Synergistic Cytokine Production by ATP and PGE2 via P2X4 and EP3 Receptors in Mouse Bone-Marrow-Derived Mast Cells

ATP is an important intercellular messenger in the extracellular space. In mast cells (MCs), ATP stimulates the ionotropic P2X4 receptor (P2X4R), resulting in enhanced degranulation and exacerbation of acute allergic reactions. In this study, we investigate whether ATP regulates inflammatory cytokine production in MCs. Gene expression was analyzed by quantitative RT-PCR, and cytokine production was measured using ELISA. The stimulation of mouse bone-marrow-derived MCs (BMMCs) with ATP alone had little effect on cytokine secretion. However, the co-stimulation with prostaglandin (PG) E2 resulted in a marked increase in the secretion of various cytokines, such as tumor necrosis factor-α, interleukin (IL)-6, and IL-13, accompanied by an increase in their mRNA levels. The effects of ATP were inhibited by P2X4R antagonists and diminished in BMMCs derived from P2X4R-deficient mice, suggesting that P2X4R mediated the reaction. The effects of PGE2 were mimicked by an EP3 receptor (EP3R) agonist and blocked by an EP3R antagonist. The synergistic cytokine mRNA elevations induced by ATP and PGE2 were blocked by nuclear factor-κB and Ca2+-calcineurin signaling inhibitors. Altogether, these results suggest that combining P2X4R and EP3R signaling enhances acute degranulation and the subsequent cytokine secretion, exacerbating allergic inflammation.


Introduction
Adenosine triphosphate (ATP), an important intracellular energy source constantly supplied by cell metabolism under cellular respiration, is present in high concentrations in the cytosol. However, cells release ATP into the extracellular space as an intercellular mediator to regulate various physiological functions [1]. Extracellular ATP accumulation is recognized by various receptors called P2 receptors. Seven subtypes of ionotropic P2X receptors (P2X1-7) recognize ATP exclusively and open non-selective cation channels [2]. In contrast, eight subtypes of G protein-conjugated P2Y receptors have been identified in humans, including P2Y 1 , Y 2 , Y 4 , Y 6 , and Y [11][12][13][14] receptors, and the ligands are not only ATP but also ADP and uridine nucleotides, such as UTP, UDP, and UDP-glucose [3]. Receptor activation is transmitted into cells by the G protein-dependent signaling [3]. In addition, extracellular ATP is rapidly dephosphorylated by various extracellular ATP-degrading enzymes to adenosine, acting as an agonist to another set of P1 adenosine receptors, such as Gi-coupled A 1 and A 3 receptors, and Gs-coupled A 2A and A 2B receptors [4]. Thus, ATP

Animals
P2X4R-deficient (P2rx4 −/− ) mice were generated by Dr. Yamamoto (University of Tokyo), as described previously [23]. C57BL/6 mice were purchased from SLC Japan (Hamamatsu, Japan). All mice were maintained in the animal facility of the Takasaki University of Health and Welfare under specific pathogen-free conditions. All experiments were performed per the Animal Research Committee of Takasaki University of Health and Welfare regulations (Approval No. 2033, 1 April 2020).

Cell Culture
BMMCs were prepared from the bone marrow of C57BL/6 mice, as described previously [24]. Briefly, bone marrow cells were collected from the femur and cultured in RPMI 1640 growth medium containing 10% fetal bovine serum, 100 units/mL penicillin, 100 µg/mL streptomycin, and 10 ng/mL mouse IL-3. After 2 weeks, recombinant SCF (10 ng/mL) was added to the growth medium and cultured for 2-3 weeks. The culture medium was changed twice per week. After these treatments, almost all (> 95%) cells displayed an MC phenotype, as indicated by CD117 (c-Kit), and FcεRI expression measured using a FACSCant II flow cytometer (BD Biosciences, Tokyo, Japan).

Measurement of Cytokine Secretion
BMMCs were sensitized with 50 ng/mL anti-DNP IgE overnight in RPMI 1640 growth medium. Cells were washed twice with RPMI medium containing 0.1% bovine serum albumin and then stimulated under various conditions for different periods at 37 • C. The reactions were terminated by centrifugation, and the supernatant and cell pellets were subjected to cytokine measurement using ELISA and RNA extraction, respectively.

Quantitative RT-PCR
Total RNA was isolated using the GenElute Mammalian Total RNA miniprep kit. Firststrand cDNA was synthesized using Moloney murine leukemia virus reverse transcriptase with a 6-mer random primer. Quantitative RT-PCR was performed using an SYBR green kit (Takara bio, Tokyo Japan), as described previously [25].

Statistics
All experiments were repeated at least three times, yielding similar results. Data represent the mean ± standard error of the mean (SEM). Statistical analyses were performed using Student's t-test for two-sample comparisons and one-way analysis of variance (ANOVA) with Dunnett's two-tailed test for multiple comparisons. Statistical significance was set at p-value < 0.05.

Effects of ATP Combined with PGE 2 and Ag on Cytokine mRNA Expression and Secretion in BMMCs
First, we examined the effects of ATP on PGE 2 or antigen DNP-HSA-induced mRNA expression of inflammatory cytokines, including IL-1β, IL-6, IL-13, and TNF-α, in BMMCs by quantitative RT-PCR. The BMMCs used in this study hardly expressed the inflammatory cytokine mRNA in the unstimulated condition. However, the co-stimulation of ATP (100 µM) with PGE 2 (1 µM) and low doses of antigen DNP-HSA (10 ng/mL) markedly increased cytokine mRNA expression ( Figure 1). The increased levels were comparable to the expression levels of glyceraldehyde-3-phosphate dehydrogenase (GAPDH), a housekeeping gene used as an internal standard. Next, we examined whether the change in mRNA reflected cytokine secretion. The stimulation of the BMMCs for 3 h with ATP alone had little effect on the production of IL-6, IL-13, and TNF-α; however, co-stimulation with antigen or PGE 2 resulted in a pronounced increase in cytokine secretion. Both the rate and absolute value of the increase were more pronounced when co-stimulated with PGE 2 than with antigen ( Figure 2). In contrast, there was almost no IL-1β secretion, exhibiting increased mRNA levels (data not shown). Therefore, the subsequent experiments dealt with a detailed analysis of the production of IL-6, IL-13, and TNF-α by co-stimulation with ATP and PGE 2 . (1:1000), anti-NF-kB p65 (1:1000), anti-P2X4 (1:2000), and horseradish peroxidase (HRP)linked anti-rabbit IgG (1:10,000). Immunoreactive proteins were detected by enhanced chemiluminescence (GE Healthcare Bio-sciences, Tokyo, Japan) using Image Reader LAS-3000 (FUJIFILM, Tokyo, Japan). The density of bands was quantified using Multi Gauge version 3.0, and the fold change in phosphorylation was calculated from the amount of phosphoprotein relative to total proteins as a loading index.

Statistics
All experiments were repeated at least three times, yielding similar results. Data represent the mean ± standard error of the mean (SEM). Statistical analyses were performed using Student's t-test for two-sample comparisons and one-way analysis of variance (ANOVA) with Dunnett's two-tailed test for multiple comparisons. Statistical significance was set at p-value < 0.05.

Effects of ATP Combined with PGE2 and Ag on Cytokine mRNA Expression and Secretion in BMMCs
First, we examined the effects of ATP on PGE2 or antigen DNP-HSA-induced mRNA expression of inflammatory cytokines, including IL-1β, IL-6, IL-13, and TNF-α, in BMMCs by quantitative RT-PCR. The BMMCs used in this study hardly expressed the inflammatory cytokine mRNA in the unstimulated condition. However, the co-stimulation of ATP (100 μM) with PGE2 (1 μM) and low doses of antigen DNP-HSA (10 ng/mL) markedly increased cytokine mRNA expression ( Figure 1). The increased levels were comparable to the expression levels of glyceraldehyde-3-phosphate dehydrogenase (GAPDH), a housekeeping gene used as an internal standard. Next, we examined whether the change in mRNA reflected cytokine secretion. The stimulation of the BMMCs for 3 h with ATP alone had little effect on the production of IL-6, IL-13, and TNF-α; however, co-stimulation with antigen or PGE2 resulted in a pronounced increase in cytokine secretion. Both the rate and absolute value of the increase were more pronounced when co-stimulated with PGE2 than with antigen ( Figure 2). In contrast, there was almost no IL-1β secretion, exhibiting increased mRNA levels (data not shown). Therefore, the subsequent experiments dealt with a detailed analysis of the production of IL-6, IL-13, and TNF-α by co-stimulation with ATP and PGE2.

Time-Dependent Increase in IL-6 and TNF-α Secretion in Response to ATP, PGE2, and Their Co-Stimulation
Although an increase in cytokine secretion is accompanied by de novo synthesis in response to stimuli, some cytokines, such as TNF-α, are exited as granule content and increased immediately after stimulation. Therefore, the detailed time-course of changes in cytokine secretion was investigated ( Figure 3). ATP stimulation alone did not induce significant cytokine secretion until 24 h. In contrast, PGE2 stimulated IL-13 and TNFα secretion at 3 h after stimulation and IL-6 secretion at 6 h. When co-stimulated with ATP and PGE2, a marked increase in IL-6 and IL-13 secretion was observed 3 h after a 1 h time lag. In contrast, TNFα secretion was increased 1 h after co-stimulation with ATP and PGE2 and reached a peak secretion at 3 h. Based on the above results, the following cytokine secretion was examined 3 h after stimulation, whereas mRNA expression was determined 1 h after stimulation.

Effects of Different Concentrations of ATP and PGE2 on IL-6, IL-13, and TNF-α Secretion
The effects of different concentrations of ATP and PGE2 on IL-6, IL-13, and TNF-α were examined. ATP alone did not release cytokines up to a concentration of 100 μM;

Time-Dependent Increase in IL-6 and TNF-α Secretion in Response to ATP, PGE 2 , and Their Co-Stimulation
Although an increase in cytokine secretion is accompanied by de novo synthesis in response to stimuli, some cytokines, such as TNF-α, are exited as granule content and increased immediately after stimulation. Therefore, the detailed time-course of changes in cytokine secretion was investigated ( Figure 3). ATP stimulation alone did not induce significant cytokine secretion until 24 h. In contrast, PGE 2 stimulated IL-13 and TNFα secretion at 3 h after stimulation and IL-6 secretion at 6 h. When co-stimulated with ATP and PGE 2 , a marked increase in IL-6 and IL-13 secretion was observed 3 h after a 1 h time lag. In contrast, TNFα secretion was increased 1 h after co-stimulation with ATP and PGE 2 and reached a peak secretion at 3 h. Based on the above results, the following cytokine secretion was examined 3 h after stimulation, whereas mRNA expression was determined 1 h after stimulation. normalized to GAPDH mRNA levels. Values are shown as mean ± SEM (n = 3). *p < 0.05, **p < 0.

Time-Dependent Increase in IL-6 and TNF-α Secretion in Response to ATP, PGE2, and Their Co-Stimulation
Although an increase in cytokine secretion is accompanied by de novo synthesis response to stimuli, some cytokines, such as TNF-α, are exited as granule content a increased immediately after stimulation. Therefore, the detailed time-course of changes cytokine secretion was investigated ( Figure 3). ATP stimulation alone did not induce s nificant cytokine secretion until 24 h. In contrast, PGE2 stimulated IL-13 and TNFα sec tion at 3 h after stimulation and IL-6 secretion at 6 h. When co-stimulated with ATP a PGE2, a marked increase in IL-6 and IL-13 secretion was observed 3 h after a 1 h time l In contrast, TNFα secretion was increased 1 h after co-stimulation with ATP and PG and reached a peak secretion at 3 h. Based on the above results, the following cytok secretion was examined 3 h after stimulation, whereas mRNA expression was determin 1 h after stimulation.

Effects of Different Concentrations of ATP and PGE2 on IL-6, IL-13, and TNF-α Secretion
The effects of different concentrations of ATP and PGE2 on IL-6, IL-13, and TNF were examined. ATP alone did not release cytokines up to a concentration of 100 μ

Effects of Different Concentrations of ATP and PGE 2 on IL-6, IL-13, and TNF-α Secretion
The effects of different concentrations of ATP and PGE 2 on IL-6, IL-13, and TNF-α were examined. ATP alone did not release cytokines up to a concentration of 100 µM; however, the secretion of IL-6, IL-13, and TNF-α increased slightly but constantly at 1 mM. In the presence of PGE 2 (1 µM), ATP increased the secretion of IL-6, IL-13, and TNFα in a concentration-dependent manner from 1 to 100 µM, and the enhancement was decreased at 1 mM. In contrast, PGE 2 elicited IL-6, IL-13, and TNFα secretion at 0.1 to 1 µM even in the Cells 2022, 11, 616 6 of 14 absence of ATP; however, its effects were markedly increased in the presence of 100 µM ATP ( Figure 4). Based on these results, 100 µM ATP and 0.1 µM PGE 2 were selected as optimal concentrations for assessing the synergistic secretion of cytokines in the subsequent experiments.
however, the secretion of IL-6, IL-13, and TNF-α increased slightly but constantly at 1 mM. In the presence of PGE2 (1 μM), ATP increased the secretion of IL-6, IL-13, and TNFα in a concentration-dependent manner from 1 to 100 μM, and the enhancement was decreased at 1 mM. In contrast, PGE2 elicited IL-6, IL-13, and TNFα secretion at 0.1 to 1μM even in the absence of ATP; however, its effects were markedly increased in the presence of 100 μM ATP (Figure 4). Based on these results, 100 μM ATP and 0.1 μM PGE2 were selected as optimal concentrations for assessing the synergistic secretion of cytokines in the subsequent experiments.

Effects of Different Purinergic Receptor Agonists Combined with PGE2 on IL-6, IL-13, and TNF-α mRNA Expression
The BMMCs used in our studies expressed multiple P2 receptors stimulated by ATP, including P2X1, P2X4, P2X7, P2Y1, P2Y2, and P2Y13 [24]. To examine the receptor subtype involved in ATP enhancement of PGE2-induced cytokine secretion, the effects of various nucleotide agonists were examined. The agonists used were P2X1 agonist αβ-methylene ATP (αβmeATP), P2Y1 and P2Y13 agonist ADP, and P2Y2 agonist UTP. Among them, ADP enhanced the secretion of IL-6 and IL-13, but was much weaker than ATP. αβmeATP and UTP did not increase the effects of PGE2 ( Figure 5).

Effects of Different Purinergic Receptor Agonists Combined with PGE 2 on IL-6, IL-13, and TNF-α mRNA Expression
The BMMCs used in our studies expressed multiple P2 receptors stimulated by ATP, including P2X1, P2X4, P2X7, P2Y 1 , P2Y 2 , and P2Y 13 [24]. To examine the receptor subtype involved in ATP enhancement of PGE 2 -induced cytokine secretion, the effects of various nucleotide agonists were examined. The agonists used were P2X1 agonist α,β-methylene ATP (αβmeATP), P2Y 1 and P2Y 13 agonist ADP, and P2Y 2 agonist UTP. Among them, ADP enhanced the secretion of IL-6 and IL-13, but was much weaker than ATP. αβmeATP and UTP did not increase the effects of PGE 2 ( Figure 5). as optimal concentrations for assessing the synergistic secretion of cytokines in the subsequent experiments.

Effects of Different Purinergic Receptor Agonists Combined with PGE2 on IL-6, IL-13, and TNF-α mRNA Expression
The BMMCs used in our studies expressed multiple P2 receptors stimulated by ATP, including P2X1, P2X4, P2X7, P2Y1, P2Y2, and P2Y13 [24]. To examine the receptor subtype involved in ATP enhancement of PGE2-induced cytokine secretion, the effects of various nucleotide agonists were examined. The agonists used were P2X1 agonist αβ-methylene ATP (αβmeATP), P2Y1 and P2Y13 agonist ADP, and P2Y2 agonist UTP. Among them, ADP enhanced the secretion of IL-6 and IL-13, but was much weaker than ATP. αβmeATP and UTP did not increase the effects of PGE2 ( Figure 5).

Role of P2X4 Receptor in Enhancing Cytokine mRNA Expression and Secretion by Co-Stimulation of ATP and PGE 2
The responses to P2 agonists shown in Figure 5 were similar to the effects on antigen-and PGE 2 -induced MC degranulation via P2X4R. Therefore, the effect of NP-1815-PX, a P2X4R antagonist [26], on cytokine secretion was investigated. NP-1815-PX did not affect cytokine secretion by PGE 2 , but significantly suppressed the augmented response induced by ATP ( Figure 6). To further clarify the involvement of P2X4R, we examined the BMMCs prepared from P2rx4 −/− mice. No increase in cytokine mRNA expression and cytokine secretion was observed in P2rx4 −/− BMMCs after co-stimulation with ATP and PGE 2 (Figure 7). The responses to P2 agonists shown in Figure 5 were similar to the effects on antige and PGE2-induced MC degranulation via P2X4R. Therefore, the effect of NP-1815-PX P2X4R antagonist [26], on cytokine secretion was investigated. NP-1815-PX did not aff cytokine secretion by PGE2, but significantly suppressed the augmented response induc by ATP (Figure 6). To further clarify the involvement of P2X4R, we examined the BMM prepared from P2rx4 −/− mice. No increase in cytokine mRNA expression and cytokine cretion was observed in P2rx4 −/− BMMCs after co-stimulation with ATP and PGE2 (Figu 7).

Stimulation of ATP and PGE2
The responses to P2 agonists shown in Figure 5 were similar to the effects on antigenand PGE2-induced MC degranulation via P2X4R. Therefore, the effect of NP-1815-PX, a P2X4R antagonist [26], on cytokine secretion was investigated. NP-1815-PX did not affect cytokine secretion by PGE2, but significantly suppressed the augmented response induced by ATP (Figure 6). To further clarify the involvement of P2X4R, we examined the BMMCs prepared from P2rx4 −/− mice. No increase in cytokine mRNA expression and cytokine secretion was observed in P2rx4 −/− BMMCs after co-stimulation with ATP and PGE2 ( Figure  7).

Role of EP 3 R in Enhancing Cytokine mRNA Expression and Secretion by Co-Stimulation of ATP and PGE 2
We examined the receptors that mediate PGE 2 action. The synergistic increase in cytokine secretion by the co-stimulation of PGE 2 and ATP was reproduced only with the EP 3 R agonist ONO-AE-248, and EP 1 , EP 2 , and EP 4 receptor agonists showed no effect ( Figure 8A-C). Conversely, the synergistic increase in cytokine secretion by the co-stimulation of PGE 2 and ATP was suppressed by the EP 3 R antagonist ONO-AE3-240, but not by EP 1 , EP 2 , and EP 4 receptor antagonists ( Figure 8D-F). Consistently, the BMMCs predominantly expressed EP 3 R mRNA, which was unaffected by P2X4R deficiency (Supplementary Figure S1).

Role of EP3R in Enhancing Cytokine mRNA Expression and Secretion by Co-Stimulation of ATP and PGE2
We examined the receptors that mediate PGE2 action. The synergistic increase in cytokine secretion by the co-stimulation of PGE2 and ATP was reproduced only with the EP3R agonist ONO-AE-248, and EP1, EP2, and EP4 receptor agonists showed no effect (Figure 8A-C). Conversely, the synergistic increase in cytokine secretion by the co-stimulation of PGE2 and ATP was suppressed by the EP3R antagonist ONO-AE3-240, but not by EP1, EP2, and EP4 receptor antagonists ( Figure 8D-F). Consistently, the BMMCs predominantly expressed EP3R mRNA, which was unaffected by P2X4R deficiency (Supplementary Figure S1).

Effects of ATP and PGE2 on ERK/Akt/NF-κB Signaling Pathway
The stimulation of the BMMCs with PGE2 induced the rapid phosphorylation of both ERK1/2 and Akt within 5 min, followed by a decrease to a sustained level at 10 and 20 min. In contrast, ATP had little effect on ERK1/2 and Akt phosphorylation at 5 min, but slightly increased after 10 and 20 min. The co-stimulation of ATP and PGE2 induced a rapid and sustained increase in ERK1/2 and Akt phosphorylation at 10 and 20 min ( Figure  9A,C,D). These co-stimulating effects of ATP and PGE2 on ERK1/2 and Akt phosphorylation were absent in P2rx4 −/− BMMCs ( Figure 9B,C,D). Akt is known to activate nuclear factor-κB (NF-κB) [27], a transcription factor for various inflammatory cytokines. ATP or PGE2 alone only slightly increased the level of phosphorylated NF-κB p65, an activated form of NF-κB. The co-stimulation with ATP and PGE2 markedly increased the levels of phosphorylated NF-κB ( Figure 9A,E). The synergistic phosphorylation of NF-κB by ATP and PGE2 was absent in P2rx4 -/-BMMCs ( Figure 9B,E). In another set of analysis, we examined the role of P2X4R in the synergistic increase in the phosphorylation of ERK1/2, Akt, and NF-kB by co-stimulation with ATP and PGE2. P2X4R deficiency abolished the increased phosphorylation of ERK, Akt, and NF-κB (Supplementary Figure S2).

Effects of ATP and PGE 2 on ERK/Akt/NF-κB Signaling Pathway
The stimulation of the BMMCs with PGE 2 induced the rapid phosphorylation of both ERK1/2 and Akt within 5 min, followed by a decrease to a sustained level at 10 and 20 min. In contrast, ATP had little effect on ERK1/2 and Akt phosphorylation at 5 min, but slightly increased after 10 and 20 min. The co-stimulation of ATP and PGE 2 induced a rapid and sustained increase in ERK1/2 and Akt phosphorylation at 10 and 20 min ( Figure 9A,C,D). These co-stimulating effects of ATP and PGE 2 on ERK1/2 and Akt phosphorylation were absent in P2rx4 −/− BMMCs ( Figure 9B-D). Akt is known to activate nuclear factor-κB (NF-κB) [27], a transcription factor for various inflammatory cytokines. ATP or PGE 2 alone only slightly increased the level of phosphorylated NF-κB p65, an activated form of NF-κB. The co-stimulation with ATP and PGE 2 markedly increased the levels of phosphorylated NF-κB ( Figure 9A,E). The synergistic phosphorylation of NF-κB by ATP and PGE 2 was absent in P2rx4 −/− BMMCs ( Figure 9B,E). In another set of analysis, we examined the role of P2X4R in the synergistic increase in the phosphorylation of ERK1/2, Akt, and NF-kB by co-stimulation with ATP and PGE 2 . P2X4R deficiency abolished the increased phosphorylation of ERK, Akt, and NF-κB (Supplementary Figure S2). . The effect of co-stimulation of PGE2 and ATP on ERK 1/2, Akt, and NF-κB phosphory lation is significantly different from that of PGE2 alone. *p < 0.05, **p < 0.01.

Effects of Signal Transduction Pathway Inhibitors on the Enhanced Cytokine mRNA Expression by Co-Stimulation with ATP and PGE2
The effects of various inhibitors against ERK, p38MAPK, phosphatidylinositol 3-ki nase (PI3K), calcineurin, and NF-κB were examined. The elevation of cytokine mRNA lev els by the co-stimulation of ATP and PGE2 was attenuated by the ERK inhibitor U0126 PI3K inhibitor wortmannin, NF-κB activation inhibitor NFκBI, calcineurin inhibitor cyclo sporine A, intracellular Ca 2+ chelator BAPTA-AM and the steroidal anti-inflammatory drug dexamethasone. In contrast, the p38MAPK inhibitor SB203580 did not suppress cy tokine mRNA expression ( Figure 10) Figure 9. Effects of PGE 2 , ATP, and co-stimulation of PGE 2 and ATP on ERK 1/2, Akt and NF-κB phosphorylation in the BMMCs. (A,B) The BMMCs prepared from WT (A) and P2rx4 −/− mice (B) were stimulated with PGE 2 (0.1 µM), ATP (100 µM), and PGE 2 and ATP for 5, 10, and 20 min. Cell lysates were subjected to Western blot analysis for phospho-Akt and total Akt (upper), phospho-ERK 1/2 and total-ERK 1/2 (middle), or phospho-NF-κB and total-NF-κB (lower). Blots are representative of three independent experiments. (C) The densitometry value of protein bands of phospho-Akt (C), phospho-ERK 1/2 (D), and phospho-NF-κB (E) are shown as relative intensities, with the results obtained with unstimulated cells (Time 0) designated as 1. Data are shown as mean ± SEM (n = 3). The effect of co-stimulation of PGE 2 and ATP on ERK 1/2, Akt, and NF-κB phosphorylation is significantly different from that of PGE 2 alone. * p < 0.05, ** p < 0.01.

Effects of Signal Transduction Pathway Inhibitors on the Enhanced Cytokine mRNA Expression by Co-Stimulation with ATP and PGE 2
The effects of various inhibitors against ERK, p38MAPK, phosphatidylinositol 3-kinase (PI3K), calcineurin, and NF-κB were examined. The elevation of cytokine mRNA levels by the co-stimulation of ATP and PGE 2 was attenuated by the ERK inhibitor U0126, PI3K inhibitor wortmannin, NF-κB activation inhibitor NFκBI, calcineurin inhibitor cyclosporine A, intracellular Ca 2+ chelator BAPTA-AM and the steroidal anti-inflammatory drug dexamethasone. In contrast, the p38MAPK inhibitor SB203580 did not suppress cytokine mRNA expression ( Figure 10 Values are shown as mean ± SEM (n = 6). *p < 0.05, **p < 0.01 vs. the response to ATP and PGE2 and + p < 0.05, ++ p < 0.01 vs. the response to PGE2.

Discussion
In a previous study, we reported that ATP augments IgE-dependent and IgE-independent MC degranulation via P2X4R activation [20,21]. MCs elicit an immediate allergic reaction by releasing chemical mediators stored in the granules; however, in the delayed phase, MCs subsequently secrete diverse inflammatory cytokines and chemokines, causing the migration and activation of other immune cells, resulting in an allergic inflammation [22]. The present study showed that extracellular ATP markedly increased the gene expression and secretion of inflammatory cytokines IL-6 and TNF-α and Th-2 cytokine IL-13 by co-stimulation with antigen and PGE2. Pharmacological analysis using agonists and antagonists revealed that the effects of ATP and PGE2 were mediated by P2X4R and EP3R, respectively, as similar to the effects on degranulation responses [20,21]. The involvement of P2X4R in the ATP-induced effect was confirmed using BMMCs prepared from P2rx4 -/mice. Regarding enhancing the degranulation reaction by ATP, the maximum response was not significantly different between the co-stimulation with antigen and with PGE2 [20,21]. However, the maximum cytokine secretion was much larger in the co-stimulation of ATP with PGE2 than that with the antigen. FcεRI activation with antigen is triggered by tyrosine phosphorylation of the immunoreceptor tyrosine-based activation motif (ITAM) in the FcεRI β subunit. It recruits and activates Lyn kinase resulting in Syk kinase activation, propagating phosphorylation-dependent signals to activate various effector proteins, including phospholipase Cγ [28]. Although FcεRI β subunit is indispensable for FcεRI signaling, the ITAM in the β subunit has an inhibitory regulatory role in MC cytokine production [29], explaining the low cytokine secretion combined with ATP and antigen. In this study, we focused on increased cytokine secretion by co-stimulation with PGE2 and ATP.
Although various ATP receptors are expressed in MCs, the receptor that mediates the ATP-induced enhancement of cytokine secretion in the presence of PGE2 was suggested to be the P2X4 receptor, as described above. No increase in cytokine secretion with ATP in the BMMCs prepared from P2rx4 -/mice and the failure of αβmeATP and UTP to stimulate cytokine secretion indicates that P2X1 and P2Y2 receptors expressed abundantly in the BMMCs [24] do not affect cytokine secretion. Among the P2 receptor agonists tested, ADP slightly but significantly enhanced the gene expression of IL-6 and IL-13. Unlike ATP, ADP did not enhance TNF-αexpression; hence, it was not the result of P2X4R stimulation with low affinity but of stimulating the Gq-coupled P2Y1 receptor that recognizes ADP. This result cannot infer that Gq-mediated signals enhance cytokine production, and

Discussion
In a previous study, we reported that ATP augments IgE-dependent and IgE-independent MC degranulation via P2X4R activation [20,21]. MCs elicit an immediate allergic reaction by releasing chemical mediators stored in the granules; however, in the delayed phase, MCs subsequently secrete diverse inflammatory cytokines and chemokines, causing the migration and activation of other immune cells, resulting in an allergic inflammation [22]. The present study showed that extracellular ATP markedly increased the gene expression and secretion of inflammatory cytokines IL-6 and TNF-α and Th-2 cytokine IL-13 by co-stimulation with antigen and PGE 2 . Pharmacological analysis using agonists and antagonists revealed that the effects of ATP and PGE 2 were mediated by P2X4R and EP 3 R, respectively, as similar to the effects on degranulation responses [20,21]. The involvement of P2X4R in the ATP-induced effect was confirmed using BMMCs prepared from P2rx4 −/− mice. Regarding enhancing the degranulation reaction by ATP, the maximum response was not significantly different between the co-stimulation with antigen and with PGE 2 [20,21]. However, the maximum cytokine secretion was much larger in the co-stimulation of ATP with PGE 2 than that with the antigen. FcεRI activation with antigen is triggered by tyrosine phosphorylation of the immunoreceptor tyrosine-based activation motif (ITAM) in the FcεRI β subunit. It recruits and activates Lyn kinase resulting in Syk kinase activation, propagating phosphorylation-dependent signals to activate various effector proteins, including phospholipase Cγ [28]. Although FcεRI β subunit is indispensable for FcεRI signaling, the ITAM in the β subunit has an inhibitory regulatory role in MC cytokine production [29], explaining the low cytokine secretion combined with ATP and antigen. In this study, we focused on increased cytokine secretion by co-stimulation with PGE 2 and ATP.
Although various ATP receptors are expressed in MCs, the receptor that mediates the ATP-induced enhancement of cytokine secretion in the presence of PGE 2 was suggested to be the P2X4 receptor, as described above. No increase in cytokine secretion with ATP in the BMMCs prepared from P2rx4 −/− mice and the failure of αβmeATP and UTP to stimulate cytokine secretion indicates that P2X1 and P2Y 2 receptors expressed abundantly in the BMMCs [24] do not affect cytokine secretion. Among the P2 receptor agonists tested, ADP slightly but significantly enhanced the gene expression of IL-6 and IL-13. Unlike ATP, ADP did not enhance TNF-α expression; hence, it was not the result of P2X4R stimulation with low affinity but of stimulating the Gq-coupled P2Y 1 receptor that recognizes ADP. This result cannot infer that Gq-mediated signals enhance cytokine production, and it is necessary to activate other G proteins, such as G 12/13 , in addition to Gq [30]. Further studies are needed to validate this hypothesis.
The stimulation of the P2X4 receptor, an ion channel-type receptor, activates the non-selective cation channel, leading to Ca 2 + influx and depolarization of the membrane potential [31]. It is interesting to understand such a stimulating mechanism affecting the gene expression of inflammatory cytokines. The effect on cytokine production was markedly different among the P2X receptor subtypes expressed in BMMCs. For example, P2X1 receptor stimulation with αβmeATP has no effect, whereas P2X7 receptor stimulation with a high ATP concentration (~1 mM) alone slightly, but significantly stimulated IL-6, IL-13, and TNF-α secretion. However, unlike the P2X4R-mediated effect, activating the P2X7 receptor did not enhance PGE 2 -induced cytokine secretion, but rather inhibited P2X4R-mediated effects. In addition to ion channel function, P2X7 receptor stimulation dilate the channel pore, allowing it to pass with high molecular weight substances, affecting the MC membrane function [32]. The three ionotropic P2X1, P2X4, and P2X7 receptors are functionally expressed in human MCs [33]. The difference in the effects of these P2X receptors on cytokine secretion may be useful in elucidating the unique mechanism of P2X4R action.
Increased cytokine secretion by P2X4R stimulation combined with PGE 2 was accompanied by increased mRNA levels. The time course of cytokine secretion showed an increase in gene expression before the increase in the secretion of all three cytokines examined. Cytokine secretion was not remarkable 1 h after stimulation when a marked mRNA increase was detected. IL-6, IL-13, and TNF-α secretion increased rapidly during 1-3 h after stimulation, and no continuous increase was observed thereafter, suggesting that P2X4R signaling enhances the transcription of inflammatory cytokines. NF-κB is involved in the transcription of various inflammatory cytokines [22,34]. NF-κB is present in the cytosol as a complex with IκB. When IκB is phosphorylated by IKK, ubiquitination of IκB progresses its degradation by the proteasome, and the dissociated NF-κB translocate into the nucleus to stimulate inflammatory cytokine gene expression [35,36]. Among EP 3 R downstream signals, ERK1/2 and PI3K-Akt activation participates in NF-κB signaling pathway [26]. PGE 2 rapidly increased ERK1/2 and Akt phosphorylation, followed by a decrease to low steady levels. In addition, ATP increased phosphorylation of ERK1/2 and Akt in a timedependent manner, but this action remained in P2rx4 −/− BMMCs, suggesting effects other than P2X4R. However, our results showed that ATP did not affect PGE 2 -induced ERK1/2 and Akt phosphorylation in the early period (5 min), but significantly increased them at later periods (10-20 min). Importantly, this enhancement of PGE 2 -induced ERK1/2 and Akt phosphorylation by ATP was absent in P2rx4 −/− BMMCs. Furthermore, the PI3K inhibitor WMN, the ERK inhibitor U0126, and the NF-κB inhibitor attenuated the upregulation of inflammatory cytokine expression by ATP in the presence of PGE 2 . These results suggest that P2X4R stimulation enhances the ERK1/2 and PI3K/Akt/NF-κB signaling pathway and increases inflammatory cytokine gene expression. Consistent with this hypothesis, the steroidal anti-inflammatory dexamethasone, which potently inhibits the NF-κB pathway [37], also inhibited inflammatory cytokine mRNA expression induced by ATP and PGE 2 . More direct evidence was observed with the change in phosphorylated NF-kB p65, a form of activated NF-κB. Namely, ATP or PGE 2 alone induced only a slight increase in phosphorylated NF-κB p65, but their combination markedly increased the phosphorylated NF-κB p65. In addition, this effect was absent in P2rx4 −/− BMMCs. The phosphorylation of NF-κB p65 by ATP and PGE 2 was rapidly induced within 5 min and returned to basal levels in 20 min. These changes are consistent with the rapid induction of cytokine mRNAs. This also suggests that sustained elevations of ERK and Akt after 10 min are not responsible for NF-κB phosphorylation but may support NF-κB-dependent cytokine gene transcription.
In addition to NF-κB-dependent signaling, a recent study reported by Jordan et al. [38] demonstrated that Ca 2+ -calcineurin signaling is involved in cytokine secretion induced by co-stimulation with ATP and IL-33. In this study, we also showed that the intracellular Ca 2+ chelating agent BAPTA-AM and the calcineurin inhibitor cyclosporine A inhibited inflammatory cytokine mRNA expression induced by ATP and PGE 2 . P2X4R-mediated Ca 2+ influx may stimulate calcineurin-dependent NFAT signaling, resulting in increased inflammatory cytokine gene transcription. Since P2X1 and P2X7 receptors, which input signals qualitatively similar to P2X4R, did not show interplay with EP 3 R signaling, P2X4R may transmit unspecified and unique signals, which is considered an important issue that needs to be addressed.

Conclusions
The present study revealed that extracellular ATP enhanced inflammatory cytokine secretion from stimulated MCs via P2X4R activation. This enhancement was more pronounced with co-stimulation with PGE 2 than with antigen. Since both PGE 2 and ATP are commonly accumulated in the inflamed microenvironment, these mediators should affect MC function in various physiological and pathophysiological situations. The cross-talk between P2X4R and EP 3 R signals upregulates an acute degranulation response and following inflammatory cytokine secretion, suggesting that these receptors may be potential therapeutic targets for diseases caused by MCs.

Supplementary Materials:
The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/cells11040616/s1, Figure S1: Expression of mRNAs for EP receptor subtypes in BMMC obtained from wild type and P2X4 receptor deficient mice. Figure S2: Effects of co-stimulation of ATP and PGE 2 on the phosphorylation of ERK1/2, Akt, and NF-κB.