Background
ATP plays a prominent role in nociception. Its application onto human skin elicits pain [
1,
2]. Injection of ATP into the rat hindpaw reduces paw withdrawal latencies and produces flinching and writhing behaviors [
3‐
6]. Recent studies of purinergic receptors in primary sensory dorsal root ganglion (DRG) neurons demonstrate that ATP gives rise to nociception by activating P2X receptors in primary sensory DRG neurons [
7‐
9].
In situ hybridization assay indicates that P2X2-P2X6 mRNAs are present in DRG neurons [
10]. P2X2 and P2X3 receptors are the major receptor types selectively expressed in peripheral and central terminals and the somata of DRG neurons [
11‐
14]. These neurons are small (diameter d < 25 μm) and medium (25 < d < 40 μm) in size, bind isolectin B4 and express vanilloid TRPV1 receptors [
15‐
18]. The importance of P2X3-containing receptors in nociception is further confirmed by the findings that nociceptive behaviors become greatly diminished in P2X3 knock-out mice [
19,
20] and in animals treated with P2X3 antisense oligonuclotides [
21], small interfering RNA (siRNA) [
22] or the specific P2X3 antagonist, A-37491 [
23]. Electrophysiological studies indicate that ATP produces large inward currents by activating P2X3 homomeric and P2X2/3 heteromeric receptors, thus evoking depolarization in small and medium DRG neurons [
11,
13,
14,
24,
25]. In the spinal dorsal horn, ATP, released from the central terminals of DRG neurons, acts on presynaptic P2X receptors to promote AMPA receptor-mediated synaptic transmission in nociceptive pathways [
26‐
30]
An important characteristic of P2X3 receptor-mediated responses is its sensitivity to tissue and nerve injury. Nociceptive behaviors produced by ATP become greatly enhanced after inflammation [
24] and nerve injury [
31‐
33]. An increase in P2X3 receptor-mediated responses is one of the major reasons for the enhancement [
11,
31]. We found that the ATP currents in DRG neurons isolated from rats with inflammation or nerve injury are 2–3 fold larger [
11,
31]. The mechanisms responsible for the increase include upregulation of P2X3 receptors [
11], enhancement of trafficking of P2X3 receptors toward the membrane [
31] and activation of calmodulin protein kinase II [
34]. The chemical mediator prostaglandin E2 (PGE2) is released during inflammation and sensitizes peripheral terminals of DRG neurons [
35,
36]. It increases capsaicin-evoked currents [
37,
38], promotes the release of substance P and CGRP from sensory neurons [
39,
40]. In the spinal cord, PGE2 dis-inhibits dorsal horn neurons by blocking inhibitory glycinergic synaptic responses [
41,
42]. PGE2 was found to enhance α,β-elicited nociceptive behaviors in rats [
3]. The interaction between PGE2 and purinergic receptors has not been thoroughly investigated. Studying the action of PGE2 on ATP currents in DRG neurons, we found that PGE2 increases P2X3-receptor mediated ATP currents. Protein kinase A (PKA) mediates the potentiating action of PGE2.
Discussion
We found that PGE2 increases the amplitude of P2X3 homomeric receptor-mediated ATP currents without changing the kinetics of the currents in a majority of small and medium DRG neurons (Fig.
1, Table
1). In contrast, PGE2 had no effect on P2X2/3 heteromeric receptors. The enhancement of ATP responses is not due to a change in the apparent affinity of ATP for P2X3 receptors (Fig.
2) but the result of activation of PKA by PGE2. The conclusion is supported by the observations: (1) the PKA activators, e.g., forskolin, not only mimics the action of PGE2 but occludes the effect of PGE2 (Fig.
4) and (2) the PKA inhibitors, H89 and PKA-I, block the action of PGE2 (Fig.
5). Our conclusion is further supported by the behavioral observations that PGE2 sensitizes the α,β-meATP-induced spontaneous flinching responses [
3] and enhances α,β-meATP-induced allodynia and hyperalgesia (Fig.
6). We further showed that the enhancement could be blocked by intraplantar injection of H89 (Fig.
6). The molecular mechanism responsible for PKA-induced potentiation of ATP current is not well understood. Few studies were conducted to study the conserved PKA site on the P2X receptor molecules. There is evidence that phosphorylation at Ser431 in the C-terminus of P2X2 receptors reduces ATP currents mediated by P2X2 receptors transfected in HEK293 cells [
47]. On the other hand, the adenylyl cyclase, forskolin, was found to increase ATP currents in HEK cells expressing P2X4 receptors although the phosphorylation site on the receptor has not been identified [
48]. The factor contributing to the differential actions of PKA on different P2X receptor subtypes has yet to be determined. It is unclear how phosphorylation of the conserved PKA site changes P2X receptor activity. A number of possibilities have been proposed by studying phosphorylation of sites other than the PKA site. Phosphorylation of the PKC sites in P2X1 and P2X2 receptor slowed the rate of inactivation of ATP currents [
49,
50]. Calcium/calmodulin-dependent protein kinase II was found to enhance P2X3 receptor activity by promoting trafficking of the receptor to the membrane [
34]. Phosphorylation of regulatory proteins associated with P2X1 or P2X7 receptor increased the activity of the P2X receptors [
51,
52].
To identify the EP receptor involved in the action of PGE2 is challenging because agonists and antagonists of EP receptors are not completely selective [
53]. By using a combination of the agonists and antagonist, we conclude that EP3 mediates the potentiating action of PGE2. This conclusion is consistent with the study of Southall and Vasko [
39] who used reverse transcription polymerase chain reaction and antisense to identify the EP receptors participate in PGE2-induced sensitization of DRG neurons. They found that although EP1, 2, 3C and 4 are expressed in DRG neurons, as observed by others [
54], only EP3C and EP4 receptor subtypes mediate the PGE2-induced cAMP production and increased release of substance P and CGRP. We have not determined which EP3 splice variant participates in the potentiating action of ATP. Given that EP3A and EP3B mRNA are not found in DRGs [
39] and EP3C (i.e., EP3γ) is the only EP3 variant coupled to Gs to produce cAMP [
55], we suggest that PGE2 acts on EP3C to potentiate the response of P2X3 currents. In addition to increasing ATP currents, PGE2 was also found to decrease or have no effect on ATP currents in some DRG neurons. The reason for this variability has not been studied. DRG neurons expressing other EP receptor subtypes or devoid of EP receptor expression could contribute to the varying PGE2 actions. For example, the EP3D receptor is known to couple to Gi or Gq to reduce cAMP level in cells [
45].
It is well documented that PGE2 produces sensitization of TRPV1 receptors [
38] and TTX-resistant Na
+ channels [
36,
56] in DRG neurons. Here, we show that the P2X3 receptor is another target of PGE2. Since P2X3 receptors play an important role in pain processing [
12,
13] and their activation is sensitive to tissue and nerve injury [
3,
11], the modulation by PGE2 would provide a way for sensory neurons to specifically respond to tissue injury. Our study elucidates the mechanism of an acute action of PGE2 on ATP currents. P2X3 receptors have been shown to undergo profound changes in their expression and trafficking after inflammation and nerve injury [
11,
31,
34]. The binding properties of EP receptors [
57] and the signaling of PGE2 can change in inflamed tissues [
58‐
60]. It is of great interest to determine how PGE2 modulates P2X3 receptor-mediated responses after inflammation or nerve injury. A good understanding of the plasticity of PGE2 action on P2X receptors under injurious conditions will help us use the downstream targets of PGE2 for designing analgesics. Such therapeutic agents should be more specific and produce less side-effects, thus providing an alternative to cyclooxygenase 2 (COX2) inhibitors which can have devastating risks for patients [
61,
62].
Methods
Animals
All animal procedures were in accordance with the guidelines of the National Institutes of Health and the International Association for the Study of Pain and approved by the Institutional Animal Care and Use Committee at the University of Texas Medical Branch. Sprague Dawley rats were used in the study. Rats of 25–35 d of age were used in the electrophysiology study; older rats (8–10 weeks old) were used in behavioral experiments.
Electrophysiology
ATP currents were recorded from acutely dissociated neurons isolated from L4-5 DRGs of 25–35 d old rats. DRG neurons isolated from young rats survived better. The characteristics of ATP currents in young and adult DRGs were indistinguishable. DRGs were excised from sodium pentobarbital (50 mg/kg, i.p.) anesthetized rats and put in an ice-cold, oxygenated dissecting solution, which contained (in mM): 135 NaCl, 5 KCl, 2 KH2PO4, 1.5 CaCl2, 6 MgCl2, 10 glucose, and 10 HEPES, pH 7.2 (osmolarity, 305 mOsm). After removal of the connective tissue, the ganglia were put into a dissecting solution containing collagenase IV (1.0–1.5 mg/ml; Boehringer Mannheim, Indianapolis, IN) and trypsin (1.0 mg/ml; Sigma, St. Louis, MO) and incubated for 1 hr at 34.5°C. Afterward, DRGs were taken out from the enzyme solution, washed and put into another dissecting solution containing DNAase (0.5 mg/ml; Sigma). Ganglia were then triturated with fire-polished glass pipettes and the dissociated cells were placed on acid-cleaned glass coverslips. The experiments were performed at room temperature two hours after plating. Cells were continuously superfused (0.5 ml/min) with an external solution [130 mM NaCl/5 mM KCl/2 mM KH2PO4/2.5 mM CaCl2/1 mM MgCl2/10 mM HEPES/10 mM glucose, pH 7.3 (osmolarity, 295–300 mosM)]. In order to obtain a fast solution exchange, ATP (Sigma) was applied through an electronic valve with a solution exchange rate of 0.2 ms. This exchange rate was fast enough so that it would not limit peak ATP responses. Under voltage-clamp conditions, the whole-cell patch recording technique was used for current recordings. Membrane potential was held at -60 mV. Unless indicated, patch-clamp electrodes had a resistance of 3–5 Mohm when filled with the pipette solution, which contained (in mM): 145 K gluconate, 10 NaCl, 10 HEPES, 10 Glucose, 5 BAPTA and 1 CaCl2, pH 7.25 adjusted with KOH (osmolarity = 290 mOsm). The currents were filtered at 2–5 kHz and sampled at 100 μs per point.
Behavioral experiments
Mechanical allodynia was quantified by the responses to von Frey filament stimulation using the 50% threshold methods [
63]. Individual rats were placed on a metal mesh platform and under a plastic dome. Animals were habituated to the testing environment for at least 20 min before an experiment. A series of calibrated von Frey filaments of increasing strengths will be applied to the midplantar surface until a paw withdrawal occurred. The 50% threshold (expressed in grams), which is defined as 10 (Xf+ k × d)/10000 where Xf = value of the final von Frey filament unit used (in log units); k = tabular value of the pattern of positive/negative response; d = mean difference (in log units) between von Frey hairs, was determined. A cutoff threshold was 15 g to avoid tissue damage. All behavioral studies were performed under blind conditions. Saline (PBS) or α,β-meATP was injected into the rat paw. The α,β-meATP effect on PW threshold was measured 10 min after the injection when the response reached a peak level. To study the effect of PGE2 on the α,β-meATP response, α,β-meATP and PGE2 were applied simultaneously to the rat paw. To study the protein kinase-dependence of PGE2, the PKA antagonist, H89 was applied 20 min prior to the (PGE2+α,β-meATP) application.
Thermal hyperalgesia was examined by measuring paw withdrawal latencies (PWLs) using the radiant heat method [
64]. A lamp was placed under the plantar surface of the rat hindpaw and the time elapsed from the onset of radiant heat stimulation to the withdrawal of the paw was recorded. The heat intensity was adjusted to give a baseline latency of ≈10 s; a cutoff time of 30 s was set to prevent tissue damage. To obtain baseline PWLs, three measurements separated by a 5-min interval were made for each rat's hind paw and scores were averaged.
Drugs
Bisindolylmaleimide I (Bis), H89 dihydrochloride (H89), protein kinase A inhibitor 6–22 amide (PKA-I), were purchased from Calbiochem (La Jolla, CA). ATP, α,β-meATP, 8-Bromoadenosine 3',5'-cyclic monophosphate (8-Br-cAMP), A-316491 and forskolin were from Sigma (St. Louis, MO) ; PGE2, sulprostone, butaprost and SC-1220 were from Cayman Chemical (Ann Arbor, MI). Except for PKA-I, ATP and α,β-meATP, all other compounds were prepared in DMSO as stocks. The final concentrations used in experiments at least 1000 times lower than stocks concentrations. Stock solutions were stored at -20°C and diluted immediately before use.
Data analyses
All data are expressed as mean ± SEM. Differences between two means were analyzed with paired or unpaired Student's t-test. Rise time (Ta) and inactivation time constants were used to determine kinetic properties of ATP responses. Ta of ATP currents were obtained by measuring the activation time between 10 and 90% of the peak value. The time constants of current inactivation were obtained by fitting the decay phase of current with exponential functions using the Levenberg-Marquardt algorithm. Comparisons between multiple means were done with one-way analysis of variance (ANOVA) followed by Newman Keuls post hoc test. A P < 0.05 was considered significant.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
CW designed and performed the electrophysiological experiments and drafted the manuscript. GL carried out the behavioral experiments. LMH designed and coordinated the study and prepared the manuscript. All authors read and approved the final manuscript.