessed by applying the peak current protocol. B, typical current trace from rP2X7 receptor-expressing oocytes challenged with 100 µM free ATP. Application of 30 µM NF110 in the continued presence of ATP resulted in a slight reduction of the linearly increasing current component and a slight increase to initial values, when the NF110 application was discontinued.
Mechanism of P2X3 Receptor Antagonism by NF110. Because P2X1 and P2X3 receptors desensitize rapidly after stimulation with ATP, a binding equilibrium between ATP and a simultaneously applied antagonist cannot be reached during the transient current response (Rettinger et al., 2000). Therefore, the mechanism of inhibition by antagonists cannot be unraveled reliably by the use of nonsteady-state measurements (Burgard et al., 2000; Rettinger et al., 2000; Rettinger and Schmalzing, 2004). In an attempt to overcome this problem, we exploited a nondesensitizing rP2X2-X3 receptor chimera containing the N-terminal part (amino acids 1-47) of the rP2X2 subunit and the complementary C-terminal part of the rP2X3 subunit (amino acids 42-397). The junction points for joining the N- and C-terminal portions of the P2X2 and P2X3 subunit, Val48 and Val42, represent conserved residues in transmembrane domain 1 of both subunits. Because amino acids 1-47 comprise only intracellular portions of the rP2X2 subunit and also because ATP interacts exclusively with the ectodomain (Ennion et al., 2000; Jiang et al., 2000), the ATP binding properties of the chimera can be expected to be identical to that of the wild-type P2X3 receptor. In contrast to the rapidly desensitizing current mediated by the wild-type rP2X3 receptor (Fig. 4A), the rP2X2-X3 chimera gave rise to stationary, nondesensitizing currents at -60 mV in standard recording solution as long as superfusion with ATP was maintained (Fig. 4B). Concentration-response curves revealed that the rP2X2-X3 chimera is half-maximally activated by 17.6 ± 1.0 nM ATP or 251 ± 34 nM ,-meATP (Fig. 4C). These values correspond to an 40-fold higher ATP potency compared with that of the wild-type rP2X3 receptors (667 ± 143 nM; Fig. 4C). To account for these observations, we assume that elimination of desensitization as realized by the rP2X2-X3 chimera unmasks nanomolar ATP potency as an inherent feature of the rP2X3 receptor.
Fig. 4. Assessing the mechanism of NF110-induced inhibition of rP2X3 receptors by exploiting a nondesensitizing rP2X2-X3 chimera. A and B, traces elicited by 10 µM ATP (denoted by gray areas) in oocytes expressing the wild-type rP2X3 receptor or the rP2X2-X3 chimera. Note that the chimera deactivated very slowly upon ATP washout. Schematics illustrate the membrane topology of P2X receptors and the swapped domains originating from the rP2X2 subunit (gray) and the rP2X3 subunit (black). C, agonist concentration-response curves for the rP2X2-X3 chimera (, ATP: EC50 = 17.6 ± 1.0 nM and nH = 2.2 ± 0.1, n = 7; , ,-meATP: EC50 = 251 ± 34 nM and nH = 1.3 ± 0.1, n = 7) and the rP2X3 receptor (, ATP: EC50 = 667 ± 143 nM and nH = 1.0 ± 0.1, n = 7). Currents were normalized to those elicited by saturating agonist concentrations. D, concentration-inhibition curves. Agonist-induced stationary currents were recorded in the presence of the indicated concentrations of NF110 according to the protocol shown in Fig. 2D. , rP2X2-X3 chimera; , rP2X2-X1 chimera. E, NF110 is a competitive antagonist at the rP2X2-X3 chimera. ATP concentration-response curves for the rP2X2-X3 chimera in the absence () and presence of NF110 (, 100 nM; , 300 nM; and , 1000 nM). Inset, Schild plot derived from this data set. The solid line is a linear regression with a slope of 1.14 ± 0.15 (not significantly different from unity; p > 0.05). Constraining the slope to unity (dashed line) yielded a pA2 value of 7.0.
Next, we examined whether NF110 inhibits nondesensitizing P2X chimeras with the same potency as that observed for the rapidly desensitizing parental receptors. To this end, we also included a previously characterized nondesensitizing rP2X2-X1 chimera (Rettinger and Schmalzing, 2004). The chimeras were first activated by superperfusion with ATP at a concentration around their EC50 values (10 nM) until a steady-state inward current was achieved in the sustained presence of ATP. Then, the oocytes were superfused with a solution containing the same concentration of ATP plus NF110. As a result of the receptor channel blockade, the current declined until a steady-state level of inhibition was established. To generate concentration-inhibition curves, this procedure was repeated for an ascending series of NF110 concentrations, followed by washout of NF110 and ATP. The extent of NF110-induced inhibition of the rP2X2-X1 and rP2X2-X3 chimeras was quantified as the percentage of stationary current remaining in the presence of NF110 relative to the stationary current amplitude elicited by ATP in the absence of NF110. A fit of the Hill equation to the pooled data yielded IC50 values for NF110 of 124 ± 32 and 55 ± 12.3 nM at the rP2X2-X1 and rP2X2-X3 receptor chimera, respectively (Fig. 4D). Together, these data indicate that NF110 inhibits rP2X2-X1 and rP2X2-X3 chimeras with similar potencies as the corresponding wild-type rP2X1 and rP2X3 receptors, respectively.
Next, ATP concentration-response curves were generated in the absence and presence of NF110 at the rP2X2-X3 chimera. Increasing NF110 concentration produced a parallel rightward shift of the ATP concentration-response curve without changing the maximal response (Fig. 4E). Schild analysis (Fig. 4E, inset) yielded a regression slope of 1.14 ± 0.15, which is not significantly different from unity. On constraining the slope to unity, a pA2 value of 7.0 was obtained. Together, these data are strongly supportive of a competitive antagonism of NF110 on P2X3 receptors.
Effect of the Positional NF110 Isomers NF448 and MK3, on P2X1, P2X2, and P2X3 Receptor-Mediated Currents.The structure of NF110 differs from that of the P2X1-selective antagonist NF449 solely by the presence of only four (instead of eight) sulfonic acid groups, highlighting the importance of negatively charged groups in P2X subtype selectivity. To examine the importance of the position of the sulfonic acid groups for P2X receptor subtype selectivity, the positional NF110 isomers NF448 and MK3 were exploited. Although NF448 and MK3 differ from NF110 solely by carrying the four sulfonic acid groups in meta- or ortho- instead of para-position (Fig. 5, A and C), respectively, their selectivity profiles were markedly different: NF448 and MK3 most potently inhibited rP2X1 receptor-mediated currents, displaying IC50 values of 181 ± 45 nM (n = 6) and 65 ± 14 nM (n = 5), respectively. Inhibition of P2X2 or P2X3 receptor-mediated currents occurred only at significantly higher concentrations of NF448 and MK3 (Fig. 5, B and D). Thus, the exact position of the sulfonic acid groups is of crucial importance for P2X subtype selectivity of these suramin derivatives.
Potency of NF110 at Endogenous P2X Receptors in Rat DRG Neurons. To examine whether NF110 is capable of inhibiting native P2X3 receptors of nociceptive neurons from the rat with similar potency as oocyte-expressed recombinant rP2X3 receptors, DRG neurons of 1-day-old rats were used. P2X3 receptors are found selectively in a subpopulation of DRG neurons and are thought to contribute to the processing of pain information. ,-meATP applied in a submaximally effective concentration of 10 µM every 5 min onto cultured small-diameter (20- to 35-µm) DRG neurons evoked rapidly desensitizing currents with a peak amplitude of 675 ± 106 pA (n = 16). Sensitivity to ,-meATP and fast desensitization identifies these currents as P2X3 receptor-mediated (Rae et al., 1998). NF110 concentration-dependently inhibited the ,-meATP-evoked currents with an IC50 value of 527 ± 304 nM (n = 5-12) (Fig. 6). A Ki value of 48 nM was calculated using the Cheng-Prusoff equation and an EC50 value for ,-meATP of 1 µM at P2X3 receptors (Khakh et al., 2001).
Fig. 6. Antagonist potency of NF110 at endogenous P2X3 receptors in DRG neurons. A, patch-clamp recordings from a neuron that was activated for 1 s in 5-min intervals with 10 µM ,-meATP (thin bars) in the absence and presence of ascending NF110 concentrations (gray rectangles). B, NF110 concentration-inhibition curve. Data points are based on peak current measurements as outlined in A. IC50 = 527 ± 304 nM and nH = 0.92 ± 0.16; n 5.
Activity of NF110 toward a Panel of Drug Targets, Including P2Y Receptors. NF110 was assessed by a commercial profiling service (Cerep) for its interference with a total of 33 non-P2X receptor drug target molecules. NF110 was considered inactive (IC50 > 10 µM) toward all targets tested. NF110 was also inactive (IC50 > 10 µM) toward P2Y1, P2Y2, and P2Y11 receptors stably expressed in astrocytoma cells.
Selectivity Profile of NF110 versus Other P2X Receptor Antagonists. We have shown by two-electrode voltage-clamp electrophysiology that the tetravalent suramin analog NF110 (Fig. 1) inhibits P2X receptor subtypes recombinantly expressed in X. laevis oocytes; these receptors can be ranked by their potencies, based on Ki values (Table 1): rP2X2+3 = rP2X3 > rP2X1 > rP2X2 >> rP2X4 > rP2X7. Homomeric rP2X3 receptors are blocked by NF110 with virtually the same potency as heteromeric rP2X2+3 receptors. The P2X3 selectivity of NF110 over P2X1 receptors (2-fold) is lower than that of the Abbott compound A-317491 (115-fold) (Jarvis et al., 2002) and more reminiscent to that of TNP-ATP, which also blocks P2X1, P2X3, and P2X2+3 receptors with similar potency close to 1 nM (Virginio et al., 1998).
Suramin Derivatives Are Competitive Inhibitors of P2X Receptors. We have previously shown for the desensitizing rP2X1 receptor that the P2X1 receptor-selective suramin derivative NF279 causes not only a rightward shift of the ATP concentration-response curves but also a depression of the maximal current response to ATP, when oocytes are pre-equilibrated with high concentrations of NF279 (Rettinger et al., 2000). Although insurmountable antagonism may be indicative of noncompetitive antagonism, a similar inhibitory pattern, called a hemi-equilibrium, can also occur if a short-acting agonist is combined with a slowly dissociating competitive antagonist. In terms of fast-desensitizing P2X receptors, the short duration of the current transient leaves little time for a slowly unbinding antagonist to equilibrate with the fast binding of ATP. The rapidly binding but slowly unbinding TNP-ATP represents a pertinent example of this problem. Based on maximum current depression, TNP-ATP was first interpreted to be a noncompetitive antagonist of the fast-desensitizing recombinant rP2X3 receptor (Virginio et al., 1998). However, re-examination at nondesen-sitizing rP2X2+3 receptors revealed a competitive mechanism of inhibition for TNP-ATP (Burgard et al., 2000) that also applies to homomeric rP2X3 receptors (Neelands et al., 2003).
To resolve the mechanism of antagonism by NF110, without the bias imposed by rapid desensitization, we exploited the nondesensitizing rP2X2-X3 chimera that allowed us to determine current inhibition under favorable steady-state conditions. In this construct, desensitization is virtually fully eliminated by substitution of the cytoplasmic N-terminal tail and the first membrane domain corresponding to the first 41 amino acids of the P2X3 subunit, respectively, with a complementary part of the rP2X2 subunit. The pivotal feature of this chimera as a surrogate for the wild-type P2X3 receptor is that the comparably small rP2X2 portion is not extracellularly accessible. Therefore, the binding properties of the rP2X2-X3 chimera described here and the previously published rP2X2-X1 chimera (Werner et al., 1996; Rettinger and Schmalzing, 2004; Rettinger et al., 2005), at least for polar ligands, are determined by the receptor subunit providing the ectodomain. This view is strongly supported by several observations: 1) like the wild-type P2X3 receptor, the rP2X2-X3 chimera was activated with high potency by ,-meATP, which is a low potency agonist for rP2X2 receptors; 2) the suramin analog NF110 blocked the chimera and the corresponding wild-type rP2X3 receptor with virtually the same IC50 value, whereas more than 100-fold higher NF110 concentrations were required for blocking the rP2X2 receptor. Essentially the same conclusions were reached in a study, in which similar nondesensitizing human chimeric P2X receptor constructs were used for assessing the antagonistic mechanisms of A-317491 and TNP-ATP (Neelands et al., 2003).
The data obtained with our rP2X2-X3 chimeric receptors indicate that NF110 meets the criteria for a competitive antagonist. First, increasing NF110 concentrations produced a parallel rightward shift of the ATP concentration-response curve without changing the maximal response. Second, Schild analysis yielded a slope close to unity. The pA2 value of 7.0 for the P2X2-X3 chimera is in reasonable agreement with the estimated Ki value of 36 nM (pKi = 7.4) at wild-type P2X3 receptors (Table 1). A competitive mechanism of inhibition by suramin analogs is consistent with previous findings showing that the bivalent suramin analog NF279 acts as a competitive blocker of a nondesensitizing rP2X2-X1 chimera (Rettinger and Schmalzing, 2004).
Rat P2X3 and P2X1 Receptors Share Nanomolar ATP Potency That Is Masked by Desensitization. Strikingly, both the rP2X2-X1 and the rP2X2-X3 chimera are half-maximally activated by 200- or 40-fold lower ATP concentrations than the wild-type rP2X1 and rP2X3 receptors, respectively. Nanomolar potency for ATP is also a feature of the wild-type rP2X1 receptor, but it is obscured by desensitization (Rettinger and Schmalzing, 2004). The micromolar ATP potency deduced under nonsteady-state conditions from peak current measurements represents no more than an amalgam of activation and desensitization. Present data indicate that the rP2X3 receptor (EC50 17 nM) shares nanomolar ATP potency with the rP2X1 receptor (EC50 3 nM) (Rettinger and Schmalzing, 2004). This is not only apparent from the concentration-response curves at the two chimeras but also from the extremely slow current deactivation after agonist washout (Fig. 4B), reflecting slow agonist unbinding (Rettinger and Schmalzing, 2004). The remarkable ATP sensitivity of rP2X3 receptors has already been noted in previous experiments (He et al., 2002). Interestingly, human counterparts of these chimeras lack this high ATP sensitivity and deactivate rapidly after agonist washout (Neelands et al., 2003). Sequence variations between human and rat P2X3 receptors are likely to account for this phenomenon.
The Role of Sulfonic Acid Groups for P2X Receptor Subtype Selectivity and Potency. Using suramin as a lead, we have previously shown that the potency for the P2X1 receptor, in particular, can be increased tremendously by appropriate structural modifications. The most potent P2X1 receptor-selective antagonist identified so far, the tetravalent suramin derivative NF449, exhibits IC50 values of 0.05 and 0.3 nM at recombinant human and rat P2X1 receptors, respectively; NF449 is thus >3000-fold more potent than the bivalent parent compound suramin (Lambrecht et al., 2002; Hulsmann et al., 2003; Rettinger et al., 2005). In contrast, 6000-fold higher NF449 concentrations were needed to block rP2X3 receptors (Rettinger et al., 2005). To allow a direct comparison with NF110, Ki values of NF449 at recombinant rP2X1, rP2X2, and rP2X3 receptors were calculated from published IC50 values and included in Table 1.
Sulfonic acid groups influence the antagonistic properties of the suramin derivatives in two aspects. 1) Comparison of the structure and antagonistic potency of NF449 with NF110, NF448, or MK3 reveals that a higher number of sulfonic acid groups has a favorable effect on the potency to block P2X1 receptors. Structurally, NF110 and its isomers differ from NF449 solely by containing only four instead of eight sulfonic acid residues (Figs. 1 and 5), yet these compounds are >200-fold less potent than NF449 in inhibiting the rat P2X1 receptor. A larger number of possible electrostatic interactions of NF449 may account for this effect. Mutagenesis studies implicate that conserved positively charged amino acids close to the external entry of the channel pore (K68, R292, and K309 of the rP2X1 receptor) contribute to ATP binding by attracting the negative charges of the phosphate groups (Ennion et al., 2000; Jiang et al., 2000; Egan et al., 2004). It is likely that these and other positively charged residues play also a role in the binding of the suramin derivatives (Jacobson et al., 2004). 2) The exact position of the sulfonic acid groups seems to be of crucial importance for P2X subtype selectivity. Changing the position of the single sulfonic acid group from ortho or meta to para as realized in NF110 led to a marked increase in P2X3 receptor potency without grossly affecting P2X1 receptor potency.
The observed competitive antagonism combined with the crucial role of negatively charged groups in achieving high potency interactions suggest that the extracellular ATP binding pocket on P2X receptors accommodates at least part of the suramin molecule. At least some of the sulfonic acid groups seem to interact electrostatically with the basic amino acids that are physiologically involved in attracting the phosphate groups of ATP. Together, the present study re-emphasizes the value of suramin as a lead compound for the development of P2X receptor subtype-selective antagonists. The observation that the number and the position of sulfonic acid groups profoundly influence subtype selectivity may be helpful for structural optimization of drugs targeting P2X receptors.
Acknowledgements
We thank Ursula Braam for expert technical assistance in molecular biology experiments.
ABBREVIATIONS: ,-meATP, -methylene-adenosine 5'-triphosphate; TNP-ATP, 2',3'-O-(2,4,6-trinitrophenyl)-adenosine 5'-triphosphate; A-317491, 5-([(3-phenoxybenzyl)[(1S)-1,2,3,4-tetrahydro-1-naphthalenyl]amino]carbonyl)-1,2,4-benzenetricarboxylic acid); NF110, 4,4',4'',4'''-(carbonylbis(imino-5,1,3-benzenetriylbis (carbonylimino)))tetra-kis-benzenesulfonic acid; NF448, 3,3',3'',3'''-(carbonylbis(imino-5,1,3-benzenetriylbis (carbonylimino)))tetra-kis-benzenesulfonic acid; NF449, 4,4',4'',4'''-(carbonylbis(imino-5,1,3-benzenetriylbis(carbonylimino)))tetrakis-benzene-1,3-disulfonic acid; MK3, 2,2',2'',2'''-(carbonylbis(imino-5,1,3-benzenetriylbis (carbonylimino)))tetra-kis-benzenesulfonic acid; NF279, 8,8'-(carbonylbis(imino-4,1-phenylenecarbonylimino-4,1-phenylenecarbonylimino))bis(1,3,5-naphthalenetrisulfonic acid); ORi, sterile oocyte Ringer's; wt, wild-type; DRG, dorsal root ganglia.
【参考文献】
Aschrafi A, Sadtler S, Niculescu C, Rettinger J, and Schmalzing G (2004) Trimeric architecture of homomeric P2X2 and heteromeric P2X1+2 receptor subtypes. J Mol Biol 342: 333-343.
Barclay J, Patel S, Dorn G, Wotherspoon G, Moffatt S, Eunson L, Abdel'al S, Natt F, Hall J, Winter J, et al. (2002) Functional downregulation of P2X3 receptor subunit in rat sensory neurons reveals a significant role in chronic neuropathic and inflammatory pain. J Neurosci 22: 8139-8147.[Abstract/Free Full Text]
Burgard EC, Niforatos W, van Biesen T, Lynch KJ, Kage KL, Touma E, Kowaluk EA, and Jarvis MF (2000) Competitive antagonism of recombinant P2X2/3 receptors by 2',3'-O-(2,4,6-trinitrophenyl) adenosine 5'-triphosphate (TNP-ATP). Mol Pharmacol 58: 1502-1510.
Burnstock G (2006) Purinergic P2 receptors as targets for novel analgesics. Pharmacol Ther, in press.
Cheng Y and Prusoff WH (1973) Relationship between the inhibition constant (Ki) and the concentration of inhibitor which causes 50 per cent inhibition (I50) of an enzymatic reaction. Biochem Pharmacol 22: 3099-3108.
Chizh BA and Illes P (2001) P2X receptors and nociception. Pharmacol Rev 53: 553-568.[Abstract/Free Full Text]
Cockayne DA, Hamilton SG, Zhu QM, Dunn PM, Zhong Y, Novakovic S, Malmberg AB, Cain G, Berson A, Kassotakis L, et al. (2000) Urinary bladder hyporeflexia and reduced pain-related behaviour in P2X3-deficient mice. Nature (Lond) 407: 1011-1015.
Collo G, North RA, Kawashima E, Merlo-Pich E, Neidhart S, Surprenant A, and Buell G (1996) Cloning of P2X5 and P2X6 receptors and the distribution and properties of an extended family of ATP-gated ion channels. J Neurosci 16: 2495-2507.[Abstract/Free Full Text]
Dorn G, Patel S, Wotherspoon G, Hemmings-Mieszczak M, Barclay J, Natt FJ, Martin P, Bevan S, Fox A, Ganju P, et al. (2004) siRNA relieves chronic neuropathic pain. Nucleic Acids Res 32: e49.[Abstract/Free Full Text]
Driessen B, Reimann W, Selve N, Friderichs E, and Bültmann R (1994) Antinociceptive effect of intrathecally administered P2 purinoceptor antagonists in rats. Brain Res 666: 182-188.
Egan TM, Cox JA, and Voigt MM (2004) Molecular structure of P2X receptors. Curr Top Med Chem 4: 821-829.
Ennion S, Hagan S, and Evans RJ (2000) The role of positively charged amino acids in ATP recognition by human P2X1 receptors. J Biol Chem 275: 29361-29367.[Abstract/Free Full Text]
Gerevich Z, Muller C, and Illes P (2005) Metabotropic P2Y(1) receptors inhibit P2X(3) receptor-channels in rat dorsal root ganglion neurons. Eur J Pharmacol 521: 34-38.
He ML, Koshimizu TA, Tomic M, and Stojilkovic SS (2002) Purinergic P2X2 receptor desensitization depends on coupling between ectodomain and C-terminal domain. Mol Pharmacol 62: 1187-1197.[Abstract/Free Full Text]
Honore P, Mikusa J, Bianchi B, McDonald H, Cartmell J, Faltynek C, and Jarvis MF (2002) TNP-ATP, a potent P2X3 receptor antagonist, blocks acetic acid-induced abdominal constriction in mice: comparison with reference analgesics. Pain 96: 99-105.
Horner S, Menke K, Hildebrandt C, Kassack MU, Nickel P, Ullmann H, Mahaut-Smith MP, and Lambrecht G (2005) The novel suramin analogue NF864 selectively blocks P2X1 receptors in human platelets with potency in the low nanomolar range. Naunyn-Schmiedeberg's Arch Pharmacol 372: 1-13.
Hulsmann M, Nickel P, Kassack M, Schmalzing G, Lambrecht G, and Markwardt F (2003) NF449, a novel picomolar potency antagonist at human P2X1 receptors. Eur J Pharmacol 470: 1-7.
Jacobson KA, Costanzi S, Ohno M, Joshi BV, Besada P, Xu B, and Tchilibon S (2004) Molecular recognition at purine and pyrimidine nucleotide (P2) receptors. Curr Top Med Chem 4: 805-819.
Jarvis MF, Burgard EC, McGaraughty S, Honore P, Lynch K, Brennan TJ, Subieta A, van Biesen T, Cartmell J, Bianchi B, et al. (2002) A-317491, a novel potent and selective non-nucleotide antagonist of P2X3 and P2X2/3 receptors, reduces chronic inflammatory and neuropathic pain in the rat. Proc Natl Acad Sci USA 99: 17179-17184.[Abstract/Free Full Text]
Jiang LH, Rassendren F, Surprenant A, and North RA (2000) Identification of amino acid residues contributing to the ATP-binding site of a purinergic P2X receptor. J Biol Chem 275: 34190-34196.[Abstract/Free Full Text]
Kassack MU, Braun K, Ganso M, Ullmann H, Nickel P, Boing B, Muller G, and Lambrecht G (2004) Structure-activity relationships of analogues of NF449 confirm NF449 as the most potent and selective known P2X1 receptor antagonist. Eur J Med Chem 39: 345-357.
Kennedy C (2005) P2X receptors: targets for novel analgesics? Neuroscientist 11: 345-356.
Khakh BS, Burnstock G, Kennedy C, King BF, North RA, Seguela P, Voigt M, and Humphrey PP (2001) International union of pharmacology. XXIV. Current status of the nomenclature and properties of P2X receptors and their subunits. Pharmacol Rev 53: 107-118.[Abstract/Free Full Text]
Lambrecht G, Braun K, Damer M, Ganso M, Hildebrandt C, Ullmann H, Kassack MU, and Nickel P (2002) Structure-activity relationships of suramin and pyridoxal-5'-phosphate derivatives as P2 receptor antagonists. Curr Pharm Des 8: 2371-2399.
Lewis C, Neidhart S, Holy C, North RA, Buell G, and Surprenant A (1995) Coexpression of P2X2 and P2X3 receptor subunits can account for ATP-gated currents in sensory neurons. Nature (Lond) 377: 432-435.
Liu M, King BF, Dunn PM, Rong W, Townsend-Nicholson A, and Burnstock G (2001) Coexpression of P2X3 and P2X2 receptor subunits in varying amounts generates heterogeneous populations of P2X receptors that evoke a spectrum of agonist responses comparable with that seen in sensory neurons. J Pharmacol Exp Ther 296: 1043-1050.[Abstract/Free Full Text]
McGaraughty S, Wismer CT, Zhu CZ, Mikusa J, Honore P, Chu KL, Lee CH, Faltynek CR, and Jarvis MF (2003) Effects of A-317491, a novel and selective P2X3/P2X2/3 receptor antagonist, on neuropathic, inflammatory and chemogenic nociception following intrathecal and intraplantar administration. Br J Pharmacol 140: 1381-1388.
Mio K, Kubo Y, Ogura T, Yamamoto T, and Sato C (2005) Visualization of the trimeric P2X2 receptor with a crown-capped extracellular domain. Biochem Biophys Res Commun 337: 998-1005.
Neelands TR, Burgard EC, Uchic ME, McDonald HA, Niforatos W, Faltynek CR, Lynch KJ, and Jarvis MF (2003) 2',3'-O-(2,4,6,Trinitrophenyl)-ATP and A-317491 are competitive antagonists at a slowly desensitizing chimeric human P2X3 receptor. Br J Pharmacol 140: 202-210.
Nicke A, B?umert HG, Rettinger J, Eichele A, Lambrecht G, Mutschler E, and Schmalzing G (1998) P2X1 and P2X3 receptors form stable trimers: a novel structural motif of ligand-gated ion channels. EMBO (Eur Mol Biol Organ) J 17: 3016-3028.
North RA (2002) Molecular physiology of P2X receptors. Physiol Rev 82: 1013-1067.[Abstract/Free Full Text]
North RA (2004) P2X3 receptors and peripheral pain mechanisms. J Physiol (Lond) 554: 301-308.[Abstract/Free Full Text]
Rae MG, Rowan EG, and Kennedy C (1998) Pharmacological properties of P2X3-receptors present in neurones of the rat dorsal root ganglia. Br J Pharmacol 124: 176-180.
Rettinger J, Braun K, Hochmann H, Kassack MU, Ullmann H, Nickel P, Schmalzing G, and Lambrecht G (2005) Profiling at recombinant homomeric and heteromeric rat P2X receptors identifies the suramin analogue NF449 as a highly potent P2X1 receptor antagonist. Neuropharmacology 48: 461-468.
Rettinger J and Schmalzing G (2004) Desensitization masks nanomolar potency of ATP at the P2X1 receptor. J Biol Chem 279: 6426-6433.[Abstract/Free Full Text]
Rettinger J, Schmalzing G, Damer S, Müller G, Nickel P, and Lambrecht G (2000) The suramin analogue NF279 is a novel and potent antagonist selective for the P2X1 receptor. Neuropharmacology 39: 2044-2053.
Souslova V, Cesare P, Ding Y, Akopian AN, Stanfa L, Suzuki R, Carpenter K, Dickenson A, Boyce S, Hill R, et al. (2000) Warm-coding deficits and aberrant inflammatory pain in mice lacking P2X3 receptors. Nature (Lond) 407: 1015-1017.
Ullmann H, Meis S, Hongwiset D, Marzian C, Wiese M, Nickel P, Communi D, Boeynaems JM, Wolf C, Hausmann R, et al. (2005) Synthesis and structure-activity relationships of suramin-derived P2Y11 receptor antagonists with nanomolar potency. J Med Chem 48: 7040-7048.
Virginio C, Robertson G, Surprenant A, and North RA (1998) Trinitrophenyl-substituted nucleotides are potent antagonists selective for P2X1, P2X3 and heteromeric P2X2/3 receptors. Mol Pharmacol 53: 969-973.[Abstract/Free Full Text]
Werner P, Seward EP, Buell GN, and North RA (1996) Domains of P2X receptors involved in desensitization. Proc Natl Acad Sci USA 93: 15485-15490.[Abstract/Free Full Text]
作者单位：Molecular Pharmacology, RWTH Aachen University, Aachen, Germany (R.H., G.S.); Max-Planck-Institute of Biophysics, Frankfurt am Main, Germany (J.R.); Rudolf-Boehm Institute for Pharmacology and Toxicology, University of Leipzig, Leipzig, Germany (Z.G., P.I.); Department of Pharmaceutical Chemistry, U