Chronic elevation of endothelin-1 alone may not be sufficient to impair endothelium-dependent relaxation

Abstract

Endothelin-1 (ET-1) is a powerful vasoconstrictor peptide considered to be causally implicated in hypertension and the development of cardiovascular disease. Increased ET-1 is commonly associated with reduced nitric oxide bioavailability and impaired vascular function; however, whether chronic elevation of ET-1 directly impairs endothelium-dependent relaxation (EDR) remains elusive. Herein, we report that (i) prolonged ET-1 exposure (i.e., 48 hours) of naïve mouse aortas or cultured endothelial cells did not impair EDR or reduce endothelial nitric oxide synthase (eNOS) activity, respectively (p>0.05); (ii) mice with endothelial cell-specific ET-1 overexpression did not exhibit impaired EDR or reduced eNOS activity (p>0.05); (iii) chronic (eight weeks) pharmacological blockade of ET-1 receptors in obese/hyperlipidemic mice did not improve aortic EDR or increase eNOS activity (p>0.05); and (iv) vascular and plasma ET-1 did not inversely correlate with EDR in resistance arteries isolated from human subjects with a wide range of ET-1 levels (r=0.0037 and r=−0.1258, respectively). Furthermore, we report that prolonged ET-1 exposure downregulated vascular uncoupling protein-1 (p<0.05), which may contribute to the preservation of EDR in conditions characterized by hyperendothelinemia. Collectively, our findings demonstrate that chronic elevation of ET-1 alone may not be sufficient to impair EDR.

INTRODUCTION

Endothelin-1 (ET-1) is a potent vasoactive peptide thought to be implicated in hypertension-associated endothelial dysfunction, an important underlying factor in the development of cardiovascular disease.^1–5 Secreted primarily by endothelial cells, ET-1 modulates vascular tone through its ligation of the G protein-coupled endothelin receptor types-A and -B (ET_A and ET_B). In vascular smooth muscle cells, binding of ET-1 to ET_A and ET_B receptors results in vasoconstriction, whereas, in the endothelium, ET-1 binding to ET_B receptors leads to the release of vasodilators such as nitric oxide and prostacyclin.^{1, 6} The net effect of ET-1 signaling is, however, vasoconstriction. Indeed, studies utilizing isolated artery preparations consistently demonstrate that acute ET-1 exposure induces robust vasoconstriction in a concentration-dependent manner.^7–10 Acute ET-1 exposure has also been reported to limit endothelium-dependent relaxation (EDR) in isolated arteries¹¹ and in vivo in young healthy individuals.^{12, 13} Reciprocally, acute blockade of ET-1 receptors increases vascular conductance^14–24 and EDR^{13, 14, 16, 23, 25} in subjects at increased risk for cardiovascular disease.

These findings, combined with mounting evidence that low nitric oxide bioavailability and reduced EDR, as it occurs in obesity or diabetes, typically coexist with increased ET-1,^{14, 26–30} have led to the view that high ET-1 is a key determinant of endothelial dysfunction.^{1, 31} However, direct experimental evidence strongly supporting the notion that chronic elevation of ET-1 causes endothelial dysfunction remains equivocal with some,^{7, 32, 33} but not all,^{8–10, 34–36} reports documenting a causal link between high ET-1 and impaired EDR. Discrepancies in findings could be attributed to different vascular beds having been interrogated. An alternative view is that increased ET-1 production or action is rather a consequence of reduced nitric oxide bioavailability.³⁷ In support of this, nitric oxide synthase (NOS) inhibition increases ET-1 production in cultured endothelial cells exposed to shear stress,³⁸ as well as, in vivo, chronic NOS inhibition increases vascular expression of ET-1.³⁹ Also, elevated blood pressure caused by NOS inhibition is blunted by ET-1 receptor antagonists,^40–43 further demonstrating the prominent role of nitric oxide in regulating the production and action of ET-1.

Herein, we utilized a variety of complementary approaches, including experiments in cultured endothelial cells and in isolated arteries from mice and humans, to evaluate the hypothesis that chronic ET-1 signaling is causally coupled with reduced endothelial NOS (eNOS) activity and impaired EDR.

METHODS

Ethics and approvals

All animal study procedures received prior approval by the University of Missouri Institutional Animal Care and Use Committee. The University of Missouri is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International. All human study procedures conformed to the Declaration of Helsinki and were approved by the University of Missouri Institutional Review Board. All subjects provided written informed consent prior to participation in the study. The data to support the findings of this study can be made available by the corresponding author on reasonable request.

Experimental protocols

Experimental protocol 1: prolonged exposure to high ET-1 in culture

Aortas from eight female wild-type (wt) mice (12 weeks of age; C57BL/6NHSD; Envigo Laboratories) were excised, cleaned of perivascular adipose tissue, and incubated under standard cell culture conditions (37°C, 5% CO₂) for 48 hours in VascuLife® EnGS cell culture medium (2% FBS) containing none versus a high (i.e., supraphysiological) concentration of ET-1 (10nM), a dose used in previous studies.^44–47 Forty-eight hours was chosen to study the chronic effects of ET-1 stimulation. A set of ET-1-treated aortic rings from the same mice were co-incubated with the ET_B receptor antagonist BQ788 (1μM) to examine the role of ET_B signaling in modulating EDR. Specifically, this condition was created to limit ET_B-mediated ET-1 signaling while maintaining unrestricted signaling through the ET_A receptor. Following the 48-hour incubation, aortic rings were removed from the culture media and mounted in wire myograph organ bath chambers for the assessment of vascular function as described below. An additional set of aortic rings from eleven female wt mice (12–14 weeks of age) as well as human aortic endothelial cells (Lonza, Cat# CC-2535; Basel, Switzerland) were cultured in VascuLife® EnGS cell culture medium (2% FBS) and incubated with ET-1 (0.1–10nM) for 48 hours. The latter set of aortas and cells were prepared for Western blot analysis as described below.

Experimental protocol 2: Endothelium-restricted overexpression of ET-1 in mice

Female transgenic (tg) mice overexpressing human pre-pro ET-1 and wt littermates were generated on a C57BL/6NHSD background at the University of Missouri as previously described.⁷ Mice were sacrificed at 27 weeks of age, or as otherwise indicated, at which time vascular tissues (aorta, femoral artery, coronary artery) were excised and either used immediately for vasomotor function assessment, flash frozen in liquid nitrogen for Western blotting and vascular ET-1 content determination, or placed in formalin for immunofluorescent microscopy assessment.

Experimental protocol 3: Chronic ET-1 receptor antagonism in obese and hyperlipidemic mice

Eighteen male low-density lipoprotein receptor knockout mice (B6.129S7-Ldlrtm1Her/J on C57BL/6J background, Jackson Labs, Bar Harbor, MA) were individually housed and placed on a Western diet (Test Diet modified 58Y1, 5APC, Test Diet) at 12–13 weeks of age as a model of obesity and accelerated cardiovascular disease.⁴⁸ Concomitantly, all mice were administered ~54mg of reduced-fat peanut butter (6.7kcal/g of food, JIF, The J.M. Smucker Company, Orrville, OH) with versus without Bosentan, a dual ET-1 receptor antagonist (100mg/kg/day; Cat no. HY-A0013, MedchemExpress, Monmouth Junction, NJ) daily for eight weeks. Peanut butter, palatable to rodents, was used to keep the drug powder in suspension and ensure intake, which was monitored daily throughout the intervention. Following the eight-week intervention, aortas were harvested, and immediately used for vascular function assessment or flash frozen for Western blotting.

Experimental protocol 4: Relationship between elevated ET-1 and EDR in humans

Omental adipose tissue and plasma were obtained from 29 patients (5 males/24 females; 89.7% Caucasian; 47.8±2.2 years; BMI=45.5±1.5 kg/m²) undergoing abdominal surgery at the University of Missouri Hospital. Omental resistance arteries were isolated from the adipose tissue and immediately used for vascular function experiments, as described below, frozen for determination of vascular ET-1 content. Plasma and arteries were stored at −80°C until analysis. Patient characteristics were collected from electronic medical records.

Assessment of blood pressure

Blood pressure in mice was determined noninvasively using a CODA tail-cuff blood pressure system (CODA-HT2; Kent Scientific, Torrington, CT) within one week before euthanasia. Animals were acclimated to the restraints and tail-cuffs for four consecutive days prior to blood pressure determination. Previous evidence indicates that blood pressure in rodents is highest at the onset of the dark cycle.^{49, 50} Accordingly, blood pressure was assessed just before the start of the dark cycle (i.e., lights off) to limit the influence of the circadian rhythm. A minimum of eight blood pressure readings were averaged for each animal.

Assessment of vascular function

Mouse thoracic aortas (cranial to the diaphragm), femoral, and coronary arteries, as well as human omental adipose tissue resistance arteries, were harvested and immediately placed in cold physiological saline solution (PSS; pH ~7.4). Using a dissecting scope, all vessels were carefully cleaned of perivascular adipose tissue and surrounding connective tissue and sectioned into 1-mm (coronary arteries) or 2-mm rings (aorta, femoral artery, human resistance arteries). Mouse aortic rings, coronary arteries, and human resistance arteries were then mounted in wire myograph organ bath chambers containing warmed PSS gassed with 95% O₂ - 5% CO₂ at 37°C, as previously described.^51–53 The viability of the arteries was assessed with either 60mM (human arteries) or 80mM KCl (mouse arteries). U-46619 (20nM; mouse aorta and coronary), phenylephrine (10μM, mouse femoral) or KCl (60mM, human resistance arteries) were used to induce preconstriction for ~20min prior to assessment of vasomotor responses to acetylcholine (ACh, e-9 to e-5M), bradykinin (e-11 to e-6M) and sodium nitroprusside (SNP, e-9 to e-4M). ACh and bradykinin were used as endothelium-dependent vasodilators, while responses to SNP were considered endothelium-independent. Vasoconstriction responses to phenylephrine (e-9 to e-5M) and ET-1 (e-11 to e-7M) were also assessed on mouse aortic rings and normalized to KCl-induced constriction. Area under the curve (AUC) for each dose-response curve was calculated using the trapezoidal rule. The AUC ACh/SNP or bradykinin/SNP ratios were used as an index of endothelial function that takes into account differences in smooth muscle cell responsiveness.

Determination of plasma and vascular ET-1

ET-1 concentrations in plasma and vascular lysates were determined via a commercially available Quantikine ELISA kit per the manufacturer instructions (#DET100, R&D Systems, Minneapolis, MN). ET-1 content in whole vascular tissue lysates (mouse thoracic aorta and human omental adipose tissue resistance arteries) were normalized to total protein content as determined per commercially available BCA kit (Pierce BCA Protein Assay Kit, #23225, Thermo Scientific, Rockford, IL) per manufacturer instructions.

Formalin fixed aortic rings from wt and tg mice were paraffin embedded, sectioned (5μm), and vascular ET-1 content visualized via immunofluorescent confocal microscopy. Briefly, aortic samples were blocked with antibody buffer containing BSA and 5% horse serum. Tissues were then incubated overnight at 4°C with ET-1 primary antibody (1:200; #2786, Abcam). Next, the IgG Alexa Fluor 546 secondary antibody (1:1000; A11003; Molecular Probes, Life Technologies) was applied for one hour and then washed. Prolong Gold Antifade mountant with DAPI was applied directly to the tissue prior to coverslip placement. After curing, aortic rings were imaged for nuclei and ET-1 using a Leica SPE confocal microscope.

Determination of nitric oxide metabolites

Plasma nitrate and nitrite were measured using a commercially available fluorometric assay kit (#780051, Cayman Chemical) following the manufacturer’s instructions. Also, as per instructions, plasma was filtered in ultracentrifugation tubes (Amicon® Ultra-0.5, Millipore Sigma) for 45 minutes at 4°C prior to utilization in the assay.

Western blot

Triton X-100 aortic lysates and RIPA (R0278, Millipore Sigma, St. Louis, MO) cell lysates were prepared in 4x Laemmli buffer. Prepared protein samples (3–10 μg/lane) were separated in Criterion Tris-Glycine eXtended-PAGE precast gels (Bio-Rad). Proteins were transferred overnight onto polyvinylidene difluoride membranes and blocked with 5% bovine serum albumin. Membranes were probed for total eNOS (1:500, #32027, Cell Signaling) and phosphorylated eNOS Ser 1177 (p-eNOS; 1:500, #9570, Cell Signaling), ET_A (1:2000, #117521, Abcam), ET_B (1:1250, #117529, Abcam), uncoupling protein-1 (UCP-1; 1:1000, #u6382, Millipore Sigma), β-tubulin (1:1000, CS2146, Cell Signaling) and vinculin (1:500; #4650, Cell Signaling). The intensity of individual protein bands was quantified via densitometry using FluoroChem HD2 (AlphaView, version 3.4.0.0) and the Bio-Rad ChemiDoc XRS+ System (Bio-Rad, Hercules, CA). Proteins of interest were normalized to β-tubulin or vinculin expression. Phospho-eNOS Ser 1177 was normalized to total eNOS. Values are expressed as fold difference.

Proteomics analysis

Protein extraction

Aortas from a subset of male and female wt (n=11) and tg (n=5) mice were processed for proteomic analysis. Each sample was added to 2mL of 1X Laemmli buffer (60mM Tris-HCl, pH 6.8, 10% glycerol, 2% SDS, 100mM DTT) on ice and homogenized with Fisher Scientific 150 Homogenizer. After all samples were processed, four volumes of 100% acetone was added to each sample and incubated overnight at −80°C. The protein pellets were recovered by centrifugation for 20min at 4 × 10³ rpm at 4°C, washed twice with 80% acetone containing 10mM DTT. The pellets were dried for five min and resuspended with 40μL 6M urea, 2M thiourea, and 100mM ammonium bicarbonate. Protein concentration was determined using EZQ assay according to the manufacturer’s protocol (Invitrogen/Life Technologies). The assay paper was scanned by the FLA-5000 Fujifilm imager and quantification results were obtained using the Multi Gauge v2.3 software.

Trypsin digestion

Thirty micrograms of protein from each sample were reduced with 10mM DTT at room temperature for one hour and alkylated with 40mM iodoacetamide (IAA) for one hour at room temperature in the dark. The excess of IAA was quenched by adding 40mM DTT and the sample incubated for another one hour. Before trypsin digestion, the urea buffer was diluted to less than 1M. Trypsin (Sigma) was added to each sample at a 1:50 trypsin-to-protein mass-ratio, and the samples were incubated for 12~16 hours at 37°C. The digested peptides were purified by C18 tips according to the manufacturer’s protocol (Pierce/Thermo). Purified peptides were then lyophilized and resuspended in 5% acetonitrile and 0.1% formic acid.

Mass Spectrometry

The dissolved peptide was analyzed with a Bruker timsTOF pro instrument attached to a nanoElute LC (Bruker, MA, USA) system. The samples were loaded onto 300μm i.d. x 5mm column with C8 PepMap100, 100Å, 5μm (Thermo Fisher Scientific). The peptides were separated by an in-house packed column of 75μm i.d. x 20cm with BEH C18 130Å, 1.7μm (Waters, USA). The peptide was eluted at a flow rate of 300nl/min with the initial gradient of 2%B (A: 0.1% formic acid in water, B: 99.9% acetonitrile, 0.1% formic acid), followed by 20min ramp to 17%B, 17–25%B over 27min, 25–37%B over 11min, 37–80% B over 6min, holding at 80%B for 6min. Total running time was 70min. TimsTOF pro was operated in PASEF mode. Duty cycle was locked to 100%. Ion mobility coefficient (1/K0) value was set from 0.6 – 1.6Vs cm⁻². MS data was collected over m/z range of 100 to 1700. During MS/MS data collection, each TIMS cycle of 1.27s contained one MS and ten PASEF MS/MS scan. Exclusion was active after 0.4min.

Data Analysis

Raw data was searched using PEAKS (version 8.5, Bioinformatics Solutions Inc. Canada) with Uniprot mouse database. Data were searched with the following parameters: trypsin protease digestion with two missed cleavage allowed, precursor ion tolerance of 50ppm, and fragment ion tolerance of 0.1Da. Cysteine carbamidomethylation was set as a fixed modification, and methionine oxidation and asparagine/glutamine deamidation were set as variable modifications. Label free quantification was provided by PEAKS if a protein was identified in at least two samples in each group and PSM FDR was less than 1%.

Statistical analysis

GraphPad Prism (version 8.0, GraphPad Prism Software, La Jolla, CA) was used for statistical analysis. Statistical comparisons were performed using univariate or multivariate analysis of variance (ANOVA), as appropriate, followed by Fisher-LSD post hoc test when applicable. T-tests were utilized as appropriate for all comparisons between two independent treatment groups or conditions. Values are expressed as means ± standard error of mean (SEM). Significance was determined when p<0.05.

RESULTS

Prolonged exposure to high ET-1 did not impair EDR or reduce eNOS activity

Exposure of naïve aortic rings to ET-1 (10nM) for 48 hours reduced subsequent ET-1-induced vasoconstriction which was not prevented with the ET_B receptor antagonist BQ788 (Figure 1A). However, prolonged ET-1 exposure did not elicit impairments in ACh- or SNP-induced relaxation in isolated mouse aortas (Figure 1A, middle and right panels). Furthermore, co-incubation with the ET_B receptor blocker did not unveil an unfavorable effect of ET-1 on EDR, suggesting that augmented ET_B signaling does not contribute to the conservancy of EDR with increased ET-1. Additionally, eNOS activation, as assessed by p-eNOS Ser1177/total eNOS, was not altered in cultured human aortic endothelial cells treated with ET-1 (0.1–10nM) for 48 hours (Figure 1B).

Figure 1. Prolonged exposure to high ET-1 did not impair EDR or reduce eNOS activity.

A) Naïve aortic rings from female C57BL/6NHSD mice were cultured in VascuLife® cell culture media containing endothelin-1 (ET-1; 0nM-10nM) ± BQ788 (1μM, ET_B receptor antagonist) for 48 hours. Aortic rings were mounted on steel pins in organ bath chambers and exposed to increasing concentrations of endothelin-1 (ET-1), acetylcholine (ACh), and sodium nitroprusside (SNP); n=8/condition. U-46619 (20nM) was used to preconstrict the arteries prior to the assessment of relaxation. B) Human aortic endothelial cells were treated with ET-1 (0nM–10nM) for 48 hours under standard cell culture conditions. eNOS activation (p-eNOS Ser1177/total eNOS) was determined via Western blot; n=5–6/condition. All data are expressed as means ± SEM. *p<0.05 vs Con or 0nM ET-1. Con, control; AUC, area under the curve.

Endothelial cell-restricted ET-1 overexpression did not increase blood pressure, impair EDR, or reduce eNOS activity

A representative PCR image for wt and tg genotypes is displayed in Figure 2A. At 27 weeks of age, tg mice exhibited significant increases in vascular (aortic, ~2.5-fold) and plasma ET-1 (~10-fold) concentrations relative to wt mice (Figure 2A). Visualization of aortic ET-1 content of tg and wt mice is depicted in Figure 2B. Despite the elevation in ET-1 in tg mice, mean arterial blood pressure was unaltered (Figure 2C). Tg mice exhibited suppressed aortic vasoreactivity to phenylephrine and ET-1 (p=0.07) relative to wt mice (Figure 2D). In isolated aortic rings, ET-1 overexpression enhanced EDR (i.e., ACh, ACh/SNP) by ~2–3 fold compared with wt counterparts (Figure 2E), whereas SNP-induced relaxation was not different between genotypes. In isolated femoral and coronary arteries, ET-1 overexpression did not significantly alter relaxation responses to ACh or SNP (Figure 2E). Overexpression of ET-1 did not alter aortic eNOS content or Ser1177 phosphorylation (p-eNOS/eNOS), nor did it influence expression of ET_A and ET_B receptors relative to wt mice (Figure 2F). In addition, plasma nitrate/nitrite concentrations were not different between wt and tg mice (30.35±5.28 μM and 41.35±7.55 μM, respectively; p=0.230).

Figure 2. Endothelial cell-restricted ET-1 overexpression did not increase blood pressure, impair EDR, or reduce eNOS activity.

A) Representative image of PCR genotyping gel of wild-type (wt) and endothelin-1 (ET-1) overexpressing transgenic (tg) mice. Aortic (e.g., vascular) and plasma endothelin-1 (ET-1) concentrations as determined by ELISA; n=18–21/group. B) Visualization of aortic ET-1 content via immunofluorescent confocal microscopy in wt and tg mice. L denotes lumen; (−) denotes negative control. C) Mean arterial blood pressure (MAP) as determined by tail-cuff blood pressure monitoring; n=18–21/group. D) Aortic vasoconstriction responses to phenylephrine (PE; n=16–19/group) and ET-1 (n=5–11/group). Aortic rings were pre-treated with L-NAME (300μM for 30 min) to limit confounding effects of nitric oxide production on vascular tone and isolate the effects of ET-1 on vasoconstriction. E) Endothelium-dependent (ACh, ACH/SNP) and –independent (SNP) vasomotor function responses in the aorta, femoral artery, and coronary artery in wt and tg mice; n=12–19/genotype/artery. Vasomotor function responses were determined via wire myography (aorta and coronary artery) or pressure myography (femoral artery). F) Western blot determination of aortic eNOS activation (p-eNOS Ser1177/total eNOS) and endothelin receptor A (ET_A) and receptor B (ET_B) in aortic homogenates from wt and tg mice. Representative Western blot images are depicted. Proteins of interest were normalized to the housekeeping protein, vinculin; n=18–19/group. All data are expressed as means ± SEM. *p<0.05 vs wt. AUC, area under the curve.

Inhibition of ET-1 receptors did not improve EDR or enhance eNOS activity in obese and hyperlipidemic mice

Eight weeks of oral Bosentan treatment in obese and hyperlipidemic mice (i.e., LDLrKO mice fed a Western diet) reduced mean arterial pressure by ~6% relative to untreated mice (Figure 3B). This effect occurred despite no significant differences in body weight (Bosentan-treated = 39.6±0.5 g, untreated = 41.2±0.8 g; p>0.05). Aortic ACh- and SNP-induced relaxation was unchanged by chronic ET-1 inhibition (Figure 3C). Aortic eNOS content or Ser1177 phosphorylation (p-eNOS/eNOS) was not different between control and Bosentan-treated mice (Figure 3D). In addition, plasma nitrate/nitrite concentrations were not different between control and Bosentan-treated mice (32.06±5.14 μM and 27.26±5.13 μM, respectively; p=0.527).

Figure 3. Inhibition of ET-1 receptors did not improve EDR or enhance eNOS activity in obese and hyperlipidemic mice.

A) Graphical depiction of experimental design. Western diet-fed mice were administered either the dual ET-1 receptor antagonist, Bosentan (Bos; 100mg/kg; chronic ET-1 receptor blockade) or vehicle control (Con; reduced fat peanut butter) daily for eight weeks. B) Mean arterial pressure (MAP) determined via tail-cuff monitoring; n=9/group. C) Endothelium-dependent (ACh, ACH/SNP) and –independent (SNP) vasomotor function responses in the aorta; n=9/group. D) Western blot determination of aortic eNOS activation (p-eNOS Ser1177/total eNOS) in aortic homogenates; n=9/group. Representative Western blot images are depicted. All data are expressed as means ± SEM. *p<0.05 vs Con. AUC, area under the curve.

Elevated ET-1 was not associated with impairment in EDR in isolated resistance arteries from overweight/obese humans

Vascular ET-1 content of human resistance arteries (depicted from lowest ET-1 to highest ET-1 content) and corresponding plasma ET-1 and EDR responses are individually displayed for all 29 overweight/obese subjects (Figure 4B). No significant correlations were found between vascular or plasma ET-1 and EDR responses (i.e., bradykinin/SNP; Figure 4C–D). In addition, plasma ET-1 concentrations were not correlated with plasma nitrate/nitrite concentrations (r=0.145, p=0.488).

Figure 4. Elevated ET-1 was not associated with impairment in EDR in isolated resistance arteries from overweight/obese humans.

A) Diagram of experimental design and outcomes. Omental adipose tissue resistance arteries were isolated from 29 patients undergoing abdominal surgery. B) Individual vascular and plasma ET-1 concentrations and vasomotor function responses (bradykinin and bradykinin/SNP ratios) for 29 human subjects. C) Correlation of endothelium-dependent relaxation (bradykinin/SNP AUC) with vascular ET-1 content and D) plasma ET-1 concentrations; n=27–29. All data are expressed as individual data points. AUC, area under the curve; NA, data not available.

Chronic ET-1 exposure was associated with reduced vascular and endothelial uncoupling protein-1 content

Proteomic analysis of wt and tg mouse aortas revealed 2,386 detected proteins. Fifty-eight proteins were differentially expressed with 47 meeting our criteria for significance (fold change >1.5; Table 1). UCP-1 was the most significantly downregulated protein (4-fold in tg mice compared with wt; Figure 5A). To corroborate the proteomic analysis, human aortic endothelial cells and wt mouse aortas were exposed to high ET-1 (10nM) for 48 hours and probed for UCP-1. ET-1 treatment downregulated UCP-1 in cultured endothelial cells (Figure 5B) and in isolated mouse aortas (Figure 5C).

Table 1. Proteomic analysis of mouse aortas.

Aortas from male and female ET-1 overexpressing transgenic and wild-type mice were processed for proteomic analysis (11 wild-type, 5 transgenic). Forty-seven proteins were identified as differentially expressed as determined by a fold-change >1.5. Uniprot accession numbers, fold change, and p-values are listed. Fold change is relative to wild-type control (set to 1.0).

Accession #	Description	Fold	p-value
P04202|TGFB1_MOUSE	Transforming growth factor beta-1	1.84	0.0042
Q61160|FADD_MOUSE	FAS-associated death domain protein	1.60	0.0003
Q08879|FBLN1_MOUSE	Fibulin-1	1.59	<0.0001
Q9Z1Z0|USO1_MOUSE	General vesicular transport factor p115	0.65	0.0071
Q9ER38|TOR3A_MOUSE	Torsin-3A	0.63	0.0031
Q8K4L3|SVIL_MOUSE	Supervillin	0.62	0.0018
Q60932|VDAC1_MOUSE	Voltage-dependent anion-selective channel protein 1	0.62	0.0048
Q62348|TSN_MOUSE	Translin	0.62	0.0064
P11688|ITA5_MOUSE	Integrin alpha-5	0.60	0.0029
Q9JLN9|MTOR_MOUSE	Serine/threonine-protein kinase mTOR	0.59	0.0086
Q61001|LAMA5_MOUSE	Laminin subunit alpha-5	0.58	0.00423
Q923B6|STEA4_MOUSE	Metalloreductase STEAP4	0.57	0.0007
P59235|NUP43_MOUSE	Nucleoporin Nup43	0.57	0.0023
Q9QYA2|TOM40_MOUSE	Mitochondrial import receptor subunit TOM40 homolog	0.56	0.0020
Q6PDI5|ECM29_MOUSE	Proteasome adapter and scaffold protein ECM29	0.55	0.0076
Q9WVC3|CAV2_MOUSE	Caveolin-2	0.54	0.0006
Q9QUI0|RHOA_MOUSE	Transforming protein RhoA	0.54	0.0064
Q8BWM0|PGES2_MOUSE	Prostaglandin E synthase 2	0.54	0.0080
Q8VDD5|MYH9_MOUSE	Myosin-9	0.52	0.0002
O54890|ITB3_MOUSE	Integrin beta-3	0.52	0.0015
Q8BK72|RT27_MOUSE	28S ribosomal protein S27 mitochondrial	0.51	0.0046
Q99JI6|RAP1B_MOUSE	Ras-related protein Rap-1b	0.50	0.0096
O09118|NET1_MOUSE	Netrin-1	0.49	0.0024
P68372|TBB4B_MOUSE	Tubulin beta-4B chain	0.48	0.0020
Q8R080|GTSE1_MOUSE	G2 and S phase-expressed protein 1	0.47	0.0057
Q3U9N9|MOT10_MOUSE	Monocarboxylate transporter 10	0.46	0.0059
Q9D0I9|SYRC_MOUSE	Arginine--tRNA ligase cytoplasmic	0.45	0.0088
Q3UV70|PDP1_MOUSE	[Pyruvate dehydrogenase [acetyl-transferring]]-phosphatase 1 mitochondrial	0.43	0.0010
Q9QZI9|SERC3_MOUSE	Serine incorporator 3	0.43	0.0036
P47809|MP2K4_MOUSE	Dual specificity mitogen-activated protein kinase kinase 4	0.43	0.0052
Q64331|MYO6_MOUSE	Unconventional myosin-VI	0.40	0.0007
Q9R0H0|ACOX1_MOUSE	Peroxisomal acyl-coenzyme A oxidase 1	0.39	0.0027
Q00898|A1AT5_MOUSE	Alpha-1-antitrypsin 1–5	0.39	0.0049
Q9CXD6|MCUR1_MOUSE	Mitochondrial calcium uniporter regulator 1	0.38	0.0021
Q9DCF9|SSRG_MOUSE	Translocon-associated protein subunit gamma	0.38	0.0082
Q8BSS9|LIPA2_MOUSE	Liprin-alpha-2	0.37	0.0054
Q9JJU8|SH3L1_MOUSE	SH3 domain-binding glutamic acid-rich-like protein	0.31	<0.0001
Q924L1|LTMD1_MOUSE	LETM1 domain-containing protein 1	0.30	0.0073
Q9WTQ8|TIM23_MOUSE	Mitochondrial import inner membrane translocase subunit Tim23	0.28	0.0079
Q5U405|TMPSD_MOUSE	Transmembrane protease serine 13	0.26	0.0040
P12242|UCP1_MOUSE	Mitochondrial brown fat uncoupling protein 1	0.25	0.0040

Figure 5. Chronic ET-1 exposure was associated with reduced vascular and endothelial uncoupling protein-1 (UCP-1) content.

A) Proteomic analysis of aortas isolated from wild-type (wt, n=11) and ET-1 overexpressing transgenic (tg, n=5) mice (~ 5 months of age). Forty-seven proteins (3 upregulated, 44 downregulated) were identified as differentially expressed between the tg and wt groups. UCP-1 was identified as the most significantly downregulated protein in tg mice relative to wt controls. B) UCP-1 protein expression in cultured human aortic endothelial cell (HAEC, n=6/treatment) and C) naïve cultured mouse aortas (n=10/treatment) treated without (0nM) versus with ET-1 (10nM) for 48 hours. All data are expressed as means ± SEM. *p<0.05 vs. 0nM ET-1.

DISCUSSION

Increased ET-1 has been regarded as a key characteristic feature and underlying mechanism of endothelial dysfunction in various chronic diseases including obesity and diabetes. However, in contrast to some existing literature,^{7, 32, 33} the present investigation revealed that elevated ET-1 alone may not be sufficient to impair EDR. Furthermore, we report that prolonged ET-1 exposure downregulated vascular UCP-1, which may represent a putative contributor to the preservation of EDR in the setting of hyperendothelinemia.

Our conclusion that elevated ET-1 alone may not be sufficient to impair EDR is supported by several lines of evidence, as follows. First, we demonstrated that sustained exposure (i.e., 48 hours) of naïve mouse aortas to exogenous ET-1 (10nM) did not impair EDR, nor did ET-1 exposure (0.1–10nM) reduce eNOS content and activity in cultured human aortic endothelial cells. This finding is in contrast to a previous study demonstrating a reduction in total eNOS protein in human saphenous vein endothelial cells exposed to 100nM of ET-1 for 24 hours.⁴⁷ While the reason for these disparate findings is unclear, the supraphysiological concentration of ET-1 or the different cell type employed in that study may have contributed. Given that ET_B signaling in endothelial cells is known to lead to nitric oxide production,⁶ we reasoned that persistent endothelial ET_B signaling with ET-1 exposure could be a contributor to the preservation of EDR. To address this question, we treated mouse aortic rings with ET-1 in the presence or absence of an ET_B blocker (i.e., BQ788). In contrast to our proposition, we found that prolonged ET_B blockade did not unmask a detrimental effect of ET-1 on EDR, suggesting that preservation of EDR with increased ET-1 was independent of augmented ET_B signaling. Second, we demonstrated that mice with endothelial cell-restricted ET-1 overexpression did not exhibit impaired EDR across multiple vascular beds including aorta, femoral artery, and coronary artery or reduced eNOS activity. In fact, we report that ET-1-overexpressing mice displayed augmented EDR in the aorta, a finding that is in agreement with previous reports using other models.^{8, 9, 54} Third, we demonstrated that chronic (eight weeks) pharmacological antagonism of ET-1 receptors in a mouse model of obesity and hyperlipidemia did not improve aortic EDR or enhance eNOS activity. Lastly, in a cohort of human subjects with a wide range of ET-1 levels, we demonstrated that vascular and plasma ET-1 did not negatively correlate with EDR in isolated resistance arteries.

In this last protocol, we chose to examine EDR in small arteries isolated from omental adipose tissue instead of examining EDR in vivo via other well-established techniques with documented prognostic value (e.g., brachial artery flow-mediated dilation^55–59). The decision to use this ex vivo approach was based on the fact that a principal objective was to assess vascular content of ET-1 and relate it to EDR in the same artery. In this regard, future studies should examine if the same lack of relationship would exist when correlating brachial artery flow-mediated dilation versus ET-1 content in brachial artery endothelial cells harvested via catheterization.

As reported in the original publication describing this transgenic mouse model,⁷ here we show that congenitally ET-1-overexpressing mice do not exhibit increased blood pressure. This is in contrast to an inducible model of endothelium-restricted ET-1 overexpression in which blood pressure was shown to rise persistently for three weeks³⁶ to three months after induction.³³ Notably, in the present investigation we found that isolated arteries from ET-1-overexpressing mice were hyporeactive to ET-1 and phenylephrine, an α1-adrenergic receptor agonist. Consistent with this finding, previous work also showed that ET-1 signaling limits α1-adrenergic receptor-mediated vasoconstriction,⁶⁰ thus contributing to the regulation of blood pressure.⁶¹ Also in line with previous reports,^8–10 the observed hyporeactivity to ET-1 (similarly detected in naïve arteries that were treated with ET-1 ex vivo for 48 hours) may result from an adaptive response involving the desensitization of the endothelin receptors by G protein-coupled receptor kinases,^{62, 63} which are implicated in blood pressure regulation.⁶⁴ Therefore, it is conceivable that maintenance of normal blood pressure in the setting of hyperendothelinemia can be attributed to both, hyporeactivity to vasoconstrictors and preservation of EDR.

It should be acknowledged that the lack of impairment in EDR in this model of congenital ET-1 overexpression is in divergence with findings reported in the first publication describing these mice,⁷ more than 15 years ago, where endothelial dysfunction was detected. That is, as also found by the original investigators in subsequent studies,³⁵ it appears that a phenotype drift has occurred over the years in this mouse model. This serendipitous observation opens a unique opportunity to uncover novel therapeutic targets for protection against endothelial dysfunction and cardiovascular disease. Indeed, to begin to identify possible compensatory and adaptive molecular mechanisms responsible for the preservation of EDR in this model of high ET-1, we conducted a proteomics analysis of aortas from ET-1-overexpressing vs. wt mice. This experiment revealed a significant downregulation of UCP-1 in the aorta of mice with ET-1 overexpression. UCP-1, which is highly expressed in brown adipose tissue but also detectable in vascular cells,^{52, 65, 66} is an inner mitochondrial membrane-bound protein that regulates the proton pump. Direct effects of ET-1 on downregulating UCP-1 were confirmed in cultured endothelial cells and in naïve aortic rings stimulated with ET-1 for 48 hours. This finding, the first in vascular cells, is congruent with previous work by others demonstrating that ET-1 downregulates UCP-1 in cultured adipocytes.⁶⁷ Importantly, contemporary work demonstrates a link between UCP-1 and vascular function. Namely, we recently found that mice with systemic ablation of UCP-1 exhibit increased aortic EDR,⁶⁸ while upregulation of aortic UCP-1 is associated with aortic stiffening in a mouse model of brown adipose tissue lipectomy.⁶⁵ In addition, work by Bernal-Mizrachi et al.⁶⁹ demonstrates that mice with vascular smooth muscle cell-specific UCP-1 overexpression display an increase in blood pressure and are more susceptible to atherosclerotic lesions. Thus, given the role of UCP-1 in modulating vascular function, it is possible that ET-1-induced downregulation of UCP-1 can contribute to maintaining EDR in a setting of high ET-1 exposure, possibly by alleviating mitochondrial oxidative stress.⁶⁹ Nevertheless, further research is needed to mechanistically determine the link between vascular UCP-1 function and EDR.

Perspectives

Increased ET-1 is commonly associated with reduced nitric oxide bioavailability and impaired EDR, an early defect in the development of cardiovascular disease. However, using a variety of complementary approaches comprising experiments in cultured endothelial cells and in isolated arteries from mice and humans, the present investigation demonstrates that chronic ET-1 signaling may not always causally drive the reduction in eNOS activity and EDR across all vascular beds. This is likely the result of compensatory and adaptive protective mechanisms including the downregulation of vascular UCP-1. Identifying such counter-regulatory mechanisms can lead to novel therapeutics for the prevention and treatment of cardiovascular disease.

Novelty and Significance

What is New?

• Elevated ET-1 signaling did not impair eNOS activation or reduce endothelium-dependent relaxation (EDR) in isolated arteries from mice.

• Pharmacological blockade of ET-1 receptors in obese/hyperlipidemic mice did not improve aortic EDR or increase eNOS activity.

• Plasma or vascular ET-1 did not inversely correlate with EDR in isolated adipose tissue resistance arteries from humans.

• Sustained elevation of ET-1 likely results in compensatory adaptive mechanisms, such as the downregulation of vascular UCP-1, which may contribute to the preservation of EDR and blood pressure in conditions characterized by hyperendothelinemia.

What is Relevant?

• Increased ET-1 has been regarded as a key characteristic feature and underlying mechanism of endothelial dysfunction in various chronic diseases including obesity and diabetes.

• Here we demonstrate that chronic elevation of ET-1 alone may not be sufficient to impair EDR.

Summary

• Chronic ET-1 signaling may not always causally drive a reduction in eNOS activity and EDR likely as a result of counter-regulatory protective mechanisms including the downregulation of vascular UCP-1.

Abbreviations

ET-1

endothelin-1

EDR

endothelium-dependent relaxation

ET_A

endothelin receptor type-A

ET_B

endothelin receptor type-B

eNOS

endothelial nitric oxide synthase

UCP-1

uncoupling protein-1