Coronary Hemodynamics in Endothelial NO Synthase Knockout Mice
Abstract—For the specific analysis of endothelial NO synthase (eNOS) function in the coronary vasculature, we generated a mouse homozygous for a defective eNOS gene (eNOS−/−). Western blot as well as immunohistochemical staining revealed the absence of eNOS protein in eNOS−/− mice. Aortic endothelial cells derived from eNOS−/− mice displayed only background levels of NOx formation compared with wild-type (WT) cells (88 versus 1990 pmol NOx · h−1/mg protein−1). eNOS−/− mice were hypertensive (mean arterial pressure, 135±15 versus 107±8 mm Hg in WT) without the development of cardiac hypertrophy. Coronary hemodynamics, analyzed in Langendorff-perfused hearts, showed no differences either in basal coronary flow or in maximal and repayment flow of reactive hyperemia. Acute NOS inhibition with Nω-nitro-l-arginine methyl ester (L-NAME) in WT hearts substantially reduced basal flow and reactive hyperemia. The coronary response to acetylcholine (ACh) (500 nmol/L) was biphasic: An initial vasoconstriction (flow, −35%) in WT hearts was followed by sustained vasodilation (+190%). L-NAME significantly reduced vasodilation in WT hearts (+125%) but did not alter the initial vasoconstriction. In eNOS−/− hearts, the initial vasoconstriction was augmented (−70%), whereas the ACh-induced vasodilation was not affected. Inhibition of cyclooxygenase with diclofenac converted the ACh-induced vasodilation into vasoconstriction (−49% decrease of basal flow). This effect was even more pronounced in eNOS−/− hearts (−71%). Our results demonstrate that (1) acute inhibition of eNOS reveals a role for NO in setting the basal coronary vascular tone as well as participation in reactive hyperemia and the response to ACh; (2) chronic inhibition of NO formation in eNOS−/− mutant mice induces no changes in basal coronary flow and reactive hyperemia, suggesting the activation of important compensatory mechanisms; and (3) prostaglandins are the main mediators of the ACh-induced vasodilation in both WT and eNOS−/− mice.
Endothelial NO synthase, also called type III NO synthase, is the major NOS isoenzyme that is widely expressed in endothelial cells throughout the vascular bed. It is generally accepted that endothelium-derived NO is an important factor in the control of basal vascular tone.1 NO is also involved in receptor-mediated vasodilation in response to various agonists such as ACh, ATP, thrombin, bradykinin, and others. Through experiments using NOS inhibitors2 3 but also by use of genetically modified animals,4 5 it has been demonstrated that functional inactivation of eNOS activity results in hypertension. In addition to the control of vascular tone, NO inhibits platelet aggregation and leukocyte adhesion to the vessel wall as well as proliferation and migration of smooth muscle cells. Thus, eNOS is considered to play an important role in maintaining the antiatherogenic surface of the vessel wall.6
In the heart, eNOS is expressed primarily in the coronary and endocardial endothelia. In addition, eNOS has been localized to cardiac myocytes and the specialized cells of sinus and AV nodes.7 Here, eNOS is considered to be part of an intrinsic regulatory pathway that may control myocardial contractility in response to changes in heart rate and to mediate the accentuated antagonism of cholinergic stimulation. Neuronal (type I) NOS may also contribute to cardiac NO formation and was found to be localized in intracardiac neurons lining coronary vessels and in cells of the sinus and AV node.7
Although more than one cell type in the heart can synthesize NO, the endothelium represents the major source of NO regulating coronary blood flow. Most studies indicate that basal formation of NO exerts a tonic dilator influence in the coronary circulation.8 Pharmacological inhibition of NO synthesis reduces coronary flow in a variety of species, although this effect is quite variable.9 10 11 Shear stress is a major regulator of endothelial NO formation. Acute changes in shear stress regulate eNOS activity via intracellular pH and tyrosine phosphorylation.12 In addition, chronic changes lead to modulation of eNOS expression via shear stress response elements in the eNOS promoter.13 Thus, shear stress may be a key determinant linking alterations in coronary flow to changes in vascular resistance via modulation of eNOS formation.
In addition to the regulation of basal coronary tone, NO may also play a role in metabolic vasodilation in the heart.14 Vascular resistance in reactive hyperemia15 16 or hypoxia17 has been shown to be sensitive to NOS inhibition. Moreover, it has been postulated that the adenosine-induced vasodilation is mediated in part by NO.18
To analyze the role of eNOS-derived NO on cardiovascular function, we constructed a knockout mouse with a disrupted eNOS gene. Here we report the first functional studies on coronary flow regulation in response to short-term ischemia and ACh-mediated vasodilation in a genetically defined model of eNOS deficiency.
Materials and Methods
Cloning of the Murine eNOS Gene and Construction of the Targeting Vector
The murine eNOS gene was cloned from a 129/Sv mouse genomic library established in bacteriophage λ DashII (Stratagene). Clones (1.8×106) were screened with a 389-bp cDNA fragment of bovine endothelial NOS that was amplified from reverse-transcribed RNA from bovine aortic endothelial cells with the oligonucleotides 5′-TACGTGCCCTGCATCCTC-3′ (forward primer) and 5′-CCCTCTGTCGCCAAGATG-3′ (reverse primer). This region encodes the NADPH binding site of eNOS, which is essential for NOS activity. The exon structure of the 3′-end of eNOS was determined by sequencing 4.5 kb of genomic DNA. Exons 19 to 26 (numbering corresponding to the human eNOS gene13 ) were identified in this region. Exon-intron boundaries were identical to those identified in the human eNOS gene. Sequence comparison of the assembled coding sequences revealed 87% identity with the human eNOS cDNA and 64% with the human neuronal NOS. For the construction of the targeting vector, a 0.8-kb EcoRI fragment containing exon 24 and the 5′ part of exon 25 was replaced by the neomycin resistance gene (neoR). This results in deletion of the adenine nucleotide binding site of the NADPH binding site and of other sequences implicated in NADPH binding located in exon 24.19 The targeting vector was constructed by placing a 0.8-kb EcoRI–Sca I fragment (short arm) on the 5′ end and an 8-kb EcoRI fragment (long arm) 3′ to the neoR gene.20 The long arm was flanked by a herpes simplex thymidine kinase gene.20
ES Cell Culture and Electroporation
E14–1 cells, a subclone of the E14 ES cell line, were grown as described21 on mitomycin-inactivated embryonic fibroblast feeder cells in medium containing 1000 U/mL leukemia inhibitory factor (Life Technologies). For electroporation, 107 cells were transfected with 50 μg of linearized targeting vector by electroporation (Bio-Rad Gene Pulser set to 240 V, 500 μF). Cells were plated onto neomycin-resistant feeder cells and double-selected for G418 (350 μg/mL) and gancyclovir (2 μmol/L). Clones were isolated on day 7 to 10 after transfection and analyzed by PCR for proper targeting (eNOS primer 5′-CTGAGGACTGCACCTGTTCA-3′, neo primer 5′-GGAGAGGCTTTTTGCTTCCT-3′ for 40 cycles of 1 minute at 94°C, 2 minutes at 55°C, and 2 minutes at 72°C). The genomic structure of the targeted eNOS locus was verified on ES cell and murine tail DNA by Southern blotting.
Targeted ES cells were microinjected into the blastocoel cavity of 3.5-day-old mouse blastocysts of strain C57BL/6. Injected blastocysts were implanted into pseudopregnant foster mothers (strain C57BL/6×CBA F1). Male chimeric mice were crossed to C57BL/6 females to test for germ line transmission. F1 heterozygous mutated animals were intercrossed to generate homozygous eNOS−/− mice. WT siblings obtained in these crosses were used as control animals in the phenotypical analysis of the eNOS mutation.
Western Blotting and Immunohistochemistry
Protein extracts were prepared by homogenization of mouse tissues in 50 mmol/L Tris (pH 7.5), 1 mmol/L EDTA, and 1 mmol/L PMSF. Extracts were centrifuged for 10 minutes at 10 000 rpm. Supernatant protein (200 μg) was separated on 7.5% SDS polyacrylamide gels and electroblotted to nitrocellulose. eNOS was detected by monoclonal antibodies directed against the C-terminal end of human eNOS (Transduction Laboratories) according to the manufacturer’s instructions. Protein bands were detected by horseradish peroxidase–labeled goat anti-mouse antibodies (Sigma) by use of the enhanced chemiluminescence detection system (Amersham).
For immunohistochemistry, organs were rapidly excised and frozen in OTC reagent (Miles). Sections (10 μm) were cut on a cryostat and processed as described previously.22 In brief, sections were mounted onto siliconized slides and fixed by immersion into 4% paraformaldehyde in 0.1 mol/L PB (pH 7.4) for 10 minutes at 4°C. Sections were rinsed twice (10 minutes) in PB, followed by the inhibition of endogenous peroxidase with 2% H2O2 in PB for 30 minutes. Sections were rinsed twice in PB and twice, 10 minutes each, in TBS (50 mmol/L Tris-HCl, pH 7.4, 0.9% NaCl). Nonspecific binding was blocked by incubation in TBS containing 10% normal goat serum. For immunostaining, sections were incubated with primary polyclonal rabbit anti-eNOS antibodies directed against the C-terminal end of human eNOS (Transduction Laboratories) diluted 1:100 with TBS/1% normal goat serum for 24 hours at 4°C. Sections were rinsed three times for 5 minutes in TBS, and binding of primary antibody was visualized with a VectaStain anti-rabbit ABC kit (PK4001, Vector Laboratories) using 3′,3′-diaminobenzidine as a chromogene. Sections were briefly air-dried and counterstained with methyl green.
Blood Pressure Measurements
Mice were anesthetized with urethane (1.5 g/kg IP). The throat was incised, and a catheter of stretched PE tubing filled with PBS containing 50 U heparin/mL was inserted into the exposed right carotid artery. Pulsatile blood pressure was recorded on a Gould recorder. Data were recorded on an analog plotter and transferred simultaneously to a personal computer (one measurement per second). Mean blood pressures were calculated from the data collected for 10 minutes starting 30 minutes after onset of anesthesia. Animals were kept on a 37°C warming plate during data collection.
Isolation of Aortic Endothelial Cells and Determination of Nitrite/Nitrate Formation
For the isolation of aortic endothelial cells, aortas were prepared from WT and eNOS−/− mice, opened longitudinally, and placed with the intima onto collagen-coated six-well plates (type III collagen, Sigma Chemical Co). After 3 days, aortic tissues were removed, and endothelial cells that had emigrated from the vessel segments were grown to confluency (DMEM, 10% FCS, 1000 U endothelial cell growth factor/mL; Sigma). Cells were split 1:4, and eNOS expression was induced by shear stress applied by shaking of the plates for 6 hours at 120 rpm on a round shaker. Then, growth media were replaced by Krebs-Henseleit buffer, and cells were incubated for 3 hours at 37°C. NOx accumulating in the culture supernatant was detected by chemiluminescence with an NO analyzer (NOA280, Sievers Inc) after reduction of NOx to NO with VCl3 according to the manufacturer’s protocol.
For the determination of plasma nitrite+nitrate levels, mice were fed with an NO2−/NO3−–free diet (Altromin). Blood was collected from ether-anesthetized animals by heart punctation. Blood coagulation was inhibited by the addition of 5 μL 0.5 mol/L EDTA. Blood cells were sedimented by centrifugation in an Eppendorf microfuge (5000 rpm), and the plasma supernatant was collected. Samples were deproteinized by the addition of a twofold volume of ice-cold ethanol and centrifugation at 13 000 rpm for 30 minutes. Fifty-microliter aliquots were used for the determination of the NOx− concentration with the NO analyzer as described above.
Isolated Mouse Hearts
Mice were injected with 250 U heparin IP and anesthetized with urethane (1.5 g/kg). The hearts were rapidly excised and transferred for preparation of the aortic trunk to warm, oxygenated Krebs-Henseleit buffer. The aorta was cannulated, and hearts were perfused in a nonrecirculating Langendorff mode at constant pressure (7 kPa, ie, 70 cm H2O) with a modified Krebs-Henseleit buffer containing (in mmol/L) NaCl 116, KCl 4.6, MgSO4 1.1, NaHCO3 24.9, CaCl2 2.5, KH2PO4 1.2, glucose 10, and EDTA 0.5 equilibrated with 95% O2 and 5% CO2 (pH 7.4, 37°C). Hearts were allowed to equilibrate for 30 to 40 minutes until a constant coronary flow was reached. Coronary flow was measured with a transit-time ultrasonic flowmeter located above the aortic cannula (Transonics). Acetylcholine and adenosine were infused into the aortic cannula for 180 to 300 seconds at infusion rates ranging from 0.2% to 1% of coronary flow.
Reactive hyperemia was elicited by interrupting coronary perfusion for 20 seconds. Maximal flow after reperfusion was recorded, and total flow, which represents the area under the flow curve, was determined by collecting and weighing the coronary effluent during the hyperemic flow, which lasted for ≈2 minutes. Adenosine release into the coronary effluent was determined by HPLC analysis essentially as described.23
Disruption of the eNOS Gene by Gene Targeting
To inactivate the eNOS gene, sequences coding for the essential NADPH binding site (exons 24 and 25) were deleted (Fig 1⇓). Expression of the disrupted gene from exons 1 to 23 can be expected to result in the formation of an inactive enzyme because the NADPH-adenine-nucleotide binding site as well as other sequences implicated in NADPH binding are encoded by exon 24,19 as well as by exon 26.24 Exons 24 and 25 are located on a 0.8-kb EcoRI fragment, which was replaced by the neomycin resistance gene in the targeting vector pKO4 (Fig 1A⇓). The murine ES cell line E14–1,21 which is derived from the mouse strain 129/Ola,25 was transfected with linearized targeting vector. Correctly targeted cells were identified by PCR, and the genomic structure was verified by Southern analysis. Three of 120 clones screened were correctly targeted and microinjected into C57BL/6 blastocysts. Male chimeras were crossed to C57BL/6 females to generate 129/Ola×C57BL/6 F1-eNOS+/− heterozygous offspring. These animals were subsequently bred to homozygosity. Analysis of the genomic structure of the eNOS locus of all three genotypes of the F2− offspring revealed that the eNOS locus was properly targeted and that the eNOS exons 24 and 25 had been deleted in genomic DNA of eNOS−/− mice (Fig 1B⇓).
Characterization of the eNOS Defect in eNOS−/− Mice
Western blot analysis was performed with monoclonal antibodies directed against the C-terminal end of human eNOS. In heart, lung, kidney, and liver of WT mice, eNOS protein was detected (Fig 2⇓). However, in knockout mice, the 135-kD band, representing the full-length enzyme, was absent, which verified a correct targeting on the protein level.
Immunohistochemistry performed with the same antibody revealed intense staining of capillaries and small arteries in hearts (Fig 3a⇓ and 3c⇓) and thymus (Fig 3e⇓) of WT animals. The staining was confined to endothelial cells. At the given level of sensitivity, eNOS expression could not be identified in cardiac myocytes. In contrast, endothelial cell layers in knockout mice did not react with the antibodies used (Fig 3b⇓, 3d⇓, and 3f⇓). High eNOS expression was also found in bronchial epithelia of WT mice but not in eNOS−/− animals (data not shown). Brain sections, stained for NADPH-diaphorase activity, a cytochemical marker for NOS activity, revealed the absence of endothelial staining in eNOS−/− mice, whereas WT controls were shown to contain stained intimal cell layers (data not shown).
To directly demonstrate the functional inactivation of eNOS, aortic endothelial cells were prepared and nitrite/nitrate accumulation in the culture supernatant was determined (Fig 4⇓). WT endothelial cells were found to produce 1990±402 pmol NOx · h−1 · mg protein−1. This activity could be inhibited by L-NAME (100 μmol · L−1) by 89% to 231±34 pmol NOx · h−1 · mg protein−1. In eNOS−/− cells, only background levels of nitrite accumulation (88±16 pmol NOx · h−1 · mg protein−1) were detected, suggesting complete inactivation of the eNOS.
It was further analyzed whether loss of eNOS function influenced the plasma nitrite+nitrate levels, the physiological oxidation products of NO. NO2/NO3 levels amounted to 1.6±0.55 μmol/L in plasma from WT and 0.74±0.07 μmol/L in eNOS−/− mice (n=4 in each group).
Influence on Blood Pressure and Heart Rate
Pulsatile blood pressures were measured in WT and eNOS−/− mice under urethane anesthesia by a stretched PE 10 catheter inserted into the right carotid artery. Mean blood pressure was determined for 10 minutes starting 30 minutes after onset of anesthesia. WT mice displayed pressures of 107±8 mm Hg (±SD), whereas in eNOS−/− mice, blood pressure was elevated to 135±15 mm Hg (n=6 in each group, P<.001). Heart rate was determined in resting awake animals by means of ECG: whereas in WT mice, heart rate was 545±79 bpm, this value was significantly reduced to 494±81 bpm (n=9; P<.05) in eNOS−/− mice.
Analysis of Coronary Hemodynamics
To analyze the influence of the eNOS mutation on coronary hemodynamics, we compared flow responses in isolated Langendorff-perfused hearts from eNOS−/− mice with those in hearts from WT mice with and without NO synthase inhibition by L-NAME.
From the data summarized in the Table⇓, it can be seen that L-NAME significantly reduced basal coronary flow in WT mice by 28%. In eNOS−/− hearts, however, basal coronary flow was not different from WT hearts. No differences in heart rate and in coronary venous release of adenosine were found between WT and eNOS−/− mice. The normalized heart weight was also unchanged.
In view of the proposed role of NO in reactive hyperemia,15 16 the flow response to short-term ischemia was analyzed. From the mean flow curves shown in Fig 5⇓ and the summarized data shown in Fig 6⇓, it is evident that inhibition of NOS activity in WT hearts by L-NAME reduced peak flow by 20% and repayment flow by 33% after 20 seconds of coronary occlusion. In contrast, the respective flow responses of WT and eNOS−/− hearts revealed only minor differences compared with the WT heart (Fig 5⇓). eNOS−/− hearts displayed a slightly higher maximal flow, but the flow repayment was not significantly different from WT hearts.
Adenosine release into the coronary venous effluents during the hyperemic phase amounted to 323.15±91.51 (SD, n=4) and 333.63±216.97 (SD, n=5) pmol · min−1 · g wet wt in the WT and eNOS−/− mice, respectively. In addition, the coronary response to adenosine, an endothelium-independent vasodilator, proved not to be different between the two groups when applied at concentrations of 250 nmol/L and 1 μmol/L (data not shown).
Intracoronary infusion of ACh (500 nmol/L) induced a biphasic response in mouse hearts. As shown in the representative recordings (Fig 7⇓), an initial vasoconstriction was followed by substantial vasodilation in WT hearts. In eNOS−/− hearts, the initial flow decrease was much more pronounced than in WT hearts and was also of longer duration, whereas the extent of vasodilation was not affected. Comparison of the averaged data (Fig 8⇓) revealed a flow decrease of −35% in WT and of −71% in eNOS−/− hearts. The subsequent ACh-induced vasodilation reached the same level as in WT hearts. Acute inhibition of NOS activity by L-NAME in WT hearts significantly reduced (−36%) steady-state vasodilation but left the initial vasoconstriction unaltered (Fig 8⇓).
Because the ACh-induced vasodilation was unaffected in eNOS−/− hearts, we tested whether prostaglandins contributed to the flow response to ACh in the mouse heart. Inhibition of cyclooxygenase with diclofenac (2 μmol · L−1) did not affect basal coronary flow in WT or eNOS−/− hearts. However, as shown in the representative recording in Fig 9⇓, diclofenac abolished the dilatory action and ACh infusion caused sustained vasoconstriction (flow, −49%) (Fig 10⇓). In eNOS−/− hearts, the vasoconstriction was even more pronounced and resulted in a flow decrease by −71%. Also note that cessation of ACh infusion was followed by a substantial increase in coronary flow. This hyperemic flow response is most likely due to the ACh-induced hypoperfusion in the presence of diclofenac.
A knockout mouse with a defective gene for eNOS was constructed for a specific analysis of eNOS function in the cardiovascular system. We decided to disrupt the NADPH binding site because the NOS reaction depends strictly on this cofactor. Exons 24 and 25 contain sequences crucial for NADPH binding. A cysteine corresponding to Cys-1048 of human eNOS is located in exon 24. A point mutation of this amino acid leads to a 90% decrease in enzymatic activity.19 Exon 25 encodes a conserved sequence motif, which binds to the adenine moiety of NADPH. Within this sequence is located another conserved cysteine (Cys 1114), which also leads to a drastic decrease of eNOS activity when mutated.19 DNA sequences encoding these elements were deleted from the mouse genome. In addition, because of the insertion of the neomycin resistance gene, the C-terminal end of eNOS encoded by exon 26 can be assumed to be absent from a truncated protein synthesized from exons 1 to 23. Recently, Xie et al24 demonstrated that this exon contains a sequence element that is highly conserved among all known NOS isoforms. Functional analysis using inducible NOS revealed that this conserved sequence constitutes an essential part of the NADPH binding site in addition to the NADPH ribose and NADPH adenine binding sites identified before. Deletion of this element resulted in a complete loss of enzymatic activity.
Western analysis of protein extracts from different organs as well as immunocytochemistry using antibodies directed against the deleted C-terminal end of eNOS demonstrated eNOS expression in WT but not in eNOS−/− mice. Because our knockout strategy involved disruption of the 3′ end of the eNOS gene, the expression of a truncated protein encoded by exons 1 to 23 cannot be ruled out.
Several lines of evidence demonstrate that the eNOS was functionally inactivated in our mice. (1) Cultured aortic endothelial cells from eNOS−/− mice displayed only background levels of NOx formation. (2) NADPH diaphorase staining, which depends on functional NOS, was absent from the endothelia of eNOS−/− mice. (3) eNOS−/− mice developed hypertension. (4) Plasma NO2−/NO3− levels were reduced by >50% in eNOS−/− mice, suggesting that eNOS activity contributes significantly to the total NO release.
In agreement with earlier reports by Huang et al4 and Shesely et al,5 who also demonstrated the development of hypertension for independently generated eNOS knockout mice, mean arterial blood pressure measured under urethane anesthesia in the present study was increased by 28 mm Hg, from 107 to 135 mm Hg, compared with WT mice. A similar increase was found by Huang et al,4 who reported blood pressures of 81 and 110 mm Hg for their WT and eNOS−/− mice, respectively, measured under the same anesthesia.
Several studies in hypertensive patients have demonstrated an impaired vasodilation in response to cholinergic stimulation, implying that a defective l-arginine–NO pathway might be involved.26 27 However, controversy still exists as to whether this impairment is causative for or the consequence of hypertension. Of interest is a recent observation that offspring of essential hypertensive patients that were still normotensive were characterized by a defect in the NO pathway. This finding suggests that a genetic predisposition to hypertension might arise from a defective NO pathway.28 Thus, eNOS−/− mice, now generated in three different laboratories, may represent a clinically relevant model system for the study of a genetically defined hypertension.
Surprisingly, despite hypertension, we found no cardiac hypertrophy in eNOS−/− mice. This is in contrast to studies in which long-term inhibition of NOS activity in the rat resulted in doubling of the cardiac weight index after 8 weeks of L-NAME administration.3 Comparison of these results is difficult, however, because L-NAME inhibits all NOS isoforms. Whether lack of hypertrophy in eNOS−/− mice is due to NO-related changes in trophic factors remains to be elucidated.
In accordance with the results obtained by Shesely et al,5 we found a reduced heart rate in awake eNOS−/− mice. Whether this bradycardia is a result of reflectory changes due to blood pressure elevation or involves a specific chronotropic action is at present unclear.
Coronary vasomotion is regulated by a large number of vasoactive factors, including NO, prostacyclin, endothelium-derived hyperpolarization factor, adenosine, endothelin, angiotensin II, K+, and others.29 Acute inhibition of NOS activity revealed a contribution of basal NO release to the control of coronary vascular tone in most species analyzed, although the extent varied widely.9 10 11 Similarly, L-NAME decreased the basal coronary flow by 28% in WT mice. Thus, NO contributes substantially to the setting of the basal coronary vascular tone in mice as well. In eNOS−/− hearts, however, basal coronary flow did not differ from WT hearts, suggesting that other factors compensated for the loss of eNOS. There was also no difference in the coronary venous release of adenosine, making it rather unlikely that this vasoactive nucleoside may have compensated for the loss of NO in maintaining constant coronary flow. Also, the sensitivity of the coronary vessels to infused adenosine was not different between WT and eNOS−/− mice, ruling out the possibility that changes in adenosine receptor density and/or signal transduction may be involved.
Reactive hyperemia after brief periods of coronary occlusion is in part mediated by the concerted action of adenosine and NO.16 However, activation of KATP channels in reactive hyperemia may play a role as well.30 In WT hearts, L-NAME substantially changed the flow response to short-term ischemia. Maximal and repayment flow were reduced after inhibition of NOS activity. In contrast, eNOS−/− hearts revealed no decrease in either maximal or repayment flow. Again, the adenosine release was identical in WT and eNOS−/− mice, suggesting that the loss of eNOS function was compensated for by a release of other mediators.
ACh causes coronary vasodilation by the combined activation of eNOS, the prostaglandin pathway, and endothelium-derived hyperpolarization factor. In addition, ACh can induce vasoconstriction when the endothelium is dysfunctional or high concentrations of ACh are applied.31 Both actions are probably mediated by the activation of M3 muscarinergic receptors, which are located on both endothelial and smooth muscle cells.32 However, the relative contribution of each of these mediators to the ACh-induced flow response varies considerably among species. Whereas in the guinea pig heart, ACh-induced vasodilation is dependent mainly on NO,33 a concerted action of PGI2 and NO was reported to mediate the response in the rabbit heart.34
In the mouse heart, ACh induces an initial rapid vasoconstriction, which is evident at the beginning of the flow response, followed by a vasodilation, which dominates the flow in the steady state. Most likely, the extent of vasodilation represents the net result of both oppositely directed effects. In WT mice, acute inhibition of NOS activity by L-NAME reduced coronary vasodilation in response to ACh. Thus, as in other species, NO contributes to the ACh-induced vasodilation in the mouse heart. The initial vasoconstriction was left unchanged. In eNOS−/− mice, on the other hand, the steady-state flow response to ACh remained unaltered compared with the WT control. The initial vasoconstriction, however, was potentiated in eNOS−/− mice. Assuming that the steady-state vasodilation represents the net effect of vasoconstriction and vasodilation, this result implies that the dilatory component of the ACh effect was also augmented, because no changes in net vasodilation occurred. These findings demonstrate that there are important changes in the responsiveness of the coronary vasculature to ACh. Chronic loss of NO appears to have sensitized cholinergic signal transduction at the site of coronary resistance vessels. Alternatively, it is conceivable that vascular permeability to ACh and/or metabolism of ACh is altered in eNOS−/− mice.
The present findings clearly demonstrate that in WT mice, the ACh-induced vasodilation is mediated predominantly by prostaglandins. Inhibition of cyclooxygenase with diclofenac abolished vasodilation and resulted in sustained vasoconstriction. Thus, the contribution of NO to the ACh-dependent vasodilation, which was identified by use of L-NAME, was not sufficient to counteract the ACh-induced vasoconstriction. In the eNOS−/− mice, conversely, vasodilation was mediated exclusively by prostaglandins. Chronic loss of eNOS in the presence of pharmacologically inhibited prostaglandin synthesis resulted in severe vasoconstriction, which reduced coronary flow by 71%. Thus, most likely the lack of eNOS during ACh-mediated vasodilation was compensated for by an elevated release of prostaglandins in eNOS−/− hearts. This result does not contradict the findings of Huang et al,4 who demonstrated lack of ACh response in isolated aortic segments of eNOS−/− mice. Rather, it emphasizes that the signal transduction pathway and mediators involved in the vascular response are profoundly different in the large conduit vessels compared with the small resistance vessels, which were examined in the present study.
Selected Abbreviations and Acronyms
|L-NAME||=||Nω-nitro-l-arginine methyl ester|
|eNOS||=||endothelial NO synthase|
|NOS||=||nitric oxide synthase|
|PCR||=||polymerase chain reaction|
This work was supported by a grant from the Fritz-Thyssen-Stiftung, Köln. We are grateful to Dr R. Kühn (University of Cologne) for the gift of E14–1 cells. Further, we would like to thank S. Küsters and E. Bergschneider for excellent technical assistance and W. Teutscher and P. Nieswand for animal care.
- Received July 31, 1997.
- Accepted October 8, 1997.
- © 1998 American Heart Association, Inc.
Ribeiro M, Antunes E, deNucci G, Lovisolo SM, Zatz R. Chronic inhibition of nitric oxide synthesis: a new model of hypertension. Hypertension. 1992;20:298–303.
Shesely EG, Maeda N, Kim HS, Desai KM, Krege JH, Laubach VE, Sherman PA, Sessa WC, Smithies O. Elevated blood pressures in mice lacking endothelial nitric oxide synthase. Proc Natl Acad Sci U S A. 1996;93:13176–13181.
Kelly RA, Balligand JL, Smith TW. Nitric oxide and cardiac function. Circ Res. 1996;79:363–380.
Kelm M, Schrader J. Control of coronary vascular tone by nitric oxide. Circ Res. 1990;66:1561–1575.
Ayajaki K, Kindermann M, Hecker M, Fleming I, Busse R. Intracellular pH and tyrosine phosphorylation but not calcium determine shear stress–induced nitric oxide production in native endothelial cells. Circ Res. 1996;78:750–758.
Marsden PA, Heng HHQ, Scherer SW, Stewart RJ, Hall AV, Shi XM, Tsui LC, Schappert KT. Structure and chromosomal localization of the human constitutive endothelial nitric oxide synthase gene. J Biol Chem. 1993;268:17478–17488.
Quyyumi AA, Dakak N, Andrews NP, Gilligan DM, Panza JA, Cannon RO III. Contribution of nitric oxide to metabolic vasodilation in the human heart. Circulation. 1995;92:320–326.
Kostic MM, Schrader J. Role of nitric oxide in reactive hyperemia of the guinea pig heart. Circ Res. 1992;70:208–212.
Park KH, Rubin L, Gross SS, Levi R. Nitric oxide is a mediator of hypoxic coronary vasodilation: relation to adenosine and cyclooxygenase-derived metabolites. Circ Res. 1992;71:992–1001.
Mansour SL, Thomas KR, Capecchi MR. Disruption of the proto-oncogene int-2 in mouse embryo-derived stem cells: a general strategy for targeting mutations to non-selectable genes. Nature. 1986;336:743–748.
Kühn R, Rajewsky K, Müller W. Generation and analysis of interleukin 4 deficient mice. Science. 1991;254:707–710.
Kroll K, Decking UKM, Dreikorn K, Schrader J. Rapid turnover of the AMP-adenosine metabolic cycle in the guinea pig heart. Circ Res. 1993;66:1561–1575.
Xie QW, Cho H, Kashiwabara Y, Baum M, Weidner JR, Elliston K, Mumford R, Nathan C. Carboxyl terminus of inducible nitric oxide synthase: contribution to NADPH binding and enzymatic activity. J Biol Chem. 1994;269:28500–28505.
Linder L, Kiowski W, Bühler W, Lüscher TF. Indirect evidence for the release of endothelium-derived relaxing factor in the human forearm circulation in vivo: blunted response in essential hypertension. Circulation. 1990;8:1762–1767.
Panza JA, Casino PR, Kilcoyne CM, Quyyumi AA. Role of endothelium-derived nitric oxide in the abnormal endothelium-dependent vascular relaxation of patients with essential hypertension. Circulation. 1993;87:1468–1474.
Taddei S, Virdis A, Mattei P, Ghiadoni L, Sudano I, Salvetti A. Defective l-arginine–nitric oxide pathway in offspring of essential hypertensive patients. Circulation. 1996;94:1228–1303.
Kanatsuka H, Sekiguchi N, Sato K, Akai K, Wang Y, Komaru T, Ashikawa K, Takishima T. Microvascular sites and mechanisms responsible for reactive hyperemia in the coronary circulation of the beating canine heart. Circ Res. 1992;71:912–922.