Brief Periods of Nitric Oxide Inhalation Protect against Myocardial Ischemia–Reperfusion Injury

Male C57BL/6J mice fed a standard diet (RMH 3000, Prolab; PMI International, St. Louis, MO) were studied. All animal experimental protocols were approved by both the Subcommittee on Research Animal Care at Massachusetts General Hospital, Boston, Massachusetts, and the Institutional Animal Care and Use Committee at Boston University School of Medicine, Boston, Massachusetts.

Mice (n = 4) were placed in a chamber (PLY3211; Buxco Research Systems, Wilmington, NC) and exposed to air supplemented with 80 ppm NO for 60 min. The quantity of nitric oxide absorption was calculated from the difference between inlet and outlet nitric oxide concentrations (parts per million), multiplied by the gas flow rate (3 l/min), and divided by the molar volume of nitric oxide at standard temperature and pressure (22.4 l/mol). To account for the dilution of nitric oxide with ambient air when the chamber was opened to insert the mouse (3 s), as well as the generation of nitrogen dioxide in the chamber, “sham absorption of nitric oxide” was measured under the same conditions without inserting a mouse (3.3 ± 0.1 μmol/h, n = 4). Nitric oxide absorbed (total absorption of nitric oxide, 7.9 ± 0.6 μmol/mouse · h, minus sham absorption of nitric oxide, 3.3 μmol), was divided by body weight and was expressed as μmol NO/g body weight per hour (Additional information regarding our detailed description of nitric oxide absorption measurements is available on the Anesthesiology Web site at http://www.anesthesiology.org, in Web Enhancement 1–Materials and Methods.)

For measurement of whole body nitric oxide metabolite levels, mice that breathed air without (n = 5) or with (n = 4) 80 ppm NO for 1 h in the chamber were subsequently anesthetized with diethyl ether and killed by cervical dislocation. Rapid full body homogenization was achieved with a Waring blender (Waring Products, Torrington, CT) using a mixture of frozen and chilled phosphate-buffered saline (at 1:5, wt/vol) containing N -ethylmaleimide (10 mm) and EDTA (2.5 mm), and the resulting homogenate was immediately analyzed.

Additional mice were placed in the chamber and exposed to air or 80 ppm NO in air for 0.5, 5, 15, and 60 min (n = 4–7 per time point). After the exposure, mice were anesthetized with diethyl ether. Blood was withdrawn from the left ventricle (LV) and immediately centrifuged at 16,000g  for 3–5 min at room temperature (22°C) to separate erythrocytes from plasma. Erythrocytes were subjected to hypotonic lysis in water containing N -ethylmaleimide (10 mm) and EDTA (2.5 mm). To remove blood, tissues were perfused via  the LV with room air-equilibrated phosphate-buffered saline supplemented with N -ethylmaleimide (10 mm) and EDTA (2.5 mm) for 1 min. Brain, heart, liver, kidney, lung, and fat were harvested, homogenized, and subjected to immediate analysis. Urine was obtained via  direct puncture of the bladder. Nitrite, nitrate, RSNO, RNNO, and nitric oxide–heme species were quantified in plasma, erythrocytes, tissues, and whole body homogenates. Because of volume limitations, urine analysis was restricted to nitrite and nitrate only.

To investigate the effect of reduced oxygen availability as occurring during ischemia on nitric oxide metabolite levels in the heart, mice were placed in the chamber and breathed either 8% oxygen in nitrogen (Hypoxia, n = 5) or 80 ppm NO in 8% oxygen for 60 min (Hypoxia + nitric oxide, n = 5). After the exposure, mice were anesthetized with diethyl ether, and cardiac nitric oxide metabolite concentrations were measured immediately after the tissue perfusion with phosphate-buffered saline containing N -ethylmaleimide (10 mm) and EDTA (2.5 mm) and homogenizing.

Methods for the detection of nitroso (RSNO and RNNO) and nitrosyl (nitric oxide–heme) compounds, as well as the oxidation products of nitric oxide (nitrite and nitrate), in blood and tissues have been detailed previously.17Quantification was achieved by group-specific denitrosation after injection of biologic samples into either a triiodide-containing reaction mixture (nitroso species) or a potassium ferricyanide solution (nitrosyl species) constantly purged with nitrogen, and oxidation products of nitric oxide were measured with the aid of a gas phase chemiluminescence detector (CLD 77am sp; EcoPhysics, Ann Arbor, MI). Nitrate and nitrite concentrations were quantified by ion chromatography (ENO20 Analyzer; Eicom, San Diego, CA).

Mice were anesthetized by intraperitoneal administration of ketamine (120 mg/kg) and xylazine (5 mg/kg) and ventilated (MiniVent 845; Hugo Sachs Elektronik–Harvard Apparatus GmbH, March-Hugstetten, Germany) at a fraction of inspired oxygen (Fio²) of 0.99–1.0. Myocardial ischemia was induced by ligation of the left coronary artery for 60 min, followed by reperfusion for 24 h.6Nitric oxide was administered during ischemia for 60, 5, or 0.5 min, just before reperfusion. During the surgical procedures, an Fio²of 0.99–1.0 without or with 80 ppm NO was applied using two separate mechanical ventilators. After 24 h, the artery was re-ligated, and either fluorescent microspheres (0.25 ml, 10-μm diameter, FluoSpheres; Invitrogen Corporation, Carlsbad, CA; for 60- and 5-min nitric oxide inhalation studies) were injected into the LV, or tissue-marking dye (0.25 ml, TMD-BL; Triangle Biomedical Science Inc., Durham, NC; for 0.5-min nitric oxide inhalation study) was injected into the right carotid artery to determine the area at risk (AAR). The heart was excised, and four consecutive 1-mm cardiac slices were stained with 2,3,5-triphenyltetrazolium chloride (1% wt/vol; Sigma-Aldrich, St. Louis, MO) for the measurement of myocardial infarction (MI) size. LV, AAR, and MI area were measured by computer-assisted planimetry (NIH Image J 1.34; Bethesda, MD), and AAR/LV and MI/AAR ratios were calculated.6 

All data are presented as mean ± SEM. Data were analyzed using one-way analysis of variance with means comparison using the Bonferroni test (Origin 7.0; OriginLab Corporation, Northampton, MA). To compare the changes of nitric oxide metabolites in blood, tissues, and urine, each time point was compared with the corresponding baseline value. For the comparison of cardiac nitric oxide metabolite levels in mice breathing low oxygen concentrations, a separate one-way analysis of variance with a means comparison using the Bonferroni test was performed for the comparison of the four groups; i.e. , baseline, breathing nitric oxide in air for 60 min, breathing 8% oxygen for 60 min, and breathing nitric oxide in 8% oxygen for 60 min. To analyze the effects of breathing nitric oxide on ischemia–reperfusion injury, analyses were performed for 60-, 5-, and 0.5-min inhalations of nitric oxide and compared with each control group. P  values less than 0.05 were considered significant.

To investigate the metabolic fate of inhaled nitric oxide, we first measured the absorption of nitric oxide from ambient gas during spontaneous ventilation in awake mice. The rate of nitric oxide uptake was nearly linear with time (0.19 ± 0.02 μmol NO/g body weight · h; fig. 1).

To further characterize the pharmacokinetics of absorbed nitric oxide, we examined how much of the absorbed nitric oxide could be recovered as nitric oxide metabolites. To measure the accumulation of nitric oxide metabolites during nitric oxide inhalation (80 ppm, 1 h), mice were killed and homogenized, and individual nitric oxide metabolites in whole body extracts were quantified. Nitric oxide inhalation led to an increase in the total body concentrations of all the nitric oxide metabolites we examined (table 1). Nitrate concentrations increased 18-fold and represented 97% of the total nitric oxide metabolites measured. Levels of nitric oxide–heme increased 13-fold, RSNO 8-fold, RNNO 5-fold, and nitrite 2-fold (table 1). Fifty-three percent (0.10 ± 0.02 μmol/g) of the nitric oxide absorbed from the gas phase during inhalation for 1 h was recovered as nitric oxide metabolites in the whole body extracts.

To gain detailed insight into the dynamics of uptake, distribution, and secondary metabolism of the nitric oxide absorbed during inhalation, the concentrations of nitric oxide metabolites were measured in blood (both erythrocytes and plasma; fig. 2) and tissues (heart, lung, brain, liver, kidney, and fat; fig. 3) of mice breathing air with or without 80 ppm NO for 0, 0.5, 5, 15, and 60 min (additional information regarding each concentration, the number of animals studied, and the P  value are available on the Anesthesiology Web site at http://www.anesthesiology.org, in Web Enhancement 2–table 1 and Web Enhancement 3–table 2). Breathing air without nitric oxide for varying periods of time (5, 15, and 60 min) did not alter nitric oxide metabolite concentrations in blood or any of the tissues we studied (data not shown).

During inhalation of nitric oxide, nitrate concentrations in plasma and erythrocytes increased linearly over the first 15 min and then tended to reach a plateau level (fig. 2). Nitrate concentrations were almost identical in plasma and erythrocytes at all time points. Nitrite concentrations in plasma reached a plateau within 15 min, whereas erythrocytic nitrite peaked as early as 5 min. Nitrite levels were lower than nitrate levels in plasma and erythrocytes by two orders of magnitude. The concentrations of RSNO, RNNO, and nitric oxide–heme increased markedly in erythrocytes (610-fold for RNNO, 535-fold for nitric oxide–heme, and 85-fold for RSNO; P < 0.001 for all) and greatly exceeded the erythrocytic or plasma nitrite concentration. In contrast, breathing nitric oxide did not significantly increase RSNO or nitric oxide–heme concentrations in plasma, and plasma RNNO levels increased only 3-fold (P < 0.001 vs.  baseline). These results suggest that plasma nitrite, nitrate, and RNNO, as well as erythrocytic RSNO, RNNO, nitric oxide–heme, nitrite, and nitrate, may all contribute to the transport of bioavailable nitric oxide from the lung to the periphery.

In the heart, increased RSNO and nitric oxide–heme levels were detected at 0.5 and 5 min, respectively, and maximum levels were attained after 15 min of nitric oxide inhalation (9- and 7-fold increases, respectively; P < 0.001 vs.  baseline for both; fig. 3). Cardiac RNNO concentrations were maximal at 0.5 min (P < 0.05 vs.  baseline) and returned to baseline thereafter despite continued inhalation of nitric oxide (fig. 3). The concentration of nitrate increased as early as 0.5 min and remained elevated thereafter. In contrast, cardiac nitrite levels were not elevated significantly during the inhalation of nitric oxide.

To study the effect of hypoxia on the heart, cardiac nitric oxide metabolite levels were measured in mice breathing a low oxygen concentration, with and without 80 ppm NO. Cardiac nitrate levels markedly increased after awake mice breathed 8% oxygen for 60 min (Hypoxia; a 9-fold increase vs.  baseline; P < 0.05; fig. 4). After 80 ppm NO was breathed in 8% oxygen for 60 min (Hypoxia + nitric oxide), the levels of nitric oxide–heme, RSNO, and RNNO were markedly elevated over those of Normoxia + nitric oxide (7-, 6-, and 5-fold increases, respectively; P < 0.05; fig. 4).

Inhalation of nitric oxide led also to rapid increases in nitric oxide metabolite concentrations in the lung, brain, and liver (fig. 3) but not in kidney or fat (additional information regarding the concentrations of nitric oxide metabolites in kidney and fat are available on the Anesthesiology Web site at http://www.anesthesiology.org, in Web Enhancement 3–table 2). The highest RSNO levels were achieved in the lung with peak concentrations attained within 5 min. During inhalation of nitric oxide, RNNO concentrations increased markedly in the liver and less so in the lung. In contrast, in the brain, inhalation of nitric oxide led to the accumulation of nitric oxide–heme, but not RSNO or RNNO. Breathing nitric oxide did not significantly increase nitrite levels in any of the tissues that we studied. Taken together, marked differences in nitric oxide metabolite regulation exist between the blood, heart, and other tissues, suggesting that generation or metabolism of nitric oxide metabolites or both are quite tissue specific.

During nitric oxide inhalation, nitrite and nitrate began to accumulate in the urine as early as 0.5 min (data not shown). After 60 min of nitric oxide inhalation, concentrations of nitrite in the urine were 0.7 ± 0.2 μm (P < 0.01 vs.  baseline level; fig. 5A) and were similar to those detected in plasma (1.0 ± 0.3 μm). In contrast, in mice breathing nitric oxide for 60 min, urinary nitrate concentrations (3.5 ± 0.5 mm) were 19-fold greater than those of plasma (P < 0.0001 vs.  baseline; fig. 5B). As an estimate of the quantity of absorbed nitric oxide that was excreted in the urine, the average concentration of nitrate after 60 min of nitric oxide breathing (3.5 mm) was multiplied by the volume of urine collected (119 ± 16 μl, n = 9) at the same time point. We estimate that approximately 9% of the nitric oxide absorbed over 1 h is excreted in the urine.

In a previous study of mice subjected to 60 min of cardiac ischemia and 24 h of reperfusion, we learned that continuous breathing of 80 ppm NO for 24 h decreased MI size as a fraction of myocardial area at risk (MI/AAR).6In the current study, the observation that blood levels of nearly all nitric oxide metabolites detected after breathing nitric oxide for 5 min were similar to those detected at 60 min (fig. 2) led us to determine whether a shorter duration of nitric oxide inhalation (≤ 60 min) could modify MI/ AAR at 24 h after reperfusion. The overall 24-h mortality rate of mice in our study after ischemia–reperfusion was 8%. In mice breathing nitric oxide for 60 or 5 min before reperfusion, MI/AAR was decreased by 32% (P < 0.05; fig. 6A) or 31% (P < 0.05; fig. 6B), respectively. In contrast, breathing nitric oxide for only 30 s just before reperfusion did not alter the degree of cardiac ischemia–reperfusion injury (fig. 6C).

In the current study, we provide the first quantitative and temporal characterization of the levels of nitric oxide metabolites that accumulate in the blood and peripheral tissues during nitric oxide inhalation. Moreover, we report that inhalation of nitric oxide for as little as 5 min before reperfusion can reduce infarct size in a murine model of myocardial ischemia–reperfusion injury, which supports the notion that increased levels of one or more nitric oxide metabolites in the blood contribute to the cardioprotective effects of breathing nitric oxide.

It is increasingly appreciated that breathing nitric oxide can elicit a wide spectrum of physiologic effects in peripheral tissues18; however, the mechanisms responsible for these salutary effects are incompletely understood. One possibility is that exposure of leukocytes and platelets to high nitric oxide concentrations as they transit the lung may inhibit their activation in peripheral tissues. On the other hand, multiple research groups have observed that inhalation of nitric oxide leads to the formation of nitric oxide metabolites in the bloodstream12,16and tissues.19To better understand the mechanisms responsible for the extrapulmonary effects of inhaled nitric oxide, we quantitatively assessed the fate of inhaled nitric oxide in whole body extracts, as well as in blood and representative tissues.

In mice breathing nitric oxide (80 ppm), the rate of nitric oxide absorption was essentially linear, and approximately 0.19 μmol/g body weight was absorbed within 1 h. Of the gaseous nitric oxide absorbed, we estimate that approximately 9% (0.017 μmol/g body weight) was excreted in the urine. RSNO, RNNO, nitric oxide–heme, nitrite, and nitrate recovered from whole body extracts accounted for 53% (0.10 μmol/g body weight) of the absorbed nitric oxide. The fate of the absorbed nitric oxide that was not detected as nitric oxide metabolites is currently unknown. Some of the absorbed nitric oxide may have been converted to metabolites not readily detected by the techniques we used (such as nitrotyrosine, nitrated fatty acids, and other nitrated species or stable C - or N -nitroso compounds). Alternatively, nitric oxide may have been reduced to nitrous oxide or nitrogen and exhaled. In whole body extracts, nitrate represented nearly 97% of the nitric oxide metabolites accumulating during nitric oxide inhalation, consistent with previous studies showing that conversion of absorbed nitric oxide into nitrate represents the major metabolic pathway for inhaled nitric oxide.16,20 

Lecour et al.  19reported that breathing 100 or 200 ppm NO increased nitric oxide concentrations in peripheral tissues, as detected by electron spin resonance spectroscopy combined with a spin-trapping technique. However, in contrast to our observations demonstrating that nitric oxide–heme levels increased as early as 5 min after the start of 80 ppm NO inhalation, Lecour et al.  did not detect an increase in cardiac nitric oxide concentrations in rats breathing 100 ppm NO for 45 min. This discrepancy may be explained by the differing techniques for nitric oxide detection or trapping: Our chemiluminescence-based technique is sufficiently sensitive to quantify the steady state concentrations achieved in a given compartment at baseline while nitric oxide is bound to its natural ligands. In contrast, electron spin resonance spectroscopy–based techniques require the accumulation of nitric oxide over a period of time using a transition metal–thiol complex as the trapping agent. In the latter technique, nitric oxide is bound to an exogenous ligand rather than a natural ligand, and higher than normal tissue nitric oxide concentrations are achieved, facilitating detection. However, techniques that use exogenous nitric oxide trapping agents inevitably perturb endogenous equilibria involving nitric oxide and its metabolites, and the apparent nitric oxide concentrations achieved depend on the probe’s distribution and saturation characteristics, the stability of the nitric oxide complex formed, and the duration of nitric oxide accumulation.

In a recent study, Lang et al.  9reported that in patients undergoing liver transplantation, breathing 80 ppm NO increased plasma nitrate and nitrite concentrations, as well as erythrocytic nitrate and nitric oxide–heme levels, but did not increase erythrocytic RSNO and RNNO levels. In contrast, we observed that breathing nitric oxide markedly increased erythrocytic RNNO and RSNO levels (600- and 80-fold, respectively). The reasons for this discrepancy are unclear. It is possible that differences in thiol reactivity between rodent and human hemoglobin contribute to differences in the concentrations of nitroso products formed.21 

During inhalation of nitric oxide, the rate of nitric oxide metabolite accumulation differed depending on the tissue we studied. Increased levels of RSNO, RNNO and nitric oxide–heme were measured in the heart of mice breathing nitric oxide, and importantly, the concentrations achieved were similar to those detected in mice carrying a transgene that directs systemic expression of nitric oxide synthase 3.22Of note, this strain of mice was shown to be protected from cardiac ischemia–reperfusion injury.22The accumulation of RSNO in the lung during inhalation of nitric oxide is consistent with the observations of Moya et al.  23In brain, only nitric oxide–heme levels increased after breathing nitric oxide for 15 min. The constancy of increased nitric oxide–heme levels in the brain despite continued nitric oxide inhalation may reflect decreased import or increased export of nitric oxide metabolites and/or down-regulation of endogenous nitric oxide production. In comparison to the other tissues we studied, breathing nitric oxide led to the greatest accumulation of RNNO in the liver (a 6.6-fold increase by 15 min), and increased hepatic RSNO and nitric oxide–heme concentrations were observed. Differences in the distribution of nitric oxide metabolites detected in these tissues strongly suggest that detection of these metabolites is not attributable to blood contaminating the tissues. Moreover, these findings suggest that the uptake, metabolism, or excretion of nitric oxide metabolites is regulated in a tissue-specific manner.

In the murine model of cardiac ischemia–reperfusion injury, nitric oxide was administered only during the ischemia period, in which nitric oxide metabolites can reach the ischemic tissue via  the circulation only after subsequent coronary reperfusion. Using this model, breathing nitric oxide for 5 min (just before reperfusion) significantly decreased the cardiac injury. In contrast, breathing nitric oxide for 0.5 min just before reperfusion (a duration of breathing nitric oxide that resulted in nitric oxide metabolite concentrations in blood that were consistently lower than those measured in mice breathing nitric oxide for 5 or 60 min) did not protect against cardiac ischemia–reperfusion injury. We acknowledge that, because the dead space and cardiac output differ between spontaneous breathing and mechanically ventilated animals, the uptake and distribution of nitric oxide in awake mice may differ from that in mice undergoing a thoracotomy and transient coronary artery occlusion (during which nitric oxide was administered through the animal’s ventilator). Nevertheless, our findings correlated with the observation that brief periods of nitric oxide inhalation were capable of protecting against cardiac ischemia–reperfusion injury in mice. The substantial elevations of cardiac nitric oxide metabolites measured in mice breathing 8% oxygen supplemented with nitric oxide in this study suggests that myocardial ischemia is likely to alter the cardiac nitric oxide metabolite levels produced by nitric oxide inhalation. To our knowledge, this is the first report of the hypoxia-related effects of inhaled nitric oxide on nitric oxide metabolites in cardiac tissue.

Importantly, all of the nitric oxide metabolites that we found to be elevated in the blood are capable of producing nitric oxide–related effects in the periphery. For example, RSNOs (e.g. , SNO-hemoglobin and SNO-albumin) are known to dilate blood vessels,24,25and RNNOs can increase cyclic guanosine monophosphate concentrations26and induce relaxation of bovine coronary arteries.27Especially in the ischemic myocardium, where tissue pH may decrease below 5.5 after 30 min of ischemia, nitrite28or nitrate29may generate nitric oxide. Accordingly, although the half-life of nitric oxide is short, blood nitric oxide metabolites with longer lifetimes can deliver nitric oxide from the lung to distant organs and regenerate nitric oxide in the periphery. Moreover, it is also possible that the downstream effects of nitric oxide metabolites may themselves be longer lasting.

Because any one nitric oxide metabolite can be converted into many others, the determination of which single nitric oxide metabolite confers cardioprotection, either solely or in concert with other nitric oxide metabolites, remains a challenge. It has been previously reported that many nitric oxide donors and nitric oxide metabolites, including S -nitrosoglutathione and nitrite, can protect the heart from ischemia–reperfusion injury, perhaps via  differing pathways. Sun et al.  30reported an attenuation of cardiac ischemia–reperfusion injury by S -nitrosoglutathione, which was associated with an increase of S -nitrosylation of the mitochondrial L-type Ca²⁺channel. Nitrite protected the heart when injected into the left ventricle at 5 min before the reperfusion31and produced cardioprotection by attenuating mitochondrial respiration via  inhibiting complex I in vitro .32Moreover, a recent study indicates that S -nitrosocysteine may protect the heart through nitric oxide–independent pathways.33Therefore, although nitric oxide itself may be the active molecule in cardioprotection, RSNOs and other nitric oxide metabolites seem to have a similar capability of protecting the heart.

Although our results and those of others suggest that inhaled nitric oxide can decrease cardiac ischemia–reperfusion injury in animal models, it is not known whether breathing nitric oxide will decrease MI size in patients experiencing an acute coronary artery occlusion (ST-segment elevation MI). It is encouraging to note that breathing 80 ppm NO has shown to decrease ischemia–reperfusion injury in human studies.8,9If our observations in mice can be extrapolated to humans, the findings suggest the possibility that brief durations of nitric oxide inhalation may prove beneficial in patients at risk for cardiac ischemia–reperfusion injury.

In summary, we report that inhaled nitric oxide dynamically increases the levels of blood and tissue nitric oxide metabolites and that the degree of accumulation of each nitric oxide metabolite is tissue specific. Moreover, brief periods of nitric oxide inhalation can reduce infarct size in a murine model of cardiac ischemia–reperfusion injury with similar efficacy as much longer periods of nitric oxide inhalation, suggesting that the concentrations of nitric oxide metabolites achieved in the target tissue may be more important for protection than the absolute amounts of nitric oxide absorbed or delivered. The protective effects of breathing nitric oxide are likely to be attributable to nitric oxide metabolites that are rapidly transported in a bioactive form via  the blood from the lung to the heart.

The authors thank Nathan S. Bryan, Ph.D. (Assistant Professor, Center for Cell Signaling, The Brown Foundation Institute of Molecular Medicine for the Prevention of Human Diseases, The University of Texas, Houston, Texas), Rong Liu, M.D., Ph.D. (Research Fellow, Division of Thoracic Surgery, Brigham and Women’s Hospital, Boston, Massachusetts), and Michael J. Raher, B.S. (Technician, Department of Anesthesia and Critical Care, Massachusetts General Hospital, Boston, Massachusetts), for their advice and skillful assistance; and Hui Zheng, Ph.D. (Assistant Professor of Medicine, Department of Medicine, Massachusetts General Hospital), for statistical assistance.