Sensitization of TRPV1 by EP1 and IP reveals peripheral nociceptive mechanism of prostaglandins

Prostaglandin E2 (PGE2) and prostaglandin I2 (PGI2) are major inflammatory mediators that play important roles in pain sensation and hyperalgesia. The role of their receptors (EP and IP, respectively) in inflammation has been well documented, although the EP receptor subtypes involved in this process and the underlying cellular mechanisms remain to be elucidated. The capsaicin receptor TRPV1 is a nonselective cation channel expressed in sensory neurons and activated by various noxious stimuli. TRPV1 has been reported to be critical for inflammatory pain mediated through PKA- and PKC-dependent pathways. PGE2 or PGI2increased or sensitized TRPV1 responses through EP1 or IP receptors, respectively predominantly in a PKC-dependent manner in both HEK293 cells expressing TRPV1 and mouse DRG neurons. In the presence of PGE2 or PGI2, the temperature threshold for TRPV1 activation was reduced below 35°C, so that temperatures near body temperature are sufficient to activate TRPV1. A PKA-dependent pathway was also involved in the potentiation of TRPV1 through EP4 and IP receptors upon exposure to PGE2 and PGI2, respectively. Both PGE2-induced thermal hyperalgesia and inflammatory nociceptive responses were diminished in TRPV1-deficient mice and EP1-deficient mice. IP receptor involvement was also demonstrated using TRPV1-deficient mice and IP-deficient mice. Thus, the potentiation or sensitization of TRPV1 activity through EP1 or IP activation might be one important mechanism underlying the peripheral nociceptive actions of PGE2 or PGI2.


Background
Tissue damage and inflammation produce an array of chemical mediators such as ATP, bradykinin, prostanoids, protons, cytokines and peptides including substance P that can excite or sensitize nociceptors to elicit pain at the site of injury. Among them prostanoids were shown to influence inflammation, and their administration was found to reproduce the major signs of inflammation including augmented pain [1]. Prostaglandin E 2 (PGE 2 ) and prostaglandin I 2 (PGI 2 ) are the products of arachidonic acid metabolism through the cyclooxygenase pathway. In addition to numerous other physiological actions in vivo, previous studies have indicated important roles for PGE 2 in nociception and inflammation [2,3]. PGE 2 is generated in most cells in response to mechanical, thermal or chemical injury and inflammatory insult, resulting in sensitization or direct activation of nearby sensory nerve endings. Analgesic effects of non-steroidal anti-inflammatory drugs (NSAIDs) are attributed predominantly to inhibition of prostaglandin synthesis. Prostaglandins act upon a family of pharmacologically distinct prostanoid receptors including EP 1 , EP 2 , EP 3 , EP 4 and IP that activate several different G protein-coupled signaling pathways [2,4,5]. Primary sensory neurons in dorsal root ganglion (DRG) are known to express mRNAs encoding several prostanoid receptor subtypes, IP, EP 1 , EP 3 and EP 4 [6,7]. The role of IP in inflammation has been clearly shown by the analysis of IP-deficient mice, although the underlying cellular mechanisms still remain to be elucidated [8]. In contrast, the potential involvement of EP receptors other than IP in inflammation and pain generation has not been well studied, although some earlier studies have suggested that prostanoids contribute to the development of pain through EP receptors [9,10].
The capsaicin receptor TRPV1 is a non-selective cation channel expressed predominantly in unmyelinated C-fibers [11]. TRPV1 is activated not only by capsaicin, but also by protons or heat (with a threshold > ~43°C), both of which cause pain in vivo [11][12][13]. A prominent role of TRPV1 in nociception has been demonstrated in studies of TRPV1-deficient mice [14,15].

PGE 2 increases TRPV1 activity through EP 1 receptors
To explore the mechanism underlying the PKA-independent PGE 2 (1.5 min)-induced potentiation of the capsaicinevoked responses observed in DRG neurons, we first examine the effects of PGE 2 on capsaicin-activated currents in HEK293 cells expressing TRPV1 and each EP receptor. PGE 2 (1 µM, 1.5 min) caused a robust increase in the magnitude of low dose (20 nM) capsaicin-activated currents in HEK293 cells co-expressing TRPV1 with EP 1 (0.90 ± 0.04 fold increase, n = 9 for control (Cont.); 4.60 ± 1.03 fold, n = 17 for PGE 2 , p < 0.05) (Figures 2A and  2B). This increase lasted more than three minutes, as we previously reported for PAR-2 (proteinase activated receptor 2)-mediated potentiation of TRPV1 activity [16]. In contrast, no such potentiation was detected in cells expressing TRPV1 with other EP receptor subtypes (0.91 ± 0.09 fold increase, n = 7; 0.77 ± 0.13, n = 9; 0.72 ± 0.24, n = 5; 0.98 ± 0.18, n = 7; 0.89 ± 0.15, n = 9 for EP 2 , EP 3α , EP 3β , EP 3γ or EP 4 , respectively) ( Figure 2B). Protracted (6.5 min) treatment with PGE 2 caused a significant increase in capsaicin-activated currents in cells expressing TRPV1 and EP 4 , a phenomenon like that observed following treatment with a mixture of FSK, IBMX and dbcAMP (3.03 ± 0.48 fold increase, n = 6, p < 0.05 vs. Cont.) ( Figure 2B), suggesting that the EP 4 receptor, known to be expressed in DRG and coupled to Gs protein, is the receptor that activates a PKA-dependent signaling pathway upon prostaglandin exposure. All cells exhibiting an increase of capsaicin-activated currents upon treatment with a mixture of FSK, IBMX and dbcAMP also showed an increase in current in the presence of PMA (data not shown), suggesting that both PKA-and PKC-dependent pathways work in the same cells. To examine how PGE 2 changes TRPV1 responsiveness, we measured TRPV1 current in single cells by applying a range of concentrations of capsaicin in the absence or presence of PGE 2 . The currents were normalized to the maximal current produced by 1 µM capsaicin in each cell. Maximal current in the presence of PGE 2 was almost the same as that in the absence of PGE 2 . The resultant dose-response curves clearly demonstrate that PGE 2 enhances capsaicin action on TRPV1 by lowering EC 50 values without altering maximal responses (EC 50 from 81.0 nM to 27.6 nM) ( Figure 2C). We next examined the effects of PGE 2 on the thermal sensitivity of TRPV1. When temperature ramps were applied to HEK293 cells expressing both TRPV1 and EP 1 in the absence of PGE 2 , heat-evoked currents developed at 40.7 ± 0.3°C (n = 8) ( Figure 2D). In contrast, the temperature threshold for TRPV1 activation was significantly reduced to 30.6 ± 1.1°C in the presence of PGE 2 (n = 8, p < 0.05) ( Figure 2D) implying that under these conditions, TRPV1 could be activated at normal body temperature. A similar potentiating effect of PGE 2 was observed for proton (pH 6.2)-evoked TRPV1 current responses (0.91 ± 0.06 fold increase, n = 3 for control; 4.47 ± 1.09 fold, n = 7 for PGE 2 , p < 0.01) ( Figure 2E). These data clearly show that TRPV1 currents evoked by any of three different stimuli (capsaicin, proton, or heat) are potentiated or sensitized by PGE 2 through EP 1 receptor activation. On the other hand, the temperature threshold for TRPV1 activation was not changed upon treatment with a mixture of FSK, IBMX and dbcAMP in HEK293 cells expressing TRPV1 (40.8 ± 0.8°C, n = 4), suggesting different actions on TRPV1 by PKA and PKC.
The signaling pathway downstream of EP 1 remains to be clarified. We have reported that G q/11 -coupled metabotropic receptor activation such as ATP (P2Y), bradykinin (B2) and proteinase-activated receptor 2 (PAR2) receptors causes potentiation or sensitization of TRPV1 through the PKC-dependent phosphorylation of TRPV1 [16][17][18]25]. Therefore, we examined whether a similar signal transduction pathway is involved in the regulation of TRPV1 responses through EP 1 . When calphostin C (Calp.C), a specific PKC inhibitor, was added to the pipette solution, the effect of PGE 2 was almost completely inhibited (0.92 ± 0.15 fold increase, n = 10) ( Figure 2F). Similarly, a PKCε translocation inhibitor (PKCε-I) abolished the potentiation of TRPV1 response by PGE 2 (1.11 ± 0.25 fold increase, n = 11) ( Figure 2F). These data suggest that PGE 2 -induced potentiation of TRPV1 responsiveness develops through activation of PKCε. To further confirm the involvement of PKC-dependent phosphorylation, PGE 2 effects were examined using cells expressing a TRPV1 mutant, S502A/S800A which is insensitive to PKCdependent phosphorylation [19]. No potentiation of capsaicin-activated currents was observed upon PGE 2 treatment of cells expressing S502A/S800A (0.85 ± 0.15 fold increase, n = 5) ( Figure 2F), further indicating the involvement of PKC-dependent phosphorylation. Since S502 is a PKA-phosphorylation site as well [26], we examined the effects of treatment with a mixture of FSK, IBMX and dbcAMP on the capsaicin-activated currents in cells expressing S502A/S800A. Such treatment failed to potentiate the capsaicin-activated currents (1.13 ± 0.07 fold increase, n = 10), suggesting that S502 is a substrate for PKA-dependent phosphorylation of TRPV1 as well.

Sensitization of TRPV1 by EP 1 receptors in mouse
To examine the involvement of EP 1 in PGE 2 (1.5 min)induced potentiation of capsaicin-evoked response in native neurons, we used a specific EP 1 agonist, ONO-DI-004 [27], and a specific EP 1 antagonist, ONO-8713 [28], in mouse DRG neurons. ONO-DI-004 was found to significantly increase the capsaicin-activated currents to an extent similar to that observed with PGE 2 (3.36 ± 0.68 fold PGE 2 increases TRPV1 activity through EP 1 receptors in a PKC-dependent manner in HEK293 cells fold, n = 8 for U73433, p < 0.05) ( Figure 3B). Furthermore, PGE 2 failed to potentiate the capsaicin-activated currents when PKCε-I was included in the pipette solution (0.86 ± 0.09 fold increase, n = 12) ( Figure 3B). A robust potentiating effect of phorbol 12-myristate 13-acetate (PMA, 100 nM) also supported the involvement of PKCdependent events (16.36 ± 3.68 fold increase, n = 11, p < 0.05) ( Figure 3B). To further confirm the involvement of EP 1 receptors, DRG neurons of EP 1 deficient mice (EP 1 -/-) were subjected to patch-clamp analysis. PGE 2 failed to potentiate capsaicin-activated currents in the DRG neurons from EP 1 -/mice (1.45 ± 0.70 fold increase, n = 10) ( Figure 3B). Functional interaction of PKCε with TRPV1 prompted us to examine the expression of the two proteins in mouse DRG. Three hundred seventy eight out of 541 TRPV1 positive neurons (69.9 %) were stained with anti-PKCε antibody, supporting the TRPV1 activation pathway through PKCε ( Figure 3C).
gesia observed in wild type mice disappeared almost completely in both TRPV1-deficient (TRPV1 -/-) mice and IP-deficient (IP -/-) mice, suggesting that the functional interaction of TRPV1 with IP causes thermal hyperalgesia at the behavioral level ( Figure 6).

Discussion
The data presented herein demonstrate that TRPV1 is essential for the development of thermal hyperalgesia in vivo induced by two major inflammation-associated prostaglandins, PGE 2 and PGI 2 , and that TRPV1 and EP 1 or IP receptors can functionally interact, mainly through a PKCdependent pathway. The temperature threshold for TRPV1 activation is reduced below 35°C in the presence of prostaglandins, so that TRPV1 can be activated at normal body temperature, possibly leading to spontaneous pain sensation. This interaction might be one important underlying mechanism for the well-recognized peripheral nociceptive actions of PGE 2 or PGI 2 in the context of inflammation. In the present study, 1 µM PGE 2 or PGI 2 was found to potentiate or sensitize TRPV1 activity. It is not well known how much PGE 2 or PGI 2 is released locally at the site of inflammation. However, more than micromolar-order concentrations of PGE 2 and PGI 2 have been reported to be synthesized by macrophages upon lipopolysacharide (LPS) stimulation [33,34], suggesting that 1 µM is an attainable concentration in the context of inflammation. It has been previously reported that EP 1 is coupled to intracellular Ca 2+ mobilization in CHO cells [35]. However, the transduction events downstream of EP 1 signaling have been unclear. Together with a report suggesting the possible coupling of EP 1 with G q/11 -protein [36], our data indicate that EP 1 receptors activate a PKCdependent signal transduction pathway.
There has been extensive work demonstrating the activation of a PKA-dependent pathway by PGE 2 that influences capsaicin-or heat-mediated actions in rat sensory neurons [20][21][22]37,38] as well as interactions between cloned TRPV1 and PKA [26,[39][40][41][42]. These results suggest that PKA plays a pivotal role in the development of hyperalgesia and inflammation by prostaglandins. In our experiments using mouse DRG neurons and HEK293 cells expressing TRPV1, a PKC-dependent pathway was found to be predominantly involved in both PGE 2 (1.5 min)-and PGI 2 (1.5 min)-induced responses. The reason that there has been no study describing the involvement of a PKCdependent pathway in the regulation of TRPV1 following prostaglandin receptor activation is not clear. In the present study, it was found that both PKA-and PKCdependent pathways are involved downstream of prostaglandin actions on TRPV1 although the PKC-dependent one appears to predominate. A PKA-dependent pathway took a relatively long time to exert its potentiating effects on TRPV1 activity, suggesting some difference between PKA-and PKC-dependent phosphorylation of TRPV1. Indeed, Bhave et al. treated cells with 8-Br-cAMP for 30 min to inhibit TRPV1 desensitization through phosphorylation [39], and significant potentiation of capsaicinactivated currents in rat DRG neurons was observed upon prolonged (greater than 10 min) exposure to PGE 2 [21]. Furthermore, there is a report describing the ineffectiveness of PKA stimulation on TRPV1 currents in Xenopus oocytes treated with 8-Br-cAMP and IBMX for relatively short periods [24]. Both PKA-dependent and PKCdependent pathways might work in concert in native cells. Patch-clamp recordings in the previous studies were performed in the Ca 2+ -containing solutions, whereas we did all of our experiments under Ca 2+ -free conditions, to avoid Ca 2+ -dependent TRPV1 desensitization [43]. Potentiation of capsaicin-activated currents by PGE 2 was observed in embryonic rat DRG neurons [21] while we used adult mouse DRG neurons. Furthermore, potentiation of heat-activated currents [26], inhibition of desensitization of capsaicin-activated currents [39,41,44] or anandamide-induced cytosolic Ca 2+ increase [40] but not potentiation of capsaicin-activated current response were examined in the previous studies investigating the involvement of PKA-dependent pathway in TRPV1 activity. Thus, difference in experimental conditions or readout might also account for the different outcomes. The physiological relevance of the two different pathways downstream of prostaglandin exposure remains to be elucidated. The fact that only PKC activation leads to the reduction of temperature threshold for TRPV1 activation might be pertinent to this issue. Disruption of interaction between phosphatidylinositol-4, 5-bisphosphate (PIP 2 ) and TRPV1 has also been reported to be involved in the sensitization of TRPV1 downstream of PLC activation [45,46]. In our study, however, both PGE 2 -and PGI 2induced potentiation of TRPV1 activity was completely inhibited by treatments with two kinds of PKC inhibitors. Thus, we believe that a PKC-dependent pathway is predominantly involved in the PGE 2 -and PGI 2 -induced potentiation or sensitization of TRPV1 activity in mice.
The inhibition of PGE 2 -induced thermal hyperalgesia observed in EP 1 -/mice, while significant, was not very robust, compared with that in TRPV1 -/mice ( Figure 4). Other pathways, most likely including one involving PKA, might account for the residual component. Further, inhibition of mustard oil-induced thermal hyperalgesia observed in TRPV1 -/or EP 1 -/mice might seem not to be robust or dramatic (Figure 4). Since many inflammatory factors activating PLC-coupled receptors are involved in the inflammatory response [47,48]. In such a complicated environment, thermal hyperalgesia was significantly diminished in TRPV1 -/mice or EP 1 -/mice albeit at a few time points, suggesting the importance of the two molecules in the context of inflammatory pain sensation.
Given the fact that one of the final targets of both PGE 2 and PGI 2 is TRPV1 as shown in our study, compounds acting on EP 1 , IP or TRPV1, or interfering with their interaction could prove useful in the treatment of pain and inflammation.

Conclusions
Potentiation or sensitization of TRPV1 activity through EP 1 or IP activation, mainly through PKC-and PKAdependent mechanisms, might be important mechanism underlying the peripheral nociceptive actions of PGE 2 or PGI 2 .

Animals
Male C57BL/6-strain mice (4 weeks, SLC, Shizuoka, Japan), EP 1 -deficient mice (4 weeks, from Dr. Narumiya), IP-deficient mice (4 weeks, from Dr. Narumiya) or TRPV1-deficient mice (4 weeks, from Dr. Julius, UCSF) were used. They were housed in a controlled environment (12 h light/dark cycle, room temperature 22-24°C, 50-60% relative humidity) with free access to food and water. All procedures involving the care and use of mice were carried out in accordance with institutional (Mie University) guidelines and the National Institute of Health guide for the care and use of laboratory animals.

Behavioral study
Thermal nociceptive threshold was assessed using the paw withdrawal test. Mice were placed in a transparent Perspex box on a thin glass platform (Plantar test, Ugo Basile, Italy). They were injected intraplantarly with PGE 2 (500 pmol/ 20 µL, Sigma) with or without ONO-8713 (500 pmol/ 20 µL), or with PGI 2 (500 pmol/ 20 µL, Sigma), or applied topically to the plantar surface of right hind paw with 10% mustard oil (Sigma) (diluted with mineral oil), and the paw withdrawal latency to radiant heat applied to the plantar surface of hind paw was measured as the time from onset of the radiant heat to the withdrawal of the mouse hind paw.

Electrophysiology
Whole-cell patch-clamp recordings were performed 1 day after transfection to HEK293 cells or dissociation of the DRG neurons. Standard bath solution contained 140 mM NaCl, 5 mM KCl, 2 mM MgCl 2 , 5 mM EGTA, 10 mM HEPES, 10 mM glucose, pH7.4 (adjusted with NaOH). Pipette solution contained 140 mM KCl, 5 mM EGTA, 10 mM HEPES, pH7.4 (adjusted with KOH). All patch-clamp experiments were performed at room temperature (22°C). Thermal stimulation was applied by increasing the bath temperature at a rate of 1.0°C/sec with a preheated solution. When the heat-activated currents started to inactivate, the preheated solution was changed to a 22°C one. Chamber temperature was monitored with a thermocouple placed within 100 µm of the patchclamped cell. For this analysis, heat-evoked current responses were compared between different cells, rather than within the same cell, because repetitive heat-evoked currents show significant desensitization even in the absence of extracellular Ca 2+ [13] and because the thermal sensitivity of TRPV1 increases with repeated heat application [49]. Threshold temperature for activation was defined as the intersection where two lines approximating the stable baseline current and the clearly increasing temperature-dependent current cross in the temperatureresponse profile. The sensitivity of DRG neurons to capsaicin is slightly lower than that of TRPV1-transfected HEK293 cells as previously reported [18,50]. Therefore, we applied capsaicin at 100 nM to DRG neurons and at 20 nM to HEK293 cells.

cAMP measurement
Intracellular cAMP level was examined using 'cAMP Biotrak Enzymeimmunoassay System' according to the manufacture's direction (Amersham Biosciences). In brief, intracellular cAMP released upon membrane hydrolysis of treated cells (10,000 cells/ well) after stimulation (90 sec) was measured based on competition between unlabelled cAMP and a fix quantity of Peroxidase-labeled cAMP for a limited number of the binding sites on a cAMP specific antibody.
Immunostaining DRG was removed from male C57BL/6-strain mice and frozen in liquid nitrogen, and the frozen tissue was cut on a cryostat at a 10 µm thickness. The sections were incubated with the rabbit anti-rat TRPV1 polyclonal antibody (1: 500; Oncogene) and anti-rat PKCε monoclonal antibody (1: 250; Transduction lab) at 4°C for 2 days. Slides with the section were washed with PBS, followed by incubation with Alexa 488-conjugated goat anti-rabbit IgG (1: 700, Molecular Probes), Alexa 350-conjugated anti-mouse IgG (1: 500, Molecular Probes) and Texas Red-phalloidin (1: 500, Molecular Probes). Images were obtained using an Olympus fluorescent microscope with a cooled-CCD camera (ORCA-ER, Hamamatsu Photonics) and IP-Lab Image software (Scanalytics Inc.).

Statistics
Values are shown as the mean ± S.E. and data are analyzed using an unpaired t test. P values of < 0.05 were considered significant.