Relaxin reduces endothelium‐derived vasoconstriction in hypertension: Revealing new therapeutic insights

Abstract

Background and Purpose

Endothelium‐derived vasoconstriction is a hallmark of vascular dysfunction in hypertension. In some cases, an overproduction of endothelium‐derived prostacyclin (PGI₂) can cause contraction rather than relaxation. Relaxin is well known for its vasoprotective actions, but the possibility that this peptide could also reverse endothelium‐derived vasoconstriction has never been investigated. We tested the hypothesis that short‐term relaxin treatment mitigates endothelium‐derived vasoconstriction in spontaneously hypertensive rats (SHR).

Experimental Approach

Male Wistar Kyoto rats (WKY) and SHR were subcutaneously infused with either vehicle (20 mmol·L⁻¹ sodium acetate) or relaxin (13.3 μg·kg⁻¹·hr⁻¹) using osmotic minipumps for 3 days. Vascular reactivity to the endothelium‐dependent agonist ACh was assessed in vitro by wire myography. Quantitative PCR and LC‐MS were used to identify changes in gene expression of prostanoid pathways and PG production, respectively.

Key Results

Relaxin treatment ameliorated hypertension‐induced endothelial dysfunction by increasing NO‐dependent relaxation and reducing endothelium‐dependent contraction. Notably, short‐term relaxin treatment up‐regulated mesenteric PGI₂ receptor (IP) expression, permitting PGI₂–IP‐mediated vasorelaxation. In the aorta, reversal of contraction was accompanied by suppression of the hypertension‐induced increase in prostanoid‐producing enzymes and reduction in PGI₂‐evoked contractions.

Conclusions and Implications

Relaxin has region‐dependent vasoprotective actions in hypertension. Specifically, relaxin has distinct effects on endothelium‐derived contracting factors and their associated vasoconstrictor pathways in mesenteric arteries and the aorta. Taken together, these observations reveal the potential of relaxin as a new therapeutic agent for vascular disorders that are associated with endothelium‐derived vasoconstriction including hypertension.

What is already known

• The peptide hormone relaxin elicits vasoprotective actions in the macrovasculature and microvasculature.

What does this study add

• Short‐term relaxin treatment attenuates endothelium‐derived vasoconstriction in arteries isolated from hypertensive rats.

What is the clinical significance

• Short‐term relaxin infusion reverses hypertension‐induced endothelial dysfunction by limiting endothelium‐derived vasoconstriction in a region‐dependent manner.

• Relaxin is a potential therapeutic drug for vascular diseases associated with exaggerated endothelium‐derived vasoconstriction.

Abbreviations

EDC

endothelium‐dependent contraction

EDCFs

endothelium‐derived contracting factors

EDH

endothelium‐derived hyperpolarisation

IK_Ca

intermediate‐conductance calcium‐activated potassium channel

Indo

indomethacin

KPSS

high potassium physiological saline solution

l‐NAME

N ^ω‐nitro‐l‐arginine methyl ester

SK_Ca

small‐conductance calcium‐activated potassium channel

TRAM34

1‐[(2‐chlorophenyl)diphenylmethyl]‐1H‐pyrazole

R_max

maximum relaxation

Rn18S

ribosomal 18S

1. INTRODUCTION

Essential hypertension is associated with vascular endothelial dysfunction (Bernardo, Weeks, Pretorius, & McMullen, 2010; Vanhoutte, Zhao, Xu, & Leung, 2016). Endothelium‐derived https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2509, https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=1915 ₂), and endothelium‐derived hyperpolarisation (EDH) are major contributors to vasorelaxation in healthy blood vessels. In hypertension, the production and vasodilator capacity of NO and EDH are diminished and contribute to endothelial dysfunction (Leung & Vanhoutte, 2015; Tang & Vanhoutte, 2010; Vanhoutte et al., 2016). In addition, the endothelium favours the production of endothelium‐derived contracting factors (EDCFs) that enhance the contraction of blood vessels in hypertensive subjects. Specifically, in the aorta of spontaneously hypertensive rats (SHR), the production of https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=269‐derived PGI₂ is paradoxically increased, with an associated reduction in the expression of the https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=345 (Félétou, Verbeuren, & Vanhoutte, 2009; Gluais, Lonchampt, Morrow, Vanhoutte, & Feletou, 2005). As a consequence, the excessive endothelium‐derived PGI₂ acts on the thromboxane TP receptors and/or other prostanoid receptors to cause contraction rather than vasorelaxation, indicating that PGI₂ shifts from being a vasodilator to an EDCF (Liu et al., 2017). Targeting vascular prostanoid pathways may reveal new effective treatments for endothelial dysfunction in hypertension.

It is well established that the peptide hormone https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=3745 has vasoprotective actions in the cardiovascular system (Jelinic et al., 2018; Jelinic, Marshall, Leo, Parry, & Tare, 2019; Leo et al., 2017; Leo, Jelinic, Ng, Tare, & Parry, 2016a). These actions occur through activation of its major receptor, https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=351, which is localised to endothelial and vascular smooth muscle cells of blood vessels (Jelinic et al., 2014; Ng, Jelinic, Parry, & Leo, 2015). Short‐term relaxin infusion (2–3 days) enhances endothelial vasodilator function by up‐regulating https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1249), https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1376, and IP expression (Jelinic et al., 2017), thereby increasing NO‐mediated relaxation and PGI₂ production in healthy mesenteric arteries (Leo, Jelinic, Ng, Tare, & Parry, 2016b). In the aorta, relaxin treatment also reverses endothelial dysfunction by stimulating NO and/or PGI₂ production in the aorta (Dschietzig et al., 2012; Ng, Leo, & Parry, 2016; Pini et al., 2016). Previous studies have reported inconsistent findings regarding the potential of relaxin to reverse vascular dysfunction in SHR. For example, relaxin treatment has minimal effects on myogenic tone and flow‐mediated vasodilation in mesenteric arteries of SHR (van Drongelen et al., 2013). However, in SHR cerebral parenchymal arterioles, relaxin treatment decreases myogenic tone and increases EDH‐type relaxation (Chan, Sweet, & Cipolla, 2013). In SHR aorta, relaxin infusion reverses adverse vascular remodelling and increases arterial compliance (Xu, Chakravorty, Bathgate, Dart, & Du, 2010). However, none of these studies investigated the potential for relaxin to mitigate the detrimental effects of EDCFs in blood vessels from hypertensive individuals. Given the importance of EDCFs in the pathogenesis of vascular dysfunction in hypertension, it is essential to resolve whether or not relaxin mitigates EDCF‐mediated contraction in the setting of hypertension.

Therefore, the key objective of the present study was to investigate the effects of short‐term relaxin treatment on vascular function, focusing on EDCF pathways, in both small mesenteric arteries and aorta of SHR. We also examined whether or not short‐term relaxin treatment elicits benefits within the heart, specifically indications of left ventricle (LV) remodelling as reported with longer durations (14 days) of relaxin infusion (Lekgabe et al., 2005) in SHR. Here, our results showed that short‐term relaxin treatment mitigates endothelium‐dependent contraction (EDC) in both the mesenteric artery and aorta of SHR. Although the overall therapeutic endpoint is a reduction in EDC, relaxin has region‐specific actions on the vasculature through novel influences on EDCF generation/bioavailability and their associated vasoconstrictor pathways. To our knowledge, these findings are the first demonstration of an interaction between relaxin and EDCFs in the context of hypertension‐induced vascular dysfunction in vivo. They also provide new insights for development of relaxin‐based peptides to treat vascular disorders involving endothelium‐derived vasoconstriction.

2. METHODS

2.1. Animals

All animal care and experimental procedures were approved by the Faculty of Science, The University of Melbourne Animal Experimentation Ethics Committee (The University of Melbourne, AEC 1413186.1) and conform to BJP's Ethical Policies and the National Health and Medical Research Council of Australia code of practice for the care and use of animals for scientific purposes. Animal studies are reported in compliance with the ARRIVE guidelines (Kilkenny, Browne, Cuthill, Emerson, & Altman, 2010) and with the recommendations made by the British Journal of Pharmacology. All rats were housed in individually ventilated large cages with paper bedding (two rats per cage) at the School of BioSciences SPF1 Animal House Facility. They were maintained on an automated time cycle of 12‐hr light/dark at 20°C, with standard food pellets (Barastoc, VIC, Australia) and water available ad libitum.

2.2. Model and experimental design

The SHR is a well‐established animal model of essential (or primary) hypertension, which is one of the most studied animal models for vascular disease in the context of hypertension (Chan et al., 2013; Debrah, Conrad, Jeyabalan, Danielson, & Shroff, 2005; Parikh et al., 2013; Xu et al., 2010). In this study, male SHR (body weight: 330–410 g) and normotensive Wistar Kyoto rats (WKY; body weight: 350–420 g) aged 24–26 weeks old (Animal Resource Centre, WA, Australia) were housed and maintained under standard conditions, as described above. Equal group size of n = 15 rats was designed for each group of this study. It is estimated that approximately 10% of SHR may not develop hypertension (systolic BP > 150 mmHg); thus, two extra SHR were included in each group (n = 17 per group for SHR). Systolic BP of rats was measured using non‐invasive tail cuff plethysmography (Kent Scientific Corporation, CT, USA), as previously described (O'Sullivan et al., 2016). The systolic BP of SHR (mean: 179 ± 6 mmHg, n = 34) was significantly higher than the WKY (mean: 121 ± 7 mmHg, n = 15). In this study, all the SHR (n = 34) developed hypertension and were included. BP was not measured during relaxin infusion in this study as it has previously been shown that relaxin treatment has no effect on BP in SHR (Chan et al., 2013; Debrah, Conrad, Jeyabalan, Danielson, & Shroff, 2005; Parikh et al., 2013; Xu et al., 2010) and also in another SHR study from our laboratory (unpublished data). The SHR were randomly divided into two groups to receive either vehicle (20 mmol·L⁻¹ sodium acetate, n = 17) or recombinant human relaxin‐2 (13.3 μg·kg⁻¹·hr⁻¹, n = 17) for a period of 3 days. The WKY received an equivalent volume of the vehicle alone for 3 days (n = 15). Considering the principles of animal ethics (replacement, reduction, and refinement), we did not add two more WKY to this study to have equal group sizes (n = 17 per group) as these additional animals would not have changed the outcomes of the statistical analysis. Similarly, given that we and many others have investigated the vascular effects of relaxin in normotensive rats (e.g., WKY + placebo vs. WKY + relaxin; Jelinic et al., 2014; Jelinic et al., 2017; Leo, Jelinic, Parkington, Tare, & Parry, 2014; van Drongelen et al., 2013), we concluded that repeating this experiment with an extra cohort of relaxin‐treated WKY would not provide new knowledge of relevance to this study.

2.3. Subcutaneous infusion of relaxin

On the day of surgery, all surgical instruments were sterilised using a glass bead steriliser (Germinator 500, Medtex, Mt Waverly, VIC, Australia), and gauze was sterilised by autoclaving. Aseptic conditions were maintained throughout surgery. For subcutaneous infusion of relaxin or vehicle, all animals were anaesthetised with 2% isoflurane (Univentor 400, Agnthos, AB, Sweden) in oxygen via inhalation. Rats were determined to be within the surgical plane of anaesthesia by the absence of pedal withdrawal, toe pinch, and corneal reflexes (both left and right). Respiration rate and colour of the mucous membranes were also monitored throughout the surgery to prevent isoflurane overdose. Rats were placed on a heating pad set to 37°C during surgery to prevent hypothermia. Ocular lubricant was applied to the eyes to prevent irritation, and the surgical site was shaved and swabbed with Microshield 5 disinfectant (Clorohexidine gluconate 5% w/v; Johnson & Johnson Medical Pty Ltd., North Ryde, NSW, Australia). Once anaesthetised, a 2‐cm incision was made between the shoulder blades, and the subcutaneous tissue was separated from the cutaneous tissue via blunt dissection to make a pocket large enough for the minipumps. The rats were implanted with a 3‐day Alzet osmotic minipump (Model 1003D, Bioscientific, Gymea, NSW, Australia) under the skin to infuse either relaxin or vehicle. Bupivacaine (2–4 drops of 0.5%, Pfizer Australia Pty Ltd., West Ryde, NSW, Australia) was applied to the incision site prior to closure for analgesia with three to four simple interrupted stitches tied with a surgeon's knot using sterile 4‐0 polyglactin‐90‐coated vicryl sutures with a 19 mm 3/8 circular reverse cutting needle (Ethicon, Johnson & Johnson Medical Pty Ltd., North Ryde, NSW, Australia). After surgery, the rats were transferred to a heated clean cage to monitor for signs of pain and distress. Once the rats regained consciousness, they were transferred back to the animal facility where regular welfare‐related assessments (monitoring for any signs of pain, infection, or distress) were carried out daily. All rats recovered fully after surgery, without any indication of pain or other ailments that required intervention.

Three days post‐onset of relaxin infusion, blood samples were obtained from the LV via cardiac puncture under 2% isoflurane anaesthesia. Plasma concentrations of relaxin were measured in duplicate using the Human Relaxin‐2 Quantikine elisa Kit (R&D Systems, Minneapolis, MN, USA) following the manufacturer's protocol. Human recombinant relaxin was detected in the plasma of relaxin‐treated SHR after 3 days of infusion (mean: 101 ± 9 ng·ml⁻¹) but was below the level of detection in the plasma of placebo‐treated WKY and SHR. The limit of detection of detection was 15.6 pg·ml⁻¹, and the intracoefficient and intercoefficient of variation were 2.3–4.7% and 5.5–10.2%, respectively.

After blood collection, the animals were killed via cervical dislocation under anaesthesia. The mesenteric arcade was isolated and immediately placed in ice‐cold Krebs bicarbonate solution (mmol·L⁻¹: 120 NaCl, 5 KCl, 1.2 MgSO₄, 1.2 KH₂PO₄, 25 NaHCO₃, 11.1 d‐glucose, and 2.5 CaCl₂), bubbled with carbogen (95% O₂ and 5% CO₂). Small mesenteric arteries (third‐order branch of the superior mesenteric artery, internal diameter ~300 μm) and abdominal aortae were isolated, cleared of fat and loose connective tissue, cut into rings 2 mm in length, and mounted on a Mulvany–Halpern wire myograph (Model 610M, Danish Myo Technology, Aarhus, Denmark). For myography experiments, we were unable to achieve equal group sizes between or within animal group comparisons in some cases because we tested combinations of multiple inhibitors in vessels from WKY or SHR treated with placebo or relaxin. For example, control (absence of any inhibitors) curves will have the highest experimental number as these need to be done in every single experiment before the effects of inhibitors can be compared. The remaining arteries were snap frozen in liquid nitrogen and stored at −80°C for further analysis. The heart was isolated and immediately placed in ice‐cold Krebs bicarbonate solution. The whole heart was then blotted dry and weighed before isolating the LV. After weighing the LV, a portion of the LV was snap frozen in liquid nitrogen and stored at −80°C for further analysis.

2.4. Assessment of vascular reactivity ex vivo

After the mesenteric arteries and aortae were mounted on the myograph, the vessels were allowed to stabilise at zero tension for 15 min before normalisation, as described previously (Leo et al., 2016b). All experiments were performed at 37°C, and the organ baths were bubbled with carbogen. Vascular reactivity was assessed, as previously described, with the following modifications (Leo, Hart, & Woodman, 2011). Briefly, mesenteric arteries were maximally contracted with high potassium physiological saline solution (KPSS; mmol·L⁻¹: 25 NaCl, 100 KCl, 1.2 MgSO₄, 1.2 KH₂PO₄, 25 NaHCO₃, 11.1 d‐glucose, and 2.5 CaCl₂), and the integrity of the endothelium was subsequently determined, as described previously (Leo, Jelinic, Parkington, et al., 2014). To evaluate vascular smooth muscle reactivity to vasoconstrictors, cumulative concentration–response curves to the TxA₂ mimetic, https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=1888 (0.1 nmol·L⁻¹–1 μmol·L⁻¹), were constructed. Similarly, to assess endothelial and vascular smooth muscle function, mesenteric arteries were submaximally precontracted to a similar level (70–80% KPSS) using https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=485 (0.1–3 μmol·L⁻¹) and cumulative concentration–response curves to the endothelium‐dependent agonist, ACh (0.1 nmol·L⁻¹–10 μmol·L⁻¹), and the endothelium‐independent vasodilators, https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=9533 (0.01 nmol·L⁻¹–10 μmol·L⁻¹) and https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=1895 (1 pmol·L⁻¹–0.1 μmol·L⁻¹), were determined. In addition, responses to https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=294 were examined after 20‐min incubation with different combinations of pharmacological blockers including the NOS inhibitor, l‐https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=5213 (200 μmol·L⁻¹), the COX inhibitor https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=1909 (Indo, 1 μmol·L⁻¹), the selective COX1 inhibitor https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=10240 (1 μmol·L⁻¹), the selective COX2 inhibitor https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=8976 (1 μmol·L⁻¹), the intermediate‐conductance calcium‐activated potassium channel IK_Ca inhibitor https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2336 (1 μmol·L⁻¹), the small‐conductance calcium‐activated potassium channel (SK_Ca) inhibitor https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2311 (1 μmol·L⁻¹), the Tx receptor antagonist https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=1980 (10 μmol·L⁻¹), and the prostacyclin synthase (PGI₂S) inhibitor U51605 (1 μmol·L⁻¹).

In another separate set of experiments, the effects of relaxin treatment on endothelium‐dependent contraction were determined in endothelium‐intact mesenteric arteries and aortae. Mesenteric arteries were incubated with the combination of l‐NAME, TRAM34, and apamin, and the aorta was incubated with l‐NAME alone for 45 min. After 45 min of incubation, endothelium‐dependent contraction was evoked by cumulative addition of ACh (10 nmol·L⁻¹–100 μmol·L⁻¹). In some cases, the role of the endothelium was investigated in endothelium‐denuded arteries. The endothelium was denuded by rubbing the lumen of the vessel with a hair, and the endothelium was deemed denuded if ACh‐induced relaxation was abolished. To investigate the mechanisms of endothelium‐dependent contraction, responses to ACh were also examined in the presence of Indo, SC560, NS398, and SQ29548.

2.5. Assessment of basal NOS activity

The effects of relaxin treatment on basal levels of NO release were also examined through the addition of l‐NAME (200 μmol·L⁻¹) in endothelium‐intact rings submaximally contracted with phenylephrine (10–100 nmol·L⁻¹) to approximately ~20% of KPSS‐mediated contraction. Under these conditions, contraction to l‐NAME is considered to reflect the level of NOS activity (Kahlberg et al., 2016).

2.6. Measurement of PG metabolites

Segments of mesenteric arteries (~5 branches) and a single aortic ring (2–3 mm) were placed in 0.5 ml or 1 ml of Krebs–HEPES buffer at 37°C, respectively. The equilibration time was 1 hr during which the buffer was changed every 15 min. At the end of the incubation period, ACh (10 μmol·L⁻¹) was applied for 10 min. The mesenteric segments and aortic ring were removed, and the buffer was collected and snap frozen in liquid nitrogen and stored at −80°C for further analysis. The arteries were dried overnight, and dry tissue weight was measured. Briefly, 500 μl of buffer was diluted with 1 ml of hexane:ethyl acetate (1:1 v/v) containing 20 μl of 1‐M citric acid and 20 μl of 1% BHT. After the separation of the organic and the water layer, the organic layer was collected. The organic layer was dried down and reconstituted in 100 μl of methanol. The PG metabolites in the plasma and bathing solution of ACh (10 μmol·L⁻¹)‐stimulated mesenteric arteries were measured by LC‐MS. A total of 24 PG standards from Cayman Chemical Company (Ann Arbor, MI, USA) were used in the LC‐MS (Table S1).

Eicosanoids were detected and quantified using the Agilent 1290 LC system connected to an Agilent triple quad (QQQ 6490) mass spectrometer. Reverse‐phase separation was performed on an Agilent XDB C18 (2.7 μm) 2.1 × 150‐mm column. Eicosanoids were eluted using a mobile phase consisting of (A) 0.1% acetic acid and (B) acetonitrile/isopropanol (90/10, v/v). Gradient elution was carried out for 10 min at a flow rate of 0.6 ml·min⁻¹. Gradient conditions were as follows: 0–1.0 min, 25% B; 1.0–8.0 min, 25–95% B; 8–8.5 min, 95% B; and 8.51–10.0 min, 25% B. A 10‐μl aliquot of each sample was injected onto the column in which temperature was kept at 40°C. Throughout the analysis, samples were kept at 5°C, and MS detection was performed using multiple reaction monitoring. Mass spectrometer parameters were optimised for each analyte and detected in negative mode. Nitrogen gas was employed as the collision gas, and capillary voltage, fragmenter voltage, and collision energy were set to 4,000, 380, and 60 V, respectively. The sheath gas temperature was maintained at 350°C, and the collision gas flow was 12 L·min⁻¹. Data acquisition was performed using MassHunter acquisition software, and all data analysis was performed using MassHunter quantitative software (Agilent).

Similarly, 6‐keto‐PGF_1α, TxB₂, and 8‐isoprostane were measured in ACh (10 μM)‐stimulated aortic rings using EIA kits from Cayman Chemical Company (Ann Arbor, MI, USA) according to the manufacturer's instructions. For elisa experiments, we did not manage to achieve equal group sizes as some of the vessels from the WKY group were used for other experiments, resulting in n = 6 for WKY and n = 8 for SHR + placebo or SHR + relaxin. The buffer was diluted 1:50 and used for the measurement of 6‐keto‐PGF_1α. Undiluted samples were used for the measurement of TxB₂ and 8‐isoprostane. The absorbance counts were normalised to dry tissue weight.

2.7. RNA extraction and quantitative PCR

Frozen blood vessels and left ventricles were pulverised and resuspended in 1‐ml TriReagent (Ambion Inc., Scoresby, VIC, Australia), and total RNA was then extracted according to the manufacturer's instructions. Extraction of RNA from very small amounts of vascular tissue can result in poor quality and/or quantity of RNA (A₂₆₀:A₂₈₀ ratios <1.8). In these instances, we excluded these samples from further analysis (mesenteric arteries: 1 WKY + placebo and 1 SHR + relaxin; aorta: 1 SHR + placebo). RNA pellets were resuspended in 15‐ to 20‐μl RNA Secure™ (Ambion). Quality and quantity of RNA was analysed using the NanoDrop ND1000 Spectrophotometer (Thermo Fischer Scientific Australia Pty Ltd, Scoresby, VIC, Australia) with A₂₆₀:A₂₈₀ ratios >1.8 indicating sufficient quality for qPCR analysis. First‐strand cDNA synthesis used 1 μg of total RNA in a 20‐μl reaction containing random hexamers (50 ng·μl⁻¹) and 200 units of Superscript™ III (Invitrogen, Mulgrave, VIC, Australia).

The comparative cycle threshold (2^−ΔCt) method of quantitative real‐time PCR (qPCR) was used to analyse https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1375 (Ptgs1) and COX‐2 (Ptgs2), https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=265#1356 (Ptgis), and IP receptor (Ptgir), https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1353 (TXAS), and TP receptor (Tp), GTP cyclohydrolase I (Gch‐1), Rxfp1, and https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1249 (Nos3) gene expression in mesenteric and aorta of rats. Rat‐specific forward/reverse primers and 6‐carboxyl fluorescein‐labelled (FAM) TaqMan probes were designed and purchased from Biosearch Technologies (Novato, CA, USA). Primers were designed to span intron/exon boundaries. qPCR was performed in triplicate on the Applied Biosystems ViiA7 PCR machine (Life Technologies, Mulgrave, VIC, Australia) using 96‐well plates with 10‐μl volume reactions in triplicate containing SensiMix (Bioline) and 10 μmol·L⁻¹ of primers and FAM‐labelled probe. Ribosomal 18S (Rn18s) was used as the reference gene. Negative template controls substituting cDNA with water or RT negative controls substituting the reverse transcriptase in the cDNA synthesis were included on each plate. Similarly, qPCR was used to assess potential actions of relaxin on the LV: expression of https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=4890 (Nppb), https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=8927 (Ctgf), https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=5060 (Tgfb1), https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=5074 (Tnf), and https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=3002 (Cybb) in rat left ventricles. Rat‐specific forward/reverse primers (GeneWorks, Thebarton, SA, Australia) were generated from GenBank. Real‐time PCR reaction was determined by SYBR green chemistry using the Applied Biosystems 7500 fast real‐time PCR system with triplicate samples of 12.5 μl containing SYBR Green PCR Master Mix (Applied Biosystems, Scoresby, VIC, Australia) and 10 μmol·L⁻¹ (gene of interest) of primers. Ribosomal 18S (Rn18s) was used as the reference gene. Details of the rat specific primers are in Table S2. The mean C_T value of WKY was used as the internal calibrator and subtracted from the mean C_T value of gene of interest and then analysed using 2^−ΔCt method to reduce unwanted sources of variation, as described previously (Marshall et al., 2016; Marshall, Ng, Unemori, Girling, & Parry, 2016). These normalised data (∆C_T) were expressed as fold change relative to means of controls (WKY) and presented as mean ± SEM.

2.8. Cardiomyocyte morphology

A portion of the LV was embedded in paraffin wax, cut into sections of 5 μm, and mounted on SuperFrost PLUS slides (Menzel‐Gläser, Braunschweig, Germany). After overnight drying at 37°C, the sections were stained with haematoxylin and eosin (Australian Biostain Pty Ltd, Traralgon East, VIC, Australia) for the determination of cardiomyocyte cross‐sectional area and width. Cardiomyocyte cross‐sectional area and width were determined from the same image of 100 individual cardiomyocytes per rat, calculated from cell outlines using ImageJ (RRID:SCR_003070; Ng et al., 2017). The counting and quantification of cardiomyocyte cross‐sectional area and width was performed by another operator who was blinded to the experimental groups.

2.9. Data and statistical analyses

The data and statistical analysis comply with the recommendations of the British Journal of Pharmacology on experimental design and analysis in pharmacology. For all experimental protocols, based on 15% SD, we will be 80% powered to detect 20% changes at P < .05 with n = 9 per group. Concentration–response curves for rat mesenteric arteries were computer fitted to a sigmoidal curve using nonlinear regression (Prism version 5.0, GraphPad Software (RRID:SCR_002798), San Diego, CA, USA) to calculate the sensitivity of each agonist (pEC₅₀). Relaxation to vasodilators was measured as a percentage of precontraction to phenylephrine, with maximal relaxation, R _max. Group pEC₅₀ and R _max values were compared using one‐way ANOVA with post hoc analysis using Tukey's test or Student's unpaired t test, as appropriate. Concentration–response curves were also analysed with repeated measures two‐way ANOVA (treatment vs. concentration). Post hoc analysis was only performed when the F value was greater than F critical value, indicating that there was no variance in homogeneity. P < .05 was considered statistically significant. Certain experiments were undertaken in duplicate (elisa) and triplicate (qPCR) to ensure the reliability of single values. Data analysis and data presentation from these experiments used the single values obtained from the mean of the technical replicates. Statistical analysis was only performed using these independent values (not technical replicates) with n > 5, and any data with sample size of n < 5 are indicated as exploratory observations. n refers to number of animals or independent experiments. Outliers were excluded in data analysis and presentation, where indicated. An outlier is predefined when an individual data point is 2 SDs from the mean. The details of which experiment(s) and how many outliers were excluded are indicated within the respective figure legends.

2.10. Materials

All drugs were purchased from Sigma‐Aldrich (St. Louis, MO, USA), except for U46619, U51605, SC560, and NS398 (Cayman Chemical Company, Ann Arbor, MI, USA). All drugs were dissolved in distilled water, with the exception of indomethacin, which was dissolved in 0.1 mol·L⁻¹ sodium carbonate, and U46619, which was dissolved in 100% ethanol (final concentration less than 0.1% ethanol), as 1 mmol·L⁻¹ stock solution and subsequent dilutions were in distilled water. Furthermore, SC560 and NS398 were dissolved in 100% DMSO, and U51605 was dissolved in 100% EtOH.

2.11. Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Harding, Sharman et al., 2018), and are permanently archived in the Concise Guide to PHARMACOLOGY 2017/18 (Alexander, Christopoulos et al., 2017; Alexander, Fabbro et al., 2017; Alexander, Striessnig et al., 2017).

3. RESULTS

3.1. Relaxin treatment reverses endothelial dysfunction in the mesenteric arteries of SHR

Endothelial dysfunction was evident in the mesenteric arteries of SHR as a significant reduction in the sensitivity but not maximum relaxation in response to ACh compared with arteries from WKY (Figure 1a). Relaxin treatment for 3 days significantly reversed this endothelial dysfunction. Relaxation to the endothelium‐independent vasodilator, sodium nitroprusside, was comparable between the three groups (Figure 1b), indicating that vascular smooth muscle function was not affected. The endothelial dysfunction evident in SHR was partially explained by a reduction in basal NOS activity, which was reversed by relaxin treatment (Figure 1c). Despite changes to basal NOS activity, Nos3 was not altered by hypertension or relaxin treatment (Figure S1a). Moreover, relaxin had no effect on the expression of the rate‐limiting enzyme for tetrahydrobiopterin synthesis, GTP cyclohydrolase I (Gch‐1), the levels of which were significantly reduced in SHR (Figure S1b). Importantly, the beneficial effects of relaxin were independent of any changes in Rxfp1 expression in mesenteric arteries (Figure S1c).

Figure 1.

Short‐term relaxin treatment reverses endothelial dysfunction by increasing NO‐mediated relaxation but not endothelium‐derived hyperpolarisation‐type relaxation. Concentration–response curves to (a) ACh and (b) sodium nitroprusside (SNP) and (c) basal NOS activity of mesenteric arteries from Wistar Kyoto rats (WKY) + placebo, spontaneously hypertensive rats (SHR) + placebo, or SHR + relaxin for 3 days. Concentration–response curves to ACh in the presence of (d) indomethacin (Indo) + l‐NAME (inhibitors of COX and NOS), (e) Indo + l‐NAME + TRAM34 + apamin (inhibitors of COX, NOS, IK_Ca, and SK_Ca, respectively), and (f) Indo + TRAM34 + apamin (inhibitors of COX, IK_Ca, and SK_Ca, respectively) in endothelium‐intact mesenteric arteries of WKY + placebo, SHR + placebo or SHR + relaxin for 3 days. The number of rats used per experimental group is shown in parentheses. ^* P < .05, significantly different from WKY, ^# P < .05, significantly different from SHR + placebo; one‐way ANOVA, Tukey's post hoc test.

To explore the mechanism(s) by which relaxin reverses endothelial dysfunction in mesenteric arteries of SHR, we evaluated vascular reactivity to ACh in the presence of various pharmacological inhibitors. In the presence of Indo + l‐NAME, relaxin treatment for 3 days did not improve the impaired classical EDH‐type relaxation observed in SHR (Figure 1d). In WKY, a significant component of relaxation remained in the presence of Indo + l‐NAME + apamin + TRAM34. However, it was abolished in both placebo‐ and relaxin‐treated SHR, suggesting that hypertension per se impairs the contribution of non‐classical EDH responses (Edwards, Feletou, & Weston, 2010; Figure 1e). In contrast, the impaired NO‐mediated relaxation evident in SHR (Figure 1f) was reversed by treatment with relaxin.

A key finding was the significant contribution of COX‐derived vasoconstrictor prostanoids in SHR mesenteric arteries, demonstrated by an increase in sensitivity to ACh in the presence of a non‐selective COX inhibitor, Indo (Figure 2b). This increased ACh sensitivity was not observed in WKY (Figure 2a) or relaxin‐treated SHR (Figure 2c), indicative that relaxin is correcting the vasoconstrictor prostanoid imbalance. To further explore the involvement of NO and prostanoids, vascular reactivity to ACh was evaluated in the presence of TRAM34 + apamin to abolish the actions of the “classical” EDH (Figure S2). Interestingly, we observed EDC in response to the higher ACh concentrations in SHR. The EDC was almost completely abolished by Indo, the PGI₂ synthase inhibitor, U51605, and TP antagonist, SQ29548 (Figure S2b,c), suggesting these contractions are mediated by COX‐derived PGI₂ targeting the TP. Furthermore, the magnitude of this ACh‐evoked contraction was significantly reduced in relaxin‐treated SHR (Figure S2a), indicating that this constrictor pathway was down‐regulated by relaxin treatment.

Figure 2.

Short‐term relaxin treatment reverses the contribution of COX‐derived vasoconstrictor prostanoids. Concentration–response curves to ACh in endothelium‐intact mesenteric arteries of (a) Wistar Kyoto rats (WKY) + placebo, (b) spontaneously hypertensive rats (SHR) + placebo, or (c) SHR + relaxin in the absence (control) or in the presence of indomethacin (Indo; 1 μmol·L⁻¹, non‐selective COX inhibitor). The number of rats used per experimental group is shown in parentheses. ^* P < .05, significantly different from control; unpaired Student's t test.

3.2. Relaxin treatment inhibited endothelium‐dependent contraction in SHR

The nature of EDC in mesenteric arteries of SHR was further investigated by assessing the response to ACh in the presence of l‐NAME + TRAM34 + apamin (Figure 3). ACh‐evoked concentration‐dependent contraction was significantly increased in SHR compared with WKY; it was endothelium dependent as contraction was abolished in endothelium‐denuded arteries (data not shown). Consistent with data in Figure S2, this EDC was significantly reduced in relaxin‐treated SHR (Figure 3a). Furthermore, EDC was also significantly increased in the aorta of SHR, and this was similarly reduced by short‐term relaxin treatment (Figure 3b,c).

Figure 3.

Short‐term relaxin treatment reverses endothelium‐dependent contraction in mesenteric arteries and aorta. Concentration–response curves to ACh in endothelium‐intact, (a) resting mesenteric arteries (in the presence of l‐NAME + TRAM34 + apamin) or (b, c) aorta (in the presence of l‐NAME or indomethacin [Indo] + l‐NAME, respectively) of Wistar Kyoto rats (WKY) + placebo, spontaneously hypertensive rats (SHR) + placebo, or SHR + relaxin for 3 days. Concentration–response curves to ACh in endothelium‐intact mesenteric arteries of (d) SHR + placebo, (e) SHR + relaxin, and (f) AUC in l‐NAME + TRAM34 + apamin (control) or in the presence of Indo (1 μmol·L⁻¹, non‐selective COX inhibitor), SC560 (1 μmol·L⁻¹, COX1 inhibitor), NS398 (1 μmol·L⁻¹, COX2 inhibitor), and SQ29548 (10 μmol·L⁻¹, TP receptor antagonist). The number of rats used per experimental group is shown in parentheses. ^* P < .05, significantly different from WKY, ^# P < .05, significantly different from SHR + placebo; one‐way ANOVA, Tukey's post hoc test

3.3. Relaxin‐dependent activation of prostanoid pathways

We investigated the underlying mechanisms associated with the reduction in ACh‐induced contraction in SHR after relaxin treatment. Blockade of COX1 and TP receptors abolished ACh‐mediated contraction, indicating their involvement in EDC in the SHR (placebo and relaxin treated) mesenteric arteries (Figure 3d,e). Furthermore, inhibition of COX2 (with NS398) had minimal effects on the contraction to ACh (Figure 3d,e). Conversely, COX1‐derived ACh‐evoked contraction was significantly reduced with relaxin treatment (Figure 3f). The involvement of COX‐derived prostanoids was also demonstrated in the aorta as EDC was completely abolished by Indo (Figure 3c).

We further tested our hypothesis that relaxin treatment in SHR could directly modulate prostanoid synthesis by analysing gene expression of key enzymes in the prostanoid synthesis pathway. In the mesenteric arteries, neither hypertension nor relaxin had any significant effects on expression of prostanoid‐generating enzymes: Ptgs1, Ptgs2, Ptgis, or Tbxas1 (Figure 4a–d). In contrast, expression of Ptgs1 (but not Ptgs2) was significantly increased in the aorta of SHR and reduced by relaxin treatment (Figure 5a,b). Similarly, Ptgis and Tbxas1 in the aorta were significantly up‐regulated by hypertension; their expression was also reduced in relaxin‐treated SHR (Figure 5c,d).

Figure 4.

Gene expression of prostanoid‐synthesising enzymes in the mesenteric arteries. Quantitative analysis of (a) Ptgs1, (b) Ptgs2, (c) Ptgis, and (d) Tbxas1 expression in mesenteric arteries of Wistar Kyoto rats (WKY) + placebo, spontaneously hypertensive rats (SHR) + placebo, or SHR + relaxin for 3 days. The number of rats used per experimental group is shown in parentheses. Outlier exclusion: (a, b) 2 SHR + placebo and 1 SHR + relaxin and (d) 1 SHR + placebo

Figure 5.

Gene expression of prostanoid‐synthesising enzymes in the aorta. Quantitative analysis of (a) Ptgs1, (b) Ptgs2, (c) Ptgis, and (d) Tbxas1 expression in aorta of Wistar Kyoto rats (WKY) + placebo, spontaneously hypertensive rats (SHR) + placebo, or spontaneously hypertensive rats (SHR) + relaxin for 3 days. The number of rats used per experimental group is shown in parentheses. ^* P < .05, significantly different from WKY, ^# P < .05, significantly different from SHR + placebo; one‐way ANOVA, Tukey's post hoc test. Outlier exclusion: (a, b) 1 WKY, 1 SHR + placebo, and 1 SHR + relaxin, (c) 1 WKY and 1 SHR + placebo, and (d) 1 WKY

Our previous findings demonstrated that relaxin treatment for 3 days stimulates PGI₂ production in mesenteric arteries from normotensive Wistar rats (Leo et al., 2016b). On this basis, we used LC‐MS to screen for the potential involvement of 24 different PG metabolites (Table S2) to further explore if prostanoid production was altered by relaxin treatment in the mesenteric arteries. Eight out of 24 metabolites were detected, whereas the remaining 16 were below the detection limit. ACh‐stimulated production of https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2391 (Figure 6a), https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=1884 _2α, https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=1051, https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=6279 ₂/PGB₂, and https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2487 ₄ in the mesenteric arteries was not affected by hypertension or relaxin treatment (Figure S3a–d). In contrast, ACh‐induced release of 6‐keto‐PGF_1α (stable metabolite of PGI₂), 8‐iso‐PGF_2α (non‐enzymatic peroxidation of arachidonic acid), and https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=1883 ₂/https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=1881 ₂ was significantly increased in SHR; the increase in the latter two metabolites was not significant in relaxin‐treated SHR compared with WKY (Figure 6b–d). In the aorta, release of 6‐keto‐PGF_1α was significantly increased but marginal for TXB₂ after ACh stimulation in SHR, but relaxin treatment elicited no further effects (Figure 6e,f).

Figure 6.

Metabolite profiling of PG production in (a–d) mesenteric arteries and (e, f) aorta. Measurement of (a) arachidonic acid, (b) PGE₂/PGD₂, (c) 8‐iso‐PGF_2α, and (d, e) 6‐keto‐PGF_1α and (f) TXB₂ production in ACh (10 μmol·L⁻¹)‐stimulated (a–d) mesenteric arteries and (e, f) aorta from Wistar Kyoto rats (WKY) + placebo, spontaneously hypertensive rats (SHR) + placebo, or SHR + relaxin for 3 days. PGE₂/PGD₂ are grouped together as the retention time cannot be sufficiently distinguished. The number of rats used per experimental group is shown in parentheses. ^* P < .05, significantly different from WKY, ^# P < .05, significantly different from SHR + placebo; one‐way ANOVA, Tukey's post hoc test. Outlier exclusion: (a) 1 SHR + relaxin, (b) 1 SHR + placebo, (c) 1 SHR + placebo, and (d) 2 SHR + placebo

3.4. Relaxin‐mediated activation of prostanoid receptors in the vascular smooth muscle

Contractions to the TP agonist, U46619, were increased in mesenteric arteries of SHR compared with WKY; this hyperresponsiveness to U46619 was significantly reduced in relaxin‐treated SHR (Figure 7a). In addition, relaxin reversed the impaired vasodilation to the IP receptor agonist, iloprost (Figure 7b), suggesting that this may be underpinned by changes in receptor numbers, affinity of the ligand to the receptor or activity of the signalling pathway. We conducted a different experiment to determine if activation of IP receptors was directly contributing to the contraction in the SHR aorta. In WKY, iloprost elicited ~20% contraction (Figure 7c), but this was significantly increased to ~60% in SHR. Relaxin treatment reduced this iloprost‐evoked contraction to below 20% (Figure 7c). Similarly, the sensitivity to U46619‐induced contraction was increased in aorta of SHR compared with WKY (SHR pEC₅₀: 7.41 ± 0.05 vs. WKY pEC₅₀: 7.07 ± 0.03). However, in contrast to the mesenteric arteries, this hypersensitivity to U46619 was not significantly reduced (SHR + relaxin pEC₅₀: 7.29 ± 0.06) in the aorta of relaxin‐treated SHR (Figure S4). In support of our pharmacological data, Ptgir expression was significantly reduced in the mesenteric arteries of SHR and was up‐regulated after relaxin treatment (Figure 8a). However, there were minimal effects on Tbxa2r expression (Figure 8b). In contrast, there were significant increases in Ptgir and Tbxa2r expression in the aorta of SHR and a clear down‐regulation of both receptors after relaxin treatment (Figure 8c,d). Collectively, these data illustrate that relaxin can directly mediate vascular smooth muscle responses to prostanoids in both arteries but by distinct mechanisms.

Figure 7.

Short‐term relaxin treatment altered vascular reactivity to prostanoid receptor activation in the mesenteric arteries and aorta. Concentration–response curves to (a) U46619 and (b) iloprost in endothelium‐intact mesenteric arteries and (c) contraction to iloprost in the presence of l‐NAME (NOS inhibitor) of endothelium‐intact, resting aorta from Wistar Kyoto rats (WKY) + placebo, spontaneously hypertensive rats (SHR) + placebo, or SHR + relaxin for 3 days. The number of rats used per experimental group is shown in parentheses. ^* P < .05, significantly different from WKY, ^# P < .05, significantly different from SHR + placebo; one‐way ANOVA, Tukey's post hoc test.

Figure 8.

Gene expression of TP and IP receptors in (a, b) mesenteric arteries and (c, d) aorta. Quantitative analysis of (a, c) Ptgir and (b, d) Tbxa2r expression in the (a, b) mesenteric arteries and (c, d) aorta of Wistar Kyoto rats (WKY) + placebo, spontaneously hypertensive rats (SHR) + placebo, or SHR + relaxin for 3 days. The number of rats used per experimental group is shown in parentheses. ^* P < .05, significantly different from WKY, ^# P < .05, significantly different from SHR + placebo; one‐way ANOVA, Tukey's post hoc test. Outlier exclusion: (b) 1 SHR + placebo and (c, d) 1 SHR + placebo and 1 SHR + placebo

3.5. Relaxin had no effect on cardiomyocyte size or biomarkers of fibrosis in the LV of SHR

The effects of relaxin on the LV have previously been shown in the SHR myocardium but only after 14 days of infusion (Lekgabe et al., 2005). Therefore, in the current study, we decided to evaluate the additional beneficial effects of relaxin on the LV after a much shorter duration of treatment. Body weights and tibial length were comparable in all groups regardless of hypertension or relaxin treatment (Figure 9a,b). Whole heart and LV weights were significantly increased in SHR compared with WKY, and this did not change after relaxin treatment (Figure 9c,d). Exploratory histological assessment of the LV demonstrated that cardiomyocyte cross‐sectional area and width (Figure 10a–c, n = 4) were increased in SHR. This was accompanied by a significant increase in expression of B‐type natriuretic peptide (Nppb) in the LV (Figure 10d). Short‐term relaxin treatment had no effect on cardiomyocyte cross‐sectional area or width (Figure 10a–c), but Nppb expression was increased (Figure 10d).

Figure 9.

Short‐term relaxin treatment had no effect on heart and left ventricle weights. (a) Body weights, (b) tibial length, (c) heart weight, and (d) left ventricle weight of Wistar Kyoto rats (WKY) + placebo, spontaneously hypertensive rats (SHR) + placebo, or SHR + relaxin for 3 days. The number of rats used per experimental group is shown in parentheses. ^* P < .05, significantly different from WKY + placebo; one‐way ANOVA, Tukey's post hoc test)

Figure 10.

Short‐term relaxin had no effect on cardiac hypertrophy and biomarkers of early‐onset fibrosis. (a) Representative images and quantification of cardiomyocyte (b) cross‐sectional area and (c) width of haematoxylin and eosin‐stained cardiomyocytes in the left ventricle of Wistar Kyoto rats (WKY) + placebo, spontaneously hypertensive rats (SHR) + placebo, or SHR + relaxin for 3 days. Quantitative analysis of (d) nppb, (e) ctgf, and (f) tgfb mRNA expression in the left ventricle of WKY + placebo, SHR + placebo, or SHR + relaxin for 3 days. The number of rats used per experimental group is shown in parentheses. ^* P < .05, significantly different from WKY + placebo, ^# P < .05, significantly different from SHR + placebo; one‐way ANOVA, Tukey's post hoc test. Outlier exclusion: (d–f) 1 WKY

We also analysed expression of biomarkers of early onset of fibrosis, connective tissue growth factor (CTGF), and TGF‐β. Both Ctgf and Tgfb expressions were significantly increased in the LV of SHR compared with WKY and remained increased after relaxin treatment (Figure 10e,f). Similarly, Cybb was increased in the LV of SHR but was not reduced after relaxin treatment (Figure S5a). Neither hypertension nor relaxin treatment affected TNF‐α (Tnf; Figure S5b).

4. DISCUSSION

This study demonstrates that endothelial dysfunction in SHR is attributed to the reduction in the vasodilator effects of NO and EDH and enhanced production of EDCFs. Endothelium‐dependent contraction in mesenteric arteries is, at least in part, mediated by increased production of COX‐1‐derived PGI₂. As a result of decreased expression of IP receptors in the mesenteric arteries of SHR, the excessive production of PGI₂ acts on TP receptors to cause contraction. Short‐term relaxin infusion reverses hypertension‐induced endothelial dysfunction by increasing NO and reducing EDC, without any effect on EDH‐type relaxation. More importantly, relaxin treatment increases the expression of IP receptors, thus allowing PGI₂ to act on these receptors to elicit vasorelaxation. Endothelium‐dependent contraction is also enhanced in the aorta of SHR and reversed by relaxin treatment. However, the mechanisms of relaxin action on vasorelaxation and EDC in the two arteries are different. Specifically, relaxin suppresses the hypertension‐associated increase in prostanoid‐producing enzymes (COX1, PGIS, and TXAS) and diminishes the PGI₂‐evoked contraction in the aorta of SHR. In summary, we have shown that relaxin diminishes EDC in the vasculature in addition to stimulating vasodilator pathways (Figure S6). This important finding opens up new possibilities for relaxin as a potential therapeutic agent for vascular diseases that are underpinned by enhanced PG‐mediated EDC.

Many studies demonstrate that short‐term relaxin treatment augments endothelial vasodilator function in resistance‐size vessels under normotensive conditions; however, the vascular effects of relaxin in hypertension are less consistent (Chan et al., 2013; Xu et al., 2010). Discrepancies between studies on SHR are attributed to the severity of the disease, age of SHR, duration of treatment, and/or vessel‐specific effects of relaxin. Furthermore, the short‐term (2–3 days) effects of relaxin on EDC have not been considered. In our study that used SHR (aged 24–26 weeks old), we demonstrated that subcutaneous infusion of relaxin for 3 days improved endothelial vasodilator function in the mesenteric arteries. This is likely to be independent of a reduction in BP because relaxin has no effect on systolic BP in SHR (Chan et al., 2013; Debrah, Conrad, Jeyabalan, et al., 2005; Parikh et al., 2013; Xu et al., 2010). Consistent with previous studies (Gluais et al., 2005; Michel, Man, Man, & Vanhoutte, 2008; Tang et al., 2008; Tang & Vanhoutte, 2009), one of the important contributors to endothelial dysfunction in the SHR is the overproduction of COX‐derived vasoconstrictor prostanoids (Félétou et al., 2009; Tang & Vanhoutte, 2009) and enhanced EDC. Of significance, relaxin treatment completely counteracted this EDC in both mesenteric arteries and aorta of SHR. In all cases, EDC was abolished by Indo, confirming that a COX‐derived product was responsible for causing contraction in these arteries.

COXs are the first enzymes involved in the biosynthetic pathway that lead to prostanoid formation from arachidonic acid. There are two isoforms of COX, the constitutive isoform, COX1, and the inducible isoform, COX2 (Félétou, Huang, & Vanhoutte, 2011). In this study, we demonstrated that EDC to ACh were blocked by selective COX1 (but not selective COX2) inhibition. Furthermore, enhanced COX‐mediated contraction in SHR occurs without changes to COX1 and COX2 mRNA expression in the mesenteric arteries. In contrast, in SHR aorta, relaxin treatment reversed the increase in COX1 but not COX2 expression. Consistent with our data, EDC is abolished in aortae of COX1 knockout mice but not in COX2 knockout mice (Tang et al., 2005; Tang & Vanhoutte, 2008). Our previous work demonstrated that short‐term relaxin treatment increased COX2 expression in normotensive mesenteric arteries, suggesting the role of COX2 cannot be completely excluded (Leo et al., 2016b). Indeed, our data indicate a small contribution of COX2‐derived EDC in the mesenteric arteries as also reported in previous studies (Martelli et al., 2013; Wong et al., 2009). Given the greater contribution of COX1‐derived EDC in SHR, it is more likely that relaxin mitigates COX1‐derived vasoconstrictor prostanoids. Collectively, relaxin treatment reduces EDC in SHR arteries regardless of its action on COX1 or COX2 gene expression. However, it is important to note that no alteration in COX gene expression does not rule out any potential changes in protein levels or enzymic activities.

COX‐derived prostanoids consist of a variety of biologically active eicosanoids that are formed from the short‐lasting but biologically active PGH₂, through the action of a set of synthases, namely, PGD, PGE, PGF, PGI, and TXA synthases. These different synthases convert PGH₂ to their respective eicosanoids such as PGD₂, PGE₂, PGF_2α, PGI₂, and TXA₂. These PGs will preferentially interact with their specific GPCRs, which are classified in five subtypes DP, EP, FP, IP, and TP on the vascular smooth muscles cells (Félétou et al., 2011). Prostanoid production in aorta and renal arteries is altered by hypertension (Gluais et al., 2005; Michel et al., 2008). For example, PGI₂S is up‐regulated and accompanied by increased PGI₂, PGE₂, and 8‐iso‐PGF_2α, the product of oxidative modification of free fatty acids. In parallel, IP receptors are down‐regulated, causing these prostanoids to converge and bind to the overexpressed TP and/or EP receptors, promoting vasoconstriction (Félétou et al., 2009; Gluais et al., 2005; Liu et al., 2017; Michel et al., 2008; Tang & Vanhoutte, 2009). To characterise which eicosanoids were altered by relaxin treatment, we measured prostanoid production in mesenteric arteries and aorta using LC‐MS and elisa, respectively. Hypertension selectively increased PGI₂ and 8‐iso‐PGF_2α production. In the aorta, increased PGI₂ production was associated with increased PGI₂S gene expression. PGE₂/PGD₂ was also elevated in the SHR mesenteric arteries. In our previous study using normotensive rats, we demonstrated that relaxin infusion for 3 days enhanced bradykinin‐mediated PGI₂ production in mesenteric arteries. It is important to note that there was a compensatory increase in PGI₂ in SHR, so it was not clear if relaxin treatment would further exacerbate or mitigate PGI₂ production. Surprisingly, relaxin had minimal effects on prostanoid production in SHR aorta and mesenteric arteries. One possible explanation is that PGI₂ is already significantly elevated in SHR, preventing relaxin from causing any further increase. In rat mesenteric arteries, PGE₂ and 8‐iso‐PGF_2α evoke vasoconstriction, and some of this is mediated via interaction with TP receptors (Kobayashi, Murata, Hori, & Ozaki, 2011; Kondo, Okuno, Suzuki, & Saruta, 1980). Enhanced production of these constrictor prostanoids may contribute to the TP receptor‐mediated contraction in SHR mesenteric arteries. Our data indicate that relaxin may down‐regulate the production of some vasoconstrictor prostanoids (PGE₂/PGD₂ and 8‐iso‐PGF_2α) in SHR mesenteric arteries, but it failed to reach statistical significance.

Previous studies have demonstrated that short‐term relaxin infusion up‐regulates IP receptor expression in mesenteric arteries of normotensive rats (Jelinic et al., 2017). Conversely, IP receptor expression is decreased in the aorta of male relaxin gene‐deficient mice; this is accompanied by impaired functional responses to the PGI₂ analogue, iloprost (Ng et al., 2015). Similarly, endothelial dysfunction is associated with up‐regulation of vasoconstrictor prostanoids in the mesenteric arteries of relaxin‐deficient mice (Leo, Jelinic, Gooi, Tare, & Parry, 2014). These findings suggest that relaxin is an important regulator of PGI₂–IP receptor pathway in the vasculature of normotensive animals. In this study, IP and TP receptors were both up‐regulated in the aorta of SHR. Moreover, we also showed a decrease in IP but not TP receptors in the mesenteric arteries. Similarly, there were differential effects of relaxin on the expression of these receptors. Specifically, relaxin treatment reduced the expression of TP and IP receptors in the SHR aorta, whereas it increased IP receptors in the mesenteric arteries. Taken together, our findings support the hypothesis that relaxin regulates IP receptor expression/activity in the mesenteric arteries, which restores the ability of PGI₂ to cause vasorelaxation in SHR vessels. In addition, we have uncovered a completely different mechanism of relaxin action in the aorta. This involves a relaxin‐mediated decrease in expression of TP and IP receptors, thereby suppressing the vasoconstrictor action of PGI₂ and other prostanoids.

In this study, relaxin treatment reversed EDC in the vasculature by altering expression and/or activity of IP receptors. There is limited information in the literature to explain regulation of IP receptor expression in vascular cells. Given that several key transcriptional regulators such as oestrogen response element, sterol response element binding protein, or enhancer binding proteins are involved in the complex regulation of IP receptor expression (Reid & Kinsella, 2015), it is possible that relaxin may interact with any of these to alter IP receptor expression. However, additional experiments beyond the scope of this study will be required to prove this hypothesis. Other than expression, the activity of IP receptors can also be influenced by GPCR‐interacting proteins (Reid & Kinsella, 2015). One of such GPCR‐interacting proteins is known as postsynaptic density 95/disc large/zonula occludens‐1 domains (PDZK1; Reid & Kinsella, 2015). Unphosphorylated PDZK1 is constitutively associated in a complex with the IP receptors. PDZK1 is thought to increase IP receptor expression at the plasma membrane and enhance agonist binding. Upon receptor activation, IP receptors undergo an agonist‐induced conformational change, leading to dissociation of PDZK1. The free PDZK1 is dependent on cAMP‐dependent PKA‐mediated phosphorylation to trigger its re‐association with the IP receptor (Reid & Kinsella, 2015). In SHR, production of PGI₂ is enhanced, suggesting that there may be an increase in the proportion of PDZK1 that is not associated with IP receptors. Relaxin increases cAMP accumulation and triggers PKA‐medicated phosphorylation in cells (Singh, Simpson, & Bennett, 2015). Therefore, it is possible that relaxin treatment may cause the re‐association of free PDZK1 with the IP receptors, leading to an increase in the availability of “activatable” IP receptors and therefore IP activity in the SHR arteries. This possibility requires exploration in future studies.

In addition to EDCFs, impaired NO bioavailability and EDH also contribute to endothelial dysfunction in hypertension. Vascular oxidative stress and uncoupling of eNOS occur in SHR arteries, leading to decreases in NOS activity and NO bioavailability (Macarthur, Westfall, & Wilken, 2008; Mason et al., 2006). In the current study, relaxin reversed hypertension‐induced endothelial dysfunction in the mesenteric arteries by increasing basal NOS activity and NO‐mediated relaxation, without any changes in eNOS and GTPCH1. This suggests that relaxin may regulate eNOS function at a post‐translational level to increase NO bioavailability. We have previously shown that relaxin enhances NOS activity via reduction of vascular oxidative stress (Leo et al., 2017), activation of eNOS phosphorylation at Ser¹¹⁷⁷ (Leo et al., 2016b; Leo, Jelinic, Parkington, et al., 2014; McGuane et al., 2011), and up‐regulation of eNOS dimerisation (Jelinic et al., 2017). Furthermore, the ability of relaxin to increase NO bioavailability and prevent endothelial dysfunction is observed in a number of disease conditions, including TNF‐α incubation (Dschietzig et al., 2012), cigarette smoke (Pini et al., 2016), arteriosclerosis (Tiyerili et al., 2016), and diabetes (Ng et al., 2016; Ng et al., 2017). Previous studies also reported that acute intravenous injection of relaxin for 3 hr increases the contribution of EDH to endothelium‐dependent relaxation in the resistance‐size arteries, but it primarily involves the activation of IK_Ca (Chan et al., 2013; Leo, Jelinic, Parkington, et al., 2014; Marshall et al., 2017). The impaired EDH‐type responses in the SHR mesenteric arteries were not reversed by 3 days of relaxin treatment. In contrast, 14 days of relaxin treatment decreases myogenic tone and increases EDH‐type relaxation in the cerebral parenchymal arterioles (Chan et al., 2013). Therefore, a longer duration of relaxin treatment may be required to reverse the impaired EDH‐type relaxation in SHR (Chan et al., 2013).

Although relaxin produces vasoprotective effects in mesenteric arteries and aorta, this has not translated into a direct effect on BP in the SHR. However, this does not preclude marked changes in the circulation, such as systemic arterial resistance, global arterial compliance, or cardiac output. Indeed, many other published studies also show that relaxin's vasoprotective actions are independent of any normalisation of BP (Chan et al., 2013; van Drongelen et al., 2013; Xu et al., 2010). Of note, relaxin treatment reduces arterial load (by decreasing systemic vascular resistance and increasing arterial compliance) while also increasing cardiac output, heart rate, and stroke volume in hypertensive rats, without any effect on mean arterial pressure (Debrah, Conrad, Danielson, & Shroff, 2005).

Although the primary focus of this study was the vasculature, it provided an opportunity to assess any additional beneficial effects on the myocardium after 3 days of relaxin infusion. This is because anti‐fibrotic and anti‐hypertrophic actions of relaxin were reported in the myocardium of 9‐ to 10‐month‐old SHR after 14 days of treatment (Samuel et al., 2017; Samuel, Summers, & Hewitson, 2016). It is important to note that the SHR used in our study were relatively young (6–7 months). Nevertheless, there was evidence of cardiomyocyte hypertrophy and increased expression of biomarkers of early‐onset cardiac fibrosis in the SHR at this age. In our study, we showed that a 3‐day infusion of relaxin had no significant effect on LV cardiomyocyte morphometry or expression of four biomarkers of fibrosis and hypertrophy we selected to analyse. On this basis, we decided not to investigate fibrosis in the LV any further. However, as we did not directly measure collagen deposition or conduct a more comprehensive analysis of fibrosis, we cannot conclude that 3 days of relaxin treatment had no anti‐fibrotic effects on the myocardium. Importantly, our data clearly demonstrate that a 3‐day relaxin infusion is sufficient to modify many aspects of vascular dysfunction associated with hypertension, but the short time course of infusion was not sufficient to have significant cardiac effects.

In conclusion, short‐term relaxin treatment reverses hypertension‐induced vascular dysfunction through novel interactions with EDCFs and their associated vasoconstrictor pathways. Relaxin clearly has region‐dependent effects on the vasculature, although the overall therapeutic endpoint is a reduction in EDC by restoring endothelium‐derived PGI₂ vasodilator pathways. We propose that relaxin or its newer generation mimetics such as B7–33 or https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=8322 could be progressed as novel therapeutic agents for the treatment of the vascular complications of hypertension and/or other vascular diseases that are associated with increased endothelium‐derived vasoconstriction (Leo, Jelinic, Ng, Parry, & Tare, 2019; Ng, Leo, Parry, & Ritchie, 2018).

AUTHOR CONTRIBUTIONS

C.H.L., H.H.N., S.A.M, M.J., and C.Q. performed the research and analysed the data. C.H.L., T.R., U.R., R.H.R., M.T., and L.J.P. designed the study. C.H.L. and L.J.P. wrote the paper. All authors revised the paper critically for important intellectual content and approved the final submission of paper.

CONFLICT OF INTEREST

The authors disclose that this project was partially funded by Novartis Pharma AG, who also provided the recombinant human relaxin‐2 as a condition of an Australian Research Council Linkage Grant. L.J.P. was also a paid consultant for Novartis Pharma AG and is a co‐inventor on a patent for relaxin use in the cervix, kidney, and brain.

DECLARATION OF TRANSPARENCY AND SCIENTIFIC RIGOUR

This Declaration acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research as stated in the BJP guidelines for https://bpspubs.onlinelibrary.wiley.com/doi/full/10.1111/bph.14207 and https://bpspubs.onlinelibrary.wiley.com/doi/full/10.1111/bph.14206 and as recommended by funding agencies, publishers, and other organisations engaged with supporting research.

Supporting information

Table S1. Rat specific forward and reverse primers used in cardiac (1–6) and vascular (7–16) tissues for qPCR

Table S2. Prostaglandin standards used in liquid chromatography mass spectrometry.

Figure S1. Relaxin treatment has no effect on NOS3, GTPCH1 and RXFP1 expression in the mesenteric arteries.

Figure S2. Relaxin treatment mitigates endothelium‐dependent contraction.

Figure S3. Metabolite profiling of prostaglandin production in mesenteric arteries.

Figure S4: Relaxin treatment has no effect on vascular reactivity to thromboxane receptor activation in the aorta.

Figure S5. Relaxin treatment has no effect on NOX2 and TNFα expression in the left ventricle.

Figure S6. Proposed mechanisms of relaxin action after 3 days infusion in rat mesenteric arteries (A) and aortae (B) from SHR. A) In mesenteric arteries of SHR, there is endothelial vasodilator dysfunction. There is also an increase in endothelium‐derived PGI2 production, which targets other prostanoid receptors (TP or possibly EP) on the smooth muscle to cause contraction. Enhanced PGD/E2 and 8‐iso PGF2α production also contributes to this contraction. Relaxin administration for 3 days increases basal NOS activity and enhances ACh‐evoked NO‐mediated relaxation in SHR mesenteric arteries. It also upregulates IP expression and/or activity, leading to a decrease in prostanoid‐mediated vasoconstriction. Relaxin may also downregulate production of PGD/E2 and 8‐iso PGF2α. Overall, relaxin treatment reverses endothelial dysfunction in SHR mesenteric arteries. B) In the aortae of SHR, there is increased production of PGI2 and TXA2, which act on TP to cause contraction. Relaxin treatment for 3 days downregulates expression of prostanoid synthesizing enzymes (COX1, PGI2S and TXA2S) and receptors (TP and IP) without affecting PGI2 and TXA2 levels in SHR. In addition, relaxin decreases IP‐evoked contraction. Therefore, in the aorta relaxin treatment reduces endothelium‐derived vasoconstriction via downregulation of TP receptors and/or activity in the SHR.

Figure S7. Supporting Information

Leo CH, Ng HH, Marshall SA, et al. Relaxin reduces endothelium‐derived vasoconstriction in hypertension: Revealing new therapeutic insights. Br J Pharmacol. 2020;177:217–233. 10.1111/bph.14858