Sexual dimorphism in prostacyclin‐mimetic responses within rat mesenteric arteries: A novel role for KV7.1 in shaping IP receptor‐mediated relaxation

Background and Purpose Prostacyclin mimetics express potent vasoactive effects via prostanoid receptors that are not unequivocally defined, as to date no study has considered sex as a factor. The aim of this study was to determine the contribution of IP and EP3 prostanoid receptors to prostacyclin mimetic iloprost‐mediated responses, whether KV7.1–5 channels represent downstream targets of selective prostacyclin‐IP‐receptor agonist MRE‐269 and the impact of the oestrus cycle on vascular reactivity. Experimental Approach Within second‐order mesenteric arteries from male and female Wistar rats, we determined (1) relative mRNA transcripts for EP1–4 (Ptger 1–4 ), IP (Ptgi) and TXA2 (Tbxa) prostanoid receptors via RT‐qPCR; (2) the effect of iloprost, MRE‐269, isoprenaline and ML277 on precontracted arterial tone in the presence of inhibitors of prostanoid receptors, potassium channels and the molecular interference of KV7.1 via wire‐myograph; (3) oestrus cycle stage via histological changes in cervical cell preparations. Key Results Iloprost evoked a biphasic response in male mesenteric arteries, at concentrations ≤100 nmol·L−1 relaxing, then contracting the vessel at concentration ≥300 nmol·L−1, a process attributed to IP and EP3 receptors respectively. Secondary contraction was absent in the females, which was associated with a reduction in Ptger3. Pharmacological inhibition and molecular interference of KV7.1 significantly attenuated relaxations produced by the selective IP receptor agonist MRE‐269 in male and female Wistar in dioestrus/metoestrus, but not pro‐oestrus/oestrus. Conclusions and Implications Stark sexual dimorphisms in iloprost‐mediated vasoactive responses are present within mesenteric arteries. KV7.1 is implicated in IP receptor‐mediated vasorelaxation and is impaired by the oestrus cycle.

As the contribution of K V 7 channels to prostanoid receptormediated relaxations is unknown, we sought to determine whether K V 7 channels were involved with prostacyclin mimetic-mediated responses in rat mesenteric arteries from aged-matched male and female rats. This artery was chosen because (1), K V 7 expression has been established Mackie et al., 2008), (2) K V 7 activators are effective relaxants , (3) a role for K V 7 channels in G αs -linked responses has been identified (e.g. Lindman et al., 2018;Stott et al., 2016Stott et al., , 2018 and (4), endothelium dependent production of PGI 2 mediates concomitant IP receptor-mediated relaxation (Liu et al., 2012) and TP/EP 3 -mediated contraction (Liu et al., 2012(Liu et al., , 2017. Moreover, in line with Docherty et al. (2019), we investigated possible sex difference as nothing is known about the impact of sex on prostanoid-mediated vascular responses. To circumvent the short half-life of prostacyclin, we characterized the contribution of EP 3 and IP receptors to responses mediated by its stable analogue iloprost and defined the contribution of K V 7 channels to IP receptor-mediated vasorelaxation using a selective IP receptor agonist, MRE-269. Our data demonstrate a striking sex-dependent difference in response to iloprost and a role for K V 7.1 channels in shaping IP receptor-evoked vasorelaxation within rat mesenteric arteries, which is oestrus cycle sensitive. What is already known • The prostacyclin analogue iloprost evokes vasoactive responses through a myriad of receptors.
• K V 7 channels are targets of endogenous vasoactive signalling cascades.

What does this study add
• Iloprost evoked biphasic relaxant-contrail responses in males, which is absent in females.
• K V 7.1 inhibition impairs MRE-269 mesenteric arteries relaxation which is affected by the oestrus cycle.

What is the clinical significance
• Sex must be considered as a factor when considering prostacyclin mimetics as a therapeutic tool.
Act 1986. Animals were kept in LSB Aspen woodchip bedding. Animals were killed by cervical dislocation with secondary confirmation via cessation of the circulation by femoral artery severance in accordance with Schedule 1 of the ASPA 1986.

| Oestrus cycle stage determination
After killing, 50 μl of PSS was inserted into the vaginal canal via a 2to 200-μl pipette tip and flushed four to six times to liberate cells from the surface of the cervix and then stored on ice; 25 μl of the subsequent cell suspension was mounted on a glass slide and examined under light microscopy (Â10 to Â20 magnification). Previously described changes in cervical cell histology allowed for the determination of oestrus cycle stage (Cora et al., 2015) as either (in order of the 4-5 day cycle) pro-oestrus, oestrus, metoestrous or dioestrus. Cycle stage determination was performed post-experiment during functional investigation as a means of blinding; this was not possible during molecular techniques.

| Immunocytochemistry
The Immuno-related procedures used comply with the recommendations made by the British Journal of Pharmacology . Vascular smooth muscle cells were isolated from morpholino-transfected mesenteric arteries via incubation in isolation PSS of the following composition (mmolÁL À1 ): 120 NaCl, 6 KCl, 12 glucose, 10 HEPEs and 1.2 MgCl 2 supplemented with 1.75 mgÁml À1 Collagenase Type IA, 0.9 mgÁml À1 protease, 1 mgÁml À1 trypsin inhibitor and 1 mgÁml À1 bovine serum albumin (Sigma, UK) at 37 C for 17 min. Vessels then underwent mechanical trituration by wide bore glass pipette to liberate vascular smooth muscle cells. The subsequent cell suspension was plated onto 13-mm coverslips in a 24-well plate, supplemented with an equal volume of Ca 2+ George's University, London), and total cell fluorescence was analysed using ImageJ (RRID:SCR_003070) software.

| Cell culture
The K v 7.1 antibody was validated using Chinese Hamster Ovarian

| Materials
All drugs for the following investigation were procured from Tocris Bioscience (Oxford, UK). Excluding CAY-10441 and MRE-269, which were acquired from Cayman chemical (Michigan, USA), Rp-8-Br-cAMP which was aquired from Sigma-Aldrich (UK) and M119K which was provided by the National Cancer Institute Drug Development Programme. All drugs were dissolved in DMSO and final vehicle concentrations were always ≤0.01. For materials regarding morpholino transfection, immunodetection or RT-qPCR, see relevant sections above.

| Data and statistical analysis
All values are expressed as mean ± standard error of the mean (SEM) for no less than five independent data points, excluding measurement of total cell fluorescence during immunocytochemistry, in which 10 cells were measured per cell. For isometric tension recordings, single dose responses to iloprost are expressed as (%) change from stable tone in response to 300 nmolÁL À1 U46619, contractions from basal tone are expressed as (%) contraction when normalized to vasoconstriction to 10 μmolÁL À1 methoxamine and all cumulative concentration effect curves are expressed as (%) stable contraction in response to 300 nmolÁL À1 U46619. This is to account for changes in vessel contractility. For functional experiments involving cumulative concentration effect curves, a transformed data set was generated using; X = Log(X), to reduce representative skew. Following which, either a four parametric linear regression analysis was performed using either (Log(Agonist) vs. response variable slope [four parameters Bottom/ Hillslope/top/EC 50 ]) using GraphPad Prism (RRID:SCR_002798, Version 9.0.0) to fit a cumulative concentration effect curve to the figure.
In the presence of the EP 3 receptor antagonist L-798,106, 1 μmolÁL À1 iloprost-mediated contraction was converted to a relaxation ( Figure 2a). In mesenteric arteries from females, EP 3 receptor blockade had no effect (Figure 2a)  (Table 1) in the mesenteric arteries from both sexes. Figure 2.F shows that expression of Ptger2/4 F I G U R E 2 Iloprost-mediated contraction and relaxation is blocked by EP 3 and IP receptor-mediated inhibition, respectively. Mean data for iloprost-mediated vasoactive responses (a, 1 μmolÁL À1 ; c, 300 nmolÁL À1 ) within precontracted (300 nmolÁL À1 U46619) mesenteric arteries from male and female rats preincubated in either solvent control (DMSO; a,c; n = 6-8), 300 nmolÁL À1 L-798,106 (a; n = 6-8) or 100 nmolÁL À1 CAY-10441 (c; n = 6-8). Scatter graph showing individual and mean values of stable precontracted tone from male and female arteries prior to adding iloprost to the chamber (b,d). Mean data for vasoconstriction from base line tension in response to 3 μmolÁL À1 iloprost in male (black; n = 7) and female (red; n = 10) mesenteric arteries in the presence of DMSO or 100 nmolÁL À1 CAY-10441 normalized to peak contraction in response to 10 μmolÁL À1 methoxamine (e). Relative gene expression of prostanoid receptors (Ptger1-4 = EP 1-4 , Ptgi = IP, Tbxar2 = TXA 2 ) normalized to stable housekeeper genes (Canx, Cyc1) expressed as 2 ÀΔCq from male (black; n = 5) and female (red; n = 6-10) whole mesenteric artery lysates (f). A two-way statistical ANOVA with a post-hoc Bonferroni test was used to generate significant values ( * P < 0.05). n = number of animals used (EP 2/4 ) was negligible in mesenteric arteries from both sexes compared to Ptger3 > Ptgir > Ptger1 (EP 3 ; IP; EP 1 ), which were well expressed ( Figure 2f). However, Ptger3 was expressed at significantly lower level in mesenteric arteries from female rats compared to mesenteric arteries from males ( Figure 2f). Thus, our data demonstrate that iloprost-mediated relaxation in rat mesenteric arteries occurs predominantly via IP receptors, while contraction was caused by activation of EP 3 receptors. Additionally, the absence of a biphasic response to iloprost in female mesenteric arteries was associated with a comparably smaller effect of EP 3 receptor inhibition on iloprost-mediated relaxation and a reduction in Ptger3 expression.

| Characterizing MRE-269-mediated relaxation
As iloprost has a plethora of potential targets, we used the clinically available IP receptor agonist, selexipag (NS-304), to delineate the mechanisms underlying IP-mediated relaxation. Application of selexipag produced concentration-dependent relaxations of precontracted mesenteric arteries from male rats ( Figure S3). However, this effect was insensitive to preincubation with IP receptor antagonist CAY-10441 ( Figure S3) thus, selexipag effects are non-IP F I G U R E 3 MRE-269 relaxation is inhibited by CAY-10441 to a threshold of 1 μmolÁL À1 in male mesenteric arteries. Mean data of MRE-269 (0.01-10 μmolÁL À1 ) mediated relaxation of precontracted arterial tone (300 nmolÁL À1 ) in vessels preincubated in DMSO solvent control (black; n = 7) or 100 nmolÁL À1 CAY-10441 (n = 5) in male mesenteric arteries. Grey box demonstrates non-CAY-10441 sensitive MRE-269-mediated relaxation. All values are expressed as mean ± SEM. A two-way statistical ANOVA with a post-hoc Bonferroni test was used to generate significant values ( * P < 0.05). n = number of animals used F I G U R E 4 Linopirdine and HMR-1556 attenuate MRE-269-mediated vasorelaxation in mesenteric arteries from male and female rats. Representative traces of MRE-269-mediated (0.01-1 μmolÁL À1 ) relaxation of precontracted tone (300 nmolÁL À1 U46619) within mesenteric arteries preincubated in either DMSO solvent control (a; black) or 10 μmolÁL À1 K V 7.1 specific blocker HMR-1556 (a; green) from male Wistar rats. Mean data for MRE-269-mediated vasorelaxation (0.01-1 μmolÁL À1 ) of precontracted tone (300 nmolÁL À1 U46619) within mesenteric arteries preincubated in either DMSO (b; black) solvent control, 10 μmolÁL À1 pan-K V 7 channel blocker linopirdine (b; yellow) or HMR-1556 (b; green) or 10 μmolÁL À1 K V 7.1 seelective blocker Chromanol 293B (c; green; n = 7-10). Mean data for isoprenaline-mediated relaxation in vessels preincubated in DMSO (black) or 10 μmolÁL À1 HMR-1556 (green) in male mesenteric arteries (d; n = 9). All values are expressed as mean ± SEM (a-f). A two-way statistical ANOVA with a post-hoc Dunnet (b) or Bonferroni (c,d) test was used to generate significant values ( * P < 0.05). n = number of animals used receptor dependent. It is now known that in the body, selexipag is metabolized into the active compound, inhibited by CAY-10441 preincubation up to threshold of 1 μmolÁL À1 . At higher concentrations, the MRE-269-mediated relaxation was not sensitive to CAY-10441 and therefore does not involve IP receptor activation. This non-IP receptor-mediated relaxation is highlighted by the grey box in Figure 3 and in the following investigations, MRE-269 was used at concentrations ≤1 μmolÁL À1 to ensure only IP receptor-mediated effects were investigated.
BK Ca and K ATP channels have also been identified as downstream targets of cAMP-PKA-dependent relaxations evoked by iloprost (Schubert et al., 1997). Here, we demonstrate that BK Ca inhibitor iberiotoxin (100 nmolÁL À1 ) but not K ATP inhibitor glibenclamide (1 μmolÁL À1 ) partially inhibited MRE-269-mediated relaxation in male mesenteric arteries ( Figure S4A,B), though this failed to reach statistical significance.

| DISCUSSION
To our knowledge, the present study is the first to highlight sex as a factor in the arterial response to prostacyclin mimetics. The study shows that application of iloprost to pre-contacted mesenteric arteries from male rats produced bimodal responses, relaxation at low concentrations, followed by contraction at higher concentra-

| Iloprost-evoked vasoconstriction
While principally regarded as a vasodilator, PGI 2 mediates both relaxation and contraction of smooth muscle (Dusting et al., 1977;Liu et al., 2017;Moncada et al., 1976). PGI 2 has subsequently been identified as an endothelial-derived contracting factor produced in response to acetylcholine within rat aorta (Gluais et al., 2005), mesenteric (Liu et al., 2017), iliac (Zhang et al., 2021) and renal arteries (Zhang et al., 2021) in a process attributed to the activation of both EP and TP prostanoid receptors. Consistent with Liu et al. (2017), we show that high concentrations of iloprost-evoked contractions were blocked by the EP 3 receptor antagonist L-798,106. As iloprost has a low affinity for TP receptors (Whittle et al., 2012), TP receptor knockout has no effect on PGI 2 -mediated contraction in mesenteric arteries (Liu et al., 2017) and all vessels in this study were precontracted with U46619, a TP receptor agonist. TP receptors were not considered for the scope of this investigation. Additionally, EP 1 receptor agonists do not elicit contractions in male mesenteric arteries (Kobayashi et al., 2011), and in agreement with previous findings (Kobayashi et al., 2011), a reduced expression of Ptger1 was observed when compared to Ptger3. Furthermore, iloprost had negligible contractile effect in mesenteric arteries from female rats, which was associated with a lower expression level of Ptger3 in these arteries.  Zhu, et al., 2012;Jepps et al., 2011;Yeung et al., 2007). K V 7.4/K V 7.5 alone however are implicated in the regulation of the resting membrane potential (Mackie et al., 2008) and basal tone (Mackie et al., 2008;Ng et al., 2011). In addition, pharmacological inhibition or molecular knockdown of K V 7.4/7.5 impairs relaxations to many different relaxants including isoprenaline, calcitonin gene-related peptide (CGRP), adenosine (G s linked), atrial natriuretic peptide (ANP;
Surprisingly, MRE-269-evoked CAY-10441-sensitive relaxations in mesenteric arteries from male rats were also potently inhibited by two structurally dissimilar K V 7.1 selective inhibitors (HMR-1556, chromanol 293B) and molecular knockdown of the channel. In contrast to K V 7.4/K V 7.5, the role of K V 7.1 within the vasculature remains enigmatic. Though K V 7.1 is expressed within vascular smooth muscle cells Chadha, Zunke, Zhu, et al., 2012;Tsvetkov et al., 2017) and K V 7.1 selective activators RL-1 and ML277 are effective relaxants of precontracted arterial tone , K V 7.1 has not been identified as the downstream target of any endogenous vasoactive signalling cascades (Chadha et al., 2014;Stott et al., 2016;. Yet in the present study, HMR-1556 produced as full an inhibition as linopirdine, which suggests K V 7.1, and not K V 7.4/7.5, contributes to MRE-269-mediated relaxations. Under the same conditions, the mixed β-adrenoceptor agonist isoprenaline produced relaxations that were not HMR1556 sensitive. Thus, our findings appear not to be an off-target effect of HMR-1556. Moreover, a role for K V 7.1 in MRE-269-mediated relaxation was substantiated by morpholino-induced reduction in K V 7.1 protein levels. While further work is required to validate these findings, to our knowledge, the first to describe an effect on vascular reactivity by K V 7.1 inhibition is our data. Additionally, our data support the notion that IP receptormediated responses are cAMP-PKA mediated. In agreement with previous work done by our lab (Stott et al., 2016Stott, Povstyan, et al., 2015), we demonstrate that relaxations that are mediated PKA, but not EPAC, are also sensitive to G βγ inhibition. The identification of G βγ contribution to IP receptor-mediated relaxation adds new complexity to the vascular response and gives credence to the novel role of G βγ in the functional relationship between GPCRs and K V 7s.
To date, comparably little is known as to how K V 7 channels operate within the female. However, K V 7 has been shown to regulate both human and murine myometrium (McCallum et al., 2009) and human chorionic plate artery (Mills et al., 2015) contractility. Intriguingly, while MRE-269-mediated relaxation was attenuated by linopirdine and HMR-1556 in arteries from female rats, the effect of the latter was far smaller than in the male. When separated into oestrus cycle stages, arteries from females in dioestrus/metestrus expressed sensitivities to HMR-1556 and linopirdine equivalent to the male, whereas arteries from females in proestrus/oestrus were entirely insensitive to either.
However, K V 7.1 activator-ML277-mediated relaxation was insensitive to changes in the oestrus cycle. As the functional output of pharmacological activation of the channel remains the same, the data indicate an oestrus cycle-dependent impairment of K V 7.1 channel coupling to IP receptor-mediated relaxation. Oestrus cycle-dependent regulation of vascular reactivity is a known but incompletely understood phenomenon (Jaimes et al., 2019) largely attributed to endogenous sexhormones, primarily, 17-β oestradiol. 17-β oestradiol negatively regulates K V 7.1 in distal colic crypt cells and cardiac myocytes (Alzamora et al., 2011;O'Mahony et al., 2007;Rapetti-Mauss et al., 2013;Waldegger et al., 1996). As previous work demonstrates that within the Wistar rat, oestradiol peaks within pro-oestrus rats followed by comparably little to none in oestrus, metestrus and dioestrus (Nilsson et al., 2015), we propose that during pro-oestrus, oestradiol levels rise, impairing K V 7.1 coupling to IP receptor during proestrus/oestrus phase, thus reducing the potency of MRE-269-mediated relaxation and its HMR-1556/linopirdine sensitivity, which does not recover until dioestrus/metestrus. Our data imply an oestrus cycle.

| Perspectives
Sexual dimorphisms in cardiovascular physiology and pathophysiology are known (Pabbidi et al., 2018), whereby women express a cardioprotective factor or factors that differentiates the aetiology of vascular disease between age-matched men and women. Chronic