Expression and function of proton-sensing G-protein-coupled receptors in inflammatory pain

Background Chronic inflammatory pain, when not effectively treated, is a costly health problem and has a harmful effect on all aspects of health-related quality of life. Despite the availability of pharmacologic treatments, chronic inflammatory pain remains inadequately treated. Understanding the nociceptive signaling pathways of such pain is therefore important in developing long-acting treatments with limited side effects. High local proton concentrations (tissue acidosis) causing direct excitation or modulation of nociceptive sensory neurons by proton-sensing receptors are responsible for pain in some inflammatory pain conditions. We previously found that all four proton-sensing G-protein-coupled receptors (GPCRs) are expressed in pain-relevant loci (dorsal root ganglia, DRG), which suggests their possible involvement in nociception, but their functions in pain remain unclear. Results In this study, we first demonstrated differential change in expression of proton-sensing GPCRs in peripheral inflammation induced by the inflammatory agents capsaicin, carrageenan, and complete Freund's adjuvant (CFA). In particular, the expression of TDAG8, one proton-sensing GPCR, was increased 24 hours after CFA injection because of increased number of DRG neurons expressing TDAG8. The number of DRG neurons expressing both TDAG8 and transient receptor potential vanilloid 1 (TRPV1) was increased as well. Further studies revealed that TDAG8 activation sensitized the TRPV1 response to capsaicin, suggesting that TDAG8 could be involved in CFA-induced chronic inflammatory pain through regulation of TRPV1 function. Conclusion Each subtype of the OGR1 family was expressed differently, which may reflect differences between models in duration and magnitude of hyperalgesia. Given that TDAG8 and TRPV1 expression increased after CFA-induced inflammation and that TDAG8 activation can lead to TRPV1 sensitization, it suggests that high concentrations of protons after inflammation may not only directly activate proton-sensing ion channels (such as TRPV1) to cause pain but also act on proton-sensing GPCRs to regulate the development of hyperalgesia.


Background
Inflammation induced by tissue injury, infection or tumor growth often accompanies persistent and chronic pain that heightens a pain experience by increasing the sensitivity of nociceptors to both thermal and mechanical stimuli. This phenomenon results, in part, from the production and release of chemical mediators (e.g., protons, adenosine triphosphate, bradykinin, histamine, postaglandin, serotonin) from the primary sensory terminal and from non-neural cells in the environment [1,2]. High local proton concentrations found in inflamed tissues (tissue acidosis) contribute directly to pain and hyperalgesia. The degree of acid-associated pain or discomfort is well associated with the magnitude of acidification, which is attributable to direct excitation or modulation of nociceptive sensory neurons by protonsensing receptors [3][4][5][6][7].
The ovarian cancer G-protein-coupled receptor 1 (OGR1) family, consisting of OGR1, GPR4, TDAG8, and G2A, respond to proton stimulus with full activation at pH 6.4-6.8 [24][25][26][27]. These four receptors were found in DRG, and most (75%~82%) are present in small-diameter neurons that are responsible for nociception [28,29]. More than half of these genes are expressed in IB 4 -positive neurons that are involved in inflammatory or neuropathic pain. However, the functions of the OGR1 family subtypes in chronic pain remain unclear.
In this study, we used the inflammatory agents capsaicin, carrageenan, or complete Freund's adjuvant (CFA) to induce peripheral inflammation and examined the change in expression of OGR1 family genes in DRG neurons. OGR1 family subtypes showed differential expression in various types of inflammatory pain (neurogenic, short-term or long-term inflammatory pain). TDAG8 seems to be the major subtype involved in CFA-induced chronic inflammation. Enhanced expression of TDAG8 gene is mainly due to an increase in total number of TDAG8-expressing neurons. Further studies demonstrated that the number of DRG neurons expressing both TDAG8 and TRPV1 genes was increased as well, and TDAG8 activation sensitized TRPV1 response to capsaicin. TDAG8 could be involved in CFA-induced chronic inflammatory pain by regulating TRPV1 function.

Differential expression of proton-sensing GPCR genes after induction of peripheral inflammation
To understand whether the proton-sensing OGR1 family is involved in inflammatory pain, mRNA expression was examined in different inflammatory pain states. Peripheral inflammation was generated by intraplantar injection with capsaicin, carrageenan, CFA, or saline. Capsaicin injection induced neurogenic inflammatory pain. Carrageenan and CFA are commonly used models of short-term and long-term chronic inflammatory pain, respectively.

Neurogenic inflammatory pain
Capsaicin administered locally caused pain-related behaviors. The greatest magnitude of pain-related behaviors was perceived during the first minute after the injection and then rapidly decreased (data not shown). Within 15 minutes, injected mice showed unilateral edema, which gradually decreased. Four hours after injection, the edema lessened, with a paw thickness of 3.09 ± 0.13 mm on the injected (ipsilateral) paw and 2.85 ± 0.13 mm on the uninjected (contralateral) paw (Fig. 1A). A small degree of edema lasted for three days (2.94 ± 0.17 and 2.69 ± 0.12 mm for the ipsilateral and contralateral paws, respectively). Saline-injected controls showed no edema (2.64 ± 0.08 and 2.49 ± 0.06 mm for the ipsilateral and contralateral paws, respectively; Fig. 1D). Surprisingly, after capsaicin injection, GPR4 gene expression decreased 24 hours (2.6~3.4-fold that of the contralateral paw), and such decrease lasted for three days (2.6~4.0-fold at 72 hours). In addition, after capsaicin injection, G2A gene expression decreased (2.7~4.5-fold) at 72 hours after injection. Saline injection produced no increase or decrease in OGR1 family gene expression (Fig. 2D).

Short-term inflammatory pain
In carrageenan-induced short-term inflammation, unilateral edema was detected early, within 15 minutes, and peaked at 4 hours after injection (4.36 ± 0.29 and 2.87 ± 0.14 mm for the ipsilateral and contralateral paws, respectively, Fig. 1B). Although decreasing with time, peripheral edema extended to three days (4.18 ± 0.12 mm at 24 hours and 3.85 ± 0.09 mm at 72 hours). At 24 hours after carrageenan injection, GPR4, TDAG8 and G2A expression increased (1.4~2.2-fold for GPR4, 1.4~2.1-fold for G2A, and 1.4~1.9-fold for TDAG8; Fig. 2B). At 72 hours after injection, the expression of GPR4 and G2A returned to baseline, but that of TDAG8 remained at high levels as compared with saline injection.
Long-term inflammatory pain CFA injection induced unilateral peripheral edema 4 hours after injection (3.86 ± 0.19 mm for the ipsilateral paw), smaller than was observed after carrageenan injection (an increase of 46% with CFA injection and 65% with carrageenan injection, as compared with saline injection; Fig. 1A, D). The edema peaked at 24 hours after injection (4.29 ± 0.15 mm), then gradually decreased but remained for at least three weeks (3.33 ± 0.19 mm at 21 days after injection; data not shown). At 24 hours after CFA injection, only TDAG8 expression was increased greatly (2.5~4.3-fold), but the level was reduced at 72 hours (1.2~1.6-fold) (Fig. 2C). In contrast, G2A expression was first decreased at 4 hours after injection (2~3-fold reduction) and gradually returned to basal levels.
Both mechanical and thermal hyperalgesia developed on the ipsilateral paw in CFA-injected mice at 4 hours after injection (Fig. 3). Before injection, the paw withdrawal threshold (50%) for mechanical stimuli (mechanical hyperalgesia) was 1.63 ± 0.11 g on the ipsilateral paw and 1.78 ± 0.11 g on the contralateral paw (0 hour). After CFA injection, the threshold was decreased on the ipsilateral paw to 0.60 ± 0 g at 4 hours and 0.56 ± 0.04 g at 24 hours but remained similar to that in controls on the contralateral paw (1.84 ± 0.29 g at 4 hours and 1.70 ± 0.10 g at 24 hours; Fig. 3A). Mechanical hyperalgesia remained for three weeks (data not shown). Non-noxious mechanical stimulation did not induce hyperalgesia in saline-injected mice (Fig. 3B). The latency of paw withdrawal to heat stimuli (thermal hyperalgesia) was 13.06 ± 0.55 seconds for the contralateral paw and 13.18 ± 0.84 seconds for the ipsilateral paw before injection. At 4 hours after CFA injection, the latency of paw withdrawal to heat was decreased to 9.44 ± 0.38 seconds for the ipsilateral paw and remained at 13.20 ± 0.51 seconds for the contralateral paw (Fig. 3C, D). The decreased latency to heat stimulation lasted for three weeks (data not shown).
Gene expression change of OGR1 family genes after peripheral inflammation Figure 2 Gene expression change of OGR1 family genes after peripheral inflammation. The wild-type CD1 mice (8)(9)(10)(11)(12) week-old) were injected at right hind paws with 25 μl of capsaicin (100 μg/ml in saline containing 10% ethanol and 0.5% Tween 80, A), carrageenan (20 mg/ml, B), CFA (50% in saline, C), or saline (D). At 4, 24, and 72 hours after injection, the mice were killed. Lumbar 4-6 DRG ipsilateral and contralateral to injected paws were taken for RNA extraction for quantitative RT-PCR. The contralateral DRG was used as untreated controls. The expression of each gene on the ipsilateral DRG was first normalized to that of mGAPDH and then represented as a relative value to contralateral controls. All data are presented as mean ± SEM of quadruplicates of three experiments (N = 3, n = 3 mice). Comparison between inflammatory agent-injected and salineinjected animals was by t test. *p < 0.05, **p < 0.01, ***p < 0.001.

DRG neurons expressing TDAG8 increase in number with CFA-induced inflammation
TDAG8 was the only receptor whose expression increased at 24 hours after CFA injection (Fig. 2C). To further determine whether the increased TDAG8 expression was due to an increased number of DRG neurons, we used in situ hybridization to examine L4-5 DRG 24 hours after CFA injection. CFA injection did not change the distribution of neurons that were PERI positive (67 ± 3% and 66 ± 2% for the contralateral and ipsilateral DRGs, respectively) and N52 positive (41 ± 3% and 41 ± 2% for the contralateral and ipsilateral DRGs, respectively) ( Fig. 4 and Table 1). At 24 hours after CFA injection, 27 ± 2% of total neurons expressed TDAG8 in the contralateral DRG, which was consistent with results for untreated controls found in a previous study [28], whereas the number was increased to 38 ± 2% on the ipsilateral DRG (Fig. 4A, B, Table 1). TDAG8-expressing neurons were increased in number in small-and large-diameter neuron populations. Of PERIpositive neurons, 24 ± 3% expressed TDAG8 on the ipsilateral DRG and 20 ± 2% on the contralateral DRG (Fig.  4A, B). In N52-positive populations, 17 ± 2% showed TDAG8 expression on the ipsilateral DRG and 10 ± 3% on the contralateral DRG (Fig. 4A, B). The distribution of Mechanical and thermal hyperalgesia in mice after CFA-induced peripheral inflammation Figure 3 Mechanical and thermal hyperalgesia in mice after CFA-induced peripheral inflammation. The wild-type CD1 mice (12 week-old) were injected with 25 μl CFA (50% in saline, A, C), or saline (B, D). The threshold (A, B) and the latency (C, D) of paw withdrawal were measured before injection (t = 0) and after injection (t = 4, 24, and 72 hours). All data are mean ± SEM of total tested mice (n = 6 per group). Comparison between inflammatory agent-injected and saline-injected animals (#) or between contralateral side and ipsilateral side of agent-injected animals (*) was by t test. *#p < 0.05, **##p < 0.01, ***###p < 0.001.
TDAG8-expressing neurons shifted slightly to a population of N52-positive neurons after CFA injection (Fig.  4C), which suggests that TDAG8-expressing neurons were greater in number in medium-to large-diameter neurons.
TDAG8 expression was further examined in peptidergic (IB 4 -negative) and non-peptidergic (IB 4 -positive) subpopulations. CFA-induced inflammation did not alter the distribution of neurons in nociceptors that were IB 4 positive (52 ± 4% on the contralateral DRG and 56 ± 5% on the ipsilateral DRG) or negative (48 ± 4% and 44 ± 5% on contralateral and ipsilateral DRG, respectively) (Fig. 5A, B and Table 2). Of the neurons labeled with PERI, 33 ± 4% of IB 4 -positive neurons expressed TDAG8 on the ipsilateral DRG as compared with 21 ± 4% on the contralateral DRG at 24 hours after CFA injection. TDAG8-expressing neurons were also increased in number in the IB 4 -negative population (12 ± 3% and 21 ± 4% for the contralateral and ipsilateral DRGs, respectively) ( Fig. 5B, Table 2). Accordingly, TDAG8-expressing neurons were increased in number in both IB 4 -positive and -negative populations after CFA injection.

TDAG8 activation increases levels of intracellular cAMP
Whether increased TDAG8 expression enhances TDAG8 function after inflammation has remained unclear. We first examined proton signaling in HEK293T cells transfected with TDAG8. Consistent with previous studies [26], TDAG8-expressing cells responded to protons and induced increased levels of intracellular cAMP. The cAMP response peaked at pH 6.6~6.4 (data not shown). Levels of cAMP with pH 6.4 was 6-fold higher that with pH 7.4 (Fig. 6A). To confirm whether the signaling elicited by TDAG8 is through the Gs protein, we used two inhibitors: pertussis toxin (PTX) blocks Gi protein-mediated signaling, and U73122 inhibits phospholipase Cβ (PLCβ), which is activated by Gq or Gi protein. Treatment with PTX or U73122 did not inhibit the cAMP signaling elicited by TDAG8, which suggests that cAMP accumulation is through activation of Gs protein (Fig. 6B).
The accumulation of intracellular cAMP was further examined in primary DRG cultures. Surprisingly, cAMP accumulation declined with acid stimulation in untreated DRG cultures (Fig. 6C), which suggests a Gi-mediated signaling elicited by acid stimulation. At 24 hours after CFA injection, DRG cultures from ipsilateral paws showed higher levels of cAMP after pH 6.4 stimulation (~2-fold increase) than those from contralateral paws (Fig. 6D), although the increase was not great. Since acid-induced Gi-signaling was found in primary culture, increased cAMP levels could be resulted from enhancement of Gssignaling or reduction of Gi-signaling. To clarify this point, intracellular [Ca 2+ ] ([Ca 2+ ] i ) was examined. After pH6.4 stimulation, the same levels of [Ca 2+ ] i were found in contralateral and ipsilateral DRG cultures from CFAinjected mice (Fig. 6E), suggesting that CFA-injection did not enhance or reduce acid-induced Gi-or Gq-signaling. PTX treatment before acid stimulation inhibited [Ca 2+ ] i levels in both ipsilateral and contralateral DRG cultures (Fig. 6F), indicating that [Ca 2+ ] i increase found in contralateral and ipsilateral DRG cultures was due to Gi-signaling. Accordingly, increased cAMP levels observed in ipsilateral DRG cultures were primarily due to enhancement of Gs-signaling.

TDAG8 activation sensitizes TRPV1 response to capsaicin
We further tested whether TDAG8 activation sensitizes the TRPV1 response to capsaicin. The addition of 10 nM capsaicin to the cells produced a rapid increase in [Ca 2+ ] i levels (0.110 ± 0.016) in TRPV1-expressing cells, but the addition of 5 nM capsaicin (0.001 ± 0.011) or pH 6.4 buffer (0.000 ± 0.006) did not induce a significant response (Fig. 8A, B

Discussion
OGR1 family members were recently found in DRG, and most (75%~82%) are in small-diameter neurons that are responsible for nociception [28,29]. Here, we demonstrated that OGR1 family subtypes are expressed differently in different inflammatory pain states, which suggests that they are involved in neurogenic, short-term and longterm chronic inflammatory pain. The receptor TDAG8 with particularly increased expression after CFA-induced inflammation. Increased mRNA levels were due to the increased number of neurons expressing TDAG8. Given that the number of neurons expressing both TDAG8 and TRPV1 genes was increased as well and that TDAG8 activation sensitized TRPV1 responses to capsaicin, TDAG8 is likely involved in long-term chronic inflammatory pain by sensitizing TRPV1 function.
Capsaicin, carrageenan, or CFA administered locally causes pain-related behaviors, which is accompanied by erythema and edema. Consistent with previous studies Number of TDAG8-expressing neurons increases in both IB 4 -positive and -negative neurons after CFA-induced inflammation  http://www.molecularpain.com/content/5/1/39 [30], capsaicin injection induces neurogenic inflammation. Significant edema developed rapidly within 5 minutes after the administration of capsaicin but was reduced gradually, which indicates an acute inflammatory response. Both carrageenan and CFA injection induced chronic inflammation with higher magnitude and longer duration of edema than did capsaicin injection. Although carrageenan induced edema with a magnitude similar to that produced by CFA, the edema produced by carrageenan peaked in level earlier at 4 hours and remained for a week. The edema induced by CFA injection was fully developed at 24 hours but lasted longer, for 3 weeks. This finding suggests that CFA injection induces long-term chronic inflammation with a peak at 24 hours, whereas carrageenan causes short-term chronic inflammation with a peak at 4 hours.
Hyperalgesia development seems to be well associated with edema development. Previous studies of humans, rats or mice have demonstrated that hyperalgesia induced by capsaicin injection is dose dependent and develops rapidly within 15 minutes after injection [31][32][33][34][35]. Our observations are consistent with these studies: low-dose capsaicin (2.5 μg/paw) administered locally induced pain-related behaviors within 5 minutes; heat hyperalgesia disappeared in 15 minutes after injection, but mechanical hyperalgesia lasted for hours. Carrageenan induced a sub-chronic inflammatory pain. The peak of hyperalgesia was at 4 hours, and hyperalgesia was extended to several days, which is consistent with previous studies [36][37][38]. In the CFA-injection model, both thermal and mechanical hyperalgesia developed fully within 24 hours after injection and extended to three weeks, which is similar to previous observations [37,38].
The OGR1 family were expressed differently in these three inflammatory pain models. After capsaicin injection, GPR4 gene expression was reduced to 2.5~3-fold the basal levels at 24 hours and remained so for three days, whereas G2A transcripts were significantly reduced 72 hours after Accumulation of cAMP by cells exposed to acidic pH Figure 6 (see previous page) Accumulation of cAMP by cells exposed to acidic pH. (A) HEK293T cells were transfected with pIRES-GFP-TDAG8 or pIRES-GFP plasmids for 36 hours. Transfected cells were exposed to pH7. 4   Percentage of neurons that expressed single gene or two genes in total DRG neuron populations with 95% confidence intervals. The number of total cells counted was 150~350. PERI = peripherin injection. Capsaicin-induced edema and hyperalgesia developed within 15 minutes. Down-regulation of GPR4 and G2A is unlikely to have a direct influence on the formation of hyperalgesia or edema. Possibly, reduced levels of GPR4 and G2A mRNA prevent the extension of edema and hyperalgesia. Administration of high-dose capsaicin to adult rats or the application of capsaicin to primary sensory neurons induces a subpopulation of DRG neuron death [39][40][41]. Although capsaicin dosage used in our studies was lower than that used in rats, it is possible that such dosage may have some influence on neurons that expressed GPR4 and G2A genes. GPR4-expressing neurons could be more sensitive to capsaicin administration than neurons expressing G2A. The in situ hybridization experiments had confirmed that the number of GPR4positve neurons was decreased (9% of decrease) after capsaicin injection (data not shown). Capsaicin-injection also induced a slight reduction in the number of PERIpositive neurons. Alternatively, down-regulation of GPR4 and G2A genes is due to functional desensitization.
GPR4, G2A, and TDAG8 transcripts were increased in levels 24 hours after carrageenan injection, but only TDAG8 expression was increased after CFA injection. In both models, the transcripts of the three genes increased in levels at 24 hours when edema and hyperalgesia were already developed. GPR4, G2A, and TDAG8 seem to maintain edema and hyperalgesia rather than induce edema and hyperalgesia. TDAG8 was identified from apoptotic immature thymocytes, where its expression is up-regulated, which suggests that TDAG8 is involved in immune cell development [42]. Although TDAG8 expression is required for the production of cAMP in immune cells, mice lacking TDAG8 have normal immune cell development [43,44]. In endothelial cells, G2A expression blocks NF-kB activation and chemokine expression, thus inhibiting macrophage accumulation [45], which suggests that G2A expression may have a protective role for prevention of early events of inflammation. GPR4 is present in the endothelial cells of blood vessels, and mice lacking GPR4 show vascular abnormalities, which suggests that GPR4 has a role in vascular growth and vascular stability [46]. Vascular stability is important for leukocyte adhesion and function [47]. TDAG8, G2A, and GPR4, were all previously suggested to have pro-inflammatory or anti-inflammatory roles. Whether they have similar roles in nociceptors is unclear. OGR1 is the only receptor whose expression did not change in any inflammatory pain models we tested. Since the level of OGR1 in DRG is the highest among the family members [28], OGR1 is likely the pH sensor for physiological conditions, whereas other family members are responsible for different pathological conditions. However, OGR1 protein and function could still be enhanced after inflammation despite no change in mRNA expression.
TDAG8 is the major proton-sensing GPCR showing increased expression after CFA-induced inflammation. Figure 7 Number of DRG neurons co-expressing TDAG8 and TRPV1 after CFA-induced inflammation. DRG tissues were contiguously sectioned at 6 μm and hybridized with dig-labeled antisense cRNA probes, then co-stained with antibodies against PERI and N52. Each pair of sections was hybridized with two different gene cRNA probes. The histogram shows the percentage of total DRG neurons that expressed both TDAG8 and TRPV1 genes.

Number of DRG neurons co-expressing TDAG8 and TRPV1 after CFA-induced inflammation
Enhanced mRNA levels were due to an increase in the total number of neurons expressing TDAG8. A considerable number of A-fiber neurons began to express TDAG8 after CFA injection. We also found that the number of TRPV1-expressing neurons was increased in medium-and large-diameter neurons after inflammation, consistent with previous results in rats [16]. The total number of Afiber neurons that expressed both TDAG8 and TRPV1 was largely increased after CFA injection. The peripheral largediameter neurons are known to respond to innocuous mechanical stimuli, whereas medium-diameter (A-delta fiber) neurons are classified into two types, both of which respond to intense mechanical stimuli but have differential responsiveness to heat [1,48]. The plasticity of A-fiber neurons is an important mechanism that causes hyperalgesia following inflammation, therefore, increased TDAG8 expression in A-fiber could be involved in mechanical and thermal hyperalgesia by regulating TRPV1 function.
In small-diameter neurons, TDAG8-expressing neurons were increased in number in both IB 4 -positive and -negative nociceptors. IB 4 -positive and -negative nociceptors have not only biochemical and anatomical differences but also distinct functions owing to their physiological properties [49,50]. IB 4 -negative nociceptors are primarily responsible for the initial nociceptive response to proton, capsaicin and noxious heat; whereas the responsiveness of IB 4 -positive neurons to capsaicin is increased after proton stimulation [51]. Breese et al. [17] proposed that peripheral inflammation sensitizes the responses of IB 4 -positive neurons to protons and capsaicin, which is due to enhanced expression and function of TRPV1. The function of IB 4 -positive neurons in inflammation has recently been of interest. Increased TDAG8 expression in IB 4 -positive neurons after peripheral inflammation may increase the sensitivity of IB 4 -positive neurons to proton and capsaicin, through modulating the functions of ASICs or TRPV1. CFA-induced inflammation selectively increases the responsiveness of IB 4 -positive neurons to protons and capsaicin because of the enhanced response of TRPV1 [17]. Our observation could provide an explanation for the enhanced TRPV1 function in IB 4 -positive neurons after inflammation. In IB 4 -positive neurons, TDAG8 responds to extracellular protons leads to cAMP accumulation to activate PKA directly. PKA modulates TRPV1 function by phosphorylation [53,54]. Alternatively, TDAG8-activation may act on other second messenger kinases, such as PKCε through the cAMP-Epac-PKCε path-TDAG8 activation sensitizes TRPV1 response to capsaicin way. The cAMP-Epac-PKCε pathway is important to IB 4positive neurons to mediate inflammatory pain [55].
PKCε not only modulates TRPV1 function but also regulates prolonged response of the chronic inflammatory pain [54,[56][57][58]. Therefore, TDAG8 could be involved in sensitization of TRPV1 function in IB 4 -positive neurons and in the development of the chronic inflammatory pain state through the cAMP-PKCε signaling. Accordingly, peripheral inflammation induces a local increase of protons, leading to TDAG8 activation. TDAG8 activation elevates cAMP levels, activating PKA and PKCε. PKA or PKCε sensitizes TRPV1 or other ion channels to cause hyperalgesia.
Although the number of neurons expressing TDAG8 increased after inflammation, the cAMP accumulation by TDAG8 was not enhanced significantly. We used cAMP assays here was to measure total cAMP levels in a culture population. Given that we found only an 11% increase in TDAG8-expressing neurons after inflammation and that DRG neurons are heterogeneous, our finding of no large increase in cAMP levels in DRG cultures after CFA injection is not surprising. Interestingly, cAMP levels were lower in primary culture after stimulation with pH 6.4 than with pH 7.4. This decrease is pH dependent and raises a possibility that increased cAMP levels observed after CFA injection is due to a reduction of Gi-signaling rather than an enhancement of Gs-signaling. Given that acid-induced [Ca 2+ ] i increase did not change in contralateral or ipsilateral DRG neurons after CFA-injection and such increase was from Gi-signaling, increased cAMP levels after inflammation are more likely due to enhanced Gs-signaling.
Previous studies have reported that proton-sensing GPCRs mediate Gs and Gq signaling pathways [24][25][26][27]. Unexpectedly, we observed pH-dependent Gi responses in DRG culture. Later studies in different tissues or cells have found that OGR1 and GPR4 can mediate more than one type of G-protein signaling [59][60][61]. The OGR1 family subtypes may also mediate Gi signaling. Alternatively, OGR1 family members forming heterodimers for function provides an explanation for why pH-dependent Gi signaling is present in DRG primary culture. About 31%~40% nociceptors express at least two OGR1 family genes [28]. Such a high degree of co-localization may reflect that a heterodimer formation between OGR1 family members is necessary for their function, because a dimerization of GPCRs is now accepted as a functional unit for ligands [62,63]. A GPCR heterodimer might have signaling pathways different from those of a homodimer [62]. A heterodimeric receptor likely switches Gs or Gq signaling to Gi signaling.

Conclusion
This is the first study to systematically explore the expression changes of proton-sensing GPCRs (the OGR1 family) at different times and in different inflammatory pain models (capsaicin, carrageenan, or CFA). Each subtype of the OGR1 family was expressed differently, which may reflect differences between models in duration and magnitude of hyperalgesia. This finding also implies the complexity of the mechanism of inflammatory pain. As demonstrated here, TDAG8 activation can lead to TRPV1 sensitization, and TDAG8 expression increased after CFAinduced inflammation. Our results suggest that high concentrations of protons after inflammation may not only directly activate proton-sensing ion channels (such as ASIC3 and TRPV1) to cause pain but also act on protonsensing GPCRs to regulate the development of hyperalgesia or to enhance the sensitivity of neurons.

Inflammation experiments and tissue collection
Male CD-1 mice (8-12 weeks old) (were bred in animal house in the National Central University, Taiwan) underwent intraplantar injection with 25 μl of saline, CFA (50% in saline), carrageenan (20 mg/ml in saline), or capsaicin (100 μg/ml in saline containing 10% ethanol and 0.5% Tween 80). At 4, 24, and 72 hours after injection, the mice were killed and paw thickness was measured. Lumbar 4-6 DRG ipsilateral and contralateral to injected paws were removed for RNA extraction, with the ganglia from uninjected paws serving as negative controls. L4-5 DRG were frozen for cryosectioning. The animal experimental procedures were approved by the Animal Care and Use Committee at the National Central University, Taiwan.

Behavioral tests
To assess mechanical nociceptive responses, animals were tested for withdrawal thresholds to mechanical stimuli (von Frey filaments, Touch-Test, North Coast Medical, Inc., Morgan Hill, CA) applied to the plantar aspect of the hindpaw. Mice (n = 6 per group) were placed on a wire mesh platform in transparent plexiglas chambers (10 × 8 × 10 cm/chamber), allowed to habituate for 2 hours each day and trained for 3 days before the test. At 4, 24, and 72 hours after mice were injected with inflammatory agents or saline, we applied a series of von Frey fibers (0.4, 0.6, 1.0, 1.4, 2.0 g), in ascending order beginning with the finest fiber, through the wire mesh onto the plantar surface of both hindpaws of mice. A withdrawal response was considered valid only if the hindpaw was removed completely from the platform. If the paw withdrawal response was ambiguous, the application of fibers was repeated.
For each paw, a von Frey fiber was applied 5 times at 5-second intervals. The threshold was when paw withdrawal was observed in more than 3 of 5 applications.
Animals were also tested for thermal nociceptive response to radiant heat applied to the plantar surface of the paw [36]. Mice (n = 6 per group) were allowed to habituate for at least 2 hours in transparent plexiglass chambers (10 × 8 × 10 cm/chamber) on a glass floor before testing. At 4, 24, and 72 hours after mice were injected with inflammatory agents or saline, we stimulated the plantar surface of mouse hindpaws with a light bulb (40% intensity, 305 mW/cm 2 ). The latency to withdrawal of the paw from radiant heat was measured. Measurements from three trials at 1-minute intervals in each paw were averaged. We obtained mean basal withdrawal latencies of 12~15 seconds in uninjected mice. For each assay, three independent preparations were run in quadruplicate. The DRG pool had at least 9 ganglia. The thermal cycling conditions were 95°C for 10 min, followed by 40 cycles of 95°C for 15 s, and 60°C for 1 min. PCR reactions and product detection were carried out in the ABI Prism 7300. The amplified product was detected by measurement of SYBR green I, which was added to the initial experiment mixtures. The threshold cycle (Ct) values obtained from the experiments indicated the fractional cycle numbers at which the amount of amplified target reached a fixed threshold. The Ct values of both the targets and internal reference (mGAPDH) were measured from the same samples, and the expression of the target genes relative to that of mGAPDH was calculated by the comparative Ct method. The specimens were examined by use of a 20× objective in a fluorescence microscope (Zeiss, Axiovert 200, Germany). The digitized images were captured by AxioVixion software. A total of 1,000 neurons from 8 sections were usually counted, and 95% confidence intervals for proportions were estimated.

Primary DRG cultures
Mouse DRGs were collected and placed in pre-warmed serum-free DMEM (Invitrogen). After centrifugation at 970 × g for 2 minutes, ganglia were incubated at 37°C for 1.5 hours with 1 ml of serum-free DMEM containing 0.125% collagenase IA (Sigma) and thoroughly mixed in 15-minute intervals, then centrifuged at 970 × g for 3 minutes. Cell pellets were re-suspended in 1 ml of 0.25% trypsin (Invitrogen) and incubated at 37°C for 15 minutes with mixing in 5-minute intervals to avoid aggregation of neurons. After trypsin digestion, cells underwent sedimentation at 1224 × g for 3 minutes and were washed once by DMEM containing 10% fetal bovine serum (FBS), then once with serum-free medium. Ganglia were re-suspended in 2 ml of serum-free DMEM and then dissociated into single cells by mechanical titration 8 times through flame-polished Pasteur pipettes of decreasing tip diameter. Cell suspension was slowly dropped into 10 ml of serum-free DMEM. After 3~5 minutes, the cell suspension on the top (~10 ml) was collected and centrifuged at 1224 × g for 5 minutes. The cell pellet was suspended and mixed in 400 μl DMEM containing 10% FBS and seeded on 100 μg/ml poly-D-lysine-coated 24-mm coverslips. After incubation at 37°C for 2 hours, cells were supplemented with 1.5 ml DMEM containing 10% FBS and maintained at 37°C for 24 hours before use. For cAMP experiments, one-day DRG cultures obtained from CFAinjected mice were treated with the indicated pH HEPES/ MES buffers (pH 7.4 and 6.4) for 30 minutes at 37°C, then underwent cAMP assay.

Intracellular calcium imaging
Transfected cells were pre-incubated at 37°C with 2.5 μM Fura-2 acetoxymethyl ester (Fura-2-AM, Molecular Probes) for 40 minutes in a HEPES/MES solution (125 mM NaCl, 1 mM KCl, 5 mM CaCl 2 , 1 mM MgCl 2 , 8 mM glucose, 10 mM HEPES and 15 mM MES, pH7.4). This solution was then replaced with a fresh one without Fura-2-AM. Coverslips were assembled into culture wells and supplemented with 300 μl of the HEPES/MES solution (pH 7.4). Cells were observed by use of a Zeiss inverted microscope and illuminated with a xenon lamp to excitation fluorescence, then images were taken with use of a Zeiss Plan-Apo 63× oil-immersion objective lens. A cooled CCD camera (Photometric) was used to detect fluorescence. GFP-positive cells within a field were identified by use of a FITC filter with excitation wavelength 480 nm and emission wavelength 535 nm. In the same field, fura-2 fluorescence was measured by 10 Hz alternating wavelength time scanning, with excitation wavelength 340 and 380 nm and emission wavelength 510 nm. The fluorescence ratio at two excitation wavelengths (340/380 nm, Ca 2+ -bound Fura-2/free Fura-2) was recorded and analyzed. The HEPES/MES buffers (600 μl) were added to the culture wells to obtain the indicated pH values. The pHevoked calcium transients and the number of cells responding to the indicated pH values were recorded. Some experiments involved initial supplementation with 500 μl of the HEPES/MES buffer (pH 7.4) in culture wells, then the addition of 500 μl of the HEPES/MES buffer (pH 7.4) containing twice the final agonist concentrations.

The cAMP assay
Transfected HEK293T cells or primary cultures were preincubated for 15 minutes with serum-free DMEM containing 30 μM of the phosphodiesterase inhibitor RO201724 (Sigma), then stimulated with indicated pH buffers containing 30 μM of RO201724 for 30 minutes at 37°C or 22°C. After stimulation, cells were lysed in ethanol. The lysates were dried and cAMP in dried lysates was quantified by use of the cAMP immunoassay kit (Assay Designs, MI), according to the manufacturer's protocol. Some experiments involved pre-incubation with 100 ng/ ml of pertussis toxin (PTX) for 2-4 hours or U73122 for 15 minutes at 37°C before pH stimulation. All data are referenced to pH at room temperature. To obtain pH at 37°C, 0.05 pH units should be subtracted for HEPES buffers in the range of pH 6.8-5.0 according to our calibration experiments.

Statistical analysis
All data are presented as mean ± SEM. Paired t test was used to compare the paw volume, withdrawal threshold, and latency between inflammatory agent-treated ipsilateral paws and contralateral paws. The t test was used to compare results for control and inflammatory agent-