Nitrosyl-heme complexes are formed in the ischemic heart: evidence of nitrite-derived nitric oxide formation, storage, and signaling in post-ischemic tissues.

In addition to the generation from specific nitric-oxide (NO) synthases, NO formation from nitrite occurs in ischemic tissues, such as the heart. Although NO binding to heme-centers is the basis for NO-mediated signaling as occurs through guanylate cyclase, it is not known if this process is triggered with physiologically relevant periods of sublethal ischemia and if nitrite serves as a critical substrate. Therefore electron paramagnetic resonance studies were performed to measure nitrosylheme formation during the time course of myocardial ischemia and reperfusion and the role of nitrite in this process. Rat hearts were either partially nitrite-depleted by nitrite-free buffer perfusion or nitrite-enriched by preinfusion with 50 microm nitrite. Ischemic hearts loaded with nitrite showed prominent spectra of six-coordinate nitrosyl-heme complexes, primarily NO-myoglobin, that increased as a function of ischemic duration, whereas in nonischemic-controls these signals were not seen. Total nitrosyl-heme concentrations within the heart were 6.6 +/- 0.7 microm after 30 min of ischemia. Nitrite-depleted hearts also gave rise to NO-heme signals during ischemia, but levels were 8-fold lower. Nitrite-mediated NO-heme complex formation during ischemia was associated with activation of guanylate cyclase. Upon reperfusion, the levels of NO-heme complexes decreased 3-fold by the first 15 min but remained elevated for over 45 min. The decrease in NO-heme complex levels was paralleled by the formation of nitrate, suggesting the oxidation of heme-bound NO upon reperfusion. Thus, nitrite-mediated NO-heme formation occurs progressively during ischemia, with these complexes serving as a store of NO with concordant activation of NO signaling pathways.

Nitric oxide (NO), an inorganic free radical, is a biologically important signaling molecule involved in many physiological and pathological processes, including vasodilatation, inhibition of platelet aggregation, neurotransmission, and cytotoxic host defense (1). Increased intracellular NO formation during and following ischemia has been shown to be important in the pathogenesis of ischemia/reperfusion injury (2)(3)(4). In both the heart and the brain, extensive studies have definitively dem-onstrated markedly increased NO formation as well as the critical role of this excessive NO in the tissue damage after an ischemic event (2)(3)(4)(5)(6). In the early period of reperfusion, superoxide and superoxide-derived oxidants are also formed. It has been observed that NO reacts with superoxide to form the potent oxidant peroxynitrite that results in protein nitration and cellular injury (7,8).
It is known that NO is generated in biological tissues by specific nitric-oxide synthases (NOSs) 1 that metabolize arginine to citrulline with the formation of NO (9). However, studies over the last several years have also demonstrated additional pathways and mechanisms for NO synthesis in biological tissues. Within ischemic tissues in particular, it has been shown that NOS/L-arginine-independent pathways of NO formation occur (10,11). These are dependent on the reduction or disproportionation of the substrate nitrite, and with increasing periods of ischemia, where there is increasingly severe hypoxia and acidosis, these NOS-independent pathways become the major source of tissue NO production (10,12,13) There are several reaction mechanisms that have been shown to lead to NOS-independent NO production. These include acidotic degradation of tissue nitrite to NO (10,14) and reaction between arginine and H 2 O 2 to form NO (15) as well as the reduction of nitrite to NO catalyzed by microbial nitrite reductase (16). Recently, it has been reported that xanthine oxidoreductase-catalyzed reduction of nitrite to NO also occurs under hypoxic conditions (17)(18)(19)(20). The mitochondrial respiratory chain as well as other heme protein-containing reductase systems may also reduce nitrite, NO 2 Ϫ , to NO (21,22). The major biochemical mechanism by which NO exerts its signaling roles in the cardiovascular system is through its binding to heme proteins (23). NO activates soluble guanylate cyclase, sGC, by binding to its heme moiety, leading to the production of cGMP. cGMP is then responsible though activation of kinases for a reduction of calcium influx within the cells, leading to vasorelaxation (24). The half-life of NO in vivo typically varies from seconds to a few hundred milliseconds in tissues and is pO 2 -dependent (25). When bound to heme moieties, it forms highly stable ferrous nitrosyl-heme complexes whose distinctive spectra have been proposed to serve as a quantitative measure of NO generation (26 -29).
In prior studies we have observed the formation of five coordinate nitrosyl-heme complexes derived primarily from myoglobin, Mb, but only in necrotic heart muscle after very long periods of ischemia (10). These complexes have axial symmetry and a characteristic triplet splitting due to the hyperfine cou-pling of the nitrogen nucleus of the heme-bound NO. Up to the present time, however, intrinsic nitrosyl-heme complex formation has not been observed during short sublethal periods of ischemia, and it had been thought that these complexes would not accumulate in significant quantities except under the near anoxic conditions that occur with prolonged global ischemia (2,10,30).
In the present study we observe that with short periods of ischemia nitrite-derived nitrosyl-heme complex formation occurs and progressively increases as a function of ischemic duration. Upon reperfusion, levels of these complexes decline but remain elevated for prolonged periods. The magnitude and time course of this process is directly detected and characterized by EPR spectroscopy. It is shown that these nitrosyl-heme complexes have an important role as a store and mediator of NO-derived signaling.

EXPERIMENTAL PROCEDURES
Materials-Potassium iodide reagent ACS was obtained from Eastman Kodak Co. Vanadium (III) chloride was obtained from Aldrich. Dulbecco's phosphate-buffered saline was obtained form Invitrogen. Hydrochloric acid was purchased from Fisher. cGMP enzyme immunoassay kit was purchased from Amersham Biosciences. Sodium [ 15 N]nitrite was purchased from Cambridge Isotope Laboratories (Andover, MA). All the other reagents were obtained from Sigma.
Isolated Heart Perfusion-Female Sprague-Dawley rats (250 -300 g) were heparinized and anesthetized with intraperitoneal pentobarbital. The hearts were excised, and the aorta was cannulated. Hearts were perfused at a constant pressure of 80 mm Hg using Krebs bicarbonate buffer (17 mM glucose, 120 mM NaCl, 25 mM NaHCO 3 , 2.5 mM CaCl 2 , 0.5 mM EDTA, 5.9 mM KCl, 1.2 mM MgCl 2 ) bubbled with 95% O 2 and 5% CO 2 gas at 37°C. A sidearm in the perfusion line allowed direct infusion of NaNO 2 just proximal to the heart. To measure contractile function, a latex balloon was inserted into the left ventricular cavity and connected to a pressure transducer via a hydraulic line, and pressure was recorded with a Gould RS4000 recorder. The balloon was initially inflated to achieve an end-diastolic pressure of 8 -14 mm Hg. Left ventricular pressure, heart rate, and coronary flow were monitored throughout the period of perfusion.
Isolated hearts were initially perfused for 15 min to enable stabilization of contractile function and then subjected to given durations of 37°C global ischemia or ischemia followed by reperfusion. At the desired time points, hearts were rapidly frozen using liquid nitrogencooled Wollenberger tongs. The frozen tissue was maintained at 77 K in liquid nitrogen and fractured to 2-3-mm pieces and placed directly into an EPR finger Dewar. The EPR Dewar was filled to a sufficient height to fill the critical volume of the EPR cavity.
EPR Spectroscopy-Tissue was fractured to a particle size of ϳ2-3 mm and placed directly into the Dewar that was inserted into a TM 110 cavity of Bruker ER300 spectrometer operating at X-band. EPR spectra were recorded at 77 K using liquid nitrogen. The EPR spectra were obtained as an average of 10 scans with sweep time of 60 s with a modulation amplitude of 3.98 G, modulation frequency of 100 kHz, and 5.0 milliwatts of microwave power at a frequency of 9.786 GHz.
Quantitation of the total radical concentration in the tissue was determined from the ratio of the double integral of the observed signal to that of a myoglobin-NO standard of known concentration measured using identical EPR parameters. The relative ratios of the component signals were determined using a computer program designed to provide a best fit of the observed spectrum from a linear combination of the component signals. The component signals used were obtained from computer simulation of the components obtained on temperature annealing the tissue in a manner similar to that reported previously using an EPR simulation program for anisotropic systems (31). Both of these programs were run on Pentium PCs. The EPR data files were directly imported into the fitting program.
Chemiluminescence NO Analyzer-A Sievers 270B nitric oxide analyzer was used to measure nitrite and nitrate levels and was interfaced though a DT2821 A to D board to a PC as described previously (12,32). Data acquisition and analysis were performed using a custom-developed software (ANOMA, automated NO measurement and analysis) (12,32). The concentration of nitrite was determined by quantitative transformation of nitrite to NO performed using the reducing agent KI (1%) under acidic conditions (1 N HCl) in a glass-purging vessel equipped with a heating jacket. Measurement of the NO released into the gas phase was performed by measurement of the light emitted from the chemiluminescence reaction of NO with ozone in the analyzer. Quantification of nitrite was performed by comparison of NO release from standard solutions of nitrite to that from the assay samples (33). The concentrations of nitrate and nitrite, collectively, were determined using a reaction mixture consisting of 1% VCl 3 in 2 N HCl at 90°C. From the difference of the values obtained from this latter assay and that for nitrite the concentrations of nitrate were calculated.
cGMP Assay-cGMP was determined by enzyme-linked immunoassay using a kit from Amersham Biosciences according to the manufacture's specification. Briefly, hearts were freeze-clamped, and 300 mg of tissue was extracted in 6% trichloroacetic acid and centrifuged. Trichloroacetic acid was removed by diethyl ether (4 washes; 1 ml of extract, 12 ml of diethyl ether was used per each wash), and the samples were dried under nitrogen flow at 60°C. After appropriate dilution the samples were subjected to enzyme immunoassay utilizing a standard curve prepared with known amount of cGMP (curve range, 2-512 fmol/well).
Standards-Nitrosyl-myoglobin standards of were prepared under anaerobic conditions in an argon atmosphere by reacting sodium nitrite with excess horse deoxymyoglobin, 1 mM, in phosphate buffer in the presence of 10 mM hydrosulfite. Samples were frozen at 77 K in a 5-mm diameter cylindrical mold and then processed identically to the tissue with fracturing to a particle size of ϳ2-3 mm. These particles were then placed directly in the EPR Dewar for concentration determination. Sodium nitrite solutions were prepared in Krebs buffer immediately before use.
Statistical Analysis-Data are expressed as mean Ϯ S.E. unless noted otherwise. The statistical significance of differences between groups was calculated using two-way analysis of variance. p values of Ͻ0.05 were considered as significant.

RESULTS
Initial experiments were performed to determine whether nitrosyl-heme complexes could be detected in normal and ischemic heart tissue and to determine the role of nitrite in this process. As reported previously, nitrite concentrations in freshly isolated rats hearts are ϳ12 Ϯ 5 M and decrease about 4-fold after isolation and perfusion with nitrite-free perfusate for 15 min, the usual requisite period for stabilization of cardiac function (10). Intrinsic plasma or tissue concentrations of nitrite have been reported to vary over a wide range from submicromolar values to levels as high as 50 M (12, 34 -36). Therefore, to evaluate the role of nitrite on nitrosyl-heme complex formation, hearts were either perfused with normal nitrite-free perfusate (nitrite levels Ͻ 0.1 M) or preinfused with 50 M nitrite for 5 min before the onset of ischemia.
After cannulation and a 15-min period for functional stabi-lization and equilibration of contractile function, developed pressures of Ͼ110 mm Hg were observed, and hearts were then infused and loaded with 50 M NaNO 2 for 5 min. The hearts were then immediately freeze-clamped at 77 K or subjected to no-flow ischemia and then frozen. Normally perfused control hearts showed no detectable signal of nitrosyl-heme complexes (Fig. 1A). Only a relatively sharp signal is seen centered at about 3370 Gauss with a g value of 2.004 that has been previously identified as arising primarily from the mitochondrial ubiquinone radical (37,38) as well as a weak iron sulfur protein signal at about 3500 Gauss with a g value of 1.94 (39). When hearts not infused with nitrite were subjected to 30 min of ischemia no major nitrosyl-heme signal was seen (Fig. 1B). However, a very prominent nitrosyl-heme signal appeared when hearts preloaded with 50 M nitrite were subjected to 30 min of global ischemia (Fig. 1C). This broad signal spanning from 3200 to 3500 Gauss appears identical to that of the six coordinate NO-heme complex of NO-Mb that has characteristic g values g 1 ϭ 2.08, g 2 ϭ 2.01, and g 3 ϭ 1.98 as reported previously (40). On careful comparison of the spectrum of Fig.  1C to that in 1B, it appears that a small component of the NO-heme signal clearly seen in 1C is also present, contributing to the shoulders seen in 1B over the region from 3250 to 3450 Gauss in the non-nitrite-treated ischemic heart. Experiments were performed to determine the effect of the duration of ischemia on NO-heme complex formation. After 5 min of infusion of 50 M nitrite, hearts were subjected to periods of global 37°C ischemia of 0 -240 min. As shown in Fig.  2, whereas with control nonischemic hearts infused with nitrite no significant NO-heme signal was observed (Fig. 2, 0Ј), with even short periods of ischemia of 10 min (Fig. 2, 10Ј) prominent NO-heme signals were seen that progressively increased as a function of the duration of ischemia. In a matched series of hearts that were also preinfused with the NOS inhibitor N Gmonomethyl-L-arginine (1 mM for 5 min) that would totally block NO formation from NOS, no significant change was seen in the magnitude of NO-heme signals during ischemia. Standard complexes of NO-Mb were prepared to further verify the identity of the observed NO-heme signal and to enable quantitation of these signals in the heart tissue. As can be seen from the inset in Fig. 3, a linear relationship was observed for the EPR signal amplitude and NO-Mb complex concentration. Furthermore, the EPR spectrum of the NO-Mb standard is identical to that of the NO-heme spectrum observed in ischemic heart tissue. This was further verified by computer simulation and fitting of the observed spectra as described under "Experimental Procedures," and the NO-heme (NO-Mb) signals in heart tissue were quantitated. The magnitude of NO-heme formation as a function of ischemic duration was determined from the series of EPR measurements shown in Fig. 2. This graph shows that with nitrite loading prominent NO-Mb complex formation is present even with short periods of ischemia of 10 min with a level of 3.6 M and rises to values of 6.6 M at 30 min and 12.7 M at 240 min. For hearts that were not loaded with nitrite, NO-Mb complex formation was also observed during myocardial ischemia (Fig 4B). In a series of these hearts the levels of NO-Mb also progressively increased as a function of ischemic duration but were about 8 -10-fold lower than the nitrite-loaded hearts with concentrations of 0.5 or 1.6 M after 30 or 240 min, respectively.
Further experiments were performed to determine the effect of varying myocardial nitrite levels on the magnitude of NOheme formation during ischemia. NO-heme formation was measured in hearts subjected to 30 min of global ischemia after 5 min of preinfusion with nitrite concentrations in the range of 25-400 M. With these increasing nitrite loading-dose concentrations a linear correlation was observed between dose and amount of Mb-NO formed from 3.04 Ϯ 0.01 M in hearts loaded with 25 M nitrite to 27.0 Ϯ 0.3 M in hearts with 400 M nitrite (Fig. 5).
To further prove that the observed NO-heme complex formation was derived from nitrite, isotope tracer experiments were performed measuring nitrosyl-heme formation in hearts infused with isotopically labeled [ 15 N]nitrite. Because 15 N has a nuclear spin of 1 ⁄2, doublet hyperfine splitting rather than the smaller triplet hyperfine splitting characteristic of natural abundance 14 15 NO-Mb spectrum was seen (Fig 6B), whereas with [ 14 N]nitrite the typical 14 NO-Mb spectrum was observed (Fig. 6A).
Upon reperfusion, myocardial reoxygenation occurs, and it would be expected that the NO-heme complexes formed during ischemia would be oxidized and degraded. To determine the time course of this process and the levels of NO-heme complexes in the reperfused heart, experiments were performed in which hearts preinfused with 50 M nitrite were subjected to 30 min of global ischemia followed by periods of reperfusion from 1 to 45 min. An initial rapid decrease was seen in the observed NO-heme EPR spectra over the first 15 min of reperfusion followed by a slow decrease thereafter with 26% of the signal intensity persisting at 45 min of reperfusion (Fig. 7). As shown in Fig. 8, the NO-heme signal decays from a value of 6.8 M after 1 min to a value of 2 M at 15 min and over the following 30 min only exhibits a small decrease to 1.75 M. Thus the NO-heme signal exhibits biphasic decay with a rapid decrease in the early minutes of reperfusion and very slow decay thereafter.
In an effort to understand the overall effects of ischemia and reperfusion on the magnitude of NO-heme formation in the heart, a series of hearts (three at each time point) preloaded with 50 M nitrite were subjected to periods of control perfusion, ischemia of 10 -30 min duration, or 30 min of ischemia followed by reperfusion for periods of 1-45 min (Fig. 9). As clearly seen in this graph, the NO-heme levels rapidly rose from near 0, Ͻ0.1 M, before ischemia to high M levels after greater than 10 -30 min of ischemia. Then, upon reperfusion, these levels rapidly declined by about 70% over the first 15 min but only slowly declined thereafter and remained markedly elevated out to 45 min.
The decrease in NO-heme signal seen upon reperfusion could be due to oxidation or facilitated release/exchange of the NO bound at the heme site of Mb. With oxidation, nitrate formation would be observed, whereas for released NO, nitrite formation would be expected. Therefore, chemiluminescence NO analyzer studies were performed to measure the formation of the oxidized NO products nitrite and nitrate released from the heart in untreated or nitrite-preloaded hearts (50 M nitrite infused 5 min before the onset of ischemia). To allow wash-out of residual nitrite in the vasculature, effluent collection was only sampled after the initial 1 min of reperfusion. In the hearts preloaded with nitrite, nitrite levels were elevated only over the first minute of reperfusion. Nitrate levels remained elevated over the first 5 min of reperfusion and exhibited a time course of decrease that paralleled the observed decrease in NO-heme concentrations within the heart (Fig. 10).
As described above, the observed NO-heme formation is primarily due to the formation of NO-Mb complexes. Of note, trace amount of five coordinate NO-heme complexes were also seen as would arise from NO bound to guanylate cyclase (42, 43)

FIG. 4. Time course of Mb-NO complex formation during global 37°C ischemia.
Spectra were recorded at 77 K as described in Fig. 2, and quantification was performed by comparison of the double integral of the signals to that of Mb-NO standards of known concentration as detailed under "Experimental Procedures." Panel A shows the data for hearts loaded with 50 M nitrite for 5 min before the onset of ischemia, and panel B shows the data for a heart without nitrite loading. (Fig. 2), although these could also arise from many other heme centers. It has been reported that NO-Mb complexes are effective in activating sGC (44). Therefore, it would be expected that the nitrite-derived NO-Mb complexes formed in the ischemic heart would activate sGC. To determine whether sGC activation occurs, measurements of the sGC product cGMP were performed in hearts that were either normally perfused nonischemic, ischemic for 30 min (global 37°C ischemia), or subjected to 30 min of ischemia followed by 15 min of reperfusion. In each of these groups both untreated control and nitritetreated hearts preinfused with 50 M nitrite 5 min before the onset of ischemia were studied. At least three hearts were studied in each of these six groups. In non-ischemic hearts that were either nitrite-treated or untreated, only low levels of cGMP were seen. However, after 30 min of ischemia, increased levels of cGMP were observed (Fig. 11). A modest increase in cGMP levels was present in untreated hearts, whereas a large 4-fold increase was seen in nitrite treated hearts after ischemia compared with preischemic levels. Furthermore, in ischemic hearts the levels of cGMP were 2.5-fold higher with nitrite treatment than in otherwise identical untreated control hearts (Fig. 11). Upon reperfusion the levels of cGMP declined, and after 15 min of reperfusion more than a 2.4-fold decrease in cGMP levels was observed from ischemic values. These changes in cGMP levels paralleled the changes seen for NOheme levels. To determine whether the observed increase in cGMP levels in nitrite-treated hearts during ischemia was independent of NOS, an additional group of nitrite-treated hearts was studied that was subjected to NOS inhibition, 1 min of pre-infusion with 1.0 mM N G -monomethyl-L-arginine before the onset of 30 min of global ischemia. In these hearts, the levels of cGMP were not significantly different from hearts without NOS inhibition with a value of 1.6 Ϯ 0.3 fmol/mg wet weight compared with 2.0 Ϯ 0.4 fmol/mg wet weight, respectively.  A and B,  respectively). Hearts were infused with 250 M nitrite for 5 min before the onset of ischemia. Thus, the observed nitrite-derived NO-heme formation is paralleled by activation of myocardial-signaling pathways.

FIG. 6. EPR spectra of the NO-heme complexes formed in ischemic hearts loaded with 14 N or 15 N nitrite (panels
It has been previously reported that increased NO formation in the post-ischemic heart is accompanied by impaired recovery of contractile function (2,3). To further determine whether the observed nitrite-mediated NO-heme formation is accompanied by altered recovery of physiological function, measurements of the recovery of contractile function and coronary flow were performed in nitrite-treated and untreated control hearts, n ϭ 6 in each group. In nitrite-treated hearts, the recovery of contractile function as measured by left ventricular-developed pressure and rate-pressure product was significantly decreased, p Ͻ 0.01, compared with that in untreated control hearts, with final recovery at 30 min diminished by more than 3-fold (Fig. 12, A and B). In contrast to the impaired recovery of contractile function in nitrite-treated hearts, the recovery of coronary flow was actually slightly but not significantly higher than that in untreated control hearts (Fig. 12C). Thus, nitrite-mediated NO-heme formation during ischemia is paralleled by subsequent depressed recovery of contractile function upon reperfusion. In spite of this evidence of enhanced myocardial injury, coronary flow was not decreased but actually modestly increased. DISCUSSION For almost a decade it has been appreciated that alterations in nitric oxide formation and metabolism have a critical role in modulating the process of post-ischemic injury in biological tissues. Whereas it was initially reported that there was a loss of endothelial function and vasoreactivity in the post-ischemic heart, and this physiological data was interpreted to indicate a deficiency in NO formation, it was subsequently directly demonstrated by EPR and other analytical methods that NO generation is greatly increased in the heart as well as other tissues (2)(3)(4)(5)(6).
Whereas NO can activate cellular signaling pathways that exert beneficial cytoprotective effects, when present in excess or with the concurrent generation of oxygen radicals such as superoxide, it becomes a powerful mediator of cellular injury. In the heart it has been shown on reperfusion that concurrent superoxide and nitric oxide generation results in the formation of the potent oxidant peroxynitrite that in turn causes protein nitration and cellular injury (7,8).
In addition to the formation of NO in the heart via specific arginine-dependent NOSs, it was subsequently demonstrated that nitrite, typically considered as a NOS-derived NO oxidation product, can be reduced to form NO during ischemia. A number of enzymatic and nonenzymatic pathways have been shown to account for this nitrite reduction (10, 14 -22). It has been previously shown that large amounts of nitrite-mediated NO production induces myocardial injury (10). The role of this pathway in activating myocardial signaling, however, has been unclear.
Because NO signaling is exerted in large part through NOheme binding with activation of guanylate cyclase, in this study we undertook to characterize the process and time course of NO-heme formation in the heart during global ischemia and upon reperfusion. We had previously observed that NO-heme formation occurs in necrotic myocardium after long ischemic periods (Ͼ4 h of global 37°C ischemia). With this prolonged ischemia, prominent nitrite-derived five-coordinate NO-heme complexes were observed (10). However, NO-heme formation had not previously been detected in viable myocardium with shorter periods of ischemia. In view of the large concentrations of myoglobin in the heart and the high affinity of myoglobin for oxygen, P 50 ϭ 2.8 torr, it had been assumed that significant formation and accumulation of heme complexes might not occur until prolonged periods of ischemia, where the myocardium is almost anoxic. It has been shown, however, that the oxygen tension in the heart rapidly falls after the onset of ischemia and reaches very low levels, Ͻ200 millitorr, within 20 min (45). Therefore, one would expect that nitrite-derived NO-heme formation and accumulation could occur in viable ischemic heart tissue.
In the current study, we observed that prominent nitritederived NO-heme formation occurs even after short periods of ischemia of 10 min; however, no significant NO-heme signals could be detected in normally perfused myocardium. With increasing duration of ischemia the NO-heme signals progressively increased. The observed NO-heme signal with periods of ischemia of up to 4 h was attributable almost entirely to 6 coordinate NO-Mbs, derived primarily from nitrite. We have previously determined in freshly isolated rat hearts that intrinsic nitrite levels are 12.5 Ϯ 5 M (13). In isolated hearts the levels of nitrite were seen to drop about 4-fold with perfusion. In the current study, we observe that in untreated perfused hearts that are partially nitrite-depleted NO-heme levels are still detectable but much lower than those of hearts loaded with 50 M nitrite. Levels of NO-Mb also progressively increased as a function of ischemic duration with concentrations of 0.5 or 1.6 M after 30 or 240 min, respectively. After nitrite loading, very prominent NO-Mb complex formation was seen even with short periods of ischemia of 10 min with a level of 3.6 M and increased to 6.6 M at 30 min and 12.7 M at 240 min. Thus, the levels of NO-Mb progressively increased as a function of ischemic duration and were about 10-fold higher after loading with 50 M nitrite. Furthermore, with increasing nitrite loadingdose concentrations a linear increase was observed in the amount of Mb-NO formed (Fig. 5).
Because it has been previously demonstrated that both NOSdependent as well as NOS-independent NO generation are both stimulated in the ischemic heart, the heme-bound NO could be generated through NOS as well as through nitrite reduction. Therefore, a component of the observed NO-heme signal could arise directly from NOS-derived NO. However, clearly, nitrite is a major substrate source, as experiments with hearts subjected to NOS inhibition before the onset of ischemia showed no significant attenuation of NO-heme formation. Also, increasing levels of nitrite resulted in increased complex formation. Furthermore, isotope tracer studies using 15 N-labeled nitrite demonstrated that the NO was derived from nitrite. Thus, tissue nitrite levels are of critical importance in this process.
Plasma and tissue nitrite levels vary sharply and can be in the range of 0.5 to 50 M (12, 34 -36). Nitrite levels are con- trolled by a number of factors including diet and concentration of ambient NO in inhaled air as well as by production from NOS or other enzymes (46 -49). These nitrite levels can be raised by high dietary ingestion, pharmacological administration of organic nitrates, or other NO-donating therapeutics. In addition, pathological conditions associated with high levels of stress and inflammation such as sepsis that are accompanied by high levels of NOS induction are also accompanied by markedly increased nitrite levels (50 -52). Thus, the observed process of nitrite-derived NO-heme formation in ischemic myocardium would be expected to be even more prominent in the clinical setting of myocardial infarction, where patients are often subjected to other conditions of stress and on pharmacological treatments that raise nitrite levels.
The levels of NO-heme complexes formed in ischemic myocardium are quite high compared with the concentrations of free NO in cells of 10 to 100 nM or less (1). NO binding to the heme of sGC is the critical event triggering the activation of sGC that results in cGMP formation in the presence of GTP (24). It has been previously demonstrated that NO-Mb effectively donates NO-heme to sGC, resulting in activation of the enzyme with cGMP formation (44). Thus, the NO-Mb complexes observed as well as NO-hemes in general are both a store as well as a bioactive form of NO that would be anticipated to be able to have a signaling role. In view of this, we performed experiments to determine whether the observed NO-Mb formation in the ischemic heart was associated with activation of sGC. We observed with nitrite treatment that cGMP levels were increased to 250% that of the levels seen in similar untreated hearts. Thus, a link between the observed NO-heme formation and myocardial signaling via sGC activation was established.
It has been shown that with NO binding to sGC, a fivecoordinate heme complex with characteristic EPR spectrum with triplet splitting is observed (42,43). Indeed, with close examination of the observed NO-heme EPR spectra arising in ischemic myocardium, we observe that in addition to the large six-coordinate NO-Mb heme signal, a small component of five coordinate heme can be discerned, particularly with longer periods of ischemia (Fig. 2). Whereas this small signal could arise from NO-bound to sGC, it also could arise from a range of other five-coordinate heme complexes (53).
Upon reperfusion the concentration of NO-heme complexes decrease, as would be expected with the reoxygenation that accompanies reperfusion. Initially a rapid decrease was seen in the observed NO-heme EPR spectra over the first 15 min of reperfusion followed by a slow decrease thereafter, with 26% of the signal intensity persisting even after 45 min of reperfusion ( Figs. 7 and 8). The time course of the early decrease parallels the oxidant burst that occurs in ischemic myocardium. In the isolated rat heart, EPR spin trapping studies demonstrate that a burst of superoxide and superoxide-derived radical generation occurs over the early reperfusion period (3,31,38,54). It is possible that the early rapid decrease in NO-heme complexes is due to superoxide or other radical reaction with the bound NO. The subsequent slow decrease could be due to a slow rate of spontaneous oxidation secondary to the levels of oxygen present in the reperfused heart. Interestingly, the NO-heme complexes formed were surprisingly stable and persisted at M levels even after 45 min of reperfusion.
Chemiluminescence NO analyzer studies were performed to measure the formation of NO oxidation products during reperfusion. In hearts preloaded with nitrite, nitrite levels were elevated only over the first minute of reperfusion, whereas nitrate levels remained elevated over the first 5 min of reperfusion and exhibited a time course of decrease that paralleled the observed decrease in NO-heme concentrations within the heart. The decrease in NO-heme complexes seen upon reperfusion could be due to oxidation or facilitated release/exchange of the NO bound at the heme site of Mb. With oxidation, nitrate formation would be observed, whereas for released NO, nitrite formation would be expected. Thus, the observed data indicate that complex destruction is largely due to oxidation of the bound NO with nitrate production; however, during the first minute of reperfusion it is also possible that NO release occurs.
As noted above the observed NO-heme complexes can be attributed almost entirely to NO-Mb. Mb is present in high concentrations (0.4 -0.7 mM) in cardiac muscle (55). Intracellular concentrations of ferric or met-Mb are strictly controlled in vivo by a met-Mb reductase that maintains the heme iron in the reduced form and is able to bind molecular oxygen or form paramagnetic NO-heme complexes (56). Similar to oxyhemoglobin, oxymyoglobin is able to oxidize NO with nitrate production, and this process has been proposed to regulate and limit NO levels in skeletal and cardiac muscle (55,57). Recently, based on studies in Mb knockout mice showing no differences in exercise capacity but higher sensitivity to NO injury, a new role for Mb has been proposed as a critical regulator of NO (58).
In the current study, we observe that under ischemic conditions myoglobin binds and stabilizes nitrite-derived NO in the form of NO-Mb complexes, and these complexes serve as a store of NO. The NO-Mb formed during ischemia can subsequently exert sustained signaling actions through the activation of sGC and possibly by other mechanisms. Indeed, we observed that sGC activation accompanied the formation of NO-Mb complexes. This stored NO can also exert subsequent cellular and functional injury. Nitrite-mediated NO-heme formation was associated with impaired recovery of contractile function in the presence of a paradoxical mild elevation in coronary flow (Fig.  12). Thus, nitrite-mediated nitrosyl heme formation occurs in ischemic myocardium and can be an important regulator of myocardial signaling and injury in the post-ischemic heart.