INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE REGULATION of myocardial oxidative phosphorylation has been extensively studied, yet our understanding of this important process remains incomplete. The assessment of steady-state fluxes and metabolite levels has provided much insight into metabolic regulation; however, dynamic changes in metabolite concentrations and metabolic fluxes have proven to be essential ingredients in understanding oxidative phosphorylation. Unfortunately, the measurement of such important parameters has proven to be complex and difficult (10). Analysis of the response time of oxygen consumption in the isolated heart is a recently developed technique that, indeed, allows the study of the dynamic regulation of oxidative phosphorylation (16). In the isolated rabbit heart, for example, it has been found that oxidative phosphorylation responds within 4-12 s after an increase in ATP hydrolysis. Because data show that phosphocreatine and inorganic phosphate change within a few seconds after a step in heart rate, the 4- to 12-s response time indicates a delay between ATP hydrolysis at the myofibril and ATP synthesis via oxidative phosphorylation. The delay found in cardiac muscle, however, is considerably smaller than that found in skeletal muscle, where the response time is often >30 s. Addition of pyruvate to a glucose buffer has been found to quicken the mitochondrial response to a step in heart rate, suggesting a glycolytic interaction in the speed of the response time (6, 14). Furthermore, it has recently been demonstrated that the mitochondrial response to a heart rate step is slowed in failing hearts (2) and after myocardial stunning (21). Such information gives new insight into the role of mitochondrial function in intact myocardium during pathophysiological conditions, but the role of macromolecules needs further elucidation.

With the advent of genetically engineered mice, the determination of the mitochondrial response time to heart rate steps in the mouse heart can provide much needed information regarding the role of specific proteins (e.g., creatine kinase) in the activation of oxidative phosphorylation and the intracellular transport of high-energy phosphoryl groups. It was, therefore, the purpose of this study to develop this technique for the buffer-perfused mouse heart and, as a first comparison with results obtained from the rabbit heart, to examine the mitochondrial response time under the influence of glucose alone as substrate and the combination of glucose with various concentrations of pyruvate.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Experimental preparation. All experimental procedures were approved by the Institutional Animal Care and Use Committee of the Vrije Universiteit. We studied 10 male or female FVB mice weighing between 26 and 48 g. Each mouse was briefly, yet profoundly, anesthetized with Avertin (2,2,2-tribromo-ethanol; 0.4 g/kg ip). Absence of reflexes was achieved within 3-5 min after injection. A tracheotomy was performed, and the mouse was ventilated with 100% oxygen. After the thorax was opened, heparin (200 IU) was injected intravenously. Under a microscope, the aorta was cannulated in situ with a blunted and grooved 20-gauge needle, whereby perfusion was initiated before the heart was excised. The hearts were Langendorff-perfused at 37°C with Tyrode solution containing (in mM) 128.3 NaCl, 4.7 KCl, 2.5 CaCl₂, 1.05 MgCl₂, 20.2 NaHCO₃, 0.42 NaH₂PO₄, and 11 glucose, gassed with 95% O₂-5% CO₂. Adenosine (10 µM) was added to the perfusate to ensure maximal vasodilation of the preparation. Constant flow was maintained and was set to between 2 and 4 ml/min to achieve a perfusion pressure of 80-100 mmHg. A small flanged length of polyethylene (PE)-100 tubing was introduced through the apex of the left ventricle to ensure thebesian drainage. A fluid-filled line attached to a latex balloon was then drawn through the mitral valve and out the apex and was connected to a Statham P23 Db pressure transducer (Statham Instruments) for the measurement of left ventricular developed pressure (LVDP). Fluid was added to the balloon to achieve an end-diastolic pressure between 5 and 10 mmHg. LVDP was derived by subtracting the end-diastolic pressure from the systolic pressure. The pulmonary artery was cannulated with PE-100 tubing, and the venous effluent was drawn at a rate of 1 ml/min across a fast Clark-type oxygen electrode (time constant = 1 s) mounted in a cuvette (volume of tubing and cuvette = 75 µl) for the measurement of coronary venous oxygen tensions. The ventricular surface of the heart was submerged, and the heart was electrically stimulated via an electrode attached to the metallic aortic cannula and an electrode placed in the submersion bath. The basal pacing rate was set to 270 beats/min. Data were sampled at 100 Hz, or at 1,000 Hz for the determination of rate of change in LVDP (dP/dt), and were recorded on-line with a personal computer.

Measurement of the response time of oxygen consumption. The venous response time (t_v) is determined from the time course of the venous oxygen tension to a step in heart rate (Fig. 1). To derive the true response time of mitochondrial oxygen consumption (t_mito), we subtracted the mean time necessary for oxygen diffusion and vascular transport (t_transport) from t_v: t_mito = t_v  t_transport. t_v and t_mito are precisely mathematically defined as the time integral of the normalized change in venous oxygen concentration and mitochondrial oxygen consumption, respectively, over the duration of the transient. The final, full amplitude of the change is normalized to one. The exact mathematical definitions of t_v, t_mito, and t_transport have been described in detail previously (17). t_transport was experimentally determined in this study from the venous oxygen response to a small step (<10%) in perfusion flow, as described previously for the rabbit heart (16, 17). In addition, it is necessary to adjust the response time for a delay in the product of the heart rate and LVDP (i.e., rate-pressure product) (t_RPP), since an upward step in heart rate results in transient changes in left ventricular systolic pressure and beat-to-beat RPP (Fig. 2). An overshoot of RPP often results in a negative t_RPP, for correction with the final equation: t_mito = t_v  t_transport  t_RPP. A full description of this technique has been given for the rabbit heart (16, 17). Oxygen consumption (µmol · min¹ · g dry wt¹) was calculated as the product of the flow collected from the pulmonary artery and the arterial-venous oxygen concentration difference. An example of the recordings for left ventricular pressure and venous oxygen tensions for a step from 270 to 400 beats/min during glucose-only perfusion is shown in Fig. 2.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Schematic representation of the determination of the mean response time of mitochondrial oxygen consumption to a step in heart rate. At time t = 0, a pacing step increase is made from 270 to 350 beats/min. Increased ATP hydrolysis at the myofibril and at ion pumps leads to increased mitochondrial oxygen consumption. Integration of the time course of the venous oxygen curve, followed by subtraction of the transport time, gives the mean response time of mitochondrial oxygen consumption (tmito).

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Recordings of left ventricular pressure (LVP) and coronary venous oxygen tension after a step increase in pacing rate from 270 to 400 beats/min (at t = 20 s, left dashed line), followed by the step back from 400 to 270 beats/min (right dashed line).

Experimental protocol. The response time of oxygen consumption was studied under the influence of substrates and pacing frequencies. Steps in pacing frequencies were made either from 270 to 350 beats/min or from 270 to 400 beats/min in the presence of 11 mM glucose with or without added pyruvate (1, 3, or 6 mM). To prevent any confounding time effects, we randomly prepared the hearts in the presence of either 11 mM glucose alone or 11 mM glucose plus 1 mM pyruvate, whereafter the higher concentrations (3 and 6 mM) of pyruvate were then studied. While the hearts did become more stable (i.e., less arrhythmic) during the length of the experiment, no time effects were found on the venous response times.

Statistical analysis. Data are presented as means ± SE. Differences among all groups were analyzed statistically using one-way ANOVA and Tukey's post hoc test. Linear regression was performed to determine the slope and intercepts of the relationship between oxygen consumption and RPP. Differences were considered statistically significant at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Mean LVDP was 82.3 ± 7.4 mmHg for glucose-perfused hearts paced at 270 beats/min. Normalized values (i.e., glucose-only perfusion at 270 beats/min LVDP = 100%) are shown in Fig. 3. LVDP decreased (P < 0.05) at both 350 and 400 beats/min during glucose perfusion. No other significant differences with heart rate or substrate were found among the groups, although LVDP tended to decline when the concentration of pyruvate was increased to 6 mM from 1-3 mM. The metabolic demand on the heart, on a beat-by-beat basis, is reflected by RPP and is defined as the product of LVDP and heart rate. RPP increased (P < 0.05) from 20,900 ± 1,600 mmHg/min at 270 beats/min to 24,500 ± 2,100 mmHg/min at 350 beats/min and to 27,100 ± 2,900 mmHg/min at 400 beats/min (P < 0.05) during glucose perfusion. The addition of 1 mM pyruvate increased the RPP to 23,100 ± 2,500, 26,900 ± 3,000, and 27,800 ± 2,600 mmHg/min at 270, 350, and 400 beats/min, respectively. Addition of 3 mM pyruvate increased RPP to a similar degree (22,400 ± 2,300, 24,400 ± 3,100, and 27,100 ± 2,500 mmHg/min), while addition of 6 mM pyruvate resulted in RPPs similar to that seen with glucose-only perfusion (21,000 ± 3,500, 23,400 ± 3,700, and 24,700 ± 3,300 at 270, 350, and 400 beats/min, respectively); however, none of the differences among the heart rates tested were statistically significant when pyruvate was present. Oxygen consumption was 19.1 ± 1.4 µmol · min¹ · g dry wt¹ (270 beats/min, glucose-only buffer) and was increased (P < 0.05) during glucose perfusion when heart rate steps were made from 270 to 350 beats/min (21.6 ± 1.8 µmol · min¹ · g dry wt¹) and from 270 to 400 beats/min (24.1 ± 2.3 µmol · min¹ · g dry wt¹) (Fig. 4). Addition of 1, 3, or 6 mM pyruvate did not significantly increase oxygen consumption compared with glucose-only perfusion at any heart rate level.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Normalized left ventricular developed pressure (LVDP) measured with a latex balloon positioned in the left ventricle during pacing at 270, 350, and 400 beats/min and with glucose only, glucose + 1 mM pyruvate, glucose + 3 mM pyruvate, or glucose + 6 mM pyruvate. Data are expressed as a percentage of the value obtained at 270 beats/min with glucose-only substrate. *P < 0.05 vs. 270 beats/min, glucose only. No significant differences with heart rate were found for any pyruvate concentration.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Normalized myocardial oxygen consumption (MO2) during pacing at 270, 350, and 400 beats/min and with glucose only, glucose + 1 mM pyruvate, glucose + 3 mM pyruvate, or glucose + 6 mM pyruvate. Data are expressed as a percentage of the value obtained at 270 beats/min with glucose-only substrate. *P < 0.05 vs. 270 beats/min, glucose only. For pyruvate, there was no statistically significant difference among heart rates. Substrate levels also had no significant effect (P > 0.05).

The coronary venous oxygen pressure was ~370 mmHg, and the perfusion flow was ~40 ml · g dry wt¹ · min¹. There values are similar and higher, respectively, than in the isolated rabbit heart, and, together with the relatively high developed pressure, this indicates adequate levels of perfusion. After the perfusion flow was lowered by 8-10%, there was no significant decrease of oxygen consumption. This direct test thus indicated adequacy of oxygenation, as shown previously for the rabbit heart preparation (17).

To examine the relationship between the metabolic demand and the oxygen consumption under various substrate conditions in the mouse heart, we determined the correlation between oxygen consumption and RPP as shown in Fig. 5. During perfusion with the four substrate conditions, oxygen consumption did indeed increase with an increased RPP (P < 0.05), and the 6 mM pyruvate group exhibited higher oxygen consumption for the given RPPs, compared with that of the other groups as determined by linear regression (same slope, higher y-intercept; P < 0.05).

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Relationship between MO2 and rate-pressure product (RPP) under the influence of glucose plus various pyruvate concentrations. Regression lines have similar slopes, but the y-intercept for glucose + 6 mM pyruvate differs (P < 0.05) from that for glucose only. r2 = 0.99, 0.98, 0.99, and 0.83 for glucose only, glucose + 1 mM pyruvate, glucose + 3 mM pyruvate, and glucose + 6 mM pyruvate, respectively.

The mitochondrial response time to upward steps in heart rate, under the influence of the various substrate concentrations, is shown in Fig. 6. As detailed in the methods, t_mito = t_v  t_transport  t_RPP, where t_v is the total measured response time and t_transport is the delay of the change in the venous oxygen signal with respect to the mitochondria. t_transport varied between 5 and 15 s, decreasing with flow and the quality of the preparation (i.e., leakage along the cannula of the pulmonary artery), t_v varied between 10 and 25 s, and t_RPP varied between 1 and 4 s. During steps in heart rate from 270 to 350 beats/min, t_mito was calculated to be 9.8 ± 1.3 s during perfusion with glucose alone. This value decreased to 8.4 ± 0.6 and 8.1 ± 1.5 s during perfusion with glucose plus 1 mM pyruvate or 3 mM pyruvate, respectively (P > 0.05). During perfusion with 6 mM pyruvate plus glucose, t_mito was significantly decreased to 5.3 ± 1.1 s. Heart rate steps from 270 to 400 beats/min presented a similar picture: t_mito = 11.0 ± 1.1, 9.4 ± 1.4, 8.8 ± 1.1, and 7.4 ± 1.1 s for the glucose-only, glucose plus 1 mM, glucose plus 3 mM, and glucose plus 6 mM pyruvate groups, respectively; however, the value for the glucose plus 6 mM pyruvate group was not statistically different from that for the glucose-only group.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Mean response times of tmito to a stepwise increase in pacing rate under the influence of various concentrations of pyruvate (0, 1, 3, and 6 mM) in combination with 11 mM glucose. Steps were made from 270 to 350 beats/min (solid bars) and from 270 to 400 beats/min (open bars). *P < 0.05 vs. glucose only. Other differences with heart rate and substrate level were not statistically significant.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The purpose of this study was to develop a sensitive and reliable technique for the measurement of the dynamic regulation of oxygen consumption in the isolated mouse heart. To this end, we adapted and miniaturized an existing technique that was developed and extensively validated in the isolated rabbit heart (for review, see Ref. 16). The major finding of this study is that the mitochondrial response time (t_mito) to a step in heart rate in the mouse heart is similar to that found in the rabbit heart. We report here a mitochondrial response time of 9.8 s during perfusion with 11 mM glucose alone. t_mito has been shown to vary between 4 and 12 s during glucose perfusion at 37°C in the rabbit heart depending on age and other conditions (16). Mice are known to have a very high mitochondrial content and capacity in the myocardium. For example, 37% of the mouse heart is composed of mitochondria compared with 29% for the rabbit heart (1). Concomitantly, in vivo mouse myocardial oxygen consumption, expressed as moles of O₂ per gram of tissue per minute, is three times higher than that found for the rabbit, reflecting a much higher heart rate (1). Yet, apparently, the fact that mice have a very high myocardial oxidative capacity does not influence t_mito. This finding is consistent with the previous finding that dobutamine treatment does not change the response time to a heart rate step in the rabbit heart, despite an increased oxygen consumption and developed pressure (15). Conversely, it has also been shown that a decrease in mitochondrial aerobic capacity due to oligomycin treatment has no effect on t_mito (18). Hence, the postulation that the mitochondrial capacity is not an important determinant of t_mito, but that signaling and conduction through the cytosol play a more important role (16), is supported by these data from the mouse heart.

Furthermore, the addition of pyruvate to glucose buffer has been shown to accelerate the speed by which the mitochondria respond to a step in heart rate (6, 14). We report here a t_mito of ~10 s during perfusion with 11 mM glucose alone, which becomes significantly faster (5.7 s) when 6 mM pyruvate is added to the glucose buffer. In the rabbit heart, t_mito is 40% faster when pyruvate is given as the exogenous substrate (16). Because lactate substrate sometimes showed the same behavior as glucose substrate, it was proposed that the reducing equivalents, which result from glycolysis or the conversion of lactate to pyruvate, and which need to be transported into the mitochondria, may slow the response time. An alternative hypothesis, however, maintains that glycolysis locally buffers a sharp increase in ATP hydrolysis and, thus, slows the mitochondrial response to a heart rate step (4, 6). Because pyruvate bypasses and even inhibits glycolysis, this would explain the acceleration of the mitochondrial response time. This hypothesis is supported by the finding that titrated inhibition of glycolysis at glyceraldehyde-3-phosphate dehydrogenase with iodoacetate indeed quickened the response time (7, 16).

The contractile performance of the isolated crystalloid-perfused hearts examined in this study was similar to that found previously by others. Kameyama et al. (9) measured left ventricular peak pressures of 88 mmHg at end-diastolic pressures of 15 mmHg, while Van Dorsten et al. (19) found left ventricular pressure to decrease from 70 to 55 mmHg when pacing was increased from 400 to 600 beats/min. These values are in good agreement with our findings. The systolic left ventricular pressure in isolated perfused hearts may be slightly smaller than that found in vivo for similar end-diastolic pressures (20). Although an ejecting isolated mouse heart is feasible (3), the higher degree of control over constancy of coronary flow and cardiac diastolic volume in the isolated Langendorff-perfused heart is highly desirable for the measurement of t_mito (17).

A descending limb of a force-frequency relationship has often been noted in mouse heart preparations. This "negative staircase" phenomenon has been noted in isolated hearts (current study and Ref. 13) as well as in open- (8, 12) and closed-chest (8) preparations. Potential causes for the negative staircase that have been examined include atrial-ventricular conduction and calcium handling. Gao et al. (5) have examined contractile activation in mouse trabeculas and report positive force-frequency relations when stimulation rates were increased from 0.2 to 4.0 Hz along with proportional rises in intracellular calcium concentrations. These experiments, however, were performed at room temperature and subphysiological heart rates; therefore, the adequacy of calcium handling in experimental mouse preparations at physiological frequencies and temperature remains to be determined. Our data show a clear decrease in developed pressure when steps were made from 270 to 350 beats/min and from 270 to 400 beats/min (Fig. 3). This effect remained during perfusion with pyruvate. The decreases in developed pressure were, however, to such a slight degree (i.e., ~10-15%) that a positive increase in RPP was observed after a step in heart rate and, thus, resulted in a concomitant rise in the oxygen consumption. In mouse hearts perfused with red blood cells, the negative force-frequency relationship is shifted to higher heart rates, preceded by a positive force-frequency relationship (11). Thus the negative force-frequency relationship may be characteristic of the isolated heart perfused with saline solutions.

The value of t_mito is the same for the steps to 350 and 400 beats/min with glucose when the concentration of added pyruvate is 3 mM or lower. As is the case in the rabbit heart at 37°C, t_mito does not depend on heart rate with glucose as substrate (16). t_mito reflects the time course of aerobic ATP synthesis and does not include glycolytic contributions to ATP synthesis. The t_mito value in the mouse heart is similar to the t_mito value found in the rabbit heart (16). The adaptation of cardiac oxidative phosphorylation to metabolic demand thus occurs at a similar speed in the mouse and rabbit heart, despite the substantially higher natural heart rate in mice.

The relationship between oxygen consumption and RPP shown in Fig. 5 indicates that oxygen consumption rises linearly with increases in RPP in our mouse heart preparation. This has also been found by others in wild-type as well as creatine kinase-deficient mice (12). The hearts perfused with the highest concentration of pyruvate (i.e., 6 mM) were less economical than those at lower pyruvate concentrations (Fig. 5). In other words, the 6 mM pyruvate group had a higher oxygen consumption for the same RPP (P < 0.05). This may indicate the importance of glycolytic support of cellular metabolism, since pyruvate inhibits glycolysis by increasing citrate levels and the ATP-to-AMP ratio, which inhibit phosphofructokinase.

In summary, we have presented a technique for the study of myocardial oxidative phosphorylation in the mouse. The results show that the mean response time of oxidative phosphorylation to a step to a higher heart rate is ~10 s. The response time decreases to ~6 s when pyruvate is added to the glucose perfusate. The data are quite similar to those previously found in the rabbit heart, indicating that the high mitochondrial content and capacity of the mouse heart does not greatly influence the response time under the conditions studied here. This study lays the groundwork for further studies with genetically altered mice that will provide important information regarding the dynamic regulation of oxidative myocardial metabolism.