Cooperativity of E-prostanoid receptor subtypes in regulating signaling and growth inhibition in human airway smooth muscle

Abstract

Prostaglandin E2 (PGE₂) is produced in the airway during allergic lung inflammation and both promotes and inhibits features of asthma pathology. These mixed effects relate to 4 E-prostanoid (EP) receptor subtypes (EP1, 2, 3 and 4) expressed at different levels on different resident and infiltrating airway cells. Although studies have asserted both EP2 and EP4 expression in human airway smooth muscle (HASM), a recent study asserted EP4 to be the functionally dominant EP subtype in HASM. Herein, we employ recently-developed subtype-selective ligands to investigate singular or combined EP2 and EP4 receptor activation in regulating HASM signaling and proliferation. The subtype specificity of ONO-AE1-259-01 (EP2 agonist) and ONO-AE1-329 (EP4 agonist) was first demonstrated in human embryonic kidney 293 cells stably expressing different EP receptor subtypes. EP receptor knockdown and subtype-selective antagonists demonstrated EP2 and EP4 receptor responsiveness in HASM cells to the specific ONO compounds, whereas PGE₂ appeared to preferentially signal via the EP4 receptor. Both singular EP2 and EP4 receptor agonists inhibited HASM proliferation, and combined EP2 and EP4 receptor agonism exhibited positive cooperativity in both chronic G_s-mediated signaling and inhibiting HASM proliferation. These findings suggest both EP2 and EP4 are functionally important in HASM, and their combined targeting optimally inhibits airway smooth muscle proliferation.—Michael, J. V. Gavrila, A., Nayak, A. P., Pera, T., Liberato, J. R., Polischak, S. R., Shah, S. D., Deshpande, D. A., Penn, R. B. Cooperativity of E-prostanoid receptor subtypes in regulating signaling and growth inhibition in human airway smooth muscle.

Prostaglandin E2 (PGE₂) exerts multiple effects in the lung by regulating the function of multiple airway cell types. During the allergic lung inflammation that occurs with asthma, PGE₂ levels are increased and regulate both pro- and antiasthmatic features (1–4). In animal models and cell-based assays, PGE₂ has been demonstrated to inhibit multiple indices of allergic inflammation (1, 5–8), inhibit proliferation of cultured human airway smooth muscle (HASM) cells (9–14), relax contracted airways ex vivo (15–18), and promote bronchoprotection and bronchorelaxation in vivo (16, 18). However, PGE₂ has also been implicated as a proinflammatory mediator and is involved in inflammatory diseases (19, 20). Further, the net effect of PGE₂ on antigen-presenting cells favors a type 2 T-helper cell response, which enhances B cell IgE production and allergic inflammation (21).

The diverse effects of PGE₂ can be attributed to the existence of 4 different [E-prostanoid (EP)1–4] G protein–coupled receptors responsive to PGE₂, which are expressed at different levels on various cell types. Because EP2 and EP4 are G_s-coupled receptors—whereas EP1 and EP3 receptors couple to G_q and G_i, respectively—cellular responses to PGE₂ are highly dependent on the EP receptor subtypes expressed. In cells expressing more than 1 EP receptor subtype, PGE₂ has the potential to stimulate cooperative signaling (e.g., by G_s-coupled EP2 and EP4 receptors) or competitive signaling (e.g., between G_i-coupled EP3 and G_s-coupled EP2/EP4 receptors).

The short half-life of PGE₂ and its promiscuity for multiple receptors on multiple cell types has limited the utility of PGE₂ as a therapeutic (22). Indeed, the potential of PGE₂ as an asthma therapeutic, at least with respect to its bronchorelaxant property, has been recognized for years. The bronchodilator effect of PGE₂ has been demonstrated in healthy and asthmatic subjects (23–25). However, side effects of inhaled PGE₂ have been problematic, consisting of both cough and substernal burning (26, 27).

G_s-coupled EP receptors on HASM are particularly attractive targets for asthma treatment, given we and others have demonstrated an ability (superior to that of β-agonists) of PGE₂ to both relax contracted HASM and inhibit mitogen-stimulated HASM proliferation (14). These G_s-, cAMP-, and PKA-dependent effects occur despite a demonstrated counteracting effect of EP3 receptor signaling in HASM (14, 28). Efforts to limit the side effects of PGE₂ by developing subtype-selective EP ligands have, until recently, been disappointing. Hampered by either a lack of sufficient specificity or a lack of efficacy (29), no EP subtype–selective ligands have advanced to the clinic (30). A purported EP2-specific ligand, AH13205, was shown to relax contracted guinea pig tracheae ex vivo and induce bronchodilation in a guinea pig model of asthma (31). However, in a subsequent clinical trial, AH13205 failed to exhibit bronchodilation activity in humans, while promoting wheezing or cough (31).

More recently, ONO-AE1-259-01 and ONO-AE1-329 have been developed as subtype-selective EP2 and EP4 receptor agonists, respectively (32). Although an early study by our group examining desensitization of PGE₂-mediated signaling asserted EP2 as the predominant G_s-coupled EP receptor in HASM, a subsequent study by Buckley et al. (16) demonstrated the EP4-selective ONO-AE1-329, but not ONO-AE1-259-01, potently relaxed contracted human airways ex vivo. To clarify the expression and functional utility of EP receptor subtypes in HASM, the present study characterizes the time- and dose- dependent signaling of subtype-selective EP agonists in HASM cultures and links such signaling to the regulation of HASM proliferation.

MATERIALS AND METHODS

Materials

Antibodies against phosphorylated ERK1/2 (threonine 202, tyrosine 204; 4370), hemagglutinin (HA) tag conjugated to Alexa-488 (2350), RIPA cell lysis buffer (9806), and platelet-derived growth factor (PDGF)-BB (8912) were from Cell Signaling Technology (Danvers, MA, USA). Antibodies against β-actin (A5441) and PGE₂ were from MilliporeSigma (Burlington, MA, USA). Protease and phosphatase inhibitors were from Bimake (Houston, TX, USA). IRDye 680 or 800 secondary antibodies were from Rockland Immunochemicals (Limerick, PA, USA). Sulprostone and PF-04418948 were obtained from Cayman Chemical (Ann Arbor, MI, USA). ONO-AE1-259-01, ONO-AE-248, ONO-AE1-329, and ONO-AE3-208 were from Ono Pharmaceutical (Osaka, Japan). Insulin-transferrin-selenium was from Thermo Fisher Scientific (41400045; Waltham, MA, USA). All other materials were obtained from MilliporeSigma or from previously identified sources (33–35).

Cell culture

Human embryonic kidney 293 (HEK-293) cells were cultured in DMEM supplemented with 10% fetal bovine serum (FBS) and antibiotics and maintained in an incubator at 37°C supplemented with 5% CO₂, as previously described (36). Sequences encoding human EP2, EP3, and EP4 receptors were subcloned into pcDNA3.1 possessing an HA cassette, creating constructs encoding each receptor with an N-terminal HA tag. HEK-293 cells stably expressing HA-EP2, HA-EP3, HA-EP4, or a vector control were generated by transfection of constructs using Lipofectamine 3000 Reagent (Thermo Fisher Scientific), followed by selection with 250 µg/ml G418, as previously described (37). Primary HASM cultures were established from human tracheae and cultured in Ham’s F12 medium supplemented with 10% FBS as previously described (12). Culture passages 2–5 were used in the present study. In addition, senescence-resistant HASM cultures stably expressing hTERT (38) were also used. Data from the hTERT and the primary HASM cultures were similar and were, therefore, analyzed collectively. As previously described (39), cells were maintained in medium containing 10% FBS and then switched to serum-free medium 24 h prior to stimulation.

Cellular cAMP accumulation

For assay of cAMP accumulation, cells plated in 24-well plates were grown to near confluence in Ham’s F12 medium supplemented with 10% FBS, washed, and replaced with serum-free media for 24 h, then stimulated with plain Ham’s F12 medium containing a final concentration of 1 mM isobutylmethylxanthine, with vehicle (0.1% DMSO) or indicated agonist. To minimize the effects of within-group variance of absolute agonist-stimulated cAMP accumulation, values were calculated and reported as a percentage of cAMP accumulation stimulated by 100 μM forskolin (FSK) for each group, as previously described (35).

Intracellular calcium measurements

Cells were plated onto collagen-coated 96-well plates, grown to confluence, and loaded with 2 µM Fluo-4 acetoxymethyl ester (BD Biosciences, Franklin Lakes, NJ, USA) and probenecid for 1 h. Indicated final concentrations of agonists were added by an automated pipetting system in duplicate, and the 525 nm signals were generated by excitation at 485 nm with a Flex Station II (Molecular Devices, San Jose, CA, USA) as previously described (40). The net peak calcium (Ca²⁺) response was calculated as [(agonist-induced fluorescence units) − (basal fluorescence units)]. Maximal Ca²⁺ response was determined by stimulating the cells with ionomycin (1 µM).

Immunoblotting

Cultured HEK-293 or HASM cells were grown to near confluence in 12-well plates and switched to serum-free medium 24 h prior to stimulation. Cells were stimulated as indicated, lysed, sonicated briefly, electrophoresed on an 8 or 10% SDS-polyacrylamide gel, transferred to nitrocellulose membranes, and subsequently probed with the indicated antibodies (10). Bands were visualized and signals (infrared emission) quantified directly using the Odyssey Infrared Imaging System (Li-Cor Biosciences, Lincoln, NE, USA), as previously described (41).

Small interfering RNA–mediated knockdown of EP receptors in HASM cells

For knockdown of EP2, EP3, or EP4 receptors (On-Targetplus Smartpool PTGER2- L-005712-00, PTGER3- L-005713-02, PTGER4- L-005714-00; Horizon Discovery, Cambridge, United Kingdom), 2 µg of small interfering RNA (siRNA) duplexes or scrambled (Scr; control) sequence (5′-GCGCGCUUUGUAGGAUUCGdTdT-3′) were mixed in 1× siRNA buffer, and HASM cultures (plated 24 h previously at a density of 10⁴ cells/cm⁻²) were transfected using Dharmafect transfection reagent (Thermo Fisher Scientific) as per the manufacturer’s instructions. After 24 h, cells were passaged onto 12-well plates for immunoblot analysis and into a matched 6-well plate for quantitative PCR analysis, followed by the indicated stimulation 72 h later corresponding to the period of knockdown. Because of the lack of sufficiently sensitive EP receptor antibodies, specific EP receptor subtype knockdown was assessed by quantitative PCR as per Saxena et al. (37).

Confocal microscopy

HEK-293 cells stably expressing indicated EP receptor or empty vector were fixed on collagen-coated coverslips with ice-cold methanol. Coverslips were permeabilized with 0.01% Triton X-100 and stained with FITC-conjugated anti-HA antibody. Confocal images were obtained on an Eclipse TI2 microscope (Nikon, Tokyo, Japan) using multitrack sequential excitation (408 for DAPI, 488 for FITC-conjugated antibody) and emission filter sets.

Retroviral expression of PKA inhibitory peptide

Stable expression of green fluorescent protein (GFP) or the GFP-chimera of the PKA inhibitory peptide (PKI) [PKI-GFP, the utility of which in HASM was previously established (11, 12, 14,17)] was achieved by retroviral infection, as previously described (11, 14). Briefly, retrovirus for the expression of each was produced by cotransfecting GP2-293 cells with p-vesicular stomatitis virus (VSV)-G vector (encoding the pantropic VSV-G envelope protein) and either pLPCX-GFP or pLPCX-PKI-GFP. At 48 h after transfection, supernatants were harvested and used to infect primary HASM cultures, and cultures were subsequently selected to homogeneity with 5 μg/ml puromycin (Tocris Bioscience, Bristol, United Kingdom).

HASM growth assays

CyQuant

Cells were plated in a 96-well plate and maintained in complete Ham’s F12 medium supplemented with 10% FBS until 90% confluent. Cells were then switched to serum-free Ham’s F12 medium, and 24 h later pretreated with indicated agonists or vehicle 30 min before addition of 10 ng/ml PDGF as per (42, 43). After 72 h treatment of PDGF with vehicle or agonists, medium was changed to assay buffer containing CyQuant dye, and fluorescence intensity was measured as per the manufacturer’s instructions. Increasing HASM cell numbers tested using CyQuant dye showed a direct linear correlation of cell number to increase in fluorescence value (r² = 0.97, data not shown).

Cell counts

A total of 48,000 cells were seeded into a 6-well plate and maintained in complete Ham’s F12 medium supplemented with 10% FBS. After 48 h, cells were switched to serum-free Ham’s F12 medium with insulin-transferrin-selenium. Cells were starved for 48 h before addition of indicated conditions, with or without the presence of 10 ng/ml PDGF. After 72 h treatment of PDGF with vehicle or agonists, cells were counted using a hemocytometer.

Statistical analyses

Data are presented as mean ± sem values. Experiments employing HASM cells were repeated with cultures from different donors to create the indicated n values. Statistically significant differences among vehicle and treatment groups were assessed by 2-way ANOVA with Bonferroni post hoc analysis or Student’s t test where appropriate, with values of P < 0.05 sufficient to reject the null hypothesis using Prism software (GraphPad Software, La Jolla, CA, USA).

RESULTS

Specificity of EP receptor subtype–specific ligands assessed in HEK-293 cells expressing recombinant EP receptors

To establish the specificity of the EP subtype–specific ligands employed in the present study, we generated HEK-293 cells overexpressing HA-tagged human EP2, EP3, and EP4 receptors and a vector-transfected control line, which were confirmed by immunocytochemistry (Fig. 1A). Because we previously established that HASM cells do not express appreciable levels of EP1 receptor (44), we focused on characterization of EP2, EP3, and EP4 agonists. We first assessed the ability of agonists to stimulate phosphorylation of the PKA substrate vasodilator-stimulated phosphoprotein (VASP) as a readout for G_s-coupled receptor signaling (45, 46) using 10 and 100 nM concentrations of ligands. VASP phosphorylation by PKA at serine 157 causes a mobility shift when run on an SDS-PAGE gel (46–50 kDa), on which the top band depicts phospho-VASP (p-VASP) and culmination of both bands represents total VASP (17). In vector-expressing HEK-293 cells, ONO-AE1-259-01 and butaprost stimulated a low level of VASP phosphorylation, whereas ONO-AE1-329 and PGE₂ stimulated VASP more robustly, suggesting low levels of endogenous EP2 receptor expression and slightly higher levels of endogenous EP4 receptors in this particular HEK-293 cell line (Fig. 1B, C). In HA-EP2–expresing cells, but not HA-EP3– or HA-EP4–expressing cells, both ONO-AE1-259-01 and butaprost significantly promoted VASP phosphorylation to a greater extent than that observed in vector control cells. Conversely, in HA-EP4–expressing cells, but not HA-EP2– or HA-EP3–expressing cells, only ONO-AE1-329 promoted VASP phosphorylation beyond that observed in vector control cells. PGE₂ induced p-VASP levels in both HA-EP2– and HA-EP4–expressing cells, but not HA-EP3–expressing cells, beyond that induced in vector control cells (Fig. 1B, C). In the HA-EP2–, HA-EP4–, and HA-EP3–expressing lines, phosphorylation of ERK 1/2 was not increased by EP2- and EP4-specific agonists. ERK 1/2 phosphorylation was observed in all lines stimulated with PGE₂ and in HA-EP3–expressing cells stimulated with ONO-AE-248 and sulprostone (EP3 ligands). These results suggest that ONO-AE1-259-01 and ONO-AE1-329 possess a high level of specificity for the EP2 and EP4 receptors, respectively.

Figure 1.

Characterization and signaling of novel EP subtype–selective agonists in HEK-293 cells expressing recombinant EP receptor subtypes. Stable lines of HEK-293 were selected after transfection of vector or indicated HA-tagged EP receptor, as described in Materials and Methods. A) Cells were plated onto collagen-coated glass coverslips, fixed, and stained for the HA tag; original magnification, ×40. B) Cells were stimulated for 10 min with indicated concentrations of agonists. All concentrations are given in nM unless otherwise indicated. C) Graphic representation of mean ± sem values from 3 to 4 experiments. p-VASP values induced by 10 and 100 nM ONO-AE1-259-01 and by 5 µM butaprost were significantly greater (P < 0.05) in HA-EP2-expressing cells relative to vector control cells (vector cells: 10 nM 10 ± 2% and 100 nM 22 ± 4%; HA-EP2 cells: 72 ± 8 and 76 ± 8%; vector cells and 5 µM butaprost 24 ± 7%, HA-EP2 cells 73 ± 5%, respectively). 248, ONO-AE-248; 259, ONO-AE1-259-01; 329, ONO-AE1-329; B, butaprost; P, PGE₂; S, sulprostone.

Signaling induced by EP-specific ligands in HASM cells

Next, we characterized EP receptor subtype signaling in HASM cells. We first assessed acute signaling indicated by either intracellular cAMP accumulation (G_s-mediated signaling readout) (Fig. 2A) or calcium mobilization (G_q-mediated signaling readout) (Fig. 2B). Accumulation of cAMP (represented as percentage of the FSK-stimulated value) response in primary HASM cells stimulated with ONO-AE1-259-01 (10 nM 14 ± 5%, 100 nM 31 ± 9%) or ONO-AE1-329 (10 nM 10 ± 2%, 100 nM 24 ± 4%) was similar. Accumulation of cAMP induced by ONO-AE1-259-01 and ONO-AE1-329 was consistent with the ability of each drug to promote VASP phosphorylation at 10 min (Fig. 2C, D). In Flexstation analysis of agonist-stimulated calcium mobilization, among the EP receptor agonists, only PGE₂ and sulprostone significantly induced a calcium signal (100 nM PGE₂ 34 ± 3%, 10 nM sulprostone 18 ± 2%, and 100 nM sulprostone 63 ± 5% maximum response), whereas ONO-AE1-259-01 and ONO-AE1-329 had no detectable effect (Fig. 2B). Interestingly, all doses of ONO-AE-248 tested (10 nM–1 µM) failed to promote calcium mobilization.

Figure 2.

Effects on second messenger generation and signaling upon acute EP-specific stimulation in HASM cells. A) HASM cultures were grown to near confluence in 24-well plates and treated with indicated agonists for 10 min. cAMP generation was isolated, quantified, and represented as a percentage of 100 µM FSK as described in Materials and Methods. B) Calcium mobilization of primary HASM cells stimulated with indicated agonists measured using Flexstation analysis. C) Representative Western blot analysis of HASM cells stimulated for 10 min with indicated agonist. D) Graphic representation of data from Western blot analyses. 248, ONO-AE-248; 259, ONO-AE1-259-01; 329, ONO-AE1-329; Hist, histamine; Sul, sulprostone. E) Representative Western blot of HASM cultures infected with retrovirus enabling expression of GFP (control) or PKI-GFP selected with puromycin, as described in Materials and Methods. Cultures were plated into 12-well plates and stimulated with indicated agonist for 10 min. All concentrations are given in nM, unless otherwise indicated. Graphical data represent mean ± sem values from 3 to 6 experiments, 2-way ANOVA. *P < 0.05.

We next compared the effects of ONO-AE1-259-01 and ONO-AE1-329 with those of PGE₂ and isoproterenol (ISO) on acute (10 min) additional G_s- (p-VASP) and G_q- or G_i- dependent (ppERK1/2) signaling outcomes (Fig. 2C, D). p-VASP levels were induced by ONO-AE1-259-01, ONO-AE1-329 (alone or in conjunction with ONO-AE1-259-01), PGE₂, and ISO (Fig. 2C, D). Acute treatment with 100 nM ONO-AE1-259-01 and ONO-AE1-329 induced nearly complete VASP phosphorylation (87 ± 3 and 76 ± 10%, respectively) similar to that induced by 1 µM ISO (91 ± 2%) and 100 nM PGE₂ (79 ± 8%). A lesser, yet significant, induction of p-VASP levels was stimulated with both 10 nM ONO-AE1-259-01 (50 ± 7%) and ONO-AE1-329 (77 ± 6%) (Fig. 2C, D). ERK 1/2 phosphorylation was not significantly induced by ONO-AE1-259-01, ONO-AE1-329 (alone or in conjunction with ONO-AE1-259), or ISO, and only weakly by PGE₂ (Fig. 2C, D). Consistent with our previous studies assessing PKA-dependent PGE₂ signaling in HASM (12, 14, 17), the EP2 (ONO-AE1-259-01)- and EP4 (ONO-AE1-329)-mediated phosphorylation of VASP was reversed by heterologous expression of PKI-GFP (Fig. 2E).

To further confirm ligand specificity and assess the dominant EP receptor in PGE₂-mediated G_s signaling in HASM, we performed siRNA-mediated knockdown of EP2, EP3, and EP4 receptors. We achieved a range of 73–92% knockdown of the targeted EP receptor subtype, with specificity achieved except with Smartpool EP2 siRNA, which produced a mean 91% knockdown of EP2 receptor but a concomitant 50% knockdown of EP4 receptor (Table 1). Experiments employing alternative EP2-targeting siRNA failed to achieve superior specificity (data not shown). Knockdown of EP2 receptor, but not of EP3 or EP4 receptors, significantly (P < 0.05) inhibited the induction of p-VASP stimulated by ONO-AE1-259-01 (Scr knockdown (KD) 32 ± 6% to EP2 KD 2 ± 0.4%) or butaprost (Scr KD 76 ± 12% to EP2 KD 23 ± 24%) (Fig. 3A, B). EP4 receptor KD, but not EP2 or EP3 receptor KD, significantly (P < 0.05) inhibited the induction of p-VASP stimulated by ONO-AE1-329 (Scr KD 80 ± 7% to EP4 KD 13 ± 4%) (Fig. 3A, B). Despite the established ability for PGE₂ to signal through EP2, KD of the EP2 receptor had no significant effect on the (10 or 100 nM) PGE₂-mediated VASP phosphorylation (Scr KD 10 nM PGE₂ 48 ± 12% and 100 nM PGE₂ 88 ± 3% to EP2 KD 10 nM PGE₂ 46 ± 13% and 100 nM PGE₂ 88 ± 4%). Alternatively, knockdown of EP4 had a more profound effect on VASP phosphorylation after treatment with a low-dose (but not high-dose) PGE₂ (Scr KD 10 nM PGE₂ 48 ± 12% and 100 nM PGE₂ 88 ± 3% to EP4 KD 10 nM PGE₂ 21 ± 6% and 100 nM PGE₂ 94 ± 2%).

TABLE 1.

EP receptor knockdown efficacy

	Knockdown efficacy (%)
EP receptor	EP2	EP3	EP4
EP2	91 ± 8	1 ± 1	50 ± 7
EP3	29 ± 10	95 ± 2	42 ± 5
EP4	20 ± 7	0 ± 0	73 ± 4

Percent reduction in EP receptor subtype mRNA abundance assessed by

quantitative PCR, as described in Materials and Methods.

Figure 3.

Genetic and pharmacological modulation of EP2 or EP4 receptors in HASM reverses PKA-mediated signaling. A) Representative blots probing for VASP and β actin in Scr or indicated EP receptor knockdown. B) Graphic representation of data presented in A. C) HASM pretreated for 10 min with vehicle, PF-04418948 (1 µM), or ONO-AE3-208 (1 µM) (EP2 and EP4 antagonist, respectively) before treatment with indicated agonist (50 nM) for 10 min. Graph depicts mean ± sem values from 3 to 6 experiments. Buta, butaprost; Sul, sulprostone; Veh, vehicle. All concentrations are given in nanomolar units, unless otherwise indicated. Scr vs. specific EP KD (B), or vehicle vs. indicated antagonist (C). *P < 0.05.

To further assess the relative role of EP receptor subtypes in PGE₂-mediated signaling in HASM, we employed subtype-selective antagonists. HASM cells were pretreated with antagonists specific to EP2 (PF-04418948) or EP4 (ONO-AE3-208), and induction of p-VASP by stimulation with EP subtype–selective agonists or PGE₂ was assessed. Pretreatment of HASM cells with PF-04418948 (1 µM) for 10 min inhibited acute p-VASP induction stimulated by ONO-AE1-259-01, but it did not inhibit induction stimulated by ONO-AE1-329 or PGE₂. Interestingly, ONO-AE3-208 (1 µM) pretreatment inhibited p-VASP induction stimulated by both ONO-AE1-329 and PGE₂ (vehicle/PGE₂ 85 ± 4% to ONO-AE-208/PGE₂ 15 ± 5%), but not by stimulation by ONO-AE1-259-01 (Fig. 3C). Taken together, these data implicate EP4 as the predominant effector of PGE₂-mediated G_s signaling.

Cooperativity of EP2 and EP4 in time-dependent signaling and inhibition of mitogen-stimulated HASM proliferation

To further characterize the effect of EP-subtype signaling, we investigated the kinetics of prolonged ONO-AE1-259-01 or ONO-AE1-329 signaling. Because specific activation of either EP2 or EP4 can promote VASP phosphorylation in HASM (Figs. 2C, D and 3), we investigated whether we could detect cooperative signaling by treating HASM with both EP2 and EP4 agonists. In HASM cells stimulated for 1 h, 100 nM PGE₂ consistently induced higher p-VASP levels than 5–10 nM of either ONO-AE1-259-01 or ONO-AE1-329 alone. Stimulation with ONO-AE1-259-01 and ONO-AE1-329 (5 nM each) for 3 h induced VASP phosphorylation that was greater than that induced by ONO-AE1-259-01 (10 nM) (P < 0.05), but only modestly (P > 0.05) greater than that stimulated by ONO-AE1-329 (10 nM) alone. After 18 h stimulation, p-VASP levels induced by combined stimulation with ONO-AE1-259-01 and ONO-AE1-329 (5 nM each) were significantly (P < 0.05) greater than those stimulated by either agonist alone (10 nM) and were comparable to that stimulated by 100 nM PGE₂ (Fig. 4).

Figure 4.

Time-dependent regulation of VASP levels by subtype-selective EP agonists alone or combined. A) Representative blots probing for VASP or β-actin of HASM cells grown to near confluence in 12-well plates and stimulated with indicated agonist and time. B) Graphic representation of data presented in A. All concentrations are given in nanomolar units, unless otherwise indicated. Graph depicts mean ± sem values from 6 experiments. *P < 0.05.

We have previously reported the PKA dependence of the antimitogenic effect of both β-agonists and PGE₂ on HASM and have noted a correlation of the antimitogenic effect of these agents with their ability to sustain PKA activity (12, 14). Consistent with the ability of combined EP2 and EP4 agonists to signal chronically in an additive or greater manner, combined treatment of ONO-AE1-259-01 and ONO-AE1-329 (5 nM each) caused an inhibition of PDGF-stimulated proliferation significantly greater than that caused by either agonist alone (10 nM). Using CyQuant analysis, the percent inhibition elicited by combined (5 nM each) ONO-AE1-259-01 and ONO-AE1-329 treatment approached that elicited by 100 nM PGE₂ (74 ± 6 and 78 ± 3%, respectively), and was significantly (P < 0.05) greater than that elicited by the β-agonist ISO (49%) (Fig. 5A). Similarly, using cell counts, the percent inhibition elicited by combined (5 nM each) ONO-AE1-259-01 and ONO-AE1-329 treatment approached that elicited by 100 nM PGE₂ (100 ± 8 and 98 ± 8%, respectively), and was significantly (P < 0.05) greater than that elicited by either ONO-AE1-259-01 and ONO-AE1-329 at 5 or 10 nM (72 ± 11 and 66 ± 8%, at 10 nM concentrations respectively) or the β-agonist ISO (61 ± 7%) (Fig. 5B).

Figure 5.

Cooperative growth inhibition by EP2 and EP4 receptor agonists in HASM. HASM cultures were pretreated with indicated agonists prior to stimulation with 10 ng/ml PDGF for 72 h, and cell proliferation was assessed by CyQuant (A) or cell counts (B). All concentrations are given in nanomolar units, unless otherwise indicated. Graph represents mean ± sem values from 7 experiments. *P < 0.05.

DISCUSSION

Findings from the present study provide the first comprehensive assessment of HASM EP receptor subtype signaling and regulation of proliferation, enabled by the recent development of subtype-specific agonists of the EP2 and EP4 receptors. Both EP2 and EP4 receptor subtype–specific agonists demonstrated activity and specificity in HASM cells, mediating cAMP generation and prolonging downstream G_s signaling. We also observed positive cooperativity upon treatment with EP2 and EP4 agonists with respect to prolonged downstream G_s signaling, which was associated with a significantly greater inhibition of HASM proliferation than that mediated by either agonist alone. Interestingly, despite a clear and strong ability of the EP2 receptor agonist to signal in HASM, EP2 receptors appeared dispensable in signaling by PGE₂ in the cells. Collectively, these findings advance our understanding of EP receptor biology and pharmacology and suggest efficacious strategies for the application of EP ligands as asthma therapies.

Based on our previous study that used primarily artificial expression systems to demonstrate that the EP2 receptor, but not the EP4 receptor, was resistant to G protein-coupled receptor kinase/arrestin-mediated desensitization, we assumed that the pronounced ability (relative to that of β-agonist) of PGE₂ to sustain signaling in HASM was explained by HASM preferentially expressing the EP2 receptor (35). We further postulated that such sustained signaling contributed to the greater ability of PGE₂ to inhibit HASM proliferation. However, the recent study by Buckley et al. (16) demonstrated that ONO-AE1-329, but not ONO-AE1-259-01, relaxed contracted human airways, suggesting that EP4 was the predominant EP receptor subtype expressed in HASM. In the present study, we found that EP2 knockdown had no significant effect on PGE₂-mediated VASP signal, whereas EP4 knockdown had a robust attenuation of PGE₂-mediated VASP phosphorylation (Fig. 3). Furthermore, pretreatment with an EP4 antagonist (but not EP2 antagonist) attenuated PGE₂-mediated VASP signaling. These data support EP4 as the physiologically predominant EP receptor subtype driving PGE₂-mediated G_s signaling and bronchodilation.

Curiously, in the current study, we demonstrate the ability of a specific EP2 receptor agonist to acutely induce cAMP generation and VASP phosphorylation in HASM cells in a manner comparable to that induced by specific EP4 receptor agonism. Moreover, our studies of recombinant EP receptor subtypes in HEK-293 cells demonstrate a clear ability of PGE₂ to stimulate EP2 receptor signaling. Why EP2 receptor signaling in HASM appears dissociated from the functional effect of HASM relaxation and bronchodilation is unclear. One possible explanation relates to compartmentalized signaling; an interesting study by Bogard et al. (47) provides clear evidence of compartmentalized signaling distinguishing disparate functional effects EP2 receptor and β-2-adrenoceptor agonism in HASM. Whether EP2 and EP4 receptors signal in distinct compartments remains to established, but with respect to the regulation of HASM proliferation, unlike in the regulation of HASM contraction EP2 agonism is clearly impactful and can cooperate in a positive manner with EP4 agonism.

The extent to which the use of dual EP2 and EP4 receptor agonist therapy can be employed in the management of obstructive lung disease remains to be established. However, the current study strongly suggests that, at least for 1 important pathogenic feature of asthma (increased airway smooth muscle mass contributing to airway remodeling), EP2 and EP4 receptor signaling is therapeutically cooperative. Moreover, the approach of combined EP2 and EP4 receptor agonist dual therapy represents a logical strategy to avoid the problematic side effects of inhaled PGE₂ (cough and substernal burning), which appear to be mediated by the EP3 receptor (48). Future studies examining the efficacy and safety of this dual therapy in in vivo models of asthma are the next important step in advancing the clinical utility of targeting EP receptor subtypes.

Glossary

EP

E-prostanoid

FBS

fetal bovine serum

FSK

forskolin

GFP

green fluorescent protein

HA

hemagglutinin

HASM

human airway smooth muscle

HEK-293

human embryonic kidney 293

ISO

isoproterenol

KD

knockdown

PDGF

platelet-derived growth factor

PGE₂

prostaglandin E2

PKI

PKA inhibitor peptide

p-VASP

phosphovasodilator-stimulated phosphoprotein

Scr

scramble

siRNA

small interfering RNA

VASP

vasodilator-stimulated phosphoprotein

AUTHOR CONTRIBUTIONS

J. V. Michael and A. Gavrila performed experiments, analyzed and interpreted data, and wrote the manuscript; D. A. Deshpande and R. B. Penn designed the studies, analyzed and interpreted the data, and wrote the manuscript; and A. P. Nayak, T. Pera, J. R. Liberato, S. R. Polischak, and S. D. Shah performed experiments, analyzed data, and provided comments contributing to the final content of the manuscript.