Nitric Oxide Protects Cardiac Sarcolemmal Membrane Enzyme Function and Ion Active Transport against Ischemia-induced Inactivation

Abstract

Nitric oxide (NO.) generated from nitric oxide synthase (NOS) isoforms bound to cellular membranes may serve to modulate oxidative stresses in cardiac muscle and thereby regulate the function of key membrane-associated enzymes. Ischemia is known to inhibit the function of sarcolemmal enzymes, including the (Na+ + K+)-ATPase, but it is unknown whether concomitant injury to sarcolemma (SL)-associated NOS isoforms may contribute to this process by reducing the availability of locally generated NO.. Here we report that nNOS, as well as eNOS (SL NOSs), are tightly associated with cardiac SL membranes in several different species. In isolated perfused rat hearts, global ischemia caused a time-dependent irreversible injury to cardiac SL NOSs and a disruption of SL NO. generation. Pretreatment with low concentrations of the NO. donor 1-hydroxy-2-oxo-3-(N-3-methyl-aminopropyl)-3-methyl-1-triazene (NOC-7) markedly protected both SL NOS and (Na+ + K+)-ATPase functions against ischemia-induced inactivation. Moreover, ischemia impaired SL Na+/K+ binding, and NOC-7 significantly prevented ischemic injury to the ion binding sites on (Na+ + K+)-ATPase. These novel findings indicate that NO. can protect cardiac SL NOSs and (Na+ + K+)-ATPase against ischemia-induced inactivation and suggest that locally generated NO. may serve to regulate SL Na+/K+ ion active transport in the heart. Nitric oxide (NO.) generated from nitric oxide synthase (NOS) isoforms bound to cellular membranes may serve to modulate oxidative stresses in cardiac muscle and thereby regulate the function of key membrane-associated enzymes. Ischemia is known to inhibit the function of sarcolemmal enzymes, including the (Na+ + K+)-ATPase, but it is unknown whether concomitant injury to sarcolemma (SL)-associated NOS isoforms may contribute to this process by reducing the availability of locally generated NO.. Here we report that nNOS, as well as eNOS (SL NOSs), are tightly associated with cardiac SL membranes in several different species. In isolated perfused rat hearts, global ischemia caused a time-dependent irreversible injury to cardiac SL NOSs and a disruption of SL NO. generation. Pretreatment with low concentrations of the NO. donor 1-hydroxy-2-oxo-3-(N-3-methyl-aminopropyl)-3-methyl-1-triazene (NOC-7) markedly protected both SL NOS and (Na+ + K+)-ATPase functions against ischemia-induced inactivation. Moreover, ischemia impaired SL Na+/K+ binding, and NOC-7 significantly prevented ischemic injury to the ion binding sites on (Na+ + K+)-ATPase. These novel findings indicate that NO. can protect cardiac SL NOSs and (Na+ + K+)-ATPase against ischemia-induced inactivation and suggest that locally generated NO. may serve to regulate SL Na+/K+ ion active transport in the heart. Myocardial ischemia causes intracellular energy depletion, leading to loss of cellular Na+/K+ homeostasis through inhibition of the sarcolemmal (Na+ + K+)-ATPase, resulting ultimately in cellular Ca2+ overload (1Jennings R.B. Reimer K.A. Hosp. Pract. 1989; 24: 89-107Crossref Scopus (9) Google Scholar, 2Jennings R.B. Steenbergen C. Reimer K.A. Monogr. Pathol. 1995; 37: 47-80PubMed Google Scholar, 3Christopherson K.S. Bredt D.S. J. Clin. Invest. 1997; 100: 2424-2429Crossref PubMed Scopus (274) Google Scholar, 4Murphy J.G. Marsh J.D. Smith T.W. Circulation. 1987; 75: 15-24PubMed Google Scholar). The nature of ischemic injury to the (Na+ + K+)-ATPase and its effect on ion pump function has not been fully defined despite its fundamental importance in the maintenance of cell viability. Sarcolemma (SL) 1The abbreviations used are: SL, sarcolemma; NOS, NO synthase; iNOS, inducible NOS; eNOS, endothelial NOS; nNOS, neuronal NOS; uNOS, universal NOS; EPR, electron paramagnetic resonance; NOC-7, 1-hydroxy-2-oxo-3-(N-methyl-3-aminopropyl)-3-methyl-1-triazene; RyR, ryanodine receptor; PLB, phospholamban. contains many bound enzymes that may modify the activity of (Na+ + K+)-ATPase, including nitric oxide synthase. NO. synthase is a cell type-specific enzyme that catalyzes the synthesis of NO., a short-lived gaseous free radical, which also serves as an important physiological regulator of myocardial function (5Hess M.L. Kukreja R.C. Ann. Thorac. Surg. 1995; 60: 760-766Abstract Full Text PDF PubMed Scopus (49) Google Scholar). All NOS isoforms, including neuronal NOS (nNOS), inducible NOS (iNOS), endothelial NOS (eNOS), and neuronal NOSμ (nNOSμ), have been identified in cardiac muscle (6Kelly R.A. Balligad J.L. Smith T.W. Circ. Res. 1996; 79: 363-380Crossref PubMed Scopus (629) Google Scholar, 7Xu K.Y. Huso D.L. Dawson T.M. Bredt D.S. Becker L.C. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 657-662Crossref PubMed Scopus (418) Google Scholar, 8Kobzik L. Reid M.B. Bredt D.S. Stamler J.S. Nature. 1994; 372: 546-548Crossref PubMed Scopus (843) Google Scholar, 9Silvagno F. Xia H. Bredt D.S. J. Biol. Chem. 1996; 271: 11204-11208Abstract Full Text Full Text PDF PubMed Scopus (299) Google Scholar), However, little is known about specific NOS isoforms on the SL membrane (SL NOSs) and what effect ischemia may have on their function. NO. generated at the sarcolemma by SL NOSs could modulate the activity of SL-associated ion pumps, but whether ischemic dysfunction of SL NOSs directly affects SL (Na+ + K+)-ATPase activity and SL ion active transport has not been studied. The functional relationships among SL NOSs, (Na+ + K+)-ATPase, NO., and SL Na+/K+ ion active transport in normal and diseased hearts have remained open questions. In this study, we show that, like eNOS, nNOS is also tightly associated with the SL membrane in several species. Using an assay of NOS activity, we demonstrate that global ischemia causes a time-dependent irreversible injury to the SL NOSs and a disruption of SL NO. generation. Pretreatment of isolated rat hearts with an NO. donor prior to global ischemia markedly protects SL NOSs and (Na+ + K+)-ATPase activities and ion binding to the SL membrane, suggesting that SL NO. may play a crucial role in regulating SL membrane ion active transport. Materials—An NOS-detect assay kit, including reduced NADPH, 1-hydroxy-2-oxo-3-(N-3-methyl-aminopropyl)-3-methyl-1-triazene (NOC-7), calmodulin, FAD, FMN, and tetrahydrobiopterin, was purchased from Alexis Biochemicals; l-arginine and EDTA were from Fluka; [3H]l-arginine was purchased from Amersham Biosciences; anti-eNOS (catalog no. PA3-031), anti-universal NOS (anti-uNOS, catalog no. MA3-030; an antibody reacting with iNOS, eNOS, and nNOS isoforms), anti-phospholamban (anti-PLB, catalog no. MA3-922), and anti-ryanodine receptor (anti-RyR, catalog no. MA3-916) were from Affinity BioReagents; Laemmli sample buffer was from Bio-Rad and all other reagents were from Sigma. The anti-nNOSα antibody (specifically recognizing the major α-isoform of nNOS and not the β- or γ-isoforms) and the homozygous mutant nNOS (nNOS –/–) knockout mouse model were from the laboratory of Dr. Ted W. Dawson (10Huang P.L. Dawson T.M. Bredt D.S. Snyder S.H. Fishman M.C. Cell. 1993; 75: 1273-1286Abstract Full Text PDF PubMed Scopus (1142) Google Scholar). The laboratory of Dr. Arthur L. Burnett provided both eNOS knockout (eNOS –/–) and nNOS/eNOS double knockout (nNOS/eNOS –/–) mice for this study. The homozygous mutants nNOS knockout, eNOS knockout, and nNOS/eNOS double knockout mice were backcrossed to the C57BL/6J black mouse. They were originally generated from Dr. Paul L. Huang's laboratory at the Massachusetts General Hospital, Boston, MA (11Huang P.L. Lo E.H. Prog. Brain. Res. 1998; 118: 13-25Crossref PubMed Google Scholar, 12Huang P.L. Huang Z. Mashimo H. Bloch K.D. Moskowitz M.A. Bevan J.A. Fishman M.C. Nature. 1995; 377: 239-242Crossref PubMed Scopus (1780) Google Scholar). Isolated Rat Heart Ischemia and Reperfusion—Female Sprague-Dawley retired breeder rats (Harlan Company, Boston, MA) were used for the study, and all experiments were carefully conducted in compliance with the National Institutes of Health guidelines for the use of laboratory animals. The heart was excised after complete anesthesia (65 mg/kg pentobarbital), and the ascending aorta was rapidly cannulated and the heart perfused. Following an equilibrium period of 15 min to allow for functional stabilization, baseline values of coronary flow left ventricular end diastolic pressure, left ventricular systolic pressure, and heart rate were measured. After 15 min of baseline coronary flow, the hearts were then subjected to global (no-flow) ischemia in different time courses with or without 45 min of reperfusion. The perfusate solution contained 17 mm glucose, 120 mm NaCl, 25 mm NaHCO3, 5.9 mm KCl, 2.5 mm CaCl2, 1.2 mm MgCl2, and 0.5 mm EDTA. All perfusate solutions were routinely filtered and bubbled with 95% O2/5% CO2 gas mixture at 37 °C (13Li H. Xu K.Y. Zhou L. Kalai T. Zweier J.L. Hideg K. Kuppusamy P. J. Pharmacol. Exp. Ther. 2000; 295: 563-571PubMed Google Scholar). A side arm in the perfusion line located just proximal to the aortic cannula allowed infusion of drug solutions using a Harvard infusion pump (Harvard Apparatus, Holliston, MA). A balloon was connected via a hydraulic line to an MLT844 physiological pressure transducer with pressures amplified to a Gould four-channel strip chart recorder as well as a personal computer equipped with MacLab data acquisition software. The balloon volume was adjusted to achieve an initial left ventricular end-diastolic pressure of 8–12 mm Hg, and all subsequent measurements were performed with the same balloon volume. Studies of ischemia and pretreatment with a membrane-permeable NO. donor NOC-7 (14Zhang P. Ohara A. Mashimo T. Sun J. Shibuta S. Takada K. Kosaka H. Terada M. Yoshiya I. Circulation. 1996; 94: 2235-2240Crossref PubMed Scopus (20) Google Scholar) were performed in four groups as follows: (i) control (perfusion for 20, 40, or 60 min without ischemia or drug treatment); (ii) treatment with NOC-7 infused at concentrations of 10, 30, 60, 90, 120, or 150 ng/kg/min for 10 min; (iii) global zero flow ischemia for 20, 40, or 60 min; and (iv) pretreatment with NOC-7 [10 ng/kg/min (61.7 nm)or30 ng/kg/min (185 nm)] for 10 min prior to ischemia under the same conditions as described previously (3Christopherson K.S. Bredt D.S. J. Clin. Invest. 1997; 100: 2424-2429Crossref PubMed Scopus (274) Google Scholar). This NOC-7 compound spontaneously and non-enzymatically releases NO. at physiological pH with a half-life time of 10 min. Isolation of Cardiac Sarcolemmal Vesicles—Cardiac SL vesicles were isolated from rat, rabbit, and mouse hearts as reported previously (15Xu K.Y. Zweier J.L. Becker L.C. Ann. N. Y. Acad. Sci. 1997; 834: 680-683Crossref PubMed Scopus (23) Google Scholar). Briefly, animals were anesthetized with sodium pentobarbital. The hearts were removed immediately and rinsed in an ice-cold solution containing 10 mm histidine (free base) and 0.75 m NaCl. The hearts were cut into 5-mm pieces and homogenized for 5 s at 14,000 rpm. The homogenates were centrifuged at 10,000 rpm for 20 min, and the pellets were resuspended in 10 mm NaHCO3 and 5 mm histidine (free base). The pellets were resuspended, homogenized, and centrifuged three times as described above, and the supernatant was discarded to remove the sarcoplasmic reticulum (SR) of the cardiac tissue for the ultimate isolation of SL vesicles free of significant SR contamination. The pellets obtained from the third centrifugation were resuspended in 10 mm NaHCO3 and 5 mm histidine and homogenized three times for 30 s at 15,000 rpm. The homogenates were centrifuged at 10,000 rpm for 20 min. The supernatant containing partially purified SL vesicles was centrifuged again for 30 min at 20,000 rpm. The pellets were resuspended and layered on top of the 0.6 m sucrose in each tube. The sample was centrifuged at 60,000 rpm for 54 min in a Beckman Ti70 rotor. After the run, the SL vesicles were observed as distinct snow white protein bands at the interfaces of the 0.25/0.6 m sucrose layers. SL vesicle fractions were then sedimented at 40,000 rpm for 40 min. The final SL membrane was at SL vesicles were with J. Xu K.Y. 1987; PubMed Scopus Google Scholar) and were in anti-phospholamban and anti-ryanodine receptor were to the of SL of SR K.Y. Huso D.L. Dawson T.M. Bredt D.S. Becker L.C. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 657-662Crossref PubMed Scopus (418) Google Scholar) membrane in isolated cardiac SL vesicles was by under conditions and was used for the different were used for nNOS and eNOS The anti-nNOSα antibody was against an protein by of nNOS and the which is the major of nNOS and with Bredt D.S. Dawson T.M. 1994; Full Text PDF PubMed Scopus Google Scholar). The anti-eNOS antibody with other NOS The antibody with all three isoforms, iNOS, eNOS, and Briefly, were by the Laemmli sample buffer and by on a of for The were then to membranes and with or for 60 min at After the membranes were with or to which was The of bound antibody was by of NOS activity was by the of to K.Y. Huso D.L. Dawson T.M. Bredt D.S. Becker L.C. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 657-662Crossref PubMed Scopus (418) Google Scholar), of vesicles and 10 of were to 40 of buffer containing mm mm reduced NADPH, mm CaCl2, calmodulin, FAD, FMN, and After for 30 min at the was with of mm pH and the mixture was to a of and with of The at pH through the the in the the NOS of + + + SL vesicles were for min at with J. Xu K.Y. 1987; PubMed Scopus Google Scholar), and were removed and activity was defined as the of in the of and J. J. Biol. Chem. Full Text PDF Google Scholar). The mixture contained enzyme and mm in a final volume of 0.5 and was at 37 °C in a The was by and after 30 min by 0.75 of solution and of The was allowed to for 30 min at and the was at using a measurements of NO were performed in a cell at K.Y. Huso D.L. Dawson T.M. Bredt D.S. Becker L.C. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 657-662Crossref PubMed Scopus (418) Google Scholar). The final concentrations of SL calmodulin, NADPH, FAD, FMN, tetrahydrobiopterin, and l-arginine were 5 0.6 and The was immediately after the solution was infused using an at with a The were as follows: min; and nNOS and eNOS with Cardiac SL the specific isoforms of NOS that are associated with cardiac SL Using an which all three isoforms of NOS, iNOS, eNOS, and nNOS, that nNOS and eNOS are in rat, rabbit, and mouse cardiac SL membrane vesicles and was not not and anti-eNOS nNOS and eNOS from control mouse SL vesicles and of nNOS or eNOS was observed in cardiac SL vesicles isolated from knockout nNOS, eNOS, or nNOS/eNOS double knockout mouse and and and in isolated rat cardiac SR vesicles but to and in SL a of isolated SL vesicles and suggesting a specific of nNOS and eNOS in cardiac SL of Ischemia on SL NOS we the effect of ischemia on SL NOS function. of l-arginine to was performed in the of the NOS and cardiac SL NOS activity significantly from to 30 and 5 30, 60, and 120 min of The loss of activity ischemia was by a of SL NO. from to and 30, 60, and 120 min of whether ischemia affects the binding of nNOS and eNOS to the SL membrane and whether the low NOS activity ischemia is to loss of enzyme from the SL membrane, experiments were performed using SL vesicles isolated from rat cardiac muscle under ischemic conditions as in The concentrations of SL nNOS and eNOS in the SL membrane vesicles isolated from normal and ischemic hearts 30, 60, and 120 min of This the of NOS isoforms in SL the of ischemic dysfunction of cardiac SL NOS enzymes, experiments were performed to the of SL NOS that cardiac SL NOS activity not 45 min after 30 or 60 min of ischemia to a time-dependent irreversible injury to cardiac SL NOS function. of NO in SL whether ischemic dysfunction of SL NOSs could SL membrane ion active we whether NO. protects SL (Na+ + K+)-ATPase activity under ischemic the effect of NOC-7 on cardiac SL (Na+ + K+)-ATPase function without Isolated rat hearts were perfused with the NO. donor NOC-7 at different concentrations as in effect of NOC-7 10 ng/kg/min to 150 on SL (Na+ + K+)-ATPase function was observed as in then whether perfusion with 10 ng/kg/min NOC-7 enzyme functions under different ischemic data show that pretreatment of 10 ng/kg/min NOC-7 provided to both SL (Na+ + K+)-ATPase and SL NOS SL (Na+ + K+)-ATPase activity was and with control hearts and and for SL NOSs after 20, 40, and 60 min of However, pretreatment with 30 ng/kg/min NOC-7 markedly protected SL (Na+ + K+)-ATPase activity against ischemia-induced enzyme activities remained at and 20, 40, and 60 min of In (Na+ + K+)-ATPase activity to 45 and 15 in the of the NO. donor under the same ischemic that 40 min of global ischemia the ion binding sites on the (Na+ + K+)-ATPase by reducing the binding of and to the The and binding sites were protected by pretreatment of NOC-7, suggesting that NO. in cardiac muscle may crucial to the SL membrane ion active transport. In pretreatment of NOC-7 markedly SL NOS activity at and after 20, 40 and 60 min by free that were generated + K+)-ATPase of as a function of Cardiac SL vesicles were isolated from the hearts of control black and 40 min of ischemia with or without NOC-7 activity was in the of mm under different Na+/K+ concentrations as in the The data show that pretreatment by NOC-7 significantly protected SL Na+/K+ ion binding and (Na+ + K+)-ATPase function against ischemia-induced The reported show that nNOS, as well as eNOS, is tightly associated with cardiac SL membrane vesicles and The that and anti-eNOS nNOS and eNOS in mice and but to nNOS and eNOS and in SL vesicles from knockout nNOS or eNOS mice and that of nNOS or eNOS were observed in cardiac SL vesicles isolated from nNOS/eNOS double knockout mouse model that NOS isoforms are on the cardiac SL and are SR membrane protein and and in isolated rat SR vesicles but to and in isolated SL suggesting a specific of nNOS and eNOS in cardiac SL is that eNOS is associated with SL on endothelial and cardiac through binding with and Smith T.W. T. R.A. J. Biol. Chem. 1997; Full Text Full Text PDF PubMed Scopus Google Scholar, L. L. Smith T.W. R.A. T. J. Biol. Chem. 1996; 271: Full Text Full Text PDF PubMed Scopus Google Scholar). nNOS in rat SL but not in ventricular A. M. H. M. PubMed Scopus Google Scholar). and Burnett Huang P.L. J.A. Nature. PubMed Google Scholar), using and of mouse heart reported that eNOS with nNOS with the ryanodine was not reported Burnett Huang P.L. J.A. Nature. PubMed Google Scholar). has been that rat muscle sarcolemma contains an nNOS and NO L. Reid M.B. Bredt D.S. Stamler J.S. Nature. 1994; 372: 546-548Crossref PubMed Scopus (843) Google Scholar, D.S. Xia H. K. Bredt D.S. Cell. 1995; Full Text PDF PubMed Scopus Google Scholar). by of nNOS in cardiac SL membrane from several species. In we also eNOS in SL membrane, which is with T. J. Clin. Invest. 1997; 100: PubMed Scopus Google Scholar). have the effect of ischemia on myocardial have on whether ischemia may cardiac SL NOS function. show that is a time-dependent loss of SL NOS activity and SL NO. A and global of ischemia in an loss of SL NOS activity and NO. generation. protein concentrations of SL NOSs from control and ischemic rat hearts were observed by suggesting that SL NOSs activity was not to loss of NOS binding to the SL The that SL NOS activity not after 45 min of that global ischemia may a time-dependent irreversible injury to SL NOS function. Cardiac SL (Na+ + K+)-ATPase 1998; PubMed Scopus Google Scholar) is an membrane protein that the of directly to the transport of and the SL This enzyme the of by regulating intracellular sodium and membrane of (Na+ + K+)-ATPase in ischemic myocardial injury may contribute to intracellular overload and of heart function. that NO. generated from cardiac SL NOSs may a of SL (Na+ + K+)-ATPase function and SL and ion transport. the is the (Na+ + K+)-ATPase activity to myocardial ischemia with a NO. we the NO. donor NOC-7 in isolated rat hearts prior to global not what was by the infusion of the NO. but we to use low concentrations the in the of a of NO. at the SL membrane, which may have been in the of what under physiological Pretreatment with 10 ng/kg/min NOC-7 (61.7 protected SL (Na+ + K+)-ATPase function against 20 min of ischemia-induced of cardiac SL (Na+ + K+)-ATPase activity 40 min of ischemia was prevented by pretreatment with 30 ng/kg/min NOC-7 (185 and significant of (Na+ + K+)-ATPase activity was also 60 min of global These suggest that NO. generated from SL NOSs in of SL (Na+ + K+)-ATPase against oxidative Cardiac SL (Na+ + K+)-ATPase is for SL membrane and active transport. activity on the of which on the binding of and that 40 min of ischemia ion binding sites on the (Na+ + K+)-ATPase by SL and binding to the In pretreatment with NOC-7 markedly prevented this These the that NO. may play an important role in and binding and Na+/K+ membrane active transport. generated free can cellular resulting in and and leading ultimately to dysfunction of the J. 1996; 271: PubMed Google Scholar). and have that are generated in after myocardial ischemia H. Circ. Res. 1993; Scopus Google Scholar). of isolated isolated rat or and isolated rat or hearts to free solutions or has in function and R.C. M.L. Res. PubMed Scopus Google Scholar). These important suggest that free may for injury to the SL NOSs and (Na+ + K+)-ATPase myocardial Kukreja that causes little to the cardiac SR R.C. Zweier J.L. Kuppusamy P. M.L. Circ. Res. PubMed Scopus Google Scholar) and SL R.C. M.L. J. Google Scholar), but cardiac SL function by the (Na+ + K+)-ATPase activity and binding sites T. J. 1987; Google Scholar) also show that inhibit nNOS activity K.Y. 2000; PubMed Scopus Google Scholar) and SR function by on the binding sites (15Xu K.Y. Zweier J.L. Becker L.C. Ann. N. Y. Acad. Sci. 1997; 834: 680-683Crossref PubMed Scopus (23) Google Scholar). These important findings suggest a inactivation for both SL NOSs and (Na+ + K+)-ATPase in which generated the global may and or and ion binding sites of the enzymes, resulting in loss of SL NOSs and (Na+ + K+)-ATPase function. has been reported that NO. with both and at rate of and J. Chem. 1998; Scopus Google Scholar, S. Res. 1993; PubMed Scopus Google Scholar). In we the effect of NO. in measurements that purified nNOS was to an NO. from nNOS markedly K.Y. 2000; PubMed Scopus Google Scholar). These that NO. The effect of NO. was using the NO. donor NOC-7 K.Y. 2000; PubMed Scopus Google Scholar). NO., generated from NOC-7, from an the K.Y. 2000; PubMed Scopus Google Scholar), and prevented inactivation of both purified nNOS and (Na+ + K+)-ATPase. These in the NO. and and a for of SL NOS and SL membrane ion active transport function both SL NOSs and (Na+ + K+)-ATPase are bound to the SL membrane, NO. may in free the of the (Na+ + K+)-ATPase. pretreatment of affects is not not the of this process that may also contribute to the NO. effect K. Dawson T.M. Dawson J. Pharmacol. Exp. Ther. Google Scholar). of functional the pretreatment of NO. donor and including of and protein of the of role of NO.. The reported a significant inactivation of cardiac SL NOSs function global of cardiac SL NOSs may contribute to the of ischemic injury by a of NO., generated at the SL membrane, resulting in reduced of findings that NO. protects cardiac SL NOSs and (Na+ + K+)-ATPase against ischemia-induced inactivation suggest that locally generated NO. may serve to regulate SL membrane Na+/K+ ion active transport in the heart. and for the nNOS knockout and Becker for the eNOS knockout and nNOS/eNOS double knockout