INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

IN THIS STUDY, we examined the potential for metabolic coordination between the mitochondrial dehydrogenase reactions and metabolite influx and efflux across the mitochondrial membrane in intact functioning hearts. In particular, this work focused on the coordination between the rate-limiting dehydrogenase of the first span of the tricarboxylic acid (TCA) cycle, -ketoglutarate dehydrogenase (-KGDH), and metabolite transport via the -ketoglutarate-malate exchanger on the mitochondrial membrane. Although already established in the isolated mitochondrial preparation (12, 31), the role of the mitochondrial dehydrogenases in coordinating TCA cycle flux (V_TCA) and the metabolite exchange rates between the cytosol and mitochondria has yet to be demonstrated in the intact beating heart. Thus a combination of flux measurements from the dynamic mode ¹³C nuclear magnetic spectroscopy (NMR) of the intact heart and radiotracer analysis of isolated heart mitochondria provides new insight into coordinating intermediary metabolism between the cytosol and mitochondria to support cardiac function.

Two exchange proteins of the mitochondrial membrane provide one mechanism by which cytosolic metabolites exchange with mitochondrial metabolites: the reversible -ketoglutarate-malate exchanger and the unidirectional glutamate-aspartate exchanger. These two transporters may function separately, but when operating in tandem, they form the malate-aspartate shuttle (10). The exchange of metabolites between mitochondria and cytosol may then be coordinated with mitochondrial TCA cycle activity at the level of the enzyme -ketoglutarate dehydrogenase (-KGDH). -KGDH oxidizes -ketoglutarate to form succinyl CoA within the TCA cycle. In this manner, -KGDH regulates V_TCA through the second span of the cycle (23) and competes with the mitochondrial membrane transporter for -ketoglutarate (11). The subsequent efflux and transamination of -ketoglutarate, via the glutamate-oxaloacetate transaminase (GOT), produces cytosolic glutamate. This balance between V_TCA and metabolite exchange rates can be evaluated by analysis of dynamic ¹³C NMR spectra of carbon isotope enrichment of glutamate in the intact heart (33-35). In the isolated mitochondria, the unidirectional influx or efflux of -ketoglutarate and malate across the mitochondrial membrane can also be measured with ambiguity for comparison to the isotope kinetics in the intact heart. Such a combined information can be interpreted to define changes in the activity of the enzymes coordinating these processes.

We have previously shown that the rate of the transaminase GOT is much too fast to be a rate-determining component of the observed interconversion rate between -ketoglutarate and glutamate in intact hearts by directly measuring the activities of the mitochondrial and cytosolic isozymes of GOT in rabbit myocardium (34). Instead, the much slower rate of interconversion can be attributed not to the transaminase but rather to the transport processes that bring -ketoglutarate out of the mitochondria for transamination with the large cytosolic glutamate pool.

Subsequent work from our laboratory has determined that ¹³C NMR methods are sensitive to both V_TCA and rates of metabolite transport between the mitochondria and cytosol in intact functioning hearts (33-35). Unlike more traditional assay and isolation techniques, ¹³C NMR spectroscopy provides an approach to characterize the activity of the TCA cycle, exchange rates, and potentially the kinetics of specific enzymes in the intact tissue under normal, diseased, or altered function.

In this study, alteration of the kinetics (K_m) of the key rate-limiting enzyme of the TCA cycle, -KGDH, was studied in both intact hearts and isolated mitochondria. As the substrate concentration at which the reaction for the given enzyme reaches half-maximal velocity, K_m served as a kinetic index of the substrate-enzyme affinity. The affinity of -KGDH for the substrate -ketoglutarate was increased (i.e., a reduction in K_m) by reducing intramitochondrial pH and increasing calcium content. Combined studies of -ketoglutarate oxidation and efflux from isolated rabbit heart mitochondria and NMR assessment of stable isotope enrichment of glutamate provided a unique opportunity to study the balance between -ketoglutarate transport and oxidation as a function of the affinity of the -KGDH for its substrate. The analysis presented here indicates that changes in the kinetic parameters of -ketoglutarate oxidation rates mediate changes in -ketoglutarate efflux from mitochondria at constant V_TCA. For the first time, we establish the influence of pH and Ca²⁺ on the cardiac -KGDH in the intact heart, while demonstrating the role this dehydrogenase plays in coordinating cytosolic metabolism with that oxidative energy production pathways.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Isolated Heart Model

Perfusate pH was reduced from 7.2 to 6.6 by decreasing the buffer content of bicarbonate to 6 mM. As previously reported, there is a simultaneous reduction in LVDP and HR at low pH (32). To maintain normal V_TCA rates at this low pH, supplemental CaCl₂ was added to the perfusate to normalize the rate-pressure product (RPP) and match control levels of MO₂ (32). With the metabolic demands of the heart normalized, we then utilized ¹³C NMR spectroscopy to monitor V_TCA and the interconversion between -ketoglutarate and glutamate in response to reduced K_m of -KGDH for -ketoglutarate.

Experimental Protocol

NMR Measurements

Carbon spectra were acquired at 100 MHz with a bilevel broad-band decoupling scheme (14). Carbon spins were nutated with a 45° excitation pulse, 64-scans per 2.5-min spectrum, 10,000-Hz sweep width, and 8 K data set. Proton excitation at 400 MHz, 0.5 W during the interpulse delay, and 7.0 W for 17 µs was applied to irradiate carbon-proton coupling and produce nuclear Overhauser enhancement of carbon signal. Postprocessing of the summed FID values NMR data included 20-Hz line broadening, Fourier transformation, and phase correction. Natural abundant ¹³C NMR signal was subtracted from subsequent ¹³C-enriched spectra. Resonance signals were identified by chemical shift values (parts/min) referenced to dioxane at 67.4 parts/min. The signal intensity of each resonance was determined by curve fitting the peak to a Lorentzian curve and integrating the area of the fit (NMR1 software, Tripos, St. Louis, MO).

Tissue Chemistry

The pH response of GOT activity from the cytosol of rabbit hearts was determined using methods to measure maximal reaction velocity of the isolated isoenzymes as previously described (15, 34). Separated cytosolic fractions of homogenized rabbit myocardium were tested for GOT activity, and the K_m for each substrate was determined under two different pH conditions, 7.2 and 6.8.

High-Resolution ¹³C NMR

Kinetic Model and Analysis

Hearts are provided [2-¹³C]acetate, which contributes to the formation of acetyl CoA enriched at the C2 position for entry into the TCA cycle. As shown in Fig. 1, this places the label at C4 of -ketoglutarate. -Ketoglutarate can be transaminated to form glutamate where the label is incorporated into the C4. The bulk of the ¹³C NMR-observed glutamate pool is cytosolic (9), and the rate of the exchange has been shown to be largely determined by the physical transport of -ketoglutarate across the mitochondrial membrane (34). -Ketoglutarate could also be oxidized by the dehydrogenase. Oxidation of -ketoglutarate shifts the label to C2 or C3 of the symmetric succinate molecule. As the carbon label is recycled within the pathway and reenters the first span of the cycle, the labeled carbon will label either C2 or C3 of -ketoglutarate and glutamate.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Diagram of labeling scheme within tricarboxylic acid (TCA) cycle and related metabolites in isolated heart as observed by dynamic 13C nuclear magnetic resonance (NMR) spectroscopy. [2-13C]acetate enters TCA cycle via [2-13C]acetyl CoA. * Initial sites of 13C labeling at C4 positions of citrate, -ketoglutarate, and glutamate.  Recycling of 13C within TCA cycle results in appearance of label at secondary C2 and C3 positions with equal probability.

Kinetic equations describing the presteady-state labeling of glutamate and other key metabolic pools were derived by the principle of mass conservation. This kinetic model is composed of nine differential equations defining the flux through each compartment and includes considerations of anaplerotic flux (33, 34). Metabolic pool sizes were also required as input parameters. The concentrations of glutamate, aspartate, citrate, and -ketoglutarate were determined by enzymatic assays. The concentration of metabolite pools known to be too small to influence the flux measurements, such as malate and oxaloacetate, was taken from the literature values obtained from hearts perfused under similar substrate conditions: 0.60 mM malate and 0.04 mM oxaloacetate (23, 28, 33). V_TCA and the F₁ between cytosolic glutamate and mitochondrial -ketoglutarate were determined by nonlinear least-square fitting of the model to ¹³C NMR data of glutamate C2 and C4 enrichment.

Isolated Mitochondria Experiments

Subsequent experiments with the mitochondria were carried out in incubation media consisting of 120 mM KCl, 30 mM 3-(N-morpholino)propanesulfonic acid, 20 mM KH₂PO₄, 5 mM NaCl, 2 mM MgCl₂, 10 U/ml hexokinase, and 20 mM glucose. Phosphate concentration in the media was high because it was consumed at a rate of 2.5 µmol · ml¹ · min¹ over 4 min. However, because the dicarboxylate carrier for the exchange of malate and inorganic phosphate is virtually absent in heart mitochondria, the phosphate concentration of the buffer had no effect on the activity of the -ketoglutarate-malate exchanger (25). The pH was adjusted to pH 7.2 (medium A) or pH 6.8 (medium B), and experiments were performed at 37°C.

The relative affinity of -KGDH for the substrate -ketoglutarate was determined by following the rate of -ketoglutarate oxidation. To measure O₂ consumption, 0.5 mg mitochondrial protein was added to a stirred, temperature-equilibrated, closed glass chamber fitted with a Clark electrode. Decreases of O₂ in the chamber were measured polarographically. A double-reciprocal plot of oxygen consumption versus initial -ketoglutarate concentration was prepared. Oxygen utilization was measured at six concentrations of -ketoglutarate ranging from 0.02 to 1.0 mM, and the apparent K_m was determined from the -ketoglutarate concentration that produced one-half the maximum oxygen consumption.

Other experiments were designed to measure individual reactions involved in the metabolism of -ketoglutarate. In these experiments mitochondria were incubated under one of the following four conditions for media at pH 7.2 or pH 6.8: condition IA, 2.5 mM [2,3-³H]glutamate, 2.5 mM [2-¹⁴C]pyruvate, 0.2 mM -ketoglutarate, and 5 mM malate (pH 7.2); condition IB, same as condition IA but adjusted to pH 6.8; condition IIA, 2.5 mM [2,3-³H]glutamate, 2.5 mM pyruvate, 0.2 -[U-¹⁴C]ketoglutarate, and 5 mM malate (pH 7.2); or condition IIB, same as condition IIA but adjusted to pH 6.8. This allowed us to track the fate of -ketoglutarate generated by isocitrate dehydrogenase (from [¹⁴C]pyruvate) or by aspartate aminotransferase (from [2,3-³H]glutamate) or by transport across the mitochondrial membrane (from medium -[U-¹⁴C]ketoglutarate under similar incubation conditions with differing pH). Products of the ³H- and ¹⁴C-labeled substrates were separated by ion exchange chromatography using Dowex 1 chloride columns (17) and quantified by double-label counting in a liquid scintillation counter. The design, rationale, and procedures for calculating fluxes from these labeled substrates have been described previously in a study of effects of Ca²⁺ and H⁺ on -ketoglutarate metabolism in liver and kidney mitochondria (26).

Statistical Analysis

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Studies in Isolated Mitochondria

pH effects on -ketoglutarate metabolism and efflux.

Subsequent experiments with isolated mitochondria were designed to find out whether the increase in the affinity of -KGDH for its substrate, -ketoglutarate, significantly changes the ratio between the amount of -ketoglutarate that proceeds around the citric acid cycle compared with the amount that effluxes from the mitochondria. To establish the pH-dependent reduction in the K_m of -KGDH for heart mitochondria and to measure the influence of altered -ketoglutarate oxidation on -ketoglutarate efflux rates, we incubated isolated rabbit heart mitochondria in the standard incubation media plus 2.5 mM [2-¹⁴C]pyruvate, 0.2 mM -ketoglutarate, 2.5 mM [2,3-³H]glutamate, 5 mM malate, 1 mM ADP, 20 mM glucose, and 10 U of hexokinase. The hexokinase trap kept ADP high and respiration maximal. High malate concentration trapped ¹⁴C cycling around the citric acid cycle and ensured that no ¹⁴CO₂ was generated from pyruvate. From data shown in Table 1, the amount of [¹⁴C]glutamate formed from [¹⁴C]pyruvate in 4 min demonstrates that at both normal and low pH, a factor of seven to nine times more carbon from pyruvate leaves the mitochondria as -[¹⁴C]ketoglutarate than leaves as [¹⁴C]glutamate.

View this table: [in this window] [in a new window]   Table 1.   Formation of [14C]glutamate vs. formation of -[14C]ketoglutarate from [2-14C]pyruvate in isolated mitochondria

From the condensation reaction catalyzed by citrate synthase, the ¹⁴C from pyruvate labels citrate, isocitrate, and then -ketoglutarate. At this point, the radiolabel can enter one of three pathways, which were quantified by chromatographic separation of the ¹⁴C-labeled metabolites. One pathway involves transport of the -ketoglutarate out of the mitochondria through the -ketoglutarate-malate exchange transporter. Without any added malate in the media to enable this transport, no -ketoglutarate will accumulate in the suspension. A second pathway for the -[¹⁴C]ketoglutarate is oxidation to form succinate and malate. A third pathway is the transamination inside the mitochondria via the mitochondrial GOT enzyme to form [¹⁴C]glutamate, a small portion of which can leave the mitochondria via the slow glutamate-hydroxyl carrier. In detecting these labeled metabolites in the mitochondrial suspension, it is the metabolites that are in the media and not inside the mitochondria that are measured, because the intramitochondrial volume in the suspension is only 0.1% of the total volume. As shown in Table 1, the amount of [¹⁴C]glutamate formed from [¹⁴C]pyruvate in the mitochondria and released into the media was very small relative to the amount of -[¹⁴C]ketoglutarate that accumulated in the suspension media. Thus any glutamate formed in the mitochondria is only a small fraction of that which would be detected in the cytosol of the intact hearts as opposed to the larger pool of glutamate produced in the cytosol from -ketoglutarate.

The products of -KGDH were also monitored in similar experiments, and the results are shown in Table 2. As described above in the MATERIALS AND METHODS, the incubation conditions differed with respect to pH (either 6.8 or 7.2) and with respect to whether pyruvate or -ketoglutarate was labeled with ¹⁴C. All three substrates (pyruvate, glutamate, and -ketoglutarate) contribute to the flux through -KGDH, and flux from each was determined by measuring the amount of ¹⁴C and ³H in the -KGDH products succinate and malate. These constitute the major end products, since fumarate and succinyl CoA could not be detected in these experiments and since malate blocked further cycling. Also shown in Table 2 is the rate of efflux of -ketoglutarate generated by each substrate compared with its rate of oxidation in the citric acid cycle by -KGDH. The data indicate that lowering the K_m of -KGDH increases the percentage of -ketoglutarate that is oxidized without first entering the external medium, where in intact cells it would be immediately converted to glutamate. Generation of matrix -ketoglutarate from glutamate was lower than generation of -ketoglutarate from pyruvate or from transport of medium -ketoglutarate into the matrix. When total flux from all substrates through -KGDH is compared with total -ketoglutarate efflux, the decrease in pH from 7.2 to 6.8 changes the oxidation-to-efflux ratio from 1.02 to 1.92. 

View this table: [in this window] [in a new window]   Table 2.   Mitochondrial -[14C]KG oxidation and efflux

Isolated Heart Studies

1

1

1

1

1

1

Flux measurements from intact hearts. Representative proton-decoupled ¹³C spectra (time-enrichment curves) of an isolated heart perfused with buffer pH adjusted to 6.6 and increased CaCl₂ content and a heart perfused under control conditions are shown in Fig. 2, A and B, respectively. Control spectra are similar to data previously presented (13, 14, 35). The signal intensities arising from the glutamate C2 and C4 resonances were curve fit to a Lorentzian line shape and plotted as a function of time. Figure 3 shows this time course of glutamate enrichment. These curves were fit to a single exponential, and time constants were evaluated for the rise of each curve. The time constants for the glutamate C4 and C2 of the control group were 9.4 and 16.5 min, respectively. For hearts perfused with buffer pH 6.6 and increased CaCl₂ content, glutamate C4 and C2 enrichment time constants were 9.4 and 11.8 min, respectively. These time constants are not a direct index of V_TCA, because labeling of the glutamate is dependent on both the cycle flux and the exchange of label between -ketoglutarate and glutamate pools (34). It should be noted, however, that the time constant for the C4 for either group are similar, whereas the C2 time constants reveal a slower enrichment rate in controls.

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Dynamic 13C NMR spectra of isolated rabbit hearts perfused with 2.5 mM [2-13C]acetate for buffer pH 6.6, 3.0 ± 1.0 mM CaCl2 (A), and buffer pH 7.2, 1.5 mM CaCl2 (B). Peak assignments include GLU-C2, second carbon of glutamate; GLU-C4, fourth carbon of glutamate; GLU-C3, third carbon of glutamate; ACE-C2, second carbon of acetate. PPM, Parts/min.

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Time course of glutamate 13C enrichment from both NMR measurements of intact hearts and kinetic analysis. Open squares, glutamate C4 enrichment; filled squares, glutamate C2 enrichment; solid lines, least-squares fitting of the model to enrichment curves. A: buffer pH 7.2, 1.5 mM CaCl2; B: buffer pH 6.6, 3.0 ± 1.0 mM CaCl2.

The results of kinetic analysis of the enrichment of glutamate provided V_TCA and the F₁ between mitochondria -ketoglutarate and cytosolic glutamate pools (33, 34). The experimental perturbations and findings are summarized in Fig. 4. V_TCA was similar in control hearts and in hearts with the reduced K_m of -KGDH: control group = 10.0 ± 0.2 µmol · min¹ · g¹; low K_m group = 10.1 ± 0.4 µmol · min¹ · g¹. However, the F₁ values of interconversion between -ketoglutarate and glutamate, which represent -ketoglutarate efflux, were significantly reduced with the reduction in the K_m of -KGDH. Control values for F₁ were 9.3 ± 0.6 µmol · min¹ · g¹ versus 2.8 ± 0.4 µmol · min¹ · g¹ at the low K_m (P < 0.05). These results indicate that mitochondrial -ketoglutarate exchange with cytosolic glutamate is significantly reduced when the affinity of -KGDH for -ketoglutarate is increased. This result of reduced F₁ is attributed to the K_m change in the dehydrogenases, because GOT is not sufficiently affected by pH changes over physiological range, as discussed below and elsewhere (5), and the transporter is also insensitive to pH (22).

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Schematic summary of experimental findings for intact hearts. Relative affinity, Km, of -ketoglutarate dehydrogenase (-KGDH) for -ketoglutarate (-KG) was significantly reduced (P < 0.05) in isolated mitochondria incubated with media at pH 7.2 (*) compared with mitochondria incubated with media pH 6.8 (). In intact hearts, (perfusion condition: *, buffer pH 7.2, 1.5 mM CaCl2; and , buffer pH 6.6, 3.0 ± 1.0 mM CaCl2), reduction in -KG exchange rate (P < 0.05), along with observed drop in cytosolic glutamate (P < 0.05), indicates that at a fixed TCA cycle rate lowered Km of -KG dehydrogenase for -KG reduced availability of -KG for transporter-dependent efflux to cytosol and conversion to glutamate.

Metabolite content and in vitro measurements. Steady-state metabolite contents are listed in Table 3. Control values are in agreement with those previously reported (15, 34). Citrate, -ketoglutarate, and aspartate levels were not statistically different between experimental groups. However, total glutamate content was 38% reduced in the hearts perfused at low pH and increased CaCl₂. As shown in previous studies (15, 34) and confirmed in Table 1, the cytosolic GOT enzyme is the critical isoform of the transaminase that accounts for the observed exchange of -[¹³C]ketoglutarate with the NMR-detected glutamate pool in the cytosol. Despite published data showing only a very shallow dependence on pH across a wider range from 6.0 to 8.0 (5), in our experiments the kinetic values for the cytosolic GOT in both groups were actually reduced slightly by 24% from a mean of 0.46 ± 0.07 µmol · min¹ · mg total heart protein¹ at pH 7.2 (n = 6) to 0.35 ± 0.08 µmol · min¹ · mg total heart protein¹ at pH 6.8 (n = 6) (P < 0.05). From the double-displacement reaction kinetics, this small but significant drop in the activity of the cytosolic GOT corresponded to calculated flux rates of 105 µmol · min¹ · g dry wt¹ at pH 7.2 and 80 µmol · min¹ · g dry wt¹ at pH 6.8 (15). The slightly reduced GOT flux clearly remained too fast, by 29-fold in comparison with F₁, to account for the 70% reduction of F₁ at low pH. These data are consistent with our previous observations that flux through GOT is much too fast to account for the observed values of interconversion between -ketoglutarate and glutamate (15, 34).

In addition, high-resolution in vitro ¹³C NMR spectra of tissue extracts revealed that substrate utilization under these conditions was not affected by the percentage of acetyl groups entering the TCA cycle at citrate synthase that are enriched with ¹³C. The percentage of total acetyl CoA derived from exogenous ¹³C-enriched acetate was 92% in controls and 90% at low pH. The values were not different between groups, and moreover, this value does not affect the measurements of flux through the TCA cycle (33, 34). The relative contribution of anaplerotic flux into the TCA cycle versus that at citrate synthase was not different in either group (only ranging from 4 to 8%) and did not contribute to the differences observed in the flux measurements from intact hearts.

Comparison of -ketoglutarate oxidation and efflux in the two preparations. Corresponding values for -ketoglutarate oxidation and efflux are catalogued in Table 4. Values for F₁ from intact hearts are shown as measurements of -ketoglutarate efflux rates, because -ketoglutarate influx-efflux across the mitochondrial membrane is known to predominate the rate contributions to F₁ (34). The relative changes in the ratio of -ketoglutarate efflux to oxidation are consistent with a change in the K_m of -KGDH in the intact heart, as was confirmed for the isolated mitochondria.

View this table: [in this window] [in a new window]   Table 4.   Effect of altered Km at fixed tricarboxylic acid cycle flux rates on -KG oxidation vs. efflux in intact hearts

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Early work by Randle et al. (23) suggested that the TCA cycle operated along two spans: acetyl CoA to -ketoglutarate controlled by citrate synthase, and -ketoglutarate to oxaloacetate controlled by -KGDH. This scheme permits a redistribution of carbons between cycle intermediates in the mitochondria during transition states of heart workload. Thus -KGDH balances flux through the two spans of the TCA cycle. In addition, -ketoglutarate is a common intermediate among different pathways within the mitochondria and cytosol. Thus competition for -ketoglutarate between further oxidation in the TCA cycle and efflux to the cytosol is partly regulated by -KGDH. This adjustment of cytosolic and mitochondrial -ketoglutarate levels by -KGDH is an important factor in coordinating flux with the -ketoglutarate-malate transporter.

In the present study, the effects of altering the enzyme-substrate affinity of a rate-limiting dehydrogenase enzyme of the TCA cycle on metabolite exchange across the mitochondrial membrane were studied in both intact hearts and isolated mitochondria. In experiments on intact perfused hearts, metabolic rates were determined by fitting a kinetic model of isotope enrichment to the dynamic ¹³C spectra acquired from hearts. We have already demonstrated in a previous study that the measured activity of the glutamate oxaloacetate transaminase is much too fast to account for the observed rates of interconversion between -[¹³C]ketoglutarate from the mitochondria and glutamate in the cytosol (34), indicating that -ketoglutarate efflux from mitochondria is the rate-determining step in the ¹³C labeling of the large cytosolic glutamate pool. By combining additional ¹³C NMR experiments on intact hearts with radioisotope studies of isolated mitochondria, we were able to examine the balance between rates of -ketoglutarate oxidation and efflux in the current study.

The results from isolated hearts are consistent with those from the isolated mitochondria to indicate that increasing the affinity of -KGDH for the substrate -ketoglutarate (i.e., a decrease in K_m) increased -ketoglutarate oxidation relative to the rate of efflux from the mitochondria for exchange with cytosolic glutamate. With the same workload in both groups of hearts maintained, oxygen consumption and V_TCA were held relatively constant despite altering the enzyme affinity with pH. The influence of shifting the balance between -ketoglutarate oxidation and -ketoglutarate efflux from the mitochondria was evident in the reduced values for the F₁ between -ketoglutarate and glutamate and the observed drop in glutamate. These results demonstrate the link between mitochondrial and cytosolic metabolites at this branch point between oxidation metabolism and metabolic communication between the mitochondria and the cytosol.

Competition for substrate by the -ketoglutarate transporter and -KGDH exists by virtue of their apparent K_m values (11, 26, 27). Earlier work reported that the reversible, nonelectrogenic, pH-insensitive -ketoglutarate-malate transporter of the mitochondrial membrane has an apparent K_m of 1.5 mM for -ketoglutarate on the matrix side of the carrier (3, 22, 24), whereas the K_m of -KGDH for -ketoglutarate was reported as 0.67 mM (11). This makes both oxidation and efflux very sensitive to regulation by the -ketoglutarate concentration in the mitochondrial matrix. The transamination via GOT has been shown by others to be pH insensitive from pH 6 to 8 (5). In this study, we actually did find that the cytosolic GOT isozyme showed a 24% reduction in activity at low pH, but this relatively small drop in activity is not sufficient to produce any measurable effect on the interconversion rate between -ketoglutarate and glutamate (F₁) in the heart. Instead, this drop in cytosolic GOT activity at pH 6.8 would need to be at least 96.5% to affect the observed F₁. In addition, the -ketoglutarate transporter is already known to be pH independent (8, 22). Therefore, the observed changes were due to the effect of pH in reducing the K_m of the -KGDH.

In this study, we report on the effect of pH on the K_m of -KGDH in isolated cardiac mitochondria. These values have been previously reported for liver and kidney mitochondria (16, 26). The pH of the bathing medium was reduced from 7.2 to 6.8, and the rate of -ketoglutarate oxidation was followed as a function of oxygen consumption. Under our conditions, the relative index K_m of -KGDH affinity was 0.23 mM at pH 7.2, and at pH 6.8, K_m reduced to 0.10 mM. In addition, the K_m of -KGDH for the substrate, -ketoglutarate, has been shown to be largely influenced by changes in intramitochondrial Ca²⁺ (6, 19, 29). Wan et al. (29) increased matrix free Ca²⁺ from 0 to 0.64 mM and reported the apparent K_m for -ketoglutarate decreased from 2.5 to 0.6 mM. Thus increasing either H⁺ or Ca²⁺ content increased the activity of -KGDH, as reported by others (26, 29). In the present study, ¹³C-NMR spectroscopy provided a noninvasive means of studying such metabolic regulation in the intact beating heart.

The analogous experiment to alter the K_m of -KGDH for -ketoglutarate was performed in the intact functioning heart. Energy demand and oxygen consumption were held constant by careful regulation of buffer pH and CaCl₂ content. Decreasing buffer pH resulted in an initial 30 ± 10% reduction in heart RPP. This is consistent with results previously reported for the isolated perfused heart (32). The reduction may be due to the influence of H⁺ on contractile processes that occur after the interaction between Ca²⁺ and troponin (4). At the low buffer pH 6.6, ³¹P NMR spectra revealed an intracellular pH of 6.89. Increasing CaCl₂ content of the buffer returned the RPP, energy demand, and MO₂ to normalized values. The increase in intracellular Ca²⁺ and H⁺ has been shown to increase both intramitochondrial Ca²⁺ and H⁺ content (7, 29). In the isolated, perfused rabbit heart experiment, the affinity of -KGDH for -KG was altered by changing intracellular pH and Ca²⁺ content without changing net flux through the enzyme reaction.

Kinetic analysis of dynamic ¹³C NMR data provided V_TCA and F₁ between -ketoglutarate and glutamate. Figure 4 shows the results with reference to the TCA cycle and transport pathways first illustrated in Fig. 1. Several important points can be inferred from these data. First, V_TCA was driven by the energetic demands of the matched workloads in both groups of hearts (control group, RPP = 17,000 ± 4,100 mmHg · beats · min¹; reduced K_m group, RPP = 18,200 ± 5,600 mmHg · beats · min¹). Therefore, the perturbations of pH and Ca²⁺ were observed at fixed TCA cycle rates (control group, V_TCA = 10.0 ± 0.2 µmol · min¹ · g¹; reduced K_m group, V_TCA = 10.1 ± 0.4 µmol · min¹ · g¹). Second, the reduction in the interconversion rate between -ketoglutarate and glutamate (control group, F₁ = 9.3 ± 0.6 µmol · min¹ · g¹; reduced K_m group, F₁ = 2.8 ± 0.4 µmol · min¹ · g¹), along with the observed drop in cytosolic glutamate (control group, [Glu] = 29.1 ± 3.0 mmol/g; reduced K_m group, [Glu] = 17.7 ± 3.0 mmol/g), suggests that the dehydrogenase was able to out compete the membrane transporter for -ketoglutarate. This would be possible if the affinity of the dehydrogenase for -ketoglutarate had increased (i.e., a decrease in K_m) with the drop in pH and increase in Ca²⁺. At a constant V_TCA, this probably slowed the rate of -ketoglutarate efflux for interconversion with glutamate in the cytosol. The observed reduction in the total glutamate pool in response to lowered pH is consistent with this hypothesis.

The result of the relative increase in the ratio -ketoglutarate oxidation to efflux from the mitochondria followed by conversion to cytosolic glutamate is that labeled -ketoglutarate is recycled to succinate at a rate faster than that at which it is converted to glutamate. This change accelerates the mixing of the label from the C4 to the C2 position within the TCA cycle relative to the exchange of -ketoglutarate with the glutamate pool. This condition, along with the reduction in glutamate pool size, caused the time difference between labeling at C4 and C2 of glutamate to become reduced without changes in net V_TCA. This perturbation resulted in little if any difference in the rate of appearance of ¹³C at the C4 position of glutamate (time constant = 9.4 min), whereas labeling at the C2 position occurred at an increased rate that was associated with a reduction in the time constant from 16.5 to 11.8 min. The results of our kinetics analysis also accounted for this perturbation by demonstrating no change in V_TCA with the reduction in F₁.

Although experiments with isolated mitochondria were intended to confirm the effectiveness of the protocol to manipulate K_m (26), it is of interest that the pH response in the ratios of -ketoglutarate oxidation to efflux were closely matched between intact hearts and isolated mitochondria metabolizing labeled glutamate. In these instances (see Table 2), oxidation-to-efflux ratios at normal pH were 1.28 in isolated mitochondria and 1.08 in intact hearts, and oxidation-to-efflux ratios at low pH were 3.76 in isolated mitochondria versus 3.57 in intact hearts. In the isolated mitochondria, the ratio of -ketoglutarate oxidation to efflux varied somewhat depending on the source of the -ketoglutarate and on pH. However, the pH effect was largest when -ketoglutarate was generated by aspartate aminotransferase via [2,3-³H]glutamate. This is probably due to compartmentation within the mitochondria. As noted in the results, the production of matrix -ketoglutarate from glutamate in isolated mitochondria was lower than the production of -ketoglutarate from pyruvate or from the -ketoglutarate entering from the medium. The aminotransferase, therefore, may have had access to a particular pool of -KGDH that was not available to the other substrates and was therefore less saturated than pools of -KGDH near the -ketoglutarate translocase or near isocitrate dehydrogenase. Thus the effect of the pH change may have been largest on this least saturated pool of -KGDH, as may be expected for the intact functioning heart, which showed strikingly similar ratios to that observed in mitochondria metabolizing labeled glutamate. In any event, it seems likely that the substrate-related differences in the effect of pH on the oxidation-to-efflux ratio was due to lack of mixing and slow diffusion in the mitochondrial matrix. In each instance, independent of the source of the -ketoglutarate generated in the matrix, a decrease in pH substantially changed the percentage of the -ketoglutarate, which continues on in the citric acid cycle without exciting the mitochondria.

In conclusion, sequential ¹³C NMR spectra were obtained from intact hearts under conditions of normal and altered -KGDH kinetics during perfusion with [2-¹³C]acetate. Dynamic ¹³C NMR analysis revealed the balance between mitochondria TCA cycle rates and metabolite exchange between the mitochondria and cytosol. The data indicate that elevated H⁺ and Ca²⁺ content increased -ketoglutarate oxidation and reduced -ketoglutarate efflux from the mitochondria and isotope exchange with cytosolic glutamate. These results indicate that ¹³C NMR is sensitive to changes in kinetic parameters of the TCA cycle dehydrogenases within the mitochondria of intact hearts as confirmed by our parallel experiments in mitochondria.