The deleterious role of the prostaglandin E₂ EP₃ receptor in angiotensin II hypertension

Abstract

Angiotensin II (ANG II) plays a key role in regulating blood pressure and inflammation. Prostaglandin E₂ (PGE₂) signals through four different G protein-coupled receptors, eliciting a variety of effects. We reported that activation of the EP₃ receptor reduces cardiac contractility. More recently, we have shown that overexpression of the EP₄ receptor is protective in a mouse myocardial infarction model. We hypothesize in this study that the relative abundance of EP₃ and EP₄ receptors is a major determinant of end-organ damage in the diseased heart. Thus EP₃ is detrimental to cardiac function and promotes inflammation, whereas antagonism of the EP₃ receptor is protective in an ANG II hypertension (HTN) model. To test our hypothesis, male 10- to 12-wk-old C57BL/6 mice were anesthetized with isoflurane and osmotic minipumps containing ANG II were implanted subcutaneously for 2 wk. We found that antagonism of the EP₃ receptor using L798,106 significantly attenuated the increase in blood pressure with ANG II infusion. Moreover, antagonism of the EP₃ receptor prevented a decline in cardiac function after ANG II treatment. We also found that 10- to 12-wk-old EP₃-transgenic mice, which overexpress EP₃ in the cardiomyocytes, have worsened cardiac function. In conclusion, activation or overexpression of EP₃ exacerbates end-organ damage in ANG II HTN. In contrast, antagonism of the EP₃ receptor is beneficial and reduces cardiac dysfunction, inflammation, and HTN.

NEW & NOTEWORTHY This study is the first to show that systemic treatment with an EP₃ receptor antagonist (L798,106) attenuates the angiotensin II-induced increase in blood pressure in mice. The results from this project could complement existing hypertension therapies by combining blockade of the EP₃ receptor with antihypertensive drugs.

INTRODUCTION

Angiotensin II (ANG II) is the major effector peptide of the renin-angiotensin system (87). ANG II plays an important role in both blood pressure regulation and inflammation along with other important biological functions (i.e., regulating cell growth and apoptosis, extracellular matrix deposition, cell migration, vascular motor tone). ANG II is released from circulating hepatic-derived angiotensinogen via the actions of renin and angiotensin-converting enzyme (21, 65, 83). ANG II is a potent vasoconstrictor, causing an increase in blood pressure and vasculature remodeling in chronic hypertension (HTN; Ref. 28). ANG II also promotes increased vascular permeability leading to increased inflammatory processes (9, 10, 86). The various actions of ANG II are mediated by specific intracellular signaling pathways downstream of ANG II binding to its surface receptors. In mammals, ANG II binds to two G protein-coupled receptors, ANG II types 1 (AT₁) and 2 (AT₂; Refs. 59, 80). The AT₁ receptor typically activates phospholipase C through the G_q protein. AT₁ is highly expressed in smooth muscle cells, whereas AT₂ receptors are present in the kidneys, pancreas, heart, adrenals, brain, and vasculature (90, 94, 95). Both receptors signal via complex intracellular signaling cascades, although it is the consensus that AT₂ antagonizes the effects of AT₁ (17, 100).

ANG II plays a critical role in cardiovascular disease (43). It has been shown that ANG II promotes pathological cardiac hypertrophy in both neonatal rat cardiomyocytes (6, 7, 77, 78) and adult cardiomyocytes (49, 73, 96). Cardiomyocytes express both AT₁ and AT₂ receptors, although there is an abundance of evidence to support that ANG II primarily signals through AT₁ in the cardiomyocytes (12, 13). Additionally, AT₂ receptors were undetectable in both neonatal and adult rat cardiac fibroblasts (18, 93). The actions of ANG II on the cardiac fibroblasts increase extracellular matrix proteins and collagen deposition, an important factor in pathological left ventricle remodeling.

In recent years, a link between the immune system and hypertension has been described (34, 75). In various cell types and tissues, ANG II can directly activate immune cells (i.e., chemotaxis and differentiation; Ref. 42), promote the transcription of proinflammatory mediators (i.e., chemokines and cytokines; Refs. 67, 86), and activate proinflammatory signaling pathways (i.e., NF-κB and MAPK; Refs. 8, 31). It has been reported that ANG II promotes the production of reactive oxygen species (ROS) through stimulation of NADPH oxidase, via signaling through the AT₁ receptor (25, 26). This increase in oxidative stress promotes endothelial dysfunction and vascular inflammation and contributes to increased activity of transcription factors like NF-κB combined with increases in chemokines, cytokines, and adhesion molecules (63, 84).

Prostaglandin E₂ (PGE₂) is produced from three known prostaglandin E synthase isoforms, a cytosolic form (cPGES) and two microsomal forms (mPGES-1 and mPGES-2), which are membrane bound. PGE₂ signals through four distinct G protein-coupled receptor subtypes (EP₁, EP₂, EP₃, and EP₄) to produce a variety of physiological and pathological effects. EP₂ and EP₄ increase cAMP levels in the cell via activation of adenylate cyclase, whereas EP₃ inhibits cAMP production and EP₁ increases Ca²⁺ levels in the cell (76, 85). However, the cardiac effects of these receptors have not been fully elucidated. We (32) have previously published that mPGES-1-null mice were more susceptible to the deleterious effects of chronic ANG II infusion on the heart, suggesting a cardioprotective role for PGE₂. Also, we (29) previously published that PGE₂ via its EP₃ receptor could reduce contractility of isolated myocytes and in the whole heart by mechanisms that appeared to involve decreased phospholamban phosphorylation. In a model of arachidonic acid (AA)-induced inflammation, Morimoto and colleagues (57) discovered that PGE₂ was the most abundantly produced AA metabolite and further showed in EP₃-null mice that PGE₂-EP₃ signaling was mediating the inflammatory response by increasing vascular permeability. Additionally, it has been reported that PGE₂ can directly trigger the degranulation of mast cells and subsequent release of proinflammatory IL-6 in mouse bone marrow-derived mast cells. This was shown to be dependent on EP₃-G_i signaling and by the influx of Ca²⁺ (5) and phosphatidylinositol 3-kinase (PI3K)-Akt pathways (45, 62). We (15) recently reported that activation of the EP₄ receptor in isolated adult cardiac fibroblasts prevented an increase in the proinflammatory chemokine MCP-5 in response to lipopolysaccharide stimulation by blocking PI3K-Akt signaling and subsequent NF-κB activation. Furthermore, we (14) have reported that overexpression of EP₄ in the cardiomyocyte improves cardiac function and reduces cardiac inflammation after myocardial infarction. Thus we hypothesize that the relative abundance of EP₃ and EP₄ receptors is a crucial determinant of end-organ damage in the diseased heart. We also hypothesize that overexpression of EP₃ is detrimental to cardiac function and promotes inflammation, whereas inhibition of the EP₃ receptor is protective in a mouse ANG II HTN model.

METHODS

Animal Use

For experimental protocol 1 (Mouse Experimental Protocols section below), male 10- to 12-wk-old C57BL/6 mice were purchased from Jackson. For protocol 2, male EP₃-transgenic (EP₃ Tg) and their wild-type littermates were kindly provided to the Harding Laboratory by Dr. Thomas Hohlfield at the University of Dusseldorf (Germany). All animal experiments were approved by the Henry Ford Health System Institutional Animal Care and Use Committee in accordance with federal guidelines.

Mouse Experimental Protocols

Effect of an EP₃ antagonist in ANG II HTN and in an isoproterenol model.

Male 10- to 12-wk-old C57BL/6 mice were treated with either vehicle or ANG II (1.4 mg·kg⁻¹·day⁻¹) for a period of 2 wk via osmotic minipump. Additionally, mice were treated with either vehicle (ethanol diluted in 0.9% NaCl) or an EP₃ antagonist (L798,106; 40 µg·kg⁻¹·day⁻¹ ip) daily. The dose of L798,106 was chosen based on the K_i values of the compound (0.3, 916, 5,000, and >5,000 nM for EP₃, EP₄, EP₁, and EP₂, respectively). In separate experiments, mice received infusion of isoproterenol (30 mg·kg⁻¹·day⁻¹) or vehicle (0.9% NaCl) for 10 days via osmotic minipump. Mice were also treated with either vehicle or an EP₃ antagonist (L798,106; 40 µg·kg⁻¹·day⁻¹ ip) daily.

Effect of EP₃ overexpression in ANG II HTN.

Male 10- to 12-wk-old EP₃ Tg mice overexpressing EP₃ in the cardiomyocytes and their wild-type littermates were treated with either vehicle (0.01 M acetic acid in 0.9% NaCl) or ANG II (1.4 mg·kg⁻¹·day⁻¹) for a period of 2 wk via osmotic minipump.

For experimental procedures, echocardiography was performed on conscious mice before surgery and at the conclusion of the study. Systolic blood pressure was monitored weekly to ensure that all animals receiving ANG II exhibited the expected increase in systolic blood pressure. At the end of the study, animals were anesthetized with pentobarbital sodium (100 mg/kg ip), and the heart was injected with 15% KCl, causing its arrest in diastole. The heart was excised and washed thoroughly in ice-cold Hanks’ balanced salt solution, followed by dissection of the atria and right ventricle. Hearts used for RNA isolation and real-time reverse transcription-polymerase chain reaction (RT-PCR) were stored in TRI Reagent at −80°C. For histological analysis, left ventricle (LV) sections were frozen in optimum cutting temperature media prechilled in isopentane and stored at −80°C. For protein extraction, LV were snap-frozen in liquid N₂ and stored at −80°C until future analysis.

Cell Treatment Protocols

Cardiomyocytes (AVM) and cardiac fibroblasts (AVF) were isolated from 10- to 12-wk-old C57BL/6 mice as previously described by us (33). AVM were serum-starved before pretreatment for 1 h with vehicle, PGE₂ (1 µM), or an EP₄ agonist (CAY10598; 1 µM). After 1 h, cells were treated with either vehicle or ANG II (10⁻⁷ M) for 24 h. Cells were maintained at 37°C, 5% CO₂. Similarly, AVF were serum-starved before pretreatment for 1 h with vehicle, PGE₂ (1 μM), or CAY10598 (0.01 or 1 μM). After 1 h, cells were treated with either vehicle or ANG II (10⁻⁷ M). Protein lysate for Western blot analysis was obtained by scraping the cells on ice with lysis buffer containing protease (cat. no. 4693159001; Roche) and phosphatase inhibitors (cat. no. 05892970001; Roche) and then flash-frozen in liquid N₂ and stored at −80°C.

Human coronary artery endothelial cells (HCECs) were purchased from American Type Culture Collection (ATCC; Manassas, VA). HCECs were maintained in Vascular Cell Basal Medium (cat. no. PCS-100-030; ATCC) containing the following components from the Endothelial Cell Growth Kit-BBE (cat. no. PCS-100-040; ATCC): bovine brain extract (0.2%), recombinant human EGF (5 ng/mL), l-glutamine (10 mM), heparin sulfate (0.75 U/mL), hydrocortisone hemisuccinate (1 µg/mL), ascorbic acid (50 µg/mL), and fetal bovine serum (2%). Cells were maintained at 37°C, 5% CO₂. For treatment, HCECs were serum-starved for 24 h before pretreatment with PGE₂ (1 µM) or CAY10598 (1 µM) for 1 h. After 1 h, HCECs were treated with vehicle or ANG II (10⁻⁷ M) for 24 h. Protein lysates for Western blot analysis were obtained by scraping the cells on ice with lysis buffer containing protease and phosphatase inhibitors and then flash-frozen in liquid N₂ and stored at −80°C. All protein lysates were sonicated on ice for 10 s at 20 kHz.

Blood Pressure Measurements

Systolic blood pressure was measured on a weekly basis using tail-cuff plethysmography. Pressure was obtained using a MC4000 Multi Channel Blood Pressure Analysis System for mice (Hatteras Instruments, Cary, NC). Briefly, mice were trained for at least 1 wk before analysis. They were maintained at 38–40°C and subjected to 10 cycles of blood pressure analysis. At least 7 valid readings in a cycle of 10 measurements were needed to accept the measurements. In addition, a minimum systolic value of 75 mmHg was set as an exclusion criterion. Blood pressure measurements always occurred at the same time of day to eliminate any variability due to the rodent circadian cycle. In separate experiments, blood pressure was confirmed by measuring mean arterial pressure in the carotid artery. Mice were anesthetized with 3% isoflurane and maintained at 0.8–1.0% during the surgical procedure. PE-10 tubing fused to PE-50 was inserted into the left carotid artery. Pressure recordings were acquired for an equilibration period of 15 min, followed by 10 min of data acquisition using LabScribe software v2 (iWorx, Dover, NH). Mean arterial pressure was averaged across the 10-min data acquisition period.

Echocardiography

Echocardiography was performed before surgery and at the conclusion of the study (2 wk after infusion). The cardiac function of all mice was assessed by echocardiography using an Acuson 256 system (Mountain View, CA) with a 15-MHz linear transducer, as reported previously (46, 66). All echocardiography assessment was performed on conscious mice to eliminate the depressant effect of anesthesia on respiration and cardiac function. Diastolic measurements were made at the maximum left ventricle cavity dimension, whereas systolic parameters were measured during maximum anterior motion of the posterior wall. All echocardiography was performed by the same investigator who was blinded to the genotype. Additionally, all mice were trained for 2–3 days before assessment to eliminate heart rate variability.

Implantation of Osmotic Minipumps

Osmotic minipumps contained ANG II, isoproterenol, or vehicle solution. ANG II was dissolved in 0.01 M acetic acid diluted in 0.9% NaCl saline and administered chronically at a dose of 1.4 mg·kg⁻¹·day⁻¹ sc (32). Control mice for these experiments received minipumps containing the vehicle (0.01 M acetic acid diluted in physiological 0.9% saline). Isoproterenol was dissolved in 0.9% NaCl and administered at a dose of 30 mg·kg⁻¹·day⁻¹ sc. As a control for the isoproterenol experiments, mice received minipumps containing the vehicle (0.9% NaCl). Mice at 10–12 wk of age were anesthetized with pentobarbital sodium (10 mg/kg), and osmotic minipumps were implanted under the skin subcutaneously. Postoperatively, mice were allowed a recovery period on a heated pad followed by administration of buprenorphine as an analgesic (0.05 mg/kg).

Histology

Mouse hearts were harvested and sectioned transversely into three parts from apex to base. The sections were frozen in prechilled isopentane and stored at −80°C for determination of myocyte cross-sectional area (MCSA). For MCSA, sections of the heart were stained with fluorescein-labeled peanut agglutinin to delineate the myocytes. Four radially oriented microscope fields were selected from each section and photographed under the ×20 objective. MCSA was measured by computer-based planimetry (NIH ImageJ) and averaged across all four fields of the sections. The mean area was then calculated for each animal. All assessments were performed by observers who were blinded to animal identification and treatment groups.

Immunohistochemistry of Macrophages and T Cells

The number of macrophages and T cells in the heart was assessed using CD68⁺ staining (rat anti-CD68⁺ antibody; Bio-Rad) and CD3⁺ staining (rabbit anti-CD3⁺ antibody; Abcam), respectively, on frozen heart sections as we (14) previously described. Briefly, frozen sections were fixed in acetone for 10 min, followed by incubation with fresh 0.3% hydrogen peroxide for 30 min. Sections were then blocked in 5% centrifuged milk in Tris-buffered saline (TBS) and incubated with either anti-CD68⁺ or anti-CD3⁺ overnight at 4°C. Biotinylated secondary antibodies were incubated on sections for 1 h at room temperature. Sections were then incubated with a ready-to-use horseradish peroxidase-streptavidin reagent (Vector, Burlingame, CA) for 40 min at room temperature followed by aminoethyl carbazole single solution (Vector). Slides were rinsed in water and counterstained with Harris hematoxylin solution for 1 min and then rinsed again in water and mounted onto slides using an aqueous mounting solution (Aqua-Mount; Lerner). Negative controls consisted of sections incubated in the absence of primary antibody. Photographs of four randomly chosen fields per section were taken under ×20 objective, and the number of positively stained cells (per square millimeter) was measured by a blinded observer using NIH ImageJ software.

Real-Time Reverse Transcription-Polymerase Chain Reaction

Measurement of mRNA expression was performed by real-time reverse transcription-polymerase chain reaction (RT-PCR) using a SYBR Green method as follows. Two micrograms of DNase I-treated RNA sample was reverse transcribed using random primers and the Omniscript reverse transcriptase kit (Qiagen, Valencia, CA). Two microliters of the reverse transcription reaction was amplified in a Roche version 2.0 LightCycler PCR machine (Indianapolis, IN) using SYBR Green dye and specific primers against interleukin-1β (Il1b), tumor necrosis factor-α (Tnfa), prostaglandin E₂ receptor subtype 3 (Ptger3), prostaglandin E₂ receptor subtype 4 (Ptger4), monocyte chemoattractant protein-5 (Mcp5), interleukin-10 (Il10), arginase-1 (Arg1), or the reference gene glyceraldehyde-3-phosphate dehydrogenase (Gapdh). Fold induction of mRNA transcripts was calculated by correcting threshold cycle (CT) values to GAPDH using the 2^-ΔΔCT method (50). Il1b, Tnfa, and Ptger3 primers were from Qiagen. Mcp5, Il10, Arg1, and Gapdh primers were synthesized by TIB MOLBIOL (Adelphia, NJ). All primer sequences are provided in Table 1.

Table 1.

Primer sequences

ID	Sense	Antisense	Acc. No.
Il1b			NM_008361
Tnfa			NM_013693.3
Ptger3			NM_011196.2
Ptger4	5ʹ-gtgcagagatccagatggtca-3ʹ	5ʹ-atcygggtttctgctgatgtc-3ʹ
Mcp5	5ʹ-caagagratcaccagcagcagg-3ʹ	5ʹ-tgcttgaggtggttgtggaa-3ʹ
Il10	5ʹ-gctgcctgctcttactgact-3ʹ	5ʹ-ctgggaagtgggtgcagtta-3ʹ
Arg1	5ʹ-gaacacggcagtggctttaac-3ʹ	5ʹ-tgcttagctctgtctgctttgc-3ʹ
Gapdh	5ʹ-caaggtcatcccagagctg-3ʹ	5ʹ-tgtcatcatacttggcaggtt-3ʹ

Primer sequences for Il1b, Tnfa, and Ptger3 are from Qiagen and are proprietary, thus only the accession numbers for the mRNA sequence are available. ID, identifier.

Western Blot Analysis

Anti-RhoB (cat. no. 2098; 1:1,000), anti-RhoA (cat. no. 2117; 1:1,000), and anti-GAPDH (cat. no. 5174; 1:2,000) were purchased from Cell Signaling Technology (Danvers, MA). Anti-Nox2/gp91 (cat. no. 611415; 1:1,000) was purchased from BD Transduction/BD Biosciences (San Jose, CA). For use with HCECs, anti-Nox2/gp91 from EMD Millipore (cat. no. 07-024; 1:1,000; Burlington, MA) was used. Protein levels were measured by Western blot using a method previously described by us (29). For mouse studies, all results were corrected to GAPDH and expressed as fold of vehicle or wild type. For AVM and HCECs experiments, results were corrected to GAPDH and expressed as fold of vehicle control.

Multiplex ELISArray Analysis

For the simultaneous detection of numerous cytokines or chemokines, we performed a multianalyte enzyme-linked immunosorbent assay (ELISA) using a commercially available kit from Qiagen (cat. no. MEM-005A) according to the manufacturer’s instructions and as previously described by us (14). The limit of detection for the panel of cytokines and chemokines is as follows (in pg/mL): IL-1β, 55.2; IL-4, 32.8; IL-6, 58.8; IL-10, 59.4; IL-12, 21.1; IL-17A, 83.3; IFN-γ, 39.0; TNF-γ, 30.5; TGF-β₁, 85.4; MCP-1, 25.5; MIP-1γ, 35.1; MIP-1β, 26.7.

Chemicals

EP₄ agonist (CAY10598) and prostaglandin E₂ (PGE₂) were from Cayman Chemical (Ann Arbor, MI). EP₃ antagonist (L798,106; K_i values are 0.3, 916, >5,000 and >5,000 nM at EP₃, EP₄, EP₁, and EP₂, respectively) was from Tocris Bioscience (Minneapolis, MN). Angiotensin II was from Bachem Bioscience (Torrance, CA). TRI Reagent was from Molecular Research Center (Cincinnati, OH). Vascular Cell Basal Medium and the Endothelial Cell Growth Kit-BBE were from ATCC. All other compounds were obtained from Sigma-Aldrich (St. Louis, MO).

Statistical Analysis

All statistics were performed by a statistician in the Department of Public Health Sciences of Henry Ford Hospital using the statistical package SAS version 9.4. For all tests, a two-sample Wilcoxon test with the Fligner–Policello correction for unequal variances was used. We also used Hochberg method for multiple testing. A P value <0.05 was considered as evidence of a statistically significant difference for experimental data with the P values being two-sided unless otherwise stated.

RESULTS

Expression of the EP₃ Receptor Increases after ANG II Infusion

We (29) have previously shown that mRNA expression of the EP₃ receptor increases in the infarcted heart and activation of EP₃ leads to reduced cardiac function. To determine expression levels of the EP₃ and EP₄ receptors in the left ventricle and to examine whether EP₃ receptor expression increases in the ANG II HTN model, we implanted osmotic minipumps containing either vehicle or ANG II for 2 wk. Figure 1A shows the expression of both receptors for both time points. After 2 wk, the expression of EP₃ increased 2.5-fold compared with vehicle-infused animals (1.02 ± 0.13 in vehicle-treated vs. 2.59 ± 0.63 in ANG II-treated mice; P < 0.05), whereas expression of the EP₄ receptor was unchanged [1.01 ± 0.10 in vehicle-treated vs. 0.96 ± 0.11 in ANG II-infused mice; P = not significant (ns)]. In additional experiments, we sought to examine whether the changes in EP₃ expression were occurring in cardiomyocytes or fibroblasts, the two major cell types of the heart. We infused ANG II or vehicle control for 2 wk and then isolated cardiomyocytes (AVM) and fibroblasts (AVF) from each animal simultaneously. Real-time RT-PCR was used to determine Ptger3 (EP₃) and Ptger4 (EP₄) mRNA expression. Figure 1B shows that EP₃ mRNA expression increases ~5-fold in the AVM (5.14 ± 1.52 in ANG II vs. 1.20 ± 0.46 in vehicle-treated; P < 0.05), whereas it increases only 25% in the fibroblasts (1.25 ± 0.05 in ANG II vs. 1.00 ± 0.07; P < 0.05). Figure 1C shows there was no significant increase in EP₄ in the AVM (2.08 ± 0.92 in ANG II vs. 1.08 ± 0.28 in vehicle-treated; P = ns); however, EP₄ significantly increased in the fibroblasts in response to ANG II (1.80 ± 0.05 in ANG II vs. 1.01 ± 0.09 in vehicle; P < 0.005).

Fig. 1.

A: real-time RT-PCR analysis of Ptger3 and Ptger4 mRNA expression in the left ventricle after 2 wk of infusion with vehicle or angiotensin II (ANG II; 1.4 mg·kg⁻¹·day⁻¹) via osmotic minipumps. B and C: real-time RT-PCR analysis of Ptger3 (B) and Ptger4 (C) mRNA expression in adult cardiomyocytes (AVM) or adult cardiac fibroblasts (AVF) after 2 wk of vehicle or ANG II infusion (1.4 mg·kg⁻¹·day⁻¹). *P < 0.05, ***P < 0.005 vs. respective vehicle. n = 3 Mice per group. For statistical analysis, a 2-sample Wilcoxon test with the Fligner–Policello correction for unequal variances was used.

EP₃ Antagonist L798,106 Improves Cardiac Function and Hypertrophy in ANG II Hypertension

Since the expression of the EP₃ receptor increases in ANG II hypertension, we hypothesized that antagonism of the EP₃ receptor would prevent a decline in cardiac function. L798,106 is a potent and specific antagonist of the EP₃ receptor (38). Figure 2A shows that there was a modest but significant reduction in ejection fraction after 2 wk of ANG II and this was attenuated when animals were treated with L798,106 (77.26 ± 0.47% in vehicle-treated vs. 69.82 ± 1.01% in ANG II-treated, P < 0.005; and 72.61 ± 0.51% in ANG II + L798,106-treated, P < 0.05). Similarly, Fig. 2B shows that shortening fraction was reduced with ANG II and improved with L798,106 treatment (51.61 ± 1.71% in ANG II-treated vs. 56.97 ± 1.01% in ANG II + L798,106-treated; P < 0.05). Moreover, Fig. 2C shows there was a significant increase in posterior wall thickness at systole after ANG II infusion, which was attenuated with an EP₃ antagonist (1.19 ± 0.02 mm in vehicle-treated vs. 1.42 ± 0.02 mm in ANG II-treated, P < 0.01; and 1.30 ± 0.03 mm in ANG II + L798,106-treated, P < 0.05). We did not observe any change in cardiac output (Fig. 2D) between the groups (14.43 ± 1.33 mL·min⁻¹·10 g⁻¹ in vehicle-treated vs. 12.38 ± 0.98 mL·min⁻¹·10 g⁻¹ in ANG II-treated, P = ns; and 13.12 ± 1.10 mL·min⁻¹·10 g⁻¹ in ANG II + L798,106-treated, P = ns).

Fig. 2.

Cardiac function assessed by echocardiography in conscious mice 2 wk after vehicle or angiotensin II (ANG II; 1.4 mg·kg⁻¹·day⁻¹) ± daily injections of L798,106 (40 µg·kg⁻¹·day⁻¹) or diluted ethanol vehicle control. Ejection fraction (EF; A), shortening fraction (SF; B), posterior wall thickness at systole (PWT_s; C), and cardiac output (CO; D) are shown. **P < 0.01, ***P < 0.005 vs. vehicle; +P < 0.05, +++P < 0.005 vs. ANG II. n = 11–13 Mice. E: heart weight (HW)-to-body weight (BW) ratio. Mice were anesthetized, and hearts were removed, weighed, and presented as milligrams per 10 g body wt. *P < 0.05 vs. vehicle. F: systolic blood pressure measured by tail-cuff plethysmography in conscious mice at baseline and 1 and 2 wk after vehicle or ANG II (1.4 mg·kg⁻¹·day⁻¹) ± daily injections of L798,106 (40 µg·kg⁻¹·day⁻¹) or diluted ethanol vehicle control for 2 wk. ***P < 0.005 vs. vehicle, +P < 0.05 vs. ANG II. n = 11–13 Mice per group. G: mean arterial pressure (MAP) measured in the carotid artery 2 wk after vehicle or ANG II (1.4 mg·kg⁻¹·day⁻¹) ± daily injections of L798,106 (40 µg·kg⁻¹·day⁻¹) or diluted ethanol vehicle. *P < 0.05 vs. vehicle, ++P < 0.01 vs. ANG II. n = 4–8 Mice per group. For statistical analysis, a 2-sample Wilcoxon test with the Fligner–Policello correction for unequal variances was used. To correct for multiple testing, we used Hochberg correcting method.

ANG II is known to produce cardiac hypertrophy (32), and we observed an increase in heart weight-to-body weight ratio (Fig. 2E) with ANG II infusion (39.86 ± 1.08 in vehicle-treated vs. 45.31 ± 1.68 in ANG II-treated; P < 0.05). This increase was prevented when animals were treated with L798,106 (36.92 ± 1.74 in vehicle + L798,106-treated vs. 41.08 ± 1.31 in ANG II + L798,106; P = 0.06, P = ns).

Treatment with L798,106 Significantly Diminishes Pressor Response of ANG II

To determine the effect of L798,106 treatment in ANG II HTN, we measured blood pressure by tail cuff (Fig. 2F). As expected, ANG II infusion increased blood pressure significantly by 1 wk, and this remained elevated throughout the duration of the study (113 ± 3.2 mmHg in vehicle-treated at 1 wk vs. 151 ± 9.9 mmHg in ANG II-treated at 1 wk, P < 0.005; and 110 ± 2.7 mmHg in vehicle-treated at 2 wk vs. 168 ± 5.0 mmHg in ANG II-treated at 2 wk, P < 0.005). Daily administration of the EP₃ antagonist L798,106 substantially attenuated the hypertensive effect of chronic ANG II treatment after 2 wk (168 ± 5.0 mmHg in ANG II to 136 ± 9.7 mmHg in ANG II + L798,106; P < 0.05). Since there was no significant effect of L798,106 on cardiac output, this suggested that the effect of L798,106 on blood pressure was not occurring via alterations in cardiac function. To confirm the blood pressure results, we repeated these experiments and measured mean arterial pressure (MAP) by carotid artery catheterization (Fig. 2G). MAP was increased with 2 wk of ANG II infusion, as expected (90 ± 4.79 in vehicle vs. 104 ± 1.08 in ANG II; P < 0.05). This increase was significantly reduced when animals were treated with L798,106 (94 ± 3.27 in ANG II + L798,106 vs. 104 ± 1.08 in ANG II; P < 0.01). These data confirm the tail-cuff blood pressure results.

L798,106 has no Effect on Blood Pressure or Cardiac Function with β-Adrenergic Stimulation

Infusion of the β-adrenergic receptor agonist, isoproterenol, is a nonischemic model that has been used to induce heart failure in rodents (1, 69, 88, 97, 103). To determine whether blockade of the EP₃ receptor would ameliorate the effects of isoproterenol similar to ANG II, we infused isoproterenol for 10 days along with L798,106 treatment. Figure 3A shows that when mice received isoproterenol infusion for 10 days, there was a significant increase in heart rate, as expected (594 ± 24 beats/min in vehicle-treated vs. 739 ± 33 beats/min in isoproterenol-treated; P < 0.005). This was evident as early as 4 days after treatment. Treatment with an EP₃ antagonist had no effect on this increase (670 ± 37 beats/min in isoproterenol + L798,106 group vs. 739 ± 33 beats/min in isoproterenol group; P = ns). Figure 3B shows that isoproterenol infusion also resulted in a significant, albeit modest increase in systolic blood pressure compared with vehicle-infused mice (122 ± 2.8 mmHg in isoproterenol vs. 108 ± 4.4 mmHg in vehicle; P < 0.05). EP₃ antagonist treatment did not reduce this increase in blood pressure (114 ± 7.3 mmHg in isoproterenol + L798,106 vs. 122 ± 2.8 mmHg in isoproterenol; P = ns). Cardiac function was assessed in these mice by echocardiography. Figure 3C shows that there was a modest but significant reduction in shortening fraction (55.16 ± 1.3% in isoproterenol vs. 61.81 ± 1.0% in vehicle; P < 0.005). However, this reduction was not altered when mice were treated with L798,106 (53.87 ± 1.2% in isoproterenol + L798,106 vs. 55.16 ± 1.3% in isoproterenol; P = ns). Figure 3D also shows a significant increase in left ventricular dimension at systole and a corresponding increase in left ventricle mass (Fig. 3E). Neither of these parameters of hypertrophy was attenuated when mice received L798,106 treatment. Altogether, these data suggest the protective effects of EP₃ antagonism are unique to the ANG II model and may not be relevant in other nonischemic models of heart failure.

Fig. 3.

Heart rate (A) and systolic blood pressure (B) measured by tail-cuff plethysmography in conscious mice throughout the infusion protocol of vehicle or isoproterenol (30 mg·kg⁻¹·day⁻¹) ± daily injections of L798,106 (40 µg·kg⁻¹·day⁻¹) or diluted ethanol vehicle control. *P < 0.05, ***P < 0.005 vs. vehicle. n = 4–6 Mice per group. Cardiac function was assessed by echocardiography in conscious mice at 10–12 wk of age 10 days after vehicle or isoproterenol (ISO) infusion (30 mg·kg⁻¹·day⁻¹). Shortening fraction (SF; C) and left ventricular dimension at systole (LVD_s; D) are shown. E: left ventricle-to-body weight (BW) ratio was obtained using heart weights at euthanasia. Mice were anesthetized, and hearts were removed, weighed, and presented as milligrams per 10 g body wt. **P < 0.01, ***P < 0.005 vs. vehicle. n = 7–8 Mice. For statistical analysis, a 2-sample Wilcoxon test with the Fligner–Policello correction for unequal variances was used. To correct for multiple testing, we used Hochberg correcting method. bpm, Beats per minute.

Treatment with L798,106 Does Not Alter Cardiac AT₁ mRNA Expression

It is known that upregulation of the AT₁ receptor plays an important role in mediating the pathophysiology of increased blood pressure (55). Therefore, we examined whether L798,106 administration is affecting mRNA levels of the ANG II AT₁ receptor in the left ventricle. Figure 4A shows that after ANG II infusion there was no change in AT₁ mRNA levels (1.02 ± 0.09 in vehicle-treated vs. 1.16 ± 0.13 in ANG II-treated; P = ns). Similarly, there was no significant change between mice that received ANG II infusion with vehicle injection and those that received ANG II infusion with L798,106 treatment (1.16 ± 0.13 in ANG II-treated vs. 0.86 ± 0.10 in ANG II + L798,106-treated; P = ns).

Fig. 4.

A: real-time RT-PCR analysis of angiotensin II (ANG II) type 1 receptor (AT₁) mRNA expression in left ventricle after 2 wk of ANG II infusion ± L798,106 (40 µg·kg⁻¹·day⁻¹ ip). n = 6 Mice. B: Western blot analysis for RhoB expression in mouse mesenteric arteries after 2 wk of vehicle or ANG II (1.4 mg/kg/day) ± L798,106 (40 µg/kg/day ip). ***P < 0.005 vs. vehicle, +++P < 0.005 vs. ANG II. n = 3–5 Mice per group. C, top, shows a representative Western blot for RhoA and the housekeeping protein GAPDH. Bottom shows quantitative analysis for RhoA expression in mouse mesenteric arteries (Mes. Art). D, top, shows a representative Western blot for RhoA and the housekeeping protein GAPDH in left ventricle homogenates after 2 wk of vehicle (Veh) or ANG II (1.4 mg⁻¹·kg⁻¹·day⁻¹) ± L798,106 (L798; 40 µg⁻¹·kg⁻¹·day⁻¹). Bottom shows quantitative analysis for RhoA expression in mouse left ventricle (LV). *P < 0.05 vs. vehicle. n = 3 Mice per group. For statistical analysis, a 2-sample Wilcoxon test with the Fligner–Policello correction for unequal variances was used. To correct for multiple testing, we used Hochberg correcting method.

Treatment with L798,106 Reduces ANG II-Induced RhoB Expression in Mesenteric Arteries

Changes in Rho-kinase signaling have been shown to contribute to increases in total peripheral resistance in hypertension, through alterations in resistance vessels such as the mesenteric arteries (4, 91, 98). Western blot analysis using mesenteric arteries from the mice revealed a significant increase in RhoB protein expression after 2 wk of ANG II infusion (Fig. 4B; 1.00 ± 0.14 in vehicle-treated vs. 3.84 ± 1.39 in ANG II-treated; P < 0.005). This was dramatically attenuated when mice were treated with L798,106 (0.94 ± 0.32 in ANG II + L798,106 vs. 3.84 ± 1.39 in ANG II-treated; P < 0.005). These results suggest that ANG II increases RhoB expression, presumably increasing contraction in the vasculature. RhoB expression is reduced by L798,106 treatment, possibly serving as a mechanism for the reduction in blood pressure we observed. RhoB expression was undetectable in left ventricle homogenates 2 wk after ANG II infusion.

Figure 4C shows that in contrast to RhoB, RhoA expression was unchanged in the mesenteric arteries with infusion of ANG II (1.00 ± 0.20 in vehicle vs. 1.08 ± 0.40 in ANG II; P = ns). Although expression of RhoB was undetectable in the LV, RhoA expression was detectable in the LV homogenates. Interestingly, there was a small but significant decrease in RhoA expression with ANG II infusion (Fig. 4D), and this was unchanged with L798,106 treatment (1.00 ± 0.03 in vehicle vs. 0.82 ± 0.04 in ANG II, P < 0.05; 0.82 ± 0.04 in ANG II vs. 0.91 ± 0.14 in ANG II + L798,106, P = ns). This significant reduction in RhoA expression is likely not physiologically relevant and does not explain changes in blood pressure with ANG II infusion.

EP₃ Overexpression Reduces Cardiac Function

Since the increase in the EP₃ receptor in ANG II HTN is occurring mainly in the cardiomyocytes, we acquired mice that overexpress the porcine analog of human EP₃ in the cardiomyocytes (EP₃ Tg) and their wild-type littermates (WT). To assess the effect of EP₃ overexpression in the cardiomyocytes on blood pressure, we measured systolic blood pressure in EP₃ Tg and their WT littermates under baseline conditions and after infusion of ANG II for 2 wk. Figure 5A shows that under baseline conditions, the EP₃ Tg mice have normal systolic blood pressure and this increases significantly after ANG II infusion, albeit to a similar extent to that observed in C57BL/6 mice (106 ± 2.01 mmHg at baseline vs. 152 ± 3.13 mmHg after ANG II infusion; P < 0.05). These data suggest that overexpression of EP₃ in the cardiomyocytes has no direct effect on blood pressure regulation.

Fig. 5.

A: systolic blood pressure measured by tail-cuff plethysmography in conscious EP₃-transgenic (EP₃ Tg) mice 2 wk after vehicle or angiotensin II (ANG II) infusion (1.4 mg·kg⁻¹·day⁻¹). ***P < 0.005 vs. vehicle. n = 4–5 Mice per group. Cardiac function was assessed by echocardiography in conscious male wild-type (WT) or EP₃ Tg mice at 10–12 wk of age under baseline conditions. Ejection fraction (EF; B), shortening fraction (SF; C), left ventricular dimension at diastole (LVD_d; D), and posterior wall thickness at systole (PWTs; E) are shown. **P < 0.01 vs. WT. n = 3 WT and n = 9 EP₃ Tg mice. F: myocyte cross-sectional area (MCSA) analysis. Sections of left ventricle were stained with peanut agglutinin to outline the cells. Myocyte area was quantified using NIH ImageJ software. ***P < 0.005 vs. WT. n = 5 Mice. G: representative MCSA images obtained from cross section of left ventricles stained with peanut agglutinin under ×20 objective. Scale bar is indicated in the lower right-hand corner of each panel. For statistical analysis, a 2-sample Wilcoxon test with the Fligner–Policello correction for unequal variances was used.

Figure 5B shows that as early as 10–12 wk of age, EP₃ Tg mice have reduced ejection fraction (56.93 ± 4.94% in EP₃ Tg vs. 76.18 ± 1.25% in WT; P < 0.05) and reduced shortening fraction at baseline (Fig. 5C; 38.49 ± 3.46% in EP₃ Tg vs. 61.69 ± 2.93% in WT; P < 0.01). Additionally, EP₃ Tg hearts display robust dilatation as shown by the left ventricular dimensions in diastole (Fig. 5D; 3.55 ± 0.23 mm in EP₃ Tg vs. 2.62 ± 0.01 mm in WT; P < 0.01). The posterior wall thickness is unchanged between EP₃ Tg and WT (Fig. 5E), thus these data suggest that the hearts display a phenotype of eccentric hypertrophy (1.13 ± 0.02 mm in EP₃ Tg vs. 1.21 ± 0.01 mm in WT; P = ns).

To measure the hypertrophy of the EP₃ Tg hearts, we performed MCSA analysis on frozen sections of LV at baseline. Figure 5F shows that the average myocyte area is increased in EP₃ Tg compared with the WT littermates (191.4 ± 2.6 in WT vs. 218.9 ± 2.8 in EP₃ Tg; P < 0.005). These data suggest that the EP₃ Tg hearts have hypertrophy under baseline conditions and confirm the echocardiography data showing dilatation of the ventricles.

We hypothesized that the decline in cardiac function observed in EP₃ Tg is exacerbated in ANG II HTN. Figure 6A shows that in response to ANG II, the hearts of WT mice are well compensated and maintain ejection fraction (76.49 ± 0.70% in vehicle vs. 75.30 ± 0.80% in ANG II; P = ns) and shortening fraction (Fig. 6B; 63.19 ± 0.98% in vehicle vs. 63.21 ± 1.37% in ANG II; P = ns). In contrast, the EP₃ Tg mice have reduced ejection fraction and shortening fraction at baseline and cardiac function in response to ANG II is worse than control mice (ejection fraction: 75.30 ± 0.83% in WT + ANG II vs. 45.92 ± 9.63% in EP₃ Tg + ANG II, P < 0.01; shortening fraction: 63.21 ± 1.37% in WT + ANG II vs. 34.77 ± 7.33% in EP₃ Tg + ANG II, P < 0.01). Similarly, there was an increase in left ventricle dimensions at systole (Fig. 6C) and diastole (Fig. 6D; 0.88 ± 0.04 in WT + ANG II vs. 2.40 ± 0.54 in EP₃ Tg + ANG II, P < 0.005; and 2.42 ± 0.05 in WT + ANG II vs. 3.50 ± 0.42 in EP₃ Tg + ANG II, P < 0.005, respectively).

Fig. 6.

Cardiac function assessed by echocardiography in conscious male wild-type (WT) or EP₃-transgenic (EP₃ Tg) mice at 10–12 wk of age after 2 wk of vehicle (Veh) or angiotensin II (ANG II) infusion (1.4 mg·kg⁻¹·day⁻¹). Ejection fraction (EF; A), shortening fraction (SF; B), left ventricular dimension at systole (LVDs; C), and left ventricular dimension at diastole (LVD_d; D) are shown. *P < 0.05, ***P < 0.005 vs. WT + vehicle; $$P < 0.01 vs. WT + ANG II. n = 3–5 Mice per group. For statistical analysis, a 2-sample Wilcoxon test with the Fligner–Policello correction for unequal variances was used. To correct for multiple testing, we used Hochberg correcting method.

EP₃-Transgenic Mice Have Increased Markers of Inflammation in Their Left Ventricle

We hypothesized that the depressed cardiac function observed in EP₃ Tg mice occurs with a concomitant increase in cardiac inflammation. We performed real-time RT-PCR for several proinflammatory cytokines and chemokines in left ventricles from EP₃ Tg and their WT littermates under baseline conditions. Figure 7A shows that there was a significant increase in TNF-α mRNA levels in EP₃ Tg mice (1.01 ± 0.08 in WT vs. 1.45 ± 0.22 in EP₃ Tg; P < 0.05). Additionally, we observed an increase in IL-1β (Fig. 7B; 1.04 ± 0.21 in WT vs. 1.87 ± 0.46 in EP₃ Tg; P < 0.05). Interestingly, the mRNA levels of the chemokine MCP-5 were unchanged between the two groups (Fig. 7C; 1.02 ± 0.15 in WT vs. 1.10 ± 0.06 in EP₃ Tg; P = ns). We also observed a significant reduction in the anti-inflammatory cytokine IL-10 (Fig. 7D; IL-10: 1.01 ± 0.11 in WT vs. 0.71 ± 0.02 in EP₃ Tg; P < 0.05). There was no statistically significant difference in the anti-inflammatory M2 macrophage marker arginase-1 (Fig. 7E; 1.02 ± 0.13 in WT vs. 0.80 ± 0.07 in EP₃ Tg; P = 0.26). Together, these data suggest that the EP₃ Tg mice have increased proinflammatory signaling in their left ventricles by upregulating proinflammatory genes and downregulating anti-inflammatory genes. To confirm our real-time RT-PCR results, we examined the protein expression of several cytokines and chemokines using a multiplex ELISA array. Figure 7F shows that there was a reduction in IL-10, confirming our real-time RT-PCR results [0.04 ± 0.01 arbitrary units (AU) in WT vs. 0.02 ± 0.01 AU in EP₃ Tg; P < 0.05]. There was also a significant reduction in IL-4 in the EP₃ Tg mice (Fig. 7G; 0.03 ± 0.01 AU in WT vs. 0.02 ± 0.0 AU in EP₃ Tg; P < 0.005). Levels of other cytokines were below the limits of detection for the assay.

Fig. 7.

Real-time RT-PCR analysis of cytokines in the left ventricles of wild-type (WT) and EP₃-transgenic (EP₃ Tg) mice under baseline conditions. Tumor necrosis factor-α (Tnfa; A), interleukin-1β (Il1b; B), monocyte chemoattractant protein-5 (Mcp5; C), interleukin-10 (Il10; D), and arginase-1 (Arg1; E) are shown. *P < 0.05 vs. WT. n = 3 Mice per group. Multiplex ELISA array analysis of interleukin-10 (IL-10; F) and interleukin-4 (IL-4; G) in the left ventricles of WT and EP₃ Tg mice under baseline conditions is shown. Values were corrected for the total micrograms of protein per well. *P < 0.05, ***P < 0.005 vs. WT. n = 3 Mice per group. For statistical analysis, a 2-sample Wilcoxon test with the Fligner–Policello correction for unequal variances was used. Abs, absorbance.

EP₃-Transgenic Mice Have Increased T Cell Infiltration

Since the EP₃ Tg mice have increased markers of inflammation with a reduction in anti-inflammatory markers, we hypothesized that this would result in increased infiltration of immune cells. In this study, we investigated the presence of CD68⁺ macrophages and CD3⁺ T cells in the left ventricle of WT and EP₃ Tg mice under baseline conditions by immunohistochemistry. Figure 8B shows that there was no change in the number of macrophages in the left ventricle of EP₃ Tg mice compared with WT littermates (90.8 ± 4.0 cells per field in WT vs. 80.0 ± 2.4 cells per field in EP₃ Tg; P = ns). However, Fig. 8D shows that there was a significantly higher number of CD3⁺ T cells in EP₃ Tg left ventricles (17.2 ± 1.3 cells per field in WT vs. 33.5 ± 3.0 cells per field in EP₃ Tg; P < 0.005). These data confirm our findings with the cytokine/chemokine mRNA levels as a lack of chemokines specific for macrophages like MCP-5 correlates with a lack of increased macrophages. Similarly, the increase in IL-1β could attribute to the increase in T cells.

Fig. 8.

A: representative immunohistochemistry sections for macrophage (CD68⁺) staining in frozen sections of left ventricle under baseline conditions. B: representative immunohistochemistry for T cell (CD3) staining in left ventricle sections under baseline conditions. C: quantitative CD68 analysis for WT and Tg mice. D: quantitative CD3⁺ analysis. Images were taken under ×20 objective and analyzed using ImageJ software. The number of cells were counted by a blinded observer and shown as number of positive cells per field. Data are presented as means ± SE. ***P < 0.005 vs. WT. n = 5–10 Mice per group. Scale bars are indicated in the lower right-hand corner of each panel. For statistical analysis, a 2-sample Wilcoxon test with the Fligner–Policello correction for unequal variances was used.

EP₃-Transgenic Mice Have Increased Expression of NADPH Oxidase 2 (Nox2/gp91)

Left ventricle hypertrophy and heart failure have been associated with an increase in ROS in various models (58, 72). To determine whether the EP₃ Tg mice have increased production of ROS in their left ventricles, we performed Western blot for Nox2, a major enzyme responsible for ROS production. Figure 9A shows that compared with WT, EP₃ Tg mice have a twofold increase in Nox2 expression in their left ventricles (1.00 ± 0.19 AU in WT vs. 2.02 ± 0.42 AU in EP₃ Tg; P < 0.05). In contrast, expression of Nox4 was not altered between the two strains. Although ROS generation was not measured per se, increased Nox2 expression suggests there is increased oxidative stress in the EP₃ Tg mice.

Fig. 9.

A: Western blot analysis for Nox2 in left ventricles of wild-type (WT) and EP₃-transgenic (EP₃ Tg) mice. Top is a representative Western blot showing Nox2/gp91 expression and the housekeeping protein GAPDH. Bottom displays quantitative analysis of Nox2 expression. Nox2 was corrected for protein loading by GAPDH and presented as fold of WT. *P < 0.05 vs. WT. n = 3 Mice per group. Western blot analysis for Nox2/gp91 expression in angiotensin II (ANG II)-treated (0.1 µM) adult cardiac fibroblasts in the presence of EP₄ agonist (Ag.) CAY10598 (0.01 or 1 µM; B) or PGE₂ (0.01 or 1 µM; C) is shown. Nox2 results were corrected to GAPDH and presented as fold of vehicle. ***P < 0.005 vs. vehicle; +P < 0.05, +++P < 0.005 vs. ANG II. n = 4–7 Mice per group. For statistical analysis, a 2-sample Wilcoxon test with the Fligner–Policello correction for unequal variances was used. To correct for multiple testing, we used Hochberg correcting method.

EP₄ Agonist Reduces ANG II-Induced Nox2/gp91 Expression in Vitro

Since the EP₃ Tg mice have increased levels of Nox2, we sought to determine whether activation of the EP₄ receptor could antagonize the ANG II-induced increase in Nox2. Interestingly, ANG II did not stimulate Nox2 expression in AVM. For this reason, we analyzed the effects of EP₄ activation in AVF and HCECs. Figure 9B shows that in AVF, treatment with ANG II (10⁻⁷ M) stimulated Nox2 expression (2.34 ± 0.10 AU in ANG II-treated vs. 1.00 ± 0.0 AU in vehicle-treated; P < 0.005). This increase was attenuated when AVF was pretreated with 1 µM CAY10598 (0.78 ± 0.08 AU in ANG II + CAY10598-treated vs. 2.34 ± 0.10 AU in ANG II-treated; P < 0.005) or with 0.01 µM CAY10598 (0.97 ± 0.10 AU in ANG II + CAY10598-treated vs. 2.34 ± 0.10 AU in ANG II-treated; P < 0.005). Similarly, Fig. 9C shows that pretreatment with 1 or 0.01 µM PGE₂ significantly reduced ANG II-induced Nox2 expression (2.48 ± 0.50 AU in ANG II-treated vs. 0.45 ± 0.09 AU in ANG II + 1 µM PGE₂-treated, P < 0.005; and 2.48 ± 0.50 AU in ANG II-treated vs. 0.73 ± 0.15 in ANG II + 0.01 µM PGE₂-treated, P < 0.01). Additionally, there was no significant effect of either PGE₂ or CAY10598 alone on Nox2 expression under basal conditions. These data suggest that activation of the EP₄ receptor can potentially reduce reactive oxygen species production by inhibiting Nox2 expression in AVF.

Conflicting with the data from AVF, treatment with ANG II did not produce a significant increase in Nox2 in HCECs (0.83 ± 0.16 AU to 1.07 ± 0.14 AU). However, Nox2 expression in the presence of ANG II was substantially attenuated by pretreatment with the EP₄ agonist (1.07 ± 0.14 AU vs. 0.54 ± 0.10 AU; P < 0.05). Similar results were observed with PGE₂ (1.07 ± 0.14 AU vs. 0.72 ± 0.19 AU; P = ns). Additionally, neither PGE₂ nor the EP₄ agonist alone affected basal Nox2 expression. These results are similar to those observed with AVF.

DISCUSSION

The results of this study show, for the first time, that administration of an EP₃ antagonist prevented a decline in cardiac function with ANG II infusion and substantially attenuated the increase in blood pressure. Furthermore, we report here that mice overexpressing EP₃ in the cardiomyocytes have reduced cardiac function at baseline, a decline in cardiac function in response to ANG II, and increased inflammatory cytokines/chemokines in their left ventricle.

Previous studies have documented the role of the EP₃ receptor in hypertension by performing studies on vascular reactivity (16). With the use of global EP₃ knockout mice, Chen et al. (16) showed in isolated mesenteric arteries that the mice lacking EP₃ display reduced systolic blood pressure at baseline and that the vasoconstrictive action of ANG II is ameliorated, suggesting an important role for EP₃ in the control of blood pressure within the resistance vessels. Our current study is novel in that it utilizes systemic administration of the pharmacological EP₃ antagonist and employs the cardiomyocyte-specific EP₃-overexpressing mice. Our data suggest that the EP₃ receptor may act in the same signaling pathway as ANG II to increase blood pressure. Although the response to L798,106 in ANG II hypertension in the kidneys and/or central nervous system was not examined in the present study, it is known that renal AT₁ receptors play a role in ANG II hypertension (19). It was also shown that EP₃ receptor activation in the rostral ventrolateral medulla mediates the pressor effects of PGE₂ (71). It was previously reported that administration of the EP₁/EP₃ agonist sulprostone invokes contraction of guinea pig aorta and induces Ca²⁺ influx (82). Other studies have confirmed the ability of EP₃ receptor activation to mobilize Ca²⁺ (2, 36, 40). The calcium dynamics of EP₃ antagonism were not examined in this study.

The Rho/Rho-kinase cascade has been shown to play an important role in regulating cardiovascular function (23, 35, 44, 68, 81, 101). For example, Guan et al. (30) have shown in rats with transverse aortic constriction pressure overload, cardiac function was improved when the rats were treated with the Rho-kinase inhibitor fasudil. Moreover, they showed this was due to a reduction in ROS. Similar cardioprotective effects were observed in a mouse model of autoimmune myocarditis when mice were treated with fasudil (20). Interestingly, in the aforementioned study, Dai et al. found a shift in the cardiac T cell population from T helper type 17 to regulatory T cell, suggesting Rho-kinase may directly affect the T cell population and IL-17 production. In our current study, we observed an increase in RhoB expression in the mesenteric arteries in response to ANG II, and this was reduced with L798,106 treatment. There was no change in RhoA expression in the mesenteric arteries. In the left ventricle, RhoB expression was completely undetectable, whereas RhoA was robustly detected. This agrees with the literature showing only the RhoA isoform is expressed in cardiomyocytes (3). Interestingly, when animals received ANG II infusion for 2 wk, there was a significant decrease in RhoA expression in the LV. It is possible that 2 wk of ANG II infusion may cause a downregulation of RhoA, as RhoA is typically shown to be fast-acting in response to ANG II; however, this is purely speculative. Furthermore, the RhoA/Rho-kinase signaling cascade in response to ANG II treatment has been studied in depth (3, 41, 51, 61, 81, 101, 102). However, little is known about the role of RhoB in ANG II HTN, therefore, making our results a novel finding. RhoB could play a role in promoting inflammation in ANG II HTN as it has been shown that RhoB, but not RhoA, is a mediator of Rho-kinase-induced NF-κB activation (74).

Isoproterenol is a nonselective β-adrenergic receptor agonist that induces positive ionotropic and chronotropic effects in the heart (11). Isoproterenol has been used to induce heart failure in rodents for many years by causing ventricular remodeling and hypertrophy (37, 69, 104). Treatment with PGE₂ has been shown to antagonize the contractile effects of isoproterenol in a Langendorff heart preparation. However, the receptor mediating this effect was not clear (48). Therefore, to determine whether EP₃ receptor blockade could ameliorate the effects of isoproterenol, we infused isoproterenol with concomitant treatment of L798,106. We observed a robust increase in heart rate, as expected with β-adrenergic stimulation, and a modest but significant increase in systolic blood pressure after 10 days of isoproterenol infusion. However, there was no effect of L798,106 on heart rate, blood pressure, or cardiac function. These data suggest that there may not be any cross talk between EP₃ receptor signaling and β-adrenergic signaling.

To study the effects of EP₃ in the heart specifically, we obtained transgenic mice that overexpress EP₃ in the cardiomyocytes. It was previously reported that the hearts of these mice are markedly hypertrophied, which is obvious as early as 5–6 wk of age, and that they have reduced ejection fraction at baseline. It was also reported that overexpression of EP₃ is protective in a mouse ischemia-reperfusion model due to the reduced contractility of the left ventricle, which would normally result in depletion of ATP (53, 54). However, we are the first to show that these mice have inflammation in their left ventricles as demonstrated by increased TNF-α and IL-1β, two major cytokines associated with heart failure (22, 89, 92). There was also reduction of anti-inflammatory cytokines IL-10 and IL-4. Interestingly, it was also observed that mRNA levels of the chemokine MCP-5 were unchanged between WT and EP₃ Tg mice. MCP-5 is a potent chemoattractant molecule for monocytes and macrophages, and we speculate that this could account for the lack of increased macrophages in the EP₃ Tg heart, as determined by immunohistochemistry. Although the number of CD68⁺ cells was unchanged, it remains to be investigated whether there is a difference in the proinflammatory M1 versus inflammation-resolving M2 macrophage phenotypes in the EP₃ Tg heart. In contrast to the macrophages, there was a significant increase in the number of infiltrating CD3⁺ T cells in EP₃ Tg mouse hearts.

Oxidative stress has been shown to modulate several aspects of pathological cardiac remodeling (24, 26, 27, 39, 47, 60, 64). A large source of reactive oxygen species comes from the family of NADPH oxidases (24). In particular, loss of the Nox2 isoform has been shown to protect the heart from adverse cardiac remodeling after myocardial infarction (52, 64). Our results show that treatment with ANG II significantly increases the expression of Nox2 in AVF. Interestingly, treatment with ANG II did not increase Nox2 expression in the cardiomyocytes and only modestly increased expression in HCECs. To our knowledge, we are the first to show that activation of the EP₄ receptor reduces ANG II-induced Nox2 expression in AVF and HCECs. Santiago et al. (79) recently reported that antagonism of the EP₄ receptor increased the contractile effect of hydrogen peroxide in coronary arteries, suggesting the vasodilatory effect of PGE₂ acting on the EP₄ receptor may protect the coronary arteries from the effects of ROS. Our study differs in that we have shown activation of EP₄ in the endothelial cells may directly reduce ROS levels via a decrease in Nox2 expression. It has also been reported in noncardiac cell types that the EP₄ receptor plays a role in the reduction of ROS. For example, it was shown in myoblast cells that treatment with an EP₄ antagonist increased ROS production, an effect that was reversed after cotreatment with antioxidants (56). In a mouse aortic dissection model, EP₄ knockout resulted in more severe vascular inflammation and increased oxidative stress, via increased Nox1 expression (99). In contrast, it was shown in the M-1 collecting duct cell line that activation of the EP₄ receptor increased Nox4 expression and subsequent ROS production (70). The mechanism of PGE₂ on ROS production is likely cell-specific depending on the specific Nox isoform that is expressed.

Conclusions

The results of this project show the importance of the EP₃ receptor in ANG II HTN and could lead to development of new drugs that target the ANG II receptors as well as EP₃ for the treatment of HTN. Furthermore, it is important to understand the deleterious effects of EP₃ in the heart during HTN as well as the detrimental effects of EP₃ on cardiac function independent of changes in blood pressure.

GRANTS

This study was funded by Henry Ford Health System internal funds (A10259) to P. Harding. T. D. Bryson was supported by the predoctoral NIH T32 Detroit Cardiovascular Training Grant 5-T32-HL-12082205.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

T.D.B. and P.H. conceived and designed research; T.D.B., T.S. Pandrangi, S.Z.K., and J.X. performed experiments; T.D.B., T.S. Pandrangi, S.Z.K., J.X., T.S. Pavlov, P.A.O., E.P., and P.H. analyzed data; T.D.B., T.S. Pandrangi, T.S. Pavlov, P.A.O., and P.H. interpreted results of experiments; T.D.B. prepared figures; T.D.B. and T.S. Pandrangi drafted manuscript; T.D.B., T.S. Pandrangi, S.Z.K., T.S. Pavlov, P.A.O., and P.H. edited and revised manuscript; T.D.B., T.S. Pandrangi, S.Z.K., J.X., T.S. Pavlov, P.A.O., E.P., and P.H. approved final version of manuscript.