Background The pyruvate dehydrogenase (PDH) enzyme complex determines the extent of carbohydrate oxidation in the myocardium. PDH is in a largely inactive state during early reperfusion of postischemic myocardium. The resultant decrease in pyruvate oxidation in postischemic hearts has been documented with 13C nuclear magnetic resonance (NMR) spectroscopy. This study demonstrates that counteracting depressed pyruvate oxidation can enhance contractile recovery in the absence of increases in either glycolytic activity or glucose oxidation. The findings indicate that increased incorporation of carbon units from pyruvate into the intermediates of the oxidative pathways by PDH influences the metabolic efficiency and mechanical work of postischemic hearts.
Methods and Results Isolated rabbit hearts were situated in an NMR magnet and perfused or reperfused (10 minutes of ischemia) with 2.5 mmol/L [3-13C]pyruvate as sole substrate to target PDH directly and bypass the glycolytic pathway. Hearts were observed with or without activation of PDH with dichloroacetate. Mechanical function and oxygen consumption (MV̇o2) were monitored. 13C and 31P NMR spectroscopy allowed observations of pyruvate oxidation and bioenergetic state in intact, functioning hearts. Metabolite content and 13C enrichment levels were then determined with in vitro NMR spectroscopy and biochemical assay. PDH activation did not affect performance of normal hearts. Postischemic hearts with augmented pyruvate oxidation (dichloroacetate-treated) sustained improved mechanical function throughout 40 minutes of reperfusion. Rate-pressure-product (RPP) increased from 8300±1800 (mean±SEM) in untreated postischemic hearts to 21 300±2400 in hearts treated with dichloroacetate (P<.05). Oxygen use per unit work [MV̇o2 multiplied by 104 divided by RPP] was improved from 1.50±0.13 to 1.14±0.11 (P<.05) without differences in high-energy phosphate content between treated and untreated hearts. Values of dP/dt were also consistently higher, by as much as 185%, during reperfusion with dichloroacetate. Postischemic hearts displayed reduced pyruvate oxidation from the incorporation of 13C into the tissue glutamate pool. With the tissue alanine level as a marker of 13C-enriched pyruvate availability in the cell, the ratio of labeled glutamate to alanine was only 58% of the control value during early reperfusion. With dichloroacetate, that ratio was 167% greater than that of untreated hearts (P<.05). By the end of the reperfusion period, the 13C enrichment of the tissue glutamate pool by pyruvate oxidation was elevated from dichloroacetate treatment (untreated, 62±7%; DCA-treated, 81±6%; P<.05), but glycogen content was similar in both groups and 13C enrichment of tissue alanine remained unchanged, near 60%, indicating no increases in glycolytic end-product formation.
Conclusions Metabolic reversal of contractile dysfunction was achieved in isolated hearts by counteracting depressed PDH activity in the postischemic myocardium. Improved cardiac performance did not result from, nor require, increased glycolysis secondary to the activation of PDH. Rather, restoring carbon flux through PDH alone was sufficient to improve mechanical work by postischemic hearts.
The pyruvate dehydrogenase (PDH) enzyme complex stands at the crossroads of glycolytic end-product formation (ie, pyruvate) in the cytosol and the oxidation of carbohydrate metabolites in mitochondria. During early reperfusion of the postischemic myocardium, destined to suffer postischemic contractile dysfunction, the PDH enzyme complex is largely in the phosphorylated inactive state. From isotope-release experiments in isolated rat heart, Patel and Olson1 determined that PDH activation was 45% of total enzyme content in ischemic myocardium. Kobayashi and Neely2 performed in vitro assay of PDH enzyme from postischemic myocardium and determined that 45% of the myocardial PDH was inactive during the first 2 minutes of reperfusion, but then PDH activity normalized to control levels after continued reperfusion. Evidence for these changes in PDH activity within intact, reperfused hearts was then provided by Lewandowski and Johnston3 from 13C nuclear magnetic resonance (NMR) spectroscopy observations of the relative extent of [3-13C]pyruvate oxidized to form 13C-labeled glutamate versus the nonoxidative conversion of labeled pyruvate to [3-13C]alanine.4 This observation of pyruvate metabolism with 13C NMR in reperfused hearts was later confirmed in the presence of free fatty acids as an alternative fuel.5 Although the inactivation of PDH was found to be only a transient phenomenon, work from other laboratories6 7 has demonstrated that stimulation of glucose oxidation by PDH activation within postischemic myocardium sustains improved contractile recovery throughout reperfusion.
Recent work by McVeigh and Lopaschuk7 and Lopaschuk et al8 suggests a beneficial effect of increased flux through PDH on cardiac performance that is more directly related to the extent of carbohydrate oxidation within the mitochondria than to associated increases in glycolysis and glycolytic energy production. While such stimulation of carbohydrate oxidation has been found to alleviate depressed cardiac performance during reperfusion of isolated hearts,6 7 8 whether in the absence of commensurate glycolytic activity the activation of PDH itself could counter postischemic contractile dysfunction has yet to be determined. Reperfusion with pyruvate, in relatively high concentrations of 5 to 10 mmol/L, has been reported to improve contractile recovery of reperfused hearts,9 10 11 but in some cases only in the presence of added glucose.9 10 The investigators of these studies also cite the possibility that substrate-induced activation of PDH1 12 may account for the beneficial effect of high-pyruvate concentrations with supplemental glucose in the reperfusion media.
In view of the recent findings cited above, we sought to examine the role of PDH oxidation on its immediate substrate, pyruvate, in supporting the mechanical work performance of postischemic, reperfused hearts in the absence of commensurate increases in glycolysis. Thus, the experiments described here were designed to focus on the influence of PDH on postischemic contractile recovery and the respiratory efficiency of mechanical work in reperfused myocardium, without limitations of substrate availability or the complications of accompanying glycolytic flux. Particular emphasis was placed on the utility of the PDH kinase inhibitor dichloroacetate (DCA) in reversing the observed trends of reduced pyruvate oxidation and postischemic contractile dysfunction that were reported in our previous 13C NMR analysis of substrate oxidation in postischemic rabbit hearts.3 In addition, other studies have indicated a potential inotropic action of pyruvate metabolism in enhancing the contractile recovery of reperfused myocardium through elevation of bioenergetic potential within the myocytes, as evidenced by high-energy phosphate levels.9 10 Therefore, an additional aim was to examine the energetic state of the intact myocardium in response to DCA-induced changes in mechanical work.
A combination of detection of high-energy phosphate levels with 31P NMR and 13C NMR spectroscopy of 13C-labeled pyruvate metabolism was examined in the intact, isolated rabbit heart during physiological recordings of function and respiratory rate under normal and postischemic conditions. Since DCA is known to increase the carbon isotope enrichment of the oxidative metabolites of isotope-labeled pyruvate,4 13 tissue concentrations and isotope enrichment levels of key metabolites by in vitro analyses were also monitored for proper validation and interpretation of the potential metabolic changes reflected in 13C NMR spectra of intact functioning hearts. As presented below, the findings demonstrate a beneficial effect of countering the initial reduction of PDH activity in the reperfused myocardium that is not related either to glycolytic activity or to heightened phosphorylation potential.
Isolated, Perfused Rabbit Heart Preparation
Hearts were perfused in an NMR magnet by use of previously described methods.4 14 Use of rabbits conformed to the guiding principles of the American Physiological Society and the Massachusetts General Hospital. Briefly, hearts were excised from Dutch belted rabbits (600 to 750 g body wt) that were given an injection of heparin 1200 IU IP and then anesthetized with an injection of sodium pentabarbital 100 mg/kg IP. Immediately upon excision, the heart was immersed in a solution containing 20 mmol/L KCl and 120 mmol/L NaCl for cardioplegia at 0°C. The aorta was attached to a 100-cm hydrostatic perfusion column, and retrograde perfusion was begun. Hearts were perfused with a modified Krebs-Henseleit buffer at 37°C, containing (mmol/L) 116 NaCl, 4 KCl, 1.5 CaCl2, 1.2 MgSO4, 1.2 NaH2PO4, and 25 NaHCO3. The buffer was equilibrated with 95% O2/5% CO2 and recirculated. The initial perfusate supply contained 5 mmol/L glucose within a 2-L reservoir. This perfusate supply was changed at the start of each 13C enrichment protocol to a 450-mL reservoir of Krebs-Henseleit buffer containing 2.5 mmol/L [3-13C]pyruvate (Isotec Inc) with no glucose. Heart temperature was maintained at 37°C by a control unit interfaced to the NMR system console and by the temperature of the perfusion medium. Hearts beat spontaneously, contracting against a fluid-filled intraventricular balloon that was connected to a pressure transducer. The balloon was inflated to an end-diastolic pressure of 5 to 10 mm Hg. Heart rate (HR) and left ventricular developed pressure (LVDP) were recorded continually. Mechanical work was assessed by the rate-pressure-product (RPP equals HR multiplied by LVDP), and dP/dt was determined from LVDP as a measure of contractility. Oxygen content (Po2) of the perfusion medium and coronary effluent was determined with a blood-gas analyzer (International Laboratories). During either normal or postischemic perfusion, myocardial oxygen consumption (MV̇o2) was determined for each heart from the difference between the oxygen contents of the perfusate at the aortic cannula and the coronary effluent, as described elsewhere.14 15
Perchloric acid extraction of tissue metabolites was performed on the frozen ventricles. The acid extracts were neutralized, and the tissue contents of alanine, glutamate, and lactate were determined by UV spectrophotometric analysis.4 16 Tissue glycogen from frozen ventricle muscle was extracted and assayed according to the methods outlined by Bartley and Dean.17 Acid extracts of 1-g tissue samples were lyophilized and then reconstituted in 0.5 mL D2O for in vitro NMR analysis described below.
NMR data were collected on a Bruker series MSL400 spectrometer interfaced to a 9.4-T vertical-bore superconducting magnet (Bruker Instrument). Field homogeneity was adjusted by shimming on proton signal from the sample to a proton (H2O) line width of 8 to 20 Hz for isolated hearts and 1 to 2 Hz for in vitro samples. 31P and 13C NMR spectra were obtained from isolated hearts perfused within a broadband, 20-mm NMR probe (Bruker Instruments) equipped with a proton decoupling coil. In vitro 13C and proton spectra were acquired with a 5-mm probe (Bruker Instruments). Signal intensity of resonance peaks was determined by integration of the area under each peak by use of the analysis subroutine within NMR-dedicated software (DISMSL, Bruker Instruments, and NMR1, Tripos Associates, Inc).
13C NMR spectra were obtained sequentially from hearts during perfusion with 13C enriched pyruvate; 31P spectra were obtained at the midpoints and end points of the perfusion periods. NMR data were collected and processed as described in previous reports.3 5 13C spectra were acquired at 101 MHz from intact hearts in 5-minute time blocks (152 scans, 45° pulses). Such 13C NMR signals were proton-decoupled with power-gated decoupling (0.5 to 6 W) to avoid sample heating. Before 13C enrichment, a 13C natural abundance was acquired from each heart for digital background subtraction from spectra collected during 13C enrichment. NMR signals were processed by exponential filtering with a line broadening of 20 Hz followed by Fourier transformation. Peak assignments were referenced to the known resonance of the exogenous, 13C-enriched substrate and the well-documented glutamate and alanine resonance signals relative to dioxane at 67.4 ppm. Changes in signal intensities due to nuclear Overhauser enhancement or relaxation effects were minimal under these pulsing conditions, as previously discussed.18 19 20 The 13C NMR signal that arose from the initial 4-carbon site of labeling within glutamate, as [3-13C]pyruvate was oxidized, was then monitored over time and reference to the invariant signal arising from the 3-carbon position of alanine, which serves as a relative index of pyruvate availability in the cell.3 4 5 Thereby, 13C spectra from intact, functioning rabbit hearts could be examined over time for the continual oxidation of intracellular pyruvate to form metabolites of the oxidative, intermediary pathways.
31P spectra were recorded at 161 MHz by accumulating 32 transient signals (45° pulse) for 1 minute, as performed in previously published studies from this laboratory.3 Raw 31P signals were converted into frequency domain spectra by Fourier transformation after application of an exponential filter (15-Hz line-broadening) to enhance the signal-to-noise ratio. Chemical shift assignments for 31P signals were made relative to 85% phosphoric acid at 0 ppm, a convention set by the International Union of Pure and Applied Chemists (IUPAC). With this convention, the phosphocreatine peak is observed to be −2.5 ppm. Relative ATP levels over the course of each protocol were assessed from the intensity of the β-phosphate peak at −18.5 ppm within each spectrum.
13C NMR spectra were collected from tissue extracts that were lyophilized and then reconstituted in 0.5 mL D2O. The raw 13C NMR signal was collected at 37°C by use of 45° excitation pulses with 2-second interpulse delays during broadband proton decoupling. The composite free induction decay was initially obtained within a 8-kiloword data set, which was then increased to 32 Kw (zero-filling) to improve digital resolution of the transformed data and processed with a gaussian filter. From in vitro samples, the multiplet structure of the four-carbon site of glutamate (Glu C-4) resonances within high-resolution 13C spectra allowed the percentage of labeled acetyl groups entering the TCA cycle (Fc) to be calculated19 for each heart.
The fractional enrichment of the initial site of labeling within Glu C-4 was determined as previously described.20 13C enrichment at the Glu C-4 was determined by comparison of the signal intensity from the corresponding resonance peak within each 13C spectra to a standard 100-mmol/L (1.1 mmol/L 13C natural abundance) glutamate solution. The amount of labeled material was referenced to known total tissue concentration of glutamate from assay.20
In vitro proton spectra were collected over 30 minutes (45° pulse, 4-second interpulse interval) from the lyophilized tissue extracts reconstituted in 0.5 mL D2O. High-resolution proton spectra were used to determine the fractional 13C enrichments of alanine and lactate by the extent of proton signal from the 3-carbon methyl groups that were split due to J-coupling to 13C versus unsplit signals arising from unlabeled methyl groups, as described elsewhere.4 21 The enrichment values obtained by this method were verified by comparison of 13C signal from the 3-carbon position of alanine (Ala C-3) to the known concentration of alanine from enzymatic assay, as performed for determination of glutamate enrichment. A 13C NMR spectrum from a standard 100-mmol/L (1.1 mmol/L 13C natural abundance) alanine solution was used as a reference signal for 13C-enriched alanine levels in the tissue extracts, as described for glutamate. Values obtained from proton spectra were within 10% of values determined by comparison of 13C spectra from samples to the standard.
Experiments were performed on both normal and postischemic hearts. At the start of the protocol, the perfusate supply to normal hearts was switched from the original glucose-containing buffer to a similar buffer containing 2.5 mmol/L 99% [3-13C]pyruvate without glucose. Hearts were perfused for 40 minutes with 13C-enriched pyruvate alone (n=10) or in combination with 5 mmol/L DCA (n=10), an inhibitor of PDH kinase.22 In the postischemic groups, after initially being perfused with glucose, hearts were subjected to 10 minutes of global zero-flow ischemia followed by 40 minutes of reperfusion with either 2.5 mmol/L 99% [3-13C]pyruvate (n=10) or 2.5 mmol/L [3-13C]pyruvate combined with 5 mmol/L DCA (n=10). In this manner, the influence of PDH activity alone was examined on the metabolic and contractile recovery of postischemic hearts. At the onset of reperfusion, coronary effluent was collected and discarded for the first 2 minutes to avoid recirculation of accumulated lactate that was washed out of the myocardium, as determined from previously acquired data of lactate washout in this reperfusion model.5
At the end point of each experiment, all hearts were rapidly frozen within a liquid nitrogen–cooled clamp. The frozen ventricular tissue was used as described above to determine the tissue concentrations of key metabolites, the 13C enrichment levels of NMR detectable metabolites, and the extent of labeled substrate utilization at the citrate synthase step of the tricarboxylic acid (TCA) cycle.
A separate set of experiments was also performed to determine the glutamate levels in isolated hearts at several additional time points within the ischemia-reperfusion protocol. In vitro biochemical assays were performed on ventricle tissue harvested immediately after the brief, 10-minute episode of zero-flow ischemia (n=4) and at 3 minutes of reperfusion with either pyruvate alone (n=5) or pyruvate plus DCA (n=5). This time point was selected to correspond to the midpoint of the first time-averaged 13C NMR spectrum of the intact hearts. Particular attention was devoted to detection of potential changes in glutamate content that may have influenced changes in 13C NMR signal intensities over the course of the experimental protocol.
Statistical analysis of the data was performed with a computer software program for scientific applications (instat, GraphPad). Intergroup analyses were performed by Student’s unpaired, two-tailed t test. Comparison of intragroup data sets was performed by Student’s paired, two-tailed t test. Changes in NMR data at each time over the 30-minute perfusion period in each group of hearts were evaluated by repeated-measures ANOVA. Differences in mean values were considered statistically significant at P<.05. Data are presented as mean±SEM.
Recovery of Contractile Function During Reperfusion
Mechanical function, assessed by HR, LVDP, and dP/dt, was similar among untreated control hearts and DCA-treated hearts during normal perfusion. The mean value for dP/dt at the midpoint of the protocols was 2113±394 mm Hg/min for the untreated normal hearts and 2465±223 mm Hg/min for the group that received DCA during normal perfusion. This lack of influence of DCA on cardiac performance in normal myocardium is reflected in the RPP values shown in Table 1⇓, and is consistent with a previous finding from this laboratory.4 Table 1⇓ and Fig 1A⇓ and 1B⇓ also display mechanical performance indexes of hearts reperfused after 10 minutes of global, zero-flow ischemia. Unlike normal myocardium, reperfused hearts displayed a significant mechanical response to treatment with DCA. Hearts reperfused with pyruvate in combination with DCA displayed marked improvements in both dP/dt and rate pressure product (RPP) compared with untreated hearts. Activation of PDH with DCA induced normal levels of dP/dt during reperfusion, whereas untreated hearts subjected to the same duration of ischemia displayed significant depression of dP/dt, reaching a mean end-point value that was only 54% of the corresponding value from the DCA-treated postischemic group (Fig 1A⇓). Hearts treated with DCA during reperfusion retained a significantly greater percentage of their preischemic dP/dt values than did untreated postischemic hearts (Fig 1B⇓). At the midpoint of reperfusion, DCA-treated hearts displayed a 158% increase (P=.0062) in RPP over that in the untreated postischemic group (Table 1⇓). Enhanced recovery of RPP during reperfusion was the result of increased LVDP among DCA-treated hearts, with no difference in HR between any of the experimental groups. These data demonstrate the ability of pyruvate without added glucose to support enhanced mechanical performance in postischemic myocardium in response to reperfusion with DCA.
Metabolic Efficiency of Mechanical Work
The phosphocreatine-to-ATP content ratio of the postischemic reperfused myocardium was significantly higher than that of hearts perfused under normal conditions (Table 1⇑) owing to a reduction in ATP content over the course of the 10-minute ischemia, as reported in earlier studies.3 5 15 Also consistent with a previous study,4 the phosphocreatine to ATP ratio in the normal heart was not affected by DCA treatment. This lack of influence on high-energy phosphate content by DCA was also observed in postischemic hearts, despite the improved contractile recovery among hearts reperfused with DCA. The similarity in 31P spectra between both the untreated and DCA-treated postischemic hearts is demonstrated in Fig 2⇓. During the ischemic period, ATP levels dropped to 55% to 66% of preischemic levels. Neither phosphocreatine nor ATP changed significantly over the course of reperfusion in either group, despite marked differences in mechanical function between untreated and DCA-treated postischemic hearts.
Mean values for oxygen consumption ranged from 23 to 26 μmol/min per gram dry wt of tissue, for both untreated and DCA-treated normal hearts and for postischemic group that was treated with DCA. A significant difference was only in evidence in the untreated postischemic group, which consumed oxygen at the reduced rate of 12 μmol/min per gram dry wt (P=.0017). The respiratory efficiency of mechanical function (MV̇o2/RPP) of the untreated postischemic hearts was also significantly impaired (P=.0016). MV̇o2/RPP values are displayed in Table 1⇑. DCA treatment was effective for improving the respiratory efficiency of postischemic hearts compared with hearts in the untreated postischemic group (P=.041). Postischemic hearts reperfused with DCA showed no statistically significant difference in MV̇o2/RPP compared with the nonischemic hearts.
13C Enrichment and Metabolite Content
A representative 13C spectrum acquired at steady state isotope enrichment from a normal heart perfused with [3-13C]pyruvate is displayed in Fig 3⇓. No particular changes in the detectable resonances that arose from metabolism of the 13C-labeled pyruvate were in evidence as a result of DCA treatment of normal hearts. Also, no qualitative changes were observed among the 13C spectra of the normal and postischemic hearts. In particular, the resonance signal from Ala C-3 (17 ppm) remained consistent over the course of both normal perfusion and reperfusion (see Fig 4⇓), demonstrating the utility of this resonance signal as an internal reference of the intracellular availability of 13C-labeled pyruvate in the intact heart.3 4 The constancy of the 13C NMR signal of alanine also served to establish that the tissue content of 13C-enriched alanine did not fluctuate from the onset of reperfusion to the end of the protocol. By use of this invariant alanine signal as an internal reference, the extent of 13C incorporation into the glutamate pool was examined by means of the ratio of the signal intensity from the initial site of glutamate labeling, the 4-carbon position, to that of the unchanging 3-carbon resonance signal of 13C-enriched alanine (Glu C-4 to Ala C-3). As described in a previous publication, the examination of this Glu C-4 to Ala C-3 value from 13C spectra of intact hearts demonstrates the enhanced entry of 13C label into oxidative metabolism by the DCA-induced activation of PDH versus nonoxidative metabolism of pyruvate.4
Fig 5⇓ graphically illustrates the 13C-enrichment of the glutamate pool over the course of both normal perfusion (Fig 5A⇓) and reperfusion (Fig 5B⇓), as indicated by the Glu C-4 to Ala C-3 ratios. As found in a previous study,3 the 13C signal emanating from labeled glutamate was significantly reduced during early reperfusion as a result of the transient reduction in PDH activity. As shown in Fig 5⇓, the parameter of Glu C-4 to Ala C-3 at 3 minutes of reperfusion is significantly reduced to 58% of the control value (P<.05). However, the initial reduction in [3-13C]pyruvate conversion into the glutamate pool by oxidative metabolism that occurs during reperfusion can be countered by stimulation of PDH. DCA treatment during reperfusion resulted in a elevation of Glu C-4 to Ala C-3 ratios to similar to those of DCA-treated hearts under normal perfusion conditions. Fig 6⇓ clearly illustrates this difference in the content of 13C-enriched Glu C-4 which is evident in the 13C spectra acquired during the initial period of reperfusion with DCA (Fig 6B⇓) and without DCA (Fig 6A⇓). Steady state spectra from the same hearts are shown in Fig 7⇓ and demonstrate that the enhanced signal from the Glu C-4 persisted throughout reperfusion with DCA.
Further analysis of the metabolite enrichment levels by in vitro methods afforded additional insight into the effectiveness of PDH activation, in the absence of increased glycolytic contributions, to both induce and support improvements in cardiac performance upon reperfusion. Table 2⇓ lists the total tissue metabolite pools for glutamate and alanine within each group of hearts along with corresponding percentages of 13C enrichment. No significant differences in total metabolite pools are evident in comparison of normal and reperfused hearts with or without DCA treatment. At steady state, the isotopic enrichment of glutamate was consistently higher with DCA treatment among normal and postischemic hearts. However, alanine content was not significantly different between any of the experimental heart groupings.
Neither the contributions of glycolysis to alanine production nor glycogen breakdown were different between the untreated and DCA-treated reperfusion groups, despite the observed differences in postischemic function. 13C enrichment of Ala C-3 was particularly unaffected by the different perfusion protocols (Table 2⇑), demonstrating conclusively that the glycolytic contribution from endogenous carbohydrate sources (unlabeled portion) was essentially the same in all groups. Furthermore, tissue glycogen levels were found to be similar in both reperfusion groups: untreated reperfusion, 319±58 μmol glucose equivalents per gram dry wt; DCA-treated reperfusion, 324±11 μmol glucose equivalents per gram dry wt. Although glycogen values can vary significantly according to the fed or fasted state in rabbits,23 these values are in the range of glycogen levels detected in rat hearts perfused with pyruvate, as reported by Bricknell and Opie.24 These glycogen levels were reduced in comparison to our published control values, as expected, and were consistent with previously reported data of glycogen content in the rabbit heart after 10 minutes of global ischemia.3 The similarity in glycogen content between the untreated and DCA-treated hearts supports our isotope enrichment data, which indicate no additional component of glycolysis from glycogen entering the alanine pool during DCA treatment. These results all indicate that enhanced pyruvate oxidation in the presence of DCA did not affect the availability of the continually infused [3-13C]pyruvate, as shown by alanine enrichment, but that augmented oxidation of pyruvate was accommodated by increased incorporation of 13C label into the glutamate pool.
Total glutamate content of the myocardium was not affected by the short duration of ischemia, a finding consistent with the previously published results of Peuhkurinen et al.25 Glutamate content immediately after 10 minutes of global ischemia was 12.9±0.9 μmol/g dry wt (n=4). After 3 minutes of reperfusion, glutamate content remained essentially unchanged at 10.4±0.8 μmol/g dry wt in untreated hearts (n=5) and 11.0±0.9 μmol/g dry wt in hearts treated with DCA (n=5), to indicate that the observed reduction in Glu C-4 to Ala C-3 ratios during early reperfusion were not due to reductions in tissue glutamate content but resulted from a reduction or delay in the contribution of label from pyruvate oxidation. Furthermore, the early increase in 13C enrichment of glutamate during reperfusion in the presence of DCA was not a result of elevated tissue glutamate content.
Also shown in Table 2⇑ are the values of 13C enrichment of the acetyl groups entering the TCA cycle at the citrate synthase reaction (Fc) at the end of the perfusion period for each protocol. Activation of PDH activity is in evidence from a small but statistically significant elevation of Fc values in both groups of DCA-treated hearts, normal and postischemic, over that in the respective untreated groups. The increased Fc values observed in hearts treated with DCA indicate that the contribution of unlabeled glucose (or glycogen) and endogenous lipid to oxidative metabolism was effectively reduced. Therefore, no increase in glucose oxidation or augmented glycolytic flux occurred to provide additional energetic support during the recovery of contractile function in hearts reperfused with DCA.
Functional and Metabolic Considerations
The experiments performed in this study, in which pyruvate was supplied as the sole carbon-based substrate during reperfusion, demonstrate that the improvements in postischemic cardiac performance in response to the activation of PDH occurred in the absence of increased glycolysis. The lack of glycolytic contributions to improved energetic and respiratory efficiency in postischemic hearts treated with DCA (Table 1⇑) is evident from the carbon isotope data (Table 2⇑). As evidence, these data indicate that the appearance of unlabeled 12C in the glycolytic production of alanine4 26 remained unchanged in both normal and postischemic hearts after treatment with DCA. The similar glycogen content in both the DCA-treated and untreated hearts also confirms that no differences in glycolysis between the groups was present. Further evidence that glycolytic activity did not contribute to augmented oxidation of endogenous, unlabeled carbohydrate is provided by an actual increase in 13C label from pyruvate contributing to the formation of acetyl groups that then enter the TCA cycle (Table 2⇑). In fact, 13C NMR observations of intact hearts (Fig 5⇑) and the enrichment data from in vitro NMR (Table 2⇑) both serve to confirm enhanced carbon flux from pyruvate into the resulting metabolites of pyruvate oxidation over that of the unchanged availability of pyruvate in the myocyte.4 Therefore, this study is the first demonstration of a direct link between counteracting the reduced PDH enzyme complex activity in reperfused myocardium with sustained improvements in contractile performance. In this sense, activation of PDH was observed to achieve a metabolic reversal of the now well-observed phenomenon of myocardial stunning.27
These results are generally consistent with the findings of McVeigh and Lopaschuk7 on hearts metabolizing glucose but are in contrast to a single previous study of pyruvate, supplied at much higher concentrations and in various combination with glucose during reperfusion with DCA.10 Surprisingly, this earlier, contrasting report also shows a complete lack of contractile recovery during reperfusion with pyruvate alone after only 10 minutes of ischemia. This poor functional recovery in the presence of pyruvate alone is contrary to both our experience3 5 and that of other laboratories28 and is possibly artifact from the decision to administer insulin during reperfusion with pyruvate alone and no exogenous glucose.10
In this study, activation of PDH enzyme successfully improved the metabolic efficiency of cardiac performance while also reversing the trend for reduced 13C incorporation into the myocardial glutamate pool as a consequence of reduced pyruvate oxidation in the postischemic heart.3 High-energy phosphate levels in the postischemic hearts belied the differences observed in contractile recovery between the hearts treated with DCA and those that were untreated (Table 1⇑). However, only static levels of phosphocreatine and ATP were detected and not the associated turnover rates, which were likely to be different, as suggested by oxygen consumption measurements. Although oxygen consumption was lowered in untreated hearts that showed evidence of postischemic contractile dysfunction, the extent of oxygen used per unit of mechanical work (Table 1⇑) was significantly increased by augmenting PDH activity in the postischemic hearts. Therefore, countering the postischemic inactivation of PDH also served to limit the wastage of oxygen utilization, which has been observed previously in both isolated hearts and whole animals.29 30
However, this improved respiratory efficiency that was afforded by reperfusing postischemic hearts with DCA does not appear related to an increase in energy production by stimulation of glycolytic flux (Table 2⇑), as discussed above, but is the result of increased activity of the PDH enzyme. Removal of accumulated lactate by activation of PDH has been one mechanism suggested for the improved recovery of hearts reperfused with glucose and DCA,6 but we virtually eliminated the potential for retaining the bulk of accumulated lactate by not recirculating the coronary effluent during early reperfusion. In addition, no evidence currently exists to support the notion that the mere presence of lactate is deleterious to otherwise viable, well-perfused myocardium, whether it be in the postischemic state or not.31 The only potential (and theoretical) deleterious effects of lactate in the well-perfused myocardium would be inhibition of glycolysis, which again was clearly not in evidence from the striking similarity in alanine-enrichment data for each of the experimental groups represented in Table 2⇑. We therefore consider the potential for any additional inhibition of glycolysis by lactate in the presence of exogenous pyruvate to have been insignificant.
13C NMR Spectroscopy
As demonstrated by the current study, the continued development of 13C NMR spectroscopy enables investigation of intracellular metabolic events that occur in response to altered pathophysiological state of the intact, functioning myocardium. In this work, a combination of 13C NMR spectroscopy of intact hearts and in vitro samples has provided insight into the shifts in metabolic processes that have been induced by pharmacological intervention and the ischemia/reperfusion protocol. The influence of DCA activation of PDH on the entry of 13C into glutamate from [1-13C]glucose has been previously demonstrated in isolated rat hearts.32 By targeting the activity of PDH with the direct substrate for that enzyme complex, 13C-enriched pyruvate, we have been able to measure quantitative changes in the incorporation of carbon isotope label into the glutamate pool through the oxidation of pyruvate.4 We were then able to reference the progressive incorporation of 13C label into oxidative metabolism to the extent of pyruvate availability within the cell by taking advantage of the rapid isotope equilibrium that exists between the pyruvate and alanine pools, as opposed to the isotopic enrichment of lactate, which does not indicate a rapid isotope equilibrium with pyruvate.4 33 34 35 These studies, and our most recent finding that glutamate levels did not change appreciably over the course of this ischemia and reperfusion protocol (see “Results”), served to confirm the validity of our finding that reduced isotope enrichment of glutamate during early reperfusion of postischemic myocardium occurred as a consequence of PDH inactivation.
As the conditions for resolving multiplet patterns within individual 13C resonance peaks in a spectrum are somewhat limited by the experimental conditions for in vivo NMR observations,36 the use of the Glu C-4 to Ala C-3 ratio as a relative index of 13C entry into oxidative metabolism versus nonoxidative metabolism of pyruvate in this study (Fig 5⇑) allowed us the means to monitor changes in a single enzyme system within intact, functioning hearts. This ratio is not to be confused with the turnover of label within the glutamate pool, which leads to multiple labeling of the glutamate molecule at the 2- and 3-carbon positions as well. Use of the Glu C-4 to Ala C-3 ratio in previous studies allowed us to detect the inactivation of PDH in reperfused hearts.3 5 These earlier findings lead to the notion of reversing the trend of reduced conversion of [3-13C]pyruvate into the glutamate pool by pharmacological activation of pyruvate oxidation with DCA. As explored in this study, activation of PDH not only reversed the reduced incorporation of 13C into oxidative metabolism but also served to counter the commonly observed phenomenon of postischemic contractile dysfunction. Therefore, 13C NMR studies of intact myocardium can contribute to both the detection of altered metabolic events associated with pathophysiological processes and to then document mechanisms of therapeutic intervention.
Limitations of the Study
The experiments in this study were designed to focus on the influence of a single enzyme, which can be monitored by 13C NMR spectroscopy, on metabolic and functional paradigms of recovery in postischemic reperfused myocardium. In doing so, we have targeted the NMR observations to the activity of PDH by supplying 13C-enriched pyruvate to an isolated perfused heart model to eliminate the confounding variables associated with activation of PDH in the presence of active glycolytic metabolism. The actual conditions of in vivo ischemia and reperfusion are not likely to offer the opportunity for such precise manipulation of substrate supply. Additionally, the purpose of this study was to examine the link between cardiac performance and the activity of PDH in the otherwise stunned myocardium and not to test the therapeutic value of this pharmacological intervention. The pharmacological effects of DCA have been examined in detail by others,22 although little is known of the potential clinical benefits of DCA treatment for the heart. Therefore, the use of both exogenous pyruvate and a pharmacological agent was incidental to the aims of this current study.
Significance of This Work
The significance of the current findings stems from the metabolic support of contractile function in the postischemic heart and the notion that carbon flux through specific enzyme complexes at the level of enzyme activation or even perhaps expression plays an active role in homeostasis of the myocyte and overall cardiac performance. This study demonstrates that the activity of the PDH enzyme complex does indeed play a significant role in the recovery of the postischemic heart. From the data presented above, activation of PDH was effective not by merely facilitating additional metabolic support by the recruitment of glycolysis but rather through a yet undetermined mechanism that counters postischemic contractile dysfunction without affecting performance of normal myocardium. Obviously, additional work will be needed to further characterize the relations between the activation state of PDH and the contractile apparatus that are implied from these data. The opportunity to observe such mechanisms of enzyme function in intact functioning myocardium through the application of 13C NMR spectroscopy methods suggests future opportunities for in vivo enzymology in the evaluation of pathophysiology and contractile performance of myocardium at a molecular level.
This work was aided in part by NHLBI grant RO1-HL-49244 (E.D.L.) and Grant-in-Aid 13-522-912 from the American Heart Association, Massachusetts Affiliate, Inc (E.D.L.) and was done during the tenure of an Established Investigator Award from the American Heart Association to E.D.L. The authors are grateful to Dr Thomas J. Brady for his continued support of this work.
Presented in part as an abstract at the 64th Scientific Sessions of the American Heart Association, Anaheim, Calif, November 11 to 14, 1991.
- Received June 16, 1994.
- Revision received September 19, 1994.
- Accepted November 13, 1994.
- Copyright © 1995 by American Heart Association
Patel TB, Olson MS. Regulation of pyruvate dehydrogenase complex in ischemic rat heart. Am J Physiol. 1984;246:H858-H864.
Lewandowski ED, Johnston DL. Reduced substrate oxidation in post-ischemic myocardium: 13C and 31P NMR analyses. Am J Physiol. 1990;58:H1357-H1365.
Johnston DL, Lewandowski ED. Fatty acid metabolism and contractile function in the reperfused myocardium: multinuclear NMR studies of isolated rabbit hearts. Circ Res. 1991;68:714-725.
McVeigh JJ, Lopaschuk GD. Dichloroacetate stimulation of glucose oxidation improves recovery of ischemic rat hearts. Am J Physiol. 1990;259:H1079-H1085.
Lopaschuk GD, Spafford MA, Davies NJ, Wall SR. Glucose and palmitate oxidation isolated working rat hearts reperfused after a period of transient global ischemia. Circ Res. 1990;66:546-553.
Latipaa PM, Hiltunen JK, Peuhkurinen KJ, Hassinen IE. Regulation of fatty acid oxidation in heart muscle. Effects of pyruvate and dichloroacetate. Biochem Biophys Acta. 1983;752(part 1):162-171.
Lewandowski ED. Nuclear magnetic resonance evaluation of metabolic and respiratory support of work load in intact rabbit hearts. Circ Res. 1992;70:576-582.
Lewandowski ED, Chari MV, Roberts R, Johnston DL. NMR studies of β-oxidation and short chain fatty acid metabolism during recovery of reperfused hearts. Am J Physiol. 1991;261:H354-H363.
Williamson JR, Corkey BE. Assays of intermediates of the citric acid cycle and related compounds by flourometric enzyme analysis. In: Lowenstein JM, ed. Methods of Enzymology. New York, NY: Academic Press, Inc; 1967:434-512.
Malloy CR, Sherry AD, Jeffrey FMH. Evaluation of carbon flux and substrate selection through alternate pathways involving the citric acid cycle of the heart by 13C NMR spectroscopy. J Biol Chem. 1988;265:6964-6971.
Lewandowski ED, Johnston DL, Roberts R. Effects of inosine on glycolysis and contracture during myocardial ischemia. Circ Res. 1991;68:578-587.
Schneider CA, Nguyen VT, Taegtmeyer H. Feeding and fasting determine postischemic glucose utilization in isolated working rat hearts. Am J Physiol. 1991;260:H542-H548.
Bricknell OL, Opie LH. Effects of substrates on tissue metabolic changes in the isolated rat heart during underperfusion and on release of lactate dehydrogenase and arrhythmyia during reperfusion. Circ Res. 1978;43:102-115.
Peuhkurinen KJ, Takala TES, Nuutinen EM, Hassinen IE. Tricarboxylic acid cycle metabolites during ischemia in isolated perfused rat heart. Am J Physiol. 1983;233:H281-H288.
Taegtmeyer H, Peterson MB, Ragavan VV, Ferguson AG, Lesch M. De novo alanine synthesis in isolated oxygen-deprived rabbit myocardium. J Biol Chem. 1977;252:5010-5018.
Braunwald E, Kloner RA. The stunned myocardium: prolonged, post-ischemic ventricular dysfunction. Circulation. 1982;66:1146-1149.
Cavallini L, Valente M, Rigobello MP. The protective action of pyruvate on recovery of ischemic rat heart: comparison with other oxidizable substrates. J Mol Cell Cardiol. 1990;22:143-154.
Russell RR, Nguyen VT, Mrus JM, Taegtmeyer H. Fasting and lactate unmask insulin responsiveness in the isolated working rat heart. Am J Physiol. 1992;263:E556-E561.
Peuhkurin KJ, Hassinen. Pyruvate carboxylation as an anaplerotic mechanism in the isolated perfused rat heart. Biochem J. 1982;202:67-76.
Peuhkurinen KJ, Hiltunen JK, Hassinen IE. Metabolic compartmentation of pyruvate in the isolated perfused rat heart. Biochem J. 1983;210(part 1):193-198.
Bunger R. Compartmented pyruvate in perfused working heart. Am J Physiol. 1985;249(part 2):H439-H449.