Investigation of Vascular Responses in Endothelial Nitric Oxide Synthase/Cyclooxygenase-1 Double-Knockout Mice: Key Role for Endothelium-Derived Hyperpolarizing Factor in the Regulation of Blood Pressure in Vivo Ramona S. Scotland, Melanie Madhani, Sharmila Chauhan, Salvador Moncada, Jørgen Andresen, Holger Nilsson, Adrian J. Hobbs and Amrita Ahluwalia Circulation 2005;111;796-803; originally published online Feb 7, 2005; DOI: 10.1161/01.CIR.0000155238.70797.4E Circulation is published by the American Heart Association. 7272 Greenville Avenue, Dallas, TX 72514 Copyright © 2005 American Heart Association. All rights reserved. Print ISSN: 0009-7322. Online ISSN: 1524-4539

The online version of this article, along with updated information and services, is located on the World Wide Web at: http://circ.ahajournals.org/cgi/content/full/111/6/796

Subscriptions: Information about subscribing to Circulation is online at http://circ.ahajournals.org/subscriptions/ Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Fax: 410-528-8550. E-mail: [email protected] Reprints: Information about reprints can be found online at http://www.lww.com/reprints

Downloaded from circ.ahajournals.org by on April 23, 2008

Vascular Medicine Investigation of Vascular Responses in Endothelial Nitric Oxide Synthase/Cyclooxygenase-1 Double-Knockout Mice Key Role for Endothelium-Derived Hyperpolarizing Factor in the Regulation of Blood Pressure in Vivo Ramona S. Scotland, PhD; Melanie Madhani, PhD; Sharmila Chauhan, PhD; Salvador Moncada, MD, PhD; Jørgen Andresen; Holger Nilsson, PhD; Adrian J. Hobbs, PhD; Amrita Ahluwalia, PhD Background—Endothelium-dependent dilatation is mediated by 3 principal vasodilators: nitric oxide (NO), prostacyclin (PGI2), and endothelium-derived hyperpolarizing factor (EDHF). To determine the relative contribution of these factors in endothelium-dependent relaxation, we have generated mice in which the enzymes required for endothelial NO and PGI2 production, endothelial NO synthase (eNOS) and cyclooxygenase-1 (COX-1), respectively, have been disrupted (eNOS⫺/⫺ and COX-1⫺/⫺ mice). Methods and Results—In female mice, the absence of eNOS and COX-1 had no effect on mean arterial blood pressure (BP), whereas BP was significantly elevated in eNOS⫺/⫺/COX-1⫺/⫺ males compared with wild-type controls. Additionally, endothelium-dependent relaxation remained intact in the resistance vessels of female mice and was associated with vascular smooth muscle hyperpolarization; however, these responses were profoundly suppressed in arteries of male eNOS⫺/⫺/COX-1⫺/⫺ animals. Similarly, the endothelium-dependent vasodilator bradykinin produced dose-dependent hypotension in female eNOS⫺/⫺/COX-1⫺/⫺ animals in vivo but had no effect on BP in male mice. Conclusions—These studies indicate that EDHF is the predominant endothelium-derived relaxing factor in female mice, whereas NO and PGI2 are the predominant mediators in male mice. Moreover, the gender-specific prevalence of EDHF appears to underlie the protection of female eNOS⫺/⫺/COX-1⫺/⫺ mice against hypertension. (Circulation. 2005;111: 796-803.) Key Words: endothelium 䡲 hypertension 䡲 nitric oxide

T

he release of paracrine vasodilators from endothelial cells is a critical determinant of vascular tone and, hence, an integral regulatory mechanism involved in the maintenance of local blood flow and systemic blood pressure (BP). Endothelium-dependent agonists such as acetylcholine (ACh) and bradykinin (BK), in addition to physical vascular stimuli such as shear stress, stimulate endothelial cells to release the dilators nitric oxide (NO), prostacyclin (PGI2), and endothelium-derived hyperpolarizing factor (EDHF).1–3 The relative contribution of each of these mediators to endotheliumdependent dilatation is inversely related to vessel caliber, such that NO- and PGI2-mediated responses predominate in conduit vessels, whereas EDHF is more prominent in resistance arteries.4 Because the latter vessels are the main contributors to peripheral resistance and organ perfusion,5 this pattern of reactivity intimates that EDHF is likely to play a pivotal role in vascular homeostasis.

See p 724 Despite this perceived importance, the identity of EDHF remains elusive, although several candidates have been proposed, including K⫹ ions,6 cAMP,7 cytochrome P450 2C products,8 H202,9 spread of electrotonic current,10 and, most recently, C-type natriuretic peptide (CNP).11 Because EDHF remains unidentified, selective antagonists of EDHF are lacking, and therefore the pathophysiological role of EDHF remains equivocal. As a consequence of this, its involvement in regulating vascular reactivity is defined as the response that persists in the presence of combined inhibition of NO and PGI2 synthesis (ie, by blocking NO synthase [NOS] and cyclooxygenase [COX], respectively). This definition, however, has proved problematic, because inhibitors of NOS and COX are not isoform selective, do not result in complete inhibition, and often have complicating inherent hemodynam-

Received June 21, 2004; revision received September 27, 2004; accepted October 5, 2004. From the Wolfson Institute for Biomedical Research (R.S.S., M.M., S.M., A.J.H.), University College London, London, United Kingdom; Clinical Pharmacology (S.C., A.A.), William Harvey Research Institute, London, United Kingdom; and the Institute of Physiology and Biophysics (J.A., H.N.), University of Aarhus, The Water and Salt Research Center, Aarhus, Denmark. Correspondence to Dr Amrita Ahluwalia, Clinical Pharmacology, William Harvey Research Institute, Barts and The London, Charterhouse Square, London EC1M 8BQ, UK. E-mail [email protected] © 2005 American Heart Association, Inc. Circulation is available at http://www.circulationaha.org

DOI: 10.1161/01.CIR.0000155238.70797.4E

796 Downloaded from circ.ahajournals.org by on April 23, 2008

Scotland et al

Gender-Specific EDHF in eNOS/COX-1–Knockout Mice

797

Figure 1. Genotyping of dKO mice. Electrophoretic analysis of polymerase chain reaction products derived from genomic DNA of female mice used in this study. Left, 300-bp band amplified from eNOS gene; 400-bp band amplified from NEO cassette. Right, 600-bp band amplified from COX-1 gene, 700-bp band amplified from NEO cassette. Lanes 1 and 5, WT; lanes 2 and 6, eNOS⫺/⫺; lanes 3 and 7, COX-1⫺/⫺; lanes 4 and 8, dKO.

ic effects in vivo. Moreover, NO and PGI2 can also elicit dilatation via hyperpolarization of vascular smooth muscle,12,13 which results in difficulties in separating the actions of these endothelium-derived vasodilators. Thus, the physiological significance of EDHF remains unsubstantiated, and its potential as a novel therapeutic target has yet to be explored. Herein, we describe the vascular phenotype of mice with targeted disruption of both the predominant endothelial isoforms of NOS (eNOS) and COX (COX-1) and demonstrate that this “EDHF mouse” is the ideal model with which to investigate the identity and physiological role of EDHF, both in vitro and in vivo. Moreover, using this model, we describe a novel, gender-specific, cardioprotective role for EDHF that may account for the relative protection of premenopausal females from cardiovascular disorders such as hypertension and atherosclerosis.

Methods Generation of eNOSⴚ/ⴚ/COX-1ⴚ/ⴚ Mice Animals with targeted disruption of the eNOS (eNOS⫺/⫺)14 or COX-1 (COX-1⫺/⫺)15 gene, both on a C57BL6 background, were mated to yield animals heterozygous for both loci (eNOS⫹/⫺/COX1⫹/⫺). These animals were then crossed with eNOS⫺/⫺ mice to yield eNOS⫺/⫺/COX-1⫹/⫺, which in turn were brother-sister mated to yield mice doubly deficient in both genes (eNOS⫺/⫺/COX-1⫺/⫺, or dKO). The offspring resulted in a 1:1 population of eNOS⫺/⫺/COX-1⫹/⫺ and dKO, which were genotyped by polymerase chain reaction (Figure 1). Litters from dKO⫻dKO matings were stillborn.

Genotyping DNA was extracted with a DNeasy tissue kit (Qiagen). The presence of eNOS, neomycin (NEO), or COX-1 was determined by polymerase chain reaction with the following primers (Figure 1): eNOS⫹/⫹ forward: 5⬘-GGTGTTTGGCTGCCGATGC-3⬘; NOS⫹/⫹ reverse: 5⬘-GCACAGCACACGGTGAACC-3⬘; NEO forward 5⬘-GCATACGCTTGATCCGGCTACC-3⬘ NEO reverse 5⬘-AAGGCGATGCGCTGCGAATC-3⬘; COX-1⫹/⫹ forward: 5⬘AGGAGATGGCTGCTGAGTTGG-3⬘; COX-1⫺/⫺ forward: 5⬘GCAGCCTCTGTTCCACATACAC-3⬘; COX-1 reverse: 5⬘-AATCTGACTTTCTGAGTTGCC-3⬘.

Resistance Artery Experiments Male and female mice, 8 to 12 weeks old (weight 20 to 27 g), were killed by neck dislocation; the mesentery was removed and placed in physiological salt solution (PSS) composed of (in mmol/L) NaCl 119, KCl 4.7, CaCl2.2 H2O 2.5, MgSO4 · 7 H2O 1.2, NaHCO3 25, KH2PO4 1.2, and glucose 5.5. Second-order arteries were mounted in a tension myograph (Danish Myotechnology) and bathed in PSS (37°C gassed with 5% CO2 in O2). After equilibration,16 vessels were stimulated with thromboxane A2 mimetic U46619 (1 ␮mol/L) until responses were reproducible. Vessels were precontracted with a submaximal concentration of U46619 (50% to 70% of response to 1 ␮mol/L), and the endothelium was stimulated by a single application of ACh (1 ␮mol/L) or BK (1 ␮mol/L). Full concentration-response

curves were not constructed because typically these small arteries do not hold a stable level of tone for prolonged periods.17 The contribution of NO and prostanoids was assessed with the NOS and COX inhibitors NG-nitro-L-arginine methyl ester (L-NAME; 300 ␮mol/L, 30 minutes) and indomethacin (5 ␮mol/L, 30 minutes,) respectively. Involvement of EDHF was determined with agents thought to block EDHF release and activity: KCl (30 mmol/L),18 apamin (100 nmol/L) and charybdotoxin (ChTx; 100 nmol/L),19 or Ba2⫹ (30 ␮mol/L) and ouabain (1 mmol/L).6 In some arteries, the endothelium was removed by passing a human hair twice through the lumen.20 In some experiments, tension and membrane potential were measured simultaneously as described previously.21 Small mesenteric arteries from female dKO mice were stimulated with a single application of ACh (10 ␮mol/L) either in the absence of or after precontraction with U46619 (10 to 20 nmol/L). Membrane potential was measured with aluminum silicate microelectrodes with a resistance of 50 to 100 M⍀, when filled with 3 mol/L KCl, and recorded on an Intra-676 amplifier (WPI Inc).

Conduit Artery Experiments Segments of thoracic aortas were mounted in organ baths under a basal tension of 0.3g. Relaxation response curves to ACh were constructed in phenylephrine (PE; 10 to 30 nmol/L, EC70)precontracted vessels in the absence or presence of L-NAME and indomethacin. At the end of each experiment, sodium nitroprusside (SNP; 1 ␮mol/L) was applied to test the integrity of the vascular smooth muscle.

Measurement of EDHF Responses In Vivo Female and male dKO mice were anesthetized with isoflurane (2%). The left common carotid artery was cannulated for measurement of arterial BP and the right femoral artery for drug administration. BK was used for examination of EDHF responses in vivo, because unlike ACh, it has little direct cardiac activity. BK (0.1 to 10 ␮g/kg) and SNP (3 ␮g/kg) were given as single bolus doses of 50 ␮L. Responses were calculated as the maximum change in BP minus the response to bolus injection of 50 ␮L of saline.

Measurement of BP in Conscious Mice Systolic BP (SBP) was measured by tail-cuff plethysmography (XBP 1007, Kent Scientific) in conscious animals in a heated room (26°C to 27°C). Mean arterial BP (MABP) in conscious female dKO and wild-type (WT) animals was also assessed with a tethered mouse model.22 After a 12-hour recovery period after surgery, MABP was recorded over the following 12 hours.

Plasma 6-Keto-Prostaglandin F1␣ Measurement Blood was collected from the tail vein and mixed 1:10 with 3.15% (vol/vol) citrate. Plasma PGI2 was assessed by measuring PGI2 metabolite 6-keto-prostaglandin F1␣ by enzyme immunoassay (Amersham Biosciences).

Materials and Solutions Drugs were purchased from Sigma, except U46619 (Biomol) and apamin and charybdotoxin (Alomone Labs). Stock solutions were made in water, except U46619 (ethanol), ouabain (DMSO), and indomethacin (5% NaHCO3).

Downloaded from circ.ahajournals.org by on April 23, 2008

798

Circulation

February 15, 2005

Comparison of Structural and Functional Properties of Mesenteric Resistance Arteries From Mice With Different Genotypes Male WT Diameter, ␮m

171⫾6.25

Basal tension, mN

⫺/⫺

Female ⫺/⫺

eNOS

COX-1

168⫾8.00

183⫾6.40

dKO

WT

162⫾8.30

181⫾7.60

eNOS

⫺/⫺

167⫾5.10

COX-1⫺/⫺

dKO

173⫾6.90

171⫾3.41

2.5⫾0.19

2.6⫾0.16

2.7⫾0.12

2.1⫾0.25

2.8⫾0.16

2.7⫾0.11

2.4⫾0.14

2.5⫾0.09

Contraction to 1 ␮mol/L U46619, mN

11.1⫾1.70

11.4⫾0.80

9.8⫾0.48

10.6⫾0.72

10.9⫾0.61

10.0⫾0.51

9.3⫾0.58

10.9⫾0.43

Relaxation to 1 ␮mol/L SNP, %

90.6⫾2.30

92.5⫾2.41

93.2⫾3.30

97.0⫾1.16

93.0⫾2.42

93.3⫾1.26

94.0⫾2.42

99.0⫾1.00

n⬎12. Data are mean⫾SEM.

Statistical Analysis Relaxation responses were calculated as percentage reversal of precontraction. Data shown are mean⫾SEM. Data were analyzed by 1- or 2-way ANOVA followed by Bonferroni multiple comparison test.

Results Resistance Artery Studies Endothelium-Dependent Relaxation in Arteries of WT Animals There was no significant difference in diameter or basal tension of arteries between the groups of mice (Table). Furthermore, the contractile response to U46619 (1 ␮mol/L) and the relaxant response to the endothelium-independent

vasodilator SNP (1 ␮mol/L) were similar between groups. Together, these data suggest no gross morphological differences between arteries of the different genotypes. To compare endothelial function between genotypes, we assessed responses to ACh. ACh caused relaxation of 84.3⫾5.1% (n⫽4; Figure 2A) in mesenteric resistance arteries of female WT mice. Pharmacological blockade of NOS or COX individually did not alter this response (n⫽4, P⬎0.05; data not shown); however, combined NOS/COX inhibition attenuated ACh-induced relaxation by 29.3⫾4.0% (n⫽4; Figure 2A). ACh-induced relaxation in arteries from male WT mice was not significantly different from female vessels

Figure 2. EDHF is predominant in female but not male resistance arteries. A, Relaxation to ACh in absence and presence of endothelium, inhibitors of NO and PGI2, or inhibitors of EDHF (apamin/ChTx, or K⫹, or Ba2⫹/ouabain). *P⬍0.05 vs control (n⫽4 to 8). B, Relaxation of arteries of male WT or COX-1⫺/⫺ animals is abolished by inhibition of NO synthesis. In arteries of eNOS⫺/⫺ animals, responses to ACh are abolished by inhibition of PGI2 release. ACh produces weak relaxation of arteries from male dKO mice (n⫽4 to 7). C, ACh-induced relaxation of dKO mesenteric resistance arteries is associated with change in membrane potential. D, Endothelium-dependent relaxation of female aortas is mediated exclusively by NO. Application of ACh produces concentration-dependent relaxation of aortic rings from WT and COX⫺/⫺ that is completely blocked by NO inhibition. Aortas that lack eNOS (eNOS⫺/⫺ or dKO) do not relax to ACh (n⫽4 to 7).

Downloaded from circ.ahajournals.org by on April 23, 2008

Scotland et al

Gender-Specific EDHF in eNOS/COX-1–Knockout Mice

(71.4⫾5.1%; n⫽14, P⬎0.05 versus female arteries) but was abolished by NOS inhibition (Figure 2B). Effect of eNOS Gene Disruption on Vascular Responses ACh caused relaxation of 42.8⫾4.8% (n⫽5; Figure 2A) in mesenteric resistance arteries of female eNOS⫺/⫺ mice. This response was unaffected by L-NAME or indomethacin treatment either alone or in combination (Figure 2A). Block of EDHF release and activity, using raised [K⫹] or Ba2⫹/ ouabain, respectively, abolished relaxations to ACh in these arteries. In contrast, ACh produced only weak relaxation of mesenteric arteries from male eNOS⫺/⫺ animals (26.1⫾4.7%; n⫽7; Figure 2B). This response was not affected by treatment with L-NAME but was markedly attenuated by indomethacin. Effect of COX-1 Gene Disruption on Vascular Responses ACh caused relaxation of 58.1⫾14.1% (n⫽4; Figure 2A) in mesenteric resistance arteries of female COX-1⫺/⫺ animals. These responses were resistant to inhibition of NOS and COX but markedly attenuated by inhibitors of EDHF (ie, raised [K⫹] or Ba2⫹/ouabain). In contrast, ACh-induced relaxation of male COX-1⫺/⫺ mesenteric arteries (45.5⫾11.9%, n⫽5; Figure 2B) was abolished by L-NAME (Figure 2B). Vascular Responses in dKO Mice ACh caused relaxation of 59.6⫾6.8% (n⫽13; Figure 2A) in mesenteric resistance arteries of female dKO mice. In accord with observations in female eNOS⫺/⫺ and COX-1⫺/⫺ arteries, ACh evoked responses were unaffected by combined NOS/ COX blockade and were mediated exclusively by EDHF, because raised [K⫹], ChTx/apamin, or Ba2⫹/ouabain profoundly suppressed these responses (Figure 2A). BK (1 ␮mol/L) also relaxed these arteries, a response that was unaffected by combined L-NAME and indomethacin treatment (54.8⫾10.2% and 50.8⫾13.0%, respectively; n⫽4 for both; P⬍0.05 versus control). In contrast, ACh caused comparatively weak relaxations in mesenteric arteries of male dKO mice (Figure 2B) that were L-NAME and/or indomethacin resistant (data not shown). A relaxant response to BK was not observed in these arteries (data not shown). Electrophysiology Basal membrane potential of arteries of female DKO mice was ⫺55⫾1.6 mV (n⫽6). In the absence of tone, ACh produced a hyperpolarization of ⫺10⫾2.7 mV (n⫽3). Simultaneous recording of membrane potential and isometric tension showed that contraction to U46619 (⬇EC50 10 to 20nmol/L) was associated with a depolarization of 14⫾3.8 mV. Application of ACh produced a repolarization of 9⫾3 mV (n⫽3; Figure 2C).

Conduit Artery Studies Relaxation to ACh in aortas of male or female WT and COX-1⫺/⫺ animals were entirely mediated by NO (n⫽4 to 7; see Figure 2D for female data). Aortas from male or female eNOS⫺/⫺ and dKO animals did not relax in response to ACh (n⫽4 to 7; see Figure 2D for female data). There were no gender differences in the responses to ACh in aortas (male data not shown).

799

Assessment of EDHF Responses In Vivo and Correlation to Systemic BP BK produced a dose-dependent decrease in BP in anesthetized female dKO mice (n⫽4; Figures 3A and 3C) but had little or no effect in male dKO animals (n⫽3; Figures 3B and C); however, SNP (3 ␮g/kg) produced similar hypotension in female (⫺29.2⫾0.7 mm Hg, n⫽4) and male (⫺24.0⫾4.84 mm Hg, n⫽3; P⬎0.05) dKO mice. There was no significant difference in SBP, measured by tail-cuff plethysmography, between different genotypes of female mice compared with WT (Figure 4A). In marked contrast, SBP was significantly elevated in male eNOS⫺/⫺ and dKO animals, respectively. There was no difference in SBP between male COX-1⫺/⫺ and WT animals. In addition, continuous BP measurement in vivo in conscious animals demonstrated that MABP in female dKO and WT mice was not significantly different (Figure 4B).

Plasma 6-Keto-Prostaglandin F1␣ Measurement

6-Keto-PGF1␣ was substantially elevated in male eNOS⫺/⫺ compared with WT mice (Figure 5); however, these enhanced levels were absent in male dKO mice. In contrast, 6-ketoPGF1␣ levels were equivalent to WT in female eNOS⫺/⫺ animals. However, 6-keto-PGF1␣ levels were significantly lower in COX-1⫺/⫺ and dKO animals, which confirms that COX-1 is the source of PGI2 in these animals and that there is no upregulation of COX-2 to compensate in both males and females. Western blot analysis of mesenteric arteries from animals of each genotype and gender confirmed a lack of COX-2 expression (data not shown).

Discussion Herein, we demonstrate a crucial role for EDHF in maintaining normal levels of BP and highlight a potential cardioprotective effect of this factor in females. These novel physiological roles for EDHF are supported by several observations. The targeted disruption of both eNOS and COX-1 genes results in elevated BP of male, but not female, mice. Moreover, significant vasodilator endothelial function, determined both in vitro and in vivo, remains in female, but not male, dKO animals. Hence, the generation of this “EDHF mouse” has created an ideal model to characterize the identity and biological actions of EDHF in vivo and in vitro (without the need for pharmacological inhibitors of NOS and COX) and revealed novel, gender-specific cardioprotective functions of this mediator. Further characterization of the vascular physiology and pathology of this animal is likely to identify new therapeutic targets to treat cardiovascular disease.

EDHF Is the Predominant Endothelium-Derived Relaxing Factor in Female Mice Using specific gene-knockout technology, we were able to directly assess the involvement of eNOS and COX-1 in endothelium-dependent relaxation of resistance arteries. The results of this study show a clear distinction between the mediators involved in female and male arteries. In female WT mesenteric small arteries, combined NOS and COX inhibition suppressed ACh-induced relaxation by only 30%, which indicates a predominant role for EDHF in this response.

Downloaded from circ.ahajournals.org by on April 23, 2008

800

Circulation

February 15, 2005

Figure 3. BK reduces BP in female but not male dKO mice in vivo. Original trace depicting typical responses to bolus injections of BK in (A) female and (B) male dKO mice. C, Mean dosedependent decrease in BP to BK in female (n⫽4) and male dKO mice (n⫽3). **Significantly different (P⬍0.01) from female.

Disruption of either eNOS or COX-1, or both genes, resulted in a residual EDHF-dependent relaxation of between 50% and 90% of that in WT animals. In addition, BK also caused NO- and PGI2-independent relaxation of female dKO arteries, which implicates EDHF as the common mediator of endothelium-dependent dilatation in these mice. Confirmation of EDHF as the mediator of these non-NO, nonprostanoid responses was achieved with classic inhibitors of EDHF release and activity. A definitive feature of EDHFmediated responses is the sensitivity to combined inhibition

of endothelial SKCa (with apamin) and IKCa (with charybdotoxin), which is thought to inhibit EDHF responses by blocking the passage of EDHF from the generating endothelial cell to the target underlying smooth muscle.19 Pretreatment with these K⫹ channel blockers, or mechanical removal of the endothelium, virtually abolished relaxation to ACh in female dKO mesenteric arteries. Similarly, elevation of extracellular [K⫹], which nonselectively blocks K⫹ channel activity,18 eliminated non-NO, nonprostanoid relaxations in female dKO arteries, which demonstrates the crucial role of

Figure 4. Female mice are not hypertensive. A, Measurement of systolic BP by tail-cuff plethysmography in conscious female and male mice (n⫽6 to 9). *Significantly different (P⬍0.01) from corresponding WT and female. B, Measurement in vivo of MABP in conscious female dKO and WT mice (n⫽5).

Downloaded from circ.ahajournals.org by on April 23, 2008

Scotland et al

Gender-Specific EDHF in eNOS/COX-1–Knockout Mice

Figure 5. COX-2 does not compensate for loss of COX-1. Plasma PGI2 levels (measured as 6-keto PGF1␣ by enzyme immunoassay) are reduced in female and male COX-1⫺/⫺ and dKO mice (n⫽6 to 9). **P⬍0.01 vs WT and #P⬍0.05 vs female.

hyperpolarization in these responses. Indeed, direct measurement of membrane potential demonstrated that ACh hyperpolarizes smooth muscle of arteries from female dKO animals in the absence of a constrictor and that relaxations of arteries contracted (and depolarized) with U46619 occur simultaneously with a repolarization of membrane potential. Moreover, concomitant blockade of KIR and Na⫹/K⫹ATPase by Ba2⫹ and ouabain, respectively (believed to underlie EDHF-mediated hyperpolarization of vascular smooth muscle6,23), blocked relaxant responses to ACh in female arteries of all genotypes. Together, these findings demonstrate that relaxant responses to endotheliumdependent stimuli (ie, ACh, BK) in female resistance arteries are dependent on EDHF and involve smooth muscle hyperpolarization.

Male Mice Exhibit Dramatically Less EDHF Activity Than Female Mice Unlike resistance arteries of female mice, NO and PGI2 are the predominant endothelium-derived dilators in resistance arteries of male mice. In male WT and COX-1⫺/⫺ animals, NO accounts almost entirely for endothelium-dependent relaxation to ACh. The finding that endothelium-dependent relaxation of eNOS⫺/⫺ arteries is markedly suppressed (⬇75%) highlights the pivotal role of NO in male arteries. Coupled with the fact that responses to ACh were suppressed in COX-1⫺/⫺ male arteries (⬇45%) compared with WT arteries, these data suggest that in male arteries, only weak compensatory mechanisms exist. In clear contrast, the absolute capacity of resistance arteries of females to relax to endothelium-dependent stimuli, irrespective of genotype, is largely maintained. It is thought that one of the physiological roles of EDHF is to function as a salvage pathway in conditions of insufficient NO release and therefore that EDHF activity is only evident in the presence of NOS inhibition24; however, in male eNOS⫺/⫺ mesenteric arteries, in which NO-mediated responses are absent, no EDHF activity is observed. Rather, COX activity partially compensates for the loss of NO, as demonstrated by the sensitivity of ACh-induced responses to indomethacin. This upregulation of COX activity is likely to be a general phenomenon in arteries of male eNOS⫺/⫺ mice,

801

because relaxation of coronary25 or superior mesenteric arteries17 to ACh and flow-mediated dilation of skeletal muscle arterioles26 are also sensitive to COX inhibition. Indeed, in the present study, assay of plasma 6-keto-PGF1␣, a stable PGI2 metabolite, demonstrated an elevation in the levels of this product in male eNOS⫺/⫺ mice. Western blot analysis of arteries taken from these animals showed no COX-2 expression, which suggests that the enhanced PGI2 synthesis likely relates to elevated COX-1 rather than COX-2 activity. Because COX-1 activity appears to be upregulated in male eNOS⫺/⫺ mice, we speculated that in the absence of both eNOS and COX-1, EDHF activity would be induced or upregulated to normalize vasomotion. Surprisingly, ACh produced negligible relaxation of male dKO arteries. This difference between endothelium-derived mediators in male and female mice suggests that sex hormones are likely to be important for EDHF activity. Indeed, there is evidence hinting at a link between estrogen and EDHF activity. EDHF-mediated responses to ACh,9,27 shear stress,28 ADP,29 or Ca2⫹ ionophore30 are all reduced by ovariectomy; however, E2 is known to target several signal transduction pathways in the vasculature, including the muscarinic receptor M2 and M3 pathways,31 which suggests that alternative mechanisms may be responsible at least for the changes in EDHF activity in response to ACh. It is unlikely that this latter mechanism is at play in the present study, because BK also produced relaxation of dKO arteries that persisted in the presence of NOS and COX inhibition, which demonstrates that EDHFmediated relaxation is a general phenomenon of these arteries and is not due to an increase in sensitivity to ACh. Mechanisms by which estrogens might increase EDHF responses are unclear, but possibilities include induction or upregulation of an EDHF synthase or enhancement of EDHF signal-transduction pathways. Communication between endothelial and vascular smooth muscle layers via myoendothelial gap junctions has been implicated as an integral part of EDHF-mediated signaling.10,32 Interestingly, there is evidence that E2 enhances the expression of connexin-43, one of the connexin protein components that form the gap junctions of resistance arteries, in ovariectomized rats.30 Such changes may be one process by which estrogen stimulates changes in EDHF activity. In addition, we have recently demonstrated that CNP is an EDHF in mesenteric resistance arteries11; estrogen enhances transcription of CNP in mouse uterus,33 and relaxations to CNP in female porcine coronary artery are greater than in males,34 which suggests that increased CNP release/activity may account for enhanced EDHF activity in female mice. Further studies investigating the effects of ovariectomy on vascular responses in these DKO mice are required to more fully examine the mechanisms of the gender-specific EDHF response. An alternative explanation for the apparent gender differences may be related to an effect of male sex hormones rather than estrogen. Recent evidence demonstrated that testosterone suppresses EDHF responses in rat middle cerebral arteries.35 Whether it is the presence of estrogen upregulating EDHF activity or the absence of a suppressive influence of testosterone on EDHF in the female DKO mice is uncertain and warrants further investigation.

Downloaded from circ.ahajournals.org by on April 23, 2008

802

Circulation

February 15, 2005

In contrast to mesenteric resistance arteries, there is no gender difference in relaxations of aortic rings to ACh. In agreement with previous studies9,14,17 we found that NO release accounts for ACh-induced relaxation of thoracic aortas (ie, responses in WT or COX-1⫺/⫺ are abolished by NOS inhibition, and responses are absent in eNOS⫺/⫺ and dKO mice). These observations are consistent with a fundamental role of EDHF in resistance arteries,4 whereas NO is the sole mediator of endothelium-dependent relaxation in conduit arteries.

Female dKO Mice Are Not Hypertensive A substantial body of evidence demonstrates a lower incidence of hypertension and heart disease among premenopausal women36 than among age-matched men. This difference has been attributed to an effect of estrogens, although this theory has come under considerable scrutiny of late and is highly controversial.37 Nevertheless, evidence in a number of animal models supports the concept that the BP-lowering effects of estrogen are due in part to its ability to stimulate eNOS-derived NO release (for review, see Mendelsohn and Karas38); whether NO is solely responsible for this specific protection is uncertain. Transgenic mice that lack the estrogen receptor [ER]-␤, which is more prevalent in females than males,39 are hypertensive40; however, evidence supports a predominant role for ER-␣ in estrogen-mediated NO signal transduction.41,42 Together, these data suggest that the effects of estrogen on BP may not be entirely dependent on NO release. Indeed, the BP-lowering effects of E2 are unaltered by targeted disruption of eNOS.43 In the present study, noninvasive measurement of SBP demonstrated no differences between female mice of different genotypes. This observation is consistent with the thesis that the BP-lowering effects of gender are independent of NO release. The BP of male and female COX-1⫺/⫺ mice were similar but not significantly different from WT despite a tendency toward lower BP, as reported previously.44 Additional loss of functional COX-1 protein in dKO animals has no further effect on BP compared with eNOS⫺/⫺ mice, which confirms that release of COX-1 products has minimal impact on BP regulation under basal conditions. Although there appeared to be a tendency for SBP to be higher in female dKO mice, as measured by tail cuff plethysmography, this was not significantly different from female WT. Indeed, in vivo measurements of MABP in conscious animals confirmed that female dKOs are normotensive. This finding demonstrates an important role for EDHF in regulating systemic BP. Moreover, because cardiovascular risk is reduced in females over males, and the present studies demonstrate that EDHF activity is prevalent in female animals but largely absent in males, it is likely that EDHF (rather than NO) activity contributes to gender-specific protection apparent in cardiovascular diseases, including hypertension.

EDHF In Vivo Although there is some evidence that EDHF contributes to regulation of blood flow in vivo,45– 48 extensive in vivo characterization of the physiological role of EDHF has not been possible to date because of the complicating effects of

the agents used to explore EDHF responses. The use of currently available tools for isolation of EDHF responses in vivo necessitates NOS and COX inhibition, which can have profound hemodynamic effects per se, which complicates measurement of EDHF activity. Furthermore, the toxin combinations that block EDHF release or activity, although useful in vitro, have many in vivo actions unrelated to EDHF block. In the present study, we have circumvented the need to use pharmacological inhibitors by creating eNOS/COX-1 dKO mice and thereby have, for the first time, permitted characterization of the physiological role of EDHF in vivo. Administration of BK in female dKO animals produced a dosedependent decrease in MABP. In contrast, BK had little or no effect on MABP in male dKO animals. Such an observation represents the first definitive demonstration of an EDHFevoked change in BP in vivo and indicates that EDHF is an important modulator of vascular tone physiologically. Moreover, the present studies show a clear EDHF-mediated regulation of BP apparent in females but not males.

Implications for Study of EDHF and Novel Therapeutic Targets By generating animals that lack both eNOS and COX-1, ie, the EDHF mouse, we have produced animals that display substantial in vitro and in vivo EDHF activity. These animals are an invaluable new tool for the study of the nature and actions of EDHF. The results of the present study demonstrate a clear link between gender and EDHF-mediated responses in resistance arteries. Stimulation of endothelial cells of female arteries, regardless of their genotype, resulted in significant EDHF activity. Moreover, this ability to release EDHF appears to be integral to BP regulation. Together, these findings suggest that EDHF may contribute to the lower incidence of cardiovascular disease in premenopausal women.

Acknowledgments This work is funded by the Wellcome Trust. We thank Dr Derek Gilroy for assistance with 6-keto-PGF1␣ assays.

References 1. Garland CJ, Plane F, Kemp BK Cocks TM. Endothelium-dependent hyperpolarization: a role in the control of vascular tone. Trends Pharmacol Sci. 1995;16:23–30. 2. Komori K, Lorenz RR, Vanhoutte PM. Nitric oxide, ACh, and electrical and mechanical properties of canine arterial smooth muscle. Am J Physiol. 1988;255:H207–H212. 3. Taylor SG, Weston AH. Endothelium-derived hyperpolarizing factor: a new endogenous inhibitor from the vascular endothelium. Trends Pharmacol Sci. 1988;9:272–274. 4. Shimokawa H, Yasutake H, Fujii K, Owada MK, Nakaike R, Fukumoto Y, Takayanagi T, Nagao T, Egashira, K, Fujishima M, Takeshita A. The importance of the hyperpolarizing mechanism increases as the vessel size decreases in endothelium-dependent relaxations in rat mesenteric circulation. J Cardiovasc Pharmacol. 1996;28:703–711. 5. Mulvany MJ, Aalkjaer C. Structure and function of small arteries. Physiol Rev. 1990;70:921–961. 6. Edwards G, Dora KA, Gardener MJ, Garland CJ, Weston AH. K⫹ is an endothelium-derived hyperpolarizing factor in rat arteries. Nature. 1998; 19:396:269 –272. 7. Griffith TM, Chaytor AT, Taylor HJ, Giddings BD, Edwards DH. cAMP facilitates EDHF-type relaxations in conduit arteries by enhancing electrotonic conduction via gap junctions. Proc Natl Acad Sci U S A. 2002; 99:6392– 6397.

Downloaded from circ.ahajournals.org by on April 23, 2008

Scotland et al

Gender-Specific EDHF in eNOS/COX-1–Knockout Mice

8. Fisslthaler B, Popp R, Kiss L, Potente M, Harder DR, Fleming I, Busse R. Cytochrome P450 2C is an EDHF synthase in coronary arteries. Nature. 1999;401:493– 497. 9. Matoba T, Shimokawa H, Nakashima M, Hirakawa Y, Mukai Y, Hirano K, Kanaide H, Takeshita A. Hydrogen peroxide is an endotheliumderived hyperpolarizing factor in mice. J Clin Invest. 2000;106: 1521–1530. 10. Sandow SL, Hill CE. Incidence of myoendothelial gap junctions in the proximal and distal mesenteric arteries of the rat is suggestive of a role in endothelium-derived hyperpolarizing factor-mediated responses. Circ Res. 2000;86:341–346. 11. Chauhan SD, Nilsson H, Ahluwalia A, Hobbs AJ. Release of C-type natriuretic peptide accounts for the biological activity of endotheliumderived hyperpolarizing factor. Proc Natl Acad Sci U S A. 2003;100: 1426 –1431. 12. Corriu C, Feletou M, Edwards G, Weston AH, Vanhoutte PM. Differential effects of prostacyclin and iloprost in the isolated carotid artery of the guinea-pig. Eur J Pharmacol. 2001;426:89 –94. 13. Cohen RA, Plane F, Najibi S, Huk I, Malinski T, Garland CJ. Nitric oxide is the mediator of both endothelium-dependent relaxation and hyperpolarization of the rabbit carotid artery. Proc Natl Acad Sci U S A. 1997; 94:4193– 4198. 14. Huang PL, Huang Z, Mashimo H, Bloch KD, Moskowitz MA, Bevan JA, Fishman MC. Hypertension in mice lacking the gene for endothelial nitric oxide synthase. Nature. 1995;377:239 –242. 15. Langenbach R, Morham SG, Tiano HF, Loftin CD, Ghanayem BI, Chulada PC, Mahler JF, Lee CA, Goulding EH, Kluckman KD, Kim HS, Smithies O. Prostaglandin synthase 1 gene disruption in mice reduces arachidonic acid–induced inflammation and indomethacin-induced gastric ulceration. Cell. 1995;83:483– 492. 16. Mulvany MJ, Halpern W. Contractile properties of small arterial resistance vessels in spontaneously hypertensive and normotensive rats. Circ Res. 1977;41:19 –26. 17. Chataigneau T, Feletou M, Huang PL, Fishman MC, Duhault J, Vanhoutte PM. Acetylcholine-induced relaxation in blood vessels from endothelial nitric oxide synthase knockout mice. Br J Pharmacol. 1999; 126:219 –226. 18. Adeagbo AS, Triggle CR. Varying extracellular [K⫹]: a functional approach to separating EDHF- and EDNO-related mechanisms in perfused rat mesenteric arterial bed. J Cardiovasc Pharmacol. 1993;21: 423– 429. 19. Petersson J, Zygmunt PM, Hogestatt ED. Characterization of the potassium channels involved in EDHF-mediated relaxation in cerebral arteries. Br J Pharmacol. 1997;120:1344 –1350. 20. Kakuyama M, Vallance P, Ahluwalia A. Endothelium-dependent sensory NANC vasodilatation: involvement of ATP, CGRP and a possible NO store. Br J Pharmacol. 1998;123:310 –316. 21. Mulvany MJ, Nilsson H, Flatman JA. Role of membrane potential in the response of rat small mesenteric arteries to exogenous noradrenaline stimulation. J Physiol. 1982;332:363–373. 22. Rees DD, Monkhouse JE, Cambridge D, Moncada S. Nitric oxide and the haemodynamic profile of endotoxin shock in the conscious mouse. Br J Pharmacol. 1998;124:540 –546. 23. Busse R, Edwards G, Feletou M, Fleming I, Vanhoutte PM, Weston AH. EDHF: bringing the concepts together. Trends Pharmacol Sci. 2002;23: 374 –380. 24. Bauersachs J, Popp R, Hecker M, Sauer E, Fleming I, Busse R. Nitric oxide attenuates the release of endothelium-derived hyperpolarizing factor. Circulation. 1996;94:3341–3347. 25. Godecke A, Decking UK, Ding Z, Hirchenhain J, Bidmon HJ, Godecke S, Schrader J. Coronary hemodynamics in endothelial NO synthase knockout mice. Circ Res. 1998;82:186 –194. 26. Sun D, Huang A, Smith CJ, Stackpole CJ, Connetta JA, Shesely EG, Koller A, Kaley G. Enhanced release of prostaglandins contributes to flow-induced arteriolar dilation in eNOS knockout mice. Circ Res. 1999; 85:288 –293. 27. Liu MY, Hattori Y, Fukao M, Sato A, Sakuma I, Kanno M. Alterations in EDHF-mediated hyperpolarization and relaxation in mesenteric arteries of female rats in long-term deficiency of oestrogen and during oestrus cycle. Br J Pharmacol. 2001;132:1035–1046.

803

28. Huang A, Wu Y, Sun D, Koller A, Kaley G. Effect of estrogen on flow-induced dilation in NO deficiency: role of prostaglandins and EDHF. J Appl Physiol. 2001;91:2561–2566. 29. Xu HL, Santizo RA, Baughman VL, Pellegrino DA. ADP-induced pial arteriolar dilation in ovariectomized rats involves gap junctional communication. Am J Physiol Heart Circ Physiol. 2002;283:H1082–H1091. 30. Liu MY, Hattori Y, Sato A, Ichikawa R, Zhang XH, Sakuma I. Ovariectomy attenuates hyperpolarization and relaxation mediated by endothelium-derived hyperpolarizing factor in female rat mesenteric artery: a concomitant decrease in connexin-43 expression. J Cardiovasc Pharmacol. 2002;40:938 –948. 31. Abdalla FM, Abreu LC, Porto CS. Effect of estrogen on intracellular signaling pathways linked to activation of M(2)- and M(3)-muscarinic acetylcholine receptors in the rat myometrium. Mol Cell Endocrinol. 2000;160:17–24. 32. Hutcheson IR, Chaytor AT, Evans W, Griffith TM. Nitric oxideindependent relaxations to acetylcholine and A23187 involve different routes of heterocellular communication: role of gap junctions and phospholipase A2. Circ Res. 1999;84:53– 63. 33. Acuff CG, Huang H, Steinhelper ME. Estradiol induces C-type natriuretic peptide gene expression in mouse uterus. Am J Physiol. 1997;273: H2672–H2677. 34. Barber DA, Burnett JC Jr, Fitzpatrick LA, Sieck GC, Miller VM. Gender and relaxation to C-type natriuretic peptide in porcine coronary arteries. J Cardiovasc Pharmacol. 1998;32:5–11. 35. Gonzales RJ, Krause DN, Duckles SP. Testosterone suppresses endothelium-dependent dilation of rat middle cerebral arteries. Am J Physiol Heart Circ Physiol. 2004;286:H552–H560. 36. Reckelhoff JF. Gender differences in the regulation of blood pressure. Hypertension. 2001;37:1199 –1208. 37. Couzin J. Estrogen research: the great estrogen conundrum. Science. 2003;302:1136 –1138. 38. Mendelsohn ME, Karas RH. The protective effects of estrogen on the cardiovascular system. N Engl J Med. 1999;340:1801–1811. 39. Hodges YK, Tung L, Yan X, Graham JD, Horwitz KB, Horwitz LD. Estrogen receptors alpha and beta: prevalence of estrogen receptor beta mRNA in human vascular smooth muscle and transcriptional effects. Circulation. 2000;101:1792–1798. 40. Zhu Y, Bian Z, Lu P, Karas RH, Bao L, Cox D, Hodgin J, Shaul PW, Thoren P, Smithies O, Gustafsson JA, Mendelsohn ME. Abnormal vascular function and hypertension in mice deficient in estrogen receptor beta. Science. 2002;295:505–508. 41. Rubanyi GM, Freay AD, Kauser K, Sukovich D, Burton G, Lubahn DB, Couse JF, Curtis SW, Korach KS. Vascular estrogen receptors and endothelium-derived nitric oxide production in the mouse aorta: gender difference and effect of estrogen receptor gene disruption. J Clin Invest. 1997;99:2429 –2437. 42. Darblade B, Pendaries C, Krust A, Dupont S, Fouque MJ, Rami J, Chambon P, Bayard F, Arnal JF. Estradiol alters nitric oxide production in the mouse aorta through the alpha-, but not beta-, estrogen receptor. Circ Res. 2002;90:413– 419. 43. Hodgin JB, Knowles JW, Kim HS, Smithies O, Maeda N. Interactions between endothelial nitric oxide synthase and sex hormones in vascular protection in mice. J Clin Invest. 2002;109:541–548. 44. Athirakul K, Kim HS, Audoly LP, Smithies O, Coffman TM. Deficiency of COX-1 causes natriuresis and enhanced sensitivity to ACE inhibition. Kidney Int. 2001;60:2324 –2329. 45. Halcox JP, Narayanan S, Cramer-Joyce L, Mincemoyer R, Quyyumi AA. Characterization of endothelium-derived hyperpolarizing factor in the human forearm microcirculation. Am J Physiol Heart Circ Physiol. 2001; 280:H2470 –H2477. 46. Brandes RP, Schmitz-Winnenthal FH, Feletou M, Godecke A, Huang PL. Vanhoutte PM, Fleming I, Busse R. An endothelium-derived hyperpolarizing factor distinct from NO and prostacyclin is a major endothelium-dependent vasodilator in resistance vessels of wild-type and endothelial NO synthase knockout mice. Proc Natl Acad Sci U S A. 2000;97: 9747–9752. 47. Nishikawa Y, Stepp DW, Chilian WM. In vivo location and mechanism of EDHF-mediated vasodilation in canine coronary microcirculation. Am J Physiol. 1999;277:H1252–H1259. 48. De Vriese AS, Van de Voorde J, Lameire NH. Effects of connexinmimetic peptides on nitric oxide synthase– and cyclooxygenaseindependent renal vasodilation. Kidney Int. 2002;61:177–185.

Downloaded from circ.ahajournals.org by on April 23, 2008