Leukotriene D₄ and prostaglandin E₂ signals synergize and potentiate vascular inflammation in a mast cell–dependent manner through cysteinyl leukotriene receptor 1 and E-prostanoid receptor 3

Abstract

Background

Although arachidonic acid metabolites, cysteinyl leukotrienes (cys-LTs; leukotriene [LT] C₄, LTD₄, and LTE₄), and prostaglandin (PG) E₂ are generated at the site of inflammation, it is not known whether crosstalk exists between these 2 classes of inflammatory mediators.

Objective

We sought to determine the role of LTD₄-PGE₂ crosstalk in inducing vascular inflammation in vivo, identify effector cells, and ascertain specific receptors and pathways involved in vitro.

Methods

Vascular (ear) inflammation was assessed by injecting agonists into mouse ears, followed by measuring ear thickness and histology, calcium influx with Fura-2, phosphorylation and expression of signaling molecules by means of immunoblotting, PGD₂ and macrophage inflammatory protein 1β generation by using ELISA, and expression of transcripts by using RT-PCR. Candidate receptors and signaling molecules were identified by using antagonists and inhibitors and confirmed by using small interfering RNA.

Results

LTD₄ plus PGE₂ potentiated vascular permeability and edema, gearing the system toward proinflammation in wild-type mice but not in Kit^W-sh mice. Furthermore, LTD₄ plus PGE₂, through cysteinyl leukotriene receptor 1 (CysLT₁R) and E-prostanoid receptor (EP) 3, enhanced extracellular signal-regulated kinase (Erk) and c-fos phosphorylation, inflammatory gene expression, macrophage inflammatory protein 1β secretion, COX-2 upregulation, and PGD₂ generation in mast cells. Additionally, we uncovered that this synergism is mediated through Gi, protein kinase G, and Erk signaling. LTD₄ plus PGE₂–potentiated effects are partially sensitive to CysLT₁R or EP₃ antagonists but completely abolished by simultaneous treatment both in vitro and in vivo.

Conclusions

Our results unravel a unique LTD₄-PGE₂ interaction affecting mast cells through CysLT₁R and EP₃ involving Gi, protein kinase G, and Erk and contributing to vascular inflammation in vivo. Furthermore, current results also suggest an advantage of targeting both CysLT₁R and EP₃ in attenuating inflammation.

Mast cells (MCs) are recognized as critical components of our immune system. They are vital in the initiation and amplification of acute inflammatory responses and play an important role in triggering asthma exacerbations through the elaboration of several soluble inflammatory mediators.^1,2 MCs reside in connective tissues and are located in close proximity to the blood vessels. Activation of MCs stimulate the formation of leukotrienes (LTs) and prostaglandins (PGs), both of which initiate vascular changes.³ Kit^W-sh/W-sh mice have the W-sash (W^sh) inversion mutation and remarkable deficiency in MCs, providing a great model system to analyze MC function in vivo.⁴ Cysteinyl leukotrienes (cys-LTs; LTC₄, LTD₄, and LTE₄) are arachidonic acid derivatives generated by MCs, eosinophils, basophils, and macrophages⁵ through the action of 5-lipoxygenase enzyme. All PGs are derived from PGH₂ and generated through arachidonic acid through the action of PGH synthase (also known as COX). MCs express both COX-1 and COX-2. COX-2 is upregulated by inflammatory stimuli driving PGD₂ generation under inflammatory conditions. In MCs PGH₂ derived from both COX-1 and COX-2 is converted to PGD₂ by a terminal hematopoietic PGD₂ synthase.⁶ Although not a product of MCs, PGE₂, a metabolite of PGH₂ through the action of PGE₂ synthase,⁷ is the most ubiquitous PG, with prominent and complex functions in inflammation, asthma, and allergic diseases. Remarkably, MCs not only generate cys-LTs but also express corresponding receptors and respond to them.⁸ Two known G protein–coupled receptors, termed cysteinyl leukotriene receptor 1 (CysLT₁R) and cysteinyl leukotriene receptor 2 (CysLT₂R), specifically recognize cys-LTs and mediate their biologic functions. CysLT₁R binds LTD₄ with higher affinity than LTC₄, whereas CysLT₂R has equal affinity for LTD₄ and LTC₄.⁵ GPR17, another cys-LT receptor, has been identified and is expressed primarily in the brain,⁹ and GPR99 has been recently identified as a cys-LT receptor with a preference for LTE₄.¹⁰ Mice lacking LTC₄ synthase have reduced numbers of MCs in the airway mucosa after sensitization and challenge by allergen,¹¹ suggesting the prominence of cys-LTs in MC function. We have previously demonstrated that stimulation of human cord blood–derived mast cells (hMCs), LAD2 cells, or both with LTD₄ potently induces calcium flux and cytokine generation⁸ through CysLT₁R. Additionally, MC proliferative and inflammatory responses are modulated by LTD₄ and stem cell factor (SCF) signaling interactions.¹² PGs have also been shown to elicit vasodilation and an increase in blood flow. Among PGs, PGE₂ is the most abundantly synthesized PG at the inflammation site and is regarded as an important regulator of inflammation.¹³ The decisive effect of PGE₂ is the outcome of specific E-prostanoid receptor (EP) 1 to 4 activation through which the signal is transduced.¹⁴ EP₁ is coupled to intracellular calcium mobilization through Gq; however, EP₂ and EP₄ are coupled to stimulation of adenylyl cyclase through Gs, and EP₃ is coupled to the inhibition of adenylyl cyclase through Gi. Different splice variants are generated by means of alternative splicing of the C-terminal tail of the EP₃ receptor and can couple to different signal transduction pathways. Eight human EP₃ isoforms are known thus far, which are identical, except for their carboxyl termini.¹⁴

The interactions among various mediator systems that participate in inflammatory responses are complex, and it is difficult to define the unique contribution of any single element. In the current study we show that LTD₄ and PGE₂ synergistically potentiate peripheral inflammation in vivo and MC activation in vitro through CysLT₁R-, EP₃-, Gi-, protein kinase (PK) G–, and extracellular signal-regulated kinase (Erk)–dependent pathways. Furthermore, our results indicate that blocking EP₃ together with CysLT₁R could be a better therapeutic target to control inflammation.

METHODS

Animals

Six- to 8-week-old BALB/c mice, C57BL/6 mice, and Kit^W-sh mice (W-sh) were obtained from Jackson Laboratories and maintained at the Comparative Medicine Unit, Northeast Ohio Medical University. All animal experiments were done in accordance with standard guidelines, as approved by the Animal Care and Use Committee of Northeast Ohio Medical University.

Reagents

LTD₄, PGE₂, MK571, BayCysLT₂, iloprost, butaprost, sulprostone, L-798, ONO-871, L-161, and PGD₂ ELISA kits were purchased from Cayman Chemicals (Ann Arbor, Mich). KT5823, PD98059, pertussis toxin (PTX), H7, GF109203X, Rp-cAMPS, and H89 inhibitors were from Tocris Bioscience (Minneapolis, Minn). Fura-2 AM was from Molecular Probes (Eugene, Ore), phospho-specific antibodies were from Cell Signaling Technology (Danvers, Mass), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody was from Fitzgerald (Acton, Mass). All secondary antibodies were obtained from Jackson ImmunoResearch (West Grove, Pa). Nonspecific small interfering RNA (siRNA) and isoform-specific siRNAs for CysLT₁R, EP₃, and PKG were obtained from Dharmacon (Lafayette, Colo), and the macrophage inflammatory protein 1β (MIP-1β) ELISA kit was from R&D Systems (Minneapolis, Minn). Cytokines for hMC cultures were obtained from PeproTech (Rocky Hill, NJ).

Intradermal injection of agonists and assessment of ear edema

Mice anesthetized with ketamine/xylazine received intradermal injections of 0.5 μmol/L LTD₄, PGE₂, and LTD₄ plus PGE₂ (in a 10-μL volume) in the right ear and 10 μL of saline in the left ear in the presence or absence of MK571, L-798, or both. At 0, 30, 60, 120, 240, and 300 minutes after the intradermal injection, ear thickness was measured with a caliper. Mice were killed 60 minutes after the indicated treatment, ear tissues were fixed in 4% paraformaldehyde and embedded in paraffin, and 4-μm-thick sections were cut and stained for hematoxylin and eosin and toluidine blue (to detect MCs). Total (toluidine blue–positive cells that are compact) and degranulated MCs (toluidine-positive cells with no clearly defined cell membrane and diffuse) in the toluidine blue–stained sections were visualized at ×60 magnification and presented in Fig 1, B; counted in each section by a blinded observer; and expressed as the number of MCs per millimeter. Representative images of intact and degranulated MCs were shown in Fig 1, B.

FIG 1.

Effect of LTD₄ and PGE₂ on ear edema in mice in vivo and MC activation in vitro. Wild-type (WT) BALB/c mice were treated with saline (Sal), 0.5 μmol/L LTD₄, 0.5 μmol/L PGE₂, or LTD₄ plus PGE₂. A, Ear thickness. B, Hematoxylin and eosin staining and toluidine blue staining. C, Quantification (blind analysis) of MCs per millimeter. D, Ear thickness in C57BL/6 and W-sh mice treated with 0.5 μmol/L LTD₄, PGE₂, and LTD₄ plus PGE₂. Results are means ± SEMs from 4 to 6 mice per group per experiment and 3 experiments performed. E–I, LAD2 cells (Fig 1, E–H) and hMCs (Fig 1, I) were stimulated with 0.5 μmol/L LTD₄, 0.5 μmol/L PGE₂, or both, and calcium flux (Fig 1, E and F), c-fos (Fig 1, G and I), and MIP-1β (Fig 1, H) were analyzed. *P < .05, **P < .01, and ***P < .001.

Cell culture

The LAD2 MC leukemia line¹⁵ was a kind gift from Dr Arnold Kirshenbaum (National Institutes of Health) and cultured as described previously.⁸ Primary hMCs were derived from cord blood mononuclear cells cultured for 6 to 9 weeks in RPMI supplemented with SCF, IL-6, and IL-10.¹⁶

Calcium flux

LAD2 cells (0.5 to 1 × 10⁶/sample) were washed and labeled with Fura 2-AM for 30 minutes at 37°C. Cells were stimulated with PGE₂ (0.5 μmol/L) with or without LTD₄ (0.5 μmol/L) priming, and the changes in intracellular calcium levels measured by using excitation at 340 and 380 nm and emission at 510 nm were recorded in a fluorescence spectrophotometer (Hitachi F-4500).⁸

Cell activation

LAD2 cells were stimulated with 0.5 μmol/L of LTD₄, PGE₂, or both for 15 minutes for the phosphorylation of Erk or 1 hour for expression of c-fos, 2 hours for expression of inflammatory gene transcripts, 3 hours for COX-2 protein expression, and 6 hours for measurement of cytokine and PGD₂ levels. LTD₄ responses were dose dependently inhibited by MK571 (with maximum inhibition at 1 μmol/L), and PGE₂ responses were attenuated by L-798 (in a dose-dependent manner, with maximum inhibition at 100 nmol/L). Therefore 1 μmol/L MK571 and 100 nmol/L L-798 were used in all the subsequent experiments. Transfection of isoform-specific siRNA smart pool constructs from Dharmacon (10 nmol/L) were carried out with siLentFect transfection reagent (Bio-Rad Laboratories, Hercules, Calif) for 48 hours, according to the manufacturer’s protocol.

Cell lysates and Western blotting

After stimulation with the respective agonists, LAD2 cells, hMCs, or both (0.5 × 10⁶) were lysed with lysis buffer (BD Biosciences, San Jose, Calif) supplemented with protease inhibitor cocktail (Roche, Mannheim, Germany) and phosphatase inhibitor cocktail (Pierce, Rockford, Ill). Immunoblotting was performed, as described previously.¹⁷ Western blots were incubated with ECL, and the bands were visualized with an imager (ProteinSimple, San Jose, Calif) and quantified by using Alpha View SA (ProteinSimple). The blots were stripped and reprobed with GAPDH antibody.

Real-time quantitative PCR

Expressions of CysLT₁R, CysLT₂R, EP₁, EP₂, EP₃, EP₄, MIP-1β, TNF-α, IL-8, COX-1, COX-2, and PKG transcripts were determined with real-time PCR performed on the LightCycler 480 (Roche). Total RNA was isolated from LAD2 cells after respective treatment with an RNeasy Mini Kit (Qiagen, Hilden, Germany), and cDNA was synthesized with the cDNA synthesis kit from Quanta BioSciences (Gaithersburg, Md). Real-time PCR was performed by using primers mentioned in Table E1 in this article’s Online Repository at www.jacionline.org, the levels of respective genes relative to GAPDH were analyzed, and the ΔΔ cycle threshold values relative to control were expressed as the fold change.

ELISA

Concentrations of PGD₂ and MIP-1β in the supernatants were assayed by using the PGD₂-MOX and MIP-1β ELISA kits purchased from Cayman Chemicals (Ann Arbor, Mich) and R&D Systems, respectively, according to the manufacturer’s instructions. PGD₂ secreted into medium was first converted into PGD₂-MOX and then assessed by using the PGD₂-MOX ELISA kit.

Statistical analysis

Blots presented are representative of 3 experiments performed, and data are expressed as means ± SEMs from at least 3 experiments, except where otherwise indicated. Significance was determined by using the Student t test, as well as 1-way ANOVA, followed by Tukey post hoc analysis.

RESULTS

Combined treatment with PGE₂ and LTD₄ synergistically potentiates peripheral vascular inflammation in mice

To elucidate the possible interaction between LTD₄ and PGE₂, we first evaluated the increase in vascular permeability and edema in the mouse ear, a widely used model to assess peripheral vascular inflammation,^18,19 in response to LTD₄ plus PGE₂ compared with agonists alone. LTD₄ or PGE₂ and LTD₄ plus PGE₂ (0.5 μmol/L, a dose at which both the agonists generated maximum response; data not shown) were injected into the ears of BALB/c mice, and tissue edema was assessed. The patterns of ear edema induced by LTD₄, PGE₂, and LTD₄ plus PGE₂ were similar, peaking at 30 minutes and returning to baseline at 300 minutes. However, we observed a significant enhancement in the magnitude of ear edema with LTD₄ plus PGE₂ compared with LTD₄ or PGE₂ alone (Fig 1, A). The increase in ear thickness with LTD₄ plus PGE₂ was rapid, transient, and approximately 6-fold higher compared with control values (0.32 ± 0.05 vs 0.05 ± 0.01) in the first half hour. We also observed an increase in ear thickness in response to both LTD₄ (approximately 3-fold, 0.15 ± 0.01) and PGE₂ (approximately 1.6-fold, 0.08 ± 0.01) individually during the first 30 minutes compared with control values. Histologic analysis of ear tissues revealed expansion of the extracellular space correlating with ear thickness (Fig 1, B). Interestingly, we found a robust increase in the number of degranulating MCs with LTD₄ plus PGE₂ treatment (Fig 1, B, middle and bottom right panel, arrows point to degranulating MCs, and Fig 1, C) compared with other groups, suggesting that LTD₄ plus PGE₂ can synergistically potentiate peripheral inflammation through their action on MCs. To elucidate the role of MCs, we repeated the above experiment in C57BL/6 and W-sh mice (on C57BL/6 background). Although we observed a similar pattern of ear inflammation in C57BL/6 wild-type mice as in BALB/c mice with all the agonists, the response was more transient and smaller (Fig 1, D). LTD₄ plus PGE₂ synergistically enhanced the inflammation in C57BL/6 wild-type mice, which is significantly attenuated in W-sh mice (Fig 1, D).

LTD₄ primes PGE₂-dependent calcium flux and potentiates c-fos phosphorylation and MIP-1β production in MCs

Next, we investigated the molecular mechanisms through which LTD₄ and PGE₂ activate MCs and potentiate inflammation. The expression pattern of cys-LT receptor transcript in LAD2 cells is similar to that in hMCs, with high levels of CysLT₁R compared with CysLT₂R (see Fig E1, C, in this article’s Online Repository at www.jacionline.org). Additionally, LAD2 cells express all 4 EPs (EP₃ > EP₂ > EP₄ > EP₁; see Fig E1, C). Furthermore, Western blot analysis of LAD2 cells and hMCs confirmed the expression of CysLT₁R, CysLT₂R, and EP_1-4 (see Fig E1, D). hMCs expressed higher CysLT₁R levels compared with LAD2 cells, and the expression pattern of EP₃ is comparable.

First, we analyzed the ability of LAD2 cells to flux calcium in response to PGE₂ and LTD₄. We observed that PGE₂ induced modest calcium flux (Fig 1, E, top panel, and Fig 1, F) and LTD₄ priming before PGE₂ stimulation led to an approximately 2.5-fold increase (0.42 ± 0.05 vs 0.16 ± 0.04; Fig 1, E, bottom panel, and Fig 1, F). Furthermore, LTD₄ and PGE₂ together increased phosphorylation and expression of c-fos (Fig 1, G) and secretion of MIP-1β (Fig 1, H). Importantly, in hMCs also, LTD₄ plus PGE₂ treatment significantly enhanced c-fos phosphorylation and expression (Fig 1, I). Stimulating LAD2 cells with a constant dose of LTD₄ (0.5 μmol/L) and varying concentrations of PGE₂ (0.001, 0.01, 0.1, and 0.5 μmol/L), we observed dose-dependent c-fos phosphorylation and expression and MIP-1β secretion (see Fig E1, A and B).

CysLT₁R inhibition partially attenuates LTD₄-primed PGE₂ responses

We then sought to identify the receptors involved in mediating this synergy by using MK571 (CysLT₁R antagonist⁸) and BayCysLT₂ (CysLT₂R antagonist²⁰). MK571 pretreatment abolished LTD₄-induced calcium influx but attenuated LTD₄-primed, PGE₂-generated calcium influx by 30% (Fig 2, A). LTD₄ plus PGE₂–stimulated c-fos phosphorylation and expression and MIP-1β secretion were reduced with MK571 preincubation (Fig 2, B-D), whereas BayCysLT₂ had no effect (Fig 2).

FIG 2.

LTD₄-primed and PGE₂-mediated calcium, c-fos, and MIP-1β responses in MCs are partly sensitive to CysLT₁R inhibition. LAD2 cells were stimulated with 0.5 μmol/L LTD₄, 0.5 μmol/L PGE₂, or both in the presence or absence of MK571 (1 μmol/L) and BayCysLT₂ (1 μmol/L), and calcium flux (A), c-fos (B and C), and MIP-1β (D) were analyzed. *P < .05 and ***P < .001.

Effects of LTD₄ plus PGE₂ are partially mediated through EP₃

We next investigated which of the putative EPs were interacting with CysLT₁R and enhancing PGE₂ responses. Stimulation of LAD2 cells with iloprost (EP₁/EP₃ agonist) and sulprostone (EP₃ agonist) in the presence of LTD₄ enhanced calcium flux (data not shown), c-fos (Fig 3, A and B), and MIP-1β expression (Fig 3, C), which is similar to what is seen with LTD₄ plus PGE₂, suggesting the involvement of EP₃ and a possible involvement of EP₁ in this synergism. The EP₂ agonist butaprost had no effect (Fig 3, A-C). Because EP₁ and EP₃ act dominantly through Gq- and Gi-mediated signaling pathways, respectively,¹⁴ we used PTX, a Gαi inhibitor, and evaluated LTD₄ plus PGE₂ synergy. Pretreatment of LAD2 cells with PTX (100 ng/mL for 18 hours) completely abolished the enhanced activation of c-fos and MIP-1β by LTD₄ and PGE₂ (see Fig E2, A-C, in this article’s Online Repository at www.jacionline.org), suggesting the involvement of EP₃. Furthermore, we found that the synergistic responses by LTD₄ plus PGE₂ were partially sensitive to the inhibition of EP₃ by L-798 (EP₃ antagonist, 100 nmol/L; Fig 3, D-F), which is similar to what is seen with CysLT₁R inhibition. Neither ONO-8711 (EP₁ antagonist, 2 nmol/L) nor L-161 (EP₄ antagonist, 100 nmol/L) had any effect on LTD₄ plus PGE₂–mediated c-fos expression/phosphorylation or MIP-1β generation (see Fig E2, D and E).

FIG 3.

EP₃ relays synergistic activation of c-fos and MIP-1β in response to PGE₂ and LTD₄. LAD2 cells were treated with LTD₄ (0.5 μmol/L) ± PGE₂ (0.5 μmol/L) or iloprost (10 μmol/L), butaprost (5 μmol/L), or sulprostone (100 nmol/L) ± L-798 (100 nmol/L), and c-fos (A, B, D, and E) and MIP-1β (C and F) were analyzed. *P < .05, **P < .01, and ***P < .001.

Synergistic responses to LTD₄ and PGE₂ were completely attenuated by blocking both CysLT₁R and EP₃ simultaneously

Next, we speculated that blocking both CysLT₁R and EP₃ simultaneously might completely block the synergistic responses. We observed that although the effects mediated by LTD₄ and PGE₂ are partially blocked by CysLT₁R and EP₃ antagonists alone (40% and 35% for c-fos and 58% and 57% for MIP-1β, respectively), combined treatment with MK571 and L-798 completely blocked the effects of LTD₄ plus PGE₂ (78% for c-fos phosphorylation and expression and 92% for MIP-1β; Fig 4, A-C). The combination of MK571 and L-798 also attenuated augmentation of c-fos and MIP-1β generation by LTD₄ plus sulprostone (data not shown). Importantly, in hMCs MK571 plus L-798 totally inhibited c-fos phosphorylation and expression (Fig 4, D and E), suggesting a functional relevance for this interaction and associated signaling.

FIG 4.

Combined effect of CysLT₁R antagonist and EP₃ antagonist on synergistic responses to LTD₄ and PGE₂. LAD2 cells (A–C) and hMCs (D and E) were preincubated with MK571 (1 μmol/L) and L-798 (100 nmol/L) separately or in combination and treated with LTD₄ (0.5 μmol/L) ± PGE₂ (0.5 μmol/L), and c-fos (Fig 4, A, B, D, and E) and MIP-1β (Fig 4, C) were analyzed. *P < .05, **P < .01, and ***P < .001.

LTD₄ and PGE₂ treatment upregulated inflammatory genes, PGD₂ generation, and Erk phosphorylation in LAD2 cells

We then investigated whether stimulation of MCs with LTD₄ plus PGE₂ would upregulate any other inflammatory chemokines and cytokines. LTD₄ treatment upregulated MIP-1β, TNF-α, IL-8, and COX-2, whereas PGE₂ upregulated COX-2 alone. LTD₄ plus PGE₂ treatment significantly upregulated MIP-1β (145 ± 30; Fig 5, A), TNF-α (90 ± 11; Fig 5, B), IL-8 (72 ± 20; Fig 5, C), and COX-2 (60 ± 10; Fig 5, D) transcripts, and the expression of COX-1 remained unchanged (Fig 5, E). Consistently, we observed a significant 3-fold upregulation of COX-2 protein with LTD₄ plus PGE₂ treatment (Fig 5, F). Furthermore, LTD₄ plus PGE₂ treatment upregulated PGD₂ secretion compared with control values (1630 ± 51 vs 1000 ± 75 pg/mL), and combined treatment with both MK571 and L-798 inhibited this secretion (Fig 6, A). Then we investigated the ability of LTD₄, PGE₂, and a combination of both in inducing Erk phosphorylation. PGE₂ and LTD₄ stimulation enhanced phosphorylation of Erk, and LTD₄ plus PGE₂ further potentiated this effect, which is sensitive to the MEK inhibitor PD989059 (50 μmol/L; Fig 6, B).

FIG 5.

Inflammatory gene induction and COX-2 protein by LTD₄ and PGE₂ in LAD2 cells. LAD2 cells were stimulated with 0.5 μmol/L LTD₄, PGE₂, or both; transcript levels of MIP-1β, TNF-α, IL-8, COX-2, and COX-1 were analyzed by means of real-time PCR at 2 hours of stimulation (A–E); and COX-2 protein levels were determined (F) in cell lysates. *P < .05, **P < .01, and ***P < .001.

FIG 6.

LTD₄ and PGE₂ crosstalk causes a synergistic increase in PGD₂ secretion and Erk phosphorylation and is mediated through PKG and Erk. LAD2 cells were stimulated with 0.5 μmol/L LTD₄, 0.5 μmol/L PGE₂, or both in the presence or absence of MK571 (1 μmol/L), L-798 (100 nmol/L), KT5823 (KT; 5 μmol/L), or PD98059 (PD; 50 μmol/L) before treatment, and PGD₂ secretion (A and F), Erk phosphorylation (B), c-fos (C and D), and MIP-1β generation (E) were analyzed. *P < .05, **P < .01, and ***P < .001.

LTD₄ plus PGE₂ signals operate through PKG and Erk-dependent pathway

We next examined the signaling involved in LTD₄ plus PGE₂ synergism downstream of receptor activation. PKs are known to be activated downstream of CysLT₁R activation.^8,21 A general PK inhibitor, H7 (10 μmol/L), completely blocked LTD₄, as well as LTD₄ plus PGE₂–induced effects (see Fig E3, A, E, and F, in this article’s Online Repository at www.jacionline.org). However, the general PKC inhibitor GF109203X (2 μmol/L) inhibited LTD₄ signals but modestly blocked enhanced c-fos and MIP-1β effects by LTD₄ plus PGE₂ (see Fig E3, B, E, and F). Furthermore, the PKA blockers Rp-cAMPS and H-89 had no effect on synergistic LTD₄ plus PGE₂ effects (see Fig E3, C-F). Interestingly, pretreatment of LAD2 cells with the PKG inhibitor KT5823 (5 μmol/L) or PD98059 attenuated LTD₄ plus PGE₂ synergistic responses (Fig 6, C-F). We then knocked down CysLT₁R, EP₃, both CysLT₁R and EP₃, and PKG in LAD2 cells by using protein-specific siRNAs (10 nmol/L) and nonspecific siRNAs as a control and analyzed LTD₄ plus PGE₂ effects. Transfection of MCs with CysLT₁R, EP₃, and PKG siRNAs significantly and specifically downregulated CysLT₁R (55%), EP₃ (60%), and PKG (64%) expression (see Fig E4 in this article’s Online Repository at www.jacionline.org). Importantly, downregulation of CysLT₁R or EP₃ partially inhibited both LTD₄- and LTD₄ plus PGE₂–induced responses, but knockdown of both CysLT₁R and EP₃ completely inhibited LTD₄ plus PGE₂–induced synergy (Fig 7, A-C). Also, knockdown of PKG significantly inhibited LTD₄ plus PGE₂–induced effects (Fig 7, A-C), whereas the control siRNA did not affect LTD₄- or PGE₂-induced inflammatory responses.

FIG 7.

Simultaneous inhibition of CysLT₁R and EP₃ on LTD₄ plus PGE₂–mediated MC activation in vitro and ear edema in vivo. CysLT₁R, EP₃, or both and PKG were knocked down in LAD2 cells by corresponding siR-NAs (10 nmol/L), and c-fos (A and B) and MIP-1β (C) were analyzed with LTD₄ plus PGE₂ treatment. D, Ear thickness in BALB/c mice treated with 0.5 μmol/L LTD₄ plus PGE₂ ± MK571 (1 μmol/L) or L-798 (100 nmol/L). Results in Fig 7, D, are means ± SEMs from 4 to 6 mice per group per experiment and 3 experiments performed. *P < .05, **P < .01, and ***P < .001.

Combined treatment with CysLT₁R and EP₃ antagonists attenuates vascular inflammation induced by LTD₄ plus PGE₂

Next, we determined the effect of blocking both CysLT₁R and EP₃ simultaneously in evoking vascular inflammation. The synergistic ear edema response caused by combined treatment with PGE₂ and LTD₄ was substantially attenuated with MK571 plus L-798 treatment (Fig 7, D).

DISCUSSION

Edema formation is a prominent feature of the inflammatory response and serves an important function in local host defense and tissue repair. Inflammatory responses result in the movement of fluid and plasma proteins into the extracellular space from leaky blood vessels in response to various chemical mediators.²² In the current study, for the first time, we demonstrate that 2 major eicosanoids, LTD₄ and PGE₂, both derived through alternate pathways from arachidonic acid, synergistically enhanced vascular inflammation in vivo (edema formation) and MC activation in vitro. Although W-sh mice have been well characterized as MC deficient, they have additional defects beyond just MC deficiency, and hence there exists a possibility that the in vivo vascular permeability attenuated in W-sh mice could be mediated through other cell types in addition to MCs. A number of mediators have been implicated in the regulation of inflammation in an MC-mediated mechanism.²³ Although MCs are primarily known to be activated in an antigen-dependent manner, evidence also suggests a role for them in antigen-independent activation.^19,22,24 They are regulatory cells throughout the course of acute inflammation, from its initiation to resolution,²⁵ and they contribute to the development of allergic disease.²⁶ MC activation in patients with various kinds of inflammatory diseases is significant from a clinical perspective. Translational implications of LTD₄ plus PGE₂ crosstalk provoked us to examine the candidate molecules and signaling mechanisms involved. We identified that LTD₄-PGE₂ synergism was mediated through CysLT₁R and EP₃. It is known that G protein–coupled receptors interact with one another, modulating each other’s function both positively and negatively.^27,28 Also, PGE₂ relays differential responses depending on the receptor subtype, receptor coupling, and the cell type. For instance, PGE₂ blocks FcεRI-mediated exocytosis of rat peritoneal MCs²⁹ and human lung MCs³⁰ in vitro through an increase in cyclic AMP levels (EP₂, EP₄, or both) but activates murine bone marrow-derived mast cells through EP₃ in vitro.³¹ On the contrary, PGE₂ in hMCs did not suppress FcεRI-dependent exocytosis but caused exocytosis on its own when the cells were primed with IL-4.¹⁶ It is likely that the suppressive effects of PGE₂ on MC activation largely reflects EP₂ signaling, and in agreement with this, EP₂ protein expression is downregulated in MCs in the nasal polyp mucosa of patients with aspirin-exacerbated respiratory disease.³² PGE₂ has been shown to stimulate chemotaxis and adhesion of MCs through EP.^33,34 In a recent report Morimoto et al¹⁹ demonstrated that EP₃ signaling in MCs generated PGE₂-induced vasodilatation and subsequent edema formation and speculated that EP₃-mediated MC activation might be involved in antigen-independent innate immune reactions. Interestingly, we identified that the LTD₄ plus PGE₂ synergism is also EP₃ dependent, strengthening the idea that EP₃ activation might contribute to a proinflammatory role of PGE₂.

LTD₄ plus PGE₂ treatment also upregulated transcripts for inflammatory mediators, such as TNF-α and IL-8, both of which are chemoattractants for neutrophils. Interestingly, we observed synergistic COX-2 induction by LTD₄ plus PGE₂ treatment and PGD₂ generation, revealing another major finding that LTD₄, in concert with PGE₂, can generate proinflammatory metabolite PGD₂ and amplify associated signaling. MCs express both COX-1 and COX-2, and COX-2 is upregulated by inflammatory stimuli, suggesting a mechanism that can amplify PGD₂ generation under inflammatory conditions. Interestingly, PGD₂ has been recently shown to synergize with LTE₄ to stimulate diverse T_H2 functions and T_H2 cell and neutrophil crosstalk.^35,36 It is possible that during inflammation, LTD₄ plus PGE₂ treatment not only activates MCs and initiates inflammation, but also the product of this signaling, PGD₂, can potentially perpetuate inflammation through combined action with LTE₄. Both LTD₄ and PGE₂ have been demonstrated to phosphorylate Erk in MCs.^8,16 In agreement, we noted phosphorylation of Erk by both agonists and observed enhanced Erk phosphorylation with LTD₄ plus PGE₂. Furthermore, we identified that LTD₄-PGE₂ synergy is relayed through PKG and Erk downstream of CysLT₁R and EP₃. Surprisingly, we found that PKC signals, which are vital in various aspects of CysLT₁R,^17,21,37 or PKA signals, which mediate EP-induced effects,¹⁶ are dispensable for this crosstalk, but it is dependent on PKG and Erk.

We have recently shown that LTD₄ can influence SCF responses, potentiating MC proliferation.¹² We also found that LTD₄ potentiates endothelial cell adhesion mediated by TNF-α,²⁰ suggesting that cys-LTs can modulate inflammation by acting in concert with other inflammatory mediators. Indeed, findings from the present study revealed that LTD₄ and PGE₂ produced at the inflammation site can synergistically activate MCs and further enhance inflammation. Notably, this enhanced inflammation by LTD₄ and PGE₂ both in vitro and in vivo is blocked only by the combined treatment of CysLT₁R and EP₃ antagonists and not by either antagonist alone. This argues for the possibility that the synergy is achieved through the contribution of a specific signaling component from LTD₄ and another component from PGE₂ signaling, rather than crosstalk at the receptor level. Although we found PKG and Erk are effector molecules in this signal, the convergence of these signaling events are still elusive.

Studies from EP knockout mice suggest that EP₃ is mainly responsible for PGE₂-induced MC activation and associated proinflammatory signaling pathways.³¹ High expression of EP₃ is observed in the brain, and recent findings suggest an injurious role for the PGE₂-EP₃ signaling axis in modulating brain injury and inflammation after intracerebral hemorrhage.³⁸ Along similar lines, enhanced EP₃ expression and signaling were found in patients with diabetes and have been speculated as a new therapeutic target for β-cell dysfunction in patients with type 2 diabetes.³⁹

All the above studies point to the pathologic role of EP₃. Current work alludes to the advantage of blocking CysLT₁R along with EP₃, and it is tempting to speculate that the combined treatment of EP₃ and CysLT₁R antagonists might contribute effectively to targeting inflammation compared with the available CysLT₁R antagonists alone and could open new avenues in clinical approaches for MC-mediated inflammatory diseases. Although PGE₂ alone did not induce significant inflammatory transcripts and MIP-1β generation, it potentiated all the inflammatory readouts in concert with LTD₄, suggesting that cys-LTs could modulate PGE₂ signaling and blocking LTD₄ plus PGE₂ could be a potential therapeutic target to combat initiation and progression of inflammation.

Key messages.

• LTD₄ synergizes with PGE₂ and activates MCs in vitro and vascular inflammation in vivo through CysLT₁R and EP₃, which might contribute to MC-mediated allergic inflammation.

• LTD₄-PGE₂ synergism triggers diverse MC inflammatory responses through Gi-, PKG-, and Erk-dependent pathways.

• LTD₄ plus PGE₂ crosstalk is efficiently blocked by simultaneous inhibition of CysLT₁R and EP₃ but not by either of the agonists alone, suggesting that vascular inflammation can be targeted efficiently by combining currently available CysLT₁R antagonists with EP₃ antagonists.

Abbreviations used

cys-LT

Cysteinyl leukotriene

CysLT₁R

Cysteinyl leukotriene receptor 1

CysLT₂R

Cysteinyl leukotriene receptor 2

EP

E-prostanoid receptor

Erk

Extracellular signal-regulated kinase

GAPDH

Glyceraldehyde-3-phosphate dehydrogenase

hMC

Human cord blood–derived mast cell

LT

leukotriene

MC

Mast cell

MIP-1β

Macrophage inflammatory protein 1β

PG

Prostaglandin

PK

Protein kinase

PTX

Pertussis toxin

SCF

Stem cell factor

siRNA

Small interfering RNA