INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

THE PRESENCE OF MYOGLOBIN (Mb) only in myocytes has always raised unsettling questions about its functional role (2, 28). Given the orthodox view of Mb as an O₂ storage protein ready to compensate any cellular O₂ deficit or as a facilitator of O₂ diffusion, one might expect hypoxic-sensitive brain tissue to sequester the protein. Brain tissue does not, and its absence raises the possibility that Mb may have other functions, which recent experimental observations have begun to intimate: under oxygenation conditions that far exceed mitochondrial demand, nitrite or CO inhibition of Mb function still produces a decline in myocardial O₂ consumption (MO₂) and phosphocreatine (PCr) (10, 13, 26). Under such conditions, Mb should presumably contribute insignificantly to facilitating O₂ diffusion, and respiration is not O₂ limited. Yet, if PCr level still declines, it would suggest that Mb inactivation has also removed a role for Mb in directly modulating respiration.

Much of the supporting data for a direct role of Mb in regulating respiration originates from CO or nitrite inhibition studies of myocytes. CO binds more tightly to the heme Fe than O₂ does, and nitrite oxidation of the heme iron from the physiological 2⁺ state [Fe(II)] to the 3⁺ state [Fe(III)] prevents any O₂ binding, because O₂ does not bind to Fe(III) heme. The nitrite experiments in perfused myocardium, however, have produced equivocal results (6, 9, 23). In a recent ¹H nuclear magnetic resonance (NMR) study that has followed the extent of nitrite inhibition of Mb in myocardium, infused nitrite concentration must exceed 10 mM before any noticeable Mb oxidation appears (6). The stoichiometry is much greater than the stoichiometry previously reported for Mb inhibition in vivo and far greater than the expected stoichiometric amount required for the corresponding in vitro reaction (6, 9, 10, 13, 21, 23). Although the rate-pressure product (RPP) and PCr levels progressively decline with increasing nitrite levels, MO₂ is relatively constant and even rises slightly. The results are in contrast to the myocyte observations and raise the question of whether nitrite acts directly on cellular function or acts indirectly by inhibiting Mb function and whether the disparity arises from any functional differences in the model systems.

Separating the contribution from Mb inactivation and nitrite is possible with CO experiments, because CO binds more tightly to the heme than O₂ does and inhibits the nitrite oxidation of the heme Fe(II) to Fe(III) state (2). Moreover, the CO inactivation experiments can yield another perspective into the functional role of Mb. If Mb plays a significant role in directly regulating O₂ consumption or oxidative phosphorylation, then CO inactivation of Mb, under conditions when O₂ supply is ample and the Mb role in facilitated O₂ diffusion is minimal, should still produce a physiological response that will appear to signal O₂ limitation.

Measuring the extent of CO binding to Mb in the myocardium is then a crucial step. We report herein that the distinct MbCO signal of the -CH₃ Val E11 at 2.26 parts per million (ppm) is detectable in the myocardium, separated from the corresponding MbO₂ signal at 2.76 ppm (15, 20). Increasing the partial pressure of CO (PCO) in the perfusate enhances the MbCO signal intensity and at the same time depresses the MbO₂ signal intensity.

The physiological and metabolic impact of CO in the cell is somewhat unexpected. Up to 80% of MbCO saturation, the MO₂ remains constant, whereas RPP is depressed by 10%. The ³¹P NMR spectra reveal that PCr, ATP, and P_i levels are not significantly disturbed. Although pH is constant, lactate formation rate has increased by a factor of two. At the CO concentration to saturate 80% MbCO, cytochrome oxidase activity should not be impaired, yet MO₂ begins to decline. On transient infusion of 50 mM nitrite in the presence of CO, MbCO is not oxidized; yet, RPP and lactate level shift dramatically. The results are consistent with a potential direct role of Mb in regulating respiration and a direct route for nitrite action.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Animal preparation and heart perfusion. Male Sprague-Dawley rats (350-400 g) were anesthetized by an intraperitoneal injection of pentobarbital sodium (60 mg/kg) and heparinized (1,000 U/kg body wt). The heart was quickly isolated and placed in an ice-cold buffer solution until aortic cannulation. It was then perfused in a retrograde mode, as prescribed by a modified Langendorff model, and was maintained at 35°C with a Lauda MT-3 water bath and temperature-jacketed reservoirs and tubings. A peristaltic pump (Rainin Rabbit) maintained a constant, nonrecirculating perfusion flow of 18 ml/min. A saline-filled latex balloon inserted in the left ventricle monitored the heart rate (HR) and left ventricular pressure (LVP) via a strain-gauge transducer (Statham P23XL) connected to an oscillographic recorder (Gould RS 3200). The balloon volume was adjusted to give an end-diastolic pressure (EDP) of 6-8 mmHg. RPP values were calculated from HR times the LV developed pressure (LVDP).

Perfusate O₂ measurement. The heart was placed in an NMR tube and isolated with a Teflon plug with holes to permit perfusate overflow. Approximately 50% of the perfusate was withdrawn via a polyethylene (PE) catheter inserted close to the pulmonary artery. A Yellow Springs Instrument (YSI) 5300 meter monitored the perfusate O₂ concentration with two YSI 5331 O₂ electrodes in a temperature-jacketed chamber (one for inflow and the other for outflow perfusate). The remaining 50% of the perfusate exited the chamber above the Teflon plug as an overflow.

CO infusion protocol. CO was introduced into the perfusion buffer in a stepwise manner. Two flowmeters controlled the flow of 95% O₂-5% CO₂ and 95% CO-5% CO₂ gases, which entered a temperature-jacketed gas-mixing chamber and equilibrated with the perfusate passing through 50 ft of gas-permeable Dow Corning Silastic tubing (ID 0.058 in., OD 0.077 in.). The CO flow rate varied stepwise at 0.0, 0.1, 0.3, 0.5, and 1.0 l/min, while the 95% O₂-5% CO₂ flow rate remained constant at 5.0 l/min. After correction was made for loss in the tubings, the resulting PCO values at the catheter tip were 0.0, 12.6, 36.3, 58.4, and 107.0 Torr, respectively. O₂ measurement of perfusate exiting the gas-mixing chamber confirmed that equilibration was essentially complete within 2 min after the CO flow rate was adjusted at each step.

Nitrite infusion protocol. In the transient infusion protocol, periods corresponding to a CO infusion, a nitrite infusion, and an O₂ reperfusion followed the control interval. During the control period, the myocardium was perfused with nitrite-free, O₂-saturated buffer flowing at 18 ml/min. CO was then introduced for 30 min at a PCO sufficient to saturate 86% of the Mb. The buffer was then switched to one containing 50 mM nitrite buffer, which was kept in a separate reservoir, which contained a specific amount of NaNO₂ dissolved in O₂-saturated perfusate. The Na⁺ concentration was adjusted to maintain the proper ion balance. After the transient nitrite infusion, reflow with O₂-saturated, nitrite-free buffer began at 18 ml/min. During the entire protocol the ³¹P NMR followed the high-energy metabolite changes. Physiological monitors continuously tracked the EDP, LVDP, HR, and MO₂.

Lactate measurement. A YSI 2700 Bioanalyzer determined the perfusate lactate concentration. Samples were measured in triplicate, and the analyzer's linear response was calibrated against a set of standard lactate solutions, ranging from 0.01 to 20 mM. The membrane current stabilized at <2 nA before any measurement commenced. An additional calibration curve, derived from buffer perfusate at different nitrite concentrations, corrected for any nitrite-dependent interference.

Curve fitting and statistical analysis. Linear regression analysis, using a least-squares method (SigmaPlot, Jandel Scientific), determined the correlation coefficient, slope, and intercept. Errors were noted as standard error. Student's t-test indicated statistical significance when P < 0.05.

NMR. An AMX 400-MHz Bruker spectrometer recorded ¹H/³¹P signals with a 20-mm ¹H-(X) probe, where X represented nuclei from ¹⁵N to ³¹P. A modified binomial pulse sequence suppressed the H₂O line and selectively excited the MbO₂ and MbCO Val E11 resonances at 2.76 and 2.26 ppm, respectively (5, 15). The ¹H 90° pulse was 65 µs, calibrated against the perfusate H₂O signal. Observing the MbO₂ and MbCO signals required a 40-ms acquisition time and a 45° pulse. The spectral width was set at 8,065 Hz; the data block size was 512. Six thousand transients were averaged for a typical ¹H spectrum, requiring 5 min of signal accumulation. The free induction decays (FID) were then zero-filled to 2 K and multiplied by an exponential-Gaussian window function, F(t) = exp(t/a) + exp(t²/b), where a and b are input constants. A nonlinear spline fit (Bruker UXNMR algorithm), based on zero points set at regions well removed from the peaks of interest (data points at least 5 times the half-height line-width excursion from the peak maximum), then smoothed the baseline. All spectral lines were referenced to the H₂O resonance at 4.67 ppm at 35°C. The chemical shift was in turn calibrated against sodium-3-(trimethylsilyl)propionate-2,2,3,3-d⁴ as 0 ppm. The integrated area of the Val E11 signal at 18 ml/min flow rate was normalized as 100% MbO₂ saturation. For the ³¹P spectra, a typical spectrum utilized a 45° pulse angle, a 0.5-s repetition time, and 512 scans/block (4.3 min). The ³¹P 90° pulse was 72 µs, calibrated against a 0.1 M phosphate solution. Spectral width was set at 6,494 Hz; the data size was 4 K. FID were apodized with an exponential function to improve the ³¹P signal-to-noise ratio. The ³¹P signals were referenced to PCr as 0 ppm and apodized with a 15-Hz exponential function.

2

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

The myocardial response to increasing PCO is reflected in the ¹H and ³¹P spectra (Fig. 1, A and B, respectively). During the control period the ¹H NMR spectra from well-oxygenated myocardium, perfused with 95% O₂-5% CO₂ saturated buffer flowing at 18 ml/min, exhibit a distinct -CH₃ Val E11 signal of MbO₂ at 2.76 ppm (Fig. 1A). Under these perfusion conditions, the signal reflects the fully saturated state (5, 14, 15). As the PCO of the perfusate increases, the MbO₂ signal decreases, while the corresponding MbCO peak at 2.26 ppm increases (Fig. 1A, spectra b and c). At PCO of 12.6 Torr, the MbO₂ signal decreases to 53.5% of control level, while the MbCO signal increases correspondingly (Fig. 1A, spectrum b). Increasing the PCO to 58.4 Torr decreases further the MbO₂ signal to 20.3% of control level, while MbCO peak rises to 84.9% (Fig. 1A, spectrum c). On reperfusion with 95% O₂-5% CO₂ saturated buffer, the MbO₂ signal recovers as the MbCO signal falls (Fig. 1A, spectrum d). In contrast the ³¹P NMR spectra show no response to the infused CO (Fig. 1B, spectra a'-d').

View larger version (15K): [in this window] [in a new window]   Fig. 1.   1H and 31P nuclear magnetic resonance (NMR) spectra of myocardium perfused under varying partial pressures of CO (PCO). A: 1H NMR spectra. a) PCO 0 Torr: control spectrum from myocardium perfused with CO-free, O2-saturated buffer flowing at 18 ml/min shows the -CH3 Val E11 signal of oxymyoglobin (MbO2) appearing at 2.76 parts per million (ppm), which reflects the fully oxygenated state. b) PCO 12.6 Torr: with 12.6 Torr CO infusion, the 1H spectrum shows a modest decrease in the -CH3 Val E11 signal of MbO2. However, a signal corresponding to the -CH3 Val E11 of MbCO emerges at 2.26 ppm. c) PCO 58.4 Torr: MbO2 signal is now significantly reduced, whereas signal of MbCO at 2.3 ppm becomes prominent. d) PCO 0 Torr: on reperfusion with CO-free, O2-saturated buffer, MbCO signal intensity decreases, whereas the -CH3 Val E11 signal of MbO2 recovers its intensity. The MbCO and MbO2 signals are in dynamic equilibrium. Signal intensity loss in 1 peak corresponds to signal intensity gain in the other. B, a'-d': 31P NMR spectra under conditions that correspond, respectively, to those in A, a-d.

Throughout a range of PCO, a dynamic equilibrium exists between MbO₂ and MbCO with no significant intervening Mb species (Fig. 2A). Changes in the MbCO signal intensity are balanced by corresponding alterations in the MbO₂ signal, such that the sum of the MbO₂ and MbCO signals is relatively constant, as denoted in Fig. 2A (dotted line). At PCO of 20 Torr, 50% of the Mb has converted to MbCO. A plot of the fractional MbCO/MbO₂ vs. PCO/PO₂ yields a linear relationship and a partition coefficient of 36, which is consistent with the values from in vitro studies (Fig. 2B).

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Relationship of intracellular MbO2 and carbon monoxymyoglobin (MbCO) as a function of PCO (Torr). A: concomitant change in MbO2 and MbCO as a function of PCO. Dotted line indicates sum of MbCO and MbO2 intensities, which remain constant. B: PCO/PO2 vs. MbCO/MbO2. The partition coefficient is 36.

The in vivo MbCO off-rate kinetics are shown in Fig. 3A. At PCO of 58.4 Torr, 84.9% of intracellular Mb is sequestered as MbCO. On reperfusion with oxygenated buffer, the MbCO declines to 50% of its original intensity within 10 min. After 40 min, the MbCO signal is no longer detectable, whereas the MbO₂ signal has returned to its control level. The time for equilibration and dead space volume clearance is <2 min, a time frame sufficient for the PCr signal to recover fully (A Glabe, S. Huang, and T. Jue, unpublished observations).

View larger version (10K): [in this window] [in a new window]   Fig. 3.   Kinetics of MbCO release in cell (A) and rate-pressure product (RPP) as well as myocardial O2 consumption (MO2) relationship as a function of MbCO saturation (B). A: time course of MbCO dissociation in myocardium on reperfusion with O2-saturated buffer. Apparent 1st-order off-rate time constant is 1.16 × 103. B: MO2 and RPP vs. %Mb saturation. MO2 shows no change until MbCO saturation exceeds 80%, whereas RPP has already exhibited a significant decline.

The MO₂ and RPP response to varying PCO is shown in Fig. 3B. MO₂ remains constant up to 76.8% MbCO saturation. At 87.6% MbCO saturation, MO₂ shows a significant decline (34.0 ± 1.3 µmol · min¹ · g dry wt¹). In contrast, RPP has already dropped significantly at 53.5% MbCO saturation (27,436 ± 2,483 mmHg/min) and remains at this depressed level up to 87.6% MbCO saturation.

Despite the increasing PCO, the high-energy phosphate signals (ATP, PCr, and P_i) show no alteration (Fig. 4A). Although pH remains constant, the lactate formation rate has increased sharply to 191% (0.495 ± 0.149 µmol · min¹ · g dry wt¹) of control, when MbCO is at 76.8% saturated (Fig. 4B). The metabolic and physiological responses to varying levels of PCO under non-O₂-limiting conditions are summarized in Table 1.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Metabolic responses in myocardium as a function of %MbCO. A: ATP, phosphocreatine (PCr), and Pi show very little change until MbCO saturation is well above 80%. B: although pH remains constant, lactate formation rate has increased dramatically when MbCO is above 60% saturated.

Nitrite does not significantly oxidize Mb in CO-treated myocardium, as shown in Fig. 5A. Infusion of 50 mM NaNO₂ in the continuing presence of CO, sufficient to saturate 87.6 ± 3.7% MbCO, does not produce any metmyoglobin (metMb) signal at 3.9 ppm nor does it decrease the MbCO signal significantly. On reperfusion with oxygenated, nitrite- and CO-free buffer, the MbCO signal disappears as the MbO₂ signal recovers (Fig. 5A).

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Effect of nitrite on MbCO and on myocardial metabolism when Mb is not oxidized to metmyoglobin (metMb). A: after control, CO is introduced to saturated Mb at 86%. Sodium nitrite (50 mM) is then introduced. It does not significantly perturb the MbCO signal intensity. On reperfusion with CO and nitrite-free, oxygenated buffer, MbCO signal gradually disappears and MbO2 signal emerges. B: ATP, PCr, and Pi response indicates that CO does not perturb these parameters. However, on nitrite infusion, PCr level drops, whereas Pi concentration increases. During reperfusion, both Pi and PCr do not recover fully to their control levels. ATP remains constant throughout the entire experimental protocol.

During nitrite perturbation in the presence of CO, some of the high-energy phosphate levels are altered. With only CO infusion, the PCr, ATP, and P_i levels are constant. Once nitrite is infused, the PCr level declines to 59.5 ± 3.4% of control, whereas P_i concentration rises to 124 ± 17.6%. ATP concentration, however, remains constant (Fig. 5B). After 30 min of reperfusion, the PCr level reaches 113 ± 2.1% of control, whereas P_i falls to 41.7% of control.

The pH and lactate formation response to nitrite infusion in the presence of CO is shown in Fig. 6A. pH remains constant at 7.14 throughout the control and CO addition period but declines to 7.07 on NaNO₂ infusion. It recovers to 7.14 after reperfusion with oxygenated, nitrite-free buffer (Fig. 6A). The lactate formation rate stays constant during the control period (0.184 ± 0.024 µmol · min¹ · g dry wt¹) but increases during CO addition to 0.590 ± 0.059 µmol · min¹ · g dry wt¹. With NaNO₂ infusion the lactate concentration rises dramatically to 44.6 ± 7.91 µmol · min¹ · g dry wt¹. With reperfusion the lactate level declines rapidly and approaches the control level. After 30 min of reperfusion the lactate level reaches 1.06 ± 0.230 µmol · min¹ · g dry wt¹.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Nitrite effect on myocardial metabolic and physiological responses when Mb is 86% inhibited by CO and cannot convert to metMb. A: CO does not perturb pH but does elevate lactate formation rate by a factor of 2, from 0.184 ± 0.024 to 0.590 ± 0.059 µmol · min1 · g dry wt1. With 50 mM sodium nitrite infusion, pH drops from 7.15 to 7.06, while lactate formation rate rises dramatically. On reperfusion with CO and nitrite-free, oxygenated buffer, pH recovers to control level, but lactate formation rate does not. B: corresponding MO2 and RPP responses indicate that nitrite itself reduces RPP level and gradually boosts MO2.

MO₂ and RPP drop slightly during CO perfusion, from 36.8 ± 1.7 to 34.0 ± 1.3 µmol · min¹ · g dry wt¹ and from 29,760 ± 1,444 to 27,796 ± 1,198 mmHg/min, respectively. With 50 mM NaNO₂ infusion, MO₂ rises to 117% of control level after 30 min, whereas RPP declines to 39% of control. pH drops to 7.06 as PCr drops to 59.5 ± 3.4% of control. With reperfusion MO₂ returns to its control level; however, RPP remains depressed at 74% of control. The metabolic and physiological parameters are listed in Table 2.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Partitioning of CO and O₂. Analyzing the cellular function of Mb requires a characterization of its ligand as well as oxidation states and entails a correlation with the physiological-biochemical response (5, 6, 14). Although optical techniques can distinguish MbO₂ saturation in vitro, they are not as successful in vivo in discriminating the overlapping MbCO and MbO₂ bands or distinguishing the Fe(II) from the Fe(III) states in a beating heart (6, 22). In contrast the ¹H NMR CH₃ Val E11 signal offers a unique opportunity to observe directly the MbO₂, MbCO, and metMb states in the myocardium. The CH₃ Val E11 signal can mark both the intracellular PO₂ and the PCO. At 2.76 ppm the signal of MbO₂ reflects the oxygenated state and decreases its intensity on deoxygenation (15). With increasing PCO, the MbO₂ signal declines as the corresponding MbCO signal emerges at 2.26 ppm (Fig. 1A). Moreover, the metMb reporter signal at 3.9 ppm reflects any nitrite oxidation of MbCO to the Fe(III) state (6, 16).

Critical O₂ level. The Mb and cytochrome oxidase have contrasting ligand binding affinities for O₂ and CO. For CO the Mb [PCO]₅₀ is 0.06 Torr, whereas cytochrome oxidase [PCO]₅₀ is ~0.1 Torr. For O₂ the cytochrome oxidase [PO₂]₅₀ is ~10 times lower than the corresponding Mb [PO₂]₅₀ (19, 30). A similar conclusion emerges on comparison of the inhibition ratio R = CO/O₂, which will produce 1:1 binding of CO:O₂ to the protein. For Mb the reported ratio is 0.025-0.04, whereas for cytochrome oxidase it is 5-15 (7, 8, 25, 29). At the Mb [PCO]₅₀, CO binds insignificantly to cytochrome oxidase and therefore does not perturb the MO₂ or the PCr level. At 87.6% MbCO saturation, MO₂ begins to decline significantly; yet, CO/O₂ is only 1/10, well below the 6/1 required to detect any decline in respiration arising from CO inhibition of cytochrome oxidase (7, 8, 27). It would appear that the drop in MO₂ does not arise from any direct CO inhibition of cytochrome oxidase.

Intracellular ligand kinetics. Although the MbCO/MbO₂ partition coefficient in vivo is identical to the one in vitro and supports a nonselective CO/O₂ transport into the cell, the CO off-rate can reveal insight into the property of Mb in the cell. In particular, the difference in Mb ligand binding properties in solution vs. in the cell remains an open question. During reperfusion with oxygenated buffer, MbCO level drops to 50% of its original level within 10 min, indicating an apparent first-order rate constant (k_off) of 1.2 × 10³ s¹ at 37°C. The in vivo k_off value is somewhat lower than the in vitro value of 1.7-4 × 10² s¹ at 22°C. Additional experiments will be required to determine the origin of the MbCO kinetics, which may indicate the presence of an unidentified cellular effector (11, 18).

CO and contractile energy coupling. With CO, 76.8% saturation of MbCO does not significantly impair MO₂. Yet, the lactate formation rate has increased by a factor of two, well below the reported value of the ratio CO/O₂ required to inhibit cytochrome oxidase (8, 19, 22). No shift in MO₂ appears until the MbCO saturation reaches 87.6%. PCr, ATP, pH, and P_i levels still remain constant. These observations are not completely consonant with the myocyte results, which note a decline in both MO₂ and PCr above a 40% MbCO saturation threshold (6, 10, 13, 26).

CO and mitochondrial energy coupling. In the second phase of the CO-induced effect, lactate formation rises dramatically despite constant energy production and utilization. Lactate oxidation is assumed to be constant. Above 53.5% MbCO saturation, MO₂, PCr, and ATP remain unaltered, whereas RPP maintains its depressed, but steady-state, level. The lactate formation rate appears to increase as a function of MbCO saturation in a dose-dependent manner and implicates potentially the presence of anaerobic ATP production. Consistent with such a view is an enhanced glycolytic ATP production to meet an energy deficit, which can arise if the ratio P/O has shifted. From the MO₂ and the ³¹P spectra, oxidative phosphorylation activity appears normal. Under all experimental conditions, the cellular PCO is insufficient to arrest the cytochrome oxidase activity. This view of Mb interaction is consistent with previous CO myocyte studies, which have suggested that Mb may play a direct role in regulating respiration (26).

Nitrite vs. metMb effect in modulating respiration. Many experiments supporting a direct Mb role in respiration utilize myocytes and nitrite oxidation (10, 13). Perfused hearts subjected to nitrite have not responded exactly as myocytes. In fact, nitrite-infused myocardium reveals a complex interaction (5). Even with 30 mM infused nitrite, a concentration that far exceeds the in vitro stoichiometry to oxidize Mb from Fe(II) to Fe(III), MO₂ shows no significant alteration, whereas RPP and PCr levels decrease as lactate formation rate increases. Because MO₂ is constant, the subsequent PCr loss in the face of a stimulated glycolytic ATP production or altered redox state suggests strongly that Mb may indeed mediate energy coupling in respiration (6).

1

1