Inflammatory mediators potentiate ATP-gated channels through the P2X(3) subunit.

The P2X(3) receptor is an ATP-gated ion channel predominantly expressed in nociceptive neurons from the dorsal root ganglion. P2X(3) receptor channels are highly expressed in sensory neurons and probably contribute to the sensation of pain. Kinetics of P2X(3) currents are characterized by rapid desensitization (<100 ms) and slow recovery (>20 s). Thus, any mechanism modulating rate of desensitization and/or recovery may have profound effect on susceptibility of nociceptive neurons expressing P2X(3) to ATP. Here we show that currents mediated by P2X(3) receptor channels and the heteromeric channel P2X(2/3) composed of P2X(2) and P2X(3) subunits are potentiated by the neuropeptides substance P and bradykinin, which are known to modulate pain perception. The effect is mediated by the respective neuropeptide receptors, can be mimicked by phorbol ester and blocked by inhibitors of protein kinases. Together with data from site-directed mutagenesis our results suggest that inflammatory mediators sensitize nociceptors through phosphorylation of P2X(3) and P2X(2/3) ion channels or associated proteins.

The P2X 3 receptor is an ATP-gated ion channel predominantly expressed in nociceptive neurons from the dorsal root ganglion. P2X 3 receptor channels are highly expressed in sensory neurons and probably contribute to the sensation of pain. Kinetics of P2X 3 currents are characterized by rapid desensitization (<100 ms) and slow recovery (>20 s). Thus, any mechanism modulating rate of desensitization and/or recovery may have profound effect on susceptibility of nociceptive neurons expressing P2X 3 to ATP. Here we show that currents mediated by P2X 3 receptor channels and the heteromeric channel P2X 2/3 composed of P2X 2 and P2X 3 subunits are potentiated by the neuropeptides substance P and bradykinin, which are known to modulate pain perception. The effect is mediated by the respective neuropeptide receptors, can be mimicked by phorbol ester and blocked by inhibitors of protein kinases. Together with data from site-directed mutagenesis our results suggest that inflammatory mediators sensitize nociceptors through phosphorylation of P2X 3 and P2X 2/3 ion channels or associated proteins.
P2X receptors mediate fast responses of excitable cells to application of ATP. So far, seven different P2X subunits have been cloned (P2X 1-7 ), which form a gene family (1). They have two transmembrane domains with intracellular termini and a rather large extracellular loop (2,3) and form non-selective cation channels with a high permeability to Ca 2ϩ . Various P2X receptors show differences both in the affinity to ATP and in the kinetics of activation and inactivation. P2X 1 and P2X 3 show high agonist affinity (EC 50 Ϸ 1 M) and rapid activation and desensitization with full activation in less than 10 ms and almost complete desensitization in less than 1 s (4, 5). Moreover, repeated application of the agonist leads to the disappearance of the ATP-gated currents (4,5). In contrast, P2X 2 and P2X 4 through P2X 7 have a low agonist affinity (EC 50 Ϸ 10 M), are slowly activating with time constants in the order of 100 ms, and show only partial desensitization (1,6). Heterooligomeric P2X 2/3 channels show a mixed phenotype with high af-finity to the agonist ␣,␤-methylene ATP (␣,␤-meATP) 1 like P2X 3 receptors and incomplete desensitization like P2X 2 receptors (5).
The desensitization properties of the rapidly activating and inactivating P2X 1 and P2X 3 are likely to be of physiological importance as synaptic transmission takes place in a few milliseconds. Modulation of the rate of desensitization and/or recovery of these receptors may contribute to synaptic efficacy. P2X 3 is almost exclusively expressed in sensory neurons (7,8), mostly in capsaicin-sensitive nociceptors (9 -11). P2X 3 -mediated currents in sensory neurons are large (12), and P2X 3 shows the strongest expression level in dorsal root ganglion (DRG) neurons compared with other P2X receptors (10). The P2X 3 protein is found on the sensory endings as well as on the presynaptic membrane in inner lamina II of the spinal horn (9), and its activation by ATP might contribute to the sensation of pain. In the dorsal horn a presynaptic mechanism is involved, which leads to a potentiation of the excitatory postsynaptic potential (13).
Here we show that the inflammatory mediators substance P (SP) and bradykinin (Bk) can potentiate currents through P2X 3 and P2X 2/3 expressed in Xenopus oocytes. Both induce an increase in peak current as well as in steady-state current with repeated application of ATP. Phorbol ester had a similar, nonadditive effect, suggesting a pathway involving protein kinase C. Mutagenesis experiments suggest that a conserved threonine at the intracellular N terminus, which had been shown to be phosphorylated in P2X 2 (19), is crucial for this effect. Together, our results suggest that inflammatory mediators potentiate ATP-gated currents in nociceptors through a mechanism involving either direct phosphorylation of P2X 3 receptors or phosphorylation of a so far unidentified protein.

EXPERIMENTAL PROCEDURES
cDNAs and Site-directed Mutagenesis-cDNAs coding for P2X 2 and P2X 3 are contained in the BamHI/NotI (P2X 2 ) or the EcoRI/XhoI (P2X 3 ) site of vector pCDNA3 (Invitrogen, Groningen, Netherlands). Splice variant P2X 2-2 has been described previously (20). cDNAs for SP and Bk receptors are contained in the HincII (SP receptor) or the EcoRI (B2 receptor) site of vector pBluescript. Chimeric molecules between P2X 2 * This work was supported in part by a grant from the "Graduiertenkolleg Zellbiologie in der Medizin" (to M. P.) and Grant FG 1-0-0 of the Attempto Research Group Program of the Universitä tsklinikum Tü bingen (to S. G.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
¶ Present address: Center for Hearing Sciences, Dept. of Otolaryngology/Head and Neck Surgery, Johns Hopkins University School of Medicine, Baltimore, MD 21205.
Xenopus laevis oocytes were surgically removed from adult animals and manually dissected. For the investigation of a possible effect of inflammatory mediators on currents through P2X receptors, we coexpressed the rat SP receptor (NK-1 receptor) or the human Bk (B2) receptor with the P2X receptors in oocytes. We injected either 5 ng of P2X 3 or 20 pg of P2X 2 , together with 5 ng of SP or Bk receptor; for expression of the P2X 2/3 heteromer, we co-expressed a mix of 0.5 ng of P2X 2 and 5 ng of P2X 3 with the neuropeptide receptors. cRNA was injected into stage V or VI Xenopus oocytes and oocytes kept in OR-2 medium (concentrations (in mM): 82.5 NaCl, 2.5 KCl, 1.0 Na 2 HPO 4 , 5.0 HEPES, 1.0 MgCl 2 , 1.0 CaCl 2 , and 0.5 g/liter polyvinylpyrrolidon, pH 7.3) at 19°C. One day after injection, oocytes were treated with collagenase type II (0.33 mg/ml; Sigma, Deisenhofen, Germany) for 40 min, and on the 2nd day the follicular layer was removed, as folliculated oocytes are known to have intrinsic P2 receptors (21). Defolliculated oocytes from every donor animal, which had been injected only with the cRNA encoding the SP receptor, were tested for endogenous ATPactivated currents.
P2X receptors showed the pharmacological and kinetic characteristics reported in previous studies using heterologous expression systems and as known from native tissue (Fig. 1). We used 100 nM ATP for activation of P2X 3 , 300 M for activation of P2X 2 and 10 M ␣,␤-meATP or 300 M ATP for activation of P2X 2/3 , in order to get sufficient current amplitude at low desensitization rate for all subunit combinations injected. To investigate intracellular regulation of ATP-gated currents, we repeatedly applied the agonist for 50 s with 50-s intervals. During the first four to eight ATP pulses, current peak amplitudes decreased; thereafter, they remained constant (Fig. 1). We took three equal current peak amplitudes as the base line for judging regulatory effects on current amplitude.
Data Analysis-Data on graphs indicate the median of the maximal increase of ATP-elicited peak current after stimulation by SP or Bk. Values were normalized to the ATP stimulation just before application of SP or Bk. Error bars indicate the range of the results as these were not symmetrically distributed. Moreover, we compared the amplitude at the end of the 50-s ATP application (quasi-steady state) and report this as the I 50 s .
Desensitization time constants were fitted to a single exponential function. They are reported as mean Ϯ standard deviation.
Statistical analysis was done with the unpaired Student's t test.

Currents through ATP Receptor Channels Containing the P2X 3 Subunit Are Potentiated by the Inflammatory Mediators
Substance P and Bradykinin-We used Xenopus oocytes as an expression system to investigate differential regulation of P2X receptor subtypes. SP or Bk application to oocytes expressing the SP or the Bk receptor, respectively, yields an immediate transient (Ϸ20 s) inward current (arrows in Figs. 2-4) resulting from activation of endogenous Ca 2ϩ -activated chloride channels. This represents a typical response of the oocyte membrane in response to Ca 2ϩ release from internal stores, providing evidence for the functional expression of the receptors and their potency to rise intracellular Ca 2ϩ levels.
Besides this transient current, SP and Bk had a regulatory effect on the amplitude of ATP-induced currents mediated by P2X 3 receptors. As shown in Fig. 2A, the application of SP (100 nM for 50 s) to oocytes expressing P2X 3 together with the SP receptor caused an immediate and transient (Ϸ150-s duration) increase in peak amplitude (median increase: 61%, range: 33-FIG. 1. Kinetic properties of P2X receptors expressed in Xenopus oocytes. Agonist application is indicated by horizontal bars. A, oocytes expressing P2X 2 channel receptors showed slowly and incompletely desensitizing currents after application of 300 M ATP. B, oocytes expressing P2X 3 channel receptors showed fast and almost completely desensitizing currents after application of 100 nM ATP. C, oocytes expressing P2X 2/3 channel receptors showed slowly and incompletely desensitizing currents after application of 10 M ␣,␤-meATP. Membrane potential was Ϫ70 mV. The SP receptor was always co-expressed.

FIG. 2.
Behavior of ATP-gated currents through P2X channels after activation of SP receptor or Bk receptor. P2X receptors were activated by a pulse protocol as shown in Fig. 1. Once ATP-gated currents showed reproducible peak currents, inflammatory mediators were applied. Arrowheads indicate Cl Ϫ currents due to the activation of endogenous Ca 2ϩ -activated Cl Ϫ channels. A, transient potentiation of currents through P2X 3 homomeric channels following application of SP (upper panel, 100 nM agonist) or Bk (lower panel; 100 nM agonist). B, P2X 2/3 heteromeric channels showed a similar potentiation by SP (upper panel) and Bk (lower panel). C, P2X 2 homomeric channels showed only very weak potentiation. D, application of SP to oocytes that had only been injected with cRNA coding for P2X 2 and P2X 3 receptors but not for SP receptors induced no current potentiation. E, columns indicate median increase of peak current amplitudes after application of SP (filled columns) or Bk (open columns) normalized to the last ATP stimulation before SP or Bk application. The error bars represent the range between which the results were distributed. The number, n, of independent measurements is indicated above each column.
165%, n ϭ 10; Fig. 2E) and steady state current (I 50 s , current amplitude at the end of the 50-s ATP pulse, median increase: 50%, range: 4 -129%, n ϭ 10) of ATP induced currents through P2X 3 receptors. After the potentiation phase, current amplitudes occasionally became smaller than before SP stimulation (not quantified). Short (2-5 s) applications of SP, repeated after about 10 min, induced each time current increase, suggesting that the underlying alteration of P2X receptors by SP was reversible (data not shown). The current increase following SP application could also be elicited using saturating ATP concentrations (300 M), demonstrating that SP does not potentiate ATP currents by changing P2X receptors' affinity to ATP.
The desensitization rate of the whole cell current can be well fitted with a single exponential function. The time constant of desensitization was not significantly different before and after treatment with SP (3.9 Ϯ 1.7 s compared to 4.6 Ϯ 2.3 s; p ϭ 0.44, n ϭ 10).
Using oocytes expressing the P2X 2/3 heteromer (obtained by coexpression of the P2X 2 and P2X 3 subunits) together with the SP receptor, application of SP yielded a similar transient and reversible increase of current (median increase of peak amplitude: 38%, range: 21-64%, n ϭ 6, Fig. 2 (B and E); median increase of I 50 s : 56%, range: 29 -68%, n ϭ 6). The effect was independent of the P2X receptor agonist; it could be elicited by 300 M ATP as well as by 10 M ␣␤-meATP. A similar effect was again seen with application of Bk when the Bk receptor was coexpressed (median increase of peak amplitude: 112%, range: 22-140%, n ϭ 9; Fig. 2 (B and E); median increase of I 50 s : 136%, range: 17-199%, n ϭ 9). The desensitization rate could not be well fitted but showed an apparent decrease after application of inflammatory mediators (see Fig. 2B).
Potentiation by Inflammatory Mediators Is due to Activation of a Protein Kinase-It is known that the desensitization of nicotinic acetylcholine receptors is directly accelerated by SP without the need of the SP receptor (22). In oocytes expressing only P2X 2/3 receptors but no SP receptor, the neuropeptide was not able to sensitize currents through P2X 2/3 (median increase of peak amplitude: Ϫ2%, range: Ϫ5% to 10%, n ϭ 6, Fig. 2 (D  and E); median increase of I 50 s : 6%, range: Ϫ3% to 17%, n ϭ 6).  Fig. 2. Arrowheads indicate Cl Ϫ currents due to the activation of endogenous Ca 2ϩ -activated Cl Ϫ channels. A, staurosporine blocked current increase through P2X 3 receptors. Oocytes had been incubated in extracellular solution containing 10 M staurosporine for 10 -30 min before the measurements were done. SP induced no longer any current increase. B, phorbol ester mimicked the effect of inflammatory mediators. Application of 4␤-PMA (10 nM) for 50 s yielded current potentiation through P2X 3 receptors very similar to that evoked by SP or Bk. C, if SP was applied during 4␤-PMA application, SP elicited no further stimulation. The strong rundown of ATP-gated currents was only observed with long application of PMA. D, columns indicate median increase of peak current amplitudes after application of effector substances normalized to the last ATP stimulation before effector application. The error bars represent the range between which the results were distributed. The number, n, of independent measurements is indicated above each column. Fig. 2. Arrowheads indicate Cl Ϫ currents due to the activation of endogenous Ca 2ϩ -activated Cl Ϫ channels. A, a P2X 2 receptor whose C terminus had been replaced by the P2X 3 C terminus gained potentiation by SP. B, a P2X 3 receptor whose C terminus had been replaced by the P2X 2 C terminus lost potentiation by SP. This mutant has been expressed as a heteromer together with P2X 2 . C, the P2X 2-2 splice variant that lacks 69 amino acids within the C terminus gained potentiation by SP. D, the C-terminally truncated P2X 2 ⌬401 receptor gained potentiation by SP. The respective P2X receptors are schematically shown at the top of each trace. E, columns indicate median increase of peak current amplitudes after application of SP (100 nM) normalized to the last ATP stimulation before SP application. The error bars represent the range between which the results were distributed. The number, n, of independent measurements is indicated above each column.

FIG. 4. Effect of the cytoplasmic C terminus of P2X receptors on potentiation of currents. Protocols were as described in legend to
This shows that the effect is mediated by the SP receptor and suggests pathways downstream of the receptor to be responsible for sensitization of P2X receptors.
Since SP and Bk are supposed to act via the phospholipase C/protein kinase C pathway, we tested different pharmacological agents interacting with kinases for their effect on sensitization of P2X receptors. Oocytes expressing P2X 3 and the SP receptor were incubated with the serine/threonine-kinase inhibitor staurosporine (10 M) for 10 to 30 min. After this incubation, application of SP no longer increased ATP-activated peak or steady state currents (median increase in peak amplitude: Ϫ6%, range: Ϫ16% to 9%, n ϭ 10, Fig. 3 (A and D); median increase in I 50 s : 3%, range: Ϫ12% to 16%, n ϭ 7), suggesting the involvement of protein kinase in the potentiating process. In addition, the effect of SP could be mimicked using the phorbol ester 4␤-PMA (10 nM for 50 s; median increase of peak current: 66%, range: 7-337%, n ϭ 22; Fig. 3 (B  and D); median increase of I 50 s : 94%, range: 17-265%, n ϭ 22) but not using the biologically inactive stereoisomer 4␣-PMA (data not shown). Moreover, as shown in Fig. 3C, during PMA application SP was unable to further sensitize the current (median increase of peak current: 4%, range: Ϫ6% to 10%, n ϭ 4; Fig. 3 (C and D); median increase of I 50 s : 8%, range: Ϫ9% to 42%, n ϭ 4), suggesting that the same mechanism underlies both effects. Together, these experiments suggest that protein kinase C mediates the stimulation of ATP-gated currents through P2X 3 and P2X 2/3 .

The Intracellular C Terminus of P2X Receptors Controls Potentiation by Inflammatory
Mediators-Intracellular signaling cascades most likely act on intracellular parts of the P2X receptors. To identify these regions, we made use of the differential effect of SP on P2X 2 and P2X 3 designing chimeras between P2X 2 and P2X 3 . Exchanging the intracellular C termini yielded active channels. The P2X 2 receptor with P2X 3 C terminus (P2X 2 -C-X 3 ) formed rather slowly desensitizing ion channels, which were, however, faster than P2X 2 wild type channels (mean of desensitization rates Ϯ S.D. 16.7 Ϯ 0.7 s, n ϭ 6, in comparison to 95.6 Ϯ 32.4 s with P2X 2 wild type, n ϭ 8). As shown in Fig. 4A, the P2X 2 -C-X 3 receptor gained potentiation by SP. It showed an increase in the steady state current that lasted longer than the effect on P2X 3 but was still transient (median increase of peak current: 62%, range: 48 -92%, n ϭ 10, Fig. 4 (A and E); median increase of I 50 s : 71%, range: 48 -101%, n ϭ 10). P2X 3 receptors with P2X 2 C terminus (P2X 3 -C-X 2 ) showed faster desensitization than the P2X 3 receptor and high agonist affinity as the P2X 3 receptor. Because of the fast desensitization of the P2X 3 -C-X 2 chimera, we investigated this chimera as heteromer with P2X 2 . The P2X 3 -C-X 2 /P2X 2 receptor activated either with ␣␤-meATP or with ATP lost SP regulation (median increase of peak current: Ϫ7%, range: Ϫ21% to Ϫ3%, n ϭ 4; Fig. 4 (B and E)), implicating the cytoplasmic C terminus as an important region for regulation by SP.
To further investigate the role of the P2X 2 C terminus we tested the P2X 2-2 receptor, a P2X 2 splice variant in which 69 amino acids of the cytoplasmic C terminus are missing (20). Fig. 4C shows that this splice variant was also sensitized by SP (median increase of peak current: 36%, range: 18 -49%, n ϭ 4, Fig. 4E; median increase of I 50 s : 33%, range: 23-50%, n ϭ 4). Thus, the P2X 3 C terminus is not necessary for SP modulation. Rather, it seems that the wild type P2X 2 C terminus inhibits the modulation by SP and that this inhibition is relieved in the splice variant. This hypothesis is further supported by a gainof-regulation with a P2X 2 receptor truncated at position 401 (P2X 2 ⌬401) deleting 72 of the about 116 cytoplasmic amino acids at the C terminus. This mutant showed kinetics similar to that of the P2X 2 wild type but gained SP regulation (median increase of peak current: 21%, range: 16 -23%, n ϭ 3, Fig. 4 (D and E); median increase of I 50 s : 21%, range: 16 -24%, n ϭ 3).
N-terminal chimeras were also functional. A P2X 2 receptor with P2X 3 N terminus (P2X 2 -N-X 3 ) was not regulated by SP, whereas a P2X 3 receptor with P2X 2 N terminus (P2X 3 -N-X 2 ) was regulated by SP (data not shown), confirming that the C terminus controls regulation of P2X receptors by SP.
A Conserved N-terminal Consensus Site Is the Most Likely Target for Phosphorylation-Regulation by SP might be due to direct phosphorylation of the channel protein. We, therefore, constructed a series of point mutations, either single or in combination, of cytoplasmic serine or threonine residues in the P2X 3 receptor. These mutations include all intracellular serines or threonines that are in a consensus sequence for phosphorylation by protein kinase A or C (R/K(X 0 -3 )S/ T(X 0 -3 )R/K). The effect of these mutations is summarized in Fig. 5. Combined mutation of all serines or threonines that are in a consensus sequence for phosphorylation at the cytoplasmic C terminus (P2X 3 T364A/T365A/T369A/S371A/T382A/S387A/ T388A) leads to a functional, SP-regulated channel, rendering it rather unlikely that direct phosphorylation of the C terminus mediates the regulation. Moreover, there is no C-terminal consensus site conserved between P2X 3 and P2X 2-2 .
Single or combined amino acid substitutions of the serines or threonines at the N terminus lead in all cases to functional, regulated channels with the exception of P2X 3 T12A, which was not measurable. This threonine residue is conserved in all known P2X receptors, contained in a conserved consensus sequence for phosphorylation (TXK/R) and has recently been shown to be phosphorylated in the P2X 2 receptor (19). Mutation of lysine 14 in P2X 3 contained in the consensus sequence (P2X 3 K14Q) similarly leads to not measurable channels, whereas mutation of threonine 13 leads to functional, regulated channels (Fig. 5), identifying the consensus motif TXK as crucial for normal channel function. A T12E mutant mimicking constitutive phosphorylation was also not measurable, demonstrating that it is not just the presence of a negative charge at this position, which is important. Mutation of the corresponding threonine in either the P2X 2 -C-X 3 or the P2X 3 -N-X 2 chimera, or the P2X 2-2 splice variant leads always to not measurable channels, rendering it impossible to directly assess functional consequences of this mutation. P2X 2-2 possesses, apart from this completely conserved threonine, only one other serine/threonine at its cytoplasmic N terminus. Mutation of FIG. 5. Site-directed mutagenesis of putative PKC-phosphorylation sites within the intracellular N and C termini of the P2X 3 receptor. Left, scheme of the intracellular N and C termini of the P2X 3 receptor. All serine/threonine residues are enlarged. The residues that were mutated are highlighted in bold. Residues that had been mutated in combination are marked by lines. Positively charged amino acids (Arg, Lys) are indicated by a ϩ sign. Right, table summarizing the phenotype of the mutants. Left column, list of mutants that were generated. Right column, results of tests for regulation by SP using a measurement protocol as described in Fig. 2. this second serine (serine 11) leads to functional, regulated channels.
Together, these results suggest that the completely conserved threonine at the cytoplasmic N terminus of P2X receptors is the best candidate for an amino acid directly phosphorylated after stimulation by SP. Alternatively, protein kinases might phosphorylate an unrelated protein, which controls activity of P2X receptors. DISCUSSION Both P2X 3 and P2X 2/3 receptors are expressed in sensory neurons, and there is accumulating evidence that they have a specific role in nociception (17,(23)(24)(25)(26). Moreover, activation of P2X 3 leads to a much stronger nociceptive effect in inflamed compared with normal tissue (27,28) and the P2X 2/3 heteromer might mediate mechanical allodynia (11). Thus, it seems that sensitization of nociceptors leads to a bigger response to ATP. Our study addressed the underlying mechanism by reconstituting the relevant signaling cascades in Xenopus oocytes.
Sensitization of nociceptors is due to the release of inflammatory mediators such as the neuropeptides SP (29) and Bk (30). SP is expressed in small diameter afferent neurons (31,32) and is released upon peripheral nociceptive stimulation in the periphery as well as in the superficial dorsal horn (33). It binds with high affinity to the tachykinin receptor (NK-1), which is expressed in neurons of the spinal cord (34,35). There are conflicting data, however, on expression of the SP (NK-1) receptor in primary afferent neurons from the DRG (34, 36 -38). Although the receptor has not been identified by immunocytochemistry (34) SP can elicit an inward current in DRG neurons, which is mediated by a non-selective cation channel (37). Moreover, it had been shown that application of SP to sensory neurons can potentiate currents, which are gated by ATP (39). This suggests a modulation of P2X receptors by SP also in sensory cells. So far, neither the P2X subunit involved nor the intracellular signaling cascade was known.
The main receptor for Bk in primary afferent neurons is the B2 receptor, which couples similar to the SP receptor to a signaling pathway using G q and phospholipase C and leading to a rise in intracellular Ca 2ϩ .
Our study demonstrates that the inflammatory mediators SP and Bk can potentiate currents through P2X 3 and P2X 2/3 receptors in Xenopus oocytes.
Heteromeric P2X 2/3 receptor channels showed a decrease of the desensitization rate after treatment with inflammatory mediators, which could explain the potentiation. As the potentiation was only transient, we were not able to measure the time constant for recovery. Rate of desensitization of P2X 3 was not significantly changed, but the fast desensitization rate of this channel may not be reliably determined in whole oocytes. We, therefore, propose that an effect on desensitization rate is the underlying mechanism of current potentiation.
Evidence that protein kinase C mediates potentiation of P2X receptors in oocytes is severalfold. (i) Both the SP and the Bk receptor are coupled to phospholipase C activation; (ii) phorbol ester can mimic the effect; (iii) stimulation by inflammatory mediators and phorbol ester is not additive; and, finally, (iv) the serine/threonine kinase inhibitor staurosporine blocks potentiation of P2X receptors.
We could not unequivocally identify a phosphorylation site on P2X 3 receptors responsible for the modulation. Our chimeras between P2X 2 and P2X 3 implicated the cytoplasmic C terminus as crucial. The results from site-directed-mutagenesis suggest, however, that there is no phosphorylation site on the cytoplasmic C terminus of P2X 3 implicated in the effect. (i) Mutation of all the serine/threonine residues in any consensus site at the same time (7 out of a total of 12 serine/threonine) yielded functional, regulated channels; (ii) there is no serine or threonine in any consensus site conserved between P2X 2-2 and P2X 3 ; and (iii) there is only one serine/threonine that is contained in both P2X 2-2 and P2X 2 ⌬401. Still, it might be that phosphorylation sites are not contained in a classic consensus site and are not conserved between P2X 2 and P2X 3 .
Mutants at the N terminus were either functional and regulated or not measurable. The non-measurable mutants all concerned a highly conserved consensus site for phosphorylation by PKC (TXR/K). Interestingly, Séguéla and colleagues (19) recently demonstrated phosphorylation of P2X 2 on this threonine. Their results suggested that the receptor is constitutively phosphorylated and that this phosphorylation leads to the slow desensitization rate of P2X 2 . Accelerated desensitization of the already rapidly desensitizing P2X 3 receptor would explain that we were not able to measure the corresponding P2X 3 T12A mutant. These findings strongly support phosphorylation of this site also in P2X 3 and speak in favor of a decreased desensitization rate as the underlying mechanism of current potentiation. This would imply that the C terminus interacts with the N terminus to control phosphorylation. In this model the C terminus of P2X 2 would stabilize phosphorylation at the N-terminal TXK site leading to constitutive phosphorylation, whereas the C terminus of P2X 3 and the splice variant P2X 2-2 would destabilize phosphorylation, allowing phosphorylation to be regulated. This would also imply that splicing at the C terminus of P2X 2 would be a means to control modulation of receptor activity by phosphorylation at the N terminus. Interaction of cytoplasmic termini has already been shown for other structurally related channels (40). As we cannot directly prove the implication of the N-terminal phosphorylation site in the regulation of P2X receptor activity, we still must consider, however, the possibility that a different, so far unidentified protein may be phosphorylated and control activity of P2X receptors.
Hu et al. (39) reported that application of 100 nM SP leads to a transient potentiation by 127.2 Ϯ 6.7% of ATP-gated currents in sensory neurons. This potentiation was blocked by the protein kinase inhibitor H7 and the SP (NK1) receptor antagonist spantide, resembling our own results and suggesting that the underlying mechanism is active in sensory neurons.
For Bk it is well established that it acts directly on primary sensory neurons, and several studies have suggested that PKC mediates this effect (41)(42)(43). Moreover, Bk sensitizes heatactivated currents in isolated nociceptors, and this sensitization can be mimicked by phorbol ester and blocked by staurosporine (44). Our results now suggest that another target for PKC in nociceptors may be P2X 3 and that Bk may sensitize different pain-related ion channels using similar mechanisms. At the same time, we show that different inflammatory mediators may converge on the same ion channel.
Together, we show that modulation of P2X 3 and P2X 2/3 activity by the inflammatory mediators SP and Bk may account for sensitization of nociceptors to the action of ATP.