1. Cell Biology
Download icon

Membrane estrogen receptor alpha (ERα) participates in flow-mediated dilation in a ligand-independent manner

  1. Julie Favre
  2. Emilie Vessieres
  3. Anne-Laure Guihot
  4. Coralyne Proux
  5. Linda Grimaud
  6. Jordan Rivron
  7. Manuela CL Garcia
  8. Léa Réthoré
  9. Rana Zahreddine
  10. Morgane Davezac
  11. Chanaelle Fébrissy
  12. Marine Adlanmerini
  13. Laurent Loufrani
  14. Vincent Procaccio
  15. Jean-Michel Foidart
  16. Gilles Flouriot
  17. Françoise Lenfant
  18. Coralie Fontaine
  19. Jean-François Arnal
  20. Daniel Henrion  Is a corresponding author
  1. Angers University, MITOVASC, CNRS UMR 6015, INSERM U1083, France
  2. CARFI facility, Angers University, France
  3. INSERM U1297, Paul Sabatier University (Toulouse III) , University Hospital (UHC) of Toulouse, France
  4. University Hospital (CHU) of Angers, France
  5. Groupe Interdisciplinaire de Génoprotéomique Appliquée, Université de Liège, Belgium
  6. INSERM U1085, IRSET (Institut de Recherche en Santé, Environnement et Travail), University of Rennes, France
Research Article
  • Cited 0
  • Views 345
  • Annotations
Cite this article as: eLife 2021;10:e68695 doi: 10.7554/eLife.68695

Abstract

Estrogen receptor alpha (ERα) activation by estrogens prevents atheroma through its nuclear action, whereas plasma membrane-located ERα accelerates endothelial healing. The genetic deficiency of ERα was associated with a reduction in flow-mediated dilation (FMD) in one man. Here, we evaluated ex vivo the role of ERα on FMD of resistance arteries. FMD, but not agonist (acetylcholine, insulin)-mediated dilation, was reduced in male and female mice lacking ERα (Esr1-/- mice) compared to wild-type mice and was not dependent on the presence of estrogens. In C451A-ERα mice lacking membrane ERα, not in mice lacking AF2-dependent nuclear ERα actions, FMD was reduced, and restored by antioxidant treatments. Compared to wild-type mice, isolated perfused kidneys of C451A-ERα mice revealed a decreased flow-mediated nitrate production and an increased H2O2 production. Thus, endothelial membrane ERα promotes NO bioavailability through inhibition of oxidative stress and thereby participates in FMD in a ligand-independent manner.

Editor's evaluation

Using multiple genetically modified mouse models, the authors have demonstrated a novel role of membrane associated estrogen receptor alpha (ERα) signaling to modulate flow-mediated dilation (FMD) in a ligand-independent manner. Specifically, the results indicate that non-nuclear actions of membrane estrogen receptor α in endothelial cells support flow-mediated vasodilatation in animals of both sexes via mechanisms that are independent of estrogenic ligands, involving NO production and an attenuation of the NO-inactivating effects of reactive oxygen species. These findings highlight a novel role of ligand-independent activation of membrane estrogen receptor α in regulation of vascular physiology and possibly in disease, adding to the recently introduced paradigm shift in the understanding of estrogen and estrogen receptor function.

https://doi.org/10.7554/eLife.68695.sa0

Introduction

Resistance arteries are the small blood vessels located upstream of capillaries. Alteration of their structures or functions can raise capillary pressure, which exacerbates organ damage due to cardio- and cerebro-vascular risk factors and associated organ disorders. The basal tone of resistance arteries allows for tight control of local blood flow. This tone results from the interaction between pressure-induced smooth muscle contraction and flow-mediated dilation (FMD) due to the activation of endothelial cells by shear stress. FMD measured in the human forearm depends mainly on the acute production of NO by endothelial cells in response to an acute increase in shear stress (Joannides et al., 1995; Green et al., 2014; Zhou et al., 2014) and reduced FMD is a hallmark of endothelium dysfunction (Zhou et al., 2014; Rizzoni and Agabiti Rosei, 2006; Stoner and Sabatier, 2012).

Epidemiological investigations have shown that, prior to menopause, women are less affected by cardiovascular disorders than men (Simoncini, 2009; Arnal et al., 2017). Estrogens protect against atherosclerosis (Billon-Galés et al., 2009) and neointimal proliferation (Smirnova et al., 2015), and accelerate re-endothelialization of injured arteries (Brouchet et al., 2001). Numerous actions of 17-beta-estradiol (E2) are mediated by estrogen receptor alpha (ERα), which acts in the nucleus as a transcription factor. E2 is strongly involved in the outward remodelling of the uterine blood vessels during pregnancy (Mandala and Osol, 2012). Indeed, we have previously shown that E2 and ERα, and more precisely its nuclear activating function AF2, are both essential for the arterial outward remodeling induced by a chronic rise in blood flow in vivo (Tarhouni et al., 2013; Tarhouni et al., 2014a; Tarhouni et al., 2014b).

However, a subpopulation of ERα is also associated with the plasma membrane and activates non-nuclear signaling (Arnal et al., 2017; Banerjee et al., 2014; Lu et al., 2017). The acute effect initially described in 1967 was a rapid increase in AMPc production in the rat uterus in response to E2 (Szego and Davis, 1967). E2 binding to the plasma membrane was subsequently reported in endometrial cells and hepatocytes (Pietras and Szego, 1977), suggesting that a fraction of ERα could be located to the membrane and contributes to the rapid effects of E2, possibly through the rapid activation of G proteins and kinases such as ERK1-2, PI3K, or P21ras (Arnal et al., 2017). In ovine fetal pulmonary artery endothelial cells, E2 stimulates eNOS activity through activation of ERα leading to increased intracellular Ca2+ within minutes (Lantin-Hermoso et al., 1997). By contrast, in HUVECs, E2 induces a rapid production of NO and cGMP independent of an increase in intracellular Ca2+ (Caulin-Glaser et al., 1997). This rapid nongenomic activation of eNOS involves Akt/PKB (Florian et al., 2004) and MAP kinase-dependent mechanisms (Chen et al., 1999). Estradiol-induced endothelium-independent dilation was also described in canine coronary arteries (Sudhir et al., 1995) and in rat cerebral microvessels (Florian et al., 2004). This dilation is also mediated by ERα located at the level of the plasma membrane. Using a mouse model lacking membrane-associated ERα, we demonstrated that the acute vasodilator effect of E2 and its accelerative effect on re-endothelialization are mediated by membrane-associated ERα (Adlanmerini et al., 2014; Zahreddine et al., 2021). On the other hand, E2 exerts protective effects against atheroma, angiotensin 2-induced hypertension, and neointimal hyperplasia through its nuclear effects (Guivarc’h et al., 2018).

The 7-transmembrane G-protein-coupled estrogen receptor (GPER, formerly known as GPR30) is another receptor located not only at the plasma membrane but also on the membrane of the endoplasmic reticulum that can be activated by E2. It was found in both human and animal arteries (Prossnitz and Barton, 2011; Barton et al., 2018). The combination of GPER-selective agonists and antagonists with the use of GPER-knock-out mice allowed to elucidate more specifically its biological effects arteries (Prossnitz and Barton, 2011; Barton et al., 2018). In the rat, GPER activation by its agonist G-1 reduces uterine vascular tone during pregnancy through activation of endothelium-dependent NO production (Tropea et al., 2015). Likewise, the G-1-induced relaxation of the mesenteric resistance arteries in both male and female rats is mainly mediated by the PI3K-Akt-eNOS pathway (Peixoto et al., 2017). Noteworthy, GPER partly contributes to E2-dependent vasodilation of mouse aortae (Fredette et al., 2018). Thus, both ERα and GPER could contribute to the rapid actions of E2, although their respective roles according to vessel type, species and pathophysiological context remain to be established.

The risk of cardiovascular diseases differs between men and women, and the protection of women is progressively lost after menopause. For instance, endothelium-dependent dilation of subcutaneous arteries is reduced in post-menopausal women compared to pre-menopausal women (Kublickiene et al., 2005; Kublickiene et al., 2008). This protection involves NO production in response to estrogens. Similarly, diet phytoestrogens could have protective actions in postmenopausal women suffering coronary artery disease (Cruz et al., 2008) and selective estrogen receptor modulators (SERMs) such as raloxifene exert protective actions in female rats through eNOS activation (Chan et al., 2010). Besides the activation of eNOS, hormonal replacement therapy also activates endothelium-derived hyperpolarizing factor (EDHF)-mediated vasodilation as shown in rat mesenteric and uterine arteries (Burger et al., 2009) as well as in the rat gracilis muscle artery with increased Epoxyeicosatrienoic acids (EETs) production involved in E2-mediated increase in FMD in hypertensive or old rats (Huang et al., 2001; Sun et al., 2004). Furthermore, estrogen therapy reduces pressure (myogenic) (Kublickiene et al., 2005; Kublickiene et al., 2008) and adrenergic-dependent contraction (Meyer et al., 1997). FMD is also improved by E2 in rat gracilis muscle arteries (Huang et al., 1998). Although there is an increase in the amplitude of FMD in women among the menstrual cycle with a greater dilation during the luteal or follicular phase (Hashimoto et al., 1995), FMD is similar in healthy young men and women (Sullivan et al., 2015). Importantly, the first disruptive mutation in the gene encoding ERα, reported in 1994 in a man who was only 30 years old (Smith et al., 1994), was found to be associated with a total absence of FMD (Sudhir et al., 1997). This single yet major clinical observation suggests that ERα-dependent signal transduction could play a role in FMD in males. Of note, conversion of testosterone into estradiol by aromatase which is expressed in the arterial wall has been shown to reduce early atherogenesis in male mice (Nathan et al., 2001).

In the present study, we investigated the role of ERα and its different subfunctions on FMD in isolated mouse resistance arteries. To this aim, we used different mouse models that were: (i) fully deficient in ERα (Esr1-/- mice), (ii) deleted in seven amino acid in the helix 12 and thus deficient in activation function (AF)–2 necessary for the nuclear transcriptional activity of ERα (AF20ERα mice), and (iii) invalidated for plasma membrane-associated signaling. To explore the role of membrane ERα, we used: (1) mice that carry a mutation of the codon encoding the cysteine (Cys) 451 palmitoylation site of ERα to alanine (C451A-ERα mice) so that the anchoring of the receptor to the plasma membrane is impossible (Adlanmerini et al., 2014) and (2) a knock-in mouse model of ERα mutated for the arginine 264 (R264A-ERα mice) so that the interaction of the membrane-located ERα with Gαi involved in rapid NO production is suppressed (Adlanmerini et al., 2020). Moreover, because the absence of membrane-associated ERα strongly reduced FMD, additional experiments were conducted to investigate: (i) the involved mechanisms, (ii) its counterpart in female mice, and (iii) the potential role of ERα activation by its ligands.

Results

FMD is reduced in mice lacking ERα but unaffected by exogenous estrogens

In mice lacking ERα (see the scheme in Figure 1A), FMD was significantly reduced in resistance arteries isolated from male Esr1-/- mice compared to littermate Esr1+/+ mice (Figure 1B). Precontraction prior to FMD (Figure 1C) and arterial diameter (Figure 1D) were not significantly affected by the absence of ERα. Agonist-mediated endothelium-dependent (acetylcholine and insulin) and endothelium-independent (SNP) dilation were not significantly affected by the absence of ERα (Figure 1E–G).

Involvement of ERα in flow-mediated dilation (FMD).

FMD was measured in mesenteric resistance arteries isolated from male mice lacking ERα (Esr1-/-) and male wild-type littermates (Esr1+/+) (A). (B) FMD was determined in response to stepwise increases in luminal flow in male Esr1-/- and Esr1+/+ mice. (C) Precontraction with phenylephrine (Phe) before measurement of FMD. (D) Basal diameter of the arteries used for FMD measurment. Besides FMD, acetylcholine- (E), insulin- (F), and sodium nitroprusside- (SNP, G) mediated dilation was measured in mesenteric resistance arteries isolated from male Esr1-/- and Esr1+/+ mice. FMD was also measured in wild-type (WT) mice in the presence (20 min incubation) or absence of 17-β-estradiol (E2, 0.01 µmol/L, H), estetrol (E4, 1 µmol/L, H), ICI 182 780 (1 µmol/L, H) and the GPER antagonist G-36 (10 µM, I). (I) G-1 (10 µM)- and E2 (0.01 µM)-mediated dilation in the presence or absence of G-36 (1 µM). FMD was then measured in mesenteric arteries isolated from intact (K) and ovariectomized (OVX, L) female Esr1-/- and Esr1+/+ mice as well as in and uterine arteries from Esr1-/- and Esr1+/+ mice (M). Flow rate rate was 3, 6, 9, 12, 15, 30, and 50 µl/min corresponding to 0.8, 1.2, 2, 2.8, 4, 8, and 12 dyn/cm2. Means ± the SEM are shown (n = 7–18 mice per group). Two-way ANOVA for repeated measurements: p = 0.0072 (interaction: p < 0.0001, B), p = 0.0087 (interaction: p < 0.0001, K), p = 0.0030 (interaction: p < 0.0001, L), p = 0.0119 (interaction: 0.0107, M). NS: two-way ANOVA for repeated measurements, panel E to I. NS: Two-tailed Mann-Whitney test, panels C and D. See source data in Figure 1—source data 1.

Figure 1—source data 1

Data and statistical analysis from experiments plotted in Figure 1B—M.

https://cdn.elifesciences.org/articles/68695/elife-68695-fig1-data1-v2.xlsx

To directly investigate the influence of estrogens on FMD, mesenteric resistance arteries isolated from male WT mice were incubated (20 min) with E2, which activates both membrane-associated and nuclear ERα, or with another natural estrogen, estetrol (E4), which activates only nuclear ERα (Abot et al., 2014). These exogenous estrogens did not affect FMD (Figure 1H). Furthermore, the estrogen receptor downregulator and GPER agonist fulvestrant (ICI-182780) (Meyer et al., 2010; Jacenik et al., 2016) did not alter FMD after 20 min of incubation with isolated arteries (Figure 1H). Similarly, the GPER antagonist G-36 did not alter FMD (Figure 1I) although G-36 inhibited the dilation induced by both E2 and the GPER agonist G-1 (Figure 1J).

We also found that FMD in mesenteric resistance arteries was similarly reduced in both intact (Figure 1K) and ovariectomized female Esr1-/- mice (Figure 1L) compared to their respective Esr1+/+ littermate controls. The similar levels of FMD in intact and ovariectomized female WT mice, as well as in male mice, suggest that circulating endogenous estrogens do not influence FMD in young healthy mice. FMD was also reduced in the uterine artery isolated from female Esr1-/- mice in comparison with Esr1+/+ mice (Figure 1M).

FMD was not altered by the inactivation of Esr2, encoding ERβ in mice (Figure 2A). Similarly, arterial precontraction, basal diameter and acetylcholine-mediated dilation were not significantly affected by the absence of ERß (Figure 2B–D).

Involvement of ERβ and endothelial ERα in flow-mediated dilation (FMD).

(A to D) FMD, precontraction, basal diameter and acetylcholine-mediated dilation measured in male mice lacking ERβ (Esr2-/-) and their littermate control (Esr2+/+). (E to H) FMD, precontraction, basal diameter and acetylcholine-mediated dilation measured in TekCre/+:Esr1-/- male mice lacking endothelial ERα (EC-ERα) and TekCre/-:Esr1lox/lox their littermate controls (WT). Flow rate rate was 3, 6, 9, 12, 15, 30, and 50 µl/min corresponding to 0.8, 1.2, 2, 2.8, 4, 8, and 12 dyn/cm2.+ source data 2. Means ± the SEM are shown (n = 6 or 7 mice per group). Two-way ANOVA for repeated measurements: p = 0.0273 (interaction: p = 0.0069**, E). NS: two-way ANOVA for repeated measurements, panel A, D, and H. NS: two-tailed Mann-Whitney test, B, C, F, and G. Data and analysis in Figure 2—source data 1.

Figure 2—source data 1

Data and statistical analysis from experiments plotted in Figure 2A—H.

https://cdn.elifesciences.org/articles/68695/elife-68695-fig2-data1-v2.xlsx

As FMD depends on the response of the endothelium to shear stress, we next investigated FMD in mice lacking ERα in endothelial cells (TekCre/+: Esr1f/f mice). FMD was reduced in arteries isolated from these mice compared to littermate WT mice (Figure 2E). Arterial precontraction, basal diameter and acetylcholine-mediated dilation were not significantly affected by the absence of endothelial ERα (Figure 2F–H).

Altogether, these results demonstrate a crucial role of ERα in FMD in both males and females and probably in a ligand independent manner. We, therefore, decided to use male mice for the remainder of the study.

FMD in mice lacking the nuclear activation function 2 (AF2) of ERα

As the AF2 nuclear function mediates several protective effects of ERα on the vasculature (Guivarc’h et al., 2018), we first investigated FMD in AF20ERα mice (Figure 3A). The gene expression level of Esr1 (encoding ERα) in endothelial cells was not affected by invalidation of the AF2 function of ERα (Figure 3B). Quality of mRNA endothelial enrichissment was attested using analysis of Tek expression as a marker of endothelial cells and Cnn1 expression as a marker of smooth muscle cells (Figure 3—figure supplement 1).

Figure 3 with 1 supplement see all
Flow-mediated dilation in mice lacking nuclear or membrane-associated ERα.

Esr1 expression level in aortic endothelial cells (expression relative to the housekeeping genes Gapdh, Hprt and Gusb), flow-mediated dilation (FMD) acetylcholine-mediated dilation were measured in mesenteric resistance arteries isolated from AF2-WT and AF20ERα male mice (A to D), C451A-WT and C451A-ERα male mice (E to H) and R264A-WT and R264A-ERα male mice (I to L). Means ± the SEM is shown (n = 13 AF20ERα, n = 5 AF2-WT mice, n = 8 C451A-ERα, n = 6 C451A-WT mice, n = 9 R264A-WT and n = 10 R264A-ERα mice). Flow rate rate was 3, 6, 9, 12, 15, 30, and 50 µl/min corresponding to 0.8, 1.2, 2, 2.8, 4, 8, and 12 dyn/cm2. Two-way ANOVA for repeated measurements: panel C, p = 0.2681 (interaction: p = 07302), panel G, p = 0.0114 (interaction: p = 0.002), panel K, p = 0.0015 (interaction: p = 0.0002). Panels D, H, and L: NS. NS, two-tailed Mann-Whitney test (panels B, F and J).

We observed that FMD was not significantly reduced in mesenteric resistance arteries isolated from AF20ERα male mice compared to littermate AF20WT animals (Figure 3C). Acetylcholine-mediated dilation (Figure 3D) was not altered by loss of the nuclear AF2 function of ERα. Thus, we demonstrated that FMD was preserved despite the loss of AF2 nuclear function of ERα.

FMD in mice lacking membrane-located ERα effects

We thus evaluated the role of the membrane ERα in FMD, thanks to two complementary models, C451A-ERα mice[24] and R264A-ERα mice (Adlanmerini et al., 2020), that allowed us to previously investigate the ERα membrane-initiated steroid signaling (MISS) pathway in the accelerative effect of E2 on re-endothelialization following arterial injury and in the acute dilation induced by E2 through rapid eNOS activation.

First, we investigated FMD in C451A-ERα male mice that lack the capacity to anchor ERα to the plasma membrane through palmitoylation (Adlanmerini et al., 2014; Figure 3E). The gene expression level of Esr1 was not significantly altered in aortic endothelial cells isolated from C451A-ERα mice compared to wild-type mice (Figure 3F). FMD was significantly reduced compared to WT littermate animals (Figure 3G) while acetylcholine-mediated dilation was not significantly affected by the absence of membrane-associated ERα (Figure 3H).

We then investigated FMD in R264A-ERα male mice that lack the capacity to activate the Gαi involved in acute NO production upon membrane-ERα activation (Figure 3I; Adlanmerini et al., 2020). The gene expression level of Esr1 was not significantly altered in aortic endothelial cells isolated from R264A-ERα mice compared to wild-type mice (Figure 3J). We found that FMD was significantly reduced in R264A-ERα male mice compared to WT littermate controls (Figure 3K). On the other hand, acetylcholine-mediated dilation was not significantly affected by the absence of membrane-associated ERα effects in R264A-ERα male mice (Figure 3L).

Thus, we demonstrated that FMD was altered as a consequence of either the inactivation of palmitoylation site of ERα (C451A-ERα mice) or the impairment of the activation of the Gαi protein by ERα (R264A-ERα mice), thus preventing membrane-associated ERα activation of FMD in male mice.

Role of ERα in the activation of the NO pathway in FMD

As NO produced by endothelial NOS plays a key role in endothelium-dependent dilation and thus in FMD, we investigated the effect of the inhibition of NO synthesis by L-NNA on FMD and acetylcholine-mediated dilation. FMD was significantly reduced by L-NNA in mesenteric resistance arteries of the four groups of littermate WT mice (Figure 4A–D), whereas L-NNA had no significant effect on FMD in Esr1-/- (Figure 4A), C451A-ERα (Figure 4B), and R264A-ERα mice (Figure 4C). In contrast, FMD was reduced to a similar extent by L-NNA in AF20ERα and WT mice (Figure 4D).

Figure 4 with 1 supplement see all
with one supplement: Effect of the blockade of NO synthesis and cyloxygenase on flow-mediated dilation.

Flow-mediated dilation (FMD) was determined in pressurized mesenteric resistance arteries isolated from male Esr1+/+ and Esr1-/- (A), C451A-WT and C451A-ERα (B), R264A-WT and R264A-ERα (C), AF20WT and AF20ERα mice (D), before and after addition of the NO synthesis blocker L-NNA (100 µM, 30 min) and then of the combination of L-NNA plus indomethacin (indo, 10 µM, 30 min). Acetylcholine-mediated relaxation was measured in the same groups in the presence and in the absence of L-NNA and of L-NNA plus indomethacin (E to H). Flow rate rate was 3, 6, 9, 12, 15, 30, and 50 µl/min corresponding to 0.8, 1.2, 2, 2.8, 4, 8, and 12 dyn/cm2. Means ± the SEM are shown (n = 6–8 per group). ***p < 0.001, two-way ANOVA for repeated measurements, L-NNA or L-NNA+ indo versus untreated arteries within each group. Data and analysis in Figure 4—source data 1.

Figure 4—source data 1

Data and statistical analysis from experiments plotted in Figure 4A–H.

https://cdn.elifesciences.org/articles/68695/elife-68695-fig4-data1-v2.xlsx

L-NNA strongly and similarly reduced acetylcholine-mediated relaxation in Esr1-/- (Figure 4E), C451A-ERα (Figure 4F), R264A-ERα (Figure 4G), AF20ERα (Figure 4H) and the corresponding littermate WT mice (Figure 4E–H), thus showing that the alteration of membrane ERα activation affected selectively the flow-mediated NO-dependent dilation, but not the acetylcholine-mediated NO-dependent dilation. The addition of either the inhibitor of cyclooxygenase indomethacin (Figure 4A–H) or the inhibitor epoxyeicosatrienoic acids (EETs) synthesis MSPPOH (Figure 4—figure supplement 1A–F) did not further reduced FMD or acetylcholine-mediated relaxation in all the groups.

Western-blot analysis of eNOS and phosphorylated eNOS was then performed on isolated resistance arteries that had been mounted in an arteriograph and then submitted or not to flow during 2 min (Figure 5). This flow rate is equivalent to the maximal response to flow observed in arteriography. In WT mice, the phosphorylation of eNOS at Ser1177 by flow (shear stress) was greater in arteries submitted to flow than in control (no flow) arteries as evidenced by a greater ratio of phosphorylated eNOS/total eNOS (Figure 5A). This ratio was not significantly greater in arteries submitted to flow than in unstimulated arteries in C451A-ERα mice (Figure 5A). The expression level of total eNOS was similar in C451A-ERα and in WT mice (Figure 5B). A similar pattern was observed in R264A-ERα mice (Figure 5D and E). By contrast, the ratio of phosphorylated eNOS/total eNOS was similarly increased by flow in WT and AF2-ERα mice without any change in total eNOS level between the two strains (Figure 5G and H).

eNOS and Akt phosphorylation in response to flow in perfused isolated mesenteric resistance arteries.

As illustrated on the scheme shown on the top of the figure, mesenteric resistance arteries were cannulated in vitro on glass micropipettes and perfused with physiological salt solution. Flow (50 µl/min or 12 dyn/cm2) was applied for 2 min before quick freezing of the artery. In control experiments no flow was applied. Western-blot analysis of eNOS, phospho (Ser1177)-eNOS (P-eNOS), Akt, phospho-Akt and β-actin in mesenteric arteries isolated from male C451A-ERα mice (C451A, A to C), R264A-ERα (R264A, D to F), AF20ERα (AF2, G to I) and their littermate control (WT) was then performed. The ratio of P-eNOS / eNOS is shown in A, D and G. The expression level of eNOS/β-actin in unstimulated arteries is shown in B, E and H. The ratio of P-Akt / Akt is shown in C, F and I. Means ± the SEM are shown (n = 6 C451A-WT, n = 9 C451A-ERα, n = 5 R264A-ERα, n = 5 R264A-WT, n = 6 AF20ERα and n = 4 AF2-WT mice). *p < 0.05 (panel C: p = 0.0374, panel D: p = 0.015, panel F: p = 0.0177, panel G: WT, p = 0.0234, AF2, p = 0.0465, panel I: p = 0.0152) and **p < 0.01 (panel A: p = 0.0045), two-tailed Mann-Whitney test. Data and analysis in Figure 5—source data 1.

Western-blot analysis of Akt and phosphorylated Akt was then performed on the same samples and a similar pattern was observed (Figure 5C,F,I). All the blots are shown in Figure 5—source data 2.

These results show that the absence of membrane-associated ERα affects flow-mediated eNOS activation pathway, at least in part, by preventing the activation of its upstream activator Akt/PKB.

Flow-mediated dilation and NO2/NO3 production in the isolated perfused kidney

As the kidney is a well-known autoregulated organ with a dense microvascular network, we investigated flow-mediated responsiveness in perfused kidneys isolated from C451A-ERα and WT mice (Figure 6A). First, the flow-pressure relationship was shifted leftward in C451A-ERα mice compared to WT mice (Figure 6B), suggesting reduced endothelial responsiveness to flow. Acetylcholine-mediated dilation in perfused kidneys was equivalent in C451A-ERα and WT mice (Figure 6C), suggesting that the response to flow was probably selectively reduced in perfused kidney of C451A-ERα mice, as shown above for the mesenteric artery. Similarly, phenylephrine-mediated contraction was not affected by the absence of membrane-ERα (51.1 ± 2.9 vs 57.4% ± 5.7% contraction, C451A-ERα and WT mice, n = 5 per group, p > 0.9999, Mann-Withney test).

Isolated and perfused kidney from C451A-ERα mice.

In the isolated and perfused kidney (A), the flow-pressure relationship was determined in C451A-ERα and WT mice (B). (C) Acetylcholine (1 µM)-mediated relaxation. The levels of nitrate-nitrite (D), ATP (E) and H2O2 (F) level were quantified in the perfusate collected from the kidney. Means ± the SEM are shown (n = 5 C451A-WT and 7 C451A -ERα mice). *p < 0.05, two-way ANOVA for repeated measurements (panel B, C451 vs WT: p = 0.0308 Interaction: p = 0.0008). Two-tailed Mann-Whitney tests (panels C to F: p > 0.999, p = 0.0317, p = 0.0079 and p = 0.0317, respectively). Data and analysis in Figure 6—source data 1.

Figure 6—source data 1

Data and statistical analysis from experiments plotted in Figure 6B–F.

https://cdn.elifesciences.org/articles/68695/elife-68695-fig6-data1-v2.xlsx

Then, we measured nitrate and nitrite concentration in the kidney perfusate and found that it was reduced in C451A-ERα compared to WT mice (Figure 6D). Interestingly, ATP production measured in the kidney perfusate was also reduced in C451A-ERα mice (Figure 6E) whereas H2O2 production was higher in C451A-ERα than in WT mice (Figure 6F).

Thus, these results suggest that FMD reduction due to the absence of membrane-ERα also affects the capacity of the renal vasculature to produce NO and ATP, whereas the increased H2O2 production suggests an excessive oxidative stress in response to flow in C451A-ERα mice.

Loss of membrane-associated ERα did not affect gene expression

Finally, we analyzed in mesenteric resistance arteries the expression of 44 genes that may be involved in the rapid endothelial response to acute changes in flow (FMD). No significant difference was observed between C451A-ERα and WT mice (Figure 7—figure supplements 1 and 2), in line with the prominent or even exclusive role of rapid, non genomic, membrane-ERα.

Acute pharmacological ROS reduction restored FMD in C451A-ERα mice

As the production of H2O2 in the kidneys from C451A-ERα mice was higher than in WT mice, we measured FMD in arteries from C451A-ERα and WT mice after pretreatment with various antioxidants. First, the addition of PEG-SOD plus catalase to the bath containing mesenteric resistance arteries isolated from WT mice did not alter FMD (Figure 7A). By contrast, PEG-SOD plus catalase enhanced FMD in arteries from C451A-ERα mice (Figure 7B).

Figure 7 with 3 supplements see all
Flow-mediated dilation and oxidative stress.

Flow-mediated dilation was determined in mesenteric resistance arteries isolated from male WT and C451A-ERα mice before and after addition of PEG-SOD and catalase (SOD-catalase, A and B), catalase (C and D) or Mito-Tempo (E and F). Flow rate was 3, 6, 9, 12, 15, 30, and 50 µl/min corresponding to 0.8, 1.2, 2, 2.8, 4, 8, and 12 dyn/cm2. Means ± the SEM are shown (n = 3–9 mice per group, see details in Figure 7—source data 1). *p < 0.05, two-way ANOVA for repeated measurements (panel A to F: p = 0.5887, p = 0.0321, p = 0.7170, p = 0.0311, p = 0.7641 and p0.0354, respectively).

Figure 7—source data 1

Data and statistical analysis from experiments plotted in Figure 7A–L.

https://cdn.elifesciences.org/articles/68695/elife-68695-fig7-data1-v2.xlsx

Catalase alone did not alter FMD in arteries from WT mice (Figure 7C), whereas it reduced FMD in arteries from C451A-ERα mice (Figure 7D).

Inhibition of mitochondrial ROS production by Mito-Tempo did not affect FMD in arteries from WT mice (Figure 7E), while it enhanced FMD in arteries isolated from C451A-ERα mice (Figure 7F).

The acute response to ATP and to the Piezo1 agonist YODA-1 were not affected by the absence of membrane ERα (C451A-ERα mice). In addition, the mechanosensitive channel blocker GsMTx4 similarly affected FMD in C451A-ERα and WT mice, suggesting that the defect in FMD is located downstream flow sensing (Figure 7—figure supplement 3).

This pharmacological approach suggests that reactive oxygen species could reduce FMD in arteries from C451A-ERα, while H2O2 could maintain in part the dilatory response induced by flow in the absence of membrane ERα.

In vivo pharmacological ROS reduction restored FMD inC451A-ERα mice

As acute antioxidant drugs restored FMD in mesenteric arteries isolated from C451A-ERα mice, we further explored the involvement of oxidative stress in the alteration of FMD by the use of two different antioxidant treatments in vivo.

After 2 weeks of treatment with the antioxidant TEMPOL, there was no longer a discernible difference between the response of WT and C451A-ERα mice to flow (Figure 8A). Mice body weight, arterial diameter, phenylephrine-, and KCl-mediated contraction as well as acetylcholine-mediated dilation were not different between TEMPOL-treated WT and C451A-ERα mice (Figure 8B–F).

FMD after antioxidant treatments in mice lacking membrane-ERα.

FMD was determined in mesenteric resistance arteries isolated from male WT and C451A-ERα mice treated for 2 weeks with the anti-oxidant TEMPOL (A to F) or with a combination of vitamin E and vitamin C for 4 weeks (G to L). At the end of the treatments arteries were collected and mounted in an arteriograph for the measurement of FMD (A and G), body weight (B and H), arterial diameter (C and I), phenylephrine (1 µM, D and J)- and KCl (80 mM, E and K)-mediated contraction and acetylcholine (1 µM)-mediated dilation (F and L). Flow rate was 3, 6, 9, 12, 15, 30, and 50 µl/min corresponding to 0.8, 1.2, 2, 2.8, 4, 8, and 12 dyn/cm2. Means ± the SEM are shown (n = 4 C451A-WT and 6 C451A-ERα mice treated with TEMPOL and n = 5 mice per group treated with vitamin E and vitamin C). NS, two-way ANOVA for repeated measurements (panel A: p = 0.6345 and G: p = 0.6482). NS, Two-tailed Mann-Whitney tests (panels B to F and H to L). Data and analysis in Figure 8—source data 1.

Figure 8—source data 1

Data and statistical analysis from experiments plotted in Figure 8A–L.

https://cdn.elifesciences.org/articles/68695/elife-68695-fig8-data1-v2.xlsx

A similar pattern was observed in mice treated for 4 weeks with vitamin E and vitamin C with no difference in FMD between WT and C451A-ERα mice (Figure 8G to L).

Thus, antioxidant treatment normalized FMD in C451A-ERα mice to the level of FMD in WT mice. Altogether, these data suggest that the absence of membrane-associated ERα increases oxidative stress, which in turn could be responsible for a large part of the alteration of NO-dependent FMD.

Discussion

We report here that endothelial membrane-associated ERα contributed to optimize flow (shear stress)-mediated dilation in young healthy mouse resistance arteries in a ligand-independent manner.

Previous experimental studies have reported the vascular benefit of estrogens on blood flow homeostasis, and E2 improves endothelium-dependent relaxation when it is reduced in diseased conditions. (Huang et al., 2001; Al-Khalili et al., 1998; Huang et al., 2000; Svedas et al., 2002; LeBlanc et al., 2009). Nevertheless, no difference in FMD has been observed between healthy men and women (Sullivan et al., 2015). In agreement, our results show that acute (20 min) incubation with exogenous E2, E4, or ICI-182780 did not affect FMD. Similarly, incubation with the GPER antagonist G-36 did not affect FMD, excluding ligand-activated GPER actions in FMD. In addition, endogenous estrogens in female WT mice had no impact on FMD as it was equivalent in male, female and ovariectomized female mice. These data suggest that FMD involves unliganded ERα activation in response to shear stress. In agreement, a recent study has reported another action of unliganded ERα, namely its inhibitory action on endothelial cell proliferation and migration (Lu et al., 2017).

Although FMD was reduced in Esr1-/- mice, agonist-mediated endothelium dependent (acetylcholine and insulin) and independent (SNP) dilation was not affected suggesting a selective reduction in flow (shear stress)-dependent signaling without a change in receptor-dependent dilation in the endothelium and in the smooth muscle.

The present study also showed that FMD involves membrane-associated ERα. FMD was similarly reduced in Esr1-/- mice and in both C451A-ERα and R264A-ERα mice. Although acute response (FMD and agonist-dependent dilation) can only be attributed to membrane-associated events, the expression level of the enzymes involved in the process could be modulated by the nuclear effects of ERα. Thus, endothelium-dependent dilation was investigated in mice lacking either the nuclear activating function AF2 of ERα or membrane-dependent action of ERα. Membrane-ERα is located at the level of the caveolae through either a binding to caveolin-1 or to striatin, thus creating a link with the Gαi and Gβγ proteins (Arnal et al., 2017). In order to abrogate the membrane effects of ERα, we used two different models. First, we used C451A-ERα mice that lack the palmitoylation site (cysteine at position 451) of the receptor so that the anchorage of ERα to the plasma membrane and the link to caveolin-1 is prevented (Adlanmerini et al., 2014). We also used a knock-in mouse model of ERα mutated for the arginine 264 (R264A-ERα mice) suppressing its interaction with Gαi involved in rapid eNOS activation (Adlanmerini et al., 2020). The fact that FMD was similarly reduced in Esr1-/-, C451A-ERα, and R264A-ERα mice without change in receptor-dependent dilation, strongly supports that membrane-associated ERα is involved in FMD. This is further supported by the absence of reduction in FMD observed in AF2-ERα mice which only lack the AF2 nuclear function of ERα. AF2 is also involved in the vascular response to a chronic increase in flow (flow-mediated remodeling) which is absent in AF20-ERα mice but fully present in C451A-ERα and R264A-ERα mice (Guivarc’h et al., 2018). This remodeling is a chronic adaptation of the vascular wall associated with changes in arterial diameter and wall mass within 2 weeks after a chronic rise in blood flow in vivo (Chehaitly et al., 2021). A chronic increase in blood supply, such as that needed for collateral growth in ischemic disorders, induces an increase in diameter together with wall thickening so that both shear and tensile stress are normalized within 1 week following the flow increase (Silvestre et al., 2001). This remodeling involves an early inflammatory phase allowing cell growth and reorganization in the arterial wall (Caillon et al., 2016) and a dilatory phase involving NO, prostaglandins and CO production (Dumont et al., 2007; Belin de Chantemèle et al., 2010; Freidja et al., 2011). Noteworthy, flow-mediated remodeling is absent in ovariectomized rats and mice and in Esr1-/- mice (Tarhouni et al., 2013) whereas this remodeling is preserved in ovariectomized rats treated with E2 (Tarhouni et al., 2014b) or resveratrol (Petit et al., 2016). More recently, we have shown that this remodeling requires activation of AF2 and is independent on membrane-located ERα (Guivarc’h et al., 2018).

Another membrane receptor for E2 located at the plasma membrane is GPER (Prossnitz and Barton, 2011). Both ligand-dependent and ligand-independent activation of GPER have been reported (Meyer et al., 2016). GPER is involved in regulation of reproductive functions, endocrine regulation and metabolism, cardiovascular, kidney, neuroendocrine and cerebral functions function as well as immune cell function. Furthermore, previous studies suggest a role for GPER in hypertension, kidney diseases, diabetes, and immune diseases. Consequently, GPER is a potential therapeutic target for the treatment of these diseases (Prossnitz and Barton, 2011). In the present study, incubation with the GPER antagonist G-36 did not affect FMD, ruling out the role of ligand-activated GPER in FMD. However, a possible role of unliganded GPER activation cannot be excluded in case of a crosstalk between membrane-dependent ERα and GPER activation.

To characterize the effect of membrane-ERα on the NO pathway which is involved in FMD, we investigated the effect of L-NNA-mediated inhibition of NO-synthesis on FMD. L-NNA inhibited FMD in arteries from WT and AF20ERα, but not arteries from Esr1-/-, C451A-ERα and R264A-ERα mice, suggesting that membrane-associated ERα is involved in NO-dependent FMD. This is in agreement with a previous study that used the ERα Neo-KO model with incomplete deletion, thereby showing that the NO pathway (dilation sensitive to L-NAME) was reduced in response to flow in the gracilis artery of male mice (Sun et al., 2007). In contrast to FMD, L-NNA strongly reduced acetylcholine-mediated dilation in WT, Esr1-/- AF20ERα, C451A-ERα and R264A-ERα mice. Thus, the NO-pathway can be activated in response to receptor stimulation in C451A-ERα and R264A-ERα mice, whereas only its activation by flow was reduced in these mice lacking only membrane-ERα signaling. FMD can involve prostaglandins and EDHF such as EETs (Sun et al., 2007) and EDHF was shown to mediate estrogen-mediated dilation of the uterine arteries (Burger et al., 2009). Nevertheless, in the present study, cyclooxygenase inhibition with indomethacin and EETs production inhibition with MSPPOH did not affect FMD in both WT and Esr1-/-, AF20ERα, C451A-ERα and R264A-ERα mice, suggesting a limited role of the pathway in mesenteric arteries of male mice. Consequently, the remaining FMD following the addition of L-NNA, indomethacin and MSPPOH relies probably on other hyperpolarizing agents. Indeed, endothelium-dependent hyperpolarization (EDH) has a major role in resistance arteries homeostasis (Brandes et al., 2000; Garland and Dora, 2017) and COX-derivatives can also induce EDH in resistance arteries and in the carotid when submitted to flow (Ohlmann et al., 2005; Bergaya et al., 2001). In humans, FMD measured in the brachial artery relies mainly on the production of NO (Alexander et al., 2021). Furthermore, changes in FMD in the brachial artery predict well the endothelial dysfunction in human resistance arteries (Park et al., 2001). Nevertheless, the difference in the nature of the agents involved in FMD besides NO between humans and mice could be a limitation of the present study. As stated above, the involvement of EDH in FMD is greater in mouse resistance arteries than in humans when measured at the level of the brachial artery.

As FMD has a key role in blood flow delivery to organs (Hill et al., 2010), we investigated the flow-pressure relationship in the mouse kidney. In agreement with the reduction in FMD observed in resistance arteries, we found a leftward shift of the flow-pressure relationship in C451A-ERα mice further confirming the decreased sensitivity to flow of the resistance vasculature. Recent studies have shown that flow activates Piezo1-dependent release of ATP through pannexin hemi-channel followed by P2Y2 activation and NO production by endothelial cells (Wang et al., 2015; Wang et al., 2016). We found that both nitrate-nitrite and ATP productions were lower in the kidney perfusate from C451A-ERα than in WT mice, in agreement with the reduced NO-dependent FMD (sensitive to L-NNA) observed in isolated arteries. As in mesenteric arteries, we have previously shown that eNOS expression level in the kidney is not altered by the absence of membrane ERα in C451A-ERα (Guivarc’h et al., 2020). Although NO and ATP production were reduced in kidneys from C451A-ERα, the acute response to ATP and to the Piezo1 agonist YODA-1 was not affected by the absence of membrane ERα. In addition, the mechanosensitive channel blocker GsMTx4 similarly affected FMD in C451A-ERα and WT mice, suggesting that the defect in FMD associated with the absence of membrane ERα is probably located downstream flow sensing.

The reduction in NO-dependent FMD found in arteries from mice lacking membrane-ERα could be due to an excessive ROS production as the chronic treatment of C451A-ERα mice with an antioxidant treatment restored FMD to control level. In agreement, we found that H2O2 production by the kidney was higher in C451A-ERα than in WT mice. This observation is in agreement with a previous work that has shown that E2 increases the release of bioactive NO by inhibition of superoxide anion production in bovine endothelial cells (Arnal et al., 1996). Accordingly, we found that reducing total ROS or mitochondrial ROS production improved FMD in C451A-ERα mice. By contrast, catalase which eliminates H2O2 reduced FMD in C451A-ERα mice suggesting that an excessive ROS production due to the absence of membrane-associated ERα had a dual effect with (1) a reduction of NO bioavailability and (2) an increase in H2O2 production contributing to some vasodilating effect. Noteworthy, kidney perfusates showed higher levels of H2O2 in C451A-ERα mice. An excessive ROS production could also alter eNOS activation as previously shown through increased phosphatase activation (Ding et al., 2020). Previous studies have also shown that shear stress induces a more quiescent and less oxidative phenotype in endothelial cells (Doddaballapur et al., 2015; Wu et al., 2018), thus reducing the oxidative products of mitochondrial origin. Nevertheless, in a context of known dysfunctional FMD, a previous work has shown that H2O2 could mediate FMD in human coronary arteries from patients suffering coronary artery disease although H2O2 remains deleterious (Freed et al., 2014).

Conclusion

To conclude, these data demonstrate for the first time a major role of ERα, and more precisely of non liganded endothelial membrane-located ERα for optimal FMD and thereby a potential role in local blood flow homeostasis. The mechanism appears to involve an optimization of NO activation and/or a decrease in ROS production as depicted in Figure 9. The functional consequences in terms of arteriolar and tissue protection should now be investigated.

Schematic representation of the known E2-mediated ERα-dependent protective effects (upper panel) and of the new pathways described in the present study (lower panel).

Previous works (upper panel) have demonstrated the role of E2 and the nuclear activating function AF2 of ERα against atherosclerosis and hypertension (Guivarc’h et al., 2018) as well as in flow-mediated outward remodeling (Tarhouni et al., 2013; Guivarc’h et al., 2018). E2-stimulated membrane-located ERα is involved in E2-dependent NO production and in endothelial healing (Adlanmerini et al., 2014). New pathway described in the present work (lower panel): Flow, by stimulation of the surface of the endothelial cell by shear stress, activates the NO pathway (e.g. phosphorylation of eNOS: P-eNOS). This results in the production of NO, which in turn induces relaxation of the smooth muscle and thus dilation. In parallel, flow activates membrane-associated ERα, which reduces oxidative stress (O2-. and H2O2) due to NADPH-oxidase activity or of mitochondrial origin. This results in enhanced NO bioavailability. The absence of membrane-associated ERα could lead to the production of O2-., which attenuates NO-dependent dilation despite a remaining dilation due to a rise in H2O2 production.

Materials and methods

Key resources table
Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional information
Strain, strain background(Mus musculus, males and females)Esr1-/-C57BL/6 J(Symbol: Esr1tm1.1Mma, Synonyme: ERalpha Knockout)Mouse Clinical Inst., Strasbourg, France, Dupont et al., 2000MGI:2386760
Strain, strain background(Mus musculus, males)AF2°ERα,C57BL/6 J(Symbol: Esr1tm1.1Ohl Synonym: ERalpha-AF2°)Mouse Clinical Inst., Strasbourg, France, Billon-Galés et al., 2009MGI:4950046
Strain, strain background(Mus musculus, males)C451A-ERα, C57BL/6 N(Symbol: Esr1tm1.1Ics Synonyme: C451A-ERalpha knock-in)Mouse Clinical Inst., Strasbourg, France, Adlanmerini et al., 2014MGI:5574591
Strain, strain background(Mus musculus, males)TekCre/+:Esr1f/f, C57BL/6(B6.Cg-Tg(Tek-cre)12Flv/J backcrossed with Esr1tm1.2MmaSynonym:Tie2Cre ERαlox/lox)Esr1lox/lox: Mouse Clinical Institut, Strasbourg, France.TekCre: Jackson Lab (Bar Harbor, Me), Billon-Galés et al., 2009TekCre:Koni et al., 2001Esr1lox/lox:Dupont et al., 2000TekCre:Esr1lox/lox:MGI:3775510
Strain, strain background(Mus musculus, males)Esr2-/-,C57BL/6 J (Symbol: Esr2tm1MmaSynonym: ERbeta)Mouse Clinical Inst., Strasbourg, France, Dupont et al., 2000MGI:2386761
Strain, strain background(Mus musculus, males)R264A-ERα, C57BL/6 NMouse Clinical Inst., Strasbourg, France, Adlanmerini et al., 2020No MGI ID yet
AntibodyAnti-eNOS, (mouse monoclonal, clone3)BD BiosciencesCat# 610297, RRID:AB_397691WB (1:1000)
AntibodyAnti-phospho-eNOS, pS1177 (Mouse monoclonal,Clone 19/eNOS/S1177)BD BiosciencesCat# 612392, RRID:AB_399750WB (1:1000)
AntibodyAnti-beta-actin, (Mouse monoclonal, clone AC-74)Sigma-AldrichCat#: 5316; RRID:AB_476743WB (1:5000)
AntibodyAnti-Akt Pan, (rabbit monoclonal, clone C67E7)Cell signalling technologyOzymeCat#: 4691; RRID:AB_915783WB (1:1000)
AntibodyAnti-phospho-Akt, S473, (rabbit monoclonal, clone D9E)Cell signalling technology OzymeCat#: 4060; RRID:AB_2315049WB (1:2000)
AntibodyAnti-mouse IgG (H + L) Secondary antibody HRP (Goat polyclonal)Thermo scientificCat#: 31430; RRID:AB_228307WB (1:5000)
AntibodyAnti-rabbit IgG(H + L) Secondary antibody HRP (Goat polyclonal)Thermo scientificCat#: 31460; RRID:AB_228341WB (1:10000)
Chemical compound, drugvitamin CSigma Aldrich Merck, Favre et al., 2011A5960
Chemical compound, drugvitamin ESigma Aldrich Merck, Favre et al., 2011T3251
Chemical compound, drugMito-tempoSigma Aldrich Merck, Freed et al., 2014SML0737
Chemical compound, drugcatalaseSigma Aldrich Merck, Bouvet et al., 2007C3155
Chemical compound, drugPEG-superoxide dismutase (SOD)Sigma Aldrich Merck, Bouvet et al., 2007S9549
Chemical compound, drugEstetrol (E4)Sigma Aldrich Merck, Abot et al., 2014SML1523
Chemical compound, drugICI 182 780Tocris Biotechne, Meyer et al., 20101047
Chemical compound, drugG-1 ((±)–1-[(3aR*,4S*,9bS*)–4-(6-Bromo-1,3-benzodioxol-5-yl)–3 a,4,5,9b-tetrahydro-3H-cyclopenta[c]quinolin-8-yl)]- ethanoneCayman chemical Bertin Bioreagent, Meyer et al., 201010008933
chemical compound, drugG-36 ((±)-(3aR*,4S*,9bS*)–4-(6-Bromo-1,3-benzodioxol-5-yl)–3 a,4,5,9b-tetrahydro-8-(1-methylethyl))–3H-cyclopenta[c]quinolineCayman chemical Bertin Bioreagent, Meyer et al., 201614,397
Sequence-based reagentN-(methylsulfonyl)–2-(2-propynyloxy)-benzenehexanamide (MSPPOH)Cayman chemical Bertin Bioreagent, Dietrich et al., 200975,770
Chemical compound, drugGrammostola spatulata mechanotoxin 4 (GsMTx4)Alomone Labs, John et al., 2018STG-100
Chemical compound, drugYODA1Bertin Bioreagent, Lhomme et al., 2019SML1558
Chemical compound, drugATPγSTocris Biotechne, Kukulski et al., 20094080
Chemical compound, drug4-hydroxy-2,2,6,6-tetramethylpiperidine (TEMPOL)Sigma Aldrich Merck, Freidja et al., 2014176,141
commercial assay or kitNitric oxide metabolite detection kitCayman Chemical780,051
commercial assay or kitHydrogen peroxide assay kitAbcamAb102500
commercial assay or kitATP determination kitInvitrogen Molecular ProbesA22066

Animal protocol

Request a detailed protocol

We used 5–6 month-old male mice lacking the gene encoding ERα (Esr1-/-) (Antal et al., 2008) or ERβ (Esr2-/-) (Antal et al., 2008), mice lacking ERα selectively targeted to the endothelium (TekCre/+:ERf/f) (Toutain et al., 2009), mice lacking the nuclear activation function AF2 (AF20ERα mice) (Billon-Galés et al., 2011), mice in which the codon for the cysteine (Cys451) palmitoylation site of ERα had been mutated to alanine (C451A-ERα mice) (Adlanmerini et al., 2014) and mice mutated for the arginine 264 of ERα (R264A-ERα mice) (Adlanmerini et al., 2020). Littermate +/+ mice were used as controls (designated wild-type, WT, or +/+) in each group.

In a separate series of experiments, 5–6 month-old female Esr1-/- and Esr1+/+ mice were used for FMD measurements. The mice had been ovariectomized or left intact (with only a sham surgery), as previously described (Toutain et al., 2009).

In another series of experiments, 5–6 month-old male C451A-ERα and C451A-WT mice were treated with the antioxidant 4-hydroxy-2,2,6,6-tetramethylpiperidine (TEMPOL, 10 mg/kg per day, 2 weeks in drinking water) (Belin de Chantemèle et al., 2009) or with the antioxidants vitamin E (1 % in chow) and vitamin C (0.05 % in water) for 4 weeks. (Favre et al., 2011; Contreras-Duarte et al., 2018).

The mice were anesthetized with isoflurane (2.5%) and euthanized with CO2. The mesentery and the uterus were quickly removed and placed in ice-cold physiological salt solution (PSS) (Tarhouni et al., 2013). Several segments of second-order arteries were collected for the functional study and for biochemical studies.

The experiments complied with the European Community standards for the care and use of laboratory animals and the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85–23, revised in 1996). The protocol was approved by the regional ethics committee (permits #14335, #16740, and #16108).

Flow-mediated dilation in mesenteric arteries in vitro

Request a detailed protocol

Arterial segments, with internal diameters of approximately 200 µm, were cannulated at both ends on glass micro-cannulas and mounted in a video-monitored perfusion system (Living System, LSI, Burlington, VT, USA) (Iglarz et al., 1998; Bolla et al., 2002). Individual artery segments were bathed in a 5 ml organ bath containing PSS (pH: 7.4, pO2: 160 mmHg, and pCO2: 37 mmHg) and perfusion of the artery was carried out with two peristaltic pumps, one controlling the flow rate and the other under the control of a pressure-servo control system. The pressure was set at 75 mmHg and flow (3–50 µl per min) was generated through the distal pipette with a peristaltic pump. Flow steps were 3, 6, 9, 12, 15, 30 and 50 µl/min which correspond to 0.8, 1.2, 2, 2.8, 4, 8 and 12 dyn/cm2.

FMD was determined before and after pretreatment with N(omega)-nitro-L-arginine (L-NNA, 100 µM, 30 min), L-NNA plus indomethacin and then with L-NNA plus indomethacin (10 µmol/L) plus N-(methylsulfonyl)–2-(2-propynyloxy)-benzenehexanamide (MSPPOH, 10 µmol/L).

In a separate series of experiments, the effect the mechanosensitve ionic channels blocker Grammostola spatulata mechanotoxin 4 (GsMTx4) (5 µmol/L, delivered intraluminally and incubated for 45 min) (John et al., 2018).

The impact of ex vivo modulation of ERα on FMD was evaluated after 20 minutes of incubation with the ERα agonists E2 (10 nM) or E4 (1 µmol/L), the GPER agonist G-1 (1 µmol/L) (Meyer et al., 2010), the GPER antagonist G-36 (1 µmol/L) (Yu et al., 2018) or the estrogen receptor downregulator and GPER agonist ICI 182 780 (0.1 µmol/L) (Meyer et al., 2010).

In another series of experiments, FMD was measured before and after incubation (20 min) of the arteries with PEG-superoxide dismutase (SOD, 120 U/mL) plus catalase (80 U/mL) (Bouvet et al., 2007), catalase (80 U/mL), or Mito-Tempo (1 µmol/L) (Freed et al., 2014).

Pharmacological profile of isolated mesenteric arteries

Request a detailed protocol

Segments of mesenteric arteries were mounted in a wire-myograph (Danish Myo Technology, Denmark) as previously described (Loufrani et al., 2002) in order to obtain cumulative concentration-response curves (CRCs) to acetylcholine (ACh) before and after pretreatment with L-NNA (10 µmol/L) and then with L-NNA (100 µmol/L) plus indomethacin (10 µmol/L).

In a separate series of experiments, CRCs to YODA1, ATPγS (ATP) were performed.

Prior to each CRC, the arteries were submitted to phenylephrine to obtain approximately 50 % of the maximal contractile response of the vessel assessed by KCl (80 mM)-mediated contraction at the beginning of the experiment.

Western-blot analysis

Request a detailed protocol

Arterial segments were cannulated under pressure (75 mmHg), and flow (50 µl/min) was applied after precontraction with phenylephrine (1 µM). After 2 minutes, the arteries were quick-frozen. Due to the limited size of the resistance arteries segments were pooled before analysis. Protein expression (eNOS, phospho-eNOS, Akt and phospho-Akt) was then determined using Western blot (Bouvet et al., 2007).

Perfused isolated mouse kidney

Request a detailed protocol

In a separated series of experiments, the right renal artery was cannulated in anesthetized mice and the kidney was excised and perfused at 37 °C with PSS as previously described (Begorre et al., 2017). The right renal artery was cannulated in anesthetized mice (as described above) with a polyethylene catheter (PE-10, 0.28 mm internal diameter, 0.61 mm external diameter, Intramedic, Evry, France). The kidney was then excised and perfused without interruption of kidney flow at 37 °C with PSS. The perfusion solution was dialyzed and the pH was adjusted to 7.4. Perfusion rate was 600 µl/min and perfusion pressure was measured continuously (PT-F pressure transductor, Living System, Burlington, VT). Endothelium-mediated dilation was tested using ACh (1 µmol/L) after precontraction with Phe (1 µmol/L). Flow-pressure relationship was assessed through an stepwise increase in perfusion flow associated with the continuous measurement of the perfusion pressure.

The PSS perfusing the kidney (perfusate in the scheme shown in Figure 6) was collected in baseline conditions (flow = 600 µl/min) and immediately frozen in liquid N2 and then stored at –80 °C.

Determination of nitrate and nitrite, ATP and H2O2 levels in the kidney perfusate

Request a detailed protocol

To determine flow-induced nitrate-nitrite, ATP or H2O2 release from the perfused mouse kidney, 500 µl of perfusate was collected. The perfusate was then centrifuged 10 min at 14 000 rpm and the supernatant added into a spin column with 10 kDa molecular weight cut-off filter for ultrafiltration (10KD Spin Column Abcam ab93349) and centrifuged at 10 000 rpm for 10 min. The centrifuged solutions was then used for nitrate-nitrite, ATP and H2O2 measurement.

Nitrate and nitrite levels in kidney perfusate were determined using a nitrate/nitrite fluorometric assay kit from Cayman Chemical (Nitric Oxide Metabolite Detection Kit Nb°780051) according to the manufacturer’s instructions.

Hydrogen peroxyde (H2O2) level was determined using the fluorimetric method of a hydrogen peroxide assay kit from Abcam (Hydrogen Peroxide Assay Kit Colorimetric/Fluorometric NbAb102500) according to the manufacturer’s instructions.

ATP level was measured using ATP determination kit from Invitrogen Molecular Probes (ATP Determination Kit Nb A22066) according to the manufacturer’s instructions.

Preparation of endothelial cells enriched fraction for transcriptional analysis

Request a detailed protocol

Endothelial cells enriched fractions were obtained as previously described (Briot et al., 2014). Briefly, 5-week-old female mice were perfused with PBS. The descending thoracic aorta was dissected and perfused with RLT buffer (Qiagen, Valencia, CA) containing 1 % beta-mercaptoethanol. Endothelial cells enrichment was confirmed by the increased endothelial marker Tek expression level and the absence of smooth muscle cell marker Cnn1 (Kalluri et al., 2019) compared to the total aorta.

Evaluation of gene expression by quantitative real-time PCR in mesenteric arteries

Request a detailed protocol

Gene expression was investigated using quantitative polymerase chain reaction after reverse transcription of total RNA (RT-qPCR). Mesenteric arteries were stored at −20 °C in RNAlater Stabilization Reagent (Qiagen, Valencia, CA, USA) until use. RNA was extracted using the RNeasy Micro Kit (Qiagen, Valencia, CA, USA) following manufacturer instructions. RNA extracted (300 ng) was used to synthesize cDNA using the QuantiTect Reverse Transcription Kit (Qiagen, Valencia, CA, USA). RT-qPCR was performed with Sybr Select Master Mix (Applied Biosystems Inc, Lincoln, CA, USA) reagent using a LightCycler 480 Real-Time PCR System (Roche, Branchburg, NJ, USA). Primer sequences are shown in the Supplementary file 1. Gapdh, Hprt and Gusb were used as housekeeping genes. Analysis was not performed when Ct values exceeded 35. Results were expressed as: 2(Ct target-Ct housekeeping gene).

Statistical analysis

Request a detailed protocol

The results are expressed as means ± the SEM. The significance of the differences between groups was determined by analysis of variance (two-way ANOVA for consecutive measurements) followed by Bonferroni’s test for the FMD and the agonist-mediated concentration-response curves. A two-tailed Mann-Whitney test (when comparing two groups) or a Kruskal-Wallis test (more than two groups) was used for the other comparisons as indicated in the figure legends. Probability values less than 0.05 were considered significant.

Data availability

All data generated or analysed during this study are included in the manuscript and supporting files. Source data files are provided for each figure and supplement.

References

    1. Dupont S
    2. Krust A
    3. Gansmuller A
    4. Dierich A
    5. Chambon P
    6. Mark M
    (2000)
    Effect of single and compound knockouts of estrogen receptors alpha (ERalpha) and beta (ERbeta) on mouse reproductive phenotypes
    Development 127:4277–4291.

Decision letter

  1. Noriaki Emoto
    Reviewing Editor; Kobe Pharmaceutical University, Japan
  2. Matthias Barton
    Senior Editor; University of Zurich, Switzerland
  3. Philip W Shaul
    Reviewer; University of Texas Southwestern Medical School, United States

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Decision letter after peer review:

Thank you for submitting your article "Membrane estrogen receptor alpha (ERα) optimizes flow-mediated dilation in both sexes, in a ligand-independent manner" for consideration by eLife. Your article has been reviewed by 4 peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by a Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Philip W Shaul (Reviewer #4).

The reviewers have discussed their reviews with one another, and this letter is to help you prepare a revised submission.

Essential revisions:

1) The ligand-independent activation of membrane ERα is a novel finding. However, simply ignoring a potential role of GPER in such ligand-independent regulation of FMD is a major flaw. The authors should carry out some straight-forward studies to look at the effects of pharmacological inhibition using GPER antagonist, G-36, which turns out to not only inhibit GPER but also via inhibition of constitutive GPER expression, to reduce abundance (and activity) of Nox1. The authors are advised to perform in vitro studies of vessels from WT mice, evaluating dilation in response to flow +/- GPER antagonist, as long as a parallel control study is done evaluating dilation in response to GPER agonist +/- GPER antagonist.

2) In the Introduction section, the authors are encouraged to cite original works related to non-nuclear signaling by the subpopulation of ER associated with the plasma membrane in endothelium.

3) In the Introduction section, the background information about cardiovascular disease risk in men versus women and the impact of estrogens on risk is not as clear-cut as the authors imply. A more balanced presentation of the clinical evidence is warranted.

4) In the Introduction, a more detailed account of the published literature in the field including studies in humans should be provided The Authors may consider citing some studies from among PMID 19126786; 7642876; 20185791; 26471832; 9176294; 17367797; 28645860; 26734763; 9791075; 16131583; 18319309; 19086257; 9930647.

5) In the Results section, Page 6, line 106 incorrectly refers the reader to Figure 1B.

6) In the Results section. Page 6, line 117: fulvestrant is an ERα and ERβ antagonist (pKi = 9 for both) as well as a GPER full agonist (pKi = 7).

7) In the Results section, Figures 2B, 2F and 2J: ERα protein abundance in endothelial cells should be evaluated, even if only feasible from vasculature that would yield more endothelial cells than mesenteric resistance arteries.

8) In the Results section, Figure 5B: the flow-pressure relationship was shifted left, or up, in C451A-ERα mice versus WT according to the group labeling, and not to the right.

9) In the Results section Page 17, line 281: the use of the term autoregulation here may be confusing, and more specific interpretation would be helpful (and better placed in the Discussion section).

10) In the Results section, how does flow increase eNOS Ser1177 phosphorylation and alter H2O2 via ERα mechanistically ? Are flow-related changes in Akt phosphorylation in endothelium altered in the absence of ERα?

11) In the Results section, Do non-nuclear actions of plasma membrane-associated ERα influence endothelial cell production of prostaglandins in the setting of FMD?

12) In the Results section, do non-nuclear actions of plasma membrane-associated ERα influence mechanosensitive ion channel localization or function in endothelial cells?

13) In the Results section, why was the antioxidant TEMPOL specifically selected? Would the Authors expect a similar outcome following treatment with other antioxidants such as e.g. quercetin or dimethyl fumarate?

14) In the Results section, does Tempol or PEG-catalase acutely restore FMD ex vivo?

15) In the Results section, Line 312: Chronic for a 2-week treatment does not sound very appropriate.

16) In the Results section, flow rate may be changed to shear stress in Dyn/cm2: this would help comparing between published works in the topic.

17) The Discussion section should be shortened, providing more succinct focused discussion of the interpretation of the findings, their implications, and possible explanations to fill the new knowledge gaps that result from the work.

18) In the Conclusion section, Figure 7: the schematic is only in part helpful because how non-nuclear actions of membrane-associated ERα in endothelial cells govern eNOS and ROS in response to flow is not addressed.

Reviewer #1:

Flow mediated dilation (FMD) is a response to acute shear stress and its reduction is known to be a hallmark of endothelium dysfunction, associated with aging and with cardiovascular and metabolic disorder. Estrogen receptor alpha (ERα) was previously found to be associated with FMD in mouse and human. This manuscript further explore which ERα variant was important for FMD. The authors reported a novel ligand-independent pathway leading to flow mediated dilation attributed to the membrane-bound ERα. The authors used mouse model with fully deficient ERα, as well as mouse with specific inactivation of either membrane-bound or nuclear ERα and showed that only disruption of membrane-bound ERα greatly inhibit flow mediated dilation. The author further revealed the mechanism involving shear stress-induced attenuation of ROS levels which increase eNOS activity. Therefore, membrane-bound ERα could be a potential target therapy to reduce oxidative damage in endothelium of resistance arteries.

Strengths

This study presented for the first time that flow mediated dilation produced by estrogen receptor alpha is attributed to its membrane receptor via ligand-independent pathway.

Weaknesses

The mechanism of membrane receptor, ligand independent pathway of estrogen receptor alpha that the authors reported lacks novelty as it was already previously described.

Reviewer #2:

Multiple animal models have been used to test the contribution of membrane estrogen receptors on the vascular dilation induced by flow (flow-mediated dilation, FMD).

The authors propose that the presence of membrane estrogen receptors optimizes flow-mediated dilation. However, the conclusion that ER promotes NO production and inhibits oxidative stress is not fully supported by the data since basal ROS production is not altered in the models without functional ER, and an antioxidant treatment normalizes the dilatory response, and effect that can be independent of a direct NOS activation. The data show that in the absence of functional membrane receptors, flow-mediated dilation is reduced but that it is restored by a treatment with an antioxidant. Thus, these receptors seem not to be necessary for FMD. The involvement of these receptors in FMD remains therefore questionable.

Reviewer #3:

The Authors report that blood flow in mouse resistance arteries activates membrane-associated ERα, which reduces oxidative stress (O2-.), resulting in enhanced NO bioavailability. The absence of membrane-associated ERα may lead to the production of O2-. which attenuates NO-dependent dilation as shown by a rise in H2O2 generation in the perfused kidney.

Strengths

(1) Use of cutting-edge animal models to explore the role of ERα in the vascular endothelium.

(2) Robust experimental procedures including kidney perfusion.

(3) Comprehensive set of experiments including appropriate controls.

(4) Disclosure of a new ligand-independent vascular ERα effect.

Weaknesses

(1) Ignoring a possible role of GPER and signaling pathways thereof in the study endpoints is a major flaw in the experimental design.

(2) The regulation of flow-mediated dilation by estrogen has been widely investigated in previous studies.

(3) The gender claim in the title is relatively weak as it is based on just one set of experiments (Figure 1; lines 137-138).

Although the study objective was stated concisely, the Authors generated findings of potential impact in the field that deserve further investigation. A cross-talk between blood flow, NO pathway and membrane-associated ERα appears to emerge from the present work and represents a conceptual advance. However, the role of GPER in this setting deserves to be assessed as well.

Reviewer #4:

Using a variety of genetically-manipulated mouse models, the authors study how membrane associated estrogen receptor alpha (ERα) impacts flow-mediated dilation (FMD). Their collective findings indicate that non-nuclear actions of plasma membrane-associated ERα in endothelial cells support FMD in both sexes via mechanisms that are independent of estrogens. The process likely involves the promotion of NO production and blunting of impact of reactive oxygen species (ROS).

Strengths include the use of multiple genetic manipulations in mice, which allow testing of ERα function as a transcription factor and as a modulator of extra-nuclear signaling initiated by a subpopulation of plasma membrane-associated ERα. The specific requirement for ERα in endothelial cells is also evaluated. FMD is primarily tested in isolated mesenteric resistance arteries, but some key findings are confirmed in uterine arteries. The complementary use of an isolated perfused kidney model is also a strength.

One weakness is the evaluation of levels of ERα expression by quantifying transcript abundance in whole arteries when ERα protein abundance in endothelial cells is of prime importance. In addition, although alterations in NO and ROS are implicated, no insights are gained into how these are impacted by ERα presumably independent of estrogens. Despite the lack of more mechanistic interrogation, the overall observation of an important role for non-nuclear function of plasma membrane-associated ERα in endothelial cells in FMD is important to the field.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for submitting your revised manuscript "Membrane estrogen receptor alpha (ERα) optimizes flow-mediated dilation in both sexes, in a ligand-independent manner" for consideration by eLife. Your revised article has been evaluated by a Reviewing Editor and a Senior Editor. the Reviewing Editor has drafted this letter to help you prepare a revised submission.

Essential revisions:

1) The wording "optimization" in the title may be "optimized". The reviewer thinks that e.g. an experimental procedure but not a biological process such as FMD may be optimized. ER participates in or supports FMD.

2) The p values and other statistical parameters could be more appropriately moved from figures to figure legends, especially when asterisks are present.

3) qPCR experiments: "relative expression" may refer to either an housekeeping gene or a reference control sample. Please indicate the first option in figure legends.

4) Figure 5: Western blots should be shown for phospho-eNOS and total eNOS, and for phospho-Akt and total Akt.

5) Line 63-65: The authors state that flow-mediated dilatation depends mainly on NO, citing studies looking at the forearm circulation in humans. This should be stated/clarified. However, in mice (the organism used for experiments in this study) flow-dependent vasodilatation in mesenteric resistance arteries (the vascular bed investigated in this study) is mediated by both, NO and EDH (PMID 16055522; 11282893). Brandes and associates also found that EDH is the main mediator of endothelium-dependent relaxation in murine resistance arteries (PMID 10944233). Moreover, the EDH-shear-dependent response is partly sensitive to non-selective COX-inhibition (PMID 16055522). The authors should discuss these important differences between humans and mice and also should mention the limitation of their study that experiments were conducted in the absence of COX inhibition, and that one cannot fully exclude the involvement of COX-dependent / endothelial-derived prostanoid effects in the effects observed.

6) Lines 112-115: With regard to estrogen effects in men the authors should discuss intracrine, aromatase-mediated production of estradiol which is converted locally in the vasculature from testosterone (PMID 11248122) by which estrogen partly provides protection in male mice from atherosclerosis.

7) Lines 145-147: The authors suddenly introduce compounds targeting GPER, without having made any reference to this receptor in the introduction. It would be helpful if the authors could add a little section to the introduction discussing this receptor as well, also citing a good overview article, such as PMID 21844907

8) Line 144, Line 465-466, : The authors state "the ERα-ERb antagonist and GPER agonist fulvestrant (ICI-182780)." This is partly correct. Fulvestrant is a SERD (selective estrogen receptor downregulator or selective estrogen receptor degrader), which down-regulates/degrades the receptors it targets. This should be corrected.

9) Line 262: findings for the evaluation of changes of gene expression with C451A-ERα are not shown in Figure 7.

10) Lines 335-336: the statement is misleading because chronic effects of E2 (recognizing that the term "chronic" is vague) can be mediated by non-nuclear actions of ERα.

11) Lines 402-403: there is no basis for stating that the decrease in FMD in mice lacking membrane ERα could reflect a feature of premature aging of the endothelium.

12) Figure 9 schematic: there is no evidence that Gai mediates the role of membrane-associated ERα in FMD.

13) P. 21: The reference for G36 provided in the table is not quite correct, the one cited was published to 2010, at a time when G36 had not yet been published. It is suggested to list PMID 27803283 as reference instead.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "Membrane estrogen receptor alpha (ERα) participates in flow-mediated dilation in a ligand-independent manner" for further consideration by eLife. Your revised article has been evaluated by a Reviewing Editor and a Senior Editor.

The manuscript has been improved but there are some remaining issues that need to be addressed, as outlined below:

1) Previous comment "Lines 145-147: The authors suddenly introduce compounds targeting GPER, without having made any reference to this receptor in the introduction. It would be helpful if the authors could add a little section to the introduction discussing this receptor as well, also citing a good overview article, such as PMID 21844907"

Authors Response: We have added a paragraph in the introduction stating the role of GPER in endothelium-dependent dilation (lines 87-90):

"E2-dependent vasodilation has also been shown to involve the G-protein-coupled estrogen receptor (GPER) in both human and animal arteries [24]. In the rat, GPER activation reduces uterine vascular tone during pregnancy through activation of endothelium-dependent NO production [25]."

It appears that there was a misunderstanding with regard to the previous comment. It was not requested to describe solely the role of GPER in endothelial cell function and NO release. Rather, the authors were expected to expand this section, briefly describing the nature of GPER (namely that it is a 7-transmembrane GPCR located at the endoplasmic reticulum), its natural and synthetic ligands, and its functions (release of NO is just one). Also, the authors should mention that there is both, ligand-dependent and ligand-independent activation (PMID 27803283) of GPER, which should also be discussed in the Discussion section.

2) The "Key Resources Table" in the Methods section includes a number of references:

Hypertension 2007;50:248-54

Hypertension. 2007;50:248-54

EMBO Mol Med. 2014;6:1328-46

Pharmacology 2010: 86, 58- 64

Sci Signal. 2016;9:ra105

J Vasc Res. 2009;46:2 53-64

Am J Physiol 2018;315:H1019-H1026

Cytokine 2009;46,166-70

These references should be added in full to the References section

3) The whole manuscript should be checked for grammar and typos, including "contribute to relaxation of the mesenteric resistance arteries in both male and female through, at least in part, a PI3K-Akt-eNOS pathway [28]." (lines 98-99, page 5) and corrections should be made as needed.

https://doi.org/10.7554/eLife.68695.sa1

Author response

Essential revisions:

1) The ligand-independent activation of membrane ERα is a novel finding. However, simply ignoring a potential role of GPER in such ligand-independent regulation of FMD is a major flaw. The authors should carry out some straight-forward studies to look at the effects of pharmacological inhibition using GPER antagonist, G-36, which turns out to not only inhibit GPER but also via inhibition of constitutive GPER expression, to reduce abundance (and activity) of Nox1. The authors are advised to perform in vitro studies of vessels from WT mice, evaluating dilation in response to flow +/- GPER antagonist, as long as a parallel control study is done evaluating dilation in response to GPER agonist +/- GPER antagonist.

We agree with your comment. Ligand-independent activation of membrane ERα has been previously demonstrated. However, the present finding shows, for the first time its involvement in the endothelium response to flow in resistance arteries.

We performed the experiments requested and added the data to figure 1 (panels I and J) and to the corresponding paragraph in the manuscript (Results section on figure 1 and discussion). We found that G36 incubation did not affect flow-mediated dilation in arteries from WT mice suggesting that GPER may not have a role in this vascular response although it is now clear that GPER has a full place in cardiovascular homeostasis and pathophysiology.

As G36 has an IC50 for G1 and E2 (estradiol) which are very close (165 nM and 112 nM, respectively: Dennis et al., J Steroid Biochem Mol Biol. 2011 Nov;127(3-5):358-66), we tested the effect of G36 on both compounds and observed that G36 inhibited both dilation indued by G1 and E2 (panel J, figure 1).

2) In the Introduction section, the authors are encouraged to cite original works related to non-nuclear signaling by the subpopulation of ER associated with the plasma membrane in endothelium.

We have extended this part of the discussion (Introduction, lines 74 to 93).

3) In the Introduction section, the background information about cardiovascular disease risk in men versus women and the impact of estrogens on risk is not as clear-cut as the authors imply. A more balanced presentation of the clinical evidence is warranted.

We have added more studies describing the protective effect of estrogens with a focus on FMD. This is mainly in the introduction between line 92 and line 111.

4) In the Introduction, a more detailed account of the published literature in the field including studies in humans should be provided The Authors may consider citing some studies from among PMID 19126786; 7642876; 20185791; 26471832; 9176294; 17367797; 28645860; 26734763; 9791075; 16131583; 18319309; 19086257; 9930647.

We have included most of these articles in the introduction and/or in the discussion.

Several articles are from our group and show that flow-mediated outward remodeling involves E2-dependent activation of nuclear function (AF2) of ERα. This remodeling occurs in response to a chronic increase in flow and affects all the layers of the arterial wall. By contrast with FMD which is an acute response involving non-nuclear ERα. We have added a paragraph on this difference to avoid confusion (lines 70-74 and in the discussion: lines 337-349).

5) In the Results section, Page 6, line 106 incorrectly refers the reader to Figure 1B.

We have corrected this error (lines 135-37).

6) In the Results section. Page 6, line 117: fulvestrant is an ERα and ERβ antagonist (pKi = 9 for both) as well as a GPER full agonist (pKi = 7).

We have completed the corresponding text as follows:

“Furthermore, the ERα-ERb antagonist and GPER agonist fulvestrant (ICI-182780) did not alter the FMD” Now in line 144.

Same correction in the Material and Methods section.

7) In the Results section, Figures 2B, 2F and 2J: ERα protein abundance in endothelial cells should be evaluated, even if only feasible from vasculature that would yield more endothelial cells than mesenteric resistance arteries.

We have performed these experiments in endothelial cells isolated from the aorta as too few cells can be obtained from the mesenteric arteries. This new data is shown in Figure 3 B, F and J.

Besides markers of endothelial cells (Tie2 or Tek) and of smooth muscle cells (Cnn1) were quantified in order to control for the quality of the endothelial cells isolation.

Tek and Cnn1 expression levels are in Figure 3, supplement figure 1.

8) In the Results section, Figure 5B: the flow-pressure relationship was shifted left, or up, in C451A-ERα mice versus WT according to the group labeling, and not to the right.

We have corrected this text. Thank you for having noticed this error.

9) In the Results section Page 17, line 281: the use of the term autoregulation here may be confusing, and more specific interpretation would be helpful (and better placed in the Discussion section).

We have modified the sentence as follows:

“Thus, these results suggest that FMD reduction due to the absence of membrane-ERα also affects the capacity of the renal vasculature to produce NO and ATP”

Now in Lines 254 – 257.

10) In the Results section, how does flow increase eNOS Ser1177 phosphorylation and alter H2O2 via ERα mechanistically ? Are flow-related changes in Akt phosphorylation in endothelium altered in the absence of ERα?

To respond to these questions, we have added several sets of new data.

We have added new data showing the level of phosphorylation of Akt in AF20ERα, C451A-ERα and R264-ERα mice (and the corresponding WT groups) (Figure 5, panels C, F and I + Figure 5 source data 2 showing all the blots for eNOS, Ph-eNOS, Akt and Ph-Akt).

In addition, we have added new data showing that mito-tempo which reduces ROS production by the mitochondria improved FMD in C451A-ERα mice (figure 7). In figure 7 we also added new data showing that the combination of superoxide dismutase and catalase also improved FMD in arteries from C451A-ERα mice (no effect in WT) and that catalase alone (elimination of H2O2 into H2O) reduced FMD in C451A-ERα mice without affecting FMD in WT.

Thus, we may state that ERα reduces mitochondria functioning and its production of ROS and H2O2. Indeed, several studies have shown that flow (shear stress) induces a more quiescent and less oxidative phenotype in endothelial cells (i.e.: Doddaballapur et al. Arterioscler Thromb Vasc Biol 2015, 35, (1), 137-45). This reduction in mitochondrial ROS production would thus allow NO to be more efficient (less ROS scavenging NO) and allow a more efficient eNOS activation. Of course, this later issue remains to be further investigated. Nevertheless, an excessive ROS production has been shown to reduce eNOS phosphorylation through increased phosphatase activation (Ding et al. Front Physiol 2020, 11, 566410).

In addition, the effect of catalase on FMD in C451A-ERα mice agrees with previous studies by D. Gutterman’s group showing that H2O2 produced by the mitochondria can dilate coronary arteries in response to flow (FMD) in arteries from patients with coronary artery disease. Although this allows keeping some FMD, the dilation remains low (as in C451A-ERα mice in the present study) and in the long term, H2O2 remains deleterious.

We have also modified the scheme shown in figure 9 to include these new data.

11) In the Results section, Do non-nuclear actions of plasma membrane-associated ERα influence endothelial cell production of prostaglandins in the setting of FMD?

We have added new data showing that the cyclooxygenase inhibitor indomethacin does not further reduce FMD in the 4 groups of mice studied: ERα-/-, AF20ERα, C451A-ERα and R264-ERα mice (and the corresponding WT groups) (new data added to Figure 4).

In addition, we observed no change in COX1, COX2 and prostacyclin expression level in mesenteric arteries isolated from C451A-ERα mice (Figure 7—figure supplement 2: J, K and L).

EDHF is the third major agent produced by the endothelium in response to flow and it is also involved in the activation of endothelium-dependent dilation by estrogens (reference 31 and 32). Thus, we have also added to the manuscript new data obtained with EETs blocker MSPPOH as EETs are major members of the EDHF family. EETs have been shown to mediate, at least in part, the protective effect of E2 on FMD in hypertensive or old rats (reference 31 and 32, introduction, lines 103-109).

This data (indomethacin and MSPPOH) is discussed lines 360-365.

In Figure 4 we show only L-NNA and indomethacin to avoid overloading the figure.

Figure 4 supplement figure 1 shows the 3 blockers, L-NNA, indomethacin and MSPPOH.

12) In the Results section, do non-nuclear actions of plasma membrane-associated ERα influence mechanosensitive ion channel localization or function in endothelial cells?

Based on the experiments described above, it is most likely that non-nuclear actions of plasma membrane associated ERα involve a reduction in oxidative stress and subsequently a better action of eNOS.

Nevertheless, it is important to decipher the link between flow (shear stress) activation of the extracellular matrix and cell surface and the plasma membrane associated ERα.

As flow has been recently shown to activate Piezo1-dependent release of ATP through pannexin hemi-channels followed by ATP-dependent activation of purinergic receptors which induce NO production by endothelial cells (Wang et al. J Clin Invest 2015, 125, (8), 3077-86; Wang et al., J Clin Invest 2016, 126, (12), 4527-4536), we investigated this pathway in C451A-ERα mice:

First, we found no change in the expression level of Piezo1 and Piezo2 channels in mesenteric arteries isolated from C451A-ERα mice. Similarly, no change in the level of polycystin1 and 2, TRPV4 or integrin α and β was found and no change in purinergic receptors level was found (Figure 7—figure supplement 1: panel G to N, Q, R and S).

Second, we have also performed additional experiments and added the new data to the manuscript showing that the acute response to ATP and to the piezo1 agonist YODA-1 was not affected by the absence of membrane ERα in C451A-ERa mice.

Similarly, the mechanosensitive channel blocker GsMTx4 similarly affected FMD in C451A-ERα and WT mice suggesting that the defect in FMD is probably located down-stream flow sensing. New data added to the discussion, lines 376-381 and shown in Figure 7—figure supplement 3, A, B and C.

As discussed above (point 11), we have also included data obtained with the EETs synthesis blocker MSPPOH. Indeed, as stated above, we used indomethacin to assess the role of prostaglandins in FMD in the mouse models used in the present work. We also used MSPPOH as EETs are produced by the endothelium and activate transient receptor potential (TRP) channel (Campbell WB, Fleming I. Epoxyeicosatrienoic acids and endothelium-dependent responses. Pflugers Arch. 2010 May;459(6):881-95). Nevertheless, this compound did not significantly change FMD after L-NNA and indomethacin blockade suggesting that this pathway and the related channels may not be involved (Figure 4 -, figure supplement 1, shown above in response to point 11).

13) In the Results section, why was the antioxidant TEMPOL specifically selected? Would the Authors expect a similar outcome following treatment with other antioxidants such as e.g. quercetin or dimethyl fumarate?

We used TEMPOL as we have already used it in the past with success. In addition, it is easy to use in the drinking water and mice keep drinking normally.

Nevertheless, we have added new data obtained with mice treated with vitamin C and vitamin E (4 weeks of treatment). This treatment has been shown efficient (Favre J et al. Coronary endothelial dysfunction after cardiomyocyte-specific mineralocorticoid receptor overexpression Am J Physiol Heart Circ Physiol. 2011 Jun;300(6):H2035-43) in reducing ROS.

This new set of experiments shows a similar result than with Tempol with a restoration of FMD in treated C451A-ERα mice treated with the antioxidant cocktail vitamin C and vitamin E. This new data is presented in figure 8 (G to L).

Importantly, we obtained a similar effect with 2 different unrelated antioxidant treatments.

14) In the Results section, does Tempol or PEG-catalase acutely restore FMD ex vivo?

We have added new data showing that the combination of superoxide dismutase and catalase restores FMD in C451A-ERa mice. By contrast, catalase which reduces H2O2 level improved FMD in C451A-ERa mice. This result supports the assumption that H2O2 may be responsible for the remaining FMD in C451A-ERa mice. This was not observed in WT mice. In addition, the effect of superoxide dismutase + catalase which reduces more globally the production of reactive oxygen species improved FMD in C451A-ERa mice. A similar result was obtained with mito-tempo which reduces mitochondrial ROS. This is discussed in more details above (point 10). New data shown in figure 7.

15) In the Results section, Line 312: Chronic for a 2-week treatment does not sound very appropriate.

We have removed the word chronic and left the duration of the treatment (2 weeks).

Of note, the new antioxidant treatment (vitamin E/vitamin C) added to the revised manuscript was supplied for a longer duration (4 weeks). This is now in figure 8.

16) In the Results section, flow rate may be changed to shear stress in Dyn/cm2: this would help comparing between published works in the topic.

We have added the corresponding shear stress in the Material and Methods and in the figure legends.

17) The Discussion section should be shortened, providing more succinct focused discussion of the interpretation of the findings, their implications, and possible explanations to fill the new knowledge gaps that result from the work.

We shortened the initial discussion. Nevertheless, we also added some text to respond to the comments raised above. Altogether, the discussion remains shorter and hopefully better focused on the results of the present work.

18) In the Conclusion section, Figure 7: the schematic is only in part helpful because how non-nuclear actions of membrane-associated ERα in endothelial cells govern eNOS and ROS in response to flow is not addressed.

We have modified this scheme (now Figure 9) and hope that it now better supports the conclusion of the discussion.

Reviewer #1:

[…] The mechanism of membrane receptor, ligand independent pathway of estrogen receptor alpha that the authors reported lacks novelty as it was already previously described.

We agree that ligand independent involvement of estrogen receptor alpha has been described in other fields, mainly in the field of breast cancer. Nevertheless, its involvement in the acute response to flow of resistance arteries is new and potentially of interest in the field of cardiovascular diseases.

Reviewer #2:

Multiple animal models have been used to test the contribution of membrane estrogen receptors on the vascular dilation induced by flow (flow-mediated dilation, FMD).

The authors propose that the presence of membrane estrogen receptors optimizes flow-mediated dilation. However, the conclusion that ER promotes NO production and inhibits oxidative stress is not fully supported by the data since basal ROS production is not altered in the models without functional ER, and an antioxidant treatment normalizes the dilatory response, and effect that can be independent of a direct NOS activation. The data show that in the absence of functional membrane receptors, flow-mediated dilation is reduced but that it is restored by a treatment with an antioxidant. Thus, these receptors seem not to be necessary for FMD. The involvement of these receptors in FMD remains therefore questionable.

We agree that the link between membrane ERα and eNOS activation by flow remains an open question. To address this question, we have added new data to the manuscript which should help better defining the relation between membrane ERα and the balance between eNOS activity and ROS.

First, we have added data showing that ROS inhibition is also active acutely and restored FMD in vitro in arteries isolated from mice lacking membrane-ERα (C451A-ERα mice) in agreement with the in vivo data obtained previously (initial version of the manuscript) showing that a 2-weeks long treatment with antioxidant TEMPOL restored FMD in C451A-ERα mice (we have also included another treatment with vitamins C and E, 4 weeks). This is also in agreement with our measurement of H2O2 in the perfusate of isolated kidneys from C451A-ERα mice. Thus, membrane-ERα is necessary to reduce ROS production, and this allows a better FMD. Therefore, FMD is reduced in the absence of membrane-ERα (no more break on ROS production) and ROS reduction restored FMD. Of course, this could exclude a direct involvement of membrane-ERα in flow-sensing and signal transduction to eNOS as you point out in your comment. To further address this question, we used blockers of mechanosensitive channels (similar effect on C451A-ERα and WT mice) and we tested Yoda1 and ATP-dependent dilation in C451A-ERα and WT mice (no difference).

Thus, it seems that flow activates on one hand the NO pathway and on the other hand flow reduces ROS production through activation of the membrane located ERα. This effect on ROS could involve a reduction in mitochondrial activity as we also found that the inhibition of mitochondrial ROS production with Mito-Tempo restored FMD in C451A-ERα mice (new data added to the manuscript). We have extended the discussion on this mechanism. Of course, the pathway linking flow to membrane-ERα and to the mitochondria remains to be further investigated.

Altogether, these data show that membrane-ERα is involved in the acute response to flow of resistance arteries through a reduction in ROS production.

We have changed the conclusion of the abstract as follows:

“Thus, endothelial membrane ERα promotes NO production through inhibition of oxidative stress and thereby helps to optimize FMD in a ligand-independent manner”

Reviewer #3:

[…] (1) Ignoring a possible role of GPER and signaling pathways thereof in the study endpoints is a major flaw in the experimental design.

We have tested the effect of G36 as suggested. Although G36 did not affect FMD in the present study, GPER is certainly an important player in the control of vascular tone in different conditions. Although GPER does not seem to have a role in FMD in the present work, its role in the pathophysiology of the vascular tree is now well recognized.

(2) The regulation of flow-mediated dilation by estrogen has been widely investigated in previous studies.

We agree that estrogen has major role in restoring FMD in many pathological (cardiovascular and metabolic disorders) or physiological (menopause) conditions. We have added more references on this effect of estrogen in the introduction (lines 96-111).

Nevertheless, the present study does not address the effect of estrogen on FMD. On the contrary, we show in the present study that membrane ERα is involved in FMD independently of its ligand. This effect was observed in healthy young mice, both in females and in males. Importantly, this effect is reminiscent of the first case of ERα gene deficiency in a young man: “The first disruptive mutation in the ERα gene, reported in 1994 in a man who was only 30 years old, was found to be associated with a selective and total absence of FMD. This single yet major clinical observation suggests that ERα-dependent signal transduction could play a role in FMD in males”.

Smith et al. N Engl J Med 1994, 331, (16), 1056-61.

Sudhir et al. Lancet 1997, 349, (9059), 1146-7.

3) The gender claim in the title is relatively weak as it is based on just one set of experiments (Figure 1; lines 137-138)

We agree that most of the experiments were conducted on male mice. Nevertheless, the main effect of the absence of membrane-ERα is a reduction of the amplitude of FMD in both male and female mice. This was observed in mesenteric arteries from intact and ovariectomized mice as well as in the uterine arteries of female mice. This later data was shown in the supplement files, and we moved it to figure 1. Nevertheless, we have removed “male and female mice” from the title.

Although the study objective was stated concisely, the Authors generated findings of potential impact in the field that deserve further investigation. A cross-talk between blood flow, NO pathway and membrane-associated ERα appears to emerge from the present work and represents a conceptual advance. However, the role of GPER in this setting deserves to be assessed as well.

We agree with your comment and have performed the requested experiments. We found that G36 did not affect FMD in the mouse mesenteric artery. We have added the data to figure 1.

Reviewer #4:

[…] One weakness is the evaluation of levels of ERα expression by quantifying transcript abundance in whole arteries when ERα protein abundance in endothelial cells is of prime importance. In addition, although alterations in NO and ROS are implicated, no insights are gained into how these are impacted by ERα presumably independent of estrogens. Despite the lack of more mechanistic interrogation, the overall observation of an important role for non-nuclear function of plasma membrane-associated ERα in endothelial cells in FMD is important to the field.

We have isolated endothelial cells from the aorta to quantify ERα gene expression (data in figure 3). We used the aorta as the quantity of cells that would be obtained from mesenteric resistance arteries would be too small and the isolation techniques used for the aorta does not apply to small resistance arteries. Although more cells are obtained from the aorta, the quantity of cells and consequently the quantity of total tissue and proteins obtained remains very limited. Analysis of the amount of protein in endothelial cells isolated from the mouse vessels is not feasible due to the insufficient amount of biological material combined to the low efficiency of ERα antibody in mice. However, we have adapted the protocol described by Iruela Arispe's team to analyze the expression of ERα in endothelial cells isolated from aorta by Q-PCR (Briot et al., Repression of Sox9 by Jag1 is continuously required to suppress the default chondrogenic fate of vascular smooth muscle cells. Dev Cell. 2014;31(6):707-721).

The results obtained demonstrated that there is no difference in the expression of ERα in the endothelial cells, whatever the genotype of the mice. Besides markers of endothelial cells (Tie2) and of smooth muscle cells (Cnn1) were quantified in order to control for the quality of the endothelial cells isolation. ERα is shown in figure 3. Tie2 (Tek) and Cnn1 are shown in the supplemental figures (Figure 3—figure supplement 1).

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Essential revisions:

1) The wording "optimization" in the title may be "optimized". The reviewer thinks that e.g. an experimental procedure but not a biological process such as FMD may be optimized. ER participates in or supports FMD.

As recommended, the title is now:

“Membrane estrogen receptor alpha (ERα) participates in flow-mediated dilation in a ligand-independent manner”.

We have also modified the last sentence of the abstract.

2) The p values and other statistical parameters could be more appropriately moved from figures to figure legends, especially when asterisks are present.

As recommended, p values are now given in the figure legends.

Data and full statistical analysis are also shown in the source data files.

3) qPCR experiments: "relative expression" may refer to either an housekeeping gene or a reference control sample. Please indicate the first option in figure legends.

As recommended, the text of the figure legend (figure 3) is now:

“Esr1 expression level in aortic endothelial cells (expression relative to the housekeeping genes Gapdh, Hprt and Gusb)”

4) Figure 5: Western blots should be shown for phospho-eNOS and total eNOS, and for phospho-Akt and total Akt.

As recommended, Western-blots (selected bands) are now shown in figure 5.

Western blots for all the mice used in the study are shown in the Figure 5 source data 2.

5) Line 63-65: The authors state that flow-mediated dilatation depends mainly on NO, citing studies looking at the forearm circulation in humans. This should be stated/clarified. However, in mice (the organism used for experiments in this study) flow-dependent vasodilatation in mesenteric resistance arteries (the vascular bed investigated in this study) is mediated by both, NO and EDH (PMID 16055522; 11282893). Brandes and associates also found that EDH is the main mediator of endothelium-dependent relaxation in murine resistance arteries (PMID 10944233). Moreover, the EDH-shear-dependent response is partly sensitive to non-selective COX-inhibition (PMID 16055522). The authors should discuss these important differences between humans and mice and also should mention the limitation of their study that experiments were conducted in the absence of COX inhibition, and that one cannot fully exclude the involvement of COX-dependent / endothelial-derived prostanoid effects in the effects observed.

We have modified the text as recommended.

More precisely:

a. Line 63-65: it is now stated that this sentence refers to human studies.

The sentence is now: “FMD measured in the human forearm depends mainly on the acute production of NO…” Line 64 now.

b. In mice FMD in mesenteric resistance arteries is mediated by both, NO and EDH:

We have added a paragraph in the discussion stating the limitations of the study as suggested by your comment. Line 365 to 374. As regard to COX-derivatives, we have added to figure 4 (in response to the previous comments) data obtained with indomethacin. Indeed, the addition of indomethacin to LNNA did not further reduced FMD and acetylcholine-mediated dilation. Similarly, the addition of MSPPOH which block EETs production, did not further reduced FMD and acetylcholine-mediated dilation (supplement figure to Figure 4). Although EETs are major component of the EDHF family, this result suggests that the remaining FMD in the presence of L-NNA, indomethacin and MSPPOH is due to other EDHFs.

6) Lines 112-115: With regard to estrogen effects in men the authors should discuss intracrine, aromatase-mediated production of estradiol which is converted locally in the vasculature from testosterone (PMID 11248122) by which estrogen partly provides protection in male mice from atherosclerosis.

We have added the following sentence in the introduction (now: lines 118-120)

“Of note, testosterone has been shown to reduce early atherogenesis in male mice through its conversion to estrogen by aromatase which is expressed in the arterial wall [42]”.

7) Lines 145-147: The authors suddenly introduce compounds targeting GPER, without having made any reference to this receptor in the introduction. It would be helpful if the authors could add a little section to the introduction discussing this receptor as well, also citing a good overview article, such as PMID 21844907

We have added a paragraph in the introduction stating the role of GPER in endothelium-dependent dilation (lines 87-90):

“E2-dependent vasodilation has also been shown to involve the G-protein-coupled estrogen receptor (GPER) in both human and animal arteries [24]. In the rat, GPER activation reduces uterine vascular tone during pregnancy through activation of endothelium-dependent NO production [25].”

8) Line 144, Line 465-466, : The authors state "the ERα-ERb antagonist and GPER agonist fulvestrant (ICI-182780)." This is partly correct. Fulvestrant is a SERD (selective estrogen receptor downregulator or selective estrogen receptor degrader), which down-regulates/degrades the receptors it targets. This should be corrected.

We have corrected the text: “the estrogen receptor downregulator and GPER agonist fulvestrant (ICI-182780)”. Now: line 149 and line 463

9) Line 262: findings for the evaluation of changes of gene expression with C451A-ERα are not shown in Figure 7.

These findings are shown in the supplement figures attached to Figure 7. This is written in the text as requested in the instruction to the authors: “Figure 7, figure supplement figures 1 and 2” These two supplemental figures were attached to the manuscript (merged files) after the figures (page 68 and 69).

10) Lines 335-336: the statement is misleading because chronic effects of E2 (recognizing that the term "chronic" is vague) can be mediated by non-nuclear actions of ERα.

We removed the sentence “The nuclear functions of ERα are mainly involved in the chronic effects of E2 and ERα activation such as atheroma prevention [7]”

Indeed, you are right, it is a misleading shortcut. In addition, the transition to flow-mediated remodeling is better now.

11) Lines 402-403: there is no basis for stating that the decrease in FMD in mice lacking membrane ERα could reflect a feature of premature aging of the endothelium.

We removed the paragraph. Aging is certainly more complex than a single reduction in membrane-ERα signaling even though it might contribute. A more in-depth investigation is needed before stating this.

12) Figure 9 schematic: there is no evidence that Gai mediates the role of membrane-associated ERα in FMD.

We removed Gai from the scheme.

13) P. 21: The reference for G36 provided in the table is not quite correct, the one cited was published to 2010, at a time when G36 had not yet been published. It is suggested to list PMID 27803283 as reference instead.

We apology for the error. We added the right reference to the table.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

The manuscript has been improved but there are some remaining issues that need to be addressed, as outlined below:

1) Previous comment "Lines 145-147: The authors suddenly introduce compounds targeting GPER, without having made any reference to this receptor in the introduction. It would be helpful if the authors could add a little section to the introduction discussing this receptor as well, also citing a good overview article, such as PMID 21844907".

Authors Response: We have added a paragraph in the introduction stating the role of GPER in endothelium-dependent dilation (lines 87-90):

"E2-dependent vasodilation has also been shown to involve the G-protein-coupled estrogen receptor (GPER) in both human and animal arteries [24]. In the rat, GPER activation reduces uterine vascular tone during pregnancy through activation of endothelium-dependent NO production [25]."

It appears that there was a misunderstanding with regard to the previous comment. It was not requested to describe solely the role of GPER in endothelial cell function and NO release. Rather, the authors were expected to expand this section, briefly describing the nature of GPER (namely that it is a 7-transmembrane GPCR located at the endoplasmic reticulum), its natural and synthetic ligands, and its functions (release of NO is just one). Also, the authors should mention that there is both, ligand-dependent and ligand-independent activation (PMID 27803283) of GPER, which should also be discussed in the Discussion section.

We apologize for misunderstanding your previous comment.

We have extended the paragraph on GPER in the introduction (lines 93-104):

“The 7-transmembrane G-protein-coupled estrogen receptor (GPER, formerly known as GPR30) is another receptor located not only at the plasma membrane but also on the membrane of the endoplasmic reticulum that can be activated by E2. […] Thus, both ERα and GPER could contribute to the rapid actions of E2, although their respective roles according to vessel type, species and pathophysiological context remain to be established.”

We have also extended the paragraph on GPER in the discussion including the various physiological functions involving GPER and the various diseases which could benefit from a targeting of GPER (lines 357-366):

“Another membrane receptor for E2 located at the plasma membrane is GPER [27]. Both ligand-dependent and ligand-independent activation of GPER have been reported [61]. […] However, a possible role of unliganded GPER activation cannot be excluded in case of a crosstalk between membrane-dependent Erα and GPER activation.”

2) The "Key Resources Table" in the Methods section includes a number of references:

Hypertension 2007;50:248-54

Hypertension. 2007;50:248-54

EMBO Mol Med. 2014;6:1328-46

Pharmacology 2010: 86, 58- 64

Sci Signal. 2016;9:ra105

J Vasc Res. 2009;46:2 53-64

Am J Physiol 2018;315:H1019-H1026

Cytokine 2009;46,166-70

These references should be added in full to the References section

References are now in full in the References Section.

3) The whole manuscript should be checked for grammar and typos, including "contribute to relaxation of the mesenteric resistance arteries in both male and female through, at least in part, a PI3K-Akt-eNOS pathway [28]." (lines 98-99, page 5) and corrections should be made as needed.

We have rechecked the manuscript.

https://doi.org/10.7554/eLife.68695.sa2

Article and author information

Author details

  1. Julie Favre

    Angers University, MITOVASC, CNRS UMR 6015, INSERM U1083, Angers, France
    Contribution
    Conceptualization, Data curation, Investigation, Methodology, Resources, Validation, Writing – original draft, Writing – review and editing
    Contributed equally with
    Emilie Vessieres
    Competing interests
    No competing interests declared
  2. Emilie Vessieres

    1. Angers University, MITOVASC, CNRS UMR 6015, INSERM U1083, Angers, France
    2. CARFI facility, Angers University, Angers, France
    Contribution
    Data curation, Formal analysis, Methodology
    Contributed equally with
    Julie Favre
    Competing interests
    No competing interests declared
  3. Anne-Laure Guihot

    1. Angers University, MITOVASC, CNRS UMR 6015, INSERM U1083, Angers, France
    2. CARFI facility, Angers University, Angers, France
    Contribution
    Investigation, Methodology
    Competing interests
    No competing interests declared
  4. Coralyne Proux

    1. Angers University, MITOVASC, CNRS UMR 6015, INSERM U1083, Angers, France
    2. CARFI facility, Angers University, Angers, France
    Contribution
    Investigation, Methodology
    Competing interests
    No competing interests declared
  5. Linda Grimaud

    Angers University, MITOVASC, CNRS UMR 6015, INSERM U1083, Angers, France
    Contribution
    Data curation, Methodology, Resources
    Competing interests
    No competing interests declared
  6. Jordan Rivron

    1. Angers University, MITOVASC, CNRS UMR 6015, INSERM U1083, Angers, France
    2. CARFI facility, Angers University, Angers, France
    Contribution
    Investigation, Methodology, Validation
    Competing interests
    No competing interests declared
  7. Manuela CL Garcia

    1. Angers University, MITOVASC, CNRS UMR 6015, INSERM U1083, Angers, France
    2. CARFI facility, Angers University, Angers, France
    Contribution
    Investigation, Validation
    Competing interests
    No competing interests declared
  8. Léa Réthoré

    Angers University, MITOVASC, CNRS UMR 6015, INSERM U1083, Angers, France
    Contribution
    Investigation
    Competing interests
    No competing interests declared
  9. Rana Zahreddine

    INSERM U1297, Paul Sabatier University (Toulouse III) , University Hospital (UHC) of Toulouse, Toulouse, France
    Contribution
    Investigation, Validation, Writing – original draft
    Competing interests
    No competing interests declared
  10. Morgane Davezac

    INSERM U1297, Paul Sabatier University (Toulouse III) , University Hospital (UHC) of Toulouse, Toulouse, France
    Contribution
    Data curation, Methodology, Performed experiments requested for the revised manuscript
    Competing interests
    No competing interests declared
  11. Chanaelle Fébrissy

    INSERM U1297, Paul Sabatier University (Toulouse III) , University Hospital (UHC) of Toulouse, Toulouse, France
    Contribution
    Investigation, Validation, Visualization
    Competing interests
    No competing interests declared
  12. Marine Adlanmerini

    INSERM U1297, Paul Sabatier University (Toulouse III) , University Hospital (UHC) of Toulouse, Toulouse, France
    Contribution
    Conceptualization, Investigation, Validation
    Competing interests
    No competing interests declared
  13. Laurent Loufrani

    1. Angers University, MITOVASC, CNRS UMR 6015, INSERM U1083, Angers, France
    2. University Hospital (CHU) of Angers, Angers, France
    Contribution
    Formal analysis, Investigation, Supervision, Validation, Writing – original draft
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-3397-2335
  14. Vincent Procaccio

    1. Angers University, MITOVASC, CNRS UMR 6015, INSERM U1083, Angers, France
    2. University Hospital (CHU) of Angers, Angers, France
    Contribution
    Methodology, Project administration, Validation
    Competing interests
    No competing interests declared
  15. Jean-Michel Foidart

    Groupe Interdisciplinaire de Génoprotéomique Appliquée, Université de Liège, Liège, Belgium
    Contribution
    Methodology, Resources
    Competing interests
    No competing interests declared
  16. Gilles Flouriot

    INSERM U1085, IRSET (Institut de Recherche en Santé, Environnement et Travail), University of Rennes, Rennes, France
    Contribution
    Conceptualization, Methodology, Validation
    Competing interests
    No competing interests declared
  17. Françoise Lenfant

    INSERM U1297, Paul Sabatier University (Toulouse III) , University Hospital (UHC) of Toulouse, Toulouse, France
    Contribution
    Conceptualization, Methodology, Supervision, Validation, Writing – original draft, Writing – review and editing
    Competing interests
    No competing interests declared
  18. Coralie Fontaine

    INSERM U1297, Paul Sabatier University (Toulouse III) , University Hospital (UHC) of Toulouse, Toulouse, France
    Contribution
    Methodology, Validation, Writing – original draft
    Competing interests
    No competing interests declared
  19. Jean-François Arnal

    INSERM U1297, Paul Sabatier University (Toulouse III) , University Hospital (UHC) of Toulouse, Toulouse, France
    Contribution
    Conceptualization, Funding acquisition, Writing – original draft
    Competing interests
    No competing interests declared
  20. Daniel Henrion

    1. Angers University, MITOVASC, CNRS UMR 6015, INSERM U1083, Angers, France
    2. CARFI facility, Angers University, Angers, France
    3. University Hospital (CHU) of Angers, Angers, France
    Contribution
    Conceptualization, Funding acquisition, Methodology, Project administration, Validation, Writing – original draft
    For correspondence
    daniel.henrion@univ-angers.fr
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-1094-0285

Funding

Fondation pour la Recherche Médicale (FRM - DPC20171138957)

  • Daniel Henrion

Fondation pour la Recherche Médicale (Equipe FRM DEQ20160334924)

  • Jean-François Arnal

Agence Nationale de la Recherche (ANR-18-CE14-0016-01)

  • Chanaelle Fébrissy
  • Jordan Rivron

Fondation Lefoulon Delalande

  • Julie Favre

Institut National de la Santé et de la Recherche Médicale

  • Daniel Henrion
  • Jean-François Arnal

Université de Toulouse

  • Jean-François Arnal

Fondation de France (00086486)

  • Jean-François Arnal

Région Occitanie Pyrénées-Méditerranée

  • Jean-François Arnal

Institut Universitaire de France

  • Jean-François Arnal

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

Esr2-/-, Esr1-/- and AF20ERα mice and the corresponding littermate WT mice were kindly provided by Prof. P Chambon and Dr. A Krust (Institute of Genetics and Molecular and Cellular Biology, Strasbourg, France). We thank J.M. Foidart and Mithra Pharma for providing Estetrol (E4). This work was supported in part by the foundation for Medical Research (Fondation pour la Recherche Médicale, Mitoshear project, contract FRM - DPC20171138957), the National Agency for Research (Agence Nationale de la Recherche, Estroshear project, contract # ANR-18-CE14-0016-01), INSERM, University of Toulouse III, the Fondation pour la Recherche Médicale (Equipe FRM DEQ20160334924), the Fondation de France (contract 00086486), the Région Occitanie, and the Institut Universitaire de France in Toulouse. JF was supported by the Lefoulon-Delalande Foundation. ChF and JR were supported by the National Agency for Research (Agence Nationale de la Recherche, Estroshear project, contract # ANR-18-CE14-0016-01).

Ethics

The investigation was conducted in accordance with the guidelines from Directive 2010/63/EU of the European Parliament for the protection of animals used for scientific purposes (authorization of the laboratory: # 00577). The protocol was approved by the Institutional Animal Care and Use Committee (IACUC): Committee on the Ethics of Animal Experiments (CEAA) of "Pays de la Loire" (permits #14335, #16740, and #16108). The mice were anesthetized with isoflurane (2.5%) and euthanized using a CO2 chamber and every effort was made to minimize suffering.

Senior Editor

  1. Matthias Barton, University of Zurich, Switzerland

Reviewing Editor

  1. Noriaki Emoto, Kobe Pharmaceutical University, Japan

Reviewer

  1. Philip W Shaul, University of Texas Southwestern Medical School, United States

Publication history

  1. Received: March 23, 2021
  2. Accepted: November 26, 2021
  3. Accepted Manuscript published: November 29, 2021 (version 1)
  4. Version of Record published: December 16, 2021 (version 2)

Copyright

© 2021, Favre et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 345
    Page views
  • 59
    Downloads
  • 0
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

Further reading

    1. Cell Biology
    Claudio Andrés Carril Pardo et al.
    Research Article

    Restoring damaged b-cells in diabetic patients by harnessing the plasticity of other pancreatic cells raises the questions of the efficiency of the process and of the functionality of the new Insulin-expressing cells. To overcome the weak regenerative capacity of mammals, we used regeneration-prone zebrafish to study b-cells arising following destruction. We show that most new insulin cells differ from the original b-cells as they coexpress Somatostatin and Insulin. These bihormonal cells are abundant, functional and able to normalize glycemia. Their formation in response to b-cell destruction is fast, efficient and age-independent. Bihormonal cells are transcriptionally close to a subset of d-cells that we identified in control islets and which are characterized by the expression of somatostatin 1.1 (sst1.1) and by genes essential for glucose-induced Insulin secretion in β-cells such as pdx1, slc2a2 and gck. We observed in vivo the conversion of monohormonal sst1.1-expressing cells to sst1.1+ ins+ bihormonal cells following b-cell destruction. Our findings support the conclusion that sst1.1 d-cells possess a pro-b identity enabling them to contribute to the neogenesis of Insulin-producing cells during regeneration. This work unveils that abundant and functional bihormonal cells benefit to diabetes recovery in zebrafish.

    1. Cell Biology
    2. Microbiology and Infectious Disease
    Jeffrey Y Lee et al.
    Research Article

    Despite an unprecedented global research effort on SARS-CoV-2, early replication events remain poorly understood. Given the clinical importance of emergent viral variants with increased transmission, there is an urgent need to understand the early stages of viral replication and transcription. We used single-molecule fluorescence in situ hybridisation (smFISH) to quantify positive sense RNA genomes with 95% detection efficiency, while simultaneously visualising negative sense genomes, subgenomic RNAs, and viral proteins. Our absolute quantification of viral RNAs and replication factories revealed that SARS-CoV-2 genomic RNA is long-lived after entry, suggesting that it avoids degradation by cellular nucleases. Moreover, we observed that SARS-CoV-2 replication is highly variable between cells, with only a small cell population displaying high burden of viral RNA. Unexpectedly, the B.1.1.7 variant, first identified in the UK, exhibits significantly slower replication kinetics than the Victoria strain, suggesting a novel mechanism contributing to its higher transmissibility with important clinical implications.